Introduction
Recessively inherited loss of function mutations in the
PINK1 gene have been linked to familial Parkinson’s disease (PD). The PINK1 protein bears a 34 amino acid mitochondrial targeting domain [
1] and has been shown to localize within mitochondria [
2]. Mitochondrial dysfunction has long been thought to play a key role in PD pathogenesis, based in part on postmortem studies that showed mitochondrial impairment (
e.g. reduced complex I activity) and oxidative damage in idiopathic PD brains [
3]. This is further supported by observations that mitochondrial complex I inhibitors, such as MPTP [
4] and rotenone [
5] produce parkinsonian syndromes in humans and experimental animal models.
Genetic studies in
Drosophila showed that PINK1 is involved in the maintenance of mitochondrial morphology by interacting with components of the mitochondrial fission and fusion machinery [
6‐
9]. Loss of PINK1 in
Drosophila appears to promote mitochondrial fusion, though the effects of PINK1 inactivation on mitochondrial morphology in cultured mammalian cells are less consistent, ranging from promotion of mitochondrial fragmentation or fusion to no effects [
10‐
14]. Despite the controversial findings on the effects of PINK1 inactivation on mitochondrial morphology in mammalian culture systems, several functional defects have been reported consistently, including impairment of mitochondrial respiration [
15‐
20] and reduction of mitochondrial transmembrane potential [
1,
11,
15,
16,
21]. Our previous analysis of
PINK1−/− mice led to the first report showing that PINK1 is required for mitochondrial respiratory function
in vivo[
18]. However the cause of these functional defects remains to be elucidated.
To determine the pathogenic cascade of events in intact mitochondria, we derived primary mouse embryonic fibroblasts (MEFs) and cortical neuronal cultures from our PINK1−/− mice. Similarly to what we previously reported in isolated mitochondria from the brain, mitochondrial respiration is impaired in PINK1−/− cells. While the enzymatic activity of each complex composing the electron transport system is normal, mitochondrial transmembrane potential (ΔΨm) is reduced in PINK1−/− MEFs and neurons. Interestingly, the reduction of ΔΨm in PINK1−/− cells is associated with increased opening of the mitochondrial permeability transition pore (mPTP). Inhibition of the mPTP reverses the depolarization of the mitochondrial inner membrane and respiration defects seen in PINK1−/− cells. We did not find evidences of increased oxidative stress, a common inducer of mPTP opening. In addition, we found no detectable changes in mitochondrial morphology in PINK1−/− cells. Together our findings highlight a role of PINK1 in the regulation of the mitochondrial permeability transition pore and suggest that increased opening of the pore in the absence of PINK1 may be responsible for the reduced mitochondrial transmembrane potential and the reduced respiratory activities.
Materials and methods
Primary MEF and cortical cultures
Mouse embryonic fibroblasts (MEFs) were derived from embryos at embryonic day 14.5. After removing the head and the inner organs embryos were individually minced with scissors, treated with trypsin (1% v/v) for twelve minutes at 37°C and dispersed mechanically and plated with MEF media (Dulbecco′s Modified Eagle Medium (DMEM), 10% fetal bovine serum (FBS), penicillin/streptomycin (Gibco, Life Technologies, Grand Island, NY, USA)). After they reached ~100% confluency, cells were frozen down in DMEM containing 10% DMSO (Sigma, St Louis, MO, USA). The number of MEF samples used in each experiment is specified in the figure legends and reflects the number of individual cultures derived from individual embryos used to derive MEFs.
Primary cortical cultures were prepared and maintained as described previously [
22]. Experiments were performed at 14 ± 1 days
in vitro. Cortices from different pups were not pooled, and the number of experiments specified in the legend reflects the number of different cultures derived from individual pups.
