In silico study of selected alkaloids as dual inhibitors of β- and γ-secretases for Alzheimer's disease

Abstract

Background

Alzheimer's disease (AD) has become common as the number of aged people increases making it as a socioeconomic problem lately. To date, no success is recorded for disease-modifying therapies for AD but only drugs for symptomatic relief exist. Research has been centered on the role of amyloid-β on the pathogenesis of AD, which has led to the development of drugs that target Aβ (β- and γ-secretase inhibitors) to reduce the amount of Aβ formed. However, the existing β and γ-secretase inhibitors were associated with harmful side effects, low efficacy, and inability to cross the blood-brain barrier.

Objective

This study therefore used in silico approach to predict the inhibitory properties of alkaloids as potential drug targets against AD.

Methods

Thus, in this current study, 54 alkaloids from the PhytoHub server (phytohub.eu), and two approved drugs were docked against β-secretases. Additionally, galantamine and 5 alkaloids with the utmost binding potential with β-secretase were subjected to pharmacokinetics evaluation and docked against γ-secretase.

Results

From the result, 5 compounds displayed for both docking periods, with demissidine, solasodine, tomatidine, and solanidine having better BE than the control drugs. Based on the pharmacokinetics evaluation, 4 compounds possessed good pharmacokinetic evaluation and biological activities than galantamine.

Conclusions

This study suggests that demissidine, solasodine, tomatidine, and solanidine are promising dual inhibitors against β- and γ-secretase proteins in silico. However, there is an urgent need to carry out in vitro and in vivo experiments on these new leads to validate the findings of this study.

Keywords

alkaloids Alzheimer's disease β- and γ-secretase inhibitors virtual docking

Introduction

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by the presence of senile plaques in the brain, which are primarily composed of aggregated amyloid-β (Aβ) that results from the amyloid-β protein precursor (AβPP).^1,2 Extracellular Aβ is produced from the proteolytic degradation of AβPP and is one of these proteins. AβPP is a type 1 transmembrane protein that can be found in a variety of tissues but is most commonly found in neurons. It has been proposed that AβPP is involved in creating synapses and the healing of neurons.³ Research focuses on three proteases—α-, β-, and γ-secretases—due to their roles in the modulation and production of Aβ and their potential as drug targets for managing AD.^2,4 This study emphasizes β- and γ-secretases, as they are involved in producing insoluble Aβ, which forms Aβ plaques.

The common symptoms of AD include cognitive dysfunction (such as memory loss and communication difficulties), developmental issues (like depression and hallucinations, often known as the non-markers), and challenges with daily activities.^2,5,6 The primary causes of AD are genetic predisposition, age, and abnormal protein accumulation within and outside of neurons.^6,7 Aβ is a protein, produced from the proteolytic breakdown of AβPP, a type 1 transmembrane protein predominantly found in neurons and a variety of tissues, and is thought to be involved in synapse formation and neuronal repairs.³ The key pathophysiological mechanism of AD is the external development of amyloid proteins, known as amyloid plaques. It is also the primary drug benchmark for the inhibitory activity of Aβ yield in AD.^7,8 It is produced by the protease activity of the β-secretase enzyme. Before the symptoms occur, unexpected changes and improvements relative to the brain's biochemistry are thought to occur.^5,9

In recent times, more information has accumulated addressing the influence of AD caused by significant neuronal damage and neurodegenerative abnormalities in numerous areas of the brain.^9,10 Amyloid plaque is one of the neuropathological hallmarks of AD. It is mainly comprised of Aβ peptide, a 38 to 43 amino acid peptide derived from the cleavage of the AβPP.¹¹ This latter make protein undergoes two distinct and competing mechanisms, usually referred to as the amyloidogenic and non-amyloidogenic pathways, in which the proteolytic mechanisms are largely dependent on three kinds of secretase enzymes (α-, β-, and γ-secretase).^12–14 In addition, almost all drugs developed over the years to treat AD were reported to have failed to exhibit the expected efficacy. Furthermore, those targeting these proteases as either enhancers, modulators, or inhibitors were discontinued before phase III trials. Hence, there is an urgent need to discover small molecules from natural sources such as plants to avert or slow down the progression of AD.

Alkaloids are nitrogen-containing plant molecules found majorly among plant families such as Papaveraceae (poppies family), Amaryllidaceae (amaryllis), Solanaceae (nightshades), and Ranunculaceae (buttercups).^15,16 Plant alkaloids exhibit a wide range of biological and pharmacological potentials, including neuroprotective potential.^15,17 Due to its neuroprotective potential, we explored some plant alkaloids in this current in silico study. Interestingly, two drugs approved by the FDA as cholinesterase inhibitors are from this class of plant molecules (galantamine and rivastigmine) to treat AD.^16,18 This suggests the importance of this class of compounds. However, little research has looked at alkaloids as potential inhibitors against β- and γ-Secretases to treat AD. This study therefore used in silico approach to predict the inhibitory properties of alkaloids as potential drug targets against AD.

Methods

Selection and preparation of protein targets

The 3D structures of β- and γ-secretases from the Protein Data Bank were retrieved with PDB codes 4LXM and 7D8X respectively.¹⁹ Then, they were subjected to cleaning and optimization through the dock prep module of UCSF-Chimera Software Ver. 1.13.1. Here, the proteins were freed from all heteroatoms such as co-crystallized ligands and water. Followed by adding hydrogen atoms and charges using gasteiger charge. Finally, the proteins were minimized under the influence of amber force field 94 fs (Amberff94 fs).

Selection and optimization of ligands

86 compounds of alkaloid extraction (Table 1) were selected as our ligands under investigation with two approved drugs to treat AD from Drugbank (galantamine and rivastigmine). These alkaloids were selected based on literature that alkaloids possess neuroprotective ability.^15,16 Their 3D conformers were obtained through the phytohub webserver (phytohub.eu) in structure data format (SDF). Open Babel housed within the Python prescription (PyRx) suit was deployed to filter the compounds through their molecular weights (<500), then those having a molecular weight less than 500 were optimized under the influence of the Universal Force field (UFF) and converted to their best energetic and stable configurations.

Table 1.

Selected alkaloids in food.

