Protection against arsenic trioxide-induced autophagic cell death in U118 human glioma cells by use of lipoic acid

Protection against arsenic trioxide-induced autophagic cell death in U118 human glioma cells by use of lipoic acid

Tain-Junn Cheng

, Ying-Jan Wang

, Wei-Wan Kao

, Rong-Jane Chen

, Yuan-Soon Ho

aChimei Medical Center Tainan, Taiwan

bDepartment of Environmental and Occupational Health, National Cheng Kung University Medical College, 138 Sheng-Li Road, Tainan 704, Taiwan

cGraduate Institute of Biomedical Technology, Taipei Medical University, No. 250 Wu-Hsing Street, Taipei 110, Taiwan Received 25 November 2005; accepted 12 December 2006

Abstract

Arsenic is an environmental toxicant found naturally in ground water. Epidemiological studies have suggested a correlation between chronic arsenic exposure and potential brain tissue damage in clinical case and animal experiments. Lipoic acid (LA) is a thiol-compound naturally occurring in plants and animals, which is thought to be a strong antioxidant and possess neuroprotective effects. The objective of this study was to determine if the AS

O

-induced glial cell toxicity could be prevented by LA. The human malignant glioma cell (U118) was selected as a research model. By using acridine orange staining and flow cytometry analysis, we found that autophagic, but not apoptotic, cell death was significantly induced by AS

O

in U118 cells, and that AS

O

-mediated autophagic cell death was nearly completely attenuated by LA. Down-regulation of p53 and Bax proteins and the up-regulation of Bcl-2 and HSP-70 proteins were observed by western blot in AS

O

-mediated autophagic cell death. Our results implied that LA completely inhibited U118 cells auto- phagic cell death induced by AS

O

. We suggested that LA may emerge as a useful protective agent against arsenic-induced glial cell toxicity and reversing arsenic-induced damage in human brain.

 2007 Elsevier Ltd. All rights reserved.

Keywords: Lipoic acid; Arsenic trioxide; Neurotoxicity; Autophage

1. Introduction

Higher doses or chronic exposure to arsenic is still a glo- bal health problem affecting many millions of people (Rat-

naike, 2003). Arsenic at a nonlethal level in drinking water consumed over a period of time may result in the manifesta- tions of toxicity in practically all systems of the body (Hant- son et al., 2003). Neurotoxic effects have been reported in clinical cases and animal experiments with chronic exposure to arsenic (Rao and Avani, 2004). Epidemiological studies also have suggested a correlation between arsenic exposure and potential neurotoxicity (Hall, 2002). For example, higher concentrations of AS

O

were detected in the plasma and cerebrospinal fluid of Alzheimer’s (Basun et al., 1991), mental health burden (Fujino et al., 2004), and Parkinson’s disease (Larsen et al., 1981) patients.

Lipoic acid (LA) is a thiol-compound naturally occurring in plants and animals (Sohal et al., 1994). It is consumed in the daily diet, absorbed through the blood-brain barrier, and taken up and transformed in cells and tissues into

0278-6915/$ - see front matter  2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.fct.2006.12.014

Abbreviations: AS2O3; arsenic trioxide; DHLA, dihydrolipoic acid;

ERK, extracellular regulated kinases; FACS, fluorescence-activated cell sorter; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HSPs, heat shock proteins; JNK, c-jun terminal kinase; LA, lipoic acid; MAP, mitogen-activated protein; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyl-2H-tetrazolium bromide; ROS, reactive oxygen species; (ST), staurosporine.

* Corresponding authors. Tel.: +886 6 235 3535x5804; fax: +886 6 2752484 (Y.-J. Wang). Tel.: +886 2 27361661x33271; fax: +886 2 2739 3422 (Y.-S. Ho).

E-mail addresses: [email protected] (Y.-J. Wang), [email protected](Y.-S. Ho).

www.elsevier.com/locate/foodchemtox

dihydrolipoic acid (DHLA) (Packer et al., 1997). Both LA and DHLA are thought to be strong antioxidants. Aside from acting as a potent antioxidant, LA increases or main- tains levels of other low molecular weight antioxidants such as ubiquinone, glutathione, and ascorbic acid (Kozlov et al., 1999). Therefore, it appears that LA could be a potential agent in the prevention of different diseases that may be related to an imbalance of the oxidoreductive cellular status.

This occurs in cases of neurodegeneration, ischemia-reper- fusion, polyneuropathy, diabetes, AIDS, and hepatic disor- der status (Packer et al., 1995).

Human fetal brain explants exposed to arsenic in tissue culture showed the characteristics of cell death, neuronal network damage, loss of ground matrix, cell loss and apop- tosis in isolated brain cells and neighboring cells. The arsenic toxicity appears to act through interference the tis- sue homeostasis in the brain rather than only affect neuron cells (Chattopadhyay et al., 2002). Other studies also indi- cated that neurological system is the major target of toxic effects of heavy metals such as arsenic (Lee et al., 2001).

Pathological alterations of glial cells have also been indi- cated associated with brain disease. However, the exact mechanism of the toxicity of arsenic in glial cells is not well studied. The objective of this study was to determine if the AS

O

-induced glial cell toxicity could be prevented by LA.

The human malignant glioma cell (U118) was selected as a research model. We found that autophagic, but not apop- totic, cell death was significantly induced by AS

O

in U118 cells, and that AS

O

-mediated autophagic cell death was nearly completely attenuated by LA. Our results pro- vide the molecular basis for the LA prevention of AS

O

- induced cell death in U118 cells; such observations may have significance in clinical application.

2. Material and methods 2.1. Cell line and cell culture

Human glioblastoma cell lines U118MG (ATCC HTB-15) and U937 cells, a human pre-monocytic leukemia cell line, were obtained from the American Type Culture Collection (ATCC). U937 cells were cultured in RPMI 1640 medium supplemented with antibiotics containing 100 U/ml penicillin, 100 lg/ml streptomycin (Life Technology, Grand Island, NY), and 10% heat-inactivated fetal calf serum (FCS) (HyClone, South Logan, UT, USA). U118 cells were cultured in DMEM containing supplement as well as U937 cells with additional nonessential amino acids and MEM sodium pyruvate (Gibco, Grand Island, NY). Incubation both cells were performed in a humidified atmosphere containing 5% CO2 at 37C.

