A targeted drug-repurposing strategy identifies Tavaborole (Kerydin) as a potent fungistatic agent against Candida auris
1Max Perutz Labs Vienna, Vienna BioCenter, Center for Medical Biochemistry, Medical University of Vienna, Vienna, Austria
*Corresponding author: Rounik Mazumdar (rounik.mazumdar@meduniwien.ac.at)Abstract
Candidozyma auris (Candida auris) is an emerging multidrug-resistant fungal pathogen posing a major global health threat. In this study, we employed a targeted drug-repurposing strategy to identify novel indications for existing FDA-approved compounds against C. auris, leading to the identification of Tavaborole as a potent fungistatic agent. Tavaborole displayed robust activity across all five tested clades of C. auris, as well as against Candida albicans and Candida glabrata. To investigate drug resistance mechanisms of C. auris, we applied quantitative proteomics analyses following exposure to Tavaborole and Amphotericin B (AmB), complemented by electron microscopy. Proteomic profiling revealed that C. auris mounts distinct but overlapping adaptive responses to antifungal stress, involving stress response pathways, metabolic reprogramming and amino acid biosynthesis. While Tavaborole primarily induced targeted stress adaptation, AmB triggered a broader, multi-pronged resistance response including oxidative stress mitigation, osmolyte production and metabolic remodeling. Shared alterations in glycogen metabolism and amino acid biosynthesis suggest conserved antifungal adaptation mechanisms. Altogether, this study highlights Tavaborole as a promising antifungal candidate against C. auris, sheds novel insights into drug resistance mechanisms employed the pathogen and delivers a drug-repurposing procedure highly customizable to target other microorganisms.
Importance
Candida auris is an emerging multidrug-resistant fungal pathogen responsible for healthcare-associated infections representing a high-priority antimicrobial resistance (AMR) threat due to its limited treatment options, high transmissibility, and capacity to cause severe and often fatal outbreaks. The slow pace of antifungal drug development underscores the urgent need for alternative strategies to expand the antifungal arsenal against priority pathogens such as C. auris. In this study, we demonstrate that a targeted drug-repurposing approach can efficiently identify antifungal activity from a small, curated set of FDA-approved compounds, leading to the discovery of Tavaborole as a fungistatic agent with broad activity across multiple C. auris clades. By integrating a customizable drug screening procedure with quantitative proteomics and electron microscopy, this work provides insights into antifungal resistance mechanisms. This study highlights how rational drug-repurposing strategies can rapidly identify clinically relevant drug candidates to counter emerging pathogens and address antifungal resistance.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
1Introduction
In an era of infectious disease outbreaks the ability to rapidly respond to emerging pathogens is critical and depends not only on the availability of molecular data but also the extrapolation of such data into clinical practice. Invasive fungal infections represent a growing global health burden, with increasing morbidity and mortality driven in part by the emergence of antifungal drug resistance. Such infections are an increasing threat to immunosuppressed patients and the elderly with growing number of incidences every year (1). Among clinically prevalent fungal pathogens, Candida and Aspergillus species account for the majority of infections (1). The over use of antifungals such as azoles and echinocandins has caused a shift to the epidemiology of pathogenic fungal species, leading to the emergence of antimicrobial drug resistance (AMR) (1, 2). The growing public health burden of AMR is well illustrated by the emergence of Candidozyma auris (Candida auris) which has been flagged by the US Centers for Disease Control and Prevention (CDC) as a serious public health threat (3, 4). The World Health Organization (WHO) has classified C. auris as a ‘critical priority’ pathogen in its fungal priority pathogens list (FPPL) underscoring urgent public health action and the European Union (EU) has launched coordinated initiatives to through a One Health approach to curb AMR (2, 5).
