Translational Quality Control by Bacterial Threonyl-tRNA Synthetases*

Translational fidelity mediated by aminoacyl-tRNA synthetases ensures the generation of the correct aminoacyl-tRNAs, which is critical for most species. Threonyl-tRNA synthetase (ThrRS) contains multiple domains, including an N2 editing domain. Of the ThrRS domains, N1 is the last to be assigned a function. Here, we found that ThrRSs from Mycoplasma species exhibit differences in their domain composition and editing active sites compared with the canonical ThrRSs. The Mycoplasma mobile ThrRS, the first example of a ThrRS naturally lacking the N1 domain, displays efficient post-transfer editing activity. In contrast, the Mycoplasma capricolum ThrRS, which harbors an N1 domain and a degenerate N2 domain, is editing-defective. Only editing-capable ThrRSs were able to support the growth of a yeast thrS deletion strain (ScΔthrS), thus suggesting that ScΔthrS is an excellent tool for studying the in vivo editing of introduced bacterial ThrRSs. On the basis of the presence or absence of an N1 domain, we further revealed the crucial importance of the only absolutely conserved residue within the N1 domain in regulating editing by mediating an N1-N2 domain interaction in Escherichia coli ThrRS. Our results reveal the translational quality control of various ThrRSs and the role of the N1 domain in translational fidelity.

Aminoacyl-tRNA synthetase (aaRS) 3 catalyzes tRNA aminoacylation and generates aminoacyl-tRNA for protein biosynthesis (1)(2)(3). This process requires amino acid activation by condensation with ATP, forming the aminoacyl-adenylate, and subsequently transferring the aminoacyl moiety to the 3Ј terminus of the cognate tRNA.
Faithful translation of the genetic code largely relies on the correct selection and recognition of substrates by aaRSs (4). It is challenging for some aaRSs to sufficiently discriminate between cognate and noncognate amino acids because these molecules are biochemically and/or structurally similar (5). To address this difficulty, approximately half of the extant aaRSs incorporate a proofreading (editing) activity to clear misactivated aminoacyl-adenylate (pretransfer editing) and/or mischarged tRNAs (post-transfer editing) (6). Pretransfer editing is further classified into tRNA-independent or tRNA-dependent editing on the basis of the presence or absence of tRNA in stimulating editing (7,8).
In most cases, correct aminoacyl-tRNA generation is critical for the faithful transduction of genetic information, which is supported by the high levels of amino acid conservation in editing active sites of specific aaRSs across the three domains of life. When this editing function is impaired or abolished, cellular dysfunction or disease phenotypes are often observed (9,10). However, in a limited number of instances, the synthesis of mischarged tRNA (from either compromised editing or tRNA identity alteration) has conferred unexpected advantages. For example, translational ambiguity at the CUG Leu codons derived from the synthesis of Leu-tRNA Ser by Candida albicans leucyl-tRNA synthetase (LeuRS) expands the protein primary structure on a proteome-wide scale and improves protein plasticity with morphopathogenesis (11,12). LeuRS and phenylalanyl-tRNA synthetase (PheRS) from some Mycoplasma parasites carry point mutations or deletions in their editing domains, thus yielding mischarged tRNA Leu or tRNA Phe that is likely to escape host defense systems, because of antigen diversity from mistranslation (13)(14)(15).
Threonyl-tRNA synthetases (ThrRSs) can be divided into eukaryotic/bacterial and archaeal types (16,17). ThrRSs belong to class II aaRS and form dimers. The domain composition of the bacterial and eukaryotic (cytoplasmic and mitochondrial) ThrRSs is well conserved. The domains include the N1 (with unknown function), N2 (for editing), aminoacylation (for amino acid activation and tRNA charging), and C-terminal (for tRNA binding) domains (CTDs) (16), along with a newly evolved N-terminal extension (N-extension) in eukaryotic enzymes (for catalysis and protein stability) (18). With the exception of Saccharomyces cerevisiae mitochondrial ThrRS (ScmtThrRS) and ThrRSs from the mitochondria of other yeasts (such as C. albicans and Schizosaccharomyces pombe) naturally lacking the editing domain (19,20), all other known ThrRSs are capable of editing; they include Escherichia coli ThrRS (EcThrRS) (21), S. cerevisiae cytosolic ThrRS (ScThrRS) (22), and human mitochondrial ThrRS (hmtThrRS) (23). Editing by ThrRSs takes place in the N2 editing domain (21). The reason for the editing deficiency of the ScmtThrRS is unknown but may be similar to the reasons underlying the lack of editing in other mitochondrial aaRSs, such as LeuRS (24) and PheRS (25). The protein translation system of the Mycoplasma species is unique in that several aaRSs (class I LeuRS and class II PheRS) are editing-deficient, thus leading to mistranslation and the subsequent accumulation of a statistical proteome, which is likely to be responsible for antigenic diversity and phenotypic plasticity (13). In addition, on the basis of sequence analysis, ThrRS from some Mycoplasma species have been suggested to be editing-deficient (13). Here, we show that the ThrRS from Mycoplasma mobile (MmThrRS) without the N1 domain is editing-capable, whereas the ThrRS from Mycoplasma capricolum (McThrRS) with both the N1 and N2 domains is an error-prone aaRS, and provide a rationale for the function of MmThrRS without the N1 domain. Eukaryotic yeast cells are sensitive to editing catalyzed by introduced bacterial ThrRSs. For this reason, the yeast ThrRS knock-out strain is an excellent tool for assessing the editing of bacterial ThrRSs. Although the function of N1 has not been previously clearly assigned, our results show that the editing catalyzed by EcThrRS is mediated by the communication between the N2 editing and N1 domains.

