Int. J. Cancer: 119, 1666–1671 (2006) ' 2006 Wiley-Liss, Inc.
Host–bacterial interaction in the development of gastric precancerous lesions in a high risk population for gastric cancer in Venezuela Ikuko Kato1,2*, Leen-Jan van Doorn3, Federico Canzian4,5, Martyn Plummer4, Silvia Franceschi4, Jorge Vivas6, ~oz4 Gladys Lopez6, Yanhui Lu4, Lydie Gioia-Patricola4, Richard K. Severson1,7, Ann G. Schwartz1,8 and Nubia Mun 1 Karmanos Cancer Institute, Wayne State University, Detroit, MI 2 Department of Pathology, Wayne State University, Detroit, MI 3 DDL Diagnostic Laboratory, Fonteijnenburghlaan Voorburg, The Netherlands 4 International Agency for Research on Cancer, Lyon, France 5 Genomic Epidemiology Group, Division of Molecular Genetic Epidemiology, German Cancer Research Center, Deutsches Krebsforschungszentrum (DKFZ), Heidelberg, Germany 6 Cancer Control Center of the Tachira State, San Cristobal, Venezuela 7 Department of Family Medicine, Wayne State University, Detroit, MI 8 Department of Internal Medicine, Wayne State University, Detroit, MI Helicobacter pylori (HP) infection affects over 50% of the world’s population. The prevalence is over 90% in populations at high risk for gastric cancer, but clinical outcomes of the infection are highly variable and thus host genetic factors have been suggested to play a role in its outcomes in addition to bacterial factors. In this study, we examined the effects of common functional genetic polymorphisms of several proinflammatory cytokines known to be overexpressed in HP-infected gastric mucosa on the risk of various stages of gastric premalignant lesions. The odds ratios (ORs) and 95% confidence intervals (CI) for atrophic gastritis, intestinal metaplasia and dysplasia were estimated by multinominal logistic regression analysis among 2,033 Venezuelan subjects. There was a significant effect of IL8 -251A allele on the prevalence of dysplasia (p 5 0.021). The OR associated with the A-allele was 1.34 (95% CI: 0.82–2.18) for heterozygotes and 2.00 (95% CI: 1.13–3.56) for homozygotes, compared with the TT genotype. Furthermore, there was a statistically significant interaction between the number of A-alleles and HP cag A genotype (p 5 0.009), suggesting that the A-allele increased the risk of dysplasia only when cag A was present. The OR for the AA compared with TT genotype was 3.22 (95% CI: 1.60–6.52) in this group. There were no associations with other proinflammatory cytokines studied, i.e., IL1b, IL6, monocyte chemoattractant protein 1 (MCP1) and TNFa, or with other stages of premalignant lesions. The present study provides important evidence suggesting host–bacterial interactions in the development of gastric precancerous lesions. ' 2006 Wiley-Liss, Inc. Key words: cytokines; genetic polymorphisms; stomach cancer; premalignant lesions; Helicobacter pylori
Helicobacter pylori (HP) infection affects over 50% of the world’s population. The prevalence ranges from 20% in developed or industrialized countries to over 90% in developing countries.1 Although most infected individuals remain asymptomatic throughout their lives, sustained infection with this organism has been firmly established as a causative factor for peptic ulcers, gastric adenocarcinoma and mucosa-associated lymphoid tissue lymphoma.2 HP infection in the stomach entails a complex immune reaction in the gastric mucosa, characterized by upregulation of a variety of proinflammatory cytokines and inflammatory mediators,3 as detailed later. However, this immune response is often not sufficient to eliminate the bacteria and thus leaves the host susceptible to complications from chronic inflammation. These include a series of gastric premalignant conditions which represent a continuum of changes from normal to carcinoma.4 Interleukin (IL)8 is a potent neutrophil-activating chemotactic chemokine that is most often upregulated in HP-infected gastric mucosa and produced by gastric epithelial cells,5 as well as monocytes and macrophages.6 IL8 regulates neutrophil infiltration, free radical generation and the release of proteolytic enzymes from activated neutrophils,7 which increases oxidative stress to the gastric mucosa.8 Reactive oxygen or nitrogen metabolites from infilPublication of the International Union Against Cancer
trating neutrophils mediate the formation of carcinogenic compounds in gastric mucosa and lumen, induce DNA damage to the adjacent epithelial cells and stimulate cell proliferation,9,10 which eventually leads to neoplastic transformation. Monocyte chemoattractant protein 1 (MCP1), which is also named CCL2, on the other hand, is a chemokine which regulates macrophage infiltration and release of superoxide anion and lysosomal enzymes from monocytes. Its role in acute and chronic inflammatory states has been well established.11 Increased expression of both chemokines in gastric epithelial cells has been noted, especially with cag A1 HP strains.5,12 IL1b exerts a broad range of biological effects to enhance the host’s inflammatory response by inducing expression of many other genes, such as tumor necrosis factor (TNF), IL6, IL8, MCP1, cyclooxygenase-2 and nitric oxide synthase 2.11,13 Its direct effects on gastric mucosa include inhibition of gastric acid secretion and stimulation of gastrin release.14 TNFa is the most abundantly produced cytokine in activated macrophages and plays a central role in initiating and regulating the cytokine cascade during an inflammation response.15 Its biological properties overlap those of IL1b.13 TNFa induction is also correlated with the degree of apoptosis and its role in tumor promotion is suggested in a TNFa-deficient mice model.16 IL6 is another cytokine which is primarily responsible for the proliferation and differentiation B- and T-lymphocytes and the differentiation of macrophages, osteoclasts and megakaryocytes. It is produced by T-cells, monocytes, fibroblasts and some epithelial cells.17 Findings from experimental animal models indicate that IL6 is important in either maintaining acute inflammation or converting into a chronic phase, while evidence exists in support of both pro- and antiinflammatory activities of IL6.17 Interestingly, genes encoding these cytokine molecules have been found to contain common functional polymorphisms which lead to either enhanced or reduced activities. These findings may explain the highly variable clinical outcomes of HP infection and the fact that only a small fraction of infected people experience the progression of gastric premalignant lesions. In this study, we seek to address whether common functional polymorphisms in the selected proinflammatory cytokine genes are associated with the prevalence of advanced stages of gastric premalignant lesions and
This paper is dedicated to the memory of Dr. Julio Torrado, who died in June 2005. He was a Spanish pathologist and had participated in the research project in Venezuela. Grant sponsor: US National Cancer Institute; Grant number: CA 98309. *Correspondence to: Karmanos Cancer Institute, 110 East Warren Avenue, Detroit, MI 48201, USA. Fax: 11-313-831-7806. E-mail:
[email protected] Received 20 January 2006; Accepted 14 February 2006 DOI 10.1002/ijc.21979 Published online 2 May 2006 in Wiley InterScience (www.interscience. wiley.com).
