Oxidative stress is involved in the permeabilization of the inner membrane of brain mitochondria exposed to hypoxia/reoxygenation and low micromolar Ca 2+ Lorenz Schild 1 and Georg Reiser 2 1 Bereich Pathologische Biochemie des Instituts fu ¨ r Klinische Chemie und Pathologische Biochemie, Otto-von-Guericke-Universita ¨ t Magdeburg, Germany 2 Institut fu ¨ r Neurobiochemie, Medizinische Fakulta ¨ t, Otto-von-Guericke-Universita ¨ t Magdeburg, Germany Oxidative stress seems to be involved in the patho- genesis of neurodegenerative processes such as tissue infarction resulting from transient ischemia in stroke [1,2]. Investigations of in vivo ischemia showed that brains of transgenic mice which overexpressed copper ⁄ zinc superoxide dismutase or manganese superoxide Keywords brain mitochondria; hypoxia, calcium; membrane permeabilization; oxidative stress Correspondence L. Schild, Bereich Pathologische Biochemie, Institut fu ¨ r Klinische Chemie und Pathologische Biochemie, Medizinische Fakulta ¨ t, Otto-von-Guericke-Uneversita ¨ t Magdeburg, Leipziger Strasse 44, 39120 Magdeburg, Germany Fax: +49 391 67 190176 Tel: +49 391 67 13644 E-mail: Lorenz.Schild@Medizin. Uni-Magdeburg.de (Received 23 March 2005, revised 11 May 2005, accepted 19 May 2005) doi:10.1111/j.1742-4658.2005.04781.x From in vivo models of stroke it is known that ischemia ⁄ reperfusion indu- ces oxidative stress that is accompanied by deterioration of brain mito- chondria. Previously, we reported that the increase in Ca 2+ induces functional breakdown and morphological disintegration in brain mitochon- dria subjected to hypoxia⁄ reoxygenation (H ⁄ R). Protection by ADP indica- ted the involvement of the mitochondrial permeability transition pore in the mechanism of membrane permeabilization. Until now it has been unclear how reactive oxygen species (ROS) contribute to this process. We now report that brain mitochondria which had been subjected to H ⁄ Rin the presence of low micromolar Ca 2+ display low state 3 respiration (20% of control), loss of cytochrome c, and reduced glutathione levels (75% of control). During reoxygenation, significant mitochondrial generation of hydrogen peroxide (H 2 O 2 ) was detected. The addition of the membrane permeant superoxide anion scavenger TEMPOL (4-hydroxy-2,2,6,6-tetra- methylpiperidine-N-oxyl) suppressed the production of H 2 O 2 by brain mitochondria metabolizing glutamate plus malate by 80% under normoxic conditions. TEMPOL partially protected brain mitochondria exposed to H ⁄ R and low micromolar Ca 2+ from decrease in state 3 respiration (from 25% of control to 60% of control with TEMPOL) and permeabilization of the inner membrane. Membrane permeabilization was obvious, because state 3 respiration could be stimulated by extramitochondrial NADH. Our data suggest that ROS and Ca 2+ synergistically induce permeabilization of the inner membrane of brain mitochondria exposed to H ⁄ R. However, per- meabilization can only partially be prevented by suppressing mitochondrial generation of ROS. We conclude that transient deprivation of oxygen and glucose during temporary ischemia coupled with elevation in cytosolic Ca 2+ concentration triggers ROS generation and mitochondrial permeabili- zation, resulting in neural cell death. Abbreviations CSA, cyclosporin A; DCFH, dichlorofluorescin; GSH, reduced glutathione; GSSH, oxidised glutathione; H ⁄ R, hypoxia ⁄ reoxygenation; mPTP, mitochondrial permeability transition pore; RCR, respiratory control ratio; ROS, reactive oxygen species; TEMPOL, 4-hydroxy-2,2,6,6- tetramethylpiperidine-N-oxyl. FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS 3593 dismutase were protected against deleterious conse- quences of stroke [3,4]. Moreover, the increase in lipid peroxidation after ischemia also points towards the involvement of oxidative stress [5,6]. Further investigations using animal models of stroke made it clear that mitochondria are injured during cer- ebral ischemia and postischemic reperfusion. Decreased respiratory capacity and changes in ultrastructure of mitochondria have been reported [7,8]. Cortical neur- onal mitochondria after transient ischemia showed condensation, increased matrix density, and deposits of electron-dense material, and finally disintegration. It was further shown that mitochondria swell due to transient focal ischemia, whereas permanent ischemia causes loss of matrix density [9]. Another mitochond- rial response to cerebral ischemia is membrane permea- bilization resulting in the release of mitochondrial proteins, such as cytochrome c, caspase 9, and SMAC- Diablo [10,11]. It is conceivable that reactive oxygen species (ROS) generated outside the mitochondria can cause damage of these organelles finally resulting in their rupture. In fact, this pathway has been demonstrated in isolated mitochondria, using iron ⁄ ascorbate as an external ROS generating system [12]. Mitochondria themselves are generators of ROS which may also cause damage. Up to 2% of the oxygen consumed by mitochondria can be converted to superoxide anion radicals by the mitochondrial respiratory chain under reduced condi- tions, which have been shown to occur when complex III is blocked with antimycin A. Subsequently, these superoxide anion