Fungal secondary metabolites as a potential inhibitor of T315I- BCR::ABL1 mutant in chronic myeloid leukemia by molecular docking, molecular dynamics simulation and binding free energy exploration approaches Authors: Dilinazi Abulaiti, Niluopaer Tuerxun, Huan Wang, Lina Ma, Fang Zhao, Yang Liu, Jianping Hao Published: 2024 ABSTRACT BackgroundChronic Myeloid Leukemia (CML) is particularly challenging to treat due to the T315I BCR::ABL1 mutation. Although fungal metabolites are known for their pharmaceutical potential, none are approved for CML. Our study screened approximately 2000 fungal secondary metabolites to discover inhibitors targeting the T315I- BCR::ABL1 mutant protein.MethodsWe conducted comprehensive analyses to elucidate the interactions between the T315I-BCR::ABL1 mutant protein and selected fungal metabolites. These analyses included molecular docking, ADMET assessment, molecular dynamics simulations, principal components analysis, exploration of free energy landscapes, and per-residue decomposition.ResultsWe identified a range of binding affinities for fungal secondary metabolites, from −11.2 kcal/mol to −2.90 kcal/mol, with the co-crystal ponatinib showing a binding affinity of −9.9 kcal/mol. Notably, twenty seven fungal metabolites had affinities ≤ -10.0 kcal/mol, surpassing ponatinib. Eight compounds, including Phellifuropyranone A and Meshimakobnol B, showed favorable drug-likeness. Molecular dynamics parameters, including RMSD, RMSF, Rg, and SASA, confirmed that Phellifuropyranone A and Meshimakobnol B bind stably to the T315I-BCR::ABL1 mutant protein. Additionally, PCA, DCCM, and free energy landscapes analyses validated the consistency of the molecular dynamics parameters. MM/PBSA analysis indicated that Phellifuropyranone A (–22.88 ± 4.28 kcal/mol) and Meshimakobnol B (−25.86 ± 3.51 kcal/mol) bind similarly to ponatinib (−25.54 ± 6.31 kcal/mol). Per-residue decomposition explored residues MET290, VAL299, ILE315, and PHE359 as crucial for binding to the T315I-BCR::ABL1 mutant protein.ConclusionsPhellifuropyranone A and Meshimakobnol B show significant potency as inhibitors of the T315I-BCR::ABL1 mutant protein, comparable to ponatinib. These compounds may serve as effective alternatives or synergistic agents with ponatinib, potentially overcoming drug resistance and improving treatment outcomes in Chronic Myeloid Leukemia. Background Chronic myeloid leukemia (CML) is a clonal expansion of progenitor hematopoietic stem cells, resulting from the fusion BCR::ABL1 oncogene.1 This oncogene accounts for 15 %-20 % of newly diagnosed leukemia cases. The BCR::ABL1 oncogene encodes a BCR::ABL1 tyrosine kinase (TK) that activates various downstream targets, such as c-Myc, STAT5, and CrkL.2, 3 These targets play crucial roles in regulating cell differentiation, proliferation, migration, and other cellular behaviors. For sustained disease control in chronic myeloid leukemia (CML), the use of tyrosine kinase inhibitors targeting BCR::ABL1 proves effective. BCR::ABL1 inhibitor, stands as the exemplar of targeted therapy in CML, offering a safe and effective treatment for most patients diagnosed with chronic phase disease.4 Under the selective pressure of TKI, approximately 50 % of chronic phase (CP) cases develop drug resistance. This resistance arises from recurrent BCR::ABL1 amino-acid exchanges, altering the conformation of the ABL1 tyrosine kinase pocket. Consequently, the activity of the most commonly prescribed TKI is impaired at various levels. Notably, the T315I BCR::ABL1 mutation poses the most significant challenge, as it entirely disrupts contact between the ABL tyrosine kinase domain and these three TKI.5, 6, 7 The T315I mutation represents a unique alteration in the ATP binding region of the BCR::ABL1 protein. This mutation alters the conformation of the binding site in such a way that most currently available tyrosine kinase inhibitors (TKI) are unable to effectively bind and inhibit the activity of the BCR::ABL1 protein. As a result, leukemia cells harboring this mutation are resistant to the majority of TKI therapies.8, 9 In chronic phase chronic myeloid leukemia, mutations in BCR::ABL1 are a well-characterized common mechanism of resistance to first-line drug treatments.10 Analysis of BCR::ABL1 mutations is recommended to aid in selecting the appropriate therapy for chronic myeloid leukemia patients post-imatinib treatment failure, as certain commonly encountered mutations can lead to clinical resistance to nilotinib and/or dasatinib.11 The presence of the T315I mutation in patients with chronic myeloid leukemia has been associated with reduced efficacy of nilotinib, which can complicate efforts to manage the disease effectively. This mutation presents a significant hurdle in achieving favorable treatment responses, necessitating careful consideration and potentially alternative therapeutic strategies12 The frequency of ABL gene mutations among chronic myeloid leukemia patients who develop resistance to imatinib is a critical aspect in understanding treatment outcomes. Investigating the results of switching to second-generation tyrosine kinase inhibitors (TKIs) following resistance to imatinib provides valuable insights into the efficacy of alternative therapies in managing the disease.13 The prominence of the T315I mutation underscores a substantial obstacle in addressing chronic myeloid leukemia (CML), as it severely constrains the available therapeutic avenues for patients carrying this mutation. This limitation presents a formidable challenge in managing the disease effectively and emphasizes the urgent need for innovative treatment strategies tailored to address such resistant mutations. Over the past few decades, fungal metabolites have emerged as key players in the realm of cancer drug discovery, contributing significantly to the development of numerous therapeutic agents.14 The captivating realm of fungus-derived secondary metabolites offers immense biomedical potential and chemical diversity. Exploring endophytic fungi for drug candidates remains an ongoing and promising endeavor within the domains of drug discovery and medicinal chemistry.15 Derived from fungal secondary metabolites, numerous compounds offer essential alternatives to strengthen the arsenal of chemotherapeutic agents against cancer. This diverse array of compounds presents promising avenues for innovation in cancer treatment strategies.16 In recent decades, research into fungal-derived materials to combat cancer has made significant progress in terms of drug delivery vehicles, active anticancer ingredients, and cancer immunotherapy. Various components of fungi have been effectively utilized to accomplish specific tasks. In addition to identifying novel anticancer compounds from fungi, their native capsular structures can also serve as drug carriers to achieve increased oral bioavailability.17 At both preclinical and clinical trial stages, certain well-characterized fungal metabolites exhibit surprisingly potent antitumor properties. While their precise mechanisms of action are still under investigation, substantial evidence indicates that they functionally target core regulatory pathways and enzymes dysregulated during cancer pathogenesis.18 Kuttuvan Valappil Sajna et al revealed that fungal secondary metabolites possess significant therapeutic potential as anticancer agents. The secondary metabolites derived from fungus exhibit an antiproliferative role by inducing intracellular ROS production and mitochondrial membrane potential loss, ultimately leading to apoptosis.19 In silico studies on a fungal metabolite against skin cancer proteins have unveiled its potential activity in combating skin cancer.20 The evaluation of secondary metabolites from an endophytic fungus through in vitro and in silico studies reveals promising anticancer properties.21 Altogether, fungal metabolites collectively represent a promising class of potential anticancer agents for combating cancer. Despite fungi producing a wide array of structurally diverse metabolites, many of which have proven to be valuable sources of pharmaceuticals, no fungal-derived agent has yet been approved as a cancer drug in CML. This is despite extensive research efforts aimed at identifying fungal secondary metabolites with promising anticancer activities. Our objective was to gather approximately two thousand fungal secondary metabolites for screening their inhibitory activity against the T315I-BCR::ABL1 mutant in chronic myeloid leukemia (CML). Subsequently, we conducted comprehensive analyses, including molecular docking, ADMET assessment, molecular dynamics simulations, principal components analysis, exploration of free energy landscapes, and per-residue decomposition, to elucidate the interactions between the T315I-BCR::ABL1 mutant protein and selected fungal metabolites. Methods Collection of fungal secondary metabolites MeFSAT, the first comprehensive manually curated database on Medicinal Fungi Secondary Metabolites And Therapeutics, serves as a valuable resource, compiling information on 184 medicinal fungi, 1830 secondary metabolites, and 149 therapeutic uses. The Medicinal Fungi Secondary Metabolites and Therapeutics (MeFSAT) database (https://cb.imsc.res.in/mefsat/home) provided us with the 3D structures of 1830 fungal secondary metabolites. This extensively curated resource offers comprehensive details on the structures, properties, and diverse applications of these compounds derived from medicinal fungi.22 We downloaded the 3D structure of secondary metabolites from the download module of MeFSAT. In addition, we downloaded the 3-dimensional structure of co-crystal Ponatinib (AP24534) from the collaboratory for structural bioinformatics protein data bank (RCSB PDB) (https://www.rcsb.org/structure/3ik3). Ligand preparation for docking study The ligands underwent energy minimization using the Universal Force Field (UFF) with the optimization algorithm set to Conjugate, comprising a total of 200 steps. This process was conducted within Open Babel, integrated into the PyRx 0.8 tool. Subsequently, all ligands subjected to energy minimization were converted into the PDBQT format. Downloading and preparation of T315I-BCR::ABL1 mutant protein The crystal structure of the T315I-BCR::ABL1 mutant protein (PDB ID: 3IK3) with a resolution of 1.90 Å was downloaded from the Research Collaboratory for Structural Bioinformatics (RCSB) − Protein Data Bank (https://www.rcsb.org).4 Utilizing high-resolution protein 3D structures as targets, we conducted screening and evaluation of inhibitors. The retrieved proteins underwent preparation, involving the removal of co-crystallized ligands, ions, and water molecules. Additionally, polar hydrogen atoms were added to enhance accuracy, utilizing BIOVIA Discovery Studio software (https://discover.3ds.com/discovery-studio-visualizer-download). Finally, this structure served as the receptor molecule for screening with secondary metabolites library. Molecular docking study For our molecular docking study, we employed PyRx software (https://pyrx.sourceforge.io/). Initially, we utilized the PrankWeb tool (https://prankweb.cz/) to identify the binding pocket of the selected protein. PrankWeb utilizes a machine learning-based approach for predicting ligand binding sites from protein structures. Following this, we converted the protein into a protein data bank with charges and atom types (PDBQT) format and selected the amino acid residues associated with the binding site to generate a grid. Subsequently, we conducted the docking study to assess the docking score between the selected receptor and ligands. ADMET analysis of fungal secondary metabolites We utilized the SwissADME web tool ((https://www.swissadme.ch)) and pkCSM, which offers free access to a wide range of predictive models for physicochemical properties, pharmacokinetics, drug-likeness, medicinal chemistry evaluation, toxicity.23, 24 This tool includes both fast and robust predictive methods, including in-house proficient techniques, providing comprehensive insights into drug-related properties and behavior. Molecular dynamics simulations Utilizing GROMACS 2023.2, a 100 ns molecular dynamics simulation was conducted on the protein–ligand complexes.25 The simulations were performed under the influence of the selected CHARMM27 force field, enabling a detailed investigation into their structural stability and dynamic behavior over an extended period. The topology for the protein was generated using the pdb2gmx command, ensuring accurate representation of its molecular structure and interactions. For the ligands, topologies were obtained from the SwissParam server (https://www.swissparam.ch/).26 The unit cells containing the native protein and protein–ligand complexes were defined using the gmx_editconf command, employing a cubic box of 10 Å dimensions and incorporating periodic boundary conditions (PBC) to simulate an infinite system. Subsequently, the solvation of the unit cells was carried out to mimic the physiological environment of the protein. This was achieved using the gmx_solvate module, which utilized the TIP3P water model to add water molecules to the system. The system underwent charge neutralization using the gmx_genion module, which replaced solvent molecules with appropriate counter-ions (Na + for positive and Cl- for negative charges) to achieve electrostatic neutrality. Following this, system minimization was performed using the steepest descent algorithm until the maximum force reached < 100.0 kJ/mol/nm within a maximum of 50,000 steps to alleviate potential stresses. Subsequently, solvent and ions were equilibrated at 300 K in two steps: first in the NVT ensemble maintaining constant particle number, volume, and temperature, followed by the NPT ensemble with constant particle number, pressure, and temperature. Finally, production molecular dynamics (MD) was conducted for 100 ns, with each step lasting 2 fs, using the gmx_mdrun module. Trajectory analysis of simulated complexes Following the production MD run, the obtained trajectory file underwent correction to remove the periodic boundary conditions (PBC) using the gmx_trjconv module in GROMACS. This process yielded a new corrected trajectory file, which was subsequently utilized for further analysis using the built-in tools available in GROMACS.25 Essential parameters, including root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), and solvent-accessible surface area (SASA), were computed using various GROMACS tools. Specifically, the gmx_rms, gmx_rmsf, gmx_gyrate, and gmx_SASA commands were employed for this purpose. The analysis of hydrogen bond formation was conducted using the gmx hbond tool, providing valuable insights into molecular interactions. Graphical representations of the data were visualized using Grace Software, a powerful tool for generating clear and visually appealing plots (https://plasma-gate.weizmann.ac.il/Grace/). Principal components analysis Following this, Principal Component Analysis (PCA) was conducted to analyze the protein dynamics. This involved utilizing the gmx_covar and gmx_anaeig modules to determine the principal components and eigenvalues essential for understanding the underlying motions. To further explore the conformational space, the gmx_anaeig module was utilized again, this time with the −extr command. This allowed for the calculation of extreme projections along the trajectory based on the eigenvalues, which were then exported in PDB format. These extreme PDB files were subsequently used to generate a visual representation of protein motion using the ModeVectors.py script in PyMOL, resulting in a porcupine plot. Exploring free energy landscape and dynamics cross-correlation matrix analysis (DCCM) analysis The xvg file generated during the PCA analysis served as a foundation for assessing the free energy landscape (FEL) using the gmx_sham module. This analysis provided valuable information about the energetics of both native and protein–ligand complexes for exploring their dynamic behavior. In the final step, dynamics cross-correlation Matrix analysis (DCCM) was conducted using the Bio3D R package.27 To conduct the DCCM analysis, GROMACS trajectory files (.xtc) were transformed into Nanoscale Molecular Dynamics (NAMD) format (.dcd).28 This conversion process was executed using VMD software, ensuring compatibility between different molecular dynamics simulation platforms for subsequent analysis.29 This analysis provides insights into the correlated and anti-correlated motions within the protein regions of interest in protein–ligand complexes. Binding free energy (MM-PBSA approach) and per-residue decomposition analysis Employing the Molecular Mechanics/Poisson-Boltzmann Surface Area (MM-PBSA) approach,30 the g_mmpbsa package is integral to GROMACS for assessing the binding free energy (BE) of ligand-bound complexes. This crucial tool calculates BE accurately, shedding light on the energetic landscape of ligand–protein interactions. The representation of the binding energy (BE) for the lead-bound protein complex was calculated using the MD simulation trajectories spanning 100 ns. The BE of the test complexes was determined utilizing these trajectories, employing the following equation:ΔGbinding = Gcomplex − (Gprotein + Gligand) In the given equation, Gcomplex