C. auris has caused serious outbreaks worldwide since its discovery in 2009, reported in over 40 countries (3, 4, 6). Genomic analysis of C. auris clinical strains have identified six clades, based on geographic location including clade I (South Asia), clade II (East Asia), clade III (Africa), clade IV (South America), clade V (Iran) and the very recent clade VI (Singapore), with clade I being the most prevalent (6–8). The pathogen has been associated with deep-seated infections of the bloodstream (candidemia), wound, respiratory tract, urinary tract, and ear infections, with high mortality rates ranging from 28% to 56% (3, 4). Furthermore, several outbreaks of C. auris have been reported worldwide since the emergence of coronavirus disease 2019 (COVID-19) (9), and co-infections with COVID-19 has resulted in a mortality rate of over 80% in some instances (9, 10).
The clinical challenges posed by C. auris stems from several sources including the fact that it is often misdiagnosed with other Candida species and requires rigorous molecular biology techniques for identification (11). Another bottleneck is the pathogens’ high transmissibility, where rapid patient to patient transmission has been reported (3, 4). In addition, C. auris possess a high degree of environmental persistence including tenacious contaminations of inanimate objects such as catheters and persistent colonization of the skin, thus making it difficult to control and eradicate outbreaks from affected areas (3, 4). For such reasons outbreaks can last for several months or even years before they are completely uprooted from hospitals.
A major impediment to combat C. auris is its multi-drug resistance (MDR) profile. It has been reported that 90% of C. auris clinical isolates are resistant to fluconazole, 35% to Amphotericin B (AmB), 7% to echinocandins, 3% to flucytosine and over 40% were reported to be resistant to more than 2 classes of antifungals (12). Treatment options therefore can be limited by intrinsic and secondary resistance and due to the availability of low number of antifungal families. This is particularly relevant in case of AmB resistance, which comes at a high cost because such a resistance essentially eliminates the last therapeutic option for treatment (13). Despite the effectiveness of the polyene AmB, its clinical use is limited in most countries due to its severe side-effects, such as nephro- and hepatotoxicity (14). The mode of action of this drug still remains enigmatic. The paradigm seems to be fixed at the pore formation theory, which states that AmB acts by forming pores on the cell membrane after binding to ergosterol thereby effectuating osmolysis (14). However, within this context, the fungicidal effect of AmB has been proposed to be more complex with induction of apoptosis and oxidative damage being implicated (14). To combat C. auris the above challenges need to be addressed in order to enable better infection control and preventive measures against the pathogen. Therefore, the quest for novel compounds with promising anti-Auris activity remains relentless. Compounds with potent anti-Auris can not only be used with a scope of potential therapy but can also be employed to limit the spread of the microorganism under clinical setting including in hospitals.
Drug repurposing (drug repositioning) is an attractive strategy to expand the application of existing approved drugs, considering the high attrition rates, extensive costs and time-consuming process for de-novo drug development (15). Even with a fast-tracked approval process of de-novo molecules as seen during the COVID-19 vaccine development, public mistrust often places such efforts back to time-consuming process and hinders our ability to combat pathogens. Repurposed drugs can be considered ‘de-risked’ compounds that have already passed safety assessment in preclinical models and in humans if early-stage trials have been completed, therefore are more likely to pass in subsequent efficacy trials.
Therefore, in order to expand the molecular knowledge spectrum of C. auris and address some of the key challenges associated with the pathogen, in this study, we applied a tailored drug repurposing strategy to identify FDA-approved compounds with activity against C. auris, coupled with quantitative proteomic analyses and electron microscopy, to elucidate potential antifungal resistance mechanisms employed by C. auris including those associated with AmB.
2Materials and Method
2.1In-silico drug screening strategy and generation custom drug library
The drug-repurposing strategy was guided by an in-silico drug screening approach based on Basic Local Alignment Search Tool (blast) (Figure 1) (16). In brief, drug target sequences were downloaded from DrugBank database (version_2021) consisting of protein targets of FDA approved drugs (17). The reference proteome sequences of C. auris were downloaded from the UniProt database and a blastp algorithm was carried out against the DrugBank target sequences (18, 19). Compounds identified with E-value: 0.00 were utilized to construct a custom drug library, from which at least one representative drug per unique C. auris protein ID was selected, yielding a targeted narrowed down set of 14 candidate compounds to be tested for primary in-vitro screening.
