Modulation of Aminoacylation and Editing Properties of Leucyl-tRNA Synthetase by a Conserved Structural Module*

Background: A structural module following the KMSKS catalytic loop is conserved in most class I synthetases. Results: This module contributes to aminoacylation and editing of leucyl-tRNA synthetases (LeuRS). Conclusion: This module affects the activities of LeuRS in both a structure- and sequence-dependent manner. Significance: This work further extends the function of stem-contact fold in LeuRS. A conserved structural module following the KMSKS catalytic loop exhibits α-α-β-α topology in class Ia and Ib aminoacyl-tRNA synthetases. However, the function of this domain has received little attention. Here, we describe the effect this module has on the aminoacylation and editing capacities of leucyl-tRNA synthetases (LeuRSs) by characterizing the key residues from various species. Mutation of highly conserved basic residues on the third α-helix of this domain impairs the affinity of LeuRS for the anticodon stem of tRNALeu, which decreases both aminoacylation and editing activities. Two glycine residues on this α-helix contribute to flexibility, leucine activation, and editing of LeuRS from Escherichia coli (EcLeuRS). Acidic residues on the β-strand enhance the editing activity of EcLeuRS and sense the size of the tRNALeu D-loop. Incorporation of these residues stimulates the tRNA-dependent editing activity of the chimeric minimalist enzyme Mycoplasma mobile LeuRS fused to the connective polypeptide 1 editing domain and leucine-specific domain from EcLeuRS. Together, these results reveal the stem contact-fold to be a functional as well as a structural linker between the catalytic site and the tRNA binding domain. Sequence comparison of the EcLeuRS stem contact-fold domain with editing-deficient enzymes suggests that key residues of this module have evolved an adaptive strategy to follow the editing functions of LeuRS.

A conserved structural module following the KMSKS catalytic loop exhibits ␣-␣-␤-␣ topology in class Ia and Ib aminoacyl-tRNA synthetases. However, the function of this domain has received little attention. Here, we describe the effect this module has on the aminoacylation and editing capacities of leucyl-tRNA synthetases (LeuRSs) by characterizing the key residues from various species. Mutation of highly conserved basic residues on the third ␣-helix of this domain impairs the affinity of LeuRS for the anticodon stem of tRNA Leu , which decreases both aminoacylation and editing activities. Two glycine residues on this ␣-helix contribute to flexibility, leucine activation, and editing of LeuRS from Escherichia coli (EcLeuRS). Acidic residues on the ␤-strand enhance the editing activity of EcLeuRS and sense the size of the tRNA Leu D-loop. Incorporation of these residues stimulates the tRNA-dependent editing activity of the chimeric minimalist enzyme Mycoplasma mobile LeuRS fused to the connective polypeptide 1 editing domain and leucinespecific domain from EcLeuRS. Together, these results reveal the stem contact-fold to be a functional as well as a structural linker between the catalytic site and the tRNA binding domain. Sequence comparison of the EcLeuRS stem contact-fold domain with editing-deficient enzymes suggests that key residues of this module have evolved an adaptive strategy to follow the editing functions of LeuRS.
Aminoacyl-tRNA synthetases (aaRSs) 2 are a large and diverse family of enzymes that catalyze the attachment of amino acids to their cognate tRNAs in a two-step aminoacyla-tion reaction as follows: 1) amino acid activation by ATP hydrolysis to form an aminoacyl-adenylate intermediate, and 2) transfer of the aminoacyl moiety from the intermediate to the cognate tRNA isoacceptor to form aminoacyl-tRNA (aa-tRNA) (1)(2)(3). Based on sequence homology and the structures of the catalytic active sites, aaRSs are divided into two classes of 10 members each. Class I synthetases are further divided into three subclasses, a, b, and c, according to sequence homology (4 -6). Leucyl-tRNA synthetase (LeuRS) belongs to class I aaRSs that include a typical Rossmann dinucleotide-binding fold active site architecture with the signature sequence modules HIGH and KMSKS (6). According to evolutionary models, the primitive catalytic core is extended by the insertion and/or fusion of additional domains (also called modules) in LeuRSs (7), most of which have inserted a large connective polypeptide 1 (CP1) domain that is responsible for amino acid editing. To ensure translation accuracy, LeuRSs have evolved a mechanism to remove aminoacyl AMP (aa-AMP; pre-transfer editing) and aa-tRNA (post-transfer editing) (8). Although post-transfer editing is carried out by the CP1 domain in most LeuRSs, this domain has been naturally deleted in LeuRS from Mycoplasma mobile (MmLeuRS) (9), and MmLeuRS is therefore unable to maintain its catalytic fidelity in post-transfer editing functions. However, once the CP1 domain of LeuRS from Escherichia coli (EcLeuRS) is inserted into MmLeuRS, the engineered enzyme functions like a typical prokaryotic LeuRS with tRNA-dependent editing activity (10).
