Sage Journals: Discover world-class research

Abstract

Background

Drug repurposing offers a rapid, cost-effective approach for discovering therapies against multiple targets.

Objective

Here, we screen virtual ligand libraries consisting of 3468 approved drugs against 11 protein targets associated with Alzheimer's disease (AD).

Methods

We employ blind molecular docking, and target amyloid-β (Aβ), microtubule-associated protein tau (MAPT), Apolipoprotein E4 (APOE4), acetylcholinesterase (AChE), butyrylcholinesterase (BChE), amyloid-β protein precursor (AβPP), β-secretase (BACE1), brain-derived neurotrophic factor (BDNF), presenilin 1 (PSEN1) and 2 (PSEN2), and α-synuclein (SNCA) proteins using AutoDock Vina.

Results

Notably, multitarget binding recurs among the top-10 ligands with Ergotamine and Dihydroergotamine potentially binding 8; Dutasteride 7; Drospirenone and Nilotinib 6; Adapalene and Conivaptan 5; Bromocriptine 4; and Rolapitant, Irinotecan, Plerixafor, Saquinavir, and Telmisartan 3, out of 11 protein targets. As such, we reveal potential binding sites for ergot alkaloids, steroids, retinoids, antivirals, angiotensin receptor blockers, and Neurokinin 1 (NK1) receptor antagonists on multiple AD targets. Importantly, the therapeutic potential of the top-scoring ligands is confounded by pharmacokinetics and adverse-effects. For example, poor blood-brain barrier (BBB) penetration, and vasoconstriction, discount ergot-alkaloid use in AD. Likewise, potential toxicity limits prolonged use of steroids, Nilotinib, Adapalene, and Irinotecan. Conversely, BBB penetration, neuronal protection, oral availability, anti-inflammation, and anti-hypertension, admit Angiotensin receptor blockers (ARB), (Telmisartan/Candesartan); Antidiuretic hormone (ADH) inhibitors (Conivaptan/Tolvaptan); and of the NK1 receptors antagonists (Rolapitant/Netupitant) use in AD.

Conclusions

Our multitarget screening identifies selective synergistic AD modulators, such as ARB, ADH and NK1 receptor inhibitors, and simplifies drug discovery by focusing on the most promising candidates for experimental validation.

Keywords

Alzheimer’s disease angiotensin receptor blocker antidiuretic hormone inhibitors drug repurposing ergot alkaloid FDA molecular docking NK1 receptor steroid virtual screening

Introduction

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by the accumulation of amyloid-β (Aβ) and tau proteins in the brain.¹ AD is the most common chronic neurodegenerative disease worldwide, and the fifth leading cause of death in the adult population.² Despite the significant public health threat posed by AD, current therapeutic options are limited. Only a handful of drugs are approved by the Food and Drug Administration (FDA), and their efficacy primarily focuses on managing symptoms, rather than halting disease progression.³ Notably, treatment of comorbid conditions associated with AD has been linked with lower disease prevalence over the last decade.⁴ Comorbid conditions linked with AD include hypertension, pain, cardiovascular disease, dyslipidemia, atherosclerosis, and type 2 diabetes.⁵ Addressing these conditions has been shown to significantly improve the proteopathic hallmarks of AD.

Multiple proteins have been linked to AD pathology, and promote the complex etiology and progression of the disease.⁶ For example, Aβ peptides, central to the amyloid cascade hypothesis, are derived from the amyloid-β protein precursor (AβPP), through sequential cleavage by β-secretase (BACE1), and γ-secretase, the latter comprising presenilin 1 (PSEN1) and 2 (PSEN2) subunits.⁷ Oligomerization of Aβ is thought to initiate a cascade of events leading to neuronal dysfunction and death.⁸ In addition, microtubule-associated protein tau (MAPT), undergoes abnormal hyperphosphorylation in AD, leading to its dissociation from microtubules, and aggregation into neurofibrillary tangles.⁹ Aggregation of Aβ and tau has also been linked with depletion of brain-derived neurotrophic factor (BDNF), a protein associated with synaptic plasticity in learning and memory.¹⁰ One of the strongest risk factors of AD, is Apolipoprotein E4 (APOE4), an allele believed to promote BBB dysfunction, Aβ aggregation, and tau phosphorylation.^11,12 In turn, Aβ accumulation, tau hyperphosphorylation, and APOE4 allele have been linked with high levels of the presynaptic protein α-synuclein (SNCA) in the cerebrospinal fluid (CSF).¹³ Finally, acetylcholinesterase (AChE) and butyrylcholinesterase (BChE) are reported to promote the formation of amyloid fibrils and tau hyperphosphorylation.¹⁴ These proteins offer a rich array of targets for screening potential AD drugs.

Computational methods play a key role in modern drug discovery and significantly accelerate the process by providing powerful tools to enhance various stages. They enable virtual high-throughput screening of chemical libraries, significantly reducing the time and cost associated with experimental screening.¹⁵ Virtual screening along with computational ADME evaluation techniques helps prioritize compounds for experimental testing. In the past, we have used such techniques to identify potential MCL1 inhibitors in a virtual library containing FDA-approved drugs.¹⁶ Notably, these predictions were confirmed experimentally in a later study aimed at drug repurposing.¹⁷

Drug repurposing is an innovative strategy that identifies new therapeutic applications for existing drugs. This approach offers several advantages over traditional drug discovery as it reduces time and cost of drug development process. More importantly, the approach reduces the risk of drug failure in later clinical stages as FDA-approved drugs being repurposed have already passed safety and toxicity studies.¹⁸ To further increase success rate, multitarget virtual screening approaches have been developed. For example, Iqbal et al. neatly identified natural-like compounds that bind multiple protein targets associated with AD such as BACE, AChE, BChE MAOs, GSK3β, and NMDA.^19–21 This approach supports our notion that AD is a spectrum of disorders rather than a single disease.²² This approach has also led to several studies targeting proteins less commonly associated with AD, such as ROCK2,²³ CALHM1,²⁴ and Arginase.²⁵ Recently, two papers describing the potential repurposing of the FDA-approved drugs ergotamine, dihydroergotamine, and bromocriptine for the treatment of AD were published.^26,27 Unfortunately, these publications ignore the adverse-effects of ergot alkaloids, such as vasoconstriction, and their potential aggravation of hypertension, a frequent AD comorbidity. The studies even neglect basic pharmacological properties such as blood-brain barrier (BBB) penetration. These studies encouraged us to perform our own analysis.

Here, we virtually screen for multitarget drugs with repurposing potential in the treatment of AD. In particular, we use molecular docking to identify ligands in approved drug libraries that potentially bind multiple AD protein-targets.

Methods

Ligand preparation

To prepare virtual ligand libraries for molecular docking, a dataset containing 1379 drugs approved by the FDA was downloaded from the Zinc database (zinc.docking.org/substances/subsets/fda). A separate dataset including 2068 drugs approved by other drug-agencies of the world was also downloaded (zinc.docking.org/substances/subsets/world-not-fda). Taken together, these libraries encompassed 3447 drugs currently approved worldwide. All library ligands were prepared following standard protocols, and involved adding polar hydrogens, optimizing geometry, assigning rotatable bonds, adding Gasteiger charges, and converting the files to PDBQT format for virtual screening with AutoDock Vina.²⁸ Ligand preparation was performed using Open Babel 3.1.1 with a Python script to automate the process.²⁹

Target preparation

To prepare targets for molecular docking, the 3D structures of 11 proteins associated with AD were obtained from the Protein Data Bank (https://www.rcsb.org/). These proteins included Aβ (PDB ID 2LFM:A), AChE (PDB ID 1C2B:A), APOE4 (PDB ID 6NCO:A), AβPP (PDB ID 2FKL:A and 2FJZ:A), BACE (PDB ID 2G94:A), BChE (PDB ID 4TPK:A), BDNF (PDB ID 1BND:A), MAPT (PDB ID 5E2W:P), PSEN1 (PDB ID 7Y5T:B), PSEN2 (PDB ID 7Y5X:B) and SNCA (PDB ID 1XQ8:A). All protein targets contained the human sequence of amino acid, except for AChE which contained that of Electrophorus electricus with 90% and 94% sequence identity and homology, respectively. To remove native ligands, non-standard residues, and non-target chains, visualization tools such as PyMol(Schrödinger), and ChimeraX were used.³⁰ To remove water molecules, add polar hydrogens and Kollman charges, and convert the files to PDBQT formats, AutoDockTools 1.5.7 was utilized.²⁸

Grid-box settings

To support blind docking, the entire volume of the target protein was encompassed within the grid-box, with an additional 8–10 Å added in all directions. This measure accounted for potential ligand binding anywhere on the protein surface, or within it. Table 1 summarizes the grid-box parameters used with AutoDockTools 1.5.7 for virtual screening with the FDA-library. Similarly, Supplemental Table 1 shows the grid-box parameters used for virtual screening with the world-library. The two grid-boxes were centered slightly differently, but both encompassed the entire protein volume for blind docking.

