Prioritization of driver genes in cancer-associated copy number alterations identifies B4GALT5 as a glycooncogene

Abstract

Cancer is a disease resulting from aberrant communication between cells of a multicellular organism. The glycan coat that surrounds the cells is an important player in cellular communication. While altered cell surface glycans are known biomarkers for cancer, glycan biosynthesis itself has not been considered a potential oncogenic pathway. So, to understand the oncogenic potential of the glycan biosynthetic pathways we have analyzed the copy number alterations (CNA) of genes encoding for glycosylation regulators (glycogenes) in cancer genome datasets and identify novel glyco-oncogenes and glyco-tumor suppressor genes (TSGs). CNA of oncogenes and TSGs is an important cancer-associated genetic alteration that associates with worst prognosis. Nevertheless, identity of the driver genes in the copy number altered segments of the genome remains obscure in most cases. We developed a prioritization pipeline based on bioinformatic and experimental criteria to identify putative driver genes. In addition to correctly identifying several well-established oncogenes/TSGs, this pipeline discerns several novel oncogenes and TSGs, some of which are glycogenes. Further, among glyco-oncogenes there is an enrichment for glycosphingolipid biosynthetic pathway and trans-Golgi associated lysosomal sorting machinery and among glyco-TSGs there is an enrichment for early N-glycan biosynthetic enzymes. As a proof-of-principle we show that one of the identified glycooncogene B4GALT5, encoding a key enzyme in the glycosphingolipid pathway exhibits oncogenic property of promoting increased growth of hepatocellular carcinoma cells. Thus, this study identifies glycosylation pathways with oncoregulatory properties and opens up a new group of enzymes as potential therapeutic targets for cancer.