Ovarian carcinoma is the most lethal gynecological cancer in the United States. Most women are diagnosed with advanced stage disease that is infrequently cured with existing therapies. Consequently, we propose to use a unique non-gel based protein separation method to search for protein biomarkers of ovarian carcinoma that may be useful in diagnosis, assessment of prognosis, and as potential drug targets. This new and efficient method will involve a two-dimensional (2-D) liquid separation to detect and compare large numbers of proteins in lysates from established ovarian carcinoma cell lines (n=35) and fresh frozen ovarian tumors (n=60). The 2-D liquid separation is based upon pl in the first dimension and hydrophobicity in the second dimension and will produce a 2-D map of the protein content of each sample under investigation. We will be able to identify marker proteins that are differentially expressed between samples either by level of expression or by structural changes due to posttranslational processing. The highly reproducible protein maps will be used to identify changes in protein expression between samples using differential image software. We will compare 2-D protein maps from normal ovarian surface epithelium primary cultures to ovarian carcinoma derived cell lines to identify marker proteins of ovarian carcinoma. In addition, using 2-D maps generated from ovarian carcinoma tissue, we will compare high stage/grade and low stage grade tumors to identify markers associated with poor prognosis and maps will be compared among the four major histological types of ovarian carcinoma to identify markers associated with each histological subtype. Putative marker proteins will be identified using molecular weight determination via electrospray (ESI)-mass spectrometry and peptide mapping using MALDI-TOF mass spectrometry. The protein expression results will be compared to existing mRNA microarray data from the same cell lines and tissues to check for correlations between protein and gene expression. Tissue microarrays (TMAs) will be used to validate candidate ovarian carcinoma biomarkers by staining the TMAs immunohistochemically with antibodies directed against specific marker proteins. The TMAs will include normal tissues and ovarian carcinomas with all stage/grade/morphologic subtype combinations represented, for the purpose of determining global expression patterns, cell type specificity, and subcellular localization of candidate biomarkers.
