key: cord-0954821-04jsj9sb
authors: Bodnar, Livia; Lorusso, Eleonora; Di Martino, Barbara; Catella, Cristiana; Lanave, Gianvito; Elia, Gabriella; Bányai, Krisztián; Buonavoglia, Canio; Martella, Vito
title: Identification of a novel canine norovirus
date: 2017-08-31
journal: Infection, Genetics and Evolution
DOI: 10.1016/j.meegid.2017.04.020
sha: 2e0eef7888633184f85b8afd59c032c90c7892d8
doc_id: 954821
cord_uid: 04jsj9sb

Abstract By screening a collection of fecal samples from young dogs from different European countries, noroviruses (NoVs) were found in 13/294 (4.4%) animals with signs of enteritis whilst they were not detected in healthy dogs (0/42). An informative portion of the genome (3.4kb at the 3′ end) was generated for four NoV strains. In the capsid protein VP1 region, strains 63.15/2015/ITA and FD53/2007/ITA were genetically related to the canine GVI.2 strain C33/Viseu/2007/PRT (97.4–98.6% nt and 90.3–98.6% aa). Strain FD210/2007/ITA displayed the highest identity to the GVI.1 canine strain Bari/91/2007/ITA (88.0% nt and 95.0% aa). Strain 5010/2009/ITA displayed only 66.6–67.6% nt and 75.5–81.6% aa identities to the GVI.1 canine strains FD210/2007/ITA and Bari/91/2007/ITA and the GVI feline strain M49-1/2012/JPN. Identity to the other canine/feline NoVs strains in the VP1 was lower than 67.6% nt and 62.7% aa. Based on the full-length VP1 amino acid sequence and the criteria proposed for distinction of NoV genotypes, the canine NoV 5010/2009/ITA could represent the prototype of a third GVI genotype, thus providing further evidence for the genetic heterogeneity of NoVs in carnivores.

Noroviruses (NoVs) are members of the Caliciviridae family, and are recognized as one of the leading causes of acute gastroenteritis in humans (Atmar and Estes, 2006; Glass et al., 2009) . Epidemics associated with NoVs occur frequently in hospitals, cruise ships and childcare centers (Glass et al., 2009; Hall et al., 2013) and NoV infection has become a major public health concern (Siebenga et al., 2009; Vega et al., 2014) .

NoVs have also been identified in several mammalian species, including cows, pigs, lion, cats, dogs, bats, sea lions and harbor porpoises (DeGraaf et al., 2017; Li et al., 2011; Martella et al., 2007 Martella et al., , 2008 Pinto et al., 2012; Scipioni et al., 2008; Sugieda et al., 1998; Wu et al., 2016) .

NoVs are small, non-enveloped round viruses of approximately 30 to 35 nm in diameter with a positive-sense RNA genome. NoV genome is organized into three open reading frames (ORFs) (Green, 2007) . ORF1 encodes a polyprotein that is cleaved by the virus-encoded protease to produce several nonstructural proteins, including the RNA dependent RNA polymerase (RdRp). ORF2 encodes a major capsid protein (VP1) and ORF3 encodes a small basic protein (VP2) that has been associated with capsid stability (Bertolotti-Ciarlet et al., 2003) .

Based on the full-length VP1 amino acid sequence, NoVs can be classified into at least seven genogroups (GI to GVII) and N40 genotypes (Green, 2013; Kroneman et al., 2013; Vinjé, 2015; Zheng et al., 2006) . Only GI, GII, and GIV NoVs infect humans, with GII strains being the most prevalent worldwide (Green, 2007) . NoVs genetically similar to human NoVs have been recently found in carnivores (Martella et al., 2007 (Martella et al., , 2008 Pinto et al., 2012; Soma et al., 2014; Summa et al., 2012) , raising public health concerns of potential cross-species transmission due to the strict social interactions between humans and pets. Interestingly, significant sequence variation has been found across different canine/feline NoV strains identified to date, allowing the classification of at least 4 genotypes from three distinct genogroups, i.e. GIV.2, GVI.1 and GVI.2 and GVII ( Table 1 ).

