Lower Concentrations of Amphotericin B Combined with Ent-Hardwickiic Acid Are Effective against Candida Strains

Life-threatening Candida infections have increased with the COVID-19 pandemic, and the already limited arsenal of antifungal drugs has become even more restricted due to its side effects associated with complications after SARS-CoV-2 infection. Drug combination strategies have the potential to reduce the risk of side effects without loss of therapeutic efficacy. The aim of this study was to evaluate the combination of ent-hardwickiic acid with low concentrations of amphotericin B against Candida strains. The minimum inhibitory concentration (MIC) values were determined for amphotericin B and ent-hardwickiic acid as isolated compounds and for 77 combinations of amphotericin B and ent-hardwickiic acid concentrations that were assessed by using the checkerboard microdilution method. Time–kill assays were performed in order to assess the fungistatic or fungicidal nature of the different combinations. The strategy of combining both compounds markedly reduced the MIC values from 16 µg/mL to 1 µg/mL of amphotericin B and from 12.5 µg/mL to 6.25 µg/mL of ent-hardwickiic acid, from isolated to combined, against C. albicans resistant to azoles. The combination of 1 µg/mL of amphotericin B with 6.25 µg/mL of ent-hardwickiic acid killed all the cells of the same strain within four hours of incubation.


Introduction
Invasive candidiasis is the cause of unacceptable high mortality rates ranging from 30 to 70% in different parts of the world [1][2][3], and the treatment of life-threatening Candida infections has been limited to just three drug classes [2,4].
The emergence of Candida resistance to the available antifungal drugs has compromised the clinical management of this disease [1,4,5], and failure of antifungal treatment is due to multifactorial events involving molecular modifications related to drug mechanism of action and over-expression of efflux pumps, among other factors [6,7].
One mechanism of resistance of Candida species to azoles, for example, is the occurrence of point mutations in the ERG11 gene [7,8]. Azoles interfere with the ergosterol biosynthesis pathway in fungal membranes by inhibiting the cytochrome P450-dependent enzyme 14α-demethylase, which is synthesized by the ERG11 gene [7]. Mutations that resulted in amino acid substitutions decreased azoles susceptibility [7,8].
Therapeutic failures with echinocandins are also reported for Candida infections [6,9]. Candida strains with reduced susceptibility to echinocandins showed mutations in the FKS genes that correlated with amino acid substitutions in the 1,3-β-D-glucan synthase, the target of echinocandins [6,9].
Resistance of Candida species to polyenes is still uncommon compared to resistance to other antifungal drugs [6,10]. However, Candida species resistant to amphotericin B have been reported for clinical isolates [6,10]. A different ergosterol structure that prevents Copaifera oleoresins are traditionally used by people from the Brazilian Amazonian region as an anti-inflammatory [43], antimicrobial [44] and antiparasitic [45], and literature data support the ethnopharmacological uses of this crude material [43].
The diterpene ent-hardwickiic acid has been highlighted as a lead compound in the search for bioactive compounds [46,47] and has been reported due to its anti-inflammatory [43], antibacterial [44], antifungal [47] and schistosomicidal activities [48].
Despite having several biological activities, it should be pointed out that this diterpene did not show cytotoxic activity against normal and cancer human cell lines [49].
This study reports for the first time the in vitro interaction between ent-hardwickiic acid and amphotericin B by using the checkerboard microdilution method against C. albicans and C. krusei strains, including a C. albicans strain resistant to azoles isolated from bloodstream infections in a tertiary care hospital in Brazil [50]. In addition, time-kill assays were performed to assess the fungistatic or fungicidal nature of different combinations of ent-hardwickiic acid and amphotericin B concentrations.
Considering that only one new azole and two new formulations of posaconazole have been launched in the market in the last decade [51] and that the need for new antifungal drugs is urgent, the strategy of combining ent-hardwickiic acid with amphotericin B was shown to be potentially effective at a low concentration of amphotericin B against Candida strains.