Table 1.

AutoDock grid-box settings of protein targets used with the FDA-library.

Protein	PDB ID	Grid Centre (Å)			Grid Size (Å)
Protein	PDB ID	X	Y	Z	X	Y	Z
Aβ	2lfm	3.130	−7.037	−17.411	80	124	90
AChE	1c2b	21.608	83.202	20.482	126	126	126
APOE4	6nco	−5.532	−12.714	18.786	126	90	126
AβPP	2fkl	12.632	2.957	−0.370	82	110	90
BACE	2g94	2.326	−0.044	37.697	126	126	126
BChE	4tpk	−3.730	10.322	17.222	108	124	102
BDNF	1bnd	16.045	−4.278	44.803	122	78	126
MAPT	5e2w	32.858	−0.762	−7.034	60	35	30
PSEN1	7y5t	170.826	173.948	157.618	118	126	126
PSEN2	7y5x	109.460	108.919	92.307	90	126	126
SNCA	1xq8	240.749	70.530	−13.091	70	126	112

Molecular docking

To virtually screen ligand-libraries against AD protein-targets, molecular docking was performed with AutoDock Vina 1.1.2.²⁸ To automate the screening process, an in-house Python script was employed. All prepared ligands were docked against the 11 AD protein targets, and between 9 and 10 poses were calculated. For each ligand, the pose with the lowest binding energy was picked as the best bound ligand. For each target, the 10 best bound ligands, with the lowest binding energy, were taken as the top-10 ligands (of a target). To compare the binding sites and modes of interaction, the top 10 ligands were superimposed and visualized with PyMol. The top 10 ligands were screened for recurrence across targets, and the most frequently occurring ligands, that target multiple proteins, were considered multitarget ligands.

Heatmap analysis

To identify drug sets targeting the exact same combination of proteins, we performed a heatmap analysis on the multitarget ligands, focusing on protein sets that included at least 3 proteins, and shared at least 2 drugs. This classified the top drugs that target the same subset of proteins, and vice versa.

Interaction analysis

To better understand the binding mechanism of drug-ligands to target-proteins, we employed the Protein-Ligand Interaction Profiler (PLIP),³¹ an automated web-based tool (https://plip-tool.biotec.tu-dresden.de/plip-web/plip/index). PLIP analyzes protein-ligand interactions based on the PDB file format of protein-ligand complexes generated using PyMol. The tool identified and characterized the interaction types as well as the involved residues.

MD simulations

Heavy-atom MD simulation of protein-ligand complex of AChE (PDB ID 1C2B) with 13 top drug ligands was performed with NAMD (Nanoscale Molecular Dynamics program; version 3.0.1)³² using CHARMM36 forcefield (toppar_c36_jul19.tgz).³³ To generate protein parameter files and ligand parameter files VMD (Visual molecular dynamics; version 2.0.0),³⁴ and CHARMM-GUI Ligand Reader and Modeler module³⁵ were used, respectively. Solvation and ionization of the protein-ligand complex were performed using VMD, and a standard TIP3 water model was added as with periodic boundary conditions, padded by 10 Å in each dimension.³⁶ Na + and Cl− ions were used to neutralize the charge. The CgenFF version of CHARMM-GUI³⁷ was used to generate topology and parameter files for the 13 ligands. The particle-mesh Ewald method was used for electrostatic interactions of the system.³⁸ Minimization was performed for 1000 steps at a constant temperature of 298 K, and production was run using NPT (constant particle number, pressure, and temperature). Thirteen sets of simulations were performed for 50 ns each. RMSD trajectory plugin tool in VMD was used to calculate the RMSD of protein and ligand.

Pharmacokinetic analysis

To evaluate the pharmacological potential of drugs in the treatment of AD, the ability to penetrate the BBB is critical. Therefore, we predicted the BBB permeability, in addition to other parameters, such as volume of distribution at steady state (VD_SS), and maximum daily tolerated dose (MTD), for select top multitarget ligands using pqCSM with their SMILES annotation (https://biosig.lab.uq.edu.au/pkcsm/prediction).³⁹

Results

Molecular docking

In this study, we screen two virtual libraries corresponding to 1379 and 2068 drugs, approved by the FDA, and by other agencies of the world, respectively (Total of 3447 drugs). Ligand preparation increases the number of ligand isoforms to 1400 and to 2115 in the FDA and world-library respectively (Total of 3515 ligands). Ligands are screened using blind molecular docking against 11 protein-targets highly associated with AD, namely Aβ, AChE, APOE4, AβPP, BACE, BChE, BDNF, MAPT, PSEN1, PSEN2, and SNCA. Figure 1 shows the top 10 ligands of the FDA-library docked to 11 protein targets, with the binding energies obtained from AutoDock Vina, and the chemical structures of each ligand. (Supplemental Figure 2 shows the top 10 ligands of the world-library.) Notably, our blind docking simulations result in ligand clusters located in the major protein ligand-binding site. The consistent ligand localization across most proteins underscores the robustness of the docking simulations and highlights the presence of well-defined and energetically favorable binding pockets. Notable exceptions are observed for protein targets, such as AβPP which demonstrate a contiguous diverging binding groove across its surface, and PSEN1 and PSEN2 which exhibit 3 major binding regions with distinct binding sites. This divergence suggests the presence of multiple druggable pockets within these proteins.

Figure 1.

Molecular docking of FDA-approved drug-ligands onto protein-targets linked with Alzheimer's disease (AD). Shown are the top 10 ligands docked to (A) Aβ (B) AChE (C) APOE4, (D) AβPP, (E) BACE, (F) BChE, (G) BDNF, (H) MAPT, (I) PSEN1, (J) PSEN2 and (K) SNCA. For each complex, the protein-target is represented as a grey surface, and the drug-ligands are represented as sticks. Around each complex is a clockwise display of the top 10 drug-ligands ordered by binding affinity. The display includes chemical structure, generic name, and binding energy of each drug ligand. The figure was prepared using PyMol (Schrödinger).

The top 10 ligands of the cholinesterases potentially bind very strongly to AChE (ΔG between −11.4 and −10.2 kcal/mol) and BChE (ΔG between −12.3 and −11.2 kcal/mol). The drugs bind at an allosteric site, at the opening of the gorge leading into the catalytic site (Supplemental Figure 1, A2, A5). The drugs occlude access to the active site and prevents catalytic breakdown of the native ligand acetylcholine, thus potentially reducing cholinergic dysfunction associated with AD.^40,41 The top 10 ligands of APOE4 potentially form strong interactions with the protein (ΔG between −9.4 and −8.6 kcal/mol) inside a pocket lined with hydrophilic residues W34 and L149 of the LDL receptor binding region (Supplemental Figure 1, A3, D2, F1). The drugs occupy the same pocket as another APOE4 stabilizer, which was designed using fragment-based discovery.⁴² As such, the top 10 drugs could stabilize the APOE4 allele variant, which has the most damaging effects on neurons.⁴³ The top 10 ligands of BACE1 potentially bind strongly to the protein (ΔG between −10.4 and −9.3 kcal/mol), and form electrostatic interactions and hydrogen bonds with residues Y71, Q73, I105, and I118, at the entrance of the active site domain (Supplemental Figure 1, C2). As such, the drugs prevent access of native ligands to catalytic site residues D32 and D228 of BACE1, and potentially prevent AβPP cleavage by β-secretase.⁴⁴ The top 10 ligands of PSEN1 potentially bind strongly to the protein (ΔG between −10.9 and −9.8 kcal/mol) and form interactions with residues lining the active site pocket, thus inhibiting catalysis of native ligands, except for Indocyanine Green, Olaparib, and Ciclesonide. For instance, Dutasteride, Ergotamine, and Dihydroergotamine form interactions with V77, V272, I282 (Supplemental Figure 1, C4, B4, A6), and thus potentially allosterically prevent AβPP cleavage by γ-secretase.⁴⁵ The top 10 ligands of PSEN2 potentially bind strongly to the protein (ΔG between −9.8 and −8.6 kcal/mol), and form interactions with residues lining the active site pocket for native ligands, except for Indocyanine Green, Netupitant, and Tolvaptan. For example, Dutasteride, Drospirenone, and Dihydroergotamine form interactions residues, L179 and I235 (Supplemental Figure 1, B5, C5, D5) and potentially prevent AβPP cleavage by γ-secretase.⁴⁶ The fact that Bromocriptine inhibits both PSEN1 and PSEN2, the major components of γ-secretase, agrees with earlier experiments showing that Bromocriptine reduces amyloid-β production in cell lines.⁴⁷ The top 10 ligands of SNCA potentially bind to the protein (ΔG between −8.3 and 7.2 kcal/mol), and form multiple interactions with α-synuclein. The native monomeric form of α-synuclein is intrinsically disordered, and the ligands potentially stabilize monomeric conformations less prone to aggregation. While α-synuclein is more associated to Parkinson's disease, than to AD, it also co-aggregates with Aβ.⁴⁸ The drugs bind to α-synuclein, in various positions and potentially solubilize it, and prevent aggregation. The top 10 ligands of BDNF potentially binds near the interface of its dimeric form (ΔG between −8.9 and −7.9 kcal/mol), and potentially stabilize this construct, which binds to TrkB tyrosine kinase receptor, to improve neuronal plasticity.⁴⁹ The ligands do not bind to BDNF domain interacting with its receptor, and potentially do not interfere with beneficial receptor activation. The top 10 ligands of MAPT potentially bind weakly to the peptide (ΔG between −4.4 and −4.0 kcal/mol) and form interactions with the phosphorylated residues of MAPT, potentially stabilizing the monomeric state and preventing tau aggregation.⁵⁰ The top 10 ligands of AβPP potentially bind weakly (ΔG between −7.5 and −6.7 kcal/mol) to this copper-binding domain, consisting of residues 133–189.⁵¹ The top 10 ligands insert into a contiguous curved groove on its surface. At one end of the binding groove, Dihydroergotamine and Drospirenone form interactions with residues V129, and F135 (Supplemental Figure 1, B2 and D2). At another end of the binding groove, Ergotamine forms contacts with D167 and M170 (Supplemental Figure 1, A4). These binding modes potentially stabilize the native isoform of the copper-binding domain of AβPP and potentially prevent unfolding and cleavage. The low-range affinity of ligands bound to MAPT and AβPP, express their low significance in our study of druggable AD targets.

