Intergenotype Recombination among New Norovirus GII.4 Variants in the Complete Genome

The norovirus (NoV), a member of the Caliciviridae family, is the leading cause of acute viral gastroenteritis, and is estimated to be the cause of almost half of all cases of gastroenteritis globally. Currently, NoVs are classified into 6 major groups, genogroup I (GI) to genogroup VI (GVI), based on amino acid sequence diversities of the VP1 [1]. GII is detected more frequently in clinical samples, based on the nucleotide sequence of RdRp (in ORF1) and VP1 (in ORF2). Twenty-seven and 22 genotypes of RdRp and VP1, respectively, have been established in GII [2]. The dominant global genotype that causes outbreaks and sporadic cases of gastroenteritis is genogroup II, genotype 4 (GII.4) [3]. Recombination occurs in many RNA viruses, and can be of major evolutionary significance [4]. Intragenotype recombination is widespread within the pandemic NoV GII.4 lineage [5]. Previous studies show that intragenotype recombination contributed to the emergence of the recent pandemic GII.4 variant in New Orleans in 2009, and a newly identified GII.4 variant, termed Sydney 2012 [6]. However, there are few reports about the role of intergenotype recombination in the emergence of the GII.4 variant in the complete genome.

In this study, the recombination and phylogenetic analyses showed that the Sydney 2012 variants, Kawasaki194 and CA3477, were recombinant NoV strains with a GII.4 capsid and a GII.2 polymerase gene. The Kawasaki194 strain (GenBank No. LC175468) was detected in the diarrheal feces of 9 patients in an outbreak in Kawasaki City, Japan, in 2016 [7]. The other NoV GII.4 variant, CA3477 strain (GenBank No. KX907727), was detected in Atlanta City, USA, in 2016. In the 2015-2016 season, the GII.4-GII.P16 polymerase emerged in November 2015, causing 60% of all GII.4 outbreaks in the USA [8]. To identify intergenotype recombination among GII.4 variants, the NoV GII.4 variant and GII.2 genotype complete genome sequences were analyzed using the Recombination Detection Program (RDP) v.4.55 and the SimPlot program.

The studied sequences comprised the available 643 complete genome sequences of the NoV GII genogroup from GenBank, dated October 2016. These were screened to exclude patented and artificial mutants, aligned in the MAFFT program [9], and manually adjusted for the correct reading frame. Redundant sequences were removed from all complete genome sequences at 97% sequence identification, and 63 representative complete genomes with T-RECs were obtained [10]. These representative sequences were separately used for phylogeny of the ORF1 and ORF2 regions. Phylogenetic trees were constructed using the neighbor-joining (NJ) method, and evaluated using the interior branch test method, with Mega 6.0 software [11]. The percent bootstrap support was indicated at each node, while the GenBank accession number was indicated at each branch. The ORF1 phylogenetic tree (Fig. 1a) showed that Kawasaki194 and CA3477 strains represent a distantly related genotype from the other GII.P4 variant in the ORF1 gene phylogenies. These 2 strains and the GII.P16 strains are closely related, and they formed 1 cluster (Fig. 1a). In Figure 1b, the Kawasaki194 and CA3477 strains and the GII.4 Sydney 2012 variants are closely related, and they formed 1 cluster in the ORF2 gene phylogenies.

Potential recombinant sequences in the 643 complete genomes were detected. The identification of potential parental sequences and the localization of possible recombination break points were determined using the RDP [12] methods embedded in RDP4 [13]. A multiple-comparison corrected p value cutoff of 0.05 was used throughout. The recombination events were further confirmed along with breakpoints using the SimPlot Program [14]. RDP results showed that 1 significant recombination signal was detected, with a high degree of confidence, and was judged by the 7 recombination detection methods. The results indicated that the recombination occurs in the Kawasaki194 and A3477 strains. When we further limited the analyses to the 2 strains, the 2 strains of the GII.4 lineage were identified as recombinants by the RDP. The major parental strain was Jiangsu2, isolated from an acute flaccid paralysis patient in Jiangsu, China, and the minor parental strain was GII.4 HS255, with nucleotide identities of 96.4 and 95.1%, respectively (Fig. 2a). This recombination event gave rise to the recombinant GII.4-GII.P16 strains, with a breaking point as shown in Figure 2b.

