Glycoprotein screening in colorectal cancer based on differentially expressed Tn antigen

Colorectal cancer (CRC) is one of the most common cancers worldwide, and the identification of new biomarkers for CRC is valuable for its diagnosis and treatment. We aimed to screen differentially expressed glycoproteins (especially O-glycoproteins) and to identify diagnostic or therapeutic candidates for colorectal cancer (CRC) based on different Tn antigen expression levels. Fresh cancer tissues and adjacent healthy tissues were obtained from CRC patients and classified into three groups based on their Tn antigen expression: CRC with negative Tn expression (CRC Tn‑), CRC with positive Tn expression (CRC Tn+) and normal control without Tn expression (NC). Protein extractions were separated and identified by iTRAQ technology. Glycoproteins and O-glycoproteins were selected using UniProt and DAVID. Deep bioinformatic analysis, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KO), was used to annotate this O-glycoprotein interaction network. Subsequently, two O‑glycoproteins were verified by western blotting and immunohistochemistry in either LS174T cells or CRC tissues. We found that 330 differentially expressed proteins were identified by iTRAQ between CRC Tn‑ and NC tissues, 317 between CRC Tn+ and NC tissues, and 316 between CRC Tn‑ and Tn+ tissues. Of the 316 proteins, 55 glycoproteins and 19 O‑glycoproteins were identified and analyzed via deep informatics. Namely, different Tn antigen expression levels in CRC led to differential protein expression patterns, especially for glycoproteins and O‑glycoproteins. Decorin and SORBS1, two representative functional O-glycoproteins, were significantly downregulated in the CRC Tn+ tissues compared with the level in the CRC Tn‑ or NC tissues. Based on this deep bioinformatic analysis, Decorin and SORBS1 are hypothesized to be involved in the TGF‑β and PPAR‑γ signaling pathways, respectively.