Fig 2: Parkin Ser108phosphorylation is rapidly induced following AO and CCCP and precedes activation of PINK1 and TBK1.(A) Schematic of overall Parkin domain structure with inset of ClustalW alignment of residues 59 to 115 of human Parkin. The evolutionary conservation of the PINK1 site Ser65 and ULK1 site Ser108 in Parkin across vertebrates is shown in red. Ser108 is one of only two fully conserved serines in the recently described ACT element (7) involved in Parkin activation (residues 101 to 110). At right, structural model of Ser108 in ACT element shown in red, Arg104 shown in green is mutated in a familial Parkinson’s patient (R104W). (B) Immunoblots of HEK293T cells stably expressing YFP-Parkin was subjected to AO (2.5 μM antimycin A and 5 μM oligomycin) treatment for the times indicated. (C) Immunoblots of HEK293T cells stably expressing YFP-Parkin were subjected to 10 μM CCCP treatment for the times indicated.

Fig 3: ULK1-dependent phosphorylation of Parkin is required for maximal Parkin activity.(A) Immunoblots of HEK293T cells transfected with YFP-Parkin-C431S or YFP-Parkin-C431S/SA3. Twenty-four hours after transfection, cells were treated with CCCP (20 μM), valinomycin (5 μM), AO (2.5 μM antimycin A + 5 μM oligomycin) or vehicle (DMSO) for 2 hours. Red arrow indicates higher molecular weight Parkin-ubiquitin thioester species. (B) YFP-Parkin-C431S–expressing HEK293T cells treated as in (A) ± SBI-0206965 (10 μM) for 2 hours. (C) Functional analysis of Parkin-mediated mitophagy activity in HEK293T cells stably bearing mito-Keima reporter. The graph indicates the mitophagy-positive cell population quantified by flow cytometry. Data are shown as the means ± SEM of two independent experiments. *P < 0.05 when compared to the control cells at 2 or 4 hours, respectively, by two-way ANOVA. (D) Immunoblot analysis of endogenous Parkin in hESC-derived induced neurons treated with AO for indicated times or 60 min with 10 or 50 μM 6965 pretreatment. (E) Model for Parkin activation. CCCP treatment causes rapid increases in AMP and mtROS, which activate AMPK to phosphorylate and activate ULK1 within minutes. ULK1, in turn, phosphorylates Parkin maximally at Ser108 in the cytoplasm within 2 min of CCCP treatment. Meanwhile, PINK1 is slowly becoming stabilized on the mitochondrial outer membrane following CCCP, with detection of its phosphorylation of ubiquitin and Parkin Ser65 appearing around 30 min and becoming maximal at 60 min or later.

Fig 4: Mutation of S17 in BNIP3 modulates its LC3B interaction and mitophagy. (a) Pulldown of GFP-LC3 stably expressed in HEK-293 T cells with transiently expressed HA-BNIP3 (WT) and different HA-BNIP3 mutants (W18A, S17A, S17E) or empty vector (EV) control, in the presence or absence of 100 nM bafilomycin A1. Inputs to the pulldown are shown on the left and the result of the pulldown on the right. Fold changes in protein levels of BNIP3 dimer and BNIp3 monomer are shown relative to WT. (b) Pulldown of GFP-LC3 with HA-BNIP3, as described in (a), in the presence (lanes 6–9, 16–19) or absence (lanes 1–5, 10–15) of exogenous FLAG-ULK1. Fold change in protein levels of BNIP3 dimer and BNIp3 monomer are shown. Fold changes in protein levels of BNIP3 dimer and BNIp3 monomer are shown relative to WT. (c–f) Immunofluorescent staining for TOMM20 (green, mitochondria), LC3B (magenta, autophagosomes), HA-BNIP3 (red) and DAPI (blue) in U2OS cells transiently expressing HA-BNIP3 (c), HA-BNIP3W18A (d), HA-BNIP3S17A (e) or HA-BNIP3S17E (f). Cells expressing exogenous HA-BNIP3 are asterisked (*) and LC3B/TOMM20 overlap is detected as white puncta (green and magenta overlap). (g) Quantification using Image J of LC3B/TOMM20 overlap per cell for at least at least 15 cells per field for each of the different forms of BNIP3 compared to cells not expressing BNIP3. Data were statistically analyzed as described in “Materials and methods” section. All data are shown as the mean ± s.e.m. Values of p ≤ 0.05 are considered significant. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001.

Fig 5: Recruitment of Parkin and downstream mitophagy effectors to mitochondria is controlled by AMPK/ULK1 signaling following CCCP.(A) Immunoblots of Mito-tag purified mitochondria from HEK293T cells stably expressing WT or SA3 YFP-Parkin treated with 20 μM CCCP for indicated times. All cells also treated with 10 μM MG132 for 4 hours. (B) Representative immunocytochemistry of NDP52 (white) and Tom20 (red) in HeLa cells stably expressing WT or SA3 YFP-Parkin. Cells treated with DMSO or CCCP for 4 hours. Nuclei stained with 4′,6-diamidino-2-phenylindole (blue). NDP52 puncta number and puncta size were analyzed by ImageJ with 10 individual images for each condition. Data are shown as the means ± SEM. **P < 0.001; ****P < 0.0001 by two-way ANOVA. Scale bar, 20 μm. n.s., not significant. (C) Immunoblots of Mito-tag purified mitochondria from WT YFP-Parkin expressing cells treated with DMSO or 10 μM 6965 for 15 min before CCCP treatment for indicated times. Cells treated with MG132 as in (A). (D) Immunoblots of cell lysates from WT YFP-Parkin expressing cells treated with CCCP ± 15 min pretreated with 50 μM 991 or 10 μM 6965 as indicated. (E) Immunoblots of purified mitochondria from cells stably expressing WT or SA3 YFP-Parkin treated with CCCP ± 50 μM 991 for 1 hour. All cells also treated with 10 μM MG132 for 1 hour. (F) Immunoblots of supernatant (top) and purified mitochondria (bottom) from control or CRISPR-mediated AMPK KO cells treated with 20 μM CCCP for indicated times. Cells treated with MG132 as in (A).