Repurposing approach identifies pitavastatin as a potent azole chemosensitizing agent effective against azole-resistant Candida species

The limited number of antifungals and the rising frequency of azole-resistant Candida species are growing challenges to human medicine. Drug repurposing signifies an appealing approach to enhance the activity of current antifungal drugs. Here, we evaluated the ability of Pharmakon 1600 drug library to sensitize an azole-resistant Candida albicans to the effect of fluconazole. The primary screen revealed 44 non-antifungal hits were able to act synergistically with fluconazole against the test strain. Of note, 21 compounds, showed aptness for systemic administration and limited toxic effects, were considered as potential fluconazole adjuvants and thus were termed as “repositionable hits”. A follow-up analysis revealed pitavastatin displaying the most potent fluconazole chemosensitizing activity against the test strain (ΣFICI 0.05) and thus was further evaluated against 18 isolates of C. albicans (n = 9), C. glabrata (n = 4), and C. auris (n = 5). Pitavastatin displayed broad-spectrum synergistic interactions with both fluconazole and voriconazole against ~89% of the tested strains (ΣFICI 0.05–0.5). Additionally, the pitavastatin-fluconazole combination significantly reduced the biofilm-forming abilities of the tested Candida species by up to 73%, and successfully reduced the fungal burdens in a Caenorhabditis elegans infection model by up to 96%. This study presents pitavastatin as a potent azole chemosensitizing agent that warrant further investigation.


Results and Discussion
Screening of Pharmakon drug library and identification of fluconazole adjuvants hit compounds. We performed an initial screen of the Pharmakon 1600 drug library, at a 16 µM fixed concentration, to identify potential fluconazole adjuvants, for which we used a standard broth microdilution method following the guidelines of the Clinical and Laboratory Standards Institute (CLSI). The screen was performed twice against the azole-resistant C. albicans NR-29448, in the presence or absence of 8 µg/ml fluconazole. This high fluconazole concentration was opted to maximize the initial pool of positive hits. Positive hits were identified as hit compounds that caused significant growth inhibition (by >50%) of the test strain only in the presence of fluconazole. Positive hits were initially determined by visual inspection then further confirmed spectrophotometrically by measuring the absorbance of Candida culture at OD 490 nM. The primary screen identified a list of 44 positive hits (2.75% initial hit rate) that exhibited synergistic interactions with fluconazole against the azole-resistant strain C. albicans NR-29448. These initial hits were sub-grouped into seven antineoplastic agents, eight antiparasitics, eight topical agents and 21 drugs that were considered potential fluconazole adjuvants for treating systemic infections and thus were termed "repositionable drugs" (Fig. 1). Notably, several hit compounds that were classified as topical agents and antiparasitics (Supplementary Table ST1) could hold promising clinical potential for treating topical Candida infections. For example, bufexamac, a topical anti-inflammatory drug, may worth further investigation as part of a future study to treat mucosal and skin infections, especially those caused by azole-resistant Candida species.
However, since the main focus of this study was to identify potent systemic fluconazole adjuvants, we directed our attention to study the fluconazole chemosensitizing activities of the repositionable drugs for their aptness for systemic administration and for their relatively low toxicity profiles. Next, we determined the minimum fluconazole-chemosensitizing concentrations of these drugs against C. albicans NR-29448, in the presence or absence of fluconazole (at 8 µg/ml). Interestingly, the antihyperlipidemic agent simvastatin demonstrated a significant fluconazole chemosensitizing activity at 8 µM while all other hit compounds showed activities only at16 µM (Table 1).