2.2Candida strains and culture
In this study we used C. auris isolates belonging to five clades including AR389 (clade I), CBS10913 (clade II), AR383 (clade III), AR385 (clade IV) and AR1097 (clade V). Additionally, Candida albicans (SC5314) and Candida glabrata (ATCC2001) were also utilized. The candida cells were routinely cultured in YPD liquid media (1% yeast extract, 2% peptone, and 2% glucose) at 30°C with constant shaking at 200 rpm.
2.3In-vitro screening assay
Primary screening of selected compounds was performed using spot assays to determine fungal growth inhibition. In brief, plates were prepared using media RPMI-1640 (Sigma-Aldrich, USA) buffered with 0.165mol/L 3-(N-morpholino) propanesulfonic acid (MOPS) (Sigma-Aldrich; USA), 2% glucose at pH 7.0 and 2% agar. Prior to experimentation, fungal cells were streaked onto YPD agar plates and allowed to propagate for 3 days, followed by cultures in YPD broth overnight at 30°C in 200 rpm shaker. Next fresh cultures were prepared and allowed to reach an optical density of 600 nm (OD600) corresponding to 1.0. Following which fungal cells were washed 3x in RPMI-MOPS media and adjusted to working inoculums of OD600 0.2 cell suspension, from which 200μL was seeded into 96 well plates to be used by a robot for spotting. All the tested compounds were prepared in dimethyl sulfoxide (DMSO). Plates were created with a final drug concentration of 50μM (20) and 2μg/mL for AmB. The following compounds were selected from the in-silico screening library: Carboxin (mw: 235.30), Cladribine (mw: 285.69), Colchicene (mw: 399.43), Dasatinib (mw: 488.01), Enasidenib (mw: 473.38), Fostamatinib (mw: 580.46), Griseofulvin (mw: 352.77), Mupirocin (mw: 500.62), Pemetrexed (mw: 427.41), Pimecrolimus (mw: 810.45), Selinexor (mw: 443.31), Sulfinpyrazone (mw: 404.12), Tavaborole (mw: 151.93), Triclabendazole (mw: 359.65). After spotting by robotic arm, the plates were incubated at 30°C and growth inhibition was visually evaluated at 24h and 48h. Any compound displaying fungal growth inhibition across clades over 48h was monitored for up to 120h. Following which antifungal susceptibility test (AST) by broth dilution was carried out to determine minimum inhibitory concentration (MIC). In brief, AST was performed using media RPMI-1640 buffered with MOPS and 2% glucose at pH 7.0. Fungal cell suspensions of OD600 0.2 were generated from which 100μL was seeded into 96-well plates, followed by the addition of 100μL 2-fold serial dilution series of positive hit compound. Controls wells contained no drug and experiments were performed as triplicates. The 96-well plates were then incubated at 30°C in 200 rpm shaker and growth inhibition was evaluated after 24h by measuring the OD600 using a Victor Nivo plate reader (PerkinElmer, USA).
