Crystal structure of a transfer-ribonucleoprotein particle that promotes asparagine formation
2010; Springer Nature; Volume: 29; Issue: 18 Linguagem: Inglês
10.1038/emboj.2010.192
ISSN1460-2075
AutoresMickaël Blaise, Marc Bailly, Mathieu Fréchin, Manja A. Behrens, Frédéric Fischer, Cristiano L. P. Oliveira, H. D. Becker, Jan Skov Pedersen, Søren Thirup, Daniel Kern,
Tópico(s)DNA Repair Mechanisms
ResumoArticle17 August 2010free access Crystal structure of a transfer-ribonucleoprotein particle that promotes asparagine formation Mickaël Blaise Corresponding Author Mickaël Blaise Department of Molecular Biology, CARB Centre, University of Aarhus, Århus C, Denmark Search for more papers by this author Marc Bailly Marc Bailly Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Mathieu Frechin Mathieu Frechin Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Manja Annette Behrens Manja Annette Behrens Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, Denmark Search for more papers by this author Frédéric Fischer Frédéric Fischer Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Cristiano L P Oliveira Cristiano L P Oliveira Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, DenmarkPresent address: Complex Fluids Group, Department of Experimental Physics, University of São Paulo, São Paulo 05314-970, Brazil Search for more papers by this author Hubert Dominique Becker Hubert Dominique Becker Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Jan Skov Pedersen Jan Skov Pedersen Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, Denmark Search for more papers by this author Søren Thirup Søren Thirup Department of Molecular Biology, CARB Centre, University of Aarhus, Århus C, Denmark Search for more papers by this author Daniel Kern Corresponding Author Daniel Kern Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Mickaël Blaise Corresponding Author Mickaël Blaise Department of Molecular Biology, CARB Centre, University of Aarhus, Århus C, Denmark Search for more papers by this author Marc Bailly Marc Bailly Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Mathieu Frechin Mathieu Frechin Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Manja Annette Behrens Manja Annette Behrens Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, Denmark Search for more papers by this author Frédéric Fischer Frédéric Fischer Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Cristiano L P Oliveira Cristiano L P Oliveira Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, DenmarkPresent address: Complex Fluids Group, Department of Experimental Physics, University of São Paulo, São Paulo 05314-970, Brazil Search for more papers by this author Hubert Dominique Becker Hubert Dominique Becker Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Jan Skov Pedersen Jan Skov Pedersen Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, Denmark Search for more papers by this author Søren Thirup Søren Thirup Department of Molecular Biology, CARB Centre, University of Aarhus, Århus C, Denmark Search for more papers by this author Daniel Kern Corresponding Author Daniel Kern Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France Search for more papers by this author Author Information Mickaël Blaise 1, Marc Bailly2, Mathieu Frechin2, Manja Annette Behrens3, Frédéric Fischer2, Cristiano L P Oliveira3, Hubert Dominique Becker2, Jan Skov Pedersen3, Søren Thirup1 and Daniel Kern 2 1Department of Molecular Biology, CARB Centre, University of Aarhus, Århus C, Denmark 2Institut de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, Architecture et Réactivité de l'ARN, Université de Strasbourg, Strasbourg, Cedex, France 3Department of Chemistry, iNANO Interdisciplinary Nanoscience Center, University of Aarhus, Århus C, Denmark *Corresponding authors: Department of Molecular Biology, CARB Centre, University of Aarhus, Gustav Wieds Vej, 10 c, DK-8000 Århus C, Denmark. Tel.: +45 8942 5265; Fax: +45 8612 3178; E-mail: [email protected] de Biologie