In the past, our docking methodology has been validated through high correlation between predicted and experimental binding energies.^16,52,53 Here too, predicted and experimental binding energies are closely related. For example, the experimental binding constant of irinotecan to AChE is Ki = 51 nM.⁵⁴ Using the equation Gibbs free energy equation ΔG° = −RT ln K, with a temperature of 298°K, and the gas constant 1.9872 × 10⁻³, the experimental binding energy corresponds to ΔG°_Experimental = −9.77 Kcal/mol. Remarkably, the experimental binding energy is close to that predicted by Vina, ΔG°_Predicted = −9.6 Kcal/mol. The close relationship between predicted and experimental binding energy attests to the strength of our molecular docking methodology.

Frequency analysis

A histogram of the top 10 FDA-ligands is shown in Figure 2. Notably, the top 10 ligands exhibit a significant trend towards multitarget interactions. Twenty-three ligands target 2 or more proteins (Figure 2A), while thirteen ligands target 3 or more proteins (Figure 2B). Collectively, these thirteen ligands target all the 11 proteins studied. Among the multitarget ligands, two of them, Ergotamine and Dihydroergotamine, each target 8 out of the 11 proteins. Together, the pair targets 10 of the 11 proteins, demonstrating a remarkable versatility in their binding capabilities. Following closely behind is Dutasteride, targeting 7 proteins, then Drospirenone and Nilotinib, each targeting 6 proteins, and then Adapalene and Conivaptan, each targeting 5 proteins. The multitarget ligand similarities is remarkable and reveals a common denominator previously unknown. The multitarget ligand similarities suggest either: (A) A convergent evolution of 11 protein targets associated with AD. Or (B) An almost complete lack of selectivity of ligands recurring in 4 or more targets (i.e., false positives). Option B is more plausible than option A based on the principle of Occam's razor, and the dilemma awaits experimental resolution. Consequently, the multitarget ligands are double-edged swords, with the potential to act in synergy, or in complete disarray; and to either drive, or mitigate AD progression.

Figure 2.

Histograms of top 10 drugs targeting proteins linked to Alzheimer’s disease (AD). Shown are (A) the number of proteins targeted by 23 of the top 10 drugs, and (B) the number of top 10 drug-ligands accommodated by the 11 protein-targets. Included in A are only 23 drugs that target two or more proteins, and in B only 13 drugs that target three or more proteins.

Interestingly, drugs that target Aβ are also most recurrent, and consist of 8 multitarget ligands shared with multiple protein targets (Figure 2B). This highlights the importance of Aβ as a representative AD target, and potentially suggests that drugs bound to Aβ, also bind other AD proteins. Similarly, drugs that target AβPP, BChE, PSEN2, and SNCA, consist of 7 multitarget ligands shared with several other proteins (Figure 2B). AβPP shares 7 of its top 10 ligands with other protein-targets, but the absence of a well-defined non-contiguous binding site, and its average affinity for ligands preclude any meaningful binding (Figure 2B). As such, we consider Ergotamine and Dihydroergotamine, drugs bound to AβPP as unselective due to their low binding affinity. Both target AβPP and interact with most AD proteins, including those with the fewest common drugs, such as APOE4 and BACE (Figure 2B). This highlights the ergot alkaloids’ structure as the most common denominator shared by the targets, and the least selective multitarget drugs among the top 10 ligands. PSEN2, a subunit of the γ-secretase complex responsible for the final cleavage of AβPP to produce Aβ, was also among the most frequently targeted proteins, but ligands bind with high affinity, and in one 3 well defined binding sites. A significant number of top 10 ligands also strongly target BChE, responsible for the breakdown of acetylcholine, a neurotransmitter central to the cholinergic dysfunction hypothesis.⁵⁵ Cholinesterase inhibition is expected to compensate for the cholinergic dysfunction by rescuing available acetylcholine.⁵⁶ Finally, APOE4, widely recognized as the strongest genetic risk factor for AD,¹¹ was also targeted by Ergotamine, an ergot alkaloid. Our multi-target approach is particularly relevant in a complex disorders like AD, where multiple pathways can co-contribute to disease progression.

Heatmap analysis

To better understand the interaction patterns of protein-targets and drug-ligands, a heatmap analysis is rendered in Figure 3. To screen for significant patterns, the heatmap includes only 11 proteins sharing 2 or more drugs, and 13 drugs targeting 3 or more proteins. Accordingly, the heatmap identifies robust multi-target interaction patterns, and produces a total of 26 union-sets, of 3, 4, 5, or 6 proteins sharing 2, 3 or 4 drugs. Of the 26 protein-sets, 3 union sets contain 6 proteins that share 2 drugs, namely: (1) Ergotamine and Dihydroergotamine, (2) Nilotinib Dihydroergotamine, and (3) Dutasteride and Dihydroergotamine. Following closely behind are 2 union sets containing 5 proteins that share 2 drugs; 9 union sets comprising 4 proteins that share 2–3 drugs; and 12 union sets containing 3 proteins that share 2–4 drugs.

Figure 3.

Heatmap of interactions. Shown are the union sets of proteins sharing bound drugs. The number of shared drugs is indicated by the red color intensity.

Interaction analysis

To analyze the structural differences among the best hits, a drug-protein interaction analysis is shown in Supplemental Figure 1 for selected targets and top-ligands to understand the contributing factors to binding energy. These interactions are important and provide insight into the binding mechanism. Various interactions are identified, including hydrogen-bonds, salt bridge, π-cation, π-stacking, and hydrophobic interactions. The drugs show higher binding affinity (i.e., lower binding energy) for the cholinesterases than for other protein targets. The high affinity associated with cholinesterase ligands is primarily due to the deep gorge of the enzyme active site, lined with aromatic residues which form an electrostatic potential pit with multiple binding interactions. BChE has a larger active site gorge, resulting in a more accommodating binding site. As shown Supplemental Figure 1 G4, the interaction profile of Drospirenone with PSEN1 and PSEN2 shows a high variety of interactions, both hydrogen-bonds and hydrophobic interactions. These examples clearly illustrate the contribution of buried hydrophobic surface to binding energy. This is also shown in the case of Dutasteride's equal affinity for BACE and PSEN1 (Supplemental Figure 1 G3), with an even number of hydrophobic interactions, but uneven amount of hydrogen bond, and salt bridges.

MD simulations

Molecular dynamics simulations up to 50 ns were carried out, to assay the top drug-ligand interactions with AChE. Hydrogen bond formation, solvent accessible surface, gyration radius, RMSD, and RMSF analyses of protein-ligand complexes were assessed to evaluate stability (Figure 4 and Supplemental Figure 3). Remarkably, all the ligands, Ergotamine, Dihydroergotamine, Dutasteride, Drospirenone, Nilotinib, Adapalene, Conivaptan, Bromocriptine, Rolapitant, Irinotecan, Plerixafor, Saquinavir, and Telmisartan remained in their binding sites for 50 ns. Remarkably, the top AChE ligands, Adapalene, Ergotamine, Drospirenone, Rolapitant, and Dutasteride, deviated less than 2 Angstrom in the binding site for the entire length of the simulation (50 ns). In contrast, Bromocriptine exhibited the most variation in the AChE binding site. Thus, our predictions should be taken with a grain of salt, and they require experimental validation.

Figure 4.

MD simulation of drugs bound to AChE. Shown are RMSD values of the top 13 drugs bound to AChE during the MD simulations. Ligand RMSD is displayed using different colors.