To the best of our knowledge, the primary screen revealed novel fluconazole chemosensitizing agents that have never been reported before, such as aripiprazole, perhexiline, phenelzine, quinestrol, dienestrol, hexestrol, norgestimate, meclocycline, tolfenamic acid, and sulfaquinoxaline. In addition and as expected, the primary screen identified several drugs with known fluconazole chemosensitizing activities, such as the cholesterol-lowering agents; simvastatin, atorvastatin, and lovastatin, artemisinin, amiodarone, sulfamethoxazole, doxycycline, and the calcineurin inhibitors nisoldipine and tamoxifen 19-25 . Synergistic interactions between fluconazole and the antihyperlipidemic statin drugs against C. albicans NR-29448. The observation that simvastatin demonstrated a significant fluconazole chemosensitizing activity was encouraging to assess the activity of other pharmacologically related antihyperlipidemic statin drugs. Microdilution checkerboard assays were used to assess the interactions between eight statin derivatives and fluconazole against C. albicans NR-29448 strain. Interestingly pitavastatin, whose activity as a fluconazole chemosensitizing agent has not been previously reported, displayed the most potent fluconazole chemosensitizing activity (ΣFICI = 0.05) and was superior to all other tested statin drugs ( Table 2). The pitavastatin's fluconazole chemosensitizing activity was even more superior than the other pharmacologically-related statin drugs, which were reported to have fluconazole-chemosensitizing activities 26,27 . Pitavastatin, at 0.25 µg/ml, was able to reduce the MIC of fluconazole by 64-fold against C. albicans NR-29448. Except for pravastatin, all other statin drugs demonstrated synergistic interactions with fluconazole against the tested strain (ΣFICI = 0.13-0.26), Table 2. Of note, pitavastatin was shown to reach a peak blood concentration of 0.23 µg/ml following a single oral dose of 4 mg, suggesting that its indication as a fluconazole adjuvant is rationally conceivable 28 . Due to its potent fluconazole chemosensitizing activity and its potential clinical importance, pitavastatin was selected for subsequent experimental investigation. out of 18 tested Candida strains (~89%), resulting in significant reductions in the fluconazole's MIC values (4-64 folds). Notably, the pitavastatin-fluconazole combination displayed variable activities against strains displaying different azole resistance mechanisms. Pitavastatin interacted synergistically with fluconazole against C. albicans TWO7243 strain, which is known to exhibit increased mRNA levels of ERG11, CDR1 (an ABC-type transporter) and MDR1 (an MFS-type transporter). Similarly, pitavastatin interacted synergistically with fluconazole against C. albicans SC-TAC1 G980E strain, which has a gain of function mutation in TAC1, a positive transcription regulator for the ABC (ATP Binding Cassette) membrane transporters [29][30][31] . However, the pitavastatin-fluconazole combination failed to display similar interactions against strain TWO7241, which exhibits increased mRNA levels of both ERG11 and MDR1, and strain SC-MRR1 P683S which has a gain of function mutation in MRR1, a positive transcription regulator for the MFS (Major Facilitator Superfamily) membrane transporters [29][30][31] . These results indicate that the azole chemosensitizing activity of pitavastatin is dictated by the underlying azole resistance mechanisms and suggest a possible role for the membrane efflux transporters.
Pitavastatin was also evaluated in combination with other azole antifungals including voriconazole and itraconazole. Similar to its effect with fluconazole, pitavastatin possessed broad-spectrum synergistic interactions with voriconazole against 16 strains of C. albicans, C. glabrata, and C. auris (ΣFICI ranged from 0.15 to 0.50, Supplementary Table 2). However, pitavastatin displayed a more narrow-spectrum synergistic relationship with itraconazole, as only 9 out of 18 of the tested Candida strains (50%) responded to the pitavastatin-itraconazole combination (Supplementary Table 3).
Of note, although pitavastatin was able to demonstrate broad-spectrum synergistic interactions with fluconazole, these interactions were not sufficient to restore the antifungal activity of fluconazole in several fluconazole-resistant isolates. Considering the current resistance breakpoints for azole drugs, two C. albicans and four C. auris isolates maintained their resistance profiles to fluconazole [32][33][34][35][36][37] . However, combining pitavastatin with either voriconazole or itraconazole displayed better outcomes against isolates displaying a lower susceptibility to either agent, suggesting a potential clinical significance for treating invasive infections caused by voriconazole (or itraconazole) resistant isolates.
