Mitochondrial Hydrogen Peroxide Activates PTEN and Inactivates Akt Leading to Autophagy Inhibition-Dependent Cell Death in Neuronal Models of Parkinson’s Disease

Abstract

Defective autophagy relates to the pathogenesis of Parkinson’s disease (PD), a typical neurodegenerative disease. Our recent study has demonstrated that PD toxins (6-OHDA, MPP⁺, or rotenone) induce neuronal apoptosis by impeding the AMPK/Akt-mTOR signaling. Here, we show that treatment with 6-OHDA, MPP⁺, or rotenone triggered decreases of ATG5/LC3-II and autophagosome formation with a concomitant increase of p62 in PC12, SH-SY5Y cells, and primary neurons, suggesting inhibition of autophagy. Interestingly, overexpression of wild-type ATG5 attenuated the inhibitory effect of PD toxins on autophagy, reducing neuronal apoptosis. The effects of PD toxins on autophagy and apoptosis were found to be associated with activation of PTEN and inactivation of Akt. Overexpression of dominant negative PTEN, constitutively active Akt and/or pretreatment with rapamycin rescued the cells from PD toxins-induced downregulation of ATG5/LC3-II and upregulation of p62, as well as consequential autophagosome diminishment and apoptosis in the cells. The effects of PD toxins on autophagy and apoptosis linked to excessive intracellular and mitochondrial hydrogen peroxide (H₂O₂) production, as evidenced by using a H₂O₂-scavenging enzyme catalase, a mitochondrial superoxide indicator MitoSOX and a mitochondria-selective superoxide scavenger Mito-TEMPO. Furthermore, we observed that treatment with PD toxins reduced the protein level of Parkin in the cells. Knockdown of Parkin alleviated the effects of PD toxins on H₂O₂ production, PTEN/Akt activity, autophagy, and apoptosis in the cells, whereas overexpression of wild-type Parkin exacerbated these effects of PD toxins, implying the involvement of Parkin in the PD toxins-induced oxidative stress. Taken together, the results indicate that PD toxins can elicit mitochondrial H₂O₂, which can activate PTEN and inactivate Akt leading to autophagy inhibition-dependent neuronal apoptosis, and Parkin plays a critical role in this process. Our findings suggest that co-manipulation of the PTEN/Akt/autophagy signaling by antioxidants may be exploited for the prevention of neuronal loss in PD.

Introduction

Parkinson’s disease (PD) is a neurodegenerative movement disorder characterized by the loss of dopaminergic neurons in the substantia nigra pars compacta (SNpc) [1, 2]. To understand the molecular mechanism of neuronal cell death in PD for development of effective neuroprotective therapies, numerous studies have been carried out, using postmortem brains, experimental cell/animal models of PD [3–5]. Increasing evidence has pointed to the association of neuronal loss with apoptotic cell death triggered by oxidative stress, impairment of mitochondrial respiration, and abnormal protein aggregation in PD [6–9]. Especially, recent findings have shown that defective autophagy is a hallmark of neurodegenerative diseases in which misfolded proteins or dysfunctional mitochondria accumulate in neurons [10]. Defective autophagy relates to the pathogenesis of PD [2, 7, 10, 11]. However, how the defective autophagy initiates and contributes to neuronal cell loss in the context of PD remains to be determined.

Autophagy, as a highly conserved homeostasis mechanism from yeast to human, delivers cytoplasmic nonfunctional or unwanted organelles and aggregate-prone and oxidized proteins to lysosomes for degradation and recycling [10, 12, 13]. Numerous reports have indicated that autophagy plays a crucial role in maintaining cellular homeostasis and protects cells from varying insults, which is particularly pivotal in neuronal survival [6, 14, 15]. However, autophagy is in fact a double-edged sword in the pathogenesis of many human diseases [14]. In many circumstances, autophagy can be adaptive to stimuli and promote cell survival, yet under certain conditions, it can also lead to cell death [14, 16]. Phosphatase and tensin homologue on chromosome 10 (PTEN), a lipid/protein dual phosphatase, antagonizes the phosphatidylinositol 3-kinase (PI3K) and is a negative regulator of Akt, controlling cell proliferation/growth, survival, and apoptosis in cells [17–20]. Though inhibition of the Akt pathway due to activation of PTEN promotes autophagy in many cases [21, 22], inactivation of Akt can also block autophagic process through impairing autophagic flux, thereby leading to cell death in skeletal myoblast cells [23]. Of note, multiple studies have documented that excessive reactive oxygen species (ROS)-induced neuronal apoptosis links to dysfunction of PTEN, Akt, and/or autophagy signaling [24–27]. For example, PTEN is involved in ROS production and neuronal death in in vitro models of stroke and PD [26]. ROS can activate Akt by inactivating PTEN, but ROS can also inhibit Akt at high concentrations [27]. Defective autophagy leads to oxidative stress and lysosomal rupture, triggering different types of cell death [24]. Our recent studies have demonstrated that PD toxins (6-OHDA, MPP⁺ or rotenone) induce neuronal apoptosis by induction of hydrogen peroxide (H₂O₂), impeding the AMPK/Akt-mTOR signaling [25]. Based on the above findings, we hypothesized that PD stress may inhibit autophagy via H₂O₂-mediated PTEN-Akt signaling pathway, thereby leading to neuronal apoptosis.

Parkin, an E3 ubiquitin ligase mainly present in the cytoplasm, is not only a stress-protective protein, but also a stress-inducible protein [28]. Parkin can provide specificity for degradation of accumulated proteins through the classic ubiquitin proteasome system [29–32]. Parkin dysfunction will lose its E3 ubiquitin ligase activity, which leads to abnormal accumulation of various proteins, and thus triggers neuronal cell death [33]. Merged data have pointed out that loss-of-function of Parkin results in mitochondrial turnover/dysfunction and dopaminergic neuronal loss involved in oxidative stress, which is a core pathogenic process in PD [32, 34, 35]. Especially, Parkin mutations can cause defective autophagy on dysfunctional mitochondria, known as mitophagy in PD [32, 34]. This prompted us to test whether downregulation or upregulation of Parkin impedes H₂O₂-PTEN/Akt/autophagy signaling in PD stress.

Here we demonstrate that PD toxins-induced mitochondrial H₂O₂ activates PTEN and inactivates Akt leading to autophagy inhibition-dependent neuronal apoptosis, and Parkin plays a critical role in this process. The results provide new insights into the mechanism behind the neuronal cell death in PD. Our findings suggest that co-manipulation of the PTEN/Akt/autophagy signaling by antioxidants may be exploited for the prevention of neuronal loss in PD.

