Identification of Kinase inhibitors as PINK1 activators for the treatment of Parkinson’s Disease

doi:10.21203/rs.3.rs-2754116/v1

Download PDF

Article

Identification of Kinase inhibitors as PINK1 activators for the treatment of Parkinson’s Disease

https://doi.org/10.21203/rs.3.rs-2754116/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Mutations in PINK1 kinase are known to be responsible for the early-onset of Parkinson's disease. The protein resides on the outer-membrane of healthy mitochondria, and acts as a quality controller by regulating and maintaining dysfunctional mitochondria also known as mitophagy. Limited knowledge of the mechanism of protein has made it cumbersome to elucidate an effective treatment. In this paper, we screen a set of kinase inhibitors using high-throughput screening for investigating potential activators. We have for the first time shown the plausible binding and selectivity towards the kinase inhibitors in the mutant PINK1 binding site. We also could highlight the ligand bind/unbinding pathway identifying the most stable binding pose utilizing the residues from the P-loop and DFG motif for drugs Tepotinib and Pyrotinib. Interestingly, no direct interaction of mutated residues with the ligand molecules were observed. Furthermore, we highlight that the Ins3 region and SER228 of the N-lobe display distinct conformational states that are inline with crystal structure for substrate binding specificity. Collectively, our findings provide a molecular basis for the PINK1 potential kinase activator, which would further facilitate as a starting point to probe new therapeutics by using the structure based drug designing for treating neurological disorder.

Biological sciences/Computational biology and bioinformatics

Biological sciences/Drug discovery

Parkinson’s disease (PD) is a common progressive neurodegenerative disorder affecting 1–2 per 1000 of the population worldwide. Currently there are no treatments that can completely stop or slow down the progression of disease. PD is clinically characterized by the degeneration of brain cells and leading to loss of motor controls¹. Although, the majority of PD cases are sporadic, occurring in people aged 60 and above while in some cases detected in early 50’s referred as early-onset PD². EOPD is an under-recognized condition that affects around 5%-10% of PD cases and often carries a more aggressive progression rate, with symptoms developing more rapidly and severely than in typical PD patients. Furthermore, individuals with EOPD may experience an atypical set of motor symptoms not commonly seen in traditional PD patients, such as postural instability and gait disturbance. The actual onset of the disease is still not clear, but cell biologists have been able to explore the involvement of mutations in 20 different PARK genes responsible for progression of disease. The impairment of PARK genes has been suggested to have directly linked to mitochondrial dysfunctioning^3–5. However, the molecular basis of dysfunctioning of the PARK gene linked to PD is still unknown.

PARK6 gene that encodes the enzyme PINK1 has the ability to maintain mitochondrial health by a process known as mitophagy (removal of damaged mitochondria)^6–8. In the PINK1-mediated mitophagy signaling pathway, as shown in Fig. 1(a) upon mitochondrial damage and loss of mitochondrial potential leads to stabilization of the PINK1 protein on the outer membrane^9,10. PINK1 kinase phosphorylates and gets activated due to mitochondrial depolarization leading to binding and phosphorylation of Ubiquitin(Ub), which inturn interacts with Parkin leading to conformational changes and phosphorylation¹¹. The activated Parkin further promotes ubiquitination of outer mitochondrial membrane proteins. This initiates a series of downstream signaling events including the involvement of autophagy receptors and autophagosomes^12,13. PINK1 and Parkin are frequently autosomal recessive mutations^14–16, and observed in patients with early onset of Parkinson’s disease¹⁷. Therefore, PINK1 can be considered as a prime drug target and upon activation would maintain the mitochondrial potential and initiate mitophagy.

