A novel enterovirus species identified from severe diarrheal goats

Backgrounds The Enterovirus genus of the family of Picornaviridae consists of 9 species of Enteroviruses and 3 species of Rhinoviruses based on the latest virus taxonomy. Those viruses contribute significantly to respiratory and digestive disorders in human and animals. Out of 9 Enterovirus species, Enterovirus E-G are closely related to diseases affecting on livestock industry. While enterovirus infection has been increasingly reported in cattle and swine, the enterovirus infections in small ruminants remain largely unknown. Methods Virology, molecular and bioinformatics methods were employed to characterize a novel enterovirus CEV-JL14 from goats manifesting severe diarrhea with morbidity and mortality respectively up to 84% and 54% in China. Results CEV-JL14 was defined and proposed as a new Enterovirus species L within the genus of Enterovirus of the family Picornaviridae. CEV-JL14 had a complete genome sequence of 7461 nucleotides with an ORF encoding 2172 amino acids, and shared 77.1% of genomic sequence identity with TB4-OEV, an ovine enterovirus. Comparison of 5’-UTR and structural genes of CEV-JL14 with known Enterovirus species revealed highly genetic variations among CEV-JL14 with known Enterovirus species. VP1 nucleotide sequence identities of CEV-14 were 51.8%-53.5% with those of Enterovirus E and F, 30.9%-65.3% with Enterovirus G, and 43.8–51. 5% with Enterovirus A-D, respectively. CEV-JL14 was proposed as a novel species within the genus of Enterovirus according to the current ICTV demarcation criteria of enteroviruses. Conclusions CEV-JL14 clustered phylogenetically to neither Enterovirus E and F, nor to Enterovirus G. It was defined and proposed as novel species L within the genus of Enterovirus. This is the first report of caprine enterovirus in China, the first complete genomic sequence of a caprine enterovirus revealed, and the unveiling of significant genetic variations between ovine enterovirus and caprine enterovirus, thus broadening the current understanding of enteroviruses.

. Those viruses are the etiological agents contributing to neurological, respiratory and digestive disorders in human and animals [2][3][4][5][6][7][8]. Out of 9 Enterovirus species, Enterovirus E (EV-E), Enterovirus F (EV-F) and Enterovirus G (EV-G) are closely related to diseases affecting on the livestock industry, where EV-E and EV-F (formerly named bovine enterovirus A and B) are the causative agents of enterovirus infections in cattle manifesting clinical signs varying from respiratory diseases to enteritic, reproductive disease and infertility [4,5,[7][8][9][10][11]. The EV-G (formerly named porcine enterovirus B) is the causative agents of swine enterovirus infections [6,12,13]. It has been demonstrated that EV-E and EV-F were usually isolated from the cattle manifesting symptoms of digestive and respiratory diseases, suggesting the potential pathogenicity of enterovirus related to those infections. EV-E and EV-F were also occasionally isolated or detected from cattle of subclinical infection or environment [14][15][16][17], which leads to the conclusion that EV-E and EV-F were not significant agents to the livestock industry. With increasing number of enteroviruses isolated from animals developing respiratory and digestive diseases, the pathogenicity of enteroviruses to animal has been drawn much attention [7,8,11,13,18]. While infection of enteroviruses and their sequences have been increasingly reported in cattle and swine recently, the enterovirus infections in small ruminants such as sheep or goat remain largely unknown. Recently, the first complete genome sequence of an ovine enterovirus (TB4-OEV) was reported in 2012 [5], a partial 5'-UTR sequence of a caprine enterovirus in Japan and several BEV-like 5'-UTR sequences in Thailand were determined from diarrheal goats or unidentified goats [8,19]. Although the phylogenetic analysis has clustered those sequences to clades of BEV-like based on the 5'-UTR sequences [19], such an analysis need to be reevaluated since the paucity of enough enterovirus sequences from petite ruminants including sheep and goats. It was reported that TB4-OEV strain is likely an interspecies recombinant between bovine enterovirus and porcine enterovirus based on the virus genome sequence analysis, where its 5'-UTR sequence has a high sequence identity with EV-E and EV-F, and the rest of its genome has a high sequence identity to EV-G [5]. Phylogenetic analysis grouped the TB4-OEV as a member of the species EV-G currently.
Classification for enterovirus species is usually based on virus genetic variability and molecular difference, in which the sequences of 5'-UTR or the capsid protein regions were normally used. It is generally accepted that picornavirus serotypes are molecularly defined by the diversity of the capsid proteins, while the enterovirus species were determined by the less diverse non-structural protein regions [20][21][22][23]. These criteria have been used successfully to classify bovine enteroviruses to EV-E, EV-F and porcine enterovirus to EV-G as well as other enterovirus classification. Previously, we defined an enterovirus HY12 isolate from cattle manifesting severe digestive and respiratory diseases as EV-E [7]. Here, we reported the isolation and characterization of the enterovirus isolate from goats showing severe diarrhea with high morbidity and mortality in China, and proposed that CEV-JL14, along with TB4-OEV, shall be grouped to a novel designated enterovirus species L (EV-L) within the genus of Enterovirus. To our knowledge, this is the first complete caprine enterovirus genome revealed, which is potentially associated with enteric disease characterized with severe watery diarrhea with high morbidity and mortality.

