Insights into catalysis by a knotted TrmD tRNA methyltransferase

The crystal structure of Escherichia coli tRNA (guanosine-1) methyltransferase (TrmD) complexed with S-adenosyl homocysteine (AdoHcy) has been determined at 2.5A resolution. TrmD, which methylates G37 of tRNAs containing the sequence G36pG37, is a homo-dimer. Each monomer consists of a C-terminal domain connected by a flexible linker to an N-terminal AdoMet-binding domain. The two bound AdoHcy moieties are buried at the bottom of deep clefts. The dimer structure appears integral to the formation of the catalytic center of the enzyme and this arrangement strongly suggests that the anticodon loop of tRNA fits into one of these clefts for methyl transfer to occur. In addition, adjacent hydrophobic sites in the cleft delineate a defined pocket, which may accommodate the GpG sequence during catalysis. The dimer contains two deep trefoil peptide knots and a peptide loop extending from each knot embraces the AdoHcy adenine ring. Mutational analyses demonstrate that the knot is important for AdoMet binding and catalytic activity, and that the C-terminal domain is not only required for tRNA binding but plays a functional role in catalytic activity.     

Ligand-mediated anticodon conformational changes occur during tRNA methylation by a TrmD methyltransferase

Orthologs of TrmD, G37 tRNA methyltransferases, have been analyzed with regard to post-tRNA binding events required to move the residue G37 in proximity to bound AdoMet for catalysis. This was approached initially by probing tRNA with T2 nuclease or Pb acetate in the presence, then absence, of Escherichia coli TrmD protein. Cleavage patterns clearly show that portions of the anticodon loop phosphodiester backbone are protected from cleavage only in the presence of sinefungin, a potent AdoMet analogue. This demonstrates that there must be considerable movement of the loop region and/or protein as the AdoMet site is occupied. Florescence energy transfer experiments were employed to better assess the movement of the G37 and G36 base residues in response to occupancy of the AdoMet site. When the Streptococcus pneumoniae TrmD protein was bound to synthetic tRNA(1)(Leu) substituted with 2-aminopurine at positions 36 and 37, fluorescence energy transfer analysis showed that a decrease in 2-aminopurine fluorescence occurs only when AdoMet is present. Taken together, these results suggest that the base to be methylated by the TrmD protein is mobilized into the active center after tRNA binding only when the AdoMet site is occupied.     

Recognition of guanosine by dissimilar tRNA methyltransferases

Guanosines are important for biological activities through their specific functional groups that are recognized for RNA or protein interactions. One example is recognition of N(1) of G37 in tRNA by S-adenosyl-methionine (AdoMet)-dependent tRNA methyltransferases to synthesize m(1)G37-tRNA, which is essential for translational fidelity in all biological domains. Synthesis of m(1)G37-tRNA is catalyzed by TrmD in bacteria and by Trm5 in eukarya and archaea, using unrelated and dissimilar structural folds. This raises the question of how dissimilar proteins recognize the same guanosine. Here we probe the mechanism of discrimination among functional groups of guanosine by TrmD and Trm5. Guanosine analogs were systematically introduced into tRNA through a combination of chemical and enzymatic synthesis. Single turnover kinetic assays and thermodynamic analysis of the effect of each analog on m(1)G37-tRNA synthesis reveal that TrmD and Trm5 discriminate functional groups differently. While both recognize N(1) and O(6) of G37, TrmD places a much stronger emphasis on these functional groups than Trm5. While the exocyclic 2-amino group of G37 is important for TrmD, it is dispensable for Trm5. In addition, while an adjacent G36 is obligatory for TrmD, it is nonessential for Trm5. These results depict a more rigid requirement of guanosine functional groups for TrmD than for Trm5. However, the sensitivity of both enzymes to analog substitutions, together with an experimental revelation of their low cellular concentrations relative to tRNA substrates, suggests a model in which these enzymes rapidly screen tRNA by direct recognition of G37 in order to monitor the global state of m(1)G37-tRNA.     

