Enhancement of Commercial Antifungal Agents by Kojic Acid

Natural compounds that pose no significant medical or environmental side effects are potential sources of antifungal agents, either in their nascent form or as structural backbones for more effective derivatives. Kojic acid (KA) is one such compound. It is a natural by-product of fungal fermentation commonly employed by food and cosmetic industries. We show that KA greatly lowers minimum inhibitory (MIC) or fungicidal (MFC) concentrations of commercial medicinal and agricultural antifungal agents, amphotericin B (AMB) and strobilurin, respectively, against pathogenic yeasts and filamentous fungi. Assays using two mitogen-activated protein kinase (MAPK) mutants, i.e., sakAΔ, mpkCΔ, of Aspergillus fumigatus, an agent for human invasive aspergillosis, with hydrogen peroxide (H2O2) or AMB indicate such chemosensitizing activity of KA is most conceivably through disruption of fungal antioxidation systems. KA could be developed as a chemosensitizer to enhance efficacy of certain conventional antifungal drugs or fungicides.


Introduction
Kojic acid (KA, Figure 1) is a natural pyrone produced by certain filamentous fungi, mainly species of Aspergillus and Penicillium. It is a common by-product in the fermentation of soy sauce, sake and rice wine, and is widely used as a food additive to prevent oxidative browning, or in cosmetics as a depigmenting agent [1][2][3]. Genes involved in KA biosynthesis were recently identified [4,5]. Cellular immunity is enhanced by KA through stimulating phagocytosis and generation of reactive oxygen species (ROS) in macrophages, and potentiation of phytohemagglutinin-based proliferation of lymphocytes [6,7]. KA is fungistatic against the pathogenic yeast, Cryptococcus neoformans, by inhibiting melanin production required for infectivity [8]. Derivatives of KA also have antimicrobial activity against a variety of other fungi and bacteria [9], showing its potential as a polyfunctional backbone for new antimicrobial agents [10].  [11]. A. oryzae is used widely in the food industry. However, A. flavus and A. parasiticus are opportunistic pathogens of various crops, and a concern since they produce carcinogenic aflatoxins that can contaminate food. A. flavus is also an agent for human invasive aspergillosis (IA). Of note, the chief agent of IA, A. fumigatus, and a third IA agent, A. terreus, do not produce KA [12][13][14].
Co-application of certain types of compounds can enhance efficacy of conventional antimicrobial agents through a process termed "chemosensitization." With regard to microbial pathogens, a chemosensitizer functions by debilitating the ability of a pathogen to completely activate a defense mechanism to an antimicrobial agent [15,16]. We investigated if KA, as a chemosensitizer, could improve activity of commercial antifungal agents against pathogenic strains of Aspergillus and yeasts (See Table 1). We tested this chemosensitizing potential by co-applying KA with hydrogen peroxide (H 2 O 2 ) to mimic host ROS, and with a commercial antimycotic, amphotericin B (AMB) and agricultural fungicides, fludioxonil (FLUD) and strobilurin (kresoxim methyl (Kre-Me)).   . Interactions were defined as: "synergistic" (FICI or FFCI ≤ 0.5) or "indifferent" (FICI or FFCI > 0.5-4) [21].
Synergistic FICIs and FFCIs between KA and H 2 O 2 only occurred in AF293. Despite the absence of calculated "synergism" as depicted by "indifferent" interactions (by definition) ( Table 2), there was enhanced antifungal activity (i.e., chemosensitization) in the remaining A. fumigatus and A. terreus strains. This enhancement was indicated by lower MICs and MFCs for either or both KA and H 2 O 2 when co-applied. Also, the A. fumigatus MAPK mutants had half the MICs and MFCs of AF293 ( Table 2; Figure 3a), suggesting that, in the wild type fungi, MAPKs in the oxidative/osmotic stress responsive pathway play protective roles against the antimycotic activity of KA + H 2 O 2 .

Enhanced Antimycotic Activity of AMB with KA in Filamentous Fungi and Yeasts
AMB is an antimycotic drug against filamentous or yeast pathogens. However, AMB can be associated with significant side effects resulting in nephrosis and other tissue-damage in invasive pulmonary aspergillosis [23]. Therefore, we reasoned that use of chemosensitizing agents from natural sources could enhance the effectiveness of AMB, while lowering toxicity of this polyene drug to human cells. The main mode of action of AMB is disruption of the fungal plasma membrane, resulting in ion leakage. However, AMB also induces oxidative damage [24][25][26][27] by stimulating ROS production [28]. Since KA contributed to oxidative stress when combined with H 2 O 2 in Aspergillus (See Table 2), we surmised it might also enhance AMB activity.