PINK1(PTEN Induced Serine/threonine Kinase-1) is one of the most divergent serine/threonine human kinase proteins. The 581 amino acid polypeptide consists of N-terminal with mitochondrial targeting signal (1–60), unconserved transmembrane region (80–110), conserved Kinase domain and conserved C-terminal region (CTR) as shown in Fig. 1(a)¹⁸. Unlike other protein kinases, transmembrane domain anchors it to the outer mitochondrial membrane, where it can sense and respond to mitochondrial stress. The kinase domain is further divided into generic N-lobe and C-lobe, having Catalytic and ATP binding sites concealed in the cleft between N and C lobes as highlighted in Fig. 1(b) and (c)¹⁹. The N-lobe majorly consists of beta-sheets and is involved in the binding of ATP. Additionally, several residues around the enzyme's active site have been identified as essential for substrate recognition, suggesting ways in which they may be regulated by other factors during cell signaling pathways. The PINK1 are unique from other kinases, due to the presence of 3 insertion regions (Ins1, Ins2 and Ins3) in the N-lobe and CTR as shown in Fig. 1(c). Majority of CTE is alpha-helical in nature, forming tightly packed E, G and H alpha-helices in the C-lobe. Although these regions are important in regulating the PINK1 activity and substrate binding, their precise function remains less explored. Structural and biochemical data of human orthologues of PINK1 (TcPink1 and PhPINK1) have highlighted the role of catalytic kinase domain mediating the activation of the protein as the first step of mitophagy cascade^20,21. Interestingly, the crystal structure of human PINK1 has not yet been identified, making it more intriguing to understand the kinase activity. Interestingly, in insect ortholog deletion of residues in Ins3 regions resulted in complete loss of substrate recognition and binding causing kinase inactivation¹⁹. Most of the mutations are present in the kinase domain, having a large impact on structure and activity of the protein. In addition, a very recent paper highlighted the variant G411S diminishes the Ub kinase activity leading to substrate receptivity^22,23. Frequently observed mutations in early PD patients are G309D and L347P^7,15, which reduces interactions with Parkin in turn reducing ubiquitination and protein-degradation^7,24. Point Mutation at residue G309 (dbSNP = rs74315355) and at L347 (dbSNP = rs28940285) resides in Ins3 and C-lobe respectively. Drosophila models of mutated PINK1 have verified direct correlation of inactive kinase in progression of neurological disorders in comparison to wildtype²⁵. Hence, activation of PINK1 kinase protein may be a key target in the treatment of early onset of PD.

Until now, efforts have been taken to screen small molecules^26–28such as Niclosamide which is an approved anthelmintic drug used in treating tapeworm infections²⁹. An alternate strategy was to use neosubstrate kinetin triphosphate which is an ATP analogue, shown to increase the activity of mutant G309D and wildtype. The neosubstrates and its derivatives showed better catalytic efficiency than its endogenous substrate ATP³⁰. Despite having activation properties on PINK1 protein these molecules^{27,28, 31–33} are shown to have irritation and environmental toxicity. Limited knowledge on the molecular mechanism of PINK1 protein and small molecules targeting the protein has made the problem statement more intriguing. Over the years the crystal structures have highlighted kinase small molecule inhibitors bound to ATP sites have shown to be controlling the conformational specificity of kinases³⁴. These inhibitors are structurally diverse and are known to be highly selective in mode of action. Apparently, some studies have paradoxically shown kinase inhibitors can act as agonists activating the protein. One such example, kinase inhibitor binding to AKT induced hyperphosphorylation, largely by rearrangement of PH domain to initiate membrane localization and increase binding affinity towards PIP3 in the membrane³⁴. The interplay whether the kinase small molecule behaves as agonist or antagonist hasn't been explored for the PINK1 kinase protein.

Currently there are approximately 300 kinase molecules FDA approved and or are in clinical trials(as reported in PKID database)³⁵, none of which has been tested against PINK1 protein. Our study highlights the testable case of these small molecules as potential activators. We chose to study mutants G309D and L347P that are predominantly associated with parkinsonism, reported to be stable but lack activity that hinders the binding of substrates. Using high-throughput screening by Molecular docking, Molecular dynamics simulations and Free energy calculations we explore the protein-ligand interactions and energetically favorable binding poses of the potential molecules. We also highlighted the prime role of N-lobe structural elements mediating the interactions with the ligands. Due to lack of studies on PINK1 mutants and their ligands, our results will provide new insights into designing development and screening testable molecules which can be further validated experimentally.

In this work, we have combined high throughput screening of molecular dynamics simulations and enhanced sampling techniques to elucidate the mechanism underlying the binding of potential activators to PINK1 protein. Till date no structure has been elucidated for human PINK1 protein, making it difficult to study the mechanistic behavior. As no crystallography data are available, we chose the modeled human PINK1 structure from the AlphaFold database. Recently elucidated structures of orthologous insect PINK1 strongly resemble the AlphaFold predicted structure especially having high confidence level for the kinase domain(N-lobe and C-lobe). The mutants G309D and L347P that reside on the N-lobe and C-lobe of the kinase domain are shown in Fig. 1(b) and (c) that tend to decrease the activity by affecting protein structure, dynamics and substrate selectivity.

Identification of small molecule activators for PINK1 mutants

To discover the potential drugs selective against the PINK1 mutant protein, we screened PKIDB database having 300 kinase specific molecules by performing molecular docking implemented in PRinS3 Suite. The molecules are docked in the cleft formed between the N-lobe and C-lobe consisting of residues as the key binding site. In Fig. 2a distribution of docking binding energies were shown as box and whisker plot. The top and bottom of whiskers represent the minimum and maximum of docking energies. Interestingly, no significant differences in energies were observed among the two mutants. Finally selected the top 7 molecules that were common among both the mutants having appreciable binding energies in the range of ~-10.69 to ~-8.12 kcal/mol and the type of docked interactions. The list of selected drugs with their disease indication, protein target and the stage of clinical phase have been mentioned in Supp Table 1. All of these drugs have been mainly used for treating cancer related disease conditions and never have been tested against PINK1 protein. The Tepotinib drug showed the strongest binding energy with ~ -10.69 kcal/mol for mutant G309D and − 10.37 kcal/mol for mutant L347P. In particular 4 drugs Lapatinib, Zandelisib, Tepotinib and Refametinib showed best docking energy (~-10 kcal/mol) for mutant L347P. The compounds expressed significant hydrophobic and hydrogen bond interactions in the binding pocket as shown in Supp. Tables 2 and 3. The P-loop residues(168ALA) and Beta-2 residue(170VAL) shown to be forming hydrophobic interactions in all the 5 protein-drug combinations. Except for the Refametinib-L397P complex, and other 4 complexes showed hydrophobic interaction with Beta-7 residue LEU369 present in the C-lobe. Interestingly, Zandelisib-G309D, Pyrotinib-G309D, Selatinib-G309D and Sitravatinib-L347P interacted with DFG motif residue ASP384 through hydrogen bonding. The docked conformation and their interactions of residues within the binding pocket are shown in Supp. Figure 1. To validate our docked poses we attempted to re-dock the crystal structure of TcPINK1 protein to AMP to its binding site using the PRinS3 X-ESS app (PDB ID: 7MP9). The re-docking results for AMP were consistent with respect to pose obtained from X-ESS. The RMSD values of the pose generated relative to that of the original pose was 1.5 Å respectively as shown in Supp. Figure 1. Previous computational studies have suggested that a pose with an RMSD value of < 2 Å is a good and acceptable pose⁴⁶.

Insight into stability and interactions between PINK1 mutants and ligands

To further investigate the stability of protein-ligand complexes and their binding mode at an atomic level we performed molecular dynamics simulations. The simulations were run for all 7 protein-ligand combinations for 200ns each totalling a simulation time of 2.8microseconds. To check for the flexibilities of the individual residues that may have contributed to the overall fluctuations in the system, RMSF(Root Mean Square Fluctuations) was computed for all the protein–ligand complexes and it was observed to be similar as shown in Supp. Figure 2. Analysis revealed distinct peaks at the loop which is between Beta-2 and Ins1 region to be very dynamic for ligands Zandelisib and Tepotinib in both mutants. However, the Refametinib-G309D complex showed high flexibility in the Ins3 region. Interestingly, Pyrotinib, Lapatinib and Sitravatinib complexed with G309D mutant protein showed most stable conformation, with restricted motions. In contrast, L347P mutant protein showed high flexibility overall for all ligands, and was more pronounced in Zandelisib, Tepotinib and Lapatinib. Together, these results demonstrated the overall stability of protein-ligand complexes with acceptable fluctuations in the N-lobe.

Further, to quantitatively characterize the relevant interactions between the PINK1-ligand complex the minimum distance between them was computed as shown in Fig. 3. For all complexes the interactions majorly overlapped with functionally important motifs within the protein. It was observed that key interactions were formed with residues GLY163, LYS164 and GLY165 of P-loop with all 7 ligands in both the mutants. More specifically, gatekeeper residue M318 is shown to interact with ligands Refametinib, Lapatinib, Tepotinib and Selatinib similar in both the mutants. However, catalytic loop interactions of residue D366 from the catalytic loop were observed in all G309D mutants, except for Tepotinib and Selatininb ligands in L347P mutant. In contrast, D384 from the DFG motif is shown to be interacting with ligands Refametinib, Pyrotinib, Sitravatnib, Lapatinib and Zandelisib in both mutants. These residues are known to be very close to each other and orient to form an active site for substrate binding. In particular, mutations at the D366(catalytic loop) and D384(DFG motif) have shown to affect kinase activity. Interestingly, the L347P mutant is present quite close to the D384 of the DFG motif. The above results indicate that ATP binding relevant residues are involved in interaction between the ligand and residues D366 and D384, becoming key to stabilizing the orientation of ligands within the binding pocket.

Free energy landscape of small molecule unbinding from binding pocket

As discussed above, the interactions between the ligands and the PINK1 have been depicted by docking and MD simulations, but lacking the binding affinities/selectivity. The ability to accurately predict the binding free energies of small molecules to proteins is vitally important for drug discovery. Metadynamics simulations offer a powerful tool for achieving this, allowing unprecedented detail into the positions of potential binding sites. In order to achieve this, the center of mass distance between the binding site and the ligand has been chosen as the collective variable. As shown in Figs. 4 and 5, the most favorable free energy of binding was observed for ligand Tepotinib and Lapatinib in G309D mutant and Pyrotinib and Lapatinib in L347P mutant. Free energy of binding for the rest of the protein-ligand complexes are shown in Supp. Figures 4 and 5. In G309D mutant, there is a -2kJ/mol difference in binding free energy between Lapatinib and Tepotinib, suggesting that Lapatinib has higher affinity towards the mutant. Similarly, in L347P mutant Pyrotinib has higher binding affinity than Lapatinib with a difference of -3kJ/mol. Interestingly, the ligand Pyrotinib binds with a large and shallower energy minimum at a center of mass distance of 0.48nm to 0.75nm. Alternatively, Lapatinib binds to both the mutants via multiple equally probable states and smaller energetic barriers compared to the other ligands. In Tepotinib, two distinct minima are observed at a distance of 0.45nm and 1.1nm separated by a small energy barrier.

One of the important favorable interactions formed between the ligand and the binding sites of protein is the hydrogen bond which was explored for the most probable state/minima. In G309D mutants, the Tepotinib forms a hydrogen bond with GLN327, LYS219, GLN205 and ASP384 whereas Lapatinib forms hydrogen bonds with LYS164,CYS166,SER167 and ASP366. Similarly, in mutant L347P Pyrotinib forms a hydrogen bond with ASP366, GLN327, ILE162, LYS164 and ASP384 while Lapatinib forms LYS164,GLN327 and ASP366. Interestingly, ligand Tepotinib in G309D and Pyrotinib in L347P mutants interact at overlapping binding sites by utilizing LYS from ALK motif, GLN from alpha-D and ASP from DFG motif. Furthermore, Lapatinib in both mutants form strong hydrogen bond interaction with P-loop and Catalytic loop. For most of the kinases, primary regions involved in controlling the activation process are the DFG motif and activation loop. Basically, rearrangement of activation loop has been suggested to accommodate ATP and substrate for activation mechanism. In an active conformational state, the ASP from DFG motif points to the ATP binding site while interacting with ATP.

Organizing N-lobe for substrate binding

Most kinases when activated lead to the role rearrangement of key residues within the protein for their substrate specificity. In particular, in PINK1 protein the Ins3 contributes to formation of a bowl-shaped binding site for ubiquitin that is very important for PINK1 mediated ubiquitin and Parkin Ubl phosphorylation^19,47. To quantitatively explore the groove formed in PINK1, we calculated the minimum distance between the CYS166 from P-loop and THR454 as seen in Fig. 6. The CYS166 is unusual in kinases and shown to be conserved in human PINK1. The Lapatinib ligand in both mutants formed a very wide and broad groove with a distance of ~ 1.9 and 2.1nm. Similarly, the distance was observed to be 1.8nm in Tepotinib G309D mutant and Pyrotinib in L347P. Experimentally it has been shown that orthologous TcPINK1 forms a 2.2nm groove to recognize and accommodate Ub substrate¹⁹. The functional importance of this groove to Ub phosphorylation was observed by introducing the mutations of residues on the groove and testing their effects on the kinase activities for Ub. Alongside, the SER228 and SER230 residues are highly conserved in human PINK1 and PhPINK1 and shown to be structurally important which is present in catalytic domain. The crystal structure of PhPINK1 revealed the role of SER228 in coordinating with the Ins3 region for substrate binding. Basically, SER228 is shown to form a close interaction with the end point of the Ins3 residues securing the ends in place. To quantify this we calculated the minimum distance between the residues HIS310 and ARG312 from the Ins3 to the SER228 as shown in Fig. 7. The closest distance between the SER228 and Ins3 ARG312 is seen for both G309D Tepotinib and L347P Lapatinib at around 0.45nm, whereas for Ins3 HIS310 all three G309D Tepotinib, G309D Lapatinib and L347P Lapatinib are at 0.52nm. Furthermore, Ins3 residues VAL284 and ARG312 distances are at 0.5nm for all the 3 ligands in both the mutants. Henceforth, the Ins3 in the N-lobe mediates a close interaction within the helix initiating a rigid organization. Such organization of the Ins3 region by SER228 and formation of a wide N-lobe groove are shown to be crucial for kinase integrity and activity, and also binding specificity of substrate ubiquitin to the N-lobe.

PINK1 (PTEN-induced kinase 1) proteins are one of the most diverse protein kinases seen in the human kinome. Recent crystallography structures elucidated for the fruit fly Drosophila melanogaster and mosquito Anopheles gambiae PINK1^18,19,47 have shown similarities to human PINK1 especially in the conserved kinase domain and transmembrane domain that infer this divergence. The PINK1 gene is known to control the specific elimination of dysfunctional mitochondria, thus fine-tuning mitochondrial networks and preserving energy metabolism, suggesting that the activation of the PINK1/Parkin pathway may be a potential therapeutic target for Parkinson's disease. Mutations in PINK1, such as G309D and L347P, are associated with recessive familial type 6 of Parkinson's disease (PARK6)^48,49. Studies have shown that the G309D mutation may disrupt key interactions between PINK1 and its binding partners, potentially affecting its ability to phosphorylate substrates and regulate mitochondrial quality control and the L347P mutation appears to have a more localized effect on the conformation of the PINK1 protein, potentially altering its stability and/or activity⁵⁰. Recent years, there has been an increase in the number of clinical trials targeting different issues related to mitochondrial dysfunctioning, but in perspective PINK1 is still lacking.

One of the key reasons that the treatment and understanding of the human PINK1 protein has been difficult is the unavailability of its crystal structure. Alphafold's predicted structure of the human PINK1 protein, which has a high confidence score in the kinase domain, has therefore been used in our study. Another reason is that Alphafold's structure predicts ordered insertion regions that have no homology to other kinase proteins showing close alignment between several experimental elucidated PINK1 structures and the predicted structure^51,52. Repurposing approved drugs can be a cost-effective and time-efficient approach to developing therapeutics for emerging diseases or conditions for which there are no effective treatments. In the case of PINK1 activation, there are several potential approaches that have been explored, including the use of ATP analogs and small molecule inhibitors of kinases that have some specificity towards PINK1. There are approximately over 300 kinase inhibitors that have been approved or are in clinical trials, none of which have been tested against PINK1 protein. Therefore, it would be necessary to conduct screening assays to identify compounds that have the desired activity and selectivity towards PINK1. To efficiently screen large datasets we employed the strength of Insilico methods: Molecular docking, Molecular dynamics simulations and Free energy calculations. After screening, only 3 molecules Tepotinib, Pyrotinib and Lapatinib were chosen as potential activators against mutant PINK1 protein. Lapatinib a quinazoline derivative ⁵³ and Pyrotinib ⁵⁴ a quinoline derivative both have shown to strongly inhibit HER-1 and HER-2 mutant proteins in breast cancer. Alongside, Tepotinib⁵⁵ an pyridazinones derivative has shown activity against MET kinase protein in treating lung cancer. Although no literature is present regarding the possible use of these three molecules as PINK1 activators. Interestingly, the most favorable interaction found for Lapatinib is between the P-loop regions and catalytic loop. The P-loop region is known to bind and coordinate the ATP molecule. Recent studies have highlighted that kinase activity gets inhibited due to the mutation in LYS from ALK motif, ASP from catalytic loop and ASP from DFG motif. They are also key residues involved in binding and orientation the ATP molecule. In our study we observe that Lapatinib, Tepotinib and Pyrotinib molecules form strong hydrogen bond interactions with these key determinants of PINK1 activity. Most of the kinases confer an active conformation defined by DFG ASP in conformation. In both mutants, DFG ASP in confirmation directs inside the protein, which we observed for Tepotinib and Pyrotinib molecules.

Once the PINK1 protein is activated it requires a specific conformation organization of the N-lobe and C-lobe for the substrate to bind and phosphorylate. The key structural elements were suggested to be the Ins3 region and SER228 residue, organization of Ins3 region can lead to dissociation of dimers and binding of substrates. All the 3 molecules in our study suggest a formation of groove spanning the N-lobe and C-lobe that could accommodate Ub substrates. Similarly residue SER228 initiates interactions with the termini residues of Ins3 region locking the conformation. It has been suggested that Ub directly does not interact with residue SER228 and its location is a key determinant in early PINK1 activation. Overall, this study provides valuable insights into possible small molecule-based drugs selective against the PINK1 mutant protein and may serve as a platform for future investigations on therapeutic uses.

Protein and Ligand Preparation

The starting structure of PINK1 was retrieved from AlphaFold Protein Structure Database [Database Identifier : AF-Q9BXM7] ³⁶. The residue length 111–575 was chosen based on the per-residue confidence level (> 75 pLDDT) of the protein model. The N-helix region and kinase domain were considered, excluding the disordered mitochondrial targeting and transmembrane region as shown in Fig. 1 (b) and (c). Mutations were introduced in the respective regions; G309D and L347P in protein structure. The ligand dataset was obtained from a protein kinase inhibitor database also referred as PKID in SDF file format (https://www.icoa.fr/pkidb/)³⁵. The database contains 300 FDA approved and or clinical trials that were well annotated and curated protein kinase inhibitors.

Molecular Docking and Molecular Dynamics Simulation using PRinS3 ( X-ESS )

PRinS3 ³⁷ has an automated pipeline designed for drug discovery which holds multi-ligand multi-protein docking, molecular dynamics (MD) and well tempered metadynamics (wt-metaD) for free energy calculation. Large number of calculations were upscaled using a data connector (DC) from PRinS3. The two mutant protein and ligand was prepared in X-ESS app by addition of hydrogens and preprocessing before the docking. Finally 600 combinations of protein and ligands were generated and used for docking. Molecular docking procedures were carried out by defining a grid corresponding to the ligand binding site. Grid maps were generated using grid box of size 60 x 60 x 60 Å³ with 0.375 Å spacing. The X-ESS app of PRinS3 uses Lamarkian global genetic algorithm³⁸ for searching best possible ligand poses. Based on the docking energy, Hydrogen bond patterns 7 best docked protein ligand combinations were chosen and further subjected to molecular dynamics(MD) simulations. A total of 14 combinations; 7 ligands for each mutant were chosen. The idea behind performing MD simulations is to determine the stability of the selected small molecules or ligands and to understand the binding selectivity of residues within the protein. The topology parameters for protein and ligands were generated using CHARMM27 forcefield ^39,40. All the combinations were solvated using the TIP3P water model, followed by applying Periodic Boundary conditions. Each combination was neutralized by adding counter ions. The input parameters were set in the X-ESS app for energy minimization and production MD runs. The Steepest descent minimization algorithm with 50000 steps was used for energy minimization. Following this step, NVT and NPT performed by maintaining the temperature and pressure at 300K and 1 atm respectively. For NVT, Modified Berendsen thermostat ⁴¹were used and for NPT Parrinello-Rahman barostat ⁴². The long range Coulombic interactions were calculated using Particle-Mesh Ewald(PME) summation⁴³, and for short-range interactions was 1.0 nm. For 14 combinations of protein-ligand production run was performed for 200 ns, totalling simulation data of 2.8 µs. The RMSD and RMSF analysis were carried out using the PRinS3 platform. Further enhanced sampling technique Well-tempered metadynamics was performed by using the equilibrated structure from the MD simulation data. The Plumed 2.8.0 version⁴⁴ combined with GROMACS 2018⁴⁵ has been integrated into the PRinS3 platform. The reaction coordinate chosen here is the distance between the center of mass of ligand heavy atoms and center of mass of protein binding site. Temperature was maintained at 300 K, Gaussian hill height 1.5 kJ/mol, hill width 0.01 nm, bias factor of 10. For each protein-ligand combination, 10 independent runs were carried out to calculate the free energy barrier.

