Virus attenuation by genome re-encoding is a pioneering approach for generating effective live-attenuated vaccine candidates. Its core principle is to introduce a large number of synonymous substitutions into the viral genome to produce stable attenuation of the targeted virus. Introduction of large numbers of mutations has also been shown to maintain stability of the attenuated phenotype by lowering the risk of reversion and recombination of re-encoded genomes. Identifying mutations with low fitness cost is pivotal as this increases the number that can be introduced and generates more stable and attenuated viruses. Here, we sought to identify mutations with low deleterious impact on the in vivo replication and... More
Virus attenuation by genome re-encoding is a pioneering approach for generating effective live-attenuated vaccine candidates. Its core principle is to introduce a large number of synonymous substitutions into the viral genome to produce stable attenuation of the targeted virus. Introduction of large numbers of mutations has also been shown to maintain stability of the attenuated phenotype by lowering the risk of reversion and recombination of re-encoded genomes. Identifying mutations with low fitness cost is pivotal as this increases the number that can be introduced and generates more stable and attenuated viruses. Here, we sought to identify mutations with low deleterious impact on the in vivo replication and virulence of yellow fever virus (YFV). Following comparative bioinformatic analyses of flaviviral genomes, we categorised synonymous transition mutations according to their impact on CpG/UpA composition and secondary RNA structures. We then designed seventeen re-encoded viruses with 100-400 synonymous mutations in the NS2A-to-NS4B coding region of YFV Asibi and Ap7M (hamster-adapted) genomes. Each virus contained a panel of synonymous mutations designed according to the above categorisation criteria. The replication and fitness characteristics of parent and re-encoded viruses were compared in vitro using cell culture competition experiments. In vivo laboratory hamster models were also used to compare relative virulence and immunogenicity characteristics. Most of the re-encoded strains showed no decrease in replicative fitness in vitro. However, they showed reduced virulence and, in some instances, decreased replicative fitness in vivo. Importantly, the most attenuated of the re-encoded strains induced robust, protective immunity in hamsters following challenge with Ap7M, a virulent virus. Overall, the introduction of transitions with no or a marginal increase in the number of CpG/UpA dinucleotides had the mildest impact on YFV replication and virulence in vivo. Thus, this strategy can be incorporated in procedures for the finely tuned creation of substantially re-encoded viral genomes.
