GSK-3β downregulates Nrf2 in cultured cortical neurons and in a rat model of cerebral ischemia-reperfusion

The NF-E2-related factor 2 (Nrf2)/antioxidant response element (ARE) pathway plays a critical role in protecting against oxidative stress in brain ischemia and reperfusion injury. Glycogen synthase kinase 3β (GSK-3β) may play a critical role in regulating Nrf2 in a Kelch-like ECH-associated protein 1 (Keap1)-independent manner. However, the relationship between GSK-3β and Nrf2 in brain ischemia and reperfusion injury is not clear. In this study, we explored the mechanisms through which GSK-3β regulates Nrf2 and Nrf-2/ARE pathways in vitro and in vivo. We used oxygen and glucose deprivation/reoxygenation (OGD/R) in primary cultured cortical neurons and a middle cerebral artery occlusion-reperfusion (MCAO/R) rat model to mimic ischemic insult. In this study, GSK-3β siRNA and inhibitors (SB216763 and LiCl) were used to inhibit GSK-3β in vitro and in vivo. After inhibiting GSK-3β, expression of total and nuclear Nrf2, Nrf2-ARE binding activity, and expression of Nrf2/ARE pathway-driven genes HO-1 and NQO-1 increased. Overexpression of GSK-3β yielded opposite results. These results suggest that GSK-3β downregulates Nrf2 and the Nrf2/ARE pathway in brain ischemia and reperfusion injury. GSK-3β may be an endogenous antioxidant relevant protein, and may represent a new therapeutic target in treatment of ischemia and reperfusion injury.

Scientific RepoRts | 6:20196 | DOI: 10.1038/srep20196 tyrosine 216. Regulation of GSK-3β usually depends on phosphorylation within the amino-terminal domain of GSK-3β (Ser 9) and results in inactivation of GSK-3β by several kinases including Akt, PKA, and PKC 11 . Recent studies have shown that inhibiting GSK-3β can reduce Nrf2 by nuclear export and degradation of Nrf2 in liver cancer cells and improve the rate of cell survival during the late phase of oxidative stress 9,10 . However, there have been no studies showing how GSK-3β regulates Nrf2 in brain ischemia and reperfusion injury.
In this study, oxygen-glucose deprivation followed by recovery (OGD/R) and middle cerebral artery occlusion-reperfusion (MCAO/R) were used to mimic ischemic reperfusion insult in vitro and in vivo, respectively. GSK-3β siRNA and inhibitors (SB216763 and LiCl) were used to inhibit GSK-3β . We found that GSK-3β downregulates expression levels of Nrf2, Nrf2-ARE binding activity, and expression levels of genes downstream of Nrf2/ARE in brain ischemic and reperfusion injury.
In parallel experiments, we found that Nrf2 expression presented a reverse trend (Fig. 1A). After 0.5 h of reoxygenation, the expression level of Nrf2 was elevated approximately 3-fold. After 1 h, 4 h, and 6 h of reoxygenation, expression of Nrf2 decreased to the normal level. The expression of p-GSK-3β (tyr216) initially reached its highest level after 1 h of reoxygenation. Therefore, to analyze GSK-3β regulation of Nrf2 in brain ischemic and reperfusion injury, we chose 1 h of reoxygenation as the optimum time for this study.
GSK-3β regulates Nrf2 in cultured neurons after OGD/R. Prior to isolating total protein, nuclear protein, and RNA from neurons at 6 d, GSK-3β inhibitors (SB 216763 and LiCl) were continuously applied from 6 h, and GSK-3β siRNA and the overexpression lentivirus (GSK-3β ) were continuously applied from 72 h. Under normal conditions, treatment with GSK-3β siRNA or GSK-3β inhibitors remarkably decreased expression of both GSK-3β and p-GSK-3β (tyr216) compare with the normal group ( Fig. 2A). No statistical difference in total and nuclear Nrf2 expression was observed in the GSK-3β siRNA and inhibitor groups (Fig. 2D) compared with the normal group. This suggests that GSK-3β does not regulate Nrf2 under normal conditions. However, after OGD/R, in the GSK-3β siRNA + OGD/R group and GSK-3β inhibitors + OGD/R groups, expression of total and nuclear Nrf2 significantly increased compared with the OGD/R group (Fig. 3D). The GSK-3β + OGD/R group showed opposite results. The results from real-time fluorescence quantitative (Q-PCR) were consistent with those from western blots (Fig. 4). These results suggest that, after OGD/R, GSK-3β may impose negative regulation on Nrf2 in neurons.
GSK-3β regulates Nrf2-ARE binding activity in neurons after OGD/R. Neuronal nuclear extracts were subject to EMSA for measurement of Nrf2-ARE binding activity. There was no statistically significant difference between normal, OGD/R, and control siRNA (con siRNA) + OGD/R groups. Treatment with GSK-3β siRNA + OGD/R and inhibitors + OGD/R resulted in a higher Nrf2 binding activity compared with the OGD/R group (Fig. 5). Opposite results were obtained with the GSK-3β + OGD/R group. These results suggest that after OGD/R, GSK-3β may exert negative control on Nrf2-ARE binding.
GSK-3β regulates Nrf2/ARE-driven gene expression in neurons after OGD/R. Western blot analysis and Q-PCR were used to investigate the effects of GSK-3β on expression of the Nrf2/ARE-driven genes, HO-1 and NQO1 (Fig. 6). In the GSK-3β siRNA + OGD/R group, expression levels of HO-1 and NQO1 increased by about 1.8-fold and 2.2-fold, respectively, compared with the OGD/R group. In the GSK-3β inhibitors + OGD/R groups, HO-1 and NQO1 expression levels were elevated by about 2-fold and 1.83-fold, respectively, compared with the OGD/R group (Fig. 6A). In the GSK-3β + OGD/R group, HO-1 and NQO1 expression levels decreased by about 2.8-fold and 2.2-fold, respectively. Results from Q-PCR were consistent with those from western blot analysis (Fig. 6D,E). These results suggest that after OGD/R, GSK-3β downregulates expression of Nrf2/ ARE-driven genes, including HO-1 and NQO1.
Scientific RepoRts | 6:20196 | DOI: 10.1038/srep20196 GSK-3β regulates Nrf2 in the cerebral cortex of rats after MCAO/R. Under normal conditions, treatment with GSK-3β siRNA or GSK-3β inhibitors significantly decreased both GSK-3β and p-GSK-3β (tyr216) levels compared with the normal group (Fig. 8A). Total and nuclear Nrf2 expression showed almost no significant change (Fig. 8A). Thus, GSK-3β does not regulate Nrf2 under normal conditions. However, after 1 h of MCAO followed by 6 h of reperfusion, total Nrf2 expression significantly increased approximately 1.8-fold in the GSK-3β siRNA + MCAO/R and GSK-3β inhibitors + MCAO/R groups, compared with the MCAO/R group (Fig. 9A). In addition, nuclear Nrf2 expression significantly increased approximately 2-fold (Fig. 9A). The results from Q-PCR were consistent with those from western blot analysis (Fig. 10). These results suggest that GSK-3β does   GSK-3β regulates Nrf2-ARE binding activity in the cerebral cortex of rats after MCAO/R. Nuclear extracts from the cerebral cortex were subjected to EMSA for measurement of Nrf2-ARE binding. Inhibiting GSK-3β by transfecting with GSK-3β siRNA and treating with inhibitors significantly increased Nrf2-ARE binding activity after MCAO/R (Fig. 11). These results suggest that GSK-3β negativity regulates Nrf2-ARE binding in the cerebral cortex of rats after MCAO/R. This result is consistent with our in vitro experiments.
GSK-3β regulates expression of Nrf2/ARE-driven genes in the cerebral cortex of rats after MCAO/R. After 6 h of reperfusion, expression levels of the Nrf2/ARE-driven genes, HO-1 and NQO1, were analyzed by western blot and Q-PCR (Fig. 12). In the GSK-3β siRNA + MCAO/R group, expression levels of HO-1 and NQO1 significantly increased approximately 1.5-fold and 2-fold, respectively, compared with the MCAO/R group (Fig. 12A). In the GSK-3β inhibitors + MCAO/R groups, HO-1 expression levels significantly increased about 1.5-fold, and NQO1 expression levels significantly increased about 1.9-fold (Fig. 12A). The results from Q-PCR were consistent with those from western blot analysis (Fig. 12D,E). These results suggest that GSK-3β downregulates expression of Nrf2/ARE-driven genes, including HO-1 and NQO1 in the cerebral cortex of rats after MCAO/R. These results are consistent with our in vitro experiments.

Discussion
In the present study, we explored the relationship between GSK-3β and Nrf2 in neurons that were subjected to OGD/R and in the cerebral cortex of rats that sustained MCAO/R. We showed that the activity of GSK-3β in Similarly, the activity of GSK-3β in the cerebral cortex of rats decreased at 1 h of reperfusion and then increased at 6 h of reperfusion. Nrf2 expression showed an opposite trend in vitro and in vivo. We employed siRNA knockdown and inhibition of GSK-3β in vivo and in vitro, which increased expression of Nrf2, Nrf2-ARE binding, and expression of antioxidant proteins HO-1 and NQO1 that are downstream from ARE. Overexpression of GSK-3β in cerebral neurons of rats demonstrated a reverse tendency. Our results suggest that increasing the level of activated GSK-3β inhibits the Nrf2/ARE signaling pathway of ischemia-reperfusion in the cerebral cortex and OGD/R in neurons. Thus, GSK-3β is a negative regulatory factor for Nrf2 in cerebral ischemia-reperfusion and OGD/R in neurons.
Nrf2, a key transcription factor, is involved in expression of many cytoprotective genes. It is well known that Nrf2 is normally retained in the cytoplasm by Keap1 [13][14][15] . In the present study, we focused on GSK-3β , which is a multifunctional kinase. GSK-3β has drawn considerable attention in recent years. It is a key kinase that is involved in several cellular signaling pathways and sensitizes cells for cell death. Recent research indicates that GSK-3β , as a negative regulator of Nrf2, participates in the distribution of Nrf2 inside and outside of the nucleus 16,17 . GSK-3β regulation of Nrf2 transcription activity is independent of the expression of Keap1 9,18 . In addition, inhibition of GSK-3β before ischemia or just before reperfusion has been shown to reduce myocardial infarct size [19][20][21]35 .
Moreover, research has also demonstrated that inhibition of GSK-3β improves cognition during oxidative stress in a mouse model of Alzheimer's disease. This effect may coincide with reduced nuclear translocation of Nrf2 22-24 . In this study, we demonstrated that inhibiting GSK-3β could induce accumulation of Nrf2 in the nucleus. However, the effects of GSK-3β on Nrf2 and the Nrf2/ARE signaling pathway in cerebral ischemia-reperfusion had not been studied previously. Nrf2 translocates to the nucleus where it activates the ARE of phase II detoxifying enzymes and antioxidant stress genes such as NQO1, and accelerates their transcription and expression. Nrf2 accumulation in the nucleus can guard against oxidative stress. However, Nrf2 cannot always be located in the nucleus; Nrf2 will be exported out of the nucleus and degraded. Studies have shown that GSK-3β leads to Nrf2 nuclear export and degradation through 2 pathways. Suryakant (2011) reported that activated GSK-3β controls accumulation of Src kinases in the nucleus. Nuclear accumulation of Src kinases results in phosphorylation of Nrf2 (Tyr568), which leads to Nrf2 nuclear export and degradation in mouse hepatoma (Hepa-1) cells 25 . Partricia (2011) reported that activation of GSK-3 based on phosphorylation of the Neh6 domain of Nrf2 and ubiquitination degraded Nrf2 through the β -TrCP/Cullin1 E3 ligase complex in human embryonic kidney (HEK) 293 T cells 9 . OGD/R was used to mimic ischemic reperfusion insult in vitro. GSK-3β was activated by phosphorylation of Tyr-216 [p-GSK-3β (tyr216)]. The activity of GSK-3β was detected indirectly by examining β -catenin. We observed that activated GSK-3β decreased after 0.5 h of reperfusion, yet increased after 1 h of reperfusion. It is possible that the