| Background:Gastric cancer (GC) causes more than700,000deaths each year all over the world,and the amount in developing countries such as China is nearly3-fold as in developedcountries, most likely because of the combination of differences of the environmental andgenetic factors. What’s more, the cure of GC is very difficult, mainly because of mostpatients’ advanced diseases. The survival rates of GC mainly depend on the earlydiagnosis. Therefore, it is very nessecery to improve the rates of early diagnosis andtreatment for GC. Looking for specific biomarkers of GC has become our concerns.Current evidence showed that genetic factors, especially germ line single nucleotidepolymorphisms (SNP) are crucial for GC carcinogenesis. To identify SNPs may contributeto the understanding of GC carcinogenesis mechanisms and help to personalizedpreventing and therapy. It is very necessary to specifically identify SNPs associated withGC. Genetic variations in prostate stem cell antigen (PSCA) have been reported to beassociated with GC risk in genome-wide association studies (GWAS). But the results oflater studies were conflicting. So the further clarification and characterization of these variations were needed, especially in specific population and specific area. Furthermore,the relationship between genetic variation and gene expression and function needing to beexplored.Prupose:To explore the SNPs of PSCA associated with GC risk among Chinese northwesternHan population, establishing the early diagnosis standard and premonitory urge.Methods:We hypothesized and tested the association between candidate SNPs in PSCA and therisk of GC in a northwestern Chinese Han population through a comprehensiveassociation analysis in patients’ peripheral blood and tissue samples respectively. Thepositive genotypes/haplotypes of PSCA and their association with gene transcription andexpression was analyzed. Then MDR was conducted to analyze locus-locus interaction.Results:By MassARRAY platform, peripheral blood samples from476GC cases and481controls were screened in a panel of4tSNPs in PSCA. We identified two significanttSNPs (rs2976392and rs2294008) associated with the risk of GC. The rs2294008(C>T)was significantly associated with an increased GC risk in dominant model (OR,1.30;95%CI,1.01-1.68; P=0.042). In addition, the rs2976392(G>A) showed association withelevated risk of GC in two genetic models (For GA: OR,1.37;95%CI,1.04-1.81; P=0.024. For GA/AA OR,1.36;95%CI,1.04-1.77; P=0.023). Then th LD was analyzedfor the4tSNPs, the results showed that three tSNPs rs229400, rs2976392and rs10216533were in LD. A protective haplotype CGG with frequency of64%(OR=0.78;95%CI,0.64-0.96; P=0.020) and a risk haplotype TAG with frequency of4%(OR=12.28;95%CI,3.75-40.27; P <0.0001) in our population were identified. Further associationstudy was conducted in various kinds of surgical resected tumor and paired adjacentnoncancerous tissues. The results showed that there was no difference of genetic variationdistribution frequencies between GC and paired adjacent noncancerous tissues. While thedifference was significant between GC and three non-digestive tract cancer tissues, suggesting that genetic variations of PSCA were polymorphisms rather than somaticmutation, and they were highly related to with GC risk, other than some non-digestivesystem cancers.Haplotype-based real-time PCR and IHC results revealed that the transcription andprotein expression of PSCA were both decreased in GC tissues compared with theiradjacent noncancerous tissues. The transcription and protein in risk haplotype wassignificantly higher than those in protective haplotype both in GC and adjacentnoncancerous tissues.Tissue microarray which was carried out with anti-PSCA antibody showedsimilarity and/or disparity of expression distribution of PSCA in digestive organ andnon-digestive organs, between GC and adjacent noncancerous tissues, which reminded usthat, PSCA might have complex functions and might play different roles in digestivesystem cancers and non-digestive system cancers.MDR analysis identified3models of locus-locus interaction includingPSCA_rs10216533-XRCC1_rs25487[OR=3.566,95%CI (2.614,4.863)], MTHFR_rs1801131-PSCA_rs2976392-CYP17A1_rs6163-YP19A1_rs4646-CYP19A1_rs1902586-MMP2_s243865[OR=7.257,95%CI (5.337,9.870)], and IL1B_rs16944-PSCA_rs10216533-CYP17A1_rs6163-NQO1_rs1800566-ENOSF1_rs2298581-XRCC1_rs25487[OR=15.837,95%CI (11.252,22.289)].Conclusion:Our comprehensive analysis confirmed that different degrees of variations of PSCAexisted between blood and tissue of GC patients to be associated with GC risk innorthwestern Chinese Han population. In addition, the genotypes/haplotypes mayinfluence the expression of PSCA and differentiation of GC tissues. The combination oflogistic regression and MDR model may be helpful to reveal the risks by one ormultiple-factor method in early diagnosis standard and premonitory urge in gastric cancer. |
PDF Full Text Request