The first canine NoV was reported in a 2 month-old dog with a fourday history of gastroenteritis and in co-infection with canine parvovirus type-2 (CPV-2) in 2007 in Italy (Martella et al., 2008) . Since then, canine NoVs have been reported in several countries in Europe, the Americas and Asia (Azevedo et al., 2012; Mesquita et al., 2010; Nascimento, 2012a, 2012b; Ntafis et al., 2010; Soma et al., 2014; Tse et al., 2012) but the pathogenic role of NoVs in dogs has not been demonstrated firmly. The prevalence of canine NoV in dogs with clinical signs of gastroenteritis has been estimated to vary from 2.1% (Martella et al., 2009) to 40% (Mesquita et al., 2010) . A study in the US identified canine NoV at a prevalence of 11% in canine diarrhea samples (Azevedo et al., 2012) . Canine NoV has also been detected in healthy dogs (Mesquita et al., 2010) , a pattern that is commonly observed in human NoV infections (Ozawa et al., 2007) . Serological investigations have confirmed the presence of NoV-specific antibodies in dogs, with various prevalence rates. In a small serological survey in Italy, antibodies to GIV.2 NoV were identified in 5% of the dogs (Di Martino et al., 2010) . In a study in UK, antibodies specific for different NoV genotypes were identified in 38.1% of samples collected between 1999 and 2001 and 60.1% of samples collected in 2012-2013 (Caddy et al., 2013) . A European-wide serological investigation revealed antibodies to GVI.2 NoV in 36% of the analysed dogs (Mesquita et al., 2014) .

In order to draw a more complete picture of NoV molecular epidemiology in dogs, in this study a collection of fecal specimens from diarrheic and healthy animals obtained from different European countries was screened using both broadly-reactive primers for caliciviruses and primers specific for NoVs. The sequence of a large portion of the genome at the 3′ end of two canine NoV strains was determined. The sequence of an additional two canine NoV strains identified in 2007 in Italy was determined and analysed in detail.

During the years of 2009, 2014, 2015 and 2016, a total of 336 dog fecal samples (n = 294 from dogs with mild to severe gastroenteritis and n = 42 from asymptomatic animals) were collected from domestic animals aged 0-12 months of four European countries (Germany (n = 3), France (n = 26), Spain (n = 26) and Italy (n = 239). The samples were submitted to the laboratory of Animal Infectious Diseases, Department of Veterinary Medicine, University of Bari, Italy, for diagnostic examinations. All the samples were stored at −80°C until use. In addition, two stool samples positive for NoV and collected in 2007 in Italy were included in the analysis.

Fecal specimens were diluted 10% w/v in phosphate-buffered saline, pH 7.2, and the debris were removed by centrifugation at 8000 ×g for 5 min. Viral nucleic acid was extracted from 200 μl of each clarified stool suspension by the QIAamp cador Pathogen Mini Kit (Qiagen S.p.A., Milan, Italy), following the manufacturer's protocol and the nucleic acid templates were stored at −80°C until use.

To assess the presence of caliciviruses (CVs), the samples were screened using a broadly reactive primer pair, p289-p290, amplifying a band of 315 bp for NoV and a band of 330 bp for vesivirus and sapovirus. The primers are targeted to the highly conserved motifs DYSKWDST and YGDD of the RNA-dependent RNA polymerase (RdRp) region of the polymerase complex (Jiang et al., 1999) . In the samples yielding amplicons of the expected sizes, the presence of NoV was confirmed using the NoV-specific primer pair JV12Y-JV13I (Vennema et al., 2002) . All the fecal samples collected from dogs with enteritis signs were also tested for the presence of common canine viral pathogens such as CPV-2 and canine coronavirus (CCoV) either by gel-based PCR or by quantitative PCR and RT-PCR (Buonavoglia et al., 2001; Decaro et al., 2005; Pratelli et al., 1999; Vennema et al., 2002) .

