Conventional protein kinase Cβ-mediated phosphorylation inhibits collapsin response-mediated protein 2 proteolysis and alleviates ischemic injury in cultured cortical neurons and ischemic stroke-induced mice

Primary cortical neuron culture and oxygen–glucose deprivation treatment

The postnatal day 0 BABL/c mice were killed within 24 h. The brains were removed and placed in ice-cold 10 mM phosphate-buffered saline (PBS). The cortices were then dissected out, cut into small pieces, and transferred into 0.25% (w/v) Trypsin–EDTA (Gibco Inc., Grand island, NY, USA) for 20 min at 37°C. High glucose Dulbecco's modified Eagle's medium (5 mL; Gibco Inc., Rockville, MD, USA) containing 10% (v/v) fetal bovine serum, 10% (v/v) equine serum, and 2 mM l-glutamine was added to terminate the action of trypsin. Mild trituration using a Pasteur pipette was used to create a cell suspension. Cells were plated onto poly-l-lysine-coated 60-mm culture dishes (2 × 10⁶ cells per dish) for Western blot assay, 24-well plates with cover slips (2 × 10⁵ cells per well) for immunofluorescent staining, or 96-well plates (1.5 × 10⁴ cells per well) for Thiazolyl blue tetrazolium bromide (MTT) assay. The medium was replaced by serum-free neurobasal medium containing 2% B27 supplement (Gibco Inc.) 4 h later.

After culturing in neurobasal medium containing 2% B27 for 7 days, the cortical neurons were subjected to oxygen–glucose deprivation (OGD) treatment. The hypoxia incubator (Thermo Scientific, Marietta, OH, USA) was used in the OGD experiment. The incubator is capable to maintain O₂ at a certain level through persistently pumping N₂ into the incubator. The neurons were replaced with glucose-free Dulbecco's modified Eagle's medium and placed in the incubator. The time point when the atmosphere in the incubator was stabilized (5% CO₂/1% O₂/94% N₂) was recorded as the starting time. At the end of 1 h OGD exposure, cells were maintained in regular neurobasal medium containing 2% B27 under normoxic conditions for 24 h reoxygenation.

cPKCβ and CRMP2 modification in OGD-injured primary cultured cortical neurons

cPKCβ was activated by the addition of 0.1 μM 12-deoxyphorbol 13-phenylacetate 20-acetate (DOPPA; Enzo Life Sciences Inc., Farmingdale, NY, USA) (Macfarlane and Manzel 1994). Calpeptin (0.2, 0.5, 1.0, and 2.0 μM; Enzo Life Sciences Inc.) was used to inhibit calpain-mediated cleavage of CRMP2. CRMP2 cleavage was also blocked by the addition of a fusing peptide containing TAT peptide and the calpain cleavage site of CRMP2 (TAT-CRMP2) (0.1, 0.25, and 0.5 μM, sequence: YGRKKRRQRRRGVPRGLYDGPVCEV, California Bioscience Inc., Sacramento, CA, USA), a fusion peptide containing TAT peptide (transduction domain of HIV-1 trans-activator transcription protein) and the calpain cleavage site of CRMP2 (Xu et al. 2007). All the chemicals above were added in the medium 1 h before and during the induction of OGD.

Animal models and treatments

The repetitive autohypoxia-induced hypoxic preconditioning (HPC) mouse model was performed as our previous reports (Lu and Liu 2001; Bu et al. 2011; Zhang et al. 2011). Briefly, mice were placed in a 125-mL jar full of normoxic air and then the jar was sealed with a rubber plug so that continued breathing induced progressive hypoxia. The autohypoxia-treated mice were removed from the jar to normoxic conditions when the first signs of gasping appeared and recovered for 30 min. After the recovery period, mice were switched to another hermetically sealed jar. This procedure was repeated four times to obtain the HPC mice; normoxic control mice were placed in open jars for the same amount of times. The oxygen concentration inside the jar at the moment of the first gasping appearance was about 6.6% in the first run and 3.8% in the fourth run, respectively.

