C-terminal Domain of Leucyl-tRNA Synthetase from Pathogenic Candida albicans Recognizes both tRNASer and tRNALeu*

Leucyl-tRNA synthetase (LeuRS) is a multidomain enzyme that catalyzes Leu-tRNALeu formation and is classified into bacterial and archaeal/eukaryotic types with significant diversity in the C-terminal domain (CTD). CTDs of both bacterial and archaeal LeuRSs have been reported to recognize tRNALeu through different modes of interaction. In the human pathogen Candida albicans, the cytoplasmic LeuRS (CaLeuRS) is distinguished by its capacity to recognize a uniquely evolved chimeric tRNASer (CatRNASer(CAG)) in addition to its cognate CatRNALeu, leading to CUG codon reassignment. Our previous study showed that eukaryotic but not archaeal LeuRSs recognize this peculiar tRNASer, suggesting the significance of their highly divergent CTDs in tRNASer recognition. The results of this study provided the first evidence of the indispensable function of the CTD of eukaryotic LeuRS in recognizing non-cognate CatRNASer and cognate CatRNALeu. Three lysine residues were identified as involved in mediating enzyme-tRNA interaction in the leucylation process: mutation of all three sites totally ablated the leucylation activity. The importance of the three lysine residues was further verified by gel mobility shift assays and complementation of a yeast leuS gene knock-out strain.

CaSerRS, a key aaRS in CUG reassignment, contains only one CUG codon-encoded residue at position 197 in the dimerization interface of the enzyme (22). Replacement of Ser 197 by Leu 197 in CaSerRS causes a local structure rearrangement and induces slightly higher catalytic activity without affecting Ser activation. The N-terminal catalytic domain of CaSerRS interacts with the long variable arm of tRNA Ser with its CTD stabilizing the intramonomer interaction (22,23). Thus, the element in CaSerRS that recognizes tRNA Ser has been revealed. Another critical aaRS involved in CUG decoding ambiguity, CaLeuRS, with 1,097 residues, also contains a single CUG codon-encoded residue at position 919, located in its CTD (from Gly 894 to Glu 1097 ) (24). The two isoforms of CaLeuRS (CaLeuRS-Ser 919 and -Leu 919 ) do not differ in amino acid activation activity; however, CaLeuRS-Leu 919 has higher efficiency in leucylating CatRNA Leu (24). CaLeuRS-Ser 919 predominates in vivo; therefore, this isoform is designated as the wild-type (WT) enzyme. For the interaction between CaLeuRS and CatRNA Ser , the elements in CaLeuRS that recognize CatRNA Ser (CAG) have not been identified.
On the basis of crystal structures of bacterial and archaeal LeuRSs, the CTD folds into a separated domain and is disordered in the absence of tRNA (7,25). Indeed, deletion analysis showed that the CTDs of Escherichia coli LeuRS (EcLeuRS), Thermus thermophilus LeuRS (TtLeuRS), Pyrococcus horikoshii LeuRS (PhLeuRS), and Natrialba magadii LeuRS (NmLeuRS) are all indispensable for leucylating tRNA Leu (6, 7, 10, 26 -28). However, deletion of the CTD enhanced the leucylation activity of yeast mitochondrial LeuRS, emphasizing its adaptation in RNA splicing (28). Several studies have explored the concrete recognition elements in the CTD of LeuRSs for cognate tRNA Leu . In EcLeuRS, Ala or Asp mutations of several conserved residues had only minimal effects on the aminoacylation activity as did the corresponding double and triple sites mutants (28). In another bacterial LeuRS, Mycobacterium tuberculosis LeuRS (MtLeuRS), several residues were shown to maintain the hydrophobic environment to stabilize the conformation of its CTD and to orient the tRNA (27). In addition, the main chains of the C-terminal Pro 962 and Glu 967 of PhLeuRS are crucial for recognizing the A47c and G47d of PhtRNA Leu (10). Our previous study showed that, in addition to CaLeuRS, other eukaryotic LeuRSs, including Saccharomyces cerevisiae LeuRS (ScLeuRS) and Homo sapiens LeuRS (HsLeuRS), can also leucylate CatRNA Ser ; however, both bacterial and archaeal LeuRSs, including EcLeuRS, C. albicans mitochondrial LeuRS (CamtLeuRS), and PhLeuRS, could not (24). In particular, eukaryotic LeuRS and archaeal PhLeuRS only exhibit obvious divergence in their CTD, indicating the potential significance of the CTD of eukaryotic LeuRS in tRNA Ser (CAG) recognition.
