Toeprinting analysis of translation arrest in the synthetic ORF RST2 (left) and the natural ORF ermBL (right) mediated by PrAMPs. The toeprint bands corresponding to the ribosomes arrested by Api137 or the natural apidaecin 1a (Api1a) at the stop codon of the ORF are indicated with orange arrowheads; the bands representing the ribosome arrested by Onc112 at the start codon are marked with blue arrowheads. Sequencing lanes are shown. The nucleotides corresponding to the toeprint bands are indicated in the gene sequence on the side of the gels; orange brackets indicate codons positioned in the P- and A- sites of the Api-stalled ribosome; blue brackets indicate codons in the P- and A- sites of the Onc112-stalled ribosome. The gels are representatives of two independent experiments.

a, Effect of the newly-isolated (marked by asterisks) or tested mutations on sensitivity of E. coli cells to Api137. The RF1 mutation is highlighted in orange; RF2 mutations, teal; rRNA mutations. gray; uL16, brown; uL4, blue; uL22, purple. Each MIC was determined in at least two independent experiments. b-d, Location of resistance mutations within the context of the terminating ribosome. b, Transverse section of the 50S ribosomal subunit (gray) of the 70S ribosome (30S subunit, yellow) showing the location of ribosomal proteins uL4, uL16, uL22 or 23S rRNA nucleotides (gray) whose mutations confer resistance to Api137. The region enlarged in (c) is boxed. c-d, Location of Api137 resistance mutations (spheres) in 23S rRNA (gray), ribosomal proteins uL4 (blue), uL16 (brown) and uL22 (purple), as well as (c) RF1 (orange) or (d) RF2 (teal). The GGQ motif of RF1 and RF2 is colored red in (c) and (d).

a, Dissociation of RF1(D241G) from the PostHC in the presence of Api137. RF1(D241G)_Qsy was incubated with PreHC_Flu (0.05 μM) to generate PostHC_Flu and then mixed with a 10-fold excess of unlabeled RF1 and RF3·GTP in the absence (gray) or in the presence (black) of Api137 (1 μM). The traces represent the average of up to eight technical replicates. No dissociation of wt RF1 in the presence of Api137 was observed under the same experimental conditions (Fig. 2f). b, Peptide hydrolysis by K12 strain-specific RF2(Ala246Thr) at turnover conditions in the absence (open circles) or in the presence (closed circles) of Api137 (1 μM). In the presence of Api137, the peptide hydrolysis reaction proceeds faster when it is catalyzed by the K12 strain RF2, compared to the B strain RF2 (Fig. 2c).

a, In silico sorting was performed with the FreAlign 9.11 software package (as described in Grigorieff, N., J. Struct. Biol. 157, 117-125 (2007)). Initial alignment of 116,212 particles was followed by 3D classification, resulting in six different classes. Class 1 (38,203 particles) was further refined, yielding a (b) final reconstruction consisting of 36,826 particles, with (c) an average resolution of 3.4 Å (based on the Fourier shell correlation (FSC) curve at FSC 0.143). d, Validation of the fit of molecular models to cryo-EM map for the Api137-RF1-70S complex. FSC curves calculated between the refined model and the final map (blue), with the self- and cross-validated correlations in orange and black, respectively. Information beyond 3.4 Å was not used during refinement and preserved for validation. (e) Side view and (f) transverse section of the cryo-EM map of Api137-RF1-70S complex colored according to local resolution as shown previously (Kucukelbir, A., Sigworth, F. J. & Tagare, H. D., Nat. Methods 11, 63-65 (2014)). g-h, Cryo-EM density for Api137 (g) colored according to local resolution and (h) shown as gray mesh with molecular model for residues 5-18.

a, RF1 (orange), deacylated P-site tRNA (green) and Api137 (salmon) in the Api137-RF1-70S complex. The position of RF1 during canonical termination is shown in blue (PDBID 5J30; Pierson, W. E. et al., Cell Rep. 17, 11-18 (2016)). Boxed regions are zoomed in the panels (b) and (c). b, Interaction of the PAT motif of RF1 (orange) with the UAG stop codon of the mRNA (cyan) in the Api137-RF1-70S complex. c, A2602 of the 23S rRNA is in the rotated conformation as observed in previous RF1-70S structures (Korostelev, A. et al., EMBO J. 29, 2577-2585 (2010); Pierson, W. E. et al., Cell Rep. 17, 11-18 (2016); Laurberg, M. et al., Nature 454, 852-857 (2008); Svidritskiy, E. & Korostelev, A. A., Structure 23, 2155-2161 (2015)). Conformation of A2602 (gray) in Api137-RF1-70S complex compared to A2602 (blue) during canonical termination (PDBID 5J30; Pierson, W. E. et al., Cell Rep. 17, 11-18 (2016)) and A2602 (slate) from the pre-attack state (PDBID 1VY4; Polikanov, Y. S., Steitz, T. A. & Innis, C. A., Nat. Struct. Mol. Biol. 21, 787-793 (2014)). Api137 (salmon) and P-site tRNA (green) are shown for reference. d, e, The binding position of Api137 (salmon) relative to the (d) MifM nascent chain (dark green; Sohmen, D. et al., Nat. Commun. 6, 6941 (2015)) or (e) antimicrobial peptide Onc112 (slate; Seefeldt, A. C. et al., Nat. Struct. Mol. Biol. 22, 470-475 (2015)). In (d) and (e) the orientations of the peptides are indicated.

(a-c) Location of residues in (a) RF1 (orange) and (b, c) RF2 (teal) that increase resistance to Api137 when mutated. Site of mutations are shown in stick and surface representation and glutamines of the GGQ motif (Gln235 in RF1 and Gln252 in RF2) are shown as sticks for reference. (d-f) Location of Api137 resistance mutations in ribosomal proteins. (d) Lys63 in uL4 (blue) interacts with 23S rRNA residues G2061 and G2444. (e) Deletion of ₈₂MKR₈₄ (outside of the figure boundaries) in uL22 (purple) confers resistance to Api137 presumably by changing the geometry of the uL22 exit tunnel loop and disrupting Api137 interaction with neighboring 23S rRNA nucleotides (gray), such as A751. (f) The mutation of Arg81 in uL16 (brown) may relieve Api137-mediated RF1 and RF2 trapping by indirectly destabilizing interactions of deacylated tRNA with the P-site mediated by G2251 of the 23S rRNA.