Identification of four compounds from the Pharmakon library with antifungal activity against Candida auris and species of Cryptococcus

Abstract

There is an urgent need to develop novel antifungals. In this study, we screened 1600 compounds for antifungal activity against Cryptococcus neoformans and Candida auris. We evaluated 4 promising compounds against 24 additional isolates of Cr. neoformans, Ca. auris, Cr. deuterogattii, and Cr. gattii. The four compounds, dequalinium chloride (DQC), bleomycin sulfate (BMS), pentamidine isethionate salt (PIS), and clioquinol (CLQ), varied in their efficacy against these pathogens but were generally more effective against cryptococci. The compounds exerted their antifungal effect via multiple mechanisms, including interference with the capsule of cryptococci and induction of hyphal-like morphology in Ca. auris. Our results indicate that DQC, BMS, PIS, and CLQ represent potential prototypes for the future development of antifungals.

Lay Summary

Fungal infections can be lethal and the options to fight them are scarce. We tested 1600 molecules for their ability to control the growth of two important fungal pathogens, namely Candida auris and species of Cryptococcus. Four of these compounds showed promising antifungal activities.

Introduction

The need for novel antifungals is indisputable. Fungal infections kill more than 1 million people every year, and the therapeutic options are frequently toxic and/or unaffordable in countries where neglected populations are more numerous.¹ There are currently no licensed antifungal vaccines and the pace of innovation in antifungal development is much slower than in other pharmaceutical areas.^1,2 Indeed, the generation of knowledge in the field of antifungals is significantly slower than that observed in antibiotic drug development (Fig. 1). A direct consequence of this discrepancy is that there are fewer classes of antifungal drugs compared to antibacterial drugs.

Figure 1.

Analysis of the number of publications (Scholarly Output) in the areas of antibacterial and antifungal development from 2016 to 2021. This demonstrates the markedly accelerated pace of knowledge generated in the field of antibacterial compared to antifungal drug development. Data were obtained from searches of publications combining the terms ‘antifungal activity’ or ‘antibacterial activity’ in the SciVal platform (www.scival.com, Elsevier) using the ‘Trends’ and ‘Benchmarking’ tools (data cutoff January 28, 2022).

The development of novel drugs is time-consuming (up to 20 years before clinical use) and expensive; additionally, there is a low rate of success for bringing novel compounds into clinical utilization.³ This is an obvious obstacle in the fight against fungal diseases. However, several other hurdles have arisen, including the emergence of new pathogens and the spread of antifungal resistance.⁴  Candida auris is an example of a recently identified fungal pathogen that is the first fungus identified by both the United States Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) as a major threat to global public health. Candida auris was first identified in Japan in 2009 from the external ear canal of a female patient.⁵ In only one decade, Ca. auris has become a global threat because it exhibits resistance to multiple antifungals, including pan-resistance to standard antifungals, and the rapid dissemination of this pathogen globally.⁶

The problem of antifungal resistance is not exclusive to Ca. auris. For example, cryptococcal meningitis kills almost 200 000 people every year.⁷ Amphotericin B is the most effective anti-cryptococcal agent.⁸ However, the less toxic lipid-based forms of amphotericin B are too expensive,⁹ and the drug is not available in regions where deaths due to Cryptococcus and cryptococcosis are more concentrated.¹⁰ This situation has led to the use of other less potent drugs to fight cryptococcosis, but the consequences are very deleterious. For instance, not only are there more treatment failures, but clinical isolates of Cryptococcus neoformans from South Africa have become twofold more resistant to fluconazole in only one decade.¹¹ Therefore, new tools to fight existing and emerging fungal pathogens are needed.

Screening large compound collections with the goal of identifying approved drugs to fight fungal infections has been a promising approach.^12–16 In this manuscript, we screened a collection of 1600 clinically tested compounds for antifungal activity against Ca. auris and pathogenic species of Cryptococcus. Our study differs from the previously reported screens because our main criterion for drug selection was how effective the tested compounds were against three lethal fungal pathogens (Cr. neoformans, Cr. deuterogattii, and Ca. auris). Based on this experimental approach, we identified four promising compounds: dequalinium chloride (DQC), bleomycin sulfate (BMS), pentamidine isethionate salt (PIS), and clioquinol (CLQ). These four compounds differed in their effect on fungal morphology and their antifungal efficacy, but all four compounds controlled the growth of Cr. neoformans, Cr. deuterogattii, and Ca. auris.

Materials and methods

Compounds

We screened the Pharmakon 1600 compound library (MicroSource Discovery Systems, Inc., Gaylordsville, CT, USA, http://www.msdiscovery.com/index.html) for antifungal activity. This library contains 1600 clinically tested drug compounds approved for use in the United States and internationally. The compounds were provided in 96-well plates at a concentration of 1 m m in dimethyl sulfoxide (DMSO).

