Mycotoxins produced by the Fusarium molds can cause a variety of human diseases and economic losses in livestock. Fusaria produce predominantly two types of mycotoxins: the nonestrogenic trichothecenes including T-2 toxin and the mycoestrogens such as zearalenone (ZEN). In a previous report, we demonstrated that the hepatotoxicity of these mycotoxins involves the mitochondrial pathway of apoptosis. Here, we observed that both fusarotoxins induced cell death by a mitochondria-dependent apoptotic process which includes opening of the mitochondrial permeability transition pore complex (PTPC), loss of mitochondrial transmembrane potential, increase in O2·− production, mitochondrial relocalization of Bax, cytochrome c release, and caspase activation. Studies performed on isolated mouse liver mitochondria showed that both ZEN and T-2 toxin might act directly on mitochondria to induce a PTPC-dependent permeabilization of mitochondrial membranes. Moreover, they may target different members of PTPC. Indeed, although the inner membrane protein adenine nucleotide translocase could be the target of T-2 toxin, ZEN seems to target the outer membrane protein voltage-dependent anion channel. Cells pretreatment with the p53 inhibitor pifithrin-α suggested that ZEN but not T-2 toxin triggered a p53-dependent mitochondrial apoptotic pathway. Finally, mitochondrial alterations induced by ZEN and T-2 toxin are mediated by Bcl-2 family proteins, such as Bax, and prevented by Bcl-xL and to a lesser extent by Bcl-2. Taken together, these data indicate that mitochondria play a pivotal role in both ZEN- and T-2 toxin–induced apoptosis and that PTPC members and proteins of Bcl-2 family should be interesting targets to overcome fusarotoxin toxicity.
Mycotoxins belong to a group of hazardous compounds that occur simultaneously in food or raw materials. The toxin species from the genus Fusarium are particularly hazardous for human health. Fusarium species occur widely on plants and are found in a variety of agricultural products such as corn, wheat, and other cereal grains used for human and animal consumption. Trichothecenes and zearalenone (ZEN) constitute the most relevant fusarial toxins, which have been identified as important contaminants in foodstuff for human or animal nutrition. Mainly, due to the diversity of their chemical structures, these fusarial toxins elicit a wide range of toxicological effects (Coulombe, 1993).
When ingested, trichothecenes cause mainly gastrointestinal, irritative, and vague symptoms. Even if ingestion of contaminated food is considered as the main exposure route for mycotoxins, exposure through inhalation and skin contact has recently received an increased attention (Austwick, 1983; Hendry and Cole, 1993; WHO, 1990). These sesquiterpenoid fungal metabolites have a characteristic 12,13-epoxytrichothecene-9-ene ring structure, which has an essential function for the toxicity (Swanson et al., 1987, 1988). Besides, the typical 12,13-epoxytrichothecene-9-ene ring structure, the number and the position of the acetyl groups have also a distinct effect on the toxicity of trichothecenes. A cleavage of the esters results in a reduction of the toxicity in different cell culture experiments (Oldham et al., 1980). Among the trichothecenes, T-2 toxin is the most toxic compound (Fig. 1A) (Gutleb et al., 2002) and is considered to be a major causative agent in fatal alimentary toxic aleukia in humans by affecting the mucosa and the immune system. T-2 toxin is rapidly absorbed after ingestion in most animal species and is distributed in the organism with little or no accumulation in any specific organ. Leucopenia and necrotic lesions of the oral cavity, esophagus, and stomach are the main pathological findings (Canady et al., 2001).
ZEN [6-(10-hydroxy-6-oxo-trans-1-undecenyl)-β-resorcyclic acid lactone] is a nonsteroidal estrogenic mycotoxin extracted from several varieties of Fusarium fungi (Fig. 1B). Studies in various species (e.g., rodents, pigs, and monkeys) have shown that ZEN and its metabolites exhibit estrogenic and anabolic activities (Etienne and Jemmali, 1982). Its strong estrogenic effects are due to its competition with 17-β-estradiol in the binding to cytosolic estrogen receptors present in the uterus, mammary gland, hypothalamus, and pituitary gland (Kuiper-Goodman et al., 1987). Thus, ZEN is frequently associated with hyperestrogenism and several physiological alterations of the reproductive tract in several laboratory animals such as mice, rats, guinea pigs, hamsters, rabbits (Creppy, 2002), and domestic animals (Allen et al., 1981; Olsen et al., 1986). ZEN also induces lipid peroxidation, inhibits protein and DNA syntheses, and induces cell death (Abid-Essefi et al., 2003, 2004). It has been shown to be genotoxic and to induce DNA-adduct formation (Lioi et al., 2004; Pfohl-Leszkowicz et al., 1995), DNA fragmentation, and micronuclei production (Abid-Essefi et al., 2003; Ouanes et al., 2003).