Materials and Methods

Reagents

Rotenone (Cat. No. R8875), 6-hydroxydopamine (6-OHDA) (Cat. No. 162957), 1-methyl-4-phenylpyridin-1-ium (MPP⁺) (Cat. No. M0896), 2′7′-dichlorodihydrofluorescein diacetate (H₂DCFDA) (Cat. No. D6883), catalase (CAT) (Cat. No. C9322), poly-D-lysine (PDL) (Cat. No. P6407), 4′,6-diamidino-2-phenylindole (DAPI) (Cat. No. D8417), and protease inhibitor cocktail (Cat. No. P8340) were purchased from Sigma (St Louis, MO, USA). Mito-TEMPO (Cat. No. ALX-430–171-M005) and rapamycin were from ALEXIS Biochemicals Corporation (San Diego, CA, USA). Dulbecco’s modified Eagle’s medium (DMEM) (Cat. No. 30030), 0.05% Trypsin-EDTA (Cat. No. 25300062), NEUROBASAL^™ Media (Cat. No. 25300062), and B27 Supplement (Cat. No. 17504044) were from Invitrogen (Grand Island, NY, USA). Fetal bovine serum (FBS) (Cat. No. SH30022) and horse serum were from Hyclone (Logan, UT, USA). MitoSox (Cat. No. 40778ES) was bought from YEASEN (Shanghai, China). CellTiter 96^®AQ_ueous One Solution (Cat. No. MR1007) Cell Proliferation Assay kit was from Promega (Madison, WI, USA). Enhanced chemiluminescence solution (Cat. No. MR1004) was from Sciben Biotech Company (Nanjing, China). The following antibodies were used: p-Akt (Thr308) (Cat. No. 9275), p-Akt (Ser473) (Cat. No. 9271), cleaved caspase-3 (Cat. No. 9664) (Cell Signaling Technology, Beverly, MA, USA), p-PTEN (Thr366) (Cat. No. 2195–1), PTEN (Cat. No. 5171–1) (Epitomics, Burlingame, CA, USA), LC3B (Cat. No. L7543), ATG5 (Cat. No. SAB5700062), SQSTM1/p62 (Cat. No. P0067) (Sigma), Akt (Cat. No. AT10694), Parkin (Cat. No. AT20255), β-tubulin (Cat. No. AT1001), HA (Cat. No. AT16321) (Sciben Biotech Company), goat anti-rabbit IgG-horseradish peroxidase (HRP) (Cat. No. 31402), and goat anti-mouse IgG-HRP (Cat. No. 31431) (Pierce, Rockford, IL, USA). Other chemicals were from local commercial sources and were of analytical grade, unless stated elsewhere.

Cell Culture

Rat pheochromocytoma (PC12) (Cat. No. CRL-1721) and human neuroblastoma SH-SY5Y (Cat. No. CRL-2266) cell lines were from American Type Culture Collection (ATCC) (Manassas, VA, USA), which were seeded in a 6-well plate (5 × 10⁵ cells/well) or 96-well plate (1 × 10⁴ cells/well) pre-coated with (for PC12) or without (for SH-SY5Y) PDL (0.2 μg/ml). PC12 cells were cultured in antibiotic-free DMEM supplemented with 10% horse serum and 5% FBS, whereas SH-SY5Y cells were grown in antibiotic-free DMEM supplemented with 10% FBS, in a humidified incubator (37 °C, 5% CO₂). Primary murine neurons were isolated from fetal mouse cerebral cortexes of 16–18 days of gestation in female ICR mice as described [36], and seeded in a PDL (10 μg/ml)-coated 6-well plate (5 × 10⁵ cells/well) or 96-well plate (1 × 10⁴ cells/well) for experiments after 6 days of culture. The experiments involving animals in this study were approved by the Institutional Animal Care and Use Committee of Nanjing Normal University (Certificate No. 200408), and were conducted 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

Recombinant adenovirus expressing human dominant negative PTEN (Ad-PTEN-C/S), and the control adenovirus expressing β-galactosidase (Ad-LacZ) were described previously [37, 38]. Recombinant adenovirus encoding HA-tagged myristoylated, constitutively active Akt (Admyr-Akt) was generously provided by Dr. Kenneth Walsh (Boston University School of Medicine, Boston, MA, USA) [39]. For experiments, PC12 cells were cultured in the growth medium, and infected with the individual adenovirus for 24 h at 5 of multiplicity of infection (MOI = 5). Subsequently, the cells were used for experiments and the cells infected with Ad-LacZ alone served as a control. Expression of HA-tagged myr-Akt was determined by Western blotting with anti-HA antibody.

Lentiviral Cloning, Production, and Infection

To generate lentiviral shRNA to Parkin, oligonucleotides containing the target sequences were synthesized, annealed, and inserted into FSIPPW lentiviral vector via the EcoR1/BamH1 restriction enzymes [40]. Lentiviral shRNA to GFP (for control) was produced as described [41]. To make FLAG-tagged wild-type Parkin (FLAG-Parkin) construct (for Parkin overexpression), wild-type ATG5 (FLAG-ATG5) construct (for ATG5 overexpression) and EGFP construct (for control), PCR template for Parkin, ATG5 or EGFP was from PC12 cells’ cDNA generated by RT-PCR using PrimeScript II 1st Strand cDNA Synthesis Kit (Takara Bio, Kusatsu, Japan) and plasmid PX458 (Addgene, Cambridge, MA, USA), respectively. The PCR products of FLAG-Parkin, FLAG-ATG5, and EGFP were cloned into pSin4-EF2-IRES-Pur vector via EcoRI/BamHI double-digestion. The above primers used are listed in Table 1. To generate lentivirus, the above constructed plasmids were co-transfected together with pMD2.G and psPAX2 (Addgene, Cambridge, MA, USA) to 293TD cells using MegaTran 1.0 reagent (OriGene Technologies, Rockville, MD, USA). Each supernatant containing viral particles was collected 48 h and 60 h post-transfection and filtered through a 0.45-μm filter, and stored at − 80 °C. For use, a monolayer of PC12 cells, when grown to about 70% confluency, were infected with the corresponding lentivirus-containing medium in the presence of 8 μg/ml polybrene for 12 h, and reinfected after 6 h. The transduced cells were selected by 48-h treatment with 2 μg/ml puromycin. After 5 days of culture, the cells were used for experiments.

Table 1.

The sequences of oligonucleotides for Parkin, EGFP, FLAG-ATG5, and FLAG-Parkin

Name	Sense	Anti-sense
Parkin	5′-AATTCCCATCACCTGACAGTACAGAACTTGCAAGAGAAGTTCTGTACTGTCAGGTGATTTTTTG-3′	5′-GATCCAAAAAATCACCTGACAGTACAGAACTTCTCTTGCAAGTTCTGTACTGTCAGGTGATGGG-3′
EGFP	5′-CCGGAATTCATGGTGAGCAAGGGCGAGGAGCT-3′	5′-CGCGGATCCGTTACTTGTACAGCTCGTCCATG-3′
FFAG-ATG5	5′-CGGAATTCATGGATTACAAGGATGACGACGATAAGATGACAGATGACAAAGATGTGC-3′	5′-CGGGATCCGTTAGGAGATCTCCAAGGGTATG-3′
FFAG-Parkin	5′-CCGGAATTCATGGATTACAAGGATGACGACGATAAGATGATAGTGTTTGTCAGGTT-3′	5′-GGACTAGTCCCTACACGTCAAACCAGTGATCACCC-3′

GFP-LC3 Assay

PC12, SH-SY5Y cells and primary neurons, PC12 cells infected with lentiviral FLAG-Parkin, FLAG-ATG5 or EGFP, or with lentiviral shRNA to Parkin or GFP, or PC12 cells infected with Ad-PTEN-C/S, Ad-myr-Akt, and/or Ad-LacZ, respectively, were infected with Ad-GFP-LC3 and seeded in a 6-well plate (5 × 10⁵ cells/well) containing a PDL-coated or PDL-uncoated glass coverslip per well. Next day, cells were treated with/without 6-OHDA (30–240 or 120 μM), MPP⁺ (0.5 and/or 1 mM), or rote-none (0.5 and/or 1 μM) for 24 h, or treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h following pre-incubation with/without rapamycin (100 ng/ml) for 2 h, or a H₂O₂-scavenging enzyme CAT (350 U/ml) or a mitochondria-targeted antioxidant Mito-TEMPO (10 μM) for 1 h, with 5 replicates of each treatment. Afterwards, the cells on the coverslips were fixed with 4% paraformaldehyde in PBS for 30 min at 4 °C and then washed 3 times with PBS, followed by photographing under a fluorescence microscope (Leica DMi8, Wetzlar, Germany) equipped with a digital camera and counting the numbers of GFP-LC3 puncta (green) per cell to estimate autophagosome formation. At least 50 cells were scored in each experiment.

Cell Viability Assay

The above indicated cells, respectively, were seeded and cultured in a PDL-coated 96-well plate (1 × 10⁴ cells/well). Next day, cells were treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h, or treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h following pre-incubation with/without rapamycin (100 ng/ml) for 2 h, with 5 replicates of each treatment. Subsequently, MTS reagent (one solution reagent) (20 μl/well) was added and incubated for an additional 3 h. Finally, the values of optical density (OD) at 490 nm were determined using a Victor X3 Light Plate Reader (PerkinElmer, Waltham, MA, USA).

DAPI and TUNEL Staining

The above indicated cells, respectively, were seeded and cultured in a 6-well plate (5 × 10⁵ cells/well) containing a PDL-coated or PDL-uncoated glass coverslip per well. Next day, cells were treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h, or treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h following pre-incubation with/without rapamycin (100 ng/ml) for 2 h, or CAT (350 U/ml) or Mito-TEMPO (10 μM) for 1 h, with 5 replicates of each treatment.

Then, the apoptotic cells with fragmented and condensed nuclei were monitored using DAPI staining as described [42]. For TUNEL staining, TUNEL reaction mixture (TdT enzyme solution and labeling solution) was added according to the manufacturer’s protocol of In Situ Cell Death Detection Kit^® (Cat. No. A111-03) (Roche, Mannheim, Germany). Finally, slides were mounted in glycerol/phosphate-buffered saline (PBS) (1:1, v/v) containing 2.5% 1,4-diazabiclo-(2,2,2) octane. Photographs were taken under a fluorescence microscope (200 ×) (Leica DMi8, Wetzlar, Germany) equipped with a digital camera. For quantitative analysis of the fluorescence intensity using TUNEL staining, the integral optical density (IOD) for 200–300 cells per graph was determined by Image-Pro Plus 6.0 software (Media Cybernetics Inc., Newburyport, MA, USA).

Immunofluorescence and Imaging

PC12, SH-SY5Y cells, and primary neurons, respectively, were seeded in a 6-well plate (5 × 10⁵ cells/well) containing a PDL-coated or PDL-uncoated glass coverslip per well. Next day, cells were treated with/without 6-OHDA (30–240 μM), MPP⁺ (0.5 and 1 mM), or rotenone (0.5 and 1 μM) for 24 h, with 5 replicates of each treatment. Then, the cells on the coverslips were fixed with 4% paraformaldehyde and blocked with 3% normal goat serum with 0.3% Triton X-100 for 1 h, incubated with anti-phospho-PTEN (Ser366) antibody (1:50, diluted in PBS containing 1% BSA) overnight at 4 °C. After incubation, coverslips were washed 3 × 5 min with PBS and then incubated with FITC-conjugated goat anti-rabbit IgG (Santa Cruz Biotechnology, Dallas, TX, USA, 1:500, diluted in PBS containing 1% BSA) for 1 h at room temperature. After three rinses, slides were fixed and imaged as described [43]. The IOD for fluorescence intensity was quantitatively analyzed as described above.

Intracellular H₂O₂ and Mitochondrial ROS Imaging

According to the information provided by the supplier, H₂DCFDA and MitoSOX are able to trace intracellular H₂O₂ and mitochondrial superoxide levels, respectively. H₂DCFDA is a stable non-fluorescent probe with peroxide-selective dye that can passively diffuse into the intracellular matrix of cells, where it is sheared by esterase and oxidized by H₂O₂, forming fluorescent DCF [44]. MitoSOX is a superoxide indicator dye that can specifically recognize mitochondrial superoxide and produce red fluorescence in live cells. In brief, PC12, SH-SY5Ycells and primary neurons, or PC12 cells infected with lentiviral shRNA to Parkin or GFP or with lentiviral FLAG-Parkin or EGFP, after treatment, were loaded with H₂DCFDA (20 μM) for 1 h or with MitoSOX (5 μM) for 10 min at 37 °C. Subsequently, all stained specimens were rinsed 3 times with PBS, followed by imaging under a fluorescence microscope, and quantitatively measuring IOD of the fluorescence intensity as described above.

Western Blot Analysis

The indicated cells, after treatments, were briefly washed with cold PBS, and then on ice, lysed in the radioimmunoprecipitation assay buffer. Afterwards, Western blotting was performed as described previously [36], and the blots for detected proteins were semi-quantified using NIH ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Statistical Analysis

All results were expressed as mean values ± standard error (mean ± SE). Student’s t-test for non-paired replicates was used to identify 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. The criterion for the statistical significance was p < 0.05.

Results

PD Toxins Induce Decreases of ATG5/LC3-II, Autophagosome Formation, and Parkin with a Concomitant Increase of p62 in Neuronal Cells

ATG5 and LC3 are two essential components for the canonical autophagy [45, 46]. The conversion of LC3-I to LC3-II of LC3 is a protein marker for autophagosome formation [45]. Additionally, p62 protein (a substrate that is degraded by autophagy), also called sequestosome 1 (SQSTM1), is commonly used as a marker for execution of autophagy [47]. To test autophagic manifestation in PD toxins-exposed neuronal cells, PC12, SH-SY5Y cells, and primary neurons were treated with/without 6-OHDA (30–240 μM), MPP⁺ (0.5 and 1 mM) or rotenone (0.5 and 1 μM) for 24 h, followed by analyzing the cellular protein levels of ATG5, LC3-II, and p62 using Western blotting. The results showed that treatment with 6-OHDA, MPP⁺, or rotenone triggered decreases of ATG5 and LC3-II in a concentration-dependent manner in the cells (Fig. 1A–D). Interestingly, the toxins elicited robust expression of p62 protein dose-dependently in the cells as well (Fig. 1A–D). Sequentially, we monitored neuronal autophagic vacuoles with GFP-LC3 localization. When PC12, SH-SY5Y cells, and primary neurons, infected with Ad-GFP-LC3, were exposed to the toxins for 24 h, autophagic vacuoles with GFP-LC3 (in green) significantly decreased in a concentration-dependent fashion (Fig. 1E–H), implying that PD toxins impaired autophagosome formation. Of note, treatment with 6-OHDA, MPP⁺, or rotenone also reduced the protein level of Parkin dose-dependently in the cells (Fig. 1A and B). Collectively, these results indicate that treatment with PD toxins induces decreases of Atg5/LC3-II, autophagosome formation, and Parkin with a concomitant increase of p62 in neuronal cells, suggesting inhibition of autophagy.

Fig. 1.

