Molecular Epidemiology of Human Sapovirus among Children with Acute Gastroenteritis in Western Canada

Study participants and specimen acquisition.

Ethics approval was obtained from the University of Alberta and the University of Calgary research ethics boards. The Alberta Provincial Pediatric EnTeric Infection TEam (APPETITE) study protocol has been previously published (16). Briefly, children less than 18 years of age with >3 episodes of vomiting or diarrhea within 24 h and <7 days of symptoms were eligible to participate and were classified as AGE cases. Children were recruited in two urban, pediatric tertiary care emergency departments (ED) (Alberta Children’s Hospital, Calgary, and Stollery Children’s Hospital, Edmonton) and through Health Link (Alberta Health Services), a province-wide nursing advice telephone line serving the province of Alberta, Canada. Controls with no infectious symptoms (i.e., no vomiting, diarrhea, fever, rhinorrhea, or cough) were also enrolled in the ED and at a public health clinic attended by children presenting for their routine infant immunizations in Calgary, Alberta. Consent was obtained from the caregivers of eligible children, along with assent when appropriate. Children enrolled in the previous 14 days, those who could not complete follow-up, and those with preexisting psychiatric illness or neutropenia or in need of emergent medical intervention were ineligible.

Rectal swabs and stool specimens from AGE and control cases were collected between December 2014 and August 2018. FLOQSwabs (Copan Italia, Brescia, Italy) were inserted into the rectum, rotated 360°, and then placed into sterile sheaths. Stool specimens were collected in fecal collectors (V302-F; Starplex Scientific, Ontario, Canada). Specimens collected in the ED were maintained at 4°C and delivered to the laboratory within 24 h of collection and immediately frozen at −70°C. Stool samples not provided prior to discharge from ED, and samples from participants recruited through Health Link and the public health clinic were collected at home and transported by a study-funded courier in coolers with ice packs to the laboratory for processing and freezing within 12 h of collection.

Nucleic acid extraction and molecular pathogen detection.

Total nucleic acid was extracted from a 60% rectal swab or 10% stool suspension in NucliSENS lysis buffer (bioMérieux, Laval, Quebec, Canada). Nucleic acid was extracted and eluted in 70 μl using the NucliSENS easyMag system (bioMérieux, Laval, Quebec, Canada) according to the manufacturer’s instructions. All resultant nucleic acid extracts were immediately stored at −70°C until testing. Viral cDNA was generated with the SuperScript II (Thermo Fisher Scientific, MA, USA) reverse transcription system as previously described (17).

Sapovirus was detected using an in-house multiplex reverse transcription–real-time quantitative PCR (RT-qPCR) assay that also detects norovirus (genogroups I and II), group A rotavirus, adenovirus (all serotypes), and astrovirus (18). The primers and probes for sapovirus detection were originally designed by Oka et al. in 2006 (19). Threshold cycle (C_T) values were used as a semiquantitative measure that inversely correlates with viral load in the specimen. The C_T value cutoff for a positive result was set at ≤38. Specimens also underwent enteropathogen detection using the Luminex Gastroenteritis Pathogen Panel (GPP), which detects three viruses (norovirus GI and GII, rotavirus, and adenovirus 40/41), nine bacteria (Clostridioides difficile, Escherichia coli O157, enterotoxigenic E. coli [ETEC], Shiga toxin-producing E. coli [STEC], Salmonella, Campylobacter, Vibrio cholerae, Shigella, Yersinia enterocolitica) and three parasites (Cryptosporidium, Giardia, Entamoeba histolytica) (17). Rectal swabs and stool specimens collected from the same patient were tested individually with the Gastroenteritis Virus Panel (GVP) and GPP.

Seasonality was assessed using the following definitions: for winter, 1 December to 28 February; spring, 1 March to 31 May; summer, 1 June to 31 August 31; and autumn, 1 September to 30 November. Average monthly temperature data were obtained by courtesy of Environment and Climate Change of Canada on the Edmonton weatherstat website (https://edmonton.weatherstats.ca/metrics/temperature.html).

Sapovirus genotype determination.

