Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Ischemic neuronal death including development of an infarct has been recently ascribed in part to the programmed cell death (Linnik et al., 1995; Chopp and Li, 1996). Transient focal ischemia initiates a cascade of detrimental events including accumulation of intracellular calcium, formation of free radicals and cytokines (tumor necrosis factor- and interleukin-1), which lead to disruption of cellular homeostasis and structural damage of ischemic brain tissue (Kochanek and Hallenbeck, 1992; Feuerstein et al., 1994). Ischemia results in the activation of cysteine proteases of the caspase family, alterations in plasma membrane phospholipids, and nuclear DNA condensation and fragmentation (Bredesen, 1995). During apoptosis, free radicals are known to induce lipid peroxidation, DNA damage (Dirnagl et al., 1999), and open the mitochondrial membrane permeability transition pore opens, mediating release of cytochrome c from mitochondria that activates caspases, finally producing apoptosis (Chen et al., 1997; Kluck et al., 1997). Therefore, treatment with antioxidants is effectively useful in suppressing neuronal damage (Huh et al., 2000).

During apoptosis, Bcl-2 allows cells to adapt to an increased state of oxidative stress by suppressing the programmed cell death, either by counteracting the radical overproduction imposed by cell death stimuli or by fortifying the cellular antioxidant defenses (Hockenbery, 1995; Chen et al., 1997). On the other hand, Bax is involved in the programmed cell death (Oltvai et al., 1993; Hockenbery, 1995).

Cilostazol was introduced to increase the intracellular level of cyclic AMP by blocking its hydrolysis by type III phosphodiesterase (Kimura et al., 1985) and is approved for use for treating intermittent claudication by the Food and Drug Administration (Dawson et al., 1998). Its principal actions include inhibition of platelet aggregation (Kimura et al., 1985; Kohda et al., 1999), antithrombosis in feline cerebral ischemia, and vasodilation via mediation of increased cyclic AMP level (Tanaka et al., 1989).

In the current study, we sought to examine the potential neuroprotective effects of cilostazol on the cerebral infarct size and on the DNA fragmentation after subjecting the rats to 2-h occlusion of MCA and 24-h reperfusion. We performed DNA fragment assay to identify the apoptotic cell damage, and analyzed the changes in Bcl-2, as well as Bax level and cytochrome c release from mitochondria after treatment with cilostazol.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Preparation of Animals. All animal studies conformed to the guidelines outlined in the Guide for Animal Experiments edited by the Korean Academy of Medical Sciences and were approved by the Animal Experimental Committee of College of Medicine, Pusan National University.

Analysis of Cerebral Infarct. At 24 h of reperfusion after 2-h MCA occlusion, rats were given an overdose of thiopental sodium and decapitated, and then the brain was quickly removed and frozen in liquid nitrogen. The brain was cut into a 2-mm thick coronal block. The brain slices were immersed in 2% solution of 2,3,5-triphenyltetrazolium chloride in normal saline at 37°C for 30 min and then fixed in 10% phosphate-buffered formalin at 4°C. The 2,3,5-triphenyltetrazolium chloride-stained brain slices were photographed using a charge-coupled device video camera and the size of an infarct was calculated with the image analysis system (Image-Pro Plus; Media Cybernetics LP, Silver Spring, MD) and expressed as the percentage of infarcted tissue in comparison with the area of the ipsilateral hemisphere. Infarct volume (in cubic millimeters) was determined by multiplying the appropriate area by the interval thickness of each section. Rats received 3 or 10 mg/kg cilostazol dissolved in dimethylsulfoxide at 5 min, 1 h, or 3 h after the completion of 2-h MCA occlusion, respectively. Control rats received 300 µl of 30% dimethylsulfoxide solution without cilostazol.

Cyclic AMP and TNF- Assay. Each brain sample was quickly removed at 1, 4, 12, 24 h reperfusion after 2-h occlusion of MCA. Then, the samples were homogenized with 0.32 M sucrose containing 200 µM phenylmethylsulfonyl fluoride and each 5 µg/ml leupeptin, pepstatin, and antipain. Cell debris was removed by centrifugation at 15,000g for 10 min, and the supernatant was used for enzyme-linked immunosorbent assay. The contents of cyclic AMP and TNF- of the homogenized brain extracts were measured using the immunoassay kits (DE0355 and RTA00; R&D Systems Inc., Minneapolis, MN). Protein concentration was determined using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA) with bovine serum albumin as a standard.

Determination of Hydroxyl Radical Scavenging. Hydroxyl radical scavenging efficacy of cilostazol was determined in air-saturated phosphate buffer (10 mM) at room temperature. The reaction was initiated by addition of a small aliquot (5 µl) of Fe²⁺ solution (10 mM FeSO₄ in 10 mM HCl) to a buffer containing cilostazol, H₂O₂ (0.12 mM), and 5,5-dimethyl-1-pyrroline-N-oxide (DMPO, 1 mM). The sample was transferred quickly to flat quartz EPR cell, and measurements were started immediately. In EPR study, 0.1% dimethylsulfoxide used as a vehicle did not influence the EPR signals.

Peroxyl Radical Absorbing Capacity (PRAC) Assay. According to the method described by Cao et al. (1993), the assay was based on the production of peroxyl radicals by 2,2'-azobis(2-amidinopropane) hydrochloride (3 mM), a peroxyl radical generator, with subsequent oxidation of the reporter protein -phycoerythrin (16.7 nM) in 24-well plates. Into each sample well, either 20 µl of 0.1% dimethylsulfoxide in phosphate buffer or 20 µl of each concentration of cilostazol was included. After adding 2,2'-azobis(2-amidinopropane) hydrochloride, the reaction mixture was incubated at 37°C. Loss of fluorescence was measured every 5 min at the emission of 590 nm and excitation of 485 nm using the Fluorescence Plate Reader (Bio-Tek Instruments Inc., Winooski, VT). The PRAC value of the compound is reflected by the increase of area under curve of fluorescence versus time. Trolox was used as a reference for PRAC assay. The fluorescence just prior to addition of the 2,2'-azobis(2-amidinopropane) hydrochloride was estimated as the 100% value for that sample. The PRAC values were calculated as follows: PRAC = [area of compound  area of blank]/[area of 1 µM trolox  area of blank], where 1 PRAC unit is equivalent to the value of 1 µM trolox.

DNA Fragmentation Assay. After 2-h MCA occlusion/24-h reperfusion, samples were dissected from the region corresponding to the penumbra zone and contralateral control areas. The brain was cut in a cryostat to produce a standard coronal block. For oligonucleosomal fragmentation of genomic DNA, cells were lysed in 1 ml of lysis buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM EDTA, 1% sodium dodecyl sulfate, and 0.5 mg/ml proteinase K). Digestion was continued for 1 to 3 h at 55°C, followed by addition of ribonuclease A to 0.1 mg/ml and running dye (10 mM EDTA, 0.25% bromophenol blue, 50% glycerol). Equivalent amounts of DNA (15-20 µg) were loaded into wells of a 1.6% agarose gel and electrophoresed in 0.5 × Tris-acetate EDTA buffer (40 mM Tris-acetate, 1 mM EDTA) for 2 h at 6 V/cm. DNA was visualized by ethidium bromide staining. Gel pictures were taken by the UV transillumination with a Polaroid camera.

Western Blot Analyses. After 2-h MCA occlusion/24-h reperfusion, the samples corresponding to the penumbra zone were homogenized, and cells were lysed in lysis buffer containing 50 mM Tris-Cl (pH 8.0); 150 mM NaCl; 0.02% sodium azide; 100 µg/ml phenylmethylsulflonyl fluoride; 1 µg/ml aprotinin, and 1% Triton X-100. Following centrifugation at 12,000 rpm, 50 µg of protein of each sample was loaded into 12% SDS-polyacrylamide electrophoresis gel and transferred to nitrocellulose membrane (Amersham Biosciences, Piscataway, NJ). The blocked membranes were then incubated with the antibody to Bcl-2 and Bax (Santa Cruz Biotechnology, Inc., Santa Cruz, CA).

Drugs. Cilostazol (6-[4-(1-cyclohexyl-1H-tetrazol-5-yl)butoxy]-3,4-dihydro-2-(1H)-quinolinone) was generously donated from Otsuka Pharmaceutical Co. Ltd (Tokushima, Japan). Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid; Alexis Biochemicals, San Diego, CA) was dissolved in dimethylsulfoxide as a stock of 10 mM. DMPO (Sigma-Aldrich, Seoul, Korea) was purified by double distillation and stored at 70°C before use. -Phycoerythrin (Sigma-Aldrich) and 2,2'-azobis(2-amidino-propane) dihydrochloride (Wako Pure Chemical Co., Osaka, Japan) were dissolved in 75 mM phosphate buffer (pH 7.0).

Statistical Analysis. Differences between data of infracted area and volume in each section between groups were evaluated by performing the Wilcoxon test. Two-way repeated measures analysis of variance were used for comparison of the results of PRAC assay. Other data were analyzed with Student's t test for comparison of two means. Results are expressed as means ± S.E.M. Differences were considered to be significant when P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Cilostazol on Infarct Size and Volume. The ischemic zone was consistently identified in the cortex and striatum of the left cerebral hemisphere as a distinct pale-stained area in the rats subjected to 2-h ischemia/24-h reperfusion. The infarct area was significantly reduced when the animals received cilostazol (10 mg/kg) 5 min or 1 h after the completion of 2-h ischemia, respectively. It was, however, not the same case when rats received the drug 3 h after 2-h ischemia (Fig. 1). Accordingly, the infarct volume (vehicle, 162.1 ± 31.1 mm³) was significantly diminished to 67.2 ± 28.2 mm³ and 70.5 ± 17.6 mm³ in the cilostazol-treated group when administered at 5 min or 1 h after the completion of 2-h ischemia, respectively (Fig. 1, Inset).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Inhibitory effect of cilostazol on the reduction in infarct size identified according to the time of cilostazol administration. When cilostazol (10 mg/kg) was administered at 5 min or 1 h, but not 3 h, after the completion of 2-h ischemia, infarct size significantly decreased in the 2nd, 3rd, and 4th coronal sections in comparison to vehicle group. Inset, cerebral infarct volume was analyzed. Results are expressed as means ± S.E.M. of six to seven experiments. *, P < 0.05; **, P < 0.01 versus vehicle (Veh).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Dose-dependent effect of cilostazol on the infarct area of each coronal section between vehicle- and cilostazol-treated rats subjected to 2-h ischemia/24-h reperfusion. Cilostazol (3 or 10 mg/kg, i.v.) was injected at 5 min after the completion of 2-h ischemia. Infarct area significantly decreased in the 2nd, 3rd, and 4th coronal sections by treatment with 10 mg/kg cilostazol. Results are expressed as means ± S.E.M. of six to seven experiments. *, P < 0.05; **, P < 0.01 versus vehicle.

Antiapoptotic Effect. At 24-h of reperfusion after MCA occlusion, the samples corresponding to the penumbra zone were obtained. A strong staining for terminal deoxynucleotidyl transferase-mediated dUTP-biotin in situ nick-end labeling was present in a moderate to large number of cells in the vehicle-treated ischemic brain, which became more conspicuous at 48 h after reperfusion. The number of terminal deoxynucleotidyl transferase-mediated dUTP-biotin in situ nick-end labeling-positive stained cells was significantly reduced in cilostazol-treated ischemic brains (data not shown).

View larger version (75K): [in this window] [in a new window]   Fig. 3.   Representative findings of agarose gel electrophoresis showing DNA laddering. At 24-h of reperfusion after 2-h MCA occlusion, the samples corresponding to the penumbra zone were obtained. DNA from penumbra lesion showed signs of internucleosomal fragmentations evidenced by a laddering pattern. A, rats received cilostazol (10 mg/kg) at 5 min, 1 h, or 3 h after the completion of 2-h ischemia. Reduction in DNA fragmentation was more prominent in the samples obtained from rats that received cilostazol at 5 min rather than 1 h or 3 h after ischemia. B, cilostazol (3 and 10 mg/kg; CSZ-3 and CSZ-10) was administered at 5 min after 2-h MCA occlusion. The experiment was done in duplicate with different samples from different rats. M, Lambda DNA/HindIII marker; CSZ, cilostazol.

