Detection of rotavirus and norovirus in children under 5 years of age in Pretoria, South Africa and molecular characterization of rotavirus strains

Abstract

Background: The most important causes of viral gastroenteritis include rotaviruses, caliciviruses, adenoviruses and astroviruses. Globaly, rotavirus infection has been estimated to result in 527 000 deaths of young children annually. In South Africa rotaviruses were found to be responsible for 20 – 30% of diarrhoeal diseases in children younger than 5 years of age. Noroviruses are the leading cause of non-bacterial acute gastroenteritis outbreaks worldwide, with human noroviruses accounting for more than 86.0% of all outbreaks caused by viruses. Aim: To investigate the prevalence of rotaviruses and noroviruses and to determine the genotypes of rotaviruses circulating in children under 5 years of age attending the private health facilities in the Tshwane region of South Africa. Material and methods: A total of 1227 stool specimens and the corresponding patient‟s data were obtained from the Lancet Pathology Laboratory (Pretoria, South Africa) from children under the age of 5 years with gastroenteritis who submitted stools for testing. The MRC-DPRU detected rotaviruses using the commercially available enzyme immunoassay DAKO IDEIATM assay whereas the Lancet Pathology Laboratory detected rotaviruses using VIKIA assay. The rotavirus genome was extracted from EIA positive samples using phenol-chloroform and analyzed utilizing polyacrylamide gel electrophoresis (PAGE). Molecular characterization of the rotavirus-positive samples was done using reverse transcriptase polymerase chain reaction (RT-PCR) and genotyping assays. Samples were sequenced for further confirmation and to type the nontypeable samples. Due to lack of sufficient raw stool materials, only 673 of the specimens were screened for noroviruses using the RIDASCREEN® Norovirus testing assay. No molecular characterization of noroviruses isolates was performed. In addition, the Lancet Pathology Laboratory detected other diarrhoeal pathogens such as bacteria, parasites fungi and viruses, and the results were used to determine the relative significance of rotaviruses and noroviruses as the causes of diarrhoeal disease in comparison to other potential enteric pathogens in the Tshwane region. Results: The Lancet Laboratory reported 216 (17.6%) rotavirus positives with VIKIA assay and the MRC-DPRU reported 534 (43.5%) rotavirus positives with DAKO assay, clearly indicating that the DAKO assay has superior sensitivity. The use of a diagnostic assay which is not sensitive such as VIKIA assay could lead to misdiagnosis, under-reporting and x mismanagement of patients. Rotavirus infection was higher in children below 12 months of age (58.2% vs 41.7% for all other age groups). There was no significant difference between black and white children infected with rotaviruses [(249 or 46.6% for black versus 279 or 52.3% for white children)]. However, only 5 (0.9%) of rotavirus positives were Indians most probably due to small sample size. Rotaviruses were mostly predominant during the months of May (66.0%), June (71.0%) and July (57.4%). The prevalence of norovirus was relatively small compared to rotavirus for the same period. Only 53 (7.9%) samples were norovirus positive by ELISA. Noroviruses were detected in 11.1% of Indian children, followed by 7.9% of black children and 7.8% in white children. There was no significant difference amongst black and white children infected with noroviruses. A total of 74.2% of the EIA rotavirus positives were confirmed by PAGE. Nine PAGE patterns were obtained, displaying six long and three short electrophoretypes. A total of 319 stool specimens were genotyped for both VP4 and VP7. Of these, 289 (90.6%) were typeable for the VP4 [P] gene and 298 (93.4%) were typeable for the VP7 [G] gene. The most predominant single P genotypes were P[8] at 179 (56.1%), P[9] at 31 (9.7%) and P[4] at 25 (7.8%). Among the genotyped samples, the most predominant single G genotypes were G3 at 54 (16.9%), G8 at 48 (15.1%), G9 at 31 (9.7%) and G2 at 27 (8.5%). Mixed P and G rotavirus genotypes were detected in 12.5% and 29.5% for VP4 and VP7 respectively. Dual, triple and quadra rotavirus infections were detected for both VP4 and VP7 genotyping. The most predominant G and P genotypes found in this study were G8P[8] (8.9%), G3P[8] (8.3%) and G9P[8] (6.0%). Phylogenetic analysis indicated that of the 8 samples sequenced for VP4, all were P8 genotypes. Of these, five (DPRU1570, DPRU1754, DPRU7596, DPRU7310, and DPRU7335) clustered with Hun9-like lineage and 3 (DPRU7581, DPRU2030 and DPRU7173) clustered with Wa-like lineage on the P[8] phylogenetic tree. The only VP7 sample that was successfully sequenced (DPRU7381) clustered with lineage II of G3 on the phylogenetic tree. Children infected with rotavirus (83.0%) were also found to be co-infected with other diarrhoeal pathogens and these included Candida spp (33.2%), E. coli (19.8%), Cryptosporidium spp (8.9%), Adenoviruses types 40 and 41 (8.2%) and Giardia lamblia (7.7%). Viruses were the most predominant enteric pathogens detected in the samples at 63.0%, followed by fungi at 19.0%, bacteria at 15.0% and parasites at 3.0%. xi Conclusion: Compared to other enteric pathogens, rotaviruses are the leading cause of diarrhoea in children attending private health care facilities in Tshwane region. It was interesting to note that the most predominant G and P genotypes were G8P[8] (8.9%), G3P[8] (8.3%) and G9P[8] (6.0%). The predominance of these strains clearly indicate that the rotavirus vaccination programme that started in August 2009 in South Africa will be effective in protecting children against major circulating strains. However, more studies should be done to monitor the epidemiological trends of all enteric pathogens in South Africa, especially that rotavirus is now a vaccine-preventable disease.