Outer Membrane Disruption Overcomes Intrinsic, Acquired, and Spontaneous Antibiotic Resistance

Abstract

Disruption of the outer membrane (OM) barrier allows for the entry of otherwise inactive antimicrobials into Gram-negative pathogens. Numerous efforts to implement this approach have identified a large number of OM perturbants that sensitize Gram-negative bacteria to many clinically available Gram-positive active antibiotics. However, there is a dearth of investigation into the strengths and limitations of this therapeutic strategy, with an overwhelming focus on characterization of individual potentiator molecules. Herein, we look to explore the utility of exploiting OM perturbation to sensitize Gram-negative pathogens to otherwise inactive antimicrobials. We identify the ability of OM disruption to change the rules of Gram-negative entry, overcome preexisting and spontaneous resistance, and impact biofilm formation. Disruption of the OM expands the threshold of hydrophobicity compatible with Gram-negative activity to include hydrophobic molecules. We demonstrate that while resistance to Gram-positive active antibiotics is surprisingly common in Gram-negative pathogens, OM perturbation overcomes many antibiotic inactivation determinants. Further, we find that OM perturbation reduces the rate of spontaneous resistance to rifampicin and impairs biofilm formation. Together, these data suggest that OM disruption overcomes many of the traditional hurdles encountered during antibiotic treatment and is a high priority approach for further development.IMPORTANCE The spread of antibiotic resistance is an urgent threat to global health that necessitates new therapeutics. Treatments for Gram-negative pathogens are particularly challenging to identify due to the robust outer membrane permeability barrier in these organisms. Recent discovery efforts have attempted to overcome this hurdle by disrupting the outer membrane using chemical perturbants and have yielded several new peptides and small molecules that allow the entry of otherwise inactive antimicrobials. However, a comprehensive investigation into the strengths and limitations of outer membrane perturbants as antibiotic partners is currently lacking. Herein, we interrogate the interaction between outer membrane perturbation and several common impediments to effective antibiotic use. Interestingly, we discover that outer membrane disruption is able to overcome intrinsic, spontaneous, and acquired antibiotic resistance in Gram-negative bacteria, meriting increased attention toward this approach.

Keywords: Gram-negative; antibiotic adjuvant; antibiotic resistance; outer membrane.

FIG 1

Identifying changes in permeability by outer membrane perturbation. (a) Heat map showing antimicrobials potentiated (reduction in MIC >4-fold shown in green) or unaffected (reduction in MIC ≤ 4-fold shown in black) by five OM-perturbing conditions. (b) Physicochemical space of 3,645 compounds screened for bacterial growth inhibition, visualized by molecular weight and calculated logD (cLogD) at pH 7.4. Compounds are colored by growth-inhibitory activity, E. coli control (blue), E. coli with SPR741 (red), no activity in either condition (gray) and activity in both conditions (purple). (c and d) Density plots of molecular weight and cLogD for growth-inhibitory compounds in the E. coli control (blue) and SPR741 condition (red). SPR741 significantly alters the hydrophobicity of compounds compatible with growth inhibition (****, P < 0.0001 by the Kolmogorov-Smirnov test) (d) but not molecular weight (c) (P > 0.05 by the Kolmogorov-Smirnov test). ns, not significant. (e) Venn diagram showing the number and overlap of compounds with growth-inhibitory activity in the E. coli control (blue) and SPR741 condition (red).

FIG 2

Perturbation of the outer membrane overcomes resistance by antibiotic inactivation. (a to c) Potency analysis of erythromycin in E. coli harboring plasmid control (a), mphA (b), or ermC (c) in the presence and absence of SPR741. Data shown represent the means ± standard errors of the means (SEM) (error bars) for at least two biological replicates. (d) Fold reduction of MIC by SPR741 for erythromycin, clarithromycin, and rifampicin in the presence of various resistance elements. Fold reduction is calculated by dividing the MIC of an antibiotic alone by its MIC in the presence of an OM perturbant.

FIG 3

Gram-positive antibiotics are potentiated to therapeutic levels in clinical E. coli isolates by OM perturbation. (a) Resistance genes for rifamycin, aminocoumarin, and macrolide antibiotics predicted in 120 clinical E. coli isolates. Genes are sorted by mechanism into antibiotic efflux (purple), inactivation (orange), and target modification (green). Pie charts represent the total number of unique resistance genes predicted in the strains separated by their corresponding resistance mechanisms. (b) Histograms showing the distribution of rifampicin, novobiocin, and clarithromycin MICs in the presence and absence of SPR741. A dashed line marks the potentiation breakpoint concentration for each antibiotic.

FIG 4

OM perturbation overcomes horizontally acquired macrolide resistance. (a and b) E. coli strains were divided into two groups. Strains predicted to contain macrolide-specific resistance elements (resistant [red, blue, purple, orange]) and no macrolide-specific resistance (sensitive [gray]). Resistant strains are subdivided by color into their predicted resistance genes. (a) MIC values of clarithromycin were significantly increased (****, P < 0.0001 by Mann-Whitney U test) in strains predicted to harbor macrolide resistance (Resistant) in the presence and absence of SPR741. The dotted line marks the potentiation breakpoint value of 2 μg/ml. (b) Fold reduction of clarithromycin MIC was not significantly different (ns) between the sensitive and resistant strains (P > 0.05, Mann-Whitney U test).

FIG 5

Disruption of the OM reduces spontaneous resistance to rifampicin and attenuates biofilm formation. (a) Frequency of resistance to rifampicin after 24 h and 48 h for E. coli, E. coli and SPR741, and E. coli ΔwaaC. (b) Rifampicin MIC during serial passage of E. coli (blue), E. coli ΔwaaC (green), and E. coli with SPR741 (red). (c) Crystal violet (CV) biofilm assay. Absorbance calculated by crystal violet (OD₅₇₀)/growth (OD₆₀₀). All data are shown with SEM and representative of at least two biological replicates. (c) Conditions are compared to the untreated control and significance calculated using the Mann-Whitney U test (ns, P ≥ 0.05; * P < 0.05; **, P < 0.01; ***, P < 0.001).