|Home | About | Journals | Submit | Contact Us | Français|
Host genetic factors play an important role in modifying the risk of human disease, including cancers of the upper gastrointestinal tract, with increasing interest in Toll-like receptor (TLR) signaling and the impact of genetic polymorphisms in these systems. The CD14-159C/T and the TLR9-1237T/C promoter polymorphisms have previously been shown to be associated with various inflammatory conditions including Helicobacter pylori-induced gastritis in Caucasian populations. In this study, we assessed the association of these two functional single nucleotide polymorphisms with gastric cancer in two independent Caucasian population-based case–control studies of upper gastrointestinal tract cancer, initially in 312 noncardia gastric carcinoma cases and 419 controls and then in 184 noncardia gastric carcinomas, 123 cardia carcinomas, 159 esophageal cancers, and 211 frequency-matched controls. Odds ratios were computed from logistic models and adjusted for potential confounding factors. No significant association was found between the CD14-159C/T and the TLR9-1237T/C promoter polymorphisms and increased risk of gastric cancer. Neither single nucleotide polymorphism has been assessed in a Caucasian gastric cancer case–control study before; although the CD14-159C/T polymorphism has been reported to show no apparent association with H. pylori-related gastric malignancy in a Taiwanese Chinese population. In conclusion, although our earlier preliminary studies suggested that the CD14-159C/T and the TLR9-1237T/C promoter polymorphisms increase the risk of precancerous outcomes, they do not seem to increase the risk of gastric cancer itself. This discrepancy merits further examination.
Helicobacter pylori is the most important acquired risk factor for gastric cancer and its interaction with the innate immune response, especially Toll-like receptors (TLRs) is key to clinical outcome (Hold et al., 2007). A single nucleotide polymorphism in the lipopolysaccharide (LPS) recognition receptor complex –such as TLR4 (TLR4+896A/G, rs4986790) has been shown to be a risk factor at various stages of H. pylori-induced gastric carcinogenesis (Hold et al., 2003, 2006, 2007; Kato et al., 2007). Another variant within the LPS recognition receptor complex, the CD14-159C/T (rs2569190), is situated close to a binding site for the Sp1 transcription factor, which is critical for CD14 expression. Carriage of the variant CD14-159 T allele is associated with increased soluble CD14 expression (Baldini et al., 1999) and inversely associated with levels of total serum IgE (Baldini et al., 1999; Koppelman et al., 2001) and has been studied in various inflammatory conditions (Hubacek et al., 1999; Arnott et al., 2004; Torok et al., 2004).
We have also studied a polymorphism within another TLR involved in bacterial recognition, such as TLR9 (TLR9-1237T/C, rs574383). In a work published in abstract form, we found that the polymorphism was associated with the development of premalignant gastric changes (Hold et al., 2006). TLR9 is responsible for initiating responses to bacterial CpG DNA in human inflammatory cells (Hemmi et al., 2000; Bauer et al., 2001), with upregulation of TLR9 mRNA expression noted in murine macrophages in response to bacterial LPS (An et al., 2002). Within the gastric environment, the presence of H. pylori has been shown to induce expression of TLR9 on gastric epithelial cells, and in-silico analysis of the TLR9-1237T/C promoter polymorphism indicates that the single nucleotide polymorphism (SNP) lies within a putative nuclear factor κB-binding site and is therefore a biologically plausible candidate SNP (Schmausser et al., 2004). The aim of our study was to assess the effect of the TLR9-1237 T/C and CD14-159C/T polymorphisms on the risk of upper gastrointestinal cancer in Caucasians.
The CD14-159C/T and TLR9-1237T/C SNPs were genotyped in two previously described independent Caucasian population-based case–control studies of upper gastrointestinal tract cancer (El Omar et al., 2000). The first was a gastric cancer case–control study derived from a Caucasian population in Warsaw, Poland in which there were DNA samples available from 327 noncardia gastric adenocarcinoma patients and 406 controls (Chow et al., 1999). The second was a multicenter esophageal and gastric cancer study conducted in three geographic areas of the United States with population-based tumor registries (Gammon et al., 1997); DNA samples were available from 306 patients with gastric adenocarcinoma (122 cardia and 184 noncardia, 90% Caucasian), 159 patients with esophageal cancer (52 with squamous cell carcinoma and 107 with adenocarcinoma, 90% Caucasian), and 211 population controls (94% Caucasian). The institutional review boards of the participating centers approved the study, and written informed consent was obtained from all participants.
