Translocation of tRNA and mRNA through the ribosome is one of the most dynamic events during protein synthesis. In the cell, translocation is catalysed by EF-G (elongation factor G) and driven by GTP hydrolysis. Major unresolved questions are: how the movement is induced and what the moving parts of the ribosome are. Recent progress in time-resolved cryoelectron microscopy revealed trajectories of tRNA movement through the ribosome. Driven by thermal fluctuations, the ribosome spontaneously samples a large number of conformational states. The spontaneous movement of tRNAs through the ribosome is loosely coupled to the motions within the ribosome. EF-G stabilizes conformational states prone to translocation and promotes a conformational rearrangement of the ribosome (unlocking) that accelerates the rate-limiting step of translocation: the movement of the tRNA anticodons on the small ribosomal subunit. EF-G acts as a Brownian ratchet providing directional bias for movement at the cost of GTP hydrolysis.
- anticodon stem–loop
- bacterial ribosome
- Brownian ratchet
- GTP hydrolysis
- peptidyltransferase centre
- protein synthesis
Ribosomes are supramolecular factories that synthesize proteins in all cells. Bacterial ribosomes consist of two subunits, the small subunit, 30S, and the large subunit, 50S. Each ribosomal subunit is composed of rRNA and a number of proteins. The functional centres of the ribosome, comprising the decoding site on the 30S subunit, the peptidyltransferase centre on the 50S subunit and the binding sites for translation factors on both subunits, are built mostly of rRNA. As a machine that has to select its substrate accurately according to the sequence of mRNA codons and produce the proteins of amino acids delivered by aminoacyl-tRNAs, the ribosome has evolved intricate mechanisms for RNA recognition and catalysis. Furthermore, protein elongation is a cyclic process, which requires that the tRNA molecules move through the ribosome from the A site, where decoding takes place, to the P site, where the peptidyl-tRNA is placed before transferring the nascent peptide on to the next aminoacyl-tRNA, and further to the E (exit) site before leaving the ribosome. The movement of tRNAs takes place during the translocation step of protein synthesis, a process which is driven by large-scale stochastic conformational fluctuations of the ribosome.
During translocation, two tRNAs move by large distances from one site of the ribosome to the next adjacent site, with the coupled movement of mRNA by one codon (Figure 1). The reaction is catalysed by EF-G (elongation factor G), which hydrolyses GTP in the process . EF-G consists of five domains: domain 1 that binds GTP/GDP, domain 2 which is common to other translational GTPases, and domains 3–5 which are specific to EF-G and form a unit that moves relative to domains 1 and 2 during translocation. Fundamental questions in understanding translocation are (i) how do the tRNAs move through the ribosome, (ii) how does EF-G accelerate the movement, and (iii) how are GTP hydrolysis and Pi release coupled to movement. The aim of this review is to summarize the recent insights into the molecular mechanism of the movement. Detailed accounts of different aspects of translocation can also be found in recent reviews [2–4].
Spontaneous movement of tRNAs
Translocation can proceed spontaneously in both forward and backward directions, implying that its basic features are inherent to the ribosome–tRNA complex itself [5–8]. Retro-translocation is driven by the gain in affinity when a cognate E-site tRNA moves into the P site, which compensates for the affinity loss accompanying the movement of peptidyl-tRNA from the P to the A site. The preferential direction of movement depends on the nature of tRNAs and their affinities for the respective ribosomal sites. The existence of spontaneous forward and backward translocation strongly suggests that the tRNAs can move through the ribosome directed solely by thermal fluctuations and the thermodynamic gradient; however, compared with the reaction promoted by EF-G, the spontaneous movement is very slow.
In the transition from one canonical ribosome-binding site to the other, i.e. from the A to the P site and from the P to the E site, the tRNAs acquire intermediate configurations, so-called hybrid states , and the ribosome undergoes dynamic structural changes [10–13]. In the hybrid states, the anticodon stem–loops of the tRNAs reside in the A and P sites of the 30S subunit, while the respective acceptor ends are oriented towards the P and E sites of the large 50S subunit. In some cases, only deacylated P-site tRNA adopts a hybrid configuration, whereas the A-site tRNA appears to remain in the classic state . At the same time, the subunits undergo a rotational (ratchet-like) movement relative to each other and the formation of hybrid states correlates with the rotated orientation of subunits [10,13]. Single-molecule FRET (fluorescence resonance energy transfer) studies indicated that the ribosome spontaneously fluctuates between classic/non-rotated and hybrid/rotated states [11,15]. Furthermore, ribosome parts move upon transitions between the two states (reviewed in detail in ), implying a high degree of flexibility and spontaneous motions.
