Intracellular GSH levels rather than ROS levels are tightly related to AMA-induced HeLa cell death

Available online at www.sciencedirect.com

Chemico-Biological Interactions 171 (2008) 67–78

Intracellular GSH levels rather than ROS levels are tightly related to AMA-induced HeLa cell death Yong Hwan Han, Suhn Hee Kim, Sung Zoo Kim, Woo Hyun Park ∗ Department of Physiology, Medical School, Research Institute of Clinical Medicine, Center for Healthcare Technology Development, Chonbuk National University, JeonJu 561-180, Republic of Korea Received 26 May 2007; received in revised form 31 August 2007; accepted 31 August 2007 Available online 6 September 2007

Abstract Antimycin A (AMA) inhibits succinate oxidase and NADH oxidase, and also inhibits mitochondrial electron transport between cytochromes b and c. We investigated the involvement of ROS and GSH in AMA-induced HeLa cell death. AMA increased the intracellular H2 O2 and O2 •− levels and reduced the intracellular GSH content. ROS scavengers (Tempol, Tiron, Trimetazidine and NAC) did not down-regulate the production of ROS and inhibit apoptosis in AMA-treated cells. Treatment with NAC and N-propylgallate showing the enhancement of GSH depletion in AMA-treated cells significantly intensified the levels of apoptosis. Calpain inhibitors I and II (calpain inhibitor III) and Ca2+ -chelating agent (EGTA/AM) significantly reduced H2 O2 levels in AMAtreated HeLa cells. However, treatment with calpain inhibitor III intensified the levels of O2 •− in AMA-treated cells. In addition, calpain inhibitor III strongly depleted GSH content with an enhancement of apoptosis in AMA-treated cells. Conclusively, the changes of ROS by AMA were not tightly correlated with apoptosis in HeLa cells. However, intracellular GSH levels are tightly related to AMA-induced cell death. © 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Antimycin A; ROS; Apoptosis; HeLa; ROS scavenger; GSH

1. Introduction Antimycin A (AMA) is a product that is predominantly composed of antimycin A1 and A3, which

Abbreviations: AMA, antimycin A; CMFDA, 5-chloromethylfluorescein diacetate; DHE, dihydroethidium; FBS, fetal bovine serum; FITC, fluorescein isothiocyanate; GSH, glutathione; H2 DCFDA, 2 ,7 -dichlorodihydrofluorescein diacetate; NAC, N-acetylcysteine; NADPH, nicotine adenine diphosphate; PBS, phosphate buffer saline; PI, propidium iodide; ROS, reactive oxygen species; SOD, superoxide dismutase; XO, xanthine oxidase. ∗ Corresponding author. Tel.: +82 63 270 3079; fax: +82 63 274 9892. E-mail address: [email protected] (W.H. Park).

are derived from Streptomyces kitazawensis [1]. AMA inhibits succinate oxidase and NADH oxidase, and also inhibits mitochondrial electron transport between cytochromes b and c [2–5]. The inhibition of electron transport causes a collapse of the proton gradient across the mitochondrial inner membrane, thereby breaking down the mitochondrial membrane potential (Ψ m ) [2,4,6]. This inhibition also results in the production of reactive oxygen species (ROS) [6,7]. Evidence indicates that either the presence of ROS or the collapse of mitochondrial membrane potential (Ψ m ) opens the mitochondrial permeability transition pore, which is accompanied by the release of proapoptotic molecules such as cytochrome c into the cytoplasm [8–10]. Because AMA acts directly on the mitochondria, AMA-induced

0009-2797/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.cbi.2007.08.011

68

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

apoptosis has been reported in many experiments, including our report on As4.1 juxtaglomerular cells [11–16]. ROS include hydrogen peroxide (H2 O2 ), superoxide anion (O2 •− ), hydroxyl radical (• OH) and peroxynitrite (ONOO− ). ROS are formed as by-products of mitochondrial respiration or oxidases such as nicotine adenine diphosphate (NADPH) oxidase, xanthine oxidase (XO) and certain arachidonic acid oxygenases [17]. A change in the redox state of the tissue implies a change in ROS generation or metabolism. Cells possess antioxidant systems to control the redox state, which is important for their survival. Excessive production of ROS gives rise to activation of events, which lead to death and survival in several types of cells [18–20]. The exact mechanisms involved in cell death induced by ROS are not fully understood, and the protective effect mediated by some antioxidants is controversial. AMA produces rapid ATP depletion that correlated to rapid and sustained increased in cytosolic free Ca2+ [21]. It has been suggested that an increase in cytosolic free Ca2+ mediate cell death following anoxia/hypoxia and chemical exposure [22,23]. Prolonged increases in cytosolic free Ca2+ activate degradative enzymes such as calpains, Ca2+ -activated cysteine proteases, which have a number of substrates including cytoskeletal proteins, procaspase-3 and PARP [24,25]. Therefore, it is worthy to investigate the effect of cytosolic free Ca2+ on AMAtreated cells. We have recently demonstrated that AMA inhibited the growth of HeLa cells with an IC50 of about 50 ␮M [26]. AMA efficiently induced apoptosis, as evidenced by flow cytometric detection of sub-G1 DNA content, annexin V binding assay and DAPI staining. In the present study, we evaluated the involvement of ROS and GSH in AMA-induced HeLa cell death and investigated whether ROS scavengers, calpain inhibitor and Ca2+ chelating agent rescue HeLa cells from AMA-induced death. 2. Materials and methods 2.1. Cell culture Human cervical adenocarcinoma HeLa cells were maintained in humidified incubator air containing 5% CO2 at 37 ◦ C. HeLa cells were cultured in RPMI-1640 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (GIBCO BRL, Grand Island, NY). Cells were routinely grown in 100-mm plastic tissue culture dishes (Nunc, Roskilde, Denmark), and were harvested with a solution of trypsin-EDTA (0.05%

