Molecular mechanisms in human hepatocellular carcinoma

Abstract

Hepatocellular carcinoma (HCQ is one of the commonest cancers worldwide. There is, however, a marked geographical variation in incidence and it has been suggested that the pathogenesis may vary in different parts of the world. A retrospective analysis of 110 HCC patients was initially undertaken which confirmed that only 29% of British patients had markers of hepatitis B infection, suggesting a possible role for other environmental agents in the pathogenesis, and that 80% of patients had underlying cirrhosis. The nature of the strong relationship between HCC and cirrhosis has not been established but it has been postulated that increased hepatocyte turnover in the cirrhotic liver may predispose to DNA damage by environmental mutagens. Cell proliferation is required to express the strongly promutagenic DNA base lesion 0'-methylguanine, produced by alkylating agents, as a mutation. &- methylguanine is repaired by the DNA repair enzyme 06-methylguanine-DNA methyltTansferase (06-MT). A microassay was developed which could reliably measure 06-MT levels in liver biopsy samples. Using this approach 06-MT levels were found to be significantly lower in cirrhotic liver when compared to non-cirrhotic and normal liver tissue. No correlation was found between lymphocyte and liver levels from individual patients with liver disease indicating that the deficiency in DNA repair is disease-a nd tissue-specific. Three polyclonal antibodies were subsequently raised to 06-MT peptides and characterised by immunoblotting in an attempt to establish the tissue distribution of the enzyme in liver. Although none of the antisera were able to detect &-MT in tissue sections they were used to analyse structural differences in the enzyme between cirrhotic and non-cirrhotic liver using SDS-PAGE followed by immunoblotting and fluorography. A band of M, 24,000r,e presentingn ative enzyme, was visualised by fluorography in all liver extracts. Densitometry of these bands correlated with the enzyme activity determined by the direct enzyme assay, validating the assay findings. Other small molecular weight bands were seen in all liver extracts and comparison with immunoblots suggested that these bands represent C-terminal truncated enzyme. The spectrum of smaller molecular weight enzyme forms was similar in cirrhotic and non-cirrhotic liver. It was, thus, concluded that although 06-MT levels were lower in'cirrhosis this was not accounted for by structural differences in the enzyme. DNA mutations (G to A) produced by the failure to repair 06-methylguanine are known to activate oncogenes and turnour suppressor genes such as p53. However only 5/55 (9%) of HCC expressed mutant p53. Other factors potentially involved in hepatocarcinogenesis include the growth factor TGF-a and a growth factor receptor encoded by the c-erb B-2 proto-oncogene. Expression of TGF-a and the C-erbB -2 oncoprotein were seen in 8/28 (28%) and 2/26 (8%) of HCC respectively, findings which differ from those observed in HCC from the Far East. Deficient DNA repair by &-MT provides one possible reason why cirrhosis is an important risk factor for the development of HCC. However, failure to repair 06-mothylguanine does not result in mutations within the p53 gene in British HCC. Furthermore, the finding of low expression of mutant p53, TGF-a and the c-erb B-2 oncoprotein in HCC from Britain compared to HCC from the Far East and Africa suggests geographical differences in the molecular mechanisms involved in hepatocarcinogenesis between areas of high and low HCC prevalence.