Recently, time-resolved cryo-EM (cryoelectron microscopic) analysis of spontaneous tRNA retro-translocation  revealed trajectories of the tRNAs and accompanying conformational fluctuations of the ribosome. From the 50 cryo-EM reconstructions of tRNAs at different stages of retro-translocation, eight most distinct states of tRNA movement were classified (Figure 2). Each of these states consists of ensembles of up to 11 substates that share similar tRNA positions and tRNA–ribosome contacts, but vary in ribosome conformation. tRNA movement entails the sequential step-by-step rupture and formation of ribosome–tRNA contacts. Several ribosomal elements, such as helices 38 and 69 of the 50S subunit, ribosomal protein L5 and the L1 stalk move over long distances (between 8 and 40 Å; 1 Å=0.1 nm) together with the tRNAs, thereby confining the space accessible for tRNA movement and reducing the need to form or break too many contacts at a time. Notably, the movements of the ribosomal subunits are larger and more complex than anticipated from previous work. In addition to the rotation movement of the 30S subunit body relative to the 50S subunit, a movement of the 30S subunit head relative to the body was observed, reminiscent of the ‘swivel’ movement found in ribosome crystals . For each of the eight states of tRNA movement, there is a characteristic spectrum of nearly continuous 30S body rotations and 30S head movements . On the other hand, it was also possible to observe a particular tRNA state over a wide range of different 30S conformations and, vice versa, to observe ribosomes in similar overall conformations that differ significantly in their tRNA positions. Although 30S body and head movements as well as tRNA movement and global changes in ribosome conformation are correlated, coupling between them is loose.
Quantitative analysis of particle distribution between substrates suggested that the five states within the ensemble of pre-translocation states are in rapid equilibrium; the same is true for the three post states  (Figure 2). From the distribution frequencies, values of the equilibrium constants (Keq) could be calculated. The rate-limiting step along the translocation reaction is the movement of the tRNA anticodon stem–loops on the 30S subunit, which is very slow . Most likely, anticodon movement is also rate-limiting during spontaneous translocation in the forward direction, as the reported rates of the overall reaction  are several orders of magnitude lower than the rates of spontaneous hybrid-state formation [11,14,15]. It is mainly this step on which EF-G acts to accelerate translocation, although the movement of the 3′ ends of tRNAs on the 50S subunit that are necessary to attain the final post-translocation position are also facilitated by EF-G .
Kinetic mechanism of EF-G-dependent translocation
Kinetic studies provide a quantitative description of the elemental steps of translocation [19–22]. EF-G is recruited to the ribosome through the interactions with ribosomal protein L7/12 , of which the Escherichia coli ribosome contains four copies. EF-G binding stabilizes the ratchet/hybrid conformation of the ribosome, whereas the tRNA anticodons do not move at this stage  (Figure 3). L7/12 together with the sarcin–ricin loop of 23S RNA, and possibly other yet unidentified elements of the ribosome, activates GTP hydrolysis by EF-G [1,23,25,26]. GTP hydrolysis drives a rearrangement of the ribosome (referred to as unlocking) that precedes and limits the rate of tRNA–mRNA movement on the 30S subunit . The structural basis for unlocking is not known, as a structural model of this transient ribosome–EF-G complex is not available, but it is likely to involve a change in the mobility of elements of the 30S subunit, possibly by facilitating the swivelling movement of the 30S head [27,28]. GTP hydrolysis by EF-G is coupled to the acceleration of unlocking [20,29], but the mechanism of coupling between GTP hydrolysis and the conformational changes in the complex is unknown. The movement of tRNA–mRNA that takes place in the unlocked state of the ribosome appears to be rapid and spontaneous [20,21]. As in the spontaneous reaction, during EF-G-catalysed translocation the tRNAs are likely to move through a quasi-continuous landscape of intermediate states, and some of these intermediate states have been characterized [22,30]. The release of Pi from EF-G, which was shown to be independent of and concomitant to the tRNA movement on the 30S subunit, induces another conformational change in EF-G that is required for the ribosome to return to the locked state and for EF-G to dissociate from the ribosome .
tRNA–mRNA movement and Pi release are intrinsically rapid and take place at random, as either step can be inhibited without affecting the other. For instance, tRNA–mRNA movement, but not Pi release, is blocked by antibiotics that bind in the 30S decoding region, such as viomycin, paromomycin or hygromycin B  or by engineering a disulfide bridge in EF-G that restricts domain movement . A point mutation at the tip of domain 4 of EF-G, H583K, has the same effect . Conversely, several point mutations in the C-terminal domain of protein L7/12 inhibit Pi release with no effect on movement . In summary, these results indicate that tRNA–mRNA movement takes place after the unlocking rearrangement.
