Antifungal In Vitro Activity of Pilosulin- and Ponericin-Like Peptides from the Giant Ant Dinoponera quadriceps and Synergistic Effects with Antimycotic Drugs

Venoms from ants comprise a rich source of bioactive peptides, including antimicrobial peptides. From the proteome and peptidome of the giant ant Dinoponera quadriceps venom, members of five known classes of antimicrobial peptides were disclosed (e.g., dermaseptin-, defensin-, ICK-, pilosulin- and ponericin-like types). Based on comparative analysis, these family members have structural determinants that indicate they could display antimicrobial activities. In previous works, pilosulin- and ponericin-like peptides were demonstrated to be active against bacteria, fungi, and parasites. Herein, the antifungal activity of ponericin- and pilosulin-like peptides were assessed, aiming at the expansion of the knowledge about AMPs in predatory ants and the development of new microbicide strategies to deal with difficult-to-treat fungal infections. Synthetic pilosulin- (Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides were evaluated for their fungicide and fungistatic activities against different species of Candida, including a drug-resistant clinical strain. The MICs and MLCs were determined for all peptides individually and in combination with general antifungal drugs by the microdilution method. The time-kill kinetic curves were set up by means of a luminescent reagent, of which the light signal is proportional to the number of viable cells. The candicidal synergism observed by the combination of subinhibitory concentrations of peptides and general antimycotic drugs were quantified by the checkerboard test and fluorescent dye permeation assay. The influence of ergosterol on the antifungal activity was verified by supplementation of culture medium. The pilosulin- (Dq-2562 and Dq-1503) and ponericin-like (Dq-3162) were the most active peptides, displaying a broad spectrum of antifungal activity in vitro, with MICs in the range of 0.625 to 10 µM. The combination of peptides and conventional antimycotic drugs displayed a synergistic reduction in the MIC values of individual peptides and drugs, while soluble ergosterol in the culture medium increased the MICs. The fungicide and fungistatic activity of the individual peptides and peptides in combination with antimycotics were time-dependent with a rapid onset of action and long-lasting effect, which involved membrane disruption as an underlying mechanism of their action. Altogether, pilosulin- and ponericin-like peptides from the giant ant D. quadriceps venom display a broad-spectrum of candicidal activity, what allows their inclusion in the row of the antifungal peptides and gives support for further studies on the development of strategies to fight candidiasis.


Peptides
The synthetic pilosulin-and ponericin-like peptides derived from D. quadriceps venom studied in this work are represented in Table 1. Accordingly, the synthetic peptides are linear, short sequence peptides identical to parental venom-derived peptides, which are the products of post-translational modification that released encrypted peptides from a longer preproprecursor, as is the case of pilosulin-like peptides. Synthetic ponericin-like peptide (Dq-3162) is identical to the natural post-translationally amidated-peptide found in the venom. Both pilosulin-and ponericin-like peptides virtually adopt alpha-helical topological structures, as seen by predicted models, with localized hydrophobic faces (Dq-2562 and Dq-3162) and values of hydrophobic moment (µ H ) that are indicative of high degree of helicity and amphiphaticity with membrane-seeking properties, affinity for lipid membranes and propensity to form pores (higher H values and lower µ H values). All peptides have net positive charges, ranging from +2 to +5 (Table 1). Table 1. Structures, predict models and physicochemical characteristics of synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides from the venom of giant ant Dinoponera quadriceps.

Peptide Primary and Secondary Structures (a) Helical Wheel Plot (b) Physicochemical Properties (c,d)
Dq-1319

Minimum Inhibitory Concentration (MIC) and Minimum Lethal Concentration (MLC)
The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 (Table 2). MW) were as reported in our previous work [33].

Minimum Inhibitory Concentration (MIC) and Minimum Lethal Concentration (MLC)
The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 ( The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 (Table 2). The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 ( The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 (Table 2). The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 ( The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 ( The synthetic pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides presented a broad-spectrum of antifungal activity in vitro, with fungistatic and fungicidal effects against different species and strains of Candida, including a clinical strain resistant to amphotericin B and fluconazole. The most active peptide with the lowest MIC and MLC was Dq-3162. The only peptide that showed antifungal activity restricted to only two strains and with only fungistatic action was Dq-1319 (Table 2).