CD14-159C/T and TLR9-1237T/C SNPs were genotyped with Taqman assays using minor groove binder AQ2probes. For CD14-159C/T, forward primer 5′ CTAGATGCCCTGCAGAATCCTT-3′ and reverse primer 5′ CCCTTCCTTTCCTGGAAATATTGCA-3′ were used along with wild-type probe VIC: CTGTTACGGCCCCCCT and variant allele probe FAM: CTGTTACGGTCCCCCT. For TLR9-1237T/C, forward primer 5′-CAGAGACATAATGGAGGCAAAGGA-3′ and reverse primer 5′-GCCTTGGGATGTGCTGTTC-3′ were used along with wild-type probe VIC: CTGCCTGAAAACT and variant allele probe FAM: TCTGCCTGGAAACT. Sequencing of selected genotypes was carried out to validate Taqman results.
Hardy–Weinberg equilibrium of alleles at individual loci was assessed by χ2 statistics. Odds ratios (ORs) with Cornfield 95% confidence intervals were computed by logistic regression using STATA version 7.0 software (STATA Press, College Station, Texas, USA). ORs for each cancer were adjusted for age (categorized as younger than 50, 50–59, 60–69, and 70 years or older), sex, and race (categorized as white and all other). Estimates of study power were assessed using Quanto (http://hydra.usc.edu/gxe/). The Polish gastric cancer study had 80–90% power to detect ORs of 1.6 or greater at a 5% significance level, and 50–60% power to detect ORs as low as 1.4.
In both control populations, the alleles of CD14-159C/T and TLR9-1237T/C were in Hardy–Weinberg equilibrium, with nonsignificant χ2 values. The allele frequencies in the control populations were similar to other documented Caucasian studies, CD14 159C/T variant allele (43–50%), TLR9-1237T/C variant allele (11–16%) (Tables 1 and and2).2). Neither SNP was associated with risk for gastric cancer in the Polish gastric cancer study (Table 1) nor with the various types of upper gastrointestinal cancer in the US-based case–control study (Table 2) in analyses adjusted for age, sex, and race (only Caucasians are reported as other ethnic groups were very small). The additional models, adjusted for other factors, gave qualitatively similar results (data not shown).
As a result of the known and putative functional effects of the CD14-159C/T and TLR9-1237T/C promoter polymorphisms, respectively, and our preliminary findings relating to their association with the development of H. pylori-induced premalignant gastric changes (Hold et al., 2006), it was important to assess their relevance to gastric cancer. Neither SNP has been assessed in a Caucasian gastric cancer case–control study before; although the CD14-159C/T polymorphism has been reported to show no apparent association with H. pylori-related gastric malignancy in a Taiwanese Chinese population (Wu et al., 2006). The CD14-159T allele has, however, been associated with increased risk of intestinal metaplasia in a Venezuelan population (Kato et al., 2007). Intestinal metaplasia, however, may not be the most appropriate surrogate marker of gastric cancer risk, as malignant potential depends on the subtype and this is often not defined in genetic association studies. Yet, in our study, neither SNP seemed to be risk factors for gastric cancer in either of the large Caucasian study populations tested. Ethnic-specific host susceptibility in gastric cancer development has been reported previously (Canedo et al., 2008).
On the basis of these findings, it is possible that the CD14-159C/T and TLR9 1237T/C polymorphisms are only risk factors at the early stages of the disease process. They may be relevant in defining the host immune response to H. pylori infection, but it would appear that they do not determine subsequent events in carcinogenic progression. Interestingly, variants in other genes that play a critical role in H. pylori-induced gastric cancer have also been identified as risk factors in the precursor stages of the disease process but not at the cancer stage (Savage et al., 2006).
It must also be considered that the CD14-159C/T and TLR9-1237T/C polymorphisms are not relevant markers in the study populations tested. Alternatively, it is possible that a more detailed assessment of the genes with more markers may show an association. The fact that two reasonably sized independent gastric cancer case–control studies have, however, failed to show a positive finding with these markers suggests that the results are genuine and that an association was not missed owing to low study power.