RESULTS AND DISCUSSION
Of the 158 tested serum samples from dogs, 17.7% (n=28) were positive for P1, 16.5% (n=26) for PCVL258, and 8.9% (n=14) for PCVL264 based on PCR. In 41 samples from cats, 17.1% (n=7) were positive for P1, 9.8% (n=4) for P4, and 14.6% (n=6) for PCVL258. No PCV2, PCV3, and PCV4 strains were identified in the PCR assays (data not shown). The results also revealed a co-infection of P4 and PCVL258 in a cat.
To further study the molecular epidemiology of these agents, we obtained the full-length genomic sequences from 18 strains, representing 4 P1, 3 PCVL258, and 2 PCVL264 strains in dogs, and 5 P1, 2 P4, and 2 PCVL258 strains in cats, all of which have been deposited in GenBank (accession nos. MH379143, MH379144, MT318811-MT318826). All 9 P1 viruses in this study had a circular genome length of 648 nt and shared 96.8 to 100% nucleotide identity with other available P1 genomes in GenBank. Compared with the nucleotide sequence of the prototype P1 strain (GenBank EF514716), the dog-origin P1 strain NJ01 (MH379143) and the cat-origin P1 strain NJ03 (MT318822) were similar to the prototype elongated ORF1 strain (KU323639) and had a nucleotide substitution at position 77 (T to C), which would result in a codon shift of the putative ORF1 starting at nt position 419 and a stop codon at aa position 122. The ORF1 sequence of other strains in this study had a stop codon at aa position 114. Alignment of the whole genomic sequences of the 21 P1 strains showed that there were multiple point or region mutations, for example, variations in 6 positions appeared specific for the cat-origin strain HeB02 (MT318824), including residues 331 (A to T), 337 (T to C), 344 (C to T), 371 (T to G), 376 (G to T), and 379-383 (CCCCC to TTGTT). The 2 P4 cat-origin strains NJ01 and NJ02 had a genome length of 710 nt. The nucleotide sequence of the cat-origin strain NJ01 was identical to that of NJ02. The overall nucleotide identity of strains NJ01 and NJ02 in the whole genome to other pig-origin strains ranged from 98.3% (MF802838) to 100% (MF802836). Whether of pig-origin or not, all PCVL258 and PCVL264 sequences were identical.
A corresponding phylogenetic tree (including 12 reference P1 sequences, 2 reference P4 sequences, 1 reference PCVL258 sequence, and 1 reference PCVL264 sequence) was constructed based on the full-length sequence. Phylogenetic analysis suggested that all P1 sequences from this study were scattered within the P1 viral lineage except one cat-origin strain (MT318824), the sequence of which showed more variation and was located in a separate branch. All P1 strains in this study were more closely related to other PCVL258 and PCVL264 strains than to P4 strains (Figure 1).
PCV2 is responsible for several multifactorial disease syndromes (especially for postweaning multisystem wasting syndrome (PMWS) in pigs. Furthermore, PCV2 can be transmitted across species barriers, and may cause severe fatal disease in other species (such as mink) (Wang et al., 2016). In addition, porcine circovirus-like agents, newly emerging causative agents in pigs, have a close evolutionary relationship with PCV2, based on phylogenetic analysis of genomic DNA sequences. The P1 viruses detected in dogs and cats in our study had high homology with P1 strains from pigs, goats, cattle, and rabbits.