Purpose: To identify the genes involved in the carcinogenesis and progression of colorectal cancer, we analyzed the gene-expression profiles of colorectal cancer cells from 12 tumors with corresponding noncancerous colonic epithelia using a cDNA microarray representing 4,08 genes.
Methods: We classified both samples and genes by using a two-way clustering analysis and identified genes that were differentially expressed in the cancerous and noncancerous tissues. Genes associated with lymph node metastasis were identified by means of the supervised learning technique.
Results: Differentially expressed genes (77 up-regulated and 45 down-regulated genes) were identified in more than 75 percent of the tumors. The functional categories of these genes belonged to signal transduction (19 percent), metabolism (17 percent), cell structure/motility (14 percent), cell cycle (13 percent), and gene protein expression (13 percent). The gene expression pattern of reverse transcriptase polymerase chain reaction (RT-PCR) results from randomly selected genes shows a pattern similar to that of cDNA microarray. Moreover, the gene expression patterns observed were similar to those reported previously, suggesting rare racial differences. Sixty genes possibly associated with lymph node metastasis in colorectal cancer were selected on the basis of clinicopathological data obtained by performing signal-to-noise calculations. "Leave-one-out" cross-validation testing correctly classified 10 of 12 patients (83.3 percent) as having colorectal cancer with lymph node metastasis vs. those without metastasis.
Conclusions: These results provide not only a new molecular basis for understanding the biologic properties of colorectal cancer, including lymph node metastasis, but also provide a resource for future development of therapeutic targets and diagnostic markers for colorectal cancer.