A screening of the MMV Pathogen Box® reveals new potential antifungal drugs against the etiologic agents of chromoblastomycosis.

Chromoblastomycosis (CBM) is a chronic subcutaneous mycosis caused by traumatic implantation of many species of black fungi. Due to the refractoriness of some cases and common recurrence of CBM, a more effective and less time-consuming treatment is mandatory. The aim of this study was to identify compounds with in vitro antifungal activity in the Pathogen Box® compound collection against different CBM agents. Synergism of these compounds with drugs currently used to treat CBM was also assessed. An initial screening of the drugs present in this collection at 1 μM was performed with a Fonsecaea pedrosoi clinical strain according to the EUCAST protocol. The compounds with activity against this fungus were also tested against other seven etiologic agents of CBM (Cladophialophora carrionii, Phialophora verrucosa, Exophiala jeanselmei, Exophiala dermatitidis, Fonsecaea monophora, Fonsecaea nubica, and Rhinocladiella similis) at concentrations ranging from 0.039 to 10 μM. The analysis of potential synergism of these compounds with itraconazole and terbinafine was performed by the checkerboard method. Eight compounds inhibited more than 60% of the F. pedrosoi growth: difenoconazole, bitertanol, iodoquinol, azoxystrobin, MMV688179, MMV021013, trifloxystrobin, and auranofin. Iodoquinol produced the lowest MIC values (1.25-2.5 μM) and MMV688179 showed MICs that were higher than all compounds tested (5 - >10 μM). When auranofin and itraconazole were tested in combination, a synergistic interaction (FICI = 0.37) was observed against the C. carrionii isolate. Toxicity analysis revealed that MMV021013 showed high selectivity indices (SI ≥ 10) against the fungi tested. In summary, auranofin, iodoquinol, and MMV021013 were identified as promising compounds to be tested in CBM models of infection.

Introduction Chemical Corporation, St. Louis, MO, USA). Seven-day-old cultures incubated at 30˚C were used in the assays.

The Pathogen Box1 compounds
The Pathogen Box1 was kindly provided by Medicines for Malaria Venture (MMV, Geneva, Switzerland). It contains 400 different compounds tested for cytotoxicity with values within levels considered acceptable for an initial drug discovery programme [19]. The Pathogen Box1 contains compounds with proven activity against neglected pathogens, including those causing tuberculosis, malaria, helminthiasis, cryptosporidiosis, toxoplasmosis, and dengue [20][21][22][23]. The compounds were supplied in 96-well microtiter plates containing 10 μl/well of 10 mM compound solutions in dimethylsulfoxide (DMSO). The plates were diluted to a final drug concentration of 1 mM in DMSO (Sigma Chemical Corporation, St. Louis, MO, USA), as recommended by the fabricant, for the drug screening. The first and last columns in each plate were left as blank for negative and positive controls, respectively. All plates were stored at −20˚C until their use in the following experiments.

Screening for antifungal activity
For the initial screening, all compounds were tested in 96-well plates (Kasvi Ltda, São José dos Pinhais, PR, Brazil) at a final concentration of 1μM in 100 μl of RPMI 1640 medium, with phenol red, with L-glutamine, and without sodium bicarbonate (Sigma Chemical Corporation), buffered with morpholine propanesulfonic acid (MOPS) (Vetec Química Fina Ltda, Rio de Janeiro, RJ, Brazil) at pH 7.0, and supplemented for a final 2% glucose (Neon Comercial Ltda, São Paulo, SP, Brazil) concentration. The antifungal drugs available as reference compounds in the MMV collection were used as controls of fungal growth inhibition. DMSO concentration in all wells, including those used for fungal control growth, at this point, corresponded to 1%. For preparation of the fungal inoculum, the F. pedrosoi strain CFP 00791 was grown as described above and then its conidia were suspended in sterile distilled water supplemented with 0.1% Tween 20 (Sigma Chemical Corporation) and vortexed, with the suspension turbidity adjusted to the 0.5 McFarland scale. This suspension was further diluted 1:10 and then 100 μl of the fungal inoculum was added to each well containing the compounds, generating a final working inoculum density of 2-5 × 10 5 CFU/ml and a final DMSO concentration of 0.5% in each well, including controls [21]. Plates were incubated at 35˚C for 72-96 hours. The optical density at 530 nm (OD530) was recorded using the Epoch microplate reader (Biotek Instruments Inc, Winooski, VT, USA). The percentage of inhibition of fungal growth (%IG) was calculated according to the formula: %IG = (1-(OD1 / OD2)) × 100, where OD1 = fungal optical density in the presence of the drug; OD2 = optical density of the fungal growth control well without any drug. The %IG data of the 400 compounds was plotted using the Prism 8 software (GraphPad Software, San Diego, CA, USA). Compounds that presented %IG values greater than 60% were selected for further assays.

