Figure 1.
Activation, domains, structures and models of human Parkin.
A) Schematic representation of Parkin's 2D structure and its activation mode. Top: PINK1-dependent phosphorylation of Ser65 has been shown to activate Parkin. During activation, an Ub-loaded E2 enzyme binds Parkin and catalyzes the Ub transfer onto the active site, Cys431, in order to charge Parkin with the small modifier protein. Bottom: Shown are color-coded functional domains of human Parkin (residues 1–465): Ubiquitin-like (UBL, red), flexible linker (gray), RING0/Unique Parkin domain (R0/UPD, green), RING1 (R1, blue), in-between RING (IBR, purple), Repressor element of Parkin (REP, yellow), RING2 (R2, pink). The putative E2 binding site in RING1 is indicated by a black line. Gray lines indicate the position of the Zn2+ coordinating cysteine/histidine residues in the different RING domains. B) Superposition of Parkin's molecular structures. Table lists recently resolved X-ray structures that have been used to generate a model for human full-length Parkin with all-atom resolution. Key residues Ser65 and Cys431 are shown as sticks with carbon in gray, nitrogen in blue, oxygen in red, and sulfur in yellow. C) Molecular modeling of Parkin pSer65. The ribbon diagram for an all-atom molecular structure of Parkin is given, presenting an in silico model of a PINK1-phosphorylated, and thus activated conformation of Parkin. Color matches that of the domain key indicated, and as above. The phosphorylated Ser65 is shown along with the active site Cys431 as Van der Waals (VdW) spheres colored by domain. Zn2+ atoms are shown as blue spheres.
Figure 2.
phospho-Ser65 triggers cleft widening in the N-terminal region.
A) Center view of a newly defined pocket in Parkin with Ser65 positioned towards the middle. The cleft is formed between the flexible linker region [cleft wall 1: Arg97, Ser110, Val105, Val111, Leu112, Asp115 Ser116, Val117, and Gly118] and the UBL domain [cleft wall 2: Met1, Ile2, Val3, Phe4, Ser19, Leu61, Asp62, Gln63, Gln64, Ile66, and Val67]. Parkin is rendered in solvent accessible surface colored by atom type (nitrogen-blue, oxygen-red, carbon-cyan). Yellow double-sided arrow indicates cleft regions that were used for center-of-mass calculations: CoM1 (Ser110, Val111, Asp115) to CoM2 (Met1, Ile2, Val3, Phe4, Gln63). The initial cleft width is given for time equal zero. B) The ribbons diagram for Parkin is shown for comparison with same orientation. Relevant domain labels are given. C) Parkin pSer65 is shown after 20 ns of unbiased MDS. The yellow arrow indicates the increased cleft distance. D) Superposition of Parkin structures after 100 ns MDS. Shown are structures for Ser65 and pSer65 as well as for S65A, S65D and S65E variants. Arrows indicate the cleft distances of Ser65 and pSer65. E) Plot for CoM1 to CoM2 distance over time from MDS. Graph shows a relatively closed and stable cleft for unmodified Ser65 (black) of about 8 Å in distance, while pSer65 (green) shows a much more wider cleft from the start, ranging from 11–14 Å over time. Phospho-mimic mutants S65D (magenta) or S65E (blue) as well as the phospho-dead mutant S65A (red) show a strong increase in cleft size over time reaching distance observed with pSer65. F) Solvent-Accessible-Surface-Area (SASA-Å2) within the pocket enclosing Ser65 measured in Å2 units. While Ser65 maintain a relatively stable SASA, values for pSer65 strongly increase over time. An increase in SASA over time is also observed for the substitutions S65D and S65E as well as for S65A after an initial decrease. G) Shown are numbers of H2O molecules within the cavity surrounding Ser65 during MDS. While Ser65 constantly maintains twelve H2O molecules, pSer65, S65D, S65E, and S65A show a greater solvation of the cavity, consistent with an increased cleft distance and improved SASA.
