Dispatch R369

Ribosomal translocation: EF-G turns the crankRachel Green

Recent results from cryo-electron microscopy haveshown that substantial structural rearrangements in bothelongation factor EF-G and the ribosome occur duringtRNA translocation. The observed sites of interactionbetween EF-G and the ribosome are consistent withmolecular mimicry models for EF-G function.

Address: Department of Molecular Biology and Genetics, Johns HopkinsUniversity School of Medicine, Baltimore, Maryland 21205, USA.

Current Biology 2000, 10:R369–R373

0960-9822/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.

In an intricate and precise molecular dance, the ribosometranslates the genetic information in the cell into theencoded polypeptides. Peptide-bond formation takesplace on the large subunit of the ribosome (50S inbacteria), and the decoding of the genetic informationthrough codon–anticodon interactions takes place on thesmall subunit (30S). The integration of these eventsoccurs at the ribosomal subunit interface, where the tRNAsubstrates bind and act. The ‘hybrid states’ model oftranslation describes the independent movement of thetRNA substrates with respect to the two subunits of theribosome [1]. Movement of the acceptor end of the tRNAs— where the amino acid or growing peptide is attached —with respect to the large subunit is coordinated withpeptide-bond formation, and movement of the anticodonend of the tRNAs and the associated mRNA with respectto the small subunit is promoted by the elongation factorEF-G in a step known as translocation (Figure 1). Both

steps are coordinated with the cleavage of a high energybond: the aminoacyl ester bond of a charged tRNA andthe phosphoanhydride linkage of GTP, respectively.

How the ribosome ratchets its massive tRNA substratesthrough the interface cavity is a fascinating problem.Translocation in the ribosome occurs spontaneously atsome very low rate in the absence of exogenous EF-G andGTP hydrolysis [2]. EF-G acts as a catalyst to increase thebasal rate of translocation. The mechanism of transloca-tion is thus dictated by ribosomal structure and motion.EF-G (and EF-G:GDP) itself is remarkably similar instructure to the EF-Tu:GTP:tRNA ternary complexwhich is responsible for loading tRNA substrates into theA site of the ribosome (Figure 2) [3]. Domain IV of EF-Gis a mimic of the anticodon end of a tRNA in the ternarycomplex, suggesting that these factors bind and act at ahomologous site.

Startling progress recently has led to the solution of X-raystructures of the ribosome that approach atomic resolu-tion, yielding unprecedented views of its molecular com-ponents [4–7]. Although even more detailed structures areanticipated, a dynamic view of how these componentsmove and interact is essential for understanding transla-tion. Cryo-electron microscopy readily explores the inter-mediate functional steps in translation. In this approach,different ribosomal complexes are formed in solution,frozen and fixed on a grid for microscopy. Computerreconstruction and a comparison of complexes identifiesdifferences in electron density that can be attributed

Figure 1

E/E P/P

Peptidyltransfer

EF-GGTP

P/P A/A P/E A/P

E P A

50S

30S

State: P/P

EF-TuGTP

tRNA-aa

Current Biology

Minimal hybrid states model for the translational elongation cycle [1].The 50S and 30S subunits are divided into three tRNA binding sites:the A (aminoacyl), P (peptidyl) and E (exit) sites, shown as rectangles[4]. tRNAs are represented as vertical bars and amino acids by smallcolored circles. mRNA is represented as an orange line bound to the

30S subunit. The cartoon depicts the vectorial movement of theacylated and deacylated tRNAs through the ribosome, catalyzed byfactors EF-Tu and EF-G during translation. Intermediate states (notshown) have been proposed for the factor-dependent steps oftranslation — decoding and translocation.

bb10j10.qxd 10/5/00 1:54 pm Page R369

either to added components — for example, to tRNA sub-strates or various translational factors — or to structuraldifferences induced in the ribosome by such components.Distinguishing between these possibilities is not alwaysstraightforward, but in the ribosome field, where much isalready known, such analysis becomes tenable.

The structure of the ribosome:EF-G complex during theprocess of translocation has been visualized for the firsttime in several recent cryo-electron microscopy studies[8–10]. It is clear from these studies that both theribosome and EF-G undergo substantial rearrangementsduring the process of tRNA translocation, and that thereexist multiple post-translocational ribosome conformers.Translation is dynamic and trapping its intermediates is asophisticated game. In these studies, two different translo-cation inhibitors, thiostrepton and fusidic acid, were usedto trap closely related intermediate states in the trans-lational cycle for analysis by cryo-electron microscopy.

Fusidic acid is an antibiotic that blocks release of theEF-G:GDP complex from the ribosome, but does notaffect GTP hydrolysis or translocation. Thiostrepton is anantibiotic that is thought to trap an intermediate state intranslocation preceding that frozen by fusidic acid. Itbinds directly to 23S rRNA and does not interfere with

EF-G binding or with GTP hydrolysis, but prevents thelater steps of tRNA translocation and EF-G:GDP release[11]. The distinctions between the different complexesare important, because the order of the events of transloca-tion — GTP hydrolysis, tRNA movement, phosphate andEF-G:GDP release — is still actively debated [12]. Trap-ping the intermediates of translation using antibioticinhibitors of the process depends on a precise understand-ing of the role played by these antibiotics in translation.And, even in these cases, the data should be interpretedcautiously as these agents may disrupt the naturalsequence of events.

In several initial studies, fusidic acid was used to stabilizethe presumed final state in translocation preceding therelease of the EF-G:GDP complex from the ribosome — a‘post-translocation’ ribosome, referred to below as the‘post(fus)’ state [8,9]. The binding orientation of EF-G onthe ribosome in this complex was strikingly consistentwith that derived by directed hydroxyl radical probingmethods from a similar complex [13]. Domain IV, themimic of the tRNA anticodon, is buried deep in the 30Ssubunit, apparently in contact with the decoding region of16S rRNA and an arc-like connection is observed betweenthe base of the ‘L7/L12 stalk’ — where L7 and L12 areproteins of the large, 50S subunit — and the G′ domain ofEF-G. Structural contacts between EF-G and the ribo-some in the stalk or ‘factor binding region’ are consistentwith biochemical data linking 23S rRNA and protein ele-ments in this region to translocation.

Wintermeyer and colleagues [10] trapped novel inter-mediate translocational ribosomal complexes by incubat-ing a pre-translocation hybrid ribosomal complex —carrying two tRNAs — and EF-G:GTP in the presence ofthiostrepton. Rapid freezing of the reaction stalls apopulation of ‘pre-translocational’ (pre(thio)) ribosomalcomplexes, and a longer incubation stalls a population of‘post-translocational’ (post(thio)) complexes. Previouskinetic analysis with thiostrepton indicated that the ‘pre-translocational’ complex described here containsEF-G:GDP [11]. Wintermeyer and colleagues [10] arguethat this pre-translocational complex represents a riboso-mal state that directly precedes the transition state fortRNA translocation, but is distinct from an early pre-translocation state where GTP has not yet beenhydrolyzed. A ribosomal complex with EF-G:GDP andfusidic acid was also generated in this study for compari-son with earlier work.

The reconstructions of the ribosome revealed substantialdifferences between these ‘pre-translocational’ and ‘post-translocational’ complexes (Figure 3). In the pre(thio)ribosomes, EF-G density clearly bridges the cleft betweenthe large and small subunits of the ribosome. Domain I ofEF-G, the GTPase domain, interacts extensively with the

R370 Current Biology Vol 10 No 10

Figure 2

Molecular mimicry. Crystal structures of EF-Tu:GTP:Phe-tRNA ternarycomplex and EF-G:GDP [3,14]. In both structures, domain I (theGTPase domain) is shown in magenta, domain II in blue, and domain III(incomplete in the EF-G structure) in green; domain IV of EF-G and thetRNA of the ternary complex are shown in yellow; and domain V ofEF-G is shown in red. Bound nucleotide is shown in orange.

Current Biology

EF-Tu•GTP•Phe-tRNA EF-G•GDP

bb10j10.qxd 10/5/00 1:54 pm Page R370

base of the L7/L12 stalk of the large subunit, whiledomain IV, the anticodon mimic, touches the shoulder ofthe small subunit (where protein S4 has been localized).In the post(thio) complex, EF-G itself has undergone sig-nificant conformational changes and its interactions withthe ribosome are completely changed. Interactions withthe L7/L12 stalk are now minimal, and the body of EF-Gis submerged in the intersubunit cavity. Domain IV isengaged with the decoding center of the small subunit,and domains I and II form multiple contacts with the headand body of the 30S subunit.

Interestingly, this thiostrepton ‘post’ complex differssignificantly from the previously observed fusidic acid‘post-translocational’ complex. In the latter complex,domain IV of EF-G also interacts with the decodingregion of the small subunit (as mimicry dictates), butmuch more of the factor’s density protrudes from theintersubunit cavity. Further, in the post(fus) complex

there are now substantial contacts between domains I andV of EF-G with electron density that is attributed to thelarge subunit (in part with the sarcin ricin loop of 23SrRNA). Indeed, the fusidic acid complex appears to betrapped with EF-G exiting the subunit interface prior todissociation from the ribosome.

The conformation of EF-G itself appears to be distinct ineach complex (fitting of the EF-G crystal structure to theobserved electron density in each case was achieved byrotating domains I and II of EF-G with respect to domainsIII, IV and V). In no case does the conformation ofribosome-bound EF-G match that of the known EF-Gcrystal structure [14,15], revealing the limitations of ourcurrent structural knowledge. Indeed, there is still no highresolution structure of EF-G:GTP. The structures of ‘pre-translocational’ and ‘post-translocational’ ribosomes indi-cate that the ribosome structure changes substantiallyduring translocation. The ribosome is not an inert surface

Dispatch R371

Figure 3

Current Biology

Control Pre(thio) Post(thio) Post(fus)Shoulder

Pre(thio) Post(thio) Post(fus)

Head

Body

L7/12

30S 50S

Three-dimensional reconstructions of ribosomes complexed with EF-G,and fitting of EF-G to density difference. In the upper panel, ribosomesare depicted with the 30S subunit to the left and the 50S subunit tothe right. Density attributed to EF-G based on difference density isshown in blue. Control, pre-translocation complex in the absence ofEF-G; pre(thio), pre-translocation complex with EF-G:GTP stalled earlywith thiostrepton; post(thio), post-translocation complex withEF-G:GDP stalled late with thiostrepton; post(fus), post-translocationcomplex trapped with fusidic acid. Several morphological features are

indicated: L7/L12 stalk on the large subunit, and the shoulder, headand body of the small subunit. In the lower panel, EF-G (color schemeas in Figure 2) is fit to the observed density differences, allowingmovement of domains I/II with respect to domains III/IV/V. Theorientation of EF-G with respect to the ribosome is substantiallydifferent in each case. Here the EF-G structures are aligned so thatdomains I and II can be superimposed. Ribosome contacts areindicated by open (30S) or closed (50S) arrowheads. Ribosomedensity is marked with an asterisk. (Compiled from [10].)

bb10j10.qxd 10/5/00 1:54 pm Page R371

traversed by the tRNA substrates and translation factors.Structural changes in the large subunit are subtle and arelimited to the L7/L12 stalk or ‘factor binding’ region. Inthe small subunit, however, structural differences betweenthe pre(thio) and post(thio) states are observed in severaljoint regions, including the neck (a thin element thoughtto be formed by a single RNA helix), the outer junction ofthe head and body, and the head at its connection with thebeak. The relative arrangements of the two subunits issimilar in the two states, though the A site side (near theL7/L12 stalk) opens up by 5–10° in the ‘post’ complex.

Previous cryo-electron microscopy studies of pre-trans-locational and post-translocational ribosomes (where EF-Gis not bound) did not reveal such differences in the struc-ture of the two subunits [16]. The suggestion is that the‘pre-translocational’ state observed here is a first glance atan intermediate state in translocation — with intermediatestructural properties. It will be interesting to determinewhether the structural transitions observed here are analo-gous to those recently observed in ribosomes that arelocked into phenotypically distinct conformers — the lowtranslational fidelity ‘ram’ and high fidelity ‘restrictive’conformers — associated with the ‘912 switch region’ of16S rRNA [17].

The molecular components of the ribosome involved intranslocation are not well characterized (reviewed in [18]).Several elements of the large subunit of the ribosome areknown to be critical for translocation, including the sarcin-ricin loop and the 1070 region of the 23S rRNA, and theproteins located in the stalk region (L7, L12 and L11).There are also data linking elements in the small subunitto translocation. Logically, the hybrid (A/P) tRNA and itsinteractions with the small subunit A site decoding regionare at the heart of translocation — the interaction of thetRNA with its codon in the mRNA is centrally responsiblefor the maintenance of the reading frame during transla-tion. An important clue might also be found in earlyexperiments demonstrating that ribosomes lacking thesmall subunit protein S12 (or having a chemically modi-fied version of S12) exhibit enhanced levels of sponta-neous translocation [2]. Movements of the recentlyidentified ‘helical switch’ region of 16S rRNA, geneticallyinteractive with both the S5 and S12 proteins of the smallsubunit, may also be critical [19].

EF-G is a member of the well-studied family of GTP-binding proteins which use the energy of GTP hydrolysisto induce conformational changes that modulate bindingaffinities for effector molecules, in this case the ribosome(reviewed in [20]). In classical models for translocation,EF-G:GTP first binds the ribosome, translocation occursand subsequent GTP hydrolysis — stimulated by theribosomal effector — results in conformational changesleading to release of EF-G:GDP from the ribosome.

However, recent rapid kinetic data indicate that GTPhydrolysis precedes translocation, suggesting that theenergy of GTP hydrolysis is coupled to this event [21].What such coupling might mean in molecular terms is notyet fully resolved.

At its simplest level, translocation should be consideredfrom a chemical perspective. Brownian motion is responsi-ble for the sampling by the ribosome of different confor-mational states. The pathways to various states are more orless energetically accessible, and so such states are differ-entially occupied. EF-G alters the equilibria of differentribosomal states by binding and stabilizing particularconformations. While the details of EF-G catalyzedtranslocation are actively debated, the following model isbroadly consistent with what is currently known. EF-Gbinds to the ribosome and GTP hydrolysis is stimulated,establishing a tight complex between the ribosome andEF-G. In this complex, conformational changes in EF-Gare coupled to changes in ribosomal conformational equi-libria, leading first to translocation and then to EF-Grelease. Somehow, the binding of EF-G to the ribosomeallows the energy of GTP hydrolysis to be captured andused in multiple steps. The cryo-electron microscopyimages discussed here provide us with snapshots of threeof these conformational steps [8–10]. As more snapshots ofthe translational cycle are taken, and as the kinetics andthermodynamics of these steps are described, we mayfinally understand how the energy of hydrolysis is used bythe ribosome to promote the ratcheting directional move-ments central to translation.

AcknowledgementsThanks to J. Lorsch for stimulating discussions, to B. Cormack and K. Wilsonfor helpful comments on the manuscript, and to M. Rodnina and W. Winter-meyer for help with the figures.

References1. Moazed D, Noller HF: Intermediate states in the movement of

transfer RNA in the ribosome. Nature 1989, 342:142-148.2. Spirin AS: Ribosomal translocation: facts and models. Prog

Nucleic Acid Res Mol Biol 1985, 32:75-114.3. Nissen P, Kjeldgaard M, Thirup S, Polekhina G, Reshetnikova L,

Clark BF, Nyborg J: Crystal structure of the ternary complex ofPhe-tRNAPhe, EF-Tu, and a GTP analog. Science 1995,270:1464-1472.

4. Cate JH, Yusupov MM, Yusupova GZ, Earnest TN, Noller HF: X-raycrystal structures of 70S ribosome functional complexes. Science1999, 285:2095-2104.

5. Clemons WM Jr, May JL, Wimberly BT, McCutcheon JP, Capel MS,Ramakrishnan V: Structure of a bacterial 30S ribosomal subunit at5.5 Å resolution. Nature 1999, 400:833-840.

6. Ban N, Nissen P, Hansen J, Capel M, Moore PB, Steitz TA:Placement of protein and RNA structures into a 5 Å-resolutionmap of the 50S ribosomal subunit. Nature 1999, 400:841-847.

7. Tocilj A, Schlunzen F, Janell D, Gluhmann M, Hansen HA, Harms J,Bashan A, Bartels H, Agmon I, Franceschi F, Yonath A: The smallribosomal subunit from Thermus thermophilus at 4.5 Å resolution:pattern fittings and the identification of a functional site. Proc NatlAcad Sci USA 1999, 96:14252-14257.

8. Agrawal RK, Penczek P, Grassucci RA, Frank J: Visualization ofelongation factor G on the Escherichia coli 70S ribosome: themechanism of translocation. Proc Natl Acad Sci USA 1998,95:6134-6138.

R372 Current Biology Vol 10 No 10

bb10j10.qxd 10/5/00 1:54 pm Page R372

9. Agrawal RK, Heagle AB, Penczek P, Grassucci RA, Frank J: EF-G-dependent GTP hydrolysis induces translocation accompanied bylarge conformational changes in the 70S ribosome. Nat Struct Biol1999, 6:643-647.

10. Stark H, Rodnina MV, Wieden HJ, van Heel M, Wintermeyer W:Large-scale movement of elongation factor G and extensiveconformational change of the ribosome during translocation. Cell2000, 100:301-309.

11. Rodnina MV, Savelsbergh A, Matassova NB, Katunin VI, SemenkovYP, Wintermeyer W: Thiostrepton inhibits the turnover but not theGTPase of elongation factor G on the ribosome. Proc Natl AcadSci USA 1999, 96:9586-9590.

12. Rodnina MV, Savelsbergh A, Wintermeyer W: Dynamics oftranslation on the ribosome: molecular mechanics oftranslocation. FEMS Microbiol Rev 1999, 23: 317-333.

13. Wilson KS, Noller HF: Mapping the position of translationalelongation factor EF-G in the ribosome by directed hydroxylradical probing. Cell 1998, 92:131-139.

14. Czworkowski J, Wang J, Steitz TA, Moore PB: The crystal structureof elongation factor G complexed with GDP, at 2.7 Å resolution.EMBO J 1994, 13:3661-3668.

15. Evarsson A, Brazhnikov E, Garber M, Zheltonosova J, Chirgadze Y,al-Karadaghi S, Svensson LA, Liljas A: Three-dimensional structureof the ribosomal translocase: elongation factor G fromThermus thermophilus. EMBO J 1994, 13:3669-3677.

16. Stark H, Orlova EV, Rinke-Appel J, Junke N, Mueller F, Rodnina M,Wintermeyer W, Brimacombe R, van Heel M: Arrangement of tRNAsin pre- and posttranslocational ribosomes revealed by electroncryomicroscopy. Cell 1997, 88:19-28.

17. Gabashvili IS, Agrawal RK, Grassucci R, Squires CL, Dahlberg AE,Frank J: Major rearrangements in the 70S ribosomal 3D structurecaused by a conformational switch in 16S ribosomal RNA. EMBOJ 1999, 18:6501-6507.

18. Wilson KS, Noller HF: Molecular movement inside the translationalengine. Cell 1998, 92:337-349.

19. Lodmell JS, Dahlberg AE: A conformational switch in Escherichiacoli 16S ribosomal RNA during decoding of messenger RNA.Science 1997, 277:1262-1267.

20. Czworkowski J, Moore PB: The elongation phase of proteinsynthesis. Prog Nucleic Acid Res Mol Biol 1996, 54:293-332.

21. Rodnina MV, Savelsbergh A, Katunin VI, Wintermeyer W: Hydrolysisof GTP by elongation factor G drives tRNA movement on theribosome. Nature 1997, 385:37-41.

Dispatch R373

bb10j10.qxd 10/5/00 1:54 pm Page R373