radicals are converted by the man- ganese-superoxide dismutase to H 2 O 2 . After diffusion into the cytosol H 2 O 2 mediates the damage of cellular components such as proteins, nucleic acids and lipids. The generation of ROS by the electron transport chain has been mainly attributed to complex III [13]. An additional source of superoxide anion radicals is com- plex I [14]. The degree of production of ROS by the respiratory chain depends on the type of tissue, the kind of substrates driving oxidative phosphorylation (either complex I-dependent or complex II-dependent substrates) and the polarization of the mitochondrial membrane [15,16]. Until now it is unclear how mito- chondrially generated ROS contribute to mitochond- rial damage by causing membrane permeabilization upon brain ischemia ⁄ reperfusion. A second factor determining mitochondrial damage upon ischemia ⁄ reperfusion is the increase in cytosolic Ca 2+ concentra- tion, which causes swelling of mitochondria, induction of ROS generation and functional failure [17,18]. Investigations on isolated mitochondria subjected to hypoxia ⁄ reoxygenation (H ⁄ R) allow the study of the effect of single factors such as elevated Ca 2+ and hyp- oxia on mitochondria [19,20]. In previous work we showed that in isolated brain mitochondria H ⁄ R in the presence of low micromolar Ca 2+ concentrations pro- vokes the permeabilization of the inner membrane [21]. In this study we report that isolated brain mitochon- dria exposed to H ⁄ R in the presence of low micro- molar Ca 2+ generate significant amounts of H 2 O 2 during reoxygenation. After this treatment reduced state 3 respiration with glutamate plus malate as sub- strate, increased cytochrome c release, and reduced glutathione (GSH) levels in comparison to normoxic controls were found. Detoxification of mitochondrially generated supoxide anion radicals by the membrane permeant superoxide anion scavenger TEMPOL (4- hydroxy-2,2,6,6-tetramethylpiperidine-N-oxyl) resulted in increased state 3 respiration and reduced membrane permeabilization. Membrane permeabilization was measured as stimulation of state 3 respiration by extra- mitochondrial NADH. Results We first investigated the effect of H ⁄ R and extramito- chondrial Ca 2+ on mitochondrial respiration. There- fore, isolated rat brain mitochondria were exposed either to 10 min hypoxia followed by 5 min reoxygena- tion or to 3.5 lm Ca 2+ , or to the combination of both. After each treatment, 5 mm glutamate, 5 mm malate, and 200 lm ADP were added to the incubation to induce state 3 respiration and oxygen consumption was analysed. The corresponding values are shown in Fig. 1. After 15 min of incubation in the presence of 3.5 lm Ca 2+ in an air atmosphere, state 3 respiration of brain mitochondria was decreased to 81.4 ± 1.2% (n ¼ 5) of normoxic control. Exposure to 10 min hypoxia and 5 min reoxygenation also resulted in a suppression of state 3 respiration (61.8 ± 2.8% of normoxic control; n ¼ 5). Most impressively, the com- bination of the two treatments caused tremendous effects on mitochondrial function measured as state 3 respiration (21.1 ± 3.3% of normoxic control; n ¼ 5). A high degree of protection was reached in the pres- ence of 5 mm ADP (84.9 ± 2.7% with ADP; n ¼ 5) but not of 5 mm AMP during the treatment. The damage to respiration of brain mitochondria exposed to H ⁄ R and ⁄ or Ca 2+ was always accompanied by the release of cytochrome c into the medium (Fig. 2). Under normoxic control conditions no cytochrome c was found in the incubation medium (lane 5). In this case cytochrome c is localized in the intermembrane space of mitochondria. H ⁄ R (lane 2), 3.5 lm Ca 2+ (lane 6), and the combination of both (lane 3) induced Mitochondrial permeabilization by transient hypoxia L. Schild and G. Reiser 3594 FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS cytochrome c release, indicating the permeabilization at least of the mitochondrial outer membrane. Neither Ca 2+ -induced cytochrome c release nor H ⁄ R and Ca 2+ -induced cytochrome c release was prevented by cyclosporin A (lane 7 and lane 4, respectively). To test the possibility that mitochondrially generated ROS could be involved in the mechanism of membrane permeabilization, we determined H 2 O 2 in the incuba- tion medium during H ⁄ R performed in the presence or absence of 3.5 lm Ca 2+ . The mitochondrial incuba- tions were carried out in the cuvette of the lumines- cence spectrophotometer at 30 °C in the presence of dichlorofluorescin (DCFH) and horseradish peroxi- dase. Before the experiment, the fluorescence signal was calibrated using standard H 2 O 2 solutions. To achieve hypoxic conditions, 1 mL of the incubation medium was bubbled with N 2 for 5 min. The cuvette was closed after mitochondria had been added. Reoxy- genation was performed by opening the cuvette and adding 1 mL of air saturated medium. In Fig. 3A, a typical protocol observed with five mitochondrial pre- parations is shown. In the hypoxic phase no relevant increase in H 2 O 2 concentration in the medium was detected. However, significant increase in H 2 O 2 con- centration was seen after reoxygenation in incubations with 3.5 lm Ca 2+ . Under these conditions, the rate of H 2 O 2 increase