The importance of intergenic recombination in norovirus GII.3 evolution

Abstract

Norovirus genotype II.3 (GII.3) strains are a major cause of sporadic gastroenteritis. Intergenic recombination between the capsid and RNA-dependent RNA polymerase (RdRp) genes is common and results in the acquisition of an alternative RdRp genotype. This study aimed to explore the evolution of the GII.3 capsid gene, focusing on the influence of intergenic recombination. The capsid genes from six GII.3 norovirus strains, collected from Australian children between 2001 and 2010, were sequenced and aligned with 66 GII.3 capsid sequences from GenBank, spanning 1975 to 2010. The GII.3 capsid gene evolved at a rate of 4.16 × 10(-3) to 6.97 × 10(-3) nucleotide substitutions/site/year from 1975 to 2010 and clustered into five temporally sequential lineages. Clustering of the GII.3 capsid gene sequences was associated with intergenic recombination and switches between RdRp genotypes GII.3, GII.a, GII.b, GII.12, and an undefined ancestral RdRp. Comparison of the substitution rate of the GII.3 and GII.b RdRps suggested that RdRp switching allows a higher evolutionary rate, leading to increased genetic diversity and adaptability. Alignment of GII.3 capsid sequences revealed 36 lineage-specific conserved amino acid substitutions, four of which were under positive selection. Many conserved substitutions were within predicted antibody binding regions and close to host attachment factor binding sites. In conclusion, evolution of GII.3 noroviruses was primarily driven by intergenic recombination. The acquisition of new RdRps may lead to a faster mutation rate and increased genetic diversity, improving overall GII.3 fitness.

Fig 1

Maximum clade credibility tree of GII.3 VP1 gene sequences. The maximum clade credibility tree is shown for 72 GII.3 capsid gene sequences from 1975 to 2010, as prepared using BEAST (strict clock) and FigTree. Where sequences were obtained from GenBank, the accession number is indicated in the taxon label. Strains sequenced for this study are italicized. An asterisk beside the taxon label indicates a strain for which the RdRp genotype was not determined. Previously defined clusters (I to III), new genetic lineages (A to F), and RdRp genotypes are indicated. Major nodes are labeled with the posterior probability that the associated taxa clustered together. Major branches are labeled (in boldface) with the tMRCA for all taxa within the cluster. tMRCA is expressed as the number of years prior to the collection date of the most recent samples included in the tree (2010). The tree is drawn to scale with branch lengths in units of time (years).

Fig 2

Homology model of the norovirus GII.3 capsid dimer. The structural homology model of the GII.3 capsid dimer (strain 693/425/2008/AU) is shown with the protruding domains (P1 and P2) and the shell domain (S) indicated. The dimer is shown from the top view (A) and two side views (B and C) with 90° rotation between the two side views. The HBGA binding sites are indicated in red, and the residues involved in antibody binding epitopes, as described for GII.4 noroviruses (sites 311, 312, 381, and 415), are displayed as spheres. Conserved substitutions are indicated by coloration. Sites of conserved substitution are colored according to the lineage where a substitution at this site last occurred: lineage B, green; lineage C, teal; GII.a RdRp group, light gray; lineage D, blue; lineage E, purple. Conserved substitutions occurring within or in proximity of antibody binding epitopes A and E are circled.

Fig 3

Mutation rates of GII.3 and GII.b RdRp gene sequences. Each graph displays the nucleotide mutation rate, in substitutions per site per year (y axis), of GII.3 (n = 66) and GII.b (n = 110) RdRp gene sequences (region A, 274 bp) (x axis). Rate predictions were given using the three clock models, strict clock (A), UCED (B), and UCLD (C). The diamond indicates the mean mutation rate, and the error bars indicate the 95% HPD intervals.

Fig 4

Bayesian skyline plots of relative effective population size over time. (A and B) Relative effective population size, N_eτ (relative genetic diversity), of norovirus strains (y axis) plotted against time (x axis). (A) Population size was determined based on 72 norovirus GII.3 VP1 gene sequences spanning 1975 to 2010. The black line represents the median posterior value, and the gray lines represent the 95% HPD intervals. (B) Relative population size was determined based on region A of the RdRp gene using 66 GII.3 sequences (gray line) spanning the years 1983 to 2007 and 110 GII.b sequences (black line) spanning the years 2000 to 2011. The 95% HPD intervals were removed from this plot for simplicity.