Whole-genome analysis of rotavirus G4P[6] strains isolated from Korean neonates: association of Korean neonates and rotavirus P[6] genotypes

Background Group A rotaviruses are the major causative agents of pediatric gastroenteritis worldwide. Several studies have reported the predominance of G4P[6] rotavirus genotypes in Korean neonates, which is uncommon in other countries. Therefore, the purposes of this study were to determine the genotype constellations of complete genomes of G4P[6] rotavirus strains isolated from Korean neonates using next-generation sequencing, to compare these sequences with other G4P[6] strains in other countries, and to determine the reason for the predominance of G4P[6] genotypes in Korean neonates. Results Twenty rotavirus G4P[6] strains, isolated from January 2013 to January 2016, were selected for whole-genome sequencing. Eleven rotavirus genes were amplified using specific primer sets, and sequencing was carried out using an Ion S5 XL next-generation sequencing platform. Genotypes of each gene were determined, and phylogenetic analyses were performed to investigate genetic distances between genes of rotaviruses in this study and those of other rotavirus G4P[6] strains whose whole-genome sequences were previously published. All 20 rotavirus strains in this study had the same genotype: G4-P[6]-I1-R1-C1-M1-A1-N1-T1-E1-H1, representing the Wa-like genotype constellation. BLAST searches of 20 G4P[6] rotavirus strains revealed that all G4 sequences in this study showed the highest nucleotide identity to G4 sequences of G4P[6] rotavirus strains isolated in Korea in 2008 (GenBank accession number: FJ603447). Additionally, P[6] gene sequences in this study showed the highest nucleotide identity to P[6] sequences of G4P[6] strains detected in Korea in 2002 (AY158093). Phylogenetic and nucleotide sequence analyses showed that G4P[6] strains in this study and previously reported G4P[6] strains in Korea were mostly detected in neonates and had similar G4 and P[6] sequences compared with other G4P[6] strains detected in other countries. Conclusions This study showed that the whole-genome constellation of rotavirus G4P[6] strains from Korean neonates resembled a Wa-like genotype constellation. Additionally, rotavirus genotypes detected in Korean neonates had unique P[6] sequences, which may be the cause of Korean neonatal rotavirus infection. Electronic supplementary material The online version of this article (10.1186/s13099-019-0318-5) contains supplementary material, which is available to authorized users.

Next-generation sequencing (NGS) technology has recently been applied to viral genome research and human genome research [14]. NGS can generate large amounts of viral sequence data simultaneously within a short time through massively parallel sequencing. NGS technology reduces time, effort, and cost compared with conventional Sanger sequencing techniques, particularly when sequencing many genes or dealing with large numbers of samples.
In this study, we attempted to determine the genotype constellation of the complete genome of G4P [6] rotavirus strains characteristically isolated from Korean neonates using NGS and to compare the sequences of Korean G4P [6] strains with G4P [6] strains in other countries. In addition, we tried to determine the reason for the predominance of G4P [6] genotypes in Korean neonates.
We performed nucleotide sequence and phylogenetic analyses of the genotypes G4, P [6], I1, R1, C1, M1, A1, N1, T1, E1, and H1 among the strains in this study and previously reported G4P [6] strains with whole-genome sequences registered in GenBank [8,10,18]. For the 20 strains reported in this study, there were 98-100% sequence similarities among the same genes in all 20 rotavirus strains. However, there were 82-96% sequence similarities between the G4 gene of the RN-001 sample and the G4 genes detected in other countries. There were also lower sequence similarities of 84-95% for P [6], 83-96% for I1, 85-94% for R1, 85-94% for C1, 85-91%    (Table 2). In contrast, the G4 and P [6] genotypes in this study were more similar to the G4 and G [6] genotypes previously reported in Korea. Interestingly, we recently reported that G8P [6] genotypes were also found in neonates in the neonatal intensive care unit of the same hospital as this study [15]. These P [6] nucleotide sequences of G8P [6] genotypes were not different from the P [6] gene sequences of the G4P [6] strains in this study, and these P [6] sequences of G8P [6] and G4P [6] strains also showed higher identity with the nucleotide P [6] gene sequences of G12P [6] and G2P [6] strains (Gen-Bank no. AY158093) in Korea [18] than with the P [6] sequences in G4P [6] strains detected in other countries (Fig. 2). We investigated whether the G4P [6] strains in this study were related to porcine G4P [6] strains because several papers have provided molecular evidence that many G4P [6] strains are human-porcine RVA reassortants or even porcine RVA having directly infected children [19][20][21]. All G4 sequences of G4P [6] strains in this study were more similar to G4 sequences of Korean G4P [6] or G4P [8] strains in previous studies than to G4 sequences of G4P [6] strains in other countries or porcine G4P[6] strains (Fig. 1). Similarly, all P[6] sequences of G4P [6] strains in this study were more similar to P [6] sequences of Korean G4P [6], G8P [6], or G12P [6] strains in previous studies than to P [6] sequences of G4P [6] strains in other countries or porcine G4P[6] strains (Fig. 2). Therefore, the G4P[6] strains reported since 1999 in Korea can be considered endemic G4P [6] strains in Korea, not strains imported from other countries. In addition, analyses of the VP6, NSP4, and NSP5/6 genes of G8P [6] strains in a previous study showed I2, E2, and H2 genotypes, indicating the DS-1-like constellation rather than the Wa-like constellation [15]. Therefore, these new rotavirus G8P [6] strains in Korea were estimated to be derived from reassortment events between the G8-P[8]-I2-R2-C2-M2-A2-N2-T2-E2-H2 strains imported from the Asian region and the P [6] gene of endemic G4 [6] strains detected in Korea [15]. The phenomenon that all 20 G4P [6] strains in this study showed the same genotype constellation (G4-P[6]-I1-R1-C1-M1-A1-N1-T1-E1-H1) and high genetic similarities suggested the possibility of persistent infection with the same rotavirus strain over 3 years in one hospital. However, 11 of the 20 G4P [6] rotavirus cases were detected on the first admission day and were transferred from other hospitals or clinics, indicating the occurrence of outside infection because rotavirus infection requires an incubation period for at least 2 days. Additionally, G4P [6] rotavirus infection in Korean neonates has been reported in several studies in other cities in Korea since 1999 [9][10][11][12][13], suggesting that rotavirus G4P [6] infection is not a local phenomenon occurring only at one hospital, but could occur throughout all of South Korea.
Both genotypes of G8P [6] and G4P [6] were frequently detected in Korean neonates, and sequence similarities were observed between P[6]s in G8P [6] strains and P [6]s in G4P [6] strains, whereas differences were found in P [6] sequences from G4P [6] strains detected in other countries. These findings suggested that selective infection by rotaviruses with these unique P [6] sequences occurred in Korean neonates. Moreover, previous reports have demonstrated that the VP8 portion of VP4 attaches to the human blood group antigen (HBGA) in the intestinal epithelium and that there is an association between the antigenicity of VP4 (VP8) and HBGA [25]. Therefore, unique P [6] sequences and the unique antigenicities of G8P [6] and G4P [6] strains may be related to HBGA in the intestinal epithelium in Korean neonates. Further studies are needed to determine the mechanism through which P [6] genotypes easily infect Korean neonates. Current rotavirus vaccination programs (e.g., RotaTeq or Rotarix), which begin after 6 weeks of age, cannot prevent neonatal rotavirus infection [9]. However, a recently developed neonatal rotavirus vaccine (RV3-BB, G3P [6]), which has P [6] antigenicity and is first given 0-5 days after birth, may be effective against Korean neonatal rotavirus G4P [6] infection [26].

