574 gastric carcinoma patients and 574 healthy controls were included. All subjects reside in Jiangsu Province of China, and are unrelated ethnic Han Chinese. Patients were consecutively recruited from July 2005 to July 2009, at the First Affiliated Hospital of Nanjing Medical University. All cases were those newly histologically diagnosed with gastric carcinoma without previous chemotherapy or radiotherapy, which was staged and classified according to the American Joint Committee on Cancer Staging Manual, and Lauren’s classification, when diagnosed [24, 25]. All healthy controls were recruited from individuals living in the same residential areas who attended routine medical examination at the same hospital without no abnormal findings at the examination and were age- (±5 years) and sex-matched to the cases. All subjects who had a history of tumors including gastric or other tumors or had undergone eradication treatment for H. pylori were excluded.
Venous blood (2 ml) was collected from each subject at the entry of the study for the genotyping and detection of H. pylori infection. Gastric carcinoma patients were followed-up every 3 months by telephone or patient’s reexamination to the hospital to update the progress of the disease.
This study has been approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University, and the written informed consent was obtained from all subjects, and the institutional ethics committees approving this research comply with acceptable international standards (the Treaty of Helsinki).
Detection of H. pylori infection
The indirect solid phase immunochromatographic assay was applied to detect the infection of H. pylori according to its IgG antibodies (Genelabs Diagnostics Pty Ltd, Singapore), which has been validated in Chinese population with an accuracy of 92.3% .
Genomic DNA was extracted by proteinase K digestion and followed by phenol–chloroform extraction and ethanol precipitation from a leukocyte pellet. Polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) was used to determine the genotype of MDM2 SNP309 . In brief, the primer sequences were 5P-CGCGGGAGTTCAGGGTAAAG-3P (forward) and 5P-AGCTGGAGACAAGTCAGGACTTAAC-3P (reverse), which generated the 237-bp fragment. The PCR product was then digested by MSPA1I (New England Biolabs, Bevely, MA, USA). The wild-type (SNP309T) allele produces a single 237-bp fragment and the variant (SNP309G) allele produced two fragments of 189- and 48 bp. Genotyping was performed without knowledge of the case or control status. The gel images were read independently by two research assistants. If a consensus was not reached on the tested genotypes, then the genotyping was repeated independently until a consensus was reached. In addition, to validate the RFLP method, 100 (50 from cases and 50 from controls) PCR products were selected randomly for direct sequencing with ABI 3700 sequencer, and the concurrence rate of these two methods was 99%.
Construction of reporter plasmids
Because the association between MDM2 SNP309 (T > G) and gastric carcinoma susceptibility, we then determined whether this polymorphism had an effect on gene expression in vitro. The MDM2 promoter-luciferase reporter plasmids containing either 551 T or 787 G sequence were prepared by amplifying the 237 -bp MDM2 promoter region by using primers with restriction sites. The primers were 5P-GACGCTAGCTTTGTCG-CGCAGTTTCCACCG-3P (forward) and 5P-CGCGAAGCTT CTTCTTG-CTCCATCTTTCC-3P (reverse), including the NheI and HindIII restriction sites. We confirm the amplified fragments by sequencing. The amplified fragments and pGL3-basic vector (Promega, Beijing, China) were cleaved by using the NheI and HindIII enzymes (TaKaRa Biotech Co., Dalian, China). After the fragments were cloned into the pGL3-basic vector, the vectors were sequenced to confirm that there were no wrong.
Human gastric cancer cell lines AGS and MKN45 and mouse embryo fibroblast cell NIH-3 T3 (Institute of Cellular Research, Chinese Academy of Science, Shanghai, China) were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS). All the cells were incubated in a stable environment with 5% CO2 at 37°C in a humidified incubator (Thermo, Forma Scientific, Inc., USA).
