Induction of Oxidative Stress Response by the Mycotoxin ... - CiteSeerX

TOXICOLOGICAL SCIENCES 95(2), 340–347 (2007) doi:10.1093/toxsci/kfl156 Advance Access publication November 7, 2006

Induction of Oxidative Stress Response by the Mycotoxin Patulin in Mammalian Cells Biing-Hui Liu,*,1 Ting-Shuan Wu,* Feng-Yih Yu,* and Ching-Chyuan Su† *Department of Biomedical Sciences, Chung Shan Medical University, Taichung, Taiwan, ROC; and †Tian-Sheng Memorial Hospital, Tong Kong, Ping-Tong, Taiwan, ROC Received August 5, 2006; accepted October 17, 2006

Patulin (PAT), a mycotoxin mainly produced by Penicillium and Aspergillus, is found in various foods and feeds. In the present study, its effects on oxidative stress in various mammalian cell lines were investigated. When cell-permeating fluorescent dyes were used as indicators of the generation of reactive oxygen species (ROS), we found that PAT treatment directly increased intracellular oxidative stress in human embryonic kidney (HEK293) and human promyelocytic leukemia (HL-60) cells. Lipid peroxidation levels were also significantly increased in HL-60 cells and mouse kidney homogenates treated with PAT. Suppression of CuZn– superoxide dismutase (SOD) expression in mammalian cells by small interfering RNA resulted in an increase in PAT-mediated membrane damage, while overexpression of human CuZn-SOD or catalase led to a reduction in damage, indicating the involvement of ROS in PAT toxicity. Pretreatment of HEK293 cells with Tiron, a free radical scavenger, reduced the phosphorylation levels of extracellular signal–regulated kinase (ERK) 1/2 elicited by PAT. The ERK1/2 signaling pathway inhibitor, U0126, also significantly decreased the levels of ROS associated with PAT treatment. These findings indicate that PAT treatment results in the ROS production in mammalian cells, and ROS partially contributes to PAT-induced cytotoxicity. Activation of ERK1/2 signaling pathway is correlated with PAT-mediated ROS. Key Words: patulin; reactive oxygen species; siRNA; CuZn-SOD; catalase; ERK1/2 signaling; mycotoxin.

The mycotoxin patulin (PAT), a secondary metabolite of fungal species, including Penicillium and Aspergillus, is a deleterious contaminant of certain feeds and foods, especially apple juice and related products (Doores, 1983). A safety level of 50 lg/l of PAT in apple juice was established by the World Health Organization and is followed by some countries (Van Egmond, 1989). PAT-induced nephropathy and gastrointestinal tract malfunction have been demonstrated in several animal models (Mahfoud et al., 2002; McKinley et al., 1982). In 1

To whom correspondence should be addressed at Department of Biomedical Sciences, Chung Shan Medical University, No. 110, Sec. 1, Chien-Kuo N. Road, Taichung, 402 Taiwan, ROC. Fax: þ886 4-24757412. E-mail: [email protected].

addition, the toxin is considered to be a clastogen, mutagen, and teratogen (Liu et al., 2003; Smith et al., 1993). In terms of biochemical mechanism, the toxic effects of PAT on various cells are closely associated with its activity on SH groups (Liu et al., 2006; Mahfoud et al., 2002). Barhoumi and Burghardt (1996) demonstrated that glutathione depletion by PAT results in the generation of reactive oxygen species (ROS) in a rat hepatocyte cell line; however, the link between ROS perturbation and PAT-induced cellular changes has not been fully examined. Aerobic organisms, which use oxygen in the energygenerating process, are susceptible to damage caused by small amounts of superoxide anion, hydroxyl radical, hydrogen peroxide (H2O2), and unstable intermediates of lipid peroxidation (Henle and Linn, 1997). These ROS are also generated by sources including ionizing radiation (Leach et al., 2001; Narayanan et al., 1997) and toxic chemicals and drugs (Srinivasan et al., 2001). For their defense, mammalian cells possess a family of antioxidant enzymes, superoxide dismutases (SOD), that convert harmful superoxide anion to H2O2, which, in turn, is metabolized to water and oxygen by catalase and glutathione peroxidase (Fridovich, 1986). Free radicals and other ROS generated by an imbalance between the radicalgenerating and -scavenging systems lead to oxidative stress, which directly causes cell changes during aging, transformation, and differentiation. Furthermore, it is hypothesized that ROS serve as subcellular messengers in gene regulation and signal transduction pathways (Allen and Tresini, 2000). Both the mitogen-activated protein kinase (MAPK) and nuclear factor jB signal transduction pathways are considered to be redox sensitive (Cakir and Ballinger, 2005; Milligan et al., 1998). We have previously found that treatment of mammalian cell cultures with PAT leads to the activation of MAPKs, including extracellular signal–regulated kinase (ERK) 1/2, p38 kinase, and c-jun N-terminal kinase (JNK) (Liu et al., 2006). Phosphorylation of ERK1/2 is a major factor contributing to PATinduced genotoxicity (Wu et al., 2005). In the present study, we demonstrated that PAT induces ROS generation in several mammalian cells and that overexpression of CuZn-SOD or catalase partially blocks PAT-induced cellular membrane

The Author 2006. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For Permissions, please email: [email protected]

GENERATION OF ROS BY PATULIN

damage. In addition, to understand whether ROS plays a role in signaling transduction, we also investigated the relationship between PAT-induced ROS and the PAT-activated ERK1/2 pathway in human cells. MATERIALS AND METHODS Reagents. Cell culture media and serum were obtained from Life Technologies (Grand Island, NY). U0126 and polyclonal rabbit antibodies against phospho-ERK1/2 (Thr202/Tyr204) and ERK1/2 were purchased from Cell Signaling (Beverly, MA). Polyclonal antibodies specific to CuZn-SOD and catalase were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and Calbiochem (La Jolla, CA), respectively. 2#, 7#-dichlorodihydro-fluorescein diacetate (H2DCF-DA) and dihydroethidium were obtained from Molecular Probes (Eugene, OR). PAT (4-hydroxy-4H-furo[3,2-c]pyran- 2(6H)-one) and all other reagents were from Sigma Chemical (St. Louis, MO). PAT was dissolved at a concentration of 10mM in 15% ethanol and stored at 20C. Cell culture. The human embryonic kidney (HEK293), human promyelocytic leukemia (HL-60), and Chinese hamster ovary (CHO-K1) cell lines were obtained from Bioresources Collection and Research Center in Taiwan. HEK293 cells were cultured in minimum essential medium supplemented with 10% horse serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin at 37C in a humidified 5% CO2 incubator. HL-60 cells and CHO-K1 were maintained in RPMI 1640 and Ham’s F12 medium, respectively, and supplemented with 10% fetal bovine serum and antibiotics as described above. Measurement of ROS in cells with oxidation-sensitive fluorescent dyes. HEK293 cells (104/per well on a 96-well tissue culture plate) were cultured for at least 24 h in 50 ll of complete medium, and then 50 ll of H2DCFDA (20lM) in Krebs-Ringer N-2-hydroxyethylpiperazine-N#-2-ethanesulfonic acid (KRH) buffer was added. Thirty minutes after incubation, the probecontaining medium was removed and various concentrations of PAT or H2O2 in KRH buffer were added and incubated for another 45 min at 37C. Since H2DCF-DA is hydrolyzed and oxidized by intracellular ROS to 2,7-dichlorofluorescein (DCF) (Rothe and Valet, 1990), the fluorescence of DCF was analyzed in an HTS 7000 Bio Assay Fluorescent Plate Reader (PerkinElmer life Sciences, Wellesley, MA) at an excitation wavelength of 485 nm and emission at 530 nm. To evaluate the ROS levels in HL-60 cultures, cells in complete medium were incubated with H2DCF-DA at a final concentration of 10lM for 30 min. After centrifugation at 100 3 g for 5 min, the medium of cultures were replaced with either solvent or 100lM PAT in phosphate-buffered saline (PBS) solution and then incubated at 37C for another 45 min before flow cytometry. Intracellular superoxide anion production was measured using the cellpermeable dye dihydroethidium, which binds to nuclear DNA when oxidized by superoxide anion and emits red fluorescence (Bindokas et al., 1996). HL-60 cells (106/per well on a 24-well tissue culture plate) in complete medium were treated with vehicle or PAT (25–100lM) for 100 min. After centrifugation at room temperature, the supernatant fluid was replaced with 2lM dihydroethidium in PBS solution and further incubated for 15 min at 37C before flow cytometry. The oxidized forms of all ROS-sensitive dyes could be excited with the 488nm laser of the Becton-Dickinson FACSalibur flow cytometer (Franklin Lakes, NJ). Emission of DCF fluorescence was detected in channel FL-1 (530-nm filter and 30-nm bandpass); emission of ethidium fluorescence after dihydroethidium oxidation was quantified in channel FL-2 (585-nm filter and 42-nm bandpass). Determination of thiobarbituric acid–reactive substances. The formation of lipid peroxidation products was evaluated as thiobarbituric acid–reactive substances (TBARS) (Janero, 1990). The mouse kidneys were minced in 10% (v/w) of cold 10mM PBS with a homogenizer. An aliquot of this homogenate was diluted 1:10 with PBS and centrifuged at 3000 3 g for 10 min; supernatants were collected for the following PAT treatment. The HL-60 cells or kidney homogenates were treated with various concentrations of PAT for 90 min at 37C and then centrifuged at 5000 3 g for 5 min. Two hundred microliters of

341

supernatant fluid was transferred to a sample tube and 250 ll of 10% trichloroacetic acid and 125 ll of 0.7% thiobarbituric acid were promptly added. To minimize peroxidation during the subsequent assay procedure, 10 ll of 4% butylated hydroxytoluene was also added to the thiobarbituric acid reagent mixture. The mixture was heated to 90C for 15 min, cooled on ice, and centrifuged at 5000 3 g for 5 min. The supernatant containing TBARS was collected for fluorescence analysis with excitation wavelength set at 535 nm and emission at 585 nm (HTS 7000, PerkinElmer life Sciences). The fluorescence intensity of TBARS was first converted to malondialdehyde (MDA) equivalents and then normalized to the cell protein content. MDA standards were prepared from 1,1,3,3-tetraethoxypropane. Plasmid construction. The plasmid expressing CuZn-SOD small interfering RNA (siRNA) was constructed by cloning the sequence 5#-GCAGATGACTTGGGCAAAG-3# into pSilencer2.1-U6 neo expression vector (Ambion, Austin, TX). This sequence was selected based on the report of Maxwell et al. (2004). The oligonucleotides containing the sense and antisense strands connected with a loop TTCAAGAGA were chemically synthesized, annealed, and inserted downstream of human U6 RNA pol III promoter. The plasmids for overexpression of human CuZn-SOD and catalase cDNAs were constructed using pcDNA 3.1 (þ) (Invitrogen, Carlsbad, CA) as the backbone plasmid. Both cDNAs were cloned by reverse transcription–polymerase chain reaction amplification and then was ligated into vector pcDNA 3.1. Briefly, total RNA was isolated from HEK293 cells and reversely transcribed into cDNA, which was then used as template for polymerase chain reaction. The sense and antisense primers used for CuZn-SOD amplification were 5#-ATTAAGCTTTAGCGAGTTATGGCGACGAA-3# and 5#-TTGAATTCTTATTGGGCGATCCCAATTAC-3#, respectively. The sense and antisense primers used for catalase amplification were 5#-ATAGGTA CCAAACCGCACGCTATGGCTGA-3# and 5#-ATATCTAGAA TCCAGTGATGAGCGGGTTA-3#, respectively. All the above constructed plasmids were sequence verified. Transfections. Transfections were performed with Lipofectamine 2000 (Invitrogen) according to the manufacturer’s recommendations. HEK293 and CHO-K1 cells, grown in 3.5-cm culture plates with serum-free medium, were at 80–90% confluence at the time of transfection. Cells were transfected with 2 lg of constructed plasmid or empty vector (as control) and incubated in the CO2 incubator for 16 h prior to replacement with fresh medium containing 10% serum and antibiotics. Twenty-four hours after medium replacement, the transfected cells were subcultured and selected with G418 (500 lg/ml) for 2–3 weeks till the formation of colonies. Stable cell lines were established by picking up single colonies and continuously cultured for the following experiments. The transfection efficiency of siRNA plasmid into HEK293 cells was estimated by cotransfection of pGFP-C1 (Clontech, Palo Alto, CA) and found to be > 90%. Measurement of extracellular lactate dehydrogenase activity. Lactate dehydrogenase (LDH) released into the medium was assayed using a LDH Cytotoxicity Detection Kit (TAKARA BIO Inc., Japan). Transfected HEK293 or CHO-K1 cells were seeded at 5 3 103 cells/well in quadruplicate in 96-well tissue culture plates and allowed to attach for at least 18 h to obtain monolayer cultures. After replacing the medium with F12 medium, the vehicle (15% ethanol in PBS) alone or various concentrations of PAT (final concentration 20–100lM) were promptly added and incubated for 4 h at 37C. An aliquot (100 ll) of cell-free medium from each well was removed for LDH assay according to the manufacturer’s protocol. Western blot analysis. For preparation of whole-cell extracts, the cultures were rinsed with 0.01M PBS and lysed by addition of extraction buffer (PBS containing 5% glycerol, 1mM dithiothreitol, 1mM ethylenediaminetetraacetic acid, pH 8.0, 0.5% Triton X-100, 0.8lM aprotinin, 1mM [4-(2-aminoethylbenzensulfonyl fluoride hydrochloride)], 20lM leupeptin, 40lM bestatin, 15lM pepstain A, 14mM E-64, and 1mM phenylmethylsulfonyl fluoride). The cell lysate was kept on ice for 10 min and then centrifuged at 16,000 3 g for 20 min at 4C. The protein concentration of the supernatant solution was determined using the Bradford protein assay (Bio-Rad, Hercules, CA) with bovine serum albumin as the standard.

342

LIU ET AL.

To study the effects of Tiron (1,2-dihydroxybenzene-3,5-disulfonic acid disodium salt) or U0126 (1,4-diamino-2,3-dicyano-1,4-bis(o-aminophenylmercapto) butadiene) on ERK1/2 activation, HEK293 cells at 80% confluence were pretreated with Tiron for 2 h or with U0126 for 30 min before coexposure to PAT. Preparation of whole-cell extracts was conducted as described above. Equal amounts of proteins (40 lg) from each sample preparation were subjected to Western blotting according to Wu et al. (2005). Polyclonal antibodies against phospho-ERK1/2 (Thr202/Tyr204), ERK1/2, CuZn-SOD, and catalase and monoclonal antibodies against tubulin were used as the probes. Bound antibody on the membrane was detected using an enhanced chemiluminescence detection system according to the manufacturer’s manual (Amersham Pharmacia Biotech, Amersham, United Kingdom). Statistical analysis of data. All statistical analyses were carried out using the software program GraphPad Prism Version 4.0 (GraphPad Software, San Diego, CA). Experimental data grouped by one variable were analyzed by unpaired two-tailed t-test or one-way ANOVA followed by Tukey post test. Experiments with two variables were analyzed by two-way ANOVA in combination with Tukey post test. A p value < 0.05 was considered significant.

RESULTS

Intracellular ROS Generation Increases after PAT Treatment Fluorescence spectrophotometry and cytometry using H2DCF-DA as the probe were used, respectively, to measure intracellular ROS production in HEK293 and HL-60 cells. Intracellular H2DCF is oxidized to fluorescent DCF by several oxidants, including H2O2, superoxide, hydroxyl radicals, and cellular peroxidases (Hempel et al., 1999). After treatment of HEK293 cells for 45 min with various concentrations of PAT, intracellular DCF fluorescence increased in a dose-dependent manner (Fig. 1A). PAT at 50 and 100lM caused a marked increase of fluorescence to 1.9-fold and 2.6-fold of control levels, respectively. Hydrogen peroxide at 100lM worked as a positive control. However, HEK293 cells treated with 100lM PAT in the presence of cysteine (200 or 500lM) or treated with 100lM PAT-cysteine adduct showed no significant increase of DCF fluorescence compared with solvent-treated control (data not shown). To evaluate whether the ROS-inducing ability of PAT is celltype specific, HL-60 cells were also exposed to PAT and then examined by fluorocytometry. Incubation of HL-60 cells with 100lM PAT led to a 7.4-fold increase in fluorescence compared with solvent-treated cells (Fig. 1B). To further understand the type of ROS induced by PAT, dihydroethidium, a dye generally used as probe for superoxide anions, was applied in HL-60 cells, and a dose-dependent increase in ethidium fluorescence was seen in PAT-treated cultures (Fig. 1C). Exposure of HL-60 cells to 100lM PAT resulted in a significant 2.3-fold increase in superoxide anion production compared with controls. These data indicate that PAT treatment is able to generate ROS, including superoxide anion, in different human cell lines. Induction of Lipid Peroxidation by PAT Previous evidence suggests that the endogenously produced ROS are, at least in part, responsible for the formation of

FIG. 1. PAT increases intracellular ROS levels in HEK293 and HL-60 cells. HEK293 cells (A) or HL-60 cells (B) were incubated with 10lM H2DCFDA for 30 min, and then the medium was replaced with various concentrations of PAT or 100lM H2O2 in PBS for another 45 min. For the HEK293 cultures, DCF fluorescence was analyzed by spectrofluorometry, and the fluorescence in HL-60 cells was measured by fluorocytometry. In the chromatograph in (B), the open and solid areas represent the cultures treated with 100lM PAT and vehicle, respectively. (C) HL-60 cells were treated with vehicle or PAT for 100 min to evaluate the intracellular superoxide anion levels. The treated cells were incubated with 2lM dihydroethidium for 15 min and then ethidium fluorescence was determined using a fluorocytometer as described in Materials and Methods. All the data are the mean ± SEM for five independent experiments. Significant difference (*p < 0.05; **p < 0.01; ***p < 0.001) compared with the vehicletreated group (0lM).


343

TABLE 1 Induction of Lipid Peroxidation by PAT in HL-60 Cells and Mouse Kidney Homogenates MDAb (nmole/mg protein)

Treatmenta PAT (lM) 50 100 200 H2O2c (lM) 100 200

HL-60 cells

Kidney homogenates

— 0.66 ± 0.08 1.04 ± 0.31

0.76 ± 0.10 1.38 ± 0.22 2.72 ± 0.24

— —

1.01 ± 0.14 2.22 ± 0.44

a

HL-60 cells or mouse kidney homogenates were treated with various concentrations of PAT or H2O2 for 90 min, then subjected to TBARS assay. b The formation of lipid peroxidation products was evaluated as TBARS, and the fluorescence intensity of TBARS was converted to MDA equivalents. The data are mean ± SEM for at least four independent experiments. c H2O2 was used as a positive control.

oxidized lipids/lipoproteins (Bartsch and Nair, 2004). Therefore, we examined the lipid peroxidation in cultured cells and tissue homogenates by measuring the production of the lipid peroxidation end product, MDA, using the TBARS assay. In HL-60 cultures treated for 90 min with 100 or 200lM PAT, MDA levels (nmol/mg protein) were 0.66 ± 0.08 or 1.04 ± 0.31, respectively (Table 1). A similar result was obtained using PATtreated mouse kidney homogenates, in which there was a dosedependent increase in MDA, indicating an enhancement of lipid peroxidation.

FIG. 2. Downregulation of CuZn-SOD in HEK293 cells and its effect on cell damage. HEK293 cells were transfected with vector or CuZn-SOD siRNA plasmid, and then stable clones were selected as described in the Materials and Methods. (A) Western blot analysis of stable clones for CuZn-SOD and tubulin. (B) Vector-treated and siRNA clone 1 cultures were exposed to various concentrations of PAT for 4 h, and then plasma membrane damage was estimated by measuring LDH activity released into the culture medium. The data are the mean ± SEM for four independent experiments and expressed as a fold value of that seen with vehicle-treated cells (0lM PAT). Significant difference (**p < 0.01; ***p < 0.001) compared with the vector-treated group.

Roles of CuZn-SOD and Catalase in PAT-Induced Cytotoxicity It has been suggested that SODs and catalase work in concert to detoxify superoxide anion and H2O2. To determine the role of superoxide anion in PAT-induced cytotoxicity, we decreased CuZn-SOD levels in HEK293 cells using siRNA. Three weeks after transfection, three stable clones containing the CuZn-SOD siRNA plasmid were selected and subjected to Western blotting for CuZn-SOD levels, and all showed a dramatic reduction compared with vector-transfected cells (Fig. 2A). Since in Table 1 PAT was able to activate lipid peroxidation, cellular membrane damage was used as a parameter to evaluate the cytotoxicity of PAT treatment. As shown in Figure 2B, significant LDH leakage was detected when clone 1 was treated for 4 h with PAT concentrations of 50 and 100lM. After incubation with 50lM PAT, LDH leakage from clone 1 (14.89 ± 1.74) was 16-fold higher than that from vector-transfected cells (0.92 ± 0.04). On the other hand, to confirm the effect of CuZn-SOD on PAT-induced cytotoxicity, stable CuZn-SOD–overexpressing clones were established from CHO-K1 cells (Fig. 3A).

Compared with HEK293 cells, CHO-K1 showed a much lower level of endogenous CuZn-SOD, so it was chosen for easily observing the effect of overexpressed CuZn-SOD. As shown in Fig. 3B, overexpression of CuZn-SOD significantly reduced LDH leakage from cells treated with 50 or 100lM PAT by ~50% compared with PAT-treated vector-transfected cells. To determine whether H2O2 was also attributable to PATinduced cytotoxicity, we measured the levels of cellular LDH leakage in CHO-K1 cultures with or without overexpression of catalase. Two stable CHO-K1 clones (cat 7 and cat 14) expressing high levels of catalase showed significant protection from PAT-induced cell membrane damage (Figs. 4A and B). When clone 7 was treated for 4 h with 50 or 100lM PAT, LDH leakage were significantly reduced to 35% or 57%, respectively, of that seen using the PAT-treated vector control. Similarly, when catalase-overexpressing HEK293 clones were exposed to 100lM PAT, it also led to a reduction in extracellular LDH activity to 17.4% of PAT-treated control levels (data not shown).

344

LIU ET AL.

FIG. 3. Overexpression of CuZn-SOD in CHO-K1 cells and the protection against cell damage. CHO-K1 cells were transfected with the pcDNA 3.1 vector or the human CuZn-SOD–expressing plasmid and stable clones selected. (A) Cell lysates of clones subjected to Western blot analysis for CuZn-SOD and tubulin. (B) Vector-treated and CuZn-SOD–overexpressing clone 1 cultures were exposed to PAT for 4 h, and then extracellular LDH activity was measured. The data, expressed as a fold value of that in cells exposed to vehicle (0lM PAT), are the mean ± SEM of four independent experiments. ***p < 0.001, significant difference compared with the paired group.

FIG. 4. Overexpression of catalase in CHO-K1 cells and the protection against cell damage. CHO-K1 cells cells were transfected with pcDNA 3.1 vector or the human catalase-expressing plasmid and stable clones were selected. (A) Cell lysates of clones subjected to Western blot analysis for catalase and tubulin. (B) Cultures of clones 7 and 14 were exposed to PAT (20, 50, or 100lM) for 4 h, and then extracellular LDH activity was measured. The data, expressed as a fold value of that in cells exposed to vehicle, are the mean ± SEM for four separate experiments. **p < 0.01, significant difference compared with the paired group.

These results strongly suggest that, in our model system, both superoxide anion and H2O2 contribute to the cytotoxicity and oxidative stress associated with PAT treatment. The Relationship between PAT-Induced ROS Generation and ERK1/2 Activation in Human Cells We have previously shown that PAT activates the ERK1/2 signaling pathway in various mammalian cell lines (Wu et al., 2005). Therefore, we would like to understand whether there is a cross talk between ERK1/2 pathway and ROS generation in PAT-treated cultures. As shown in Fig. 5, when HEK293 cells were preincubated with Tiron, a free radical scavenger, before coexposure to PAT, the presence of Tiron for either 30 or 60 min significantly inhibited the ERK1/2 phosphorylation induced by PAT, suggesting that ROS may play a role in PATelicited ERK1/2 activation. On the other hand, we also examined whether ERK1/2 activation led to the ROS generation. It was found that the elevated phospho-ERK level caused by PAT was dramatically decreased by pretreatment of HEK293 cultures with the MEK1/2 inhibitor, U0126 (Fig. 6A). When

FIG. 5. Effect of Tiron on PAT-activated ERK1/2 phosphorylation. HEK293 cells were pretreated for 2 h with or without 5mM Tiron and then coincubated with vehicle or 30lM PAT for another 30 or 60 min. Whole-cell extracts were prepared immediately and subjected to Western blotting using phospho-ERK1/2 or ERK1/2 antibodies. The relative phospho-ERK1/2 (pERK) levels shown in the lower panels are the mean ± SEM, which were normalized by arbitrarily setting the value for Tiron-treated cells as 1. *p < 0.05, significant difference compared with the paired group.


FIG. 6. Effects of U0126 on PAT-induced ERK phosphorylation (A) and ROS generation (B). HEK293 cells were pretreated for 30 min with or without U0126 and then coincubated with either vehicle or 50lM PAT for another 30 min. Whole-cell extracts were subjected to Western blotting using phosphoERK1/2 or ERK1/2 antibodies. In (B), cells were left untreated (h) or treated (n) with 10lM U0126 for 30 min. After incubation for 30 min with 10lM H2DCF-DA, the medium was removed, and various concentrations of PAT were added for another 45 min. The levels of DCF fluorescence were measured by fluorocytometry, and the data are expressed as the mean ± SEM for four independent experiments. *p < 0.05, significant difference compared with the paired group.

HEK293 cells were treated with 100lM PAT alone, the DCF fluorescence value was 2.63 ± 0.19; this value was significantly reduced to 1.67 ± 0.17 in the presence of U0126 (Fig. 6B). Downregulation of ERK1/2 phosphorylation by U0126 seems to result in a partial blockage of ROS production in PAT-treated cultures.

DISCUSSION

PAT, which is often found in moldy fruits and their derivatives (Doores, 1983; Piemontese et al., 2005), is mutagenic and teratogenic in certain animal and cell models (Riley and Showker, 1991; Smith et al., 1993). It is generally believed that PAT, which is electrophilic, exerts its toxicity by covalently binding to sulfhydryl groups in macromolecules (Liu et al., 2006; Pfeiffer et al., 2005). Studies on liver cells or liver slices have also demonstrated that PAT treatment results in glutathione depletion (Barhoumi and Burghardt, 1996; Mansfield et al., 2004). It is known that disturbance of glutathione homeostasis can either lead to or result from oxidative stress (Chen et al.,

345

2000; Schulz et al., 2000). In the present study, using different approaches, we provided evidence that both HEK293 and HL60 cultures rapidly generated ROS, including superoxide anion, after PAT treatment. Our findings are consistent with those of Barhoumi and Burghardt (1996) showing that ROS is generated in PAT-treated liver cell cultures. Furthermore, we provide another evidence of ROS generation by showing that product of lipid peroxidation, MDA, was found in HL-60 cells and mouse kidney homogenates incubated with PAT (Table 1). In addition, either the mouse liver or brain homogenates showed similar MDA levels as in kidney homogenates after PAT treatment (data not shown), implying that PAT did not prefer a specific tissue to exert its lipid peroxidation ability. Oxidative/nitrosative stress caused by ROS and reactive nitrogen species is a major cause for the formation of DNA-reactive MDA (Bartsch and Nair, 2004). Although MDA-modified DNA adducts have been detected in animal and human tissues and may be a marker of human cancer risk (Khanzode et al., 2004; Zhang et al., 2002), there is no direct evidence to correlate its formation and PAT-induced DNA damage. There are two main types of SOD in mammalian cells, CuZnSOD, located in the cytoplasm, and Mn-SOD, located in the mitochondria. Previous studies of CuZn-SOD overexpression or defects in animal models have demonstrated that CuZn-SOD may be involved in decreased myocardial ischemia/perfusion or axonal injury in mice and in maintaining the normal life span in Drosophila and mice (Liu et al., 1999; Phillips et al., 1989; Reaume et al., 1996). In the present study, reduction of CuZnSOD expression in HEK293 cells using siRNA led to an increase in LDH release after PAT treatment (Fig. 2), suggesting the involvement of superoxide anion in PAT-induced cytotoxicity. Although CuZn-SOD serves as the primary cellular defense against superoxide anion, lower levels of CuZn-SOD in siRNA transfectants did not result in cell death or morphological changes until two months after transfection (data not shown). The role of superoxide anion in PAT-induced toxicity was confirmed by the fact that the cell membrane of CuZnSOD–overexpressing cells was more resistant to short-term (4 h) PAT damage than that of the vector-transfected control (Fig. 3). Nevertheless, after a long term (24 h) exposure to PAT, the viability of SOD-overexpressing cultures was not significantly elevated compared with the vector controls in 3-(4,5-dimethylthiazole-2-yl)-2,5-biphenyl tetrazolium bromide viability assay (data not shown). This unexpected observation indicated that a high CuZn-SOD level failed to rescue the cell death after 24 h incubation with PAT, even though it prevented the membrane damage caused by 4 h exposure; the mechanism associated with the discrepancy is not clear. In theory, scavenging of superoxide anion by CuZn-SOD may lead to an increase in H2O2. Without a concomitant increase in peroxide-scavenging enzymes, excess H2O2 will react with Fe2þ to generate ROS via the Fenton reaction (Zhong et al., 1997). The hydroxyl radicals, rather than the superoxide anions, generated by this reaction are the major cause of damage to

346

LIU ET AL.

biological macromolecules. Catalase is responsible for protecting cells from accumulation of H2O2 by converting it to H2O and O2 (Fridovich, 1986). As shown in Fig. 5, overexpression of catalase in the cytosol protected cells from PAT-induced LDH leakage, indicating that H2O2 was involved in the membrane damage caused by PAT. This result is consistent with previous reports that overexpression of catalase in either the cytosolic or mitochondrial compartment protects HepG2 cells and human prostate cancer cells against various oxidative injuries (Ahmad et al., 2005; Bai et al., 1999). Nevertheless, after 24 h of PAT treatment, the viability of transfectants with high levels of catalase was no different from that of vector-treated clones (data not shown). The observed different effects on stress resistance and cell viability may result from the fact that oxidative stress is not the only factor contributing to PATinduced cytotoxicity. This possibility is supported by several documents showing that increased levels of catalase provide protection against oxidative damage in Drosophila but are not able to prolong the life spans of flies (Mockett et al., 2003; Sun and Tower, 1999). We showed that the presence of a ROS inhibitor, Tiron, partially prevented the ERK1/2 phosphorylation caused by PAT (Fig. 5), and an ERK1/2 pathway inhibitor, U0126, led to lower levels of ROS generation in PAT-treated human cells (Fig. 6). Tiron is considered to be a nontoxic chelator of metals as well as a cell membrane–permeating scavenger of superoxide and protects cultured cells against H2O2- and superoxide anion– induced cytotoxicity (Krishna et al., 1992). ROS generation may change the redox status of cells with subsequent effects on specific kinases, phosphatases, and transcription factors (Cakir and Ballinger, 2005) and is known to be associated with MAPK pathway activation. Treatment of mammalian cells, including Jurkat T, NIH3T3, and vascular smooth muscle cells, with H2O2/ superoxide anion activates the ERK1/2 signaling pathways (Lee et al., 2000; Song and Lee, 2003; Susa and Wakabayashi, 2003). On the other hand, MAPKs play a pivotal role in ROS generation (Krishna et al., 1992; Woo et al., 2002). Our results indicate that, in PAT-treated cultures, the ERK1/2 signaling pathway is not only activated by ROS but may also play an upstream role in the process of ROS formation. A similar phenomenon has been reported for the JNK signaling pathway, which is activated during oxidative stress and further controls the expression of proteins that contribute to oxidative stress (Lee et al., 2000). We have previously shown that treatment of HEK293 cells with PAT-cysteine adduct leads no cytotoxicity and MAPK pathway activation (Liu et al., 2006). We herein also found that the ROS-inducing ability of PAT can be abolished either by the copresence of cysteine or by blocking the thiol-reacting sites on the toxin with cysteine (data not shown). The thiol group of cysteine is known to be preferred for PAT-mediated crosslink reactions (Fliege and Metzler, 1999). Therefore, these evidences strongly suggest that toward cellular SH-containing macromolecules by electrophilic PAT results in the formation of ROS in cell cultures. It is also supported by the finding of Barhoumi and

Burghardt (1996) that treatment of rat liver cells with PAT leads to glutathione depletion right before the ROS generation. In conclusion, our data show that PAT, a mycotoxin commonly found in apple juice and related products, damaged the cell membrane through ROS generation. In addition, the ERK1/2 pathway was not only activated by ROS but may also mediate ROS generation in PAT-treated cultures. Whether the generation of ROS by PAT in mammalian cells correlates with the mutagenic and clastogenic properties of PAT need to be further investigated.

ACKNOWLEDGMENTS The authors thank Dr. Tisha King Heiden for her assistance in statistics analysis. This work was supported by grants NSC 93-2313-B-040-003 and 942313-B-040-001 from the National Science Council of the Republic of China, Taiwan and by fundings from Tian-Sheng Memorial Hospital, Tong kong, Ping-Tong, Taiwan.

REFERENCES Ahmad, I. M., Aykin-Burns, N., Sim, J. E., Walsh, S. A., Higashikubo, R., Buettner, G. R., Venkataraman, S., Mackey, M. A., Flanagan, S. W., Oberley, L. W., et al. (2005). Mitochondrial O2*- and H2O2 mediate glucose deprivation-induced stress in human cancer cells. J. Biol. Chem. 280, 4254–4263. Allen, R. G., and Tresini, M. (2000). Oxidative stress and gene regulation. Free Radic. Biol. Med. 28, 463–499. Bai, J., Rodriguez, A. M., Melendez, J. A., and Cederbaum, A. I. (1999). Overexpression of catalase in cytosolic or mitochondrial compartment protects HepG2 cells against oxidative injury. J. Biol. Chem. 274, 26217–26224. Barhoumi, R., and Burghardt, R. C. (1996). Kinetic analysis of the chronology of patulin- and gossypol-induced cytotoxicity in vitro. Fundam. Appl. Toxicol. 30, 290–297. Bartsch, H., and Nair, J. (2004). Oxidative stress and lipid peroxidation-derived DNA-lesions in inflammation driven carcinogenesis. Cancer Detect. Prev. 28, 385–391. Bindokas, V. P., Jordan, J., Lee, C. C., and Miller, R. J. (1996). Superoxide production in rat hippocampal neurons: Selective imaging with hydroethidine. J. Neurosci. 16, 1324–1336. Cakir, Y., and Ballinger, S. W. (2005). Reactive species-mediated regulation of cell signaling and the cell cycle: The role of MAPK. Antioxid. Redox. Signal. 7, 726–740. Chen, Z., Oberley, T. D., Ho, Y., Chua, C. C., Siu, B., Hamdy, R. C., Epstein, C. J., and Chua, B. H. (2000). Overexpression of CuZnSOD in coronary vascular cells attenuates myocardial ischemia/reperfusion injury. Free Radic. Biol. Med. 29, 589–596. Doores, S. (1983). The microbiology of apples and apple products. Crit. Rev. Food Sci. Nutr. 19, 133–149. Fliege, R., and Metzler, M. (1999). The mycotoxin patulin induces intra- and intermolecular protein crosslinks in vitro involving cysteine, lysine, and histidine side chains, and alpha-amino groups. Chem. Biol. Interact. 123, 85–103. Fridovich, I. (1986). Biological effects of the superoxide radical. Arch. Biochem. Biophys. 247, 1–11. Hempel, S. L., Buettner, G. R., O’Malley, Y. Q., Wessels, D. A., and Flaherty, D. M. (1999). Dihydrofluorescein diacetate is superior for detecting intracellular oxidants: Comparison with 2#,7#-dichlorodihydrofluorescein


347

diacetate, 5(and 6)- carboxy-2#,7#-dichlorodihydrofluorescein diacetate, and dihydrorhodamine 123. Free Radic. Biol. Med. 27, 146–159.

Pfeiffer, E., Diwald, T. T., and Metzler, M. (2005). Patulin reduces glutathione level and enzyme activities in rat liver slices. Mol. Nutr. Food Res. 49, 329–336.

Henle, E. S., and Linn, S. (1997). Formation, prevention, and repair of DNA damage by iron/hydrogen peroxide. J. Biol. Chem. 272, 19095–19098. Janero, D. R. (1990). Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid peroxidation and peroxidative tissue injury. Free Radic. Biol. Med. 9, 515–540.

Phillips, J. P., Campbell, S. D., Michaud, D., Charbonneau, M., and Hilliker, A. J. (1989). Null mutation of copper/zinc superoxide dismutase in Drosophila confers hypersensitivity to paraquat and reduced longevity. Proc. Natl. Acad. Sci. U.S.A. 86, 2761–2765.

Khanzode, S. S., Muddeshwar, M. G., Khanzode, S. D., and Dakhale, G. N. (2004). Antioxidant enzymes and lipid peroxidation in different stages of breast cancer. Free Radic. Res. 38, 81–85. Krishna, C. M., Liebmann, J. E., Kaufman, D., DeGraff, W., Hahn, S. M., McMurry, T., Mitchell, J. B., and Russo, A. (1992). The catecholic metal sequestering agent 1,2-dihydroxybenzene-3,5-disulfonate confers protection against oxidative cell damage. Arch. Biochem. Biophys. 294, 98–106. Leach, J. K., Van Tuyle, G., Lin, P. S., Schmidt-Ullrich, R., and Mikkelsen, R. B. (2001). Ionizing radiation-induced, mitochondria-dependent generation of reactive oxygen/nitrogen. Cancer Res. 61, 3894–3901. Lee, Y. J., Galoforo, S. S., Sim, J. E., Ridnour, L. A., Choi, J., Forman, H. J., Corry, P. M., and Spitz, D. R. (2000). Dominant-negative Jun N-terminal protein kinase (JNK-1) inhibits metabolic oxidative stress during glucose deprivation in a human breast carcinoma cell line. Free Radic. Biol. Med. 28, 575–584. Liu, B. H., Wu, T. S., Yu, F. Y., and Wang, C. H. (2006). Mycotoxin patulin activates the p38 kinase and JNK signaling pathways in human embryonic kidney cells. Toxicol. Sci. 89, 423–430. Liu, B. H., Yu, F. Y., Wu, T. S., Li, S. Y., Su, M. C., Wang, M. C., and Shih, S. M. (2003). Evaluation of genotoxic risk and oxidative DNA damage in mammalian cells exposed to mycotoxins, patulin and citrinin. Toxicol. Appl. Pharmacol. 191, 255–263. Liu, D., Liu, J., and Wen, J. (1999). Elevation of hydrogen peroxide after spinal cord injury detected by using the Fenton reaction. Free Radic. Biol. Med. 27, 478–482. Mahfoud, R., Maresca, M., Garmy, N., and Fantini, J. (2002). The mycotoxin patulin alters the barrier function of the intestinal epithelium: Mechanism of action of the toxin and protective effects of glutathione. Toxicol. Appl. Pharmacol. 181, 209–218. Mansfield, K. D., Simon, M. C., and Keith, B. (2004). Hypoxic reduction in cellular glutathione levels requires mitochondrial reactive oxygen species. J. Appl. Physiol. 97, 1358–1366. Maxwell, M. M., Pasinelli, P., Kazantsev, A. G., and Brown, R. H., Jr (2004). RNA interference-mediated silencing of mutant superoxide dismutase rescues cyclosporin A-induced death in cultured neuroblastoma cells. Proc. Natl. Acad. Sci. U.S.A. 101, 3178–3183. McKinley, E. R., Carlton, W. W., and Boon, G. D. (1982). Patulin mycotoxicosis in the rat: Toxicology, pathology and clinical pathology. Food Cosmet. Toxicol. 20, 289–300. Milligan, S. A., Owens, M. W., and Grisham, M. B. (1998). Differential regulation of extracellular signal-regulated kinase and nuclear factor-jB signal transduction pathways by hydrogen peroxide and tumor necrosis factor. Arch. Biochem. Biophys. 352, 255–262.

Piemontese, L., Solfrizzo, M., and Visconti, A. (2005). Occurrence of patulin in conventional and organic fruit products in Italy and subsequent exposure assessment. Food Addit. Contam. 22, 437–442. Reaume, A. G., Elliott, J. L., Hoffman, E. K., Kowall, N. W., Ferrante, R. J., Siwek, D. F., Wilcox, H. M., Flood, D. G., Beal, M. F., Brown, R. H., Jr, et al. (1996). Motor neurons in Cu/Zn superoxide dismutase-deficient mice develop normally but exhibit enhanced cell death after axonal injury. Nat. Genet. 13, 43–47. Riley, R. T., and Showker, J. L. (1991). The mechanism of patulin’s cytotoxicity and the antioxidant activity of indole tetramic acids. Toxicol. Appl. Pharmacol. 109, 108–126. Rothe, G., and Valet, G. (1990). Flow cytometric analysis of respiratory burst activity in phagocytes with hydroethidine and 2#,7#-dichlorofluorescin. J. Leukoc. Biol. 47, 440–448. Schulz, J. B., Lindenau, J., Seyfried, J., and Dichgans, J. (2000). Glutathione, oxidative stress and neurodegeneration. Eur. J. Biochem. 267, 4904–4911. Smith, E. E., Duffus, E. A., and Small, M. H. (1993). Effects of patulin on postimplantation rat embryos. Arch. Environ. Contam. Toxicol. 25, 267–270. Song, J. J., and Lee, Y. J. (2003). Catalase, but not MnSOD, inhibits glucose deprivation-activated ASK1-MEK-MAPK signal transduction pathway and prevents relocalization of Daxx: Hydrogen peroxide as a major second messenger of metabolic oxidative stress. J. Cell. Biochem. 90, 304–314. Srinivasan, A., Lehmler, H. J., Robertson, L. W., and Ludewig, G. (2001). Production of DNA strand breaks in vitro and reactive oxygen species in vitro and in HL-60 cells by PCB metabolites. Toxicol. Sci. 60, 92–102. Sun, J., and Tower, J. (1999). FLP recombinase-mediated induction of Cu/Znsuperoxide dismutase transgene expression can extend the life span of adult Drosophila melanogaster flies. Mol. Cell. Biol. 19, 216–228. Susa, S., and Wakabayashi, I. (2003). Extracellular alkalosis activates ERK mitogen-activated protein kinase of vascular smooth muscle cells through NADPH-mediated formation of reactive oxygen species. FEBS Lett. 554, 399–402. Van Egmond, H. P. (1989). Current situation on regulations for mycotoxins. Overview of tolerances and status of standard methods of sampling and analysis. Food Addit. Contam. 6, 139–188. Woo, C. H., You, H. J., Cho, S. H., Eom, Y. W., Chun, J. S., Yoo, Y. J., and Kim, J. H. (2002). Leukotriene B(4) stimulates Rac-ERK cascade to generate reactive oxygen species that mediates chemotaxis. J. Biol. Chem. 277, 8572– 8578. Wu, T. S., Yu, F. Y., Su, C. C., Kan, J. C., Chung, C. P., and Liu, B. H. (2005). Activation of ERK mitogen-activated protein kinase in human cells by the mycotoxin patulin. Toxicol. Appl. Pharmacol. 207, 103–111.

Mockett, R. J., Bayne, A. C., Kwong, L. K., Orr, W. C., and Sohal, R. S. (2003). Ectopic expression of catalase in Drosophila mitochondria increases stress resistance but not longevity. Free Radic. Biol. Med. 34, 207–217.

Zhang, Y., Chen, S., Hsu, T., and Santella, R. M. (2002). Immunochemical detection of malondialdehyde-DNA adducts in human oral mucosa cells. Carcinogenesis 23, 207–211.

Narayanan, P. K., Goodwin, E. H., and Lehnert, B. E. (1997). Alpha particles initiate biological production of superoxide anions and hydrogen peroxide in human cells. Cancer Res. 57, 3963–3971.

Zhong, W., Oberley, L. W., Oberley, T. D., and St Clair, D. K. (1997). Suppression of the malignant phenotype of human glioma cells by overexpression of manganese superoxide dismutase. Oncogene 14, 481–490.