P7C3 inhibits GSK3β activation to protect dopaminergic neurons against neurotoxin-induced cell death in vitro and in vivo

Parkinson’s disease (PD) is the second most prevalent neurodegenerative disease. Although its pathogenesis remains unclear, mitochondrial dysfunction plays a vital role in the pathology of PD. P7C3, an aminopropyl carbazole, possesses a significant neuroprotective ability in several neurodegenerative disorders, including PD. Here, we showed that P7C3 stabilized mitochondrial membrane potential, reduced reactive oxygen species production, and inhibited cytochrome c release in MES23.5 cells (a dopaminergic (DA) cell line) exposed to 1-methyl-4-phenylpyridinium (MPP+). In MES23.5 cells, P7C3 inhibited glycogen synthase kinase-3 beta (GSK3β) activation induced by MPP+. P7C3 also inhibited p53 activity and repressed Bax upregulation to protect cells from MPP+ toxicity. In addition, the activation of p53 was significantly attenuated with the inhibition of GSK3β activity by P7C3. Furthermore, P7C3 blocked GSK3β and p53 activation in the midbrain, and prevented DA neuronal loss in the substantia nigra in 1-methyl-4-phenyl-1,2,3,4-tetrahydropyridine mice. Thus, our study demonstrates that P7C3 protects DA neurons from neurotoxin-induced cell death by repressing the GSK3β-p53-Bax pathway both in vitro and in vivo, thus providing a theoretical basis for P7C3 in the potential clinical treatment of PD.

P7C3 is a small aminopropyl carbazole compound that was first identified to have a remarkable neuroprotective ability for newborn hippocampal neurons from apoptosis. 1 Thereafter, P7C3 was demonstrated to possess neuroprotective efficacy in many diseases or brain injury. P7C3 and its variant P7C3A20 exhibit neuroprotective efficacy in traumatic brain injury 2,3 and stroke. 4,5 This compound can also induce postnatal hippocampal neurogenesis, showing antidepressant efficacy in mice 6,7 and improve the scopolamine-induced learning and memory impairment. 8 Importantly, P7C3 and its variants exhibit strong neuroprotective efficacy in many neurodegenerative disease models. In the G93A-SOD1 mutant mouse model of amyotrophic lateral sclerosis, the early administration of P7C3 and its variant P7C3A20 at or before disease onset effectively blocks motor neuronal death. 9 P7C3 also significantly relieves the tyrosine hydroxylase (TH) neuronal loss in the substantia nigra (SN) and striatum in 1-methyl-4-phenyl-1,2,3,4-tetrahydropyridine (MPTP) mouse or 6-hydroxydopamine (6-OHDA) rat model of Parkinson's disease (PD). 10,11 Recently, P7C3 and its variants were shown to bind to and enhance the activity of nicotinamide phosphoribosyltransferase, thereby increasing nicotinamide adenine dinucleotide levels to protect cells. 12 PD is one of the most common neurodegenerative disorders, following only Alzheimer's disease (AD). 13 The preferential loss of dopaminergic (DA) neurons in the SN leads to dopamine depletion in the striatum, resulting in symptoms of bradykinesia, resting tremor, muscle rigidity, and flexed posture. 14 Although the pathogenesis of PD remains obscure, evidence of the etiology of PD is increasing. For example, oxidative stress 15 and mitochondrial dysfunction 16 are considered significant elements of DA neuronal death in PD. Mitochondrial dysfunction is an important factor associated with PD pathogenesis. In the SN of PD brains, mitochondrial complex I activity is significantly decreased, 17 thus providing a link between mitochondrial dysfunction and PD.
The direct link between mitochondrial dysfunction and PD was demonstrated by evidence suggesting that MPTP induces PD symptoms in drug-abuse patients. 18 MPTP, a special neurotoxin that can selectively damage DA neurons, is widely applied to establish PD animal models. 19 When MPTP penetrates the blood-brain barrier, it can be metabolized into 1-methyl-4-phenylpyridinium (MPP + ) by monoamine oxidase in glial cells; MPP + is then taken up into DA neurons through the dopamine transporter. 20 MPP + inhibits mitochondrial complex I to interfere with the electron transport chain, resulting in ATP depletion and reactive oxygen species (ROS) production. 21 ROS induces mitochondrial damage, leading to the loss of membrane potential and release of cytochrome c (Cyt c) from mitochondria, which activates caspase-3 and induces apoptosis. 22 Although P7C3 exerts protective effects in animal models of PD, the mechanisms underlying its protection in PD remain unclear. As P7C3 has multiple protective effects in many different neuronal diseases, it is likely that this compound may function in different pathways in different diseases, including PD. In this study, we showed that P7C3 suppresses the activity of glycogen synthase kinase-3 beta (GSK3β) induced by MPP + and MPTP, leading to the inhibition of p53 activation and Bax expression, thus impeding mitochondrial damage and eventually protecting DA neurons from death in MPTP mouse models of PD.

