S-Nitrosylation of parkin as a novel regulator of p53-mediated neuronal cell death in sporadic Parkinson's disease

Abstract

Background: Mutations in the gene encoding parkin, a neuroprotective protein with dual functions as an E3 ubiquitin ligase and transcriptional repressor of p53, are linked to familial forms of Parkinson's disease (PD). We hypothesized that oxidative posttranslational modification of parkin by environmental toxins may contribute to sporadic PD.

Results: We first demonstrated that S-nitrosylation of parkin decreased its activity as a repressor of p53 gene expression, leading to upregulation of p53. Chromatin immunoprecipitation as well as gel-shift assays showed that parkin bound to the p53 promoter, and this binding was inhibited by S-nitrosylation of parkin. Additionally, nitrosative stress induced apoptosis in cells expressing parkin, and this death was, at least in part, dependent upon p53. In primary mesencephalic cultures, pesticide-induced apoptosis was prevented by inhibition of nitric oxide synthase (NOS). In a mouse model of pesticide-induced PD, both S-nitrosylated (SNO-)parkin and p53 protein levels were increased, while administration of a NOS inhibitor mitigated neuronal death in these mice. Moreover, the levels of SNO-parkin and p53 were simultaneously elevated in postmortem human PD brain compared to controls.

Conclusions: Taken together, our data indicate that S-nitrosylation of parkin, leading to p53-mediated neuronal cell death, contributes to the pathophysiology of sporadic PD.

Figure 1

Parkin functions as a repressor of p53 transcription and protein levels. A-B, p53 promoter activity measured by luciferase assay in HEK-293 and SH-SY5Y cells transiently transfected with parkin or the control vector pcDNA. C, p53 promoter activity in SH-SY5Y cell line stably expressing parkin (P-SY5Y). D-E, Western blot analyses of parkin, p53 and actin in SH-SY5Y cells transiently (D) or stably (E) expressing parkin. Quantification of p53 signal (right panels) was made relative to actin and expressed in arbitrary units (a.u.). All values are mean + SEM, n = 6 (A), 12 (B), 18 (C), 3 (D) and 9 (E); * p < 0.01.

Figure 2

Effect of NO on p53 transcription and protein levels in parkin-overexpressing SH-SY5Y cells. A, p53 promoter activity in cells transfected with the control (pcDNA) or Parkin-expression plasmid. The transfected cells were subsequently exposed to 200 μM NO donor GSNO or the control compound GSH. Values are expressed as a percentage of the control (pcDNA-transfected GSH-treated cells). B, Protein levels of parkin, p53 and actin by western blot (left panel) in cells transiently transfected with parkin or pcDNA, and exposed to 200 μM GSNO or GSH. Quantification of p53 levels (right panel) was made relative to actin and expressed in arbitrary units (a.u.). Values are mean + SEM, n = 3; * p < 0.01, ** p < 0.05.

Figure 3

Effect of NO on subcellular localization of parkin and its physical interaction with p53 promoter (Pp53). A, Histological images of SH-SY5Y cells transiently transfected with myc-tagged parkin and exposed to 200 μM old or fresh SNOC. Cells were probed with anti-myc antibody (green) and Hoechst stain (blue). Quantification of the intensity ratio of parkin in the nucleus relative to cytoplasm (right panel) indicates that parkin is excluded from the nucleus after SNOC exposure. Scale bar: 10 μm. Values are mean + SEM, n = 9 from triplicate experiments; * p < 0.01. B, Chromatin immunoprecipitation (ChIP) assay using a specific parkin antibody (MAB5512) and oligonucleotide probe encoding the human p53 promoter sequence. Control or parkin-overexpressing SH-SY5Y cells were exposed to 200 μM GSNO (or GSH) for 4 hours prior to performing the ChIP assay. Relative fold enrichment was assessed and normalized to control. Values are mean + SEM, n = 4–5; * p < 0.05. C, Electrophoretic mobility shift assay (EMSA) using recombinant parkin protein and oligonucleotide probe encoding the human p53 promoter sequence. Parkin proteins were pre-exposed to 200 μM old or fresh SNOC prior to incubation with the oligonucleotide probe. The optical density of shifted bands was quantified and expressed as percent control. Values are mean + SEM, n = 4; * p < 0.01.

Figure 4

Role of p53 in NO-induced death in SH-SY5Y cells overexpressing parkin. A, Validation of p53-shRNA as a gene-silencing tool. SH-SY5Y cells were transfected with control (ctrl)- or p53 shRNA encoding plasmids, and the cell lysates subjected to western blotting using anti-p53 or actin antibodies. The optical density of the p53 bands was quantified and normalized to actin, n = 3; * p < 0.01. B, Histological images of SH-SY5Y cells transiently transfected with pcDNA or parkin plasmid. Cells were co-transfected with ctrl- or p53-shRNA plasmids that also encoded GFP (green). Cells were exposed to 600 μM old or fresh SNOC 24 h prior to TUNEL (red) staining. Scale bar: 20 μm. C, Percentage of TUNEL-positive cells among GFP-positive cells under the designated conditions. Values are mean + SEM of 41–83 cells per condition performed in triplicate experiments; * p < 0.05. D, Schematic representation of the pathway linking NO to apoptosis in this system.

Figure 5

Role of p53 in herbicide/fungicide-induced apoptosis in SH-SY5Y cells overexpressing parkin. A, Histological images of cells transiently co-transfected with plasmids for pcDNA or parkin, together with shRNA vectors coexpressing GFP (green). Cells were exposed to a combination of 100 μM paraquat (PQ) and 10 μM maneb (MB) or vehicle for 6 h, fixed, and subjected to TUNEL staining (red). Scale bar: 20 μm. B, Percentage of TUNEL-positive cells among GFP-positive cells under the indicated conditions. Values are mean + SEM of 183–656 cells per condition performed in triplicate experiments; * p < 0.05.

Figure 6

Role of NO in herbicide/fungicide-induced neuronal death in mesencephalic primary cultures. A, Histological images of mesencephalic mixed cultures. Immunostaining for dopamine transporter (DAT, red), parkin (green), and Hoechst dye (blue). Scale bars: 20 μm. B, Protein levels of parkin measured by western blot of cell lysates. C, Images of primary cultures exposed to a combination of 100 μM paraquat (PQ) and 10 μM maneb (MB), with 1 mM N_ω-nitro-L-arginine (NNA) or vehicle for 6 h. Staining for TUNEL (red) and Hoechst (blue). Scale bar: 10 μm. Quantification of the percentage of TUNEL-positive nuclei in each experimental condition shows PQ/MB induces cell death while NNA inhibits PQ/MB-induced death. Values are mean + SEM of 100–341 nuclei per condition performed in triplicate experiments; * p < 0.05.

Figure 7

Increased S-nitrosylation of parkin and p53 levels in a mouse model of PD. Levels of S-nitrosylated parkin (SNO-parkin), total parkin, p53, and actin were examined by biotin-switch and western blot in mice treated with the nNOS inhibitor 3-Br-7-NI, PQ/MB, or PQ/MB with 3-Br-7-NI (left panel). Ratios of SNO-parkin and total parkin were quantified to indicate the extent of parkin S-nitrosylation under the indicated conditions (center panel). Quantifications of p53 signal were normalized to actin (right panel). Values are mean + SEM, n = 3 mice per condition; * p < 0.05.

Figure 8

Dopamine synthesis, neuronal survival, neurogenesis and astrocytosis in a mouse model of PD. A, Histological images of brain slices from mice treated with PQ/MB, the nNOS inhibitor 3-Br-7-NI, or PQ/MB with 3-Br-7-NI. Staining for TH, NeuN, MAP2, PCNA or GFAP are shown in the indicated brain regions. Scale bars: 75 μm (TH), 35 μm (NeuN), and 25 μm (MAP2, PCNA, GFAP). B, Quantification of number of cells staining for TH, NeuN and PCNA; quantitative confocal immunoreactivity for MAP2; and quantitative optical density for GFAP under the conditions indicated in the specified brain regions. Values are mean + SEM, n = 5–6 mice per condition; * p < 0.05.

Figure 9

Increased S-nitrosylation of parkin and p53 levels in postmortem human brains from Parkinson’s disease (PD) and Incidental Lewy body disease (ILBD). A, Levels of SNO-parkin, total parkin, p53, and actin examined by biotin-switch and western blot in lysates prepared from unaffected control brains (ctrl) and brains with PD and ILBD. B, Ratios of SNO-parkin relative to total parkin were calculated to indicate the extent of parkin S-nitrosylation in ctrl, PD and ILBD samples. Values are mean + SEM of samples from 2 control, 5 PD and 2 ILBD human brains; * p < 0.05. C, p53 levels are quantified and normalized to actin for ctrl, PD and ILBD brains. Values are mean + SEM; * p < 0.01. D, Ratios of SNO-parkin/parkin and p53/actin were plotted for each sample. The two values are positively correlated with a Pearson correlation coefficient of 0.682 and p < 0.05.

Figure 10

Schematic representation of proposed mechanism whereby S-nitrosylation of parkin regulates p53-mediated neuronal death in sporadic PD. A, Under physiological conditions, parkin is neuroprotective by repressing p53 transcription. B, During nitrosative stress, for example due to pesticide exposure, parkin becomes S-nitrosylated. SNO-parkin no longer binds to the p53 promoter and is excluded from the nucleus. This results in activation of the p53 gene and subsequent p53-mediated neuronal death.