Stapled Phd Peptides Inhibit Doc Toxin Induced Growth Arrest in Salmonella

Bacterial toxin inhibition is a promising approach to overcoming antibiotic failure. InSalmonella, knockout of the toxin Doc has been shown to significantly reduce the formation of antibiotic-tolerant persisters. Doc is a kinase that is inhibited in nontolerant cells by its cognate antitoxin, Phd. In this work, we have developed first-in-class stapled peptide antitoxin mimetics based on the Doc inhibitory sequence of Phd. After making a series of substitutions to improve bacterial uptake, we identified a lead stapled Phd peptide that is able to counteract Doc toxicity in Salmonella. This provides an exciting starting point for the further development of therapeutic peptides capable of reducing antibiotic persistence in pathogenic bacteria.


■ INTRODUCTION
Protein−protein interactions (PPIs) are increasingly recognized as tractable drug targets, despite their initial "undruggable" reputation. 1,2Over recent years, a number of strategies have emerged for their successful modulation, with peptide "stapling" becoming a frequent method of choice. 3,4"Stapled" peptides are derived from the binding sequence of one protein partner of an α-helix-mediated PPI.Outside of the stabilizing environment of a whole protein, peptide sequences may not adopt a well-defined helical conformation.This results in susceptibility to proteolytic degradation and, often, limited cell uptake. 5To circumvent these issues, amino acids capable of chemical cross-linking to one another can be introduced into the active sequence at a spacing of one (i, i + 4) or multiple (i, i + 7 or i, i + 11) helical turns. 6The formation of a covalent bond, a "staple", between these two residues constrains the peptide into a permanent helical conformation, improving affinity, stability, and cell penetration, 7 including in Gramnegative bacteria. 8,9Since their initial introduction, stapled peptides have risen in prominence, providing validated inhibitors for multiple targets in cancer, 10,11 including some progressing to clinical trials. 12,13Beyond cancer, stapled peptides have also shown promise in malaria 14 and for bacterial targets, with a number of examples with efficacy against drug transporters, 15 cell division, 16 and gene transcription. 8−21 Bacterial toxins are a class of ubiquitous proteins which act in response to stress and represent an untapped pool of targets for inhibition by peptide ligands. 22Toxins inhibit key cellular processes, leading to bacterial growth arrest, and have been linked to increased survival of bacteria to host immune defense, antibiotic treatment, and bacteriophages. 23,24They are expressed alongside a cognate antitoxin, which, for Type II toxin-antitoxin systems, forms an inhibitory PPI in nonstressed cells. 25Activation of the toxin through degradation or reduced expression of the antitoxin results in bacterial growth arrest, enabling survival under stress conditions. 26Therefore, toxin inhibition, through mimicry of the mode of action of the antitoxin, could provide a mechanism to reduce population survival under antibiotic stress. 27,28n Salmonella enterica serovar typhimurium (S. typhimurium), growth-arrested antibiotic persisters are formed upon macrophage internalization and complicate clearance of the infection.The number of macrophage-induced antibiotictolerant cells is significantly reduced upon knockout of the phddoc toxin-antitoxin module encoding for the toxin Doc and its antitoxin partner, Phd. 23Doc functions as a kinase, phosphorylating the translation elongation factor EF-Tu and subsequently inhibiting protein synthesis. 29,30We have previously carried out a comprehensive characterization of the Phd-Doc interface and demonstrated that antitoxin peptides can effectively mimic the activity of the full length Phd protein both in vitro and when expressed in S. typhimurium. 31ere, we report the development of stapled Phd peptides capable of the rescue of Doc-induced growth arrest when administered to S. typhimurium (Figure 1).Using the Cterminus from S. typhimurium Phd (Phd STm 52−73 ) with pM affinity for Doc (Doc STm ) as a template, we generated a library of analogues.A combination of residue substitutions and hydrocarbon stables was used to reduce the overall negative charge of the sequence and enhance bacterial uptake while minimizing negative effects on affinity and Doc STm inhibition activity.S. typhimurium cultures were then treated with a subset of optimized peptides, which proved capable of counteracting the effects of Doc STm toxicity.This study provides the first example of extracellularly administered inhibitors of toxin-induced growth arrest and promising starting points for the optimization of stapled peptide toxin inhibitors as agents to reduce antibiotic persistence.

■ RESULTS
Arginine Scan of Phd 52−73 Peptide to Reduce the Negative Charge of the Wild-Type Sequence.We previously characterized the C-terminal domain of S. typhimurium Phd antitoxin peptide (Phd STm 52−73 ) as a high affinity inhibitor of Doc STm toxin, exhibiting pM binding affinity and Doc STm inhibition activity comparable to fulllength Phd STm 1−73 protein. 31Based on this activity, the Phd STm 52−73 peptide holds great promise as a Doc STm toxin  inhibitor to tackle and study antibiotic persistence in S. typhimurium.However, due to its high negative net charge of −4.9 at neutral pH 7 (Figure 2A), it was expected that the Phd STm 52−73 peptide would exhibit very low penetration into Gram-negative bacteria, necessitating the development of a cell-permeable Phd STm 52−73 variant able to target the intracellular toxin.
As a starting point for the generation of cell-permeable Phd STm 52−73 peptides, we used the wild-type Phd 52−73 sequence with N-terminal acetylation and an additional N-terminal tryptophan residue to allow spectrophotometric quantification, as previously described. 31All peptides were prepared by an automated microwave-assisted solid-phase peptide synthesis (SPPS) using the Fmoc/t-Bu strategy.As an initial modification, Met52 and Met60 of Phd 52−73 were replaced with norleucine to prevent oxidation problems and interference with hydrocarbon stapling, resulting in peptide 1 (Figure 2A).The binding of peptide 1 to recombinantly produced Doc STm was verified by a thermal shift assay (Figure 2A,B), revealing a large thermal stabilization of Doc STm by 1 with a positive melting temperature (T m ) shift of 28.8 ± 2.7 °C.The observed thermal shift was the same as previously measured for wild-type Phd 52−73 (ΔT m = 28.8 ± 0.9 °C), confirming that methionine replacement by norleucine did not alter the Doc STm binding affinity of the peptide.Therefore, peptide 1 was used as a template for the generation of further Phd 52−73 analogues.
In a first development cycle toward cell-permeable Phd STm 52−73 peptides, we performed an arginine scan of 1 to identify residues suitable for the introduction of positive charges without loss of Doc STm binding affinity.Peptide variants with single arginine substitutions of all nonarginine residues except for the previously identified hot-spot residues for Doc STm binding and inhibition (Phe56, Ile59, Nle60, His63, Leu67, and Leu70), 31 as well as Nle52 and Lys73, were prepared (2−13, Figure 2A).Thermal shift analysis of Doc STm in the presence of each peptide revealed that only the substitution of Glu55, Val62, and Glu66 was not fully tolerated, as displayed by the reduced thermal stabilization of Doc STm for these peptides (ΔT m ≈ 15−23 °C).
To check the effect of multiple arginine substitutions on the Doc STm binding of 1, peptides 14 and 15 with five and six welltolerated arginine substitutions, respectively, were synthesized.However, the introduction of this larger number of arginine residues in 14 and 15 led to a significant loss of Doc STm stabilization (ΔT m ≈ 11−13 °C).

N-Terminal Arginine-Spiking and Stapling Retain In Vitro Toxin Inhibition and Enable Bacterial Uptake.
Since the introduction of a larger number of arginine residues in peptide 1 proved to be detrimental to Doc STm binding and single arginine substitutions are unlikely to be sufficient to improve bacterial uptake, we investigated if the substitution of a few selected residues by arginine in combination with i, i + 4 hydrocarbon stapling would generate Phd 52−73 analogues with reduced negative charge and simultaneous staple-induced improved bacterial penetration.Pentenylalanine residues stapled using Grubbs metathesis were selected due to their prevalence in the field and ease of synthesis. 32esidues Asp53 and Ala61 or Asp53, Ala57, and Ala61 in the N-terminal part of peptide 1 were substituted for arginine to generate analogues 16 and 19 with two or three additional arginines and net charges of −1.9 and −0.9, respectively (Figure 3A).These substitutions were combined with i, i + 4 hydrocarbon stapling between residues 4 and 8 in the Nterminal half of the peptide (17 and 20) as well as stapling between residues 15−19 in the C-terminal half of the peptide (18 and 21) to obtain analogues with net charges of −0.9 to 0.1 (Figure 3A,B).For modified peptides, 16−21 a loss in Doc STm stabilization compared to 1 was observed by thermal shift (ΔT m ≈ 15−20 °C).In addition, the Doc inhibition activity of the peptide analogues was assessed in a phosphorylation assay with recombinant EF-Tu STm and Doc STm and analyzed by dot blot as previously described. 31eptide 1 and analogues 16 and 19 fully inhibited EF-Tu STm phosphorylation by Doc STm when present at three times the concentration of Doc STm in the assay (1 μM), while analogues 17, 18, 20, and 21 required a higher concentration to achieve full inhibition (Figure 3C).
As the net charge of peptides 16−21 at neutral pH was still negative or zero, charge reversal of the peptide sequence by additional removal of negatively charged aspartic acid and glutamic acid residues was investigated.Arginine-spiked peptide sequences 16 and 19 were further modified by substitution of Asp54 and Asp72 with asparagine and substitution of Glu55 and Glu66 with glutamine to obtain analogues 22 and 24 with positive net charges of 2.1 and 3.1, respectively.This was additionally combined with i, i + 4 hydrocarbon stapling between residues 15−19 in the Cterminal half of the peptide to gain the positively charged stapled analogues 23 and 25 (Figure 3A).Peptides 22−25 still bound to Doc STm but exhibited a further decreased stabilization compared to 16−21 by thermal shift (ΔT m of 22−25 ≈ 12−15 °C).Furthermore, a concentration of 22−25 higher than three times the Doc STm concentration was required for full Doc STm inhibition in the EF-Tu STm phosphorylation assay (Figure 3D).Due to the very high evolutionary conservation of Glu55 in Phd antitoxin across bacterial species and the significant loss in Doc interaction when replacing this residue with arginine (peptide 4, Figure 2A), we speculated that substitution of Glu55 for glutamine might have caused the observed loss in Doc binding affinity in peptides 22−25.Accordingly, we prepared analogues 26−29, which replicate 22−25 but still contain residue Glu55, resulting in peptides with positive net charges of 1.1−3.1 (Figure 3A).The presence of Glu55 was sufficient to regain Doc STm binding affinity, as analogues 26−29 demonstrated a significantly higher thermal stabilization of Doc STm in the thermal shift assay (ΔT m of 26− 29 ≈ 20−21 °C).In addition, the unstapled peptides 26 and 28 demonstrated a Doc STm inhibition activity similar to that of 1 in the EF-Tu STm phosphorylation assay (Figure 3D), while the stapled peptides 27 and 29 still required a higher concentration for full Doc STm inhibition.
Peptide analogues 16−25 as well as base peptide 1 were subsequently analyzed for their cellular uptake into E. coli as a Gram-negative model bacterium, to check if the introduced sequence modifications and stapling translated into a higher bacterial cell penetration of the Phd 52−73 peptide.For the uptake studies, E. coli MG1655 cells were treated with 5 μM of fluorescein (FAM)-labeled peptide variants for 2 h, and the penetration of the peptides into the bacterial cells was determined by flow cytometry.As expected, the highly negatively charged peptide FAM-1 displayed extremely low uptake into E. coli, while FAM-labeled 16, 18, 19, and 22−24 exhibited an increased bacterial uptake with penetration efficiencies of around 7−14% (Figure 4).For the FAM-labeled Residues Asn65/Glu69 or Asn65/Glu69/Asp72 in peptide 1 were substituted for arginine and mixed with different combinations of substitutions of asparagine/glutamine equivalents of Asp53, Asp54, Glu66, and Asp72 to yield peptides 30−33 with net charges ranging from 1.1 to 3.1 (Figure 5A).The substitutions in 31−33 were well tolerated (ΔT m ≈ 20 °C) similar to 26−29, while 30 exhibited a reduced stabilization of Doc STm (ΔT m ≈ 14 °C).A concentration higher than three times the Doc STm concentration was required for full Doc STm inhibition by 30, 31, and 33 in the EF-Tu STm phosphorylation assay, while 32 displayed a slightly higher Doc inhibition activity close to the activity of peptide 1 (Figure 5B).
Peptide 33 with good Doc STm stabilization and the highest positive net charge was subsequently used for stapling evaluation: i, i + 4 hydrocarbon stapling was performed in all possible positions in the N-terminal half of the peptide up to the central glycine (Figure 5A), while keeping the hot-spot residues intact.The resulting stapled analogues 34−37 displayed a Doc inhibition activity similar to that of unstapled 33 (Figure 5C), i.e., reduced activity compared to peptide 1.In addition, 34−36 stabilized Doc STm to the same extent as 33 by thermal shift (ΔT m ≈ 18−21 °C), while analogue 37 with the staple located closest to the suggested structural glycineinduced kink in the peptide displayed a reduced stabilization (Figure 5A).
FAM-labeled variants of the new peptide series 30−37 were tested for their cellular uptake into E. coli; however, precipitation was observed for all peptides in the assay conditions, resulting in large errors and an overestimation of the uptake (Figure S1).
Cellular Uptake of Phd 52−73 Peptide Analogues in S. typhimurium and Intracellular Doc STm Inhibition Activity.Selected Phd 52−73 analogues with the highest Doc STm binding affinities were investigated for their cellular uptake into Gram-negative target bacterium S. typhimurium.Bacteria were treated with fluorescein-labeled versions of peptide 1, unstapled 33, and stapled peptides 18, 27, 29, 34, and 36 at two different concentrations (2 and 10 μM).Uptake of peptides into the bacteria after 4 and 22 h, to determine both initial and longer-term uptake, was then measured by flow cytometry.As in the E. coli studies, highly negatively charged peptide 1 displayed very low uptake into S. typhimurium was present at both tested concentrations and incubation times.Additionally, at 2 μM, all modified peptides displayed very low to no uptake, except for stapled analogue 36, which was taken up into around 7% of bacteria after 22 h of incubation (Figure 6A).At 10 μM concentration, peptides 18, 27, 29, and 36 displayed higher uptake, with varying penetration efficiencies of around 9−30% (Figure 6B).
To investigate whether our Phd peptides could inhibit Doc toxin activity inside S. typhimurium, we performed growth rescue experiments.Peptide 1 and peptides 18 and 26−29 from the series of N-terminally arginine-spiked and Cterminally stapled Phd STm 52−37 analogues, as well as peptides 33−37 from the series of N-terminally stapled and Cterminally arginine-spiked analogues, were tested in the growth rescue assay.A culture of Doc STm -expressing S. typhimurium strain was diluted to an OD 600 of 0.1 and treated with 2 or 10 μM peptide for 24 h at 37 °C while growth of the bacteria was monitored via OD 600 .Untreated S. typhimurium displayed a clear growth arrest after 24 h of culture due to the activity of Doc STm toxin (Figure 6C,D).Coexpression of a wild-type Phd STm 52−37 peptide in S. typhimurium as a positive control for Doc inhibition resulted in growth rescue, as displayed by a final OD 600 of around 0.9 after 24 h of culture (Figure 6C,D).Despite the low cell uptake, consistent and full rescue from Doc STm -induced growth inhibition was obtained for base peptide 1 after 24 h of treatment when used at 10 μM concentration (Figure 5C,E).The stapled peptide with the most promising uptake profile in S. typhimurium, 36, reproducibly rescued Doc STm -induced growth arrest (Figure 6C,F).For all other peptides, we observed a larger variation and predominantly lower growth rescue activity (Figures S2  and S3).Analysis of the growth curves from independent experiments (Figure 6D−F) shows that growth rescue kinetics varied between experiments, with peptide 36 displaying faster growth rescue than peptide 1 in all experiments.Growth rescue by peptide 36 was slower than coexpression of a Phd STm 52−37 peptide, which is expected due to the time required for bacterial cell penetration.

■ DISCUSSION
The generation of cell-permeable ligands for bacterial targets is a significant challenge, particularly for Gram-negative organisms that present a peptidoglycan layer sandwiched between an inner and an outer membrane as a barrier.For peptide ligands, the majority of effort has been focused on the development of antimicrobial peptides which destabilize these membranes to bring about a bactericidal effect. 33However, for many targets and applications, bacterial cell penetration without any associated toxicity would be highly desirable.One such class of targets is bacterial toxins, which are involved in stress responses.Peptide ligands have been developed which work to activate toxins with the aim of initiating cell death. 9,34,35owever, we propose an alternative approach.Given that toxin activation may result in the increased formation of antibiotic tolerance persister cells, 36 we have instead focused on the development of toxin inhibitors.
In this work, we sought to develop cell-penetrant peptide inhibitors of the bacterial toxin Doc, a target implicated in the survival of Salmonella to antibiotic treatment. 23Taking the Doc STm binding sequence from its cognate antitoxin, Phd, as a starting point, we carried out a series of modifications with the aim of improving uptake into Gram-negative Salmonella while retaining the high affinity and inhibition activity of the wild type.As our initial sequence, 1 had a net charge of −4.9 at pH 7, and knowing how crucial positive charges are for cell uptake, we initially performed an arginine scan to determine the tolerance of the sequence for positive charge substitutions.With this and our previous work characterizing the interaction hot spot residues 31 in hand, we then synthesized a library of peptides with two or three arginine substitutions in combination with the replacement of native carboxylic acid residues (Asp and Glu) for equivalent amide side chains (Asn and Gln) and hydrocarbon peptide stapling.In all cases, this resulted in varying degrees of reduction in Doc STm stabilization and inhibition activity in vitro when compared to peptide 1.Fortunately, in many cases these reductions were tolerated, yielding peptides capable of stabilizing Doc STm with a melting temperature shift (ΔT m ) of more than 20 °C and successfully inhibiting the toxin, despite possessing net charges of between +1.1 and +3.1 at pH 7 (18, 26−29, 33−37).
A subset of fluorescein-labeled analogues of these peptides showed promising improvements in uptake in S. typhimurium in comparison with peptide 1, which displayed less than 1% fluorescent cells after 22 h at a concentration of 10 μM.In contrast, when treated with 2 μM of peptide 36, approximately 10% of cells contained peptide after 22 h, and when treated with 10 μM of peptide 36, more than 20% of cells showed evidence of internalized peptide (Figure 5A,B).We then treated S. typhimurium cells expressing Doc STm with peptides 1, 18, 26−29, 33−37.In the absence of Doc STm inhibition, bacterial growth and replication were halted, and OD 600 remained static over the time course of the experiment.When Phd STm 52−73 was coexpressed with Doc STm , growth was rescued, and the cultures reached an OD 600 of ∼0.9 after 24 h (Figure 5D).When treated with either peptide 1 or 36, growth rescue was also observed and the same maximal OD 600 was reached after 24 h in both cases, albeit with a longer lag time (Figure 5E).Stapled peptide 36 was able to rescue growth more rapidly than 1, indicating that the improved cell uptake enhances the in vivo activity (Figure 5F).Given the poor uptake of FAM-1, the observed Doc inhibition was somewhat surprising.It is possible that the hydrophobic fluorescein hinders bacterial cell penetration, resulting in an underestimation in peptide cell penetration, and/or that the higher affinity and activity of this sequence (closest to the native antitoxin) ensured that any which does permeate a bacterial cell is highly effective in binding to and inhibiting the toxin.

■ CONCLUSIONS
We have developed a novel class of Phd STm peptides that effectively inhibit, both in vitro and in vivo, the toxin Doc STm , which contributes to the antibiotic survival of S. typhimurium.By using a combination of amino acid substitutions and hydrocarbon stapling, we have significantly improved the uptake of the wild-type sequence to enable the extracellular administration of our Doc STm inhibitors.The most effective peptide, 36, fully rescued Doc STm -induced growth inhibition at a faster rate than the unmodified peptide 1.This paves the way for the application of stapled peptides as toxin inhibitors to reduce the antibiotic tolerance of pathogenic bacteria.
To generate hydrocarbon-stapled peptides, N-terminally Fmocprotected resin-bound peptides (50 μmol scale) with (S)-2-(4pentenyl)-alanine residues in positions i, i + 4 were washed with DCM and anhydrous DCE.Ring-closing metathesis was then performed by treatment of the peptide resins with 10 mM Grubbs first-generation catalyst in anhydrous DCE (1 mL) for 2 × 2−3 h at RT under nitrogen atmosphere.Following the reaction, peptide resins were extensively washed with DCE and DCM, the N-terminal Fmoc group was removed, and peptides were either acetylated or coupled to 5(6)-carboxyfluorescein as described above.
Cleavage from the resin and simultaneous side chain deprotection were accomplished using a mixture of TFA/TIS/H 2 O (95/2.5/2.5, 3 mL) for 45 min at 40 °C in the CEM Razor rapid peptide cleavage system or for 3 h at RT.The cleavage mixture was concentrated to around 1 mL under a nitrogen stream, and crude peptides were precipitated and washed with ice-cold diethyl ether.The peptide pellets were dissolved in ACN/H 2 O and subsequently lyophilized.Crude peptides were purified on a Shimadzu LC-20AR preparative HPLC system using a preparative reversed-phase Phenomenex Aeris Peptide XB-C18 column (150 × 21 mm, 5 μm, 100 Å) with a flow rate of 20 mL/min, different linear gradients of eluent B1 [0.08% (v/ v) TFA in ACN] in eluent A1 [0.1% (v/v) TFA in water], and detection at 220 nm.The purity of the peptides was determined on a Shimadzu LC-2030C 3D HPLC system using an analytical Phenomenex Aeris Peptide XB-C18 column (150 × 4.6 mm, 3.6 μM, 100 Å) with a flow rate of 1.5 mL/min, a linear gradient of 20 to 95% eluent B1 in eluent A1 over 15 min, and detection at 220 nm.The correct identity of the peptides was confirmed on a Waters LC− MS system (2545 quaternary gradient module, 2767 sample manager, system fluidics organizer, and 3100 mass detector) using an analytical reversed-phase Waters XBridge C18 column (100 × 4.6 mm, 5 μm, 130 Å, 1.2 mL/min), a linear gradient of 20 to 98% eluent B2 [0.1% (v/v) formic acid in ACN] in eluent A2 [0.1% (v/v) formic acid in water] over 10 min, and mass detection in the range from 400 to 2000 m/z, as well as by MALDI-ToF mass spectrometry (Micromass, Waters).The observed masses were in agreement with the calculated masses, and a purity of >93% could be obtained for all compounds by LC−MS analysis (Tables S1 and S2).
Protein Expression and Purification.C-terminally His-tagged Salmonella Typhimurium Doc toxin and EF-Tu protein were recombinantly produced in E. coli as previously described. 31iochemical and Biological Methods.Thermal Shift Assay.The thermal shift assay was performed using a Mx3005P qPCR System (Agilent) collecting fluorescence data with a temperature ramp of 25 to 95 °C.Samples were prepared in a buffer containing 20 mM K 2 HPO 4 and 50 mM (NH 4 ) 2 SO 4 at pH 8.0, with a final concentration of recombinant Doc protein at 5 μM and Phd peptides at 50 μM.SYPRO Orange dye (Sigma-Aldrich, 5000X stock in DMSO) was used to monitor protein denaturation in a final concentration of 3×.The final sample volume was 20 μL.Each condition was prepared in triplicate in each independent experiment.The melting curves were plotted using GraphPad Prism 9 (GraphPad Software, USA), and the melting temperatures were obtained by fitting the sigmoidal section of the curves to a Boltzmann sigmoid function.For the calculation of thermal shifts ΔT m , the average melting temperature of free Doc STm toxin as determined in the respective peptide measurement cycles was subtracted from the average melting temperature of Doc STm in the presence of the peptide (Table S3).
Dot Blot Phosphorylation Assay.Samples were prepared in assay buffer [50 mM HEPES (pH 7.5), 25 mM (NH 4 ) 2 SO 4 , 2 mM TCEP, 2 mM MgCl 2 , and 1 mM ATP] with a final concentration of recombinant Doc at 1 μM, recombinant EF-Tu at 3 μM, and varying concentrations of synthetic Phd peptides (final concentrations: 10 μM to 5 nM).EF-Tu (3 μM) in assay buffer was used as a negative control (no Doc and Phd peptide), while EF-Tu (3 μM) with Doc (1 μM) in assay buffer was used as a positive control (no Doc inhibitor).The final sample volume was 10 μL.Samples were incubated for 16 h at RT and subsequently spotted on a nitrocellulose membrane.Phosphorylated EF-Tu was detected by immunodecoration using a rabbit monoclonal antiphosphothreonine antibody (Abcam, ab218195) at 1:2000 dilution (1 h at 4 °C), followed by incubation with a goat antirabbit IgG (H + L) HRP conjugate antibody (Advansta) at 1:10,000 dilution (1 h at RT). Chemiluminescence was developed using the HRP Luminata Kit (Merck, WBLUR0100) and captured with an ImageQuant LAS4000 Western blot imaging system (GE HealthCare).Each peptide was tested in two independent experiments.
Peptide Uptake in E. coli.E. coli MG1655 strain was grown in LB medium to an OD 600 of around 0.5.Approximately 10 7 cells were washed once with PBS and subsequently incubated with 5 μM of the indicated 5(6)-carboxyfluorescein-labeled Phd 52−73 peptide in PBS (500 μL) for 2 h at 37 °C.Bacteria were washed twice with PBS, and trypan blue (1 mg mL −1 ) in PBS was added and incubated for 10 min at RT. Bacteria were washed an additional time with PBS and resuspended in 2 mL of PBS.Flow cytometry analysis was performed using a ThermoFisher Attune NxT and a blue laser (BL1) for fluorescein excitation.
Peptide Uptake in S. typhimurium.S. typhimurium (14028s) glmS::mCherry strain was grown overnight in LB medium containing 1% glucose.The culture was then diluted to an OD 600 of approximately 0.1 into 200 μL of fresh M9 minimal medium supplemented with 0.5% arabinose containing 2 or 10 μM of the indicated 5(6)-carboxyfluorescein-labeled Phd 52−73 peptide.The samples were incubated on the bench at RT for 4 or 22 h with no access to light.Bacteria were then spun down, washed with 1 mL of PBS solution, and finally resuspended in 1 mL of PBS solution for flow cytometry analysis (BD LSR II).Constitutively expressed mCherry was used to discriminate bacteria from the debris in each sample.
Growth Rescue Experiments in S. typhimurium.S. typhimurium (14028s) Δphd-doc::Km strains carrying pCA24N (empty vector or encoding for the Phd 52−73 variant) and pBAD33::doc plasmids were grown overnight in LB medium containing 1% glucose and supplemented with 100 μg/mL carbenicillin and 34 μg/mL chloramphenicol antibiotics.Cultures were then diluted to an OD 600 of approximately 0.1 into 200 μL of fresh M9 minimal medium supplemented with 0.5% arabinose, 100 μg/mL carbenicillin, 34 μg/mL chloramphenicol, and containing 2 or 10 μM of the indicated synthetic Phd 52−73 peptide.Untreated Doc STm -expressing S. typhimurium as well as S. typhimurium coexpressing Doc and Phd 52−73 peptides were included as controls in each experiment.After 2 h of incubation at RT, samples were transferred to a flat-bottom 96-well plate (Greiner), and OD 600 was monitored every 15 min for 24 h at 37 °C using an Infinite M Plex plate reader (Tecan LifeScience) and orbital shaking with 1 mm amplitude.Peptides were tested in three independent experiments.

Figure 1 .
Figure 1.Under stress, Phd is degraded or no longer expressed releasing free Doc toxin which phosphorylates EF-Tu, resulting in Salmonella growth arrest and persister formation.Stapled Phd peptides mimic the Doc-binding domain of the antitoxin, inhibiting Doc and preventing persister formation.

Figure 2 .
Figure 2. Arginine scan of Phd 52−73 peptide analogue 1. (A) Peptide sequences, peptide net charge at pH 7, and ΔT m (thermal shift assay) values of the interaction of Doc STm to peptides Phd 52−73 , 1 and arginine-scan analogues 2−15.Hot-spot residues important for Doc binding and inhibition are marked in orange in the Phd 52−73 peptide.ΔT m values are shown as the mean ± SD.B = L-norleucine.(B) Thermal shift denaturation curves of free Doc STm at 5 μM (blue) and in the presence of 50 μM Phd 52−73 peptide 1 (green).The sigmoidal sections selected for the melting temperature fit are shown in black.

Figure 3 .
Figure 3. N-terminally arginine-spiked and hydrocarbon-stapled Phd 52−73 peptides with in vitro Doc inhibitory activity.(A) Peptide sequences, peptide net charge at pH 7, and ΔT m (thermal shift assay) values of the interaction of Doc STm to Phd 52−73 peptide 1 and analogues 16−29.ΔT m values are shown as mean ± SD.B = L-norleucine, X = (S)-2-(4-pentenyl)-alanine.(B) Schematic depiction of C-terminally i, i + 4 hydrocarbonstapled Phd 52−73 peptide.The shown peptide structure is derived from the homology model of Doc STm bound to Phd STm 52−73 , as previously described.(C) Dot blot detection of phosphorylated EF-Tu STm in the presence of Doc STm and Phd 52−73 peptides 1 and 16−21.Peptides were tested at eight concentrations, ranging from 10 μM to 5 nM (3-fold dilutions).Negative (EF-Tu STm 3 μM) and positive (EF-Tu STm 3 μM + Doc STm 1 μM) phosphorylation controls of the assay are shown on the right.(D) Dot blot detection of phosphorylated EF-Tu STm in the presence of Doc STm and Phd 52−73 peptides 1 and 22−29.Assay controls are shown on the right.

Figure 4 .
Figure 4. Bacterial uptake of first series of Phd 52−73 analogues in E. coli.Cellular uptake of 5 μM fluorescein (FAM)-labeled Phd 52−73 peptides 16−25 in E. coli MG1655 after incubation for 2 h at 37 °C as determined by flow cytometry.Data are shown as mean ± SD.Precipitation was observed for peptides FAM-20, FAM-21, and FAM-25 in the assay.

Figure 5 .
Figure 5. N-terminally hydrocarbon-stapled and C-terminally arginine-spiked Phd 52−73 peptides with in vitro Doc STm inhibitory activity.(A) Peptide sequences, peptide net charge at pH 7, and ΔT m (thermal shift assay) values of the interaction of Doc STm to Phd 52−73 peptide 1 and analogues 30−37.ΔT m values are shown as mean ± SD.B = L-norleucine, and X = (S)-2-(4-pentenyl)-alanine.(B) Dot blot detection of phosphorylated EF-Tu STm in the presence of Doc STm and Phd 52−73 peptides 1 and 30−33.Peptides were tested at eight concentrations, ranging from 10 μM to 5 nM (3-fold dilutions).Negative (EF-Tu STm 3 μM) and positive (EF-Tu STm 3 μM + Doc STm 1 μM) phosphorylation controls of the assay are shown on the right.(C) Dot blot detection of phosphorylated EF-Tu STm in the presence of Doc STm and Phd 52−73 peptides 1 and 34− 37. Assay controls are shown on the right.

Figure 6 .
Figure 6.Cellular uptake and Doc STm inhibition activity of selected peptides in S. typhimurium.Cellular uptake of (A) 2 and (B) 10 μM selected fluorescein (FAM)-labeled Phd 52−73 analogues in S. typhimurium after incubation for 4 or 22 h at RT as determined by flow cytometry.(C) Growth end points (measured as OD 600 ) of a S. typhimurium (14028) Δphd-doc:Km strain expressing Doc STm (pBAD33) after treatment with 2 or 10 μM selected peptides for 24 h at 37 °C.Untreated Doc STm -expressing S. typhimurium (negative control) as well as S. typhimurium (14028s) Δphddoc:Km strain coexpressing Doc STm and Phd STm 52−73 antitoxin peptide (positive control) were included in each measurement.Data are shown as means ± SD (D) growth curves of untreated Doc STm -expressing S. typhimurium (black/gray) and S. typhimurium coexpressing Doc STm and Phd STm 52−73 peptide (blue) from three independent experiments.(E,F) Growth curves of Doc STm -expressing S. typhimurium were treated with 10 μM peptide 1 (E) or peptide 36 (F) from three independent experiments.