Minimal inhibitory concentration (MIC) of selected drugs
Assays for MIC determination were performed in sterile polystyrene flat-bottom 96-well microtiter plates using the broth microdilution method according to the EUCAST guidelines [24]. The final concentration of the selected compounds of the Pathogen Box1 ranged from 0.015 to 10 μM. All isolates were tested at a final cellular density of 2 to 5 × 10 5 CFU/ml. Plates were incubated at 35˚C for 72-96 h. All experiments were repeated at least two times. The MIC was defined as the lowest drug concentration (μM) that inhibited 100% of fungal growth.

Fungicidal activity
The minimal fungicidal concentration (MFC) was determined by transferring an aliquot of 5 μl of each well without fungal growth of the microdilution plates used for the determination of MIC, as described above, on Sabouraud 2% glucose agar (Sigma Chemical Corporation). The MFC was determined as the lowest drug concentration without fungal growth on the Sabouraud agar after five days of incubation at 35˚C. When the MFC:MIC ratio of an agent is 1 or 2, the compound is considered fungicidal against the pathogen. If the ratio is higher than 2, the mode of action is probably fungistatic [25,26].

Evaluation of compounds applicability in CBM treatment
Cytotoxicity data of the compounds was provided by the MMV. The selectivity index (SI) was calculated as follows: SI = CC 50 (μM) / MIC (μM), where CC 50 is the drug concentration that kills 50% of the cells tested. Compounds with selectivity index lower than 1 were not considered for the synergistic assays. The original use of the compounds and their known side-effects were also verified in the literature, to evaluate their safety for human use. Compounds found to have dangerous side-effects were further excluded from the analysis.

Synergism evaluation
Synergistic activity between the selected compounds and standard antifungal drugs used in the treatment of CBM (itraconazole and terbinafine) was tested by the checkerboard method [27]. Two drugs were loaded into a single 96-well plate, so that in each of the wells there were different concentrations of the compound-antifungal combination. Compound/antifungal dilutions were prepared following the methodology proposed by the EUCAST, starting from a 100-fold concentrated stock compound/antifungal solution according to the MIC determination methodology [24]. Final concentrations of the standard drugs corresponded to 0.06-4 μg/ml (itraconazole) and 0.015-1 μg/ml (terbinafine). Serial two-fold dilutions of the selected Pathogen Box1 compounds were performed so that the MIC of that compound was in a central position of the plate (lines 5-7). Fungal inocula and incubation conditions were the same that were described for the previous experiments. The drug interaction was classified according to the fractional inhibitory concentration index (FICI). The FICI was obtained by the formula:  [28]. The type of interaction between the antifungal agents in combination was classified as the following: synergism if FICI � 0.5; indifference if 0.5 < FICI < 4 and antagonism if FICI � 4 [27,28].

