Antifungal Drug Repurposing

Control of fungal pathogens is increasingly problematic due to the limited number of effective drugs available for antifungal therapy. Conventional antifungal drugs could also trigger human cytotoxicity associated with the kidneys and liver, including the generation of reactive oxygen species. Moreover, increased incidences of fungal resistance to the classes of azoles, such as fluconazole, itraconazole, voriconazole, or posaconazole, or echinocandins, including caspofungin, anidulafungin, or micafungin, have been documented. Of note, certain azole fungicides such as propiconazole or tebuconazole that are applied to agricultural fields have the same mechanism of antifungal action as clinical azole drugs. Such long-term application of azole fungicides to crop fields provides environmental selection pressure for the emergence of pan-azole-resistant fungal strains such as Aspergillus fumigatus having TR34/L98H mutations, specifically, a 34 bp insertion into the cytochrome P450 51A (CYP51A) gene promoter region and a leucine-to-histidine substitution at codon 98 of CYP51A. Altogether, the emerging resistance of pathogens to currently available antifungal drugs and insufficiency in the discovery of new therapeutics engender the urgent need for the development of new antifungals and/or alternative therapies for effective control of fungal pathogens. We discuss the current needs for the discovery of new clinical antifungal drugs and the recent drug repurposing endeavors as alternative methods for fungal pathogen control.


Standard Dilution Methods: CLSI and EUCAST
The majority of articles (Table S1, Supplementary Materials) performed drug repurposing via experimental testing approaches. These include CLSI, EUCAST, or other microdilution/agar assays such as biofilm bioassay, high-throughput ATP content assay, microdilution, and fluorescent microscopic analysis, metabolism and hyphal inhibitory assays, drug diffusion susceptibility testing, human neutrophils, epithelial cell adhesion and invasion assays, murine model, and macrophage assay, among others.
The standard dilution methods developed by the CLSI and EUCAST quantitatively determine (1) minimum inhibitory concentrations (MICs) and minimum fungicidal concentrations (MFC) of drugs/compounds via the microdilution assay settings, as well as whether the antifungal efficacy of drugs/compounds is fungicidal or fungistatic, where fungicidal indicates a ratio of MFC/MIC ≤4 [211], and (2) the levels and types of drug/compound interactions when two drugs/compounds are co-applied, thus calculating the fractional inhibitory concentration indices (FICI) using MIC values or the fractional fungicidal concentration indices (FFCI) using MFC values; synergism indicates FICI or FFCI values ≤0.5, while indifference indicates FICI or FFCI values >0.5-4 [212].
The other method termed the "disc diffusion" test is a relatively inexpensive assay compared to the standard dilution method, for which a few standard assay protocols have been documented in CLSI. CLSI M44 was validated only for azoles and echinocandins for the isolates of Candida species, while CLSI M51-A and supplement M51-S1 qualitatively analyze the efficacy of CAS, triazoles, AMB, etc. [213]. There is a lower agreement between the disc diffusion test results and that of the standard dilution assays, especially in the values from Aspergillus flavus (AMB and VOR) and A. fumigatus (AMB), suggesting that the standard dilution assay seems useful to determine the interpretative breakpoints for both Candida and Aspergillus species.

Biofilm Analysis
Studies have shown that many fungi can alternate planktonic (freely floating, homogeneous cells) and sessile (surface-aggregated, heterogeneous cells) growth, which significantly affects fungal pathogenesis and human infection [214][215][216]. The sessile, multicellular communities of fungi, also known as biofilms, are highly structured fungal communities, which are either adherent to biological or physical surfaces, such as oral mucosa, denture acrylic substrates, and catheters, or form aggregates within the protective extracellular matrix (ECM). The majority of clinically important fungi can produce biofilms, which include filamentous fungal pathogens (Aspergillus, Fusarium, and zygomycetes), yeast pathogens (Blastoschizomyces, Saccharomyces, Malassezia, Trichosporon, Cryptococcus, and Candida species), and Pneumocysitis [214][215][216]. There are also variations in biofilm morphology depending on the types of fungi, namely, (1) C. albicans forms complex morphology with blastospores, hyphae, and ECM, (2) C. neoformans forms an organized structure having yeast cells with a matrix, (3) A. fumigatus forms hyphal cells with ECM, (4) Trichosporon asahii forms yeast and hyphal cells with ECM, etc. [216]. Therefore, in addition to the standard in vitro testing for the planktonic cells, such as CLSI or EUCAST assays, a highly reproducible microtiter plate-based colorimetric measurement determining metabolic activities of the pathogen's biofilm are also used [217].
In general, the development cycle of a fungal biofilm consists of initial adhesion, colonization, proliferation with ECM production, biofilm maturation, and dispersion [214][215][216]. Various environmental factors induce surface attachment and biofilm formation of fungi, including the flow of body fluids (urine, blood, saliva, mucus), pH, temperature, and host immune factors, whereby biofilms protect fungi from the harsh environments including antifungal drugs (e.g., ECM shields fungal cells from drugs and reduces drug penetration), chemical and physical stress, etc., or enable a community-coordinated gene regulation or metabolism [214][215][216].
Biofilm-forming fungal infections are very difficult to treat, which often involve increased drug-resistance phenotypes [218]. In particular, the triazoles and traditional formulations of polyene drugs are considered inactive against fungal biofilms [219]. Differential antifungal activity of drugs has also been documented against biofilms from different fungal pathogens. For instance, echinocandins and AMB lipid formulations exhibited in vitro and in vivo antifungal activities against C. albicans biofilms, while other fungal biofilms, such as A. fumigatus or C. auris biofilms, were resistant to echinocandin drugs including CAS [214,219] In addition to biofilm formation, the phenotypic variability of fungi could also play an important role in clinical outcomes of therapeutic interventions, including the repurposed drugs/compounds ( Table 2). In C. albicans, the morphological switch from yeast cells to hyphae (filamentous forms) serves as a crucial virulence factor, which promotes infection and invasion in hosts. There was also a positive correlation between the level of azole resistance and the capability to form a hyphal structure; under hypha-inducing conditions, only the C. albicans resistant to azoles could form hyphae while the susceptible isolates could not [220]. In the antifungal drug screenings, the manganese nitrosyl [Mn(PaPy 3 )(NO)](ClO 4 ) ({Mn-NO}), a biocompatible NO-donating reagent that delivers NO under visible light, has been determined more effective against the hyphal form of C. albicans, when compared to the yeast cells [221]. Triclosan has also been used in oral hygiene products with a broad-spectrum antimicrobial activity. Of note, in C. albicans, triclosan at subinhibitory concentrations antagonized the antifungal activity of the azole drug FLU, which was specific under hypha-inducing conditions [222]. This antagonism could be due to the membranotropic characteristic of triclosan and also the unique composition of hyphal membranes [222]. Meanwhile, in A. fumigatus, the conidia and hyphal forms of fungal fragments were equally susceptible to the AMB and azole drugs, while hyphal clumps were only susceptible to the relatively high concentrations of AMB [223]. Table 2. Examples of repurposed drugs negatively affecting biofilm formation, hyphal filamentation, or yeast growth in fungi.

Fungi Effects
Ref.

Animal Model Systems
Animal model systems (mammalian and nonmammalian models) are important components for antifungal drug development/discovery including the validation of repurposed drugs [224] (Table 3). Table 3. Examples of animal models used in drug repurposing (see also Beauvericin BALB/c (inbred) mice, female, 7 weeks old