Mitochondrial respiration
Mitochondrial respiration was assayed as the O2 consumption of cell suspension using a Clark electrode (Rank Brothers Ltd, Cambridge, England). Cells were resuspended to a final density of 2.106 cells/ml in respiration buffer (0.137 M NaCl, 5 mM KCl, 0.7 mM NaH2PO4, 25 mM Tris, pH 7.4 at 25°C). Endogenous respiration activity was measured after addition of glucose (10 mM, Sigma). For complex driven respiration, plasma membranes were permeabilized by addition of digitonin at a final concentration of 0.01% (Sigma). Cells were supplemented with substrates for either complex I (10 mM glutamate/malate, Sigma), II (10 mM succinate, Sigma) or III (1 mM TMPD/1 mM ascorbate, Sigma) together with adenosine diphosphate (ADP, 1 mM, Sigma) to the recording chamber. State 3 respiration activity was then measured. ADP independent respiration activity (State 4) was monitored after addition of oligomycin (2 μM, Sigma).
Enzymatic activity of ETS complexes and ATP synthase
All assays were performed on mitochondria isolated from MEFs according to a previously established method [
23]. For each complex 5 μg of mitochondrial proteins and 100 μl of each assay buffer were used. Complex I (NADH: ubiquinone oxidoreductase) buffer (35 mM NaH
2PO
4 pH 7.2, 5 mM MgCl
2, 0.25% BSA, 2 mM KCN, 1 μM antimycin, 97.5 μM ubiquinone-1, 0.13 mM NADH, Sigma). Only the rotenone sensitive activity was monitored by following the oxidation of NADH at 340 nm (OD 6220 M
-1.cm
-1). Complex II (succinate dehydrogenase) buffer (25 mM KH
2PO
4, 5 mM MgCl
2, pH 7.2, 20 mM succinate, 50 μM 2,6-dichlorophenolindophenol (DCPIP), 0.25% BSA, 2 mM KCN, 1 μM antimycin, Sigma). Enzymatic activity was monitored by the reduction of DCPIP/PES at 600 nm (OD 19100 M
-1.cm
-1) after addition of 65 μM ubiquinone 1. Complex III activity (decylubiquinol/ferricytochrome C oxidoreductase) buffer (3 mM sodium azide, 1.5 μM rotenone, 50 μM ferricytochrome C, and 50 mM phosphate buffer, pH 7.2, Sigma). Reaction was followed as the increase in reduced Cytochrome C absorbance at 550 nm (OD 18500 M
-1.cm
-1) after the addition of 35 μM of freshly prepared ubiquinol 2. Complex IV (Cytochrome C oxidase) activity and Complex II + III (succinate-Cytochrome C reductase) activities were previously described [
18]. Levels of Cytochrome C were measured by western blot using a commercial antibody (Cell signalling Technology, Danvers, MA, USA).
Measurement of mitochondrial transmembrane potential and mPTP opening
Mitochondrial ΔΨ was measured with the non-quenching Tetramethylrhodamine, methyl ester (TMRM) fluorescence methods (Molecular Probes, Life Technologies). MEFs were stained with TMRM (50 nM) in DMEM for 30 min at 37°C in the dark. The cells were then washed twice with PBS. Mitochondrial PTP opening was assessed by the quenching of calcein-AM fluorescence by cobalt [
24]. Thirty min after cells were loaded with Calcein-AM (1 μM, Molecular Probes, Life Technologies) at 37°C in the dark, CoCl
2 (1 mM, Sigma) was added and cells incubated for another 10 min. Then, fluorescence of 30,000 cells for each experiment was measured with a flow cytometer (FACSCalibur), and the data were processed with the CellQuest program (BD Biosciences, San Jose, CA, USA). Neurons were incubated for 30 min with TMRM (50 nM) in neuronal extracellular buffer with calcium or with Calcein (1 μM) for 45 min in the dark after 30 min CoCl
2 (1 mM) was added. Then cells were washed and imaged on a Leica DMI6000 Microscope. Imaging processing and data analysis were performed using LASAF software (Leica, Wetzlar, Germany). In some experiments cells were pre-incubated for 1 hr with atractylate (20 μM, Sigma), Cyclosporine A (CsA, 1 μM, Sigma), Bongkrekic acid (BkA, 10 μM, Sigma), FK-506 (5 μM, Sigma), 0.1% vehicle (DMSO), Tocopherol (50 μM, 4 hr, Sigma) or NAC (1 mM, 2 hr, Sigma).
For imaging expreriments, MEFs were cultured on glass bottom culture dishes. Cells were loaded for Δψm with TMRM (50 nM) and Mitotracker Green (200 nM) (Molecular Probes, Life Technologies) with or without Oligomycin (1 μM), FCCP (1 μM) or CoCl2 (1 mM). For the mPTP opening assay, cells were loaded with calcein-AM (1 μM) and Mitotracker Red (150 nM) (Molecular Probes, Life Technologies), with or without CoCl2 (1 mM) both in HBSS 1X (Gibco, Life Technologies) for 20 min at 37°C. Then, cells were washed three times in HBSS 1X. Live images of the cells were captured with the Olympus FluoView FV1000 Confocal Microscope (Olympus Imaging America Inc, Center Valley, PA, USA) and analyzed using ImageJ software.
Oxidative stress assay
To measure H2O2 production mitochondria were isolated using the mitochondrial isolation kit from Sigma according to the manufacturer instructions. The experiment was started by adding 100 μl of assay buffer (HBSS containing 10 μM Amplex Red, 10 mM succinate, 0.2 units/ml Horse Radish Peroxidase) and followed over time on a fluorescence plate reader. The same conditions were used to determine the production of superoxide anion using the Dihydroethidium (10 μM, DHEt) method. The protein carbonyl contents in cell lysates were detected by the OxyBlot protein oxidation detection kit (Millipore, Billerica, MA) using the instructions provided by the manufacturer. Lipid peroxidation was analyzed using the ThioBarbituric Acid Reactive Species (TBARS) assay and according to the manufacturer's instructions (Cayman Chemical, Ann Arbor, MI, USA).
Analysis of mitochondrial morphology
For visualization of mitochondria, MEFs were either stained with Mitotracker Red (250 nM) or infected with a retrovirus expressing mt-DsRed [
25]. Primary cultured neurons were only imaged with MitoTracker Red (100 nM). Then cells were washed for 10 min and fixed with 4% PFA (Electron Microscopy Science, Hatfield, PA, USA) for 20 min. After fixation coverslips were mounted on glass slides and imaged by epifluorescence on a Leica DMI6000 Microscope (Leica Microsystems GmbH, Wetzlar, Germany). For live imaging, after staining cells were mounted on a perfusion chamber in culture media containing HEPES (1 mM) and imaged at 22°C. Regardless of the staining method, cells were then scored by eye into four different categories according to the morphology of their mitochondrial network previously described [
25]. The automatic analysis of the size and branching of the mitochondrial network was done using the particle analysis function of ImageJ according to a previously described methods using ImageJ [
26].
Calcium imaging
FCCP releasable pool was measured by adapting a previously described method [
27]. Briefly, MEFs and primary cortical neurons cultures were loaded with Fura-2 AM (5 μM, 45 min at 37°C) (Molecular probes, Life Technologies), and imaged with a Leica DMI6000 Microscope. Imaging processing and data analysis were performed using LAS AF software (Leica). FCCP (1 μM) was applied using an 8-channel gravity perfusion system (ALA Scientific Instrument, Farmingdale, NY, USA).
Statistical analysis
Statistical analysis was performed using Prism 5 (Graph-Pad Software) and Excel (Microsoft). Pooled results were expressed as means ± SEM. Significance was determined by the non paired Student t-test.
Discussion
In the current study, we investigated the mechanism underlying the mitochondrial respiration defects caused by loss of PINK1. We established primary MEFs and cortical neuronal cultures from our
PINK1−/− mice to evaluate mitochondrial functions in intact cells. Similar to what we previously reported in mitochondria isolated from mouse brains [
18], mitochondrial respiration is impaired in
PINK1−/− MEFs, indicating that these cells represent a valid cellular model to study the detailed mechanisms underlying respiratory defects seen in
PINK1−/− mice (Figure
1). Although respiration impairment can be caused by defects in mitochondrial transmembrane potential or the electron transport system, we found that only the mitochondrial transmembrane potential is reduced in
PINK1−/− cells (Figure
2), while enzymatic activities of the complexes composing the electron transfer system are unaffected (Figure
1). In search for mechanisms underlying the reduction of the transmembrane potential, we found that opening of the mitochondrial permeability transition pore is increased in the absence of PINK1 and that this defect can be rescued by inhibitors of the mPTP (Figures
3 and
4). Furthermore, mitochondrial transmembrane potential and respiration defects caused by loss of PINK1 were also reversed specifically by inhibitors of the mPTP (Figures
5 and
6), suggesting that increased opening of the mPTP underlies the defects in mitochondrial transmembrane potential and respiration observed in
PINK1−/− cells. These mitochondrial functional defects occur in the absence of elevated oxidative stress (Figure
7) and mitochondrial morphological changes (Figure
8), but mitochondrial calcium is increased in
PINK1−/− cells, suggesting that elevated mitochondrial calcium underlies the increase in mPTP opening (Figure
9).