Two in vivo models of ischemia were used. The first was induction of 1 h transient MCAO with sacrifice at 24 h. The second was permanent MCAO with sacrifice at 6 h. Briefly, using methods previously described (Bu et al. 2011; Zhang et al. 2011; Feng et al. 2013), mice were anesthetized with sodium pentobarbital (0.06 g/kg i.p.). A ventral midline incision was then made to expose the left common carotid artery, and external and internal carotid arteries. The left common carotid artery and external carotid artery were ligated. For transient MCAO, an arteriotomy was made in the left external carotid artery, and then a blunt-tipped 5-0 surgical nylon filament (0.22 ± 0.01 mm in diameter; Beijing Cinontech, Beijing, China) was inserted through the external carotid artery to the internal carotid to a point approximately 12-mm distal to the carotid bifurcation to occlude the middle cerebral artery. After 1 h occlusion, the filament was carefully withdrawn to restore blood flow and the mice were kept for 24 h. For permanent MCAO, the same filament was inserted via the common carotid artery. Mice were kept for 6 h before being killed. The ischemic penumbra or peri-infarct region exits more apparently, and the effects of cPKCβ on CRMP2 phosphorylation and proteolysis are obvious at this time point. In the sham groups, mice received the same surgical exposure without occlusion. During the surgery, body temperature was maintained at 37°C using a heating lamp and thermal blanket. Mice were placed in a postoperative cage, kept warm, and undisturbed for at least 2 h with regular observation.

cPKCβ and CRMP2 modification in MCAO mice by intracerebroventricularly injection

Dimethylamine analog LY333531 (200 nM, 5 μL, cPKCβ-specific inhibitor) (Jirousek et al. 1996), DOPPA (2 μM, 5 μL, cPKCβ-selective activator), calpeptin (20 μM, 5 μL, calpain inhibitor), and TAT-CRMP2 (10 μM, 5 μL), which can prevent the cleavage of endogenous CRMP2, were intracerebroventricularly injected just before MCAO as in our previous reports (Bu et al. 2011; Zhang et al. 2011). Briefly, mice were anesthetized with pentobarbital sodium (0.06 g/kg i.p.) and fixed in a stereotaxic frame. The cannula (28-gauge, inner diameter 0.18 mm, outer diameter 0.36 mm) tip was inserted into the left cerebral ventricle at a point 0.5 mm posterior to bregma, 1.0 mm lateral to bregma, and 3.5 mm below the skull surface. The dosages of LY333531, DOPPA, calpeptin, and TAT-CRMP2 were determined by our preliminary experiments.

Measurement of cerebral blood flow

Regional cerebral blood flow (CBF) was monitored by laser Doppler flowmetry (Perimed Periflux system 5000, Järfälla-Stockholm, Sweden). The tip of the probe was fixed to the surface of skull over the MCA supplied core area (2 mm posterior to bregma and 6 mm lateral to midline). Cerebral blood flow during MCAO and reperfusion was expressed as a percentage of the baseline value.

Neurological deficit scoring and behavioral test

Neurological deficit scores were evaluated as reported previously (Longa et al. 1989) by an observer blinded to the treatment: 0, no deficit; 1, flexion of contralateral forelimb upon lifting of the whole animal by the tail; 2, circling to the contralateral side; 3, falling to the contralateral side; and 4, no spontaneous motor activity.

Behavioral tests were performed by an observer blinded to the treatment. The wire hanging test and pole test were performed as described previously (Ji et al. 2009). The wire hanging test was used to assess the motor deficits of mice. Mice were trained to cling with their forepaws to a steel wire (1 mm) stretched between two posts 60 cm above a thick protective pad 2 days before the actual test. On the day of testing, latency to fall was tested three times and results were averaged. The pole test was carried out using a vertical steel pole covered with tape to create a rough surface. Mice were placed near the top of the pole with head upward. The time taken to turn completely downward and the total time to reach the floor with all four paws were recorded. Each animal was tested three times and the average time was taken as the results.