Results
Sequence Analysis of Mycoplasma ThrRSs-ThrRSs from bacteria (e.g., EcThrRS) and eukaryotes (either cytoplasmic (e.g. ScThrRS) or mitochondria (e.g. hmtThrRS)) contain well conserved N1, N2, aminoacylation, and CTDs; the eukaryotic enzymes have an additional N-extension domain. The Scmt-ThrRS is an exception because it consists of only the aminoacylation and CTDs (19). Detailed primary sequence analysis of ThrRSs from Mycoplasma species has revealed the diversity of their domains. Despite having a domain composition similar to that of EcThrRS, the active site of the McThrRS N2 editing domain was degenerate (Fig. 1A). Counterparts of crucial residues with functional side chains in the editing active site of EcThrRS (including His 73 , His 77 , Lys 156 , Asp 180 , Cys 182 , and His 186 ) (bold and light green in Fig. 1B) (21) have been extensively mutated to Leu 69 , Phe 73 , Gln 150 , Met 174 , Val 176 , and Leu 180 (bold and red in Fig. 1B), thus raising the question of how this enzyme maintains translational quality control. MmThrRS lacks an N1 domain (with unknown evolutionary advantages) but maintains the intact editing active site of the N2 domain (bold and yellow in Fig. 1B). Excluding ThrRSs containing only the aminoacylation and CTDs (such as yeast mitochondrial ThrRSs), to our knowledge, the MmThrRS is the first example of a ThrRS naturally lacking an N1 domain. Similarly to Scmt-ThrRS, the Mycoplasma wenyonii ThrRS (MwThrRS) has retained only the aminoacylation and CTDs. Although the ThrRS from Mycoplasma hemocanis (MhThrRS) has only aminoacylation and CTDs, its N terminus possesses an additional domain with no significant homolog and for which no function has been assigned. Thus, the ThrRSs from Mycoplasma species exhibit the most distinct diversity in the composition and arrangement of their domains (Fig. 1A).
MmThrRS Is Editing-capable, Whereas McThrRS Is an Errorprone aaRS-The features of ThrRSs harboring only the aminoacylation and CTDs have been studied using ScmtThrRS as a model (19,20). Thus, in this work, we focused on the study of McThrRS (with a degenerate editing domain) and MmThrRS (with no N1 domain). Because these two enzymes are modified near or at the editing domain, we initially intended to study whether they have editing capability. The editing requirement is based on whether an aaRS is able to discriminate between cognate and noncognate amino acids with a sufficient discrimination factor (DF) (Յ1/3300) (26). The DF can be calculated from kinetic constants using the equation DF ϭ (k cat /K m ) cognate amino acid /(k cat /K m ) noncognate amino acid . We purified MmThrRS and McThrRS to homogeneity (Fig.  1C). The kinetics of MmThrRS and McThrRS for cognate Thr and noncognate Ser was determined with an ATP-PPi exchange reaction. We found that the two enzymes displayed DFs of ϳ1/900 (1/944 and 1/848 for McThrRS and MmThrRS, respectively) ( Table 1), which are significantly larger than the acceptable threshold (1/3300) (26). Thus, we concluded that each of these enzymes misactivates noncognate Ser and therefore is likely to require an editing function to ensure the correct Thr-tRNA Thr formation.
To directly monitor the editing activities of the two enzymes, we prepared Ser-tRNA Thr by using the editing-deficient EcThrRS-H73A/H77A (21). We found that MmThrRS, but not McThrRS, was able to hydrolyze Ser-tRNA Thr and remove noncognate Ser ( Fig. 2A), thus indicating that the degeneration of the crucial editing active sites of McThrRS impairs its posttransfer editing. In contrast, MmThrRS, which has an intact N2 domain, has post-transfer editing activity. We further mutated two conserved His residues of the MmThrRS N2 domain (His 9 and His 13 ), whose counterparts participate in the editing activities of EcThrRS (His 73 and His 77 ) (21) and ScThrRS (His 151 and His 155 ) (22). The MmThrRS-H9A/H13A mutant had aminoacylation activity (data not shown) similar to that of MmThrRS, but its post-transfer editing was not detected in vitro, thus indicating that the N2 domain is involved in post-transfer editing ( Fig. 2A).
