Comprehensive gene expression analysis as a tool for hunting cancer genes

Abstract

Nearly three million messenger RNA transcripts have been catalogued by tissue origin to create one of the largest free databases for cancer gene expression analysis. Seventy different cell populations were studied by serial analysis of gene expression to provide precise gene expression levels in brain, breast, colon, ovarian and prostate cancers and corresponding normal tissues. Additionally, several model systems have been analyzed, providing insight into the genes transcriptionally regulated by fundamental oncogenic processes, such as angiogenesis, oncogene amplification and telomere maintenance. These data, digital northern displays and comparison tools for virtual experiments are made available by the Cancer Genome Anatomy Project at http://www.ncbi.nlm.nih.gov/SAGE/. Despite the many challenges inherent in applying large-scale gene expression data, we focus on two examples in which application of project data has yielded progress. First, mining this data readily provides candidate tumor markers or tumor antigens for the major tumors represented in the database. Using fluorescent real-time polymerase chain reaction and western blotting we showed that the mined tumor markers are expressed in an independent set of tumors and not in other normal tissues. Large-scale gene expression analysis may therefore provide a means for identifying therapeutic or prognostic targets. Second, analysis of a cell culture model for telomere maintenance (an important function of transformed cells) has provided insight into genes not previously implicated in this process. This example helps demonstrate that functional genomics on the basis of comprehensive gene expression analysis is a useful tool for hunting cancer-related genes.