Programmed cell death in yeast by thionin-like peptide from Capsicum annuum fruits involving activation of caspases and extracellular H+ flux

CaThi is a thionin-like peptide isolated from fruits of Capsicum annuum, which has strong antimicrobial activity against bacteria, yeasts and filamentous fungi, and induced reactive oxygen species (ROS) in fungi. ROS are molecules that appear in the early stages of programmed cell death or apoptosis in fungi. Due to this fact, in this work we analyzed some events that may be related to process of apoptosis on yeast induced by CaThi. To investigate this possibility, we evaluated phosphatidylserine (PS) externalization, presence of active caspases and the ability of CaThi to bind to DNA in Candida tropicalis cells. Additionally, we investigated mitochondrial membrane potential, cell surface pH, and extracellular H+ fluxes in C. tropicalis cells after treatment with CaThi. Our results showed that CaThi induced PS externalization in the outer leaflet of the cell membrane, activation of caspases, and it had the ability for DNA binding and to dissipate mitochondrial membrane potential. In addition, the cell surface pH increased significantly when the C. tropicalis cells were exposed to CaThi which corroborates with ~96% inhibition on extracellular H+ efflux. Taking together, these data suggest that this peptide is capable of promoting an imbalance in pH homeostasis during yeast cell death playing a modulatory role in the H+ transport systems. In conclusion, our results strongly indicated that CaThi triggers apoptosis in C. tropicalis cells, involving a pH signaling mechanism.


Introduction
Antimicrobial peptides (AMPs) are common molecules of the immune defense system in virtually all life forms, with representatives in organisms ranging from bacteria to plants and mammals [1][2][3]. They participate in an ancient defense system, innate immunity, which is the first line of defense for most living organisms during the early stages of an infection [4,5]. These peptides generally exhibit a wide range of inhibitory activity against viruses, bacteria, filamentous fungi, yeasts, protozoa, insects, and others [6][7][8].
AMPs have ∼12-100 amino acid residues, molecular mass less than 10 kDa, and many positively charged lysine or arginine residues that confer a net positive charge at physiological pH [9][10][11]. The presence of a large number of cysteine residues (4, 6, or 8) that are connected in pairs forming disulphide bonds is an important characteristic for these peptides and gives them a high level of stability to extreme physicochemical conditions [12,13]. In general, AMPs present a mechanism of action that involves interaction with different membrane structures, which leads to the formation of pores and culminates in increased permeability, imbalance of ion homeostasis via inhibition on primary H + transport systems and results in cell death [3,[14][15][16][17]. However, the complete mechanism of action of AMPs remains unclear. 20 min of incubation at 30 • C with constant orbital agitation at 500 rpm, the cells were washed in 500 μl PBS and resuspended in 20 μl PBS. Negative (incubated in the absence of CaThi) and positive (incubated with 300 mM acetic acid) control cells had the same treatment as did cells treated with the peptide. Cells were analyzed by DIC using an optical microscope equipped with a fluorescence filter to detect fluorescein (excitation wavelength: 450-490 nm, emission wavelength: 500 nm).

Analysis of mitochondrial functionality
Mitochondrial functionality was assessed by the fluorescent dye Rhodamine 123 (Sigma). Rhodamine 123 is a cationic fluorescent dye that has high affinity with the electrical potential of membranes; thus, it marks active mitochondria in living cells, i.e. the loss of mitochondrial membrane potential is observed as a decrease in the fluorescent signal. Initially, C. tropicalis yeast cells (1 × 10 4 cells.ml −1 ) were incubated in Sabouraud broth containing 10 μg.ml −1 CaThi, with the final assay volume adjusted to 200 μl. The assay was performed on 96-well cell culture plates (Nunc) at 30 • C for 24 h. After incubation with CaThi, C. tropicalis cells were resuspended in the Saboraud medium and incubated with 10 μg.ml −1 Rhodamine 123, with constant orbital agitation at 500 rpm for 2 h and protected from light, and then analyzed by DIC under an optical microscope equipped with a fluorescence filter (excitation wavelength: 506 nm, emission wavelength: 530 nm). Control cells were treated in the same manner, except CaThi was excluded.

