Virulenzminderung eines onkogenen Herpesvirus durch rechnerunterstützte Viruskonstruktion
Virologie
Zusammenfassung der Projektergebnisse
Codon pair bias deoptimization enables highly efficient attenuation of viruses. In contrast to other attenuation methods, live attenuated virus vaccine candidates can be rationally designed and produced within days. The technique involves recoding of viral genes, while preserving their codon bias and amino acid sequence. Recoding increases the number of codon pairs that are statistically underrepresented in protein coding sequences of the viral host, and involves swapping of available synonymous codons. While the exact mechanisms that cause the attenuation of recoded viruses is unknown, the prevailing theory proposes that underrepresented codon pairs interfere with efficient mRNA translation and/or protein processing, and this is manifested in a reduced level of protein production. Viruses with codon pair deoptimized genes cannot express their genetic information as efficiently as their wild-type parents, and therefore reproduce at reduced levels. The reduction of the reproductive fitness enables the host to gain the upper hand in controlling virus replication by innate and adaptive immune response. Codon pair deoptimization has led to attenuation of many different RNA viruses, but it has never been applied on large double-stranded DNA viruses, such as poxviruses, asfarviruses, or herpesviruses. We used codon pair deoptimization to attenuate an oncogenic Marek’s disease virus (MDV). To determine if MDV can get attenuated by codon pair deoptimization we constructed MDV mutants that carried recoded genes. The recoded genes were either codon pair optimized, randomized, or deoptimized. The codon pair optimized genes contained an elevated number of codon pairs that are overrepresented in host genome, and the deoptimized genes contained an increased number of underrepresented codon pairs of the host. We studied RNA and protein production from the recoded MDV genes first by transient transfection experiments. We observed that the mRNA levels in transfected cells correlated with the codon pair bias of the recoded genes: the mRNA levels were in general higher in cells transfected with codon pair optimized genes and lower in cells transfected with codon pair deoptimized genes. Correspondingly, and in line with our expectations, codon pair optimized genes produced more, and codon pair deoptimized genes produced less protein than the parental genes. Viruses with codon pair randomized and optimized genes grew equally good, or better than the parental virus in cell culture. On the contrary, viruses with codon pair deoptimized genes showed moderate to severe growth defects in replication. In addition, the level of codon pair deoptimization directly correlated with the level of viral attenuation. These results suggested that some of the recoded viruses might also be attenuated in vivo. Viruses that carried the recoded gene of viral DNA polymerase UL30 or the principal MDV oncogene meq were evaluated for their pathogenicity by animal experiments. In the case of UL30, the severity of viral attenuation in vivo correlated with the viral replication in cell culture and also the level of codon pair deoptimization. Nonetheless, the deoptimized UL30 MDV mutants, despite being severely attenuated in vitro, remained moderately oncogenic for its host. Interestingly, codon pair optimized UL30 viruses, which grew better in vitro than the parental virus, were not more pathogenic than the parental virus in chickens. Surprisingly, neither codon pair deoptimization, randomization, or optimization affected pathogenicity of MDV viruses that contained the recoded meq oncogene in vivo. Our results showed that only small amount of oncoprotein meq is sufficient for MDV transformation of infected cells, because not even a severe reduction of meq protein production has affected oncogenic potential of MDV. The results of this project imply that codon pair deoptimization might is an applicable strategy for attenuation of other herpesviruses, and other large double-stranded DNA viruses. However, this strategy might not be ideal for attenuation of tumor viruses.
Projektbezogene Publikationen (Auswahl)
-
(2018) Attenuation of Viruses by Large-Scale Recoding of their Genomes: the Selection Is Always Biased. Current clinical microbiology reports 5 (1) 66–72
Osterrieder, Nikolaus; Kunec, Dusan
-
(2016) Codon Pair Bias Is a Direct Consequence of Dinucleotide Bias. Cell Rep. 14 (1):55-67
Kunec D, Osterrieder N
