|Home | About | Journals | Submit | Contact Us | Français|
From October 2003 through September 2004, a total of 289 stool samples were collected from children less than 5 years of age with severe diarrhoea on admission or visiting the emergency room of the Navrongo War memorial Hospital in rural Ghana during a rotavirus burden of disease study. Rotavirus antigen was detected in 115 (39.8%) of stools tested for rotaviruses. Four rotavirus-positive samples were found to bear G10P specificity by RT-PCR, PCR-ELISA and oligonucleotide microarray hybridization. Two of these viruses, further exhibited serotype G10 specificity by neutralization, subgroup II specificity by EIA and possessed long electropheretic patterns by PAGE. Their VP7 genes shared a much closer nucleotide identity with other African human G10 strains (>97%) than with human G10 strain from Asia or South America (<86%)) or animal strains (<85%). The VP8* genes of the Ghanaian G10 strains exhibited more than 94% identity to that of human P viruses and belonged to the P lineage 1a. The deduced VP7 amino acid sequence showed the Ghanaian strains to be more closely related to human G10 strains than animal G10 strains. The possession of the typical human subgroup II specificity and the P specificity (commonly found in Ghana and the rest of Africa) and the marked similarity in the VP7 antigenic sites suggests that these G10 strains may have evolved through genetic reassortment between bovine and human strains
Diarrhoea disease is a leading cause of preventable death in children under 5 years of age in developing countries. Diarrhoea is estimated to be responsible for 22% of the 10 million annual under-five deaths worldwide and ranks second to neonatal deaths as the major cause of mortality in children . Rotavirus has been recognized as a cause of severe diarrhoea in children worldwide and a major cause of morbidity in the developed world and both morbidity and mortality in the developing world. Parashar et. al. estimate that there are approximately 2 million rotavirus related diarrhoea hospitilizations and more than 500,000 rotavirus-related deaths in children under 5 years of age annually with the majority occurring in developing countries .
Rotaviruses are members of the Reoviridae family and possess a genome that consists of 11 segmented double-stranded RNA genes that code for both structural and non-structural proteins. They have a triple-layered structure consisting of a core encapsulating the segmented RNA genome, an inner capsid which bears group determinant antigens and an outer capsid formed by two proteins involved in virus neutralization and protective immunity and whose antigenic characteristics have determined their serotype specificity. [3,4]. To date, 7 groups of rotaviruses (A–G) have been isolated and characterised  with group A rotaviruses identified as responsible for more than 90% of all human rotavirus infections. Of the 16 G genotypes and 27 P genotypes detected, 11 G genotypes (G1–G6, G8, G9, G10–G12) and 11 P types (P–P, P–P, P, P, P) have been identified in humans [5,6]. The globally most common rotavirus strains possess G genotype G1–G4 and G9 specificities. It should be noted however that the geographical distribution of rotavirus strains varies; whilst strains bearing the G1P genotype are predominant in Europe and the United states, rotavirus strains with G9 G5 and G8 VP7 genotype strains are increasing being detected with increased frequency in Africa, Asia and Latin America . The increased detection of rotavirus G9 strains throughout the world has already made this genotype the 4th most common cause of rotavirus diarrhoea in children [6,8–10]. Furthermore rotavirus G5, G6, G8 and G10 strains, known as major pathogens of pigs or cattle, recently have also been sporadically detected in humans [4,6,11–17]. Rotavirus strains bearing the G5 specificity, for example, have been described as an important pathogen in humans in Brazil . In addition, the epidemiologic importance of G6 viruses in Hungary , G8 viruses in Africa [12,19,20] and G10 viruses in India , appears to be increasing. Studies from Ghana and other developing countries have also shown the increased detection of rotavirus strains with unusual G and P combinations [6,17,19,21]. These reports have also cited large numbers of mixed infections in the population and the existence (especially in rural Africa, Asia and South America) of farm/domestic animals and humans living in close proximity to each other. The existence of these unusual strains is hypothesized to be due to the genetic reassortment between human and animal strains and that reassortment may be the driving force behind the large diversity of rotavirus strains observed [17,22–24].
As the incidence of rotavirus diarrhoea does not differ dramatically between industrialized and developing countries, disease control may not be achieved by improvements in water supply, hygiene and sanitation. Hence, the most plausible intervention likely to have a great impact on disease is vaccination. Present rotavirus vaccines are targeted at the most common global genotypes G1,G2,G3 and G4. However, as stated earlier, genotypes other than the globally common strains G1, G2, G3 and G4 are epidemiologically important in West Africa and other African countries. There is therefore the need to continue strain surveillance to aid in providing answers to issues of vaccine efficacy when they are introduced in Africa. We report here the detection and characterization of rotavirus G10 strains in hospitilized children with diarrhoea during a hospital based surveillance of severe diarrhoea in a rural population in Northern Ghana. This is the first report, to our knowledge of the detection of rotavirus G10 strains in hospitalized African children with severe diarrhoea.
