WO2025201510A1 - Use of a fkbp51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities - Google Patents

Use of a fkbp51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities

Info

Publication number
WO2025201510A1
WO2025201510A1 PCT/CN2025/085699 CN2025085699W WO2025201510A1 WO 2025201510 A1 WO2025201510 A1 WO 2025201510A1 CN 2025085699 W CN2025085699 W CN 2025085699W WO 2025201510 A1 WO2025201510 A1 WO 2025201510A1
Authority
WO
WIPO (PCT)
Prior art keywords
pink1
fkbp51
hsp90α
cdc37
hsp90
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/CN2025/085699
Other languages
French (fr)
Inventor
Li-Zhi Mi
Xiaohong Qin
Junrong XUE
Yanfeng Zhang
Xinrong YU
Jia GUO
Jianing DUAN
Weida ZHANG
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University
Original Assignee
Tianjin University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Publication of WO2025201510A1 publication Critical patent/WO2025201510A1/en
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61PSPECIFIC THERAPEUTIC ACTIVITY OF CHEMICAL COMPOUNDS OR MEDICINAL PREPARATIONS
    • A61P9/00Drugs for disorders of the cardiovascular system
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61PSPECIFIC THERAPEUTIC ACTIVITY OF CHEMICAL COMPOUNDS OR MEDICINAL PREPARATIONS
    • A61P25/00Drugs for disorders of the nervous system
    • A61P25/14Drugs for disorders of the nervous system for treating abnormal movements, e.g. chorea, dyskinesia
    • A61P25/16Anti-Parkinson drugs

Definitions

  • HSP90s undergo chaperone cycles to promote the folding, translocation, and degradation of its client proteins 4, 5 .
  • Co-chaperones are engaged into this cycle at different stages to act as a scaffold for loading clients (i.e. CDC37) 6-8 , modulating the conformation and ATPase activity of HSP90 (i.e. CDC37, Aha1) 6, 9, 10 , promoting the release of matured clients (i.e. P23, HOP) 11, 12 , regulating the degradation of unfolded clients (i.e. FKBP51, FKBP52, CHIP) 13-16 , or controlling the activities of clients via other enzymatic functions (i.e. PP5) 17, 18 .
  • CDC37 is essential for the folding and stabilization of many kinase clients. Serving as a scaffold, CDC37 loads clients from HSP70 onto HSP90, inhibits the ATPase activity of HSP90, and stabilizes unfolded clients in the HSP90 chaperone complexes 6-10, 19, 20 .
  • Immunophilins such as FKBP51 and FKBP52, represent another class of co-chaperones for HSP90 13-15 .
  • FKBP51 and FKBP52 are two TPR (TetratricoPeptide Repeat) domain containing peptidyl-prolyl cis/trans isomerase (PPIase) with mutually antagonizing functions against each other 13-15, 21 . They are involved in the inflammation, innate immunity, stress response, and neuronal protection, and thus are actively explored as potential druggable targets for treating stress-related diseases 14, 22, 23 .
  • Recent structural studies revealed how they are involved in the regulation of the glucocorticoid receptor activation and maturation 13, 24 . But it remains largely elusive how they are involved in the folding, activation, and maturation of kinase clients.
  • PINK1 a nucleus-encoded mitochondrial kinase, is critical for the quality control and autophagy of mitochondria 25-28 .
  • expressed PINK1 is translocated into mitochondria through the TOM complex 29 , cleaved by PARL 30, 31 and MPP 32 , retro-translocated back to the outer membrane of mitochondria, and eventually degraded by proteasome 33 . As such, it is kept at the lower protein and activity levels.
  • PINK1 Upon the stimulation by various signals of mitochondrial damage, the full-length PINK1 is arrested and accumulated on the outer membrane of mitochondria, activated through dimerization and association with TOM complex 29, 34 , recruits PARKIN (an E3 ligase) to damaged mitochondria via phosphorylation of PARKIN and ubiquitylated proteins on mitochondria 35-37 . recruited PARKINs are fully activated and trigger the clearance of damaged mitochondria via proteasome 36-38 . In these series of cascades, precise regulation of PINK1’s quantity and activity on mitochondria is the key for coordinating these events in mitophagy. However, the structures and molecular details for this regulation is largely undefined.
  • PINK1 Loss-of-function mutations of PINK1 are linked to the early onset of autosomal recessive Parkinson’s disease (PD) 39-41 . Therefore, PINK1 could be a potential target for the treatment of PD 42, 43 . Indeed, early studies focused on the development of ATP analogs to up-regulate the kinase activities of PINK1 mutants 43 . But these efforts were still under debate in proofing the concept in animal models 44, 45 . Development of alternative strategies in the treatment of PINK1-deficient PD is challenging but worth to be explored.
  • the present invention determined the Cryo-EM structures and molecular recognition for the HSP90 ⁇ /CDC37/FKBP51/PINK1 complexes, and these data disclosed a general mechanism for client recognition by HSP90 ⁇ /CDC37/FKBP51 complex and the molecular basis for coupling PINK1 stability, folding, and degradation in its activity regulation. Moreover, the present invention revealed an alternative avenue and a potential target for treating PINK1-deficient Parkinson’s disease. For example, the present invention identified that specific inhibition of FKBP51 could promote PINK1-mediated PARKIN recruitment in response to mitochondrial depolarization.
  • the present invention provided use of a FKBP51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities.
  • the disease is Parkinson disease.
  • the Parkinson disease is PINK1-dificient Parkinson disease.
  • the disease is cardiomyositis, coronary heart disease or liver fibrosis.
  • the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof.
  • the disease is Parkinson disease.
  • the Parkinson disease is PINK1-dificient Parkinson disease.
  • the disease is cardiomyositis, coronary heart disease or liver fibrosis.
  • the disease is Parkinson disease.
  • the Parkinson disease is PINK1-dificient Parkinson disease.
  • the disease is cardiomyositis, coronary heart disease or liver fibrosis.
  • the HSP90 inhibitor is geldanamycin or a pharmaceutically acceptable salt thereof.
  • the disease is Parkinson disease.
  • the Parkinson disease is PINK1-dificient Parkinson disease.
  • the disease is cardiomyositis, coronary heart disease or liver fibrosis.
  • the present invention provided a pharmaceutical composition
  • a pharmaceutical composition comprising a FKBP51 inhibitor or a HSP90 inhibitor and one or more pharmaceutically acceptable excipients, wherein the composition is used for treating or preventing diseases caused by mitochondrial abnormalities, preferably Parkinson disease.
  • the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof
  • the HSP90 inhibitor is geldanamycin or a pharmaceutically acceptable salt thereof.
  • the present application may be embodied in any other forms without departing from the spirit or scope thereof.
  • the present application encompasses any and all combinations of the above aspects and embodiments. It is to be understood that any embodiment may be combined with any other embodiment (s) to describe an additional embodiment. It is also to be understood that an individual element from any embodiment may be combined with any and all other elements from any other embodiment (s) to describe an additional embodiment.
  • Figure 1 shows biochemical characterization and Cryo-EM structures of hPINK1/HSP90 ⁇ /CDC37 in complex with or without FKBP51, wherein:
  • b Size-exclusion chromatography profile of the Strep-affinity purified complex. Molecular mass of the protein markers is shown on the top of the profile and marked with arrows. Each elution peak is numbered and marked by red dots at the bottom of the profile (upper) . Each peak fraction is subjected to electrophoresis on a 12%SDS-PAGE gel stained with Coomassie blue (lower left) and analyzed by Western blotting with the anti-FLAG antibody (lower right) . hPINK1 is indicated by red arrow.
  • HSP90 ⁇ subunit D The active site of HSP90 ⁇ subunit D.
  • the ATP molecule bound at the NTD of HSP90 ⁇ and the highly conserved residues V2, Y4, and W7 from CDC37 (malate) are shown as sticks covered with Cryo-EM map.
  • F134 of the subunit D forms a stronger ⁇ -cation interaction with R400;
  • R46 forms a salt-bridge contact with E47 to facilitate the deprotonation of E47 for ATP hydrolysis.
  • HSP90 ⁇ subunit C is presented in light yellow surface;
  • HSP90 ⁇ subunit D is in light blue cartoon.
  • HSP90 ⁇ subunit C The active site of HSP90 ⁇ subunit C.
  • the ATP molecule bound at the NTD of HSP90 ⁇ is shown as sticks covered with Cryo-EM map.
  • HSP90 ⁇ subunit D the interaction between F134 and R400 is impaired as the benzene ring of F134 moved away.
  • the amino group of R46 is away from the carboxyl group of E47, which is not favorable for ATP hydrolysis.
  • HSP90 ⁇ subunit C is in cartoon; HSP90 ⁇ subunit D is in surface representation.
  • TcPINK1 Superposition of the C-lobe from the unfolded hPINK1 in hPINK1/HSP90/CDC37/FKBP51 complex and that from TcPINK1 (PDB: 8ATQ) showing the conformational changes of the kinase.
  • TcPINK1 is shown as gray cartoon with ⁇ 5 and ⁇ K colored in golden.
  • hPINK1 is shown in surface representation colored by sequence conservation as in Fig. 3b.
  • HPNI motif on CDC37 (blue) mimics the configuration of the folded, functional PINK1 ⁇ i- ⁇ 4 loop to stabilize the C-lobe of the unfolded kinase.
  • the interface of PINK1 for CDC37 binding is highly conserved.
  • TcPINK1 is colored in gray with ⁇ K, ⁇ L, ⁇ M colored in golden.
  • hPINK1 is colored in magenta.
  • the activation loop (G386-C412) is invisible in electron density map.
  • the position of the activation loop of PINK1 is in the proximity to the ligand binding site of the FK1 domain.
  • Figure 4 shows that inhibition of the ATPase activity of HSP90 ⁇ by geldanamycin promotes the degradation of PINK1 under stress, wherein:
  • SAFit2 has little time-dependent effect on the phosphorylation of poly-ubiquitin in transiently transfected HEK293T cells.
  • the quantity of phosphorylated poly-ubiquitin is analyzed by Western blotting with an anti-pUb antibody (upper) . The quantitative results were shown in the lower panel. Relative phosphorylation was normalized against the average of total phosphorylated poly-ubiquitin densities.
  • SAFit2 could promote PINK1-mediated PARKIN recruitment.
  • Figure 6 shows model for hPINK1 folding and functional regulation by HSP90 ⁇ /CDC37/FKBP51 chaperone system under resting and stressed states, wherein: Under normal physiological conditions, the majority of PINK1 undergoes degradation via ubiquitin-proteasome pathway. Meanwhile, other unfolded PINK1s associated with CDC37 are stabilized by HSP90 and then transported into mitochondria, where they could be further stabilized by TRAP1 in the mitochondrial matrix. These two pathways in together prevent uncontrolled PINK1/PARKIN-mediated mitochondrial autophagy for the health of cells (left) . Under stressed conditions, the syntheses of HSP90 ⁇ and FKBP51 are stimulated, whereas they could form a complex with CDC37-associated PINK1.
  • PINK1 The translocation of PINK1 into damaged mitochondria is stalled so that increasing amount of PINK1 are accumulated on the outer membrane of mitochondria. Subsequently, via not fully-defined mechanisms, PINK1s are activated to initiate PARKIN recruitment for mitophagy. Consequently, the HSP90 inhibitor GA could suppress the folding and maturation of PINK1 by inhibiting the ATPase activity of HSP90, resulting in PINK1 degradation. During this process, FKBP51 negatively regulates PINK1 functions to prevent excessive autophagy of mitochondria. Thus, the specific inhibitor for FKBP51, the SAFit2, could enhance the maturation of PINK1 on mitochondria and thereby promotes the clearance of damaged mitochondria (right) .
  • Figure 7 shows biochemical characterization and Cryo-EM structures of the hPINK1 (130-581) /HSP90 ⁇ /CDC37 in complex with or without FKBP51, wherein:
  • Figure 10 shows multiple sequence alignment of FKBPs (a) and the kinase clients of HSP90/CDC37 (b) .
  • the conserved residues are highlighted with solid shades. Every 10th residue is marked with black dots on top of the sequences.
  • the ⁇ helices and ⁇ strands of FKBP51 (a) or hPINK1 (b) are shown with rods and arrows, respectively.
  • the secondary structures of FK1, FK2, TPR domains on FKBP51 are colored by light gray, white, dark gray, respectively. Residues located at each interface of interactions are marked at the bottom of the sequences as shown. EOPD mutations were marked with red stars on top of the sequences (b) .
  • Figure 11 shows multiple sequence alignment of HSP90s.
  • the conserved residues are highlighted with solid shades. Every 10th residue is marked with black dots on top of the sequences.
  • the ⁇ helices and ⁇ strands of HSP90 ⁇ are shown with rods and arrows, respectively.
  • the secondary structures of NTD, MD, CTD domain are colored by light gray, white, dark gray, respectively. Residues located at the interface of interactions with PINK1, FKBP51, CDC37 are marked at the bottom of the sequences.
  • Figure 12 shows detailed interactions in the recognition of hPINK1 by the HSP90 ⁇ /CDC37/FKBP51 complex, wherein:
  • hPINK1 magenta and red
  • HSP90 ⁇ D cyan
  • the H310-Y321 region of hPINK1 is shown as sticks and covered with cryo-EM map.
  • P374 on the ⁇ 7- ⁇ 8 loop of hPINK1 protrudes into a hydrophobic cavity formed by W606, M610, M614, M625 of hHSP90 ⁇ D;
  • F385 of hPINK1 forms a ⁇ -cation interaction with K534 of HSP90 ⁇ D;
  • L372 of hPINK1 forms a hydrophobic interaction with Y604 of HSP90 ⁇ D.
  • the ⁇ 7- ⁇ 8 could be stabilized in the place.
  • PINK1 C-lobe structure ConSurf analysis for the PINK1 C-lobe structure.
  • the 3D structure of hPINK1 C-lobe is rendered as surface.
  • the surface is color-coded by its conservation grade with cyan-through-maroon representing the nonconserved to conserved residues.
  • the interfaces on PINK1 involved in the binding to HSP90, CDC37, and FKBP51 are highly conserved.
  • the plasmid encoding the EGFP-PARKIN was constructed as described.
  • HEK293T cells were cultured in 15 cm dishes until their confluency reached ⁇ 80%.
  • 75 ⁇ g of plasmid and 225 ⁇ g of PEI were separately dissolved in PBS. Then, the PEI solution was added into the plasmid solution in droplets. The mixture was incubated at room temperature for 10 minutes before it was added into the culture media of the cells. After 5 hours, the culture media of the transfected cells were replaced with fresh media.
  • the transfected cells were treated with 1 ⁇ M CCCP overnight after 40 hours of transfection.
  • Cells were harvested after 48 hours of transfection. The culture media were gently removed; and 2 mL of PBS was added into each dish to resuspend the cells. The cell suspensions were centrifuged at 1,000 x g for 5-10 minutes at 4°C. Then, the supernatant was removed. The cells were rapidly frozen in liquid nitrogen and stored at -80°C for future application. Frozen cells were thawed and resuspended with the lysis buffer (20 mM HEPES at pH 7.5, 300 mM NaCl, 10%glycerol, 1 mM PMSF, 0.5%Triton X-100, 1 mM EDTA) .
  • the lysis buffer (20 mM HEPES at pH 7.5, 300 mM NaCl, 10%glycerol, 1 mM PMSF, 0.5%Triton X-100, 1 mM EDTA
  • the column was washed successively with 30 column volumes (cv) of Wash Buffer I (20 mM HEPES at pH 7.5, 150 mM NaCl, 10%glycerol, 1 mM PMSF, and 0.02%Triton X-100) and 30 cv of Wash Buffer II (20 mM HEPES at pH 7.5, 150 mM NaCl, and 0.02%Triton X-100) . Finally, the bound proteins were eluted from the column using Elution Buffer (20 mM HEPES at pH 7.5, 100 mM NaCl, 0.01%Digitonin, and 5mM biotin) .
  • Affinity-purified samples were further purified with size-exclusion chromatography using a Superose 6 increase 3.2/300 column (Cytiva) equilibrated in 20 mM HEPES (pH 7.5) , 100 mM NaCl, 0.01%Digitonin, and 1mM DTT.
  • the peptides were loaded onto a 150 ⁇ m ⁇ 2 cm self-packed C18 trap column (particle size 3 ⁇ m, Dr. MASCH GmbH, Germany) and separated on a 150 ⁇ m ⁇ 30 cm self-packed C18 analytical column (particle size 1.9 ⁇ m, Dr. MASCH GmbH) .
  • Mobile phase A contains 0.1%formic acid
  • mobile phase B contains 80%acetonitrile and 0.1%formic acid.
  • the program for the mobile gradients was set as follows: the gradient for the mobile phase B was running from 8 to 12%over 10 min, 12 to 27%over 69 min, 27 to 45%over 28 min, 45 to 95%over 3 min, and then 95%over 10 min.
  • the mass spectrometer was operated in the data-dependent acquisition mode using Xcalibur 4.0 software.
  • a single full-scan mass spectrum was collected in the Orbitrap (350–1800m/z, 120,000 resolution) followed by data-dependent MS2 scans at 30%collision energy (HCD) in an ion trap.
  • MS/MS spectra from each LC-MS/MS run was searched against the Swiss human protein database (version released in August 2018, containing 20, 325 sequence entries) using Proteome Discoverer (Version 2.2) searching algorithm.
  • the search criteria were set as follows: full tryptic specificity was required; two missed cleavages were allowed; carbamidomethylation was set as fixed modification; oxidation was set as dynamic modifications; precursor ion mass tolerance was 20 ppm for all MS acquired in the Orbitrap; and fragment ion mass tolerance was 0.6 Da for all MS2 spectra acquired in the ion trap.
  • High confidence score filter FDR ⁇ 1%) was used to select the target peptides and their corresponding MS/MS spectra were manually inspected.
  • hPink1 (1-581) -chaperone complex (0.45 mg/ml) or hPink1 (130-581) -chaperone complex (0.25 mg/ml) was loaded on a glow-discharged (Gatan Solarus Plasma Cleaning System 955) holy gold grid (R1.2/1.3, 300 mesh, X-Pivot, China) .
  • Grids were blotted for 1 s and flash-frozen in liquid ethane cooled with liquid nitrogen using a Vitrobot Mark IV (Thermo Fisher Scientific Inc. ) , which was operated at 4 °C and 100 %humidity.
  • the grids were subsequently transferred to a Titan Krios electron microscope (Thermo Fisher Scientific Inc. ) operated at 300 kV and equipped with a Gatan K3 Summit direct electron detector and a GIF Quantum energy filter.
  • Movie stacks were automatically collected using EPU with a preset defocus ranged from -1.2 ⁇ m to -1.8 ⁇ m in super-resolution mode. Data collection was performed at a normal magnification of 130,000 with a pixel size of on the specimen.
  • Each movie stack containing 32 frames was exposed for 1.11 s.
  • the slit width on the energy filter was 20 eV; and the total dose was for each micrograph stack.
  • Image processing was carried out in cryoSPARC 59 , and the strategies were shown in Fig. S1, S8 and S10.
  • the movie stacks were motion-corrected using MotionCor2 60 , and the defocus values were estimated with Patch CTF estimation.
  • Micrographs with contaminations or the maximum resolution worse than were excluded from calculation, resulting in a total of 18,161 micrographs (hPink1 1-581 ) and 8,111 micrographs (hPink1 130-581 ) for structural determination.
  • particles were auto-picked using blob picker, extracted with a box size of 400 pixels from 500 micrographs, and classified into 50 classes by 2D classification. 19,832 particles from 500 micrographs with representative 2D class averages were selected as the training dataset for Topaz Train 61 . Then, 4,995,315 particles were auto-picked from 18,161 micrographs using Topaz 61 . After two rounds of 2D classification, 3,693,128 particles were selected and then subjected to Ab-initio Reconstruction into five classes, which were used as the templates for Heterogeneous Refinement of all selected particles. After refinement, particles were selected from the density maps with detailed features and the highest resolution.
  • hPink1 130-581 dataset a total of 80,034 particles were auto-picked using blob picker, extracted with a box size of 400 pixels from 500 micrographs, and classified into 50 classes by 2D classification. 1,827 particles from 100 micrographs with representative 2D class averages were selected as the training dataset for Topaz Train 61 . Then, 1,861,310 particles were auto-picked from 8, 111 micrographs using Topaz 61 . After one round of 2D classification, 1,724,632 particles were selected and split into two sets randomly, and then subjected to Ab-initio Reconstruction in four classes, which were used as the templates for Heterogeneous Refinement of all selected particles.
  • 658,543 particles were selected from the density maps with detailed features and the highest resolution. These particles were subsequently subjected to a second round of 2D classification to select 554, 347 particles with better quality.
  • the selected particles were refined with CTF refinement and reconstructed with Non-uniform Refinement to improve the resolution of the density to To separate individual conformations from these particles, we performed a 3D variability analysis in simple mode (20 low-resolution density maps for further movie creation) and cluster mode, respectively.
  • the 407,577 particles selected from cluster 1 and cluster 2 (representing ‘prime-activation’ conformation) were subjected to Non-uniform refinement, yielding a map with resolution.
  • cluster 3 (representing ‘autoinhibited’ conformation) were further split into 2 clusters.
  • Non-uniform refinement of cluster 1 (75,286 particles) and cluster 2 (71, 484 particles) yielded density maps with and resolutions, respectively.
  • HSP90 complexes and TcPINK1 were individually docked into the Cryo-EM density maps using Chimera 65 .
  • the models were iteratively rebuilt in COOT and refined in Phenix 66, 67 .
  • the geometry of the refined models was validated with MolProbity 68 .
  • the statistics for the Cryo-EM data collection and model refinement were reported in supplemental Table 1.
  • the refined coordinates and Cryo-EM data were deposited into PDB and EMDateBank, respectively.
  • HEK293T cells were cultured in DMEM supplemented with 10%FBS at 37 °C in a CO 2 incubator. Upon ⁇ 80%confluency, the cells in a 10-cm dish were transfected with 25 ⁇ g of the plasmid encoding the Flag-tagged full-length hPINK1 using 75 ⁇ g of PEI dissolved in PBS. After 5 hours, the transfected cells were resuspended in the fresh media. And they were evenly passed to the wells of a 24-well plate. After 36 hours, 2 ⁇ M CCCP was added into each well. Next, 10 ⁇ M geldanamycin was separately added into each of 7 groups of triplicated wells at different time points ranging from 12 hours to 1 hour.
  • the cells in each well were washed with 500 ⁇ L of PBS and then were lysed with 80 ⁇ L RIPA supplemented with 1 mM PMSF and 1 mM Na 3 VO 4 .
  • the lysed cells were frozen at -20 °C for over 30 minutes before they were thawed and cleared by centrifugation at 12,000 rpm for 10 min at 4°C.
  • the supernatants were mixed with 6 ⁇ SDS loading buffer for Western-Blotting analysis.
  • the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit (Abmart Cat #: M21002) and goat anti-mouse (Abmart Cat #: M21001) IgG antibodies, respectively, for 1 h at 4 °C.
  • the wild type and hPINK1 knock-out HeLa cells were cultured in DMEM supplemented with 10%FBS in 35 mm confocal dishes at 37 °C in a CO 2 incubator. Just before the transfection, the culture media was replaced with fresh media. Then, the cells in each well at ⁇ 80%of confluency were transiently transfected with 2.5 ⁇ g /well of the plasmid encoding the EGFP-Parkin using 7.5 ⁇ g/well of PEI. After 4 hours, the culture media was replaced with the fresh media.
  • WT Hela cells were passed into 24-well plates and cultured in DMEM supplemented with 10%FBS at 37 °C in a CO 2 incubator. Upon ⁇ 80%confluency, the cells were treated respectively with none, 10 ⁇ M CCCP, 20 ⁇ M geldanamycin, 2 ⁇ M geldanamycin, 20 ⁇ M geldanamycin plus 10 ⁇ M CCCP, 2 ⁇ M geldanamycin plus 10 ⁇ M CCCP, 2 ⁇ M SAFit2, 200 nM SAFit2, 2 ⁇ M SAFit2 plus 10 ⁇ M CCCP, and 200 nM SAFit2 plus 10 ⁇ M SAFit2 for 12 hours.
  • the Pearson’s R values for the co-localization between EGFP-Parkin and MitoTracker were calculated from images of randomly selected EGFP-Parkin positive cells collected from replicated transfection experiments for each sample. The statistics and distribution of these values were analyzed using GraphPad Prism. The statistical significance was determined with unpaired Student’s T test and was expressed as: *, p ⁇ 0.05; **, p ⁇ 0.005. The error bar represents the mean ⁇ SD.
  • hPINK1 activity regulation we transiently transfected the plasmids encoding the full-length hPINK1 or N-terminally truncated hPINK1 (130-581) into HEK293T cells and purified the hPINK1 complexes by affinity chromatography. Affinity-purified hPINK1 complex was eluted in a sharp and symmetric peak from Superose 6 increase column. This peak contains multiple protein bands on Coomassie blue staining SDS-PAGE (Figs. 1a, b, Fig 7a, b) . LC-MS/MS analysis of these bands revealed that the major components of the complex include HSP90 ⁇ , CDC37, FKBP51, and PINK1 (Fig. 1c, Fig. 7c) .
  • hPINK1 In FKBP51-bound hPINK1 complex, two HSP90 ⁇ s in closed conformation form a pseudo-symmetric but elongated dimer (Fig. 1f, Fig. 7e) .
  • the N-lobe of hPINK1 is unfolded with its ⁇ 5 strand threading through the lumen of HSP90 ⁇ dimer.
  • the N-terminal domain of CDC37 interacts with both the C-lobe of hPINK1 and one HSP90 ⁇ protomer, while the M domain of CDC37 binds to the other HSP90 ⁇ protomer located on the other side of the complex.
  • hPINK1 The C-lobe and CTD of hPINK1 interact with one HSP90 ⁇ protomer, the FK1 domain of FKBP51, and the N-terminal domain of CDC37. As such, CDC37 orchestrates the interactions between HSP90 ⁇ and hPINK1.
  • HSP90 ⁇ serves as a central organizer to coordinate the interactions between FKBP51 and its client.
  • HSP90 ⁇ In the class lacking FKBP51, HSP90 ⁇ , CDC37, and hPINK1 adopt nearly identical interactions and assembly as they are in the complex of HSP90 ⁇ /CDC37/FKBP51/PINK1 (The RMSD of 1685 C ⁇ atoms between these two structures is ) .
  • Fig. 1g, Fig. 7f The RMSD of 1685 C ⁇ atoms between these two structures is .
  • R46 forms a salt-bridge contact with the catalytic residue, E47, to deprotonate the E47 (Fig. 2a) . It had been shown that the salt-bridge contact between R46 and E47 is a critical switch point for coupling the ATP hydrolysis with the global conformational changes of HSP90 46 .
  • FKBP51 binds to the C-terminal domains of HSP90 ⁇ .
  • the ⁇ 7E helix from the TPR domain of FKBP51 is anchored to the groove at the dimeric interface of HSP90 ⁇ C-terminal domains through hydrophobic and electrostatic interactions (Fig. 2c, Fig. 10) .
  • the periodicity of the helix is broken at A410 to accommodate its mismatch with the symmetry of HSP90 ⁇ dimer (Fig. 2c) .
  • the C-terminal tail from the CDC37-engaged HSP90 ⁇ protomer (the subunit D) forms a L-shaped clamp to escort and lock the ⁇ 7E helix in place (Fig. 2c) .
  • the conserved MEEVD motif from HSP90 ⁇ forms a short helix and is embedded at the cavity of the TPR domain (Fig. 2d, Fig. 11) .
  • the overall conformation of FKBP51 is flexible for adapting to the recognition of different clients (Fig. 2e) .
  • the individual TPR, FK1 and FK2 domains of the FKBP51 remain nearly identical with RMSDs of C ⁇ s ranging from to But the two linkers connecting these three domains are relatively flexible.
  • the FK1 domains are rotated by 19° relative to each other when PINK1-bound and GR-bound FKBP51 structures are superimposed on their FK2 domains.
  • FK2 domains rotate 7° relative to each other when their TPR domains are superimposed.
  • the folding of the kinase N-lobe is regulated by the association of HSP90 ⁇ with PINK1 ⁇ 5 strand.
  • the ⁇ 5 strand is the central component for organizing the hydrogen bonding network of the kinase N-lobe ⁇ sheet.
  • its neighboring ⁇ 4 and ⁇ 3 strands could flip over under different conformational states (Fig. 3c) 52 . Therefore, retracting ⁇ 5 strand will destroy the hydrogen bonding network required for the properly folding of the kinase N-lobe (Fig. 3c) . Consistently, the density for PINK1 N-lobe in its HSP90 chaperone complex was not resolvable.
  • the ⁇ 15 ⁇ in the C-domain of HSP90 ⁇ becomes disordered and moves away, creating a hydrophobic cavity to accommodate the protrusion of PINK1’s ⁇ 7- ⁇ 8 loop (Fig. 12c) .
  • three mutations associated with PD are in this ⁇ 7 strand.
  • one of these mutations (I368N) had been shown to be able to lower the levels of PINK1 associated with HSP90/CDC37 complex and to impair the import of PINK1 into TOM40 complex under stress 53 .
  • CDC37 The interaction of CDC37 with its clients are regulated by its phosphorylation 8 .
  • the phosphorylation on Ser13 of CDC37 stabilizes the proper orientation of the CDC37 fragment (a.a. 13-34) so that this fragment could insert into a highly conserved pocket on PINK1 C-lobe and in the proximity of the ⁇ 7- ⁇ 8 strands (Fig. 3d) .
  • the HPNI motif on this fragment could mimic the configuration of the folded functional PINK1 ⁇ i- ⁇ 4 loop to stabilize the C-lobe of the unfolded kinase (Fig. 3c, d) .
  • HSP90/CDC37/FKBP51 binds to three important but highly conserved structural elements of PINK1: the ⁇ 5 strand, the activation loop, and the CTD.
  • the ⁇ 5 strand is essential for coordinating the hydrogen-bonding network of the functional kinase N-lobe ⁇ -sheet 49, 50, 54 . Clamping this strand in the lumen of HSP90 dimer would disrupt such hydrogen-bonding network and stabilize the kinase in an inactive, unfolded state, which is required for the import of PINK1 into mitochondria via TOM complex or degradation of PINK1 via proteasome 56 .
  • the CTD of PINK1 is involved in the regulation of PINK1 kinase activities 52 .
  • the binding of HSP90 to the ⁇ L helix dislodges the ⁇ K helix from its association with the ⁇ L helix and impairs the dimeric assembly and activation of the kinase 52 .
  • many PD-associated mutations are enriched in the ⁇ K helix (Fig. 10b) .