The pitavastatin-fluconazole combination significantly reduces the biofilm-forming abilities of Candida species. Candida species are known for their remarkable capabilities of forming robust adherent structures (i.e., biofilms) on surfaces of different abiotic surfaces, such as catheters, and medical implants [38][39][40] . Biofilms limit the penetration of antifungal drugs and can contribute to treatment failure and chronic infections 41 . Fungal cells residing in biofilms have been reported to have increased expression of efflux genes 42,43 . Biofilms were also reported to trigger the formation of Candida persisters, which can tolerate very high doses of the antifungal agents 44 . Collectively, these factors contribute significantly to the remarkable ability of Candida's biofilms to resist the effect of antifungal drugs, especially azoles 45,46 . Thus, there is a pressing necessity for novel antifungal adjuvants with activity against Candida biofilms. Here, we investigated whether the synergistic relationship between azole drugs and pitavastatin could interfere with the biofilm-forming ability of Candida. Compared to single treatments with either fluconazole or pitavastatin, incubating the tested Candida species with pitavastatin (at 0.5 × MIC) in the presence of a subinhibitory concentration of fluconazole (2 µg/ml) resulted in a significant reduction in the biofilm-forming abilities of C. albicans NR-29448 (by ~92%, Fig. 2a), C. glabrata HM-1123 (by ~70%, Fig. 2b), and C. auris 385 (by ~41%, Fig. 2c). These findings indicate potent inhibitory activities of the pitavastatin-fluconazole combination against different Candida biofilms. However, when tested against preformed biofilms, the pitavastatin-fluconazole combination failed to disrupt mature biofilms suggesting poor penetrating abilities of the tested combination (data not shown). www.nature.com/scientificreports www.nature.com/scientificreports/ Pitavastatin significantly interferes with the ABC-mediated efflux activity of Candida. Notably, the azole chemosensitizing activity of statins has been attributed to their ability to interfere with the fungal ergosterol biosynthesis 26,47 . However, this mechanism does not explain their inconsistent effects against the efflux-activated strains. As shown in Table 3, pitavastatin demonstrated significant fluconazole chemosensitizing activity against strains whose efflux mechanisms involve a significant role of the ABC-type transporters (SC-TAC1 G980E and TWO7243) but failed to do so against strains whose efflux mechanisms are solely mediated by the MFS-type transporters (SC-MRR1 P683S and TWO7241). Additionally, we noticed a significant reduction in the intrinsic antifungal activity of pitavastatin (by 4-8 fold) against C. albicans strains exhibiting increased mRNA levels of ABC-type efflux transporters (Table 3). Moreover, Candida species that are known for their hyperactive ABC-transporters such as C. glabrata and C. auris displayed significantly reduced susceptibility to pitavastatin as compared to the wild type C. albicans strain [48][49][50] . These observations suggest a high affinity of pitavastatin towards the fungal ABC efflux pumps. Therefore, we postulated that pitavastatin may enhance the antifungal activity of fluconazole through a competitive interference with Candida's ABC-type membrane transporters. To investigate this premise, we first used nile red efflux assay. Nile red is a known substrate for the two major membrane transporters (ABC and MFS) which have been reported as major contributors to azole resistance in Candida [51][52][53] . Therefore, nile red can be used efficiently as a non-specific reporter dye to measure drug effects on the efflux activities of C. albicans strains, regardless of their efflux mechanisms. As shown in Fig. 3a, pitavastatin (at 0.25 × MIC) significantly maintained a high level of nile red fluorescence intensity in the ABC efflux-activated strain (SC-TAC1 G980E ), compared to the non-treated control. However, the nile red fluorescence intensity was greatly diminished in the MFS efflux-activated strain (SC-MRR1 P683S ), and the signal was comparable to the non-treated control (Fig. 3b). These results suggest a significant ability of pitavastatin to interfere specifically with the ABC efflux-mediated activities in Candida. These effects were confirmed using flow cytometry analysis. C. albicans SC-TAC1 G980E exhibited a significant increase in the nile red fluorescence intensity following exposure to pitavastatin at 0.25 × MIC. However, C. albicans SC-MRR1 P683S was indifferent to the pitavastatin effect, and the nile red fluorescence intensity was comparable to the non-treated control (Fig. 4a). Pitavastatin treatment resulted in a significant increase (65%) in the mean fluorescent intensity only in the ABC-efflux activated strain (SC-TAC1 G980E) , as compared to the non-treated control (Fig. 4b), which supports our previous observation.
