Association of respiratory syncytial virus glyco- and non-structural proteins with disease severity in infected children

Abstract

Respiratory Syncytial Virus (RSV) is a major cause of bronchiolitis and pneumonia in infants, immunocompromised and the elderly in both developed and developing countries. Re-infections are common and G protein variability is one mechanism to overcome herd immunity. This is illustrated by the appearance of the BA genotype with a 60 nucleotide duplication dominating the subtype B genotypes in epidemics worldwide. To investigate the evolution of subtype A and B in South Africa (SA) since 2002 the genetic variability in the G protein was analysed in all recent strains isolated over four years (2006-2009) in SA hospitals. Bayesian analysis revealed a replacement of all subtype B genotypes previously identified in SA with the BA genotype since 2006, while subtype A genotypes identified in previous years are still circulating. The evolutionary rate of the SA BA genotype was shown to be 2.305 x 10-3 nucleotide substitutions/site/year while subtype A was shown to have 3.382 x 10-3 substitutions/site/year and drift was evident. The most recent common ancestor (MRCA) of the SA BA viruses was determined to date back to 1997. All SA BA isolates clustered with the BA-IV sub-genotype and the appearance of new sub-genotypes within this branch may occur if drift continues. Sequencing of the complete G protein of selected BA SA strains revealed an additional 6 nucleotide deletion. When comparing G protein variability with disease severity, the dominating subtype/genotype in each season was also the subtype/genotype that was associated with increased hospitalizations. No direct association between G protein variability and disease severity was seen. Subtype/genotype switching was evident over the four years most probably because of herd immunity. G protein variability may play a role in RSV‟s ability to re-establish annual epidemics by allowing immune evasion. Certain substitutions or alterations may enhance the fitness of viruses as is evident with the BA strains that replaced all other B genotypes previously identified in SA. The G protein‟s ability to accommodate such substantial changes and facilitate immune evasion may complicate vaccine development. It remains to be seen if this BA genotype will remain dominant or if the dominance will eventually fade because of herd immunity. Subtyping of RSV strains over the four years identified G protein PCR amplicons significantly reduced in size in 2 out of 209 clinical specimens. Sequence analysis revealed subtype B strains lacking nearly the entire G protein ectodomain in one HIV positive and one HIV exposed child hospitalized with pneumonia. G protein deletion mutants replicate effectively in vitro but have not been detected in nature. This study suggests that RSV clinical strains that lack most of the G protein gene may occur in immunocompromised patients with lower respiratory tract infection (LRTI). The molecular mechanism whereby this occurs is not clear, however reduced immune pressure in these patients may allow these strains to utilise the F protein for binding and replication. Further characterization of such strains may elucidate the replication and pathogenic potential, however low viral load and long term storage of specimens complicated the isolation of such strains. Acquisition of the 60 nucleotide duplication appeared to have improved the fitness of the BA viruses and more recent subtype B strains may need to be included in experimental vaccines to evaluate their efficacy in the current setting of evolved circulating strains. In addition, the association of clinical strains lacking most of the G protein with LRTI may have implications for the utilisation of certain attenuated strains in immunocompromised children. RSV is unique among paramyxoviruses in having two non-structural (NS) proteins that play a major role in inhibiting the host‟s immune response. Sequence and quantification analysis of these proteins were performed to determine its role in disease severity. We were unable to attribute specific protein polymorphisms with differences in disease severity but identified genome heterogeneity (quasispecies) within patients which may have an influence on the NS protein function, and also have an influence on the innate immune response and thus an effect on disease severity. When comparing patients with mild and severe disease, higher expression levels of NS1 were seen in the hospitalized group compared to the out-patient group, supporting our hypothesis that increased levels of NS may lead to enhanced suppression of the interferon pathway and in effect result in more severe disease. However, the opposite was found for NS2 with higher expression levels in the mild group. Genetic variability within the gene-end and gene-start sequences were not seen thus could not explain the differences in expression levels observed, although variability within the promoter area may need to be investigated.