Additionally, LeuRSs recruit the tRNA anticodon binding domain to bind tRNA Leu isoacceptors (11,12). The refined crystal structure of the EcLeuRS and tRNA Leu complex revealed a tRNA anticodon binding domain composed of a cylindrical ␣-helical bundle domain of five helices that is common to all class Ia synthetases. EcLeuRS establishes nonspecific backbone contacts between tRNA Leu nucleotides 12, 13, 22-26, and 42 with amino acid residues 667-686 and 749 -760 of enzyme (PDB entry 4AQ7 and 4ARC, see Ref. 11). Interestingly, a highly conserved structural module following the KMSKS loop exhibits the characteristic ␣-␣-␤-␣ topology in most class Ia and Ib aaRSs (Fig. 1A). Both the KMSKS loop and this structural mod-ule constitute the stem contact (SC)-fold (13)(14)(15)(16)(17)(18). The last ␣-helix of the SC-fold is almost coaxial to the first ␣-helix of the helix bundle domain but is separated by a short linker segment in most class I synthetases (13). However, the SC-fold of LeuRS lacks this segment, and its last ␣-helix instead overlaps with the first ␣-helix of the helix bundle domain (Fig. 1A). A previous study suggested that the effect of this class Ia/b SC-fold domain on tRNA recognition is structure-dependent rather than sequence-dependent, because sequence is much less conserved than structure in this case (17). Furthermore, the SC-fold is likely to dock against the inner side of the L-shaped tRNA, thereby positioning the anticodon stem (13).
LeuRSs also employ other appended domains, such as the C-terminal domain and leucine-specific domain (LSD), to facilitate tRNA Leu binding during the tRNA Leu -charging reaction (7,20,21). Meanwhile, elements embedded in the sequence of each tRNA promote specific interactions with its cognate aaRS (22); for example, the so-called discriminator nucleotide A73 in tRNA Leu is critical for both the aminoacylation and editing activities of LeuRS (23)(24)(25). The tertiary interactions between the D-and T-loops that form the elbow region of L-shaped tRNA Leu are critical elements for leucine identification (26), and the nucleotide G42 in the anticodon arms of tRNA Leu from Saccharomyces cerevisiae is crucial during the editing process (27,28). All these results were consistent with the seminal studies performed on isoleucyl-tRNA synthetase showing that the editing activity was influenced by the D-loop size of tRNA Ile (29).
Here, we systematically investigate the key residues of the conserved SC-fold structural module in prokaryotic and eukaryotic LeuRSs (Fig. 1B). Crucial basic and Gly residues in the third ␣-helix (␣3) of the SC-fold were found to modulate the aminoacylation and editing activities of LeuRS. Additionally, several key residues of helix ␣3 interact with the tRNA Leu anticodon stem, whereas acidic residues of the ␤-strand control the tRNA-dependent editing activity by sensing the size of the tRNA Leu D-loop. These findings were further confirmed by experiments performed on the chimeric MmLeuRS-CP1/LSD enzyme, a minimalist LeuRS enzyme fused with the E. coli CP1 domain and LSD (21). The role of the ␤-strand in enhancing the editing activity of the chimera was further confirmed by substituting the crucial residues for the E. coli enzyme. Together, these results support an influential role for the SC-fold module in both aminoacylation and editing functions of LeuRSs.

EXPERIMENTAL PROCEDURES
Expression and Purification of LeuRSs and Their Mutants from Different Species-LeuRSs from human cytoplasm (hcLeuRS), Aquifex aeolicus (AaLeuRS), Pyrococcus horikoshii (PhLeuRS), E. coli (EcLeuRS), and M. mobile (MmLeuRS) along with their mutants were obtained and purified as reported previously by our laboratory (28,30,31). Final concentrations were determined using Bradford protein assays according to the manufacturer's protocol (Bio-Rad). Mutations were engineered using the KOD Plus mutagenesis kit and confirmed by DNA sequencing (BioSun Bioscience).
ATP-Pyrophosphate (PPi) Exchange, tRNA Charging, and Deacylation-The ATP-PPi exchange of EcLeuRS (37°C) and its mutants was performed as described previously (10,34). AMP Formation-Because the net effect of the editing reaction is the consumption of ATP, editing can be measured by monitoring AMP formation in the presence of a noncognate amino acid. AMP formation rates for MmLeuRS-CP1/LSD, EcLeuRS, hcLeuRS, AaLeuRS, and their mutants were measured as described previously (20,26,28). MmtRNA CAA Leu and its mutants were used to study the editing ability of the chimeric MmLeuRS-CP1/LSD. The reaction mixture contained 100 mM Tris-HCl (pH 7.  Filters were then washed twice with 0.3 ml of the washing buffer and air-dried before radioactive quantification. All data were analyzed using GraphPad PRISM software.

