key: cord-0798898-ffex7dw9
authors: Escutenaire, S.; Isaksson, M.; Renström, L. H. M.; Klingeborn, B.; Buonavoglia, C.; Berg, M.; Belák, S.; Thorén, P.
title: Characterization of divergent and atypical canine coronaviruses from Sweden
date: 2007-05-29
journal: Arch Virol
DOI: 10.1007/s00705-007-0986-1
sha: 9d0887c03a8e79e16c1b6ed8760a4f47efd95125
doc_id: 798898
cord_uid: ffex7dw9

Field canine coronaviruses (CCVs) identified during a series of outbreaks of gastroenteritis in Swedish dogs were subjected to genetic analysis involving the open reading frame 1b (ORF1b) and the membrane (M) and spike (S) protein genes. Four field viruses originating from the Stockholm region presented identical sequences and segregated separately from other CCVs characterized so far and from GOT/05, the variant recovered in Western Sweden. A recombinant origin of the fifth virus identified in the Stockholm region is suggested. In addition, the five viruses originating from the same geographical area displayed atypical 5′ S gene sequences.

RNA of 27-31 kb in length. It contains in its 5 0 region two overlapping reading frames (ORFs), ORF1a and ORF1b, encoding polyproteins leading to the replicase complex. Downstream ORFs code for the structural proteins, i.e. spike (S), envelope (E), membrane (M) and nucleocapsid (N) proteins, respectively.

Canine coronaviruses (CCVs) responsible for gastroenteritis and feline coronaviruses (FCoVs) belong to the group 1 coronaviruses. FCoVs are subdivided into types I and II, differing by immunological properties and by ORF1b and S gene sequences [5, 13] . Both types include feline enteric coronaviruses, causative agents of unapparent to mild transient enteritis, and feline infectious peritonitis viruses, which are associated with fatal systemic disease [13] . FCoV type II was suggested to originate from homologous recombination between CCV and FCoV type I [4] . CCVs have also been differentiated into types I and II, based on genetic similarity with the respective FCoV types [9, 10] . CCV type II includes most canine reference strains, and CCV type I constitutes a new genotype represented by Italian viruses [8, 10, 11] . Both CCV types I and II have been associated with mild to moderate gastroenteritis in dogs with more severe clinical signs in young animals [7] .

In this study, we characterized field CCVs originating from a series of outbreaks of gastroenteritis in Swedish dogs. Sequences from three different genes were determined, including the highly variable 5 0 regions of the M and S genes.

Most gastroenteritis cases in dogs occurred in December 2004 and January 2005, in different regions of Sweden including the Stockholm area. The animals presented vomiting, inappetence and recurrent diarrhea for several weeks. Abundant haemorrhagic diarrhea was also recorded, and more severe clinical signs with occasional fatal outcome occurred in pups. The present analysis was based on six faecal samples that first tested positive for group 1 coronaviruses by real-time reverse transcriptionpolymerase chain reaction (RT-PCR) [3] . Of the six dogs, four originated from the Stockholm area: Å KERSBERGA=05 (AKBG=05), S € O ODERT € A ALJE= 05 (SDT=05), UPPSALA1=04 (UPPS1=04) and UPPSALA2=04 (UPPS2=04). Only two of these animals (UPPS1=04 and UPPS2=04) were epidemiologically related: they had the same owner, a veterinarian whose clinic recorded during the same period 20-30 cases of canine gastroenteritis. The fifth dog [G € O OTEBORG=05 (GOT=05)] was from Gothenburg, a town located in western Sweden. All five animals were adults. The sixth sample was collected from a 7-day-old pup (MORA=04) originating from central Sweden and which died after severe gastroenteritis. The pup was born to a bitch which also presented diarrhea and vomiting. The bitch had supposedly been infected a few days before in a kennel housing dogs that had participated in the Stockholm international dog show. Several gastroenteritis cases had been recorded during the exhibition. Viruses identified in this study will be given the same reference name as the dog from which they were recovered.

Total RNA was extracted from faecal homogenates by means of the QIAamp viral RNA kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Consensus degenerate primers targeting a region of the ORF1b and primers specific for CCV and FCoV M and S gene se-quences were selected. The reverse transcription and amplification were carried out as described in Ref. [2] . Sequences of the primers and cycling conditions are available upon request. DNA samples generated from two different RT-PCR runs were sequenced in both directions using the PCR primers. Longer PCR products amplified from the 5 0 S gene were cloned with the pGEM-T Cloning kit (Promega) and sequenced automatically with primers M13 Forward and M13 Reverse.

