Mechanism of virus attenuation by codon pair deoptimization

Published: 20-04-2020| Version 1 | DOI: 10.17632/rsxsyy5wrd.1
Nicole Groenke,
Jakob Trimpert,
Sophie Merz,
Andelé Conradie,
Emanuel Wyler,
Hongwei Zhang,
Orsalia-Georgia Hazapis,
Sebastian Rausch,
Markus Landthaler,
Klaus Osterrieder,
Dusan Kunec


Codon pair deoptimization (CPD), also known as synthetic attenuation virus engineering (SAVE), is an attenuation strategy that is based on genetic recoding of virus genomes. The strategy has the potential to revolutionize the development of viral vaccines, because one can rapidly produce efficacious and non-reverting virus vaccines. Yet, the mechanism behind attenuation is unknown. The recoding rearranges the positions of synonymous codons in viral genomes to create suboptimal codon pairs. However, CPD also results in an unintentional increase of the frequency of CpG dinucleotides in recoded sequences, because suboptimal codon pairs often contain CpG dinucleotides at the codon pair boundary. We systematically dissected the contribution of underrepresented codon pairs from that of CpG dinucleotides by studying a series of recoded influenza A virus mutants in which the two features were independently varied. We show unequivocally that suboptimal codon pairs were responsible for attenuation of recoded influenza viruses in tissue culture and in vivo, while an increase of CpG dinucleotides had no effect. Next, we identified how suboptimal codon pairs induce attenuation. We show that suboptimal codon pairs reduce mRNA stability and reduce translation efficiency of recoded codon pair- deoptimized genes. This in turn decreases protein output and directly causes virus attenuation. Several recent studies showed that codon optimality is a major determinant of mRNA stability. Here we demonstrate that the identity of codon pairs is another critical determinant of mRNA stability, which profoundly affects protein output and, ultimately, fitness of recoded viruses. Our work explains why exchanging the positions of synonymous codons can have dramatic consequences on gene expression of recoded genes. Furthermore, our work demonstrates that codon pair (de)optimization can be employed to modulate mRNA stability and protein output of synthetic genes for a broad range of biotechnological and pharmaceutical applications.