Time-Dependent Antifungal Activity of D. quadriceps Peptides
The fungistatic and fungicidal effects of D. quadriceps peptides were demonstrated through the microbial growth and time-kill kinetics assay using the luminescent BacTiter-Glo®reagent, which reinforced the data obtained from the determination of the MICs and MLCs of all peptides. The onset of peptide action was as fast as 30 min, in all treatments, as measured by the reduction in the luminescent signal and, consequently, in the reduction of microbial cell viability. After 2 h of exposure, the reduction in microbial cell viability, in all treatments, showed statistically significant differences, compared to untreated Candida cells. The antifungal activity of all peptides was sustained up to 24 h of exposure and incubation. The peptides were active against both strains of susceptible and drug-resistant yeasts, and no difference in the pattern of growth kinetics and microbial death were observed between the sensitive (C. albicans ATCC 90028) and amphotericin B-and fluconazole-resistant Candida (C. albicans

Checkerboard Test
The checkerboard test demonstrated that all peptides were able to interact synergistically with antimycotics used in clinical settings against most Candida strains. The combinations prepared with the peptides Dq-3162, Dq-2562, Dq-1503, and Dq-1319 and antimycotic drugs showed synergistic effects in the order of 60%, 52%, 64%, and 40%, respectively. It is important to note that, for the reference strains (C. albicans ATCC 90028), all combinations of these peptides with amphotericin B and nystatin were also synergic, with a reduction in the MICs of these drugs up to eight-and sixteen-fold, respectively. For the amphotericin B-and fluconazole-resistant C. albicans CA1 strain, only the interactions with amphotericin B were synergistic, with a four-fold reduction in the MIC of this drug. For the drug-resistant strain C. krusei ATCC 40095, which also has an intrinsic resistance to fluconazole, the peptide Dq-1503 was able to potentiate the action of fluconazole, reducing its MIC by sixteen-fold (Table 3). Dq-1319

Time-Kill Effect of the Combinations of D. quadriceps Antimicrobial Peptides and Amphotericin B on the Candida Cell Viability
The synergism observed with combinations of peptides and conventional antimycotic drugs was also evaluated by microbial growth and time-kill kinetics using the luminescent BacTiter-Glo®reagent. The combinations that included individual peptides and amphotericin B were selected for this analysis. The time-kill curves obtained in this test series showed similar patterns observed in the kinetics' test for individual peptides alone, i.e., the peptide and antimycotic combinations also reduced the cell viability of both susceptible and resistant strains, in the first 30 min of exposure. The contrastive differences between treated and untreated Candida cells were significant in a time course of 2 h of exposure. Between 2 h and 24 h of treatment, the combination of the four peptides individually and amphotericin B reduced the cell viability of both susceptible and resistant strains, to a minimum of cell viability, close to the detection limit of the assay, indicating a long-lasting fungicidal effect (Figure 3, Figure 4).
In Figure 3, the time-kill curves show the antifungal activity of the synthetic peptides in combination with amphotericin B against C. albicans ATCC 90028, at their subinhibitory concentrations. As can be seen, the four individual peptides in combination with amphotericin B are able to reduce the microbial viability within 8 h of exposure, even at concentrations that are 16-fold (for Dq-3162, Dq-2562, and Dq-1503) and 2-fold (for Dq-1319) lower than the MICs for the peptides alone. As noticed in Figure 4, after 4 h (for Dq-2562 and Dq-1503) or 8 h (for Dq-3162), at sub-MIC concentrations, the activity of peptides that were not associated with amphotericin B against C. albicans CA1, was still fungistatic. The fungistatic effect is also observed with amphotericin B at sub-MIC concentrations within the same time-frame.

