Graphics & Design

A new understanding of the decoding principle on the ribosome

A new understanding of the decoding principle on the ribosome
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  LETTER  doi:10.1038/nature10913 A new understanding of the decoding principle onthe ribosome Natalia Demeshkina 1 * , Lasse Jenner 1 * , Eric Westhof  2 , Marat Yusupov 1 & Gulnara Yusupova 1 During protein synthesis, the ribosome accurately selects transfer RNAs (tRNAs) in accordance with the messenger RNA (mRNA)triplet in the decoding centre. tRNA selection is initiated by elongation factor Tu, which delivers tRNA to the aminoacyltRNA-binding site (A site) and hydrolyses GTP upon establishing codon–anticodon interactions in the decoding centre 1–9 . At thefollowing proofreading step the ribosome re-examines the tRNAand rejects it if it does not match the A codon 2,3,10–14 . It was sug-gested that universally conserved G530, A1492 and A1493 of 16SribosomalRNA,criticalfortRNAbindingintheAsite 15–17 ,actively monitor cognate tRNA 18 , and that recognition of the correctcodon–anticodon duplex induces an overall ribosome conforma-tional change (domain closure) 19 . Here we propose an integratedmechanism for decoding based on six X-ray structures of the 70Sribosome determined at 3.1–3.4A˚resolution, modelling cognateor near-cognate states of the decoding centre at the proofreading step. We showthat the 30S subunitundergoes anidentical domainclosureuponbindingofeithercognateornear-cognatetRNA.Thisconformational changeofthe30Ssubunitformsadecodingcentrethat constrains the mRNA in such a way that the first two nucleo-tides of the A codon are limited to form Watson–Crick base pairs.When U  N G and G N U mismatches, generally considered to formwobble basepairs,areat thefirstorsecondcodon–anticodon posi-tion, the decoding centre forces this pair to adopt the geometry close to that of a canonical C N G pair. This by itself, or with distor-tions in the codon–anticodon mini-helix and the anticodon loop,causes the near-cognate tRNA to dissociate from the ribosome. WedeterminedsixX-raystructuresofthe70Sribosomeat3.1–3.4A˚resolution (Supplementary Tables 1 and 2) programmed by 30-nucleotide-long mRNAs with the AUG codon and tRNA fMet in thepeptidyl tRNA-binding site (P site) and the A site occupied by tRNA 2Leu ortRNA Tyr (Fig.1aandMethods).Inonesetofexperiments,tRNA 2Leu and tRNA Tyr were bound totheirrespectivecognatecodonsCUC and UAC in the A site. In a second set of experiments, wemodelled near-cognate states of the ribosome (Supplementary Fig. 1).These states of the ribosome naturally occur during protein synthesisbutwithlowprobabilitybecausebindingofcognatetRNAiskinetically  1 De´partement de Biologie et de Ge´nomique Structurales, Institut de Ge´ne´tique et de Biologie Mole´culaire et Cellulaire, Illkirch 67400, France; CNRS, UMR7104, Illkirch 67400, France; INSERM, U964,Illkirch67400,France;Universite´ deStrasbourg,Strasbourg67000,France. 2 ArchitectureetRe´activite´ del’ARN,Universite´ deStrasbourg,InstitutdeBiologieMole´culaireetCellulaire,CNRS,Strasbourg67084, France. * These authors contributed equally to this work. mRNA DC30S50SPEA S12C1054C518G530C(+6)G34S12C1054C518G530G34U(+6) A1493G36C(+4) A1492G530 A(+5)U35 A1492G530G(+5)U35 c Cognate U(+6) G34 A-tRNA P-tRNA U(+4)U(+5)G36 A35P/A-kink16S rRNA  f tRNA  2Leu   3 ′ -GAG-5 ′ Near-cognatemRNA/Phe 5 ′ -UUU-3 ′ First A1493U(+4)G36 deha bg tRNA  2Leu   3 ′ -GAG-5 ′ mRNA/Leu 5 ′ -CUC-3 ′ FirstCognatetRNA  Tyr   3 ′ -AUG-5 ′ Near-cognatemRNA/Cys 5 ′ -UGC-3 ′ tRNA  Tyr   3 ′ -AUG-5 ′ mRNA/Tyr 5 ′ -UAC-3 ′ CognatetRNA  2Leu   3 ′ -GAG-5 ′ Near-cognatemRNA/Phe 5 ′ -UUU-3 ′ tRNA  2Leu   3 ′ -GAG-5 ′ mRNA/Leu 5 ′ -CUC-3 ′ SecondSecondThirdThird Figure 1  |  Codon–anticodon interactions in the decoding centre on the 70Sribosome. a  ,ThemRNApathonthe70Sribosomewiththedecoding(DC)areaindicated. b ,Close-upviewofthemRNAP/Akinkwithnear-cognatetRNA 2Leu .Magnesiumionsareingreen. c , d ,Thefirstbasepairsofthenear-cognate( c )andcognate( d )codon–anticodonduplexesandtheirinteractionswithA1493of16SrRNA. e , f  ,Thesecondbasepairsofthenear-cognate( e )andcognate( f  )codon–anticodon duplexes and their interactions with G530 and A1492 of 16S rRNA. g  ,  h , A classical wobble U N G pair ( g  ) versus canonical C N G interactions ( h ); amagnesium ion interacting with the basepairiscoordinatedbyprotein S12 andpart of 16S rRNA. All graphical representations were prepared with PyMol.Probable hydrogen bonds within 3A˚ distance are indicated by dashed lines;2 F  o 2 F  c  electron density maps are contoured at 1.2 sigma. 2 5 6 | N AT U R E | V O L 4 8 4 | 1 2 A P R I L 2 0 1 2 Macmillan Publishers Limited. All rights reserved  ©2012  favoured.Wemadetheribosomeacceptnear-cognatetRNAbygivingitonly one type of tRNA carrying a mismatch to the A codon along withtRNA fMet forthePsite.InthesecomplexestheAsitewasfilledeitherby tRNA 2Leu andcodonUUUwithaU N Gmismatchinthefirstpositionof thecodon–anticodonhelix orbytRNA Tyr andcodonUGCwitha G N Umismatch in the second position.As shown earlier, the mRNA forms a kink between the P and Acodons (P/A kink), a universal feature of the mRNA path on the 70Sribosome that is stabilized by the P site tRNA, 16S ribosomal RNA(rRNA) and magnesium ions 7,20,21 (Fig. 1b and Supplementary Fig. 2).The single mismatch states described above represent bona fide near-cognate complexes expected to have standard wobble U N G and G N Ubasepairs.Atourdataresolution(3.1–3.5A˚)wecanconfidentlyassignthe general base pairing (Supplementary Figs 3 and 4). The electrondensity maps unambiguously demonstrate that U4 and G5 of the Acodons UUU and UGC do not show the anticipated wobble inter-actions with G36 in tRNA 2Leu and U35 in tRNA Tyr , respectively.Instead, U4 N G36 and G5 N U35 at the first and second positions of the codon–anticodon duplexes form base pairs similar to a standardWatson–Crick G N C pair (Fig. 1c, e and Supplementary Fig. 5). G N C-like G N U or G N T pairs have been shown earlier for RNA and DNA inotherX-raystructures 22,23 . WhenU N Gisatthethirdcodon–anticodonposition we observe standard U N G wobble pairing (Fig. 1g, h).Unexpectedly, nucleotides A1493, A1492 and G530 of 16S rRNA inhelix 44 (h44), which contact the first and the second pairs of thecodon–anticodon helix, interact with these unusual U4 N G36 andG5 N U35 pairs identically to the way they interact with canonicalWatson–CrickbasepairsC4 N G36andA5 N U35(Fig.1c–f).Thesefind-ings are in contradiction with studies where these nucleotides weregiven a role as monitors and discriminators of canonical Watson–Crick pairs in the decoding process 18,19 . Our structures show thatG530, A1492 and A1493 form a static part of the decoding centre,defining its spatial and stereochemical properties (Fig. 2a, b).The observed non-wobble U N G differs from previous X-ray studiesthatwerebasedonthe30Ssubunitalone 19 (Fig.2candSupplementary Figs 5 and 6). For example, for the study of the mismatch at the first  A-codon Anticodon A1493 A1492G5303 ′ 5 ′ 3 ′ 5 ′ a b CognateNear-cognateCognateNear-cognateCognate