Different cDNAs encoding putative AMPs from the venom gland of scorpions T. obscurus, Tityus costatus, Hadrurus gertschi, and Opisthacanthus cayaporum were characterized by expressed sequence tag (EST). The amino acid sequences deduced from ESTs as described in literature (Diego-Garcia et al., 2005; Schwartz et al., 2007; Silva et al., 2009; Guerrero-Vargas et al., 2012) were considered to obtain the synthetic peptides as specified bellow (Table 1).

Comparative analyses of amino acids sequences were performed employing the Basic Local Alignment Search Tool (BLAST¹) and Collection of Anti-Microbial Peptides (CAMP²) databases to identify peptides with at least 40% of identity to our queries sequence and to which antimicrobial activity was previously determined. The multiple sequence alignments were done using the M-Coffee tool³ on the sequences identified in the previous step, and the consensus sequences were represented as a sequence logo, generated at Weblogo web server⁴.

Peptides were chemically synthesized using FMOC-Butila (AminoTech Pesquisa e Desenvolvimento LTDA, Brazil), and further purified by reverse-phase high performance liquid chromatography (HPLC). Peptides ToAP1, ToAP2S1, ToAP3, NDBP-4.23, ToAP4, and NDBP-5.7 were synthesized with C-terminal amidation. To assess the purity and sequence correctness, peptides were submitted to matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF/TOF MS; UltraFlex III, Bruker Daltonics, Germany), under reflector (MS) and LIFT^TM (MS/MS) positive modes. Matrix solution was prepared using 5.0 mg of α-cyano-4-hydroxycinnamic acid (Sigma, USA) and solubilized with 250 μL of acetonitrile, 200 μL of deionized water, and with 50 μL of an aqueous trifluoroacetic acid solution (at 3% by volume). The aqueous solution of the peptide was mixed with a saturated matrix solution (1:3 peptide/matrix) and air-dried. The monoisotopic molecular mass of the ion corresponding to the peptide of interest was determined by the ratio between m/z peaks in the spread profile (m/z ratio from 600 to 3,000). Additionally, interpretation of the MS/MS mass spectra and de novo sequencing of peptides was performed using the FlexAnalysis 3.0 software (Bruker Daltonics, Germany). Throughout the experiment, the stability of each peptide was reevaluated using the same mass spectrometry parameters detailed above. Only samples containing more than 95% of the peptide (high purity) were used for biological experiments. Peptides were stored at -20°C and dissolved in Milli-Q water before each experiment.

The putative secondary structures and the α-helix content (%) were predicted using the software PHD as described previously (Guo et al., 2013)⁵. The hydrophobicity and the helical hydrophobic moment of the peptides were determined in the HeliQuest web server⁶. Other physicochemical properties were determined by ProtParam tool⁷ (Wilkins et al., 1999). Helical wheel projections were generated at the respective web server⁸.

Candida albicans strain SC 5314 (ATCC MYA-2876) was provided by Dr. Joshua Nosanchuk (New York, NY, USA). C. glabrata ATCC 90030, C. parapsilosis ATCC 22019, and C. tropicalis ATCC 750 were provided by Dr. Érika Kioshima (Universidade Estadual de Maringá, Paraná, Brazil). C. neoformans var. grubii strain H99 (serotype A, ATCC 208821) was provided by Dr. John Perfect (Durham, NC, USA). C. neoformans var. neoformans B3501 (serotype D, ATCC 34873) was obtained from ATCC (Manassas, VA, USA). All strains were stored as frozen stocks in 35% glycerol at -80°C. Before each experiment, all Candida spp. strains were grown in Sabouraud’s dextrose broth overnight at 30°C with agitation. C. neoformans strains were grown in the same conditions for 24 h. Fungal cells were collected by centrifugation, washed three times in phosphate buffered saline (PBS) (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 2 mM KH₂PO₄) and inoculated in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with L-glutamine and buffered to pH 7.0 with 165 mM 3-(N-morpholino)propanesulfonic acid for experiments described below.

In vitro antifungal assays were performed according to broth microdilution susceptibility test from Clinical and Laboratory Standards Institute (CLSI) M27-A3 guidelines with some modifications. Briefly, twofold serial dilutions of each peptide were prepared in 96-well polystyrene microplates to a final volume of 50 μL. Amphotericin B (Amp B; A2942, Sigma-Aldrich), a known antifungal drug, was used as a positive control. Final concentrations of each peptide and Amp B ranged from 100 to 0.78 μM and 16 to 0.03 μg/mL, respectively, except in tests with C. glabrata and C. parapsilosis, where the final peptide concentration ranged from 400 to 0.78 μM. In each plate, wells were included without peptide as a growth control. Then, 50 μL of the adjusted inoculum in RPMI-1640 medium were added to each well to a final concentration of 2 × 10³ cells/mL for Candida spp. and 10⁴ cells/mL for C. neoformans strains. The plates were incubated at 37°C for 24 and 48 h, respectively. The strains of C. neoformans were also incubated with shaking (200 rpm). The minimum inhibitory concentration (MIC) was defined as the lowest AMP concentration that completely inhibited visible fungal growth at the end of the incubation period. The experiments were performed at least three times on separate dates.

Considering the potential use of the AMPs as treatment against pathogens, even if in a more distant future, some of their properties against host cells should be addressed. To achieve this goal, we have performed hemolytic assay to evaluate if these molecules are toxic or harmless to red blood cells. Further, we have also tested the possible cytotoxicity of those peptides against other mammalian cell types by following the cell viability score after treatment. The solution of human red blood cells was treated with the peptide at levels ranging from 100 to 0.78 μM. Whole human blood was collected from healthy donors in vials with ethylenediaminetetraacetic acid (EDTA). The red blood cells were separated by centrifugation and resuspended in PBS. After the centrifugation (800 g, 5 min at 4 ± 1°C) the supernatant was removed and the red blood cells pellet was resuspended to a 3% hematocrit in sterile water (considered as 100% hemolysis) and PBS (taken as blank for spectrophotometric measurement). Fifty microliters of the 3% cell suspension in PBS were added to different final concentrations of the peptide (v/v) in a 96-well polystyrene microplate and then incubated for 1 h (Nahar et al., 2008; Zhao et al., 2009; Qi et al., 2010). The microplates were centrifuged and the supernatant was collected and analyzed at 540 nm using a microplate spectrophotometer. The percentage of hemolysis was calculated relative to the positive control as performed by Nahar et al. (2008). The procedures were approved by the Ethics Committee on Human Research of Faculty of Medicine/University of Brasília (UnB) and conform to FDA standards (UnBDoc no. 66704/2016).

BALB/c mice were injected with thioglycolate 72 h before peritoneal macrophages were harvested. They were diluted to a concentration of 10⁶ cell/mL in RPMI-1640 supplemented with penicillin/streptomycin (100 U/mL, 100 g/mL), 2 mM L-glutamine, 2 mM non-essential amino acids, 1 mM sodium pyruvate (all reagents from Sigma-Aldrich, St Louis, MO, USA) and 10% fetal bovine serum (FBS), and incubated in 96-well polystyrene microplates at 37°C, 5% CO₂ for 24 h. The peptides were added then and cells were further incubated for 24 h as described previously (Cruz et al., 2005). Three independent experiments were performed. Cytotoxic activity was evaluated by the release of cytoplasmic enzyme lactate dehydrogenase due to cell lysis using the Cytotox kit (Promega). Briefly, at the end of the incubation period, supernatants were collected, the adhered cells were washed and the toxicity was determined according to the manufacturer’s protocol. The procedures were approved by the Ethics Committee on Animal Research of Institute of Biology (UnBDOC 52657/2011)/University of Brasília (UnB).

The effects of the peptides NDBP-4.23, ToAP1, ToAP2, and NDBP-5.7 on C. albicans biofilms were analyzed according to the protocol described by Pierce et al. (2008). Briefly, C. albicans SC 5314 (10⁶ cells/mL) was grown on 96-well polystyrene microplates and 4 h after seeding the peptides or Amp B were added to assess their effects on the first step of biofilm formation, the initial cell adherence. After 24 h of treatment, biofilm formation was measured by the metabolic assay based on the reduction of 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT). For the purpose of investigating effects of peptides on mature biofilms, the Candida cells were allowed to grow for 24 h prior to addition of the peptides or Amp B and the XTT assay was performed in the same way. All results were analyzed for statistical significance using analysis of variance (ANOVA) followed by Tukey post-test using GraphPad Prism 6 and expressed in terms of biofilm viability as XTT-readings percentage normalized by the control groups. Each compound was tested in triplicate and the experiments were performed at least twice on separate dates.

