Abstract
Antibiotic resistance ABC-Fs (ARE ABC-Fs) are translation factors that provide resistance against clinically important ribosome-targeting antibiotics which are proliferating among pathogens. Here, we combine genetic and structural approaches to determine the regulation of streptococcal ARE ABC-F gene msrD in response to macrolide exposure. We show that binding of cladinose-containing macrolides to the ribosome prompts insertion of the leader peptide MsrDL into a crevice of the ribosomal exit tunnel, which is conserved throughout bacteria and eukaryotes. This leads to a local rearrangement of the 23 S rRNA that prevents peptide bond formation and accommodation of release factors. The stalled ribosome obstructs the formation of a Rho-independent terminator structure that prevents msrD transcriptional attenuation. Erythromycin induction of msrD expression via MsrDL, is suppressed by ectopic expression of mrsD, but not by mutants which do not provide antibiotic resistance, showing correlation between MsrD function in antibiotic resistance and its action on this stalled complex.
Similar content being viewed by others
Introduction
ABC-F ATPases are members of the ATP-Binding Cassette (ABC) superfamily. Unlike other ABC proteins, the ABC-Fs are translation factors and some of them, termed antibiotic resistance ABC-Fs (ARE ABC-Fs), confer resistance to clinically important antibiotics that target the ribosomal peptidyl transferase center (PTC) and/or the nascent peptide exit tunnel (NPET)1,2,3,4,5,6,7. ABC-F proteins are composed of two ABC domains joined by a linker domain called P-site tRNA-interaction motif (PtIM)8,9, also termed antibiotic resistance determinant (ARD) for ARE ABC-Fs1,2,3. The two ABC domains dimerize upon binding of two ATP molecules and in this conformation the factor can bind to the ribosomal E site9, where the PtIM adopts an α-helical hairpin conformation that directly interacts with the peptidyl-tRNA and extends toward the PTC/NPET. Three antibiotic resistance phenotypes are associated with ARE ABC-Fs: (i) Macrolides, Ketolides, Streptogramins group B (MKSB); (ii) Pleuromutilins, Lincosamides, Streptogramins group A (PLSA); (iii) Phenicols, Oxazolidinones (PhO)10,11. However, despite structural investigations1,2,3,5,7, some questions remain about the exact molecular mechanism of action of ARE ABC-Fs.
Over the last 40 years, biochemical and structural investigations demonstrated that the nascent chains (NC,) which elongate into the ribosomal tunnel, can interact with metabolites or antibiotics, thus adapting protein synthesis to environmental cues12,13,14,15,16. In human pathogens, antibiotic-dependent formation of stalled ribosome complexes (SRCs) on regulatory ORFs, named leader peptides, can subsequently allow the regulation of downstream resistance genes in response to antibiotic exposure17,18,19,20.
The ARE ABC-F gene msrD is a member of the msr (macrolide and streptogramin B resistant) gene group previously referred to as mel21,22. This gene is generally found in operon with a macrolide efflux facilitator (mefA or mefE) gene and the two corresponding proteins act synergistically to confer macrolide resistance23,24,25,26. This operon is a part of the Macrolide Efflux Genetic Assembly (MEGA)27,28 that disseminates in human pathogens on numerous mobile genetic elements: conjugative transposons (tn916-type)29 and conjugative prophages (Φ1207.3)30 among Streptococci; conjugative plasmids (pMUR050)31 among Enterobacteria. In Streptococci, the operon is transcribed in the presence of erythromycin (ERY) as a polycistronic mRNA from a single promoter located ∼350 bp upstream of the mefA32,33 gene (Fig. 1a). It is also regulated by ribosome-mediated transcriptional attenuation via the MefAL leader peptide (MTASMRLR), which is closely related to the ErmDL leader peptide (MTHSMRLRFPTLNQ), both polypeptides harboring the characteristic macrolide-arrest motif Arg/Lys-X-Arg/Lys, so called “+x + ” motif14,33,34,35.
a Organization of mefA/msrD macrolide resistance operon. In the presence of ERY, ribosomes stall during translation of mefAL leading to transcription anti-attenuation, both mefA and msrD being then transcribed. Similar mechanism occurs during translation of msrDL, which regulates transcription of msrD only. b Sequence alignment of various msr homologs visualized with Jalview according to percentage of identity. Positions of tested mutations are highlighted in red and indicated by arrows. Main features of ABC-F proteins are indicated on the schematic. Genebank accession numbers are indicated between brackets. See also Supplementary Fig. 1b. c Relative growth of E. coli DB10 expressing msrD variants in the presence of ERY after 24 h. Optical densities were normalized relative to optical densities of E. coli DB10 pBAD-Control grown in the absence of ERY. Error bars represent mean ± s.d. for triplicate experiments. See also Supplementary Table 1. Source data are provided as a Source Data file.
Here, we report the presence of a second transcriptional attenuator on the mefA/msrD operon regulating exclusively the msrD expression upon macrolide exposure. The presence of ERY induces ribosomal stalling on the previously identified msrDL leader peptide (encoding MYLIFM)34, allowing RNA polymerase (RNAP) to bypass an intrinsic terminator. Our findings demonstrate that the stalling occurs due to hampered translation termination on the UAA stop codon, and inhibition of action of both release factors 1 and 2 (RF1 and RF2). Our results are supported by the cryo-electron microscopy (cryo-EM) structure of MsrDL-SRC that provides molecular insights on how the PTC precludes productive accommodation of RF1/RF2 and how the NC discriminates tunnel-bound cladinose-containing macrolides. The path of the NC within the tunnel greatly differs from previously described leader peptides12,13,14,15,16 with the NC latching into a ribosomal crevice, that was shown to be conserved from prokaryotes to eukaryotes. This crevice is delimited by the 23 S rRNA nucleotides U2584, U2586, G2608 and C2610 that may form a ribosomal functional site. Finally, our results demonstrate that the two ATPase sites of MsrD are functionally asymmetric and the protein can negatively self-regulate its own synthesis in the presence of ERY.
Results
MsrD provides macrolide and ketolide resistance
The mefA/msrD macrolide resistance operon (Fig. 1a) is part of the MEGA element currently spreading among clinical and livestock isolates33,36,37,38. A phylogenetical analysis (Supplementary Fig. 1a) shows that the operon disseminates predominantly in Gram-positive firmicutes (mostly Streptococci) and in some Gram-negative proteobacteria (such as Haemophilus influenzae, Neisseria gonorrhoeae or Escherichia coli). The gene msrD shares ~62 % identity with its closest homolog msrE7, both factors exhibiting the canonical ABC-F domains organization (Fig. 1b and Supplementary Fig. 1b).
The model bacterium Escherichia coli (E. coli) presents undeniable advantages for genetic and molecular biology, but the presence of multidrug efflux systems greatly limits its use in antibiotic research39,40. To circumvent this limitation, we took advantage of the E. coli DB10 strain, which exhibits exacerbated sensitivity to macrolide antibiotics41,42. Heterologous expression of the wild-type (WT) gene msrD from an arabinose-inducible promoter conferred macrolide and ketolide resistance phenotype to the strain (Table 1), demonstrating the functionality of the factor in this organism. Expression of msrDWT resulted in eightfold minimal inhibitory concentration (MIC) increase for the 14-membered macrolide ERY, 8-fold for the 15-membered macrolide azithromycin (AZI) and 2-fold for telithromycin (TEL, a ketolide antibiotic derived from the 14-membered macrolides). Similar results were reported for Streptococcus pneumoniae clinical isolates23. However, no change in MIC was observed for the 16-membered macrolides tylosin (TYL) and spiramycin (SPI), or the non-macrolide antibiotics lincomycin (LNC), linezolid (LNZ) and retapamulin (RTP) (Table 1).
To understand the function of the ATPase activity of MsrD, we constructed ATPase-deficient variants MsrDE125Q, MsrDE434Q and MsrDE125Q/E434Q (hereafter referred as MsrDEQ2) by replacing the catalytic glutamic acid residue (E) in the Walker B motif of each or both ABC domains by a glutamine (Q) (Fig. 1b and Supplementary Fig. 1b). This mutation strongly reduces the ATP hydrolysis while preserving the local stereochemistry of the active site and has been extensively used to trap ABC-F factors on the ribosome in ATP-bound conformation1,2,3,5,9. Protein expression levels have been assessed by western blot for each tested variants (Supplementary Fig. 1c).
We evaluated the influence of E-to-Q mutations of MsrD on bacterial growth and antibiotic resistance phenotypes and showed that all the mutations impair both phenotypes (Fig. 1c). While MsrDE125Q and MsrDEQ2 mutations had a similar effect on bacterial growth in the absence or at low ERY concentration, the MsrDE434Q mutant was more toxic. However, MsrDE125Q maintained some resistance (MIC = 4 µM) compared to control (MIC = 2 µM). This observation was supported by the half maximal inhibitory concentration (IC50) measurement of the MsrDE125Q strain which was ~8 times higher than Control (Fig. 1c and Supplementary Table 1). The two ATP hydrolysis sites (Supplementary Fig. 1b) were therefore concluded to perform different tasks and are functionally asymmetric, as previously reported for the ARE ABC-F Vga(A)43.
We also evaluated the importance of other MsrD residues. The truncation of the PtIM (replaced by three glycines) or just the Loop (∆PtIM, ∆Loop respectively) completely abolished the resistance phenotype in vivo (Fig. 1c and Supplementary Fig. 1e). This has been previously reported in vitro for the MsrE7. Punctual mutations in the Loop (Fig. 1c) affected the resistance phenotype to a different extent. Mutation of residue R241 to alanine had almost no effect on the phenotype, but equivalent mutation for residues L242 or H244 reduced the IC50 by ~3 fold (Fig. 1c and Supplementary Table 1). This demonstrated that these residues are important, but are not essential for MsrD function despite being predicted to be located at the vicinity of the antibiotic (Supplementary Fig. 1d) in the MsrE-ribosome complex7. Interestingly, this finding contrasts with the mutagenesis assays for MsrE, showing that the mutation R241A reduced the resistance by more than 50%7. Replacement of residue H244 by the larger residue tryptophan should displace the drug by steric occlusion, but the variant lost most of its antibiotic resistance phenotype (Fig. 1c). Similarly, MsrDWT does not provide resistance to TYL nor SPI (Table 1), despite predicted clash of both antibiotics with MsrD according to the MsrE structure (Supplementary Fig. 1d). The same observation was made with LNC, LNZ and RTP (Table 1 and Supplementary Fig. 1d). Therefore, MsrD’s action on the antibiotic is indirect, and direct steric occlusion to displace the antibiotic seems unlikely.
Ribosome-mediated transcriptional attenuation regulates msrD expression
Analysis of the mefA/msrD operon sequence indicates a stringent ribosome binding site (RBS) only 8 bp downstream of the gene mefA stop codon. This sequence was proposed to be required for the translation of the downstream putative small ORF, msrDL (Supplementary Fig. 2a). In order to test the influence of msrDL on msrD expression, a fluorescent reporter with the sequence spanning from the first nucleotide downstream mefA stop codon to msrD first three codons fused to the Yellow Fluorescent Protein gene (YFP) was cloned in the low copy pMMB plasmid44 under the control of an IPTG-inducible PLlacO-1 promoter45. The resulting plasmid, pMMB-msrDL-msrD(1-3):yfp, was used to directly follow the fluorescence signal reflecting the yfp expression (Fig. 2a and Supplementary Fig. 2b).
a ERY-dependent induction of msrD(1-3):yfp. Fluorescent reporters shown on schematics have been introduced in E. coli DB10 and grown in presence of 1 mM IPTG and increasing sublethal ERY concentrations during 17 h. Fluorescence has been plotted against OD600, error bars for both axes represent mean ± s.d. for triplicate experiments. b Northern analysis of msrD transcript. RNAs were extracted before adding or not 1 mM IPTG (T0’), 10 min after adding IPTG (T10’), and 5, 15, 30, 60, 120 min after adding or not 100 nM ERY. Location of probe in msrD(1-3):yfp 3’ UTR is shown on the schematic. The presence of a second band was also observed, but we hypothesized that it was an abortive transcript resulting from the construct insofar as its presence correlated with induction by IPTG and ERY. Note the presence of leaky transcription in absence of IPTG, that is slightly amplified in presence of ERY. c Deletion of the intrinsic terminator between msrDL and msrD(1-3):yfp lead to a constitutive induction in absence of ERY. Error bars for both axes represent mean ± s.d. for triplicate experiments. Source data are provided as a Source Data file.
A basal fluorescence level was found to be stable in the E. coli DB10 strain containing pMMB-Control independently of ERY concentration (Fig. 2a), while a fluorescence increase that correlated with the increase in ERY concentration was observed in cells containing the plasmid pMMB-msrDL-msrD(1-3):yfp. This indicated an ERY-dependent induction of msrD. Optimal induction was found at 100 nM ERY, however, 6.25 nM was sufficient to significantly induce msrD(1-3):yfp expression, demonstrating high sensitivity of the system toward its inducer. A reduced, but significant fluorescence signal was detected at 0 nM ERY indicating an imperfect repression of the regulation. Inactivation of the ORF msrDL by replacing the start codon with an amber stop codon (pMMB-msrDL(no_ORF)-msrD(1-3):yfp) resulted in the elimination of the fluorescence induction (Fig. 2a). This demonstrated that msrDL translation is an important and necessary cis-acting feature that regulates msrD(1-3):yfp expression as previously reported for ermC46.
In bacteria, antibiotic-dependent gene induction relies mainly on translational or transcriptional attenuation47. To determine if msrDL prompts a transcriptional or translational attenuation, northern blot analysis with a probe located in msrD(1-3):yfp 3’ UTR was performed on total RNA extracted from bacteria with or without ERY and/or IPTG exposure. A 1.2-kb band corresponding to the msrDL-msrD(1-3):yfp mRNA was detected in the presence of IPTG and started to increase incrementally 5 min after exposure to 100 nM ERY (Fig. 2b). Consistent with this ERY-dependent transcript accumulation, we concluded that msrD regulation by msrDL occurs at the transcription level.
Transcription attenuation employs premature transcription termination via: (i) Rho-dependent terminator (RDT) by binding of Rho factor, a RecA-type ATPase, to a C-rich and G-poor sequence known as Rho utilization site48; (ii) Rho-independent terminator (RIT) where a stable GC-rich stem-loop followed by a poly-uridine sequence causes RNAP to drop off49. Treatment with bicyclomycin, a Rho inhibitor, did not result in constitutive expression of the reporter, thus excluding the possibility of a RDT attenuation (Supplementary Fig. 2c). Search for secondary structures in the intergenic region between the genes msrDL and msrD using ARNold server50 revealed the presence of a RIT with a ∆G of −7.9 kcal.mol−1. We also identified a consensus “TTNTTT” NusG-dependent RNAP pausing site embedded in the msrDL sequence (Fig. 2c and Supplementary Fig. 2a), as previously described for vmlR51,52. In order to test possible transcription termination activity, we generated a msrDL-msrD construct that lacks the putative terminator (referred as pMMB-msrDL(no_term)-msrD(1-3):yfp). The deletion of the terminator resulted in a constitutive expression of the fluorescent reporter in the absence of ERY, indicating that the terminator is a necessary cis-acting regulatory feature that terminates transcription (Fig. 2c). These results validate the presence of a second transcriptional attenuator on the mefA/msrD operon that regulates exclusively the gene msrD.
Selective drug-sensing by MsrDL prevents its translation elongation and termination
To determine how this regulation occurs via a ribosome-mediated mechanism, translation of msrDL mRNA with the PURE system53 was subjected to various antibiotics and its propensity to form SRC was analyzed by toeprinting54 (Fig. 3 and Supplementary Fig. 3). A strong toeprint signal was observed in the presence of ERY and AZI, while it was absent in the presence of TEL, TYL and SPI (Fig. 3a). This in vitro result correlates to the in vivo observations made with the msrD(1-3):yfp reporter gene which was also only induced by ERY and AZI, but not by TEL, TYL, SPI and non-macrolide antibiotics (Fig. 3b).
a Toeprinting assay of msrDL in the absence (-) or in the presence of 50 µM of various macrolide antibiotics. White arrow indicates initiation codon. Black arrows indicate ribosome stalling, with M6 codon in the P site. Chemical structure of antibiotics is shown, C3 cladinose sugar of ERY and AZI being highlighted. See also Supplementary Fig. 3. b In vivo induction of msrD(1:3):yfp by various PTC/NPET targeting antibiotics. Bacteria containing pMMB-msrDL-msrD(1:3):yfp were grown for 17 h in the presence of 1 mM IPTG, in the absence (gray histograms) or in the presence of 50 nM antibiotics (red histograms). c Toeprinting assay depleted of release factors was performed in the absence of ribosome (line 1), in the presence of 50 µM RTP to assess start codon (line 2), without or with 50 µM ERY (line 3 and 4), without or with 50 µM ERY then supplemented with 100 µM puromycin (line 5 and 6), without or with 50 µM ERY in the presence of RF1/RF2/RF3 (line 7 and 8). The schematic indicates position of toeprint signal on the synthetic mRNA, P site codon of MsrDL-SRC is underlined. See also Supplementary Fig. 3. d Effects of msrDL variants on the expression of msrD(1:3):yfp. Bacteria were grown during 17 h in the presence of 1 mM IPTG, in the absence (gray histograms) or in the presence of 100 nM ERY (red histograms). Square boxes show fold of induction. e Toeprinting assay performed on the MsrDLWT, MsrDLL3A and MsrDLI4A constructs in the absence or presence of 50 µM ERY or 100 µM puromycin. Open arrow indicates the initiation codon. Black arrows indicate ribosomal stalling, with the M6 codon in the P site. f Toeprinting assay performed on the MsrDL7A-iso construct in the absence or presence of ERY. Addition of a sense codon after M6 codon lead to translational arrest with M6 codon in the P site (black arrows) and A7 codon in the A site in the presence of ERY. A faint toeprint was also observed with A7 codon in the P site (gray arrows) and the stop codon in the A site. b, c Error bars represent mean ± s.d. for triplicate experiments and the p values were determined by unpaired two-sided t-test without adjustment. Source data are provided as a Source Data file.
We used an inhibitor of the first peptide bond formation, which allows the identification of an ORF’s start codon55, to further characterize the translation of msrDL mRNA. We showed that in presence of 50 µM RTP, a clear toeprint can be observed (Fig. 3c). This toeprint corresponds to a stalled ribosome displaying the initiation codon in its ribosomal P site, which confirms that these ribosomes can initiate msrDL translation. The toeprint observed at positions +31, +32 and +33 relative to msrDL 5’ end (Fig. 3a, c) in presence of ERY and AZI indicates SRC formation located at the termination step of MsrDL synthesis with the C-terminal methionine (M6) codon in the P site and an UAA stop codon in the A site. In accordance with this, reactions that are depleted of release factors (RF1, RF2, RF3) (Fig. 3c) showed the same toeprint results. Moreover, the addition of puromycin led to the loss of this toeprint signal, while preserving it in the presence of ERY (Fig. 3c). Puromycin is an aminonucleoside antibiotic mimicking an A site substrate tyrosyl-tRNA and causing premature chain termination. While actively translating ribosomes are sensitive to this antibiotic, stalled ones tend to be refractory15,16,20,56,57. The resistance of MsrDL-SRC to puromycin is consistent with an ERY-sensing NC that leads to PTC silencing and stalling. In consequence, the stalled ribosome on msrDL covers the first half of the hairpin of the RIT and therefore prevents its formation and transcription termination (Fig. 3c and Supplementary Fig. 3c).
The main dissimilarity between tested macrolides resides in the presence or absence of a cladinose moiety on C3 of the macrocyclic lactone ring which appeared to be critical for MsrDL-SRC formation (Fig. 3a) similarly to ErmCL20. Absence of significant toeprint in the presence of TEL (whose C3 cladinose sugar is replaced by a keto group) suggested that msrDL is translated efficiently without stalling. TYL and SPI differ from the other macrolides by the presence of disaccharide mycaminose-mycarose in place of the C5 monosaccharide desosamine. When bound to the ribosome, this extended sugar moiety protrudes into the PTC of the ribosome58. Absence of intermediate toeprint and increase of the toeprint corresponding to initiation complex in the presence of these two drugs imply that these antibiotics stabilize the initiation complex. Previous results58,59,60 have shown that they inhibit the formation of the peptide bond, our results suggest that it occurs mainly for the first peptide bond formation (Fig. 3a).
We replace individually each amino acid of MsrDL sequence by an alanine to investigate the importance of each residue and its ability to induce the msrD(1-3):yfp reporter, which indirectly reports the formation of an ERY-dependent SRC (Fig. 3d). The scan showed a strong reduction of inducibility for all the mutated residues with the exception of residue Y2 (Fig. 3d). Since MsrDL is a short hexapeptide, it is very likely that most of the residues would have certain importance in maintain the conformation that is able to sense ERY. A complete loss of inducibility was observed when the residues L3 and I4 were replaced by alanine, suggesting that these two residues are directly involved in drug-sensing. Toeprinting assays conducted on these two constructs showed that both have less MsrDL-SRC formation in the presence of ERY and the stalled complexes were not resistant to puromycin (Fig. 3e). The importance of these two residues were structurally confirmed and is detailed in the next section.