Crucial Basic Residues of the Connecting Structural Module
Are Essential for the Aminoacylation Activity of LeuRSs-The refined structure of EcLeuRS revealed that the third ␣-helix (Ser 661 to Lys 683 ; helix ␣3; PDB entry 4ARC and 4AQ7; Fig. 1, B and C) is the only region of the SC-fold that interacts directly with tRNA in both aminoacylation and editing conformations (11,12), indicating that this ␣-helix may be critical for both processes. To investigate the effect of the ␣-helix on enzymatic activity, we constructed a deletion mutant lacking the third ␣-helix. Despite the 22-amino acid deletion, CD spectroscopy data also indicated the deletion mutant was well expressed in a soluble form and had a definite stable secondary structure ( Fig.  2A). However, this construct completely lacked both aminoacylation and editing activities, supporting the hypothesis that this helix is critical for these functions ( Fig. 2B and Tables 1 and 2).
Considering the interaction between tRNA and aaRS is predominantly mediated by crucial basic residues (37), in EcLeuRS, we investigated the function of residues Arg 668 , Lys 671 , and Arg 672 in the third ␣-helix of the SC-fold by site-directed mutagenesis (PDB codes 4ARC and 4AQ7; Fig. 1, B and C). EcLeuRS-K671A did not affect the catalytic efficiency (k cat /K m ) at all (Table 1). However, the k cat /K m values of EcLeuRS-R668A and EcLeuRS-R668E had decreased to 23 and 5% of the k cat /K m values of wild-type (WT) EcLeuRS (Table 1). The k cat /K m values of the double mutant EcLeuRS-R668A/R672A was only 6.4% of that of the native enzyme, and the catalytic efficiency of the EcLeuRS-R668E/R672E double-site mutant was severely diminished, with only 1.4% of tRNA aminoacylation activity remaining (Table 1). Although mutations at Arg 668 or Arg 672 decreased the aminoacylation activity to different extents, all of these mutants, including the almost inactive EcLeuRS-R668E/ R672E, exhibited intact Leu activation activity comparable with the WT enzyme (data not shown), suggesting these Arg residues are not involved in the amino acid activation step but rather the second tRNA transfer step. These results also suggest that the enzyme mutations alter interactions with the tRNA Leu that are critical for the stabilization of the transition state of the aminoacylation reaction but not for the affinity of enzyme for tRNA.
We next investigated the two crucial basic residues of helix ␣3 of the SC-fold of LeuRS from A. aeolicus (AaLeuRS) and M. mobile (MmLeuRS). AaLeuRS forms a unique heterodimer with its SC-fold located in the ␤-subunit. The counterparts of EcLeuRS-Arg 668 and EcLeuRS-Arg 672 in AaLeuRS are ␤Arg 94 and ␤Arg 98 , respectively, and mutating these residues in AaLeuRS had a similar effect on aminoacylation; AaLeuRS-␤R98A did not alter the catalytic efficiency, whereas mutating ␤Arg 94 to either Ala or Glu decreased k cat /K m values to 34 and 22% of that of WT AaLeuRS (Table 1). When the Ala mutants were combined in ␤R94A/R98A, the aminoacylation activity dropped to only 17% of that of WT AaLeuRS, and the double Glu mutant ␤R94E/R98E exhibited an even lower k cat /K m value (only 4.4% of that of WT AaLeuRS; Table 1).
In MmLeuRS, Lys 452 and Arg 456 correspond to the Arg 668 and Arg 672 residues of EcLeuRS, respectively. Substitution of Lys 452 with either Ala or Glu had only a minimal effect on aminoacylation activity. However the tRNA catalytic efficiency of these mutants was significantly decreased to 25% for MmLeuRS-R456A and to 21% for MmLeuRS-R456E of that of the WT MmLeuRS (Table 1). Notably, the K m values for the above MmLeuRS mutants were not significantly altered. These results suggest that helix ␣3 in the SC-fold of MmLeuRS interacts with tRNA in a different manner to that of EcLeuRS and AaLeuRS.
The crucial basic residues of helix ␣3 in the archaeal P. horikoshii LeuRS (PhLeuRS) (PDB entry 1WZ2) and eukaryotic human cytoplasmic LeuRS (hcLeuRS) enzymes were investigated. PhLeuRS contains five positively charged residues (Lys 692 , Lys 696 , Arg 698 , Lys 699 , and Arg 703 ) in helix ␣3, and alanine-scanning mutagenesis revealed that the K m value for tRNA of PhLeuRS-K696A was 0.65 M, an ϳ3-fold increase compared with the WT value (0.20 M). The k cat of PhLeuRS-R703A (0.0055 s Ϫ1 ) was much lower than that of native PhLeuRS (0.020 s Ϫ1 ; Table 1). The other PhLeuRS mutants had no effect on tRNA charging activity.
Arg 766 is the only basic residue in helix ␣3 of hcLeuRS SCfold (Fig. 1C). The mutant hcLeuRS-R766A decreased the k cat /K m value to less than 10% that of the native hcLeuRS (Table  1). These results demonstrate that basic residues on helix ␣3 of the SC-fold are critical for catalytic efficiency, which may explain their high sequence conservation in both prokaryotic and eukaryotic LeuRSs.
Crucial Basic Residues Contribute to the Fidelity of LeuRSs-TLC assays were performed to check whether the mutations of the basic residues affected the editing activity of LeuRS against Nva. AMP formation rates (k obs ) of EcLeuRS-R668A/R672A and -R668E/R672E were ϳ11 and 9.1% that of the WT EcLeuRS in the presence of EctRNA GAG Leu and Nva ( Table 2). The k obs values of the single-site mutants EcLeuRS-R668A and EcLeuRS-R672A were 67 and 59% that of wild-type EcLeuRS in the presence of EctRNA GAG Leu and Nva (Table 2 and Fig. 3). Remarkably, mutating each basic residue to the acidic Glu severely impacted the AMP formation of LeuRS. The k obs values of EcLeuRS-R672E and EcLeuRS-R668E were decreased significantly to 20 and 12% that of WT EcLeuRS for Ect-RNA GAG Leu in the presence of Nva (Table 2 and Fig. 3). These results suggested that the basic residues contribute to both the AMP formation and catalytic efficiency of EcLeuRS.
To identify which editing pathway was affected by these residues, we measured the k obs value of the mutants in the presence of Nva but in the absence of tRNA. The k obs value of the editing-defective EcLeuRS-R672A mutant was 1.6-fold that of the native enzyme in the absence of tRNA (Table 3). In contrast, k obs values without tRNA for the other mutants were similar to the WT values (Table 3). These results reveal that mutation of the basic residues largely impacts the tRNA-dependent transfer editing capacity of EcLeuRS. Further deacylation results indicated the impaired tRNA-dependent transfer editing of the mutants resulted from their decreased hydrolysis ability (Fig.  2B).
Finally, we probed the basic residues of AaLeuRS and hcLeuRS. Mutations of either Arg 94 or Arg 98 of AaLeuRS to Ala or Glu markedly decreased the editing activity of the enzyme, suggesting that both residues are essential for the editing process. The k obs values were ϳ0.20 s Ϫ1 in the presence of AatRNA GAG Leu and Nva, which was around 10% that of WT AaLeuRS, and the double mutant ␤R94E/R98E was even lower (0.13 s Ϫ1 ; Table 2). Consistently, mutation of the unique basic residue Arg 766 to Ala in hcLeuRS (hcLeuRS-R766A) led to a 90% decrease in AMP formation in the presence of hctRNA CAG Leu ( Table 2). Overall, our results show that both prokaryotic and eukaryotic LeuRSs employ crucial basic residues of helix ␣3 of the SC-fold to modulate the catalytic efficiency and fidelity of the enzyme.
Basic Residues Participate in the LeuRS Affinity for the Anticodon Stem of tRNA Leu -Filter binding assays revealed that the lower catalytic efficiencies (k cat /K m ) of the above mutants were due to a lower affinity for tRNA Leu . Indeed, the dissociation constants (k d ) of native EcLeuRS and hcLeuRS for their cognate tRNAs were 0.20 and 0.18 M, respectively, and this was increased more than 10-fold for EcLeuRS-R668A/R672A and 27-fold for EcLeuRS-R668E/R672E (Table 4). The k d of mutant hcLeuRS-R766A was 15-fold higher than hcLeuRS for hctRNA Leu ( Table 4). As the k d value reflects the affinity of LeuRS for tRNA Leu in the ground state before the aminoacylation catalysis, these results indicated that these residues were involved in the initial binding of enzyme with tRNA Leu . Moreover, the unchanged K m values of mutants for tRNA Leu suggested that the affinity of enzyme for tRNA Leu in the transition state of the aminoacylation reaction was unaffected.
