Mutant Pink1 induces mitochondrial dysfunction in a neuronal cell model of Parkinson's disease by disturbing calcium flux

Abstract

Parkinson's disease (PD) is characterized by accumulation of alpha-synuclein (alpha-syn) and degeneration of neuronal populations in cortical and subcortical regions. Mitochondrial dysfunction has been considered a potential unifying factor in the pathogenesis of the disease. Mutations in genes linked to familial forms of PD, including SNCA encoding alpha-syn and Pten-induced putative kinase 1 (PINK1), have been shown to disrupt mitochondrial activity. We investigated the mechanisms through which mutant Pink1 might disrupt mitochondrial function in neuronal cells with alpha-syn accumulation. For this purpose, a neuronal cell model of PD was infected with virally-delivered Pink1, and was analyzed for cell survival, mitochondrial activity and calcium flux. Mitochondrial morphology was analyzed by confocal and electron microscopy. These studies showed that mutant (W437X) but not wildtype Pink1 exacerbated the alterations in mitochondrial function promoted by mutant (A53T) alpha-syn. This effect was associated with increased intracellular calcium levels. Co-expression of both mutant Pink1 and alpha-syn led to alterations in mitochondrial structure and neurite outgrowth that were partially ameliorated by treatment with cyclosporine A, and completely restored by treatment with the mitochondrial calcium influx blocker Ruthenium Red, but not with other cellular calcium flux blockers. Our data suggest a role for mitochondrial calcium influx in the mechanisms of mitochondrial and neuronal dysfunction in PD. Moreover, these studies support an important function for Pink1 in regulating mitochondrial activity under stress conditions.

Fig. 1

Characterization of levels of expression in B103 neuronal cells infected with LV α-syn and Pink1. (A) In samples infected with empty vector no fluorescent signal was detected. (B) Abundant signal was detected in cells infected with LV-GFP. (C and D) Detection by immunoflourescence microscopy of α-syn expression in cells infected with LV-α-syn wt or LV-α-syn A53T. (E and F) Immunoflourescence detection of Pink1 with an antibody against HA in cells infected with LV-Pink1 wt or LV-Pink1 W437X. (G) Western blot analysis of homogenates from neuronal cells infected with LV-GFP, LV-α-syn or LV-Pink1. Scale bar (20μm) in panel (F) applies to all photographs.

Fig. 2

Analysis of mitochondrial morphology by Mitotracker in neuronal cells infected with LV expressing α-syn and Pink1. (A-C) Baseline appearance of mitochondria (arrows) in neuronal cells infected with LV-empty control, α-syn wt and Pink1 wt. (D-F) Enlarged mitochondria (arrows) in neuronal cells infected with LV-α-syn A53T and/or Pink1 W437X. (G) Image analysis of mean mitochondrial diameter. An average of 200 cells per condition were recorded by laser confocal microscopy and analyzed with the NIH Image J program. Cells expressing α-syn A53T or Pink1 W437X showed a 25–50% increase in mitochondrial size. These alterations were exacerbated in cells co-infected with LV-α-syn A53T and LV-Pink1 W437X. *p < 0.05 compared to cells infected with LV-control by one-way ANOVA followed by post-hoc Dunnett’s test; #p < 0.05 compared to cells singly-infected with LV-α-syn A53T or LV-Pink1 W437X by one-way ANOVA followed by post-hoc Tukey-Kramer test. Scale bar (5μm) in panel (F) applies to all photographs.

Fig. 3

Ultrastructural analysis of neuronal cells infected with LV expressing α-syn and Pink1. (A and B) Cells infected with LV-control display preservation of the baseline structure of mitochondria (M), rough endoplasmic reticulum, and lysosomes (L). (D-F) In neuronal cells infected with either LV-α-syn A53T (C and D) or Pink1 W437X (E and F) the mitochondria (M) were elongated and often surrounded by filaments (arrows). In addition, in the cytoplasm there were abundant electrodense laminar structures reminiscent of autophagolysosomes (AP). (G and H) Neuronal cells co-infected with LV-α-syn A53T and Pink1 W437X displayed the presence of enlarged mitochondria with redundant cristae (arrows) and abnormal lysosomes (L) with electrodense material and autophagolysosomes (AP). Scale bar (0.5μm) in panel (G) applies to panels A, C, E, G; scale bar (1μm) in panel (H) applies to panels B, D, F, H.

Fig. 4

Characterization of the mitochondrial pathology in a neuronal cell line infected with LV expressing mutant α-syn and Pink1. (A) Preserved mitochondria (M) morphology in cells infected with the LV-control. (B, C) Enlarged and elongated mitochondria were found in the proximity of autophagolysosomes (AP) in cells co-infected with LV-α-syn A53T and Pink1 W437X. (D) Accumulation of abnormal electrodense granules (arrows) in the mitochondrial matrix in cells co-infected with LV-α-syn A53T and Pink1 W437X. (E, F) Abnormal mitochondria surrounded by filaments (arrowheads) and autophagolysosomes in cells co-infected with LV-α-syn A53T and Pink1 W437X. Insets show higher-power image of peri-mitochondrial filaments. Scale bar (0.25μm) in panel (F) applies to all micrographs.