1667
HOST–BACTERIAL INTERACTION IN GASTRIC PRECURSORS TABLE I – PCR PRIMERS, PROBES AND LABELS FOR TaqMan GENOTYPING ASSAYS Gene
Polymorphism
dbSNP
PCR primers (50 –30 )
TaqMan probes (50 –30 )
CCCCTTTCCTTTAACTTGATTGTG AGGTTTGGTATCTGCCAGTTTCTC TGACGACCTAAGCTGCACTTTTC GGGCTGATTGGAAACCTTATTAAGA TAAAATACTGAAGCTCCACAATTTGG ATCTTGTTCTAACACCTGCCACTCT GGGAGGGCATCTTTTCTTGAC GGTGAAGGGTATGAATCAGAAAAGA GAAATGGAGGCAATAGGTTTTGAG GGCCACTGACTGATTTGTGTGTAG
VIC-TTTGAAAGCTATAAAAACAG FAM-TTTGAAAGCCATAAAAA VIC-TCTTGCGATGCTAAA FAM-TCTTGCCATGCTAAA VIC-CATACAATTGATAATTCA FAM-CATACATTTGATAATTCA VIC-ACAGCTATCACTTTC FAM-AGACAGCTGTCACTTT VIC-CCGTCCCCATGCC FAM-CCGTCCTCATGCC
IL1B
C-31T
rs1143627
IL6
G-174C
rs1800795
IL8
T-251A
rs4073
MCP1
G-2518A
rs1024611
TNF
G-308A
rs1800629
whether bacterial factors modify the effects of such polymorphisms in a population known to have extremely high HP infection and gastric cancer rates.
none of the above-listed lesions was detected in any biopsy. The HP-positive rate based on histological staining was 67.3% for normal/SG, 97.5% for CG, 94.1% for AG, 94.4% for IM and 86.4% for dysplasia according to the global diagnosis.
Material and methods Study population Details concerning the study design and characteristics of study participants have been published elsewhere.18 Briefly, the original study population consisted of 2,200 participants in a randomized placebo-controlled trial for gastric precancerous lesions using b-carotene, a-tocopherol and ascorbic acid. They were recruited from participants in the gastric cancer screening program of the state of Tachira, Venezuela, who were between 35 and 69 years of age in general good health. They were all classified as Hispanic for ethnicity, but no further race information was obtained. At baseline, participants underwent gastroscopic examination with biopsies, collection of blood and urine specimens and completion of questionnaires with written informed consent. All data and specimens used in the present study were obtained at baseline before any antioxidant treatment was started.
Helicobacter genotyping Total DNA was released from gastric biopsy specimens using proteinase K treatment. Briefly, biopsies were incubated with 250 ll of 10 mM Tris HCl (pH 8.0), 5 mM EDTA, 0.1 % SDS and 0.1 mg/ml proteinase K for at least 2 hr at 51°C. Proteinase K was inactivated by incubation at 95°C for 10 min. Ten microliters of the above reaction mixture was used for a multiplex PCR reaction. Presence of cag A was determined by hybridization of the multiplex PCR products in a microtiter plate, using 2 cag A-specific probes.20 The presence of H. pylori (irrespective of the cag A status) was checked by analyzing the presence of vac A s-region amplimers by hybridization of the same multiplex PCR products in a microtiter plate, using a specific general probe for the vac A s-region.20
Gastroscopic and histological examination At gastroscopic examination 5 biopsies for histological assessment were taken from the lesser curvature of the antrum 1 cm from the pylorus, from the greater curvature of the antrum 1 cm above the pylorus, from a mid-portion of the lesser curvature of the antrum, from the lesser curvature of the antrum immediately below the incisura and from the middle corpus 2 cm from the lesser curvature. They were fixed in buffered formalin and stained with hematoxylin–eosin and with Giemsa stain to define H. pylori infection. Subjects’ slides were assigned to 1 of the 3 pathologists at the pathology laboratory of the Cancer Control Center, for their review. Before the project started, these pathologists participated in training sessions with consultant pathologists, expert in gastric pathology (Drs. P. Correa, Louisiana State University, USA and H. Ramirez, Universidad de Valle, Columbia), and advanced lesions (dysplasia in particular) were re-reviewed by Dr. Ramirez in the middle of the project. Histological diagnoses were classified as described in details previously19: Superficial gastritis (SG), characterized by infiltration of lymphocytes, plasmocytes and polymorphonuclear leukocytes in the superficial portion of the lamina propria and by focal necrosis and regenerative changes in the glandular neck region; Chronic gastritis (CG), lymphoplasmocytic cells infiltrate deeper layers of the mucosa; Chronic atrophic gastritis (AG), characterized by loss of normal glands which are partly replaced by connective tissue and inflammatory cells; Intestinal metaplasia (IM), the gastric glands are replaced by cells normally present only in the intestine: absorptive cells, goblet cells, argentaffin cells and Paneth’s cells; IM subtypes were classified according to Filipe and Jass; Dysplasia: characterized by closely packed tubular glands with diminished mucus secretion, scant cytoplasm, and large, hyperchromatic, crowded and elongated nuclei. The most advanced among all evaluable biopsies was used as the global diagnosis for a given subject. A global diagnosis of ‘‘normal’’ was applied when
Human genotyping analysis Buffy coat or other high-yield sources of DNA were not available for this study; therefore, we decided to use plasma as the primary source of genomic DNA (96%). For the rest, a DNA aliquot from a frozen biopsy which was extracted for HP genotyping was used. Details of procedures for DNA extraction from plasma as well as for whole genome amplification (WGA) were described in our previous publication.21 DNA samples obtained from WGA were genotyped by the 50 exonuclease (TaqMan) assay for 5 single nucleotide polymorphisms (SNPs) on the IL1B, IL6, IL8, MCP1 and TNF genes, which encode IL1b, IL6, IL8, MCP1 and TNFa proteins respectively, using the primers and probes as described in Table I. The reaction mix included 10 ng of post-WGA DNA, 5 pmol of each primer, 1 pmol of each probe and 4 ll of 23 master mix (Applied Biosystems) in a final volume of 8 ll. The thermocycling was performed in 96-well plates and included 40 cycles with 30 sec at 95°C followed by 60 sec at 60°C. PCR plates were read on an ABI PRISM 7900HT instrument (Applied Biosystems, Foster City, CA). All readings were performed independently by 2 operators, and only concordant calls were kept. All samples that did not give a reliable result in the first round were resubmitted for up to 2 additional rounds of genotyping. All genotypes that could not be obtained after 3 attempts were left blank. Statistical analysis Out of the 2,200 original study subjects, 1 was removed because of no evaluable gastric biopsy specimens. Human genotyping was possible for a total of 2,180 subjects. A total of 2,033 subjects (92.4%) were genotyped for HP cag A. Accordingly, analyses were limited to these 2,033 subjects for whom both bacterial and human genotyping results were available. The response variable in this study was global histological diagnosis, which was divided into 4 groups: dysplasia, IM, atrophic gastritis and normal to chronic gastritis. The last group served as the control group in this study because the combined frequency of normal and superficial gastritis in this population was still less than 5%. Multinominal logistic regression analysis was employed, using the SAS CATMOD procedure,22 to estimate odds ratios (ORs) and
1668
KATO ET AL. TABLE II – GENOTYPE AND ALLELE FREQUENCIES OF THE POLYMORPHISMS STUDIED AMONG THE STUDY POPULATION IN VENEZUELA Genes
IL1B IL6 IL8 MCP1 TNF
Polymorphisms (1 position 2)1
Number genotyped
C-31T G-174C T-251A G-2518A G-308A
1,962 1,951 1,992 1,946 1,963
Genotypes (%) 1–1
29.15 59.30 32.98 28.88 84.92
Alleles (%)
1–2
2–2
1
2
48.47 34.75 49.85 47.58 14.01
22.38 5.95 17.17 23.54 1.07
53.39 76.68 57.91 52.67 91.93
46.61 23.32 42.09 47.33 8.07
1
1 indicates a base for a major allele and 2 a base for a minor allele.
TABLE III – ODDS RATIOS (OR)1 AND 95% CONFIDENCE INTERVALS (CI) FOR GASTRIC PRECANCEROUS LESIONS ACCORDING TO GENETIC POLYMORPHISMS RELATED TO INFLAMMATORY RESPONSES Gene/genotype
IL1B TT CT CC Dose-effect Dominant2 Recessive3 IL6 GG CG CC Dose-effect Dominant Recessive IL8 TT AT AA Dose-effect Dominant Recessive MCP1 AA AG GG Dose-effect Dominant Recessive TNF GG AG/AA
No. of controls
Atrophic gastritis No
234 508 301
64 142 79
809 301
211 64
623 344 69
175 93 11
413 69
104 11
348 542 172
93 142 50
714 172
192 50
307 503 228
79 127 68
731 228 894 153
OR
Intestinal metaplasia (95% CI)
1.00 – 1.01 (0.72–1.41) 0.91 (0.63–1.33) p 5 0.613 0.97 (0.71–1.33) 0.91 (0.68–1.22)
No
115 236 169 405 169
1.00 – 0.98 (0.73–1.30) 0.57 (0.29–1.11) p 5 0.223 0.91 (0.69–1.20) 0.58 (0.30–1.11)
293 201 26
1.00 – 0.98 (0.73–1.32) 1.07 (0.72–1.58) p 5 0.817 1.00 (0.75–1.32) 1.09 (0.76–1.53)
187 251 91
146 239 134
195 68
1.00 – 1.02 (0.74–1.40) 1.18 (0.82–1.71) p 5 0.399 1.07 (0.79–1.44) 1.17 (0.85–1.60)
230 50
1.00 1.27
– (0.89–1.80)
227 26
OR
(95% CI)
1.00 – 0.96 (0.72–1.27) 1.12 (0.82–1.51) p 5 0.413 1.02 (0.78–1.32) 1.15 (0.91–1.46) 1.00 – 1.22 (0.96–1.54) 0.82 (0.50–1.34) p 5 0.560 1.15 (0.92–1.44) 0.76 (0.47–1.24)
Dysplasia No
26 65 23 88 23 66 40 10 50 10
OR
(95% CI)
1.00 – 1.18 (0.71–1.96) 0.72 (0.39–1.33) p 5 0.295 1.02 (0.63–1.65) 0.64 (0.39–1.06) 1.00 – 0.96 (0.62–1.49) 1.41 (0.66–3.01) p 5 0.628 1.03 (0.68–1.54) 1.43 (0.68–3.01)
1.00 – 0.89 (0.70–1.14) 0.99 (0.72–1.37) p 5 0.773 0.92 (0.73–1.15) 1.06 (0.79–1.42)
29 58 29
30 57 28
373 134
1.00 – 1.00 (0.77–1.30) 1.19 (0.87–1.60) p 5 0.293 1.06 (0.83–1.34) 1.19 (0.92–1.53)
85 28
1.00 – 1.15 (0.70–1.87) 1.15 (0.65–2.04) p 5 0.628 1.15 (0.72–1.82) 1.05 (0.66–1.70)
447 74
1.00 0.91
96 19
1.00 1.06
342 91
– (0.67–1.25)
87 29
1.00 – 1.34 (0.82–2.18) 2.00 (1.13–3.56) p 5 0.021 1.50 (0.95–2.38) 1.66 (1.03–2.68)
– (0.61–1.84)
1 Adjusted for age, sex, educational level, length of refrigerator use, fruit and starchy vegetable intakes, cigarette smoking and PCR results for cag A1 and cag A2 HP strains.–2Dominant model indicates a comparison between subjects with 1 or 2 variant alleles and no alleles.–3Recessive model indicates a comparison between subjects with two variant alleles and others.