Results
The minimum inhibitory concentrations (MICs) of amphotericin B and ent-hardwickiic acid were first determined using the broth microdilution method. Amphotericin B showed MIC values of 8 µ g/mL against C. albicans ATCC 10231 and C. krusei ATCC 6258 and 16 µ g/mL against a C. albicans strain resistant to azoles. The MIC values of ent-hardwickiic acid were smaller than those found for amphotericin B (6.25 µ g/mL against C. albicans ATCC 10231, 3.12 µ g/mL against C. krusei ATCC 6258 and 12.5 µ g/mL against a C. albicans strain resistant to azoles).
MIC values of fluconazole were also determined for C. albicans ATCC 10231 (12.5 µg/mL) and the quality control strain Candida parapsilosis ATCC 22019 (4 µ g/mL) to assure that the antifungal microdilution test was performed appropriately [52,53]. Our results were reproducible, and the MIC values are within the proposed range for these strains.
It should be pointed out that C. krusei is considered intrinsically resistant to fluconazole [53], and the clinical isolate of C. albicans used in this study also showed resistance to fluconazole [50]. In this study, the MIC values of fluconazole against C. krusei ATCC 6258 and C. albicans resistant strain were 25 ug/mL and greater than 100 ug/mL, respectively.
The combination of amphotericin B (1) and ent-hardwickiic acid (2) was then assessed by using the checkerboard microdilution method and synergistic (ƩFIC ≤ 0.5) and additive (ƩFIC > 0.5) interactions of compounds 1 and 2 were found for tested strains (Table 1). Antagonism was not detected. Copaifera oleoresins are traditionally used by people from the Brazilian Amazonian region as an anti-inflammatory [43], antimicrobial [44] and antiparasitic [45], and literature data support the ethnopharmacological uses of this crude material [43].
The diterpene ent-hardwickiic acid has been highlighted as a lead compound in the search for bioactive compounds [46,47] and has been reported due to its anti-inflammatory [43], antibacterial [44], antifungal [47] and schistosomicidal activities [48].
Despite having several biological activities, it should be pointed out that this diterpene did not show cytotoxic activity against normal and cancer human cell lines [49].
This study reports for the first time the in vitro interaction between ent-hardwickiic acid and amphotericin B by using the checkerboard microdilution method against C. albicans and C. krusei strains, including a C. albicans strain resistant to azoles isolated from bloodstream infections in a tertiary care hospital in Brazil [50]. In addition, time-kill assays were performed to assess the fungistatic or fungicidal nature of different combinations of ent-hardwickiic acid and amphotericin B concentrations.
Considering that only one new azole and two new formulations of posaconazole have been launched in the market in the last decade [51] and that the need for new antifungal drugs is urgent, the strategy of combining ent-hardwickiic acid with amphotericin B was shown to be potentially effective at a low concentration of amphotericin B against Candida strains.