Editing and translation of sequences were done with the BioEdit package (version 7.0.1). The Clustal X program (version 1.83) was used for sequence alignment. Percentages of nucleotide and amino acid identity were determined with the MegAlign program (version 5.08) (DNASTAR Inc.). Maximum-likelihood trees were constructed from sequence alignments by means of the program package TREE-PUZZLE (version 5.2) and visualized with the program TreeView (version 1.6.6). Nucleotide positions refer to FIPV 79-1146 (Accession number: DQ010921) for the ORF1b sequences and to CCV INSAVC-1 (Accession number: D13096) for the M and S genes.

Sequences generated in this study were deposited in the GenBank database. The designations and GenBank accession numbers are as follows: Based on the genomic regions studied, AKBG= 05, SDT=05, UPPS1=04 and MORA=04 viruses displayed identical nucleotide sequences and were named SWE1 viruses. The six Swedish viruses shared higher identity with CCVs type II based on alignments of nucleotide and deduced amino acid 1508 S. Escutenaire et al.

sequences of the M gene ( Table 1 ). The SWE1 viruses displayed 96.3 and 96.0% nucleotide identity with UPPS2=04 and GOT=05, respectively (Table 1) . Inferred amino acid sequences were mostly divergent within the amino terminus of the M protein where the Swedish viruses presented one unique amino acid site (V instead of L or I; Fig. 1 ). In addition, the SWE1 viruses displayed one common unique amino acid deletion in the amino terminus ectodomain. These viruses differed from GOT=05 and UPPS2=04 by 8 and 9 amino acids, respectively, and by 2 amino acid deletions. Six of the amino acid mutations were non-conservative.

In the alignment based on 3 0 S gene sequences, the Swedish field viruses displayed higher ranges of nucleotide and amino acid identity with both CCVs and FCoVs of type II (Table 2 ). In contrast with data from the M gene, the 3 0 S regions of SWE1 viruses and UPPS2=04 were more closely related. In addition, SWE1 viruses and UPPS2=04 presented one common unique amino acid site (L instead of V) when compared with all CCVs type II and differed from GOT=05 by 5 amino acids. All five substitutions were conservative.

Amplification of the highly variable 5 0 region of the S gene failed in all but one sample (GOT=05) using CCV type II-specific primers [6] . By means of newly designed and broadly reactive primers, a successful amplification was obtained from all six Swedish field samples. Whereas a PCR product of the expected size (897 bp; nucleotide positions: 469-1365) was generated from GOT=05, the frag- In the phylogenetic tree based on partial ORF1b sequences (nucleotide positions: 13796-13982), the Swedish viruses constituted with the CCVs a lineage well separated from FCoVs, transmissible gastroenteritis virus (TGEV), porcine epidemic diarrhea virus (PEDV) and human coronaviruses within the genogroup 1 (data not shown). The viruses from the Stockholm region also clustered separately from GOT=05.

In the tree based on M gene sequences, the viruses from Sweden segregated into two distinct subgroups among the CCVs type II (Fig. 2a) . GOT=05 and UPPS2=04 branched together with two viruses originating from Italy, separately from the SWE1 variants. In the tree based on the 3 0 region of the S gene, the Swedish viruses displayed a different clustering from the M gene tree (Fig. 2b) . Whereas GOT=05 grouped together with the main CCV reference strains, UPPS2=04 and the SWE1 viruses constituted a sublineage distinct from those represented by the CCVs and FCoVs of type II. The contradictory clustering of UPPS2=04 supports the above data on genetic diversity and sequence analysis and suggests that the virus was derived from a recombination between SWE1and GOT=05-like viruses with a crossover site located between the S and M genes. In the tree based on the 5 0 S gene, GOT=05 branched together with CCV UCD-2, separately from the other CCVs type II (data not shown). Given the scarcity of genomic sequences available from CCVs type I, no phylogenetic tree was constructed from the 5 0 S gene sequences generated from SWE1 viruses and UPPS2=04.

The genetic relatedness among the field CCVs mirrored their geographical origin: the four SWE1 variants originating from the Stockholm region presented identical sequences and grouped separately from GOT=05, recovered in the western part of the country. The clustering pattern of the fifth virus from the Stockholm region, UPPS2=04, differed according to the genomic region examined, suggesting a recombinant origin of its genome. Deduced from phylogenetic data, this hypothesis is supported by amino acid markers both in the S and M proteins.

