Norway rats (Rattus norvegicus) have a global distribution and live in close proximity to humans and human feces. Rats may therefore consume or contaminate human food products and can be reservoirs and vectors of several pathogens, including Leptospira spp., Salmonella spp. and Campylobacter spp. [1, 2]. Noroviruses are considered to be the single most common etiological agent of non-bacterial gastroenteritis in humans [3]. Noroviruses are also known to cause severe diseases in immunocompromised laboratory mice and have recently been detected in wild mice and rats [4, 5]. So far, all known rodent norovirus strains belong to genogroup V. Genogroup V noroviruses are genetically distant from human noroviruses, which have only been found in genogroups I, II and IV.

Recently, rat-associated hepatitis E virus (ratHEV) has been identified in wild rats in Germany, the US and Vietnam using serological methods, RT-PCR and real-time RT-PCR [6–11]. Although this genotype was shown to be unable to infect non-human primates [8], results from a serosurvey of forestry workers in Germany may suggest a human transmission of ratHEV or a related virus [12].

Here, we report the simultaneous detection of a genogroup I norovirus and ratHEV in a Norway rat, using reverse-transcription PCR, quantitative reverse-transcription PCR and transmission electron microscopy, including conventional negative staining (NS) with uranyl acetate and immunogold electron microscopy.

As part of an ongoing survey for rodent-borne pathogens in Europe, 11 Norway rats (Rattus norvegicus), 6 male and 5 female, were trapped by pest-control personnel underground in the Copenhagen (Denmark) sewer system at five sites between January and March 2012. During necropsy, no obvious pathomorphological alterations could be observed in any of the rats. Intestinal contents were screened for calicivirus by RT-PCR as described previously [13]. Calicivirus RNA was identified in 1 of 11 specimens (specimen KS12/1305 from a male rat). Comparison of the 278-bp PCR product (excluding the primer sequences) with other sequences in GenBank using tblastx revealed that the virus showed the highest amino acid sequence identity (97 % identity and 221 max score) to the recombinant human norovirus strain Hu/GI.b-GI.6/Roosendaal180/2007/NL (GenBank accession no. JN176919). Subsequently, the 3,986-nt 3′-terminus of the genome sequence (GenBank accession no. KC294198), which included the 3′ end of open reading frame (ORF) 1 and the full-length sequences of ORFs 2 and 3, was determined using a 3′-RACE approach as described elsewhere [13]. Based on ClustalW nucleotide sequence alignments of the partial RNA-dependent RNA polymerase (RdRp) and the full-length capsid genes, two maximum-likelihood trees were constructed (Fig. 1A-B), and the rat-derived virus was assigned the recombinant GI.b/GI.6 genotype. A norovirus genogroup I-specific real-time RT-PCR and serial dilutions of prequantified DNA plasmids bearing the target sequence were used to estimate the concentration of genome equivalents in the intestinal contents of the rat as described elsewhere [14]. Not accounting of losses and inefficiencies during the extraction steps, the RT reaction and real-time PCR, the virus titer was estimated to be 5 × 107 genome equivalents per gram of fresh feces. The presence of norovirus was confirmed by transmission electron microscopy, and more specifically by immunogold electron microscopy using polyclonal rabbit anti-capsid protein VP1 antibodies (Abcam, Cambridge, UK) and goat anti-rabbit IgG linked with 5-nm gold particles (BBinternational, Cardiff, UK) (Fig. 2A-B).

Fig. 1
figure 1

Molecular phylogenetic analysis of noroviruses on the basis of ClustalW nucleotide sequence alignments of a partial RNA-dependent RNA polymerase gene (A) and of the full-length capsid gene (B) and a phylogenetic tree of hepatitis E viruses based on a ClustalW nucleotide sequence alignment of the partial open reading frame 1 (C). Norovirus and ratHEV strains KS12/1305 (boldface) are described in the present study. The trees were constructed using the maximum-likelihood method based on the Tamura-Nei model of MEGA5. The percentages of replicate trees in which the associated viruses clustered together in the bootstrap tests (1,000 replicates) are shown next to the branches. The trees are drawn to scale, with branch lengths measured as the number of substitutions per site

Full size image
Fig. 2
figure 2

Transmission electron microscopy micrographs of two noroviruses (A-B) (approximately 38 nm and 34 nm in diameter, respectively – arrows) and two HEV-like particles (C-D) (C, an empty particle of approximately 33 nm in diameter; D, a filled particle of approximately 29 nm in diameter – arrows) in the intestinal contents of rat KS12/1305 by immunogold electron microscopy using rabbit anti-capsid protein VP1 polyclonal antibodies and goat anti-rabbit IgG linked with 5-nm gold particles. Gold granules are closely associated with the norovirus particles (arrowheads), whereas the HEV-like particles showed no immune reactions. Bars = 50 nm