2.4Proteome isolation
In brief, C. auris clade I cells were grown to OD600 0.8 in RPMI-MOPS with 2% glucose at 30°C in 200 rpm shaker followed by cell count by a CASY counter (Roche, Swiss). Following which, 5mL fresh cultures using the same media were seeded with 1 x 107 candida cells for treatment with 2μM AmB or Tavaborole for 4hrs. All samples including untreated controls contained 0.002% DMSO. Experiments were performed as biological triplicates per treatment group. Following the treatment, protein was isolated from fungal cells. Briefly, candida cells were centrifuged at 3000 rpm for 5 min, followed by ice cold PBS wash-3x after which the cell pellet was resuspended in 500μL of ice-cold candida lysis buffer (1% sodium deoxycholate (SDC), 100mM Tris-HCl, 150mM NaCl, 1mM PMSF, 1mM EDTA, 1x cOmplete™ Protease Inhibitor (Roche, Swiss)) in 1.5mL screw-cap tubes (Sarstedt, Germany). Next, the candida cells were subjected to mechanical disruption by glass beads using FastPrep™-24 5G Bead Beating (Fisher Scientific, USA). Following which, a small hole was punctured at the bottom of the screw-cap tube using red-hot flamed needle and the screw-cap tube was then inserted into 1.5mL Eppendorf tube and centrifuge at 3000 rmp for 2 min at 4°C to separate the lysate from the glass beads. The lysate collected in the Eppendorf tube was then subjected to acetone precipitation by adding four volumes of ice-cold 100% acetone followed by 2h incubation at -20°C. Post incubation, the samples was centrifuged at 3000 rpm for 5min to obtain the resulting protein pellet and supernatant acetone was discarded. The protein pellet was air dried on ice and was further processed for mass spectrometric analysis.
2.5Mass spectrometry sample preparation
Briefly, protein pellets were resuspended in 50µL 4%(w/v) SDS, 100mM Tris/HCl pH 8.5, 0.1M DTT, shaken at RT until fully resuspended and incubated at 95°C for 4 min. Lysates were clarified by centrifugation at 16000g for 10 min at 20°C. Supernatants were transferred to new tubes and protein concentration was measured using 600nm protein assay kit (Pierce) with SDS compatibility reagent. Solutions were diluted 1:10 with dH2O, transferred to FASP filters in two steps, and centrifuged for 20 min at 12000g each step. Filters were washed with 200µL 8M urea in 100mM Tris/HCl pH 8.5 for 20 min at 12000g, 100µL 50mM iodoacetamide in 100mM Tris/HCl pH 8.5 were added, shaken for 1 min and incubated 30 min in the dark at room temperature. Filters were centrifuged for 10 min at 12000g followed by washing three times with 200µL 8M urea in 100mM Tris/HCl pH 8.5 and three times with 100µL 100mM Tris/HCl pH 8.5. Filters were transferred to new collection tubes, 40µL 50mM ABC containing 1µg trypsin (Promega) were added and kept at 37°C overnight in a wet chamber to prevent evaporation. Digested peptides were collected by centrifuging for 20 min at 12000g. Filters were washed with 40µL 100mM Tris/HCl pH 8.5, centrifuged for 15 min at 12000g resulting in pooled eluates. Digested peptides were acidified with 10µL 10% TFA and the peptides were desalted using an MCX 96 well plate (Waters).
2.6Liquid chromatography separation coupled to mass spectrometry
Peptides were separated on an Ultimate 3000 RSLC nano-flow chromatography system (Thermo-Fisher), using a pre-column for sample loading (Acclaim PepMap C18, 2 cm × 0.1 mm, 5μm, Thermo-Fisher), and a C18 analytical column (Acclaim PepMap C18, 50 cm × 0.75mm, 2μm, Thermo-Fisher), applying a segmented linear gradient from 2% to 35% and finally 80% solvent B (80 % acetonitrile, 0.1 % formic acid; solvent A 0.1 % formic acid) at a flow rate of 230nL/min over 120 min. Eluting peptides were analyzed on an Exploris 480 Orbitrap mass spectrometer (Thermo Fisher), which was coupled to the column with a FAIMS pro ion-source (Thermo-Fisher) using coated emitter tips (PepSep, MSWil).
2.7Mass spectrometry data acquisition in data-independent acquisition mode (DIA) and raw data analysis
The mass spectrometer was operated in DIA mode with the FAIMS CV set to -45, the survey scans were obtained in a mass range of 350-1200 m/z, at a resolution of 30k at 200 m/z and a normalized AGC target at 300%. 31 MSMS spectra with variable isolation width between 14 and 27 m/z covering 399.5-899.5 m/z range including 1 m/z windows overlap, were acquired in the HCD cell at 30% collision energy at a normalized AGC target of 1000% and a resolution of 30k. The max. injection time was set to auto. Raw data were processed using Spectronaut software (version 16.1.220730.53000, https://biognosys.com/software/spectronaut/) with the DirectDIA workflow. The Uniprot C. auris reference proteome (version 2022_02, www.uniprot.org), as well as a database of most common contaminants were used. The searches were performed with full trypsin specificity and a maximum of 2 missed cleavages at a protein and peptide spectrum match false discovery rate of 1%. Carbamidomethylation of cysteine residues were set as fixed, oxidation of methionine and N-terminal acetylation as variable modifications. The global normalization and imputation were done in Spectronaut - all other parameters were left at default.
2.8Data analysis using R scripts
Spectronaut output tables were further processed using Cassiopeia_LFQ 4.6.4 (https://github.com/maxperutzlabs-ms/Cassiopeia_LFQ). Contaminant proteins, protein groups identified only by one peptide and protein groups with less than two quantitative values in one experimental group, were removed for further analysis. Differences between groups were statistically evaluated using the LIMMA package (21) at 5% FDR (Benjamini-Hochberg).
2.9Proteomics data deposition
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository (22) with the dataset identifier PXD057542.
2.10Transmission electron microscopy (TEM)
An aliquot of fungal cells used for proteomics was also subjected to electron microscopy. In brief, following the drug treatment, candida cells were pelleted at 3000 rpm for 5min. Cell pellets were then fixed using a mixture of 2% glutaraldehyde (Agar Scientific, UK) and 2% paraformaldehyde (Electron Microscopy Sciences, USA) in 0.1mol/l sodium cacodylate buffer, pH 7.2 at room temperature overnight followed by 3 rinsing steps with the same and a post-fixation in 2% osmium tetroxide (Agar Scientific, UK) in 0.1mol/l sodium cacodylate buffer. Dehydration was performed in a graded series of acetone and samples were embedded in Agar 100 resin (Agar Scientific, UK). 70-nm sections were cut and post-stained with 2% uranyl acetate and Reynolds lead citrate (Delta Microscopies, France). Micrographs were recorded on an FEI Morgagni 268D (FEI, Netherlands) operated at 80 keV, equipped with a Mega View III CCD camera (Olympus-SIS).
3Results
3.1In-silico drug screening
Following the in-silico drug screening procedure up to 43 unique compounds were identified targeting 24 C. auris protein targets (Supplementary Table 1). Established frontline antifungals such as the Azole group (Fluconazole, Oxiconazole, Terconazole, Clotrimazole, Voriconazole, Tioconazole, Miconazole, Econazole, Sertaconazole, Posaconazole, Bifonazole, Luliconazole), Echinocandin class (Anidulafungin, Caspofungin, Micafungin) and Flucytosine (5-FC) were identified via the in-silico screening strategy, serving as an important reference benchmark to validate the procedure. The narrowed down custom list for the primary in-vitro screen targeting at least one C. auris protein included 14 compounds such as Carboxin, Cladribine, Colchicene, Dasatinib, Enasidenib, Fostamatinib, Griseofulvin, Mupirocin, Pemetrexed, Pimecrolimus, Selinexor, Sulfinpyrazone, Tavaborole and Triclabendazole.
3.2In-vitro primary drug screening
Primary in-vitro spot assays performed on the 14 repurposed compounds identified two positive hits as novel indication of an existing drug; Triclabendazole and Tavaborole, with the latter emerging as the lead compound exhibiting sustained inhibition across C. auris clades beyond 48h. The compound Tavaborole (Kerydin/AN2690) (mw: 151.93) (Figure 2) displayed robust effect against all five tested clades of C. auris, plus on C. albicans and C. glabrata for up to 120h; a threshold time-point to determine fungicidal vs fungistatic effect (Figure 3). Drug susceptibility test confirmed Tavaborole activity against Candida species with a low MIC range of 3µM (Figure 4). The anthelmintic drug Triclabendazole exhibited modest anti-Candida activity (Supplementary Material 1), however, its effect was weaker than that of Tavaborole. Additionally, a recent study (23), has already identified and confirmed the antifungal activity of Triclabendazole against C. auris clades, as such it was not investigated further and was beyond the scope of this current study.
