Do mutant woodchuck hepatitis viruses play a role in hepatocellular carcinogenesis?

Although hepadnaviruses are implicated in the aetiology of hepatocellular carcinoma, the pathogenic mechanisms involved remain uncertain. Clonally propagated integrations of hepadnaviral DNA into cellular DNA can be demonstrated in most virally induced hepatocellular carcinomas. Integration occurs at random sites in cellular DNA, but the highly preferred sites in viral DNA are adjacent to the directly repeated sequence, DR1, less often DR2, or in the cohesive overlap region. Integrants invariably contain simple deletions or complex rearrangements that have been thought to occur after integration. We report here the detection, in the serum of woodchucks with hepatocellular carcinoma, of mutant woodchuck hepatitis viruses that are strikingly similar to the rearranged genomes found previously as integrated sequences in cellular DNA. Of the four mutants studied, two had large inverted duplications, one a 219 nucleotide direct duplication, and one a 219 nucleotide deletion. Virus-virus DNA junctions occurred either adjacent to DR1 or DR2 or in the cohesive overlap region at topoisomerase I cleavage sites. Thus, it is possible that rearrangement of the hepadnavirus genome precedes integration of viral DNA into cellular DNA and that mutant genomes that are preferentially integrated into cellular DNA have an aetiological role in hepatocarcinogenesis.