Immunofluorescent staining of CRMP2 and cell viability assay in cultured cortical neurons

After the treatment of OGD, cells grown on cover slips were fixed in 4% paraformaldehyde and goat serum in PBST was used for 1 h as a block to prevent non-specific antibody binding. Cells were then incubated with rabbit anti-CRMP2 (1 : 500, #9393; Cell Signaling Technology, Beverly, MA, USA) overnight. After washing, the secondary Alexa Fluor^® 488-conjugated goat anti-rabbit IgG (1 : 500, green fluorescent; Invitrogen, Carlsbad, CA, USA) was added and incubated for 2 h. Finally, 4′,6-diamidino-2-phenylindole was used to stain the nuclei. Cover slips were dried at 4°C overnight. Slides were mounted and observed under a fluorescence microscope system (Leica DM4000B, Wetzlar, Germany).

According to the manufacturer's instruction, the cell viability of cultured cortical neurons was assessed by Thiazolyl blue tetrazolium bromide (MTT, 0.5 mg/mL; Applichem Inc., Omaha, NE, USA). The OD₄₉₀ for the control group was set to 100%, and the other groups were presented as a percentage of control.

Measurements of infarct size by TTC staining

After permanent and transient MCAO, blind measurements of the infarct volume were taken. Mice were killed, and the brains were quickly removed and cut into 1.5-mm thickness coronal sections. Brain sections were incubated in 0.5% 2,3,5-triphenyltetrazolium chloride (TTC) solution at 37°C for 20 min, and then the images of the slices were scanned into a computer. The infarct volume was analyzed according to the evaluation procedure of our previous reports (Bu et al. 2011; Zhang et al. 2011), and expressed as the percentage of total brain through the equation: I = [∑VI × (1 − E)/∑VT × (1 − B)] × 100%. ∑VI is the volume of tissue that is not stained with TTC, ∑VT is the total brain volume, E is edema ratio that was calculated from (∑VL − ∑VR)/(∑VL + ∑VR) × 100%, ∑VL and ∑VR are the volume of left ischemic and right non-ischemic hemisphere, B is background (∑VS/∑VT) × 100%, ∑VS is the volume of the unstained white matter in the sham group, and ∑VT is the total brain volume.

Nissl staining

Nissl staining was used to assess the neural cell damage in brain sections. Mice for this measurement were deeply anesthetized and transcardially perfused with 100 mM PBS containing 4% paraformaldehyde. Brains were removed and postfixed in 4% paraformaldehyde followed by graded sucrose solutions (20% for 24 h and 30% for another 24 h) to dehydrate. Then brains were cut into 20-μm thick slices, washed in fresh PBS, stained with 0.04% cresyl violet (Sigma-Aldrich), and dissolved in acetate buffer for 1 h. Images were photographed under a Nikon 50i light microscopy (Nikon, Tokyo, Japan).

Sample preparation and Western blot analysis

At the end of 1 h OGD followed by 24 h reoxygenation, cells in 60-mm dishes were washed three times with PBS and then stored at −70°C. For the required regions of cortexes, the brains from killed mice were immediately placed into ice-cold artificial cerebral spinal fluid (in mM: NaCl 125.0, KCl 2.5, CaCl₂ 2.0, NaHCO₃ 26.0, NaH₂PO4 1.25, MgCl₂ 1.0, glucose 5.0, pH 7.4) bubbled with 95% O₂/5% CO₂ and dissected as previous reports (Ashwal et al. 1998; Niu et al. 2005). In brief, 2 mm from the anterior tip of the frontal lobe was removed and the remaining brain was sectioned to four 2-mm thickness slices. Each hemisphere was longitudinally cut 1 mm from the midline to remove the tissue supplied by the anterior cerebral artery. A 45-degree angle transverse diagonal cut was then made to separate the ischemic core and peri-infarct region. The corresponding regions from the non-ischemic hemisphere were dissected as contralateral controls. All the tissues were frozen in liquid nitrogen and kept frozen at −70°C for later analysis.