Full size image

Moreover, electron microscopy also suggested the presence of HEV-like particles in specimen KS12/1305 (Fig. 2C-D). We then used primers (HEV-for, 5′-GCNCTGTTYGGCCCNTGGTT and HEV-rev, 5′-GGYTCACCRGARTGYTTYTTCCA) modified from a previously published broadly reactive nested HEV RT-PCR assay [7] in a single two-step RT-PCR setup. Using this assay, a product of the expected length (~325 bp) was obtained. Sequence analysis of the 283-bp fragment (without primers) with other sequences in the GenBank database using tblastx showed that the virus was most closely related (>99 % amino acid sequence identity) to a rat-associated HEV strain that had been detected in a wild rat in Germany (hepatitis E virus rat/Mu09/0685/DEU/2010; GenBank accession no. JN167537). This was confirmed by the construction of a phylogenetic tree, which was based on a ClustalW nucleotide sequence alignment (GOP 15 and GEP 6.66, respectively) of the amplified region of KS12/1305 (GenBank accession no. KC294199) and the corresponding regions of other HEV sequences of various HEV genotypes (Fig. 1C). The Copenhagen HEV KS12/1305 strain grouped with the recently identified novel ratHEV clade.

Noroviruses of genogroups I, II and IV are known to cause gastroenteritis in humans and are commonly associated with water- and food-borne transmission via the fecal-oral route. Rodents such as wild rats (Rattus spp.) have been shown to harbor and transmit a range of important human pathogens. Although there have been recent reports on the detection of novel genogroup V norovirus strains in wild rats [4, 5], so far there has been no convincing evidence that rats could be reservoirs for human noroviruses. Here, we report the detection of a recombinant GI.b/GI.6 human norovirus in the intestinal contents of a wild Norway rat. The concentration of norovirus genome equivalents found by real-time RT-PCR in rat specimen KS12/1305 was in the middle range of reported viral loads in norovirus-infected humans [15], and therefore replication of the virus in the rat may be considered. The potential for norovirus transmission from animal to man or vice versa has been a matter of debate [3]. It has been shown that human norovirus strains can replicate in pigs and cattle [16, 17]. In addition, human norovirus of genotype GII.4 was detected in pig fecal specimens and in retail meat in Canada [18], although similar findings have not been reported since. Further, even though a norovirus recombination event on the genogroup level has not been reported, there is still a risk that co-infection in rats with human norovirus and rat strains from genogroup V [4] may lead to the emergence of novel recombinant strains with unknown effects on host range and virulence. Due to their habitats and feeding behaviour, sewer rats are generally exposed to human feces, and the intake of human pathogens is likely to be substantial. In fact, the trapping position of rat KS12/1305 was within the sewer system of a larger hospital in the Copenhagen area, where the concentration of certain human pathogens in the clinical discharges might at times be above average. Although the norovirus particles may have merely passed through the rat’s intestinal tract by way of its close contact to large amounts of human feces, the high virus titer is still remarkable. Either way, our finding indicates a possible route for a rat-mediated norovirus transmission. Clearly, the screening of larger numbers of samples in a future study using RT-PCR, serological tests and immunohistochemical assays will elucidate the significance of this finding, allow the actual carriage rates of norovirus in rats to be determined, and help to evaluate their potential as reservoirs for zoonotic transmissions of norovirus.

A further finding of interest was the simultaneous detection of ratHEV in the intestinal contents of the norovirus-positive rat. HEV is a major causative agent of hepatitis in humans, both in developing and developed countries. HEV genotypes 1 and 2 exclusively infect humans, whereas strains of genotypes 3 and 4 can cause sporadic cases of hepatitis in humans and are zoonotically transmitted, with domestic pigs, wild boars, deer and other mammals serving as reservoirs. Recently, Johne and colleagues described novel HEV strains in rats (ratHEV) from several German cities that were genetically distinct from the human pathogenic genotypes [6, 7]. These ratHEV strains have been suggested to represent a novel HEV genotype. Genetically similar strains have also been detected in the US and Vietnam [8–10]. The detection of ratHEV in a Copenhagen Norway rat in this study underlines its global presence and suggests the necessity for future studies on its potential for human infections.