Celastrol ameliorates Cd-induced neuronal apoptosis by targeting NOX2-derived ROS-dependent PP5-JNK signaling pathway

Abstract

Celastrol, a plant-derived triterpene, has neuroprotective benefit in the models of neurodegenerative disorders that are characterized by overproduction of reactive oxygen species (ROS). Recently, we have reported that cadmium (Cd) activates c-Jun N-terminal kinase (JNK) pathway leading to neuronal cell death by inducing ROS inactivation of protein phosphatase 5 (PP5), and celastrol prevents Cd-activated JNK pathway against neuronal apoptosis. Therefore, we hypothesized that celastrol could hinder Cd induction of ROS-dependent PP5-JNK signaling pathway from apoptosis in neuronal cells. Here we show that celastrol attenuated Cd-induced expression of NADPH oxidase 2 (NOX2) and its regulatory proteins (p22^phox, p40^phox, p47^phox, p67^phox, and Rac1), as well as the generation of ROS in PC12 cells and primary neurons. Also, N-acetyl-L-cysteine (NAC), a ROS scavenger, potentiated celastrol’s inhibition of the events in the cells triggered by Cd, implying neuroprotection by celastrol via blocking Cd-evoked NOX2-derived ROS. Further research revealed that celastrol was involved in the regulation of PP5 inactivation and JNK/c-Jun activation induced by Cd, as celastrol prevented Cd from reducing PP5 expression, and over-expression of wild-type PP5 or dominant negative c-Jun strengthened celastrol’s inhibition of Cd-induced phosphorylation of JNK and/or c-Jun, as well as apoptosis in neuronal cells. Of importance, inhibiting NOX2 with apocynin or silencing NOX2 by RNA interference enhanced the inhibitory effects of celastrol on Cd-induced inactivation of PP5, activation of JNK/c-Jun, ROS and apoptosis in the cells. Furthermore, we noticed that expression of wild-type PP5 or dominant negative c-Jun, or pretreatment with JNK inhibitor SP600125 reinforced celastrol’s suppression of Cd-induced NOX2 and its regulatory proteins, and consequential ROS in neuronal cells. These findings indicate that celastrol ameliorates Cd-induced neuronal apoptosis via targeting NOX2-derived ROS-dependent PP5-JNK signaling pathway. Our data highlight a beneficial role of celastrol in the prevention of Cd-induced oxidative stress and neurodegenerative diseases.

Graphical Abstract

We proposed that celastrol, a plant-derived triterpene, ameliorated cadmium (Cd)-elicited neuronal apoptosis by preventing Cd from upregulation of ROS-generating NOX2 and its regulatory proteins (p22^phox, p40^phox, p47^phox, p67^phox, and Rac1), thus suppressing ROS inactivation of PP5 and activation of JNK pathway. The findings highlight a beneficial role of celastrol in the prevention of Cd-induced oxidative stress and neurodegenerative diseases.

Introduction

Cadmium, one of the most toxic environmental and industrial pollutants, can accumulate in the brain by penetrating the blood-brain barrier, which leads to overproduction of reactive oxygen species (ROS) in neurons (Grandjean and Landrigan 2006; Mendez-Armenta and Rios 2007; Genovese and Cuzzocrea 2008; Wang and Du 2013). Excessive intracellular ROS induced by Cd can directly oxidize lipids, proteins, and nucleic acids, thereby leading to damage of the basic cell structures and consequential dysfunction of the central nervous system (CNS) (Bertin and Averbeck 2006; Genovese and Cuzzocrea 2008; Wang and Du 2013), for instance, headache and vertigo, olfactory dysfunction, slowing of vasomotor functioning, peripheral neuropathy, decreased equilibrium, neurobehavioral defects in attention, psychomotor speed, and learning disabilities (Pihl and Parkes 1977; Jarup et al. 1993; Wright et al. 2006; Wang and Du 2013). Emerging evidence suggests Cd-induced ROS as a pathogenic factor in the development of neurodegenerative disorders, such as Parkinson’s disease (PD), Alzheimer’s disease (AD) and Huntington’s disease (HD) (Lopez et al. 2006; Chen et al. 2008a; Hossain et al. 2009; Goncalves et al. 2010; Wei et al. 2015).

NADPH oxidases (NOXs), including NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1, and DUOX2, are a family of transmembrane multiunit enzymes that share the capacity to transport electrons across the plasma membrane and dedicate to ROS generation (Bedard and Krause 2007; Block and Gorin 2012). NOX2, also known as gp91^phox, is essential in innate host defense (Brown and Griendling 2009), and its expression has been detected in phagocytes, endothelium, vascular smooth muscle cells, fibroblasts, cardiomyocytes, skeletal muscle, hepatocytes, hematopoietic stem cells and CNS (Bedard and Krause 2007; Brown and Griendling 2009). NOX2 exists in close association with p22^phox, and its activation involves interaction with p40^phox, p47^phox, p67^phox and the small GTPase Rac1 (Bedard and Krause 2007; Brown and Griendling 2009; Block and Gorin 2012). Once assembled, NOX2 will be active and fuse with phagosomes and the plasma membrane to form NOX2-containing vesicles. The active enzyme complex generates ROS by transporting electrons from cytoplasmic NADPH to extracellular or phagosomal oxygen (Bedard and Krause 2007). NOX2 gene expression is inducible in response to multiple stimuli (Bedard and Krause 2007). Recently, our group has demonstrated that the expression of NOX2 and its regulatory proteins is upregulated by Cd, which is associated with Cd-induced ROS-dependent apoptosis in neuronal cells (Chen et al. 2011).