Measurements of H + flux using ion-selective vibrating probe system
C. tropicalis was grown in 40 ml Sabouraud broth for 24 h at 30 • C and 250 rpm. After 24 h, an aliquot of 1 μl of cellular suspension of C. tropicalis was transferred to the center of a Petri dish (60 × 15 mm) containing 1 ml Sabouraud agar, and grown for 24 h at 30 • C. After this growth period, 5 ml Sabouraud broth was gently added into the Petri dish, and in order to determine the extracellular voltage differences, proton flux, and cell surface pH, measurements of the H + flux were performed by H + selective vibrating probe.
H + -specific vibrating microelectrodes were manufactured as described by Feijó et al. [32] and Ramos et al. [33]. Micropipettes composed of 1.5 mm borosilicate glass capillaries were stretched and treated with dimethyl dichlorosilane (Sigma-Aldrich, U.K.). After silanization, the microelectrodes were back-loaded with electrolyte solution (15 mM KCl and 40 mM KH 2 PO 4 , pH 6.0 for H + , from 15 to 20 mm microelectrode column) and then front-loaded with respective ion-selective liquid exchange cocktail (Fluka, from 20 to 25 μm electrolyte column). An Ag/AgCl wire electrode holder (World Precision Instruments) was inserted into the back of the microelectrode for electrical contact with the bathing solution. The ground electrode was used as a dry reference (DRIREF-2, World Precision Instruments) and was inserted into the sample bath. The microelectrodes were calibrated by measuring the background signal at the beginning and end of each experiment using standard solutions covering the experimental range of each ion. Both the slope and intercept of the calibration line were used to calculate the respective ion concentration from the mV values measured during the experiments.
The treatment with CaThi was performed in C. tropicalis cells after determination of each H + flux at each colony (assay was done twice and in triplicate, n=6). Data acquisition was stopped and the respective treatments were added in the Petri dishes and the H + flux was measured for a minimum of 5 min or until they reached the steady state. After that, a background reference was taken and H + flux was recorded again.

Analysis of DNA-binding capability of CaThi
DNA from C. tropicalis was extracted using the DNeasy plant mini kit (Qiagen) from a culture of C. tropicalis grown in Sabouraud broth for 24 h. The extracted DNA was quantitated using NanoDrop 2000 (Applied Biosystems). The test is based on the alteration of electrophoretic mobility due to the binding of DNA to protein. This binding results in the formation of a complex resulting in a different electrophoretic mobility than free DNA.
The assay was performed as described by Park et al. [34]. Briefly, a 0.8% agarose gel was prepared according to Sambrook and Russel [35]. One hundred nanograms of DNA from C. tropicalis was incubated with 10 μg.ml −1 of CaThi plus 20 μl binding buffer (5% glycerol, 10 mM Tris/HCl pH 8.0, 1 mM EDTA, 1 mM DTT, 20 mM KCl, and 50 μg.ml −1 BSA), and the mixture was incubated for 30 min at 30 • C. Water was used as a negative control, and 10 μg.ml −1 of poly-l-lysine (Sigma), which has the ability to bind to DNA, served as a positive control. Gel images were captured using an ImageQuant LAS 500 (GE Healthcare).

PS externalization in S. cerevisiae yeast cell by CaThi
Positive labeling for PS induced by CaThi was observed, visualized by the green fluorescent contour on the spheroplast of S. cerevisiae, suggesting that the translocation of PS from the internal leaflet to the external leaflet of the plasma membrane occurred. This result indicates that CaThi may cause an apoptotic effect on S. cerevisiae cells, as observed by PS exposure. A similar effect was observed in the positive control, where the cells were incubated with hydrogen peroxide. However, in the negative control (heated), propidium iodide labeling was observed, indicating necrotic death ( Figure 1).

Caspase activity detection
The detection of caspase activity was performed using the CaspACE FITC-VAD-FMK in situ marker. The FITC-VAD-FMK marker is an analog of the caspase inhibitor Z-VAD-FMK (carbobenzoxy-valyl-alanylaspartyl-[O-methyl] fluoromethylketone). The N-terminus carbobenzoxy (Z) is replaced by FITC, creating a fluorescent label for apoptosis. This complex enters the cell, where it acts as a pseudosubstrate that irreversibly inhibits caspases by binding to the cysteine residue at its active site and becomes fluorescent. To verify the occurrence of apoptotic events induced by CaThi in C. tropicalis cells, the activity of caspase type enzymes was investigated. Our results show an induction of caspase activity, suggesting a programmed cell death occurring by an apoptotic pathway ( Figure 2).

Analysis of mitochondrial functionality in cells of C. tropicalis
The mitochondrial functionality assay was evaluated through the fluorescent dye Rhodamine 123. C. tropicalis cells, after 24 h of treatment with 10 μg.ml −1 CaThi, presented with diminished mitochondrial signals, as observed by the weak fluorescence of the dye Rhodamine 123. Control cells with functional mitochondria evinced a strong signal of Rhodamine 123 fluorescence ( Figure 3). This result indicates that after treatment with CaThi, cells of C. tropicalis lost their electrical membrane potential, causing the dysfunction of their mitochondria.