The amplicons were excised from the gel and purified using the QIAquick gel extraction kit (Qiagen GmbH, Hilden, Germany). The amplicons were subjected to direct sequencing using BigDye Terminator Cycle chemistry and 3730 DNA Analyzer (Applied Biosystems, Foster, CA). Basic Local Alignment Search Tool (BLAST; http://www.ncbi. nlm.nih.gov) and FASTA (http://www.ebi.ac.uk/fasta33) with default values were used to find homologous hits. The sequence of~3.4-kb fragment of the NoV genome (the 3′ end of ORF1, the full-length ORF2, ORF3, and the non-coding region through the poly-A tail of 4 canine (Zheng et al., 2006) . , 2006) . cDNA was synthesized by Super Script III First-Strand cDNA synthesis kit (Invitrogen Ltd., Milan, Italy) with primer QT. PCR was then performed with TaKaRa La Taq polymerase (Takara Bio Europe, Saint-Germain-en-Laye, France) with forward primer JV12Y and reverse primer JV13I. The cDNA was purified and cloned by using TOPO® XL Cloning Kit (Invitrogen Ltd., Milan, Italy). Additional primers were designed to determine the complete 3.4-kb sequences by a primer walking strategy (Table 2) . Sequence editing and multiple alignments were carried out with the Geneious software package version 9.1.6 (Biomatters Ltd., New Zealand). Genome sequences of NoVs strains were retrieved from GenBank and aligned using Clustal W (Larkin et al., 2007) . Phylogenetic trees were generated using Bayesian analysis with MrBayes (Huelsenbeck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003) .

Using the consensus primer pair p289-p290, CV RNA was found in 37 of 336 samples (11.01%). A total of 13 samples generated amplicons of the expected size for NoV (315 bp) and were subsequently confirmed to contain NoV RNA using NoV-specific primers, either alone (1.4%, 5/ 336) or in mixed infections with CPV-2 (1.4%, 5/336), CCoV (0.89%, 3/ 336), canine sapovirus (0.29%, 1/336) and vesivirus (0.29%, 1/336). Seventeen samples (5.05%, 17/336) were found to contain canine sapovirus alone (1.19%, 4/336) or in mixed infection with CPV-2 (0.6%, 2/336), CCoV (2.4%, 8/336), CPV-2/CCoV (0.6%; 2/336) and NoV (0.29%, 1/ 336). All the NoV and sapovirus positive samples were identified from diarrheic dogs with a prevalence rate of 4.4% (13/294) for NoVs and 5.7% (17/294) for sapovirus, whilst they were not detected from asymptomatic animals (0/42). Eleven NoV-positive samples were identified in the Italian collection of samples (11/239, 4.6%), whilst two NoV-positive samples were identified in the Spanish collection (2/26, 7.6%). NoV RNA was not detected in the German (0/3) and French (0/26) samples.

The sequence of~3.4-kb fragment at the 3′ end of the genome, including the partial RdRp and the complete ORF2 and ORF3 of the strains 63.15/2015/ITA, 5010/2009/ITA, FD210/2007/ITA and FD53/2007/ITA was determined and then submitted to GenBank under accession no. JF939046, JF930689, KY486328 and KY486329. The sequences were analysed with cognate sequences of animal and human NoVs reference strains available in the GenBank database.

As shown in displayed between the canine strains from this study and other animal and human GIV NoVs was b 55.7% aa and it was b45.7% aa to the Chinese GVII canine NoV strains.

