Quality control by trans-editing factor prevents global mistranslation of non-protein amino acid α-aminobutyrate

Abstract

Accuracy in protein biosynthesis is maintained through multiple pathways, with a critical checkpoint occurring at the tRNA aminoacylation step catalyzed by aminoacyl-tRNA synthetases (ARSs). In addition to the editing functions inherent to some synthetases, single-domain trans-editing factors, which are structurally homologous to ARS editing domains, have evolved as alternative mechanisms to correct mistakes in aminoacyl-tRNA synthesis. To date, ARS-like trans-editing domains have been shown to act on specific tRNAs that are mischarged with genetically encoded amino acids. However, structurally related non-protein amino acids are ubiquitous in cells and threaten the proteome. Here, we show that a previously uncharacterized homolog of the bacterial prolyl-tRNA synthetase (ProRS) editing domain edits a known ProRS aminoacylation error, Ala-tRNA^Pro, but displays even more robust editing of tRNAs misaminoacylated with the non-protein amino acid α-aminobutyrate (2-aminobutyrate, Abu) in vitro and in vivo. Our results indicate that editing by trans-editing domains such as ProXp-x studied here may offer advantages to cells, especially under environmental conditions where concentrations of non-protein amino acids may challenge the substrate specificity of ARSs.

Keywords: Aminoacyl-trna synthetases; ProRS; Rhodopseudomonas palustris; trans-editing.

Figure 1.

Docking study of ProXp-x substrate specificity. Final docked model of (A) CCA-Ala (teal) and (B) CCA-Abu (green) bound in the T. thermophilus ProXp-x active site (gray), and (C) a previously published model of CCA-Ala (magenta) bound in the E. faecalis active site (gray).²⁸ Black dotted lines in all panels denote potential hydrogen bond interactions and red dotted lines in (B) indicate close proximity of the Abu sidechain to hydrophobic residues A37 and L136. Residue numbers in parentheses denote corresponding positions in R. palustris ProXp-x for (A) and (B), and in E. coli ProRS in panel (C). Non-interacting residues and the two 5′ cytidines of the ligand are not shown for clarity. (D) Theoretical binding free energies for different CCA-conjugated amino acid substrates determined using MM-PBSA method with error bars representing the standard error of the calculation.

Figure 2.

Editing activities of R. palustris ProRS. (A) Pre-transfer editing time course showing formation of AMP in the presence of 3 mM Pro (•), 500 mM Ala (▪), 3 mM Cys (♦), and 300 mM Abu (▴). (B) Deacylation of 0.5 μM Pro-tRNA^Pro by 0.5 μM R. palustris ProRS (•) or 120 mM NaOH (○), 0.5 μM Ala-tRNA^Pro by 0.5 μM R. palustris ProRS (▪) or 1 μM E. coli ProRS (□), 0.5 μM Cys-tRNA^Pro by 0.5 μM R. palustris ProRS (♦) or 0.5 μM R. palustris YbaK (⋄), and 0.5 μM Abu-tRNA^Pro by 0.5 µM R. palustris ProRS (▴) or 120 mM NaOH (▵). All reactions were performed in triplicate as described in the Experimental Methods. Lines represent exponential fits of the data and error bars indicate SD.

Figure 3.

Deacylation of aa-tRNA by R. palustris ProXp-x. Deacylation time course of (A) aa-tRNA^Pro and (B) aa-tRNA^Val substrates. Reactions were performed with 0.75 µM R. palustris ProXp-x and 0.1 µM Pro-tRNA^Pro (•), 0.1 µM Cys-tRNA^Pro (▴), 0.1 µM Abu-tRNA^Pro or Abu-tRNA^Val (♦), 0.1 µM Ala-tRNA^Pro or Ala-tRNA^Val (▪), 0.1 µM Val-tRNA^Pro or Val-tRNA^Val (▾), and 0.1 µM Thr-tRNA^Val (). All reactions were performed in triplicate as described in the Experimental Methods. Lines represent exponential fits and error bars indicate SD.

Figure 4.

Effect of R. palustris ProXp-x on Abu misacylation by R. palustris ProRS. Formation of mischarged Abu-tRNA^Pro by 0.5 µM R. palustris ProRS alone (○), 0.5 μM R. palustris ProRS and 0.1 μM ProXp-x (▪), 0.5 μM R. palustris ProRS and 0.5 μM ProXp-x (▵), 0.5 μM R. palustris ProRS and 1.5 μM ProXp-x (▾), 0.5 μM R. palustris ProRS and 1.5 μM K45A ProXp-x (□). Reactions were performed at 30 °C with 6 μM E. coli tRNA^Pro. Lines represent exponential fits of the data.

Figure 5.

Deacylation of Abu-tRNA variants by R. palustris ProXp-x. (A) Deacylation of 0.5 μM Abu-tRNA^Pro (•), A73C Abu-tRNA^Pro (▵), and C73/G1:C72 Abu-tRNA^Pro (▪) by 0.5 μM Rp ProXp-x. (B) Comparison of acceptor stem compositions of E. coli tRNA^Pro and tRNA^Val. Acceptor stem sequence elements shared between the two tRNAs, other than the universal CCA end, are boxed. All reactions were performed in triplicate as described in the Experimental Methods. Lines represent exponential fits of the data and error bars indicate SD.

Figure 6.

R. palustris ProXp-x rescues E. coli valS:T222P Abu sensitivity. The indicated bacterial strains were stamped onto minimal media agarose plates containing 20 mg/mL kanamycin and 0.3 mM thymidine. Plates were pre-streaked with 20 µl of 0.1 mM Abu vertically along the center of the plate and (A) lacked IPTG or (B) contained 20 μL of 100 mM IPTG. (1) E. coli valS⁺, (2) E. coli valS:T222P, (3) E. coli valS:T222P + ProXp-x (pET15b), (4) E. coli valS:T222P + pET15b (empty vector), (5) E. coli valS:T222P + K45A ProXp-x (pET15b).