PINK1 is activated by mitochondrial membrane potential depolarization and stimulates Parkin E3 ligase activity by phosphorylating Serine 65

Abstract

Missense mutations in PTEN-induced kinase 1 (PINK1) cause autosomal-recessive inherited Parkinson's disease (PD). We have exploited our recent discovery that recombinant insect PINK1 is catalytically active to test whether PINK1 directly phosphorylates 15 proteins encoded by PD-associated genes as well as proteins reported to bind PINK1. We have discovered that insect PINK1 efficiently phosphorylates only one of these proteins, namely the E3 ligase Parkin. We have mapped the phosphorylation site to a highly conserved residue within the Ubl domain of Parkin at Ser(65). We show that human PINK1 is specifically activated by mitochondrial membrane potential (Δψm) depolarization, enabling it to phosphorylate Parkin at Ser(65). We further show that phosphorylation of Parkin at Ser(65) leads to marked activation of its E3 ligase activity that is prevented by mutation of Ser(65) or inactivation of PINK1. We provide evidence that once activated, PINK1 autophosphorylates at several residues, including Thr(257), which is accompanied by an electrophoretic mobility band-shift. These results provide the first evidence that PINK1 is activated following Δψm depolarization and suggest that PINK1 directly phosphorylates and activates Parkin. Our findings indicate that monitoring phosphorylation of Parkin at Ser(65) and/or PINK1 at Thr(257) represent the first biomarkers for examining activity of the PINK1-Parkin signalling pathway in vivo. Our findings also suggest that small molecule activators of Parkin that mimic the effect of PINK1 phosphorylation may confer therapeutic benefit for PD.

Keywords: PINK1; Parkin; Parkinson's disease.

Figure 1.

(Overleaf.) TcPINK1 phosphorylates human Parkin at Ser⁶⁵ in vitro. (a) The indicated PD-linked proteins (1 μM) were incubated with either full-length MBP-fusion of wild-type TcPINK1 (1–570) or kinase-inactive (KI) TcPINK1 (D359A) (0.5 μg) and [γ-³²P] ATP for 30 min. Assays were terminated by addition of SDS loading buffer and separated by SDS-PAGE. Proteins were detected by Colloidal Coomassie blue staining (lower panel) and incorporation of [γ-³²P] ATP was detected by autoradiography (upper panel). Similar results were obtained in three independent experiments. Fine dividing lines indicate that reactions were resolved on separate gels. The substrate bands on the Coomassie gel are denoted with a small red asterisk. All substrates were of human sequence and expressed in E. coli unless otherwise indicated. Tags on the substrates used for this experiment were glutathione s-transferase (GST)-α-synuclein, Parkin (no tag as His-SUMO tag cleaved off), His-UCHL1, GST-DJ1, GST-LRRK2 KI (1326-end D2017A), MBP-ATP13A2, GST-Omi, MBP-PLA2G6, GST-FBX07, GST-GAK-kinase-inactive (D191A), VPS35 (no tag as GST-tag cleaved off). (b) As in (a) except that proteins reported to interact with PINK1 were tested as PINK1 substrates. Human DJ1, Omi, TRAP1, PARL, NCS1 and Miro2 were expressed in E. coli with an N-terminal GST tag. Similar results were obtained in three independent experiments. (c) Timecourse of phosphorylation of Parkin by wild-type TcPINK1. MBP-TcPINK1 (0.5 μg) was incubated in the presence of GST-Parkin (1 μg) and [γ-³²P] ATP for the times indicated and assays terminated by addition of SDS loading buffer. Samples were subjected to SDS-PAGE and proteins detected by Colloidal Coomassie blue staining (lower panel) and incorporation of [γ-³²P] ATP was detected by autoradiography (upper panel). Gel pieces were quantified by Cerenkov counting for calculation of the stoichiometry of Parkin phosphorylation. Similar results were obtained in two independent experiments. (d) Mapping of phosphopeptides on Parkin after phosphorylation by TcPINK1 in vitro. Full-length GST-Parkin (1 μg) was incubated with 2 μg of either wild-type TcPINK1 (1–570) or KI TcPINK1 (D359A) in the presence of Mg²⁺[γ-³²P] ATP for 60 min. Assays were terminated by addition of LDS loading buffer and separated by SDS-PAGE. Proteins were detected by Colloidal Coomassie blue staining and phosphorylated Parkin was digested with trypsin. The resultant peptides were separated by reverse phase HPLC on a Vydac C₁₈ column (Vydac 218TP5215) equilibrated in 0.1% (v/v) trifluoroacetic acid and the column developed with an acetonitrile gradient (diagonal line). The flow rate was 0.2 ml min⁻¹ and fractions (0.1 ml each) were collected and analysed for ³²P radioactivity by Cerenkov counting. Two major ³²P-labelled peaks (P1, P2) were identified following incubation with wild-type TcPINK1 (left). No peaks were identified following incubation with kinase-inactive TcPINK1 (right). (e) Schematic of domain organization of Parkin illustrating that Ser⁶⁵ lies within the Ubl domain (upper panel) and sequence alignment of residues around Ser⁶⁵ in human Parkin and a variety of lower organisms showing high degree of conservation. Abbreviations: Ubl, ubiquitin-like; IBR, in-between-RING; RING, really interesting new gene. (f) Mutation of Ser⁶⁵Ala (S65A) abolishes Parkin phosphorylation by TcPINK1. Full-length wild-type TcPINK1 (1–570) and kinase inactive TcPINK1 (D359A) against wild-type or S65A mutants of full-length Parkin, or the isolated Ubl-domain-containing N-terminal fragment (residues 1–108). The indicated substrates (2 μM) were incubated in the presence of the indicated enzyme (1 μg) and [γ-³²P] ATP for 30 min. Assays were terminated by addition of SDS loading buffer and separated by SDS-PAGE. Proteins were detected by Colloidal Coomassie blue staining (lower panel) and incorporation of [γ-³²P] ATP was detected by autoradiography (upper panel).

