Association of two ERCC4 tagSNPs with susceptibility to atrophic gastritis and gastric cancer in Chinese

Gene 519 (2013) 335–342

Contents lists available at SciVerse ScienceDirect

Gene journal homepage: www.elsevier.com/locate/gene

Association of two ERCC4 tagSNPs with susceptibility to atrophic gastritis and gastric cancer in Chinese☆ Yantao Gong a, b, Caiyun He a, Zhipeng Duan a, Liping Sun a, Qian Xu a, Chengzhong Xing a, b,⁎, Yuan Yuan a,⁎⁎ a Tumor Etiology and Screening Department of Cancer Institute and General Surgery, the First Affiliated Hospital of China Medical University, and Key Laboratory of Tumor Etiology and Prevention, Shenyang 110001, Liaoning Province, China b Department of Surgical Oncology, the First Affiliated Hospital of China Medical University, Shenyang 110001, Liaoning Province, China

a r t i c l e

i n f o

Article history: Accepted 25 January 2013 Available online 13 February 2013 Keywords: ERCC4/XPF Susceptibility Atrophic gastritis Gastric cancer Helicobacter pylori

a b s t r a c t Genetic polymorphisms in excision repair cross-complementing group 4 (ERCC4) may contribute to the risk of cancer development. However, there are few reports regarding to susceptibility to gastric cancer (GC) or its precursor, atrophic gastritis (AG). Thereby, we investigated the association between two tag single nucleotide polymorphisms (tagSNPs) rs6498486 and rs254942, which represents the majority of common SNPs of ERCC4 gene, and the risks of GC and AG development in a sex- and age-matched case–control designed study. We found that rs6498486 polymorphism was associated with a reduced AG risk in total population (for AC vs. AA: OR = 0.69, 95%CI = 0.52–0.94, P = 0.016; for AC/CC vs. AA: OR = 0.68, 95%CI = 0.51–0.92, P = 0.010) as well as in the subpopulation of youngers (age b 60 years) (for AC/CC vs. AA: OR = 0.67, 95%CI = 0.45–0.99, P = 0.048). For the rs254942 polymorphism, compared with the common TT genotype, the genotypes of CT and CT/CC were only observed to reduce AG risk in the subgroups of males (for CT vs. TT: OR = 0.64, 95%CI = 0.45–0.90, P = 0.012; for CT/CC vs. TT: OR = 0.66, 95%CI = 0.47–0.92, P = 0.016) and youngers (for CT vs. TT: OR = 0.72, 95%CI = 0.53–0.97, P = 0.035; for CT/CC vs. TT: OR = 0.74, 95%CI = 0.55–0.99, P = 0.045). However, no significant statistical association of the two SNPs with GC susceptibility was observed in the total population. Only rs6498486 AC and AC/CC genotypes were found to be marginally associated with a reduced GC risk in the subgroup of males (for AC vs. AA: OR = 0.69, 95%CI = 0.49–0.99, P = 0.043; for AC/CC vs. AA: OR = 0.71, 95%CI = 0.50–0.99, P = 0.046). Our findings suggested that the ERCC4 rs6498486 and rs254942 may be associated with AG risk. Further validation of our results in larger populations and additional studies evaluating their molecular function are required. © 2013 Elsevier B.V. All rights reserved.

Abbreviations: AG, atrophic gastritis; CHB, Chinese Han Beijing; CI, confidence interval; ELISA, enzyme-linked immunosorbent assay; H. pylori, Helicobacter pylori; HWE, Hardy–Weinberg Equilibrium; IgG, immunoglobin G; ERCC4/XPF, excision repair cross-complementing group 4; GC, gastric cancer; LD, linkage disequilibrium; NER, nucleotide excision repair; OR, odds ratio; PCR–RFLP, polymerase chain reaction restriction fragment length polymorphism; ROS, reactive oxidative species; SNP, single nucleotide polymorphism; UV, ultraviolet light; XPA, xeroderma pigmentosum complementation group A. ☆ Author contributions: Yantao Gong and Caiyun He performed the genotyping, statistical analysis, data interpretation and wrote the paper, and they contributed equally to this work and thus are regarded as the co-first author. Zhipeng Duan, Liping Sun, and Qian Xu were responsible for the collection of the serum/biopsy samples for the study and serological testing. Chengzhong Xing was responsible for collection of clinical information and revised the manuscript. Yuan Yuan conceived and designed this study and revised the manuscript. ⁎ Correspondence to: C. Xing, Tumor Etiology and Screening Department of Cancer Institute and General Surgery, the First Affiliated Hospital of China Medical University, 155# North Nanjing Street, Heping District, Shenyang City, 110001, Liaoning Province, China. Tel.: +86 24 83283555; fax: +86 24 83282292. ⁎⁎ Correspondence to: Y. Yuan, Tumor Etiology and Screening Department of Cancer Institute and General Surgery, the First Affiliated Hospital of China Medical University, 155# North Nanjing Street, Heping District, Shenyang City, 110001, Liaoning Province, China. Tel.: +86 24 83282153; fax: +86 24 83282292. E-mail addresses: [email protected] (C. Xing), [email protected] (Y. Yuan). 0378-1119/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2013.01.059

