Candesartan cilexetil disrupts methicillin-resistant Staphylococcus aureus membrane and potentiates gentamicin and polymyxin B activity

Nagendran Tharmalingam; Robert Wilson Kovacs; Suelen Scarpa de Mello; Philip Rupert Baldwin; Harikrishna Sekar Jayanthan; Kulandaisamy Arulsamy; Rajmohan Rajmuthiah; Fernanda Cristina Possamai Rossatto; Katherine E Manz; Joseph A DeGiorgis; Orlando Acevedo; Michael S Gilmore; Steven J Ludtke; Kurt D Pennell; Frederick M Ausubel; Eleftherios Mylonakis

doi:10.1038/s41467-026-70173-0

. 2026 Mar 15;17:4012. doi: 10.1038/s41467-026-70173-0

Candesartan cilexetil disrupts methicillin-resistant Staphylococcus aureus membrane and potentiates gentamicin and polymyxin B activity

Nagendran Tharmalingam ^1,^2,³, Robert Wilson Kovacs ^4,^#, Suelen Scarpa de Mello ^5,^6,^#, Philip Rupert Baldwin ^7,^#, Harikrishna Sekar Jayanthan ⁸, Kulandaisamy Arulsamy ⁹, Rajmohan Rajmuthiah ³, Fernanda Cristina Possamai Rossatto ³, Katherine E Manz ¹⁰, Joseph A DeGiorgis ¹¹, Orlando Acevedo ⁴, Michael S Gilmore ^5,⁶, Steven J Ludtke ⁷, Kurt D Pennell ¹⁰, Frederick M Ausubel ^5,¹², Eleftherios Mylonakis ^1,^2,^3,^✉

PMCID: PMC13136397 PMID: 41833947

Abstract

Methicillin-resistant Staphylococcus aureus (MRSA) is a leading cause of bacterial infections, but treatment options are limited due to MRSA multi-drug resistance. The anti-hypertensive drug Candesartan cilexetil (CC) exhibits potent anti-MRSA activity. It permeabilizes the membrane of MRSA cells and potentiates the activity of aminoglycoside antibiotics. Here, we used a variety of methods to elucidate the mechanism by which CC disrupts membrane homeostasis. We show that CC binds to bilayer lipid molecules, decreases membrane fluidity, down-regulates cell membrane and cell wall related genes and related metabolites, and decreases C20 fatty acids (C20:0). Decreasing C20:0 fatty acids confers CC- resistance, which can be reversed by C20:0 supplementation. Structural activity relationship analysis shows that the tetrazole ring and ester carbonic acid of CC are critical for antibacterial activity. Finally, CC reduces MRSA-MW2 replication in a murine MRSA abscess model, supporting a potential role of CC as a lead antimicrobial compound/potentiator against MRSA.

Subject terms: Antibiotics, Drug discovery and development, Antimicrobial resistance

In this work, authors show that candesartan cilexetil acts as a potent antibiotic potentiator, disrupting Staphylococcus aureus membranes and markedly enhancing the activity of clinically used antibiotics against drug-resistant infections.

Introduction

Staphylococcus aureus is a leading cause of hospital- and community-acquired infections^1–4. The broad repertoire of S. aureus virulence factors contributes to the wide spectrum of clinical manifestations and infection types, while persistence and tolerance mechanisms allow the pathogen to survive or continue limited growth in the presence of antibiotics, complicating treatment^5,6. Increasing levels of antibiotic resistance further limits treatment options for S. aureus. The Centers for Disease Control and Prevention estimated that methicillin-resistant S. aureus (MRSA) is associated with more than 300,000 life-threatening infections and more than 10,000 fatalities per year in the United States⁷.

The high cost of antibiotic development has engendered interest in repurposing existing Food and Drug Administration (FDA)-approved medications to treat bacterial infections. Our laboratory is particularly interested in antibacterial agents that affect key physical properties of the bacterial membrane, such as curvature and fluidity, thereby disrupting membrane architecture and compromising its functional integrity^8–15. Membrane-disrupting compounds can also interact with membrane-embedded proteins or alter the proton motive force¹⁶. The particular effects of membrane-disrupting compounds are highly dependent upon the specific composition of the lipid bilayer¹⁷, key components of which include cardiolipins and negatively charged phospholipids, such as phosphatidylglycerol¹⁸. In MRSA, alterations in membrane lipid composition, particularly reduced phosphatidylglycerol or altered cardiolipin distribution, have been associated with decreased drug penetration and increased tolerance to membrane-active antibiotics. These adaptive changes stabilize the bilayer and create a barrier to antibiotic entry, thereby contributing directly to drug resistance^19–21. Thus, targeting membrane lipid composition and fluidity may represent an effective strategy to overcome intrinsic and acquired resistance mechanisms in MRSA.

Here we show, that candesartan cilexetil (CC), an angiotensin II receptor blocker (ARB) used for treating hypertension²², exhibited anti-staphylococcal activity in an MRSA-C. elegans high-throughput screen of ~82,000 synthetic compounds²³, consistent with published work by Xu et al.²⁴. Here, we explored the mechanism of action of CC using a variety of assays, microscopic techniques including cryo- computed tomography, multi-OMICS analysis, as well as in silico all-atom molecular dynamics (MD) simulations, and gas chromatography-mass spectrometry (GC-MS) analysis. We also evaluated the efficacy of CC in a murine S. aureus infection model. A key finding of our work is that CC treatment unexpectedly decreased S. aureus membrane fluidity and reduced the levels of C20 fatty acids (C20:0). Importantly, decreasing the levels of C20:0 fatty acids by cerulenin treatment greatly decreased CC susceptibility and supplementation of the cerulenin-treated cells with C20:0 fatty acids significantly restored CC susceptibility, suggesting that C20:0 fatty acids play an important mechanistic role in determining the susceptibility of MRSA to CC treatment. Our data suggest that CC has potential for development as S. aureus antibiotic potentiator.

Results

CC inhibits the growth of MRSA

The Microsource-1 library of chemical compounds was among the chemical libraries that our laboratory previously screened for anti-MRSA compounds using a C. elegans–MRSA infection model^11,23. An antimicrobial compound identified in this screen was the antihypertensive drug Candesartan Cilexetil (CC; Fig. 1A)²³.

Fig. 1 — A Structure of CC and candesartan. Killing kinetics of CC against exponential (B) or stationary phase cells (C). Data are presented as mean values ± SD. P-value for CC against exponential phase (B; 8x, 4x, 2x, 1x or gentamicin vs untreated) is less than p < 0.0001 and for stationary phase (C; 8x, 4x vs control) is p < 0.0001; however no-significance was observed against gentamicin treatment; 2-way-ANOVA. D CC inhibits biofilm formation and disrupts mature biofilm. Data are presented as mean values ± SD. P-value for CC (8x, 4x, 2x, 1x, 0.5x, or tetracycline vs untreated) is less than p < 0.0001; 2-way-ANOVA. E CC permeabilizes exponentially growing MRSA-MW2 membranes to Sytox Green (ex=435 nm/em=530 nm). Data are presented as mean values ± SD. P value for CC (8x, 4x, 2x, 1x vs untreated) is less than p < 0.0001; 2-way-ANOVA). F CC decreases MRSA-MW2 membrane potential demonstrated using DiSC₃(5) assay and cellular fluorescence was measured periodically (ex =620 nm/em=660 nm; CCCP-control). Data are presented as mean values ± SD. P-value for CC (2x, 1x, or valinomycin or gramicidin vs untreated) is less than p < 0.0001; 2-way-ANOVA). G CC causes a rapid, dose-dependent increase in intracellular pH in MRSA-MW2 cells (ex = 440, em = 490 nm). Experimental pH values (n = 3) were plotted on a calibrated standardized assay graph, using the fluorescence units of known pH solutions. H All atom molecular dynamic (MD) simulations showed that CC (yellow) rapidly perturbs simulated *S. aureus* membranes on contact (DOPC: DOPG (7:3) (1,2-dioleoyl-sn-glycero-3-phosphocholine/glycerol)). All simulations were repeated in triplicate. I SEM images of MRSA-MW2 cells treated with CC. The assay was done in triplicate and representative micrograph is shown. J AFM image of MRSA-MW2 cells. K TEM image with yellow arrows indicate deformations/mesosome-like structural deformities/membrane-deformation/cell-shrinkage/condensation. The assay was done in triplicate and representative micrograph is shown. L Cryo-computed tomography of 3D-projections of untreated MRSA-MW2 cells with intact membrane (green) or CC-treated (blue; 10xMIC). The segmented overlay of membrane damage is shown in the bottom row. All graphs (B–G) show representative data with means ± SD for biologically independent samples (n = 3). Source data are provided as a Source Data file.

We found that the minimal inhibitory concentration (MIC) of CC is 8 mg/L against MRSA-MW2 and several S. aureus clinical strains and 16 mg/L against Enterococcus faecium (E007) (Supplementary Tables 1 and 2). CC is also active against vancomycin-intermediate resistant strains (Supplementary Table 3). In humans, CC is a prodrug²². The active moiety, candesartan (Fig. 1A-inner panel), which is generated in gastric epithelial cells²⁵, did not exhibit antibacterial activity (Supplementary Table 4). Killing kinetics analysis showed that CC (≥2xMIC) was bactericidal against exponential-phase MRSA-MW2 cells (5-log₁₀-CFU/mL reduction to limit of detection (LOD), Fig. 1B) and stationary-phase MRSA-MW2 cells with a persister-like phenotype¹¹ (4-log₁₀-CFU reduction over a 4-h treatment period at 4xMIC and 8xMIC, Fig. 1C). Notably, different inocula were used to represent distinct bacterial growth phases: ~10⁷ CFU/ml for log-phase cultures (actively dividing conditions; Fig. 1B) and ~10⁸ CFU/ml for stationary-phase cultures (metabolically less active and tolerant to antibiotics; Fig. 1C). In contrast, gentamicin (25xMIC) had no effect on persisters (Fig. 1C), consistent with published results^9,11. Biofilm studies showed that CC treatment resulted in a dose-dependent reduction in biofilm formation and a decrease in metabolically active cells within mature biofilms, as assessed by Alamar Blue fluorescence (Fig. 1D). We found that CC inhibited bacterial biofilm formation at concentrations as low as 0.5xMIC (4 mg/L) (Fig. 1D-right) and disrupted mature biofilms at concentrations as low as 2xMIC (16 mg/L, Fig. 1D-left) as demonstrated by using Alamar Blue to monitor the levels of metabolically active cells. While the reduction at sub-MIC concentrations was modest, these results indicate that CC can affect biofilm metabolic activity.

CC disrupts the S. aureus membrane

Consistent with its rapid killing kinetics, CC (≥0.5xMIC) permeabilized MRSA-MW2 to Sytox Green (Fig. 1E). Experiments using the fluorescent probe DiSC₃(5) showed that 1xMIC and 2xMIC CC decreased MRSA-MW2 membrane potential similar to the positive control carbonyl cyanide m-chlorophenyl hydrazone (CCCP) (Fig. 1F). Unexpectedly, however, experiments with the fluorescent probe BCECF-AM showed that CC caused a rapid dose-dependent increase in intracellular pH in contrast with the positive control CCCP, which decreased intracellular pH (Fig. 1G). These latter results suggest that rapid intracellular alkalization may be a key feature underlying the mechanism by which CC disrupts the S. aureus membrane.

Unbiased all-atom molecular dynamic (MD) simulations (500 ns) of equilibrated bilayer-drug interactions using a model DOPC:DOPG (7:3) (1,2-dioleoyl-sn-glycero-3-phosphocholine/glycerol) lipid bilayer, showed that the tetrazole group of CC anchors the drug to the membrane surface where it binds to hydrophilic head groups of membrane phospholipids (glycerol moiety of DOPG), allowing CC to penetrate and rapidly perturb the membrane (Fig. 1H, and Supplementary movie S1). The all-atom MD simulations of CC interacting with a model S. aureus membrane were carried out independently at Vanderbilt University and the University of Miami with similar results (compare Fig. 1H with Supplementary Fig. 1). Although candesartan also penetrates DOPC/DOPG-simulated S. aureus membranes in these simulations, it does not interact with membrane lipids and does not appear to perturb the membrane (Supplementary Fig. 1F).

Scanning electron microscopy (SEM) studies (Fig. 1I) showed normal morphology in DMSO but that multiple blisters formed on the MRSA-MW2 cell surface after CC treatment at 4xMIC, 2xMIC, 1xMIC (right to left, respectively). Atomic force microscopy (AFM) confirmed the SEM observation that CC at 10xMIC caused changes in membrane topography and mechanical properties followed by cell-leakage (Fig. 1J). Transmission electron microscopy (TEM) studies showed that CC treatment caused morphological changes, including mesosome-like structural deformities at 8xMIC or membrane deformation and cell shrinkage or condensation at 10xMIC (Fig. 1K). Cryo-computed tomography was used to generate a 3D model of CC-damaged membranes as well as mesosome-like formation (Fig. 1L). Cryo-computed tomography with 3D segmentation revealed that, compared to untreated (Supplementary Movie S2), CC treatment caused membrane-associated irregularities such as membrane thinning, disruption of the bilayer structure, and the formation of perforations or gaps (Fig. 1L, and Supplementary Movies S3 and S4), consistent with MDS analysis.

Comprehensive molecular multi-OMICS analysis

To gain further insights into the MOA by which CC kills S. aureus and to identify potential CC targets, we carried out comprehensive transcriptomic, metabolomic, and lipidomic analyzes. Using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, an integrated analysis of significantly downregulated genes, metabolites, and lipids from multi-omics data identified several key metabolic pathways that were altered by treatment of MRSA-MW2 with a sublethal concentration of CC (4 mg/L), including pathways leading to fatty acid precursors, as well as cell membrane and cell wall components (Fig. 2A). Volcano plots of the RNA-seq data revealed upregulated (left) and downregulated (right) genes (Supplementary Fig. 2A) while lollipop plots highlighted the number of genes modulated in particular pathways (Supplementary Fig. 2B). Downregulated genes included ones involved in fatty acid synthesis (lip1, lip2), cell wall biosynthesis (wecB, walk, glmU, murA, murB, murC), and cell membrane integrity (fakB2, uhpT) (Fig. 2B). In addition, the transcriptomics analysis showed down-regulation of genes responsible for purine biosynthesis (purN, purM, purF, purH), amino acid metabolism (bioA, bioB, ilvC), stress responses (arcA, arcD, sigS), and energy metabolism (atpA, kdpA), indicating that CC has a broad impact on cellular physiology (Fig. 2B, and Supplementary Fig. 2C, D).

Metabolomic studies of MRSA-MW2 treated with 4 mg/L CC identified several enriched pathways (Supplementary Fig. 2E-right) and a Volcano plot (Supplementary Fig. 2E left) shows metabolites significantly altered by CC treatment, with a fold change (FC) > 2 (or <0.5 for decreases) and a p-value < 0.05 (Supplementary Fig. 2E-left). The metabolomic data are consistent with the transcriptomics data, which showed that CC causes downregulation of metabolites such as mevalonic acid (likely influenced by menD, and menH in menaquinone biosynthesis) and UDP-N-acetylglucosamine (associated with wecA, glmU, murA, murB, murC) (Fig. 2C). A heatmap of the top 25 affected metabolites (Fig. 2C) shows distinct metabolic shifts driven by CC, with downregulated metabolites including guanosine, mevalonic acid, L-lactic acid, as well as intermediates in glycolysis and pentose phosphate pathways (G6P, fructose-1,6-diphosphate, sedoheptulose 7-phosphate, D-ribose-5-phosphate). These metabolites are crucial for energy metabolism, nucleotide biosynthesis, and stress responses.

PLS-DA analysis of the lipidomics data detected an average of 132 lipids, with full separation between control and treated groups, and 20 lipids showing significant differences (p < 0.05) (Supplementary Fig. 2F-right) and a volcano plot highlights lipids significantly altered by CC treatment, with a fold change (FC) > 2 (or <0.5 for decreases) and a p-value < 0.05 (Supplementary Fig. 2F-left). We found that reduction of key membrane lipids, including phosphatidylglycerol (PG(15:0_15:0), PG(15:0_16:0), PG(15:0_20:2), PG(15:0_17:0), PG(14:0_18:0)) and phosphatidylethanolamine (PE(15:0_16:0)) (Fig. 2D), which are crucial for maintaining membrane integrity and function²⁶. This reduction may be linked to the metabolic activity of ilvC, bioA, bioB, and bioD, although the precise mechanism remains to be fully elucidated. Lipidomic profiling identified 15 lipids (Supplementary Fig. 2E, and Fig. 2D) and two key lipids, phosphatidylglycerol (PG) (15:0_15:0) and diacylglycerol (DAG) (15:0_17:0), that were significantly decreased. Heatmaps of the top 25 lipids (Fig. 2D) illustrated distinct lipid expression patterns, highlighting a reduction in specific phospholipids (PG(15:0_15:0), PG(15:0_16:0), PE(15:0_20:4)) essential for membrane structure and function. These findings suggest altered membrane integrity and lipid metabolism following CC treatment.

The multi-omics data indicate that CC disrupts S. aureus cell wall and membrane integrity and suggest that CC, even at concentrations below its MIC, may have deleterious effects on the viability and pathogenicity of S. aureus, in addition to its ability to rapidly kill S. aureus cells. The downregulation of genes critical for the synthesis of cell wall precursors and membrane components appears to lead to a reduction in metabolites that are essential for these processes. Lipidomic data further confirms the impact on membrane structure by showing a decrease in key phospholipids needed for membrane integrity. This disruption in cellular structure is likely to compromise cell function, and the findings suggest that targeting these pathways may identify additional therapeutic strategies.

Transcriptomics showed down-regulation of wecA, which is involved in the synthesis of cell wall constituents and metabolomics analysis showed that CC treatment reduced levels of UDP-NAG, a key precursor in peptidoglycan biosynthesis. Lipidomics further revealed decreases in phospholipid abundance and changes in fatty acid composition, consistent with impaired cell-wall and membrane homeostasis. To test whether depletion of cell-wall precursors contributes to CC-elicited membrane defects, we supplemented cultures with UDP-N-acetylglucosamine (UDP-NAG) or N-acetylglucosamine (NAG). CC-treated MRSA-MW2 shows impaired growth and membrane permeability to CC. However, both supplements partially rescued CC-inhibited growth (Fig. 2E; p < 0.0001; 2-way Analysis of Variance (ANOVA)) and reduced Sytox Green permeability (Fig. 2F, G; p < 0.0001; 2-way ANOVA), supporting a functional link between CC-mediated depletion of UDP-NAG and loss of cell-wall/membrane integrity (Fig. 2H).

CC synergize with gentamicin, kanamycin, streptomycin, & polymyxin B

Testing various classes of antibiotics (including beta-lactams, aminoglycosides, tetracycline, macrolides, glycopeptides, and cyclic peptides) using checkerboard MIC analysis^13,27, showed that CC synergizes well with aminoglycosides (gentamicin, kanamycin, streptomycin), as well as with the cyclic peptide (polymyxin B) with respect to inhibiting the growth of MRSA-MW2 (Fractional Inhibitory Concentration Index ≤ 0.5, Fig. 3A). Killing kinetics analysis demonstrated that CC+gentamicin (1xMIC for both) eradicated MRSA-MW2 logarithmic cells in 60 min (Fig. 3B), whereas CC alone at 1xMIC did not eradicate logarithmic-cells and gentamicin alone at 1xMIC eradicated them only after 240 min. Furthermore, gentamicin (10xMIC) + CC (2xMIC) decreased a persister cell population by 4-log₁₀-CFU, whereas CC or gentamicin alone at these concentrations had no effect (Fig. 3C). Similar to CC + gentamicin synergism, CC at 1xMIC + polymixin B at 1xMIC eradicated log phase MRSA-MW2 within 30 min, whereas CC (1xMIC) or polymyxin B (1xMIC) alone exhibited modest killing activity (Fig. 3B). Likewise, CC (2xMIC) + polymyxin B (1xMIC) enhanced the killing of stationary-phase MRSA-MW2 cells (1.5-log₁₀-CFU) compared to CC (2xMIC) or polymyxin B (1xMIC) alone which had no observable effect (Fig. 3C).

Fig. 3 — A CC synergizes with aminoglycosides or lipopeptide (synergy: FICI ≤ 0.5; no interaction: 0.5 < FICI ≤ 4; antagonism: FICI > 4). Individual biologically independent experiments were carried out in triplicate and representative results are shown. B CC is a potent bactericidal agent when used in combination with gentamicin or polymyxin B against exponential-phase MRSA-MW2 cells. Total viable counts were quantified using the spot-plate method. The graphs show data sharing the same DMSO control with means ± SD for n = 3 biologically independent samples. P-value (CC vs untreated) is less than p < 0.0001; 2-way-ANOVA. C CC kills MRSA-MW2 persister cells when combined with gentamicin or polymyxin B. Stationary-phase cells were treated with CC alone or combined with gentamicin or polymyxin B. Drug-treated cells were aliquoted, serially diluted, and the total viable count was quantified. The graphs show data sharing the same DMSO with means ± SD for n = 3 biologically independent samples. P-value (CC vs untreated) is less than p < 0.0005; 2-way-ANOVA. D CC in combination with polymyxin B causes membrane permeation. CC (yellow) combined with polymyxin B (red) immediately damages simulated *S. aureus* membranes. These MD simulation studies of the CC+polymyxin B complexes interacting with a simulated *S. aureus* bacterial membrane bilayer. PMF calculations revealed rapid insertion of the complex and synchronous membrane disruption. The simulations were repeated three times with similar results, and a representative simulation is shown. E Topography analysis of MRSA-MW2 using atomic force microscopy. MRSA-MW2 cells were treated with a combination of CC+ polymyxin B or CC alone or polymyxin B alone at 10x MIC for 4 h. The images show structural deformity in cells treated with the combination of CC+ polymixin B or CC compared to DMSO (Fig. 1J). The AFM data in Fig. 1J and in this Figure were collected in the same experiment. All graphs (B, C) show representative data with means ± SD for biologically independent samples (n = 3). Source data are provided as a Source Data file.

Since both CC and aminoglycosides disrupt membrane homeostasis^28,29, we tested CC+aminoglycoside combinations using DiSC₃(5) to determine the effect of the combination on MRSA-MW2 membrane potential. The combinations decreased the membrane potential more rapidly than CC alone or the aminoglycosides alone. Valinomycin/CCCP³⁰ was used as a positive control (Supplementary Fig. 3A–D). Interestingly, MD analyzes of CC+polymyxin B (Fig. 3D) showed that CC and polymyxin B spontaneously formed defined and stable (< −5 kcal mol⁻¹ ΔG_complex) bi- and tri-molecular complexes at 1:1, 1:2 and 2:1 ratios of CC to polymyxin B moieties. Further simulations of the 1:1 complex interacting with a simulated S. aureus bilayer revealed similar antimicrobial behavior to polymyxin B, including similar free energy minima of penetration within the membrane’s headgroup region (Fig. 3D, and Supplementary Movie S5). However, the added bulk from the complexation of CC (2:1) enabled immediate large-scale membrane deformation during membrane permeabilization, particularly at the distal face to the drug penetration (the side of the membrane away from where the drug is entering), which was not observed in the simulations with CC or polymyxin B alone. Analysis of the 1:2 complex revealed a slightly shallower penetration depth than polymyxin alone or the 1:1 complex (Fig. 3D, and Supplementary Movie S6). The 1:2 complex caused 170% more membrane disruption at the proximal face and 270% more disruption at the distal face compared to the 1:1 complex, indicating greater long-range effects on bilayer integrity. The dramatic increase in deformation at the distal face indicates long-range reorganization of the membrane, reducing the bilayer integrity far away from the penetration site. These findings point to an interesting steric-driven, on-contact mechanism for membrane disruption, occurring on the nanosecond and single-molecule scales. In contrast, the antibacterial action of polymyxin B involves the gradual amphiphilic aggregation of several penetrated molecules (Supplementary Movie S7)³¹. This proposed mechanism is consistent with the combinatorial action of CC + polymyxin enhancing cellular damage using AFM (Figs. 1J and 3E) or membrane permeability to Sytox Green (Supplementary Fig. 3E–H). Note that the AFM data in Figs. 1J and 3E were collected in the same experiment.

Mechanism by which CC disrupts the S. aureus membrane

To further explore the mechanism by which CC disrupts the S. aureus membrane, we quantified membrane fluidity using the Laurdan generalized polarization (L-GP) metric in which a decrease in L-GP correlates with an increase in membrane fluidity³². Unexpectedly, we found that CC at 32 mg/L (4xMIC) or 64 mg/L (8xMIC) caused a significant increase in the L-GP in MRSA-MW2-treated cells (Fig. 4A), suggesting that CC makes the membrane more rigid, potentially retarding the rate of cell division³³.

Fig. 4 — A CC decreases MRSA-MW2 membrane fluidity demonstrated using the formula=Laurdan-GP (I₄₄₀ − I₄₉₀)/(I₄₄₀ + I₄₉₀) when excited at 350 nm. Data are presented as mean values ± SD. P-value for CC (8x, 4x) is less <0.0001(****) or BA is 0.09(**) against untreated using one-way-ANOVA. B–E Exogenous lipids affect the interaction of CC (8x, 4x) with the MRSA-MW2 membrane. MRSA-MW2+stearic acid (P-value of BA, CC (8x, 4x) and polymyxin B is 0.01(*), 0.024(*), 0.89, 0.02(**) respectively) or oleic acid (0.05)(*), <0.0001(****), ns respectively; (One-way-ANOVA) were used to determine Laurdan GP and Sytox Green uptake (D, E; 2-way-ANOVA). F–J Exogenous lipids affect the bactericidal activity of CC. Killing kinetics of CC was evaluated in exponentially growing control cells (F) or cells+stearic acid (G), or cells+oleic acid (H). Next, we tested the lipid biosynthesis inhibitor-cerulenin (J) was compared to untreated (I). K–O CC affects membrane fluidity. Laurdan microscopy of (low power magnification; scale-bar unavailable; check method) dye-treated MRSA-MW2 cells were embedded in an agarose-gel on a microscope slide with either one of CC-8xMIC-O; 4xMIC-N; DMSO-K; CCCP-H, vancomycin-L. The assay was repeated twice, and representative data are shown (M). P–Y Membrane C20:0 levels positively correlated with susceptibility to CC. P Log-MRSA-MW2 cells were treated with CC, cerulenin, or vancomycin. Total lipids were converted to fatty acid methyl esters, and analyzed using GC-MS. A representative graph is shown from three biological replicates. Q–S Cerulenin decreases membrane permeability. MRSA-MW2 cells were grown exponentially in the presence of cerulenin (R), cerulenin+C20:0 fatty acid (S), or no treatment (Q), and added with Sytox Green (ex=435 nm, em=530 nm) to measure cellular fluorescence. T–X MRSA-MW2 cells were grown exponentially in the presence of cerulenin (U), cerulenin+C20:0 fatty acid (V), CC at sub-MIC (W), or CC at sub-MIC + C20:0 fatty acid (X) and killing kinetics were assayed. Y CC binds to C20:0 fatty acid. The C20:0 fatty acid+BODIPY TR cadaverine fluorescent probe+Tris buffer tested for Bodipy displacement demonstrated via fluorescence spectrophotometer. All graphs (A–J, Q–Y) show representative data with means ± SD for biologically independent samples (n = 3). Source data are provided as a Source Data file.

Reasoning that CC may target membrane lipids, we studied the effect of CC on MRSA-MW2 membranes containing different lipid compositions. Because bacterial membrane lipid composition can be affected by growing bacteria in the presence of different lipids³⁴, we cultured MRSA-MW2 cells in the presence of 25 mg/L oleic acid (long-chain unsaturated fatty acid) or 25 mg/L stearic acid (long-chain saturated fatty acid), treated them with CC (4xMIC or 8xMIC), and measured membrane fluidity. Figure 4B shows that growth in stearic acid increased L-GP (increase in rigidity) and that this level of rigidity was modestly increased by treatment with CC. In contrast, Fig. 4C shows that growth in oleic acid increased membrane fluidity but that treatment with CC increased rigidity to about the same levels as in cells grown in the absence of long-chain fatty acids or the presence of stearic acid. Consistent with these results, we found that MRSA-MW2 cell growth in oleic acid (Fig. 4D and H) behaved like wild type cells (Fig. 4F), but that MRSA-MW2 cell growth in stearic acid decreased the ability of CC to permeabilize MW2 cells to Sytox Green (Fig. 4E) and decreased the susceptibility of MRSA-MW2 to CC-mediated killing (Fig. 4G). Fluorescence microscopy of Laurdan dye-treated MRSA-MW2 cells³⁵ revealed that dimethyl sulfoxide (DMSO, Fig. 4K) or vancomycin (Fig. 4L) or CCCP-treated cells (Fig. 4M) only exhibited fluorescence in the dividing septum, whereas CC-treated cells showed a fluorescence signal all over the membrane at 4xMIC (Fig. 4N) and 8xMIC (Fig. 4O).

To identify potential specific lipid targets of CC in the bacterial bilayer of MRSA-MW2 cells, we used cerulenin³⁶, which is known to modify bilayer fatty acid composition^37–39. MRSA-MW2 cells grown for 4 h with 40 mg/L cerulenin were dramatically more resistant to CC-mediated killing (Fig. 4J) than control cells (Fig. 4I). Since cerulenin modifies the bacterial membrane bilayer lipid composition, we hypothesized that the target of CC is either missing in these cells or another membrane lipid blocks CC activity. Hence, we analyzed the lipid compositions of untreated and cerulenin-treated MRSA-MW2 cells via GC-MS using the MIDI protocol⁴⁰. We also analyzed lipid composition of sub-MIC CC-treated MRSA-MW2 cells, reasoning from the “omics” data presented in Fig. 2 that such treatment may affect lipid biosynthetic pathways.

Cerulenin treatment showed a significant shift in the MRSA-MW2 membrane fatty acid profile, including a significant decrease in 20 carbon atom (C20:0) fatty acids (Fig. 4P), suggesting that C20:0 lipids plays a critical role in membrane susceptibility to CC. Consistent with this conclusion, cerulenin-treated cells were less susceptible to CC-mediated permeabilization (fivefold increase in CC-mediated Sytox Green uptake) compared to untreated cells (tenfold increase in CC-mediated Sytox Green uptake, Fig. 4Q, R). Moreover, we found that supplementing cerulenin-grown cells with arachidic acid, a C20:0 lipid, at least partially rescued the ability of CC to permeabilize the MRSA-MW2 cells (15-fold increase in CC-mediated Sytox Green uptake, Fig. 4S).

To provide additional data in support of the hypothesis that C20:0 lipids are required for CC susceptibility, we cultured the MRSA-MW2 cells in the presence of cerulenin with or without C20:0 supplementation. Killing kinetics analysis showed that compared to DMSO-treated cells (Fig. 4T), cerulenin-treated cells lacking C20:0 were less susceptible to CC (Fig. 4U) and that C20:0 supplementation partially (p < 0.001; 2-way ANOVA) restored the susceptibility of MRSA-MW2 cells to CC at 30 and 60 min (Fig. 4V). In addition, we found a similar pattern for treatment with sub-MIC levels of CC, with a significant decrease in C20:0 fatty acids in the MRSA-MW2 membrane (Fig. 4P), resulting in the cells being less susceptible to CC (Fig. 4W), similar to cerulenin-treated cells. As was the case with cerulenin, C20:0 supplementation of the sub-MIC CC-treated cells, partially (p < 0.001; 2-way ANOVA) restored CC-sensitivity at 60 min (Fig. 4X). Finally, consistent with the hypothesis that CC directly interacts with C20:0 fatty acids, we found that CC displaces the lipid-interacting fluorescent probe BODIPY TR cadaverine⁴¹ coupled to C20:0 (Fig. 4Y). Using the BODIPY TR cadaverine probe⁴¹, we found that CC interacts with a variety of lipid-containing membrane components including lipoteichoic acid, lipopolysaccharide, L-alpha phosphatidylglycerol, 1,2-dipalmitoyl-sn-glycero-3-phospho-rac-(1-glycerol), and cardiolipin (Supplementary Fig. 4A–E).

We next tested different pH levels to assess how variations in environmental acidity or alkalinity influence antibacterial activity of CC and membrane interactions, as pH can significantly impact drug efficacy, bacterial physiology, and membrane properties^42–44. Specifically, we found that CC significantly enhanced membrane rigidity only at pH 7.0 and pH 8.0, but permeabilized the membrane at both alkaline and mildly acidic conditions (pH 6), but not at lower pH levels (Supplementary Fig. 5A, B). The anti-S. aureus activity of CC was highest at pH 4 with an MIC of 2 µg/mL, even though CC did not permeabilize the membrane at pH 4, and CC retained significant antibacterial activity from pH 5 to 9 (Supplementary Table 5). At pH 4.0, the protonation of CC may enhance its interaction with intracellular targets, increasing its potency without significant membrane disruption. CC was bactericidal at pH 7 and 8 but bacteriostatic at pH <6.0 and >8.0 (Supplementary Fig. 5C), correlating with its ability to enhance membrane rigidity at neutral to alkaline pH and permeabilize cells at pH 6.0 and higher (Supplementary Fig. 5A–C). Interestingly, despite the absence of detectable membrane permeabilization at pH 4, CC displayed its highest potency (MIC = 2 mg/L) under these acidic conditions. This suggests that CC may exert antibacterial effects through membrane-independent mechanisms at low pH. Together, these observations suggest that the mechanisms by which CC kills S. aureus may be complicated and involves more than simple physical disruption of the membrane.

Isolation of CC-resistant mutants

Our observations showing that a cerulenin-induced decrease in C20:0 membrane lipids results in significantly enhanced resistance to CC led us to hypothesize that MRSA MW2 mutants with decreased C20:0 levels should exhibit at least some level of resistance to CC-mediated killing. To isolate such C20:0 mutants and in general to identify MRSA-MW2 CC-resistant mutants that might help identify the target(s) of CC, we carried out a 50-day serial passage experiment using sub-lethal concentrations of CC to identify CC-resistant mutants as described in previous publications from our laboratory^9,11. The MIC of CC remained unchanged until about day 25, with an approximately 32-fold increase by day 50 (Supplementary Fig. 4F). Whole-genome sequencing of purified clones from different days of passage identified single-nucleotide polymorphisms (SNP) in genes purR, folC, farR, and tarS (Supplementary Table 6). One particular CC-resistant mutant clone carrying mutations in purR, folC, farR, and tarS (strain MRSA-MW2-S50 from day 50 of serial passage (Supplementary Tables 6)) was chosen for further analysis (as described in the section below on structure-activity relationship studies). Interestingly, however, MRSA-MW2-S50 did not exhibit decreased levels of C20:0 lipids (Supplementary Fig. 4G), suggesting that a variety of mechanisms could lead to CC-resistance. In this regard, the observation that the CC-resistant strain MRSA-MW2-S50 contained a mutation in the tarS gene, which is required for cell-wall teichoic acid biosynthesis, prompted us to test whether a well-studied S. arueus tar mutant exhibits CC resistance. Specifically, we tested a previously constructed deletion of the tarO gene, an isogenic derivative of the tarO mutant carrying wild-type tarO expressed on plasmid pTP63, and S. aureus strain Newman, the parent of the tarO mutants. As shown in Supplementary Fig. 4H, the MICs of the Newman strain, the tarO mutant, and the complemented mutants were unexpectedly 8 mg/L, 4 mg/L and 8 mg/L, respectively. That is, the tarO mutant was more susceptible to CC than the parent strain, in contrast to our expectation.

Since the ΔtarO mutant was more sensitive to CC, it is unlikely that TarO or TarS are direct molecular targets of the compound. Instead, these results indicate that TarO–TarS–dependent pathways preserve cell wall architecture and mitigate CC-induced stress. Mutations in tarS may alter WTA composition or surface charge, potentially reducing CC binding or penetration. Collectively, these findings suggest that CC activity is modulated by the integrity and glycosylation status of wall teichoic acids.

On the basis of these tarO data and the fact that our passaging experiment did not yield mutants with reduced levels of C20:0 fatty acids, we decided that comprehensive genetic analysis of CC resistance is likely to be complicated and therefore beyond the scope of this current publication.

