Evolution of multiple, mutually orthogonal prolyl-tRNA synthetase/tRNA pairs for unnatural amino acid mutagenesis in Escherichia coli

Abstract

The site-specific incorporation of unnatural amino acids (UAAs) into proteins in living cells relies on an engineered tRNA/aminoacyl-tRNA synthetase (tRNA/aaRS) pair, orthogonal to the host cell, to deliver the UAA of choice in response to a unique nonsense or frameshift codon. Here we report the generation of mutually orthogonal prolyl-tRNA/prolyl-tRNA synthase (ProRS) pairs derived from an archaebacterial ancestor for use in Escherichia coli. By reprogramming the anticodon-binding pocket of Pyrococcus horikoshii ProRS (PhProRS), we were able to identify synthetase variants that recognize engineered Archaeoglobus fulgidus prolyl-tRNAs (Af-tRNA(Pro)) with three different anticodons: CUA, AGGG, and CUAG. Several of these evolved PhProRSs show specificity toward a particular anticodon variant of Af-tRNA(Pro), whereas others are promiscuous. Further evolution of the Af-tRNA(Pro) led to a variant exhibiting significantly improved amber suppression efficiency. Availability of a prolyl-tRNA/aaRS pair should enable site-specific incorporation of proline analogs and other N-modified UAAs into proteins in E. coli. The evolution of mutually orthogonal prolyl-tRNA/ProRS pairs demonstrates the plasticity of the tRNA-aaRS interface and should facilitate the incorporation of multiple, distinct UAAs into proteins.

Fig. 1.

Suppression efficiency of tRNA^Pro/ProRS pairs is affected by the anticodon. (A) Different anticodon variants of the tRNA^Pro used in this study. In the box at the far left is shown the wild-type tRNA_NGG^Pro, which has an anticodon sequence similar to tRNA_AGGG but not to the other two variants. (B) Prolyl-tRNA_CUA is not charged efficiently by wild-type ProRS. The efficiency of Af-tRNA_CUA^Pro in TAG suppression was measured as ChlorR using the pRepCM3b reporter plasmid with an empty pBK vector (control) or with a pBK vector expressing PhProRS. (C) Af-tRNA_AGGG^Pro is charged efficiently by wild-type ProRS. The efficiency of Af-tRNA_AGGG^Pro in suppressing CCCT was measured as ChlorR using the pRepCM3b reporter plasmid with an empty pBK vector (control) or with a pBK vector expressing six different wild-type ProRSs from the following archaeal species: P. horikoshii, M. mazei, Sulfolobus solfataricus, Methanocaldococcus jannaschii, Halobacterium sp. (NRC-1), and T. thermophilus.

Fig. 2.

Evolution of the anticodon-binding pocket of PhProRS to charge different Af-tRNA^Pro anticodon variants. (A) Anticodon-binding site of ProRS in the crystal structure from T. thermophilus. The consensus sequence of the proline anticodon (GG), shown in magenta, interacts strongly with the amino acid residues in the binding pocket. Highlighted amino acid residues, involved in the recognition of this GG motif, were randomized, and the corresponding residues in PhProRS are shown. (B) Sequences of evolved PhProRS mutants that efficiently charge different anticodon variants of Af-tRNA^Pro. PhProRS variants highlighted in magenta are anticodon permissive and efficiently charge Af-tRNA^Pro bearing any of the three anticodons, CUA, AGGG, or CUAG. (C) PhPRS_CUA-h1/Af-tRNA_CUA^Pro, PhPRS_CUAG-h1/Af-tRNA_CUAG^Pro, and wtPhPRS/Af-tRNA_AGGGa^Pro are mutually orthogonal tRNA/aaRS pairs suppressing TAG, CTAG, and CCCT, respectively: Af-tRNA_CUA^Pro, Af-tRNA_AGGGa^Pro, and Af-tRNA_CUAG^Pro in the pREP, pREP4b-ccct, and pREP4b-ctag reporter plasmids, respectively, were used to determine the TAG, CCCT, and CTAG suppression activity measured as ChlorR. Using these reporter plasmids the TAG (blue), CCCT (red), and CTAG (green) suppression activities of two anticodon-specific PhProRS mutants (CUA-h1, CUAG-h1), wild-type PhProRS, and one anticodon-permissive PhProRS mutant (CUA-h3) were measured. The control experiment measured ChlorR elicited by the reporter plasmid in the absence of any PhProRS variants (empty pBK vector) (D) Specific PhProRS variants (identified at the top) allow E. coli to survive in the presence of 120 μg/mL chloramphenicol only when paired with the Af-tRNA^Pro bearing the cognate anticodon (identified at the right).

Fig. 3.

Evolution of a highly efficient amber suppressor, Af-tRNA_CUA^Pro. (A) The highlighted segments of Af-tRNA_CUA^Pro were randomized. A highly efficient variant incorporating the indicated mutations, Af-tRNA_CUA^Pro-h8, was identified. (B) Suppression activity of the wild type and h8 Af-tRNA_CUA^Pro, measured as ChlorR, in the presence (black bar) and absence (white bar) of the cognate PhPRS_CUA-h1. The h8-tRNA variant has enhanced activity relative to wild-type tRNA and remains orthogonal to E. coli. (C) Expression levels of GFP-Tyr151TAG using pEvol-Pro encoding the original (wild type) or the evolved h8-tRNA. Expression levels of the same GFP-Tyr151TAG construct using pEvol-MjYRS (wild type) and of the corresponding wild-type GFP construct (GFPwt), not carrying a nonsense codon, are provided for comparison.