Genome analysis of canine astroviruses reveals genetic heterogeneity and suggests possible inter-species transmission

Abstract

Canine astrovirus RNA was detected in the stools of 17/63 (26.9%) samples, using either a broadly reactive consensus RT-PCR for astroviruses or random RT-PCR coupled with massive deep sequencing. The complete or nearly complete genome sequence of five canine astroviruses was reconstructed that allowed mapping the genome organization and to investigate the genetic diversity of these viruses. The genome was about 6.6kb in length and contained three open reading frames (ORFs) flanked by a 5' UTR, and a 3' UTR plus a poly-A tail. ORF1a and ORF1b overlapped by 43 nucleotides while the ORF2 overlapped by 8 nucleotides with the 3' end of ORF1b. Upon genome comparison, four strains (HUN/2012/2, HUN/2012/6, HUN/2012/115, and HUN/2012/135) were more related genetically to each other and to UK canine astroviruses (88-96% nt identity), whilst strain HUN/2012/126 was more divergent (75-76% nt identity). In the ORF1b and ORF2, strains HUN/2012/2, HUN/2012/6, and HUN/2012/135 were related genetically to other canine astroviruses identified formerly in Europe and China, whereas strain HUN/2012/126 was related genetically to a divergent canine astrovirus strain, ITA/2010/Zoid. For one canine astrovirus, HUN/2012/8, only a 3.2kb portion of the genome, at the 3' end, could be determined. Interestingly, this strain possessed unique genetic signatures (including a longer ORF1b/ORF2 overlap and a longer 3'UTR) and it was divergent in both ORF1b and ORF2 from all other canine astroviruses, with the highest nucleotide sequence identity (68% and 63%, respectively) to a mink astrovirus, thus suggesting a possible event of interspecies transmission. The genetic heterogeneity of canine astroviruses may pose a challenge for the diagnostics and for future prophylaxis strategies.

Keywords: Astroviridae; Dog; Hungary; Semiconductor sequencing.

Fig. 1

Distribution of eukaryotic viral sequences in 26 fecal specimens collected from sheltered dogs. The numbers of virus-specific reads are indicated above the columns.

Fig. 2

Schematic representation of the genome of canine AstV strain HUN/2012/115. The upper line represents the complete nt length of the genome from the 5′ untranslated region (UTR) through the 3′ UTR and the poly-A tail (pA). The relative position of the slippery sequence AAAAAAC and downstream stem loop structure is indicated. Also, the position of the subgenomic RNA (sgRNA) promoter and of the stem loop-II-like (s2m) structure are indicated. Rectangles represent the three possible open reading frames (ORFs) (light grey) and the alternate ORF1b and ORF2 products due to in-frame start codons (dark grey). The starting aa motifs of the various ORFs and the ending aa motif of ORF2 are marked. The potential trans-membrane regions (TM), the predicted protease motif (PRO), the genome-linked protein (VPg) and the polymerase region (POL) are also shown.

Fig. 3

Phylogenetic trees based on the full-length ORF1a (3A), ORF1b (3B) and ORF2 (3C) of canine AstVs with a selection of mammalian AstVs. The tree was generated using the neighbor-joining method, with the Jukes Cantor algorithm of distance correction and bootstrapping over 1000 replicates. The scale bar indicates the number of nt substitutions per site. Black circles indicate the AstV strains detected in this study.

Fig. 3

Phylogenetic trees based on the full-length ORF1a (3A), ORF1b (3B) and ORF2 (3C) of canine AstVs with a selection of mammalian AstVs. The tree was generated using the neighbor-joining method, with the Jukes Cantor algorithm of distance correction and bootstrapping over 1000 replicates. The scale bar indicates the number of nt substitutions per site. Black circles indicate the AstV strains detected in this study.

Fig. 3

Phylogenetic trees based on the full-length ORF1a (3A), ORF1b (3B) and ORF2 (3C) of canine AstVs with a selection of mammalian AstVs. The tree was generated using the neighbor-joining method, with the Jukes Cantor algorithm of distance correction and bootstrapping over 1000 replicates. The scale bar indicates the number of nt substitutions per site. Black circles indicate the AstV strains detected in this study.