Bovine viral diarrhea virus (BVDV) is a member of the genus Pestivirus of the family Flaviviridae, together with the classical swine fever virus (CSFV) and sheep border disease virus (BDV). Based on antigenic and nucleotide sequence differences, two genotypes, categorized as BVDV-1 or BVDV-2, have been discriminated in the past years. So far, 21 subgenotypes (BVDV-1a to BVDV-1u) have been reported worldwide [1–4]. In addition to BVDV-1 and BVDV-2, genetically distinct isolates including Th/04_KhonKaen virus, D32/00-‘HoBi’, and CHKaHo/cont may be members of a third putative species and are referred to as “HoBi-like”, “BVDV-3”, or “atypical pestiviruses” [5, 6]. The BVDV genome comprises a positive single-strand RNA (ssRNA) molecule of approximately 12.3 kilobases (kb) in length, with one open reading frame (ORF) flanked by 5′ and 3′-untranslated regions (UTR). The BVDV ORF encodes a polyprotein that is cleaved into four structural proteins (C, Erns, E1, and E2) and seven non-structural proteins (Npro, p7, NS2-3, NS4A, NS4B, NS5A, and NS5B). In addition, based on their ability to cause a cytopathic effect, BVDV strains are further categorized as cytopathogenic (CP) or noncytopathogenic (NCP). In the past years, BVDV 1b, 1m, and 1q have commonly been detected in Chinese cattle and pigs [7–11]. However, the prevalence of BVDV-1d in cattle in China has rarely been reported [12]. This subtype is also occasionally detected in yaks [13]. In the current study, we determined the full-length genome sequences of three BVDV isolates and analyzed the specific genomic information and phylogenetic relationship to other recently reported novel and representative pestiviruses.

Three BVDV strains (BJ1201, BJ1305, and BJ1308) were isolated from cows between 2012 and 2013 in Beijing, China, and stored in our laboratory. These three viruses were isolated from serum samples as described in a previous report [14]. Further historical and epidemiological details about these isolates are shown in Table S1. The three isolates were propagated in Madin-Darby bovine kidney (MDBK) cells, and no cytopathic effect was observed. Viral RNA was then extracted from the cellular fraction using a Quick-RNA Kit (Aidlab Biotechnologies Co., Ltd, China), following the manufacturer’s instructions. The three complete genome sequences have been deposited in the GenBank database under accession numbers KT943518, KT951840, and KT951841. Sequence alignments and identity analysis were performed using the MegAlign program (DNASTAR) and NCBI BLAST (BLAST: Basic Local Alignment Search Tool) online.

To investigate the evolutionary relationship of BVDV isolates, phylogenetic analysis was performed using Molecular Evolutionary Genetics Analysis (MEGA) software, version 6.06 [15]. Three genome sequences were aligned with reference strains retrieved from GenBank, including BVDV, CSFV, and other representative strains. Phylogenetic trees were reconstructed using MEGA software, based on the full-length genome, 5′-UTR, and Npro gene. The bootstrap values were obtained for 1000 replicates using the neighbor-joining algorithm (NJ), and evolutionary distances were determined using the Kimura two-parameter model. Based on BLAST analysis and known characteristics of putative post-transcriptional processing sites of other BVDVs [7, 8], the positions of the coding genes and UTRs in the genome of the three isolates were determined (Table S2).

A phylogenetic tree constructed based on full-length genome sequences revealed that the isolates BJ1201, BJ1305, and BJ1308 clustered with previous BVDV-1 isolates (Fig. 1a). The nucleotide sequence identity among the three isolates was 99.0 %, and the deduced polypeptide sequences of BJ1201 and BJ1308 were 100 % identical. The three genome sequences shared nucleotide sequence identities of 73.4 % to 98.3 % with the other BVDV-1 strains, about 70.0 % with BVDV-2 strains, about 67.0 % with BVDV-3, about 66.3 % with CSFV strains, 66.3 % with the BDV strain X818, and 60.2 % with the pronghorn strain. Furthermore, the three isolates clustered with Korean strain 10JJ-SKR (Fig. 1a), which belongs to BVDV-1d subgenotype [16]. Twelve genes of BJ1305 were compared with those of seven BVDV-1 strains and three BVDV-2 strains (Table 1). The sequences of BVDV-2 viruses had a higher degree of divergence from those of BJ1305 than from those of the BVDV-1 strains. In the coding sequences, the highest degree of shared identity was observed between the BJ1305 and 10JJ-SKR strains. The full-length genome sequence of 10JJ-SKR was 98.3 % identical to those of the three isolates. However, the 3′UTR gene of BJ1305 was significantly different from that of 10JJ-SKR. The 3′UTR sequences of BJ1305 and 10JJ-SKR were 12.1 % identical. In a comparison of the coding sequences of BVDV-1 and BVDV-2, the ZM-95 virus was found to have a six-nucleotide insertion in the Erns gene, and 890, CP7 and VEDEVAC were found to have insertions of 228, 27 and 45 nucleotides, respectively, in the NS2 gene (Table 1). These results show that the BJ1305 virus differs genetically from other BVDV-1 strains.

Fig. 1
figure 1

Phylogenetic analysis based on the nucleotide sequences of (a) the full-length genome, (b) the 5′-UTR, and (c) the Npro gene

Full size image
Table 1 Nucleotide sequence identity (%) of the coding sequences of the BJ1305 virus to those of other isolates of BVDV-1 and BVDV-2
Full size table

To confirm the subtype assignment based on the full-length genome sequence and to compare our isolates to other reference strains, phylogenetic trees based on 5′-UTR and Npro gene were reconstructed (Fig. 1b, c). For this analysis, sequences from members of 21 genetic subgroups of BVDV-1 from different regions of the world were used [1, 3, 7, 8, 17]. The two phylogenetic trees showed that isolates BJ1201, BJ1305, and BJ1308 clustered in the same phylogenetic branches as in the phylogenetic tree based on the full-length genome (Fig. 1b, c). The 5′-UTRs of the 16-111 and F-AU viruses had 96.7 % and 97.5 % sequence identity, respectively, to those of the three isolates in the present study. The Npro the isolates of 721 and F-AU had 89.7 % and 89.9 % sequence identity to those of the three isolates. In a previous study, Weng et al. reported the 5′-UTR and Npro sequences of 18 BVDV isolates, including BJ1201, BJ1305, BJ1308 [14], but the results differed from those of this study. First, the samples for virus isolation in this study were different from those used in the previous study (Table S1). Second, there were distinct differences in the nucleotide sequence of these isolates, as shown in Table S3. Finally, the subgenotype assignment for BJ1305 and BJ1308 in our study was different from that reported by Weng et al. (BJ1305-1m, BJ1308-1a).

In conclusion, we sequenced the complete genomes and analyzed the phylogenies of isolates BJ1201, BJ1305, and BJ1308 and obtained detailed genomic information about these three strains. Phylogenetic and sequence analysis based on the full-length genome, 5′-UTR, and Npro gene showed that these viruses are strains of the BVDV-1d subgenotype. This is the first report of the genomic sequences of BVDV 1d viruses in China.