signifies the energy associated with the lead secondary compound or standard inhibitor bound to the test protein complex. Moreover, Gprotein and Gligand represent the energies of the protein and ligand, respectively, within an aqueous environment. Results Fungal secondary metabolites potentially interacted with T315I- BCR::ABL1 mutant protein The study conducted virtual screening on a selection of 1830 fungal secondary metabolites sourced from the MeFSAT database. The aim was to evaluate their potential as inhibitors against the T315I-BCR::ABL1 mutant protein, a known oncogenic enzyme associated with chronic myeloid leukemia (CML). By employing molecular docking approach, we identified the most promising compounds capable of effectively inhibiting the activity of this mutant protein (3IK3) with appreciable binding affinity. Prior to conducting molecular docking, we employed the PrankWeb tool to predict the binding sites of a known inhibitor (Fig. 1A). This step was crucial for facilitating optimal interactions between the compounds and its target at the molecular level. In addition, we identified the interacting residues with co-crystal ligand (AP24534 or ponatinib) (Fig. 1B), and these were combined with PrankWeb identified residues to generate a grid box. Interestingly, within the identified top-ranked pocket, we found that the co-crystal ligand (AP24534 or ponatinib) is embedded (Fig. 1A). Consequently, we selected this pocket for the molecular docking study of the T315I-BCR::ABL1 mutant protein with fungal secondary metabolites. We discovered a range of binding affinities among fungal secondary metabolites, spanning from −11.2 kcal/mol to −2.90 kcal/mol (Supplementary Table 1). Furthermore, upon re-docking the co-crystal ponatinib, we observed a binding affinity of −9.9 kcal/mol. Our analysis unveiled that 27 fungal secondary metabolites exhibited lower binding affinities (≤ −10.0) compared to the reference co-crystal ponatinib (−9.9 kcal/mol), as outlined in Table 1. The binding affinities of all fungakl metabolites tabulated in Supplementary Table 1. Molecular docking elucidates both the location and manner in which a ligand interacts with a protein. This computational technique offers valuable insights into the intricate interplay between molecules, exploring the precise binding sites and the specific configurations adopted during the binding process. By simulating the docking process, we can elucidate the molecular mechanisms underlying biological functions and design more effective drugs or therapeutic interventions targeting specific protein–ligand interactions.Fig. 1Identification of Binding residues of T315I- BCR::ABL1 Mutant Protein. A. Exploring the amino acid residues in the top-ranked binding pocket by PrankWeb tool. B. Exploration of binding residues with Ponatinib in co-crystal structure of T315I-BCR::ABL1 mutant protein.Table 1Binding affinity comparison of co-crystal ponatinib and top-ranked fungal secondary metabolites.LigandName of the compoundFormulaMWBinding Affinityrmsd/ubrmsd/lbMSID001279Phelligridin IC33H20O13624.5−11.200MSID001271Phellifuropyranone AC21H14O7378.33−11.000MSID000080(+)-(R)-grifolinone CC44H60O5668.94−10.900MSID001420Sterhirsutin IC45H54O9738.9−10.900MSID000939Ganoleuconin OC51H74O7799.13−10.800MSID001720Inonotusin BC23H18O8422.38−10.700MSID001178C20H12O8380.3−10.600MSID001179Meshimakobnol BC20H12O7364.31−10.600MSID001081Inoscavin CC23H16O8420.37−10.500MSID001561Caputmedusin AC42H50N2O11758.85−10.500MSID001082Inoscavin DC21H16O8396.35−10.400MSID001200Methylinoscavin CC24H18O8434.39−10.400MSID001201Methylinoscavin DC22H18O8410.37−10.400MSID001535AmauroamoieninC46H54O7718.92−10.400MSID000963C51H72O8813.11−10.300MSID001266Phellibaumin AC19H12O7352.29−10.300MSID001792Lepiotaprocerin GC30H38O5478.62−10.300MSID001277Phelligridin GC32H18O12594.48−10.200MSID001290PhellinstatinC39H26O15734.61−10.200MSID001385Schizine BC26H26N2O5446.5−10.200MSID000321AlbatrellinC44H56O6680.91−10.100MSID001245Palustrisolide AC38H58O8642.86−10.100MSID001278Phelligridin HC33H18O13622.49−10.100MSID001417Sterhirsutin FC30H33ClO6525.03−10.100MSID001568Demalonyl fomitopsin HC32H50O5514.74−10.100MSID000814Ganocin DC20H22O3310.39−10.000MSID001268Phellibaumin CC33H20O13624.5−10.0000LIPonatinibC29H27F3N6O532.6−9.900 Identification of Binding residues of T315I- BCR::ABL1 Mutant Protein. A. Exploring the amino acid residues in the top-ranked binding pocket by PrankWeb tool. B. Exploration of binding residues with Ponatinib in co-crystal structure of T315I-BCR::ABL1 mutant protein. Table 1Binding affinity comparison of co-crystal ponatinib and top-ranked fungal secondary metabolites.LigandName of the compoundFormulaMWBinding Affinityrmsd/ubrmsd/lbMSID001279Phelligridin IC33H20O13624.5−11.200MSID001271Phellifuropyranone AC21H14O7378.33−11.000MSID000080(+)-(R)-grifolinone CC44H60O5668.94−10.900MSID001420Sterhirsutin IC45H54O9738.9−10.900MSID000939Ganoleuconin OC51H74O7799.13−10.800MSID001720Inonotusin BC23H18O8422.38−10.700MSID001178C20H12O8380.3−10.600MSID001179Meshimakobnol BC20H12O7364.31−10.600MSID001081Inoscavin CC23H16O8420.37−10.500MSID001561Caputmedusin AC42H50N2O11758.85−10.500MSID001082Inoscavin DC21H16O8396.35−10.400MSID001200Methylinoscavin CC24H18O8434.39−10.400MSID001201Methylinoscavin DC22H18O8410.37−10.400MSID001535AmauroamoieninC46H54O7718.92−10.400MSID000963C51H72O8813.11−10.300MSID001266Phellibaumin AC19H12O7352.29−10.300MSID001792Lepiotaprocerin GC30H38O5478.62−10.300MSID001277Phelligridin GC32H18O12594.48−10.200MSID001290PhellinstatinC39H26O15734.61−10.200MSID001385Schizine BC26H26N2O5446.5−10.200MSID000321AlbatrellinC44H56O6680.91−10.100MSID001245Palustrisolide AC38H58O8642.86−10.100MSID001278Phelligridin HC33H18O13622.49−10.100MSID001417Sterhirsutin FC30H33ClO6525.03−10.100MSID001568Demalonyl fomitopsin HC32H50O5514.74−10.100MSID000814Ganocin DC20H22O3310.39−10.000MSID001268Phellibaumin CC33H20O13624.5−10.0000LIPonatinibC29H27F3N6O532.6−9.900 Binding affinity comparison of co-crystal ponatinib and top-ranked fungal secondary metabolites. Fungal secondary metabolites are drug-like compounds to inhibit T315I- BCR::ABL1 mutant protein Drug-likeness prediction plays a crucial role in the virtual screening of drug candidates, as it is believed to be closely linked to the comprehensive array of properties required for successful progression through clinical trials. This association underscores the importance of accurately assessing drug-likeness in early stages of drug discovery.31 We conducted an analysis employing Lipinski, Ghose, Veber, Egan, and Muegge rules to assess the drug-likeness of selected compounds identified through molecular docking. Remarkably, our investigation unveiled that eight compounds, namely Phellifuropyranone A (MSID001271), (+)-(R)-grifolinone C (MSID000080), Meshimakobnol B (MSID001179), Inoscavin D (MSID001082), Methylinoscavin C (MSID001200), Phellibaumin A (MSID001266), Schizine B (MSID001385), and Ganocin D (MSID000814), exhibited favorable drug-likeness properties (Table 2). In contrast, the co-crystal Ponatinib did not demonstrate drug-likeness upon evaluation using the Lipinski, Ghose, Veber, Egan, and Muegge rules, as indicated in Table 2. We further present the comprehensive physicochemical properties, including lipophilicity, water solubility, pharmacokinetics, and drug-likeness characteristics, in Supplementary Table 2.Table 2Metabolite ID, name, Drug-likeness, Structure, and bioavailability radar plot of the selected fungal secondary metabolites and standard ponatinib.Secondary Metabolite ID and nameFormulaMWDrug-likeness (Lipinski, Ghose, Veber, Egan, and Muegge rule)StructureRadar plotMSID001271 (Phellifuropyranone A)C21H14O7378.33YesMSID000080 ((+)-(R)-grifolinone C)C44H60O5668.94YesMSID001179(Meshimakobnol B)C20H12O7364.31YesMSID001082 (Inoscavin D)C21H16O8396.35YesMSID001200 (Methylinoscavin C)C24H18O8434.39YesMSID001266 (Phellibaumin A)C19H12O7352.29YesMSID001385 (Schizine B)C26H26N2O5446.5YesMSID000814 (Ganocin D)C20H22O3310.39Yes0LI (Ponatinib)C29H27F3N6O532.56No Table 2Metabolite ID, name, Drug-likeness, Structure, and bioavailability radar plot of the selected fungal secondary metabolites and standard ponatinib.Secondary Metabolite ID and nameFormulaMWDrug-likeness (Lipinski, Ghose, Veber, Egan, and Muegge rule)StructureRadar plotMSID001271 (Phellifuropyranone A)C21H14O7378.33YesMSID000080 ((+)-(R)-grifolinone C)C44H60O5668.94YesMSID001179(Meshimakobnol B)C20H12O7364.31YesMSID001082 (Inoscavin D)C21H16O8396.35YesMSID001200 (Methylinoscavin C)C24H18O8434.39YesMSID001266 (Phellibaumin A)C19H12O7352.29YesMSID001385 (Schizine B)C26H26N2O5446.5YesMSID000814 (Ganocin D)C20H22O3310.39Yes0LI (Ponatinib)C29H27F3N6O532.56No Metabolite ID, name, Drug-likeness, Structure, and bioavailability radar plot of the selected fungal secondary metabolites and standard ponatinib. We utilized the pKCSM server to predict the toxicity of these eight drug-like compounds. The toxicity profile of Phellifuropyranone A (MSID001271), (+)-(R)-grifolinone C (MSID000080), Meshimakobnol B (MSID001179), Inoscavin D (MSID001082), Methylinoscavin C (MSID001200), Phellibaumin A (MSID001266), Schizine B (MSID001385), and Ganocin D (MSID000814) are tabulated in Table 3.Table 3Toxicity of the selected fungal secondary metabolites and standard ponatinib.Name of the toxicity modelMSID00080MSID000814MSID001082MSID001179MSID001200MSID001266MSID001271MSID0013850LI (Ponatinib)AMES toxicityNoNoNoNoNoYesYesNoYesMax. tolerated dose (human) ((log mg/kg/day))0.08−0.4640.5430.6340.3810.4460.51−0.9260.126hERG I inhibitorNoNoNoNoNoNoNoNoYeshERG II inhibitorYesNoYesYesYesYesYesYesYesOral Rat Acute Toxicity (LD50) (mol/kg)3.0092.2742.3172.792.5312.2922.6962.9942.529Oral Rat Chronic Toxicity (LOAEL) (log mg/kg_bw/day)2.1731.9712.0141.9132.0312.4912.1851.681−0.117HepatotoxicityYesNoNoNoNoNoNoNoYesSkin SensitisationNoNoNoNoNoNoNoNoNoT.Pyriformis toxicity (log ug/L)0.2851.5420.2860.2880.2890.3130.2970.3220.285Minnow toxicity (log mM)−4.1090.5521.3641.2530.3841.0591.2410.2685.257 Table 3Toxicity of the selected fungal secondary metabolites and standard ponatinib.Name of the toxicity modelMSID00080MSID000814MSID001082MSID001179MSID001200MSID001266MSID001271MSID0013850LI (Ponatinib)AMES toxicityNoNoNoNoNoYesYesNoYesMax. tolerated dose (human) ((log mg/kg/day))0.08−0.4640.5430.6340.3810.4460.51−0.9260.126hERG I inhibitorNoNoNoNoNoNoNoNoYeshERG II inhibitorYesNoYesYesYesYesYesYesYesOral Rat Acute Toxicity (LD50) (mol/kg)3.0092.2742.3172.792.5312.2922.6962.9942.529Oral Rat Chronic Toxicity (LOAEL) (log mg/kg_bw/day)2.1731.9712.0141.9132.0312.4912.1851.681−0.117HepatotoxicityYesNoNoNoNoNoNoNoYesSkin SensitisationNoNoNoNoNoNoNoNoNoT.Pyriformis toxicity (log ug/L)0.2851.5420.2860.2880.2890.3130.2970.3220.285Minnow toxicity (log mM)−4.1090.5521.3641.2530.3841.0591.2410.2685.257 Toxicity of the selected fungal secondary metabolites and standard ponatinib. We explored the 3-dimensional interactions of the T315I-BCR::ABL1 mutant protein with eight fungal secondary drug-like compounds: Phellifuropyranone A (MSID001271), (+)-(R)-grifolinone C (MSID000080), Meshimakobnol B (MSID001179), Inoscavin D (MSID001082), Methylinoscavin C (MSID001200), Phellibaumin A (MSID001266), Schizine B (MSID001385), and Ganocin D (MSID000814) (Fig. 2). Additionally, Fig. 2 shows the 3-dimensional interaction of the T315I-BCR::ABL1 mutant protein with the standard drug Ponatinib. Furthermore, we explored the 2-dimensional interactions of the T315I-BCR::ABL1 mutant protein with the eight fungal secondary drug-like compounds mentioned earlier (Fig. 3). In this evaluation, detailed information such as the number of amino acids involved and the types of bonds formed by these compounds with the mutant protein are depicted (Fig. 3). Additionally, the 2-dimensional interaction of the T315I-BCR::ABL1 mutant protein with the standard drug Ponatinib is illustrated in Fig. 2.Fig. 2The 3-dimensional interactions of 8 drug-like fungal secondary metabolites with the T315I- BCR::ABL1 mutant protein.Fig. 3The 2-dimensional interactions of 8 drug-like fungal secondary metabolites with the T315I- BCR::ABL1 mutant protein. The 3-dimensional interactions of 8 drug-like fungal secondary metabolites with the T315I- BCR::ABL1 mutant protein. The 2-dimensional interactions of 8 drug-like fungal secondary metabolites with the T315I- BCR::ABL1 mutant protein. Trajectory analysis of simulated complexes Considering both the binding affinity of fungal secondary metabolites with the T315I-BCR::ABL1 mutant protein and the drug-likeness characteristics of these compounds, we decided to focus on the top three protein–ligand complexes, including protein-Phellifuropyranone A, protein-(+)-(R)-grifolinone C, and protein-Meshimakobnol B for molecular dynamics simulation. Additionally, we included the protein-Ponatinib (co-crystal compound) complex in our selection for molecular dynamics simulation. This approach aims to comprehensively explore their interactions and dynamics, providing valuable insights into their potential therapeutic efficacy. The mean Root Mean Square Deviation (RMSD) (Fig. 4A) for C-alpha atoms is 0.28 nm, with a maximum value of 0.42 nm and a minimum value of 0.0005 nm. For the Protein-Phellifuropyranone A complex, the RMSD averages at 0.20 nm, with a maximum of 0.38 nm and a minimum of 0.0005 nm (Fig. 4A) In contrast, the Protein-(+)-(R)-grifolinone C complex shows a mean RMSD of 0.29 nm, with a maximum of 0.52 nm and a minimum of 0.0005 nm (Fig. 4A). The Protein-Meshimakobnol B complex exhibits a mean RMSD of 0.17 nm, with a maximum of 0.27 nm and a minimum of 0.0004 nm (Fig. 4A). Notably, the standard ponatinib and protein complex displays slightly higher RMSD values compared to the other three complexes, with a mean RMSD of 0.86 nm, a maximum of 1.20 nm, and a minimum of 0.0005 nm (Fig. 4A).The RMSD values indicate the level of structural deviation between the protein complexes and their respective ligands or standards. Lower RMSD values (such as 0.17 nm for Protein-Meshimakobnol B) suggest closer similarity or less deviation between the protein structure and the ligand. Higher RMSD values (such as 0.86 nm for ponatinib and protein complex) indicate greater structural deviation, possibly suggesting a less stable or more flexible interaction between the protein and ligand.Fig. 4The comparative RMSD (A) and RMSF (B) plots for Mutant protein-Phellifuropyranone A, Mutant protein-(+)-(R)-grifolinone C, Mutant protein-Meshimakobnol B, and Mutant protein-Ponatinib complex. The comparative RMSD (A) and RMSF (B) plots for Mutant protein-Phellifuropyranone A, Mutant protein-(+)-(R)-grifolinone C, Mutant protein-Meshimakobnol B, and Mutant protein-Ponatinib complex. The Root Mean Square Fluctuation (RMSF) of C-alpha atoms was analyzed across several protein complexes, revealing varied levels of structural flexibility (Fig. 4B). The mean RMSF for C-alpha atoms was observed to be 0.11 nm, with a minimum value of 0.04 nm and a maximum of 0.88 nm. Specifically, the Protein-Phellifuropyranone A complex, Protein-(+)-(R)-grifolinone C complex, and Protein-Meshimakobnol B complex exhibited mean RMSF values of 0.15 nm, with maximum values ranging from 0.92 nm to 1.27 nm and minimum values consistently at 0.04 nm (Fig. 4B). In contrast, the standard ponatinib and protein complex demonstrated slightly higher RMSF values overall, averaging at 0.15 nm with a maximum of 0.97 nm (Fig. 4B). These results highlight significant variations in the flexibility of C-alpha atoms among different protein–ligand complexes, suggesting diverse structural dynamics that may influence their functional roles and interactions within biological systems or drug binding mechanisms. The radius of gyration (Rg) of C-alpha atoms was analyzed across several protein complexes, revealing varied levels of structural compactness. The mean Rg for C-alpha atoms across the complexes was found to be 2.06 nm, with individual complexes showing minor deviations: the minimum Rg observed was 2.00 nm and the maximum was 2.12 nm (Fig. 5A). Specifically, the Protein-Phellifuropyranone A complex exhibited an average Rg of 2.08 nm, ranging from a minimum of 2.02 nm to a maximum of 2.15 nm. Similarly, the Protein-(+)-(R)-grifolinone C complex and the Protein-Meshimakobnol B complex showed mean Rg values of 2.09 nm and 2.07 nm, respectively, with maximum values of 2.15 nm and 2.13 nm, and minimum values consistently around 2.02 nm (Fig. 5A). Notably, the standard ponatinib and protein complex displayed almost similar Rg values to the other three complexes, with a mean Rg of 2.09 nm, a maximum of 2.15 nm, and a minimum of 2.03 nm (Fig. 5A). These findings highlight slight variations in the compactness of protein structures across different complexes, indicating that while they share similar overall dimensions, subtle differences in structural organization may exist that could influence their functional properties or interactions within biological contexts.Fig. 5The comparative Rg (A) and SASA (B) plots for Mutant protein-Phellifuropyranone A, Mutant protein-(+)-(R)-grifolinone C, Mutant protein-Meshimakobnol B, and Mutant protein-Ponatinib complex. The comparative Rg (A) and SASA (B) plots for Mutant protein-Phellifuropyranone A, Mutant protein-(+)-(R)-grifolinone C, Mutant protein-Meshimakobnol B, and Mutant protein-Ponatinib complex. The analysis of solvent-accessible surface area (SASA) across various protein complexes reveals nuanced differences in their surface exposure. The mean SASA for the protein structures was calculated at 159.38 nm2, with values ranging from a minimum of 149.32 nm2 to a maximum of 173.23 nm2. Specifically, the Protein-Phellifuropyranone A complex exhibited an average SASA of 155.52 nm2, ranging between 146.16 nm2 and 167.46 nm2 (Fig. 5B). Similarly, the Protein-(+)-(R)-grifolinone C complex showed a mean SASA of 158.41 nm2, with a minimum of 149.55 nm2 and a maximum of 167.68 nm2. The Protein-Meshimakobnol B complex displayed a mean SASA of 159.50 nm2, ranging from 148.56 nm2 to 170.19 nm2 (Fig. 5B). Notably, the standard ponatinib and protein complex demonstrated SASA values close to the other three complexes, with a mean of 157.81 nm2, a maximum of 170.50 nm2, and a minimum of 148.0 nm2 (Fig. 5B). These findings suggest that while the overall surface accessibility (SASA) of these proteins and their complexes remains relatively consistent, subtle variations exist among different complexes. Such differences may reflect distinct structural conformations or binding interactions with their respective ligands. Understanding SASA variations is crucial for elucidating how these proteins interact with their environment or potential ligands, influencing their functional roles in biological systems or drug design applications. Phellifuropyranone A, a fungal secondary metabolite, forms nearly 3 hydrogen bonds with the active site of the mutant protein (Fig. 6A). the (+)-(R)-grifolinone C is observed to establish 2 hydrogen bonds with the same active site of the mutant protein (Fig. 6B). Comparing these interactions, the highest number of hydrogen bonds, amounting to 4, is found in the complex between the mutant protein active site and Meshimakobnol B, as depicted in Fig. 6C. In contrast, the interaction between Ponatinib and the protein exhibits fluctuations during simulation, as illustrated in Fig. 6D. The number of hydrogen bonds is very low between 0 to 60 ns, and then there is a slight increase in hydrogen bonds from 60 ns to 100 ns for Ponatinib-protein complex. This variability suggests that other complexes may be more stable than the Ponatinib-protein complex over the course of the simulation period.Fig. 6The dynamics of intermolecular hydrogen bonds between the mutant protein and lead fungal metabolites. A. Number of hydrogen bond between Protein and Phellifuropyranone A. B. Number of hydrogen bond between protein and (+)-(R)-grifolinone C. C. Number of hydrogen bond between protein and meshimakobnol B. D. Number of hydrogen bond between protein and ponatinib. The dynamics of intermolecular hydrogen bonds between the mutant protein and lead fungal metabolites. A. Number of hydrogen bond between Protein and Phellifuropyranone A. B. Number of hydrogen bond between protein and (+)-(R)-grifolinone C. C. Number of hydrogen bond between protein and meshimakobnol B. D. Number of hydrogen bond between protein and ponatinib. Principal components analysis Principal Component Analysis (PCA) is a powerful statistical technique used to analyze and visualize the collective motions of molecules or complexes based on their structural fluctuations. Initially, PCA through MD simulation identifies changes in protein motion over its trajectory. This involves calculating eigenvectors (total atom movement) and eigenvalues (atomic contributions to each movement), providing insights into structural and conformational changes in protein induced by ligand presence. Upon closely examining each docked complex, it becomes evident that there is a significant drop in eigenvalues beyond the first three out of the twenty eigenvalues. This observation suggests that the structure of the docked protein–ligand complex exhibits substantial flexibility during the initial stages of simulation. Principal Component Analysis (PCA) was conducted to assess the stability of various mutant protein complexes. In the simulation, the blue dots depict the initial conformations, followed by white dots representing intermediate states, and finally, red dots indicating conformations towards the end of the simulation period (Fig. 7). The result indicated that the first three PCs represented 49.6 %, 43.9 %, 50.2 %, and 46.9 % in the movement variance observed in the trajectories of Protein-Phellifuropyranone A, Protein-(+)-(R)-grifolinone C, Protein- Meshimakobnol B, Protein-ponatinib, respectively (Fig. 7). Protein- Meshimakobnol B compelx had the highest value followed by the Protein-Phellifuropyranone A (Fig. 7). During the 100 ns molecular dynamics (MD) simulations, the conformations of each system exhibited dynamic behavior, with significant fluctuations before eventually stabilizing into a dominant state. The results indicate that the mutant protein forms more stable complexes with Phellifuropyranone A (Fig. 7A), (+)-(R)-grifolinone C (Fig. 7B), and Meshimakobnol B (Fig. 7C). In contrast, the complex between the mutant protein and Ponatinib (Fig. 7D) is less stable state. This suggests that Phellifuropyranone A, (+)-(R)-grifolinone C, and Meshimakobnol B might have stronger or more favorable interactions with the mutant protein, leading to greater stability, as compared to Ponatinib.Fig. 7Principal component analysis. 2-diensional projection of trajectories into the top three eigenvectors (PC1, PC2, and PC3) for a mutant protein in complex with Phellifuropyranone A (A), (+)-(R)-grifolinone C (B), Meshimakobnol B (C), and Ponatinib (D). The red dots and blue dots stood for the stable state. And the intermediate state among these two states was shown by light red or light blue dot. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Principal component analysis. 2-diensional projection of trajectories into the top three eigenvectors (PC1, PC2, and PC3) for a mutant protein in complex with Phellifuropyranone A (A), (+)-(R)-grifolinone C (B), Meshimakobnol B (C), and Ponatinib (D). The red dots and blue dots stood for the stable state. And the intermediate state among these two states was shown by light red or light blue dot. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) A dynamic cross-correlation matrix (DCCM) was generated to analyze correlated or anti-correlated motions within the mutant protein when bound to various fungal secondary metabolites. Fig. 8A-D illustrates the DCCM maps depicting these protein–ligand interactions. The correlation coefficients in the matrix range from −1 to 1, with corresponding color intensities indicating the strength of correlation or anti-correlation. In the DCCM maps, cyan colors denote positive correlation values, ranging from 0.25 to 1, highlighting regions where protein and ligand movements are synchronized (Fig. 8A-D). Conversely, magenta colors represent negative correlation values, ranging from −0.25 to −1, indicating regions where movements are anti-correlated. Lighter shades of cyan and magenta indicate weaker or negligible correlations, with values between −0.25 and +0.25 (Fig. 8A-D). These color-coded representations allow for a visual comparison of the dynamic interactions across different protein–ligand complexes. Specifically, cyan and magenta regions can be compared to discern variations in correlated or anti-correlated motions induced by the various studied fungal secondary metabolites.Fig. 8DCCM plots for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. DCCM plots for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. In the analysis of mutant-protein complexes with Phellifuropyranone A (Fig. 8A), distinct correlation patterns are observed in the DCCM. Specifically, regions highlighted in cyan and magenta indicate anti-correlated movements, which are notably more pronounced in the mutant-protein-Phellifuropyranone A complex compared to other complexes. In these same regions, other complexes exhibit negative correlations, although less intense than those observed with mutant-protein-Phellifuropyranone A (Fig. 8A). Comparatively, in complexes involving Protein-Phellifuropyranone A, Protein-(+)-(R)-grifolinone C (Fig. 8B), and Protein-Meshimakobnol B (Fig. 8C), anti-correlated movements in analogous regions are less prominent when contrasted with Protein-ponatinib complex (Fig. 8D). However, positive correlations are consistently present across all systems, albeit with variations among the complexes. Comparing complexes of proteins reveals changes in residue movements, indicating that interactions with ligands affect these movements. These findings underscore the dynamic nature of protein–ligand interactions, where specific compounds induce distinct patterns of correlated and anti-correlated motions. The varying intensities and types of correlations observed provide insights into the structural dynamics influenced by different fungal secondary metabolites. Porcupine Analysis was conducted to examine the principal motions within ligand-bound complexes. In this analysis, the spikes correspond to the direction of fluctuations along PC1 and PC2, while their length indicates the magnitude of these fluctuations observed in the Cα atom backbone of the protein. This method provides a visual representation of the essential dynamic movements induced by the ligand binding, highlighting significant directional changes and their relative intensities across the protein structure. The fluctuations in PC1 and PC2 for the Mutant protein complexes bound to Phellifuropyranone A, (+)-(R)-grifolinone C, Meshimakobnol B, and Ponatinib are depicted in Fig. 9. The analysis reveals distinct levels of fluctuation among these complexes specifically along PC1. This variation suggests that each mutant protein complex exhibits unique patterns of dynamic movements influenced by their respective ligands. The differing levels of fluctuations along PC1 and PC2 highlight how each ligand induces specific structural dynamics within the protein backbone. These insights from Fig. 9 provide valuable information regarding the functional implications and binding effects of these ligands on the mutant protein structure.Fig. 9Porcupine analysis for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex of PC1. The red spikes represent the direction of fluctuations presented by PC1, and their size represents the magnitude of the fluctuation observed in the Cα atom backbone of the protein. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Porcupine analysis for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex of PC1. The red spikes represent the direction of fluctuations presented by PC1, and their size represents the magnitude of the fluctuation observed in the Cα atom backbone of the protein. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Free energy landscape analysis Free energy landscapes offer valuable insights into the range of conformations adopted by a protein structure. Using the first two principal components, Gibbs free energy values were computed for each conformer of the protein. Fig. 10 displays contour plots illustrating the transition from high to low energy states, highlighting energy minima basins. These basins are depicted by a color gradient shifting from red to blue, with concentrated blue spots indicating metastable states of the protein and its complexes.Fig. 10Free energy landscape in three dimensional and two dimensional representations for (A-B) Mutant protein-Phellifuropyranone A, (C-D) Mutant protein-(+)-(R)-grifolinone C, (E-F) Mutant protein-Meshimakobnol B, and (G-H) Mutant protein-Ponatinib complex. Free energy landscape in three dimensional and two dimensional representations for (A-B) Mutant protein-Phellifuropyranone A, (C-D) Mutant protein-(+)-(R)-grifolinone C, (E-F) Mutant protein-Meshimakobnol B, and (G-H) Mutant protein-Ponatinib complex. In the analysis of mutant protein complexes with various ligands, distinct patterns in energy basin distribution were observed. Specifically, for the Mutant protein-Phellifuropyranone A complex, the energy basin divides into two clusters of low-energy conformers, with one cluster exhibiting concentrated minima (Fig. 10A-B). This indicates that while the protein maintains stability throughout the simulation, it also displays inherent flexibilities. Similarly, in the Mutant protein-(+)-(R)-grifolinone C complex, a single cluster with concentrated minima is observed (Fig. 10C-D). In contrast, the Mutant protein-Meshimakobnol B complex shows an energy basin divided into several distinct minima, with one cluster demonstrating concentrated minima (Fig. 10E-F). Lastly, the Mutant protein-Ponatinib complex displays only one cluster with concentrated minima within its energy basin (Fig. 10G-H). These observations suggest that each mutant protein–ligand complex exhibits unique energy landscape characteristics, reflecting varying levels of stability and flexibility induced by their respective ligands. Binding free energy analysis The g_mmpbsa module was utilized for performing binding free energy calculations using an MM/PBSA-based approach. This method was applied to determine the binding free energies of complexes involving mutant proteins. Analysis focused on whole 100 ns of the overall trajectory. To estimate the binding of the ligands with the proteins throughout the 100 ns simulation, the free binding energy of these compounds was calculated. The results of these calculations are summarized in Fig. 11. Fig. 11 presents a comprehensive analysis of the thermodynamic interactions between the fungal compounds and the protein, highlighting their potential to inhibit mutant proteins. The figure also provides a summary of additional energy contributions, specifically detailing the electrostatic energy, non-polar solvation energy, solvation energy, gas-phase energy, and van der Waals intermolecular components (Fig. 11). These insights are crucial for understanding how these compounds interact with the protein at a molecular level, offering valuable information for drug development and therapeutic targeting of specific protein mutants (Figure). In this study, the binding free energies (ΔGbind) of four compounds were evaluated: Phellifuropyranone A (–22.88 ± 4.28 kcal/mol) (Fig. 11A), (+)-(R)-grifolinone C (–23.29 ± 12.82 kcal/mol) (Fig. 11B), Meshimakobnol B (−25.86 ± 3.51 kcal/mol) (Fig. 11C), and Ponatinib (−25.54 ± 6.31 kcal/mol) (Fig. 11D). Notably, Phellifuropyranone A and Meshimakobnol B demonstrated binding energies comparable to the reference compound Ponatinib. The van der Waals contribution for each ligand was also examined, revealing negative values for Phellifuropyranone A (−46.75 ± 2.96), (+)-(R)-grifolinone C (−76.82 ± 4.46), Meshimakobnol B (−46.23 ± 2.68), and Ponatinib (−50.48 ± 3.38) (Fig. 11A-D). These negative values indicate favorable van der Waals interactions between the ligands and the protein, contributing significantly to their binding affinities.Fig. 11Calculated Energies in kcal/mol and its Components Using MM-PBSA Method Obtained from the whole trajectory analysis. The energy calculation plot for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. Calculated Energies in kcal/mol and its Components Using MM-PBSA Method Obtained from the whole trajectory analysis. The energy calculation plot for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. However, the binding profile of (+)-(R)-grifolinone C exhibited greater variability (Fig. 12B). During the simulation, instances were observed (at 69.01 ns, 87.01 ns, 88.01 ns) where positive values were recorded in the protein–ligand interaction plot (Fig. 12B). This suggests that at these specific time points, the ligand was notably distant from the protein, possibly indicating transient or less stable binding interactions. Interestingly, despite these fluctuations, Phellifuropyranone A (Fig. 12A) and Meshimakobnol B (Fig. 12C) consistently showed favorable binding energies comparable to Ponatinib (Fig. 12D) throughout the simulation period. This suggests their potential as effective ligands in the context of the study. In summary, while (+)-(R)-grifolinone C displayed variability in its binding behavior, Phellifuropyranone A and Meshimakobnol B exhibited stable and favorable binding energies akin to the reference compound Ponatinib, highlighting their potential utility in similar studies or applications.Fig. 12Binding free energy plot for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. Binding free energy plot for (A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. Per-residue decomposition analysis Next, the binding free energy for each contributing residue was calculated using the g_mmbpsa program (Fig. 13). For Mutant protein-Phellifuropyranone A complex, the Lys271 (−1.63 kcal/mol), VAL289 (−0.79 kcal/mol) MET290 (−1.53 kcal/mol), VAL299 (−0.77 kcal/mol), ILE315 (−1.63 kcal/mol), PHE317 (−-1.16 kcal/mol), MET318 (−1.10 kcal/mol), LEU370 (−-0.99 kcal/mol) and PHE382 (−1.23 kcal/mol) residues were the top contributors to bind with Phellifuropyranone A (Fig. 13A). For Mutant protein-(+)-(R)-grifolinone C complex, Lys271 (−0.87 kcal/mol), VAL289 (−1.11 kcal/mol), MET290 (−1.39 kcal/mol) and PHE359 (−1.53 kcal/mol) residues were the top contributors for binding with (+)-(R)-grifolinone C (Fig. 13B). ASP381 (8.42 kcal/mol), ASP363 (6.50 kcal/mol), GLU286 (4.74 kcal/mol), HIS361 (3.47 kcal/mol), and ARG362 (1.23 kcal/mol) were the top residues contributing positive energy to the binding with −(+)-(R)-grifolinone C (Fig. 13B). The positive energy values indicate that these residues contribute to the instability of the interaction with the target protein, suggesting a weak or unfavorable binding affinity. For Mutant protein-Meshimakobnol B complex, HIS361 (−2.26 kcal/mol), ALA269 (−1.75 kcal/mol), VAL289 (−1.28 kcal/mol), MET290 (−2.14 kcal/mol), and ILE315 (−1.27 kcal/mol) residues were the top contributors for binding with Meshimakobnol B (Fig. 13C). The control Mutant protein-Ponatinib complex displayed the least values for binding with VAL289 (−1.87 kcal/mol), MET290 (−1.80 kcal/mol), ILE315 (−1.03 kcal/mol), PHE359 (−1.14 kcal/mol) residues of mutated protein (Fig. 13D). Interestingly, our findings revealed that the mutated residue ILE315 binds to Phellifuropyranone A and Meshimakobnol B with significantly lower energy compared to the approved drug ponatinib. This indicates that both Phellifuropyranone A and Meshimakobnol B have a crucial interaction with the T315I-BCR::ABL1 mutant protein.Fig. 13Residue wise binding free contribution energy for A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. Residue wise binding free contribution energy for A) Mutant protein-Phellifuropyranone A, (B) Mutant protein-(+)-(R)-grifolinone C, (C) Mutant protein-Meshimakobnol B, and (D) Mutant protein-Ponatinib complex. Discussion In this study, our focus was on understanding the structural implications of fungal secondary metabolites binding to the T315I-BCR::ABL1 mutant protein, a significant mutation associated with chronic myeloid leukemia (CML). This specific mutation alters the BCR::ABL1 protein, making it resistant to tyrosine kinase inhibitors, which are a crucial class of drugs used to treat CML. To conduct our investigation, we retrieved the three-dimensional structure of the T315I-BCR::ABL1 mutant protein from the RCSB Protein Data Bank and meticulously prepared it for molecular docking experiments. Molecular docking allowed us to simulate and analyze how these fungal secondary metabolites interact with the mutant protein at a molecular level. Understanding these interactions is pivotal for potentially developing new therapeutic strategies or modifying existing drugs to overcome resistance mechanisms in CML treatment. Initially, we screened compounds based on their binding affinity, prioritizing those that exhibited superior binding affinity compared to the approved drug ponatinib (Table 1). Through molecular docking and ADMET studies, we identified eight compounds with promising binding affinities (Table 2) and drug-like properties (Table 2). These compounds are Phellifuropyranone A (MSID001271), (+)-(R)-grifolinone C (MSID000080), Meshimakobnol B (MSID001179), Inoscavin D (MSID001082), Methylinoscavin C (MSID001200), Phellibaumin A (MSID001266), Schizine B (MSID001385), and Ganocin D (MSID000814). Our findings revealed that these compounds exhibit significant binding affinities comparable to or greater than ponatinib, an approved drug for targeting the T315I-BCR::ABL1 mutant protein in chronic myeloid leukemia. Importantly, these compounds also demonstrated acceptable toxicity levels and possess drug-like properties, making them potential candidates for further development. Based on these results, we selected the top three complexes formed between these compounds and the mutant protein, along with the standard protein-ponatinib complex, for molecular dynamics simulation studies. These simulations aim to investigate the stability and dynamic behavior of these complexes, providing insights into their potential efficacy as therapeutic agents against the resistant T315I-BCR::ABL1 mutant protein. Next, we conducted a 100 ns molecular dynamics simulation on the complexes involving the free protein T315I-BCR::ABL1 bound to Phellifuropyranone A, (+)-(R)-grifolinone C, and Meshimakobnol B, alongside the control complex of protein-ponatinib. The RMSD graphs revealed that the trajectories of T315I-BCR::ABL1 bound to Phellifuropyranone A and Meshimakobnol B converged to stable and uniform patterns (Fig. 4A). T315I-BCR::ABL1-(+)-(R)-grifolinone C exhibited relatively less stable behavior throughout the simulation. In contrast, the RMSD curve of the standard protein-ponatinib complex showed noticeable fluctuations over the simulation period. These results suggest that Phellifuropyranone A and Meshimakobnol B may offer more stable binding interactions with the mutant protein T315I-BCR::ABL1 compared to both (+)-(R)-grifolinone C and the approved drug ponatinib. The RMSF analysis was performed to evaluate the inherent flexibilities of the protein, revealing a consistent trend in protein flexibility across all complexes studied upon binding with ligands. The observed RMSF values showed minor fluctuations but remained fairly similar to each other (Fig. 4B). Next, we employed Rg analysis to examine the overall compactness of the protein structures across the studied complexes (Fig. 5A). The analysis revealed a consistent pattern among the complexes, where the observed Rg values exhibited minor fluctuations but remained fairly similar to each other. In contrast, control protein-ponatinib showed a slightly higher Rg during the simulation period. This suggests localized changes in protein compactness upon ligand binding. Since the protein is present in a solvent for the simulation, we have studied the SASA (Fig. 5B) for the protein and its complexes which showed a trend of Protein-Meshimakobnol B > protein > Protein-(+)-(R)-grifolinone C > Protein-ponatinib > Protein-Phellifuropyranone A, which shows that upon ligand binding significant changes occur in the tertiary structure of the protein. The high SASA value of Protein-Meshimakobnol B is credited to the larger size of the ligand. Next, principal component analysis (PCA) was employed to identify essential motions amidst global dynamics. The analysis revealed significant decreases in Cα fluctuations in ligand-bound complexes, indicating substantial alterations in protein structure upon ligand binding. Two-dimensional plots of PC1, PC2, and PC3 illustrated notable changes in essential mobilities within the ligand-bound complexes (Fig. 7), suggesting distinct conformational shifts in the protein under investigation. Moreover, the contribution of PC1, PC2, and PC3 to residue movements was assessed, exploring specific regions of the protein that undergo essential motion in the presence of ligands. Dynamic cross-correlation matrix (DCCM) analysis further underscored changes in correlations and anti-correlations among residues in the complexes, aligning with observations from PCA regarding altered essential mobilities (Fig. 8). Additionally, Porcupine analysis consistently identified essential motions within the protein structures (Fig. 9). Taken together, PCA analysis indicates that ligand binding induces significant changes in the structural dynamics of the protein, particularly within its catalytic pocket. The analysis of free energy landscapes through Gibbs free energy computations using the first two principal components revealed distinct energy basin distributions among mutant protein–ligand complexes (Fig. 10). These contour plots illustrate transitions between high and low energy states, with colors shifting from red to blue denoting metastable states (Fig. 10). These findings underscore the unique energy landscape characteristics of each mutant protein–ligand complex. They not only highlight varying stability levels induced by different ligands but also suggest diverse modes of interaction and conformational dynamics within the catalytic pocket of mutant proteins. Such insights are crucial for understanding how these ligands modulate protein function and for guiding future drug design efforts targeting mutant proteins in therapeutic contexts. Finally, the results from binding free energy calculations support the findings observed in docking simulations. The protein-phellifuropyranone A complex exhibited a binding free energy of 22.88 ± 4.28 kcal/mol, while the protein-(+)-(R)-grifolinone C complex showed a value of –23.29 ± 12.82 kcal/mol, indicating instability in some instances (Fig. 11). In contrast, the protein-Meshimakobnol B complex demonstrated a more favorable binding free energy of −25.86 ± 3.51 kcal/mol (Fig. 11). The analysis of binding free energies further highlighted that the protein-(+)-(R)-grifolinone C complex displayed positive values in certain frames, suggesting potential instability (Fig. 12). Furthermore, in our analysis of per-residue decomposition, we identified specific residues crucial