Evolutionary Phylodynamics of Korean Noroviruses Reveals a Novel GII.2/GII.10 Recombination Event

Viral gastroenteritis is the most common causal agent of public health problems worldwide. Noroviruses cause nonbacterial acute gastroenteritis in humans of all ages. In this study, we investigated the occurrence of norovirus infection in children with acute gastroenteritis admitted to university hospitals in South Korea. We also analyzed the genetic diversity of the viruses and identified novel recombination events among the identified viral strains. Of 502 children with acute gastroenteritis admitted to our three hospitals between January 2011 and March 2012, genotyping of human noroviruses was performed in 171 (34%) norovirus-positive samples. Of these samples, 170 (99.5%) were in genogroup II (GII), while only one (0.5%) was in genogroup I (GI). The most common GII strain was the GII.4-2006b variant (n = 96, 56.5%), followed by GII.6 (n = 23, 13.5%), GII.12 (n = 22, 12.9%), GII.3 (n = 20, 11.8%), GII.2 (n = 6, 3.5%), GII.b (n = 2, 1.2%), and GII.10 (n = 1, 0.6%). Potential recombination events (polymerase/capsid) were detected in 39 GII strains (22.9%), and the most frequent genotypes were GII.4/GII.12 (n = 12, 30.8%), GII.4/GII.6 (n = 12, 30.8%), GII.4/GII.3 (n = 8, 20.5%), GII.b/GII.3 (n = 3, 7.7%), GII.16/GII.2 (n = 2, 5.1%), GII.4/GII.2 (n = 1, 2.6%), and GII.2/GII.10 (n = 1, 2.6%). For the first time, a novel GII.2/GII.10 recombination was detected; we also identified the GII.16/GII.2 strain for the first time in South Korea. Our data provided important insights into new recombination events, which may prove valuable for predicting the emergence of circulating norovirus strains with global epidemic potential.


Introduction
Viral gastroenteritis is the most common causal agent of public health problems worldwide. Using novel sensitive diagnostic methods, researchers have shown that noroviruses (NoVs) cause acute nonbacterial gastroenteritis in humans of all ages. NoVs are transmitted from person to person through contaminated foods such as raw shellfish and other routes such as aerosols and water [1]. Approximately 23 million cases of gastroenteritis caused by NoVs are estimated to occur in the United States each year, and about 60-85% of all gastroenteritis outbreaks within the United States, Europe, and Japan are associated with NoVs [2,3].
NoVs comprise a distinct genus within the family Caliciviridae and are nonenveloped viruses harboring a positive-sense RNA genome of about 7.7 kb [4]. This genome includes three open reading frames (ORFs): ORF1 encodes nonstructural proteins, including proteases, RNA nucleoside triphosphatases (NTPases), and RNA-dependent RNA polymerase (RdRp), while ORF2 and ORF3 encode the major capsid proteins VP1 and VP2, respectively [5]. Genome characterization studies based on the RdRp or VP1 gene sequences have classified NoVs into five distinct genogroups (GI-GV), which can be further subdivided into 44 genotypes, including the 16 GI, 23 GII, 2 GIII, 2 GIV, and 1 GV genotypes [6]. Recently, a sixth genogroup has been also proposed [7]. The GI, GII, and GIV genotypes are known to infect humans; GII.4 is the most common genotype and has caused global outbreaks and acute gastroenteritis in humans since the mid-1990s [8,9]. Other genotypes are found less frequently, but have been shown to cause sporadic outbreaks or epidemics [10].
The evolution of new RNA viruses and viral genomes with genetic variation is attributed to genetic recombination, the key mechanism mediating sequence exchange [11]. Viral recombination can change phylogenetic inference, virulence, and epidemiology and can have major implications in vaccination strategies [12]. Several recombinant NoV strains have caused outbreaks of acute gastroenteritis [13,14,15]. Additionally, phylogenetic analysis of RdRp-and capsid-encoding genes has supported that NoVs undergo recombination [16]. Predictive models have indicated that common NoV recombinants have breakpoints located either within or close to the ORF1/ORF2 junction. Consequently, this region is thought to constitute a recombination hotspot in NoVs [12].
In South Korea, after rotaviruses, NoVs are the next most important cause of nonbacterial acute gastroenteritis in children. Surveillance studies have indicated that the emergence of NoVs may be associated with both the season and the geographic area. The GII.4 genotype was the most prevalent strain between 2007 and 2008, while reports from 2001 to 2005 showed that GII.1 was the most predominant strain [17,18].
In this study, we investigated the occurrence of NoV infections in children with acute gastroenteritis at three university hospitals located in South Korea between January 2011 and March 2012. We also analyzed the genetic diversity of NoVs according to genogrouping and genotypes and identified novel recombination events among the identified strains.

