Characterization of pUbl and RING2 interactions on RING0

A. SEC assay showing depletion of RING2 (383-465) from phospho-Parkin stabilize pUbl-linker (1-140) binding with Parkin (141-382) after treatment with 3C protease. Fractions that were pooled for subsequent proteolysis are highlighted in the box.

B. SEC assay showing depletion of pUbl-linker (1-140) from phospho-Parkin stabilize RING2 (383-465) binding with Parkin (R0RB, 141-382) after treatment with TEV protease. Fractions that were pooled for subsequent proteolysis are highlighted in the box.

C. Crystal structure of pUbl-linker (1-140) depleted Parkin (141-465) in complex with pUb (brown). Different domains of Parkin are colored, as shown in the left panel. Catalytic Cys431 is highlighted. pUbl-linker (1-140) depleted Parkin (141-465) complex structure (colored as in the left panel) is superimposed with R0RBR structure (PDB 4I1H, grey) in the right panel.

Untethering of the linker between IBR-RING2 allows Parkin and phospho-ubl interaction in trans

A. Binding assay between phospho-Parkin (Lys211Asn) and ΔUbl-Parkin. A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown in the lower panel. A schematic representation is used to explain SEC data. In the lower panel, the Isothermal Titration Calorimetry assay between phospho-Parkin (Lys211Asn) and ΔUbl-Parkin is shown. N.D. stands for not determined.

B. Binding assay between phospho-Parkin (Lys211Asn) and ΔUbl-Parkin (untethered RING2 by TEV treatment). A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown in the lower panel. A schematic representation is used to explain SEC data. In the lower panel, the Isothermal Titration Calorimetry assay between phospho-Parkin (Lys211Asn) and ΔUbl-Parkin (untethered RING2 by TEV treatment) is shown. The dissociation constant (Kd) is shown.

C. SEC assay to test binding between R0RBR (untethered RING2 by TEV treatment) and phospho-Parkin (Lys211Asn), and displacement of RING2 (383-465) from R0RBR, the left panel. The peak1 (black) containing R0RB (141-382) and phospho-Parkin (Lys211Asn) complex was incubated with pUb, followed by HRV 3C, to purify ternary trans-complex of phospho-Parkin (1-140 + 141-382 + pUb) on SEC, the right panel. The concentrated fractions from the shoulder (highlighted with a dashed line) of the peak in the right panel were loaded on SDS PAGE to confirm complex formation.

D. Crystal structure of the trans-complex of phospho-Parkin with pUb (brown) showing phospho-Ubl domain (wheat) bound with RING0 (cyan) domain of Parkin (cyan).

Parkin dimerization and trans-activation of native Parkin are mediated by phosphorylation of the Ubl domain of Parkin

A. SEC assay between phospho-Parkin and untethered ΔUbl-Parkin. A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown in the lower panel. TEV protein contamination is indicated (*).

B. SEC assay between phospho-Parkin and untethered WT-Parkin (upper panel) or untethered Lys211N-Parkin (lower panel). TEV protein contamination is indicated (*).

C. SEC-MALS assay to confirm the complex formation between untethered WT-Parkin and native phospho-Parkin.

D. Ubiquitination assays to check the WT-Parkin activation (right panel) with increasing concentrations of phospho-Parkin (T270R, C431A). A non-specific, ATP-independent band is indicated (*). The lower panel shows a Coomassie-stained loading control.

Analysis of Parkin mutant recruitment to mitochondria in HeLa cells

A. Immunofluorescence of HeLa cells co-transfected with either wild-type (WT) mCherry-Parkin or mCherry-Parkin S65A and GFP-Parkin C431F or, B. GFP-Parkin H302A-C431F and, C. GFP-Parkin K211N-C431F. Cells were treated for 1 h with 10 μM CCCP, DMSO was used as control. Mitochondria were labelled with anti-TOMM20 antibody (blue). Scale bars, 10 μm. D. Quantification of GFP-Parkin (WT and mutants) on mitochondria. The colocalization of GFP-Parkin (WT and mutants) with TOMM20 (mitochondria) was assessed using Pearson’s correlation coefficient. Errors are represented as S.D. and statistical differences in Pearson’s correlation coefficient were assessed by one-way ANOVA and Tukey’s multiple comparisons post-test. Statistical significance is indicated as follows: *, p < 0.05; ****, p < 0.0001.