Considering that MsrDL inhibits the action of both RF1 and RF2 in the presence of ERY, we mutated UAA stop codon into RF1-specific UAG and RF2-specific UGA stop codons to investigate a putative preferential inhibition as previously shown for TnaC61. Mutation of UAA into UAG or UGA reduced the level of inducibility of the reporter gene (Supplementary Fig. 2d) demonstrating the importance of the stop codon. This observation was corroborated by our phylogenetic analysis, which revealed a large prevalence of UAA stop codon among the different msrDL variants (Supplementary fig. 2a). Most likely the UAA stop codon has been selected by evolution to maximize msrD expression in the presence of an inducer. We designed a construct of MsrDLWT reporter where the codons of msrDL were replaced by synonymous codons, as described in the Methods section. This construct, MsrDLWT-iso, showed a reduction of the yfp expression, possibly due to the lack of NusG pausing site, but the regulation is mostly preserved (Fig. 3d). Therefore, the amino acid composition of the peptide is responsible for the translational stalling. Addition of an extra alanine before the stop codon in the MsrDL7A-iso construct reduced the regulatory potential of the leader peptide (Fig. 3d). Toeprinting assay revealed that this construct has two stalling positions: one occurs with the additional A7 codon in the ribosomal A site and the second with the stop codon in the A site (Fig. 3f). Thus, MsrDL peptide seems to induce efficient ribosomal stalling by preventing elongation and termination. The MsrDL7A-iso construct showed less induction of the yfp reporter by ERY compared to MsrDLWT-iso construct (Fig. 3d) suggesting that when inhibition of elongation and termination does not occur at the same location, the regulation is less efficient.
Molecular mechanism of ribosome stalling by MsrDL
To understand how MsrDL inhibits translation termination in the presence of ERY, a synthetic mRNA containing a single msrDL copy was translated in vitro with purified E. coli DB10 ribosomes in the presence of ERY (as described in Supplementary Fig. 4a) and the resulting sample was subjected to cryo-EM. After image processing and particle images sorting two nearly identical subclasses were obtained and refined, one containing only a P-site tRNAs and one containing P- and E-site tRNAs (11% and 28.1% of total particles, respectively), as described in Methods and in Supplementary Fig. 4b. However, the ribosome (observed in the non-rotated state) and the MsrDL peptide conformations in both classes are identical, therefore they were merged together in order to improve the final resolution. The resulting initial map of the 70 S ribosome (EMD-13805) with an average resolution of 3.0 Å that was subjected to multibody refinement, generating reconstructions for the body and the head of the small subunit (EMD-13807 and EMD-13808) and the large subunit (EMD-13806), presenting average resolutions of 3.08, 3.3 and 2.97 Å, respectively (Supplementary Fig. 4b, 4c and Supplementary Table 2). The resolutions of the derived reconstructions were consistent with unambiguous assignment of ribosomal proteins side chains, rRNA nucleotides and some post-transcriptional modifications (Supplementary Fig. 4d–4f). The local resolution of the codon-anticodon allowed unambiguous identification of the P-site tRNA as an elongator MettRNAMet (Supplementary Fig. 4e), which was also confirmed by the presence of distinctive elements such as N4-acetylcytidine at position 34 (ac4C34) and N6-threonylcarbamoyladenosine at position 37 (t6A37) (Supplementary Fig. 4e, f)62,63, consistent with the toeprint results which assigned the M6 codon in the P site and the UAA stop codon in the A site (Fig. 3c).
A clear density corresponding to an ERY molecule bound in its canonical position was found at the entrance of the NPET (Fig. 4a, b and Supplementary Fig. 4d), allowing attribution of macrocyclic lactone ring, cladinose and desosamine sugars. A continuous density at the 3’ end of the P-site tRNA extending within the entrance of the ribosome tunnel was identified as MsrDL-NC and the local resolution of ~3 Å allowed modeling of the peptide de novo (Fig. 4b and Supplementary Fig. 4d). The leader peptide adopted a hook-like shape with its N-terminal extremity protruding into a cavity at the entrance of the tunnel delimited by 23 S rRNA nucleotides U2584, U2586, G2608 and C2610 (Fig. 4b). One noticeable feature of MsrDL interaction with the ribosome is its residue Y2 that forces the base U2584 to bulge out and Y2 forms a π-stacking with the base G2583 (Fig. 4c, d) in place of U2584. This unique interaction should stabilize the peptide conformation, while it appears structurally significant, it is not strictly essential, since replacement of the Y2 by an alanine reduces the induction of the reporter gene by less than 50% (Fig. 3d). In addition, this residue is substituted in some MsrDL variants (Supplementary Fig. 2a). MsrDL did not show numerous significant electrostatic interactions with the ribosome, since it is mostly composed of hydrophobic amino acids. It is likely stabilized by hydrophobic interactions.
a Transverse section of the cryo-EM map showing the 30 S (gray) and 50 S (white) ribosomal subunits, mRNA (purple), ERY (red) and MsrDL-NC (gold) bound to P-site tRNAMet (beige). b Close-up of MsrDL-NC within ribosomal tunnel showing experimental density and modeled structure, colored as Fig. 4a while 23 S rRNA nucleotides are shown in light gray. c, d Presence of MsrDL residue Y2 displace nucleotide U2584 out of its position when compared to RF1- (PDB 5J30, light blue) and RF2-containing (PDB 5CZP, blue) termination complex, thus forming a π-stacking with G258364. e Conformation of U2584 and U2585 prevented productive accommodation of RF1 (PDB 5J30, light blue) and RF2 (PDB 5CZP, blue) while catalytic methylated glutamine clashed with MsrDL residue F564. f PTC in MsrDL-SRC (light gray) is stabilized in an uninduced state (PDB 1VQ6, pink) rather than in an induced state (PDB 1VQN, green) as U2585 is pushed back by MsrDL residue Y265. g–i Molecular basis for C3 cladinose sugar recognition by MsrDL. Residue L3 at proximity of cladinose sugar while residue I4 at proximity of desosamine sugar. TEL lacking cladinose sugar and failing to form MsrDL-SRC has been aligned (PDB 4V7S)76. j Presence of residue I4 avoided rotation of A2062 to form an Hoogsteen base pairing with m2A2503 as is the case for ErmCL-SRC (PDB 3J7Z, green)12. Light blue dashed lines indicate hydrogen bonds formed by Hoogsteen base pairing. For the whole figure, structures were aligned on domain V of 23 S rRNA. Spheres represent van der Waals radii.
Our biochemical investigations showed that the MsrDL-SRC was formed due to hampered translation termination on the UAA stop codon. This stop codon is recognized by both RF1 and RF2, which subsequently catalyze peptide hydrolysis via their conserved GGQ motif protruding within the PTC. Structural alignment of our atomic model to RF1- and RF2-containing ribosome structures64 revealed that proper accommodation of the RFs is prevented by steric clashes of the methylated glutamine and the glycines of the GGQ motif against the residue F5 of MsrDL and 23 S rRNA bases U2584 and U2585, respectively (Fig. 4e). The conformation of MsrDL peptide stabilized the PTC in an uninduced conformation65 which prevents the opening of the active site necessary for peptide bond formation and hydrolysis (Fig. 4f). This observation suggests that the MsrDL-SRC cannot catalyze peptide bond formation, as confirmed by the toeprinting performed on the MsrDL7A-iso construct (Fig. 3e). Structural alignment of the MsrDL-SRC with ErmBL- and ErmCL-SRCs demonstrated that 23 S rRNA nucleotides U2506 and U2585 are not in the same conformation (Supplementary Fig. 5a, b). These bases in MsrDL-SRC adapt a conformation similar to ErmDL-, SpeFL- and TnaC(R23F)-SRCs (Supplementary Fig. 5c–e)12,13,14,15,16. However, the unusual position of base U2584 seems to be a unique feature of MsrDL-SRC.
Our in vivo and in vitro experiments demonstrated that the MsrDL monitors the presence of cladinose-containing drugs. Residues L3 and I4 being critical for drug recognition and/or conformation of the peptide (Fig. 3d, e). Accordingly, in our atomic model we identified residue L3 as the closest amino acid to the cladinose moiety and residue I4 as the closest to the desosamine moiety of the ERY molecule (Fig. 4b, g). However, similarly to ErmBL, no close contact between the nascent chain and the drug was observed13,66. The residue L3 being ~3.8 Å away from cladinose sugar and the residue I4 being ~3.9 Å away from desosamine sugar, consistent with a hydrophobic interaction between MsrDL and the drug (Fig. 4g–i). This observation is also in accordance with previous descriptions of pentapeptides conferring ERY resistance in which the presence of a leucine or isoleucine at position 3 was critical for drug recognition67,68. Moreover, the importance of 23 S rRNA residue A2062 acting as an ERY-sensor and contributing in silencing the PTC, has been demonstrated for some regulatory leader peptides69,70. In the case of ErmCL, the critical feature for macrolide-dependent stalling is the rotation of the base A2062 that forms an Hoogsteen base pairing with m2A250312,69,70. The structure of the MsrDL-SRC showed that residue I4 restrains rotation of base A2062 (Fig. 4j). Therefore, MsrDL induces an A2062/m2A2503-independent ribosome stalling70.
The path of MsrDL-NC within the ribosomal tunnel contrasts with all previous descriptions of elongating polypeptides (Figs. 4b and 5). Indeed, ERY-sensing ErmBL, ErmCL and ErmDL leader peptides12,13,14 engage in the NPET and bypass the tunnel-bound drug, while MsrDL curves before encountering it and then engages into a dead-end crevice at the entrance of the NPET (Figs. 4b and 5b). The same observation (Fig. 5c) can be made with metabolite-sensing SpeFL/TnaC(R23F) leader peptides15,16, which avoid the crevice and enter the NPET. The first methionine of MsrDL-NC latches into the dead-end crevice, which is delimited by 23 S rRNA nucleotides U2584, U2586, G2608 and C2610 (Fig. 5a–c). This crevice located at the base of helix 93 (h93) is part of domain V of 23 S rRNA and is conserved from bacteria to human (mitoribosome and cytosolic ribosome) (Fig. 5d, Supplementary Fig. 5f and 5g). We termed it “proximal crevice”.
a Secondary structure of the E. coli 23 S rRNA domain V showing location of the proximal crevice at the base of h93. For the whole figure, nucleotides delimitating proximal crevice are shown in red. b, c Comparison of MsrDL path (gold) with ERY-dependent leader peptides ErmBL (PDB 5JTE, blue), ErmCL (PDB 3J7Z, green), ErmDL (PDB 7NSO, teal) as well as L-ornithine-sensing SpeFL (PDB 6TC3, pink) and L-tryptophan-sensing TnaC(R23F) (PDB 7O1A, purple)12,13,14,15,16. See also Supplementary Fig. 5a–e. d Proximal crevice in domain V is conserved from bacteria to human (MsrDL-SRC, gray; S. aureus PDB 6YEF, light green; H. sapiens 55 S mitoribosome PDB 7A5F, light blue; H. sapiens 80 S ribosome PDB 6OLI, marine blue)93,94,95. See also Supplementary Fig. 5g and h. For the whole figure, atomic model of MsrDL is shown with its experimental density. Structures were aligned on domain V of 23 S rRNA.