3.3Proteomics
Proteomics analysis identified 4555 C. auris proteins which covers 84% of the C. auris UniProt proteome database of 5409 sequences. The treatment with Tavaborole caused the differential regulation of 151 proteins (Supplementary Table 1) with the increased abundance of proteins including LEU1, LEU4, orf19.1502, GLT1, GCN4, HPA2, orf19.813, SOD1, YHB1 and GLG21 (Figure 5A). Upon AmB treatment 50 proteins were differentially regulated (Supplementary Table 1) with the increased abundance of proteins including DDR48, orf19.7085, orf19.813, RHR2, LEU1, and GLG21 (Figure 5B).
3.4TEM imaging analysis
Electron microscopy was performed on C. auris cells treated with Tavaborole and AmB (Figure 6). Untreated candida cells displayed well defined cell-wall and plasma membrane with homogeneous cytoplasm and intact intracellular organization. Candida treated with Tavaborole exhibited pronounced intracellular alterations characterized by enlarged vacuoles and compromised internal architecture, suggesting intracellular stress and potential perturbation cellular homeostasis, while the cell-wall and plasma membrane remained largely intact. In contrast, candida cells treated with AmB showed structural damage characterized by extensive vacuolization, separation between plasma membrane and cytoplasmic contents, and signs of compromised membrane integrity.
4Discussion
Infections caused by C. auris pose a major challenge in regards to its treatment, particularly due to its drug resistance profile (24). Importantly, resistance to the polyene AmB is of particular concern as it is rare among fungal pathogens including Candida species (25). Resistance to AmB essentially eliminates the last therapeutic option against fungal pathogens (13). The mechanism of AmB action remains enigmatic, though reports suggest that the drug interacts with fungal cell membrane ergosterol thereby effectuating osmolysis (14). Notably, AmB resistance has been reported across C. auris clades and pose a massive health threat across the globe. As such, a deeper understanding of C. auris drug resistance mechanism as well as the quest for novel antifungal alternatives is of immense importance to combat this enigmatic pathogen.
In an era of precision medicine and artificial intelligence (AI) driven drug discovery, the demand for molecules with a high likelihood of therapeutic efficacy is of monumental value. To this end, we applied a tailored drug repurposing strategy to generate a custom drug screening library consisting of 14 compounds, predicted to display anti-Candida activity with high probability. Our approach was guided by utilizing the C. auris proteome sequences to perform a blast screen against FDA approved drug target sequences from DrugBank database. With this procedure we were able to dramatically reduce the number of drugs to be tested from thousands to only a handful. In general, antifungal drug screening libraries are expensive and consist of over thousands of compounds to be tested, and the screening usually result in a few successful ‘hit’ compounds (26). For example, the Johns Hopkins Clinical Compound Library consisted of over 1500 compounds to be tested with six successful hits, the Prestwick library comprised of 1200 compounds with seven hits and the TargetMol library of FDA-approved drugs contained 1068 compounds for screening with one successful hit (26, 27). In contrast, our drug screening library consisted of 14 compounds to be tested leading to two successful hits, therefore the ratio of the number of drugs tested to yield one successful hit is relatively high. Large-scale repurposing screens typically yield hit rates below 1%, including the Johns Hopkins Clinical Compound Library (0.40%), Prestwick Chemical Library (0.58%), and the TargetMol FDA-approved drug library (0.094%), whereas our targeted screening strategy identified 2 hits from 14 compounds (14.29%), representing an approximately 25 to 150-fold enrichment in hit rate. The drug Tavaborole identified as the lead hit, displayed potent fungistatic activity against clinical isolates of five C. auris clades, and C. albicans and C. glabrata. The second identified drug Triclabendazole also displayed antifungal activity against C. auris corroborating with the recent study (23).
Tavaborole is an FDA approved drug is sold under the name Kerydin, which is used to treat fungal infection onychomycosis caused by Trichophyton rubrum and Trichophyton mentagrophytes. The drug exerts its antifungal activity by inhibiting Leucyl-tRNA synthetase (LeuRS) to disrupt fungal protein synthesis (28). Proteomics analysis revealed that upon exposure of C. auris cells to Tavaborole several proteins were over abundant namely LEU1, LEU4, orf19.1502, GLT1, GCN4, HPA2, orf19.813, SOD1, YHB1 and GLG21. Tavaborole inhibits Leucyl-tRNA synthetase which catalyzes ligation of the amino acid L-leucine to tRNA (29). Here, the exposure of C. auris to Tavaborole induced the upregulation of leucine biosynthesis associated protein LEU1 and LEU4 (30). Furthermore, the orf19.1502 associated protein which has a predicted aminoacyl-tRNA hydrolase activity was upregulated (31). An aminoacyl-tRNA is an tRNA with its cognate amino acid. Every amino acid has its own specific aminoacyl-tRNA synthetase which chemically binds it to tRNA in order to be transferred to a growing peptide (32). Under certain circumstances where a wrong amino acid forms a cognate tRNA, it must be hydrolyzed in order to prevent incorrect protein synthesis (32, 33). The upregulation of the above C. auris proteins suggests an intricate control of amino acid biosynthesis in response to Tavaborole. This is further consolidated by the display of general amino acid control, also known as the GCN response, a phenomenon where deprivation of a particular amino acid induces the expression of genes of all amino acid biosynthesis pathways to increase (34). This is evident in our study by the positive regulation of GLT1 responsible for glutamate biosynthesis (35). In C. albicans GCN4 functions as a transcriptional regulator of amino acid biosynthesis including coordinating responses to amino acid starvation (34). The transcription modulator GCN4 can employ histone acetyltransferase (HATs) complexes to confer transcriptional activation (36). The HPA2 is a member of HATs that can modify histones by acetylating lysine residues at histone tails or at histone globular domains (37–39). Whereby, histone modifications can regulate the transcriptional state of genes and can confer various advantages including responding to external stimuli (37–40). For C. albicans it has been reported that HATs regulate genes to respond to external stimuli in the context of virulence, oxidative stress and antifungal drug tolerance (41). Additionally, the upregulation of small heat shock protein associated with orf19.813, superoxide dismutase SOD1 and nitric oxide dioxygenase YHB1 indicates a possible oxidative stress response (42–44). Furthermore, the role of glycogen metabolism was highlighted with the upregulation of GLG21, a homolog of S. cerevisiae GLG2 encoding the enzyme glucosyltransferase mediating glycogen metabolism (31, 45). Glycogen metabolism in candida has been linked to virulence, survival under stress environment and synthesis of cell-wall (45–47). The upregulation of GLG21 upon Tavaborole exposure thus seem to be a protective mechanism employed by C. auris to circumvent potential stress induced by the drug.