Above all, the only unexplored question concerning interaction between CaLeuRS and CatRNA Ser (CAG) seems to be identification of the elements in CaLeuRS that recognize CatRNA Ser . Besides, although the function of the CTD of bacterial and archaeal LeuRSs in recognition of tRNA Leu has been studied extensively by structural and biochemical methods, the potential role of the CTD of eukaryotic LeuRSs in the recognition of tRNA Leu has never been reported. Here, our results first showed that CTD of CaLeuRS (CaCTD) is indispensible for leucylating both CatRNA Ser (CAG) and CatRNA Leu (CatRNAs). Additionally, three highly conserved lysine residues within CaCTD were identified as important for leucylating both tRNAs in an additive manner both in vitro and in vivo. In combination, our data identified the specific mechanism concerning recognition of CatRNA Ser (CAG) by CaLeuRS and improved our understanding of the recognition of tRNA Leu by eukaryotic LeuRS.
tRNA Gene Transcription, Methylation, and 32 P Labeling-In C. albicans, only CatRNA Ser (CAG) bears a CAG anticodon, and no tRNA Leu (CAG) exists. Therefore, in this study, we used the CatRNA Leu (UAA) isoacceptor to measure the leucylation activity of CaLeuRS. pTrc99b-CatRNA Leu (UAA) and pTrc99b-CatRNA Ser (CAG) were constructed in our laboratory (22). Detailed T7 in vitro runoff transcription of CatRNA Leu and CatRNA Ser was performed according to our previously described method (31). The purified CatRNA Ser transcript was methylated at position G37 by EcTrmD because the methyl group of m 1 G37 of CatRNA Ser is a critical element for recognition by CaLeuRS (13,32). All CatRNA Ser used in this study refers to m 1 G37 CatRNA Ser (CAG). In vitro transcription of HstRNA Leu (CAG) was performed as described previously (30). 32  In Vitro Activity Assays-For the first step of the aminoacylation reaction, amino acid activation was measured by an ATP-PP i exchange assay at 30°C in a reaction mixture containing 60 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 2 mM DTT, 4 mM ATP, 2 mM [ 32 P]tetrasodium pyrophosphate (22 cpm/mol), 1 mM Leu, and 20 nM CaLeuRS or its variants. Samples of the reaction mixture (9 l) were removed at 5-min intervals and immediately added to 200 l of quenching solution (2% activated charcoal, 3.5% HClO 4 , and 50 mM tetrasodium pyrophosphate) and mixed by vortexing for 20 s. The solution was filtered through a Whatman GF/C filter followed by washing with 20 ml of Milli-Q water and 10 ml of 95% ethanol. The filter was dried, and [ 32 P]ATP was measured using a scintillation counter (Beckman Coulter).
Leucylation of CatRNA Leu was performed at 30°C in a reaction mixture containing 60 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 2 mM DTT, 4 mM ATP, 20 M [ 3 H]Leu, 10 M CatRNA Leu , and 20 nM CaLeuRS or its variants. Samples of the reaction mixture (9 l) were removed onto a Whatman filter at 2-min intervals. After washing with 5% trichloroacetic acid (three times) and 95% ethanol (twice), the filters containing precipitated [ 3 H]leucyl-tRNA Leu were dried, and radioactivity was quantified using a scintillation counter. Leucylation of CatRNA Leu by ScLeuRS and its variant was performed under the same conditions, and that of HstRNA Leu (CAG) by HsLeuRS and its variant was performed as described previously (33). The kinetics of CatRNA Leu aminoacylation by CaLeuRS and its variants were determined in the presence of varying concentrations of tRNA Leu (0.5-32 M).
Misleucylation of [ 32 P]CatRNA Ser was carried out at 30°C in a reaction mixture containing 60 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 2 mM DTT, 4 mM ATP, 40 mM Leu, 5 M "cold" CatRNA Ser , 1 M [ 32 P]CatRNA Ser , and 1 mM CaLeuRS or its variants. Aliquots (9 l) were removed at specific time points for ethanol precipitation with sodium acetate (pH 5.2) at Ϫ20°C overnight. The precipitated samples were centrifuged at 12,000 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 1 h at 37°C. After treatment with nuclease S1, leucyl-[ 32 P]tRNA should produce leucyl-[ 32 P]AMP, and free [ 32 P]tRNA should produce [ 32 P]AMP. Quenched aliquots (2 l) of the digestion mixture were spotted on a thin polyethyleneimine cellulose plate and separated by the thin layer chromatography (TLC) in 0.1 M ammonium acetate and 5% acetic acid. Known amounts of [␣-32 P]ATP were diluted stepwise and spotted onto the plate for quantification. After visualization by phosphorimaging, the data were analyzed using Multi-Gauge Version 3.0 software (Fujifilm). Misaminoacylation of [ 32 P]CatRNA Ser by ScLeuRS, HsLeuRS, and their variants was performed under the same conditions except that misaminoacylation by HsLeuRS and its variant was performed at 37°C.