MsrD negatively regulates its own expression
The protein MsrD provides antibiotic resistance by interacting with antibiotic-stalled ribosome while msrDL translation in the presence of ERY leads to the formation of a SRC. If MsrD is also able to rescue this SRC, it will repress its own expression and form a feedback loop similar to the ARE ABC-F Vga(A)71. We tested this hypothesis by using an E. coli DB10 strain double transformed with pMMB-msrDL-msrD(1-3):yfp and the compatible pBAD plasmid expressing various msrD variants. Strains were grown in the presence of 300 nM of ERY with their optical density and fluorescence monitored (Fig. 6a and Supplementary Fig. 6). The strain carrying the pBAD-Control showed a high fluorescence and a low OD600 as observed in Fig. 2a, reflecting an induction of the reporter and a susceptibility to ERY. On the contrary, expression of msrDWT led to a higher cell growth and a stronger reduction of fluorescence, which correspond respectively to antibiotic resistance and a repression of the expression of the reporter. These results support a model where MsrDWT can rescue MsrDL-SRC and therefore alleviate the RIT-mediated repression on the msrD transcription.
a Effects of MsrD variants on MsrDL. E. coli DB10 containing pMMB-msrDL-msrD(1-3):yfp and expressing various msrD mutants were grown in the presence of 0.2 % L-Arabinose, 1 mM IPTG and 300 nM ERY, both OD600 and fluorescence being recorded over 24 h. Fluorescence was plotted against OD600, error bars for both axes represent mean ± s.d. for triplicate experiments. Color code is same as Fig. 1c. Light green rectangle indicates bacterial growth over control plasmid. b Model of MsrD regulating its own expression and providing antibiotic resistance. The presence of a NusG-dependent pause site might stall the RNAP and explains why the system works in bacteria where transcription and translation are not so tightly coupled. In the absence of ERY, RNAP dropped off at Rho-independent transcription terminator. In the presence of ERY, the ribosome following the RNAP (presumably paused) stalled and unwound the terminator leading to msrD transcription. Once translated, MsrD negatively regulates its own expression on one side, and provides antibiotic resistance on the other side. Source data are provided as a Source Data file.
The strain expressing msrDE125Q had a low ERY resistance (Fig. 1c). When it was co-transformed with the pMMB-msrDL-msrD(1-3):yfp plasmid, no resistance was observed, possibly because the presence of the two plasmids created some toxicity. Overall, the results showed that MsrD variants which lost the ability to provide antibiotic resistance (∆PtIM, ∆Loop and ATPase deficient mutants) also lost the ability to rescue MsrDL-SRC (Fig. 6a and Supplementary Fig. 6). Point mutations in the Loop of the PtIM indicated some discrepancy between antibiotic resistance phenotype and MsrDL-SRC rescue. All the mutants showed an intermediary phenotype with less antibiotic resistance and less repression of the reporter compared to MsrD. The mutant MsrDH244w had the same repression effect on the reporter as MsrDH244A for an equivalent OD600, but it provided more than 3 times less antibiotic resistance (Figs. 1c and 6a). The mutant MsrDR241A provides a resistance close to MsrD, but does not rescue the MsrDL-SRC as efficiently as the latter one (Figs. 1c and 6a). Mutants MsrDL242A and MsrDH244A provided similar resistance, but MsrDH244A conferred a stronger repression of the reporter (Figs. 1c and 6a). Together these observations demonstrate that the determinants of the dual function of MsrD in conferring antibiotic resistance and rescuing MsrDL-SRC are overall similar.
Discussion
We used genetic, molecular biology and structural approaches to dissect the regulation of the streptococcal ARE ABC-F gene msrD. We characterized the msrDL regulatory ORF which induces ribosome stalling in the presence of cladinose-containing macrolides (Figs. 3, 4 and 6b). Our structural data shows that this inhibition results from the conjoint effects of: (i) MsrDL maintaining 23 S rRNA bases U2506 and U2585 in an uninduced conformation (Fig. 4f), (ii) the steric occlusion by U2584, U2585 and MsrDL residue F5 that precludes the productive accommodation of RF1 and RF2 (Figs. 3c and 4e). The MsrDL-stalled ribosome then hampers the formation of a RIT and thus allows for the transcription of msrD (Figs. 2c, 3a, 3c and 6b). The coupling between translation and transcription necessary for this mechanism in Streptococci is possibly mediated by a NusG-dependent RNAP pausing site in msrDL (Fig. 2c).
Alongside ErmBL, ErmCL and ErmDL, MsrDL constitutes the fourth ERY-sensing NC that has been structurally described, however some mechanisms of drug-sensing and PTC silencing differ. Firstly, MsrDL and ErmCL need the presence of the C3 cladinose sugar of ERY for ribosomal stalling (Fig. 3a, b), this differs to ErmBL and ErmDL which induce stalling with drugs lacking this sugar such as oleandomycin, solithromycin or telithromycin12,14,66,72. Secondly, similar to ErmBL, no close contact between the drug and MsrDL was observed (Fig. 4h, i)66. Thirdly, MsrDL is the only described leader peptide that exploits inhibition of its hydrolysis by RFs to induce stalling in the presence of antibiotic, comparable to the metabolite sensors TnaC and SpeFL15,16. ErmBL13, ErmCL12 and ErmDL14 employ only elongation inhibition in physiological condition. Finally, MsrDL-NC conformation and path within the ribosome differ from previously described ligand-sensing leader peptides that elongate within the tunnel (Fig. 5b, c)12,13,14,15,16. It adopts a hook-like shape possibly due to hydrophobic interactions with the drug (Fig. 4b, g–i), while its initiating methionine engages in a dead-end crevice at the NPET entrance (Fig. 5). This crevice named “proximal crevice” is conserved in prokaryotes and eukaryotes (Supplementary Fig. 5f and 5g) and is a functional site within the ribosome.
ARE ABC-F genes tend to be transcriptionally regulated as illustrated by msrD, vmlR, vgaA, lmo0919 and lmrC51,71,73,74,75,76. However, MsrD can regulate its own expression by creating a negative feedback-loop which is not the case for VmlR51. The gene msrD forms an operon with the efflux pump mefA (Fig. 1a), which is also regulated via ribosome-mediated transcriptional attenuation, translational stalling occurring on the mefAL33. Therefore, msrD is under the dual-regulation of two ERY-sensing leader peptides. One explanation for this redundancy may reside in the need for the bacteria to tightly regulate msrD because of its toxicity. The expression of msrDWT in E. coli DB10 led to a 20% growth defect compared to control condition (Fig. 1c), suggesting that msrDWT expression has a fitness cost for the bacteria and can be beneficial only in the presence of antibiotic. The presence of two attenuators creates a double-lock system to avoid any basal expression in the absence of inducer and the negative feedback-loop maintains a minimal amount of MsrD production. Thus, as long as MefA can export enough antibiotic to provide resistance, msrD should be kept repressed. The fitness cost of the expression of ARE ABC-F genes is a potential Achilles heel of the bacteria containing those genes and may be exploited to fight antibiotic resistance. MsrD provides resistance to ERY, AZY and TEL23 (Table 1), but only the first two induce its expression, indicating that MsrD can provide some resistance for an antibiotic which does not induce its expression. Thus, MsrDL facilitates the use of MsrD by the bacteria and maintains its evolution under its control.
Methods
Construction of plasmids
Strains and plasmids used in this study are listed in Supplementary Table 3. Oligonucleotides are listed in Supplementary Table 4. For the whole study, the considered reference sequence for mefA/msrD macrolide resistance operon was a Tn916-type transposon inserted in Streptococcus pneumoniae strain 23771 genome (Genbank accession number FR671415.1)77. Plasmids pBAD-Control and pBAD-msrDWT containing a C-terminal hexahistidine tag (originally referred as pBAD33 and pVN50) were retrieved from Olivier Chesneau at Institut Pasteur (Paris, France)24. Both mefA/msrD macrolide resistance operon and pBAD-msrDWT encode the same protein MsrD (Uniprot accession number A0A496M710). pBAD plasmids allow a tight and stringent expression control via catabolite repression, inserted gene being repressed in presence of 0.4% (w/v) Glucose and induced in the presence of 0.2% (w/v) L-Arabinose. Transformed clones were selected with 20 µg.ml−1 chloramphenicol. In this study, all msrD mutants contain a C-terminal hexahistidine tag. Catalytic mutant msrDEQ2 was generated by amplifying pBAD-msrDWT with primer pairs 5/6 and 7/8, both fragments being assembled with NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). Catalytic mutants msrDE125Q and msrDE434Q were generated via quickchange mutagenesis with primers pairs 9/10 and 11/12. Mutants msrD∆Loop and msrD∆PtIM were generated based on phylogenetic alignments by deleting residues between K216/K254 and E189/A274 respectively, a three-glycine linker being added to allow flexibility, via quickchange mutagenesis using primer pairs 13/14 and 15/16. Variants msrDR241A, msrDL242A, msrDH244A and msrDH244W were generated via fusion PCR by amplifying msrDWT with primers pairs 1/18, 1/20, 1/22, 1/24 and 2/17, 2/19, 2/21, 2/23 respectively. Backbone was amplified with primer pair 3/4 and fragments were assembled with NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). All pMMB constructs originated from a low copy IPTG-inducible plasmid pMMB67EH-yfp, consisting of the original pMMB67EH44 containing optimized venus-yfp under the control of Ptac promoter (manuscript in preparation). Transformed clones were selected with 100 µg.ml−1 ampicillin. To test msrDL functions in vivo, several fluorescent reporter genes have been designed. First, pMMB67EH-yfp native Ptac promoter was replaced by PLlacO-1 promoter as previously described45:
5’- ATAAATGTGAGCGGATAACATTGACATTGTGAGCGGATAACAAGATACTGAGCACA −3’ (lac operators, shaded gray; transcription start, bold). This IPTG-inducible promoter allows a tight and stringent transcription regulation compared to Ptac promoter, insofar as no regulatory element is found in the 5’ UTR. Promoter was replaced by amplifying pMMB67EH-yfp with primer pairs 25/26, PCR fragment being then re-circularized via NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). The resulting plasmid was named pMMBpLlacO-1-67EH-yfp. Plasmid pMMB-msrDL-msrD(1−3):yfp has been designed by introducing the sequence spanning from the first nucleotide downstream mefA stop codon to msrD three first codons fused to yfp. The introduced sequence is as follows: 5’-ACAATATTGGAGGAATATTTATGTATCTTATTTTCATGTAACTCTTCCTGCTAAAATCGCAGGGTTTTCCCTGCATACAAGCAAATGAAAGCATGCGATTATAGACAGGAGGAAATGTTATGGAATTA−3’ (RBS, bold; msrDL, italic, msrD three first codons are underlined). To clone such construct, yfp gene was amplified from pMMB67EH-yfp with primer pairs 27/29 then with 28/29 to generate the insert, backbone was amplified with primer pair 30/31 using pMMBpLlacO-1-67EH-yfp as matrix, both fragments being then assembled with NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). Plasmid pMMB-control was built by removing the msrDL-msrD(1−3):yfp cassette via quickchange mutagenesis using primer pair 32/33, the resulting construction containing only the PLlacO-1 promoter followed by the plasmid endogenous rrnB transcription terminator. Plasmid pMMB-msrDL(no_term)-msrD(1-3):yfp (where the RIT between msrDL and msrD is deleted) was generated by amplifying pMMB-msrDL-msrD(1−3):yfp with primer pair 34/35, PCR fragment being then re-circularized via NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). The various msrDL mutants were cloned via quickchange mutagenesis by amplifying pMMB-msrDL-msrD(1−3):yfp with primers 36 to 51. Plasmid pMMB-msrDL(MYLIFMA-isocodons)-msrD(1−3):yfp was generated by replacing msrDLWT sequence (ATGTATCTTATTTTCATGTAA) by a recoded sequence (ATGTACCTGATCTTCATGGCCTAA) using isocodons and introducing an extra alanine codon (7 A codon) before stop codon. Sequence was recoded using isocodons because mutating the WT sequence introduced a new promoter. To do so, pMMB-msrDL-msrD(1−3):yfp was amplified with primer pair 52/53, PCR fragment being then re-circularized via NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs). Recoded sequence without the 7 A codon (ATGTACCTGATCTTCATGTAA) was generated via quickchange mutagenesis by amplifying pMMB-msrDL(MYLIFMA-isocodons)-msrD(1−3):yfp with primer pair 54/55, leading to plasmid pMMB-msrDL(WT-isocodons)-msrD(1−3):yfp.
Antibiotic susceptibility testing, MIC and IC50 determination
A saturated preculture of E. coli DB10 transformed with pBAD plasmid was grown overnight at 37 °C under vigorous shaking in Luria-Bertani Miller broth (LB), 20 µg.ml−1 chloramphenicol and supplemented with 0.4% (w/v) Glucose. Antibiotic susceptibility testing assay was performed in Mueller-Hinton broth (MH, Sigma Aldrich), antibiotics being diluted via serial dilutions. A 96-wells flat-bottom plate (Flacon) was filled with a final volume per well of 200 µl, containing 20 µg.ml−1 chloramphenicol and 0.2% (w/v) L-Arabinose and antibiotic to test. Wells were inoculated at OD600 ~ 0.03-0.04 prior to addition of 60 µl mineral oil (Sigma Aldrich) avoiding evaporation but not oxygen diffusion. Plates were therefore incubated for 24 h in CLARIOstar Plus plate reader (BMG Labtech) at 37 °C with 600 rpm double-orbital shaking, OD600 being measured each 30 min. Optical densities at 24 h were then normalized relative to optical densities of E. coli DB10 pBAD-Control grown in the absence of antibiotic in Prism 7(GraphPad). MIC was determined as absence of growth compared to blank, IC50 was calculated in Prism using equation Y=Bottom + (Top-Bottom)/(1 + ((X^HillSlope)/(IC50^HillSlope))) and the standard deviation was calculated by multiplying standard error by square root of n (n being at least 3 replicates). To generate curves, for bacteria grown in absence of antibiotic (0 µM), since coordinates are plotted as logarithms and since log(0) is undefined, this point has been approximated 2 log units below the lowest tested value (i.e. 0.0625 µM) consistently with Prism user guide78. Curve fitting was performed with non-linear fitting function “log(inhibitor) vs. response -- Variable slope” using Y=Bottom + (Top-Bottom)/(1 + 10^((LogIC50-X)*HillSlope)). Both equations gave strictly the same result for IC50.
In vivo induction assay
A saturated preculture of E. coli DB10 transformed with pMMB plasmid was grown overnight at 37 °C under vigorous shaking in LB with 100 µg.ml−1 ampicillin. For bacteria double-transformed with pMMB and pBAD plasmids, media was also supplemented with 20 µg.ml−1 chloramphenicol and supplemented with 0.4% (w/v) Glucose. In vivo induction assay was performed in Mueller-Hinton broth (MH, Sigma Aldrich). A 96-wells flat-bottom plate (Flacon) was filled with a final volume per well of 200 µl, containing 100 µg.ml−1 ampicillin, 1 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG). Growth medium was supplemented with 20 µg.ml−1 chloramphenicol and 0.2 % (w/v) L-Arabinose if pBAD plasmid is present. Induction of fluorescent reporters being antibiotic-dependent, growth media were supplemented with required antibiotic accordingly. Plates were therefore incubated in CLARIOstar Plus plate reader (BMG labtech) at 37 °C with 600 rpm double-orbital shaking up to 24 h, OD600 and fluorescence (excitation: 497-15 nm, emission: 540-20 nm, gain 1600) being measured each 30 min.
Immunoblotting
Samples were loaded on 12.5 % acrylamide SDS-PAGE gels. They were resolved by migrating at 0.04 A and then applied on Immun-Blot PVDF membrane (Bio-Rad) that had been activated in absolute ethanol then washed in electro-transfer buffer (25 mM Tris base, 192 mM glycine, 0.1% (w/v) SDS, 10% (v/v) absolute ethanol). Proteins were electro-transferred at 100 V during 1 h. Membranes were blocked by incubating 1 h in 1X PBS supplemented with 0.5% nonfat dry milk then washed with 1X PBS. C-term His-tagged MsrD variants were detected using anti-6xHis-tag primary antibody (Covalab, 1:2 000 dilution in 1X PBS, 0.1% Tween-20) combined with anti-mouse-HRP secondary antibody (Covalab, 1:20 000 dilution in 1X PBS, 0.1% Tween-20). Immunoblots were revealed by performing an ECL detection using Clarity Western ECL Substrate (Bio-Rad) and imaged with Li-Cor Odyssey FC imaging system (Li-Cor).
Preparation of E. coli DB10 ribosomes
An overnight saturated E. coli DB10 preculture was used to inoculate growth medium (LB, dilution 1:100) and cells were grown at 37 °C and 180 rpm. When OD600 = 0.5, cells were washed and harvested by two centrifugations during 20 min, 5 000 xg at 4 °C followed by resuspensions in Buffer A (20 mM Tris pH 7.4, 10 mM Mg(OAc), 100 mM NH4(OAc), 0.5 mM EDTA). Pellets were resuspended in Buffer A supplemented with 6 mM β-mercaptoethanol, 25 µg.ml−1 lysozyme, 0.001% (v/v) protease inhibitor cocktail (Sigma Aldrich) and lysed three times at 2.5 kBar using a cell disrupter (Constant Systems Limited). Lysate was clarified by two centrifugations during 15 min, 22 000 xg at 4 °C then spun for 20 h, 34 700 rpm at 4 °C in a Type 70 Ti rotor (Beckmann Coulter) through a 37.7% (w/v) sucrose cushion in Buffer B (20 mM Tris pH 7.4, 10 mM Mg(OAc), 500 mM NH4(OAc), 0.5 mM EDTA, 6 mM β-mercaptoethanol). Sucrose cushions were decanted and ribosome pellets resuspended in Buffer C (20 mM Tris pH 7.4, 7.5 mM Mg(OAc), 60 mM NH4(OAc), 0.5 mM EDTA, 6 mM β-mercaptoethanol). Finally, 12 mg of ribosomes were loaded onto 10-40 % (w/v) sucrose gradients in Buffer C and centrifuged 14 h at 20 000 rpm, 4 °C in SW28 rotor (Beckman Coulter). Gradients were fractionated using Biocomp Piston Gradient Fractionator (BioComp Instruments), absorbance being measured at 254 nm. Fractions containing 70 S ribosomes were pooled, washed and concentrated in Amicon 50k (Merckmillipore) using Buffer C. Ribosomes concentration was adjusted to 18-20 µM, then aliquoted and flash frozen in liquid nitrogen.
RNA extraction and northern blotting
Total RNA extraction was realized using the RNAsnap method as previously described79. In brief, a preculture of E. coli DB10 containing pMMB-msrDL-msrD(1−3):yfp was grown overnight at 37 °C under vigorous shaking in LB supplemented with 100 µg.ml−1 ampicillin. Growth medium (MH, 100 µg.ml−1 ampicillin) was inoculated at OD600 = 0.05, and cells were grown at 37 °C and 180 rpm. When OD600 = 0.5, cells were treated (or not) with 1 mM IPTG during 10 min, then treated (or not) with 100 nM erythromycin. Samples of 2 ml were collected (T0’ = before IPTG was added, T10’ = 10 min after IPTG was added, and 5 min, 15 min, 30 min, 60 min and 120 min after erythromycin was added), then spun during 1 min at 15 000 rpm. Growth media was removed and cells were resuspended in 100 µl RNAsnap buffer (95% (v/v) formamide, 18 mM EDTA, 0.025% (v/v) SDS, 1% (v/v) β-mercaptoethanol). Samples were heated for 10 min at 95 °C then clarified by centrifuging 10 min, at 16 000 rpm at 16 °C. For northern blotting analysis, 6 µg of total RNAs were resolved on a 1% (w/v) agarose gel then transferred onto Amersham Hybond-N+ Membrane (GE Healthcare) by capillary transfer. Radioactive probe was prepared using 40 pmol of primer 59 and 5’-labeled with 10 U of T4 Polynucleotide Kinase (New England Biolabs) and [γ32P]ATP (150 µCi). Probe was hybridized overnight at 42 °C using ULTRAhyb-Oligo hybridization buffer (Thermo Fisher Scientific). Membrane was washed three times at 42 °C during 15 min (once in 2x SSC + 0.1% (v/v) SDS, once in 1x SSC + 0.1% (v/v) SDS and finally in 0.1x SSC + 0.1% (v/v) SDS). Radioactive signal was visualized by exposing 4 h a Storage Phosphor Screen BAS-IP MS 2040 (Fujifilm) then imaged with a Typhoon FLA 9500 (GE Healthcare).
