A 5′-flanking region polymorphism in toll-like receptor 4 is associated with gastric cancer in a Chinese population

Objective Inflammation induced by H.pylori colonization in the stomach is related to the development of gastric cancer and the genetic variations of the genes involved in the immune responses modify the host response to the infection. The aim of this study was to evaluate whether polymorphisms in the toll-like receptor 4 (TLR4) gene, a key regulator of both innate and adaptive immunity, were related to the susceptibility to gastric cancer in a Chinese population. Methods Two variations in the 5′-flanking region of TLR4 (rs1927914 T > C and rs10759932 T > C) were genotyped by using the PCR-restriction fragment length polymorphism (RFLP) assay in a case-control study of 1,053 incident gastric cancer cases and 1,100 cancer-free controls in a Chinese population. Results Individuals carrying the C allele of rs10759932 had a significantly reduced risk of gastric cancer (adjusted OR = 0.81; 95%CI = 0.67-0.96), compared with the wild-type homozygote (TT), and the protective effect was not significantly different among subgroups stratified by age, sex, smoking, drinking and H.pylori infection status (P for heterogeneity > 0.05). No significant association was observed between rs1927914 and gastric cancer risk in this study population. Conclusion The T to C allele substitution of rs10759932 may play a protective role in gastric carcinogenesis in a Chinese population. Large studies with different ethnic populations are warranted to confirm these findings.


INTRODUCTION
Helicobacter pylori (H. pylori), a gram-negative microaerophilic bacterium, mainly colonizes the human stomach and induces various gastric lesions, ranging from gastritis to cancer. Generally, most individuals with H. pylori infection develop chronic gastritis, but only few of them (<1%) ultimately result in gastric cancer. Therefore, it is well recognized that host factors, especially genetic heterogeneity, may be important in gastric carcinogenesis interacting with H. pylori infection. Innate immunity plays a crucial role in host protection against pathogens, relying on pattern recognition receptors, such as the toll-like receptors (TLRs), to alter the immune system response to existing or invading bacteria. The TLRs are cellsurface signaling molecules that recognize pathogenassociated molecular patterns expressed on infectious agents. Once activated, the TLRs lead not only to the induction of an inflammatory response, but also to the development of antigen-specific adaptive immunity [1] . H. pylori lipopolysaccharide (LPS) is considered to be one of the virulence factors involved in the gastritis, inducing secretion of proinflammatory cytokines, inducible NO and oxygen radicals by activating monocytes and gastric epithelial cells. TLR4, a member of the TLR family and the main receptor of LPS [2,3] , is involved in recognition of H. pylori and plays an important role in gastroduodenal diseases and gastric carcinogenesis [3,4] .
Two non-synonymous polymorphisms of TLR4, Asp299Gly (rs4986790) and Thr399Ile (rs4986791), have been shown to be extensively involved in H. pylori infection-related disease, including gastric cancer. For example, Hold et al [5] found that the 399Ile allele was associated with an increased susceptibility to gastric cancer in a multi-stage study, which was supported by some other studies [4,6,7] , although there were some conflicting results [8] . However, the above two single nucleotide polymorphisms (SNPs) were absent in populations of eastern Asians [9][10][11] , including Chinese [10] . Alternatively, we found two common polymorphisms (minor allele frequency≥0.05, rs1927914 and rs10759932) located in the 5' flanking region of the TLR4 gene in a Chinese population by searching the NCBI SNP database, and these two polymorphisms may influence transcriptional factor binding site and modify the promoter activity and gene expression. In the present study, we hypothesized that the above two SNPs in TLR4 were associated with the risk of gastric cancer in the high-risk Chinese population. To test this hypothesis, we performed genotyping analysis for the two SNPs in the 5'-flaking region of TLR4, rs1927914 and rs10759932, in a casecontrol study with 1,053 gastric cancer cases and 1,100 cancer-free controls from a high incidence area of gastric cancer, Jiangsu Province, China.