C. albicans
Preventive deferasirox treatment significantly reduced the fungal burden in tongue tissue [137] N-Acetylcysteine C57/BL6 mice, female, 6 to 8 weeks old C. gattii N-Acetylcysteine + AMB 1 achieved higher survival than the control and reduced morbidity in murine-induced cryptococcosis; reduced fungal burden in lungs/brain and lower concentrations of proinflammatory cytokines in the lungs [138] Cisplatin BALB/c mice, female, 4 to 6 weeks old C. neoformans Cisplatin significantly inhibited C. neoformans growth in a mouse model [168] Panobinostat Galleria mellonella larvae C. albicans Panobinostat and FLU combination enhanced survival rate of G. mellonella [143] Pilocarpine hydrochloride G. mellonella larvae C. albicans Pilocarpine hydrochloride protects G. mellonella larvae from C.
albicans-induced mortality in a dose-dependent manner [154] Pitavastatin Caenorhabditis elegans animal model

C. albicans
Pitavastatin-FLU combination reduced the biofilm formation of Candida species and the fungal burdens in a C. elegans infection model Mammalian models are represented by murine, rat, guinea pigs, and rabbits, including both naïve and compromised mice, whereby the antifungal activity of the repurposed drugs can be examined for pharmacokinetics (PK) (e.g., tissue distribution, excretion), pharmacodynamics (PD), immune responses elicited by fungal pathogens, and vaccination attempts [224,225] (see [138] in Table S1, Supplementary Materials). There are various animal models and infection routes, together with various immune suppression regimens. Examples include, but are not limited to, drug screening murine model, murine neutropenic thigh model for determining PK/PD of antifungal drugs, murine model for testing mucocutaneous candidiasis, diabetic murine model of disseminated mucormycosis, murine pulmonary mucormycosis, cyclophosphamide/cortisone immunocompromised murine model of pulmonary mucormycosis, persistently neutropenic rabbit model for investigating acute, invasive pulmonary aspergillosis (IPA), and persistently granulocytopenic rabbit model for characterizing the efficacy of L-AMB against IPA [224,225]. It is important to note that animal testing should be compliant with animal welfare regulation, including Institutional Animal Care and Use Committee (IACUC) review, for all proposed animal experiments [226].
Nonmammalian models currently consist of Drosophila melanogaster (fruit fly), Caenorhabditis elegans (free-living nematode) and Galleria mellonella (wax moth). Examples include biofilm formation in D. melanogaster, melanization and toxicity testing in G. mellonela, and slow and fast killing testing in C. elegans, among others. [227]. When compared to the mammalian models, the nonmammalian models are considered affordable and easy to handle (see [154] in Table S1, Supplementary Materials). While studies have shown that many data from nonmammalian models are in parallel with those obtained from mammalian models, some results did not correlate well. The major drawback of nonmammalian models lies in their unsuitability for microbial vaccination, colonization assessment, challenge research, and immune response. Accordingly, mammalian models are considered to better represent the human condition during the efficacy assessment of the repurposed drugs [224], as also described in Section 4.1.
There have been several types of candidate drugs used for antifungal repurposing (Table S1, Supplementary Materials). These include antipsychiatric, estrogen modulator, antidepressant, antiplatelet aggregation, enzyme (serine palmitoyl-transferase) inhibitor, anticardiovascular, antiarthritis, antistroke, antiatherosclerosis, anticancer, and anthelmintic drugs. However, the use of various drug libraries in the antifungal drug repurposing process is also increasing recently; drug libraries have been prepared/preserved by either public or commercial institutes/vendors, as shown in Table 4.

