Chlorate Specifically Targets Oxidant-Starved, Antibiotic-Tolerant Populations of Pseudomonas aeruginosa Biofilms

Abstract

Nitrate respiration is a widespread mode of anaerobic energy generation used by many bacterial pathogens, and the respiratory nitrate reductase, Nar, has long been known to reduce chlorate to the toxic oxidizing agent chlorite. Here, we demonstrate the antibacterial activity of chlorate against Pseudomonas aeruginosa, a representative pathogen that can inhabit hypoxic or anoxic host microenvironments during infection. Aerobically grown P. aeruginosa cells are tobramycin sensitive but chlorate tolerant. In the absence of oxygen or an alternative electron acceptor, cells are tobramycin tolerant but chlorate sensitive via Nar-dependent reduction. The fact that chlorite, the product of chlorate reduction, is not detected in culture supernatants suggests that it may react rapidly and be retained intracellularly. Tobramycin and chlorate target distinct populations within metabolically stratified aggregate biofilms; tobramycin kills cells on the oxic periphery, whereas chlorate kills hypoxic and anoxic cells in the interior. In a matrix populated by multiple aggregates, tobramycin-mediated death of surface aggregates enables deeper oxygen penetration into the matrix, benefiting select aggregate populations by increasing survival and removing chlorate sensitivity. Finally, lasR mutants, which commonly arise in P. aeruginosa infections and are known to withstand conventional antibiotic treatment, are hypersensitive to chlorate. A lasR mutant shows a propensity to respire nitrate and reduce chlorate more rapidly than the wild type does, consistent with its heightened chlorate sensitivity. These findings illustrate chlorate's potential to selectively target oxidant-starved pathogens, including physiological states and genotypes of P. aeruginosa that represent antibiotic-tolerant populations during infections.IMPORTANCE The anaerobic growth and survival of bacteria are often correlated with physiological tolerance to conventional antibiotics, motivating the development of novel strategies targeting pathogens in anoxic environments. A key challenge is to identify drug targets that are specific to this metabolic state. Chlorate is a nontoxic compound that can be reduced to toxic chlorite by a widespread enzyme of anaerobic metabolism. We tested the antibacterial properties of chlorate against Pseudomonas aeruginosa, a pathogen that can inhabit hypoxic or anoxic microenvironments, including those that arise in human infection. Chlorate and the antibiotic tobramycin kill distinct metabolic populations in P. aeruginosa biofilms, where chlorate targets anaerobic cells that tolerate tobramycin. Chlorate is particularly effective against P. aeruginosalasR mutants, which are frequently isolated from human infections and more resistant to some antibiotics. This work suggests that chlorate may hold potential as an anaerobic prodrug.

Keywords: Nar; Pseudomonas aeruginosa; antibiotic tolerance; biofilms; chlorate; nitrate reduction; prodrug.

FIG 1

Chlorate kills oxidant-starved P. aeruginosa cells displaying physiological tolerance to tobramycin. Viable-cell plate counts from P. aeruginosa cultures that were incubated for 4 h without (untreated) or with 40 μg/ml tobramycin, 10 mM chlorate, 40 μg/ml tobramycin plus 10 mM chlorate, or 10 mM chlorite. Cultures were incubated with these compounds under oxic conditions (black), anoxic conditions with 40 mM nitrate (dark gray), or anoxic conditions without nitrate (light gray). Data show the means of results of 9 biological replicates from 3 independent experiments, and error bars indicate standard errors.

FIG 2

Chlorate consumption is correlated with cell death during oxidant starvation. P. aeruginosa cultures were incubated without (untreated) or with 1 mM chlorate (chlorate treated) under oxic conditions (A) or under anoxic conditions with 40 mM nitrate (B). Cultures were monitored for 72 h to determine viable-cell counts (top) and chlorate and nitrate concentrations (bottom) over time. The dashed line in panel B shows the time when nitrate concentrations approximate zero in chlorate-treated cultures. Data show the means of results from 3 biological replicates, and error bars indicate standard errors. In some cases, error bars are smaller than the size of the symbols.

FIG 3

nar genes are required for chlorate reduction and sensitivity. Strains were grown aerobically, resuspended in fresh medium containing or lacking 1 mM chlorate, and incubated in an anaerobic glove box for 72 h. (A) The percent survival was calculated for each strain as the viable-cell counts in chlorate-treated cultures divided by the mean viable-cell count in untreated cultures at the end of the incubation, multiplied by 100. (B) The percentage of chlorate remaining was calculated for each strain as the concentration of chlorate in each culture at the end of the incubation divided by the initial concentration in the medium, multiplied by 100. Strains with pEV, pnar, and pnap, indicate strains that contain an empty vector, a vector with narGHJI, and a vector with napEFDABC, respectively. Three biological replicates of both treated and untreated cultures were used in this experiment, and error bars indicate standard errors.

FIG 4

The ΔlasR mutant has increased rates of nitrate respiration and chlorate consumption and is more sensitive to chlorate. Cell density (OD₅₀₀, filled circles) and nitrate concentrations (open circles) were monitored in the WT and ΔlasR mutant cultures growing under oxic conditions (A), oxic conditions with 40 mM nitrate (B), and anoxic conditions with 40 mM nitrate (C). (D) The WT and ΔlasR cultures were incubated without (squares) or with (filled circles) 1 mM chlorate under anoxic conditions for 72 h, over which time viable-cell counts (solid lines) and chlorate concentrations (dashed lines) were monitored. Data from all experiments show the means of results from three biological replicates, and error bars indicate standard errors.

FIG 5

Tobramycin and chlorate target distinct populations in aggregate biofilms. (A) Cartoon of the agar block biofilm assay (ABBA), where cells suspended in agar medium grow as aggregate biofilms. At early incubations, aggregates are uniform in size, but oxygen gradients develop over time, both within the aggregate population and within individual aggregates, leading to a metabolically heterogeneous population. After 12 h of growth, aggregates are incubated with drugs for 6 h before they are stained and imaged via confocal microscopy. (B) Representative images of untreated aggregates and those treated with 40 μg/ml tobramycin, 10 mM chlorate, or 40 μg/ml tobramycin plus 10 mM chlorate (combined) are shown at three depths, where cells are stained with SYTO 9 (cyan; live) and propidium iodide (red; dead). The scale bar is 50 μm for all images. (C) Confocal images were used to generate a sensitivity profile for each treatment condition, where the proportion of dead cells (propidium iodide intensity divided by the sum of the propidium iodide and SYTO 9 intensities) was determined at each depth. The dashed arrow highlights a shift in the depth, where 50% of cells are killed by chlorate in chlorate-only samples and compared to combined-treatment samples. Data show the means from 6 independent experiments, and error bars indicate standard errors.

FIG 6

Oxygen profiles in treated aggregate biofilm populations. Oxygen profiles of P. aeruginosa aggregate biofilms after a 6-h incubation without a treatment (black) or with 40 μg/ml tobramycin (red), 10 mM chlorate (blue), or 40 μg/ml tobramycin plus 10 mM chlorate (combined; purple). The yellow region highlights measurements taken above the surface of the agar, and the dashed black line shows oxygen profiles from uninoculated agar samples. Data show the means from 3 independent experiments, and colored regions indicate standard errors.