Translocation as continuous movement through the ribosome

ABSTRACT In each round of translation elongation, tRNAs and mRNA move within the ribosome by one codon at a time. tRNA–mRNA translocation is promoted by elongation factor G (EF-G) at the cost of GTP hydrolysis. The key questions for understanding translocation are how and when the tRNAs move and how EF-G coordinates motions of the ribosomal subunits with tRNA movement. Here we present 2 recent papers which describe the choreography of movements over the whole trajectory of translocation. We present the view that EF-G accelerates translocation by promoting the steps that lead to GTPase-dependent ribosome unlocking. EF-G facilitates the formation of the rotated state of the ribosome and uncouples the backward motions of the ribosomal subunits, forming an open conformation in which the tRNAs can rapidly move. Ribosome dynamics are important not only in translocation, but also in recoding events, such as frameshifting and bypassing, and mediate sensitivity to antibiotics.

During the elongation phase of translation the ribosome moves along the mRNA while synthesising the nascent polypeptide. The movement takes place during the translocation step of elongation and entails the displacement of 2 tRNAs bound to their respective codons through the ribosome by one codon in each round of elongation. It is now more than 2 decades ago since researchers obtained the first evidence suggesting that the ribosome not only provides the environment but also actively participates in translocation, with the dynamics of the ribosomal subunits (the small and large subunits, SSU and LSU, respectively), and of the tRNAs playing a key role in promoting the movement. 1,2 Structural studies provided high-resolution snapshots of ribosome motions while translocating the tRNAs through the A, P and E sites. They identified the types of motions that play a key role in translocation: the rotation of the SSU relative to the LSU and the swiveling motion of the SSU subdomain, the SSU head, relative to the SSU body. [3][4][5][6][7][8][9][10][11] In parallel, ensemble and single-molecule kinetics and molecular dynamics simulations provided insights into the link between the movement of the tRNA-mRNA complex and the motions of the subunits. [12][13][14][15][16] The current working model for translocation suggests that each time a peptide bond is formed, the CCA ends of the A-site and P-site tRNA move toward the P and E site on the LSU, respectively, into the A/P and P/E hybrid state. [17][18][19][20][21] This step can proceed spontaneously, driven purely by thermal energy, and is reversible. At the same time, the SSU rotates in counterclockwise (CCW) direction relative to the LSU, and the ribosome alternates between the rotated and the non-rotated state which is loosely coupled to the formation of the hybrid and the classical tRNA state, respectively (Fig. 1A). In the next step, the anticodons of the tRNAs together with the mRNA move relative to the SSU, and the CCA ends complete their movement on the LSU, reaching the classical P/P and E/E state and completing translocation. The SSU moves clockwise (CW) into their initial state relative to the LSU. Movement of the tRNAs on the SSU comprises the main kinetic barrier to translocation, 15,22 which is overcome by elongation factor G (EF-G), a GTPase, at the cost of GTP hydrolysis. 23 However, many fundamental questions remain unanswered, such as: Does EF-G bind to the ribosome in the non-rotated-classical or rotated-hybrid state? How are the movements of the ribosome and the tRNA-mRNA complex coordinated? How do they enable translocation and how is the energy of thermal fluctuations rectified into forward movement? How does EF-G overcome the kinetic barrier of translocation? What does GTP hydrolysis accomplish? Answering these questions is of fundamental interest not only because they address key questions about molecular machines, but also because translation is an important target for antibiotics. 24 Furthermore, movements in molecular machines are interesting from a more broad evolutionary perspective, also given the recent observation that eukaryotic ribosome can employ a peculiar rolling motion. 25 Several recent studies now provide answers to these pressing questions. 26,27,28 We used a toolbox of fluorescence-labeled translocation components and looked at the moving parts of the ribosome complex in real time, covering multiple angles of translocation, much as a film director would do to better cover its main scene under different perspectives 26,27 (Fig. 1B). Movements, defined as changes in distances between the main players, i.e., tRNAs, mRNA, and different ribosome components, are inferred from the changes in the intensity of fluorescence resonance energy transfer (FRET) between fluorescence reporters attached to different positions. For example, FRET pairs on the SSU and LSU allow us to monitor subunit rotations and SSU swiveling, whereas different combinations of labels on EF-G, tRNA, mRNA, and the ribosome reveal tRNA-mRNA displacement and EF-G dissociation. Because the activity of these fluorescence-labeled components is indistinguishable from their non-labeled counterparts, one can integrate the information from all reports into a global model showing the choreography of molecular movements during translocation. In parallel, Blanchard and colleagues 28 used a similar toolbox of fluorescence reporters to study translocation by singlemolecule FRET (smFRET) techniques. In the following, we compare the 3 studies and suggest a picture of translocation as seen from these multicolor, multi-angle and multi-perspective studies.
