| Nucleic Acids Research | Pages |
Efficient aminoacylation of tRNALys,3 by human lysyl-tRNA synthetase is dependent on covalent continuity between the acceptor stem and the anticodon domain
Introduction
Materials And Methods
Protein purification and assays
Ribonucleic acids
Native polyacrylamide gel electrophoresis
Results
Anticodon mutations
Aminoacylation of minihelixLys,3
Aminoacylation of semi-synthetic tRNALys,3
Discussion
References
Efficient aminoacylation of tRNALys,3 by human lysyl-tRNA synthetase is dependent on covalent continuity between the acceptor stem and the anticodon domain
Received July 28, 1999; Revised September 15, 1999; Accepted October 25, 1999
ABSTRACT In this work, we probe the role of the anticodon in tRNA recognition by human lysyl-tRNA synthetase (hLysRS). Large decreases in aminoacylation efficiency are observed upon mutagenesis of anticodon positions U35 and U36 of human tRNALys,3. A minihelix derived from the acceptor-T[Psi]C stem-loop domain of human tRNALys,3 was not specifically aminoacylated by the human enzyme. The presence of an anticodon-derived stem-loop failed to stimulate aminoacylation of the minihelix. Thus, covalent continuity between the acceptor stem and anticodon domains appears to be an important requirement for efficient charging by hLysRS. To further examine the mechanism of communication between the critical anticodon recognition elements and the catalytic site, a two piece semi-synthetic tRNALys,3 construct was used. The wild-type semi-synthetic tRNA contained a break in the phosphodiester backbone in the D loop and was an efficient substrate for hLysRS. In contrast, a truncated variant that lacked nucleotides 8-17 in the D stem-loop displayedseverely reduced catalytic efficiency. The elimination of key tRNA tertiary structural elements has little effect on anticodon-dependent substrate binding but severely impacts formation of the proper transition state for catalysis. Taken together, our studies provide new insights into human tRNA structural requirements for effective transmission of the anticodon recognition signal to the distal acceptor stem domain.
INTRODUCTION
The relationship between aminoacyl-tRNA synthetases and tRNAs establishes the genetic code whereby specific amino acids are esterified to cognate anticodon-bearing tRNAs during translation. All tRNAs have similar secondary and tertiary structures. Synthetase interaction with the cognate tRNA is based on the presence of identity elements within the tRNA that allow for specific binding and catalysis (1). Both the bacterial and yeast recognition systems have been extensively studied. In general, tRNA identity elements are contained primarily within the acceptor stem and anticodon domains (1). These two domains are located at opposite ends of the L-shaped tRNA tertiary structure (2). Precisely how communication between the critical anticodon recognition elements and the catalytic site located ~76 Å away occurs is in general an open question for the synthetases, and undoubtedly involves a complex set of protein-RNA interactions outside the primary recognition sites (1).
Based on both primary sequence and structural data the aminoacyl-tRNA synthetases have been divided into two major classes each containing 10 members (3). The two classes of synthetases have been further organized into distinct subclasses. Class II aminoacyl-tRNA synthetases are characterized by a highly degenerate set of sequence elements known as motifs 1, 2 and 3. The enzymes that belong to subclass IIb (asparaginyl-, aspartyl- and lysyl-tRNA synthetases) are further distinguished by an N-terminal oligonucleotide-binding (OB) fold that recognizes anticodons with uridine at the second position (U35) (4). The specific molecular interactions between this N-terminal domain and the anticodon have been characterized using biochemical and structural methods for yeast aspartyl-tRNA synthetase (AspRS) (5-7) and both Escherichia coli and Thermus thermophilus lysyl-tRNA synthetases (LysRS) (8-10). Moreover, in the case of E.coli LysRS, mutagenesis of anticodon nucleotides 35 and 36 resulted in [ge]100-fold decrease in in vitro aminoacylation efficiency, whereas mutagenesis of U34 to G had a less dramatic effect (17-fold) on activity (11).