PD toxins-induced decreases of ATG5/LC3-II, autophagosome formation, and Parkin with a concomitant increase of p62 in neuronal cells. PC12, SH-SY5Y cells, and primary neurons infected with/without Ad-GFP-LC3, respectively, were treated with/without 6-OHDA (30–240 μM), MPP⁺ (0.5 and 1 mM) or rotenone (0.5 and 1 μM) for 24 h. A, C Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. B, D The relative densities for ATG5, LC3-II, p62, Parkin to β-tubulin were semi-quantified using NIH ImageJ. E, G Representative GFP-LC3 puncta imaging (in green) in the cells was shown by using GFP-LC3 assay. Scale bar: 2 μm. F, H The number of GFP-LC3 puncta per cell was quantified. At least 50 cells were scored in each experiment. All data were expressed as mean ± SE (n = 3 for B, D; n = 5 for F, H). Using one-way ANOVA, *p < 0.05, **p < 0.01, difference vs control group

Overexpression of ATG5 Attenuates PD Toxins-Induced Autophagy Inhibition and Apoptosis in Neuronal Cells

Deletion of ATG5 can completely inhibit autophagy [46]. We have observed that ATG5 protein level decreased in PD toxins-induced neuronal cells (Fig. 1). To validate the importance of ATG5 in PD toxins-induced autophagy inhibition and neuronal apoptosis, ATG5 in PC12 cells was overexpressed. As shown in Fig. 2A, infection with lentiviral FLAG-tagged wild-type ATG5 (FLAG-ATG5), but not lentiviral EGFP (control), resulted in a robust expression of FLAG-tagged ATG5, regardless of absence or presence of 6-OHDA, MPP⁺, or rotenone. Interestingly, overexpression of ATG5 conferred high resistance to decreases of LC3-II and Parkin, as well as increases of p62 and cleaved caspase-3 in PC12 cells treated with 6-OHDA, MPP⁺, or rotenone (Fig. 2 A and B). Consistent with this, overexpression of ATG5 potently blocked the toxins-induced autophagosome decline, cell viability reduction and apoptosis in the cells, as evidenced by GFP-LC3 assay (Fig. 2C), MTS assay (Fig. 2D), DAPI staining (Fig. 2E and F), and TUNEL staining (Fig. 2G and H). The results demonstrate that PD toxins-downregulated ATG5 contributes to PD toxins-induced autophagy inhibition-dependent apoptosis in neuronal cells.

Fig. 2.

Overexpression of ATG5 prevents PD toxins from inducing autophagy inhibition and apoptosis in neuronal cells. PC12 cells, infected with lentiviral FLAG-ATG5 or EGFP (as control) and infected with/without Ad-GFP-LC3, respectively, were treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. A Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. B The relative densities for ATG5, LC3-II, p62, Parkin, cleaved caspase-3 to β-tubulin were semi-quantified using NIH ImageJ. C The number of GFP-LC3 puncta per cell was quantified by GFP-LC3 assay. At least 50 cells were scored in each experiment. D The relative cell viability was determined by the MTS assay. E, G Apoptotic cells were evaluated by using DAPI staining for nuclear fragmentation and condensation (arrows) and TUNEL staining for fragmented DNA (in green), respectively. Scale bar: 20 μm. F, H The percentage of cells with fragmented nuclei and IOD values of TUNEL-positive cells were quantified. All data were expressed as mean ± SE (n = 3 for B; n = 5 for C, D, F, H). Using one-way ANOVA or Student’s t-test, ^ap < 0.05, difference vs control group; ^bp < 0.05, FLAG-ATG5 group vs EGFP control group

PD Toxins Activates PTEN and Inactivates Akt, Resulting in Autophagy Inhibition and Apoptosis in Neuronal Cells

It is well known that PTEN antagonizes PI3K by catalyzing phosphatidylinositol (3,4,5)-trisphosphate (PIP3) to phosphatidylinositol 4,5-bisphosphate (PIP2), thereby inhibiting Akt, regulating cell proliferation/growth, survival, and apoptosis [17–20]. Akt inactivation can block autophagic process via impairing autophagic flux, causing cell death [23]. Our previous studies have found that PD toxins inactivate Akt signaling, leading to neuronal apoptosis [48]. Therefore, we reasoned that PTEN might be involved in PD toxins-induced Akt inactivation, autophagy inhibition, and apoptosis in neuronal cells. To this end, we checked the phosphorylation status of PTEN and Akt in PC12, SH-SY5Y cells, and primary neurons exposed to 6-OHDA (30–240 μM), MPP⁺ (0.5 and 1 mM), or rotenone (0.5 and 1 μM) for 24 h. Western blot analysis showed that treatment with these toxins repressed the phosphorylation of PTEN (Thr366) and Akt (Thr308 and Ser473) in the cells dose-dependently (Fig. 3A–D). Similarly, our immunofluorescence staining also showed a dose-dependent decrease in p-PTEN (Thr366) (in green) in the cells treated with 6-OHDA, MPP⁺, or rotenone (Fig. 3E–H). Additionally, an experiment of different time points (0, 6, 12, and 24 h) after the PD toxin treatment was conducted to see time-dependent manifestations of PTEN, Akt, ATG5, LC3-II, and p62 in PC12 cells and primary neurons. The results showed that decreased p-PTEN (Thr366) in the cells was seen starting in 6–12 h of postexposure to 6-OHDA, MPP⁺, or rotenone and became more profound in 24 h, whereas decreased p-Akt (Ser473 and Thr308)/ATG5/LC3-II and increased p62 were observed in 12–24 h of postexposure of the cells to the PD toxins (Fig. 3I and J), clearly indicating that p-PTEN changes earlier than p-Akt and the autophagy flux do, and PTEN is the upstream factor. Taken together, the findings imply that PD toxins activate PTEN and inactivate Akt in neuronal cells.

Fig. 3.

PD toxins induce activation of PTEN and subsequent inactivation of Akt/inhibition of autophagy in neuronal cells. PC12, SH-SY5Y cells, and/or primary neurons, respectively, were treated with/without 6-OHDA (30–240 μM or 120 μM), MPP⁺ (0.5 and/or 1 mM), or rotenone (0.5 and/or 1 μM) for 6, 12, and/or 24 h. A, C, I Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. B, D, J The relative densities for p-PTEN (Thr366), p-Akt (Ser473), p-Akt (Thr308), ATG5, LC3-II, p62 to β-tubulin were semi-quantified using NIH ImageJ. E, G Expression of p-PTEN (Thr366) was stained and imaged using immunofluorescence, showing that treatment of the cells with PD toxins for 24 h caused lower p-PTEN expression (in green). Scale bar: 20 μm. F, H IOD for fluorescence intensity of p-PTEN expression was determined. All data were expressed as mean ± SE (n = 3 for B, D, J; n = 5 for F, H). Using one-way ANOVA, *p < 0.05, **p < 0.01, difference vs control group

Next, PC12 cells, infected with Ad-PTEN-C/S, Ad-PTEN-C/S/Ad-myr-Akt, Ad-LacZ (as control), and/or Ad-GFP-LC3, were exposed to 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. We showed that infection with Ad-PTEN-C/S and Ad-myr-Akt, but not Ad-LacZ, increased the total protein levels of PTEN and Akt, respectively (Fig. 4A). Overexpression of PTEN-C/S in PC12 cells increased the basal level of p-Akt and rendered remarkable resistance to 6-OHDA, MPP⁺ or rotenone-induced dephosphorylation of Akt, decreases of ATG3/LC3-II/autophagosomes, increases of p62 and cleaved caspase-3 (Fig. 4A–C, Fig. S1A), as well as cell viability reduction and apoptosis (Fig. 4D and E). Of importance, the toxins-triggered events were ameliorated in the cells co-infected with Ad-PTEN-C/S/Ad-myr-Akt more potently than those infected with Ad-PTEN-C/S (Fig. 4A–E, Fig. S1A) or Ad-myr-Akt (data not shown) alone. The results suggest that the PTEN/Akt pathway mediates PD toxin-induced neuronal autophagy impairment and death. However, overexpression of PTEN-C/S and/or myr-Akt failed to rescue PD toxins-induced decrease of Parkin in PC12 cells (Fig. 4A and B), implying that the PTEN/Akt signaling does not mediate the expression of Parkin in PD toxins-induced neuronal cells.