Membrane Permeabilization Induced by Pilosulin-and Ponericin-like Peptides
According to data from Table 4 and Figure S1, on the accumulation of the fluorescent dye SYTOX®Green in microbial cells with disrupted cytoplasmic membranes, the most active peptides pilosulin-(Dq-2562) and ponericin-like (Dq-3162) peptides, at their MICs, caused an increase in the cell number that stained positive for this cell impermeant nucleic acid stain, as consequence of the increased number of cells with compromised cytoplasmic membranes (54 to 76%, in 4 h of exposure). The synergistic combinations of peptides and amphotericin B, at their sub-MICs, caused similar effect of compromising Candida cell membranes and cell death (50 to 73%). Untreated Candida cells showed low uptake of SYTOX®Green, reflecting the low percentage of dead cells (4 to 5%). The average cell size of Candida cells treated with either individual peptides, amphotericin B or the combinations of peptides and amphotericin B, was reduced by up to 50% for both standard and drug-resistant clinical isolate strains in comparison to untreated cells (Table S1).     In Figure 6, the percentage of hemolysis is shown for (1)   In Figure 6, the percentage of hemolysis is shown for (1) the individual peptides and amphotericin B at their MIC-equivalent concentrations for the drug-resistant clinical strain C. albicans CA1 (for Dq-3162, Dq-2562, and Dq-1503) and for the standard strain C. albicans ATCC 90028 (for Dq-1319); and (2) for the combinations of individual peptides and amphotericin B, at their subinhibitory MICs (SUBMIC), which are equivalent to those concentrations that synergistic effects were observed for the same strains used in the checkerboard test. As can be seen, the combinations of the individual peptides and amphotericin B, at their subinhibitory concentrations (sub-MICs), in which that peptide and antimycotic drug together are synergistically fungicide, hemolysis activity was negligible, in vitro.
Antibiotics 2020, 9, x 2 of 24 MICs (SUBMIC), which are equivalent to those concentrations that synergistic effects were observed for the same strains used in the checkerboard test. As can be seen, the combinations of the individual peptides and amphotericin B, at their subinhibitory concentrations (sub-MICs), in which that peptide and antimycotic drug together are synergistically fungicide, hemolysis activity was negligible, in vitro.

Effect of Soluble Ergosterol on the Antifungal Activity of Peptides
The presence of increasing amount of soluble ergosterol in the culture medium altered proportionally the MICs of all peptides for the different species of Candida sp tested, from 1-to 2order (Table 5). For instance, for the peptide Dq-3162, the increase of the MICs increased from 0.0625 to 10 μM, in the presence of 0 to 800 μg mL −1 of ergosterol, respectively. Similarly, the MIC of amphotericin B, used as a positive control, was affected by the presence of increasing concentrations of soluble ergosterol in the culture medium; the amphotericin MIC increased from 0.5 μM to 32 μM, at concentrations of soluble ergosterol of 0 and 800 μg mL −1 , respectively, in the culture medium.

Effect of Soluble Ergosterol on the Antifungal Activity of Peptides
The presence of increasing amount of soluble ergosterol in the culture medium altered proportionally the MICs of all peptides for the different species of Candida sp tested, from 1-to 2-order (Table 5). For instance, for the peptide Dq-3162, the increase of the MICs increased from 0.0625 to 10 µM, in the presence of 0 to 800 µg mL −1 of ergosterol, respectively. Similarly, the MIC of amphotericin B, used as a positive control, was affected by the presence of increasing concentrations of soluble ergosterol in the culture medium; the amphotericin MIC increased from 0.5 µM to 32 µM, at concentrations of soluble ergosterol of 0 and 800 µg mL −1 , respectively, in the culture medium. The hyphen denotes lack of activity ("-": inactive). Concentration is in µM and (µg mL −1 ) for peptides and amphotericin B.