ACT plays a key role in enzyme kinetics

A. Size-exclusion chromatography (SEC) assay to test the binding of E2~Ub with phospho-Parkin (left panel) or phospho-Parkin (ΔACT) (right panel). Assays were done using Parkin in complex with pUb. A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown in the lower panel.

B. Size-exclusion chromatography (SEC) assay to check displacement of the RING2 domain after phosphorylation of Parkin (ΔACT). The upper panel shows a schematic representation of the Parkin (ΔACT) construct used for the RING2 displacement assay. Conformational changes in Parkin, as suggested by the SEC experiment, are shown schematically.

C. Ubiquitination assay to check the effect of ACT deletion on Parkin activity, phospho-Parkin (ΔACT) activity shows higher activity in the lanes where reactions were incubated at longer time points as indicated. A non-specific, ATP-independent band is indicated (*). The middle panel shows a Coomassie-stained loading control. In the lower panel, the bar graph shows the integrated intensities of ubiquitin levels from three independent experiments (mean ± s.e.m.). Statistical significance was determined using pair-wise student’s t-test (**P< 0.01, ***P < 0.001, ns-nonsignificant).

ACT is more efficient in cis

A. Crystal structure of ternary trans-complex of phospho-Parkin with pUb (1-140 + 141-382 + pUb), left panel. pUbl (wheat) of 1-140 fragment of Parkin and RING0 (cyan) of 141-382 fragment of Parkin are shown. The right panel shows superimposed structures of ternary trans-complex of phospho-Parkin with pUb, colored as the left panel, and the phospho-Parkin complex with pUb (PDB 6GLC) is shown in grey.

B. SEC assay to check the binding between ΔUbl-Parkin (untethered RING2 by TEV) and phospho-Ubl (1-76). A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown in the lower panel.

C. Crystal structure of trans-complex of phospho-Parkin with cis ACT (1-76+ 77-382 + pUb) shows ACT (cyan) present in the pocket on RING0 (Cyan) and pUbl (wheat) in the vicinity.

D. Comparison of R0RBR and ΔUbl-Parkin activation using the increasing concentrations of pUbl (1-76). A non-specific, ATP-independent band is indicated (*). The middle panel shows a Coomassie-stained loading control. The lower panel shows Miro1 ubiquitination for the respective proteins shown in the upper lane. Coomassie-stained gel showing Miro1 is used as the loading control of substrate ubiquitination assay.

E. Ubiquitination assay of ΔUbl-Parkin with increasing concentrations of pUbl (1-76), pUbl-linker (1-140), pUbl-linker-ΔACT (1-140, Δ101-109). A non-specific, ATP-independent band is indicated (*). The middle panel shows a Coomassie-stained loading control. The lower panel shows Miro1 ubiquitination for the respective proteins shown in the upper lane. Coomassie-stained gel showing Miro1 is used as the loading control of substrate ubiquitination assay.

F. Comparison of R0RBR activation using the increasing concentrations of pUbl (1-76), pUbl-linker (1-140), pUbl-linker-ΔACT (1-140, Δ101-109). ΔUbl-Parkin activation reactions using increasing concentrations of pUbl (1-76) are used as control. A non-specific, ATP-independent band is indicated (*). The middle panel shows a Coomassie-stained loading control. The lower panel shows Miro1 ubiquitination for the respective proteins shown in the upper lane. Coomassie-stained gel showing Miro1 is used as the loading control of substrate ubiquitination assay.

Linker (408-415) of Parkin binds with donor ubiquitin (Ubdon) of E2-Ub.

A. The asymmetric unit of the crystal structure of pUbl-linker (1-140) depleted Parkin (141-465, Arg163Asp, Lys211Asn) in complex with pUb. Domains (different colors) of Parkin molecule-1 and pUb (brown) are shown. Parkin molecule-2 (grey) and pUb (orange) are shown. The interface of two Parkin molecules is highlighted (dashed line).

B. The 2Fo-Fc map (grey) of the linker region between REP and RING2. 2Fo-Fc map is contoured at 1.5 σ. Water molecules are represented as w.

C. Crystal structure showing interactions between the linker (408-415) and ubiquitin. Different regions are colored as in panel A. Hydrogen bonds are indicated as dashed lines.

D. Sequence alignment of Parkin from various species highlighting conservation in the linker (408-415) region. Residue numbers shown on top of sequence alignment are according to human Parkin.