Data from the 3 0 region of the S gene suggested that the SWE1 viruses and UPPS2=04 constituted a distinct sublineage, well separated from the CCVs type II identified so far. In a previous analysis of the highly variable 5 0 S gene, the Australian isolate UWSMN-1 was found as an outlier among the CCVs, with 13.9-17.0% nucleotide divergence from CCV type II reference strains [6] . Comparable or even higher percentages of nucleotide variation were observed between GOT=05 and the same reference strains, thus suggesting that GOT=05 also constitutes a genetically distinct variant among CCVs. In addition, 5 0 S gene sequences related to CCV type I were identified among the SWE1 viruses and UPPS2=04. A dual infection of the dogs with both CCVs types I and II may not be excluded. However, as there was no evidence of 5 0 S gene sequence of CCV type II or M gene sequence of CCV type I in their stools, the dogs were more likely infected with a single viral type presenting an atypical S gene. Taken together, these data suggest that the Stockholm viruses arose from recombination between CCV type I-and CCV type II-like or possibly a third type of CCV. Results generated lately also strengthen the hypothesis of a recombinant origin of UPP2=04 by confirming its relatedness to the SWE1 viruses for the S gene. Although the TGEV-like 5 0 S segment of the CCV isolate UCD-1 has been postulated to originate from recombination [15] , no such event was previously reported among field CCVs. Our data suggest that for the SWE1 viruses and UPPS2=04, a template switch would have taken place within the S gene and that for UPPS2=04 another crossover site would be located between the S and the M genes. The identification of hotspots for recombination within the S gene of MHV or IBV corroborates our assumptions [1, 14] . In addition, a recombination event also arose between the S and M genes of FCoVs type II [4] . Sequence analysis from UPPS1=04 and UPPS2=04, recovered from dogs living in the same house, further suggests that viral populations within a same environment may be complex and represented by several distinct variants. The infection of an animal with genetically different viruses allows the occurrence of recombination events. It is unknown whether the recombination associated with UPPS2= 04 took place among the dogs from Uppsala or whether the recombinant virus originated from another location. Data from the M gene confirm previous observations of accumulation of non-synonymous substitutions within the amino terminus ectodomain of the protein, which is exposed on the outside of the virion [11, 12] . As the amino terminus of the M protein is highly immunogenic, these features could result in different antigenic properties among CCVs recovered in Sweden, including the ability of the viruses to escape the host immune response.

There was no apparent relationship between the severity of the clinical signs in the dogs and the differences observed among the CCVs identified in this study. The death of the pup infected with MORA=04 was most probably related to the age of this animal as a higher susceptibility to develop severe clinical signs has been reported in young dogs infected with CCV [7] . [11] Divergent and atypical canine coronaviruses from Sweden 1513

A clustering of RNA recombination sites adjacent to a hypervariable region of the peplomer gene of murine coronavirus

SYBR Green real-time reverse transcription-polymerase chain reaction assay for the generic detection of coronaviruses

One-tube fluorogenic reverse transcriptionpolymerase chain reaction for the quantitation of feline coronaviruses

Feline coronavirus type II strains 79-1683 and 79-1146 originate from a double recombination between feline coronavirus type I and canine coronavirus

Comparison of the amino acid sequence and phylogenetic analysis of the peplomer, integral membrane and nucleocapsid proteins of feline, canine and porcine coronaviruses

Identification of canine coronavirus strains from feces by S gene nested PCR and molecular characterization of a new Australian isolate

Genetic evolution of canine coronavirus and recent advances in prophylaxis

Two genotypes of canine coronavirus simultaneously detected in the fecal samples of dogs with diarrhea

Cloning and expression of two fragments of the S gene of canine coronavirus type I

Genetic diversity of a canine coronavirus detected in pups with diarrhoea in Italy

Identification of coronaviruses in dogs that segregate separately from the canine coronavirus genotype

Molecular characterization of a virulent canine coronavirus BGF strain

A comparison of the genomes of FECVs and FIPVs and what they tell us about the relationships between feline coronaviruses and their evolution

Evidence of natural recombination within the S1 gene of infectious bronchitis virus

The S gene of canine coronavirus, strain UCD-1, is more closely related to the S gene of transmissible gastroenteritis virus than to that of feline infectious peritonitis virus

We thank Lihong Liu for assisting sequencing and useful advice. The work was supported and financed by project 22.1= 2003-1059 of the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS).