Exploring phytochemical inhibitors of protein kinase C alpha for therapeutic targeting of Alzheimer's disease

Abstract

Background

Alzheimer's disease (AD) is characterized by neurodegeneration linked to amyloid-β (Aβ) plaques and tau protein tangles. Protein kinase C alpha (PKCα) plays a crucial role in modulating amyloid-β protein precursor (AβPP) processing, potentially mitigating AD progression. Consequently, PKCα stands out as a promising target for AD therapy.

Objective

Despite the identification of numerous inhibitors, the pursuit of more effective and precisely targeted PKCα inhibitors remains crucial.

Methods

In this study, we employed an integrated virtual screening approach of molecular docking and molecular dynamics (MD) simulations to identify phytochemical inhibitors of PKCα from the IMPPAT database.

Results

Molecular docking screening via InstaDock identified compounds with strong binding affinities to PKCα. Subsequent ADMET and PASS analyses filtered out compounds with favorable pharmacokinetic profiles. Interaction analysis using Discovery Studio Visualizer and PyMOL further elucidated binding conformations of selected compounds with PKCα. Top hits underwent 200 ns MD simulations using GROMACS to validate stability of the interactions. Finally, we propose two phytochemicals, Kammogenin and Imperialine, with appreciable drug-likeliness and binding potential with PKCα.

Conclusions

Taken together, the findings suggest Kammogenin and Imperialine as potential PKCα inhibitors, highlighting their therapeutic promise for AD after further validation.

Keywords

ADMET analysis Alzheimer's disease amyloid-β PASS analysis phytochemicals protein kinase C alpha virtual screening

Introduction

Alzheimer's disease (AD) is a neurodegenerative disease that progressively deteriorates over time. The presence of toxic amyloid-β plaques (Aβ), the appearance of neurofibrillary tangles (NFTs), and the formation of hyperphosphorylated tau are some of the main features of AD, which account for 60% of dementia cases.¹ There aren't any apparent clinical symptoms during the early stages of AD, but other neuropsychiatric symptoms like mood swings and episodes of agitation, confusion, disorientation, and anger are shown in the severe stage of AD.² The presence of toxic Aβ species, NFT, and the resulting neurodegeneration are the distinct hallmarks of AD, which might be triggered by microglial activation, leading to the release of neurotoxins and inflammatory agents.³

The amyloid-β protein precursor (AβPP) is a transmembrane glycoprotein member with a small intracytoplasmic domain, a membrane-spanning region containing the Aβ peptide, and a large additional cytoplasmic domain. The AβPP molecule undergoes cleavage to produce an amyloidogenic fragment when sequences diverge at its internal Aβ location.⁴ Among many isoforms of Aβ, the primary Aβ component in senile plaques is Aβ₄₀, and in the cerebrovascular lesions, Aβ₄₂ makes up the majority of Aβ.^5,6

Proteolytic processing of the AβPP results in distinct pathways, making its metabolic fate crucial to the pathogenesis of AD. In the non-amyloidogenic pathway, α-secretase cleaves AβPP at the cell surface, disrupting the formation of amyloidogenic fragments by releasing soluble AβPPα (sAβPPα). The protein kinase C (PKC) family is essential in directing AβPP through various pathways; their activation elevates the secretion of sAβPP and decreases the secretion of Aβ peptide. PKC is important in AD development, as seen by the lower levels and activity of PKC in AD brains. The choice between non-amyloidogenic and amyloidogenic pathways for AβPP processing is influenced by extracellular signals and second messengers. Phenol esters activate PKC, which increases sAβPP secretion via a-secretase and emphasizes the role of PKC in the regulated processing of AβPP.^7,8

The PKC family has been previously reported for the development of diabetes, cancer,⁹ ischemic heart disease,¹⁰ several autoimmune diseases,¹¹ Parkinson's disease,¹² and AD. The PKC protein family functions in a variety of cell types, such as neurons, endothelial cells, and fibroblasts. They are activated by diacylglycerol (DAG) and calcium (Ca²⁺) and are involved in several signal transduction pathways that are connected to differentiation, apoptosis, and proliferation. Protein kinase C (PKC) is a phospholipid-dependent family of serine/threonine kinases divided into three subfamilies based on structural and activation features, i.e., i) classical PKCs (α, βI, βII, and γ), ii) novel PKCs (δ, ε, η, and θ), and iii) atypical PKCs (ζ, ι, and λ).¹³ PKC isoforms are essential for several cognitive processes, such as memory and learning, and pathological damage in the brain during AD might be caused by PKCα signaling pathways.¹⁴ When α-secretase is activated, AβPP is degraded, facilitating the formation of soluble AβPPα and stopping Aβ accumulation. However, dysfunctional PKCα does not activate α-secretase, leading to disruption of AβPP processing and the resulting accumulation of Aβ.¹⁵

In addition, PKCα activation is controlled through the phosphorylation of three specific residues in its kinase domain. These include Thr497, which is located at the activation-loop site. Thr638, which serves as the autophosphorylation site, and Ser657, found at the hydrophobic C-terminal site. Phosphorylation at these sites is crucial for PKCα to exhibit any significant activity. Ser657 phosphorylation is commonly used to indicate PKCα activation (Figure 1). A previous study in double transgenic mice found that by stimulating AβPPα processing pathways and Aβ degradation, activation of the PKCα isoform can increase Aβ production and associated dementia in AD.¹⁶ Mutations in the PKCα gene can cause a gain of function or an abnormally active enzyme. Excessive PKCα activity may alter AβPP processing and promote the formation of Aβ plaques and neurodegeneration observed in AD may be exacerbated.¹⁷ Research has shown that the M489V variant of PKCα can have a significant impact on the brain phosphoproteome, leading to cognitive impairment and synaptic degeneration in mouse models, suggesting that targeting PKCα may hold promise as a potential therapy for AD.¹⁸

Figure 1.

Schematic diagram of PKCα and its domains structure. The figure was generated through PyMOL using the PDB coordinates with ID: 4RA4.

PKC inhibitors are important for the treatment of cancer and viral infections. Gö 6983, bisindolylmaleimide I, enzastaurin, sotrastaurin, and rottlerin are PKC inhibitors that reduce SARS-CoV-2 replication and support stem cell self-renewal.^13,19 These PKC isoform inhibitors have been extensively studied for their effects on intracellular signaling pathways and disease management. While PKC inhibitors faced setbacks in cancer clinical trials, repurposing them for neurodegenerative diseases like AD presents a promising avenue for therapeutic intervention.²⁰ The increased activity of the complex signaling pathways involving PKCα in AD, leading to synaptic dysfunction, Aβ generation, and cognitive decline, highlights its importance as a target for therapeutic interventions in neurodegenerative diseases such as AD.^21,22

Currently, there is a deficiency in disease-modifying strategies for AD, and the existing therapeutics only offer restricted alleviation for the related deficits. Given the significant and increasing influence on the affected individuals and the wider community, it is essential to identify potential leads for developing effective therapeutics against AD.²¹ Over the centuries, various traditional medical systems have used medicinal plants to treat neurological disorders. Several natural products have been found to be toxic to humans, although most of the natural products characterized are either therapeutic or nontoxic. Therefore, it is important to carefully consider the source and level of toxicity of the natural product. Traditional Indian medical systems such as Ayurveda and Siddha have long used medicinal plants from India to treat a variety of human ailments. A FAIR-compliant, non-redundant, in silico stereo-enabled library of 17,967 phytochemicals is provided by Indian Medicinal Plants, Phytochemistry, and Therapeutics (IMPPAT 2.0), an expanded and improved phytochemical atlas of Indian medicinal plants that we used in our study.^23,24

This study targets PKCα through small-molecule phytochemicals and identifies a few potential inhibitors from the IMPPAT 2.0 library utilizing a structure-based drug-discovery approach. A systematic workflow of structure-based virtual screening is portrayed in Figure 2. There are a few known inhibitors of PKCα, such as 3KZ, midostaurin, and ellagic acid. This study also includes a comparison with the co-crystal ligand of PKCα, 3KZ, to validate the findings.

Figure 2.

Workflow of methods used in the identification of potential inhibitors of PKCα.

Methods

Computer environment and web resources

A Dell^® workstation with a configuration of a 6-core CPU and 32 GB RAM, running on Windows 11, was used to conduct this study. We used a steady power source and a fast-wired Ethernet internet connection. Various computational tools as well as online resources such as InstaDock,²⁵ PyMOL, and PyMod for virtual screening and homology modeling^26,27; UniProt,²⁸ RCSB Protein Data Bank (PDB),²⁹ IMPPAT library³⁰ for data retrieval, Discovery studio³¹ for visualizing 2D interaction, pkCSM,³² SwissADME,³³ and Pass server³⁴ for evaluation of the small molecules were used. GROMACS software package³⁵ was used for MD simulations to validate the binding potential of the elucidated compounds.

Receptor and library preparation

The sequence, structural, and functional information of PKCα was retrieved from the UniProt database. The 3D structure of PKCα was retrieved from the RCSB Protein Data Bank (PDB ID: 4RA4, resolution: 2.63 Å) and homology modeling was performed through MODELLER 10.2 suite embedded in PyMod 3.0 to model the missing residues in the structure. We have extracted the kinase domain from the whole three-dimensional structure using PyMOL. From the IMPPAT database, we downloaded and filtered the phytochemical compound library based on Lipinski's rule of five to ensure that the screened molecules are physiologically active chemicals.

Molecular docking-based virtual screening

Molecular docking-based virtual screening is a useful tool in drug development. Virtual screening facilitates the simple screening of a vast collection of drug-like substances against a predefined target as it is accessible at many free and commercial sources. We employed molecular docking on phytoconstituents from the IMPPAT database and the 3D structure of PKCα in receptor-based virtual screening to find PKCα inhibitors. InstaDock was used in the virtual screening process. Using blind docking, we identified the optimal binding pocket for the molecules by using the whole structure of the receptor protein. InstaDock sets up the docking grid and creates the ligand file and the configuration file. The ligands undergo docking one by one within the defined coordinates in the configuration file.

ADMET prediction

After docking, we used the ADMET characteristics to filter out the compounds that showed high binding affinity. SwissADME (http://www.swissadme.ch/) and pkCSM (http://biosig.unimelb.edu.au/pkcsm) web servers were used to evaluate the ADMET characteristics, and the PAINS filter (pan-assay interference compounds) was employed. The PAINS filter eliminates substances with a stronger tendency to bind to many targets. The ability of ADMET characteristics to profile substances according to their pharmacokinetic properties. Compounds lacking a PAINS pattern and exhibiting significant ADMET characteristics were selected for further evaluation.

PASS analysis

When evaluating the biological properties of a small molecule and its interactions, PASS (Prediction of Activity Spectra for Substances) analysis is an extremely useful approach. After conducting ADMET analysis, the selected compounds are screened using PASS analysis to examine their biological characteristics further. The PASS server provides results with two labels: “probability to be active (Pa)” and “probability to be inactive (Pi).” A higher Pa value for a compound indicates a greater likelihood that the compound possesses the related biological activity.

Interaction analysis

Discovery Studio Visualizer was used to evaluate the two-dimensional interactions of the selected ligands with PKCα following ADMET and PASS analyses. InstaDock was employed to generate various binding conformations of the ligand split files. This docking and interaction analysis identified the optimal binding conformation for each ligand. Additionally, the PyMOL visualizer was used to generate the three-dimensional binding patterns of each ligand with PKCα.

MD simulations

After docking and subsequent analysis, the selected phytochemicals underwent all-atom MD simulation for 200 ns using the GROMACS suite. The simulation was conducted with the CHARMM36 force field^36–38 at a constant temperature of 300 K, where the ligand topology and force field parameters for each small molecule were generated by CGenFF webserver. Each protein-ligand system was solvated in a 10 Å virtual cubic box of water using the gmx solvate module with the simple point charge (SPC) water model. Energy minimization was performed using the steepest descent method, followed by charge neutralization. Under periodic boundary conditions, the temperature of each system was raised from 0 to 300 K during a 100 ps equilibration phase at constant volume. The simulations were validated using quality control metrics such as density, enthalpy, kinetic energy, and volume. The resulting simulations for PKCα and its ligand complexes were further analyzed. GROMACS trajectories were generated to study various parameters, including root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), and solvent-accessible surface area (SASA). Principal component analysis (PCA) and free energy landscape (FEL) analysis were also performed. Graphs and figures evaluating PKCα residual interaction and stability after protein-ligand interaction were plotted using QtGrace.³⁹

Results

Molecular docking-based virtual screening

Phytochemical substances from the IMPPAT database were subjected to molecular docking to determine their binding affinities with the PKCα protein. Out of the 11,708 phytochemical compounds in the IMPPAT library, InstaDock produced docked conformations and binding affinities for each one. The top 20 hits with binding affinities greater than or equal to −10.2 kcal/mol against PKCα were selected for further analysis. The results indicate that these selected phytochemical compounds exhibit a significant affinity for PKCα. For reference, the co-crystal ligand (3KZ) of the PKCα PDB structure was also docked, showing a binding affinity of −9.7 kcal/mol. The binding affinities of the top 20 compounds and the reference molecule are presented in Table 1. All the compounds from the IMPPAT screening demonstrated higher binding affinities than the reference molecule 3KZ, indicating the strong binding potential of these screened molecules against PKCα.

Table 1.

The binding affinities of the selected top 20 hits against PKCα.

S. No.	Compound ID	Molecular Formula	Affinity (kcal/mol)
1.	IMPHY008586	C22H14O6	−10.9
2.	IMPHY011162	C27H42O4	−10.8
3.	IMPHY004234	C27H44O4	−10.8
4.	IMPHY002700	C28H34O6	−10.6
5.	IMPHY010711	C28H38O7	−10.6
6.	IMPHY010989	C28H34O6	−10.6
7.	IMPHY006882	C23H13N3O2	−10.5
8.	IMPHY005667	C30H48O5	−10.5
9.	IMPHY003078	C25H26O6	−10.5
10.	IMPHY002121	C27H40O	−10.3
11.	IMPHY003911	C27H43NO3	−10.3
12.	IMPHY010403	C27H43NO3	−10.3
13.	IMPHY014742	C27H42O4	−10.3
14.	IMPHY002751	C27H40O5	−10.3
15.	IMPHY008834	C28H36O7	−10.3
16.	IMPHY001575	C27H40O4	−10.2
17.	IMPHY007613	C26H28O8	−10.2
18.	IMPHY012275	C27H42O4	−10.2
19.	IMPHY010851	C27H42O4	−10.2
20.	IMPHY011402	C27H42O5	−10.2
21.	3KZ	C24H27FN4O2	−9.7

ADMET properties

Following the molecular docking analysis, the top 20 phytochemical compounds were further screened for their ADMET analysis. The ADMET evaluation is crucial to ensure that the compounds not only bind effectively to PKCα but also possess drug-like properties suitable for therapeutic use. From the ADMET prediction, five compounds were selected due to their favorable attributes and lack of hazardous patterns. These selected compounds demonstrated optimal absorption, distribution, metabolism, and excretion profiles while exhibiting minimal toxicity, making them promising candidates for further development. The detailed ADMET evaluation of these selected hits is presented in Table 2. The ADMET screening process was instrumental in narrowing down the initial pool of 20 top-performing phytochemicals to five potential drug candidates. The selected compounds showed excellent pharmacokinetic properties, including good oral bioavailability and BBB permeability, which is essential for practical therapeutic applications. Furthermore, the excretion profiles of these compounds suggest they can be efficiently removed from the body, reducing the risk of accumulation and potential toxicity.

Table 2.

ADMET properties of the top 5 compounds.

Compound ID	Phytochemical	Absorption	Distribution	Metabolism	Excretion	Toxicity
Compound ID	Phytochemical	GI Absorption	BBB Permeation	CYP2D6 Inhibitor	OCT2 substrate	AMES
IMPHY002121	Kammogenin	High	Yes	No	No	No
IMPHY003911	Imperialine	High	Yes	No	No	No
IMPHY001575	9-Dehydrohecogenin	High	Yes	No	Yes	No
IMPHY004035	Ruscogenin	High	Yes	No	Yes	No
IMPHY004082	Peiminine	High	Yes	No	No	No
3KZ	N/A	High	Yes	No	Yes	Yes

PASS analysis

Evaluating the key biological activities of selected compounds is essential to ensure their effectiveness as therapeutic agents. In this study, the biological activity of the identified compounds was examined using the PASS server. Following PASS analysis, two compoundsImperialine and Kammogeninwere found to exhibit the desired biological activities. The results indicate that these phytochemicals possess activities beneficial for dementia treatment, including neurotrophic factor enhancement, anti-inflammatory effects, and protein kinase inhibition. These properties suggest that Imperialine and Kammogenin can potentially modulate neurodegenerative processes and provide therapeutic benefits for AD. The detailed results of the PASS analysis, including a comparison with the reference co-crystal ligand, are presented in Table 3. The findings underscore the promise of Imperialine and Kammogenin as multi-functional compounds with significant potential for further development in dementia therapeutics.

Table 3.

PASS analysis of the top 2 compounds and known inhibitor.

Compound ID	Phytochemical	Pa	Pi	Activity
IMPHY002121	Kammogenin	0.744	0.002	Neurotrophic factor enhancer
		0.797	0.003	Phosphatase inhibitor
		0.817	0.005	Antiinflammatory
IMPHY003911	Imperialine	0.429	0.037	Dementia treatment
		0.330	0.023	Vascular dementia treatment
		0.304	0.204	Neurotransmitter uptake inhibitor
3KZ	N/A	0.714	0.037	Nootropic
		0.367	0.186	Antineurotic
		0.364	0.041	Antidepressant

Interaction analysis

Discovery Studio Visualizer was used to visualize the 2D interactions of the compounds. Following PASS analysis, PyMOL was employed to confirm the binding and interaction patterns of the two selected compounds and the co-crystallized ligand. The analysis generated 27 docked conformations for the three phytochemicals. Hydrogen bonding and covalent interactions, such as pi-pi bonds, are crucial in drug development because they enhance structural stability, catalysis, and the biological activity of receptors upon ligand binding. Comparative analysis of each ligand with PKCα, using PKCα-3KZ as a reference molecule, revealed unique interactions with significant residues of PKCα (Figure 3). While analyzing in detail, Figure 3A shows the localization of the compounds within the PKCα binding cavity, whereas Figure 3B provides a magnified view of the interactions. A surface potential representation of the compounds within the PKCα binding pocket showed that Kammogenin, Imperialine, and 3KZ occupied the deep binding pocket of the protein (Figure 3C). The compounds showed deep localization and high complementarity, suggesting that they may successfully block the PKCα binding sites and stop them from being accessible to ATP. These visualizations highlight the specific binding interactions and the potential effectiveness of the selected phytochemicals in modulating PKCα activity, underscoring their therapeutic promises.

Figure 3.

PKCα residual interaction with the elucidated compounds Kammogenin (blue), Imperialine (green), and 3KZ (magenta). (A) Localization of compounds in PKCα binding cavity. (B) Magnified interaction. (C) Surface potential representation of compounds with PKCα binding pocket.

The compounds Kammogenin and Imperialine were found to interact significantly with key residues of the PKCα binding site. Figure 4 illustrates the interaction and binding mechanism of these two selected compounds alongside the reference molecule. Both phytochemical compounds demonstrated strong interactions with critical residues of PKCα, including Val420 and Lys368, suggesting their potential as inhibitors of PKCα activity (Figure 4A, B). The interactions between Kammogenin and Imperialine with these specific residues of PKCα resulted in the formation of stable complexes, altering the protein's structure and potentially reducing its functionality. Comparative analysis of the selected compounds with 3KZ revealed areas of improvement in terms of residue interactions (Figure 4C), indicating their enhanced potential as effective inhibitors of PKCα activity.

Figure 4.

The 2D plots of PKCα binding pocket residues and their interactions with the screened molecules. (A) Kammogenin, (B) Imperialine, and (C) 3KZ.

(Table 4)

Table 4.

The chemical properties of the selected compounds along with the reference molecule.

S. No.	Compound ID	Compound name	Affinity (kcal/mol)
1.	IMPHY002121	Kammogenin	−10.3
2.	IMPHY003911	Imperialine	−10.3
3.	3KZ	(1R)-9-[(3S,4S)-1,3-dimethylpiperidin-4-yl]-8-(2-fluorophenyl)-1-methyl-3,5-dihydro[12,4]triazino[3,4-c][1,4]benzoxazin-2(1H)-one	−9.7

MD simulations

We conducted MD simulations to explore the structural dynamics and stability of PKCα-ligand complexes following the virtual screening procedure.^40–44 The two hits from the IMPPAT library were included in these simulations, employing specific docked conformation complexes under defined solvent conditions. Each 200 ns simulation utilized the initial orientation of the chosen compound as the starting position. Throughout the simulations, various structural factors and traits were meticulously tracked and examined. This comprehensive investigation provided insights into how these interactions influenced the protein's stability and conformational changes over time. Additionally, it elucidated how PKCα's dynamic behavior altered both before and after contact with the ligands. By employing MD simulations, we gained a deeper understanding of the molecular mechanisms underlying the interactions between PKCα and the selected ligands, shedding light on their potential therapeutic efficacy and providing valuable insights for drug development efforts.

Structural deviations

We employed RMSD analysis to assess the structural stability and conformational changes of PKCα and its complexes with Kammogenin, Imperialine, and the co-crystal ligand 3KZ over a 200 ns MD simulation. The average RMSD values for PKCα, PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ were 0.27 nm, 0.26 nm, 0.34 nm, and 0.27 nm, respectively (Table 5). Throughout the simulations, PKCα exhibited stable structural dynamics across all complexes, with minimal fluctuations (Figure 5 A). Notably, the PKCα-Kammogenin complex demonstrated heightened stability compared to other complexes and the free form, indicated by lower RMSD values. This observation suggests a more rigid and less variable conformation for PKCα when bound to Kammogenin.

Figure 5.

Structural dynamics of PKCα upon binding with the three selected compounds. (A) RMSD plot of PKCα in complex with Kammogenin, Imperialine, and 3KZ. (B) RMSF plot of PKCα in complex with Kammogenin, Imperialine, and 3KZ. Lower panels show the probability distribution function of values as PDF. # represents a number.

Table 5.

Different MD parameters (average) calculated after 200 ns simulations.

System	RMSD (nm)	RMSF (nm)	Rg (nm)	SASA (nm²)	#H-bonds
PKCα	0.27	0.13	1.9	146.6	162
PKCα-Kammogenin	0.26	0.12	1.9	145.8	166
PKCα-Imperialine	0.34	0.14	2	147.4	160
PKCα- 3KZ	0.27	0.15	1.9	141	168

The RMSF analysis elucidates the local flexibility of each residue within the protein structure. Our analyses, revealed distinctive RMSF profiles for free PKCα, PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ. These profiles exhibit occasional irregular peaks indicative of localized fluctuations around the mean residue positions. The average RMSF values for PKCα, PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ were calculated to be 0.13 nm, 0.12 nm, 0.14 nm, and 0.15 nm, respectively. The ligand Kammogenin binds to the PKCα protein at Val420 and Leu345. At these positions, the RMSF plot coincides with the RMSF of the free PKCα protein, suggesting minimal residual fluctuation. Imperialine interacts with the protein at Lys368, and it also shows minimal residual fluctuation compared to the free structure. Figure 5B illustrates consistent variations in the compounds throughout the analysis. Notably, all three complexes share a similar and stable RMSF pattern, which initiates from amino acid residue 400 and remains constant after that.

Structural compactness

The radius of gyration (Rg) is a metric for assessing the overall conformation and stability of proteins, reflecting their folding into tertiary structures.⁴⁵ Derived from the RMS distances from the collective center of mass of atoms, Rg evaluates the compactness of protein structures, with lower values indicating more stable folding during complex formation. Notably, there was a significant consistency in the average Rg values across free PKCα, PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ. The time-based evolution of Rg, as depicted in Figure 6 A, showcases stable dynamics and folding for all complexes characterized by a consistent minimum Rg. Among the selected small lead-like phytochemical compounds, kammogenin exhibited the highest stable Rg value.

Figure 6.

Structural compactness and folding of PKCα with the three selected compounds. (A) Rg plot and (B) SASA plot of PKCα with Kammogenin, Imperialine, and 3KZ. Lower panels show the probability distribution function values as PDF.

The solvent-accessible surface area (SASA) quantifies the surface area of a protein interacting with its solvent environment, reflecting its level of folding and compactness through hydrophobic and hydrophilic residue interactions.⁴⁶ Despite the simulations, SASA remained largely unaffected. The average SASA values for PKCα, PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ were computed, showing minimal variation (Figure 6B). Specifically, the average SASA values were 146.6 nm², 145.8 nm², 147.4 nm², and 141 nm², respectively (Table 5). The consistent behavior observed in both Rg and SASA metrics underscores the stability and structural integrity of PKCα and its complexes under various conditions. These findings are essential for understanding the molecular mechanisms underlying protein-ligand interactions and can enhance drug discovery efforts targeting PKCα-related pathways.

Dynamics of hydrogen bonds

The analysis of hydrogen bond formation is pivotal for understanding the conformational dynamics of proteins.⁴⁷ In this study, we predicted intramolecular hydrogen bonds for both the free PKCα structure and PKCα bound to Kammogenin, Imperialine, and 3KZ. The folding dynamics of these systems over a 200 ns simulation period are illustrated in Figure 7. The plots indicate that the number of intramolecular hydrogen bonding interactions remains relatively stable throughout the simulations for all four systems (Figure 7). The average number of hydrogen bonds formed before and after binding to Kammogenin, Imperialine, and the co-crystal ligand 3KZ were found to be 162, 166, 160, and 168, respectively (Table 5). To validate the interaction reliability, we calculated the PDF of intramolecular hydrogen bonds. The results affirm the stability of these interactions across all systems throughout the simulation duration. Overall, the plots demonstrate that intramolecular hydrogen bonds within PKCα exhibit stability for the simulation of each of the studied complexes. This suggests robust structural integrity and reliable interactions between PKCα and its ligands, underscoring the potential significance of these interactions in modulating PKCα function.

Figure 7.

Dynamics of intramolecular hydrogen bonds within (A) PKCα, (B) PKCα-Kammogenin, (C) PKCα-Imperialine, and (D) PKCα-3KZ.

Dynamics of intermolecular interactions

The examination of intermolecular bonds resulting from ligand-protein interactions revealed stable interactions across all complexes (Figure 8). Notably, the PKCα-Kammogenin complex exhibited a higher number of intermolecular interactions compared to other complexes, with two of these interactions remaining stable throughout the 200 ns simulation (Figure 8A). In the temporal evolution, PKCα-Imperialine demonstrated up to four intermolecular interactions, with two robust connections (Figure 8B). Conversely, the co-crystalized ligand displayed a maximum of eight intermolecular connections (Figure 8C). These stable intermolecular interactions contribute to maintaining the initial docking positions of the complexes throughout the simulation period. The presence of stable intermolecular interactions underscores the structural integrity and reliability of ligand-protein interactions within the complexes. These findings provide valuable insights into the dynamics of ligand binding and its impact on the stability of protein-ligand complexes, which are crucial considerations in drug discovery and molecular design endeavors.

Figure 8.

Time evaluation and stability of intermolecular hydrogen bonds formed between PKCα and (A) Kammogenin, (B) Imperialine, and (C) 3KZ. The lower panels show the probability distribution function plot.

Evaluation of secondary structure

Evaluating the conformational behavior of a protein and its dynamics can be accomplished by examining the changes of its secondary structure content. After binding to Kammogenin, Imperialine, and 3KZ, we computed the secondary structural changes in PKCα. After 200 ns of simulation, all the systems show balanced secondary structure content with no significant change (Figure 9). In the cases of PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ complexes, the average number of residues involved in secondary structure formation was not distinctly different from that of free PKCα (Figure 9 and Table 6). These findings suggest that binding to Kammogenin, Imperialine, and 3KZ does not induce substantial changes in the secondary structure content of PKCα, indicating the maintenance of its structural integrity despite ligand interactions. These insights are crucial for understanding the molecular mechanisms underlying protein-ligand interactions and can suggest rational drug design strategies targeting PKCα-related pathways.

Figure 9.

Secondary structure analysis. (A) Free PKCα and PKCα in complex with (B) Kammogenin, (C) Imperialine, and (D) 3KZ during simulation. # represents the number.

Table 6.

Number of amino acid residues participating in secondary structure content of free PKCα and PKCα in complex with Kammogenin, Imperialine, and 3KZ during simulation.

Component	PKCα	PKCα-Kammogenin	PKCα-Imperialine	PKCα-3KZ
Coil	66	59	60	61
β-sheet	37	38	36	40
β-bridge	6	5	6	5
Bend	29	30	28	27
Turn	26	26	28	31
α-helix	93	93	93	88
π-helix	0	0	0	0
3₁₀-helix	2	3	4	5
$κ$ -Helix	0	5	4	2

Principal component analysis

Protein and protein-ligand complex conformational changes and collective movements were evaluated using PCA.⁴⁸ PCA was employed to explore the conformational dynamics of both free PKCα and its complexes with Kammogenin, Imperialine, and 3KZ using simulated trajectories. Figure 10 illustrates the conformational sampling by projecting the Cα atoms onto the principal components. Remarkably, the crucial subspaces of the free PKCα structure are closely aligned with those of the protein-ligand complexes. However, all complexes exhibited structural instability, extending beyond the eigenvectors (EVs) observed in the free PKCα structure. The PKCα-Kammogenin complex covered a smaller subspace in both EV1 and EV2 compared to the other complexes. This observation suggests that the PKCα-Kammogenin complex displays higher stability, as it occupies a more confined conformational space in the PCA. Overall, PCA analysis highlights the conformational dynamics and stability differences between free PKCα and its complexes with various ligands.

Figure 10.

Conformational subspace in the principal component analysis of (A) PKCα (B) PKCα-Kammogenin, (C) PKCα-Imperialine, and (D) PKCα-3KZ.

Free energy landscape analysis

We employed FEL analysis to delve into the energetics and folding dynamics of protein-ligand complexes under various solvent conditions, aiming to identify global and local energy minima.⁴⁹ It is a crucial computational tool in thermodynamics and statistical mechanics, and it provides deep insights into the behaviour of biological systems. It gives understanding about the conformational dynamics and provides understanding about different energy substates. The Gibbs free energy landscape shows how binding of a ligand affects the stability of protein, implicating the affinity and specificity of binding. The FELs provide insights into the stability and folding pathways of the complexes.⁵⁰ Figure 11 displays the FELs of the free PKCα and the complexes PKCα-Kammogenin, PKCα-Imperialine, and PKCα-3KZ. In the FEL diagrams, deep blue hues represent extremely low energy levels, indicating proximity to native states. In the free state of PKCα, one large basin, corresponds to a global minimum and a single minimum (Figure 11A). Upon binding with Kammogenin, three separate basins were detected, whereas binding with Imperialine revealed three smaller basins and one large basin (Figure 11 B, C). Similarly, the complex with the co-crystal ligand 3KZ displayed four distinct and distributed basins (Figure 11D). The achievement of global minima in all the cases indicated that the screened molecules had minimal impact on the PKCα folding mechanism. The FEL analysis demonstrates that the binding interactions between Kammogenin and Imperialine with PKCα did not induce protein unfolding throughout the 200 ns simulation. This indicates the stability and structural integrity of the PKCα-ligand complexes, supporting their potential as promising leads for further research and therapeutic advancement of AD.

Figure 11.

The Gibbs free energy landscapes of (A) free PKCα, (B) PKCα-Kammogenin, (C) PKCα-Imperialine, and (D) PKCα −3KZ.

Discussion

This study was aimed to discover potential inhibitors of PKCα using advanced computational methods. In a previous study, 18 concentrate agave sap samples were studied for their physiochemical and antiproliferative properties where Kammogenin showed anticancer potential.⁵¹ In another study, to identify apoptotic saponins from an acetonic extract of concentrated agave sap, fast centrifugal partition chromatography was employed in conjunction with in vitro tests on HT-29 cells as a screening procedure and Kammogenin was reported for its apoptotic potential.⁵² Imperialine derived from Bulbus Fritillaria cirrhosa D. Don (BFC) was shown to exhibit anti-cancer effects against non-small lung cancer both in vitro and in vivo. This is linked to an inflammation-cancer feedback loop that centered around NF-κB.⁵³ A study on Imperialine isomers, steroidal alkaloids from the cevanine group, derived from Fritillaria wabuensis bulbs, suggests that Imperialine may hold promise for developing treatments for inflammatory diseases.⁵⁴

The reference molecule 3KZ showed a docking score of −9.7 kcal/mol. The most favorable conformation was selected by carefully analyzing each docked position of the compounds compared to 3KZ. The results suggest that Imperialine and Kammogenin have an interaction at the ATP-binding sites of PKCα, particularly at Lys368, Asp481, and Val420, along with several other notable residues, with a docking score of −10.3 kcal/mol for both. Detailed interaction analyses were performed to improve our understanding of the noncovalent interactions between Imperialine and Kammogenin and the co-crystallized reference inhibitor 3KZ with PKCα. The MD simulation study also reflects the correlation between the two ligands and the co-crystal ligand 3KZ with comparable structural deviation, compactness, and FEL. Taken together, the study suggested that Imperialine and Kammogenin showed high potential to be used as promising leads for therapeutic development against PKCα-associated neurodegeneration, including AD.

Conclusions

PKCα has emerged as a promising target for AD therapy. Despite the identification of numerous inhibitors, the quest for more effective and precisely targeted PKCα inhibitors remains crucial. In our study, we employed in-silico methods, encompassing virtual screening and all-atom MD simulations. From phytochemicals extracted from the IMPPAT database, two compounds, Kammogenin and Imperialine, were selected via virtual screening, with the co-crystal ligand 3KZ serving as a reference molecule. Kammogenin and Imperialine exhibited notable selectivity and robust binding affinity for the active site residues of PKCα. The results of PCA, FEL analyses, and time-evolution observations in all-atom MD simulations indicate that Kammogenin and Imperialine establish stable and promising binding interactions with PKCα. These compounds hold significant potential in unlocking novel therapeutic avenues by addressing the complexities associated with PKCα inhibition. Moving forward, rigorous in vitro validation of these compounds is indispensable in the development of effective therapeutics aimed at targeting PKCα and advancing AD treatment strategies.

Footnotes

Acknowledgments

MIH acknowledges the Council of Scientific and Industrial Research for financial support [Project No. 27(0368)/20/EMR-II]. A.S. acknowledges Ajman University, UAE for supporting the publication. MFA and AH acknowledge the researchers supporting project number (RSP2024R122) at King Saud University, Riyadh, Saudi Arabia.

ORCID iDs

Afzal Hussain

Anas Shamsi

Md Imtaiyaz Hassan

Author contributions

Azad (Conceptualization; Data curation; Formal analysis; Validation; Writing – original draft); Arunabh Choudhury (Data curation; Investigation; Methodology; Project administration; Visualization; Writing – review & editing); Afzal Hussain (Data curation; Methodology; Project administration; Resources; Writing – review & editing); Mohamed F AlAjmi (Data curation; Project administration; Software; Validation; Visualization); Taj Mohammad (Data curation; Project administration; Supervision; Writing – original draft); Sneh Prabha (Data curation; Formal analysis; Methodology; Software; Writing – original draft); Manoj Kumar Sharma (Conceptualization; Formal analysis; Investigation; Methodology; Writing – review & editing); Anas Shamsi (Data curation; Methodology; Supervision; Writing – review & editing); Md Imtaiyaz Hassan (Conceptualization; Formal analysis; Investigation; Project administration; Supervision; Writing – review & editing).

Funding

This work is supported by Central Council for Research in Unani Medicine (CCRUM), Ministry of AYUSH, Government of India (Grant No. 3-69/2020- CCRUM/Tech). Researchers supporting project number (RSP2024R122) at King Saud University, Riyadh, Saudi Arabia.

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 underlying this article is available within the manuscript.

References

Selkoe

. Alzheimer's disease: genes, proteins, and therapy. Physiol Reviews 2001; 81: 741–766.

Fantuzzi

. Agitation and aggressiveness in Alzheimer’s disease: focus on neurotransmission mechanisms and pharmacological treatments. Università di Parma. Dipartimento di Medicina e Chirurgia, Parma, Italy, 2024.

Jolly-Tornetta

Wolf

. Regulation of amyloid precursor protein (APP) secretion by protein kinase cα in human ntera 2 neurons (NT2N). Biochemistry 2000; 39: 7428–7435.

Bayer

Cappai

Masters

, et al. It all sticks together—the APP-related family of proteins and Alzheimer's disease. Mol Psychiatry 1999; 4: 524–528.

Ekundayo

Obafemi

Adewale

, et al. Oxidative stress, endoplasmic reticulum stress and apoptosis in the pathology of Alzheimer’s disease. Cell Biochem Biophys 2024; 82: 457–477.

Kumar

Singh

. A review on Alzheimer's disease pathophysiology and its management: an update. Pharmacol Rep 2015; 67: 195–203.

Racchi

Mazzucchelli

Pascale

, et al. Role of protein kinase calpha in the regulated secretion of the amyloid precursor protein. Mol Psychiatry 2003; 8: 209–216.

Chen

Zhu

Xiong

, et al. PKC And Ras are involved in M1 muscarinic receptor-mediated modulation of AMPA receptor GluA1 subunit. Cell Mol Neurobiol 2020; 40: 547–554.

Inagaki

Churchill

Mochly-Rosen

. Epsilon protein kinase C as a potential therapeutic target for the ischemic heart. Cardiovasc Res 2006; 70: 222–230.

10.

Dempsey

Newton

Mochly-Rosen

, et al. Protein kinase C isozymes and the regulation of diverse cell responses. Am J Physiol Lung Cell Mol Physiol 2000; 279: L429–L438.

11.

Zhang

Anantharam

Kanthasamy

, et al. Neuroprotective effect of protein kinase Cδ inhibitor rottlerin in cell culture and animal models of Parkinson's disease. J Pharmacol Exp Ther 2007; 322: 913–922.

12.

Landgraf

Malmberg

Falke

. Effect of PIP2 binding on the membrane docking geometry of PKCα C2 domain: an EPR site-directed spin-labeling and relaxation study. Biochemistry 2008; 47: 8301–8316.

13.

Kawano

Inokuchi

Eto

, et al. Protein kinase C (PKC) isozymes as diagnostic and prognostic biomarkers and therapeutic targets for cancer. Cancers 2022; 14: 5425.

14.

Alkon

Sun

M-K

Nelson

. PKC Signaling deficits: a mechanistic hypothesis for the origins of Alzheimer's disease. Trends Pharmacol Sci 2007; 28: 51–60.

15.

de Barry

Liégeois

Janoshazi

. Protein kinase C as a peripheral biomarker for Alzheimer’s disease. Exp Gerontol 2010; 45: 64–69.

16.

Zhao

, et al. Inhibition of PKCδ reduces amyloid-β levels and reverses Alzheimer disease phenotypes. J Exp Med 2018; 215: 1665–1677.

17.

Lordén

Wozniak

Doré

, et al. Enhanced activity of Alzheimer disease-associated variant of protein kinase cα drives cognitive decline in a mouse model. Nat Commun 2022; 13: 7200.

18.

Callender

Yang

Lordén

, et al. Protein kinase Cα gain-of-function variant in Alzheimer’s disease displays enhanced catalysis by a mechanism that evades down-regulation. Proc Natl Acad Sci U S A 2018; 115: E5497–E5505.

19.

Huang

Feng

Shi

, et al. Protein kinase C inhibitors reduce SARS-CoV-2 replication in cultured cells. Microbiol Spectr 2022; 10: e01056–e01022.

20.

Ghosh

Durgvanshi

Agarwal

, et al. Current status of drug targets and emerging therapeutic strategies in the management of Alzheimer's disease. Curr Neuropharmacol 2020; 18: 883–903.

21.

Singh

Nandy

Jyoti

, et al. Protein kinase C (PKC) in neurological health: implications for Alzheimer’s disease and chronic alcohol consumption. Brain Sci 2024; 14: 554.

22.

Khan

Hassan

Shahid

, et al. Nature's toolbox against tau aggregation: an updated review of current research. Ageing Res Rev 2023; 87: 101924.

23.

Vivek-Ananth

Mohanraj

Sahoo

, et al. IMPPAT 2.0: an enhanced and expanded phytochemical atlas of Indian medicinal plants. ACS Omega 2023; 8: 8827–8845.

24.

Mohanraj

Karthikeyan

Vivek-Ananth

, et al. IMPPAT: a curated database of Indian medicinal plants, phytochemistry and therapeutics. Sci Rep 2018; 8: 4329.

25.

Mohammad

Mathur

Hassan

. Instadock: a single-click graphical user interface for molecular docking-based virtual high-throughput screening. Brief Bioinform 2021; 22: bbaa279.

26.

Janson

Paiardini

. Pymod 3: a complete suite for structural bioinformatics in PyMOL. Bioinformatics 2021; 37: 1471–1472.

27.

DeLano

. The PyMOL molecular graphics system. San Carlos: Delano Scientific, 2002, http://www pymol org/ .

28.

UniProt Consortium. Uniprot: a worldwide hub of protein knowledge. Nucleic Acids Res 2019; 47: D506–D515.

29.

Rose

Prlić

Altunkaya

, et al. The RCSB protein data bank: integrative view of protein, gene and 3D structural information. Nucl Acids Res 2017; 45: D271–DD28.

30.

Anjum

Mohammad

Almalki

, et al. Phytoconstituents and medicinal plants for anticancer drug discovery: computational identification of potent inhibitors of PIM1 kinase. OMMICS 2021; 25: 580–590.

31.

Dassault Systèmes. BIOVIA Discovery Studio, 2008.

32.

Pires

Kaminskas

Ascher

. Prediction and optimization of pharmacokinetic and toxicity properties of the ligand. Methods Mol Biol 2018; 1762: 271–284.

33.

Daina

Michielin

Zoete

. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci Rep 2017; 7: 42717.

34.

Parasuraman

. Prediction of activity spectra for substances. J Pharmacol Pharmacother 2011; 2: 52.

35.

Huang

MacKerell Jr

. CHARMM36 All-atom additive protein force field: validation based on comparison to NMR data. J Comput Chem 2013; 34: 2135–2145.

36.

Siveen

Prabhu

Achkar

, et al. Role of non receptor tyrosine kinases in hematological malignances and its targeting by natural products. Mol Cancer 2018; 17: 1–21.

37.

Sahu

August

. ITK Inhibitors in inflammation and immune-mediated disorders. Curr Top Med Chem 2009; 9: 690–703.

38.

Huang

Bai

, et al. A unique role for ITK in survival of invariant NKT cells associated with the p53-dependent pathway in mice. J Immunol 2012; 188: 3611–3619.

39.

Berendsen

van der Spoel

van Drunen

. GROMACS: a message-passing parallel molecular dynamics implementation. Comput Phys Commun 1995; 91: 43–56.

40.

Mohammad

Batra

Dahiya

, et al. Identification of high-affinity inhibitors of cyclin-dependent kinase 2 towards anticancer therapy. Molecules 2019; 24: 4589.

41.

Mohammad

Khan

Lobb

, et al. Identification and evaluation of bioactive natural products as potential inhibitors of human microtubule affinity-regulating kinase 4 (MARK4). J Biomol Struct Dyn 2019; 37: 1813–1829.

42.

Mohammad

Shamsi

Anwar

, et al. Identification of high-affinity inhibitors of SARS-CoV-2 main protease: towards the development of effective COVID-19 therapy. Virus Res 2020; 288: 198102.

43.

Mohammad

Siddiqui

Shamsi

, et al. Virtual screening approach to identify high-affinity inhibitors of serum and glucocorticoid-regulated kinase 1 among bioactive natural products: combined molecular docking and simulation studies. Molecules 2020; 25: 823.

44.

Naqvi

AAT

Mohammad

Hasan

, et al. Advancements in docking and molecular dynamics simulations towards ligand-receptor interactions and structure-function relationships. Curr Top Med Chem 2018; 18: 1755–1768.

45.

Noor

Choudhury

Raza

, et al. Probing baicalin as potential inhibitor of Aurora kinase B: a step towards lung cancer therapy. Int J Biol Macromol 2024; 258: 18.

46.

Ali

Hassan

Islam

, et al. A review of methods available to estimate solvent-accessible surface areas of soluble proteins in the folded and unfolded states. Curr Protein Pept Sci 2014; 15: 456–476.

47.

Shakeel

Khan

Roy

, et al. Investigating potential of cholic acid, syringic acid, and mangiferin as cancer therapeutics through sphingosine kinase 1 inhibition. Int J Biol Macromol 2023; 253: 1.

48.

Stein

SAM

Loccisano

Firestine

, et al. Principal components analysis: a review of its application on molecular dynamics data. Annu Rep Comput Chem 2006; 2: 233–261.

49.

Anwar

Khan

Hussain

, et al. Investigating pyruvate dehydrogenase kinase 3 inhibitory potential of myricetin using integrated computational and spectroscopic approaches. ACS Omega 2024; 9: 29633–29643.

50.

Papaleo

Mereghetti

Fantucci

, et al. Free-energy landscape, principal component analysis, and structural clustering to identify representative conformations from molecular dynamics simulations: the myoglobin case. J Mol Graph Model 2009; 27: 889–899.

51.

Santos-Zea

Rosas-Pérez

Leal-Díaz

, et al. Variability in saponin content, cancer antiproliferative activity and physicochemical properties of concentrated agave sap. J Food Sci 2016; 81: H2069–H2075.

52.

Santos-Zea

Fajardo-Ramírez

Romo-López

, et al. Fast centrifugal partition chromatography fractionation of concentrated agave (Agave salmiana) sap to obtain saponins with apoptotic effect on colon cancer cells. Plant Foods Hum Nutr 2016; 71: 57–63.

53.

Lin

Patra

, et al. Mechanistic and therapeutic study of novel anti-tumor function of natural compound imperialine for treating non-small cell lung cancer. J Ethnopharmacol 2020; 247: 112283.

54.

Xiao

, et al. Imperialine and verticinone from bulbs of Fritillaria wabuensis inhibit pro-inflammatory mediators in LPS-stimulated RAW 264.7 macrophages. Planta Med 2015; 81: 821–829.