Following our initial report of mitochondrial respiration defects in
PINK1−/− mouse brains [
18], a growing consensus has been building on the importance of PINK1 in mitochondrial respiration [
15‐
19,
44], though the underlying mechanism remained unclear. Defects in the activity of the electron transport system complexes have been suggested as a possible mechanism underlying the respiratory defects resulting from the loss of PINK1, as silencing PINK1 expression by siRNA in SH-SY5Y cells affected mitochondrial ATP synthesis and activity of ETC complexes [
19]. However, enzymatic activities of the ETC complexes in our primary
PINK1-deficient cells are normal. Instead we found that loss of PINK1 increased opening of the mitochondrial permeability transition pore, and that blocking mPTP opening occluded the difference between
PINK1−/− and control MEFs for endogenous and State 3 respiration. These results suggest that increased mPTP opening is responsible for reduced mitochondrial respiration in
PINK1−/− cells. Previous reports showed that mPTP opening triggered by elevated calcium concentrations leads to reduced state 3 respiratory activities [
45,
46], an effect that can be prevented by pretreatment with CsA [
45,
47,
48].
Reduced transmembrane potential in
PINK1-deficient cells has been reported in a wide variety of cells [
15‐
17,
21,
49]. In accordance with these previous reports, we also found that ΔΨm is reduced in primary
PINK1−/− fibroblasts and neurons. It has been proposed that reduced enzymatic capacity of complex I of the mitochondrial electron transport system might be the underlying cause of the defects in ΔΨm [
11]. However, similar to other previous studies [
17,
50,
51], we found that complex I enzymatic activity as well as the activity of all other complexes composing the ETS are normal in our cell models in the absence of PINK1. Rather, we found that increased opening of the mPTP likely plays an important role in the mitochondrial depolarization observed in
PINK1−/− cells, as inhibitors of the mPTP, CsA and BkA, rescued the ΔΨm defects in
PINK1−/− cells. The opening of the mPTP allows free diffusion of small ions across the mitochondrial inner membrane as a corrective mechanism for cation overload [
52‐
54]. Hence, increased opening of the mPTP allows a partial depolarization of the mitochondrial membrane, and this defect can be reversed by inhibition of mPTP opening such as CsA [
24,
52]. The stronger rescuing effect observed with CsA might relate to the fact that, in addition to blocking the mPTP, it also hampers mitochondrial calcium uptake, which may be increased in
PINK1-deficient cells, as suggested by higher mitochondrial calcium levels in these cells.
A possible role of PINK1 in modulating mitochondrial morphology and dynamics emerged from studies in
Drosophila. Loss of PINK1 function in flies results in abnormally large mitochondria with fragmented cristae and reduced capacity to generate ATP [
6,
7]. This mitochondrial phenotype is suppressed by genetically promoting mitochondrial fission or decreasing mitochondrial fusion, inferring that perturbed mitochondrial fission in
PINK1-deficient models underlies functional defects [
9,
40,
41]. However, whether and how PINK1 may regulate mitochondrial morphology and dynamics in mammalian cells is much less clear. The effects of PINK1 deficiency on mitochondrial morphology and dynamics seem to depend on the cell type studied and range from inducing mitochondrial fission [
10,
42] or fusion [
14] to no effect [
11,
16,
19,
50]. Consistent with these studies [
11,
16,
19,
50], our analysis of primary cultured
PINK1−/− MEFs and neurons did not show overt changes in mitochondrial morphology in fixed or live cells (Figure
8). These findings are also in agreement with our earlier EM study showing that no drastic ultrastructural changes in mitochondrial number and integrity in
PINK1−/− brains at 3–4 and 22–24 months of age [
18]. Thus, loss of PINK1 function causes mitochondrial functional defects, in the absence of morphological changes, suggesting that the morphological abnormalities observed in mammalian cell lines may be downstream consequences of these mitochondrial functional defects.
How does loss of PINK1 lead to increased mPTP opening? Opening of the mPTP is primarily induced by oxidative stress and/or elevated intramitochondrial calcium concentrations [
45]. We did not find any evidence of oxidative damage or increased production of ROS in
PINK1-deficient cells (Figure
7). However, we found that mitochondrial calcium concentration measured indirectly in the cytosol following FCCP treatment is increased in
PINK1−/− MEFs and neurons (Figure
9). This observation is consistent with a recent study showing that loss of PINK1 reduces the activity of the mitochondrial Na
+/Ca
2+ exchanger (NCX), which regulates Na
+-dependent Ca
2+ efflux [
15]. Pharmacologic inhibition of NCX activity leads to accumulation of calcium in isolated mitochondria [
55]. It is therefore possible that impaired NCX activity in
PINK1−/− cells may lead to accumulation of mitochondrial calcium, which in turn increases the opening of the mPTP. In this context, the mPTP may serve as a Ca
2+-activated Ca
2+ release channel [
56]. However, it is unclear how loss of PINK1 affects the activity of the mitochondrial Na
+/Ca
2+ exchanger. A direct regulation of the mitochondrial Na
+/Ca
2+ exchanger by PINK1-mediated phosphorylation is possible but difficult to demonstrate, since the molecular nature of the Na
+/Ca
2+ exchanger is unknown.
Mitochondrial dysfunction has long been thought to play an important role in the pathogenesis of Parkinson’s disease [
57]. This is based on earlier studies using postmortem idiopathic PD brains showing mitochondrial respiration impairment and oxidative damage [
3], and on findings that mitochondrial complex I inhibitors, such as MPTP and rotenone, produce parkinsonian syndromes in humans and experimental animal models [
4,
5]. Our prior reports showing mitochondrial respiration defects in
Parkin−/− and
PINK1−/− mouse brains and linking these recessive PD genes to mitochondrial function provided experimental evidence in support of a causal role of mitochondrial functional impairment in PD pathogenesis [
18,
58]. The current study highlights the importance of mitochondrial permeability transition pore opening in PINK1 mediated mitochondrial respiration and function. Our recent unpublished work further showed that loss of Parkin or DJ-1 also leads to increases in mPTP opening (EG and JS, unpublished data). Thus, increased mPTP may be a common mechanism leading to PD pathogenesis.
In summary, our study highlights an important role of PINK1 in the regulation of mitochondrial permeability transition pore. Our findings suggest that dysregulation of the opening of the mPTP likely underlies impairment of mitochondrial respiration and reduction of mitochondrial transmembrane potential. Future studies will be needed to elucidate the mechanism by which PINK1 regulates mitochondrial calcium homeostasis and opening of the mPTP. Given the importance of mPTP opening in the regulation of the release of proapoptotic factors from mitochondria to the cytosol, it will be important to determine whether alteration of mPTP opening is a key mechanism underlying increased vulnerability of
PINK1-deficient cells to exogenous stressors [
11,
59]. In addition, future studies are needed to determine whether increased opening of the mPTP is a feature common to other genetic forms of the disease, and whether modulation of its opening may provide a novel therapeutic strategy for the treatment of Parkinson’s disease.
Competing interests
The authors declare that they have no competing interests.
Authors’ contribution
CG and JS conceived and designed the study and wrote the manuscript. CG and EG carried out the experiments and obtained the data for Figures 1-9. LM, EC and CV participated in experimental design for Figure 9, and ZS and DC carried out the mitochondrial morphological analysis using the retroviral vectors mt-DsRed in Figure 8. All authors read and approved the final manuscript.