A hospital-based burden of disease study was set up at Navrongo War Memorial Hospital in the district in October 2003 to assess the impact of rotavirus-associated diarrhoea in under 5 year olds in the district and to document strains associated with the hospitalizations. The Navrongo War Memorial Hospital is the main referral hospital in the district and serves a rural population of 140,000 consisting of mainly subsistence farmers. It is served by four health clinics all of which are within a 30 Km radius of the hospital. Stool samples, clinical and demographic data were obtained from all children less than 5 years of age with severe non-bloody diarrhoea visiting the emergency room and/or admitted to the wards for testing for rotavirus shedding. Diarrhoea was defined as the passage of 3 or more looser than normal (as described by care giver/mother on presentation at clinic) stools within a 24 hour period. The protocol was approved by the ethical committees of Noguchi Memorial Institute for Medical Research and the Navrongo Health Research Centre.
Group A rotaviruses were detected by enzyme immuno assay (EIA) with the commercial rotavirus IDEA kit (DAKO, UK) on 10% suspensions of faecal material as per manufacturer’s instructions.
The viral dsRNA was extracted and purified from 10% stool suspensions as previously described  and electrophoresed for 20 hours at 100V in a 10% vertical acrylamide slab gel. The RNAs were visualised by the silver staining technique as described by Herring et al .
Rotavirus strains identified by both EIA and PAGE were further characterized by RT-PCR with genotype specific primers as described earlier . Briefly, a semi-nested multiplex RT-PCR was performed on purified ds RNA extracted from EIA and PAGE positive stool samples after specific priming with VP7 and VP4 consensus primer pairs followed with a second round multiplex PCR which incorporated G  and P  type specific primers in separate G and P genotyping reactions. All the PCR products were examined by gel electrophoresis in a 1.2% agarose gels (Seakem, Flowgen) containing 4 µg/ml ethidium bromide. A Subset of the EIA, PAGE and RT-PCR untypable samples were sent to the National Institutes of Health (NIH), USA, and were further analyzed by RT-PCR using 3 different sets (i.e., the H2 pool, C pool, and A pool) of type-specific primers for the VP7 gene and one set of type-specific primers for the VP4 gene as described previously .
Virus isolation was done as previously described  Briefly, approximately 100 µl of clarified (at 10,000 × g for 5 min) and trypsin-activated 10% stool suspension of strains 86 and 402 were inoculated onto tube cultures of MA104 cells. After an hour adsorption period at 30°C, tubes were washed once and fed with Eagle’s minimum essential medium containing 0.5 µg of trypsin per ml. Cells were incubated for seven days on a roller apparatus at 37°C and were observed for CPE. Culture-adapted Ghanaian G10 strain 86 was triply plaque-purified, and then analyzed by a 60% plaque reduction neutralization (PRN) assay against guinea pig hyperimmune antiserum raised to selected prototype rotavirus strains belonging to each of the 14G serotypes as described previously  Subgroup specificity of culture-adapted strains 86 and 402 was determined by subgroup-specific monoclonal antibody-based enzyme-linked immunosorbent assay as previously described 
A full-length cDNA of the VP7 or VP8* gene of the culture-adapted strains 86 and 402 was amplified, purified with the Wizard SV Gel and PCR Clean-Up System (Promega) and sequenced as described previously (Hoshino et al., A porcine G9 rotavirus strain shares neutralization and VP7 phylogenetic sequence lineage 3 characteristics with contemporary human G9 rotavirus strains. Virology 2005;332:177–188). Two independent cDNA clones with the desired VP7 gene or VP8* gene insert were sequenced at least twice using the BigDye Terminator Cycle Sequencing Read Reaction v3.1 kit (Applied Biosystems) with M13 forward and reverse primers and the ABI PRISM3100 automated DNA sequencer. Sequence alignment was carried out with DNASIS software (Hitachi software) and phylogenetic analyses were made using UPGMA method with the MacVecter 7.1 program (Accelrys Inc.).
Oligonucleotide microarray hybridization assay was performed according to methods described by Chizhikov  with a slight modification at the National Institute of Health (NIH), USA. In this modification, a single primer extension was employed instead of a nested PCR for generation of the fluorescent labeled samples . The fluorescent labeled single-stranded DNA samples for hybridization were generated by the single primer extension reaction and the Cy5-labeled samples separated from the non-incorporated Cy5-dCTP by centrifugation. PCR-ELISA was performed as previously described . Briefly, the biotinylated amplicons of VP7 or VP4 gene immobilized onto streptavidin-coated 96-well microplates were hybridized to digoxigenin-labeled G or P type-specific oligoprobes (3 probes/type). The hybrids were then detected using antidigoxigenin Fab fragment labeled with peroxidase and the reaction was measured spectrophotometrically.
