Generating permissive site-specific unnatural aminoacyl-tRNA synthetases.

Genetically incorporated unnatural amino acid (UAA) technologies are powerful tools that are greatly enhancing our ability to study and engineer biological systems. Using these techniques, researchers can precisely control the position and number of novel chemical moieties in a protein, via introducing the novel R group of UAAs, that are genetically encoded in the protein's primary structure. The substrate recognition properties of a natural aminoacyl-tRNA synthetase (aaRS) must be modified in order to incorporate UAAs into proteins. Protocols to do so are technically simple but require time and optimization, which has significantly limited the accessibility of this important technology. At present, engineered unnatural aminoacyl-tRNA synthetases (UaaRS) are evaluated on their translational efficiency (the extent to which they allow for incorporation of UAAs into protein) and fidelity (the extent to which they prevent incorporation of natural amino acids). We propose that a third parameter of substrate recognition, permissivity, is equally important. Permissive UaaRSs, whose relaxed substrate recognition properties allow them to incorporate multiple unnatural amino acids (but not natural amino acids), would eliminate the need to generate new UaaRSs for many new UAAs. Here, we outline methods for quickly and easily assessing the permissivity of existing UaaRSs and for generating permissive UaaRSs. In proof of principle experiments, we determined the degree of permissivity of two UaaRSs for a family of structurally related fluorinated UAAs ((19)F-UAAs). We then increased the permissivity of the initial UaaRSs to allow for incorporation of the family of (19)F-UAAs. Finally, we validated the utility of these new (19)F-UAAs as probes for fluorine NMR studies of protein structure and dynamics. We expect that results of this work will increase the accessibility of UAA technology and the use of new UAAs in proteins.