In vitro transcription
In vitro transcription was carried out in T7 RiboMAX Large Scale RNA Production System kit (Promega) according to manufacturer instructions. Briefly, to generate DNA matrix (Supplementary Table 5), T7 promoter (5’-GCGAATTAATACGACTCACTATAGGG-3’) was added by PCR using primers pair 56/57 (Supplementary Table 4) and pMMB plasmids as templates. Transcription reactions were incubated 4 h at 37 °C and transcripts were purified with TRIzol Reagent (Thermo Fisher Scientific) and Direct-zol RNA Miniprep kit (Zymo Research), samples being eluted in THE Ambion RNA Storage Solution (Thermo Fisher Scientific). Final concentration was adjusted to 5 pmol.µl−1.
Toeprinting assay
Position of stalled ribosomes on mRNA was determined by toeprinting assay, slightly adapted from previously described methods15,80. In vitro translation reactions were performed using PURExpress ∆RF123 kit and PURExpress ∆Ribosome (New England Biolabs) according to manufacturer instructions. Briefly, prior to in vitro translation reactions, 0.5 µl of 500 µM ligand (e.g. erythromycin etc) was dried in a micro-centrifuge tube, using a SpeedVac vacuum concentrator (Thermo Fisher Scientific), final concentration being 50 µM once resuspended in a 5 µl reaction. Reactions were incubated during 15 min at 37 °C using 5 pmol of RNA templates (generated as described above) and 3.3 µM of purified E. coli DB10 ribosomes. When needed, samples were treated with 100 µM final puromycin and incubated 3 min more. Immediately, 2 pmol of CY5-labeled primer 58 complementary to NV1 sequence20 was added and incubated for 5 min at 37 °C. For each reaction, reverse transcription was performed with 0.5 µl (corresponding to 5 U) of Avian Myeloblastosis Virus Reverse Transcriptase (Promega), 0.1 µl dNTP mix (10 mM), 0.4 µl Pure System Buffer (5 mM K-phosphate pH 7.3, 9 mM Mg(OAc), 95 mM K-glutamate, 5 mM NH4Cl, 0.5 mM CaCl2, 1 mM spermidine, 8 mM putrescine, 1 mM DTT) and incubated 20 min at 37 °C. Reactions were quenched with 1 µl NaOH 5 M and incubated 15 min at 37 °C. Alkali were therefore neutralized by adding 0.7 µl HCl 25% immediately supplemented with 20 µl toe-printing resuspension buffer (300 mM Na(OAc) pH 5.5, 5 mM EDTA, 0.5% (v/v) SDS). Finally, cDNAs were purified using QIAquick Nucleotide Removal kit (Qiagen), vacuum dried and resuspended in 6 µl formamide dye (95% (v/v) formamide, 20 mM EDTA, 0.25% (w/v) bromophenol blue). Sanger sequencing was performed on DNA matrix used for in vitro transcription. Briefly, each 20 µl reaction contained 7.5 nM DNA matrix, 75 nM CY5-labeled primer 58, 40 µM of each dNTPs, 0.025 U Taq Pol (New England Biolabds), 1x Thermo Pol Buffer (New England Biolabs) and corresponding ddNTPs (625 µM ddCTP/ddTTP/ddATP or 50 µM ddGTP). After PCR, 20 µl formamide dye was added to each sequencing reaction. Samples were denatured 3 min at 80 °C, then 5 µl of sequencing reaction and 1.5 µl of toe-printing reaction were loaded on a 6% acrylamide/bis-acrylamide (19:1) sequencing gel containing 8 M urea. Gel was resolved by migrating 90 min at 50 W, then imaged using a Typhoon FLA 9500 (GE Healthcare Life Sciences) using CY5 mode, LPR Ch.2 filter and 635 nm laser.
Preparation of MsrDL-SRC
A DNA matrix (Supplementary Table 5) was prepared with CloneAmp HiFi PCR Premix (Takara) using primer pair 56/60 (Supplementary Table 4) and plasmid pMMB-msrDL-msrD(1−3):yfp as matrix. The corresponding mRNA was generated as described above. The complete sequence of MsrDL-SRC mRNA is:
5’- GGAGCGGAUAACAAGAUACUGAGCACAACAAUAUUGGAGGAAUAUUUAUGUAUCUUAUUUUCAUGUAACUCUUCCUGCUAAAAUCGCAGGGUUUUCCCUGC −3’ (RBS, bold; msrDL, shaded gray; the ATG codon in the P site of stalled ribosomes is underlined). Prior to in vitro translation reaction, 1 µl of 500 µM erythromycin (50 µM final concentration once resuspended in 10 µl) has been dried in a micro-centrifuge tube, using a SpeedVac vacuum concentrator (Thermo Fisher Scientific). Dried erythromycin has been resuspended in a 10 µl in vitro translation reaction carried out in PUREfrex 2.0 kit (Genefrontier), containing 1.83 µM of purified E. coli DB10 ribosomes and 3.6 µM mRNA (molar ratio 1:2). The reaction has been incubated for 10 minutes at 37 °C, before adding 100 µM final puromycin, and incubated 3 min more. Reaction was therefore diluted in ribosome purification Buffer C (20 mM Tris pH 7.4, 7.5 mM Mg(OAc), 60 mM NH4(OAc), 0.5 mM EDTA, 6 mM β-mercaptoethanol) to reach a concentration of 150 nM ribosomes. Cryo-EM grids were immediately prepared.
Cryo-EM grids preparation
Safematic ccu-010 HV carbon coater was used to coat Quantifoil carbon grids (QF-R2/2-Cu) with a thin carbon layer of approximate thickness of 2 nm. Grids were then glow discharged for 20 sec at 3 mA. Then, 4 µl of in vitro translation reaction diluted to 150 nM were applied, and after a 2 sec blotting (force 5) and 30 sec waiting time, grids were vitrified in liquid ethane using a Vitrobot Mark IV (FEI) set to 4 °C and 100 % humidity.
Image acquisition and processing
Data collection was performed on a Talos Arctica instrument (FEI Company) at 200 kV using the SerialEM software for automated data acquisition. Data were collected at a nominal defocus of −0.5 to −2.7 µm at a magnification of 120,000 X yielding a calibrated pixel size of 1.2 Å. Micrographs were recorded as movie stack on a K2 Summit direct electron detector (Gatan), each movie stack were fractionated into 65 frames for a total exposure of 6.5 sec corresponding to an electron dose of 64 e-/Å2. MotionCor281 was used for dose weighting, drift and whole-frame motion correction. A dose weighted average image of the whole stack was used to determine the contrast transfer function with the software Gctf82. Particles were picked using a Laplacian of gaussian function (min diameter 260 Å, max diameter 320 Å). A total of ~254k particles were extracted from a subset of motion-corrected images (1523 micrographs) presenting a resolution equal or better than 4 Å with a box size of 360 pixels. The particles were binned three folds for 2D and subsequent 3D classification. After 2 rounds of 2D classification in RELION 383, ~158k particles were selected and submitted to Relion 3D classification83. A class of non-rotated 70 S depicting high-resolution features and bearing a tRNA with various occupancies was selected, representing ~104k particles. This class was further 3D-classified with a spherical mask engulfing the tRNAs binding sites (Supplementary Fig. 4b) into 4 classes, thus yielding to P- and E-tRNAs, P-tRNAs only, A- and P-tRNAs and E-tRNAs 70 S reconstructions. Only P- and E-tRNAs and P-tRNAs only reconstructions were used for further processing, which represents ~62k particles displaying the same global conformation of the 70 S. These particles where re-extracted at the full pixel size (1.2 Å) and refined through the 3D auto-refinement performed in RELION 383 resulting in a 3 Å reconstruction (Supplementary Fig. 4c), after CTF-refinement, Bayesian particle polishing and post-processing in RELION 3. In spite of the sufficient local resolution at the vicinity of MsrDL peptide at the PTC ( ~ 3 Å), residual movements of the 30 S around the 50 S can be deduced by the lower local resolutions of the head and the body of the 30 S. In order to improve their resolution so to derive a complete atomic model of the entire 70 S complex, we applied RELION 3 multi-body refinement by defining three bodies; 50 S, 30S-head and 30S-body. After completion of the refinement, the reconstructions of the 30S-head, 30S-body and 50 S reached average resolutions of 3.3, 3.08 and 2.97 Å, respectively. The local resolution (Supplementary Fig. 4b, d) was estimated using ResMap84.
Model building and refinement
The atomic model of erythromycin-stalled Escherichia coli 70 S ribosome with streptococcal MsrDL nascent chain was built into cryo-EM maps using Coot and Phenix85,86. Insofar as E. coli DB10 hasn’t been sequenced yet, we assumed that ribosomal proteins and rRNAs were strictly identical to E. coli K12. Furthermore, we did not notice any significant features in the map. Structure of SpeFL-SRC in response to L-ornithine was used as initial model (PDB 6TC3)15 and has been fitted in 70 S ribosome map EMD-13805. Then, each part of the ribosome (50 S, 30 S Body and 30 S Head) has been individually inspected in the corresponding map (respectively EMD-13806, EMD-13807 and EMD-13808) modified if necessary in Coot. P-site elongator tRNA MettRNAMet was modeled de novo based on E. coli K12 MG1655 metT gene, posttranscriptional modifications were added consistently with Modomics database87 (http://genesilico.pl/modomics/). E-site tRNA PhetRNAPhe was derived from crystal structure of phenylalanine tRNA from E. coli (PDB 6Y3G) with minor adjustments88. ERY, MsrDL leader peptide and mRNA were modeled de novo in Coot. Final model was refined in map EMD-13805 using Phenix86.
Figure preparation
Growth curves and histograms were generated using GraphPad Prism 7 (GraphPad). Sequence alignments were visualized with JalView89. Western blotting, northern blotting and toe-printing gels were analyzed using Fiji90. Figures depicting molecular structures or electronic density maps were prepared using PyMOL Molecular Graphics System, Chimera and ChimeraX91,92.
