Utilizing Sense Codon Reassignment to Understand the Plasticity of the E. coli Genetic Code

The genetic code was set and has remained largely unchanged since the evolution of the last universal common ancestor of life on earth. Several lines of evidence suggest that this probable â??frozen accidentâ? is potentially fairly plastic. Although the meaning of the 64 codons has not changed in 2 billion years of evolution, huge variation in the use of synonymous codons exists. The components that are employed to translate the code are also highly variable across species. The orthogonality of certain tRNA/aminoacyl tRNA synthetase pairs between species has been exploited to expand the genetic code, most commonly by incorporating non-canonical amino acids in response to the amber stop codon. Unfortunately, the number of stop codons is limited, and the method of nonsense codon suppression is not readily extendable to additional codons. In contrast, sense codon reassignment should be a broadly extendable, general method for genetic code expansion. The ability to expand the number of available genetic codes from the 100s presently available (employing 21 amino acids) to 10,000s (22 amino acids) to 1,000,000s (23 amino acids) would open vast new vistas of protein sequence and functional space for exploration and exploitation.

Engineering the genetic code requires a detailed understanding of the molecular and systems level interactions governing the entire process of translation. To evaluate the potential for sense codon reassignment in E. coli, we have evaluated 20 of the 21 codons read via a wobble interaction. Our approach evaluates the ability of a tyrosine-charging M. jannaschii aminoacyl tRNA synthetase to recognize M. jannaschii tRNAs with altered anticodons and the extent to which Tyr-aminoacylated M. jannaschii tRNAs can compete with endogenous E. coli tRNAs to decode and reassign the meaning of sense codons. Variants of the M. jannaschii tyrosine pair have been evolved to incorporate ~100 different non-canonical amino acids into proteins in response to the amber stop codon.

Our work augments studies of natural translational systems and examines the plasticity and fitness costs imposed by redefining the E. coli genetic code. We show that to some extent every wobble codon in E. coli can be reassigned to another amino acid. We examine sense codon reassignment efficiency in the context of the prevalence of each codon in the E. coli genetic code and the predicted effect on tRNA aminoacylation as a result of altering the M. jannaschii tRNA anticodon to decode a given sense codon. We discuss observed growth defects of the E. coli strains as a result of reassignment, which are not always correlated with increased reassignment. We evaluate the ability of orthogonal tRNAs to decode only the desired sense codon and show that consideration of possible tRNA modifications by endogenous E. coli enzymes is of high importance for maintaining the fidelity of orthogonal tRNAs.