amounted to about 1.22 ± 0.35 pmolÆ s )1 Æmg )1 of mitochondrial protein. The decrease in the fluorescence signal recorded at the moment of reoxy- genation is caused by dilution due to the addition of 1 mL medium. To investigate whether the generation of ROS by brain mitochondria during H ⁄ R at low micromolar extramitochondrial Ca 2+ is associated with oxidative stress, we analysed changes in levels of GSH. The expo- sure of brain mitochondria to 3.5 lm Ca 2+ caused a decrease in GSH of 0.81 ± 0.02 nmolÆmg )1 of mito- chondrial protein (Fig. 3B). Hypoxia ⁄ reoxygenation had a smaller effect on GSH concentration. This treat- ment resulted in a decrease of 0.29 ± 0.19 nmolÆmg )1 of mitochondrial protein. The combination of 3.5 lm Ca 2+ and H ⁄ R led to the substantial decrease in GSH of 1.27 ± 0.15 nmolÆmg )1 of mitochondrial protein. This correlates well with increased amounts of H 2 O 2 ,as shown in Fig. 3A. Freshly isolated rat brain mitochon- dria contained 5.73 ± 0.23 nmolÆmg )1 of mitochondrial protein GSH. The reduction in the GSH concentration was partially reflected by increased amounts of oxidized glutathione (GSSG) (data not shown). To directly test whether mitochondrially derived ROS are involved in damaging the organelles under H ⁄ R and Ca 2+ , we used TEMPOL which permeates biological membranes and scavenges superoxide anions. First, we Fig. 2. Induction of cytochrome c release from rat brain mitochon- dria by H ⁄ R and Ca 2+ . Brain mitochondria ( 0.5 mg proteinÆmL )1 ) were incubated at 30 °C under the conditions indicated: –Ca 2+ , 15 min in incubation medium; Ca 2+ , 15 min in incubation medium plus 3.5 l M Ca 2+ ;H⁄ R, 10 min hypoxia followed by 5 min reoxy- genation; CSA, 2 l M cyclosporin A. A volume of 2 mL of the incu- bation mixture was centrifuged at 12 000 g for 10 min and the resulting supernatant at 100 000 g for 15 min at 4 °C. Cyto- chrome c was detected in the supernatant by western blot analy- sis. As positive controls 6 ng and 15 ng cytochrome c were applied to the gel (lane 1 and lane 8, respectively). The western blot pre- sented is typical for four preparations of mitochondria. Fig. 1. Effect of H ⁄ R and low micromolar Ca 2+ on state 3 respir- ation of brain mitochondria. Rat brain mitochondria (0.5 mg pro- teinÆmL )1 ) were incubated at 30 °C. The substrates (5 mM glutamate plus 5 mM malate) and 200 lM ADP were added before oxygen consumption (state 3 respiration) of mitochondria was measured. Control, 15 min incubation of mitochondria in air-satur- ated medium; Ca 2+ , 3.5 lM Ca 2+ ;H⁄ R, 10 min hypoxia followed by 5 min reoxygenation; ADP, 5 m M ADP; AMP, 5 mM AMP. The res- piration of untreated rat brain mitochondria (100%) corresponds to 71 nmol O 2 min )1 Æmg )1 from protein. Data represent mean val- ues ± SEM from five preparations of mitochondria. *Significant at P < 0.05. L. Schild and G. Reiser Mitochondrial permeabilization by transient hypoxia FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS 3595 evaluated the degree by which TEMPOL reduces extra- mitochondrial H 2 O 2 accumulation caused by brain mitochondria metabolizing glutamate plus malate under normoxic conditions (incubation in air saturated med- ium). Therefore, the fluorescence of dichlorofluorescein formed by H 2 O 2 -dependent oxidation of DCFH in the presence of horseradish peroxidase was measured. For quantitative analysis, the fluorescence signal was calib- rated by using H 2 O 2 standard solutions which were added to the incubation medium. The corresponding calibration is shown in Fig. 4A. A typical result simi- larly obtained in five mitochondrial preparations dem- onstrating the effect of the addition of substrates on H 2 O 2 production is shown in Fig. 4B. Considerable amounts of H 2 O 2 were released into the incubation medium, when 5 mm glutamate and 5 mm malate were added. The rates of H 2 O 2 accumulation presented as mean values ± SEM were: before substrate addition, 7.0 ± 1.5 pmol min )1 Æmg )1 of mitochondrial protein (n ¼ 5); after the addition of 5 mm glutamate and 5mm malate, 88.2 ± 5.7 pmol min )1 Æmg )1 of mitoch- ondrial protein (n ¼ 5); and after subsequent addition of 10 mm TEMPOL, 21.8 ± 2.1 nmol min )1 Æmg )1 of mitochondrial protein (n ¼ 5). The reduction of the increase in fluorescence intensity by TEMPOL by about 75% demonstrates the power of TEMPOL to effectively scavenge superoxide anion radicals within the mito- chondria. Fig. 4. Effect of substrate supply and TEMPOL on H 2 O 2 generation by brain mitochondria. (A) Calibration curve of H 2 O 2 measurements. At the points indicated 88 pmol H 2 O 2 standard solution were added to 2 mL incubation medium at 30 °C and the fluorescence intensity (excitation at 488 nm, emission at 525 nm) was monitored. (B) Rat brain mitochondria (0.5 mgÆmL )1 ) were incubated at 30 °C in incu- bation medium. At the times indicated 5 m M glutamate plus 5 mM malate or 10 mM TEMPOL were added. The rates of H 2 O 2 produc- tion are given in the text. The data are presented as mean ± SEM from five mitochondrial preparations. A B Fig. 3. Effect of H ⁄ R and low micromolar Ca 2+ on mitochondrial H 2 O 2 production and GSH. (A) H 2 O 2 accumulation was followed as described. When Ca 2+ was present, as indicated, the concentration was 3.5 l M. The traces shown are typical for four preparations of brain mitochondria. (B) GSH concentrations were determined as described. Incubations were: Ca 2+ , 15 min in incubation medium plus 3.5 l M Ca 2+ ;H⁄ R+Ca 2+ , 10 min hypoxia followed by 5 min reoxygenation in the presence of 3.5 l M Ca 2+ . The data are presen- ted as mean values ± SEM. *Significant at P < 0.05. Mitochondrial permeabilization by transient hypoxia L. Schild and G. Reiser 3596 FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS In the next series of experiments, we applied 10 mm TEMPOL to brain mitochondria exposed to 10 min hypoxia followed by 5 min reoxygenation (H ⁄ R) in the presence of 3.5 lm Ca 2+ and analysed the respiration. The data in Fig. 5 are presented as mean values ± SEM from five mitochondrial preparations. Under control conditions, that is, incubation of mitochondria in Ca 2+ - free and air saturated medium in the presence of 5 mm glutamate and 5 mm malate, oxidative phosphorylation was well coupled [state 4 respiration, 14.0 ± 1.7 nmol O 2 Æmin )1 Æmg )1 ; state 3 respiration, 77 ± 7.1 nmol O 2 Æmin )1 Æmg )1 ; respiratory control ratio (RCR), 5.49]. Exposure to H ⁄ R in the presence of 3.5 lm Ca 2+ resul- ted in dramatically decreased state 3 respiration and decreased RCR value (state 4 respiration: 11.8 ± 1.8 nmol O 2 Æmin )1 Æmg )1 ; state 3 respiration, 20.6 ± 1.6 nmol O 2 Æmin )1 Æmg )1 ; RCR, 1.74). The pres- ence of 10 mm TEMPOL during H ⁄ R partially preven- ted the decrease in state 3 respiration (state 3 respiration, 40.4 ± 7.4 nmol O 2 Æmin )1 Æmg )1 ; state 4 respiration, 14.0 ± 2.8, nmol O 2 Æmin )1 Æmg )1 ; RCR, 2.88). The most important experiment which helped us to understand the mechanism of damage was to study the sensitivity of state 3 respiration to extramitoch- ondrial NADH. Thus, we were able to detect permea- bilization of the inner mitochondrial membrane. Under control conditions and in the presence of 10 mm TEMPOL, extramitochondrial NADH had no influence on state 3 respiration (Fig. 5) which is due to the tight membrane. In contrast, when NADH was added to mitochondria which had been subjected to 10 min hypoxia followed by 5 min reoxygenation and 3.5 lm Ca 2+ , a substantial stimulation of respiration was observed (from 20.5 ± 1.6 to 77.0 ± 5.9 nmol O 2 Æmin )1 Æmg )1 ), indicating permeabilization of the membrane system. Even in the additional presence of 10 mm TEMPOL, significant permeabilization of the mitochondrial membrane occurred, which was shown by the NADH sensitivity of state 3 respiration (40.0 ± 7.4 vs. 84.5 nmol O 2 Æmin )1 Æmg )1 after NADH addition). Discussion In animal models of stroke, mitochondria are injured upon ischemia ⁄ reperfusion. This injury is characterized by swelling [22] and membrane perturbation which results in release of cytochrome c [23]. Attempts to prevent mitochondrial membrane permeabilization in in vivo models of stroke were only partially successful. Applying the immunosuppressive compound cyclospo- rin A which is known to prevent opening of the mito- chondrial permeability transition pore (mPTP) resulted in an incomplete protection from neurodegeneration. Cyclosporin A diminished the size of the infarct, but was not able to prevent general necrotic cell death [24]. These findings demonstrate, however, the involvement of mitochondrial membrane permeabilization in the damage of mitochondria upon ischemia ⁄ reperfusion. Only in vitro studies on mitochondria allow investiga- tion of the impact of single factors and their interplay, such as Ca 2+ and ROS. Therefore, we exposed isola- ted rat brain mitochondria to H ⁄ R and ⁄ or Ca 2+ and determined H 2 O 2 concentration, membrane permeabil- ity, and, as a parameter of mitochondrial function, oxygen consumption. Indirect evidence for the suggestion that oxidative stress may also contribute to permeabilization of the mitochondrial membrane and subsequently to the impairment of mitochondria upon ischemia⁄ reper- Fig. 5. Influence of Ca 2+ increase in combination with H ⁄ Rand extramitochondrial NADH on respiration of brain mitochondria. Rat brain mitochondria (0.5 mgÆmL )1 ) were incubated at 30 °C in incu- bation medium. The substrates (5 m M glutamate plus 5 mM malate) were added before oxygen consumption of mitochondria was measured. Control measurements and H ⁄ R were performed as described in the Experimental procedures. State 4 respiration was measured after the addition of substrates. State 3 respiration was induced by the addition of 200 l M ADP. The state 3 respiration of untreated rat brain mitochondria (100%) corresponds to 71 nmol O 2 Æmin )1 Æmg protein )1 . To estimate membrane permeability, 5 mM NADH was added to the incubations after the state 3 respiration had been analysed. Designations: H ⁄ R+Ca 2+ ± 10 min hypoxia and 5 min reoxygenation in the presence of 3.5 l M Ca 2+ ; TEMPOL, 10 m M TEMPOL. Data represent mean values ± SEM from five preparations of mitochondria. §, Difference between state 4 and state 3 respiration is significant at P < 0.05; *Difference between state 3 respiration with and without NADH significant at P < 0.05. L. Schild and G. Reiser Mitochondrial permeabilization by transient hypoxia FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS 3597 fusion comes from in vivo investigations using gene knock-out animals, which were deficient in the mito- chondrial antioxidant enzyme manganese superoxide dismutase. In this model, increased cytochrome c release was found after ischemia ⁄ reperfusion in com- parison to wild type animals [25,26]. Remarkable protection from brain injury was achieved by administration of metalloporphyrin catalytic antioxi- dants [27]. However, investigations of the therapeutic efficacy of antioxidant compounds both in animal models and humans [28–31] generated contradictory results. Consequently, the initial enthusiasm for the use of antioxidants to treat acute brain injury subsi- ded. As a reason for the failure, the bioavailability of antioxidants was discussed. However, the protective effect varied depending on the type of brain ischemia (focal or global) and the animal species (rat or mouse). The in vivo application of the Mito Tracker Red CMH(2)Xros, a rosamine derivative used for the detec- tion of mitochondrial free radicals, identified mito- chondria as generators of free radicals primarily in vulnerable neurons following focal cerebral ischemia [32]. Another piece of evidence for mitochondrial ROS production during ischemia ⁄ reperfusion comes from in vivo models of stroke using hydroethidine oxidation by superoxide anion radicals [33]. Thus, it may be con- cluded that mitochondria contribute to the induction of oxidative stress during ischemia ⁄ reperfusion in brain. There is a growing body of information concerning the mechanism of ROS generation by the respiratory chain in mitochondria. Complex I and complex III were identified as generators of superoxide anion radi- cals in brain mitochondria [34–38]. Depending on the animal species and incubation conditions, different fractions for the generation of ROS were attributed to the two respiratory chain complexes. When succinate was used as substrate, almost all superoxide anion radicals are produced in complex I of the respirarory chain by reversed electron transfer [38]. Although it seems that the substrate pair glutamate and malate induces the production of relatively small amounts of ROS, in our experiments we determined H 2 O 2 genera- tion by brain mitochondria in the presence of these electron donors. This is of relevance in brain, because in this tissue glucose is metabolized providing the NADH-yielding substrate pyruvate. Under these con- ditions, some succinate is formed by the citric acid cycle, oxidizing malate, even in the case of the NADH-depending substrate supplementation. In our experiments, we subjected isolated brain mitochondria to H ⁄ R in the presence of low micro- molar Ca 2+ . This experimental design mimics the situation during ischemia ⁄ reperfusion in which mitochondria have to endure transient interruption of oxygen supply and increased cytosolic Ca 2+ concentra- tions. Permeabilization of at least the outer mitoch- ondrial membrane, indicated by cytochrome c release, and dramatic functional injury seen as decrease of ADP-stimulated respiration, are consequences of this treatment. The high degree of protection by ADP sug- gests that the mitochondrial injury is caused by open- ing of the mPTP. There is evidence that ADP inhibits opening of the mPTP by occupying binding sites located in the inner and outer mitochondrial mem- brane [39,40]. The binding of ADP stabilizes the matrix conformation of the adenine nucleotide translo- cator which is known to prevent pore opening [36]. The depolarization of the mitochondrial membrane which occurs within the hypoxic phase and probably also during reoxygenation supports opening of the mPTP. It is a particular property of brain mitochon- dria that opening of the mPTP can happen even in the presence of CSA [41]. Two different mechanisms seem to be responsible for the release of cytochrome c induced either by Ca 2+ under normoxic conditions (air saturated medium) or by H ⁄ R and Ca 2+ (Fig. 2). In the first case, only the outer membrane was pemea- bilized, as reported earlier [42]. This process was not sensitive to CSA. In contrast, H ⁄ R in combination with low micromolar Ca 2+ caused CSA insensitive permeabilization of the inner mitochondrial mem- brane, indicated by the stimulatory effect of NADH on state 3 respiration (Fig. 5). In this situation it is likely that mitochondria swell, which then results in the rupture of the outer membrane accompanied by the release of cytochrome c. It had already been shown that especially in brain, CSA-insensitive permeabiliza- tion of the mitochondrial membrane may occur [21,41,43]. We report here that brain mitochondria subjected to H ⁄ R in the presence of low micromolar Ca 2+ generate H 2 O 2 during reoxygenation that is released into the extramitochondrial space. This is in line with in vivo studies of ischemia ⁄ reperfusion showing increased ROS production by mitochondria [32,33,44]. The decreased levels of GSH found after exposure to H ⁄ R in the presence of low micromolar Ca 2+ points towards the induction of oxidative stress. Although decrease in GSH may be caused by H 2 O 2 -dependent oxidation and ⁄ or by decrease in the reduction of GSSG due to permeabilization of the mitochondrial