Our findings indicate that PKoV-infected pigs are not restricted geographically but distributed worldwide regardless of clinical conditions. In the present study, we determined PKoV infection status and prevalence in healthy pigs and in those suffering diarrhea from Sichuan Province, China. PKoV was detected in 17/18 sampled cities. Around 53% (87/163) of samples contained PKoV, with 64.3% (72/112) of diarrhea samples and 29.4% (15/51) of normal samples containing the virus. These results indicate general circulation and endemic infection of PKoV in Sichuan domestic pigs. We also confirmed that PKoV is common in apparently healthy pigs, which agrees with the findings presented by researchers in other countries. The infection rate in healthy pigs (29.4%) was similar to that seen in Lulong County (30.12%), slightly higher than that for Shanghai (22.4%) , and the United States (21.7%) , and much lower than that for Hungary (65%) . In pigs with diarrhea, prevalence was 64.3% and in accordance with that previously reported in Shanghai (61.2%) , and much lower than that observed in Thailand (97–99%) [21, 28], Brazil (78.4%)  and Korea (84.5%) . These differences in prevalence can be largely attributed to sampling time, sampling range, fecal consistency, and age of evaluated pigs.
Statistical analysis of PKoV incidence suggests that PKoV infection correlates with diarrhea (χ
2 = 17.126, p = 3.5 × 10-5). Similar results have been reported in Shanghai (p = 0.000), Brazil (p = 0.0002) and Korea (p = 3.2 × 10-17) using Pearson’s chi-square test [19, 29, 31]. However, it was not possible to conclusively show that PKoV was the etiological agent of diarrhea. The existence of other pathogens that can cause diarrhea could not be ruled out. In three diarrhea samples from Korea that tested positive for PKoV, other enteric pathogens were not detected . There was a high prevalence of virus in pigs with diarrhea from Thailand (97–99%) [21, 28] and Korea (84.5%) . These observations imply that PKoV might have some association with diarrhea in pigs. The lack of a cell culture system to propagate PKoV in vitro limits further study regarding the biology and pathogenicity of this emerging virus.
Among the four tested age groups, piglets under the age of 4 weeks were more likely to be infected with PKoV (χ
2 = 10.941, p = 9.4 × 10-4), which is similar to what has been reported in Hungary, Shanghai, the United States and Brazil. These reports indicate that young piglets are highly susceptible to PKoV infection [19, 25, 27, 31]. A study conducted by Barry et al. indicated that this might be possibly be due to an inefficient immune response or other intrinsic age-related factors .
Prevalence of PKoV in suckling, weaned, and growing/finisher pigs decreased as host age increased. Similar patterns have been observed in swine herds from Hungary, Korea, the United States, Brazil, and Japan [23–25, 27, 29]. In the Czech Republic, higher PKoV prevalence in samples from post-weaning pigs and nursing piglets were seen compared with those obtained from an abattoir . Therefore, PKoV infection might have an association with host age. Further epidemiological studies from other geographical areas will be required to clarify this.
The overall frequency (40.0%) of PKoV in sows is much lower than that reported by Dufkova et al. (90.9%), but much higher than that reported by Barry et al. (11.8%) [19, 26]. According to these researchers, unproductive infection and passive shedding of the virus might result in the low prevalence among sows living in the same breeding environment. However, these conditions can lead to a high prevalence in piglets, as infected sows might act as a reservoir of PKoV and cause continuous infection of piglets.
Kobuvirus was not detected in any of the Tibetan pigs we sampled. It is possible that PKoV infection is not as frequent in Tibetan pigs compared with that observed in domestic pigs. However, as only four pigs were examined it is more likely that this number is not representative of the actual infection status.
The VP1 sequences identified in Sichuan Province formed four large clusters, suggesting multiple PKoV strains are circulating in Sichuan Province. Previous studies have already revealed the existence of multiple PKoV lineages in China and Korea [21, 24, 31]. Our phylogenetic analyses confirmed high levels of genetic diversity for the VP1 gene, which has been previously reported [28, 32, 40].
We found that two pigs (pigs 31 and 32) were co-infected with multiple PKoV strains. Three different PKoV strains were found in pig 31, and two different PKoV strains in pig 32. This is the first report, with supporting evidence, of multiple strains of PKoV co-infecting a single pig. On a Brazilian farm, different PKoV strains were suspected to exist in a serum sample from an individual pig, but only one strain was detected. Identification of the strain was based on a gene that was more conserved than that for VP1. The authors of this report suggested it was likely that there were indeed two different strains and those they were not only a result of constant changes in the RNA genome of kobuviruses . CHN/SC/31-A1, CHN/SC/31-A2 and CHN/SC/31-A3 (obtained from pig 31) formed a cluster. CHN/SC/32-B1 and CHN/SC/32-B2 belonged to different branches, sharing 84.5% nucleotide sequence identity. This would suggest genetic diversity of PKoV within the same host and between different hosts. This is also the first report describing marked sequence diversity of PKoV within the same host.
