Structural basis of early translocation events on the ribosome

Abstract

Peptide-chain elongation during protein synthesis entails sequential aminoacyl-tRNA selection and translocation reactions that proceed rapidly (2-20 per second) and with a low error rate (around 10^-3 to 10^-5 at each step) over thousands of cycles¹. The cadence and fidelity of ribosome transit through mRNA templates in discrete codon increments is a paradigm for movement in biological systems that must hold for diverse mRNA and tRNA substrates across domains of life. Here we use single-molecule fluorescence methods to guide the capture of structures of early translocation events on the bacterial ribosome. Our findings reveal that the bacterial GTPase elongation factor G specifically engages spontaneously achieved ribosome conformations while in an active, GTP-bound conformation to unlock and initiate peptidyl-tRNA translocation. These findings suggest that processes intrinsic to the pre-translocation ribosome complex can regulate the rate of protein synthesis, and that energy expenditure is used later in the translocation mechanism than previously proposed.

Fig. 1. Early kinetic and structural intermediate of tRNA₂–mRNA translocation.

a, Schematic of the translocation reaction coordinate in bacteria depicting SSU body-rotation (blue) and head-swivel (purple). tRNAs are coloured on a gradient from the A (green) to P (yellow) to E (orange) sites. The states enclosed in dashed boxes were characterized in this study. Green (donor, uS13, LD550) and red (acceptor, uL1, LD650) circles denote fluorophore positions (see b). FA, fusidic acid. b, Population FRET histograms showing FRET evolution over time upon EF-G injection with buffer, SPC (3 mM) or fusidic acid (FA, 400 μM). n represents the number of observed molecules. c, Overview of the INT1 ribosome structure captured by SPC, coloured as in a.

Fig. 2. Unlocking of the peptidyl-tRNA decoding centre.

a, Locally filtered electron density illustrating shape-specific recognition of the A-site codon–anticodon pair by EF-G (red) in its active, GTP-bound conformation (INT1). b, c, Unlocking of the tRNA₂–mRNA decoding centre in the PRE-H1 (b) to INT1 (c) transition. Peptidyl-tRNA, green; mRNA, pink; H69, grey; h44, blue; h18, cyan; uS12, light blue. Threshold σ = 5.

Fig. 3. Overview of the active, GTP-bound conformation of EF-G.

a, Domain architecture of EF-G in its active, GTP-bound conformation (INT1, coloured) and in a post-hydrolysis conformation (INT2, grey, G-domain alignment). b, Locally filtered electron density (mesh) in the nucleotide-binding pocket for INT1. c, Elongated switch-I (residues 38–68, yellow) contacts with the SRL (grey), the SSU (light blue), DII (orange) and DIII (pink). The conformational change of DII is indicated with an arrow. Threshold σ = 6.

Fig. 4. Non-uniform tRNA₂–mRNA movement during translocation.

a, Overlay of the tRNA₂–mRNA module from the A (green) to P (yellow) to E (orange) sites. b, Overlay from a, viewed from the codon–anticodon interface. Circles on the tRNAs at position 34 N1 (deacyl-tRNA, left) and N3 (peptidyl-tRNA, right) depict the tRNA trajectories during translocation. c, tRNA anticodon–mRNA codon movement during translocation, same perspective as b.

Extended Data Fig. 1. smFRET investigations of translocation.

a–g, smFRET data on translocation inhibition by SPC. a, Schematic of translocation indicating the positions of the donor (LD550, uS13 N terminus) and acceptor (LD655, peptidyl-tRNA U47) fluorophores, and FRET efficiency values for the indicated states. b–e, Population FRET histograms showing time evolution of FRET between uS13 and peptidyl-tRNA at a time resolution of 400 ms on delivery of 5 μM EF-G in either buffer A (b, d) or B (c, e), without (b, c) or with 3 mM SPC (d, e). N indicates the number of observed molecules. f, Example smFRET trace from the data in b–e. EF-G, injected approximately 4 s after the start of data acquisition, rapidly binds to the PRE complex (0.14 FRET), converting it into INT1, and subsequently INT2 (both 0.3 FRET), before achieving the POST state (0.5 FRET). g, Kinetic analysis of data as in b–e, demonstrating an approximately tenfold potentiating effect of buffer B on SPC inhibition. The points represent mean accumulation of translocated ribosomes with time, defined as reaching the 0.5 FRET state. Solid lines represent fits of bi-exponential functions to the data. Error bars indicate standard deviations derived from three technical experimental replicates. h–o, Approximate positions of published donor (green sphere) and acceptor (red sphere) dyes on the INT1 structure. h, Donor tRNA^Phe (Cy3, s⁴U8), acceptor fMet-Phe-Lys-tRNA^Lys (Cy5, acp³U47)^,. i, Donor tRNA^Phe (Cy3, acp³U47), acceptor fMet-Phe-Lys-tRNA^Lys (Cy5, acp³U47). j, Donor uS13 (LD550, N-terminal ACP), acceptor fMet-Phe-Lys-tRNA^Lys (Cy5, acp³U47). k, Donor uL11 (Cy3, residue 87), acceptor fMet-Phe-Lys-tRNA^Lys (Cy5, acp³U47). l, Donor uS13 (LD550, N-terminal ACP), acceptor uL5 (LD650, N-terminal ACP). m, Donor bL9 (Cy3, residue N11C), acceptor bS6 (Cy5, residue D41C). n, Donor fMet-Phe-Lys-tRNA^Lys (Cy3, acp³U47), acceptor EF-G (Cy5, C-terminal SFP). o, Donor EF-G (bifunctional rhodamine, residues 467–474), acceptor uL11 (Cy5, residue 87). p, Schematic of translocation indicating the positions of donor (LD550, uS13 N terminus) and acceptor (LD650, uL1) fluorophores and FRET efficiency values for the indicated states in q–t. q–t, Population FRET histograms showing time evolution of FRET between uS13 and uL1 upon injection of 5 μM EF-G with either 1 mM GTP (q, Apo), 1 mM GTP and 3 mM SPC (r), 1 mM GTPγS (s) or 1 mM GTP and 400 μM fusidic acid (t, FA), revealing that the non-hydrolysable GTP analogue GTPγS stalls the ribosome in both the same states as SPC (INT1) and fusidic acid (INT2), whereas SPC stalls only in INT1 and fusidic acid stalls only in INT2. N indicates the number of observed molecules.

Extended Data Fig. 2. Fourier shell correlation and local resolution for cryo-EM structures along the translocation reaction coordinate.

a–f, Fourier shell correlation (FSC) resolution curves for PRE-C (a), PRE-H2* (b), PRE-H1 (c), POST (d), INT1 (e) and INT2 (f) structures obtained by masking the two half maps and calculating the cross-resolution between the masked volumes in Relion. Resolution was estimated using the 0.143 cut-off criterion (dotted line). g–l, Local resolution electron density maps for PRE-C (g), PRE-H2* (h), PRE-H1 (i), POST (j), INT1 (k) and INT2 (l). CP, central protuberance. LSU, grey background; SSU, blue background. Threshold σ = 5. See also Methods and Supplementary Table 1.

Extended Data Fig. 3. Structural evidence of tRNA identity for cryo-EM structures along the translocation reaction coordinate.

a, Locally filtered electron density for cognate tRNA anticodon–mRNA codon (pink) interactions with deacyl-tRNA^Phe (AAG, left, yellow/orange) and peptidyl-tRNA^Lys (UUU, right, green) from PRE-C (top) to POST (bottom). SSU head (purple) and body (blue). Threshold σ = 6. b–d, Locally filtered electron density for modified tRNA bases in support of tRNA assignment from PRE-C (top) to POST (bottom): deacyl-tRNA^Phe(AAG) ms²i⁶A37 (MIA, b); peptidyl-tRNA^Lys(UUU) t⁶A37 (T6A, c); peptidyl-tRNA^Lys(UUU) mnm⁵s²34 (U8U, d). Arrows designate defining density for each modification. Coloured as in a. e, Locally filtered electron density for the nascent peptide (fMet-Phe-Lys, red). Deacyl-tRNA, yellow/orange; peptidyl-tRNA, green; A2602, dark green; P loop, dark blue. Threshold σ = 5 for PRE-C, INT1 and INT2. Threshold σ = 7 for PRE-H2*, PRE-H1 and POST.

Extended Data Fig. 4. Global conformational changes within the ribosome that define the translocation reaction coordinate.

a, Schematic of the translocation reaction coordinate in bacteria depicting SSU rotation (blue) with respect to the LSU (grey) and SSU head-swivel (purple) processes. tRNAs are coloured on a gradient from the A (green) to P (yellow) to E (orange) sites. Deacyl-tRNA dissociation (orange) can occur at multiple steps after INT2. tRNA positions are depicted in chimeric-hybrid notation (ssu head SSU BODY/LSU). b, SSU conformational changes accompanying each sequential translocation step, viewed from inside the intersubunit space (left) and towards the intersubunit space from the head domain (inset), coloured by r.m.s.d. at each SSU residue for each transition. Degree of shoulder domain closure, SSU body-rotation and SSU head-swivel as compared to POST is indicated as ‘total’. Cylindrical axes and bolded text indicate the degree of SSU body-rotation (transparent black axis, right, LSU core alignment) and SSU head-swivel (solid black axis, right, SSU body alignment) as compared to the previous state on the reaction coordinate. Threshold σ = 5. c, Deacyl-tRNA (yellow/orange), peptidyl-tRNA (green/yellow) and EF-G (red) movements during translocation. Current tRNA and EF-G positioning (solid coloured, outlined), previous position (transparent colour, no outline) and next position (white, solid outline). Alignment on the LSU core. Camera perspective is identical for all images. See also Supplementary Table 2.

Extended Data Fig. 5. LSU interactions with tRNA and the SSU in each characterized ribosome structure along the translocation reaction coordinate.

a, Location of intersubunit bridge B1 on the ribosome (top, left) and overlay of uL31 model (dark blue) depicting the conformational change during translocation (top, right). Sphere is positioned on residue Ile66 (Cα). Bottom, bridge B1 interactions between the LSU central protuberance and the SSU head domain (purple) in each structural state, including the ASF (H38, dark green) to uS13 (B1a), uL5 to uS13 (B1b) and uL31 to uS13/uS19 (B1c). Threshold σ = 4. b, Location of the L1 stalk, uL5 (dark grey), uL16 (light grey) and ASF (dark green) on the INT1 ribosome (top). L1 stalk positioning and interaction with deacyl-tRNA (yellow/orange) elbow in each classified structure (bottom). c, Bridge B1 interactions between the central protuberance and deacyl-/peptidyl-tRNA from PRE-C (top) to POST (bottom). P-site tRNA G19–C56 pair contacts include LSU H84 base A2309 and uL5 N-terminal Arg80. A-site tRNA contacts include G19–C56 pair coordination by ASF base A896 and TΨC stem packing against uL16 residues Arg6 and Lys7. A site (green circle), P site (yellow circle), E site (orange circle) regions are shown. d, Locally filtered electron density illustrating peptidyl transferase centre interactions along the translocation interaction coordinate. Watson–Crick pairing between the 3′-CCA tRNA ends and the peptidyl transferase centre. P loop (LSU rRNA 2251–2253, 2450–2451), dark blue; A loop (LSU rRNA 2553–2555, 2582–3585, dark grey; peptide, red; LSU base A2602, dark green; deacyl-tRNA, yellow-orange; peptidyl-tRNA, green; Threshold σ = 7. Camera perspective is identical for images within each panel. Alignment on the LSU core. See also Supplementary Video 2.

Extended Data Fig. 6. Changes in tRNA conformation during translocation.

Alignment of deacyl-tRNA (rows 1 and 2, orange) and peptidyl-tRNA (bottom, green) globally (left column), on the tRNA acceptor arm domain (positions 1–6 and 50–72, middle column, grey oval) or on the tRNA anticodon stem loop (positions 30–40, right column, pink circle). Deacyl-tRNA is coloured on a gradient from yellow to orange based on the position on the reaction coordinate (PRE-C, PRE-H2*, PRE-H1, INT1, INT2 and POST). Peptidyl-tRNA is coloured on a gradient from green to yellow. Blue circle annotates the position of the tRNA elbow. Alignment is on the outlined circle. See also Supplementary Table 2 and Video 4.

Extended Data Fig. 7. SSU unlocking on the leading-edge during translocation.

a, Locally filtered electron density illustrating the interaction between the SSU shoulder domain (cyan) A532 base, the SSU head domain (purple) G1206 of h34 and uS3 (dark purple) 193 loop near the A-site mRNA and EF-G (red) binding site. b, Locally filtered electron density illustrating the points of contact involved in unlocking the tRNA₂–mRNA module from the SSU body domain monitoring bases (A1492, A1493) in the A site. Peptidyl-tRNA, green; mRNA, pink; H69, grey; h44, blue; h28, dark blue; h18, cyan; uS12, light blue; h34, purple; EF-G, red. c, Locally filtered electron density and molecular models from b, viewed from beneath the SSU A site. Coloured as in b. d, Cartoon schematic depicting A-site unlocking. Dotted lines indicate weak electron density. Coloured as in b, c. Camera perspective is identical for images in each panel. Alignment on LSU core. Threshold σ = 5. See also Fig. 2.

Extended Data Fig. 8. Overview of the EF-G conformations and interactions in INT1 and INT2.

a, EF-G binding site (left) and domain architecture (middle and right) in its active, GTP conformation (INT1). Model for INT2 EF-G is shown as transparent in the right panel. b, Interaction between EF-G DIV (red) and the mRNA (pink) peptidyl-tRNA (green) codon–anticodon minihelix in the A site in the INT1 (left) and INT2 (right) structures (top). Locally filtered electron density illustrating the interaction between EF-G DIV loop I Lys504 and peptidyl-tRNA (row 2), EF-G DIV loop II Gly542 and C1210 of SSU h34 (purple, row 3) and EF-G DIV loop II His584 and peptidyl-tRNA (row 4). Threshold σ = 3. c, GTP/GDP-P_i conformation of EF-G (INT1, coloured) compared to a published structure of an EF-G homologue (grey, EF-G-2, PDB ID: 1WDT). Degree of bend angle between the G domain and DIV compared to INT1 is indicated. Conformational change in DII is indicated with an arrow. Coloured as in a. d, EF-G DII conformational change (arrow) in the INT1 (left) to INT2 (right) transition, oriented by the G domain (Asn37), switch I (Trp52) and DII (His367, Asn369 and Arg371), rearranging DII contact with SSU h5 via Phe329 and Arg362. Coloured as in a. e, EF-G rotation of approximately 15° around the SRL (grey) and into the A site in the INT1 (left) to INT2 (right) transition. This movement placed SSU head nucleotide C1210 (h34, purple) in direct contact with EF-G DIV in INT2. Coloured as in a. Threshold σ = 5. f, Conformational rotation of the superdomain DI–DIII of EF-G in the INT1 (solid) to INT2 (transparent) transition (approximately 17°, left). Models of EF-G were superimposed on the G domain. r.m.s.d. of EF-G in the INT1 to INT2 transition mapped on the INT1 structure (right). Density absent in the INT2 structure is depicted in dark blue. Alignment on the LSU core, unless otherwise stated.

Extended Data Fig. 9. E-site mRNA codon interactions with SSU h23, uS7 and uS11.

Locally filtered electron density illustrating the interaction between mRNA and the SSU in the E site. The black arrow annotates the −1 mRNA position, which flips in base orientation during translocation. The orange arrow annotates the gap between SSU proteins uS7 and uS11. mRNA, pink; uS7, purple; uS11, cyan; h23, blue; h24, light blue; h28, dark blue; deacyl-tRNA, yellow; peptidyl-tRNA, green. A putrescine molecule (PUT, grey) is modelled proximal to the −1 ribose in the PRE-H2* and PRE-H1 states. Alignment on the LSU core. Threshold σ = 5.

