Characterization of GBV-CHGV viral genome: comparison among different isolates for a ∼ 2 kb-sequence that covers entire E1 and most of 5′UTR and E2

Abstract

It is yet not fully uncovered where and how translation starts in GB virus C or hepatitis G virus ($\text{GBV-}\text{C}\text{HGV}$), and, accordingly, information on structural proteins of this virus has been scarce. To know a vertebral architecture of $\text{GBV-}\text{C}\text{HGV}$ genome at least for its 5′ portion, we analyzed 31 Japanese isolates of $\text{GBV-}\text{C}\text{HGV}$. Our results indicated that most of the Japanese isolates were classifiable into ‘Japanese type’, two isolates into ‘North American type’, one into ‘African type’, and another into the fourth distinct type. Despite such sequence diversity across individual types, however, several common characteristics existed. (1) The AUG that resides just downstream of an oligopyrimidine tract was conserved by all isolates as the first in-frame initiation codon for the polyprotein. (2) The sequence upstream of this AUG had many covariant nucleotide substitutions, suggesting conserved secondary structure. (3) The ORF that starts with this AUG did not code for core. (4) Predicted amino acid sequence of putative E1 and E2 was highly conserved among different types, and did not contain hypervariable regions.