Circular dichroism (CD) assays were carried out using Jasco J-815 spectropolarimeter equipped with a Peltier-type temperature controller. Peptides (0.2 mg/mL) were analyzed in water at 37°C in absence or the presence of trifluoroethanol (TFE) (10, 30, and 50%). The mean spectrum of three consecutive experiments was corrected for the baseline buffer contribution. The observed ellipticities were converted into molar ellipticity ([θ]) based on molecular mass per residue of 115 Da (Adler et al., 1973). The α helix secondary structure content (fθ) was estimated considering the values of [θ]_{208 nm} as a function of TFE concentration using the following equation (Greenfield and Fasman, 1969):

The primary structures of ToAP2, ToAP1, ToAP3, ToAP4, and ToAcP peptides from T. obscurus were used as queries on BLAST. Data analysis revealed that ToAP2 is similar to peptides previously described as belonging to NDBP subfamily 3 and to AMPs from ant venom (Figures 1A,B). ToAP1, ToAP3, and ToAP4 showed similarity to AMPs belonging to NDBP subfamily 4 (Figures 1C,D). ToAP4 also has 100% identity to putative AMP clone 6 from T. costatus [UniProt: Q5G8B3]. ToAcP did not show similarity to any AMPs described before and shares 90% identity to Toxin Tp3 [UniProt: P0DL22], a peptide from Tityus pachyurus with no toxicity to mammalian cells and without known biological activity (Barona et al., 2006). The result of these analyses indicates that peptide ToAP2 belong to NDBP subfamily 3 and ToAP1, ToAP3, ToAP4 belongs to NDBP subfamily 4.

The results of in silico secondary structure and physicochemical property prediction for the peptides are summarized in Table 2. Except for ToAP2 and ToAcP, which have net charges at pH 7.0 of +6 and -4, respectively, all others showed a net charge of +1. All peptides have a predicted helical content of at least 53.85% and, except for ToAcP, had positive hydrophobicity values, which means that in theory they are slightly hydrophobic. ToAP2S1, an analog from ToAP2, had different physicochemical parameters. Its net charge was reduced from +6 to +4 as a result of the substitution of two lysines for leucines; and both its hydrophobicity and hydrophobic moment were increased. ToAP2S1 also had its predicted helical content increased from 57.69 to 88.46% relative to its prototype ToAP2. Helical wheel projections predict that ToAP2, ToAP2S1, ToAP1, ToAP3, and ToAP4, once adopting a helical conformation, are amphipathic peptides, thus presenting distinct hydrophobic and hydrophilic faces (Figure 2). ToAcP, however, presents only a hydrophilic face. The lack of a distinct hydrophobic face compromises its amphipathicity.

Antifungal assays were performed against different species of Candida and C. neoformans (Table 3).

Seven peptides were active against Candida spp. and C. neoformans. ToAP2 and Con10, both from NDBP subfamily 3, had antifungal activity against all strains tested with MICs values ranging from 3.12 to 200 μM and 12.5 to 200 μM, respectively. Among peptides from NDBP subfamily 4, only ToAP4 has not shown antifungal activity in the tested concentrations (data not shown). The other peptides from this subfamily presented antimicrobial activity against all Candida spp. and C. neoformans in the same MIC range (6.25–200 μM), except for C. glabrata, which is insensitive to them. C. glabrata is also resistant to all peptides from NDBP subfamily 5. Among those subfamily peptides, NDBP 5.7 presented antifungal activity with MIC values ranging from 12.5 to 25 μM and NDBP 5.8 inhibited the growth of pathogens tested at concentrations ranging from 50 to 200 μM. NDBP-5.6 and ToAcP did not present antifungal activity in this assay (data not shown).

Since ToAP2 showed the lower MIC against all pathogens tested in this work, derivatives of its structure were produced in order to increase the α-helical content. The modified peptide, ToAP2S1 (Table 1) was synthesized and tested against two Candida species (C. albicans and C. glabrata) and C. neoformans H99. ToAP2S1 did not show antifungal activity in the tested concentrations (data not shown), demonstrating that the changes in peptide sequence abolished antifungal activity relative to the reference peptide.

Based on results from antimicrobial assay, we decided to proceed only with those peptides (at least one per family) that showed best antimicrobial activity.

