Co-Infection by Waterborne Enteric Viruses in Children with Gastroenteritis in Nepal

Enteric viruses are highly contagious and a major cause of waterborne gastroenteritis in children younger than five years of age in developing world. This study examined the prevalence of enteric virus infection in children with gastroenteritis to identify risk factors for co-infections. In total, 107 stool samples were collected from patients with acute gastroenteritis along with samples of their household drinking water and other possible contamination sources, such as food and hand. The presence of major gastroenteritis-causing enteric virus species (group A rotaviruses, enteroviruses, adenoviruses, and noroviruses of genogroup I) in stool and water samples was examined using quantitative polymerase chain reaction. Among the 107 stool samples tested, 103 (96%) samples contained at least one of the four tested enteric viruses, and the combination of group A rotaviruses and enteroviruses was the most common co-infection (52%, n = 54/103). At least one viral agent was detected in 16 (16%) of 103 drinking water samples. Identical enteric viruses were detected in both the stool and water samples taken from the same patients in 13% of cases (n = 13/103). Group A rotaviruses were most frequently found in children suffering from acute diarrhea. No socio-demographic and clinical factors were associated with the risk of co-infection compared with mono-infection. These less commonly diagnosed viral etiological agents in hospitals are highly prevalent in patients with acute gastroenteritis.


Introduction
Acute gastroenteritis is one of the leading causes of morbidity and mortality in developing countries such as Nepal [1][2][3][4][5][6]. Usually, gastroenteritis caused by enteric viruses: group A rotaviruses (RVAs), noroviruses (NoVs), astroviruses, and adenoviruses (AdVs) 40 and 41 comprise a significant proportion of gastroenteritis cases in developed as well as in developing countries [7,8]. Enteric viruses are a group of viruses that can cause acute watery diarrhea and are a major threat to human health worldwide [9]. They are usually prevalent in countries with issues of poor hygiene and sanitation [10]. Each year, more than 1.4 million children die as a consequence of waterborne gastroenteritis [11], with a high proportion of mortality caused due to the lack of timely intervention [12,13]. Despite gastroenteritis being easily preventable and treatable, a higher number of children suffer from the

Ethical Review
For the enrollment of the acute watery diarrheal children from Kanti Children's Hospital, ethical approval was taken from the Ethical Review Board of Nepal Health Research Council (reference number 925), Kathmandu, Nepal. An informed written consent from the caretakers of the children was taken and documented in the questionnaire form before stool specimen collection.

Collection and Processing of Fecal Samples
Stool samples were collected with a spatula in a clean, dry, disinfectant-free, wide mouthed screw-capped container. Prior to collection of the stool samples, patients were provided with instructions for sample collection technique and were requested to provide 20 g of stool specimen in the container. A total of 107 stool samples were collected during the study period. Ten percent fecal suspensions were prepared in phosphate-buffered saline and preserved at −25 • C until DNA and RNA extraction. The fecal samples were analyzed for enteric viruses but not for other fecal indicator bacteria and parasites.

Collection and Processing of Hand Swabs, Water, and Food Samples
Beside stool samples, other samples, such as hand swabs (n = 97), drinking water (n = 103), and food (n = 22), were also collected from the enrolled children. During the questionnaire survey, detailed demographic information, socio-economic data, contact number, sanitation, and hygiene data were taken. Hand swabs of the caretakers or children, depending on whether the children were fed by their caretaker or could eat by themselves, were collected. The hand swabs were collected using a BM Fukitool A kit (GSI Creos, Tokyo, Japan) containing 10 mL of phosphate-buffered saline and a cotton swab. The swab was dipped into the buffer. Drinking water and other possible sources of infection such as fruits and vegetables (food sources) were also collected from the household of each patient. Eleven liters of water samples was collected in 1 L sterile, well-labelled, screw-capped plastic bottles. The water samples were de-chlorinated with 1/100 volume of 5000 mg/L sodium thiosulfate. Similarly, 20 g of food sample was also taken from the children's house in a well-labelled plastic sandwich bag (27.3 × 26.8 × 0.06 cm). All the collected samples were transported to the Public Health Research Laboratory, Institute of Medicine, Tribhuvan University, maintaining the cold chain.
Ten grams of food samples were weighed and washed in 100 mL of normal saline (0.85 g of sodium chloride in 100 mL of sterile de-ionized water) in a plastic sandwich bag. Subsequently, the food samples were rinsed with normal saline, which was then mixed uniformly by shaking the plastic sandwich bag for about 10 min. The final volume obtained from washed food samples was approximately 100 mL. The water, food, and hand swab samples collectedwere first analyzed for fecal indicator bacteria (total coliforms and E. coli) by the most probable number (MPN) method using a Colilert reagent (IDEXX Laboratories, Westbrook, CA, USA). For food and hand swabs, E. coli concentration values were expressed as MPN/g and MPN/hands, respectively, depending on the sample type.

