Combining serial analysis of gene expression and array technologies to identify genes differentially expressed in breast cancer

Several methods have been used recently to determine gene expression profil es of cell populations. Here we demonstrate the strength of combining two a pproaches, serial analysis of gene expression (SAGE) and DNA arrays, to hel p elucidate pathways in breast cancer progression by finding genes consiste ntly expressed at different levels in primary breast cancels, metastatic br east cancers, and normal mammary epithelial cells. SAGE profiles of 21PT an d 21MT, two well-characterized breast tumor cell lines, were compared with SAGE profiles of normal breast epithelial cells to identify differentially expressed genes. A subset of these candidates was then placed on an array a nd screened with clinical breast tumor samples to find genes and expressed sequence tags that are consistently expressed at different levels in diseas ed and normal tissues, In addition to finding the predicted overexpression of known breast cancer markers HER-2/neu and MUC-1, the powerful coupling o f SAGE and DNA arrays resulted in the identification of genes and potential pathways not implicated previously in breast cancer, Moreover, these techn iques also generated information about the differences and similarities of expression profiles in primary and metastatic breast tumors. Thus, combinin g SAGE and custom array technology allowed for the rapid identification and validation of the clinical relevance of many genes potentially involved in breast cancer progression. These differentially expressed genes may be use ful as tumor markers and prognostic indicators and may be suitable targets for various forms of therapeutic intervention.