Pharmacokinetics

The primary goal of pharmacology is to simultaneously obtain a maximal therapeutic effect, and a minimal toxic effect. Predicted pharmacokinetics are shown in Table 2, and include BBB permeability, MTD, and VD_SS. The predicted pharmacokinetic values are based on theoretical calculations, and do not replace experimental values. The predicted values are shown solely for easy comparison. BBB permeability is a measure of brain absorption, and is of great importance in AD. A positive LogBB value indicates good BBB penetration, while a negative one suggests poor brain accessibility. The MTD is another measure of safety. High MTD values indicate that high drug doses are safe and can help achieve therapeutic concentrations at the brain target site, in case of low BBB permeability. Finally, the VD_SS is a measure of drug distribution. Low VD_SS is indicative of strong plasma binding, while high VD_SS suggests extensive distribution into tissues. For drugs with low BBB penetration, low VD_SS can help maintain higher plasma concentrations, potentially enhancing delivery to the brain.

Table 2.

Predicted pharmacokinetics of top drugs.

Drug	BBB* (Log BB)	MTD^† (Log mg/kg/day)	VD_SS^# (Log L/kg)
Rolapitant	0.137	−0.47	0.983
Drospirenone	0.033	−0.019	0.065
Telmisartan	−0.072	0.42	0.843
Dutasteride	−0.222	−0.979	0.052
Plerixafor	−0.239	0.711	0.985
Adapalene	−0.334	0.338	−1.588
Ergotamine	−0.575	−0.327	1.426
Dihydroergotamine	−0.589	−0.327	1.428
Nilotinib	−0.848	0.293	−0.032
Bromocriptine	−0.897	−0.631	0.733
Conivaptan	−0.926	0.383	−0.275
Saquinavir	−0.939	−0.533	0.51
Irinotecan	−1.053	−0.039	1.416

*Blood-Brain Barrier Penetration.

^†

Maximum Tolerated Dose.

Volume of Distribution at Steady State Concentration.

Despite the strong binding of Ergotamine, and Dihydroergotamine to multiple AD targets, their modest brain penetration severely limits their clinical use. Theoretically, ergotamine and dihydroergotamine can cross into the brain of AD, in which BBB integrity is compromised. Yet, even if they cross, Ergotamine and Dihydroergotamine have sustained vasoconstrictor actions, and are contraindicated in patients with coronary, cerebral, and peripheral vascular disease, thus potentially worsening AD symptoms. It is surprising that Ergotamine has been recommended as potential treatment of AD.²⁶ Like Ergotamine and Dihydroergotamine, Bromocriptine poorly penetrates the BBB, and it is surprising, therefore, that these ergot alkaloids have been suggested as a potential treatment of AD.²⁷ Likewise, other top candidates could also be counterproductive. Irinotecan is a toxic chemotherapeutic agent, and not recommended for AD patients. Saquinavir is a HIV-1 protease inhibitor that fails to penetrate the BBB. Adapalene is synthetic retinoid used topically in the treatment of acne; however, the oral route presents high toxicity, and could also be unrecommended for AD. As such, top-ranking candidates may not necessarily be effective in the treatment of Alzheimer's disease (AD), highlighting the need for careful evaluation and critical analysis.

Discussion

In this study, we virtually screen a library of approved drugs against 11 AD targets using molecular docking. Then, we analyze the top 10 drugs for each target for recurrence. Top drugs that bind 3 or more AD targets are: (1) Ergotamine, (2) Dihydroergotamine, (3) Dutasteride, (4) Drospirenone, (5) Nilotinib, (6) Adapalene, (7) Conivaptan, (8) Bromocriptine, (9) Rolapitant, (10) Irinotecan, (11) Plerixafor, (12) Telmisartan, and (13) Saquinavir (Figure 2A). These drugs can be classified broadly into ergot alkaloids, steroid hormones, retinoids, NK1 receptor antagonists, antivirals, antidiuretic hormone (ADH) inhibitors, and angiotensin receptor blockers (ARB) as discussed henceforth. The ergot alkaloids, and steroid hormones are indiscriminate binders likely driving the disease, and unfit for AD repurposing, as discussed hereafter. The remaining drugs potentially halt disease progression, but Nilotinib, Adapalene, Irinotecan, and Saquinavir suffer from high toxicity or BBB impermeability, and are bad candidates for AD repurposing. Good candidates include ARBs (i.e., Telmisartan and Candesartan), NK1 receptor antagonists (i.e., Rolapitant and Netupitant), and ADH inhibitors (i.e., Conivaptan and Tolvaptan). The good candidates readily cross the BBB and/or protect its integrity, and exhibit low toxicities. Interestingly, the good candidates reduce hypertension, a comorbid trait associated with AD, by blocking angiotensin II response that is also involved in learning and memory.^57–60 The good candidates also reduce inflammation and are orally available.

Ergotamine, Dihydroergotamine, and Bromocriptine (Ergot alkaloids)

Among the top 10 ligands, the ergot alkaloids, Ergotamine, Dihydroergotamine, and Bromocriptine, emerge as highly recurrent in both frequency distribution and heatmap analyses. The ergot alkaloids are dopaminergic receptor agonist associated with vasoconstriction, and can exacerbate hypertension associated with AD.⁶¹ In addition, these drugs poorly cross the BBB, which is a critical requirement for treating this central nervous system (CNS) disease.⁶² While the CNS effect of Ergotamine, and Dihydroergotamine in AD patients is limited due to poor brain penetration, it is expected to produce vasoconstriction with adverse effects such as ischemic hypoxia (i.e., ergotic ‘brain gangrene’) in AD patients with compromised BBB.⁶³ Overall, the use of ergot-alkaloids, such as Hydergine, an Ergoloid Mesylate, have shown no long-term benefit in the treatment of dementia.^64,65 On the opposite, epidemiological studies link AD with migraine, and the use of ergot alkaloids.⁶⁶ As such, iatrogenic AD could be linked with repeated administration of ergot alkaloid in migraine patients suffering from compromised BBB. AD has also been linked to fungal infections with Aspergillus spp., that produce ergot mycotoxins.^67,68 As such, sporadic AD can be linked with prolonged exposure to ergot-alkaloids found in airborne fungal spores, that cross compromised BBB or bypass it entirely in the nasal cavity.⁶⁹ To add insult to injury, Ergot alkaloids also exhibit anti-prolactin effects, and suppress proliferation of oligodendrocyte precursor cells in the CNS.⁷⁰ These findings all question the recent proposals to repurpose ergot alkaloids as a potential drug against AD.^26,27 Notably, the clinical trials using multiple ergoloid mesylates,^64,65 do not support assertions that dihydroergocristine mesylate^71,72 and Bromocriptine mesylate⁷² attenuate AD pathologies in mice. As such, the ergot alkaloids, which together potentially bind all 11 AD protein targets (i.e., Aβ, AChE, APOE4, AβPP, BACE, BChE, BDNF, MAPT, PSEN1, PSEN2, and SNCA), do not halt AD progression, but rather drive it.

Dutasteride and drosperinone (steroids)

The binding sites can also accommodate steroid hormones, and the top 10 ligands include Dutasteride and Drospirenone with high recurrence. Dutasteride is a 5α-reductase inhibitor, and Drospirenone is a synthetic progesterone. These sterols readily cross the BBB; however, clinical findings associate them with an increased risk of AD. Notably, men taking dutasteride, or finasteride, were at increased risk of dementia (HR 1.10) in general, and of AD (HR 1.28) in particular.⁷³ Persistently, AD has also been associated with BPH (HR 1.10–1.20), and 5α-reductase inhibitor users, compared with the general population⁷⁴ As such, iatrogenic AD could be linked to prolonged administration of dutasteride in men diagnosed with BPH.⁷³ Drospirenone is a synthetic progesterone used in combined hormonal contraception, and in postmenopausal hormone replacement therapy. Both estrogen (Figure 1C) and progesterone (Figure 1B-D, F, J-K) independently and interactively modulate AD neuropathology.⁷⁵ When co-administered, progesterone blocks the beneficial effect of estrogen on Aβ accumulation in ovariectomized mice. Also, co-administration of Drospirenone and estrogen fail at improving working memory of ovariectomized rats⁷⁶ AD has also been linked with postmenopausal hormone therapy, both with estrogen (HR 1.05–1.14) and with estrogen-progestogen combinations (HR 1.13–1.21).⁷⁷ As such, iatrogenic AD could be linked to administration of (combined) hormone replacement therapies in postmenopausal women. Finally, Drospirenone-containing pills are contraindicated in patients susceptible to venous thromboembolism, due to a high risk of blood-clots in the arms or legs, lung, heart, and brain, compared to other progestin-containing pills.⁷⁸ As such, steroid hormones which potentially bind 6–7 out of the 11 AD protein targets (Dutasteride: Aβ, BDNF, PSEN1, PSEN2, SNCA; and Drospirenone: AChE, BChE APOE4, AβPP, PSEN2, SNCA; Estradiol: APOE4), do not halt AD progression, but rather drive it.

