Genomics of hepatocellular carcinoma: Hepatitis virus infection and hepatocarcinogenesis

Chronic infection of the liver by HBV or HCV constitute the most important risk factors for the development of hepatocellular carcinoma (HCC). Adeno-associated virus 2, is a third virus recently found to be associated with HCC, although the world-wide prevalence of this virus is not yet known. HBV and AAV2 both integrate into the host genome where they can disrupt normal cellular pathways relating to cancer. Integration of HBV has been the subject of several detailed studies by next generation sequencing and these have uncovered integration sites that could affect hundreds of genes. However, among these studies, integration of only 6 genes are replicated in more than one study: TERT and MLL4, are the most common sites of integration, followed by CCNE1, CCNA2, SOX5 and SENP5. It’s an interesting feature of the HBV-associated disease that among this small set of genes, only TERT is recurrently mutated in tumors without virus thereby suggesting the activation of alternative routes to tumorigenesis. SENP5, in particular, is a peptidase that removes SUMO peptides from post-translationally conjugated proteins revealing novel regulatory mechanisms at work in these tumors. But what role, if any, is played by the hundreds of sites observed only once across all cohorts studied so far? Listed among the genes observed only once with an integrated virus, that is, singleton integrations, are many genes known to play a role in cancer, included among them lincRNA and antisense RNA. Gene expression analysis shows both overexpressed oncogenes such as GLI2 and CCND1 and down-regulation of tumor suppressors such as NCOR2 and ZFPM2 strengthening the notion that at least some of the singleton viral integrations are playing a role in the individual patient’s disease. These results also highlight the importance of multi-platform analysis of viral integration data. Surprisingly, fibronectin (FN1) is also a recurrent integration target, resulting in its overexpression, but is observed almost exclusively in normal or cirrhotic tissue adjacent to tumors rather than the tumor itself, suggesting a different role for this gene in liver pathogenesis. Thus integrated HBV appears to play multiple roles in HCC tumorigenesis, activating or inactivating key cancer genes, with a wide variety of functional targets in play, much of which goes undetected in screens relying only on recurrent events in large cohorts.