Discussion
The venom of hymenopterans (ants, bees, and wasps) and other arthropods (e.g., spiders, scorpions, and centipedes) have been recognized as a rich source of antimicrobial peptides among other types of biologically active peptides, with a vast diversity of structures and broad-spectrum of molecular and cellular targets [21,40,41]. The crude venom of the giant ant Dinoponera quadriceps was previously observed to display in vitro antiparasitic effect against Leishmania amazonensis and Trypanosoma cruzi, mediated by apoptosis and necrosis [35]. The antibacterial activity of D. quadriceps crude venom against Staphylococcus aureus was also reported and the underlined mechanism of antimicrobial action was suggested to involve the interference with cell membrane permeability [34]. Synthetic dinoponeratoxin-derived peptides that are related to ponericins and pilosulins from the D. quadriceps venom showed antiparasitic activity in vitro, which were also mediated by apoptosis, in T. cruzi [36]. In disk diffusion test, the potential antimicrobial activity of isolated D. quadriceps peptides was observed in Gram-positive and Gram-negative bacteria, dermatophyte fungi and yeasts [32].
Recently, we have described five families of peptides present in the D. quadriceps venom peptidome and characterized the mast cell-degranulating, hemolytic, and antimicrobial activities of representative synthetic dinoponeratoxin-related peptides [33]. In that work, it was verified that such pilosulin-(Dq-2562) and ponericin-like (Dq-3162) peptides were active against bacteria and yeasts. Based on these previous reports, herein, the antifungal activity was forward investigated with the synthetic version of the linear, processed encrypted peptides, related to ponericin family of peptides, the peptides Dq-2562, Dq-1503, and Dq-1319, and the 28-residue ponericin-like peptide, Dq-3162. As above mentioned, revealed by the peptidome analysis of crude D. quadriceps venom, the peptides Dq-1319 and Dq-1503 are the processed post-translational peptide products of the parental Dq-2562, while Dq-3162 is the amidated peptide product of larger precursor of 30 amino acid residues [25,33]. Thus, post-translational modification operates in the venom gland of the giant ant D. quadriceps to increase the diversity and stability of peptide structures and, supposedly, to augment the range of biological activities in the venom.
Regarding the antifungal effects of pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides of D. quadriceps, the results from the determination of the MICs and MLCs (Table 2) are corroborated by the time-kill kinetics (Figures 1 and 2) and pointed out that the peptides Dq-3162, Dq-2562, and Dq-1503 display the most efficient fungicide activity. In general, membranolytic antimicrobial peptides (AMPs) are characterized by their short size, presence of hydrophobic and polar sides (i.e., amphipathicity), as well as helicity [42], and, indeed, these D. quadriceps peptides display most of these structural determinants. However as aforementioned, the topological structures and mechanisms of antimicrobial activity are numerous [17]. The same is valid for the antifungal peptides [18]. With exception of Dq-3162 (a ponericin-related), the pilosulin-like peptides Dq-2562, Dq-1503, and Dq-1319 are largely hydrophobic (Table 1) with the highest values of µH (hydrophobic moment). All peptides possess net positive charges. It is known that amphipathicity, cationicity, hydrophobicity, and charge distribution are important attributes for membrane interaction, membrane insertion, membranolysis, antimicrobial activity, and the level of cytotoxicity (e.g., hemolysis) [43,44]. As the first step of microbial killing, positively charged peptide usually interact with the pathogen membrane by electrostatic attraction via negatively charged components on bacterial (or yeast) plasma membrane. This is followed by peptide-membrane insertion, destabilization, and, eventually, pore formation, lysis, and cytoplasmic extravasation [45,46]. In concordance with these overall structural features and physicochemical properties, the peptide Dq-3162 and Dq-2562 displayed a considerable candicidal activity (Table 2) and proportionally better efficacy/cytotoxicity ratios. Such efficacy/cytotoxicity ratios are improved by the combinations of individual peptides and antimycotic drugs (Table 3). Indeed, by using a cell membrane impermeant, molecular probe that is specific for staining nucleic acids (SYTOX®Green), which permeate only cells with compromised cytoplasmic membranes, it was verified that membrane disruption is involved in the mechanism of action of these D. quadriceps ponericin-and pilosulin-like peptides alone, as well as in combinations with amphotericin B (Figure S1). Notably, once exposed to D. quadriceps antifungal peptides, the average size of Candida cells that were treated with either the individual peptides or the peptides in combinations with amphotericin B, was substantially reduced, as determined by cell size (contour) analysis in the Countess™ II FL Automated Cell Counter. This observation suggests a level of shrinkage of Candida cells that were exposed to peptides and a membrane disruption effect mediated by the peptides and their combinations (Table S1). The reduction of cell size and cell shrinkage that result from membrane permeability have been observed for other antimicrobial peptides that act by disrupting membrane integrity of microbes, such as Bac8c -a synthetic cationic peptide derived from bactenecin, found in bovine neutrophils [47]. Moreover, as demonstrated by the time-kill kinetics with a luminescent reagent, the pilolusin-and ponericin-like peptides, as seen here, displayed a rapid onset of fungicide action (~30 min) and a long-lasting (up to 24 h) antifungal effect in vitro (Figures 1 and 2).
Despite the appreciable antifungal activity of these peptides, a possible problem with all antimicrobial peptides for future development as therapeutic alternatives against microbial infection is their inherent cytotoxicity, like exemplified by hemolysis. This is particularly true whether the concentrations required to kill or inhibit the microbial growth are closed to the concentrations that provoke (cyto-)toxicity. In fact, venom and venom-peptides from ants and insects commonly cause cell lysis, like hemolysis, and tissue damage that are intrinsic biological properties that contribute to the overall envenomation symptoms [48,49]. Indeed, a number of AMPs are active due to their membranolytic mechanism of action [50,51], therefore these peptides are invariably and potentially cytotoxic whether they are not selective to the pathogen targets. Thus, thinking about the potential use of AMPs from venom, two immediate practical concepts could be envisioned and solutions carried out: One is the reduction of the potential cytotoxicity by amino acid substitution in a peptide-design program [52,53]. The second, more straightforward, is the reduction in the concentration of