Figure 3.
Cell-based high content imaging of Parkin Ser65 mutations.
HeLa cells transiently expressing GFP-Parkin wild type, S65A, S65D, or S65E mutations were left untreated (0 h) or treated with the uncoupler CCCP for 2 h or 4 h. Cells expressing GFP only or the catalytically inactive GFP-Parkin C431S mutations served as controls. A) Images have been acquired using automated microscopy and show mitochondria (TOM20) in red, GFP-Parkin in green and nuclei (Hoechst) in blue. Quantification of Parkin re-localization is assessed by measuring maximal intensity in a cytoplasmic ring around the nucleus divided by the mean intensity of the nuclear GFP signal. Cytoplasmic ring and inner nuclear regions are schematically shown in the merge image. The GFP ratio is given for each GFP-positive cell in white as an example and reflects Parkin translocation to mitochondria. Untransfected cells or cells below threshold expression of Parkin are marked with a white asterisk and have been excluded from the analysis. B) The average GFP ratios of Parkin wild type and Ser65 mutations per well are shown as a heat map. C–E) Bar graphs give the percentages of cells with a defined mitochondrial translocation of Parkin at 0 h (C), 2 h (D), and 4 h (E) of CCCP treatment. A GFP threshold of 1.8 was chosen that corresponds to the average ratio of Parkin wild type translocation after 2 h CCCP treatment (n>4 plates with at least 4 wells per condition, one-way ANOVA, Tukey's post-hoc, p<0.0001, F = 147.3, ns – not significant, *** p<0.0005). F) Given are representative merge images at higher magnification that show co-localization of GFP-Parkin (green) and mitochondria (anti-TOM20, red). Nuclei (Hoechst) are shown in blue. Scale bars correspond to 10 µM. For images of the individual channels at all time points, see Figure S4.
Figure 4.
MdMD excursions along the pathway of the UBL domain across Parkin.
Representative pathways via MdMD excursions between LC-MOD/MC generated conformers for the movement of the N-terminal region (residues 1–140) in MDS state 1 (panel A) to the active site region in MDS state 5 (panel E). Panels A–E and A′–E′ represent the five key points from the guideposts that MdMD was able to drive the structures toward. RMSD between the structural model and the guideposts were within 3 Å in each case. Color-coded ribbon structures are given. A) The initial structure for Parkin is mostly unperturbed after 3 ns of MdMD sampling with global variable based on LCMOD sampling between generated Parkin conformers. B) Opening of the N-terminal region is shown after>10 ns of MdMD sampling. C) Midpoint for the UBL domain movement towards the active site region following>15 ns of MdMD sampling. The linker helix region is more dynamic, exposing the E2 binding site in RING1. D) UBL domain adjustment as it approaches the C-terminal region after 30 ns of MdMD sampling. E) The UBL domain is in final position and occupies region around the active site (Cys431) after 35 ns of MdMD sampling. A′–E′) Superposition of the four replicates from MdMD that match the relative time point/stage from panels A–E. Ensemble of structures from replicates gives a common relative pathway between guideposts generated conformers.
Figure 5.
Phosphorylation of Ser65 releases the safety belts of Parkin.