Identification of compounds with antifungal activity against CBM agents
The internal reference drug posaconazole inhibited 99% of the F. pedrosoi growth, validating the assay. Amphotericin B, which is also present in the drug collection, inhibited F. pedrosoi growth by 29%. Among the 400 compounds present in the Pathogen Box1, another eight drugs were found to have antifungal activity against the F. pedrosoi CFP00791 strain (Fig 1). These compounds were identified as difenoconazole (%IG = 98%), bitertanol (%IG = 98%), iodoquinol (%IG = 99%), azoxytrobin (%IG = 95%), MMV688179 (%IG = 73%), MMV021013 (%IG = 62%), trifloxystrobin (%IG = 82%), and auranofin (%IG = 82%). The chemical structure of these compounds is depicted in Fig 2. Their original applications, as described by Medicines for Malaria Venture, are as follows: difenoconazole, bitertanol, MMV688179, and trifloxystrobin have anti-kinetoplastid action, iodoquinol has anti-onchocerciasis action, azoxystrobin has anti-malaria action, MMV021013 has anti-tuberculosis action, and auranofin, a regulated drug for rheumatoid arthritis treatment, is another reference compound present in the collection. Table 1 depicts the MIC and MFC values for each selected compound and the eight CBM agents tested. In brief, iodoquinol (MMV002817) presented the lowest MIC values among the compounds that we tested for antifungal activity. Seven out of the eight strains had MIC values corresponding to 1.25 μM and one MIC was equal to 2.5 μM to this compound. MMV688179 showed the highest MIC values against all species (5 ->10μM). Auranofin (MMV688978) was fungicidal against P. verrucosa and E. dermatitidis, and presented fungistatic activity against the other species. Iodoquinol (MMV002817) and bitertanol (MMV688942) were fungicidal against P. verrucosa and fungistatic against all other species. MMV688179 was fungicidal against P. verrucosa and F. monophora, and fungistatic against the other species. All other compounds had fungistatic activity against the set of isolates tested in this study.

Analysis of synergistic effects
The compound MMV688179 was excluded from further analyses because it showed the highest MIC values (MIC >10 μM) and low selectivity index (Table 2). Trifloxystrobin was found to be toxic to keratinocytes [29], discouraging its use to treat CBM, which affects the skin. From the set of the other agricultural fungicides, we selected bitertanol for synergistic studies, on the basis of its higher selectivity index (Table 2). Therefore, iodoquinol, bitertanol, MMV021013, and auranofin were selected for the determination of FICI in the in vitro combination test with itraconazole and terbinafine. The concentration ranges of the compounds tested corresponded to 0.01-6 μg/ml for auranofin, 0.003-2.5 μg/ml for MMV021013, 0.005-3 μg/ml for bitertanol, and 0.007-4 μg/ml for iodoquinol. The results obtained from the checkerboard method with these four compounds in combination with itraconazole or terbinafine   Tables 3 and 4, respectively. The auranofin and itraconazole combination produced the single synergistic interaction (FICI < 0.5), against the C. carrionii strain. The combinations of iodoquinol, bitertanol, and MMV021013 with itraconazole produced indifferent interactions (FICI between 0.5 and 4) for all isolates tested. All compounds in combination with terbinafine produced indifferent interactions.