S/No.	Compounds	Phytohub ID	Pubchem CID
1.	15-decarboxy-betanin	PHUB000416	12300103
2.	17-decarboxy-betanin	PHUB000413	101333461
3.	17-decarboxy-neobetanin	PHUB000415	102401025
4.	2-decarboxy-betanin	PHUB000410	131751981
5.	3-Methoxy-tyramine-betaxanthin	PHUB000431	135871118
6.	5-Hydroxynorvaline-betaxanthin	PHUB000456	246598
7.	Alanine-betaxanthin	PHUB001470	24779335
8.	Arginine-betaxanthin	PHUB000451	447276923
9.	Asparagine-betaxanthin	PHUB000434	447276924
10.	Betalamic acid	PHUB000404	5281176
11.	Betanidin	PHUB000407	135449343
12.	Betanin	PHUB000405	12300103
13.	glutamine-betaxanthin	PHUB000432	135438599
14.	Glycine-betaxanthin	PHUB000444	135438598
15.	Gomphrenin I	PHUB000429	6096868
16.	Gomphrenin II	PHUB001477	131752748
17.	Gomphrenin III	PHUB001476	101105498
18.	Histamine-betaxanthin	PHUB000446	774
19.	Histidine-betaxanthin	PHUB000442	6274
20.	Hylocerenin	PHUB000423	101168510
21.	Isobetanidin	PHUB000408	135612764
22.	Isobetanin	PHUB000406	6325438
23.	Isogomphrenin I	PHUB000430	90658633
24.	Isogomphrenin II	PHUB001478	101105497
25.	Isogomphrenin III	PHUB001479	101105499
26.	Isolampranthin II	PHUB000428	157010344
27.	Isoleucine-betaxanthin	PHUB000436	157010345
28.	Isophyllocactin	PHUB000426	101168512
29.	Lampranthin II	PHUB000427	11953909
30.	Leucine-betaxanthin	PHUB000435	447276973
31.	Neobetanin	PHUB000409	102401026
32.	Phenylalanine-betaxanthin	PHUB000448	52923786
33.	Phyllocactin	PHUB000425	101056997
34.	Prebetanin	PHUB000422	6325833
35.	Proline-betaxanthin	PHUB000447	57513848
36.	Tryptophan-betaxanthin	PHUB000454	136728070
37.	Tyramine-betaxanthin	PHUB000440	157010370
38.	tyrosine-betaxanthin	PHUB000443	135887852
39.	Valine-betaxanthin	PHUB000450	447277025
40.	γ-Aminobutyric acid-betaxanthin	PHUB000445	447277027
41.	13-Hydroxylupanine	PHUB001464	250873
42.	Adenine	PHUB001480	190
43.	Albine	PHUB000825	442936
44.	Calystegin A3	PHUB000826	442999
45.	Calystegin B2	PHUB000827	443000
46.	Gentianine	PHUB000828	354616
47.	Harman	PHUB000829	5281404
48.	Harmine	PHUB000830	5280953
49.	Hyoscyamine	PHUB000831	154417
50.	lupanine	PHUB001465	91471
51.	lupinine	PHUB001467	91461
52.	Multiflorine	PHUB001468	6918763
53.	Perlolyrine	PHUB000832	160179
54.	Physoperuvine	PHUB000833	443008
55.	Salsolinol ((-)-)	PHUB000834	91588
56.	Scopolamine	PHUB000835	3000322
57.	Sparteine	PHUB001469	644020
58.	Trigonelline	PHUB001378	5570
59.	1,3-dimethyluric acid	PHUB002107	70346
60.	Caffeine	PHUB000789	2519
61.	Convicine	PHUB000793	88000
62.	Dihydrozeatin	PHUB000795	32021
63.	Theobromine	PHUB000790	5429
64.	Theophylline	PHUB000791	2153
65.	Vicine	PHUB000792	135413566
66.	Zeatin (trans-)	PHUB000794	449093
67.	Arecaidine	PHUB000836	10355
68.	Arecoline	PHUB000837	2230
69.	Carpaine	PHUB000838	442630
70.	Cucurbitine	PHUB000839	442634
71.	Fagomine	PHUB000840	72259
72.	Guvacine	PHUB000841	3532
73.	Pelletierine ((-)-)	PHUB000842	92987
74.	Piperine	PHUB000843	638024
75.	Trichostachine	PHUB000845	636537
76.	Chaconine (alpha-)	PHUB000846	442971
77.	Dehydrotomatine	PHUB000847	131751243
78.	Demissidine	PHUB000848	101379
79.	Demissine	PHUB000849	442975
80.	Solamargine	PHUB000850	73611
81.	Solanidine	PHUB000851	65727
82.	Solanine (alpha-)	PHUB000852	159233
83.	Solasodine	PHUB000853	442985
84.	Solasonine	PHUB000854	119247
85.	Tomatidine	PHUB000855	65576
86.	Tomatine	PHUB000856	28523

Virtual docking simulation of β-secretase and ligands

Before the docking simulation, the protocol for docking was validated by re-docking the co-crystallized ligand into its binding region in β-secretase to regenerate a similar binding post like that in the X-ray crystallized structure of β-secretase (4LXM). Auto dock vina was used for the docking and PyMOL 2.4 was used to align the two complex structures, consequently obtaining the root mean square deviation (RMSD). The virtual docking of all selected ligands against β-secretase was achieved by employing the flexible procedure of docking previously used.²⁰ PyRx 0.8, a suite integrated with Auto Dock Vina, was applied for the molecular docking simulation.²¹ The specific target site for β-secretase corresponding to the binding region used in AβPP processing was adjusted using the grid box with dimensions (23.1362 × 19.9375 × 23.7691) Å and the center was attuned based on the site of binding in the protein consisting of the following amino acids: Asp228, Asp32, Gln73, Thr72, and Gln12. The compounds with better binding energy than the control drugs at the end of the experiment were subjected to drug-likeness screening using the SwissAdme server, and molecular interaction analysis with the aid of PyMOL^© Molecular Graphics (version 2.4, 2016, Shrodinger LLC) and LigPlot ⁺. The generated compounds are regarded as potential hits.

