Address correspondence and reprint requests to De-Maw Chuang, Molecular Neurobiology Section, National Institute of Mental Health, National Institutes of Health, 10 Center Drive, MSC 1363, Bethesda, Maryland 20892-1363, USA. E-mail: email@example.com
Growing evidence from in vitro studies supports that valproic acid (VPA), an anti-convulsant and mood-stabilizing drug, has neuroprotective effects. The present study investigated whether VPA reduces brain damage and improves functional outcome in a transient focal cerebral ischemia model of rats. Subcutaneous injection of VPA (300 mg/kg) immediately after ischemia followed by repeated injections every 12 h, was found to markedly decrease infarct size and reduce ischemia-induced neurological deficit scores measured at 24 and 48 h after ischemic onset. VPA treatment also suppressed ischemia-induced neuronal caspase-3 activation in the cerebral cortex. VPA treatments resulted in a time-dependent increase in acetylated histone H3 levels in the cortex and striatum of both ipsilateral and contralateral brain hemispheres of middle cerebral artery occlusion (MCAO) rats, as well as in these brain areas of normal, non-surgical rats, supporting the in vitro finding that VPA is a histone deacetylase (HDAC) inhibitor. Similarly, heat shock protein 70 (HSP70) levels were time-dependently up-regulated by VPA in the cortex and striatum of both ipsilateral and contralateral sides of MCAO rats and in these brain areas of normal rats. Altogether, our results demonstrate that VPA is neuroprotective in the cerebral ischemia model and suggest that the protection mechanisms may involve HDAC inhibition and HSP induction.
Valproic acid (VPA), a short-chain fatty acid, is widely used for the treatment of seizures and bipolar mood disorder (for review, Tunnicliff 1999; Johannessen 2000). Despite intensive research, the mechanisms underlying the therapeutic effects of VPA remain unclear. It has been recently reported that VPA directly inhibits histone deacetylase (HDAC) at therapeutic levels (with an IC50 = 0.4 mm), causing histone hyperacetylation (Phiel et al. 2001; Göttlicher et al. 2001). HDAC has been strongly implicated in the modulation of gene expression as well as life span in a variety of organisms such as yeast, Caenorhabditis elegans and Drosophila (Chang & Min 2002). In addition, VPA has been shown to activate the cell survival factor, Akt (De Sarno et al. 2002), presumably through inhibition of HDAC (De Sarno et al. 2002). In rat cortical neurons, long-term VPA treatment blocks glutamate-induced excitotoxicity (Hashimoto et al. 2002) and prolongs life span of these cortical cultures (Jeong et al. 2003). Similar to the effect of another mood stabilizing drug, lithium, VPA protects mature rat cerebellar granule cells in cultures from NMDA receptor-mediated excitotoxicity and this action is mimicked by other histone HDAC inhibitors such as butyrate and trichostatin A (Kanai et al. 2002). A growing body of reports also demonstrate that VPA is neuroprotective against a variety of other insults (Mark et al. 1995; Mora et al. 1999; Bown et al. 2000; Wang et al. 2003).
Stroke is one of the leading causes of mortality and morbidity world-wide. Although our knowledge concerning the molecular and cellular pathophysiology of brain injury after focal ischemia has advanced greatly, the development of new treatment drugs for acute ischemic stroke has not progressed as rapidly. The use of intravenous recombinant tissue-type plasminogen activator (rt-PA) for stroke patients has been approved in the US and more recently in other countries. However, the restrictive therapeutic time window of rt-PA limits its clinical effectiveness for stroke (National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group 1995). Another strategy for treating acute stroke patients is the development of neuroprotective drugs. Unfortunately, clinical trials using various pre-clinically neuroprotective drugs for stroke have proven unsuccessful (Broderick and Hacke 2002). Because of increasing prevalence of stroke in the general population and lack of adequate therapies, the research and development of novel and more efficacious drugs for this neurodegenerative disease is imperative. Middle cerebral artery occlusion (MCAO) of rodents provides an excellent model that is relevant to ischemic stroke in humans. Earlier, our laboratory showed that long-term lithium pre-treatment decreases both infarct volume and alleviates neurological abnormalities in a permanent MCAO rat model (Nonaka and Chuang 1998). Moreover, we have shown that post-insult administration with therapeutic doses of lithium also decreases the infarct volume and neurological deficits in a transient ischemia model using MCAO followed by reperfusion (Ren et al. 2003). Additionally, chronic lithium pre-treatment was reported to reduce brain damage in the MCAO/reperfusion model by anti-apoptotic mechanisms (Xu et al. 2003).