Mitogen-activated protein kinases (MAPKs) are evolutionarily highly conserved cascade of serine/threonine protein kinases, which connect cell surface receptors to regulatory targets in response to various stimuli (Kyriakis and Avruch 2001; Li et al. 2004; Kyriakis and Avruch 2012). Mammalian cells express at least three distinct groups of MAPKs, including extracellular signal-regulated protein kinase 1/2 (Erk1/2), c-Jun N-terminal kinase (JNK), and p38 MAPK (Kyriakis and Avruch 2012). In neuronal cells, activation of JNK signaling cascades by environmental stress or other stimuli has been shown to promote neuronal cell death (Moon et al. 2013; Moon and Park 2015). Protein phosphatase 5 (PP5) negatively regulates JNK cascade, involved in stress responses (Morita et al. 2001; Huang et al. 2004). In our previous studies, we have demonstrated that Cd activates JNK pathway leading to neuronal cell death by inducing ROS inactivation of PP5 (Chen et al. 2008a). However, whether Cd induces neuronal apoptosis by targeting NOX2-derived ROS inactivation of PP5 and activation JNK pathway is still unknown. Especially, it is important to find effective interventions for Cd-induced the events in the neuronal cells. Hence, we proposed that a compound that can prevent Cd-induced NOX2-derived ROS from inactivating PP5 and/or activating JNK pathway might be useful to prevent the neurotoxicity of Cd.

Celastrol, a pentacyclic triterpene, is extracted from the roots of the Tripterygium wilfordii (thunder god vine) plant (Corson and Crews 2007). As a traditional Chinese medicine, celastrol is typically used for treating rheumatoid arthritis, however, recently celastrol has attracted great attentions for its potent anticancer, anti-obesity and neuroprotective effects (Allison et al. 2001; Tao et al. 2001; Corson and Crews 2007; Liu et al. 2015a). Celastrol has been considered as a natural anti-oxidant, which can effectively suppress exogenous and endogenous oxidative stress in a variety of cells, including neuronal cells (Allison et al. 2001; Yu et al. 2010; Gu et al. 2013; Choi et al. 2014; Guan et al. 2016). Furthermore, numerous studies have shown that celastrol may inhibit NOX isoforms, especially NOX1 and NOX2 (Jaquet et al. 2011; Altenhofer et al. 2015). Celastrol targets JNK pathway as well (Li et al. 2012; Li et al. 2015; Lo Iacono et al. 2015). In rat cerebral ischemia model, the protective effect of celastrol is partly attributed to suppressing JNK pathway (Li et al. 2012). Interestingly, we have also unveiled that celastrol protects against Cd-induced neuronal apoptosis by inhibiting JNK pathway (Chen et al. 2014a). Therefore, we hypothesized that celastrol could hinder Cd-induced ROS-generating NOX2 family from ROS-dependent PP5-JNK signaling pathway contributing to apoptosis in neuronal cells. Here, for the first time, we demonstrated that celastrol attenuated Cd-induced upregulation of NOX2 and its regulatory proteins (p22^phox, p40^phox, p47^phox, p67^phox, and Rac1), as well as overproduction of ROS in neuronal cells, and identified that celastrol ameliorated Cd-induced neuronal apoptosis via targeting NOX2-derived ROS-dependent PP5-JNK signaling pathway. Our findings highlight that celastrol has a potential to prevent Cd-induced oxidative stress and neurodegeneration.

Materials and methods

Materials

Cadmium chloride, celastrol, poly-D-lysine (PDL), N-acetyl-L-cysteine (NAC), SP600125, apocynin, and protease inhibitor cocktail were purchased from Sigma (St. Louis, MO, USA). 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate (CM-H₂DCFDA) was purchased from MP Biomedicals (Solon, OH, USA). Dulbecco’s modified Eagle medium (DMEM), 0.05% Trypsin-EDTA, NEUROBASAL™ Media and B27 Supplement were purchased from Invitrogen (Grand Island, NY, USA). Horse serum and fetal bovine serum (FBS) were supplied by Hyclone (Logan, UT, USA). Enhanced chemiluminescence reagent was from Millipore (Billerica, MA, USA). CellTiter 96^® AQ_ueous One Solution Cell Proliferation Assay kit was from Promega (Madison, WI, USA). Other chemicals were purchased from local commercial sources and were of analytical grade, unless stated elsewhere.

Cell culture

Rat pheochromocytoma (PC12) cell line (RRID: CVCL_0481) was obtained from American Type Culture Collection (ATCC) (Manassas, VA, USA). Because of the replicative nature and cost-effectiveness, the cell line is widely used as neuronal cell models, so it was employed in this study. For culture, PC12 cells, seeded in a 6-well plate (5×10⁵ cells/well) or 96-well plate (1×10⁴ cells/well) pre-coated with 0.2 µg/ml PDL, were grown in antibiotic-free DMEM supplemented with 10% horse serum and 5% FBS. Cells were maintained in a humidified incubator of 5% CO₂ at 37°C. To verify the data obtained from PC12 cells, primary neurons were also used in this study. For this, primary murine neurons were isolated from fetal mouse cerebral cortexes of 16–18 days of gestation in female ICR mice (being pregnant) as described (Chen et al. 2010), and seeded in a 6-well plate (5×10⁵ cells/well) or 96-well plate (1×10⁴ cells/well) coated with 10 µg/ml PDL for experiments after 6 days of culture. All procedures used in this study were approved by the Institutional Animal Care and Use Committee, and were in compliance with the guidelines set forth by the Guide for the Care and Use of Laboratory Animals.

Recombinant adenoviral constructs and infection of cells

Hemagglutinin (HA)-tagged wild type human PP5 plasmid (Morita et al. 2001) was a gift from Dr. Hidenori Ichijo (University of Tokyo, Tokyo, Japan). The recombinant adenovirus encoding HA-tagged PP5 (Ad-PP5) was generated using the ViraPower Adenoviral Expression System (Invitrogen, Carlsbad, CA), as described (Chen et al. 2008a). The recombinant adenovirus expressing FLAG-tagged dominant negative c-Jun (FLAG-△169) (Ad-dn-c-Jun) (Whitfield et al. 2001) (a gift from Dr. Jonathan Whitfield, Eisai London Research Laboratories, University College London, London, UK), and the control adenovirus expressing the green fluorescent protein (GFP) (Ad-GFP) or β-galactosidase (Ad-LacZ) were described previously (Chen et al. 2008a; Chen et al. 2008b; Liu et al. 2010). For experiments, PC12 cells were grown in the growth medium and infected with the individual adenovirus for 24 h at 5 of multiplicity of infection (MOI = 5). Afterwards, cells were used for experiments. Cells infected with Ad-GFP or Ad-LacZ served as a control. Expression of HA-tagged PP5 and FLAG-tagged dn-c-Jun was assessed by Western blot analysis with antibodies to HA and FLAG, respectively.