Study subjects
Incident patients with histologically confirmed gastric adenocarcinoma were consecutively recruited from Jiangsu Province, including the cities of Yangzhong, Yixing, Yangzhou and Nanjing, between January 2003 and August 2007. Gastric cardia cancers were defined as tumors located within 20 mm distal to the gastro-esophageal junction, as we previously described [12] . The pathohistology was classified according to Lauren's classification [13] by senior pathologists. The exclusion criteria included reporting previous cancer history and having undergone radiotherapy or chemotherapy. The cancerfree controls were selected from individuals living in the same residential areas as the cases and were frequency-matched to the cases by age (±5 years) and sex. All the subjects were genetically unrelated Han Chinese.
Each participant was scheduled for an interview after a written informed consent was obtained and a structured questionnaire was administered by trained interviewers to collect information on demographic data and environmental exposure history including tobacco smoking and alcohol consumption. Those who had smoked <1 cigarette per day for less than 1 year in their lifetime were defined as non-smokers, otherwise, they were considered as smokers, and those who consumed 3 or more alcohol drinks a week for over 6 months were considered as alcohol drinkers. After an interview, an approximately 5-ml venous blood sample was collected from each subject. The serum antibodies against H. pylori CagA was determined by enzyme-linked immunosorbent assay (rat antihuman Cytotoxin Associated Gene A IgG, Adlitteram Diagnostic Laboratories, Shanghai, China) in 767 cases and 916 controls based on the availability of serum samples of the subjects. The cutoff value for positive and negative results was determined according to the manufacture's instructions. The OD value≥3.5 was defined as serum positive. The study was approved by the Institutional Review Board of Nanjing Medical University.

SNP selection and genotyping
The NCBI public SNP database was used to search all common (minor allele frequency≥0.05 in Chinese population) and potentially functional polymorphisms of the TLR4 gene, i.e., located at the 5' flanking region, 5' untranslated region or exons with amino acid change. Two SNPs in the 5' flanking region (rs1927914 T > C and rs10759932 T > C) were finally selected.
The SNP rs1927914 T > C was genotyped by the PCR-restriction fragment length polymorphism (R FL P) method. The pri mers were sense-5'-TGGGATTAAATGAACTGG-3' and antisense-5'-TGCTTGGAGGATATTACAG-3', which generated a 155 bp product. The PCR product was then digested by Nla Ⅲ (Fermentas International Inc, Canada) and separated on a 3% agarose gel stained with ethidium bromide. The C allele results in two fragments of 116 bp and 39 bp, and the T allele produce one fragment of 155 bp. The rs10759932 T > C variation was detected by a primer-introduced restriction analysis (PIRA)-PCR assay and a mismatched C was introduced to the anti-sense primer to replace T at + 2-bp from the polymorphic site to create a Hha I restriction site. The primers were sense-5'-TTTGTATAATTTGACTACCATTGCGT-3'and antisense-5'-CATTTTTTCACATCTTCACCAGC-3', and the amplified product (139 bp) was then digested by Hha I (Fermentas International Inc, Canada). The C allele produced two fragments of 117 bp and 22 bp, while the T allele resulted in a single 139 bp.
Genotyping was performed without knowing the subject's case or control status and the results were independently reviewed by two research assistants. To validate the genotyping results, 10% of the samples were randomly selected to confirm the assays, and the results were 100% concordant.

Statistical analyses
Goodness-of-fit χ 2 test was used to test the Hardy-Weinberg equilibrium for each SNP among controls. Differences in selected demographic variables, smoking and drinking status, serum CagA status, and frequencies of the genotypes and alleles between the cases and controls were evaluated by the χ 2 tests. The association between the genotypes of TLR4 SNPs and gastric cancer risk were estimated by computing the odds ratios (ORs) and 95% confidence intervals (CIs). The heterogeneity between stratifications was assessed with the Chi-square-based Q test. All the statistical analyses were performed by Statistical Analysis System software (version 9.1.3; SAS Institute, Cary, NC).

RESULTS
The distribution of selected characteristics between 1,053 gastric cancer patients and 1,100 cancer-free controls are summarized in Table 1. There were no significant differences in terms of distribution in age and sex between the two groups (P = 0.45 and 0.21, respectively), suggesting that our frequencymatching was adequate. The mean age for cases was 60.63±10.18 years and 60.05±10.36 years for controls. As shown in Table 1, we did not find significant differences in smoking and drinking status between the cases and controls (P = 0.22 for smoking and P = 0.29 for drinking). The positive rate of CagA in cases was 28.6% (219/767), while the control group was 25.7% (235/916), and the difference was not statistically significant between the two groups (P = 0.18). Among the gastric cancer cases, 472 (44.8%) were cardia gastric cancer, 482 (45.8%) noncardia gastric cancer, and 99 (9.4%) others (including mixed and unknown). For the histological subtypes, there were 735 (69.8%) intestinal-type, 188 (17.9%) diffuse-type, and 130 (12.3%) others (including mixed and unknown).
The genotype distribution of TLR4 rs1927914 and rs10759932 in the cases and controls are shown in Table 2. The observed genotype frequencies for the    In the stratified analysis (Table 3), the protective effect of the rs10759932 TC/CC genotype remained significant among young subjects (aged <60), males, drinkers, individuals with the negative serum CagA, and both intestinal and diffuse type cases. However, there were no significant differences between different subgroups stratified by these variables (P > 0.05 for heterogeneity test). In addition, we did not find any significant effect for rs1927914 TC/CC genotypes on risk of gastric cancer in different strata (data not shown).