Gel Mobility Shift Assay-100 nM m 1 G37 CatRNA Ser (CAG) or CatRNA Leu (UAA) was incubated with a range of CaLeuRS (0 -12 M) or its mutants in 20 l of buffer (20 mM Tris-HCl (pH 8.0), 100 mM NaCl, and 15% glycerol) on ice for 20 min. After incubation, 1 l of loading buffer (0.25% bromphenol blue and 30% glycerol) was added to the sample, and then the sample was loaded into a 6% polyacrylamide native gel. The electrophoresis was carried out at 4°C at a constant voltage of 100 V for 80 min using 50 mM Tris-glycine buffer. The gel was then stained with ethidium bromide. The RNA bands were quantified by using a Fujifilm imaging analyzer.
Yeast Complementation Assay-The yeast leuS gene was amplified by PCR using pET28a-ScleuS as a template, digested with BamHI and EcoRI, and then ligated into the p416GPD plasmid (predigested with BamHI and EcoRI) to generate p416GPD-ScleuS (leuS ϩ , ura ϩ ) (Fig. 2). The diploid yeast strain BY4743-ScleuS ϩ/Ϫ (Dharmacon), which contains only one chromosomal copy of leuS with the other replaced by a G418 resistance gene, was transformed with the rescue plasmid p416GPD-ScleuS. The transformants were cultured on synthetic dropout medium without uracil (SD/Ura Ϫ ). Dissection of spores and further genotype confirmation were performed as described previously (34). The spores that grew on the SD/Ura Ϫ plate and failed to grow on SD/Ura Ϫ /5-fluoroorotic acid plate were selected as the yeast leuS gene knock-out strain, which was designated Sc⌬leuS (Fig. 2).
To decrease the expression level, the promoter of p425TEF was replaced by that of endogenous ScleuS gene in a stepwise manner. First, the TEF promoter was deleted in one round of mutagenesis. Second, an 800-bp DNA fragment upstream of the yeast leuS open reading frame was amplified and ligated into the vector to obtain reconstructed shuttle vector p425ScPr. The gene encoding CaLeuRS was subcloned from pET28a-CaleuS and then inserted between the PstI and XhoI sites of p425ScPr to obtain p425ScPr-CaleuS. The gene encoding a His 6 tag was inserted at the 5Ј-terminus of CaleuS to obtain recombinant plasmid p425ScPr-His 6 -CaleuS. Mutations were engineered by site-directed mutagenesis using p425ScPr-His 6 -CaleuS as a template (Fig. 2). The constructs were introduced into Sc⌬leuS strain using the lithium acetate method. Transformants were selected on SD/Ura Ϫ /Leu Ϫ plates, and a single colony of each transformant was cultured in liquid SD/Leu Ϫ medium. To compare the growth rate of each transformant, yeast cultures were diluted to a concentration equivalent to A 600 ϭ 1, and a 10-fold dilution of the yeast was dropped onto the SD/Leu Ϫ /5-fluoroorotic acid plate to induce the loss of the rescue plasmid (p416GPD-ScleuS). Complementation was observed by comparing the growth rates of Sc⌬leuS expressing CaLeuRS or its different variants.
Preparation of Yeast Lysates and Western Blotting Assay-Yeast transformants were grown in 20 ml of liquid SD/Leu Ϫ medium. The yeast were then harvested; resuspended in icecold lysis buffer containing 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 1 mM EDTA (pH 8.0), 1% Triton X-100, 0.1% sodium deoxycholate, 1 mM PMSF, and 1 mM DTT; mixed with a few glass beads; vortexed rigorously for 10 s (three times at 1-min intervals); and then centrifuged at 10,000 ϫ g for 10 min. The supernatant was separated by SDS-PAGE (10% gel) and transferred to a PVDF membrane. The membrane was blocked with PBST (phosphate-buffered saline containing 0.05% Tween 20) with 5% nonfat dried milk for 3 h at room temperature. Membranes were hybridized overnight with anti-His 6 antibody (Abmart, M20001) or anti-␣-tubulin antibody (Cell Signaling Technology, 3873S) separately at 4°C. After incubating with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature, immunoreactivity was detected using LAS4000 (Fujifilm) and the SuperSignal West Pico Trial kit (Thermo Fisher Scientific). The results were quantified using Multi-Gauge Version 3.0 software. The total amount of protein was normalized according to that of ␣-tubulin, and the amount of CaLeuRS was designated as 100%.

Results
The CaCTD Is Indispensable for Leucylating Both CatRNA Ser -(CAG) and CatRNA Leu -Our previous study showed that only eukaryotic LeuRSs (e.g. CaLeuRS, ScLeuRS, and HsLeuRS) could charge CatRNA Ser (CAG); however, archaeal LeuRS (e.g. PhLeuRS) and bacterial/mitochondrial LeuRS (e.g. EcLeuRS and CamtLeuRS) could not (24). The archaeal and eukaryotic LeuRSs appear to be derived from the same ancestor with the most divergence in the CTD, implying that this region may be critical for recognizing CatRNA Ser . To address this question, we truncated the CTD of CaLeuRS (from Gly 894 to Glu 1097 ), based on the crystal structure of PhLeuRS (Protein Data Bank code 1WKB), obtaining a CTD-truncated mutant (CaLeuRS-⌬C) (Fig. 1B) (6). CaLeuRS-⌬C could be purified to high homogeneity and retained intact leucyl-adenylate synthesis activity (data not shown), suggesting that deletion of CTD had no direct effect on the secondary structure of CaLeuRS and the catalytic active site for amino acid activation. However, it totally lacked the leucylation activity for CatRNA Ser , indicating the indispensable role of the CTD in Leu-CatRNA Ser formation (Fig. 3, A and   B). Similar results were obtained in assays of the leucylation of CatRNA Leu by CaLeuRS-⌬C (Fig. 3C), supporting the hypothesis that the CTD was critical for aminoacylation of tRNA. To further determine the importance of CaCTD in leucylating CatRNA Ser and CatRNA Leu , we substituted CaCTD for PhCTD in PhLeuRS to obtain a mosaic enzyme, PhLeuRS-CaCTD. Compared with PhLeuRS, the mosaic enzyme gained the ability to leucylate CatRNA Ser and CatRNA Leu , further indicating the importance of CaCTD in capturing CatRNAs for aminoacylation (Fig. 3, D, E, and F).
PhLeuRS shows high homology with CaLeuRS (Fig. 4A), and simultaneous deletion of its C-terminal Ile 966 and Glu 967 completely abolishes the leucylation activity (10). Primary sequence alignment showed that Ile 1093 of CaLeuRS is highly conserved among yeast and archaeal LeuRSs and corresponds to Ile 966 of PhLeuRS (Fig. 4A). To explore whether CaLeuRS interacts with tRNAs in a similar manner to PhLeuRS, we performed progressive deletions from the C terminus of CaLeuRS, obtaining five mutants, CaLeuRS-⌬C1 to -⌬C5, all of which retained intact amino acid activation activity (data not shown). Further data showed that CaLeuRS-⌬C1 and -⌬C2 showed similar leucylation activity to the WT in the presence of CatRNA Ser ; CaLeuRS-⌬C3 and -⌬C4 gradually decreased their leucylation activity, and CaLeuRS-⌬C5 abolished the leucylation activity (Figs. 4B). Similar results were obtained in assays of the leucylation of CatRNA Leu by the five deletion mutants (Fig. 4C). However, the substitution of Ala for the five terminal residues (Glu 1097 , Val 1096 , Asn 1095 , Lys 1094 , and Ile 1093 ) of CaLeuRS showed similar leucylation activities to the native enzyme (data not shown), suggesting that either the interaction mode between CatRNA and residues at the C terminus of CaLeuRS is similar to that of the PhLeuRS-tRNA Leu system (10) or the C terminus is essential for maintaining proper conformation (the latter issue was subsequently addressed using the Sc⌬leuS strain). Overall, the results show that CaCTD is critical for leucylating both CatRNA Ser and CatRNA Leu ; however, the specific elements within CaCTD for charging tRNAs remain to be explored.
Identification of Important Residues within CaCTD-The above data showed that the C-terminal five residues, especially Ile 1093 , are important for the leucylation activity of CaLeuRS for both CatRNA Ser (CAG) and CatRNA Leu . Despite PhLeuRS having Ile 966 corresponding to Ile 1093 , it still failed to recognize both CatRNAs, indicating that the terminal five residues of CaLeuRS may act to affect the leucylation activity in a manner different from that of PhLeuRS. Therefore, CaCTD must harbor other specific elements for charging CatRNAs. Phylogenetic analysis of 13 LeuRSs from various yeast species revealed a high degree of conservation in their CTDs. To identify key residues within CaCTD that may interact with tRNA directly, 10 conserved basic residues (Arg 923 , Lys 997 , Lys 1007 , Arg 1009 , Arg 1021 , and 938 KKKKGK 943 ) were mutated to Ala, and the leucylation activities of these mutants were assayed. Additionally, we substituted the conserved Pro residues (Pro 963 , Pro 1086, and Pro 1089 ), Tyr 921 , and Gln 966 , which are either polar or potentially structure-crucial, with Ala (Fig. 5). Among the 15 Ala mutants, only CaLeuRS-R923A, -K938A, -K941A, -K1007A, and -P1086A showed marked decreases in leucylation activity  FEBRUARY 12, 2016 • VOLUME 291 • NUMBER 7 JOURNAL OF BIOLOGICAL CHEMISTRY 3617 compared with the WT (Table 1). To further investigate their mode of interaction with tRNA, we replaced the five residues with Asp, obtaining CaLeuRS-R923D, -K938D, -K941D, -K1007D, and -P1086D. The residue Arg 923 may contribute to maintaining stability/conformation of the enzyme because CaLeuRS-R923D mainly produced inclusion bodies during expression of its gene in E. coli. All the other four mutants retained intact Leu activation activity comparable with the WT (data not shown), indicating that these mutations had no effect on the catalytic active site. In the subsequent aminoacylation reaction, we assayed and compared the kinetic constants of the WT and the mutants (Table 1). CaLeuRS-K938D, -K941D, -K1007D, and -P1086D all showed increased K m values (5.43, 10.95, 12.36, and 12.08 M, respectively) compared with the WT (2.01 M), showing that the affinity between the mutants and tRNA was disrupted. In addition, the k cat values of the four mutants decreased to different extents, accounting for 33, 63, 18, and 25% of that of the WT, respectively. Because of the increased K m and decreased k cat values, their catalytic efficiencies (k cat /K m ) were approximately 12, 12, 3, and 4% of that of the WT, respectively. Substituting of these three lysine residues with acidic Asp affects the tRNA Leu -leucylating activity, indicating that the positive charge of the three lysines may be important during the leucylation process. This hypothesis was confirmed by the data that the Gln mutants (CaLeuRS-K938Q, -K941Q, and -K1007Q) also showed decreased catalytic efficiency (about 53, 29, and 38% of that of the WT, respectively) although not as obvious as that of their corresponding Asp mutants (Table 1). And the K m values of all three Gln mutants increased (3.33, 5.54, and 4.01 M, respectively), indicating a reduction of the affinity between the mutants and tRNA (Table  1).

Recognition of CatRNA Ser and CatRNA Leu by CaLeuRS
To explore whether the residues important for Leu-tRNA Leu synthesis also affect Leu-tRNA Ser formation, we investigated leucylation of CatRNA Ser by CaLeuRS-K938D, -K941D, -K1007D, and -P1086D. All the mutants showed obvious decreased k obs values (0.95 ϫ 10 Ϫ4 , 0.14 ϫ 10 Ϫ3 , 0.25 ϫ 10 Ϫ4 , and 0.28 ϫ 10 Ϫ4 s Ϫ1 ) in CatRNA Ser leucylation activity, representing about 50, 70, 10, and 10% of that of the WT (0.20 ϫ 10 Ϫ3 s Ϫ1 ), respectively (Fig. 6A). However, their corresponding Ala mutants, CaLeuRS-K938A, -K941A, and -P1086A, showed a subtle decrease in the CatRNA Ser -charging activity (Fig. 6B), indicating that introduction of negatively charged residues at Lys 938 , Lys 941 , and Pro 1086 hindered the formation of Leu-tRNA Ser . However, CaLeuRS-K1007A still lost almost half of the CatRNA Ser -charging activity (k obs ϭ 0.14 ϫ 10 Ϫ3 s Ϫ1 ) compared with the WT (Fig. 6B). These results indicated that the basic residues (Lys 938 , Lys 941 , and Lys 1007 ) contribute significantly to the leucylation of both CatRNA Leu and CatRNA Ser , affecting the affinity of enzyme for the tRNA and/or the efficiency of transition in the aminoacylation reaction, which further explains their high conservation within the CTD of yeast LeuRSs.
Lys 938 , Lys 941 , and Lys 1007 Made an Additive Contribution to Leucylation-We identified three basic residues, Lys 938 , Lys 941 , and Lys 1007 , that were implicated as tRNA-interacting elements within the CTD based on the positive charge of their side chains. However, none of their Asp mutants completely lost the    and -K941D/K1007D were even completely devoid of the ability to charge both CatRNA Ser and CatRNA Leu , emphasizing the additive effect of the three lysine residues in aminoacylation (Fig. 7, A and B).
The reduced aminoacylation activity of these mutants was possibly due to their decreased tRNA binding capacity. To verify this possibility, we performed gel mobility shift assays to compare the dissociation constants (K d ) of the interactions between the WT, CaLeuRS-K938D/K941D, -K938D/K1007D, -K941D/K1007D, or -K938D/K941D/K1007D and tRNAs. A shifted band for the CaLeuRS-tRNA Ser complex was observed in the presence of 1 M CaLeuRS or more with a calculated K d value of 2.7 M (Fig. 8, A and B). However, under the same conditions, no shifted band was observed for CaLeuRS-K938/ 941D, -K938D/K1007D, -K941D/K1007D, and -K938D/K941D/ K1007D with tRNA Ser even in the presence of 12 M enzyme (Fig. 8A), suggesting that the double and triple site mutations disrupted the formation of the enzyme-tRNA Ser complex. A similar shifted band was observed with tRNA Leu with a calculated K d value of 2.6 M (Fig. 8, C and D); however, no shifted band was observed for CaLeuRS-K938D/K941D, -K938D/ K1007D, -K941D/K1007D, and -K938D/K941D/K1007D with tRNA Leu (Fig. 8C). Thus, it is postulated that the decrease or loss of leucylation ability by these mutants may be partially due to the reduction of the tRNA binding ability during catalysis. Furthermore, it is notable that CaLeuRS binds tRNA Ser and tRNA Leu with similar affinities in vitro despite the divergence of the main bodies of tRNA Ser and tRNA Leu , indicating the importance of the anticodon arm/loop in the interaction with CaLeuRS.
The Counterparts of Lys 1007 in Other Eukaryotic LeuRSs Are Also Important for Leucylating tRNA-The K1007D mutation showed the most severe impact on the leucylation activity. Primary sequence alignment revealed that Lys 1007 is highly conserved in eukaryotic LeuRSs, ranging from yeast species such as S. cerevisiae and Candida tropicalis to higher organisms, such as H. sapiens, Mus musculus, and Xenopus laevis. The counterparts of CaLeuRS-Lys 1007 are identified as Lys 1002 in ScLeuRS and Lys 997 in HsLeuRS, respectively (Fig. 9A). To assess any functional conservation of this site, we replaced them with Asp in ScLeuRS and HsLeuRS to obtain ScLeuRS-K1002D and HsLeuRS-K997D, respectively. In the presence of CatRNA Ser , the k obs value of ScLeuRS-K1002D (0.35 ϫ 10 Ϫ3 s Ϫ1 ) was about 40% of that of ScLeuRS (0.92 ϫ 10 Ϫ3 s Ϫ1 ), and that of HsLeuRS-K997D (0.11 ϫ 10 Ϫ2 s Ϫ1 ) was about 70% of that of HsLeuRS (0.16 ϫ 10 Ϫ2 s Ϫ1 ) (Fig. 9B). Further assay of their tRNA Leucharging activity showed similar results. The k obs value of ScLeuRS-K1002D (0.03 s Ϫ1 ) was approximately 10% of that of ScLeuRS (0.26 s Ϫ1 ) (Fig. 9C), and that of HsLeuRS-K997D (0.28 s Ϫ1 ) was about 40% of that of HsLeuRS (0.67 s Ϫ1 ) (Fig. 9D). These data showed that the counterparts of CaLeuRS-Lys 1007 in ScLeuRS and HsLeuRS are also important for the recognition of both CatRNA Ser (CAG) and their cognate tRNA Leu s by CTD, emphasizing the importance of this highly conserved basic residue for the recognition function of CTD.
Confirmation of tRNA Binding Capacity Using Sc⌬leuS-Previous studies showed that CaLeuRS was able to rescue an Sc⌬leuS strain because CaLeuRS showed high sequence homology with ScLeuRS (64.9%) and could aminoacylate SctRNA Leu efficiently (34,35). Hence, the Sc⌬leuS strain is a good tool to test the in vivo effect of residues identified in in vitro screens.
CaLeuRS-Leu 919 and CaLeuRS-Ser 919 are two isoforms differing at position 919, and CaLeuRS-Leu 919 has a higher catalytic efficiency for CatRNA Leu in the aminoacylation assay (24). Here, both isoforms supported Sc⌬leuS strain growth to a similar level (Fig. 10A). However, by comparing steadystate protein levels of CaLeuRS-Leu 919 and -Ser 919 from yeast transformants containing their genes, we found that the amount of CaLeuRS-Leu 919 was only 32% of that of CaLeuRS-Ser 919 , indicating that the insertion of Ser or Leu at position 919 may affect gene expression or protein stability (Fig. 10, C and D).
The growth of Sc⌬leuS containing CaLeuRS-⌬C1, -⌬C2, -⌬C3, and -⌬C4 was similar to that containing CaLeuRS-Ser 919 (Fig. 10A). However, the growth of Sc⌬leuS containing CaLeuRS-⌬C5 was severely disrupted (Fig. 10A). Western blotting showed that the genes encoding all the truncated proteins were expressed to a similar level in yeast (Fig. 10, C and D). Thus, the inability of CaLeuRS-⌬C5 to support Sc⌬leuS viability was probably caused by the total loss of aminoacylation capacity, not by non-or decreased expression. Compared with CaLeuRS-Ser 919 , the five deletion mutants were expressed at lower levels, suggesting that the terminal five residues may act to stabilize the conformation of CaCTD and affect protein expression.

JOURNAL OF BIOLOGICAL CHEMISTRY 3621
cially CaLeuRS-K938D/K1007D (Fig. 10B). The Sc⌬leuS containing CaLeuRS-K938D/K941D/K1007D could not grow on SD/Leu Ϫ /5-fluoroorotic acid plates, suggesting that the triple site mutant had totally lost tRNA binding ability even when present in sufficient amount. All the results indicated the significance of Lys 938 , Lys 941 , and Lys 1007 in binding tRNA in vivo, which is consistent with the results obtained in vitro.

Discussion
CTDs from Three Kingdoms Are Essential for Recognizing tRNA Leu -aaRSs aminoacylate their cognate tRNAs to produce materials for protein biosynthesis. Structural studies have shown that they evolved N-/C-domains appended to the catalytic center to capture tRNAs for efficient and accurate aminoacylation. For example, the N-terminal domain, containing lysine-rich motifs, significantly enhances tRNA binding in yeast aspartyl-tRNA synthetase and mammalian lysyl-tRNA synthetase (36,37). The CTDs of Staphylococcus aureus isoleucyl-tRNA synthetase, T. thermophilus valyl-tRNA synthetase, and E. coli histidyl-tRNA synthetase are also responsible for recognizing tRNA (38 -40).
