SEARCH

SEARCH BY CITATION

Keywords:

  • gastric cancer;
  • macrophage-derived chemokine;
  • CCL22;
  • genetic polymorphism;
  • Helicobacter pylori infection

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

BACKGROUND:

Gastric carcinoma is widely considered to be related to Helicobacter pylori infection, and the chemokine (C-C motif) ligand 22 (CCL22) plays an important role in suppressing immune responses against H. pylori and tumor cells. In this study, the authors examined the association between single nucleotide polymorphisms (SNPs) in the CCL22 gene and the risk of gastric carcinoma.

METHODS:

Information on SNPs in the CCL22 coding region was obtained from the HapMap Project database. Genotypes were determined in a case-control cohort that consisted of 1001 patients with gastric carcinoma and 1066 controls, and odds ratios (ORs) and 95% confidence intervals (95% CIs) were computed by using a logistic regression model. Serum H. pylori antibody levels were measured by using an enzyme-linked immunosorbent assay.

RESULTS:

The 16C[RIGHTWARDS ARROW]A SNP (reference SNP no. 4359426) in exon 1 of the CCL22 gene, which causes a 2 aspartate (2Asp) to 2 alanine (2Ala) substitution in the CCL22 protein, was associated with a significantly increased risk of gastric carcinoma. Individuals who were homozygous for the Ala/Ala genotype had an OR of 2.27 (95% CI, 1.28-4.02) compared with individuals who had the Asp/Asp genotype. Stratification analysis indicated that the association was more pronounced among men (OR, 2.64; 95% CI, 1.29-5.41) and among younger individuals (OR, 2.85; 95% CI, 1.36-5.96) compared with women and older individuals. Moreover, a multiplicative joint effect between the CCL22 SNP and H. pylori infection that intensified the risk was observed (OR for the presence of both Ala/Ala genotype and H. pylori infection, 18.37; 95% CI, 2.30-146.67).

CONCLUSIONS:

The results from this study suggested that the CCL22 polymorphism is associated with an increase risk of developing H. pylori infection-related gastric carcinoma. Cancer 2009. © 2009 American Cancer Society.

Gastric cancer is 1 of the most common cancers and is the second leading cause of cancer death worldwide.1 It has been established that Helicobacter pylori infection plays an important role in the initiation and development of gastric cancer. Chronic gastritis caused by H. pylori infection tends to develop into atrophy gastritis and, consequently, gastric cancer.2 Although the exact mechanism underlying H. pylori-induced gastric carcinogenesis remains to be elucidated, it is believed that inflammatory cells, such as macrophages, dendritic cells, monocytes, and neutrophils, are involved in the process by producing and releasing various inflammatory mediators, such as cytokines, chemokines, and matrix metalloproteinases.3 However, the clinical outcome of H. pylori infection is highly variable, and only a small fraction of infected individuals develops gastric cancer,4 suggesting that susceptibility to gastric cancer may be influenced by both microbial factors and host factors. The identification of such factors would provide better insight into the mechanisms of gastric carcinogenesis and, thus, would aid in the early detection and prevention of the cancer.

Accumulating evidence has demonstrated that genetic polymorphisms in genes that encode inflammatory mediators and the innate immunity system, the first-line defense against microbial invasion, may have a significant influence on the risk of developing gastric cancer related to H. pylori infection. For example, studies in Caucasian populations reported that genetic polymorphisms in interleukin 1B (IL-1B), which encodes IL-1β, and IL-1RN, which encodes the IL-1 receptor antagonist, that may enhance the production of IL-1β, were associated with hypochlorhydria induced by H. pylori infection and an increased risk of gastric cancer,5, 6 whereas this association was not confirmed in Asian populations.7, 8 In a Chinese population, it was reported that genetic polymorphisms in IL-8, IL-10, and tumor necrosis–α (TNF-α) may play important roles in the development of gastric cancer.9 Recently, we demonstrated that a functional genetic polymorphism in the promoter of CD14, a pattern-recognition receptor that plays a central role in innate immunity and directs the adaptive immune responses, also is associated with increased susceptibility to H. pylori infection-related gastric carcinoma.10

Chemokine (C-C motif) ligand 22 (CCL22 or macrophage-derived chemokine [MDC]) is a chemokine that is produced mainly by macrophages and dendritic cells and is also secreted by many types of tumor cells.11, 12 It is well known that CCL22 has important effect on suppressing immune responses against microbial infections and tumor cells by its ability to recruit T-regulator cells and T-helper type 2 cells, thereby promoting microbial persistence and tumor development.13, 14 For instance, it was reported previously that the expression of CCL22 in H. pylori-infected mucosa was significantly greater that its expression in uninfected healthy mucosa,15 and the elevated expression of CCL22 was correlated with an increased population of T-regulator cells in early gastric cancer.16 These results indicate that CCL22 may play a role in the development of H. pylori-related gastric cancer. In view of these findings, we hypothesized that genetic polymorphisms in the CCL22 gene that affect CCL22 function may be associated with susceptibility to H. pylori-related gastric cancer. Here, we report a significant contribution of a single nucleotide polymorphism (SNP) in the CCL22 coding region to the risk of gastric cancer in a case-control study among Chinese. We also demonstrate an interaction between the CCL22 SNP and H. pylori infection in intensifying risk of the cancer.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

Individuals for Case-control Analysis

This study included 1001 patients with gastric cancer and 1066 cancer-free population controls, all of whom were unrelated ethnic Han Chinese. Because this is an ongoing molecular epidemiologic study of gastric cancer, a report on part of this case-control panel (both n = 470) was published previously.10 However, DNA samples from 18 patients and 8 controls no longer were available. In the current study, we added 549 more cases and 604 more controls to extend the sample size. Patients were recruited consecutively between January 2000 and July 2006 at the Cancer Hospital, Chinese Academy of Medical Sciences (Beijing), and all had histopathologically verified primary gastric adenocarcinoma. All eligible patients who were diagnosed at the hospital during the study period were recruited, and the response rate was 92%. Patients were from the city of Beijing and its surrounding regions, and there were no age, sex, or disease stage restrictions. The exclusion criteria included other primary cancer and previous chemotherapy or radiotherapy. The pathologic stage of gastric carcinoma was evaluated according to the International Union Against Cancer tumor-lymph node-metastasis (TNM) classification at diagnosis on the basis of postoperative pathologic examination of specimens. Population controls were recruited from a community nutritional survey that was conducted in the same region during the period of patient recruitment. The controls were selected randomly from a database that consisted of 2500 individuals based on a physical examination. The selection criteria included no history of other primary cancer and frequency matching to the cases on sex and age (±5 years). At recruitment, informed consent was obtained from each participant. This study was approved by the institutional review board of the Chinese Academy of Medical Sciences Cancer Institute.

Single Nucleotide Polymorphism Screening and Genotyping

SNPs in the CCL22 coding region were searched in the National Center for Biotechnology Information (NCBI) gene and SNP databases (available at: http://www.ncbi.nlm.nih.gov/SNP/ accessed on October 10, 2007) and the HapMap Project database (release 21a). There are 4 coding region SNPs (cSNPs) registered in the database, ie, 16C/A (alanine 2 aspartate [Ala2Asp], reference SNP number [rs] 4359426) in exon 1, 1639T/C (tyrosine 27 tyrosine [Tyr27Tyr], rs34057704) in exon 2, 1676C/T (histidine 40 tyrosine [His40Tyr], rs41398344) in exon 2, and 4732T/C (Ala77Ala, rs2229161) in exon 3. Among these cSNPs, only the Ala2Asp SNP has a minor allelic frequency >1% in the Chinese population (HapMap data release 21a) and, thus, was selected for genotyping in the current study.

Genomic DNA was extracted from peripheral blood lymphocytes in all participants. Genotypes of the Ala2Asp SNP were determined by polymerase chain reaction (PCR)-based restriction fragment length polymorphism. The PCR primer pairs that were used to amplify the DNA sequence containing Ala2Asp were 5′-TGGGAGG TAGTTCTTCTTTTGA-3′ and 5′-CCACAGCAAG GAGGACGA-3′. PCR was performed with a 10-μL reaction mixture containing 10 ng of template DNA, 0.2 μM of each primer, 0.3 mM of each deoxynucleotide triphosphate, 2.0 mM of MgCl2, and 0.5 U of Taq DNA polymerase (TaKaRa, Dalian, China) with 1 × reaction buffer. The PCR profile consisted of an initial melting step of 2 minutes at 94°C; followed by 35 cycles of 30 seconds at 94°C, 30 seconds at 55°C, and 30 seconds at 72°C; and a final elongation step for 7 minutes at 72°C. The PCR product was digested with 2.0 U of restriction endonuclease MboI (Fermentas, Vilnius, Lithuania) overnight to identify 3 different genotypes. Genotyping was performed without knowledge of the case/control status of the participants. A 10% random sample of cases and controls was tested twice by different individuals, and all results were 100% concordant.

Serologic Determination of Anti-H. pylori Antibody

Determination of serum anti-H. pylori antibody status was performed in the laboratory of Dr. Weicheng You (Beijing Institute for Cancer Research, Beijing, China) as described previously.17 In brief, H. pylori stains that were cultured in gastric biopsies from 5 patients were used to provide a local antigen preparation for serology tests. Serum levels of anti-H. pylori immunoglobulin G (IgG) and IgA were measured separately in duplicate using an enzyme-linked immunosorbent assay method. Quality-control samples were assayed using the same method in the laboratory of Dr. M. J. Blaser laboratory at Vanderbilt University (Nashville, Tenn). An individual was determined to be positive for H. pylori infection when the mean absorbance for either IgG or IgA was >1.0, a cutoff value that was determined from the examination of a group of H. pylori-negative individuals and reference sera.

Statistical Analysis

Chi-square tests were used to examine the differences in demographic variables and genotype distributions between patients and controls. The associations between genotypes and risk of gastric cancer were estimated by odds ratios (ORs) and their 95% confidence intervals (95% CIs), which were computed by using logistic regression models. All ORs were adjusted for age, sex, and CD14 −260C[RIGHTWARDS ARROW]T genotype, as appropriate. We tested the null hypothesis of a multiplicative gene-H. pylori infection interaction. A greater than multiplicative joint effect was suggested when OR11 > OR10 × OR01.18 Stratification analyses also were done with age, sex, tumor differentiation, invasion, and metastasis. All analyses were performed with the Statistical Analysis System software package (version 9.0; SAS Institute, Cary, NC).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

The baseline characteristics of patients and controls are summarized in Table 1. The distributions of sex and age were fairly identical in cases and controls, indicating that the matching of controls to cases was adequate. In this case-control panel, only 452 patients and 462 controls had H. pylori infection data. Among the patients, 73.7% of individuals were anti-H. pylori antibody positive, whereas this value was 56.5% among controls (P < 1.0 × 10−7). Individuals with H. pylori infection had an OR of 2.19 (95% CI, 1.65-2.90) of developing gastric cancer. Of 1001 patients, 958 (95.7%) underwent gastrectomy and had detailed tumor TNM classification data, whereas the remaining patients (4.3%) had not undergone gastrectomy and had unknown tumor TNM classification.

Table 1. Baseline Clinical Characteristics of Patients and Controls
 Patients (n=1001)Controls (n=1066) 
VariableNo.%No.%P*
  • *

    Two-sided chi-square test.

  • According to the 1997 International Union Against Cancer TNM classification for gastric carcinoma.

Sex    .190
 Men69369.276671.9 
 Women30830.830028.1 
Age, y    .062
 ≤6058958.867062.9 
 ≥6141241.239637.1 
Helicobacter pylori infection     
 Unknown54954.860456.7 
 Known45245.246243.3<.0001
  Positive33373.726156.5 
  Negative11926.320143.5 
Tumor stage     
 IA10310.3   
 IB747.4   
 II12912.9   
 IIIA20520.5   
 IIIB14714.7   
 IV30029.9   
 Unknown434.3   
Differentiation     
 High323.2   
 Intermediate15615.6   
 Low73573.4   
 Unknown787.8   

Genotyping results (Table 2) indicated that the frequencies of the 3 CCL22 genotypes, ie, Asp/Asp, Ala/Asp, and Ala/Ala, were 71.5%, 26.8%, and 1.7%, respectively, among controls, consistent with what was expected from Hardy-Weinberg equilibrium (P = .141). The corresponding genotype frequencies among patients were 70.5%, 25.7%, and 3.8%, respectively, which differed significantly from the values among controls (chi-square test, 8.79; P = .012; 2 degrees of freedom), and the Ala/Ala genotype was more prevalent in patients than in controls. Multivariate logistic regression analysis demonstrated that individuals with the CCL22 Ala/Ala genotype had a >2-fold increased risk for developing gastric carcinoma compared with individuals with the CCL22 Asp/Asp genotype (adjusted OR, 2.27; 95% CI, 1.28-4.02; P = .005). However, the heterozygous Asp/Ala genotype was not associated with increased risk (adjusted OR, 0.97; 95% CI, 0.80-1.18 [P = .778]), suggesting a recessive effect of this SNP on gastric cancer susceptibility.

Table 2. Genotype Frequencies of the Chemokine (C-C Motif) Ligand 22 (CCL22) Aspartate-2-Alanine Polymorphism Among Patients and Controls and Their Association With the Risk of Gastric Cancer
 Patients, n=1001Controls, n=1066  
GenotypeNo.%No.%OR (95% CI)*P
  • OR indicates odds ratio; 95% CI, 95% confidence interval; Asp; asparagine; Ala, alanine.

  • *

    Data were calculated by logistic regression analysis and were adjusted for age and sex.

Asp/Asp70670.576271.51.00 (Reference) 
Asp/Ala25725.728626.80.97 (0.80-1.18).778
Ala/Ala383.8181.72.27 (1.28-4.02).005

Stratification analyses of the CCL22 Asp/Ala polymorphism and gastric cancer risk demonstrated that the Ala/Ala genotype was associated significantly with an increased risk of cancer only among men (adjusted OR, 2.64; 95% CI, 1.29-5.41 [P = .008]) and among individuals aged ≤60 years (adjusted OR, 2.85; 95% CI, 1.36-5.96 [P = .003]), but not among women (adjusted OR, 1.85; 95% CI, 0.73-4.71 [P = .20]) or among individuals aged ≥60 years (adjusted OR, 1.47; 95% CI, 0.60-3.65 [P = .40]). However, the Ala2Asp genotype was associated neither with tumor differentiation nor with TNM status (data not shown).

Because H. pylori infection is an established etiologic factor in gastric cancer and CCL22 may play a role in promoting microbial persistence, we examined the joint effect of this CCL22 polymorphism and H. pylori infection on the risk of developing gastric cancer. We previously demonstrated that the CD14 −260C[RIGHTWARDS ARROW]T polymorphism also is associated with increased risk of gastric cancer in this study population.10 Thus, we included the CD14 −260 genotypes in our logistic regression model for adjustment. It appeared that H. pylori infection and CCL22 polymorphism had significant interaction in intensifying the risk (Table 3). Individuals who carried either the Ala/Ala genotype or H. pylori infection alone had an OR of 4.38 (95% CI, 1.07-17.87) or 1.96 (95% CI, 1.40-2.74), respectively; however, the OR was increased to 18.37 (95% CI, 2.30-146.67) in individuals who had both H. pylori infection and the variant Ala/Ala genotype, which is greater than the product of the OR for the Ala/Ala genotype and the OR for H. pylori infection alone (18.37 > 4.38 × 1.96). Similar results also were apparent for the heterozygous Asp/Ala genotype. Together, these results suggest that there is a joint effect between the CCL22 polymorphism and H. pylori infection.

Table 3. Risk of Gastric Cancer Associated With the Chemokine (C-C Motif) Ligand 22 (CCL22) Genotypes by Helicobacter pylori Infection Status
  Patients, n=452Controls, n=462  
GenotypeH. pylori InfectionNo.%No.%OR (95% CI)*P
  • OR indicates odds ratio; 95% CI, 95% confidence interval; Asp; asparagine; Ala, alanine.

  • *

    Data were calculated by logistic regression analysis and were adjusted for age, sex, and CD14 −260C[RIGHTWARDS ARROW]T genotypes.

Asp/AspNegative8819.5143311.00 (Reference) 
Ala/AspNegative245.35511.90.69 (0.39-1.23).208
Ala/AlaNegative71.530.74.38 (1.07-17.87).040
Asp/AspPositive23652.219742.61.96 (1.40-2.74)<.0001
Ala/AspPositive8518.86313.62.32 (1.49-3.59).0002
Ala/AlaPositive122.710.218.37 (2.30-146.67).006

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

Previous studies have indicated that genetic polymorphisms in the genes involved in inflammation and innate immunity may confer susceptibility to gastric cancer. In the current study, we observed that the ASP[RIGHTWARDS ARROW]Ala polymorphism in CCL22 was associated with a >2-fold increased risk for developing gastric cancer in a Chinese population. This genetic polymorphism had a significant joint effect with H. pylori infection on intensifying the risk of the cancer, suggesting that CCL22 may play a role in the development of H. pylori-related gastric cancer. To the best of our knowledge, this is the first study investigating the association between genetic polymorphism in CCL22 and the risk of developing gastric cancer.

