One of the undisputed keys to the successful treatment of cancer is early diagnosis. This strongly argues for participation in routine screening programs. However, those that are in place suffer from significant drawbacks. A major thrust of applied cancer research in the coming years will be a search for improved diagnostic methodologies. This will rely on the identification of a novel set of early tumor markers that can be detected through non-invasive approaches. This application proposes to identify secreted and cell surface proteins that are overexpressed in early stage cancers of the colon and breast. We will begin by distilling from a highly complex cDNA library sequences that represent secreted and cell surface proteins (Secrets) using a powerful biological selection. By focussing exclusively on """"""""trafficked"""""""" proteins we greatly limit the set of genes that must be surveyed. Tumor and normal cells purified by laser-capture microdissection will be used for the preparation of specific probes for differential expression analysis on DNA microarrays. When coupled with sequence information, this will create a database of expression patterns for secreted and cell surface proteins in early stage lesions of the colon and breast. Those clones that show differential expression will be tested using RNA microarrays for tumor and tissue specificity. A set of potential markers will be chosen based upon these expression patterns. Both the DNA sequences and immunological probes for promising candidates will be passed to the Marker Validation Laboratories for further evaluation.
