Prevalence of noroviruses in children hospitalized for acute gastroenteritis in Hohhot, China, 2012-2017

Background Noroviruses (NVs) are an important cause of acute gastroenteritis (AGE) worldwide. There are limited data on the prevalence and molecular characterization of NVs in children in Hohhot, China. Methods Between January 2012 and December 2017, 1863 stool samples were collected at Maternal and Child Health Hospital in Hohhot. All samples were screened for NVs by real-time reverse transcription polymerase chain reaction (real-time RT-PCR). Results NVs were detected in 24.15% of these inpatient cases, ranging from 12.78 to 32.92% in different years. NV was detected throughout the year, with a peak in winter. Based on sequence analysis of the partial VP1 gene, the 306 identified NV strains were divided into six genotypes: GII.3 (71.24%), GII.4 (23.53%), and GII.2, GII.5, GII.6, and GII.13 (total 5.23%). Based on further sequence analysis of the RNA-dependent RNA polymerase (RdRp), GII.P12/GII.3, GII.Pe/GII.4, and GII.P4/GII.4 were identified as predominant genotypes, accounting for 92.6% of genotyped strains. The median age of the children with NV infection was 8.0 (range 0–59) months. However, children infected with GII.3 were younger (median 7.0 months) than GII.4-positive patients (median 10.0 months). Conclusion NV contributed greatly to AGE among hospitalized children in Hohhot in China. Continuous surveillance is important for understanding the local prevalence and characterization of NV.


Background
NVs are a leading cause of epidemic gastroenteritis [1], responsible for at least 90% of all non-bacterial gastroenteritis cases and 50% of all gastroenteritis outbreaks worldwide [2,3]. All age groups can be infected by NVs, especially children, the old, and the immunosuppressed. The acute gastroenteritis which caused by NVs has become the second infectious disease due to its highly transmission with fecal-oral route and lowly infectious dose [4,5]. In the developing countries, more than 200, 000 deaths per year are caused by NVs infection [6].
NVs are belonging to Caliciviridae, with a 7.5-kb-long ssRNA genome. Three different open reading frames (ORFs) encode different proteins of NVs [7]. Six nonstructural proteins are encoded by ORF1, such as p28, VPg, 3CL pro , and the structural proteins are encoded by ORF2 (VP1) and ORF3 (VP2) to form the capsid of NVs. [7]. Total seven genogroups (GI, GII, GIII, GIV, GV, GVI, and GVII) NVs are classified by the identities of VP1 and RdRp amino acid sequences of NVs [8,9]. Human beings are the major host to genogroups GI, GII, and GIV NVs, however, other genogroups NV are mainly infected with animals. The epidemic of human acute gastroenteritis are mainly caused by GI or GII NVs [8,10] and the recombination between ORF1 and ORF2 [11], antigenic variation of P domain of VP1 [12] are contributed to generate novel NV strains. The emerging of NVs leading to the epidemic of human acute gastroenteritis in the worldwide,at hence, it is important to determine both the capsid and RdRp genotypes for strain characterization.
Despite the large number of NV genotypes cocirculating in children over the past decade, genotypes GII.4 and GII. 3 have predominated worldwide [13][14][15][16][17]. Up to nowadays, six outbreaks of acute gastroenteritis are associated with at least seven distinct GII.4 NV variants [14]. Three major clusters (I-III) and five smaller genetic lineages (A-E) have been identified in the VP1 region of GII.3 NV [13,18]. Global surveillance data over the years indicated that the majority of outbreaks of NV were caused by GII.4, but the recent NV outbreak strains in Asia were GII.2 and GII.17 [1,19,20]. Furthermore, other genotypes were co-circulating, with different epidemiological patterns in different regions [1,17].
China is a vast country, and the prevalence of NV varies by region [1,19,21,22]. However, reports and knowledge of routine sporadic NV disease in northern China have been limited. Furthermore, in several recent years, rare NV genotypes (GII.2, GII.6, and GII.17) have caused prevalent outbreaks in China [19,20] and other parts of Asia [23,24]. Therefore, it is important to explore the NV strains in sporadic AGE cases, which may also aid in understanding the prevalence of human NV in the population. In the present study, we explored the epidemiological and molecular characteristics of NVs in children younger than 5 years hospitalized for AGE from 2012 through 2017 in Hohhot, China.

Specimen collection
Between January 2012 and December 2017, 1863 fecal samples were collected from children younger than 5 years admitted to Maternal and Child Health Hospital in Hohhot with acute gastroenteritis (AGE). AGE was defined as two or more loose stools within a 24-h period. Stool specimens were collected within 7 days after the onset of the disease. All AGE cases admitted to the hospital were enrolled. All specimens were sent to the Inner Mongolia Centers for Disease Control and Prevention for NV detection, and the NV-positive samples were sent to the Institute for Viral Disease Control and Prevention, China CDC, for further testing and sequencing. Demographic and clinical data were collected using a standard questionnaire. Oral informed consent was obtained from the guardians of all hospitalized children for samples and data collection.

RNA extraction and viral detection
Stool samples were prepared with phosphate-buffered saline solution to a final concentration of 10%. Viral RNA was extracted from 200 μL of a 10% stool suspension using the Viral Nucleic Acid Extraction Kit II (Geneaid, China), according to the manufacturer's instructions. The RNA was then subjected to one-step real-time RT-PCR targeting NV sequences at the ORF1/ORF2 junction for NV detection, as described previously [25].

Capsid and RdRp gene amplification
Following NV detection, positive samples were tested by conventional RT-PCR for amplification of the partial RdRp and VP1 gene regions of NV using the QIAGEN One-Step RT PCR Kit (QIAGEN). The 5′-end capsid gene (region C in ORF2) was amplified with oligonucleotide primer sets G1SKF/R and COG2F/G2SKR for 330 bp of the GI and 344 bp of the GII capsid genes respectively, as previously described [26]. The 3′-end of the RdRp gene (region A in ORF1) was amplified using a published oligonucleotide primer set, P290/P289 for a 319-bp fragment, as previously described [27]. The PCR products were visualized on agarose gels.

Sequencing and phylogenetic analysis
All PCR products were purified and sequenced by Invitrogen Trading (Shanghai). The sequence identity was searched using the Basic Local Alignment Search Tool (BLAST) (https://blast.ncbi.nlm.nih.gov/Blast.cgi) multiple sequence alignment, and a phylogenetic tree was constructed using MEGA 7 [28]. The dendrogram was determined using the neighbor-joining method in MEGA7. Bootstrap resampling (1000 replications) was used and bootstrap values ≥75% are shown. The partial nucleotide sequences determined in this study were submitted to GenBank (acc. no. MK319217 to MK319535).

Statistical analysis
Data were analyzed using SPSS 19.0 and Excel (MS office 2007). Numbers and percentages were computed for categorical variables. Categorical variables and proportions were compared using a two-sided chi-square test or Fisher's exact test where appropriate. The 95% confidence intervals (95% CI) were estimated using the binomial exact method and P-values < 0.05 were considered statistically significant.
NV infection can occur throughout the year (Fig. 2). Every year, NV infection had three peaks. The highest

Discussion
In the present study, we conducted long-term surveillance for NV-associated gastroenteritis in hospitalized children younger than 5 years of age in Hohhot, Inner Mongolia, China. From 2012 to 2017, the NV positivity rate was 24.15% in Hohhot, which was close to that in Wuhan (2007-2010, 25.9%), but higher than that in other parts of China [29]. This result indicates the high activity of NV in recent years in Hohhot. Furthermore, 93% (1741/1863) of AGE cases were children aged 0 to 23 months, and NV positivity was higher in the 0-23month age group than in children older than 24 months, which implies that young children are more vulnerable to NV infection. This result is consistent with previous studies [15,22,30,31], indicating that intervention approaches such as a vaccine for NV infection should be implemented for children under 24 months of age.
According to a systematic review of the global seasonality of NV, in the Northern Hemisphere NV usually peaks in cold seasons, whereas it does not exhibit a consistent summer or winter peak in the Southern Hemisphere [32]. In southeastern China, NV infections were detected mainly in the spring and summer [26], whereas in North China, NV infections occurred commonly in the winter [33,34]. However, in this study we found three peaks of NV infection per year in Hohhot. The highest peak was observed in winter, and two small peaks occurred in spring and fall. Hohhot is located in the northern part of China in the continental Mongolian plateau, with short, hot, rainless summers, and long, severe, cold, snowless winters. This differs from the climate in the other northern regions in China. The seasonal pattern of NV in Hohhot might be related to this unique local climate.
The oldest GII.3 NV strain identified to date is from a 1983 sample isolated in the Goulburn Valley of Australia [13], and GII.3 has been reported to be the major cause of sporadic pediatric NV infection [15,17,35,36]. In China, early reports on NV showed GII.3 was prevalent in 1999 [37,38], and until now, GII.3 was reported as the dominant genotype next to GII.4 in sporadic NV infection [39,40]. However, in this study, we found that NV GII.3 was the most prevalent strain among children hospitalized for AGE from 2012 to 2017. Hohhot, located in the North China Plateau, is vast and sparsely populated, and lags behind the developed cities of our country in socioeconomic development and the convenience of transportation. Population flow and economic trade in Hohhot are infrequent. During this period, the GII.P12/GII.3 recombinant was the most common genotype, and it was also detected more frequently in China and Japan than in other parts of the world [14,34,36]. The GII.3 strains identified in our study were grouped into two clusters (I and II) in a phylogenetic analysis of the capsid gene sequences. Lacking early NV surveillance data from Hohhot, it is not yet possible to determine whether GII.3/cluster II strains emerged and spread before 2012. However, our results suggested the hypothesis of an alternating epidemic pattern between GII.3/cluster I strains and GII.3/cluster II strains in Hohhot. Because only partial capsid gene sequences were obtained, rather than the complete VP1 region, or even the longer genome including the ORF1/ORF2 overlap region, we cannot clearly delineate the evolutionary   [19,21]. Previous studies show GII.4 usually caused epidemics in adult populations, whereas GII.3 NV was prevalent in young children and infants [45,46]. The reasons for the difference between GII.4 and GII.3 are still unknown. Perhaps a variety of factors account for it, including cross protection from heterologous strains or the minimal infectious dose necessary to cause disease. More studies are needed to address this issue.
Unlike neighboring cities and other parts of China, where emerging GII.P17-GII.17 NVs caused outbreaks during 2014-2015, no GII.P17-GII.17 was found in children hospitalized for AGE in Hohhot in 2015. Previous reports showed that infection with the emergent GII.P17-GII.17 was more likely to occur in older children and adults [47], which may be the reason for negative detection of GII.P17-GII.17 in Hohhot, with 93% of AGE cases among children ages 0 to 23 months. However, the exact mechanism for the emergence, prevalence, and age preference of GII.P17-GII.17 remains unknown.
NV recombination was genetically characterized via analysis of regions from both the capsid gene and RdRp gene. In our study, lacking the overlap genes from the RdRp and VP1 regions, the NV recombinants were not conformed and may actually have come from two different coinfecting NV genotypes. However, since 2018, a conventional RT-PCR, as developed by the US Centers for Disease Control and Prevention (Atlanta, GA, USA) (J. Vinjé, pers. comm., 2017 Mar 1, [48], has been used for amplification of partial RdRp and VP1 gene regions. This RT-PCR will be used in our future NV surveillance.

Conclusions
We found high NV prevalence in children younger than 5 years in Hohhot who were hospitalized for AGE. GII.3 NV was identified as the most common cause of AGE from 2012 to 2017, and circulating genotypes changed over time with the emergence of new genetic variants. Continuous monitoring with molecular genotyping is necessary for NV diseases in children and will be helpful for the local control and prevention of AGE.