Lentiviral shRNA cloning, production and infection

To generate lentiviral shRNA to NOX2, oligonucleotides containing the target sequences were synthesized, annealed and inserted into FSIPPW lentiviral vector (Kanellopoulou et al. 2005) via the EcoR1/BamH1 restriction enzyme site. The sequences of oligonucleiotides used were: NOX2 sense: 5’-AATTCCCGGGATGAATCTCAGGCCAATCTGCAAGAGAGATTGGCCTGAGATTCATCCCTTTTTG-3’, anti-sense: 5’-GATCCAAAAAGGGATGAATCTCAGGCCAATCTCTCTTGCAGATTG GCCTGAGATTCATCCCGGG-3’. Lentiviral shRNA to GFP (for control) was described previously (Liu et al. 2006). To produce lentiviral shRNAs, above constructs were co-transfected together with pMD2G and psPAX2 (Addgene, Cambridge, MA, USA) to 293TD cells using MegaTran 1.0 reagent (OriGene Technologies, Rockville, MD, USA). Each virus-containing medium was collected at 48 h and 60 h post-transfection, respectively. For use, monolayer PC12 cells, when grown to about 70% confluence, were infected with the corresponding lentivirus containing medium in the presence of 8 mg/ml polybrene for 12 h twice at an interval of 6 h. Uninfected cells were eliminated by exposure to 2 mg/ml puromycin for 48 h before use. After 5 days of culture, cells were used for experiments.

Analysis for cell viability

PC12 cells infected with Ad-PP5, Ad-dn-c-Jun or Ad-GFP, or PC12 cells infected with lentiviral shRNA to NOX2 or GFP, respectively, were seeded in a PDL-coated 96-well plate (1×10⁴ cells/well). The next day, cells were treated with/without Cd (10 µM) for 24 h following pre-incubation with/without celastrol (1 µM) for 1 h with 5 replicates of each treatment. In some cases, cells were pretreated with/without a JNK inhibitor SP600125 (20 µM) or a pan caspase inhibitor zVAD-fmk (100 µM) for 1 h, and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 24 h. Subsequently, cell viability, after incubation with MTS reagent (one solution reagent) (20 µl/well) for 3 h, was evaluated by measuring the optical density (OD) at 490 nm using a Victor X3 Light Plate Reader (PerkinElmer, Waltham, MA, USA).

Assays for cell caspase-3/7 activity

PC12 cells and primary neurons were seeded in a PDL-coated 96-well plate (1×10⁴ cells/well). The next day, cells were pretreated with/without an antioxidant and ROS scavenger NAC (5 mM) for 1 h, and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 24 h with 5 replicates of each treatment. Subsequently, caspase-3/7 activity was determined using Caspase-Glo^® 3/7 Assay Kit (Promega, Madison, WI, USA), following the instructions of the supplier.

TUNEL staining

The terminal deoxynucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate (dUTP) nick-end labeling (TUNEL) staining was performed according to the manufacturer’s protocols of In Situ Cell Death Detection Kit^® (Roche, Mannheim, Germany). In brief, PC12 cells and primary neurons, or PC12 cells infected with Ad-PP5, Ad-dn-c-Jun or Ad-LacZ, or PC12 cells infected with lentiviral shRNA to NOX2 or GFP, respectively, were seeded at a density of 5 × 10⁵ cells/well in a 6-well plate containing a PDL-coated glass coverslip per well. The next day, cells were treated with/without Cd (10 µM) for 24 h following pre-incubation with/without celastrol (1 µM) for 1 h with 5 replicates of each treatment. In some cases, cells were pretreated with/without NAC (5 mM), SP600125 (20 µM), zVAD-fmk (100 µM), or a NOX2 inhibitor apocynin (50 µM) for 1 h, and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 24 h. Afterwards, cells were fixed with 4% paraformaldehyde prepared in PBS for 2 h at 4°C. The fixed cells of each slide were washed 3 times with PBS, and then incubated in the permeabilization solution (0.1% Triton 100-X, 0.1% sodium citrate) for 2 min on ice, followed by addition of TUNEL reaction mixture (TdT enzyme solution and labeling solution) to the samples and incubation for 1 h in a dark and humidified incubator at 37°C. After TUNEL labeling reaction, all stained specimens were rinsed 3 times with PBS and mounted with coverslips containing a mounting medium. Finally, photographs were taken under a fluorescence microscope (Leica DMi8, Wetzlar, Germany) equipped with a digital camera. For quantitative analysis of the fluorescence intensity, five image fields from each treatment were randomly selected, and the integral optical density (IOD) was measured by Image-Pro Plus 6.0 software (Media Cybernetics Inc., Newburyport, MA, USA).

Cell ROS assay and ROS imaging

PC12 cells and primary neurons, or PC12 cells infected with Ad-PP5, Ad-dn-c-Jun or Ad-LacZ, or PC12 cells infected with lentiviral shRNA to NOX2 or GFP, respectively, were seeded in a PDL-coated 96-well plate (1×10⁴ cells/well) or in a 6-well plate (5 × 10⁵ cells/well) containing a PDL-coated glass coverslip per well. The next day, cells were treated with/without Cd (10 and/or 20 µM) for 24 h following pre-incubation with/without celastrol (1 µM) or apocynin (10–100 µM) for 1 h with 5 replicates of each treatment. In some cases, cells were pretreated with/without NAC (5 mM), SP600125 (20 µM) or apocynin (50 µM) for 1 h, and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 24 h. Subsequently, intracellular ROS fluorescence was detected and imaged using an oxidant-sensitive probe, CM-H₂DCFDA (MP Biomedicals), and quantitatively analyzed as described (Xu et al., 2016). Images were randomly taken from five fields of each treatment, followed by quantitative analysis.