cytotoxic compounds through additive and, synergistic combinations and formulations, including combination of peptides and drugs. This later maneuver makes possible not only the maintenance of the therapeutic potency exerted by the individual substances in the drug combinations, but also makes some improvements due to the use of lower concentrations of the active substances. Moreover, the combination of drugs, including peptides, with distinct mechanisms of action is granted for fighting the emergence of drug resistance in pathogenic microbes. Thus, the search for synergistic combinations of AMPs and general antibiotics used in clinical settings are advantageous in decreasing the toxicity toward nontargeted healthy cells and restoring the susceptibility of microbes to the current drugs in the arsenal of antifungal chemotherapeutics [8,54,55]. In fact, examples of the combination of AMPs and general antifungals (e.g., amphotericin B and fluconazole) have been reported, as exemplified by β-peptide, crotalicidin-derived, and thionin-like peptides [56][57][58][59].
Herein, the pilosulin-(Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides were all able to act synergistically with amphotericin B, by reducing their respective MICs and restoring the susceptibility of drug-resistant strains of Candida (Table 3). The synergistic effect was evidenced, by the checkerboard test, through the combination of peptides with the polyenic antifungals (nystatin and amphotericin B), which act by binding to ergosterol and, consequently, by altering the membrane permeability of yeasts [55]. The interaction of peptides with amphotericin B against C. albicans is interesting, seen that such an antifungal compound is a drug-of-choice in the treatment of systemic fungal infections, despite causing nephrotoxicity [8,55]. For this reason, the time-kill kinetics of the combinations of peptides and antimycotics were assessed with amphotericin B against strains of C. albicans. This kind of assessment indicated that fungicide activity, through the exposure to peptides and amphotericin B combinations (Figures 3 and 4), displayed, similarly, a rapid onset of action and prolonged fungicidal effect, as observed for all individual peptides. Furthermore, the peptide-amphotericin B combinations, at their sub-MICs, preserved the effectiveness to kill Candida cell mediated by membrane-disruption (Table 4), but also reduced substantially the in vitro cytotoxicity as ascertained by the hemolysis assay (Figures 5 and 6). Thus, the combination of existing therapies with adjuvant substances that improve antifungal activity, reduce the toxicity of the therapeutics, restore the MICs of resistant strains to the MICs of susceptible microbes, could contribute to overcome the resistance mechanisms of microorganisms to drugs. Moreover, they prolong the shelf life of current antifungals and, therefore, they are highly advantageous and should be further explored [54]. The analysis of the time-kill kinetic curves in Figures 3 and 4 also call the attention to the effect exerted by the peptides alone, in subinhibitory concentrations. Although the candicidal effect was not total within 24 h, the subinhibitory concentrations of the peptides were sufficient to inhibit the yeast growth and/or to reduce the microbial cell viability for a considerable period of time (for 4 to 8 h). In addition, at these subinhibitory concentrations, the peptides did not show an appreciable in vitro cytotoxicity to red blood cells. These experimental evidences give support for further studies and usage of the combination of peptides and drugs at concentrations inferior to their MICs [60].
Finally, as above mentioned and experimentally confirmed for D. quadriceps peptides, by a fluorescent dye permeation assay with SYTOX®Green, membrane-targeting and permeabilization are intrinsic properties of most AMPs to exert their biological effect. The action on pathogen membranes can be accompanied of either membrane lysis and cytoplasmic extravasation or intracellular penetration and interaction with intracellular targets, among other particular mechanisms of action. Several reports have detailed the action of AMPs on the cell membranes of microorganisms [18,50,51,61,62] and pilosulin-and ponericin-like peptides from D. quadriceps could be added to these groups of membranolytic peptides. Thus, since the cell membrane of yeasts is mostly made up of ergosterol, it was important to investigate the influence of soluble ergosterol in the culture medium, on the activity of peptides, as the primary approach. Here, it was evidenced that the presence of soluble ergosterol in the culture medium caused changes in the MICs of all peptides, as well as the MIC of the polyenic antifungal amphotericin B, in a concentration-dependent fashion (Table 5). Thus, it is speculated the peptides might interact with or be sequestered by the soluble ergosterol in the culture medium, instead of binding directly to the membrane ergosterol on the surface of yeast membrane, consequently, decreasing the peptide to cell ratio and requiring higher concentrations of peptides to produce the antifungal effect. The influence of ergosterol in the activity of antifungal substances and extracts were previously observed by others [63,64]. In fact, the metabolism of ergosterol is the target for most clinically relevant antifungals and this sterol might have other biological roles in fungi than a structural participation in the composition of membranes [65]. Further, biophysical experiments should be conduct to investigate the peptide and lipid interaction with the aim of deep understanding the membranolytic mechanism of action of D. quadriceps peptides.
Altogether, the linear, amphiphilic pilosulin-(Dq-2562, Dq-1503, and Dq-1319), and ponericin-like (Dq-3162) peptides perform an efficient antifungal effect against susceptible and azole-resistant clinical strain of Candida albicans. Moreover, in combination with general antimycotic compounds (e.g., amphotericin) the cytotoxicity is reduced to a negligible level. In such peptide and antifungal drug combinations, the effective MICs for individual peptides and antifungal drug combined are decreased to sub-MICs, restoring the amphotericin B MIC to values found for susceptible strains of Candida. Moreover, the membrane perturbation mediated by D. quadriceps peptides appears as an important underlying mechanism of Candida cell death, in reason of the direct increase of membrane permeation to fluorescent dye and reduction of average cell size and cell shrinkage, as well as indirectly by the influence of soluble ergosterol on the antifungal activity of peptides. Further studies are invaluable to investigate the detailed mechanisms of peptide interaction with biological membranes and their components, as well as with other cellular and molecular targets. Last but not least, the formulation of D. quadriceps peptides and antimicrobial chemotherapeutics in combinations would be a valuable strategy for future research.

Pilosulin-and Ponericin-Like Peptides from the Giant Ant D. quadriceps
The pilosulin-like peptides, Dq-2562, Dq-1503, and Dq-1319, were commercially prepared by solid phase peptide synthesis (China Peptides Co. Ltd., Shangai, China), with a minimum purity grade of 95%, as ascertained by HPLC and mass spectrometry analysis. The ponericin-like peptide Dq-3162 was also synthesized commercially (GeneScript Bioech Corp., Piscataway, NJ, USA), but it was a kind gift from Dr. K. Konno, the Institute of Natural Medicine, University of Toyama, Toyama, Japan. The peptides were weighed, solubilized in sterile deionized water to obtain a 1 mM stock solution, and stored in aliquots at -20 • C, until use.

Microorganisms
The microorganisms used here were obtained from the Collection of Reference Microorganisms in Health Surveillance of the National Institute for Quality Control in Health, of the Oswaldo Cruz Foundation (CMRVS, FIOCRUZ-INCQS, Rio de Janeiro, Brazil). The amphotericin B-and fluconazole-resistant strain Candida albicans CA1 was obtained from a regional clinical laboratory.

Determination of Minimum Inhibitory Concentration (MIC) and Minimum Lethal Concentration (MLC)
The determination of the MICs was as recommended by the Clinical and Laboratory Standards Institute in the reference method M27-A3 [66]. In order to prepare the inoculum, the Candida strains were cultured in Sabouraud broth, at 35 • C, for 24 h. Aliquots of these cultures were transferred to sterile test tubes containing 0.85% saline in order to obtain a turbidity equivalent to 0.5 McFarland standard (~10 6 CFU/mL). This suspension was diluted to obtain inoculum with 10 4 CFU/mL, which was used in the analysis. For determination of the MIC, 100 µL of Sabouraud broth with serial dilution of peptides (from 20 µM to 0.009 µM) and 100 µL of the microbial inoculum were mixed together in the wells of sterile 96-well microplates. The microplates were incubated at 35 • C for 24 h. After this period, the microbial growth was inspected by eyes. The MIC was considered the lowest concentration of the substance capable of inhibiting the fungal growth. Control of medium sterility (culture medium only), growth control (microbial suspension in culture medium) and control with a known antifungal drug (e.g., microbial suspension in culture medium containing amphotericin B) were included in the MIC determination. To find the MLC, 5 µL-aliquots of peptide and microbial suspension mixtures from the individual wells, which did not show microbial growth, in the MIC assay, were transferred to Petri-dishes containing Plate-Count agar and incubated at 35 • C for 24 h. Colonies were counted and the MLC was considered e the lowest concentration of peptide that caused, at least, 99.9% of cell death, compared to the inoculum used initially [67,68]. The MICs and MLCs of amphotericin B, miconazole, fluconazole, nystatin and cyclopyrox were also determined for all strains, according to the same methodology described above. The tests were performed in triplicate and the data were used to select the peptides and strains used in the subsequent tests.

