Identification and Characterization of Porcine Kobuvirus Variant Isolated from Suckling Piglet in Gansu Province, China

Kobuviruses comprise three species, the Aichivirus A, Aichivirus B, and Aichivirus C (porcine kobuvirus). Porcine kobuvirus is endemic to pig farms and is not restricted geographically but, rather, is distributed worldwide. The complete genomic sequences of four porcine kobuvirus strains isolated during a diarrhea outbreak in piglets in the Gansu province of China were determined. Two of these strains exhibited variations relative to the traditional strains. The potential 3C/3D cleavage sites of the variant strains were Q/C, which differed from the Q/S in the traditional porcine kobuvirus genome. A 90-nucleotide deletion in the 2B protein and a single nucleotide insertion in the 3′UTR were found in the variant strains. The VP1 regions of all four porcine kobuviruses in our study were highly variable (81%–86%). Ten common amino acid mutations were found specifically at certain positions within the VP1 region. Significant recombination sites were identified using SimPlot scans of whole genome sequences. Porcine kobuviruses were also detected in pig serum, indicating that the virus can escape the gastrointestinal tract and travel to the circulatory system. These findings suggest that mutations and recombination events may have contributed to the high level of genetic diversity of porcine kobuviruses and serve as a driving force in its evolution.

Porcine kobuvirus was first detected in fecal samples from domestic pigs in Hungary, in 2007 [4]. Recently, porcine kobuviruses have also been detected at a high frequency in stool samples from healthy pigs in Hungary, China [11], Japan [12], Thailand [13], Brazil [14], The Netherlands [14], and Korea [15]. In addition, the prevalence of these viruses in pigs with diarrhea was shown to be very high in China during the last two years. Since December, 2010, there has been a large-scale outbreak of diarrheal disease in suckling piglets, characterized by watery diarrhea, dehydration, and vomiting. This outbreak led to 80% morbidity and 60% mortality in affected piglets in several farms in the Gansu Province of China. A recent report suggested that porcine kobuvirus infection correlated with diarrhea in pigs [15]. Porcine kobuviruses are a newly recognized viral agent in animals. Little is known about this virus, and only eight porcine kobuvirus genome sequences are available in GenBank. For this reason, it is important to increase the awareness of this disease and to stimulate further clinical, epidemiological, and molecular studies regarding porcine kobuviruses. In this study, we report four complete genome sequences and describe the detailed genomic organization of two porcine kobuvirus variant strains.

Prevalence and Complete Genomic Sequences of Porcine Kobuviruses
Overall, five (31.3%) of 17 fecal samples and two (33.3%) of six serum samples tested positive for kobuvirus infection using Kv primers. A total of 11 overlapping cDNA clones spanning the entire genome of K-11/2012/CH, K-4/2012/CH, GS-1/2012/CH, and GS-2/2012/CH were obtained, and their nucleotide sequences have been deposited into the GenBank database. Their accession numbers are KC414936, KC424638, KC424639, and KC424649, respectively. The complete RNA genomes of the K-11/2012/CH and K-4/2012/CH strains were similar to the sequence of the Hungarian strain, which consists of 8210 nucleotides (nt), excluding the poly (A) tail. A large open reading frame (ORF) of 7467 nt encoding a potential polyprotein precursor of 2488 amino acids (aa) was found. This ORF was located at the 5′ end and was preceded by 576 nt; 167 nt followed the ORF prior to the poly (A) tail. Relative to the genomes of the traditional strains, the genomes of the GS-1/2012/CH and GS-2/2012/CH variant strains contained 8121 nt, excluding the poly (A) tail; a 90-nucleotide deletion in the 2B protein and a single nucleotide insertion in the 3′UTR were also observed ( Table 1).

Analysis of the UTRs and Coding Regions of the Four Strains
An alignment of the sequences of the four genomic sequences with porcine kobuvirus strain S-1-HUN, revealed that the genome sequences of K-4/2012/CH, K-11/2012/CH, GS-1/2012/CH, and GS-2/2012/CH shared 89%, 86%, 88%, and 87% identity to the S-1-HUN at the nucleotide level, respectively ( Table 1). The highest nucleotide identities (96% and 98%) between the variant strains and the S-1-HUN strain were observed in the 3′UTR (Table 1), and the lowest nucleotide identities (76% and 87%) between the variant strains and the S-1-HUN strain were observed in the 2B coding region (Table 1). K-4/2012/CH and K-11/2012/CH polyprotein sequences were analyzed for potential cleavage sites, and the sites were identical to those in theS-1-HUN strain (Table 1). However, the potential 3C/3D cleavage site for the GS-1/2012/CH and GS-2/2012/CH variant strains was Q/C, which differed from the Q/S sequence in the traditional porcine kobuvirus strains.