Synergism between Repurposed Agents and Conventional Antifungals
Combination therapy in controlling fungal pathogens is defined as a co-application of two or more antifungal drugs to treat fungal infections [230]. Combination therapy has been developed on the basis that co-administration of antifungal drugs having different cellular/molecular targets could effectively eliminate fungal pathogens, especially those resistant to conventional drugs. However, the efficacy of combination therapies often varies depending upon the types of drugs co-applied. Although many studies determined better results for fungal pathogen control with drug combinations, other data exhibited no added merit of drug co-treatment over the individual application of each drug alone, which may be associated with drug antagonism [231]. Risk factors also exist during combination therapy, such as multidrug interactions and cytotoxicity.
Studies have shown the synergistic interaction between repurposed agents and conventional antifungal drugs, such as FLU, AMB, or CAS (see Table 5 for a summary). For example, Spitzer et al. determined that the antifungal capacity of chemicals can be systematically enhanced via the combined application of known commercial drugs, such as FLU, with other types of bioactive compounds from drug repurposing [23]. They found that the repurposed compounds did not have to possess potent antifungal activity on their own, but that the compounds potentiated the FLU antifungal activity Antibiotics 2020, 9,812 13 of 29 with considerable species specificity. These synergistic drug combinations were different from the traditional combination therapies mentioned above [23]. Table 5. Summary of the drugs and compounds exerting synergism during co-application (see also Similarly, the anticholesterol drug lovastatin has been repurposed as a synergistic antifungal modulator to the azole drug ITR against the planktonic cells and biofilms of the yeast pathogen C. albicans; the lovastatin regulation of the ergosterol biosynthetic pathway has been the proposed mechanism of antifungal action [45]. The antifungal mechanism of lovastatin in A. fumigatus also involved ergosterol biosynthesis, which was controlled further by cellular iron homeostasis [65]. In A. fumigatus, iron starvation induces the production of the siderophore triacetylfusarinine C (TAFC), for which mevalonate is the key intermediate for the synthesis of both ergosterol and TAFC, the critical virulence factors [65]. Of note, the expression of the enzyme 3-hydroxy-3-methyl-glutaryl (HMG)-CoA reductase (Hmg1), responsible for the production of mevalonate, was increased under iron starvation, while the synthesis of TAFC was reduced following the lovastatin-mediated inhibition of Hmg1 [65].
In our prior study, considerable augmentation of the control of Cryptococcus species was achieved by the co-application of repurposed compounds, such as octyl gallate or benzaldehydes, with conventional antifungal agents [232]. The Cryptococcus species exhibited higher susceptibility to the inhibition of mitochondrial respiration compared to other yeast pathogens Candida species [232]. This "species-specific" enhancement of sensitivity to the co-treatments resulted from the inability of Cryptococcus species to produce cellular energy (ATPs) via the fermentation process. Studies by Spitzer et al. mentioned above indicated that the differential susceptibility of pathogens to the newly developed drugs or interventions is triggered by the differences in physiological/genetic settings of test strains, which could result in species-specific antifungal responses.
Related antifungal "chemosensitization" has been developed recently as a new intervention strategy, where co-application of a repurposed compound (chemosensitizer), such as food additives, with conventional drugs enhanced the antifungal efficacy of the co-applied drugs [77,233]. A chemosensitizer causes the target pathogen to be more susceptible to the co-applied conventional drug via the modulation of the pathogen's defense system, such as the oxidative stress signaling system or cell-wall integrity pathway. Considering that the chemosensitizers could also function as probe-like chemicals by negatively affecting specific cellular targets such as antioxidant systems, types of "drug-compound (repurposed) combinations" enable target-specific control of fungal pathogens, including augmentation of the activity of the echinocandin drug CAS by the cell-wall-targeting octyl gallate [234]. The chemosensitization strategy has been applied further to the development of a high-efficiency drug repurposing protocol that could enhance the sensitivity of target pathogens to the drug candidates, thus reducing time/costs for screening new antifungal drugs, as well as overcoming drug/fungicide resistance of fungal pathogens [104].

Pioglitazone: Needs for In Vivo Drug Validation
Pioglitazone (PIO) has been used as an adjuvant of AMB for the treatment of cryptococcosis. AMB causes excessive generation of reactive oxygen species linked to compromised renal function. PIO is an agonist of peroxisome proliferator-activated receptor γ, which is used to treat type 2 diabetes and is also used as an adjuvant of many drugs triggering side-effects due to its redox-active and anti-inflammatory characteristics [235]. In a murine model, co-application of PIO and AMB exhibited higher efficacy than AMB alone for the inhibition of yeast pathogens, whereby the combination (PIO + AMB) disrupted yeast transmission from the lungs to the brain, which also eliminated yeasts that reached the central nervous system [236]. PIO did not exhibit in vitro antifungal activity, nor did it affect the AMB-mediated fungicidal activity of macrophages; however, PIO as a therapeutic adjuvant counteracted the oxidative bursts after the reduction of the fungal burden, thus relieving the oxidative stress damages to the host (in vivo) and preventing the establishment of meningoencephalitis [236]. The fact that the in vitro antifungal activity of PIO and AMB co-application was determined as "indifferent" while that of in vivo administration increased the survivability of the animals, compared to AMB alone, strongly suggests the importance of the in vivo validation of drug repurposing. The in vitro CLSI or EUCAST testing alone could miss identifying highly effective antifungal adjuvants such as PIO (also observed in other drug developments, personal communication [237], American Chemical Society National Meeting, 2020).

Resistance to Repurposed Drugs/Compounds: Cinnamic and Benzoic Derivatives
The model yeast Saccharomyces cerevisiae has been used as a useful screening system for identifying antifungal agents in view that (1) the S. cerevisiae genome has been sequenced and well-characterized ( [238], accessed 3 September 2020), (2) S. cerevisiae gene deletion mutants have been very useful for investigating the mechanisms or target genes of screened leads [239], and (3) many genes in S. cerevisiae are homologous to those of fungal pathogens [240]. The antifungal compounds screened via S. cerevisiae (wild type or mutants) also exhibited broad-spectrum antifungal activities against pathogenic yeasts (Candida and Cryptococcus species) and filamentous fungi (Aspergillus, Fusarium, and Scedosporium species) [2].
Caution should be exercised during the high-throughput repurposing process so as not to overlook the tolerant response of certain mutants. For instance, cinnamic acids are generally recognized as safe (GRAS) compounds, which have been used as food additives [241]. In recent repurposing studies, cinnamic acid derivatives have been investigated as antifungal alternatives which target fungal cell-wall biosynthesis and integrity [242][243][244]. While the wild type and cell-wall integrity mutants of S. cerevisiae showed a sensitive response to the selected cinnamic acids, such as 3-or 4-methoxycinnamic acids, the glutathione reductase mutant (glr1∆) was hyper-tolerant to 4-methoxycinnamic acid when compared to other test strains [243] (Figure 1). This type of hyper-tolerance was eliminated by 4-methylcinnamic acid, which is the structural derivative of 4-methoxycinnamic acid having a deoxygenated para methyl moiety (Figure 1). Glutathione reductase is necessary for the reduction of the oxidized glutathione (GSSG) to reduced glutathione (GSH) to maintain cellular redox homeostasis [245]. While the study showed the structure-activity relationship of cinnamic derivatives in targeting fungal cell-wall components where the para methyl moiety is critical to overcoming the glr1∆ hyper-tolerance to 4-methoxycinnamic acid, the investigation also highlighted the importance of comprehensive determination of the "gene-compound" interaction/response, thus avoiding unfavorable outcomes including fungal tolerance to the repurposed agents (Figure 1; Figure 2b for the scheme) during drug repurposing.
highlighted the importance of comprehensive determination of the "gene-compound" interaction/response, thus avoiding unfavorable outcomes including fungal tolerance to the repurposed agents (Figure 1; Figure 2b for the scheme) during drug repurposing. Edible plants including herbs are rich sources of bioactive metabolites that possess various hepatoprotective, antihypertensive, antitumor, or immunomodulatory effects. However, natural ingredients in plant extracts, such as benzoic derivatives, could negatively affect the fungal signaling mutants (for example, A. fumigatus antioxidant mitogen-activated protein kinase (MAPK) mutants sakA and mpkC [246,247]) where fungal MAPK mutants showed tolerance to the benzoic ingredients, while the wild-type strains remain susceptible to the molecules (our unpublished observation; see Figure 2a for fungal bioassay and Figure 2b for the scheme). Collectively, studies proved the significance of "gene-compound" interaction analysis during the preclinical stage of drug repurposing (regardless of the sources of the drugs or compounds whether synthetic, natural, crude extracts, or purified), thus circumventing the unfavorable downside of repurposed drugs.   Edible plants including herbs are rich sources of bioactive metabolites that possess various hepatoprotective, antihypertensive, antitumor, or immunomodulatory effects. However, natural ingredients in plant extracts, such as benzoic derivatives, could negatively affect the fungal signaling mutants (for example, A. fumigatus antioxidant mitogen-activated protein kinase (MAPK) mutants sakA and mpkC [246,247]) where fungal MAPK mutants showed tolerance to the benzoic ingredients, while the wild-type strains remain susceptible to the molecules (our unpublished observation; see Figure 2a for fungal bioassay and Figure 2b for the scheme). Collectively, studies proved the significance of "gene-compound" interaction analysis during the preclinical stage of drug repurposing (regardless of the sources of the drugs or compounds whether synthetic, natural, crude extracts, or purified), thus circumventing the unfavorable downside of repurposed drugs. Edible plants including herbs are rich sources of bioactive metabolites that possess various hepatoprotective, antihypertensive, antitumor, or immunomodulatory effects. However, natural ingredients in plant extracts, such as benzoic derivatives, could negatively affect the fungal signaling mutants (for example, A. fumigatus antioxidant mitogen-activated protein kinase (MAPK) mutants sakA∆ and mpkC∆ [246,247]) where fungal MAPK mutants showed tolerance to the benzoic ingredients, while the wild-type strains remain susceptible to the molecules (our unpublished observation; see Figure 2a for fungal bioassay and Figure 2b for the scheme). Collectively, studies proved the significance of "gene-compound" interaction analysis during the preclinical stage of drug repurposing (regardless of the sources of the drugs or compounds whether synthetic, natural, crude extracts, or purified), thus circumventing the unfavorable downside of repurposed drugs.

Summary
Current antifungal intervention strategies often encounter limited efficiency in controlling fungal pathogens. Infections of the bloodstream or lungs by Candida species or the airborne Cryptococcus/Aspergillus species, respectively, that are resistant to conventional drugs cause serious health issues. Resistance to drugs develops via the repeated usage of antifungal agents over time (acquired resistance), while certain fungal species are intrinsically resistant to the conventional drugs (intrinsic resistance); examples of intrinsic resistance include azole resistance (C. glabrata, C. krusei, and C. auris), echinocandin resistance (Cryptococcus and Fusarium species), and polyene resistance (C. auris and A. terreus) [248,249].
Drug repurposing for fungal treatments is an alternative strategy for developing new antifungals. In this paper, two types of drug repurposing approaches were discussed, in silico/computational approaches and experimental approaches. While the majority of drug repurposing studies adopted experimental repurposing platforms, it is expected that the numbers of in silico/computational investigations will be increased in the future considering the increasing numbers of in silico data including omics, drug target, and chemical structural data. The preclinical validation of the efficacy of the repurposed drugs will require the testing of molecules in the model systems with the unbiased determination of breakpoints via the standard microdilution protocols developed by CLSI or EUCAST.
While repurposed drugs could be applied independently as novel antifungal agents for treating fungal pathogens, they can also function as effective synergists/adjuvants in formulations to conventional antifungal drugs such as FLU [23]. In contrast to traditional combination therapy, co-application with a repurposed drug could avoid drug antagonism, multidrug interactions, and cytotoxicity. The One Health approach acknowledges that human, animal, and environmental health is closely linked [250], for which drug repurposing could provide solutions to eliminate resistant fungi such as pan-azole-resistant Aspergillus species [251].
In summary, drug repurposing could provide promising alternatives to current antifungal practices. Future inclusion of additional resources, in addition to the one described in this paper, such as the DrugCentral database [252] and Aggregate Analysis of ClinicalTrials.gov (AACT) database [253], would improve the antifungal drug repurposing processes that have the potential to benefit agriculture, food security, and animal and human health.