Computational identification of potential tau tubulin kinase 1 (TTBK1) inhibitors: a structural analog approach

Abstract

Abnormal deposition or aggregation of protein alpha-synuclein and tau in the brain leads to neurodegenerative disorders. Excessive hyperphosphorylation of tau protein and aggregations destroys the microtubule structure resulting in neurofibrillary tangles in neurons and affecting cytoskeleton structure, mitochondrial axonal transport, and loss of synapses in neuronal cells. Tau tubulin kinase 1 (TTBK1), a specific neuronal kinase is a potential therapeutic target for neurodegenerative disorders as it is involved in hyperphosphorylation and aggregation of tau protein. TTBK inhibitors are now the subject of intense study, but limited numbers are found. Hence, this study involves structure-based virtual screening of TTBK1 inhibitor analogs to obtain efficient compounds targeting the TTBK1 using docking, molecular dynamics simulation and protein-ligand interaction profile. The initial analogs set containing 3884 compounds was subjected to Lipinski rule and the non-violated compounds were selected. Docking analysis was done on 2772 compounds through Autodock vina and Autodock 4.2. Data Warrior and SwissADME was utilized to filter the toxic compounds. The stability and protein-ligand interaction of the docked complex was analyzed through Gromacs and VMD. Molecular simulation results such as RMSD, Rg, and hydrogen bond interaction along with pharmacokinetic properties showed CID70794974 as the potential hit targeting TTBKl prompting the need for further experimental investigation to evaluate their potential therapeutic efficacy in Alzheimer’s disease.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s40203-024-00242-z.

Introduction

Treatment of neurodegenerative diseases (NDs) is one of the biggest challenges currently because of its part in causing mortality and morbidity in older people. NDs include accumulation of cerebral proteinopathies which is caused by the clumping of insoluble aggregates of protein such as tau in nerve cells and non-nerve cells that leads to a motor, cognitive and behavioral deficit (Goedert 2018). Alzheimer’s disease (AD) and progressive supranuclear palsy (PSP) are among the 20 clinicopathological entities of tauopathies. Tau pathology occurs almost in all neurodegenerative diseases and seen often in Dementia with Lewy body (DLB) (Burbank and Mitchison 2000). Tau, which is microtubule, associated protein in its normal phosphorylation state regulates stability and movement of microtubule (MTs), axonal transport and neurite outgrowth. Post translational modifications such as phosphorylation, nitration, truncation, acetylation, glycation, and methylation are involved in regulating tau (Cohen et al. 2011; Funk et al. 2015) (Burbank and Mitchison 2000). Tau is a highly soluble brain protein but in neurodegenerative disorders, due to modifications such as high phosphorylation, tau gets aggregated and becomes insoluble. Tau serves as a substrate for kinases such as protein kinase A (PKA) (Chatterjee et al. 2009), TTBK1 and glycogen synthase kinase 3 (GSK3) (Cohen et al. 2011). The mechanism of tau deposition remains unexplored but tau phosphorylation and tau truncation result in conformational alterations in polypeptides further leading to disease related tau aggregation and accumulation (Hanger and Wray 2010). There are six tau isoforms in the human brain of adults with three or four MT binding repeats (3R or 4R). These repetitive domains facilitate the binding of tau to MTs and, they play a role in the nucleation process of tau aggregation (Goode et al. 2000; Mylonas et al. 2008). Inhibition of aggregation is done by Phenylthiazolylhydrazide N-Phenylamines Rhodanine-based inhibitors and benzothiazoles, Anthraquinones, phenothiazines, perphenazine, flavonoid polyphenols and porphyrins which has special characteristics such as permeability, cytotoxicity, and disaggregation ability (Bulic et al. 2009).