To confirm this recombination event, the relevant strains were analyzed by NJ trees using MEGA.6.0. Figure 3a and b show 2 trees constructed on the recombinant region (position 7771-5228) and nonrecombinant region (position 5228-7771), respectively. In Figure 3a, the daughter recombinant strains, CA3477 and Kawasaki194, were clustered closely to the major parent, the Jiangsu2 strain. Figure 3b, in which CA3477 and Kawasaki194 strains were clustered to the minor parent HS255 strain, shows discordant relationships with Figure 3a.

This recombination event was confirmed using SimPlot. When the standard similarity plots were constructed using CA3477, Kawasaki194, Jiangsu2, and HS255 sequences as queries, the SimPlot results were similar to those of the RDP analyses. The sequence of CA3477 and Kawasaki194 showed a high similarity to that of Jiangsu2, except 5,228-7,771 nt in the ORF1/ORF2 overlap, which had a high similarity to that of Jiangsu2, suggesting the presence of recombination. The results are shown in Figure 4a. To obtain a more accurate overview of the recombination, a subsequent bootscanning analysis was carried out with a SimPlot subprogram. Bootstrap values of over 80% are considered to be significant. As shown in Figure 4b, when the sequences of CA3477, Kawasaki194, Jiangsu2, and HS255 were studied as queries, reliable bootstrap values in the bootscanning analysis were produced. These results indicated that the recombination occurs in the CA3477 and Kawasaki194 strains. The results of SimPlot supported those of RDP.

Genetic recombination is a widespread phenomenon in NoVs. This has a major impact on their evolution and genotype diversity, and has been associated with the emergence of new genotypes. Most recombinations in NoVs occur in a single hotspot breakpoint located in the ORF1/ORF2 overlap [15,16,17]. In finding viral recombinants, RAT (Recombination Analysis Tool) indicates that large-scale recombination occurs in the NoV. It also shows a “hot spot” for recombination between the end of ORF1 and the start of ORF2 [18]. In addition, the T-RECs was specifically evaluated on NoV sequences, as demonstrated by an analysis of 555 NoV complete genomes and 2,500 sequence fragments, where a recombination hotspot was identified at the ORF1-ORF2 junction. A combined characterization of both the polymerase and capsid regions is important to monitor the emergence of new NoV genotypes and recombinant strains [19,20,21,22]. NoV intragenotype recombinants are frequently identified in molecular epidemiological studies [23,24,25,26]. This study also suggested that intergenotype recombination might have also played a role in the emergence of some GII.4 variants. GII.2 and GII.4, characterized as recombinant GII.4 variants Kawasaki194 and CA3477, were first described in the complete genome. In the 2015-2016 season, a GII.4 Sydney virus with a novel GII.P16 polymerase emerged in 2015, causing 60% of all GII.4 outbreaks in the USA [8]. The GII.P16 sequences shared by the GII.2 and GII.4 Sydney genotypes were nearly identical to the GII.P16-GII.4 Sydney sequences reported in Japan in January 2016 [7]. GII.2 viruses are known to harbor GII.P16 polymerases. GII.P16-GII.2 viruses previously reported in China, Japan, and Australia and New Zealand [27,28,29] are genetically distinct from the recently emerging GII.P16-GII.2 viruses, which caused a sharp increase in the number of NoV infections in Germany and China in late 2016 [30,31]. In this recombinant event, the major parent, Jiangsu2 strain, was isolated during 2011-2012 in Jiangsu, China. The cause of this recombination event, although unclear, may have been an increase in international communication.

This work was supported by the China Postdoctoral Science Foundation (2016M591790), Jiangsu Province Postdoctoral Science Foundation (1601072C), and Zhenjiang Health Science and Technology key project (SHW2015015).

The authors declare that they have no conflicts of interest.

J.X. and T.L. contributed equally to this work and should be considered joint first authors.