Heart and Circulatory Physiology

Atorvastatin protects against cerebral infarction via inhibition of NADPH oxidase-derived superoxide in ischemic stroke

Hua Hong, Jin-Sheng Zeng, David L. Kreulen, David I. Kaufman, Alex F. Chen


Statins have recently been shown to exert neuronal protection in ischemic stroke. Reactive oxygen species, specifically superoxide formed during the early phase of reperfusion, augment neuronal injury. NADPH oxidase is a key enzyme for superoxide production. The present study tested the hypothesis that atorvastatin protects against cerebral infarction via inhibition of NADPH oxidase-derived superoxide in transient focal ischemia. Transient focal ischemia was created in halothane-anesthetized adult male Sprague-Dawley rats (250–300 g) by middle cerebral artery occlusion (MCAO). Atorvastatin (Lipitor, 10 mg/kg sc) was administered three times before MCAO. Infarct volume was measured by triphenyltetrazolium chloride staining. NADPH oxidase enzymatic activity and superoxide levels were quantified in the ischemic core and penumbral regions by lucigenin (5 μM)-enhanced chemiluminescence. Expression of NADPH oxidase membrane subunit gp91phox and membrane-translocated subunit p47phox and small GTPase Rac-1 was analyzed by Western blot. NADPH oxidase activity and superoxide levels increased after reperfusion and peaked within 2 h of reperfusion in the penumbra, but not in the ischemic core, in MCAO rats. Atorvastatin pretreatment prevented these increases, blunted expression of membrane subunit gp91phox, and prevented translocation of cytoplasmic subunit p47phox to the membrane in the penumbra 2 h after reperfusion. Consequently, cerebral infarct volume was significantly reduced in atorvastatin-treated compared with nontreated MCAO rats 24 h after reperfusion. These results indicate that atorvastatin protects against cerebral infarction via inhibition of NADPH oxidase-derived superoxide in transient focal ischemia.

  • transient focal ischemia
  • neuronal injury
  • inducible nitric oxide synthase
  • oxidative stress
  • statins

the cellular mechanisms involved in ischemic brain injury are multifactorial and remain incompletely understood. Reactive oxygen species (ROS) formed during ischemia and reperfusion contribute to neuronal injury, and suppression of ROS alleviates oxidative stress-induced neuronal damage (1). Superoxide is the first ROS generated in the oxygen free radical chain during the early phase of reperfusion and contributes to neuronal injury (9). NADPH oxidase is a principal enzyme for the production of superoxide (4). It is expressed in neural and nonneural cell types of the central nervous system, including cerebral cortical neurons, hippocampal pyramidal neurons, and cerebellar Purkinje cells, as well as in microglial cells, monocytes, and neutrophils (3, 17). NADPH oxidase is a complex enzyme that consists of the membrane subunit cytochrome b558 (p22phox and gp91phox), multiple cytoplasmic subunits (p47phox, p67phox, and p40phox), and the small G protein Rac-1 (2). Expression of gp91phox, which is the catalytic subunit of the enzyme, has been shown to be augmented in ischemic stroke (7). Conversely, gp91phox-mutant (X-CGD) mice exhibit reduced cerebral infarction (21). These findings indicate that NADPH oxidase plays a key role in ROS-induced neuronal damage in ischemic stroke.

3-Hydroxy-3-methylglutaryl-CoA reductase inhibitors (statins), the mainstay cholesterol-lowering agents, have recently been shown to protect against ischemic stroke independent of their lipid-lowering effect (13). An important mechanism by which statins protect against ischemic stroke is upregulation of endothelial nitric oxide (NO) synthase (13). Additionally, some studies have shown that statins also have an antioxidant effect and can downregulate NADPH oxidase in cultured rat aortic smooth muscle cells (14). Therefore, the present study tested the hypothesis that statins protect against cerebral infarction by inhibiting NADPH oxidase-derived superoxide and increasing cerebral blood flow (CBF) in transient focal ischemia. Our findings indicate that atorvastatin suppresses NADPH oxidase enzymatic activity by inhibiting gp91phox and p47phox expression, resulting in decreased superoxide levels in the early phase of ischemia-reperfusion and reduced cerebral infarction.


Animal and drug preparations.