Furthermore, the misaminoacylation of tRNA Thr with Ser by McThrRS, MmThrRS, and EcThrRS (as an editing-capable control) was performed. We found that Ser-tRNA Thr was apparently formed by McThrRS, whereas only trace amounts of Ser-tRNA Thr were formed by MmThrRS, which was comparable with EcThrRS (Fig. 2, B and C).
These data indicate that MmThrRS is an editing-capable ThrRS, whereas the McThrRS has lost its post-transfer editing activity, probably because of its degenerate editing active site in the N2 domain. Thus, because it clearly synthesized mischarged tRNA Thr , McThrRS is an error-prone tRNA synthetase.
MmThrRS and McThrRS Have Negligible tRNA-dependent Pretransfer Editing Capacities-In addition to their post-transfer editing capacity, eukaryotic cytoplasmic ThrRSs (e.g. ScThrRS) also exhibit tRNA-independent and tRNA-dependent pretransfer editing activities, contributing the highest pro-portion in the overall editing pathways (22). A Ser-included AMP formation assay was used to clarify whether McThrRS and MmThrRS also possess pretransfer editing activities.
According to this approach, the k obs values of AMP formation for MmThrRS were 0.093 Ϯ 0.011 and 0.086 Ϯ 0.010 s Ϫ1 in the absence and presence of tRNA Thr , respectively (Fig. 3, A and C), whereas the k obs values of AMP formation for McThrRS were 0.224 Ϯ 0.027 and 0.225 Ϯ 0.032 s Ϫ1 , respectively, which were larger than those for MmThrRS (Fig. 3, B and D, and Table 2). We also performed Thr-included AMP formation assays in the absence of tRNA Thr with MmThrRS and McThrRS. The result showed that k obs value of MmThrRS for AMP formation was too low to be accurately determined (Fig. 4A), whereas that of McThrRS was 0.013 Ϯ 0.004 s Ϫ1 with Thr ( Fig. 4, B and C), significantly lower than that with Ser (0.224 Ϯ 0.027 s Ϫ1 ), indi-      AMP formation of MmThrRS-H9A/H13A was 0.021 Ϯ 0.007 s Ϫ1 in the presence of tRNA Thr , which was larger than that of MmThrRS (Table 2). These data clearly show that tRNA Thr does not stimulate tRNA-dependent pretransfer editing of The inability of the McThrRS to rescue Sc⌬thrS growth might have originated from its failure to aminoacylate yeast tRNA Thr isoacceptors or to perform post-transfer editing. To distinguish between these two possibilities, we assayed the aminoacylation of the three yeast isoacceptors tRNA Thr (UGU), tRNA Thr (CGU), and tRNA Thr (AGU) by McThrRS. Our data showed that McThrRS aminoacylated each yeast tRNA Thr isoacceptor with a higher catalytic rate than did MmThrRS (Fig. 5B), thus indicating that the defective editing function of McThrRS prevents the yeast growth.
To determine whether editing produces phenotype differences, we replaced the N1 and  (Fig. 5D). Furthermore, the chimeric enzyme in Sc⌬thrS was able to rescue cell growth, further suggesting that the defective editing of McThrRS blocks yeast growth in vivo (Fig. 5E).
To further reveal whether the editing-defective ThrRS leads to a growth failure in yeast, we introduced MmThrRS-H9A/ H13A into Sc⌬thrS. Mutations at these two editing-crucial residues resulted in slow growth, thus clearly indicating that impaired post-transfer editing impeded normal yeast growth The post-transfer editing of MmThrRS-N2M was abolished by these mutations (Fig. 5D), and Sc⌬thrS harboring MmThrRS-N2M did not grow (Fig. 5F). However, in the absence of 5-FOA, by Western blot analysis, we detected no differences in either the in vivo amounts of MmThrRS, MmThrRS-H9A/H13A, and MmThrRS-N2M in Sc⌬thrS on the basis of a His 6 tag at the C terminus of each protein (Fig. 5G) or the in vitro aminoacylation of tRNA Thr by these three enzymes (data not shown). Thus, the loss of editing capacity in these mutants is responsible for their failure to rescue the growth defects of Sc⌬thrS.
These observations suggest a correlation between the growth phenotype of Sc⌬thrS and the editing capacities of various native or mutant ThrRSs (Fig. 5C). The findings also clearly show that Sc⌬thrS cells are highly sensitive to post-transfer editing catalyzed by introduced bacterial ThrRSs. Therefore, Sc⌬thrS is an ideal tool to study not only aminoacylation activity (22) but also the post-transfer editing activity of exogenous bacterial ThrRSs in vivo in subsequent studies.
