PTEN-L is a novel protein phosphatase for ubiquitin dephosphorylation to inhibit PINK1-Parkin-mediated mitophagy

Abstract

Mitophagy is an important type of selective autophagy for specific elimination of damaged mitochondria. PTEN-induced putative kinase protein 1 (PINK1)-catalyzed phosphorylation of ubiquitin (Ub) plays a critical role in the onset of PINK1-Parkin-mediated mitophagy. Phosphatase and tensin homolog (PTEN)-long (PTEN-L) is a newly identified isoform of PTEN, with addition of 173 amino acids to its N-terminus. Here we report that PTEN-L is a novel negative regulator of mitophagy via its protein phosphatase activity against phosphorylated ubiquitin. We found that PTEN-L localizes at the outer mitochondrial membrane (OMM) and overexpression of PTEN-L inhibits, whereas deletion of PTEN-L promotes, mitophagy induced by various mitochondria-damaging agents. Mechanistically, PTEN-L is capable of effectively preventing Parkin mitochondrial translocation, reducing Parkin phosphorylation, maintaining its closed inactive conformation, and inhibiting its E3 ligase activity. More importantly, PTEN-L reduces the level of phosphorylated ubiquitin (pSer65-Ub) in vivo, and in vitro phosphatase assay confirms that PTEN-L dephosphorylates pSer65-Ub via its protein phosphatase activity, independently of its lipid phosphatase function. Taken together, our findings demonstrate a novel function of PTEN-L as a protein phosphatase for ubiquitin, which counteracts PINK1-mediated ubiquitin phosphorylation leading to blockage of the feedforward mechanisms in mitophagy induction and eventual suppression of mitophagy. Thus, understanding this novel function of PTEN-L provides a key missing piece in the molecular puzzle controlling mitophagy, a critical process in many important human diseases including neurodegenerative disorders such as Parkinson's disease.

Fig. 1

PTEN-L resides at the outer mitochondrial membrane. a Domain structure of PTEN-L protein. b Cell fractionation was performed to isolate mitochondria in HeLa cells. Tim23 and GAPDH were used as mitochondrial and cytosolic markers, respectively. C cytosol, M mitochondria. c Relative ratios of PTEN-L and PTEN in cytosol and mitochondria were calculated with normalization of volume. d Topology assay showing PTEN-L localization at the OMM. Mitochondria were isolated from YFP-Parkin-HeLa cells stably expressing Flag-PTEN-L and treated with different doses of proteinase K and digitonin. e Immunogold EM. YFP-Parkin-HeLa cells stably expressing Flag-PTEN-L were treated without (A, B) or with CCCP (5 µM) (C, D) for 4 h. Cells were subjected to immunoelectron microscopy with anti-Flag antibody. Panels (B) and (D) are the magnified images of boxes in (A) and (C), respectively. Arrows indicate ER structure

Fig. 2

PTEN-L negatively regulates mitophagy induced by various mitochondria-damaging agents. a YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with CCCP (5 µM) for 24 h and immunoblotting for mitochondrial proteins was performed as indicated. b Quantification of mitochondrial proteins from a. c Degradation of mtDNA. Representative images of YFP-Parkin-HeLa cells with PTEN-L stable expression and control vector immunostained to label mtDNA (red) after treatment with CCCP (5 µM) for 24 h. d Quantification of mtDNA from > 300 cells per group. Scale bar, 10 µm. e Wild-type (WT) and PTEN-L KO YFP-Parkin-HeLa cells were treated with CCCP (4 µM) for 24 h and immunoblotting was performed as indicated. f Quantification of data from e. Data in b, d, f are presented as mean ± SD from three independent experiments. ***P < 0.001 (two-way ANOVA, b and f), ***P < 0.0001 (Student’s t-test, d)

Fig. 3

PTEN-L prevents Parkin mitochondrial translocation. a YFP-Parkin-HeLa cells transiently transfected with Flag-PTEN-L or control vector were treated with CCCP (5 µM) for 2 h. YFP-Parkin (green), Tom20 (red), Flag-PTEN-L (cyan). Scale bar, 10 µm. b Percentage of cells with Parkin mitochondrial translocation was quantified by counting at least 300 cells. c Wild-type (WT) and PTEN-L KO YFP-Parkin-HeLa cells were treated with CCCP (4 µM) for 30 and 90 min. YFP-Parkin (green). Scale bar, 10 µm. d Percentage of cells with Parkin mitochondrial translocation from c (CCCP 90 min) was quantified by counting at least 300 cells. e PTEN-L KO YFP-Parkin-HeLa cells were transiently transfected with mCherry-PTEN-L or control vector and treated with CCCP (4 µM) for 90 min. YFP-Parkin (green), mCherry (red). Scale bar, 10 µm. Data in b, d are presented as mean ± SD from three independent experiments. **P < 0.01, ***P < 0.001 (two-way ANOVA test)

Fig. 4

PTEN-L impairs Parkin E3 ligase activity and reduces pSer65-Parkin. a YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with CCCP (5 µM) for indicated hours. YFP-Parkin was immunoprecipitated with anti-GFP antibody. Immunoprecipitants (IPs) and whole-cell lysates (WCLs) were analyzed for YFP-Parkin, mitofusin-2 (MFN2), Tom20, PTEN-L, PTEN and tubulin. b YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with O/A (10 nM and 100 nM) for indicated hours and the WCLs were analyzed by immunoblotting as indicated. c Wild-type (WT) and PTEN-L KO YFP-Parkin-HeLa cells were treated with CCCP (5 µM) for indicated hours and the WCLs were analyzed by immunoblotting as indicated. d Wild-type (WT) and PTEN-L KO YFP-Parkin-HeLa cells were treated with O/A (10 nM and 100 nM) for indicated hours and the WCLs were analyzed by immunoblotting as indicated. e The MS/MS spectra of the Parkin peptide containing phospho-Ser65. YFP-Parkin-HeLa cells were treated with CCCP (10 µM) for 4 h. YFP-Parkin was pulled down with GFP beads and subjected to MS/MS analysis. f pSer65-Parkin was quantified using MS-based relative quantification analysis in YFP-Parkin-HeLa cells with or without PTEN-L stable expression after CCCP (10 µM) and O/A (25 nM and 250 nM) treatment. Data are presented as mean ± SD from 3 independent experiments. **P < 0.01 (Student’s t-test)

Fig. 5

PTEN-L keeps Parkin in closed conformation by enhancing the interaction of Parkin UBL and RING1 domains in a protein phosphatase activity-dependent manner. a Construction of PTEN-L truncations. PTEN-L contains an ATR region, a phosphatase domain and a C-terminal region with a C2 domain and a C-Tail domain. PTEN-L-C297S is a dual lipid-protein phosphatase-defective mutant, while PTEN-L-G302R is a lipid phosphatase-defective mutant. b, c HEK293T cells transfected with GFP-Parkin and different constructs of Flag-tagged PTEN-L were treated without or with CCCP (5 µM) for 4 h. PTEN-L was immunoprecipitated with anti-Flag beads followed by immunoblotting for GFP and Flag. d Construction of Flag-tagged Parkin truncations, including Parkin-FL (full length) and truncated Parkin constructs: UBL-R0, Exon4, R1-IBR-R2, UBL-R1-IBR-R2 (deletion of R0 domain) and IBR-R2. e HEK293T cells transfected with GFP-PTEN-L and different constructs of Flag-tagged Parkin were treated without or with CCCP (5 µM) for 4 h. PTEN-L was then immunoprecipitated with anti-GFP beads followed by immunoblotting for Flag and GFP. f HEK293T cells were transfected with Flag-PTEN-L or the two Flag-PTEN-L mutants, together with Parkin truncation mutants GFP-UBL and GST-RING1 (R1) or GST empty vector (pEBG). Cells were then treated with or without CCCP (20 µM) for 4 h. RING1 was immunoprecipitated with anti-GST beads followed by immunoblotting for GFP, Flag and GST. g YFP-Parkin-HeLa cells transiently transfected with mCherry-PTEN-L or the two mCherry-PTEN-L mutants were treated with CCCP (5 µM) for 2 h. YFP-Parkin (green), mCherry (red). Scale bar, 10 µm

Fig. 5

PTEN-L keeps Parkin in closed conformation by enhancing the interaction of Parkin UBL and RING1 domains in a protein phosphatase activity-dependent manner. a Construction of PTEN-L truncations. PTEN-L contains an ATR region, a phosphatase domain and a C-terminal region with a C2 domain and a C-Tail domain. PTEN-L-C297S is a dual lipid-protein phosphatase-defective mutant, while PTEN-L-G302R is a lipid phosphatase-defective mutant. b, c HEK293T cells transfected with GFP-Parkin and different constructs of Flag-tagged PTEN-L were treated without or with CCCP (5 µM) for 4 h. PTEN-L was immunoprecipitated with anti-Flag beads followed by immunoblotting for GFP and Flag. d Construction of Flag-tagged Parkin truncations, including Parkin-FL (full length) and truncated Parkin constructs: UBL-R0, Exon4, R1-IBR-R2, UBL-R1-IBR-R2 (deletion of R0 domain) and IBR-R2. e HEK293T cells transfected with GFP-PTEN-L and different constructs of Flag-tagged Parkin were treated without or with CCCP (5 µM) for 4 h. PTEN-L was then immunoprecipitated with anti-GFP beads followed by immunoblotting for Flag and GFP. f HEK293T cells were transfected with Flag-PTEN-L or the two Flag-PTEN-L mutants, together with Parkin truncation mutants GFP-UBL and GST-RING1 (R1) or GST empty vector (pEBG). Cells were then treated with or without CCCP (20 µM) for 4 h. RING1 was immunoprecipitated with anti-GST beads followed by immunoblotting for GFP, Flag and GST. g YFP-Parkin-HeLa cells transiently transfected with mCherry-PTEN-L or the two mCherry-PTEN-L mutants were treated with CCCP (5 µM) for 2 h. YFP-Parkin (green), mCherry (red). Scale bar, 10 µm

Fig. 6

PTEN-L dephosphorylates ubiquitin. a YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with CCCP (5 µM) for 3 h. Whole-cell lysates were analyzed by immunoblotting. b YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with CCCP (5 µM) for 3 h. Immunofluorescence staining against pSer65-Ub was performed and observed by fluorescent microscopy. pSer65-Ub (red), YFP-Parkin (green), Nucleus (DAPI, blue). Scale bar, 10 µm. c YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with O/A (25 nM and 250 nM) for indicated hours and immunoblotting was performed. d Wild-type (WT) and PTEN-L KO YFP-Parkin-HeLa cells were treated with O/A (25 nM and 250 nM) for indicated hours and immunoblotting was performed. e In vitro dephosphorylation assay. Purified pSer65-Ub was incubated with purified Flag-PTEN-L, Flag-PTEN-L-C297S or Flag-PTEN-L-G302R in phosphatase reaction buffer and ubiquitin phosphorylation level was evaluated by immunoblotting

Fig. 7

PTEN-L disrupts the feedforward mechanism in mitophagy by targeting the pSer65-Ub chains. a In vitro dephosphorylation assay using purified pSer65-tetra-Ub. Purified Flag-PTEN-L was incubated with pSer65-tetra-Ub in the phosphatase reaction buffer for 1 h at 30 °C. Calf intestinal phosphatase (CIP) was used as a positive control. b In vitro dephosphorylation assay using purified pSer65-poly-Ub chains, following the same procedure in a. λPP was used as a positive control. c YFP-Parkin-HeLa cells with PTEN-L stable expression or control vector were treated with CCCP (10 µM) for 4 h. YFP-Parkin was pulled down by GFP beads and subjected to immunoblotting. d The MS/MS spectra of the ubiquitin peptide containing phospho-Ser65. YFP-Parkin-HeLa cells were treated with CCCP (10 µM) for 4 h and YFP-Parkin was pulled down with GFP beads. e pSer65-Ub was quantified using MS-based relative quantification analysis in YFP-Parkin-HeLa cells with or without PTEN-L stable expression after CCCP (10 µM) and O/A (25 nM and 250 nM) treatment. Data are presented as mean ± SD from 3 independent experiments. ***P < 0.0001 (Student’s t-test)

Fig. 8

Illustration of the novel function of PTEN-L as a protein phosphatase in suppression of mitophagy. PTEN-L is able to counteract PINK1-mediated ubiquitin phosphorylation (pSer65-Ub), which then leads to prevention of Parkin recruitment to damaged mitochondria, reduction of pSer65-Parkin level and maintenance of Parkin in its closed inactive conformation to impair Parkin E3 ligase activity, all resulting in the disruption of the feedforward loop to inhibit mitophagy