E. Ubiquitination assay of Parkin mutants in the linker region. The middle panel shows a Coomassie-stained loading control. The lower panel shows Miro1 ubiquitination for the respective proteins shown in the upper lane. Coomassie-stained gel showing Miro1 is used as the loading control of substrate ubiquitination assay.

F. Size-exclusion chromatography (SEC) assay to compare the binding of E2~Ub with phospho-Parkin (upper panel) or phospho-Parkin Ile411Ala (lower panel). Assays were done using Parkin in complex with pUb. A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown.

Model showing different modes of Parkin activation.

The cis-activation model uses binding of pUbl in the same molecule, thus resulting in the displacement of RING2 (1). The trans-activation model uses the binding of pUbl of fully-activated parkin (phospho-parkin complex with pUb) with partially-activated parkin (WT-parkin and pUb complex), thus resulting in the displacement of RING2 in trans (2). Recruitment and activation of parkin isoforms lacking Ubl (Isoform 10) or RING2 domain (Isoform 5), thus complementing each other using trans-activation model (3). Catalytic cysteine on RING2 is highlighted.

Data collection and Refinement statistics

Summary of Parkin conformations/model.

A. Conformation of Parkin observed in various crystal structures solved so far B. Proposed conformation of phospho-Parkin 24, 29 C. Schematic representation of Parkin domains. HRV 3C and TEV sites incorporated in the Parkin construct are marked with black and green arrows, respectively

pUbl and RING2 have a competitive mode of binding on RING0.

A. Superimposition of the structure of WT-Parkin (PDB 5C1Z) and phospho-Parkin (PDB 6GLC) structure. RING2 (blue), pUbl (brown), RING0 (red), and ACT (black) are shown. For clarity, other domains of Parkin are not included. B. Ubiquitination assay on Parkin construct with TEV and HRV 3C sites. A non-specific, ATP-independent band is indicated (*). The lower panel shows a Coomassie-stained loading control. C. Size-exclusion chromatography (SEC) assay showing the binding/displacement of Ubl-linker (1-140) under native or phosphorylated conditions. A colored key for each trace is provided. Coomassie-stained gels of indicated peaks are shown in the lower panel. A schematic representation is used to explain SEC data. D. Size-exclusion chromatography (SEC) assay showing binding/displacement of RING2 (383-465) under native or phosphorylated conditions. Coomassie-stained gels of indicated peaks are shown in the lower panel. TEV as contamination is indicated (*).

Density map of pUbl-linker (1-140) depleted Parkin (141-465) complex with pUb structure.

The 2Fo-Fc map (blue) of Parkin molecules (shown in different colors) in the crystal structure of the pUbl-linker (1-140) depleted Parkin (141-465) complex with pUb (grey). The 2Fo-Fc map is contoured at 1.5 σ.

K211N mutation affects RING2 displacement but not pUbl.

A. Size-exclusion chromatography (SEC) assay to test the displacement of RING2 (left panel) or pUbl-linker (right panel) after phosphorylation of Parkin (Lys211Asn). B. Ubiquitination assay to test the activity of Parkin Lys211Asn in the presence of pUb or using phospho-Parkin Lys211Asn. The middle panel shows a Coomassie-stained loading control. A non-specific, ATP-independent band is indicated (*). The lower panel shows Miro1 ubiquitination for the respective proteins shown in the upper lane. Coomassie-stained gel showing Miro1 is used as the loading control of substrate ubiquitination assay. C. Crystal structure of pUbl-linker (1-140) depleted Parkin (141-465, Lys163Asp, Lys211Asn) in complex with pUb (brown), different domains of Parkin are colored as panel E, superimposed apo R0RBR structure (PDB 4I1H) is shown in grey.

Parkin treatment with TEV does not affect native interactions.

A. Schematic representation of the R0RBR construct. TEV sites incorporated in the Parkin construct are marked green arrows, respectively B. Purification of untethered R0RBR Parkin over Hiload 16/600 Superdex 75pg column. Fractions from the highlighted region of the peak were loaded on SDS-PAGE (lower panel). C. Crystal structure of untethered R0RBR Parkin. The 2Fo-Fc map is shown in grey. Parkin domains are shown in different colors.

Electron density map of the ternary trans-complex of Parkin.

The 2Fo-Fc map (grey) of the ternary trans-complex of phospho-Parkin (1-140 (wheat) + 141-382 (cyan) + pUb (brown)). The 2Fo-Fc map is contoured at 1.5 σ.

