PURPOSE: Successful prevention of colorectal cancer (CRC) would benefit from a rapid serum screening for early detection. Here, a novel strategy for CRC biomarker discovery and validation exclusively based on MS procedures is reported. EXPERIMENTAL DESIGN: Identification of CRC serum biomarkers is initially made using label-free quantification on pooled serum samples from different CRC stages followed by two consecutive steps of targeted parallel reaction monitoring assays in different serum cohorts. Relevance of different protein depletion and peptide fractionation extent is investigated. Absolute quantification of a selected peptide is performed as a proof-of-concept. RESULTS: A total of 945 proteins showed differential abundance in the discovery phase. Based on their statistical significance and relative expression in disease stages, 123 potential biomarkers are selected for a training step. In the final validation step, five peptides belonging to four proteins are consistently quantified in individual CRC serum samples and controls. Different statistical analyses indicate that peptides GWVTDGFSSLK (APOC3) and LCNNPTPQFGGK (THBS1) are candidate biomarkers. Absolute quantification of LCNNPTPQFGGK shows statistical significance for the diagnosis of early respect to late CRC stages. CONCLUSIONS AND CLINICAL RELEVANCE: Two peptides from APOC3 and THBS1 are validated by PRM as potential biomarkers for non-invasive diagnosis of colorectal cancer.
PURPOSE: Successful prevention of colorectal cancer (CRC) would benefit from a rapid serum screening for early detection. Here, a novel strategy for CRC biomarker discovery and validation exclusively based on MS procedures is reported. EXPERIMENTAL DESIGN: Identification of CRC serum biomarkers is initially made using label-free quantification on pooled serum samples from different CRC stages followed by two consecutive steps of targeted parallel reaction monitoring assays in different serum cohorts. Relevance of different protein depletion and peptide fractionation extent is investigated. Absolute quantification of a selected peptide is performed as a proof-of-concept. RESULTS: A total of 945 proteins showed differential abundance in the discovery phase. Based on their statistical significance and relative expression in disease stages, 123 potential biomarkers are selected for a training step. In the final validation step, five peptides belonging to four proteins are consistently quantified in individual CRC serum samples and controls. Different statistical analyses indicate that peptides GWVTDGFSSLK (APOC3) and LCNNPTPQFGGK (THBS1) are candidate biomarkers. Absolute quantification of LCNNPTPQFGGK shows statistical significance for the diagnosis of early respect to late CRC stages. CONCLUSIONS AND CLINICAL RELEVANCE: Two peptides from APOC3 and THBS1 are validated by PRM as potential biomarkers for non-invasive diagnosis of colorectal cancer.