Transient transfection and luciferase assays
For transfection, NIH-3 T3, AGS and MKN45 cells were seeded onto 24-well plates (100,000 cells per well), and each well was cotransfected with 2.25 μg of the vector DNA with either SNP309T or SNP309G allele and internal control 10 ng pRL-SV40, in which a cDNA could encode the Renilla luciferase, and the pGL3-basic vector without any insert was applied as the negative control. After 24 h transfection, cells were lysed and luciferase activity was assayed by using the Dual-Luciferase Reporter Assay System (Promega) and normalized according to the Renilla luciferase activity. Each plasmid construct were done for independent triplicate.
Purification and utilization of LPS
LPS was extracted from H. pylori strains (26695) using the phenol-water extraction procedure previously described , and was quantitated using the purpald assay .
AGS cells were cotransfected with either MDM2 SNP309T or SNP309G luciferase reporters with Renilla luciferase reporter (as internal control) for 24 h, followed by incubation with two concentrations (0.1 and 1 μg/ml) of H. pylori LPS for 24 h. Cells were lysed and luciferase activity performed by using the Dual-Luciferase Reporter Assay System (Promega) and normalized according to the Renilla luciferase activity. Each plasmid construct were done for independent triplicate.
Hardy-Weinberg equilibrium of alleles was evaluated by chi-square test. Comparison of age between cases and controls was evaluated by the Mann–Whitney U test, and the difference in the distribution of genotypes between cases and controls was determined by chi-square test. The association between the MDM2 SNP309 and the risk for gastric carcinoma was estimated by odds ratios (OR) with the 95% confidence intervals (CI). Logistic regression was used to control for selected potential confounders (sex, age, and H. pylori infection) and to assess crude and adjusted OR and 95% CI. The difference in the levels of luciferase reporter gene expression between different constructs was examined by student’s test. Kaplan-Meier was used to construct the cumulative overall survival curves and the difference was tested by the log-rank test, and Cox’s proportional hazards model was used to perform the multivariate analysis to adjust for age, sex, TNM stage and H. pylori. All data analyses were done using SPSS software (version 11.0, Chicago, IL, USA). A P value of < 0.05 was considered statistically significant. Our sample size of 574 gastric carcinoma patients and 574 controls provided a 93% power to test an OR of 1.5 at the first error of 0.05 in a 2-sided test.
We identified related studies by searching the PubMed, Embase, Web of Science, China National Knowledge Infrastructure and Wanfang databases up to August 2012 with the search phrases of “murine double minute or MDM2” and “polymorphism(s) or SNP309” and “gastric or stomach” and “cancer or carcinoma or tumor or neoplasm”. Additional studies were checked by screening reference lists of studies and reviews. The following criteria were used for literature selection: (a) the case–control study of the association between MDM2 SNP309 polymorphism and gastric carcinoma susceptibility; (b) genotypic frequencies in both cases and controls available. Only the study of most recent and complete was used in this meta-analysis. Hence, we included 5 studies in our meta-analysis, containing 1935 cases and 2953 controls. The following information was collected from each publication: the first author’s name, year of publication, country, number of cases and controls, and MDM2 genotype information.
The odds ratios (ORs) with 95% confidence intervals (CIs) were pooled to assess the association between MDM2 SNP309 and gastric carcinoma risk. The fixed effect model (Mantel-Haenszel method) and the random effect model (the Dersimonian and Laird method) were used to pool data from different studies [29, 30]. If the heterogeneity between studies is absent, the fixed effect model was used; otherwise, the random effect model was applied. The Z test was used to determine the statistical significance of the OR. The chi-square-based Q test and I2 were used to assess the statistical heterogeneity between studies, heterogeneity was considered statistically significant when P ≤ 0.05 , and Ι2 was used to qualify variation in OR attributable to heterogeneity. The publication bias was estimated by the Egger’s test . P values < 0.05 are considered statistically significant. All analysis was done by using the Review Manage (v.4.2) and Statistical Analysis System software (v.9.1.3, SAS Institute, Cary, NC).