Ethics statement
The stool samples used in this study were collected and analyzed under our protocol (number #2010-10-02), which was approved by the Human Subjects Institutional Review Board (IRB) of Chung-Ang University College of Medicine, Seoul, Korea. For children enrolled in this study, written informed consent was obtained from parents or legal guardians. This consent also included permission to use the data for future research purposes.

Stool specimens
Stool samples were collected from 502 children hospitalized with acute gastroenteritis at three university hospitals in Seoul between January 2011 and March 2012. All samples were diluted 10-fold with phosphate-buffered saline (pH 7.4) and centrifuged at 10,0006g for 15 min. The supernatant was used as the fecal suspension.

RNA extraction
Viral RNA was extracted using TRI reagent (Gibco BRL Life Technologies, Grand Island, NY, USA), according to the manufacturer's instructions. In brief, 0.3 mL supernatant from fecal suspensions was mixed with 0.7 mL TRI reagent and 0.2 mL chloroform/isoamylalcohol (24:1). After centrifugation at 12,0006g for 15 min, an equal volume of isopropanol was added to the aqueous phase containing RNA. The RNA was precipitated by centrifugation at 12,0006g for 10 min, washed with 70% ethanol, dissolved in 20 mL RNase-free water, and stored at 280˚C until use in reverse transcription-polymerase chain reaction (RT-PCR).

RT-PCR
RT was performed with 1-mM concentrations of primer GIR1M for GI capsid or primer GIIR1M for GII [19] and NV35/Reverse for RdRp, as described previously [20]. Nested RT-PCR was performed using the primers listed in Table 1. For capsid genes, the first PCR was performed in a 25-mL reaction tube containing 2 mL cDNA, 2.5 U Ex Taq DNA Polymerase (TaKaRa Bio Co., Shiga, Japan), and 20 pM GIF1M/forward and GIF1M/reverse primers for NoV capsid GI or GIIF1M/forward and GIIR1M/reverse primers for capsid GII. PCR amplification was performed with denaturing at 95˚C for 5 min, annealing at 50˚C for 1 min, extension at 72˚C for 1 min for 30 cycles, and a final extension at 72˚C for 10 min. Only the samples were weak or not amplified after the first PCR, the second PCR was performed in a 25-mL reaction containing 1 mL of the first PCR product, 2.5 U Ex Taq DNA polymerase, and 20 pM GIF2/forward and GIR1M/reverse primers for NoV GI capsid or GIIF3M/forward and GIIR1M/reverse primers for NoV GII capsid to generate 307-and 288-bp products, respectively. Thirty-five cycles of PCR amplification were completed under the same conditions as for the first PCR. For the RdRP gene, NV32 and NV36 were used as the primers for the first amplification reaction, while NV33 and NV35 primers were used for the second amplification reaction. PCR products were separated on 1.0% SeaKem LE agarose gels (FMC Bioproducts, Rockland, ME, USA) and stained with ethidium bromide. Gels were imaged using a GelDoc XR image-analysis system (BioRad, Hercules, CA, USA).

Nucleotide sequencing and phylogenetic analysis
Amplified PCR products were cleaned using a QIAquick PCR purification kit (Qiagen, Westburg, Germany) and then sequenced using a BigDye Terminator Cycle Sequencing Kit and an automated DNA sequencer (Model 3730, Applied Biosystems, Foster City, CA, USA). The resulting nucleotide sequences were aligned using CLUSTAL-X 2.1 program with parameters set against the corresponding NoV sequences from the NCBI GenBank database [21]. The nucleotide sequences of the capsid and RdRp genes from the study strains were compared with cogent genes of NoVs available in the public database. Phylogenetic trees were constructed using neighbor-joining algorithms from the PHYLIP suite [22] and the Kimura two-parameter model using MEGA6 software. Evolutionary distances for neighbor-joining analyses were based on the model developed by Jukes and Cantor [23]. Tree topology was evaluated using the bootstrap resampling method with 1,000 replicates of the neighbor-joining dataset with the CONSENSE and SEQBOOT programs from the PHYLIP suite. The recombination nucleotide sequences obtained in this study were deposited in the GenBank under the accession numbers KC110854 (4CAU-14), KC10855 (Gil-67), KC110856 (Gil-70), and KC110857 (4CAU-38).

