Peroxin Pex21p interacts with the C-terminal noncatalytic domain of yeast seryl-tRNA synthetase and forms a specific ternary complex with tRNA Ser Vlatka Godinic 1 , Marko Mocibob 1 , Sanda Rocak 1 , Michael Ibba 2 and Ivana Weygand-Durasevic 1 1 Department of Chemistry, Faculty of Science, University of Zagreb, Croatia 2 Department of Microbiology, The Ohio State University, Columbus, OH, USA Accurate aminoacylation of tRNA by aminoacyl-tRNA synthetases (aaRSs) is a crucial step in the faithful translation of mRNA. In addition to their fundamental role in translation, aaRSs are now known to participate in a wide variety of other cellular processes [1]. These noncanonical activities vary with the type of aaRS and its cellular location [2]. The expansion of synthetase function well beyond protein synthesis, and in many cases the participation of synthetases in the control of the efficiency and fidelity of the catalytic reaction, depend on terminally appended or inserted noncatalytic domains [3–5]. For example, specialized tRNA-binding domains attached to the N-terminal or C-terminal ends guide the productive docking of cognate tRNAs [6–8]. Editing domains, appended to the ends or inserted into the core domains, catalyze the tRNA-dependent hydro- lysis of incorrectly attached amino acids [9]. The role of catalytically dispensable peptides is also to promote association between the synthetases [10]. Such multi- synthetase complexes are a common feature of all higher eukaryotic cells or tissues [3,11,12]. These assem- blies often involve different nonsynthetase proteins [13,14], which may affect the efficacy of the aminoacy- lation reaction in trans. In addition, many novel func- tions unrelated to protein synthesis have been ascribed to aaRS-interacting multifunctional proteins and aaRSs in higher eukaryotes [2]. For example, fragments of TyrRS stimulate angiogenesis, whereas those of TrpRS Keywords peroxin; protein biosynthesis; protein– protein interaction; seryl-tRNA synthetase; yeast two-hybrid Correspondence I. Weygand-Durasevic, Department of Chemistry, Faculty of Science, University of Zagreb, Horvatovac 102a, 10000 Zagreb, Croatia Fax: +385 1 460 6401 Tel: +385 1 460 6230 E-mail: weygand@chem.pmf.hr (Received 22 December 2006, revised 24 February 2007, accepted 27 March 2007) doi:10.1111/j.1742-4658.2007.05812.x The seryl-tRNA synthetase from Saccharomyces cerevisiae interacts with the peroxisome biogenesis-related factor Pex21p. Several deletion mutants of seryl-tRNA synthetase were constructed and inspected for their ability to interact with Pex21p in a yeast two-hybrid assay, allowing mapping of the synthetase domain required for complex assembly. Deletion of the 13 C-terminal amino acids abolished Pex21p binding to seryl-tRNA synthe- tase. The catalytic parameters of purified truncated seryl-tRNA synthetase, determined in the serylation reaction, were found to be almost identical to those of the native enzyme. In vivo loss of interaction with Pex21p was con- firmed in vitro by coaffinity purification. These data indicate that the C-ter- minally appended domain of yeast seryl-tRNA synthetase does not participate in substrate binding, but instead is required for association with Pex21p. We further determined that Pex21p does not directly bind tRNA, and nor does it possess a tRNA-binding motif, but it instead participates in the formation of a specific ternary complex with seryl-tRNA synthetase and tRNA Ser , strengthening the interaction of seryl-tRNA synthetase with its cognate tRNA Ser . Abbreviations aaRS, aminocyl-tRNA synthetase; EMSA, electrophoretic mobility shift assay; Gal-X, 5-bromo-4-chloroindol-3-yl b- D-galactopyranoside; Gal-ONp, 2-nitrophenyl b- D-galactopyranoside; RS, tRNA synthetase; Sc, Saccharomyces cerevisiae; SerRS, seryl-tRNA synthetase; Zmc, Zea mays cytosolic. 2788 FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS inhibit angiogenesis [15]. Thus, the unusual versatility of aaRSs in higher eukaryotes is further increased by their interactions with nonsynthetase proteins. The association of the synthetases into multicompo- nent complexes has also been described in lower eukary- otes [16,17], archaea [18,19], and bacteria [20,21]. As in higher eukaryotes, some of these complexes also include nonsynthetase accessory proteins. In yeast, the tRNA- binding protein Arc1p is associated with MetRS and GluRS [17,22,23], and enhances the binding affinity for cognate tRNAs. The association of an aaRS with a puta- tive metabolic protein Mj1338 has recently been identi- fied in archaea [19], whereas Trbp111 and CsaA are the best known bacterial tRNA chaperones [24–26]. On the other hand, the recent discovery of autonomous editing proteins that are efficient at hydrolyzing misacylated products provides a direct link between ancestral aaRSs consisting solely of the catalytic core and extant enzymes to which functionally independent modules are appen- ded [27,28]. Factors unrelated to the translation machin- ery have also been found to associate with aaRSs. Yeast TyrRS interacts with Knr4, a protein involved in cell wall synthesis [29], and heat shock protein 90 interacts with human glutamyl-prolyl-tRNA synthetase and medi- ates protein–protein interactions during the association of several human synthetases [30,31]. We previously identified the yeast peroxin Pex21p as a protein that interacts with seryl-tRNA synthetase (SerRS) (EC 6.1.1.11), and this was confirmed by an in vitro binding assays using truncated Pex21p fused to glutathione-S-transferase [32]. Pex21p is part of a two- member peroxin family (Pex18p and Pex21p) specific- ally required for peroxisomal targeting of the Pex7p peroxisomal signal recognition factor and import of peroxisomal targeting signal 2-type peroxisomal matrix proteins [33,34]. The observation that yeast (Saccharo- myces cerevisiae, Sc) SerRS, like all eukaryotic