In Silico Identification of Lepiotaprocerin C as a Promising PIM-1 Kinase Inhibitor: An Integrated Docking,Molecular Dynamics,MM/PBSA,QSAR,and ADMET Study

Abstract

Proviral Integration site for Moloney murine leukemia virus-1 (PIM-1) kinase, a serine/threonine kinase overexpressed in various malignancies, plays a critical role in promoting cell survival and proliferation, making it a promising target for anticancer therapy. This study employed an integrated in silico approach to evaluate Lepiotaprocerin derivatives (A to L) from Macrolepiota procera as potential PIM-1 inhibitors. Molecular docking of 12 Lepiotaprocerins revealed Lepiotaprocerin C as the most potent compound, exhibiting superior binding affinity (−11.4 kcal/mol) compared with the reference inhibitor AZD1208. Binding site validation using CASTp, PrankWeb, and blind docking confirmed the ATP-binding pocket as the active cavity. The Lepiotaprocerin C-PIM-1 complex demonstrated enhanced stability during 200 ns molecular dynamics simulations, maintaining low RMSD and strong hydrogen-bond interactions, supported by a favorable Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA) binding free energy (−22.0 ± 2.1 kcal/mol). Based on quantitative structure activity relationship (QSAR) analysis, the calculated pIC₅₀ value of Lepiotaprocerin C was 8.67. QSAR modeling (R² = .74, Q² = 0.90) confirmed robust predictive capacity, while absorption, distribution, metabolism, and elimination and PerMM analysis indicated favorable pharmacokinetic and permeability profiles. Prediction of Activity Spectra for Substances and toxicity predictions further revealed high antineoplastic potential (Pa = 0.881) and a nontoxic safety profile. These results highlight Lepiotaprocerin C as a promising, stable, and safe inhibitor of PIM-1 kinase, warranting further in vitro and in vivo validation for potential anticancer drug development.

Keywords

Anticancer activity cell membrane permeability Lepiotaprocerin C molecular docking PIM kinase

Introduction

Cancer remains one of the leading causes of morbidity and mortality worldwide, accounting for nearly 10 million deaths annually according to the World Health Organization.¹ Among various molecular targets implicated in tumorigenesis, Proviral Integration site for Moloney murine leukemia virus-1 (PIM-1) kinase, a serine/threonine kinase, has garnered significant attention. Overexpression of PIM-1 is associated with poor prognosis in several malignancies, including leukemia, prostate, breast, and colorectal cancers. It promotes tumor cell survival, proliferation, and resistance to apoptosis through modulation of cell cycle regulators and anti-apoptotic proteins.^2,3 In experimental models, suppression of PIM-1 reduced cell migration and proliferation, triggered apoptotic cell death, and worked in concert with other chemotherapeutic drugs.⁴ Given its pivotal role in cancer progression, PIM-1 kinase has emerged as a compelling target for selective inhibition in anticancer drug discovery.

Several synthetic inhibitors of PIM-1, including AZD1208^5,6 and SGI-1776,^7-9 have demonstrated preclinical efficacy. However, their clinical translation has been hindered by limited bioavailability, off-target toxicity, and pharmacokinetic constraints. These limitations have redirected focus toward the discovery of safer, natural small-molecule inhibitors with improved selectivity and tolerability profiles. Natural products derived from medicinal plants and fungi offer structurally diverse scaffolds with potent bioactivity and reduced systemic toxicity, making them ideal candidates for anticancer drug development.^10,11

Macrolepiota procera, commonly known as the Parasol mushroom and a member of the Agariaceae family, offers diverse bioactive compounds, including triterpenoids (Lepiotaprocerins), phenolic acids, and alkaloids, which may target complex cellular pathways with potentially reduced systemic toxicity.¹² Its bioactive profile is less explored than that of other mushrooms.¹³ Several isolates from M procera have demonstrated cytotoxic properties, with in vitro IC₅₀ values (6.18 to 80.27 mg/mL) reported in different cancer cell lines, indicating potential anti-proliferative activity.^14,15 In vitro studies also showed anti-inflammatory effects by Lepiotaprocerins constituents, supporting ongoing interest in its therapeutic potential.¹⁶

The present study focuses on Lepiotaprocerin derivatives (A to L), major bioactive triterpenoid from M procera, as potential inhibitors of PIM-1 kinase. The research integrates computational approaches—molecular docking, binding site validation, molecular dynamics (MD) simulations, and Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA) binding energy calculations—to elucidate its structural stability and binding mechanism. In addition, quantitative structure-activity relationship (QSAR) modeling; absorption, distribution, metabolism, and elimination (ADME) profiling; and toxicity prediction were performed to assess its pharmacokinetic feasibility and safety. The study aims to identify the most potent Lepiotaprocerin derivative targeting PIM-1 kinase, validate its dynamic and energetic stability through long-term MD simulations, and evaluate its drug-likeness and biological activity through predictive modeling.

Materials and Methods

Selection of phytochemical

A series of bioactive phytoconstituents, specifically Lepiotaprocerins (A to L) (Figure 1), were selected for this study based on their documented anticancer potential and natural origin from Lepiota species as reported in prior ethnopharmacological and phytochemical literature.^14-18 The selection criteria included previously reported cytotoxicity, structural novelty, and drug-likeness properties. The chemical structures of these compounds, along with a reference drug (AZD1208), were retrieved in .sdf format from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Various articles reported that AZD1208 showed significant PIM kinase inhibitory activity.^19-21 Because of reported PIM kinase inhibitory activity, we used AZD1208 as reference drug in our study.

Figure 1.

Selected Lepiotaprocerin derivatives and reference drug with their PubChem ID and 2D structures.

Compound and target protein preparation

All ligand structures underwent energy minimization using MMFF94 force field and Open Babel v2.4.1 for geometry optimization and conversion into pdbqt format. This step ensured that the ligands attained their lowest energy conformations prior to docking simulations.²²

For the protein target, PIM-1 kinase, a serine/threonine kinase implicated in oncogenesis and cell survival, was selected due to its well-established role in hematological malignancies and solid tumors.⁷ The specific crystal structure (PDB ID: 4JX3) was chosen from the Protein Data Bank (https://www.rcsb.org/) based on its high resolution (2.50 Å).²³ The energy minimization of selected protein was conducted using Chimera v1.16 at default setting. The protein structure was pre-processed by removing all water molecules, ions, and heteroatoms, followed by the addition of polar hydrogens and assignment of Kollman charges using AutoDock Tools v1.5.7. The prepared structure was then saved in one folder for molecular docking simulations.²⁴

Protein dynamics and validation by normal mode analysis

To evaluate the structural flexibility and stability of the target protein, normal mode analysis (NMA) was performed using the iMODS online server (https://imods.iqf.csic.es/).²⁵ The analysis was conducted using default parameters, which included assessment of the deformability, eigenvalues, B-factor (mobility), and covariance matrix of the PIM-1 kinase structure. These parameters collectively provided insights into the intrinsic motions and dynamic behavior of the protein in a simulated environment. Prior to docking, the quality and stereochemical reliability of the selected protein structure (PDB ID: 4JX3) were validated using the PROCHECK online server.²⁶ The Ramachandran plot confirmed the structural integrity and proper folding of the protein by assessing the phi (φ) and psi (ψ) torsion angles of amino acid residues. Furthermore, the physicochemical characteristics of the PIM-1 kinase were analyzed using the Expasy ProtParam tool (https://web.expasy.org/protparam/). Parameters such as molecular weight, theoretical isoelectric point (pI), instability index, aliphatic index, and grand average of hydropathicity (GRAVY) were computed to assess the protein’s biochemical stability.²⁷

Binding pocket prediction

Identification of potential binding sites on the PIM-1 kinase structure was done using the Computed Atlas of Surface Topography of Proteins (CASTp 3.0) server (http://sts.bioe.uic.edu/castp/index.html?1bxw).²⁸ The analysis used the 3-dimensional crystal structure of the protein (PDB ID: 4JX3) to predict and visualize topographical features including solvent-accessible pockets, internal cavities, and surface depressions. The server generated a detailed list of potential binding cavities along with associated surface area, volume, and constituent amino acid residues. The largest and most accessible pocket, located in the ATP-binding region, was selected for subsequent molecular docking based on its structural compatibility with ligand binding.

Binding site validation

Potential binding sites on the PIM-1 kinase structure were identified using a dual-algorithm approach to ensure reliability. The Computed Atlas of Surface Topography of Proteins (CASTp 3.0) served as the primary prediction tool, using alpha complex and discrete flow algorithms to map surface topography and internal cavities.²⁸

For validation, “PrankWeb (https://prankweb.cz/) is a template-free machine learning method based on the prediction of local chemical neighborhood ligandability centered on points placed on a solvent-accessible protein surface.”²⁹ In addition, blind docking validation was conducted using AutoDock Vina v1.2.0 across the entire protein surface without predefined constraints. The grid box encompassed the complete protein structure with exhaustiveness set to 8 to ensure comprehensive sampling.

Binding site validation criteria included (1) high overlap between CASTp, blind docking, and PrankWeb predictions; (2) clustering of blind docking poses within predicted cavities; and (3) conservation of key interaction residues across methodologies. The primary binding cavity, located in the ATP-binding region and validated by all 3 approaches, was selected for targeted molecular docking studies.

Molecular docking methodology

Molecular docking simulations were conducted using AutoDock Vina integrated within the PyRx v0.8 software suite, which incorporates Open Babel for file conversion and ligand optimization.³⁰ The docking procedure aimed to predict the binding affinity and orientation of Lepiotaprocerin compounds (A-L) within the active site of the PIM-1 kinase protein (PDB ID: 4JX3). The optimized structures were also converted to pdbqt format using Open Babel. A site-specific docking approach was employed to explore potential binding modes across the entire binding cavity. The docking grid box was defined with the parameters given in Table 1. We analyzed the cavity dimension of blind docking by generating a grid box around the docked complex after docking and found near similar to other 2 grid dimension. This also confirm the selected binding site is accurate.

Table 1.

Docking grid box dimensions and parameters.

Grid center coordinates
PrankWeb	Blind docking (for docking)	Blind docking (after docking)	CASTp
center_x = 18.35	center_x = 8.41	center_x = 18.31	center_x = 16.94
center_y = −35.96	center_y = −40.33	center_y = −36.59	center_y = −32.96
center_z = 0.70	center_z = 1.48	center_z = -5.47	center_z = 3.89

These coordinates ensured coverage of the major binding pockets predicted by CASTp, particularly around the ATP-binding domain. Docking simulations were performed for each Lepiotaprocerin compound, and the resulting binding affinities (in kcal/mol) and poses were analyzed by BIOVIA Discovery Studio Visualizer 2021 to identify the most promising inhibitor candidates.

MD simulation

Protein-ligand interactions involving the protein structure (PDB ID: 4JX3) were evaluated using molecular docking approach. Following docking, MD simulations were conducted using the AMBER software suite, applying the ff19SB force field to accurately represent protein behavior. The systems were solvated in a 12 Å orthorhombic periodic box using the TIP3P explicit water model to fully encompass the protein-ligand complexes. A 2-step energy minimization protocol was employed, beginning with positional restraints on heavy atoms, followed by an unrestrained minimization to optimize system stability. The minimized systems were gradually heated from 0 K to 300 K under the NVT ensemble, then subjected to equilibration. Equilibration was done over 200 ns, with system snapshots taken every 5 ns to monitor stability. Subsequently, a 200 ns production MD simulation was performed under the NPT ensemble to maintain realistic thermodynamic conditions.³¹ Trajectory analysis was conducted using the CPPTRAJ module from AmberTools v24,³² and protein-ligand complex visualizations were generated with VMD.³³

MM/PBSA binding free energy calculation

To evaluate the binding free energy of the Lepiotaprocerin C-PIM-1 kinase complex following MD simulation, the MM/PBSA method was applied using the MMPBSA.py module from the AMBER Tools suite v24. The topology files for the complex (protein-ligand), receptor (protein), and ligand were generated using LEaP, while the last 20 ns of the production trajectory (extracted every fifth frame) were selected for analysis. Water molecules and counterions were stripped using cpptraj to ensure accurate energy computation. The binding free energy (ΔG_bind) was calculated according to the standard MM/PBSA formula, incorporating molecular mechanics energies (electrostatic and van der Waals interactions), polar solvation energy (using the Poisson-Boltzmann model), and nonpolar solvation energy (estimated from solvent-accessible surface area). The MM/PBSA calculations were done under physiological ionic strength (0.15 M) using default dielectric constants.³⁴

Visualization and interpretation

The resulting energy components and total binding free energy were plotted using XMGrace for visual representation. The time-dependent energy fluctuations and stability trends were analyzed, helping confirm the thermodynamic stability of the complex and the contribution of each interaction to overall binding. Line plots of total ΔG_bind, van der Waals, and electrostatic components were generated to highlight dominant interactions.³⁵

QSAR model development and validation

The methodology of QSAR model development and validation was based on Rosell-Hidalgo et al,³⁶ Nguyen et al,³⁷ and Daoui et al.³⁸

Studied compounds

A total of 28 PIM-1 kinase inhibitors were retrieved from the ChEMBL database (https://www.ebi.ac.uk/chembl) along with their experimentally reported half-maximal inhibitory concentration (IC₅₀) values. All IC₅₀ values were converted into molar units (M) and subsequently transformed into their negative logarithmic scale, pIC₅₀ = −log₁₀(IC₅₀[M]), to normalize the activity data. The initial energy minimization was conducted using MMFF94 force field. The molecular structures of all compounds were downloaded in .sdf format for descriptor calculation and QSAR modeling. The 2-dimensional (2D) structures of all PIM-1 kinase inhibitors are given in supplementary file (Figure 1S).

Calculation of molecular descriptors

Molecular descriptors were calculated using CheMaster v1.2 software. The program automatically generated descriptors from the molecular structures. From the initially computed set, 7 statistically significant descriptors were selected using the software’s built-in feature selection algorithm, which considers inter-descriptor correlation and relevance to biological activity. The selected descriptors were subsequently used for model construction.

Dataset splitting

To ensure balanced representation and diversity, the dataset was divided into training and test sets using the k-clustering method implemented in CheMaster v1.2. The dataset was partitioned in a 75:25 ratio, with 75% of compounds forming the training set for model development and 25% reserved as the external test set for validation. This method ensures that both sets uniformly represent the chemical space of the studied compounds.

Model development

A multiple linear regression (MLR) model was developed using the training set, where pIC₅₀ served as the dependent variable and the 7 selected molecular descriptors as independent variables. The regression equation was obtained using the ordinary least squares method. Descriptor multicollinearity was examined using the variance inflation factor (VIF), and descriptors with VIF > 5 were excluded to ensure statistical stability. Model adequacy was verified by analyzing residual plots and verifying assumptions of linearity, independence, and normal distribution of residuals.

Internal validation

The internal predictivity and robustness of the QSAR model were evaluated using statistical parameters: the correlation coefficient (R), coefficient of determination (R²), and the cross-validated coefficient (Q²) obtained through leave-one-out cross-validation. A model was considered statistically acceptable when R² > .6 and Q² > 0.5, in accordance with established QSAR guidelines.

External validation

The external predictive ability of the model was evaluated using the independent test set. Parameters including R, R²_pred, Q²_pred, and the slopes k and k′ (from regressions through the origin between observed and predicted pIC₅₀ values) were computed. A reliable model was considered to meet the Golbraikh-Tropsha criteria:

R² > 0.6

Q²_pred > 0.5

0.85 ⩽ k ⩽ 1.15 and 0.85 ⩽ k′ ⩽ 1.15

|R₀²—R₀′²| < 0.3

Y-randomization

To confirm that the model was not a result of chance correlation, Y-randomization (response permutation) tests were conducted. The dependent variable (IC₅₀) was randomly shuffled while maintaining the descriptor matrix constant, and new MLR models were built for each permutation (n = 30). The significantly lower R² and Q² values of the randomized models compared with the original model confirmed the robustness and nonrandom nature of the developed QSAR model.

Applicability domain

The applicability domain (AD) of the QSAR model was evaluated using a Williams plot, which relates the standardized residuals to the leverage (h_i) values of each compound to identify structural and response outliers. The leverage threshold (h*) was calculated as equation (1).

h * = \frac{3 (p + 1)}{n}

where p represents the number of descriptors (7) and n the total number of compounds (28). Compounds with leverage values greater than h* are considered structurally influential, whereas those with standardized residuals exceeding ±3 are regarded as response outliers. The graph was generated using MatLab R2024b.

Cell membrane permeability and translocation pathways analysis

The PerMM (Permeability of Molecules across Membranes) web server was used to evaluate the membrane permeability characteristics of Lepiotaprocerin C. Parameters, including permeability coefficient (logP), energy barriers for translocation, and passive diffusion pathways across a biomimetic lipid bilayer model (BLM) were calculated. Experimental conditions were maintained at default settings, with a physiological temperature of 310 K and pH 7.4. Drug optimization mode and BLM-based permeability were selected to ensure a realistic simulation environment for translocation dynamics.³⁹

Physicochemical, pharmacokinetic properties evaluations

To assess the drug-likeness and pharmacokinetic potential of Lepiotaprocerin C, a comprehensive in silico ADME evaluation was done. The compound’s SMILES notation was input into the SwissADME server to compute key physicochemical parameters (such as molecular weight, lipophilicity, and number of hydrogen bond donors/acceptors), drug-likeness, and bioavailability scores.⁴⁰

Prediction of Activity Spectra for Substances analysis

To predict the potential biological activities of Lepiotaprocerin compounds, the Prediction of Activity Spectra for Substances (PASS) server was employed. The input SMILES of each phytoconstituent was analyzed to generate probability scores for activity (Pa) and inactivity (Pi) across a broad range of pharmacological classes. Compounds with Pa > 0.7 were considered to have high confidence in exhibiting the predicted biological effects.⁴¹

Toxicity study

Toxicity predictions were conducted using 2 independent platforms. ProTox-II and the ADMETlab 2.0 server. These tools provided estimated values of toxicity classes, hepatotoxicity, carcinogenicity, and other relevant safety profiles.^42,43

Results and Discussion

Protein dynamics and validation

Figure 2 presents a comprehensive analysis of the structural dynamics and validation of the PIM-1 kinase protein (PDB ID: 4JX3). The deformability plot (Figure 2A) revealed multiple peaks across the atom indices, indicating flexible regions within the protein structure, which are essential for ligand accommodation and active-site mobility. In the mobility comparison (Figure 2B), the NMA data were overlaid with the PDB data, and both profiles exhibited similar fluctuation patterns, suggesting that the modeled dynamics were consistent with the experimental structure. The covariance map (Figure 2C) demonstrated correlated (red), anti-correlated (blue), and uncorrelated (white) motions among the residues, with prominent red diagonal regions indicating strong intra-domain interactions that contribute to structural stability. The eigenvalue obtained from the NMA (2.53 × 10^–4) in Figure 2D reflected the energy required for conformational deformation; a relatively low value signified greater flexibility and favorable binding adaptability. Finally, the Ramachandran plot (Figure 2E) validated the stereochemical quality of the protein model, with most residues clustered in the most favored regions (deep red), and very few outliers observed, confirming that the backbone dihedral angles (φ and ψ) were within permissible limits. Collectively, these findings confirmed that the protein structure was dynamically stable, conformationally flexible, and stereochemically valid for subsequent molecular docking and simulation studies.

Figure 2.

Illustration of (A) deformability, (B) atom index, (C) residue index, (D) eigenvalue, and (E) Ramachandran plot of 4JX3 protein.

Physiochemical properties of PIM-1 protein

The physicochemical properties of the PIM-1 kinase (PDB ID: 4JX3) were evaluated using the ExPASy ProtParam tool, and the results are summarized in Table 2. The protein was composed of 313 amino acid residues, with a calculated molecular weight of 35 685.78 Da and a theoretical isoelectric point (pI) of 5.72, suggesting that the protein is slightly acidic in nature. The sequence contained 44 negatively charged residues (aspartic acid and glutamic acid) and 34 positively charged residues (arginine and lysine), resulting in a net negative charge under physiological conditions. The aliphatic index was calculated to be 95.88, indicating a high proportion of aliphatic side chains and suggesting that the protein may be thermostable. Furthermore, the GRAVY was −0.247, implying that the protein is overall hydrophilic and likely to be soluble in aqueous environments. These physicochemical parameters supported the protein’s structural stability and its suitability for molecular docking and simulation studies.

Table 2.

Physicochemical parameters of PIM-1 protein.

SN	Parameters	Value
1.	Number of amino acids	313
2.	Molecular weight	35 685.78
3.	Theoretical pI	5.72
4.	Total number of negatively charged residues (Asp + Glu)	44
5.	Total number of positively charged residues (Arg + Lys)	34
6.	Formula	C₁₆₀₇H₂₄₉₅N₄₄₃₀O₄₅₉S₁₀
7.	Total number of atoms	5014
8.	Aliphatic index	95.88
9.	Grand average of hydropathicity (GRAVY)	–0.247

Binding pocket analysis

The binding pocket analysis of the PIM-1 kinase protein (PDB ID: 4JX3) was conducted using the CASTp server, which identified a prominent cavity with high potential for ligand binding. A total of 28 amino acid residues were mapped within this primary binding pocket, including Leu44, Gly45, Ser46, Gly47, Gly48, Phe49, Val52, Ala65, Lys67, Val69, Arg73, Ile74, Ser75, Asp76, Pro87, Met88, Glu89, Val90, Leu92, Leu93, Ile104, Leu118, Leu120, Glu121, Arg122, Pro123, Val126, Gln127, Asp128, and Asp131. These residues are predominantly located near the ATP-binding domain, which is functionally critical for kinase activity. In addition, a set of secondary residues—Glu171, Asn172, Leu174, Ile185, Asp186, and Phe187—were identified as part of the extended binding cavity, as visualized in Figure 3.

Figure 3.

Binding site amino acids are covered with grid box for accurate docking. Grid box of binding site predicted from CASTp server (A), blind docking (B), and Prank web server presenting in brown color (C).

Furthermore, 3 methods show strong agreement on a common binding site, validating the CastP pocket. Nine residues are identified by PrankWeb, CASTp and blind docking (after docking): Leu44, Gly45, Ser46, Lys67, Leu120, Glu121, Arg122, Pro123, Val126—these form the primary consensus interaction motif. Several additional residues are shared by 2 methods (supporting the cavity): Gly47, Gly48, Val69, Arg73, Ile74, Ser75, Asp76, Gln127, Phe49, Val52, Ala65, Ile104, Asp128, Leu93, Glu171, Leu174. Differences reflect each method’s sensitivity: PrankWeb highlights surface-exposed and polar/charged contributors (eg, Glu89, Ile185, Asp186), CASTp maps the full geometric cavity (adds Pro87, Met88, Leu118, Val90, Leu92, Asp131), and blind docking pinpoints ligand-contact residues (eg, Val69, Arg73, Ile74, Ser75, Gln127, Glu171, Leu174). Blind docking residues were cross-validated against the post-docking cavity evaluation, confirming strong agreement and validating the identified binding pocket.

CASTp

Binding site amino acids: Leu12, Phe17, Val20, Lys35, Leu44, Gly45, Ser46, Gly47, Gly48, Phe49, Val52, Ala65, Lys67, Val69, Arg73, Ile74, Ser75, Asp76, Pro87, Met88, Glu89, Val90, Leu92, Leu93, Ile104, Leu118, Leu120, Glu121, Arg122, Pro123, Val126, Gln127, Asp128, and Asp131.

PrankWeb

Binding site amino acids (brown color): Ile104, Leu120, Glu121, Arg122, Pro123, Val126, Asp128, Glu171, Leu174, Ile185, Asp186, Leu44, Gly45, Ser46, Phe49, Val52, Ala65, Lys67, Glu89, Leu93.

Blind docking

Binding site amino acids after docking: Leu44, Gly45, Ser46, Gly47, Gly48, Lys67, Val69, Arg73, Ile74, Ser75, Asp76, Leu120, Glu121, Arg122, Pro123, Val126, Gln127, Glu171, Leu174.

Molecular docking analysis

The docking study of Lepiotaprocerin derivatives against PIM-1 kinase revealed distinct variations in binding affinities across different docking approaches (CASTp, PrankWeb, and blind docking), allowing a comprehensive assessment of ligand-protein interactions. Overall, most Lepiotaprocerin derivatives demonstrated comparable or superior binding affinities to the reference inhibitor AZD1208, which showed docking scores of −9.7, −9.6, and −9.6 kcal/mol across the 3 validation methods (Table 3).

Table 3.

Comparative docking scores of Lepiotaprocerin derivatives and reference inhibitor AZD1208 against PIM-1 kinase obtained from CASTp, PrankWeb, and blind docking analysis.

Compounds	Docking score (kcal/mol)
Compounds	CASTp	PrankWeb	Blind docking
Lepiotaprocerin A	−9.8	−10.0	−8.2
Lepiotaprocerin B	−9.7	−9.7	−7.9
Lepiotaprocerin C	−11.4	−11.4	−10.7
Lepiotaprocerin D	−10.6	−11.0	−9.5
Lepiotaprocerin E	−9.9	−9.8	−9.6
Lepiotaprocerin F	−10.5	−10.5	−9.5
Lepiotaprocerin G	−10.5	−10.5	−10.5
Lepiotaprocerin H	−9.2	−9.2	−9.4
Lepiotaprocerin I	−9.3	−9.5	−9.9
Lepiotaprocerin J	−9.1	−9.2	−9.2
Lepiotaprocerin K	−9.6	−9.8	−9.7
Lepiotaprocerin L	−9.9	−9.9	−9.9
AZD1208 (reference)	−9.7	−9.6	−9.6

Among all derivatives, Lepiotaprocerin C exhibited the highest binding affinity across all docking protocols (−11.4 kcal/mol in CASTp and PrankWeb; −10.7 kcal/mol in blind docking), suggesting stronger and more stable interactions with the PIM-1 kinase active site. Its high negative binding energy indicates efficient accommodation within the binding pocket, likely due to optimal hydrophobic contacts and hydrogen-bonding interactions with key residues. Other derivatives, including Lepiotaprocerin D (−10.6 to −11.0 kcal/mol), Lepiotaprocerin F (−10.5 kcal/mol), and Lepiotaprocerin G (−10.5 kcal/mol), also displayed high docking scores, closely approaching the potency of Lepiotaprocerin C. These values suggest that several derivatives exhibit favorable conformational adaptability and strong hydrophobic or π-π stacking interactions within the catalytic pocket. In contrast, compounds such as Lepiotaprocerin H, I, J, and K demonstrated slightly lower docking scores (−9.1 to −9.9 kcal/mol), although still within an active range, indicating stable but comparatively weaker binding energies.

Structural insights into ligand-PIM-1 interactions

The docking visualization from the CASTp cavity binding site reveals that Lepiotaprocerin C (Figure 4A) fits deeply within the PIM-1 kinase pocket, forming a stable hydrogen bond with Glu121 (3.1 Å), while hydrophobic residues contribute to stabilizing the ligand through van der Waals interactions. The reference inhibitor AZD1208 (Figure 4B), in contrast, interacts through multiple hydrogen bonds with Asp128 and Asp131, supported by π-π and hydrophobic contacts within the same cavity.

Figure 4.

The interaction is from the CASTp cavity binding site. Molecular docking interaction of Lepiotaprocerin C with PIM-I kinase (A). Molecular docking interaction of AZD1208 with PIM-I kinase (B). Interaction is showing in 3D surface and cartoon model with conventional hydrogen bonds.

The blind docking analysis confirms that Lepiotaprocerin C (Figure 5A) occupies the same predicted pocket, forming stable hydrogen bonds with Asp131 (3.4 Å) and Lys67 (2.6 Å), consistent with the residues identified in the CastP and PrankWeb cavities. AZD1208 (Figure 5B) also interacts with Asp128, Asp131, and Leu44, validating this site as the true active pocket. The docking at the PrankWeb binding site shows that Lepiotaprocerin C (Figure 6A) establishes hydrogen bonds with Lys67, Glu89, and Asp186, reinforcing its strong electrostatic and hydrogen-bonding interactions with catalytically relevant residues. AZD1208 (Figure 6B) forms key hydrogen bonds with Asn172, Glu171, and Asp128, showing similar anchoring within the active cleft. The hydrogen-bonding distances (2.0-3.4 Å) suggest energetically stable interactions for both ligands. The 2D molecular interaction of all derivatives is presented in supplementary file (Tables 1S to 3S).

Figure 5.

The interaction is from the blinding docking. Molecular docking interaction of Lepiotaprocerin C with PIM-I kinase (A). Molecular docking interaction of AZD1208 with PIM-I kinase (B). Interaction is showing in 3D surface and cartoon model with conventional hydrogen bonds.

Figure 6.

The interaction is from the Prank web binding site. Molecular docking interaction of Lepiotaprocerin C with PIM-I kinase (A). Molecular docking interaction of AZD1208 with PIM-I kinase (B). Interaction is showing in 3D surface and cartoon model with conventional hydrogen bonds.

Across all 3 docking approaches—CastP, blind docking, and PrankWeb—Lepiotaprocerin C consistently demonstrated the strongest binding affinity and most stable orientation within the validated catalytic pocket of PIM-1 kinase. It formed multiple hydrogen bonds with key active-site residues, including Glu121, Lys67, Asp128, and Val126, along with hydrophobic contacts that reinforced complex stability. Compared with the reference inhibitor AZD1208, Lepiotaprocerin C showed deeper binding, stronger electrostatic interactions, and a more favorable docking score, indicating higher affinity and selectivity for the PIM-1 kinase binding cavity. Therefore, the CASTp-docked Lepiotaprocerin C complex—exhibiting the most stable binding conformation and strongest intermolecular interactions—was selected for MD simulation to further validate the structural stability and dynamic behavior of the complex under physiological conditions.

MD simulation and binding analysis of protein-ligand complexes

Figures 7 and 8 illustrate the structural configuration of the most populated cluster obtained from the 200 ns MD simulation of the Lepiotaprocerin C and AZD1209-PIM-1 kinase (4JX3) complex. The dominant conformational state of the complex was extracted using clustering analysis, representing the most stable binding mode over the simulation trajectory. In order to understand how ligand binding influences the structural stability and dynamic properties of protein-ligand complexes, MD simulations were conducted. Parameters such as RMSD, RMSF, Rg, SASA, hydrogen bonding, and binding free energy were analyzed to gain insight into conformational behaviors and interaction profiles in both ligand-bound and unbound (apo) form.

Figure 7.

(A) The root means square deviations and (B) root means square fluctuations of ligands bound complexes. (C) Radius of gyrations and (D) SASA of complexes.

Figure 8.

The representation structure of the most populated cluster of protein (4JX3) with Lepiptaprocerin C during 200 ns molecular dynamics simulation binding site of Lepiptaprocerin C-4JX3 complex and (A) interaction of with Lepiptaprocerin C ligand in protein (B). The ligand with stick and protein complex shown as cartoon.

RMSD and RMSF analysis

To investigate structural changes throughout the simulation over time, RMSD analysis was performed on the Cα atoms. Comparisons between the protein (4JX3) with Azd1208 ligand and protein (4JX3) with Lepiptaprocerin C revealed that both complexes were stable during 200 ns MD simulations (Figure 7A). To further support the RMSD observations, RMSF analysis was done to evaluate residue-specific flexibility (Figure 7B). The Protein (4JX3) with Lepiptaprocerin C complex showed lower fluctuations in compared with protein (4JX3) with Azd1208 ligand complex. The RMSF plot compares the residue-wise flexibility of protein 4JX3 when bound to 2 different ligands—Azd1208 and Lepiptaprocerin C. Both complexes show similar overall fluctuation patterns, indicating comparable structural dynamics across most residues. However, the Lepiptaprocerin C complex (red) shows slightly higher fluctuations at several regions, particularly near the C-terminal (around residue 250), suggesting increased flexibility in these segments upon ligand binding. The RMSD remain within the range of below 2Å throughout the run. This enhanced flexibility may indicate conformational rearrangements induced by Lepiptaprocerin C binding, potentially affecting protein stability and interaction dynamics compared with Azd1208.

Radius of gyration and SASA analysis

Analysis of Rg and SASA (Figure 7C and D) offered additional insights into the structural compactness and solvent interactions. Higher SASA values in the complexes pointed to increased surface exposure and looser protein folding. The both ligand-bound complexes had lower SASA values, implying improved compactness and structural integrity.

Cluster analysis and representative conformations

To explore dominant binding conformations, trajectory clustering was performed, resulting in 5 unique clusters. The most populated cluster, comprising about 70% of the total simulation time was used to extract representative structures for interaction analysis. These structures (Figure 8A and B)/(Figure 9A and B) served as a basis for detailed interaction mapping, highlighting the residues crucial to stable ligand binding.

Figure 9.

The representation structure of the most populated cluster of protein (4JX3) with ligand Azd1208 during 200 ns molecular dynamics simulation. Binding site of Azd1208 in protein (A) interaction of Azd1208 with protein (B).

Hydrogen bonding analysis

Figure 10 provides comparative insights into the dynamic stability and interaction strength of Lepiotaprocerin C versus the reference inhibitor AZD1208 in complex with PIM-1 kinase (4JX3) over the course of the 200 ns MD simulation. In Figure 10A, the hydrogen bond distribution revealed that Lepiotaprocerin C (red line) formed a higher and more sustained number of interactions with the target protein compared with AZD1208 (black line). Most notably, Lepiotaprocerin C consistently formed 5 to 7 hydrogen bonds, with a few frames reaching up to 10 to 12 bonds (Figure 10A). This indicates stronger and more persistent ligand-protein interactions, contributing to complex stability.

Figure 10.

(A) The number of hydrogen bonds between ligands and target protein (B) The RMSD of ligand in 2 different complexes.

Ligand RMSD and binding stability

The RMSD profiles of ligand Azd1208 and Lepiptaprocerin C were analyzed to assess their positional dynamics within the active site (Figure 10B). The Lepiptaprocerin C remained firmly bound with minor fluctuations, confirming strong, stable engagement with the active-site residues. On the other hand, Azd1208 was highly stable in the active site of the protein. The ligand Lepiptaprocerin C showed a small deviation from the initial structure; however, this ligand was also never left its active site during simulations.

Binding free energy estimation

To further assess the binding stability of the ligand-protein complexes, the MM/PBSA approach implemented in the AMBER18 software suite was employed to estimate the binding free energy by combining various energetic contributions, including van der Waals interactions, electrostatic forces, polar solvation, and nonpolar solvation energies. The results revealed that the 4JX3-Lepiptaprocerin C complex exhibited a stronger binding affinity (−22.0 ± 2.1 kcal/mol) compared with the 4JX3-Azd1208 complex (−19.82 ± 1.5 kcal/mol), suggesting that Lepiptaprocerin C forms a more stable complex with the protein (Table 4). The electrostatic energy (ΔE_EEL) was significantly more favorable for Lepiptaprocerin C (−40.68 kcal/mol) than for Azd1208 (−21.85 kcal/mol), indicating stronger electrostatic interactions in the former complex. Conversely, the van der Waals energy (ΔE_VDW was more favorable for Azd1208 (−34.67 kcal/mol) compared with Lepiptaprocerin C (−16.77 kcal/mol), reflecting better hydrophobic interactions in the Azd1208 complex. The polar solvation energy (ΔE_PB) contributed unfavorably to binding for both ligands (≈40 kcal/mol), whereas the nonpolar solvation (ΔE_PBnp) and dispersion (ΔE_Disper) terms provided minor stabilizing effects. Overall, the more negative total binding free energy (ΔG) for Lepiptaprocerin C indicates that its enhanced electrostatic and solvation contributions overcome the relatively weaker van der Waals interactions, resulting in a more stable and energetically favorable complex with the protein 4JX3.

Table 4.

Binding free energy components of protein (4JX3) with Azd1208 ligand and Lepiptaprocerin C.

Parameters, kcal/mol	Value of Azd1208	Value of Lepiptaprocerin C
∆EELEC	−21.85	−40.68
∆EVDW	−34.67	−16.77
∆EPB	40.42	40.57
∆EPBnp	−3.7	−5.06
∆EDisper	0.0	0.0
∆G	−19.82 ± 1.5	−22.0 ± 2.1

Binding free energy (∆G) of protein (4JX3) and ligand (Azd1208) complex was calculated from the 200 ns simulation. The molecular-mechanical energy calculations were performed using MM/PBSA, and entropy calculations using nmode analysis. ∆EELEC, ∆EVDW, ∆EPBnp, and ∆EPBsolv are referred to the electrostatic, van der Waals, polar, the nonpolar contribution to the solvation energy, and the electrostatic contribution to the solvation energy, respectively.

QSAR Model Development and Validation

Molecular descriptor selection

The dataset of 28 PIM-1 kinase inhibitors exhibited pIC₅₀ values between 7.22 and 9.69, indicating moderate to strong inhibitory potency. Compounds with higher LogP (3.3-4.8) and molar refractivity (>160) generally showed greater activity, suggesting that lipophilicity and molecular polarizability enhance kinase inhibition. The molecular weights ranged from 401 to 644 Da, with most compounds containing 2 H-bond donors and 8 to 10 acceptors, supporting favorable interactions with the enzyme’s active site. Moderate flexibility (3-7 rotatable bonds) was common, indicating structural adaptability during binding. The dataset was evenly divided into training (75%) and test (25%) sets with comparable descriptor ranges, ensuring representative sampling. Overall, the results indicate that hydrophobicity, molecular size, and polarizability play key roles in governing PIM-1 kinase inhibitory activity. Using the k-means clustering method, the compounds were divided into a training set of 21 molecules and a test set of 7 molecules, ensuring uniform chemical space distribution (supplementary file, Table 4S).

MLR

The MLR model established for the PIM-2 kinase inhibitors was expressed as follows:

\begin{array}{l} {pIC}_{50} = 8.308 + 0.57 * LogP - 3.270 E - 02 * MR - 0.387 * RB \\ + 1.42 E - 02 * MW + 0.762 * HBD - 0.179 * HBA - 0.30 * RC \end{array}

The predicted versus experimental pIC₅₀ plot shows a strong linear correlation, indicating high predictive accuracy of the MLR model. Most compounds cluster near the ideal line, confirming minimal deviation between observed and predicted values. The close alignment demonstrates the model’s robustness, reliability, and strong external predictivity for PIM-1 kinase inhibitors (Figure 11). Furthermore, the comparison between experimental and predicted pIC₅₀ values demonstrates a strong correlation, confirming the reliability of the developed MLR model. Most compounds exhibit minimal deviation (ΔpIC₅₀ ⩽ 0.3) between observed and predicted values, indicating good internal consistency. A few compounds, such as CHEMBL1092653 and CHEMBL4518369, showed slightly higher deviations but remained within acceptable limits (Tables 5 and 6). The close alignment between experimental and predicted activities across the dataset confirms the model’s accuracy, robustness, and ability to capture key structural features influencing PIM-1 kinase inhibition.

Figure 11.

Predicted versus experimental pIC₅₀ plot for the MLR model of PIM-1 kinase inhibitors.

Table 5.

Experimental and predicted pIC₅₀ values of the training set compounds used for MLR model development.

Observation	Obs (pIC₅₀)	Pred (pIC₅₀)
CHEMBL1092045	9.000	9.264
CHEMBL1089219	7.220	7.325
CHEMBL1088972	7.450	7.886
CHEMBL1088971	7.850	7.886
CHEMBL1092034	8.300	8.249
CHEMBL1093182	8.690	8.774
CHEMBL1092642	8.600	8.600
CHEMBL1092041	8.630	8.754
CHEMBL1092653	9.300	8.751
CHEMBL1092043	9.090	9.712
CHEMBL1093181	9.690	9.009
CHEMBL1092908	8.550	8.484
CHEMBL1092036	8.720	8.843
CHEMBL4518369	7.670	8.082
CHEMBL1091665	8.440	8.609
CHEMBL1092909	8.580	8.355
CHEMBL1088974	8.690	8.430
CHEMBL1092046	9.150	9.420
CHEMBL1092035	8.950	8.687
CHEMBL1092039	9.300	8.915
CHEMBL1092040	9.300	9.136

The close agreement between observed and predicted activities indicates the high accuracy and internal consistency of the developed QSAR model for PIM-1 kinase inhibitors.

Table 6.

Experimental and predicted pIC₅₀ values of the test set compounds used for MLR model development.

Compounds	Obs (pIC₅₀₎	Cal (pIC₅₀)
CHEMBL1092914	9.22	9.47586
CHEMBL1088919	7.35	7.42933
CHEMBL1091666	8.01	8.97115
CHEMBL1092038	8.77	8.71395
CHEMBL1092044	8.77	9.30699
CHEMBL1088970	8.77	8.38747
CHEMBL1092042	9.15	9.15221

Internal and external validation

The internal and external validation results of the developed MLR model demonstrate its robustness and predictive reliability. The internal validation parameters (R = .86, R² = .74, Q² = 0.90) satisfy the recommended QSAR acceptance thresholds (R² > .6 and Q² > 0.5), confirming strong model fitting and internal consistency. The external validation results (R = .79, R² = .63, Q² = 0.51) also meet the Golbraikh-Tropsha criteria, indicating good predictive performance on the test set. The regression slopes (k = 1.022, k₁ = 0.954) are close to unity, reflecting minimal systematic deviation between experimental and predicted pIC₅₀ values. The low mean squared errors (MSE = 0.1 for internal and 0.42 for external validation) further support the model’s accuracy and reliability (Table 7).

Table 7.

Internal and external validation parameters for the MLR-based QSAR model of PIM-1 kinase inhibitors.

Parameters	Internal validation	External validation
R	0.86	0.79
R2	0.74	0.63
Rcv2 (Q2)	0.90	0.51
MSE	0.1	0.42
K	—	1.022
K1	—	0.954

All statistical parameters meet the accepted criteria, confirming the model’s robustness, internal stability, and external predictive ability.

Y-randomization

The Y-randomization (response permutation) test was performed to verify that the developed MLR model was not a result of chance correlation. The randomized models yielded significantly lower statistical values (average r = .618, R² = .394, Q² = −1.253) compared with the original model (r = .8839, R² = .7813, Q² = 0.3469), confirming the true structure-activity relationship captured by the model (Table 8 and 9). Moreover, the corrected correlation parameter (cR_p² = 0.558) exceeded the recommended threshold (>0.5), further validating the robustness and nonrandom nature of the model.

Table 8.

Y-randomization results for the MLR model.

Model	R	R ²	Q ²
Original	.8839	.7813	0.3469
Random 1	.7296	.5323	−0.7123
Random 2	.547	.2992	−0.7981
Random 3	.7168	.5138	−0.0166
Random 4	.5855	.3428	−1.2481
Random 5	.5989	.3587	−1.3936
Random 6	.7736	.5984	−0.2881
Random 7	.5525	.3053	−1.1494
Random 8	.4292	.1842	−2.6023
Random 9	.5831	.34	−1.5246
Random 10	.6339	.4018	−0.6742
Random 11	.367	.1347	−1.3045
Random 12	.7168	.5138	−0.775
Random 13	.4427	.196	−2.4142
Random 14	.6376	.4065	−1.2146
Random 15	.7188	.5166	−1.5059
Random 16	.6512	.424	−1.5601
Random 17	.4886	.2388	−2.4755
Random 18	.4523	.2045	−1.2553
Random 19	.5884	.3462	−2.4073
Random 20	.6722	.4519	−1.0538
Random 21	.8386	.7033	0.2635
Random 22	.7513	.5645	−0.5976
Random 23	.6402	.4099	−0.756
Random 24	.6839	.4677	−1.7816
Random 25	.629	.3957	−0.8698
Random 26	.664	.4409	−1.3712
Random 27	.5703	.3253	−1.5085
Random 28	.5352	.2864	−1.6253
Random 29	.7694	.592	−1.3547
Random 30	.5811	.3377	−1.627

The original model shows considerably higher R, R², and positive Q² values than randomized models, confirming that the developed QSAR model is statistically significant, robust, and not due to random chance.

Table 9.

Y-randomization validation of the MLR model.

Parameters	Observed	Original
Average R	.61829	.8839
Average R²	.39443	.7813
Average Q²	−1.2534	0.3469
cRp2	0.55838

The lower R, R², and Q² values in randomized trials compared to the original model, along with cR_p² > 0.5, confirm that the model is statistically significant and not generated by random correlation. The lower r, R², and Q² values in randomized trials compared with the original model, along with cR_p² > 0.5, confirm that the model is statistically significant and not generated by random correlation.

Applicability domain

In the present model, all training (red) and test (blue) compounds were found within the acceptable leverage and residual limits (|standardized residual| < 3, h_i < 0.8571), indicating the absence of both structural and response outliers. The even distribution of test compounds within the training space confirms that the model’s predictions are reliable and fall within its defined chemical AD, demonstrating the robustness and predictive stability of the developed QSAR model (Figure 12).