ACKNOWLEDGMENTS

ABP, SB and SR thanks Indian Institute of Technology Kanpur and Prescience for financial support. JKS would like to acknowledge SERB, India, for their support.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Hoehn, M. M. & Yahr, M. D. Parkinsonism: onset, progression, and mortality. Neurology 17, 427–427 (1967).
Valente, E. M. et al. Hereditary Early-Onset Parkinson’s Disease Caused by Mutations in PINK1. Science 304, 1158–1160 (2004).
Nguyen, T. N., Padman, B. S. & Lazarou, M. Deciphering the Molecular Signals of PINK1/Parkin Mitophagy. Trends Cell Biol. 26, 733–744 (2016).
Harper, J. W., Ordureau, A. & Heo, J.-M. Building and decoding ubiquitin chains for mitophagy. Nat. Rev. Mol. Cell Biol. 19, 93–108 (2018).
Pickrell, A. M. & Youle, R. J. The Roles of PINK1, Parkin and Mitochondrial Fidelity in Parkinson’s Disease. Neuron 85, 257–273 (2015).
Hereditary Early-Onset Parkinson’s Disease Caused by Mutations in PINK1. Science, 304(5674), 1158–1160 | 10.1126/science.1096284. https://sci-hub.ru/https://www.science.org/doi/10.1126/science.1096284?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed.
Gandhi, S. & Plun-Favreau, H. Mutations and mechanism: how PINK1 may contribute to risk of sporadic Parkinson’s disease. Brain 140, 2–5 (2017).
Han, H. et al. PINK1 phosphorylates Drp1S616 to regulate mitophagy-independent mitochondrial dynamics. EMBO Rep. 21, e48686 (2020).
Gandhi, S. et al. PINK1-Associated Parkinson’s Disease Is Caused by Neuronal Vulnerability to Calcium-Induced Cell Death. Mol. Cell 33, 627–638 (2009).
Bayne, A. N. & Trempe, J.-F. Mechanisms of PINK1, ubiquitin and Parkin interactions in mitochondrial quality control and beyond. Cell. Mol. Life Sci. 76, 4589–4611 (2019).
Geisler, S. et al. The PINK1/Parkin-mediated mitophagy is compromised by PD-associated mutations. Autophagy 6, 871–878 (2010).
McWilliams, T. G. & Muqit, M. M. PINK1 and Parkin: emerging themes in mitochondrial homeostasis. Curr. Opin. Cell Biol. 45, 83–91 (2017).
Mills, R. D. et al. Biochemical aspects of the neuroprotective mechanism of PTEN-induced kinase-1 (PINK1). J. Neurochem. 105, 18–33 (2008).
Abramov, A. Y. et al. Bioenergetic Consequences of PINK1 Mutations in Parkinson Disease. PLoS ONE 6, e25622 (2011).
Beilina, A. et al. Mutations in PTEN-induced putative kinase 1 associated with recessive parkinsonism have differential effects on protein stability. Proc. Natl. Acad. Sci. 102, 5703–5708 (2005).
Deas, E., Plun‐Favreau, H. & Wood, N. W. PINK1 function in health and disease. EMBO Mol. Med. 1, 152–165 (2009).
Sha, D., Chin, L.-S. & Li, L. Phosphorylation of parkin by Parkinson disease-linked kinase PINK1 activates parkin E3 ligase function and NF- B signaling. Hum. Mol. Genet. 19, 352–363 (2010).
Kumar, A. et al. Structure of PINK1 and mechanisms of Parkinson’s disease-associated mutations. eLife 6, e29985 (2017).
Okatsu, K. et al. Structural insights into ubiquitin phosphorylation by PINK1. Sci. Rep. 8, 10382 (2018).
Sim, C. H., Gabriel, K., Mills, R. D., Culvenor, J. G. & Cheng, H.-C. Analysis of the regulatory and catalytic domains of PTEN-induced kinase-1 (PINK1). Hum. Mutat. 33, 1408–1422 (2012).
Rasool, S. et al. Mechanism of PINK1 activation by autophosphorylation and insights into assembly on the TOM complex. Mol. Cell 82, 44-59.e6 (2022).
Puschmann, A. et al. Heterozygous PINK1 p.G411S increases risk of Parkinson’s disease via a dominant-negative mechanism. Brain 140, 98–117 (2017).
Fiesel, F. C. et al. Substitution of PINK1 Gly411 modulates substrate receptivity and turnover. Autophagy 0, 1–22 (2022).
Prestel, J. et al. Clinical and molecular characterisation of a Parkinson family with a novel PINK1 mutation. J. Neurol. 255, 643–648 (2008).
Julienne, H., Buhl, E., Leslie, D. S. & Hodge, J. J. L. Drosophila PINK1 and parkin loss-of-function mutants display a range of non-motor Parkinson’s disease phenotypes. Neurobiol. Dis. 104, 15–23 (2017).
Shiba-Fukushima, K. et al. A Cell-Based High-Throughput Screening Identified Two Compounds that Enhance PINK1-Parkin Signaling. iScience 23, 101048 (2020).
Zhang, C. et al. Sorafenib targets the mitochondrial electron transport chain complexes and ATP synthase to activate the PINK1–Parkin pathway and modulate cellular drug response. J. Biol. Chem. 292, 15105–15120 (2017).
Zhang, S. et al. Inhibition of the PINK1-Parkin Pathway Enhances the Lethality of Sorafenib and Regorafenib in Hepatocellular Carcinoma. Front. Pharmacol. 13, 851832 (2022).
Barini, E. et al. The Anthelmintic Drug Niclosamide and Its Analogues Activate the Parkinson’s Disease Associated Protein Kinase PINK1. ChemBioChem 19, 425–429 (2018).
Hertz, N. T. et al. A Neo-Substrate that Amplifies Catalytic Activity of Parkinson’s-Disease-Related Kinase PINK1. Cell 154, 737–747 (2013).
Li, N. et al. Gefitinib facilitates PINK1/Parkin-mediated mitophagy by enhancing mitochondrial recruitment of OPTN. Fundam. Res. 2, 807–816 (2022).
Miller, S. & Muqit, M. M. K. Therapeutic approaches to enhance PINK1/Parkin mediated mitophagy for the treatment of Parkinson’s disease. Neurosci. Lett. 705, 7–13 (2019).
Osgerby, L. et al. Kinetin Riboside and Its ProTides Activate the Parkinson’s Disease Associated PTEN-Induced Putative Kinase 1 (PINK1) Independent of Mitochondrial Depolarization. J. Med. Chem. 60, 3518–3524 (2017).
Dar, A. C. & Shokat, K. M. The Evolution of Protein Kinase Inhibitors from Antagonists to Agonists of Cellular Signaling. Annu. Rev. Biochem. 80, 769–795 (2011).
Carles, F., Bourg, S., Meyer, C. & Bonnet, P. PKIDB: A Curated, Annotated and Updated Database of Protein Kinase Inhibitors in Clinical Trials. Molecules 23, 908 (2018).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Prescience Insilico Solution Suite (PRinS3). https://www.prescience.in/prins.
Morris, G. M. et al. AutoDock4 and AutoDockTools4: Automated Docking with Selective Receptor Flexibility. J. Comput. Chem. 30, 2785–2791 (2009).
MacKerell Jr., A. D., Banavali, N. & Foloppe, N. Development and current status of the CHARMM force field for nucleic acids. Biopolymers 56, 257–265 (2000).
Zoete, V., Cuendet, M. A., Grosdidier, A. & Michielin, O. SwissParam: A fast force field generation tool for small organic molecules. J. Comput. Chem. 32, 2359–2368 (2011).
Bussi, G., Donadio, D. & Parrinello, M. Canonical sampling through velocity-rescaling. J. Chem. Phys. 126, 014101 (2007).
Parrinello, M. & Rahman, A. Polymorphic transitions in single crystals: A new molecular dynamics method. J. Appl. Phys. 52, 7182–7190 (1981).
Essmann, U. et al. A smooth particle mesh Ewald method. J. Chem. Phys. 103, 8577–8593 (1995).
Bonomi, M. et al. PLUMED: a portable plugin for free-energy calculations with molecular dynamics. Comput. Phys. Commun. 180, 1961–1972 (2009).
Abraham, M. J. et al. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1–2, 19–25 (2015).
Sándor, M., Kiss, R. & Keserű, G. M. Virtual Fragment Docking by Glide: a Validation Study on 190 Protein−Fragment Complexes. J. Chem. Inf. Model. 50, 1165–1172 (2010).
Schubert, A. F. et al. Structure of PINK1 in complex with its substrate ubiquitin. Nature 552, 51–56 (2017).
Narendra, D. P. & Youle, R. J. Targeting Mitochondrial Dysfunction: Role for PINK1 and Parkin in Mitochondrial Quality Control. Antioxid. Redox Signal. 14, 1929–1938 (2011).
Cardona, F., Sánchez‐Mut, J. V., Dopazo, H. & Pérez‐Tur, J. Phylogenetic and in silico structural analysis of the Parkinson disease‐related kinase PINK1. Hum. Mutat. 32, 369–378 (2011).
Xiong, H. et al. Parkin, PINK1, and DJ-1 form a ubiquitin E3 ligase complex promoting unfolded protein degradation. J. Clin. Invest. 119, 650–660 (2009).
Gan, Z. Y. et al. Activation mechanism of PINK1. Nature 602, 328–335 (2022).
Kakade, P. et al. Mapping of a N-terminal α-helix domain required for human PINK1 stabilization, Serine228 autophosphorylation and activation in cells. Open Biol. 12, 210264 (2022).
Opdam, F. L., Guchelaar, H.-J., Beijnen, J. H. & Schellens, J. H. M. Lapatinib for Advanced or Metastatic Breast Cancer. The Oncologist 17, 536–542 (2012).
Jacobson, A. Pyrotinib Improves Survival in Previously Treated HER2-Positive Metastatic Breast Cancer. The Oncologist 27, S5–S6 (2022).
Paik, P. K. et al. Tepotinib in Non-Small-Cell Lung Cancer with MET Exon 14 Skipping Mutations. N. Engl. J. Med. 383, 931–943 (2020).

No competing interests reported.

ABPSupp.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of Kinase inhibitors as PINK1 activators for the treatment of Parkinson’s Disease

Status:

Version 1

Abstract

Figures

Introduction

Results And Discussion

Identification of small molecule activators for PINK1 mutants

Insight into stability and interactions between PINK1 mutants and ligands

Free energy landscape of small molecule unbinding from binding pocket

Organizing N-lobe for substrate binding

Conclusion

Methodology

Protein and Ligand Preparation

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1