Absence of the N1 Domain of MmThrRS Increases in Vivo Activity and Optimizes Protein Structure/Stability-To better understand the consequences of the absence of the N1 domain from MmThrRS, we fused the N1 domain of McThrRS (Met 1 -Leu 67 ) at the N terminus of MmThrRS, resulting in the chimeric mutant MmThrRS-ƒMcN1 (Fig. 5C). MmThrRS-ƒMcN1 was unable to rescue cell growth in the presence of 5-FOA (Figs. 5C and 6A), indicating that the activity and/or protein level of the chimeric mutant had been significantly decreased. Western blot analysis of the yeast cell lysate showed that under the same conditions, the amount of MmThrRS-ƒMcN1 dropped sharply to ϳ20% of that of MmThrRS, indicating that maintenance of the N1 domain of McThrRS has a negative effect on MmThrRS protein structure and/or stability. However, the MmThrRS-ƒMcN1 expression level was comparable with that of ScThrRS (Fig. 6, B and C), implying that the protein activity was also influenced by the presence of the N1 domain of McThrRS. We  Because of the editing deficiency of McThrRS and its inability to support the growth of Sc⌬thrS, we were unable assess the contribution of the N1 domain to the editing activity of ThrRS and its in vivo function. Therefore, we further used EcThrRS, which has editing function (21) and was able to replace yeast ThrRS for in vivo aminoacylation (Fig. 6A). An EcThrRS-⌬N1 (Lys 63 -Glu 642 ) expression vector was constructed, and the enzyme was purified (Fig. 1C). Consistently with the in vitro data of aminoacylation kinetics from McThrRS and McThrRS-⌬N1 (Table 4), N1 truncation did not have an observable influence, thus further showing that the N1 domain contributes little to substrate binding and tRNA charging. To determine whether this N1 truncation influences the enzyme structure, we determined the dissociation constant (k d ) values of the native EcThrRS and EcThrRS-⌬N1 with tRNA Thr by using a filter-binding assay. The k d values for EcThrRS and EcThrRS-⌬N1 were 1.92 Ϯ 0.23 and 1.54 Ϯ 0.20 M, respectively, indicating that the N1 truncation has a negligible effect on the protein structure (Fig. 7A). However, the EcThrRS-⌬N1 enzyme did not support the growth of Sc⌬thrS (Figs. 5C and 7B). Thus, we speculated that editing is impaired in EcThrRS-⌬N1. Indeed, the hydrolysis of Ser-tRNA Thr revealed that EcThrRS-⌬N1 had no post-transfer editing (Fig. 7C). The editing deficiency and failure to rescue Sc⌬thrS containing EcThrRS-⌬N1 were reminiscent of the characteristics of McThrRS and MmThrRS-N2M, thus strongly suggesting that Sc⌬thrS cell viability is directly regulated by the ThrRS editing function. Both MmThrRS and McThrRS lack tRNA-dependent pretransfer editing capacity. To elucidate whether EcThrRS has tRNAindependent and tRNA-dependent pretransfer editing activities, along with any potential role of the N1 domain, we assayed AMP formation catalyzed by EcThrRS and EcThrRS-⌬N1, respectively. The amount of AMP produced by EcThrRS through a TLC-based editing assay was significantly stimulated in the presence of tRNA (Fig. 7, D and E), and the calculated k obs values were 1.81 Ϯ 0.21 s Ϫ1 with tRNA and 0.33 Ϯ 0.05 s Ϫ1 without tRNA, thus indicating strong tRNA-dependent editing, including the pre-and post-transfer editing of EcThrRS (Table  5). However, under the same conditions, the k obs values of the EcThrRS-⌬N1 in the presence and absence of tRNA Thr were only slightly different (0.34 Ϯ 0.06 s Ϫ1 with tRNA and 0.25 Ϯ  0.04 s Ϫ1 without tRNA) (Fig. 7, F and G, and Table 5). These data show that tRNA triggers little AMP formation by EcThrRS-⌬N1 without post-transfer editing and that EcThrRS has negligible tRNA-dependent pretransfer editing activity, similarly to MmThrRS and McThrRS (Fig. 3). In addition, EcThrRS and EcThrRS-⌬N1 have similar tRNA-independent pretransfer editing activity levels, thus suggesting that the N1 domain does not influence tRNA-independent pretransfer editing, which is considered to occur at the synthetic active site (19,27), and further implying that the N1 domain truncation does not affect the protein structure.