versus first mismatchCognate versus second mismatch Anticodon A1493 A1492G530 A-codon e 1491PARH69h44 A-tRNA  G530  A1493 A19131Å Near-cognate / PARNear-cognatemRNA 3635343 ′ 3 ′ +4Near-cognate / tRNA  Tyr Cognate / tRNA  Tyr +5+6>0.7Å 1Å 373332383536345 ′ 3 ′ G36P/A-kinkP-codon A-codon5 ′ -UU(+4)70S / long mRNA30S / short mRNA mimic c df PAR1.6Å  A1493 A1492B2a A-codon Figure 2  |  Thenatureofthedecodingcentre. a  , b ,Theoverallconformationsof universally conserved G530, A1492 and A1493 of 16S rRNA in the cognatestructures are identical to those in the near-cognate models when themismatches are at the first ( a  ) or second ( b ) codon–anticodon positions. c , Differences betweenthepositionof thefirst uridine in theUUUcodon base-pairedtotheGAGanticodonoftRNA 2Leu fromour70Sstructure andfromthe30S model 19 .  d , Comparison of the anticodon loops of tRNA Tyr in the cognate(red) and near-cognate (cyan) states.  e , Rearrangements of rRNA helices h44and H69 in the near-cognate state upon binding of the aminoglycosideparomomycin (PAR). The near-cognate structures with tRNA Tyr are shown (asimilar effect of PAR is observed with tRNA 2Leu , see also Supplementary Fig.10).  f  , Magnifiedview of thechanges in theA1493 phosphate position inducedby PAR. Superimpositions in  d ,  e  and  f   were performed using 23S rRNA asreference. LETTER RESEARCH 1 2 A P R I L 2 0 1 2 | V O L 4 8 4 | N AT U R E | 2 5 7 Macmillan Publishers Limited. All rights reserved  ©2012  position,30Scrystalssoakedwithananticodonstem-loopoftRNA 2Leu and a hexauridine (U 6 ) mRNA displayed a classical wobble U N G pair.TherethePcodonwasmimickedbythe3 9 -endof16SrRNA,sotheU 6 mRNA could only bind to the A site and downstream, leading to asituation where the mRNA was not covalently linked between the Pand A codons.Bysuperimposing the Asite GAG anticodonsfromournear-cognatetRNA 2Leu structureand the 30Smodelwefoundthatthefirst nucleotide of the A codon is positioned differently (Fig. 2c andSupplementary Fig. 6). Because it does not have the natural restraintcoming from being covalently bound to the P codon, this first nucleo-tide in the A codon has the freedom to move so it can form a wobbleU N G pair. However, in our structure the P/A mRNA kink specifically directs the first nucleotide of the A codon to form Watson–Crick likeinteractions with G36 of tRNA 2Leu (Fig. 1c).The positional restrictions imposed by the decoding centre on thefirst two near-cognate U N G codon–anticodon pairs may result in dif-ferences in geometry compared with the corresponding cognatehelices (Supplementary Figs 7 and 8). Although, the resolution of the data sets does not allow us to determine the exact value for thesedeviations in the base pairs, the tendencies are clear. A noticeablechangeisanincreaseinbucklingatthefirstandthirdcodon–anticodonpositionswiththeU4 N G36mismatch(SupplementaryFig.7)andatthesecond and third positions with the G5 N U35 mismatch (Supplemen-tary Fig. 8). These deviations could disturb the base stacking network within the near-cognate codon–anticodon helices and deform theentire anticodon loop structure 24 (Fig. 2d, Supplementary Fig. 9 andSupplementary Movie 1). This deformation might influence theposition of helix 69 (H69) of 23S rRNA, whose universally conservedA1913 protrudes into the decoding centre 20,21 (Fig. 2e).Additional structures of the near-cognate states determined in thepresenceofthemiscodingaminoglycosideparomomycin(Supplemen-tary Tables 1 and 2) reveal a movement of H69 accompanied by