Despite the fact that they are produced by the same fungal species, the various fusarial toxins might involve different signaling pathways of cytotoxicity and apoptosis. Apoptosis is a tightly regulated cell death program, which plays an essential role during the development and homeostasis of most organisms (Wyllie, 1997). This process can be triggered by two major signaling cascades, the extrinsic pathway (or death receptor pathway) and the intrinsic pathway (or mitochondrial pathway). The intrinsic pathway can be triggered by various endogenous signals and involves mitochondria as central integrators and coordinators of the apoptotic process and is characterized by mitochondrial membrane permeabilization (MMP), which results in the release of apoptogenic factors from the intermembrane space to the cytosol, such as cytochrome c (cyt c), and promotes caspases activation. Briefly, Bcl-2 family proteins, which includes antiapoptotic (Bcl-2, Bcl-xL, Bcl-w …) and proapoptotic members (Bax, Bad, Bak, Bik, Bcl-xS, …), and the permeability transition pore complex (PTPC) have been involved as crucial regulators of MMP (for review, see Kroemer et al., 2007). As the expression of the majority of these proteins is tissue specific, development specific, and/or disease specific, the mechanism of MMP is based on the activation of particular proapoptotic proteins and the inhibition of antiapoptotic ones. Thus, the identification of the precise actors of the apoptotic-signaling pathway is necessary for the characterization of toxin cytoxicity.
The aim of the present work was to study the implication of mitochondria, PTPC, and Bcl-2 family members in the apoptotic process induced by the two fusarial toxins, ZEN and T-2 toxin. Our study was performed on human cervix carcinoma cells (HeLa) overexpressing or not the antiapoptotic members of Bcl-2 family of proteins in order to look for possible similarities or differences in the cell response to our mycotoxins. This cell line constitutes a model for responding to both mycotoxins ZEN and T-2 toxin. Indeed, toxicokinetic studies concerning T-2 toxin have shown that this compound does not accumulate in any specific organ and may attack all the tissues in the organism (Swanson and Corley, 1989). Besides, ZEN possesses known estrogenic effects (Osweiler, 2000). By the combination of cellular and subcellular approaches, we found that the two mycotoxins might directly activate various intramitochondrial constitutive targets.
MATERIALS AND METHODS
Cell lines, cell culture, and chemicals.
HeLa, HeLa Bcl-2, HeLa Bcl-xL, and HeLa vMIA (human cervix carcinoma, generously given by V. Goldmacher, ImmunoGen, Cambridge, MA) and HCT116, HCT116-Bax-KO, HCT116-Bak-KO (human colorectal cancer, generously given by B. Vogelstein, Johns Hopkins University School of Medecine, Baltimore, MD) were grown as monolayer culture in Dulbecco’s modified Eagle’s medium:F12 supplemented with 100 μg/ml penicillin, 100 U/ml streptomycin, 1% Glutamax, and 10% fetal bovine serum (all from Invitrogen, Cergy-Pontoise, France), under 5% CO2/95% air. ZEN and T-2 toxin were obtained from Sigma Chemical Company (St Louis, MO) and were dissolved in pure ethanol. Calcein-AM (C-3100) and Hanks’ balanced salt solution were purchased from Invitrogen. Fluorescein diacetate (FDA) was from Polysciences (Lille, France). Pifithrin-α (PFT) was from Biomol (France). z-VAD-fmk (N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone) 10μM was from Bachem GmbH (Well am Rhein, Germany). 3,3′-Dihexyloxacarbocyaniniodode (DiOC(6)3), Hoechst 33258, and dihydroethidine (DHE) were from Molecular Probes (Cergy-Pontoise, France). For immunofluorescence assay, anti-Bax antibody (rabbit polyclonal, N20) was from Santa Cruz Biotechnology (Santa Cruz, CA), the anti-cyt c (mAb 6H2.B4) was from Pharmingen (BD Biosciences, San Jose, CA) and the antiapoptosis-inducing factor (AIF) (rabbit polyclonal) was from Chemicon (Hampshire, UK). Goat antimouse immunoglobulin G (IgG) conjugated with Alexa fluor 350 or goat anti-rabbit IgG Alexa fluor 488 were from Invitrogen. The kit CaspACE assay system colorimetric (G-7351) was purchased from Promega (Madison, WI). All others compounds were purchased from Sigma.
Effect of ZEN and T-2 toxin on HeLa cells viability.
For cell viability studies, cells were cultured in 24-well multidishes and treated with increasing concentrations of ZEN (2.5–60μM) and T-2 toxin (0.5–30nM) dissolved in the mixture ethanol:water (vol:vol) for 24 h. Then, they were incubated 5 min at 37°C with 0.2 μg/ml FDA, a nonfluorescent compound that becomes fluorescent when cleaved by esterases of living cells, as described by Rincheval et al. (2002). Controls were performed at the same time with ethanol:water (vol:vol). Propidium iodide (PI) was added just before analysis with XL3C flow cytometer (Beckman-Coulter, France). Corresponding IC50s for each mycotoxin was determined as the concentration inducing the loss of 50% cell viability.
Determination of nuclear apoptosis.
For the detection of nuclear apoptosis (sub-G1 population), cells were cultured in six-well multidishes and incubated with ZEN (30μM) and T-2 toxin (7.5nM) dissolved in the mixture ethanol:water (vol:vol) at 37°C. Controls were performed at the same time with ethanol:water (vol:vol). After 48 h of incubation, cells were harvested, fixed, and permeabilized with cold 70% ethanol, washed three times with PBS, and stained with 50 μg/ml PI in the presence of 250 μg/ml of RNAse A. The percentage of cell in sub-G1 was analyzed with XL3C flow cytometer (Beckman-Coulter).
Flow cytometry analysis of mitochondrial transmembrane potential and PTPC opening.
Cells (3 × 105 cells per well) were treated with ZEN (30μM) and T-2 toxin (7.5nM) dissolved in the mixture ethanol:water (vol:vol) for 6, 16, 24, and 48 h. A control was performed with ethanol:water (vol:vol). For inhibition experiments, cells were pretreated for 1 h with either ebselen (EBS) 10μM or PFT 50μM or z-VAD-fmk 10μM. For analysis of mitochondrial transmembrane potential (ΔΨm), Hela cells were stained with 100nM DiOC(6)3 for 15 min at 37°C, and the percentage of cells with low DiOC(6)3 fluorescence was determined by flow cytometry. Necrosis was estimated by adding 10 μg/ml of PI just before analysis. PTPC opening was assessed as previously described (Deniaud et al., 2008). Briefly, cells (2 × 105 cells/ml) were preincubated for 15 min at 37°C with 1μM calcein-AM (C-3100, Invitrogen) and 1mM CoCl2 in Hanks’ balanced salt solution (without phenol red and without sodium bicarbonate; Invitrogen) supplemented with 1mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.3. Hanks’ solution was then replaced by complete culture medium during apoptosis induction.
Isolation of mouse liver mitochondria and measurements of depolarization and swelling.
Mitochondria were isolated from mouse liver (C57BL/6J and Swiss, female, 6–12 weeks old, Charles River, France) by differential centrifugations and purified on Percoll gradient according to Jacotot et al. (2001). Protein concentration was determined using the micro-bicinchoninic acid assay (Pierce, Rockford, IL). All assays were performed in 96-well plates (200 μl) and were performed at 37°C in a spectrofluorimeter (TECAN genios; TECAN, Austria). For optical density (OD) measurement, transparent microplates were used, and for fluorescent measurement, black microplates were used. For swelling and depolarization measurements, mitochondria (25 μg of proteins) were diluted in a hypo-osmotic buffer (10mM Tris-Mops, pH 7.4, 5mM succinate, 200mM sucrose, 1mM Pi, 10μM ethylene glycol tetraacetic acid, 2μM rotenone) in the presence or in the absence of various concentrations of ZEN or T-2 toxin. The mitochondrial swelling was immediately measured by the decrease in OD at 540 nm for 100 min. Several pharmacological inhibitors, such as cyclosporin A (CsA), which interferes with the adenine nucleotide translocase (ANT)-cyclophilin D (Cyp-D) binding (Belzacq et al., 2003), bongkrekic acid (BA), an inhibitor of ANT, and 4,4′-diisothiocyanatostilbene-2,2′-disulphonic acid (DIDS), which targets voltage-dependent anion channel (VDAC) (Block et al., 1979), were added to mitochondrial suspension concomitantly with the mycotoxins. Similarly, the depolarization of the mitochondria was measured by the rhodamine 123 (1μM) fluorescence dequenching assay (λexc: 485 nm, λem: 535 nm; Molecular Probes) (Baker et al., 2004). When inhibitors are used, results are expressed in time of half-effect for T-2 toxin’s inhibition and in percentage of maximal effect for ZEN’s inhibition.
Detection of O2·− by DHE.
To measure the relative levels of cellular O2·−, the dye DHE (Molecular Probes) was used because O2·− oxidizes DHE to red-fluorescent ethidium (Eth) that can be detected by flow cytometry. Cells (2 × 105 cells/ml) were treated with ZEN (30μM) and T-2 (7.5nM) dissolved in the mixture ethanol:water (vol:vol) for 24 h. Then, the cells were centrifuged, washed with PBS, and incubated with DHE (5μM) for 15 min. O2·− level was immediately analyzed by flow cytometry with an excitation/emission wavelengths of 480 and 570 nm according to Peshavariya et al. (2007). The setting of flow cytometry was the same for control and treated samples.
To determine Bax, cyt c, and AIF localization, cells were seeded on slides in six-well multidishes, treated with ZEN (30μM) and T-2 toxin (7.5nM) dissolved in the mixture ethanol:water (vol:vol) for 24 h, washed with PBS, fixed with 3.7% paraformaldehyde for 10 min at room temperature (RT), and permeabilized 3 min at −20°C with cold acetone. After two washes with PBS, cells were saturated with PBS/bovine serum albumin (BSA) (3%) for 30 min, incubated for 1 h at RT with anti-Bax and anti-cyt c antibodies diluted in PBS/BSA 1%. After two washes, the secondary antibodies (goat antimouse IgG conjugated with Alexa fluor 350 or goat anti-rabbit IgG Alexa fluor 488) were added in PBS-BSA 1% for 45 min at RT. Nuclei were then stained with 2.5 μg/ml Hoechst 33348 for 5 min. Micrographs were taken on a LEICA fluorescence microscope (DMRH type, Leica, Rueil-Malmaison, France).