Tau tubulin kinase TTBK ((EC 2.7.11.26) is a member of casein kinase I family and are abundantly expressed in hippocampus, midbrain, substansia nigra, cerebellum Purkinje cells and granular cell layer (Ahamad et al. 2021). The tau protein kinase comprises of the N-terminal region, proline rich domain (40–49) and C-terminal region. The N-terminal and C-terminal regions are connected through hinge region (108–111) (Goedert et al. 1989). TTBK isoforms TTBK1 and TTBK2 are highly homologous with 88.7% identity in their catalytic pocket but they vary in tissue distribution. TTBK1 is a neuron specific kinase and TTBK2 resides in heart, muscle, liver, thymus, spleen, lung, kidney, testis, and ovaries (Takahashi et al. 1995). TTBK1 that can phosphorylate tau protein and are also expressed uniquely in neurons in central nervous system (CNS) has been identified by screening brain-specific kinases through the complementary DNA library (Ikezu and Ikezu 2014; Sato et al. 2006). TTBK2 has been an attractive target in many types of cancer. However, TTBK1 role has not been well established but it has been said to phosphorylate Tau, Gsk3β, CDK5, TDP43 and tubulin. The upregulation of TTBK1 found in patients of Alzheimer disease depicts the vital role of TTBK1 in disease progression and it is also responsible for causing amyotrophic lateral sclerosis (ALS) and childhood onset schizophrenia (John et al. 2019; Sato et al. 2008). TTBK1 phosphorylates the tau protein at Ser422 where pS422 is the critical biomarker for PHF (Paired helical filaments). In a study, mutant human TTBK1 and FTLD-causing tau expressed in double transgenic mouse model was observed to induce and increase aggregation of oligomeric tau, loss of motor neurons of spinal cords and enhanced motor deficiency (Katrachanca and Koleske 2017). The above-mentioned changes are related with elevated neuroinflammation by TTBK1-dependent mononuclear phagocytes activation and this proves that TTBK1 may induce and increase the toxicity of tau in terms of both direct phosphorylation and pro-inflammatory pathways (Asai et al. 2014). These data finally commend a major role of TTBK1 in stimulating neurodegeneration associated with tau (Taylor et al. 2020). With the above said context, a good strategy would be modulating TTBK1 activity as its upregulated activity accounts to AD pathology (Halkina et al. 2021). However, their clinical application is hampered by limitations such as lack of clinical data, specificity, potential side effects and disease complexity. Until today only five drugs have been authorized for AD treatment and also no new drugs have been approved after 2003, majority of the failed drugs focused on targeting accumulation of Aβ (Amyloid-β peptide) (Panza 2019). There are numerous serious neurological diseases which includes FTLD, ALS or AD (Frontotemporal dementia, Amyotrophic lateral sclerosis or Alzheimer’s disease), in which elevation of TTBK1 activity disrupts the stability of various proteins such as TDP-43 or tau. Successful treatment for the above pathologies is lacking. Hence, compounds capable of targeting these kinases may become an effective new therapeutic alternative (Nozal and Martinez 2019). So far only 3 TTBK1 and TTBK2 inhibitors have been discovered and all the compounds exhibited affinity for both isoforms (Kiefer et al. 2014; Xue et al. 2013).

The hunt for potent drug that is capable of blocking or delaying the upcoming loss of nerve cells is an immediate need and also its one of the essential challenges of our current century Various kinases have been involved in distinct neurodegenerative diseases but the lack of inhibitors for the cure of these diseases leads to a fascinating chance for development of drugs (Wu & Nielsen 2016). Analogy plays an important role in scientific research. Analogue based approach of drug design is one of the oldest methodologies of medicinal chemistry and still is intensively exploited one. Enforced by new techniques, such as combinatorial chemistry and computer aided drug design, structural analogy is a rich source for new substances of potential medical importance (Kafarski and Lipok 2015). Hence, the aim of our current study is to identify hit compounds against the TTBK1 that help medicinal chemists to improve compounds with desired properties such as improved inhibitory efficacies and pharmacokinetics.

Materials and methods

Protein preparation

The kinase domain of TTBK1 was obtained from the (Protein data bank) PDB. The low resolution (1.42 Å) and lack of missing residues in the active site of 4NFN made it the ideal candidate for our study (Kiefer et al. 2014). TTBK1 (Uniprot ID: Q5TCY1) has a total of 1321 amino acids, while the crystal structure 4NFN has 321 amino acids of the kinase domain which is required for our study. The previous literature was used to determine the catalytic pocket of the protein molecule TTBK1 (Halkina et al. 2021; Halkina et al. 2021; Kiefer et al. 2014; Nozal 2022).