Exponentially growing cells were detached by 0.05% trypsin-EDTA (Gibco) in DMEM supplemented with 10% fetal bovine serum (Hyclone, Logan, UT) and antibiotics (Life Technologies, Grand Island, NY).

2.2. Determination of cell viability

Human U118 cells were treated with AS2O3(1–50 lM) in the presence or absence of LA (50–100 lM) for 24 h. For the time-dependent study, U118 cells were treated with AS2O3(5–10 lM) for 24, 48 and 72 h. Cell viability was determined at the indicated times based on a 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay.

Briefly, cells were seeded in a 96-well plate at a density of 1· 10⁴cells/well

and allowed to adhere overnight. After removing the medium, 200 lL of fresh medium per well, containing 10 mmol/L Hepes (pH 7.4), was added.

Then, 50 lL of MTT was added to the wells and the plate was incubated for 2–4 h at 37C in the dark. The medium was removed, and 200 lL DMSO and 25 lL Sorensens’s glycine buffer was added to the wells.

Absorbance was measured using an ELISA plate reader at 570 nm.

2.3. Western analysis

Proteins isolated from the U118 cells were loaded at 50 lg/lane on 12%

(w/v) sodium dodesylsulfate–polyacrylamide gel electrophoresis, blotted, and probed using antibodies, including anti-caspase-3 (E8), anti-caspase-8 (E20), anti-caspase-9 (H170), cyclin E, p53, p21/Cip1, Bax, Bcl-2, glyc- eraldehyde-3-phosphate dehydrogenase (GAPDH), (Santa Cruz, Inc.

CA), cyclin A, cyclin E, cyclin B, and HSP70 (Transduction Laboratories, Lexington, KY). Immunoreactive bands were visualized by incubation with colorigenic substrates, nitro blue tetrazolium and 5-bromo-4-chloro- 3-indolyl-phosphate (Sigma Chemical Co., St. Louis, MO). The expression of GAPDH was used as the control for equal protein loading.

2.4. Determination of apoptosis

Apoptosis was judged by the following criteria: (a) Cell morphology as described previously (Ho et al., 1996). (b) Translocation of phosphotidyl serine to the cell surface detected by an Annexin V-FITC apoptosis detection kit (Calbiochem, Bad Soden, Germany), according to our pre- vious paper (Liu et al., 2003). The U937 cells were selected as a positive control, and treated with either staurosporine (ST) (1 lM) or AS2O3

(20 lM) for 24 h, then harvested for Annexin V staining assay. (c) The presence of a sub-G1 peak detected by flow cytometry and measured using a fluorescence-activated cell sorter (Becton Dickinson, Heidelberg, Ger- many) (Lee et al., 2003). (d) The appearance of DNA fragmentation analyzed by the method described previously (Ho et al., 1996).

2.5. ROS production measurement

ROS production was monitored by flow cytometry using 2,7-dichlo- rodihydrofluorescein diacetate (DCFH-DA), as described by Zegura (Zegura et al., 2004). This dye is a stable nonpolar compound that readily diffuses into cells and is hydrolyzed by intracellular esterase to yield DCFH, which is trapped within the cells. Hydrogen peroxide or low molecular weight peroxides produced by the cells oxidize DCFH to the highly fluorescent compound 2⁰,7⁰-dichlorofluorescein (DCF); thus, the fluorescence intensity is proportional to the amount of peroxide produced by the cells. The cells were incubated with 20 lM DCFH-DA. After 30 min, DCFH-DA was removed and the cells were treated with AS2O3

(5 lM) in PBS for 0–60 min. H2O2 was added to the U118 cells and incubated for 60 min as a positive control.

2.6. Supravital cell staining with acridine orange for autophagy detection

Cell staining with Acridine orange (Sigma Chemical Co.) was per- formed according to published procedures (Kanzawa et al., 2003; Trag- anos and Darzynkiewicz, 1994), adding a final concentration of 1 mg/ml for a period of 20 min. AS2O3(10 lM) was dissolved in DMSO and added to the cells 30 min before the addition of acridine orange. Photographs were obtained with a fluorescence microscope (Axioscop) equipped with a mercury 100-W lamp, 490-nm band-pass blue excitation filters, a 500-nm dichroic mirror, and a 515-nm long-pass barrier filter. Flow cytometric analysis is also available to detect AVO percentage (Kanzawa et al., 2003).

2.7. Statistical analysis

Values are expressed as the mean ± S.E. The significance of the dif- ference of the respective controls for each experimental test condition was 1028 T.-J. Cheng et al. / Food and Chemical Toxicology 45 (2007) 1027–1038

assayed using an unpaired Student’s t-test comparing each data point to independent groups. A p value <0.05 was considered to be significant.

3. Results

In this study, U118 cells were treated with AS

O

at dif- ferent concentrations (1–50 lM) for 24 h, and the viability of the cells was determined. Fig. 1a showed that the viabil- ity of U118 cells was less than 60% at 24 h after exposure of the cells to AS

O

(5 lM). In Fig. 1b, the cytotoxic effects of the exposure of U118 cells to 5 and 10 lM AS

O

were examined in a time-dependent manner. Fig. 1b also indi- cated that, As

O

inhibited the viability of cells in a dose- and time-dependent manner. The number of viable cells treated with As

O

(5 or 10 lM) for 24–72 h decreased to a level below the initial cell number (5 · 10

) (Fig. 1b).

These results indicate that treatment with As

O

not only inhibited cell viability, but also induced cell death in U118 cells.

The morphological changes of the U118 cells treated with AS

O

(5–25 lM) in the presence or absence of LA (50 lM) were illustrated in Fig. 2. Under stimulating con- dition, the morphology of U118 cells becoming rounded.

Interestingly, the morphological changes of the AS

O

- induced cytotoxic effects were nearly completely reversed by LA treatment in the U118 cells (Fig. 2a). The viability assays revealed that the 50% mortality of the AS

O

-trea- ted cells that died within 24 h was reversed by 50 lM LA (Fig. 2b). The LA (>50 lM) treatment promoted long-term survival in cells treated with AS

O

(5 lM), and LA provided more than 80% protection during days 2–3.