Molecular docking simulation of γ-secretase and hit ligands

Similar to the above steps, the molecular docking was carried out between γ-secretase and hit compounds with good pharmacokinetic attributes based on the outcome of drug-likeness screening via the SwissAdme server. However, comparing the molecular interactions in the co-crystallized and re-docked complexes were carried out for the protocol validation. Furthermore, the specific target site for γ-secretase corresponding to the binding region exercised in AβPP processing was attuned via the grid box with dimensions (26.514 × 24.0892 × 25.4268) Å, and the center was acclimated based on the catalytic site binding in the target comprising of the listed amino acids; Leu381, Pro433, Ala434, Leu435, Lys380, Asp257, Leu432, Tyr288, Asp385, Ser289, Val261, Val272, Ser290, Gly382, Gly384, Leu282, Ile143, Ile253, Tyr256, Phe388, Thr421, Leu425, and Val379.

In silico pharmacokinetics screening

Hit compounds against the secretase enzymes were screened for their pharmacokinetics and physicochemical properties with the assistance of the SwissAdme server after uploading their canonical smiles.

Molecular dynamics simulation analyses

The hit alkaloids were arranged after the thorough computational virtual screening, molecular docking analysis, and ADMET analysis. The selected protein-ligand complexes rigid binding stability and molecular pattern in the physiological sphere were validated via molecular dynamic (MD) simulation evaluation of 100 ns.^22–24 The Schrodinger's Desmond module was deployed for MD simulation investigations.²⁵ The analysis begins with introducing the best-docked poses of demissidine, solanidine, solasodine, Tomatidine, and galantamine ligand-protein complexes for the two secretases, we are probing in this study, to Desmond module in the proper file format. The protein preparation wizard was utilized for selected ligand and protein structural optimization by removing and repairing steric hindrances, disordered and undesirable atomic geometries.²⁶

The solvated water–soaked MD simulation system was generated using the Desmond System Builder tool. The intermolecular interaction potential 3 points transferable (TIP3P) model was turned to account for solvated model inception. An orthorhombic simple point charge water box of 8 Å beyond to any boundary of the protein-ligand complex was precisely invented. The OPLS-3e force field was deployed for energy minimization.²⁷ The simulated system complex charges were kept at neutral, and isosmotic state by adding desirable amount of the counter ions. A total 59.446 mM, 50.817 mM sodium (Na⁺) and chloride (Cl⁻) ions are added into the simulation system. Then, the simulation system was equilibrated. The NPT ensemble at 300 K temperature and 1.0 bar pressure coupled with known Berendsen coupling protocol is set off for system equilibration. The SHAKE algorithm was deployed for H-bond geometry constrained analysis as a time function. The non-bonded energy computations, particularly long-range electrostatic energy, were computed with the Particle Mesh Ewald (PME) algorithm. The simulation was set for 100 ns and 1000 simulation trajectories frames were retained for the computation of root-mean square fluctuation (RMSF), RMSD, protein-ligand contacts, and interactions fractions (%). The in-built Desmond interaction diagram tool was used for plot and figure generation.

Results

The present in silico studies aimed at discovering potential dual inhibitors against two key enzymes involved with amyloid protein processing called β- and γ-secretases (3D structures provided in Figure 1) from 86 alkaloids sourced from Phytohub server. Out of these 86 compounds (Table 1), 52 were found to have molecular weights less than 500 through the open babel module in PyRx suit 0.8. consequently, these 52 compounds and two controls (2D structures provided in Figure 2) were docked against the active site of β-secretase through autodock vina. We observed that binding energies (BE) that results are from −4.8 to −9.4 kcal/mol where the two control drugs (galantamine and rivastigmine) have BE values of −7.4 and −6.1 kcal/mol, respectively. Five compounds among the 52 alkaloids displayed the utmost BE while 15 alkaloids had BE below that of the two control drugs. The details of these findings are presented in Table 2.

Figure 1.

3D structures of Secretases used in our current in silico study retrieved from the protein data bank server. (a) β-secretase (4LXM, resolved at 2.30 Å). (b) γ-secretase (7D8X, resolved at 2.60 Å).

Figure 2.

2D structures of alkaloid compounds docked against secretase.

Table 2.

Binding energy (kcal/mol) of compounds docked against β-secretase.

S/No.	Compounds	Binding Energy (kcal/mol)
1.	1,3-dimethyluric acid	−5.9
2.	3-Methoxy-tyramine-betaxanthin	−8.1
3.	5-Hydroxynorvaline-betaxanthin	−7.0
4.	13-Hydroxylupanine	−7.3
5.	Adenine	−5.3
6.	Arecoline	−4.9
7.	Asparagine-betaxanthin	−7.3
8.	Betalamic acid	−6.1
9.	Betanidin	−7.7
10.	Caffeine	−5.5
11.	Calystegin A3	−6.1
12.	Calystegin B2	−6.3
13.	Carpaine	−9.4
14.	Galantamine*	−7.4
15.	Rivastigmine*	−6.1
16.	Convicine	−7.0
17.	Cucurbitine	−4.6
18.	Demissidine	−8.9
19.	Dihydrozeatin	−6.8
20.	Fagomine	−4.7
21.	γ-Aminobutyric acid-betaxanthin	−6.7
22.	Gentianine	−6.5
23.	glutamine-betaxanthin	−7.1
24.	Glycine-betaxanthin	−6.4
25.	Guvacine	−4.9
26.	Harman	−6.9
27.	Harmine	−7.0
28.	Histamine-betaxanthin	−4.1
29.	Histidine-betaxanthin	−4.9
30.	Hyoscyamine	−7.8
31.	Isobetanidin	−8.0
32.	lupanine	−7.5
33.	lupinine	−5.7
34.	Multiflorine	−7.5
35.	Pelletierine ((-)-)	−5.3
36.	Perlolyrine	−8.2
37.	Physoperuvine	−5.0
38.	Piperine	−7.9
39.	Proline-betaxanthin	−7.5
40.	Salsolinol ((-)-)	−6.1
41.	Scopolamine	−7.8
42.	Solanidine	−8.8
43.	Solasodine	−9.0
44.	Sparteine	−7.2
45.	Theobromine	−5.8
46.	Theophylline	−5.7
47.	Tomatidine	−8.8
48.	Trichostachine	−7.6
49.	Trigonelline	−4.8
50.	Tryptophan-betaxanthin	−8.1
51.	tyrosine-betaxanthin	−8.2
52.	Valine-betaxanthin	−7.8
53.	Vicine	−7.1
54.	Zeatin (trans-)	−7.1

*Control drugs

The best control drug and 5 alkaloids with the utmost binding potential with β-secretase were subjected to evaluating drug-likeness and medicinal chemistry properties using the SwissAdme server. The results are presented in Tables 3 and 4. In addition, the molecular interaction analysis was assessed using PyMOL molecular visualization and LigPlot⁺ software for 3D and 2D presentations respectively (Figure 3). The analysis of the protocol validation showed our docking protocol was okay for use (Figure 3(a)). The binding poses of alkaloids and galantamine were observed to take a similar spot in the binding region of β-secretase (Figure 3(b)). Figures 3(c) to (h) present the 2D interaction analysis of compounds with amino acid residues of the target protein.