1. Introduction Despite a general decline in incidence rates of gastric cancer (GC), it remains the fourth most common cancer and one of the leading causes of cancer death worldwide (Siegel et al., 2011). Generally, the process of gastric carcinogenesis incorporates at least three key steps including development of superficial gastritis, precancerous conditions (i.e. atrophy, intestinal metaplasia and dysplasia) and carcinoma (Fox and Wang, 2007). Helicobacter pylori (H. pylori) is thought to be a major environmental risk factor facilitating gastric carcinogenesis. However, only a small proportion of infected people develop severe diseases (Uemura et al., 2001), strongly suggesting a determining role of host factors and their interaction with environmental exposures in this progression process (Hamajima et al., 2006). Human DNA repair systems play a critical role in maintaining genome stability and preventing carcinogenesis (Hoeijmakers, 2001). Impaired DNA repair capacity can give rise to the cancer development (Ramos et al., 2004; Wang et al., 2010a, 2010b). As one of the most versatile defense mechanisms, the nucleotide excision repair (NER) that

336

Y. Gong et al. / Gene 519 (2013) 335–342

consists of seven key genes (xeroderma pigmentosum complementation group A (XPA)–XPG) has the ability to process various lesions caused by ultraviolet light (UV), chemical carcinogens and reactive oxidative species (ROS) (de Laat et al., 1999). Emerging epidemiologic evidences demonstrate that sequence variations in key DNA repair genes of NER may modulate DNA repair capacity and alter individual susceptibility to various cancers (Berndt et al., 2006; He et al., 2012a, 2012b; Pan et al., 2009). Here we focused on a key component of NER, excision repair cross-complementing group 4 (ERCC4, also known as XPF), that functions in the irreversible dual-incision step during DNA repair by forming an obligate heterodimer complex with ERCC1 and then operating 5′ incision to the DNA lesion (Fagbemi et al., 2011). The dual incision accomplished by two sequential incisions at 3′ and 5′ sides of the damaged oligonucleotides is the first rate-limiting step in NER, and thereby is the key to successful removal of DNA lesions. The formation of heterodimer by both endonucleases is indispensable for the nuclease activity of the complex in the NER. More importantly, the actual catalytic domain locates in ERCC4, which determines the NER activity (Enzlin and Scharer, 2002). Moreover, low expression of ERCC4 has been reported to be associated with an increased risk of head and neck cancer (Wei et al., 2005). The ERCC4/ERCC1 complex also plays an important role in the repair of interstrand cross-links in a NER-independent pattern and in the maintenance of telomere integrity (Niedernhofer et al., 2004; Wu et al., 2008). Up to date, hundreds of single nucleotide polymorphisms (SNPs) in ERCC4 gene have been identified, some of which have been reported to be associated with cancer risks, including lung (Shao et al., 2008), pancreatic (McWilliams et al., 2008), and bladder cancers (L.E. Wang et al., 2010; M. Wang et al., 2010). However, information of ERCC4 genetic polymorphisms with susceptibility to gastric carcinogenesis is limited. Until recently, two reports investigated the genetic effect of rs6498486 polymorphism of ERCC4 on the risk of gastric cancer but found no positive association (He et al., 2012a, 2012b; Zhou et al., 2012). In the present study, aiming at elucidating whether ERCC4 genetic variations alter GC or AG risk, we selected two ERCC4 tagSNPs (rs6498486 and rs254942) that represent the majority of common SNP (minor allele frequency (MAF) ≥ 0.05) of this gene by using a tagging SNP approach. We investigated effects of the two tagSNPs and their interaction with H. pylori infection on susceptibilities to GC and one of its key precursors—atrophic gastritis (AG) in a sex- and age-matched case–control study in Chinese. 2. Materials and methods 2.1. Study design and study population All the subjects in the present study were retrospectively enrolled from a multicentre research in Liaoning Province, China from 1997 to 2011. The subjects included in this study were endoscopically and histologically examined and classified into three groups: GC, AG and NOR (healthy control), among which the eligible NOR were confirmed to have relative normal mucosa or only mild superficial gastritis. Data of sex, age and history of any illness were also retrospectively extracted from registered documents. Participants with other malignancies were excluded from this study. Genotypes of the two ERCC4 tagSNPs investigated in our study were separately detected with different methods. And thus the corresponding study populations for each polymorphism were not completely identical. The flow chart describing the study design and study population for each SNP study was presented in Fig. 1. Firstly, a total of 1333 (NOR: 417, AG: 404 and GC: 495) and 1854 subjects (NOR: 724, AG: 652 and GC: 478) were included for rs6498486 and rs254942 studies respectively. Furthermore, NOR were frequency-matched to cases of AG and GC respectively by gender (1:1) and age (±5 years) for individual association analysis of ERCC4 genotypic effect on AG and GC

Fig. 1. Flow chart describing the study design and study population for each SNP study.

risks. After matching, the numbers for cases and controls for the rs6498486 polymorphism were 385:385 (AG:NOR) and 400:400 (GC: NOR), while 560:560 (AG:NOR) and 378:378 (GC:NOR) for the rs254942 polymorphism. For further exploration of combined effects of these two SNPs, 798 participants with both genotypes were included in haplotype analyses, in which the numbers for cases and controls were 208:207 (AG:NOR) and 284:181 (GC:NOR). 2.2. Ethical consideration Ethical approval for this study was obtained from the Human Ethics Review Committee of China Medical University (Shenyang, China). Written informed consents were obtained from the participants. 2.3. TagSNP selection Genotype data of HapMap Chinese Han Beijing (CHB) population (Release 27, Phase I + II+ III, http://www.HapMap.org) were extracted within extended gene regions of ERCC4, which encompasses 5 kb of upstream and downstream flanking sequence. There are 33 common SNPs of ERCC4 in CHB. By using Haploview 4.2 (Barrett, 2009), TagSNPs were selected based on pairwise linkage disequilibrium (LD) information to maximally represent (r2 > 0.8) the common SNPs (minor allele frequency (MAF)>0.05). As we can observe in Fig. 2, the 33 SNPs form a closely linked block, within which only two tagSNPs were required to captured the other common SNPs (mean r2 = 0.966). Since a previous study in bladder cancer reported that rs6498486 A>C polymorphism in ERCC4 promoter region have significant impact on its gene mRNA level and binding activity of certain transcription factor (L.E. Wang et al., 2010; M. Wang et al., 2010), we selected rs6498486 as a tagSNP. Additionally, rs254942 T>C linked closely with rs31870 (r2 = 0.938) was predicted to be a splicing site by the FASTSNP software (http:// fastsnp.ibms.sinica.edu.tw/pages/input_CandidateGeneSearch.jsp), suggesting a potentially effect on this gene's splicing regulation; therefore, rs254942 was selected as another tagSNP in our study. 2.4. Genotyping of rs6498486 Genomic DNA was isolated and purified from the whole blood sample by routine phenol–chloroform method. The SNP rs6498486 was genotyped by polymerase chain reaction and restriction fragment

