Shifted positioning of the anticodon nucleotide residues of amber suppressor tRNA species by Escherichia coli arginyl-tRNA synthetase Daisuke Kiga 1,2 , Kensaku Sakamoto 2 , Saori Sato 1 , Ichiro Hirao 1 and Shigeyuki Yokoyama 1,2 1 Yokoyama CytoLogic Project, ERATO, JST c/o RIKEN, Hirosawa, Wako-shi, Saitama, Japan; 2 Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Japan Cytidine in the anticodon second position (position 35) and G or U in position 36 of tRNA Arg are required for amino- acylation by arginyl-tRNA synthetase (ArgRS) from Escheri- chia coli. Nevertheless, an arginine-accepting amber suppressor tRNA with a CUA anticodon (FTOR1D26) exhibits suppression activity in vivo [McClain, W.H. & Foss, K. (1988) Science, 241, 1804–1807]. By an in vitro kinetic study with mutagenized tRNAs, we showed that the arginylation of FTOR1D26 involves C34 and U35, and that U35 can be replaced by G without affecting the activity. Thus, the positioning of the essential nucleotides for the arginylation is shifted to the 5 0 side, by one residue, in the suppressor tRNA Arg . We found that the shifted positioning does not depend on the tRNA sequence outside the anti- codon. Furthermore, by a genetic method, we isolated a mutant ArgRS that aminoacylates FTOR1D26 more efficiently than the wild-type ArgRS. The isolated mutant has mutations at two nonsurface amino-acid residues that interact with each other near the anticodon-binding site. Keywords: tRNA identity; anticodon; aminoacyl-tRNA synthetase; genetic screen; kinetic analysis. The accurate recognition of a tRNA by its aminoacyl-tRNA synthetase (aaRS) is a vital step in ensuring the fidelity of translation. An aaRS distinguishes its cognate tRNAs from the tRNA species existing in the same cell by the recognition of the particular nucleotide residues (identity determinants) that are found only in the cognate tRNAs as one set. In most tRNA species, these nucleotides are located in the anticodon moiety and at the discriminator position (position 73) [1]. The identity of tRNA Arg involves two anticodon positions: a cytidine residue in position 35 (C35) and a G or U in position 36 both contribute to the arginine-accepting activity, and the effect of base substitutions is much larger for position 35 than for position 36 [2 – 4]. In contrast, position 34 is not thought to contribute to the reaction, because various base substitutions are allowed in this position; the naturally occurring tRNA Arg species have inosine, cytidine, and a modified uridine in this position [5], and tRNA Arg tran- scripts with A34 and C34 exhibit arginylation activity comparable to that of the fully modified tRNA Arg [4]. The irrelevance of position 34 and the ambiguous recog- nition at position 36 are necessary, because there are several tRNA species with different anticodon sequences for reading the six arginine codons, which only have G as the second letter in common. Leucine and serine are also each encoded with six codons read by several tRNA species, but their tRNAs have identity determinants located mainly outside the anticodon [1]. In addition to the anticodon residues, A20 in the D loop also contributes to the tRNA Arg identity in Escherichia coli and probably in most organisms other than yeast, but its contribution to the activity is only comparable to that of G/U36 [2,4,6]. McClain & Foss [6] used amber suppressor tRNA species to analyze the identity determinants of E. coli tRNA Arg . First, a tRNA Arg 2 variant, whose anticodon was replaced by CUA, inserted more Lys than Arg in response to the amber codon, suggesting that the amber suppressor tRNA Arg has structural features similar to those of tRNA Lys . Therefore, they tried a different approach: an amber suppressor derived from tRNA Phe was engineered by the transplantation of A20 together with the three surrounding nucleotides. Actually, the