To obtain whole cell lysates, the cultured cortical neurons and tissue samples were rapidly thawed and homogenized at 25°C in buffer C [5 mM Tris-Cl, pH 7.5, containing 2 mM dithiothreitol, 2 mM ethylenediaminetetraacetic acid (EDTA), 1 mM ethyleneglycoltetraacetic acid (EGTA), 5 μg/mL each of leupeptin, aprotinin, pepstatin A, and chymostatin, 50 mM potassium fluoride, 50 μM okadaic acid, 5 mM sodium pyrophosphate, 2% sodium dodecyl sulfate] and sonicated. Protein concentration was determined by bicinchoninic acid kit (Pierce Company, Rockford, IL, USA) with albumin diluted in buffer C as standard.

Samples loaded with the equal amount of proteins (50 μg) were electrophoresed on 10% sodium dodecyl sulfate polyacrylamide gel, and then transferred onto polyvinylidene difluoride (PVDF) membrane (GE Healthcare, Buckinghamshire, UK) at 4°C. After several rinses with Tween/Tris-buffered salt solution (20 mM Tris-Cl, pH 7.5, 0.15 M NaCl, and 0.05% Tween-20), the transferred PVDF membrane was blocked with 10% non-fat milk in Tween/Tris-buffered salt solution for 1 h. The blocked PVDF membrane was then incubated with the primary polyclonal antibodies of rabbit anti-CRMP2 (1 : 1000, #9393; Cell Signaling Technology), rabbit anti-phospho-CRMP2 (T514, 1 : 2000, ab62478; Abcam Inc., Cambridge, MA, USA), or rabbit anti-caspase-3 (1 : 1000, #9662; Cell Signaling Technology) for 3 h at 25°C. To verify equal loading of protein, the blots were reprobed with primary monoclonal antibody of mouse anti-β-actin (1 : 10 000, 60008-1-Ig; Proteintech Group, Chicago, IL, USA) or mouse anti-β-tubulin polyclonal antibody (1 : 10 000, 10094-1-AP; Proteintech Group). The horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse (1 : 5000; Stressgen Biotechnologies Corporation, Victoria, BC, Canada) were incubated as second antibodies for 1 h. After incubation with the secondary antibodies, the enhanced chemiluminescence kit (GE Healthcare) was used to detect the signals.

Statistics and analysis

The Gel Doc-2000 imaging system (Bio-Rad Laboratories, Hercules, CA, USA) was used to perform the quantitative analysis of Western blot results. For CRMP2 protein level, the ratio of intact CRMP2 densities to β-actin band density in the control group was expressed as 1.0; for the changes in CRMP2 phosphorylation, the ratio of p-CRMP2 band density to total CRMP2 band densities was expressed as 1.0 in control group. The ratio of cleaved caspase-3 band density to procaspase-3 and cleaved caspase-3 band densities was normalized to 1.0 in control. The data from other groups were represented as ratio of that from control group. The presented values were expressed as mean ± SE. Statistical analysis was conducted by one-way anova followed by all pairwise multiple comparison procedures using Bonferroni test, with significance cut-off set as p < 0.05.

cPKCβ-mediated phosphorylation inhibits CRMP2 proteolysis and protects cultured cortical neuron against OGD-induced ischemic injury

CRMP2 is abundantly expressed in neuron cytosol and has been proved to produce a 55-kD BDP of CRMP2 in cultured neurons under simulated ischemic conditions (Bretin et al. 2006; Zhang et al. 2007). In this study, cortical neurons from postnatal mice were cultured for 7 days, and then subjected to 1 h OGD followed by 24 h reoxygenation in regular medium. We found that the intact CRMP2 (I-CRMP2, Fig. 1a and b, one-way anova followed by Bonferroni test, p < 0.001, n = 6 per group) protein levels reduced significantly, while BDP appeared and the CRMP2 phosphorylation level decreased (p-CRMP2, Fig. 1a and b, p < 0.05, n = 6 per group) in 1 h OGD/24 h reoxygenation-treated cortical neurons. However, the cPKCβ-selective activator DOPPA (0.1 μM) promoted higher p-CRMP2 and intact CRMP2 levels, and inhibited the production of BDP when compared with that of the 1 h OGD/24 h reoxygenation group (Fig. 1a and b, p < 0.05, n = 6 per group). In addition, the MTT assay results in Fig. 1(c) showed that DOPPA could significantly increase the cell survival rate of 1 h OGD/24 h reoxygenation-treated cortical neurons (Fig. 1c, p < 0.05, n = 8 per group). These results suggested that cPKCβ-mediated CRMP2 phosphorylation protects cortical neurons against 1 h OGD/24 h reoxygenation-induced ischemic injuries through inhibiting CRMP2 proteolysis.

Inhibition of calpain-mediated CRMP2 proteolysis protects cultured cortical neurons against OGD-induced ischemic injury and cell apoptosis

To determine the role of CRMP2 proteolysis in 1 h OGD/24 h reoxygenation-induced ischemic injuries, the calpain inhibitor calpeptin and TAT-CRMP2 were used to block the endogenous CRMP2 proteolysis.

As shown in Fig. 2(a) and (b), calpeptin significantly reduced the endogenous CRMP2 proteolysis at 0.2, 0.5, 1.0, and 2.0 μM (p < 0.05, n = 5 per group) in 1 h OGD/24 h reoxygenation-treated cortical neurons. Similarly, TAT-CRMP2 (0.1, 0.25, and 0.5 μM) also significantly inhibits the production of BDP in 1 h OGD/24 h reoxygenation-treated cortical neurons (Fig. 2c and d, p < 0.05, n = 5 per group). The MTT assay results showed that both calpeptin and TAT-CRMP2 effectively reduced the 1 h OGD/24 h reoxygenation-induced cell death in cultured neurons in a dose-dependent manner (Fig. 2e, p < 0.05, n = 8 per group). In addition, we found that 1 h OGD/24 h reoxygenation induced a significant increase in cleaved caspase-3 (17 kD), while TAT-CRMP2 dose dependently inhibited the cleavage of procaspase-3 (34 kD, Fig. 2f and g, p < 0.05, n = 5 per group). These results suggested that calpeptin and TAT-CRMP2 protect cortical neurons against 1 h OGD/24 h reoxygenation-induced ischemic injury and cell apoptosis through inhibiting CRMP2 proteolysis.

BDP nuclear translocalization is involved in OGD-induced ischemic injury in cultured cortical neuron

To determine whether changes of CRMP2 subcellular distribution occurred after OGD, immunofluorescent staining was applied to observe CRMP2 expression in cortical neurons following 1 h OGD/24 h reoxygenation treatment. As shown in Fig. 3(a–c), CRMP2 was mainly expressed in the cytoplasm of control cortical neurons, while the 1 h OGD/24 h reoxygenation treatment caused an increase of CRMP2 nuclear translocation (Fig. 3d–f). This OGD-induced nuclear localization of CRMP2 could be prevented by pretreatment of cPKCβ-selective activator DOPPA (Fig. 3g–i), calpeptin (Fig. 3j–l), and TAT-CRMP2 (Fig. 3m–o).

cPKCβ-mediated phosphorylation inhibits CRMP2 proteolysis and protects mouse brain after ischemic stroke

cPKCβ-selective activator DOPPA was administered in mouse brain to further determine whether cPKCβ modulates phosphorylation and proteolysis of CRMP2 in vivo. As shown in Fig. 4(a) and (b), there was a significant decrease in CRMP2 phosphorylation and an increase in CRMP2 proteolysis, but DOPPA inhibited the decrease in p-CRMP2 level and the increase in BDP level both in ischemic core and peri-infarct region of mice with 6 h permanent MCAO (p < 0.05, n = 5 per group). Consistent with our previous report (Bu et al. 2011), HPC-induced cPKCβ activation played a similar role to the DOPPA in CRMP2 phosphorylation and proteolysis in ischemic core and peri-infarct region, but intracerebroventricular injection of cPKCβ-specific inhibitor LY333531 could inhibit this effect of HPC on CRMP2 phosphorylation and proteolysis (Fig. 4c and d, p < 0.05, n = 5 per group). In addition, the TTC staining results showed that DOPPA could reduce the neurological deficit scores (Fig. 4e, p < 0.05, n = 8 per group) and the infarct volume (Fig. 4f and g, p < 0.05, n = 5 per group) of mice with 1 h MCAO/24 h reperfusion, which indicates that cPKCβ-mediated CRMP2 phosphorylation can protect the brain against ischemic injury in mice with ischemic stroke through inhibiting CRMP2 proteolysis in vivo.

Inhibition of calpain-mediated CRMP2 proteolysis protects brain against ischemic injury of mice with ischemic stroke

To further determine the neuroprotective effect of CRMP2 proteolysis inhibition in vivo, calpeptin and TAT-CRMP2 were intracerebroventricularly injected to inhibit CRMP2 proteolysis in mouse brain with transient MCAO. As shown in Fig. 5(a) and (b), calpeptin effectively suppressed the production of BDP both in ischemic core and in the peri-infarct region (p < 0.05, n = 4 per group) and reduced the neurological deficit scores (Fig. 6a, p < 0.05, n = 9 per group) and infarct volume (Fig. 6b and c, p < 0.05, n = 5 per group) of mice with 1 h MCAO/24 h reperfusion.

We previously reported that TAT-CRMP2 could inhibit CRMP2 dephosphorylation and proteolysis in ischemic core of mice with 6 h permanent MCAO (Bu et al. 2011). Here, we applied transient MCAO mouse model to determine the neuroprotective effect of TAT-CRMP2 on the brain of mice with ischemic stroke. Cerebral blood flow was monitored during MCAO and 20 min of reperfusion by laser Doppler flowmetry. After the onset of MCAO, CBF sharply reduced to about 20% of baseline value and this was maintained during the 1 h of MCAO. The decrease of CBF in TAT-CRMP2 administrated mice was not significantly different from control ischemic mice (Fig. 5c, n = 4 per group). Intracerebroventricular injection of TAT-CRMP2 could inhibit CRMP2 proteolysis in ischemic core and peri-infact region of mice with 1 h transient MCAO followed by 24 h reperfusion (Fig. 5d and e, p < 0.05, n = 6 per group). Mice treated with TAT-CRMP2 achieved lower scores than the ischemia group (Fig. 6a, p < 0.05, n = 9 per group). The wire hanging test showed that ischemic mice clung to the wire for less time than the sham animals and TAT-CRMP2 treatment made mice cling to the wire for longer (Fig. 6d, p < 0.05, n = 7 per group). The pole test showed that animals from the ischemic group took more time to turn their head downward on the pole (time turn) and reach the ground (time total) compared to the sham-operated groups. TAT-CRMP2 reduced the time for mice to turn their head and reach the ground (Fig. 6e and f, p < 0.05, n = 7 per group). The TTC and Nissl staining results demonstrated that TAT-CRMP2 effectively reduce both the infarct volume (Fig. 6b and c, p < 0.05, n = 5 per group) and proportion of shrunken cells with pyknotic nuclei in peri-infact region (Fig. 6i and j) of brain with 1 h MCAO/24 h reperfusion. In addition, we observed that TAT-CRMP2 reduced the cleavage of caspase-3 in the peri-infarct region of ischemic mice (Fig. 6g and h, p < 0.05, n = 5 per group). These results suggested that calpeptin and TAT-CRMP2 protect brains against ischemic injury of mice with ischemic stroke through inhibiting calpain-mediated CRMP2 proteolysis in vivo.