Fig. 5

Co-localization of α-syn and Pink1 with mitochondria in neuronal cells infected with LV. Cells were grown on coverslips, labeled with Mitotracker Red, immunostained with antibodies against α-syn or HA to detect Pink1 and imaged with the laser confocal microscope. (A-C) Cells expressing Pink1 wt or (D-F) α-syn wt displayed a more diffuse immunoreactivity (arrows). (G-I) In cells infected with LV-Pink1 W437X or (J-L) LV-α-syn A53T the immunoreactivity for Pink1 or α-syn was in closer proximity to mitochondria stained with Mitotracker (arrows). (M-O) Close-up of the image indicated by the box. (P-R) Co-localization of mutant α-syn and Mitotracker became more evident in cells co-infected with LV-α-syn A53T and LV-Pink1 W437X (*). Scale bar (5μm) in panel (R) applies to panels A-L and P-R; scale bar (2μm) in panel (O) applies to panels M-O.

Fig. 6

Functional mitochondrial alterations, and the protective effects of cyclosporine A and Ruthenium red, in B103 cells infected with LV expressing α-syn and Pink1. All values are expressed as percent of control. Neuronal cells were treated with 5μM cyclosporine A or 10mM Ruthenium red. (A-C) Semi-quantitative analysis of the mitochondrial ΔΨ_m by Mitotracker Red fluorescence in cells treated with vehicle control (A), cyclosporine A (B), or Ruthenium red (C). (D-F) Determination (by luminescent assay kit) of ATP production in cells treated with vehicle control (D), cyclosporine A (E) and Ruthenium red (F). Cells expressing mutant α-syn and Pink1 showed a reduction in mitochondrial function and ATP production, this effect being greater in cells expressing both mutant proteins. Treatment with cyclosporine A partially reduced the deficits in cells expressing mutant α-syn and Pink1, while Ruthenium red fully reverted the effects of the combined mutant proteins. *p < 0.05 compared to LV-control by one-way ANOVA followed by post-hoc Dunnett’s test; #p < 0.05 compared to cells singly-infected with LV-α-syn A53T or LV-Pink1 W437X by one-way ANOVA followed by post-hoc Tukey-Kramer test.

Fig. 7

Neuronal alterations and the protective effects of cyclosporine A and Ruthenium red, in B103 cells infected with LV expressing α-syn and Pink1. All values are expressed as percent of control. Neuronal cells were treated with 5μM cyclosporine A or 10μM Ruthenium red. (A-C) MTT assay in cells treated with vehicle control (A), Cyclosporine A (B), or Ruthenium red (C). (D-F) Determination of neurite length by phase contrast microscopy and image analysis of cells treated with vehicle control (D), Cyclosporine A (E), or Ruthenium red (F). Cells expressing mutant α-syn and Pink1 showed a reduction in mitochondrial activity and neurite lengths, this effect being greater in cells expressing both mutant proteins. Treatment with cyclosporine A partially reduced the deficits in cells expressing mutant α-syn and Pink1, while Ruthenium red fully reverted the effects of the combined mutant proteins. *p < 0.05 compared to LV-control by one-way ANOVA followed by post-hoc Dunnett’s test; #p < 0.05 compared to cells singly-infected with LV-α-syn A53T or LV-Pink1 W437X by one-way ANOVA followed by post-hoc Tukey-Kramer test.

Fig. 8

Effects of calcium blockers on membrane potential and ATP production in a neuronal cell line infected with LV expressing α-syn and Pink1. All values are expressed as percent of control. Neuronal cells were treated with cellular calcium inhibitors cobalt chloride (5μM), or flufenamic acid (40μM) and analyzed with Mitotracker Red or by ATP assay. (A-C) Semi-quantitative analysis of the mitochondrial ΔΨ_m by Mitotracker Red fluorescence in cells treated with vehicle control (A), Cobalt chloride (B), or Flufenamic acid (C). (D-F) Determination (by luminescent assay kit) of ATP production in cells treated with vehicle control (D), Cobalt chloride (E), or Flufenamic acid (F). In contrast to the protective effects of Ruthenium red, neither of the other calcium blockers reverted the effects of expressing mutant α-syn and Pink1 on the ΔΨ_m or ATP production. *p < 0.05 compared to LV-control by one-way ANOVA followed by post-hoc Dunnett’s test; #p < 0.05 compared to cells singly-infected with LV-α-syn A53T or LV-Pink1 W437X by one-way ANOVA followed by post-hoc Tukey-Kramer test.

Fig. 9

Effects of calcium blockers on intracellular calcium levels in neuronal cells expressing α-syn and Pink1. All values are expressed as percent of control. Neuronal cells were incubated with cellular calcium inhibitors ruthenium red (10μM), cobalt chloride (5μM), or flufenamic acid (40μM) followed by treatment with calcium dye and analyzed by fluorescence microscopy on a spectrophotomer. (A) Calcium levels in cells treated with vehicle control. (B) Calcium levels in cells treated with Ruthenium red. (C) Calcium levels in cells treated with Cobalt chloride. (D) Calcium levels in cells treated with Flufenamic acid. As for ΔΨ_m, treatment with ruthenium red was able to revert the effects of co-expressed mutant proteins on calcium influx. *p < 0.05 compared to LV-control by one-way ANOVA followed by post-hoc Dunnett’s test.