resultant suppressor (‘F to R’ or FTOR1) inserted Arg about 10 times more frequently than Lys or Tyr in response to the amber codon. Furthermore, only Arg was detected with FTOR1D26, which was made by the deletion of A26 from FTOR1. As these Arg-inserting amber suppressors lack both of the anticodon identity determinants, C35 and G/U36, of tRNA Arg , C34 may be recognized by the C35- recognition site of ArgRS. (This mode of recognition is referred to hereafter as ‘shifted positioning’.) The deletion of A26 may facilitate the shift of C34 toward the location of C35, which might enhance the Arg-accepting activity, but not the Lys- and Tyr-accepting activities [6]. Alternatively, the A26 deletion may depress the Lys- and Tyr-accepting activities, but not the Arg-accepting activity, of the tRNA [6]. Recently, the crystal structure of the ArgRS : tRNA com- plex from yeast was determined, which revealed a structural basis for the anticodon recognition by ArgRS [7]. An interesting finding was that C35 does not make any specific interactions with amino-acid side chains, but only with certain backbone atoms in the C-terminal domain. Even with this structural information, two questions remained. Correspondence to S. Yokoyama, Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. Fax: 1 81 35841 8057, Tel.: 1 81 35841 4413, E-mail: yokoyama@biochem.s.u-tokyo.ac.jp Enzymes: arginyl-tRNA synthetase from Escherichia coli (SWISS-PROT entry name ¼ SYR_ECOLI; EC 6.1.1.19). (Received 2 July 2001, revised 2 October 2001, accepted 3 October 2001) Abbreviations: aaRS, aminoacyl-tRNA synthetase; ArgRS, arginyl-tRNA synthetase. Eur. J. Biochem. 268, 6207–6213 (2001) q FEBS 2001 First, the possibility of the ‘shifted positioning’ of C34 needed to be examined. Secondly, the types of amino-acid replacements that change the anticodon specificity of ArgRS needed to be identified; because of the unique manner of C35 recognition by ArgRS, it is difficult to rationally design amino-acid replacements to change the anticodon specificity on a structural basis. In the present study, we analyzed FTOR1D26 and its variants, and found that position 34, but not position 35, is crucial for the arginylation in E. coli. In addition, this shift of the crucial position does not depend on the sequence outside the anticodon. Finally, by genetic selection, we isolated a mutant ArgRS that arginylates the amber suppressor tRNAs better than the wild-type enzyme. MATERIALS AND METHODS DNA manipulation, sequencing, and PCR amplification Standard techniques were used for restriction endonuclease digestion, ligation, and gel electrophoresis [8]. The nucleo- tide sequence was determined using a Sequencing Pro Autosequencer core kit for labeled primers (Toyobo, Tokyo, Japan). PCR was performed using the Gene Amp PCR system 9700 (Applied Biosystems). Preparation of ArgRS The wild-type and mutant ArgRS species were purified by a histidine-tag system from overproducing cells, to exclude any contamination by the endogenous wild-type enzyme. Thus, each ArgRS preparation has 10 tandem histidine residues at the N-terminus (the C-terminal tag inactivates the enzyme). The tagged wild-type enzyme exhibited an aminoacylation activity similar to the reported values [4,9]. To tag the enzymes, the argS coding sequences were amplified by PCR with the primers, 5 0 -GGGAATTCCATA TGAATATTCAGGCTCTTC-3 0 and 5 0 -GAGCGGATCCA AGCTTCCATTTTCAGAATACATTTAGATGGC-3 0 ,and were then ligated between the Nde I–Bam HI sites of pET26b (Novagen). The E. coli BLR (DE3) cells (Novagen) were transformed with this plasmid expressing the tagged ArgRS, and were then grown at 25 8C in Luria– Bertani media supplemented with 6% glucose. At the late log phase of cell growth, isopropryl thio-b- D-galactoside was added to the media to a 0.5 m M concentration, and after 1 h of induction, the cells were harvested. The enzyme