As the refined structure of the EcLeuRS-EctRNA UAA Leu complex (PDB codes 4ARC and 4AQ7) displayed disorder in the anticodon loop, there was no clear evidence to show whether residues Arg 668 and Arg 672 from helix ␣3 of the SC-fold participated in interactions with the D-stem and anticodon stem of EctRNA UAA Leu (11). In an attempt to verify and identify the determinants of the editing reaction on the tRNA, we constructed a set of mutants of tRNA Leu . Because EcLeuRS recognizes the synthetic transcript of MmtRNA UAA Leu produced in vitro as well as the fully modified EctRNA GAG Leu purified from an E. coli overproducing strain (k cat /K m ϭ 2.6 and 2.2 M Ϫ1 s Ϫ1 , respectively), we chose to mutate the MmtRNA UAA Leu transcript, which is

Kinetic constants of LeuRSs and their mutants at basic residues from various species for their cognate tRNAs in aminoacylation reaction
Kinetic constants were determined using the tRNA charging assay described under "Experimental Procedures." All parameters are the average of three repeats with the standard deviations indicated. NM means not measurable (too low to be measured).  remarkably similar in sequence to EctRNA GAG Leu (Fig. 4). We constructed seven mutants with specific nucleotide substitutions in the stems of MmtRNA UAA Leu : -G24U and -C25U in the D-stem and -U39G, -C40U, -C41U, -C42U, and -C41U/C42U in the anticodon stem (Fig. 4). The catalytic efficiency (k cat /K m ) in the aminoacylation reaction and the k obs in the editing reaction were measured, and the values for WT MmtRNA UAA Leu and the -G24U, -C25U, and -U39G mutants were indistinguishable, indicating that these nucleotides do not critically interact with LeuRS (data not shown). However, the catalytic efficiencies (k cat /K m ) of EcLeuRS for the -C40U, -C41U, and -C42U variants were 73, 51, and 26% that of EcLeuRS for WT MmtRNA UAA Leu ( Table 5). Consistently, the k cat /K m value of the EcLeuRS for double mutant -C41U/C42U was even lower (12% that of WT MmtRNA UAA Leu ), with a value of 0.31 M Ϫ1 s Ϫ1 (Table 5). Further analysis revealed that the decreased catalytic efficiency may result from an altered affinity for the tRNA substrate. In the filter binding assay, the k d values of EcLeuRS for MmtRNA UAA Leu -C41U (1.7 M) and -C42U (1.6 M) mutants were ϳ5-fold greater than that of WT MmtRNA UAA Leu (0.27 M; Table 4). In addition, we showed that the double mutant MmtRNA UAA Leu -C41U/C42U was even less capable of editing. In the presence of 15 mM Nva, the k obs value of MmtRNA UAA Leu -C41U/C42U was 1.4 s Ϫ1 , compared with 2.6 s Ϫ1 for WT MmtRNA UAA Leu (Table 4). Together, these results suggest that LeuRS employs these basic residues of the SC-fold domain to modulate both aminoacylation and editing activities through interaction with the anticodon stem of tRNA Leu .

Enzyme
Two Gly Residues in Helix ␣3 Participate in the Aminoacylation and Editing Functions of LeuRS-Analysis of the SC-fold of EcLeuRS showed that both the ␤-strand and helix ␣3 undergo substantial conformational changes during aminoacylation and editing (PDB code 4AQ7 and 4ARC; Fig. 5A). Mutating the first two Gly residues of ␣3 to Ala (EcLeuRS-G662A/G665A) was accompanied by a substantial drop in aminoacylation activity (Table 6), whereas mutating to Pro did not impact the aminoacylation activity of EcLeuRS-G662P or EcLeuRS-G665P (Table 6). However, the catalytic efficiency of the double mutant EcLeuRS-G662P/G665P was decreased to 9.5% that of WT EcLeuRS with EctRNA GAG Leu , but exhibited similar secondary structure with WT EcLeuRS (Table 6 and Fig. 2A).
The mutations also diminished the Leu activation activity, as measured using the ATP-PPi exchange reaction (Fig. 4B). Compared with the k obs of WT EcLeuRS (57 s Ϫ1 ), the activity of the double mutant G662P/G665P (3.6 s Ϫ1 ) was impacted to a much greater extent than the double mutant G662A/G665A (16.5 s Ϫ1 ). These results suggest that the decreased aminoacylation activity may be the consequence of a reduction in the formation of adenylate. Additionally, these double mutants were also shown to be defective in their editing capacity. The k obs value of the double mutant G662P/G665P in the presence of Nva and EctRNA GAG Leu was only 9.4% that of the WT enzyme. Mutation G665P alone reduced the k obs to 12% that of the WT enzyme, whereas the G662P mutation had only a minor  (Table 2).