Western Blot Analyses. Figure 4 shows Western blot results for Bcl-2 and Bax protein and release of cytochrome c following treatment with cilostazol. Samples from normal rats showed a considerable amount of Bcl-2 protein but trace level of Bax protein and cytochrome c release. When samples were obtained from rats subjected to 2-h MCA occlusion and 24-h reperfusion, Bcl-2 protein showed a markedly reduced level, whereas Bax protein and cytochrome c release greatly increased. Following treatment with cilostazol, both Bax protein and cytochrome c release were significantly reduced with increasing doses of cilostazol. In contrast, Bcl-2 level dose dependently increased.

View larger version (72K): [in this window] [in a new window]   Fig. 4.   Representative findings of Western blot for Bcl-2, Bax proteins, and cytochrome c release from mitochondria in the penumbra regions isolated from normal control brains and cilostazol (3 and 10 mg/kg; CSZ-3 and CSZ-10)-treated brains. Cilostazol was administered at 5 min after the completion of 2-h occlusion of MCA. Both Bax protein and cytochrome c immunoreactivity were markedly increased, whereas Bcl-2 protein was decreased in the vehicle-treated ischemic brains. These variables were efficiently reversed by cilostazol treatment dose dependently. The experiment was conducted in duplicate with different samples from different rats. CSZ, cilostazol.

View larger version (41K): [in this window] [in a new window]   Fig. 5.   Densitometric analyses of Bcl-2, Bax proteins, and cytochrome c release following treatment with cilostazol (3 and 10 mg/kg; CSZ-3 and CSZ-10) in comparison with vehicle-treated group. Cilostazol was administered at 5 min after the completion of 2-h occlusion of MCA. Increased Bax protein and cytochrome c and decreased Bcl-2 protein shown in the vehicle group were substantially reversed by cilostazol treatment. Results are expressed as means ± S.E.M. of four experiments. CSZ, cilostazol. ###, P < 0.001 versus control; ***, P < 0.01 versus vehicle.

Effect of Cilostazol on Cyclic AMP Level. The cyclic AMP levels in the hemispheres of untreated, nonischemic control rats were 3.1 ± 0.5 pmol/mg of protein, which was significantly elevated to 7.2 ± 0.2 pmol/mg of protein (P < 0.01) following treatment with cilostazol (10 mg/kg, i.v.). Following 2-h occlusion, rats were decapitated at 1, 4, 12, and 24 h, respectively. The cyclic AMP levels from brains subjected to 2-h ischemia followed by 4-h (1.3 ± 0.3 pmol/mg of protein) and 12-h reperfusion (1.3 ± 0.4 pmol/mg of protein) were significantly elevated to 4.9 ± 0.4 and 3.8 ± 1.0 pmol/mg of protein (P < 0.01) by pretreatment with cilostazol, respectively (Fig. 6). However, the cyclic AMP levels after cilostazol treatment were little different between ipsilateral and contralateral side of brains subjected to MCA occlusion (data not shown).

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Effect of cilostazol on the cyclic AMP level in the ischemic brain in comparison to the vehicle-treated one. Rats received 10 mg/kg cilostazol i.v. 5 min after 2-h occlusion of MCA. Samples for cyclic AMP assay were obtained at 1, 4, 12, and 24 h during reperfusion after the completion of 2-h occlusion. Control represents normal intact rat. Results are expressed as means ± S.E.M. of four to five experiments. ##, P < 0.01 versus corresponding vehicle. *, P < 0.05; **, P < 0.01 versus control of cilostazol treatment.

Effect of Cilostazol on TNF- Levels. The level of TNF- in the untreated, nonischemic control hemispheres was 4.7 ± 8.7 pg/mg of protein. In the ischemic hemispheres, the level of TNF- was highly elevated to 48.7 ± 5.3, 56.9 ± 7.1, and 105.8 ± 4.9 pg/mg of protein at 1-, 4-, 12-h reperfusion after 2-h occlusion, respectively, and then returned to the basal level at 24-h reperfusion. Those increases were significantly suppressed by treatment with cilostazol (10 mg/kg, i.v., 5 min after the completion of 2-h ischemia) as shown in Fig. 7.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Effect of cilostazol on the TNF- level in the ischemic brain in comparison to the vehicle-treated one. Rats received 10 mg/kg cilostazol i.v. 5 min after 2-h ischemia. Samples for TNF- assay were obtained at 1, 4, 12, and 24 h during reperfusion after the completion of 2-h occlusion. TNF- significantly increased in the brains subjected to MCA occlusion from 1 to 12 h and then returned to basal level at 24 h. Cilostazol significantly inhibited the TNF- level. Results are expressed as means ± S.E.M. of five to six experiments. ##, P < 0.01; ###, P < 0.001 versus control. *, P < 0.05; **, P < 0.01 versus corresponding vehicle.

Hydroxyl Radical Scavenging Effect of Cilostazol. Figure 8 shows the EPR spectra of the spin adduct of DMPO with the hydroxyl radical, DMPO/^·OH, which was observed when DMPO reacted with hydroxyl radical generated by the Fenton system. Scavenging of the hydroxyl radicals was confirmed by using catalase (0.5-10 U/ml). Cilostazol potently inhibited the DMPO/^·OH adduct formation in a concentration-dependent manner. The signals were almost completely suppressed by 10⁵ M cilostazol. The concentration required for inhibiting the hydroxyl radical formation by 50% (IC₅₀) was 2.58 ± 0.07 µM. However, cilostazol did not inhibit the formation of DMPO/^·OOH (data not shown).

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Representative drawings of the EPR spectra of spin adduct of hydroxyl radicals. The signals of DMPO/·OH spin adduct were detected upon exposure to the H2O2-ferrous sulfate system in the presence of DMPO. Cilostazol potently and concentration dependently inhibited the DMPO/·OH spin adduct. The concentration required for inhibiting the hydroxyl radical formation by 50% (IC50) was 2.58 ± 0.07 µM.

Peroxyl Radical Absorbing Capacity (PRAC). Peroxyl radical absorbing ability of cilostazol was examined using -phycoerythrin. 2,2'-Azobis(2-amidinopropane) hydrochloride was used as a source of peroxyl radicals. Figure 9 shows the time-dependent decrease of -phycoerythrin fluorescence in the absence (blank) and presence of different concentrations of cilostazol. The decrease of the -phycoerythrin fluorescence showed a delay time dependence on the concentration of the antioxidants for 100 min. In the presence of 10⁶ and 10⁴ M cilostazol, a significant right shift of the extinction curve (P < 0.05, analysis of variance) was observed, suggestive of a peroxyl radical scavenging effect. The PRAC value for cilostazol was similar to that of trolox (Fig. 9, Inset).

View larger version (37K): [in this window] [in a new window]   Fig. 9.   Graph showing the time-dependent decrease of -phycoerythrin (-PE) fluorescence in the absence (blank) and presence of different concentrations of cilostazol and trolox. When 2,2'-azobis(2-amidinopropane) hydrochloride was used as a source of peroxyl radicals, peroxyl radical absorbing ability of cilostazol was determined by employing -PE. Inset, PRAC values were calculated. Each point represents means ± S.E.M. of four experiments. **, P < 0.01 versus 106 M trolox.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The current study shows that cilostazol decrease of infarct size was associated with decreased oligonucleosomal DNA fragmentation, increased Bcl-2, decreased Bax protein, and reduced cytochrome c release from mitochondria. Furthermore, cilostazol increased cyclic AMP production and suppressed TNF- in the ischemic cortex. This compound additionally showed a potent ability to scavenge hydroxyl and peroxyl radicals in in vitro experiments.

The present results showed that cerebral infarct volume was significantly reduced when the animals received cilostazol at 5 min and 1 h, but not 3 h, after the 2-h ischemia. Cilostazol is known to inhibit platelet aggregation and produce vasorelaxation via activation of cyclic AMP as a type III phosphodiesterase inhibitor (Tanaka et al., 1989). Cilostazol was recently approved by the Food and Drug Administration for treatment of intermittent claudication (Dawson et al., 1998). Gotoh et al. (2000) reported that cilostazol treatment achieves a considerable risk reduction (about 41.7%) in patients with recurrent cerebral infarction. An elevation of cyclic AMP was demonstrated to suppress the generation of superoxide anion and hydrogen peroxide in alveolar macrophages (Takei et al., 1998). The present results show that cilostazol effectively scavenged hydroxyl and peroxyl radicals. Previous work (Kim et al., 2002) showed that cilostazol reduces intracellular hydrogen peroxide, highlighting the ability of cilostazol to react with a wide spectrum of radical molecules.

Yang et al. (1994) showed that increased infarct size observed at 24 h after MCA occlusion was significantly decreased in transgenic mice overexpressing human copper-zinc superoxide dismutase, suggestive of the importance of the oxygen free radicals in the ischemic brain injury. Cilostazol-induced reduction of cerebral infarct size may correlate with both ROS and increasing cyclic AMP. It is likely that the unique pharmacological profile of cilostazol, both increased cyclic AMP and scavenging effect of oxygen radicals, contributes to the current findings of inhibition of apoptosis and decreased cerebral infarct. The fact that increased cyclic AMP suppressed the generation of superoxide and hydrogen peroxide (Takei et al., 1998) might suggest the synergistic effect of cyclic AMP and oxygen radical scavengers. Whether cyclic AMP is mechanistically involved is uncertain because rat cerebral cortex has low levels of phosphodiesterase III (Challiss and Nicholson, 1990).

TNF-, a deleterious cytokine in stroke, was demonstrated to mediate inflammatory, thrombogenic, and vascular changes in association with brain injury (Kochanek and Hallenbeck, 1992; Feuerstein et al., 1994). Increased level of TNF- in brain tissue has been found in cerebral ischemia (Lavine et al., 1998), causing neuronal cell death via induction of free radicals in glial cells (Hu et al., 1997) and apoptosis (Böhler et al., 2000). Recently, cyclic AMP elevating agents such as Ro-201724, amrinone, milrinone, and pentoxyphylline inhibited TNF- production in rat hearts and glial cells (Yoshikawa et al., 1999). The ROS including H₂O₂ and its derived form, hydroxyl radical (Li et al., 1997), are implicated in the signaling pathways initiated by TNF-, which is in turn involved in apoptosis (Kroemer et al., 1995; Böhler et al., 2000). In concert with these reports, it is suggested that decreased TNF- level was closely related with increased cyclic AMP levels, and the free radical-scavenging action of cilostazol further ameliorates the consequences observed after cerebral ischemia by reducing TNF- levels.

Recently, accumulating evidence points to a significant role for Bcl-2 and related proteins in promoting cell survival and cell death (Bredesen, 1995). Martinou et al. (1994) showed that overexpression of Bcl-2 in transgenic mice protects neurons from ischemia-induced cell death. ROS, including H₂O₂ and hydroxyl radical, and lipid hydroperoxides are all implicated in the processes of apoptosis (Kroemer et al., 1995; Li et al., 1997), and they mediate cytokine (i.e., TNF- and IL-1)-induced apoptosis in rat mesangial cells (Böhler et al., 2000). Bcl-2 protects the integrity of mitochondrial oxidative phosphorylation and thus limits mitochondrial dysfunction induced by several apoptosis stimuli (Kluck et al., 1997). In our results, low levels of the Bcl-2 and high levels of Bax were found in the penumbra regions of ischemic brains in association with increased cytochrome c release. Interestingly, even postischemic administration of cilostazol could reverse these increased levels of Bax and cytochrome c as well as the decreased Bcl-2 levels induced by MCA occlusion.

Bax is believed to be a cell-death effector, the activity of which is neutralized by binding of Bcl-2 (Sato et al., 1994). Our results that cilostazol strongly suppressed MCA occlusion-induced up-regulation of Bax protein provide support for the postulate that cilostazol can protect against anti-ischemia-induced apoptosis. Mitochondria is an important regulatory site of apoptosis (Kroemer, 1998), especially in relation to the rise of cytochrome c release from mitochondria to cytosol, thereby governing apoptosis (Zhang et al., 2000). Most recently, it was suggested that Bcl-2 prevents the loss of the mitochondrial membrane potential and the release of cytochrome c to cytosol (Gross et al., 1999), whereas Bax protein promotes apoptosis by triggering the release of cytochrome c from mitochondria and activation of caspase cascade (Jürgensmeier et al., 1998). ROS produced endogenously are known to enhance the permeability of the mitochondrial membrane and the release of cytochrome c to the cytosol (Shimizu et al., 1999). Consistent with other reports, our results show up-regulation of Bcl-2 and down-regulation of Bax protein and cytochrome c release that coincide with an impressive neuroprotective effect of cilostazol to suppress the DNA fragmentation and brain infarct due to ischemic injury after MCA occlusion.