Molecular docking of colossolactones into the ATP-binding site and the allosteric pocket of EGFR .... To evaluate the binding interactions and structural stability of 16 colossolactone d
OVERVIEW
Epidermal Growth Factor Receptor
1.1.1 A strategic target in cancer therapy
The epidermal growth factor receptor (EGFR), a crucial tyrosine kinase receptor, belongs to the ErbB family, which includes EGFR (ErbB1), HER2 (ErbB2), HER3 (ErbB3), and HER4 (ErbB4) [14] This receptor is involved in regulating various essential biological processes such as proliferation, differentiation, and cell survival
EGFR is crucial for normal physiological functions; however, its abnormal activation or mutations are linked to various cancers, notably non-small cell lung cancer (NSCLC), breast cancer, and colorectal cancer (Figure 1.1).
Figure 1.1 Schematic diagram of the EGFR signaling pathway
EGFR activation is initiated when ligands like EGF or TGF-α bind to its extracellular domain, triggering receptor dimerization and autophosphorylation This process activates critical signaling pathways such as Ras/MAPK, PI3K/AKT, and JAK/STAT, which are essential for cell growth and survival Dysregulation of these pathways can lead to tumor development and progression, highlighting the importance of EGFR in cancer biology and as a therapeutic target.
EGFR is a significant target in cancer treatment Somatic mutations and EGFR overexpression in cancer cells create a favorable “therapeutic window” for selective
3 targeting [18] Current therapeutic strategies include monoclonal antibodies, kinase inhibitors, ΔEGFR peptide vaccines [19], and cytotoxic agent delivery via EGFR [20]
EGFR is a key target in cancer therapy due to its central role in the cancer signaling network, where pathways like Ras/MAPK and PI3K/AKT are interconnected through complex feedback loops, reducing the effectiveness of targeting individual components Inhibiting pathways such as PI3K/AKT or mTOR can unintentionally activate alternative signaling mechanisms, like MAPK or AKT via feedback activation, highlighting the need for a more direct approach As a cell surface receptor and kinase often overexpressed or genetically altered in cancer, EGFR serves as a promising target for disrupting oncogenic signaling at its source, making it an ideal candidate for therapeutic intervention.
1.1.2 Structure of the Epidermal Growth Factor Receptor
Figure 1.2 Structural diagram of EGFR
Notes: (a) The full-length EGFR chain showing the ECD in the open conformation, the transmembrane (TM) domain, and the intracellular kinase domain Note the distance between domains II and IV The dimerization arm in domain II is exposed and ready for dimer interaction
(b) The ECD structure of EGFR in the closed state (PDB ID: 1NQL) Domain IV folds back and interacts with domain II In this closed state, the dimerization arm of domain II is not accessible for dimerization
The ErbB family proteins feature an extracellular domain (ECD) that includes a glycosylated N-terminal segment of approximately 620 amino acids, encompassing the ligand-binding site This ECD is structurally divided into four domains (I–IV): domains I and III adopt a beta-helix/solenoid configuration, while domains II and IV are rich in cysteine residues and contain disulfide modules, contributing to their stability and functionality.
A hydrophobic transmembrane domain (TMD) connects the extracellular and intracellular regions via a juxtamembrane domain (JMD) Structural analyses reveal that the JMD consists of two key parts: the extracellular juxtamembrane segment (eJM), a short seven-amino-acid stretch linking the TMD to the C-terminus of domain IV in the extracellular domain (ECD), and the intracellular juxtamembrane segment (iJM), which separates the TMD from the kinase domain.
The intracellular region of ErbB proteins consists of approximately 540 amino acids, including a crucial tyrosine kinase domain (TKD) and a carboxyterminal tail (CTT) of about 230 amino acids The C-terminal tail plays a vital role in phosphorylation due to its multiple tyrosine residues, although its three-dimensional structure remains unclear because of its high flexibility.
Figure 1.3 Structure of the EGFR tyrosine kinase domain
The tyrosine kinase domain of EGFR comprises five key regions: the N-lobe, C-lobe, hinge region, ATP-binding pocket, and allosteric pocket (Figure 1.3) [4], [31], [33] The N-lobe is particularly important because it contains the C-helix, which plays a crucial role in EGFR activation by enabling the transition from an inactive to an active conformation [34].
The N- and C-lobes of the kinase are connected by a narrow hinge region that houses the ATP-binding pocket, with Thr790 serving as the critical “gatekeeper” residue The T790M mutation, one of the most common resistance mechanisms, enhances ATP affinity within this pocket, thereby decreasing the effectiveness of tyrosine kinase inhibitors (TKIs) This mutation plays a key role in mediating resistance to targeted therapies by preventing TKIs from adequately binding to the kinase.