DISCUSSION
In this case-control study of gastric cancer, we investigated the associations of two SNPs in the 5'-flanking region of TLR4 with the risk of gastric cancer in a Chinese population. We found, for the first time, that rs10759932 variant genotypes (TC/CC) of the TLR4 gene were associated with a significantly reduced risk of gastric cancer in this high-risk population.
TLR4 was identified as a potential binding receptor for H. pylori on gastric epithelial cells by recognizing the LPS of the gram-negative bacteria. In the human stomach, TLR4 is known to be expressed by epithelial cells, which was demonstrated to become greater in H. pylori gastritis than in non-inflamed gastric mucosa [14] . TLR4, in conjunction with CD14 and MD-2, transductes signals through MyD88, Toll/ IL-1 receptor domain and TRAF6. This promotes transcription of genes, which are involved in immune activation, including the transcription factor NF-kB and MAP kinase pathways [15] The dysregulation of TLR signaling may result in an unbalanced ratio between pro-and anti-inflammatory cytokines and thus induce the development of precancerous lesions and cancer [6,16] TLR4 Asp299Gly and Thr399Ile polymorphisms, located in the coding sequence and resulting in amino acid exchanges, were shown to affect the TLR4 extracellular domain and reduce the responsiveness to H. pylori LPS [17] , while Asp299Gly polymorphism was also observed to increase the proinflammatory tumor necrosis factor response [18,19] . Alejandra et al [4] reported that these two SNPs could alter mucosal cytokine and chemokine patterns in Mexican gastroduodenal diseases. Both SNPs have been implicated in the process of gastric cancer and its precursors [4][5][6][7] . However, these two SNPs are absent or rare in the Chinese population [10,20,21] and thus may not be an important factor in determining the outcome of the H. pylori infected individuals in such areas.
In this study, we identified two SNPs located in the 5' flanking region of TLR4 and found that rs10759932 variant genotypes (TC/CC) were associated with a significantly reduced risk of gastric cancer. It has been demonstrated that the low-functioning TLR4 polymorphisms may result in a reduced inflammatory response associated with a low-damaging infection resulting in persistent infection [22][23][24] , whereas TLR4 mRNA was up-regulated in gastric epithelial cell lines infected with H. pylori [25] . Therefore, it is biologically plausible that the 5' flanking region polymorphisms may influence the binding affinity of the transcriptional factors, regulate the transcriptional level of TLR4, involve the process of inflammatory response and host immunity, and finally modulate the development of gastric cancer. However, functional evaluations need to confirm these hypotheses in further studies.
Cytotoxin-associated gene pathogenicity island (cagPAI), a 40-kb cassette of genes that codes a type IV secretory system (T4SS) involved in the injection of the CagA protein into the epithelial cells, is one of the important virulence factors with regard to H.pylori infection. Within the cytosol, CagA is phosphorylated by multiple members of the Src family of kinases, considered to play a key role in carcinogenesis, which are involved in cell proliferation and differentiation [26] . Anti-CagA antibody of H. pylori was not observed to modify the development of gastric carcinogenesis in this study. Interestingly, in the stratified analyses, the protective effect of rs10759932 TC/CC genotypes seemed prominent in individuals with negative CagA, although there was not a significant difference between subgroups with different CagA status (P for heterogeneity = 0.07). This genetic variation may be associated with other virulence factors or the CagAnegative strain may be protective elements interacting with TLR4 gene.
Several potential limitations in this study need to be addressed. First of all, because the gastric cancer cases were recruited from the hospitals and the controls were selected from the general community, inherent selection bias cannot be completely excluded. However, potential confounding factors might have been minimized by matching age, sex and residential areas, and further controlled in data analysis with adjustment for age, sex, smoking and drinking status. Secondly, the serum status of H. pylori CagA was determined by enzyme-linked immunosorbent assay, which may not represent the whole spectrum of H. pylori infection, although CagA is considered one of most important virulence factors. Thirdly, the functional significance of rs10759932 was not determined in this study. Further studies with biological assays are needed to clarify the potential mechanism of the SNP modifying gastric cancer development.
In summary, this study revealed that the SNP rs10759932 in the 5'-flanking region of TLR4 was associated with gastric cancer risk in Chinese population, which provides further important evidence that genetic variants in the TLR4 gene contribute to gastric cancer susceptibility. However, large studies with different ethnic populations and functional evaluations are warranted to confirm these findings.