Determination of Time-Kill Kinetics of Peptide Activity Against C. albicans
This assay was performed with the strain C. albicans ATCC 90028 and the drug-resistant clinical strain C. albicans CA1, the pilosulin-like peptides Dq-2562 and Dq-1503, as well as the ponericin-like peptide, Dq-3162. To sterile microtubes, 100 µL of peptide solution, at concentrations equivalent to the MIC and 2 × MIC, were added and mixed with 100 µL of microbial inoculum (0.5 McFarland standard, 10 6 CFU/mL). The microtubes were incubated at 35 • C and, in a time-frame of 0.0, 0.5, 1, 2, 4, 8, 16, and 24 h, and 50 µL were taken and transferred to 96-well flat bottom black microplates. Then, 50 µL of the BacTiter-Glo®reagent (Promega, Madison, WI, USA) was added to every well and the plates were incubated for 5 min, in the dark, under gentle agitation. Luminescence was measured with the Synergy HT multiple detection microplate reader (BioTek, Winoski, VT, USA). Untreated microbial suspensions and treated with a general antifungal drug (amphotericin B) were used as controls. Also, wells containing only Sabouraud broth were used to obtain background values of luminescence. The test was performed in triplicate.

Checkerboard Test
This test was performed with the strains C. albicans ATCC 90028, C. tropicalis ATCC 13803, C. krusei ATCC 40095, C. parapsilosis ATCC 40038, and C. albicans CA1. The reference antimycotic drugs used were amphotericin B, fluconazole, miconazole, cyclopyrox, and nystatin. The peptides were tested based on the results found in the determination of individual MICs. In 96-well microplates, 50 µL peptide solution in culture medium and 50 µL of individual antifungal were mixed and diluted in Sabouraud broth to reach final concentrations of 1 2 , 1 4 , 1/8 and 1/16 × MIC. Then, 100 µL of microbial suspension (10 4 CFU/mL) were transferred and homogenized by gently mixing. The microplates were incubated at 35 • C, for 24 h and microbial growth inspection by eyes. The test was performed in triplicate. Controls of medium sterility (culture medium only) and cell growth (microbial suspension in culture medium) were included [69]. To assess the antifungal activity of the combinations of D. quadriceps antimicrobial peptides and general antimycotics, the fractionated inhibitory concentration index (FICI) was calculated as follows: FICI = (MIC of the peptide alone / MIC of the peptide in the combination) + (MIC of the antifungal alone / MIC of the antifungal in the combination) The FICI was interpreted as: Synergistic effect, when FICI ≤ 0.5; additive effect, when FICI > 0.5 and < 1.0; and antagonistic effect, when FICI > 1.0 [70,71] To check the membrane disruption effect caused by the most active peptides, Dq-2562 and Dq-3162, a cell impermeant fluorescent dye, SYTOX®Green nucleic acid stain (Molecular Probes, Eugene, OR, USA) was used. This dye penetrates only cells with damaged cytoplasmic membranes and is impermeable to live cells with intact membranes. In this assay, the two strains of Candida albicans were tested, as described for time-kill kinetics, in 4.2.3. Aliquots of 100 µL cell suspension of the strain C. albicans ATCC 90028 and the drug-resistant clinical strain C. albicans CA1, 0.5 McFarland standard (10 6 CFU/mL), were incubated (35 • C, 4 h) with the peptides at their MICs and the sub-MICs when in combinations with amphotericin B for observation of their synergistic effect. After treatment, cells were collected by centrifugation (12,000× g, 4 • C, 2 min) and the culture medium removed and discarded. Cells were resuspended in the same original volume with 10 mM HEPES, 150 mM NaCl, pH 7.6, and SYTOX®Green (10 µM) added. After gently homogenization, the suspensions were incubated for 10 min and the dead cells, with disrupted membranes, counted with the Countess™ II FL Automated Cell Counter (ThermoFisher Scientific, Waltham, MA, EUA), in the GFP channel. Total cell number (live and dead cells) was enumerated in the brightfield. The contour (shape) analysis estimated the average Candida cell size of populations that were to the peptides and drug in comparison with untreated cells. The ratio of dead green/live cells was expressed in percentage. Untreated strain suspensions and treated with amphotericin B were used as controls.