Results
P7C3 attenuates MPP + -induced cytotoxicity in MES23.5 cells. MPP + damages DA neurons through the inhibition of complex I of the mitochondrial respiratory chain. 22 We therefore detected the protective effects of P7C3 on MPP +induced DA neuronal death in vitro. We observed that the pretreatment of P7C3 markedly attenuated the MPP +induced cytotoxicity to MES23.5 cells, as evidenced by the Hochest 33342 and propidium iodide (PI) staining (Figures 1a  and b). P7C3 strikingly decreased the number of the cells with condensed nuclei in response to treatment with MPP + (Figure 1a). Moreover, the amount of MES23.5 cell death induced by MPP + was also markedly reduced in the presence of P7C3 (Figure 1b).
P7C3 alleviates MPP + -induced oxidative stress and mitochondrial damage in MES23.5 cells. Since MPP + impairs DA neurons by inhibiting the mitochondrial complex I, resulting in the overproduction of ROS and dysfunction of mitochondria, including the loss of the mitochondrial membrane potential (MMP), 23 we examined whether P7C3 affects   (1, 5, or 10 μM) for 2 h and then exposed to MPP + (250 μM) for 24 h. After treatment, the cells were subjected to Hoechst or PI staining to detect cell death. Scale bar, 10 μm (b) MES23.5 cells were treated as in a and subjected to flow cytometry analysis ROS generation and stabilizes mitochondrial membrane to protect DA neurons against MPP + toxicity. In MES23.5 cells, MPP + largely increased ROS production, whereas pretreatment with P7C3 significantly decreased it (Figure 2a). Moreover, MPP + largely decreased MMP, as indicated by tetramethylrhodamine methyl ester (TMRM), a dye sequestered by active mitochondria (Figure 2b). However, pretreatment with P7C3 significantly restored MMP in a dosedependent manner (Figure 2b). P7C3, at a concentration of 10 μM, almost completely restored the MMP impaired by MPP + (Figure 2b). Moreover, P7C3 itself did not affect MMP ( Figure 2b). These data suggest that P7C3 stabilizes MMP and inhibits ROS production to protect MES23.5 cells exposed to MPP + .
P7C3 inhibits cytochrome c release and caspase-3 activation induced by MPP + . The mitochondria play a significant role in apoptosis. Mitochondrial damage results in the release of Cyt c, which triggers caspases activation, and eventually induces apoptosis. 24 To further confirm the effects of P7C3 on mitochondria, we performed mitochondrial fractionation assays to examine the release of Cyt c after MPP + treatment in the presence or absence of P7C3 in MES23.5 cells. P7C3 strikingly blocked MPP + -induced Cyt c release from mitochondria to the cytosol (Figures 3a-c). The inhibition of caspase-3 cleavage by P7C3 was observed in cells treated with MPP + (Figure 3d). The pro-apoptotic protein Bax, which is a Bcl-2 family protein, regulates and promotes permeability transition pore (PTP) opening. 25 Thus, we examined Bax protein levels in cells treated with MPP + in the presence of P7C3. Interestingly, P7C3 significantly decreased the Bax expression induced by MPP + (Figure 3d). These data indicate a role for P7C3 in regulating the mitochondrial apoptotic pathway in cells exposed to MPP + . influences p53. In MES23.5 cells treated with MPP + , pretreatment with P7C3 strikingly restrained p53 phosphorylation at serine 15 (Ser15) but not influenced its abundance ( Figure 4a). In addition, the nuclear fractionation experiments showed that P7C3 prevented p53 nuclear translocation in MPP + -treated MES23.5 cells (Figure 4b), indicating that P7C3 inhibited the increase of p53 activity induced by MPP + . It has been reported that mitochondrial oxidative stress activates GSK3β 27 and that an inhibition of GSK3β attenuates p53 activity. 28 We therefore examined the effects of P7C3 on GSK3β activity by evaluating its phosphorylation at sites Ser9 (a site inhibiting GSK3β enzymatic activity) 29 and Tyr216 (a site activating GSK3β enzymatic activity). 30 In MES23.5 cells that were treated with MPP + , the phosphorylation of Ser9 in GSK3β was decreased (Figure 4c). However, that of tyrosine 216 was increased (Figure 4c), suggesting that MPP + activates GSK3β. However, pretreatment with P7C3 significantly inhibited GSK3β activation mediated by MPP + (Figure 4c). Moreover, the level of β-catenin, an important substrate of GSK3β, 31 (Figure 4d), suggesting an inhibition of GSK3β activity by P7C3. To further identify the inhibiting effects of P7C3 on GSK3β activation induced by MPP + , the specific inhibitor of AKT MK-2206 was introduced, which leads to an inhibition of the phosphorylation of GSK3β at Ser9 by AKT. 32 MK-2206 treatment significantly repressed GSK3β phosphorylation at Ser9 and increased caspase-3 cleavage and Bax levels in MPP + -treated cells (Figure 4e). However, upon P7C3 treatment, the phosphorylation of GSK3β was partially restored in MPP + -treated cells that were treated with MK-2206 (Figure 4e).