Fungal pathogens

The Pharmakon 1600 compound library was evaluated for antifungal activity against the strains Cr. neoformans H99 and Ca. auris MMC2.¹⁷ For the follow-up experiments, 24 additional isolates from different fungal species were tested. The Cr. neoformans isolates were 3Pb3 (environmental), 176A1 (environmental), 23Pb2 (environmental), 19Pb4 (environmental), Cg366 (environmental), Cn161 (human liquor), Cn222 (human liquor), and Cn116 (human liquor). The Cr. deuterogattii isolates were the standard isolates R265, Cg460 (human liquor), Cg221 (human forearm secretion), Cg158 (human liquor), Cg456 (human liquor), and Cg188 (human liquor). Cryptococcus gattii isolates included Cg365 (environmental), Cg367 (unknown source), and Cg306 (unknown source). The Ca. auris isolates tested (all isolated from the blood of human patients) were MMC1 (clade I),¹⁷ and the CDC-curated isolates CDC383 (clade III), CDC388 (clade I), CDC384 (clade III), CDC390 (clade I), CDC387 (clade I), and CDC385 (clade IV) (https://www.cdc.gov/drugresistance/resistance-bank/index.html). The cryptococcal isolates were characterized previously at the genomic and phenotypic levels¹⁸ and were available in our laboratory. Fungal strains were maintained in Sabouraud agar at 30°C.

Screening of antifungal activity

The original stock solutions of the 1600 compounds were diluted 1:20 (v/v) in twice concentrated (2×) Roswell Park Memorial Institute medium (RPMI-1640, Sigma-Aldrich, Saint Louis, MO, USA) supplemented with 2% glucose, buffered (pH 7.0) with 165 m m morpholinepropanesulfonic acid (MOPS). This procedure generated intermediate plates containing the compounds at 50 μm, 5% DMSO. These intermediate plates were further diluted 1:2.5 (v/v) in 2 × RPMI-1640 medium supplemented with 2% glucose. This procedure generated working solutions at 20 μm, 2% DMSO in a volume of 100 μl. Each well of these plates was inoculated with Cr. neoformans H99 or Ca. auris MMC2 suspended in sterile water following the procedures described in the EUCAST antifungal susceptibility testing (AFST, E.DEF 7.3) protocol.¹⁹ The final fungal density was 2.5 × 10⁵ cells/ml, and the final concentration of the compounds was 10 μm, 1% DMSO. All plates had a set of controls: (1) sterility control (2 × RPMI-1640 medium supplemented with 2% glucose, buffered (pH 7.0) with 165 m m MOPS, and 2% DMSO, plus sterile water, no fungi); (2) growth control (2 × RPMI-1640 medium supplemented with 2% glucose, buffered (pH 7.0) with 165 m m MOPS, and 2% DMSO, plus Cr. neoformans H99 or Ca. auris MMC2, no compounds); and (3) antifungal activity control (amphotericin B at 0.5 mg/l). Of note, the Pharmakon plates already contained known antifungal compounds (ketoconazole, miconazole, voriconazole, fluconazole, and itraconazole), which served as additional controls for fungal growth inhibition. After inoculation, the plates were incubated at 35°C for 24 or 48 h for Ca. auris or Cr. neoformans, respectively. Fungal growth was assessed via measuring optical density at 530 nm. The compounds were considered active if they inhibited more than 50% of fungal growth. For follow-up experiments, the purified compounds were obtained from Sigma–Aldrich, Saint Louis, MO, USA.

Determination of the minimum inhibitory concentration (MIC), and minimum fungicidal concentration (MFC) for the selected compounds

The MIC was determined according to the EUCAST antifungal susceptibility testing (AFST, E.DEF 7.3) protocol.¹⁹ Briefly, 96-well plates were prepared to contain the compounds at concentrations ranging from 0.019 to 10 μm in 100 μl of RPMI-1640 medium supplemented with 2% glucose and buffered (pH 7.0) with 165 m m MOPS. These plates were inoculated with 100 μl of fungal suspensions containing 2.5 × 10⁵ cells/ml prepared in sterile water. The plates were incubated at 35°C for 24 or 48 h for Ca. auris or Cryptococcus spp., respectively. The optical density at 530 nm was determined spectrophotometrically, and the MIC was defined as the minimum concentration that inhibited >90% of fungal growth when compared to the control without any compound.

To determine the MFC, 5 μl of each well from the MIC plate was transferred to Sabouraud agar plates and incubated at 30°C for 24 or 48 h for Ca. auris or Cryptococcus spp., respectively. The MFC was categorized as the minimum concentration where no fungal growth was observed. The MIC was determined for all isolates described in the ‘Fungal pathogens’ section. The MFC was determined for C. auris isolates MMC1 and MMC2, Cr. neoformans H99, and Cr. deuterogattii R265.

Determination of the half maximal inhibitory concentration (IC50) using macrophages

The IC50 was determined in macrophages using the Cytotox 96 nonradioactive cytotoxicity assay kit (catalog number G1780; Promega, Madison, WI, USA) as we recently described.²⁰ Briefly, RAW 264.7 macrophages (10⁵ cells/well in DMEM supplemented with 10% FBS) were treated with each selected compound (0–10 μm) for 24 h at 37°C in DMEM supplemented with 10% FBS. Cytotoxicity was inferred from the determination of the levels of lactate dehydrogenase activity in the medium. Control systems included vehicle-treated cells (viability control) and macrophages lysed with the lysis solution provided by the manufacturer (death control). The IC50 was defined as the lowest drug concentration inducing death in 50% of the macrophage population.