Adult male Sprague-Dawley rats (250–300 g body wt) were housed in groups (3 per cage) under controlled conditions at 23 ± 3°C in a 12:12-h light-dark cycle. All experiments were approved by the Institutional Animal Care and Use Committees of Michigan State University and the First Affiliated Hospital of Sun Yat-Sen University. Experiments were divided into three groups: normal, middle cerebral artery occlusion (MCAO) (8), and MCAO with atorvastatin treatment. Atorvastatin (Lipitor, Pfizer) was dissolved in a PBS-10% ethanol solution at 2 mg/ml (pH 7.6) and administered by subcutaneous injection (10 mg/kg) 48, 24, and 2 h before reperfusion according to an established protocol that demonstrated the effectiveness of the same regimens (6).

Physiological measurements.

Blood pressure and heart rate were continuously measured with a transducer-recorder (Quad Bridge ML118, AD Instruments) from 15 min before to 15 min after MCAO, as we previously described (8). Regional CBF was measured with a laser-Doppler perfusion monitor (Perimed) (8). Briefly, a small incision was made at the midpoint between the left orbit and the external auditory canal under anesthesia. The temporalis muscle was retracted, and the underlying fascia was cleared. A small area of skull ∼1 mm posterior and 5 mm lateral to the bregma was thinned to allow placement of the laser-Doppler probe (8). CBF was monitored and compared before and after MCAO in the animals treated with atorvastatin or saline.

Focal cerebral ischemia.

Rats were anesthetized with 1.5% halothane and maintained on 1.0% halothane in 70% N2O-30% O2 with a face mask. Body temperature was monitored by a rectal probe and maintained at 37–37.5°C by an automatic homeothermic blanket control unit (Harvard Apparatus). The physical parameters of all rats included in the study were within normal limits. The MCAO model was established by proximal occlusion of the left middle cerebral artery with the use of a nylon monofilament, as we previously described (8). Briefly, after a 2-h occlusion, reperfusion was accomplished by careful withdrawal of the monofilament. The common carotid artery was ligated distal to the incision. The time of reperfusion (0–24 h) depended on the experimental design. After reperfusion, the brains were harvested. Infarct volume was evaluated by 2,3,5-triphenyltetrazolium chloride staining, as described elsewhere (7, 8).

Ischemic core and penumbra dissections.

The ischemic core and penumbra were dissected according to well-established protocols in rodent models of unilateral proximal MCAO (1). Briefly, each hemisphere was cut longitudinally, from dorsal to ventral ∼1.5 mm from the midline to exclude medial brain structures that were supplied primarily by the anterior cerebral artery. A transverse diagonal incision at approximately the “2 o'clock” position separated the core from the penumbra (1).

Measurements of brain NADPH oxidase activity and superoxide levels.

NADPH oxidase enzymatic activity and superoxide levels were quantified by lucigenin-enhanced chemiluminescence, as we described previously (11, 12). For NADPH oxidase assay, brain tissues from the ischemic core and penumbra were homogenized and centrifuged, and the supernatants were assayed for protein concentration (Bio-Rad). NADPH oxidase activity was measured in the supernatants in the presence of its substrate NADPH (10−4 mol/l; Sigma) and lucigenin (5 × 10−6 mol/l; Sigma). No enzymatic activity could be detected in the absence of NADPH. Reactions were initiated by the addition of 10–20 μl of tissue homogenates containing 25–50 μg of extracted proteins. The enzyme activity was expressed as millimoles per minute per milligram of protein. For superoxide assay, fresh brain tissues were placed in polypropylene tubes containing 5 μmol/l lucigenin in 1 ml of modified Krebs solution and incubated for 30 min before luminometer (model TD-20/20) readings were taken. Photomultiplier background signal was determined in brain-free preparations and automatically subtracted. The adjusted readings were converted to nanomoles per minute per milligram of tissue via a standard curve generated by cytochrome c (Sigma) reduction from superoxide derived from hypoxanthine (Sigma) and xanthine oxidase (Sigma).

Extraction of membrane-bound proteins by subcellular fractionation.

Differential centrifugation was used for isolation of membrane proteins as described elsewhere (15). Briefly, brain tissues were homogenized in ice-cold HEPES buffer (20 mM HEPES, 150 mM NaCl, and 1 mM EDTA, pH 7.4), to which the protease inhibitor cocktail (Sigma) was added just before homogenization. Cell nuclei and unbroken cells were removed by centrifugation of the homogenates at 2,800 g at 4°C for 20 min. The supernatants were divided into two parts: one was used for assay of gp91phox protein, and the other was further centrifuged at 100,000 g for 60 min at 4°C to separate the membrane proteins. The pellets, which contained the membrane protein fraction, were resuspended in HEPES buffer containing 1% Triton for 20 min on ice and used for assay of membrane-translocated p47phox and Rac-1 proteins (15).