Fig. 4.

PD toxins-induced activation of PTEN and inactivation of Akt contribute to autophagy inhibition and apoptosis in neuronal cells. PC12 cells, infected with Ad-PTEN-C/S, Ad-myr-Akt, and/or Ad-LacZ (as control) and infected with/without Ad-GFP-LC3, respectively, were treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h, or pretreated with/without rapamycin (100 ng/ml) for 2 h and then treated with/without 6-OHDA, MPP⁺ or rotenone for 24 h. A, F Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. B, G The relative densities for p-PTEN (Thr366), p-Akt (Ser473), p-Akt (Thr308), ATG5, LC3-II, p62, Parkin, cleaved caspase-3 to β-tubulin were semi-quantified using NIH ImageJ. C, H The number of GFP-LC3 puncta per cell was quantified by GFP-LC3 assay. At least 50 cells were scored in each experiment. D, I The relative cell viability was determined by the MTS assay. E, J Apoptotic cells were evaluated by nuclear fragmentation and condensation using DAPI staining. All data were expressed as mean ± SE (n = 3 for B, G; n = 5 for C, D, H–J). Using one-way ANOVA or Student’s t-test, ^ap < 0.05, difference vs control group; ^bp < 0.05, Ad-PTEN-C/S group or Ad-PTEN-C/S + Ad-myr-Akt group vs Ad-LacZ group; ^cp < 0.05, Ad-PTEN-C/S + Ad-myr-Akt grouo vs Ad-PTEN-C/S group; ^dp < 0.05, Ad-myr Akt group or Admyr-Akt + Rapamycin group vs Ad-LacZ group; ^ep < 0.05, Ad-myr-Akt + Rapamycin group vs Ad-myr-Akt group

To define the role of Akt in PD toxins-induced autophagy inhibition and apoptosis in neuronal cells, PC12 cells, infected with Ad-myr-Akt, Ad-LacZ, and/or Ad-GFP-LC3, were exposed to 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h following pre-incubation with/without rapamycin (100 ng/ml), a known autophagy inducer [49], for 2 h. As shown in Fig. 4E, expression of high HA-tagged Akt mutant was seen in PC12 cells infected with Ad-myr-Akt, but not in the cells infected with Ad-LacZ (control virus). Overexpression of myr-Akt significantly prevented 6-OHDA-, MPP⁺-, or rotenone-induced dephosphorylation of Akt, decreases of ATG3/LC3-II/autophagosomes, increases of p62 and cleavage of caspase-3 (Fig. 4F–H, Fig. S1B), as well as cell viability reduction and apoptosis (Fig. 4I and J), but not decrease of Parkin (Fig. 4F and G). Intriguingly, pretreatment of the cells with rapamycin possessed more powerful inhibitory effects on the PD toxins-induced events than ectopic expression of myr-Akt alone (Fig. 4F–J, Fig. S1B). Of interest, pretreatment with rapamycin rescued the cells from decrease of Parkin in response to 6-OHDA, MPP⁺, or rotenone (Fig. 4F and G), implying that Parkin protein level is regulated through an autophagy-dependent mechanism. Taken together, our findings underscore the concept that PD toxins induce autophagy inhibition and consequential apoptosis in neuronal cells, in part, by activation of PTEN and inactivation of Akt.

PD Toxins-Induced Intracellular and Mitochondrial H₂O₂ Activates PTEN and Inactivates Akt Contributing to Autophagy Inhibition and Apoptosis in Neuronal Cells

Our group has recently found that PD toxins’ induction of H₂O₂ impedes the AMPK/Akt-mTOR signaling pathway contributing to cell death in neuronal cells [25]. To validate whether PD toxins-impaired PTEN/Akt signaling leading to autophagy inhibition and neuronal apoptosis is due to induction of intracellular H₂O₂, we used catalase (CAT), a H₂O₂-scavenging enzyme. The results showed that pretreatment with CAT markedly mitigated the toxins-induced H₂O₂ production in PC12, SH-SY5Y cells and primary neurons, as evidenced by using a peroxide-selective probe H₂DCFDA for imaging and quantifying (Fig. 5A and B). Of importance, CAT rescued the cells from PD toxins-elicited decreases in p-PTEN, p-Akt, ATG5, LC3-II, and Parkin and increase in p62 (Fig. 5C and D). CAT also profoundly blocked the diminishment of autophagosomes, the activation of caspase-3 and the increase of fragmented nuclei in the cells exposed to the PD toxins (Fig. 5C–F, Fig. S2). The findings support that the PD toxins induce intracellular H₂O₂, which mediates autophagy inhibition leading to apoptosis by activating PTEN and inactivating Akt in neuronal cells.

Fig. 5.

PD toxins elicit intracellular H₂O₂, resulting in activation of PTEN and inactivation of Akt contributing to autophagy inhibition and apoptosis in neuronal cells. PC12, SH-SY5Y cells and primary neurons were pretreated with/without CAT (350 U/ml) for 1 h and then exposed to 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. A Cell H₂O₂ was imaged using a peroxide-selective probe H₂DCFDA. Scale bar: 20 μm. B IOD for cell H₂O₂ fluorescence intensity was quantitatively analyzed. C Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. D The relative densities for p-PTEN (Thr366), p-Akt (Ser473), p-Akt (Thr308), ATG5, LC3-II, p62, Parkin, cleaved caspase-3 to β-tubulin were semi-quantified using NIH ImageJ. E The number of GFP-LC3 puncta per cell was quantified by GFP-LC3 assay. At least 50 cells were scored in each experiment. F Apoptotic cells were evaluated by nuclear fragmentation and condensation using DAPI staining. All data were expressed as mean ± SE (n = 3 for D; n = 5 for B, E, F). Using oneway ANOVA or Student’s t-test, ^ap < 0.05, difference vs control group; ^bp < 0.05, + CAT group vs—CAT group

Next, we investigated whether the effects of PD toxins on PTEN/Akt signaling leading to autophagy inhibition and neuronal apoptosis are associated with excessive H₂O₂ production in the mitochondria of neuronal cells. For this, PC12, SH-SY5Y cells, and primary neurons were pretreated with/without a mitochondria-targeted antioxidant Mito-TEMPO (10 μM) for 1 h, and then exposed to 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. We found that pretreatment with Mito-TEMPO obviously attenuated the PD toxins-induced generation of H₂O₂ in the cells (Fig. 6A and Fig. S3A). MitoSOX, a mitochondrial superoxide indicator, was used to define mitochondrial ROS, exhibiting that the increases of MitoSOX red fluorescence were obviously diminished by Mito-TEMPO in the cells (Fig. 6B and C), clearly indicating the PD toxins’ induction of mitochondrial ROS. Consistently, Mito-TEMPO substantially reversed the PD toxins-triggered decreases of p-PTEN, p-Akt, ATG5/LC3-II/autophagosomes and Parkin and increase of p62, as well as cleavage of caspase-3 and apoptosis in the cells (Fig. 6D–G, Fig. S3B). The results indicate that the PD toxins indeed evoke mitochondrial H₂O₂, which inhibits autophagy contributing to apoptosis by impairing the PTEN/Akt signaling in neuronal cells.

