Human noroviruses (NoV) cause epidemics of acute gastroenteritis (AGE) worldwide and can be transmitted through consumption of contaminated foods. Fresh products such as shellfish can be contaminated by human sewage during production, which results in the presence of multiple virus strains, at very low concentrations. Here, we tested a targeted metagenomics approach by deep-sequencing PCR amplicons of the capsid (VP1) and polymerase (RdRp) viral genes, on a set of artificial samples and on shellfish samples associated to AGE outbreaks, to evaluate its advantages and limitations in the identification of strains from the NoV genogroup (G) II. Using artificial samples, the method allowed the sequencing of most strains, but not all, and displayed variability between replicates especially with lower viral concentrations. Using shellfish samples, targeted metagenomics was compared to Sanger-sequencing of cloned amplicons and was able to identify a higher diversity of NoV GII and GIV strains. It allowed phylogenetic analyses of VP1 sequences and the identification, in most samples, of GII.17[P17] strains, also identified in related clinical samples. Despite several limitations, combining RdRp- and VP1-targeted metagenomics is a sensitive approach allowing the study NoV diversity in low-contaminated foods and the identification of NoV strains implicated in outbreaks.
Keywords: amplicon deep sequencing; foodborne virus; metagenomics; norovirus; viral diversity.