Screening by RT-PCR of a collection of fecal samples obtained from young dogs (aged 0-12 months) with gastroenteritis identified canine NoVs in 4.4% of the symptomatic dogs, either alone or in mixed infectious with CPV-2, CCoV, canine sapovirus and vesivirus, but not in fecal samples from clinically healthy dogs. In previous molecular investigations canine NoVs have been detected in diarrheic stool samples with a prevalence rate ranging from 2.2% (Martella et al., 2009) , to 11.0% (Azevedo et al., 2012) and to 40% (Mesquita et al., 2010) . Canine NoVs have also been detected in healthy dogs with a prevalence of 9.0% (Mesquita et al., 2010) . Accordingly, the pathogenic role of these viruses in dog enteritis is not firmly established yet. The prevalence rate of 4.4% observed in this study was lower than those reported in the US and Portugal (Azevedo et al., 2012; Mesquita et al., 2010) , although different primer sets were used in the various studies. In our investigation, we preferred using broadly reactive primers designed to amplify all the CVs, p289-p290 (Jiang et al., 1999) , and subsequently a set of broadly reactive primers able to amplify the majority of NoV strains, JV12Y-JV13I (Vennema et al., 2002) . This algorithm, while allowing us to identify genetically diverse NoV strains, was likely less sensitive than other RT-PCR assays based on specific primers. The majority of the NoV-positive samples were identified in the Italian collection of samples (11/239, 4.6%), whilst 2 NoV-positive samples were identified in the Spanish collection (2/26, 7.6%). Mixed infections were identified but they were not unexpected as this eventuality is not infrequent. A limit of our study was surely the non-homogenous size of the sample collections, chiefly the small number of samples analysed for the non-Italian sample collections. Nevertheless, to our knowledge, this is the first report on the presence of NoV in dogs in Spain.

A fragment of about 3.4-kb at the 3′ end of the genome of four samples was used for analysis and comparison with a large selection of NoV strains belonging to all genogroups available on database. In the fulllength VP1 capsid gene (ORF1), two strains (FD53/2007/ITA and 63.15/2015/ITA) were characterized as GVI.2, whilst strain FD210/ 2007/ITA was characterized as GVI.1. A fourth strain, 5010/2009/ITA, although segregating within genogroup GVI, was more diverse genetically in the VP1 and could not be classified into any established GVI genotypes. According to Zheng's classification (Zheng et al., 2006) , the canine NoV strains 63.15/2015/ITA and FD53/2007/ITA were classified together with the canine strain C33/Viseu/2007/PRT and the feline strain TE/77-13/ITA, as a genotype 2 (N 85% pairwise aa identity inter genotypes), within the genogroup GVI (N 55% pairwise aa identity intergenogroups). The canine strain 5010/2009/ITA was also classified into genogroup GVI, but the aa identity to its closest relatives (the canine GVI.1 strains Bari/91/2007/ITA and FD210/2007/ITA) and the GVI feline strain M49-1/2012/JPN was lower (75.5-81.6% aa) than the 85% pairwise aa identity cut-off, proposed to discriminate between different NoV genotypes. Thereby, strain 5010/2009/ITA was tentatively classified as a third genotype (GVI.3) within the genogroup VI. Interestingly, in our analysis the canine GVI.1 strains FD210/2007/ITA, Bari/91/2007/ ITA and 5010/2009 /ITA appeared equally distant from the feline NoV strain M49-1/2012/JPN (75.5-80.4% aa) . This feline strain (GenBank accession number LC011950) is registered in NCBI sequence database as a GVI.1 strain, although it does not fit the criteria for being assigned to this genotype and it was tentatively proposed as a fourth GVI genotype, GVI.4 (Tables 1 and 3; Fig. 1 ).

NoVs genetically related have been detected in different host species. GII NoV have been found in human and pigs, although porcine NoVs belong to different genotypes (GII.11, GII.18 and GII.19 ) (Wang et al., 2005) . GIII NoV infect ruminants, with genotypes GIII.1 and GIII.2 (Liu et al., 1999; Oliver et al., 2007) being identified in cattle and genotype GIII.3 in sheep (Wolf et al., 2009) . However, circulation of NoVs strains belonging to the same genotype in different host species seems rather uncommon. By converse, it seems clear that carnivores (canids and felids) may be infected by NoV strains with the same capsid or polymerase (pol) types (Table 1) . Binding of human NoVs to human tissues seems to be mediated by antigens of the histo-blood group antigen (HBGA) family and therefore to be genetically determined (Marionneau et al., 2002) . Likewise, binding of GIV.2 and GVI NoVs in dog tissues seems mediated by antigens of the HBGA family (Caddy et al., 2014) . Interestingly, virus-like particles (VLPs) of seven different human NoV genotypes (GI.1, GI.2, GI.3, GII.3, GII.4, GII.6, and GII.12) were also found to bind to canine gastrointestinal tissues (Caddy et al., 2015) and human NoV strains of genotypes GII.4 and GII.12 were detected from infected dogs (Summa et al., 2012) . Whether dogs and cats share similar HBGAs that can be used as attachment factors by NoV could be hypothesized. Alternatively, carnivore HBGAs could be more "permissive", i.e. they could allow binding of NoVs of different genotypes. Regardless, this intriguing phenomenon could be a key factor explaining the apparent genetic diversity of carnivore NoVs.

Based on sequence and phylogenetic analysis of the 3′ partial sequence of ORF1 spanning750-nt at the COOH terminus of the RdRp, different pol genetic lineages could be distinguished among canine/feline GIV and GVI NoVs that were indicated, for the limited purpose of this study, with the letters "a" to "d". Table 1 ). A unique lineage (pol a) was formed by feline NoVs of genogroup GIV and GVI and by the lion NoV strain GIV.2/Pistoia/2006/ITA. Interestingly, inconsistencies were observed between the pol-and capsid-based phylogenies, which are suggestive of potential recombination events (Table 1) . This phenomenon has been reported repeatedly when studying NoVs from carnivores (Di Martino et al., 2016; Martella et al., 2009; Takano et al., 2015) . The canine NoV strain Bari/91/2007/ITA, resembles GIV.2 NoVs in its pol gene (pol b) whilst it possesses a GVI.1 capsid (Martella et al., 2009 ). The feline NoV strain, M49-1/2012/JPN possesses a GIV.2 pol (pol a), and a GVI capsid (Takano et al., 2015) , whilst the feline strain TE/77-13/2013/ITA possesses a GIV.2 pol (pol a) and a GVI.2 capsid (Di Martino et al., 2016) . In all the cases, the recombination site was located within the ORF1/ORF2 overlap, a region highly conserved across all NoV sequences (Bull et al., 2005) . RNA recombination is a major force driving viral evolution (Worobey and Holmes, 1999; Lai, 1992) . Recombination in viruses may substantially affect phylogenetic groupings, confusing molecular epidemiologic studies, and also can have major implications in vaccine design (Bull et al., 2005) . A recombinant NoV can be defined as a strain that clusters with 2 distinct groups of NoV strains when 2 different regions (normally the capsid and polymerase) of the genome are subjected to phylogenetic analysis (Bull et al., 2005) . Accordingly, analysis of the diagnostic regions located on the RdRp cannot be used to characterize unequivocally these animal NoVs and a definitive characterization should rely on the ORF2 as proposed for humans NoVs strains (Kroneman et al., 2013) .

Epidemiological studies are necessary to understand the relevance of newly discovered pathogens. These studies are important as they allow us to obtain information on the genetic diversity of viruses, a step that is propaedeutic to the development of accurate diagnostic tools and necessary for the understanding of viral evolution. Analysis of NoVs identified in carnivores thus far is unveiling a marked genetic diversity and suggests that the evolution of canine and feline NoV is tightly intermingled. This could have implications when enacting prophylaxis measures in animal populations (such as humane shelters, kennels or breeding facilities), as cats and dogs may be infected by a number of common pathogens (Decaro et al., 2010; Di Martino et al., 2009 Martella et al., 2002; Pratelli et al., 2003) . Finally, as some animal viruses may be transmitted to humans, generating new threats for human health, defining a baseline of the genetic diversity on animal viruses is important, to understand better and readily identify the origin of novel human viruses.