In the case of yeast AspRS, in addition to the anticodon, the acceptor stem contains important recognition elements that are known to make specific contacts with the enzyme (5,7). In particular, mutations at the so-called `discriminator' base located at position 73 result in decreases in in vitro catalytic efficiency that are similar in magnitude to those observed upon anticodon mutagenesis (6). In this system, a minihelix derived from the acceptor stem domain is a substrate for aminoacylation, albeit with a kcat that is reduced ~9000-fold relative to full-length tRNAAsp (12). Attempts to further stimulate minihelixAsp aminoacylation by addition of a separate anticodon stem-loop were unsuccessful (12). Thus, in this subclass IIb yeast system, covalent attachment of the anticodon domain to the acceptor domain appears to be a requirement for effective communication between these two key sites.
In contrast to the yeast AspRS system, previous investigations of another subclass IIb synthetase, namely human LysRS (hLysRS), showed that it is relatively insensitive to the identity of the discriminator base (13). Based on our knowledge of tRNA recognition by subclass IIb synthetases and bacterial LysRS in particular, we hypothesized that the anticodon nucleotides U35 and U36 would be critical for hLysRS recognition. Moreover, all residues observed to interact with the central U35 base of the anticodon in the known crystal structure of T.thermophilus LysRS complexed with tRNALys are conserved in the human enzyme (9,13). In the present study, we confirm the importance of the anticodon elements and investigate the mechanism of anticodon-dependent aminoacylation by a human synthetase for the first time. In particular, we elucidate tRNA structural requirements for effective communication between the critical anticodon nucleotides and the catalytic site.
MATERIALS AND METHODS
Protein purification and assays
Two forms of hLysRS were used in this study. Plasmid pM131 encodes an N-terminal truncated form comprising amino acids 66-597 (hLysRS-[Delta]N65) and was purified from E.coli strain pM131/PAL[Delta]S[Delta]UTR(DE3) that is also devoid of E.coli LysRS activity (13). Although this variant lacks the N-terminal 65 amino acid residues it still contains residues that constitute the entire N-terminal [beta]-barrel OB-fold domain. Plasmid pM368 encodes full-length hLysRS (amino acids 1-597) and an additional N-terminal sequence that includes a His6 tag (6H-hLysRS) and was purified from protease-deficient E.coli strain DE3. The proteins were purified as previously described (13). Enzyme concentrations were estimated by the Bradford method using the Bio-Rad protein assay kit and bovine serum albumin as the standard (14).
Aminoacylation assays using hLysRS were conducted at 30°C as previously reported (13). For wild-type human tRNALys,3, assays contained 12.5 nM LysRS and tRNA concentrations of 0.25-18 µM. Mutant tRNAs were assayed using 50 nM LysRS and tRNA concentrations of 0.25-7 µM (U36C), 1-12 µM (U35C, U36G and U36A) and 10-25 µM (U35A and U35G). Aminoacylation assays of native human tRNALys (partially purified) were carried out using 1.75 nM LysRS and tRNA concentrations of 0.0125-0.25 µM. Native tRNALys concentrations were estimated by carrying out plateau-level charging of the tRNA. Two sets of assay conditions were used with the semi-synthetic tRNA construct.When the 17mer and 15mer 5[prime]-oligonucleotides were used, the assays were carried out as for wild-type human tRNALys,3. The truncated semi-synthetic tRNA constructs prepared with a 5[prime]-14mer, 10mer or 7mer were assayed at 20°C using 50 nM LysRS. The semi-synthetic tRNAs were annealed by mixing a 1.5-fold excess of the 5[prime]-fragment with the 3[prime]-59mer and heating at 60°C for 3 min. MgCl2 was added to 10 mM and the sample was incubated at room temperature for 5 min before placing on ice. The minihelixLys,3 aminoacylation reactions contained 0.125-1.25 µM human LysRS and minihelixLys,3 concentrations varying from 8 to 200 µM. For minihelixLys,3 stimulation attempts an RNA stem-loop mimicking the tRNALys,3 anticodon domain was added to the aminoacylation reactions at concentrations varying from 20 to 100 µM. Inhibition assays were carried out using 2.5 nM hLysRS and unmodified tRNALys,3 concentrations of 0.25-8 µM. Concentrations of the inhibitors tested were as follows: 0-8 µM 3[prime]-oxidized tRNALys,3; 0-60 µM U35G tRNALys,3; 0-100 µM minihelixLys,3; 0-20 µM UUU anticodon stem-loop; 0-100 µM U35G anticodon stem-loop. Dixon plots were produced and the Ki values were determined from replots of the slopes versus 1/[tRNA] (15).