A) Zoom into safety belt 1: The UBL blocks RING1 and IBR domains. Key cysteine residues of the E2 binding site in RING1 are indicated. The E2 binding site was defined as follows: Ile236, Thr237, Cys238, Ile239, Thr240, Cys241, Thr242, Asp243, Val244, Arg245, Ile259, Cys260, Leu261, Asp262, Cys263, Phe264, His265, Leu266, and Tyr267 B) The distance between UBL domain (Leu26) and RING1 (Cys238) significantly increased over time MDS. C) Similarly, the distance between UBL (Leu26) and IBR (Phe364) domains significantly increased over time MDS. D) Zoom into safety belt 2: The REP region blocks the E2 binding site in RING1 (as defined in A). E) Dynamic change in REP-RING1 interaction during Parkin opening motion. Graph shows the release of the REP region from the E2-binding site in RING1 as measured by RMSD. The RING1 is released from the REP region by MdMD time of 20 ns, exposing the E2 binding site. F) Loosened interaction between the center Tyr391 in REP region and Cys238 in RING1. The distance increases from 10 to 20 Å. During longer simulations, the distance eventually collapses as the UBL domain moves away and E2 binding has transiently occurred. Across many replicates, we find that the availability of adequate space for an E2 enzyme to approach the binding site in RING1 begins somewhere between 5–22 ns. G) Zoom into safety belt 3: Cys431 is buried by RING0. H) Release of the active site (Cys431) from RING0 (Arg163 C-alpha atom) as measured by RMSD for center-of-mass. RMSD increases moderately over time indicative of a less compacted area. I) SASA for Cys431 entire residue. During MDS, more water is available to Cys431, indicating its enhanced exposure.
Figure 6.
Protein-protein docking for Parkin and a charged E2∼Ub complex.
A) Ten conformations spanning closed Parkin to fully opened Parkin (see Figure 4) were sampled for protein-protein interactions. The E2-Ub complex was docked at the midpoint UBL position (state 2/3) when the REP region liberated the binding site in RING1 (Figure 4B/C). This conformation showed fewer steric clashes and lowest energy profile. The docking in the same position is shown and rotated 180° to reveal the other side. Residues of the Ubch5a-Ub complex are indicated by color (dark green and brown, respectively). B–D) Docking at the RING1 interface is critical for E2-Ub progression towards the active site of Parkin. E2 binding at RING1 limits the UBL-linker mobility preventing the drift back to the original, auto-inhibited state. B) Same orientation as in A (left side) predicted as an optimal docking conformation after 0 ns of MDS. The distance between Gly76 of Ub and Parkin's active site (Cys431) is indicated. C) Following unbiased MD (>200 ns), the E2-Ub complex moves towards the Parkin's active center. The decreased distance is shown after 50 ns. D) Overall re-orientation of the UBL domain and the E2-Ub complex is shown after 200 ns.
Figure 7.
Activation and enzymatic function of Parkin.
A) HeLa cells were transfected with FLAG-Parkin C431S. Cells were treated with CCCP for 0, 1, 2, 4, or 16 h and harvested. Western blots were prepared to monitor the formation of a Parkin C431S-Ub oxyester, which, in contrast to unmodifed Parkin (closed arrowhead) appears as a band shift (open arrowhead) and is sensitive to NaOH treatment. The phospho-mimic mutations S65D or S65E showed some levels of Parkin C431S-Ub even in the absence of CCCP, consistent with a slight activation under steady-state conditions. B) HEK293E cells were transfected with FLAG-Parkin wild type and left either untreated or were treated with 10 µM CCCP. FLAG immunoprecipitations were performed and ubiquitinations reactions were carried out on the beads. All ubiquitination reactions contained E1, ATP, and N-terminally biotinylated Ub. Either no E2 enzyme, or UBE2D2, UBE2L3 or UBE2N plus its co-factor Uev1a were added. In order to analyze the effect of Parkin phosphorylation, some FLAG immunoprecipitates were pretreated with phosphatase before Ub reaction was carried out. UBE2D2, UBE2L3 and UBE2N can serve as co-factors for Parkin in vitro. CCCP treatment is not required, but enhances the ubiquitination reactions. Phosphatase treatment reduces the ubiquitinations to the extent observed without CCCP treatment. UBE2N shows no detectable activity towards Parkin auto-ubiquitination. A closed arrowhead indicates the position of unmodified FLAG-Parkin, an open arrowhead labels Ub modified species. An arrow labels the molecular weight of Parkin on the Streptavidin-HRP blot.