Nilotinib

Nilotinib is a kinase inhibitor used for the treatment of chronic myeloid leukemia with Philadelphia chromosome.⁷⁹ It penetrates the BBB, and several studies highlight the potential of Nilotinib in the treatment of AD. Notably, Nilotinib has been shown to reduce Aβ burden,⁸⁰ hippocampal volume loss,⁸¹ and inflammatory markers.⁸² Moreover, Nilotinib has demonstrated the ability to restore synaptic dysfunction in human embryonic stem cells.⁸³ Finally, one less favorable study found Nilotinib to significantly increase BACE1 protein expression, without affecting that of AβPP.⁸⁴ Nilotinib has been registered for phase III clinical trial by KeifeRx in December 2021. Yet, to the best of our knowledge, the trial has not started, and nilotinib for AD treatment no longer appears in the company pipeline.⁸⁵ Nilotinib safety in AD patients has been questioned, with major concerns expressed about ca,rdiovascular side effects (i.e., arrhythmias, hypertension, and heart failure), which can compound AD comorbidities.⁸⁶ As such, Nilotinib potentially binds 6 of the 11 AD protein targets (i.e., Aβ, AβPP, BDNF, MAPT, PSEN2, and SNCA), partially halting disease progression; however, major adverse effects limit its use.

Adapalene (retinoids)

Adapalene is a third-generation topical retinoid approved for the treatment of acne and cervical neoplasia. Like other retinoids, adapalene penetrates the BBB, which have shown promising potential in AD treatment.⁸⁷ Notably, Adapalene completely inhibits Aβ aggregation.⁸⁸ Adapalene's neuroprotective effects against oxidative stress in vitro remain debated, with one study supporting its efficacy⁸⁹ and another refuting it.⁹⁰ However, Adapalene topical administration has been linked with intracranial hypertension,⁹¹ and oral administration is contraindicated due to safety concerns.⁹² As such, Adapalene potentially binds 5 of the 11 AD protein targets, Aβ, AChE, AβPP, BDNF, and SNCA, indicating its potential to partially modulate disease progression. However, oral toxicity critically limits Adapalene repurposing for AD treatment.

Irinotecan

Irinotecan is a topoisomerase inhibitor, used in the treatment of cancers, like metastatic colorectal carcinoma and pancreatic adenocarcinoma. Irinotecan crosses BBB readily but is significantly toxic. The drug's mechanism of action involves DNA damage and cell death, which is beneficial for cancer treatment, but harmful for neurodegenerative diseases like AD.⁹³ Despite potential binding to AβPP, PSEN2, and SNCA, Irinotecan is not a candidate for AD repurposing.

Saquinavir, Indinavir, and Daclastavir (Antivirals)

Saquinavir and Indinavir are a HIV protease inhibitor used for the treatment of HIV infection. Likewise, Daclastavir antiviral drug that targets the non-structural protein 5A (NS5A) used in the treatment of hepatitis C virus infection. These drugs are ineffective at treating viral infection in the brain because they do not cross the BBB.⁹⁴ Saquinavir, potentially targeting Aβ, BACE, BChE, PSEN1; Indinavir targeting PSEN1; and Daclastavir targeting BDNF, are not candidates for drug repurposing in AD, unless administered intranasally to bypass the BBB.

Plerixafor

Plerixafor is a CXCR4 antagonist used in combination with granulocyte-colony stimulating factor to mobilize hematopoietic stem cells to the peripheral blood. Preclinical studies have shown that plerixafor can cross the BBB,⁹⁵ and antagonize CXCR4 expression which is significantly increased in AD patients.⁹⁶ Moreover, subcutaneous administration of Plerixafor in a mouse model of AD improved cognition, and reduced neuroinflammation.⁹⁷ As such, Plerixafor potentially binds Aβ, MAPT, and AβPP, and it is a candidate for repurposing in AD.⁹⁸ Notably, Plerixafor is orally unavailable, which could limit its use.

Conivaptan and Tolvaptan (antidiuretic hormone inhibitor)

Conivaptan and Tolvaptan are ADH inhibitors, used to raise serum sodium levels in euvolemic and hypervolemic hyponatremic patients. Conivaptan exhibits unfavorable BBB penetration, but exerts neuroprotective effects against oxidative stress mechanisms in AD.⁹⁰ Conivaptan potentially reduces aggregation in model systems by inhibiting Hexokinase-1 and 14–3-3G/γ protein interaction.⁹⁹ Notably, conivaptan attenuates cerebral edema and BBB disruption following cardiac arrest,¹⁰⁰ and it reduces blood cytokine levels in rat models of autoimmune encephalomyelitis.¹⁰¹ Tolvaptan also improves psychomotor speed domain, and is less studied than Conivaptan.¹⁰² Conivaptan is not without it risks either, of headache, and infusion site reactions, and too rapid correction of hyponatremia can cause fatal osmotic demyelination syndrome.¹⁰³ Finally, Conivaptan is readily available orally.¹⁰⁴ As such, both ADH inhibitors: (1) Conivaptan which potentially binds to Aβ, BChE, BDNF, MAPT, and PSEN2, and (2) Tolvaptan which potentially binds to BDNF and PSEN2, are good candidates for drug repurposing in AD.

Rolapitant and Netupitant (NK1 receptor antagonists)

Rolapitant and Netupitant are selective neurokinin 1 (NK1) receptor antagonist, used in the treatment of chemotherapy-induced nausea and vomiting. Rolapitant readily crosses the BBB, and blocks the action of tachykinin peptides, such as substance P which plays an important role in neuroinflammation.¹⁰⁵ Rolapitant is also approved for treating pulmonary arterial hypertension, by improving blood flow and reducing hypoxemia. While the oral formulation of Rolapitant (i.e., Varubi) was withdrawn in 2018 due to reports of anaphylactic shock, Netupitant remains orally available. Rolapitant potentially binds to BDNF, BACE, and AChE, and Netupitant to PSEN2 and BChE, and as such NK1 receptor antagonists are good candidates for drug repurposing in AD.¹⁰⁶

Telmisartan, Candesartan, Azilsartan (Angiotensin receptor blockers)

Telmisartan and Candesartan are angiotensin II receptor blockers (ARB), used orally in the treatment of hypertension, diabetic nephropathy, and heart failure. Remarkably, ARB use is associated with a reduced incidence of cognitive impairment, dementia, and AD in white men and women.¹⁰⁷ Likewise, Telmisartan is associated with a lower risk of AD in African Americans,¹⁰⁸ and with a lower risk of dementia in East Asians suffering from diabetes and hypertension.¹⁰⁹ Notably, use of ARB, and to a lesser degree use of β-adrenoceptor blockers,¹⁰⁷ but not use of angiotensin converting enzyme (ACE) inhibitors,¹¹⁰ reduce AD incidence. In AD mice models, Telmisartan also alleviates AD-related neuropathologies and cognitive impairments,¹¹¹ and intranasal Telmisartan significantly reduced amyloid burden in the cortex and hippocampus.¹¹² While all ARBs penetrate the BBB, the highly lipophilic Telmisartan attains brain concentration 10-fold higher than those of Losartan and Irbesartan,¹¹³ and Candesartan restores BBB dysfunction.¹¹⁴ Several studies have highlighted disproportionate neuroprotective benefits of Telmisartan and Candesartan over other ARBs without understanding their mechanism. Here, we propose potential mechanisms through binding multiple protein targets associated with AD. Other potential mechanisms include high lipophilicity and BBB restoration through APOE4-binding.¹¹⁵ Telmisartan potentially targets APOE4, BChE, and MAPT, Candesartan targets BACE, and Azilsartan targets AChE, and as such these ARB are good candidates for drug repurposing in AD.

Limitations

As a potential limitation to this study, it is crucial to note that our findings are based on computational work. Therefore, the beneficial or deleterious effect of these drugs in AD treatment remains to be validated experimentally and clinically. Experimental validation is required to evaluate the binding affinity of top drugs to protein targets of AD. In addition, clinical validation is required to evaluate the overall effect of good candidates for repurposing.

As another potential limitation to this study, multitarget drugs do not necessarily indicate a better chance at AD repurposing. In fact, multitarget drugs are indicators of less discriminate selectivity, and of more potential toxicity. This could help explain why highly multitarget drugs, like ergot alkaloids, and steroids are not good candidates for AD repurposing. Conversely, drugs targeting only one AD protein could be too selective and restrain their repurposing potential. Thus, a balance between too much selectivity, and too little selectivity is required to identify good candidates. In fact, ergotamine has been proposed to cure everything from colon cancer,¹¹⁶ to arbovirus infection,¹¹⁷ and in-silico it binds to 24 proteins of the coronavirus,¹¹⁸ including coronavirus protease,¹¹⁹ and coronavirus helicase,¹²⁰ as well as human MCL-1,¹⁶ NRP1,¹²¹ sortilin,¹²² Histamine N-Methyl Transferase,¹²³ TGF-β receptor type 1 (ALK5),¹²⁴ casting immense doubt on the specificity of ergot alkaloids. The latter paper calls out the false positives. We believe ergot alkaloids are not specific enough for repurposing in AD.