One further observation is of particular interest. The unlocking of the ribosome precedes, and is required for, both tRNA movement and Pi release. However, after the unlocking step, additional rearrangements have to occur that affect the tRNA movement on the 30S subunit and the Pi release in different ways. Notably, antibiotics that bind to the decoding region selectively inhibit tRNA movement and have no effect on Pi release, suggesting that conformational changes at the decoding site are required for movement. Interestingly, a specific cleavage of 16S rRNA at the decoding centre by colicin E3 leads to an acceleration of tRNA movement in translocation . This finding can be rationalized in terms of facilitated fluctuations of the 30S subunit head which allow for fast movement of the tRNA anticodon stem–loops on the 30S subunit. The effect of the H583K mutation indicates that the contact of domain 4 of EF-G with the ribosome is involved in opening up the decoding region [26,32–34]. In the presence of EF-G the directionality of tRNA movement is biased by EF-G which accompanies the displacement of tRNA in such a way that in the post-translocation state domain 4 of EF-G, which is crucial for translocation [1,33], occupies the 30S A site [12,32,35], thus effectively preventing back movement of peptidyl-tRNA while the unlocked state prevails. On EF-G, an insertion in the GTP-binding domain, the G′ subdomain, appears to be important for EF-G binding to the ribosome and the conformational coupling between GTP hydrolysis, retention of Pi, and unlocking [36–38].
EF-G as a Brownian ratchet
Given that the movement of tRNAs through the ribosome may occur spontaneously due to thermal fluctuations, what is the role of EF-G in accelerating translocation? Is EF-G an active motor that couples the energy of GTP hydrolysis to the movement through direct mechano-chemical coupling, or does it act as a Brownian ratchet that biases thermal fluctuations towards forward movement by structural anisotropy produced by GTP cleavage? EF-G accelerates translocation in three distinct ways. First, EF-G contributes to the stabilization of the hybrid/rotated state of the ribosome [11,18] and may, thereby, bias and accelerate the partial movement of the tRNAs from the classical/non-rotated state towards the hybrid/rotated state. The GTP-bound form of EF-G, but not GTP hydrolysis, is important for the reaction. Most likely, the energy of EF-G·GTP binding to the ribosome is utilized, as at this early stage, it is likely that GTP bound to EF-G has not yet been hydrolysed. Secondly, EF-G accelerates the unlocking step and probably also the following conformational changes at the decoding region required for rapid tRNA movement. When GTP hydrolysis is prevented by using non-hydrolysable analogues, the rate of unlocking is decreased by a factor of 50 [1,29], suggesting that at least part of the free energy of GTP hydrolysis is coupled to conformational changes of the ribosome. At this step, EF-G is present in the GDP·Pi-bound form, and the release of Pi is not required to drive unlocking; on the contrary, the retention of Pi may be necessary to prevent the premature collapse of EF-G into an unproductive conformation. Thirdly, EF-G is likely to bias diffusion to produce forward movement. This involves a conformational change of EF-G, brought about by GTP hydrolysis and partially stored in the system due to delayed Pi release. This suggests that EF-G functions as a Brownian ratchet that biases tRNA movement in the forward direction by both inducing conformational rearrangements and preventing backward movement [1,39]. The latter effect may be particularly important in cases where spontaneous forward movement is impaired, e.g. with tRNAs that thermodynamically favour the pre-translocation state or with mRNAs that have considerable secondary structure. Recent force measurement on ribosomes translating mRNA with a strong secondary structure suggested that translation occurs through successive translocation-and-pause cycles . Pause lengths depended on the secondary structure of the mRNA; the applied force destabilized the secondary structure and decreased the duration of pauses, but did not affect translocation times. It is generally assumed that the ability to unwind secondary structure of mRNA is inherent in the ribosome, and it has been suggested that ribosomal proteins S3, S4 and S5 that reside at the mRNA entrance site are active in unwinding the mRNA . An involvement of cellular helicases in mRNA unwinding is considered unlikely because the ribosome is able to move through mRNA hairpins in the presence of only two factors required for translation, EF-Tu and EF-G and GTP, and an additional energy source, such as ATP, is not needed. It is not known, however, whether the energy of GTP hydrolysis by EF-G is used to drive the unwinding of the mRNA, thereby coupling translocation and mRNA unwinding.
The work in our laboratories was funded by grants from the Deutsche Forschungsgemeinschaft.
We thank Holger Stark and Niels Fischer for their fruitful collaboration on the cryo-EM analysis of translocation. We also thank the past and present members of our group who have contributed to this work, Andreas Savelsbergh, Frank Peske, Andrey Konevega, Dagmar Mohr and Berthold Wilden, and our collaborators Yuri Semenkov and Vladimir Katunin.
NACON VIII: 8th International Meeting on Recognition Studies in Nucleic Acids: An Independent Meeting held at The Edge, University of Sheffield, Sheffield, U.K., 12–16 September 2010. Organized by Mike Blackburn, Mark Dickman, Jane Grasby, David Hornby, Chris Hunter, John Rafferty, Jim Thomas, David Williams and Nick Williams (Sheffield, U.K.).
Abbreviations: E, exit; EF-G, elongation factor G
- © The Authors Journal compilation © 2011 Biochemical Society