Characterization of naturally occurring mutations in the precore and core genes of hepatitis B virus

Persistent infection with hepatitis B viruses (HBV) can cause a major problem to human health including the development of hepatocellular carcinomas. However, the mechanism of establishing chronicity of HBV infection remains poorly understood. The wide spectrum of the sequelae caused by HBV infection most likely reflects the different consequences of the complicated virus-host interaction. To understand the virus-host interaction, we focused our research on the naturally-occurring HBV variants which might shed light on the evolution of virus and the pathogenesis of HBV infection. Since nucleocapsid protein, core antigen (HBcAg), is the major target for the T cell immune response during HBV infection, we have characterized HBV variants with a core internal deletion (CID) as well as an e antigen (HBeAg) negative, stop codon mutation. Worldwide, both mutants are frequently found in chronic hepatitis B patients, and have not been found in acute hepatitis patients. Most of the e antigen negative variants result from a nonsense mutation in the precore region. At nucleotide (nt) 1896, a G-to-A mutation creates a TAG stop codon at codon 28 of precore protein. Another G-to-A mutation at nt 1899 often occurs either alone or with the 1896-nt mutation. HBV e antigen negative mutants bearing a TAG stop codon mutation at codon 28 uniformly replicate at least 20 folds better than mutants bearing a TGA stop codon at the same amino acid position. A single mutation at nt 1899, changing a G to a pyrimidine (T or C) is deleterious to viral RNA encapsidation and DNA replication. Our results explain why only a purine (G or A) at nt 1899, never a pyrimidine, is observed in HBV infection in nature. The stringent selection for a highly efficient RNA encapsidation element may play a crucial role in the selection for these two closely-linked precore mutations in nature. The preferential occurrence of the G to A mutation at nt 1896 and 1899 is likely to be a combined consequence of both selection and higher intrinsic mutation frequency at these positions. Functional characterization of the CID mutants revealed all of the features of defective interfering (DI) particles originally discovered in influenza virus half a century ago. Internal deletion alone is necessary and most likely to be sufficient for the DI phenotype. Our results provided the first experimental evidence for the existence of DI variants in natural infections in humans. We hypothesize here that the virus-virus interactions between the sub-populations of the replication-competent helper viruses and the replication-defective DI particles could contribute to the fluctuation of immune targets in virus-host interactions and consequently the establishment and maintenance of HBV infection