Results
The minimum inhibitory concentrations (MICs) of amphotericin B and ent-hardwickiic acid were first determined using the broth microdilution method. Amphotericin B showed MIC values of 8 µg/mL against C. albicans ATCC 10231 and C. krusei ATCC 6258 and 16 µg/mL against a C. albicans strain resistant to azoles. The MIC values of ent-hardwickiic acid were smaller than those found for amphotericin B (6.25 µg/mL against C. albicans ATCC 10231, 3.12 µg/mL against C. krusei ATCC 6258 and 12.5 µg/mL against a C. albicans strain resistant to azoles).
MIC values of fluconazole were also determined for C. albicans ATCC 10231 (12.5 µg/mL) and the quality control strain Candida parapsilosis ATCC 22019 (4 µg/mL) to assure that the antifungal microdilution test was performed appropriately [52,53]. Our results were reproducible, and the MIC values are within the proposed range for these strains.
It should be pointed out that C. krusei is considered intrinsically resistant to fluconazole [53], and the clinical isolate of C. albicans used in this study also showed resistance to fluconazole [50]. In this study, the MIC values of fluconazole against C. krusei ATCC 6258 and C. albicans resistant strain were 25 ug/mL and greater than 100 ug/mL, respectively.
The combination of amphotericin B (1) and ent-hardwickiic acid (2) was then assessed by using the checkerboard microdilution method and synergistic (ΣFIC ≤ 0.5) and additive (ΣFIC > 0.5) interactions of compounds 1 and 2 were found for tested strains (Table 1). Antagonism was not detected. The combination of both compounds at determined concentrations markedly reduced the MIC values, and a synergistic effect was detected when 4 µg/mL of amphotericin B was combined with 3.12 µg/mL of ent-hardwickiic acid against a C. albicans strain resistant to azoles. Additive effects were detected with 1 and 2 µg/mL of amphotericin B combined with 6.25 µg/mL of ent-hardwickiic acid, with 0.125 µg/mL of amphotericin B combined with 1.00 µg/mL of ent-hardwickiic acid and with 8 µg/mL of amphotericin B combined with 1.56 µg/mL of ent-hardwickiic acid against the same resistant strain.
Synergistic effects were detected against the reference strains of C. albicans and C. krusei in the range of amphotericin B concentrations from 0.031 µg/mL to 2 µg/mL and from 0.0156 µg/mL to 2 µg/mL, respectively. The ent-hardwickiic acid concentrations varied in the same assay from 0.39 µg/mL to 3.12 µg/mL and from 0.195 µg/mL to 0.78 µg/mL, respectively.
In order to assess the fungistatic or fungicidal nature of different combinations of ent-hardwickiic acid and amphotericin B concentrations, time-kill assays were performed using four combinations of amphotericin B and ent-hardwickiic acid concentrations for each strain that resulted in growth inhibition at amphotericin B concentrations lower than the MIC value of this antifungal agent alone. The four selected combinations for each strain are presented in Table 2. The fourth combination containing 4 µg/mL of amphotericin B and 3.12 µg/mL of ent-hardwickiic acid killed all C. albicans resistant strain cells within 2 h (Figure 2a). The combination of 1 µg/mL of amphotericin B with 6.25 µg/mL of ent-hardwickiic acid killed all the cells of the same strain within 4 h of incubation. The other antifungal combinations did not show fungicidal activity within the 24 h incubation period, but exhibited a significant reduction in the growth of this strain.
Among the tested combinations of ent-hardwickiic acid and amphotericin B concentrations against the reference strain of the C. albicans, the combination of 0.5 µg/mL of amphotericin B and 0.78 µg/mL of ent-hardwickiic acid was the one that showed the highest growth reduction within 24 h (Figure 2b). During this period of 24 h, ent-hardwickiic acid and amphotericin B combinations exhibited a significant reduction in the growth of C. albicans reference strain, but the fungicidal point was not detected.
The fourth combination containing 4 μg/mL of amphotericin B and 3.12 μg/mL of ent-hardwickiic acid killed all C. albicans resistant strain cells within 2 h (Figure 2a). The combination of 1 μg/mL of amphotericin B with 6.25 μg/mL of ent-hardwickiic acid killed all the cells of the same strain within 4 h of incubation. The other antifungal combinations did not show fungicidal activity within the 24 h incubation period, but exhibited a significant reduction in the growth of this strain.  The same behavior was observed for the time-kill curves of ent-hardwickiic acid and amphotericin B combinations against the C. krusei reference strain. All the curves showed a significant reduction in growth during 24 h without achieving the fungicidal point in this period. The combination of 0.25 µg/mL of amphotericin B and 0.39 µg/mL of ent-hardwickiic acid showed the highest growth reduction within 24 h (Figure 2c).
Considering the results obtained from time-kill assays, it should be highlighted that the fungicidal activity of two combinations of ent-hardwickiic acid and amphotericin B concentrations against C. albicans resistant strain was very fast (2 to 4 h), which can be clinically relevant.

Discussion
The aim of this study was to evaluate for the first time the potential of ent-hardwickiic acid combined with amphotericin B against Candida strains.
Amphotericin B was licensed in 1959 and after more than sixty years is still the main antifungal agent used to treat invasive fungal infections [54,55]. However, its principal chronic adverse effect is nephrotoxicity, whose clinical manifestations range from hypokalemia to kidney insufficiency [56].
Among several complications that might arise after SARS-CoV-2 infection is the acute kidney injury that affects over a quarter of patients hospitalized with COVID-19 disease [57]. The clinical management of these patients includes hemodynamic support and avoidance of nephrotoxic drugs [58].
Drug combination strategies have the potential to reduce the risk of side effects due to a reduction of effective dose of each compound without loss of therapeutic efficacy [59].
In this study, the combination of ent-hardwickiic acid and amphotericin B markedly reduced the MIC values when compared with those of drugs alone. The combination was effective in using lower concentrations of each compound than those needed to achieve the same effect of each isolated compound. In addition, two combinations of ent-hardwickiic acid and amphotericin B concentrations exhibited fungicidal activity against C. albicans resistant strain after 2-4 h of incubation.
The combination of 1 µg/mL of amphotericin B with 6.25 µg/mL of ent-hardwickiic acid killed all the cells of C. albicans resistant strain within four hours of incubation. This concentration of amphotericin B in plasma has not been associated with toxic effects and drug discontinuation [60]. According to the literature, the pharmacodynamic characteristics of amphotericin B indicate that after the administration of doses of 0.6 to 3.0 mg/kg of body weight/day of amphotericin B deoxycholate (Bristol-Myers Squibb), the mean maximum concentrations (C max s) achieved in serum are 1.1 to 3.6 µg/mL [60,61]. In the presence of serum, amphotericin B loses its fungicidal activity, but remains with its fungistatic activity [62].
There are no studies yet about the stability of ent-hardwickiic acid in human serum, but this compound has been highlighted as a lead compound in the search for bioactive compounds [46,47]. In previous studies, this diterpene showed fungistatic and fungicidal effects against C. glabrata at lower concentrations than fluconazole and its derivatives obtained by biotransformation reactions exhibited potent antifungal activity [47].
Regarding the toxicity of ent-hardwickiic acid, a study carried out in normal and tumor human cell lines showed that this diterpene was not cytotoxic to the tested cell lines [43,49,63], as well as to the RAW 264.7 cells, which are monocyte/macrophage-like cells reported as an appropriate model of macrophages [45]. In addition, this compound did not affect the animal's locomotor capacity in open-field and rotarod tests [43].
Many marketed drugs have a natural product origin, and the majority of these successful natural products were formulated to interact with biological systems to achieve their therapeutic potential [64,65]. Natural products may also provide different mechanisms of action, since these compounds are optimized by evolution to be useful in the defense of organisms [66]. As an example to be cited, macrocyclic diterpenes were able to overcome multidrug resistance in C. albicans as potent inhibitors of drug efflux pumps [67].
With regard to ent-hardwickiic acid, there is only one study reporting the mechanism of action of this diterpene against Streptococcus mutans (ATCC 25175) and Porphyromonas gingivalis (ATCC 33277) [68]. The authors performed assays to determine cell membrane integrity by leakage through the bacterial membrane of nucleic acids and protein. The results indicated that the diterpene ent-hardwickiic acid damaged the S. mutans and P. gingivalis cell membranes, causing cellular component release followed by the release of cytoplasmic material [68]. Further studies are necessary to elucidate the mechanism of the antifungal action of this compound.
The interest in natural products to provide drug leads has been revitalized mainly with the aim of overcoming the resistance of microorganisms to antimicrobial agents [66].
In conclusion, the results of the present study indicate that the combination of amphotericin B and the natural product ent-hardwickiic acid has the potential to inspire the development of treatment options for life-threatening Candida infections.

Candida Strains
Candida albicans ATCC 10231, C. krusei ATCC 6258 and C. parapsilosis ATCC 22019 were acquired from American Type Culture Collection (ATCC, Rockville, MD, USA). The C. albicans strain resistant to azoles was isolated from bloodstream infections in a tertiary care hospital in Brazil using the Bactec™ 9240 system (Becton & Dickinson, Franklin Lanes, NJ, USA) and provided for this study by Prof. Dr. Márcia E. da Silva Ferreira. This strain was identified with the VITEK ® 2 system (BioMérieux, Marcy l'Étoile, France) and by using molecular techniques [50].

Antifungal Agents
Amphotericin B and fluconazole were acquired from Sigma-Aldrich, and ent-hardwickiic acid was isolated from Copaifera pubiflora oleoresin according to Teixeira and co-workers [47].
The authorizations to undertake scientific studies with C. pubiflora oleoresin were issued under the numbers 35143-1 and 010225/2014-5 from the Brazilian Council for Authorization and Information on Biodiversity (SIBIO/ICMBio/MMA/BRASIL) and Genetic Heritage Management (CGEN/MMA/BRASIL), respectively.

