Molecular Identification and Characterization of a New Type of Bovine Enterovirus

Abstract

Bovine enteroviruses belong to the family Picornaviridae. Little is known about their pathogenic potential; however, they cause asymptomatic infections in cattle and are excreted in feces. In the present study, viruses isolated from environmental samples were sequenced. According to phylogenetic analyses and standard picornavirus nomenclature, these isolates constitute a new type of bovine enterovirus serogroup A.

TEXT

Enteroviruses (EVs) constitute a large genus within the positive-stranded RNA family Picornaviridae, with 10 species, including Bovine enterovirus. Bovine enteroviruses (BEVs) are found mostly in cattle populations, but little is known about their pathogenic potential. They were first isolated in the late 1950s from cattle (11, 17) and have been isolated and studied since then. Initially, they were classified into seven serotypes (6), which were later reduced to four (1) and then two (10) serotypes. Recently, based on sequence data and phylogenetic analysis, BEVs were divided into two serogroups, BEV-A and BEV-B, containing three types (BEV-A1, BEV-A2, and BEV-A3) and six types (BEV-B1, BEV-B2, BEV-B3, BEV-B4, BEV-B5, and BEV-B6), respectively (http://www.picornaviridae.com/enterovirus/bev/bev.htm).

In the present study, we characterized seven untypeable PAK-NIH isolates as BEVs based on the 5′ untranslated region (5′ UTR) and viral protein 1 (VP1) gene. Our data showed that these isolates form a distinct genetic cluster and reveal a new type of BEV-A circulating in the country.

A total of 229 sewage samples from three different sites in Pakistan (Lahore, Karachi, and Quetta) were analyzed at the National Institute of Health, Islamabad. Samples were concentrated by a two-phase separation method (24), and 0.5-ml aliquots of concentrate were inoculated into five L20B cell flasks and one RD cell flask (15, 20) for the isolation of poliovirus and enteroviruses. Any sample exhibiting a cytopathic effect on L20B cells was passaged on RD cells and subjected to a microneutralization assay (25).

All untypeable samples were screened for human and bovine enteroviruses by reverse transcriptase PCR (RT-PCR) using the primers and probes corresponding to the 5′ UTR (8, 12). The VP1 gene of BEVs was further amplified by using primers BEV-1C587F (CCATGTGGTAYCARACIAAYATGGT) and BEV-2A82R (GATTGCCAIACTTCATTYTCCCA) (9).

Sequencing was performed in an ABI 3130 genetic analyzer (Applied Biosystems), and chromatograms were inspected in Sequencher version 4.9. VP1 and 5′-UTR sequences were subjected to online BLAST analyses. Sequences were aligned with the help of ClustalW (23). Phylogenetic trees were constructed by the neighbor-joining method using the Tamura-Nei model for nucleotide sequences (22) and the Jones-Taylor-Thornton (JTT) substitution model for amino acid sequences (7) in MEGA software (21). The statistical significance of phylogenetic trees was estimated by bootstrap analysis with 1,000 pseudoreplicate data sets.

Seven samples untypeable by microneutralization (Table 1) were confirmed as BEVs by RT-PCR. Their partial 5′-UTR sequences showed 84 to 85% nucleotide identity to sequences from different BEV strains in the online blastn program. Upon pairwise distance comparison, five PAK-NIH isolates (21E5, 48E3, 68E2, 92E1, and 123E1) were found to be identical to one another, having 93.7 to 99.4% nucleotide similarity, and were placed into group I, whereas two isolates (59E1 and 67E2) showed 86% identity and were placed into group II; a similar grouping pattern was observed upon phylogenetic analysis (Fig. 1). PAK-NIH isolates were closely matched with BEV-A2 SL305, having 82.9 to 86.3% nucleotide homology, and have also shown 81.7 to 84.0% nucleotide identity to BEV-A2 K2577.

Table 1.

Details of untypeable samples characterized in this study^a

Strain	Collection yr	Location	Accession no. for:		PCR result for:
			5′ UTR (partial)	VP1	HEV	BEV
PAK-NIH-21E5	2009	Karachi	JQ690748	JQ690741	Neg	Pos
PAK-NIH-48E3	2011	Quetta	JQ690749	JQ690742	Neg	Pos
PAK-NIH-59E1	2010	Lahore	JQ690750	JQ690743	Neg	Pos
PAK-NIH-67E2	2010	Lahore	JQ690751	JQ690744	Neg	Pos
PAK-NIH-68E2	2010	Karachi	JQ690752	JQ690745	Neg	Pos
PAK-NIH-92E1	2010	Karachi	JQ690753	JQ690746	Neg	Pos
PAK-NIH-123E1	2010	Karachi	JQ690754	JQ690747	Neg	Pos

^aAll samples were isolated from sewage. HEV, human enterovirus; Neg, negative; Pos, positive.

