Screening of Antifungal Azole Drugs and Agrochemicals with an Adapted alamarBlue-Based Assay Demonstrates Antibacterial Activity of Croconazole against Mycobacterium ulcerans

Abstract

An alamarBlue-based growth inhibition assay has been adapted for the thermosensitive and slow-growing pathogen Mycobacterium ulcerans. The standardized test procedure enables medium-throughput screening of preselected compound libraries. Testing of a set of 48 azoles with known antifungal activity led to the identification of an imidazole antifungal displaying an inhibitory dose (ID) of 9 μM for M. ulcerans.

TEXT

Mycobacterium ulcerans is the causative agent of Buruli ulcer, a disfiguring disease characterized by the formation of chronic, necrotizing skin ulcers. Buruli ulcer is a major public health problem, primarily in West Africa, where it typically affects impoverished inhabitants of remote rural areas (25). Surgical excision of lesions was the only therapeutic option until the World Health Organization released provisional guidelines in 2004 advising an 8-week combination chemotherapy regimen of oral rifampin and intramuscular streptomycin (26). Implementation of this regimen has significantly reduced recurrence rates. However, there are some drawbacks associated with the administration of these antibiotics. Streptomycin requires daily injections and is potentially nephrotoxic and ototoxic (5). Monotherapy with rifampin is not a solution because resistance can develop rapidly (14). While studies on fully oral regimens replacing streptomycin with clarithromycin yielded promising results (2, 6), identification of new drugs that could replace rifampin would be most desirable.

Since lead compounds identified by target-based approaches are frequently inactive against the pathogen grown in culture (4, 13, 19), whole-cell assays are highly valuable in the search for new antibiotics. The extremely low growth rate of M. ulcerans complicates the identification of potentially effective new scaffolds. In this report, we describe the development of an alamarBlue microplate assay which enables screening of compounds for their activity against M. ulcerans within 10 days. Here we report results of a pilot study performed with this medium-throughput screening procedure. Testing of a preselected group of azoles led to the detection of compounds with potency against M. ulcerans.

alamarBlue-based assays have been shown to be powerful and sensitive tools for the identification of growth-inhibitory compounds irrespective of the underlying mode of action and chemical classification. They are widely used for screens with both mammalian cells and microorganisms, including mycobacterial species (4, 7, 9, 10, 13, 19). In order to develop a suitable test format for M. ulcerans, the specific growth features, a long generation time of about 72 h and a low optimal growth temperature of 28 to 32°C, had to be taken into account. To reduce the problem that M. ulcerans tends to aggregate when grown in liquid medium, the inoculum was prepared in BacT/Alert culture bottles supplemented with enrichment medium according to the manufacturer's protocol (bioMérieux). NM20/02, a M. ulcerans strain isolated in 2002 from the Buruli ulcer lesion of a Ghanaian patient (22, 27), and additional strains from Togo (ITM 970680), Côte d'Ivoire (ITM 940511), and Australia (JS5147) were grown for 6 to 8 weeks at 30°C until the exponential phase was reached. The test inoculum was prepared by diluting the culture with BacT/Alert medium to a final optical density at 600 nm (OD₆₀₀) of about 0.02.

The alamarBlue growth inhibition assay was performed in clear, flat-bottomed 96-well plates (BD Falcon). Stock solutions of the tested compounds were prepared in high-grade dimethyl sulfoxide (DMSO) (Sigma) at a concentration of 10 mg/ml and diluted for the assay with DMSO to a concentration of 0.2 mg/ml. Then, samples in a 2-fold dilution series in medium from 12.5 μg/ml to 0.1953 μg/ml were prepared in a volume of 100 μl and set up in triplicate experiments. As a control, streptomycin (Sigma) at a concentration of 12.5 μg/ml to 0.1953 μg/ml was used. In addition, medium and DMSO controls were included. To each well of a 96-well plate, 100 μl of the diluted M. ulcerans suspension was added. Plates were incubated for 8 to 10 days at 30°C, and then 20 μl alamarBlue (Invitrogen) was added and the plates were further incubated and monitored at 37°C as recommended for the reduction of the alamarBlue component resazurin. During the short incubation period of 24 h, M. ulcerans viability was not affected by this elevated temperature (8). Since M. ulcerans is considered a biosafety level 3 (BSL3) pathogen in Switzerland, bacterial suspensions were afterwards inactivated for measurement of the fluorescence intensity outside the BSL3 laboratory. Samples from each well were transferred in 100-μl portions to corresponding wells of a new 96-well plate containing 100 μl neutral buffered formalin solution (Sigma) (final concentration, 5% [vol/vol]). The plates were sealed with optical adhesive film (Applied Biosystems), and the fluorescence intensity was measured using a SPECTRAmax Gemini XS device with Softmax Pro 5.2 software (λ = 540/588 nm). Signal reduction was assessed by comparison with M. ulcerans control cultures containing compound-free DMSO, and inhibitory dose (ID) values corresponding to 90% and 50% inhibition (ID₉₀ and ID₅₀, respectively) were determined.

For registered drugs, a complete toxicological profile is commonly available, but registered agrochemicals have also been broadly examined for their toxicological potential and for other features that have to be characterized prior to clinical phase I testing. Hence, compounds from both sources represent preferred candidates for medium-throughput testing for new indications. Here we have focused on azole-type fungicides, commercialized as drugs or agrochemicals. These compounds are cytochrome P450 inhibitors with partial specificity for the lanosterol 14-alpha-demethylase cytochrome P450 51 (CYP51). The genomes of M. tuberculosis and M. ulcerans (http://genodb.pasteur.fr/cgi-bin/WebObjects/GenoList.woa/wa) encode at least 20 CYP enzymes. CYP inhibitors have been shown to inhibit M. tuberculosis growth (15), and among them, CYP51 seems to be highly relevant (21). Therefore, azole-type fungicides might have potential as drug candidates for treatment of mycobacterial infections. We have retrieved 48 commercialized azoles from the BASF compound repository (see Table S1 in the supplemental material) and tested them in the alamarBlue-based screening assay for M. ulcerans growth-inhibitory activity. Of these 48 azoles, only the antifungal drug Croconazole {1-[1-[2-[(3-chlorophenyl)methoxy]phenyl]ethenyl]-1H-imidazole} (18) showed significant activity, with an ID₉₀ of 2.9 μg/ml, equivalent to 9 μM, and an ID₅₀ of 0.8 μg/ml, equivalent to 2.6 μM (Fig. 1A). In addition to the Ghanaian reference strain, we tested the activity of Croconazole on M. ulcerans isolates originating from Côte d'Ivoire, Togo, and Australia and obtained similar ID₅₀ values of 0.7 μg/ml, 1.1 μg/ml, and 0.6 μg/ml, respectively.

Fig 1.

Inhibition of M. ulcerans growth upon treatment with the azole compound Croconazole (A) and its 4-Cl analogue (B).

The characteristic feature of Croconazole, discriminating it from all other tested azoles, is a methylene substitution at the imidazole-binding carbon. To further evaluate the structure-activity relationship of this compound class, eight additional analogues possessing this structural motif have been tested. Of these, only the 4-Cl-isomer of Croconazole displayed similar activity, with an ID₉₀ of 2.8 μg/ml, equivalent to 8 μM, and an ID₅₀ of 0.95 μg/ml, equivalent to 2.7 μM (Fig. 1B and 2). While cytochrome binding via the imidazole is the essential interaction with the P450 enzyme, these data indicate also that the lipophilic side chain interaction is of critical importance, since even moderate side chain modifications lead to loss of activity. However, metabolic deactivation also cannot be ruled out. As all the other tested commercial azole drugs and agrochemicals, which show a very broad spectrum of substitution patterns and intrinsically high activity on a variety of CYPs, were inactive against M. ulcerans, the unique activity of Croconazole and its 4-Cl analogue is rather surprising.

Fig 2.

Overview of the Croconazole derivatives with corresponding CAS numbers and ID₉₀ values.

Croconazole is a potent imidazole antifungal that acts extracellularly on the cell membrane of fungi. It has been shown to inhibit predominantly CYP51 and thus interferes with the synthesis of ergosterol, the precursor of vitamine D₂ (11) The active sites of CYP51 are conserved in many mycobacterial species and virtually identical in M. tuberculosis and M. ulcerans, while there are many differences from human CYP51 (21). It remains to be elucidated whether Croconazole exerts its inhibitory activity on M. ulcerans via inhibition of CYP51 or via another target.

Interestingly, apart from being employed as a broad-spectrum fungicide (17, 24) developed by Shionogi Research Laboratories (Osaka, Japan), Croconazole is commercially also available as an active ingredient of a cream called Pilzcin (Merz and Co., Frankfurt/Main, Germany [16]), indicated for topical treatment of dermatomycoses and candidiasis. Pilzcin is in particular recommended for treatment of tinea pedis, an infection of the skin caused by fungi of the genus Trichophyton (3). Severe side effects upon treatment with Croconazole-containing drugs have not been described; however, a few cases of contact sensitization have been reported (23). Since Buruli ulcer primarily affects the skin and underlying tissue, topical treatment regimens have been considered and evaluated (1, 20). In active lesions, clusters of toxin-producing M. ulcerans are primarily found in deep layers of the subcutaneous fat tissue. Topical treatments thus may be effective only if the antimycobacterial agent is able to act in these deeper layers of the skin.

Since the development of new drugs is a very costly and time-consuming process, reevaluation of already-existing therapeutics for their potential effects against other diseases is a straightforward approach. A recent study focusing on the detection of new antituberculosis drugs by reevaluating a library of known human therapeutics led to the identification of promising new scaffolds (12). In a similar approach, we identified Croconazole as a potent compound active against M. ulcerans by revisiting a panel of 48 well-characterized azole drugs and agrochemicals. In conclusion, screening of such preselected sets of compounds increases the chance to detect scaffolds on which future drug development efforts can be focused.