Phylogeny of the genus Flavivirus using complete coding sequences of arthropod-borne viruses and viruses with no known vector

Abstract

Attempts to define the evolutionary relationships and origins of viruses in the genus Flavivirus are hampered by the lack of genetic information particularly amongst the non-vectored flaviviruses. Using a novel protocol for sequence determination, the first complete coding sequence of St Louis encephalitis virus and those of two representative non-vectored flaviviruses, Rio Bravo (isolated from bat) and Apoi (isolated from rodent), are reported. The encoded polyproteins of Rio Bravo and Apoi virus are the smallest described to date within the genus Flavivirus. The highest similarities with other flaviviruses were found in the NS3 and NS5 genes. The proteolytic cleavage sites for the viral serine protease were highly conserved among the flaviviruses completely sequenced to date. Comparative genetic amino acid alignments revealed that p-distance cut-off values of 0·330–0·470 distinguished the arthropod-borne viruses according to their recognized serogroups and Rio Bravo and Apoi virus were assigned to two distinct non-vectored virus groups. Within these serogroups, cladogenesis based on the complete ORF sequence was similar to trees based on envelope and NS5 sequences. In contrast, branching patterns at the deeper nodes of the tree were different from those reported in the previous study of NS5 sequences. The significance of these observations is discussed.