Fig 1.

Phylogenetic trees constructed through the neighbor-joining method by alignment of partial 5′-UTR (A) and complete VP1 (B) nucleotide sequences of PAK-NIH isolates (labeled ●) with reference BEV sequences retrieved from http://www.picornaviridae.com/enterovirus/bev/bev_seq.htm and evaluated by 1,000 bootstrap pseudoreplicates. Only bootstrap values over 50% are shown at branch nodes. The scale bar represents 5% nucleotide sequence divergence. First and second closest matches are labeled ▲ and ■, respectively. Each reference strain designation includes the strain name and GenBank accession number. Details of the strains used to prepare phylogenetic trees are available upon request. UC, unclassified; OEV, ovine enterovirus.

VP1 sequences of PAK-NIH isolates were analyzed online using the programs blastn (demonstrating 73 to 75% nucleotide identity) and blastx (demonstrating 82 to 89% amino acid identity) and were also aligned and compared with other BEV sequences retrieved from GenBank. The closest match was German strain BEV-A3 D14/3/96, having 71.2 to 73.8% nucleotide identity (84.9 to 87.1% amino acid identity), and the second closest was an Australian strain of BEV-A2, K-2577, with 70.4 to 72.2% nucleotide identity (82 to 85.3% amino acid identity). The phylogenetic tree (Fig. 1) showed that study isolates formed a distinct cluster segregating into two groups supported by high bootstrap values. Group I members (48E3, 68E2, 92E1, and 123E1) had 97.5 to 99.4% nucleotide similarity (98.2 to 99.3% amino acid identity) to one another, and group II members (21E5, 59E1, and 67E2) showed 82.5 to 86.1% nucleotide identity (92.6 to 96.0% amino acid identity), revealing more divergence than group I. Group I had two transmission chains; one (for 68E2, 92E1, and 123E1) was in Karachi, and the second (for 48E3) was in Quetta. Similarly, two transmission chains were found for group II; one (for 59E1 and 67E2) was in Lahore, and the second (for 21E5) was only in Karachi (Fig. 1). PAK-NIH-59E1 was the most divergent isolate of this distinct cluster, showing 83.5% mean nucleotide identity (94.1% amino acid identity) to other PAK-NIH isolates.

Sequencing of the VP1 gene provides an important source of information for enterovirus classification (4, 5, 18), especially for the differentiation of serotypes (14, 16). For demarcation of a new serotype, values of <75% nucleotide identity and <88% amino acid identity have been employed as yardsticks (18, 19). Our isolates fulfill these criteria, with 86.2% amino acid and 71.8% nucleotide mean identities in the VP1 region to the closely matched BEV-A3 strain D14/3/96, and could be considered a new type of BEV-A.

Our data suggest that the genetic clustering pattern of PAK-NIH isolates and the results of comparison with available BEV sequences in the phylogenetic analysis (Fig. 1) are similar to the pattern previously reported by Zell et al. (26). They sequenced the whole capsid region, but on the basis of data available from the present study, we propose that the same information may be obtained by sequencing the VP1 gene rather than the whole capsid region. PAK-NIH isolates have a wide geographical distribution including three cities from three different provinces of Pakistan and demonstrate four different chains of transmission. Although the number of strains isolated from each city is limited, their genetic diversity shows a long period of circulation, and it may be possible that this new type has emerged as a result of genetic evolution of circulating BEVs, which needs to be studied in detail in the future.

A difference in grouping of BEV strains correlated with serotype between the 5′-UTR and VP1 phylogenies has been observed (Fig. 1), which is due to recombination that is a normal phenomenon in enterovirus species (2, 13). However, our isolates form a distinct genetic cluster based on sequences from both regions, and analyzing two regions did not yield contradictory or conflicting information. Furthermore, the analysis of 5′ UTRs and VP1 genes of PAK-NIH isolates reveals that these isolates fulfill the enterovirus classification criteria and need to be considered a new type of BEV-A, for which we propose the label BEV-A4.

Nucleotide sequence accession numbers.

The VP1 and 5′-UTR nucleotide sequences have been deposited in the GenBank database under accession numbers JQ690741 to JQ690754.