Fig 1.
Bacterial stringent response and experimental approach.
(A) Schematics representing the onset of stringent response, (p)ppGpp accumulation, and gene expression regulation. Translation of stress-related mRNAs upon (p)ppGpp accumulating is the focus of the investigation (indicated as the red question mark). (B) Schematics of instrumentation for measuring 30S initiation complex (IC) by fluorescence Microscale Thermophoresis (MST) (provided by Nanotemper Technologies). Generally, each capillary was loaded with 1 μM 30S subunits, 4 μM mRNA, 2 μM IF1, 1 μM IF2, 1.5 μM IF3, 0.5 μM Bpy-Met-tRNAfMet (Bpy-tRNAi), 0.2 mM guanosine nucleotides, and varying concentrations of the tested ligand. Fluorophore excitation and measurement of the emitted fluorescence, together with IR-mediated perturbation of the equilibrium, are achieved through an objective and dichroic mirror. (C) Time courses of fluorescence showing the equilibrium relaxation upon thermal perturbation at varying ligand concentrations. Blue vertical lines indicate the fluorescence signal readout prior to the equilibrium perturbation (Fi), while vertical red lines indicate the fluorescence signal upon reaching the new equilibrium (Ff). Colored traces represent different concentrations of the titrating mRNA (0.1 to 4 μM). The ratio between the Ff over Fi times 1,000 is the normalized thermophoresis shift (MST). The dependency of MST shift on ligand concentration allows to estimate KD constants of the interaction using a hyperbolic or quadratic function and nonlinear regression fitting. MST measurements (closed circles) are indicated in the same colors as their respective time traces in (C). (D) Formation of 30S IC measured by MST. Time traces of thermophoresis (S1 Fig) were used to calculate MST values for all four combinations of mRNA start codons and labeled tRNAs for dependencies on mRNA concentration. Squares indicate 30S ICs programmed with AUG mRNA, while circles show that UUC was used as start site. Closed symbols indicate that Bpy-Met-tRNAfMet was used, while open symbols correspond to Bpy-Phe-tRNAPhe. Continuous lines indicate nonlinear regression fittings. Three to four measurements were performed; mean and error bars representing standard deviations are plotted (S1 Data). Bpy, BODIPY fluorophore; DksA, RNA polymerase-binding transcription factor DksA; GDP, guanosine diphosphate; GTP, guanosine triphosphate; IC, initiation complex; IF1, translation initiation factor IF1; IF2, translation initiation factor IF2; IF3, translation initiation factor IF3; IR, infrared; MST, Microscale Thermophoresis; (p)ppGpp, guanosine tetra- and pentaphosphate; RelA, (p)ppGpp synthase RelA; SpoT, Bifunctional (p)ppGpp synthase/hydrolase SpoT; tRNAi, initiator tRNA.
Fig 2.
pppGpp allows translation initiation.
(A) Formation of 30S IC measured by MST at increasing concentrations of IF2 in the presence of ppGpp (squares), pppGpp (down triangles), GTP (up triangles), GDP (circles), or in the absence (diamonds) of guanosine nucleotides. Continuous lines indicate nonlinear regression fittings with a quadratic equation. Three measurements were performed; mean and standard deviation are plotted (S1 Data). (B) as (A) to measure the concentration dependence for each tested guanosine nucleotide at constant IF2 (1 μM). (C) Formation of 70S pre-IC as measured by light scattering (LS) on a stopped-flow apparatus [29]; 0.1 μM 30S ICs were formed in the absence of (cyan) or the presence of 0.2 mM ppGpp (red), pppGpp (orange), GTP (blue), or GDP (purple) and rapidly mixed with 0.3 μM 50S subunits. Continuous lines show best fits using exponential functions. (D) Formation of 70S IC as measured by Bpy-tRNAi accommodation in the P site of the ribosome; 0.1 μM 30S ICs were formed with Bpy-tRNAi and rapidly mixed with 0.3 μM 50S subunits on a stopped-flow apparatus as described in [14]. Colors are as in (C). Continuous lines show best fits using an exponential function for a single reaction step. Each stopped-flow trace results from the mean of 5 to 7 replicates. Bpy-tRNAi, Bodipy labelled initiator tRNA; GDP, guanosine diphosphate; GTP, guanosine triphosphate; IC, initiation complex; IF2, translation initiation factor IF2; MST, Microscale Thermophoresis.
Fig 3.
MST analysis shows an mRNA dependence of 30S IC formation.
