The entire nucleotide sequences of two GB virus C/hepatitis G virus isolates of distinct genotypes from Japan.

Recently, putative viral agents responsible for human non-A to E hepatitis have been independently reported by two groups of investigators and designated GB virus C (GBV-C) and hepatitis G virus (HGV), respectively. The entire nucleotide sequences were determined for two viral genomes isolated from Japanese blood donors with GBV-C RNA. One of them (GT230) had a total genomic length of 9390 nucleotides (nt) with 5' and 3' untranslated regions of 551 and 313 nt, while the other (GT110) had genomic lengths of 9395, 281 and 315 nt, respectively. They both had a single long open reading frame, encoding 2842 amino acids (aa) in GT230 and 2933 aa in GT110. Surprisingly, they both lacked a clearly identifiable core gene, and possessed the E1/E2 gene with only four potential N-linked glycosylation sites. Pairwise comparison and phylogenetic analysis of the entire sequence indicated that the prototype GBV-C and two HGV isolates reported, as well as GT230 and GT110, are the same virus possibly of different genotypes. The five GBV-C/HGV isolates were variable up to 13.8% in the genomic nucleotide sequence, and contained deletions and insertions within the 5'-terminal 518-593 nt, which resulted in four different sizes of predicted polyproteins encoded by genomes of individual isolates. By contrast, the 3' untranslated region was well conserved. The high degree of sequence conservation within this region would favour it as a target for sensitive detection of GBV-C/HGV RNA.