Structure-activity relationship (SAR) studies

We sought to carry out a limited structure-activity relationship (SAR) study using commercially available analogs of CC as well as a few custom synthesized CC analogs. First, the Microsource 1 library that was used in the C. elegans high-throughput screening platform that identified CC contained several other angiotensin receptor blocking compounds with structures related to CC, including irbesartan, losartan, olmesartan medoxomil, telmisartan, and valsartan. These CC-related compounds possess one or more combinations of benzimidazole, biphenyl or tetrazole moieties. However, none of these compounds were identified in the screen for compounds that blocked the ability of MRSA-MW2 to kill C. elegans and none of them exhibited anti-MRSA activity (Supplementary Fig. 6A).

Second, we tested a limited set of commercially available tetrazole-containing molecules, but none of these exhibited an anti-MRSA activity (Supplementary Fig. 6B), indicating that an intact tetrazole ring alone is insufficient for antimicrobial activity. Third, we tested five commercial (Supplementary Fig. 6C) and six custom-synthesized analogs of CC (Supplementary Fig. 6D) that have benzimidazole, biphenyl, tetrazole, and cilexetil moieties. Among the five commercially available CC-analogs (Supplementary Fig. 6C), two have unmodified tetrazole rings, and three have modified tetrazole rings. The two CC analogs with unmodified tetrazole rings exhibited antibacterial activity against MRSA-MW2 with an MIC of 8 mg/L, the same as CC, (Supplementary Fig. 7) and interestingly were also active against the CC-resistant mutant MRSA-MW2-S50 generated by serial passage (Supplementary Fig. 7A). These two active compounds readily penetrated a simulated S. aureus membrane (Supplementary Fig. 7B), permeabilized MRSA-MW2 to Sytox Green (Supplementary Fig. 7C), and exhibited toxicity profiles against mammalian cells similar to CC (Supplementary Fig. 7D, E). In contrast, the three CC analogs with modified tetrazole rings (Supplementary Fig. 8) lost their antibacterial activity (Supplementary Fig. 8A), failed to penetrate the simulated bilayer (Supplementary Fig. 8B), did not permeabilize MRSA-MW2 (Supplementary Fig. 8C), and had no effect on mammalian cell survival (Supplementary Fig. 8D, E). However, an intact tetrazole ring alone is insufficient for antimicrobial activity since candesartan has an unmodified tetrazole but does not block cell growth or permeabilize MRSA-MW2 (Supplementary Fig. 8A). Consistent with these findings, MD analysis showed that the unmodified tetrazole containing CC analogs permeated a simulated S. aureus membrane with a deep free energy minimum of 26–20 Å from the bilayer center (Supplementary Fig. 9A). In contrast, the analogs with modified tetrazoles do not interact with the membrane in an energetically favorable manner and do not permeate the membrane (Supplementary Fig. 9A). Additional MD data shown in Supplementary Fig. 9 and discussed in the Supplementary Fig. 9 legend are also consistent with the biochemical and physiological data described in this paragraph.

Third, we synthesized six CC analogs with an intact tetrazole ring but modified cilexetil moieties to better understand the structural features of the cilexetil moiety that confer antibacterial activity (Fig. 5). In compound-1, replacing the cyclohexyl ring with isopropyl (Fig. 5A) (Supplementary Information 1) significantly increased the anti-MRSA-MW2 MIC from 8 to 64 mg/L (Fig. 5A). MD simulations showed compound 1 did not permeabilize the membrane (Supplementary Movie S8) and that compound 4 has poor permeability (Supplementary Movie S10). However, compounds 5, 6, 7 permeated the membrane (Fig. 5B, and Supplementary Movies S11–13). The deviations in average bilayer penetration distances Δd_pen of the analogs’ hydrophilic tetrazole groups and hydrophobic isopropyl/cyclohexyl/oxanyl groups (Supplementary Fig. 10A), and penetration distance of analogs and mean distances from the bilayer center are shown in Supplementary Fig. 10B, C. Compound 1 also did not cause MRSA-MW2 membrane permeability (Fig. 5D) and exhibited less mammalian cell toxicity (up to 64 mg/L) than the other compounds that were tested (Fig. 5E). On the other hand, compound-1 still exhibited significant bactericidal activity at 4× the MIC, which is likely due to non-specific toxicity given the high concentration (Fig. 5C). In compound 2, replacing the cyclohexyl ring with a pyran ring but maintaining the same linker length (Supplementary Information 2) caused a complete loss of anti-MRSA activity (MIC > 128 mg/L), demonstrating that the cyclohexyl ring is also critical for antibacterial activity. In compounds-4, 5, 6 and 7, retaining the cyclohexyl ring (Fig. 5A) with different length linkages (Supplementary Information 3–5) yielded compounds with the same to moderately reduced antibacterial activity as CC (MIC of 8–32 mg/L (Fig. 5A) and bactericidal activity (Fig. 5C), which largely correlated with their ability to permeabilize MRSA-MW2 membranes (Fig. 5D) and their toxicity to HepG2 cells (Fig. 5E). Interestingly, compounds 5 and 6 were moderately active against the MRSA-MW2 CC-resistant MW2-S50 mutant (Fig. 5A). In compound-7, removing two ester groups in the cilexetil linker (Supplementary Information 6) did not affect antibacterial activity, but compound-7 was not active against CC-resistant MRSA-MW2. Importantly, we evaluated the potential mammalian toxicity of all six analogs using human RBCs. The IC50 values were as follows: CC = 16 mg/L, CC4 = 50 mg/L, CC5 = 25 mg/L, CC6 = 20 mg/L, and CC7 = 16 mg/L, indicating that most analogs retain comparable or reduced toxicity relative to CC.

Fig. 5 — A Structure of CC analogs and their activity against CC-resistant MRSA-MW2. The antibacterial activity of CC analogs against MRSA-MW2 and the CC-resistant MRSA-MW2-S50 mutant using broth microdilution assays were determined. B 500 ns *NPT* MD simulations carried out at the University of Miami of tetrazole-bearing CC analogs with *S. aureus* membranes, following the protocol of the original CC simulations in Fig. 1 and performed in triplicate, revealing a similar membrane insertion process of CC. C CC analogs are bactericidal. Killing kinetics of CC analogs (4x vs untreated) against exponential phase cells. Data are presented as mean values ± SD. P-value for CC against exponential phase is less than p < 0.0001; 2-way-ANOVA. Exponentially growing MRSA-MW2 cells were treated with various concentrations of CC analogs or DMSO (negative control) and colony-forming units (CFU) were determined at various times. D CC and its analogs are membrane-active. Sytox Green was added to MRSA-MW2 cells, and the cells were added to serially diluted CC analogs. The cellular fluorescence was measured at ex 435 nm and em 530 nm. MRSA-MW2 cells treated with CC or its analogs showed an increase in cellular fluorescence. Data are presented as mean values ± SD. P-value for CC against exponential phase is less than p < 0.0001; 2-way-ANOVA. E IC₅₀ of CC analogs. Serially diluted test compounds were added to a monolayer of human hepatic-HepG2 cells. After 22 h of incubation, cell survival was measured using the WST-1 reagent and IC₅₀ values were calculated. F Mammalian cell hemolysis assay. Serially diluted test compounds were added to human RBCs and after incubation for an hour and the supernatant was examined under a spectroscopy at 540 nm. All graphs (C–F) show representative data with means ± SD for biologically independent samples (n = 3). Source data are provided as a Source Data file.

In vivo efficacy of CC against MRSA-MW2

Finally, we tested the in vivo efficacy of CC in invertebrate and vertebrate models. CC exerted dose-dependent antibacterial activity in a C. elegans – MRSA-MW2 liquid-killing assay, reaching 100% worm-survival at 6.25 mg/L^23,45 (Supplementary Fig. 11A). CC at 50 mg/kg also enhanced the survival of MW2-infected Galleria mellonella (p < 0.001) compared to untreated controls (Supplementary Fig. 11B).

Pharmacokinetic studies in mice showed a maximum bioavailability of 8 mM CC in serum 12 h after administration (Fig. 6A). Notably, 50 mg/kg CC did not exhibit gross toxicity based on liver kidney profiling (Fig. 6B, C) and H&E staining of liver, heart, and kidney tissue sections (Fig. 6E).

Injection of 10⁶ S. aureus USA300 cells expressing luciferase into the thighs of Swiss-Webster mice resulted in an active infection site after 24 h. Compared to untreated controls, 50 mg/kg CC (intraperitoneal) and 30 mg/kg gentamicin (subcutaneous) alone decreased S. aureus abundance per gram of tissue by 1-log₁₀-CFU and 2-log₁₀-CFU, respectively. Their combination decreased bacterial burden by 2.5-3.0-log₁₀-CFU (Fig. 6D). Qualitative visual inspection of the images in Fig. 6F suggested that luminescence was reduced at the infection site after CC+gentamicin treatment, whereas in untreated controls, the bacterial luminescence remained essentially unchanged during the same period.

Discussion

Our laboratory has previously reported several membrane-active agents that exhibit anti-MRSA activity^{11–13,46–48}. In this paper, we report the identification of the FDA-approved antihypertensive drug candesartan cilexetil (CC) as a potent anti-MRSA antibiotic that kills logarithmically growing as well as persister cells and decreases the metabolic activity of biofilm cells. Several lines of evidence showed that the mode of action of CC involves penetration and perturbation and/or disruption of the MRSA membrane. CC permeabilized the membrane bilayer to Sytox Green, rapidly dissipated the proton motive force, and caused rapid intracellular alkalization. All-atom molecular dynamics simulations revealed that CC penetrates a simulated S. aureus bilayer and binds to specific lipid molecules in the bilayer, disrupting its structural integrity. TEM revealed structural disruptions at the ultrastructural level, AFM demonstrated changes in membrane topography and mechanical properties, and cryo-CT provided 3D visualization of the extent of membrane compromise. Cryo-tomography with 3D segmentation revealed additional effects caused by CC such as membrane thinning, disruption of the bilayer structure, and the formation of perforations or gaps in the membrane.

Multi-omics analyzes of MRSA cells subjected to treatment with sub-MIC levels of CC, including transcriptomics, metabolomics, and lipidomics, showed that both genes and metabolites associated with membrane biosynthesis were altered by CC treatment, including metabolites such as mevalonic acid, UDP-N-acetylglucosamine, sedoheptulose 7-phosphate, phosphatidylglycerol and phosphatidylethanolamine, which are crucial for maintaining membrane integrity and function. Integrated analysis of multi-omics data revealed that downregulation of cell wall synthesis genes in the transcriptome correlates with decreased levels of key phospholipids in the lipidome, indicating that CC disrupts membrane composition while simultaneously affecting biosynthetic pathways. In addition, rescue experiments using cell-wall constituents (UDP-NAG, NAG) strengthened the connection between transcriptomic/metabolomic down-regulation of cell-wall biosynthesis and the observed lipid/lipidome changes: reduced UDP-NAG compromises peptidoglycan precursor pools, which likely contributes to membrane destabilization and increased permeability. Overall, the “omics” studies indicate and that CC functions both rapidly and over time to disrupt the integrity of the S. aureus membrane. Consistent with the multi-omics findings, the coordinated downregulation of cell wall synthesis genes and the reduction in membrane phospholipids suggest that CC perturbs both structural and biosynthetic components of the S. aureus membrane, highlighting a synergistic disruption of membrane integrity.

To further elucidate the molecular mechanism of CC-induced membrane disruption, generalized polarization measurements using Laurdan dye demonstrated that CC exposure decreases membrane fluidity, rendering the MRSA membrane more rigid. Gas chromatography-mass spectrometry revealed that sub-MIC treatment with CC modified the membrane lipid composition, notably decreasing the content of 20-carbon fatty acids (C20:0). Decreasing the levels of C20:0 fatty acids in the MRSA membrane using a variety of growth media as well as by using inhibitors of C20:0 biosynthesis increased resistance to CC, suggesting that C20:C fatty acids are a key membrane-associated CC target.

Gentamicin and polymyxin B were chosen for detailed analysis of synergistic interactions with CC because their mechanisms complement CC’s membrane-disrupting activity, producing robust and mechanistically interpretable synergy. DiSC3(5) experiments showed that CC combined with aminoglycosides depolarized the MRSA membrane more rapidly than CC alone, directly linking the synergy to enhanced membrane disruption. Molecular dynamics simulations further revealed that CC and polymyxin B spontaneously form stable bi- and tri-molecular complexes that penetrate and deform the S. aureus membrane, amplifying the antimicrobial effect beyond what either molecule achieves alone.

SAR studies showed that the CC tetrazole ring and carbonic acid ester linker moieties are critical for antibacterial activity against MRSA-MW2. Unmodified tetrazole CC analogs retained antibacterial activity. Reducing the hydrophobicity of the cyclohexyl terminus of the carbonic acid ester inhibited penetration into the hydrophobic core, whereas increasing linker hydrophobicity led to deeper and more compact embedding. Shortening the linker length facilitated penetration by creating a more rigid amphiphile, whereas increasing linker length hindered penetration.

Although we have amassed a significant amount of data indicating that CC penetrates and disrupts the S. aureus cellular membrane, some important issues remain regarding the precise mechanism that leads to membrane disruption. Membrane-active antibiotics primarily cause physical injury and/or embed in the membrane. They can also alter membrane potential and intracellular pH, which are the two main components of the proton motive force⁴⁹. As shown in Fig. 1G, CC dissipates the proton motive force, which is known to cause cell death⁵⁰, but surprisingly, CC also caused a rapid increase rather than an expected decrease in intracellular pH (Fig. 1F). The mechanism underlying this phenomenon and its significance in killing S. aureus cells remains to be elucidated.

An additional unexpected result was the finding that treatment of S. aureus with CC results in a decrease in membrane fluidity and increase in membrane rigidity. In contrast, in previous work from our laboratory, Kim et al.⁹ reported that the membrane active compounds bithionol and nTZDpa increase the fluidity of the S. aureus membrane and that their ability to kill persister cells positively correlates with their ability to enhance membrane fluidity^9,51. This apparent contradiction can be explained by membrane lipid phase transition theory⁵²: CC selectively interacts with long-chain saturated fatty acids (C20:0) and other lipids, promoting tighter packing of lipid tails and stabilizing the gel-like phase. This rigidity compromises membrane integrity dissipates the proton motive force, and increases permeability, resulting in bacterial killing^9,53,54. Thus, our finding that CC treatment results in a decrease in membrane fluidity suggests that an increase in membrane fluidity is not highly correlated with anti-persister activity^9,51. It has been reported that the fatty acid composition of membranes affects their fluidity and plays a significant role in the ability of compounds to interact with them⁵⁵. Consistent with these findings, our CC data suggest that the presence of C20 lipids in the S. aureus membrane sensitizes S. aureus to CC and that decreasing the amount of C20 lipids in the membrane may allow S. aureus to develop resistance. Analogous results reported by Boudjemaa et al.⁵⁵ showed that membrane fatty acid composition plays a significant role in the bactericidal activity of daptomycin^27,41,56. Somewhat surprisingly, our experiment designed to isolate CC-resistant S. aureus mutants did not yield mutants related to the biosynthesis of C20 fatty acids. Although CC-resistant MRSA-MW2 mutants lacked mutations in C20:0 biosynthesis genes, resistance may arise through compensatory mechanisms, such as substituting other long-chain fatty acids for C20:0 or remodeling membrane-associated proteins, maintaining membrane integrity while reducing CC susceptibility⁵⁷. Future lipidomic and proteomic studies are needed to confirm these conjectures. A final piece of this puzzle is our RNA-seq, metabolomics, and lipidomics data which showed that treatment of S. aureus with sub-MIC concentrations of CC leads to alterations in the levels of a variety of metabolites and pathways involved in cell membrane biosynthesis. Nevertheless, the “OMICs” experiments did not specifically identify significant changes in the levels of C20 fatty acids. Further experimentation is needed to fully elucidate the role of C20 fatty acids in determining resistance/susceptibility of S. aureus to CC.

CC is a complicated molecule with a variety of moieties, including biphenyl, benzimidazole, tetrazole, and cilexetil groups. We found that both an unmodified tetrazole group and an intact cyclohexyl ring in the cilexetil moiety are required for anti S. aureus activity. Interestingly, two CC analogs with shortened cilexetil group linkers exhibited significant activity against a CC-resistant MRSA-MW2 mutant, suggesting that further SAR studies could potentially identify CC analogs that would be more resistant to the development of resistance than CC.

CC is a prodrug and our in vivo data demonstrate that therapeutically relevant concentrations are achievable and effective against MRSA-MW2. Pharmacokinetic profiling in mice showed serum levels relevant to the in vitro combinatorial MIC with clinical agents, and CC treatment significantly reduced bacterial burden in a mouse thigh infection model. These findings support the translational potential of CC as a viable antibacterial or antibiotic-potentiating agent despite its prodrug nature. Although our in vivo studies demonstrate the antibacterial efficacy of CC, parameters such as blood pressure, serum potassium, and kidney function were not evaluated. Assessing these with larger cohorts and multiple time points will be addressed in future work, including evaluations of CC analogs

In summary, we have identified CC as a potential antimicrobial agent for Gram-positive pathogens. It is active against S. aureus, exhibiting bactericidal activity against both exponential-phase and persister cells. CC exhibits potent anti-MRSA activity by disrupting bacterial membranes and sub-MIC levels of CC perturb lipid homeostasis and alter essential metabolic pathways. CC significantly reduces membrane fluidity, thereby increasing rigidity upon CC exposure. Additionally, CC synergizes with aminoglycoside and lipopeptide antibiotics. The presence of C20 fatty acids in the membrane sensitizes S. aureus to CC-mediated killing. An intact tetrazole moiety as well as an intact cyclohexyl ring in the cilaxetil moiety play key roles in its antibacterial activity. CC can potentially be used to eradicate bacterial infections. These features of CC warrant its further development for use in healthcare settings.

Methods

Data reporting

Methods of this research comply with all relevant ethical regulations approved by the Institutional Biosafety Committee approved the animal protocol (Rhode Island Hospital Institutional Animal Care and Use Committee (RIH IACUC) and Houston Methodist Research Institute) and the experiments were conducted according to the Animal Welfare Act and National Institutes of Health guidelines for animal care and use.

Bacterial strains

The S. aureus methicillin-resistant strain MW2 BAA-1707 (ATCC, Manassas, VA, USA) was used throughout this study. It is a community-acquired, SCCmec Type IV, Panton-Valentine Leucocidin (PVL)-positive strain, which was isolated in 1998 from a female patient in North Dakota, USA⁵⁸. The S. aureus strain VRS1 carries a plasmid encoding the vanA gene, which confers resistance to vancomycin⁵⁹. The cell wall-defective strain MW2∆tarO expresses an inactive, truncated variant of tarO containing only 80 amino acids⁶⁰ and was grown at 37 °C in tryptic soy broth (TSB) (Becton, Dickinson (BD) and Company, Franklin Lakes, NJ, USA).

Compounds

The MicroSource-1 US Drug Collection (http://www.msdiscovery.com/usdrugs.html) is a collection of 1040 compounds that have reached clinical trial stages in the USA. The library was obtained from The Institute of Chemistry and Cell Biology at Harvard Medical School (Boston, MA, USA) in 384-well plates (rows 3 to 22). For high-throughput screening, 0.1 μL of 2 mg/mL compound stock solutions in DMSO were pin-transferred to each well such that the library was screened at a final compound concentration of 2.86 μg/mL. CC was purchased from Cayman Chemical (Ann Arbor, MI, USA), and candesartan was purchased from Tocris (Minneapolis, MN, USA). Tetrazole modified and unmodified CC analogs were purchased from Molport (Toronto Research Chemicals, Canada) and Cayman Chemical and Combiblocks (San Diego, CA, USA). The CC analogs listed in Fig. 5 were synthesized at Sai Lifesciences (Watertown, MA, USA).

Nematode strains

The C. elegans glp-4(bn2); sek-1(km4) double mutant strain was used throughout this study. Nematodes were maintained at 15 °C on a lawn of Escherichia coli strain HB101 on 10-cm plates⁶¹. The glp-4(bn2) mutation renders the strain incapable of producing progeny at 25 °C⁶², and the sek-1(km4) mutation enhances sensitivity to various pathogens⁶³, reducing assay time.

Infection assay for the compound screen

The C. elegans-MRSA screening assay was described earlier²³. Briefly, 4500 glp-4(bn2);sek-1(km4) L1 hatchlings were grown on SK-HB101 agar plates for 52 h at the restrictive temperature of 25 °C until animals became sterile young adults and were harvested with M9 buffer. S. aureus MW2 was grown overnight at 37 °C in TSB, first under aerobic conditions and later shifted to anaerobic growth conditions at 37 °C. The high-throughput screening assay was performed using 384-well plates (Corning no. 3712, Corning, NY, USA), and the total volume in each well was 70 μL, with the final composition being 70% M9 buffer, 19% Sheath solution (Union Biometrica Part no. 300-5101-000, Holliston, MA, USA), 10% TSB, and 1% DMSO or compounds dissolved in DMSO. The bacterial concentration was adjusted to a final optical density (OD)₆₀₀ of 0.04. A Union Biometrica Complex Object Parametric Analyzer and Sorter (COPASBioSort) was used to transfer 15 adult worms to each well of the assay plate. After 5 days of incubation in a humidified chamber at 25 °C, the bacteria and other debris were washed from the wells with a microplate washer, leaving 10 μL of assay volume with worms following the final aspiration step. Finally, 60 μL of 0.9 μM Sytox Orange in M9 was dispensed into each well for a final Sytox concentration of 0.7 μM. The plates were incubated overnight at 25 °C in a humidified chamber. The plates were imaged the next day using an Image Xpress Micro automated microscope (Molecular Devices, San Jose, CA, USA), capturing both transmitted light and TRITC (535 nm excitation, 610 nm emission) fluorescent images with a 2X objective. The entire assay was done in duplicate²³.

Worm survival quantification using CellProfiler and hit identification

The images were processed using the open-source image analysis software CellProfiler (http://www.cellprofiler.org/) and a pipeline of image processing and analysis modules as described previously^64,65. The software calculates the ratio of the Sytox worm area to the bright field worm area and the resultant percentage survival data for each well of the assay plates. Z-scores were calculated from the ratio data, and samples with Z > 2σ in duplicate assay plates were considered hits.

Antimicrobial susceptibility testing

Compounds (10 mg/mL stock solution in DMSO) were tested by broth microdilution and disc diffusion assays adapted from established protocols⁶⁶. Broth microdilution was done in triplicate in 96-well plates. The total volume in each well was 100 μL. Twofold serial dilutions were carried out to get compounds in the 0.125–64 µg/mL concentration range. The bacterial concentration was adjusted to an initial OD₆₀₀ of 0.03. After overnight incubation at 37 °C, the absorbance was measured to determine antimicrobial activity¹³. The disc diffusion test was done on Mueller Hinton agar (MHA) (BD), and the assay was repeated three times. Empty and sterilized discs (BD) were soaked in either DMSO or 10 mg/mL stock solution and air-dried. Three hundred microliters of an overnight culture of MW2 were spread on tryptic soy agar (TSA). After completely drying the agar plate in a laminar flow hood, DMSO or compound-impregnated discs were overlaid on the plate and incubated at 37 °C for 20 h. Antimicrobial susceptibility was determined by measuring the diameter of the zone of inhibition¹³.

Killing kinetics assays

These assays were carried out as described previously¹⁴. An inoculum of ~10⁷ CFU/ml (logarithmic phase) was used for time-kill assays (Fig. 1B) to assess CC activity against actively dividing cells, whereas a higher inoculum of ~10⁸ CFU/ml (stationary phase) was used for growth inhibition assays (Figs. 1C and 3B) to evaluate activity against metabolically less active populations. Test compounds were then added at 8x, 4x, 2x, 1x, and 0.5x MIC and incubated with agitation at 37 °C. Aliquots were periodically drawn, serially diluted, and plated onto tryptic soy agar (TSA; BD Biosciences). Colony-forming units (CFUs) were counted after overnight incubation, and assays were carried out in duplicate¹².

Membrane permeability

Sytox Green (Life Technologies, Carlsbad, CA, USA) was used to probe the effects of CC on MRSA-MW2 membrane permeabilization as previously described⁴⁸. Assays were carried out in duplicate in 96 wells of black flat transparent bottom plates (Corning). Bacterial cells were harvested from logarithmically growing cultures by centrifugation at 3724 × g for 5 min, washed twice with phosphate-buffered saline (PBS, pH 7.4), and resuspended in PBS to OD₅₉₅ nm = 0.2. Sytox Green was added at a final concentration of 5 μM, and cells were incubated in the dark for 30 min. Cell suspensions (50 μL) were added to 50 μL of compound (64 μg/mL in PBS), and the fluorescence intensity was measured (excitation 485 nm, emission 530 nm) periodically over 60 min. DMSO was included as the vehicle control. The membrane-targeting nature of CC was indicated by an increase in cellular fluorescence caused by physical injury of the bacterial membrane, resulting in DNA staining by Sytox Green.

Membrane potential

Membrane potential was measured with a DiSC₃(5) fluorescence probe (Molecular Probes) using the protocol described earlier¹⁰. MRSA-MW2 cells were added with DiSC₃(5) and incubated in dark for 30 min. After incubation, cells were added to a serially diluted compounds and the cellular fluorescence was measured at 620/660 nm.

Membrane fluidity

MRSA-MW2 overnight cultures were inoculated into 25 mL of TSB and allowed to grow to OD₆₀₀ nm ~1.5 at 37 °C with shaking at 200 rpm. After incubation, the cells were harvested by centrifugation and resuspended in PBS followed by 10-min incubation with 10 µM Laurdan (Molecular Probes) at room temperature in the dark. The stained cells were washed four times with PBS. Compounds at various concentrations in 100 μL of PBS in 96-well black transparent flat-bottomed plates (Corning) were mixed with Laurdan-treated bacterial cells. After 1-h incubation at room temperature, the Laurdan fluorescence intensities were measured in the dark at 35 °C using a spectrophotometer (SpectraMax M2) with emission wavelengths of 440 nm and 490 nm on excitation of 350 nm. The Laurdan GP was calculated using the formula GP = (I₄₄₀ – I₄₉₀)/(I₄₄₀ + I₄₉₀). All experiments were conducted in biological triplicate⁹.

Biofilm studies

The biofilm studies were carried out as described^46,67 with slight modifications. The biofilm inhibition studies were carried out using the cells grown under glucose supplementation⁶⁷. The cells were then added to MHB with serially diluted CC and incubated for 24 h. After incubation, planktonic cells were removed, and the plates were washed with sterile PBS (3X) and MHB with Alamar Blue HS was added to the wells to monitor the metabolically alive cells. A mature biofilm eradication assay was carried out using the glucose supplemented TSB grown cells to form a matured biofilm in the 96 well plates. After 24 h, planktonic cells were removed and MHB with serially diluted CC was added to the wells. After 24 h, the plated were washed with sterile PBS (3X) and MHB with Alamar Blue HS were added to the wells to monitor metabolically alive cells.

Scanning electron microscopy

MRSA-MW2 bacteria were suspended in MHB in the presence or absence of CC (1x, 2x, and 4x MIC) for 90 min with agitation. After incubation, the bacteria were harvested by centrifugation at 4000 rpm for 5 min, and harvested cells were adhered to coverslips using 1% poly-L-lysine (Sigma-Aldrich, St. Louis, MO, USA). Cells were fixed in a 5% glutaraldehyde (Sigma-Aldrich), 4% paraformaldehyde (Sigma-Aldrich), and 0.1 M sodium cacodylate (Sigma-Aldrich) solution. After fixation, the cells were washed in 0.1 M sodium cacodylate buffer and dehydrated by immersion in increasing ethanol concentration (from 30 to 100%)⁶⁸. The critical point drying method was applied to dehydrate the samples, and the coverslips were mounted on aluminum stubs and then sputter-coated with gold (Emitech K550, Ashford, Kent, UK). Images were taken on a Hitachi 2700 scanning electron microscope^10,69.

Transmission electron microscopy

CC treated bacterial cells were fixed in 2.5% glutaldehyde in 0.1 M Na Cacodylate and 2% paraformaldehyde with 2 mM calcium chloride for 2–3 h at 4 °C and washed with cold 0.15 M Sodium Cacodylate buffer with 2 mM CaCl. Cells were treated with TCH for 20 min followed by rinsing 5 times at 3 min intervals using ddH₂O at room temperature, then treated with 2% OsO₄ in ddH₂O for 30 min at room temperature followed by an overnight treatment with 1% uranyl acetate overnight at 4 °C. The cells were washed 5 times at 3 min interval using ddH₂O at room temperature and then washed with a Pb aspartate solution in a 60° oven for 30 min, and then washed with 5 times at 3 min interval using ddH₂O at room temperature. Cells were dehydrated in graded ethanol in 70%, 90%, 95%, 20 min each. Finally, cells were embedded in the resin followed by microscopy.

Atomic force microscopy

Sample fixation

The samples were fixed in a solution containing 2.5% glutaraldehyde in PBS for 15 min at room temperature. After fixation, bacterial cell samples were carefully applied to a pre-APTES-treated* silicon surface and incubated for 20 min. For cells grown in 60 mm dishes, fixation and dehydration were performed directly in the dish. Following this step, the silicon-coated surface or the cell-coated dish was rinsed thoroughly with DI water to remove any PBS residues.

Dehydration procedure

The samples were sequentially immersed in ethanol solutions of increasing concentrations (20%, 40%, 60%, 90%, and 100% in water), spending 15 min in each solution. After each step, the samples were spun down to ensure complete processing. Further dehydration was achieved by immersing the samples in tert-butanol for 15 min. All chemicals and buffers were sourced from Sigma Aldrich. After dehydration, the tert-butanol solution was removed, and the samples were left to air dry.

AFM imaging setup

For AFM imaging, the silicon wafer was mounted onto an AFM ferrometal disc and positioned on the AFM scanner. Alternatively, 60 mm dishes were secured directly to the AFM scanner using a Teflon holder. The imaging was conducted using two Bruker Nano systems: the Multimode-9 AFM system for bacterial cells on silicon wafers and the Biocatalyst AFM for regular cells on dishes. Imaging was performed under ambient conditions in contact mode. The AFM cantilever had a spring constant of 0.4 N/m, a scanning frequency of 1 Hz, and a probe-sample interaction force maintained at 0.1 nN. APTES

APTES treatment protocol

APTES, or (3-Aminopropyl)triethoxysilane, was used to prepare the silicon surface. A 0.01 v/v solution of APTES in DI water was applied to the silicon (100) surface and incubated for 2 h. After incubation, the silicon surface was rinsed three times with DI water to remove any residual APTES and then dried using a stream of dry nitrogen.

Cryo-ET of MRSA cells

MRSA-MW2 cells were treated with CC (10× MIC; 80 mg/L) and handed over to the CryoEM Core Team for CryoET preparation. Approximately 3.5 μL of the sample was applied to a Quantifoil 3.5/1 200Cu grid for vitrification. The sample was blotted using calcium-free blotting paper in a Vitrobot (Mark IV, FEI Corp) with a blot time of 2.5 s (1 blot) and immediately plunged into liquid ethane at liquid nitrogen temperature. The grids were then stored in liquid nitrogen until the imaging date. The frozen grids with cells were transferred to either a JEM 3200FSC microscope (JEOL) equipped with a K2 Summit camera and post-column energy filter or a Thermo Scientific Glacios microscope equipped with a Falcon 4 camera. Data from the 3200FSC microscope were collected using SerialEM. Images were recorded at 300 keV with a 40 eV energy filter width on a Gatan 4096 × 4096 pixel K2 Summit direct electron camera in counting mode, with an exposure time of 0.2 s per frame and a total exposure of 3 s. Single-tilt image series (±45°, 2° increments) were collected at a defocus range of 6–10 μm and a magnification of 10,000×, yielding a pixel size of 3.616 Å/pixel. The total electron dose per tomogram was approximately 60 electrons/Å², consistent with standard cellular cryo-ET protocols.

Tomogram processing

Image stacks were automatically aligned and reconstructed using e2tomogram in EMAN2, following initial quality assessment of the raw data to ensure proper tilt-series alignment. To improve contrast and reduce noise, the tomograms were averaged by a factor of four, resulting in 3D reconstructed volumes with dimensions of 1K × 1K × 1K and a resolution of 16 Å/pixel. The vitrified cells on the cryo grids were carefully screened prior to imaging to identify regions with optimal preservation and uniform distribution. The reconstructed tomograms were visualized using EMAN2, allowing for a detailed examination of structural features, including mesosome deformities and the integrity of the bacterial cell wall. Observations focused on identifying disruptions in cellular architecture, particularly in regions of potential compound-induced damage. Specific regions of interest, such as the cell wall and intracellular structures, were analyzed in greater detail. To highlight and quantify the clearest features of these findings, further segmentation and rendering were performed using UCSF Chimera, as illustrated in Fig. 1L. This analysis allowed for the precise visualization of structural abnormalities and comparative assessments between experimental groups⁷⁰.

Computational methodology: molecular dynamics (MD) simulations

All-atom MD simulations were performed using the AMBER22/AmberTools23^71,72 package to investigate the interactions between CC and polymyxin B in a simulated plasma membrane of S. aureus. The second generation Generalized Amber Force Field (GAFF2)⁷³ and AmberTools antechamber package⁷⁴ was utilized to create the non-bonded and bonded parameters for CC and polymyxin B. Partial atomic charges were computed using restrained electrostatic potentials⁷⁵. The structure of CC was built using GaussView⁷⁶, and the geometry was optimized using Gaussian16⁷⁷ with the M06-2X/6-31+G(d) method^78,79 in conjunction with the Polarizable Continuum Model (PCM)⁸⁰ to simulate an aqueous environment. The structure of polymyxin B was adapted from a complex of LSD1-CoREST1 in complex with polymyxin B (PDB entry 5L3F)^81,82 and geometry optimized in Gaussian16 using the B3LYP/6-31+G(d) method^79,83–86 and PCM. The Lipid 21⁸⁷ and ff14SB⁸⁸ forcefields were used to parameterize the DOPC and DOPG (1,2-dioleoyl-sn-glycero-3-phosphocholine/glycerol) lipid bilayer. Construction of the membranes in the xy plane was performed using the PACKMOL-Memgen software⁸⁹ (included in AmberTools23), using 200 neutral-charged DOPC and 86 negatively-charged DOPG lipids representing a DOPC:DOPG ratio of 7:3 with dimensions of 100 Å x 100 Å. All systems were charge-neutralized using sodium ions and were solvated explicitly using the TIP3P water model⁹⁰. The simulation box had an initial height of 102 Å, which, post-equilibration, stabilized to 73 Å, sufficient to prevent the membrane or solutes from interacting with their periodic images. Long-range electrostatic interactions were treated using the Particle-Mesh-Ewald (PME) method⁹¹, utilizing a default cutoff range of 10 Å. The SHAKE⁹² algorithm was used to constrain bonds involving hydrogen atoms, and the time-step for all simulations was set to 2 fs. All simulations were run using the Langevin thermostat⁹³ (set to 310 K and a collision frequency of 1 ps⁻¹) and Berendsen barostat⁹⁴ (set to 1 atm, with semiisotropic pressure scaling).

A solvated DOPC-DOPG membrane bilayer was initially constructed without the drugs present. All water molecules and ions were initially minimized while the lipid bilayer atoms were harmonically restrained using the steepest descent (SD) algorithm for 5000 steps, followed by 5000 steps of conjugate gradient (CG) minimization using mixed CPU and GPU-enabled AMBER22 pmemd engines^95,96. Subsequently, all constraints on the solvent molecules and the lipid bilayer released, and the system was minimized using the SD algorithm for 5000 steps, followed by 5000 steps of CG minimization. This was followed by a two-step thermalization, where the solvated DOPC-DOPG membrane was constrained and the system was gradually heated from 0 to 100 K using a constant NVT ensemble over 5 ps, followed by an additional heating step from 100 K to 310 K using NVT over 100 ps. To ensure correct equilibration of the membrane, a subsequent 100 ps unconstrained NVT simulation was then performed, and a density-correction step involving NPT equilibration ensured the membrane’s density was stable – a process which took up to 2.5 ns. After a 500 ns initial equilibration of the solvated lipid systems, the drugs (CC, Polymyxin B, or CC + Polymyxin B) were introduced approximately 5–6 Å above the membrane surface into the aqueous region and any overlapping water molecules were removed. A constraint of 2.5 kcal mol⁻¹ Å⁻² was placed on the drugs, and the newly complexed system was re-equilibrated using the same procedure as the original DOPC-DOPG membrane, with the final step consisting of 100 ns in the NPT ensemble. After this re-equilibration was complete, the constraints were released on the drugs and their interactions with the membrane, including penetration, were further simulated for 500 ns.

Free energy profiles were computed for the translocation of the drug molecules from aqueous environment into the DOPC-DOPG membrane, which is meant to estimate the energetic pathway that is composed of a transfer energy and membrane penetration energy barrier. Umbrella sampling simulations were set up using AMBER’s tleap ‘build’ module, placing the CC, polymyxin B, or a pre-equilibrated CC+Poly complex at the center of the equilibrated membrane. A 50 ns steered MD in the z direction (with the vector defined by the drug and the DOPC-DOPG xy centers-of-mass) from 0 to 36 Å was performed using a harmonic bias with a force constant of 1.1 kcal mol⁻¹ Å⁻². Umbrella windows were sampled at distances every 2 Å, restrained using a harmonic bias of 2.5 kcal mol⁻¹ Å⁻², and were run for a 50 ns production trajectory. Finally, the weighted histogram analysis method^97,98 was used to produce PMF profiles to elucidate the free energy change as a function of z displacement.

Freely-solvated systems of the drugs polymyxin B and CC were also constructed using USCF Chimera⁷⁰ and simulated in boxes of 5271, 6281, or 6765 TIP3P water molecules for 1:1, 1:2 and 2:1 ratios of CC:Polymyxin, respectively. These followed the same energy minimization and equilibration procedure as the DOPC-DOPG membrane, with three deviations: energy minimization was only performed once using the GPU-based AMBER22 pmemd engine, and NVT equilibrations were run without restraints and using isotropic pressure scaling. Following this, 1500 ns NPT production simulations were performed, and frames with inter-drug distances of less than 8 Å were selected for clustering. To evaluate the dominant clusters of drug aggregates, the K-means algorithm was utilized to generate 10 clusters using RMSD of all drug atoms as the clustering metric. Free energy profiles were computed for the binding energy of these clusters using umbrella sampling, following the same protocol as the umbrella sampling of drug-and-membrane simulations described above. In this case, the last frame of NPT production was chosen at the starting point, and the sampling coordinate was a scalar from the drugs’ contact distance to a COM distance of 20 Å.

Metabolomics and lipidomics analysis sample preparation

MRSA-MW2 cells were treated with CC or control (DMSO). Six biological samples were processed: control (NC-1, NC-2, NC-3) and CC-treated (CC-1, CC-2, CC-3) and submitted to the MSF Core- UTMB. Cells were extracted using a liquid-liquid extraction with MTBE/methanol/water (10:3:2.5, v/v/v). Isotopically labeled internal standards were added prior to extraction, including a commercial lipid mix (UltimateSPLASH™ ONE) for lipidomics. The aqueous phase was used for metabolomics, and the organic phase for lipidomics. The remaining protein pellet was used to quantify total protein content using a BCA assay.

LC-MS/MS analysis

Samples were analyzed using an Acquity Premier UHPLC system (Waters) coupled to a QTRAP 6500 mass spectrometer (SCIEX). Quality control (QC) measures included pooled QC (mixture of all samples), reference standard QC (122 metabolites), and internal standard QC. Pooled QCs were injected at the beginning, middle, and end of the batch to monitor signal stability and retention time drift.

Targeted metabolomics

Dried aqueous phases were resuspended in 80% acetonitrile and separated on a BEH Z-HILIC column (Waters). Mobile phases were: (A) 15 mM ammonium bicarbonate + 2.5 µM Agilent InfinityLab deactivator additive, pH 9; (B) acetonitrile/15 mM ammonium bicarbonate (90:10, v/v) + 2.5 µM additive. A total of 207 metabolites were monitored, and 121 were detected.

Targeted lipidomics

Dried organic phases were resuspended in dichloromethane/methanol (1:1, v/v) and separated by HILIC chromatography on a Luna NH2 column (Phenomenex). Mobile phases were: (A) acetonitrile/water/hexane (92:6:2, v/v/v) + 10 mM ammonium acetate, pH 9.3; (B) acetonitrile/water (1:1, v/v) + 10 mM ammonium acetate, pH 9.3. Lipids were monitored using the Lipidyzer 2.0 targeted panel⁹⁹.

Data analysis

RNA-seq preprocessing, differential expression, and pathway analysis

Raw RNA-sequencing reads (FASTQ files) were processed using Trim Galore to remove adapter sequences and low-quality bases. High-quality reads were aligned to the Staphylococcus aureus strain MW2 reference genome using the STAR aligner (10.1093/bioinformatics/bts635) with default parameters. Gene-level read counts were generated from the aligned BAM files using feature Counts. Differential gene expression analysis was performed using the DESeq2 package¹⁰⁰ in R. Genes with a P value < 0.05 were considered significantly differentially expressed between CC and DMSO samples. Expression patterns of differentially expressed genes were visualized using the pheatmap and EnhancedVolcano R packages. Pathway enrichment analysis of the differentially regulated genes was conducted using the clusterProfiler package¹⁰¹.

Metabolomics and lipidomics

Peak areas were integrated using MultiQuant v3.0.3 (SCIEX) and normalized to appropriate internal standards using a custom R script (v4.3.1). Statistical analyzes were performed in MetaboAnalyst v6.0. Data were log10-transformed and autoscaled. Partial least squares-discriminant analysis (PLS-DA) was performed to compare groups. Volcano plots used thresholds of fold change >2 and p < 0.05. Heatmaps were generated using Euclidean distance and Ward clustering.

Checkerboard assays

Antibacterial synergy was tested using checkerboard assays²⁷. CC was tested with different combinations of clinical antibiotics. MRSA-MW2 cells were adjusted to OD₆₀₀ = 0.06 and added to compound pairs that had been serially diluted in the same 96-well plates, vertically for one compound and horizontally for the other. Assays were carried out in triplicate as described in the antibacterial susceptibility assay sub-section. The fractional inhibitory concentration index (FICI) was calculated as follows: FICI = (MIC_A combination) / (MIC_A alone) + (MIC_B combination) / (MIC_B alone) FICI ≤ 0.5 indicates synergy, 0.5 < FICI < 4 suggests non-interaction, and FICI ≥ 4 indicates antagonism²⁷.

Fluorescence microscopy

Fluorescence Microscopy was carried out as described⁴⁷. Briefly, the cells were strained with fluorescence probes in agar medium entrapped in the microscopic slide. The slides were then examined under 40x in Nikon Florescence Microscope with red and green filters. For Figs. 4 and 6, the microscopes used had a software glitch in scale bars during image acquisition. As a result, the original image data do not contain embedded scale information. In these cases, we report the imaging magnification and acquisition parameters in the corresponding figure legends to provide spatial context.

Bacterial membrane lipid preparation for GC-MS analysis

The membrane lipid profile was analyzed using the MIDI protocol. Bacterial cells were grown overnight in TSB and were used to prepare the log-phase cells with or without CC. After 4 h of agitation, cells were harvested by centrifugation followed by saponification. The cells were treated with 1 ml of Reagent 1 (45 g sodium hydroxide, 150 ml methanol, and 150 ml of distilled water), which was added to each tube containing cells, vortexed briefly, and heated in a boiling water bath for 5 min with a vigorous vortex in-between the heat treatment for 5–10 s and returned to the water bath to complete the 30-min heating. After saponification, methylation was carried out by adding 2 ml of Reagent 2 (325 ml certified 6.0 N hydrochloric acid and 275 ml methyl alcohol) to cooled tubes. The tubes were capped and briefly vortexed, and the tubes were heated for 10 min at 80 °C. After heat incubation, the fatty acid methyl esters were extracted by adding 1.25 ml of Reagent 3 (200 ml hexane and 200 ml methyl tert-butyl ether) to the cooling tubes and tubes were placed in a rotator for 10 min. After incubation, the aqueous (lower) phase was pipetted out and discarded. About 3 ml of Reagent 4 (10.8 g sodium hydroxide dissolved in 900 ml distilled water) was added to the organic phase remaining in the tubes and placed in a rotator for 5 min. After incubation, about 2/3 of the organic phase was pipetted into a GC vial used for further analysis. GC-MS analyzes of the prepared fatty acid methyl esters were performed on an Agilent 7890B instrument coupled to an Agilent 5977A MSD (Agilent Technologies, Palo Alto, CA, USA). An Agilent HP-5MS (30 m × 0.25 mm ID, 0.25 μm film thickness) capillary column was used for separation, using helium as carrier gas. The instrument method was optimized using a 37-component FAME standard mix (CRM47885, Sigma-Aldrich). The temperature program was as follows: the initial temperature 50 °C was held for 1 min, then programmed to increase at 30 °C/min to 150 °C, hold for 1 min, then increase at 6 °C/min to 300 °C and hold for 5 min. Characterization and identification of fatty acids were performed by comparing the mass spectra with the NIST reference mass spectra library (NIST, Mass Spectral Database 08, National Institute of Standards and Technology, Gaithersburg, MD, USA). Data acquisition and processing were performed using Agilent MassHunter software for GC/MSD. Results were expressed as relative percentages of each fatty acid.

Bodipy TR cadeverine displacement assay

The affinity of CC for various lipid molecules was determined using the BODIPY TR cadaverine (Molecular Probes) displacement assay. Bacterial membrane constituents (lipopolysaccharide (LPS) or Lipoteichoic acid (LTA) or Lipid A (Sigma-Aldrich))/ fatty acids (strearic acid or oleic acid or arachidic acid (Sigma-Aldrich) at 5 mg/mL and BODIPY TR cadaverine (500 μM) stock solutions were prepared. The mixtures were prepared using the membrane constituents diluted in Tris buffer to yield final concentrations of lipopolysaccharide (LPS) or Lipoteichoic acid or arachidic acid or stearic acid or oleic acid or Lipid A (Sigma-Aldrich) at 50 μg/mL and 5 μM BODIPY TR cadaverine. 50 μl of the mixture (bacterial membrane constituents or fatty acids mixtures+ BODIPY TR cadaverine) was added to 50 μl of CC in serial dilutions in black-bottomed 96-well plates. Fluorescence was measured at 580/620 nm of excitation/emission. Polymyxin B, known to bind and neutralize LPS, was used as the positive control²⁷.

Cytotoxicity assays

The assays were performed as previously described^102–105. Briefly, mammalian cell lines were maintained in a conical cell culture flask and 25,000 cells/well in 96 well plates were dispensed 24 h prior to the experiment. Compounds were serially diluted in serum and antibiotic free medium and added to the monolayer and incubated for 24 h. At the end of 20 h of incubation, the WST-1 reagent was added and optical density was measured at 420 nm.

Serial passage experiments

Serial passage experiments were performed using the broth microdilution method in 96-well plates. CC was serially diluted in MHB and added to MRSA-MW2 cells, which were then incubated at 35 °C for 24 h. After incubation, optical density measurements were taken, and cells from the well with a sub-MIC concentration of the compound were harvested. The harvested cells were suspended in fresh MHB at a 1:1000 dilution, and this suspension was then added to a new set of wells containing the same sub-MIC concentration of CC, as described previously¹¹ This process was repeated every 24 h for a total of 50 days to track changes in bacterial resistance over time.

Genomic DNA isolation, whole-genome sequencing, and assembly

Clones from the serial passaging experiment in sub-lethal concentrations of CC were cultured in Mannitol Salt Agar and isolated single colonies were cultured in TSB for genomic DNA isolation. The genomes of selected CC variants of MRSA- MW2 were sequenced. Total genomic DNA was extracted from cultures of each strain using the DNeasy Blood & Tissue kit (Qiagen, Valencia, CA, USA) and quantified by Qubit dsDNA High-Sensitivity Assay (Invitrogen, Carlsbad, CA, USA). DNA libraries were prepared from 50 ng of DNA per sample using the Nextera XT DNA Sample Preparation Kit (Illumina, San Diego, CA, USA) with recommended modifications for 2 ×250 bp paired-end sequencing. Samples were multiplexed and sequenced on the MiSeq platform at the Massachusetts Eye and Ear Infirmary Ocular Genomics Institute (Boston, MA, USA). Sequence reads were assembled de novo using the CLC Genomics Workbench software (Cambridge, MA, USA). For evolved strains, variants were identified by mapping sequencing reads for each isolate to the annotated reference genome, which was the starting culture of the serial passaging protocol (MRSA-MW2-S1). For confirmation of variant detection, SNPs and insertion/deletion (indel) variants were identified using CLC Genomics Workbench software (Cambridge, MA, USA) and confirmed by Pilon (10.1371/journal.pone.0112963), using the MRSA-MW2 genome as a reference. Later, contigs were annotated through the NCBI Prokaryotic Genome Annotation Pipeline, and the genomes were deposited in the DDBJ/EMBL/GenBank. Pharmacokinetic mouse model. CFW mice (n = 4) were obtained from Charles River Laboratories (Wilmington, MA, USA) and acclimated for 72 h. CC was dissolved in a 1:1 solution of Kolliphor EL (Sigma-Aldrich, St Louis, MO, USA) and ethanol USP (RIH, RI, USA) and then diluted 1:10 in saline USP to a final concentration of 50 mg/kg. 50 mg/kg of CC was administered via IP, and blood samples were collected using the tail vein prick method. About 50 µl of blood samples were collected from each mouse at different time points (0, 1, 2, 4, 8, 14, and 24 h) and stored in −80 until use. Pooled blood (100 µL) samples in centrifuge tubes were prepared for targeted analysis of CC by addition of 200 µL acetonitrile, vortex mixing for 30 s, and incubation at 4 °C for 12 h^106,107. The samples were centrifuged at 100 × g for 10 min, and the supernatant was transferred to an analysis vial. The sample extracts were analyzed using liquid chromatography-high resolution mass spectrometry (LC-HRMS) using a Thermo QExactive Orbitrap HFX instrument equipped with a Vanquish ultra-high-performance LC (LC-Orbitrap). Extracts were injected into the LC-Orbitrap in 20 µL volumes in triplicate. Reverse-phase LC was used to separate sample components and was performed with a C18 column (Thermo Hypersil Gold Vanquish, 50 mm × 2.1 mm × 1.9 µm) at a constant temperature of 60 °C. Mobile phase A contained 2 mM aqueous ammonium acetate, and mobile phase B contained 2 mM ammonium acetate in acetonitrile. CC was eluted from the column at a constant flow rate of 0.5 mL/min using a mobile phase gradient as follows: equilibration with 2.5% B for 1 min, increase to 100% B over 11 min, and held for 2 min, and back to 2.5% B over 1 min and held for 1.5 min (total run time 16.5 min, data were collected from 0.05 to 12.5 min). The MS was operated in full scan dd-MS² mode (40 NCE) with an inclusion list for CC (m/z 609.2462). Ionization was performed in negative mode with an ionization window of 4.0 m/z. Ions included in the inclusion list were further fragmented in the Higher-energy C-trap dissociation collision cell filled with N₂ (produced by a Peak Scientific Nitrogen Generator, Genius NM32LA). For the full scan, the Orbitrap was operated with a resolution of 120,000. For dd-MS², the Orbitrap was operated with a resolution of 15,000, AGC of 2 × 10⁵, and a maximum dwell time of 400 ms. Quantification was performed in TraceFinder 5.0 General with an external seven-point calibration curve prepared by serial dilutions of a calibration standard. The calibration standard was prepared by dissolving 19.5 mg CC in 3.45 mL dimethyl sulfoxide (9.3 mM). The ion used for quantitation was m/z 609.2462, and the two ions used for confirmation were m/z 309.1031 and m/z 99.0813. The LOD was determined from seven injections of a calibration standard and calculated as: LOD = [s * t (df, 1–α = 0.99)]/m; where s is the standard deviation, t is the student’s t-value, df is the degrees freedom, α is the significance level (n = 7, α = 0.01, t = 3.14), and m is the slope of the calibration curve¹⁰⁸. The LOD was 0.315 µM in 100 µL blood.

Deep-seated thigh infection mouse model and Live imaging

Six-week-old female CFW mice (n = 5/group) were obtained from Charles River Laboratories (Wilmington, MA, USA) and allowed 72 h for acclamation. Animals were maintained in a specific pathogen–free facility with a 12 h light/12 h dark cycle, controlled ambient temperature (20–24 °C), and relative humidity (40–60%), with ad libitum access to food and water. 150 mg/kg of cyclophosphamide (RIH pharmacy, RI, USA) was injected via IP 4 days before infection, and 100 mg/kg of the same drug was administered on the day before infection to make the mice neutropenic. On the day of infection, 10⁶ cells (approx) of log-phase S. aureus (USA 300 Luciferase- a kind gift from the Hamblin lab) suspended in 50 µl of saline were injected into both sides of the thigh (n = 2/mice) of each mouse. After 24 h, the mice were anesthetized using Isoflurane USP and imaged under IVIS to confirm the illumination of bacterial load at the infection site. CC was dissolved in a 1:1 solution of Kolliphor EL (Sigma-Aldrich, St Louis, MO, USA) and ethanol USP (RIH, RI, USA) and then diluted 1:10 in saline USP to a final concentration of 50 mg/kg. 50 mg/kg of CC via IP, 25 mg/kg of vancomycin via IP, 30 mg/kg of gentamicin via SC, or CC (50 mg/kG; IP) + gentamicin (30 mg/kg; SC) were administered to the relevant groups of mice (n = 5; using both legs of each mouse gave a sample number 10) every 24 h for 2 days. Control mice were injected with 100 µl of 10% Kolliphor EL/ethanol in saline via IP. The bacterial load was monitored using the IVIS every 24 h once and before euthanization. After euthanizing mice at 48 h after treatment, the infected thighs were aseptically excised. The homogenized tissue sample in saline were diluted serially and spot plated on TSA plates. The bacterial burden was recorded as CFU/g tissue. The Institutional Biosafety Committee approved the animal protocol (Rhode Island Hospital Institutional Animal Care and Use Committee (RIH IACUC) and the experiments were conducted according to the Animal Welfare Act and National Institutes of Health guidelines for animal care and use. One-way ANOVA analyzed statistical significance among each group with posthoc Tukey test using GraphPad Prism 7 (GraphPad Software, La Jolla, CA, USA).