Ethics statement
The diseased goats were sacrificed following a standard protocol approved specifically to this study by the animal ethics committee at Jilin University.

Virus isolation
Fecal samples were collected from diarrheal goats. Spleen samples were collected from the dead or sacrificed goats and processed following the protocols approved by the animal ethics committee at Jilin University. Briefly, fecal samples were processed as described previously [7]. Spleen samples were processed in a dilution of 1:10 (W/V) with 10 mM phosphate buffered saline (PBS) (pH7.2) and homogenized. After centrifugation at 12,000 × g for 15 min at 4˚C the supernatants were filtered with 0.45 nm filter and the flow-through was used to inoculate Vero cells with a confluency of 70-80%. The inoculums were discarded after incubation with cells for 2 h before the addition of Dulbecco's modified eagle's medium (DMEM) (Invitrogen) supplemented with 2% fetal bovine serum (Sigma), 2 μg/ml Gentamycin and 2 mM Lglutamine (Invitrogen). The cells were observed every 4-6 h, and cytopathic effects (CPE) were captured using Canon Digital Camera.

Electron microscope observation
Fecal, spleen and the infected cell culture samples were processed for EM observation as previously described [7]. Briefly, fecal and spleen samples were processed in 1:10 (w/v) with 10 mM, pH7.2 PBS, and homogenized before centrifuging at 8000 × g for 20 min at 4˚C. The infected cells were frozen and thawed for 3 times before centrifugation as described above. The supernatants were then incubated with 1% phosphotungstic acid, and the viruses were examined using electron microscope (JEM-2200FS/CR)(JEOL, Tokyo, Japan).

TCID 50 titration
TCID 50 was determined as previously described [24]. Briefly, the viruses were processed in 10-fold serial dilutions and used to infect quadruplicate wells for each dilution. After 48-72 h post infection, the wells with cytopathic effects were counted and TCID 50 was determined. Physicochemical properties of CEV-JL14 were characterized as described previously [7]. Viruses were treated with organic solvents (chloroform and ether), heated at 56˚C, 65˚C and 80˚C for 1 h, or incubated with DMEM of pH 3.0 and pH5.0 before infecting the cells.

RT-PCR amplification and determination of the complete genome sequences
Reverse transcriptase reaction was performed using SuperScript™ II Reverse Transcriptase (Invitrogen, Carlsbad, CA) following manufacturer's instruction. PCR amplification was done using Taq DNA polymerase (New England Biolabs, Ipswich, MA) after the conditions being optimized. The primers used for determination of the potential pathogens and for amplification of the complete genomic sequences were listed in Table 1. The complete genome sequence was amplified by PCR using high fidelity PCR kit (Takara, Dalian, China). PCR products were either directly sequenced or cloned into pGEM-T vector (Tiangen, Beijing) before sequencing. Each product or recombinant was sequenced at least twice. The complete genome sequence was obtained by joining the nucleotide sequences of several overlapping PCR-amplified fragments using DNAstar Lasergene software. The nucleotide sequence served as a template for searching homologous sequences through GenBank (www.ncbi.nlm.nih.gov).

Phylogenetic analysis
Alignment analysis of multiple sequences was performed using the Clustal W method [25]. The amino acid sequence of CEV-JL14 was deduced from the nucleotide sequence. Briefly, the complete nucleotide sequence of CEV-JL14, the polyprotein, the 5'-UTR nucleotide sequence, the deduced amino acids of VP1, P1, P2C and P3CD of CEV-JL14 were aligned with the corresponding regions of representative virus strains of EV-E, EV-F, EV-G and other enterovirus species available in the GenBank. The representative strains were listed in Table 2.  Phylogenetic tree was initially generated by neighbor-joining methods [26], and further analysis was performed with a bootstrap value of 500.

