Structural basis for +1 ribosomal frameshifting during EF-G-catalyzed translocation

Fig. 1. +1FS of E. coli tRNAPro(UGG) in a cell-based assay. a The lacZ reporter construct. A CCX-X codon motif was inserted to the lacZ gene next to the start codon AUG, requiring a +1FS event at the motif to produce a full-length β-gal protein. b–d Measured +1FS frequency in E. coli trmD-KO/JM101 strains expressing the Ara-controlled human trm5 to produce the m1G37+ (blue) or m1G37– (orange) condition. The +1FS frequency was reported for the strain expressing the CCA-A reporter (b), the CCC-A reporter (c), and the CCC-C reporter (d), by determining the β-gal activity as Miller Units normalized by that of the corresponding control strain. Data are shown as mean ± SD, n = 4, for all but three datasets. For datasets CCA-A/proM+/proL+/m1G37–, CCA-A/proM+/proL–/m1G37–, and CCC-A/proM+/proL–/m1G37–, n = 3. P values are calculated from Welch’s two-sided t test and are shown in (b–d).

Fig. 2. +1FS of E. coli tRNAPro(UGG) in a biochemical functional assay. a The reaction scheme of the functional assay to measure the yield of +1FS as the fractional conversion of fMP to fMPV. An E. coli 70SIC programmed with the non-slippery CCA-A reporter or the slippery CCC-A reporter was mixed with a TC containing EF-Tu-GTP with a native or transcript of tRNAPro(UGG), Val-tRNA (*UAC, *U = cmo5U anticodon), and Ser-tRNA (GCU anticodon) in the HF or CE buffer containing the indicated MgCl2 concentration. Each reaction was quenched after 5 min with 0.5 M KOH. Peptides were resolved by electrophoretic TLC and quantified by phosphor-imaging. b–f The fractional conversion of fMP to fMPV (pink) was reported for tRNAPro(UGG) in the native-state at the CCC-A codon motif at 37 °C (b); in the native-state at the CCC-A codon motif at 20 °C (c); in the native-state at the CCA-A codon motif at 20 °C (d); in the transcript-state at the CCC-A codon motif at 20 °C (e); and in the transcript-state at the CCA-A codon motif at 20 °C (f). All data are presented as mean ± SD. The bars in graphs b−f are SDs of three independent (n = 3) experiments; for datasets CCC-A transcript/HF(20),CE(3.5),CE(20); CCA-A…

Fig. 3. Cryo-EM structures of pre-translocation 70S formed with fMet-tRNAfMet (P site) and Pro-tRNAPro (A site). a Overall view of the 70S structure with non-frameshifting mRNA (CCA-A; Structure I). Weaker density in the E-site than in the A and P sites suggests partial occupancy of E-tRNA (“Methods”). b Cryo-EM density (gray mesh) for codon–anticodon interaction between non-frameshifting mRNA and tRNAPro in the A site of Structure I. The view approximately corresponds to the boxed decoding center region (DC) in (a). The map was sharpened with a B-factor of -80 Å2 and is shown at 2.5 σ. c Decoding center nucleotides G530 (in the shoulder region) and A1492-A1493 (in the body region) stabilize the codon–anticodon helix in Structure I. d Overall view of the 70S structure with the slippery mRNA (CCC-A; Structure I-FS). Weaker density in the E-site than in the A and P sites suggests partial occupancy of E-tRNA (see “Methods”). e Cryo-EM density (gray mesh) for codon–anticodon interaction between the slippery mRNA codon and tRNAPro in Structure I-FS. The map was sharpened with a B-factor of -80 Å2 and is shown at 2.5 σ. f Partially open conformation of the 30S subunit due to the shifted G530…

Fig. 4. Cryo-EM structures of mid-translocation states formed with EF-G•GDPCP. a Overall view of mid-translocation Structure II with the non-frameshifting mRNA. b Cryo-EM density (gray mesh) of the non-frameshifting tRNAPro and mRNA codon near the P site. The map is sharpened by applying the B-factor of -80 Å2 and is shown with 2.5 σ. c Interaction of the EF-G loop I (Ser509-Gly511, red) with the codon–anticodon helix (space-filling surface and cartoon representation). d Overall view of mid-translocation Structure II-FS with the frameshifting mRNA. e Cryo-EM density (gray mesh) of the frameshifting tRNAPro and mRNA codon near the P site. The map is sharpened by applying the B-factor of -80 Å2 and is shown with 2.5 σ. Note the unpaired and bulged C1 nucleotide in the mRNA, also shown in (i). f Interaction of the EF-G domain IV loop I (Ser509-Gly511) with the codon–anticodon helix of the frameshifting mRNA (compare to c). g Differences in positions of tRNAPro (green) and tRNAfMet (orange) in the frameshifting structure II-FS relative to those in the non-frameshifted structure II (gray). h Adjustment of loop II of domain IV of EF-G (red) to accommodate the shifted position of tRNAPro…

Fig. 5. Cryo-EM structures of near post-translocation states formed with EF-G•GDPCP. a Overall view of the near-post-translocation Structure III with the non-frameshifting mRNA. b Cryo-EM density (gray mesh) of the non-frameshifted tRNAPro and mRNA codon at the P site. The map was sharpened by applying the B-factor of -80 Å2 and is shown at 2.5 σ. c Overall view of the near post-translocation Structure III-FS with the frameshifting mRNA. d Cryo-EM density (gray mesh) of the frameshifted tRNAPro and mRNA codon at the P site. The map was sharpened by applying the B-factor of -80 Å2 and is shown at 2.5 σ. e Comparison of mRNA and tRNA positions in the nearly translocated frameshifted (colored, III-FS) and non-frameshifted (gray, III) complexes. f Positions of loop II of EF-G (red) and tRNAPro (green) in the Structures III-FS and III (gray). Structural alignments were performed by superposition of 16S rRNAs. The color scheme is as in Fig. 4.

Fig. 6. The mechanism of + 1 frameshifting. a Schematic of canonical ribosomal translocation by EF-G and ribosome rearrangements. b Schematic of ribosomal translocation by EF-G resulting in +1 frameshifting. The second rows in a, b show local rearrangements of mRNA-tRNA and positions of the decoding-center nucleotide G530 and P-site nucleotide G926 of the 30S subunit. The color scheme is as in Fig. 4.