Recombination plays an important role in the evolution of virus genomes. It is a major driving force for the generation of new genotypes or strains of virus. Significant recombination breakpoints were observed in the polyprotein gene sequences for pig 31. Two parent strains simultaneously infecting one host is a prerequisite for recombination. Co-infection of multiple PKoV strains in the same pig may potentially facilitates the occurrence of recombination events. Possible recombination events were analyzed for CHN/SC/31-A2; these might have been generated from recombination between CHN/SC/31-A1 and CHN/SC/31-A3 in the VP1 region. Recombination events have contributed to the genetic diversity within hosts that we observed. Multiple PKoV strains in the same pig could have arisen from recombination events .
Wang et al. used Simplot for genetic analysis of PKoV strains; however, no significant recombination events were identified in SH-W-CHN, the strain they investigated. Certain possible recombination signals were identified in a small region (nt 8083–8210) at the 3′ end of the viral genome . BootScan results and phylogenetic analysis of five complete Aichi virus sequences revealed a mosaic genome sequence of Aichi virus . Phylogenetic analysis of the VP1 region and 3D region of strain H023/2009/JP suggest it may be a natural recombinant from porcine and bovine kobuviruses [23, 28]. Recombination in kobuviruses is likely to be a usual phenomenon, just as it is in other members of the Picornavirus genus .
PKoV infection is widely distributed in healthy pigs and asymptomatic pigs, providing favorable conditions for recombination . The pathogenic and zoonotic potential of PKoV remains unclear. The closely related Aichi virus is the causative agent of human gastroenteritis, and bovine kobuvirus is associated with diarrhea in cattle [3, 12, 17, 44–46]. It would appear that the pathogenesis of PKoV is similar to that for other picornaviruses, and we believe that it may have a major role in causing enteric diseases in swine.
Since the end of 2010, massive outbreaks of diarrhea have occurred in suckling piglets in China; however, the etiological agent has yet to be determined. Affected pigs exhibited signs of watery diarrhea, dehydration, and vomiting with morbidity ranging 80–100% and mortality between 50–90%. Of the suckling pigs we tested in this study, 76.5% (52/68) were suffering from diarrhea and positive for PKoV. A high frequency of PKoV in piglets with gastroenteritis has been observed in other countries. Therefore, we propose that PKoV is the likely etiological agent of these outbreaks of severe diarrhea in China that began in 2010.
Recombination in kobuviruses creates changes in virus genomes, which probably generates new virus variants. Cao et al. sequenced and analyzed the complete genome of a PKoV variant with a 30-amino acid deletion in the 2B-coding region and a threonine amino acid insertion in its VP1 region. This variant was isolated from the 2010 outbreak in China . It is possible that PKoV variants generated through recombination or other evolutionary forces are related to the large-scale outbreak of severe diarrhea in suckling piglets from Sichuan Province. Further research is required to determine the exact role of PKoV variants in swine disease.
Many emerging viruses are of zoonotic origin and cause epidemics in humans after overcoming the interspecies barrier through mutation or recombination events. Intertypic recombination in H023/2009/JP implies the possibility of cross-species transmission of kobuviruses [23, 28]. PKoV was not detected among 454 samples that were obtained from children with diarrhea in China . Given that the frequency of recombination events in RNA viruses is relatively high, the possibility of zoonotic transmission among kobuvirus cannot be excluded.
PKoV is prevalent in both healthy and diseased pigs, and we suspect that at least two different PKoV types exist in piggeries. One type likely leads to gastroenteritis in pigs and possibly acts in combination with other enteric viruses; the other type probably causes subclinical infections in animals. Similar speculation has been mentioned by Verma et al. They also speculated that PKoV pathogenicity could be related to virus load and the presence of other enteric viruses, or kobuvirus might just be an endogenous passenger virus . The difficulties in propagating PKoV in vitro limit our understanding of the growth kinetics and pathogenicity of this virus. Detection of PKoV in serum samples has been reported in Hungary and Brazil [19, 27]. Further extensive epidemiological investigation of different PKoV strains from various regions would facilitate the understanding of its clinical and epidemiological characteristics.