Fig. 6.

PD toxins evoke mitochondrial H₂O₂/ROS, resulting in activation of PTEN and inactivation of Akt contributing to autophagy inhibition and apoptosis in neuronal cells. PC12, SH-SY5Y cells and primary neurons were pretreated with/without Mito-TEMPO (10 μM) for 1 h and then exposed to 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. A Cell H₂O₂ was imaged and quantified using a peroxide-selective probe H₂DCFDA. B, C Mitochondrial ROS (in red) were imaged and quantified using a mitochondrial superoxide indicator MitoSOX. Scale bar: 20 μm. D Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. E The relative densities for p-PTEN (Thr366), p-Akt (Ser473), p-Akt (Thr308), ATG5, LC3-II, p62, Parkin, cleaved caspase-3 to β-tubulin were semi-quantified using NIH ImageJ. F The number of GFP-LC3 puncta per cell was quantified by GFP-LC3 assay. At least 50 cells were scored in each experiment. G Apoptotic cells were evaluated by nuclear fragmentation and condensation using DAPI staining. All data were expressed as mean ± SE (n = 3 for D; n = 5 for A, C, F, G). Using one-way ANOVA or Student’s t-test, ^ap < 0.05, difference vs control group; ^bp < 0.05, + Mito-TEMPO group vs—Mito-TEMPO group

Parkin Exerts a Critical Role for PD Toxins-Induced H₂O₂ Production, PTEN Activation/Akt Inactivation, Autophagy Inhibition, and Apoptosis in Neuronal Cells

Parkin is not only a stress-protective protein but also a stress-inducible protein [28], which is an essential protein for degradation of accumulated proteins through the classic ubiquitin proteasome pathway [29–32]. Having observed that PD toxins-induced autophagy inhibition and apoptosis related to declined Parkin (Fig. 1 and Fig. 2), we postulated that silencing Parkin might potentiate PD toxins-elicited neuronal apoptosis. To test this, PC12 cells, infected with lentiviral shRNA to Parkin or GFP and/or Ad-GFP-LC3, were exposed to 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. As shown in Fig. 7A, lentiviral shRNA to Parkin, but not GFP, downregulated the protein expression of Parkin by ~90% in PC12 cells, as detected by Western blotting. To our surprise, knockdown of Parkin substantially elevated the basal and/or the PD toxins-reduced p-PTEN (Thr366), p-Akt (Thr308 and Ser473), ATG5 and LC3-II (Fig. 7A and B), and suppressed the PD toxins-induced increase of p62 and cleavage of caspase-3 (Fig. 7A and B). Silencing Parkin conferred significant resistance to the PD toxins-evoked decrease of autophagosomes and increase of H₂O₂ production (Fig. 7C and D, Fig. 4S), and attenuated the PD toxins-triggered cell viability reduction and apoptosis (Fig. 7E–G) in the cells as well. The results imply an important role of Parkin in PD toxins-induced decrease of autophagosome formation and increase of apoptosis in neuronal cells, and that depletion of Parkin attenuated the effects of PD toxins on the PTEN/Akt signaling.

Fig. 7.

Depletion of Parkin attenuates PD toxins-induced H₂O₂, PTEN activation/Akt inactivation, autophagy inhibition, and apoptosis in neuronal cells. PC12 cells, infected with lentiviral shRNA to Parkin or GFP (as control) and infected with/without Ad-GFP-LC3, respectively, were treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM) or rotenone (1 μM) for 24 h. A Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. B The relative densities for p-PTEN (Thr366), p-Akt (Ser473), p-Akt (Thr308), ATG5, LC3-II, p62, cleaved caspase-3 to β-tubulin were semi-quantified using NIH ImageJ. C The number of GFP-LC3 puncta per cell was quantified by GFP-LC3 assay. At least 50 cells were scored in each experiment. D Cell H₂O₂ was imaged and quantified using a peroxide-selective probe H₂DCFDA. E The relative cell viability was determined by the MTS assay. F Apoptotic cells were evaluated by using DAPI staining for nuclear fragmentation and condensation (arrows). Scale bar: 20 μm. G The percentage of cells with fragmented nuclei was quantified. All data were expressed as mean ± SE (n = 3 for B; n = 5 for C–E, G). Using one-way ANOVA or Student’s t-test, ^ap < 0.05, difference vs control group; ^bp < 0.05, Parkin shRNA group vs GFP shRNA group

To substantiate the role of Parkin in PD toxins-induced H₂O₂ induction, PTEN activation/Akt inactivation, autophagy inhibition, and neuronal apoptosis, we further constructed lentivirus FLAG-tagged wild-type Parkin (FLAG-Parkin) to overexpress Parkin. As shown in Fig. 8A and B, overexpression of Parkin in PC12 cells strengthened the basal and/or 6-OHDA-, MPP⁺-, or rotenone-induced decreases of p-PTEN, p-Akt, ATG5, and LC3-II and increases of p62 and cleaved caspase-3 (Fig. 8A and B). Consistently, overexpression of Parkin also reinforced the basal and/or the PD toxins-induced autophagosomes’ loss, excessive H₂O₂ generation, cell viability reduction, and apoptosis in the cells (Fig. 8C–F, Fig. 5S). Taken together, the results support the notion that Parkin plays a critical role for PD toxins-induced H₂O₂ production, PTEN activation/Akt inactivation, autophagy inhibition, and apoptosis in neuronal cells.

Fig. 8.

Overexpression of Parkin potentiates PD toxins-induced H₂O₂, PTEN activation/Akt inactivation, autophagy inhibition, and apoptosis in neuronal cells. PC12 cells, infected with lentiviral FLAG-Parkin or EGFP (as control) and infected with/without Ad-GFP-LC3, respectively, were treated with/without 6-OHDA (120 μM), MPP⁺ (1 mM), or rotenone (1 μM) for 24 h. A Total cell lysates were subjected to Western blotting with indicated antibodies. The blots were probed for β-tubulin as a loading control. Similar results were obtained in at least five independent experiments. B The relative densities for p-PTEN (Thr366), p-Akt (Ser473), p-Akt (Thr308), ATG5, LC3-II, p62, cleaved caspase-3 to β-tubulin were semi-quantified using NIH ImageJ. C The number of GFP-LC3 puncta per cell was quantified by GFP-LC3 assay. At least 50 cells were scored in each experiment. D Cell H₂O₂ was imaged and quantified using a peroxide-selective probe H₂DCFDA. E The relative cell viability was determined by the MTS assay. F Apoptotic cells were evaluated by nuclear fragmentation and condensation using DAPI staining. All data were expressed as mean ± SE (n = 3 for B; n = 5 for C–F). Using one-way ANOVA or Student’s t-test, ^ap < 0.05, difference vs control group; ^bp < 0.05, FLAG-Parkin group vs EGFP control group