inner membrane, the reduction of GSH concentration indicates decrease in antioxidative capacity. Conse- quently, the concentration of free ROS may enhance oxidative stress. Mitochondrial permeabilization by transient hypoxia L. Schild and G. Reiser 3598 FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS The increases in cytosolic Ca 2+ and ROS concentra- tion are two essential factors that favour opening of the mPTP [45]. Thereby oxidation of SH-groups of the adenine nucleotide translocator stimulates pore open- ing. We hypothesize that elimination of ROS may reduce the permeabilization of the mitochondrial membrane and subsequently exert beneficial effects on mitochondria. In fact, we were able to demonstrate that the reduction of H 2 O 2 concentration in the pres- ence of TEMPOL significantly protected mitochondria from permeabilization of the inner membrane upon H ⁄ R in the presence of low micromolar Ca 2+ . From the experiments presented here we conclude that com- plete protection of mitochondria requires additional prevention of increase in extramitochondrial Ca 2+ concentration into the low micromolar range during H ⁄ R. Experimental procedures All chemicals were of analytical grade. 2¢7¢-dichlorofluore- scin-diacetate and TEMPOL were from Sigma (St. Louis, MO, USA). Horseradish peroxidase was from Boehringer (Mannheim, Germany). Mn-SOD, catalase and glutathione reductase were from Sigma (Deisenhofen, Germany). Preparation and incubation of brain mitochondria This work was conducted in accordance with the regula- tions of the National Act, the use of Experimental Animals (Germany). Mitochondria were prepared from the brains of 220–240 g male Wistar rats in ice-cold medium containing 250 mm mannitol, 20 mm Tris, 1 mm EGTA, 1 mm EDTA, and 0.3% (w ⁄ v) BSA at pH 7.4 (isolation medium) by using a modified standard procedure [21,46]. The mitochon- dria were well coupled, as indicated by a respiratory control ratio > 4 with glutamate plus malate as substrates. Protein content was measured according to the Bradford method [47] using BSA as the standard. In separate experiments we determined protein values with the Bradford and Lowry methods. From the comparison, a correction factor of 1.4 was estimated. This was used to correct the protein values of the Bradford method. Mitochondria (0.5–1.0 mg proteinÆmL )1 ) were incubated in a medium containing 10 mm KH 2 PO 4 , 0.5 mm EGTA, 60 mm KCl, 60 mm Tris, 110 mm mannitol, 1 mm free Mg 2+ at pH 7.4 and 30 °C. Extramitochondrial calcium was adjusted by using Ca 2+ ⁄ EGTA buffers. For calculating the concentration of free calcium, we used the complexing constants according to Fabiato et al. [48]. Hypoxia was produced by bubbling 2 mL of the incuba- tion medium with N 2 until oxygen was not detectable any more by means of a Clark electrode, as described previ- ously [21]. Afterwards, the mitochondria added to the med- ium further decreased the oxygen concentration via the respiratory chain until depolarization of the mitochondrial membrane was reached (not shown). A 2 mL volume of air-saturated incubation medium was added to achieve reoxygenation. Measurement of respiration Oxygen uptake of the mitochondria was measured at 30 °C in a thermostat-controlled chamber equipped with a Clark- type electrode. For the calibration of the oxygen electrode, the oxygen content of the air-saturated incubation medium was taken to be 217 nmolÆmL )1 [49]. Immunoblotting of cytochrome c A volume of 2 mL of the incubation mixture was centrifuged at 12 000 g for 10 min at 4 °C, and the resulting superna- tants were centrifuged at 100 000 g for 15 min at 4 °C. The supernatants were used for western blot analysis [50]. The mouse anti-(cytochrome c) Ig (PharMingen, Heidelberg, Germany) was applied in a dilution of 1 : 6000 and the sec- ondary sheep antimouse horseradish conjugated antibody (Chemicon, Chandlers Ford, UK) in a dilution of 1 : 3000. Detection of H 2 O 2 production Extramitochondrial H 2 O 2 peroxide produced by brain mitochondria was estimated by measuring the fluorescence (excitation at 488 nm, emission at 525 nm) caused by the H 2 O 2 -dependent oxidation of DCFH to the fluorescent compound dichlorofluorescein in the presence of horserad- ish peroxidase [51]. Prior to the experiments, DCFH was obtained from dichlorofluorescin-diacetate by treatment with alkali. Fluorescence signals were calibrated by measur- ing the fluorescence changes upon addition of known amounts of H 2 O 2 . Determination of glutathione The measurement of GSH and GSSG was based on the reaction with 5,5¢-dithio-bis-2-nitrobenzoic acid using a microplate assay according to Baker et al. [52]. Statistical analysis Data are presented as mean values ± SEM. The signifi- cance of differences was checked by using Student’s t-test of paired values. Acknowledgements The expert technical assistance of Mrs R. Widmayer is greatfully acknowledged. The work was supported by L. Schild and G. Reiser Mitochondrial permeabilization by transient hypoxia FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS 3599 grants from Bundesministerium fu ¨ r Bildung und Fors- chung (01ZZ0107). References 1 Beal MF (1998) Mitochondrial dysfunction in neuro- degenerative diseases. Biochim Biophys Acta 1366, 211– 223. 2 Halliwell B (1992) Reactive oxygen species and the cen- tral nervous system. J Neurochem 59 , 1609–1623. 3 Morita-Fujimura Y, Fujimura M, Kawase M, Chen SF & Chan PH (1999) Release of mitochondrial cyto- chrome c and DNA fragmentation after cold injury- induced brain trauma in mice: possible role in neuronal apoptosis. Neurosci Lett 267, 201–205. 4 Noshita N, Sugawara T, Fujimura M, Morita-Fujimura Y & Chan PH (2001) Manganese superoxide dismutase affects cytochrome c release and caspase-9 activation after transient focal cerebral ischemia in mice. J Cereb Blood Flow Metab 21, 557–567. 5 Hall ED (1995) Inhibition of lipid peroxidation in cen- tral nervous system trauma and ischemia. J Neurol Sci 134 (Suppl.), 79–83. 6 YuZF, Bruce-Keller AJ, Goodman Y & Mattson MP (1998) Uric acid protects neurons against excitotoxic and metabolic insults in cell culture, and against focal ischemic brain injury in vivo. J Neurosci Res 53, 613–625. 7 Anderson MF & Sims NR (1999) Mitochondrial respiratory function and cell death in focal cerebral ischemia. J Neurochem 73, 1189–1199. 8 Sims NR & Anderson MF (2002) Mitochondrial contri- butions to tissue damage in stroke. Neurochem Int 40, 511–526. 9 Solenski NJ, diPierro CG, Trimmer PA, Kwan AL & Helms GA (2002) Ultrastructural changes of neuronal mitochondria after transient and permanent cerebral ischemia. Stroke 33, 816–824. 10 Namura S, Nagata I, Takami S, Masayasu H & Kiku- chi H (2001) Ebselen reduces cytochrome c release from mitochondria and subsequent DNA fragmentation after transient focal cerebral ischemia in mice. Stroke 32, 1906–1911. 11 Sugawara T, Noshita N, Lewen A, Gasche Y, Ferrand- Drake M, Fujimura M, Morita-Fujimura Y & Chan PH (2002) Overexpression of copper ⁄ zinc superoxide dismu- tase in transgenic rats protects vulnerable neurons against ischemic damage by blocking the mitochondrial pathway of caspase activation. J Neurosci 22, 209–217. 12 Wiswedel I, Hirsch D, Nourooz-Zadeh J, Flechsig A, Luck-Lambrecht A & Augustin W (2002) Analysis of monohydroxyeicosatetraenoic acids and F2-isoprostanes as markers of lipid peroxidation in rat brain mitochon- dria. Free Radic Res 36, 1–11. 13 Cadenas E, Boveris A, Ragan CI & Stoppani AO (1977) Production of superoxide radicals and hydrogen peroxide by NADH-ubiquinone reductase and ubiqui- nol-cytochrome c reductase from beef-heart mitochon- dria. Arch Biochem Biophys 180, 248–257. 14 Turrens JF & Boveris A (1980) Generation of super- oxide anion by the NADH dehydrogenase of bovine heart mitochondria. Biochem J 191, 421–427. 15 Korshunov SS, Skulachev VP & Starkov AA (1997) High protonic potential actuates a mechanism of pro- duction of reactive oxygen species in mitochondria. FEBS Lett 416, 15–18. 16 Loschen G, Flohe L & Chance B (1971) Respiratory chain linked H 2 O 2 production in pigeon heart mito- chondria. FEBS Lett 18, 261–264. 17 Meldrum BS (1986) Cell damage in epilepsy and the role of calcium in cytotoxicity. Adv Neurol 44, 849–855. 18 Starkov AA, Chinopoulos C & Fiskum G (2004) Mito- chondrial calcium and oxidative stress as mediators of ischemic brain injury. Cell Calcium 36, 257–264. 19 Du G, Mouithys-Mickalad A & Sluse FE (1998) Gen- eration of superoxide anion by mitochondria and impairment of their functions during anoxia and reoxy- genation in vitro. Free Radic Biol Med 25, 1066–1074. 20 Schild L, Reinheckel T, Wiswedel I & Augustin W (1997) Short-term impairment of energy production in isolated rat liver mitochondria by hypoxia ⁄ reoxygena- tion: involvement of oxidative protein modification. Biochem J 328, 205–210. 21 Schild L, Huppelsberg J, Kahlert S, Keilhoff G & Reiser G (2003) Brain mitochondria are primed by moderate Ca 2+ rise upon hypoxia ⁄ reoxygenation for functional breakdown and morphological desintegra- tion. J Biol Chem 278, 25454–25460. 22 Rodrigues CM, Spellman SR, Sola S, Grande AW, Linehan-Stieers C, Low WC & Steer CJ (2002) Neuro- protection by a bile acid in an acute stroke model in the rat. J Cereb Blood Flow Metab 22, 463–471. 23 Fujimura M, Morita-Fujimura Y, Kawase M, Copin JC, Calagui B, Epstein CJ & Chan PH (1999) Manga- nese superoxide dismutase mediates the early release of mitochondrial cytochrome C and subsequent DNA frag- mentation after permanent focal cerebral ischemia in mice. J Neurosci 19, 3414–3422. 24 Matsumoto S, Friberg H, Ferrand-Drake M & Wieloch T (1999) Blockade of the mitochondrial permeability transition pore diminishes infarct size in the rat after transient middle cerebral artery occlusion. J Cereb Blood Flow Metab 19, 736–741. 25 Kim GW, Kondo T, Noshita N & Chan PH (2002) Manganese superoxide dismutase deficiency exacerbates cerebral infarction after focal cerebral ischemia ⁄ reper- fusion in mice: implications for the production and role of superoxide radicals. Stroke 33, 809–815. 26 Murakami K, Kondo T, Kawase M, Li Y, Sato S, Chen SF & Chan PH (1998) Mitochondrial susceptibility to oxidative stress exacerbates cerebral infarction that fol- Mitochondrial permeabilization by transient hypoxia L. Schild and G. Reiser 3600 FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS lows permanent focal cerebral ischemia in mutant mice with manganese superoxide dismutase deficiency. J Neurosci 18, 205–213. 27 Mackensen GB, Patel M, Sheng H, Calvi CL, Batinic- Haberle I, Day BJ, Liang LP, Fridovich I, Crapo JD, Pearlstein RD & Warner DS (2001) Neuroprotection from delayed postischemic administration of a metallo- porphyrin catalytic antioxidant. J Neurosci 21, 4582– 4592. 28 Forsman M, Fleischer JE, Milde JH, Steen PA & Michenfelder JD (1988) Superoxide dismutase and cata- lase failed to improve neurologic outcome after com- plete cerebral ischemia in the dog. Acta Anaesthesiol Scand 32, 152–155. 29 Matsumiya N, Koehler RC, Kirsch JR & Traystman RJ (1991) Conjugated superoxide dismutase reduces extent of caudate injury after transient focal ischemia in cats. Stroke 22, 1193–1200. 30 Muizelaar JP, Marmarou A, Young HF, Choi SC, Wolf A, Schneider RL & Kontos HA (1993) Improving the outcome of severe head injury with the oxygen radical scavenger polyethylene glycol-conjugated superoxide dis- mutase: a phase II trial. J Neurosurg 78, 375–382. 31 Uyama O, Matsuyama T, Michishita H, Nakamura H & Sugita M (1992) Protective effects of human recombi- nant superoxide dismutase on transient ischemic injury of CA1 neurons in gerbils. Stroke 23, 75–81. 32 Kim DY, Won SJ & Gwag BJ (2002) Analysis of mito- chondrial free radical generation in animal models of neuronal disease. Free Radic Biol Med 33, 715–723. 33 Cino M & Del Maestro RF (1989) Generation of hydrogen peroxide by brain mitochondria: the effect of reoxygenation following postdecapitative ischemia. Arch Biochem Biophys 269, 623–638. 34 Barja G & Herrero A (1998) Localization at complex I and mechanism of the higher free radical production of brain nonsynaptic mitochondria in the short-lived rat than in the longevous pigeon. J Bioenerg Biomembr 30, 235–243. 35 Kwong LK & Sohal RS (1998) Substrate and site speci- ficity of hydrogen peroxide generation in mouse mito- chondria. Arch Biochem Biophys 350, 118–126. 36 Novgorodov SA, Gudz TI, Milgrom YM & Brierley GP (1992) The permeability transition in heart mitochondria is regulated synergistically by ADP and cyclosporin A. J Biol Chem 267 , 16274–16282. 37 Starkov AA & Fiskum G (2001) Myxothiazol induces H 2 O 2 production from mitochondrial respiratory chain. Biochem Biophys Res Commun 281, 645–650. 38 Votyakova TV & Reynolds IJ (2001) DY m -Dependent and -independent production of reactive oxygen species by rat brain mitochondria. J Neurochem 79, 266–277. 39 Haworth RA & Hunter DR (1980) Allosteric inhibition of the Ca2+-activated hydrophilic channel of the mitochondrial inner membrane by nucleotides. J Membr Biol 54, 231–236. 40 Zoratti M & Szabo I (1995) The mitochondrial perme- ability transition. Biochim Biophys Acta 1241, 139–176. 41 Chinopoulos C, Starkov AA & Fiskum G (2003) Cyclosporin A-insensitive permeability transition in brain mitochondria: inhibition by 2-aminoethoxydiphe- nyl borate. J Biol Chem 278, 27382–27389. 42 Schild L, Keilhoff G, Augustin W, Reiser G & Striggow F (2001) Distinct Ca 2+ thresholds determine cyto- chrome c release or permeability transition pore opening in brain mitochondria. FASEB J 15, 565–567. 43 Brustovetsky N & Dubinsky JM (2000) Limitations of cyclosporin A inhibition of the permeability transition in CNS mitochondria. J Neurosci 20, 8229–8237. 44 Patole MS, Swaroop A & Ramasarma T (1986) Genera- tion of H 2 O 2 in brain mitochondria. J Neurochem 47 , 1–8. 45 Crompton M & Costi A (1988) Kinetic evidence for a heart mitochondrial pore activated by Ca 2+ , inorganic phosphate and oxidative stress. A potential mechanism for mitochondrial dysfunction during cellular Ca 2+ overload. Eur J Biochem 178, 489–501. 46 Schild L, Westphal S, Scheithauer S, Holfeld M, Augus- tin W & Sabel BA (1996) Isolation of functionally intact pig retina mitochondria. Acta Ophthalmol Scand 74, 354–357. 47 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein util- izing the principle of protein-dye binding. Anal Biochem 72, 248–254. 48 Fabiato A & Fabiato F (1979) Calculator programs for computing the composition of the solutions containing multiple metals and ligands used for experiments in skinned muscle cells. J Physiol (Paris) 75, 463–505. 49 Reynafarje B, Costa LE & Lehninger AL (1985) O 2 solubility in aqueous media determined by a kinetic method. Anal Biochem 145, 406–418. 50 Ghafourifar P, Schenk U, Klein SD & Richter C (1999) Mitochondrial nitric-oxide synthase stimulation causes cytochrome c release from isolated mitochondria. Evi- dence for intramitochondrial peroxynitrite formation. J Biol Chem 274 , 31185–31188. 51 Black MJ & Brandt RB (1974) Spectrofluorometric ana- lysis of hydrogen peroxide. Anal Biochem 58, 246–254. 52 Baker MA, Cerniglia GJ & Zaman A (1990) Microtiter plate assay for the measurement of glutathione and glutathione disulfide in large numbers of biological samples. Anal Biochem 190, 360–365. L. Schild and G. Reiser Mitochondrial permeabilization by transient hypoxia FEBS Journal 272 (2005) 3593–3601 ª 2005 FEBS 3601 . Oxidative stress is involved in the permeabilization of the inner membrane of brain mitochondria exposed to hypoxia/reoxygenation and low micromolar. that mitochondria contribute to the induction of oxidative stress during ischemia ⁄ reperfusion in brain. There is a growing body of information concerning the