(A) Scheme of the 30S IC highlighting ppGpp (red square) and GTP (blue triangle) competition for IF2. (B) MST dependence on the mRNA concentration for mTufA (squares), mInfA (triangles), or model messenger mMF1 (circles). MST was calculated from the respective fluorescence time dependencies (S3 Fig). Comparison of the resulting KD (C) and MST amplitudes (D) as an indicator of the efficiency of 30S IC formation for all three mRNAs. (E) Formation of 30S IC with increasing concentrations of GTP for the three mRNAs. Symbols as in (B). (F) Comparison of KD values calculated from (E). (G) ppGpp to GTP competitive assays for 30S IC formation. The 30S ICs formed with each mRNA (symbols as in B) in the presence of 50 μM GTP were subjected to increasing concentrations of ppGpp. Log of competitor concentrations is plotted and used for determining the inhibitory concentration for 50% inhibition (IC50) using a same-site competition model. (H) Bar graph comparing IC50 values for ppGpp (black) or GDP (gray) for all three mRNAs (S4 Fig). Continuous lines indicate nonlinear regression fittings with quadratic (B and E) or same-site competition (G) functions. Three measurements were performed; mean and standard deviations are plotted (S1 Data). GDP, guanosine diphosphate; GTP, guanosine triphosphate; IC, initiation complex; IF1, translation initiation factor IF1; mInfA, InfA mRNA; MST, Microscale Thermophoresis; mTufA, TufA mRNA; tRNAi, initiator tRNA.
Fig 4.
Kinetic parameters of ppGpp-mediated regulation of 70S IC progression.
(A) Scheme of 70S IC formation. The 30S complexes programmed with mTufA (red), mInfA (purple), or mMF1 (blue) and 20 μM GTP in the absence (B) or presence (C) of 200 μM ppGpp were mixed with 50S subunits. Time traces were analyzed by nonlinear regression with exponential terms. (D) Bar graph comparing amplitudes in the absence of any competing nucleotide (white) or in the presence of ppGpp (black) or GDP (gray). (E) Bar graph comparing apparent rates of 70S pre-IC formation (colors as in D). Formation of 70S IC, as measured by Bpy-tRNAi accommodation in the absence (F) or presence of 200 μM ppGpp (G). Time traces were analyzed by nonlinear regression with one exponential term. (H) Bar graph comparing amplitude variations as a function of mRNAs and guanosine nucleotides (colors as in D). (I) Bar graph comparing apparent rates of 70S IC formation for all three mRNAs (bar colors as in D). Continuous lines show best fits using an exponential function for a single reaction step. All time traces are mean values of 5 to 10 replicates. Bar graphs show mean and standard errors (error bars) derived from the nonlinear regression fitting (D,E,H,I) (S1 Data). Bpy-tRNAi, Bodipy labelled initiator tRNA; GDP, guanosine diphosphate; GTP, guanosine triphosphate; IC, initiation complex; IF3, translation initiation factor IF3; LS, light scattering; mInfA, InfA mRNA; mTufA, TufA mRNA.
Fig 5.
Permissive ppGpp-mediated inhibition of mRNA translation.
(A) Scheme of the translating 70S complex. (B) In vitro translation of mTufA and mInfA derivatives harboring a coding sequence for the Lumio labelling system (Experimental conditions and analysis). Protein synthesis reactions were started in the presence of 2 mM GTP and increasing concentrations of ppGpp. Synthetized proteins were fluorescently labelled and resolved by 20% SDS-PAGE. The resulting images were analyzed by pixel densitometry using ImageJ [37] to estimate translation efficiencies (TEs). (C) Log of ppGpp concentrations was plotted and used for determining the inhibitory concentration for 50% inhibition (IC50) using a same-site competition model. (D) Comparison of inhibitory constants for both mRNAs as measured during 30S IC formation (white) or overall translation efficiency (gray). Mean and standard deviations from three measurements are plotted (S1 Data). GTP, guanosine triphosphate; mInfA, InfA mRNA; mTufA, TufA mRNA; TE, translation efficiency.
Fig 6.
mTufA structured enhancer of translation initiation.
(A) Two-dimensional representation of the mTufA 5′ UTR. The RNAfold algorithm predicted three structured regions: SR I (pink), SR II (violet), SR IIIa (yellow), and SR IIIb (green). Blue and pink arrows indicate adenines or cytidines found to be unpaired in vivo by DMS chemical probing [38]. Blue arrows indicate a match between modelled structures and in vivo chemical probing, whereas pink arrows indicate a lacking match. Black arrows indicate the starting points of the mRNA truncations used here and in vivo active promoters. (B) Three-dimensional modelling of the three structured elements contained in the mTufA 5′ UTR. Three-dimensional modelling was performed either using the full-length (FL) UTR (gray) or the isolated structured regions (colored as in (A)). (C) Scheme representing the different mTufA truncations used in this study. FL stands for the UTR used in Figs 3–5. Translation initiation was measured by MST as a function of the mRNA (D) or GTP (E) for all mTufA truncations. (F) ppGpp to GTP competition experiments as measured by MST. (G) In vitro translation efficiencies of mTufA derivatives in the absence or presence of 2 mM ppGpp. Mean and standard deviations from three measurements are plotted in D, E, F, and G (S1 Data). DMS, dimethyl sulfide; GTP, guanosine triphosphate; MST, Microscale Thermophoresis; mTufA, TufA mRNA; SR, structured region.
Fig 7.
(A) Translation initiation efficiency as a function of mRNAs along five steps of transition from NS to S conditions. Nucleotide concentrations are indicated by the filled areas and refer to the right axis, GTP (blue), GDP (violet), ppGpp (red), and pppGpp (orange). mTufA (square), mInfA (triangles), and mMF1 (circles) represent averages of three replicates (S1 Data). (B) as in (A) for mTufA (squares) or mRnr (circles) variants containing (close) or lacking (open) the SETI element proximal to the TIR. Symbols and error bars represent averages of three replicates and standard deviations, respectively (S1 Data). (C) Comparison of ppGpp tolerance as calculated from the ratio of translation initiation efficiencies found in the presence of 0.2 mM ppGpp over the efficiencies obtained in the absence of the alarmone (S1 Data). Translation initiation efficiencies for all mRNAs tested here were calculated from thermophoresis measurements considering the highest observed MST (+200 mTufA construct) as 100% and that of free Bpy-tRNAi as 0%. All 30S IC components are as in Fig 1. Dashed line indicates the average tolerance of all mRNAs lacking the SETI element. FL stands for full-length while Tr indicates a truncation of the SETI element in mRnr or the corresponding segment in mTktB. (D) Overall folding comparison of the mTufA and mRnr SETIs. Both mRNA segments share similar folds, albeit having a primary sequence identity below 30% (see S2 File for primary and secondary structure comparisons). Bpy-tRNAi, Bodipy labelled initiator tRNA; GDP, guanosine diphosphate; GNP, guanosine nucleotide; GTP, guanosine triphosphate; IC, initiation complex; mInfA, InfA mRNA; mRnr, Rnr mRNA; MST, Microscale Thermophoresis; mTktB, TktB mRNA; mTufA, TufA mRNA; NS, non-stringent; S, stringent; SETI, structured enhancer of translation initiation; TIR, translation initiation region.
Fig 8.
Model of translation initiation during stringent response.
(A) Schematics representing translation initiation at NS conditions (i), at a high ppGpp/GTP ratio (ii), or by using pppGpp (iii) during stringent response. At high ppGpp/GTP ratio, ppGpp binds IF2, precluding start codon recognition and promoting the 30S pre-IC. In turn, the 30S pre-IC can exchange the bound mRNA for a more ppGpp-tolerable transcript, allowing GTP to replace the tetraphosphate and proceed to protein elongation. Alternatively, IF2 can bind pppGpp and proceed towards translation elongation. (B) Structural model of the 30S pre-IC programmed with mTufA SETI (green surface), structurally aligned on PDB 5LMP [42]. Ribosomal proteins uS7 (violet surface) and uS9 (cyan surface) are highlighted as potential sites for mTufA SETI interactions. The red circle indicates clashing of the mTufA SETI with the 30S pre-IC complex. (C) Structural model of the 30S IC (PDB 5LMV, [42]) and the mTufA SETI. Head movements related to 30S IC formation upon decoding the start codon and potentially enhanced by SETI elements are indicated by arrows [42,43]. GTP, guanosine triphosphate; IC, initiation complex; IF2, translation initiation factor IF2; mTufA, TufA mRNA; NS, non-stringent; PDB, Protein Data Bank; SETI, structured enhancer of translation initiation; tRNAi, initiator tRNA; uS7, universal ribosomal protein S7; uS9, universal ribosomal protein S9.