Figure 3.

Virtual docking of some alkaloids against β-secretase utilizing Auto Dock Vina, PyMOL, and LigPlot⁺ software based on the methodology. (a) Validation of virtual docking protocol. This process is crucial for the enhancement of the reliability of a computational docking experiment. Comparing the binding postures of co-crystallized ligand (green) with re-docked ligand (cyan) in stick representation. Molecular docking protocol exactly regenerated the binding conformation of crystallographically determined protein-ligand complex with an RMSD value of 0.000 Å. (b) The3D binding postures of hit compounds showing their posregiones in the binding site of β-secretase using PyMOL software. Carpaine (red), demissidine (purple), galantamine (cyan), solanidine (yellow), solasidine (green), and tomatidine (blue) were found to occupy similar spot within the binding region of β-secretase. The 2D molecular interaction analysis was carried out through LigPlot⁺ software and snapshots were taken appropriately. (c) Carpaine; (d) Demissidine; (e) Galantamine; (f) Solanidine; (g) Solasidine; (h) Tomatidine. Interaction analysis shows hydrogen bond (green dashed lines), and hydrophobic interaction (red curved lines) as compounds (purple) interact with the amino acid residues in the catalytic domain of β-secretase.

Table 3.

Druglikeness evaluation of hit compounds from the virtual docking involving β-secretase and 54 compounds.

Molecule	MW	#H-bond acceptors	#H-bond donors	Consensus Log P	Fraction Csp3
Carpaine	478.71	6	2	4.64	0.93
Demissidine	399.65	2	1	5.24	1
Solanidine	397.64	2	1	5.02	0.93
Solasodine	413.64	3	2	4.67	0.93
Tomatidine	415.65	3	2	4.9	1
Galantamine*	287.35	4	1	1.91	0.53

*Control drug

Table 4.

Prediction of the medicinal chemistry properties via the SwissAdme server.

Molecule	Bioavailability Score^#	Brenk #alerts	Leadlikeness #violations	Synthetic Accessibility⁺
Carpaine	0.55	1 (more than 2 esters)	2	6.82
Demissidine	0.55	0	2	4.89
Solanidine	0.55	1 (isolated alkene)	2	5.59
Solasodine	0.55	1 (isolated alkene)	2	6.56
Tomatidine	0.55	0	2	6.43
Galantamine*	0.55	1 (isolated alkene)	0	4.57

*Control drug; #Bioavailability score between 0 to 1.0; +Ease of synthesis between 1 and 10

We carried out the second docking against γ-secretase and the five compounds, galantamine and co-crystallized ligand from the protein data bank. The results obtained here (Table 5) were promising as we observed that demissidine, solasodine, tomatidine, and solanidine have BE of −10.2, −10.4, −11.1, and −10.2 kcal/mol respectively, outsmarting the co-crystallized ligand and galantamine (−9.3 and −7.2 kcal/mol respectively). The molecular interaction analysis and binding poses are provided in Figure 4 similarly done as Figure 3. The implication of having a very low binding energy (kcal/mol) witnessed in the four alkaloids, shows they may likely have higher binding affinity to the two target proteins in our current work.

Figure 4.

Molecular docking simulation of Hit alkaloids against γ-secretase with the aid of Auto Dock Vina, PyMOL and ligPlot⁺ software according to the methodology. (a) Validation of docking simulation protocol. (i) The binding configurations of co-crystallized ligand (yellow) with re-docked ligand (cyan) in stick illustration. Amino acid residues involved in the atoms of the Ligands in questions are rendered via LigPlot⁺. (ii) Co-crystalized L685,458; (iii) Re-docked L685,458. (b) The 3D binding positions of hit compounds showing their positions in the binding site of γ-secretase using PyMOL software. Carpaine (grey), demissidine (orange), galantamine (cyan), solanidine (yellow), solasidine (blue), and tomatidine (purple) were found to occupy similar spot within the binding region of γ-secretase. The 2D molecular interaction analysis was carried out through LigPlot⁺ software and snapshots were taken appropriately. (c) Carpaine; (d) Demissidine; (e) Galantamine; (f) Solanidine; (g) Solasidine; (h) Tomatidine. Interaction analysis shows hydrogen bond (green dashed lines), and hydrophobic interaction (red curved lines) as compounds (purple) interact with the amino acid residues in the catalytic domain of γ-secretase.

Table 5.

Binding energy (kcal/mol) of hit compounds when docked against secretase using Auto Dock Vina

S/No.	Compounds	Binding Energy (kcal/mol)
1	Demissidine	−10.2
2	Carpaine	−7.8
3	Solasodine	−10.4
4	Tomatidine	−11.1
5	Solanidine	−10.2
6	Galantamine*	−7.2
7	L685,458⁺	−9.3

*Control; + co-crystalized molecule

In a similar light, the in silico pharmacokinetic properties were evaluated using the SwissAdme server where properties such as gastrointestinal absorption, blood-brain barrier permeability, substrate of p-glycoprotein, and inhibitors of cytochrome P450 isoforms and excretion through the skin (log Kp). These outcomes are presented in Table 6 which summarily indicates that our compounds are better pharmacokinetically than galantamine.

Table 6.

In silico pharmacokinetics evaluation through the SwissAdme server.