Generally, LeuRS are classified into two types (6 -8). In fact, based on their divergent CTDs, they are further classified into three types (bacterial, archaeal, and eukaryotic LeuRSs) (6,7). CTDs of both bacterial and archaeal LeuRSs are responsible for recognizing tRNA Leu as negligible leucyl-tRNA Leu synthesis was observed in the CTD-deleted mutants of EcLeuRS, TtLeuRS, PhLeuRS, and NmLeuRS (7,10,26,28). However, their concrete recognition mechanisms are divergent. A yeast three-hybrid system and band shift assay showed that the CTD of the ␤-subunit of Aquifex aeolicus LeuRS was responsible for binding tRNA both in vivo and in vitro; however, the recognition elements remained unclear (41). Although several conserved, positively charged residues were identified within the CTD of EcLeuRS, none of them provided site-specific interaction with tRNA Leu (28). Asp scanning of the MtLeuRS CTD identified several residues (Val 914 , Leu 949 , Gln 915 , and Leu 964 ) that contributed to maintain the hydrophobic environment to orient tRNA Leu in the correct aminoacylation conformation (27). By contrast, based on the crystal structure of PhLeuRS-tRNA Leu complex (Protein Data Bank code 1WZ2), some residues (Asp 845 , Ile 849 , Pro 962 , Ile 964 , Ile 966 , and Glu 967 ) within the CTD of PhLeuRS formed base-specific interactions with tRNA Leu mainly through van der Waals interactions or hydrogen bonds (10).
Our present study, for the first time, used CaLeuRS as a eukaryotic LeuRS model, revealing that the CTD from a eukaryotic LeuRS is also indispensable for recognizing tRNA Leu . Notably, three basic residues (Lys 938 , Lys 941 , and Lys 1007 ) within CaCTD were important during the leucylation process; their positively charged side chains may interact with tRNA, consistent with the Arg 921 of bacterial MtLeuRS and Lys 961 of archaeal PhLeuRS (10,27). In addition to the three key lysine residues, the C-terminal five residues of CaLeuRS may maintain the proper conformation of the CTD during leucylation but not interact with tRNA Leu directly, which contrasts with archaeal PhLeuRS (10). In conclusion, although CTDs from the three kingdoms are all essential for recognizing tRNA Leu , the recognition mechanisms have evolved to be distinct, allowing more specific and efficient discrimination of their different cognate tRNAs.
CaLeuRS Recognizes CatRNA Ser (CAG) and CatRNA Leu Using the Same Elements-CaLeuRS is unique because it recognizes two types of tRNAs (cognate CaRNA Leu and non-cognate CatRNA Ser (CAG)) (24). In S. cerevisiae, substituting the anticodon arm of SctRNA Ser with that of SctRNA Leu endowed it with leucine charging capacity, emphasizing that the anticodon arm/loop of SctRNA Leu plays a dominant role in recognition by ScLeuRS (19). CatRNA Ser (CAG) is a naturally chimeric tRNA with the main body of tRNA Ser and the anticodon of tRNA Leu (13,14). Mutation analysis of tRNA determinants showed that the anticodon loops/arms of both CatRNA Ser (CAG) and CatRNA Leu are crucial for recognition by CaLeuRS (13, 19).
In our study, Lys 938 , Lys 941 , and Lys 1007 within CaCTD contribute to recognizing both CatRNA Leu and CatRNA Ser (CAG), suggesting a similar interaction mode between the CaLeuRS and CatRNA Leu /CatRNA Ser in the aminoacylation state. In addition, the affinities between the CaLeuRS and CatRNA Leu / CatRNA Ser were similar in vitro. All of these observations were consistent with the mutation analysis of the tRNA determinants that indicated the crucial importance of the anticodon loop/arm of CatRNA Ser and CatRNA Leu in the recognition by LeuRS (13,19). In addition, both CatRNA Ser and CatRNA Leu belong to class II tRNAs with a long variable arm, sharing similar tertiary structures. Therefore, the hypothesis that CaLeuRS recognizes the two tRNAs using the same elements is reasonable. Although the recognition elements within CaCTD for CatRNA Ser (CAG) and CatRNA Leu are similar, CatRNA Ser -(CAG) may still possess some antideterminants for CaLeuRS, leading to tRNA Ser being a poorer charging tRNA compared with tRNA Leu , contributing to the low abundance of Leu in CUG sites.
Lys 938 , Lys 941 , and Lys 1007 Affect the Binding Capacity of CaLeuRS for Both CatRNA Ser (CAG) and CatRNA Leu with Additive Effect-Although the key recognition elements of bacterial or archaeal CTD have been elucidated, neither single nor double/triple site mutants lost the aminoacylation activity completely (28). In this study, we identified three lysine residues (Lys 938 , Lys 941 , and Lys 1007 ) in CaLeuRS and first verified that their double/triple site mutants had more severe disruption or even destruction of the leucylation activity compared with the single site mutants, differing from that of EcLeuRS (28).