Minimum Inhibitory Concentration of Antifungal Compounds
The minimum inhibitory concentration values of amphotericin B and ent-hardwickiic acid against Candida strains were first determined in triplicate by using the broth microdilution method in 96-well microplates according to the recommendations of the Clinical and Laboratory Standards Institute (document M27-A4) [69]. Amphotericin B and enthardwickiic acid were dissolved in dimethyl sulfoxide (Merck, Saint Louis, USA) and diluted in RPMI 1640 medium to achieve concentrations ranging from 16 µg/mL to 0.0156 µg/mL and from 100 µg/mL to 0.19 µg/mL, respectively. The final content of DMSO was 5% (v/v), and this solution was used as negative control. The fungal inoculum was adjusted to yield a cell concentration of 2.5 × 10 3 CFU/mL. The following controls were included: one inoculated and one non-inoculated well to verify the adequacy of the broth for organism growth and the medium sterility, respectively. Fluconazole was used as positive control and its MIC value was also determined for the quality control strain C. parapsilosis ATCC 22019 to assure that the antifungal microdilution test was performed appropriately [52,53]. The 96-well microplates were incubated at 35 • C for 24 h. After the incubation period, the microorganism viability was also measured by adding 30 µL of resazurin solution (0.02%) to the microplates to confirm the MIC values determined visually [70].

Checkerboard Microdilution Method
The in vitro interactions between amphotericin B and ent-hardwickiic acid were investigated by using the checkerboard microdilution method in 96-well microplates as previously described with adaptations [71]. Amphotericin B and ent-hardwickiic acid were dissolved in dimethyl sulfoxide, and stock solutions of both compounds were prepared in RPMI 1640 medium in the range of concentrations from 4-to 8-fold more concentrated than the highest concentration of each compound to be tested.
In each well of the microplate, 100 µL of growth medium was added, and serial twofold dilutions of amphotericin B and ent-hardwickiic acid stock solutions were mixed in each well, resulting in 77 combinations ( Figure S1).
The MIC values of the isolated compounds were again determined by inoculating only amphotericin B and ent-hardwickiic acid in row H (12-2) and column 1 (A-G), respectively. One well without the antifungals was added as growth control.
The final inoculum was adjusted to yield a cell concentration of 2.5 × 10 3 CFU/mL. A mirror plate without microorganisms and with the same concentrations of compounds was prepared to be used as optical density background in a microplate reader. Both microplates were incubated at 35 • C for 24 h.
The growth in each well was quantified spectrophotometrically at 530 nm in a microplate reader, and the MIC values for each combination of compounds were defined as the concentration of compounds combination or the concentration of isolated compound that reduces microbial growth by more than 80% [71].
The interactions between amphotericin B (1) and ent-hardwickiic acid (2) in different combinations of concentrations were determined based on the calculated coefficient of the sum of fractional inhibitory concentration (ΣFIC) [72]. The ΣFIC is calculated according to the formula: where and The results can be interpreted as follow: ΣFIC ≤ 0.5: synergistic, ΣFIC > 0.5 to ≤ 1: additive, ΣFIC > 1 to ≤ 4: indifferent and ΣFIC > 4: antagonistic.

Time-Kill Assays
Time-kill assays were performed in triplicate for four combinations of amphotericin B and ent-hardwickiic acid concentrations following the procedures proposed for the timekill evaluation of antibacterial agents [73] with adaptations. The assays were also carried out with microorganisms without antifungal agents.
The final inoculum was adjusted to yield a cell concentration of 2.5 × 10 3 CFU/mL. Microplates containing the combinations of compounds and the microorganisms were incubated at 35 • C for 24 h. During this period, aliquots (20 µL) of each well were removed, diluted when necessary and spread onto Sabouraud dextrose agar for counting of viable colonies at predetermined time points (0, 2, 4, 6, 12 and 24 h). The lower limit of accurate and reproducible detectable colony counts was 100 CFU/mL. Time-kill curves were built by plotting log 10 CFU/mL versus time with the aid of the Prism software (version 5.0; GraphPad Software, Inc., Boston, MA, USA).