In Fig. 3a, DNA smearing instead of fragmentation was observed in human U118 cells, using AS

O

treatment in a dose-dependent manner. Flow cytometric analysis was fur- ther performed to confirm the apoptotic cells with Annexin V-PI staining. Our results demonstrated that apoptosis was not induced even by higher dose of AS

O

(50 lM) treat- ment in U118 cells (Fig. 3b). However, significant apopto- sis were induced in U118 cells treated with apoptosis inducer staurosporine (ST), a highly potent, nonspecific

inhibitor of PKC, which degrades DNA to oligonucleoso- mal fragments for 24 h. The model of U937 cells treated with AS

O

(Nolte et al., 2004) and ST (Grant et al., 1994) was served as apoptosis positive controls, further confirmed that AS

O

could induce apoptosis in U937 cells but not in U118 cells (Fig. 3c).

As described in the previous report, significant ROS production was the major mechanism for the induction of apoptosis in different types of human cancer cells (Chun et al., 2002; Gao et al., 2002; Maeda et al., 2001; Shen et al., 2003). To determine whether ROS production could be induced by AS

O

in U118 cells, the ROS level was then detected at different doses (5–25 lM) and time periods (0–120 min). Our results demonstrated that the ROS level in human U118 cells did not change in response to AS

O

within 2 h (Fig. 3d). It has been demonstrated that AS

O

-induced apoptosis in U937 cells was mediated through the production of H

O

(Park et al., 2003). Thus, the U118 cells treated with H

O

were adapted as a positive control for the determination of ROS (Woo et al., 2002).

The results in Fig. 4a revealed that AS

O

-induced DNA degradation was nearly completely reversed by LA pretreatment. Flow cytometric analysis was further per- formed to evaluate the percentage of DNA-damaged (sub-G1) cells induced by AS

O

, and the protective effects of LA during a 24–48 h exposure (Fig. 4b). Our results showed that AS

O

-induced sub-G1 cell formation was sig- nificantly attenuated, from 19.78 ± 0.18% to 4.34 ± 0.23%, in response to 50 lM LA for 24 h. Similar results were also observed in a long-term (48 h) experiment (Fig. 4b).

To determine the regulatory mechanisms of AS

O

- induced cytotoxicity, U118 cells were treated with AS

O

for 24 h in a dose-dependent (1–25 lM) manner. The regu- latory proteins which were related to cell cycle and apopto- sis were then determined by immunoblotting analysis (Fig. 5). As described previously, the AS

O

-induced apop- tosis was seen as an important mechanism for cancer che- motherapeutic purposes when applied in different human cancer cells. The p53 (Filippova and Duerksen-Hughes, 2003), Bax (Karlsson et al., 2004), and caspases 3, 8, and

Fig. 1. AS2O3is cytotoxic to U118 cells. (a) For a determination of the dose-dependent cytotoxic effects of AS2O3in U118 cells, cells were treated with different concentrations (1–50 lM) of AS2O3for 24 h. The viability of cells under different treatments was detected by MTT assay, and expressed as the percentage of control (survival of control). (b) Time-dependent effects of the AS2O3-induced cytotoxicity in U118 cells. Cells were treated with AS2O3(5 and 10 lM) for 24, 48 and 72 h, and the viability of cells was analyzed. Data were expressed as the mean ± S.E.^*p < 0.05 indicated a significant difference in AS2O3treated groups compared with untreated group.

9 (Jiang et al., 2001) were induced in AS

O

-induced apop- tosis. However, in our study, the p53 and Bax proteins were down-regulated while Bcl-2 was up-regulated in the U118 cells treated with AS

O

in a dose-dependent manner (Fig. 5). In addition, caspases 3, 8 and 9 were not activated by AS

O

, although PARP, the substrate of caspase 3, was

degraded in the 25 lM AS

O

-treated cells. A previous study demonstrated that the cell cycle arrest of the G0/

G1 and G2/M phases was induced by AS

O

(2 lM, for 72 h) in human myeloma cells. The expression level of cyclin A was inhibited, whereas the CDK inhibitor p21 was induced (Park et al., 2000). In this study, the protein

Fig. 2. LA protection of U118 cells from AS2O3-induced cell death. (a) Phase contrast photomicrograph observations on the morphological changes of U118 cells. Human U118 cells were treated with AS2O3(5–25 lM) in the presence or absence of LA (50 lM) for 24 h, whereas the cells treated with DMSO (0.05%) were the control. (b) LA inhibition of AS2O3-induced cell death by MTT assay. Cells were treated with 5 lM of AS2O3in the presence of LA (50 and 100 lM) treatment for 24 h. The cellular viability was detected by MTT assay as described in Section2. (c) The AS2O3-induced cell death was attenuated by LA treatment in a time-dependent manner. Cells were treated with AS2O3(5 lM) in the presence or absence of LA. The cellular viability was detected by MTT assay at the indicated time points. Data are expressed as the mean ± S.E.^*p < 0.05 indicated a significant difference between AS2O3- treated and combine-treated groups, as analyzed by unpaired Student’s t-test.

1030 T.-J. Cheng et al. / Food and Chemical Toxicology 45 (2007) 1027–1038

levels of cyclins A, E, and p21 were not changed, even by treatment with 25 lM AS

O

, but cyclin B1 was decreased in a dose-dependent manner.

As described above, the cytotoxicity induced by AS

O

was nearly completely prevented by LA (Figs. 2 and 4).

To further investigate the molecular mechanisms of the

LA affect on AS

O

-induced cytotoxicity, U118 cells were treated with AS

O

(5 lM) in the presence of LA (50 lM) for 24 h. The cell lysates were isolated, and Wes- tern blotting analysis was then performed (Fig. 6). Our results, as presented in Fig. 5, revealed that Bcl-2 and HSP70 were up-regulated in the AS

O

(5 lM)-treated

Fig. 3. Apoptotic cell death and intracellular peroxide levels were not induced by AS2O3in U118 cells. (a) U118 cells were treated with AS2O3(1–50 lM) for 24 h. Cells were harvested and DNA degradation was determined by gel electrophoresis (b). (c) The apoptotic effect of the U118 cells treated with AS2O3 (5–50 lM) or staurosporine (ST) for the indicated time points was monitored by flow cytometric analysis with Annexin V-FITC apoptosis detection kit. Human U937 cells were treated either with 10 lM AS2O3(Nolte et al., 2004) or 50 nM ST (Grant et al., 1994) for 24 h as a positive control.

(d) U118 cells were treated with AS2O3at difference doses (5–15 lM) and time period (0–120 min). For a positive control (Gamalei et al., 1998), H2O2

(400 lM) was added to the culture medium for 60 min. The level of intracellular peroxide in cells was measured by flow cytometric analysis using DCHF- DA as a fluorescence dye. Data are derived and counted from three independent experiments.^*p < 0.05.

U118 cells. The increased Bcl-2 and HSP70 level was signif- icantly reversed by LA (50 lM) treatment when U118 cells were exposed to AS

O

(Fig. 5). These results differed from those of a previous report which indicated that HSP70 was inhibited in AS

O

-induced apoptosis in human U937 pre- monocytic cells in which the down-regulated-HSP70 was attenuated by Bcl-2 over-expression (Ramos et al., 2005).

In addition, Fig. 5 demonstrated that p53 and Bax protein expressions were inhibited by AS

O

treatment in U118 cells. Such effects were also reversed by LA treatment when U118 cells were exposed to AS

O

. Moreover, pretreat- ment with protein synthesis inhibitor cyclohexamide (CXM) blocked the autophagy indicated the requirement of the protein synthesis in the induction of AVO induced by AS

O

(Fig. 6b).

However, in this study, apoptosis was judged by four independent criteria, including (a) cell morphology, as described previously (Fig. 2a) (Ho et al., 1996), (b) translo- cation of phosphotidyl serine to the cell surface, as detected by an Annexin V-FITC apoptosis detection kit (Fig. 3b), (c) the presence of a sub-G1 peak, as detected by flow cytome-

try (Fig. 4b), and (d) the appearance of DNA fragmentation analyzed by the method previously described (Fig. 3a) (Ho et al., 1996). Our results indicated that apoptosis was observed in only 14.8% of U118 cells when treated with high-dose AS2O5 (50 lM) for 48 h. And, in the 10 lM AS

O

-treated U118 cells, supravital cell staining with acri- dine orange revealed the induction of acidic vesicle organ- elles (Fig. 7b), which were completely blocked by LA (50 lM) (Fig. 7d), suggesting the involvement of autoph- agy. Calculation of the acridine orange positive-stained cells demonstrated that AS

O

-induced autophagic cell death was increased in the U118 cells in a dose-dependent manner (Fig. 7f). To further confirm arsenic induced autophagy was not mainly through ROS generation. One of the antioxidant DMSA (Dimercaptosuccinic acid) was used to test this hypothesis. Fig. 7e showed that the ROS scavenger DMSA could not reverse arsenic induced autophagy in U118 cells analyzed by flow cytometry. However, AS

O

-induced autophagy in U118 cells was significantly attenuated by LA treatment (Fig. 7f). Our results suggested that arsenic mediated autophagic cell death was not mainly through

Fig. 4. LA attenuation of AS2O3-induced cell death was not through apoptosis inhibition. (a) AS2O3-induced DNA degradation was completely prevented by LA treatment in U118 cells. Cells were treated with AS2O3(10 lM) with or without LA (50 lM) for 24 h. DNA integrity in cells was analyzed by agarose electrophoresis. (b) The DNA degradation effect on the cells treated with AS2O3was also monitored by flow cytometric analysis. Cells were treated with AS2O3(5 lM) in the presence or absence of LA (50 lM) for either 24 or 48 h. Cells were then harvested for determination of the subG1 population (representing the apoptotic or necrotic cells). The percentage of cells in the subG1 phase of the cell cycle was determined using established CellFIT DNA analysis software, as shown at the bottom. Results are shown by the means ± SD of three independent experiments.^*p < 0.05.

1032 T.-J. Cheng et al. / Food and Chemical Toxicology 45 (2007) 1027–1038

ROS generation, and the protective effect of LA was not attributed to its antioxidant activity only.

4. Discussion

Increasing evidence indicated that glial cells play critical roles in numerous brain functions such as migration, axonal growth, and terminal differentiation of different neuronal subsets, through the release of soluble factors and cell-cell contact. Pathological alterations of glial cells have also been indicated associated with brain disease that impaired normal functions of neuron (Kim and de Vellis,

2005). The main purpose of this study was to investigate the mechanisms of glial cell toxicity caused by arsenic and to test the protective effects of LA. Human glioblas- toma cell line U118 cells were used as the research model to study arsenic-induced toxicity in glial cells. Although the use of cell lines for in vivo human disease fills a neces- sary gap, there is an implicit assumption that these tumor cell lines offer a representation differ from primary cultured cells. However, tumor cell lines are still good models widely used for studying the mechanisms, functions of the cells not restricted in the oncology field.

Fig. 5. The dose-dependent effect of the cell growth and apoptosis- associated regulatory proteins expression in U118 cells treated with AS2O3. U118 cells were treated with AS2O3(1–25 lM) for 24 h, and the expression levels of cell cycle- and cell death-associated proteins were detected by Western blot analysis. Cells were mock-treated with DMSO (0.05%) as a control group. The expression level of GAPDH protein was examined and served as a loading control.

Fig. 6. The AS2O3-regulated protein expression was reversed by LA treatment in U118 cells. (a) U118 cells were treated with LA (50 lM), AS2O3(5 lM), or combine-treated with LA plus AS2O3 for 24 h. Cells were then harvested and the protein expression was detected by Western blot analysis. Cells were mock-treated with DMSO (0.05%) as a control group. The expression level of GAPDH protein was examined and served as a loading control. (b) U118 cells were treated with AS2O35 lM in the presence or absence of protein synthesis inhibitor cyclohexamide (CXM) 50 lM for 24 h. Cells were stained with acridine orange and the percentage of AVO occurred was detected by flow cytometry.

AS

O

-induced apoptosis was a general mechanism found in most of clinical applications for cancer chemo-

therapy (Park et al., 2003, 2000; Woo et al., 2002). Apop- totic neuronal cell death was also suggested as a general

Fig. 7. Microphotograph using supravital cell-stain acridine orange in AS2O3-treated U118 malignant glioma cells. (a) control; (b) 10 lM AS2O3-treated cells; (c) 50 lM LA-treated cells; (d) 10 lM AS2O3and 50 lM LA-treated cells. Note the large amount of acidic vesicular organelle (AVOs) after treatment with AS2O3(b). This is consistent with the autophagic changes described in the previous paper. Acidification of AVO was inhibited by LA (d). In control cells, the cytoplasm and nucleus basically revealed the green fluorescence, but a small accumulation of the acidic component was occasionally observed (a).

The bar appearing in (a) is 10 lM. (e) U118 cells pretreated with antioxidant DMSA 50 lM following AS2O35 lM did not show any inhibition on AVO induction. The mean ± S.E.^*p < 0.05 indicated a significant difference in AS2O3alone or combined DMSA groups compared with control group, as analyzed by unpaired Student’s t-test. (f) The AVOs positive cells were counted from 10³cells and counted from three independent experiments, and each value was presented as the mean ± S.E.^*p < 0.05 indicated a significant difference between AS2O3alone groups and combined LA groups, as analyzed by unpaired Student’s t-test.

1034 T.-J. Cheng et al. / Food and Chemical Toxicology 45 (2007) 1027–1038

mechanism underlying arsenic neurological toxicity. How- ever, in this study, when U118 cells were exposed to higher doses of AS

O

(>5 lM, Figs. 1 and 7), autophagy (pro- grammed cell death type II), but not apoptosis (pro- grammed cell death type I) was induced in U118 cells as evidenced by AVOs assays. Similar results were reported in recent studies. Kanzawa et al. demonstrated that AS

O

induced autophagic cell death, but not apoptosis, in several human malignant glioma cell lines, where U118 cells were not included (Kanzawa et al., 2003, 2005).

Autophagy is a process in which subcellular membranes undergo dynamic morphological changes for the degrada- tion and turnover of cytoplasmic organelles (Mizushima et al., 1998). Excessive autophagy may contribute to the pathogenesis of some neurodegenerative disorders such as Alzheimer’s disease by altering the processing of mutant forms of amyloid precursor proteins (Meijer and Codogno, 2004). However, the precise mechanisms regulating auto- phagic cell death in glial cells remain unclear.

AS

O

-induced apoptosis has been observed in different types of human cancer cells and in brain tissue cultures (Shila et al., 2005). In these models, increased intracellular oxidative stress was demonstrated to be one of the important mechanisms that lead to AS

O

-induced cell death (Chun et al., 2002; Liang et al., 2003; Samuel et al., 2005; Shen et al., 2003; Shila et al., 2005; Tong et al., 2001; Wang et al., 2005; Woo et al., 2002). The present results reveal that ROS generation was not induced by 5 lM AS

O

within 2 h in U118 cells (Fig. 3d), suggesting the possibility that AS

O

-induced autophagic cell death in U118 cells may not occur through the ROS-mediated signaling pathway.

Our results postulated that specific regulatory mechanisms of AS

O

-induced cytotoxicity must occur in U118 cells.

Previous studies indicated that autophagic cell death and apoptosis are pathways to the same end; a functional con- nection between both forms of cell death is likely to be oper- ative. Both forms of cell death can act as backup mechanisms of each other, under conditions where cell death is imperative (Ng and Huang, 2005). As described previ- ously, the apoptosis-associated regulatory proteins, includ- ing p53, Bax, and caspases 3, 8, and 9, were induced in AS

O

-mediated apoptosis (Liang et al., 2003) and AS

O

- induced apoptosis was more easily observed in p53 wild-type glioblastoma cells (U87MG), when compared to p53- mutated cells (T98G) (Zhao et al., 2002). The p53 status in U118 cells has been described as a mutated type in a previous report (Ohneseit et al., 2005). Interestingly, we also found that caspases 3, 8, and 9 remained unchanged, and p53 and Bax proteins were significantly down-regulated in the U118 cells treated with AS

O

in a dose-dependent manner (Fig. 5). Such results implied that sufficient or intact p53 pro- tein functions are needed to induce apoptosis in response to AS

O

in human glioma cells. Thus, arsenic induced U118 cells autophagy maybe occurred when the p53 mediated apoptotic pathway of U118 cells was blocked after arsenic exposure. However, the precise mechanisms regulating autophagic cell death in glial cells remain unclear.

While some heat shock proteins (HSPs) are constitu- tively expressed, others are induced by various stress agents. Induction of HSP-70 has been demonstrated in C6 rat glioma cells treated with 100 lM sodium arsenite for 1 h (Kato et al., 1997). It is suggested that HSP expres- sion is generated by abnormally folded, nonnative proteins produced by stress capable of triggering the oxidation of nonprotein thiols and leading to heat shock transcription factor activation. A previous report indicated that one of the three types of autophagy is chaperone mediated autophagy, in which the heat shock protein families (Hsc- 73) bind to soluble proteins and initiate their transport to prelysosomal or lysosomal compartment leading to com- plete degradation of the proteins (Ng and Huang, 2005).

In our current study, we found that HSP-70 expression was increased in arsenic-induced U118 cell autophagy and attenuated by LA treatment. These results implicate that arsenic-induced autophagic cell death mechanisms in glial cells may related to the heat shock response. Arsenic is capable of inducing HSPs of various sizes, from the smal- ler proteins (HSP-25, HSP-27, and HSP-30) to the larger HSP species, such as HSP-105 depends on cell type, dose, and metabolic state of the cell. However, the detailed mech- anisms of Hsp-70 in AS

O

induced autophagy need to be further investigated.

Current evidence supports the role of reactive oxygen species in a number of types of acute and chronic patho- logic conditions in the brain and neural tissue in response to arsenic. The study of Shila et al. also confirmed that arsenic reduced the activities of antioxidants enzymes in the rat brain region. Simultaneously treatment of LA along with arsenic may against arsenic-induced decline in the antioxidants enzymes, increased oxidants and lipid peroxi- dation (Shila et al., 2005). Samuel et al. further suggested the protective effects of LA was due to its antioxidant activ- ity to directly react with various ROS species, regulate the reproductive nature of the system to prevent oxidation of protein thiols and prevent arsenic from binding to thiols (Samuel et al., 2005). The original purpose of this study was to test the hypothesis that LA prevents AS

O

-induced autophagy through the scavenging of ROS generation.

Induction of oxidative stress, including ROS generation, lipid peroxidation, and glutathione reduction in U118 cells treated with mycotoxin fumonisin B1 has been reported before (Samuel et al., 2005). However, our result reveal that ROS generation was not induced by AS

O

within 2 h in U118 cells (Fig. 3d), suggesting the possibility that AS

O

-induced autophagic cell death in U118 cells may not occur through the ROS-mediated signaling pathway.

On the one hand, AS

O

-mediated autophagic cell death

was proved to be nearly completely attenuated by LA,

but not by the other antioxidant such as DMSA

(Fig. 7e). Thus, we suggested that the protective effect of

LA was not attributed to its antioxidant activity only and

we strongly believe that other regulating pathways could

dominate the direct effects of ROS in AS

O

-induced

autophagy in U118 cells. It is suggested that LA may

activate certain as-yet-uncharacterized signaling intermedi- ates by inducing intramolecular disulfide bond formation, and these intermediates in turn trigger the activation of transcription factors which promote the expression of several genes that can protect U118 cells from autophagy (McCarty, 2001; Prestera and Talalay, 1995; Talalay et al., 1995). Moreover, the binding of arsenic to critical thiol groups in proteins is one of its toxic mechanisms.

Since LA itself is a thiol compound, direct interaction between LA and arsenic may occur. Thus, the protective effect of LA against AS

O

-induced glial cells from autoph- agy may be attributed to the direct chelating of the com- pound or through inducing uncertain signaling pathways.

From oncologic aspect, glioblastoma is a highly malig- nant glioma and possesses resistant to many treatments, including radiotherapy, chemotherapy, and adjuvant ther- apy (Haga et al., 2005). Recently, clinical trial results have suggested that AS

O

has potential effectiveness in patients with other solid tumors, including melanoma (Kim et al., 2005), hepatocellular carcinoma (Zhang et al., 2003), and renal cancer (Vuky et al., 2002). Previous reports and the current study have demonstrated that arsenic trioxide could induce autophagy in malignant glioma cells (Kanz- awa et al., 2003, 2005). These suggested a clinical benefit by using AS

O

as anticancer agent in the treatment of gli- oma. Furthermore, the current study also found that LA could reverse AS

O

-induced autophagic cell death of U118 cells. It has been reported that treatment of antioxi- dants can protect cells against radiation and chemotherapy (Greggi Antunes et al., 2000; Sonneveld, 1978; Witenberg et al., 1999). Theses results implied that antioxidants might protect cancer cells, thereby reducing the oncologic effec- tiveness or cytotoxicity therapy. Although the protective effects of LA in inhibiting U118 cells from autophagy were not mainly through antioxidant property, we suggested that patients should avoid LA supplements during arsenic based chemotherapy or arsenic combined radiotherapy.

Taken together, the present study demonstrated for the first time that AS

O

-induced glial cells toxicity could be completely attenuated by LA. U118 cells, autophagic, but not apoptotic, cell death was induced by AS

O

, and the AS

O

-mediated autophagic cell death may not occur mainly through the ROS-triggered signal pathway.

Down-regulation of p53 and Bax proteins and up-regula- tion of Bcl-2 and HSP-70 proteins may involve in AS

O

- mediated autophagic cell death. Even the detailed mecha- nisms of LA against AS

O

-induced glial cells autophagy are still unclear and need to be further investigated. We concluded that LA could be a useful protective agent against arsenic-induced glial cells toxicity and reversing arsenic-induced damage in human brain.

Acknowledgements

This study was supported by the National Science Council grant NSC 93-2314-B-038-051 to Dr. Ho, and

NSC 93-2320-B-006-049 to Dr. Wang and by the Chi Mei Medical Center (93CM-TMU-07).

References

Basun, H., Forssell, L.G., Wetterberg, L., Winblad, B., 1991. Metals and trace elements in plasma and cerebrospinal fluid in normal aging and Alzheimer’s disease. Journal of Neural Transmission Parkinsons Disease and Dementia Section 3 (4), 231–258.

Chattopadhyay, S., Bhaumik, S., Nag Chaudhury, A., Das Gupta, S., 2002. Arsenic induced changes in growth development and apoptosis in neonatal and adult brain cells in vivo and in tissue culture.

Toxicology Letters 128 (1–3), 73–84.

Chun, Y.J., Park, I.C., Park, M.J., Woo, S.H., Hong, S.I., Chung, H.Y., Kim, T.H., Lee, Y.S., Rhee, C.H., Lee, S.J., 2002. Enhancement of radiation response in human cervical cancer cells in vitro and in vivo by arsenic trioxide (AS2O3). FEBS Letters 519 (1–3), 195–200.

Filippova, M., Duerksen-Hughes, P.J., 2003. Inorganic and dimethylated arsenic species induce cellular p53. Chemical Research in Toxicology 16 (3), 423–431.

Fujino, Y., Guo, X., Liu, J., You, L., Miyatake, M., Yoshimura, T.

Japan Inner Mongolia Arsenic Pollution Study Group, 2004.

Mental health burden amongst inhabitants of an arsenic-affected area in Inner Mongolia, China. Social Science and Medicine 59 (9), 1969–1973.

Gamalei, I.A., Kirpichnikova, K.M., Ishchenko, A.M., Zhakhov, A.V., Kliubin, I.V., 1998. Activation of astrocytes of by anaphylatoxins of the complement system. Tsitologiia 40 (8–9), 768–772.

Gao, F., Yi, J., Shi, G.Y., Li, H., Shi, X.G., Tang, X.M., 2002. The sensitivity of digestive tract tumor cells to AS2O3is associated with the inherent cellular level of reactive oxygen species. World Journal of Gastroenterology 8 (1), 36–39.

Grant, S., Turner, A.J., Bartimole, T.M., Nelms, P.A., Joe, V.C., Jarvis, W.D., 1994. Modulation of 1-[beta-D-arabinofuranosyl] cytosine- induced apoptosis in human myeloid leukemia cells by staurosporine and other pharmacological inhibitors of protein kinase C. Oncology Research 6 (2), 87–99.

Greggi Antunes, L.M., Darin, J.D., Bianchi, M.D., 2000. Protective effects of vitamin C against cisplatin-induced nephrotoxicity and lipid peroxidation in adult rats: a dose-dependent study. Pharmacological Research 41 (4), 405–411.

Haga, N., Fujita, N., Tsuruo, T., 2005. Involvement of mitochondrial aggregation in arsenic trioxide (AS2O3)-induced apoptosis in human glioblastoma cells. Cancer Science 96 (11), 825–833.

Hall, A.H., 2002. Chronic arsenic poisoning. Toxicology Letters 128 (1–3), 69–72.

Hantson, P., Haufroid, V., Buchet, J.P., Mahieu, P., 2003. Acute arsenic poisoning treated by intravenous dimercaptosuccinic acid (DMSA) and combined extrarenal epuration techniques. Journal of Toxicology Clinical Toxicology 41 (1), 1–6.

Ho, Y.S., Wang, Y.J., Lin, J.K., 1996. Induction of p53 and p21/WAF1/

CIP1 expression by nitric oxide and their association with apoptosis in human cancer cells. Molecular Carcinogenesis 16 (1), 20–31.

Jiang, X.H., Wong, B.C., Yuen, S.T., Jiang, S.H., Cho, C.H., Lai, K.C., Lin, M.C., Kung, H.F., Lam, S.K., 2001. Arsenic trioxide induces apoptosis in human gastric cancer cells through up-regula- tion of p53 and activation of caspase-3 Erratum appears in Int J Cancer 2001 September 15;93(6):916 Note: Chun-Yu Wong B [corrected to Wong BC]. International Journal of Cancer 91 (2), 173–179.

Kanzawa, T., Kondo, Y., Ito, H., Kondo, S., Germano, I., 2003.

Induction of autophagic cell death in malignant glioma cells by arsenic trioxide. Cancer Research 63 (9), 2103–2108.

Kanzawa, T., Zhang, L., Xiao, L., Germano, I.M., Kondo, Y., Kondo, S., 2005. Arsenic trioxide induces autophagic cell death in malignant glioma cells by upregulation of mitochondrial cell death protein BNIP3. Oncogene 24 (6), 980–991.

1036 T.-J. Cheng et al. / Food and Chemical Toxicology 45 (2007) 1027–1038

Karlsson, J.I.O.R., Porn-Ares, I., Pahlman, S., 2004. Arsenic trioxide- induced death of neuroblastoma cells involves activation of Bax and does not require p53. Clinical Cancer Research 10 (9), 3179–

3188.

Kato, K., Ito, H., Okamoto, K., 1997. Modulation of the arsenite-induced expression of stress proteins by reducing agents. Cell Stress and Chaperones 2 (3), 199–209.

Kim, S.U., de Vellis, J., 2005. Microglia in health and disease. Journal of Neuroscience Research 81 (3), 302–313.

Kim, K.B., Bedikian, A.Y., Camacho, L.H., Papadopoulos, N.E., McCullough, C., 2005. A phase II trial of arsenic trioxide in patients with metastatic melanoma. Cancer 104 (8), 1687–1692.

Kozlov, A.V., Gille, L., Staniek, K., Nohl, H., 1999. Dihydrolipoic acid maintains ubiquinone in the antioxidant active form by two-electron reduction of ubiquinone and one-electron reduction of ubisemiqui- none. Archives of Biochemistry and Biophysics 363 (1), 148–

154.

Larsen, N.A., Pakkenberg, H., Damsgaard, E., Heydorn, K., Wold, S., 1981. Distribution of arsenic, manganese, and selenium in the human brain in chronic renal insufficiency, Parkinson’s disease, and amyo- trophic lateral sclerosis. Journal of the Neurological Sciences 51 (3), 437–446.

Lee, Y.W., Ha, M.S., Kim, Y.K., 2001. Role of reactive oxygen species and glutathione in inorganic mercury-induced injury in human glioma cells. Neurochemical Research 26 (11), 1187–1193.

Lee, W.S., Chen, R.J., Wang, Y.J., Tseng, H., Jeng, J.H., Lin, S.Y., Liang, Y.C., Chen, C.H., Lin, C.H., Lin, J.K., Ho, P.Y., Chu, J.S., Ho, W.L., Chen, L.C., Ho, Y.S., 2003. In vitro and in vivo studies of the anticancer action of terbinafine in human cancer cell lines: G0/G1 p53- associated cell cycle arrest. International Journal of Cancer 106 (1), 125–137.

Liang, X.Q., Cao, E.H., Zhang, Y., Qin, J.F., 2003. P53-induced gene 11 (PIG11) involved in arsenic trioxide-induced apoptosis in human gastric cancer MGC-803 cells. Oncology Reports 10 (5), 1265–

1269.

Liu, J.D., Wang, Y.J., Chen, C.H., Yu, C.F., Chen, L.C., Lin, J.K., Liang, Y.C., Lin, S.Y., Ho, Y.S., 2003. Molecular mechanisms of G0/G1 cell- cycle arrest and apoptosis induced by terfenadine in human cancer cells. Molecular Carcinogenesis 37 (1), 39–50.

Maeda, H., Hori, S., Nishitoh, H., Ichijo, H., Ogawa, O., Kakehi, Y., Kakizuka, A., 2001. Tumor growth inhibition by arsenic trioxide (AS2O3) in the orthotopic metastasis model of androgen-independent prostate cancer. Cancer Research 61 (14), 5432–5440.

McCarty, M.F., 2001. Versatile cytoprotective activity of lipoic acid may reflect its ability to activate signalling intermediates that trigger the heat-shock and phase II responses. Medical Hypotheses 57 (3), 313–

317.

Meijer, A.J., Codogno, P., 2004. Regulation and role of autophagy in mammalian cells. International Journal of Biochemistry and Cell Biology 36 (12), 2445–2462.

Mizushima, N., Noda, T., Yoshimori, T., Tanaka, Y., Ishii, T., George, M.D., Klionsky, D.J., Ohsumi, M., Ohsumi, Y., 1998. A protein conjugation system essential for autophagy see comment. Nature 395 (6700), 395–398.

Ng, G., Huang, J., 2005. The significance of autophagy in cancer.

Molecular Carcinogenesis 43 (4), 183–187.

Nolte, F., Friedrich, O., Rojewski, M., Fink, R.H., Schrezenmeier, H., Korper, S., 2004. Depolarisation of the plasma membrane in the arsenic trioxide (AS2O3)-and anti-CD95-induced apoptosis in myeloid cells Erratum appears in FEBS Lett. 2005 Jul 4;579(17):3866. FEBS Letters 578 (1–2), 85–89.

Ohneseit, P.A., Prager, D., Kehlbach, R., Rodemann, H.P., 2005. Cell cycle effects of topotecan alone and in combination with irradiation.

Radiotherapy and Oncology 75 (2), 237–245.

Packer, L., Witt, E.H., Tritschler, H.J., 1995. Alpha-lipoic acid as a biological antioxidant. Free Radical Biology and Medicine 19 (2), 227–

250.

Packer, L., Tritschler, H.J., Wessel, K., 1997. Neuroprotection by the metabolic antioxidant alpha-lipoic acid. Free Radical Biology and Medicine 22 (1–2), 359–378.

Park, W.H., Seol, J.G., Kim, E.S., Hyun, J.M., Jung, C.W., Lee, C.C., Kim, B.K., Lee, Y.Y., 2000. Arsenic trioxide-mediated growth inhibition in MC/CAR myeloma cells via cell cycle arrest in associ- ation with induction of cyclin-dependent kinase inhibitor, p21, and apoptosis. Cancer Research 60 (11), 3065–3071.

Park, I.C., Park, M.J., Woo, S.H., Lee, H.C., An, S., Gwak, H.S., Lee, S.H., Hong, S.I., Bae, I.J., Seo, K.M., Rhee, C.H., 2003. Tetraarsenic oxide induces apoptosis in U937 leukemic cells through a reactive oxygen species-dependent pathway. International Journal of Oncology 23 (4), 943–948.

Prestera, T., Talalay, P., 1995. Electrophile and antioxidant regulation of enzymes that detoxify carcinogens. Proceedings of the National Academy of Sciences of the United States of America 92 (19), 8965–

8969.

Ramos, A.M., Fernandez, C., Amran, D., Sancho, P., de Blas, E., Aller, P., 2005. Pharmacologic inhibitors of PI3K/Akt potentiate the apoptotic action of the antileukemic drug arsenic trioxide via glutathione depletion and increased peroxide accumulation in myeloid leukemia cells. Blood 105 (10), 4013–4020.

Rao, M.V., Avani, G., 2004. Arsenic induced free radical toxicity in brain of mice. Indian Journal of Experimental Biology 42 (5), 495–498.

Ratnaike, R.N., 2003. Acute and chronic arsenic toxicity. Postgraduate Medical Journal 79 (933), 391–396.

Samuel, S., Kathirvel, R., Jayavelu, T., Chinnakkannu, P., 2005. Protein oxidative damage in arsenic induced rat brain: influence ofDL-alpha- lipoic acid. Toxicology Letters 155 (1), 27–34.

Shen, Z.Y., Shen, W.Y., Chen, M.H., Shen, J., Zeng, Y., 2003. Reactive oxygen species and antioxidants in apoptosis of esophageal cancer cells induced by AS2O3. International Journal of Molecular Medicine 11 (4), 479–484.

Shila, S., Kokilavani, V., Subathra, M., Panneerselvam, C., 2005. Brain regional responses in antioxidant system to alpha-lipoic acid in arsenic intoxicated rat. Toxicology 210 (1), 25–36.

Sohal, R.S., Ku, H.H., Agarwal, S., Forster, M.J., Lal, H., 1994.

Oxidative damage, mitochondrial oxidant generation and antioxi- dant defenses during aging and in response to food restriction in the mouse. Mechanisms of Ageing and Development 74 (1–2), 121–133.

Sonneveld, P., 1978. Effect of alpha-tocopherol on the cardiotoxicity of adriamycin in the rat. Cancer Treatment Reports 62 (7), 1033–

1036.

Talalay, P., Fahey, J.W., Holtzclaw, W.D., Prestera, T., Zhang, Y., 1995.

Chemoprotection against cancer by phase 2 enzyme induction.

Toxicology Letters 83, 173–179.

Tong, Q., Zeng, F., Zheng, L., Zhao, J., Lu, G., 2001. Apoptosis inducing effects of arsenic trioxide on human bladder cancer cell line BIU-87.

Chinese Medical Journal 114 (4), 402–406.

Traganos, F., Darzynkiewicz, Z., 1994. Lysosomal proton pump activity:

supravital cell staining with acridine orange differentiates leukocyte subpopulations. Methods in Cell Biology 41, 185–194.

Vuky, J., Yu, R., Schwartz, L., Motzer, R.J., 2002. Phase II trial of arsenic trioxide in patients with metastatic renal cell carcinoma. Investiga- tional New Drugs 20 (3), 327–330.

Wang, Z.H., Yu, D., Chen, Y., Hao, J.Z., 2005. Proteomic analysis of nuclear matrix proteins during arsenic trioxide induced apoptosis in leukemia K562 cells. Chinese Medical Journal 118 (2), 100–

104.

Witenberg, B., Kalir, H.H., Raviv, Z., Kletter, Y., Kravtsov, V., Fabian, I., 1999. Inhibition by ascorbic acid of apoptosis induced by oxidative stress in HL-60 myeloid leukemia cells. Biochemical Pharmacology 57 (7), 823–832.

Woo, S.H., Park, I.C., Park, M.J., Lee, H.C., Lee, S.J., Chun, Y.J., Lee, S.H., Hong, S.I., Rhee, C.H., 2002. Arsenic trioxide induces apoptosis through a reactive oxygen species-dependent pathway and loss of

mitochondrial membrane potential in HeLa cells. International Jour- nal of Oncology 21 (1), 57–63.

Zegura, B., Lah, T.T., Filipic, M., 2004. The role of reactive oxygen species in microcystin-LR-induced DNA damage. Toxicology 200 (1), 59–68.

Zhang, T., Wang, S.S., Hong, L., Wang, X.L., Qi, Q.H., 2003. Arsenic trioxide induces apoptosis of rat hepatocellular carcinoma cells in vivo.

Journal of Experimental and Clinical Cancer Research 22 (1), 61–

68.

Zhao, S., Tsuchida, T., Kawakami, K., Shi, C., Kawamoto, K., 2002.

Effect of AS2O3 on cell cycle progression and cyclins D1 and B1 expression in two glioblastoma cell lines differing in p53 status.

International Journal of Oncology 21 (1), 49–55.

1038 T.-J. Cheng et al. / Food and Chemical Toxicology 45 (2007) 1027–1038