Identification and Characterization of the First Equine Parainfluenza Virus 5

Dear Editor,

Parainfluenza virus 5 (PIV5), known as canine parain-fluenza virus in the veterinary field, is a negative-sense, nonsegmented, single-stranded RNA virus belonging to the Paramyxoviridae family (Chen 2018). The virus was first reported in primary monkey kidney cells in 1954 (Hsiung 1972), then it has been frequently discovered in various hosts, including humans, dogs, pigs, cats, rodents, calves, and lesser pandas (Chatziandreou et al. 2004; Lee and Lee 2013; Liu et al. 2015; Zhai et al. 2017; Jiang et al. 2018). So far, PIV5 has not been reported in horse. In this study, using metagenomics analysis we have identified a novel equine PIV5.

Totally 148 fecal samples, 81 sera samples, and 115 nasal swabs were collected in six different cities of Xin-jiang and Inner Mongolia, China in 2018 (Supplementary Figure S1). First, Fresh fecal samples were collected from 20 thoroughbred horses (mean 5.6 years) at an equestrian club in Hutubi County, north Xinjiang, China. The cDNA library of fecal specimens was prepared as previously described (He et al. 2013). Briefly, viral RNAs were extracted from fecal pooled samples, then reverse tran-scribed, and randomly amplified using PCR. The tagged and purified PCR products were subjected to Illumina sequencing (HiSeq X-ten, United States, Illumina) in a lane by Shanghai Personal Biotechnology Co., Ltd (Shanghai, China). In total, 76, 986, 684 reads were generated, of which 197, 056 (0.26%) being annotated to mammalian viruses. Sequence analysis showed that 146 reads were closely related to human PIV5, sharing 99.8%–100% identity with the fusion protein (F) gene of AGS strain (GenBank no. KX060176).

To confirm the next-generation sequencing results, a pair of primers amplifying the equine PIV5 F gene were designed (Supplementary Table S1). RT-PCR results showed that eight fecal samples from 20 thoroughbred horses of the same equestrian club were positive for equine PIV5, indicating that the virus was likely circulating in the thoroughbred horses. Further studies revealed that eight fecal samples from 30 Yili horses in Yining city and five fecal samples from 11 Asian wild horses in Urumqi were positive for F gene (Table 1). However, the F gene was not detected in Yanqi, Sanhe and Wushi horses (Table 1). Further investigation found that Yili and Asia wild horses were co-bred with thoroughbred horses, but Yanqi, Sanhe and Wushi horses were not co-bred with thoroughbred horses. This suggested that equine PIV5 might be an exotic virus with high prevalence and different geographic dis-tribution in north Xinjiang. In addition, no F gene was detected in equine sera and nasal swabs of all samples (Table 1), indicating intestinal cell tropism of equine PIV5. Sequence comparison showed that the F genes of all 21 equine PIV5 shared 100% nucleotide identity.

The whole genomes of equine PIV5 were successfully obtained by RT-PCR using primers listed in the Supple-mentary Table S1. The obtained genome sequence is 15, 246 nt and consists of genes encoding nucleocapsid protein (NP), V protein (V), membrane protein (M), fusion protein (F), hemagglutinin-neuraminidase protein (HN) and L protein (L) (GenBank no. MN604146). Sequence analysis indicated that the novel equine PIV5 shared 98.2%–99.9% homology with six Human PIV5, and 95.6%–98.0% homology with the animal source PIV5 strains (Table 2), indicating that equine PIV5 was closely related to human PIV5.

Phylogenetic analysis based on the whole genome sequences showed that the novel equine PIV5 and six human PIV5 reference strains formed a single clade and were closely related to the AGS strain, but diverging from the lineages of animal source PIV5 strains (Fig. 1), indi-cating it may be a human-source virus. Unfortunately, our attempts to isolate the novel equine PIV5 using Vero cells and chicken embryos failed.

Figure 1.  Phylogenetic analysis of whole genome sequence of equine PIV5. The black circle indicates the virus identified in our study.

In summary, this study provides the first molecular evidence for equine PIV5. While our observations do not permit us to conclude that the virus associates with clini-cally symptomatic disease, we do indicate that equine PIV5 has a high prevalence and distributed in different geo-graphic areas in Xinjiang, one of the major horse-produc-ing regions in China. Further research is needed to develop new methods for virus isolation, and to investigate the serological epidemiology of PIV5 in horses in China.