Figure 12.

William’s plot illustrating the AD of the developed MLR model for PIM-1 kinase inhibitors.

Calculation of Cal (pIC₅₀) of Lepiotaprocerin C from QSAR model

The docked Lepiotaprocerin C conformation obtained from the most stable CASTp docking complex was extracted and used as input for QSAR prediction to estimate its biological activity. Using the developed MLR model, the calculated pIC₅₀ value for Lepiotaprocerin C was determined to be 8.67, whereas the reference inhibitor exhibited a pIC₅₀ of 9.03 (Table 10). This result indicates that Lepiotaprocerin C possesses strong inhibitory potential against PIM-1 kinase, comparable to the reference drug, thereby validating its docking-derived binding affinity and justifying its selection for further MD simulation to assess the stability and binding efficiency of the complex under dynamic conditions.

Table 10.

Calculated pIC₅₀ values of Lepiotaprocerin C and the reference inhibitor derived from the QSAR model.

Compounds	Cal (pIC₅₀)
Lepiotaprocerin C	8.67
Reference drug	9.03

Pharmacokinetics properties evaluations of Lepiotaprocerin C and reference drug

Lepiotaprocerin C shows a higher molecular weight (478.62) and greater lipophilicity (LogP = 4.82) than the reference drug (379.14; LogP = 2.84), indicating improved membrane permeability. Lepiotaprocerin C has 5 hydrogen bond acceptors (fewer than 10) and no hydrogen bond donors (fewer than 5), meeting most requirements. The results for Lepiotaprocerin C and reference drug are shown in Table 11, which follows the Lipinski rule of 5.

Table 11.

Physicochemical properties of Lepiotaprocerin C and reference drug.

SN	Parameters	Lepiotaprocerin C	Reference Drug
1.	Molecular weight	478.62	379.14
2.	Hydrogen bond acceptor	5	5
3.	Hydrogen bond donor	0	3
4.	Consensus logP	4.82	2.84

Absorption and distribution study

Lepiotaprocerin C exhibited moderate water solubility (−5.79 log mol/L) compared with the reference drug (−4.55 log mol/L), indicating slightly lower aqueous solubility. However, it showed significantly higher CaCO-2 permeability (1.203 log Papp) than the reference (−5.67 log Papp), suggesting superior intestinal absorption. Both compounds were predicted to be well absorbed in the gastrointestinal tract and acted as P-glycoprotein substrates and inhibitors (I and II), indicating potential involvement in efflux transport. Lepiotaprocerin C displayed better skin permeability (−3.47 log Kp) than the reference (−5.36 log Kp), reflecting improved transdermal diffusion. The compound also showed slightly higher volume of distribution (VDss = 0.67 log L/kg) compared with the reference drug (0.55 log L/kg), suggesting broader tissue distribution. Its fraction unbound (5.5%) was lower than the reference (28.7%), implying stronger plasma protein binding. Moreover, Lepiotaprocerin C demonstrated marginally higher blood-brain barrier (BBB) permeability (log BB = 0.17) relative to the reference (0.14), indicating the potential for limited yet notable central nervous system (CNS) penetration (Table 12). These parameters suggest that Lepiotaprocerin C possesses favorable absorption and distribution characteristics, with enhanced permeability and tissue exposure despite reduced solubility.

Table 12.

Absorption and distribution parameters of Lepiotaprocerin C and the reference drug.

SN	Parameters name	Lepiotaprocerin C	Reference drug	Unit
1.	Water solubility	−5.79	−4.55	Numeric (log mol/L)
2.	CaCO-2 permeability	1.203	−5.67	Numeric (log Papp in 10⁻⁶ cm/s)
3.	GI absorption	Absorbed	Absorbed	—
4.	Skin permeability	−3.47	−5.36	Numeric (log Kp)
5.	P-glycoprotein substrate	Yes	Yes	Categorical (Yes/No)
6.	P-glycoprotein I inhibitor	Yes	Yes	Categorical (Yes/No)
7.	P-glycoprotein II inhibitor	Yes	Yes	Categorical (Yes/No)
8.	VDss (human)	0.67	0.55	Numeric (log L/kg)
9.	Fraction unbound (human)	5.5	28.7	Numeric (Fu %)
10.	BBB permeability	0.17	0.14	Numeric (log BB)

Lepiotaprocerin C exhibits lower water solubility but markedly higher CaCO-2 permeability (1.203), suggesting enhanced intestinal absorption. Both compounds show high GI absorption and P-glycoprotein interaction potential. Lepiotaprocerin C demonstrates slightly greater BBB permeability (0.17) and tissue distribution (VDss = 0.67), indicating better central and systemic exposure despite its lower unbound fraction.

Membrane permeability and binding energy analysis

Lepiotaprocerin C exhibited a lower membrane binding energy (−1.82 kcal/mol) than the reference drug (−4.67 kcal/mol), implying reduced lipid affinity and easier membrane passage. Its higher BLM permeability (0.63) compared with the reference (−4.04) suggests superior passive diffusion potential (Table 13). The compound showed moderate permeability across the BBB (−2.67) and CaCO-2 membrane (−3.32), both improved relative to the reference, indicating better transport efficiency through biological membranes and supporting its favorable absorption profile.

Table 13.

Comparative absorption permeability and membrane binding parameters of Lepiotaprocerin C and the reference drug.

SN	Parameters	Lepiotaprocerin C	Reference drug
1.	Free energy of binding (DOPC)	−1.82 kcal/mol	−4.67 kcal/mol
2.	Log of perm. coeff.—BLM	0.63	−4.04
3.	Log of perm. coeff.—BBB (Po)	−2.67	4.32
4.	Log of perm. coeff.—CaCO₂ (Po)	−3.32	−4.51

Lower free energy of binding (−1.82 kcal/mol) indicates weaker interaction with the DOPC membrane compared with the reference (−4.67 kcal/mol), suggesting higher mobility across the lipid bilayer.

Lepiotaprocerin C displayed a lower free energy of binding to the dioleoylphosphatidylcholine (DOPC) membrane (−1.82 kcal/mol) than the reference drug (−4.67 kcal/mol), indicating weaker lipid affinity and easier diffusion across membranes. Its BLM permeability (0.63) was markedly higher than the reference (−4.04), demonstrating superior passive transport. The compound also showed moderate BBB (−2.67) and CaCO-2 (−3.32) permeability coefficients, both better than the reference (4.32 and −4.51, respectively), signifying enhanced membrane permeability and overall absorption potential.

Metabolic stability/cytochrome P450 inhibition

Lepiotaprocerin C inhibits CYP1A2 and CYP3A4, suggesting involvement in primary metabolic pathways, while showing no inhibition of CYP2C19, CYP2C9, or CYP2D6, unlike the reference drug (Table 14). This selective inhibition pattern indicates better metabolic stability and a lower likelihood of adverse drug-drug interactions, supporting its suitability as a safer therapeutic candidate.

Table 14.

Prediction of cytochrome P450 inhibition profile and metabolic stability of Lepiotaprocerin C.

SN	Parameters	Lepiotaprocerin C	Reference drug
SN	Parameters	Prediction	Prediction
1.	Cytochrome CYP1A2 inhibitor	Yes	Yes
2.	Cytochrome CYP2C19 inhibitor	No	Yes
3.	Cytochrome CYP2C9 inhibitor	No	Yes
4.	Cytochrome CYP2D6 inhibitor	No	No
5.	Cytochrome CYP3A4 inhibitor	Yes	Yes

Excretion study

The excretion parameters of Lepiotaprocerin C and reference drug were evaluated to understand its elimination potential and renal clearance mechanism. Lepiotaprocerin C demonstrated higher plasma clearance (11.564 mL/min/kg), implying efficient excretory elimination, while maintaining a comparable half-life (1.057 h) to the reference drug (1.006 h) (Table 15). This balance suggests that Lepiotaprocerin C undergoes faster clearance without significantly affecting its duration of pharmacological activity.

Table 15.

Excretion results parameter for Lepiotaprocerin C and reference drug.

SN	Parameters	Lepiotaprocerin C	Reference drug
1.	CLplasma	11.564 mL/min/kg	3.585 mL/min/kg
2.	T1/2	1.057 hr	1.006 hr

PASS prediction of biological activity

Prediction of Activity Spectra for Substances analysis of Lepiotaprocerin C revealed a high probability of antineoplastic activity (Pa = 0.881) and apoptosis antagonism (Pa = 0.865), indicating strong potential for anticancer action. The compound also showed moderate predicted activity against specific cancer types, including breast cancer (Pa = 0.502), lung cancer (Pa = 0.451), colon cancer (Pa = 0.404), colorectal cancer (Pa = 0.403), and ovarian cancer (Pa = 0.334) (Table 16). These findings support potential activity of Lepiotaprocerin C as a multitargeted anticancer agent.

Table 16.

PASS prediction of Lepiotaprocerin C.