Molecule	MW	GI absorption	BBB permeant	Pgp substrate	CYP1A2 inhibitor	CYP2C19 inhibitor	CYP2C9 inhibitor	CYP2D6 inhibitor	CYP3A4 inhibitor	log Kp (cm/s)
Carpaine	478.71	High	No	No	No	No	No	No	No	−4.75
Demissidine	399.65	High	Yes	No	No	No	No	No	No	−3.83
Solanidine	397.64	High	Yes	No	No	No	No	No	No	−4.4
Solasodine	413.64	High	Yes	Yes	No	No	No	No	No	−5.0
Tomatidine	415.65	High	Yes	Yes	No	No	No	No	No	−4.43
Galantamine	287.35	High	Yes	Yes	No	No	No	Yes	No	−6.75

Though molecular docking studies can offer preliminary insights into protein-ligand interactions, it is necessary to explore these interactions in greater detail MD simulations. To this end, we conducted MD simulations on protein-ligand complexes to observe their dynamic behavior over time in real-time. Specifically, we ran 100-ns MD simulations for demissidine, solasodine, tomatidine, and solanidine, including reference molecules, complexed with both targets. In order to guarantee correct analysis of the ligand-protein interactions, we first assessed the stability of the protein and ligand structures by examining the trajectories for RMSD and RMSF metrics. The outcomes of these analyses are provided in Figures 5 and 6 for β-secretase complexes and Figures 9 and 10 for γ-secretase complexes. However, the interaction maps and detailed 2D interactions were analyzed and presented in Figures 7, 8, 11, and 12 of both protein complexes.

Figure 5.

Root mean square deviation of hit compound β-secretase complexes. (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine. The grey colored trajectory is for the Protein-ligand complexes while the red colored trajectory is for the ligands.

Figure 6.

Root mean square fluctuation patterns of β-secretase bounded to hit compounds. (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine. The green lines show the point of contact of the hit compounds with β-secretase.

Figure 7.

Contact histogram analysis of β-secretase bounded to hit compounds. (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine. Green-colored histograms represent hydrogen bonding, purple-colored histograms are for hydrophobic interactions whereas dark blue and red colored histograms are for water bridges and cationic interactions respectively.

Figure 8.

A detailed 2D projection of the atomic interactions that occurred within the selected trajectory (0 through 100 ns). (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine.

Figure 9.

Root mean square deviation of hit compound-γ-secretase complexes. (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine. The grey colored trajectory is for the apoprotein while the red colored trajectory is for the protein-ligand complex.

Figure 10.

Root mean square fluctuation patterns of γ-secretase bounded to hit compounds. (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine. The green lines show the point of contact of the hit compounds with β-secretase.

Figure 11.

Contact histogram analysis of γ-secretase bounded to hit compounds. (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine. Green-colored histograms represent hydrogen bonding, purple-colored histograms are for hydrophobic interactions whereas dark blue and red-colored histograms are for water bridges and cationic interactions respectively.

Figure 12.

A detailed 2D projection of the atomic interactions that occurred within the selected trajectory (0 through 100 ns). (a) Demissidine; (b) Galantamine; (c) Solanidine; (d) Solasidine; (e) Tomatidine.

Discussion

AD has been a great socioeconomic concern that poses an enormous expense to individuals along with their relatives. Several complex cascades of events result from a progressive accumulation of beta-amyloid attributed to factors such as loss of neuronal synapses, neuronal cell death, and progressive neurotransmitter deficits, thus contributing to the clinical symptoms of dementia.²⁸

The development of β-secretase and γ-secretase inhibitors portray an appealing therapeutics potential for AD. β-secretase (BACE1) is a type-1 membrane-anchored aspartyl protease responsible for the processes of Aβ peptide in the human brain in patients with AD.^29,30 The cleavage performed by β-secretase led to the activity of γ-secretase on the substrate AβPP to generate an intracellular domain (AICD) and a 48- or 49-residue transmembrane peptide (Ab48 or Ab49).³¹ Accumulation of Aβ peptides of varying lengths and byproducts leads to the formation of Aβ plaques, a fingerprint of AD.^32,33 However, the success of some β-secretase and γ-secretase small molecules inhibitors or modulators has been reported to depict inability to cross the blood-brain barrier, low efficacy or severe side effects.³⁴ Naturally occurring alkaloids such as galantamine and huperzine A have long been used as therapeutic means of providing symptomatic relief to AD patients.³⁵

In the current study, the virtual screening strategic step done was to identify selected plant alkaloids that are druggable and could serve as an option in treating AD. Thus, a total of 18 alkaloids exhibited higher affinities when compared to the selected approved drug galantamine and rivastigmine after docking them against β-secretase. All the selected hit compounds, carpaine, solasodine, tomatidine, solanidine, and demissidine including galantamine from the virtual docking involving β-secretase, were subjected to drug-likeness screening. In particular, all the 5 alkaloids show overall good drug-likeness attributes such as low molecular weight (< 500), consensus Log P (CLogP less than 5), number of hydrogen bond acceptors (< 10), and number of hydrogen bond donors (not more than 5) (Table 3). These parameters were set to predict the degree of absorption of the compounds to serve as an oral drug according to the Lipinski Rule of five (R05).³⁶ The bioavailability score, pan assay interference compound alert (PAINS alert), lead likeness, synthetic accessibility, and Brenk alerts are considered for the prediction of their medicinal chemistry properties of the 5 selected hit compounds with the control drug (galantamine) yielded a positive output (Table 4). This predicts the ability of the lead molecules to occur as promiscuous compounds,³⁷ related to PAIN's alert, to deliver druglike potential,³⁸ ability to synthesize the hit compounds easily,³⁹ and structural alert.⁴⁰

Thus, this will play a crucial role in the selection process of the compounds for biological evaluation and synthesis of the hit molecules that show a structural-activity relationship with regards to their safety and efficacy during in vivo and in vitro testing. Noteworthy, the molecular docking conducted with the selected 5 ligands after the first docking period was against γ-secretase. This is to have hit compounds that can serve as dual inhibitors of both β-secretase and γ-secretase. The binding energy scores of the ligand-receptor complex exhibited here were high. Thus, with the dual virtual docking approach in this study, we found out that five ligands, carpaine, solasodine, tomatidine, solanidine, and demissidine, are the best inhibitors of β-secretase and γ-secretase (Tables 2 and 5).

It was reported that patients with AD can benefit from switching to dual inhibitors which help in the stabilization of diseases, reduction in the burden of concomitant psychoactive treatment, and improve the function of cognitive.⁴¹ The major cause of clinical attrition in drug development in the early 1990s was known to be pharmacokinetics and poor drug metabolism.⁴² Early evaluation of absorption, distribution, metabolism, excretion, and toxicity (ADMET) using in silico approach helps to predict models for pharmacokinetics of lead compounds to be effective as drugs. Therefore, correct ADMET properties for drug candidates are very important in drug discovery. The in silico pharmacokinetics evaluation for all the 5 selected lead molecules was done through the SwissAdme server (Table 6). Gastrointestinal absorption is one of the most important factors to be considered in ADMET properties evaluation because the most frequently used route of drug delivery is the oral delivery system, and it is difficult for most oral drugs to cross the intestinal epithelial barrier which determines the bioavailability.

Here, all the 5 selected ligand molecules have a high gastrointestinal absorption rate. Thus, this could increase the efficacy of the drug. Another important ADMET parameter considered is the blood-brain barrier permeability since AD is a neurodegenerative disorder. Therefore, a good drug candidate should be able to cross the blood-brain barrier. All the hit compounds except carpaine were predicted to cross the blood-brain barrier. This show that the lead molecules are promising as the success of some of the reported β-secretase and γ-secretase inhibitors possess poor blood-brain barrier penetration, low efficacy, and severe side effect, a quick need for wet laboratory experimentation. It was reported that the isoforms of CYP enzymes including 1A2, 2C19, 2C9, 2D6, and 3A4 were responsible for about 90% of the oxidative metabolic reactions.⁴³ Also, the higher a small molecule inhibits these CYP isoforms, the higher the chances of involving in drug-drug interaction with other drugs.⁴⁴ None of the molecules were predicted to be inhibitors of CYP enzyme isoforms.

Thus, this shows that these drug candidates can be used alone or with other drugs without resulting in any adverse effects or therapeutic failure. P-glycoproteins are proteins involved in transporting many drugs through the cell membrane and they are very important in intestinal absorption, brain penetration, drug metabolism, and excretion.⁴¹ In addition, P-gp substrate or inhibition may affect the bioavailability and safety of drugs inside the cell. In this study, we included Pgp substrate in here as it helps to minimize the cost and time with integration with in vitro evaluation. However, solasodine and tomatidine were predicted to none-substrate. This shows that these two compounds may have the potential of overcoming multidrug resistance.⁴¹

Our study showed that only solasodine bound with β-secretase by forming interaction with one of the key amino acids among the catalytic dyad residue, Asp 32 (Figure G ii). This indicates an important contribution to the cleavage of the amyloid precursor protein. It is also noteworthy, that the Tyr 71 residue contributes to the inhibitor binding at the active site by changing its conformation.⁴⁵ Carpaine was able to exhibit 8 interactions with the amino acid residue within the binding cavity of β-secretase: Thr72, Tyr71, Gly230, Leu30, Trp115, Phe108, Gln73, and Ile110. Also, it formed 12 hydrophobic interactions within the binding portion of γ-secretase: Leu166, Leu286, Gly382, Asp257, Leu 268, Ile143, Ile253, Gly384, Phe388, Ile387, Leu383, and Trp165. Demissidine contributes one hydrogen bond with both β-secretase and γ-secretase: Ser325 and Gln276, respectively. Similarly, demissidine established 7 and 9 hydrophobic interactions with β-secretase and γ-secretase: Ile110, Phe108, Leu30, Trp115, Tyr71, Gln73, Thr72; and Gly382, Pro433, Val261, Ala431, Lys380, Leu282, Val272, Ile287, and Asp385, respectively.

Phe108, Trp115, Leu30, Gln12, and Ile110 interact with galantamine hydrophobically, and Gln73 via hydrogen bond with β-secretase; while Leu381, Ile287, Val272, Lys380, Leu425, and Ala430 interact with galantamine through hydrophobic interaction; and Gly382 and Leu432 with hydrogen bond interaction for γ-secretase. Solanidine established hydrophobic interaction with Phe108, Tyr71, Leu30, Trp115, Gln73, and 1 hydrogen bond interaction with Gln276 amino acid residues of β-secretase, while Val261, Pro433, Ala431, Lys380, Leu282, Val272, Ile287, Gly382, and Asp385 formed hydrophobic interaction in the catalytic domain of γ-secretase, in addition to one hydrogen bond with Gln276. Hydrophobic bond interaction was observed with 4 and 9 amino acid residues within the catalytic domain of β-secretase and γ-secretase for solasodine: Tyr71, Ile118, Asp32, and Ile110; and Asp385, Gly382, Lys380, Ala431, Val272, Leu282, Ile287, Leu268, and Gly384, respectively. However, hydrogen bond interaction with solasodine was established only with β-secretase through Gly11.

Tomatidine formed hydrophobic bond interaction with the amino acid residue of β-secretase: Gly230, Thr 231, Thr 232, Asn 233, Tyr71, and Thr72; while hydrophobic bond interaction was established with Gly384, Pro433, Ala431, Lys380, Leu282, Val272, Ile287, Leu268, and Gly382; and hydrogen bond interaction with Asp385 was formed with the amino acid residue within the catalytic domain of γ-secretase. It is imperative to know that solasodine, tomatidine, solanidine, and demissidine are Solanum alkaloids. Solanum alkaloids have been a topic of interest since decades. They were reported to have exhibited promising biological potentialities including anticancer activities, neurotoxicity and neuroprotective activity, anti-inflammatory effects, and antimicrobial activity both in vitro and in vivo. However, these best molecules were found to interact with key amino acids in the active site of β- and γ-secretases and were predicted to be safe. Therefore, these lead molecules can be investigated further to develop potent and safe anti-AD drugs.

RMSD has been widely recognized as a key metric for assessing the conformational integrity of protein-ligand complexes over time.⁴⁶ The fluctuations in RMSD values over time indicate that the protein structure may undergo conformational alterations due to the binding of the inhibitor as seen in Figures 5 and 9. Lower RMSD values are typically interpreted as a ratification of conformational stability, signifying minimal structural perturbations upon ligand binding, in contrast, higher RMSD values suggest significant structural changes.47 In this study, none of the protein-ligand complexes returned RMSD values above 2.5 Å (β-secretase complexes, Figure 5). An indication that there is structural stability as the hit compounds bind the active site of the proteins. However, the RMSD values returned for γ-Secretase complexes were between 3.5 A and 7.4 Å.