The epidemiologic and clinical importance of norovirus infection

Detection of norovirus in dogs in Arkansas

The 3′ end of Norwalk virus mRNA contains determinants that regulate the expression and stability of the viral capsid protein VP1: a novel function for the VP2 protein

Norovirus recombination in ORF1/ORF2 overlap

Evidence for evolution of canine parvovirustype2 in Italy

Serological evidence for multiple strains of canine norovirus in the UK dog population

Genogroup IV and VI canine noroviruses interact with histo-blood group antigens

Evidence for human norovirus infection of dogs in the United Kingdom

New approaches for the molecular characterization of canine parvovirus type 2 strains

Characterisation of canine parvovirus strains isolated from cats with feline panleukopenia

Norovirus infection in harbor porpoises

Characterization of a strain of feline calicivirus isolated from a dog faecal sample

Detection of antibodies against norovirus genogroup GIV in carnivores

A novel feline norovirus in diarrheic cats

Norovirus gastroenteritis

Caliciviridae: the noroviruses: specific virus families

Caliciviridae: the noroviruses

Acute gastroenteritis surveillance through the National Outbreak Reporting System, United States

MrBayes: Bayesian inference of phylogenetic trees

Design and evaluation of a primer pair that detects both Norwalk-and Sapporo-like caliciviruses by RT-PCR

Proposal for a unified norovirus nomenclature and genotyping

RNA recombination in animal and plant viruses

Clustal W and Clustal X version 2.0

The fecal viral flora of California sea lions

Molecular characterization of a bovine enteric calicivirus: relationship to the Norwalk-like viruses

Norwalk virus binds to histoblood group antigens present on gastroduodenal epithelial cells of secretor individuals

Analysis of the capsid protein gene of a feline-like calicivirus isolated from a dog

Norovirus in captive lioncub

Detection and molecular characterization of a canine norovirus

Genetic heterogeneity and recombination in canine noroviruses

Molecular epidemiology of canine norovirus in dogs from Portugal

Gastroenteritis outbreak associated with faecal shedding of canine norovirus in a Portuguese kennel following introduction of imported dogs from Russia

Novel norovirus in dogs with diarrhea

Seroprevalence of canine norovirus in 14 European countries

An outbreak of canine norovirus infection in young dogs

Complete genomic characterization and antigenic relatedness of genogroup III, genotype 2 bovine noroviruses

Norovirus infections in symptomatic and asymptomatic food handlers in Japan

Discovery and genomic characterization of noroviruses from a gastroenteritis outbreak in domestic cats in the US

Fatal coronavirus infection in puppies following canine parvovirus 2b infection

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

MrBayes 3: Bayesian phylogenetic inference under mixed models

Animal noroviruses

3′ end cDNA amplification using classic RACE

Norovirus illness is a global problem: emergence and spread of norovirus GII.4 variants

Detection of norovirus and sapovirus from diarrheic dogs and cats in Japan

Detection of Norwalk-like virus genes in the caecum contents of pigs

Pet dogs -a transmission route for human noroviruses?

Molecular characterization and pathogenicity of a genogroup GVI feline norovirus

Complete genome sequences of novel canine noroviruses in Hong Kong

Genotypic and epidemiologic trends of norovirus outbreaks in the United States

Rational optimization of generic primers used for Norwalk-like virus detection by reverse transcriptase polymerase chain reaction

Advances in laboratory methods for detection and typing of norovirus

Porcine noroviruses related to human noroviruses

Molecular detection of norovirus in sheep and pigs in New Zealand farms

Evolutionary aspects of recombination in RNA viruses

Deciphering the bat virome catalog to better understand the ecological diversity of bat viruses and the bat origin of emerging infectious diseases

Norovirus classification and proposed strain nomenclature

This study was supported by grants from the institution CAPES, Brazil, from the Momentum Program, Hungary, and from the action "Future in Research" of Apulian Region, Italy.