Evolutionarily conserved RNA secondary structures in coding and non-coding sequences at the 3' end of the hepatitis G virus/GB-virus C genome

Laboratory for Clinical and Molecular Virology, University of Edinburgh, Summerhall, Edinburgh EH9 1QH, UK 1 Author for correspondence: Peter Simmonds. Fax +44 131 650 7965. e-mail Peter.Simmonds{at}ed.ac.uk Hepatitis G virus (HGV)/GB virus C (GBV-C) causes persistent, non-pathogenic infection in a large proportion of the human population. Epidemiological and genetic evidence indicates a long-term association between HGV/GBV-C and related viruses and a range of primate species, and the co-speciation of these viruses with their hosts during primate evolution. Using a combination of covariance scanning and analysis of variability at synonymous sites, we previously demonstrated that the coding regions of HGV/GBV-C may contain extensive secondary structure of undefined function (Simmonds & Smith, Journal of Virology 73, 5787–5794, 1999 ). In this study we have carried out a detailed comparison of the structure of the 3'untranslated region (3'UTR) of HGV/GBV-C with that of the upstream NS5B coding sequence. By investigation of free energies on folding, secondary structure predictive algorithms and analysis of covariance between HGV/GBV-C genotypes 1–4 and the more distantly related HGV/GBV-C chimpanzee variant, we obtained evidence for extensive RNA secondary structure formation in both regions. In particular, the NS5B region contained long stem–loop structures of up to 38 internally paired nucleotides which were evolutionarily conserved between human and chimpanzee HGV/GBV-C variants. The prediction of similar structures in the same region of hepatitis C virus may allow the functions of these structures to be determined with a more tractable experimental model.