The mechanism underlying the association between the CCL22 Asp2Ala polymorphism and the risk of gastric cancer is not immediately evident. A plausible explanation is that the Asp-to-Ala change may alter the processing efficiency of nascent CCL22 protein in endoplasmic reticulum or its releasing kinetics and, thus, may increase the expression level of this chemokine, because alanine differs greatly from aspartate in a variety of physicochemical properties.19, 20 Because the Asp-to-Ala change occurs at the second amino acid of the N-terminal of the signal peptide of pro-CCL22, it may have an important effect on nascent protein processing. Signal peptides generally comprise 3 regions: a positively charged N-terminal region (n-region), a central hydrophobic region (h-region), and a C-terminal region (c-region) that contains the cleavage site for signal peptidases on the endoplasmic reticulum membrane.21 In their co-translational translocation and proteolysis process in endoplasmic reticulum, secretory proteins such as chemokines generally adopt the Ncyt-Cexo orientation; that is, the nascent polypeptides expose the C-terminal cleavage site to signal peptidase with the N-terminal pointing toward the cytosol.22 The amount of the positive charge in the n-region is a prominent determinant for that process, because the sufficiently positively charged N-terminal is required for correct orientation.23 It is presumed that the Asp-to-Ala change in the CCL22 molecule adds net N-terminal–positive charges, thereby contributing to the accelerated processing of nascent pro-CCL22-2Ala compared with pro-CCL22-2Asp and resulting in increased expression levels of this variant isoform. Although direct evidence supporting this assumption is not available to date, other observations have demonstrated that mutations in signal peptides alter the expression levels of secretory proteins.24

It is well known that macrophages, dendritic cells, and various tumor cells are able to secrete CCL22, which recruits T-regulatory cells and T-helper type 2 cells to the inflammatory sites and/or tumor tissues,25-28 resulting in the suppression of immune responses against infection and/or tumor.29 It has been demonstrated that naturally occurring T-regulatory cells in the stomach mucosa are associated with increased H. pylori colonization30 and that the depletion of T-regulatory cells leads to reduced H. pylori colonization in infected mice.31 T-regulatory cells also may be capable of suppressing antitumor immunity by inhibiting effector T cells, thereby promoting cancer development and progression.26, 32 Specifically, it was reported that the CCL22 levels within gastric tumor tissues were correlated with an increased population of T-regulatory cells, especially in early stages of the cancer,16 and that increased T-regulatory cells were associated with a worse prognosis and increased mortality from gastric cancer.33, 34 Conversely, it also is believed that T-helper type 2 cells are involved in the suppression of host-immune responses against infection and tumor.35 In this context, it would be expected that the Asp2Ala polymorphism, which presumably is associated with the increased production of CCL22, is a susceptibility factor for the development of H. pylori-related gastric carcinoma.

In the current study, we observed that there was a joint effect between the CCL22 Asp2Ala polymorphism and H. pylori infection on the risk of gastric cancer. This finding was expected and is biologically plausible. H. pylori is a well established etiologic factor in gastric cancer4; and, in the current study, we also observed that H. pylori infection significantly elevated the risk of developing gastric carcinoma in our study population, although only some of the individuals who had detailed H. pylori-infection data were included in our analysis. Because CCL22 plays an important role in suppressing host-immune responses against infection and tumor, we might have expected that individuals who carried the variant CCL22 2Ala/Ala genotype and had H. pylori infection would have the greatest risk for developing gastric cancer. By using stratification analysis, we also observed that the risk of gastric cancer related to the variant CCL22 genotypes was pronounced in younger individuals (aged ≤60 years) and men. These findings are in line with the observation that gastric carcinoma occurs dominantly among men rather than women36 and with the concept that genetic susceptibility often is associated with an early age of disease onset.

The current study may have some limitations because of study design. The patients for the case group in this study were recruited from the hospital, and the control group was recruited from the community; thus, selection bias and/or systematic error may have occurred. It would be important to confirm these results in a population-based cohort study. In addition, among the study participants, only 452 cases and 462 controls had H. pylori-infection data, which may limit the statistical power of our analysis of the gene-environment interaction.

In conclusion, we have demonstrated an association between Ala2Asp polymorphism in the CCL22 coding region and an increased risk of gastric cancer alone and in interaction with H. pylori infection. These results further support H. pylori infection as a major risk factor for gastric carcinoma and highlight a possible role of genetic variations in the host-immune response against infection and tumor in determining phenotypic variations of gastric cancer.

Conflict of Interest Disclosures

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

Supported by the National Natural Science Foundation (grant 30530710) and by the State Key Basic Research Program (grant 2004CB518701).

References

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References