The Importance of the N1 Domain for Editing EcThrRS Relies on the Interdomain Communication-Given the importance of the N1 domain in editing of EcThrRS and in supporting the growth of Sc⌬thrS, we subsequently studied its mechanism. The editing active site of the ThrRS is within the N2 domain; however, our above data showed that the N1 domain is involved in the editing activity of ThrRS. The N1 domain is located at the N terminus of the N2 editing domain, far away from the mischarged tRNA Thr synthesized in the aminoacylation active site (28). On the basis of its spatial localization, we propose that interdomain communication is likely between the N1 and N2 domains. We analyzed the primary and tertiary structures of the available ThrRS crystal structures. On the basis of the tertiary structure of EcThrRS (Protein Data Bank code 1QF6) (16), the Asp 46 residue in the N1 domain might potentially play a role in the N1-N2 interaction because: (i) the Asp residue is the only absolutely conserved residue within the N1 domain in all ThrRSs harboring the N1 domain (Fig. 8A), and (ii) Asp 46 is located at the interface between the N1 and N2 domains, and its side chain points toward the N2 domain (Fig. 8B). To uncover the role of Asp 46 in the in vivo function of EcThrRS, Asp 46 was mutated to nine other residues (Ala, Arg, Glu, Tyr, Asn, Cys, Leu, Pro, and Thr). Each gene encoding these variants was then expressed in Sc⌬thrS. Strikingly, none of the Asp 46 mutants were able to rescue the growth of the thrS deletion strain (Fig.  8C), including the Glu 46 substitution variant with the similarly negative charge (EcThrRS-D46E). The results indicated that both the charge and size of the side chain at residue 46 are crucial to the growth of Sc⌬thrS. To determine whether the growth deficiency was due to the loss of editing and/or aminoacylation activity, we selected EcThrRS-D46R with a positively charged residue at position 46 and EcThrRS-D46E with negatively charged residue at this position as the representative samples and then investigated their in vitro aminoacylation and editing activities relative to the EcThrRS. Although EcThrRS-D46R and EcThrRS-D46E had similar aminoacylation activities as the mutants with EcThrRS (Fig. 8D), their post-transfer editing activity was completely lost (Fig. 8E), thus suggesting that a loss of editing activity was responsible for the failure to rescue. Because Asp 46 is located in the N1 domain, it is likely that Asp 46 mediates communication between the two domains by its interaction with some residues in the N2 domain to regulate the editing activity of EcThrRS.
By carefully analyzing the structure of EcThrRS (Protein Data Bank code 1QF6) (16), we found that an oxygen atom of the carboxyl group side chain of Asp 46 potentially interacts with the side chain of Lys 136 (2.95 Å) or Tyr 173 (2.66 Å) in the N2 domain, whereas another oxygen atom of Asp 46 likely interacts with the amide group of the His 186 main chain (2.83 Å) in the N2 domain (Figs. 8B and 9A). The Lys 136 mutants substituted with Ala, Glu, and Arg were tested for their capacity to rescue the growth of the Sc⌬thrS strain. EcThrRS-K136E was unable to replace ScThrRS; however, both EcThrRS-K136A and EcThrRS-K136R compensated for the loss of ScThrRS and EcThrRS (Fig. 9B). These data indicated that the interaction of Asp 46 with other potential residues was sufficient for in vivo function in EcThrRS-K136A and EcThrRS-K136R. Indeed, introduction of a Glu residue at position 136 leads to carboxylate groups of both Glu 136 and Asp 46 into close proximity, inducing structural disruption, thus accounting for the failure of EcThrRS-K136E to rescue growth of Sc⌬thrS. In contrast, Ala 136 in EcThrRS-K136A and Arg 136 in EcThrRS-K136R are accommodated both in terms of size (a small methyl group) or complementary charge. After substitution of Tyr 173 with Ala, Asp, or Arg, none of the three mutants, EcThrRS-Y173A, -Y173R, or -Y173D, rescued the growth of Sc⌬thrS, thus implying that Tyr 173 is crucial for the in vivo function of EcThrRS (Fig. 9B). The purified EcThrRS-Y173R and EcThrRS-Y173D had modest reductions in their aminoacylation activity compared with EcThrRS (within 2-fold) (Fig. 9C); however, their post-transfer editing activity was abolished (Fig. 9D), again indicating that mutants unable to perform post-transfer editing could not complement the loss of thrS in Sc⌬thrS, similarly to EcThrRS-⌬N1, EcThrRS-D46R, and EcThrRS-D46E.
The above data show the importance of the Asp 46 in the N1 domain and the Tyr 173 in the N2 domain. If these two residues directly interact, such an interaction could be re-established by the simultaneous mutation of the two residues. We constructed five double mutants in the context of EcThrRS-D46E, including EcThrRS-D46E/Y173H, -D46E/Y173K, -D46E/Y173S, EcThrRS-D46E/Y173F, and -D46E/Y173R. Only EcThrRS-D46E/Y173F was able to support Sc⌬thrS growth, despite its slightly decreased efficiency, thus suggesting a re-established interaction between the N1 and N2 domains (Fig. 9E). EcThrRS-D46E did not rescue Sc⌬thrS growth, and EcThrRS-Y173F provided poor support for growth compared with EcThrRS, thus suggesting that simultaneous mutations at both the 46 and 173 positions are required for the recovery of interaction between the N1-N2 domains (Fig. 9E). As described above, Asp 46 potentially also interacts with the main chain amide group of His 186 . When His 186 was substituted with Gly (EcThrRS-H186G) to minimalize the side chain and thereby enhance the flexibility, the EcThrRS-H186G complemented ScThrRS very well (data not shown). Interestingly, when Gly 186 was introduced into the complementation-defective EcThrRS-D46E, the growth of the yeast containing EcThrRS-D46E/H186G was restored to levels comparable with those of the native enzyme (Fig. 9E), thus implying that the bulkier Glu at position 46 clashes with the main chain of His 186 . However, the introduction of the smallest Gly at 186 improved the communication.
The aminoacylation and post-transfer editing activities of the purified EcThrRS-D46E/Y173F and EcThrRS-D46E/ H186G were assayed. Their aminoacylation activity was decreased only slightly (Fig. 9C), whereas their post-transfer editing activity was modestly restored (Fig. 9F). This slight increase in the editing seemed sufficient to support in vivo translational quality control. Collectively, these data demonstrate that there is an interdomain communication mechanism mediated by an interaction between residues Asp 46 in the N1 domain and Lys 136 , Tyr 173 , and His 186 in the N2 domain, which is critical for the editing active site in the N2 domain and cell survival.

TABLE 5 k obs values of EcThrRS and EcThrRS-⌬N1 in AMP formation assays
The results are the averages of three independent repeats, with the standard deviations indicated.
EcThrRS (Protein Data Bank code 1QF6) (16), it is clear that the aromatic ring of Tyr 173 stacks on the imidazole group of His 186 on one face while packing against the aliphatic portion of the Lys 136 side chain on the other face. Thus, we suggest that Tyr 173 maintains a hydrophobic core and that the salt bridge between Asp 46 to Lys 136 is not sufficient to sustain the fold when a cavity introduced by only mutation at Tyr 173 . Interestingly, EcThrRS-D46E/H186G supports growth of Sc⌬thrS well, indicating that the vacant position created by Gly at position 186 is compensated by nearly residues, possibly His 73 or His 77 . Altogether, these results suggest that the network is complex and requires multiple interactions. Indeed, mutations at both Asp 46 and Tyr 173 can restore yeast growth. The assay of the editing activity of bacterial ThrRSs with Sc⌬thrS was more sensitive than with the in vitro hydrolysis of Ser-tRNA Thr . Hence, Sc⌬thrS is an excellent tool for studying the in vivo editing function of introduced bacterial ThrRSs.
Hidden Contribution of N1 Domain to the Editing Function of ThrRS-Using ScThrRS, we have previously shown that the N-extension of eukaryotic ThrRS is involved in the enzymatic activity, structure, and stability of ThrRS (18). However, the biological significance of the N1 domain of ThrRS has so far remained elusive. EcThrRS has N1, N2, aminoacylation, and CTDs. By aligning sequences from various species, we deter-mined that MmThrRS did not have an N1 domain. By constructing EcThrRS-⌬N1, we found that the N1 domain is directly involved in the editing function of ThrRS. Indeed, the effect of the N1 domain on the editing function of ThrRS had been ignored until now because of its spatial separation from the editing active site of the N2 domain. Asp 46 , the only absolutely conserved amino acid residue in the N1 domain of EcThrRS, forms an interaction network with the Lys 136 , Tyr 173 , and His 186 in the N2 domain. Substitution of Asp 46 with any other amino acid, even Glu or Asn, resulted in a loss of the editing activity of EcThrRS, thus suggesting that both the property and the size of this residue play critical roles in this interdomain communication. We propose that an interaction network among Asp 46 , Lys 136 , Tyr 173 , and His 186 controls the spatial orientation or rotation of the N2 editing domain during the translocation of the CCA 76 end from the aminoacylation active site to the editing active site. Disruption of this interaction by a mutation at Asp 46 or Tyr 173 inhibited the CCA 76 end with Ser from being captured or accommodated by the N2 domain. Indeed, the rotation of the CP1 editing domain, a prerequisite for a proper accommodation of the CCA 76 end of mischarged tRNAs, has been observed in class I aaRSs, LeuRS, and valyl-tRNA synthetase (29 -31). However, the absence of the N1 domain was beneficial for the function and structure of

MmThrRS in vivo.
Even without an N1 domain, the MmThrRS retained post-transfer editing activity. It is possible that the N2 domain of MmThrRS is more flexible than that of EcThrRS and that this flexibility is responsible for the translocation of the CCA 76 end of tRNA during editing.
Mycoplasma Species May Differ in Requirement of Editing-LeuRSs from three Mycoplasma species (M. mobile, M. agalactiae, and M. synoviae) exhibit either loss or partial truncation of the editing domain, thus leading to editing deficiency and Ile misincorporation of the Leu codons in vivo (13). Similarly, the editing active sites of PheRS from M. mobile (MmPheRS) are degenerate, thus leading to Tyr replacement at the Phe codons in vivo (15). Mistranslation in the proteome of M. mobile has been suggested to be an evolutionary strategy to confer phenotypic plasticity in Mycoplasma pathogens. However, it is notable that some Mycoplasma species have retained editing-functional LeuRSs or PheRSs (13,15).  control in the different Mycoplasma organisms. Notably, M. mobile encodes an editing-deficient LeuRS and PheRS, an evolutionary phenomenon that is probably advantageous, whereas it encodes an editing-capable ThrRS for the translation of Thr codons. Similarly, in M. capricolum, its ThrRS and PheRS are editing-deficient, whereas its LeuRS has been reported to be editing-capable (13). Thus, it is likely that different aaRSs require various levels of translational quality control at the codons of specific amino acids, even within a single Mycoplasma organism.

Experimental Procedures
Materials-L-Thr, L-Ser, dithiothreitol, tetrasodium pyrophosphate, inorganic pyrophosphate, Tris-HCl, MgCl 2 , NaCl, and activated charcoal were purchased from Sigma. [ 14 C]Thr was obtained from Biotrend Chemicals (Destin, FL), and [ 14 C]Ser and [␣-32 P]ATP were obtained from PerkinElmer Life Sciences. DNA fragment rapid purification and plasmid extraction kits were purchased from Yuanpinghao Biotech (Tianjing, China). The KOD-plus mutagenesis kit was obtained from TOYOBO. T4 DNA ligase and restriction endonucleases were obtained from Thermo Scientific (Pittsburgh, PA). Phusion high fidelity DNA polymerase was purchased from New England Biolabs. Nickel-nitrilotriacetic acid Superflow was purchased from Qiagen. Polyethyleneimine cellulose plates were purchased from Merck. Pyrophosphatase (PPiase) was obtained from Roche Applied Science. The dNTP mixture was obtained from TaKaRa (Japan). Oligonucleotide primers were synthesized by Invitrogen. E. coli BL21 (DE3) cells were purchased from Stratagene (Santa Clara, CA).
Cloning and Mutagenesis-The genomes of M. mobile and M. capricolum were kindly provided by Dr. Makoto Miyata (Osaka City University). The ORFs of the MmThrRS and McThrRS genes were amplified from genomic DNA and were separately inserted via restriction digestion into the pET28a(ϩ) vector that was cut with NheI-XhoI and NcoI-XhoI. All TGA codons (the codon encoding Trp in Mycoplasma species) in the ORFs of the MmThrRS and McThrRS genes were changed to Trp (TGG) codons (13). The recombinant plasmids containing the ScThrRS and EcThrRS genes, p425TEF-ScThrRS and p425TEF-EcThrRS, had been previously constructed in our lab (22). The ORFs encoding MmThrRS and McThrRS were inserted via restriction digestion into the yeast p425TEF expression vector that was cut with BamHI-XhoI to form p425TEF-MmThrRS and p425TEF-McThrRS, respectively. All genes with deletions and point mutations were made by following the instructions in the KOD-plus mutagenesis kit. All constructs were confirmed by DNA sequencing.
tRNA Gene Cloning, Expression, and Purification-According to our previous work, four DNA fragments covering the E. coli tRNA Thr (UGU) gene double strands were synthesized by Invitrogen, phosphorylated, and ligated into pTrc99b (precleaved by PstI and EcoRI) to construct the pTrc99b-tRNA Thr plasmid (33). E. coli MT102 cells were transformed with pTrc99b-tRNA Thr and cultured to express the tRNA Thr gene. The Thr accepting activity of the tRNA Thr isolated from the extract of the transformants reached 1500 pmol/A 260 . Because the accepting activity of pure tRNA Thr was considered to be 1600 pmol/A 260 , tRNA Thr should comprise ϳ93% of the total tRNA. 32  EcThRS and its mutants. At various time intervals, 9-l aliquots of the reaction solution were added to Whatman filter pads and quenched with cold 5% TCA. The pads were washed three times for 15 min each with cold 5% TCA and then three times for 10 min each with 100% ethanol. The pads were then dried under a heat lamp, and the radioactivity of the precipitates was quantified using a scintillation counter (Beckman Coulter).
In the ATP-PPi exchange reaction, the K m values of MmThrRS and McThrRS for Thr (3.39 and 4.40 mM, respectively) were significantly higher than those of other ThrRSs (for example, 0.21 mM for ScThrRS) (22). In addition, the signal from [ 32 P]AMP or aminoacyl-[ 32 P]AMP could be directly observed on the plates used in the TLC assays (34). To improve the accuracy of the aminoacylation assay, their kinetic parameters were measured using [ 32 P]tRNA Thr . The kinetic parame-ters of McThrRS and McThrRS-⌬N1 were measured at 30°C in a similar reaction mixture using 0.137 M [ 32 P]tRNA Thr and 200 nM enzymes. For Thr, the concentration of Thr varied from 0.1-2 mM; for ATP, the concentration was changed from 0.005 mM to 1 mM; for tRNA Thr , 0 -20 M cold tRNA Thr was used. The kinetic parameters of EcThrRS and EcThrRS-⌬N1 were measured at 37°C using 0.137 M [ 32 P]tRNA Thr and 25 nM enzyme. For Thr, the concentration was varied from 5-2,000 M, for ATP, the concentration was changed from 5 to 2000 M, and for tRNA Thr , 0 -20 M cold tRNA Thr was used.
