Metastatic disease is the major cause of treatment failure and death from breast cancer. We propose to study the biology of human breast cancer metastases by finding the genes which are responsible for the metastatic phenotype in cellular model systems, validating their significance in human tumor samples. Gene discovery will be done using cDNA expression arrays containing a large fraction (greater than 25,000) of all the cellular genes, finding those which are differentially expressed between matched cell lines that differ only in their ability to metastasize.
Our aims are: 1) To establish a suitable model system for comparative analysis of breast cancer metastases. Breast cancer cell lines which are suppressed in their metastatic ability have been identified in the lab, but further characterization is needed. 2) To compare the gene expression pattern in these matched cell lines, to identify sets of genes which are both upregulated and downregulated in the cells with the metastatic phenotype. RNA-based probes prepared from each cell line and its metastatic derivative will by hybridized sequentially the cDNA filter arrays, to find those which are differentially expressed. 3) Evaluation of the selected genes in clinical samples for their correlation with metastatic ability. Gene expression will be measured in primary breast cancers stratified for their metastatic behavior, to test whether expression tracks with the predicted phenotype. The novelty of this approach is the use of cDNA arrays, which have not been previously applied to this problem; this large-scale survey may result in the identification of hundreds of genes, many with unknown function, and thus additional schemes will be required to prioritize those that are potentially most valuable. The new genes identified will have utility for diagnostic methods (as serum markers or diagnostic antibodies), as targets of therapy to prevent metastases, and to understand the biologic basis of metastatic disease in human breast cancer.