TABLE 3 Observed rate constants of LeuRS and its mutants at basic residues for AMP synthesis in the presence of Nva and absence of tRNA Leu
All rates are the average of three repeats with the standard deviations indicated.    Table 7. D, structure of MmLeuRS (gray; PDB code 3ZIU (39). Residues shown on helix ␣3 (cyan) are equivalent to Gly 662 and Gly 665 of EcLeuRS. Other helices of the SC-fold are represented in pink, and the ␤-strand is in red.  impact (Table 7 and Fig. 5C). Additionally, the post-transfer editing activity of mutation G665P was also decreased (Fig.  2B). These observations suggest that the editing-defective mutant EcLeuRS-G665P, which exhibited a standard synthetic efficiency, could be used to produce mischarged aa-tRNA Leu molecules.

LeuRS
To further explore the hypothesis that Gly residues might facilitate conformational changes of the SC-fold module during the transition between the aminoacylation and editing stages, we examined the sequences of PhLeuRS and hcLeuRS. Both enzymes contain a single Gly in helix ␣3, albeit at different positions (PDB code 1WZ2; Fig. 1C). Mutations to Pro in PhLeuRS-G695P decreased the catalytic efficiency of aminoacylation to 5.3% that of WT PhLeuRS. Similarly, a mutant replaced Gly 763 of hcLeuRS with Pro, and hcLeuRS-G763P had only half-catalytic efficiency of hcLeuRS (Table 6). Although not conserved, these data suggest that the Gly residues play a substantial role in retaining aminoacylation activity.
Sequence alignment revealed that MmLeuRS lacks any Gly residues in helix ␣3; the two Gly residues found in EcLeuRS are replaced by Thr and Ala (PDB code 3ZIU; Figs. 1C and 5D). Additionally, the ␤-strand of the SC-fold of MmLeuRS contains two Lys residues, whereas other LeuRS enzymes usually contain two or three acidic residues in this vicinity (Fig. 1C). This suggests that MmLeuRS employs a different strategy for interaction with its cognate tRNA, which may be linked to the absence of the CP1 domain and its associated editing function. Indeed, without the need to transfer the tRNA from the synthetic to the editing site, the role of the SC-fold may have diverged in these enzymes, and this may explain the altered charge distribution and the absence of flexibility-inducing Gly residues.
␤-Strand of the SC-fold Participates in the Editing Function of LeuRS-To study the role of the key residues of the SC-fold ␤-strand, we replaced residues Asp 653 , Glu 657 , and Leu 656 of EcLeuRS with Lys. In the aminoacylation reaction, the catalytic efficiency of mutant EctRNA GAG Leu -D653K and double mutant EctRNA GAG Leu -L656K/E657K was not significantly different from that of the native enzyme (Table 8). However, in the editing step, in the presence of Nva and tRNA, the k obs value of EctRNA GAG Leu -D653K dropped to 45% that of WT EcLeuRS for AMP formation, which was only half that of the native enzyme (Table 7). For the double mutant L656K/E657K, the k obs value was decreased to 9.1% that of WT EcLeuRS in the presence of tRNA, which suggested that the mutant had lost its tRNA-dependent editing activity (Fig. 6A and Table 7). Therefore, we performed an Ile-tRNA Leu deacylation assay in the presence of mutant D653K or L656K/E657K. The deacylation curves showed that both mutants were severely affected in their capacity to deacylate mischarged EctRNA GAG Leu , suggesting that the SC-fold ␤-strand is involved in tRNA-dependent editing of LeuRS following the transfer step (Fig. 6B).
In the crystal structure of the complex containing EcLeuRS and EctRNA UAA Leu (11), nucleotides G12 and G13 are located in the vicinity of the ␤-strand of the SC-fold. Although these residues are not within bonding distance of the acidic residues 653 and 657, we mutated them to U in MmtRNA UAA Leu to probe their role during the complete aminoacylation/editing cycle. However, both aminoacylation and editing activity properties were indistinguishable from the WT tRNA (Table 4).
We further explored the possibility that, due to its appropriate placement in the structure of the complex, the SC-fold may sense the size of the D-loop during the editing step. We chose to use a different MM tRNA Leu with a CAA anticodon that contains only 10 nucleotides in the D-loop instead of the usual 11 nucleotides in EctRNA UAA Leu and MmtRNA UAA Leu (Fig. 4). Remarkably, during the editing reaction, MmtRNA CAA Leu could not stimulate AMP formation as well as MmtRNA UAA Leu ; in the presence of Nva and EcLeuRS, the k obs was only 0.91 s Ϫ1 compared with 2.6 s Ϫ1 for both MmtRNA CAA Leu and EctRNA UAA Leu . This result strongly suggested that the size of the D-loop is a key determinant of editing activity. To confirm this hypothesis, we inserted in the D-loop after cytosine 17 of MmtRNA CAA Leu (17a), just before the crucial nucleotides G18 and G19. In the resulting mutant MmtRNA CAA Leu (ϩC17a), the loop was identical to that of MmtRNA UAA Leu and EctRNA UAA Leu (Fig. 4). Consistently, in the presence of Nva, the k obs value of EcLeuRS for mutant (ϩC17a) was increased to 2.1 s Ϫ1 , reaching a level comparable with MmtRNA UAA Leu and EctRNA UAA Leu (Tables 7 and 9). This result further indicates that the ␤-strand of the SC-fold may participate in editing by sensing the size of the tRNA Leu D-loop. The editing activity of EcLeuRS was magnified when the native size of the D-loop of tRNA Leu was introduced into MmtRNA CAA Leu . These results show that the ␤-strand adjacent to the SC-fold influences the editing activities of EcLeuRS, probably via the conserved acidic residues.

TABLE 6 Kinetic constants of LeuRSs from various species and their mutants at Gly residues for their cognate tRNAs in aminoacylation reaction
Kinetic constants were determined using the tRNA charging assay described under "Experimental Procedures." All parameters are the average of three repeats with the standard deviations indicated.   MAY 8, 2015 • VOLUME 290 • NUMBER 19

JOURNAL OF BIOLOGICAL CHEMISTRY 12263
Engineering of the Chimeric MmLeuRS-CP1/LSD and MmtRNA CAA Leu to Improve Fidelity-We previously engineered a variant of MmLeuRS containing an insertion of CP1 and LSD from EcLeuRS, and this chimera exhibited improved tRNA-dependent editing activity for isoacceptor MmtRNA UAA Leu (21). In this study, the k obs value of AMP formation during the editing reaction of the chimera was 0.037 and 0.041 s Ϫ1 in the absence or presence of the isoacceptor MmtRNA CAA Leu , respectively (Table 9 and Fig. 7A). This result suggests the tRNA-dependent editing pathway was inefficient in the presence of the MmtRNA CAA Leu isoacceptor. Therefore, MmLeuRS-CP1/LSD was able to discriminate between the AUG and CAA isoacceptors during the editing reaction. Considering the contribution of the SC-fold to the catalytic function and fidelity of LeuRS, we first introduced into MmLeuRS two Glu residues in place of the two unusual basic residues (Lys 440 and Lys 441 ) found in the ␤-strand. However, the double mutant MmLeuRS-CP1/LSD-K440E/K441E did not possess any aminoacylation activity ( Table 7), suggesting that both Lys residues contribute to the catalytic activity in MmLeuRS. We further substituted Lys 440 with Leu and Lys 441 with Glu to mimic the residues in EcLeuRS. Remarkably, the k obs value of MmLeuRS-CP1/LSD-K440L/ K441E for MmtRNA CAA Leu was twice that of the starting MmLeuRS-CP1/LSD chimera in the editing reaction (0.11 s Ϫ1 versus 0.041 s Ϫ1 in Table 9 and Fig. 7, A and B). In contrast, the k cat /K m values of the double mutant -K440L/K441E exhibited only 33% compared with the starting chimera MmLeuRS-CP1/ LSD (Table 7). We next measured the editing capacity of the double mutant K440L/K441E in the presence of MmtRNA CAA Leu with the (ϩC17a) insertion in the D-loop, and the k obs value (0.31 s Ϫ1 ) was almost 3-fold higher than that of MmtRNA CAA Leu (0.11 s Ϫ1 ), which confirmed the importance of the size of the D-loop (Table 9 and Fig. 7, C and D).
Taken together, these results show that the SC-fold ␤-strand influences both aminoacylation and editing activities of EcLeuRS via conserved acidic residues. Revealingly, the editing activity of MmLeuRS that ordinarily lacks these residues was enhanced when acidic residues were introduced into this enzyme.

Conserved SC-fold Is Crucial for Both Aminoacylation and
Editing of LeuRSs-Systematic comparison of class I synthetases identified a region connecting the Rossmann fold and the anticodon binding domain (7,(15)(16)(17)(18) named the SC-fold. This module is poorly conserved in sequence and is widely considered to function in a structure-dependent manner rather than in a sequence-dependent manner. However, this study showed that several crucial SC-fold residues contribute to both aminoacylation and editing functions of LeuRS and interact specifi-   cally with the tRNA. Although archaeal and eukaryotic cytoplasmic LeuRSs are architecturally distinct from bacterial LeuRSs (38), all share crucial elements, including one or two crucial basic residues on helix ␣3 of the SC-fold. Interestingly, many bacterial and archaeal LeuRSs contain additional basic residues, such as EcLeuRS that harbors an extra Lys residue at position 671 between the two crucial basic residues. However, mutation of this additional basic residue had no effect on the aminoacylation activity of the enzyme (Table 1). Moreover, of the five basic residues on helix ␣3 of PhLeuRS, only Arg 703 and Lys 696 were essential for catalytic activity. In contrast, hcLeuRS contains only one basic residue (Arg 766 ) that is involved in both synthetic and editing catalytic steps. These results suggest that despite the evident separation in the three kingdoms of life, evolutionary pressure has caused the SC-fold to retain these functional elements. Specific Gly residues form the second critical element that modulates the aminoacylation and editing properties of LeuRS, and mutation of these residues affected the Leu activation activity in EcLeuRS, suggesting the synthetic active site pocket may undergo conformational changes during catalysis. In this study we identified a mutant (EcLeuRS-G665P) with diminished editing properties reminiscent of a previously characterized CP1editing site mutant (EcLeuRS-Y330D) (34). Compared with the basic residues, mutants with impaired aminoacylation activity, both EcLeuRS-G665P and EcLeuRS-Y330D, retained tRNA aminoacylation activity, and these variants could therefore be used in the future to mischarged aa-tRNA Leu for deacylation studies.
The crystal structure of EcLeuRS in complex with tRNA Leu (PDB codes 4ARC and 4AQ7) revealed Gly residues that may contribute to conformational changes of the SC-fold ␤-strand containing several acidic residues (Glu 657 and Asp 653 ) involved in the fidelity of EcLeuRS (Fig. 4A). This structure also high-lighted the proximity of residue Met 654 and tRNA Leu , which is an even shorter distance than between Glu 657 , Asp 653 , and the tRNA (PDB code 4AQ7) (11). However, mutation of Met 654 to Ala or Lys did not affect either the synthetic or editing properties of LeuRS, suggesting Met 654 may not be functionally important or may act in a different way (data not shown).
Enzymes Lacking the SC-fold Crucial Residues Exhibit Impaired Editing Capacity-To date, only two LeuRSs lacking the editing activity have been described (9,10,38,39) as follows: MmLeuRS lacks the entire CP1 editing domain, and human mitochondrial LeuRS (hmLeuRS) has lost only the catalytic residues of the CP1 domain. As a consequence, both enzymes are deprived of post-transfer editing activity. Sequence alignment revealed that neither MmLeuRS nor hmLeuRS contain the Gly or acidic residues found in helix ␣3 of the SC-fold (Fig. 1C). Additionally, hmLeuRS utilizes a Thr instead of the first Gly of helix ␣3, and it has a Leu in place of the crucial Glu in the ␤-strand. The lack of editing activity in these enzymes further suggested that they may be functionally important and may be involved in mediating communication between the SC-fold and the CP1 editing domain.
Improving the Aminoacylation Fidelity by Modifying the Editing Determinants of Both Synthetase and tRNA-The discovery of residues of the SC-fold that influence the editing capacity of LeuRS and the interaction with the anticodon stem and D-loop of tRNA Leu provide valuable new information that may be of use in the engineering of both synthetase enzymes and tRNAs. Incorporation of two key EcLeuRS ␤-strand residues, including a conserved acidic residue, into the chimeric MmLeuRS-CP1-LSD enzyme significantly enhanced the editing activity, which opens up exciting possibilities for improving the fidelity of this and other artificial synthetases.
Similarly, we engineered the tRNA to improve the fidelity of LeuRS by enhancing the tRNA-dependent editing capacity. We observed that the isoacceptor MmtRNA CAA Leu was only weakly able to stimulate the editing activity of the chimeric LeuRS, compared with other isoacceptors. By inserting a single nucleotide in the D-loop of MmtRNA CAA Leu that is naturally smaller than the D-loop of other tRNA Leu molecules, the editing activity of the synthetic MmLeuRS enzyme was markedly enhanced. In concert with enzyme engineering, such tRNA modification approaches are paving the way for improving the fidelity of artificial synthetase enzymes. Although error-prone systems are presumably advantageous for proteome diversity in some living organisms (9,19), constructing artificial enzymes with high fidelity from naturally occurring enzymes lacking proofreading activity remains challenging. Remarkably, the work described here suggests that at least two nonexclusive pathways may exist to achieve this goal.
In summary, this study reveals the functional importance of the conserved SC-fold module in both the aminoacylation and editing activities of LeuRS enzymes. The SC-fold is therefore not only a connecting linker between the tRNA-binding module and the catalytic site. Furthermore, the SC-fold interacts with the inner side of the L-shaped tRNA, thereby positioning the anticodon stem. Given that the KMSKS motif is presumed to contribute to the catalytic site, the SC-fold carrying this module also forms a functional link between the catalytic activity and tRNA binding. This feature senses the size of the D-loop of tRNA Leu and controls the editing reaction. In addition, engineering of the chimeric MmLeuRS-CP1/LSD and tRNA provides evidence that the SC-fold module has co-evolved with the CP1 editing domain in LeuRS enzymes.