Ligand preparation

The co-crystalized ligands 2KC (3-({5-[(4-amino-4-methylpiperidin-1-yl)methyl]pyrrolo[2,1-f][1,2,4]triazin-4-yl}amino)-5-bromophenol) was extracted from the 4NFN and used as reference ligand. The ligand library for virtual screening was developed using a PubChem tanimoto co-efficient search with a 90% similarity (Kim et al. 2023).TTBK1 inhibitors such as 2KC, DTQ, VP7 (4-(2-amino-5,6,7,8-tetrahydropyrimido[4’,5’:3,4]cyclohepta[1,2-b]indol-11-yl)-2-methylbut-3-yn-2-ol), VSY ((3 S)-1-[1-(2-aminopyrimidin-4-yl)-1 H-pyrazolo[4,3-c]pyridin-6-yl]-3-methylpent-1-yn-3-ol), 9IV (~{N}-[4-(2-chloranylphenoxy)phenyl]-7~{H}-pyrrolo[2,3-d]pyrimidin-4-amine), 9IO (~{N}-(4-methoxyphenyl)-7~{H}-pyrrolo[2,3-d]pyrimidin-4-amine) and F8E (methyl 2-bromo-5-(7 H-pyrrolo[2,3-d]pyrimidin-4-ylamino)benzoate) was utilized for similar structure search. The PubChem ID for the native ligands is CID135566890; CID5687; CID58221549; CID155908671; CID162640226; CID110875107; and CID1720884. The similar structure library was subjected to Lipinski rule of five and the compounds with zero violation were used for our study library. The force field MMFF94 was used to minimize the reference and library ligands for 500 steps using open Babel 3.3.1 (Boyle et al. 2011). Further, the ligands were converted into pdbqt in open Babel 3.3.1.

Validation study

The docking process was validated before proceeding with virtual screening using Autodock 4.2 (Morris et al. 2009). The pdbqt format of TTBK1 was prepared in Autodock 4.2 through removal of water, adding polar hydrogens and Kollman charges. The reference ligand prepared during the initial process were redocked into the experimentally proven catalytic pocket of TTBK1. The active sites of the protein TTBK1 was obtained from the literatures (Halkina et al. 2021; Kiefer et al. 2014; Nozal 2022).

The protein was kept rigid and grid parameters were set as 64 × 60 × 55 Å. The spacing was set as 0.44 Å. The Lamarckian genetic algorithm was applied to obtain 100 ligand poses. The validation process was done by superimposing the docked complex and PDB complex followed by the backbone RMSD measurement. Discovery studio 2021 and was used for visualizing the protein-ligand interaction.

Rank based screening

Rank based screening was employed for the similar structure library obtained from ligand preparation steps. The TTBK1 protein prepared for the validation process was docked against the ligand library using Autodock vina 1.1.2. The configuration parameters applied in the validation process were used here. The protein was kept rigid and the grid parameters were set as 64 × 60 × 55 Å and the spacing used was 0.44 Å at the active site. The algorithm used for screening is iterated local search algorithm. For further studies, compounds with the lowest binding energy were selected (Jaghoori et al. 2016). .

ADMET prediction

To evaluate drug-likeness and toxicity of selected compounds ADMET prediction was performed using SwissADME and Osiris data warrior (Daina et al. 2017).

Exhaustive screening

Ligands that were filtered by ADMET filter was employed to Autodock 4.2 with the same grid parameters used for the validation procedure. The algorithm used was Lamarckian genetic algorithm and 100 conformers were obtained for each ligand. The protein was kept rigid and the spacing was set as 0.44 Å. The grid parameters were set as 64 × 60 × 55 Å at the active site as like validation and fast rank screening. The ligands with the highest binding affinity, pose, cluster and key residue interaction as reported in previous studies was selected as the best hits. The binding poses and interactions was visualized in discovery studio.

Molecular simulation

Using GROMACS 2020.4 software, docked protein-ligand complexes of the best hits were subjected to molecular dynamic simulations (Kohnke et al. 2020). Molecular interaction abilities were determined by the forces acting on the system and those are generally parameterized by experimental data or by quantum chemical calculations. Molecular dynamic simulations were done using a force field called CHARMM27 (MacKerell et al. 1998). Ligand topologies were generated by Swissparam respectively (Zoete et al. 2012). TIP3P was used to dissolve the protein, which was set up in a cubic box of 1 cm in height. Ions of Na were added to the system to neutralize it. With NVT and NPT, temperature and pressure were stabilized at 300 K and 1 bar and energy minimization was done for 50,000 steps. After 100ns, XMgrace was used to analyse the RMSD (Root mean square deviation), SASA (solvent accessible surface area), hydrogen bond interaction, RMSF (Root mean square fluctuation), and Rg (radius of gyration).