The genotype of sapovirus isolates was determined in only one of the two specimen types (the specimen with the lower C_T value) collected from each sapovirus-positive participant. To identify the genotype for each sapovirus-positive sample, the cDNA underwent nested PCR amplification with primers SV-F13, SV-F14, SV-R13, SV-R14, SV-F22, and SV-R2 to generate amplicons of approximately 420 bp (20) (see Table S1 and Fig. S1 in the supplemental material). The PCR products were subjected to gel electrophoresis with 2% agarose followed by excision of the product with the expected size to undergo purification via the Qiaquick gel extraction kit (Qiagen, Germantown, MD, USA). The nucleotide sequences of each sample were retrieved using the BigDye Terminator v3.1 cycle sequencing kit (Life Technologies, Austin, TX, USA). The Sanger chromatograms were analyzed using MEGA X 10.0.5 to obtain the sequences, and the NCBI nucleotide BLAST search engine was used to assign genotypes (21, 22). The sapovirus genotyping results were verified using the Calicivirus Typing Tool (https://norovirus.ng.philab.cdc.gov) (23).

Nucleic acids of sapovirus-positive stool samples that were not successfully genotyped by the above method were reextracted using the ReliaPrep RNA Cell Miniprep system (Promega, Madison, WI, USA). Viral RNA was reverse transcribed using the SuperScriptII reverse transcription system as described above, and the cDNAs were subjected to another heminested PCR amplification using primers SLV5317, SLV5749, SVpol3′-A, and SVpol3′-B to generate amplicons of 371 bp for resequencing through agarose gel electrophoresis and gel purification (24) (Table S1 and Fig. S1).

Disease severity and participant clinical characteristics.

Clinical disease severity was assessed using the modified Vesikari scale (MVS) score (25, 26). The MVS is a composite measure that includes seven components that are each assigned a score based on data collected between symptom onset and resolution. The components include (i) duration of diarrhea, (ii) maximal number of episodes of diarrhea in any given 24-h period, (iii) duration of vomiting, (iv) maximal number of episodes of vomiting in any given 24-h period, (v) maximal participant temperature recorded, (vi) presence and type of unscheduled visits to a health care provider, and (vii) type of treatment, if any, received for gastroenteritis. MVS scores can be further categorized into severity groups as mild (0 to 8), moderate (9 to 10), and severe (≥11) (26).

We compared the MVS scores of participants with different sapovirus genogroups and genotypes. The description of illness collected included the presence of fever, duration of diarrhea and vomiting in hours, modified Vesikari score (MVS) (26), and presence of dehydration measured by clinical dehydration scores reported by parents at the index ED visit. The number of codetected pathogens, age in months, and sex were also included in the analysis.

Phylogenetic analysis of VP1-encoding region.

Partial VP1 sequences that were longer than 210 nucleotides were included in this analysis and compared to 143 near-full-length sapovirus reference sequences that were downloaded from GenBank. All sequences were analyzed and aligned with software package MUSCLE in MEGA X. A phylogenetic dendrogram was constructed using the maximum likelihood method, based on the Kimura two-parameter model with a discrete gamma distribution (1,000 bootstraps) in MEGA version X (21, 22).

Statistical analysis.

Data were summarized with frequencies and percentages for categorical variables and medians and interquartile ranges (IQRs) for continuous variables. Categorical variables were compared using the χ² test or Fisher’s exact test where appropriate. Assay concordance was analyzed with Cohen’s kappa coefficient. Continuous variables were compared using the Kruskal-Wallis test. We calculated two-tailed P values and set the significance level α at 0.05. Descriptive statistics were computed using the SPSS software 26 (27) and GraphPad Prism v8.4.2 (GraphPad Software, San Diego, California, USA; www.graphpad.com).

Accession number(s).

The partial VP1 sequences described above can be accessed in GenBank under accession no. MZ420980 to MZ421251.

Prevalence and diversity of sapovirus strains in children.

In total, 3,347 children with AGE and 1,355 controls submitted specimens. Sapovirus was detected in 9.5% (317/3347) and 2.9% (39/1355) of participants from each group, respectively (Fig. 1). Partial capsid gene sequencing was successful in 85.2% (270/317) of samples from children with AGE; the remaining 14.8% (47/317) did not yield a successful typing result (Fig. 1). Typing was unsuccessful in 17 (43.6%) controls. All known human-pathogenic genogroups (GI, GII, GIV, and GV) were identified. The most common genogroup detected in children with AGE was GI (56.8%; 180/317); GI and GII were detected with equal frequency (28.8%; 11/39) as the leading sapovirus genogroups among controls. Eleven genotypes were identified in the present study (Table 1). Among children with AGE, sapovirus GI.1 (37.2%; 118/317) was the most common genotype, followed by GI.2 (18.9%; 60/317). All other genotypes detected accounted for <10% of detections (Table 1). The concordance between rectal swabs and stool specimens for sapovirus positivity was high (kappa = 0.907). Norovirus genogroups I and II and rotavirus were detected by both the Luminex GPP and the in-house GVP (kappa = 0.87, 0.94, and 0.98, respectively, indicating excellent agreement between the GPP and GVP). Because the GPP identifies adenovirus 40/41 and the GVP detects all serotypes of adenovirus, it was not possible to directly compare the two in adenovirus detection rate.

Among children with AGE, sapovirus was detected all year-round, with positive rates near 20% in the late autumn and winter and close to 2% in the summer (Fig. 2). Sapovirus GI.1 was the predominant strain in 2015 (13%) and 2017 (11%), while genotype GI.2 had a higher prevalence (9%) in 2016, with a high rate of sapovirus AGE persisting into spring 2017 (Fig. 2; Table 1). Common genotypes GI.1, GI.2, GII.1, and GV.1 were detected in all 4 years, while GII.2 and GII.5 circulated in three (Table 1).

Among the 270 samples from children with AGE and 22 from controls that were successfully genotyped, 93.5% (273/292) of the sequences were suitable for phylogenetic analysis, with 92.3% (252/273) and 7.7% (21/273) produced by the methods of Okada et al. (20) and Monica et al. (24), respectively. Success rates in obtaining usable sequences belonging to genogroups I, IV, and V, as well as GII.1 and GII.4, were greater than 90%, whereas the proportions of usable GII.2 and GII.3 sequences were 71.4% and 62.5%, respectively (Table 1). Partial VP1 sequences of GI.1 from this study clustered with reference sequences from east Asia (China, Japan, and South Korea), the Americas (the United States and Peru), and Europe (Germany and the United States) collected between 2013 and 2016, as well as a historical strain from Sapporo, Japan, collected in 1982 (see Fig. S2 in the supplemental material). GI.2 from the present study clustered with reference sequences from Germany, the United States, and Brazil collected in recent years (2009 to 2016) and a Bangladesh sequence from 2005. GI.3 sequences clustered with reference sequences from Japan (2008) and Kenya (2006); GI.7 clustered with strains from Japan (2008). The same multicontinent circulation was observed for the GII.1, GII.2, GII.3, GII.5, and GIV.1 strains. On the other hand, GII.4 was found to cluster with reference sequences from Peru (2016) and Brazil (2011) and GV.1 with reference sequences from the United States (2014) (Fig. S2).

Clinical severity of sapovirus infections in children.

There were no differences in the MVS scores, clinical characteristics, or C_T values among the AGE cases when comparing the various sapovirus genogroups and genotypes (Fig. 3A and B; see Fig. S3 and Tables S2 to S4 in the supplemental material). Children with AGE presenting to emergency departments (ED) had more severe gastroenteritis symptoms (P < 0.001) than patients receiving care at home (Fig. 3C; see Table S5 in the supplemental material). Of the 231 cases enrolled at an ED who tested positive for sapovirus, 97.2% were discharged, none was admitted to the intensive care unit (ICU), and there were no deaths. There were no differences in sapovirus gastroenteritis severity between different age groups (Fig. 3D).

Codetections of sapovirus with other enteropathogens.

Among children with AGE who tested positive for sapovirus, 35% (111/317) had one or more codetected enteropathogens: there was one other pathogen in 29.3% (93/317), and there were two and four other pathogens in 5.4% (17/317) and 0.3% (1/317), respectively. Adenovirus, norovirus, and rotavirus were the most commonly codetected enteropathogens (Table 2). Codetection of sapovirus and other enteropathogens was found in 35.9% (14/39) of controls, with adenovirus being the most common (Table 2).

Recurrent sapovirus infections in children.

We observed repeated sapovirus infections during the study period in three children who were enrolled initially as controls and subsequently developed AGE (Table 3). Patient 1 was positive for sapovirus with an unknown genotype (C_T = 37.7) in October 2015 and had a GI.1 sapovirus (C_T = 14.2) detected when symptomatic in November 2016. Patient 2 was positive for GI.1 sapovirus (C_T = 32.7) in December 2015 and GII.5 sapovirus (C_T = 20.0) in November 2016. Patient 3 was positive for both GII.5 sapovirus (C_T = 29.5) and adenovirus species C genotype 2 (28) in December 2015 and GI.2 sapovirus (C_T = 15.3) and GII norovirus in November 2016.