for the binding interactions with Phellifuropyranone A, Meshimakobnol B, and Ponatinib. Notably, MET290, VAL299, ILE315, and PHE359 were found to contribute significantly to the binding energies across all three compounds (Fig. 13). Interestingly, the mutant residue ILE315 was observed to contribute less binding free energies for Phellifuropyranone A, Meshimakobnol B, than the standard Ponatinib, underscoring their effective interaction with the T315I-BCR::ABL1 mutant protein (Fig. 13). However, this residue showed minimal contribution to binding with (+)-(R)-grifolinone C, indicating a distinct binding profile for this compound. In addition, ASP381 (8.42 kcal/mol), ASP363 (6.50 kcal/mol), GLU286 (4.74 kcal/mol), HIS361 (3.47 kcal/mol), and ARG362 (1.23 kcal/mol) were the top residues contributing positive energy to the binding with −(+)-(R)-grifolinone C (Fig. 13). These positive energy values suggest an unstable interaction and weak binding affinity of −(+)-(R)-grifolinone C with the target protein. These findings highlight specific molecular interactions crucial for the efficacy of Phellifuropyranone A, Meshimakobnol B, and Ponatinib against the mutant protein. Understanding these residue-specific interactions provides insights into the structural basis of their binding affinities and can inform further optimization of these compounds or the development of new therapeutics targeting the T315I mutation in chronic myeloid leukemia. When compared to the approved drug ponatinib, both phellifuropyranone A and Meshimakobnol B demonstrate notable potency as inhibitors of the T315I-BCR::ABL1 mutant protein in chronic myeloid leukemia. This observation suggests that these compounds could potentially offer effective therapeutic alternatives or complementary treatments for cases where ponatinib may not be fully effective or tolerated. Their ability to inhibit the mutant protein underscores their potential utility in overcoming drug resistance and improving treatment outcomes in clinical settings. Previous studies have highlighted the medicinal potential of phellifuropyranone A and Meshimakobnol B across various diseases. For instance, phellifuropyranone A has been noted for its moderate BCR::ABL1 cytotoxic activity and its potential role in anti-tumor treatments.32 Additionally, both phellifuropyranone A and Meshimakobnol B have demonstrated antiproliferative effects against mouse melanoma cells and human lung cancer cells in vitro.33 Moreover, research has identified phellifuropyranone A as possessing cytotoxic activity against several human cancer cell lines, emphasizing its promise as a valuable compound for developing novel anticancer therapies.34 Furthermore, studies have reported that both meshimakobnol B and phellifuropyranone A exhibit antitumor effects in both in vitro and in vivo models.35. Meshimakobnol B and phellifuropyranone A, both isolated from a Japanese mushroom, have been identified as effective growth inhibitors of cancer cells. Further biological evaluation revealed that the presence of specific functional groups in these compounds is crucial for their cytotoxic effects, indicating their importance in determining anticancer activity.36 It has been reported that drug compounds exhibit synergistic cytotoxicity when combined with anticancer agents, particularly in BCR-ABL positive leukemia cells.37 This synergy can enhance the effectiveness of treatment, providing a promising approach for overcoming resistance and improving patient outcomes. This indicates the importance of fungal metabolites as a promising source of novel anticancer agents. Fungi produce a diverse array of secondary metabolites, many of which exhibit bioactive properties that can target various cancer cell pathways.38 The ability of these compounds to inhibit the proliferation of aggressive cancer types, such as melanoma and lung cancer, indicates their potential therapeutic value. Previous studies have shown that numerous fungal metabolites possess unique mechanisms of action, including the induction of apoptosis, disruption of cell cycle progression, and modulation of signaling pathways involved in tumor growth. Fungi generate a wide variety of bioactive compounds, such as terpenes, polyphenols, and alkaloids, many of which demonstrate notable anticancer properties and have the potential to be developed as candidates or lead compounds for new anticancer medications.39 By isolating and characterizing these metabolites, researcher can explore new avenues for cancer treatment, particularly for patients who have developed resistance to conventional therapies. The ongoing investigation of fungal-derived compounds not only enriches our understanding of their pharmacological potential but also paves the way for the development of innovative therapeutic strategies that utilize the unique properties of these natural products. Given these findings, the therapeutic potential of phellifuropyranone A, Meshimakobnol B, and other fungal secondary metabolites against chronic myeloid leukemia, particularly in targeting the T315I-BCR::ABL1 mutant protein, warrants further exploration through comprehensive in vivo and in vitro studies. First, it is essential to assess the activity of these compounds in various cancer cell lines, both individually and in combination with existing anticancer therapies, to evaluate their efficacy in relevant biological contexts. Additionally, a thorough investigation of potential adverse events will be crucial for establishing a comprehensive safety profile, especially in combination treatments. The importance of examining drug solubility and exploring formulation strategies to enhance bioavailability cannot be overstated, as these factors are vital for effective clinical application. Finally, understanding off-target effects will be critical to ensuring the safety and specificity of these compounds. Addressing these areas will provide a clearer pathway for translating our findings into viable cancer treatments and contribute to the ongoing search for novel therapeutic options. Conclusions This study investigated the structural influences of ligand binding on the T315I-BCR::ABL1 mutant protein in chronic myeloid leukemia (CML). We conducted a comprehensive screening of 1,830 fungal secondary metabolites, comparing their effects with the approved drug ponatinib to identify potential inhibitors for the T315I-BCR::ABL1 mutant protein. Global dynamics studies, based on RMSD, RMSF, Rg analysis, and SASA, indicated that binding with Phellifuropyranone A and Meshimakobnol B stabilizes the protein. Additionally, PCA and FEL analyses revealed highly stable states in the T315I-BCR::ABL1 mutant protein complexes with these compounds, suggesting stable transitions during simulation. Binding free energy calculations showed comparable binding energies for the approved ponatinib and the potential inhibitors Phellifuropyranone A and Meshimakobnol B, indicating their promise as inhibitors against the T315I-BCR::ABL1 mutant protein in CML patients. While these findings provide valuable insights into the structural dynamics of the T315I-BCR::ABL1 mutant protein and suggest potential new therapeutic candidates, it is essential to consider the broader treatment landscape for CML. Given the effectiveness of existing therapies, including ponatinib, we must evaluate the clinical relevance and financial viability of developing new compounds. Future studies should investigate not only the efficacy of these metabolites in relevant cell lines and combination therapies but also address their safety profiles and potential adverse effects. Additionally, it is essential to evaluate drug solubility, formulation, and off-target effects, which are crucial for assessing the feasibility of these compounds as therapeutic options. Overall, these findings provide valuable insights into the structural dynamics of T315I-BCR::ABL1 mutant protein and suggest potential new therapeutic candidates for chronic myeloid leukemia. CRediT authorship contribution statement Dilinazi Abulaiti: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Niluopaer Tuerxun: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Formal analysis, Data curation. Huan Wang: Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation. Lina Ma: Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Formal analysis, Data curation. Fang Zhao: Writing – original draft, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation. Yang Liu: Writing – original draft, Visualization, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Jianping Hao: Writing – review & editing, Writing – original draft, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.