Discussion
CBM is difficult to treat because there is no standardized drug of choice. In addition, relapses are frequent for this mycosis [13]. The taxonomic diversity of the CBM agents adds another obstacle to the development of new drugs against this disease, because most patients are diagnosed exclusively on the basis of the presence of muriform cells in infected tissues [2]. Therefore, general therapeutic decisions are made regardless of the etiology of CBM, which may be deleterious for the patient. In this study, for instance, E. dermatitidis often presented MIC values higher than other species for the studied drugs. In this context, we recommend that repurposing studies for CBM use a large number of etiologic agents of this disease in the antifungal activity experiments. The compound collection used in this study, the Pathogen Box1 library, was previously demonstrated to contain molecules with antifungal activity. Mayer and Kronstad identified five compounds (tolfenpyrad, difenoconazole, bitertanol, posaconazole, and MMV688271) that exhibited antifungal activity against Cryptococcus neoformans and Candida albicans [18]. In our study, antifungal activities were identified for three of these compounds (difenoconazole, posaconazole, and bitertanol). It is noteworthy that tolfenpyrad inhibited F. pedrosoi growth by 49%. This compound is an agricultural insecticide and with well-known poisoning effects [30]. MMV688271, which completely inhibited the growth of C. neoformans and C. albicans [18], showed only 36% inhibition against F. pedrosoi in our screening assay. Vila and Lopez-Ribot evaluated three of the compounds that we investigated (MMV687807, MMV687273, and MMV688768) against C. albicans biofilms [17]. In the present study, these compounds showed 13% (MMV687807), 37% (MMV687273) and 24% (MMV688768) inhibition of the F. pedrosoi growth. Wall et al identified three compounds with anti-C. auris activity in the Pathogen Box1 at a 5 μM concentration, including pentamidine, MMV659010, and iodoquinol. In this group, iodoquinol and MMV659010 showed %IG of 86 and 62, respectively, against plant cells. Pentamidine showed 61% inhibition in biofilms of C. auris [31]. In the present study, these compounds presented 34% (MMV659010), 49% (pentamidine), and 99% (iodoquinol) inhibition of the F. pedrosoi growth. These results clearly demonstrate that major fungal pathogens respond differently to the Pathogen Box1 compounds (S1 Table).
The MMV Pathogen Box1 contains two antifungal drugs as reference compounds. Inhibition of F. pedrosoi growth by posaconazole (99%) was observed, but amphotericin B inhibited less fungal growth (29%). This data can be explained by the high MIC values of amphotericin B (4-16 μg/ml) against this species [14]. In our study, three compounds were selected as promising anti-CBM agents: auranofin, iodoquinol, and MMV021013. Auranofin has been in clinical use as the gold standard for the oral treatment of rheumatoid arthritis since the 1980s [32,33]. The oral bioavailability of auranofin and its reduced side effects offer significant advantages over traditional injectable drugs [32,33]. Several studies were conducted to identify alternative therapeutic applications for auranofin, particularly in the area of infectious diseases. For example, antifungal activity against C. albicans [34] and C. neoformans [34][35][36] has been described. In this study, low MIC values for this drug against eight different CBM agents were observed. Furthermore, the synergism found in the combination of auranofin with itraconazole against C. carrionii, the main CBM agent in arid and semi-arid areas [37], as well as its fungicidal activity against E. dermatitidis support the use of auranofin as a therapeutic option for CBM. These findings justify future in vivo testing and clinical studies to evaluate auranofin as an antifungal drug.
Iodoquinol (diiodohydroxyquinoline) is a luminal amebicide. Its use is recommended after treatment with a tissue amebicide, such as metronidazole, to eradicate surviving parasites in the colon [38]. This drug has low absorption through the gastrointestinal tract, so only the topical formulation would be applicable against CBM. Burnett and Mitchell demonstrated that the topical formulation of 1% iodoquinol in combination with 2% hydrocortisone acetate had antifungal activity against C. albicans, dermatophytes, and Malassezia furfur [39]. The study by Wall et al. demonstrated the in vitro efficacy of iodoquinol against several Candida species under planktonic conditions. As expected from the initial results with C. auris, iodoquinol was virtually ineffective against preformed biofilms of all species tested. However this compound displayed potent activity against planktonic growth of all strains tested, with IG 50 (50% inhibition growth) values generally lower than 1 μg/ml [31]. In the present study, lower MIC values of iodoquinol (100% inhibition) were observed against all studied CBM agents, supporting its use against this group of fungi.
In summary, auranofin was identified in our study as a promising compound to be tested as a CBM therapeutic agent, on the basis of its low MIC values against CBM agents, synergism with itraconazole against a C. carrionii strain, and fungicidal action against P. verrucosa and E. dermatitidis. Iodoquinol has also proved promising for future studies involving CBM agents due to its low MIC values, fungicidal activity against P. verrucosa, and approved topical use in some countries. The 2-pyridyl-4-aminopyrimidine MMV021013 showed good selectivity and a combined activity against mycobacteria, Leishmania, Trypanosoma, and CBM agents. These data suggest the need for further in vitro and in vivo studies with these drugs aiming to develop new therapeutic tools against the CBM agents.
Supporting information S1 Table. Screening of 400 compounds present in the MMV Pathogen Box1 collection against a Fonsecaea pedrosoi strain (CFP00791). Results are presented as the percent inhibition of growth (%IG) of the fungal strain. (PDF)