Outbreak of disease with a high morbidity and mortality
In April, 2014, a disease occurred on a goat farm in Jilin province in Northeast China characterized by goats manifesting pyrexia, fever, cough at the onset. Three to five days later, the goats showed respiratory distress, dyspnea and severe diarrhea. No obvious improvement was observed after they were treated with antibiotics or sulfa drug. Majority of the sick goats succumbed to death within 7-14 days. The recovered goats showed a wasting and curved-body hair. During the outbreak, 64 out of 76 goats in a barn showed some obvious clinical signs, 35 out of 64 diseased goats were deceased. The morbidity rate is 84% (64/76) and the mortality is 54% (35/64).

EM observation and virus isolation
To determine the potential agents involved in this outbreak, the fecal and spleen samples from 4 sick/sacrificed goats were processed, respectively, to inoculate the Vero cells for virus isolation. After incubation with the inoculums, Vero cells showed cytopathic effects (CPE) as early as 8-10 h. Compared with normal Vero cells (Fig 1A), infected cells initially became rounded before the majority of the cells detached off the flask 24-48 h post infection (PI) (Fig 1B). To rule out the possibility that the CPEs were caused by the potential toxic effect of the inoculums, the infected cells were frozen, thawed and used as inoculums to infect cells at least 5 passages. Similar CPEs were also observed in Vero cells after they were infected with viruses from different passages (not shown), suggesting the CPEs were indeed caused by viral pathogens. Further examination of infected cell cultures using electron microscope (EM) revealed 20-30 nm virus particles in the samples (Fig 1C and 1D). The viruses were isolated using Vero cells or detected using RT-PCR from three out of four fecal samples and two of four spleen samples collected respectively from 4 goats examined (Table 3), indicating that the viruses are likely associated with the outbreak. The isolated virus particles were designated as CEV-JL14.

Virus titration and physiochemical properties
To determine the virus infectivity, TCID 50 was determined following the protocol as previously described [24]. Serially diluted viruses were inoculated to quadruplicate samples in 96-well microplate. After 48-72 h, the wells showing CPE were counted. The TCID 50 for CEV-JL14 strain in 5 th passage from three repeats was shown to be 10 7.4 pfu/0.1ml. Characterization of physicochemical properties of CEV-JL14 showed that it was resistant to treatment in low pH (3.0). Treatment of CEV-JL14 with chloroform/ ether had no significant effects on its infectivity, suggesting it is a non-enveloped virus.

CEV-JL14 is a caprine enterovirus
To define the viral pathogen, PCRs were performed to identify the potential viruses using primers for parvovirus, enterovirus and foot and mouth disease virus (FMDV) since those viruses had the similar sizes of 22-30 nm. No fragments were obtained using primers for parvovirus and FMDV after different amplification conditions were tried, ruling out the possibility that those viruses are associated with the disease. However, fragment was amplified using primers selected based on the genome sequence comparison for EV-E, EV-F and EV-G (Table 1), suggesting that the virus is likely an enterovirus. Sequencing the PCR-amplified fragment (partial 5'-UTR) revealed that the fragment indeed contained a nucleotide sequence of enterovirus. This partial 5'-UTR fragment shared 86% sequence identity with the corresponding region of the ovine enterovirus strain TB4-OEV, confirming that the isolated virus is enterovirus.

The complete genome sequence of CEV-JL14
To unveil the genome sequence of CEV-JL14, the primers listed in Table 1 were designed based on the nucleotide sequence of TB4-OEV, and used to amplify the viral complete genome. Complete genome sequence of CEV-JL14 was obtained by joining the nucleotide sequence of several overlapping PCR amplified fragments. Sequence analysis showed that the CEV-JL14 had the highest sequence identity to ovine enterovirus TB4-OEV Since CEV-JL14 was the first caprine enterovirus with its complete genome sequence revealed, it is natural to determine its complete sequence homology with other enteroviruses. As shown in Table 4  CEV-JL14 had a higher similar 5'-UTR secondary structure to BHM26 than other Enterovirus species To determine the secondary structure feature of CEV-JL14, its 5'-UTR sequence was analyzed using RNAfold software as described previously [27]. The predicted secondary structure of CEV-JL14 was compared with TB4-OEV and representative enterovirus from different Enterovirus species including EV-E, EV-F, EV-G and EV-A. As shown in S1 Fig

CEV-JL14 belongs to a novel species within the genus Enterovirus
The criteria for classification of the unknown virus in the genus of Enterovirus is polythetic and normally based on the sequence variation of the 5'-UTR and the capsid structural genes [20,23]. To determine the phylogenetic status of CEV-JL14, the complete nucleotide sequences of CEV-JL14 and its encoded individual gene were aligned with the corresponding region of known representative enteroviruses. As shown in Fig 2A, alignment analysis of CEV-JL14 complete genome sequence with representative enterovirus species clustered the enteroviruses to 9 clades including EV-A, EV-B, EV-C, EV-D, EV-E, EV-F, EV-G, EV-J and a new clade that only consists of CEV-JL14 and TB4-OEV. Alignment analysis with 5'-UTR sequences also demarcates the enteroviruses into 9 clades where the CEV-JL14 was neither clustered to the clade of EV-E and EV-F, nor to the clade of EV-G and EV-A-D. However, it is surprising to note that CEV-JL14, together with several partial goat enteroviruses and TB4-OEV, was clustered to a distinct clade from those of EV-E, EV-F and EV-G, EV-A, EV-B, EV-C, and EV-D, suggesting the viruses in the distinct clade is the novel species within the genus of Enterovirus. We named the enterovirus in this clade (caprine enterovirus and ovine enterovirus) as Enterovirus species L (EV-L). Similar patterns were also observed when the VP1 amino acid sequence (Fig 2C), the polyprotein sequences (Fig 2D), P1, P2C, and P3CD amino acid sequence (Fig 3) were applied for alignment analyses. The VP1 nucleotide sequence identity of CEV-JL14 was 64.7% to TB4-OEV, 51.8-53.5% to EV-E and EV-F, 30.9%-65.3% to EV-G, and 43.8-51.5% to EV-A-D. The P1 amino acid sequence identity of CEV-JL14 was 81.1% to TB4-OEV, 61.0-63.1% to EV-E and EV-F, 69.7-74.7% to EV-G, and 45.7-52.9% to EV-A-D ( Table 4). Analysis of the P1 amino acid sequence revealed higher amino acid sequence identity of P1 intraspecies for EV-E, EV-F and EV-G, respectively, where it was 86.2-96.7% among viruses within EV-E, 84.5-95.5% within EV-F, and 73.4-85% within EV-G, respectively. Those results showed that P1 sequence identity was much higher among viruses in the same species, and demarcated the CEV-JL14 and TB4-OEV from EV-E, EV-F, EV-G and other enterovirus species (EV-A-D).
To confirm the above results, the percentage of the G+C content of CEV-JL14, TB4-OEV, and representative EV-G were calculated. As shown in S1 Table, TB4-OEV had 48.75% G+C content, which had 2.5% percentage difference of G+C content over those viruses in the EV-G (45.45-46.25%), suggesting TB4-OEV should not listed as member of EV-G according to the criteria that G+C contents should be less than 2.5% difference if the enteroviruses were within the same species. The G+C content for CEV-JL14 is 48.26%, which also had 2.5% G+C content difference to PEV10 LP54 and PEV-B-KOR strain in EV-G. Those results, contradicting to the classification criteria, suggest that CEV-JL14 and TB4-OEV should be not grouped in EV-G although TB4-OEV was originally listed as member of EV-G. Instead, they should be grouped as a new enterovirus species.
To further ensure the above claim, the potential proteolytic sites for CEV-JL14, TB4-OEV, and viruses in EV-G were comparatively analyzed. As shown in S2 Table, 3 out of 10 potential proteolytic cleavage sites including VP4/VP2, P2B/P2C, P3A/P3B shared the same amino acid sequence for CEV-JL14, TB4-OEV and viruses in EV-G, while 6 out of 10 potential proteolytic processing sites shared the same proteolytic sequence between CEV-JL14 and TB4-OEV. Those results demonstrated the proteolytic sites are highly conserved between TB4-OEV and CEV-JL14, and highly variable among CEV-JL14, TB4-OEV and EV-G, indicating that CEV-JL14 and TB4-OEV were different from EV-G viruses.
Taken together, the above results demarcate CEV-JL14 from EV-E, EV-F, and EV-G and support our claim that CEV-JL14, together with TB4-OEV, should be grouped as new Enterovirus species, designated as EV-L within the genus of Enterovirus.
Sequence comparison between CEV-JL14 and TB4-OEV revealed 82.3% sequence identity for 5'-UTR, and 64.7% for VP1, respectively (Table 4). Those results suggest a high nucleotide sequence divergence even between sheep and goat enterovirus and indicate a different subgenotype for ovine enterovirus and caprine enterovirus, respectively.

Discussion
In this study, we have isolated and characterized a novel enterovirus from goats characterized by severe diarrhea with a high morbidity and mortality. Sequence analysis clearly distinguishes the CEV-JL14 from the EV-E, EV-F, EV-G, and other known enterovirus species based on the current criteria for enterovirus classification. Our phylogenetic analysis showed that CEV-JL14 clustered to neither EV-E, EV-F, EV-G nor other enterovirus species, it is clustered to a novel distinct clade together with TB4-OEV and Caprine enterovirus G1 from Japan and recently reported enterovirus in goats from Thailand. Pursuant to current criteria for classification of enterovirus species, where a range from 50 to 55% for heterologous species, 70 to 85% for heterologous serotypes/homologous species and greater than 90% for homologous serotypes [23], we proposed that CEV-JL14, together with TB4-OEV and other ovine/caprine enterovirus strains, was classified to a novel enterovirus species L (EV-L). This proposal is based on our following findings: 1) the nucleotide sequence identity of VP1 for CEV-JL14 was only 51.8-53.5% to EV-E and EV-F, 30.9%-65.3% to EV-G, and 43.8-51.5% to EV-A-D, which is well-matched with the criteria that 50-55% of nucleotide sequence belongs to heterogeneous species; 2) the deduced VP1 amino acid sequence were only 48.5%-53.3% to EV-E and EV-F, 57.0%-68.0% to EV-G, and 37.0-45.2% to EV-A-D, which is also consistent with the enterovirus species criteria; 3) CEV-JL14, together with TB4-OEV, was clustered to a distinct clade from EV-E, EV-F, EV-G, and EV-A-D based on the 5'-UTR sequence analysis and their secondary structure prediction. 4) The P1 amino acid sequence identity of CEV-JL14 was 81.1% to TB4-OEV, 61.0-63.1% to EV-E and EV-F, 69.7-74.7% to EV-G, and 45.7-52.9% to EV-A-D. 5) the percentage difference of G+C content for CEV-JL14 and TB4-OEV to EV-G is more than 2.5%, suggesting that CEV-JL14 and TB4-OEV should not listed as member of EV-G. 6) The proteolytic cleavage sites between CEV-JL14 and TB4-OEV is highly conserved, while they are highly variable between CEV-JL14/TB4-OEV and EV-G. Those results support our claim that CEV-JL14, together with other ovine/caprine enterovirus strains, ought to be considered as new species in the genus of Enterovirus. Although cladograms variations were revealed employing different capsid genes for the analysis, similar cladogram patterns were observed even under circumstance of the lack of enough virus isolates. With more and more caprine or ovine enterovirus sequences revealed, the clade will become much clearer.
TB4-OEV was the first ovine enterovirus isolated with its complete genome sequence revealed, and analysis of TB4-OEV sequence grouped this virus to EV-G, a Enterovirus species that caused swine infection mainly [5]. Our phylogenetic analysis clustered the TB4-OEV and the caprine enterovirus CEV-JL14 to a novel enterovirus named enterovirus L (EV-L) within the genus of Enterovirus. The discrepancy between our analysis and previous analysis is likely due to either lack of ovine/caprine enterovirus sequences or the majority of goat enterovirus sequences was just a partial 5'-UTR sequences [5,19]. To our knowledge, this is the first enterovirus isolate from goats in China and first caprine enterovirus with its complete genomic sequence revealed, which will enrich the current understanding of enterovirus variation and evolution.
Although enteroviruses have been demonstrated to be the leading agents related to human respiratory, digestive and neurological disorder including the poliomyelitis, coxsackievirus infection, echovirus infection and hand, food, mouth diseases (HFD), the enterovirus infection in animals remains largely unexplored. The frequent isolation or detection of a large amount of enterovirus particles from animals characterized with respiratory and digestive diseases with high morbidity and mortality should arouse attention to this emerging infection. The isolation of enterovirus from goats with severe diarrhea and detection of the virus antigens in the tissue of the diseased goats suggest that the enterovirus likely play a crucial role in the outbreak characterized in this study. Future investigation will focus on unveiling the mechanism underlying the involvement of enteroviruses in the severe respiratory and digestive diseases in animals. The finding that sequence divergence between the caprine enterovirus and ovine enterovirus indicates that enteroviruses co-evolved with the host species, which is also an interesting topic for future investigation.
Supporting information S1 Fig. The 5'-UTR secondary structure of CEV-JL14. The 5'-UTR secondary structure of CEV-JL14 was predicted using minimal free energy with the methods described previously [27]