It is well established that the heat shock protein (HSP) family has robust cytoprotective effects. HSPs including HSP70 have been shown to be induced in the brain by a variety of pathological insults including cerebral ischemia (for review, Giffard and Yenari 2004). A neuroprotective role of HSP70 in cerebral ischemia is supported by the observations that its overexpression is correlated with the survival of cells following the insult (Kinouchi et al. 1993; Yenari et al. 1998; Fredduzzi et al. 2001) and that gene-mediated HSP70 overexpression protects against brain infarction (Rajdev et al. 2000; Hoehn et al. 2001). Interestingly, the neuroprotection elicited by post-insult lithium treatment is associated with super-induction of HSP70 in the ischemic brain (Ren et al. 2003). The present study used a rat transient ischemia model to examine whether post-insult treatment with VPA has neuroprotective effects in reducing infarct volume and neurological deficits and, if so, to assess whether this neuroprotection is associated with inhibition of HDAC activity and induction of HSP70 in the brain.
Materials and methods
Rat MCAO/reperfusion model and valproic acid treatment
All procedures employing experimental rats were performed in compliance with National Institutes of Health Guide for Care and Use of Laboratory Animals. Male Sprague-Dawley rats weighing from 250 to 300 g were anesthetized with halothane (3% in a mixture of 70% N2O and 30% O2). A 4–0 nylon suture with its tip coated with silicon was inserted from the left external carotid artery into the left internal carotid artery and then to the Circle of Willis to occlude the origin of the left middle cerebral artery. One hour after the initiation of MCAO, the nylon suture was withdrawn and the ischemic brain tissue received blood reperfusion for various times before killing. Blood gas and blood pressure were maintained within their normal ranges during surgery. Body temperature was monitored with a rectal probe and maintained in the range of 36.5°C to 37.5°C with a heating pad. Laser-Doppler flowmetry was used to evaluate the effectiveness of MCAO and reperfusion. Details of the experimental procedures are as described previously (Ren et al. 2003). Rats showing tremor or seizure were excluded from the study. A dose of 300 mg/kg of VPA (Hassel et al. 2001) or normal saline as a vehicle control was injected subcutaneously into rats every 12 h, with the first injection made immediately after the onset of MCAO. For evaluation of activated caspase-3, acetylated histone H3 and HSP70 levels, the animals were killed and the brain samples were collected at 6 h, 24 h, and 48 h after the initiation of MCAO.
Neurological deficits and infarct volume evaluation
The neurological deficits in rats subjected to MCAO/reperfusion were evaluated using 10 different tests for motor, sensation and reflex abnormalities by an individual blinded to the treatment condition, using a modified procedure of Li et al. (2000), as described previously (Ren et al. 2003). In brief, evaluations were made at the time of recovery from surgery and before killing at 24 or 48 h after ischemic insult induction. Six tests of motor performance (flexion of forelimb, flexion of hindlimb, head movement more than 10° to the vertical axis within 30 s, inability to walk straight, circling toward the paralytic side, and falling down to the paralytic side) were included to evaluate hemiplegia in the extremities and trunk. In sensation tests, visual and tactile placing as well as a proprioceptive test was adopted to evaluate sensory abnormalities. In reflex tests, pinna and startle reflex were used to evaluate the loss of responses. A score of 0 (normal) or 1 (unable to perform, abnormal in task performance, or deficit in reflex) was given to each test.
At 24 or 48 h after the onset of MCAO and immediately after neurological deficit evaluation, rats were killed by CO2 asphyxiation, and brains were removed, stained with 2% 2,3,5-triphenyltetrazolium chloride (TTC) for 10 min and fixed overnight in formalin for evaluation of infarct volume. In a blinded manner, infarct volume of six adjacent coronal slices (2-mm thickness) of each brain was quantified by capturing images with a digital camera and then performing computerized analysis using Adobe Photoshop 5.5. Infarct volume determination was corrected for edema by subtracting the volume of intact tissue in the ischemic hemisphere from the volume of the contralateral hemisphere as previously described (Lin et al. 1993).
