Valnemulin restores colistin sensitivity against multidrug-resistant gram-negative pathogens

The study addresses the urgent need to overcome rising resistance to last-resort antibiotics, particularly colistin, used against MDR, XDR, and PDR gram-negative pathogens. Colistin’s efficacy is compromised by chromosomal mutations (phoPQ, pmrAB, crrAB, mgrB) and plasmid-mediated mcr genes that add phosphoethanolamine to lipid A, reducing colistin binding. Given colistin’s nephro- and neurotoxicity, strategies that restore its activity at lower doses are desirable. Antibiotic adjuvants can enhance antibiotic efficacy without strong intrinsic antibacterial activity. The research question is whether valnemulin, an FDA-approved pleuromutilin primarily active against mycoplasmas and Pasteurella but not Enterobacteriaceae, can act as a safe, effective adjuvant to restore colistin activity against colistin-resistant gram-negative bacteria, including extracellular and intracellular forms, and improve outcomes in vivo. The study further aims to elucidate mechanisms underlying any synergy observed.

Prior work has shown that adjuvants can potentiate colistin by diverse mechanisms. Natural phytochemicals such as catechol-type flavonoids and kaempferol potentiate colistin by disrupting bacterial iron homeostasis. Melatonin restores colistin activity against MCR-producing pathogens through promoting oxidative damage and inhibiting efflux. Otilonium bromide, an FDA-approved antispasmodic, enhances colistin by dissipating proton motive force (PMF) and suppressing efflux pumps. These findings support repurposing FDA-approved compounds as colistin adjuvants to reduce development cost and toxicity. Valnemulin binds the 50S ribosomal subunit to inhibit protein synthesis and has shown synergy with doxycycline against Acinetobacter baumannii. It displays low cytotoxicity to mammalian cells at clinically relevant concentrations (0–25 µg/ml). However, no colistin adjuvant is currently commercially available, underscoring the need for new agents.

- Bacterial strains and culture: Clinical and laboratory strains (listed in Supplementary Table S1) including colistin-resistant E. coli P47 (mcr-1), K. pneumoniae P30 and 80 (mgrB inactivation), and A. pittii CS17 (pmrC mutation), plus colistin-susceptible E. coli J53, K. pneumoniae MGH78578, and A. pittii P87. Cultured in LB broth/agar at 37 °C. - Antimicrobial susceptibility and synergy testing: MICs determined by broth dilution per CLSI (2023). Checkerboard assays in CAMHB with 8×6 matrices at 10^6 CFU/ml to compute FICI (FICI ≤ 0.5 indicates synergy). Isobolograms plotted from FIC values. Triplicate biological replicates. - Time-kill kinetics: Mid-log cultures of colistin-resistant strains exposed to Val, colistin, or both over 24 h; CFU quantified over time to generate killing curves (GraphPad Prism 9.0). - Resistance development: Serial passage (10 days) of colistin-susceptible E. coli BW25113, K. pneumoniae 80 (MGH78578), and A. pittii P87 with sub-MIC colistin ± 2 µg/ml Val to track fold-changes in colistin MIC. - Scanning electron microscopy (SEM): E. coli P47 (mcr-1) treated with Val (2 µg/ml), colistin (2 or 64 µg/ml), or combination (Val 2 µg/ml + colistin 2 µg/ml) for 2 h; fixed with glutaraldehyde, dehydrated, and imaged on Hitachi SU8010. - Cytotoxicity: HEK293T cells treated with Val, colistin, or both for 24 h; MTT assay read at 570 nm on EnSight microplate reader. Blank controls included. - Intracellular killing: Vero cells infected with E. coli P47 (MOI 100), centrifuged and incubated 1 h, then treated with Val, colistin, or both for 6 h. Extracellular bacteria removed with meropenem (2 µg/ml). Cells washed, lysed (DMEM + 0.1% BSA + 0.1% Triton X-100), plated for CFU. - Mouse infection model: Immunosuppressed BALB/c mice (5–8 weeks) infected intraperitoneally with colistin-resistant, carbapenem-resistant E. coli P80 (4.0 × 10^6 CFU/ml). Treatment groups (n=5/group): saline, Val (2 mg/kg), colistin (4 mg/kg), or Val+colistin, dosed every 12 h for 48 h starting 1 h post-infection. Survival monitored; IACUC approved. - Membrane permeability: SYTOX Green staining after 2 h exposure to Val, colistin, or both; fluorescence measured (ex/em 488/523 nm). - Membrane depolarization (PMF proxy): DiSC3(5) labeling of E. coli P47 followed by treatment with Val, colistin, or both; fluorescence increase indicates depolarization. - Intracellular ATP: Enhanced ATP Assay Kit used after 2 h treatments; luminescence measured on EnSight reader. - Swimming motility: Inoculation on 0.3% agar LB with Val, colistin, or both; halos imaged after 48 h at 37 °C; time-kill controls performed to rule out killing under these conditions. - Transcriptomics: E. coli P47 treated with colistin (1 µg/ml) ± Val (2 µg/ml) to OD600 ~0.6. RNA extracted and sequenced (Illumina HiSeq; Majorbio). Differential expression analyzed with DESeq2; functional annotation with GO and KEGG. Statistical analyses via unpaired t-tests or one-way ANOVA (GraphPad Prism 9.0).

- Synergy in vitro (checkerboard/isobolograms): Against colistin-resistant strains E. coli P47 (mcr-1), K. pneumoniae P30 and 80 (mgrB inactivation), and A. pittii CS17 (pmrC mutation), Val + colistin showed strong synergy with FICIs of 0.312, ≤0.0625, ≤0.0938, and ≤0.0156, respectively. In colistin-susceptible strains E. coli J53, K. pneumoniae MGH78578, and A. pittii P87, FICIs were 0.3125, ≤0.28125, and ≤0.375, indicating synergy albeit generally weaker than in resistant isolates. - Time-kill assays: Monotherapy with Val (2 µg/ml) or colistin (2 µg/ml; for K. pneumoniae P30 32 µg/ml; for A. pittii CS17 16 µg/ml) did not suppress growth over 24 h. Combination therapy effectively eradicated bacteria within 24 h with marked CFU reductions relative to either agent alone. - Resistance development: Serial passage with sub-MIC colistin caused rapid MIC increases in K. pneumoniae MGH78578 (64-fold) and A. pittii P87 (16-fold). Adding Val (2 µg/ml) mitigated these increases to 16-fold and ~1-fold, respectively; E. coli J53 did not develop resistance under either condition. - Intracellular killing: In Vero cells infected with E. coli P47, Val alone (2 µg/ml) had no effect and colistin alone (2 µg/ml) slightly reduced intracellular CFU, whereas the combination (Val 2 µg/ml + colistin 2 µg/ml) reduced viable intracellular bacteria by approximately 1000-fold versus untreated. - In vivo efficacy: In an immunosuppressed mouse sepsis model with colistin-resistant E. coli P80, survival at 48 h was 0% with saline, Val (2 mg/kg), or colistin (4 mg/kg) monotherapies; combination therapy (Val + colistin) produced 60% survival at 48 h (trend not reaching statistical significance). - Membrane effects: SEM showed morphological damage and apparent cytosolic leakage only with the combination (Val 2 µg/ml + colistin 2 µg/ml), consistent with membrane disruption. SYTOX Green assays demonstrated significantly increased membrane permeability with Val + colistin versus colistin alone across E. coli P47, K. pneumoniae P30, and A. pittii CS17 (****p < 0.0001). - PMF dissipation and ATP depletion: DiSC3(5) fluorescence increased significantly with Val + colistin compared to either agent alone, indicating enhanced membrane depolarization. Intracellular ATP levels were significantly reduced by the combination relative to monotherapies. - Motility suppression: While Val (2 µg/ml) or colistin (1 µg/ml) alone did not affect swimming motility, the combination reduced migration distance on 0.3% agar (visual qualitative assessment; statistical analysis not performed due to irregular colony morphology). CCCP, a PMF disruptor, synergized with colistin in E. coli P47 and J53 with FICIs ≤0.25 and ≤0.375, respectively. - Transcriptomics: Compared to colistin alone (1 µg/ml), adding Val (2 µg/ml) yielded 1545 upregulated and 318 downregulated genes in E. coli P47. Upregulated genes were enriched in membrane/transport functions, including ABC transporters (yddA, malk, malF, malG, phnC, yciO, malA), MFS (ydiM, setC, ydjE, entS, ydiN), and RND efflux (acrE, acrF). Downregulated genes were enriched for cell motility, especially flagellar biosynthesis and structural genes (e.g., fliT, flis, flgN, fliQ, flio, flir, fliH, flil, flij, figD; flgB, flgA, yegR, fliN, fliM, fliD, flgK, flgL, flie, flgG, flgH, flgC, fliF, flgF, flgE, fliC).

The results demonstrate that valnemulin significantly enhances colistin activity against both colistin-resistant and colistin-susceptible gram-negative bacteria, including intracellular forms and in a murine sepsis model. Mechanistically, Val potentiates colistin’s membrane-targeting action by increasing membrane permeabilization and dissipating PMF, which reduces intracellular ATP and suppresses flagellum-dependent motility. Transcriptomics supports membrane stress and transport responses together with repression of motility-related genes, consistent with observed phenotypes. The PMF dissipation likely weakens divalent cation (Mg2+, Ca2+)-mediated stabilization of LPS, facilitating colistin binding to lipid A and enhancing bactericidal effect. By restoring colistin efficacy, Val could permit lower colistin dosing, potentially mitigating nephro- and neurotoxicity. The ability to kill intracellular bacteria suggests possible benefits in hard-to-treat infections where intracellular persistence contributes to chronicity and, potentially, cancer-related pathophysiology. Overall, these findings support repurposing Val as a practical adjuvant to extend the clinical utility of colistin.

This study identifies valnemulin as a potent colistin adjuvant that restores colistin activity against MDR gram-negative pathogens. The combination displays robust in vitro synergy across resistant and susceptible isolates, reduces the emergence of colistin resistance, eradicates intracellular bacteria, and improves survival in a murine sepsis model. Mechanistically, Val enhances colistin-mediated membrane disruption, dissipates PMF, decreases intracellular ATP, and suppresses motility, with transcriptomic changes consistent with these effects. Given Val’s favorable cytotoxicity profile, the combination may allow reduced colistin dosing. Future work should validate efficacy across broader pathogen panels and infection models, optimize dosing and pharmacodynamics/pharmacokinetics, assess toxicity and therapeutic windows in vivo, and explore clinical translation.

- In vivo survival benefit (60% at 48 h for the combination) did not reach statistical significance in the reported experiment (n=5 per group). - Swimming motility assessments were qualitative; irregular colony morphology precluded precise distance measurements and formal statistical analysis. - The number of tested strains, while including clinically relevant resistance mechanisms (mcr-1, mgrB inactivation, pmrC mutation), remains limited; broader validation is needed. - Affinity, pharmacokinetic, and toxicity profiles of the Val–colistin combination were not comprehensively characterized in vivo beyond short-term survival and cell viability assays.