Background
Acute gastroenteritis (AGE) in children is one of the most significant diseases, causing morbidity and mortality worldwide [
1,
2]. Although the improvement in sanitation and prevention strategies has determined a substantial reduction in the mortality rate for diarrhea from 15% in 2008, to about 9% in 2015, equivalent to 500,000 deaths among children less than 5 years old, infectious diarrheas are still an important public health concern, both in resource-poor settings and industrialized countries [
2,
3].
Viruses are recognized as a major cause of severe AGE, particularly in children. Rotaviruses (RV) are the main cause of mortality due to diarrhea in those under 5 years old, preventable with the vaccination [
4,
5]. Yet, despite a significant reduction after RV vaccine introduction in 2006, hospitalizations for infantile AGE of viral etiology continued to be reported [
6‐
9]. The increasing use of more powerful diagnostic systems in the last few years, as the conventional polymerase chain reaction (PCR) or high-throughput technologies, for the amplification and identification of virus genomes in stool samples, has resized the study of the agents involved in childhood AGE [
7,
10,
11], and changed significantly the pathogen spectrum of community-acquired gastroenteritis [
6,
7,
12,
13].
Other viruses considered of clinical importance and frequently associated with diarrhea mortality are the human caliciviruses (HuCVs), members of the family
Caliciviridae, which have acquired importance, especially after RV vaccine implementation [
6,
10,
14]. Human adenoviruses (HAdVs), members of the family
Adenoviridae, are often reported as the second or third cause of infantile diarrhea, both sporadic or outbreak associated, and cause a wide range of disease symptoms [
14,
15]. Finally, human astroviruses (HAstVs), of the family
Astroviridae, which affect predominantly children under 2 years of age, have been involved in 0.5–15% of diarrheal outbreaks associated to severe pediatric cases [
16‐
18].
On the other hand, the clinical importance of viruses belonging to the wide family
Picornaviridae, such as enterovirus (EV), Aichi virus (AiV) and klassevirus/salivirus (KV), is up-to-date unclear, with those viruses presumably playing a rather minor epidemiological role in diarrhea [
14,
19‐
22]. Some subgroups of EVs have been involved as causative of at least 3.4% of AGE of unknown etiology [
23]. Similarly, AiV, of the genus
Kobuvirus, was initially described as cause of oyster-associated non-bacterial gastroenteritis in human [
22], and later associated with AGE, reaching detection rates between 0.5 and 0.9% in Europe, and up to 4% in Asia and Africa [
20,
24]. AiVs were recovered during a study from a major river polluted with sewage discharges in Caracas (Venezuela) during 2007–2008 [
25], but its impact on the burden of AGE in Venezuela is unknown. Finally, KVs, discovered in human stool and sewage [
22], have been significantly associated with pediatric diarrhea in different countries, especially in children less than 3 years old, with a frequency ranging from 0.1 to 8.7% [
26‐
28].
Because the information about viruses different from RV associated with diarrhea in Venezuela is limited, the present study was aimed to determine the incidence of infections caused by other conventional gastroenteritis viruses before RV vaccine implementation, and to investigate the contribution of AiV and KV to diarrheal diseases, until now unknown in Venezuela. For this purpose, children less than 5 years old with diarrhea attended at a large public hospital in Valencia City, over a 5-year period (2001–2005), were studied using molecular detection assays.
Methods
Study design
The study included stool samples collected from children with AGE under 5 years old, attended during the years 2001–2005 in the city of Valencia, Carabobo State (Venezuela), as part of a RV diarrhea surveillance program conducted at the
Hospital de Niños ‘‘Dr. Jorge Lizarraga’’ of the
Ciudad Hospitalaria ‘‘Dr. Enrique Tejera’’ (CHET) described previously [
29].
AGE was defined as three or more liquid stools over a 24-h period and for not over 14 days. To determine the epidemiologic and clinical characteristics of the AGE, information from the clinical history and the physical examination were collected: age, gender, nutritional status and type of treatment (outpatient or inpatient hospital based) were recorded for each case and used as measurement instrument for the severity of the community-acquired AGE, together with the estimation of dehydration, assessed according to WHO criteria [
30]. Inpatient treatment was defined as the admission to either the emergency room for a short stay to receive oral rehydration therapy (< 24 h) or to the regular pediatric wards of the hospital for longer time [
31]. The socioeconomic status was determined by a modified Graffar methodology [
32].