Hemolytic Assay
Fresh human blood from consented healthy donor was collected into vacutainer tubes (BD, Beckton Dickinson, São Paulo, Brazil) and centrifuged at 1000× g, for 10 min, at 4 • C. The plasma was removed and the pellet containing the red blood cells was rinsed three times with sterile PBS (35 mM phosphate, 150 mM NaCl, pH 7.4). After rinsing, cell pellet was resuspended in PBS to obtain an 8% (v/v) suspension of red blood cells. A 100 µL aliquot of this suspension was transferred to sterile microtubes containing 100 µL of individual peptide solutions (at concentrations ranging from 40 µM to 0.3125 µM). The final concentrations of red blood cells and peptides, in the hemolytic assay, were 4% (v/v) and 20 µM to 0.1562 µM of peptides, respectively. After incubation (37 • C, 60 min), under slight agitation, the microtubes were centrifuged at 1000× g for 2 min, the supernatants were collected and transferred to 96-well flat bottom clear plates and the quantity of hemoglobin released from lysed erythrocytes was estimated at 540 nm, with the Synergy HT multiple detection microplate reader (BioTek, Winoski, VT, USA). Untreated red cell suspensions and treated with 1% Triton X-100 were used as negative and positive controls, respectively [72]. The percentage of hemolysis was determined as: (DO 540nm peptide and/or drug treated cells − DO 540nm untreated cells) / (DO 540nm Triton X-100 treated cells − DO 540nm untreated cells) × 100.
The hemolytic activity of some peptide and antifungal combinations was also assessed using the same protocol described above, except by using 50 µL of each individual substance used in this study. All tests were performed in triplicate.

Influence of Ergosterol on the MIC of Peptides and Amphotericin B
The MICs of peptides for the strains C. albicans ATCC 90028, C. tropicalis ATCC 13803, C. krusei ATCC 40095, and C. parapsilosis ATCC 90038 were determined by the microdilution technique as described in 4.2.2. (reference method M27-A3, CLSI [66]), except that the medium was supplemented with increasing concentrations of ergosterol (100, 200, 400, and 800 µg/mL) in the culture medium. The assay was carried out in triplicate. Negative (culture medium) and positive growth controls (microbial suspension in culture medium) were included with amphotericin B used as a reference antifungal drug.

Conclusions
In the present work, it was evidenced that synthetic pilosulin (Dq-2562, Dq-1503, and Dq-1319) and ponericin-like (Dq-3162) peptides, identical to the parental peptides found in the venom of D. quadriceps giant ant, display a broad-spectrum of antifungal activity against all species of Candida sp. tested. With fungistatic and fungicidal effects, these D. quadriceps venom-derived peptides possess a rapid onset of action and long-lasting antifungal effect. In addition, these peptides are able to synergistically modulate the action of general antifungal drugs of different classes, as exemplified here. Although the individual peptides caused certain level of in vitro cytotoxicity to human red blood cells, at concentrations equivalent to their MICs, this was not the case when in combinations with amphotericin B, at their sub-MICs, which ensured the antifungal effect at such very low concentrations, but with negligible hemolytic activity. The peptides alone, even at subinhibitory concentrations, are also sufficient to sustain the antifungal effect for 4 to 8 h, without appreciable hemolysis. Thus, the pilosulin-and ponericin-like peptides from the venom of D. quadriceps, alone or in combination with current antimicrobial drugs, were demonstrated here to be potentially promising for the development of effective alternatives against fungal infections. The fact of these peptides exerts their effect by disrupting the plasma membrane is an advantageous mechanistic feature to make difficult the development of drug-resistance in pathogenic yeasts. Further studies are required to understand the detailed mechanisms of membrane interaction and destabilization, as well as to search for additional cellular and molecular targets of these pilosulin-and ponericin-like peptides. Moreover, from the viewpoint of pharmaceutical sciences, the study of combination and formulation of D. quadriceps peptides and antimicrobial chemotherapeutics could be promising for the antimycotic adjuvant medication.