Analysis of H + flux using an ion-selective vibrating probe
Extracellular voltage difference, proton fluxes, and cell surface pH were measured in C. tropicalis yeast using an H + -selective vibrating probe. The dose-response test showed a significant inhibition of the H + efflux in the yeast cells after treatment with CaThi ( Figure 4). An inverse behavior was observed with the extracellular pH ( Figure 4). A stable H + voltage difference was recorded in the presence or absence of the CaThi peptide ( Figure 5A). Before exposure to 40 μg.ml −1 CaThi, the yeast cells showed a steady state of extracellular H + efflux activity of 11.04 + − 1.39 ηmol.cm −2 .min −1 , but after the peptide treatment, the H + efflux had effluxes of ∼0.36 + − 0.15 ηmol.cm −2 .min −1 (P≤0.0001, t test); i.e. they presented a significant inhibition of 96.84% ( Figure 5B). Consequently, the cell surface pH had a significant increase when the yeast cells were exposed to CaThi ( Figure 5C). However, after 10 min of removal of CaThi from the medium, the basal H + efflux and cell surface pH returned to the normal level, suggesting that this inhibition is CaThi dependent (data not shown).

DNA electrophoretic mobility assay
In a previous work, we demonstrated that CaThi is able to internalize and suggested a possible target in the nucleus of C. tropicalis cells [26]. Thus, we hypothesized that CaThi could interact with the DNA of these cells. Therefore, we looked for evidence of this occurrence by verifying a possible change in DNA mobility. Our result indicates that CaThi at 10 μg.ml −1 was able to bind to the DNA molecule and that this binding results in a complex that has a different electrophoretic mobility from the free DNA, preventing its entry into the agarose gel mesh, as may be observed in the test (second lane) and in the positive control when using poly-l-lysine (third lane). This result was corroborated when we visualized the negative control only in the presence of water, where the normal electrophoretic mobility of the DNA molecule is observed (first lane) ( Figure 6).

Discussion
CaThi causes viability loss in yeast cells (C. tropicalis) and in filamentous fungi (F. solani) and increased ROS production in both fungi [26,27]. Importantly, ROS are the main inducers of apoptosis in fungi [36]. Taken together, this information led us to hypothesize that CaThi triggers an apoptosis pathway in treated fungi. Mitochondria play a crucial role in cell survival by generating ATP and controlling apoptosis, the cell cycle and other essential metabolisms.
Mitochondria generate ATP through oxidative phosphorylation during aerobic respiration, where glucose, pyruvate, and NADH molecules are oxidized, generating ROS as metabolic byproducts [37][38][39]. Therefore, mitochondria and ROS play a major role in the induction of apoptosis under normal physiological conditions. Apoptosis is characterized by noticeable biochemical and physical alterations that occur in the cytoplasm and in cell components such as the nucleus and plasma membrane. These alterations occur in an ordered time frame and are generated by two main signaling pathways, the intrinsic or mitochondrial and the extrinsic pathways; both pathways start with the activation of caspases [40,41]. To investigate the possibility of the induction of apoptosis by CaThi in yeast cells, some of the altered characteristics were evaluated as PS externalization, the presence of active caspases, and the dissipation of To verify the PS externalization, we used the protoplasts of S. cerevisiae as a model. We used this model because the technique of obtaining spheroplasts was already standardized in our group for S. cerevisiae and because of the requirement that the cell wall be absent for the experimental detection of PS externalization. The translocation of PS from the inner leaflet to the outer leaflet of the plasma membrane occurs during the early stages of apoptosis, and this externalization can be detected by the labeling with Annexin V, which binds to phospholipids with negative charges but with high specificity for PS [20]. When our protoplasts were analyzed by fluorescence microscopy, we observed only Annexin V labeling (Figure 1), suggesting that CaThi induces PS externalization. This result indicates that a programmed cell death pathway may have been activated in this yeast. Similarly, other AMPs isolated from different sources are capable of inducing apoptosis in yeast cells. Hwang et al. [20] found that papiliocin peptide, isolated from P. xuthus, induced the accumulation of ROS and free radicals that are important regulators of apoptosis in C. albicans. These cells treated with papiliocin showed a series of changes normally seen in cells undergoing early apoptosis, such as the translocation of PS from the inner side to the outer side of the plasma membrane. Lee et al. [42] demonstrated by Annexin-FITC labeling and the TUNEL assay that coprisin peptide, similar to defensin, isolated from Copris tripartitus (beetle), was involved in both the early and late stages of apoptosis, increasing the intracellular level of ROS and free hydroxyl radicals in C. albicans cells.
After verifying the PS translocation event in the presence of CaThi, we verified the presence of activated caspases and the dissipation of mitochondrial membrane potential by rhodamine 123 [43][44][45]. Caspases are specific aspartate proteases containing cysteines, are typically activated in the early stages of apoptosis and play a central role in apoptotic signaling network [29]. In these two tests, we verified that CaThi increases the level of caspase activity in C. tropicalis cells and could cause the collapse of the mitochondrial membrane potential at the same cells (Figures 2 and  3). The papiliocin AMP also showed these activities, causing the dissipation of mitochondrial membrane potential and the activation of caspases. Some late events of apoptosis have also been observed, such as the fragmentation and condensation of DNA. These data suggest that papiliocin led C. albicans cells to apoptosis via ROS accumulation [20]. Another AMP called psacotheasin, a knotin family peptide, isolated from larvae of the beetle Psacothea hilaris, was able to provoke various apoptotic effects in C. albicans, such as PS externalization, mitochondrial membrane depolarization, and increased activity of caspase, in addition to DNA fragmentation and condensation [46]. In 2011, Cho and Lee [47] suggested that areninin-1 from Arenicola marina (lugworm), has antifungal activity by inducing apoptosis in C. albicans, which caused an increased production of intracellular ROS, mitochondrial membrane depolarization, caspases activation, plasma membrane depolarization, and translocation of PS to the external leaflet of membrane surfaces. This same yeast showed morphological changes in the nucleus and structural changes in the DNA molecule. In conjunction with the results already demonstrated for other AMPs, our data suggest that CaThi possesses antifungal activity by activation of the apoptosis pathway.
The H + fluxes through the plasma membrane exert essential functions in the physiology of the fungal cell. It is normally mediated by the plasma membrane H + -ATPase and interference in this flux can lead to cell death. H + flux was monitored by measuring the H + flux in C. tropicalis cells. The results presented here showed that after treatment with CaThi, the H + efflux had a significant inhibition of ∼96%, thus clearly demonstrating that this peptide is capable of causing an imbalance in H + homeostasis (Figures 4 and 5). Diz et al. [48] and Ribeiro et al. [49] demonstrated that peptides present in a fraction obtained from seeds of Capsicum annuum also inhibited 100% of the H + efflux in S. cerevisiae at a concentration of 160 μg.ml −1 . In contrast, an LTP extracted from coffee (Coffea canephora) called Cc-LTP 1 increased the acidification of the medium, i.e. an increase in the H + efflux, in cells of S. cerevisiae, C. albicans, and C. tropicalis [50]. Therefore, apoptotic induction may alter the ionic gradient across the plasma membrane, leading to an ionic imbalance and the depolarization of this membrane [21,47,51]. Andrés et al. [17] found that a human lactoferrin (hLf), a protein of the innate immune system, affects the activity of mitochondrial and plasma membrane H + -ATPase, inducing an apoptosis-like process in yeast. Indeed, changes in membrane potential and transmembrane ion fluxes are amongst the fastest responses of the cell to external stimuli [52,53]. The H + efflux in yeast cells are linked to the activity of plasma membrane H + -ATPase, which plays a crucial role for cell growth, providing energy for nutrient uptake by channels and secondary transporters [54,55]. The formation of a proton electrochemical gradient is necessary for the maintenance of the intracellular pH and cell survival. Thus, the monitoring of the proton fluxes in yeasts could be a new tool to select innovative and more effective antifungal compounds [55].
Thionins may be a new group of DNA-binding proteins, as reported by molecular modeling studies, suggesting that they may interact with DNA within the microorganism cells [56,57]. In a previous study, Taveira et al. [26] demonstrated that CaThi is internalized in C. tropicalis cells and is localized in the nucleus of these cells. An obvious nuclear target should be DNA, and to verify if CaThi has this binding ability, we performed a CaThi binding assay on DNA. The results demonstrated that CaThi binds to DNA of C. tropicalis, thus reinforcing our hypothesis of a possible nuclear target ( Figure 6). One possibility for microbial inhibition based on DNA binding capability is the blockage of molecular synthesis, as demonstrated by the AMPs PR-39 (from pig) and cryptdin-2 (a Paneth cell defensin from mouse) [58,59]. In general, AMPs are multifunctional molecules as observed by the effects they induced after interaction with microorganism cells and, thus, they are supposed to have many cellular targets which altogether led to microorganism death [60]. Moreover, the effects of PS externalization, caspase activation, mitochondrial membrane depolarization, and H + efflux inhibition which suggest an apoptotic cell death, may be summed to the effect to the DNA binding leading to cell death. Thus, we can assume that directly binding to DNA through electrostatic interactions, as peptides have a positive overall charge while DNA molecule has a negative charge, CaThi may have promoted changes in the cellular DNA structure, leading to cell death [47]. However, whether these in vitro observations have significance in vivo is not known, and more studies will be needed for this direct connection to be fully established.
Herein, we further increase the understanding of the action mechanism of fungal inhibition by CaThi. The data presented in the present study suggest a pH signaling mechanism, in which CaThi triggers cell death in C. tropicalis based on the key markers of yeast apoptosis, such as PS externalization, presence of active caspase, and dissipation of mitochondrial membrane potential. Additionally, we originally showed that CaThi regulates the external pH of the cell, acting directly on the extracellular H + fluxes, which could be an important target to identify if any compound has antifungal function.