Concentration of Viruses in Water Samples
Assessment of the concentration of protozoa and viruses in water samples was done using the electronegative membrane-vortex method [27], as described in a previous study [28]. In brief, 10 L of the drinking water was mixed with 100 mL of a 2.5 mol/L MgCl 2 solution and then passed through a mixed cellulose membrane filter (diameter = 90 mm; pore size = 0.8 µm). The membrane was vigorously vortexed with an elution buffer containing 0.2 g/L Na 4 P 2 O 7 10H 2 O, 0.3 g/L C 10 H 13 N 2 O 8 Na 3 3H 2 O, and 0.1 mL/L Tween brand polysorbate 80 in a 50 mL plastic tube, followed by centrifugation at 2000× g for 10 min at 4 • C. The supernatant was concentrated further using a Centriprep YM-50 ultrafiltration device (Merck Millipore, Burlington, VT, USA) according to the manufacturer's protocol.

Detection of Viruses
Two hundred microliters of both viral concentrate and 10% fecal suspension were used to extract viral DNA using a QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany). Viral RNA was also extracted from 140 µL of each viral concentrate and 10% fecal suspension using a QIAamp Viral RNA Mini Kit (QIAGEN).
Extracted RNA was subjected to reverse transcription using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer's instruction. AdVs, EVs, NoVs-GI, and RVAs were determined quantitatively using qPCR from water and stool samples. DNA/cDNA (2.5 µL) was mixed with 22.5 µL of a PCR mixture containing 12.5 µL of Probe qPCR Mix (Takara Bio, Kusatsu, Japan), 0.4 pmol/µL each of forward and reverse primers, and 0.2 pmol/µL of a TaqMan probe. Subsequently, PCR tubes containing the mixtures were placed in a Thermal Cycler Dice Real Time System TP800 (Takara Bio) and incubated at 95 • C for 30 s, followed by 45 cycles of 95 • C for 5 s and 58 • C for 30 s for AdVs [29] and NoVs-GI [30], or 60 • C for 60 s for EVs [31,32] and RVAs [33]. Both positive (artificially synthesized plasmid DNA) and negative controls (PCR-grade water) were run together with the samples during qPCR analysis. Ten-fold serial dilutions of the positive control were used to generate the standard curves. Samples whose cycle threshold (Ct) value was greater than 40 were considered negative.

Process Control
To determine if inhibition occurred during qPCR amplification, artificially synthesized plasmid DNA-containing sequences amplified by qPCR assays for chicken parvovirus [34] and porcine teschovirus [35] were used as a process control as described previously [36]. These two viruses were selected because they are not commonly present in human stool and drinking water. In brief, 2.5 µL of cDNA was inoculated with 22.5 µL of a reaction buffer containing 5 × 10 4 copies of plasmid DNA of chicken parvovirus or porcine teschovirus, 12.5 µL of Probe qPCR Mix, 0.4 pmol/µL each of forward and reverse primers, and 0.2 pmol/µL of a TaqMan probe. The thermal conditions were as follows: 95 • C for 30 s, 45 cycles of 95 • C for 5 s, and 56 • C for 30 s. PCR-grade water was added to a qPCR tube instead of cDNA, which was considered to be a non-inhibition control. The qPCR amplification efficiency was calculated from ratio of the copy number of plasmid DNA in the sample qPCR tube to that in the non-inhibition control tube. The calculated efficiencies for stool and water samples ranged from 55 to 158% and from 65 to 195%, respectively, indicating that there was no inhibition during qPCR.