Sample collection
From a total of 13,026 fecal diarrhea specimens obtained from the enrolled children within 48 h following admission, 227 were randomly selected from RV negative-tested samples. All the samples had been systematically examined for the presence of RV antigen, bacteria and parasites as previously described [
29,
33] and resulted negative for all of them. All samples were stored at − 80 °C until processed.
Fecal suspensions (10% w/v in phosphate buffer saline) were prepared from each stool sample, vortexed, and clarified by centrifugation at 10,000g for 10 min. Viral RNA/DNA was extracted simultaneously from 200 µl of supernatant, using the QIAamp MinElute Virus Spin Kit (QIAGEN, Hilden, Germany), based in a spin-column procedure, and following the manufacturer’s instructions. Briefly, samples were lysed in the presence of QIAGEN Protease and Buffer AL containing RNA carrier provided by the kit. Ethanol absolute (Merck, KGaA, Darmstadt, Germany) was added to the sample that was then transferred onto a QIAamp MinElute column, where the viral nucleic acids were adsorbed onto the silica-gel membrane. Wash buffers were used to remove impurities by centrifugation, and finally, the viral nucleic acids were eluted in 50 µl of Buffer AVE (provided), for use in amplification reactions or storage at − 70 °C.
Reverse transcription (RT)
Screening for the presence of RNA viruses, such as HuCV, HAstV, AiV, KV, EV and human picobirnavirus (HPBV), was conducted firstly by RT reaction, as follows: the extracted RNA was denatured and then reverse transcribed with random primers (0.02 µg) using M-MLV reverse transcriptase (200 U) and deoxynucleoside triphosphate mix (0.2 mM), RNasin (40 U) (Invitrogen, Carlsbad, California, USA) in reverse transcription buffer to a final volume of 50 µl. The mixture was incubated at 37 °C for 1 h followed by incubation at 70 °C for 15 min, to obtain cDNA.
Polymerase chain reaction (PCR)
Single PCR reactions were performed from 5 µl of extracted DNA (for HAdV detection), or cDNA (for RNA viruses), using a selected combination of oligonucleotide primers specific for each virus previously described [
22,
34‐
40] at a final concentration of 0.2 µM each one. Two additional degenerated primers were designed for this study by multiple alignments, leading to broad target specificity for HuCV (290YM) and HAstV (MON394d) (Table
1). Cycling conditions used were adapted as shown in Table
1. All PCR reactions were done in a final volume of 50 µl and the PCR products were analyzed by agarose gel electrophoresis and ethidium bromide staining.
Table 1
Oligonucleotide primers and amplification conditions used in this study for the molecular detection of gastroenteritis viruses
Calicivirus | RNA-dependent RNA polymerase | 1st | − | 289H | TGACGATTTCATCATCACCATA | | 4865–4886a | | 34 |
| | − | 289I | TGACGATTTCATCATCCCCGTA | A | 4865–4886a | 319 | |
| | | + | 290YM | GATTACTCCAGGTGGGAYTCMAC | | 4568–4590a | | In this study |
Adenovirus | Hexon | 1st | + | hexAA1885 | GCCGCAGTGGTCTTACATGCACATC | B | 18,858–18,883b | 301 | 35 |
| | | − | hexAA1913 | CAGCACGCCGCGGATGTCAAAGT | 19,136–19,158 | | |
Astrovirus | ORF-1 | 1st | + | MON340 | CGTCATTATTTGTTGTCATACT | C | 1182–1203 | 289 | 36 |
| | | − | MON348 | ACATGTGCTGCTGTTACTATG | 1450–1470 | | |
| | 2nd | + | MON394d | GARATCCGTGATGCTAATGG | D | 1250–1269 | 220 | In this study |
| | | − | MON348 | ACATGTGCTGCTGTTACTATG | 1450–1470 | | 37 |
Aichi virus | 3C-3D | 1st | + | 6261 | ACACTCCCACCTCCCGCCAGTA | E | 6261–6282c | 519 | 38 |
| | | − | 6779 | GGAAGAGCTGGGTGTCAAGA | | 6760–6779 | | |
Klassevirus | 2C | 1st | + | LG0098 | CGTCAGGGTGTTCGTGATTA | F | 4463–4482 | 345 | 27 |
| | | − | LG0093 | AGAGAGAGCTGTGGAGTAATTAGTA | | 4783–4807 | | |
Enterovirus | 5′NTR2 | 1st | + | EV1 | CGGCCCCTGAATGCGGC | G | 454–470 | 194 | 39 |
| | | − | EV2 | CACCGGATGGCCAATCCA | | 630–647 | | |
Picobirnavirus | | | + | PicoB25 | TGGTGTGGATGTTTC | | 665–679d | 201 | |
RNA-dependent RNA polymerase | Multiplex PCR | − | PicoB43 | ARTGYT GGTCGAACTT | H | 850–865d | | 40 |
| + | PicoB23 | CGGTATGGATGTTTC | 685–699e | 369 |
| | | − | PicoB24 | AAGCGAGCCCATGTA | | 1039–1053e | | |
Statistical analysis
Data were analysed for the comparisons of variables using 2 × 2 tables with χ2 test, or Fisher’s exact test (two-tailed, 95% confidence intervals) (Epi Info™ 7.1.4.0, CDC Atlanta, GA, USA). Student’s test was applied for comparisons of variable values. Tests were considered significant when p < 0.05.
Discussion
The present study shows the epidemiology of viruses that caused pediatric AGE in Valencia (Venezuela) between 2001 and 2005 before the RV vaccine implementation. Although the population studied does not represent the entire epidemiological data of the viral diarrheal disease of this country, the results should provide a good estimation of the real impact of the viral AGE during the years 2001–2005 by causes other than RV.
The high prevalence of enteric virus found in this study is similar to that reported previously by others authors [
12,
41,
42], and showed that EV, HuCV, HAdV, HAstV, AiV and KV accounted for a significant proportion of RV-negative AGE in this locality. The rate was lower than that shown by a Japanese study where multiplex assays including a larger number of target pathogens were applied [
8], but it was higher than that described in European, Asian and African studies [
17,
43‐
45], as well as that reported in a previous study performed during the year 2003 in Valencia City [
46]. Of note, a fraction (41%) of the diarrhea cases here studied remained without a precise etiology, probably due to a low viral load, the presence of inhibitors in the samples or viruses not included in the assays. However, the relatively higher detection rate of viral agents reflects an increase of the diagnostic capabilities of the PCR-based assays used, although it could also depend on the population group studied, which included mostly children under 24 months of age, belonging to the lowest socioeconomic stratum (Graffar 5), living in the most precarious sanitary and dietary conditions, where the fecal–oral transmission of a wide range of pathogens was favored.
The significant higher frequency of viral infections, as well coinfections, shown here in children less than 24 months of age contrasts with previous data from Valencia where no age differences were observed in viral enteric infections [
46]. It is instead in agreement with data obtained by others authors elsewhere [
44,
47], and shows the highest susceptibility of the children to the viral infection during the early childhood, perhaps due to unsatisfactory protective immunity.
Previous data have reported that viral infections other than RV are clinically milder than the RV infection [
44,
46,
48]. In this study, only RV-negative stool samples were included; therefore a comparison of the clinical conditions with children infected with RV could not be done. However, the data suggest that the infections by viruses such as EV, HuCV, HAdV, HAstV, AiV and KV would be mainly associated with less severe diarrheic episodes, not necessarily demanding medical intervention or hospitalizations.
This study demonstrates the contribution of EV and HuCV as important etiologic agents of viral AGE in the setting studied, both viruses found together in mixed infections in almost a quarter of the cases studied. The detection rate obtained for EV as single infecting agent was similar to that reported in a study carried out in Maracaibo (Venezuela) during 2008–2009 [
49], and it was higher than that described in Thailand [
23]. On the other hand, this rate was similar to the RV rate detection (24.5%) reported in another study carried out in Valencia City, during the same period [
33]. Some serotypes of echovirus and coxsackievirus B have been described to be cause of diarrhea [
50,
51]. It is noteworthy that the presence of Sabin vaccine-related strains in stool samples of diarrheic children could have caused an overestimation in the EV detection rate with the PCR assay used. In addition, EVs could be occasionally shed with the feces of patients suffering a broad spectrum of other non-enteric diseases, sometimes in prolonged way [
20,
23,
50]. This would explain in part the relatively higher rate of EV found in this study in infected children older than 24 months than that of other viruses. Thus, case–control studies and further genotyping of the strains detected will be desirable, to better define the burden of EV as a cause of diarrheal disease.
The overall prevalence of HuCV observed in this study, the second most common causative agent of viral AGE, was comparable to that described by others among RV-negative samples from children with diarrhea in four distinct Thai regions under sentinel surveillance between 2006 and 2008 [
52], and higher than that reported previously in Valencia City during the 2003 [
46]. This prevalence indicates a greater ability of the primers used in the PCR assay to detect a broad diversity of strains. It ratifies also the need of monitoring the contribution of the HuCVs to the burden of the AGE after implementation of RV vaccination.
A significant observation in this study was also the relatively higher detection rate of HAdV infection, as compared to a previous study based on serologic assays from Valencia [
46], and to reports from other continents [
53‐
55] that suggest the existence of a geographic variability of the virus prevalence, as well as the important contribution of the HAdVs to the mixed infections. A similar rate of HAdV detection was reported from Korea during the years 2012–2013 [
6], but it is noteworthy that the relative high prevalence for HAdV observed in this study could have also been determined by the presence in the stools of non-enteric types that could occasionally be excreted from respiratory source, and detected by the assay used, directed to amplify a conserved portion of the hexon-coding gene, common for all the HAdV. Thus, the molecular characterization is a crucial step to define the species of HAdV mainly involved in diarrhea and to understand the true contribution to the AGE. No information about the types of HAdV that have been circulating in Venezuela is available, but preliminary results indicate that most of the HAdV strains found in Valencia City during the same period were enteric types (Blanco R., personal communication).
HAstV were involved in a modest number of episodes, mainly in mixed infections with HuCV, HAdV and EV, which evolved as a mild form of AGE, similar to that reported by other studies [
17,
41]. The HAstV detection rate found here was comparable with the data from a previous local study [
46], and those from Lebanon, France and Germany [
41,
44,
56].
Although AiV and KV have been associated with AGE in several continents [
17,
24,
27,
28,
57‐
59], to our knowledge, there have not been reports of AiV and KV causing infections in Venezuelan human population. Unfortunately, the low rate of detection in this study did not allow to evaluate their relationship with socio-demographic and clinic variables, but their presence confirms the participation as agent of childhood diarrhea and the relatively recent introduction in Venezuela.
In this study were used primers directed to the most commonly described HPBVs of genogroup I and II [
40], but no virus was found. Possibly, their high genomic diversity could have limited the detection with the RT-PCR assay available. Thus, additional efforts are required to optimize assays able to identify these and other uncommon viruses associated with AGE, as well the use of new technologies as virus microarray, sequence-independent amplification and sequencing of viral nucleic acids [
7,
11,
22], to clarify their epidemiology and possible pathogenicity.
Conclusions
This study demonstrated a high prevalence of enteropathogenic viruses other than RV in Venezuelan children suffering acute diarrhea, confirming the contribution of conventional enteric viruses in the pediatric AGE in this country. In addition, the presence of emergent viruses more recently described, such as AiV and KV is also described. Because the study included only diarrheic pediatric patients who received medical attention in Valencia City, the prevalence of virus infection reported here could represent an underestimation of the true rates of gastroenteritis associated viruses circulation in the population. Future studies should consider asymptomatic and self-limiting diarrhea cases. However, these results, obtained from five consecutive years, expand the knowledge about the spectrum of viral agents involved in acute community-acquired disease, and provide a baseline data for the molecular epidemiology study of these pathogens, which will be helpful for comparison with regional data obtained in post-RV vaccination era. Finally, they ratify the need for a long-term surveillance for such enteropathogenic viruses, following the implementation of RV vaccination, to better understand the participation of these agents in children AGE.
Authors’ contributions
EV and ACA conceived the study, participated in the design/adaptation of experimental protocols, analysis results and wrote the initial draft of the manuscript. RG, ACA and RB were engaged in sample processing and data collection. EV supervised the laboratory procedures and PCR quality control, performed the statistical analysis, data interpretation, graphic representations of the results, and drafted the final manuscript. ACA, KP and RB performed the molecular analysis. ACA, RB, RG, JEL and FL helped reviewing the manuscript. All authors read and approved the final manuscript.