Statistics and reproducibility
Statistical details can be found in the figure legends. Statistical significance was assessed using unpaired two-tailed t-test function in Prism 7 (GraphPad) without adjustments. All the experiments presented have be reproduced at least twice with the same results. This includes all the Toeprinting experiments (Fig. 3a, c and e and supplementary Fig. 3d) and the Northern Blot (Fig. 2b).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The data that support this study are available from the corresponding authors upon reasonable request. Cryo-EM map of erythromycin-stalled Escherichia coli 70 S ribosome with streptococcal MsrDL nascent chain has been deposited at the Electron Microscopy Data Bank (EMDB) with accession code EMD-13805, as well as 50 S, 30 S Body and 30 S Head maps obtained after multibody refinement with accession code EMD-13806, EMD-13807, and EMD-13808, respectively. Corresponding atomic model has been deposited in the Protein Data Bank (PDB) with accession code 7Q4K. The msrD sequence is accessible in the Genebank database under No. FR671415. All the other sequences used for the sequence alignment in Supplementary Figs. 1 and 2 were extracted from the Genebank database, the reference Nos. are indicated in Supplementary Figure 2. Source data are provided with this paper.
References
Crowe-McAuliffe, C. et al. Structural basis for antibiotic resistance mediated by the Bacillus subtilis ABCF ATPase VmlR. PNAS 115, 8978–8983 (2018).
Crowe-McAuliffe, C. et al. Structural basis of ABCF-mediated resistance to pleuromutilin, lincosamide, and streptogramin A antibiotics in Gram-positive pathogens. Nat. Commun. 12, 3577 (2021).
Crowe-McAuliffe, C. et al. Structural basis for PoxtA-mediated resistance to phenicol and oxazolidinone antibiotics. Nat. Commun. 13, 1860 (2022).
Lenart, J., Vimberg, V., Vesela, L., Janata, J. & Balikova Novotna, G. Detailed mutational analysis of Vga(A) interdomain linker: implication for antibiotic resistance specificity and mechanism. Antimicrob. Agents Chemother. 59, 1360–1364 (2015).
Mohamad, M. et al. Sal-type ABC-F proteins: intrinsic and common mediators of pleuromutilin resistance by target protection in staphylococci. Nucleic Acids Res. 50, 2128–2142 (2022).
Sharkey, L. K. R., Edwards, T. A. & O’Neill, A. J. ABC-F proteins mediate antibiotic resistance through ribosomal protection. mBio 7, e01975 (2016).
Su, W. et al. Ribosome protection by antibiotic resistance ATP-binding cassette protein. PNAS 115, 5157–5162 (2018).
Boël, G. et al. The ABC-F protein EttA gates ribosome entry into the translation elongation cycle. Nat. Struct. Mol. Biol. 21, 143–151 (2014).
Chen, B. et al. EttA regulates translation by binding the ribosomal E site and restricting ribosome-tRNA dynamics. Nat. Struct. Mol. Biol. 21, 152–159 (2014).
Fostier, C. R. et al. ABC-F translation factors: from antibiotic resistance to immune response. FEBS Lett. 595, 675–706 (2021).
Ousalem, F., Singh, S., Chesneau, O., Hunt, J. F. & Boël, G. ABC-F proteins in mRNA translation and antibiotic resistance. Res. Microbiol. 170, 435–447 (2019).
Arenz, S. et al. Drug sensing by the ribosome induces translational arrest via active site perturbation. Mol. Cell 56, 446–452 (2014).
Arenz, S. et al. A combined cryo-EM and molecular dynamics approach reveals the mechanism of ErmBL-mediated translation arrest. Nat. Commun. 7, 12026 (2016).
Beckert, B. et al. Structural and mechanistic basis for translation inhibition by macrolide and ketolide antibiotics. Nat. Commun. 12, 4466 (2021).
Herrero del Valle, A. et al. Ornithine capture by a translating ribosome controls bacterial polyamine synthesis. Nat. Microbiol. 5, 554–561 (2020).
van der Stel, A.-X. et al. Structural basis for the tryptophan sensitivity of TnaC-mediated ribosome stalling. Nat. Commun. 12, 5340 (2021).
Horinouchi, S. & Weisblum, B. Posttranscriptional modification of mRNA conformation: Mechanism that regulates erythromycin-induced resistance. PNAS 77, 7079–7083 (1980).
Ito, K. & Chiba, S. Arrest peptides: Cis-acting modulators of translation. Annu. Rev. Biochem. 82, 171–202 (2013).
Narayanan, C. S. & Dubnau, D. Evidence for the translational attenuation model: ribosome-binding studies and structural analysis with an in vitro run-off transcript of ermC. Nucleic Acids Res. 13, 7307–7326 (1985).
Vazquez-Laslop, N., Thum, C. & Mankin, A. S. Molecular mechanism of drug-dependent ribosome stalling. Mol. Cell 30, 190–202 (2008).
Gay, K. & Stephens, D. S. Structure and dissemination of a chromosomal insertion element encoding macrolide efflux in streptococcus pneumoniae. J. Infect. Dis. 184, 56–65 (2001).
Roberts, M. C. et al. Nomenclature for macrolide and macrolide-lincosamide-streptogramin B resistance determinants. Antimicrob. Agents Chemother. 43, 2823–2830 (1999).
Daly, M. M., Doktor, S., Flamm, R. & Shortridge, D. Characterization and Prevalence of MefA, MefE, and the Associated msr(D) Gene in Streptococcus pneumoniae Clinical Isolates. J. Clin. Microbiol 42, 3570–3574 (2004).
Nunez-Samudio, V. & Chesneau, O. Functional interplay between the ATP binding cassette Msr(D) protein and the membrane facilitator superfamily Mef(E) transporter for macrolide resistance in Escherichia coli. Res. Microbiol. 164, 226–235 (2013).
Tatsuno, I. et al. Functional Predominance of msr(D), Which Is More Effective as mef(A)-Associated Than mef(E)-Associated, Over mef(A)/mef(E) in Macrolide Resistance in Streptococcus pyogenes. Microb. Drug Resistance 24, 1089–1097 (2018).
Zhang, Y. et al. Predominant role of msr(D) over mef(A) in macrolide resistance in Streptococcus pyogenes. Microbiol. (Read.) 162, 46–52 (2016).
Del Grosso, M., Camilli, R., Iannelli, F., Pozzi, G. & Pantosti, A. The mef(E)-Carrying Genetic Element (mega) of Streptococcus pneumoniae: Insertion Sites and Association with Other Genetic Elements. Antimicrob. Agents Chemother. 50, 3361–3366 (2006).
Del Grosso, M., Scotto d’Abusco, A., Iannelli, F., Pozzi, G. & Pantosti, A. Tn2009, a Tn916-Like Element Containing mef(E) in Streptococcus pneumoniae. Antimicrob. Agents Chemother. 48, 2037–2042 (2004).
Chancey, S. T. et al. Composite mobile genetic elements disseminating macrolide resistance in Streptococcus pneumoniae. Front. Microbiol. 6, 26 (2015).
Iannelli, F. et al. Nucleotide sequence of conjugative prophage Φ1207.3 (formerly Tn1207.3) carrying the mef(A)/msr(D) genes for efflux resistance to macrolides in Streptococcus pyogenes. Front Microbiol 5, 687 (2014).
González-Zorn, B. et al. Genetic basis for dissemination of armA. J. Antimicrobial Chemother. 56, 583–585 (2005).
Ambrose, K. D., Nisbet, R. & Stephens, D. S. Macrolide efflux in streptococcus pneumoniae is mediated by a dual efflux pump (mel and mef) and is erythromycin inducible. Antimicrob. Agents Chemother. 49, 4203–4209 (2005).
Chancey, S. T. et al. Transcriptional attenuation controls macrolide inducible efflux and resistance in streptococcus pneumoniae and in other gram-positive bacteria containing mef/mel(msr(D)) Elements. PLOS ONE 10, e0116254 (2015).
Ramu, H., Mankin, A. & Vazquez-Laslop, N. Programmed drug-dependent ribosome stalling. Mol. Microbiol 71, 811–824 (2009).
Sothiselvam, S. et al. Macrolide antibiotics allosterically predispose the ribosome for translation arrest. Proc. Natl. Acad. Sci. USA 111, 9804–9809 (2014).
de Block, T. et al. WGS of commensal neisseria reveals acquisition of a new ribosomal protection protein (MsrD) as a possible explanation for high level azithromycin resistance in Belgium. Pathogens 10, 384 (2021).
Lu, C. et al. Phenotypic and genetic characteristics of macrolide and lincosamide resistant ureaplasma urealyticum isolated in Guangzhou, China. Curr. Microbiol 61, 44–49 (2010).
Sharma, P. et al. Comparison of antimicrobial resistance and pan-genome of clinical and non-clinical enterococcus cecorum from poultry using whole-genome sequencing. Foods 9, 686 (2020).
Nikaido, H. Multidrug resistance in Bacteria. Annu. Rev. Biochem. 78, 119–146 (2009).
Nikaido, H. & Pagès, J.-M. Broad-specificity efflux pumps and their role in multidrug resistance of Gram-negative bacteria. FEMS Microbiol Rev. 36, 340–363 (2012).
Datta, N., Hedges, R. W., Becker, D. & Davies, J. Plasmid-determined fusidic acid resistance in the enterobacteriaceae. Microbiology 83, 191–196 (1974).
Arthur, M. & Courvalin, P. Contribution of two different mechanisms to erythromycin resistance in Escherichia coli. Antimicrob. Agents Chemother. 30, 694–700 (1986).
Jacquet, E. et al. ATP Hydrolysis and Pristinamycin IIA Inhibition of the Staphylococcus aureus Vga(A), a Dual ABC Protein Involved in Streptogramin A Resistance *. J. Biol. Chem. 283, 25332–25339 (2008).
Fürste, J. P. et al. Molecular cloning of the plasmid RP4 primase region in a multi-host-range tacP expression vector. Gene 48, 119–131 (1986).
Lutz, R. & Bujard, H. Independent and tight regulation of transcriptional units in escherichia coli Via the LacR/O, the TetR/O and AraC/I1-I2 regulatory elements. Nucleic Acids Res. 25, 1203–1210 (1997).
Dubnau, D. Induction of ermC requires translation of the leader peptide. EMBO J. 4, 533–537 (1985).
Dersch, P., Khan, M. A., Mühlen, S. & Görke, B. Roles of regulatory RNAs for antibiotic resistance in bacteria and their potential value as novel drug targets. Front. Microbiol. 8, 803 (2017).
Banerjee, S., Chalissery, J., Bandey, I. & Sen, R. Rho-dependent transcription termination: more questions than answers. J. Microbiol. 44, 11–22 (2006).