Effects of the compounds on fungal morphology

For the analysis of fungal morphology, Cr. neoformans H99, Cr. deuterogattii R265, Ca. auris MMC1, and Ca. auris MMC2 were cultivated overnight on Sabouraud broth at 30°C with shaking. The cells were washed three times with phosphate-buffered saline (PBS), and the inoculum was adjusted to 2.5 × 10⁵ cells/ml. For Cryptococcus spp., the experiments were performed in capsule induction medium (10% Sabouraud diluted in MOPS to pH 7.0).²¹ For Ca. auris, Sabouraud broth was used. The compounds were added to the culture at a subinhibitory concentration (MIC divided by 2). The concentration of DMSO was maintained at 1% in all cases, including for the control system (no compound). The systems were incubated at 37°C under a 5% CO₂ for 24 h. The cultures were recovered, washed three times, and processed for fluorescence, scanning electron, light, or time-lapse microscopy.

Fluorescence microscopy

Yeast cells were fixed with 4% paraformaldehyde and washed three times with PBS. The fixed cells were blocked with 1% bovine serum albumin (BSA) in PBS for 1 h at 37°C, following chitin staining of the cell wall with 25 μm calcofluor white (CFW, Sigma-Aldrich, Saint Louis, MO, USA) for 30 min at 37°C. The cells were washed three times with PBS and incubated with 5 μg/ml wheat germ agglutinin and tetramethylrhodamine conjugate (WGA-TRITC) in PBS for 30 min at 37°C. For Cryptococcus spp., the cells were washed three times before WGA-TRITC staining; the capsule was stained after incubation with the capsule-binding 18B7 monoclonal antibody (donated by Arturo Casadevall, Johns Hopkins Bloomberg School of Public Health) at 10 μg/ml in 1% BSA for 1 h at 37°C. The cells were washed with PBS and incubated with a secondary anti-IgG antibody conjugated to Alexa 488 (Invitrogen, Waltham, MA, USA) for 1 h at 37°C. The cells were analyzed under a DMi8 microscope (Leica, Wetslar, HE, Germany). Images were recorded with LasAF software (Leica, Wetslar, HE, Germany) and processed with ImageJ.²² Since Ca. auris staining with WGA-TRITC produced weak signals of red fluorescence, these signals were amplified using Adobe Photoshop 2020. All systems were submitted to the same conditions required for signal amplification.

Scanning electron microscopy (SEM) analysis

Control or compound-treated cells were washed three times with PBS and fixed with 2.5% glutaraldehyde in 0.1 m sodium cacodylate buffer (pH 7.2) for 1 h at room temperature. The cells were washed three times with 0.1 m sodium cacodylate buffer (pH 7.2) containing 0.2 m sucrose and 2 m m MgCl₂. The cells were placed over 0.01% poly-L-lysine-coated coverslips and allowed to adhere for 1 h at room temperature. The cells were then dehydrated in ethanol (30, 50, and 70% for 5 min, followed by 90% for 10 min, and 100% twice for 10 min). Dehydrated cells were immediately critical point dried (Leica EM CPD300, Wetslar, HE, Germany), mounted on metallic bases, and coated with a gold layer (Leica EM ACE200, Wetslar, HE, Germany). The cells were visualized using an SEM (JEOL JSM-6010 Plus/LA, Akishima, Tokyo, Japan) operating at 10 keV.

Light microscopy

Compound-treated cryptococci were observed for analysis of the capsule under light microscopy. Paraformaldehyde-fixed cells were counterstained with India Ink and observed under a DMi8 microscope (Leica, Wetslar, HE, Germany). Images were recorded with LasAF software (Leica, Wetslar, HE, Germany), and the cell body and capsule dimensions were determined in digitalized images using ImageJ.²²

Time-lapse microscopy

Based on morphological effects observed through different approaches, the effect of BMS on the morphology of Ca. auris was further evaluated by time-lapse microscopy. Candida auris (MMC1 and MMC2 strains, 2.5 × 10⁵ cells/ml on Sabouraud agar) was treated with 5 μm BMS and incubated at 37°C for 12 h on an Operetta high-content imaging system (PerkinElmer, Waltham, MA, USA). During the 12 h incubation period, images were captured every 5 min with a 40 × objective. Movies of each well were prepared using the Harmony high-content imaging and analysis software (PerkinElmer, Waltham, MA, USA) and ImageJ.²²

Statistical analysis

Statistical analyses were performed using GraphPad Prism (GraphPad Software, San Diego, CA, USA), and the results were considered significant for P values < 0.05. In the capsule size assay, the results were analyzed using a one-way anova and Tukey's post-hoc tests.

Results

Antifungal activity against Cr. neoformans and Ca. auris in the Pharmakon compound collection

Among the 1600 compounds that we tested, 69 were active against Cr. neoformans H99 (Fig. 2). Thirty-two compounds inhibited growth of Cr. neoformans by 50–90% while 37 compounds inhibited fungal growth by more than 90%. Against Ca. auris MMC2, 61 out of the 1600 compounds were active (Fig. 2). Twenty-four compounds inhibited growth of Ca. auris between 50 and 90%, while 37 compounds inhibited fungal growth by more than 90%. A total of 38 out of the 1600 compounds inhibited the growth of both Cr. neoformans H99 and Ca. auris MMC2 by 50% (Table 1). These molecules corresponded to known antifungals, antibiotics, antineoplastic, and anthelmintic drugs, among others. For further tests, we eliminated compounds that were known antifungals, compounds approved only for topical use, and compounds with previously reported activity against Cr. neoformans and/or Ca. auris at the time of our screening.

Figure 2.

Antifungal activity of 1600 compounds from the Pharmakon library screened against Cr. neoformans H99 (black circles) or Ca. auris MMC2 (green triangles). Compounds with known antifungal activity are boxed on the right side. The symbols (circles and triangles) represent the antifungal effect of each compound in the collection. Most compounds exhibit activity below the 50% cutoff value for growth inhibition. We selected four compounds for further tests based on their ability to inhibit the growth of Cr. neoformans H99 and Ca. auris MMC2 by 70–100%. These compounds are indicated with arrows (dequalinium chloride, DQC; bleomycin sulfate, BMS; pentamidine isethionate salt, PIS; clioquinol, CLQ).

Table 1.

Compounds inhibiting 50% or more of the growth of Cr. neoformans H99 and Ca. auris MMC2 in the Pharmakon 1600 compound collection.*

		Growth inhibition (%)
Compound	Original reported activity	Cr. neoformans H99	Ca. auris MMC2
Dichlorophen	Anthelmintic	84.3	70.6
Triclabendazole	Anthelmintic	84.1	53
Bithionate sodium	Anthelmintic, antiseptic	98.4	74.6
Iodoquinol	Antiamebic	94.5	96.2
Alexidine hydrochloride	Antibacterial (topical mouthwashes)	100	55.1
Pyrithione zinc	Antibacterial, antifungal, Antiseborrheic (topical)	100	100
Amphotericin B	Antifungal	100	99.1
Butoconazole	Antifungal	100	98.4
Ciclopirox olamine	Antifungal	100	99.6
Climbazole	Antifungal	87.4	91.7
Clotrimazole	Antifungal	90.4	96.5
Econazole nitrate	Antifungal	93.2	91
Enilconazole sulfate	Antifungal	60.5	88.7
Fluconazole	Antifungal	74.4	91.5
Itraconazole hydrochloride	Antifungal	69.7	85.7
Ketoconazole	Antifungal	90.4	95.8
Miconazole nitrate	Antifungal	98	97
Oxiconazole nitrate	Antifungal	100	92.1
Sulconazole nitrate	Antifungal	100	98.1
Terbinafine hydrochloride	Antifungal	83.3	65.6
Terconazole	Antifungal	82.5	96.7
Tioconazole	Antifungal	90.8	90.8
Voriconazole	Antifungal	83.6	94.8
Phenylmercuric acetate	Antifungal, antimicrobial	100	99.3
Cetylpyridinium chloride	Antiinfectant (topical)	100	100
Dequalinium chloride	Antiinfectant, antineoplasic	98.6	86.8
Broxyquinoline	Antiinfectant, disinfectant	99.5	100
Methylbenzethonium chloride	Antiinfective (topical)	100	61.4
Hexachlorophene	Antiinfective (topical)	100	64
Bleomycin sulfate	Antineoplastic	94.1	94.3
Ebselen	Antioxidant, lipoxygenase inhibitor, inhibits oxidation of LDL	84.3	99.9
Pentamidine isethionate	Antiprotozoal; inhibits nucleic acid and protein synthesis	98	90.9
Cycloheximide	Antipsoriatic, protein synthesis inhibitor	100	92.8
Piroctone olamine	Antiseborrheic (topical)	100	99.3
Clioquinol	Antiseptic, antiamebic	99.7	98.8
Chloroxine	Chelating agent, antiseborrheic (topical)	100	99.6
Thonzonium bromide	Mucolytic, antibacterial, surface active agent	100	99.4
Suloctidil	Peripheral vasodilator	51.2	62.4

^*For information on the selection of compounds for further testing, see the Results section.

For selection of the most promising compounds, we eliminated standard antifungals or compounds exclusively used topically, because cryptococcosis and Ca. auris candidiasis are systemic infections. We selected molecules that were active against both Cr. neoformans and Ca. auris, preferentially causing growth inhibition equal to or higher than 90%. Based on these criteria, four compounds were selected: DQC, BMS, PIS, and CLQ. The four compounds adhered to our selection criteria with the exception of PIS (90% inhibition of Ca. auris growth but approximately 70% inhibition of Cr. neoformans growth). We selected PIS for further experiments because of its ability to inhibit growth of both Cr. neoformans and Ca. auris. The structures of these four compounds are illustrated in Figure 3.

Figure 3.

Structures of the four compounds with antifungal activity selected for further evaluation in the current study.

Antifungal susceptibility testing of the selected compounds

Using the EUCAST protocol,¹⁹ we determined the MIC values of DQC, BMS, PIS, and CLQ against 26 isolates from 4 species: Cr. neoformans (3Pb3, 176A1, 23Pb2, 19Pb4, Cg366, Cn161, Cn222, and Cn116), Cr. deuterogattii (R265, Cg460, Cg221, Cg158, Cg456, and Cg188), C. gattii (Cg365, Cg367, and Cg306), and Ca. auris (MMC1, CDC383, CDC388, CDC384, CDC390, CDC387, and CDC385). The results detailed below are summarized in Table 2.

Table 2.

MIC determinations of the four selected compounds (dequalinium chloride, DQC; bleomycin sulfate, BMS; pentamidine isethionate salt, PIS; clioquinol, CLQ) against multiple isolates of Cr. neoformans, Cr. deuterogattii, and Ca. auris.

	DQC		BMS		PIS		CLQ
Isolate	μm	μg/mL	μm	μg/ml	μm	μg/ml	μm	μg/ml
Cr. neoformans H99	1.25	0.375	10	15.1	10	5.9	1.25	0.194
Cr. neoformans 3Pb3	1.25	0.375	2.5	3.775	10	5.9	1.25	0.194
Cr. neoformans 176A1	1.25	0.375	2.5	3.775	5	2.95	1.25	0.194
Cr. neoformans 23Pb2	1.25	0.375	2.5	3.775	10	5.9	1.25	0.194
Cr. neoformans 19Pb4	1.25	0.375	5	7.55	10	5.9	1.25	0.194
Cr. neoformans Cg366	1.25	0.375	5	7.55	5	2.95	1.25	0.194
Cr. neoformans Cn161	1.25	0.375	5	7.55	10	5.9	0.625	0.097
Cr. neoformans Cn222	1.25	0.375	5	7.55	10	5.9	0.625	0.097
Cr. neoformans Cn116	1.25	0.375	5	7.55	10	5.9	0.625	0.097
Cr. deuterogattii R265	1.25	0.375	2.5	3.775	5	2.95	1.25	0.194
C. gattii Cg365	1.25	0.375	2.5	3.775	10	5.9	0.625	0.097
C. gattii Cg367	1.25	0.375	2.5	3.775	10	5.9	0.625	0.097
Cr. deuterogattii Cg460	1.25	0.375	2.5	3.775	5	2.95	0.625	0.097
Cr. deuterogattii Cg221	1.25	0.375	2.5	3.775	5	2.95	0.625	0.097
Cr. deuterogattii Cg158	1.25	0.375	2.5	3.775	5	2.95	0.625	0.097
Cr. deuterogattii Cg456	1.25	0.375	2.5	3.775	5	2.95	0.625	0.097
Cr. deuterogattii Cg188	1.25	0.375	2.5	3.775	5	2.95	0.625	0.097
C. gattii Cg306	1.25	0.375	2.5	3.775	5	2.95	1.25	0.194
Ca. auris MMC1	>10	>5.3	>10	>15.1	>10	>5.9	1.25	0.194
Ca. auris MMC2	10	5.3	10	15.1	5	2.95	1.25	0.194
Ca. auris CDC383	>10	>5.3	>10	>15.1	>10	>5.9	2.5	0.387
Ca. auris CDC388	>10	>5.3	>10	>15.1	>10	>5.9	2.5	0.387
Ca. auris CDC384	>10	>5.3	>10	>15.1	>10	>5.9	2.5	0.387
Ca. auris CDC390	>10	>5.3	>10	>15.1	>10	>5.9	2.5	0.387
Ca. auris CDC387	>10	>5.3	10	15.1	>10	>5.9	1.25	0.194
Ca. auris CDC385	>10	>5.3	>10	>15.1	>10	>5.9	2.5	0.387

Unexpectedly, the MIC results were more reliable for the Cryptococcus isolates compared to the Ca. auris isolates. For instance, the MIC of DQC was 1.25 μm (0.375 μg/ml) against all the cryptococcal isolates. However, the MIC of DQC was 10 μ m (5.3 μg/ml) against Ca. auris MMC2. Against all the other Ca. auris isolates tested, we did not observe growth inhibition at the range of concentrations that we tested. BMS possessed antifungal activity against all cryptococcal isolates tested. The MIC of BMS was 2.5 μm (3.775 μg/ml) against all Cr. gattii and Cr. deuterogattii isolates. Against the Cr. neoformans isolates, the MIC values for BMS varied from 2.5 to 10 μm (3.775 to 15.1 μg/ml). Once again, more complex results were observed for BMS against isolates of Ca. auris. The MIC of BMS was 10 μm (15.1 μg/ml) against the isolates MMC2 and CDC387, but all other Ca. auris isolates were resistant to BMS. Against all Cryptococcus isolates tested, the MIC of PIS varied from 5 to 10 μm (2.95 to 5.9 μg/ml). For Ca. auris, PIS inhibited fungal growth only against the MMC2 isolate (5 μm, 2.95 μg/ml). Of note, CLQ produced the most promising results against both genera. Against the Cryptococcus isolates, the MIC of CLQ ranged from 0.625 to 1.25 μm (0.097–0.194 μg/ml) range. All Ca. auris isolates tested were sensitive to CLQ, with the MIC values ranging from 1.25 to 2.5 μm (0.194–0.387 μg/ml).

The fungicidal activity of DQC, BMS, PIS, and CLQ was tested against the standard isolates Cr. neoformans H99 and Cr. deuterogattii R265, respectively, in addition to the Ca. auris isolates MMC1 and MMC2 (Table 3). All four compounds exhibited fungicidal activity against the Cryptococcus isolates tested, but only CLQ exhibited fungicidal activity against both Ca. auris isolates.

Table 3.

MFC determinations of the four selected compounds (dequalinium chloride, DQC; bleomycin sulfate, BMS; pentamidine isethionate salt, PIS; clioquinol, CLQ) against the standard isolates H99 and R265 of Cr. neoformans and Cr. deuterogattii R265, respectively, and the MMC1 and MMC2 isolates of Ca. auris.

		MFC
Compounds	Isolates	μm	μg/ml
Dequalium chloride (DQC)	Cr. neoformans H99	2.5	0.75
	Cr. deuterogattii R265	1.25	0.375
	Ca. auris MMC1	>10	>5.3
	Ca. auris MMC2	>10	>5.3
Bleomycin sulfate (BMS)	Cr. neoformans H99	10	15.1
	C. deuterogattii R265	10	15.1
	Ca. auris MMC1	>10	>15.1
	Ca. auris MMC2	>10	>15.1
Pentamidine isethionate salt (PIS)	Cr. neoformans H99	10	5.9
	Cr. deuterogattii R265	5	2.95
	Ca. auris MMC1	>10	>5.9
	Ca. auris MMC2	>10	>5.9
Clioquinol (CLQ)	Cr. neoformans H99	2.5	0.387
	Cr. deuterogattii R265	5	0.774
	Ca. auris MMC1	10	1.548
	Ca. auris MMC2	2.5	0.387

We also determined the IC50 for each of the compounds using macrophages as prototypes of host cells (Fig. 4). In this analysis, we limited the concentration range to 10 μm, since this was the highest concentration used in our screen. While BMS and PIS produced IC50 values (2.5 and 5 μm, respectively) close to most of their MICS, the IC50 values for DQC and CLQ were higher than 10 μm.

Figure 4.

Toxicity of the four selected compounds (A, dequalinium chloride, DQC; B, bleomycin sulfate, BMS; C, pentamidine isethionate salt, PIS; D, clioquinol, CLQ) to RAW 264.7 macrophages. Cytotoxicity (%) was inferred from the determination of the levels of lactate dehydrogenase activity in the medium. Control systems included DMSO-treated cells (viability control) and macrophages killed with the lysis buffer provided by the manufacturer (death control). The IC50 (boxed areas) was defined as the lowest drug concentration inducing death in 50% of the macrophage population. The IC50 values for DQC and CLQ were outside the 0–10 μm range.

DQC, BMS, PIS, and CLQ affect fungal morphology

We next evaluated how DQC, BMS, PIS, and CLQ affected the morphology of Cryptococcus spp. and Ca. auris. Once again, we used the standard isolates Cr. neoformans H99, Cr. deuterogattii R265, and the Ca. auris isolates MMC1 and MMC2 in these assays. We analyzed general morphological changes with a focus on the surface of fungal cells. In this context, our analyses included the cryptococcal capsule due to the key role of this cellular structure in the pathogenesis of Cryptococcus.^23,24

Fluorescence microscopy of control or compound-treated cryptococci stained for cell wall chitin (blue fluorescence), chitooligomers (red fluorescence), or capsule (green fluorescence) demonstrated that the cryptococcal isolates had regular cell walls and similar serological reactivities to the anti-capsule antibody (Fig. 5A–D). However, dramatic effects on the capsule were observed when exposed to three out of the four compounds (DQC, BMS, and PIS). Against Cr. neoformans H99 (Fig. 5A), the capsule fibers were markedly scarcer and smaller in compound-treated cells than in the control, except for CLQ. Against Cr. deuterogattii R265 (Fig. 5C), similar results were obtained with PIS. When BMS was tested, the size and number of the capsule fibers were also apparently reduced, but the effects were less dramatic than those caused by PIS. In contrast, the capsule seemed more exuberant when DQC and CLQ were tested. Of note, BMS and PIS apparently induced aggregation of Cr. neoformans H99.

Figure 5.

The effects of DMSO (control) or the four selected compounds (dequalinium chloride, DQC; bleomycin sulfate, BMS; pentamidine isethionate salt, PIS; clioquinol, CLQ) on the morphology of cryptococci. In the microscopy panels (A and C), the cells were stained for cell wall chitin (CFW-derived blue fluorescence), chitooligomers (WGA-TRITC-derived red fluorescence), and capsule (FITC-derived green fluorescence of secondary antibodies), as described in the Methods section. No apparent effects were observed either on the cell wall morphology or in the serological reactivity of the capsule. On the other hand, capsular properties were apparently affected after treatment with DQC, BMS, and PIS, mostly against C. neoformans H99. Determination of cell body and capsular dimensions in Cr. neoformans H99 (B) and Cr. deuterogattii R265 (D) after treatment with DMSO (control) or the four selected compounds. Asterisks denote statistical significance (P <0.05) for each compound compared to control conditions. Capsule size was defined as the subtraction of the cell body diameter (limited by the borders of the cell wall) from the whole cell diameter limited by the capsule borders.

The visual suggestion that the antifungal compounds affected the capsular dimensions of cryptococci was generally confirmed after the determination of cell and capsular dimensions (Fig. 5B and D). When the cell body size was determined (capsule not included), none of the compounds induced any alterations in Cr. neoformans H99. Against Cr. deuterogattii R265, PIS and CLQ caused slight reductions in the cell body sizes. However, when the capsule was measured in Cr. neoformans H99, it was clear that treatment with DQC, BMS, and PIS induced diminished capsules, while CLQ caused a modest increase in the capsule size. Against Cr. deuterogattii R265, treatment with CLQ had no effect on capsule size, while DQC induced the formation of a larger capsule. As observed against Cr. neoformans H99, BMS and PIS promoted the formation of smaller capsules in Cr. deuterogattii R265.

When we analyzed the effects of the four compounds against Ca. auris MMC1 and MMC2, DQC, PIS, and CLQ did not cause any apparent changes in fungal morphology (Fig. 6). The opposite was observed when BMS was tested. This compound induced the formation of hyphae-like morphological stages that is not typical for Ca. auris. To make the visualization of this effect clearer, we recorded the growth of both isolates of Ca. auris using time-lapse movies (Supplementary videos 1–4). In the presence of BMS, both Ca. auris isolates formed hyphae-like structures that predominated at the end of the cultivation period.

Figure 6.

The effects of DMSO (control) or the four selected compounds (dequalinium chloride, DQC; bleomycin sulfate, BMS; pentamidine isethionate salt, PIS; clioquinol, CLQ) on the morphology of Ca. auris MMC1 and MMC2. In the fluorescence panels, the cells were stained for cell wall chitin (CFW-derived blue fluorescence) and chitooligomers (WGA-TRITC-derived red fluorescence). None of the compounds caused any apparent changes in the morphology of the cell wall. However, BMS induced a hyphae-like morphological stage in both isolates. Please see supplementary videos 1–4 for an animated view of this effect.

For clarity, the results described in this section are summarized in Table 4.

Table 4.

General effects of dequalinium chloride, bleomycin sulfate, pentamidine isethionate salt, and clioquinol against Cr. neoformans, Cr. deuterogattii, and Ca. auris.

	Antifungal activity			Morphological alterations
Compound	Ca. auris	Cr. neoformans	Cr. deuterogattii	Ca. auris	Cr. neoformans	Cr. deuterogattii
Dequalinium chloride	Yesa	Yesb,c	Yesb,c	No	Yese	Yesf
Bleomycin sulfate	Yesa	Yesb,c	Yesb,c	Yesd	Yese	Yese
Pentamidine isethionate salt	Yesa	Yesb,c	Yesb,c	No	Yese	Yese
Clioquinol	Yesb,c	Yesb,c	Yesb,c	No	Yesf	No

^aActivity exclusive to the standard isolate using for drug screening.

^bActivity against multiple isolates of the same species.

^cFungicidal.

^dInduction of hyphae-like morphology.

^eCapsule reduction.

^fCapsule enlargement.

Discussion

There are limited therapeutic options for treating infections caused by Cr. neoformans, Cr. deuterogattii, and Ca. auris,^25,26 which justifies the search for novel therapeutic alternatives to the antifungal drugs currently used. In our study, we found four compounds with antifungal activity against three lethal fungal pathogens. As we will discuss further, the compounds selected in our study have important limitations. However, the identification of novel prototypes for drug development is fundamental for future initiatives aimed at introducing new antifungals into clinical use. In our screen, a major criterion for compound selection was the ability to control the growth of three fungal pathogens (Cr. neoformans, Cr. deuterogattii, and Ca. auris). A discussion of the strengths and limitations of each of the four compounds is presented below.

DQC is a quaternary ammonium salt with reported antiseptic and disinfectant activities that is effective in the treatment of wound dressings, mouth infections, and vaginal bacterial conditions.²⁷ In our study, DQC exhibited fungistatic activity against Ca. auris, but this activity was limited to one clinical isolate. In contrast, DQC exhibited fungicidal activity against multiple cryptococcal isolates. DQC exhibited a contradictory effect on the capsule, which is the major virulence factor of Cryptococcus. While DQC inhibited capsule formation in Cr. neoformans, DQC promoted a slight increase in the capsular dimensions in Cr. deuterogattii. In a previous screen, DQC showed a modest ability to inhibit spore germination in Cr. neoformans.²⁸ Taken together, these results might suggest that although DQC is unlikely to be a candidate prototype for drug development against Ca. auris, its activity against Cryptococcus could be improved. In this direction, future studies aimed at understanding the antifungal mechanism of action of DQC are necessary.

BMS is a glycopeptide produced by Streptomyces verticillus. BMS inhibits DNA metabolism and is used as an antineoplastic, particularly for solid tumors.²⁹ In our study, BMS exhibited fungistatic activity only against one clinical isolate of Ca. auris (MMC2). In contrast, BMS exhibited fungicidal activity against multiple isolates of Cr. neoformans and Cr. deuterogattii and reduced capsular dimensions. Of note, BMS induced the formation of aggregates in cryptococci and filamentation in Ca. auris. Considering that BMS also affected cell-wall separation and capsule formation in cryptococci, it seems reasonable to suggest that BMS affects the cell wall of fungi. Therefore, BMS could represent a prototype for the development of novel drugs targeting the fungal cell wall.

The morphological effects induced by bleomycin might be important for the pathophysiology of Ca. auris. Since its discovery, it had been thought that Ca. auris was incapable of filamentous growth. However, it has been recently demonstrated that the transition between yeast and filamentous phenotypes in Ca. auris was triggered by passage through the mammalian body.³⁰ Pseudohyphal growth of Ca. auris was also triggered by genotoxic stress induced by hydroxyurea, methyl methanesulfonate, and 5-fluorocytosine.³¹ Our results align with the previous finding that BMS inhibited fungal cell-wall septation and cytokinesis, which resulted in aberrant cell division and inhibition of cytokinesis in the model yeast Saccharomyces cerevisiae.³² In fact, the morphological changes in Sa. cerevisiae reported in the previous study resembled those observed in our study.

PIS is a synthetic amidine derivative with previously reported antiprotozoal and antifungal activities.^33,34 PIS appears to interact with the minor groove of AT-rich DNA regions of the pathogens’ genomes and interferes with DNA replication and function.³⁴ PIS is effective in the treatment of trypanosomiasis, leishmaniasis, and some fungi, especially Pneumocystis jerovecii.³⁴ In our study, PIS exhibited fungistatic activity against only one isolate of Ca. auris (MMC2). However, the anti-cryptococcal effect of PIS is promising. PIS exhibited fungicidal activity, was active against multiple isolates of Cr. neoformans and Cr. deuterogattii, and induced an important reduction in the capsular dimensions against both species. Pentamidines are known anticryptococcal compounds. An early study found that high concentrations of pentamidine (high concentrations 10 and 100 mg/l in vitro) inhibited the growth of cryptococci isolated from patients with acquired immunodeficiency syndrome.³⁵ In a follow-up study, pentamidine analogs were found to be active against Aspergillus fumigatus, Fusarium solani, Candida species other than Ca. albicans, and fluconazole-resistant strains of Ca. albicans and Cr. neoformans.³³ More recently, in a screen for the search of compounds inhibiting the germination of cryptococcal spores, pentamidine effectively reduced lung fungal burdens when used prophylactically (before infection) or as a treatment (after establishing an infection).²⁸ Due to its efficacy in vivo and low intranasal toxicity, pentamidine was proposed as a lead candidate for repurposing for broader use as an antigerminant to prevent spore-mediated disease in immunocompromised patients.²⁸ Together, these results indicate a great potential for the development of anticryptococcal agents based on the structure and antifungal activity of PIS.

CLQ was the only compound in our study that exhibited fungicidal activity against multiple isolates of Ca. auris, Cr. neoformans, and Cr. deuterogattii. CLQ is an orally bioavailable, lipophilic, copper-binding, halogenated 8-hydroxyquinoline with antifungal, antiparasitic, and potential antitumor activities.³⁶ CLQ forms a stable chelate with copper (copper [II] ions) and inhibits the chymotrypsin-like activity of the proteasome; consequently, ubiquitinated proteins may accumulate, which is followed by apoptosis.³⁶ However, CLQ is neurotoxic, which led to its withdrawal from the market in 1983.³⁶ However, due to its promising activity as a tool to fight Alzheimer's disease, CLQ has been explored as a prototype to generate safer derivatives. For instance, a series of novel flurbiprofen–CLQ hybrids were designed, synthesized, and characterized as multifunctional agents for treatment of Alzheimer's disease.³⁷ Due to the broad activity of CLQ against Ca. auris and cryptococci, developing analogs of CLQ could be used to develop potentially potent and safe antifungals.

Candida auris isolates were more resistant to our four compounds than Cryptococcus spp. In addition, the MIC values for the compounds against the Ca. auris isolates were highly variable. The MIC value variations observed against Ca. auris may be related to the fact that the isolates used in our study belong to four distinct clades, classified from I to V. These clades can present tens to hundreds of thousands of single nucleotide polymorphisms; therefore, these isolates of Ca. auris differ to some extent from each other.³⁸ For Ca. auris, distinct levels of susceptibility have been observed for fluconazole and voriconazole when isolates from clades I and III were compared.³⁹ In addition, a recent study investigating the antifungal activity of nitroxoline against 35 fungal isolates found that the MIC of nitroxoline was lower against Ca. auris isolates from clade I compared to clade III.⁴⁰

Our screening approach selecting compounds with activity against three major fungal pathogens identified four compounds with clear strengths but also with limitations, including high toxicity. Further studies to identify the mechanism of action, safety, and medicinal chemistry will be required for further development of these four compounds. Due to the urgent need for new tools to fight fungal diseases, knowledge of novel compounds with antifungal activity is highly desirable. Our study contributes to bridge this important gap in the field of infectious diseases.

Funding

M.L.R. is supported by grants from the Brazilian Ministry of Health (grant 440015/2018-9), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq; grants 405520/2018-2 and 301304/2017-3), and Fiocruz (grants PROEP-ICC 442186/2019-3, VPPCB-007-FIO-18, and VPPIS-001-FIO18). M.L.R. also acknowledges support from the Instituto Nacional de Ciência e Tecnologia de Inovação em Doenças de Populações Negligenciadas (INCT-IDPN). H.C.O. was supported by scholarships from the program Inova Fiocruz/Fundação Oswaldo Cruz. J.D.N. is supported in part by NIH R21AI156104. The funders had no role in the decision to publish or prepare this manuscript.

Declaration of interest

The authors have no conflict of interest to report.