Statistical Analysis
IBM SPSS Statistics Version 20.0 (IBM Corporation, Armonk, NY, USA) was used to perform the statistical analysis. Differences in percentages between categories were compared using the chi-square test. The factors associated with co-infection were analyzed further using univariable logistic regression, and an odds ratio with a 95% confidence interval (CI) was calculated. The observed odds ratio gives an indication of the risk of co-infection compared to the reference category. A p value <0.05 was considered statistically significant. Multiple regression models were not run because of the limited sample size. Table 1 summarizes the distribution and concentrations of enteric viruses according to age groups. In total, 103 out of 107 samples (96%) were positive for at least one of the four enteric viruses tested. RVAs were detected in 91% of the samples (n = 97), EVs were detected in 55% of the samples (n = 59), AdVs were detected in 28% of the samples (n = 30), and NoVs-GI were found in 5% of the samples (n = 5). There was no significant difference in positive percentages between female (n = 46/47, 98%) and male children (n = 57/60, 95%) (p > 0.05). The highest number of samples (50/107, 47%) was from the age group of 0-11 months. In the age groups of 24-35, 36-47, and 48-59 months, positive percentages were 100%. The mean of the total enteric virus concentrations, which was calculated as the arithmetic sum of the four enteric viruses tested, was 6.9 ± 2.1 log copies/g (n = 103).  Table 2 summarizes the results of detection of enteric viruses in drinking water samples. At least one of the four enteric viruses tested was detected in 16 out of 103 drinking water samples (16%). EVs were detected most frequently in water samples (9/103, 9%), followed by RVAs (5/103, 5%), AdVs (4/103, 4%), and NoVs-GI (4/103, 4%,). Arithmetic mean of the total enteric virus concentrations was 3.9 ± 3.8 log copies/L (n = 16).

Detection of Fecal Indicator Bacteria from Hand Swab and Food Samples
As summarized in Table 5, E. coli was detected in 14% (n = 14/97) of hand swab samples and 45% (n = 10/22) of food samples, with concentrations of 1.9 ± 3.0 log MPN/hands and 2.7 ± 3.9 log MPN/g food, respectively. Total coliforms were observed in 63% (n = 61/97) of the hand swab samples and 72% (n = 16/22) of the food samples, with a concentration of 1.9 ± 3.5 log MPN/hands and 3.9 ± 4.0 log MPN/g, respectively.  Table 6 shows the relationship between co-occuring viruses and demographic, clinical, and environmental factors. Odds ratios with 95% CI are presented to show the magnitude and uncertainty of the estimates. Univariable logistic regression showed no significant association between the factors studied and the outcome of co-occurance (p > 0.05). Univariable logistic regression suggested that the risk of co-occurring infections of two or more organisms was much higher if food samples from a child's home contained E. coli (p = 0.03). a The total percentages in each factor group may not be equal to 100% due to missing data. b Water from the following sources: Shallow tube well, spring, stone spout, and tanker. c Only nine vegetables samples related to children with single enteric virus occurrence and 10 vegetable samples related to children with co-occurrence were tested for E. coli and these were used to calculate the proportion with detected or not detected E. coli in the above table.

