Genomic properties of 62 field isolates of bovine viral diarrhea virus (BVD
V) collected fi-om 1974 to 1999 in Japan were investigated. The 5' untransl
ated region (UTR) was amplified by reverse transcriptase polymerase chain r
eaction (RT-PCR) and the 244 to 247 base nucleotide sequences were determin
ed. Serological properties were also characterized by the cross-neutralizat
ion test using antisera against the representative strain of the classified
genotype. Using phylogenetic tree analysis, BVDV 1 was subdivided into two
major clusters, BVDV-1a (29 isolates) and BVDV-1b (27 isolates). In group
1a, 3 differed from the other viruses, and were classified in a branch assi
gned as 1a'. However, 4 isolates (So CP/75, 190 CP, 190 NCP and KS86-1-NCP)
could not be assigned to group 1a or 1b. In comparison with the published
sequence data, KS86-1-NCP, 190 CP and 190 NCP were similar to the Southern
Africa isolates that have recently been proposed as BVDV 1c. The 5' UTR seq
uence of So CP/75 was unique among those of BVDV 1, suggesting that the iso
late should be classified into a new genotype. Only 2 out of 62 isolates co
llected in 1989 and 1990 were identified as BVDV 2. The results of the cros
s-neutralization test strongly supported these data, suggesting a close cor
relation between the 5' UTR sequence and the antigenicity of BVDV.