Ray-Soni, A., Bellecourt, M. J. & Landick, R. Mechanisms of bacterial transcription termination: all good things must end. Annu. Rev. Biochem. 85, 319–347 (2016).
Naville, M., Ghuillot-Gaudeffroy, A., Marchais, A. & Gautheret, D. ARNold: A web tool for the prediction of Rho-independent transcription terminators. RNA Biol. 8, 11–13 (2011).
Takada, H. et al. Expression of Bacillus subtilis ABCF antibiotic resistance factor VmlR is regulated by RNA polymerase pausing, transcription attenuation, translation attenuation and (p)ppGpp. Nucleic Acids Res gkac497 (2022). https://doi.org/10.1093/nar/gkac497.
Yakhnin, A. V. et al. NusG controls transcription pausing and RNA polymerase translocation throughout the Bacillus subtilis genome. Proc. Natl. Acad. Sci. USA 117, 21628–21636 (2020).
Shimizu, Y. et al. Cell-free translation reconstituted with purified components. Nat. Biotechnol. 19, 751–755 (2001).
Hartz, D., McPheeters, D. S., Traut, R. & Gold, L. Extension inhibition analysis of translation initiation complexes. in Methods in Enzymology 164 419–425. (Academic Press, 1988).
Meydan, S. et al. Retapamulin-assisted ribosome profiling reveals the alternative bacterial proteome. Mol. Cell 74, 481–493.e6 (2019).
Muto, H., Nakatogawa, H. & Ito, K. Genetically Encoded but Nonpolypeptide Prolyl-tRNA Functions in the A Site for SecM-Mediated Ribosomal Stall. Mol. Cell 22, 545–552 (2006).
Seip, B. & Innis, C. A. How widespread is metabolite sensing by ribosome-arresting nascent peptides? J. Mol. Biol. 428, 2217–2227 (2016).
Hansen, J. L. et al. The Structures of Four Macrolide Antibiotics Bound to the Large Ribosomal Subunit. Mol. Cell 10, 117–128 (2002).
Mao, J. C. & Robishaw, E. E. Effects of macrolides on peptide-bond formation and translocation. Biochemistry 10, 2054–2061 (1971).
Poulsen, S. M., Kofoed, C. & Vester, B. Inhibition of the ribosomal peptidyl transferase reaction by the mycarose moiety of the antibiotics carbomycin, spiramycin and tylosin1 1Edited by D. E. Draper. J. Mol. Biol. 304, 471–481 (2000).
Emmanuel. J. S., Sengupta, A., Gordon, E. R., Noble, J. T. & Cruz-Vera, L. R. The regulatory TnaC nascent peptide preferentially inhibits release factor 2-mediated hydrolysis of peptidyl-tRNA. J. Biol. Chem. 294, 19224–19235 (2019).
Ohashi, Z. et al. Characterization of C+ located in the first position of the anticodon of Escherichia coli tRNAMet as N4-acetylcytidine. Biochimica et. Biophysica Acta (BBA) - Nucleic Acids Protein Synth. 262, 209–213 (1972).
Varshney, U., Lee, C. P. & RajBhandary, U. L. From elongator tRNA to initiator tRNA. PNAS 90, 2305–2309 (1993).
Pierson, W. E. et al. Uniformity of Peptide Release Is Maintained by Methylation of Release Factors. Cell Rep. 17, 11–18 (2016).
Schmeing, T. M., Huang, K. S., Strobel, S. A. & Steitz, T. A. An induced-fit mechanism to promote peptide bond formation and exclude hydrolysis of peptidyl-tRNA. Nature 438, 520–524 (2005).
Arenz, S. et al. Molecular basis for erythromycin-dependent ribosome stalling during translation of the ErmBL leader peptide. Nat. Commun. 5, 3501 (2014).
Lovmar, M. et al. The molecular mechanism of peptide-mediated erythromycin resistance *. J. Biol. Chem. 281, 6742–6750 (2006).
Tenson, T., Xiong, L., Kloss, P. & Mankin, A. S. Erythromycin resistance peptides selected from random peptide libraries*. J. Biol. Chem. 272, 17425–17430 (1997).
Koch, M., Willi, J., Pradère, U., Hall, J. & Polacek, N. Critical 23S rRNA interactions for macrolide-dependent ribosome stalling on the ErmCL nascent peptide chain. Nucleic Acids Res. 45, 6717–6728 (2017).
Vázquez-Laslop, N., Ramu, H., Klepacki, D., Kannan, K. & Mankin, A. S. The key function of a conserved and modified rRNA residue in the ribosomal response to the nascent peptide. EMBO J. 29, 3108–3117 (2010).
Vimberg, V. et al. Ribosome-Mediated Attenuation of vga(A) Expression Is Shaped by the Antibiotic Resistance Specificity of Vga(A) Protein Variants. Antimicrobial Agents Chemother. 64, e00666–20 (2020).
Vázquez-Laslop, N. et al. Role of antibiotic ligand in nascent peptide-dependent ribosome stalling. Proc. Natl. Acad. Sci. USA. 108, 10496–10501 (2011).
Dar, D. et al. Term-seq reveals abundant ribo-regulation of antibiotics resistance in bacteria. Science 352, aad9822 (2016).
Koberska, M. et al. Beyond Self-Resistance: ABCF ATPase LmrC Is a Signal-Transducing Component of an Antibiotic-Driven Signaling Cascade Accelerating the Onset of Lincomycin Biosynthesis. mBio 12, e0173121 (2021).
Ohki, R., Tateno, K., Takizawa, T., Aiso, T. & Murata, M. Transcriptional termination control of a novel ABC transporter gene involved in antibiotic resistance in Bacillus subtilis. J. Bacteriol. 187, 5946–5954 (2005).
Dunkle, J. A., Xiong, L., Mankin, A. S. & Cate, J. H. D. Structures of the Escherichia coli ribosome with antibiotics bound near the peptidyl transferase center explain spectra of drug action. Proc. Natl. Acad. Sci. USA 107, 17152–17157 (2010).
Croucher, N. J. et al. Rapid pneumococcal evolution in response to clinical interventions. Science 331, 430–434 (2011).
Miller, J. R. GraphPad Prism Version 4.0 Step-by-Step Examples. vol. GraphPad Software Inc., San Diego (GraphPad Software Inc., San Diego, 2003).
Stead, M. B. et al. RNAsnapTM: a rapid, quantitative and inexpensive, method for isolating total RNA from bacteria. Nucleic Acids Res. 40, e156 (2012).
Seefeldt, A. C. et al. The proline-rich antimicrobial peptide Onc112 inhibits translation by blocking and destabilizing the initiation complex. Nat. Struct. Mol. Biol. 22, 470–475 (2015).
Zheng, S. Q. et al. MotionCor2 - anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat. Methods 14, 331–332 (2017).
Zhang, K. Gctf: Real-time CTF determination and correction. J. Struct. Biol. 193, 1–12 (2016).
Zivanov, J. et al. New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife 7, e42166 (2018).
Kucukelbir, A., Sigworth, F. J. & Tagare, H. D. The Local Resolution of Cryo-EM Density Maps. Nat. Methods 11, 63–65 (2014).
Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Cryst. D. 60, 2126–2132 (2004).
Liebschner, D. et al. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Cryst. D. 75, 861–877 (2019).
Boccaletto, P. et al. MODOMICS: a database of RNA modification pathways. 2017 update. Nucleic Acids Res. 46, D303–D307 (2018).
Bourgeois, G. et al. Structural basis of the interaction between cyclodipeptide synthases and aminoacylated tRNA substrates. RNA 26, 1589–1602 (2020).
Waterhouse, A. M., Procter, J. B., Martin, D. M. A., Clamp, M. & Barton, G. J. Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–1191 (2009).
Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).
Goddard, T. D. et al. UCSF ChimeraX: Meeting modern challenges in visualization and analysis. Protein Sci. 27, 14–25 (2018).
Pettersen, E. F. et al. UCSF Chimera—A visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612 (2004).
Desai, N. et al. Elongational stalling activates mitoribosome-associated quality control. Science 370, 1105–1110 (2020).
Golubev, A. et al. Cryo-EM structure of the ribosome functional complex of the human pathogen Staphylococcus aureus at 3.2 Å resolution. FEBS Lett. 594, 3551–3567 (2020).
Li, W. et al. Structural basis for selective stalling of human ribosome nascent chain complexes by a drug-like molecule. Nat. Struct. Mol. Biol. 26, 501–509 (2019).
Acknowledgements
This work has benefited from the facilities and expertize of the Biophysical and Structural Chemistry platform (BPCS) at IECB, CNRS UMS3033, Inserm US001, University of Bordeaux and supported by funds from the CNRS (UMR8261) and Paris Cité University. C.R.F. is funded by a doctoral grant from the French Ministère de l’Enseignement supérieur, de la Recherche et de lʼInnovation, F.O. and G.B received support from the LABEX program (DYNAMO ANR-11-LABX-0011), the ANR grants EZOtrad (ANR-14-ACHN-0027) for G.B. and ABC-F_AB (ANR-18-CE35-0010) for G.B. and Y.H., ERC-2017-STG #759120 “TransTryp” for Y.H.. E.C.L. and C.A.I. received funding for this project from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 724040). The authors would like to thank Dr. Olivier Chesneau (Département de Microbiologie, Institut Pasteur, Paris - France) for providing E. coli DB10 strain, plasmids pBAD33 and pVN50; Dr. Sylvain Durand and Dr. Maud Guillier (UMR 8261, CNRS, Université de Paris, Institut de Biologie Physico-Chimique, Paris - France) who provided bicyclomycin and equipment, respectively for toeprinting experiments. We also would like to thank Laura Monlezun and Tina Wang for proofreading the manuscript.
Author information
Authors and Affiliations
Contributions
F.O. performed the first experiments that initiated the project. C.R.F. performed most of the experiments following. S.N. performed northern blotting experiments. C.R.F. prepared the cryo-EM sample. H.S. prepared cryo-EM grids and imaged them. C.A.I. and Y.H. processed the cryo-EM data. C.R.F and E.C.L. reconstructed the atomic model. G.B. initiated the project and designed the research program with Y.H.. G.B. and C.R.F. wrote the paper with input from all authors. G.B., Y.H., S.N. and F.O. did the final manuscript revision.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Fostier, C.R., Ousalem, F., Leroy, E.C. et al. Regulation of the macrolide resistance ABC-F translation factor MsrD. Nat Commun 14, 3891 (2023). https://doi.org/10.1038/s41467-023-39553-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-023-39553-8
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