P7C3 possesses cytoprotective effects analogous to the GSK3β inhibitor. To further identify that the inhibition of GSK3β by P7C3 plays a role in p53 activation, we compared the effects of P7C3 on p53 with those of the GSK3β inhibitor SB216763 in MPP + -treated MES23.5 cells. P7C3, similar to the GSK3β inhibitor SB216763, inhibited p53 phosphorylation ( Figure 5a) and GSK3β activation (Figure 5b). In addition, P7C3 and SB216763 suppressed cell death in MPP + -treated MES23.5 cells, detected using flow cytometry ( Figure 5c). Thus, these data suggest that the suppression of GSK3β activation by P7C3 inhibits p53 activation and protects cells against MPP + toxicity.

P7C3 inhibits GSK3β activity and protects MPTP-induced DA neuronal loss in vivo.
To further identify the role of P7C3, we utilized a MPTP mouse model to examine the effects of P7C3 on GSK3β and neuroprotection. The animals were subjected to continuous P7C3 intraperitoneal injection for 21 days, followed by the intraperitoneal injection of MPTP for 5 days (Figure 6a). Pretreatment with P7C3 significantly blocked MPTP-induced TH neuronal loss in the SN (Figure 6b). In addition, P7C3 inhibited MPTP-induced GSK3β and p53 activation in the midbrain (Figure 6c). Consistent with the immunohistochemical data (Figure 6b), P7C3 recovered the TH protein levels in MPTP mice (Figure 6c).

Discussion
It has been reported that P7C3 possesses the neuroprotective efficacy in neurodegenerative disease animal models. [9][10][11] In PD animal models, P7C3 protects DA neurons from MPTP and 6-OHDA toxicity. 10,11 However, it is still unknown in which pathway P7C3 exerts protective effects in PD models. Here, the present study showed that P7C3 functions in the GSK3β/ p53/Bax pathway to protect mitochondria from MPP + -or MPTP-induced damage, thereby improving the survival of DA neurons in vitro and in vivo (Figure 7). GSK3β is associated with various neurodegenerative diseases including AD and PD. 33 GSK3β activation is involved in both neurotoxin-induced and genetic factor-induced DA neuronal loss. 34 Accordingly, the active form of GSK3β (p-GSK3β Tyr216) is increased in the striatum of postmortem samples of PD. 35 In addition, MPTP animals show the hyperphosphorylation of Tyr216 and dephosphorylation of Ser9 in GSK3β. 36 The inhibition of GSK3β protects DA neurons from MPTP toxicity. 37 In our observations, P7C3 protects DA neurons from MPP + -and MPTP-induced cell death, showing a significant protective efficacy in vitro and in vivo. However, in the MPP + cellular model, P7C3 inhibited MPP + -induced GSK3β activation, indicated as repressed phosphorylation at Tyr216 and increased phosphorylation at Ser9. In MPTP mice, P7C3 markedly blocked the dephosphorylation of GSK3β at Ser9, without an obvious influence on Tyr216. The slightly different effects of P7C3 on MPTPmediated GSK3β activation between in vitro and in vivo suggest that some factors may be differentially involved in the phosphorylation of Tyr216 in GSK3β, and P7C3 has a preferential effect on Ser9 in modulating GSK3β activation in vivo.
Damage to mitochondria by MPP + or MPTP is caused by an induction of ROS production due to an inhibition of mitochondrial complex I, which results in mitochondrial oxidative stress. 21 It has been reported that mitochondrial oxidative stress activates GSK3β and Bax, leading to mitochondrial PTP opening. 27 The opening of PTP triggers a feed-forward loop involving a progressive Ca 2+ surge, which further induces ROS production and PTP opening. 38 The inhibition of GSK3β blocks oxidative stress-induced PTP opening. 27,39 Most interestingly, GSK3β activation occurs in a ROS-dependent manner, and in turn induces PTP opening. 27 However, GSK3β inhibition also prevents ROS production. 40 As GSK3β inhibitor SB216763 also decreases its activation as P7C3 in MPP +treated cells, it seems that ROS and GSK3β influence each other. In our observations, P7C3 inhibited MPP + -induced GSK3β activation and Bax upregulation to stabilize the mitochondrial membrane and block Cyt c release, suggesting a role for P7C3 in mitochondrial protection by the abrogation of PTP sensitization via the inhibition of GSK3β.