Western blot analysis

PC12 cells and primary neurons, or PC12 cells infected with lentiviral shRNA to NOX2 or GFP, or PC12 cells infected with Ad-PP5, Ad-dn-c-Jun or Ad-GFP, respectively, after treatments, were lysed, followed by Western blotting, as described previously (Chen et al., 2010; Chen et al., 2014a). The blots for detected protein were semi-quantified using NIH Image J software (National Institutes of Health, Bethesda, MD, USA). The following antibodies were used: phosphorylated JNK (p-JNK) (Thr183/Tyr185), JNK, p-c-Jun (Ser63), c-Jun, p22^phox (Santa Cruz Biotechnology, Santa Cruz, CA, USA), caspase-3, and PARP (Cell Signaling Technology, Beverly, MA, USA), PP5 (BD Biosciences, San Jose, CA), NOX2, p67^phox (Epitomics, Burlingame, CA, USA), Rac1 (Cytoskeleton, Denver, CO, USA), p40^phox, p47^phox, FLAG, HA, and β-tubulin (Sigma), goat anti-rabbit IgG-horseradish peroxidase (HRP), goat anti-mouse IgG-HRP, and rabbit anti-goat IgG-HRP (Pierce, Rockford, IL, USA).

Statistical analysis

All data were expressed as mean values ± standard error (mean ± SE). Student’s t-test for non-paired replicates was used to identify statistically significant differences between treatment means. Group variability and interaction were compared using either one-way or two-way ANOVA followed by Bonferroni’s post-tests to compare replicate means. p-value < 0.05 was significant.

Results

Celastrol attenuates Cd-induced ROS contributing to apoptosis by preventing Cd upregulation of NOX2 and its regulatory proteins in neuronal cells

Our recent study has shown that Cd evokes ROS-dependent neuronal apoptosis by upregulating NOX2 and its regulatory proteins including p22^phox, p40^phox, p47^phox, p67^phox and Rac1 (Chen et al. 2011). Celastrol relieves Cd-induced neuronal cell death (Chen et al. 2014a). Additionally, it has been reported that celastrol is a potent inhibitor of NOX1 and NOX2 (Jaquet et al. 2011). Here, using Western blot analysis, we found that celastrol remarkably inhibited Cd-upregulated expression of NOX2, p22^phox, p40^phox, p47^phox, p67^phox and Rac1 (Fig. 1a). We also observed that celastrol obviously attenuated the intracellular ROS production in PC12 and primary neurons induced by 24-h exposure to Cd (Fig. 1b–d), as evidenced by detecting, imaging and quantifying using an oxidant-sensitive probe, CM-H₂DCFDA. This is agreement with the findings that celastrol attenuates Cd-induced cell viability reduction and apoptosis (Chen et al. 2014a).

Fig. 1.

Celastrol inhibited Cd-upregulated expression of NOX2 and its regulatory proteins as well as ROS generation in neuronal cells. PC12 cells and primary neurons were pretreated with/without celastrol (1 µM) for 1 h and then exposed to Cd (10 and/or 20 µM) for 12 h (for Western blotting) or 24 h (for ROS assay and ROS imaging). (a) Total cell lysates were subjected to Western blotting using indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were observed in at least three independent experiments. (b and c) Cell ROS assay and ROS imaging (in green) was evaluated using an oxidant-sensitive probe CM-H₂DCFDA. Scale bar: 20 µm. (d) Fluorescent intensity of ROS imaging was quantified. Results were presented as mean ± SE, n=5. ^ap < 0.05, difference with control group; ^bp < 0.05, difference with 10 µM Cd group; ^cp < 0.05, difference with 20 µM Cd group.

To confirm the relationship of celastrol’s prevention of Cd-induced ROS-dependent apoptosis with celastrol’s suppression of Cd-upregulated NOX2 family members in neuronal cells, an antioxidant and ROS scavenger NAC was employed. The results showed that NAC strikingly blocked Cd-upregulated expression of NOX2 and its regulatory proteins, as well as Cd-induced ROS generation in the cells (Fig. 2a and b). Especially, co-treatment with celastrol/NAC exhibited a stronger inhibitory effect on Cd-stimulated NOX2 family members and ROS generation in the cells (Fig. 2a and b). Consistently, co-treatment with celastrol/NAC reduced the number of TUNEL-positive cells with fragmented DNA more potently than treatment with NAC or celastrol alone in the cells (Fig. 2c and d). NAC also profoundly diminished activation of caspases 3/7 in the cells in response to Cd, and substantially enhanced the inhibitory effect of celastrol (Fig. 2e). These findings clearly indicate that celastrol attenuates Cd-induced ROS contributing to apoptosis by preventing Cd upregulation of NOX2 and its regulatory proteins in neuronal cells.

Fig. 2.

NAC potentiated celastrol’s rescue against Cd-induced NOX2 family members, ROS generation and apoptosis in neuronal cells. PC12 cells and primary neurons were pretreated with/without NAC (5 mM) for 1 h and then celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 12 h (for Western blotting) or 24 h (for ROS imaging, TUNEL staining and caspase-3/7 activity assay). (a) Total cell lysates were subjected to Western blotting using indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were observed in at least three independent experiments. (b) Cell ROS was imaged and quantified using an oxidant-sensitive probe CM-H₂DCFDA. (c) Apoptotic cells were evaluated by in situ detection of fragmented DNA (in green) using TUNEL staining. Scale bar: 20 µm. (d) The number of TUNEL-positive cells was quantified. (e) Caspase-3/7 activities were detected using Caspase-3/7 Assay Kit. Results were presented as mean ± SE, n=5. ^ap < 0.05, difference with control group; ^bp < 0.05, difference with 10 µM Cd group; ^cp < 0.05, difference with Cd/Celastrol group or Cd/NAC group.

Celastrol blocks Cd activation of JNK pathway leading to neuronal apoptosis by mitigating Cd downregulation of PP5

Given that Cd inhibition of PP5 results in activation of JNK pathway and apoptosis of neuronal cells (Chen et al. 2008a), and celastrol attenuates Cd-induced neuronal apoptosis by blocking JNK pathway (Chen et al. 2014a), we reasoned that celastrol might reverse PP5 inactivation and subsequent JNK activation induced by Cd. As predicted, Cd-reduced PP5 expression was indeed rescued by celastrol in PC12 cells and primary neurons (Fig. 3a and b). To dissect the role of PP5 in celasrol’s blockage of Cd-induced JNK activation and neuronal apoptosis, PC12 cells, infected with recombinant adenovirus expressing HA-tagged wild-type human PP5 (Ad-PP5) or Ad-GFP/Ad-LacZ (as control), were pretreated with celastrol (1 µM) or a JNK inhibitor SP600125 (20 µM) for 1 h, followed by exposure to Cd (10 µM) for 12 or 24 h. By Western blot analysis, we observed over-expression of HA-tagged PP5 in the cells infected with Ad-PP5 (Fig. 3c). In line with our previous findings (Chen et al. 2008a), exposure to Cd decreased PP5 expression, and correspondingly increased the phosphorylation of JNK, particularly the protein expression and phosphorylation of c-Jun, a substrate of JNK in the control cells infected with Ad-GFP (Lane 4 vs. Lane 1) (Fig. 3c and d). Over-expression of PP5 partially blocked Cd-induced p-JNK/p-c-Jun (Lane 10 vs. Lane 4) (Fig. 3c and d). Celastrol, but not SP600125, attenuated Cd-induced decrease in PP5 expression; both celastrol and SP600125 mitigated Cd-induced p-JNK/p-c-Jun (Lane 5 vs. Lane 4, Lane 6 vs. Lane 4) (Fig. 3c and d). Furthermore, over-expression of PP5 was able to enhance the inhibitory effects of celastrol or SP600125 on Cd-induced p-JNK/p-c-Jun (Lane 11 vs. Lane 5, Lane 12 vs. Lane 6). Moreover, over-expression of PP5 also reinforced the protective effect of celastrol or SP600125 against Cd-induced cleavage of caspase-3 (Lane 11 vs. Lane 5, Lane 12 vs. Lane 6) (Fig. 3c and d). By MTS assay and TUNEL staining, we observed that over-expression of PP5 alone partially prevented Cd-induced cell viability reduction and apoptosis in PC12 cells (Fig. 3e and f). Addition of celastrol or SP600125 elicited more potent protection against Cd-induced apoptosis (Fig. 3e and f).

Fig. 3.

Celastrol blocked Cd activation of JNK pathway and neuronal apoptosis by preventing Cd from inactivation of PP5. PC12 cells and primary neurons, or PC12 cells infected with Ad-PP5, Ad-dn-c-Jun or Ad-GFP/Ad-LacZ (as control), respectively, were pretreated with/without celastrol (1 µM) for 1 h, or with/without SP600125 (20 µM) or zVAD-fmk (100 µM) for 1 h and then with/without celastrol for 1 h, followed by exposure to Cd (10 and/or 20 µM) for 12 h (for Western blotting) or 24 h (for MTS assay and TUNEL staining). (a-d, g and h) Total cell lysates were subjected to Western blotting using indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were observed in at least three independent experiments (a, c and g), and blots for PP5, p-JNK, p-c-Jun, cleaved-caspase-3, cleaved-PARP were semi-quantified (b, d and h). (e and i) Cell viability was evaluated by MTS assay. (f and j) The number of TUNEL-positive cells was quantified by in situ detection of fragmented DNA using TUNEL staining. Results were presented as mean ± SE, n=3–5. ^ap < 0.05, difference with control group; ^bp < 0.05, difference with 10 µM Cd group; ^cp < 0.05, difference with 20 µM Cd group; ^dp < 0.05, Ad-PP5 group or Ad-dn-c-Jun group versus Ad-GFP group or Ad-LacZ group.

Next, recombinant adenovirus expressing FLAG-tagged dominant negative c-Jun (Ad-dn-c-Jun) was utilized. The results showed that ectopic expression of FLAG-tagged dn-c-Jun significantly suppressed Cd-triggered protein expression and phosphorylation of c-Jun (Fig. 3g and h). Celastrol, but not zVAD-fmk, powerfully attenuated Cd-increased c-Jun protein expression and phosphorylation (Fig. 3g and h). Interestingly, over-expression of dn-c-Jun was able to potentiate the inhibitory effects of celastrol on Cd-induced c-Jun/p-c-Jun (Fig. 3g and h). Of importance, over-expression of dn-c-Jun potently reinforced the preventive effects of celastrol or zVAD-fmk against Cd-induced cleavage of caspase-3 (Fig. 3g and h), as well as cell viability reduction and apoptosis in PC12 cells (Fig. 3i and j). Collectively, the findings support the notion that celastrol inhibits Cd-induced activation of JNK pathway and consequential cell apoptosis, at least in part, by mitigating Cd downregulation of PP5 in neuronal cells.

Celastrol prevents Cd-induced PP5-JNK signaling pathway from neuronal apoptosis by targeting NOX2-derived ROS

To uncover whether celastrol’s prevention of Cd-induced PP5-JNK signaling pathway from neuronal apoptosis is attributed to its inhibition of Cd-stimulated NOX2-derived ROS, apocynin, an inhibitor of NOXs (Touyz 2008), was employed. The results showed that apocynin (10–100 µM) inhibited Cd-induced upregulation of NOX2 and generation of ROS in PC12 cells and primary neurons in a concentration-dependent manner (Fig. 4a–c). Of note, co-treatment with apocynin (50 µM) and celastrol (1 µM) exhibited a more significant inhibitory effect on Cd-upregulated NOX2 compared to treatment with apocynin or celastrol alone in PC12 cells and primary neurons (Fig. 4d). Cd-induced downregulation of PP5 as well as activation of JNK/c-Jun and caspase-3 were reversed by co-treatment with apocynin/celastrol more potently by treatment with apocynin or celastrol alone in the cells (Fig. 4d and e). In line with this, the combination of celastrol and apocynin more powerfully inhibited cell ROS generation and apoptosis than celastrol or apocynin alone in the cells in response to Cd (Fig. 4f and g).

Fig. 4.

Pharmacological inhibition of NOX2 with apocynin strengthened celastrol’s preventive effects on Cd inactivation of PP5, activation of JNK and neuronal apoptosis. PC12 cells and primary neurons were pretreated with/without apocynin (10–100 µM) for 1 h, or pretreated with/without apocynin (50 µM) for 1 h and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 12 h (for Western blotting) or 24 h (for ROS imaging and TUNEL staining). (a, b, d and e) Total cell lysates were subjected to Western blotting using indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were observed in at least three independent experiments (a and d), and blots for NOX2, PP5, p-JNK, p-c-Jun, cleaved-caspase-3 were semi-quantified (b and e). (c and f) Cell ROS was imaged and quantified using an oxidant-sensitive probe CM-H₂DCFDA. (g) The number of TUNEL-positive cells was quantified by in situ detection of fragmented DNA using TUNEL staining. Results were presented as mean ± SE, n=3–5. ^ap < 0.05, difference with control group; ^bp < 0.05, difference with 10 µM Cd group; ^cp < 0.05, difference with Cd/Celastrol or Cd/Apocynin group.

To further corroborate the role of NOX2 in celastrol’s intervention of Cd-induced PP5-JNK signaling pathway and neuronal apoptosis, PC12 cells, infected with lentiviral shRNA to NOX2 or GFP, were exposed to Cd (10 µM) for 12 h or 24 h following pretreatment with/without celastrol (1 µM) for 1 h. As shown in Fig. 5a, NOX2 expression was downregulated by ~90% in shRNA NOX2-infected cells compared to shRNA GFP-infected cells. RNAi-mediated depletion of NOX2 did not affect the basal protein content of PP5, and activation of JNK and caspase-3. However, silencing NOX2 significantly strengthened celastrol’s suppression of Cd-induced downregulation of PP5, upregulation of c-Jun, and phosphorylation of JNK/c-Jun (Fig. 5a and b). Consistently, depleting NOX2 also enhanced the inhibitory effects of celastrol on Cd-induced cleavage of caspase-3 (Fig. 5a and b), ROS production (Fig. 5c), as well as cell viability reduction (Fig. 5d) and apoptosis (Fig. 5e) in the cells. Taken together, these data verify that celastrol prevents Cd-induced activation of JNK cascade from neuronal apoptosis by suppressing NOX2-derived ROS.

Fig. 5.

Down-regulation of NOX2 reinforced celastrol’s blockage of Cd inactivation of PP5, activation of JNK, and neuronal apoptosis. PC12 cells, infected with lentiviral shRNA to NOX2 or GFP (as control), were pretreated with/without celastrol (1 µM) for 1 h and then exposed to Cd (10 µM) for 12 h (for Western blotting) or 24 h (for ROS imaging, MTS assay and TUNEL staining). (a and b) Total cell lysates were subjected to Western blotting using indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were observed in at least three independent experiments (a), and blots for PP5, p-JNK, p-c-Jun, cleaved-caspase-3 were semi-quantified (b). (c) Cell ROS was imaged and quantified using an oxidant-sensitive probe CM-H₂DCFDA. (d) Cell viability was evaluated by MTS assay. (e) The number of TUNEL-positive cells was quantified by in situ detection of fragmented DNA using TUNEL staining. Results were presented as mean ± SE, n=3–5. ^ap < 0.05, difference with control group; ^bp < 0.05, difference with 10 µM Cd group; ^cp < 0.05, NOX2 shRNA group versus GFP shRNA group.

Modulation of JNK/c-Jun and PP5 activity interferes with celastrol’s inhibition of Cd-stimulated NOX2 family members and ROS

To pinpoint whether JNK positively mediates Cd-stimulated NOX2 family members and ROS generation, and the role of celastrol in preventing Cd-induced NOX2-derived ROS from activation of JNK, PC12 cells and primary neurons were pretreated with/without SP600125 (20 µM) for 1 h, and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 12 h or 24 h. The results showed that treatment with SP600125 alone significantly decreased Cd-elevated expression of NOX2, p22^phox, p40^phox, p47^phox, p67^phox and Rac1 (Fig. 6a and b), as well as ROS generation in the cells (Fig. 6c). Of importance, there existed more inhibitory effects on Cd-induced NOX2 family members and ROS in the cells co-treated with celastrol/SP600125 than in those treated with celastrol or SP600125 alone (Fig. 6a–c).

Fig. 6.

Modulation of JNK/c-Jun or PP5 activity interfered with celastrol’s inhibition of Cd-stimulated NOX2 family members and ROS. PC12 cells and primary neurons, or PC12 cells infected with Ad-PP5, Ad-dn-c-Jun and Ad-GFP/Ad-LacZ (as control), respectively, were pretreated with/without celastrol (1 µM) for 1 h, or pretreated with/without SP600125 (20 µM) for 1 h and then with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 12 h (for Western blotting) or 24 h (for ROS imaging). (a, b, d, e, g and h) Total cell lysates were subjected to Western blotting using indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were observed in at least three independent experiments (a, d and g), and blots for NOX2, p22^phox, p40^phox, p47^phox, p67^phox, Rac1 were semi-quantified (b, e and h). (c, f and i) Cell ROS was imaged and quantified using an oxidant-sensitive probe CM-H₂DCFDA. Results were presented as mean ± SE, n=3–5. ^ap < 0.05, difference with control group; ^bp < 0.05, difference with 10 µM Cd group; ^cp < 0.05, difference with Cd/Celastrol or Cd/SP600125 group; ^dp < 0.05, Ad-dn-c-Jun group or Ad-PP5 group versus Ad-LacZ group.

Considering that the JNK inhibitor SP600125 (20 µM) may have off-target effects, a genetic approach was taken to validate the role of JNK pathway in celastrol’s prevention of Cd-induced neuronal apoptosis. As JNK executes its pro-apoptotic action by activation of c-Jun (Kyriakis and Avruch 2001), PC12 cells were infected with Ad-dn-c-Jun or Ad-GFP/Ad-LacZ (as control), and then exposed to Cd (10 µM) for 12 h or 24 h post pre-incubation with/without celastrol (1 µM) for 1 h. The results showed that expression of dn-c-Jun partially prevented the cells from upregulation of NOX2, p22^phox, p40^phox, p47^phox, p67^phox and Rac1 as well as overproduction of ROS induced by Cd (Fig. 6d–f). Notably, expression of dn-c-Jun strengthened the inhibitory effects of celastrol on Cd-induced NOX2 family members and ROS in the cells (Fig. 6d–f).

As PP5 negatively regulates JNK cascade (Morita et al. 2001; Huang et al. 2004), we continued to study the role of PP5 in positive mediation of Cd-induced NOX2 family members and ROS. To this end, PC12 cells were infected with Ad-PP5 or Ad-GFP/Ad-LacZ (control), and then pretreated with/without celastrol (1 µM) for 1 h, followed by exposure to Cd (10 µM) for 12 h or 24 h. As expected, over-expression of PP5 obviously suppressed Cd-induced upregulation of NOX2, p22^phox, p40^phox, p47^phox, p67^phox and Rac1 as well as overproduction of ROS, and reinforced the inhibitory effects of celastrol on Cd-induced events in the cells (Fig. 6g–i). On the basis of these data, we conclude that celastrol exerts a beneficial role in preventing Cd-induced NOX2 family members and ROS from inactivation of PP5 and activation of JNK/c-Jun, which, in turn, also positively mediates Cd-induced NOX2 family members and ROS.

Discussion

Many studies have documented that Cd-induced ROS contributes to neurodegenerative diseases, such as AD, PD and HD (Lopez et al. 2006; Hossain et al. 2009; Goncalves et al. 2010; Wei et al. 2015). Recently we have shown that Cd in vitro and in vivo induces ROS generation by upregulating the expression of NOX2 and its regulatory proteins p22^phox, p67^phox, p40^phox, p47^phox, and Rac1, leading to neuronal apoptosis (Chen et al. 2011; Chen et al. 2014b). So finding effective interventions for prevention of Cd-induced NOX2 family members and ROS in neuronal cells is of great importance for treatment of Cd-induced neurodegenerative diseases. Celastrol, a natural compound extracted from the roots of Tripterygium wilfordii, has been reported to show neuroprotective effects on the models for AD, PD and ALS (Allison et al. 2001; Cleren et al. 2005; Kiaei et al. 2005). As a natural antioxidant, celastrol can effectively suppress exogenous and endogenous oxidative stress in various types of cells, including neuronal cells (Allison et al. 2001; Yu et al. 2010; Gu et al. 2013; Choi et al. 2014; Guan et al. 2016). Celastrol may inhibit NOX isoforms, especially NOX1 and NOX2 (Jaquet et al. 2011; Altenhofer et al. 2015). Our recent studies have revealed that celastrol attenuates Cd-induced neuronal apoptosis by inhibiting JNK pathway (Chen et al. 2014a). However, it is not clear how celastrol prevents Cd-activated JNK pathway related to Cd neurotoxicity, and whether this is associated with celastrol’s target of NOX family members and ROS in neuronal cells induced by Cd exposure. Here, for the first time, we provide evidence that celastrol attenuated Cd-induced upregulation of NOX2, p22^phox, p40^phox, p47^phox, p67^phox, and the small GTPase Rac1, as well as overproduction of ROS in PC12 cells and primary neurons. Celastrol was involved in the regulation of PP5 inactivation and consequential JNK/c-Jun activation induced by Cd. Further, we found that celastrol prevented the effect of Cd on PP5-JNK cascade and neuronal apoptosis by suppressing NOX2-derived ROS.

NOXs, including NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1, and DUOX2, are a family of transmembrane multiunit enzymes (Bedard and Krause 2007; Block and Gorin 2012). High expression of NOX2 occurs in the CNS (Bedard and Krause 2007; Brown and Griendling 2009). Abnormal activity of NOX2 is closely related to the development of neurodegenerative diseases (Lambeth 2007). Therefore, we firstly focused on how celastrol affects Cd-induced NOX2-derived ROS contributing to apoptosis in neuronal cells. It is well known that NOX2 constitutively associates with its regulatory protein p22^phox, a noncatalytic subunit, and is activated through a series of interactions with other regulatory proteins p40^phox, p47^phox, p67^phox, and Rac1 as cytosolic subunits, which are associated with ROS generation in phagocytes and in numerous nonphagocytic cells, including neurons (Bedard and Krause 2007; Brown and Griendling 2009; Frey et al. 2009; Block and Gorin 2012). Several studies have shown that celastrol disrupts the binding of p22^phox and p47^phox by specifically binds of p47^phox, which leads to the inhibition of NOX2 activity and ROS production (Jaquet et al. 2011; Cifuentes-Pagano et al. 2013; Altenhofer et al. 2015). In this study, we showed that, on one hand, celastrol’s attenuation of Cd-induced ROS production was closely related to its preventing Cd from upregulating the expression of ROS-generating enzyme NOX2 and its regulatory proteins in PC12 cells and primary neurons (Fig. 1a–d); on the other hand, celastrol attenuated Cd induction of ROS-stimulated expression of these ROS-generating proteins, because treatment with NAC enhanced celastrol’s suppression of Cd-induced ROS and Cd-induced expression of NOX2, p22^phox, p67^phox, p40^phox, p47^phox, and Rac1 in the cells (Fig. 2a and b). Consistently, NAC substantially attenuated the number of TUNEL-positive cells and activation of caspases 3/7 in the cells triggered by Cd, and potentiated the inhibitory effect of celastrol (Fig. 2c–e). The results suggest a positive feedback loop involved in celastrol’s mitigation of Cd-evoked ROS generation and NOX2 activity contributing to neuronal apoptosis by hindering a concomitant series of NOX2 regulatory proteins. In addition, it is worth mentioning that NOX2 gene expression is inducible in response to multiple stimuli (Bedard and Krause 2007). Also, we have observed that Cd upregulates the expression of NOX2 and its regulatory proteins by activating the mammalian target of rapamycin (mTOR) (Chen et al. 2011), which controls protein synthesis (Albert and Hall 2015). The function and activation of NOX2 is mainly attributed to combination with its regulatory proteins (Bedard and Krause 2007; Block and Gorin 2012). Therefore, it is important to further investigate special effects and correlation of celastrol on Cd-induced NOX2 and its regulatory proteins at either transcriptional and/or translational level in neuronal cells.

In this study, we found that Cd inhibition of PP5 protein expression was significantly reversed by celastrol in PC12 cells and primary neurons, suggesting celastrol’s prevention from Cd inactivation of PP5. PP5 is well-known to negatively regulate JNK pathway (Morita et al. 2001; Huang et al. 2004). Our recent studies have shown that Cd down-regulates PP5 activity, leading to activation of JNK pathway in PC12 cells (Chen et al. 2008a). Celastrol blocks Cd activation of JNK pathway associated with neuronal apoptosis (Chen et al. 2014a). Therefore, here we sought to gain more insights into the role of PP5 in celasrol’s blockage of JNK activation and apoptosis in neuronal cells in response to Cd. Our results demonstrated that celastrol mitigated Cd-induced neuronal apoptosis indeed by preventing Cd down-regulation of PP5 and consequential activation of JNK pathway. This is strongly supported by the findings that ectopic expression of wild-type PP5 or inhibition of JNK with SP600125 strengthened the inhibitory effect of celastrol on Cd-induced cell death in PC12 cells (Fig. 3c–j). In this study, we also showed that SP600125 dramatically inhibited Cd-activated c-Jun, a substrate of JNK in the cells (Fig. 3c and d). Considering that the JNK inhibitor SP600125 may have off-target effects, a genetic approach was conducted by ectopic expression of dn-c-Jun, confirming the role of JNK cascade in celastrol’s prevention of Cd-induced neuronal apoptosis (Fig. 3g–j). Our data underscore that celastrol has an ability to prevent Cd from inactivation of PP5 and activation of JNK/c-Jun, thereby attenuating Cd-induced neuronal apoptosis.

In this study, we also asked whether celastrol’s rescue against Cd-induced inactivation of PP5 and activation of JNK/c-Jun was due to celastrol’s inhibition of NOX2-derived ROS in neuronal cells induced by Cd. Using pharmacological inhibitor apocynin, which inhibits ROS-generating enzyme NOX activity by blocking the assembly of NOX complex (Aldieri et al. 2008; Touyz 2008), we found that apocynin (50 µM) effectively suppressed Cd-stimulated NOX2 activation and ROS generation (Fig. 4a–c). Of importance, inhibition of NOX2 with apocynin enhanced celastrol’s prevention of Cd-induced inactivation of PP5, activation of JNK/c-Jun, ROS and apoptosis in PC12 cells and primary neurons (Fig. 4a–g). This was further supported by the observations in NOX2-knockdown cells (Fig. 5a–e). Interestingly, we also noticed that silencing NOX2 by RNA interference almost completely rescued downregulation of PP5 and blocked phosphorylation of JNK/c-Jun in PC12 cells in response to celastrol and/or Cd, but only partially attenuated Cd-induced cleavage of caspase-3, as well as cell viability reduction and apoptosis in the cells, implying that involvement of Cd-activated other signaling pathways in neuronal apoptosis may not be substantially relieved by downregulation of NOX2. Taken together, we deduce that celastrol counteracts Cd-induced ROS by down-regulating the expression of NOX2 and its regulatory proteins, thereby preventing Cd-induced downregulation of PP5 and activation of JNK pathway.

It has been shown that endoplasmic reticulum (ER) stress and calcium imbalance are involved in Cd-induced cell death in a variety of cells, including neuronal cells (Biagioli et al. 2008; Kim et al. 2013; Liu et al. 2015b; Luo et al. 2016; Rajakumar et al. 2016). Celastrol affects ER stress and intracellular free Ca²⁺ ([Ca²⁺]_i) as well (Wang et al. 2012; Yoon et al. 2014; Liu et al. 2015a). Although this work did not involve these targets, during our research, we also observed that when celastrol-pretreated PC12 cells and primary neurons were exposed to Cd for 24 h, [Ca²⁺]_i elevation was significantly reduced compared to the vehicle-pretreated cells (data not shown), and demonstrated that celastrol hindered [Ca²⁺]_i-mediated CaMKII phosphorylation, thereby preventing Cd from activation of Akt/mTOR pathway and neuronal apoptosis (data not shown). Undoubtedly, more studies are needed to address whether celastrol prevents neuronal cell death also by alleviating Cd-induced ER stress.

Recently, we have revealed that that activated JNK may feedback positively mediate Cd-induced ROS in neuronal cells (Xu et al. 2016). In the current study, we noticed that pharmacological inhibition of JNK with SP600125 or overexpression of dominant negative c-Jun reinforced the inhibitory effects of celastrol on Cd-upregulated NOX2, p22^phox, p67^phox, p40^phox, p47^phox, and Rac1, as well as ROS generation in PC12 cells and/or primary neurons (Fig. 6a–f). Furthermore, overexpression of PP5 also strengthened celastrol’s suppression of Cd-induced events (Fig. 6g–i), suggesting that downregulated PP5 or activated JNK may feedback upregulate Cd-stimulated NOX2 family members and ROS. The findings enhance our understanding of celastrol’s intervention in Cd-stimulated ROS-generating enzyme NOX2 and its regulatory proteins, which is critical for preventing Cd inactivation of PP5, activation of JNK pathway, as well as neuronal apoptosis. As cellular ROS homeostasis is tightly controlled by the ROS-generating and -eliminating systems (Trachootham et al. 2009), further research is needed to determine whether celastrol attenuates Cd-induced ROS also by enhancing the function of ROS-eliminating systems in neuronal cells. Moreover, it would be interesting to unveil how pharmacological deactivation of JNK, or genetic modulation to activate PP5 or to deactivate c-Jun strengthens celastrol’s inhibition of Cd-induced expression of these ROS-generating proteins and production of ROS in neuronal cells.

In summary, here we have shown that celastrol prevented Cd from upregulating the expression of ROS-generating NOX2 and its regulatory proteins, thus attenuating Cd-induced ROS. This prevented Cd downregulation of PP5, thereby suppressing Cd activation of JNK-dependent apoptosis in neuronal cells (Fig. 7). Our results highlight a beneficial role of celastrol in the prevention of Cd-induced oxidative stress and neurodegenerative diseases.

Fig. 7.

A schematic diagram showing the neuroprotective effect of celastrol on Cd-induced apoptosis in neuronal cells. Celastrol ameliorated Cd-elicited neuronal apoptosis by preventing Cd from upregulation of ROS-generating NOX2 and its regulatory proteins (p22^phox, p40^phox, p47^phox, p67^phox, and Rac1), thus suppressing ROS inactivation of PP5 and activation of JNK pathway.

Abbreviations

AD

Alzheimer disease

ALS

amyotrophic lateral sclerosis

Cd

cadmium

DMEM

Dulbecco’s Modified Eagle’s Medium

FBS

fetal bovine serum

HD

Huntington’s disease

JNK

c-Jun N-terminal kinase

MAPK

mitogen-activated protein kinase

NAC

N-acetyl-L-cysteine

NOX2

NADPH oxidase 2

PBS

phosphate buffered saline

PD

Parkinson disease

PDL

poly-D-lysine

PP5

protein phosphatase 5

ROS

reactive oxygen species

TUNEL

the terminal deoxynucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate (dUTP) nick-end labeling