Phosphorylation of Parkin leads to the association of Parkin molecules in trans.

A, B, C, Schematic representation of SEC assay from Fig 3 A and B, respectively D. Phos-Tag analysis shows the effect of pUbl (1-76) or pUb (1-76) on the phosphorylation of Parkin by PINK1.

Parkin localization on mitochondria.

A. HeLa cells were transfected with plasmid for wild-type (WT) GFP-Parkin, GFP-Parkin C431F, GFP-Parkin H302A/C431F and GFP-Parkin K211N/C431F. Cells were treated for 1 h with 10 μM CCCP or DMSO was used as control. Mitochondria were labeled with anti-TOMM20 antibody (blue). Scale bars, 10μm. B. Quantification of GFP-Parkin (WT and mutants) on mitochondria***, p < 0.001. C. Quantification of mCherry-Parkin (WT and S65A) on mitochondria in Figure 4, ns=not significant.

Role of ACT in Parkin activation.

A. The 2Fo-Fc map (grey) of the trans-complex of phospho-Parkin in the crystal structure of the ternary trans-complex of phospho-Parkin (1-140 (wheat) + 141-382 (cyan)). The 2Fo-Fc map is contoured at 1.0 σ. B. The 2Fo-Fc map (grey) of the ACT region in the ternary trans-complex structure of phospho-Parkin with cis ACT (1-76 (wheat) + 77-382 (cyan)). The 2Fo-Fc map is contoured at 1.0 σ. C. Comparison of ubiquitination activity of ΔUbl-Parkin and R0RBR. A non-specific, ATP-independent band is indicated (*). The middle panel shows a Coomassie-stained loading control. The lower panel shows Miro1 ubiquitination for the respective proteins shown in the upper lane. Coomassie-stained gel showing Miro1 is used as the loading control of substrate ubiquitination assay. D. The bar graph shows the integrated intensities of Miro-1 ubiquitination levels from three independent experiments (mean ± s.e.m.). pUbl was added in 2-fold (+) or 4-fold (++) molar excess. Statistical significance was determined using pair-wise student’s t-test (P< 0.001). E. The bar graph shows the integrated intensities of Miro-1 ubiquitination levels from three independent experiments (mean ± s.e.m.). Various pUbl constructs were added in 2-fold (+) or 4-fold (++) molar excess. Statistical significance was determined using pair-wise student’s t-test(P < 0.001)

The linker connecting REP and RING2 shows conformational flexibility.

A. Superimposition of Parkin structures showing the flexible nature of the linker between REP-RING2. Different Parkin structures are colored according to their PDB code. B. Superimposition of pUbl-linker (1-140) depleted Parkin (141-465, Arg163Asp, Lys211Asn) and pUb complex with apo R0RBR (144-465, PDB 4I1H) showing conformational changes in the linker region (408-415, purple) of Parkin. Apo R0RBR (PDB 4I1H) structure is shown in grey. C. Conformation of REP-linker predicted from AlfaFold 34. A Full-length Parkin sequence was used to model the Parkin structure using AlfaFold. D. Sequence alignment of the linker connecting IBR-RING2 among various RBR family E3-ligases. The conserved hydrophobic patch among Parkin, HOIP, HHARI, and HOIL is highlighted with a dashed box. The sequence numbering is according to the sequence of Parkin.

The linker connecting REP and RING2 domain binds with ubiquitin (Ubdon) of E2-Ub.

A. Crystal structures of various RBR family E3-ligases showing linker between IBR-RING2 interact with ubiquitin (Ubdon) of E2-Ub. Only RING1 (cyan), IBR (magenta), linker (purple), and RING2 (blue) are shown interacting with E2 (orange) and ubiquitin (Ubdon) (brown) of E2~Ub B. Ub-VS probe reactivity with the catalytic Cys of RING2 of phospho-Parkin, or phospho-Parkin I411A C. Ubiquitination assay to compare Parkin activity using WT/mutant ubiquitin. A non-specific, ATP-independent band is indicated (). The lower panel shows a Coomassie-stained loading control. The bar graph shows the integrated intensities of ubiquitination levels from three independent experiments (mean ± s.e.m.). Statistical significance was determined using pair-wise student’s t-test (**P < 0.001)

Schematic representation of domain organization in various isoforms 41,42of Parkin.