Extended Data Fig. 10. EF-G nucleotide-binding pocket in INT1 and INT2.

a, INT1 EF-G domain architecture (left) and zoom in of locally filtered electron density illustrating nucleotide interactions, including catalytic residues His92 of switch II and G2661 and G2662 of the SRL (grey, right). Threshold σ = 6. b, Locally filtered electron density at different thresholds (σ = 4, 6, 8, 10) for the phosphates in the nucleotide-binding pocket of EF-G in INT1. c, Alternative modelling of GDP-P_i (dark blue) in the nucleotide-binding pocket of INT1 compared to GTP (orange). Ligand cross correlation (CC) value reported from the corresponding PDB validation reports. Distances were measured between γ phosphate, β phosphate and His92 amine group. d, Locally filtered electron density illustrating the nucleotide-binding pocket of EF-G in the INT2 structure. Threshold σ = 6. G domain, red; DII, dark orange; DIII, strawberry; DIV, hot pink; DV, yellow/orange; P loop, green; switch I, yellow; switch II, lime; Mg²⁺, dark green. See also Fig. 3.

Extended Data Fig. 11. SPC-binding sites evidenced in INT1.

a, SPC-binding site (red) on the SSU between the head (purple) and body (blue) domains beneath the SSU P site. b, Zoom-in of the primary SPC-binding site showing that the convex face of SPC sits in the major groove of h34 (purple, between A1193 and G1064) and that the concave face of SPC sits between h28 (dark blue, G1387) and uS5 (cyan). Lys26 of uS5 reaches to interact with the h28 phosphate G1387. c, Overlay of the SSU SPC-binding pocket for INT1 (coloured) and INT2 (white), showing collapse of the SPC binding site and reorientation of Lys26 during SSU head-domain swivel. SPC shown in ball-and-stick and transparent-sphere representation. Distance changes for h28 (G1387 C1′), h34 (A1196 C1′) andh35 (C1066 C1′) are indicated. d, Electron density indicative of a second SPC-binding site near the exit tunnel of the LSU between uL22 and nucleotides G488 and A1284. Threshold σ = 6. e, Overlay of the SSU from PRE-C (light grey), PRE-H2* (light blue), PRE-H1 (blue), INT1 (light purple), INT2 (purple) and POST (dark grey) aligned on the SSU body domain to illustrate head-domain swivel. See also Supplementary Video 6.

Extended Data Fig. 12. Coordinated translocation of the tRNA₂–mRNA module.

Overview of the helical architecture involved in tRNA₂–mRNA translocation (left). a–f, Interactions between the SSU and the tRNA₂–mRNA module depicted with molecular model and electron density (left) together with a schematic representation (right) for PRE-C (a), PRE-H2* (b), PRE-H1 (c), INT1 (d), INT2 (e) and POST (f). A-site contacts with the tRNA₂–mRNA module depicted here include (1) anchored h34 (plum) SSU head base C1054 stacking on tRNA at position 34, (2) intercalation of h28 (dark blue) C1397 into mRNA (pink) downstream of the A-site codon (PRE +7/+8, POST +10/+11) and (3) hydrogen bonding of the h18 (cyan) SSU shoulder G530 base with the A-site wobble position. P-site interactions include (4) anchored stacking of h31 (purple) SSU head base m²G966 at tRNA position 34, (5) h28 C1400 base stacking against the P-site wobble position, (6) A-minor interactions of SSU head bases G1338/A1339 with the tRNA minor groove, (7) electrostatic interactions of uS9 (light purple) C-terminal Arg130 residue with the anticodon U-turn motif and (8) hydrogen bonding of the h28 G926 base with the phosphate backbone between the −1 and +1 mRNA positions, anchored by A1505 (aqua). E-site interactions include (9) the anchored stacking of the SSU body h23 (light blue) G693 base against the −3 mRNA base. Distances in the schematic are not to scale. Dotted residues display weak electron density. Grey circles depict contacts that are unchanged from the previous state, yellow circles depict contacts that are different from the previous state. Peptidyl-tRNA, green; deacyl-tRNA, orange. Camera perspective is identical for all images. Alignment on the LSU core. Threshold σ = 6.