rearrangements of the intersubunit bridge B2a between h44 and H69(Fig. 2e and Supplementary Movie 2). These distortions are mostprobably caused by binding of the antibiotic that strongly shifts theA1493 phosphate group (Fig. 2f). Although this shift does not altermuch the interactions of A1493 with U4 N G36 and A1492/G530 withG5 N U35(SupplementaryFigs7gand8h),theselocalchangesmodulatethe B2a bridge. H69 is displaced towards the tRNA, which probably enhances theinteractionsurface ofH69with the tRNA D-stem. InthepresenceofparomomycinthepositionofH69isclosertothatobservedfor the cognate state (Supplementary Fig. 10). Furthermore, displace-ment of the A1493 phosphate group relaxes the decoding pocket fromthe side of the A codon (Fig. 2e) and changes the deformation of thenear-cognate codon–anticodon helix (Supplementary Movie 2). Thisnovelunderstandingoftheparomomycinactionthereforediffersfromthe previously suggested mechanism in decoding where it was pro-posed to influence the monitoring capabilities of A1492 and A1493(ref. 18). The observed moderate structural rearrangements withparomomycinareconsistentwithitsmeasuredeffectattheproofread-ing step 25 .We find that both near-cognate tRNAs induce rearrangements of the 30S subunit known as domain closure 19 (that is, shoulder move-ment and head rotation) as described for cognate tRNA 21 (Fig. 3a).Thisimpliesthatdomainclosureisaninherentqualityoftheribosomein response to binding of any tRNA to the A site 12 and is prerequisitefor formation of the decoding pocket.Initially, the mechanism underlying the decoding process wasdeduced from pioneer X-ray structures of the isolated 30S subunitwherecrystalsweresoakedwithU 6 andanticodonstem-loopmimick-ing mRNA and tRNA 18,19 . Besides those limitations, all attempts tomodel near-cognate states on the 30S subunit were performed in thepresence of paromomycin, which, on the one hand, stimulated anordered binding of the near-cognate tRNA analogs, but, on the other, 23SNear-cognateCognate16S16SNear-cognateCognateSuperposition23S16Sshoulder a b Vacant 70S In h44H69 A1493 A1492 A1913/flexibleP-tRNA bindingH69 A-tRNA bindingCognateNear-cognateNon-cognate A1493 A1492 A1913 A1492 A1493 A1913 H69 In h44 OutIn h44 Out G530/   syn h18G530/   anti  h18G530/   anti  h18 c d h44h44 h18 S12S12 G 5 ′ -3 ′ -5 ′ -3 ′ -    F   i  r  s   t   S  e  c  o  n  d   T   h   i  r  d   F   i  r  s   t   S  e  c  o  n  d   T   h   i  r  d U AGC mRNA tRNA  UGU AGUC h44h44 h18 S12S12  3 ′ - tRNA  5 ′ - 3 ′ - GUGUU A  h44h44 h18 S12S12  3 ′ - tRNA  5 ′ - 3 ′ - P/A-kink H69H69H69 Cognate, tRNA  2Leu Near-cognate, tRNA  2Leu 3 ′ - h44h44 h18 S12S12  A  3 ′ -5 ′ - 3 ′ -  A UGU tRNA  C h44h44 h18 S12S12  3 ′ - tRNA  5 ′ - 3 ′ -h44h44 h18 S12tRNA 3 ′ -  AGUCGU A UUG H69H69H69 Cognate, tRNA  Tyr Near-cognate, tRNA  Tyr GC 5 ′ - 5 ′ -5 ′ -5 ′ -5 ′ -5 ′ - head S12 Figure 3  |  The principle of decoding. a  , Superposition of 23S rRNA fromnear-cognate and cognate structures with tRNA 2Leu shows identical domainclosure in the near-cognate and cognate states.  b , Representation of conformations of the core nucleotides composing the decoding centredepending on its functional states (data with the vacant 70S ribosome areunpublished).  c ,  d , Illustration of the decoding principle: together with theconstraints imposed on the A codon by the P/A kink coordinated by amagnesium ion (green sphere), the DC (h18, h44 and protein S12 from thesmallsubunitandH69fromthelargesubunit)restrictstheallowedgeometryof the first two nucleotides of the codon. No such restraints are imposed on thethirdbasepair.Anear-cognatetRNAwithG N Uinthefirstorsecondpositionisforced to form Watson–Crick-like base pairs (middle panels). This createsrepulsion or requires energy for tautomerization (shown in pink), which by itself can be the source of the tRNA discrimination. The right panels illustratetheimpossiblesituationwhenstandardwobblebasepairs(showninred)occurin either the first or second positions of the codon–anticodon duplexes. RESEARCH LETTER 2 5 8 | N AT U R E | V O L 4 8 4 | 1 2 A P R I L 2 0 1 2 Macmillan Publishers Limited. All rights reserved  ©2012  made it difficult to distinguish the independent effect of the mismatchonthe30Sdecodingcentre 19 .Thecognateandnear-cognatemodelsof the 70S ribosome described here, with or without paromomycin,together with our previous structures 12,21 , give rise to novel insightsintodecodingwithrevisitedrolesforuniversallyconservednucleotidesG530, A1492 and A1493 of 16S rRNA (Fig. 3b). Although thesenucleotides form extensive contacts to the A-minor groove of acodon–anticodon helix,G530,A1492 and A1493 donot activelysensethe correct Watson–Crick base-pairing geometry and thus do notdiscriminate against near-cognate tRNA.We propose that upon binding of cognate or near-cognate tRNA tothe70Sribosome,thesmallsubunitundergoesdomainclosurearoundthe anticodon loop of the tRNA. The closure results in formation of atight decoding centre that restricts the first two nucleotides of the Acodon to form exclusively Watson–Crick base pairs with the tRNAanticodon(Fig.3c,d).Owingtoourcurrentdataresolution,wecannotprecisely identify the hydrogen bond pattern of the mismatches in thenear-cognatecodon–anticodon helices, but tautomerism is a plausiblechemical mechanism. An alternative explanation for the tRNA dis-crimination source could be repulsion in the U N G pair (Fig. 1c, e).Energy expenditure for formation of tautomers (or repulsion energy)could constitute the sole cause for the very efficient rejection of near-cognate tRNAs by the ribosome 3,4 . Additionally, the observeddeformation of the anticodon may lead to alterations in the codon–anticodon mini-helix and propagate through the rest of the near-cognate tRNA molecule, destabilizing it and causing the tRNA todissociate from the ribosome (Supplementary Fig. 11). This corrobo-rates the idea that evolutionarily tuned sequences of tRNAs play anactive role in the tRNA selection process 26–28 . Recent X-ray structuresof the 70S ribosome with cognate tRNA and elongation factor Tu 7,28 demonstratedthatbindingofananticodonloopoftRNAinthedecod-ing centre is nearly identical to that shown for accommodated tRNA.Thisinformationpromptsustohypothesizethatthesamemechanismdescribed here for the proofreading step governs the initial tRNAselection step. METHODS SUMMARY Ribosomes were purified from  Thermus thermophilus  cells as described before 21 .For all complexes, mRNA, tRNA fMet and tRNA 2Leu (or tRNA Tyr ) were present infivefold stoichiometric excess of the ribosome concentration. Complexes withparomomycin were obtained by including the antibiotic (30 m M) into the incuba-tion mixture. Crystals were grown at 24 u C by sitting-drop vapour diffusion asdescribedbefore 20 .Allcrystalsbelongedtospacegroup P  2 1 2 1 2 1 andcontainedtwomolecules per asymmetric unit. A very low dose mode with high redundancy wasused for data collection 29 . The structure 21 , with tRNAs, mRNA and metal ionsremoved,wasusedforrefinementwithPhenix  30 .Throughoutrefinement,nobase-pair restraints were used between tRNAs and mRNAs to avoid bias towardsstandard base-pair geometry. Full Methods  and any associated references are available in the online version ofthe paper at Received 25 August 2011; accepted 1 February 2012.Published online 21 March 2012. 1. Rodnina,M. V.& Wintermeyer, W.Fidelity ofaminoacyl-tRNA selection ontheribosome:kineticandstructural mechanisms.  Annu. Rev. Biochem.  70,  415–435(2001).2. Rodnina,M. V., Gromadski, K.B., Kothe,U. & Wieden, H.J.Recognition andselection of tRNAintranslation.  FEBS Lett.  579,  938–942 (2005).3. Zaher, H.S. & Green,R. Fidelity atthe molecular level:lessons fromproteinsynthesis.  Cell  136,  746–762 (2009).4. Gromadski,K.B.,Daviter,T.&Rodnina,M.V.Auniformresponsetomismatchesincodon-anticodon complexes ensures ribosomal fidelity. Mol. Cell  21,  369–377(2006).5. Lee,T.H.,Blanchard,S.C.,Kim,H.D.,Puglisi,J.D.&Chu,S.TheroleoffluctuationsintRNAselection bythe ribosome.  Proc. Natl Acad. Sci. USA 104,  13661–13665(2007).6. Li, W. et al.  Recognition ofaminoacyl-tRNA: a common molecularmechanismrevealed bycryo-EM.  EMBO J.  27,  3322–3331 (2008).7. Schmeing, T.M.  et al.  The crystal structure ofthe ribosome bound to EF-Tu andaminoacyl-tRNA. Science  326,  688–694 (2009).8. Schuette,J.C. et al. GTPaseactivationofelongationfactorEF-Tubytheribosomeduringdecoding. EMBO J.  28,  755–765(2009).9. Voorhees,R.M.,Schmeing,T.M.,Kelley,A.C.&Ramakrishnan,V.Themechanismfor activation ofGTP hydrolysis onthe ribosome.  Science  330,  835–838 (2010).10. Ehrenberg,M., Kurland,C. G.& Ruusala, T. CountingcyclesofEF-Tu to measureproofreading intranslation.  Biochimie  68,  261–273 (1986).11. Voorhees, R.M.,Weixlbaumer, A.,Loakes, D., Kelley,A.C. & Ramakrishnan, V.Insightsinto substratestabilizationfrom snapshots of the peptidyl transferasecenter ofthe intact70S ribosome. Nature Struct. Mol. Biol.  16,  528–533 (2009).12. Jenner,L.,Demeshkina,N.,Yusupova,G.&Yusupov,M.Structuralrearrangementsofthe ribosome atthe tRNA proofreadingstep. Nature Struct. Mol. Biol.  17, 1072–1078 (2010).13. Geggier,P. etal. Conformationalsamplingofaminoacyl-tRNAduringselectiononthe bacterial ribosome.  J. Mol. Biol.  399,  576–595(2010).14. Whitford,P.C. etal. Accommodationofaminoacyl-tRNAintotheribosomeinvolvesreversibleexcursions along multiple pathways.  RNA  16,  1196–1204(2010).15. Moazed,D.&Noller,H.F.BindingoftRNAtotheribosomalAandPsitesprotectstwo distinct setsofnucleotidesin16S rRNA.  J. Mol. Biol.  211,  135–145(1990).16. Powers, T.& Noller, H.F.Selective perturbation ofG530of 16 S rRNAbytranslational miscodingagentsand a streptomycin-dependence mutationinproteinS12.  J. Mol. Biol.  235, 156–172 (1994).17. Yoshizawa,S.,Fourmy,D.&Puglisi,J.D.Recognitionofthecodon–anticodonhelixbyribosomal RNA.  Science  285, 1722–1725 (1999).18. Ogle,J.M. etal. RecognitionofcognatetransferRNAbythe30Sribosomalsubunit. Science  292,  897–902(2001).19. Ogle,J.M.,Murphy,F.V.,Tarry,M.J.&Ramakrishnan,V.SelectionoftRNAbytheribosomerequiresatransitionfromanopentoaclosedform. Cell 111, 721–732(2002).20. Selmer,M. etal. Structureofthe70SribosomecomplexedwithmRNAandtRNA. Science  313,  1935–1942(2006).21. Jenner, L. B.,Demeshkina, N., Yusupova,G. & Yusupov, M. Structural aspectsofmessenger RNAreading framemaintenance bythe ribosome.  Nature Struct. Mol.Biol.  17,  555–560 (2010).22. BPS DatabaseofRNABase-pairStructures. Bebenek, K., Pedersen,L. C.& Kunkel, T. A.Replication infidelityvia a mismatchwith Watson–Crickgeometry.  Proc. Natl Acad. Sci. USA 108,  1862–1867 (2011).24. Auffinger,P.&Westhof,E.AnextendedstructuralsignatureforthetRNAanticodonloop.  RNA  7,  334–341 (2001).25. Pape,T.,Wintermeyer,W.&Rodnina,M.V.Conformationalswitchinthedecodingregion of16S rRNAduring aminoacyl-tRNA selection onthe ribosome.  NatureStruct. Biol.  7,  104–107 (2000).26. Cochella, L. & Green,R.An active role for tRNA indecodingbeyondcodon:anticodon pairing.  Science  308,  1178–1180(2005).27. Dale, T. & Uhlenbeck,O. C.Amino acidspecificity intranslation.  Trends Biochem.Sci.  30,  659–665 (2005).28. Schmeing,T.M.,Voorhees,R.M.,Kelley,A.C.&Ramakrishnan,V.HowmutationsintRNAdistantfromtheanticodonaffectthefidelityofdecoding. NatureStruct.Mol.Biol.  18,  432–436 (2011).29. Mueller,M.,Wang,M.&Schulze-Briese,C.Optimalfine w -slicingforsingle-photon-countingpixel detectors.  Acta Crystallogr. D  68,  42–56 (2012).30. Adams, P.D.  et al. PHENIX:a comprehensivePython-based system formacromolecular structuresolution.  Acta Crystallogr. D  66,  213–221 (2010). SupplementaryInformation  is linkedto the online version ofthe paper Acknowledgements  Weare gratefultoC. Schulze-Brieseandthe staffatthe SwissLightSource(Switzerland)forhelpduringsynchrotronX-raydatacollection.WethankS.DuclaudforribosomepreparationandthestaffoftheStructuralBiologyDepartmentcorefacilityatInstitutdeGe´ne´tiqueetdeBiologieMole´culaireetCellulaire,Universite´deStrasbourg. Thiswork was supported by ANR BLAN07-3_190451(to M.Y.),ANR-07-PCVI-0015-01(to G.Y.),Fondation pourlaRecherche Me´dicaleen France (toN.D.) andbythe European CommissionSPINE2. AuthorContributions N.D.andL.J.conductedexperimentsandperformedanalysis.Allauthorsdiscussed the results andcommentedon the manuscript. Author Information  Theatomiccoordinates andstructurefactorsfor the determinedcrystal structures are deposited intheProtein DataBank under accessionnumbers3TVFand3TVE(cognatetRNA 2Leu complex),3UYEand3UYD(near-cognatetRNA 2Leu complex),3UZ3and3UZ1(near-cognatetRNA 2Leu complexwithparomomycin),3UZ6and 3UZ9(cognate tRNA Tyr complex), 3UZGand 3UZF(near-cognate tRNA Tyr complex), and3UZLand 3UZK(near-cognate tRNA Tyr complex with paromomycin).Reprints and permissions information isavailable Theauthorsdeclarenocompetingfinancialinterests.Readersarewelcometocommentonthe online version ofthis article Correspondence andrequestsfor materials shouldbeaddressed to M.Y. ( G.Y.( LETTER RESEARCH 1 2 A P R I L 2 0 1 2 | V O L 4 8 4 | N AT U R E | 2 5 9 Macmillan Publishers Limited. All rights reserved  ©2012  METHODS Complexformationandcrystallization. Ribosomeswerepurifiedfrom Thermusthermophilus  cells as described before 21 . The 30-nucleotide-long mRNA con-structs I–IV (see below) were purchased from Dharmacon. In all sequences theAUG start codon is underlined and the A codons are in bold. Purified nativeuncharged  Escherichia coli  tRNA fMet , tRNA 2Leu and tRNA Tyr were supplied by Chemical Block. Aminoglycoside antibiotic paromomycin was purchased fromSigma-Aldrich.The cognate and near-cognate complexes were formed in 10mM tris-acetate,pH 7.0, 40mM KCl, 7.5mMMg(Ac) 2 , 0.5mMDTTby incubating 70Sribosomes(3 m M) with mRNA-I, -II, -III or -IV and tRNA fMet for 10min at 37 u C. ThentRNA 2Leu and tRNA Tyr were added to the mixtures with mRNA-I or -II andmRNA-III or -IV, respectively, and the complexes were further incubated for30min. For all complexes mRNA, tRNA fMet and tRNA 2Leu (or tRNA Tyr ) werepresent in fivefold stoichiometric excess of the ribosome concentration.Complexes with paromomycin were obtained by including the antibiotic(30 m M) into the incubation mixture containing mRNA-II with tRNA 2Leu andmRNA-IV with tRNA Tyr . Crystals were grown at 24 u C by sitting-drop vapourdiffusion as described before 20 . In agreement with previous results, initiatortRNA fMet was found in the P site of all complexes and either tRNA 2Leu ortRNA Tyr was found in the A site (tRNA fMet was easily distinguishable fromtRNA 2Leu and tRNA Tyr based on the large variable loops in those tRNAs). Inthe complexes with mRNA III and IV, the E site was occupied by tRNA Tyr .However, in complexes with mRNA I and II, the quality of the density did notallow identification of the E site tRNA which was then modelled as tRNA fMet .mRNA-I 5 9 -GGCAAGGAGGU(U) 4 AUG CUC (U) 9 -3 9  (cognate for tRNA 2Leu ).mRNA-II5 9 -GGCAAGGAGGU(U) 4 AUG UUU  (U) 9 -3 9 (near-cognatefortRNA 2Leu ).mRNA-III 5 9 -GGCAAGGAGGU(A) 4 AUG UAC (A) 9 -3 9  (cognate for tRNA Tyr ).mRNA-IV5 9 -GGCAAGGAGGU(A) 4 AUG UGC (A) 9 -3 9 (near-cognatefortRNA Tyr ). Data collection, processing and structure determination.  All crystals belong tospace group  P  2 1 2 1 2 1  and contain two molecules per asymmetric unit. Data on allsix complexes were collected at 100K at the Synchrotron Light Source,Switzerland, using the Pilatus 6M detector. A very low dose mode was used andhuge redundancy was collected 29 . The structure 21 , with tRNAs, mRNA and metalions removed, was used for refinement with Phenix  30 . The initial model wascorrectly placed within each data setby rigid body refinement with each moleculeasarigidbody.Thiswasfollowedbyrigidbodyrefinementofindividualsubunits.AfterpositionalandB-factorrefinement,theresultingelectrondensitymapswereinspectedandthetRNAsandmRNAligandswerebuiltintheseunbiasedmaps.Inalloftheseandthefollowingrefinementrounds, nobase-pairrestraintswere usedbetween tRNAs and mRNAs to avoid bias towards perfect base-pair geometry.After several cycles of manual rebuilding followed by positional and individualisotropicB-factor refinement, magnesium ions were added and a final refinementroundtookplace.Asummaryofthecrystallographicdataandrefinementstatisticsis given in Supplementary Tables 1 and 2. Supplementary Table 3 shows theaverage B-factorforthe entirestructureas well asaverage B-factorsfor thedecod-ing centre and the codon–anticodon helix. From this table it is seen that theaverage B-factors for the substructure comprising the decoding centre (G530,A1492, A1493 from 16S and G1913 from 23S), as well as the codon–anticodonhelices (nucleotides 34–36 of tRNA in the A site and the corresponding codon of mRNA), are less than the overall B-factor for the entire ribosome structure.Therefore it is clear that the decoding centre is part of the most accurately deter-mined parts of these models.To verify that the base-pair geometry described in the paper is correct, weperformedamanyextraindependentrefinementroundswithbase-pairgeometriesrestrained to various standards (Watson–Crick, wobble, etc.) so that we could beconfidentaboutthereportedgeometries.SupplementaryFigs3and4showOMIT-averaged kick maps 30,31 of the G N U mismatch for the two near-cognate complexes.These unbiased maps show that a wobble conformation of these base pairs wouldnot fit into the electron density and clearly demonstrate that a Watson–Crick conformation is the only plausible fit. 31. Praaenikar, J., Afonine,P.V., Guncar, G., Adams, P.D. & Turk,D. Averaged kickmaps:lessnoise, moresignalandprobablylessbias.  Acta Crystallogr. D  65, 921–931 (2009). RESEARCH LETTER Macmillan Publishers Limited. All rights reserved  ©2012
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks