Imidazole antibiotics inhibit the nitric oxide dioxygenase function of microbial flavohemoglobin

Abstract

Flavohemoglobins metabolize nitric oxide (NO) to nitrate and protect bacteria and fungi from NO-mediated damage, growth inhibition, and killing by NO-releasing immune cells. Antimicrobial imidazoles were tested for their ability to coordinate flavohemoglobin and inhibit its NO dioxygenase (NOD) function. Miconazole, econazole, clotrimazole, and ketoconazole inhibited the NOD activity of Escherichia coli flavohemoglobin with apparent K(i) values of 80, 550, 1,300, and 5,000 nM, respectively. Saccharomyces cerevisiae, Candida albicans, and Alcaligenes eutrophus enzymes exhibited similar sensitivities to imidazoles. Imidazoles coordinated the heme iron atom, impaired ferric heme reduction, produced uncompetitive inhibition with respect to O(2) and NO, and inhibited NO metabolism by yeasts and bacteria. Nevertheless, these imidazoles were not sufficiently selective to fully mimic the NO-dependent growth stasis seen with NOD-deficient mutants. The results demonstrate a mechanism for NOD inhibition by imidazoles and suggest a target for imidazole engineering.

FIG. 1.

Inhibition of E. coli NOD activity by imidazoles. NOD activity of E. coli flavoHb was assayed at the indicated concentrations of miconazole (line 1), econazole (line 2), clotrimazole (line 3), or ketoconazole (line 4). DMSO (lines 1-3) and methanol (line 4) were present at a final concentration of 0.1% (vol/vol).

FIG. 2.

NOD activity of E. coli flavoHb with various miconazole, O₂, and NO concentrations. NOD activity was assayed with various concentrations of O₂ and 0.75 μM NO (A) or with various concentrations of NO and 200 μM O₂ (B) in the presence of 0 μM (•), 0.1 μM (▪), 0.25 μM (○), or 0.5 μM (□) miconazole. DMSO was present at a final concentration of 0.1% (vol/vol).

FIG. 3.

Titration of oxidized and reduced flavoHb with miconazole. (A) Miconazole was added to E. coli flavoHb-Fe(III) (4.0 μM heme), and coordination was followed by measuring the absorbance differences for the maximum (417 nm) and minimum (382 nm) of difference spectra of the miconazole complex and free enzyme. (B) Miconazole was added to NADH-reduced flavoHb-Fe(II) (4.0 μM heme) under anaerobic conditions. Samples were incubated for 5 min to allow coordination, difference spectra were recorded, and the absorbance difference for the peak (412 nm) and trough (437 nm) were measured.

FIG. 4.

NADH reduction of the flavoHb-Fe(III)-miconazole complex. (A) Spectra of E. coli flavoHb (6.0 μM heme) were recorded at 1-min intervals following addition of 1 mM NADH. (B) As for panel A, except miconazole (13 μM) was added prior to the addition of NADH. Arrows indicate the direction of the absorbance change. (C) Time course for NADH-mediated reduction of flavoHb-Fe(III) (line 1) and flavoHb-Fe(III)-miconazole complex (line 2) as measured by the absorbance changes at 433 nm and 427 nm, respectively. (D) Time course for reduction of FAD in flavoHb-FAD/Fe(III) (line 1) and flavoHb-FAD/Fe(III)-miconazole (line 2) measured at 460 nm. DMSO was present at a final concentration of 1.3% (vol/vol).

FIG. 5.

Effect of miconazole on NO metabolism in S. cerevisiae. (A) NO consumption by S. cerevisiae was assayed with various concentrations of miconazole. (B) Time dependence of activity loss in the presence of 0, 2, 5, 10, or 50 μM miconazole as indicated (lines 1-5). DMSO was added as the solvent to a final concentration of 0.1% (vol/vol). Error bars represent standard deviations of the average for three independent trials.

FIG. 6.

Effect of NO and imidazoles on S. cerevisiae and C. albicans growth. (A) Cultures of S. cerevisiae strain BY4742 (lines 1-4) and the isogenic NOD-deficient mutant yhb1Δ (lines 5 and 6) were grown under a normoxic atmosphere in the absence (lines 1, 3, and 5) or presence of 960 ppm NO gas (lines 2, 4, and 6). Miconazole (lines 3 and 4) (5 μM) and NO were delivered at the time indicated by the arrow. Approximate generation times (minutes) are given in italics. (B) Cultures of C. albicans strain RM1000 and the isogenic NOD-deficient mutant yhb1Δ/yhb1Δ (ΔYHB1) were grown in modified YPD medium containing 5 mM glucose and 10 μM O₂. Cultures were exposed to 160 ppm NO (<0.3 μM in solution) with or without 5 μM econazole. After 20 h, growth was measured by the absorbance at 600 nm. Single asterisks indicate a P of <0.05 relative to the control condition. Double asterisks indicate a P of <0.05 relative to both the control condition and between strains.

FIG. 7.

Proposed mechanism for imidazole inhibition of NOD. Azoles coordinate the ferric heme iron and impair hydride transfer (k′_H) (step 1) and electron transfer (k_ET) (steps 2a and 2b). Azoles readily dissociate from the reduced ferrous heme iron, allowing O₂ binding and formation of Fe(III)-O₂ · (steps 3a and 3b). Rapid reaction of · NO with the Fe(III)-O₂ · intermediate forms the ferric-peroxynitrite intermediate (4a and 4b), which rapidly isomerizes to produce nitrate and flavoHb-Fe(III) (5a and 5b).