Finally, as a limitation of this paper, the good candidates for drug repurposing in AD include the top drugs, but not automatically any other members of their drug families.

Conclusion

Good candidates for drug-repurposing in AD include: (1) The antidiuretic hormone inhibitor, Conivaptan and Tolvaptan; (2) The NK1 receptor antagonist, Rolapitant, and Netupitant; and (3) The angiotensin receptor blocker, Telmisartan, and Candesartan. Their selective multitarget potential, antihypertensive and anti-inflammatory properties, BBB penetration or protection, and oral availability underscores their suitability for further investigation.

Supplemental Material

sj-docx-1-alz-10.1177_13872877251387167 - Supplemental material for Virtual screening of drugs against multiple targets of Alzheimer's disease

Supplemental material, sj-docx-1-alz-10.1177_13872877251387167 for Virtual screening of drugs against multiple targets of Alzheimer's disease by Karin Ben Zaken, Samuel Mesfin, Naamah Bloch and Abraham O Samson in Journal of Alzheimer's Disease

Footnotes

Acknowledgements

We thank the Katz foundation for a generous grant to AOS.

ORCID iDs

Samuel Mesfin

Abraham O Samson .

Author contribution(s)

Karin Ben Zaken: Data curation; Formal analysis; Investigation; Methodology; Software; Validation; Visualization; Writing – original draft; Writing – review & editing.

Naamah Bloch: Investigation; Project administration; Resources; Supervision.

Abraham O Samson: Conceptualization; Funding acquisition; Project administration; Resources; Supervision; Writing – original draft; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant to AOS from the Katz family foundation, and the Ginsburg Family foundation.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

Supplemental figures and tables are available on the journal website.

Supplemental material

Supplemental material for this article is available online.

References

Polis

Samson

. A new perspective on Alzheimer’s disease as a brain expression of a complex metabolic disorder. In: Wisniewski

(ed) Alzheimer’s disease. Brisbane, Australia: Codon Publications, 2019, pp.1–22.

Alzheimer’s Association . 2022 Alzheimer’s disease facts and figures. Alzheimers Dement 2022; 18: 700–789.

Xiao

Zhang

. Current therapeutics for Alzheimer’s disease and clinical trials. Explor Neurosci 2024; 3: 255–271.

Drew

. An age-old story of dementia. Nature 2018; 559: S2–S3.

Avitan

Halperin

Saha

, et al.

Towards a consensus on Alzheimer’s disease comorbidity?

J Clin Med 2021; 10: 4360.

Zhu

Liu

Walker

, et al. Associations between genetically predicted plasma protein levels and Alzheimer’s disease risk: a study using genetic prediction models. Alzheimers Res Ther 2024; 16: 8.

Pichet Binette

Gaiteri

Wennström

, et al. Proteomic changes in Alzheimer’s disease associated with progressive Aβ plaque and tau tangle pathologies. Nat Neurosci 2024; 27: 1880–1891.

Askenazi

Kavanagh

Pires

, et al. Compilation of reported protein changes in the brain in Alzheimer’s disease. Nat Commun 2023; 14: 4466.

Strang

Golde

Giasson

. MAPT Mutations, tauopathy, and mechanisms of neurodegeneration. Lab Invest 2019; 99: 912–928.

10.

Gao

Zhang

Sterling

, et al. Brain-derived neurotrophic factor in Alzheimer’s disease and its pharmaceutical potential. Transl Neurodegener 2022; 11: 4.

11.

Huang

Y-WA

Zhou

Wernig

, et al. Apoe2, ApoE3, and ApoE4 differentially stimulate APP transcription and Aβ secretion. Cell 2017; 168: 427–441.e21.

12.

Montagne

Nation

Sagare

, et al. APOE4 Leads to blood–brain barrier dysfunction predicting cognitive decline. Nature 2020; 581: 71–76.

13.

Twohig

Nielsen

. α-synuclein in the pathophysiology of Alzheimer’s disease. Mol Neurodegener 2019; 14: 23.

14.

García-Ayllón

M-S

. Revisiting the role of acetylcholinesterase in Alzheimer’s disease: cross-talk with P-tau and β-amyloid. Front Mol Neurosci 2011; 4: 22.

15.

Shoichet

. Virtual screening of chemical libraries. Nature 2004; 432: 862–865.

16.

Glantz-Gashai

Meirson

Reuveni

, et al. Virtual screening for potential inhibitors of Mcl-1 conformations sampled by normal modes, molecular dynamics, and nuclear magnetic resonance. Drug Des Devel Ther 2017; 11: 1803–1813.

17.

Bourafai-Aziez

Benabderrahmane

Paysant

, et al. Drug repurposing: deferasirox inhibits the anti-apoptotic activity of Mcl-1. Drug Des Devel Ther 2021; 15: 5035–5059.

18.

Pinzi

Bisi

Rastelli

. How drug repurposing can advance drug discovery: challenges and opportunities. Front Drug Discov 2024; 4: 1460100.

19.

Iqbal

Rizvi

SMD

Rehman

, et al. Soyasapogenol-B as a potential multitarget therapeutic agent for neurodegenerative disorders: molecular docking and dynamics study. Entropy 2022; 24: 593.

20.

Iqbal

Rehman

Alajmi

, et al. Multitargeted virtual screening and molecular simulation of natural product-like compounds against GSK3β, NMDA-receptor, and BACE-1 for the management of Alzheimer’s disease. Pharmaceuticals 2023; 16: 622.

21.

Iqbal

Rehman

Bin Dukhyil

, et al. High-throughput screening and molecular dynamics simulation of natural product-like compounds against Alzheimer’s disease through multitarget approach. Pharmaceuticals 2021; 14: 937.

22.

Polis

Samson

. Role of the metabolism of branched-chain amino acids in the development of Alzheimer’s disease and other metabolic disorders. Neural Regen Res 2020; 15: 1460–1470.

23.

Mani

Jindal

Chopra

, et al. ROCK2 Inhibition: a futuristic approach for the management of Alzheimer’s disease. Neurosci Biobehav Rev 2022; 142: 104871.

24.

Khare

Maheshwari

Rizvi

SMD

, et al. Homology modelling, molecular docking and molecular dynamics simulation studies of CALMH1 against secondary metabolites of Bauhinia variegata to treat Alzheimer’s disease. Brain Sci 2022; 12: 770.

25.

Polis

Samson

. Arginase as a potential target in the treatment of Alzheimer’s disease. Adv Alzheimer Dis 2018; 07: 119–140.

26.

Wang

Gao

, et al. A drug repositioning approach reveals ergotamine may be a potential drug for the treatment of Alzheimer’s disease. J Alzheimers Dis 2024; 101: 1355–1366.

27.

Wang

Gao

, et al. Dihydroergotamine and bromocriptine: potential drugs for the treatment of major depressive disorder and Alzheimer’s disease comorbidity. Mol Neurobiol 2024; 62: 2493–2514.

28.

Trott

Olson

. Autodock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 2010; 31: 455–461.

29.

O’Boyle

Banck

James

, et al. Open Babel: an open chemical toolbox. J Cheminform 2011; 3: 33.

30.

Meng

Goddard

Pettersen

, et al. UCSF Chimerax: tools for structure building and analysis. Protein Sci 2023; 32: 1–13.

31.

Salentin

Schreiber

Haupt

, et al. PLIP: fully automated protein–ligand interaction profiler. Nucleic Acids Res 2015; 43: W443–W447.

32.

Phillips

Hardy

Maia

JDC

, et al. Scalable molecular dynamics on CPU and GPU architectures with NAMD. J Chem Phys 2020; 153: 044130.

33.

Huang

Rauscher

Nawrocki

, et al. CHARMM36m: An improved force field for folded and intrinsically disordered proteins. Nat Methods 2017; 14: 71–73.

34.

Humphrey

Dalke

Schulten

. VMD: visual molecular dynamics. J Mol Graph 1996; 14: 33–38.

35.

Kim

Lee

, et al. CHARMM-GUI ligand reader and modeler for CHARMM force field generation of small molecules. J Comput Chem 2017; 38: 1879–1886.

36.

Jorgensen

Chandrasekhar

Madura

, et al. Comparison of simple potential functions for simulating liquid water. J Chem Phys 1983; 79: 926–935.

37.

Vanommeslaeghe

Hatcher

Acharya

, et al. CHARMM General force field: a force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields. J Comput Chem 2010; 31: 671–690.

38.

Essmann

Perera

Berkowitz

, et al. A smooth particle mesh Ewald method. J Chem Phys 1995; 103: 8577–8593.

39.

Pires DE

Blundell

Ascher

. pkCSM: predicting small-molecule pharmacokinetic and toxicity properties using graph-based signatures. J Med Chem 2015; 58: 4066–4072.