95% confidence intervals (CIs) associated with selected human genetic polymorphisms for atrophic gastritis, IM and dysplasia, in comparison with controls. The ORs were presented for high-activity genotypes (hetero- and homozygotes) according to functional data available in the literature. The ORs were also calculated using both dominant and recessive models, as long as high-activity homozygotes consisted of more than 4% of the sample. The gene-dose effect was also tested with the logit of risk according to the number of high-activity alleles. When these analyses were not feasible because of genotype frequency of less than 2%, only the ORs for either dominant or recessive model were presented. All ORs were adjusted for basic demographic variables (sex, age and educational levels), HP genotypes (cag A1 and cag A2 strains) and other environmental risk factors reported previously (cigarette smoking, quintile levels of fruit and starchy vegetable intakes, and duration of refrigerator use),23 unless a covariate was used as a stratification variable in calculating separate ORs in some analyses. The interactions between human genetic polymorphisms and bacterial and other environmental factors or between 2 human genetic polymorphisms were tested by including multiplicative interaction terms between those variables as well as by calculating strata-specific ORs.
Results As shown in Table II, genotyping for individual polymorphisms was successful for 96–97% of the samples. Except for TNF, the polymorphisms examined in this study were found to be very common in this Hispanic population in Venezuela. Minor allele frequencies were close to 50% for IL1B and MCP1, 42% for IL8 and 23% for IL6. We tested departure from Hardy–Weinberg equilibrium by use of v2 test, using p 5 0.01 as threshold. This threshold was chosen on the basis of anticonservativeness of this test, as noted by Wigginton et al.24 All SNPs were in Hardy–Weinberg equilibrium in this population. As shown in Table III, the prevalence of IL1B C-allele was not associated with risk of any type of premalignant lesions, as the ORs for the dominant model were all virtually unity. There were no dose–response relationships with increasing number of C-alleles for any type of lesions. Compared with subjects with the CT or TT genotype, those with the CC genotype tended to have higher risk of IM (OR 5 1.15, 95% CI: 0.91–1.46), but lower risk of dysplasia (OR 5 0.64, 95% CI: 0.39–1.06), although these findings did not reach statistical significance.
1669
HOST–BACTERIAL INTERACTION IN GASTRIC PRECURSORS 1
TABLE IV – ODDS RATIOS (OR) AND 95% CONFIDENCE INTERVALS (CI) FOR GASTRIC DYSPLASIA ACCORDING TO GENETIC POLYMORPHISMS OF PROINFLAMMATORY CYTOKINES AND HELICOBACTER PYLORI GENOTYPES cag A1 Gene/genotype
IL1B TT CT/CC IL6 GG/CG CC IL8 TT AT AA Dose-effect MCP1 AA AG/GG TNF GG AG/AA
No. of controls
cag A (2)
Dysplasia No
OR
(95% CI)
No. of controls
Dysplasia No
OR
(95% CI)
82 249
22 64
1.00 1.01
– (0.57–1.77)
152 559
4 24
1.00 1.53
– (0.52–4.55)
304 27
81 6
1.00 0.86
– (0.33–2.24)
662 42
25 4
1.00 2.82
– (0.89–8.93)
114 172 54
18 43 27
1.00 – 1.58 (0.85–2.93) 3.22 (1.60–6.52) p 5 0.001
234 369 118
11 15 2
1.00 – 0.98 (0.44–2.21) 0.39 (0.08–1.83) p 5 0.329
97 236
20 68
1.00 1.24
– (0.70–2.20)
210 494
10 17
1.00 0.69
– (0.31–1.57)
280 56
73 14
1.00 0.90
– (0.41–1.74)
613 97
23 5
1.00 1.18
– (0.43–3.26)
1 Adjusted for age, sex, educational level, length of refrigerator use, fruit and starchy vegetable intakes and cigarette smoking.