At specific time points, samples were taken for ethanol precipitation with NaAc (pH 5.2) at Ϫ20°C overnight. The precipitated samples were centrifuged (10,000 ϫ g) at 4°C for 30 min, dried at room temperature for 30 min, and digested with 6 l of nuclease S1 (25 units) for 2 h at 37°C. After treatment with nuclease S1, aminoacyl-[ 32 P]AMP and [ 32 P]AMP should be produced from aminoacyl-[ 32 P]tRNA and free [ 32 P]tRNA, respectively. Samples (2 l) of the digestion mixture were loaded and separated by TLC in 0.1 M NH 4 Ac and 5% acetic acid. The plates were visualized by phosphorimaging, and the data were analyzed using Multi-Gauge Version 3.0 software (FUJIFILM). Post-transfer Editing-The post-transfer editing activity of the various ThrRSs and mutants was assessed by the hydrolysis of [ 14 C]Ser-tRNA Thr and was measured at 30°C (for Mycoplasma ThrRSs) or 37°C (for EcThrRSs) in a reaction mixture containing 60 mM Tris-HCl (pH 7.5), 20 mM MgCl 2 , 30 mM KCl, 2 M [ 14 C]Ser-tRNA Thr , and 200 nM enzyme. The preformed [ 14 C]Ser-tRNA Thr was prepared by using editing-deficient EcThrRS-H73A/H77A (21).

Misaminoacylation-Misaminoacylation by
AMP Formation Assay-AMP formation was assayed by TLC at 30°C in a reaction mixture containing 60 mM Tris-HCl (pH 7.5), 20 mM MgCl 2 , 30 mM KCl, 10 units/ml PPiase, 10 mM Thr or 1 M Ser, 3 mM [␣-32 P]ATP, and 2 M Mycoplasma ThrRSs or its mutant derivatives, either in the absence or presence of 20 M tRNA Thr . For EcThrRS and its mutants, the reaction was performed at 37°C in the same reaction mixture but with the addition of 40 mM Ser and 1 M enzyme. Samples (1.5 l) were quenched in 6 l of 200 mM NaAc (pH 5.0). The quenched aliquots (1.5 l of each sample) were then spotted onto polyethyleneimine cellulose plates that had been prewashed with water. The separation of Ser-[␣-32 P]AMP, [␣-32 P]AMP, and [␣-32 P]ATP was performed in 0.1 M NH 4 Ac and 5% acetic acid. The plates were visualized by phosphorimaging, and the data were analyzed using Multi Gauge software (version 3.0, FUJIF-ILM). Quantification of [␣-32 P]AMP was achieved by densitometry relative to samples containing known concentrations of [␣-32 P]ATP. The rates were obtained using only the initial time points, where the plot of [␣-32 P]AMP versus time was linear. The data were then fit to the following equation: y ϭ b ϩ k ss t, where b and k represent the burst amplitude and the steady-state rate, respectively. The observed reaction rate con-stants (k obs ) were obtained by dividing the steady-state rate of the reaction by enzyme concentration.
Measurement of Dissociation Constants for tRNA Thr by Filter Binding Assays-The formation of the EcThrRS-⌬N1/Ec-ThrRS-[ 32 P]tRNA Thr complex was monitored using the nitrocellulose filter binding method (35). Nitrocellulose membranes (0.22 m) were presoaked in washing buffer (50 mM Hepes/ KOH, pH 6.8, 50 mM KCl) for at least 10 min before use. The [ 32 P]tRNA Thr (27,353 c.p.m., 2.74 pmol) was incubated with either EcThrRS or EcThrRS-⌬N1 at various concentrations (0.2-9 M) for 30 min at 0°C in 50 l of buffer containing 50 mM Hepes/KOH (pH 6.8), 50 mM KCl, and 12 mM MgCl 2 . The samples were then applied and filtered through the nitrocellulose membrane. The filters were washed with 0.3 ml of washing buffer and dried, and the radioactivity was measured. The data were then analyzed using GraphPad Prism.
Sc⌬thrS Complementation Assay-The Sc⌬thrS yeast strain was constructed in our lab (22). For complementation assays, the genes were inserted into p425TEF, which was modified to have an in-frame DNA sequence encoding a His 6 tag at the C terminus of the ThrRSs. The constructs were transformed into Sc⌬thrS using the LiAc method (36). Transformants were selected on SD/Ura Ϫ /Leu Ϫ /G418 plates, and a single clone was cultured in liquid SD/Leu Ϫ /G418 medium. The culture was then diluted to a concentration equivalent to 1 A 600 , and a 10-fold dilution of the yeast was plated onto SD/Leu Ϫ /G418 in the presence of 5-FOA to induce the loss of the rescue plasmid (pRS426-ScthrS). Complementation was observed by comparing the growth rates of Sc⌬thrS expressing native ScThrRS, MmThrRS, McThrRS, EcThrRS, and their mutants.