The opening of PTP results from the oligomerization of Bax or Bak on the mitochondrial membrane, leading to a loss of MMP and release of mitochondrial components. 41 The release of Cyt c activates caspases to induce cell death. 41 Bax is a proapoptotic protein that is transactivated by the transcription factor p53. 26 GSK3β acts upstream of caspases and Cyt c release. 42,43 Most importantly, GSK3β interacts with and promotes the transcriptional activity of p53. 28,42 The dominant-negative mutant of p53 blocks GSK3β-induced apoptosis. 44 suggesting that GSK3β influences p53, contributing to apoptosis. The phosphorylation of Ser15 plays a critical role in activating p53 upon stimulation. 45 Oxidative stress induces p53 phosphorylation at Ser15, but not serine 9, 20, or 392. 46 In addition, oxidative stress does not influence p53 protein levels. 46 Although there is no direct evidence to prove that GSK3β activation directly promotes p53 phosphorylation at Ser15, GSK3β inhibition reduces ROS generation. 40,47 It is possible that the inhibition of p53 phosphorylation by P7C3 may result from the inhibition of ROS generation that is associated with GSK3β activation. In our study, P7C3 and GSK3β inhibitor SB216763 effectively blocked the MPP + -induced phosphorylation of Ser15 in p53, thus repressing DA neuronal death. In addition, P7C3 blocked MPTP-induced GSK3β and p53 activation to promote the survival of DA neurons in the SN in vivo.
In summary, we showed that P7C3 inhibits GSK3β activation to protect DA neurons from MPP + -or MPTPinduced toxicity in vitro or in vivo. Thus, our study provides a mechanistic explanation for the neuroprotective efficacy of Mitochondrial membrane potential measurement. To measure the MMP, P7C3-or MPP + -treated MES23.5 cells were incubated with 100 nmol/l TMRM (Sigma) in PBS for 15 min at 37°C. Healthy mitochondria were indicated with TMRM in red, while damaged mitochondria rapidly lost the red fluorescence, implying the loss of MMP. After incubation, MES23.5 cells were immediately washed with PBS and then resuspended in PBS. Finally, the cells were observed using an inverted IX71 microscope system (Olympus).
Measurement of reactive oxygen species production. P7C3-or MPP + -treated MES23.5 cells were incubated in serum-free media with 5 μM 2',7'dichlorofluorescin diacetate (DCFH-DA) (Beyotime) at 37°C for 20 min. DCFH-DA transforms into fluorescent DCF to assess the generation of intracellular ROS. The cells were then washed with PBS and resuspended in PBS. Intracellular ROS were measured using flow cytometry.
Quantification detection of apoptosis. MES23.5 cells were administrated as described above. Apoptosis was measured by flow cytometry using the Annexin V-FITC Apoptosis Detection Kit (Beyotime) according to the manufacturer's instructions.
Immunohistochemistry. Male C57BL/6 mice, 25-30 g, were administrated P7C3 or MPTP as described in the animal experiments above. Then, the mice were perfused with 0.9% saline, followed by 4% paraformaldehyde in 0.1 M PBS (pH 7.4). After perfusion, the mice brains were harvested for post-fixation in the same fixing agent overnight at 4°C, followed by the treatment with the 30% sucrose at 4°C for another night. Serial 20 μM-thick mouse midbrain slices were prepared with a freezing microtome. Immunohistochemical staining was conducted using anti-TH antibodies (Millipore) against six slices per mouse (120 μm interval). After incubation with primary antibody at room temperature for 6 h, the slices were incubated with rhodamine (red)-or FITC (green)-conjugated secondary antibody (Invitrogen, Carlsbad, CA, USA) for 2 h. Finally, the slices were stained with DAPI for 5 min and observed using an inverted IX71 microscope system (Olympus). TH + neurons were counted with Image J software (National Institutes of Health, Bethesda, MD, USA) for each slice by two blinded investigators.
Statistical analysis. The quantification analysis of immunoblots from three independent experiments was conducted using Photoshop 7.0 (Adobe, San Jose, CA, USA). The data were analyzed using origin 6.0 (Originlab, Northampton, MA, USA). Statistical significance was evaluated using one-way ANOVA. The criterion of significance was set at Po0.05. The values are shown as the means ± S.E.M.

Conflict of Interest
The authors declare no conflict of interest.