Recombination analysis
To detect recombination events, an additional RT-PCR was performed using NV33 and GIIR1M primers to amplify a genomic fragment of 1,057 bp, which included the junction between the RdRP and capsid genes [24]. The most frequent sites for recombination flanked the capsid region and could be identified using similarity plots in SimPlot software and Bootscan analysis [25,26]. SimPlot analyzes the percent identity of a query sequence to a panel of reference sequences by position in a stepwise manner across the whole alignment [27]. Identification of recombination events without prior identification of nonrecombinant reference sequences was performed using Bootscan analysis [25].  (Table 3).

GI genogroup
The nucleotide sequence of the partial capsid gene (287 bp) of the KW-71 strain was compared with the NoV GI sequences available in the GenBank database. Phylogenetic analysis demonstrated that the capsid gene of the KW-71 strain clustered into the GI.3 genotype (Fig. 1).

GII genogroup
he nucleotide sequences of the partial RdRp gene (338 bp) and the partial capsid gene (287 bp) from 96 of the 123 NoVs GII strains were determined and compared with those of NoVs available in GenBank. Each sequence data of the remaining 27 samples were subsequently totally identical to one among these 96 strains. The phylogenetic tree showed that most of the RdRp and capsid genes from the Korean study strains were classified into the GII.4 genotype and shared nucleotide sequence identities between 94% and 99% (Fig. 2). The most circulated GII strains were the GII.4 2006b variant (56.5%), followed by GII.6 (13.5%), GII.12 ( (Fig. 2).

Analysis of recombination events
Recently, reports from several countries have demonstrated that the GII.b/GII.3 recombinant is the predominant pediatric genotype. Phylogenetic analysis indicated that the Gil-67 strain was a GII.b/GII.3 recombinant (Fig. 2). SimPlot software was used to compare the nucleotide sequence of Gil-67 to those of the GII.3 Arg320 strain isolated in Argentina and the GII.b/GII.21 PC03 strain isolated in India. The TV24 strain (U02030) was used as a reference strain where the region of genetic recombination occurred between 5,082 and 5,110 bp at the ORF1/ORF2 overlap region (Fig. 3A) [28]. Upstream of this junction, the nucleotide homology was notably different. Nucleotide sequence comparison showed that the capsid gene of the Gil-67 strain were almost identical with reference strains but shared distinctly different from the RdRp gene to the Arg320 strain (data not shown). Bootscan analysis of Gil-67 was performed to determine the genomic region involved in the recombination event at nucleotide position 5,082 of the PC03 strain as compared to that of the Arg320 strain (Fig. 3B). These strains were suspected to be recombinant NoVs according to the partial capsid and polymerase regions of one representative isolate, the Gil-67 strain. Compared with the Arg320 strain, the Gil-67 strain shared a low level of sequence identity (84%) in the RdRp gene region, but higher identity (95%) in the capsid region. In contrast, compared with the PC03 strain, Gil-67 shared a high level of sequence identity (95%) in the RdRp gene region and lower identity (65%) in the capsid region. A recombinant site was also observed at the overlap of ORF1 and ORF2 (Fig. 3B). Analysis of the full-length nucleotide sequences of the RdRp and capsid genes revealed that the Gil-70 and 4CAU-38 strains were GII.16/GII.2 recombinant strains (Fig. 4). The nucleotide sequences of the RdRp gene in Gil-70 and 4CAU-38 shared 96.7-99% identity with those of the NoV GII.16 reference strains OH10026, DHGX-2, E2116, and L53. In contrast, the capsid nucleotide sequences of Gil-70 and 4CAU-38 shared 97.2% identity with those of the NoV GII.2 reference strain Melksham (data not shown). SimPlot analysis indicated that the region of genetic recombination of these strains was found between 5,078 and 5,098 bp at the overlap of the ORF1 and ORF2 regions with the GII and GII.16/ GII.2 reference strains Melksham, L23Vannes, Neustrelit260, and DHGX-2 (Fig. 5).

GII.2 genotype
Analysis of the full-length nucleotide sequences of the RdRp and capsid genes revealed that the 4CAU-14 strain was a GII.2/GII.10 recombinant strain (Fig. 4). The nucleotide sequence of the RdRp gene in 4CAU-14 shared 93.5-100% identity with those of the NoV GII.2 reference strains 7137, Neustrelitz260, and OH08020. In contrast, the nucleotide sequence of the capsid gene in 4CAU-14 shared 95.3-100% identity with those of the NoV GII.10 reference strains 6952, Mc37, and Vietnam026 (data not shown). SimPlot analysis indicated that the region of genetic recombination of these strains was found between 5,078 and 5,098 bp at the overlap of the ORF1 and ORF2 regions with the GII and GII.2/GII.10 reference strains Mc37, Vietnam026, Neustrelitz260, and OH08020 (Fig. 6).