Results and discussion

Redocking with reference ligands

Protein TTBK1 was docked with the co-crystalized ligand 2KC obtained from PDB to validate the docking study. From this docking analysis, it was seen that the docked ligand was bound to active sites of TTBK1, i.e., Gln110, Glu77, Asp176, Asn159, Phe177, Lys87 which were proved experimentally (Kiefer et al. 2014). Poseview, a server to predict active sites and experimentally proven literatures demonstrated that these residues i.e., Gln110, Glu77, Asp176, Asn159, Glu101, Lys87, Gln134, Phe177, Gln108 are the binding pockets of TTBK1 (Kiefer et al. 2014). During redocking of co-crystalized ligand 2KC to TTBK1, it was clear that the binding mode and pose of ligand was same as their experimentally proven binding site. Reference ligand (2KC) had a binding energy of -12.29 kcal/mol and it formed hydrogen bonds with Asn159, Asp176, Glu77, Lys63, Gln110 (Fig. 1). All the above results show that the docking protocol is conclusive for further studies.

Fig. 1.

Reference ligand- 2KC binding to active site residues of TTBK1

Virtual screening

PubChem, a system maintained by the national center for biotechnology information, contains chemical molecules and their activities towards biological target. The PubChem database contained 3884 similar structures of TTBK1 inhibitors and they were screened to eliminate the compounds that did not follow Lipinski’s rule of five. Among the selected compounds, 2772 compounds followed RO5 (Table 1). Compounds containing null Lipinski violation shows that the compounds have the best drug likeliness properties and acceptable oral bioavailability property (Lipinski 2004).

Table 1.

Similar structure library against TTBK1 inhibitors

PDB ID	PubChem ID	Total number of Analogs	Lipinski filtered compounds
4BTK	CID5687	3790	2681
4BTM	CID1720884	9	9
4NFN	CID135566890	5	5
7JXX	CID58221549	44	41
7JXY	CID155908671	31	31
7Q8V	CID162640226	3	3
7Q8W	CID1108751072	2	2

Lipinski filtered molecules were rendered to fast screening approach by applying virtual screening pipeline. Fast screening process was executed to 2772 compounds in Autodock vina 1.1.2 against TTBK1. Binding energy range of compounds at first set of screening were − 5.6 kcal/mol to -10.6 kcal/mol. To narrow down the list of compounds and their chemical space, the threshold was set as lesser than − 9.0 kcal/mol (Table S1). 56 compounds were selected based on their binding energy within the above said range and subjected to test ADMET properties (Fig. 2).

Fig. 2.

Binding energy range of all the compounds subjected to fast rank screening

ADMET filter

Osiris data warrior was used to predict the toxicity properties such as mutagenic, tumorigenic, reproductive effect, solubility, and synthetic accessibility for the selected compounds. 56 compounds that had the highest binding affinity were subjected to Osiris. The 49 compounds predicted as non-toxic in Osiris were taken for exhaustive screening. (Table S1).

Exhaustive screening

The exhaustive screening was done in Autodock 4.2 to the 49 compounds for 100 runs. To find the best results from an exhaustive screening approach, multi filtration protocol was done. The criteria used were clusters with lesser binding energy, highest frequency containing cluster and key residue interaction. After Molecular interaction analysis, it was found that CID70794974 and CID134264304 showed better results than other compounds by exhibiting good interaction with the active site of TTBK1 (Fig. 3). The other compounds were eliminated due to the formation of unfavorable interaction and less binding affinity.

Fig. 3.

Protein ligand interaction of selected compounds (CID70794974, CID134264304) against TTBK1