Y. Gong et al. / Gene 519 (2013) 335–342

337

Fig. 2. Plot of information of linkage disequilibrium (r2) among common ERCC4 polymorphisms based on CHB HapMap population.

length polymorphism (PCR–RFLP) method. The forward (F) and reverse (R) primer sequences for PCR were 5′-TTTTTGGCAGCTT GAGGCTA-3′ (F) and 5′-CTCAGAAAGCCGAAGAGAGC-3′ (R) respectively. PCR was carried out in a volume of 25 μl reaction mixture containing 10× buffer 2.5 μl, 4× dNTP 2.0 μl, each primer 1.0 μl, Taq DNA polymerase 0.5 μl (2.5 U), DNA template 1.0 μl (50–100 ng), and appropriate purified H2O. The amplification was performed at 94 °C (5 min) for initial denaturation, followed by 35 cycles of denaturation at 94 °C (45 s), annealing at 62 °C (45 s) and extension at 72 °C (45 s), and ended with a final elongation at 72 °C (10 min). As a negative control, PCR mix without DNA sample was used to ensure contamination free PCR product. The PCR products was digested by 5 units of MnlI (Fermentas, GmbH, Germany) overnight at 37 °C, and then electrophoresed on 3.5% agarose gel stained with Genefinder. The restricted fragments presenting AA, AC and CC genotypes had band sizes of 204 bp, 204/160/44 bp and 160/44 bp, respectively (Fig. 3a). These genotypes were validated by direct DNA sequencing (Fig. 3b). 2.5. Genotyping of rs254942 Genomic DNA was also isolated from peripheral blood lymphocytes by routine phenol–chloroform method and was diluted to working concentrations of 50 ng/μl for genotyping. All samples were randomized on 384-well plates and blinded for disease status. Assay design and rs254942 genotyping were performed by CapitalBio (Beijing, China) using Sequenom MassARRAY platform (Sequenom, San Diego, California, USA) according to the manufacturer's instructions. 50 samples were repeated genotyped, and the results were 100% concordant.

2.6. Test for H. pylori serology The detailed method of examination of Hp serology has been described in our previous study (Gong et al., 2010). In brief, about 5 ml of fasting venous blood was collected from the participants and the serum sample was obtained after centrifugation at 3500 ×g for 10 min. Serum concentrations of H. pylori-immunoglobin (Ig) G were carried out by enzyme-linked immunosorbent assay (ELISA, H. pylori-IgG ELISA kit, BIOHIT Plc, Helsinki, Finland). A reading higher than 34 enzyme immune-units (EIU) was regarded as H. pylori seropositive. 2.7. Statistical analysis Hardy–Weinberg equilibrium for each SNP was assessed among controls. The independent-samples t test was used to determine difference of age between cases and controls. The Pearson's χ 2 test was used to compare the distributions of gender, H. pylori infection status and genotypes of rs6498486 and rs254942 between cases and controls. Unconditional logistic regression model was used to evaluate the association between genotypes and risk of GC and AG, and odds ratio (OR) and 95% confidence interval (95%CI) adjusted by age and sex were calculated. Interactive effects of genotype and H. pylori infection on risk of GC and AG were evaluated using the likelihood ratio test of full regression model. All the statistical analyses mentioned above were carried out using SPSS 13.0. LD information and haplotype analyses were analyzed by the SHEsis software (Shi and He, 2005). A two-sided P b 0.05 was considered statistically

338

Y. Gong et al. / Gene 519 (2013) 335–342

Fig. 3. Three genotypes for the ERCC4 rs6498486 polymorphism. Panel a: PCR-based restriction analysis of rs6498486 shown on 3.5% agarose electrophoresis. M: 100 bp DNA size marker; lanes 3–6: AA genotype; lanes 2–7: AC genotype, lane 1: CC genotype. Panel b: DNA sequencing analysis of different PRC products containing rs6498486 polymorphism locus.

more likely linked to H. pylori infection than NOR (both P b 0.001), indicating a positive association between H. pylori infection and development of AG and GC. The statistical powers for rs6498486 and rs254942 polymorphisms ranged from 0.78 to 0.99 by setting a significant level at 0.05 to obtain an OR from 1.5 to 2.0 under a dominant genetic model.

significant. We calculated statistical power using the PGA software (http://dceg.cancer.gov/tools/analysis/pga). 3. Results 3.1. Baseline characteristics and ERCC4 genotype frequencies of the study population

3.2. ERCC4 polymorphisms and risk of AG and GC Geographic characteristics of the matched study populations were summarized in Table 1. There were no statistical differences in the distributions of age and gender between AG and controls and GC and controls (both P > 0.05). Both AG and GC cases seemed to be