The proteome dynamics of C. auris upon exposure to AmB revealed the upregulation of several proteins associated with DDR48, orf19.7085, orf19.813, RHR2, LEU1 and GLG21. The polyene caused the increased abundance of stress response molecules including proteins associated with DDR48, orf19.7085 and orf19.813. Stress adaptation is a crucial factor for microbial survival under dynamic environments including candida species. Several studies have suggested that AmB can autoxidize to bring about oxidative stress with a fungicidal impact (25). In C. albicans it has been reported that AmB induces oxidative and nitrosative stress by production of reactive oxygen species (ROS) and nitrogen species (48–50). The stress protein DDR48 was reported to be induced in C. albicans in response to oxidative stress, antifungal drugs, exposure to cell wall-perturbing agents and DNA damage (51). The proteins associated with orf19.7085 and orf19.813 are also classified as oxidative stress molecules (31). Therefore, it can be inferred, that the above stress response molecules expressed by C. auris, underpins the effect of oxidative stress induced by AmB and reflects on the pathogens’ stress adaptation mechanism including antifungal resistance. The RHR2 which encodes for glycerol-3-phosphatase involved in glycerol biosynthesis and has been reported to be mediating fungal stress response including osmotic stress and oxidative stress (52, 53). Increased osmolarity is known to cause water loss and cell shrinking, therefore a major survival strategy is to accumulate suitable osmolytes such as glycerol to maintain water balance and restore cell volume (54). In yeast osmotic stress causes overproduction of glycerol which in turn is mediated by the high-osmolarity glycerol (HOG) mitogen-activated protein kinase (MAPK) pathway (55, 56). This osmosensing pathway has been implicated in pathogenicity and cell wall biogenesis in C. albicans (56). Previous proteomics study in C. albicans reported the overabundance of RHR2 protein upon exposure to AmB which could be indicative of glycerol buildup (55). Given that AmB functions by creating pores on fungal cell wall to effectuate osmolysis, the overabundance of RHR2 in our study suggests that C. auris could employ excess glycerol as an osmolyte to counter the osmotic stress caused by AmB. Finally, the common upregulation of GLG21, LEU1 and orf19.813 associated proteins upon exposure to both Tavaborole and AmB suggests, an intrinsic drug resistance response in C. auris, in which the organism adapts its glycogen metabolism and amino acid biosynthesis coupled with stress responses to counter antifungal agents. Furthermore, electron microscopic analysis of C. auris upon exposure to Tavaborole corroborates with the suggestive mode of action of the drug, indicating intracellular stress inducement and potential disruption of cellular homeostasis. In contrast, exposure to AmB is indicative of a more profound adverse effect on candida cells suggesting severe cellular stress and compromising membrane integrity, consistent with the polyene’s hypothesized mode of action of inducing oxidative stress and membrane disruption.
5Conclusion
In this study, we applied a custom drug-repurposing strategy to identify novel indications for existing compounds, leading to the identification of Tavaborole as a potent fungistatic agent against C. auris. Compared to commercial drug screen libraries, our targeted drug-repurposing procedure resulted in a much higher hit rate, which can be readily adapted to target other organisms and enhanced using AI tools. The lead compound Tavaborole exhibited robust anti-Candida activity across the five tested clades of C. auris, as well as C. albicans and C. glabrata. Additionally, to elucidate antifungal resistance mechanisms in C. auris, we performed quantitative proteomic analyses following fungal exposure to Tavaborole and AmB, complemented by electron microscopy to assess drug-induced ultrastructural changes. Our analyses indicated that C. auris mounts distinct yet overlapping adaptive responses to antifungal stress. Tavaborole exposure primarily induced stress response pathways and modulation of amino acid biosynthesis. In contrast, AmB triggered a broader, multi-layered resistance response involving oxidative stress adaptation, osmolyte production and metabolic reprogramming encompassing glycogen metabolism and amino acid biosynthesis. Notably, alterations in glycogen metabolism and amino acid biosynthesis were common to both Tavaborole and AmB treatments, suggesting conserved antifungal adaptation mechanisms that warrant further investigations.
Supporting information
6.#Acknowledgment
This study was funded by the Austrian Science Fund (FWF) project CandidOmics-P33425 (Grant DOI: 10.55776/P33425). Proteomics analyses were performed by the Mass Spectrometry Facility at Max Perutz Labs using the VBCF instrument pool. Electron microscopy imaging was done by the Electron Microscopy Facility, Vienna BioCenter Core Facilities (VBCF), Vienna, Austria. We would like to thank Karl Kuchler and lab members for their support.
7Contributions
Conceptualization: RM. Methodology: RM. Software: RM. Validation: RM, AB. Formal analysis: RM. Investigation: RM, AB. Resources: RM. Data curation: RM. Visualization: RM. Supervision: RM. Project administration: RM. Funding acquisition: RM. Writing-original draft: RM. Writing-Review & Editing: Both authors.
8Conflict of interest
The authors declare no competing interests.