CID70794974 (3-[[5-[[(4R)-3,4-diaminopiperidin-1-yl]methyl]pyrrolo[2,1-f][1,2,4]triazin-4-yl]amino]phenol) is a structurally similar analog (92%) of 3-({5-[(4-amino-4-methylpiperidin-1-yl)methyl]pyrrolo[2,1-f][1,2,4]triazin-4-yl}amino)-5-bromo phenol (2KC). 2KC (CID135566890) has been predicted as mutagenic in protox II and formed four hydrogen bonds (Glu77, Glu110, Asn159, Asp176) in previous study ref. However, CID70794974, an analogue of 2KC was predicted as non-toxic and formed five hydrogen bonds including Glu110 and Asn159 (Tables 2 and 3). Furthermore, the compound demonstrated favorable ADMET and physicochemical properties, with a synthetic accessibility score of 2.99 in SwissADME. CID134264304 (3-[(2-amino-6,7-dimethoxyquinazolin-4-yl)amino]-N’-hydroxybenzenecarboximidamide) is a structurally similar analog (92%) of 4-[3-hydroxyanilino]-6,7-dimethoxyquinazoline (DTQ in PDB). DTQ is the first study in which an inhibitor has been co-crystalized with TTBK1. However, DTQ has been predicted as mutagenic and carcinogenic in protox II and formed three hydrogen bonds (Lys87, Glu101 and Glu134) in experimental study (Xue et al. 2013). CID134264304, an analogue of DTQ (CID5687) was predicted as non-toxic in our study and also formed four hydrogen bonds (Tables 2 and 3). The synthetic accessibility was predicted as 2.99 in SwissADME. Based on these findings, CID70794974 (Hit 1) and CID134264304 (Hit 2) were subjected to simulation.

Table 2.

Lipinski filter and ADMET properties for selected compounds

Pubchem ID	Molecular weight	H bond donor	H bond acceptor	LogP	Mutagenic	Tumorogenic	Reproductive effect	Solubility
CID70794974	353.42	4	6	2.45	None	None	None	Soluble
CID134264304	354.36	4	6	2.14	None	None	None	Soluble

Table 3.

Structure, binding energy and interactions of selected compounds

Molecular dynamics simulation

Molecular dynamics simulation is considered a promising approach in computer-aided drug discovery subject for studying the stability of the macromolecule in a dynamic state. In this study MD trajectories of TTBK1, TTBK1 with reference inhibitor (2KC), TTBK1 complexed with Hit 1 and Hit 2 were analyzed. TTBK1 exhibited an average RMSD of 0.203 nm and showed frequent oscillations throughout the simulation. The wide deviation was observed between 40 ns to 80 ns and thereafter the graph increased at 95 ns. The reference system attained stability for 15 ns to 75 ns producing an average RMSD of 0.206 nm. A mild increase in plateau was observed at 75 ns to 100 ns. Hit 1 showed a stable plateau for 10 ns to 60 ns and a gradual increase in RMSD was observed at 65 ns. Thereafter, the system attained stability producing an average RMSD of 0.202 nm. Hit 2 displayed an increase in RMSD for the initial 10 ns and thereafter the plateau attained stability for 30 ns to 85 ns with 0.3 nm. A sudden drop in RMSD up to 0.25 nm was observed at 85 ns and maintained the stable plateau to the end of the simulation producing an average of 0.28 nm. Apoprotein, reference complex, and hit 1 followed a similar RMSD but deep insight of the graph shows that hit 1 is more stable in comparison to all the complex with minimal fluctuations (Fig. 4A).

Fig. 4.

Stability analysis of TTBK1, Hit1, Hit2, reference (A) RMSD (B) RMSF (C) Radius of gyration (D) SASA

Root mean square fluctuation (RMSF) was plotted for 100 ns to get insight about the deviation of C alpha atoms present in each system including apoprotein. The average RMSF produced for TTBK1 alone, reference complex, hit 1 and hit 2 are 0.093 nm, 0.102 nm, 0.095 nm and 0.124 nm. Binding site residues (Glu77, Lys87, Gln108, Gln110, Glu101, Glu134, Ans159, Asp176, Phe177) of the TTBK1 showed no high fluctuations indicating the stability of both reference and hit complexes. Minor variation was seen in the loop regions (40–50 and 60–70) as these regions are more flexible than the other secondary structures (Fig. 4B).

Radius of gyration (Rg) is the vital parameter that gives insight into the structural modifications and overall compactness of the tertiary structure of the protein. Additionally, it defines the target protein’s rigidity and folding. The average Rg produced for TTBK1, reference, hit 1 and hit 2 are 2.01 nm, 2.02 nm, 2.03 nm, and 2.05 nm. Though all the systems had similar average Rg, the graph showed minor differences in their trend where the TTBK1 was stable throughout the simulation with a sudden drop in Rg value for 40 to 50ns. Reference compound and hit 1 showed a similar trend of Rg throughout the simulation process. The system bearing hit 2 had higher Rg in comparison to all other systems for 20 to 80 ns and the plateau dropped at 80 ns and attained stability in accordance with the RMSD results. The graph indicates reference and hit 1 is more compact in comparison to all other systems (Fig. 4C).