Genotype frequencies of the two tagSNPs investigated in our study were in agreement with Hardy–Weinberg equilibrium in controls (both P > 0.05). Individual effects of ERCC4 polymorphisms on the

Table 1 Baseline characteristics of study population for each ERCC4 SNP. rs6498486

Gender Male Female Age b60 ≥60 H. pylori b Positive Negative No info. a b

rs254942

NOR% (N = 385)

AG% (N = 385)

255(66.2) 130(33.8)

255(66.2) 130(33.8)

204(53.0) 181(47.0)

211(54.8) 174(45.2)

75(19.5) 274(71.2) 36(9.3)

172(44.7) 159(41.3) 54(14.0)

Pa

NOR% (N = 400)

GC% (N = 400)

270(67.5) 130(32.5)

270(67.5) 130(32.5)

206(51.5) 194(48.5)

221(55.3) 179(44.7)

79(19.8) 285(71.2) 36(9.0)

93(23.3) 91(22.7) 216(54.0)

Pa

1

NOR% (N = 560)

AG% (N = 560)

304(54.3) 256(45.7)

304(54.3) 256(45.7)

388(69.3) 172(30.7)

394(70.4) 166(29.6)

87(15.5) 468(83.6) 5(0.9)

302(53.9) 253(45.2) 5(0.9)

1

0.613

GC% (N = 378)

258(68.3) 120(31.7)

258(68.3) 120(31.7)

246(65.1) 132(34.9)

250(66.1) 128(33.9)

57(15.1) 317(83.9) 4(1.0)

93(24.6) 113(29.9) 172(45.5)

Pa 1

0.696

b0.001

P value was estimated by Pearson χ2 test. P was estimated for comparison between H. pylori positive and negative subgroups.

NOR% (N = 378)

1

0.288

b0.001

Pa

0.759

b0.001

b0.001

Y. Gong et al. / Gene 519 (2013) 335–342

339

Table 2 Association of ERCC4 polymorphisms with AG risk. rs6498486

Overall

Gender Male

Female

Age (y) b60

≥60

a

rs254942

Genotype

NOR % (N = 385)

AG% (N = 385)

OR(95%CI)a

P

Genotype

NOR % (N = 560)

AG% (N = 560)

OR(95%CI)a

P

AA AC CC AC + CC

208(54.0) 159(41.3) 18(4.7) 177(46.0)

243(63.1) 129(33.5) 13(3.4) 142(36.9)

Reference 0.69(0.52–0.94) 0.62(0.30–1.30) 0.68(0.51–0.92)

0.016 0.230 0.010

TT TC CC TC + CC

331(59.1) 208(37.1) 21(3.8) 229(40.9)

358(63.9) 177(31.6) 25(4.5) 142(36.1)

Reference 0.79(0.61–1.01) 1.14(0.62–2.07) 0.82(0.64–1.04)

0.060 0.680 0.097

AA AC CC AC + CC AA AC CC AC + CC

134(52.5) 108(42.4) 13(5.1) 121(47.5) 74(56.9) 51(39.2) 5(3.9) 56(43.1)

155(60.8) 90(35.3) 10(3.9) 100(39.2) 88(67.7) 39(30.0) 3(2.3) 42(32.3)

Reference 0.72(0.50–1.03) 0.66(0.28–1.56) 0.71(0.50–1.01) Reference 0.64(0.38–1.08) 0.50(0.11–2.17) 0.63(0.38–1.04)

TT TC CC TC + CC TT TC CC TC + CC

174(57.2) 119(39.1) 11(3.7) 130(42.8) 157(61.3) 89(34.8) 10(3.9) 99(38.7)

203(66.8) 90(29.6) 11(3.6) 101(33.2) 155(60.5) 87(34.0) 14(5.5) 101(39.5)

Reference 0.64(0.45–0.90) 0.84(0.35–2.00) 0.66(0.47–0.92) Reference 0.98(0.68–1.43) 1.49(0.63–3.49) 1.03(0.72–1.47)

AA AC CC AC + CC AA AC CC AC + CC

113(55.4) 80(39.2) 11(5.4) 91(44.6) 95(52.5) 79(43.6) 7(3.9) 86(47.5)

137(64.9) 67(31.8) 7(3.3) 74(35.1) 106(60.9) 62(35.6) 6(3.5) 68(39.1)

Reference 0.69(0.45–1.04) 0.53(0.19–1.41) 0.67(0.45–0.99) Reference 0.69(0.45–1.07) 0.75(0.24–2.33) 0.70(0.46–1.07)

TT TC CC TC + CC TT TC CC TC + CC

225(58.0) 149(38.4) 14(3.6) 163(42.0) 106(61.6) 59(34.3) 7(4.1) 66(38.4)

256(65.0) 123(31.2) 15(3.8) 138(35.0) 102(61.4) 54(32.5) 10(6.1) 64(38.6)

Reference 0.72(0.53–0.97) 0.94(0.44–1.99) 0.74(0.55–0.99) Reference 0.94(0.59–1.50) 1.41(0.50–3.95) 1.00(0.64–1.55)

0.077 0.350 0.061 0.095 0.356 0.074

0.077 0.207 0.048 0.106 0.625 0.103

0.012 0.710 0.016 0.952 0.359 0.854

0.035 0.874 0.045 0.821 0.505 0.987

OR and 95%CI was calculated by logistic regression adjusted for age and sex. Analyses of results with P b 0.05 were highlighted in bold characters.

0.55–0.99, P = 0.045) in the subgroup of youngers and in males (OR = 0.66, 95%CI= 0.47–0.92, P= 0.016) (Table 2). For GC susceptibility, only rs6498486 AC genotype and (AC + CC) genotypes were observed to marginally associated with a reduce GC risk in the subgroup of males (OR = 0.69, 95%CI = 0.49–0.99, P = 0.043; OR = 0.71, 95%CI = 0.50–0.99, P = 0.046, respectively). No positive association was observed between both SNPs and GC risk in total population (Table 3). Moreover, no association was found for risk of each subtype of GC (Table 4). Combined effects of these two ERCC4 polymorphisms were assessed by haplotype analyses. No statistical difference in the distribution of these haplotypes was found between cases and controls (Table 5).

risk of AG and GC were assessed in total population and in subpopulations according to sex and age, and also in histological subtype for GC cases based on the Lauren's classification. For AG susceptibility, compared with the common AA genotype, rs6498486 AC genotype and (AC + CC) genotypes were found to be associated with reduced risk of AG in total population, with corresponding ORs of 0.69 (95%CI= 0.52–0.94, P = 0.016) and 0.68 (95%CI= 0.51– 0.92, P = 0.010), moreover, (AC + CC) genotypes were also observed to associated with a reduced AG risk in the subgroups of youngers (b 60 yr); and with regard to rs254942 polymorphism, no overall effect of this polymorphism on AG risk was observed in total population; however, compared with TT genotype, (TC + CC) genotypes were found to be associated with reduced AG risk (OR = 0.74, 95%CI =

Table 3 Association of ERCC4 polymorphisms with GC risk. rs6498486

Overall

Gender Male

Female

Age b60

≥60

a

rs254942 a

Genotype

NOR% (N = 400)

GC% (N = 400)

OR(95%CI)

AA AC CC AC + CC

222(55.5) 159(39.8) 19(4.8) 178(44.5)

237(59.3) 143(35.8) 20(5.0) 163(40.7)

Reference 0.84(0.63–1.13) 0.97(0.50–1.87) 0.86(0.65–1.14)

AA AC CC AC + CC AA AC CC AC + CC

142(52.6) 113(41.9) 15(5.5) 128(47.4) 80(61.5) 46(35.4) 4(3.1) 50(38.5)

165(61.1) 91(33.7) 14(5.2) 105(38.9) 72(55.4) 52(40.0) 6(4.6) 58(44.6)

Reference 0.69(0.49–0.99) 0.81(0.38–1.73) 0.71(0.50–0.99) Reference 1.26(0.76–2.09) 1.67(0.45–6.17) 1.29(0.79–2.12)

AA AC CC AC + CC AA AC CC AC + CC

116(56.3) 79(38.3) 11(5.4) 90(43.7) 106(54.6) 80(41.2) 8(4.2) 88(45.4)

134(60.6) 75(33.9) 12(5.5) 87(39.4) 103(57.5) 68(38.0) 8(4.5) 76(42.5)

Reference 0.82(0.55–1.23) 0.94(0.40–2.21) 0.83(0.57–1.23) Reference 0.87(0.57–1.33) 0.96(0.36–2.74) 0.89(0.58–1.33)

P

Genotype

NOR% (N = 378)

GC% (N = 378)

OR(95%CI)a

P

0.247 0.930 0.284

TT TC CC TC+ CC

221(58.5) 142(37.6) 15(4.0) 157(41.5)

226(59.8) 129(34.1) 23 (6.1) 152(40.2)

Reference 0.89(0.66–1.21) 1.49(0.76–2.95) 0.95(0.71–1.27)

0.465 0.247 0.718

TT TC CC TC+ CC TT TC CC TC+ CC

147(57.0) 101(39.1) 10(3.9) 111(43.0) 74(61.7) 41(34.2) 5(4.1) 46(38.3)

154(59.7) 87(33.7) 17(6.6) 104(40.3) 72(60.0) 42(35.0) 6(5.0) 46(40.0)

Reference 0.82(0.56–1.18) 1.60(0.71–3.63) 0.89(0.62–1.27) Reference 1.06(0.62–1.82) 1.23(0.36–4.24) 1.07(0.63–1.80)

TT TC CC TC+ CC TT TC CC TC+ CC

138(56.1) 98(39.8) 10(4.1) 108(43.9) 83(62.9) 44(33.3) 5(3.8) 49(37.1)

143(57.2) 87(34.8) 20(8.0) 107(42.8) 83(64.8) 42(32.8) 3(2.4) 45(35.2)

Reference 0.85(0.59–1.24) 1.93(0.87–4.28) 0.95(0.67–1.37) Reference 0.95(0.56–1.61) 0.51(0.11–2.30) 0.91(0.55–1.51)

0.043 0.581 0.046 0.378 0.441 0.311

0.336 0.887 0.362 0.520 0.977 0.558

OR and 95%CI was calculated by logistic regression adjusted for age and sex. Analyses of results with P b 0.05 were highlighted in bold characters.

0.290 0.255 0.536 0.820 0.733 0.787

0.423 0.104 0.819 0.866 0.384 0.730

340

Y. Gong et al. / Gene 519 (2013) 335–342

Table 4 Association of ERCC4 polymorphisms with different histological-type GC risk.

rs6498486 AA AC CC AC + CC rs254942 TT CT CC CT + CC a

NOR%

Diffuse type%

N= 400 222(55.5) 159(39.8) 19(4.7) 178(45.5) N= 378 221(58.5) 142(37.6) 15(3.9) 157(41.5)

N = 159 94(59.1) 57(35.8) 8(5.1) 65(40.9) N = 118 69(58.5) 41(34.7) 8(6.8) 49(41.5)

OR(95%CI)a

P

Intestinal type%

Reference 0.85(0.58–1.25) 0.98(0.41–2.32) 0.87(0.60–1.26)

0.403 0.957 0.447

Reference 0.93(0.60–1.45) 1.70(0.69–4.20) 1.00(0.66–1.53)

0.754 0.247 0.986

N = 127 73(57.5) 47(37.0) 7(5.5) 54(42.5) N = 83 52(62.7) 27(32.5) 4(4.8) 31(37.3)

OR(95%CI)a

P

Reference 0.87(0.57–1.33) 1.00(0.40–2.50) 0.88(0.58–1.32)

0.508 0.995 0.534

Reference 0.83(0.50–1.39) 1.18(0.37–3.77) 0.86(0.53–1.41)

0.480 0.784 0.549

OR and 95%CI were calculated by logistic regression adjusted for age and sex.

3.3. Interaction effects of H. pylori and ERCC4 polymorphisms on the risk of AG and GC Subjects with available information of status of H. pylori infection were included in the interaction analyses and no statistical differences of age and sex were found between cases and controls (data not shown). Results of interaction analyses between H. pylori infection, genotypes and risk of AG and GC were presented in Table 6. We combined the heterozygous and variant homozygous genotypes for the interaction analyses. Subjects without H. pylori infection who carry protective genotypes, either rs6498486 (AC + CC) or rs254942 (TC + CC) genotypes, were used as reference, risk of AG and GC development substantially elevated in individuals carrying hazard genotypes (rs6498486 AA or rs254942 TT) and infected by H. pylori. However, we did not identify significant interactive effects of ERCC4 genotype and H. pylori infection on the risk of AG and GC in a multiplicative interactive model, both P for interaction were larger than 0.05 (Table 6). 4. Discussion Despite a crucial role of endonuclease ERCC4 in NER capacity, information of association of its genetic polymorphisms with susceptibility of AG and GC is limited. In the present study, we investigated effects of two ERCC4 tagSNPs that represent the majority of common SNPs of ERCC4 on individual susceptibility to AG and GC. Our data showed that rs6498486 was associated with a reduced AG risk in the total population as well as in the subpopulation of youngers. In addition, rs6498486 decreased GC risk in the subpopulation of males. We also observed an association between rs254942 TC/CC genotype and a decreased AG risk in the subgroups of males and youngers. The ERCC4 gene contains 6 exons that encompass 28.2 kb on 16p13.12–p13.13. The rs6498486 (− 357A > C) polymorphism, located in the ERCC4 promoter region, was selected as a tagSNP representing 30 other common polymorphisms (i.e. rs3136038, rs1799797, rs744154 and so on) in this gene and investigated in our study. Prior to our present study, the variant allele combination (TCA) of − 673C > T (rs3136038), − 357A > C (rs6498486) and − 30T > A (rs1799797), which are in complete linkage disequilibrium, has been reported to be linked with higher transcription

activity when compared to CAT (Yu, 2008). The variant TT genotype of rs3136038 polymorphism was also reported to reduce the expression of ERCC4 mRNA levels and risk of lung cancer (Yu, 2008). Collectively, the rs6498486 variant T allele seems to be associated with suboptimal regulation of ERCC4 transcription. In consistent with the results of Zhou's and He's studies (He et al., 2012a, 2012b; Zhou et al., 2012), no significant association of rs6498486 with the overall risk of GC was found. Differently, our research also investigated whether the rs6498486 was related to the risk of a precursor of GC, i.e. AG, and our findings demonstrated a reverse association of this polymorphism with AG risk. Furthermore, in a two-stage association study, Milne et al. found that breast cancer risk remarkably decreased in individuals carrying rs744154 variant homogenous genotype (CC) (Milne et al., 2006), which is also in complete linkage disequilibrium with rs6498486 variant homogenous genotype (CC). The above described reports showed a protective effect of rs6498486 variant C allele on cancer development, and partly supported its protective role in cancer and precancerous disease development observed in this study. Contrarily, Wang and his colleagues reported that a positive association of this polymorphism with increased risk of the development and the recurrence of bladder cancer (L.E. Wang et al., 2010; M. Wang et al., 2010). These discrepancies may be partly attributed to differences in the pathogenesis of different malignances. We also found that AG risk significantly decreased in individuals carrying rs254942 variant TC and TC + CC genotypes in males and youngers. This polymorphic site located in intron 5 of ERCC4 gene near its 3′ end. According to the HapMap Chinese data, rs254942 has strong LD with another intronic rs31870 polymorphism in ERCC4 gene (r 2 = 0.938). Both rs254942 and rs31870 do not cause any amino acid change themselves, however, rs254942 locus is predicted to be an alternative splicing site or transcription factor binding site by FastSNP search (Yuan et al., 2006). Moreover, the rs254942 allele T→C change, predicted by another public tool for transcription factor search (TESS), possibly contributes to several TFBS change. No predicted functional effect was found for rs31870 polymorphic site. Collectively, rs254942 is likely to be a putatively functional polymorphic site and selected to be another tagSNP capturing rs31870 in our study. Further study on its biological function is required. Several other polymorphisms of ERCC4 gene, including rs3136038 and rs1800067, have been reported to alter the risk of lung and

Table 5 Association of haplotype of the two ERCC4 SNPs (rs6498486–rs254942) with AG and GC risks. Haplotypes

NOR% (N = 207)

AG% (N = 208)

χ2

P

OR(95%CI)

NOR% (N = 181)

GC% (N = 284)

χ2

P

OR(95%CI)

A–T A–C C–T C–Ca

225.21(0.544) 83.79(0.202) 102.79(0.248) 2.21(0.005)

237.01(0.577) 89.99(0.216) 88.99(0.214) 0.01(0.000)

0.437 0.206 1.48 /

0.508 0.649 0.223 /

1.10(0.83–1.44) 1.08(0.77–1.51) 0.82(0.59–1.13) /

196.26(0.542) 74.74(0.206) 88.74(0.245) 2.26(0.006)

305.45(0.538) 131.55(0.232) 128.55(0.226) 2.45(0.004)

0.026 0.783 0.460 /

0.871 0.376 0.497 /

0.98(0.75–1.28) 1.16(0.84–1.59) 0.90(0.66–1.23) /

a

All those frequencies b 0.05 were ignored in analyses.