Moléculaire et Cellulaire, UPR 9002 du CNRS, 15, rue René Descartes, F-67084 Strasbourg, Cedex, France. Tel.: +33 (0)3 88 41 70 92; Fax: +33 (0)3 88 60 22 18; E-mail: [email protected] The EMBO Journal (2010)29:3118-3129https://doi.org/10.1038/emboj.2010.192 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Four out of the 22 aminoacyl-tRNAs (aa-tRNAs) are systematically or alternatively synthesized by an indirect, two-step route requiring an initial mischarging of the tRNA followed by tRNA-dependent conversion of the non-cognate amino acid. During tRNA-dependent asparagine formation, tRNAAsn promotes assembly of a ribonucleoprotein particle called transamidosome that allows channelling of the aa-tRNA from non-discriminating aspartyl-tRNA synthetase active site to the GatCAB amidotransferase site. The crystal structure of the Thermus thermophilus transamidosome determined at 3 Å resolution reveals a particle formed by two GatCABs, two dimeric ND-AspRSs and four tRNAsAsn molecules. In the complex, only two tRNAs are bound in a functional state, whereas the two other ones act as an RNA scaffold enabling release of the asparaginyl-tRNAAsn without dissociation of the complex. We propose that the crystal structure represents a transient state of the transamidation reaction. The transamidosome constitutes a transfer-ribonucleoprotein particle in which tRNAs serve the function of both substrate and structural foundation for a large molecular machine. Introduction The fidelity of translation depends on accurate synthesis of aminoacyl-tRNAs (aa-tRNAs) by the aa-tRNA synthetases (aaRS). Eighteen of the 22 species of homologous aa-tRNAs participating in protein synthesis are formed by direct charging of the amino acid (aa) onto the cognate tRNA. In contrast, the cognate aa-tRNAs pairs charged with asparagine (Asn), glutamine (Gln), selenocysteine (Sec) and cysteine (Cys) are systematically (Sec) or alternatively (Asn, Gln and Cys) formed by a two-step process (Ibba and Söll, 2004; Kern et al, 2005; Sheppard et al, 2008). In these alternative pathways, the cognate aa-tRNA is formed by conversion of an esterified non-cognate amino acid (aa) first attached on the tRNA by a non-cognate aaRS (Curnow et al, 1996, 1998; Becker and Kern, 1998). This pathway raises the question of how the mischarged aa-tRNA intermediate travels from the aaRS to the aa-modifying enzyme. This process must prevent both, the use of the mischarged aa-tRNA in protein synthesis as well as the hydrolysis of the very labile ester bond linking the aa to the tRNA. Recently, we have unraveled the coupling of the two steps of indirect asparaginyl-tRNAAsn (Asn-tRNAAsn) formation in Thermus thermophilus in which the aspartate (Asp) mischarged onto tRNAAsn by a non-discriminating aspartyl-tRNA synthetase (ND-AspRS) is amidated into Asn by a GatCAB tRNA-dependent amidotransferase (Bailly et al, 2007). Both enzymes, together with the uncharged tRNA assemble into a ribonucleoprotein particle (RNP) called transamidosome in which the mischarged Asp-tRNAAsn formed by the ND-AspRS is channelled to the GatCAB that amidates the Asp esterified on the tRNAAsn (Bailly et al, 2007). The assembly of the three partners prevents both the hydrolysis of the labile, mischarged aa-tRNA intermediate and its premature release that could potentially challenge the integrity of the genetic code (Bailly et al, 2007). Given the necessity of the transamidosome for faithful translation of Asn codons, there is a possibility that all so-called indirect pathways of aa-tRNA formation might systematically use particles in which the tRNA, the mischarging aaRS and the aa-modifying enzyme assemble. Indeed, it was recently shown that in methanogenic archaea, tRNA cysteinylation is catalysed by a particle in which the tRNACys substrate is bound to O-phosphoseryl-tRNA synthase (SepRS) and Sep-tRNA•Cys-tRNA synthase (SepCysS) that form Cys-tRNACys (Zhang et al, 2008). Likewise, in mammals, tRNA selenocysteinylation proceeds through formation of supramolecular complexes including the partners that promote formation of Sec-tRNASec, tRNASec modification and incorporation of Sec into polypeptide chains (Small-Howard et al, 2006). Discovery of transfer-ribonucleoprotein (tRNP) particles in which tRNA might act as a scaffold raises a wide range of new structural and functional issues concerning the assembly of the protein and nucleic acid partners, such as the recruitment of the specific substrate by the non-discriminating aaRS, and the mechanism of channelling of the mischarged aa-tRNA from one active site to the other one. To gain insights into which structural elements are involved in the assembly of these tRNPs and in channelling of the aa-tRNA inside these particles, we solved the 3D structure of the T. thermophilus transamidosome by X-ray crystallography and its structure in solution by small angle X-ray scattering (SAXS). The crystal structure represents a transient state of the transamidation reaction. The structural and biochemical data reveal a particle in which the four-bound tRNAsAsn have two different non-overlapping functions: two are substrates and cofactors for Asn formation, whereas the two others constitute catalytically inert but essential building blocks bricks of this molecular machine. The tRNAAsn stabilizes indeed the transamidosome to prevent release of the mischarged aa-tRNA intermediate and also to prevent the GatCAB denaturation at the optimal growth temperature of T. thermophilus. The crystal structure reveals how the ND-AspRS accommodates the tRNAAsn anticodon and suggests that the tRNA acceptor carrying the mischarged Asp undergoes a 35-Å shift to switch from the aminoacylation to the transamidation active site of the tRNP. Results Crystal structure of the transamidosome The structure of the complex was solved by molecular replacement as described (Bailly et al, 2009). The refinement statistics are shown in Table I. The asymmetric unit contains a 520-kDa complex formed by two dimeric ND-AspRSs, two trimeric GatCABs and four tRNAsAsn (Figure 1A–C). The structure of the ND-AspRS resembles those described for other bacterial and eukaryotic AspRSs (Ruff et al, 1991; Delarue et al, 1994). Each subunit of the homodimer contains two main domains. The C-terminal (C-t) catalytic core, organized as a seven-stranded antiparallel β-sheet including the three consensus motifs of class II aaRSs (Charron et al, 2003), catalyses both the activation of Asp in the presence of ATP and its transfer onto the 3′ OH ribose of the terminal adenosine of tRNAAsn. The N-terminal (N-t) anticodon-binding domain (ABD) is an OB fold formed by a five-stranded β-barrel (Charron et al, 2003) (Figure 1). The two domains are linked by a 20-aa-long hinge region essential for the aminoacylation activity, as it ensures inter-domain communication and dimerization of the subunits (Ruff et al, 1991; Delarue et al, 1994). The organization of T. thermophilus GatCAB resembles that of the enzymes from Staphylococcus aureus and Aquifex aeolicus (Nakamura et al, 2006, 2010; Wu et al, 2009). The core of the GatA subunit, which exhibits the amidohydrolase activity, is organized as an 11-stranded β-sheet mixed with α-helices. GatB is formed by two domains connected by a long loop. The N-t part forms a globular cradle domain and the C-t an α-helical domain named Yqey. However, because of the lack of clear electron density, the Yqey domain has been traced as a polyalanine chain (Figure 1E; Supplementary Figure S1). Notably, an anomalous signal peak was detected in each GatB subunit and attributed to a zinc ion tetrahedrally coordinated by residues Cys22, Cys24, Cys38 and Cys41 (Supplementary Figure S2). The zinc ion seems to contribute to the stability of the complex and to the organization of the ammonia channel crossing the GatA and GatB subunits. This channel ensures the transfer of the ammonium ions formed in the amidase site of GatA to the transamidase site of GatB (Nakamura et al, 2006). Finally, the contact area between GatA and GatB is surrounded by GatC, previously described as a chaperone, which forms a belt attaching the two subunits (Figure 1). The four tRNAs of the tRNP display two different binding modes to the GatCABs. The so-called catalytic tRNAAsn (cattRNAAsn) are bound in a functional state with their anticodon loops contacting the ABD of one of the monomers of the ND-AspRS and their acceptor arms buried in the GatB active site of the GatCABs. Both cattRNAsAsn bind the GatCAB and ND-AspRS molecules in the same manner (Figure 1C–F). The second tRNA-binding mode is adopted by the so-called scaffolding tRNAsAsn (scaftRNAAsn), which binds in a non-catalytic manner. Although their anticodon loops bind the ABD of the other ND-AspRS monomer, their acceptor stems are entrapped in the interface between the GatA and GatB subunits of the second GatCAB molecule. Both scaftRNAsAsn bind the GatCAB and ND-AspRS molecules in a same manner (Figure 1C–F). This means that in the complex each GatCAB binds one scaftRNAAsn and one cattRNAAsn (Figure 1D). Because of the lack of a clear electron density, the 3′ OH-CCA end of scaftRNAsAsn was not built. Only a limited number of protein–protein contacts could be observed between the ND-AspRS and the GatCAB. Each GatCAB interacts with the two surrounding ND-AspRS monomers; Arg110, Arg111 and Arg113 from the GatB subunit contact Ser320, Ala365, Lys366, Gly367, from the ND-AspRS monomer by H-bonds or Van der Waals interactions and Glu130, Gly131 and Ala132 from the same GatB contact residues 286–288 of the other ND-AspRS monomer, by Van der Waals interactions (Figure 1G). However, as previously demonstrated, these few protein–protein interactions are not sufficient to mediate complex formation in the absence of tRNA (Bailly et al, 2007). Figure 1.Crystal structure of the transamidosome. (A–C) Composition of the asymmetric unit. The ND-AspRS appears in blue, the cattRNAAsn in red, the scaftRNAAsn in grey and the subunits of the trimeric GatCAB in green, magenta and yellow. The schematic structure follows the same colour code. (D) On this view, one GatCAB, one scaftRNAAsn and one cattRNAAsn have been removed, to show that each ND-AspRS binds one scaftRNAAsn and one cattRNAAsn in order to recruit one GatCAB. (E, F) Structure of the transamidosome in which one ND-AspRS, one cattRNAAsn and one scaftRNAAsn have been removed and which corresponds to that in solution. The dotted circles indicate the GatB catalytic (cradle) and Yqey domains and the ND-AspRS ABD. (G) Protein–protein interactions in the transamidosome. All the components are in grey, except GatB and the two ND-AspRSs that appear, respectively, in yellow and blue. The panels show a zoom of the two regions of GatB and the ND-AspRS in contact and the residues involved in the interaction; the hydrogen bond is indicated by the dashed line. Download figure Download PowerPoint Table 1. Data collection and refinement statistics Data collectiona Beamline X10SA, Swiss Light Source Wavelength (Å) 0.98 Space group P21 Resolution (Å) 50-3 Cell dimensions a, b, c (Å) a=115.9, b=214.0, c=127.8 α, β, γ (deg) α=90, β=93.3, γ=90 Resolution (Å) 50-3 Rmeas 15.4 (80.5) Rmrgd-F 14 (63.9) (I/σ(I)) 10.3 (2.4) Completeness (%) 99.9 (99.9) Redundancy 4.25 Refinement Resolution (Å) 39.18-3.00 No. of reflections 123 924 Rwork/Rfree (%) 21.1/25.1 No. of atoms 33 548 Protein 27 186 Nucleic acids 6224 Ions Mg/Zn 29/2 Water 107 B-factors Protein overall 49.2 Nucleic acids overall 66.2 Ions Mg/Zn 28.3/63.6 Water 56.6 R.m.s.d. Bond lengths (Å) 0.01 Bond angles (deg) 0.95 Ramachandran plotb (%) In core 92.2 Allowed 7.3 a Values in parenthesis are for the last resolution shell. b Values are from Molprobity. Comparison between the crystal structure and the structure in solution determined by SAXS The composition of the transamidosome in solution was previously investigated by size exclusion chromatography, static light scattering, dynamic light scattering and analytical ultracentrifugation (Bailly et al, 2008). All these approaches have predicted a molecular weight for the transamidosome ranging from 300 to 400 kDa, in agreement with a 380-kDa complex, composed of one dimeric ND-AspRS, two tRNAsAsn and two GatCABs but in contradiction with the crystal structure showing the presence of an extra ND-AspRS bound to two tRNAsAsn. Thus, to gain insights into the stoichiometry of the macromolecules that compose the transamidosome in solution, we investigated the particle by SAXS. The pair distribution function p(r) obtained by an indirect Fourier transformation of the experimental SAXS data shows that the transamidosome in solution is an elongated particle with a maximal diameter of 185 Å. The SAXS data give a molecular mass of 325±50 kDa and a radius of gyration of 55±1 Å (Supplementary Figure S3). The size of the model obtained by ab initio methods (Figure 2A and B) suggests that the complex in solution consists, as shown previously, of one dimeric ND-AspRS, two GatCABs and two tRNAs. This is further confirmed by the comparison of the calculated scattering curves for various stoichiometries of the partners with the SAXS data, which gives the best fit for the 380-kDa model (Supplementary Figure S4). Figure 2.SAXS measurements of the transamidosome. (A, B) SAXS ab initio modelling. The black curve represents the best fit from the ab initio modelling to the scattering data (circle). On the right, the average model is represented in black and the crystal structure, the 380-kDa molecule, composed of two dimeric ND-AspRSs (blue), two GatCABs (green, magenta and yellow for the GatC, A and B subunits), one cattRNAAsn (red) and one scaftRNAAsn (grey) is fitted to the SAXS average model. (C, D) Comparison of the crystal structure of the transamidosome to the experimental scattering data. (C) Crystal structure of the transamidosome depleted of one ND-AspRS bound on a cattRNAAsn and a scaftRNAAsn. The two blue arrows are displayed to mark the different orientations of the GatCAB in the crystal structure compared with the model shown in panel E. (D) The circles represent the data and the curve represents the theoretical behaviour of the model calculated with CRYSOL. (E, F) Comparison of the model of the transamidosome to the experimental scattering data. (E) The model represents one ND-AspRS bound to two equivalent GatCABs and two cattRNAsAsn. (F) The circles represent the data and the curve represents the theoretical scattering of the model calculated with CRYSOL. (G, H) SAXS rigid body refinement. (G) Best fit of the rigid body refinement performed using the SASREF program (black curve) to the scattering data (circles). (H) SASREF model. The flexible GatB C-t domain is in blue, the position of the C-t GatB in the crystal structure is in orange and the black arrow illustrates the movement of the GatB C-t domain in the absence of bound tRNA. The colour code is the same as in Figure 1. Download figure Download PowerPoint The position of the two GatCABs and the presence of the scaftRNAAsn in the crystal structure are surprising, as we were expecting that, as proposed previously, each ND-AspRS would bind two cattRNAAsn and that the two GatCABs would be bound symmetrically to these tRNAs (Bailly et al, 2007). Thus, we compared the scattering curves for several models with the SAXS data. The theoretical scattering intensity calculated from the crystal structure of the transamidosome composed of one ND-AspRS bound to one scaftRNAAsn, one cattRNAAsn and two GatCABs gives a reasonable fit to the scattering data (χ=6.8) (Figure 2C and D). In contrast, the theoretical scattering intensity of a model in which the two tRNAsAsn are bound catalytically on the dimeric ND-AspRS and the two GatCABs are equivalent and symmetrical gives a poor fit (χ=8) compared with the model extracted from the crystal structure (Figure 2E and F). As proposed previously, the C-t domain of the GatB subunit may be flexible in the absence of tRNA (Nakamura et al, 2006, 2010). We therefore performed rigid body refinement, using the SASREF program (Svergun and Petoukhov, 2005), in order to fit the GatB C-t domain, with the scattering data. A reasonable fit (χ=3.4) was obtained with a model in which the C-t domain of one GatB shifts closer to the GatB catalytic domain, whereas that of the other GatB is anchored to the bound tRNAAsn. Thus, one of the two GatB C-t domain is not bound to tRNA (Figure 2G and H). Altogether, the SAXS results clearly indicate that in solution the transamidosome comprises one ND-AspRS, two GatCABs and two tRNAsAsn. But as the χ value from model extracted from crystal structure fits with the SAXS data, the results agree with non-equivalent orientations of the two GatCABs. Finally, the rigid body refinement shows that one of the two GatBs has its C-t flexible and therefore it does not bind tRNA in its active site in solution. This tRNA is the scaftRNAAsn. To confirm these observations, we investigated the equivalence of the tRNAsAsn by kinetic experiments. GatCAB takes control of the ND-AspRS catalytic mode inside the tRNP Aminoacylation of tRNAAsn by the preformed ND-AspRS•tRNAAsn complex, in the absence of free tRNA, shows a homogeneous hyperbolic kinetic and formation of two Asp-tRNAAsn with identical rate constants. Thus, both catalytic sites of the dimeric ND-AspRS are equivalent (Figure 3A). The rate constant equals that of tRNA charging at the steady state when free tRNAAsn is added in excess (Figure 3A). In contrast, in the preformed ternary ND-AspRS•tRNAAsn•GatCAB complex, aminoacylation of tRNAAsn occurs through a biphasic kinetic formed by a hyperbolic fast phase followed by a significantly slower linear phase (Figure 3B). The slow linear phase extrapolates at to to one Asp-tRNAAsn formed per dimeric ND-AspRS. The amplitude of the burst increases linearly with the concentration of the transamidosome (not shown) but is not altered in the presence of an excess of free tRNAAsn (Figure 3B). As the rate constants derived from two phases differ by two orders of magnitude (respectively, 0.19 and 0.0018/s; Figure 3B), the stoichiometry of one aa-tRNA formed in the burst by the two active sites of the dimeric ND-AspRS indicates that inside the transamidosome, the AspRS charges one tRNAAsn faster than the second one (Kern and Lapointe, 1979). Thus, the kinetic of Figure 3B is consistent with a non-equivalence of the two tRNAAsn aminoacylation sites of the ND-AspRS triggered by GatCAB binding that contrasts with the equivalent sites displayed by the free enzyme (Figure 3A). With respect to the structure of the transamidosome showing that the ND-AspRS binds one tRNA in a functional state but not the other one (cattRNAAsn and scaftRNAAsn), the biphasic kinetic could reflect the aminoacylation of the two non-equivalently bound tRNAAsn by suggesting that, after fast aspartylation of cattRNAAsn, followed by amidation of the Asp moiety, the scaftRNAAsn is aspartylated slowly with the rate constant equaling that of the binary ND-AspRS•tRNAAsn complex (Figure 3A). Therefore, we propose a working model of the transamidation reaction in which the large complex observed in the crystal structure would represent a transient state of the catalytic process and a snapshot of the transamidation reaction in which two ND-AspRSs saturated with tRNAAsn are bound together to two GatCABs, one of the two ND-AspRS is leaving the complex while the second one is binding to it (Figure 3C). As this complex is not seen in solution, likely binding of a second ND-AspRS•tRNAAsn complex provokes the release of the first one. According to the kinetic investigation, the leaving complex contains the Asn-cattRNAAsn end product and the scaftRNAAsn, whereas the entering one is saturated with uncharged tRNAAsn. Figure 3.Presteady-state and steady-state aminoacylation kinetics of the ND-AspRS•tRNAAsn complex and of the transamidosome and mechanism of transamidation. (A) Aminoacylation catalysed by the binary ND-AspRS•tRNAAsn complex at 25°C. The reactions were conducted in the absence (1) or in the presence (2) of free tRNAAsn. (B) Aminoacylation catalysed by the binary ND-AspRS•tRNAAsn complex (1, 2) and by the transamidosome (3, 4) at 6°C in the presence of 25% glycerol. The reactions were conducted without (1, 3) or with (2, 4) free tRNAAsn. The presteady-state and steady-state rate constants are indicated. (C) A proposed model of the dynamic of the presteady-state and steady-state functioning of the transamidosome. Two GatCABs (green ovals, magenta spheres and yellow ovals for the C, A and B subunits, respectively) bind the dimeric ND-AspRS (blue triangle) saturated by two tRNAAsn (L forms) (1) and induces non-equivalence of the ND-AspRS active sites (2); only one tRNA (cattRNAAsn) is functionally bound (red L) and is asparaginylated, whereas the other one (scaftRNAAsn) not functionally bound (black L) promotes stability of the complex. This complex binds a second dimeric tRNAAsn-bound ND-AspRS (3) promoting the dissociation of the ND-AspRS bound on the newly formed Asn-tRNAAsn (4) that is replaced by a new ND-AspRS•tRNAAsn complex (5). During this exchange, the scaftRNAAsn bound on the second ND-AspRS by maintaining the transamidosome prevents its dissociation, whereas the bound cattRNAAsn is asparaginylated. The dimeric AspRS bound on Asn-tRNAAsn is then exchanged with a novel ND-AspRS•tRNAAsn complex (cycle 2). Steady-state cycles 1 and 2 refer to successive asparaginylation of the cattRNAAsn bound on each dimeric ND-AspRS. The letter N in the blue circles indicates asparaginylation of the cattRNAAsn. Download figure Download PowerPoint What happens with the released ND-AspRS•tRNAAsn complex? As most homologous aa-tRNAs, Asn-tRNAAsn is trapped by the elongation factor EF-Tu and exchanged on the ND-AspRS with free tRNAAsn. The scaftRNAAsn either remains bound on AspRS or is exchanged with another molecule. Thus, a new complex forms in which the ND-AspRS binds equivalently the two tRNAAsn. The strong affinity of tRNAAsn-bound ND-AspRS for GatCAB (Bailly et al, 2007) promotes then binding of the complex on the preformed transamidosome (Figure 3C, steps 4–5) or on free GatCABs to form a new transamidosome, whether the concentration of GatCAB is limiting or exceeds that of the ND-AspRS, respectively. The previous scaftRNAAsn can then bind functionally and become catalytic. If tRNAAsn is aspartylated before entry of the ND-AspRS complex in the transamidosome, the poor affinity of the aspartylated tRNAAsn for EF-Tu with respect to the ND-AspRS contrasting with its strong affinity when bound on the ND-AspRS for GatCAB will considerably restrict its use for protein synthesis. Structural basis of the non-discriminating behaviour of the AspRS The ND-AspRS aspartylates both tRNAAsp and tRNAAsn with similar efficiencies (Becker and Kern, 1998). The structure of the transamidosome reveals how the enzyme accommodates both the 34GUC36 anticodon of tRNAAsp and the 34GUU36 anticodon of tRNAAsn. Previous investigations have suggested that conformation of the L1-loop, connecting strands 4 and 5 of the ABD, is the key element that determines the relaxed specificity towards nucleotide 36 that distinguishes the two anticodons (Schmitt et al, 1998; Charron et al, 2003). Biochemical investigations have revealed that the T. thermophilus ND-AspRS uses G34 and U35 of tRNAAsp as major identity determinants for recognition but not C36, whereas all three nucleotides contribute strongly to recognition by the discriminating AspRS (Kern et al, 2005). Analysis of the interactions between ND-AspRS and tRNAAsn in the transamidosome shows that contacts are made between the ABD and all three nucleotides of the anticodon. The structure of the complex shows that G34 N1 establishes hydrogen bond with the Glu76 carboxyle group, whereas G34 N2 contacts both Asn68 Oδ1 and the carboxyle group of Glu76. In addition, U35 interacts with N3 and O4 of t
Referência(s)