Detection and characterisation of canine astrovirus, canine parvovirus and canine papillomavirus in puppies using next generation sequencing

Abstract

Gastroenteritis in young animals is a clinical presentation with many infectious and non- infectious aetiologies. We used next generation sequencing (NGS) to investigate the possible infectious causes of gastroenteritis in puppies from a dog kennel in Victoria, Australia. The near complete genome of a canine astrovirus was obtained from pooled faecal samples, and was found to be 94.7% identical with a canine astrovirus detected in the United Kingdom in 2012. The phylogenetic analysis of the capsid gene found similarities to those of canine astroviruses identified in Italy in 2005 and in UK and Hungary in 2012, but distant from that of a canine astrovirus previously identified in Australia in 2012. Thus, different serotypes of canine astrovirus are likely circulating in Australia. The close relationship to European astroviruses also suggested that there had been recent movements of ancestor canine astroviruses between Australia and Europe. NGS also detected other infections in the puppies including several canine papillomaviruses and a canine parvovirus (vaccine strain) as well as a very low level of campylobacter. Canine astrovirus was the probable cause of diarrhoea in these puppies, with the possible involvement of campylobacter bacteria. NGS was effective as a non-targeted method to determine the likely infectious cause of gastroenteritis.

Figure 1

Phylogenetic analysis of the canine astrovirus near-complete genome. The nucleotide sequences were aligned and analysed using the maximum likelihood method in MEGA 7.0 using the General Time Reversible (GTR + G + I) model with a bootstrapping of 1000 replicates. The analysis involved 8 sequences of the canine astrovirus genome including canine astrovirus sequence DF-BC15-CAV-AUS-2017 from the dog samples. The numbers at nodes represent bootstrap values. Branch lengths are scaled according to the numbers of nucleotide substitutions per site.

Figure 2

Similarity plot generated in SimPlot using the canine astrovirus sequence (DF-BC15- CAV-AUS-2017) from the dog samples as the query sequence against seven other reference sequences from NCBI Genbank. Divergence between DF-BC15-CAV-AUS-2017 and each reference sequence over the 6612 nucleotides genome using a 200 nucleotides sliding window at 20 nucleotides intervals and the F84 distance model with the maximum likelihood method. Percentage identities at each analysis point were plotted on a line chart. For similarity plot analysis, the y-axis shows the percentage similarity between the parental sequences and the query sequence. Different colors represent a different reference sequence.

Figure 3

Phylogenetic analysis of the nucleotide sequence of the canine astrovirus ORF1a region. The nucleotide sequence identity of DF-BC15-CAV-ORF1a-AUS-2017 was analysed using the maximum likelihood method in MEGA 7.0 using the General Time Reversible (GTR + G) model and with a bootstrapping of 1000 replicates. The analysis involved 6 sequences of canine astrovirus ORF1a region. The numbers at each node represent the bootstrap values. Branch lengths are scaled according to the numbers of nucleotide substitutions per site.

Figure 4

Phylogenetic analysis of the amino acid sequence of canine astrovirus ORF1a region. The amino acid sequence of DF-BC15-CAV-ORF1a-AUS-2017 was analysed using the maximum likelihood method in MEGA 7.0 using the Jones-Taylor-Thornton (JTT + G) model and with a bootstrapping of 500 replicates. The analysis involved 6 sequences of the canine astrovirus ORF1a region. The numbers at nodes represent bootstrap values. Branch lengths are scaled according to the numbers of amino acid substitutions per site.

Figure 5

Phylogenetic analysis of the nucleotide sequence of canine astrovirus ORF1b region. The nucleotide sequence of DF-BC15-CAV-ORF1b-AUS-2017 was analysed using the maximum likelihood method in MEGA 7.0 using the Tamura-Nei model (T93 + G) model and with a bootstrapping of 1000 replicates. The analysis involved 9 sequences of the canine astrovirus ORF1b region. The numbers at nodes represent bootstrap values. Branch lengths are scaled according to the numbers of nucleotide substitutions per site.

Figure 6

Phylogenetic analysis of the amino acid sequence of canine astrovirus ORF1b region. The amino acid sequence of DF-BC15-CAV-ORF1b-AUS-2017 was analysed using the maximum likelihood method in MEGA 7.0 using the Jones-Taylor-Thornton (JTT + G) model and with a bootstrapping of 500 replicates. The analysis involved 9 sequences of the canine astrovirus ORF1b region. The numbers at nodes represent bootstrap values. Branch lengths are scaled according to the numbers of amino acid substitutions per site.

Figure 7

Phylogenetic analysis of nucleotide sequence of canine astrovirus ORF2 region. The nucleotide sequence of DF-BC15-CAV-ORF2-AUS-2017 was analysed using the maximum likelihood method in MEGA 7.0 using the General Time Reversible (GTR + G + I) model and with a bootstrapping of 1000 replicates. The analysis involved 14 sequences of the canine astrovirus ORF2 region. The numbers at nodes represent bootstrap values. Branch lengths are scaled according to the numbers of nucleotide substitutions per site.

Figure 8

Phylogenetic analysis of amino acid sequence of canine astrovirus ORF2 region. The amino acid sequence identity of DF-BC15-CAV-ORF2-AUS-2017 was summarized using the maximum likelihood method in MEGA 7.0 using the Jones-Taylor-Thornton (JTT + G) model and with a bootstrapping of 500 replicates. The analysis involved 14 sequences of the canine astrovirus ORF2 region. The numbers at nodes represent bootstrap values. Branch lengths are scaled according to the numbers of amino acid substitutions per site.

Figure 9

Similarity plot generated in SimPlot using the DF-BC15-CAV-ORF2-AUS-2017 sequence as the query sequence against thirteen other reference sequences. Divergence between DF-BC15-CAV-ORF2-AUS-2017 and the reference virus sequences over the 2505 nucleotides of ORF2 using a 200 nucleotides sliding window at 20 nucleotides intervals and the F84 distance model with the maximum likelihood method. Percentage identities at each analysis point were plotted on a line chart. For similarity plot analysis, the y-axis shows the percentage similarity between the reference sequences and the query sequence. Different colours are indexed for different reference sequences.