Coronal sections of 10-µm thickness were cut using a cryostat and fixed with 4% paraformaldehyde for 30 min before staining with an antibody against caspase-3 (rabbit polyclonal; 1 : 200; Cell Signaling Technology, Beverly, MA, USA), which detects the activated form of caspase-3 (17–20 kDa) that results from cleavage after the aspartic acid residue at position 175. Sections were blocked in 5% normal goat serum, rinsed in 1% Triton X-100 and then incubated with the primary antibody at 4°C overnight. After three 10-min washes, brain slices were incubated with the secondary antibody conjugated to FITC (goat anti-rabbit; 1 : 500; Santa Cruz Biotechnology, Santa Cruz, CA, USA). The sections were also double-stained with anti-NeuN monoclonal antibody (mouse monoclonal; 1 : 1000; Chemicon International, Temecula, CA, USA) to identify neuronal cells, with the secondary antibody conjugated to Cy3 (donkey anti-mouse; 1 : 500; Jackson ImmunoResearch Laboratories, West Grove, PA, USA).
Immunoprecipitation and western blotting
Brain tissues samples from the ischemic cortex within left middle cerebral artery territory and the corresponding area in the contralateral side of lithium-treated and vehicle-treated rats were homogenized and sonicated for 40 s in a lysis buffer as previously described (Ren et al. 2003). Protein concentrations were determined using a bicinchoninic acid kit (Pierce, Rockford, IL, USA). For activated caspase-3, immunoprecipitation was performed prior to western blotting. An aliquot of 100 µg of total protein was incubated overnight with 4 µg of anti-activated caspase-3 antibody (rabbit polyclonal; 1 : 25; Cell Signaling Technology) at 4°C. The immuno-complexes were precipitated by incubation with 50 µL of protein A-sepharose (Pharmacia Biotech, St Alban, UK) at 4°C for 4 h.
For western blotting, aliquots of the proteins were separated by electrophoresis on sodium dodecyl sulfate–polyacrylamide gels (10%) and transferred to a polyvinylidene difluoride membrane. Activated caspase-3 was detected by using an anti-activated caspase-3 antibody (rabbit polyclonal; 1 : 1000; Cell Signaling Technology). Acetylated histone H3 was detected by using an antibody raised against acetylated histone H3 on lys9 (rabbit polyclonal; 1 : 1000; Upstate, Lake Placid, NY, USA), and HSP70 was also detected by using its specific antibody (rabbit polyclonal; 1 : 1000; Stressgene Biotechnologies, Victoria, BC, Canada). Briefly, the membranes were rinsed in 0.01 m Tris-buffered saline (pH 7.4) containing 0.1% Triton X-100 for 30 min, blocked in 5% non-fat dry milk for 30 min, and then incubated overnight at 4°C with the primary antibody in Tris-buffered saline containing 3% non-fat dry milk. Membranes were then washed three times with Tris-buffered saline and incubated overnight at 4°C with a horseradish peroxidase-conjugated secondary antibody in Tris-buffered saline containing 3% non-fat dry milk. Immunoreactivity was detected by enhanced chemiluminescent autoradiography (ECL kit; Amersham Life Science, Arlington Heights, IL, USA) in accordance with the manufacturer's instructions. The western blots were captured with a digital camera and the intensities quantified with NIH Image 1.60.
All results were expressed as mean ± SEM. Statistical analysis was performed by anova and Fisher's protected least significant difference. A p-value of ≤ 0.05 was considered to be statistically significant.
Post-insult treatment with VPA reduces MCAO/reperfusion-induced brain injury and neurological deficits in rats
Rats were subjected to MCAO for 1 h followed by reperfusion for various times. Physiological parameters of this MCAO/reperfusion model have been reported previously (Ren et al. 2003). When used, VPA (300 mg/kg) was given immediately after ischemia by subcutaneous injection followed by injections every 12 h thereafter. The body temperatures were found to be unchanged between the VPA-treated groups and saline-treated controls when measured at 20 min prior to ischemia, 20 min post-ischemia and various times (20 min, 24 h and 48 h) post-reperfusion (data not shown).
Brain infarct volumes in the cerebral cortical and striatal areas over a series of sections were determined at 24 and 48 h after MCAO/reperfusion by TTC staining. VPA treatment significantly decreased the infarct volume compared with the saline-treated vehicle control (Figs 1a and b). Neurological deficits in rats subjected to MCAO/reperfusion were assessed by motor, sensory and reflex tests (Fig. 1c). Severe neurological deficits were observed at 1 h after the onset of MCAO and these defects were reduced at 24 and 48 h, similar to our previous observations (Ren et al. 2003). Treatments with VPA caused a marked reduction in neurological deficit scores compared with vehicle control at both 24 and 48 h. Sham-operated rats did not show visible neurological deficits (data not shown).
