| The Cancer Genome Anatomy Project (CGAP) database of the National Cancer Institute contains thousands of expressed sequences, both known and novel, derived from diverse sets of normal, precancerous, and tumor cDNA libraries. This offers the possibility of using this database as a rational starting point for bioinformatics-based cancer gene discovery. Using the Digital Differential Display tool of the CGAP database, a hypothesis-driven gene discovery approach was undertaken to analyze differential expression of various solid tumor types. Two hundred known genes and five hundred novel sequences were discovered to be differentially expressed, and a comprehensive database was established to facilitate identification of cancer diagnostic and therapeutic targets. To validate the use of bioinformatics in discovering genes with organ- and tumor-selectivity, novel ESTs predicted to be colon tumor-specific were analyzed further for expression specificity. Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) analysis using matched sets of colon normal- and tumor-derived cDNAs identified one EST to be specifically expressed in the majority of colon tumors and normal small intestine. Due to this apparent specificity, the gene was termed Colon Carcinoma Related Gene (CCRG). Based on protein sequence analysis, CCRG belongs to a novel class of secreted factors. Another gene identified in this study showed homology to Single Minded 2 gene (SIM2). Involvement between SIM2 and cancer has not yet been reported. Isoform-specific expression of SIM2 short-form (SIM2-s) was seen in colon, pancreas, and prostate carcinomas but not in most normal tissues. Using a large collection of paraffin sections from colon, pancreas, and prostate tumor and normal tissues, elevated protein expression was seen in tumors compared to normal tissue specimens, demonstrating the diagnostic potential of SIM2-s. Antisense inhibition of SIM2-s expression in colon and pancreatic cancer cell lines caused inhibition of gene expression, growth inhibition, and apoptosis. Administration of SIM2-s antisense in nude mice caused inhibition of colon tumor growth without pronounced gross toxicity. Using GeneChip® technology, a gene expression profile indicative of apoptosis was observed in the colon cancer model. CCRG and SIM2-s offer both a diagnostic and therapeutic potential in select cancers and validate the use of bioinformatics approaches in the gene discovery paradigm. |