The hemolytic activities of peptides ToAP2, NDBP-4.23, ToAP1, and NDBP-5.7 were evaluated against human erythrocytes (Figure 3). ToAP2S1 was also tested to evaluate if it also has lost its hemolytic activity. NDBP-4.23 and ToAP1 might be considered of interest to future studies, because in all the tested concentrations their hemolysis curves were significantly lower than 50% of hemolysis (Conlon et al., 2007; Thaker et al., 2011; Figures 3B,C). However, NDBP-5.7 presents a hemolysis percent higher than 50% in concentrations up to 50 μM (Figure 3D). Comparing the original peptide ToAP2 and the modified ToAP2S1, the last one shows a higher hemolysis curve with concentrations up to 25 μM while the ToAP2 maintained its curve above 50% of hemolysis (Figure 3A).

Toxicity to mammalian cells was further evaluated by assessing the viability of murine peritoneal macrophages in the presence of the peptides (Figure 4). The threshold of acceptable cytotoxicity was set at 50% viability as postulated elsewhere (Thaker et al., 2011; Zhao et al., 2012). In conformity with the hemolytic test, ToAP2 and NDBP-4.23 in concentrations below 6.25 and 25 μM, respectively, have shown low cytotoxicity (Figures 4A–C). For ToAP1 and NDBP-5.7, peritoneal macrophages displayed higher viability in the whole range of peptide concentrations (Figures 4B–D). The difference in the hemolytic curves between ToAP2 and its analog (ToAP2S1) could be explained by the higher hydrophobicity of ToAP2S1 (Table 2) and consequent tendency to self-associate. Chen et al. (2007) demonstrated that there is an optimal hydrophobicity window for AMPs, below which antimicrobial activity can be lost and above which self-association becomes highly likely. Peptide aggregation prevents penetration through the capsule and cell wall of microorganisms (Jiang et al., 2008). However, the higher hydrophobicity does not affect their activity on the eukaryotic membrane (Chen et al., 2007), which explains the lower antifungal activity and the higher hemolytic activity of the modified ToAP2S1, respectively.

We visualized C. albicans biofilms at 4 and 24 h by fluorescence microscopy using uvitex 2B, which binds chitin in the fungal cell wall, and they exhibited typical features of mature biofilms, including a basal layer of yeast cells and a thicker layer of yeast and hyphae cells (Supplementary Material). This structure was very similar to the morphology of C. albicans biofilms presented in other works in the literature (Pierce et al., 2014, 2015). Additionally, the inhibition values of biofilms by Amp B were similar to those found by Ramage et al. (2001, 2002).

The effects of ToAP2, NDBP-4.23, ToAP1, and NDBP-5.7 on C. albicans biofilms were evaluated on the initial adherence step (Figure 5A) and on the mature biofilm (Figure 5C). As a control, the effect of Amp B was also shown at both time points (Figures 5B,D), with a reduction of at least 80% in biofilm viability at 0.25 μg/mL for the initial adherence and 2 μg/mL for the three-dimensional architecture. All the peptides were active against biofilm formation, albeit at a higher concentration compared to what was observed against planktonic cells (100 μM for the initial adherence and 400 μM for the three-dimensional architecture). ToAP2 was the most active peptide at the two different stages of biofilm formation, with a lower effective concentration than the others (25 μM for the initial adherence and 200 μM for the three-dimensional architecture).

The secondary structures of native (ToAP2) and modified (ToAP2S1) peptides were assessed by CD spectroscopy. FAR-UV CD spectra of ToAP2 in water at 37°C revealed a typical disordered structural pattern (Figure 6A). At increasing concentrations of 10, 30, and 50% TFE (v/v), a red shift from 200 to 208 and 222 nm and increased negative dichroic signals were observed (Figure 6A). These results show that ToAP2 lacks secondary structure in aqueous environments and has a tendency to adopt an α-helical pattern at aqueous environments in the present of the co-solvent TFE (Table 4). In contrast, an α-helical conformation was observed for the modified peptide ToAP2S1 in water at 37°C even in the absence of TFE. In the presence of 10, 30, and 50% TFE, intense negative dichroic bands at 208 and 222 nm were observed (Figure 6B), which are compatible with higher α-helix content (Table 4). Fractional helicity of peptides in water and TFE environments was calculated considering the molar ellipticity at 208 nm (Table 4). There was an increased α-helix content for both peptides as a function of TFE, with ToAP2S1 presenting approximately 17% of α-helix in water in the absence of the co-solvent.