Human serum and urine are each estimated to contain over 5,000 different proteins and peptides, many in trace concentrations. Disease-associated changes have been described for most of the less than 5% which have been quantitatively studied, but very few of these are useful cancer markers. Thus there is both an urgent need for new markers, and a forest of potential markers to be explored. We proposed concurrent development and testing of technology for systematically and sequentially fractionating proteins from serum and urine from normal individuals and cancer patients to discover useful tumor markers. The key analytical tools are quantitative high resolution two-dimensional electrophoresis under denaturing conditions, mass spectroscopy for protein identification, rapid recycling immunosubtraction, preparative and analytical gel filtration, high resolution chromatography based on ion exchange and hydrophobicity, and isoelectric focusing and conventional gel electrophoresis. In combination these can resolve thousands of proteins over a wide dynamic range. Phase I will concentrate on the demonstration that new proteins and a small set of candidate markers can indeed be found using these techniques in a manual format, while in Phase II each method will be mechanized or automated to allow higher throughput cancer marker discovery. PROPOSED COMMERCIAL APPLICATION: Identification of proteins or protein fragments unique to cancer would lead to the development of new clinical tests and to methods for imaging small tumor masses. If successful, this would revolutionize the diagnosis and treatment of cancer.