There was a significant gene-dose-effect of IL8 A-allele on the prevalence of dysplasia (p 5 0.021). The OR was 1.34 (95% CI: 0.82–2.18) for heterozygotes and 2.00 (95% CI: 1.13–3.56) for homozygotes. The OR for the homozygotes compared with the others was 1.66 (95% CI: 1.03–3.56). There were no associations between the presence of this allele and any other types of preneoplastic lesions, including advanced stage IM, types 2 and 3 (OR for AA vs. TT 0.86, 95% CI: 0.53–1.39). Furthermore, there was no statistically significant interaction between the number of IL8 A-alleles and IL1B CT/CC genotype on the prevalence of dysplasia. There was an 40% increase in risk of dysplasia and a 40% decrease in risk of atrophic gastritis associated with the IL6 Callele, a 20% increase in risk of any lesions with the MCP1 GG genotype, compared with AA, and an 30% increase in risk of atrophic gastritis associated with TNF A-allele, but none of these were statistically significant. When interaction terms between these genetic polymorphisms and bacterial and other environmental factors were tested concerning the risk of dysplasia, there were no statistically significant interactions except that with cag A HP genotype. Table IV reports the ORs for dysplasia according to HP cag A status. The cag A2 group was formed regardless of HP status on histology or by PCR because it was not possible to calculate the ORs for HP-negative group due to too few cases of dysplasia divided by genotype. There were clear differences in the OR associated with IL8 polymorphisms by cag A status. The A-allele increased the risk of dysplasia only when cag A was present, and the OR for the AA compared with TT genotype was 3.22 (95% CI: 1.60–6.52). On the other hand, a nonsignificant decrease in risk was observed when cag A was not present. As a result, the OR for the interaction between the number of A-alleles and HP cag A genotype was highly statistically significant (p 5 0.009). The ORs for dysplasia associated with IL6 CC genotype and IL1B C-allele were more pronounced when cag A was negative, but none of the ORs or their interaction terms were statistically significant. There were no particular trends in risk according to MCP1 and TNF polymorphisms by cag A status. The risk of IM and atrophic gastritis associated with the IL8 polymorphism did not significantly differ by cag A status (data not shown). Discussion In this study, we examined the effects of common polymorphisms of several proinflammatory cytokine genes on the risk of various stages of gastric premalignant lesions. These cytokines are
known to be overexpressed in HP-infected gastric mucosa, and have been suspected to play roles in gastric carcinogenesis via enhancing inflammatory activities and/or via separate biological pathways as discussed earlier. In addition, functional significance of the polymorphisms studied has been well characterized. To date, limited information is available on the associations of these genetic polymorphisms and risk of gastric premalignant lesions, whereas there have been a number of case–control studies for gastric cancer.25–36 We recommend that caution should be exercised in comparing our results with those of other cross-sectional studies for several reasons. First, because of the very high prevalence of HPrelated gastric premalignant lesions in populations at high risk for stomach cancer, controls from studies in such populations may include a substantial proportion of individuals with a high-grade preneoplastic lesion. On the other hand, cases in low-risk populations may include a significant number of patients with gastric cancer which is not HP-related. Our study, which is based on an endoscopic cohort with detailed histopathology data, was advantageous in excluding advanced stages of precancerous lesions from the controls, but the fact that the majority of the controls had CG might have diminished differences between the cases and controls. Nevertheless, in view of the fact that the majority of adults in high-risk areas have a certain degree of CG,18 our control group is considered representative of the general population in such high-risk areas. The IL8 gene was mapped to chromosome 4q12-13 and a common SNP in the promoter region T(-251)A was identified. The A-allele was associated with increased production of IL8 by LPSstimulated blood and with increased risk of bronchiolitis in infants.37 Subsequently, several investigators have reported its association with gastric cancer or precancerous lesions. An increased risk of noncardiac gastric cancer was found among individuals carrying the A-allele in Chinese25 and Japanese26 populations, where the cag A1 strain prevalence is very high. The latter study also reported no overall association of this allele with histological scores for atrophy and metaplasia. Both studies revealed that the risk of gastric cancer was almost doubled among those with the AA genotype, compared with the TT, but the risk associated with heterozygotes was inconsistent with the former, showing no association, and the latter the risk as high as homozygotes. Our results on dysplasia are in almost complete agreement with these results, but those on heterozygotes support a co-dominant model. Other studies, which are less relevant to HP infection, have shown a positive association of the A-allele with risk of gastric cardiac adenocarcinoma27 and an inverse association with diffuse type of gastric carcinoma.28
1670
KATO ET AL.