40.

Brus

Košak

Turk

, et al. Discovery, biological evaluation, and crystal structure of a novel nanomolar selective butyrylcholinesterase inhibitor. J Med Chem 2014; 57: 8167–8179.

41.

Bourne

Grassi

Bougis

, et al. Conformational flexibility of the acetylcholinesterase tetramer suggested by X-ray crystallography. J Biol Chem 1999; 274: 30370–30376.

42.

Petros

Korepanova

Jakob

, et al. Fragment-based discovery of an apolipoprotein E4 (apoE4) stabilizer. J Med Chem 2019; 62: 4120–4130.

43.

Sun

Y-Y

Wang

Huang

H-C

. Roles of ApoE4 on the pathogenesis in Alzheimer’s disease and the potential therapeutic approaches. Cell Mol Neurobiol 2023; 43: 3115–3136.

44.

Ghosh

Kumaragurubaran

Hong

, et al. Design, synthesis and X-ray structure of protein−ligand complexes: important insight into selectivity of memapsin 2 (β-secretase) inhibitors. J Am Chem Soc 2006; 128: 5310–5311.

45.

Guo

Wang

Zhou

, et al. Molecular basis for isoform-selective inhibition of presenilin-1 by MRK-560. Nat Commun 2022; 13: 6299.

46.

Soto-Ospina

Araque Marín

Bedoya

GdJ

, et al. Structural predictive model of presenilin-2 protein and analysis of structural effects of familial Alzheimer’s disease mutations. Biochem Res Int 2021; 2021: 1–20.

47.

Lei

Niu

, et al. The FDA-approved natural product dihydroergocristine reduces the production of the Alzheimer’s disease amyloid-β peptides. Sci Rep 2015; 5: 16541.

48.

Vadukul

Papp

Thrush

, et al. α-synuclein aggregation is triggered by oligomeric amyloid-β 42 via heterogeneous primary nucleation. J Am Chem Soc 2023; 145: 18276–18285.

49.

Robinson

Radziejewski

Stuart

, et al. Structure of the brain-derived neurotrophic factor/neurotrophin 3 heterodimer. Biochemistry 1995; 34: 4139–4146.

50.

Malia

Teplyakov

Ernst

, et al. Epitope mapping and structural basis for the recognition of phosphorylated tau by the anti-tau antibody AT8. Proteins 2016; 84: 427–434.

51.

Kong

GK-W

Adams

Harris

, et al. Structural studies of the Alzheimer’s amyloid precursor protein copper-binding domain reveal how it binds copper ions. J Mol Biol 2007; 367: 148–161.

52.

Ben Zaken

Bouhnik

Omer

, et al. Polyoxometalates bind multiple targets involved in Alzheimer’s disease. J Biol Inorg Chem 2025; 30: 299–309.

53.

Detroja

Samson

. Virtual screening for FDA-approved drugs that selectively inhibit arginase type 1 and 2. Molecules 2022; 27: 5134.

54.

Butini

Campiani

Borriello

, et al. Exploiting protein fluctuations at the active-site gorge of human cholinesterases: further optimization of the design strategy to develop extremely potent inhibitors. J Med Chem 2008; 51: 3154–3170.

55.

Nordberg

Ballard

Bullock

, et al. A review of butyrylcholinesterase as a therapeutic target in the treatment of Alzheimer’s disease. Prim Care Companion CNS Disord 2013; 15: PCC.12r01412.

56.

Kumar

Pintus

Petrillo

, et al. Novel 2-pheynlbenzofuran derivatives as selective butyrylcholinesterase inhibitors for Alzheimer’s disease. Sci Rep 2018; 8: 4424.

57.

Gouveia

Camins

Ettcheto

, et al. Targeting brain renin-angiotensin system for the prevention and treatment of Alzheimer’s disease: past, present and future. Ageing Res Rev 2022; 77: 101612.

58.

Center for Drug Evaluation and Research . Application Number: 22-311, Pharmacology Review(S), https://www.accessdata.fda.gov/drugsatfda_docs/nda/2008/022311s000_pharmr.pdf.

59.

Reid

Schwartz

Ben

, et al. Interactions between vasopressin and the renin–angiotensin system. Prog Brain Res 1983; 60: 475–491.

60.

Diz

Westwood

Bosch

, et al. NK1 Receptor antagonist blocks angiotensin II responses in renin transgenic rat medulla oblongata. Hypertension 1998; 31: 473–479.

61.

de Hoon

Poppe

Thijssen

, et al. Dihydroergotamine: discrepancy between arterial, arteriolar and pharmacokinetic data. Br J Clin Pharmacol 2001; 52: 45–51.

62.

Schankin

Maniyar

Seo

, et al. Ictal lack of binding to brain parenchyma suggests integrity of the blood–brain barrier for 11C-dihydroergotamine during glyceryl trinitrate-induced migraine. Brain 2016; 139: 1994–2001.

63.

Roberto

Raschi

Piccinni

, et al. Adverse cardiovascular events associated with triptans and ergotamines for treatment of migraine: systematic review of observational studies. Cephalalgia 2015; 35: 118–131.

64.

Schneider

Olin

Novit

, et al. Hydergine for dementia. Cochrane Database Syst Rev 2000; 2: CD000359.

65.

Hughes

Williams

Currier

. An Ergot alkaloid preparation (hydergine) in the treatment of dementia: critical review of the clinical literature. J Am Geriatr Soc 1976; 24: 490–497.

66.

Kim

Park

, et al. Association between migraine and Alzheimer’s disease: a nationwide cohort study. Front Aging Neurosci 2023; 15: 1196185.

67.

Panaccione

Coyle

. Abundant respirable ergot alkaloids from the common airborne fungus Aspergillus fumigatus. Appl Environ Microbiol 2005; 71: 3106–3111.

68.

Bredesen

. Inhalational Alzheimer’s disease: an unrecognized—and treatable—epidemic. Aging 2016; 8: 304–313.

69.

Miyake

Bleier

. The blood-brain barrier and nasal drug delivery to the central nervous system. Am J Rhinol Allergy 2015; 29: 124–127.

70.

Silvestrini

Liuzzi

Chiodini

. Effect of ergot alkaloids on growth hormone and prolactin secretion in humans. Pharmacology 1978; 16: 78–87.

71.

Wang

Cao

Huang

, et al. DHEC Mesylate attenuates pathologies and aberrant bisecting N-glycosylation in Alzheimer’s disease models. Neuropharmacology 2024; 248: 109863.

72.

Liu

Cheng

Z-Y

Y-F

, et al. Dopamine D2 receptor agonist Bromocriptine ameliorates Aβ1-42-induced memory deficits and neuroinflammation in mice. Eur J Pharmacol 2023; 938: 175443.

73.

Garcia-Argibay

Hiyoshi

Fall

, et al. Association of 5α-reductase inhibitors with dementia, depression, and suicide. JAMA Netw Open 2022; 5: e2248135.

74.

Nørgaard

Horváth-Puhó

Corraini

, et al. Sleep disruption and Alzheimer’s disease risk: inferences from men with benign prostatic hyperplasia. EClinicalMedicine 2021; 32: 100740.

75.

Carroll

Rosario

Chang

, et al. Progesterone and estrogen regulate Alzheimer-like neuropathology in female 3xTg-AD mice. J Neurosci 2007; 27: 13357–13365.

76.

Koebele

Poisson

Palmer

, et al. Evaluating the cognitive impacts of drospirenone, a spironolactone-derived progestin, independently and in combination with ethinyl estradiol in ovariectomized adult rats. Front Neurosci 2022; 16: 885321.

77.

Savolainen-Peltonen

Rahkola-Soisalo

Hoti

, et al. Use of postmenopausal hormone therapy and risk of Alzheimer’s disease in Finland: nationwide case-control study. Br Med J 2019; 364: l665.

78.

Choice of contraceptives. Treat Guidel Med Lett 2010; 8: 89–96.

79.

Pagan

Hebron

Wilmarth

, et al. Pharmacokinetics and pharmacodynamics of a single dose Nilotinib in individuals with Parkinson’s disease. Pharmacol Res Perspect 2019; 7: e00470.

80.

La Barbera

Vedele

Nobili

, et al. Nilotinib restores memory function by preventing dopaminergic neuron degeneration in a mouse model of Alzheimer’s disease. Prog Neurobiol 2021; 202: 102031.

81.

Turner

Hebron

Lawler

, et al. Nilotinib effects on safety, tolerability, and biomarkers in Alzheimer’s disease. Ann Neurol 2020; 88: 183–194.

82.

Stevenson

Varghese

Hebron

, et al. Inhibition of discoidin domain receptor (DDR)-1 with nilotinib alters CSF miRNAs and is associated with reduced inflammation and vascular fibrosis in Alzheimer’s disease. J Neuroinflammation 2023; 20: 116.

83.

