Isolation and Characterization of Porcine Astrovirus 5 from a Classical Swine Fever Virus-Infected Specimen

Abstract

Many new astroviruses have been identified in humans and other animals in recent years, but only a few have been successfully isolated for extensive biological study. Here, we report an unusual isolation of a porcine astrovirus 5 (PAstV5) strain from a clinical classical swine fever virus (CSFV)-infected tissue sample. Incubation of porcine PK-15 cells with an extract of the CSFV-positive tissue resulted in unexpected cytopathic effects (CPEs), and high-throughput viromic sequencing identified PAstV5 and porcine circovirus type 2 (PCV2) as well as CSFV in the culture. After clearance of CSFV and PCV2, a pure PAstV5 strain, named PAstV5-AH29-2014, was obtained. Analysis revealed virus of typical astroviral morphology with a genome of 6,448 nucleotides, sharing 84.3 to 88.9% nucleotide identity with previously published PAstV5 strains. A mechanistic study showed that CSFV coinfection was likely an important factor for successful isolation by significantly enhancing PAstV5 replication in PK-15 cells via suppression of a type I interferon response. Altogether, PAstV5-AH29-2014, as the first isolated PAstV5 strain, will provide critical material for the investigation of the biological and pathogenic properties of this virus as well as for future development of relevant biological and diagnostic reagents.IMPORTANCE Porcine astroviruses are mainly associated with gastroenteritis and neurological diseases in pigs, and five genotypes have been identified (PAstV1-5). However, the clinical manifestations of genotypes other than PAstV1 have not yet been determined because of the failure of in vitro virus isolation. Here, we report a surprising isolation of a PAstV5 strain from a clinical classical swine fever virus (CSFV)-infected tissue sample, which can stably passage in PK-15 cells, and coinfection with CSFV significantly enhanced the replication of PAstV5, possibly through suppression of beta interferon production. Thus, the first isolated PAstV5 strain will be useful for investigating the biological and pathogenic properties of this virus, and the findings obtained in this study provide new insights into defining the interaction mechanism between CSFV and PAstV5.

Keywords: Porcine astrovirus 5; characterization; classical swine fever virus; isolation.

FIG 1

Phylogenetic analysis of PAstVs compared with representative PAstVs from other mammalian hosts. The tree was constructed based on complete capsid protein sequences by MEGA 7.0 with the best fit model (GTR+G+I) and 1,000 bootstrap replicates. All PAstV5 sequences available in GenBank were used, and PAstV5-AH29-2014, obtained in this study, is indicated by a black triangle.

FIG 2

Growth of PAstV5-AH29-2014 in PK-15 cells. (A) PAtV5-AH29-2014 can stably propagate in PK-15 cells, with its titers before purification being higher than those after removal of CSFV by viral neutralizing antibody treatment. (B) Purity detection of PAstV5-AH29-2014 stock by IFA. The upper row was detected by anti-PCV2 MAb 3C1, while the lower row was detected by anti-CSFV MAb WH303. The result showed that both PCV2 and CSFV were detected at passage 4 (P4), but only CSFV, and not PCV2, was detected at passage 21 (P21) of the coinfected culture, AH29-2014. In contrast, in purified PAtV5-AH29-2014 at P21, neither virus was detected. The CSFV positive control is strain JL23-2015, stored in our laboratory, while the PCV2 positive control was provided by Xinglong Yu, from Hunan Agricultural University. (C) CPEs started to show at 48 h and became severe at 96 h postinoculation, and no visible difference was observed in the purified (upper; passage 21) and the unpurified (middle; passage 4) PAstV5 cultures at different time points. (D) Growth dynamic of PAstV5-AH29-2014 at passage 21 was assessed by qRT-PCR (black curve) and TCID₅₀ assay by IFA (red curve). The IFA of PAstV5-AH29-2014 using antiserum against PAstV5 capsid protein is shown on the right. Bars represent the averages and standard deviations from three independent measurements.

FIG 3

Electron microscopy of PAstV5-AH29-2014 particles following purification by sucrose gradient (A) and in ultrathin cell sections (B).

FIG 4

Effect of CSFV coinfection on PAstV5 propagation (A) and IFN-β expression (B). (A) Coinfection of both viruses significantly enhanced PAstV5 titer (left) and had no effect on CSFV titers (right). (B) The production of IFN-β is enhanced by PAstV5-AH29-2014 but inhibited by CSFVs. Bars represent the averages and standard deviations from three independent measurements (**, P < 0.01; ***, P < 0.001; both by two-tailed Student's t test).

FIG 5

Inhibition of PAstV5 replication by swine IFN-β added into the culture is dose dependent (A) and can be alleviated by anti-swine IFN-β antibody (B). Rabbit IgG was used as a negative antibody control. Bars represent the averages and standard deviations from three independent measurements (**, P < 0.01; ***, P < 0.001; both by two-tailed Student's t test).