Genotyping coronaviruses associated with feline infectious peritonitis

Abstract

Feline coronavirus (FCoV) infections are endemic among cats worldwide. The majority of infections are asymptomatic or result in only mild enteric disease. However, approximately 5 % of cases develop feline infectious peritonitis (FIP), a systemic disease that is a frequent cause of death in young cats. In this study, we report the complete coding genome sequences of six FCoVs: three from faecal samples from healthy cats and three from tissue lesion samples from cats with confirmed FIP. The six samples were obtained over a period of 8 weeks at a single-site cat rescue and rehoming centre in the UK. We found amino acid differences located at 44 positions across an alignment of the six virus translatomes and, at 21 of these positions, the differences fully or partially discriminated between the genomes derived from the faecal samples and the genomes derived from the tissue lesion samples. In this study, two amino acid differences fully discriminated the two classes of genomes: these were both located in the S2 domain of the virus surface glycoprotein gene. We also identified deletions in the 3c protein ORF of genomes from two of the FIP samples. Our results support previous studies that implicate S protein mutations in the pathogenesis of FIP.

Fig. 1.. Genomic organization of FCoV. Genomic ORFs are shown as boxes. Only pp1ab is shown as a translation product of the genomic RNA. The non-structural proteins nsp1–11 are translated from ORF1a (dark grey shading) and translation of the ORF1b proteins (nsp12–16) occurs following −1 ribosomal frameshifting (RFS). The nsp11 protein is not depicted as it represents a short (9 aa) carboxyl extension of nsp10. nsp9, ssRNA-binding protein (ssRBP); nsp12, RNA-dependent RNA polymerase (RdRp); nsp13, helicase (Hel) and NTPase; nsp14, 3′→5′ exoribonuclease (ExoN) and N7-methyltransferase (N7-MT); nsp15, uridylate-specific endonuclease (NendoU); nsp16, 2′-O-methyltransferase (2′-OMT).

Fig. 2.. Agarose gel electrophoresis of PCR amplicons 1–7 (lanes 1–7, respectively) for tissue lesion samples 26M, 27C and 28O and faecal samples 65F, 67F and 80F. The 1 kbp DNA plus Ladder (Invitrogen) was used as molecular size markers and the 1, 2, 3 and 6 kbp fragments are indicated.

Fig. 3.. Coverage of sequence reads across the assembled FCoV genomes from faecal and tissue lesion samples. Sequence reads were aligned against the de novo assembled 80F target genome for the faeces-derived samples 80F, 65F and 67F and the tissue lesion-derived samples 26M, 27C and 28O.

Fig. 4.. Phylogenetic analysis of the core RdRp domain of nsp12 (aa 4503–4807 in pp1ab, Supplementary Fig. S1 available in the online Supplementary Material) for FCoV strains sequenced in this study and selected FCoV genome sequences. The phylogenetic tree was reconstructed by the neighbour-joining method from an alignment made with clustal w (MacVector). GenBank accession numbers are shown for all sequences. Bootstrap values exceeded 60 % at all nodes. Bar, nucleotide substitutions per site.

Fig. 5.. Sequence assembly workflow for FCoV genomes from faecal samples 65F, 67F, 80F and tissue lesion samples 26M, 27C, 28O.