cytosolic SerRSs, contains a dispensable C-terminal extension that influences the enzyme’s stability and possibly its substrate recognition properties [35,36] prompted us to investigate possible interactions between this domain and SerRS. Our results revealed that the interaction with the peroxin is mediated by a short C-terminal peptide (consisting of 13–20 amino acids) that facilitates formation of a ternary complex with tRNA Ser . Results Pex21p participates in ternary complex formation with SerRS and tRNA Ser As shown by our previous experiments, the C-terminal extension was dispensable for cell viability [35], but truncated SerRS was unstable and displayed somewhat altered kinetic parameters towards its substrates [36]. In contrast, the interaction with Pex21p slightly increa- ses serylation by the full-length synthetase [32]. There- fore, we investigated the potential involvement of Pex21p in SerRS substrate binding. We were partic- ularly interested in whether the peroxin mediates cap- ture or positioning of tRNA by the synthetase, and whether it interacts directly with tRNA Ser . As sequence analysis of Pex21p did not reveal a tRNA-binding motif, it seemed unlikely that Pex21p was a tRNA- binding protein. On the other hand, an RNA-binding motif could be hidden due to its bipartite nature, as in other cis-acting or trans-acting synthetase cofactors, where two or more relatively weak tRNA-binding regions work in synergy [7,37–39]. It seemed plausible that both the synthetase and protein cofactor could bind tRNA, as interactions with two proteins might enable more efficient tRNA turnover. We coincubated tRNA Ser and the proteins at 25 °C, and then per- formed gel-shift analysis, monitoring by western blot whether the rate of protein migration was increased or shifted upon tRNA binding. Whereas SerRS (Fig. 1, lane 1) forms stable binary complexes both with cog- nate tRNA Ser (Fig. 1, lane 2) and with Pex21p (Fig. 1, lane 4), we did not observe Pex21p–tRNA Ser interac- tions by the gel-shift analysis (Fig. 2A, lane 3). In this experimental setup, it is not possible to better separate SerRS and SerRS–Pex21p complex (Fig. 1, compare lanes 1 and 4), but complex formation was confirmed by several other approaches [32]. The addition of Fig. 1. Gel-shift analysis of SerRS and Pex21p protein with tRNA Ser or total yeast tRNA. Full-size His 6 -Pex21p was purified from Escheri- chia coli and SerRS was purified from yeast and incubated with tRNA Ser transcript or total yeast tRNA in the absence or presence of Pex21p. Complexes were subjected to nondenaturing polyacryla- mide gel electrophoresis, transferred to nitrocellulose membrane, incubated with antibodies to SerRS, and visualized by chemilumi- nescence. The dotted arrow indicates a faster-migrating SerRS– tRNA Ser complex. The full arrow indicates a Pex21p–SerRS–tRNA Ser supershifted complex. V. Godinic et al. Interaction between SerRS and Pex21p FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS 2789 Pex21p further retarded the complex between SerRS and its cognate tRNA (Fig. 1, lane 2), indicative of ternary complex formation (Fig. 1, lanes 5 and 6). In support of this finding, a retarded complex was observed with antibodies against the His tag in His 6 - Pex21p (not shown). A ternary complex was also detected after coincubation of SerRS and Pex21p with total yeast tRNA, as indicated by the occurrence of a supershifted band of the same intensity and mobility (Fig. 1, lane 6). A supershifted band denoting a ternary complex was also obtained in the experiment conducted with 3¢-radiolabeled tRNA Ser transcript (Fig. 2A, lane 4). Equal amounts of Pex21p and SerRS were preincuba- ted for 10 min, prior to addition of 32 P-labeled tRNAs. Reaction mixtures were subjected to electrophoresis on a nondenaturing polyacrylamide gel. Free and protein- bound tRNA was visualized by radiography, and the ternary complex was only detected with cognate tRNA Ser . Faint bands in the upper part of the gel are artefacts of macromolecular aggregation occurring upon entry into the gel. Yeast tRNA Phe , which was used as a control, did not affect the migration of the protein binary complex (Fig. 2A, lane 5), suggesting that ternary complex formation was specific for tRNA Ser . The formation of binary and ternary complexes was assayed under various ionic conditions. Magnesium ions are of great importance in stabilizing the tRNA tertiary structure, and influence aaRS–tRNA complex formation; as a result, the complex between SerRS and tRNA Ser cannot be detected at low Mg 2+ concentra- tions [40]. In agreement, the complex obtained at 150 mm sodium chloride and 2.5 mm MgCl 2 was rather weak (Fig. 2A, lane 1), and it seems to be sta- bilized by addition of Pex21p. When the ionic strength was reduced to 30 mm NaCl and the concentration of Mg 2+ was increased (8 mm MgCl 2 ), the stability of the binary complex was increased (Fig. 2B). Pex21p increases tRNA binding by SerRS To further investigate the influence of Pex21p on tRNA binding by SerRS, we compared the extent of tRNA binding by isolated enzyme and by the SerRS–Pex21p binary complex. Labeled tRNA Ser was mixed with increasing amounts of preformed SerRS–Pex21p com- plex, and the resulting tRNA–protein complexes were resolved and quantified (the values for bound tRNA were estimated by deducting uncomplexed tRNA from total tRNA). The percentage of bound tRNA Ser was plotted as a function of enzyme concentration (Fig. 3), and indicated that tRNA binding by SerRS was eleva- ted in the presence of Pex21p. Mapping the Pex21p interaction domain in SerRS Eukaryotic cytosolic SeRSs comprise positively charged noncatalytic peptides appended to conserved catalytic cores. As shown in Fig. 4, extensions are characterized by high lysine content, but their primary sequences are not conserved. In order to map the interaction domain of SerRS that is involved in complex assembly with Pex21p, we prepared a number of truncated yeast SerRS vari- ants (Fig. 5A). By aligning the primary structures of A B Fig. 2. (A) Formation of a ternary complex, Pex21p–SerRS–tRNA Ser . Equal amounts of Pex21p and SerRS were preincubated for 10 min prior to addition of 3¢-radiolabeled tRNA Ser in binding buffer contain- ing 150 m M sodium chloride and 2.5 mM MgCl 2 . The complexes were analyzed by 5% native PAGE, and submitted to phosphorimag- ing. Lane 1, bottom bracket: free tRNA Ser. Lane 2, middle bracket: tRNA Ser shifted in SerRS–tRNA Ser complex (the stability of this complex is discussed further in the text). Lane 3: labeled tRNA Ser with Pex21p added in a binding reaction. Lane 4, upper bracket: larger ternary complex Pex21p–SerRS–tRNA Ser trapped at the top of the gel. Lane 5: SerRS and Pex21p in a binding reaction with radio- labeled tRNA Phe used as a control. Lane 6: tRNA Phe only. (B) tRNA Ser shifted in the SerRS–tRNA Ser complex in the presence of 30 mM NaCl and 8 mM MgCl 2 . Lane 1, bottom bracket: radiolabeled tRNA Ser. Lanes 2–5, upper bracket: SerRS–tRNA Ser complex. The concentra- tion of 32 P-tRNA Ser was 0.15 lM, and the concentration of SerRS in lanes 2–5 was 0.4, 0.7, 1.0 and 2.5 l M, respectively. In lane 5, almost all of the tRNA Ser was sequestered in SerRS–tRNA Ser complex. Interaction between SerRS and Pex21p V. Godinic et al. 2790 FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS available SerRS proteins, and on the basis of the struc- tural data available for Escherichia coli and Thermus thermophilus SerRS [41,42], we attempted to delete the C-terminal extension and subsequently the struc- tural motifs III, II, and I. Finally, the shortest trun- cated protein (ScSerRSD356) comprised only the presumed N-terminal coiled-coil domain (denoted COIL in Fig. 5A). To investigate the role of the C-terminal extension, two truncated mutants were designed. ScSerRSDC20 lacked the 20 C-terminal amino acids, whereas in ScSerRSDC13, only the fragment of 13 amino acids (containing seven lysines) was cut off. Truncated yeast SerRS constructs were fused to the C-terminal end of the LexA DNA-binding domain (LexAbd), yielding a series of LexA-fusion proteins, which were then used as baits in the two-hybrid assay. Upon transformation of baits into the yeast strain L40 coexpressing full- length Pex21p fused to the transcription activator GAL4-activation domain (GAL4ad), which yields GAL4–Pex21p, they were tested in a yeast two-hybrid system to assay for Pex21p binding in vivo. The fusion proteins denoted LexA–ScSerRS, LexA– ScSerRSDC13, LexA–ScSerRSDC20, LexA–ScSerRSDC82, Fig. 3. Binding of tRNA Ser to free SerRS and to the preformed Pex21p–SerRS complex. Labeled in vitro-transcribed tRNA Ser (50 nM) was titrated with increasing concentrations of SerRS (0.5–2.5 l M)in the presence of BSA (10 l M; j) or Pex21p (2 lM; m) in binding buffer, and electrophoresed in 5% native polyacrylamide gel (as des- cribed in Experimental procedures). The curve corresponds to the percentage of bound tRNA Ser as a function of enzyme concentration. The results were scanned using a phosphorimager and analyzed by IMAGEQUANT software. The data were fitted to a single-site binding equation. Fig. 4. Alignment of C-terminal regions of SerRS proteins from different domains of life: AA, Aquifex aeolicus; BS, Bacillus subtilis; LP, Leg- ionella pneumophila; EC, E. coli; TD, Thiobacillus denitrificans; TT, Thermus thermophilus; TP, Thermofilum pendens; AP, Aeropyrum pernix; PF, Pyrococcus furiosus; HS, Halobacterium salinarum; MT, Methanothermobacter thermautotrophicus; HS, Homo sapiens; BS, Bos taurus; MM, Mus musculus; GG, Gallus gallus; AT, Arabidopsis thaliana; ZM, Z. mays; CA, Candida albicans; SC, S. cerevisiae; SP, Schizosaccha- romyces pombe. Only eukaryotic cytosolic enzymes contain positively charged C-terminal extensions. The C-terminal sequence of S. cerevis- iae and Z. mays cytosolic SerRSs is shown in bold letters. The sequence truncated in the yeast SerRSDC13 mutant is underlined. V. Godinic et al. Interaction between SerRS and Pex21p FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS 2791 LexA–ScSerRSDC202 and LexA–ScSerRSDC356 were stably expressed in yeast, as confirmed by western blot using antibodies to LexA (Fig. 5B). In order to test the spe- cificity of t he S erRS–Pex21p i nteraction, the f ull-length maize (Zea mays cytosoli c, Zmc) Se rRS (LexA–Zmc SerRS) and its truncated varian t (LexA–Zm cSerRSDC26), lacking 26 C-terminal amino acids, were also analyzed. In a western blot, equal amounts of protein extract were loaded per well for each construct, indicating that the level of expression was the same for all d eletion mutants and full- length SerRSs, except for LexA–ScSerRSD35 6 and LexA– ZmcSerRS, whose expression was h igher (Fig. 5B). The basic C-terminal extension of yeast SerRS is required for interaction with Pex21p in vivo To test activation of the reporter gene HIS3, transfor- mants were inspected for growth after 4 days of incu- bation on minimal plates (–Trp-Leu-His) containing 30 mm 3-amino-1,2,4-triazole (Fig. 6A,B, left panels). To verify the interaction and subsequently quantify it, we tested the activation of the second reporter gene lacZ by colony-lift filter assays (Fig. 6A,B, middle pan- els). The expression of the lacZ reporter gene, indica- ting the strength of protein–protein interactions, was quantified using 2-nitrophenyl b-d-galactopyranoside (Gal-ONp) as a substrate for b-galactosidase (Fig. 6C). Screening variants for 3-amino-1,2,4-triazole resistance and a lacZ-positive phenotype revealed that C-terminal truncation of SerRS abolished Pex21p binding. The interaction is lost upon deletion of the positively charged C-terminal fragment (mutant C13) within the extension. The deletion of the whole C-terminal exten- sion of yeast SerRS also resulted in noninteracting trun- cated SerRS (mutant C20). Accordingly, all other more truncated variants (see deletion scheme in Fig. 5A) failed to interact with Pex21p (Fig. 6A,C). Interestingly, maize cytosolic SerRS, which also possesses the C-ter- minal extension, did not interact with Pex21p in the two-hybrid assays (Fig. 6B). It is pertinent to note that positively charged C-terminally appended fragments of yeast and maize SerRSs do not share sequence homo- logy. Thus, the C-terminal domain of yeast SerRS func- tions as the specific Pex21p-binding site. Truncation of a short C-terminal fragment of SerRS abolishes Pex21p binding without affecting catalytic activity A pull-down assay was used to verify that Pex21p spe- cifically recognizes the C-terminal appendix of yeast SerRS in vitro.Ni 2+ –nitrilotriacetic acid agarose was A B Fig. 5. Schematic representation and west- ern blot analysis of full-length and deletion constructs of yeast (S. cerevisiae, Sc) and maize cytosolic (Z. mays, Zmc) SerRS. (A) The names of the constructs used as baits in the two-hybrid system are shown on the left. The number of truncated amino acids from the C-terminus of full-length ScSerRS or ZmSerRS is indicated. (B) Western blot of protein extracts comprising fusion proteins in the L40 yeast strain. Lane 1: LexAbd (25.5 kDa). Lane 2: LexA–ScSerRSDC356 (38 kDa). Lane 3: LexA–ScSerRSDC202 (55.3 kDa). Lane 4: LexA–ScSerRSDC82 (69.4 kDa). Lane 5: LexA–ScSerRSDC20 (76.5 kDa). Lane 6: LexA–ZmcSerRS (a full- length Z. mays cytosolic SerRS; 77.2 kDa). Lane 7: LexA–ZmcSerRSDC26 (74.9 kDa). Lane 8: S. cerevisiae strain L40, nontrans- formed. Lane 9: LexA–ScSerRS (a full-length S. cerevisiae SerRS; 78.8 kDa). Lane 10: LexA–ScSerRSDC13 (77.3 kDa). Calculated molecular masses of the LexA-fusion proteins are in parentheses. Interaction between SerRS and Pex21p V. Godinic et al. 2792 FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS saturated with crude E. coli extract containing recom- binant Pex21p, which was immobilized on the resin by its N-terminal His-tag. Ni 2+ –nitrilotriacetic acid agarose precharged with Pex21p was incubated with yeast or maize SerRS variants. Proteins bound to resin were eluted with buffer containing 300 mm imidazole and analyzed by SDS ⁄ PAGE followed by Coomassie Brilliant Blue staining (Fig. 7). As expected, only full-length yeast SerRS binds to resin previously saturated with Pex21p (Fig. 7, lane 6). Truncated yeast SerRSDC13 or maize SerRS were not pulled down on the resin precharged with Pex21p (Fig. 7, lanes 7 and 8). In the absence of Pex21p, purified SerRS deriva- tives were also not retained on the resin (Fig. 7, lanes 3 and 4). To ensure that SerRSDC13 was stable and correctly folded in pull-down experiments, we determined kinetic parameters for full-length Saccharomyces cerevisiae (Sc)SerRS and truncated enzyme in standard amino- acylation reactions. K m and k cat values with respect to both tRNA and serine (K m ¼ 51±6lm and k cat ¼ 0.47 ± 0.02 s )1 for serine; K m ¼ 0.65 ± 0.07 lm and k cat ¼ 0.54 ± 0.02 s )1 for tRNA Ser ) were unchanged as compared to those obtained for intact ScSerRS (K m ¼ 47 ± 5 lm and k cat ¼ 0.46 ± 0.02 s )1 for serine; K m ¼ 0.71 ± 0.11 lm and k cat ¼ 0.58 ± 0.03 s )1 for A B C Fig. 6. In vivo interaction of Pex21p with homologous and heterologous SerRS variants. Yeast strain L40 coexpressing GAL4–Pex21p and full-length or C-terminally truncated yeast (A) or maize (B) LexA–SerRS variants were plated onto medium that lacks histidine to test for histidine prototrophy (far left plates). The same transformants were also subjected to b-galactosidase colony-lift filter assay using Gal-X as a substrate (see middle plates). The appearance of a blue color indicates protein–protein interactions. Yeast cells were transformed with various constructs as indicated on the right part of the panel, which shows the orientation on the plates that were tested. b-Galactosidase activity was quantified using the Gal-ONp assay (C). The bars indicate Miller units showing the strength of interaction of bait proteins with GAL4–Pex21p. LexA ⁄ GAL4–Pex21p and ScSerRS–GAL4 were used as controls. V. Godinic et al. Interaction between SerRS and Pex21p FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS 2793 tRNA Ser ). As ScSerRSDC13 is catalytically fully active, we assume that the overall structure of SerRS is unaf- fected by deletion of its C-terminal appendix. The similarity of kinetic parameters for the full-length SerRS and the ScSerRSDC13 truncation mutant indi- cates that the C-terminal peptide, composed of 13 amino acids, was not important in substrate recogni- tion, but was primarily involved in protein cofactor binding. Thus, 13 C-terminal amino acids of yeast SerRS function as the binding domain for Pex21p, as revealed by yeast two-hybrid and pull-down assays. Discussion Role of the C-terminal extensions of eukaryotic SerRSs We previously identified the peroxin Pex21p as an inter- action partner of SerRS [32]. Here we examined whether the positively charged C-terminal region of the synthe- tase, a characteristic of all eukaryotic SerRSs, is involved in protein binding. Our studies revealed that the removal of 13 amino acids from the C-terminus gen- erates a stable ScSerRSDC13 variant that exhibits essen- tially the same kinetic parameters for serylation as the wild-type, but has lost the ability to interact with Pex21p. Thus, in contrast to the idiosyncratic extensions of several eukaryotic aaRSs, which serve as additional tRNA-binding domains, due to their high lysine content and overall positive charge [6,7,43], the yeast SerRS C- terminal extension does not participate directly in sub- strate binding, but instead mediates protein binding. Whereas SerRS binds both tRNA and Pex21p, the stoichiometry of the ternary complex is not yet known. In agreement with our previously reported findings that dimeric yeast SerRS binds cognate tRNA Ser anti- cooperatively [44,45], and that the serylation efficiency is moderately enhanced when approximately two mole- cules of Pex21p are bound per dimeric SerRS [32], our current model suggests cross-subunit binding of one tRNA per two SerRS subunits, whereas each subunit interacts with one molecule of Pex21p. In this complex, Pex21p may possibly contribute to tRNA binding by enhancing the affinity of the enzyme for the second tRNA molecule. However, whether the peroxin affects the affinity, stability or stoichiometry of the heterotri- meric Pex21p–SerRS–tRNA Ser complex remains an open question. Role of Pex21p in enhancing cognate tRNA binding by SerRS Pex21p does not bind tRNA Ser (Fig. 2A, lane 3) or any other component of total yeast tRNA (not shown), and thus cannot be considered a tRNA-binding protein. On the other hand, SerRS binds both cognate tRNA Ser and Pex21p, resulting in the formation of a ternary complex. The Haemophilus influenzae YbaK protein, which hydrolyzes misacylated Cys-tRNA Pro in trans, also appears to lack specific tRNA recognition capabil- ity [45], but, like Pex21p, forms a binary complex with the corresponding synthetase (ProRS) and a ternary complex with the synthetase–tRNA pair. It is possible that Pex21p induces a conformational change in the synthetase, facilitating additional contacts between the tRNA and SerRS. Moreover, higher levels of tRNA (15–20%) were found in the heterotrimeric than in the binary complex, in agreement with our previous obser- vation that Pex21p moderately stimulates the amino- acylation efficiency of SerRS [32]. This suggests that Pex21p-induced contacts between tRNA and the syn- thetase may contribute to tRNA binding, in synergy with the major interaction between the N-terminal a-helical coiled coil of SeRS and the long extra arm of tRNA Ser . The failure to obtain discrete upshifted bands in the gel mobility shift assay, which is probably a consequence of the rapid dissociation kinetics of the components in the complex, precluded K d determin- ation, and the precise mechanism by which ternary Fig. 7. Pull-down assay of SerRSs on Ni 2+ –nitrilotriacetic acid agarose precharged with crude E. coli extract containing recombin- ant His-tagged Pex21p. Lanes 1 and 2: purified yeast SerRS (lane 1) and ZmcSerRS (lane 2). Lanes 3 and 4: proteins eluted from Ni 2+ – nitrilotriacetic acid agarose saturated with crude E. coli extract con- taining no Pex21p and incubated with purified yeast SerRS (lane 3) or maize SerRS (lane 4). Lanes 5–8: proteins eluted from Ni 2+ – nitrilotriacetic acid agarose precharged with crude E. coli extract con- taining recombinant His-tagged Pex21p (lane 5) and incubated with full-length yeast SerRS (lane 6), truncated SerRSDC13 (lane 7) or maize SerRS (lane 8). Only full-length yeast SerRS binds to Pex21p immobilized on Ni 2+ –nitrilotriacetic acid agarose (lane 6). Nonspecific adsorption of yeast SerRS or ZmcSerRS to Ni 2+ –nitrilotriacetic acid agarose was not observed (lanes 3 and 4, respectively). Positions of molecular mass markers are indicated on the left. Interaction between SerRS and Pex21p V. Godinic et al. 2794 FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS complex formation stimulates serylation remains unclear. In vivo, the lack of Pex21p does not signifi- cantly perturb peroxisomal biogenesis, as it is function- ally redundant with Pex18p [34]. The level of SerRS in a Dpex21 strain is only slightly decreased compared to the wild-type, as quantified by immunoblot analysis (data not shown). It seems, therefore, that the two genes are not coregulated, and that Pex21p does not act as a transactivator of SES1 gene transcription. Available RNA microarray data [47–49] indicate that Pex21p is overexpressed under oxidative stress condi- tions [50,51], whereas the level of SerRS is decreased. On the other hand, upon stress-related overexpression of Pex21p, the peroxin may bind SerRS (or SerRS– tRNA Ser ) and target this component of the transla- tional machinery for degradation by an as yet unknown mechanism. Interestingly, in addition to the catalysis of the aminoacylation reaction, SerRS has also been found to participate in the synthesis and turnover of diadenosine oligophosphates (Ap n A) [52], which plays an important role in the response of bacterial and eukaryotic cells to a variety of stress conditions. These adenylylated nucleotides may be alarmones, i.e. regula- tory molecules, alerting cells to the onset of oxidation stress [53]. Therefore, aaRSs could be important coor- dinators in stress signaling networks. Experimental procedures Plasmid constructions The full-length gene encoding S. cerevisiae SeRS was inserted in-frame to the 3¢-end of the coding sequence of the tran- scription factor LexA-binding domain (LexAbd) in the yeast expression vector pAB151 [32]. Furthermore, all truncated constructs for bait proteins were created using PCR, and checked by sequencing. Inserted genes for C-terminal-dele- tion mutants were expressed in-frame with the LexA DNA- binding domain. The gene for Pex21p was cloned in pET15b (Novagen, Madison, WI, USA) for protein purification, and in pACT2 (Clontech, Mountain View, CA, USA) for two- hybrid analysis. Full-length and truncated SerRS genes were inserted in pCJ11 for protein overexpression and purifica- tion. The Zea mays SerRS gene was inserted in pET28b (Novagen). Yeast two-hybrid analysis Recombinant plasmids were introduced by the lithium acetate transformation procedure as previously described [32] into the L40 S. cerevisiae strain (MATa his3-D200 trp1-D901 leu2–3, 112 ade2 LYS2::(lexAop) 4 -HIS3 URA3:: (lexAop) 8 -lacZ), kindly provided by I. Stagljar (Department of Biochemistry, University of Toronto, Canada). Trans- formants were allowed to grow at 30 °C for 2–3 days on – Trp-Leu (absence of the amino acids Trp and Leu) plates, and then transferred to selective – Trp-Leu-His (absence of the amino acids Trp, Leu and His) plates, with 30 mm 3-amino-1,2,4-triazole for testing activation of the reporter gene HIS3. Then, transformants were tested for b-galactosidase activity by colony-lift filter assay using 5-bromo-4-chloro-3-indolyl-b-d-galactopyranoside (Gal-X). The filters were incubated at room temperature, and checked periodically for the appearance of a blue color that developed between 30 min and 8 h. b-Galactosidase activity was quantified using Gal-ONp as a substrate in assays car- ried out according to the manufacturer’s instructions (Clon- tech), and expressed in Miller units under the trial conditions pH 7.0 and temperature 30 °C. Each determin- ation was performed in triplicate. Western blotting To probe the expression levels of fusion proteins, yeast whole cell protein lysates were separated on a 9% SDS ⁄ PAGE gel, transferred to nitrocellulose membrane, and immunoblotted with rabbit anti-LexA (Invitrogen, Carls- bad, CA, USA) and rabbit anti-SerRS raised against yeast SerRS. Anti-LexA was diluted 1 : 5000 (v ⁄ v) and anti- SerRS 1 : 500 in Tris-buffered saline (NaCl ⁄ Tris) contain- ing 0.2% (v ⁄ v) Tween-20. Secondary anti-rabbit IgG (Novagen) conjugated with horseradish peroxidase was diluted 1 : 10 000 (v ⁄ v), and anti-mouse IgG (Novagen) conjugated with horseradish peroxidase was diluted 1 : 5000 (v ⁄ v). Immunoreactive bands were subsequently visualized using chemiluminescence (KPL, Gaithersburg, MD, USA). Nondenaturing gels were also subjected to western blot analysis using antibodies to His-tag (Novagen) for detection of His-tagged Pex21p, or antibodies to SerRS for detection of yeast SerRS. Purification of proteins pET15bPEX21 plasmid was introduced into the bacterial strain BL21(DE3)pLysS (Novagen). Cells were harvested, resuspended in ice-cold lysis buffer [50 mm NaCl, 50 mm Tris ⁄ HCl, pH 7.5, 5% (v ⁄ v) glycerol, 5 mm dithiothreitol and 0.2 mm phenylmethanesulfonyl fluoride], lysed on ice by mild sonication, and centrifuged (10 000 g, 15 min, 4 °C, 6K1s centrifuge, Sigma, Osterode am Hartz, Germany). The lysate was subjected to centrifugation (20 000 g , 30 min at 4 °C, 6K1s centrifuge) to remove cell debris, and protein was purified on Ni 2+ –nitrilotriacetate agarose (Qiagen Inc. Valencia, CA, USA) according to the manufacturer’s proto- col. For analysis of purified Pex21p, aliquots were boiled in sample buffer and loaded onto SDS–polyacrylamide gel. The eluant was dialyzed against 1 L of 20 mm Tris ⁄ HCl V. Godinic et al. Interaction between SerRS and Pex21p FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS 2795 (pH 7.5), 100 mm NaCl, 1 mm dithiothreitol and 20% (v ⁄ v) glycerol, and used for electrophoretic mobility shift assay (EMSA). The overproduction of ScSerRSDC13 and full-length ScSerRS enzymes was achieved as described [34]. The enzymes were purified by a two-step chromato- graphic procedure on FPLC MonoQ and MonoS columns (Pharmacia Biotech Inc., Uppsala, Swedan). Z. mays SerRS was overexpressed in E. coli strain BL21(DE3) after 2.5 h of induction with 1 mm isopropyl thio-b-d-gal- actoside at 30 °C, and purified by ion exchange chroma- tography on a MonoQ HR 10 ⁄ 10 column (Pharmacia Biotech). Aminoacylation assays Reaction mixtures contained 50 m m Tris ⁄ HCl (pH 7.5), 4mm dithiothreitol, 15 mm MgCl 2 ,5mm ATP, 0.4 mgÆmL )1 BSA, variable concentrations of [ 14 C]l-serine (32–40 CiÆmol )1 ), and unfractionated yeast tRNA. Kinetic parame- ters (k cat and K m ) for serine were determined in the presence of 4.3 mgÆmL )1 unfractioned yeast tRNA (containing 5 lm tRNA Ser ) and 10–400 lm [ 14 C]l-serine. For tRNA kinetic parameter determination, the concentration of [ 14 C]l-serine was kept constant (100 lm) and the unfractioned yeast tRNA concentration was varied (0.17–3.4 mgÆmL )1 , corres- ponding to 0.2–4.0 lm tRNA Ser ). The amount of tRNA Ser isoacceptors in unfractioned yeast tRNA was determined from the maximal acceptor activity of unfractioned yeast tRNA towards purified ScSerRS and [ 14 C]l-serine. The enzyme concentration was 7.5–10 nm for serine and 4 nm for tRNA kinetic parameter determination. Reactions were per- formed at 30 °C. Kinetic parameters were calculated from initial velocities for different substrate concentrations using nonlinear regression. EMSA For radioactive EMSA, yeast tRNA Ser transcript was gener- ated and purified as before [54]. The tRNA transcript was charged with [ 14 C]l-serine, and the aminoacylation plateau was measured with homologous SerRS, giving a serine-accept- ing activity of 700 pmol of serylated tRNA per A 260 .The CCA 3¢-end from tRNA Ser was removed prior to the labeling reaction by incubation for 2 h at room temperature with 73 lgÆmL )1 snake venom exonuclease (phosphodiesterase I) from Crotalus atrox (Sigma-Aldrich, St Louis, MO, USA) in 40 mm sodium glycinate (Na-Gly) buffer (pH 9.0) and 10 mm magnesium acetate. The reaction product was extracted with phenol ⁄ chloroform, desalted by gel filtration through a Se- phadex G25 column (Amersham Biosciences, Piscataway, NJ, USA), and precipitated with ethanol. The CCA 3¢-end of the tRNA Ser transcript was reconstituted and labeled with [ 32 P]ATP[aP] by incubation for 10 min at 37 °Cwith0.5lm snake venom-treated tRNA in 50 mm Na-Gly (pH 9.0), 10 mm MgCl 2 ,10lm CTP, 9 lm ATP, 1 lm [ 32 P]ATP[aP] and 3 lgÆmL )1 E. coli tRNA-terminal nucleotidyltransferase in a final volume of 20 lL. The reaction was stopped by the addition of one volume of phenol, and the resulting mixture was gel filtered twice through a G25 column. Prior to complex formation, tRNA Ser was freshly renatured by heating to 80 °C for 2 min; MgCl 2 was then added to 10 mm, and the reaction mixture was further placed on ice. Proteins Pex21p, SerRS (250 nm) and ⁄ or BSA were incu- bated for 10 min at 25 °C in a binding buffer containing 20 mm Tris ⁄ HCl (pH 8.0), 8 mm (or 2.5 mm) MgCl 2 ,30mm (or 150 mm) NaCl, and 5% (v ⁄ v) glycerol. After addition of 32 P-labeled tRNA Ser or 32 P-labeled tRNA Phe (150 nm), incu- bation was continued for an additional 15 min at room temperature in a final binding reaction volume of 25 lL. Finally, tRNA Ser –protein complexes were resolved by elec- trophoresis in a 5% native polyacrylamide gel (37.5 :1 acryl- amide ⁄ bisacrylamide) in 0.5 · Tris-borate buffer (90 mm Tris, pH 8.3, and 65 mm boric acid) with 5 mm MgCl 2 . The gel was vacuum-dried for 30 min in a gel-dryer, and tRNA was detected by Phosphorimager (Amersham Biosciences). For nonradioactive EMSA, the protein concentration was 2 lm, the tRNA concentration was 3 lm, and complexes were formed by coincubating tRNA Ser and proteins for 10 min at 25 °C in the presence of 30 mm KCl, 5 mm MgCl 2 , and 20 mm Tris ⁄ HCl (pH 8.0). After electrophor- esis, proteins were transferred to nitrocellulose membrane and subjected to western blotting using specific antibodies. In vitro binding assay Ni 2+ –nitrilotriacetic acid agarose (Qiagen) was equilibrated with lysis buffer containing 10 mm imidazole, and saturated with E. coli BL21(DE3) crude extract containing expressed His-tagged Pex21p. The resin was washed extensively with buffer containing 40 mm imidazole to remove unbound pro- teins, and this was followed by equilibration with buffer for SerRS binding (40 mm imidazole, 5 mm MgCl 2 , and 5 mm 2-mercaptoethanol). Resin saturated with Pex21p was dis- pensed in small batches (15 lL) and incubated with 30 lgof purified ScSerRS, ScSerRSDC13 or ZmcSerRS for 10 min at room temperature. The resin was thoroughly washed with binding buffer (40 mm imidazole, 5 m m MgCl 2 , and 5 mm 2-mercaptoethanol), and bound proteins were eluted with 300 mm imidazole. The eluate was analyzed by SDS ⁄ PAGE. Nonspecific binding of SerRS to Ni 2+ –nitrilotriacetic acid agarose and the presence of impurities bound to resin were tested on resin saturated with E. coli BL21(DE3) crude extract without any recombinant protein expressed. Acknowledgements This work was supported by grants from the Ministry of Science, Education and Sports of the Republic of Cro- atia (I. Weygand-Durasevic), and the National Institute Interaction between SerRS and Pex21p V. Godinic et al. 2796 FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS of General Medical Sciences (M. Ibba). We thank Pro- fessor I. Stagljar (Department of Biochemistry, Univer- sity of Toronto, Canada) for the yeast two-hybrid strain and plasmid. We are grateful to J. Jaric and J. Ling for generous gifts of tRNA transcripts. References 1 Francklyn C, Perona JJ, Puetz J & Hou JM (2002) Aminoacyl-tRNA synthetases: versatile players in the changing theater of translation. RNA 8, 1363–1372. 2 Park SG, Ewalt KL & Kim S (2005) Functional expan- sion of aminoacyl-tRNA synthetases and their interact- ing factors: new perspectives on housekeepers. Trends Biochem Sci 30, 569–574. 3 Mirande M (1991) Aminoacyl-tRNA synthetase family from prokaryotes and eukaryotes: structural domains and their implications. Prog Nucleic Acid Res Mol Biol 40, 95–142. 4 Cusack S (1995) Eleven down and nine to go. Nat Struct Biol 2, 824–831. 5 Mirande M (2005) Multi-aminoacyl-tRNA complexes. In The Aminoacyl-tRNA Synthetases (Ibba M, Francklyn C & Cusack S, eds), pp. 298–308. Landes Bioscience, Georgetown, TX. 6 Frugier M, Moulinier L & Giege R (2000) A domain in the N-terminal extension of class IIb eukaryotic ami- noacyl-tRNA synthetases is important for tRNA bind- ing. EMBO J 19, 2371–2380. 7 Francin M, Kaminska Kerjan P & Mirande M (2002) The N-terminal domain of mammalian lysyl-tRNA synthetase is a functional tRNA-binding domain. J Biol Chem 277, 1762–1769. 8 Kim S, Landro JA, Gale AJ & Schimmel P (1993) C-terminal peptide appendix in a class I tRNA synthe- tase needed for acceptor–helix contacts and microhelix aminoacylation. Biochemistry 32, 13026–13031. 9 Geslain R & Ribas de Pouplana L (2004) Regulation of RNA function by aminoacylation and editing? Trends Genet 20, 604–610. 10 Ling C, Yao YN, Zheng YG, Wei H, Wang L, Wu XF & Wang ED (2005) The C-terminal appended domain of human cytosolic leucyl-tRNA synthetase is indispen- sable in its interaction with arginyl-tRNA synthetase in the multi-tRNA synthetase complex. J Biol Chem 280, 34755–34763. 11 Cerini C, Kerjan P, Astier M, Gratecos D, Mirande M & Semeriva M (1991) A component of the multisynthe- tase complex is a multifunctional aminoacyl-tRNA synthetase. EMBO J 10, 4267–4277. 12 Filonenko VV & Deutscher MP (1994) Evidence for similar structural organization of the multienzyme ami- noacyl-tRNA synthetase complex in vivo and in vitro. J Biol Chem 269, 17375–17378. 13 Quevillon S, Agou F, Robinson JC & Mirande M (1997) The p43 component of the mammalian multi- synthetase complex is likely to be the precursor of the endothelial monocyte-activating polypeptide II cytokine. J Biol Chem 272, 32573–32579. 14 Rho SB, Kim MJ, Lee JS, Seol W, Motegi H, Kim S & Shiba K (1999) Genetic dissection of protein–protein interactions in multi-tRNA synthetase complex. Proc Natl Acad Sci USA 96, 4488–4493. 15 Tzima E & Schimmel P (2006) Inhibition of tumor angiogenesis by a natural fragment of a tRNA synthe- tase. Trends Biochem Sci 31, 7–10. 16 Harris CL & Kolanko CJ (1995) Aminoacyl-tRNA synthetase complex in Saccharomyces cerevisiae. Bio- chem J 309, 321–324. 17 Deinert K, Fasiolo F, Hurt EC & Simos G (2001) Arc1p organizes the yeast aminoacyl-tRNA synthetase complex and stabilizes its interaction with the cognate tRNAs. J Biol Chem 276, 6000–6008. 18 Praetorius-Ibba M, Rogers TE, Samson R, Kelman Z & Ibba M (2005) Association between Archaeal prolyl- and leucyl-tRNA synthetases enhances tRNA(Pro) aminoacylation. J Biol Chem 280, 26099–26104. 19 Lipman RS, Chen J, Evilia C, Vitseva O & Hou YM (2003) Association of an aminoacyl-tRNA synthetase with a putative metabolic protein in archaea. Biochemis- try 42, 7487–7496. 20 Harris CL (1987) An aminoacyl-tRNA synthetase com- plex in Escherichia coli. J Bacteriol 169, 2718–2723. 21 Buddha MR, Keery KM & Crane BR (2004) An unu- sual tryptophanyl tRNA synthetase interacts with nitric oxide synthase in Deinococcus radiodurans. Proc Natl Acad Sci USA 101, 15881–15886. 22 Simos G, Segref A, Fasiolo F, Hellmuth K, Shevchenko A, Mann M & Hurt EC (1996) The yeast protein Arc1p binds to tRNA and functions as a cofactor for the methionyl- and glutamyl-tRNA synthetases. EMBO J 15, 5437–5448. 23 Simos G, Sauer A, Fasiolo F & Hurt EC (1998) A con- served domain within Arc1p delivers tRNA to aminoa- cyl-tRNA synthetases. Mol Cell 1, 235–242. 24 Morales AJ, Swairjo MA & Schimmel P (1999) Struc- ture-specific tRNA-binding protein from the extreme thermophile Aquifex aeolicus. EMBO J 18, 3475–3483. 25 Nomanbhoy T, Morales AJ, Abraham AT, Vortler CS, Giege R & Schimmel P (2001) Simultaneous binding of two proteins to opposite sides of a single transfer RNA. Nat Struct Biol 8, 344–348. 26 Kawaguchi S, Muller J, Linde D, Kuramitsu S, Shibata T, Inoue Y, Vassylyev DG & Yokoyama S (2001) The crystal structure of the ttCsaA protein: an export-related chaperone from Thermus thermophilus. EMBO J 20, 562–569. V. Godinic et al. Interaction between SerRS and Pex21p FEBS Journal 274 (2007) 2788–2799 ª 2007 The Authors Journal compilation ª 2007 FEBS 2797 [...]... Stability of the complex between yeast seryl-tRNA synthetase and tRNASer under different electrophoretic conditions Croat Chem Acta 77, 599–604 41 Cusack S, Berthet-Colominas C, Hartlein M, Nassar N ¨ & Leberman R (1990) A second class of synthetase structure revealed by X-ray analysis of Escherichia ˚ coli seryl-tRNA synthetase at 2.5 A Nature 347, 249–255 42 Fujinaga M, Berthet-Colominas C, Yaremchuk AD,... cerevisiae due to altered mechanism of substrate recognition FEBS Lett 439, 235–240 37 Kaminska M, Deniziak M, Kerjan P, Barciszewski J & Mirande M (2000) A recurrent general RNA binding domain appended to plant methionyl-tRNA synthetase acts as a cis-acting cofactor for aminoacylation EMBO J 19, 6908–6917 38 Kaminska M, Shalak V & Mirande M (2001) The appended C -domain of human methionyl-tRNA synthetase. .. aminoacyl-tRNA synthetases J Biol Chem 275, 31682–31688 31 Han JM, Lee MJ, Park SG, Lee SH, Razin E, Choi EC & Kim S (2006) Hierarchical network between the components of the multi-tRNA synthetase complex: implications for complex formation J Biol Chem 281, 38663–38667 32 Rocak S, Landeka I & Weygand-Durasevic I (2002) Identifying Pex21p as a protein that specifically interacts with yeast seryl-tRNA synthetase. .. steps of the peroxisomal PTS2 import pathway EMBO Rep 2, 1035–1039 35 Weygand-Durasevic I, Lenhard B, Filipic S & Soll D (1996) The C-terminal extension of yeast seryl-tRNA synthetase affects stability of the enzyme and its substrate affinity J Biol Chem 271, 2455–2461 36 Lenhard B, Praetorius-Ibba M, Filipic S, Soll D & Weygand-Durasevic I (1998) C-terminal truncation of yeast SerRS is toxic for Saccharomyces... AD, Tukalo MA & Cusack S (1993) Refined crystal structure of the seryl-tRNA synthetase from Thermus ˚ thermophilus at 2.5 A resolution J Mol Biol 234, 222–233 43 Hammamieh R & Yang DC (2001) Magnesium ion-mediated binding to tRNA by an amino-terminal peptide of a class II tRNA synthetase J Biol Chem 276, 428–433 44 Gruic-Sovulj I, Ludemann HC, Hillenkamp F, Weygand-Durasevic I, Kucan Z & Peter-Katalinic... synthetase has a tRNA-sequestering function Biochemistry 40, 14309–14316 2798 39 Shalak V, Kaminska M, Mitnacht-Kraus R, Vandenabeele P, Clauss M & Mirande M (2001) The EMAPII cytokine is released from the mammalian multisynthetase complex after cleavage of its p43 ⁄ proEMAPII component J Biol Chem 276, 23769–23776 40 Gruic-Sovulj I, Rokov-Plavec J, Mocibob M, Kamenski T & Weygand-Durasevic I (2004) Stability... Detection of noncovalent tRNA.aminoacyltRNA synthetase complexes by matrix-assisted laser desorption ⁄ ionization mass spectrometry J Biol Chem 272, 32084–32091 45 Gruic-Sovulj I, Landeka I, Soll D & Weygand-Durasevic I (2002) tRNA-dependent amino acid discrimination by yeast seryl-tRNA synthetase Eur J Biochem 269, 5271–5279 46 An S & Musier-Forsyth K (2004) Trans-editing of Cys-tRNAPro by Haemophilus...Interaction between SerRS and Pex21p V Godinic et al 27 Ahel I, Korencic D, Ibba M & Soll D (2003) Trans-editing of mischarged tRNAs Proc Natl Acad Sci USA 100, 15422–15427 28 Korencic D, Ahel I, Schelert J, Sacher M, Ruan B, Stathopoulos C, Blum P, Ibba M & Soll D (2004) A ¨ freestanding proofreading domain is required for protein synthesis quality control in Archaea Proc Natl Acad Sci USA 101,... ª 2007 The Authors Journal compilation ª 2007 FEBS V Godinic et al 51 Lopez-Huertas E, Charlton WL, Johnson B, Graham IA & Baker A (2000) Stress induces peroxisome biogenesis genes EMBO J 19, 6770–6777 52 Belrhali H, Yaremchuk A, Tukalo M, Berthet-Colominas C, Rasmussen B, Bosecke P, Diat O & Cusack S ¨ (1995) The structural bases for seryl-adenylate and Ap 4A synthesis by seryl-tRNA synthetase Structure... synthetase FEMS Microbiol Lett 214, 101–106 33 Purdue PE, Yang X & Lazarow PB (1998) Pex18p and Pex21p, a novel pair of related peroxins essential for peroxisomal targeting by the PTS2 pathway J Cell Biol 143, 1859–1869 34 Einwachter H, Sowinski S, Kunau WH & Schliebs W (2001) Yarrowia lipolytica Pex20p, Saccharomyces cerevisiae Pex18p ⁄ Pex21p and mammalian Pex5pL fulfil a common function in the early . Peroxin Pex21p interacts with the C-terminal noncatalytic domain of yeast seryl-tRNA synthetase and forms a specific ternary complex with tRNA Ser Vlatka. formation of a specific ternary complex with seryl-tRNA synthetase and tRNA Ser , strengthening the interaction of seryl-tRNA synthetase with its cognate tRNA Ser . Abbreviations aaRS,