Until very recently MNVs were only described in laboratory mice and with limited genetic diversity compared to other noroviruses, either indicating a recent introduction of MNVs to mouse (Mus musculus) and/or limited evolutionary pressure on MNVs compared to noroviruses of other hosts.
In 2010, we initiated a study for the molecular detection of caliciviruses in stool samples of wild and laboratory rodents. Based on literature search and sequence alignments, two primer pairs, a broadly reactive and an MNV specific, were selected for calicivirus detection. Primer pair P289/P290 targets nucleotide sequences encoding conserved amino acid motifs in the calicivirus RdRps and it was successfully used in our previous studies for the molecular detection of noro-, sapo-, vesi- and recoviruses (Farkas et al., 2010
; Farkas et al., 2000
; Farkas et al., 2008
; Farkas et al., 2004
). Tests with a few MNV positive samples indicated that P289/P290 is also able to detect MNVs. Despite its broader detection range, none of the wild mouse cDNA samples tested positive with P289/P290, while two of the striped field mice (Apodemus agrarius
) samples yielded MNV specific amplicons with the MNV specific primers. Since the two primer sets target different regions of the MNV genome, for comparative reasons all of the laboratory mouse samples were tested only with the MNV specific primers. Ten of the 41 pooled samples collected from laboratory mice yielded MNV sequences. Since BLAST search and phylogenetic analysis indicated that all of the MNV sequences obtained from the CCHMC mouse colony were genetically closely related to each other and to previously described MNVs (), individual samples of the positive pools were not retested. Wild mouse individual samples were not available. Phylogenetic analyses of the 356 nt partial ORF2 sequences indicated that the two striped field mice (Apodemus agrarius
) MNVs, TF5WM and TF7WM are more closely related to MNVs recently described in wood mouse (Apodemus sylvaticus
) in the UK (Smith et al., 2011
), than to laboratory mouse MNVs (). However, while the difference among the laboratory mouse MNVs was less than 5 % nt positions, TF5WM/TF7WM and Apo455/Apo960 differed from each other in 8–11% nt positions indicating the existence of higher genetic diversity among MNVs in wild mice than described in laboratory mice. Since norovirus classification is based on the analyses of complete capsid (VP1) protein sequences full length ORF2 (VP1) sequences were obtained for both TF5WM and TF7WM. Analysis of the 1623 nt (541 aa) long ORF2 (VP1) sequences of TF5WM/TF7WM revealed 22–24% nt (15–18% aa) differences with the laboratory mouse MNV strains. More interestingly, the differences between TF5WM/TF7WM and Apo455/Apo960 (20–22% nt and 13–14% aa) ranged at a very similar level. For comparison, the diversity among the laboratory mouse MNVs in the full length ORF2 (VP1) was 4–13% nt (2–7% aa) (). According to a large scale analysis of norovirus VP1 sequences, norovirus strains within the same genetic cluster (genoptype) differ at 0–14.1 % aa positions while the difference among the different genotypes of the same genetic group (genogroup) is 14.3–43.8 % and among the various genogroups is 44.9–61.4 % (Zheng et al., 2006
). Based on these, the Apodemus
MNVs are clearly separated from the laboratory mouse MNVs at the genetic cluster/type level (15–18%), while the difference among the UK and Hungarian Apodemus
strains (13–14%) is slightly below the cut-off value proposed for genotype distinction. Pairwise distance calculations including over 70 GI and GII human NoVs corroborated this observation (data not shown). Based on these observations, we propose the classification of Mus
MNVs into two genetic clusters, GV.1 and GV.2, respectively (). To clearly establish whether genotype level variation exist among the Apodemus
MNVs more sequence data is needed.
Both the TF5WT and TF7WT ORF2 sequences contained a second reading frame (ORF4) of the same length (639 nt) as in laboratory mouse MNVs which is 21 nt longer than the Apo455 and Apo960 ORF4s (618 nt). Recently, McFadden et al., demonstrated that ORF4 is efficiently translated during MNV infection and ORF4-derived protein, VF1 plays a role in promoting virus replication by interfering with interferon mediated host response pathways and apoptosis (McFadden et al., 2011
). The ORF4/VF1 coding region almost completely overlaps with the N-terminal and S-domain (N/S) of the ORF2/VP1 with a +1 frame shift between the 2 ORFs. Interestingly, in the ORF2/N/S the differences among the various MNVs were higher at the nt than at the aa level (4–15% nt vs. 2–8% aa), while in the ORF4/VF1 the nt differences were significantly lower than the aa differences (5–16% nt vs. 10–35% aa) (). Multiple sequence alignments revealed that most of the nt substitutions (~75%) occurred at positions corresponding to the 3rd
codon position in ORF2/N/S and the 2nd
codon position in ORF4/VF1. This evolutionary conservation of the N/S domain of a structural protein (VP1) vs. a protein (VF1) with role in virulence remains to be explained. The larger number of non-synonymous substitutions in ORF4 may also suggest that the active, functional domain(s) of VF1 might be limited to a few conserved regions.
In summary, here we provided further evidence for the existence of genetically diverse MNVs in wild mice. Evaluation of more diverse host species is necessary.