The bacterial cytoplasmic membrane is a major inhibitory target for antimicrobial compounds. Commonly, although not exclusively, these compounds unfold their antimicrobial activity by disrupting the essential barrier function of the cell membrane. As a consequence, membrane permeability assays are central for mode of action studies analysing membrane-targeting antimicrobial compounds. The most frequently used in vivo methods detect changes in membrane permeability by following internalization of normally membrane impermeable and relatively large fluorescent dyes. Unfortunately, these assays are not sensitive to changes in membrane ion permeability which are sufficient to inhibit and kill bacteria by membrane depolarization. In this manuscript, we provide experimental advice how membrane potential, and its changes triggered by membrane-targeting antimicrobials can be accurately assessed in vivo. Optimized protocols are provided for both qualitative and quantitative kinetic measurements of membrane potential. At last, single cell analyses using voltage-sensitive dyes in combination with fluorescence microscopy are introduced and discussed.

Antibiotics that interfere with the bacterial cytoplasmic membrane have long-term potential for the treatment of infectious diseases as this mode of action is anticipated to result in low resistance frequency. Vancoresmycin is an understudied natural product antibiotic consisting of a terminal tetramic acid moiety fused to a linear, highly oxygenated, stereochemically complex polyketide chain. Vancoresmycin shows minimum inhibitory concentrations (MICs) from 0.125 to 2 μg/mL against a range of clinically relevant, antibiotic-resistant Gram-positive bacteria. Through a comprehensive mode-of-action study, utilizing Bacillus subtilis reporter strains, DiSC(5) depolarization assays, and fluorescence microscopy, we have shown that vancoresmycin selectively targets the cytoplasmic membrane of Gram-positive bacteria via a non-pore-forming, concentration-dependent depolarization mechanism. Whole genome sequencing of the producing strain allowed identification of the 141 kbp gene cluster encoding for vancoresmycin biosynthesis and a preliminary model for its biosynthesis. The size and complex structure of vancoresmycin could confound attempts to generate synthetic analogues. To overcome this problem and facilitate future studies, we identified, cloned, and expressed the 141 kbp biosynthetic gene cluster in Streptomyces coelicolor M1152. Elucidation of the mode-of-action of vancoresmycin, together with the heterologous expression system, will greatly facilitate further studies of this and related molecules.

Rather than being homogenous diffusion-dominated structures, biological membranes can exhibit areas with distinct composition and characteristics, commonly termed as lipid domains. Arguably the most comprehensively studied examples in bacteria are domains formed by cardiolipin, which have been functionally linked to protein targeting, the cell division process and the mode of action of membrane-targeting antimicrobials. Cardiolipin domains were originally identified in the Gram-negative model organism Escherichia coli based on preferential staining by the fluorescent membrane dye nonylacridine orange (NAO), and later reported to also exist in other Gram-negative and -positive bacteria. Recently, the lipid-specificity of NAO has been questioned based on studies conducted in E. coli. This prompted us to reanalyse cardiolipin domains in the Gram-positive model organism Bacillus subtilis. Here we show that logarithmically growing B. subtilis does not form microscopically detectable cardiolipin-specific lipid domains, and that NAO is not a specific stain for cardiolipin in this organism.

Rather than being a homogenous diffusion-dominated structure, biological membranes can exhibit areas with distinct composition and characteristics commonly termed as lipid domains. Arguably the most comprehensively studied examples in bacteria are domains formed by cardiolipin, which have been functionally linked to protein targeting, cell division process, and mode of action of membrane targeting antimicrobials. Cardiolipin domains were originally identified in the Gram-negative model organism Escherichia coli based on preferential staining by the fluorescent membrane dye nonyl acridine orange (NAO), and later reported to exist also in other Gram-negative and -positive bacteria. Recently, the lipid-specificity of NAO has been questioned based on studies conducted in E. coli. This prompted us to re-analyse cardiolipin domains also in the Gram-positive model organism Bacillus subtilis. Here we show that logarithmically growing B. subtilis does not form microscopically detectable cardiolipin-specific lipid domains, and that NAO is not a specific stain for cardiolipin in this organism.AbbreviationsNAO10-nonyl acridine orange bromideNile Red9-diethylamino-5-benzo[±]phenoxazinone