Monitoring of group C rotavirus in children with acute gastroenteritis in Brazil: An emergent epidemiological issue after rotavirus vaccine?^†

INTRODUCTION

The Group C rotavirus (GpCRV) was first detected from piglets in 1980 [Saif et al., 1980] and confirmed later as a human pathogen by Bridger et al. [1986]. Since then, GpCRV have been detected from both sporadic episodes and outbreaks of gastroenteritis throughout the world [Cunliffe et al., 2001; Phan et al., 2004; Schnagl et al., 2004; Steyer et al., 2006; Abid et al., 2007]. The GpCRV infection in Brazil was first detected in the city of Rio de Janeiro by Pereira et al. [1983]. Thereafter, the GpCRV infections have been identified sporadically or as outbreaks at various times and locations [Gabbay et al., 1989, 1999; Timenetsky et al., 1993; Souza et al., 1998; Teixeira et al., 1998].

There are several unresolved questions about the ecology of the GpCRV. Its epidemiology has not been described fully due to the lack of samples, sensitive diagnostic assays [Castello et al., 2000], and the sporadic pattern of detection. The description of localized outbreaks may be interpreted as a spillover introduction of GpCRV strains from an unidentified source or reservoir [Bányai et al., 2006; Luchs et al., 2009]. A possible zoonotic role has been postulated based on the elevated GpCRV seroprevalence rates in human populations living in rural settings [Iturriza-Gómara et al., 2004]. In fact, Gabbay et al. [2008] documented the carriage of porcine GpCRV in children in Belém, Brazil, and established this virus as another emerging zoonotic infection in humans. The possibility of zoonotic spread of GpCRV has also been considered based on the results of a seroepidemiological survey conducted in the United Kingdom in which the prevalence of antibodies was higher in rural population than in urban population [Iturriza-Gómara et al., 2004].

The GpCRV was detected both in sporadic cases and outbreaks affecting adults and children from rural areas in the states of São Paulo (SP) and Goiás (GO) during the 2007–2008 national Brazilian rotavirus surveillance [Luchs et al., 2009]. This was the first report of GpCRV detection in Southeastern Brazil in over 15-year period of monitoring, however, no genetic information was described in the report. It seems important to improve the GpCRV laboratory surveillance system, especially after 2006 when the RIX4414 rotavirus vaccine was included in the Brazilian Immunization Program, preventing severe rotavirus gastroenteritis, and inducing significant reduction in the frequency of Group A rotavirus (GpARV) detection in children with gastroenteritis [Gurgel et al., 2008]. In the future, the GpCRV may assume an epidemiological importance.

The aim of this study was to intensify the monitoring of GpCRV detection in fecal samples from children ≤15 years with acute gastroenteritis during the 2007–2010 national rotavirus surveillance. In addition undertake the molecular characterization of the major capsid protein (VP6) of the GpCRV strains detected from children during this period, in order to obtain further information on the genetic relationships between human and animal.

MATERIALS AND METHODS

Samples

The aim of the Acute Diarrhea Disease Monitoring Program (ADDM), with a national range is the early detection of diarrhea outbreaks. Stool samples from patients with acute gastroenteritis are sent to Enteric Diseases Laboratory of the Adolfo Lutz Institute, reference center for rotavirus surveillance and a member of ADDM, in order to conduct the viral investigation. The clinical samples tested in this study were obtained from the following states: São Paulo (SP), Mato Grosso do Sul (MS), Paraná (PR), Tocantins (TO), Goiás (GO), Santa Catarina (SC), and the Brazilian Federal District (FD) (Fig. 1). The samples studied were part of ADDM, acquired from a convenient retrospective sampling, without inclusion or exclusion criteria, with no characterization of the participating health centers, and without patients' clinical evaluation. Therefore, the study may not be representative of the actual epidemiological scenario. This retrospective study was conducted from September 2007 to September 2010 with convenient surveillance specimens collected from children under the age of 15 years presenting with acute gastroenteritis.

RESULTS

RT-PCR Amplification and Phylogenetic Analyses of the VP6 Genes

Of the eight GpCRV positive samples, six specimens were amplified successfully by the RT-PCR assay, and their sequences were determined. All the six GpCRV strains (SP22/Hu/BR/2007, SP09/Hu/BR/2008, SP12/Hu/BR/2008, SP28/Hu/BR/2008, SP35/Hu/BR/2008, SP36/Hu/BR/2008) belong to SP and are listed in Table I.

All the GpCRV SP strains were very similar to the GpCRV samples on BLAST search based on the VP6 protein. The phylogenetic analysis indicated that the GpCRV SP strains belonged to the human lineage. None of the GpCRV SP sequences were related closely to animal GpCRV strains (Fig. 2). The GpCRV SP strains were related to each other, since they shared 97–100% amino acid (aa) identities in the VP6 gene.

The GpCRV SP strains analyzed in this study showed a genetic relationship with GpCRV strains from Japan, especially from the city of Yokohama (AB533513) isolated in 2009. The other humans GpCRV Brazilian strains available in GenBank clustered with strains from Nigeria (AF325806), Thailand (EF641110), United Kingdom (M94156, X59843), Spain (AM118018, AM118020), India (AY786570, AY786571, AY795898), and Slovenia (DQ439863). In addition, 12 porcine Korean GpCRV strains were placed on separated branches from the human GpCRV strains, Cowden strain and porcine-like human Brazilian children strains (Fig. 2).

DISCUSSION

This study was designed in order to investigate the frequency of GpCRV infection in children ≤15 years old with acute gastroenteritis and addressed the genetic relationship of the GpCRV detected with others human, porcine, and bovine strains. The frequency of GpCRV infections detected in this work (0.3%) was similar to that observed in other studies carried out with pediatric population in South Korea (0.7%) [Moon et al., 2011] and Turkey (0.8%) [Mitui et al., 2009]. However, the prevalence of the infection was lower compared to Nigeria (1.8%) [Adah et al., 2002], Argentina (2.8%) [Castello et al., 2002], Spain (16%) [Sánchez-Fauquier et al., 2003], Malawi (3.3%) [Cunliffe et al., 2001], and Japan (1.2%) [Kuzuya et al., 2007]. This study does not suggest that the GpCRV have currently a major epidemiological impact, even after the introduction of a GpARV vaccine. Indeed, the decrease in the detection has fallen to zero in 2009 and 2010.

Considering the GpARV detection, the annual frequency was similar to that reported in studies carried out in Brazil [Linhares, 2000; Carmona et al., 2006]; however, a low percentage of GpARV infection was observed in 2009. The Brazilian States of Midwest, Southeast, and South regions exhibit a temperate-like climate, with positive GpARV specimens appearing to peak during the winter or dry season (from May to September) [Pereira et al., 1993]. According to the National Meteorology Institute (INMET) records, during the year of 2009, Brazil experienced an atypical raining winter season with higher average rainfall record due to the El Niño phenomenon (http://www.inmet.gov.br) therefore, this climatic condition could be involved in the low detection frequency observed.

The median age of GpARV and GpCRV detections in the current study were 2.4 and 4.5 years old, respectively, agreeing with seroepidemiological investigations and detection data that suggest a difference between age distribution in children with diarrhea caused by GpARV, which typically infects children before 3 years of age, and GpCRV, which affects older children [Nilsson et al., 2000; Kuzuya et al., 2001; Iturriza-Gómara et al., 2004; Castello et al., 2009].

Seven of eight GpCRV samples were detected in SP, suggesting that the GpCRV occurs more frequently in this state. However, this may also be an outcome from the efficient health surveillance program conducted in the state, and the higher number of samples sent to laboratory for viral diagnosis. Serological investigations also suggest a higher GpCRV incidence in rural than in urban populations [Iturriza-Gómara et al., 2004; Mukhopadhya et al., 2010]. A similar pattern was observed in this study, once the eight samples were collected from people who live in countryside areas. However, there is no serological information about GpCRV from diverse geographical areas in Brazil, and the common source of these infections was unable to be identified during the study.

The high aa identity level of the GpCRV SP strains in the VP6 protein, suggest that these strains were originated from a single or a group of related strains. Also, the partial VP6 gene analyzed showed a relationship between the GpCRV strains from SP and Japan, especially from the city of Yokohama, Kanagawa prefecture. Currently, Brazil is the country with the largest number of Japanese outside Japan, and the Brazilian community in Japan for its part, is the third largest community of foreign workers residing in Japan [Beltrão and Sugahara, 2006]. The fact that this migration flow may lead to a possible exchange of GpCRV strains can be speculated, however, additional analysis of full-length of the VP6 gene, and/or VP7 and VP4 genes are required to fully validate this finding.

The molecular analysis showed that human strains were related distantly to the bovine and porcine strains as observed by Mitui et al. [2009]. The human GpCRV strains appear to constitute a single linage supported by high bootstrap values, irrespective of the geographic origin. The highly conservation among the human GpCRV genetic sequences could suggest a recent evolution.

In this study, all samples from SP cluster with other human samples and did not show any evidence of animal ancestry. However, the human porcine-like samples from Belém clustered with the prototype porcine Cowden in agreement with the data showed by Gabbay et al. [2008]. There are evidences that the transmission of the GpARV can occur from animal-to-human as well as from animal-to-animal by direct transmission of the virus or the contribution of one or several genes to reassortants [Martella et al., 2006; Ghosh et al., 2007]. Compared to the GpARV, there is a paucity of information regarding the sequence and phylogenetic data on all 11 genomic segments of the GpCRV. Therefore, it is largely unknown whether these genomic segments are totally similar to those of the original species or reassortants from other species. However, the GpCRV are thought to be able to cause interspecies transmission [Jeong et al., 2009].

In fact, the oldest GpCRV samples were collected from pigs, which strongly suggest that the origin of this pathogen could be porcine. The acquisition of genes with human origin could facilitate the adaptation of such strain in order to infect the human population. The fact that the GpCRV strains do not remain circulating in humans (like GpARV), may suggest that this group does not achieve the fitness required to become a successful human pathogen. An alternative hypothesis is that the GpCRV may cause a subclinical infection in humans. Based on Meleg et al. [2008] study conducted in Hungary, the GpCRV were detected from raw sewage samples, suggesting that the virus is in circulation; however, a significant increase in the number of sporadic cases or outbreaks was not observed.

In conclusion, this study adds further evidence that the GpCRV is a minor cause of acute childhood gastroenteritis in Brazil during the surveillance period. In addition, small-scale of GpCRV outbreaks in restricted communities might occur and go undetected, particularly if the severity of the disease is minor and hospitalization is not required [Bányai et al., 2006]. It is noted that the GpCRV incidence and associated disease remain unclear once sensitive tests for its detection are not available to clinical laboratories. The diagnosis is difficult since most of the ELISA assays do not recognize the GpCRV specific VP6 antigen. The RT-PCR using GpCRV specific primer is a convenient option [Castello et al., 2000]; however, it has not been used widely due to the large costs for routine surveillance. The PAGE analysis of dsRNA requires the presence of 10⁸–10¹⁰ viral particles/ml for a positive result [Castello et al., 2000] nevertheless; this method has sufficient sensitivity to detect GpCRV with low costs.