Burden of rotavirus disease and molecular characterization rotaviruses at Dr George Mukhari Hospital from 2003-2005

Abstract

Background: Rotavirus infection remains a significant clinical problem throughout the world, infecting almost every child younger than 5 years of age, despite socio-economic status or environmental conditions. Rotavirus is the most common cause of severe dehydrating gastroenteritis in infants and young children. Implementation of an effective vaccine programme could reduce the incidence and severity of rotavirus disease. Decisions about new candidate rotavirus vaccines require reliable data on disease impact in both developed and developing countries. The aim of this study was to assess the burden of rotavirus associated disease at the tertiary care Dr George Mukhari Hospital, Ga-Rankuwa and the secondary care hospital Brits Hospital, Madibeng and to describe the genetic diversity of rotavirus strains circulating in Ga-Rankuwa and Brits communities over a similar time period as the testing of Rotarix@ vaccine. The broad objectives included; to perform a hospital-based burden of rotavirus disease in two different hospitals in the North West of Pretoria area, to conduct molecular characterization of rotaviruses circulating in the Pretoria region and lastly to devise an alternative molecular typing method to detect rotavirus VP6 subgroups.

Materials and Method: To investigate the hospital-based burden of rotavirus disease, diarrhoeal stool samples were collected at Dr George Mukhari and Brits Hospitals from children less than 5 years of age. Group A rotavirus antigen was detected from the samples using commercially available rotavirus enzyme immunoassay IDEIA TM Rotavirus test (DAKO, Dakocytomation, Denmark). Genetic analyses of rotavirus strains were determined by polyacrylamide gel electrophoresis (PAGE) to characterize the electrophoretic patterns followed by analysis of the P and G genotypes by RT -PCR and multiplex PCR amplification of specific sequences of VP7 and VP4 genes. To devise an alternative molecular typing method to detect rotavirus VP6 subgroups, with subgroup specificities determined by both VP6 monoclonal antibodies and restriction fragment length polymorphism using restriction endonuclease Acil, Odel and Rsa I. Selected PCR amplicons (VP7 and VP6 genes) were purified, cloned and sequenced. Consensus sequences of the VP7 and VP6 genes were aligned and analysed manually with Chromaslite and BioEdit software packages. Multiple sequence alignment was implemented by Mafft software packages. The nucleotide and deduced amino acid sequences of the VP7 and VP6 genes were compared with reference strains available from GenBank. Multiple methods were used to construct phylogenetic trees and included neighbor¬ joining, maximum parsimony analysis and maximum likelihood distance. Bootstrap values were computed using 1000 replicates with Phylip and the MEGA softwares. The graphic representation of each phylogenetic tree was displayed with the Treeview program.

Results: Between 2003 and 2005, a total of 2 514 diarrhoeal stool samples were collected. Of these, 527 (21%) were positive for group A rotavirus and the majority of children hospitalized were less than 2 years of age. The annual peak prevalences of group A rotavirus were 56%, 59% and 56% for 2003, 2004 and 2005, respectively and were observed during the autumn and winter months. The estimated incidence of gastroenteritis associated with rotavirus indicates that one in every 50 to 70 children in the area is likely to be hospitalized with rotavirus diarrhoea between birth and 2 years of age. During the three-year study period, ten, six and seven different RNA electrophoretic patterns were identified in 2003, 2004 and 2005, respectively. The VP6 genes of the representative strains (G1, G2, G3, G9, G8 and G12) were ana lysed with restriction endonuclease Acil, Ode! and Rsa!. The restriction endonucleases produced 11 unique restriction profiles (A-K). The VP6 RFLP results correlated well with strains displaying long RNA electropherotypes and VP6 subgroup 1/ specificity and also with strains displaying short RNA electropherotypes and exhibiting VP6 subgroup I specificity as determined with VP6 monoclonal antibodies.

The genotypic distribution varied remarkably and major rotavirus strains detected in circulation during the study period included G2P[4] in 2003, G1 P[8] in 2004 and G3P[8]/ G3P[6] in 2005. Rotavirus strains carrying G8P[8] specificities and unusual G 12P[6] strains were also detected at low frequency The consensus VP7 nucleotide sequences, exhibited the greatest homology and identity (97-99%), when compared against corresponding international reference strains. The nucleotide sequence datasets were closely related to strains from South Africa, Vietnam, Bangladesh, East India, Republic of Congo, China, Russia, Thailand and Japan. The phylogenetic tree revealed the South African strains (G1-G3, G8-G9 and G12) clustered with international strains whereas the G1 strains clustered into two different lineages. Phylogenetic analysis of the VP6 gene revealed four lineages with international reference strains. The VP6 gene showed 97-99% identity at the deduced amino acids level with strains from Taiwan, Bangladesh, the United States and Brazil.

Conclusion: This is the first study to estimate the disease burden associated with rotavirus diarrhoea in South Africa. The overall results confirm that rotavirus is the most common cause of severe diarrhoea. The epidemiology of rotavirus diarrhoea in South Africa correlates well with what has been reported in other countries. The proportion of hospitalization of rotavirus infection in children less than 5 years was estimated to an annual prevalence of 22.8% (95%CI 21.2%, 24.5%) at Dr George Mukhari Hospital, while at Brits Hospital was estimated at 18.2% (95%CI 14.9%, 22.1 %). Rotavirus genotypes circulating at Dr George Mukhari Hospital showed a high degree of diversity and the emergence of uncommon rotavirus strains such as G12. The emergence of novel rotaviruses in the region needs to be taken into account where vaccine efficacy is concerned. It is, thus, important to continue with such studies to monitor the rotavirus strains associated with severe

gastroenteritis in a hospital setting before and after the introduction of a rotavirus vaccine. Results also indicated that RFLP analysis of VP6 amplicons might be a simple and reliable, alternative to MAb subgrouping. The sequence analysis of the partial length VP6 gene confirmed the location and the recognition sites of the restriction enzymes The RFLP analysis proved to have more potential to accurately detect different rotavirus subgroups.