Identification of two divergent swine Noroviruses detected at the slaughterhouse in North East Italy.

Norovirus (NoV) has emerged as one of the major causative agents of non-bacterial, food- and water-borne gastroenteritis in humans, with the main genogroup involved in human outbreaks (GII), which has been detected worldwide in different animal species including swine. A four-month investigation at the slaughterhouse aiming to examine the presence of NoV in the swine in North-Eastern Italy, enabled the detection of two divergent Noroviruses (NoVs) (GII.P11) in two swine farms. This represents the first study in the swine population of North-Eastern Italy, which has paved the way for future integrated virological and epidemiological investigations on swine NoVs.


Background
Noroviruses (NoVs) are a leading pathogen of foodborne diseases, which cause an estimated 23 million cases of illness annually in the USA, and > 90% of nonbacterial epidemic gastroenteritis cases worldwide [1,2]. NoVs are transmitted through the ingestion of contaminated food or water, either via the faecal-oral route or via air-borne particles and contact with contaminated surfaces [3]. NoVs belonging to the Caliciviridae family are classified into ten genogroups (GI-GX), based on the variation of the major capsid protein (VP1), and are further divided into 49 genotypes [4]. The genetic diversity of the RNA-dependent-RNA-polymerase (RdRp) dictates the number of p-types [4]. NoVs detected in naturally infected pigs are classified as GII, which can also infect humans [5][6][7][8][9]. Since the first report of GII NoVs in pigs in the USA [10], other countries have reported the presence of this genogroup in diseased and healthy pigs [11]. However, little is known about the GII distribution and the epidemiological characteristics of NoVs in swine. Currently, no data are available on the NoVs prevalence in swine in Italy, although, belonging to the family Caliciviridae, they are listed among zoonotic agents that European member states should report on [12][13][14][15]. The aim of this study was to investigate the presence of NoVs in swine in two North East Italian regions, Veneto and Friuli Venezia Giulia (FVG).

Materials and methods
Seventy-nine faecal samples from 76 swineherds were collected at three slaughterhouses between March and June 2017. Forty swine herds were distributed in Veneto (eighteen fattening, six farrow to wean and sixteen farrow to finish) and 36 in FVG (thirty fattening farms, one farrow to wean and five farrow to finish).
One-step RT-PCRs were performed using the Super-Script III One-Step RT-PCR System with Platinum Taq DNA polymerase kit (Invitrogen, Carlsbad, CA, USA). For NoV detection an approach based on a universal one-step RT-PCR for Caliciviridae family [10,16] targeting a 300 bp RdRp fragment, paired with Sanger sequencing and BLAST search was adopted. Primers pair NVG4F and VN3T20 [10] were used to amplify a genomic portion of about 2500 bp containing the VP1 coding region, followed by sanger sequencing and BLAST search. A primer-walking approach based on newly designed primers (Table 1) was used to amplify and sequence NoV full genome applying the thermal cycling profile reported in Table 2.
Nucleotide (nt) sequences of the partial RdRp and of the VP1 were aligned with the MAFFT online software version 7 (https://mafft.cbrc.jp/alignment/software/). The best substitution model was identified using MEGA 6.0 and the phylogenetic analysis was performed with PhyML 3.0 with 100 replicate of bootstrap.

Results and discussion
Two out of seventy-nine faecal samples (2.53%) resulted NoV positive and characterized as GII.P11 based on the RdRp sequence ( Fig. 1). NoVs were identified in two different fattening farms located in Veneto provinces in March and June 2017, namely: NoV/Italy/swine/ 17DIAPD90019/2017 (90019/2017) from Padova province and NoV/Italy/swine/17DIAPD90078/2017 (90078/  The phylogenetic analysis based on 300 bp of the RdRp highlighted that the two Italian NoVs belong to the GII.P11 p-type but cluster into two different GII.P11 subgroups (Fig. 1a) The nearly complete genome sequence was obtained only for 90078/2017 and it was 7045 nt in length, with 18 and 500 nt missing toward the 5′ and the 3′ ends respectively, and organized in three open reading frames (ORFs). ORF1 showed the conserved cleavage sites and aa motifs typical of NoVs, including GRPGIGKT (nucleoside-triphosphatase (NTPase)), EYDEY (viral protein genome-linked (VPg)), GDCG (protease), and DYSRWDST, GLPSG and YGDD (RdRp). In accordance with the genome organization of other NoVs, ORF1 and ORF2 showed an overlap of twelve nt, with the latter encoding for the capsid protein in a + 1 frame.
Recombination analysis showed the presence of a possible recombination event in the ORF1 with GII.P18 strain NoV/ swine/GII.18/OH-QW125/03/US (AY823305.2). The recombination event was supported (p < 0.05) by seven out of eight methods with PhylPro being the exception (Fig. 2); however, the presence of double peaks in the chromatograms in the area surrounding the potential break point (1847 (G/C), 1852 (C/A), 1858 (A/G), 1863 (T/C) and 1875 (T/C)), hinders its confident identification. Sequencing of the genomic region of interest was repeated using freshly isolated RNA and adopting different primer combinations. However, it was not possible Up to date, the only detection of NoV in swine reported in Italy resulted from a retrospective analysis on archived faecal samples collected in 2006, which did not disclose sample characteristics, source and geographic origin and produced merely a 300 bp of the RdRp swine NoV sequence [17]. A follow-up study attempting to identify NoVs in swine in Italy proved unsuccessful [18]. Our investigation adds new and updated information on swine NoVs in Italy, obtained through sampling at the slaughterhouses. Although slaughterhouse sampling does have its limits, as primarily older animals are surveyed, it also presents several advantages in terms of animals and farms that can be potentially sampled, time and resources necessary, appearing a cost-effective approach to investigate the presence of NoVs for which epidemiological background information are missing in Italy.
Based on the partial RdRp sequence, the Italian strains belong to the GII.P11 p-type, proving, however, to be genetically distinct from GII.P11 NoVs reported worldwide. The two strains, sampled only few months apart from animals originating from different provinces, showed a low RdRp genetic identity (85.31%) between each other. This is suggestive of a high genetic variability among GII.P11 NoVs circulating in the area. This evidence, combined with the previous findings, which report that other age and production categories can harbour NoVs [19,20,23,24], suggests that modifying the sampling strategy may have a remarkable impact on the number and type of NoVs detected in swine in Italy, possibly elucidating the association between NoV detection, animal categories and farm typology.
We reported the first European swine NoV nearly full genome sequence, obtained using an in-house primerwalking strategy. Full genome analysis revealed the presence of a possible recombination event in the ORF1. This process normally occurs in the ORF1/ORF2 overlap region [25], although recombination events in other genomic regions have also been reported [26]. The putative minor parental strain was identified as a virus belonging to the GII.P18/GII.18 genotype. Recombination events between NoV genotypes had been observed and represent a major threat for human and veterinary health, as novel recombinants might have different antigenic properties compared to their parental strains [27,28]. However, our findings are not conclusive, as it was not possible to identify the recombination break point. Coinfections of NoVs belonging to different genotypes have been observed [25], hence the authors cannot exclude to be in the presence of a co-infection, rather than of a recombinant virus. The detection of GII.P18 genomic sequences suggests the circulation of this p-type in the area, making future investigations worthwhile.
NoV is largely studied as a human virus, only recently receiving attention for its capability of infecting terrestrial animals. Studies on NoV in swine remain isolated efforts, meaning that available information is disconnected by spatial-temporal gaps. Enhanced surveillance is uttermost necessary to explain the evolution and ecology of swine NoVs.