In vivo imaging: The S. aureus bacterial strain used in the in vivo imaging experiments was USA300: LUC. Briefly, the cultured strain was injected into each thigh of the mouse. After 24 h, the mouse was anesthetized in a chamber that had a continuous supply of isofluorane (1.75 LPM). After 2 min, the mouse was placed in the IVIS imaging platform and maintained under oxygen (2.5 LPM) and isofluorane anesthesia was supplied (1.75 LPM) through a nose cone. The imaging chamber was closed, and imaging was initiated to see the bioluminescence of bacterial presence in the thigh^109,110.

Statistical analysis and reproducibility

All statistical analysis was carried out using GraphPad Prism version 10.02 (GraphPad Software, La Jolla CA, USA) and p values of < 0.05 were considered significant. All experiments were repeated a minimum of three times with technical replicate. The representative of experimental data and micrographs values are used to plot or show the experimental data.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary information^{(13.9MB, pdf)}

41467_2026_70173_MOESM2_ESM.pdf^{(9.2KB, pdf)}

Description of Additional Supplementary Files

Supplement Movie S1^{(43.7MB, mp4)}

Supplement Movie S2^{(7.5MB, mp4)}

Supplement Movie S3^{(7.1MB, mp4)}

Supplement Movie S4^{(5.9MB, mp4)}

Supplement Movie S5^{(38.3MB, mp4)}

Supplement Movie S6^{(40.9MB, mp4)}

Supplement Movie S7^{(48.3MB, mp4)}

Supplement Movie S8^{(44.3MB, mp4)}

Supplement Movie S9^{(50.3MB, mp4)}

Supplement Movie S10^{(42MB, mp4)}

Supplement Movie S11^{(47.7MB, mp4)}

Supplement Movie S12^{(42.4MB, mp4)}

Supplement Movie S13^{(50.9MB, mp4)}

Reporting summary^{(1.6MB, pdf)}

Transparent Peer Review file^{(1.2MB, pdf)}

Source data

Source data^{(34.5MB, xlsx)}

Acknowledgements

The study was supported by NIH grant P01 AI083214 initially awarded to EM and then to EM upon renewal in 2022. Additional support came from pilot grant P20GM121344 awarded to NT. NT thanks the HMRI-AFM core and Dr. Jianhua Gu for assistance. The Thermo LC-Orbitrap MS was supported by an NSF Major Research Instrumentation (MRI) grant CBET-1919870 to KDP. NT sincerely thanks Jenna Port and Dr. Tun-Li Shen at Brown University for their assistance with SEM analysis and GC-MS analysis, respectively. SSM and MSG thank Andrew Van Camp for confirming the mutations (using Pilon) identified in CC-resistant MRSA-MW2 mutants. We thank William Russell and Trevor Romsdahl and the CPRIT Supported UTMB Mass spectrometry facility for help with the targeted lipidomic and metabolomic sample analysis (RP250644).

Author contributions

N.T. and E.M. conceived the project, designed the study, led the project. N.T. performed MIC assays, killing kinetics, fluorescence probe assays, biofilm, checkerboard studies, scanning electron microscopy experiments, mammalian cell models for toxicity, animal studies and live animal imaging. N.T., R.W.K., H.S.R., and O.A. designed the in-silico MD studies and R.W.K., H.S.R., and O.A. performed the MD simulations. N.T. designed and performed bacterial membrane lipid analysis. GC-MS analysis was done in a Mass Spec-core at Brwon University. N.T. designed, performed and analyzed the selection of resistant mutants and S.S.M. isolated genomic DNA, performed whole genome analysis and bioinformatic analysis of variant calling. N.T. performed bacterial treatments for AFM and cryo-CT. CryoEM microscopy was performed in Cryo-EM core at Baylor. P.R.B. acquired the data from Cryo-EM core at Baylor and performed segmentation, 3D data preparation. A.F.M. performed in HMRI-AFM core. R.R. performed C. elegans-high throughput screening and C. elegans liquid killing assay. N.T. and F.C.P.R. designed and performed G. mellonella larvae rescue assay. K.E.M. designed and performed LC-HRMS. N.T. designed and prepared samples for TEM; J.A.D.G. performed TEM. N.T. designed CC analogs in Fig. 6 and synthesized by SAI Lifesciences. N.T., J.A.D.G., O.A., K.D.P., S.L., M.S.G., E.M. contributed reagents, equipment, materials and/or analysis tools. N.T., F.M.A., and E.M. analyzed data and wrote the manuscript. E.M. supervised the project.

Peer review

Peer review information

Nature Communications thanks Yong Wu, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The whole genome sequencing data generated in this study have been deposited in the NCBI database under accession code, https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA659249. The RNA seq, metabolomics, lipidomics data are available under 10.5281/zenodo.17610396. The processed RNA data code is available at 10.5281/zenodo.17610396. The RNA seq, metabolomics, lipidomics data generated in this study are provided in the Supplementary Information/Source Data file. The raw data that are used to generate graphs in this work are available in the Source Data file. Source data are provided with this paper.

Competing interests

FMA had a financial interest in Octagon Therapeutics, Inc., a company that was previously engaged in developing antimicrobial compounds. None of the other authors declare financial conflicts of interest.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Robert Wilson Kovacs, Suelen Scarpa de Mello, Philip Rupert Baldwin.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70173-0.

References

1.Boucher, H. W. & Corey, G. R. Epidemiology of methicillin-resistant Staphylococcus aureus. Clin. Infect. Dis.46, S344–S349 (2008). [DOI] [PubMed] [Google Scholar]
2.Jones, R. N. Global epidemiology of antimicrobial resistance among community-acquired and nosocomial pathogens: a five-year summary from the SENTRY Antimicrobial Surveillance Program (1997-2001). Semin. Respir. Crit. Care Med.24, 121–134 (2003). [DOI] [PubMed] [Google Scholar]
3.Gottlieb, G. S. et al. Staphylococcus aureus bacteremia in the surgical patient: a prospective analysis of 73 postoperative patients who developed Staphylococcus aureus bacteremia at a tertiary care facility. J. Am. Coll. Surg.190, 50–57 (2000). [DOI] [PubMed] [Google Scholar]
4.Benfield, T. et al. Increasing incidence but decreasing in-hospital mortality of adult Staphylococcus aureus bacteraemia between 1981 and 2000. Clin. Microbiol. Infect.13, 257–263 (2007). [DOI] [PubMed] [Google Scholar]
5.Feng, Y. et al. Evolution and pathogenesis of Staphylococcus aureus: lessons learned from genotyping and comparative genomics. FEMS Microbiol. Rev.32, 23–37 (2008). [DOI] [PubMed] [Google Scholar]
6.Lindsay, J. A. Genomic variation and evolution of Staphylococcus aureus. Int. J. Med. Microbiol.300, 98–103 (2010). [DOI] [PubMed] [Google Scholar]
7.CDC. Biggest Threats and Data (Centers for Disease Control and Prevention, Georgia. USA, 2019).
8.Kim, W. et al. The neutrally charged diarylurea compound PQ401 kills antibiotic-resistant and antibiotic-tolerant Staphylococcus aureus. mBio11, e01140-20 (2020). [DOI] [PMC free article] [PubMed]
9.Kim, W. et al. A selective membrane-targeting repurposed antibiotic with activity against persistent methicillin-resistant Staphylococcus aureus. Proc. Natl. Acad. Sci. USA116, 16529–16534 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Tharmalingam, N., Port, J., Castillo, D. & Mylonakis E. Repurposing the anthelmintic drug niclosamide to combat Helicobacter pylori. Sci. Rep.8, 3701 (2018). [DOI] [PMC free article] [PubMed]
11.Kim, W. et al. A new class of synthetic retinoid antibiotics effective against bacterial persisters. Nature556, 103–107 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Tharmalingam, N. et al. Antibacterial properties of four novel hit compounds from a methicillin-resistant Staphylococcus aureus–Caenorhabditis elegans high-throughput screen. Microb. Drug Resist.24, 666–674 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Tharmalingam, N. et al. Activity of a novel protonophore against methicillin-resistant Staphylococcus aureus. Future Med. Chem.9, 1401–1411 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Rajamuthiah, R. et al. Repurposing salicylanilide anthelmintic drugs to combat drug resistant Staphylococcus aureus. PLoS ONE10, e0124595 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Tharmalingam, N., Jayanthan, H. S., Port, J., Rossatto, F. C. P. & Mylonakis, E. Mefloquine reduces the bacterial membrane fluidity of Acinetobacter baumannii and distorts the bacterial membrane when combined with polymyxin B. mBio16, e04016-e04024 (2025). [DOI] [PMC free article] [PubMed]
16.Straus, S. K. & Hancock, R. E. Mode of action of the new antibiotic for Gram-positive pathogens daptomycin: comparison with cationic antimicrobial peptides and lipopeptides. Biochim. Biophys. Acta Biomembranes1758, 1215–1223 (2006). [DOI] [PubMed] [Google Scholar]
17.Mingeot-Leclercq, M.-P. & Décout, J.-L. Bacterial lipid membranes as promising targets to fight antimicrobial resistance, molecular foundations and illustration through the renewal of aminoglycoside antibiotics and emergence of amphiphilic aminoglycosides. MedChemComm7, 586–611 (2016). [Google Scholar]
18.Epand, R. M., Walker, C., Epand, R. F. & Magarvey, N. A. Molecular mechanisms of membrane targeting antibiotics. Biochim. Biophys. Acta Biomembranes1858, 980–987 (2016). [DOI] [PubMed] [Google Scholar]
19.Mishra, N. N. et al. Analysis of cell membrane characteristics of in vitro-selected daptomycin-resistant strains of methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother.53, 2312–2318 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Tsai, M. et al. Staphylococcus aureus requires cardiolipin for survival under conditions of high salinity. BMC Microbiol.11, 13 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Nikolic, P., Mudgil, P., Harman, D. G. & Whitehall, J. Untargeted lipidomic differences between clinical strains of methicillin-sensitive and methicillin-resistant Staphylococcus aureus. Infect. Dis.54, 497–507 (2022). [DOI] [PubMed] [Google Scholar]
22.Gleiter, C. H., Jagle, C., Gresser, U. & Morike, K. Candesartan. Cardiovasc. Drug Rev.22, 263–284 (2004). [DOI] [PubMed] [Google Scholar]
23.Rajamuthiah, R. et al. Whole animal automated platform for drug discovery against multi-drug resistant Staphylococcus aureus. PLoS ONE9, e89189 (2014). [DOI] [PMC free article] [PubMed]
24.Xu, L. et al. Repurposing Candesartan Cilexetil as antibacterial agent for MRSA infection. Front. Microbiol.12, 688772 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.MacGregor, G. A., Viskoper, J. R., Antonios, T. F. T. & He, F. J. Efficacy of Candesartan Cilexetil alone or in combination with amlodipine and hydrochlorothiazide in moderate-to-severe hypertension. Hypertension36, 454–460 (2000). [DOI] [PubMed] [Google Scholar]
26.Kuhn, S., Slavetinsky, C. J. & Peschel, A. Synthesis and function of phospholipids in Staphylococcus aureus. Int. J. Med. Microbiol.305, 196–202 (2015). [DOI] [PubMed] [Google Scholar]
27.Zheng, Z. et al. Synergistic Efficacy of Aedes aegypti Antimicrobial Peptide Cecropin A2 and Tetracycline against Pseudomonas aeruginosa. Antimicrob. Agents Chemother.61, e00686-17 (2017). [DOI] [PMC free article] [PubMed]
28.Bryan, L. E. & Kwan, S. Roles of ribosomal binding, membrane potential, and electron transport in bacterial uptake of streptomycin and gentamicin. Antimicrobial Agents Chemother.23, 835–845 (1983). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Arrow, A. S. & Taber, H. W. Streptomycin accumulation by Bacillus subtilis requires both a membrane potential and cytochrome aa3. Antimicrob. Agents Chemother.29, 141–146 (1986). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Benarroch, J. M. & Asally, M. The microbiologist’s guide to membrane potential dynamics. Trends Microbiol.28, 304–314 (2020). [DOI] [PubMed] [Google Scholar]
31.Hancock, R. E. W. & Chapple, D. S. Peptide antibiotics. Antimicrob. Agents Chemother.43, 1317–1323 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Wenzel, M., Vischer, N. O. E., Strahl, H. & Hamoen, L. W. Assessing membrane fluidity and visualizing fluid membrane domains in bacteria using fluorescent membrane dyes. Bio-Protoc.8, e3063 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Zielińska, A. et al. Flotillin-mediated membrane fluidity controls peptidoglycan synthesis and MreB movement. eLife9, e57179 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Sohlenkamp, C. & Geiger, O. Bacterial membrane lipids: diversity in structures and pathways. FEMS Microbiol. Rev.40, 133–159 (2015). [DOI] [PubMed] [Google Scholar]
35.Omardien, S. et al. Bactericidal activity of amphipathic cationic antimicrobial peptides involves altering the membrane fluidity when interacting with the phospholipid bilayer. Biochim. Biophys. Acta Biomembranes1860, 2404–2415 (2018). [DOI] [PubMed] [Google Scholar]
36.Saleh, I. et al. Design, synthesis, and antibacterial activity of N-(trifluoromethyl) phenyl substituted pyrazole derivatives. RSC Med. Chem.12, 1690–1697 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Altenbern, R. A. Cerulenin-inhibited cells of Staphylococcus aureus resume growth when supplemented with either a saturated or an unsaturated fatty acid. Antimicrob. agents Chemother.11, 574–576 (1977). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Omura, S. The antibiotic cerulenin, a novel tool for biochemistry as an inhibitor of fatty acid synthesis. Bacteriol. Rev.40, 681–697 (1976). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Parsons, J. B., Frank, M. W., Subramanian, C., Saenkham, P. & Rock, C. O. Metabolic basis for the differential susceptibility of Gram-positive pathogens to fatty acid synthesis inhibitors. Proc. Natl. Acad. Sci. USA108, 15378–15383 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Fernandes, M. F., Saxena, J. & Dick, R. P. Comparison of whole-cell fatty acid (MIDI) or phospholipid fatty acid (PLFA) extractants as biomarkers to profile soil microbial communities. Microb. Ecol.66, 145–157 (2013). [DOI] [PubMed] [Google Scholar]
41.Wood, S. J., Miller, K. A. & David, S. A. Anti-endotoxin agents. 1. Development of a fluorescent probe displacement method optimized for the rapid identification of lipopolysaccharide-binding agents. Comb. Chem. High. Throughput Screen.7, 239–249 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Damper, P. D. & Epstein, W. Role of the membrane potential in bacterial resistance to aminoglycoside antibiotics. Antimicrob. Agents Chemother.20, 803–808 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Taber, H. W., Mueller, J. P., Miller, P. F. & Arrow, A. Bacterial uptake of aminoglycoside antibiotics. Microbiol. Rev.51, 439–457 (1987). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Lebeaux, D. et al. pH-mediated potentiation of aminoglycosides kills bacterial persisters and eradicates in vivo biofilms. J. Infect. Dis.210, 1357–1366 (2014). [DOI] [PubMed] [Google Scholar]
45.Sifri, C. D., Begun, J., Ausubel, F. M. & Calderwood, S. B. Caenorhabditis elegans as a model host for Staphylococcus aureus pathogenesis. Infect. Immun.71, 2208–2217 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Zheng, Z. et al. Antimicrobial activity of 1, 3, 4-oxadiazole derivatives against planktonic cells and biofilm of Staphylococcus aureus. Future Med. Chem.10, 283–296 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Mylonakis, E., Tharmalingam, N., Khader, R. & Fuchs, B. B. The anti-virulence efficacy of 4-(1, 3-dimethyl-2, 3-dihydro-1H-benzimidazol-2-yl) phenol against methicillin-resistant Staphylococcus aureus. Front. Microbiol.10, 1557 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Kim, W. et al. Identification of an antimicrobial agent effective against methicillin-resistant Staphylococcus aureus persisters using a fluorescence-based screening strategy. PLoS ONE10, e0127640 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Bakker, E. P. & Mangerich, W. E. Interconversion of components of the bacterial proton motive force by electrogenic potassium transport. J. Bacteriol.147, 820–826 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Bartek, I. L. et al. Antibiotic bactericidal activity is countered by maintaining pH homeostasis in Mycobacterium smegmatis. mSphere1, 10.1128/msphere.00176-00116 (2016). [DOI] [PMC free article] [PubMed]
51.Kim, W. et al. Discovery and optimization of nTZDpa as an antibiotic effective against bacterial persisters. ACS Infect. Dis.4, 1540–1545 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Bramkamp, M. Fluidity is the way to life: lipid phase separation in bacterial membranes. EMBO J.41, e110737 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Gohrbandt, M. et al. Low membrane fluidity triggers lipid phase separation and protein segregation in living bacteria. EMBO J.41, e109800 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Scheinpflug, K. et al. Antimicrobial peptide cWFW kills by combining lipid phase separation with autolysis. Sci. Rep.7, 44332 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Boudjemaa, R. et al. Impact of bacterial membrane fatty acid composition on the failure of daptomycin to kill Staphylococcus aureus. Antimicrob. Agents Chemother.62, e00023–00018 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Woodall, B. M. et al. Enterococcus faecalis readily adapts membrane phospholipid composition to environmental and genetic perturbation. Front. Microbiol.12, 616045 (2021). [DOI] [PMC free article] [PubMed]
57.Kénanian, G. et al. Permissive fatty acid incorporation promotes Staphylococcal adaptation to FASII antibiotics in host environments. Cell Rep.29, 3974–3982.e3974 (2019). [DOI] [PubMed] [Google Scholar]
58.Baba, T. et al. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet359, 1819–1827 (2002). [DOI] [PubMed] [Google Scholar]
59.Fox, P. M., Lampen, R. J., Stumpf, K. S., Archer, G. L. & Climo, M. W. Successful therapy of experimental endocarditis caused by vancomycin-resistant Staphylococcus aureus with a combination of vancomycin and beta-lactam antibiotics. Antimicrob. Agents Chemother.50, 2951–2956 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Campbell, J. et al. Synthetic lethal compound combinations reveal a fundamental connection between wall teichoic acid and peptidoglycan biosyntheses in Staphylococcus aureus. ACS Chem. Biol.6, 106–116 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Tan, M. W., Mahajan-Miklos, S. & Ausubel, F. M. Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis. Proc. Natl. Acad. Sci. USA96, 715–720 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Beanan, M. J. & Strome, S. Characterization of a germ-line proliferation mutation in C. elegans. Development116, 755–766 (1992). [DOI] [PubMed] [Google Scholar]
63.Tanaka-Hino, M. et al. SEK-1 MAPKK mediates Ca2+ signaling to determine neuronal asymmetric development in Caenorhabditis elegans. EMBO Rep.3, 56–62 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Kamentsky, L. et al. Improved structure, function and compatibility for CellProfiler: modular high-throughput image analysis software. Bioinformatics27, 1179–1180 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Moy, T. I. et al. High-throughput screen for novel antimicrobials using a whole animal infection model. ACS Chem. Biol.4, 527–533 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Jorgensen, J. H. & Ferraro, M. J. Antimicrobial susceptibility testing: a review of general principles and contemporary practices. Clin. Infect. Dis.49, 1749–1755 (2009). [DOI] [PubMed] [Google Scholar]
67.Gwisai, T. et al. Repurposing niclosamide as a versatile antimicrobial surface coating against device-associated, hospital-acquired bacterial infections. Biomed. Mater.12, 045010 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Vinod, N. et al. Chemically induced Salmonella enteritidis ghosts as a novel vaccine candidate against virulent challenge in a rat model. Vaccine32, 3249–3255 (2014). [DOI] [PubMed] [Google Scholar]
69.Maruthamuthu, M. K. et al. Development of bisphenol A-removing recombinant Escherichia coli by monomeric and dimeric surface display of bisphenol A-binding peptide. Bioprocess Biosyst. Eng.41, 479–487 (2018). [DOI] [PubMed] [Google Scholar]
70.Pettersen, E. F. et al. UCSF Chimera-a visualization system for exploratory research and analysis. J. Comput. Chem.25, 1605–1612 (2004). [DOI] [PubMed] [Google Scholar]
71.Case, D. A. et al. Amber 2023 (University of California, San Francisco, USA, 2023).
72.Case, D. A. et al. AmberTools. J. Chem. Inf. Model63, 6183–6191 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
73.He, X., Man, V. H., Wang, W., Lee, T.-S. & Wang, J. A fast and high-quality charge model for the next generation general AMBER force field. J. Chem. Phys.153, 114502 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Wang, J., Wang, W., Kollman, P. A. & Case, D. A. Automatic atom type and bond type perception in molecular mechanical calculations. J. Mol. Graph.25, 247–260 (2006). [DOI] [PubMed] [Google Scholar]
75.Bayly, C. I., Cieplak, P., Cornell, W. & Kollman, P. A. A well-behaved electrostatic potential based method using charge restraints for deriving atomic charges: the RESP model. J. Phys. Chem.97, 10269–10280 (1993). [Google Scholar]
76.Dennington R., Keith T. A. & Millam J. M. Semichem Inc. (Shawnee Mission, KS, USA, 2016).
77.Frisch, M. J. et al. Gaussian, Inc. (Wallingford CT, 2016).
78.Zhao, Y. & Truhlar, D. G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor. Chem. Acc.120, 215–241 (2007). [Google Scholar]
79.Hehre, W. J., Ditchfield, R. & Pople, J. A. Self-consistent molecular orbital methods. XII. Further extensions of Gaussian-type basis sets for use in molecular orbital studies of organic molecules. J. Chem. Phys.56, 2257–2261 (1972). [Google Scholar]
80.Miertuš, S., Scrocco, E. & Tomasi, J. Electrostatic interaction of a solute with a continuum. A direct utilization of ab initio molecular potentials for the prevision of solvent effects. Chem. Phys.55, 117–129 (1981). [Google Scholar]
81.Speranzini, V. et al. Polymyxins and quinazolines are LSD1/KDM1A inhibitors with unusual structural features. Sci. Adv.2, e1601017 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Speranzini, V. et al. LSD1-CoREST1 in Complex with Polymyxin B (Worldwide Protein Data Bank, 2016). https://www.wwpdb.org/pdb?id=pdb_00005l3f.
83.Becke, D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys.98, 5648–5652 (1993). [Google Scholar]
84.Lee, C., Yang, W. & Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B37, 785 (1988). [DOI] [PubMed] [Google Scholar]
85.Vosko H., Wilk L. & Nusair M. Accurate spin-dependent electron liquid correlation energies for local spin density calculations: a critical analysis. Can. J. Phys.58, 8 (1980).
86.Stephens, J., Devlin, F. J., Chabalowski, C. F. & Frisch, M. J. Ab initio calculation of vibrational absorption and circular dichroism spectra using density functional force fields. J. Phys. Chem.98, 11623–11627 (1994). [Google Scholar]
87.Dickson, C. J., Walker, R. C. & Gould, I. R. Lipid21: complex lipid membrane simulations with AMBER. J. Chem. Theory Comput.18, 1726–1736 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Maier, J. A. et al. ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB. J. Chem. Theory Comput.11, 3696–3713 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Schott-Verdugo, S. & Gohlke, H. PACKMOL-Memgen: a simple-to-use, generalized workflow for membrane-protein-lipid-bilayer system building. J. Chem. Inf. Model59, 2522–2528 (2019). [DOI] [PubMed] [Google Scholar]
90.Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys.79, 926–935 (1983). [Google Scholar]
91.Darden, T. A., York, D. M. & Pedersen, L. Particle mesh Ewald: an N⋅log(N) method for Ewald sums in large systems. J. Chem. Phys.98, 10089–10092 (1993). [Google Scholar]
92.Ryckaert, J.-P., Ciccotti, G. & Berendsen, H. J. C. Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput Phys.23, 327–341 (1977). [Google Scholar]
93.Loncharich, R. J., Brooks, B. R. & Pastor, R. W. Langevin dynamics of peptides: the frictional dependence of isomerization rates of N-acetylalanyl-N’-methylamide. Biopolymers32, 523–535 (1992). [DOI] [PubMed] [Google Scholar]
94.Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., DiNola, A. & Haak, J. R. Molecular dynamics with coupling to an external bath. J. Chem. Phys.81, 3684–3690 (1984). [Google Scholar]
95.Salomon-Ferrer, R., Goetz, A. W., Poole, D., Le Grand, S. & Walker, R. C. Routine microsecond molecular dynamics simulations with AMBER - Part II: Particle Mesh Ewald. J. Chem. Theory Comput.9, 3878–3888 (2013). [DOI] [PubMed] [Google Scholar]
96.Le Grand, S., Goetz, A. W. & Walker, R. C. SPFP: speed without compromise - a mixed precision model for GPU accelerated molecular dynamics simulations. Comp. Phys. Comm.184, 374–380 (2013). [Google Scholar]
97.Kumar, S., Bouzida, D., Swendsen, R. H., Kollman, P. A. & Rosenberg, J. M. The weighted histogram analysis method for free-energy calculations on biomolecules. I. The method. J. Comput Chem.13, 1011–1021 (1992). [Google Scholar]
98.Grossfield, A. An implementation of WHAM: The Weighted Histogram Analysis Method, Version 2.0.9.) (2014).
99.Romsdahl, T. B., Cocuron, J.-C., Pearson, M. J., Alonso, A. P. & Chapman, K. D. A lipidomics platform to analyze the fatty acid compositions of non-polar and polar lipid molecular species from plant tissues: examples from developing seeds and seedlings of pennycress (Thlaspi arvense). Front. Plant Sci.13, 2022 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
100.Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol.15, 550 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
101.Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics16, 284–287 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
102.Selvamani, V. et al. Hierarchical micro/mesoporous copper structure with enhanced antimicrobial property via laser surface texturing. Adv. Mater. Interfaces7, 1901890 (2020). [Google Scholar]
103.Tharmalingam, N. et al. Inhibitory effect of piperine on Helicobacter pylori growth and adhesion to gastric adenocarcinoma cells. Infect. Agent Cancer9, 1–10 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
104.Vetha, B. S. S. et al. Curcumin encapsulated micellar nanoplatform for blue light emitting diode induced apoptosis as a new class of cancer therapy. Macromol. Res.27, 1179–1184 (2019). [Google Scholar]
105.Swami Vetha, B. S., Oh, P.-S., Kim, S. H. & Jeong, H.-J. Curcuminoids encapsulated liposome nanoparticles as a blue light emitting diode induced photodynamic therapeutic system for cancer treatment. J. Photochem. Photobiol. B Biol.205, 111840 (2020). [DOI] [PubMed] [Google Scholar]
106.Yuce, M. & Capan, Y. Development and validation of hplc analytical methods used-for determination of assay, content uniformity and dissolution of immediate release candesartan cilexetil 32 Mg tablets. Acta Pol. Pharm.74, 357–367 (2017). [PubMed] [Google Scholar]
107.Peeplİwal A. K., Bonde C. G., Mohanraj K. Bioanalytical method development and its validation for determination of candesartan cilexetil by high performance liquid chromatography with UV detection. Acta Pharm. Sci.52, 247–253 (2010).
108.MacDougall, D. & Crummett, W. B. Guidelines for data acquisition and data quality evaluation in environmental chemistry. Anal. Chem.52, 2242–2249 (1980). [Google Scholar]
109.Liu, H. et al. Auranofin releasing antibacterial and antibiofilm polyurethane intravascular catheter coatings. Front. Cell. Infect. Microbiol.9, 37 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
110.Dai, T. et al. Blue light eliminates community-acquired methicillin-resistant Staphylococcus aureus in infected mouse skin abrasions. Photomed. Laser Surg.31, 531–538 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary information^{(13.9MB, pdf)}

41467_2026_70173_MOESM2_ESM.pdf^{(9.2KB, pdf)}

Description of Additional Supplementary Files

Supplement Movie S1^{(43.7MB, mp4)}

Supplement Movie S2^{(7.5MB, mp4)}

Supplement Movie S3^{(7.1MB, mp4)}

Supplement Movie S4^{(5.9MB, mp4)}

Supplement Movie S5^{(38.3MB, mp4)}

Supplement Movie S6^{(40.9MB, mp4)}

Supplement Movie S7^{(48.3MB, mp4)}

Supplement Movie S8^{(44.3MB, mp4)}

Supplement Movie S9^{(50.3MB, mp4)}

Supplement Movie S10^{(42MB, mp4)}

Supplement Movie S11^{(47.7MB, mp4)}

Supplement Movie S12^{(42.4MB, mp4)}

Supplement Movie S13^{(50.9MB, mp4)}

Reporting summary^{(1.6MB, pdf)}

Transparent Peer Review file^{(1.2MB, pdf)}

Source data^{(34.5MB, xlsx)}

Data Availability Statement

[CR1] 1.Boucher, H. W. & Corey, G. R. Epidemiology of methicillin-resistant Staphylococcus aureus. Clin. Infect. Dis.46, S344–S349 (2008). [DOI] [PubMed] [Google Scholar]

[CR2] 2.Jones, R. N. Global epidemiology of antimicrobial resistance among community-acquired and nosocomial pathogens: a five-year summary from the SENTRY Antimicrobial Surveillance Program (1997-2001). Semin. Respir. Crit. Care Med.24, 121–134 (2003). [DOI] [PubMed] [Google Scholar]

[CR3] 3.Gottlieb, G. S. et al. Staphylococcus aureus bacteremia in the surgical patient: a prospective analysis of 73 postoperative patients who developed Staphylococcus aureus bacteremia at a tertiary care facility. J. Am. Coll. Surg.190, 50–57 (2000). [DOI] [PubMed] [Google Scholar]

[CR4] 4.Benfield, T. et al. Increasing incidence but decreasing in-hospital mortality of adult Staphylococcus aureus bacteraemia between 1981 and 2000. Clin. Microbiol. Infect.13, 257–263 (2007). [DOI] [PubMed] [Google Scholar]

[CR5] 5.Feng, Y. et al. Evolution and pathogenesis of Staphylococcus aureus: lessons learned from genotyping and comparative genomics. FEMS Microbiol. Rev.32, 23–37 (2008). [DOI] [PubMed] [Google Scholar]

[CR6] 6.Lindsay, J. A. Genomic variation and evolution of Staphylococcus aureus. Int. J. Med. Microbiol.300, 98–103 (2010). [DOI] [PubMed] [Google Scholar]

[CR7] 7.CDC. Biggest Threats and Data (Centers for Disease Control and Prevention, Georgia. USA, 2019).

[CR8] 8.Kim, W. et al. The neutrally charged diarylurea compound PQ401 kills antibiotic-resistant and antibiotic-tolerant Staphylococcus aureus. mBio11, e01140-20 (2020). [DOI] [PMC free article] [PubMed]

[CR9] 9.Kim, W. et al. A selective membrane-targeting repurposed antibiotic with activity against persistent methicillin-resistant Staphylococcus aureus. Proc. Natl. Acad. Sci. USA116, 16529–16534 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Tharmalingam, N., Port, J., Castillo, D. & Mylonakis E. Repurposing the anthelmintic drug niclosamide to combat Helicobacter pylori. Sci. Rep.8, 3701 (2018). [DOI] [PMC free article] [PubMed]

[CR11] 11.Kim, W. et al. A new class of synthetic retinoid antibiotics effective against bacterial persisters. Nature556, 103–107 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Tharmalingam, N. et al. Antibacterial properties of four novel hit compounds from a methicillin-resistant Staphylococcus aureus–Caenorhabditis elegans high-throughput screen. Microb. Drug Resist.24, 666–674 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Tharmalingam, N. et al. Activity of a novel protonophore against methicillin-resistant Staphylococcus aureus. Future Med. Chem.9, 1401–1411 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Rajamuthiah, R. et al. Repurposing salicylanilide anthelmintic drugs to combat drug resistant Staphylococcus aureus. PLoS ONE10, e0124595 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Tharmalingam, N., Jayanthan, H. S., Port, J., Rossatto, F. C. P. & Mylonakis, E. Mefloquine reduces the bacterial membrane fluidity of Acinetobacter baumannii and distorts the bacterial membrane when combined with polymyxin B. mBio16, e04016-e04024 (2025). [DOI] [PMC free article] [PubMed]

[CR16] 16.Straus, S. K. & Hancock, R. E. Mode of action of the new antibiotic for Gram-positive pathogens daptomycin: comparison with cationic antimicrobial peptides and lipopeptides. Biochim. Biophys. Acta Biomembranes1758, 1215–1223 (2006). [DOI] [PubMed] [Google Scholar]

[CR17] 17.Mingeot-Leclercq, M.-P. & Décout, J.-L. Bacterial lipid membranes as promising targets to fight antimicrobial resistance, molecular foundations and illustration through the renewal of aminoglycoside antibiotics and emergence of amphiphilic aminoglycosides. MedChemComm7, 586–611 (2016). [Google Scholar]

[CR18] 18.Epand, R. M., Walker, C., Epand, R. F. & Magarvey, N. A. Molecular mechanisms of membrane targeting antibiotics. Biochim. Biophys. Acta Biomembranes1858, 980–987 (2016). [DOI] [PubMed] [Google Scholar]

[CR19] 19.Mishra, N. N. et al. Analysis of cell membrane characteristics of in vitro-selected daptomycin-resistant strains of methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother.53, 2312–2318 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Tsai, M. et al. Staphylococcus aureus requires cardiolipin for survival under conditions of high salinity. BMC Microbiol.11, 13 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Nikolic, P., Mudgil, P., Harman, D. G. & Whitehall, J. Untargeted lipidomic differences between clinical strains of methicillin-sensitive and methicillin-resistant Staphylococcus aureus. Infect. Dis.54, 497–507 (2022). [DOI] [PubMed] [Google Scholar]

[CR22] 22.Gleiter, C. H., Jagle, C., Gresser, U. & Morike, K. Candesartan. Cardiovasc. Drug Rev.22, 263–284 (2004). [DOI] [PubMed] [Google Scholar]

[CR23] 23.Rajamuthiah, R. et al. Whole animal automated platform for drug discovery against multi-drug resistant Staphylococcus aureus. PLoS ONE9, e89189 (2014). [DOI] [PMC free article] [PubMed]

[CR24] 24.Xu, L. et al. Repurposing Candesartan Cilexetil as antibacterial agent for MRSA infection. Front. Microbiol.12, 688772 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.MacGregor, G. A., Viskoper, J. R., Antonios, T. F. T. & He, F. J. Efficacy of Candesartan Cilexetil alone or in combination with amlodipine and hydrochlorothiazide in moderate-to-severe hypertension. Hypertension36, 454–460 (2000). [DOI] [PubMed] [Google Scholar]

[CR26] 26.Kuhn, S., Slavetinsky, C. J. & Peschel, A. Synthesis and function of phospholipids in Staphylococcus aureus. Int. J. Med. Microbiol.305, 196–202 (2015). [DOI] [PubMed] [Google Scholar]

[CR27] 27.Zheng, Z. et al. Synergistic Efficacy of Aedes aegypti Antimicrobial Peptide Cecropin A2 and Tetracycline against Pseudomonas aeruginosa. Antimicrob. Agents Chemother.61, e00686-17 (2017). [DOI] [PMC free article] [PubMed]

[CR28] 28.Bryan, L. E. & Kwan, S. Roles of ribosomal binding, membrane potential, and electron transport in bacterial uptake of streptomycin and gentamicin. Antimicrobial Agents Chemother.23, 835–845 (1983). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Arrow, A. S. & Taber, H. W. Streptomycin accumulation by Bacillus subtilis requires both a membrane potential and cytochrome aa3. Antimicrob. Agents Chemother.29, 141–146 (1986). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Benarroch, J. M. & Asally, M. The microbiologist’s guide to membrane potential dynamics. Trends Microbiol.28, 304–314 (2020). [DOI] [PubMed] [Google Scholar]

[CR31] 31.Hancock, R. E. W. & Chapple, D. S. Peptide antibiotics. Antimicrob. Agents Chemother.43, 1317–1323 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Wenzel, M., Vischer, N. O. E., Strahl, H. & Hamoen, L. W. Assessing membrane fluidity and visualizing fluid membrane domains in bacteria using fluorescent membrane dyes. Bio-Protoc.8, e3063 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Zielińska, A. et al. Flotillin-mediated membrane fluidity controls peptidoglycan synthesis and MreB movement. eLife9, e57179 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Sohlenkamp, C. & Geiger, O. Bacterial membrane lipids: diversity in structures and pathways. FEMS Microbiol. Rev.40, 133–159 (2015). [DOI] [PubMed] [Google Scholar]

[CR35] 35.Omardien, S. et al. Bactericidal activity of amphipathic cationic antimicrobial peptides involves altering the membrane fluidity when interacting with the phospholipid bilayer. Biochim. Biophys. Acta Biomembranes1860, 2404–2415 (2018). [DOI] [PubMed] [Google Scholar]

[CR36] 36.Saleh, I. et al. Design, synthesis, and antibacterial activity of N-(trifluoromethyl) phenyl substituted pyrazole derivatives. RSC Med. Chem.12, 1690–1697 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Altenbern, R. A. Cerulenin-inhibited cells of Staphylococcus aureus resume growth when supplemented with either a saturated or an unsaturated fatty acid. Antimicrob. agents Chemother.11, 574–576 (1977). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Omura, S. The antibiotic cerulenin, a novel tool for biochemistry as an inhibitor of fatty acid synthesis. Bacteriol. Rev.40, 681–697 (1976). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Parsons, J. B., Frank, M. W., Subramanian, C., Saenkham, P. & Rock, C. O. Metabolic basis for the differential susceptibility of Gram-positive pathogens to fatty acid synthesis inhibitors. Proc. Natl. Acad. Sci. USA108, 15378–15383 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Fernandes, M. F., Saxena, J. & Dick, R. P. Comparison of whole-cell fatty acid (MIDI) or phospholipid fatty acid (PLFA) extractants as biomarkers to profile soil microbial communities. Microb. Ecol.66, 145–157 (2013). [DOI] [PubMed] [Google Scholar]

[CR41] 41.Wood, S. J., Miller, K. A. & David, S. A. Anti-endotoxin agents. 1. Development of a fluorescent probe displacement method optimized for the rapid identification of lipopolysaccharide-binding agents. Comb. Chem. High. Throughput Screen.7, 239–249 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Damper, P. D. & Epstein, W. Role of the membrane potential in bacterial resistance to aminoglycoside antibiotics. Antimicrob. Agents Chemother.20, 803–808 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Taber, H. W., Mueller, J. P., Miller, P. F. & Arrow, A. Bacterial uptake of aminoglycoside antibiotics. Microbiol. Rev.51, 439–457 (1987). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Lebeaux, D. et al. pH-mediated potentiation of aminoglycosides kills bacterial persisters and eradicates in vivo biofilms. J. Infect. Dis.210, 1357–1366 (2014). [DOI] [PubMed] [Google Scholar]

[CR45] 45.Sifri, C. D., Begun, J., Ausubel, F. M. & Calderwood, S. B. Caenorhabditis elegans as a model host for Staphylococcus aureus pathogenesis. Infect. Immun.71, 2208–2217 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Zheng, Z. et al. Antimicrobial activity of 1, 3, 4-oxadiazole derivatives against planktonic cells and biofilm of Staphylococcus aureus. Future Med. Chem.10, 283–296 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Mylonakis, E., Tharmalingam, N., Khader, R. & Fuchs, B. B. The anti-virulence efficacy of 4-(1, 3-dimethyl-2, 3-dihydro-1H-benzimidazol-2-yl) phenol against methicillin-resistant Staphylococcus aureus. Front. Microbiol.10, 1557 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Kim, W. et al. Identification of an antimicrobial agent effective against methicillin-resistant Staphylococcus aureus persisters using a fluorescence-based screening strategy. PLoS ONE10, e0127640 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Bakker, E. P. & Mangerich, W. E. Interconversion of components of the bacterial proton motive force by electrogenic potassium transport. J. Bacteriol.147, 820–826 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Bartek, I. L. et al. Antibiotic bactericidal activity is countered by maintaining pH homeostasis in Mycobacterium smegmatis. mSphere1, 10.1128/msphere.00176-00116 (2016). [DOI] [PMC free article] [PubMed]

[CR51] 51.Kim, W. et al. Discovery and optimization of nTZDpa as an antibiotic effective against bacterial persisters. ACS Infect. Dis.4, 1540–1545 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Bramkamp, M. Fluidity is the way to life: lipid phase separation in bacterial membranes. EMBO J.41, e110737 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Gohrbandt, M. et al. Low membrane fluidity triggers lipid phase separation and protein segregation in living bacteria. EMBO J.41, e109800 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Scheinpflug, K. et al. Antimicrobial peptide cWFW kills by combining lipid phase separation with autolysis. Sci. Rep.7, 44332 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Boudjemaa, R. et al. Impact of bacterial membrane fatty acid composition on the failure of daptomycin to kill Staphylococcus aureus. Antimicrob. Agents Chemother.62, e00023–00018 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Woodall, B. M. et al. Enterococcus faecalis readily adapts membrane phospholipid composition to environmental and genetic perturbation. Front. Microbiol.12, 616045 (2021). [DOI] [PMC free article] [PubMed]

[CR57] 57.Kénanian, G. et al. Permissive fatty acid incorporation promotes Staphylococcal adaptation to FASII antibiotics in host environments. Cell Rep.29, 3974–3982.e3974 (2019). [DOI] [PubMed] [Google Scholar]

[CR58] 58.Baba, T. et al. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet359, 1819–1827 (2002). [DOI] [PubMed] [Google Scholar]

[CR59] 59.Fox, P. M., Lampen, R. J., Stumpf, K. S., Archer, G. L. & Climo, M. W. Successful therapy of experimental endocarditis caused by vancomycin-resistant Staphylococcus aureus with a combination of vancomycin and beta-lactam antibiotics. Antimicrob. Agents Chemother.50, 2951–2956 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Campbell, J. et al. Synthetic lethal compound combinations reveal a fundamental connection between wall teichoic acid and peptidoglycan biosyntheses in Staphylococcus aureus. ACS Chem. Biol.6, 106–116 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Tan, M. W., Mahajan-Miklos, S. & Ausubel, F. M. Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis. Proc. Natl. Acad. Sci. USA96, 715–720 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Beanan, M. J. & Strome, S. Characterization of a germ-line proliferation mutation in C. elegans. Development116, 755–766 (1992). [DOI] [PubMed] [Google Scholar]

[CR63] 63.Tanaka-Hino, M. et al. SEK-1 MAPKK mediates Ca2+ signaling to determine neuronal asymmetric development in Caenorhabditis elegans. EMBO Rep.3, 56–62 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Kamentsky, L. et al. Improved structure, function and compatibility for CellProfiler: modular high-throughput image analysis software. Bioinformatics27, 1179–1180 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Moy, T. I. et al. High-throughput screen for novel antimicrobials using a whole animal infection model. ACS Chem. Biol.4, 527–533 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR66] 66.Jorgensen, J. H. & Ferraro, M. J. Antimicrobial susceptibility testing: a review of general principles and contemporary practices. Clin. Infect. Dis.49, 1749–1755 (2009). [DOI] [PubMed] [Google Scholar]

[CR67] 67.Gwisai, T. et al. Repurposing niclosamide as a versatile antimicrobial surface coating against device-associated, hospital-acquired bacterial infections. Biomed. Mater.12, 045010 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Vinod, N. et al. Chemically induced Salmonella enteritidis ghosts as a novel vaccine candidate against virulent challenge in a rat model. Vaccine32, 3249–3255 (2014). [DOI] [PubMed] [Google Scholar]

[CR69] 69.Maruthamuthu, M. K. et al. Development of bisphenol A-removing recombinant Escherichia coli by monomeric and dimeric surface display of bisphenol A-binding peptide. Bioprocess Biosyst. Eng.41, 479–487 (2018). [DOI] [PubMed] [Google Scholar]

[CR70] 70.Pettersen, E. F. et al. UCSF Chimera-a visualization system for exploratory research and analysis. J. Comput. Chem.25, 1605–1612 (2004). [DOI] [PubMed] [Google Scholar]

[CR71] 71.Case, D. A. et al. Amber 2023 (University of California, San Francisco, USA, 2023).

[CR72] 72.Case, D. A. et al. AmberTools. J. Chem. Inf. Model63, 6183–6191 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR73] 73.He, X., Man, V. H., Wang, W., Lee, T.-S. & Wang, J. A fast and high-quality charge model for the next generation general AMBER force field. J. Chem. Phys.153, 114502 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR74] 74.Wang, J., Wang, W., Kollman, P. A. & Case, D. A. Automatic atom type and bond type perception in molecular mechanical calculations. J. Mol. Graph.25, 247–260 (2006). [DOI] [PubMed] [Google Scholar]

[CR75] 75.Bayly, C. I., Cieplak, P., Cornell, W. & Kollman, P. A. A well-behaved electrostatic potential based method using charge restraints for deriving atomic charges: the RESP model. J. Phys. Chem.97, 10269–10280 (1993). [Google Scholar]

[CR76] 76.Dennington R., Keith T. A. & Millam J. M. Semichem Inc. (Shawnee Mission, KS, USA, 2016).

[CR77] 77.Frisch, M. J. et al. Gaussian, Inc. (Wallingford CT, 2016).

[CR78] 78.Zhao, Y. & Truhlar, D. G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor. Chem. Acc.120, 215–241 (2007). [Google Scholar]

[CR79] 79.Hehre, W. J., Ditchfield, R. & Pople, J. A. Self-consistent molecular orbital methods. XII. Further extensions of Gaussian-type basis sets for use in molecular orbital studies of organic molecules. J. Chem. Phys.56, 2257–2261 (1972). [Google Scholar]

[CR80] 80.Miertuš, S., Scrocco, E. & Tomasi, J. Electrostatic interaction of a solute with a continuum. A direct utilization of ab initio molecular potentials for the prevision of solvent effects. Chem. Phys.55, 117–129 (1981). [Google Scholar]

[CR81] 81.Speranzini, V. et al. Polymyxins and quinazolines are LSD1/KDM1A inhibitors with unusual structural features. Sci. Adv.2, e1601017 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR82] 82.Speranzini, V. et al. LSD1-CoREST1 in Complex with Polymyxin B (Worldwide Protein Data Bank, 2016). https://www.wwpdb.org/pdb?id=pdb_00005l3f.

[CR83] 83.Becke, D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys.98, 5648–5652 (1993). [Google Scholar]

[CR84] 84.Lee, C., Yang, W. & Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B37, 785 (1988). [DOI] [PubMed] [Google Scholar]

[CR85] 85.Vosko H., Wilk L. & Nusair M. Accurate spin-dependent electron liquid correlation energies for local spin density calculations: a critical analysis. Can. J. Phys.58, 8 (1980).

[CR86] 86.Stephens, J., Devlin, F. J., Chabalowski, C. F. & Frisch, M. J. Ab initio calculation of vibrational absorption and circular dichroism spectra using density functional force fields. J. Phys. Chem.98, 11623–11627 (1994). [Google Scholar]

[CR87] 87.Dickson, C. J., Walker, R. C. & Gould, I. R. Lipid21: complex lipid membrane simulations with AMBER. J. Chem. Theory Comput.18, 1726–1736 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR88] 88.Maier, J. A. et al. ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB. J. Chem. Theory Comput.11, 3696–3713 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR89] 89.Schott-Verdugo, S. & Gohlke, H. PACKMOL-Memgen: a simple-to-use, generalized workflow for membrane-protein-lipid-bilayer system building. J. Chem. Inf. Model59, 2522–2528 (2019). [DOI] [PubMed] [Google Scholar]

[CR90] 90.Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys.79, 926–935 (1983). [Google Scholar]

[CR91] 91.Darden, T. A., York, D. M. & Pedersen, L. Particle mesh Ewald: an N⋅log(N) method for Ewald sums in large systems. J. Chem. Phys.98, 10089–10092 (1993). [Google Scholar]

[CR92] 92.Ryckaert, J.-P., Ciccotti, G. & Berendsen, H. J. C. Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput Phys.23, 327–341 (1977). [Google Scholar]

[CR93] 93.Loncharich, R. J., Brooks, B. R. & Pastor, R. W. Langevin dynamics of peptides: the frictional dependence of isomerization rates of N-acetylalanyl-N’-methylamide. Biopolymers32, 523–535 (1992). [DOI] [PubMed] [Google Scholar]

[CR94] 94.Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., DiNola, A. & Haak, J. R. Molecular dynamics with coupling to an external bath. J. Chem. Phys.81, 3684–3690 (1984). [Google Scholar]

[CR95] 95.Salomon-Ferrer, R., Goetz, A. W., Poole, D., Le Grand, S. & Walker, R. C. Routine microsecond molecular dynamics simulations with AMBER - Part II: Particle Mesh Ewald. J. Chem. Theory Comput.9, 3878–3888 (2013). [DOI] [PubMed] [Google Scholar]

[CR96] 96.Le Grand, S., Goetz, A. W. & Walker, R. C. SPFP: speed without compromise - a mixed precision model for GPU accelerated molecular dynamics simulations. Comp. Phys. Comm.184, 374–380 (2013). [Google Scholar]

[CR97] 97.Kumar, S., Bouzida, D., Swendsen, R. H., Kollman, P. A. & Rosenberg, J. M. The weighted histogram analysis method for free-energy calculations on biomolecules. I. The method. J. Comput Chem.13, 1011–1021 (1992). [Google Scholar]

[CR98] 98.Grossfield, A. An implementation of WHAM: The Weighted Histogram Analysis Method, Version 2.0.9.) (2014).

[CR99] 99.Romsdahl, T. B., Cocuron, J.-C., Pearson, M. J., Alonso, A. P. & Chapman, K. D. A lipidomics platform to analyze the fatty acid compositions of non-polar and polar lipid molecular species from plant tissues: examples from developing seeds and seedlings of pennycress (Thlaspi arvense). Front. Plant Sci.13, 2022 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR100] 100.Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol.15, 550 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR101] 101.Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics16, 284–287 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR102] 102.Selvamani, V. et al. Hierarchical micro/mesoporous copper structure with enhanced antimicrobial property via laser surface texturing. Adv. Mater. Interfaces7, 1901890 (2020). [Google Scholar]

[CR103] 103.Tharmalingam, N. et al. Inhibitory effect of piperine on Helicobacter pylori growth and adhesion to gastric adenocarcinoma cells. Infect. Agent Cancer9, 1–10 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR104] 104.Vetha, B. S. S. et al. Curcumin encapsulated micellar nanoplatform for blue light emitting diode induced apoptosis as a new class of cancer therapy. Macromol. Res.27, 1179–1184 (2019). [Google Scholar]

[CR105] 105.Swami Vetha, B. S., Oh, P.-S., Kim, S. H. & Jeong, H.-J. Curcuminoids encapsulated liposome nanoparticles as a blue light emitting diode induced photodynamic therapeutic system for cancer treatment. J. Photochem. Photobiol. B Biol.205, 111840 (2020). [DOI] [PubMed] [Google Scholar]

[CR106] 106.Yuce, M. & Capan, Y. Development and validation of hplc analytical methods used-for determination of assay, content uniformity and dissolution of immediate release candesartan cilexetil 32 Mg tablets. Acta Pol. Pharm.74, 357–367 (2017). [PubMed] [Google Scholar]

[CR107] 107.Peeplİwal A. K., Bonde C. G., Mohanraj K. Bioanalytical method development and its validation for determination of candesartan cilexetil by high performance liquid chromatography with UV detection. Acta Pharm. Sci.52, 247–253 (2010).

[CR108] 108.MacDougall, D. & Crummett, W. B. Guidelines for data acquisition and data quality evaluation in environmental chemistry. Anal. Chem.52, 2242–2249 (1980). [Google Scholar]

[CR109] 109.Liu, H. et al. Auranofin releasing antibacterial and antibiofilm polyurethane intravascular catheter coatings. Front. Cell. Infect. Microbiol.9, 37 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR110] 110.Dai, T. et al. Blue light eliminates community-acquired methicillin-resistant Staphylococcus aureus in infected mouse skin abrasions. Photomed. Laser Surg.31, 531–538 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Candesartan cilexetil disrupts methicillin-resistant Staphylococcus aureus membrane and potentiates gentamicin and polymyxin B activity

Nagendran Tharmalingam

Robert Wilson Kovacs

Suelen Scarpa de Mello

Philip Rupert Baldwin

Harikrishna Sekar Jayanthan

Kulandaisamy Arulsamy

Rajmohan Rajmuthiah

Fernanda Cristina Possamai Rossatto

Katherine E Manz

Joseph A DeGiorgis

Orlando Acevedo

Michael S Gilmore

Steven J Ludtke

Kurt D Pennell

Frederick M Ausubel

Eleftherios Mylonakis

Abstract

Introduction

Results

CC inhibits the growth of MRSA

Fig. 1. CC disrupts MRSA-MW2 membranes.

CC disrupts the S. aureus membrane

Comprehensive molecular multi-OMICS analysis

Fig. 2. Transcriptome, metabolome, and lipidomic analysis.

CC synergize with gentamicin, kanamycin, streptomycin, & polymyxin B

Fig. 3. CC synergizes with clinical antibiotics against MRSA-MW2.

Mechanism by which CC disrupts the S. aureus membrane

Fig. 4. MRSA-MW2 membrane rigidity and CC as a bactericidal and lipophilic agent.

Isolation of CC-resistant mutants

Structure-activity relationship (SAR) studies

Fig. 5. SAR of CC analogs.

In vivo efficacy of CC against MRSA-MW2

Fig. 6. In vivo efficacy CC against MRSA-MW2 in invertebrate and vertebrate infection models.

Discussion

Methods

Data reporting

Bacterial strains

Compounds

Nematode strains

Infection assay for the compound screen

Worm survival quantification using CellProfiler and hit identification

Antimicrobial susceptibility testing

Killing kinetics assays

Membrane permeability

Membrane potential

Membrane fluidity

Biofilm studies

Scanning electron microscopy

Transmission electron microscopy

Atomic force microscopy

Sample fixation

Dehydration procedure

AFM imaging setup

APTES treatment protocol

Cryo-ET of MRSA cells

Tomogram processing

Computational methodology: molecular dynamics (MD) simulations

Metabolomics and lipidomics analysis sample preparation

LC-MS/MS analysis

Targeted metabolomics

Targeted lipidomics

Data analysis

RNA-seq preprocessing, differential expression, and pathway analysis

Metabolomics and lipidomics

Checkerboard assays

Fluorescence microscopy

Bacterial membrane lipid preparation for GC-MS analysis

Bodipy TR cadeverine displacement assay

Cytotoxicity assays

Serial passage experiments

Genomic DNA isolation, whole-genome sequencing, and assembly

Deep-seated thigh infection mouse model and Live imaging

Statistical analysis and reproducibility

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions