Parkin interacts with Ambra1 to induce mitophagy

Abstract

Mutations in the gene encoding Parkin are a major cause of recessive Parkinson's disease. Recent work has shown that Parkin translocates from the cytosol to depolarized mitochondria and induces their autophagic removal (mitophagy). However, the molecular mechanisms underlying Parkin-mediated mitophagy are poorly understood. Here, we investigated whether Parkin interacts with autophagy-regulating proteins. We purified Parkin and associated proteins from HEK293 cells using tandem affinity purification and identified the Parkin interactors using mass spectrometry. We identified the autophagy-promoting protein Ambra1 (activating molecule in Beclin1-regulated autophagy) as a Parkin interactor. Ambra1 activates autophagy in the CNS by stimulating the activity of the class III phosphatidylinositol 3-kinase (PI3K) complex that is essential for the formation of new phagophores. We found Ambra1, like Parkin, to be widely expressed in adult mouse brain, including midbrain dopaminergic neurons. Endogenous Parkin and Ambra1 coimmunoprecipitated from HEK293 cells, SH-SY5Y cells, and adult mouse brain. We found no evidence for ubiquitination of Ambra1 by Parkin. The interaction of endogenous Parkin and Ambra1 strongly increased during prolonged mitochondrial depolarization. Ambra1 was not required for Parkin translocation to depolarized mitochondria but was critically important for subsequent mitochondrial clearance. In particular, Ambra1 was recruited to perinuclear clusters of depolarized mitochondria and activated class III PI3K in their immediate vicinity. These data identify interaction of Parkin with Ambra1 as a key mechanism for induction of the final clearance step of Parkin-mediated mitophagy.

Figure 1.

Parkin binds to Ambra1. A, TAP flow chart. B, Schematic of untagged and TAP-tagged Parkin. RING, RING finger domain; IBR, in-between-ring domain. C, Enrichment of TAP-tagged Parkin during TAP. Extracts of HEK293 cells transfected with untagged or TAP-tagged Parkin were subjected to TAP. Extract, Initial cell extract; FT, flow-through; El, eluate. The successive fractions were analyzed by SDS-PAGE and Western blot with anti-Parkin. The same amount of total protein (0.2 μg) was loaded on the gel for each fraction. D, A total of 4000 cm² of HEK293 cell culture was transfected with untagged or TAP-tagged Parkin, and extracts were subjected to TAP. Final TAP eluates were analyzed by SDS-PAGE and Western blot with anti-Parkin or Coomassie blue staining. 1, Untagged Parkin; 2, empty lane; 3, TAP-tagged Parkin; asterisk, Parkin band; arrow, Ambra1-containing band, as determined by mass spectrometry. E, F, Control experiments for the anti-Ambra1 antibody (Strategic Diagnostics) used for WB in this study. E, HEK293 cells were transiently transfected with empty vector or FLAG-tagged Ambra1 (20 ng/cm² cDNA). Protein extracts were analyzed with SDS-PAGE and WB with anti-Ambra1 or anti-FLAG. Anti-Ambra1 recognized endogenous Ambra1 and transfected Ambra1, whereas anti-FLAG only detected transfected Ambra1. F, HEK293 cells were untransfected (Untransf.) or transfected with control (Ctrl.) siRNA, Ambra1 siRNA 1, or Ambra1 siRNA 2. At 24 h after transfection, SDS-PAGE and WB were performed with the indicated antibodies. The observed knockdown of the endogenous anti-Ambra1 band after Ambra1 siRNA 1 or 2 transfection confirmed the specificity of this antibody. G, Transfection of TAP-tagged versus untagged Parkin in HEK293 cells followed by TAP, SDS-PAGE of the TAP eluates, and WB with anti-Ambra1 or anti-β-actin. H, I, Extracts from untransfected HEK293 cells (H) or adult mouse brain (I) were subjected to immunoprecipitation (IP) with anti-Ambra1 or control IgG, followed by WB of input and IP fractions with the indicated antibodies. Presenilin 1 in H and calretinin in I are shown as negative controls. J, K, Extracts from untransfected HEK293 cells (J) or mouse brain (K) were subjected to IP with anti-Parkin or control IgG, followed by WB of input and IP fractions with the indicated antibodies. L, Parkin deletion constructs used in this study. All deletion constructs had an N-terminal FLAG tag. M, HEK293 cells were cotransfected with untagged full-length Ambra1 and each of the indicated FLAG-tagged Parkin deletion constructs. Cell extracts were subjected to IP with anti-Ambra1, followed by WB of input and IP fractions with the indicated antibodies. L, Linker domain. N, Ambra1 deletion constructs used in this study. All constructs had a C-terminal Myc–FLAG tag. WD40, WD40 repeat-containing domain. O, Extracts from HEK293 cells cotransfected with untagged Parkin and each of the indicated FLAG-tagged Ambra1 deletion constructs were subjected to IP with anti-Parkin, followed by WB of input and IP fractions with the indicated antibodies.

Figure 2.

Anatomical, cellular, and subcellular distribution of Ambra1. A, Extracts from 2-month-old mouse brain regions were analyzed using SDS-PAGE and Western blot with the indicated antibodies. Mes., Mesencephalon; Cbl., cerebellum; Thal., thalamus; Striat., striatum; Olf., olfactory bulb. B, C, Control experiments for the anti-Ambra1 antibody (Covalab) used for immunocytochemistry in this study. B, HEK293 cells were transiently transfected with FLAG-tagged Ambra1 and double labeled with anti-Ambra1 and anti-FLAG. Note that endogenous Ambra1 in HEK293 cells is below the immunocytochemical detection threshold of this anti-Ambra1 antibody. C, SH-SY5Y cells were untransfected (Untransf.) or transfected with control (Ctrl.) siRNA, Ambra1 siRNA 1, or Ambra1 siRNA 2. At 24 h after transfection, cells were immunostained with anti-Ambra1. Note that endogenous Ambra1 is detectable in this neural cell line. The observed suppression of the endogenous anti-Ambra1 signal after Ambra1 siRNA 1 or 2 transfection confirmed the specificity of this antibody. D–H, Embryonic (D, E, G, H) or postnatal (F) mesencephalic cultures on day 6 in vitro were double labeled with anti-Ambra1 and anti-MAP2, a neuronal marker (D), anti-tyrosine hydroxylase (TH), a marker of dopaminergic neurons (E, F), anti-GFAP, an astrocyte marker (G), or anti-CNPase, an oligodendrocyte marker (H). Arrowhead in D indicates a non-neuronal (MAP2-negative) cell. Microscope settings in G and H are identical to those used in D–F, so that the intensity of Ambra1 staining in glial cells can be compared with that in neurons. I, Embryonic mesencephalic cultures were double labeled with anti-Ambra1 and anti-Parkin, revealing partial colocalization of the endogenous proteins in the neuronal cytosol. J–N, Embryonic mesencephalic cultures were double labeled with anti-Ambra1 and either anti-cytochrome c, a mitochondrial marker (J), anti-KDEL, an ER marker (K), anti-GM130, a Golgi marker (L), anti-LC3, an autophagosome marker (M), or anti-LAMP-1, a lysosomal marker (N). Cell nuclei were stained with DAPI or TOTO-3. Scale bars, 10 μm.

Figure 3.

Parkin does not ubiquitinate Ambra1. HeLa cells (A, B) or HEK293 cells (C) were transfected with various combinations of HA-tagged ubiquitin (HA-Ub), FLAG-tagged IKKγ, Parkin, and Ambra1, as indicated. At 24 h after transfection, extracts were made in denaturing conditions. After dilution in non-denaturing buffer, immunoprecipitation (IP) was performed with anti-FLAG, anti-Ambra1, or control IgG. The IP and input samples were resolved by SDS-PAGE and Western blot with the antibodies indicated to the right of the blots. In B, a short and a more prolonged film exposure of the same anti-HA blot is shown.

Figure 4.

Mitochondrial depolarization promotes the interaction of endogenous Parkin and Ambra1. A, B, Untransfected HEK293 cells (A) or untransfected SH-SY5Y cells (B) were treated with CCCP (10 μm in A; 25 μm in B) or DMSO for 12 h, followed by immunoprecipitation (IP) of the cell extracts with anti-Ambra1 or control IgG. The input and IP samples were resolved by SDS-PAGE and Western blot with the indicated antibodies. C, Extracts from SH-SY5Y cells were fractionated into a cytosolic (C) and a mitochondria-enriched (M) fraction. After loading the same total amount of protein on the gel for each fraction, SDS-PAGE and immunoblotting were performed for the mitochondrial marker Tom20 and the cytosolic marker GAPDH. D, Untransfected SH-SY5Y cells were treated with DMSO or CCCP (25 μm) for 16 h. Next, cell extracts were fractionated into cytosolic and mitochondrial fractions, as in C. Immunoprecipitation with anti-Ambra1 was then performed on the same total amount of protein for the DMSO- and CCCP-treated fractions, followed by WB of the immunoprecipitates with anti-Parkin.

Figure 5.

Ambra1 is not required for Parkin translocation to depolarized mitochondria. A, B, At 24 h after transfection with Parkin, HeLa cells were treated with DMSO, CCCP (10 μm), or valinomycin (1 μm) for 3 h and immunostained for Parkin and Tom20, a mitochondrial marker. B, Quantification of the percentage of Parkin-expressing cells in which Parkin strongly colocalized with mitochondria (n = 6 for DMSO; n = 3 for CCCP and for valinomycin). *p < 0.001 compared with DMSO. C, At 24 h after cotransfection with Parkin and Ambra1, HeLa cells were treated with DMSO or CCCP (10 μm) for 3 h and immunostained for Parkin, Ambra1, and Tom20. D, Untransfected SH-SY5Y cells were treated with DMSO or CCCP (25 μm) for 3 h and immunostained for Parkin and Tom20. Arrowheads in D indicate clusters of Parkin immunoreactivity that colocalize with mitochondria. E, F, HeLa cells were untransfected (No transf.) or transfected with control siRNA, Ambra1 siRNA 1, or Ambra1 siRNA 2. At 24 h after transfection, SDS-PAGE and Western blot were performed with the indicated antibodies. F, The Ambra1 protein level after siRNA treatment was quantified and normalized to the level after control siRNA transfection (n = 4). *p < 0.001 compared with control siRNA. G, H, HeLa cells were cotransfected with Parkin and the indicated siRNA. After 24 h, cells were treated for 3 h with CCCP (10 μm) and immunostained for Parkin and Tom20. H, Quantification of the percentage of Parkin-expressing HeLa cells in which Parkin strongly colocalized with mitochondria (n = 3). I, J, SH-SY5Y cells were transfected with the indicated siRNAs. After 24 h, cells were treated for 3 h with CCCP (25 μm) and immunostained for endogenous Parkin and Tom20. Arrowheads in I indicate clusters of Parkin immunoreactivity that colocalize with mitochondria. J, Quantification of the percentage of SH-SY5Y cells with Parkin clusters colocalizing with mitochondria (n = 3). Scale bars, 10 μm.

Figure 6.

Mitochondrial depolarization induces Parkin-dependent mitophagy. A, B, At 24 h after transfection with Parkin, HeLa cells were treated with DMSO, CCCP (10 μm), or valinomycin (1 μm) for 24 h and immunostained for Parkin and Tom20. B, Percentage of Parkin-positive and Parkin-negative cells without detectable Tom20 immunoreactivity after 24 h treatment with CCCP or DMSO (n = 11 for DMSO; n = 8 for CCCP; n = 3 for valinomycin). ^#p < 0.001 compared with CCCP-treated Parkin-negative cells. *p < 0.001 compared with DMSO-treated Parkin-positive cells. ^§p < 0.001 compared with valinomycin-treated Parkin-negative cells. C, D, At 24 h after transfection with Parkin, HeLa cells were treated with DMSO or CCCP (10 μm) for 24 h and immunostained for Parkin and cytochrome c. D, Percentage of Parkin-positive and Parkin-negative cells without detectable cytochrome c immunoreactivity after 24 h treatment with CCCP or DMSO (n = 3). *p < 0.001 compared with CCCP-treated Parkin-negative cells. ^#p < 0.001 compared with DMSO-treated Parkin-positive cells. E, F, At 24 h after transfection with Parkin, HeLa cells were treated with DMSO or CCCP (10 μm) for 24 h and immunostained for Parkin and DNA polymerase γ (POLG). F, Percentage of Parkin-positive and Parkin-negative cells without detectable POLG immunoreactivity after 24 h treatment with CCCP or DMSO (n = 3). *p < 0.001 compared with CCCP-treated Parkin-negative cells. ^#p < 0.001 compared with DMSO-treated Parkin-positive cells. Arrows in A, C, and E indicate Parkin-positive cells, and arrowheads indicate Parkin-negative cells. G, At 24 h after transfection with Parkin, HeLa cells were treated for 24 h with CCCP (10 μm) alone, with CCCP (10 μm) and bafilomycin A1 (100 nm), or with CCCP (10 μm) and 3-methyladenine (3MA; 10 mm) and immunostained for Parkin and Tom20. The percentage of Parkin-positive and Parkin-negative cells without detectable Tom20 immunoreactivity was quantified (n = 3). *p < 0.001 compared with Parkin-positive cells treated with CCCP alone. H, At 24 h after transfection with Parkin, HeLa cells were treated with CCCP (10 μm) for 24 h and immunostained for Parkin and Pex14p, a peroxisomal marker. I, Untransfected SH-SY5Y cells were treated with DMSO or CCCP (25 μm) for 24 h and immunostained for Tom20. J, K, Untransfected SH-SY5Y cells were treated for 24 h with DMSO, with CCCP (25 μm), or with CCCP (25 μm) and bafilomycin A1 (Bafilom.; 100 nm). Cell extracts were analyzed by SDS-PAGE and Western blot with the indicated antibodies. K, The amount of COX IV was quantified and normalized to the amount of COX IV in the DMSO condition (n = 7 for DMSO and CCCP; n = 4 for CCCP + bafilomycin A1). *p < 0.005 compared with DMSO. ^#p < 0.001 compared with CCCP + bafilomycin A1. Scale bars, 10 μm.

Figure 7.

Ambra1 contributes to Parkin-mediated mitophagy. A, B, HeLa cells were cotransfected with Parkin and the indicated siRNAs, treated for 24 h with CCCP (10 μm) and stained for Parkin and Tom20. Arrows in A indicate Parkin-positive cells, and arrowheads indicate Parkin-negative cells. B, Quantification of the effect of siRNA-mediated Ambra1 downregulation on CCCP-induced mitophagy in Parkin-positive cells (n = 6–9). *p < 0.001 compared with control siRNA condition after CCCP. C, D, SH-SY5Y cells were transfected with the indicated siRNAs, treated for 24 h with CCCP (25 μm), and stained for Tom20. Microscope settings in C are identical for the different siRNA conditions. Arrow in C indicates a cell with no detectable mitochondrial staining. D, Quantification of the effect of siRNA-mediated Ambra1 downregulation on CCCP-induced mitochondrial clearance in SH-SY5Y cells (n = 3). *p < 0.05 compared with control siRNA condition after CCCP. E, At 24 h after transfection with Ambra1 alone, HeLa cells were treated with DMSO or CCCP (10 μm) for 24 h and then immunostained for Ambra1 and Tom20. Arrows in E indicate Ambra1-overexpressing cells. F, At 24 h after cotransfection with Myc-tagged Ambra1 and Parkin, HeLa cells were treated with DMSO or CCCP (10 μm) for 24 h and immunostained for Myc, Parkin, and Tom20. Arrows in F indicate cells overexpressing both Ambra1 and Parkin, and arrowheads indicate cells overexpressing only Parkin. G, HeLa cells were transfected with various combinations of Myc-tagged Ambra1, Parkin, and empty vector (EV), as indicated. At 24 h after transfection, cells were treated with CCCP (10 μm) for 24 h. After immunostaining for Myc, Parkin, and Tom20, the percentage of transfected cells devoid of mitochondrial staining was quantified (n = 3–5). *p < 0.001 compared with cells transfected with EV alone. ^#p < 0.001 compared with cells transfected with Parkin and EV. Scale bars, 10 μm.

Figure 8.

Activation of class III PI3K around clusters of depolarized mitochondria. A, At 24 h after transfection with p40(phox)PX–EGFP, HeLa cells were treated for 1 h with 75 nm of the PI3K inhibitor wortmannin or DMSO. B, HeLa cells were cotransfected with p40(phox)PX–EGFP and Parkin. At 24 h after transfection, cells were treated with DMSO or CCCP (10 μm) for 16 h and immunostained for Parkin and Tom20. The bottom row is a magnification of the boxed area in the row above. C, SH-SY5Y cells were transfected with p40(phox)PX–EGFP. After 24 h, the cells were treated with DMSO or CCCP (25 μm) for 16 h and immunostained for Tom20. D, Quantification of the experiment illustrated in B. The number of HeLa cells with perinuclear mitochondrial clusters was quantified as a percentage of the total number of mitochondria-containing cells (black bars; n = 3). Also, the number of cells with perinuclear mitochondrial clusters in contact with p40(phox)PX–EGFP hotspots was quantified as a percentage of the total number of mitochondria-containing cells (white bars; n = 3). *p < 0.001 compared with DMSO-treated Parkin-positive cells. ^#p < 0.001 compared with CCCP-treated Parkin-negative cells. Scale bars, 10 μm.

Figure 9.

Ambra1 is recruited in a Parkin-dependent manner to perinuclear clusters of depolarized mitochondria. A, At 24 h after transfection with Ambra1 alone (rows 1, 3) or cotransfection with Ambra1 and Parkin (rows 2, 4, 5), HeLa cells were treated with DMSO or CCCP (10 μm) for 16 h and immunostained for Parkin, Tom20, and Ambra1. The bottom row is a magnification of the boxed area in the row above. B, Untransfected SH-SY5Y cells were treated with DMSO or CCCP (25 μm) for 16 h and immunostained for cytochrome c and Ambra1. The bottom row is a magnification of the boxed area in the row above. Scale bars, 10 μm.

Figure 10.

Ambra1 locally activates class III PI3K around perinuclear clusters of depolarized mitochondria. A, HeLa cells were cotransfected with Parkin, p40(phox)PX–EGFP, and the indicated siRNA. At 24 h after transfection, cells were treated for 16 h with CCCP (10 μm) and stained for Parkin and Tom20. B, Quantification of the experiments illustrated in A. The number of Parkin-positive cells with perinuclear mitochondrial clusters was quantified as a percentage of the total number of mitochondria-containing Parkin-positive cells (black bars; n = 3). Also, the number of Parkin-positive cells with perinuclear mitochondrial clusters in contact with p40(phox)PX–EGFP hotspots was quantified as a percentage of the total number of mitochondria-containing Parkin-positive cells (white bars; n = 3). *p = 0.001 compared with control siRNA-treated cells. Scale bars, 10 μm.