Western blotting of gp91phox, membrane-translocated p47phox, and Rac-1 proteins.

The sample proteins, which were extracted as described above, were subjected to SDS-PAGE. The separated proteins were transferred by electrophoresis to polyvinylidine difluoride membranes, as we described previously (11). The membranes were incubated with anti-gp91phox antibody (1:1,000 dilution; BD Transduction Laboratories), anti-p47phox antibody (1:1,000 dilution; Upstate Biotechnology), and anti-Rac-1 antibody (1:5,000, Upstate Biotechnology), respectively, at 4°C overnight and then incubated with a horseradish peroxidase-linked secondary antibody (1:5,000 dilution; Santa Cruz Biotechnology) for 40 min. The positive bands were revealed using Western blotting detection reagents (Pierce) and autoradiography film. Relative molecular band intensity was determined by densitometry (NIH Scion Image).

Statistical analysis.

Values are means ± SE. One-way ANOVA with Turkey's post hoc test was used to compare multiple-group values (i.e., measurements of NADPH oxidase activity, superoxide levels, and gp91phox, p47phox, and Rac-1 proteins). Unpaired Student's t-test was used to compare two-group data (i.e., cerebral infarct volume). P < 0.05 was considered to be statistically significant.


Physiological parameters.

Blood pressure and heart rate, as monitored from 15 min before to 15 min after MCAO, were not significantly different throughout the procedure (Table 1). Regional CBF was reduced to ∼15% of the baseline levels (i.e., 100%) measured before MCAO. There was no significant difference in CBF levels between the atorvastatin and control groups before MCAO.

View this table:
Table 1.

Physiological parameters of MCAO rats

Atorvastatin reduces infarct volume 24 h after reperfusion in MCAO rats.

Reproducible infarcts in the cerebral cortex and striatum were observed 24 h after MCAO in control rats (Fig. 1A) and were significantly reduced with atorvastatin pretreatment: 44.6 ± 2.7 vs. 21.4 ± 5.1% (P < 0.01, n = 5–6; Fig. 1B).

Fig. 1.

A: typical sections of ischemic brain in rats subjected to 2 h of middle cerebral artery occlusion (MCAO) followed by 24 h of reperfusion with or without atorvastatin pretreatment. MCAO produced an ipsilateral focal infarct (white area), as shown by 2,3,5-triphenyltetrazolium chloride staining. B: atorvastatin pretreatment significantly reduced infarct volume compared with nontreated MCAO rats. Values are means ± SE of 5–6 rats in each group. *P < 0.01 vs. MCAO.

Time-dependent changes of NADPH oxidase activity in MCAO rats.

NADPH oxidase activity decreased during the 2-h ischemic period but increased quickly on reperfusion in the core and penumbral regions, reaching peak levels within 2 h after reperfusion (Fig. 2). It was significantly higher in the penumbra than in the normal brain tissues of MCAO rats (62.55 ± 7.61 vs. 32.91 ± 1.54 mmol·min−1·mg protein−1, P < 0.01, n = 4–6) but significantly lower in the ischemic core.

Fig. 2.

Time course of NADPH oxidase activity in penumbral and core regions of MCAO rats. NADPH oxidase activity was assayed during 2 h of ischemia and the following reperfusion periods. Values are means ± SE of 4 rats in each group. *P < 0.01; **P < 0.001; #P < 0.01 vs. baseline.

Atorvastatin impedes peak NADPH oxidase activity and superoxide levels during reperfusion.

Atorvastatin treatment significantly reduced penumbral NADPH oxidase activity of MCAO rats (41.64 ± 2.51 vs. 62.55 ± 7.61 mmol·min−1·mg protein−1, P < 0.05, n = 4–6; Fig. 3A) 2 h after reperfusion, resulting in decreased superoxide levels (0.3 ± 0.07 vs. 0.6 ± 0.04 nmol·min−1·mg tissue−1, P < 0.001, n = 3–4; Fig. 3B).

Fig. 3.

Effect of atorvastatin on NADPH oxidase activity (A) and superoxide levels (B) in penumbral and core regions of MCAO rats 2 h after reperfusion. NADPH oxidase activity and superoxide levels increased significantly in penumbra compared with normal brain tissue and were significantly reduced with atorvastatin treatment. Values are means ± SE of 4–6 rats in each group. *P < 0.001; **P < 0.01 vs. normal. #P < 0.05 vs. MCAO.

Atorvastatin inhibits overexpression of gp91phox and membrane-translocated p47phox.

As shown in Fig. 4, gp91phox and membrane-translocated p47phox proteins increased significantly in the penumbra compared with normal brain tissues of MCAO rats on reperfusion: 149.1 ± 9.3 vs. 100.0 ± 15.4% for gp91phox (P < 0.01) and 577.5 ± 190.5 vs. 99.9 ± 24.9% for p47phox (P < 0.05, n = 6–7). Atorvastatin treatment significantly inhibited gp91phox overexpression (116.0 ± 8.8 vs. 149.1 ± 9.3%, P < 0.01, n = 6–7) and membrane translocation of p47phox (190.1 ± 3.5 vs. 577.5 ± 190.5%, P < 0.05, n = 6–7). In contrast, the membrane-bound Rac-1 protein was not significantly changed with or without atorvastatin treatment.

Fig. 4.

Western blot analyses of gp91phox (A), membrane-translocated p47phox (B), and membrane-bound Rac-1 (C) protein expression. Both gp91phox and membrane-translocated p47phox proteins were significantly increased in penumbra of MCAO rats after 2 h of reperfusion compared with normal rats and were significantly inhibited with atorvastatin treatment. Values are means ± SE of 6–7 rats in each group. *P < 0.05 vs. normal. #P < 0.05 vs. MCAO. Membrane-bound Rac-1 protein level in MCAO rats was not significantly different from that in normal rats. Values are means ± SE of 4–5 rats in each group. P > 0.05.


The major new findings of the present study are as follows: 1) After a transient decrease in response to ischemia, NADPH oxidase activity in the penumbral and core regions rapidly and markedly increased on reperfusion and peaked within 2 h. 2) Treatment with atorvastatin three times before ischemia blunted peak enzymatic activity in the penumbra by impeding the overexpression of its membrane subunit gp91phox and membrane translocation of its cytosolic subunit p47phox, resulting in reduced superoxide levels 2 h after reperfusion but not CBF. 3) Inhibition of NADPH oxidase by atorvastatin accounts, at least in part, for the protection against cerebral infarction in ischemic stroke.

ROS are well known to play a pivotal role in cerebral ischemia-reperfusion-induced cell injury (22). Superoxide is the first ROS generated in the oxygen free radical chain during the early phase of reperfusion after cerebral ischemia (5). The interaction of superoxide with NO results in the production of peroxynitrite, one of the most harmful ROS species, which causes neuronal tissue damage via lipid peroxidation, protein oxidation, nitration, and DNA breakage (8, 20). Recent studies indicate that NADPH oxidase is a predominant source of superoxide generation in the vasculature (4) as well as the central nervous system (17). However, the profile of NADPH oxidase activity during cerebral ischemia-reperfusion is unknown. In the present study, we found that NADPH oxidase activity changes in a time-dependent manner during ischemia-reperfusion and is significantly higher in the penumbral than in the core region. Throughout the course of ischemia, NADPH oxidase activity remained at low levels compared with the normal brain tissues. The low activity of NADPH oxidase may be attributable to the complete blockade of blood flow to the ischemic core area, where most cells die quickly from energy exhaustion. In contrast, penumbral NADPH oxidase activity remained relatively stable during the ischemic phase and sharply increased on reperfusion, before it subsided somewhat 2 h after reperfusion. Although the time course of superoxide formation during the entire period of ischemia-reperfusion was not determined in the present study, we found markedly and significantly higher superoxide levels in the penumbral than in the core region 2 h after reperfusion. The latter findings are also consistent with previously published reports (5, 9). The increased penumbral NADPH oxidase activity and the resultant superoxide generation may thus play an important role in neuronal damage and contribute to enlargement of the infarct size in the late phase of reperfusion. Accordingly, prevention of NADPH oxidase activation in the early phase of reperfusion may represent a useful strategy for protection against oxidative stress-induced neuronal injury in ischemic stroke.