Discussion
In this study, enteric viruses were found to be a major causative agent of gastroenteritis in children suffering from diarrhea. Our analysis demonstrated an overall prevalence of enteric viruses of 96%, which was much higher than the figures reported in previous studies conducted in Korea (44%) [37], Venezuela (59%) [38], Greece (39%) [39], Brazil (11%) [40], and Germany (59%) [41]. High positive percentage of enteric viruses might be caused by poor hygiene practices and inadequate sanitation infrastructures. The etiological agents for viral gastroenteritis have not been recorded in clinical laboratories of Nepal; therefore, a comparison of prevalence data of viral agents is nearly impossible. However, there have been studies on the prevalence of rotaviruses and other enteric viruses in Nepal within privately funded projects [1,3,4,6,[42][43][44][45][46]. These studies showed that children under five years of age frequently suffered from viral gastroenteritis.
RVAs appear to be a major causative agent of gastroenteritis in Nepal, and the prevalence may range from 15 to 45% of the total identified gastroenteritis cases [3,4,6,[42][43][44][45]. Similarly, the current study showed a very high prevalence of RVAs (91%), followed by EVs (55%), AdVs (28%), and NoVs-GI (5%) in children with gastroenteritis. The higher detection rates of viral agents may be attributed to the use of highly sensitive qPCR. Previous studies used comparatively less sensitive enzyme-linked immunosorbent assays for the detection of RVAs. In other studies, NoVs have been reported as the second leading cause of viral gastroenteritis in children from Japan [47] and Spain [48], and the occurrence of NoVs in our study also signifies the presence of NoVs in Nepal.
Additionally, a case-control study conducted at Kanti Children's Hospital in Maharajgunj, Nepal, also showed 8% of children less than five years old were positive for NoVs [21]. Detection of RVAs was comparatively higher in children suffering from diarrhea due to the extremely high number of viruses shed from infected individuals (possibly as high as 10 11 viruses/g of stool) [45]. The study also found high concentrations of RVAs in stool, ranging from 4.5 to 10.7 log copies/g. RVAs are highly contagious and persistent and they can survive in the air or water long enough to pose a threat to humans and animals [37]. Therefore, vaccines for RVAs should be part of Nepal's immunization programs; however, this is currently not the case. Introduction of a safe, effective, and affordable RVA vaccine in Nepal is essential to reduce RVA infections.
The concentrations of enteric viruses in children were comparable across all age groups, which suggests that children under 15 years of age are all equally vulnerable to contracting an enteric virus infection. Moreover, routine evaluation of gastroenteritis patients for the presence of enteric viruses is essential to avoid misdiagnosis and inappropriate treatment. In addition, since breastfeeding and oral rehydration therapy have been proven to reduce the incidence of diarrhea, the increased susceptibility of children to infection during early childhood may be due to poorly developed immunity related to eating and nutritional habits [49]. An increase in outdoor activities and a decrease in maternal care and feeding may also be linked to the increased risk of gastroenteritis in children 24 months old and older [50].
Fecal-oral route is a major source of transmission of gastroenteritis. Therefore, this study was conducted to find the presence of causative agents in drinking water, hands, and foods of children with gastroenteritis. Occurrence of E. coli and total coliforms from the tested hand swabs and food samples signifies that children had a poor hygienic status, and the above vehicles might be some of the potential sources of infection. This study also showed that chances of multiple infection with enteric viruses was significantly higher in the children who consumed E. coli-positive food (p < 0.05). To the best of the authors' knowledge, this is the first study conducted in both hospital and household to identify the potential source of gastroenteritis by taking water samples together with food and hand swabs from each individual gastroenteritis-affected child. Sixteen percent of the water samples tested were positive for at least one of the four viruses considered. This was lower than expected, which might be because of the presence of concentrations of enteric viruses in the water samples below the detection limit of the qPCR. Therefore, large volumes of water samples are frequently required for the detection of enteric viruses, as they are present in very low concentrations in water.
In this study, the majority of children consumed water from tap (47%), followed by jar water (35%). A lower number of children consumed water from tankers (5%), stone spouts (0.9%), and shallow tube wells (0.9%). Although two enteric viruses (RVAs and EVs) were detected in spring water, the rates were not sufficient to predict quality of the entire sources of spring water. The second highest positive percentage of AdVs was found in tanker water. The presence of AdVs in drinking and recreational water have been reported to be the potential sources of infections and viral outbreaks [17]. The presence of all four enteric viruses tested in tap water suggested that the water that is being supplied by the municipality was contaminated. Because of the poor quality of the water supplied by the municipality of the Kathmandu Valley to households, people are likely to prefer relying on private vendors of water from tankers and jar water. However, to date, regulations and oversight for the tanker water suppliers have not been established; therefore, viral contamination may be a risk for those who consume tanker water. The water supplied by jar was labelled as treated; however, detection of fecal indicator bacteria in jar water samples brings into question the quality control methods followed by the suppliers [51]. The presence of high concentrations of fecal indicator bacteria and other enteric pathogens in the municipal water supply indicates that water distributed by the municipality may likely pose a serious threat to human health. Therefore, the government should take a suitable action to reduce further contamination in pipe line water sources. Treated drinking water using chlorine, filtration, and boiling has still not produced a guaranteed supply of water that is safe for drinking. Thus, awareness programs and appropriate hygiene practices are essential to improve the treatment, storage, and transportation of water. Similarly, water supplied by pipe was not provided on a continuous basis; therefore, storage of water in the home is popular in developing countries like Nepal. This might be another possible way of contamination, due to microbial infiltration of poorly maintained systems [14,52].
The occurrence of co-infections (64%) was higher than that of mono-infections (34%) in the children suffering from diarrhea. Combined RVAs and EVs appeared as the most frequent mixed infection, followed by combined RVAs, AdVs, and EVs. The high prevalence of co-infections might be due to the fact that these pathogens share same mode of acquisition through the fecal-oral route, and ultimately susceptibility is increased following a primary infection. Poor sanitation, unsafe drinking water, and other common predisposing factors within developing countries enhances the severity of co-infections. Enteric infections involving a single pathogen are serious, and those involving multiple pathogens are even worse. They have a greater effect on weakening immunity as well as on compromising the nutritional status of the children by displacing water and creating a serious ionic imbalance and multiple other physical malfunctions.
Univariable logistic regression suggests that there was no significant association between the factors which were considered in this study and co-occurrence (p > 0.05) of infective organisms. In our study, the clinical picture of children with co-infection was similar to that of children with mono-infection for all clinical signs taken into examination, particularly dehydration. The presence of E. coli in the raw food consumed by children, however, increased the risk of co-infections in children (p = 0.04). This may be due to the practice of using river water, sewage, and groundwater for agricultural irrigation in the Kathmandu Valley. Several studies have suggested that these water sources are even not suitable for agricultural irrigation [53][54][55], and it is essential that the government take action to enhance wastewater treatment and monitoring.
In the study, there was an evidence of the presence of the same enteric viruses in both stool and water in 13% of the tested samples (n = 13/103). The result obtained was lower than expected, but the collected water was only used for drinking purposes; therefore, theoretically it should be free from any potential microbial contaminant. Six percent of household water samples that were collected from the houses of the children were found positive for RVAs. These data could provide evidence of transmission of enteric viruses from drinking water to the children. Enteric viruses are transmitted to natural bodies of water through either human fecal matter or animal sources. To get more information regarding the viral etiological agents and possible transmission factors in the environment, further studies should be conducted using larger sample size and considering a wider variety of risk factors to identify important barriers to safe water access and to warn the government regarding a safe drinking water policy.

Conclusions
Enteric viruses were detected in both stool samples and drinking water used by children with gastroenteritis. Positive percentages for at least one of the four enteric viruses were found to be comparable across all age groups. The rate of co-infection of two or more enteric viruses was found to be high in children with gastroenteritis, but their clinical presentation was not more severe than that of children with mono-infection with the same pathogens (p > 0.05). RVAs were found to be the most frequently occurring enteric viruses, suggesting they are a major cause of diarrhea in Nepal. The World Health Organization recommends that rotavirus vaccines should be included in all national immunization programs, but they have not yet been included in the national immunization schedule of Nepal. The occurrence of identical genera of enteric viruses in water and in stool samples indicates that drinking water is a contributing factor for gastroenteritis. An increased risk of co-infection associated with any other factor was statistically not significant, while E. coli-positive food samples showed significant association with the risk of co-infection. Therefore, more variables should be considered to find out the actual causes of gastroenteritis and coinfection.