These findings were further confirmed using rhodamine 6G efflux assay. Rhodamine 6G has been shown to display a substrate specificity to the ABC membrane transporters 54 . Similarly, pitavastatin at a subinhibitory concentration (0.25 × MIC) significantly reduced the percentage of effluxed rhodamine in the ABC-efflux activated strain C. albicans SC-TAC1 G980E , as compared to the non-treated control (Supplementary Fig. 1). Once again, this result indicates that the azole chemosensitization activities displayed by pitavastatin can be attributed, at least in part, to their ability to interfere with the function of Candida's ABC transporters. efficacy of the pitavastatin-fluconazole combination in Caenorhabditis elegans infection model. In the field of antimicrobial drug discovery, it is quite frequent to notice that several promising  www.nature.com/scientificreports www.nature.com/scientificreports/ antimicrobial compounds fail when assessed in vivo in animal models, despite potent in vitro activities. Given the in-vitro promising activity of the pitavastatin-fluconazole combination, together with its potent antibiofilm activity against different Candida species, it was necessary to assess its activity in vivo. C. elegans is a satisfactory animal model for the initial assessment of promising antimicrobial agents prior to their evaluation in mammalian models. In order to validate our in vitro results, C. elegans was utilized as an animal model to investigate the fluconazole chemosensitizing activity of pitavastatin. As shown in Fig. 5, treating C. elegans infected nematodes with pitavastatin (at 0.5 x MIC) combined with three different concentrations of fluconazole (2, 8, and 32 µg/ml) displayed variable outcomes depending on the fluconazole concentration and the infectious strain. Compared to the untreated control which accumulated 233 ± 21 CFU/worm, pitavastatin-fluconazole combinations significantly reduced the mean fungal CFU burdens of C. albicans NR-29448 in the infected nematodes by ~82-96% (Fig. 5a). Likewise, pitavastatin-fluconazole combinations reduced the fungal burdens of C. glabrata ATCC MYA-2950 by ~84-93% compared to the untreated control which accumulated 344 ± 19 CFU/worm (Fig. 5b). Against C. auris 390, pitavastatin-fluconazole combinations reduced the CFU burdens in the infected nematodes by 14-92% compared to the untreated control which accumulated 250 ± 25 CFU/ml (Fig. 5c). As expected, single fluconazole treatments failed to reduce the CFU burdens in nematodes infected with the fluconazole-resistant isolates (C. albicans NR-29448 or C. auris 390). However, single treatments with fluconazole at 8 or 32 µg/ml were able to reduce CFU burdens of C. glabrata ATCC MYA-2950 by only 26 and 57% respectively, though more potent activities were attained with combination treatments as shown earlier. Altogether, these results are encouraging for future evaluation of the pitavastatin-fluconazole combination in higher animal models. www.nature.com/scientificreports www.nature.com/scientificreports/  www.nature.com/scientificreports www.nature.com/scientificreports/ conclusion The present study characterized pitavastatin as a promising agent for sensitizing azole-resistant Candida species to the antifungal effect of azoles. Pitavastatin, exhibited broad-spectrum synergistic interactions with fluconazole against a variety of clinically-relevant Candida species, including emerging multi-drug resistant C. auris isolates. Moreover, the pitavastatin-fluconazole combination significantly interfered with Candida's biofilm-forming abilities. Additionally, the pitavastatin-fluconazole combination significantly reduced Candida's CFU burdens in infected C. elegans, suggesting potential clinical importance. Finally, the mechanism of synergy displayed by pitavastatin and fluconazole embroils, at least in part, significant interference with Candida's efflux machinery. Further in vivo studies in higher animals are required to assess the potential of pitavastatin to be repurposed as a promising fluconazole adjuvant for controlling invasive Candida infections in humans. elegans strain AU37 genotype [glp-4(bn2) I; sek-1(km4) X], was co-incubated with cell suspensions of (a) C. albicans NR-29448, (b) C. glabrata ATCC MYA-2950, or (c) C. auris 390 using an inoculum size of ~5 × 10 7 CFU/ml for 3 hours at room temperature. Infected nematodes were washed with PBS and then treated with the pitavastatin-fluconazole combination at the respective concentration. Treatment with PBS, pitavastatin alone, or fluconazole alone served as controls. After 24 hours of treatment, worms were lysed to determine the fungal burden (CFU/worm) after treatment. *Indicates a significant difference between each treatment compared to the non-treated control. Whereas # indicates significant differences between the tested pitavastatin-fluconazole combinations relative to the single treatment with the respective fluconazole concentration. The statistical significance was considered for P < 0.05 as determined by one-way ANOVA with posthoc Dunnet's test for multiple comparisons. Screening of pharmakon library and structurally-related compounds. The Pharmakon 1600 drug library was screened against C. albicans NR-29448, a strain that displayed high-level resistance to several azole antifungal drugs. Briefly, C. albicans NR-29448 was diluted to approximately 0.5-2.5 × 10 3 cells/ml in RPMI 1640 medium buffered with 0.165 M MOPS reagent. An aliquot (100 µl) of the fungal suspension was transferred to the wells of a round-bottomed 96-well microtitre plate containing 16 µM of each drug. The plates were then incubated for 24 hours at 35 °C. Drugs that only inhibited the growth of C. albicans in the presence of fluconazole were identified as "positive hits".