Recombination Analysis of the Variant Stains
The four complete kobuvirus genome sequences were analyzed using Simplot. The similarity plot displays the consecutive nucleotide identity (%) among the queried strain and parental strains. The

Analysis of Coding Region Amino Acid Sequences
The structural and non-structural protein regions of the both K-4/2012/CH and K-11/2012/CH strains contain 2329aa and 4353aa, respectively. The amino acid identity identities among the of structural regions (VP0-VP3) are 90%-97%, and the non-structural regions (2A-3D) share 93%-100% amino acid identity with that of the S-1-HUN strain ( Table 2). Compared with the K-4/2012/CH and K-11/2012/CH strains, the variant strain has a 30-animo acid deletion within the 2B protein. The amino acid identities among the structural regions (VP0-VP3) are 89%-97%, and the non-structural regions (2A-3D) share 90%-100% amino acid identity with that of the S-1-HUN strain ( Table 2). The 2B amino acid sequence alignment for the 12 strains analyzed in this study and the eight strains from the GenBank database (K-30-HUN/2008 and S-1-HUN are Hungarian strains; others are Chinese strains) is shown in Figure 2A. Ten amino acid mutations were observed at positions 7, 11, 15, 28, 29, 30, 33, 37, 41, and 44. Though the function of the porcine kobuvirus 2B protein is not known, protein 2B has been shown to assume different functions in different kobuviruses [15].  VP1 is highly exposed and is the most immunodominant part of the kobuvirus capsid protein [3], and this protein is the most variable among the structural proteins [16]. Eight VP1 regions were analyzed in this study, and more than three amino acid mutations were found specifically at

Phylogenetic Analysis
We compared the entire genome sequences of K-4/2012/CH, K-11/2012/CH, GS-1/2012/CH, and GS-2/2012/CH with those of other kobuviruses. A phylogenetic tree based on whole genome sequences indicted that all four of the study sequences were placed in the porcine kobuvirus genetic clade, which was separate from the other kobuviruses (   The sequences of the 3D region have been particularly useful for placing viruses within species or genera and for comparing viruses of different genera or families [17]. A phylogenetic analysis shows two genetic lineages of porcine kobuviruses, however, these viruses are phylogenetically distinct from both bovine kobuviruses and Aichi viruses based on partial sequences of the 3D region ( Figure 4). The four porcine kobuviruses isolated in our study and the Hungarian strains were in the same lineage, and the South Korean strains were placed in another lineage. Figure 4. Phylogenetic tree of porcine kobuviruses based on the sequences from the 3D region (588 nt). The phylogenetic tree was constructed using the neighbor-joining clustering method; distance was calculated using the maximum composite likelihood correction for evolutionary rate in MEGA version 5.0. Bootstrap values (based on 1,000 replicates) for each node are provided when >50%.

Viral Culture
No cytopathic effects (CPE) were observed in Vero, BHK-21 (Baby Hamster Kidney), or PK-15 (Pig Kidney) cells following serial passage of cultures. Porcine kobuvirus RNA replication were not detected in the cells and supernatants in every passage were collected.

Fecal and Serum Samples
Seventeen fecal and serum samples were collected in July, 2012, from dead piglets; these piglets showed signs of watery diarrhea and dehydration for <10 days and were from two farms (located in Jiayuguan and Linxia City) in Gansu Province. Six serum samples from a one-year-old pig in Jiuquan City within Gansu Province were also obtained. Fecal samples were prepared as 10% homogenates in PBS and then centrifuged at 10,000 ×g for 20 min. The resultant supernatants were applied to Vero, PK, or BHK cells prior to RT-PCR analysis.

RT-PCR, Cloning, and Sequencing
Total RNA was extracted from fecal and serum samples using TRIzol reagent according to the manufacturer's instructions (Invitrogen Life Technologies, Grand Island, NY, USA). RNA was reverse transcribed with avian myeloblastosis virus reverse transcriptase (Promega, Beijing, China) in the presence of random primersand oligo dT15 (TaKaRa Bio Inc., Otsu, Japan). The cDNA was immediately amplified with primers that were designed based on the sequences of kobuvirus reference strains ( Table 3). The 5′ and 3′ ends of the genome were confirmed using a SMARTer rapid amplification of cDNA ends (RACE) cDNA amplification kit (Clontech, Japan). PCR reactions contained 1 µL of FastPfu DNA Polymerase (TransGen Biotech, Beijing, China), 10 µL of 5× FastPfu Buffer, 5 µL of dNTP mixture (2.5 mM), 1 µL of each primer, 2 µL of template, and sterile H 2 O to yield a final volume of 50 µL. The PCR conditions were as follows: 94 °C for 5 min; 35 cycles of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 1 min; and a final extension at 72 °C for 10 min. Amplicons were separated in a 1.5% agarosegel. Table 3. Primers designed from contigs to acquire the porcine kobuvirus genomes.

Sequencing and Phylogenetic Analysis
PCR products were cloned into the pMD-18 T vector (TaKaRa, Otsu, Japan). Competent E. coli DH5a cells (TaKaRa) were transformed with the recombinant plasmid. Three clones were sequenced using the Applied Biosystems (ABI) 3730xl DNA analyzer. Sequences were analyzed using the ClustalX version 2.0 [18], DNASTAR, and MegAlign version 5.0 (DNAStar Inc., Madison, WI, USA) software packages. A dendrogram was constructed using the neighbor-joining clustering method in MEGA 5.0 [18]. Possible recombination events in the coding region were analyzed using Simplot software [20]. All kobuvirus strains used in the study are listed in Table 2.

Viral Culture
Original fecal samples that tested positive for porcine kobuvirus by RT-PCR were propagated separately for virus isolation. Vero, PK-15, and BHK-21 cells were used as general-purpose cell lines for propagating picornaviruses.

Conclusions
The first porcine kobuvirus was identified in Hungary in 2007. Since then, porcine kobuviruses have been reported in several Asian countries. Porcine kobuviruses are not geographically restricted and are widely distributed in pigs worldwide. Porcine kobuviruses were also detected at a high frequency in pigs without clinical diarrhea [16]. Here, we report the complete nucleotide sequences and genetic organization of two traditional strains (K-4/2012/CH and K-11/2012/CH) and two variant strains (GS-1/2012/CH and GS-2/2012/CH).
The complete genome sequences of four strains were compared to the prototype porcine kobuvirus strain (S-1-HUN). The genomes of the K-11/2012/CH and K-4/2012/CH strains were similar to that of the Hungarian strain, which consists of 8210 nucleotides. However, the variant strains, GS-1/2012/CH and GS-2/2012/CH, contain 8121 nucleotides, with a 90-nucleotide deletion within the 2B protein and a single nucleotide insertion within the 3′UTR. Though the function of the porcine kobuvirus 2B protein is unknown, the 2B protein has many different functions in different kobuviruses [21]. In our study, ten amino acid mutations were observed in the 2B protein-coding region, though further studies are required to address whether the variant strains cause serious diarrheal outbreaks in pigs.
The VP1 regions of the four porcine kobuviruses in our study were highly variable [22]. The overall nucleotide identity among the VP1 sequences was 81%-86% (Table 1). Ten amino acid mutations were consistently observed at certain positions within the VP1 region. The mechanisms responsible for the evolution of the viral RNA sequences remain largely unknown. Host, viral, and environmental factors may play roles in this process [23].
Natural recombination has been reported in most picornavirus genera, and it is a key genetic feature of these infectious agents [24]. By exchanging sequences, the virus may induce changes in its genome sequence. This may lead to the acquisition of many key adaptive mutations [25], thus generating genetic variation and even producing new viral variants. One striking difference between porcine kobuvirus and the members of the other two kobuvirus species is that the 5′UTR of porcine kobuvirus contains an IRES related to that of hepatitis C virus [26,27] , whereas the 5′UTRs of the other species contain IRESs that are structurally and mechanistically unrelated [28,29]. In our study, significant recombination signals were identified using SimPlot scans of whole genome sequences. However, kobuvirus recombination cannot be discussed without also considering other evolutionary properties, such as global prevalence and its dynamic epidemiology.
Kobuviruses are transmitted primarily by the fecal-oral route, and other picornaviruses may use a similar mechanism [16]. Porcine kobuviruses are thought to be confined to the intestine, but porcine kobuvirus viremia has also been recently reported [5]. We also detected porcine kobuvirus sequences in pig serum. This finding suggests that porcine kobuvirus can escape the gastrointestinal tract and travel to the circulatory system. In addition to fecal-oral transmission, further experiments are necessary to investigate the possibility of kobuvirus transmission through breastfeeding and via blood-borne infections.
In summary, porcine kobuviruses are a newly recognized viral agent in animals. Porcine kobuviruses maybe persist for long periods because this viral infection can be asymptomatic. The presence of mutations and recombination events may contribute to the high levels of genetic diversity in porcine kobuviruses. The genomic changes and viremia may make this virus a potentially dangerous disease-causing agent in general, particularly in humans.