Nishioka

Tooi

Isobe

, et al. BMS-708163 and Nilotinib restore synaptic dysfunction in human embryonic stem cell-derived Alzheimer’s disease models. Sci Rep 2016; 6: 33427.

84.

Srivastava

Renna

Johnson

, et al. Nilotinib as a prospective treatment for Alzheimer’s disease: effect on proteins involved in neurodegeneration and neuronal homeostasis. Life 2024; 14: 1241.

85.

Simuni

Fiske

Merchant

, et al. Efficacy of nilotinib in patients with moderately advanced Parkinson disease. JAMA Neurol 2021; 78: 312.

86.

Tan

Baggio

Shortt

, et al. Cardiovascular safety of nilotinib in Alzheimer disease. Ann Neurol 2021; 89: 196.

87.

Alsharief

. How do retinoids affect Alzheimer’s disease and can they be novel drug candidates? Cureus 2024; 16: e57548.

88.

Habchi

Chia

Limbocker

, et al. Systematic development of small molecules to inhibit specific microscopic steps of Aβ42 aggregation in Alzheimer’s disease. Proc Natl Acad Sci U S A 2017; 114: E200–E208.

89.

Kolarcik

Bowser

. Retinoid signaling alterations in amyotrophic lateral sclerosis. Am J Neurodegener Dis 2012; 1: 130–145.

90.

Zhou

Jiao

, et al. Uncovering the oxidative stress mechanisms and targets in Alzheimer’s disease by integrating phenotypic screening data and polypharmacology networks. J Alzheimers Dis 2024; 99: S139–S156.

91.

Givre

Fleischman

. Intracranial hypertension in a patient using topical Adapalene. J Neuroophthalmol 2008; 28: 156–158.

92.

Rusu

Tanase

Pascu

G-A

, et al. Recent advances regarding the therapeutic potential of Adapalene. Pharmaceuticals 2020; 13: 217.

93.

Anthony

. Irinotecan toxicity. Curr Opin Support Palliat Care 2007; 1: 35–39.

94.

Mahajan

Roy

, et al. Enhancing the delivery of anti retroviral drug saquinavir across the blood brain barrier using nanoparticles. Curr HIV Res 2010; 8: 396–404.

95.

Lee

Duda

Muzikansky

, et al. Phase I and biomarker study of plerixafor and bevacizumab in recurrent high-grade glioma. Clin Cancer Res 2018; 24: 4643–4649.

96.

Parachikova

Cotman

. Reduced CXCL12/CXCR4 results in impaired learning and is downregulated in a mouse model of Alzheimer disease. Neurobiol Dis 2007; 28: 143–153.

97.

Gavriel

Rabinovich-Nikitin

Ezra

, et al. Subcutaneous administration of AMD3100 into mice models of Alzheimer’s disease ameliorated cognitive impairment, reduced neuroinflammation, and improved pathophysiological markers. J Alzheimers Dis 2020; 78: 653–671.

98.

Shin

Lee

, et al. Combined effects of hematopoietic progenitor cell mobilization from bone marrow by granulocyte colony stimulating factor and AMD3100 and chemotaxis into the brain using stromal cell-derived factor-1α in an Alzheimer’s disease mouse model. Stem Cells 2011; 29: 1075–1089.

99.

Ganne

Mainali

Balasubramaniam

, et al. Ezetimibe lowers risk of Alzheimer’s and related dementias over sevenfold, reducing aggregation in model systems by inhibiting 14-3-3G::hexokinase interaction. Aging Biol 2024; 2: 20240028.

100.

Nakayama

Amiry-Moghaddam

Ottersen

, et al. Conivaptan, a selective arginine vasopressin V1a and V2 receptor antagonist attenuates global cerebral edema following experimental cardiac arrest via perivascular pool of aquaporin-4. Neurocrit Care 2016; 24: 273–282.

101.

Quintanar-Stephano

Viñuela-Berni

Macías-Segura

, et al. Effects of arginine vasopressin (AVP) deficiency, conivaptan and desmopressin on clinical symptoms and blood cytokine levels in rats with experimental autoimmune encephalomyelitis. FASEB J 2018; 32: 741.7–741.7.

102.

Verbalis

Ellison

Hobart

, et al. Tolvaptan and neurocognitive function in mild to moderate chronic hyponatremia: a randomized trial (INSIGHT). Am J Kidney Dis 2016; 67: 893–901.

103.

Tandukar

Sterns

Rondon-Berrios

. Osmotic demyelination syndrome following correction of hyponatremia by ≤10 mEq/L per day. Kidney360 2021; 2: 1415–1423.

104.

Ghali

Koren

Taylor

, et al. Efficacy and safety of oral conivaptan: a V1A/V2 vasopressin receptor antagonist, assessed in a randomized, placebo-controlled trial in patients with euvolemic or hypervolemic hyponatremia. J Clin Endocrinol Metab 2006; 91: 2145–2152.

105.

Philipp

. Substance P and antagonists of the neurokinin-1 receptor in neuroinflammation associated with infectious and neurodegenerative diseases of the central nervous system. J Neurol Neuromedicine 2016; 1: 29–36.

106.

Johnson

Young

Marriott

. The therapeutic potential of targeting substance P/NK-1R interactions in inflammatory CNS disorders. Front Cell Neurosci 2017; 10: 296.

107.

Walker

Davies

Martin

, et al. Comparison of antihypertensive drug classes for dementia prevention. Epidemiology 2020; 31: 852–859.

108.

Zhang

Hou

, et al. Population-based discovery and Mendelian randomization analysis identify telmisartan as a candidate medicine for Alzheimer’s disease in African Americans. Alzheimers Dement 2023; 19: 1876–1887.

109.

Liu

C-H

Sung

P-S

Y-R

, et al. Telmisartan use and risk of dementia in type 2 diabetes patients with hypertension: a population-based cohort study. PLoS Med 2021; 18: e1003707.

110.

Barthold

Joyce

Wharton

, et al. The association of multiple anti-hypertensive medication classes with Alzheimer’s disease incidence across sex, race, and ethnicity. PLoS One 2018; 13: e0206705.

111.

X-X

Wei

Cao

H-M

, et al. Telmisartan alleviates Alzheimer’s disease-related neuropathologies and cognitive impairments. J Alzheimers Dis 2023; 94: 919–933.

112.

Torika

Asraf

Cohen

, et al. Intranasal telmisartan ameliorates brain pathology in five familial Alzheimer’s disease mice. Brain Behav Immun 2017; 64: 80–90.

113.

Hazlewood

Chen

Clark

, et al. Differential effects of angiotensin II type I receptor blockers on reducing intraocular pressure and TGFβ signaling in the mouse retina. PLoS One 2018; 13: e0201719.

114.

Hammer

Bahramnejad

Watkins

, et al. Candesartan restores blood–brain barrier dysfunction, mitigates aberrant gene expression, and extends lifespan in a knockin mouse model of epileptogenesis. Clin Sci 2024; 138: 1089–1110.

115.

Ouk

C-Y

Rabin

, et al. The use of angiotensin-converting enzyme inhibitors vs. angiotensin receptor blockers and cognitive decline in Alzheimer’s disease: the importance of blood-brain barrier penetration and APOE ε4 carrier status. Alzheimers Res Ther 2021; 13: 43.

116.

Chandrasekhar

Gor

Ramalingam

, et al. Repurposing FDA-approved compounds to target JAK2 for colon cancer treatment. Discov Oncol 2024; 15: 226.

117.

Montes-Grajales

Puerta-Guardo

Espinosa

, et al. In silico drug repurposing for the identification of potential candidate molecules against arboviruses infection. Antiviral Res 2020; 173: 104668.

118.

Jamir

Sarma

Priyadarsinee

, et al. Polypharmacology guided drug repositioning approach for SARS-CoV2. PLoS One 2023; 18: e0289890.

119.

Gurung

Ali

Lee

, et al. In silico screening of FDA approved drugs reveals ergotamine and dihydroergotamine as potential coronavirus main protease enzyme inhibitors. Saudi J Biol Sci 2020; 27: 2674–2682.

120.

White

Lin

Cheng

. Discovery of COVID-19 inhibitors targeting the SARS-CoV-2 Nsp13 helicase. J Phys Chem Lett 2020; 11: 9144–9151.

121.

Gül

. In silico drug repositioning against human NRP1 to block SARS-CoV-2 host entry. Turk J Biol 2021; 45: 442–458.

122.

Almoyad

MAA

Wahab

Mohanto

, et al. Repurposing drugs to modulate sortilin: structure-guided strategies against atherogenesis, coronary artery disease, and neurological disorders. ACS Omega 2024; 9: 18438–18448.

123.

Jiménez

Żołek

Hernández Perez

, et al. Drug repurposing to inhibit histamine N-methyl transferase. Molecules 2023; 28: 576.

124.

Chun Hao

Chau Ling

Harith

, et al. Drug repurposing of clinically approved drugs to target epithelial-mesenchymal transition using molecular docking approach. Malays J Med Health Sci 2023; 19: 15–23.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

8.20 MB