Human respiratory syncytial virus (RSV) is an important agent of pediatric respiratory tract disease worldwide and is responsible for a huge burden of morbidity and significant mortality. There is no licensed vaccine. Obstacles to vaccine development include the poor growth of the virus in cell culture, the semi- permissive nature of the infection in convenient experimental animals, and the difficulty of achieving an appropriate balance between immunogenicity (which depends on reasonable levels of virus replication) and attenuation (which depends on reduced levels of virus replication). We recently developed a method for producing RSV by the intracellular coexpression of cDNAs encoding a complete RSV replicative intermediate RNA (antigenome) and the N, P, L and M2-1 proteins, which together constitute a nucleocapsid that is fully competent for RNA synthesis. This provides an important tool for basic molecular and pathogenesis studies as well as a method for fine-tuning the level of attenuation of candidate vaccine viruses. RSV encodes ten mRNAs encoding eleven proteins (the M2 mRNA contains two overlapping ORFs encoding two separate proteins, M2- 1 and M2-2). We investigated whether individual RSV genes could be ?knocked out? without ablating the ability of the virus to grow in cell culture. To date, four RSV genes can be individually knocked out without loss of infectivity, namely NS1, NS2, SH and G. This was done in two cases (NS1 and NS2) by introducing stop codons into the translational open reading frame (ORF), and in all four cases also was done by completely deleting the gene, such that the genome is shorter and encodes one fewer mRNA. The NS1 knockout virus grows somewhat less well than the parent in cell culture, and its analysis is in progress. The NS2 knockout virus grows somewhat more slowly than does the wild type parent in a single step growth curve and forms pinpoint plaques. The version of the NS2 knockout virus in which stop codons had been introduced exhibits reversion to an NS2+ phenotype by reversion of the stop codons to sense. This was not observed with the gene deletion, illustrating its stability against reversion. The SH knockout virus forms plaques that were larger than wild type, and it grows as well or slightly better than wild type in cell culture in a single step growth curve. In mice, the SH knockout grew as well as wild type in the lower respiratory tract but was significantly restricted in the upper respiratory tract. It was equivalent to wild type with regard to the ability to induce RSV-specific serum antibodies and protection against challenge virus replication. In chimpanzees, the only experimental animal which approaches humans with regard to permissiveness for RSV infection and disease, the SH knockout virus was only slightly attenuated. Interestingly, however, it was associated with significantly reduced disease, i.e. rhinorrhea, which would be an ideal feature for a live-attenuated vaccine virus. The G knockout virus grows less well than its wild type parent in cell culture. Nonetheless, it does form plaques and can be propagated. The fact that its attachment activity is not essential for in vitro growth implies that one of the other surface glycoproteins can serve an auxiliary attachment function. We showed that RSV can accept the insertion of sequence encoding an additional, foreign mRNA. Chimeric genes were constructed in which the ORF encoding chloramphenicol acetyl transferase (CAT), green fluorescent protein (GFP) or luciferase (LUC) was engineered to be flanked by the RSV gene-start (GS) and gene-end (GE) transcription signals. Each transcription cassette was inserted into the leader-NS1, SH-G or F-M2 junctions in the antigenomic cDNA. Infectious recombinant viruses which each contained one foreign insert were recovered. Each foreign gene was expressed as an additional mRNA, and each protein was expressed to a level comparable to that of the RSV proteins. The presence of the short CAT or GFP genes (~750 bp) resulted in a small decrease in plaque size and a 20-fold reduction in virus yield. In contrast, the presence of the longer (~1750 bp) LUC gene reduced virus growth to the point that the virus could be propagated but not significantly amplified. This shows that the insertion of an additional gene can be a method of attenuation that appears to be length-dependent. We hypothesize that this restriction occurs at the level of packaging, which is currently being investigated. Insert stability was investigated for the CAT gene and found to be remarkably high. Deletion of foreign sequence was not observed, and the accumulation of point mutations was not rapid, such that after seven passages each gene had on average a single nucleotide substitution which, in the 25 plaques examined, did not affect protein expression or enzymatic activity. The gene deletions and insertions described above each resulted in a change in genome length and in the number of expressed mRNAs. We are examining transcription by these viruses to determine what effects, in any, these changes have on sequential transcription. We also are examining the effects of introducing longer-than-natural intergenic regions as possible attenuating mutations and to shed light on polymerase activities at the gene junctions. We previously showed that one of the attenuating mutations of the biologically-derived cptsRSV vaccine virus is a single nucleotide change (A9 to G, negative-sense) in the GS signal of the M2 gene. This mutation has now been inserted into the GS signal of the NS1, or NS2, or both genes, and evaluation of the resulting mutants is in progress. We previously showed that the naturally-occurring GE signals for these two genes are only 60% as efficient in directing transcriptional termination as are the signals for the other eight RSV genes. This is associated with an increased frequency of synthesis of the following readthrough mRNAs: NS1-NS2, NS2-N, NS1-NS2-N. Since internal ORFs of eukaryotic mRNAs are not efficiently translated, this would have the effect of reducing the amount of mRNA capable of synthesizing the NS2 and N proteins.
Showing the most recent 10 out of 36 publications