Discussion
In South Korea, NoV-related gastroenteritis has been a major public health concern since the identification of the virus in 2005, as evidenced by epidemiological studies of the outbreaks that occurred from contaminated groundwater on Jeju Island [18,29]. Here, we analyzed the prevalence of NoV infections among children hospitalized with acute gastroenteritis at three different localities in South Korea between January 2011 and March 2012. The prevalence of NoV infection in this study was approximately 34%, which was concordant with those of previous reports on NoV epidemiology worldwide [30][31][32][33]. NoV prevalence ranges from 10% to 60% or even more in pediatric patients hospitalized for acute gastroenteritis, indicating that NoV is an important enteropathogen causing diarrhea in Korean children.
Frequent global outbreaks of NoV-associated acute gastroenteritis have been reported since the late 1990s, and NoV has proven to be the etiological agent in many sporadic cases and outbreaks of gastroenteritis across the globe [34]. To date, three NoV genogroups, GI, GII, and GIV, have been recognized as the main contributors to common outbreaks and are associated with acute gastroenteritis infecting human populations of all ages [35]. In South Korea, while the NoV GII.1 genotype was frequently detected between 2001 and 2005 [17], GII.4 is most predominant genotype reported to date [18,29]. In this study, the most predominant genotypes were GII.4 (56.5%), followed by GII.6 (13.5%), GII.12 (13%), and GII.3 (11.8%). These data indicated that changes in the patterns of predominant NoV genotypes have occurred over time, suggesting that continuous The occurrence of a variety of different NoVs circulating within a population is a risk factor for recombination events, which in turn can lead to mixed infections [36]. These events strongly impact molecular epidemiological studies, viral vaccine design, and viral control programs [13]. The co-circulation of two potential parental strains may facilitate recombination when the nucleic acid sequences of the strains physically interact in infected cells during copy-choice recombination [37]. RNA recombination is responsible for a large proportion of viral diversity, which is influenced by various immunological and intracellular constraints, leading to the production of viable recombinants [38]. In 1999, Jiang and colleagues were the first to report a naturally occurring NoV recombinant, in which the recombinant site was found between the RdRp and capsid genes in the RNA region [16]. Since then, several recombinant strains have been sporadically reported around the globe. Between 2004 and 2007, 21 NoV recombinant types Novel Norovirus GII.2/GII.10 Recombination were identified; the NoV recombinant type GII.12/GII.13 was the first isolate identified in South Korea during this period [39][40][41]. In our current phylogenetic analysis of the RdRp/capsid regions, we identified 39 recombinant strains, which were then divided into seven different subtypes: GII.2/GII.10 (n51), GII.b/GII.3 (n53), GII.4/GII.2 (n51), GII.4/GII.3 (n58), GII.4/GII.6 (n512), GII.4/GII.12 (n512), and GII.16/GII.2 (n52). The NoV GII.16/GII.2 recombinant, reported in South Korea for the first time in this study, was the predominant strain in Japan during the 2009-2010 season [42]. Interestingly, the NoV GII.2/GII.10 recombinant (4CAU-14 strain) could be characterized as a novel intergenotype recombinant, which highlights the genetic variation in the RdRp/capsid genes among strains that have been circulating across the country. Therefore, understanding the recombination events occurring in NoV strain is important for tracking the emergence of new NoV variants [40].
Similarity Plots, otherwise known as SimPlots, can be used to recognize the percent similarity between the query sequences and a panel of reference sequences. Recombinant breakpoints were found between the overlap of ORF1 and ORF2 at nucleotide positions 5,078 and 5,098 bp for the GII.16/GII.2 (Gil-70 and 4CAU38) new recombinant strains and at nucleotide position 5,082 and 5,110 bp for the GII.b/GII.3 (Gil-67) recombinant strain. As seen in other recombinant NoVs, the results from the current study indicate that the region between the ORF1 and ORF2 junction may play a role in recombination events [27].
The findings of our present study support the hypothesis that certain NoV strains circulate in specific areas without clinical manifestations and may occasionally change their genetic properties through recombination events. Since recombination allows a virus to increase its genetic fitness, evolve, spread among the population, and probably escape the host immune response, our findings Novel Norovirus GII.2/GII.10 Recombination suggest that NoVs have a huge capacity for genetic change, which will facilitate the generation of new recombinants. NoV infections continue to be important gastroenteritis-associated pathogens infecting the Korean population, with GII.4-2006b variants currently being the most predominant strains. Here, we report the first finding of NoV GII.16/GII.2 recombinants in Korea, with GII.2/GII.10 being the first recombinant described thus far. Continuous surveillance is required to understand the continuously evolving status of NoVs infections in the Korean population.