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Medicinal Chemistry (AREA)
  • Organic Chemistry (AREA)
  • Neurosurgery (AREA)
  • Veterinary Medicine (AREA)
  • Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • General Chemical & Material Sciences (AREA)
  • Biomedical Technology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Neurology (AREA)
  • Pharmacology & Pharmacy (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Psychology (AREA)
  • Cardiology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Peptides Or Proteins (AREA)

Abstract

Provided is the use of a FKBP51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities and a method for treating or preventing diseases caused by mitochondrial abnormalities, comprising administering to a patient in need thereof a FKBP51 inhibitor.

Description

Use of a FKBP51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities Technical Field
The present application is directed to use of a FKBP51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities.
Background
Parkinson’s disease is a chronic degenerative disease of the central nervous system that mainly affects the motor system. Common symptoms of the disease include tremor, rigidity, slowness of movement and difficulty with walking. These motor problems are commonly denominated “parkinsonism” or “parkinsonian syndrome” . In addition, non-motor related symptoms may occur such as depression, dysautonomia, sleep disorder, anxiety, fatigue and dementia.
Dysfunctional mutations of PINK1, a key regulator of mitophagy, are associated with familial Parkinson’s disease. Therefore, the quantity and activity of PINK1 need to be tightly regulated. Meanwhile, in response to stress, the expression of HSP90α and FKBP51 are stimulated to get into the folding, maturation, and degradation of their clients, including kinases and nuclear receptors.
Precise regulation of the quantity and activity of cellular proteins in a spatiotemporally controlled manner is critical for the normal physiology of the cells. Dysregulation of such process leads to the misfolding, aggregation, and degradation of functional proteins, which in turn are associated with various human diseases1, 2. Central to this quantity and activity control machinery are the molecular chaperons represented by the HSP90s1, 2. These chaperons, aided by their co-chaperons in a cell-specific and context-dependent manner, facilitate the folding, stabilization, and molecular interactions of its ‘client’ proteins within the crowded cellular environment1-3.
Driven by the ATP hydrolysis, HSP90s undergo chaperone cycles to promote the folding, translocation, and degradation of its client proteins4, 5. Co-chaperones are engaged into this cycle at different stages to act as a scaffold for loading clients (i.e. CDC37) 6-8, modulating the conformation and ATPase activity of HSP90 (i.e. CDC37, Aha1) 6, 9, 10, promoting the release of matured clients (i.e. P23, HOP) 11, 12, regulating the degradation of unfolded clients (i.e. FKBP51, FKBP52, CHIP) 13-16, or controlling the activities of clients via other enzymatic functions (i.e. PP5) 17, 18. Among these co-chaperones, CDC37 is essential for the folding and stabilization of many kinase clients. Serving as a scaffold, CDC37 loads clients from HSP70 onto HSP90, inhibits the ATPase activity of HSP90, and stabilizes unfolded clients in the HSP90 chaperone complexes6-10, 19, 20.
Immunophilins, such as FKBP51 and FKBP52, represent another class of co-chaperones for HSP9013-15. FKBP51 and FKBP52 are two TPR (TetratricoPeptide Repeat) domain containing peptidyl-prolyl cis/trans isomerase (PPIase) with mutually antagonizing functions against each other13-15, 21. They are involved in the inflammation, innate immunity, stress response, and neuronal protection, and thus are actively explored as potential druggable targets for treating stress-related diseases14, 22, 23. Recent structural studies revealed how they are involved in the regulation of the glucocorticoid receptor activation and maturation13, 24. But it remains largely elusive how they are involved in the folding, activation, and maturation of kinase clients.
PINK1, a nucleus-encoded mitochondrial kinase, is critical for the quality control and autophagy of mitochondria25-28. Under the resting and non-stressed conditions, expressed PINK1 is translocated into mitochondria through the TOM complex29, cleaved by PARL30, 31 and MPP32, retro-translocated back to the outer membrane of mitochondria, and eventually degraded by proteasome33. As such, it is kept at the lower protein and activity levels. Upon the stimulation by various signals of mitochondrial damage, the full-length PINK1 is arrested and accumulated on the outer membrane of mitochondria, activated through dimerization and association with TOM complex29, 34, recruits PARKIN (an E3 ligase) to damaged mitochondria via phosphorylation of PARKIN and ubiquitylated proteins on mitochondria35-37. Recruited PARKINs are fully activated and trigger the clearance of damaged mitochondria via proteasome36-38. In these series of cascades, precise regulation of PINK1’s quantity and activity on mitochondria is the key for coordinating these events in mitophagy. However, the structures and molecular details for this regulation is largely undefined.
Loss-of-function mutations of PINK1 are linked to the early onset of autosomal recessive Parkinson’s disease (PD) 39-41. Therefore, PINK1 could be a potential target for the treatment of PD42, 43. Indeed, early studies focused on the development of ATP analogs to up-regulate the kinase activities of PINK1 mutants43. But these efforts were still under debate in proofing the concept in animal models44, 45. Development of alternative strategies in the treatment of PINK1-deficient PD is challenging but worth to be explored.
Herein, we determined Cryo-EM structures of HSP90α/CDC37/PINK1 complex in the presence and absence of FKBP51. These structures revealed the molecular details for how HSP90α/CDC37/FKBP51complex stabilizes unfolded hPINK1 kinase to regulate the activity and stability of hPINK1. In addition, we studied how HSP90 and FKBP51 inhibitors regulate the stability and PINK1-mediated PARKIN recruitment in response to mitochondrial damage. Collectively, our study provided a framework for understanding chaperone-mediated PINK1 regulation and a clue for the development of strategies in the treatment of PINK1-dificient PD.
Summary of the Present Application
Herein, the present invention determined the Cryo-EM structures and molecular recognition for the HSP90α/CDC37/FKBP51/PINK1 complexes, and these data disclosed a general mechanism for client recognition by HSP90α/CDC37/FKBP51 complex and the molecular basis for coupling PINK1 stability, folding, and degradation in its activity regulation. Moreover, the present invention revealed an alternative avenue and a potential target for treating PINK1-deficient Parkinson’s disease. For example, the present invention identified that specific inhibition of FKBP51 could promote PINK1-mediated PARKIN recruitment in response to mitochondrial depolarization.
In the first aspect, the present invention provided use of a FKBP51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities.
In one embodiment of the first aspect, the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof. SAFit2 is represented by the following formula:
In one embodiment of the first aspect, the disease is Parkinson disease. In a preferred embodiment of the first aspect, the Parkinson disease is PINK1-dificient Parkinson disease. In one embodiment of the first aspect, the disease is cardiomyositis, coronary heart disease or liver fibrosis.
In the second aspect, the present invention provided a method for treating or preventing diseases caused by mitochondrial abnormalities, comprising administering to a patient in need thereof a FKBP51 inhibitor.
In one embodiment of the second aspect, the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof. In one embodiment of the second aspect, the disease is Parkinson disease. In a preferred embodiment of the second aspect, the Parkinson disease is PINK1-dificient Parkinson disease. In one embodiment of the second aspect, the disease is cardiomyositis, coronary heart disease or liver fibrosis.
In the third aspect, the present invention provided use of a HSP90 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities.
In one embodiment of the third aspect, the HSP90 inhibitor is geldanamycin or a pharmaceutically acceptable salt thereof, geldanamycin is represented by the following formula:
In one embodiment of the third aspect, the disease is Parkinson disease. In a preferred embodiment of the third aspect, the Parkinson disease is PINK1-dificient Parkinson disease. In one embodiment of the third aspect, the disease is cardiomyositis, coronary heart disease or liver fibrosis.
In the fourth aspect, the present invention provided a method for treating or preventing diseases caused by mitochondrial abnormalities, comprising administering to a patient in need thereof a HSP90 inhibitor.
In one embodiment of the third aspect, the HSP90 inhibitor is geldanamycin or a pharmaceutically acceptable salt thereof. In one embodiment of the fourth aspect, the disease is Parkinson disease. In a preferred embodiment of the fourth aspect, the Parkinson disease is PINK1-dificient Parkinson disease. In one embodiment of the fourth aspect, the disease is cardiomyositis, coronary heart disease or liver fibrosis.
In the fifth aspect, the present invention provided a pharmaceutical composition comprising a FKBP51 inhibitor or a HSP90 inhibitor and one or more pharmaceutically acceptable excipients, wherein the composition is used for treating or preventing diseases caused by mitochondrial abnormalities, preferably Parkinson disease. In one embodiment of the fifth aspect, the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof and the HSP90 inhibitor is geldanamycin or a pharmaceutically acceptable salt thereof.
The present application may be embodied in any other forms without departing from the spirit or scope thereof. The present application encompasses any and all combinations of the above aspects and embodiments. It is to be understood that any embodiment may be combined with any other embodiment (s) to describe an additional embodiment. It is also to be understood that an individual element from any embodiment may be combined with any and all other elements from any other embodiment (s) to describe an additional embodiment.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 shows biochemical characterization and Cryo-EM structures of hPINK1/HSP90α/CDC37 in complex with or without FKBP51, wherein:
a. Schematic diagram showing the domain organization of the proteins presented in the complex. The stripped/dashed parts indicate regions where there is no interpretable density. Red stars indicate ubiquitinated residues; and red circles indicate phosphorylated residues (S13 and T118) identified by mass spectrometry.
b. Size-exclusion chromatography profile of the Strep-affinity purified complex. Molecular mass of the protein markers is shown on the top of the profile and marked with arrows. Each elution peak is numbered and marked by red dots at the bottom of the profile (upper) . Each peak fraction is subjected to electrophoresis on a 12%SDS-PAGE gel stained with Coomassie blue (lower left) and analyzed by Western blotting with the anti-FLAG antibody (lower right) . hPINK1 is indicated by red arrow.
c. Mass spectrometry analysis of the composition of the protein bands. Endogenous HSP90α, CDC37 and FKBP51 proteins were co-purified with hPINK1.
d. Cryo-EM map of the hPINK1/HSP90α/CDC37/FKBP51 complex. The map is colored according to the color of each subunit shown in Fig. 1a, except two monomers of the HSP90αdimer are colored in yellow and cyan, respectively.
e. Cryo-EM map of the hPINK1/HSP90α/CDC37 complex. The map is colored as Fig. 1d.
f/g. Overview of the hPINK1/HSP90α/CDC37/FKBP51 and hPINK1/HSP90α/CDC37 complexes bound with ATP.
Figure 2 shows detailed interactions and assembly of the HSP90α/CDC37/FKBP51 complex, wherein:
a. The active site of HSP90α subunit D. The ATP molecule bound at the NTD of HSP90α and the highly conserved residues V2, Y4, and W7 from CDC37 (malate) are shown as sticks covered with Cryo-EM map. F134 of the subunit D forms a stronger π-cation interaction with R400; R46 forms a salt-bridge contact with E47 to facilitate the deprotonation of E47 for ATP hydrolysis. HSP90α subunit C is presented in light yellow surface; HSP90α subunit D is in light blue cartoon.
b. The active site of HSP90α subunit C. The ATP molecule bound at the NTD of HSP90α is shown as sticks covered with Cryo-EM map. Compared with that of HSP90αsubunit D, the interaction between F134 and R400 is impaired as the benzene ring of F134 moved away. The amino group of R46 is away from the carboxyl group of E47, which is not favorable for ATP hydrolysis. HSP90α subunit C is in cartoon; HSP90α subunit D is in surface representation.
c. The TPR domain of FKBP51 (green) binds to the CTDs of HSP90α (yellow/cyan) . The α7 helix is shown as sticks covered with Cryo-EM map. It is anchored to the groove at the dimeric interface of HSP90α CTDs. The periodicity of the helix is broken at A410 when it interacts with HSP90α dimer with a mismatched symmetry.
d. The Cryo-EM map around the conserved MEEVD fragment from HSP90α subunit D (cyan) . It forms a short helix to interact with the TPR domain of FKBP51 (green) .
e. Comparison of the structures of HSP90/FKBP51 complexed with or without clients. The FKBP51/HSP90/P23 complex (PDB: 7l7i, left) , PINK1/FKBP51/HSP90/CDC37 complex (PDB: 8XOZ, middle) , and GR/FKBP51/HSP90 complex (PDB: 8ffw, right) are shown in cartoon.
Figure 3 shows the recognition of hPINK1 by HSP90α/CDC37/FKBP51 complex, wherein:
a. The interface for hPINK1/HSP90α D interaction. HSP90α D is depicted in surface representation colored by hydrophobicity (cyan: polar, yellow: nonpolar) . hPINK1 shown in cartoon is in magenta. The stretched β5 of hPINK1 threading through the lumen of HSP90 is shown as sticks, while the β7-β8 of hPINK1 is shown in red.
b. The sequence conservation for PINK1 β5. Sequence conservation of PINK1 is mapped onto the atomic model of hPINK1 from the hPINK1/HSP90/CDC37/FKBP51 complex shown in surface. Conservation is colored from maroon (conserve) to cyan (nonconserve) . L314, F315, L316 at the central region of the β5 strand are highly conserved in PINK1 family.
c. Superposition of the C-lobe from the unfolded hPINK1 in hPINK1/HSP90/CDC37/FKBP51 complex and that from TcPINK1 (PDB: 8ATQ) showing the conformational changes of the kinase. TcPINK1 is shown as gray cartoon with β5 and αK colored in golden. hPINK1 is shown in surface representation colored by sequence conservation as in Fig. 3b.
d. The HPNI motif on CDC37 (blue) mimics the configuration of the folded, functional PINK1αi-β4 loop to stabilize the C-lobe of the unfolded kinase. The interface of PINK1 for CDC37 binding is highly conserved.
e. Superposition of the CTD from the unfolded hPINK1 (magenta) and that from TcPINK1 (golden) . The secondary structures of the αL and αK are partially disrupted in the unfolded hPINK1 structure, forming a flower-shaped ring with αM. These three helixes are reorganized in the folded TcPINK1 structure and packed against each other.
f. Superposition of the unfolded hPINK1 and the dimeric TcPINK1. The reorganized αK helix interacts with the N-lobe of another PINK1 to stabilize the dimeric assembly of the kinase. TcPINK1 is colored in gray with αK, αL, αM colored in golden. hPINK1 is colored in magenta.
g. The Cryo-EM density map showing the interaction between the FK1 domain of FKBP51 (green) and hPINK1 (magenta) . The activation loop (G386-C412) is invisible in electron density map. The position of the activation loop of PINK1 is in the proximity to the ligand binding site of the FK1 domain.
Figure 4 shows that inhibition of the ATPase activity of HSP90α by geldanamycin promotes the degradation of PINK1 under stress, wherein:
a. Geldanamycin (GA) promotes the degradation of PINK1 under mitochondrial stress in transiently transfected HEK293T cells. The result from Western blotting and statistics from triplicated technical repeats were shown on the upper and lower panels, respectively. Relative expression was normalized against the average of the band 2 densities. The statistical significance was represented (ns, no significance; *, p <0.05; **, p <0.01; ***, p <0.001; ****, p<0.0001) .
b. The level of phosphorylated poly-ubiquitin displayed a trend of increase but was not statistically significant. The quantity of phosphorylated poly-ubiquitin is analyzed by Western blotting with an anti-pUb antibody (upper) . The quantitative results were shown in the lower panel. Relative phosphorylation was normalized against the average of total phosphorylated poly-ubiquitin densities.
c/d. The effect of GA on PINK1-mediated PARKIN recruitment. Representative confocal images of the WT (c) or PINK1 KO (d) HeLa cells transiently transfected with the plasmid encoding the EGFP-PARKIN. Transfected cells were treated with none, 5 μM CCCP, 10 μM geldanamycin, 10 μM geldanamycin plus 5 μM CCCP, respectively, for 1 hour before they were fixed. Mitochondria were stained with 100 nM Mitotracker.
e. Statistical analysis of the colocalization of EGFP-PARKIN and Mitotracker under different conditions. Over 70 images of EGFP-PARKIN+ cells were selected from replicated experiments for each sample and manually masked with ImageJ. The Pearson’s R values for the colocalization of EGFP-PARKIN and Mitotracker in each cell were calculated with the Coloc 2 program in ImageJ.
Figure 5 shows that SAFit2 promotes PINK1-mediated PARKIN recruitment in mitophagy, wherein:
a. SAFit2 has little time-dependent effect on the degradation of PINK1 in transiently transfected HEK293T cells. The result from Western blotting and statistics from triplicated technical repeats were shown on the upper and lower panels, respectively. Relative expression was normalized against the average of the band 2 densities.
b. SAFit2 has little time-dependent effect on the phosphorylation of poly-ubiquitin in transiently transfected HEK293T cells. The quantity of phosphorylated poly-ubiquitin is analyzed by Western blotting with an anti-pUb antibody (upper) . The quantitative results were shown in the lower panel. Relative phosphorylation was normalized against the average of total phosphorylated poly-ubiquitin densities.
c/d. SAFit2 could promote PINK1-mediated PARKIN recruitment. Representative confocal images of the WT (c) or PINK1 KO (c) HeLa cells transiently transfected with the plasmid encoding the EGFP-PARKIN. Transfected cells were treated with none, 2 μM CCCP, 4 μM SAFit2, 4 μM SAFit2 plus 2 μM CCCP, respectively, for 1 hour before they were fixed. Mitochondria were stained with 100 nM Mitotracker.
e. Statistical analysis of the colocalization of EGFP-PARKIN and Mitotracker under different conditions. Over 70 images of EGFP-PARKIN+ cells were randomly selected from replicated experiments for each sample and manually masked with ImageJ. The Pearson’s R values for the colocalization of EGFP-PARKIN and Mitotracker in each cell were calculated with the Coloc 2 program in ImageJ.
f. Co-admission of SAFit2 but not GA could enhance the expression and/or stabilization of the endogenous PINK1 in WT Hela cells stimulated with CCCP overnight. The quantity of PINK1 is analyzed by Western blotting with an anti-hPINK1 antibody (upper) . The quantitative results were shown in the lower panel.
Figure 6 shows model for hPINK1 folding and functional regulation by HSP90α/CDC37/FKBP51 chaperone system under resting and stressed states, wherein: Under normal physiological conditions, the majority of PINK1 undergoes degradation via ubiquitin-proteasome pathway. Meanwhile, other unfolded PINK1s associated with CDC37 are stabilized by HSP90 and then transported into mitochondria, where they could be further stabilized by TRAP1 in the mitochondrial matrix. These two pathways in together prevent uncontrolled PINK1/PARKIN-mediated mitochondrial autophagy for the health of cells (left) . Under stressed conditions, the syntheses of HSP90α and FKBP51 are stimulated, whereas they could form a complex with CDC37-associated PINK1. The translocation of PINK1 into damaged mitochondria is stalled so that increasing amount of PINK1 are accumulated on the outer membrane of mitochondria. Subsequently, via not fully-defined mechanisms, PINK1s are activated to initiate PARKIN recruitment for mitophagy. Consequently, the HSP90 inhibitor GA could suppress the folding and maturation of PINK1 by inhibiting the ATPase activity of HSP90, resulting in PINK1 degradation. During this process, FKBP51 negatively regulates PINK1 functions to prevent excessive autophagy of mitochondria. Thus, the specific inhibitor for FKBP51, the SAFit2, could enhance the maturation of PINK1 on mitochondria and thereby promotes the clearance of damaged mitochondria (right) .
Figure 7 shows biochemical characterization and Cryo-EM structures of the hPINK1 (130-581) /HSP90α/CDC37 in complex with or without FKBP51, wherein:
a. Size-exclusion chromatography profile of the Strep-affinity purified complex. Molecular mass of the protein markers is shown on the top of the profile and marked with arrows. Each elution peak is numbered and marked by red dots at the bottom of the profile.
b. The peak fractions from size-exclusion chromatography were subjected to electrophoresis on a 12%SDS-PAGE gel stained with Coomassie blue (left) and analyzed by Western blotting with an anti-FLAG antibody (right) . hPINK1 is indicated by red arrow.
c. Mass spectrometry analysis of the composition of the protein bands. Endogenous HSP90α, CDC37 and FKBP51 proteins were co-purified with hPINK1 (130-581) .
d. Cryo-EM map of the hPINK1 (130-581) /HSP90α/CDC37/FKBP51 complex. The map is colored according to the color of each subunit shown in Fig. 1a, except two monomers of the HSP90α dimer are colored in yellow and cyan, respectively.
e/f. Overview of the complex of hPINK1 (130-581) /HSP90α/CDC37/FKBP51 (e) and hPINK1 (130-581) /HSP90α/CDC37 (f) .
Figure 8 shows overall strategies and statistics in processing the Cryo-EM data of HSP90α/CDC37/FKBP51/hPINK1.
Figure 9 shows overall strategies and statistics in processing the Cryo-EM data of HSP90α/CDC37 /hPINK1.
Figure 10 shows multiple sequence alignment of FKBPs (a) and the kinase clients of HSP90/CDC37 (b) . The conserved residues are highlighted with solid shades. Every 10th residue is marked with black dots on top of the sequences. The α helices and β strands of FKBP51 (a) or hPINK1 (b) are shown with rods and arrows, respectively. The secondary structures of FK1, FK2, TPR domains on FKBP51 are colored by light gray, white, dark gray, respectively. Residues located at each interface of interactions are marked at the bottom of the sequences as shown. EOPD mutations were marked with red stars on top of the sequences (b) .
Figure 11 shows multiple sequence alignment of HSP90s. The conserved residues are highlighted with solid shades. Every 10th residue is marked with black dots on top of the sequences. The α helices and β strands of HSP90α are shown with rods and arrows, respectively. The secondary structures of NTD, MD, CTD domain are colored by light gray, white, dark gray, respectively. Residues located at the interface of interactions with PINK1, FKBP51, CDC37 are marked at the bottom of the sequences.
Figure 12 shows detailed interactions in the recognition of hPINK1 by the HSP90α/CDC37/FKBP51 complex, wherein:
a. The main interactions between hPINK1 (magenta and red) and HSP90α D (cyan) . the H310-Y321 region of hPINK1 is shown as sticks and covered with cryo-EM map. P374 on the β7-β8 loop of hPINK1 protrudes into a hydrophobic cavity formed by W606, M610, M614, M625 of hHSP90α D; F385 of hPINK1 forms a π-cation interaction with K534 of HSP90α D; L372 of hPINK1 forms a hydrophobic interaction with Y604 of HSP90α D. As such, the β7-β8 could be stabilized in the place.
b. The stretched β5s from the clients of HSP90/CDC37 system are shown in sticks and surface. The surface is colored according to the hydrophobicity (cyan: polar, yellow: nonpolar) . The hydrophobic residues L316 (hPINK1) , W423 (RAF1) , V92 (CDK1) are aligned at the center of the strand.
c. Comparison of the structures of hPINK1 (magenta) with RAF1 (blue, left) or CDK4 (green, right) superimposed on their associated-HSP90s. The α15` of HSP90α becomes disordered and moves away to accommodate the protrusion of PINK1’s β7-β8 loop. The HSP90/CDC37 complexed with RAF1 or CDK4 is colored in gray except the client. The HSP90/CDC37 associated with PINK1 is colored as follows: HSP90α C, yellow; HSP90α D, cyan; CDC37, slate.
d. The interaction between PINK1 (magenta) and HSP90α C (light yellow) (left) . The PINK1 residues involved in the interaction are mainly from the disordered αL and αJ helixes. K555, F558, L559, and L517 on the interface are highly conserved in PINK1 family (right) . W379 forms hydrophobic interaction with HSP90α C to stabilize the PINK1 β8.
e. ConSurf analysis for the PINK1 C-lobe structure. The 3D structure of hPINK1 C-lobe is rendered as surface. The surface is color-coded by its conservation grade with cyan-through-maroon representing the nonconserved to conserved residues. The interfaces on PINK1 involved in the binding to HSP90, CDC37, and FKBP51 are highly conserved.
Examples
Materials and Methods
Plasmid construction
The gene encoding the hPINK1 (a.a. 1-581) was subcloned into the pEF-1 vector at the BamHI and XhoI sites. In the construct, hPINK1 (a.a. 1-581) was fused with a 3×FLAG-tag followed by a Strep II-tag at the protein C-terminus. The genes encoding the hPINK1 (a.a. 130-581) and hTRAP1 (a.a. 60-704) were subcloned into a bicistronic vector, the pET3, at the BamHI and XhoI sites. These two genes were linked together by an IRES element for co-expression of the two proteins from the same transcribed mRNA. In the construct, hPINK1 (a.a. 130-581) was also fused with a 3×FLAG-tag followed by a Strep II-tag at the protein C-terminus.
The plasmid encoding the EGFP-PARKIN was constructed as described.
Expression and purification of hPINK1 (1-581) and hPINK1 (130-581)
HEK293T cells were cultured in 15 cm dishes until their confluency reached ~ 80%. In transfection, 75 μg of plasmid and 225 μg of PEI were separately dissolved in PBS. Then, the PEI solution was added into the plasmid solution in droplets. The mixture was incubated at room temperature for 10 minutes before it was added into the culture media of the cells. After 5 hours, the culture media of the transfected cells were replaced with fresh media.
For expressing hPINK1 (130-581) , the transfected cells were treated with 1 μM CCCP overnight after 40 hours of transfection.
Cells were harvested after 48 hours of transfection. The culture media were gently removed; and 2 mL of PBS was added into each dish to resuspend the cells. The cell suspensions were centrifuged at 1,000 x g for 5-10 minutes at 4℃. Then, the supernatant was removed. The cells were rapidly frozen in liquid nitrogen and stored at -80℃ for future application. Frozen cells were thawed and resuspended with the lysis buffer (20 mM HEPES at pH 7.5, 300 mM NaCl, 10%glycerol, 1 mM PMSF, 0.5%Triton X-100, 1 mM EDTA) . The resuspended cell mixture was passed through a 20-Gauge syringe for 20-30 times. Subsequently, the mixture was cleared by centrifugation at 20,000 x g for 30 minutes at 4℃. The cleared supernatant was mixed with Strep II-affinity resin and incubated on a shaker for 1 hour. Then, the resin was loaded onto a column. The column was washed successively with 30 column volumes (cv) of Wash Buffer I (20 mM HEPES at pH 7.5, 150 mM NaCl, 10%glycerol, 1 mM PMSF, and 0.02%Triton X-100) and 30 cv of Wash Buffer II (20 mM HEPES at pH 7.5, 150 mM NaCl, and 0.02%Triton X-100) . Finally, the bound proteins were eluted from the column using Elution Buffer (20 mM HEPES at pH 7.5, 100 mM NaCl, 0.01%Digitonin, and 5mM biotin) .
Affinity-purified samples were further purified with size-exclusion chromatography using a Superose 6 increase 3.2/300 column (Cytiva) equilibrated in 20 mM HEPES (pH 7.5) , 100 mM NaCl, 0.01%Digitonin, and 1mM DTT.
Analysis of mass spectrometry data
The protein bands on Coomassie-blue stained SDS-PAGE gel was cropped and digested in Eppendorf tube with 0.5μM Trypsin for 16 hours. The digested peptides were analyzed with a Fusion Lumos mass spectrometer coupled with an Easy-nLC 1200 system (Thermo Fisher Scientific) .
In the analysis, the peptides were loaded onto a 150 μm × 2 cm self-packed C18 trap column (particle size 3 μm, Dr. MASCH GmbH, Germany) and separated on a 150 μm × 30 cm self-packed C18 analytical column (particle size 1.9 μm, Dr. MASCH GmbH) . Mobile phase A contains 0.1%formic acid, while mobile phase B contains 80%acetonitrile and 0.1%formic acid. The program for the mobile gradients was set as follows: the gradient for the mobile phase B was running from 8 to 12%over 10 min, 12 to 27%over 69 min, 27 to 45%over 28 min, 45 to 95%over 3 min, and then 95%over 10 min.
The mass spectrometer was operated in the data-dependent acquisition mode using Xcalibur 4.0 software. A single full-scan mass spectrum was collected in the Orbitrap (350–1800m/z, 120,000 resolution) followed by data-dependent MS2 scans at 30%collision energy (HCD) in an ion trap. MS/MS spectra from each LC-MS/MS run was searched against the Swiss human protein database (version released in August 2018, containing 20, 325 sequence entries) using Proteome Discoverer (Version 2.2) searching algorithm. The search criteria were set as follows: full tryptic specificity was required; two missed cleavages were allowed; carbamidomethylation was set as fixed modification; oxidation was set as dynamic modifications; precursor ion mass tolerance was 20 ppm for all MS acquired in the Orbitrap; and fragment ion mass tolerance was 0.6 Da for all MS2 spectra acquired in the ion trap. High confidence score filter (FDR < 1%) was used to select the target peptides and their corresponding MS/MS spectra were manually inspected.
Cryo-EM sample preparation and data acquisition
For cryo-EM sample preparation, 4 μL of hPink1 (1-581) -chaperone complex (0.45 mg/ml) or hPink1 (130-581) -chaperone complex (0.25 mg/ml) was loaded on a glow-discharged (Gatan Solarus Plasma Cleaning System 955) holy gold grid (R1.2/1.3, 300 mesh, X-Pivot, China) . Grids were blotted for 1 s and flash-frozen in liquid ethane cooled with liquid nitrogen using a Vitrobot Mark IV (Thermo Fisher Scientific Inc. ) , which was operated at 4 ℃ and 100 %humidity. The grids were subsequently transferred to a Titan Krios electron microscope (Thermo Fisher Scientific Inc. ) operated at 300 kV and equipped with a Gatan K3 Summit direct electron detector and a GIF Quantum energy filter. Movie stacks were automatically collected using EPU with a preset defocus ranged from -1.2 μm to -1.8 μm in super-resolution mode. Data collection was performed at a normal magnification of 130,000 with a pixel size ofon the specimen. Each movie stack containing 32 frames was exposed for 1.11 s. The slit width on the energy filter was 20 eV; and the total dose was for each micrograph stack.
Image processing
Image processing was carried out in cryoSPARC59, and the strategies were shown in Fig. S1, S8 and S10. The movie stacks were motion-corrected using MotionCor260, and the defocus values were estimated with Patch CTF estimation. Micrographs with contaminations or the maximum resolution worse thanwere excluded from calculation, resulting in a total of 18,161 micrographs (hPink11-581) and 8,111 micrographs (hPink1130-581) for structural determination.
For hPink11-581 dataset, particles were auto-picked using blob picker, extracted with a box size of 400 pixels from 500 micrographs, and classified into 50 classes by 2D classification. 19,832 particles from 500 micrographs with representative 2D class averages were selected as the training dataset for Topaz Train61. Then, 4,995,315 particles were auto-picked from 18,161 micrographs using Topaz61. After two rounds of 2D classification, 3,693,128 particles were selected and then subjected to Ab-initio Reconstruction into five classes, which were used as the templates for Heterogeneous Refinement of all selected particles. After refinement, particles were selected from the density maps with detailed features and the highest resolution. These particles were subsequently subjected to a second round of Ab-initio Reconstruction and Heterogeneous Refinement. The selected particles were refined with Non-uniform Refinement to improve the resolution of the density toAfter applying C2 symmetry and Local CTF Refinement, a map with a resolution ofwas obtained 62, 63. All reported resolutions were estimated using the gold-standard FSC 0.143 criterion. The local-resolution distributions of the 3D reconstructions were evaluated using ResMap64.
For hPink1130-581 dataset, a total of 80,034 particles were auto-picked using blob picker, extracted with a box size of 400 pixels from 500 micrographs, and classified into 50 classes by 2D classification. 1,827 particles from 100 micrographs with representative 2D class averages were selected as the training dataset for Topaz Train61. Then, 1,861,310 particles were auto-picked from 8, 111 micrographs using Topaz61. After one round of 2D classification, 1,724,632 particles were selected and split into two sets randomly, and then subjected to Ab-initio Reconstruction in four classes, which were used as the templates for Heterogeneous Refinement of all selected particles. After refinement, 658,543 particles were selected from the density maps with detailed features and the highest resolution. These particles were subsequently subjected to a second round of 2D classification to select 554, 347 particles with better quality. The selected particles were refined with CTF refinement and reconstructed with Non-uniform Refinement to improve the resolution of the density toTo separate individual conformations from these particles, we performed a 3D variability analysis in simple mode (20 low-resolution density maps for further movie creation) and cluster mode, respectively. The 407,577 particles selected from cluster 1 and cluster 2 (representing ‘prime-activation’ conformation) were subjected to Non-uniform refinement, yielding a map withresolution. The particles from cluster 3 (representing ‘autoinhibited’ conformation) were further split into 2 clusters. Non-uniform refinement of cluster 1 (75,286 particles) and cluster 2 (71, 484 particles) yielded density maps withandresolutions, respectively.
Model building and refinement
To generate initial models, the structures of HSP90 complexes and TcPINK1 (PDB ID 5YJ9 or 7MP8) were individually docked into the Cryo-EM density maps using Chimera65. The models were iteratively rebuilt in COOT and refined in Phenix66, 67. The geometry of the refined models was validated with MolProbity68. The statistics for the Cryo-EM data collection and model refinement were reported in supplemental Table 1. The refined coordinates and Cryo-EM data were deposited into PDB and EMDateBank, respectively.
Inhibition of the ATPase activity of HSP90 by geldanamycin
HEK293T cells were cultured in DMEM supplemented with 10%FBS at 37 ℃ in a CO2 incubator. Upon ~ 80%confluency, the cells in a 10-cm dish were transfected with 25 μg of the plasmid encoding the Flag-tagged full-length hPINK1 using 75 μg of PEI dissolved in PBS. After 5 hours, the transfected cells were resuspended in the fresh media. And they were evenly passed to the wells of a 24-well plate. After 36 hours, 2 μM CCCP was added into each well. Next, 10 μM geldanamycin was separately added into each of 7 groups of triplicated wells at different time points ranging from 12 hours to 1 hour. After treating the cells with CCCP for 12 hours, the cells in each well were washed with 500 μL of PBS and then were lysed with 80 μL RIPA supplemented with 1 mM PMSF and 1 mM Na3VO4. The lysed cells were frozen at -20 ℃ for over 30 minutes before they were thawed and cleared by centrifugation at 12,000 rpm for 10 min at 4℃. The supernatants were mixed with 6× SDS loading buffer for Western-Blotting analysis.
In Western-blotting assay, the samples were separated by electrophoresis on 8%SDS–PAGE gels. Then, the proteins on the gels were transferred to PVDF membranes. Next, the membranes were blocked for 1 h with 5%non-fat dry milk in TBST and individually incubated with anti-Flag or anti-phosphorylated Ubiquitin (Cell signaling Technology Cat #: 62802S, 1: 1000 dilution) antibody overnight at 4 ℃. After washing with TBST, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit (Abmart Cat #: M21002) and goat anti-mouse (Abmart Cat #: M21001) IgG antibodies, respectively, for 1 h at 4 ℃.
The Western-Blotting images were collected with LAS-4000 imager (GE Health) and the density for each band was measured with the software from the manufacture. The statistics was calculated for each time point of treatment from triplicated technical repeats.
Inhibition of the PPIase activity of FKBP51 by SAFit2
The experiments for inhibition of the PPIase activity of FKBP51 were carried out similarly as described above in the Inhibition of the ATPase activity of HSP90 except that 10 μM geldanamycin was replaced by 10 μM SAFit2.
Using confocal microscopy to characterize the recruitment of Parkin to mitochondria For analyzing the effects of geldanamycin on Parkin recruitment, the wild type and hPINK1 knock-out HeLa cells were cultured in DMEM supplemented with 10%FBS in 35 mm confocal dishes at 37 ℃ in a CO2 incubator. Just before the transfection, the culture media was replaced with fresh media. Then, the cells in each well at ~80%of confluency were transiently transfected with 2.5 μg /well of the plasmid encoding the EGFP-Parkin using 7.5 μg/well of PEI. After 4 hours, the culture media was replaced with the fresh media. After 24 hours, transfected cells were treated with none, 5 μM CCCP, 10 μM geldanamycin, and 5 μM CCCP plus 10 μM geldanamycin for 1 hour in the presence of 100 nM MitoTracker Deep Red FM (Yeasen Biotechnology) . Cells were fixed and stained with 10 μg/mL DAPI as described by the manufacturer. Images were taken with Leica SP8 confocal microscopy using a HC PL APO CS2 63x/1.40 OIL objective at 23 ℃. Three-channel images were sequentially collected with a 405 nm, a 488 nm, and a 638 nm laser to detect the signals from DAPI, EGFP-Parkin, and MitoTracker, respectively.
To analyze the recruitment of Parkin to mitochondria, the Pearson’s R values were used to quantify the co-localization of EGFP-Parkin with MitoTracker. In the analysis, the boundary of each EGFP-Parkin positive cell (green channel) was individually masked. The Pearson’s R value for the co-localization between EGFP-Parkin (green channel) and MitoTracker (red channel) in each cell was calculated using the Coloc 2 program in the ImageJ software (NIH) . The confocal images were taken from replicated transfection experiments. The distribution and statistics of the Pearson’s R values were quantified from over 70 randomly selected cells for each sample.
In analyzing the effects of SAFit2 on Parkin recruitment, similar experiments were performed as described above except that 10 μM geldanamycin was replaced by 4 μM SAFit2. The distribution and statistics of the Pearson’s R values were quantified from over 70 randomly selected cells for each sample.
Endogenous PINK1 stability assays
WT Hela cells were passed into 24-well plates and cultured in DMEM supplemented with 10%FBS at 37 ℃ in a CO2 incubator. Upon ~ 80%confluency, the cells were treated respectively with none, 10 μM CCCP, 20 μM geldanamycin, 2 μM geldanamycin, 20 μM geldanamycin plus 10 μM CCCP, 2 μM geldanamycin plus 10 μM CCCP, 2 μM SAFit2, 200 nM SAFit2, 2 μM SAFit2 plus 10 μM CCCP, and 200 nM SAFit2 plus 10 μM SAFit2 for 12 hours. After treatment, the cells were washed with 500 μL/well of PBS and then were lysed with 80 μL/well RIPA supplemented with 1 mM PMSF. The lysed cells were frozen at -20 ℃ for over 30 minutes before they were thawed and cleared by centrifugation at 12,000 rpm for 10 min at 4℃. The supernatants were mixed with 6× SDS loading buffer for Western-Blotting analysis.
In Western-blotting assay, the samples were separated by electrophoresis on 8%SDS–PAGE gels. Then, the proteins on the gels were transferred to PVDF membranes. Next, the membranes were blocked for 1 h with 5%non-fat dry milk in TBST and incubated with anti-PINK1 antibody (Cell signaling Technology Cat #: 6946, 1: 500 dilution) overnight at 4 ℃. After washing with TBST, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit (Abmart Cat #: M21002) IgG antibodies for 1 h at 4 ℃.
The Western-Blotting images were collected with LAS-4000 imager (GE Health) and the density for each band was measured with the software from the manufacture. The statistics was calculated for each treatment from triplicated technical repeats.
Quantification and statistical analysis
In Western-Blotting experiments, the images of the gels were photographed with GE LAS-4000 imager. The densities for the protein bands were quantified with the software coming from the manufacture. The means and standard deviations were determined from triplicated experiments. The statistical significance was calculated using paired T-test of the results from triplicated technical repeats (ns, no significance; *, p <0.05; **, p <0.01; ***, p <0.001; ****, p<0.0001) . The data were processed and presented with Origin.
For analyzing the confocal images, the Pearson’s R values for the co-localization between EGFP-Parkin and MitoTracker were calculated from images of randomly selected EGFP-Parkin positive cells collected from replicated transfection experiments for each sample. The statistics and distribution of these values were analyzed using GraphPad Prism. The statistical significance was determined with unpaired Student’s T test and was expressed as: *, p < 0.05; **, p < 0.005. The error bar represents the mean ± SD.
Results
The overall architecture of the HSP90α/CDC37/FKBP51/PINK1 complex
To understand the molecular basis for hPINK1 activity regulation, we transiently transfected the plasmids encoding the full-length hPINK1 or N-terminally truncated hPINK1 (130-581) into HEK293T cells and purified the hPINK1 complexes by affinity chromatography. Affinity-purified hPINK1 complex was eluted in a sharp and symmetric peak from Superose 6 increase column. This peak contains multiple protein bands on Coomassie blue staining SDS-PAGE (Figs. 1a, b, Fig 7a, b) . LC-MS/MS analysis of these bands revealed that the major components of the complex include HSP90α, CDC37, FKBP51, and PINK1 (Fig. 1c, Fig. 7c) .
The Cryo-EM structures of hPINK1 in complex with HSP90α and its co-chaperones were determined at 2.8 - (Fig. 1d, e, Figs. 7d, 8, 9) . These structures reveal two different assemblies of the hPINK1 complex: one is bound with FKBP51, the other is not (Fig. 1f, g, Fig. 7e, f) .
In FKBP51-bound hPINK1 complex, two HSP90αs in closed conformation form a pseudo-symmetric but elongated dimer (Fig. 1f, Fig. 7e) . The N-lobe of hPINK1 is unfolded with its β5 strand threading through the lumen of HSP90α dimer. The N-terminal domain of CDC37 interacts with both the C-lobe of hPINK1 and one HSP90α protomer, while the M domain of CDC37 binds to the other HSP90α protomer located on the other side of the complex. The C-lobe and CTD of hPINK1 interact with one HSP90α protomer, the FK1 domain of FKBP51, and the N-terminal domain of CDC37. As such, CDC37 orchestrates the interactions between HSP90α and hPINK1.
The α7e helix of FKBP51 is hold by the groove formed at the dimeric interface of HSP90αC-terminal domains (Fig. 1f, Fig. 7e) . The FK2 domain leaves the main body of HSP90α dimer, having no direct interactions with other part of the complex. Twisting around the linker between the FK1 and FK2 domain, FKBP51 uses its FK1 domain to grasp its client kinase, hPINK1 (Fig. 1f, Fig. 7e) . Through such interactions and assemblies, HSP90α serves as a central organizer to coordinate the interactions between FKBP51 and its client.
In the class lacking FKBP51, HSP90α, CDC37, and hPINK1 adopt nearly identical interactions and assembly as they are in the complex of HSP90α/CDC37/FKBP51/PINK1 (The RMSD of 1685 Cα atoms between these two structures is ) . (Fig. 1g, Fig. 7f)
Detailed interactions and assembly of the HSP90α/CDC37/FKBP51 complex
To analyze how CDC37 regulates the conformations and activities of HSP90, we compared the structural details around the ATP hydrolysis sites of HSP90αs. In the protomer of HSP90α(subunit D) , which binds to the N-terminus of CDC37, the V2, Y4, and W7 residues of CDC37 are engaged into a hydrophobic cavity of HSP90α, displacing the other protomer of HSP90α (subunit C) away from the dimeric interface (Fig. 2a) . Through such engagement, F134 of the subunit D is pushed inward, forming a stronger π-cation interaction with the Arg finger, R400, from the M domain of the subunit D. Moreover, R46 forms a salt-bridge contact with the catalytic residue, E47, to deprotonate the E47 (Fig. 2a) . It had been shown that the salt-bridge contact between R46 and E47 is a critical switch point for coupling the ATP hydrolysis with the global conformational changes of HSP9046.
In contrast, on the other side of HSP90α dimer, the two subunits interact with each other. The interaction between F134 and R400 as well as the interaction between R46 and E47 are impaired, leading to a different conformational state of the subunit C from its CDC37-engaged partner (Fig. 2b) .
FKBP51 binds to the C-terminal domains of HSP90α. The α7E helix from the TPR domain of FKBP51 is anchored to the groove at the dimeric interface of HSP90α C-terminal domains through hydrophobic and electrostatic interactions (Fig. 2c, Fig. 10) . The periodicity of the helix is broken at A410 to accommodate its mismatch with the symmetry of HSP90α dimer (Fig. 2c) . The C-terminal tail from the CDC37-engaged HSP90α protomer (the subunit D) forms a L-shaped clamp to escort and lock the α7E helix in place (Fig. 2c) . The conserved MEEVD motif from HSP90α forms a short helix and is embedded at the cavity of the TPR domain (Fig. 2d, Fig. 11) .
The overall conformation of FKBP51 is flexible for adapting to the recognition of different clients (Fig. 2e) . Comparing the structures of FKBP51 in FKBP51/HSP90/GR13, FKBP51/HSP90/P2347, and our FKBP51/HSP90α/CDC37/PINK1 complex, the individual TPR, FK1 and FK2 domains of the FKBP51 remain nearly identical with RMSDs of Cαs ranging fromtoBut the two linkers connecting these three domains are relatively flexible. For example, the FK1 domains are rotated by 19° relative to each other when PINK1-bound and GR-bound FKBP51 structures are superimposed on their FK2 domains. Similarly, FK2 domains rotate 7° relative to each other when their TPR domains are superimposed.
The recognition of hPINK1 by the HSP90α/CDC37/FKBP51 complex.
As many other kinase clients in association with the closed HSP90 complex, the β5 strand of PINK1 becomes disordered and embedded in a tightly-packed hydrophobic tunnel constituted by the residues from HSP90α M-domain dimer (Fig. 3a, Fig. 12a) . The alignment of known client-HSP90 structures8, 48, which have clearly resolved densities for the lumen of the HSP90 dimer, showed an extended conformation and a conserved pattern in this client fragment for HSP90 recognition (Figs. 10b, 12b) . The hydrophobic residue at the position of PINK1’s L316 is anchored at the center of the tunnel for binding HSP90α. Polar residues are located at both sides (Fig. 12b) . This configuration might create an energy barrier to prevent the sliding of the unfolded fragment in the tunnel. In addition, in PINK1 family, hydrophobic residues at the central region of this β5 strand are highly conserved, while residues at both ends are variable (Fig. 3b) .
The folding of the kinase N-lobe is regulated by the association of HSP90α with PINK1 β5 strand. In reported insect PINK1 structures49-51, the β5 strand is the central component for organizing the hydrogen bonding network of the kinase N-lobe β sheet. In addition, its neighboring β4 and β3 strands could flip over under different conformational states (Fig. 3c) 52. Therefore, retracting β5 strand will destroy the hydrogen bonding network required for the properly folding of the kinase N-lobe (Fig. 3c) . Consistently, the density for PINK1 N-lobe in its HSP90 chaperone complex was not resolvable.
Besides the lumen, the α15` in the C-domain of HSP90α becomes disordered and moves away, creating a hydrophobic cavity to accommodate the protrusion of PINK1’s β7-β8 loop (Fig. 12c) . Notably, three mutations associated with PD are in this β7 strand. And one of these mutations (I368N) had been shown to be able to lower the levels of PINK1 associated with HSP90/CDC37 complex and to impair the import of PINK1 into TOM40 complex under stress53.
The interaction of CDC37 with its clients are regulated by its phosphorylation8. The phosphorylation on Ser13 of CDC37 stabilizes the proper orientation of the CDC37 fragment (a.a. 13-34) so that this fragment could insert into a highly conserved pocket on PINK1 C-lobe and in the proximity of the β7-β8 strands (Fig. 3d) . As such, the HPNI motif on this fragment could mimic the configuration of the folded functional PINK1 αi-β4 loop to stabilize the C-lobe of the unfolded kinase (Fig. 3c, d) .
The binding of HSP90α to PINK1 alters the conformation of the kinase CTD, and thus regulates the dimeric assembly of the kinase52. The binding of HSP90α to PINK1 CTD disrupts the secondary structure of the αL helix and displaces the αK helix away (Fig. 3e, Fig. 12d) . The unfolded αL helix is associated with the C subunit of HSP90α, while the αK helix is essential for the dimeric assembly of the kinase (Fig. 3f) 52. Indeed, 5 out of 7 PD-associated mutants identified on PINK1 CTD are enriched on this αK helix52, emphasizing the functional importance of this helix in the regulation of the kinase activities. The FK1 domain of FKBP51, which has the PPIase activity, directly binds to the C-lobe of PINK1 (Fig. 3g) . Strikingly, the well-ordered PINK1 activation loop resolved in all early reported PINK1 structures becomes disappeared in electron density, indicating this loop becomes flexible in conformation (Fig. 3g) 49, 50, 52, 54. As the activation loop is in the proximity to the ligand binding site of the FK1 domain, we tempted to consider the possibility that the conformation of the activation loop is regulated by the PPIase activity of the FK1 domain. In support of this possibility, sequence alignment of PINK1 family members discloses a conserved LPF motif on the activation loop, which was proposed as a potential peptidyl inhibitor for FKBP51 in molecular docking55.
Collectively, our structures reveal how HSP90α/CDC37/FKBP51 recognize highly conserved interfaces on PINK1 to regulate its folding, conformation, and assembly (Fig. 12e) .
Inhibition of the ATPase activity of HSP90α promotes the degradation of PINK1 under stress
To validate the cellular impact of PINK1-HSP90α chaperone complex assembly on PINK1 functions, we used geldanamycin to inhibit the ATPase activity of HSP90 and studied how such inhibition affects the degradation, ubiquitin phosphorylation, and PARKIN recruitment of PINK1.
In transiently transfected HEK293T cells, adding 10 μM geldanamycin promoted the degradation of PINK1 under mitochondrial stress induced by 2 μM CCCP in a time-dependent manner. Compared with only CCCP treated sample, adding geldanamycin for 12 hours enhanced the degradation of PINK1 by over 80 % (Fig. 4a) . In accompany with PINK1 degradation, however, the level of phosphorylated poly-ubiquitin displayed a trend of increase but was not statistically significant (Fig. 4b) .
The treatment of 10 μM geldanamycin could down-regulate the recruitment of transfected EGFP-PARKIN to mitochondria under the stimulation of 5 μM CCCP in wild type Hela cells (Fig. 4c, d, e) . In comparison, either adding geldanamycin by itself or knocking out PINK1 in Hela cells could abolish such down-regulatory effect as visualized in confocal microscopy images and analyzed by the statistics of Pearson`s colocalization coefficients from randomly selected cells (Fig. 4c, d, e) .
Specific inhibition of the PPIase activity of FKBP51 promotes PINK1-mediated PARKIN recruitment in mitophagy
Then, we used a specific inhibitor for the PPIase of FKBP51, the SAFit2, to study the functions of FKBP51 on PINK1 regulation.
Unlike geldanamycin, 10 μM SAFit2 treatment had little time-dependent effect on the degradation of PINK1 (Fig. 5a) or phosphorylation of poly-ubiquitin (Fig. 5b) in PINK1-transfected HEK293T cells stimulated with 2μM CCCP. However, adding 4 μM SAFit2 could enhance CCCP-induced recruitment of EGFP-PARKIN to mitochondria in wild type Hela cells (Fig. 5c, d, e) . Such effect is abolished in PINK1 knock-out Hela cells, indicating SAFit2-enhanced PARKIN recruitment in response to mitochondrial stress is PINK1 dependent. Consistently, co-admission of SAFit2 could enhance the expression and/or stabilization of the full-length, endogenous PINK1 in WT Hela cells stimulated with 10 μM CCCP overnight (Fig. 5f) .
Discussion
The Cryo-EM structures determined in our study provided molecular details for understanding the recognition of PINK1 by HSP90α/CDC37/FKBP51 chaperone complex. The HSP90α/CDC37/FKBP51 complex recognizes various highly-conserved interfaces on PINK1 to regulate its folding, conformation, and dimeric assembly, which in turn ultimately control the quality and autophagy of mitochondria. In addition, we found that selective inhibition of the PPIase activity of FKBP51 by SAFit2 could promote PINK1-mediated PARKIN recruitment in response to mitochondrial membrane depolarization.
Our study established a connection from chaperone-mediated PINK1 folding and degradation to PINK1 kinase activity regulation under mitochondrial stress (Fig. 6) . Under normal physiological conditions, PINK1 synthesized in the cytoplasm needs to be unfolded for transportation into mitochondria via TOM complex29, 56. Misfolded PINK1 and some of those retro-translocated back to the outer membrane of mitochondria are degraded so that the level and activity of PINK1 are kept at a lower steady state to protect the healthy mitochondria30, 31. But upon mitochondrial damage, the transportation of PINK1 via TOM complex is stalled, and thus the full-length PINK1 is accumulated and activated on the outer membrane of mitochondria to elicit the recruitment of PARKIN and clearance of damaged mitochondria29, 34. In these processes, HSP90/CDC37 complex had been suggested to be a key regulator for PINK1 stability and degradation53, 57. But it remains not fully defined how HSP90/CDC37 complex interacts with PINK1 and whether other components are required for the assembly and regulation of this PINK1-chaperone complex. From our cryo-EM structures and mass-spec analysis, we identified that FKBP51 is one of the co-chaperones involved in the regulation of PINK1 activity.
In consistence with our structures, using geldanamycin to disrupt the association of HSP90 with TPR-containing proteins, including FKBP51, TOM70 and CHIP, promotes the degradation of overexpressed PINK1 in response to CCCP treatment in a time-dependent manner. Moreover, specific inhibition of the PPIase activity of FKBP51 enhanced PINK1-mediated PARKIN recruitment under CCCP-induced mitochondria stress. These data in together support the functional linkage for HSP90/CDC37/FKBP51-mediated PINK1 unfolding and stabilization. Through such a linkage, the unfolding and stabilization of PINK1 are tightly coupled with PINK1 activation upon mitochondrial damage. As such, the cells could adapt properly to environmental stress for protecting the fidelity of mitochondria or clearing damaged organelles for the health of the cell.
The recognition of PINK1 by HSP90/CDC37/FKBP51 complex is highly conserved and adaptive. HSP90/CDC37/FKBP51 binds to three important but highly conserved structural elements of PINK1: the β5 strand, the activation loop, and the CTD. The β5 strand is essential for coordinating the hydrogen-bonding network of the functional kinase N-lobe β-sheet49, 50, 54. Clamping this strand in the lumen of HSP90 dimer would disrupt such hydrogen-bonding network and stabilize the kinase in an inactive, unfolded state, which is required for the import of PINK1 into mitochondria via TOM complex or degradation of PINK1 via proteasome56.
The CTD of PINK1 is involved in the regulation of PINK1 kinase activities52. The binding of HSP90 to the αL helix dislodges the αK helix from its association with the αL helix and impairs the dimeric assembly and activation of the kinase52. Intriguingly, many PD-associated mutations are enriched in the αK helix (Fig. 10b) .
In all early reported PINK1 structures, the activation loop is well-ordered and extended, representing an active conformation being ready for substrate binding49-51. It remains puzzling how the conformation of PINK1 activation loop is regulated. In our cryo-EM structures, the density for the activation loop becomes disappeared and is positioned to the proximity to the ligand-binding site of the PPIase of FKBP51. The sequence of PINK1 activation loop contains a conserved LPF motif, which was proposed as a potential FKBP51 inhibitor55. Using a selective inhibitor for FKBP51, we further validated the requirement of the PPIase activity of FKBP51 in the regulation of PINK1 protein levels and activities under stress. These data do not only provide insights for understanding the regulation of PINK1 activities, but also have broad impact for understanding the client-recognition by HSP90/CDC37/FKBP51 complex.
Development of targeted therapeutics for the treatment of PD becomes increasingly demanded in modern society. Early strategy in developing ATP analogs to boost the kinase activity of PINK1 was under debate44, 45. Therefore, it is warranted to develop alternative strategies for the treatment of PD associated with PINK1 deficiency. In our study, we identified FKBP51 is a potential target for enhancing the stability of PINK1 as well as PINK1-mediated PARKIN recruitment under stress. Indeed, it had been shown that SAFit2 has clinical potentials in the treatment of neuroinflammation, depression, neuropathic pain, and Huntington disease14, 58. Therefore, our study revealed an alternative avenue for the treatment of PINK1-deficient PD.
In summary, our study laid a foundation for understanding the structure, recognition, and regulation of PINK1 by the HSP90/CDC37/FKBP51 complex. In addition, we provided insights for developing alternative strategies in the treatment of PD.
Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it is apparent to those skilled in the art that certain minor changes and modifications will be practiced. Therefore, the description and examples should not be construed as limiting the scope of the invention.
The disclosures of all publications, patents, patent applications and published patent applications referred to herein are hereby incorporated herein by reference in their entirety.
References
1 Schopf, F.H., Biebl, M.M. &Buchner, J. The HSP90 chaperone machinery. Nat Rev Mol Cell Biol 18, 345-360, doi: 10.1038/nrm. 2017.20 (2017) .
2 Soto, C. &Pritzkow, S. Protein misfolding, aggregation, and conformational strains in neurodegenerative diseases. Nat Neurosci 21, 1332-1340, doi: 10.1038/s41593-018-0235-9 (2018) .
3 Zhao, R. et al. Navigating the chaperone network: an integrative map of physical and genetic interactions mediated by the hsp90 chaperone. Cell 120, 715-727, doi: 10.1016/j. cell. 2004.12.024 (2005) .
4 Prodromou, C. et al. The ATPase cycle of Hsp90 drives a molecular ‘clamp’ via transient dimerization of the N‐terminal domains. The EMBO Journal 19, 4383-4392-4392, doi: https: //doi. org/10.1093/emboj/19.16.4383 (2000) .
5 Prodromou, C. et al. Identification and structural characterization of the ATP/ADP-binding site in the Hsp90 molecular chaperone. Cell 90, 65-75, doi: 10.1016/s0092-8674 (00) 80314-1 (1997) .
6 Roe, S. M. et al. The Mechanism of Hsp90 regulation by the protein kinase-specific cochaperone p50 (cdc37) . Cell 116, 87-98, doi: 10.1016/s0092-8674 (03) 01027-4 (2004) .
7 Keramisanou, D., Vasantha Kumar, M. V., Boose, N., Abzalimov, R. R. &Gelis, I. Assembly mechanism of early Hsp90-Cdc37-kinase complexes. Sci Adv 8, eabm9294, doi: 10.1126/sciadv. abm9294 (2022) .
8 Verba, K.A. et al. Atomic structure of Hsp90-Cdc37-Cdk4 reveals that Hsp90 traps and stabilizes an unfolded kinase. Science 352, 1542-1547, doi: 10.1126/science. aaf5023 (2016) .
9 Xu, W. et al. Dynamic Tyrosine Phosphorylation Modulates Cycling of the HSP90-P50CDC37-AHA1 Chaperone Machine. Molecular Cell 47, 434-443, doi: https: //doi. org/10.1016/j. molcel. 2012.05.015 (2012) .
10 Siligardi, G. et al. Regulation of Hsp90 ATPase activity by the co-chaperone Cdc37p/p50cdc37. J Biol Chem 277, 20151-20159, doi: 10.1074/jbc. M201287200 (2002) .
11 Noddings, C. M., Wang, R. Y., Johnson, J. L. &Agard, D. A. Structure of Hsp90-p23-GR reveals the Hsp90 client-remodelling mechanism. Nature 601, 465-469, doi: 10.1038/s41586-021-04236-1 (2022) .
12 Wang, R. Y. et al. Structure of Hsp90-Hsp70-Hop-GR reveals the Hsp90 client-loading mechanism. Nature 601, 460-464, doi: 10.1038/s41586-021-04252-1 (2022) .
13 Noddings, C. M., Johnson, J. L. &Agard, D. A. Cryo-EM reveals how Hsp90 and FKBP immunophilins co-regulate the glucocorticoid receptor. Nat Struct Mol Biol 30, 1867-1877, doi: 10.1038/s41594-023-01128-y (2023) .
14 Hahle, A., Merz, S., Meyners, C. &Hausch, F. The Many Faces of FKBP51. Biomolecules 9, doi: 10.3390/biom9010035 (2019) .
15 Habara, M. et al. FKBP52 and FKBP51 differentially regulate the stability of estrogen receptor in breast cancer. Proc Natl Acad Sci U S A 119, e2110256119, doi: 10.1073/pnas. 2110256119 (2022) .
16 Ko, H. S. et al. CHIP regulates leucine-rich repeat kinase-2 ubiquitination, degradation, and toxicity. Proc Natl Acad Sci U S A 106, 2897-2902, doi: 10.1073/pnas. 0810123106 (2009) .
17 Oberoi, J. et al. HSP90-CDC37-PP5 forms a structural platform for kinase dephosphorylation. Nat Commun 13, 7343, doi: 10.1038/s41467-022-35143-2 (2022) .
18 Jaime-Garza, M. et al. Hsp90 provides a platform for kinase dephosphorylation by PP5. Nat Commun 14, 2197, doi: 10.1038/s41467-023-37659-7 (2023) .
19 Silverstein, A.M., Grammatikakis, N., Cochran, B.H., Chinkers, M. &Pratt, W.B. p50 (cdc37) binds directly to the catalytic domain of Raf as well as to a site on hsp90 that is topologically adjacent to the tetratricopeptide repeat binding site. J Biol Chem 273, 20090-20095, doi: 10.1074/jbc. 273.32.20090 (1998) .
20 Vaughan, C. K. et al. Structure of an Hsp90-Cdc37-Cdk4 complex. Mol Cell 23, 697-707, doi: 10.1016/j. molcel. 2006.07.016 (2006) .
21 Wochnik, G. M. et al. FK506-binding proteins 51 and 52 differentially regulate dynein interaction and nuclear translocation of the glucocorticoid receptor in mammalian cells. J Biol Chem 280, 4609-4616, doi: 10.1074/jbc. M407498200 (2005) .
22 Zgajnar, N. R. et al. Biological Actions of the Hsp90-binding Immunophilins FKBP51 and FKBP52. Biomolecules 9, doi: 10.3390/biom9020052 (2019) .
23 Storer, C. L., Dickey, C. A., Galigniana, M. D., Rein, T. &Cox, M. B. FKBP51 and FKBP52 in signaling and disease. Trends Endocrinol Metab 22, 481-490, doi: 10.1016/j. tem. 2011.08.001 (2011) .
24 Baischew, A., Engel, S., Taubert, M.C., Geiger, T.M. &Hausch, F. Large-scale, in-cell photocrosslinking at single-residue resolution reveals the molecular basis for glucocorticoid receptor regulation by immunophilins. Nat Struct Mol Biol 30, 1857-1866, doi: 10.1038/s41594-023-01098-1 (2023) .
25 Narendra, D. P. et al. PINK1 is selectively stabilized on impaired mitochondria to activate Parkin. PLoS Biol 8, e1000298, doi: 10.1371/journal. pbio. 1000298 (2010) .
26 Vives-Bauza, C. et al. PINK1-dependent recruitment of Parkin to mitochondria in mitophagy. Proc Natl Acad Sci U S A 107, 378-383, doi: 10.1073/pnas. 0911187107 (2010) .
27 Wang, X. et al. PINK1 and Parkin target Miro for phosphorylation and degradation to arrest mitochondrial motility. Cell 147, 893-906, doi: 10.1016/j. cell. 2011.10.018 (2011) .
28 Gandhi, S. et al. PINK1-associated Parkinson's disease is caused by neuronal vulnerability to calcium-induced cell death. Mol Cell 33, 627-638, doi: 10.1016/j. molcel. 2009.02.013 (2009) .
29 Lazarou, M., Jin, S. M., Kane, L. A. &Youle, R. J. Role of PINK1 binding to the TOM complex and alternate intracellular membranes in recruitment and activation of the E3 ligase Parkin. Dev Cell 22, 320-333, doi: 10.1016/j. devcel. 2011.12.014 (2012) .
30 Sekine, S. et al. Reciprocal Roles of Tom7 and OMA1 during Mitochondrial Import and Activation of PINK1. Mol Cell 73, 1028-1043 e1025, doi: 10.1016/j. molcel. 2019.01.002 (2019) .
31 Jin, S.M. et al. Mitochondrial membrane potential regulates PINK1 import and proteolytic destabilization by PARL. J Cell Biol 191, 933-942, doi: 10.1083/jcb. 201008084 (2010) .
32 Greene, A.W. et al. Mitochondrial processing peptidase regulates PINK1 processing, import and Parkin recruitment. Embo Reports 13, 378 (2012) .
33 Yamano, K. &Youle, R.J. PINK1 is degraded through the N-end rule pathway. Autophagy 9, 1758-1769, doi: 10.4161/auto. 24633 (2013) .
34 Okatsu, K. et al. A dimeric PINK1-containing complex on depolarized mitochondria stimulates Parkin recruitment. J Biol Chem 288, 36372-36384, doi: 10.1074/jbc. M113.509653 (2013) .
35 Matsuda, N. et al. PINK1 stabilized by mitochondrial depolarization recruits Parkin to damaged mitochondria and activates latent Parkin for mitophagy. J Cell Biol 189, 211-221, doi: 10.1083/jcb. 200910140 (2010) .
36 Kane, L.A. et al. PINK1 phosphorylates ubiquitin to activate Parkin E3 ubiquitin ligase activity. J Cell Biol 205, 143-153, doi: 10.1083/jcb. 201402104 (2014) .
37 Koyano, F. et al. Ubiquitin is phosphorylated by PINK1 to activate parkin. Nature 510, 162-166, doi: 10.1038/nature13392 (2014) .
38 Trempe, J.F. et al. Structure of parkin reveals mechanisms for ubiquitin ligase activation. Science 340, 1451-1455, doi: 10.1126/science. 1237908 (2013) .
39 Valente, E.M. et al. Hereditary early-onset Parkinson's disease caused by mutations in PINK1. Science 304, 1158-1160, doi: 10.1126/science. 1096284 (2004) .
40 Yang, Y. et al. Mitochondrial pathology and muscle and dopaminergic neuron degeneration caused by inactivation of Drosophila Pink1 is rescued by Parkin. Proc Natl Acad Sci U S A 103, 10793-10798, doi: 10.1073/pnas. 0602493103 (2006) .
41 Park, J. et al. Mitochondrial dysfunction in Drosophila PINK1 mutants is complemented by parkin. Nature 441, 1157-1161, doi: 10.1038/nature04788 (2006) .
42 Petit, A. et al. Wild-type PINK1 prevents basal and induced neuronal apoptosis, a protective effect abrogated by Parkinson disease-related mutations. Journal of Biological Chemistry 280, 34025-34032 (2005) .
43 Hertz, N.T. et al. A neo-substrate that amplifies catalytic activity of parkinson's -disease-related kinase PINK1. Cell 154, 737-747, doi: 10.1016/j. cell. 2013.07.030 (2013) .
44 Orr, A.L. et al. Long-term oral kinetin does not protect against alpha-synuclein-induced neurodegeneration in rodent models of Parkinson's disease. Neurochem Int 109, 106-116, doi: 10.1016/j. neuint. 2017.04.006 (2017) .
45 Chin, R.M. et al. Pharmacological PINK1 activation ameliorates Pathology in Parkinson's Disease models. bioRxiv, doi: 10.1101/2023.02.14.528378 (2023) .
46 Mader, S.L. et al. Conformational dynamics modulate the catalytic activity of the molecular chaperone Hsp90. Nat Commun 11, 1410, doi: 10.1038/s41467-020-15050-0 (2020) .
47 Lee, K. et al. The structure of an Hsp90-immunophilin complex reveals cochaperone recognition of the client maturation state. Mol Cell 81, 3496-3508 e3495, doi: 10.1016/j. molcel. 2021.07.023 (2021) .
48 Garcia-Alonso, S. et al. Structure of the RAF1-HSP90-CDC37 complex reveals the basis of RAF1 regulation. Mol Cell 82, 3438-3452 e3438, doi: 10.1016/j. molcel. 2022.08.012 (2022) .
49 Gan, Z.Y. et al. Activation mechanism of PINK1. Nature 602, 328-335, doi: 10.1038/s41586-021-04340-2 (2022) .
50 Kumar, A. et al. Structure of PINK1 and mechanisms of Parkinson's disease-associated mutations. Elife 6, doi: 10.7554/eLife. 29985 (2017) .
51 Rasool, S. et al. Mechanism of PINK1 activation by autophosphorylation and insights into assembly on the TOM complex. Mol Cell 82, 44-59 e46, doi: 10.1016/j. molcel. 2021.11.012 (2022) .
52 Xu, H., Liu, X., Jiang, Y., Zhang, Y., Xue, J., Wang, S., Chen, S., Chen C., Yu, X., Li, B., Sui, S.F., Qin, X., Mi, L.Z., Liu Z. Dynamic conformations and assembies of PINK1 kinases. unpublished
53 Ando, M. et al. The PINK1 p. I368N mutation affects protein stability and ubiquitin kinase activity. Mol Neurodegener 12, 32, doi: 10.1186/s13024-017-0174-z (2017) .
54 Rasool, S. &Trempe, J. F. New insights into the structure of PINK1 and the mechanism of ubiquitin phosphorylation. Crit Rev Biochem Mol Biol 53, 515-534, doi: 10.1080/10409238.2018.1491525 (2018) .
55 Han, J.T. et al. Discovery of pentapeptide-inhibitor hits targeting FKBP51 by combining computational modeling and X-ray crystallography. Comput Struct Biotechnol J 19, 4079-4091, doi: 10.1016/j. csbj. 2021.07.015 (2021) .
56 Sekine, S. &Youle, R.J. PINK1 import regulation; a fine system to convey mitochondrial stress to the cytosol. BMC Biol 16, 2, doi: 10.1186/s12915-017-0470-7 (2018) .
57 Moriwaki, Y. et al. L347P PINK1 mutant that fails to bind to Hsp90/Cdc37 chaperones is rapidly degraded in a proteasome-dependent manner. Neurosci Res 61, 43-48, doi: 10.1016/j. neures. 2008.01.006 (2008) .
58 Bailus, B.J. et al. Modulating FKBP5/FKBP51 and autophagy lowers HTT (huntingtin) levels. Autophagy 17, 4119-4140, doi: 10.1080/15548627.2021.1904489 (2021) .
59 Punjani, A., Rubinstein, J.L., Fleet, D.J. &Brubaker, M.A. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat Methods 14, 290-296, doi: 10.1038/nmeth. 4169 (2017) .
60 Zheng, S. Q. et al. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat Methods 14, 331-332, doi: 10.1038/nmeth. 4193 (2017) .
61 Bepler, T. et al. Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs. Nat Methods 16, 1153-1160, doi: 10.1038/s41592-019-0575-8 (2019) .
62 Punjani, A., Zhang, H. &Fleet, D.J. Non-uniform refinement: adaptive regularization improves single-particle cryo-EM reconstruction. Nat Methods 17, 1214-1221, doi: 10.1038/s41592-020-00990-8 (2020) .
63 Zivanov, J., Nakane, T. &Scheres, S.H.W. Estimation of high-order aberrations and anisotropic magnification from cryo-EM data sets in RELION-3.1. IUCrJ 7, 253-267, doi: 10.1107/S2052252520000081 (2020) .
64 Kucukelbir, A., Sigworth, F.J. &Tagare, H.D. Quantifying the local resolution of cryo-EM density maps. Nat Methods 11, 63-65, doi: 10.1038/nmeth. 2727 (2014) .
65 Pettersen, E.F. et al. UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25, 1605-1612, doi: 10.1002/jcc. 20084 (2004) .
66 Emsley, P., Lohkamp, B., Scott, W.G. &Cowtan, K. Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66, 486-501, doi: 10.1107/S0907444910007493 (2010) .
67 Liebschner, D. et al. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol 75, 861-877, doi: 10.1107/S2059798319011471 (2019) .
68 Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr D Biol Crystallogr 66, 12-21, doi: 10.1107/S0907444909042073 (2010) .

Claims (10)

  1. Use of a FKBP51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities.
  2. The use according to claim 1, wherein the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof.
  3. The use according to claim 1 or 2, wherein the disease is Parkinson disease.
  4. The use according to claim 3, wherein the Parkinson disease is PINK1-dificient Parkinson disease.
  5. The use according to claim 1 or 2, wherein the disease is cardiomyositis, coronary heart disease or liver fibrosis.
  6. A method for treating or preventing diseases caused by mitochondrial abnormalities, comprising administering to a patient in need thereof a FKBP51 inhibitor.
  7. The method according to claim 6, wherein the FKBP51 inhibitor is SAFit2 or a pharmaceutically acceptable salt thereof.
  8. The method according to claim 6 or 7, wherein the disease is Parkinson disease.
  9. The method according to claim 8, wherein the Parkinson disease is PINK1-dificient Parkinson disease.
  10. The method according to claim 6 or 7, wherein the disease is cardiomyositis, coronary heart disease or liver fibrosis.
PCT/CN2025/085699 2024-03-29 2025-03-28 Use of a fkbp51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities Pending WO2025201510A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CNPCT/CN2024/085021 2024-03-29
CN2024085021 2024-03-29

Publications (1)

Publication Number Publication Date
WO2025201510A1 true WO2025201510A1 (en) 2025-10-02

Family

ID=97218981

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2025/085699 Pending WO2025201510A1 (en) 2024-03-29 2025-03-28 Use of a fkbp51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities

Country Status (1)

Country Link
WO (1) WO2025201510A1 (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015039758A1 (en) * 2013-09-19 2015-03-26 MAX-PLANCK-Gesellschaft zur Förderung der Wissenschaften e.V. Selective fkbp51 ligands for treatment of psychiatric disorders
WO2022049005A1 (en) * 2020-09-04 2022-03-10 Technische Universität Darmstadt High affinity macrocyclic fkb51-inhibitors for treatment of psychiatric disorders

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015039758A1 (en) * 2013-09-19 2015-03-26 MAX-PLANCK-Gesellschaft zur Förderung der Wissenschaften e.V. Selective fkbp51 ligands for treatment of psychiatric disorders
WO2022049005A1 (en) * 2020-09-04 2022-03-10 Technische Universität Darmstadt High affinity macrocyclic fkb51-inhibitors for treatment of psychiatric disorders

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
BUFFA VANESSA, KNAUP FABIAN H., HEYMANN TIM, SPRINGER MARGHERITA, SCHMIDT MATHIAS V., HAUSCH FELIX: "Analysis of the Selective Antagonist SAFit2 as a Chemical Probe for the FK506-Binding Protein 51", ACS PHARMACOLOGY & TRANSLATIONAL SCIENCE, vol. 6, no. 3, 10 March 2023 (2023-03-10), US, pages 361 - 371, XP093359036, ISSN: 2575-9108, DOI: 10.1021/acsptsci.2c00234 *
CRUZ BRYAN, VOZELLA VALENTINA, CARPER BENJAMIN A., XU JOY C., KIRSON DEAN, HIRSCH SHAWN, NOLEN TRACY, BRADLEY LAUREN, FAIN KATIE, : "FKBP5 inhibitors modulate alcohol drinking and trauma-related behaviors in a model of comorbid post-traumatic stress and alcohol use disorder", NEUROPSYCHOPHARMACOLOGY, vol. 48, no. 8, 1 July 2023 (2023-07-01), pages 1144 - 1154, XP093359041, ISSN: 0893-133X, DOI: 10.1038/s41386-022-01497-w *
QIU BIN, ZHONG ZHAOHUI, DOU LONGYU, XU YUXUE, ZOU YI, WELDON KORRI, WANG JUN, ZHANG LINGLING, LIU MING, WILLIAMS KENT E., SPENCE J: "Elimination of FKBP51 attenuates CCl4-induced liver injury via enhancement of mitochondrial function by increased Parkin activity", RESEARCH SQUARE, 16 March 2023 (2023-03-16), pages 1 - 22, XP093359052, Retrieved from the Internet <URL:https://www.researchsquare.com/article/rs-2679397/v1> DOI: 10.21203/rs.3.rs-2679397/v1 *

Similar Documents

Publication Publication Date Title
Li et al. Template-assisted covalent modification underlies activity of covalent molecular glues
Walter et al. Cryo-EM structures and functional characterization of murine Slc26a9 reveal mechanism of uncoupled chloride transport
Zhao et al. Cryo-EM structures of apo and antagonist-bound human Cav3. 1
Chang et al. Bidirectional control of autophagy by BECN1 BARA domain dynamics
Greco et al. Nuclear import of histone deacetylase 5 by requisite nuclear localization signal phosphorylation
Saleh et al. The structural basis of Cdc7-Dbf4 kinase dependent targeting and phosphorylation of the MCM2-7 double hexamer
Li et al. Template-assisted covalent modification of DCAF16 underlies activity of BRD4 molecular glue degraders
Bacic et al. Structure and dynamics of the chromatin remodeler ALC1 bound to a PARylated nucleosome
Lees et al. Applications of Cryo-EM in small molecule and biologics drug design
Yang et al. GSK1702934A and M085 directly activate TRPC6 via a mechanism of stimulating the extracellular cavity formed by the pore helix and transmembrane helix S6
Hu et al. Dynamic molecular architecture and substrate recruitment of cullin3–RING E3 ligase CRL3KBTBD2
Fan et al. Structural pharmacology of TRPV4 antagonists
Thebault et al. Structural and functional insights into the role of the N-terminal Mps1 TPR domain in the SAC (spindle assembly checkpoint)
Adriaenssens et al. Reconstitution of BNIP3/NIX-mitophagy initiation reveals hierarchical flexibility of the autophagy machinery
Liang et al. Formation of I2+ III2 supercomplex rescues respiratory chain defects
Wei et al. Selective inhibition reveals the regulatory function of DYRK2 in protein synthesis and calcium entry
Nörpel et al. Structure of the human C9orf72-SMCR8 complex reveals a multivalent protein interaction architecture
Yu et al. Cryo-EM structure of human HCN3 channel and its regulation by cAMP
Gamerdinger et al. Mechanism of cotranslational protein N-myristoylation in human cells
Chan et al. Cysteine allostery and autoinhibition govern human STING oligomer functionality
WO2025201510A1 (en) Use of a fkbp51 inhibitor in the manufacture of a medicament for treating or preventing diseases caused by mitochondrial abnormalities
Puckerin et al. Similar molecular determinants on Rem mediate two distinct modes of inhibition of CaV1. 2 channels
De Jesús-Pérez et al. Structural mechanism of TRPV5 inhibition by econazole
Li et al. Structural insights into human EMC and its interaction with VDAC
Zhou et al. Two types of peptides derived from the neurotoxin GsMTx4 inhibit a mechanosensitive potassium channel by modifying the mechanogate

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 25776125

Country of ref document: EP

Kind code of ref document: A1