- Email: [email protected]
Reactive oxygen species involved in the glutamate toxicity of C6 glioma cells via XC¯ antiporter system
~
Pergamon
Neuroscience Vol. 73, No. 1, pp. 201-208, 1996 Copyright© 1996 IBRO. Publishedby ElsevierScienceLtd Printed in Great Britain S03 06-4522(96)00025- 5 0306-4522/96 $15.00+ 0.00
REACTIVE OXYGEN SPECIES INVOLVED IN THE GLUTAMATE TOXICITY OF C6 GLIOMA CELLS VIA Xt~ ANTIPORTER SYSTEM K. MAWATARI,* Y. Y A S U I , t K. SUGITANI,* T. TAKADERA:~ and S. KATOt§ *Department of Laboratory Sciences, tDepartment of Neurobiology, NIRI, University of Kanazawa Faculty of Medicine, Kanazawa 920, Japan :~Hokuriku University, Faculty of Pharmaceutical Science, Kanazawa 920-11, Japan Abstract--We recently demonstrated that continuous L-glutamate exposure led to cell death in C6 glioma cells over a period of 24-36 h, due to inhibition of cystine uptake through the cystine/glutamate (xr) antiporter. The antioxidant vitamin E provided protection against this effect, supporting the hypothesis that depletion of glutathione might be responsible, resulting from insufficient cystine uptake. To clarify the content of oxidative stress after glutathione depletion, the present study was done to investigate accumulation and target molecules of reactive oxygen species induced by glutamate treatment. The accumulation of reactive oxygen species was increased three-fold as compared to a control culture. Membrane oxidation, as judged by lipid peroxidation, was increased two-fold after glutamate treatment. Cellular ATP content was significantly reduced by glutamate exposure. For the two cytosolic enzymes examined, activity of glyceraldehyde 3-phosphate dehydrogenase was slightly enhanced by glutamate treatment, while activity of glutamine synthetase was not changed. Impairment of nuclear DNA after glutamate exposure was also revealed by nuclear chromatin condensation with DNA fragmentation. Thus, the multiple targets (membrane, cytoplasm and nuclei) of oxygen radicals in glutamate toxicity through the xe antiporter system were evaluated for the first time. Furthermore, prevention from cell death and from cellular toxicity induced by oxygen radicals could be seen using three specific oxygen radical scavengers, catalase, 3,3,5,5-tetramethyl-pyrroline N-oxide and ~-phenyl-N-t-butylnitrone, without restoring the glutathione deficit. This indicates that radical scavengers did not interact with the xe antiporter system, but directly scavenged the oxygen radicals. Taken together, the data strongly suggest that O ; , H202 and 'OH accumulate in response to oxidative stress after glutathione depletion, resulting in glutamate cell death of C6 glioma cells. Copyright © 1996 IBRO. Published by Elsevier Science Ltd. Key words: cell death, glutathione depletion, H202 accumulation, lipid peroxidation, radical scavengers.
Recent review articles have separated glutamate toxicity into receptor-mediated and transporter-mediated types. 7's'27 The former (receptor-mediated) is subdivided into N-methyl-D-aspartate (NMDA) and non-NMDA receptor types. The mechanism for the toxicity is thought to be due to membrane depolarization (excitation) accompanying cation influx, especially of calcium ions, by glutamate receptor activation. 7 The latter (transporter-mediated) is unique and requires cellular expression of the cystine/glutamate antiporter (x~). 2 The x~ system transports cystine inward and glutamate outward with the glutamate gradient providing the driving §To whom correspondence should be addressed. Abbreviations: DCF, 2',7'-dichlorofluorescin; DCFH-DA, 2,7-dichlorofluorescin diacetate; DMEM, Dulbecco's modified Eagle's medium; EDTA, ethylenediamine tetra-acetate; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GSH, glutathione (7-glutamylcysteinylglycine); HEPES, N-2-hydroxyethylpiperazine-N'-2ethanesulfonic acid; LDH, lactate dehydrogenase; NMDA, N-methyl-D-aspartate; PBN, ~t-phenyl-N-tbutylnitrone; ROS, reactive oxygen species; TMPO, 3,3,5,5-tetramethyl-pyrroline N-oxide.
force. The cystine taken up is intracellularly reduced to cysteine and then utilized to synthesize glutathione (y-glutamylcysteinylglycine, GSH). Therefore, cellular levels of GSH are maintained through the xE antiporter system in some cells. ~ The non-excitatory mechanism for the glutamate toxicity through the xE antiporter is due to inhibition of cystine uptake in the constant ( ~ 2 4 h ) and high (depending upon the concentration of extracellular cystine 19,29) exposure of glutamate. This inhibition leads to GSH depletion, finally resulting in a delayed type cell death, s,19 It is expected that glutamate cell death results from oxidative stress after GSH depletion, because GSH is a major non-protein thiol compound in the living cells and is a key molecule for protection and conjugation against various oxidative stress or toxins. 26 Recently, glutamate toxicity through the x~ antiporter has been reported in many CNS neurons and glia. 6'17'19'29'3°'32 Pathogenesis of neurodegenerative disorders including stroke, 4,21 Parkinson's disease9.43 and Alzheimer's disease3,12 has implicated oxygen radicals.34,41 The brain has a large oxygen consumption and thus has many opportunities to be exposed to
201
202
K. Mawatari et al.
oxidant overload. However, which reactive oxygen species (ROS) are most involved in the disorders remain to be clarified. To answer this question, we developed a n in vitro cell system. W e previously used a rat clonal cell line o f C6 glioma cells to investigate the toxicity a n d protective m e c h a n i s m s for glutamate cell d e a t h mediated via the x~ a n t i p o r t e r system. '9'4° The toxic action o f glutamate (10 m M ) o n C6 glioma cells could be prevented by cystine (0.1-1 m M ) a n d vitamin E ( 1 0 0 / t M ) . The decrease o f cellular G S H was initiated early (4-6 h) after glutamate t r e a t m e n t a n d the c o n c e n t r a t i o n o f G S H was decreased 10% of control at 12 h after glutamate exposure. However, m o r p h o l o g i c a l or biochemical signs of cell d e a t h were initiated later ( 1 3 - 1 4 h ) after g l u t a m a t e t r e a t m e n t a n d almost all cells died at 24-36 h after g l u t a m a t e exposure. 19 By utilizing the relatively long delay between G S H depletion a n d cell death, the present study has clarified oxidants accumulated in the course o f cell death. A preliminary a b s t r a c t o f this work has been published elsewhere. 25
EXPERIMENTAL PROCEDURES
Chemicals
Monosodium L-glutamate was purchased from Ajinomoto-Takara Corp. (Tokyo, Japan). Vitamin E was obtained from Wako Pure Chemicals (Osaka, Japan). Catalase, superoxide dismutase, 3,3,5,5-tetramethylpyrroline N-oxide (TMPO) and ct-phenyl-N-t-butylnitrone (PBN) were obtained from Sigma (St Louis, MO, U.S.A.). Cell culture
C6 glioma cells were grown in Dulbecco's modified Eagle's medium (DMEM; Gibco) containing 5% fetal calf serum with penicillin (100units/ml) and streptomycin (100/~g/ml) at 37°C in a humidified atmosphere of 90% air/10% CO2, as described previously26 The concentration of extracellular cystine in the standard DMEM was about 200#M. Exponentially growing cells were dissociated and plated on 35-mm dishes (Falcon) at a density of 4 x 10~cells/dish [for lactate dehydrogenase (LDH) assay and DNA staining] or on 60-mm dishes (Falcon) at a density of 1.5 × 106 cells/dish (for other assays). Test reagents such as glutamate or other chemicals were simply added to the culture medium at the beginning of the culture. Cell viability (cytotoxicity) was assessed by the ratio of released LDH into the medium per total LDH activity, as described previously.4° The effects of drugs on C6 glioma cells were also observed morphologically with a phase-contrast microscope (Nikon, TMD-2). Biochemical assay Reactive oxygen species. ROS were assayed using 2',7'dichlorofluorescin diacetate (DCFH-DA; Molecular Probes, Eugene, OR, U.S.A.) which is de-esterified within cells to the ionized free acid, 2',7'-dichlorofluorescin. This is trapped within cells and thus accumulated. 2',7'-Dichlorofluorescin is capable of being oxidized to fluorescent 2',7'-dichlorofluorescein (DCF) by ROS. ROS accumulation was measured by a modification of the previously described method of Lebel and Bondy) 3 In brief, C6 glioma cells were collected, washed with Hank's balanced salt solution, pH 7.4, and resuspended in a total volume of 2.0 ml of the same buffer. The cell suspension was divided into two tubes, one for ROS assay and the other for protein
quantification. The cells were diluted in the same volumes of Hank's balanced salt solution and then incubated with 10pM DCFH-DA at 37°C for 15 min. After loading with DCFH-DA, the cell suspension was centrifuged at 3000 g for 5 min (4°C) and the pellet was resuspended in 2 ml of the same ice-cold buffer, and then the fraction was incubated for a further 60 min (37°C). At the beginning and end of incubation, fluorescence was measured at 488 nm excitation and 525 nm emission (Hitachi, F-2000). Autofluorescence was corrected by the inclusion of parallel blanks without DCFH-DA. The correction for autofluorescence was always less than 10% of the total. ROS accumulation was quantitated from a DCF (Wako Pure Chemicals) standard curve (10-600nM) and results were expressed as nmol DCF formed/h//~g protein. Protein concentration was determined by the method of Lowry et al. 24 Glutathione. Cellular levels of GSH were assayed by a high-performance liquid chromatography method with thiol group-specific fluorogenic reagents, as described elsewhere.19 Lipid peroxide. The content of lipid peroxides was determined fluorometrically using the thiobarbituric acid method described previously. 4° A TP. The level of ATP was determined using a luciferindependent bioluminescence method. 44 C6 glioma cells were collected and an aliquot of cell suspension was diluted twice in 0.6 N perchloric acid solution with 2 mM EDTA, then centrifuged at 12,000g for 2min at 4°C and the supernatant was neutralized with 3 M K2CO 3. The assay mixture consisted of 20 pl of the supernatant, 180/~1 of 25mM HEPES-NaOH buffer, pH7.8, and 100/~1 of luciferin luciferase reagents (Kikkoman Ltd). The concentration of ATP was calculated from an ATP standard curve (0.1-100 nM). Glutamine synthetase. The activity of glutamine synthetase was assayed by a high-performance liquid chromatography method with postcolumn derivatization using o-phthalaldehyde.tS Glyceraldehyde 3-phosphate dehydrogenase. The activity of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was assayed by a modification of the previously described method of Giblin et al. It In brief, cells were collected, centrifuged at 3000 g for 5 min and homogenized in a total volume of 1.0ml of cold 20mM triethanolamine buffer, pH 7.6, containing 0.4 mM EDTA. The assay mixture consisted of 75 mM triethanolamine, 1.5 mM EDTA, 6.0 mM 3-phosphoglyceric acid (Boehringer Mannheim), 1,I mM ATP, 1.3mM MgSO4, 0.2mM reduced form of nicotinamide adenine dinucleotide (NADH; Oriental Kobo, Ltd) and 20 units of phosphoglycerate kinase (Boehringer Mannheim). The activity of GAPDH was assayed spectrophotometrically and expressed as the ratio of the absorbance reduction of NADH at 340 nm as compared with that of the standard GAPDH enzyme (8000U/ml; Boehringer Mannheim). DNA staining. Nuclear DNA staining was performed with a specific fluorescent probe, H33258. 3~ Prior to cultivation, a small cover glass (22 x 22 mm 2) was put into a 35-mm dish, onto which C6 glioma cells were seeded and cultured. After removal of the medium, cells attached on the cover glass were gently fixed with 1.0 ml of 10% formalin neutral buffer solution (pH 7.4) for 5 min, washed with distilled water and 2.0 ml of water added. Cells were stained with H33258 (Molecular Probes, Eugene, OR, U.S.A.) at a final concentration of 8 #g/ml for 5 min. The cells were then washed in water and mounted with glycerol. Fluorescence was observed and photographed using a fluorescence microscope (Nikon UFX-IIA) at 370 nm excitation and 500 nm emission. Statistics
Three to five different experiments for biochemical assays were repeated. Results are expressed as the mean + S.D. and statistical differences were evaluated using Student's t-test.
Oxidative stress and glutamate toxicity Table I. Increased rate of 2',7'-dichlorofluorescin formation after glutamate treatment DCF formed (nmol/#g protein/h) Time (h) 8 16 22
C6
L-Glutamate, 10 mM
0.21 +_.0.06 0.21 __.0.05 0.19 + 0.03
0.30 _.+0.06 0.58 _.+0.16" 0.63 + 0.14"
C6 glioma cells were cultured in standard DMEM with 10mM L-glutamate and DCF formation was measured with a fluorescent probe, DCFHDA. The rate of DCF formation was significantly increased (about three times) 16 and 22 h after glutamate treatment. E a c h value is the mean _+S.D. (n = 3-5). *P < 0.01 compared with control.
RESULTS
Accumulation of oxygen radicals in glutamate toxicity To clarify the involvement of oxygen radicals in the glutamate toxicity of C6 glioma cells, we first measured accumulation of ROS after glutamate exposure. ROS accumulation was estimated utilizing a converting reaction of DCFH-DA to DCF oxidized by ROS. The rate of generation of DCF in the C6 cells increased time-dependently (Table 1). ROS accumulation was enhanced 2.8-fold 16 h and maximally 3.3-fold 22 h after glutamate treatment. To further confirm the participation of ROS in the glutamate toxicity, we next investigated the effects of oxygen radical scavengers on DCF formation. We used TMPO and PBN as spin-trapping agents specific to O~- and "OH. In the presence of these scavengers (TMPO, 3mM; PBN, 2mM) with glutamate (10mM), the increase of ROS accumulation was clearly suppressed 22 h after treatment (Table 2). Vitamin E (100/~M), a powerful antioxidant, was even more effective than TMPO and PBN (Table 2). These data strongly suggest that oxygen radicals substantially accumulate within C6 cells after GSH depletion.
Table 2. Protective effect of radical scavengers and antioxidant on the rate of 2',7'-dichlorofluorescin formation after glutamate treatment Treatment Control L-Glutamate (10 mM) L-Glutamate 4- TMPO (3 mM) L-Glutamate 4- PBN (2 mM) L-Glutamate 4- vitamin E (100~M)
DCF formed (nmol/pg protein/h) 0.19 + 0.03 0.63 + 0.14 0.43 ___0.07* 0.38 + 0.03* 0.15 __+0.03**
C6 cells were cultured with 10 mM L-glutamate or with 10 mM L-glutamate plus either TMPO, PBN or vitamin E. TMPO (3mM), PBN (2mM) and vitamin E (100/~M) significantly suppressed the increased rate of DCF formation 22 h after glutamate treatment. Each value is the mean +__S.D. (n = 4-6). *P < 0.05 and **P < 0.01 compared with glutamate treatment.
203
Table 3. Protective effect of radical scavengers and antioxidant on glutamate cell death Treatment Control L-Glutamate (10 mM) L-Glutamate + catalase (100 pg/ml) L-Glutamate + TMPO (3 mM) L-Glutamate + PBN (2 mM) L-Glutamate 4- superoxide dismutase (450 units/ml)
Cell death (%) 3.3 _+ 1.2 65.1 _+4.7 13.0 + 4.0* 15.0 __+4.0*
6.4 _+ 1.6" 62.8 _+5.1
C6 cells were cultured with 10mM L-glutamate or with l0 mM L-glutamate plus catalase, TMPO, PBN or superoxide dismutase. When 10mM L-glutamate was added to the medium at the beginning of the culture, almost all cells died 24-36 h later. The presence of catalase (100/tg/ml), TMPO (3 mM) or PBN (2 mM) suppressed this glutamate-induced cell death, whereas the cell death was unaffected by the presence of superoxide dismutase (450 units/m1).Cell death (%) 36 h after culture was determined by LDH assay. Each value is the mean +_S.D. of three to five different experiments with duplicate samples. *P < 0.01 compared with glutamate treatment.
Protection from glutamate cell death by oxygen radical scavengers In order to examine the protective effect of radical scavengers, either catalase (100/~g/ml), an H202removing enzyme, TMPO (3 mM) or PBN (2 mM) was added to the culture medium in combination with 10mM glutamate at the beginning of culture. Cell viability (cell death) of C6 glioma cells was evaluated by LDH release into the medium 36 h after culture (see Experimental Procedures). Addition of these agents completely suppressed glutamateinduced cell death (Table 3). The minimum effective doses of catalase, TMPO and PBN were 30/~g/ml, 1 mM and 1 mM, respectively. When C6 glioma cells were treated with these scavengers alone, no significant change in morphology, growth or LDH release could be seen (data not shown). In contrast to catalase, superoxide dismutase (1-500units/ml), a superoxide anion-removing enzyme, did not suppress cell death (Table 3). These data strongly suggest that accumulation of ROS (O~-, H202 and 'OH) participates in the glutamate cell death of C6 glioma cells.
Lack of restoration of the glutathione deficit with oxygen radical scavengers To address the mechanism by which oxygen radical scavengers prevented glutamate-induced cell death, we measured cellular levels of GSH to exclude the possibility of restoration of GSH level by their scavengers, as occurs in the case of cystineJ 9 Cellular levels of GSH 20 h after combined treatment with glutamate and scavengers are shown in Table 4. These scavengers did not restore the GSH deficit induced by glutamate. The data strongly suggest that the oxygen radical scavengers do not modulate the x~ antiporter system, but directly interact with either the ROS and/or their intermediates.
K. Mawatari et al.
204
Table 4. Cellular levels of glutathione in C6 glioma cells treated with glutamate and antioxidative agents Treatment Control L-Glutamate (I0 raM) L-Glutamate + catalase (I00/~g/ml) L-Glutamate + TMPO (3 mM)
GSH (pmol//~g protein) 33.3 + 1.9 + 1.6 _ 3.0 +
9.3 0.3 0.1 1.4
C6 glioma cells were cultured with glutamate alone or in combination with catalase or TMPO. Cellular levels of GSH 20 h after treatment were measured by a highperformance liquid chromatography method (see Experimental Procedures). Catalase and TMPO did not restore GSH depletion by glutamate. Each value is the mean +__S.D. (n = 4).
stained nuclear D N A with H33258 in the presence or absence of glutamate. Glutamate exposure (10 mM) induced condensation and aggregation of chromatin masses and subsequent D N A fragmentation (Fig. 1B) 20h after treatment as compared with control (Fig. 1A). The morphological changes in nuclei were initiated 14-16 h after glutamate treatment and thereafter the number of abnormal nuclei with condensation was gradually increased (Fig. 1B). In the presence of radical scavengers (catalase, TMPO, PBN) and vitamin E, the morphology of nuclei was almost the same as in controls (Fig. 1C). DISCUSSION
Accumulation of lipid peroxides in the glutamate toxicity To identify the cellular targets of oxygen radicals we measured biochemical oxidations using subcellular fractions (membrane, cytosol, nuclei). First, membrane oxidation was assayed using membrane lipid peroxides. Addition of glutamate induced about twofold increase of lipid peroxide accumulation as compared with control 20 h after culture (Table 5). The scavengers catalase and TMPO almost completely blocked the increase of lipid peroxide accumulation induced by glutamate. However, scavengers alone did not affect lipid peroxide accumulation (data not shown).