Furthermore, the gel mobility shift assays showed that CaLeuRS-K938D/K941D, -K938D/K1007D, -K941D/K1007D, and -K938D/K941D/K1007D could not form obvious enzyme-tRNA Leu shifted bands with incalculable K d values, indicating the importance of the three lysine residues in binding tRNA. However, in vivo complementation assays showed that the three double site mutants still supported yeast growth to some extent. This was probably caused by the double site mutants also binding yeast tRNA Leu in vivo because the K d value may not be accurately captured/reflected by the gel mobility shift assay, which was measured under limited enzyme concentration. Besides, it is possible that expression of double site mutants by the high copy p425ScPr plasmid would produce excess protein to complement the impaired tRNA binding capacity of the mutants. Hence, despite the fact that the double site mutants showed negligible aminoacylation activity and K d in vitro, they could still complement yeast growth to varying degrees. However, both the in vitro and in vivo data supported the view that the combination of the triple site mutant (CaLeuRS-K938D/ K941D/K1007D) had completely lost the aminoacylation activity, suggesting an additive effect of these lysine residues on leucylating CatRNA Leu .
The Leucylating Capacity of the Mutants Was Further Confirmed Using Sc⌬leuS-Our research focused on exploring the interaction between CaLeuRS and CatRNA Ser /CatRNA Leu . The in vitro data showed that Lys 938 , Lys 941 , and Lys 1007 are important for leucylating both tRNAs. As CaLeuRS can com-plement ScLeuRS in vivo (34), we used an Sc⌬leuS strain in which the gene encoding LeuRS was deleted to investigate the catalytic function of the various mutants mentioned in vitro.
In C. albicans, CaLeuRS has two isoforms, CaLeuRS-Ser 919 and -Leu 919 , because the CUG codon at position 919 could be decoded by CatRNA Ser (CAG) as Ser or Leu. CaLeuRS-Ser 919 is the main form; therefore, we used it as the wild type in the in vitro aminoacylation assay. However, in the complementation assay, both isoforms were introduced into the Sc⌬leuS to compare their expression levels and ability to rescue the yeast. In our study, although there was a 68% decrease in the relative expression level of CaLeuRS-Leu 919 compared with that of CaLeuRS-Ser 919 , Sc⌬leuS transformed with the gene encoding CaLeuRS-Leu 919 grew slightly better than that transformed with CaLeuRS-Ser 919 , which reflects the higher catalytic efficiency of CaLeuRS-Leu 919 compared with CaLeuRS-Ser 919 (24). This strict regulation of the relative amounts of CaLeuRS-Ser 919 /CaLeuRS-Leu 919 might balance the CUG decoding with CaSerRS-Ser 919 /CaSerRS-Leu 919 as well as relieve the potential effect on phenotypic diversity caused by excessive Leu misincorporation (16). Moreover, mutants of CaLeuRS with low aminoacylation activity (such as CaLeuRS-K938D, -K941D, -K1007D, -P1086D, and -K938D/K941D) complemented the loss of ScLeuRS as well as the native enzyme, suggesting that mutants exogenously expressed from high copy plasmids were sufficient to support yeast growth although with minimal activity. A similar phenotype was also observed in complementation of a yeast threonyl-tRNA synthetase knock-out strain (42).
Sc⌬leuS containing CaLeuRS-K938D/K1007D and -K941D/ K1007D showed a growth-retarded phenotype, and Sc⌬leuS containing CaLeuRS-K938D/K941D/K1007D showed a lethal phenotype, which further emphasized the significance and additive effect of Lys 938 , Lys 941 , and Lys 1007 in leucylating tRNA in vivo. In addition, CaLeuRS-K938D/K941D/K1007D and -⌬C5 could not rescue Sc⌬leuS, which suggests that both the three lysine residues and the terminal five residues are indispensable for leucylation of tRNA through either direct interaction or maintenance of the proper conformation. Therefore, we believe that the CTD acts to recognize tRNA through complex interactions.
Concluding Remarks-C. albicans, an opportunistic pathogen, has evolved a chimeric CatRNA Ser (CAG) that leads to protein translation ambiguity. The resultant ambiguity at the CTG codon is crucial in the morphological switch and virulence of this pathogen. The elements in CaLeuRS that mediate this crucial genetic code ambiguity were undetermined. In this study, we have provided the first evidence and clarified the interaction mechanism of the indispensable function of the CTD of CaLeuRS in recognizing both CatRNA Ser and CatRNA Leu . Our results deepen our understanding of the mechanism that meditates CTG codon ambiguity in C. albicans and improves our knowledge concerning tRNA Leu recognition by eukaryotic LeuRSs.