Y. Gong et al. / Gene 519 (2013) 335–342

341

Table 6 Association of interaction of ERCC4 polymorphisms and H. pylori infection with AG and GC risk.

For rs6498486 H. pylori negative AC + CC AA H. pylori positive AC + CC AA Interactiona For rs254942 H. pylori negative TC + CC TT H. pylori positive TC + CC TT Interactiona a

NOR%

AG%

N = 349

N= 331

113(32.4) 132(37.8)

53(16.0) 85(25.7)

OR(95%CI)

P

Reference 1.37(0.90–2.10)

0.143

45(12.9) 65(19.6) 59(16.9) 128(38.7) OR = 1.093, P = 0.789 N = 555 N= 555

3.07(1.86–5.07) 4.61(2.94–7.27)

b0.001 b0.001

187(33.7) 250(45.0)

Reference 1.31(0.93–1.84)

0.119

7.16(4.61–11.14) 7.251(5.00–10.52)

b0.001 b0.001

76(13.7) 133(24.0)

42(7.6) 122(22.0) 76(13.7) 224(40.3) OR = 0.772, P = 0.362

NOR%

GC%

N = 364

N= 184

OR(95%CI)

P

113(31.0) 142(39.0)

28(15.2) 57(31.0)

Reference 1.60(0.95–2.68)

0.075

46(12.6) 39(21.2) 63(17.4) 60(32.6) OR = 0.710, P = 0.377 N = 374 N= 206

3.34(1.84–6.07) 3.80(2.20–6.55)

b0.001 b0.001

127(34.0) 167(44.7)

36(17.5) 68(33.0)

Reference 1.42(0.79–1.91)

0.139

30(8.0) 44(21.4) 50(13.3) 58(28.1) OR = 0.551, P = 0.124

5.15(2.84–9.32) 4.04(2.38–6.86)

b0.001 b0.001

Interaction effects were estimated by likelihood ratio rest under a full model.

pancreatic cancers. However, in the present study, genetic impact of ERCC4 variations did not exhibit significant influence on the risk of GC in total population, except a protective effect of rs6498486 AC/CC genotypes in a subpopulation of males. Considering a significantly positive association of both ERCC4 tagSNPs with AG investigated in our study and a closely etiologic link connecting AG with intestinaltype GC, we further explored the ERCC4 genetic influence on risk of histologic subtype of GC according to Lauren's classification (Lauren, 1965), but no positive association was found. These results indicated that variant genotypes of ERCC4 SNPs investigated in our study have protective effect on AG development, but do not have obvious impact on gastric mucosa transition from AG to GC. Haplotype analyses demonstrated that no joint effect of rs6498486 and rs254942 on the development of AG and GC. It has been well known that gastric carcinogenesis is closely associated with H. pylori infection. Chronic H. pylori infection induces persistent inflammation response and a great deal of ROS, which can cause a great of different types of DNA lesion (De Bont and van Larebeke, 2004; Obst et al., 2000). Under normal condition, there is a balance between DNA damage and repair. However, reduced DNA repair capacity caused by genetic polymorphisms and increased DNA damage generated by H. pylori infection may disequilibrate this status and thereby lead to an accumulation of genome damage. In this context, polymorphisms of DNA repair genes may interact with H. pylori infection in the development of AG and GC. In the present study, however no obvious interactive effects between the two ERCC4 SNPs investigated in our study and H. pylori infection were found on the risk of AG and GC. Of course, there were several limitations in our study. Our study population is relatively small, especially for the stratified analyses in subpopulation; thereby, impact of ERCC4 polymorphisms on the risk of different histological types of GC and their interaction effects with H. pylori infection might be probably underestimated. We only considered the statistical results of our subgroup analyses as exploratory and hence did not adjust for multiple testing, as recommended previously (Bender and Lange, 2001). In addition, the significance of these two variants in AG and GC patients needs to be elucidated by functional assays, despite that the rs6498486 polymorphism has been reported to be linked with transcription regulation and the expression of ERCC4 mRNA level. In conclusion, our findings suggested that the ERCC4 tagSNPs, rs6498486 and rs254942, may play protective roles in gastric carcinogenesis, especially in the development of AG. Further validation of our results in larger populations and additional studies evaluating their function impact are required.

Acknowledgments This study is supported by grants from the National Basic Research Program of China (973 Program ref no. 2010CB529304), the Science Technology Project in Liaoning Province (ref no. 2011225002) and the grants of the Science Project of Liaoning Province (ref [2008]621). Disclosure: The authors have no conflict of interest. References Barrett, J.C., 2009. Haploview: visualization and analysis of SNP genotype data. Cold Spring Harb. Protoc. 4 (10) (pdb ip71). Bender, R., Lange, S., 2001. Adjusting for multiple testing—when and how? J. Clin. Epidemiol. 54, 343–349. Berndt, S.I., Platz, E.A., Fallin, M.D., Thuita, L.W., Hoffman, S.C., Helzlsouer, K.J., 2006. Genetic variation in the nucleotide excision repair pathway and colorectal cancer risk. Cancer Epidemiol. Biomarkers Prev. 15, 2263–2269. De Bont, R., van Larebeke, N., 2004. Endogenous DNA damage in humans: a review of quantitative data. Mutagenesis 19, 169–185. de Laat, W.L., Jaspers, N.G., Hoeijmakers, J.H., 1999. Molecular mechanism of nucleotide excision repair. Genes Dev. 13, 768–785. Enzlin, J.H., Scharer, O.D., 2002. The active site of the DNA repair endonuclease XPF– ERCC1 forms a highly conserved nuclease motif. EMBO J. 21, 2045–2053. Fagbemi, A.F., Orelli, B., Scharer, O.D., 2011. Regulation of endonuclease activity in human nucleotide excision repair. DNA Repair (Amst) 10, 722–729. Fox, J.G., Wang, T.C., 2007. Inflammation, atrophy, and gastric cancer. J. Clin. Invest. 117, 60–69. Gong, Y.H., Sun, L.P., Jin, S.G., Yuan, Y., 2010. Comparative study of serology and histology based detection of Helicobacter pylori infections: a large population-based study of 7,241 subjects from China. Eur. J. Clin. Microbiol. Infect. Dis. 29, 907–911. Hamajima, N., Naito, M., Kondo, T., Goto, Y., 2006. Genetic factors involved in the development of Helicobacter pylori-related gastric cancer. Cancer Sci. 97, 1129–1138. He, J., et al., 2012a. Polymorphisms in the XPG gene and risk of gastric cancer in Chinese populations. Hum. Genet. 131, 1235–1244. He, J., et al., 2012b. Polymorphisms in ERCC1 and XPF genes and risk of gastric cancer in an eastern Chinese population. PLoS One 7, e49308. Hoeijmakers, J.H., 2001. Genome maintenance mechanisms for preventing cancer. Nature 411, 366–374. Lauren, P., 1965. The two histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. An attempt at a histo-clinical classification. Acta Pathol. Microbiol. Scand. 64, 31–49. McWilliams, R.R., et al., 2008. Polymorphisms in DNA repair genes, smoking, and pancreatic adenocarcinoma risk. Cancer Res. 68, 4928–4935. Milne, R.L., et al., 2006. ERCC4 associated with breast cancer risk: a two-stage case– control study using high-throughput genotyping. Cancer Res. 66, 9420–9427. Niedernhofer, L.J., et al., 2004. The structure-specific endonuclease Ercc1–Xpf is required to resolve DNA interstrand cross-link-induced double-strand breaks. Mol. Cell. Biol. 24, 5776–5787. Obst, B., Wagner, S., Sewing, K.F., Beil, W., 2000. Helicobacter pylori causes DNA damage in gastric epithelial cells. Carcinogenesis 21, 1111–1115. Pan, J., et al., 2009. Genetic susceptibility to esophageal cancer: the role of the nucleotide excision repair pathway. Carcinogenesis 30, 785–792. Ramos, J.M., Ruiz, A., Colen, R., Lopez, I.D., Grossman, L., Matta, J.L., 2004. DNA repair and breast carcinoma susceptibility in women. Cancer 100, 1352–1357.

342

Y. Gong et al. / Gene 519 (2013) 335–342

Shao, M., et al., 2008. Polymorphisms in excision repair cross-complementing group 4 (ERCC4) and susceptibility to primary lung cancer in a Chinese Han population. Lung Cancer 60, 332–339. Shi, Y.Y., He, L., 2005. SHEsis, a powerful software platform for analyses of linkage disequilibrium, haplotype construction, and genetic association at polymorphism loci. Cell Res. 15, 97–98. Siegel, R., Ward, E., Brawley, O., Jemal, A., 2011. Cancer statistics, 2011: the impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J. Clin. 61, 212–236. Uemura, N., et al., 2001. Helicobacter pylori infection and the development of gastric cancer. N. Engl. J. Med. 345, 784–789. Wang, L.E., et al., 2010a. Reduced DNA repair capacity for removing tobacco carcinogen-induced DNA adducts contributes to risk of head and neck cancer but not tumor characteristics. Clin. Cancer Res. 16, 764–774. Wang, M., et al., 2010b. A novel XPF −357A>C polymorphism predicts risk and recurrence of bladder cancer. Oncogene 29, 1920–1928.

Wei, Q., Wang, L.E., Sturgis, E.M., Mao, L., 2005. Expression of nucleotide excision repair proteins in lymphocytes as a marker of susceptibility to squamous cell carcinomas of the head and neck. Cancer Epidemiol. Biomarkers Prev. 14, 1961–1966. Wu, Y., Mitchell, T.R., Zhu, X.D., 2008. Human XPF controls TRF2 and telomere length maintenance through distinctive mechanisms. Mech. Ageing Dev. 129, 602–610. Yuan, H.Y., et al., 2006. FASTSNP: an always up-to-date and extendable service for SNP function analysis and prioritization. Nucleic Acids Res. 34, W635–W641. Yu, D.K., 2008. FunctionaI ERCCl and XPF variants are associated with susceptibility to lung cancer but resistance to platinum-based chemotherapy. Peking Union Medical College, [D]Beijing. Zhou, R.M., Niu, C.X., Wang, N., Chen, Z.F., Lei, S.E., Li, Y., 2012. ERCC1 gene +262A/C polymorphism associated with the risk of gastric cardiac adenocarcinoma in nonsmokers. Arch. Med. Res. 43, 67–74.