The first questions that we asked were: How rapid is the spontaneous subunit rotation at conditions of rapid translation? And how does EF-G affect the subunit rotation? Using a combination of rapid ensemble kinetics and single-molecule FRET, we estimated the rates of subunit rotations for different tRNA pairs. 27 After the peptide bond is formed, the SSU rotates in CCW direction relative to the LSU and the rate of this spontaneous rotation is 40 s -1 (at 37 C) with any tRNA used (Fig. 2). However, the distribution between the rotated and non-rotated states depends on the tRNA, 27,29,30 due to differences in the rate of spontaneous reverse rotation. 27 EF-G can bind to either the rotated or the nonrotated state of the ribosome. 18,[31][32][33][34] But what happens when EF-G binds to a ribosome that remained in the non-rotated state? It turned out that binding of EF-G dramatically accelerates the rate of CCW rotation to 200 s -1 (ref. 27 ) and stabilizes the rotatedhybrid state. 3,8,28,31,35,36 At cellular concentrations of EF-G, the life-time of the rotated-hybrid state is very short, because EF-G binds very rapidly and drives the ribosome toward translocation.
Nevertheless, the rotated-hybrid state is an authentic translocation intermediate which serves to accelerate tRNA movement through the ribosome. 37-40 EF-G rapidly equalizes a potentially heterogeneous population of ribosome complexes to a uniform state poised for translocation. The high velocity of EF-G-catalyzed CCW rotation may explain why some smFRET studies did not capture the translocation route via the non-rotated state of the ribosome 16,28 : As soon as EF-G binds to the ribosome in the non-rotated state, the complex is rapidly converted into the rotated state, followed by translocation and CW rotation.
The next question is how the movements of the ribosome and the tRNA-mRNA complex are coordinated. We addressed this question using rapid ensemble kinetics. 26 The advantage of ensemble kinetics is its high temporal resolution. This allowed us to collect unperturbed translocation trajectories in the time range from milliseconds to seconds, which is the optimal time window for looking at tRNA movements. We designed a novel approach of unbiased fitting to decipher the choreography of movements: instead of trying to understand the nature of each of the FRET and fluorescence changes separately, we performed global data fitting using a minimum model that accounts for all data points (100,000) of the datasets collected with 9 different observables. The minimum model comprises 5 main kinetic steps of translocation; this does not mean that translocation cannot entail further steps, but those steps are unlikely to be rate-limiting, i.e. they should be rapid compared to the 5 steps identified by the global analysis. Another important outcome of the analysis are the values for FRET and fluorescence changes for each reporter and at each step (i.e., intrinsic Positions of fluorescent reporter groups on ribosomal proteins, tRNA, mRNA, and EF-G are indicated. The fluorescence change of the label on protein S13 reports on EF-G binding. FRET changes monitor the SSU rotation relative to the LSU (S6-L9), SSU head swiveling (S13-L33), or EF-G binding and dissociation (EF-G-L12). FRET changes between the P-site tRNA and SSU or LSU monitor the transit of the deacylated tRNA through the E site and its dissociation from the ribosome. SSU, light green; LSU, light cyan.
fluorescence intensities, IFIs), which are analogous to FRET values obtained by single-molecule measurements. These values indicate the directions of motions of the ribosomal subunits and the tRNAs and allow us to identify the point in time when unlocking of the ribosome takes place. 22 In the following, we describe these new findings against the background of previous reports identifying translocation intermediates by rapid kinetics 22,41-43 and single-molecule FRET. 31,35 We denote the EF-Ginduced translocation intermediates as chimeric states (CHI) in order to distinguish them from the non-rotated-classical (N-C) and the rotated-hybrid (R-H) pre-translocation (PRE) state or the post-translocation (POST) state formed after the dissociation of EF-G.
The first step identified by the kinetic analysis is the initial EF-G binding to the ribosome, which has an association rate constant of 55 mM -1 s -1 , and is reversible, with the dissociation constant of 65 s -1 (Fig. 2). In the next step, which proceeds at the rate of 85 s -1 (at 37 C), EF-G hydrolyzes GTP, engages in the translocation, and uncouples SSU head and body movements. The SSU body starts to move in CW direction, whereas the SSU head continues to swivel in CCW direction. At this step, movements of SSU head and body become uncoupled, which may remove the physical hurdles on the way of the tRNAs and thereby start the unlocking of the ribosome. The resulting CHI1 state converts to CHI2 at a rate of 43 s -1 , which corresponds to the rate-limiting step of unlocking and is then followed by rapid tRNA translocation and Pi release from EF-G. 22 The SSU body continues to move into CW direction, but now also the SSU head starts swiveling backward (Fig. 2,  inset). Single-molecule FRET studies suggested that the A-site tRNA moves toward the P site, as seen from the decrease in FRET between the tRNA and ribosomal protein L11. 31 This may coincide with the partial movement of the P-site tRNA toward the E site. 41 The exact structural rearrangements leading to unlocking and the following rapid movement of tRNAs and mRNA are not known. Engagement of EF-G domain 4 may promote translocation by altering the conformation of the SSU, e.g. by opening the mRNA-binding cleft, by stabilizing the open conformation of the E-site gate, or by displacing ribosome elements that act as hurdles for 30S translocation, such as intersubunit bridges B1a, B4, B7a, and B8. 4,7,8,33,[44][45][46] The sequence of the 2 following rapid steps is known from earlier ensemble kinetic work. 22,42,43,47 After unlocking, EF-G releases the GTPase product inorganic phosphate, Pi; in parallel, the tRNAs move toward, but not yet quite into, the P site on the LSU (Fig. 2), whereas translocation on the SSU is lagging behind. The resulting CHI3 state is short-lived, because the motions of the tRNA anticodon and CCA ends are normally synchronized with respect to both subunits 42 ; the CHI3 state was identified when subsequent steps were blocked by using EF-G mutants, non-hydrolyzable GTP analogs, or antibiotics. 31,42 During unperturbed translocation, the formation of CHI4 signifies the placement of the CCA end of the A-site peptidyl-tRNA in the P site of the LSU where it becomes reactive with puromycin, which is a diagnostic test for the completion of tRNA translocation. 42 Importantly, Pi release and tRNA translocation are 2 parallel, non-coupled reactions independent of one another. 22,47 Blocking tRNA translocation by viomycin, spectinomycin, streptomycin, paromomycin, hygromycin B, or tetracycline does not inhibit Pi release. 22,48 On the other hand, interfering with Pi release by introducing mutations in ribosomal protein L12 or mimicking the Pi-bound state by adding Pi analogs does not block single-round translocation. 22,49 Because the 2 parallel pathways (Pi release first, then translocation, or translocation first, then Pi release) eventually merge, further reactions on the translocation pathway, including EF-G dissociation, are blocked. 49,50 In the subsequent step, the tRNA, which has been displaced from the P to the E site, moves further through the ribosome via at least one distinct intermediate state (CHI5), and finally The swiveling motions of the SSU head are depicted by a color gradient from light green (classical non-swiveled SSU head position) to forest green (maximum degree of swiveling relative to the SSU body). 26 Peptidyl-and deacylated-tRNA in the PRE complex are shown in magenta and blue, respectively. EF-G (purple) is depicted in 2 conformations, a compact 33 and an extended one after engagement with the ribosome. 8,61 The rates of transitions between PRE(N) and PRE(R) and PRE(N)-EF-G and PRE(R)-EF-G are from ref. 27 The rates of EF-G binding and dissociation (step ①) are ensemble rate constants obtained for a mixture of N and R states 26 in which the PRE(R) state is predominant. 27 All other rate constants for the kinetically defined steps ②, ③, ④, and ⑤ are from ensemble kinetics studies with the PRE(fMF) complexes at 37 C. 26 The existence of rapid steps between steps ③ and ④ was demonstrated previously. 22,42 Translocation intermediates (CHI1 to CHI4) are adopted from an smFRET study 31 and are consistent with other smFRET data, 28,35 ensemble kinetics 22,26 and structural studies 3,8,61 . An additional intermediate, CHI5, was identified by ensemble kinetics 26 and smFRET. 28 The POST state may entail further conformational sub-states. 28 Kinetics of GTP hydrolysis and Pi release were described earlier. 22,23 The light red background indicates complexes undergoing unlocking; the light green background shows complexes that move toward relocking. Inset: Distinct timing of CCW and CW movements of the SSU body relative to LSU (blue symbols) and of the SSU head (green symbols) as indicated by normalized intrinsic fluorescence intensities (IFI).
dissociates from the ribosome 26 (Fig. 2). In parallel, the SSU head continues moving backward (the SSU CW rotation is almost complete at CHI4). Concomitantly, EF-G is released from the translocated complex in a stepwise manner -first from the ribosomal interface which then allows a smooth back swiveling of the SSU head and subsequently from the L12 stalk. The ribosomenow in the POST statepresents the SSU head and body in their non-swiveled, non-rotated states, the peptidyl-tRNA in the P site and empty E and A site, with the latter exposing the next mRNA codon which is immediately decoded by the next aminoacyl-tRNA. The dance of the ribosome along the translation axis can then continue with the next round of elongation. In summary, these data provide a picture of a smooth motion along the translocation axis, with a continuum of intermediate states that guide the tRNAs from the PRE to the POST state.
A recent single-molecule FRET study which utilized a similar approach of multi-angle fluorescence detection also suggests that the movements of the SSU head and body are de-coupled by EF-G. 28 The key conclusion, i.e. that the SSU CW body rotation precedes the CW head swiveling, is consistent with the ensemble kinetics study. 26 However, some details of the kinetic models 26,28 are difficult to reconcile due to the use of different observables in the 2 studies. Interestingly, the authors observed exaggerated SSU motions monitored by the S13-L5 FRET pair; it is not certain whether we observe a similar motion upon transition to CHI1 or this is yet a different type of SSU dynamics, as different labeling positions used in the 2 studies may report different motions. The exaggerated head-domain motions may represent the intrinsic motions of the protein L5 following the tRNAs movement, 15 or an as yet unidentified motion of the head outside of the swiveling trajectory (SSU head tilting) or even reflect a rolling motion similar to the one visualized for eukaryotic ribosomes. 25 In fact, subunit rolling was recently observed in SecM-stalled POST complexes from E. coli, 51 which would fit to the placement of this intermediate late on the translocation trajectory. 28 It appears that our model of translocation (Fig. 2) is more detailed with respect to the early unlocking steps, whereas ref. 28 identifies additional steps during relocking.
EF-G-mediated GTP hydrolysis plays a fundamental role in translocation. GTP hydrolysis accelerates the unlocking step leading to A to P and P to E tRNA movement (step 3) by 40fold, as demonstrated by the analysis of the translocation pathway with a slowly hydrolyzable analog (i.e., GTPgS) or a GTPase-deficient EF-G mutant 23,26,42,43 (Fig. 3A). Also, the rate of step 5 (release of EF-G and the E-site tRNA) is dramatically reduced when GTP hydrolysis is inhibited. However, even more strikingly, the comparison of the intrinsic fluorescence intensities (IFI) for the intermediates formed with either GTP or GTPgS suggest that preventing GTP hydrolysis leads to the appearance of a series of intermediates which are structurally different from those on the pathway with GTP hydrolysis (Fig. 3B, C). Thus, our data support the view that uncoupling of the motions of the ribosomal subunits from the movement of the tRNA-mRNA complex is a fundamental consequence of GTP hydrolysis and subsequent ribosome unlocking. EF-G appears to work as a wedge to uncouple the SSU head swiveling  ) and GTPgS (red symbols) for the SSU rotation (B) and SSU head swiveling (C). Intrinsic fluorescence intensities (IFI) reflect structural differences of the intermediates formed in steps ① to ⑤ (Fig. 2). from the SSU body rotation. EF-G may also actively push the A-site tRNA toward the P site 52 or act as a "door-stop" by following the spontaneous movement of the tRNA out of the A site with its domain 4 and restricting the backward movement of the tRNA. 7,8,53 While it is clear that a precise balance between synchronization and de-synchronization in SSU head and body motions is fundamental for optimum translocation, it would be of great interest to look at cases when their harmonious dance is disturbed. Antibiotics have already been used to uncouple tRNA-mRNA movements on the SSU and LSU, 42 but how the movements of these ligands correlate with the motions of the SSU body and head in the presence of inhibiting antibiotics is still insufficiently understood. Indeed, we know which ribosome conformations are stabilized by most of the known antibiotics and how ribosome dynamics are affected, 28,29,54,55 but what is still missing are the kinetics and the sequence of motions determining translocation inhibition and consequent translation arrest.
Less dramatic than a translational arrest by antibiotics, but of equally great importance, is the occurrence of translational pauses and their release during programmed ribosome frameshifting. In ¡1 programmed frameshifting the impaired backward rotation of the SSU head, assisted by a late release of EF-G, drives the slippage of the mRNA into the new reading frame. [56][57][58] Another exciting example of dynamics is ribosome bypassing, whereupon actively translating ribosomes "hop" over a large distance on the mRNA. 59,60 Understanding the mechanisms of such events from the molecular point of view could help in the design of molecules specifically targeting such motions. In the future, it would be even more challenging to investigate ribosome motions in the context of polysomes to understand whether the neighboring ribosomes in a polysome are synchronized, de facto linking ribosome dynamics and regulation of gene expression. Finally, insights into the fundamental principles of translocation on bacterial ribosomes will help to understand the more complex translational dynamics of the eukaryotic ribosome and to test whether a larger ribosome is an equally good dancer.

Disclosure of potential conflicts of interest
No potential conflicts of interest were disclosed.