Statins, among the most effective cholesterol-lowering agents, have been shown to protect against ischemic brain damage in several recent studies. The cerebrovascular protective effects of statins have been attributed to upregulation of endothelial NO synthase, reduction of inflammatory responses, increase in plaque stability, and inhibition of thrombus formation (6). In addition, statins exhibit superoxide-scavenging properties in the vasculature (16), which may, in part, be due to their action on NADPH oxidase (19). In the present study, we tested the hypothesis that atorvastatin, one of the most widely prescribed statins, protects against cerebral infarction by suppressing NADPH oxidase activity in focal cerebral ischemia. Our findings demonstrate that atorvastatin treatment (10 mg/kg) three times before reperfusion by subcutaneous injection effectively reduced cerebral infarct volume. This dose regimen has been shown to result in significant neuronal protection (6). We chose the time point of 2 h after reperfusion to assess the effect of atorvastatin on NADPH oxidase-derived superoxide levels, because we found that NADPH oxidase activity had reached its peak levels by then. Our data show that atorvastatin significantly inhibited the increased NADPH oxidase activity and subsequent production of superoxide. Because superoxide is known to interact with NO to form peroxynitrite (7, 8), which also appears at the early phase of reperfusion (18), the cellular injury and death caused by the latter might be alleviated in the penumbral region. Consistent with this notion, infarct volume was significantly smaller in atorvastatin-treated MCAO rats 24 h after reperfusion, perhaps reflecting the dynamic processes in the penumbra during reperfusion.

NADPH oxidase is a complex enzyme that consists of a number of subunits, including the membrane-bound catalytic subunit gp91phox and the cytoplasmic subunits p47phox and small G protein Rac-1, which translocate to the membrane on enzyme activation (2). Accordingly, the expression of gp91phox has been shown to be augmented in ischemic stroke (7), and gp91phox-mutant mice exhibit reduced cerebral infarct volume (20). In the present study, we determined the protein expression profile of the catalytic subunit gp91phox and the membrane-translocated p47phox and Rac-1 in penumbral tissues 2 h after reperfusion. Our results indicate overexpression of gp91phox and membrane-translocated p47phox, but not Rac-1, in parallel with augmented NADPH oxide enzymatic activity. Importantly, atorvastatin pretreatment impeded the expression of gp91phox protein and membrane translocation of p47phox protein, which may account for the decrease of NADPH oxidase activity and resultant superoxide formation. Taken together, these findings suggest that the protection of atorvastatin against cerebral infarction is mediated, at least in part, by its inhibitory action on NADPH oxidase-derived superoxide.

Despite the aforementioned experimental observations, a few issues could not be addressed by this single study. We observed a relatively small but significant increase of gp91phox protein level 4 h after MCAO, along with an increase of the translocated p47phox (membrane-bound). Inasmuch as the latter may not equal an increase in new protein expression, these findings suggest that translocation of the cytosolic subunit of NADPH oxidase (e.g., p47phox) and its subsequent interaction with membrane-bound gp91phox at the protein level, rather than the de novo new enzyme protein synthesis, may account for the increase of NADPH oxidase enzyme activity. In addition, the unchanged Rac-1 level at a single time point may not reflect the whole picture of the Rac-1 expression dynamics over time in this model. Hence, future studies are needed to establish a time course of protein expression for various NADPH oxidase subunits, including the time course of translocated Rac-1 levels in this model. In addition, because our data did not show total protection with atorvastatin, we could not exclude the possible involvement of other pathways and/or cellular systems that may also produce superoxide, which could account for the residual damage to the brain in the face of treatment with atorvastatin.

In conclusion, the present study suggests that NADPH oxidase activity increases quickly and peaks within 2 h after reperfusion in the penumbral region in MCAO rats, which may represent an important mechanism for enlargement of the ischemic core. The study demonstrates further that the protective effects of atorvastatin are, at least partially, mediated by its inhibitory action on NADPH oxidase activity during the early phase of reperfusion in the penumbra, which has been suggested to be a main therapeutic target in acute ischemic stroke.


This work was supported in part by American Heart Association Grants 0130537Z, 0225408Z, and 0455594Z (A. F. Chen), China Medical Board Grant 00730, a 2002 Teaching and Research Award for Outstanding Young Faculty in Higher Education, the Ministry of Education of China, and National Natural Sciences Foundation of China Grants 39940012 and 30271485 (J. S. Zeng).


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract