Molecular characterization of gastrointestinal cancer through serial analysis of gene expression and GICHIP and its diagnostic implication

点击翻译按钮获取中文摘要

Serial analysis of gene expression (SAGE) is a powerful technique to allow genome-wide analysis of gene expression in a quantitative manner without a prior knowledge of the sequence of the gene. We performed on 5 samples of gastric cancer with different histology and stages, and created the largest SAGE libraries in the world, containing a total of 137,706 expressed tags including unique 38,903 tags (GEO accession number GSE 545). By comparing gene expression profiles between gastric cancer and normal gastric mucosa in combination with quantitative RT-PCR,APOC1,CEACAM6, and YF13H12 were found to be frequently overexpressed in gastric cancer, and FUS,CDH17,COL1A1,COL1A2, and APOE were associated with tumor progression. We picked up candidate genes that were detected in our gastric cancer libraries but not in the libraries from various normal tissues, especially important or crucial organs. The expression of these genes was examined in gastric cancers and normal human tissues by quantitative RT-PCR and found that at least 8 genes were specifically expressed in gastric cancers. We have developed custom-made oligo-DNA microarray, called Ex-STOMACHIP, with 475 cDNAs including specific genes identified by our SAGE analysis, known genes related to development and progression of cancer, and marker genes for chemosensitivity. By examining gene expression profiles of 20 gastric cancers, we identified gene clusters which can differentiate tumor stage and histology. Regarding the relation between genetic polymorphism of tumor-related genes and gastric cancer, we have performed case-control and case-case studied using about 500 subjects, and found a significant association with SNPs of HER2 and MMP-1. SNP (G/A) in 5'UTR of the CLDN18, newly identified as down-regulated gene in gastric cancer, was found to affect gastric cancer risk (OR-5.17,95%CI:2.01-13.3).