Biofilm inhibition assay. Three Candida species, C. albicans NR-29448, C. glabrata HM-1123, and C. auris 385 demonstrated a prominent ability to form robust adherent biofilms. As such, these strains were used to study the antibiofilm activity of the pitavastatin-fluconazole combination. The microtiter biofilm formation assay using crystal violet was used, as previously described 4,21 . Briefly, overnight cultures of the tested Candida strains, grown in YPD broth, were diluted in RPMI 1640 medium to approximately 1 × 10 5 CFU/ml. Then 100 µl aliquots of each suspension were transferred to wells of tissue-culture treated polystyrene 96-well plates. Pitavastatin (at 0.5 × MIC) was added either individually or in combination with fluconazole (2 µg/ml) and the plates were then incubated for 24 h at 35 °C. Following incubations, adherent biofilms were then rinsed twice with phosphate-buffered saline (PBS) and left to dry at room temperature. Air-dried biofilms were stained with crystal violet (0.01%). Stained biofilms were rinsed thrice with PBS and then air-dried. The resultant biofilm biomasses were quantified by dissolving the crystal violet-stained biofilms in absolute ethanol before recording absorbance values (OD 595 ).
Nile Red efflux assay and flow cytometry. Nile red efflux assay was performed following a previously reported protocol [59][60][61] . Briefly, exponential phase Candida cells were harvested by centrifugation (3,000 × g, 5 minutes), washed thrice with PBS, and incubated for an additional 2 hours at 35 °C with shaking (200 rpm). Cells were incubated overnight on ice, then resuspended at a concentration of ~1 × 10 7 cells per ml in HEPES-NaOH (50 mM; pH 7.0) containing 7.5 mM nile red and incubated at 35 °C for 30 minutes. Stained cells were washed three times with cold HEPES-NaOH (50 mM; pH 7.0). Cell suspensions were transferred onto opaque 96-well plates containing two-fold serial dilutions of the test agents. Glucose at final concentration 10 mM was used to initiate the nile red efflux. Detection of nile red fluorescence intensity was commenced about 15 seconds after glucose addition (T 0 ) and then in one-minute intervals for 10 minutes. Nile red fluorescence intensity was measured at an excitation wavelength of 485/9 and an emission wavelength of 528/15 using the SpectraMax i3x microplate reader (Molecular Devices, CA, USA). For flow cytometric analysis, pitavastatin (at 0.25 × MIC) was added to nile red-loaded cells as previously described, then glucose (at final concentration 10 mM) was used to initiate the nile red efflux. After 10 minutes of adding glucose, cells were fixed in 2% paraformaldehyde and were examined in a Canto II flow cytometer (BD Bioscience, San Jose, CA), following a previously reported protocol 62 . Data were analyzed using FlowJo software v10 (Tree Star, Ashland, OR).
Rhodamine Rh6G efflux assay. Rhodamine 6G efflux assay was conducted following a previously reported protocol 63 . Briefly, exponential growth phase Candida cells (SC-TAC1 G980E ) were harvested as described earlier, washed thrice with PBS, and incubated for additional 2 hours at 35 °C with shaking (200 rpm) to induce starvation. The cells were then resuspended at a concentration of ~1 × 10 7 cells per ml in HEPES-NaOH (50 mM; pH 7.0) buffer containing rhodamine 6G (10 mM) and 2-deoxyglucose (5 mM). Cells were incubated with shaking for 90 minutes at 30 °C, to permit rhodamine accumulation under energy-depleting conditions. Rhodamine-stained cells were harvested and washed at least five times with HEPES-NaOH to remove extracellular rhodamine. Pitavastatin at 0.25 × MIC, clorgyline (5 µg/ml), or the vehicle (1% DMSO) was added to the cells and incubated for 5 minutes at 30 °C. Rhodamine efflux was induced by glucose addition at a final concentration of 10 mM. After 10 minutes of adding glucose, cells were harvested by centrifugation, and 100 µl aliquots of cell supernatants were transferred to 96-well plates for detecting the amount of effluxed rhodamine. The rhodamine fluorescence