T7 RNA polymerase was purified according to Grodberg and Dunn (16)from E.coli strain BL-21/pAR 1219, which was a gift from Dr F. William Studier.
Ribonucleic acids
Unmodified wild-type human tRNALys,3 was prepared by in vitro transcription from FokI-linearized plasmid pLYSF119 as previously described (13). Mutant tRNALys,3 variants as well as minihelixLys,3 derived from the acceptor-T[Psi]C stem-loop domain of tRNALys,3 were prepared by overlap extension PCR mutagenesis (17) and verified by sequencing (18). The gene for 3[prime]-59mer tRNALys preceded by a T7 RNA polymerase promoter was assembled by ligating together six deoxyoligonucleotides as previously described (19). This DNA insert was then ligated into EcoRI/BamHI-digested pLYSF119. DEAE-grade human tRNALys was purchased from Bio S and T (Montreal, Canada). Synthetic RNA oligonucleotides mimicking the human tRNALys,3 anticodon stem-loop domain and 5[prime]-regions of the D loop and acceptor stems were prepared by automated solid phase chemical synthesis on a Gene Assembler Plus (Pharmacia) using the phosphoramidite method (20).
To determine concentrations of RNAs at 260 nm the following extinction coefficients were used: 5[prime]-7mer, 6.8 × 104 M-1; 5[prime]-10mer, 8.8 × 104 M-1; 5[prime]-14mer, 11.6 × 104 M-1; 5[prime]-15mer, 12.3 × 104 M-1; 5[prime]-17mer, 13.6 × 104 M-1; anticodon stem-loops (17mer), 13.6 × 104 M-1; minihelixLys,3 (35mer), 26.0 × 104 M-1; 3[prime]-59mer, 42.5 × 104 M-1; full-length tRNAs, 60.4 × 104 M-1.
Native polyacrylamide gel electrophoresis
Native gels consisted of a resolving gel containing 12% acrylamide (29:1 acrylamide:bisacrylamide), 25 mM Tris, 0.12 mM glycine (pH 7.8) and 10 mM MgCl2. The stacking gel contained 5% acrylamide (29:1 acrylamide:bisacrylamide), 25 mM Tris, 0.12 mM glycine (pH 5.8) and 10 mM MgCl2. RNA samples were annealed using the protocols described above and glycerol was added to 25% just before loading. Approximately 7 µg of annealed RNA was applied to each lane. Gels were run at 30 mA for 24 h at room temperature and stained with ethidium bromide for visualization.
RESULTS
We have previously shown that in vitro transcripts of human tRNALys,3 are substrates for both the N-terminal truncated and full-length forms of hLysRS. Both forms of hLysRS were able to efficiently aminoacylate both native and unmodified E.coli tRNALys indicating that the base modifications are not critical for cross-species aminoacylation (13). In the present study, we used partially purified native human tRNALys in aminoacylation assays to examine the effect of base modifications on cognate charging by hLysRS. The kcat/Km shows a 10-fold decrease for the unmodified transcript compared to the native human tRNALys, indicating that the base modifications make a modest contribution to recognition of human tRNALys by hLysRS.
Anticodon mutations
Figure 1 shows the structure of unmodified human tRNALys,3 and the individual mutations at the last two positions of the anticodon that were prepared and tested using hLysRS-[Delta]N65. The results show that mutations at either of these two positions have a significant effect on aminoacylation efficiency, with decreases of 150- to >3000-fold at position 35, and 11- to 180-fold at position 36. The most severe effect was observed upon U35G substitution, which resulted in a complete loss of aminoacylation activity. In general, purine substitutions at both positions resulted in more severe decreases than pyrimidine substitutions (Fig. 1 and Table 1). As shown in Table 1, whereas pyrimidine substitution had little effect on Km and resulted in moderate decreases in kcat, purine substitutions affected both kcat and Km. Since no aminoacylation activity was detected for the U35G variant, the effect of this substitution on synthetase binding was examined by carrying out inhibition studies. As shown in Table 2, whereas a 3[prime]-periodate-oxidized tRNALys,3 variant is an excellent inhibitor of tRNALys,3 aminoacylation with a Ki of 1.3 µM, the U35G variant has an elevated Ki (42 µM). Thus, this variant is significantly defective in both binding and catalysis.
Figure 1. Secondary structure of unmodified human tRNALys,3 used in this study. The arrows point to individual mutations at positions 35 and 36 that were tested in this work. The numbers represent the fold decrease in kcat/Km relative to the wild-type unmodified tRNALys,3.
Table 1. Kinetic parameters determined from in vitro aminoacylation of human tRNALys,3 variants using hLysRS-[Delta]N65
| tRNALys,3 variant | kcat (s-1) | Km (µM) | kcat/Km (relative) | Fold decrease |
| Wild-type | 0.38 | 3.4 | 1 | |
| U35A | ND | ND | 0.0003 | ~3000 |
| U35C | 0.0079 | 7.8 | 0.0065 | 153 |
| U35G | - | - | - | >3000 |
| U36A | 0.0068 | 9.5 | 0.0055 | 182 |
| U36C | 0.046 | 3.7 | 0.093 | 11 |
| U35G | 0.012 | 16 | 0.0056 | 179 |
Table 2. Inhibition constants determined from in vitro aminoacylation of unmodified human tRNALys,3 using 6H-hLysRS
| Inhibitor | Ki (µM) |
| tRNALys,3 (3[prime]-oxidized) | 1.3 |
| tRNALys,3 (U35G) | 42 |
| Anticodon stem-loop (WT) | 5 |
| Anticodon stem-loop (U35G) | 33 |
Although the kinetic parameters reported in Table 1 were determined using the N-terminal truncated form of hLysRS, two of the mutant tRNA variants (U35G and U36G) were tested with the full-length his-tagged enzyme and the results obtained were very similar (data not shown).
Aminoacylation of minihelixLys,3
Acceptor-T[Psi]C stem-loop-derived minihelices are substrates for many aminoacyl-tRNA synthetases despite the fact that they lack the anticodon and D stem-loop domains (21). Even in cases where the anticodon is a major recognition element, minihelices have been shown to be specifically aminoacylated. However, very few human synthetases have been tested for minihelix aminoacylation to date. We prepared the minihelix corresponding to the acceptor-T[Psi]C stem-loop domain of human tRNALys,3 (Fig. 2) and tested it in aminoacylation assays with hLysRS. The minihelix was not detectably aminoacylated by the human enzyme even when high concentrations of RNA (200 µM) and hLysRS (1 µM) were included in the assay. The minihelix weakly inhibited aminoacylation of tRNALys,3 (data not shown). Based on inhibition experiments we estimate that the Ki is [ge]100 µM; however, due to the relatively weak binding, it was only possible to determine a lower limit for the Ki value.Thus, the lack of charging even at high minihelix concentrations may, in part, be attributed to a lower binding affinity.
Figure 2. Structure of stem-loop domains derived from human tRNALys,3 used in this work. The anticodon stem-loop domain (boxed) was chemically synthesized while the acceptor-T[Psi]C stem-loop minihelixLys,3 (shaded) was prepared by in vitro transcription.
We next determined whether the addition of an RNA stem-loop mimicking the anticodon domain of tRNALys,3 (Fig. 2) could stimulate minihelixLys,3 aminoacylation. We first established that the wild-type anticodon was a good inhibitor of tRNALys,3 aminoacylation, with a Ki of 5 µM (Table 2), indicating that binding to LysRS was similar to the full-length tRNA. Moreover, the specificity of anticodon binding is maintained in the context of the RNA stem-loop, since a U35G anticodon stem-loop variant is a significantly weaker inhibitor with a Ki of 33 µM (Table 2). Nevertheless, we were unable to observe any detectable minihelixLys,3 aminoacylation even upon addition of 100 µM anticodon stem-loop domain.
Aminoacylation of semi-synthetic tRNALys,3
Aminoacylation experiments with the minihelixLys,3 and anticodon stem-loops suggest that covalent continuity is required for communication between the critical anticodon elements and the site of aminoacylation. To elucidate whether the entire tRNA structure is required for the anticodon to trigger aminoacylation of the acceptor stem domain, semi-synthetic tRNA constructs were prepared. The first semi-synthetic construct tested contained a break in the sugar-phosphate backbone between nucleotides 17 and 18 of the D loop (Fig. 3). A similar construct was shown to be active in the E.coli ProRS system (22). Aminoacylation assays with hLysRS showed that this semi-synthetic tRNALys,3 (3[prime]-59mer + 5[prime]-17mer) was indeed an efficient substrate, with a kcat/Km within a factor of 2 of the wild-type tRNA (Table 3 and Fig. 4). In contrast, the 3[prime]-59mer alone (Fig. 3, top right), which contains the CCA 3[prime]-end, was not aminoacylated by the synthetase (Fig. 4). A second semi-synthetic tRNA was prepared using only a 5[prime]-7mer oligomer annealed to the 3[prime]-59mer (Fig. 3). This construct maintains covalent continuity between the acceptor stem and anticodon domains but disrupts the tRNA tertiary core structure by deletion of nucleotides 8-17. This construct was a very poor substrate for the synthetase with a kcat/Km that was 2700-fold reduced relative to the intact tRNA (Table 3 and Fig. 4). Comparison of the individual kinetic parameters shows that the decreased efficiency of the truncated semi-synthetic construct is primarily due to a 6500-fold decrease in kcat, while the Km is 2-fold lower (Table 3). Upon introduction of the U35G mutation into the 3[prime]-59mer fragment the aminoacylation activity of both semi-synthetic tRNA constructs was eliminated (data not shown). Thus, despite maintenance of anticodon binding specificity, the truncated semi-synthetic construct was severely defective in formation of the transition state for catalysis.
Figure 3. Scheme showing preparation of semi-synthetic tRNAs. A chemically synthesized 5[prime]-oligonucleotide (17mer, indicated in larger font and italics) is annealed to an in vitro transcribed 3[prime]-59mer fragment. The solid black line in the folded L-shaped representation indicates a break in the sugar-phosphate backbone in the wild-type construct prepared with the 5[prime]-17mer. The truncated semi-synthetic variants are shown by numbered arrows indicating the length of each fragment from the 5[prime]-end. The 5[prime]-7mer is also boxed. Dotted lines in the L-shaped representation indicate tertiary interactions and are based on the known structure of yeast tRNAPhe (2).
Figure 4. Comparison of aminoacylation activity of full-length unmodified tRNALys,3 and semi-synthetic tRNA constructs. The semi-synthetic tRNAs are indicated by the length of the 5[prime]-oligonucleotide. The inset shows an expanded view of the results obtained with the shorter constructs. All RNAs were assayed at 5 µM, with the exception of the 3[prime]-59mer alone, which was present at 2 µM.
Table 3. Kinetic parameters determined from in vitro aminoacylation of full-length and semi-synthetic tRNALys,3 using 6H-hLysRS
| tRNALys,3variant | Temperature (°C) | kcat (s-1) | Km (µM) | kcat/Km (relative) | Fold decrease |
| Wild-type | 30 | 2.0 | 3.4 | 1 | |
| 20 | 1.7 | 3.7 | 1 | ||
| 3[prime]-59mer | |||||
| + 5[prime]-17mer | 30 | 1.5 | 5.1 | 0.63 | 1.6 |
| + 5[prime]-14mer | 20 | 0.88 | 8.9 | 0.22 | 4.5 |
| + 5[prime]-7mer | 20 | 0.00026 | 1.5 | 0.00037 | 2700 |
To more precisely determine the structural elements that contribute to the kcat defect in the truncated semi-synthetic tRNA, three additional constructs were tested (Fig. 3). In all cases, a truncated 5[prime]-oligomer was annealed to the 3[prime]-59mer. When a 5[prime]-15mer was used, wild-type levels of aminoacylation were observed (Fig. 4). In contrast, deletion of one more nucleotide from the 3[prime]-end (5[prime]-14mer) resulted in a significant decrease in catalytic efficiency (Fig. 4). This semi-synthetic tRNA lacks the capability to make a key tertiary contact, namely the G15:C48 Levitt base pair (23). A more careful kinetic analysis of this variant showed that the defect was due to an ~2-fold reduction in kcat and a 3-fold elevation in Km (Table 3). Finally, truncation to a 5[prime]-10mer resulted in even lower levels of activity, similar to the 5[prime]-7mer construct (Fig. 4).
To determine if the 5[prime]-oligonucleotides were annealing to the 3[prime]-59mer in the semi-synthetic constructs, migration of the annealed tRNA molecules on a native polyacrylamide gel run at room temperature was compared to that of the full-length tRNALys,3. The native gel shown in Figure 5 compares the mobility of full-length tRNALys,3 (lane 1, arrow), the 3[prime]-59mer alone (lane 2), and the five annealed semi-synthetic tRNA constructs (lanes 3-7). The 3[prime]-59mer (lane 2) migrates as several smeared bands, presumably due to conformational flexibility, with the major band migrating significantly slower than the full-length tRNA (lane 1). Each of the annealed semi-synthetic tRNAs (lanes 3-7) migrates as a single compact band and to a similar distance as the full-length tRNA. These results suggest that the semi-synthetic constructs are properly annealed and fold into a conformation that closely resembles that of the full-length tRNA molecule.
Figure 5. Ethidium stained, native 12% polyacrylamide gel showing: full-length tRNALys,3 (lane 1); 3[prime]-59mer alone (lane 2); 3[prime]-59mer + 5[prime]-17mer (lane 3); + 5[prime]-15mer (lane 4); + 5[prime]-14mer (lane 5); + 5[prime]-10mer (lane 6); + 5[prime]-7mer (lane 7).
DISCUSSION
A previous in vitro study of E.coli tRNALys recognition found that E.coli LysRS required the discriminator base and the anticodon for efficient aminoacylation (11). Moreover, a 140-fold difference in catalytic efficiency was observed between unmodified and fully modified E.coli tRNALys. Multiple alignments of 21 class II LysRS sequences previously indicated that the catalytic domain is one of the most highly conserved domains among synthetases (13). Nevertheless, in contrast to the E.coli enzyme, our earlier in vitro work with hLysRS showed it to be insensitive to the identity of the discriminator base (13). The fact that the human enzyme could efficiently aminoacylate both native and unmodified E.coli tRNALys also suggested that this enzyme is relatively insensitive to base modifications. The human tRNALys,3 isoacceptor is extensively modified, containing 13base modifications.In this work, using a crude preparation of native human tRNALys, we observed only a 10-fold contribution of base modifications to aminoacylation catalytic efficiency of hLysRS. Thus, modified bases are not essential for hLysRS recognition, but contribute modestly to overall catalytic efficiency.
In this study, we also probed anticodon recognition by hLysRS. It is well established that the subclass IIb synthetases recognize the anticodon via an N-terminal OB-fold domain (4). The N-terminal anticodon binding domain shows only moderate sequence conservation (8%) among the 21 LysRS sequences previously examined (13). Our results show that the effect of mutations at U36 of human tRNALys,3 was similar to the decreases in aminoacylation observed in the E.coli tRNALys system (11). On the other hand, mutagenesis of U35 to G or A has a larger effect on aminoacylation in the human system than in the E.coli system. To date, no major recognition elements have been identified in the acceptor stem of human tRNALys,3, and our results suggest that the eukaryotic system has an even greater anticodon dependence than the prokaryotic system.
In the three-dimensional L-shaped tRNA structure, the site of aminoacylation and the anticodon are ~76 Å apart (2). The mechanistic details of how communication is achieved between the anticodon and the amino acid attachment site are not well understood for most aminoacylation systems, and a complex set of RNA-protein interactions are likely to be involved. To gain some insights into this question, attempts have been made in some systems to stimulate charging of an acceptor stem-derived minihelix with the anticodon stem-loop domain. Most of this work has been carried out using class I synthetases from E.coli or yeast (24-28). For two of the systems studied, E.coli isoleucyl-tRNA synthetase and yeast valyl-tRNA synthetase, small increases (2-3-fold) in minihelix aminoacylation were observed upon anticodon addition (24,28). The yeast AspRS system is the only subclass IIb system where minihelix aminoacylation has been reported (12). Although minihelixAsp alone is aminoacylated by yeast AspRS, a stem-loop mimicking the anticodon of tRNAAsp was not capable of stimulating the aminoacylation further (12).
In our experiments with the human lysine system, we find that a minihelix derived from the acceptor-T[Psi]C stem-loop of human tRNALys,3 is not charged and that the anticodon stem-loop fails to trigger minihelixLys,3 aminoacylation. These data suggest that covalent continuity between these two key elements is required for aminoacylation by hLysRS. In support of the notion that covalent continuity is sufficient for at least low levels of aminoacylation in the lysine system, yeast and rabbit liver LysRS were reported to weakly aminoacylate a substrate consisting of 5[prime]-G(U20)CCA-3[prime] (29). In the case of the yeast enzyme, which displayed higher activity than the rabbit enzyme in these previous studies, the 3[prime]-U in the single-stranded substrate may mimic N73 of the full-length yeast tRNALys and the 5[prime]-region of the oligomer may mimic the UUU anticodon. This result suggests that simply by maintaining covalent continuity between the two distal elements, the lysine anticodon trinucleotide is able to trigger acceptor stem aminoacylation.
To probe this idea further, we have dissected human tRNALys,3 in more subtle ways and carried out a kinetic analysis of semi-synthetic tRNA variants (Fig. 3 and Table 3). Simply introducing a break in the sugar-phosphate backbone after C17 does not significantly affect aminoacylation catalytic efficiency. However, when sequence elements in the D stem-loop known to be involved in core tRNA tertiary interactions are deleted, hLysRS catalytic efficiency is severely affected despite the fact that the global tRNA-like fold is maintained (Fig. 5). Most strikingly, the 2700-fold decrease in catalytic efficiency observed with the 3[prime]-59mer + 5[prime]-7mer construct was entirely due to a decrease in kcat, with binding actually slightly improved relative to the full-length tRNA (Table 3). Thus, the anticodon domain of this semi-synthetic tRNA construct is likely to be properly bound to the synthetase. Specificity of the binding interaction is supported by the fact that a U35->G substitution in the semi-synthetic construct abolishes activity. The inhibition trends observed with stem-loops bearing UUU and UGU anticodon sequences also suggest that hLysRS is able to specifically recognize the major (U35) recognition element in the context of a simple RNA stem-loop (Table 2). Thus, the primary effect of deleting key tertiary structural elements is on the transition state of the aminoacylation reaction. The similarity in Ki values determined for both full-length and truncated U35G variants is in accord with the critical nature of the interactions between the central anticodon base and the N-terminal OB-fold (9).
The presence of D stem elements and tRNA tertiary structure was also shown to be important for aminoacylation by class I yeast methionyl-tRNA synthetase, although whether deletion of the D arm affected primarily kcat or Km was not reported in this study (26). These researchers concluded that although most of the binding energy is localized in the anticodon region, protein contacts with the D arm are required for optimal catalysis. In the case of another class I enzyme, E.coli glutaminyl-tRNA synthetase, specific amino acids that interact with the inside of the L-shaped tRNA have been identified. This is one of the few systems where a high resolution co-crystal structure of the synthetase complexed with tRNA is known (30). Synthetase mutations at these interaction sites result in relaxed anticodon recognition, thus providing a connection between the acceptor end and the anticodon domain (31).
Subclass IIb yeast AspRS is one of two class II enzymes for which high a resolution X-ray structure in complex with the tRNA has been reported (5). In this case, the acceptor stem contains key recognition elements (of approximately equal importance to the anticodon) and a minihelix is aminoacylated by yeast AspRS. The addition of a separate anticodon domain was not able to stimulate catalysis, presumably due to missing RNA-protein contacts present in the intact L-shaped structure (12). Based on the co-crystal structure, enzyme contacts to backbone functional groups of U11 and U12 in the D stem are observed (7). However, mutations at a neighboring identity element, the G10:U25 base pair, affect primarily the Km (6), and it is not clear whether these contacts are needed to communicate anticodon recognition to the active site. In the human lysine system studied here, we have shown that the anticodon makes a much larger relative contribution to aminoacylation catalytic efficiency than the acceptor stem. Based on this observation, it is not too surprising that minihelixLys,3 is not aminoacylated by hLysRS. We show that effective communication of the anticodon binding energy to the active site requires the presence of a relatively small number of key tertiary structural elements. More specifically, the presence of an intact Levitt base pair (G15:C48) contributes 0.88 kcal/mol to the apparent free energy of transition state stabilization, while nucleotides 8-14 in the D arm contribute an additional 4.6 kcal/mol.
In summary, in this work we show that the anticodon is a critical recognition element for hLysRS and examine the mechanism of communication between the anticodon and the amino acid acceptor stem domain in a human system for the first time. As reported previously for several E.coli and yeast aminoacylation systems, covalent continuity between the acceptor and anticodon domains appears to be an important requirement for efficient charging by the human synthetase. Elimination of tertiary structural elements of tRNALys,3 has little effect on anticodon-dependent substrate binding, but severely impacts formation of the transition state for catalysis. Taken together, our results suggest that a relatively small number of specific tertiary structural elements of human tRNALys,3 are responsible for conferring effective communication between the crucial anticodon recognition elements and the acceptor stem.
REFERENCES
*To whom correspondence should be addressed. Tel: +1 612 624 0286; Fax: +1 612 626 7541; Email: musier{at}chem.umn.edu
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. McCulley and C. D. Morrow Complementation of Human Immunodeficiency Virus Type 1 Replication by Intracellular Selection of Escherichia coli Formula Supplied in trans J. Virol., October 1, 2006; 80(19): 9641 - 9650. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. SISSLER, M. HELM, M. FRUGIER, R. GIEGE, and C. FLORENTZ Aminoacylation properties of pathology-related human mitochondrial tRNALys variants RNA, May 1, 2004; 10(5): 841 - 853. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Javanbakht, R. Halwani, S. Cen, J. Saadatmand, K. Musier-Forsyth, H. Gottlinger, and L. Kleiman The Interaction between HIV-1 Gag and Human Lysyl-tRNA Synthetase during Viral Assembly J. Biol. Chem., July 18, 2003; 278(30): 27644 - 27651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Francin and M. Mirande Functional Dissection of the Eukaryotic-specific tRNA-interacting Factor of Lysyl-tRNA Synthetase J. Biol. Chem., January 10, 2003; 278(3): 1472 - 1479. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Ambrogelly, D. Korencic, and M. Ibba Functional Annotation of Class I Lysyl-tRNA Synthetase Phylogeny Indicates a Limited Role for Gene Transfer J. Bacteriol., August 15, 2002; 184(16): 4594 - 4600. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Javanbakht, S. Cen, K. Musier-Forsyth, and L. Kleiman Correlation Between tRNALys3 Aminoacylation and Its Incorporation into HIV-1 J. Biol. Chem., May 10, 2002; 277(20): 17389 - 17396. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Soll, H. D. Becker, P. Plateau, S. Blanquet, and M. Ibba Context-dependent anticodon recognition by class I lysyl-tRNA synthetases PNAS, December 19, 2000; 97(26): 14224 - 14228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Miller, B. Ehresmann, U. Hubscher, and S. F. J. Le Grice A Novel Interaction of tRNALys,3 with the Feline Immunodeficiency Virus RNA Genome Governs Initiation of Minus Strand DNA Synthesis J. Biol. Chem., July 13, 2001; 276(29): 27721 - 27730. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Francin, M. Kaminska, P. Kerjan, and M. Mirande The N-terminal Domain of Mammalian Lysyl-tRNA Synthetase Is a Functional tRNA-binding Domain J. Biol. Chem., January 11, 2002; 277(3): 1762 - 1769. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||