OBJECTIVE: To systematically understand the cellular and molecular mechanism of gastric cancer (GC) development and to discover early diagnosis and predictive biomarkers, which will be used for early diagnosis and novel treatment targets. METHODS: 70 mer 22 K-oligonucleotide microarrays and bioinformatic analysis were conducted to recognize gene expression profiles in GC and normal appearing tissue (NAT). The control group was collected from non-tumor patients including 20 specimen mixture as a common reference (CR) and 5 individuals as additional control. Our results showed that 837 different expression genes (DEGs) were identified in GC while 570 DEGs were in NATs by Bayesian analysis (P<0.001, Fold change>2.0) as compared respectively with CR. An interesting finding is that we identified 67 over-expressed genes in both GC and NAT tissues, and these gene expression alterations could not be detected by comparison of GC with NATs, which were normally used in routine experiment design. Most of these genes were involved in the control of cell proliferation, metabolism and differentiation. RESULTS: These differential expressed genes were confirmed at mRNA and protein levels in primary tumors using RT-PCR and immunohistochemistry (IHC). The results showed that three genes, EGR1, CYR61 and ADAMTS1 were over expressed in both GC and NATs at mRNA level. These results were consistent with oligo microarray data. Another interesting finding is that these three genes were also over-expressed in intestinal metaplasia (IM) and dysplasia (DYS), which indicated that these three genes might be potential biomakers for early detection of GC. CONCLUSION: Through the systematic analysis of gene expression profiles in GC tissues, NAT and CR normal tissues, we identified a group of genes over-expressed both in GC and precancerous lesions, which might be potential biomarkers for early GC diagnosis.
OBJECTIVE: To systematically understand the cellular and molecular mechanism of gastric cancer (GC) development and to discover early diagnosis and predictive biomarkers, which will be used for early diagnosis and novel treatment targets. METHODS: 70 mer 22 K-oligonucleotide microarrays and bioinformatic analysis were conducted to recognize gene expression profiles in GC and normal appearing tissue (NAT). The control group was collected from non-tumorpatients including 20 specimen mixture as a common reference (CR) and 5 individuals as additional control. Our results showed that 837 different expression genes (DEGs) were identified in GC while 570 DEGs were in NATs by Bayesian analysis (P<0.001, Fold change>2.0) as compared respectively with CR. An interesting finding is that we identified 67 over-expressed genes in both GC and NAT tissues, and these gene expression alterations could not be detected by comparison of GC with NATs, which were normally used in routine experiment design. Most of these genes were involved in the control of cell proliferation, metabolism and differentiation. RESULTS: These differential expressed genes were confirmed at mRNA and protein levels in primary tumors using RT-PCR and immunohistochemistry (IHC). The results showed that three genes, EGR1, CYR61 and ADAMTS1 were over expressed in both GC and NATs at mRNA level. These results were consistent with oligo microarray data. Another interesting finding is that these three genes were also over-expressed in intestinal metaplasia (IM) and dysplasia (DYS), which indicated that these three genes might be potential biomakers for early detection of GC. CONCLUSION: Through the systematic analysis of gene expression profiles in GC tissues, NAT and CR normal tissues, we identified a group of genes over-expressed both in GC and precancerous lesions, which might be potential biomarkers for early GC diagnosis.