Figure 2.

PINK1 phosphorylation of Parkin at Ser⁶⁵ mediates activation of Parkin E3 ligase activity. Wild-type (a) but not kinase-inactive (b) PINK1 activates wild-type Parkin, but does not affect the activity of Ser⁶⁵Ala (S65A) mutant Parkin (c). Two micrograms of wild-type or S65A Parkin were incubated with indicated amounts of wild-type or kinase-inactive (D359A) MBP-TcPINK in a kinase reaction (50 mM Tris-HCl (pH 7.5), 0.1 mM ethylene glycol tetraacetic acid (EGTA), 10 mM MgCl₂, 1% 2-mercaptoethanol and 0.1 mM [γ-³²P] ATP (approx. 500 cpm pmol⁻¹) (in parallel to confirm the phosphorylation) for 60 min. The ubiquitylation reaction was then initiated by addition of ubiquitylation assay components (50 mM Tris-HCl (pH 7.5), 0.05 mM EGTA, 10 mM MgCl₂, 0.5% 2-mercaptoethanol, 0.12 μM human recombinant E1 purified from Sf21 insect cell line, 1 μM human recombinant UbcH7 purified from E. coli, 0.05 mM Flag-Ubiquitin (MW approx. 9.5 kDa) (Boston Biochem) and 2 mM ATP). Reactions were terminated after 60 min by addition of SDS-PAGE loading buffer and resolved by SDS-PAGE. Ubiquitin, Parkin and PINK1 were detected using anti-FLAG, anti-Parkin and anti-MBP antibodies, respectively. Incorporation of [γ-³²P] ATP was detected by autoradiography (lower panel). Ubiquitin attached to the E1 (Ub-Ube1) and ubiquitin dimer (Ub₂) formation occurred in the assay in all conditions (a–c). Ubiquitylation of PINK1 (Ub-PINK1) is indicated (a). Formation of polyubiquitin chains (poly-Ub) upon Parkin activation (a) is indicated. As mentioned in §4, further work is required to establish the nature of these chains and whether they are linked to UbcH7. Representative of five independent experiments.

Figure 3.

Human Parkin Ser⁶⁵ is a substrate of human PINK1 upon CCCP stimulation. (a) Confirmation by mass spectrometry that Ser⁶⁵ of human Parkin is phosphorylated by CCCP-induced activation of human wild-type PINK1-FLAG. Flp-In T-Rex HEK293 cells expressing FLAG-empty, wild-type PINK1-FLAG, and kinase-inactive PINK1-FLAG (D384A) were co-transfected with HA-Parkin, induced with doxycycline and stimulated with 10 μM of CCCP for 3 h. Whole-cell extracts were obtained following lysis with 1% Triton and approximately 30 mg of whole-cell extract were subjected to immunoprecipitation with anti-HA-agarose and run on 10% SDS-PAGE and stained with colloidal Coomassie blue. Coomassie-stained bands migrating with the expected molecular mass of HA-Parkin were excised from the gel, digested with trypsin, and subjected to high performance liquid chromatography with tandem mass spectrometry (LC-MS-MS) on an LTQ-Orbitrap mass spectrometer. Extracted ion chromatogram analysis of Ser¹³¹ and Ser⁶⁵ phosphopeptide (3⁺ R.NDWTVQNCDLDQQSIVHIVQRPWR.K+P). The total signal intensity of the phosphopeptide is plotted on the y-axis and retention time is plotted on the x-axis. The m/z value corresponding to the Ser¹³¹ phosphopeptide was detected in all conditions whilst that of the Ser⁶⁵ phosphopeptide was only detected in samples from wild-type PINK1-FLAG-expressing cells following CCCP treatment. (b) Characterization of Parkin phospho-Ser⁶⁵ antibody. Flp-In T-Rex HEK293 cells expressing FLAG-empty, wild-type PINK1-FLAG, and kinase-inactive PINK1-FLAG were co-transfected with untagged wild-type (WT) or Ser⁶⁵Ala (S65A) mutant Parkin, induced with doxycycline and stimulated with 10 μM of CCCP for 3 h. 0.25 mg of 1% Triton whole-cell lysate were subjected to immunoprecipitation with anti-Parkin antibody (S966C) covalently coupled to protein G Sepharose and then immunoblotted with anti-phospho-Ser⁶⁵ antibody in the presence of dephosphorylated peptide. Ten per cent of the immunoprecipitate (IP) was immunoblotted with total anti-Parkin antibody. Twenty five micrograms of whole cell lysate was immunoblotted with total PINK1 antibody.

Figure 4.

Knock-down of endogenous PINK1 abrogates Parkin Ser⁶⁵ phosphorylation. (a) Timecourse of endogenous PINK1 stabilization by CCCP treatment. HEK293 cells were stimulated at the indicated time points with 10 μM of CCCP. One milligram of whole-cell lysates were immunoprecipitated with anti-PINK1 antibody (S085D) or pre-immune IgG covalently coupled to protein G Sepharose and resolved by 8% SDS-PAGE. Immunoblotting was performed with total PINK1 antibody (Novus). Representative of three independent experiments. (b) Knock-down of endogenous PINK1 abrogates Parkin Ser⁶⁵ phosphorylation. HEK293 cells were co-transfected with PINK1 siRNA (#1 or #2) or scrambled siRNA (scrambled) and untagged wild-type (WT) or Ser⁶⁵Ala (S65A) mutant Parkin as indicated using TransFectin reagent (Bio-Rad). Forty-eight hours post-transfection, cells were treated with or without 10 μM CCCP for 3 h. 0.25 mg of 1% Triton whole-cell lysate were subjected to immunoprecipitation with GST-Parkin antibody (S966C) covalently coupled to protein G Sepharose and then immunoblotted with anti-phospho-Ser⁶⁵ antibody in the presence of dephosphorylated peptide. Five per cent of the IP was immunoblotted with total anti-Parkin antibody. 0.25 mg of whole-cell lysates were immunoprecipitated with anti-PINK1 antibody (S085D) and immunoblotted with anti-PINK1 antibody (Novus). Representative of three independent experiments.

Figure 5.

Human PINK1 directly phosphorylates Parkin Ser⁶⁵ upon CCCP stimulation in vitro. Flp-In T-Rex HEK293 cells expressing wild-type PINK1-FLAG, and kinase-inactive PINK1-FLAG were induced to express protein by addition of 0.1 μg ml⁻¹ of doxycycline in the culture medium for 24 h. Cells were then treated with 10 μM of CCCP for 3 h and lysates subjected to sub-cellular fractionation. Five milligrams of mitochondrial lysate was subjected to immunoprecipitation with anti-FLAG agarose and used in an in vitro radioactive kinase assay with [γ-³²P]-Mg²⁺ATP and E. coli expressed recombinant GST-Parkin Ubl domain (residues 1–108) (Ubl) and mutant GST-Parkin (residues 1–108) Ser⁶⁵Ala (Ubl S65A), purified from E. coli. One half of the assay reaction was run on a 10% SDS-PAGE and was subjected to autoradiography. Colloidal Coomassie stained gel shows equal loading of recombinant substrate. The other half of the reaction was immunoblotted with anti-phospho-Thr²⁵⁷ PINK1 and total PINK1 antibodies following 8% SDS-PAGE.

Figure 6.

Identification and characterization of a novel autophosphorylation site of PINK1 induced by the mitochondrial uncoupling agent CCCP. (a) CCCP induces a band-shift in wild-type but not kinase-inactive PINK1. Flp-In T-Rex HEK293 cell lines stably expressing FLAG alone, wild-type or kinase-inactive PINK1-FLAG were induced to express protein by addition of 0.1 μg ml⁻¹ of doxycycline in the culture medium for 24 h. Cells were then treated with 10 μM of CCCP for 3 h and lysates subjected to sub-cellular fractionation. Twenty-five micrograms of cytoplasmic or mitochondrial lysate were resolved by 8% SDS-PAGE. Relative purity of the fractions was confirmed using cytoplasmic and mitochondrial markers, namely GAPDH and HSP60, respectively. Whole-cell extracts from the same cells were also made in parallel using 1% Triton lysis as described in the methods. In mitochondrial and whole-cell extracts, both wild-type and kinase-inactive PINK1 became stabilized by CCCP but a band-shift was noted for wild-type PINK1 which was revealed to be a doublet on lower exposure. The upper band was absent from kinase-inactive PINK1 treated with CCCP. (b) Identification of Thr²⁵⁷ phosphorylation site on PINK1. Flp-In T-Rex HEK293 cell lines stably expressing FLAG alone, or wild-type PINK1-FLAG were treated with DMSO or 10 μM of CCCP for 3 h. Recombinant PINK1 was immunoprecipitated from 10 mg of mitochondrial extract for each condition using anti-FLAG-agarose and subjected to 4–12% gradient SDS-PAGE and stained with colloidal Coomassie blue. The Coomassie-stained bands migrating with the expected molecular mass of PINK1-FLAG were excised from the gel, digested with trypsin, and subjected to precursor-ion scanning mass spectroscopy. The major phosphopeptide that is indicated ‘Thr²⁵⁷’ was seen from cells expressing wild-type PINK1-FLAG treated with CCCP and this was not seen in bands from the other two conditions. The figure shows the signal intensity (cps, counts of ions per second) of the HPO₃⁻ ion (−79 Da) seen in negative precursor ion scanning mode versus the ion distribution (m/z) for the Thr²⁵⁷ phosphopeptide. The observed values of 722.4 and 788.4 are for the VALAGEYGAVTYR and VALAGEYGAVTYRK variants, respectively, of the Thr²⁵⁷ peptide as [M-2H]²⁻ ions. Other phosphopeptides marked with an asterisk were observed but we were unable to assign phosphorylation site(s). (c) Evidence that CCCP induces PINK1 auto-phosphorylation using a phospho-specific Thr²⁵⁷ antibody. 0.5 mg of mitochondrial extracts (treated with DMSO or 10 μM of CCCP for 3 h) of Flp-In T-Rex stable cell lines expressing FLAG empty, wild-type PINK1-FLAG, kinase-inactive PINK1-FLAG (D384A) and phospho-mutant Thr²⁵⁷Ala (T257A) were immunoprecipitated with anti-FLAG agarose and resolved by 8% SDS-PAGE. Blots were probed with anti-phospho-Thr²⁵⁷ PINK1 antibody and anti-PINK1 antibody. (d) Mutation of Thr²⁵⁷Ala PINK1 does not affect Parkin Ser⁶⁵ phosphorylation. Flp-In T-Rex HEK293 cells expressing FLAG-empty, wild-type PINK1-FLAG, kinase-inactive PINK1-FLAG and T257A PINK1-FLAG were co-transfected with untagged wild-type (WT) or Ser⁶⁵Ala (S65A) mutant Parkin, induced with doxycycline and stimulated with 10 μM of CCCP for 3 h. 0.25 mg of 1% Triton whole-cell lysate were subjected to immunoprecipitation with anti-Parkin antibody (S966C) covalently coupled to protein G Sepharose and then immunoblotted with anti-phospho-Ser⁶⁵ Parkin antibody in the presence of dephosphorylated peptide. Ten per cent of the IP was immune-blotted with total anti-Parkin antibody. One milligram of 1% Triton whole-cell lysate were immunoprecipitated with anti-FLAG agarose and resolved by 8% SDS-PAGE. Blots were probed with anti-phospho-Thr²⁵⁷ PINK1 and anti-PINK1 antibody. (e) PINK1 dephosphorylation by lambda phosphatase inhibits PINK1 activity. C-terminal-FLAG tagged wild-type or kinase-inactive (D384A) PINK1 were immunoprecipitated from 5 mg of mitochondrial enriched extracts using anti-FLAG agarose beads. Wild-type PINK1 was incubated with or without 1000 U of lambda phosphatase or treated with lambda phosphatase along with 50 mM EDTA. Kinase-inactive PINK1 was incubated in buffer alone without lambda phosphatase. The beads were washed thrice in 50 mM Tris pH 7.5, 0.1 mM EGTA and then used in an in vitro kinase assay with GST-Parkin Ubl (1–108) as the substrate. Samples were analysed as described in legend to figure 1.

Figure 7.

Timecourse of CCCP-induced activation of PINK1. (a) Timecourse of PINK1 autophosphorylation in vivo. Flp-In T-Rex HEK293 cells stably expressing PINK1-FLAG wild-type and kinase-inactive (D384A) were stimulated at the indicated time points with 10 μM of CCCP. 0.5 mg of mitochondrial extracts were immunoprecipitated with anti-FLAG agarose and resolved by 8% SDS-PAGE. Immunoblotting was performed with anti-phospho-Thr²⁷⁵ antibody or total PINK1 antibody. (b) No time-dependent activation of cytoplasmic PINK1 in vivo. As in (a) cytoplasmic extracts were obtained at the indicated time-points and immunoprecipitated with anti-FLAG agarose and resolved by 8% SDS-PAGE. Immunoblotting was performed with PINK1 anti-phospho-Thr²⁷⁵ antibody or total PINK1 antibody. (c) Timecourse of Parkin Ser⁶⁵ phosphorylation in vivo. Flp-In T-Rex HEK293 cells stably expressing wild-type PINK1-FLAG were co-transfected with untagged wild-type (WT) or Ser⁶⁵Ala (S65A) mutant Parkin; induced with doxycycline and stimulated with 10 μM of CCCP at the indicated time points. 0.25 mg of 1% Triton whole-cell lysate were subjected to immunoprecipitation with anti-Parkin antibody (S966C) covalently coupled to protein G Sepharose and then immunoblotted with anti-phospho-Ser⁶⁵ antibody in the presence of dephosphorylated peptide. One per cent of the IP was immunoblotted with total anti-Parkin antibody. 1.5 mg of whole-cell extracts were immunoprecipitated with anti-FLAG agarose and resolved by 8% SDS-PAGE. Immunoblotting was performed with total PINK1 antibody.

Figure 8.

PINK1 is specifically activated by inducers of mitochondrial membrane depolarization. (a) Table of agonists tested. (b) PINK1 phosphorylation of Parkin at Ser⁶⁵ is specific to mitochondrial uncouplers. Flp-In T-Rex HEK293 cells expressing wild-type PINK1-FLAG were co-transfected with untagged wild-type Parkin, induced with 0.1 μg ml⁻¹ doxycycline for 24 h and stimulated with the indicated agonists for 3 h except for Deferiprone (24 h treatment). 0.25 mg of 1% Triton whole-cell lysate was subjected to immunoprecipitation with anti-Parkin antibody (S966C) covalently coupled to protein G Sepharose and then immunoblotted with anti-phospho-Ser⁶⁵ antibody in the presence of dephosphorylated peptide. Ten per cent of the IP was immunoblotted with total anti-Parkin antibody. Twenty-five micrograms of whole-cell lysate was immunoblotted with total PINK1 antibody (Novus), phospho-JNK (CST), total-JNK (CST), phospho-ERK1/2, total-ERK1/2 (CST), phospho-ACC (CST), total ACC (CST), phospho-AMPK (CST) and total AMPK (CST). Representative of two independent experiments.

Figure 9.

Model of Parkin activation by PINK1. Under basal conditions Parkin is kept in a closed inactive conformation by Ubl-mediated autoinhibition. Following mitochondrial depolarization, PINK1 phosphorylates Parkin at Ser⁶⁵ thereby relieving Parkin autoinhibition and enabling Parkin to become active to ubiquitylate target substrates.