Target molecules of reactive oxygen species in the cytoplasm Three biochemical markers (ATP content, glutamine synthetase, G A P D H ) were selected to evaluate the cytoplasmic targets of ROS. First, we measured cellular levels of ATP after glutamate treatment to monitor the function of mitochondria. A significant reduction of cellular level of ATP (59.8% of control) could be seen 20 h after glutamate exposure (Table 6). This reduction of ATP was blocked by catalase, TMPO and vitamin E. Second, we assayed activity of glutamine synthetase, which is a cytosolic glial marker enzyme. Glutamine synthetase activity was not changed 20h after glutamate exposure (Table 6). Radical scavengers (catalase, TMPO and vitamin E) also did not modify glutamine synthetase activity. Third, we measured the activity of G A P D H after glutamate treatment. G A P D H is a key enzyme of glycolysis and is one of the most enriched cytosolic enzymes. The activity of G A P D H was not decreased but rather slightly enhanced 20 h after glutamate exposure. Radical scavengers did not block the increase of G A P D H activity.
Nuclear DNA damage in glutamate toxicity In a previous paper, 19 we reported fragmentation of cellular D N A in glutamate-treated C6 cells by agarose gel electrophoresis. In the present study, we
Accumulation of oxygen radicals in glutamate toxicity of C6 glioma cells Our previous work ~9suggested that glutamate toxicity of C6 glioma cells through the x~ antiporter system was due to oxidative stress after GSH depletion. The purpose of this work was to further define cytotoxic oxidative stress. A significant increase of ROS accumulation could be seen at 16 h after glutamate treatment (Table 1). Using specific radical scavengers (catalase, TMPO and PBN), we could detect the involvement of O~-, H202 and "OH in glutamate toxicity through the xE antiporter system. Nitrones such as TMPO and arylnitrones such as PBN are the only spin traps currently suitable for detection of hydroxyl and superoxide radicals.I° They are also cell-permeable and can directly trap O~- and "OH. Recent reports have demonstrated the neuroprotective effects of TMPO and PBN on other forms of neurotoxicity, such as N M D A - or ischemiainduced cell death in rodent cerebellar neurons. 22,46 Exogenously applied catalase, an H202-removing enzyme, efficiently protected C6 cells against glutamate toxicity; superoxide dismutase, an 02-removing enzyme, did not (Table 3). Since H202 is freely diffusible across membranes, ~3the reduction of extracellular H202 by catalase would effectively lower H202 within cells. On the other hand, the life span of O~- is very short and thus 0 2 cannot penetrate the membrane. Therefore, exogenously applied superoxide dismutase cannot scavenge excessive 0 2 Table 5. Effect of antioxidative agents on lipid peroxidation in the glutamate toxicity of C6 glioma cells Treatment Control L-Glutamate (10 mM) L-Glutamate + catalase (100 #g/ml) L-Glutamate + TMPO (3 mM)
Lipid peroxide (pmol/#g protein) 0.9 + 0.08 1.7 + 0.24* 1.12 ___0.12"* 1.05 _ 0.18'*
Lipid peroxides in the C6 cells were determined 20 h after treatment with 10mM L-glutamate or 10mM L-glutamate plus catalase or TMPO. L-Glutamate induced an increase of accumulation of lipid peroxides and this increase was suppressed by catalase or TMPO. Each value is the mean + S.D. (n = 3). *P < 0.01 compared with control. **P < 0.05 compared with L-glutamate.
Oxidative stress and glutamate toxicity
205
Table 6. Cytosolic markers after glutamate treatment with or without ant±oxidative agents
Treatment Control L-Glutamate (10 mM) L-Glutamate + catalase (100 #g/ml) L-Glutamate + TMPO (3 mM) L-Glutamate + vitamin E (100/~M)
Glutamine synthetase ATP content activity GAPDH activity (pmol//~g protein) (pmol//~g protein per h) (mU/#g protein) 44.8 _ 26.8 _ 36.3 ± 35.8 ± 34.1 ±
6.1 8.2* 7.9 4.2 5.1
2.98 _ 0.87 2.37 __+0.56 3.08 __+0.48 2.35 __+0.52 2.65 ± 0.67
8.2 ± 12.1 _ 9.5 + 12.0 ± 10.5 ±
1.1 2.2** 1.3 2.0** 1.1"*
C6 glioma cells were cultured with glutamate alone or in combination with catalase, TMPO or vitamin E. Cellular ATP content and activities of glutamine synthetase and GAPDH were measured at 20 h after treatment using bioluminescence, high-performance liquid chromatographic and spectrophotometric methods, respectively (see Experimental Procedures). Glutamate induced a reduction of ATP content and a slight increase of GADPH activity, whereas glutamine synthetase activity was not changed by glutamate treatment. The reduction of cellular ATP content was prevented by radical scavengers and ant±oxidant. Each value is the mean ___S.D. (n = 34). *P < 0.01 and **P < 0.05 compared with control. generated inside of the cell. Sfiez et al. 39 reported in cervical ganglion neurons in culture that superoxide dismutase protects cultured neurons against death by substrate deprivation. This discrepancy with our data may be ascribed to different experimental conditions and different cell types. Cellular target molecules sensitive to oxygen radicals Membranes. Membranes consist mainly of lipids and proteins. Membrane lipid peroxidation has been proposed to be a consequence of oxidative stress. Unsaturated fatty acids in membrane lipids are particularly susceptible to oxygen radicals, because double bonds are thought to undergo peroxidation through oxidative chain reactions. 42 A two-fold increase of lipid peroxide accumulation in C6 cells was detected 20 h after glutamate treatment. The increase of lipid peroxides was completely protected by radical scavengers (Table 5) or the ant±oxidant vitamin E. 4° This fact suggests that membrane lipid is one of the main targets of oxygen radicals in glutamate toxicity. In brain ischemia, whose pathogenesis is also connected to excess release of glutamate, accumulation of lipid peroxides has been reportedJ '28 Cytoplasm. The susceptibility of cytoplasm to oxygen radicals was tested by measuring cellular levels of ATP and activities of the cytosolic enzymes, glutamine synthetase and GAPDH. ATP content was measured to estimate the function of mitochondria, as mitochondrial respiration is an important source of O{ and H202 generation. 5 Therefore, mitochondria might be expected to be target organelles for oxygen radicals. Indeed, GSH deficiency leads to mitochondrial damage in rat brain. 15 In our present study, cellular ATP (40.2%) was significantly reduced 20 h after glutamate treatment. This reduction of ATP was almost restored by oxygen radical scavengers or the ant±oxidant vitamin E. The activities of two cytosolic enzymes, glutamine synthetase and GAPDH, were selected to detect representative target molecules of oxygen radicals. The histidine residue of glutamine synthetase and the cysteine residue of G A P D H are well known to be inactivated by oxygen radicals, especially H2 02.35,38 A NSC 73/1--H
glial marker enzyme, glutamine synthetase has also been used as a biochemical indicator of glial cell damage induced by the glutamate analog ~-aminoad±pate in the fish retina. 2° The activity of glutamine synthetase in C6 glioma cells was not decreased 20 h after glutamate treatment (Table 6), suggesting that glutamine synthetase is not a target molecule of oxygen radicals. Our previous study showed that the glutamine synthetase activity of C6 glioma cells was much lower (1%) than that of retinal glial cells. TM Therefore, function of glutamine synthetase between glial and glioma cells may be quite different in different environments. Surprisingly, activity of G A P D H was not decreased but rather slightly increased after glutamate treatment. In C6 glioma cells, G A P D H may not be a target for oxygen radicals. If the reduction of cellular ATP after glutamate treatment as mentioned above indicates mitochondrial dysfunction, the G A P D H activation by glutamate treatment may reflect a compensatory production of ATP through anerobic glycolysis. Nuclei. It is well known that oxygen radicals directly attack the nuclei and cause single- and double-strand D N A breaks. 12 Furthermore, oxygen radicals induce an increase of intracellular Ca 2+, causing an activation of calcium-dependent endonuclease resulting in D N A fragmentations, a6 Recently, Ratan et al. 37 reported that oxidative stress also induced apoptosis in embryonic cortical neurons. In the present study, we demonstrated apoptotic-like D N A fragmentation with H33258 staining (Fig. 1). The apoptotic D N A fragmentation showed typical internucleosomal D N A 45 cleavage with de novo protein and RNA synthesis. Whether or not the glutamate cell death in our C6 glioma cells is an apoptotic cell death remains unresolvedJ 4 CONCLUSION In this study, we demonstrate a strong correlation between the ROS accumulation and glutamateinduced cell death of C6 glioma cells. Three types of ROS ( 0 2 , H202 and 'OH) are generated from the four-step reduction of molecular oxygen. These
206
K. Mawatari et al.
reactions are catalysed by transition metal ions such as iron, which provide electrons at each step. 33 Highly reactive 'OH is formed from H202 with reaction with
iron ions through the Fenton type Harber-Weiss reaction. ~3These reactions and sequent accumulation of cytotoxic ROS may be promoted in glutamate
Fig. 1. DNA specific staining with H33258 after glutamate treatment. (A) Cells (4.0 x 105) were plated and incubated in standard DMEM for 20 h. (B) When 10 mM L-glutamate was added to the medium at the beginning of the culture, condensation of chromatin masses and aggregation at the nuclear membrane progressed and almost all nuclei changed 20 h after treatment. (C) In the presence of catalase (100 #g/ml), TMPO (3 mM), PBN (3 mM) or vitamin E (100 p M) with glutamate, the morphology of nuclei was similar to that of controls. Scale bar = 10/~m.
207
Oxidative stress and glutamate toxicity
Catalase
.2~ ~
1
.~TMPO
I
R 0 S - . ~ . ...... G S H
2",H202"~ OH) •
Mitochondria
~
C6 CELL
A Cysteine
A Cystine
ROS : Reactive Oxygen Species Vit. E : Vitamin E
Cystine Fig. 2. A scheme for oxidative stress by glutamate toxicity of C6 glioma cells via the x~ antiporter system and for protection by radical scavengers and antioxidants. Mitochondrial respiration is an important source of 0 2 and H202. After GSH depletion, the excessively accumulating H202 is catalysed by iron ion to form highly reactive "OH by the Fenton type Harber-Weiss reaction. In the present study, catalase may scavenge excessive H202 outside the cell, while TMPO and PBN can directly trap O~- and "OH inside the cell. Vitamin E is effective as a radical chain-breaking antioxidant. Therefore, these antioxidative agents can protect C6 cells against oxidative stress by glutamate exposure.
toxicity of C6 glioma cells. In Fig. 2, we summarize ROS accumulation and multiple cellular injuries by ROS during glutamate cell death of C6 glioma cells via the x~ antiporter system. After GSH depletion, ROS accumulation was increased, leading through multiple cellular damage to cell death. Cell death and cell damage were completely prevented by oxygen radical scavengers and an antioxidant. Our in vitro C6 glioma cell system offers a very useful model to analyse both the cellular target molecules to ROS and
protection mechanisms for ROS-induced neurodegenerative disorders)
Acknowledgements--We thank Mr S. Ishita for technical
assistance and Mrs Tami Urano for secretarial assistance. This work was supported in part by research grants (Nos 06808084 and 07558105 to S.K.) from the Ministry of Education, Science and Culture of Japan, from the Sasakawa Health Science Foundation and from the Mochida Pharmaceutical Science Foundation, 1993.
REFERENCES
1. Bannai S. (1986) Exchange ofcystine and glutamate across plasma membrane of human flbroblasts. J. biol. Chem. 261, 2256-2263. 2. Bannai S. and Kitamura E. (1980) Transport interaction of L-cystine and L-glutamate in human diploid fibroblasts in culture. J. biol. Chem. 255, 2372-2376. 3. Behl C., Davis J. B., Lesley R. and Schubert D. (1994) Hydrogen peroxide mediates amyloid fl protein toxicity. Cell 77, 817-827. 4. Bromont C., Marie C. and Bralet J. (1989) Increased lipid peroxidation in vulnerable brain regions after transient cortex ischemia in rats. Stroke 20, 918-924. 5. Chance B., Sies H. and Boveris A. (1979) Hydroperoxide metabolism in mammalian organs. Physiol. Rev. 59, 527-605. 6. Cho Y. and Bannai S. (1990) Uptake of glutamate and cystine in C6 glioma cells and in cultured astrocytes. J. Neurochem. 53, 2091-2097. 7. Choi D. W. (1988) Glutamate neurotoxicity and disease of the nervous system. Neuron 1, 623-634. 8. Coyle J. T. and Puttfarcken P. S. (1993) Oxidative stress, glutamate and neurodegenerative disorders. Science 262, 689-694. 9. Dimonte D. A., Chan P. and Sandy M. S. (1992) Glutathione in Parkinson's disease: a link between oxidative stress and mitochondrial damage? Ann. Neurol. 32, S1 l l-S115. 10. Finkelstein E., Rosen G. M. and Rauckman E. J. (1980) Spin trapping of superoxide and hydroxyl radical: practical aspects. Archs Biochem. Biophys. 200, 1-16. 11. Giblin F. J., Reddan J. R., Schrimscher L., Dziedzic D. C. and Reddy V. N. (1990) The relative roles of the glutathione redox cycle and catalase in the detoxification ofH202 by cultured rabbit lens epithelial cells. ExplEye Res. 50, 795-804. 12. Halliwell B. (1992) Reactive oxygen species and the central nervous system. J. Neurochem. 59, 1609-1623.
208
K. Mawatari et al.
13. Halliwell B. and Gutteridge J. M. C. (1989) Free Radicals in Biology and Medicine, pp. 188-266. Oxford University Press, Oxford. 14. Higuchi Y., Kato S. and Matsukawa S. (1994) Involvement of protein kinase C in apoptosis induced by glutamate in C6 rat glioma cells. Proc. XVlth Int. Cancer Congress 857-861. 15. Jain A., Martensson J., Stole E., Auld P. A. M. and Meister A. (1991) Glutathione deficiency leads to mitochondrial damage in brain. Proc. natn. Acad. Sci. U.S.A. 88, 1913-1917. 16. Kato S., Higashida H., Higuchi Y., Hatakenaka S. and Negishi K. (1984) Sensitive and insensitive states of cultured glioma cells to glutamate damage. Brain Res. 303, 365-373. 17. Kato S., Ishita S., Sugawara K. and Mawatari K. (1993) Cystine/glutamate antiporter expression in retinal Miiller glial cells: implications for DL-Ct-aminoadipate toxicity. Neuroscience 57, 473~,82. 18. Kato S., Negishi K., Honma K., Sakai K. and Shimada Y. (1989) A high performance liquid chromatography assay for glutamine synthetase. Neurochem. Int. 14, 491496. 19. Kato S., Neglshi K., Mawatari K. and Kuo C. H. (1992) A mechanism for glutamate toxicity in the C6 glioma cells involving inhibition of cystine uptake leading to glutathione depletion. Neuroscience 48, 903-914. 20. Kato S., Sugawara K., Matsukawa T. and Negishi K. (1990) Gliotoxic effects of ct-aminoadipic acid isomers on the carp retina: a long-term observation. Neuroscience 36, 145-153. 21. Kontos H. A. (1989) Oxygen radicals in CNS damage. Chem. Biol. Interact. 72, 229-255. 22. Lafon-Cazal M., Pietri S., Culcasi M. and Bockaert J. (1993) NMDA-dependent superoxide production and neurotoxicity. Nature 364, 535-537. 23. Lebel C. P. and Bondy S. C. (1990) Sensitive and rapid quantitation of oxygen reactive species formation in rat synaptosomes. Neurochem. Int. 17, 435-440. 24. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-275. 25. Mawatari K., Sugitani K., Yasui Y. and Kato S. (1994) Cellular mechanism of glutamate damage in C6 glioma cells. Bull. Jap. Soc. Neurochem. 33, 252-253. 26. Meister A. (1988) Glutathione metabolism and its selective modification. J. biol. Chem. 263, 17205-17208. 27. Meldrum B. (1993) Amino acids as dietary excitotoxins: a contribution to understanding neurodegenerative disorders. Brain Res. Rev. 18, 293-314. 28. Michel H. S., Vaishnav Y. N., Kempski O., von Lubitz D., Weiss J. F. and Feuerstein G. (1987) Breathing 100% oxygen after global brain ischemia in Mongolian gerbils resulting in increasedlipid peroxidation and increased mortality. Stroke 18, 426-430. 29. Murphy T. H., Miyamoto M., Sastre A., Schnaar R. L. and Coyle J. T. (1989) Glutamate toxicity in a neuronal cell line involves inhibition of cystine transport leading to oxidative stress. Neuron 2, 1547-1558. 30. Murphy T. H., Schnaar R. L. and Coyle J. T. (1990) Immature cortical neurons are uniquely sensitive to glutamate toxicity by inhibition of cystine uptake. Fedn Am. Socs exp. BioL 4, 1624-1633. 31. Oberhammer F. A., Pavelka M., Sharma S., Tiefenbacher R., Purchio A. F., Bursh W. and Schulte-Hermann R. (1992) Induction of apoptosis in cultured hepatocytes and in regressing liver by transforming growth factor fl I. Proc. natn. Acad. Sci. U.S.A. 89, 5408-5412. 32. Oka A., Belliveau M. J., Rosenberg P. A. and Volpe J. J. (1993) Vulnerability of oligodendroglia to glutamate: pharmacology, mechanism and prevention. J. Neurosci. 13, 1441-1453. 33. Olanow C. W. (1992) An introduction to the free radical hypothesis in Parkinson's disease. Ann. Neurol. 32, Suppl., $2-$9. 34. Olanow C. W. (1994) A radical hypothesis for neurodegeneration. Trends Neurosci. 16, 439-444. 35. Oliver C. N., Starke-Reed P. E., Stadtman E. R., Liu G. J., Carney J. M. and Floid R. A. (1990) Oxidative damage to brain proteins, loss of glutamine synthetase activity, and production of free radicals during ischemia/reperfusion-induced injury to gerbil brain. Proc. natn. Acad. Sci. U.S.A. 87, 5144-5147. 36. Orrenius S., McConkey D. J., Bellomo G. and Nicotera P. (1989) Role of Ca 2+ in toxic cell killing. Trends pharmac. Sci. I0, 281-285. 37. Ratan R. R., Murphy T. H. and Baraban J. M. (1994) Oxidative stress induces apoptosis in embryonic cortical neurons. J. Neurochem. 62, 376-379. 38. Reddan J. R., Sevilla M. D., Giblin F. J., Padgaonkar V., Dziedzic D. C., Leverenz V., Misra I. C. and Peters J. L. (1993) The superoxide dismutase mimic TEMPOL protects cultured rabbit lens epithelial cells from hydrogen peroxide insult. Expl Eye Res. 56, 543-554. 39. S~ez J. C., Kessler J. A., Bennett M. V. L. and Spray D. C. (1987) Superoxide dismutase protects cultured neurons against death by starvation. Proc. natn. Acad. Sci. U.S.A. 84, 3056-3059. 40. Shinagawa S. (1994) Serotonin protects C6 glioma cells from glutamate toxicity. Neuroscience 59, 1043-1050. 41. Siesj6 B. K., Aghardh C. D. and Bengtsson F. (1989) Free radicals and brain damage. Cerebr. Brain Metab. Rev. 1, 165-211. 42. Slater T. F. (1984) Free radical mechanisms in tissue injury. Biochem. J. 222, 1-15. 43. Sofic E., Paulus W., Jellinger K., Riedever P. and Youdium M. B. (1991) Selective increase of iron in substantia nigra zona compacta of parkinsonian brain. J. Neurochem. 56, 978-982. 44. Stanley P. E. and Williams S. G. (1969) Use of the liquid scintillation spectrometer for determining adenosine triphosphate by the luciferase enzyme. Analyt. Biochem. 29, 381-392. 45. Wyllie A. H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555-556. 46. Yue T. L., Gu J. L., Lysko P. G., Cheng H. Y., Barone F. C. and Feuerstein G. (1992) Neuroprotective effects of phenyl-t-butyl-nitrone in gerbil global brain ischemia and in cultured rat cerebellar neurons. Brain Res. 574, 193-197. (Accepted 10 January 1996)
Pergamon
Neuroscience Vol. 73, No. 1, pp. 201-208, 1996 Copyright© 1996 IBRO. Publishedby ElsevierScienceLtd Printed in Great Britain S03 06-4522(96)00025- 5 0306-4522/96 $15.00+ 0.00
REACTIVE OXYGEN SPECIES INVOLVED IN THE GLUTAMATE TOXICITY OF C6 GLIOMA CELLS VIA Xt~ ANTIPORTER SYSTEM K. MAWATARI,* Y. Y A S U I , t K. SUGITANI,* T. TAKADERA:~ and S. KATOt§ *Department of Laboratory Sciences, tDepartment of Neurobiology, NIRI, University of Kanazawa Faculty of Medicine, Kanazawa 920, Japan :~Hokuriku University, Faculty of Pharmaceutical Science, Kanazawa 920-11, Japan Abstract--We recently demonstrated that continuous L-glutamate exposure led to cell death in C6 glioma cells over a period of 24-36 h, due to inhibition of cystine uptake through the cystine/glutamate (xr) antiporter. The antioxidant vitamin E provided protection against this effect, supporting the hypothesis that depletion of glutathione might be responsible, resulting from insufficient cystine uptake. To clarify the content of oxidative stress after glutathione depletion, the present study was done to investigate accumulation and target molecules of reactive oxygen species induced by glutamate treatment. The accumulation of reactive oxygen species was increased three-fold as compared to a control culture. Membrane oxidation, as judged by lipid peroxidation, was increased two-fold after glutamate treatment. Cellular ATP content was significantly reduced by glutamate exposure. For the two cytosolic enzymes examined, activity of glyceraldehyde 3-phosphate dehydrogenase was slightly enhanced by glutamate treatment, while activity of glutamine synthetase was not changed. Impairment of nuclear DNA after glutamate exposure was also revealed by nuclear chromatin condensation with DNA fragmentation. Thus, the multiple targets (membrane, cytoplasm and nuclei) of oxygen radicals in glutamate toxicity through the xe antiporter system were evaluated for the first time. Furthermore, prevention from cell death and from cellular toxicity induced by oxygen radicals could be seen using three specific oxygen radical scavengers, catalase, 3,3,5,5-tetramethyl-pyrroline N-oxide and ~-phenyl-N-t-butylnitrone, without restoring the glutathione deficit. This indicates that radical scavengers did not interact with the xe antiporter system, but directly scavenged the oxygen radicals. Taken together, the data strongly suggest that O ; , H202 and 'OH accumulate in response to oxidative stress after glutathione depletion, resulting in glutamate cell death of C6 glioma cells. Copyright © 1996 IBRO. Published by Elsevier Science Ltd. Key words: cell death, glutathione depletion, H202 accumulation, lipid peroxidation, radical scavengers.
Recent review articles have separated glutamate toxicity into receptor-mediated and transporter-mediated types. 7's'27 The former (receptor-mediated) is subdivided into N-methyl-D-aspartate (NMDA) and non-NMDA receptor types. The mechanism for the toxicity is thought to be due to membrane depolarization (excitation) accompanying cation influx, especially of calcium ions, by glutamate receptor activation. 7 The latter (transporter-mediated) is unique and requires cellular expression of the cystine/glutamate antiporter (x~). 2 The x~ system transports cystine inward and glutamate outward with the glutamate gradient providing the driving §To whom correspondence should be addressed. Abbreviations: DCF, 2',7'-dichlorofluorescin; DCFH-DA, 2,7-dichlorofluorescin diacetate; DMEM, Dulbecco's modified Eagle's medium; EDTA, ethylenediamine tetra-acetate; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GSH, glutathione (7-glutamylcysteinylglycine); HEPES, N-2-hydroxyethylpiperazine-N'-2ethanesulfonic acid; LDH, lactate dehydrogenase; NMDA, N-methyl-D-aspartate; PBN, ~t-phenyl-N-tbutylnitrone; ROS, reactive oxygen species; TMPO, 3,3,5,5-tetramethyl-pyrroline N-oxide.
force. The cystine taken up is intracellularly reduced to cysteine and then utilized to synthesize glutathione (y-glutamylcysteinylglycine, GSH). Therefore, cellular levels of GSH are maintained through the xE antiporter system in some cells. ~ The non-excitatory mechanism for the glutamate toxicity through the xE antiporter is due to inhibition of cystine uptake in the constant ( ~ 2 4 h ) and high (depending upon the concentration of extracellular cystine 19,29) exposure of glutamate. This inhibition leads to GSH depletion, finally resulting in a delayed type cell death, s,19 It is expected that glutamate cell death results from oxidative stress after GSH depletion, because GSH is a major non-protein thiol compound in the living cells and is a key molecule for protection and conjugation against various oxidative stress or toxins. 26 Recently, glutamate toxicity through the x~ antiporter has been reported in many CNS neurons and glia. 6'17'19'29'3°'32 Pathogenesis of neurodegenerative disorders including stroke, 4,21 Parkinson's disease9.43 and Alzheimer's disease3,12 has implicated oxygen radicals.34,41 The brain has a large oxygen consumption and thus has many opportunities to be exposed to
201
202
K. Mawatari et al.
oxidant overload. However, which reactive oxygen species (ROS) are most involved in the disorders remain to be clarified. To answer this question, we developed a n in vitro cell system. W e previously used a rat clonal cell line o f C6 glioma cells to investigate the toxicity a n d protective m e c h a n i s m s for glutamate cell d e a t h mediated via the x~ a n t i p o r t e r system. '9'4° The toxic action o f glutamate (10 m M ) o n C6 glioma cells could be prevented by cystine (0.1-1 m M ) a n d vitamin E ( 1 0 0 / t M ) . The decrease o f cellular G S H was initiated early (4-6 h) after glutamate t r e a t m e n t a n d the c o n c e n t r a t i o n o f G S H was decreased 10% of control at 12 h after glutamate exposure. However, m o r p h o l o g i c a l or biochemical signs of cell d e a t h were initiated later ( 1 3 - 1 4 h ) after g l u t a m a t e t r e a t m e n t a n d almost all cells died at 24-36 h after g l u t a m a t e exposure. 19 By utilizing the relatively long delay between G S H depletion a n d cell death, the present study has clarified oxidants accumulated in the course o f cell death. A preliminary a b s t r a c t o f this work has been published elsewhere. 25
EXPERIMENTAL PROCEDURES
Chemicals
Monosodium L-glutamate was purchased from Ajinomoto-Takara Corp. (Tokyo, Japan). Vitamin E was obtained from Wako Pure Chemicals (Osaka, Japan). Catalase, superoxide dismutase, 3,3,5,5-tetramethylpyrroline N-oxide (TMPO) and ct-phenyl-N-t-butylnitrone (PBN) were obtained from Sigma (St Louis, MO, U.S.A.). Cell culture
C6 glioma cells were grown in Dulbecco's modified Eagle's medium (DMEM; Gibco) containing 5% fetal calf serum with penicillin (100units/ml) and streptomycin (100/~g/ml) at 37°C in a humidified atmosphere of 90% air/10% CO2, as described previously26 The concentration of extracellular cystine in the standard DMEM was about 200#M. Exponentially growing cells were dissociated and plated on 35-mm dishes (Falcon) at a density of 4 x 10~cells/dish [for lactate dehydrogenase (LDH) assay and DNA staining] or on 60-mm dishes (Falcon) at a density of 1.5 × 106 cells/dish (for other assays). Test reagents such as glutamate or other chemicals were simply added to the culture medium at the beginning of the culture. Cell viability (cytotoxicity) was assessed by the ratio of released LDH into the medium per total LDH activity, as described previously.4° The effects of drugs on C6 glioma cells were also observed morphologically with a phase-contrast microscope (Nikon, TMD-2). Biochemical assay Reactive oxygen species. ROS were assayed using 2',7'dichlorofluorescin diacetate (DCFH-DA; Molecular Probes, Eugene, OR, U.S.A.) which is de-esterified within cells to the ionized free acid, 2',7'-dichlorofluorescin. This is trapped within cells and thus accumulated. 2',7'-Dichlorofluorescin is capable of being oxidized to fluorescent 2',7'-dichlorofluorescein (DCF) by ROS. ROS accumulation was measured by a modification of the previously described method of Lebel and Bondy) 3 In brief, C6 glioma cells were collected, washed with Hank's balanced salt solution, pH 7.4, and resuspended in a total volume of 2.0 ml of the same buffer. The cell suspension was divided into two tubes, one for ROS assay and the other for protein
quantification. The cells were diluted in the same volumes of Hank's balanced salt solution and then incubated with 10pM DCFH-DA at 37°C for 15 min. After loading with DCFH-DA, the cell suspension was centrifuged at 3000 g for 5 min (4°C) and the pellet was resuspended in 2 ml of the same ice-cold buffer, and then the fraction was incubated for a further 60 min (37°C). At the beginning and end of incubation, fluorescence was measured at 488 nm excitation and 525 nm emission (Hitachi, F-2000). Autofluorescence was corrected by the inclusion of parallel blanks without DCFH-DA. The correction for autofluorescence was always less than 10% of the total. ROS accumulation was quantitated from a DCF (Wako Pure Chemicals) standard curve (10-600nM) and results were expressed as nmol DCF formed/h//~g protein. Protein concentration was determined by the method of Lowry et al. 24 Glutathione. Cellular levels of GSH were assayed by a high-performance liquid chromatography method with thiol group-specific fluorogenic reagents, as described elsewhere.19 Lipid peroxide. The content of lipid peroxides was determined fluorometrically using the thiobarbituric acid method described previously. 4° A TP. The level of ATP was determined using a luciferindependent bioluminescence method. 44 C6 glioma cells were collected and an aliquot of cell suspension was diluted twice in 0.6 N perchloric acid solution with 2 mM EDTA, then centrifuged at 12,000g for 2min at 4°C and the supernatant was neutralized with 3 M K2CO 3. The assay mixture consisted of 20 pl of the supernatant, 180/~1 of 25mM HEPES-NaOH buffer, pH7.8, and 100/~1 of luciferin luciferase reagents (Kikkoman Ltd). The concentration of ATP was calculated from an ATP standard curve (0.1-100 nM). Glutamine synthetase. The activity of glutamine synthetase was assayed by a high-performance liquid chromatography method with postcolumn derivatization using o-phthalaldehyde.tS Glyceraldehyde 3-phosphate dehydrogenase. The activity of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was assayed by a modification of the previously described method of Giblin et al. It In brief, cells were collected, centrifuged at 3000 g for 5 min and homogenized in a total volume of 1.0ml of cold 20mM triethanolamine buffer, pH 7.6, containing 0.4 mM EDTA. The assay mixture consisted of 75 mM triethanolamine, 1.5 mM EDTA, 6.0 mM 3-phosphoglyceric acid (Boehringer Mannheim), 1,I mM ATP, 1.3mM MgSO4, 0.2mM reduced form of nicotinamide adenine dinucleotide (NADH; Oriental Kobo, Ltd) and 20 units of phosphoglycerate kinase (Boehringer Mannheim). The activity of GAPDH was assayed spectrophotometrically and expressed as the ratio of the absorbance reduction of NADH at 340 nm as compared with that of the standard GAPDH enzyme (8000U/ml; Boehringer Mannheim). DNA staining. Nuclear DNA staining was performed with a specific fluorescent probe, H33258. 3~ Prior to cultivation, a small cover glass (22 x 22 mm 2) was put into a 35-mm dish, onto which C6 glioma cells were seeded and cultured. After removal of the medium, cells attached on the cover glass were gently fixed with 1.0 ml of 10% formalin neutral buffer solution (pH 7.4) for 5 min, washed with distilled water and 2.0 ml of water added. Cells were stained with H33258 (Molecular Probes, Eugene, OR, U.S.A.) at a final concentration of 8 #g/ml for 5 min. The cells were then washed in water and mounted with glycerol. Fluorescence was observed and photographed using a fluorescence microscope (Nikon UFX-IIA) at 370 nm excitation and 500 nm emission. Statistics
Three to five different experiments for biochemical assays were repeated. Results are expressed as the mean + S.D. and statistical differences were evaluated using Student's t-test.
Oxidative stress and glutamate toxicity Table I. Increased rate of 2',7'-dichlorofluorescin formation after glutamate treatment DCF formed (nmol/#g protein/h) Time (h) 8 16 22
C6
L-Glutamate, 10 mM
0.21 +_.0.06 0.21 __.0.05 0.19 + 0.03
0.30 _.+0.06 0.58 _.+0.16" 0.63 + 0.14"
C6 glioma cells were cultured in standard DMEM with 10mM L-glutamate and DCF formation was measured with a fluorescent probe, DCFHDA. The rate of DCF formation was significantly increased (about three times) 16 and 22 h after glutamate treatment. E a c h value is the mean _+S.D. (n = 3-5). *P < 0.01 compared with control.
RESULTS
Accumulation of oxygen radicals in glutamate toxicity To clarify the involvement of oxygen radicals in the glutamate toxicity of C6 glioma cells, we first measured accumulation of ROS after glutamate exposure. ROS accumulation was estimated utilizing a converting reaction of DCFH-DA to DCF oxidized by ROS. The rate of generation of DCF in the C6 cells increased time-dependently (Table 1). ROS accumulation was enhanced 2.8-fold 16 h and maximally 3.3-fold 22 h after glutamate treatment. To further confirm the participation of ROS in the glutamate toxicity, we next investigated the effects of oxygen radical scavengers on DCF formation. We used TMPO and PBN as spin-trapping agents specific to O~- and "OH. In the presence of these scavengers (TMPO, 3mM; PBN, 2mM) with glutamate (10mM), the increase of ROS accumulation was clearly suppressed 22 h after treatment (Table 2). Vitamin E (100/~M), a powerful antioxidant, was even more effective than TMPO and PBN (Table 2). These data strongly suggest that oxygen radicals substantially accumulate within C6 cells after GSH depletion.
Table 2. Protective effect of radical scavengers and antioxidant on the rate of 2',7'-dichlorofluorescin formation after glutamate treatment Treatment Control L-Glutamate (10 mM) L-Glutamate 4- TMPO (3 mM) L-Glutamate 4- PBN (2 mM) L-Glutamate 4- vitamin E (100~M)
DCF formed (nmol/pg protein/h) 0.19 + 0.03 0.63 + 0.14 0.43 ___0.07* 0.38 + 0.03* 0.15 __+0.03**
C6 cells were cultured with 10 mM L-glutamate or with 10 mM L-glutamate plus either TMPO, PBN or vitamin E. TMPO (3mM), PBN (2mM) and vitamin E (100/~M) significantly suppressed the increased rate of DCF formation 22 h after glutamate treatment. Each value is the mean +__S.D. (n = 4-6). *P < 0.05 and **P < 0.01 compared with glutamate treatment.
203
Table 3. Protective effect of radical scavengers and antioxidant on glutamate cell death Treatment Control L-Glutamate (10 mM) L-Glutamate + catalase (100 pg/ml) L-Glutamate + TMPO (3 mM) L-Glutamate + PBN (2 mM) L-Glutamate 4- superoxide dismutase (450 units/ml)
Cell death (%) 3.3 _+ 1.2 65.1 _+4.7 13.0 + 4.0* 15.0 __+4.0*
6.4 _+ 1.6" 62.8 _+5.1
C6 cells were cultured with 10mM L-glutamate or with l0 mM L-glutamate plus catalase, TMPO, PBN or superoxide dismutase. When 10mM L-glutamate was added to the medium at the beginning of the culture, almost all cells died 24-36 h later. The presence of catalase (100/tg/ml), TMPO (3 mM) or PBN (2 mM) suppressed this glutamate-induced cell death, whereas the cell death was unaffected by the presence of superoxide dismutase (450 units/m1).Cell death (%) 36 h after culture was determined by LDH assay. Each value is the mean +_S.D. of three to five different experiments with duplicate samples. *P < 0.01 compared with glutamate treatment.
Protection from glutamate cell death by oxygen radical scavengers In order to examine the protective effect of radical scavengers, either catalase (100/~g/ml), an H202removing enzyme, TMPO (3 mM) or PBN (2 mM) was added to the culture medium in combination with 10mM glutamate at the beginning of culture. Cell viability (cell death) of C6 glioma cells was evaluated by LDH release into the medium 36 h after culture (see Experimental Procedures). Addition of these agents completely suppressed glutamateinduced cell death (Table 3). The minimum effective doses of catalase, TMPO and PBN were 30/~g/ml, 1 mM and 1 mM, respectively. When C6 glioma cells were treated with these scavengers alone, no significant change in morphology, growth or LDH release could be seen (data not shown). In contrast to catalase, superoxide dismutase (1-500units/ml), a superoxide anion-removing enzyme, did not suppress cell death (Table 3). These data strongly suggest that accumulation of ROS (O~-, H202 and 'OH) participates in the glutamate cell death of C6 glioma cells.
Lack of restoration of the glutathione deficit with oxygen radical scavengers To address the mechanism by which oxygen radical scavengers prevented glutamate-induced cell death, we measured cellular levels of GSH to exclude the possibility of restoration of GSH level by their scavengers, as occurs in the case of cystineJ 9 Cellular levels of GSH 20 h after combined treatment with glutamate and scavengers are shown in Table 4. These scavengers did not restore the GSH deficit induced by glutamate. The data strongly suggest that the oxygen radical scavengers do not modulate the x~ antiporter system, but directly interact with either the ROS and/or their intermediates.
K. Mawatari et al.
204
Table 4. Cellular levels of glutathione in C6 glioma cells treated with glutamate and antioxidative agents Treatment Control L-Glutamate (I0 raM) L-Glutamate + catalase (I00/~g/ml) L-Glutamate + TMPO (3 mM)
GSH (pmol//~g protein) 33.3 + 1.9 + 1.6 _ 3.0 +
9.3 0.3 0.1 1.4
C6 glioma cells were cultured with glutamate alone or in combination with catalase or TMPO. Cellular levels of GSH 20 h after treatment were measured by a highperformance liquid chromatography method (see Experimental Procedures). Catalase and TMPO did not restore GSH depletion by glutamate. Each value is the mean +__S.D. (n = 4).
stained nuclear D N A with H33258 in the presence or absence of glutamate. Glutamate exposure (10 mM) induced condensation and aggregation of chromatin masses and subsequent D N A fragmentation (Fig. 1B) 20h after treatment as compared with control (Fig. 1A). The morphological changes in nuclei were initiated 14-16 h after glutamate treatment and thereafter the number of abnormal nuclei with condensation was gradually increased (Fig. 1B). In the presence of radical scavengers (catalase, TMPO, PBN) and vitamin E, the morphology of nuclei was almost the same as in controls (Fig. 1C). DISCUSSION
Accumulation of lipid peroxides in the glutamate toxicity To identify the cellular targets of oxygen radicals we measured biochemical oxidations using subcellular fractions (membrane, cytosol, nuclei). First, membrane oxidation was assayed using membrane lipid peroxides. Addition of glutamate induced about twofold increase of lipid peroxide accumulation as compared with control 20 h after culture (Table 5). The scavengers catalase and TMPO almost completely blocked the increase of lipid peroxide accumulation induced by glutamate. However, scavengers alone did not affect lipid peroxide accumulation (data not shown).
Target molecules of reactive oxygen species in the cytoplasm Three biochemical markers (ATP content, glutamine synthetase, G A P D H ) were selected to evaluate the cytoplasmic targets of ROS. First, we measured cellular levels of ATP after glutamate treatment to monitor the function of mitochondria. A significant reduction of cellular level of ATP (59.8% of control) could be seen 20 h after glutamate exposure (Table 6). This reduction of ATP was blocked by catalase, TMPO and vitamin E. Second, we assayed activity of glutamine synthetase, which is a cytosolic glial marker enzyme. Glutamine synthetase activity was not changed 20h after glutamate exposure (Table 6). Radical scavengers (catalase, TMPO and vitamin E) also did not modify glutamine synthetase activity. Third, we measured the activity of G A P D H after glutamate treatment. G A P D H is a key enzyme of glycolysis and is one of the most enriched cytosolic enzymes. The activity of G A P D H was not decreased but rather slightly enhanced 20 h after glutamate exposure. Radical scavengers did not block the increase of G A P D H activity.
Nuclear DNA damage in glutamate toxicity In a previous paper, 19 we reported fragmentation of cellular D N A in glutamate-treated C6 cells by agarose gel electrophoresis. In the present study, we
Accumulation of oxygen radicals in glutamate toxicity of C6 glioma cells Our previous work ~9suggested that glutamate toxicity of C6 glioma cells through the x~ antiporter system was due to oxidative stress after GSH depletion. The purpose of this work was to further define cytotoxic oxidative stress. A significant increase of ROS accumulation could be seen at 16 h after glutamate treatment (Table 1). Using specific radical scavengers (catalase, TMPO and PBN), we could detect the involvement of O~-, H202 and "OH in glutamate toxicity through the xE antiporter system. Nitrones such as TMPO and arylnitrones such as PBN are the only spin traps currently suitable for detection of hydroxyl and superoxide radicals.I° They are also cell-permeable and can directly trap O~- and "OH. Recent reports have demonstrated the neuroprotective effects of TMPO and PBN on other forms of neurotoxicity, such as N M D A - or ischemiainduced cell death in rodent cerebellar neurons. 22,46 Exogenously applied catalase, an H202-removing enzyme, efficiently protected C6 cells against glutamate toxicity; superoxide dismutase, an 02-removing enzyme, did not (Table 3). Since H202 is freely diffusible across membranes, ~3the reduction of extracellular H202 by catalase would effectively lower H202 within cells. On the other hand, the life span of O~- is very short and thus 0 2 cannot penetrate the membrane. Therefore, exogenously applied superoxide dismutase cannot scavenge excessive 0 2 Table 5. Effect of antioxidative agents on lipid peroxidation in the glutamate toxicity of C6 glioma cells Treatment Control L-Glutamate (10 mM) L-Glutamate + catalase (100 #g/ml) L-Glutamate + TMPO (3 mM)
Lipid peroxide (pmol/#g protein) 0.9 + 0.08 1.7 + 0.24* 1.12 ___0.12"* 1.05 _ 0.18'*
Lipid peroxides in the C6 cells were determined 20 h after treatment with 10mM L-glutamate or 10mM L-glutamate plus catalase or TMPO. L-Glutamate induced an increase of accumulation of lipid peroxides and this increase was suppressed by catalase or TMPO. Each value is the mean + S.D. (n = 3). *P < 0.01 compared with control. **P < 0.05 compared with L-glutamate.
Oxidative stress and glutamate toxicity
205
Table 6. Cytosolic markers after glutamate treatment with or without ant±oxidative agents
Treatment Control L-Glutamate (10 mM) L-Glutamate + catalase (100 #g/ml) L-Glutamate + TMPO (3 mM) L-Glutamate + vitamin E (100/~M)
Glutamine synthetase ATP content activity GAPDH activity (pmol//~g protein) (pmol//~g protein per h) (mU/#g protein) 44.8 _ 26.8 _ 36.3 ± 35.8 ± 34.1 ±
6.1 8.2* 7.9 4.2 5.1
2.98 _ 0.87 2.37 __+0.56 3.08 __+0.48 2.35 __+0.52 2.65 ± 0.67
8.2 ± 12.1 _ 9.5 + 12.0 ± 10.5 ±
1.1 2.2** 1.3 2.0** 1.1"*
C6 glioma cells were cultured with glutamate alone or in combination with catalase, TMPO or vitamin E. Cellular ATP content and activities of glutamine synthetase and GAPDH were measured at 20 h after treatment using bioluminescence, high-performance liquid chromatographic and spectrophotometric methods, respectively (see Experimental Procedures). Glutamate induced a reduction of ATP content and a slight increase of GADPH activity, whereas glutamine synthetase activity was not changed by glutamate treatment. The reduction of cellular ATP content was prevented by radical scavengers and ant±oxidant. Each value is the mean ___S.D. (n = 34). *P < 0.01 and **P < 0.05 compared with control. generated inside of the cell. Sfiez et al. 39 reported in cervical ganglion neurons in culture that superoxide dismutase protects cultured neurons against death by substrate deprivation. This discrepancy with our data may be ascribed to different experimental conditions and different cell types. Cellular target molecules sensitive to oxygen radicals Membranes. Membranes consist mainly of lipids and proteins. Membrane lipid peroxidation has been proposed to be a consequence of oxidative stress. Unsaturated fatty acids in membrane lipids are particularly susceptible to oxygen radicals, because double bonds are thought to undergo peroxidation through oxidative chain reactions. 42 A two-fold increase of lipid peroxide accumulation in C6 cells was detected 20 h after glutamate treatment. The increase of lipid peroxides was completely protected by radical scavengers (Table 5) or the ant±oxidant vitamin E. 4° This fact suggests that membrane lipid is one of the main targets of oxygen radicals in glutamate toxicity. In brain ischemia, whose pathogenesis is also connected to excess release of glutamate, accumulation of lipid peroxides has been reportedJ '28 Cytoplasm. The susceptibility of cytoplasm to oxygen radicals was tested by measuring cellular levels of ATP and activities of the cytosolic enzymes, glutamine synthetase and GAPDH. ATP content was measured to estimate the function of mitochondria, as mitochondrial respiration is an important source of O{ and H202 generation. 5 Therefore, mitochondria might be expected to be target organelles for oxygen radicals. Indeed, GSH deficiency leads to mitochondrial damage in rat brain. 15 In our present study, cellular ATP (40.2%) was significantly reduced 20 h after glutamate treatment. This reduction of ATP was almost restored by oxygen radical scavengers or the ant±oxidant vitamin E. The activities of two cytosolic enzymes, glutamine synthetase and GAPDH, were selected to detect representative target molecules of oxygen radicals. The histidine residue of glutamine synthetase and the cysteine residue of G A P D H are well known to be inactivated by oxygen radicals, especially H2 02.35,38 A NSC 73/1--H
glial marker enzyme, glutamine synthetase has also been used as a biochemical indicator of glial cell damage induced by the glutamate analog ~-aminoad±pate in the fish retina. 2° The activity of glutamine synthetase in C6 glioma cells was not decreased 20 h after glutamate treatment (Table 6), suggesting that glutamine synthetase is not a target molecule of oxygen radicals. Our previous study showed that the glutamine synthetase activity of C6 glioma cells was much lower (1%) than that of retinal glial cells. TM Therefore, function of glutamine synthetase between glial and glioma cells may be quite different in different environments. Surprisingly, activity of G A P D H was not decreased but rather slightly increased after glutamate treatment. In C6 glioma cells, G A P D H may not be a target for oxygen radicals. If the reduction of cellular ATP after glutamate treatment as mentioned above indicates mitochondrial dysfunction, the G A P D H activation by glutamate treatment may reflect a compensatory production of ATP through anerobic glycolysis. Nuclei. It is well known that oxygen radicals directly attack the nuclei and cause single- and double-strand D N A breaks. 12 Furthermore, oxygen radicals induce an increase of intracellular Ca 2+, causing an activation of calcium-dependent endonuclease resulting in D N A fragmentations, a6 Recently, Ratan et al. 37 reported that oxidative stress also induced apoptosis in embryonic cortical neurons. In the present study, we demonstrated apoptotic-like D N A fragmentation with H33258 staining (Fig. 1). The apoptotic D N A fragmentation showed typical internucleosomal D N A 45 cleavage with de novo protein and RNA synthesis. Whether or not the glutamate cell death in our C6 glioma cells is an apoptotic cell death remains unresolvedJ 4 CONCLUSION In this study, we demonstrate a strong correlation between the ROS accumulation and glutamateinduced cell death of C6 glioma cells. Three types of ROS ( 0 2 , H202 and 'OH) are generated from the four-step reduction of molecular oxygen. These
206
K. Mawatari et al.
reactions are catalysed by transition metal ions such as iron, which provide electrons at each step. 33 Highly reactive 'OH is formed from H202 with reaction with
iron ions through the Fenton type Harber-Weiss reaction. ~3These reactions and sequent accumulation of cytotoxic ROS may be promoted in glutamate
Fig. 1. DNA specific staining with H33258 after glutamate treatment. (A) Cells (4.0 x 105) were plated and incubated in standard DMEM for 20 h. (B) When 10 mM L-glutamate was added to the medium at the beginning of the culture, condensation of chromatin masses and aggregation at the nuclear membrane progressed and almost all nuclei changed 20 h after treatment. (C) In the presence of catalase (100 #g/ml), TMPO (3 mM), PBN (3 mM) or vitamin E (100 p M) with glutamate, the morphology of nuclei was similar to that of controls. Scale bar = 10/~m.
207
Oxidative stress and glutamate toxicity
Catalase
.2~ ~
1
.~TMPO
I
R 0 S - . ~ . ...... G S H
2",H202"~ OH) •
Mitochondria
~
C6 CELL
A Cysteine
A Cystine
ROS : Reactive Oxygen Species Vit. E : Vitamin E
Cystine Fig. 2. A scheme for oxidative stress by glutamate toxicity of C6 glioma cells via the x~ antiporter system and for protection by radical scavengers and antioxidants. Mitochondrial respiration is an important source of 0 2 and H202. After GSH depletion, the excessively accumulating H202 is catalysed by iron ion to form highly reactive "OH by the Fenton type Harber-Weiss reaction. In the present study, catalase may scavenge excessive H202 outside the cell, while TMPO and PBN can directly trap O~- and "OH inside the cell. Vitamin E is effective as a radical chain-breaking antioxidant. Therefore, these antioxidative agents can protect C6 cells against oxidative stress by glutamate exposure.
toxicity of C6 glioma cells. In Fig. 2, we summarize ROS accumulation and multiple cellular injuries by ROS during glutamate cell death of C6 glioma cells via the x~ antiporter system. After GSH depletion, ROS accumulation was increased, leading through multiple cellular damage to cell death. Cell death and cell damage were completely prevented by oxygen radical scavengers and an antioxidant. Our in vitro C6 glioma cell system offers a very useful model to analyse both the cellular target molecules to ROS and
protection mechanisms for ROS-induced neurodegenerative disorders)
Acknowledgements--We thank Mr S. Ishita for technical
assistance and Mrs Tami Urano for secretarial assistance. This work was supported in part by research grants (Nos 06808084 and 07558105 to S.K.) from the Ministry of Education, Science and Culture of Japan, from the Sasakawa Health Science Foundation and from the Mochida Pharmaceutical Science Foundation, 1993.
REFERENCES
1. Bannai S. (1986) Exchange ofcystine and glutamate across plasma membrane of human flbroblasts. J. biol. Chem. 261, 2256-2263. 2. Bannai S. and Kitamura E. (1980) Transport interaction of L-cystine and L-glutamate in human diploid fibroblasts in culture. J. biol. Chem. 255, 2372-2376. 3. Behl C., Davis J. B., Lesley R. and Schubert D. (1994) Hydrogen peroxide mediates amyloid fl protein toxicity. Cell 77, 817-827. 4. Bromont C., Marie C. and Bralet J. (1989) Increased lipid peroxidation in vulnerable brain regions after transient cortex ischemia in rats. Stroke 20, 918-924. 5. Chance B., Sies H. and Boveris A. (1979) Hydroperoxide metabolism in mammalian organs. Physiol. Rev. 59, 527-605. 6. Cho Y. and Bannai S. (1990) Uptake of glutamate and cystine in C6 glioma cells and in cultured astrocytes. J. Neurochem. 53, 2091-2097. 7. Choi D. W. (1988) Glutamate neurotoxicity and disease of the nervous system. Neuron 1, 623-634. 8. Coyle J. T. and Puttfarcken P. S. (1993) Oxidative stress, glutamate and neurodegenerative disorders. Science 262, 689-694. 9. Dimonte D. A., Chan P. and Sandy M. S. (1992) Glutathione in Parkinson's disease: a link between oxidative stress and mitochondrial damage? Ann. Neurol. 32, S1 l l-S115. 10. Finkelstein E., Rosen G. M. and Rauckman E. J. (1980) Spin trapping of superoxide and hydroxyl radical: practical aspects. Archs Biochem. Biophys. 200, 1-16. 11. Giblin F. J., Reddan J. R., Schrimscher L., Dziedzic D. C. and Reddy V. N. (1990) The relative roles of the glutathione redox cycle and catalase in the detoxification ofH202 by cultured rabbit lens epithelial cells. ExplEye Res. 50, 795-804. 12. Halliwell B. (1992) Reactive oxygen species and the central nervous system. J. Neurochem. 59, 1609-1623.
208
K. Mawatari et al.
13. Halliwell B. and Gutteridge J. M. C. (1989) Free Radicals in Biology and Medicine, pp. 188-266. Oxford University Press, Oxford. 14. Higuchi Y., Kato S. and Matsukawa S. (1994) Involvement of protein kinase C in apoptosis induced by glutamate in C6 rat glioma cells. Proc. XVlth Int. Cancer Congress 857-861. 15. Jain A., Martensson J., Stole E., Auld P. A. M. and Meister A. (1991) Glutathione deficiency leads to mitochondrial damage in brain. Proc. natn. Acad. Sci. U.S.A. 88, 1913-1917. 16. Kato S., Higashida H., Higuchi Y., Hatakenaka S. and Negishi K. (1984) Sensitive and insensitive states of cultured glioma cells to glutamate damage. Brain Res. 303, 365-373. 17. Kato S., Ishita S., Sugawara K. and Mawatari K. (1993) Cystine/glutamate antiporter expression in retinal Miiller glial cells: implications for DL-Ct-aminoadipate toxicity. Neuroscience 57, 473~,82. 18. Kato S., Negishi K., Honma K., Sakai K. and Shimada Y. (1989) A high performance liquid chromatography assay for glutamine synthetase. Neurochem. Int. 14, 491496. 19. Kato S., Neglshi K., Mawatari K. and Kuo C. H. (1992) A mechanism for glutamate toxicity in the C6 glioma cells involving inhibition of cystine uptake leading to glutathione depletion. Neuroscience 48, 903-914. 20. Kato S., Sugawara K., Matsukawa T. and Negishi K. (1990) Gliotoxic effects of ct-aminoadipic acid isomers on the carp retina: a long-term observation. Neuroscience 36, 145-153. 21. Kontos H. A. (1989) Oxygen radicals in CNS damage. Chem. Biol. Interact. 72, 229-255. 22. Lafon-Cazal M., Pietri S., Culcasi M. and Bockaert J. (1993) NMDA-dependent superoxide production and neurotoxicity. Nature 364, 535-537. 23. Lebel C. P. and Bondy S. C. (1990) Sensitive and rapid quantitation of oxygen reactive species formation in rat synaptosomes. Neurochem. Int. 17, 435-440. 24. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-275. 25. Mawatari K., Sugitani K., Yasui Y. and Kato S. (1994) Cellular mechanism of glutamate damage in C6 glioma cells. Bull. Jap. Soc. Neurochem. 33, 252-253. 26. Meister A. (1988) Glutathione metabolism and its selective modification. J. biol. Chem. 263, 17205-17208. 27. Meldrum B. (1993) Amino acids as dietary excitotoxins: a contribution to understanding neurodegenerative disorders. Brain Res. Rev. 18, 293-314. 28. Michel H. S., Vaishnav Y. N., Kempski O., von Lubitz D., Weiss J. F. and Feuerstein G. (1987) Breathing 100% oxygen after global brain ischemia in Mongolian gerbils resulting in increasedlipid peroxidation and increased mortality. Stroke 18, 426-430. 29. Murphy T. H., Miyamoto M., Sastre A., Schnaar R. L. and Coyle J. T. (1989) Glutamate toxicity in a neuronal cell line involves inhibition of cystine transport leading to oxidative stress. Neuron 2, 1547-1558. 30. Murphy T. H., Schnaar R. L. and Coyle J. T. (1990) Immature cortical neurons are uniquely sensitive to glutamate toxicity by inhibition of cystine uptake. Fedn Am. Socs exp. BioL 4, 1624-1633. 31. Oberhammer F. A., Pavelka M., Sharma S., Tiefenbacher R., Purchio A. F., Bursh W. and Schulte-Hermann R. (1992) Induction of apoptosis in cultured hepatocytes and in regressing liver by transforming growth factor fl I. Proc. natn. Acad. Sci. U.S.A. 89, 5408-5412. 32. Oka A., Belliveau M. J., Rosenberg P. A. and Volpe J. J. (1993) Vulnerability of oligodendroglia to glutamate: pharmacology, mechanism and prevention. J. Neurosci. 13, 1441-1453. 33. Olanow C. W. (1992) An introduction to the free radical hypothesis in Parkinson's disease. Ann. Neurol. 32, Suppl., $2-$9. 34. Olanow C. W. (1994) A radical hypothesis for neurodegeneration. Trends Neurosci. 16, 439-444. 35. Oliver C. N., Starke-Reed P. E., Stadtman E. R., Liu G. J., Carney J. M. and Floid R. A. (1990) Oxidative damage to brain proteins, loss of glutamine synthetase activity, and production of free radicals during ischemia/reperfusion-induced injury to gerbil brain. Proc. natn. Acad. Sci. U.S.A. 87, 5144-5147. 36. Orrenius S., McConkey D. J., Bellomo G. and Nicotera P. (1989) Role of Ca 2+ in toxic cell killing. Trends pharmac. Sci. I0, 281-285. 37. Ratan R. R., Murphy T. H. and Baraban J. M. (1994) Oxidative stress induces apoptosis in embryonic cortical neurons. J. Neurochem. 62, 376-379. 38. Reddan J. R., Sevilla M. D., Giblin F. J., Padgaonkar V., Dziedzic D. C., Leverenz V., Misra I. C. and Peters J. L. (1993) The superoxide dismutase mimic TEMPOL protects cultured rabbit lens epithelial cells from hydrogen peroxide insult. Expl Eye Res. 56, 543-554. 39. S~ez J. C., Kessler J. A., Bennett M. V. L. and Spray D. C. (1987) Superoxide dismutase protects cultured neurons against death by starvation. Proc. natn. Acad. Sci. U.S.A. 84, 3056-3059. 40. Shinagawa S. (1994) Serotonin protects C6 glioma cells from glutamate toxicity. Neuroscience 59, 1043-1050. 41. Siesj6 B. K., Aghardh C. D. and Bengtsson F. (1989) Free radicals and brain damage. Cerebr. Brain Metab. Rev. 1, 165-211. 42. Slater T. F. (1984) Free radical mechanisms in tissue injury. Biochem. J. 222, 1-15. 43. Sofic E., Paulus W., Jellinger K., Riedever P. and Youdium M. B. (1991) Selective increase of iron in substantia nigra zona compacta of parkinsonian brain. J. Neurochem. 56, 978-982. 44. Stanley P. E. and Williams S. G. (1969) Use of the liquid scintillation spectrometer for determining adenosine triphosphate by the luciferase enzyme. Analyt. Biochem. 29, 381-392. 45. Wyllie A. H. (1980) Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555-556. 46. Yue T. L., Gu J. L., Lysko P. G., Cheng H. Y., Barone F. C. and Feuerstein G. (1992) Neuroprotective effects of phenyl-t-butyl-nitrone in gerbil global brain ischemia and in cultured rat cerebellar neurons. Brain Res. 574, 193-197. (Accepted 10 January 1996)