VPA treatment protects neurons from ischemia-induced caspase-3 activation
Because caspase-3 activation has been implicated in neuronal damage resulting from focal cerebral ischemia (Benchoua et al. 2001; Manabat et al. 2003), the role of activated caspase-3 in ischemia-induced brain damage and its inhibition by VPA-induced neuroprotective effects was also investigated. At 48 h after the onset of MCAO, activated caspase-3 staining (green) was found in the cells within the ischemic cortical area (Fig. 2a1). A large majority of these caspase-3-positive cells expressed NeuN, a neuronal marker, producing yellowish staining as shown by co-labeling with anti-NeuN antibody (Fig. 2a2 and 3). Activated caspase-3 was not detected in the corresponding brain area of sham-operated rats (data not shown). Immunoprecipitation of activated caspase-3 followed by western blotting revealed that activated caspase-3 exisited in the two forms of 17 and 19 kDa (Fig. 2b) at 48 h after the onset of cerebral ischemia. Levels of both forms of activated caspase-3 were robustly suppressed by post-insult VPA treatments (Fig. 2c).
VPA treatment increases levels of acetylated histone H3 in the brain
In vitro studies have shown that VPA is a direct inhibitor of HDAC (Göttlicher et al. 2001; Phiel et al. 2001). However, it is much less clear as to whether VPA also inhibits HDAC activity in vivo. We measured brain levels of acetylated histone H3 levels at lys9, an index of HDAC inhibition, at various times after the onset of MCAO. In the ipsilateral cortical area, VPA treatment increased levels of acetylated histone H3 at 6, 24 and 48 h after ischemia, compared with the corresponding vehicle control (Figs 3a and b). In the ipsilateral side of vehicle-treated control, acetylated histone H3 levels were relatively unchanged between 6 h and 24 h and this was followed by a decline at 48 h. In the contralateral intact hemisphere of vehicle-treated rats, levels of acetylated histone H3 were markedly decreased at 24 and 48 h, compared with 6 h (Figs 3c and d). VPA treatment caused an increase in acetylated histone levels at 6, 24 and 48 h, compared with corresponding vehicle controls. Protein levels of the housekeeping gene β-actin were unchanged in both ipsilateral and contralateral sides during the 48-h period. Treatment of normal, non-surgical rats with VPA also elicited a time-dependent increase in the cortical levels of acetylated histone H3 with no change in β-actin levels (Figs 3e and f). Similar to the results in the cortex, VPA treatment also induced an increase in acetylated histone H3 levels in the striatum of ischemic rats and normal rats. (Figs 4a–f).
VPA treatment up-regulates HSP70 levels in the brain
Several studies reported that HSPs including HSP70 are up-regulated shortly after ischemia and persist in the ischemic brain penumbra where recovery of neurons is most pronounced (for review, Giffard and Yenari 2004), suggesting that there is a correlation between HSP induction and resistance to damage. To evaluate the role of HSP70 in mediating the neuroprotective effects of VPA, we performed western blotting of HSP70 in both ipsilateral and contralateral hemispheres of ischemia. In ipsilateral brain of vehicle-treated rats subjected to MCAO, HSP70 levels were detected at 6 h, reached a maximum at 24 h and declined at 48 h in the ischemic brain area (Figs 5a and b). VPA treatment elicited a further increase in levels of HSP70 at every time point compared with the corresponding vehicle controls. As expected, HSP70 levels were unchanged in the contralateral brain hemisphere of vehicle-treated rats (Figs 5c and d). However, VPA treatment elicited a robust increase in HSP70 levels at all three time points. In the cortex of normal rats, VPA treatment also time-dependently increased levels of HSP70 with little or no change in β-actin levels (Figs 5e and f). Moreover, HSP70 protein levels in the striata of the ipsilateral and contralateral hemisphere of ischemic rats and of normal rats were similarly increased by VPA treatment (Figs 6a–f).
The present study showed that post-insult treatment with VPA reduces brain infarct size when measured at 24 or 48 h after the onset of MCAO-induced ischemia (Fig. 1). VPA also facilitates the functional recovery from neurological deficits under these experimental conditions (Fig. 1). The VPA dose used in the present study, 300 mg/kg, is close to that used in animal studies to control seizures. VPA-induced neuroprotection was further demonstrated by a reduction in MCAO-induced caspase-3 activation in the ischemic area, as shown by immunohistochemistry and western blotting analysis (Fig. 2). These results suggest that VPA neuroprotection in the MCAO model involves anti-apoptotic actions in the ischemic penumbra. It is noteworthy that lithium-induced neuroprotection in the rat MCAO/reperfusion model is also associated with suppression of ischemia-induced caspase-3 activation (Ren et al. 2003; Xu et al. 2003).