was then purified by chromatography on Ni-nitrilotriacetic acid agarose (Qiagen), followed by FPLC using a Resource-Q column (Amersham Pharmacia Biotech). In vitro aminoacylation analyses Transfer RNA variants were prepared using T7 RNA polymerase [10]. According to the standard procedure [10], the tRNA aminoacylation assays were performed at 37 8Cin 40 mL of a buffer [100 m M Tris/HCl, pH 7.5, 15 mM MgCl 2 , 4m M ATP, 60 mM [ 14 C]arginine (332.1 pCi : pmol 21 ,New England Nuclear)] containing various concentrations of the enzyme and the tRNA. Before the reaction, the tRNA was renatured at 65 8C for 5 min in the reaction buffer. The kinetic parameters were obtained from Lineweaver–Burk plots. The obtained data are the averages of at least two independent assays. As E. coli ArgRS has a K m value of 12 m M for arginine [9], its concentration in our assay is nearly saturating. Bacterial strains and plasmids for the genetic study The E. coli arginine auxotroph, which we designated as NR101, is an argE(Am) strain provided in an Interchange kit (Promega), and confers the tetracycline resistance. The arginine-inserting amber suppressor tRNA, FTOR1D26, reported by McClain & Foss [6], had been expressed from the gene cloned in pUC18 (which we desig- nated as pUCsupR) under the control of the lpp promoter and the rrnC terminator. This gene was cloned in the low-copy vector, pAp102, which has a ColIbP9 replication origin, and a copy number of 1.7 per cell [11], to generate pApsupR. The pAp102 vector was a gift from K. Mizobuchi (Department of Applied Physics and Chemistry, University of Electro-Communications, Tokyo, Japan). The E. coli argS gene, with its native promoter and SD sequence, was inserted between the Bam HI–HindIII sites of the pACYC184 plasmid to generate pACargS. Construction of a mutant argS library and genetic selection In order to prepare a mutant ArgRS library, part of the argS sequence (0.5 kb), from the Eco RV site (amino-acid residue 318) to the C-terminal end, was amplified by error-prone PCR with the primers, 5 0 -CACCACTGATATCGCCTG TGCG-3 0 and 5 0 -GAGCAAGCTTCCATTTTCAGAATAC ATTTAGATGGC-3 0 , where the HindIII site is underlined [12]. After digestion with these restriction enzymes, the PCR product was cloned in pACargS, in place of its counterpart in the wild-type argS sequence, to generate the pACargS mu plasmids. In order to isolate mutant ArgRS species that amino- acylate FTOR1D26, strain NR101 was transformed with pApsupR together with pACargS mu , and was then grown for 36 h at 37 8C on an M9 plate supplemented with proline, methionine, and thiamine (1 m M each), but not with argi- nine. The colonies that formed on the plate were isolated, and after the pACargS mu plasmids were extracted, they were tested again for the potential to suppress the arginine auxotroph. The Eco RV – HindIII fragments of the recovered argS genes from the plasmids thus selected were cloned again in pACargS, in place of the counterpart in the wild- type argS, and were then examined again for their suppressing potential. Finally, the Eco RV – HindIII frag- ments thus selected were subjected to sequence determination. RESULTS AND DISCUSSION ArgRS recognition of tRNA Arg species with a CUA anticodon In E. coli and yeast, C35 and G or U in position 36 are crucial for the arginylation [2,4,13]. In contrast, position 34 is not thought to contribute to the reaction [4,5]. With this in mind, we investigated the manner by which ArgRS recognizes FTOR1D26. A wild-type tRNA Arg , tRNA Arg 2 [14], and FTOR1D26 variants with various 6208 D. Kiga et al. (Eur. J. Biochem. 268) q FEBS 2001 anticodon sequences were prepared by in vitro transcription with T7 RNA polymerase, and were subjected to in vitro kinetic analyses. The wild-type tRNA Arg 2 exhibits an arginine-accepting activity similar to those reported so far [4,9], while FTOR1D26 exhibits a 6 Â 10 5 -fold lower activity than the wild-type (Table 1). The replacement of C34 by either A or U in FTOR1D26 further reduced the arginine-accepting activity. This result is in sharp contrast to the wild-type tRNA Arg with C35, in which position 34 is not relevant, but position 35 is strictly required to be C for the arginine-accepting activity [4,5]. For position 35 of FTOR1D26, the U35A substitution also reduces the activity, but U35G does not affect it, which is reminiscent of the effects of the base substitutions in position 36 of the wild- type tRNA Arg species. As the specificities of ArgRS for positions 34 and 35 of FTOR1D26 are exactly the same as those for positions 35 and 36, respectively, of the wild-type tRNA Arg , this indicates that C34 and U35 of FTOR1D26 are recognized by the binding pockets for C35 and G/U36, respectively, of the enzyme, as previously proposed [6]. We then examined the effect of the deletion of A26. The arginine-accepting activity of FTOR1 is higher than that of FTOR1D26 in vitro (Table 1). However, FTOR1 inserts Lys and Tyr, in addition to Arg, but FTOR1D26 predominantly inserts Arg, in response to the amber codon [6]. As previously proposed [6], the two possible reasons why the deletion of A26 makes the tRNA much more specific to arginine are (a) the arginine-accepting activity, but not the lysine- and tyrosine-accepting activities, of the tRNA was increased, and (b) the recognitions of the tRNA by LysRS and TyrRS were more impaired than that by ArgRS. The present results clearly indicate that the latter reason is the case for the A26 deletion of FTOR1. The C34A and C34U substitutions of FTOR1 both reduce the arginine-accepting activity, as in the case of FTOR1D26 (Table 1). Therefore, the shifted positioning occurred in both the presence and absence of A26 in FTOR1. We next assayed the amber suppressor tRNA derived from tRNA Arg 2 , whose nucleotide sequence is significantly different from that of FTOR1D26. This amber suppressor tRNA, tRNA Arg 2 (CUA), shows a higher arginine-accepting activity in vitro than those of FTOR1D26 and FTOR1 (Table 1). In contrast, tRNA Arg 2 (CUA) inserts more lysine than arginine in vivo in response to the amber codon, while the other two tRNAs mainly insert arginine [6]. The nucleo- tide sequence of tRNA Lys is more similar to that of tRNA Arg 2 (CUA) than to those of FTOR1D26 and FTOR1, which are derived from tRNA Phe . Again, the substitution of C34 in tRNA Arg 2 (CUA) decreased the arginine acceptance (Table 1), indicating that neither the irregular conformation of FTOR1D26 nor the tRNA sequence outside the anticodon is important for the shifted positioning. In summary, the change in the tRNA framework from tRNA Arg 2 (CUA) to FTOR1, as well as the A26 deletion of FTOR1, increased the arginine specificity by decreasing the recognition by LysRS and TyrRS much more than that by ArgRS. Genetic selection of mutant ArgRS molecules To explore the accommodation of the suppressor tRNA by ArgRS, mutant enzymes facilitating the amber suppression by the tRNA were isolated by genetic methods. In the complex structure of ArgRS : tRNA, C35 is recognized only by the main-chain atoms of the enzyme [7], and this finding raised the question of what types of amino-acid replace- ments change the anticodon specificity of the enzyme. For the mutant isolation, we constructed a genetic system to detect the increased activity of a mutant enzyme for FTOR1D26. When expressed from a high-copy plasmid (pUCsupR), FTOR1D26 by itself exhibits a suppression activity. When expressed from a single copy plasmid (pApsupR), FTOR1D26 does not complement the poor growth of the E. coli argE(Am) mutant on minimal medium (data not shown): this cannot be rescued by the over- production of ArgRS from the plasmid pACargS (a multi- copy plasmid carrying argS ), which allowed us to isolate mutant argS genes that promote the amber suppression by FTOR1D26. Table 1. The kinetic parameters for the aminoacylation of the tRNA Arg variants by the wild-type ArgRS and the argS1 mutant enzyme. ND, not determined because of the very low activation of the tRNA by the enzyme variant. “Framework” indicates a moiety of tRNA outside the anticodon. tRNA Wild-type ArgRS argS1 Framework Anticodon k cat (s 21 ) K m (mM) k cat /K m (M 21 : s 21 ) k cat (s 21 ) K m (mM) k cat /K m (M 21 : s 21 ) tRNA Arg 2 (wild-type) ACG 17 1.0 1.7 Â 10 7 2.8 0.44 6.4 Â 10 6 FTOR1D26 CUA 9.3 Â 10 24 34 27 3.1 Â 10 23 32 97 FTOR1D26 UUA ND ND ,10 ND ND ,10 FTOR1D26 AUA ND ND ,10 ND ND ,10 FTOR1D26 CGA 1.1 Â 10 23 38 29 3.7 Â 10 23 32 1.2 Â 10 2 FTOR1D26 CAA ND ND ,10 ND ND ,10 FTOR1 CUA 2.8 Â 10 23 14 2.0 Â 10 2 6.4 Â 10 23 12 5.3 Â 10 2 FTOR1 UUA ND ND ,20 1.1 Â 10 23 19 58 FTOR1 AUA ND ND ,20 8.6 Â 10 24 11 78 tRNA Arg 2 CUA 3.1 Â 10 22 4.5 6.9 Â 10 3 8.4 Â 10 22 5.6 1.5 Â 10 4 tRNA Arg 2 UUA ND ND ,700 ND ND ,700 tRNA Arg 2 AUA ND ND ,700 ND ND ,700 q FEBS 2001 Unusual manner of anticodon recognition (Eur. J. Biochem. 268) 6209 Fig. 1. Amino-acid replacements found in the ArgRS mutants. The amino-acid sequences from residue 318 to the C-terminus (residue 577) of the wild-type E. coli ArgRS and its mutants are shown. The secondary structures are assigned according to the crystal structure of yeast ArgRS determined Cavarelli et al. [20]. 6210 D. Kiga et al. (Eur. J. Biochem. 268) q FEBS 2001 On average, a few base substitutions were introduced in random positions in the C-terminal region of ArgRS by error-prone PCR; this region ranges from residue 318, a position between the HIGH and KMSKS sequences, to the C-terminus. Of the 10 7 argE(am) cells with pApsupR that were transformed by these randomly mutagenized argS genes, 10 2 cells formed colonies on the minimal plates without arginine after an incubation at 37 8C for 36 h. After excluding the revertant cells, 26 different mutant argS genes were isolated with the amino-acid replacements listed in Fig. 1, which will be discussed later. Recognition of the suppressor tRNA species by a mutant ArgRS To characterize the mutant ArgRSs, we tried to purify the mutant enzymes. The histidine tag was added to the N-terminus of the enzyme to exclude the contamination by the endogenous wild-type ArgRS. As all but one of the mutant enzymes with the histidine tag were found in the insoluble fraction of the cell lysates, we hereafter focused on the only soluble mutant enzyme, argS1, with two amino- acid replacements, Met to Val in position 460 and Tyr to Asp in position 524. In order to investigate the effects of these replacements individually, the argS mutants with either replacement were constructed by site-directed mutagenesis, and were subjected to the same test used for the mutant selection; neither of the single mutants could form colonies on the plate for the selection. Both replacements were thus shown to be necessary for the ability of argS1 to suppress argE(Am) when FTOR1D26 is expressed simultaneously (data not shown). AsshowninTable1,theargS1 mutant enzyme arginylates FTOR1D26 3.6-fold more efficiently, and the wild-type tRNA Arg 2.7-fold less efficiently, than the wild- type ArgRS. Thus, the tRNA specificity of argS1 changes by 9.7-fold, compared with that of the wild-type ArgRS. The activities of argS1 for the other two tRNAs with the CUA anticodon, FTOR1 and tRNA Arg 2 (CUA), were also higher than that of the wild-type enzyme. The effects of the anticodon base substitutions in FTOR1D26 were examined for argS1, and the shifted positioning of the anticodon residues was also observed for this mutant enzyme (Table 1); the substitutions of C34U, C34A, and U35A all reduce the arginine-accepting activity, whereas U35G has no effect on the activity. In addition, the substitutions of C34, in both FTOR1 and tRNA Arg 2 (CUA), all reduce the arginine-accepting activity by the mutant enzyme. Between the wild-type and mutant enzymes, the differ- ence in the aminoacylation activity of the suppressor tRNA species is mainly derived from the difference in the k cat values, not in the K m values. It reminds us that the substi- tution of the identity determinants in the anticodon of tRNA Arg mainly decreases the k cat values [2,4]. The muta- tions therefore seem to facilitate the unusual signal Fig. 2. Mapping of the yeast ArgRS residues that correspond to the two amino-acid replacements in the argS1 mutant of E. coli ArgRS on the yeast ArgRS : tRNA complex structure. (A) The crystal structure of the yeast ArgRS : tRNA complex [7] used as a working model of the E. coli system. In the argS1 mutant of E. coli ArgRS, Met460 and Tyr524 are replaced by Val and Asp, respectively. The E. coli Met460 and Tyr524 correspond to Leu489 and Phe555, respectively, of yeast ArgRS. On the ribbon model of yeast ArgRS, the positions of Leu489 and Phe555 are indicated with the side chains (ball- and-stick) in green and magenta, respectively. The tRNA main chain is shown by a wire model, mainly colored yellow, while a part of the anticodon stem (positions 39–41), which interact with the H18 and H22 helices bearing Leu489 and Phe555, respectively, are shown in red. The three nucleotide residues of the anticodon are indicated by orange sticks. (B) The detailed structure around Leu489 and Phe555 in the yeast ArgRS : tRNA complex. The H18 helix and the Leu489 side chain are shown in green, and the H22 helix and the Phe555 side chain are shown in magenta. The side chains of Tyr488 on the H18 helix and those of His559, Ser562, and Ser563 on the H22 helix, shown by balls and sticks, form a protein surface that contacts the anticodon-stem residues in positions 39–41, shown by sticks. The E. coli ArgRS residues that correspond to the yeast ArgRS residues are indicated in parentheses. All the figures were prepared using the programs MOLSCRIPT [21] and RASTER3D [22]. q FEBS 2001 Unusual manner of anticodon recognition (Eur. J. Biochem. 268) 6211 transduction from the C35- and G/U36-binding pockets upon the binding of some nucleotides in the noncognate CUA anticodon to the catalytic domain of the enzyme. Structural elements for the change in the tRNA specificity of argS1 The argS1 mutant enzyme involves two mutations, M460V and Y524D, as described above. In the sequence alignment of the ArgRSs from E. coli and yeast, the E. coli Met460 corresponds to the yeast Leu489, which is located on an a helix, H18 (Fig. 2), in the reported structure of the yeast ArgRS : tRNA complex [7]. The other argS1 mutation, Y524D, of the E. coli ArgRS occurs in the position that corresponds to position 555 in the yeast ArgRS. This position is located on another a helix, H22 (Fig. 2). These two helices form part of the a helix bundle that constitutes the anticodon-binding site. In the helix bundle of the yeast ArgRS, these two amino-acid residues are below the protein surface, and interact directly with each other (Fig. 2). The mutations in these positions should induce some structural change on the surface of the helices, which contact the backbones of the nucleotide residues in positions 39–41 within the anticodon stem of the tRNA (Fig. 2). Upon binding of the anticodon of the wild-type tRNA Arg to the site on the yeast ArgRS, the anticodon loop undergoes a large conformational change [7]; upon the putative binding of C34 and U35 of the suppressor tRNA species to the C35- and G/U36-binding sites, respectively, of the enzyme (‘shifted positioning’), a somewhat different conformational change of the anticodon loop is likely to be required. This alternative anticodon-loop conformational change of the suppressor tRNA is probably much less favorable for amino- acylation than that of the wild-type tRNA. In the argS1 mutant enzyme, the unusual structure due to the mutations of the anticodon-stem-docking site may facilitate the alter- native anticodon-loop conformational change of the sup- pressor tRNA, but it may hamper the anticodon-loop conformational change of the wild-type tRNA. It has been reported that mutations in positions other than the anticodon-interacting amino-acid residues increase the aminoacylation of a nonsense suppressor tRNA; for E. coli glutaminyl-tRNA synthetase, the mutations are outside the anticodon-binding site, and are located near the core region of tRNA in the enzyme : tRNA complex [15]. It was argued that these mutations affect the process of transmitting the signal from the anticodon binding domain to the active site, and make the enzyme bind C35 in the pocket for U35, which is conserved in all of the glutamine tRNAs. On the other hand, on the surface of the putative anticodon-recognition helix of E. coli methionyl-tRNA synthetase, there are two acidic residues that reportedly serve as negative discrimi- nants against noncognate tRNA anticodons through a direct electrostatic repulsion; the replacement of one of these residues does not affect the activity of the cognate tRNA, but rather increases the activity of the noncognate ones [16]. The mutations in the 26 ArgRS mutants isolated in this work exhibit a tendency to converge at certain positions (Fig. 1), although their contributions to the mutant pheno- type have not yet been confirmed, and the mutant enzymes have not been characterized. Structural and genetic studies have suggested that the H16, H18, and H22 helices and the V loop of yeast ArgRS are probably involved in the recognition of the anticodon moiety [7,17]. The conver- gence of the above mutations in these domains suggests that their effects are associated with the anticodon recognition by the E. coli enzyme. Misrecognition of an apparently irrelevant anticodon sequence has been also reported for yeast ArgRS [13,18,19], which misarginylates tRNA Asp transcripts with an efficiency similar to to that of the arginylation of the cognate tRNA Arg species. This misrecognition is due to the existence, in the anticodon region, of an alternative set of nucleotides con- tributing to the arginylation. In addition, for the yeast ArgRS, the replacement of Y491H (position 462 in the E. coli numbering) probably affects its anticodon specificity, causing a misarginylation of the noncognate tRNA Asp with a GUC anticodon sequence [17]. Notably, Tyr491 is also located on the H18 helix, and is only two residues down- stream of Leu489, the counterpart of Met460 found in the E. coli argS1 mutant. ACKNOWLEDGEMENTS We thank Dr J. Cavarelli (Biologie et Genomique Structurales, Institut de Genetique et de Biologie Moleculaire et Cellulaire, CNRS/INSERM/ ULP) and coworkers, for allowing us to use the published coordinates of the ArgRS : tRNA complex before public release. We also thank Dr K. Mizobuchi for providing us the pAp102 vector. This work was supported in part by the Japan Society for the Promotion of Science under the Research for the Future program (JSPS-RFTF 96I00101). REFERENCES 1. Giege, R., Sissler, M. & Florentz, C. (1998) Universal rules and idiosyncratic features in tRNA identity. Nucleic Acids Res. 26, 5017–5035. 2. Schulman, L.H. & Pelka, H. (1989) The anticodon contains a major element of the identity of arginine transfer RNAs. Science. 246, 1595–1597. 3. McClain, W.H., Foss, K., Jenkins, R.A. & Schneider, J. (1990) Nucleotides that determine Escherichia coli tRNA (Arg) and tRNA (Lys) acceptor identities revealed by analyses of mutant opal and amber suppressor tRNAs. Proc. Natl Acad. Sci. USA 87, 9260–9264. 4. Tamura, K., Himeno, H., Asahara, H., Hasegawa, T. & Shimizu, M. (1992) In vitro study of E. coli tRNA (Arg) and tRNA (Lys) identity elements. Nucleic Acids Res. 20, 2335–2339. 5. Sprinzl, M., Horn, C., Brown, M., Ioudovitch, A. & Steinberg, S. (1998) Compilation of tRNA sequences and sequences of tRNA genes. Nucleic Acids Res. 26, 148–153. 6. McClain, W.H. & Foss, K. (1988) Changing the acceptor identity of a transfer RNA by altering nucleotides in a ‘variable pocket’. Science. 241, 1804–1807. 7. Delagoutte, B., Moras, D. & Cavarelli, J. (2000) tRNA amino- acylation by arginyl-tRNA synthetase: induced conformations during substrates binding. EMBO J. 19, 5599 –5610. 8. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor University Press, Cold Spring Harbor, New York. 9. Lin, S.X., Shi, J.P., Cheng, X.D. & Wang, Y.L. (1988) Arginyl- tRNA synthetase from Escherichia coli, purification by affinity chromatography, properties, and steady-state kinetics. Biochem- istry. 27, 6343–6348. 10. Nureki, O., Niimi, T., Muramatsu, T., Kanno, H., Kohno, T., Florentz, C., Giege, R. & Yokoyama, S. (1994) Molecular recog- nition of the identity-determinant set of isoleucine transfer RNA from Escherichia coli. J. Mol. Biol. 236, 710–724. 6212 D. Kiga et al. (Eur. J. Biochem. 268) q FEBS 2001 11. Clewell, D.B. & Helinski, D.E. (1970) Existence of the colicinogenic factor-sex factor ColI-b-P9 as a supercoi led circular DNA-protein relaxation complex. Biochem. Biophys. Res. Com- mun. 41, 150–156. 12. Cadwell, R.C. & Joyce, G.F. (1992) Randomization of genes by PCR mutagenesis. PCR Methods & Applications. 2, 28–33. 13. Sissler, M., Giege, R. & Florentz, C. (1996) Arginine aminoacyla- tion identity is context-dependent and ensured by alternate recog- nition sets in the anticodon loop of accepting tRNA transcripts. EMBO J. 15, 5069–5076. 14. Chakraburtty, K. (1975) Primary structure of tRNA Arg II of E. coli B. Nucleic Acids Res. 2, 1787–1792. 15. Rogers, M.J., Adachi, T., Inokuchi, H. & Soll, D. (1994) Functional communication in the recognition of tRNA by Escherichia coli glutaminyl-tRNA synthetase. Proc. Natl Acad. Sci. USA 91, 291–295. 16. Schmitt, E., Meinnel, T., Panvert, M., Mechulam, Y. & Blanquet, S. (1993) Two acidic residues of Escherichia coli methionyl-tRNA synthetase act as negative discriminants towards the binding of non-cognate tRNA anticodons. J. Mol. Biol. 233, 615 –628. 17. Geslain, R., Martin, F., Delagoutte, B., Cavarelli, J., Gangloff, J. & Eriani, G. (2000) In vivo selection of lethal mutations reveals two functional domains in arginyl-tRNA synthetase. RNA 6, 434–448. 18. Sissler, M., Giege, R. & Florentz, C. (1998) The RNA sequence context defines the mechanistic routes by which yeast arginyl- tRNA synthetase charges tRNA. RNA 4, 647–657. 19. Sissler, M., Eriani, G., Martin, F., Giege, R. & Florentz, C. (1997) Mirror image alternative interaction patterns of the same tRNAwith either class I arginyl-tRNA synthetase or class II aspartyl-tRNA synthetase. Nucleic Acids Res. 25, 4899–4906. 20. Cavarelli, J., Delagoutte, B., Eriani, G., Gangloff, J. & Moras, D. (1998) L-Arginine recognition by yeast arginyl-tRNA synthetase. EMBO J. 17, 5438–5448. 21. Kraulis, P.J. (1991) MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946–950. 22. Merritt, E.A. & Bacon, D.J. (1997) Raster3D. Photorealistic Molecular Graphics. Methods Enzymol. 277, 505–524. q FEBS 2001 Unusual manner of anticodon recognition (Eur. J. Biochem. 268) 6213 . Shifted positioning of the anticodon nucleotide residues of amber suppressor tRNA species by Escherichia coli arginyl -tRNA synthetase Daisuke Kiga 1,2 , Kensaku. the backbones of the nucleotide residues in positions 39–41 within the anticodon stem of the tRNA (Fig. 2). Upon binding of the anticodon of the wild-type tRNA Arg to the site on the yeast ArgRS, the anticodon. for the aminoacylation of the tRNA Arg variants by the wild-type ArgRS and the argS1 mutant enzyme. ND, not determined because of the very low activation of the tRNA by the enzyme variant. “Framework”