Distinct determinants of tRNA recognition by the TrmD and Trm5 methyl transferases

TrmD and Trm5 are, respectively, the bacterial and eukarya/archaea methyl transferases that catalyze transfer of the methyl group from S-adenosyl methionine (AdoMet) to the N1 position of G37 in tRNA to synthesize m1G37-tRNA. The m1G37 modification prevents tRNA frameshifts on the ribosome by assuring correct codon-anticodon pairings, and thus is essential for the fidelity of protein synthesis. Although TrmD and Trm5 are derived from unrelated AdoMet families and recognize the cofactor using distinct motifs, the question of whether they select G37 on tRNA by the same, or different, mechanism has not been answered. Here we address this question by kinetic analysis of tRNA truncation mutants that lack domains typically present in the canonical L shaped structure, and by evaluation of the site of modification on tRNA variants with an expanded or contracted anticodon loop. With both experimental approaches, we show that TrmD and Trm5 exhibit separate and distinct mode of tRNA recognition, suggesting that they evolved by independent and non-overlapping pathways from their unrelated AdoMet families. Our results also shed new light onto the significance of the m1G37 modification in the controversial quadruplet-pairing model of tRNA frameshift suppressors.     

Distinct origins of tRNA(m1G37) methyltransferase

The enzyme tRNA(m1G37) methyltransferase catalyzes the transfer of a methyl group from S-adenosyl-l-methionine (AdoMet) to the N1 position of G37 in the anticodon loop of a subset of tRNA. The modified guanosine is 3' to the anticodon and is important for maintenance of reading frame during decoding of genetic information. While the methyltransferase is well conserved in bacteria and is easily identified (encoded by the trmD gene), the identity of the enzyme in eukarya and archaea is less clear. Here, we report that the enzyme encoded by Mj0883 of Methanocaldococcus jannaschii is the archaeal counterpart of the bacterial TrmD. However, despite catalyzing the same reaction and displaying similar enzymatic properties, MJ0883 and bacterial TrmD are completely unrelated in sequence. The catalytic domain of MJ0883, when aligned with the five known structural folds (I-V) that have been described to bind AdoMet, is of the class I fold, similar to the ancient Rossmann fold that binds nucleotides. In contrast, the catalytic domain of the bacterial TrmD has the unusual class IV fold of a trefoil knot structure. Thus, both the sequence and structural arrangements of tRNA(m1G37) methyltransferase have distinct evolutionary origins among primary kingdoms, revealing an unexpected but remarkable non-orthologous gene displacement to achieve an important tRNA modification.     

Differentiating analogous tRNA methyltransferases by fragments of the methyl donor

Bacterial TrmD and eukaryotic-archaeal Trm5 form a pair of analogous tRNA methyltransferase that catalyze methyl transfer from S-adenosyl methionine (AdoMet) to N(1) of G37, using catalytic motifs that share no sequence or structural homology. Here we show that natural and synthetic analogs of AdoMet are unable to distinguish TrmD from Trm5. Instead, fragments of AdoMet, adenosine and methionine, are selectively inhibitory of TrmD rather than Trm5. Detailed structural information of the two enzymes in complex with adenosine reveals how Trm5 escapes targeting by adopting an altered structure, whereas TrmD is trapped by targeting due to its rigid structure that stably accommodates the fragment. Free energy analysis exposes energetic disparities between the two enzymes in how they approach the binding of AdoMet versus fragments and provides insights into the design of inhibitors selective for TrmD.     

Control of Catalytic Cycle by a Pair of Analogous tRNA Modification Enzymes

Enzymes that use distinct active site structures to perform identical reactions are known as analogous enzymes. The isolation of analogous enzymes suggests the existence of multiple enzyme structural pathways that can catalyze the same chemical reaction. A fundamental question concerning analogous enzymes is whether their distinct active-site structures would confer the same or different kinetic constraints to the chemical reaction, particularly with respect to the control of enzyme turnover. Here, we address this question with the analogous enzymes of bacterial TrmD and its eukaryotic and archaeal counterpart Trm5. TrmD and Trm5 catalyze methyl transfer to synthesize the m1G37 base at the 3′ position adjacent to the tRNA anticodon, using S-adenosyl methionine (AdoMet) as the methyl donor. TrmD features a trefoil-knot active-site structure whereas Trm5 features the Rossmann fold. Pre-steady-state analysis revealed that product synthesis by TrmD proceeds linearly with time, whereas that by Trm5 exhibits a rapid burst followed by a slower and linear increase with time. The burst kinetics of Trm5 suggests that product release is the rate-limiting step of the catalytic cycle, consistent with the observation of higher enzyme affinity to the products of tRNA and AdoMet. In contrast, the lack of burst kinetics of TrmD suggests that its turnover is controlled by a step required for product synthesis. Although TrmD exists as a homodimer, it showed half-of-the-sites reactivity for tRNA binding and product synthesis. The kinetic differences between TrmD and Trm5 are parallel with those between the two classes of aminoacyl-tRNA synthetases, which use distinct active site structures to catalyze tRNA aminoacylation. This parallel suggests that the findings have a fundamental importance for enzymes that catalyze both methyl and aminoacyl transfer to tRNA in the decoding process.     

Crystal structures of the tRNA:m2G6 methyltransferase Trm14/TrmN from two domains of life

Methyltransferases (MTases) form a major class of tRNA-modifying enzymes needed for the proper functioning of tRNA. Recently, RNA MTases from the TrmN/Trm14 family that are present in Archaea, Bacteria and Eukaryota have been shown to specifically modify tRNA(Phe) at guanosine 6 in the tRNA acceptor stem. Here, we report the first X-ray crystal structures of the tRNA m(2)G6 (N(2)-methylguanosine) MTase (TTC)TrmN from Thermus thermophilus and its ortholog (Pf)Trm14 from Pyrococcus furiosus. Structures of (Pf)Trm14 were solved in complex with the methyl donor S-adenosyl-l-methionine (SAM or AdoMet), as well as the reaction product S-adenosyl-homocysteine (SAH or AdoHcy) and the inhibitor sinefungin. (TTC)TrmN and (Pf)Trm14 consist of an N-terminal THUMP domain fused to a catalytic Rossmann-fold MTase (RFM) domain. These results represent the first crystallographic structure analysis of proteins containing both THUMP and RFM domain, and hence provide further insight in the contribution of the THUMP domain in tRNA recognition and catalysis. Electrostatics and conservation calculations suggest a main tRNA binding surface in a groove between the THUMP domain and the MTase domain. This is further supported by a docking model of TrmN in complex with tRNA(Phe) of T. thermophilus and via site-directed mutagenesis.     

Exploring GpG bases next to anticodon in tRNA subsets

Transfer RNA (tRNA) structure, modifications and functions are evolutionary and established in bacteria, archaea and eukaryotes. Typically the tRNA modifications are indispensable for its stability and are required for decoding the mRNA into amino acids for protein synthesis. A conserved methylation has been located on the anticodon loop specifically at the 37^th position and it is next to the anticodon bases. This modification is called as m1G37 and it is catalyzed by tRNA (m¹G37) methyltransferase (TrmD). It is deciphered that G37 positions occur on few additional amino acids specific tRNA subsets in bacteria. Furthermore, Archaea and Eukaryotes have more number of tRNA subsets which contains G37 position next to the anticodon and the G residue are located at different positions such as G36, G37, G38, 39, and G40. In eight bacterial species, G (guanosine) residues are presents at the 37^th and 38^th position except three tRNA subsets having G residues at 36^th and 39^th positions. Therefore we propose that m1G37 modification may be feasible at 36^th, 37^th, 38^th, 39^th and 40^th positions next to the anticodon of tRNAs. Collectively, methylation at G residues close to the anticodon may be possible at different positions and without restriction of anticodon 3^rd base A, C, U or G.     

Structural alterations of the tRNA(m1G37)methyltransferase from Salmonella typhimurium affect tRNA substrate specificity

In Salmonella typhimurium, the tRNA(m1G37)methyltransferase (the product of the trmD gene) catalyzes the formation of m1G37, which is present adjacent and 3' of the anticodon (position 37) in seven tRNA species, two of which are tRNA(Pro)CGG and tRN(Pro)GGG. These two tRNA species also exist as +1 frameshift suppressor sufA6 and sufB2, respectively, both having an extra G in the anticodon loop next to and 3' of m1G37. The wild-type form of the tRNA(m1G37)methyltransferase efficiently methylates these mutant tRNAs. We have characterized one class of mutant forms of the tRNA(m1G37)methyltransferase that does not methylate the sufA6 tRNA and thereby induce extensive frameshifting resulting in a nonviable cell. Accordingly, pseudorevertants of strains containing such a mutated trmD allele in conjunction with the sufA6 allele had reduced frameshifting activity caused by either a 9-nt duplication in the sufA6tRNA or a deletion of its structural gene, or by an increased level of m1G37 in the sufA6tRNA. However, the sufB2 tRNA as well as the wild-type counterparts of these two tRNAs are efficiently methylated by this class of structural altered tRNA(m1G37)methyltransferase. Two other mutations (trmD3, trmD10) were found to reduce the methylation of all potential tRNA substrates and therefore primarily affect the catalytic activity of the enzyme. We conclude that all mutations except two (trmD3 and trmD10) do not primarily affect the catalytic activity, but rather the substrate specificity of the tRNA, because, unlike the wild-type form of the enzyme, they recognize and methylate the wild-type but not an altered form of a tRNA. Moreover, we show that the TrmD peptide is present in catalytic excess in the cell.     

Catalysis by the second class of tRNA(m1G37) methyl transferase requires a conserved proline

The enzyme tRNA(m1G37) methyl transferase catalyzes the transfer of a methyl group from S-adenosyl methionine (AdoMet) to the N1 position of G37, which is 3' to the anticodon sequence and whose modification is important for maintaining the reading frame fidelity. While the enzyme in bacteria is highly conserved and is encoded by the trmD gene, recent studies show that the counterpart of this enzyme in archaea and eukarya, encoded by the trm5 gene, is unrelated to trmD both in sequence and in structure. To further test this prediction, we seek to identify residues in the second class of tRNA(m1G37) methyl transferase that are required for catalysis. Such residues should provide mechanistic insights into the distinct structural origins of the two classes. Using the Trm5 enzyme of the archaeon Methanocaldococcus jannaschii (previously MJ0883) as an example, we have created mutants to test many conserved residues for their catalytic potential and substrate-binding capabilities with respect to both AdoMet and tRNA. We identified that the proline at position 267 (P267) is a critical residue for catalysis, because substitution of this residue severely decreases the kcat of the methylation reaction in steady-state kinetic analysis, and the k(chem) in single turnover kinetic analysis. However, substitution of P267 has milder effect on the Km and little effect on the Kd of either substrate. Because P267 has no functional side chain that can directly participate in the chemistry of methyl transfer, we suggest that its role in catalysis is to stabilize conformations of enzyme and substrates for proper alignment of reactive groups at the enzyme active site. Sequence analysis shows that P267 is embedded in a peptide motif that is conserved among the Trm5 family, but absent from the TrmD family, supporting the notion that the two families are descendants of unrelated protein structures.     

S-adenosylhomocysteine hydrolases as the primary target enzymes in androgen regulation of methylation complexes

tRNA methylation complexes consisting of S-adenosylmethionine (AdoMet) synthetase, tRNA methylases, and S-adenosylhomocysteine (AdoHcy) hydrolase have been prepared from rat Novikoff hepatoma cells. The existence of the ternary enzyme complex is supported by dissociation and reconstitution of the ternany tRNA methylation complexes. In rat prostate and testis, two isozymes each for AdoMet synthetase and AdoHcy hydrolase are detected. The K_m (methionine) values for the two AdoMet synthetases are 3.1 and 23.7 μm and the K_m (adenosine) values for the two AdoHcy hydrolases are 0.33 and 1.8 μm. Correspondingly, two groups of methylation complexes are detectable, sedimenting in a sucrose gradient as 7 S and 8 S. The 7 S complexes are composed of AdoMet synthetase and AdoHcy hydrolase with the higher K_m values, and the 8 S complexes are composed of the respective isozymes with the lower K_m values. tRNA methylation complexes belong to the 8 S group. In hormone-depleted rat prostates and testes following hypophysectomy, the specific activities of AdoMet synthetases, tRNA methylases, and AdoHcy hydrolases are decreased severely, but are restored promptly after administration of testosterone. Thus, methylation enzymes are responsive to the regulation by steroid hormone. AdoHcy hydrolases from hormone-depleted tissues are unstable, and ternary tRNA methylation complexes are easily dissociable into individual activities. The stability of AdoHcy hydrolases is markedly improved by testosterone, and the integrity of ternary tRNA methylation complexes is maintained in the presence of testosterone. These results suggest that AdoHcy hydrolases are the primary target enzymes in adrogen regulation of methylation complexes.     

Crystal structure of archaeal tRNA(m(1)G37)methyltransferase aTrm5

Methylation of the N1 atom of guanosine at position 37 in tRNA, the position 3'-adjacent to the anticodon, generates the modified nucleoside m(1)G37. In archaea and eukaryotes, m(1)G37 synthesis is catalyzed by tRNA(m(1)G37)methyltransferase (archaeal or eukaryotic Trm5, a/eTrm5). Here we report the crystal structure of archaeal Trm5 (aTrm5) from Methanocaldococcus jannaschii (formerly known as Methanococcus jannaschii) in complex with the methyl donor analogue at 2.2 A resolution. The crystal structure revealed that the entire protein is composed of three structural domains, D1, D2, and D3. In the a/eTrm5 primary structures, D2 and D3 are highly conserved, while D1 is not conserved. The D3 structure is the Rossmann fold, which is the hallmark of the canonical class-I methyltransferases. The a/eTrm5-defining domain, D2, exhibits structural similarity to some class-I methyltransferases. In contrast, a DALI search with the D1 structure yielded no structural homologues. In the crystal structure, D3 contacts both D1 and D2. The residues involved in the D1:D3 interactions are not conserved, while those participating in the D2:D3 interactions are well conserved. D1 and D2 do not contact each other, and the linker between them is disordered. aTrm5 fragments corresponding to the D1 and D2-D3 regions were prepared in a soluble form. The NMR analysis of the D1 fragment revealed that D1 is well folded by itself, and it did not interact with either the D2-D3 fragment or the tRNA. The NMR analysis of the D2-D3 fragment revealed that it is well folded, independently of D1, and that it interacts with tRNA. Furthermore, the D2-D3 fragment was as active as the full-length enzyme for tRNA methylation. The positive charges on the surface of D2-D3 may be involved in tRNA binding. Therefore, these findings suggest that the interaction between D1 and D3 is not persistent, and that the D2-D3 region plays the major role in tRNA methylation.     

Mechanisms for auto-inhibition and forced product release in glycine N-methyltransferase: crystal structures of wild-type, mutant R175K and S-adenosylhomocysteine-bound R175K enzymes

Glycine N-methyltransferase (S-adenosyl-l-methionine: glycine methyltransferase, EC 2.1.1.20; GNMT) catalyzes the AdoMet-dependent methylation of glycine to form sarcosine (N-methylglycine). Unlike most methyltransferases, GNMT is a tetrameric protein showing a positive cooperativity in AdoMet binding and weak inhibition by S-adenosylhomocysteine (AdoHcy). The first crystal structure of GNMT complexed with AdoMet showed a unique "closed" molecular basket structure, in which the N-terminal section penetrates and corks the entrance of the adjacent subunit. Thus, the apparent entrance or exit of the active site is not recognizable in the subunit structure, suggesting that the enzyme must possess a second, enzymatically active, "open" structural conformation. A new crystalline form of the R175K enzyme has been grown in the presence of an excess of AdoHcy, and its crystal structure has been determined at 3.0 A resolution. In this structure, the N-terminal domain (40 amino acid residues) of each subunit has moved out of the active site of the adjacent subunit, and the entrances of the active sites are now opened widely. An AdoHcy molecule has entered the site occupied in the "closed" structure by Glu15 and Gly16 of the N-terminal domain of the adjacent subunit. An AdoHcy binds to the consensus AdoMet binding site observed in the other methyltransferase. This AdoHcy binding site supports the glycine binding site (Arg175) deduced from a chemical modification study and site-directed mutagenesis (R175K). The crystal structures of WT and R175K enzymes were also determined at 2.5 A resolution. These enzyme structures have a closed molecular basket structure and are isomorphous to the previously determined AdoMet-GNMT structure. By comparing the open structure to the closed structure, mechanisms for auto-inhibition and for the forced release of the product AdoHcy have been revealed in the GNMT structure. The N-terminal section of the adjacent subunit occupies the AdoMet binding site and thus inhibits the methyltransfer reaction, whereas the same N-terminal section forces the departure of the potentially potent inhibitor AdoHcy from the active site and thus facilitates the methyltransfer reaction. Consequently GNMT is less active at a low level of AdoMet concentration, and is only weakly inhibited by AdoHcy. These properties of GNMT are particularly suited for regulation of the cellular AdoMet/AdoHcy ratio.     

Mechanism of molecular interactions for tRNA(Val) recognition by valyl-tRNA synthetase

The molecular interactions between valyl-tRNA synthetase (ValRS) and tRNA(Val), with the C34-A35-C36 anticodon, from Thermus thermophilus were studied by crystallographic analysis and structure-based mutagenesis. In the ValRS-bound structure of tRNA(Val), the successive A35-C36 residues (the major identity elements) of tRNA(Val) are base-stacked upon each other, and fit into a pocket on the alpha-helix bundle domain of ValRS. Hydrogen bonds are formed between ValRS and A35-C36 of tRNA(Val) in a base-specific manner. The C-terminal coiled-coil domain of ValRS interacts electrostatically with A20 and hydrophobically with the G19*C56 tertiary base pair. The loss of these interactions by the deletion of the coiled-coil domain of ValRS increased the K(M) value for tRNA(Val) 28-fold and decreased the k(cat) value 19-fold in the aminoacylation. The tRNA(Val) K(M) and k(cat) values were increased 21-fold and decreased 32-fold, respectively, by the disruption of the G18*U55 and G19*C56 tertiary base pairs, which associate the D- and T-loops for the formation of the L-shaped tRNA structure. Therefore, the coiled-coil domain of ValRS is likely to stabilize the L-shaped tRNA structure during the aminoacylation reaction.     

Class I tyrosyl-tRNA synthetase has a class II mode of cognate tRNA recognition

Bacterial tyrosyl-tRNA synthetases (TyrRS) possess a flexibly linked C-terminal domain of approximately 80 residues, which has hitherto been disordered in crystal structures of the enzyme. We have determined the structure of Thermus thermophilus TyrRS at 2.0 A resolution in a crystal form in which the C-terminal domain is ordered, and confirm that the fold is similar to part of the C-terminal domain of ribosomal protein S4. We have also determined the structure at 2.9 A resolution of the complex of T.thermophilus TyrRS with cognate tRNA(tyr)(G Psi A). In this structure, the C-terminal domain binds between the characteristic long variable arm of the tRNA and the anti-codon stem, thus recognizing the unique shape of the tRNA. The anticodon bases have a novel conformation with A-36 stacked on G-34, and both G-34 and Psi-35 are base-specifically recognized. The tRNA binds across the two subunits of the dimeric enzyme and, remarkably, the mode of recognition of the class I TyrRS for its cognate tRNA resembles that of a class II synthetase in being from the major groove side of the acceptor stem.     

Mechanism of N-methylation by the tRNA m1G37 methyltransferase Trm5

Trm5 is a eukaryal and archaeal tRNA methyltransferase that catalyzes methyl transfer from S-adenosylmethionine (AdoMet) to the N(1) position of G37 directly 3' to the anticodon. While the biological role of m(1)G37 in enhancing translational fidelity is well established, the catalytic mechanism of Trm5 has remained obscure. To address the mechanism of Trm5 and more broadly the mechanism of N-methylation to nucleobases, we examined the pH-activity profile of an archaeal Trm5 enzyme, and performed structure-guided mutational analysis. The data reveal a marked dependence of enzyme-catalyzed methyl transfer on hydrogen ion equilibria: the single-turnover rate constant for methylation increases by one order of magnitude from pH 6.0 to reach a plateau at pH 7.0. This suggests a mechanism involving proton transfer from G37 as the key element in catalysis. Consideration of the kinetic data in light of the Trm5-tRNA-AdoMet ternary cocrystal structure, determined in a precatalytic conformation, suggests that proton transfer is associated with an induced fit rearrangement of the complex that precedes formation of the reactive configuration in the active site. Key roles for the conserved R145 side chain in stabilizing a proposed oxyanion at G37-O(6), and for E185 as a general base to accept the proton from G37-N(1), are suggested based on the mutational analysis.     

Evolutionary repair reveals an unexpected role of the tRNA modification m1G37 in aminoacylation

The tRNA modification m¹G37, introduced by the tRNA methyltransferase TrmD, is thought to be essential for growth in bacteria because it suppresses translational frameshift errors at proline codons. However, because bacteria can tolerate high levels of mistranslation, it is unclear why loss of m¹G37 is not tolerated. Here, we addressed this question through experimental evolution of trmD mutant strains of Escherichia coli. Surprisingly, trmD mutant strains were viable even if the m¹G37 modification was completely abolished, and showed rapid recovery of growth rate, mainly via duplication or mutation of the proline-tRNA ligase gene proS. Growth assays and in vitro aminoacylation assays showed that G37-unmodified tRNA^Pro is aminoacylated less efficiently than m¹G37-modified tRNA^Pro, and that growth of trmD mutant strains can be largely restored by single mutations in proS that restore aminoacylation of G37-unmodified tRNA^Pro. These results show that inefficient aminoacylation of tRNA^Pro is the main reason for growth defects observed in trmD mutant strains and that proS may act as a gatekeeper of translational accuracy, preventing the use of error-prone unmodified tRNA^Pro in translation. Our work shows the utility of experimental evolution for uncovering the hidden functions of essential genes and has implications for the development of antibiotics targeting TrmD.