In our MCAO/reperfusion paradigm, HSP70 levels were maximally increased at 24 h and the elevation continued through 48 h in the ischemic cortical and striatal areas. During this time period, HSP70 levels were further markedly increased from as early as 6 h by VPA treatment (Fig. 4), also similar to the results of our lithium study (Ren et al. 2003). However, unlike lithium, HSP70 was also robustly up-regulated by VPA in the contralateral hemisphere where ischemic damage did not occur. Thus, VPA-induced up-regulation of HSP70 is insult-independent, dissimilar to the action of lithium at the dosage of 1 mEq/kg, which is insult-dependent. HSP70 may display its anti-apoptotic effects by multimechanisms. It inhibits cytochrome c-dependent activation of caspase-3 (Pandey et al. 2000) and its downstream effector such as phospholipase A2 (Jaattela et al. 1998), as well as suppresses the activity of apoptosis-inducing factor (Ravagnan et al. 2001). HSP has also been shown to inactivate c-Jun-N-terminal kinase, which has been linked to excitotoxic cell death, by direct binding to this kinase or via a dephosphorylation mechanism (Gabai et al. 2000; Park et al. 2001). Moreover, the heat shock response inhibits NF-kB activation, inflammatory gene expression and macrophage/microglial activation in the rat brain (Heneka et al. 2003). Finally, HSP is a molecular chaperon capable of binding to malfolded protein to prevent protein aggregation and hence cell death (Hendrick and Hartl 1993). Thus, VPA-induced HSP70 over-expression may have a major role in the neuroprotective effects observed in this study, and could be a convergent mechanism with lithium-induced neuroprotection in the same experimental paradigm.
The observed increase in acetylated histone levels in the cortex and striatum of ischemic and normal rats treated in VPA (Figs 3 and 4) suggests that HDAC is indeed inhibited by this drug, confirming the in vitro study that VPA is an inhibitor of HDAC (Göttlicher et al. 2001; Phiel et al. 2001). The delayed loss in acetylated histone levels in vehicle-treated, MCAO/reperfusion rats could be due to stress-induced activation of HDAC or inhibition of histone acetyltransferase. Hyperacetylation of histones at lysine residues weakens the interaction of histone with DNA, leading to relaxation of the nucleosome structure and facilitation of transcription factor binding to DNA elements (Berger 1999). Interestingly, HDAC inhibitors, trichostatin A and sodium butyrate, were reported to increase HSP70 gene transcription in Drosophila melanogaster (Chen et al. 2002). It has also been shown that VPA and trichostatin A induce slow activation of the cell survival factor Akt and inactivation of glycogen synthase kinase-3β (GSK-3β) (De Sarno et al. 2002). Moreover, GSK-3β inactivation, which is cytoprotective, has been reported to activate heat shock factor, the transcription factor of HSP70 (Bijur and Jope 2000). Taken together, it seems possible that HSP70 induction by VPA and other HDAC inhibitors is related to GSK-3 inhibition. A recent study shows that HDAC inhibitors enhance the acetylation of SP1, a cytoprotective transcription factor, resulting in induction of anti-oxidant enzymes and cytoprotection in cortical neurons (Ryu et al. 2003). These results suggest a potential mechanism whereby SP-1 acetylation mediates VPA-induced HSP70 expression and neuroprotection in MCAO/reperfusion rats. Other anti-epileptic drugs such as lamotrigine, phenytoin, tiagabine and topiramate are also reported to have neuroprotective effects in rat and gerbil models of cerebral ischemia (for review, Calabresi et al. 2003; Leker and Neufeld 2003). The roles of HDAC inhibition and HSP induction in the neuroprotection elicited by these anti-convulsants require further investigation.
In addition to rapid induction of HSP70 shown in this report, chronic treatment with VPA has also been reported to induce other neuroprotective proteins. For example, VPA enhances the expression of the anti-apoptotic protein Bcl-2 (Chen et al. 1999), glucose-regulated protein 78 (Bown et al. 2000) and brain-derived neurotrophic factor (Fukumoto et al. 2001) in rats following weeks of treatment. It is unclear as to whether the expression of these proteins is regulated by HDAC inhibition and if they play a role in mediating protection against ischemia-induced brain damage. Future knockdown studies are necessary to elucidate the role of these proteins in mediating neuroprotection in our ischemic paradigm. VPA has a relatively broader therapeutic dose window and has become increasingly popular in the treatment of bipolar mood disorder (Post et al. 2002). Considering that VPA is safe, well tolerated and has a wide spectrum of neuroprotective actions, it seems promising to further investigate its beneficial effects in the animal stroke model and to perform clinical trials using this drug to treat stroke patients.