ANALYSIS OF GENES ASSOCIATED WITHLYMPHATIC METASTASIS IN PANCREATIC CARCINOMA USING cDNA MICROARRAY

Objective: To identify new markers for prediction of lymph node metastasis. Methods: cDNA probes were prepared by labeling mRNA from samples of four pancreatic carcinoma tissues with Cy5-dUTP and mRNA from adjacent normal tissues with Cy3-dUTP respectively through reverse transcription. The mixed probes of each sample were then hybridized with 4,096 cDNA arrays (4,000 unique human cDNA sequences), and the fluorescent signals were scanned by ScanArray 3000 scanner (General Scanning, Inc.). The values of Cy5-dUTP and Cy3-dUTP on each spot were analyzed and calculated by ImaGene 3.0 software (BioDiscovery, Inc.). Genes that differentially expresses in each cancerous tissue were sought out according to the standard that the absolute value of natural logarithm of the ratio of Cy5 to Cy3 is greater than 0.69, i. e., more than 2 times change of gene expression, and the signal value of either Cy3 and Cy5 need to be greater than 600. Then, the genes differently expressed in cancer with and without lymphatic metastasis were screened out for further analysis. Results: Among 2 samples with lymphatic metastasis and 2 samples without metastasis, 56 genes, which accounted for 1.40% of genes on the microarray slides, exhibited differentially expression in cancerous tissues with lymphatic metastasis. There were 32 over-expressed genes including 11 having been registered in Genebank, and 24 under-expressed genes including 3 in Genebank. Conclusion: Microarray analysis may provide invaluable information to identify specific gene expression profile of lymphatic metastasis in pancreatic cancer.