Microtiter Plate (microdilution) Bioassay: Filamentous Fungi
Checkerboard assays of KA (0.2-12.8 mM) and AMB (0.125-32 µg/mL) (See Experimental Section) were initially used to assess antifungal interactions against the Aspergillus strains, by using CLSI M38-A protocol [20]. In assays of the Aspergillus strains, co-application of KA increased AMB activity only in strains of A. fumigatus, where FICIs and FFCIs were synergistic in the A. fumigatus MAPK mutant strains (Table 3; Figure 3b). Among the Candida and Cryptococcus strains tested, KA enhanced AMB activity in C. albicans CAN276, C. krusei ATCC6258, C. neoformans CN24 ( Table 3). Synergism of KA + AMB was observed in C. krusei ATCC6258 and C. neoformans strains (Table 3; Figure 3c).
In parallel checkerboard assays of S. cerevisiae, the wild type and two MAPK cell wall integrity mutant strains, i.e., slt2∆ (MAPK deletion; cell wall integrity pathway) and bck1∆ (MAPK kinase kinase deletion; cell wall integrity pathway) were included. We tried to determine whether the MAPK system for cell wall integrity plays a protective role against the antimycotic activity of KA + AMB.
These mutants previously showed hypersensitivity to certain environmental stresses [30,31]. However, the mutants were not more sensitive than the wild type to co-application of either compound (Table 4), indicating Slt2p and Bck1p (viz., cell wall integrity pathway) do not participate in yeast cell homeostasis under KA + AMB treatment.

No Enhancement of Antimycotic Activity of H 2 O 2 with KA in Yeasts
KA (5 mM) and H 2 O 2 (2 and 3 mM) co-application was tested against yeast in agar bioassays, including five clinical strains of Candida, one of C. neoformans and non-pathogenic, S. cerevisiae. Yeast cells (1 × 10 6 ) were serially diluted (10-fold), spotted onto Synthetic Glucose (SG) agar incorporated with KA and/or H 2 O 2 , and incubated at 30 °C, S. cerevisiae, or 35 °C, Candida/Cryptococcus (See [32] for methods). These assays revealed no effect (data not shown) and hence, checkerboard assays to determine MICs, FICIs, etc., were not performed.
The results of all chemosensitization tests (i.e., KA + H 2 O 2 or AMB in filamentous and yeast strains) are summarized in Table 5.

Enhanced Antimycotic Activity of Strobilurin with KA in A. fumigatus
We also tested combinations of KA with agricultural fungicides, fludioxonil (FLUD) or Kre-Me (strobilurin), fungicides that target different components of the oxidative stress response system [33,34], by using A. fumigatus wild type and MAPK (sakA∆, mpkC∆) mutants. Certain fungi with mutations in genes involved in signal transduction of stress response, e.g., MAPK signaling pathway, can escape toxicity of the commercial fungicide FLUD [34]. In a prior study we found redox-active benzo derivatives co-applied with either of these fungicides reduced effective dosages and prevented tolerance of A. fumigatus sakA∆ and mpkC∆ mutants to FLUD [35]. However, in our present study, co-application of KA with FLUD did not overcome tolerance of these mutants to this fungicide (Figure 4a).
In a parallel study, we tested combinations of KA with Kre-Me. Kre-Me is an inhibitor of complex III of the mitochondrial respiratory chain (MRC), the key route system for cellular energy (ATP) production [36]. Moreover, disruption of complex III of the MRC results in an abnormal release of electrons that additionally cause cellular oxidative stress [37]. Therefore, antioxidant enzymes play important roles in protecting cells from oxidative damage triggered by MRC inhibitors. KA improved antimycotic activity of Kre-Me against all A. fumigatus strains (Figure 4b), where A. fumigatus sakA∆ and mpkC∆ mutants showed relatively higher tolerance to Kre-Me than the wild type (AF293). Thus, results indicated that the chemosensitizing mechanism of KA might not involve glutathione/superoxide dismutase-based oxidative stress response, differing from redox-active benzo derivatives [35]. We speculated that, in addition to inhibiting ATP production, co-application of KA and Kre-Me might involve responses of other types of antioxidant enzymes/systems. Comprehensive chemosensitization tests using KA with additional strobilurins are currently underway in various filamentous fungi, including Aspergillus, Penicillium, Acremonium, Scedosporium, and others (Note: There was no chemosensitization effect of KA with any azole drug, such as fluconazole, ketoconazole, itraconazole, in Aspergillus or yeasts (data not shown)).

Fungal Strains and Culture Conditions
Aspergillus strains (See Table 1

Agar Plate Bioassay: Filamentous Fungi
In the plate bioassay, measurement of sensitivities of filamentous fungi to the antifungal agents was based on percent (%) radial growth of treated compared to control ("No treatment") fungal colonies (See text for test concentrations.) [38]. Minimum inhibitory concentration (MIC) values on agar plates were determined based on triplicate bioassays, and defined as the lowest concentration of agents where no fungal growth was visible on the plate. For the above assays, fungal conidia (5 × 10 4 CFU/mL) were diluted in phosphate-buffered saline (PBS) and applied as a drop onto the center of PDA plates with or without antifungal compounds. Growth was observed for three to seven days to determine cellular sensitivities to drugs/compounds. were determined by using checkerboard bioassays in microtiter plates (with RPMI 1640 medium, except SG for S. cerevisiae; Sigma Co., Madrid, Spain). To determine changes in MICs of antifungal agents (i.e., drugs and chemosensitizers) in microtiter wells, checkerboard bioassays (0.5 × 10 5 to 2.5 × 10 5 CFU/mL) were performed using broth microdilution protocols according to methods outlined by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) [29]. MICs for chemosensitization were defined as the concentrations where no fungal growth was visible at 24 and 48 h. All bioassays