Using the GROMACS analytical utility tool g_sas, the solvent accessible surface area (SASA) was interpreted. SASA (Solvent Accessible Surface area) is a measure of a protein’s surface area that is accessible to water molecules both with and without ligands. The average SASA value observed for TTBK1, reference, hit 1 and hit 2 are 157.78 nm², 157.78 nm2, 160.52 nm² and 160 nm². The SASA profile of all the systems followed the same trend with no significant difference (Fig. 4D).

The occupancy of H-bond contacts between protein-ligands during the simulation was assessed to further support the molecular interaction stability of ligands with protein. H-bonds are essential for the stability of the protein-ligand complex in addition to conserving the conformational structure and stability of protein structure. The average number of hydrogen bonds formed during the MD simulation denotes their persistent hydrogen bonding interactions with the target protein’s binding site. Reference compound formed 3–4 hydrogen bonds throughout the simulation process and were reduced to 2 bonds between 40 ns – 60ns. The maximum number of hydrogen bonds are formed by key residues such as Gln77 (57%), Ser158 (49.1%) and Asp176 (44.16%) Hit 1 system formed 4 hydrogen bonds and was consistent throughout the simulation process (Fig. 5). The higher occupancy was contributed by the vital residue Gln77 (100%), Ser158(42.75%) and Gln110 (26.79%). This observation aligns with the binding interactions observed in crystal structures of TTBK1 with inhibitors like 2KC, VP7, VSY, and VNG2.73 (PDB: 4NFN, 7JXX, 7JXY, and 7Q8V), as confirmed by X-ray diffraction experiments (Halkina et al. 2021; Kiefer et al. 2014; Nozal 2022). Whereas, hit 2 system formed an average of 1–3 hydrogen bonds and are broken for 0–80 ns. At the end of the simulation (90 ns -100 ns), hit2 formed five hydrogen bonds (Fig. 6). The hydrogen bond was contributed by Gln77(88.18%), Leu175 (58.75%), Asp176 (11.65) and Ala61 (10.55%). The distinct characteristic of high number of consistent hydrogen bonds formation with Gln77 (100%) throughout the simulation time indicates that hit1 has a stronger ability to bind to TTBK1 than hit2 for effective inhibition.

Fig. 5.

Hydrogen bond analysis of TTBK1 with Reference compound, Hit1 and Hit2

Fig. 6.

Hydrogen bond occupancy of hits and reference

Conclusion

TTBK1 inhibitors hold promise as potential treatments for neurodegenerative diseases, particularly Alzheimer’s and Parkinson’s disease, by targeting the hyperphosphorylation of tau protein. In this study, Tanimoto similarity search was performed and 3884 analogs of already reported TTBK1 inhibitors were obtained to find the alternate chemical compound for TTBK1 inhibition. The series of steps such as Lipinski filtration, rank-based screening, ADMET filtration and exhaustive screening resulted in four hit compounds. Further two hits with key residue interaction, better physiochemical properties and high solubility were subjected to simulation. Hit 1 (CID70794974) followed a similar trend of RMSD, Rg and SASA in comparison to the reference. However, the insight into hydrogen bond count and occupancy showed Hit 1 has the potential and stable interaction with TTBK1 than reference and Hit 2. Furthermore, our simulation analysis reveals that Hit 1 forms stable hydrogen bonds with key residues of TTBK1, such as Glu77 and Gln110, which are critical for TTBK1 inhibition. Notably, these findings align with binding interactions observed in crystal structures of TTBK1 with known inhibitors, further validating the inhibitory potential of Hit 1. Although direct experimental validation through biophysical methods may pose challenges, the structural similarity and binding interactions observed in silico provide compelling evidence for the therapeutic efficacy of our hit compound against TTBK1. Further experimental studies will be valuable for validating these findings and elucidating the therapeutic efficacy of Hit 1 in the context of neurodegenerative diseases.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Study Conception and Design: KP, ES, MK, KK; Data collection, analysis and result interpretation: KP, ES, MK, KK; Manuscript preparation: KP, ES, MK, KK; Figures preparation: ES; Manuscript revision and editing: KK and RM. The final manuscript was verified by RM and approved for submission.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interest

The authors have no conflicts of interest to declare that are relevant to the content of this article.