Compound	Pa	Pi	Bioactivity
Lepiotaprocerin C	0.881	0.005	Antineoplastic
	0.865	0.005	Apoptosis antagonist
	0.502	0.019	Breast cancer
	0.451	0.017	Lung cancer
	0.404	0.019	Colon cancer
	0.403	0.021	Colorectal cancer
	0.334	0.015	Ovarian cancer

Toxicity study

In silico toxicity predictions indicated that Lepiotaprocerin C is likely to be nontoxic across major biological systems. The compound was predicted to be non-hepatotoxic (probability = 0.83), non-neurotoxic (0.81), non-cardiotoxic (0.56), and non-nephrotoxic (0.56). It also showed no evidence of carcinogenicity (0.52), immunotoxicity (0.78), mutagenicity (0.89), or cytotoxicity (0.70). The prediction for overall clinical toxicity was also inactive (0.57) (Table 17). These results suggest that Lepiotaprocerin C has a favorable toxicity profile, supporting its potential for further preclinical development.

Table 17.

Predicted toxicity profile of Lepiotaprocerin C based on in silico analysis.

SN	Targeted toxicity	Lepiotaprocerin C	Probability
1.	Hepatotoxicity	Inactive	0.83
2.	Neurotoxicity	Inactive	0.81
3.	Nephrotoxicity	Inactive	0.56
4.	Cardiotoxicity	Inactive	0.56
5.	Carcinogenicity	Inactive	0.52
6.	Immunotoxicity	Inactive	0.78
7.	Mutagenicity	Inactive	0.89
8.	Cytotoxicity	Inactive	0.70
9.	Clinical toxicity	Inactive	0.57

Limitations and future prospects

While this study provides strong computational evidence supporting Lepiotaprocerin C as a potential PIM-1 kinase inhibitor, it is not without limitations. All results were derived from in silico approaches, including molecular docking, MD simulations, ADMET predictions, and PASS analysis, which rely on theoretical models and algorithms. These methods, although powerful, may not fully capture the complexity of biological systems such as metabolism, off-target interactions, and intracellular signaling pathways. In addition, the lack of in vitro or in vivo validation restricts the ability to confirm the compound’s real-world efficacy, bioavailability, and toxicity profile.

Future work should focus on experimental validation of Lepiotaprocerin C, including enzyme inhibition assays, cytotoxicity studies in cancer cell lines, and in vivo pharmacokinetic and toxicity assessments. Moreover, the compound’s impact on apoptosis and downstream signaling pathways should be explored through proteomic and transcriptomic profiling. Integration of machine learning models for activity prediction and multitarget docking may further strengthen the therapeutic evaluation. Ultimately, such studies will help establish Lepiotaprocerin C as a viable candidate for preclinical and clinical development in cancer therapy.

Conclusion

This comprehensive computational investigation identifies Lepiotaprocerin C from Macrolepiota procera as a promising natural PIM-1 kinase inhibitor with potent anticancer potential. Docking and cavity validation confirmed its selective binding at the ATP-binding site, while 200 ns MD simulations and MM/PBSA analysis substantiated its dynamic stability and favorable energetics. The developed QSAR model provided reliable predictive accuracy for kinase inhibition, aligning with experimental benchmarks. ADMET and permeability analyses supported its drug-likeness, metabolic stability, and safety, while PASS results indicated strong antineoplastic activity. Compared with AZD1208, Lepiotaprocerin C exhibited deeper active-site penetration and higher electrostatic interaction energy, underscoring its superior affinity and pharmacological promise. Overall, the study establishes a mechanistic and pharmacokinetic rationale for further experimental validation of Lepiotaprocerin C as a potential lead compound in targeted anticancer therapy through PIM-1 kinase inhibition.

Supplemental Material

sj-docx-1-bbi-10.1177_11779322251410083 – Supplemental material for In Silico Identification of Lepiotaprocerin C as a Promising PIM-1 Kinase Inhibitor: An Integrated Docking, Molecular Dynamics, MM/PBSA, QSAR, and ADMET Study

Supplemental material, sj-docx-1-bbi-10.1177_11779322251410083 for In Silico Identification of Lepiotaprocerin C as a Promising PIM-1 Kinase Inhibitor: An Integrated Docking, Molecular Dynamics, MM/PBSA, QSAR, and ADMET Study by Keshava KS, Faten Qais Ibraheem, Shankar Thapa, Somashekhar M Metri, Sadik Shaik, Santosh Prasad Chaudhary Kurmi, Abhishek Chowdhury, Vipin Kumar Mishra and Pramila T in Bioinformatics and Biology Insights

Footnotes

Acknowledgements

We thank the Department of Pharmacology, GM Institute of Pharmaceutical Sciences and Research, Davangere, Karnataka, India, for providing resources and platform.

ORCID iDs

Keshava KS

Faten Qais Ibraheem

Shankar Thapa

Sadik Shaik

Santosh Prasad Chaudhary Kurmi

Ethical Considerations

Ethical approval was not required for this study as it did not involve any human participants or animal subjects.

Author Contributions

Keshava K S: Conceptualization; Methodology; Software; Formal analysis; Writing—review & editing.

Faten Qais Ibraheem: Software; Visualization; Writing—review & editing.

Shankar Thapa: Conceptualization; Validation; Writing—review & editing; Software.

Somashekhar M Metri: Methodology; Formal analysis; Visualization.

Sadik Shaik: Software; Formal analysis; Validation.

Santosh Prasad Chaudhary Kurmi: Methodology; Software; Data curation; Writing—review & editing.

Abhishek Chowdhury: Software; Formal analysis; Visualization.

Vipin Kumar Mishra: Software; Validation; Formal analysis; Visualization.

Pramila T: Supervision; Visualization; Writing—review & editing; Data curation.

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.

Supplemental Material

Supplemental material for this article is available online.

References

Cancer. WHO. Published 2025. Accessed February 20, 2025. https://www.who.int/news-room/fact-sheets/detail/cancer

Bellon

Nicot

Targeting Pim kinases in hematological cancers: molecular and clinical review. Mol Cancer. 2023;22:1-25. doi:10.1186/s12943-023-01721-1

Golestanifar

Garkani-Nejad

In silico design and ADMET evaluation of new inhibitors for PIM1 kinase using QSAR studies, molecular docking, and molecular dynamic simulation. Heliyon. 2024;10:e38309. doi:10.1016/j.heliyon.2024.e38309

Arrouchi

Lakhlili

Ibrahimi

Re-positioning of known drugs for Pim-1 kinase target using molecular docking analysis. Bioinformation. 2019;15:116-120. doi:10.6026/97320630015116

Julson

Quinn

Nazam

Bownes

Stewart

Beierle

EA.

PIM kinase inhibition sensitizes neuroblastoma to doxorubicin. J Pediatr Surg. 2024;59:1334-1341. doi:10.1016/j.jpedsurg.2024.03.014

Julson

Quinn

Butey

, et al. PIM kinase inhibition attenuates the malignant progression of metastatic hepatoblastoma. Int J Mol Sci. 2024;25:427. doi:10.3390/ijms25010427

Choudhury

Bahadi

Ray

, et al. PIM1 kinase and its diverse substrate in solid tumors. Cell Commun Signal. 2024;22:529. doi:10.1186/s12964-024-01898-y

Ren

Liu

Pang

, et al. ZIKV prM hijacks PIM1 kinase for phosphorylation to prevent ubiquitin−mediated degradation and facilitate viral replication. Front Cell Infect Microbiol. 2024;14:1502770. doi:10.3389/fcimb.2024.1502770

Zhou

Wang

, et al. PIM1 kinase promotes EMT-associated osimertinib resistance via regulating GSK3β signaling pathway in EGFR-mutant non-small cell lung cancer. Cell Death Dis. 2024;15:644. doi:10.1038/s41419-024-07039-0

10.

Mali

SB.

Cancer treatment: role of natural products. Time to have a serious rethink. Oral Oncol Rep. 2023;6:100040. doi:10.1016/j.oor.2023.100040

11.

Naeem

Yang

, et al. Natural products as anticancer agents: current status and future perspectives. Molecules. 2022;27:8367. doi:10.3390/molecules27238367

12.

Adamska

Tokarczyk

Possibilities of using macrolepiota procera in the production of Prohealth food and in medicine. Int J Food Sci. 2022;2022:5773275. doi:10.1155/2022/5773275

13.

Kousar

Naeem

Jamaludin

, et al. Exploring the anticancer activities of novel bioactive compounds derived from endophytic fungi: mechanisms of action, current challenges and future perspectives. Am J Cancer Res. 2022;12:2897-2919.

14.

Kosanić

Ranković

Rančić

Stanojković

Evaluation of metal concentration and antioxidant, antimicrobial, and anticancer potentials of two edible mushrooms Lactarius deliciosus and Macrolepiota procera. J Food Drug Anal. 2016;24:477-484. doi:10.1016/j.jfda.2016.01.008

15.

Zara

Rasul

Sultana

Jabeen

Selamoglu

Identification of Macrolepiota procera extract as a novel G6PD inhibitor for the treatment of lung cancer. Saudi J Biol Sci. 2022;29:3372-3379. doi:10.1016/j.sjbs.2022.02.018

16.

Chen

Zhao

, et al. Anti-proliferative and anti-inflammatory lanostane triterpenoids from the Polish edible mushroom Macrolepiota procera. J Agric Food Chem. 2018;66:3146-3154. doi:10.1021/acs.jafc.8b00287

17.

Secme

Kaygusuz

Eroglu

Dodurga

Colak

Atmaca

Potential anticancer activity of the parasol mushroom, Macrolepiota procera (Agaricomycetes), against the A549 human lung cancer cell line. Int J Med Mushrooms. 2018;20:1075-1086. doi:10.1615/IntJMedMushrooms.2018028589

18.

Roszczenko

Szewczyk-Roszczenko

Gornowicz

, et al. The anticancer potential of edible mushrooms: a review of selected species from Roztocze, Poland. Nutrients. 2024;16:1-22. doi:10.3390/nu16172849

19.

Anwar

Lepore

Czerniecki

Koski

Showalter

LE.

PIM kinase inhibitor AZD1208 in conjunction with Th1 cytokines potentiate death of breast cancer cells in vitro while also maximizing suppression of tumor growth in vivo when combined with immunotherapy. Cell Immunol. 2024;397-398:104805. doi:10.1016/j.cellimm.2024.104805

20.

Kirschner

Wang

van der Meer

, et al. PIM kinase inhibitor AZD1208 for treatment of MYC-driven prostate cancer. J Natl Cancer Inst. 2015;107:dju407. doi:10.1093/jnci/dju407

21.

Lee

Min

, et al. Pan-PIM kinase inhibitor AZD1208 suppresses tumor growth and synergistically interacts with AKT inhibition in gastric cancer cells. Cancer Res Treat. 2019;51:451-463. doi:10.4143/crt.2017.341

22.

Pandey

Thapa

Biradar

, et al. LC-MS profiling and cytotoxic activity of Angiopteris helferiana against HepG2 cell line: molecular insight to investigate anticancer agent. PLoS ONE. 2024;19:e0309797. doi:10.1371/journal.pone.0309797

23.

Møller

Windahl

Sim

, et al. Oligosaccharide and substrate binding in the starch debranching enzyme barley limit dextrinase. J Mol Biol. 2015;427:1263-1277. doi:10.1016/j.jmb.2014.12.019

24.

Thapa

Biradar

Nargund

, et al. Synthesis, molecular docking, molecular dynamic simulation studies, and antitubercular activity evaluation of substituted benzimidazole derivatives. Adv Pharmacol Pharm Sci. 2024;2024:9986613. doi:10.1155/2024/9986613

25.

López-Blanco

Aliaga

Quintana-Ortí

Chacón

IMODS: internal coordinates normal mode analysis server. Nucleic Acids Res. 2014;42:W271-W276. doi:10.1093/nar/gku339

26.

Laskowski

MacArthur

Moss

Thornton

JM.

PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Crystallogr. 1993;26:283-291. doi:10.1107/S0021889892009944

27.

Thapa

Nargund

Biradar

MS.

Molecular design and in-silico analysis of trisubstituted benzimidazole derivatives as Ftsz inhibitor. J Chem. 2023;2023:1-9. doi:10.1155/2023/9307613

28.

Tian

Chen

Lei

Zhao

Liang

CASTp 3.0: computed atlas of surface topography of proteins. Nucleic Acids Res. 2018;46:W363-W367. doi:10.1093/nar/gky473

29.

Jendele

Krivak

Skoda

Novotny

Hoksza

PrankWeb: a web server for ligand binding site prediction and visualization. Nucleic Acids Res. 2019;47:W345-W349. doi:10.1093/nar/gkz424

30.

Thapa

Nargund

Biradar

Banerjee

Karati

In-silico investigation and drug likeliness studies of benzimidazole congeners: the new face of innovation. Inform Med Unlocked. 2023;38:101213. doi:10.1016/j.imu.2023.101213

31.

Wang

Wolf

Caldwell

Kollman

Case

DA.

Development and testing of a general amber force field. J Comput Chem. 2004;25:1157-1174. doi:10.1002/jcc.20035

32.

Roe

Cheatham

TE.

PTRAJ and CPPTRAJ: software for processing and analysis of molecular dynamics trajectory data. J Chem Theory Comput. 2013;9:3084-3095. doi:10.1021/ct400341p

33.

Humphrey

Dalke

Schulten

VMD: visual molecular dynamics. J Mol Graph. 1996;14:33-38. doi:10.1016/0263-7855(96)00018-5

34.

Singh

Mishra

Das

AK.

Crystal structure of a thiolase from archaeal pyrococcus furiosus and its in silico functional annotation. Biochem Biophys Res Commun. 2024;693:149377. doi:10.1016/J.BBRC.2023.149377

35.

Mishra

Flipped regiospecificity in L434F mutant of 8-lipoxygenase. Phys Chem Chem Phys. 2020;22:16013-16022. doi:10.1039/D0CP02351E

36.

Rosell-Hidalgo

Moore

Ghafourian

Prediction of drug-induced mitochondrial dysfunction using succinate-cytochrome c reductase activity, QSAR and molecular docking. Toxicology. 2023;485:153412. doi:10.1016/j.tox.2022.153412

37.

Nguyen

Kim

MS.

Identification of promising inhibitory heterocyclic compounds against acetylcholinesterase using QSAR, ADMET, biological activity, and molecular docking. Comput Biol Chem. 2023;104:107872. doi:10.1016/j.compbiolchem.2023.107872

38.

Daoui

Elkhattabi

Chtita

Elkhalabi

Zgou

Benjelloun

AT.

QSAR, molecular docking and ADMET properties in silico studies of novel 4,5,6,7-tetrahydrobenzo[D]-thiazol-2-Yl derivatives derived from dimedone as potent anti-tumor agents through inhibition of C-Met receptor tyrosine kinase. Heliyon. 2021;7:e07463. doi:10.1016/j.heliyon.2021.e07463

39.

Lomize

Hage

Schnitzer

, et al. PerMM: a web tool and database for analysis of passive membrane permeability and translocation pathways of bioactive molecules. J Chem Inf Model. 2019;59:3094-3099. doi:10.1021/acs.jcim.9b00225

40.

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. doi:10.1038/srep42717

41.

Lagunin

Stepanchikova

Filimonov

Poroikov

PASS: prediction of activity spectra for biologically active substances. Bioinformatics. 2000;16:747-748. doi:10.1093/bioinformatics/16.8.747

42.

Drwal

Banerjee

Dunkel

Wettig

Preissner

ProTox: a web server for the in silico prediction of rodent oral toxicity. Nucleic Acids Res. 2014;42:W53-58. doi:10.1093/nar/gku401

43.

Xiong

, et al. ADMETlab 2.0: an integrated online platform for accurate and comprehensive predictions of ADMET properties. Nucleic Acids Res. 2021;49:W5-W14. doi:10.1093/nar/gkab255

Supplementary Material

Please find the following supplemental material available below.

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

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

0.00 MB

1.16 MB

In Silico Identification of Lepiotaprocerin C as a Promising PIM-1 Kinase Inhibitor: An Integrated Docking,Molecular Dynamics,MM/PBSA,QSAR,and ADMET Study

Abstract

Keywords

Introduction

Materials and Methods

Selection of phytochemical

Compound and target protein preparation

Protein dynamics and validation by normal mode analysis

Binding pocket prediction

Binding site validation

Molecular docking methodology

MD simulation

MM/PBSA binding free energy calculation

Visualization and interpretation

QSAR model development and validation

Studied compounds

Calculation of molecular descriptors

Dataset splitting

Model development

Internal validation

External validation

Y-randomization

Applicability domain

Cell membrane permeability and translocation pathways analysis

Physicochemical, pharmacokinetic properties evaluations

Prediction of Activity Spectra for Substances analysis

Toxicity study

Results and Discussion

Protein dynamics and validation

Physiochemical properties of PIM-1 protein

Binding pocket analysis

CASTp

PrankWeb

Blind docking

Molecular docking analysis

Structural insights into ligand-PIM-1 interactions

MD simulation and binding analysis of protein-ligand complexes

RMSD and RMSF analysis

Radius of gyration and SASA analysis

Cluster analysis and representative conformations

Hydrogen bonding analysis

Ligand RMSD and binding stability

Binding free energy estimation

QSAR Model Development and Validation

Molecular descriptor selection

MLR

Internal and external validation

Y-randomization

Applicability domain

Calculation of Cal (pIC50) of Lepiotaprocerin C from QSAR model

Pharmacokinetics properties evaluations of Lepiotaprocerin C and reference drug

Absorption and distribution study

Membrane permeability and binding energy analysis

Metabolic stability/cytochrome P450 inhibition

Excretion study

PASS prediction of biological activity

Toxicity study

Limitations and future prospects

Conclusion

Supplemental Material

sj-docx-1-bbi-10.1177_11779322251410083 – Supplemental material for In Silico Identification of Lepiotaprocerin C as a Promising PIM-1 Kinase Inhibitor: An Integrated Docking, Molecular Dynamics, MM/PBSA, QSAR, and ADMET Study

Footnotes

Acknowledgements

ORCID iDs

Ethical Considerations

Author Contributions

Funding

Declaration of Conflicting Interests

Supplemental Material

References

Supplementary Material

Calculation of Cal (pIC₅₀) of Lepiotaprocerin C from QSAR model