Out of a total of 289 stool samples collected from children with severe diarrhoea during the study period and examined for rotaviruses, 115 (39.8%) were found to be shedding rotaviruses in their stools. Seven children (2.4%) were found to be shedding non-group A rotaviruses; these were negative in the rotavirus EIA assay but showed typical rotavirus electrophoresis pattern by PAGE. Ninety samples tested rotavirus positive by both PAGE and ELISA. Sixty nine out of these 90 (76.7%) isolates had enough faecal material for further analysis and hence were subjected to RT-PCR for the determination of their genotypes (Table 1). The common G and P types detected were G1 (42.0%) and P (36.2%) and the common circulating strain was G1P (27.5%). Rotavirus strains with the unusual genotypes G1P,G3P,G2P constituted 24.6% of strains genotyped. Four strains could not be assigned a G genotype and one a P genotype.
A subset of 17 samples made up of the 5 untypable strains and 12 which could be assaigned a G/P specificity by our assays were sent to the National Institutes of Health (NIH), USA, for confirmation and further anaylsis by RT-PCR using 3 different VP7type-specific primers sets (i.e., the H2 pool, C pool, and A pool) as described previously . Four samples (86, 152, 402 and 404) were found to bear G10P specificity. These results were further confirmed by (i) PCR-ELISA and (ii) oligonucleotide microarray hybridization (data not shown) . The G10 strains, 86 and 152 were recovered from a 7 months old girl and a 9 months old boy, respectively, seen at the emergency clinic, whilst strains 402 and 404 were recovered from a 5 months old baby boy and a 4 months old baby girl, respectively, with severe diarrhoea on admission at the paediatric ward of the same hospital.
Culture-adapted and triply plaque-purified G10 strain 86 exhibited (i) a significant two-way cross-neutralization relationship with the prototype G10 strains B223 and KC-1; (ii) no significant neutralization relationship to viruses belonging to G1–G9, and G11–G14 (Table 2). Culture-adapted strains 86 and 402, which displayed similar, if not identical, long electrophoretic patterns by PAGE (Fig. 1), were shown to belong to subgroup II (data not shown).
VP7 nucleotide sequence comparison (DNAMAN for WINDOWS, Lynnon Corporation) of the isolated Ghanaian G10 strains and published VP7 sequences of other rotavirus G10 strains showed more than 99% identity between the Ghanaian isolates and more than 97% nucleotide sequence identity to other West African G10 strains (Table 3). Both strains showed between 88.8% and 88.9% identity to the Thai human G10 strain Mc35. They however showed less than 86% identity to human G10 strains from India (I321), Brazil (R239), the prototype G10 bovine strain from USA (B233), the lamb strain from China and the pig G10 strain (Thai). Similarly the West African strains exhibited a high amino acid identity of >94% amongst themselves. A comparison of the amino acid compositions of the VP7 neutralization domain A (aa87–101); B (aa143–152) and C (aa208 –221) showed a high conservation of amino acid composition between the Ghanaian isolates and other G10 strains except for the substitution of the negatively charged Glutamic acid (E) with the positively charged lysine (K) at amino acidic position 149 in antigenic site B (data not shown).
A comparison of the nucleic acid composition of the VP8* gene of the Ghanaian G10P strains with other published P strains (Table 4) showed a nucleic acid identity of between 93.5–96.7% with reference P human rotavirus strains MW23, M37, ST3 and US1205. The Ghanaian G10 strains, 86/Hu/Ghana and 402/Hu/Ghana exhibited >93% amino acid identity with G10 strains MW23, M37, ST3 and US1205. They, however, showed more than 99% amino acid identity between themselves.
Phylogenetic analysis of the VP7 gene of Ghanaian G10 isolates was carried out with 8 reference G10 VP7 gene sequences obtained from available databases. The dendrogram indicates four distinct lineages with a common progenitor strain (Fig. 2). One lineage was formed by the West African human G10 strains and the Thai human G10 strain Mc35. Within this lineage, two sub-lineages were formed by the West African strains and the other by the Thai strain Mc35. The Ghanaian G10 strains were found to clustering together in the West African G10 lineage (Fig. 2). The second lineage was formed by the prototype US bovine strain B223, the Indian human strain I321 and the Thai pig G10 strain whilst the Chinese lamb G10 and the Brazilian human G10 strains R239 formed the third and fourth lineages, respectively.
A phylogenetic analysis of the VP8* gene of strains 86 and 402 carried out with reference human and animal P strains showed the Ghanaian strains being genetically more related to human P strains clustering with strains belonging to the P lineage 1a [37,38] (Fig. 3).
The anticipated licensure and introduction of rotavirus vaccines in the developing world will help to reduce the disease burden due to rotavirus diarrhoea . Serum neutralizing antibody responses following either primary  or experimental oral [41–43] rotavirus infection appear to be homotypic, broadening after subsequent infections eventually leading to the development of antibodies broadly-reactive to all genotypes [13,44]. Whether serotype-specific neutralizing antibodies play an important role in protection against rotavirus diarrheoa is still under discussion. However, studies in experimental animals indicate that VP4 and VP7 are independent protective antigens, and that antibodies to either protein can confer resistance to virulent rotavirus in a type-specific manner [43,45–49]. Whilst the first rotavirus infection is usually the most severe, subsequent infections provide less severe outcomes . Rotavirus strains G1, G2, G3 and G4 constitute more than 88.5% of globally identified strains . Present generation rotavirus vaccines, designed to mimic the protection provided against severe outcomes, are hence targeted at these strains. However it must be noted that there are regional and geographical differences in strain distribution that may impact on the efficacy of these vaccines. For example, whilst the strains G1P, G2P, G3P and G4P represent >90% of North American and European strains, they represent only 68% of strains identified in Asia and South America and 50% in Africa . Furthermore, strains with unusual G/P combinations constitute more than 14% of reported rotavirus isolates from Asia, 27% in Africa, 11% in South America, 5% in North America, 1.4% in Europe and 0.1% in Australia . Results from rotavirus surveillance studies across Africa, spearheaded by the African Rotavirus Surveillance Network (AFRN) over the past 8 years, are beginning to identify increased diversity of strains, uncommon strains and strains typically found in animals [19,20,51]. Rotavirus strain diversity evolves by point mutations and/or genetic reassortment between strains concurrently infecting an individual . The detection of G10 strains in Ghana and subsequently in other West African countries of Cameroon and Cote d’Ivoire is very interesting since rotavirus G10 strains are commonly found in animals, particularly cattle, pigs and lambs . Sharing of genomic similarities between rotaviruses from different animal species is regarded as evidence of interspecies transmission of rotaviruses that may occur as a whole viron or genetic reassortment. Keeping of domestic animals is very common in Ghana and animals and humans usually live in close proximity and share (in some cases) same water bodies in very rural settings. This condition provides a perfect opportunity for dual infections, which is manifested by the high incidence of mixed detected infections in Africa  and genetic reassortment. Zoonotic transmission of animal rotavirus strains is recognized as an important vehicle contributing to the diversity of rotaviruses in humans . Although there is no available data on circulating bovine rotavirus strains in Ghana, the association of the Ghanaian G10 strains with the typical human subgroup II specificity, the P specificity common in Africa, the genetic relatedness of their VP8* genes to human P strains and the marked similarity in their VP7 antigenic sites suggests that these G10 strains may be due to genetic reassortment occurring between bovine and human strains. A total genome analysis and comparison with other isolated bovine strain from Ghana will help will confirm our hypothesis. Rotavirus strains 86 and 402 were found to be much more closely related to the West African human G10 strains than to the other Asian human G10 strains suggesting that they might have occurred through different events. Recent trends from surveillance studies in Africa have shown an increasing proportion of untypable rotavirus (Abstracts, Symposium on Epidemiology and Surveillance of Rotaviruses in Africa, African Health Sciences Congress, Mauritius, July 2008) For lack of resources and technology, these have remained untyped except for a few which have been sequenced to determine their genotypes, thus underestimating the detection of rotaviruses of zoonotic origin in Africa. Monitoring the rotavirus types circulating in Ghana and the rest of Africa is crucial for the evaluation of introduction of future rotavirus vaccination programmes in addition to shedding some more light on the spread of newly emerging and unusual genotypes in the sub-region. The use of expanded genotyping primer sets and sequence analysis (although much more expensive) will help to determine the G and/or P genotypes of these untypable strains giving more information on how widespread animal and unusual strains are in human populations in Africa.
We are grateful to the field staff of the Navrongo Health Research Centre, Navrongo, Ghana and staff of the Department of Electron Microscopy and Histopathology, Noguchi Memorial Institute for Medical Research, Legon, Ghana for their technical support. We also express our thanks and gratitude to all mothers who allowed their children to participate in the study. The project was approved by the Ethics Committees of the Noguchi Memorial Institute for Medical Research and the Navrongo Health Research Centre, Ghana. This research was supported in part by the Rotavirus Vaccine Programme, PATH, MIE Hospital Diarrhoea Research Programme, the World Health Organization and the Intramural Research Program of the National Institute of Allergy and Infectious Diseases, National Institutes of Health, USA.
Nucleotide sequence accession number
The nucleotide sequences of the Ghanaian strains have been deposited at the GenBank nucleotide sequence databases with the following accession numbers AY843333 (86, VP7), AY843335 (86, VP8*), AY843332 (402, VP7) and AY843334 (402, VP8*).