For the RMSF values of both complexes, a consistent observation through complex simulations (Figures 6 and 10) is the relative flexibility of the intervening loop regions without compromising overall complex stability. These fluctuations are markedly isolated from the active binding site, suggesting a spatial segregation of dynamic behavior within the protein. Meaningfully, fluctuations at the binding sites (green vertical lines) remain within a restrained amplitude, not exceeding 4.0 Å, indicative of a balanced dynamic state for both protein complexes. This controlled dynamism is in concordance with the RSMD values observed in the structural stability plots from Figures 5 and 9. There, the stability within the core binding area maintained below 3.5 Å throughout the 100-ns simulation trajectory, underscoring the protein's ability to sustain its functional conformation despite localized fluctuations elsewhere.

In Figures 7, 8, 11 and 12, there were range of interactions observed during the course of simulation, highlighting hydrogen bonds in green, hydrophobic interactions in purple, ionic interactions in magenta, and water bridges in blue. The stacked bar outlines in Figures 7 and 11 excellently communicated the distribution of simulation time dedicated to sustaining each specific interaction. For example, a value of 0.35 indicated that 35% of the simulation time was devoted to maintaining a particular interaction. In certain cases, values exceeding 1.0 were plausible, indicating multiple contacts between a ligand and the same protein sub-class due to structural conformational changes. Moreover, Figures 8 and 12 presented 2D interaction charts for the top four molecules, providing insights into the persistence of contacts across the entire simulation trajectory.

Interestingly, during and after the MD simulation, some key amino acid residues were found among those that interacted with the atoms of our four hit compounds bound to β-secretase, such as Tyr71, Phe108, Gly230, Thr232 and Trp115 (Figure 8). They were also found interacting with the compounds after molecular docking (Figure 3(c) to (h)). similarly, the atoms of compounds bound to γ-Secretase were observed to interact with Lys380, Gly382, Gln276, Asp3855, Ile287 and Leu432 during and after 100 ns MD simulation (Figure 12). These amino acids were also involved in contributing to the binding mechanism of these four compounds earlier in our docking experiment. (Figure 4(c) to (h))

Conclusion

Finally, phytocompounds, especially alkaloids, were known for their therapeutic value applications, sometimes considered to reduce AD. In this current study, we analyzed 54 alkaloids in stepwise computational virtual docking approaches, pharmacokinetics evaluation, drug-likeness prediction, and medicinal chemistry evaluation to identify potential dual inhibitors of β-secretase and γ-secretase with good efficacy and safety. The binding energy values of carpaine, solasodine, tomatidine, solanidine, and demissidine after docking against the active site of β-secretase are −9.4, −8.9, −8.8, −8.8, and −8.9 kcal/mol respectively. However, carpaine was unable to match the same feat against γ-secretase (−.7.8 kcal/mol). But solasodine, tomatidine, solanidine, and demissidine returned −10.4, −11.1, −10.2, and −10.2 kcal/mol respectively. Ultimately, 4 ligands, solasodine, tomatidine, solanidine, and demissidine, were selected as the best potential inhibitors. However, there will be an urgent need for in vitro and in vivo studies with the appropriate model to give more insight into the hit compounds.

Footnotes

Acknowledgments

The authors have no acknowledgments to report.

ORCID iDs

Isreal Ayobami Onifade

Haruna Isiyaku Umar

Ifeoluwapo Deborah Jegede

Bunmi Helen Adeleye

Dorcas Oladayo Fatoba

Ridwan Opeyemi Bello

Sodiq Fakorede

Nike Idowu

Author contributions

Isreal Onifade (Conceptualization; Data curation; Formal analysis; Funding acquisition; Software; Supervision; Validation; Visualization; Writing – original draft; Writing – review & editing); Haruna Umar (Formal analysis; Investigation; Resources; Software; Supervision; Validation; Visualization; Writing – original draft; Writing – review & editing); Abdullahi Aborode (Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Software; Supervision; Visualization; Writing – original draft; Writing – review & editing); Ifeoluwa Jegede (Formal analysis; Funding acquisition; Methodology; Software; Supervision; Validation; Visualization; Writing – original draft); Aeshah Awaji (Data curation; Formal analysis; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing – original draft); Bunmi Adeleye (Formal analysis; Funding acquisition; Resources; Software; Supervision; Validation; Visualization; Writing – review & editing); Dorcas Fatoba (Formal analysis; Funding acquisition; Resources; Supervision; Validation; Writing – review & editing); Ridwan Opeyemi Bello (Resources; Writing – review & editing); Sodiq Fakorede (Resources; Writing – review & editing); Nike Idowu (Resources; Writing – review & editing).

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

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

The data available have been presented in the paper.

References

Andrews

Renton

Fulton-Howard

, et al. The complex genetic architecture of Alzheimer's disease: novel insights and future directions. EBioMedicine 2023; 90: 104511.

Wilson

Segawa

Boyle

, et al. The natural history of cognitive decline in Alzheimer's disease. Psychol Aging 2012; 27: 1008–1017.

Wang

Geng

. Alzheimer's disease hypothesis and related therapies. Transl Neurodegeneration 2018; 7: 1–4.

Gaire

Koronyo

Fuchs

, et al. Alzheimer's disease pathophysiology in the retina. Prog Retin Eye Res 2024; 101: 101273.

Andrade-Guerrero

Santiago-Balmaseda

Jeronimo-Aguilar

, et al. Alzheimer's disease: an updated overview of its genetics. Int J Mol Science 2023; 24: 3754.

Mendez

. Early-onset Alzheimer's disease: nonamnestic subtypes and type 2 AD. Arch Med Res 2012; 43: 677–685.

Korczyn

Grinberg

. Is Alzheimer disease a disease? Nat Rev Neurology 2024; 20: 245–251.

Humpel

. Chronic mild cerebrovascular dysfunction as a cause for Alzheimer's disease? Exp Gerontol 2011; 46: 225–232.

Lesné

Sherman

Grant

, et al. Brain amyloid-β oligomers in ageing and Alzheimer's disease. Brain 2022; 145: 72–76.

10.

Jack

Jr Andrews

Beach

, et al. Revised criteria for diagnosis and staging of Alzheimer's disease: Alzheimer's Association Workgroup. Alzheimers Dement 2024; 20: 5143–5169.

11.

Żukowska

Moss

Subramanian

, et al. Molecular basis of selective amyloid-β degrading enzymes in Alzheimer's disease. FEBS J 2024; 291: 2999–3029.

12.

