Characterization of the Genomic Diversity of Norovirus in Linked Patients Using a Metagenomic Deep Sequencing Approach

Abstract

Norovirus (NoV) is the leading cause of gastroenteritis worldwide. A robust cell culture system does not exist for NoV and therefore detailed characterization of outbreak and sporadic strains relies on molecular techniques. In this study, we employed a metagenomic approach that uses non-specific amplification followed by next-generation sequencing to whole genome sequence NoV genomes directly from clinical samples obtained from 8 linked patients. Enough sequencing depth was obtained for each sample to use a de novo assembly of near-complete genome sequences. The resultant consensus sequences were then used to identify inter-host nucleotide variations that occur after direct transmission, analyze amino acid variations in the major capsid protein, and provide evidence of recombination events. The analysis of intra-host quasispecies diversity was possible due to high coverage-depth. We also observed a linear relationship between NoV viral load in the clinical sample and the number of sequence reads that could be attributed to NoV. The method demonstrated here has the potential for future use in whole genome sequence analyses of other RNA viruses isolated from clinical, environmental, and food specimens.

Keywords: antigenic variation; genetic variation; linked patients; next-generation sequencing (NGS); norovirus; recombination.

Figure 1

Correlation between the viral titre (genome copies/μl) and depth of coverage (reads mapped to norovirus genome).

Figure 2

Distribution of NoV reads across the sequenced genomes. Coverage was calculated as the total number of reads covering a given nucleotide and was normalized by the sum of total coverage across the genome. i.e., at each residue, the coverage was divided by the total coverage and the sum of normalized coverage equals one. (A–D) read coverage for the linked patients from families (A–D) (Table 1). Schematic representation of the NoV encoded proteins is shown below the graphs

Figure 3

Phylogenetic analysis of the NoV genomes identified in this study. The consensus sequences of the ORF1 for GII.4 (A) and GII.6 (B) sequences respectively, and ORF2 for GII.4 (C) and GII.6 (D) from each assembly were then aligned with highly similar sequences from GenBank database. The robustness of the phylogeny was assessed through bootstrap analysis of 1000 pseudo-replicates. The accession numbers for the reference genomes are provided in parentheses. The scale bars represent the number of substitutions per site.

Figure 4

Distribution of single nucleotide variants (SNVs) across the sequenced NoV genomes. Schematic representation of the NoV encoded proteins is shown above the graphs. (A–D), graphs represent SNV frequency at each position for families (A–D).

Figure 5

Non-synonymous differences within the structural domains of the capsid protein (VP1, ORF2), which are the N-terminal (N), shell (S), P1, and P2 domains. (A) Non-synonymous differences between the sequenced GII.4 Sydney genomes (families A and C) and their references (B) amino acid differences between the sequenced GII.6 genomes (families B and D) and their references. Individual epitope sites are highlighted in different colors: Yellow, epitope site A; blue, epitope site B; gray, epitope site C; orange, epitope site D; pink, epitope site E. Stretches of amino acid differences are shown in red. Residues are numbered according to the NoV GII4/2012/Sydney (Accession No: KM258128).

Figure 6

Recombination analyses of the full genomes of norovirus. A sliding window of 200 nt was utilized to make the SimPlot. SimPlot was generated for the linked recombinant genomes 15–65 and 15–66. Two breakpoints were identified, which leads to production of a mosaic genome that has a GII.6 backbone with an insertion from a GII.4 Sydney. Schematic representation of the NoV genome is shown above the graph.