PI3K/Akt pathway is activated during the early phase of antioxidant stress, involving short-term activation of Akt and inhibition of GSK-3β by phosphorylation of GSK-3β (Ser 9) 26 . After long-term antioxidant stress, inhibition of GSK-3β by the PI3K/Akt pathway decreases, possibly due to an unknown tyrosine kinase that activates GSK-3β , as well as increased expression of GSK-3β 25 . Expression of Nrf2 changed in the opposite direction. However, Jaiswal (2006) reported that H 2 O 2 activated GSK-3β after 4 h in HepG2 cells 13 . It is possible that the differences in time course were due to differences in sensitivity to oxidative stress between cerebral neurons and HepG2 cells. GSK-3β interference or inhibition increased total Nrf2 protein, accumulation in the nucleus, as well as Nrf2 mRNA. These results agree with a study by Abhinav (2007), which showed that continuous activation of GSK-3β prevented accumulation of Nrf2 in the nucleus when HepG2 cells were transfected with GSK-3β siRNA 16 . Knockdown or inhibition of GSK-3β increased Nrf2-ARE binding as shown by EMSA. Furthermore, knockdown or inhibition of GSK-3β activated expression of genes downstream from ARE, including HO-1 and NQO1. Overexpression of GSK-3β yielded opposite results in neurons. These results are similar to those of Salazar et al. (2006), who demonstrated that transcription of phase II detoxifying enzymes and antioxidants significantly decreased when cotransfected with Nrf2 and GSK-3β in HEK 293 T cells 27,36 . In addition, we observed that GSK-3β was activated after 6 h of reperfusion in vivo. The effects of interfering or inhibiting GSK-3β before MCAO in rats showed changes similar to the study in vitro. The results described above were not observed under normal conditions both in vivo and in vitro. It is possible that under normal conditions, GSK-3β is inhibited by phosphorylation at GSK-3β Ser-9, and regulation and degradation of Nrf2 occurs primarily through Keap1/Cullin 3/Rbx1 complexes 9 .
In this study, we demonstrated negative regulation of the transcription factor Nrf2 by GSK-3β after oxidative stress induced by OGD/R in vitro and cerebral ischemia-reperfusion in vivo. Our research suggests a potential new direction for treatment of stroke. However, the specific mechanisms through which GSK-3β regulates Nrf2 have not been clarified in cerebral ischemia-reperfusion. Whether or not GSK-3β regulates Nrf2 through the Src subfamily of kinases and the β -TrCP pathway requires further study.

Methods and Materials
Experimental Animals and Chemicals. Adult male Sprague-Dawley rats (60-80 d old, 240-300 g) were used for the in vivo study. Newborn Sprague-Dawley rats (0-24 h old) were used to culture primary cortical neurons. The animal protocol was approved by the Chongqing Medical University Biomedical Ethics Committee. All experimental procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All efforts were made to minimize the number of animals used and their suffering.
SB216763 and LiCl were purchased from Sigma-Aldrich (St Louis, MO, USA). SB216763 was dissolved in dimethyl sulfoxide (DMSO) and diluted with saline. LiCl was dissolved in saline.

Primary Culture of Rat Cortical Neurons and OGD/R. Neurons were cultured as described in our
previous studies 28,29 . Cortical neurons were obtained from the cerebral cortex of 24-h-old rats. Approximately 2 × 10 6 cells in 2 mL of Neurobasal Medium containing glutamine (1 mM), 1% penicillin and streptomycin (Pen/ Strep) (penicillin 100 U/mL, streptomycin 100 μ g/mL), and 2% B27 supplement were seeded per well. Neurons were cultured in a humidified incubator with 5% CO 2 /balanced with air (result: 20% O 2 ) at 37 °C. The cells were cultured for 6-7 d in vitro. Cultured cells were examined using NeuN and GFAP staining to ensure that more than 90% of the cells were neurons. OGD/R was conducted as previously described 30,31 . Briefly, after neurons were cultured for 6 d, they were washed 3 times with glucose-free DMEM. The glucose-free DMEM had been previously equilibrated with 1% O 2 , 5% CO 2 , and 94% N 2 at 37 °C in an incubator. Neurobasal Medium was then replaced with glucose-free DMEM, and the cells were transferred to an incubator with 1% O 2 , 5% CO 2 , and 94% N 2 for 1.5 h at 37 °C. The medium was then changed back to Neurobasal Medium and the cultures were returned to the normal incubator for recovery times of 0.5 h, 1 h, 4 h, or 6 h. An appropriate time of reoxygenation was selected for subsequent studies.

MCAO and Design of In Vivo
Experiments. Rats were given free access to food and water in optimal surroundings before the operation. Adult rats were divided randomly into 11 groups. Under normal conditions, the groups were: normal group, scramble group, GSK-3β siRNA (siRNA) group, SB216763 (SB) group (20 μ g/kg, intracerebroventricular injection), and the LiCl group (50 mg/kg, intraperitoneal injection). After MCAO/R, the groups were: sham-operated (sham) group, MCAO/R group, scramble + MCAO/R group, GSK-3β siRNA (siRNA) + MCAO/R group, SB + MCAO/R group, and the LiCl MCAO/R group. Transient cerebral ischemia (MCAO) was described in detail in our previous study 33,34 . Rats were anesthetized with chloral hydrate (350 mg/kg, intraperitoneal injection) and subjected to the operation. A nylon filament (diameter 0.24-0.28 mm) was inserted into the middle cerebral artery for 1 h. The nylon filament was carefully removed to allow blood to return to the ischemic artery, and then was sutured to establish reperfusion. Regional cerebral blood flow was detected by an ultrasonic blood flow meter before ischemia, during MCAO, and during reperfusion. Sham-operated rats were subjected to the same surgical procedure as MCAO rats except for occlusion of the common carotid arteries. Animals that had blood reperfusion below 70% or that died during reperfusion were excluded from analysis. GSK-3β Interference in Rats. GSK-3β siRNA was constructed by Shanghai GenePharma Co., Ltd (forward 5ʹ-GGAGAGCCCAAUGUUUCAUTT-3ʹand reverse 5ʹ-AUGAAACAUUGGGCUCUCCTT-3ʹ). SiRNAs were dissolved in RNase-free water to a final concentration of 2 μ g/μ L. Forty-eight hours before MCAO, 7 μ L of GSK-3β siRNA was injected ipsilaterally into the left lateral cerebral ventricle. Transfection efficiency was confirmed under a fluorescence microscope. As a control, rats were injected with the scramble siRNA (forward 5ʹ-GCGCCAGUGGUACUUAAUATT-3ʹ and reverse 5ʹ-UAUUAAGUACCACUGGCGCTT-3ʹ) using the same procedure as for GSK-3β siRNA. Sustained GSK-3β downregulation was confirmed by qRT-PCR and western blot analysis 48 h after transfection.
qRT-PCR. Total  Electrophoretic Mobility Shift Assay (EMSA). EMSA was conducted using BiotinLight TM EMSA Kit (Exprogen, China). For EMSA, The Nrf2 consensus oligonucleotide probe (Bio-5ʹ -TGG GGA ACC TGT GCT GAG TCA CTG GAG-3ʹ ) was end-labeled with [c-32 P] ATP (Sangon Biotech) using T4-polynucleotide kinase. Five micrograms of neuronal total nuclear protein or 8 μ g of cerebral nuclear proteins were incubated in a binding buffer for 20 min at room temperature. Samples were then loaded on a non-denaturing 6.5% polyacrylamide gel and electrophoretically separated in 0.25X TBE buffer. The gel was vacuum-dried and exposed to X-ray film (Fuji Hyperfilm, Tokyo, Japan) at 80 °C with an intensifying screen. Levels of Nrf2 DNA binding activity were quantified by computer-assisted densitometric analysis.

Gene product
Forward primer Reverse primer Fragment size (bp)