trypsin and 0.53 mM EDTA) when they were in a logarithmic phase of growth. Cells were maintained at the above-described culture conditions for all experiments. 2.2. Reagents AMA (Sigma–Aldrich Chemical Company, St. Louis, MO) was dissolved in ethanol at 2 × 10−2 M as a stock solution. The cell-permeable O2 •− scavengers, 4-hydroxy-TEMPO (4-hydroxyl-2,2,6,6-tetramethylpierydine-1-oxyl) (Tempol), 4,5-dihydroxyl1,3-benzededisulfonic acid (Tiron), 1-[2,3,4-trimethoxibenzyl]-piperazine (Trimetazidine), N-acetylcysteine (NAC) and N-propylgallate were obtained from Sigma. These were dissolved in ethanol or water solution buffer at 1 × 10−1 M as a stock solution. The cell-permeable inhibitor of calpains I and II (calpain inhibitor III; carbobenzoxy-valyl-phenylalanial) and also cellpermeable of the Ca2+ -chelating agent (EGTA/AM; ethyleneglycol-bis(␤-aminoethyl)-N,N,N ,N tetraacetoxymethyl Ester) was purchased from Calbiochem (San Diego, CA). These were dissolved in DMSO at 1 × 10−1 M as a stock solution. All of the stock solutions were wrapped in foil and kept at 4 or −20 ◦ C. 2.3. Detection of intracellular H2 O2 and O2 •− concentration Intracellular H2 O2 concentration was detected by means of an oxidation-sensitive fluorescent probe dye, 2 ,7 -Dichlorodihydrofluorescein diacetate (H2 DCFDA) (Invitrogen Molecular Probes, Eugene, OR). H2 DCFDA was intracellularly deacetylated by nonspecific esterase, which was further oxidized by cellular peroxides to the fluorescent compound, 2,7-dichlorofluorescein (DCF) (Ex /Em = 485 nm/535 nm). Dihydroethidium (DHE) (Invitrogen Molecular Probes) is a fluorogenic probe that is highly selective for superoxide anion radical detection. DHE is cell-permeable and reacts with superoxide anion to form ethidium, which, in turn, intercalates in the deoxyribonucleic acid, thereby exhibiting a red fluorescence (Ex /Em = 515 nm/595 nm). In brief, cells were incubated with AMA with or without ROS scavengers, calpain inhibitor and Ca2+ -chelating agent for the indicated times. Cells were then washed in PBS and incubated with 20 ␮M H2 DCFDA or 5 ␮M DHE at 37 ◦ C for 30 min according to the instructions of the manufacturer. DCF fluorescence and red fluorescence were detected using a FACStar flow cytometer (Becton Dickinson). For each sample, 5000 or 10,000 events were collected. H2 O2 and O2 •− production were expressed as mean fluorescence

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

intensity (MFI), which was calculated by CellQuest software. 2.4. Detection and measurement of intracellular glutathione (GSH) Cellular GSH levels were analyzed using 5chloromethylfluorescein diacetate (CMFDA, Molecular Probes) or the Glutathione Assay Kit (Sigma, St. Louis, MO). CMFDA is a membrane-permeable dye for determining levels of intracellular glutathione [27,28]. In brief, cells were incubated with AMA with or without ROS scavengers, calpain inhibitor and Ca2+ -chelating agent for 72 h. Cells were then washed in PBS and incubated with 5 ␮M CMFDA at 37 ◦ C for 30 min according to the instructions of the manufacturer. Cytoplasmic esterases convert nonfluorescent CMFDA to fluorescent 5-chloromethylfluorescein, which can then react with the glutathione (Ex /Em = 522 nm/595 nm). Propidium Iodide (PI; (1 ␮g/ml)) (Ex /Em = 488 nm/617 nm) was subsequently added, and CMF fluorescence and PI staining intensity were determined using a FACStar flow cytometer (Becton Dickinson) and calculated by CellQuest software. PI was used to differentiate dead and normal cells. This agent is membrane impermeant and generally excluded from viable cells. For each sample, 5000 or 10,000 events were collected. To measure the level of total GSH in AMA-treated HeLa cells using the Glutathione Assay Kit, cells were incubated with the designated dose of AMA for 72 h. 1 × 107 cells were then washed in PBS and suspended in 3 vol. of the 5% 5-sulfosalicylic acid (SSA) solution. The samples were freezed and thawed three times using liquid nitrogen and 37 ◦ C water bath. They were then centrifuged at 10,000 × g for 10 min and total GSH in supernatant was measured. In brief, 150 ␮l of 100 mM potassium phosphate buffer, pH 7.0, with 1 mM EDTA was mixed with 5 ␮l of the glutathione reductase (6 units/ml) and 5 ␮l of DTNB (5,5 -dithiobis(2-nitrobenzoic acid)) solution (1.5 mg/ml), and to this mixture were added 1 ␮l of sample and 50 ␮l of NADPH solution (0.16 mg/ml in 150 ␮l of 100 mM potassium phosphate buffer, pH 7.0, with 1 mM EDTA). A standard calibration curve was prepared using 0, 2, 5, and 10 ␮l of 50 ␮M GSH in 5% SSA solution. The optical density of each sample was measured at 412 nm using a microplate reader (Spectra MAX 340, Molecular Devices Co, Sunnyvale, California, USA) after the incubation of 5 min at room temperature. Each plate contained multiple wells of a given experimental condition and multiple control wells.

69

2.5. Cell viability assay The in vitro cell viability effect of AMA with or without ROS scavengers on HeLa cells was determined by measuring 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) dye absorbance of living cells as described previously [29]. In brief, cells (2 × 105 cells per well) were seeded in 96-well microtiter plates (Nunc, Roskilde, Denmark). After exposure to AMA and/or ROS scavengers, calpain inhibitor and Ca2+ -chelating agent for 72 h, 50 ␮l of MTT (Sigma) solution (2 mg/ml in PBS) were added to each well, and the plates were incubated for additional 3 or 4 h at 37 ◦ C. MTT solution in medium was then aspirated off. To achieve solubilization of the formazan crystal formed in viable cells, 100 or 200 ␮l of DMSO were added to each well. The optical density of each well was measured at 570 nm using a microplate reader (Spectra MAX 340, Molecular Devices Co., Sunnyvale, California, USA). Each plate contained multiple wells of a given experimental condition and multiple control wells. This procedure was replicated for two to four plates per condition. 2.6. Quantification of caspase-3 and casase-8 activity The activity of caspase-3 and casase-8 was assessed using the caspase-3 and casase-8 Colorimetric Assay Kits (R&D systems Inc., Minneapolis, MN, USA), which are based on the spectrophotometric detection of the color reporter molecule p-nitroanaline (pNA) after cleavage from the labeled substrate DEVD-pNA (caspase-3) and IETD-pNA (caspase-8) as an index, respectively. Briefly, cells were incubated with the designated dose of AMA for 72 h. The cells were then washed in PBS and suspended in 5 vol. of lysis buffer (20 mM HEPES pH 7.9, 20% glycerol, 200 mM KCl, 0.5 mM EDTA, 0.5% NP40, 0.5 mM DTT, 1% Protease inhibitor cocktail (from Sigma)). The lysates were then collected and stored at −20 ◦ C until use. Protein concentration was determined by the Bradford method. Supernatant samples containing 100 ␮g of total protein were used for determination of caspase-3 and caspase-8 activity. These are added to each well in 96-well microtiter plates (Nunc, Roskilde, Denmark) with the DEVD-pNA and IETD-pNA at 37 ◦ C for 1–2 h. The optical density of each well was measured at 405 nm using a microplate reader (Spectra MAX 340, Molecular Devices Co., Sunnyvale, California, USA). Each plate contained multiple wells of a given experimental condition and multiple control wells. The activity of caspase-3 and caspase-8

70

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

was expressed in arbitrary absorbance units (absorbance at a wavelength of 405 nM). 2.7. Cell cycle and Sub-G1 analysis The cell cycle and sub-G1 distribution were determined by staining DNA with PI (Sigma–Aldrich) as described previously [30]. PI is also a fluorescent biomolecule that can be used to stain DNA. In brief, 1 × 106 cells were incubated with the designated doses of AMA with or without ROS scavengers, calpain inhibitor and Ca2+ -chelating agent for 72 h. Cells were then washed with phosphate-buffered saline (PBS) and fixed in 70% ethanol. Cells were again washed with PBS and then incubated with propidium iodine (PI: 10 ␮g/ml) with simultaneous treatment of RNase at 37 ◦ C for 30 min. The percentages of cells in the different phases of the cell cycle or having sub-G1 DNA content were measured with a FACStar flow cytometer (Becton Dickinson, San Jose, CA, USA) and analyzed using lysis II and Cellfit software (Becton Dickinson) or ModFit software (Verity Software Inc.). 2.8. Annexin V/PI staining Apoptosis was determined by staining cells with annexin V-fluorescein isothiocyanate (FITC) (Ex /Em = 488 nm/519 nm) and PI labeling, because annexin V can identify the externalization of phosphatidylserine during the progression of apoptosis and, therefore, can detect cells in early apoptosis. In brief, 1 × 106 cells were incubated with AMA with or without ROS scavengers, calpain inhibitor and Ca2+ -chelating agent for 72 h. The prepared cells were washed twice with cold PBS and then resuspended in 500 ␮l of binding buffer (10 mM HEPES/NaOH pH 7.4, 140 mM NaCl, and 2.5 mM CaCl2 ) at a concentration of 1 × 106 cells/ml. Five microliters of annexin V-FITC (PharMingen, San Diego, CA, USA) and PI (1 ␮g/ml) were then added to these cells, which were analyzed with a FACStar flow cytometer (Becton Dickinson). Viable cells were negative for both PI and annexin V; apoptotic cells were positive for annexin V and negative for PI, whereas late apoptotic dead cells displayed both high annexin V and PI labeling. Non-viable cells, which underwent necrosis, were positive for PI and negative for annexin V. 2.9. Measurement of mitochondrial membrane potential (Ψ m ) Mitochondrial membrane was monitored using the fluorescent dye, Rhodamine 123, a cell-permeable

cationic dye that preferentially enters into mitochondria based on highly negative mitochondrial membrane potential (Ψ m ). Depolarization of mitochondrial membrane potential (Ψ m ) results in the loss of Rhodamine 123 from the mitochondria and a decrease in intracellular fluorescence (Ex /Em = 485 nm/535 nm). In brief, 1 × 106 cells were incubated with AMA with or without ROS scavengers, calpain inhibitor and Ca2+ -chelating agent for 72 h. Cells were washed twice with PBS and incubated with Rhodamine 123 (0.1 ␮g/ml; Sigma) at 37 ◦ C for 30 min. PI (1 ␮g/ml) was subsequently added, and Rhodamine 123 and PI staining intensity were determined by flow cytometry. 2.10. Statistical analysis Results represent the mean of at least three independent experiments; bar, S.D. Microsoft Excel or Instat software (GraphPad Prism4, San Diego, CA, USA) was used to analyze the data. Student’s t-test or one-way analysis of variance (ANOVA) with post hoc analysis using Tukey’s multiple comparison test was used for parametric data. Statistical significance was defined as P < 0.05. 3. Results 3.1. Effect of AMA on ROS and GSH production in HeLa cells To assess the production of intracellular H2 O2 in the AMA-treated HeLa cells, we used H2 DCFDA fluorescence dye. As shown in Fig. 1A and B intracellular H2 O2 levels were increased in HeLa cells treated with AMA (2, 10 and 50 ␮M) at 72 h. In contrast, treatment with 100 and 200 ␮M AMA did not increase the intracellular H2 O2 levels. When we treated cells with 50 ␮M AMA for the short time, we observed an increase in H2 O2 levels within 2 min (Fig. 1C). At 120 min, the level of H2 O2 in AMA-treated cells was significantly increased about 22 times higher than that of the control cells (Fig. 1C). Next, we investigated the change of intracellular O2 •− in the AMA-treated HeLa cells. Red fluorescence derived from DHE, reflecting the accumulation of O2 •− , was increased in the AMA-treated HeLa cells (Fig. 1D, E and F). The level of O2 •− was increased approximately three times in cells treated with the concentration of 50 ␮M AMA at 72 h (Fig. 1D and E). The accumulation of O2 •− was also observed in 50 ␮M AM-treated cells at the early time of 2 min. At 120 min, the level of O2 •− was increased by approximately four times (Fig. 1F). A slow decrease phase in ROS levels in 50 ␮M AM-treated cells was observed after about 120 min (data not shown).

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

71

Fig. 1. Effects of AMA on ROS levels, H2 O2 and O2 •− in HeLa cells. Exponentially growing cells were treated with the indicated amount of AMA for the designated time. (A) Intracellular H2 O2 level (DCF fluorescence) in AMA-treated HeLa cells for 72 h was determined by a FACStar flow cytometer, as described in Section 2. (B) The graph shows the levels of mean DCF fluorescence of A. (C) The graph shows the levels of mean DCF fluorescence in AMA-treated cells. (D) Intracellular O2 •− level (DHE fluorescence) in AMA-treated HeLa cells for 72 h was determined by a FACStar flow cytometer. (E) The graph shows the levels of mean DHE fluorescence of D. (F) The graph shows the levels of mean DHE fluorescence in AMA-treated cells. *, P < 0.05 compared with the control group (Student’s t-test). AMA stands for antimycin A.

Cellular GSH has been shown to be crucial for regulation of cell proliferation, cell cycle progression and apoptosis [31,32]. Therefore, we analyzed the GSH level of HeLa cells by using CMF fluorescence. The M1 region of histograms in AMA-treated HeLa cells shows a lower level of intracellular GSH content (Supplement 1). AMA significantly elevated the percentages of cells residing in the M1 population at 72 h (Supplement 1 and Fig. 2B), which indicated the depletion of intracellular GSH content in HeLa cells by AMA. The noteworthy changes in the depletion of intracellular GSH content were observed at concentrations of approximately 50–100 ␮M AMA. To evaluate whether the M1 population of cells in the negative CMF fluorescence region were dead, we stained the cells with additional PI for verifying disruption of the plasma membrane. As shown in Fig. 2A and C, the negative CMF fluorescence cells mostly showed PI-

positive staining, which indicated that cells showing GSH depletion were dead. In addition, we observed that AMA decreased GSH levels (mean CMF fluorescence) in HeLa cells in a dose-dependent manner (Fig. 2D). Furthermore, we confirmed that AMA treatment significantly reduced GSH levels in HeLa cells, as evidenced by the direct measurement of glutathione content in cells (Fig. 2E). These data support that CMFDA is a membrane-permeable dye for determining the levels of intracellular glutathione [27,28]. 3.2. Effects of ROS scavengers, calpain inhibitor and Ca2+ -chelating agent on ROS production in AMA-treated HeLa cells To determine whether ROS production in AMAtreated HeLa cells was changed by ROS scavengers,

72

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

Fig. 2. Effects of AMA on GSH level (CMF fluorescence) and viability (PI fluorescence) in HeLa cells. Exponentially growing cells were treated with the indicated amount of AMA for 72 h. (A) CMF fluorescence cells and PI-positive staining cells were measured with a FACStar flow cytometer, as described in Section 2. (B) The graph shows the percent of CMF negative staining cells from A. (C) The graph shows the percent of CMF negative and PI-positive staining cells from A. (D) The graph shows the levels of mean CMF fluorescence of A. (E) The graph shows the level of GSH content. *, P < 0.05 compared with the control group (Student’s t-test).

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

73

Fig. 4. Effects of ROS scavengers, calpain inhibitor and Ca2+ chelating agent (EGTA/AM) on cell viability in HeLa cells. Exponentially growing cells were treated with the indicated amounts of ROS scavengers, calpain inhibitor and Ca2+ -chelating agent in addition to AMA for 72 h. Cell viability inhibition was assessed by MTT assays. *, P < 0.05 compared with the control group and #, P < 0.05 compared with the cells treated with only AMA (ANOVA test).

Fig. 3. Effects of ROS scavengers, calpain inhibitor and Ca2+ chelating agent (EGTA/AM) on intracellular ROS and GSH levels in AMA-treated HeLa cells. Exponentially growing cells were treated with the indicated amounts of ROS scavengers, calpain inhibitor and Ca2+ -chelating agent in addition to AMA for 72 h. The graphs show the levels of mean DCF fluorescence (H2 O2 level) (A), mean DHE fluorescence (O2 •− level) (B) and the percent of CMF negative (GSH depleted) cells (C). *, P < 0.05 compared with the control group and #, P < 0.05 compared with the cells treated with only AMA (ANOVA test).

cell-permeable ROS scavengers (Tempol and Tiron [33]), and a well-known antioxidants (NAC and N-propylgallate [34]), were co-incubated with the AMAtreated HeLa cells for 72 h. An anti-ischemic and metabolic agent, Trimetazidine, was also used as an indirect antioxidant [35,36]. The concentration of 500 ␮M ROS scavengers (Tempol, Tiron and Trimetazidine) we used in this experiment was considered to be optimal dose since the higher dose of ROS scavengers slightly inhibited cell growth (Fig. 4). To measure the accurate intracellular fluorescence level of ROS, we used only the cells residing in the R2 region; these cells are considered to have intact plasma membrane (Supplements 2A and B). In contrast to our expectation, none of the scavengers except N-propylgallate significantly reduced the intracellular H2 O2 levels in AMA-treated cells (Fig. 3A). These scavengers were also unable

to decrease the O2 •− levels in those cells (Fig. 3B). However, treatment with N-propylgallate intensified the increase of O2 •− levels in AMA-treated cells (Fig. 3B). The control cells treated with ROS scavengers (Tempol and Tiron) show the decrease of H2 O2 levels (Fig. 3A). ROS scavengers except N-propylgallate did not alter the basal levels of O2 •− in control cells (Fig. 3B). Treatment with N-propylgallate increased O2 •− levels in control cells (Fig. 3B). It is possible that the ROS scavengers can reduce ROS levels in AMA-treated HeLa cells in the early stage of incubation. Therefore, we assessed the levels of ROS in HeLa cells treated with AMA and ROS scavengers at an early time point (2, 20, 30 and 60 min). Treatment with Tiron and N-propylgallate inhibited an increase in H2 O2 levels in 50 ␮M AMA-treated cells at an early time point of 2 min (Supplement 3). All the ROS scavengers except Tempol prevented the increased levels of H2 O2 in AMA-treated cells from 20 to 60 min (Supplement 3). The higher dose of NAC (2 mM) did not show a strong effect on the decrease of H2 O2 levels compared with other lower doses of NAC (100 and 500 ␮M) (Supplement 3). Interestingly, treatment with NAC (500 ␮M and 2 mM) increased H2 O2 levels in control cells while Npropylgallate decreased the H2 O2 levels in the early time stage (Supplement 3). Treatment with Trimetazidine (500 ␮M) and NAC (100 ␮M, 500 ␮M and 2 mM) inhibited an increase in O2 •− levels in 50 ␮M AMA-treated cells from the early time point of 2 min (Supplement 4). Other scavengers of Tempol and Tiron decreased the levels of O2 •− in AMA-treated cells at 60 min while N-propylgallate strongly augmented the levels of O2 •− (Supplement 4).

74

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

Furthermore, we investigated whether calpains I and II inhibitor (calpain inhibitor III) and Ca2+ -chelating agent (EGTA/AM) inhibit an increased ROS levels in AMA-treated HeLa cells. Both of them significantly reduced H2 O2 levels in AMA-treated or -untreated HeLa cells (Fig. 3A). Treatment with calpain inhibitor III intensified the levels of O2 •− in AMA-treated or -untreated HeLa cells while EGTA/AM slightly reduced the levels of O2 •− in AMA-treated cells (Fig. 3B). 3.3. Effects of ROS scavengers, calpain inhibitor and Ca2+ -chelating agent on GSH depletion in AMA-treated HeLa cells To determine whether GSH depletion in AMA-treated HeLa cells was altered by ROS scavengers, calpain inhibitor and Ca2+ -chelating agents, these agents were co-incubated with the AMA-treated HeLa cells for 72 h. The scavengers did not inhibit the depletion of GSH content (negative CMF fluorescence) in AMAtreated HeLa cells (Fig. 3C and Supplement 2C (M1 region cells)). Unexpectedly, treatment with NAC and Npropylgallate showed a significant enhancement of GSH depletion by AMA (Fig. 3C). Calpain inhibitor III also strongly depleted GSH content in AMA-treated cells. N-propylgallate and calpain inhibitor III depleted GSH content in the control HeLa cells (Fig. 3C). Next, we evaluated whether the M1 population of cells in the negative CMF fluorescence region were alive or dead. As shown in Supplements 5A and B, the negative CMF fluorescence cells mostly showed PI-positive staining, which supported that cells showing GSH depletion were generally dead. In addition, the number of PI negative cells was small among the negative CMF fluorescence cells (Supplements 5A and C). Interestingly, the proportion of CMF negative and PI-positive cells in NAC-treated cells was significantly increased, indicating that NAC exaggerated the disruption of the plasma membrane.

inhibitor III and EGTA/AM strongly reduced the viability of AMA-treated cells (Fig. 4). N-Propylgallate and calpain inhibitor III decreased the viability of HeLa control cells (Fig. 4). Suppression of HeLa cell growth or viability by AMA and ROS scavengers can be explained by cell cycle arrest. As shown in Supplement 6, DNA flow cytometric analysis indicated that 50 ␮M AMA markedly increased the population of S phase cells compared with AMA-untreated control cells (approximately 40% versus 23%) for 72 h. None of the scavengers (Tempol, Tiron, Trimetazidine and NAC) significantly altered the cell cycle distribution in AMA-treated cells. Next, we examined whether ROS scavengers, calpain inhibitor and Ca2+ -chelating agent prevent AMAinduced HeLa cell death. Before doing this experiment, we confirmed that AMA induced apoptosis by observation that AMA increased the activity of caspase-3 and

3.4. Effects of ROS scavengers, calpain inhibitor and Ca2+ -chelating agent on AMA-treated HeLa cell growth We examined the effect of the AMA and ROS scavengers, calpain inhibitor and Ca2+ -chelating agent on the viability of HeLa cells using an MTT assay. Cell viability was reduced by about 50% in HeLa cells after treatment with 50 ␮M AMA for 72 h (Fig. 4). While ROS scavengers, Tempol, Tiron, and Trimetazidine did not effect on the viability of AMA-treated cells, NAC and N-propylgallate significantly intensified the inhibition of cell viability (Fig. 4). In addition, treatment with calpain

Fig. 5. Effects of ROS scavengers, calpain inhibitor and Ca2+ chelating agent (EGTA/AM) on AMA-induced apoptosis in HeLa cells. Exponentially growing cells were treated with the indicated amounts of ROS scavengers, calpain inhibitor and Ca2+ -chelating agent in addition to AMA for 72 h. Sub-G1 cells (A), annexin Vstaining cells (B) and Rhodamine 123 negative staining cells (C) were measured with a FACStar flow cytometer. *, P < 0.05 compared with the control group and #, P < 0.05 compared with the cells treated with only AMA (ANOVA test).

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

caspase-8 in HeLa cells (Supplement 7). None of the scavengers were able to prevent apoptosis induced by AMA in HeLa cells (Fig. 5 and supplement 8). In contrast, NAC and N-propylgallate significantly increased apoptosis induced by AMA in view of sub-G1 cells and annexin V-staining cells (Fig. 5A and B and Supplements 8A and B). Additionally, the higher concentration (2.5 mM) of ROS scavengers could not significantly change the levels of apoptosis compared with 500 ␮M ROS scavengers (data not shown). Treatment with calpain inhibitor III and EGTA/AM significantly enhanced apoptosis in AMA-treated cells (Fig. 5A and B). In addition, N-propylgallate and calpain inhibitor III induced apoptosis in HeLa control cells (Fig. 5A and B). We also examined whether ROS scavengers, calpain inhibitor and Ca2+ -chelating agent prevent the loss of mitochondrial transmembrane potential (Ψ m ) induced by AMA. None of the scavengers attenuate the loss of mitochondrial membrane potential (Ψ m ) in AMA-treated cells (Fig. 5C and Supplement 8C). Interestingly, NAC showing the augmentation of apoptosis in AMA-treated HeLa cells did not significantly amplify the loss of mitochondrial membrane potential (Ψ m ) (Fig. 5C and Supplement 8C). In contrast, treatment with N-propylgallate intensified the loss of mitochondrial membrane potential (Ψ m ) in AMA-treated or -nontreated HeLa cells (Fig. 5C). Calpain inhibitor III also intensified the loss of mitochondrial membrane potential (Ψ m ) in AMA-treated or -nontreated HeLa cells (Fig. 5C). 4. Discussion In this study, we focused on the involvement of GSH and ROS such as H2 O2 and O2 •− in AMA-induced HeLa cell death, investigated whether ROS scavengers calpain inhibitor and Ca2+ -chelating agent rescue cells from AMA-induced apoptosis and studied its mechanism, since we have recently observed that AMA inhibited the growth of HeLa cells with an IC50 of about 50 ␮M and efficiently induced apoptosis [26]. In this experiment, we used a concentration of 50 ␮M AMA because this concentration was considered to be suitable dose to differentiate the levels of apoptosis in the presence of AMA versus apoptosis in the presence of AMA and ROS scavengers. Our data showed that the intracellular H2 O2 and O2 •− levels were significantly increased in AMA-treated HeLa cells at 72 h. In fact, 50 ␮M AMA efficiently induced apoptosis in HeLa cells, as evidenced by sub-G1 cells and annexin V-staining cells and this concentration of AMA was an IC50 in the growth of HeLa cells using

75

an MTT assay in the previous data [26]. These results are consistent with other reports, which showed that increased intracellular H2 O2 plays an important role in AMA-induced cell death in liver cells [37,38], A549 human lung cancer cells [7] and As4.1 juxtaglomerular cells [11]. An increase in O2 •− levels by AMA was also reported in human lung epithelial cells [39] and As4.1 juxtaglomerular cells [11]. These data suggest that the apoptotic effects of AMA are comparative to intracellular ROS levels. Conversely, our other results indicated that the changes of ROS by AMA were not entirely correlated with apoptosis in HeLa cells, since 100 and 200 ␮M AMA, which showed strong apoptosis (data not shown), did not show an increased level of ROS compared to that in 50 ␮M AMA-treated cells. However, we cannot exclude the possibility that no increase in ROS levels by 100 and 200 ␮M AMA resulted from a leakage of the dye from the cells. We have also shown that NAC known as an antioxidant agent intensified apoptosis without an increase of ROS levels in AMA-treated HeLa cells. Treatment with N-propylgallate augmented the levels of apoptosis in AMA-treated cells with strongly reducing H2 O2 levels and enhancing O2 •− levels. Therefore, the exact cell death mechanisms through intracellular ROS in AMA-treated HeLa cells must be defined further. We attempted to determine whether ROS scavengers prevent AMA-induced cell death by reducing intracellular ROS levels. In contrast to our expectation, none of the scavengers (Tempol, Tiron, Trimetazidine and NAC) significantly reduced the level of ROS in HeLa cells treated with 50 ␮M AMA at 72 h. In addition, these scavengers could not reduce apoptosis in AMA-treated cells at 72 h. It is possible that ROS scavengers did not prevent cell death because the dose of 50 ␮M AMA is too high. However, we could not observe the anti-apoptotic effects of these ROS scavengers on the experiment at a lower dose of AMA (data not shown). Interestingly, recent studies suggest that some substances that have been regarded as antioxidant agents, such as ascorbic acid and tannic acid, can stimulate the generation of ROS and show a pro-oxidant effect under certain circumstances [40,41]. In our data, NAC showing the significant augmentation of apoptosis in AMA-treated HeLa cells did not significantly amplify the loss of mitochondrial membrane potential (Ψ m ) and alter the ROS levels. In addition, we observed that NAC intensified apoptosis accompanied with an increase in O2 •− levels in AMA-treated As4.1 juxtaglomerular cells [11] and Calu-6 lung cancer cells (unpublished data). These results suggest that NAC plays a role as a pro-apoptotic agent or an oxidant rather than as an anti-apoptotic agent or an antioxidant in AMA-treated

76

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

cells. In contrast, we could observe that ROS scavengers including NAC and N-propylgallate inhibited an increase in H2 O2 levels in 50 ␮M AMA-treated cells at the early time points (20, 30 and 60 min) and the ROS scavengers except N-propylgallate reduced O2 •− levels in 50 ␮M AMA-treated cells at the early time point (60 min). Treatment with N-propylgallate intensified O2 •− levels in 50 ␮M AMA-treated cells at an early and late time. These results suggest the possibility that ROS scavengers except N-propylgallate have a reduced effect on ROS levels in AMA-treated cells at the only early time point. With regard to intracellular GSH, a main nonprotein antioxidant in the cell, it was able to clear away the superoxide anion free radical and provide electrons for enzymes such as glutathione peroxidase, which reduces H2 O2 to H2 O. It has been reported that the intracellular GSH content has a decisive effect on anticancer druginduced apoptosis, indicating that apoptotic effects are inversely comparative to GSH content [42,43]. Likewise, our result clearly indicated the depletion and reduction of intracellular GSH content by AMA in HeLa cells. We have shown that the negative CMF fluorescence (GSH depletion) cells were all dead, as evidenced by PI-positive staining (cell viability). These results support that intracellular GSH levels are tightly related to a decisive factor in AMA-induced cell death. In fact, NAC and N-propylgallate showing a pro-apoptotic effect in AMA-treated cells augmented GSH depletion and increased PI-positive staining in AMA-treated HeLa cells. It is of note that NAC as a well-known thiol-donor could not recover the depletion of GSH in AMA-treated HeLa, As4.1 cells [11] and Calu-6 cells (unpublished data). These data also supported the notion that NAC plays a role as a pro-apoptotic agent or an oxidant in AMA-treated cells. In view of the cell cycle distribution by AMA and ROS scavengers, ROS scavengers could not alter S phase accumulation during the cell cycle induced by AMA, indicating that the changes in ROS levels and GSH content by these scavengers are not tightly correlated with the cell cycle regulation in HeLa cells. It has been known that AMA produces rapid ATP depletion that correlated to rapid and sustained increased in cytosolic free Ca2+ [21] and an increase in cytosolic free Ca2+ mediate cell death following anoxia/hypoxia and chemical exposure [22,23]. The prolonged increases in cytosolic free Ca2+ activate degradative enzymes such as calpains, Ca2+ -activated cysteine proteases, which have a number of substrates including cytoskeletal proteins, procaspase-3 and PARP [24,25]. Therefore, it is worthy to investigate the effect of cytosolic free Ca2+ on

AMA-treated HeLa cells. We used calpain inhibitors I and II (calpain inhibitor III) and Ca2+ -chelating agent (EGTA/AM) to know the role of cytosolic free Ca2+ in AMA-treated HeLa cells although we did not assess the changes of cytosolic free Ca2+ and the degradation of calpains in these cells. According to our data, calpain inhibitors I and II (calpain inhibitor III) and Ca2+ -chelating agent (EGTA/AM) significantly reduced H2 O2 levels in AMA-treated HeLa cells. However, treatment with calpain inhibitor III intensified the levels of O2 •− in AMA-treated while EGTA/AM slightly reduced the levels of O2 •− in AMA-treated cells. In addition, both of them did not inhibit the levels of apoptosis in AMA-treated cells. These data supported the previous suggestion that the changes of ROS by AMA were not entirely correlated with apoptosis in HeLa cells. Calpain inhibitor III strongly depleted GSH content with an enhancement of apoptosis in AMA-treated cells. Treatment with EGTA/AM mildly depleted GSH content without a significant increase in apoptosis in view of annexin V staining and the loss of mitochondrial membrane potential (Ψ m ) in AMA-treated cells. These results also support that intracellular GSH levels are tightly related to AMA-induced cell death. These data related to effect of cytosolic free Ca2+ on AMAtreated HeLa cells suggest that AMA does not increase the cytosolic free Ca2+ in HeLa cells or the increased cytosolic free Ca2+ by AMA induce apoptosis via a calpain-independent manner. We cannot rule out the possibility that calpains I and II inhibitor (calpain inhibitor III) can trigger apoptosis regardless of the activation of calpain, because calpain inhibitor III alone strongly caused apoptosis and this agent intensified the loss of mitochondrial membrane potential (Ψ m ) in control HeLa cells. In summary, we have demonstrated that AMA potently generates ROS, H2 O2 and O2 •− , and induces the depletion and reduction of GSH content in HeLa cells. ROS scavengers did not prevent the apoptosis triggered by AMA. It seems that AMA does not increase the cytosolic free Ca2+ in HeLa cells or the increased cytosolic free Ca2+ by AMA induce apoptosis via a calpain-independent manner. Conclusively, the changes of ROS by AMA were not tightly correlated with apoptosis in HeLa cells. However, intracellular GSH levels are tightly related to AMA-induced cell death.

Conflict of interest None declared.

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

Acknowledgements This research was supported by the Korean Science and Engineering Foundation (R01-2006-000-10544-0) and the Korea Research Foundation Grant funded by the Korean Government (MOEHRD). Contract grant sponsor: The Korea Research Foundation Grant funded by the Government of the Republic of Korea (MOEHRD) and the Korean Science and Engineering Foundation Contract grant number: R012006-000-10544-0. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j. cbi.2007.08.011. References [1] K. Nakayama, F. Okamoto, Y. Harada, Antimycin A: isolation from a new Streptomyces and activity against rice plant blast fungi, J. Antibiot. 9 (2) (1956) 63–66. [2] M.L. Campo, K.W. Kinnally, H. Tedeschi, The effect of antimycin A on mouse liver inner mitochondrial membrane channel activity, J. Biol. Chem. 267 (12) (1992) 8123–8127. [3] J.J. Maguire, V.E. Kagan, L. Packer, Electron transport between cytochrome c and alpha tocopherol, Biochem. Biophys. Res. Commun. 188 (1) (1992) 190–197. [4] N.A. Pham, B.H. Robinson, D.W. Hedley, Simultaneous detection of mitochondrial respiratory chain activity and reactive oxygen in digitonin-permeabilized cells using flow cytometry, Cytometry 41 (4) (2000) 245–251. [5] A. Alexandre, A.L. Lehninger, Bypasses of the antimycin a block of mitochondrial electron transport in relation to ubisemiquinone function, Biochim. Biophys. Acta 767 (1) (1984) 120–129. [6] R.S. Balaban, S. Nemoto, T. Finkel, Mitochondria, oxidants, and aging, Cell 120 (4) (2005) 483–495. [7] V. Panduri, S.A. Weitzman, N.S. Chandel, D.W. Kamp, Mitochondrial-derived free radicals mediate asbestos-induced alveolar epithelial cell apoptosis, Am. J. Physiol. 286 (6) (2004) L1220–L1227. [8] V. Petronilli, D. Penzo, L. Scorrano, P. Bernardi, F. Di Lisa, The mitochondrial permeability transition, release of cytochrome c and cell death. Correlation with the duration of pore openings in situ, J. Biol. Chem. 276 (15) (2001) 12030–12034. [9] P. Costantini, B.V. Chernyak, V. Petronilli, P. Bernardi, Modulation of the mitochondrial permeability transition pore by pyridine nucleotides and dithiol oxidation at two separate sites, J. Biol. Chem. 271 (12) (1996) 6746–6751. [10] J.G. Pastorino, M. Tafani, R.J. Rothman, A. Marcinkeviciute, J.B. Hoek, J.L. Farber, Functional consequences of the sustained or transient activation by Bax of the mitochondrial permeability transition pore, J. Biol. Chem. 274 (44) (1999) 31734–31739. [11] Y.W. Han, S.Z. Kim, S.H. Kim, W.H. Park, The changes of intracellular H2O2 are an important factor maintaining mitochondria membrane potential of antimycin A-treated As4.1 juxtaglomerular cells, Biochem. Pharmacol. 73 (6) (2007) 863–872.

77

[12] M.A. King, Antimycin A-induced killing of HL-60 cells: apoptosis initiated from within mitochondria does not necessarily proceed via caspase 9, Cytometry A 63 (2) (2005) 69–76. [13] E.J. Wolvetang, K.L. Johnson, K. Krauer, S.J. Ralph, A.W. Linnane, Mitochondrial respiratory chain inhibitors induce apoptosis, FEBS Lett. 339 (1–2) (1994) 40–44. [14] G.P. Kaushal, N. Ueda, S.V. Shah, Role of caspases (ICE/CED 3 proteases) in DNA damage and cell death in response to a mitochondrial inhibitor, antimycin A, Kidney Int. 52 (2) (1997) 438–445. [15] A.O. de Graaf, J.P. Meijerink, L.P. van den Heuvel, R.A. DeAbreu, T. de Witte, J.H. Jansen, J.A. Smeitink, Bcl-2 protects against apoptosis induced by antimycin A and bongkrekic acid without restoring cellular ATP levels, Biochim. Biophys. Acta 1554 (1–2) (2002) 57–65. [16] W.H. Park, Y.W. Han, S.W. Kim, S.H. Kim, K.W. Cho, S.Z. Kim, Antimycin, A induces apoptosis in As4.1 juxtaglomerular cells, Cancer Lett. 251 (1) (2007) 68–77. [17] D.B. Zorov, M. Juhaszova, S.J. Sollott, Mitochondrial ROSinduced ROS release: an update and review, Biochim. Biophys. Acta 1757 (5–6) (2006) 509–517. [18] G. Dasmahapatra, M. Rahmani, P. Dent, S. Grant, The tyrphostin adaphostin interacts synergistically with proteasome inhibitors to induce apoptosis in human leukemia cells through a reactive oxygen species (ROS)-dependent mechanism, Blood 107 (1) (2006) 232–240. [19] S.B. Wallach-Dayan, G. Izbicki, P.Y. Cohen, R. Gerstl-Golan, A. Fine, R. Breuer, Bleomycin initiates apoptosis of lung epithelial cells by ROS but not by Fas/FasL pathway, Am. J. Physiol. 290 (4) (2006) L790–L796. [20] H.U. Simon, A. Haj-Yehia, F. Levi-Schaffer, Role of reactive oxygen species (ROS) in apoptosis induction, Apoptosis 5 (5) (2000) 415–418. [21] J.F. Harriman, X.L. Liu, M.D. Aleo, K. Machaca, R.G. Schnellmann, Endoplasmic reticulum Ca(2+) signaling and calpains mediate renal cell death, Cell Death Differ. 9 (7) (2002) 734– 741. [22] T. Jiang, R.L. Grant, D. Acosta, A digitized fluorescence imaging study of intracellular free calcium, mitochondrial integrity and cytotoxicity in rat renal cells exposed to ionomycin, a calcium ionophore, Toxicology 85 (1) (1993) 41–65. [23] A. Kribben, E.D. Wieder, J.F. Wetzels, L. Yu, P.E. Gengaro, T.J. Burke, R.W. Schrier, Evidence for role of cytosolic free calcium in hypoxia-induced proximal tubule injury, J. Clin. Invest. 93 (5) (1994) 1922–1929. [24] D.E. Croall, G.N. DeMartino, Calcium-activated neutral protease (calpain) system: structure, function, and regulation, Physiol. Rev. 71 (3) (1991) 813–847. [25] K.M. McGinnis, M.E. Gnegy, Y.H. Park, N. Mukerjee, K.K. Wang, Procaspase-3 and poly(ADP)ribose polymerase (PARP) are calpain substrates, Biochem. Biophys. Res. Commun. 263 (1) (1999) 94–99. [26] W.H. Park, Y.W. Han, S.H. Kim, S.Z. Kim, An ROS generator, antimycin A, inhibits the growth of HeLa cells via apoptosis, J. Cell. Biochem. 102 (1) (2007) 98–109. [27] A. Macho, T. Hirsch, I. Marzo, P. Marchetti, B. Dallaporta, S.A. Susin, N. Zamzami, G. Kroemer, Glutathione depletion is an early and calcium elevation is a late event of thymocyte apoptosis, J. Immunol. 158 (10) (1997) 4612–4619. [28] D.W. Hedley, S. Chow, Evaluation of methods for measuring cellular glutathione content using flow cytometry, Cytometry 15 (4) (1994) 349–358.

78

Y.H. Han et al. / Chemico-Biological Interactions 171 (2008) 67–78

[29] W.H. Park, J.G. Seol, E.S. Kim, J.M. Hyun, C.W. Jung, C.C. Lee, B.K. Kim, Y.Y. Lee, Arsenic trioxide-mediated growth inhibition in MC/CAR myeloma cells via cell cycle arrest in association with induction of cyclin-dependent kinase inhibitor, p21, and apoptosis, Cancer Res. 60 (11) (2000) 3065– 3071. [30] W.H. Park, C.W. Jung, J.O. Park, K. Kim, W.S. Kim, Y.H. Im, M.H. Lee, W.K. Kang, K. Park, Trichostatin inhibits the growth of ACHN renal cell carcinoma cells via cell cycle arrest in association with p27, or apoptosis, Int. J. Oncol. 22 (5) (2003) 1129–1134. [31] M. Poot, H. Teubert, P.S. Rabinovitch, T.J. Kavanagh, De novo synthesis of glutathione is required for both entry into and progression through the cell cycle, J. Cell. Physiol. 163 (3) (1995) 555–560. [32] T. Schnelldorfer, S. Gansauge, F. Gansauge, S. Schlosser, H.G. Beger, A.K. Nussler, Glutathione depletion causes cell growth inhibition and enhanced apoptosis in pancreatic cancer cells, Cancer 89 (7) (2000) 1440–1447. [33] J. Yamada, S. Yoshimura, H. Yamakawa, M. Sawada, M. Nakagawa, S. Hara, Y. Kaku, T. Iwama, T. Naganawa, Y. Banno, S. Nakashima, N. Sakai, Cell permeable ROS scavengers, Tiron and Tempol, rescue PC12 cell death caused by pyrogallol or hypoxia/reoxygenation, Neurosci. Res. 45 (1) (2003) 1–8. [34] S. Gorsdorf, K.E. Appel, G. Obe, A.G. Hildebrandt, Inhibition of bleomycin-induced DNA strand breaks in V 79 Chinese hamster cells by the antioxidant propylgallate, Toxicol. Lett. 38 (1–2) (1987) 135–143.

[35] A.K. Tikhaze, V.Z. Lankin, E.A. Zharova, S.V. Kolycheva, Trimetazidine as indirect antioxidant, Bull. Exp. Biol. Med. 130 (10) (2000) 951–953. [36] W.C. Stanley, M. Marzilli, Metabolic therapy in the treatment of ischaemic heart disease: the pharmacology of trimetazidine, Fund. Clin. Pharmacol. 17 (2) (2003) 133–145. [37] H.M. Chen, H.H. Ma, X.J. Yan, Inhibitory effect of agarohexaose on antimycin A induced generation of reactive oxygen species, Yao xue xue bao (Acta Pharm. Sin.) 40 (10) (2005) 903–907. [38] H.M. Chen, X.J. Yan, Antioxidant activities of agarooligosaccharides with different degrees of polymerization in cell-based system, Biochim. Biophys. Acta 1722 (1) (2005) 103–111. [39] C. Li, M.M. Wright, R.M. Jackson, Reactive species mediated injury of human lung epithelial cells after hypoxia-reoxygenation, Exp. Lung Res. 28 (5) (2002) 373–389. [40] J.J. Li, Q. Tang, Y. Li, B.R. Hu, Z.Y. Ming, Q. Fu, J.Q. Qian, J.Z. Xiang, Role of oxidative stress in the apoptosis of hepatocellular carcinoma induced by combination of arsenic trioxide and ascorbic acid, Acta Pharmacol. Sin. 27 (8) (2006) 1078–1084. [41] N.S. Khan, A. Ahmad, S.M. Hadi, Anti-oxidant, pro-oxidant properties of tannic acid and its binding to DNA, Chem.-Biol. Interact. 125 (3) (2000) 177–189. [42] J.M. Estrela, A. Ortega, E. Obrador, Glutathione in cancer biology and therapy, Crit. Rev. Clin. Lab. Sci. 43 (2) (2006) 143–181. [43] Y. Higuchi, Glutathione depletion-induced chromosomal DNA fragmentation associated with apoptosis and necrosis, J. Cell. Mol. Med. 8 (4) (2004) 455–464.