While more pronounced cytokine induction by certain HP strains has been reported,5,38 to date there have been limited publications concerning interactions between host genetic polymorphisms and bacterial genotypes, except for those on IL1B polymorphisms which showed joint effects rather than synergetic interactions.39,40 Our study is the first to address the interaction between host IL8 polymorphisms and HP cag A genotypes. Our results were very intriguing, suggesting that the effects of host genetic polymorphisms are context-dependent. In fact, IL8 high-activity genotype did not have any effect without HP cag A and was rather implicative of protective effects. The latter may be possible if enhanced IL8 induction leads to successful elimination of HP from the stomach or to reinforced mucosal defense through its angiogenic and proliferative properties. However, the number of dysplasia patients in cag A or HP-negative groups was too small to draw any reliable conclusions, including those on negative interactions with IL6 and IL1B polymorphisms. It should also be noted that these subgroups analyses are prone to chance findings. Nevertheless, biological interaction between cag A and IL8 high-activity genotype seems quite plausible. First, enhanced IL8 induction has often been described with cag A1 HP strains,5,38 although strong linkage of cag A with other virulence factors, such as vac A and opi A, often makes it difficult to isolate the impact of individual virulence factors. It has been postulated that this induction involves activation of nuclear factor jB and interferon regulatory factor-1.41,42 Second, cag A is also capable of inducing cytoskeletal rearrangement and morphological changes in gastric epithelial cells.43,44 Such effects may be compounded by increased oxidative damage8 and mitogenic activity45 caused by IL8 stimulation and thus may ultimately lead to dysplasia. Correspondingly, in a meta-analysis by Huang et al. it was found that infection with cag A1 strain further increased the risk of noncardiac gastric cancer by 2-fold among HP-positive populations46 and both cag A antibody and increased serum IL-8 levels were predictive of gastric cancer in a case–control study in Poland.47 Collectively, the data from our study and others suggest that the involvement of IL8 is likely to be in the late stage of the gastric carcinogenic process, which takes at least 2 decades. The absence of an association with earlier stages of premalignant lesions is not necessarily contradictory, because altered gene expression by HP infection appears to be specific to stages of gastric premalignant and malignant lesions.48 The present study did not confirm associations between genetic polymorphisms in the IL1B and TNF genes that were reported in some earlier publications. The IL1B polymorphism was the first that was linked to gastric cancer. The initial observation by ElOmar et al. that the C-allele at position -31 and the T-allele at
position -511 were in almost complete linkage disequilibrium and associated with increased risk of gastric cancer29 was confirmed by several subsequent studies conducted in different populations.30–33 But, conflicting observations exit,25,34,35 so do functional data.29 Among several SNPs in the TNF promoter, functional significance has been best established for G-308A and the A-allele has been designated as a high-activity allele that leads to increased transcription and production.49 The A-308 allele was found to be associated with gastric cancer risk in 3 studies,25,30,32 while others33,35 failed to find any association. To our knowledge, there have been no reports for the significant positive associations between genetic polymorphisms at the MCP1 -2518 or at the IL6-174 and gastric cancer or precursors, despite their clear functionality.50,51 There are several limitations in this study. First, the effect of possible misclassification in histological diagnoses was discussed in details elsewhere.23 It is likely to be nondifferential and to result in an underestimation of ORs. Second, there is high likelihood that some of the associations observed in this study were chance findings due to multiple comparisons, although the number of genes tested in this study is approximately equal to the number of environmental risk factors typically studied in epidemiological investigations. Third, the polymorphisms examined in this study are located in promoter regions and are thought to exert their effects through changes in transcriptional levels. However, there are many other factors at posttranscriptional levels which may modify biological activities and thus overall contribution of these polymorphisms on net biological activities remains to be elucidated. In summary, while no associations with most of cytokine genes, our findings on IL8 suggest that host–bacterial interactions play an important role in the development of gastric precancerous lesions. Our results indicate that some genetic polymorphisms exert their effects only in the presence of certain environmental factors. This may help to explain the inconsistent findings from previous studies of human common genetic polymorphisms (increasingly apparent in recent years), and calls for special attention to such modifiers in evaluating the results from different populations. Acknowledgements This study was supported by the US National Cancer Institute through a research grant to Dr. I Kato (CA 98309). We thank Drs. Simon Peraza, Dennis Castro, Victor Sanchez, Elsa Cano, Olga Andrade, and Rita Garcia for their contribution to endoscopic and pathological evaluation, Dr. Walter Oliver for institutional support for the original trial and Mr. Didier Colin, Mr. Bernhard Kleter and Ms. Bianca Gieze for their excellent technical assistance in statistical and laboratory analyses.
References 1. 2. 3.
4.
5.
6.
Arora S, Czinn SJ. Vaccination as a method of preventing Helicobacter pylori-associated gastric cancer. Cancer Epidemiol Biomarkers Prev 2005;14:1890–1. IARC. IARC monographs on the evaluation of carcinogenic risks to humans, vol. 61: schistosomes, liver flukes and Helicobacter pylori. Lyon: IARC, 1994. Jung HC, Kim JM, Sing IS, Kim CY. Helicobacter pylori induces an array of pro-inflammatory cytokines in human gastric epithelial cells: quantification of mRNA for interleukin-8, -1 a/b, granulocyte-macrophage colony-stimulating factor, monocyte chemoattractant protein-1 and tumor necrosis factor-a. J Gastroenterol Hepatol 1997;12:473–80. Correa P, Haenszel W, Cuello C, Zavala D, Fontham E, Zarama G, Tannenbaum S, Collazos T, Ruiz B. The gastric precancerous process in a high risk population: cross-sectional studies. Cancer Res 1990;50: 4731–6. Martin Guerrero JM, Hergueta Delgado P, Esteban Carretero JJ, Rivera Hueto FF, Pellicer Bautista FJ, Herrerias Gutierrez JM. Pathogenic implications of interleukin-8 activity and bacterial phenotype in antral gastritis associated with Helicobacter pylori. Rev Esp Enferm Dig 2000;92:301–15. Harris PR, Mobley HL, Perez-Perez GI, Blaser MJ, Smith PD. Helicobacter pylori urease is a potent stimulus of mononuclear phagocyte activation and inflammatory cytokine production. Gastroenterology 1996;111:419–25.
7. 8. 9. 10. 11.
12. 13. 14. 15.
Baggiolini M, Dewald B, Moser B. Interleukin-8 and related chemotactic cytokines—CXC and CC chemokines. Adv Immunol 1994;55: 97–179. Li C-Q, Pignatelli B, Ohshima H. Increased oxidative and nitrative stress in human stomach associated with cagA1 Helicobacter pylori infection and inflammation. Dig Dis Sci 2001;46:836–44. Naito Y, Yoshikawa T. Molecular and cellular mechanisms involved in Helicobacter pylori-induced inflammation and oxidative stress. Free Radic Biol Med 2002;33:323–36. De Luca A, Iaquinto G. Helicobacter pylori and gastric diseases: a dangerous association. Cancer Lett 2004;213:1–10. Mukaida N, Harada A, Yasumoto K, Matsushima K. Properties of pro-inflammatory cell type-specific leukocyte chemotactic cytokines, interleukin 8 (IL-8) and monocyte chemotactic and activating factor (MCAF). Microbiol Immunol 1992;36:773–89. Mori N, Ueda A, Geleziunas A, Wada A, Hirayama T, Yoshimura T, Yamamoto N. Induction of monocyte chemoattractant protein 1 by Helicobacter pylori involves NF-j B. Infect Immun 2001;69:1280–6. Dinarello CA. The biology of interleukin 1. Chem Immunol 1992;51: 1–32. El-Omar EM. The importance of interleukin 1b in Helicobacter pylori associated disease. Gut 2001;48:743–7. Makhatadze NJ. Tumor necrosis factor locus: genetic organization and biological implications. Hum Immunol 1998;59:571–9.
HOST–BACTERIAL INTERACTION IN GASTRIC PRECURSORS
16. Suganuma M, Okabe S, Marino MW, Sakai A, Sueoka E, Fujiki H. Essential role of tumor necrosis factor a (TNF-a) in tumor promotion as revealed by TNF-a-deficient mice. Cancer Res 1999;59:4516–18. 17. Wong PKK, Campbell IK, Egan PJ, Ernst M, Wicks IP. The role of the interleukin-6 family of cytokines in inflammatory arthritis and bone turnover. Arthritis Rheum 2003;48:1177–89. 18. Mu~ noz N, Vivas J, Buiatti E, Kato I, Oliver W. Chemoprevention trial on precancerous lesions of the stomach in Venezuela: summary of the study design and baseline data. In: Stewart BW, McGregor D, Kleihues P, eds. Principles of chemoprevention. Lyon: IARC, 1996. IARC Science Publication No. 139,125–33. 19. Mu~ noz N, Kato I, Peraza S, Lopez G, Carrillo E, Ramirez H, Vivas J, Castro D, Sanchez C, Buiatti E, Oliver W. Prevalence of precancerous lesions of the stomach in Venezuela. Cancer Epidemiol Biomarkers Prev 1996;5:41–6. 20. van Doorn LJ, Figueiredo C, Rossau R, Jannes G, van Asbroeck M, Sousa JC, Carneiro F, Quint W. Typing of the Helicobacter pylori vacA gene and detection of the cagA gene by PCR and reverse hybridization. J Clin Microbiol 1998;36:1271–6. 21. Lu Y, Gioia-Patricola L, Gomez JV, Plummer M, Franceschi S, Kato I, Canzian F. Use of whole genome amplification to rescue DNA from plasma samples. Biotechniques 2005;39:511–15. 22. Allion PD. Logistic regression using the SAS system: theory and application. Cary, NC: SAS Institute, 1999. Ch. 5,103–32. 23. Kato I, Vivas J, Plummer M, Lopez G, Peraza S, Castro D, Sanchez V, Cano E, Andrade O, Garcia R, Franceschi S, Oliver W, Mu~noz N. Environmental factors in Helicobacter pylori-related gastric precancerous lesions in Venezuela. Cancer Epidemiol Biomarkers Prev 2004;13:468–76. 24. Wigginton JE, Cutler DJ, Abecasis GR. A note on exact test of Hardy–Weinberg equilibrium. Am J Hum Genet 2005;76:887–93. 25. Lu W, Pan K, Zhang L, Lin D, Miao X, You W. Genetic polymorphisms of interleukin (IL)-1B, IL-1RN, IL-8, IL-10 and tumor necrosis factor a and risk of gastric cancer in Chinese population. Carcinogenesis 2005;26:631–6. 26. Ohyauchi M, Imatani A, Yonechi M, Asano N, Miura A, Iijima K, Koike T, Sekine H, Ohara S, Shimosegawa T. The polymorphism interleukin 8-251 A/T influences the susceptibility of Helicobacter pylori related gastric diseases in the Japanese population. Gut 2005;54:330–5. 27. Savage SA, Abnet CC, Mark SD, Qiao YL, Dong ZW, Dawsey SM, Taylor PR, Chanock SJ. Variants of the IL-8 and IL-8RB genes and risk for gastric cardia adenocarcinoma and esophageal squamous cell carcinoma. Cancer Epidemiol Biomarkers Prev 2004;13:2251–7. 28. Lee WP, Tai DI, Lan KH, Li AF, Hsu HC, Lin EJ, Lin YP, Sheu ML, Li CP, Chang FY, Chao Y, Yen SH et al. The -251T allele of the interleukin-8 promoter is associated with increased risk of gastric carcinoma featuring diffuse-type histopathology in Chinese population. Clin Cancer Res 2005;11:6431–41. 29. El-Omar EM, Carrington M, Choe W-H, McColl KEL, Bream JH, Young HA, Herrera J, Lissowska J, Yuan C-C, Rothman N, Lanyon G, Martin M et al. Interleukin-1 polymorphisms associated with increased risk of gastric cancer. Nature 2000;304:398–402. 30. Machado JC, Figueiredo C, Canedo P, Pharoah P, Carvalho R, Nabais S, Castro Alves C, Campos ML, van Doorn LJ, Caldas C, Seruca R, Carneiro F et al. A proinflammatory genetic profile increases the risk for chronic atrophic gastritis and gastric carcinoma. Gastroenterology 2003; 125:364–71. 31. Zeng ZR, Hu PJ, Hu S, Pang RP, Chen MH, Ng M, Sung JJ. Association of interleukin 1B gene polymorphism and gastric cancers in high and low prevalence regions in China. Gut 2003;52:1684–9. 32. El-Omar EM, Rabkin CS, Gammon MD, Vaughan TL, Risch HA, Schoenberg JB, Stanford JL, Mayne ST, Goedert J, Blot WJ, Fraumeni JF, Jr, Chow WH. Increased risk of noncardia gastric cancer associated with proinflammatory cytokine gene polymorphisms. Gastroenterology 2003;124:1193–201. 33. Garza-Gonzalez E, Bosques-Padilla FJ, El-Omar E, Hold G, TijerinaMenchaca R, Maldonado-Garza HJ, Perez-Perez GI. Role of the polymorphic IL-1B, IL-1RN and TNF-A genes in distal gastric cancer in Mexico. Int J Cancer 2005;114:237–41.
1671
34. Kato S, Onda M, Yamada S, Matsuda N, Tokunaga A, Matsukura N. Association of the interleukin-1b genetic polymorphism and gastric cancer risk in Japan. J Gastroenterol 2001;36:696–9. 35. Wu MS, Wu CY, Chen CJ, Lin MT, Shun CT, Lin JT. Interleukin-10 genotypes associated with the risk of gastric carcinoma in Taiwanese Chinese. Int J Cancer 2003;104:617–23. 36. Hwang IR, Hsu PI, Peterson LE, Gutierrez O, Kim JG, Graham DY, Yamaoka Y. Interleukin-6 genetic polymorphisms are not related to Helicobacter pylori-associated gastroduodenal diseases. Helicobacter 2003;8:142–8. 37. Hull J, Thomson A, Kwiatkowski D. Association of respiratory syncytial virus bronchiolitis with the interleukin 8 gene region in UK families. Thorax 2000;55:1023–7. 38. Crabtree JE, Covacci A, Farmery SM, Xiang Z, Tompkins DS, Perry S, Lindley IJ, Rappuoli R. Helicobacter pylori induced interleukin-8 expression in gastric epithelial cells is associated with CagA positive phenotype. J Clin Pathol 1995;48:41–5. 39. Figueiredo C, Machado JC, Pharoah P, Seruca R, Sousa S, Carvalho R, Capelinha AF, Quint W, Caldas C, van Doorn LJ, Carneiro F, Sobrinho-Simoes M. Helicobacter pylori and interleukin 1 genotyping: an opportunity to identify high-risk individuals for gastric carcinoma. J Natl Cancer Inst 2002;94:1680–7. 40. Rad R, Prinz C, Neu B, Neuhofer M, Zeitner M, Voland P, Becker I, Schepp W, Gerhard M. Synergistic effect of Helicobacter pylori virulence factors and interleukin-1 polymorphisms for the development of severe histological changes in the gastric mucosa. J Infect Dis 2003; 188:272–81. 41. Yamaoka Y, Kudo T, Lu H, Casola A, Brasier AR, Graham DY. Role of interferon-stimulated responsive element-like element in interleukin-8 promoter in Helicobacter pylori infection. Gastroenterology 2004;126:1030–43. 42. Brandt S, Kwok T, Hartig R, Konig W, Backert S. NF-jB activation and potentiation of proinflammatory responses by the Helicobacter pylori CagA protein. Proc Natl Acad Sci USA 2005;102:9300–5. 43. Backert S, Moese S, Selbach M, Brinkmann V, Meyer TF. Phosphorylation of tyrosine 972 of the Helicobacter pylori CagA protein is essential for induction of a scattering phenotype in gastric epithelial cells. Mol Microbiol 2001;42:631–44. 44. Higashi H, Nakaya A, Tsutsumi R, Yokoyama K, Fujii Y, Ishikawa S, Higuchi M, Takahashi A, Kurashima Y, Teishikata Y, Tanaka S, Azuma T et al. Helicobacter pylori CagA induces Ras-independent morphogenetic response through SHP-2 recruitment and activation. J Biol Chem 2004;279:17205–16. 45. Zhu YM, Webster SJ, Flower D, Woll PJ. Interleukin-8/CXCL8 is a growth factor for human lung cancer cell. Br J Cancer 2004;91:1970–6. 46. Huang JQ, Zheng GF, Sumanac K, Irvine EJ, Hunt RH. Meta-analysis of the relationship between cagA seropositivity and gastric cancer. Gastroenterology 2003;125:1636–44. 47. Konturek SJ, Starzynska T, Konturek PC, Karczewska E, Marlicz K, Lawniczak M, Jaroszewicz-Heigelman H, Bielanski W, Hartwich A, Ziemniak A, Hahn EG. Helicobacter pylori and CagA status, serum gastrin, interleukin-8 and gastric acid secretion in gastric cancer. Scand J Gastroenterol 2002;37:891–8. 48. Sun Y, Li J-Y, He J-S, Zhou L-X, Chen K. Tissue microarray analysis of multiple gene expression in intestinal metaplasia, dysplasia and carcinoma in the stomach. Histopathology 2005;46:505–14. 49. Kroeger KM, Carville KS, Abraham LJ. The -308 tumor necrosis factor-a promoter polymorphism effects transcription. Mol Immunol 1997;34:391–9. 50. Brull DJ, Montgomery HE, Sanders J, Dhamrait S, Luong L, Rumley A, Lowe GD, Humphries SE. Interleukin-6 gene -174g>c and -572g>c promoter polymorphisms are strong predictors of plasma interleukin-6 levels after coronary artery bypass surgery. Arterioscler Thromb Vasc Biol 2001;21:1458–63. 51. Rovin BH, Lu L, Saxena R. A novel polymorphism in the MCP-1 gene regulatory region that influences MCP-1 expression. Biochem Biophys Res Commun 1999;259:344–8.