Discussion

PD is a typical neurodegenerative disease characterized by the loss of dopaminergic nigrostriatal neurons [1, 2]. Deep brain stimulation in PD onsets represents a paradigmatic cross-talk between mammalian disease models and clinical evidence in humans [50]. The cellular and rodent models for PD toxins (6-OHDA, MPTP/MPP⁺, and/or rotenone) have contributed to understanding of the PD pathology [25, 48, 50, 51]. Data show that PD patients suffer only a 60–70% loss of SNc neuropils [50]. PD early stages are probably dominated by the subtle impairment of synaptic transmission and the derangement of α-synuclein oligomers [50, 52]. In agreement with these findings, during the course of this research, we also noticed increased levels of α-synuclein in PC12, SH-SY5Y cells, and primary neurons exposed to 6-OHDA, MPP⁺, or rotenone in a concentration- and time-dependent manner. Autophagy dysfunction has been implicated as a hallmark of several neurodegenerative diseases [10]. Especially, numerous studies have shown defective autophagy in PD [2, 7, 10, 11]. However, little is known about how the defective autophagy initiates and contributes to neuronal cell loss in the context of PD. Recently, our group has shown that 6-OHDA, MPP⁺, or rotenone induces neuronal apoptosis by induction of H₂O₂, impeding the AMPK/Akt-mTOR signaling [25]. Here, we provide evidence that PD toxins induce excessive intracellular and mitochondrial H₂O₂, which elicits autophagy inhibition contributing to neuronal apoptosis via activating PTEN and inactivating Akt. We also identified that Parkin plays a critical role in the process.

Autophagy has been considered as a double-edged sword for cells [14]. Autophagy plays an important role in the removal of intracellular misfolded proteins and damaged organelles to maintain the survival and homeostasis of neuronal cells [6, 14, 15]. Yet, under stress state, it can cause cell death as well [14, 16]. Studies have shown that ATG5 and LC3, as two essential components for the canonical autophagy, are involved in the formation of autophagosomes [45, 46], and especially the conversion of LC3-I to LC3-II of LC3 is a protein indicator for autophagic activity [45]. In addition, the p62 protein is a marker for execution of autophagy [47]. It has been reported that treatment with MPP⁺ (1 mM, 24 h) reduced the levels of ATG5 and LC3-II, but increased the level of p62 in SH-SY5Y cells [53], and treatment with rotenone (1 μM, 24 h) decreased the level of LC3-II and increased the level of p62 in PC12 cells [54], suggesting inhibition of autophagy, while other studies have shown that the PD toxins may induce impaired autophagic flux in neuronal cells [55–57]. The discrepancy may be due to different experimental conditions used in these experiments. In line with the above reports, in this study, we found that treatment with 6-OHDA, MPP⁺, or rotenone triggered decreases of ATG5/LC3-II and autophagosomes with a concomitant increase of p62 in a concentration-dependent fashion in PC12, SH-SY5Y cells and primary neurons, as detected by Western blotting and GFP-LC3 assay (Fig. 1). Using genetic rescue experiments for ATG5, we showed that overexpression of wild-type ATG5 substantially attenuated the inhibitory effect of PD toxins on autophagy (Fig. 2A–C). Concurrently, overexpression of ATG5 conferred high resistance to PD toxins-induced apoptotic cell death, as evidenced by less cell viability reduction, decreased percentages of cells with nuclear fragmentation and condensation (Fig. 2E and F), as well as the number of TUNEL-positive cells with fragmented DNA (Fig. 2G and H) in PC12 cells. Collectively, these observations support that there exists ATG5 deficiency in PD toxins-induced oxidative stress, which causes autophagy inhibition-dependent neuronal cell death.

PTEN is a negative regulator for Akt activity [17–20]. In the current study, we observed that 6-OHDA, MPP⁺, or rotenone activated PTEN and inactivated Akt, as the toxins reduced the phosphorylation levels of PTEN and Akt in PC12, SH-SY5Y cells, and primary neurons by using Western blotting and/or immunofluorescence staining (Fig. 3). In general, PTEN activation and Akt inactivation have been thought to promote autophagy [21, 22]. For example, PTEN positively regulates macroautophagy by inhibiting the PI3K/Akt pathway [21]. PTEN attenuates the PI3K/Akt/mTOR signaling pathway to enhance autophagy [22]. However, under certain circumstances, inactivation of Akt can result in impaired autophagic flux leading to cell death [23]. Accordingly, we reasoned that there may exist a crosstalk between PTEN, Akt, and autophagy pathways in neuronal cells exposed to 6-OHDA, MPP⁺, or rotenone, i.e., the toxins-induced activation of PTEN and concurrent deactivation of Akt may cause inhibition of autophagy. At first, we noticed that the cells exhibited decreased p-PTEN (Thr366) in 6–12 h, while decreased p-Akt (Ser473 and Thr308)/ATG5/LC3-II and increased p62 in 12–24 h of post exposure to 6-OHDA, MPP⁺, or rotenone (Fig. 3I and J), suggesting that PTEN, as an upstream factor, regulates Akt activity and autophagy in neuronal models of PD. Importantly, here, for the first time, we present evidence that 6-OHDA, MPP⁺, or rotenone induced inhibition of autophagy pathway indeed by activation of PTEN and inactivation of Akt, resulting in neuronal apoptosis. This is strongly supported by the findings that ectopic expression of dominant negative PTEN (PTEN-C/S) and/or myr-Akt, or myr-Akt and/or pretreatment with rapamycin dramatically rescued the cells from the toxins-induced downregulation of ATG5/LC3-II and autophagosome formation and upregulation of p62, as well as consequential apoptosis in PC12 cells (Fig. 4, Fig. S1). Our findings underscore that PD toxins induce activation of PTEN and inactivation of Akt, leading to autophagy inhibition and eventually apoptosis in neuronal cells.

Many studies have shown that ROS can alter the structures and functions of cellular proteins, and also activate or inhibit related signaling pathways, leading to defects in their physiological function and subsequently more ROS production and ultimately SNpc neuronal cell death/loss in PD [58–60]. Excessive ROS-induced neuronal apoptosis links to dysfunction of PTEN, Akt, and/or autophagy signaling [24–27]. Oxidation by H₂O₂ can result in inhibition of protein tyrosine phosphatase 1B (PTP1B) and PTEN [61], but accumulation of mitochondrial superoxide anions can activate PTP1B and PTEN, leading to inhibition of insulin-like growth factor-1 receptor (IGF-1R) and Akt in murine dermal fibroblasts [62]. Of note, treatment with rotenone (500 μM, 3 h, inducing mitochondrial superoxide anions) did not obviously alter the levels of PTEN and PTP1B, but significantly increased the activity of the two phosphatases, resulting in inhibition of IGF-1R and Akt in murine dermal fibroblasts [62]. Our recent studies have demonstrated that 6-OHDA, MPP⁺, or rotenone induces intracellular and mitochondrial H₂O₂ overproduction [25]. In this study, we also observed that when PC12, SH-SY5Y cells, and primary neurons were exposed to 6-OHDA, MPP⁺, or rotenone for 24 h, cellular H₂O₂ level was significantly elevated (Figs. 5 and 6). Importantly, we revealed that excessive intracellular and mitochondrial H₂O₂ due to PD toxins activated PTEN and inactivated Akt contributing to autophagy inhibition and apoptosis in the cells. This is supported by the observations that pretreatment with CAT, a H₂O₂-scavenging enzyme, and Mito-TEMPO, a mitochondria-selective superoxide scavenger, markedly ameliorated 6-OHDA-, MPP⁺-, or rotenone-induced generation of H₂O₂, decreases of p-PTEN, p-Akt, ATG5/LC3-II/autophagosomes, and increase of p62, as well as cleavage of caspase-3 and apoptosis in the cells (Figs. 5 and 6, Fig. S2 and S3). Our observations are in agreement with the above findings. Collectively, our results support that PD toxins-induced mitochondrial ROS may inhibit autophagy contributing to apoptosis by activating PTEN and inactivating Akt in neuronal cells. However, it is worth mentioning that some studies have shown that H₂O₂ can oxidize PTEN resulting in PTEN inactivation [63, 64], which is in contrast to our findings here. Likely, activation of PTEN is attributed to PD toxins-induced mitochondrial H₂O₂ in this study. Further research is needed to address whether this is the case, or due to different cell lines used or other factors, and whether PD toxins-induced inhibition of Akt is a consequence of activating PTEN alone, activating PTEN/inactivating Akt jointly, or more mechanisms involved.

Studies have demonstrated a close relationship of Parkin dysfunction, as a critical etiological factor of neuronal loss, to pathogenic process of PD [32, 34, 35]. Parkin mutations can impact autophagy in PD [32, 34] and especially, PTEN-induced kinase 1 (PINK1)/Parkin mitophagy is a key mechanism to contribute mitochondrial quality control, and the defects are thought to be a cause of those PD onsets [65]. Akt is significantly decreased in Parkin-knockout compared with wild-type synaptosomes in mouse brain [66]. These facts prompted us to search for the logic that ties Parkin, PTEN/Akt and autophagy together. In this study, we observed that PD toxins (6-OHDA, MPP⁺ or rotenone) concentration-dependently elicited decreases of Parkin in PC12, SH-SY5Y cells, and primary neurons (Fig. 1A–D). This is in agreement with the report that L-DOPA-induced phospho-ubiquitin formation causes Parkin degradation [67]. Interestingly, overexpressing ATG5, or pretreatment with CAT or Mito-TEMPO to scavenge H₂O₂ conferred great resistance to PD toxins-induced reduction of Parkin (Figs. 2, 5, and 6), implying that the protein level of Parkin is regulated by H₂O₂ and autophagy. To gain more insights into the role and significance of Parkin in PD toxins-evoked H₂O₂-mediated PTEN/Akt signaling, autophagy, and apoptosis, we extended our experiments using lentiviral shRNA to Parkin for knockdown of Parkin, or lentiviral FLAG-tagged wild-type Parkin (FLAG-Parkin) for overexpression of Parkin. Surprisingly, knockdown of Parkin markedly weakened rather than strengthened PD toxins-induced H₂O₂ production, PTEN activation/Akt inactivation, autophagy inhibition and apoptosis in PC12 cells (Fig. 7, Fig. S4), yet overexpression of Parkin potentiated the above effects of PD toxins (Fig. 8, Fig. S5). These results suggest the involvement of Parkin in the PD toxins-induced oxidative stress and a feedback mechanism for loss of Parkin involved. Thus, we tentatively propose that though PD toxins downregulate Parkin expression, a certain level of Parkin is required for PD toxins to exert the action on H₂O₂-mediated PTEN/Akt signaling, autophagy and apoptosis in neuronal cells. Undoubtedly, more studies are needed to further understand the mechanism in depth.

In this study, we have demonstrated that abnormal H₂O₂-PTEN/Akt signaling leads to autophagy inhibition-dependent cell death in neuronal models of PD, and Parkin plays an essential role in this process. Because the physiological and pathophysiological status in the context of in vivo PD is more complex, this might be very different from the models used in this study. It has been reported that inactivation of Parkin from dopamine neurons results in activation of the neuroinflammation, contributing to the death of dopamine neurons and neurodegeneration in PD, and loss of Parkin causes the death of dopamine neurons, accompanied by the increase of ROS in a mouse model of PD [68]. Decreased Parkin solubility reflects diminished Parkin function, which is associated with impairment of autophagy in the nigrostriatum of sporadic PD [69]. Although the data support our observations, further research in vitro and in vivo is necessary to establish stronger links between PD toxins-induced oxidative stress-PTEN/Akt-autophagy-Parkin signaling pathways and human PD.

It would be interesting to explore the role and significance of Parkin in the transcriptome level in a PD model. The latest research shows that Parkin participates in the assembly of inflammatory body NLRP3 and triggers downstream inflammatory reaction, leading to the death of dopaminergic neurons in a PD model [68]. In the follow-up research, we will investigate the relevant mechanism in the PD animal model and Parkin-knockout mice. We will detect the autophagy of dopaminergic neurons by injecting 6-OHDA into the substantia nigra to further understand the molecular mechanism of Parkin in the PD model.

In conclusion, we have shown that PD toxins evoke ATG5 deficiency, contributing to autophagy inhibition-dependent apoptosis in neuronal cells. PD toxins elevate the level of intracellular/mitochondrial H₂O₂, which causes activation of PTEN and inactivation of Akt, convergently inhibiting autophagy leading to apoptosis in neuronal cells. We have also identified that Parkin is essential for PD toxins-mediated H₂O₂-PTEN/Akt-autophagy pathway and apoptosis in neuronal cells (Fig. 9). Our findings suggest that co-manipulation of the PTEN/Akt/autophagy signaling by antioxidants may be exploited for the prevention of neuronal loss in PD.

Fig. 9.

A schematic model of how PD toxins (6-OHDA/MPP⁺/rotenone) inhibit autophagy leading to neuronal apoptosis. PD toxins evoke intracellular/mitochondrial H₂O₂. This causes activation of PTEN and inactivation of Akt, convergently inhibiting autophagy contributing to apoptosis in neuronal cells. Parkin is essential for PD toxins-induced H₂O₂-PTEN/Akt-autophagy pathway and apoptosis in neuronal cells

Abbreviations

6-OHDA

6-Hydroxydopamine

AD

Alzheimer disease

AMPK

AMP-activated protein kinase

Atg

Autophagy-related

CAT

Catalase

DAPI

4′,6-Diamidino-2-phenylindole

DMEM

Dulbecco’s modified Eagle’s medium

FBS

Fetal bovine serum

HA

Hemagglutinin

HD

Huntington’s disease

H₂DCFDA

2′7′-Dichlorodihydrofluorescein diacetate

H₂O₂

Hydrogen peroxide

IGF-1R

Insulin-like growth factor-1 receptor

IOD

Integral optical density

LC3

Microtuble-associated protein 1 light chain 3

MPP⁺

1-Methyl-4-phenylpyridin-1-ium

mTOR

Mammalian target of rapamycin

PBS

Phosphate-buffered saline

PD

Parkinson’s disease

PDL

Poly-D-lysine

PI3K

Phosphatidylinositol 3-kinase

PKB/Akt

Protein kinase B

PTEN

Phosphatase and tensin homologue on chromosome 10

PINK1

PTEN-induced kinase 1

PTP1B

Protein tyrosine phosphatase 1B

ROS

Reactive oxygen species

TUNEL

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

Data Availability

The data used to support the findings of this study are available from the corresponding author upon reasonable request.