Binding to serine 65-phosphorylated ubiquitin primes Parkin for optimal PINK1-dependent phosphorylation and activation

Abstract

Mutations in the mitochondrial protein kinase PINK1 are associated with autosomal recessive Parkinson disease (PD). We and other groups have reported that PINK1 activates Parkin E3 ligase activity both directly via phosphorylation of Parkin serine 65 (Ser(65))--which lies within its ubiquitin-like domain (Ubl)--and indirectly through phosphorylation of ubiquitin at Ser(65). How Ser(65)-phosphorylated ubiquitin (ubiquitin(Phospho-Ser65)) contributes to Parkin activation is currently unknown. Here, we demonstrate that ubiquitin(Phospho-Ser65) binding to Parkin dramatically increases the rate and stoichiometry of Parkin phosphorylation at Ser(65) by PINK1 in vitro. Analysis of the Parkin structure, corroborated by site-directed mutagenesis, shows that the conserved His302 and Lys151 residues play a critical role in binding of ubiquitin(Phospho-Ser65), thereby promoting Parkin Ser(65) phosphorylation and activation of its E3 ligase activity in vitro. Mutation of His302 markedly inhibits Parkin Ser(65) phosphorylation at the mitochondria, which is associated with a marked reduction in its E3 ligase activity following mitochondrial depolarisation. We show that the binding of ubiquitin(Phospho-Ser65) to Parkin disrupts the interaction between the Ubl domain and C-terminal region, thereby increasing the accessibility of Parkin Ser(65). Finally, purified Parkin maximally phosphorylated at Ser(65) in vitro cannot be further activated by the addition of ubiquitin(Phospho-Ser65). Our results thus suggest that a major role of ubiquitin(Phospho-Ser65) is to promote PINK1-mediated phosphorylation of Parkin at Ser(65), leading to maximal activation of Parkin E3 ligase activity. His302 and Lys151 are likely to line a phospho-Ser(65)-binding pocket on the surface of Parkin that is critical for the ubiquitin(Phospho-Ser65) interaction. This study provides new mechanistic insights into Parkin activation by ubiquitin(Phospho-Ser65), which could aid in the development of Parkin activators that mimic the effect of ubiquitin(Phospho-Ser65).

Keywords: PINK1; Parkin; Parkinson's disease; phosphorylation; ubiquitin.

Figure 1

Ubiquitin^{Phospho-Ser65} primes Parkin for phosphorylation by PINK1

1. Ubiquitin^{Phospho-Ser65} enhances Parkin phosphorylation by PINK1. The effects of ubiquitin^{Phospho-Ser65} (left), wild-type (WT) (right) and Ser65Ala (S65A) (middle) ubiquitin on Parkin phosphorylation were investigated in a kinase assay. The indicated ubiquitin species was incubated with WT or kinase-inactive (KI) MBP-TcPINK1, Parkin and Mg²⁺ [γ-³²P] ATP for 60 min. Parkin concentration was kept constant, whilst ubiquitin concentration was varied, reaching molar ratios indicated above the gel. Assays were terminated by the addition of LDS loading buffer, and products were analysed by SDS–PAGE. Proteins were detected by Colloidal Coomassie Blue staining (top panel), and incorporation of [γ-³²P] ATP was detected by autoradiography (bottom panel). Data show mean of two trials ± s.d., n = 2 for each condition.

2. Selectivity of ubiquitin^{Phospho-Ser65}-enhanced Parkin phosphorylation. Under similar conditions, the effects of ubiquitin^{Phospho-Ser65} on phosphorylation of various PINK1 substrates were assessed in a kinase assay, where equimolar amounts of Parkin, GST-Ubl and GST-HAX1 (all expressed in E. coli) were used in the presence or absence of ubiquitin^{Phospho-Ser65} and WT TcPINK1. The incorporation of radioactive [γ-³²P] ATP into substrate proteins was measured and is displayed above the panels. Proteins were detected by Colloidal Coomassie Blue staining (top panel), and incorporation of [γ-³²P] ATP was detected by autoradiography (bottom panel). Broken dividing lines indicate separate gels; asterisks denote the kinase band. The molecular mass in kDa is indicated. Data show mean of two trials ± s.d., n = 2 for each condition.

3. Ubiquitin^{Phospho-Ser65}, but not Ubl^{Phospho-Ser65}, enhances full-length Parkin phosphorylation by PINK1. Kinase assays in the presence or absence of ubiquitin^{Phospho-Ser65} or Ubl^{Phospho-Ser65} were carried out as in (A). Data are representative of two independent experiments.

Ubiquitin dimers, tetramers or a mono-ubiquitylated substrate are capable of priming Parkin phosphorylation by PINK1

1. A, B (A) The effects of ubiquitin dimers of each linkage type (Met1 (M1), Lys6 (K6), Lys11 (K11), Lys27 (K27), Lys 29 (K29), Lys33 (K33), Lys48 (K48) and Lys63 (K63), ubiquitin tetramers with M1, K6, K11, K29, K33, K48 and K63 linkages and (B) the model mono-ubiquitylated substrate, Dac-ubiquitin (Dac-ub) were assessed in a PINK1 kinase assay similar to that in Fig1. Assays were analysed by SDS/PAGE. Proteins were detected by Colloidal Coomassie Blue staining (top panel), and incorporation of [γ-³²P] ATP was detected by autoradiography (bottom panel).

Figure 2

Identification of Parkin histidine 302 and lysine 151 as critical residues mediating ubiquitin^{Phospho-Ser65} interaction and activation of Parkin

1. Structure of Parkin modified from displaying the location of sulphate-containing pockets surrounded by the following residues: Pocket 1 (K161/R163/K211); Pocket 2 (K151/H302/R305/Q316); and Pocket 3 (R455).

2. His302 and Lys151 are critical for mediating ubiquitin^{Phospho-Ser65}-enhanced phosphorylation of Parkin by TcPINK1. Wild-type (WT), full-length Parkin or the indicated Alanine mutant for each residue of the sulphate-containing pocket was incubated with wild-type (WT) MBP-TcPINK1, Parkin and Mg²⁺ [γ-³²P] ATP in the presence or absence of ubiquitin^{Phospho-Ser65}. Proteins were detected by Colloidal Coomassie Blue staining (top panel) and Parkin phosphorylation levels assessed by incorporation of [γ-³²P] ATP detected by autoradiography (bottom panel) and displayed above the panel. Data show mean of two trials ± s.d., n = 2 for each condition.

3. Parkin His302Ala and Lys151Ala mutants disrupt ubiquitin^{Phospho-Ser65} mediated activation of Parkin E3 ligase activity. WT Parkin and putative Pocket 2 mutants (K151A, H302A, R305A, Q316A) were assessed for activity via ubiquitylation assays. Each reaction contained 0.05 mM ubiquitin comprising 25 μg of FLAG-ubiquitin (Boston Biochem) mixed with 5 μg of ubiquitin^{Phospho-Ser65} or non-phospho-ubiquitin. Parkin activity was evaluated by subjecting ubiquitylation reactions to analysis by SDS–PAGE and immunoblotting as follows: ubiquitin (anti-FLAG-HRP antibody), Parkin (anti-Parkin antibody) and Miro1 (anti-SUMO1 antibody). Data are representative of five independent experiments.

4. Parkin His302Ala and Lys151Ala mutants hinder Parkin’s ability to interact with ubiquitin^{Phospho-Ser65} and discharge ubiquitin from a loaded E2. WT Parkin and Pocket 2 mutants were assessed for their ability to discharge ubiquitin from a loaded UbcH7 (E2-Ub) enzyme with or without ubiquitin^{Phospho-Ser65} or non-phospho-ubiquitin. Reactions were subjected to SDS–PAGE analysis in the absence of any reducing agent. The activity was assessed by change in UbcH7 (E2)/UbcH7-Ub (E2-Ub) ratio. The quantification of Coomassie bands was performed by LICOR and is presented above the panel. Data show mean of three trials ± s.d., n = 1 for each condition.

Analysis of stability and ubiquitin^{Phospho-Ser65}-independent E3 ligase activity of “Pocket 2” mutants of Parkin

1. Thermal denaturation curves obtained by differential scanning fluorimetry of wild-type (WT), and His302Ala (H302A)- and Lys151Ala (K151A)-mutant Parkin. Results are displayed as the differential of the fluorescence in arbitrary units divided by the differential of the temperature, plotted against temperature. Inset: the minimum of each curve indicates the melting point (Tm). Summary of melting points (Tm) of each protein as follows: WT (59°C); H302A (59.5°C); and K151A (59.5°C).

2. PINK1-dependent full-length Parkin E3 ligase activity mediated via phosphorylation of Ubl Ser65 and constitutive basal E3 ligase activity mediated by Ubl-deleted Parkin (ΔUbl; residues 80-465) are not affected by His302Ala (H302A) mutation. A 2 μg amount of wild-type full-length or ΔUbl-Parkin-biotin and H302A full-length or ΔUbl-Parkin was incubated with 1 μg of wild-type (WT), kinase-inactive (KI) or no TcPINK1 in an E3 ligase assay. Reactions were terminated after 60 min by the addition of LDS loading buffer and analysed by SDS/PAGE. Ubiquitin and Parkin were detected using anti-FLAG and anti-Parkin antibodies, respectively.

Figure 3

Parkin His302 and Lys151 are required for optimal binding with ubiquitin^{Phospho-Ser65}

1. Parkin H302A and K151A mutants exhibit marked reduction in binding to ubiquitin^{Phospho-Ser65} compared to wild-type (WT) Parkin. Isothermal calorimetry analysis of Parkin WT (left), K151A mutant (middle) and H302A mutant (right) with ubiquitin^{Phospho-Ser65}.

2. Table showing the K_d values (in μM), ΔH values (in kcal/mol), ΔS values (in kcal/mol deg) and −TΔS (in kcal/mol) derived from the graphs. Data are representative of two independent experiments.

Analysis of heterodimer formation between Ser⁶⁵-phosphorylated Dac-ubiquitin and Parkin using gel filtration chromatography About 100 μg of Dac-ubiquitin or Dac-ubiquitin^{Phospho-Ser65} (both ∼38.3 kDa) was incubated with or without 100 μg of wild-type Parkin (WT) or Parkin H302A (both ∼51.6 kDa) for 30 min and subjected to chromatography on a Superdex 200 Increase column (GE Healthcare Life Sciences). Parkin eluted at 13.8 ml and Dac-ubiquitin at 15.1 ml. The dimer eluted at 12.9 ml. The x-axis is millilitre elution volume, and the y-axis is arbitrary milli absorption units.

1. Profile of Dac-ubiquitin.

2. Profile of Dac-ubiquitin^{Phospho-Ser65}.

3. Profile of WT Parkin.

4. Profile of Parkin H302A mutant.

5. Profile of WT Parkin and Dac-ubiquitin.

6. Profile of WT Parkin and Dac-ubiquitin ^{Phospho-Ser65}.

7. Profile of H302A Parkin and Dac-ubiquitin^{Phospho-Ser65}.

8. Profile of H302A Parkin and Dac-ubiquitin.

9. Profile of S65A Parkin.

10. Profile of S65A Parkin and Dac-ubiquitin.

11. Profile of S65A Parkin and Dac-ubiquitin^{Phospho-Ser65}.

Identification of Parkinson’s disease-associated mutants that disrupt ubiquitin^{Phospho-Ser65}-enhanced phosphorylation of Parkin by TcPINK1

Schematic of Parkin domain and location of disease-associated Parkin mutants (upper panel). Wild-type (WT) full-length Parkin or the indicated disease point mutant was incubated with wild-type MBP-TcPINK1 and Mg²⁺ [γ-³²P] ATP in the presence or absence of ubiquitin^{Phospho-Ser65}. Proteins were detected by Colloidal Coomassie Blue staining (top panel) and Parkin phosphorylation levels assessed by incorporation of [γ-³²P] ATP detected by autoradiography (bottom panel) and displayed above the panel (lower Panel).

Figure 4

Parkin His302 is required for optimal phosphorylation of Parkin at Ser⁶⁵ in cells upon CCCP-stimulated PINK1 activation

1. Parkin H302A-mutant displays marked decrease in Parkin Ser⁶⁵ phosphorylation upon PINK1 activation. Wild-type HeLa (upper panel) or PINK1 knockout HeLa cells (lower panel) were transfected with untagged wild-type (WT), and Ser65Ala (S65A)- or His302Ala-mutant Parkin and stimulated with 10 μM of CCCP or DMSO for 6 h in triplicates. The lysates were subjected to immunoblotting as follows: Parkin Ser⁶⁵ phosphorylation (anti-phospho-Ser65 antibody), Parkin (anti-Parkin antibody), actin (anti-actin antibody) and PINK1 (anti-PINK1 antibody). Data are representative of three independent experiments.

2. Parkin H302A-mutant disrupts mitochondrial accumulation of Parkin Ser⁶⁵ phosphorylation. Wild-type HeLa cells stably expressing untagged wild-type (WT) (top row and second row), and Ser65Ala (S65A) (third row)- or His302Ala (H302A) (fourth row)-mutant Parkin were stimulated with 10 μM of CCCP for 6 h. HeLa cells expressing WT Parkin were also treated with DMSO for 6 h (second row). Cells were stained for Parkin Ser65 phosphorylation (anti-phospho-Ser65 antibody) or total Parkin (anti-Parkin antibody); mitochondria were labelled using MITO-ID® Red. Data representative of four independent experiments.

Parkin His302 is required for optimal activation of Parkin ubiquitin E3 ligase activity at mitochondria in response to PINK1 activation by CCCP

Wild-type HeLa cells were transfected with untagged wild-type (WT), and Ser65Ala (S65A)-, His302Ala (H302A)- or Cys431Phe (C431F)-mutant Parkin and stimulated with 10 μM of CCCP or DMSO for 6 h. Mitochondrial enriched extracts were incubated with a ubiquitin-binding resin derived from His-Halo-Ubiquilin UBA-domain tetramer (UBA^UBQLN1). Captured ubiquitylated proteins were subjected to immunoblotting with CISD1 and ubiquitin antibodies. In parallel, mitochondrial input extracts were immunoblotted with Tomm70 antibody.

Figure 5

Ubiquitin^{Phospho-Ser65}-mediated disruption of the Ubl domain with ΔUbl Parkin is dependent on residue His302 An AlphaScreen™ binding assay of Parkin Ubl domain and C-terminus of Parkin were established. GST-Ubl (residues 1-76) and wild-type C-terminal Parkin-biotin (residues 80-465) were incubated with streptavidin-coated donor beads and glutathione acceptor beads for 60 min.

1. A, B Ubiquitin^{Phospho-Ser65} (ub^PhosSer65) disrupts maximal binding signal of Parkin and Ubl interaction of wild-type (wt) (A), but not H302A (H302A) Parkin (B).

2. C, D Ubl^{Phospho-Ser65} (Ubl^PhosSer65) exhibits substantially reduced ability to disrupt interaction between GST-Ubl and wild-type (C) and H302A Parkin (D).

3. E Table showing the IC₅₀ values (in nM) derived from graphs, where the interaction inhibition is incomplete, and not-determined (ND) is indicated.

Data information: Data show mean of one trial ± s.d., n = 3 for each condition (A–D).

Figure 6

Purified Parkin phosphorylated at Ser65 exhibits constitutive E3 ligase activity that is no longer sensitive to ubiquitin^{Phospho-Ser65}

1. Parkin phosphorylated at Ser65 exhibits significant constitutive activity. About 2 μg of full-length wild-type (Parkin WT) or phosphorylated at Ser65 (P-Parkin WT) Parkin was analysed using E3 ligase assay with increasing amounts of non-phospho-ubiquitin or ubiquitin^{Phospho-Ser65} as indicated. Parkin activity was evaluated by immunoblotting as follows: ubiquitin (anti-FLAG-HRP antibody), Parkin (anti-Parkin antibody) and Miro1 (anti-SUMO1 antibody). The molecular mass in kDa is indicated. Data are representative of three independent experiments.

2. Dephosphorylation of Parkin at Ser65 leads to reversal of Parkin E3 ligase constitutive activity. About 1 μg of full-length Parkin phosphorylated at Ser65 was subjected to increasing amounts of alkaline phosphatase as indicated. Reactions were then incubated with ubiquitylation assay components (E1 and UbcH7) in the presence of 0.05 mM FLAG-ubiquitin. Reactions were terminated after 60 min by the addition of LDS loading buffer. The effects on Parkin E3 ligase activity were evaluated by ubiquitin chain formation ubiquitylation and Parkin autoubiquitylation evaluated by immunoblotting as follows: ubiquitin (anti-FLAG-HRP antibody) and Parkin (anti-Parkin antibody). The degree of Parkin Ser⁶⁵ de-phosphorylation was monitored using anti-pSer65 Parkin antibody. Data are representative of four independent experiments.

Figure 7

Priming model of Parkin activation by PINK1-dependent phosphorylation of ubiquitin Upon activation of PINK1 by mitochondrial depolarisation, PINK1 can phosphorylate ubiquitin to generate phospho-ubiquitin^{Phospho-Ser65}. Binding of ubiquitin^{Phospho-Ser65} to non-phosphorylated Parkin can disrupt intramolecular interaction of Ubiquitin-like (Ubl) domain to Parkin C-terminus. The Ser65 residue on Ubl becomes more accessible for PINK1-dependent phosphorylation leading to an open and active conformation of Parkin.