Engineered orthogonal translation systems from metagenomic libraries expand the genetic code

Abstract

Genetic code expansion with non-canonical amino acids (ncAAs) opens new opportunities for the function and design of proteins by broadening their chemical repertoire. Unfortunately, ncAA incorporation is limited both by a small collection of orthogonal aminoacyl-tRNA synthetases (aaRSs) and tRNAs and by low-throughput methods to discover them. Here, we report the discovery, characterization, and engineering of a UGA suppressing orthogonal translation system mined from metagenomic data. We developed an integrated computational and experimental pipeline to profile the orthogonality of >200 tRNAs, test >1,250 combinations of aaRS:tRNA pairs, and identify the AP1 TrpRS:tRNATrpUCA as an orthogonal pair that natively encodes tryptophan at the UGA codon. We demonstrate that the AP1 TrpRS:tRNATrpUCA is highly active in cell-free and cellular contexts. We then use Ochre, a genomically recoded Escherichia coli strain that lacks UAG and UGA codons, to engineer an AP1 TrpRS variant capable of 5-hydroxytryptophan incorporation at an open UGA codon. We anticipate that our strategy of integrating metagenomic bioprospecting with cell-free screening and cell-based engineering will accelerate the discovery and optimization of orthogonal translation systems for genetic code expansion.