Patient samples
Rotavirus-positive stool samples were collected from neonates younger than 1 month of age in a 650-bed hospital from January 2013 to January 2016. Twenty G4P [6] rotavirus-positive samples were successfully genotyped for whole-gene genotyping using NGS (11 specimens in 2013, six specimens in 2014, two specimens in 2015, one specimen in 2016). During this period, 270 rotavirus antigen-positive samples from neonates with symptomatic diarrhea were collected, and 56 samples were arbitrarily selected for this G4P [6] whole-genome sequencing study. Forty-nine samples from these 56 samples (87.5%) were genotyped as G4P [6] strains using G and P typing (seven samples were non-G4P [6] strains). Of 49 G4P [6] strains, 20 samples were successfully amplified for all 11 rotavirus genes evaluated in whole-genome sequencing. Clinical data, including age and sex, were collected from patient medical records. Eleven (55.0%) samples were collected from males, and the overall median age of the donors was 11 days (range 5-28 days). This study was approved by the Institutional Review Board of Hallym University Dongtan Sacred Heart Hospital (IRB nos. 2013-030, 2017-08-007).

Whole-genome sequencing of rotaviruses using NGS
Whole-genome sequencing of rotaviruses was carried out using reverse transcription polymerase chain reaction (RT-PCR) and NGS. Viral RNA was extracted from fecal suspensions using a QIAamp Viral RNA Mini kit (Qiagen, Hilden, Germany) and the QIAcube platform (Qiagen). The RNA was denatured and reverse transcribed using the SuperScript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA). Eleven rotavirus genes were amplified from the double-stranded RNA genome using specific primer sets described in Additional file 10: Table S1 [27]. All 20 RT-PCR products for each genome were pooled in equimolar amounts, sheared using an Ion Xpress Plus Fragment Library Kit (Thermo Fisher Scientific, Waltham, MA, USA), and then ligated to barcoded adaptors using Ion Express Barcode Adapter kits (Thermo Fisher Scientific), to create about 300-bp sized fragment libraries. Template preparation, including emulsion PCR, was performed using Ion 510 and Ion 520 and Ion 530 kit-Chef (Thermo Fisher Scientific) and an Ion Chef system (Thermo Fisher Scientific). NGS was performed using the Ion Torrent S5 XL NGS platform (Thermo Fisher Scientific) and Ion S5 Sequencing kit on a 520 chip. Sequenced reads were quality checked and trimmed using Ion Torrent Suite version 5.0.4. Raw sequence data were processed using the CLC genomics workbench (http://www.clcbi o.com/). Sequenced reads were trimmed and mapped to the rotavirus reference sequence (ASM265499v1 or ASM268153v1), and consensus sequences of each gene were obtained. Because we could not obtain the sequences of VP7 genes by NGS, VP7 genotyping was carried out using RT-PCR and Sanger sequencing with another specific primer set (46F/911R; Additional file 10: Table S1).

Rotavirus genotypes and constellation
The genotypes of gene sequences were obtained using the Rota C v2.0 online automated genotyping tool [28], and whole-genome constellations were obtained. The closest nucleotide sequences to each gene were obtained using the Basic Local Alignment Search Tool (BLAST) on the National Center for Biotechnology Information (NCBI) website. Sequence similarities between the genes in this study and other G4P [6] strains with whole-genome sequence data in GenBank were compared using BLAST on the NCBI website.