The major focus of the proposed project is the characterization and subsequent application of potential biomarkers of prevalent epithelial cancers in humans: the prostate cancer-specific carbohydrate antigen, F77, and for breast cancer, the more broadly-expressed epithelial cancer-associated carbohydrate antigen, AE3, both recognized by monoclonal antibodies. The project brings together a team of scientists with hugely complementary expertise and technical capabilities in carbohydrates, immunobiology, cancer cell biology, oncology, proteomics, and computational biology, as well as unique and well-annotated clinical specimens. We intend to adopt two main approaches to the structural characterization of F77 and AE3 antigens. The first is carbohydrate microarray analysis coupled with mass spectrometry using glycan arrays generated from the prostate cancer cell line, PC-3, and from AE3 antigen-positive epithelial mucins. The second is cell transfection of specific glycosyltransferases and other glyco-modifying enzymes such as sulfotransferases, an approach that has recently met with considerable success. Once characterized, the F77 and AE3 antigenic glycan sequences will be included as key probes in a carbohydrate microarray platform that we have established that presently includes more than 700 glycan sequences of glycoproteins and glycolipids. This will open the way to analysis of cancer patient sera for the presence of autoantibodies to F77 and AE3 and any other glycan biomarkers and to the design of new analysis procedures for the detection of aberrant glyco-antigenemia; one such approach will be by immune-proteomics. The expression of F77 and AE3 antigens will also be evaluated as tissue-based prognostic markers to identify aggressive primary prostate cancer. Finally, the mechanistic role of the F77 antigen in metastatic capabilities of cancer cells will be investigated to link the presence of the biomarker to the biology of the tumor. For prostate and breast cancers, existing screening methods for detection and characterization are variously unsatisfactory and unreliable. The application of biomarkers to detect prostate and breast cancer and to provide information about the prognosis of individual patients would be extremely useful and provide a very desirable alternatives to the current low sensitivity and/or specificity screening methods.