The cells were cultured (106 cells/ml) in 25-cm2 flasks for 24 h, in the absence or in the presence of mycotoxins (ZEN 30μM, T-2 toxin 7.5nM) at 37°C. Controls were performed at the same time with ethanol:water (vol:vol). Cells were harvested and centrifuged at 250 × g, and the pellet was incubated in ice-cold lysis buffer as previously mentioned for 10 min and then centrifuged at 250 × g for 20 min. Supernatants (cell extracts containing caspase-3) were retrieved and aliquots corresponding to 50 μg total protein, along with acetylated tetrapeptide (acetyl-Asp-Glu-Val-Asp) substrate labeled with the chromophore p-nitroaniline (pNA) in the presence of caspase-3 buffer, were added in a 96-well flat-bottomed microplate. In the presence of active caspase-3, cleavage and release of pNA from the substrate occurs. Free pNA produces a yellow color detected by a spectrophotometer at 405 nm. Additional controls, some free from cell lysates and others lacking substrate, were included. A standard curve was realized in order to determine the correspondence between OD and pNA concentration, then the results were expressed as caspase-3-specific activity (pmol pNA/h/μg protein) calculated as indicated by the manufacturers.
Determination of the Effect of Fusarial Toxins on HeLa Cells Viability
The effect of mycotoxins on HeLa cells viability was analyzed by FDA assay. Growing cultures of HeLa cells were exposed to increasing ZEN and T-2 toxin concentrations for 24 h. After incubation, results obtained for treated cells in the different assays were significantly different from the control (p < 0.05). Indeed, a concentration-dependent decrease of the percentage of viability was observed (Fig. 2). The highest concentrations of T-2 toxin induced more pronounced toxicity than ZEN, with a percentage of cells remaining viable of 6.1 ± 4.8% in the case of T-2 toxin and 23.65 ± 1.4% in the case of ZEN. The corresponding inhibitory concentrations that yield to 50% of growth inhibition (IC50s) were 30μM for ZEN and 7.5nM for T-2 toxin, demonstrating that even at a very low concentration, T-2 toxin induced a strong diminution of cell viability. The two mycotoxins thus induced cell death in HeLa cells, the most toxic compound being T-2 toxin, with maximal effects in the nanomolar range.
Determination of the Mode of Cell Death Induced by Mycotoxins
In order to determine the mode of cell death induced by mycotoxins, that is, apoptosis versus necrosis, flow cytometric experiments were performed. Condensation and fragmentation of the chromatin, considered as hallmarks of the apoptotic process (Ardestani et al., 2008), were determined by analyzing DNA content, the apoptotic cells corresponding to the hypoploid sub-G1 peak were detected after DNA staining. In contrast, necrosis is characterized by an early permeabilization of the plasma membrane. Therefore, necrosis was evaluated by labeling cells with the membrane-impermeant fluorochrome PI (10 μg/ml), which is excluded from viable and apoptotic cells but stained necrotic cells. HeLa cells were treated with ZEN (30μM) or T-2 toxin (7.5nM), and the percentage of apoptotic and necrotic cells was quantified (Fig. 3). The sub-G1 population significantly increased from 11.1 ± 1.55% in vehicle control–treated cells to 38.7 ± 2.33% and 63.3 ± 1.97% with ZEN and T-2 toxin, respectively. Whatever the toxin used, we observed only a low percentage of necrosis (PI+ cells), the maximum reaching about 5 ± 0.07% with the highest concentration of drugs. These results demonstrate that the cytoxicity of the two mycotoxins observed in Figure 2 is mediated by an apoptotic process rather than a necrotic process.
Mycotoxins Induce Opening of PTPC and Decrease of the ΔΨm
To evaluate the role of mitochondria in fusarial toxins–induced apoptosis, we investigated their ability to modify the ΔΨm. Cells were treated with ZEN (30μM) or T-2 toxin (7.5nM), for 6, 16, 24, and 48 h, and ΔΨm was assessed with the specific probe DiOC6(3). The drop in DiOC6(3) fluorescence, indicative of ΔΨm dissipation, was measured by flow cytometry. Our results demonstrated that both ZEN and T-2 toxin induced a time-dependent loss of ΔΨm (Fig. 4A). After 48 h of incubation, the percentage of cells with low ΔΨm was 88.2 ± 4.5% and 95.95 ± 0.49% for ZEN (30μM) and for T-2 toxin (7.5nM), respectively. Then, we used the calcein/cobalt assay to examine opening of PTPC and inner membrane (IM) permeabilization (Petronilli et al., 1999; Poncet et al., 2003). This method relies on the loading of cells with the fluorescent probe calcein and its quencher cobalt (Co2+). When loaded into cells in its acetoxymethyl ester form, calcein is liberated by the action of nonspecific esterases in all subcellular compartments including the mitochondrial matrix. Conversely, Co2+ is excluded from the mitochondrial matrix due to the IM impermeability to this ion. As a consequence, when the barrier provided by IM is functional, cells are fluorescent due to the presence of nonquenched calcein in the mitochondrial matrix. Upon PTPC opening, calcein is no longer entrapped in mitochondria, and thus, its fluorescence is quenched by Co2+ (Deniaud et al., 2008). As observed in Figure 4B, after 24 h of treatment, ZEN and T-2 toxin induced a loss of calcein fluorescence (calcein-) in about 50% of cells, indicating the opening of PTPC. Taken together, these results suggest that fusarotoxins trigger the mitochondrial pathway of apoptosis and that MMP can be associated with opening of PTPC.
T-2 Toxin and ZEN Can Directly Target Mitochondrial PTPC
Based on the fact that several toxins, which induce the signs of MMP in cellulo act directly on mitochondria, we wondered whether mycotoxins might have direct MMP-inducing effects. Mitochondria were therefore purified from C57BL/6J and Swiss mouse liver, incubated in the absence or in the presence of various concentrations of mycotoxins, and the kinetics of the loss of ΔΨm (increase in rhodamine 123 fluorescence; Figs. 5A and 5E) and mitochondrial matrix swelling (decrease in OD; Figs. 5B and 5F) were concomitantly recorded (Belzacq et al., 2003; Jacotot et al., 2001). Incubation with ZEN-induced depolarization and swelling of mitochondria. ZEN provoked a rapid and total depolarization of the inner mitochondrial membrane (Fig. 5A) in parallel with a concentration-dependent matrix swelling in mitochondria isolated from C57BL/6J mouse liver (Fig. 5B). In Swiss mice, both depolarization and swelling were dose dependent (Figs. 5E and 5F). However, T-2 toxin triggered a concentration-dependent loss of ΔΨm and matrix swelling for concentrations above 250nM (Figs. 5A and B) in C57BL/6J mouse liver. Only depolarization was observed in Swiss mice liver mitochondria (Fig. 5H).
These results indicate different sensitivities to our mycotoxins depending on the specie. Mitochondrial matrix swelling has been demonstrated to be a consequence of PTPC opening in many systems (Block et al., 1979). The involvement of PTPC in the induction of MMP by the mycotoxins was thus investigated using well-known pharmacological inhibitors of PTPC members: CsA, which interferes with the ANT-Cyp-D binding (Baker et al., 2004); BA, which inhibits ANT; and DIDS, which targets VDAC. The sigmoid dose-response curves obtained after T-2 toxin treatment in mitochondria from C57BL/6J mice allowed us to calculate the time of half-effect of this toxin in the presence of the different inhibitors, an increase in the time of half-effect corresponding to the inhibition of the mitochondrial depolarization or swelling (Fig. 5C). According to the kinetics of the ZEN effects on MMP, no time of half-effect could be calculated, and the data of the inhibitory experiments are thus presented as the percentage of ZEN effect alone (Fig. 5D). We observed that T-2 toxin- and ZEN-induced depolarization and swelling were greatly inhibited by CsA, indicating that PTPC is a target of these mycotoxins. The MMP provoked by T-2 toxin in mitochondria from C57BL/6J mice was delayed by incubation with the ANT inhibitor BA but not with DIDS. In contrast, mitochondrial events induced by ZEN were partially reduced by the VDAC inhibitor DIDS but not by BA in mitochondria from C57BL/6J and Swiss mice. Altogether, these results indicate that although T-2 toxin and ZEN might directly act on mitochondria to induce PTPC-dependent MMP, they appeared to target different members of PTPC, whereas ANT could be the target of T-2 toxin, ZEN seems to target VDAC.
Fusarial Toxins Trigger Relocalization of Bax and Release of Cyt c
To further characterize the mitochondrial events involved in fusarotoxins-induced apoptosis, we examined the mitochondrial relocalization of Bax and the release of cyt c by immunocytochemistry. After 24 h of exposure to ZEN (30μM) and T-2 toxin (7.5nM), Hela cells were fixed, permeabilized, and stained with fluorochrome-labeled Bax or cyt c antibodies and Hoechst 33258 to detect apoptotic nuclei. This dye allowed us to visualize cells with condensed of fragmented chromatin corresponding to apoptotic cells. As depicted in Figure 6, cells showing mycotoxin-induced DNA fragmentation or condensation (brighter nucleus than normal cells) also display a relocalization of the proapoptotic protein Bax to the mitochondria and the redistribution of the apoptogenic protein cyt c from mitochondria to the cytosol. Therefore, in response to fusarial toxins, Bax appears to favor the PTPC-dependent MMP, which leads to the release of cyt c.
Fusarial Toxins Induce Caspase-Dependent Apoptotic Pathway
To determine whether the cytotoxic activities of mycotoxins rely on a caspase-dependent process, HeLa cells were incubated with ZEN or T-2 toxin for 24 h, the whole-cell lysates were prepared and the activity of caspase-3 was investigated. As shown in Figure 7A, an increase in caspase-3 activity was found in HeLa cells as compared with the vehicle control–treated cells in response to both ZEN and T-2 toxin. This activity reached 36 pmol pNA/h/μg of protein at 30μM of ZEN and to 49 pmol pNA/h/μg of protein at 7.5nM of T-2 toxin. Results are significantly different when compared with the vehicle-treated control (p < 0.05). Thus, the fusarotoxin-induced MMP is followed by caspase-3 activation. HeLa cells were also pretreated for 1 h with 50μM of z-VAD-fmk, a pan-caspase inhibitor, and the ΔΨm was recorded (Fig. 7B). The given histograms (upper panel) are those obtained from a representative experiment, whereas the bar graph (lower panel) represents the mean ± SD of three distinct experiments. These data demonstrate that inhibition of caspases greatly reduced the mitochondrial alterations and suggest that a caspase could act upstream of mitochondria to trigger or amplify the apoptotic process.
Induction of Reactive Oxygen Species Production by ZEN and T-2 Toxin
Reactive oxygen species (ROS) produced by oxidative stress have been implicated in apoptosis as possible signaling molecules (Chen and Yan, 2005). The level O2·− was therefore analyzed using DHE, a fluorescent probe reported to be relatively specific for this ROS (Bindokas et al., 1996; Walrand et al., 2003). Indeed, when DHE is oxidized by O2·−, it originates Eth, a fluorescent compound, that stains nucleic acid. Incubation with mycotoxins resulted in an oxidization of the DHE probe, and we observed an induction of ∼17-fold of control in the presence of ZEN and 11-fold of control in presence of T-2 toxin (Fig. 8A). Thus, the two fusarial toxins strongly increase the level of O2·− in HeLa cells. ROS have been demonstrated to be a cause or a consequence of PTPC opening and mitochondrial alterations in apoptosis (reviewed in Le Bras et al., 2005). In order to determine whether ROS production occurs upstream or downstream of mitochondrial alterations, we analyzed the effect of the antioxidant EBS on mycotoxins-induced loss of ΔΨm. As shown in Figure 8B, pretreatment of cells with 10μM EBS totally prevented the dissipation of ΔΨm triggered by ZEN but was not very efficient in the case of T-2 toxin, suggesting that ROS are produced upstream of mitochondria in response to ZEN, whereas their production could be a consequence of mitochondrial alterations in the apoptotic process induced by T-2 toxin.
ZEN, but not T-2 Toxin, Induces p53-Dependent Apoptosis
Mycotoxins were demonstrated to be genotoxic and to induce DNA-adduct formation (Lioi et al., 2004; Pfohl-Leszkowicz et al., 1995). Because p53 has been clearly implicated as a central initiator of the apoptotic process in response to genotoxic stress, in particular by regulating the expression of genes of the Bcl-2 family that are directly involved in the initiation of mitochondria-induced apoptosis (Wu and Deng, 2002), we next investigated whether ZEN- and T-2 toxin–induced mitochondrial alterations depends upon p53 activity. To this aim, we examined the effect of PFT, an inhibitor of p53 transcriptional activities, on ZEN- and T-2 toxin–induced mitochondrial alterations. HeLa cells were pretreated 1 h with 50μM of PFT, incubated for 24 h in the presence of mycotoxins, and mitochondrial depolarization was analyzed by flow cytometry. PFT totally prevents the dissipation of ΔΨm provoked by ZEN, whereas no significant protection was observed in the presence of T-2 toxin (Fig. 9). Our results suggested that in HeLa cells, the mitochondriotoxic activity of ZEN implicates a p53-dependent signaling pathway, whereas T-2 toxin induced a p53-independent apoptosis.
The MMP Triggered by ZEN and T-2 Toxin Is Regulated by Bcl-2 Family Members
To determine the specific contribution of Bcl-2 family members to the MMP process induced by fusarial toxins, various cell lines, which have been genetically modified, were used. The role of antiapoptotic protein was determined in HeLa cells stably expressing Bcl-2 and Bcl-xL (HeLa Bcl-2 and HeLa Bcl-xL), whereas the role of the proapoptotic members was studied using HeLa cells expressing the viral protein vMIA (HeLa vMIA), which exerts its antiapoptotic function by neutralizing Bax (Poncet et al., 2006), and HCT116 cells genetically invalidated for Bax or Bak (HCT116-Bax-KO, HCT116-Bak-KO). These cells were treated with ZEN and T-2 toxin for 24 h and the transmembrane inner potential was analyzed (Fig. 10). In comparison to parental cells (HeLa), cells overexpressing Bcl-xL were greatly protected against dissipation of ΔΨm, the percentage of DIOC− cells being reduced from 93.67 ± 4.17% in HeLa to 53 ± 3.32% in HeLa Bcl-xL for ZEN and from 79.2 ± 0.28% to 33.5 ± 2.40% for T-2 toxin. Conversely, the enforced expression of Bcl-2 did not significantly prevent the mitochondrial alterations induced by ZEN and prevented those induced by T-2 toxin. The invalidation (HCT116-Bax-KO) or the neutralization (HeLa vMIA) of Bax strongly inhibited the mitochondrial depolarization, demonstrating the important role of this proapoptotic protein in the mitochondriotoxic activity of ZEN and T-2 toxin. In contrast to Bax, we observed a difference between ZEN and T-2 toxin concerning the role of Bak. Indeed, although Bak invalidation did not significantly modify the percentage of DIOC-cells in the presence of T-2 toxin, it provided protection to a similar extent as Bax invalidation against the loss of ΔΨm provoked by ZEN (29.9 ± 1.83% vs. 60.25 ± 2.61% in HCT116 cells). Our results therefore demonstrate the role of Bcl-xL and Bax in the regulation of mitochondrial alterations induced by the fusarial toxins, and point to Bak as a mediator of ZEN, but not T-2 toxin, cytotoxic activity.
In order to gain further insight into the mechanism of fusarotoxin-mediated cytotoxicity, we compared the molecular events implicated in cell death triggered by ZEN and T-2 toxin. Our results showed a concentration-dependent cytotoxic effect of ZEN and T-2 toxin as revealed by the FDA assay (Fig. 2). We demonstrated that this cell death process is accompanied by condensation and fragmentation of nuclear DNA, caspase activation but no early plasma membrane permeabilization and therefore corresponds to apoptosis rather than necrosis (Figs. 3, 6, and 7). In addition, we observed an opening of PTPC, a dissipation of ΔΨm, a production of ROS, and the release of cyt c, clearly establishing the central role of mitochondria in the cytotoxic activity of the two fusarial toxins.
Both ZEN- and T-2 toxin–induced MMP, a phenomenon that implies the formation of pores or channels, which cause the dissipation of the ΔΨm built across the mitochondrial IM (for review, see Kroemer et al., 2007). The induction of the permeability of the IM to calcein by ZEN and T-2 toxin demonstrates the involvement of PTPC in response to fusarial toxins (Fig. 4B). Opening of this pore is known to trigger the permeability transition, a sudden increase of the IM permeability to solutes with molecular mass up to 1.5 kDa. Following IM permeabilization, an increase in mitochondrial matrix volume occurs as a consequence of the massive entry of solutes and water. This matrix swelling gives rise to a distension and disorganization of the cristae and leads to disruption of the outer membrane (OM) (Petit et al., 1998; Scorrano et al., 2002) and to the release of proteins from the intermembrane space to the cytosol (Scarlett and Murphy, 1997). The exact molecular nature of the PTPC is still a matter of debate and actually may depend on the experimental system and proapoptotic stimuli, although an emerging consensus considers that this polyprotein complex is based on the dynamic interaction between VDAC (the most abundant protein of the OM), ANT (the most abundant protein of the IM), and Cyp-D (in the matrix) (Zoratti and Szabo, 1994, 1995 and for recent reviews, see Brenner and Grimm, 2006). The opening of VDAC is a regulated process, and it was shown that this channel, firstly considered as a low specific pore, may exhibit some degree of specificity in the mitochondrial import/export of molecules (e.g., ATP, Ca2+, and other ions). It was also hypothesized that the interaction of VDAC with proapoptotic Bcl-2 family members provokes the closure of the channel and thus could modulate OM permeabilization (Kroemer et al., 2007). ANT belongs to a large family of structurally related proteins, namely, the family of mitochondrial carriers (Wohlrab, 2005). Some of these proteins share the capacity to convert into nonspecific pores. Thus, ANT has been proposed to be a major player of IM permeabilization during apoptosis because it can switch from a vital function (stoichiometric ADP/ATP exchange through IM) to a lethal one under activation by a variety of molecules such as atractyloside, ROS, chemotherapeutic agents, and viral proteins and peptides (for reviews, see Belzacq and Brenner, 2003; Halestrap, 2004). Because opening of PTPC can occur after the direct interaction of drugs with a member of this pore (Rotem et al., 2005; Vieira et al., 2001) or can result from the activation of an upstream pathway (Deniaud et al., 2008), we tested whether ZEN and T-2 toxin directly induced alterations on isolated mitochondria. Our results indicate that both ZEN and T-2 toxin act directly on mitochondria to induce PTPC-dependent MMP. The sensitivity to T-2 toxin depends on the tested specie. We have demonstrated a loss of ΔΨm in mitochondria isolated from Swiss mice liver less evident than the depolarization observed in mitochondria isolated from liver of C57BL/6 mice. Indeed, other studies have demonstrated that sensitivities to chemicals are highly strain dependent. For instance, Schauwecker (2002) has demonstrated that C57BL/6 and FVB/N mice exhibit different susceptibilities to excitotoxin-induced cell death. Moreover, using specific inhibitors of VDAC, ANT, and Cyp-D, we demonstrated that despite the direct effect of fusarotoxins on mitochondria, they target different members of PTPC, whereas ZEN seems to target VDAC, ANT might constitute a target of T-2 toxin.
Many reports have demonstrated that the caspase family plays an important role in apoptosis. These proteases are unique in their requirement for an Asp residue at the cleavage site in their substrates. Among the 14 known members of the caspase family of proteases, caspase-3 plays a critical role in the execution of the apoptotic process (Chen et al., 2001). In the mitochondrial pathway of apoptosis, this caspase is activated downstream of mitochondria by a multiprotein complex called apoptosome including Apaf-1, caspase-9, dATP, and cyt c (Zou et al., 1999). In our system, we observed that the PTPC-dependent MMP triggered by ZEN and T-2 toxins leads to the release of cyt c from the intermembrane space to the cytosol and subsequent caspase-3 activation (Figs. 6 and 7), indicating the central role of caspase-3 in the cytotoxicity of the fusarotoxins. The dependence of caspases activation on mycotoxins-triggered apoptosis was also evidenced by the complete abrogation of fusarotoxin-induced mitochondrial apoptosis in cells pretreated with the broad-range peptide inhibitor of caspases, zVAD-fmk. This latter observation suggests that a caspase-activated upstream of mitochondria could favor MMP or that z-VAD-fmk inhibits a caspase-3–mediated feedback amplification loop that amplifies MMP and mitochondrial alterations, as proposed previously (Chen et al., 2000).
Differences in the molecular mechanisms responsible for the toxicity of the two mycotoxins were highlighted by our study of the effect of the pharmacological inhibitors of p53 and ROS, that is, PFT and EBS, respectively. Indeed, we observed that inhibition of p53 transcriptional activities by PFT prevented ZEN-induced MMP, whereas PFT was devoid of effect in the presence of T-2 toxin (Figs. 8 and 9). Therefore, in addition to its ability to directly activate PTPC, the genotoxic stress induced by ZEN (Lioi et al., 2004; Pfohl-Leszkowicz et al., 1995) could activate p53, which in turn transactivates apoptotic response genes, such as bax, puma, or noxa, whose products trigger the mitochondrial pathway of apoptosis (Wu and Deng, 2002). Conversely, the absence of PFT effect in the presence of T-2 toxin indicates that this fusarotoxin promotes a p53-independent apoptotic process. Another difference between the two toxins was observed in the presence of the antioxidant EBS. Although the two mycotoxins induced O2·− production, MMP was prevented by EBS in the presence of ZEN but not in the presence of T-2 toxin. Based on previous reports demonstrating that ROS can be a cause or a consequence of the mitochondrial alterations (Le Bras et al., 2005), our data indicate that ROS are generated upstream of mitochondrial alterations in response to ZEN, suggesting that this mycotoxin induce an oxidative stress that causes MMP and subsequent apoptosis, whereas in the case of T-2 toxin, ROS production occurs as a consequence of the PTPC-mediated mitochondrial alterations.
In cells overexpressing antiapoptotic proteins of the Bcl-2 family, Bcl-xL and to a lesser extent Bcl-2 prevents the mitochondrial alterations triggered by ZEN or T-2 toxins (Fig. 10). Such a specificity of Bcl-xL (vs. Bcl-2) protective activity has been previously reported in MDA-MB-231 as well as in HeLa cells in response to TRAIL (TNF-related apoptosis-inducing ligand)–induced cell death (Kim et al., 2003). In addition, we showed that treatment of cells with mycotoxins stimulates the relocalization of Bax from the cytosol to mitochondria (Fig. 6) and that knockout (HCT116 Bax−/− cells) or neutralization (HeLa vMIA cells; Poncet et al., 2006) of this proapoptotic protein impedes the ΔΨm loss induced by fusarotoxins (Fig. 10). These results indicate that Bax favors the apoptotic process and thus is critical for the cytotoxic activities of ZEN and T-2 toxin. Conversely, the involvement of the proapoptotic protein Bak depends on the toxin used. Indeed, Bak deficiency impairs mitochondrial alterations triggered by ZEN but was devoid of effect in response to T-2 toxin. Although the requirement for Bax, but not for Bak, in response to T-2 toxin needs further investigations, these data highlights another difference in the molecular mechanism of the mitochondriotoxic activities of ZEN and T-2 toxin.
The observed toxic effects induced by ZEN might be attributed to its structural similarity with other estrogenic compounds inducing diverse toxic effects such as mitochondrial apoptotic cell death (Klinge, 2008; Ouanes et al., 2005; Stopper et al., 2005). In isolated liver mitochondria of female rats, estradiol has been shown to induce deleterious toxic effects leading to mitochondrial impairment (Moreira et al., 2007). In another hand, published data have demonstrated that trichothecenes, to which belongs T-2 toxin, have mitochondriotoxic activities. Acetoxyscirpenol toxins such as 15-acetoxyscirpenol, 4,15-diacetoxyscirpenol, and 3alpha-acetyldiacetoxyscirpenol are trichothecenes, which have been shown to induce mitochondrial apoptotic cell death in jurkat T cells involving disturbance of the ΔΨm (Lee et al., 2006).
In conclusion, our results provide for the first time important mechanistic insights in the signaling pathways of the cell death process induced by commonly encountered fusarial toxins (Fig. 11). We demonstrated that the toxicity of ZEN and T-2 toxins is mediated by a PTPC-dependent activation of the mitochondrial pathway of apoptosis, which is regulated by certain Bcl-2 family members. By using a cell-free system of purified mitochondria, we demonstrated that PTPC opening can occur independently of any other subcellular component and that VDAC and ANT might constitute potential targets of ZEN and T-2 toxin, respectively. Finally, although the two mycotoxins share some common events of the mitochondrial pathway of apoptosis, p53, ROS, and Bak were shown to play different role according to the toxin used.
“Ministère Tunisien de l'enseignement Supérieur, de la Recherche Scientifique et de la Technologie through the Laboratoire de Recherche sur les Substances Biologiquement Compatibles: LRSBC” and “La Direction Générale de la Recherche Scientifique et de la Rénovation Technologique: DGRST et le Centre National de la Recherche Scientique: CNRS (Action d'Echanges Tuniso-Française: DGRST/CNRS 04/R 0803)”.