The chemical synthesis of nucleic acids enables the generation of recoded genes with novel properties. The synonymous genome recoding introduces a large number of mutations into recoded genes without changing the amino acid composition of encoded proteins. The goal of the recoding is to alter codon usage, codon pair usage or dinucleotide frequencies of the recoded genes and thereby change different biological properties. Codon optimization and codon pair optimization are used to improve protein production from recoded genes. On the contrary, the goal of codon deoptimization and codon pair deoptimization is to reduce the protein expression of recoded genes, which can be used for the production of experimental live-attenuated virus vaccine candidates. While codon and codon pair deoptimization are capable of producing attenuated viruses, the molecular basis behind the attenuation is not well understood. In my work, I confirmed that codon and codon pair deoptimization are both excellent tools for virus attenuation, and therefore suitable for the development of viral vaccines. These live-attenuated virus vaccines, prepared by both methods, induced in mice model a remarkable protection against WT virus challenge. The main goal of this project was to elucidate the molecular mechanism of virus attenuation by codon pair deoptimization. Due to low frequencies of CpG dinucleotides in the human genome, codon pair deoptimization unintentionally increases the number of CpG dinucleotides in the recoded genes. Therefore, it was assumed that CpG dinucleotides have a substantial impact on virus attenuation. I showed that increased numbers of CpG dinucleotides do not influence influenza A virus. However, I discovered that RNAs produced by codon pair deoptimized genes have higher degradation rates. Additionally, I found out that codon pair deoptimization reduces protein production of recoded genes also by decreasing their translation efficiency. These findings elucidate the highly complex molecular mechanism underlying the synonymous genome recoding. In the second part of my work, I showed that the commonly used codon optimization is a suitable method to increase the protein expression of target genes. More importantly, I demonstrated that also codon pair optimization has a great potential to improve RNA stability and protein production. Therefore, codon pair optimization might be a new powerful tool to enhance protein expression for diverse applications in biotechnology.