Eisele

Monteiro

Fearns

, et al. Targeting protein aggregation for the treatment of degenerative diseases. Nat Rev Drug Discov 2015; 14: 759–780.

13.

Wolfe

. γ-Secretase: once and future drug target for Alzheimer's disease. Expert Opin Drug Discov 2024; 19: 5–8.

14.

Ali

Martins

AMC

Alam

, et al. Neuroprotective effects of alkaloids. In: Khan

Aschner

Mirzaei

(eds) Phytonutrients and neurological disorders. London: UK: Academic Press, 2023, pp.245–257.

15.

Rao

Descamps

John

, et al. Ayurvedic medicinal plants for Alzheimer's disease: a review. Alzheimers Res Ther 2012; 4: 22.

16.

De Strooper

. Proteases and proteolysis in Alzheimer disease: a multifactorial view on the disease process. Physiol Review 2010; 90: 465–494.

17.

Yang

Zhou

Guo

, et al. Structural basis of γ-secretase inhibition and modulation by small molecule drugs. Cell 2021; 184: 521–533.

18.

Umar

Siraj

Ajayi

, et al. Molecular docking studies of some selected gallic acid derivatives against five non-structural proteins of novel coronavirus. J Genet Eng Biotechnology 2021; 19: 16.

19.

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.

20.

Bappy

MNI

Robin

Prome

, et al. Subtractive proteomics analysis to uncover the potent drug targets for distinctive drug design of Candidaauris. Heliyon 2023; 9: e17026.

21.

Moin

Patil

Tabassum

, et al. Immunoinformatics approach to design novel subunit vaccine against the epstein-barr virus. Microbiol Spectr 2022; 10: e0115122.

22.

Aborode

Kumar

Olowosoke

, et al. Predictive identification and design of potent inhibitors targeting resistance-inducing candidate genes from E. coli whole-genome sequences. Front Bioinform 2024; 4: 1411935.

23.

Madhavi Sastry

Adzhigirey

Day

, et al. Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments. J Comput Aided Mol Des 2013; 27: 221–234.

24.

Harder

Damm

Maple

, et al. OPLS3: a force field providing broad coverage of drug-like small molecules and proteins. J Chem Theory Comput 2016; 12: 281–296.

25.

Kashyap

Kalaiselvan

Kumar

, et al. Ajmalicine and reserpine: indole alkaloids as multi-target directed ligands towards factors implicated in Alzheimer's disease. Molecules 2020; 25: 1609.

26.

Sabbah

Zhong

. Modeling the protonation states of β-secretase binding pocket by molecular dynamics simulations and docking studies. J Mol Graph Model 2016; 68: 206–215.

27.

Aborode

Pustake

Awuah

, et al. Targeting oxidative stress mechanisms to treat Alzheimer's and Parkinson's disease: a critical review. Oxid Med Cell Longev 2022; 2022: 7934442.

28.

Hakem

Fouad

Arafa

. Gamma secretase as an important drug target for management of Alzheimer's disease: a comprehensive review. Curr Top Med Chem 2024; 24: 109–127.

29.

Imbimbo

Ippati

Watling

, et al. Role of monomeric amyloid-β in cognitive performance in Alzheimer's disease: insights from clinical trials with secretase inhibitors and monoclonal antibodies. Pharmacol Res 2023; 187: 106631.

30.

Nordvall

Lundkvist

Sandin

. Gamma-secretase modulators: a promising route for the treatment of Alzheimer's disease. Front Mol Neurosci 2023; 16: 1279740.

31.

. Plant alkaloids as drug leads for Alzheimer's disease. Neurochem Int 2015; 89: 260–270.

32.

Lipinski

Lombardo

Dominy

, et al. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev 2001; 46: 3–26.

33.

Baell

Holloway

. New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays. J Med Chemistry 2010; 53: 2719–2740.

34.

Danazumi

Umar

. You must be flexible enough to be trained, mr. Dynamics Simulator. Mol Divers 2024; 28: 2731–2733.

35.

Ertl

Schuffenhauer

. Estimation of synthetic accessibility score of drug-like molecules based on molecular complexity and fragment contributions. J Cheminform 2009; 1: 8.

36.

Brenk

Schipani

James

, et al. Lessons learnt from assembling screening libraries for drug discovery for neglected diseases. Chem Med Chem 2008; 3: 435–444.

37.

Broccatelli

Carosati

Neri

, et al. A novel approach for predicting P-glycoprotein (ABCB1) inhibition using molecular interaction fields. J Med Chem 2011; 54: 1740–1751.

38.

Yang

Lou

Sun

, et al. admetSAR 2.0: web-service for prediction and optimization of chemical ADMET properties. Bioinformatics 2019; 35: 1067–1069.

39.

Williams

Hyland

Jones

, et al. Drug-drug interactions for UDP-glucuronosyltransferase substrates: a pharmacokinetic explanation for typically observed low exposure (AUCi/AUC) ratios. Drug Metab Dispos 2004; 32: 1201–1208.

40.

Cheng

Zhou

, et al. Insights into molecular basis of cytochrome p450 inhibitory promiscuity of compounds. J Chem Inf Model 2011; 51: 2482–2495.

41.

Barman

Prabhakar

. Computational insights into substrate and site specificities, catalytic mechanism, and protonation states of the catalytic asp dyad of β-secretase. Scientifica (Cairo) 2014; 2014: 598728.

42.

Umar

Ajayi

Mukerjee

, et al. Discovery of novel HSP27 inhibitors as prospective anti-cancer agents utilizing computer-assisted therapeutic discovery approaches. Cells 2022; 11: 2412.

43.

Ghahremanian

Rashidi

Raeisi

, et al. Molecular dynamics simulation approach for discovering potential inhibitors against SARS-CoV-2: a structural review. J Mol Liquid 2022; 354: 118901.

44.

White

Lawrence

Brough

, et al. Inflammasomes as therapeutic targets for Alzheimer's disease. Brain Pathol 2017; 27: 223–234.

45.

Wisniewski

Drummond

. Developing therapeutic vaccines against Alzheimer's disease. Expert Rev Vaccines 2016; 15: 401–415.

46.

Jiang

Sun

Chen

. Oxidative stress: a major pathogenesis and potential therapeutic target of antioxidative agents in Parkinson's disease and Alzheimer's disease. Prog Neurobiol 2016; 147: 1–19.