The allosteric pocket, adjacent to the ATP-binding site, only appears when the α-
The outward movement of the C helix causes EGFR to assume its inactive state, disrupting its active conformation Key structural elements such as the Lys745–Glu762 salt bridge and the DFG motif (D855, F856, and G857) are crucial for closing the allosteric pocket, thereby stabilizing the active form of EGFR Understanding these conformational shifts is essential for targeting EGFR in therapeutic interventions.
1.1.3 Monoclonal Antibodies and Tyrosine Kinase Inhibitors
Currently, two effective strategies are used to inhibit EGFR: monoclonal antibodies (MAbs) and tyrosine kinase inhibitors (TKIs) [39] Monoclonal antibodies like Cetuximab [40] and Panitumumab [41] target the extracellular domain of EGFR, blocking ligand binding and preventing receptor activation Additionally, some MAbs can enhance immune responses by stimulating antibody-dependent cellular cytotoxicity (ADCC) to eliminate cancer cells [39].
Unlike monoclonal antibodies (MAbs), tyrosine kinase inhibitors (TKIs) specifically target the intracellular kinase domain of EGFR, effectively inhibiting receptor phosphorylation and disrupting downstream signaling pathways Currently, four generations of EGFR-TKIs have been developed to improve treatment outcomes for non-small cell lung cancer (NSCLC).
First-generation TKIs, such as Gefitinib (FDA-approved in 2003), Erlotinib, Lapatinib, and Icotinib, are quinazoline-based compounds that selectively inhibit the epidermal growth factor receptor (EGFR) by reversibly binding to the ATP-binding pocket of its kinase domain This binding prevents receptor phosphorylation and subsequent downstream signaling pathways, effectively blocking tumor cell proliferation and survival.
Colossolactone compounds and their applications
Natural products have historically been a vital resource in the pharmaceutical industry, driving drug innovation and development Research shows that naturally derived compounds can effectively inhibit cancers with EGFR mutations by modulating key signaling pathways As synthetic drugs often cause adverse effects and resistance within the EGFR tyrosine kinase domain, exploring natural EGFR-TKIs offers a promising alternative, particularly for countries like Vietnam with abundant natural resources.
Medicinal mushrooms of the genus Ganoderma (family Ganodermataceae) have been utilized in traditional Asian medicine for centuries, with Ganoderma lucidum
Reishi, also known as Lingzhi mushroom, is a prime example of medicinal fungi with significant health benefits (Figure 1.5a) [68] Its chemical constituents have been extensively studied, highlighting lanostane-type triterpenoids as the major secondary metabolites responsible for its diverse biological activities [69].
Ganoderma colossum (Figure 1.5b) is a distinctive mushroom species characterized by its yellowish surface and notable bioactivities, including antibacterial, anticancer, and anti-HIV effects Research by Isaka et al has shown that G colossum is a rich source of structurally diverse, physiologically active triterpenoids, particularly lanostane-type triterpenes with six- or seven-membered lactone rings called colossolactones These unique compounds contribute to the mushroom's significant medicinal potential and make G colossum a valuable subject for further pharmacological studies.
Figure 1.5 Representative species of the genus Ganoderma
Colossolactones possess a steroidal backbone and exist in various forms such as oxa-A-homo-steroidal or oxa-A,B-dihomo-steroidal derivatives To date, a total of 16 colossolactones have been isolated from G colossum, including colossolactones A–H
[6], [7], colossolactones I–VIII [74], [77] and most recently, colossolactone J [76] Their chemical structures and related information are summarized in Table 1.1
The pharmacological potential of these colossolactones has been validated in multiple studies Specifically, colossolactones V, VI, and E exhibited inhibitory activity against HIV-1 protease with IC50 values of 9.0, >100, and 8.0 μg/mL, respectively [74] While colossolactone E and 23-hydroxy-colossolactone E demonstrated antibacterial activity against Bacillus subtilis and Pseudomonas syringae, colossolactone B showed no such effects [72]
Colossolactone E also exhibited antimalarial activity against Plasmodium falciparum with an IC50 of 10.0 àM [78] Notably, colossolactone IV has been shown to exert anticancer effects on various cancer cell lines, including breast cancer (MCF-7), cervical cancer (HeLa), colorectal cancer (HCT-116), and hepatocellular carcinoma (HepG2), with IC50 values ranging from 4.9 to 64.2 àM and resistance fractions (R- fraction) below 47.5% [79]
The structures retrieved from the PubChem database [80], as presented in Table
Colossolactone G has been experimentally demonstrated to enhance the effectiveness of existing chemotherapeutic agents, notably potentiating the anticancer activity of 5-fluorouracil (5-FU) and gemcitabine (GCB) in colorectal cancer cells When used in combination, colossolactone G produces a synergistic effect that promotes cell cycle arrest and induces apoptosis, thereby improving therapeutic outcomes.
Table 1.1 Structures of 16 colossolactone derivatives sourced from the PubChem database
Beyond colorectal cancer, colossolactones H and G have also been reported to enhance the therapeutic efficacy of targeted agents such as gefitinib and sorafenib in NSCLC and hepatocellular carcinoma, respectively [7], [82] These findings not only underscore the therapeutic potential of naturally occurring colossolactones in supporting cancer treatment regimens but also lay a foundation for further mechanistic and clinical investigations into this class of compounds.
Computational methods for simulating and evaluating biomolecular
Molecular docking is a vital in silico technique extensively used in modern drug discovery to predict how small molecules, such as drug candidates or chemical compounds, interact with biological targets like enzymes or proteins Its primary aim is to identify the best binding pose of a ligand within a protein's active site, enabling estimation of binding affinity and interaction specificity Based on the "lock-and-key" model, molecular docking evaluates geometric and physicochemical complementarity to simulate molecular interactions effectively.
Figure 1.6 Illustration of molecular docking
Notes: The ligand is represented as an orange stick, the protein as a grey surface, and the active site is highlighted in purple
A typical molecular docking workflow involves predicting the ligand’s spatial conformation within the protein's active site and scoring the binding energy to assess complex stability Non-covalent interactions, including hydrogen bonds, hydrophobic interactions, electrostatic forces, and Van der Waals forces, are crucial in determining the binding affinity between the ligand and the protein.
Various software tools have been developed for molecular docking AutoDock and AutoDock Vina are two widely used open-source programs due to their efficiency
Advanced docking programs utilize genetic algorithms and Monte Carlo methods to explore optimal ligand binding conformations, followed by energy-based scoring functions to identify the best binding poses, enhancing accuracy and reliability in predicting ligand-receptor interactions [83] Widely used software such as GOLD, Glide, MOE-Dock, FlexX, Surflex, DOCK, LigandFit, FRED, and MCDock employ diverse search algorithms—including genetic, systematic, and stochastic methods—to efficiently optimize ligand-target configurations and accurately estimate binding energies [84,86], thereby improving the success rate of virtual screening and drug discovery efforts.
Molecular docking is essential for virtual screening, as it simulates the interactions between vast numbers of compounds and target proteins to identify potential drug candidates This technique provides valuable insights into the molecular mechanisms of action and aids in the chemical optimization of lead compounds to improve their biological activity.
Molecular docking has limitations, as most software treats proteins as rigid bodies, despite their inherent high flexibility in biological systems Additionally, scoring functions often fail to accurately predict binding free energy because they overlook solvent effects, ions, and pH fluctuations that occur in physiological environments To enhance the reliability of docking results, it is essential to validate them using advanced computational techniques like molecular dynamics simulations, complemented by in vitro or in vivo experimental confirmation.
Molecular Dynamics (MD) simulation is a vital computational technique in biochemistry and drug development, enabling detailed analysis of biomolecular behavior By solving Newton’s equations of motion for individual atoms, MD accurately models the structural dynamics of proteins, nucleic acids, lipids, and complex biomolecular assemblies This technique allows researchers to observe how these molecules evolve over time under conditions that closely mimic the physiological environment, providing essential insights for understanding biological functions and designing effective drugs.
Molecular dynamics (MD) offers dynamic insights into structural fluctuations and molecular interactions, unlike static methods like molecular docking This approach is crucial for understanding key biological processes such as protein folding, allosteric transitions, ligand-binding mechanisms, and the effects of mutations on protein structure and function [9], [88-90].
A standard molecular dynamics (MD) workflow involves system preparation—including proteins, ligands, solvents, and ions—followed by simulation execution, which encompasses energy minimization, NVT and NPT equilibration, and the production run, culminating in result analysis to extract dynamic insights Widely used MD software packages such as GROMACS, AMBER, NAMD, and CHARMM facilitate these processes, each supporting various force fields like AMBER ff14SB, CHARMM36, and OPLS, to ensure accurate and reliable simulations.
A typical molecular dynamics (MD) simulation involves three key phases: system preparation, simulation execution, and result analysis During system preparation, the system is solvated, parameterized, and neutralized to set up accurate initial conditions The process continues with energy minimization, heating, and equilibration to stabilize the system before the production run The simulation is then performed for a specified duration to generate trajectory data, which is subsequently analyzed to understand the system's dynamic behavior.
Molecular dynamics (MD) plays a crucial role in drug design by addressing the limitations of molecular docking, especially when analyzing flexible targets and weak interactions A primary application of MD is to evaluate the stability of protein–ligand complexes after docking, providing insights into whether the binding pose is maintained under physiological conditions By monitoring key interactions such as hydrogen bonds, hydrophobic interactions, salt bridges, and geometrical fluctuations, MD offers a detailed assessment of complex stability, enhancing the accuracy of drug development processes [9], [94], [95].
Molecular Dynamics (MD) simulations reveal hidden structural states and intermediate conformations essential for understanding protein mechanisms and ligand-binding pathways that are not visible in crystal structures or static models These simulations enable the study of allosteric transitions, active site gating, and the intrinsic motions of proteins, providing deep insights into complex molecular mechanisms crucial for drug discovery and biochemistry research.
Molecular Dynamics (MD) simulations play a crucial role in advancing free energy calculation methods like MM/PBSA, MM/GBSA, and FEP, which deliver more precise binding affinity estimates compared to traditional docking scoring functions [98-100] These techniques are increasingly employed in drug discovery for compound screening, optimizing drug candidates, and assessing the impact of mutations on ligand binding affinity, thereby enhancing the accuracy and reliability of computational predictions in pharmacological research.
Molecular dynamics (MD) simulations are essential in analyzing amino acid variations within active sites and internal interaction networks, which play a crucial role in understanding drug resistance mechanisms By providing detailed insights into these molecular interactions, MD guides the development of mutation-resistant drug derivatives, addressing the growing challenge of drug resistance Additionally, MD is utilized to simulate membrane interactions, further aiding in the design of more effective and resilient therapeutic agents.
13 permeation, evaluate pharmacokinetic properties such as permeability, and assess ligand interactions with lipid bilayers, membrane receptors, or ion channels—challenging yet important drug targets [90]
Molecular Dynamics (MD) simulations offer valuable insights into biomolecular interactions but face significant challenges related to modeling limitations, computational costs, and accuracy A primary limitation is the simulation timescale, as critical biological processes like protein folding and allosteric transitions occur over microseconds to milliseconds, which are often beyond the reach of classical MD simulations typically limited to nanoseconds to microseconds This limitation can result in missing essential conformations or structural states, impacting the accuracy and comprehensiveness of the simulation results.
Current force fields like AMBER, CHARMM, and OPLS rely on classical approximations that overlook quantum mechanical effects, dynamic electronic polarization, and covalent bonding, which are critical for accurately determining binding energies and reaction mechanisms Additionally, modeling complex biological environments—including solvent effects, pH, ions, and cofactors—remains limited compared to in vivo conditions, impacting the overall accuracy of simulations.
To overcome these challenges, several enhanced methodologies are being developed:
- Enhanced sampling techniques: Methods such as Accelerated MD (AMD), Hyperdynamics, Replica Exchange MD (REMD), and Adaptive Sampling help traverse energy barriers and access rare conformational states within practical timescales [106]
- Quantum Mechanics/Molecular Mechanics (QM/MM): This hybrid method allows more accurate simulations of chemical reactions, covalent bonding, and charge transfers within enzyme active sites while maintaining computational feasibility [107]
- Next-generation force fields: Polarizable force fields and machine learning– based models are being developed to improve the accuracy of complex molecular interaction descriptions [107-109]
- Artificial Intelligence (AI): Machine learning models can predict transition-state structures, reduce simulation time, and support analysis of large-scale MD datasets [109]
MATERIALS AND METHODS
Materials and equipments
The three-dimensional (3D) structures of the epidermal growth factor receptor (EGFR) tyrosine kinase domain (EGFR-TK) were obtained from the RCSB Protein Data Bank Specifically, the active state structure is identified by PDB ID 2GS6, while the inactive state structure corresponds to PDB ID 2GS7 These high-resolution crystal structures provide crucial insights into the conformational differences of EGFR-TK, essential for understanding its role in signal transduction and targeted cancer therapies.
The molecular structures of gefitinib and sixteen colossolactone derivatives (I– VIII, A–H) were obtained from the PubChem database [80] in 2D format and were subsequently geometry-optimized via quantum chemical calculations.
High-performance workstation systems powered our simulations and data processing, each featuring Dual Intel® Xeon® Platinum 8180 CPUs with 56 cores and 112 threads for exceptional computational power Equipped with 256GB Samsung DDR4-3200 ECC RDIMM RAM and NVIDIA® GeForce RTX™ 4080 SUPER 16GB GPUs, these systems delivered robust performance for complex molecular simulations and large-scale data analysis Additionally, Samsung 990 PRO NVMeTM M.2 1TB SSDs ensured rapid data access and storage stability, enabling efficient and reliable scientific computing.
Research scope
This study encompasses the following major components
- Data preparation: Retrieval and preprocessing of 3D structures of EGFR and sixteen colossolactone derivatives; geometry optimization and charge assignment via DFT; generation of parameter files for molecular simulations
Molecular docking simulations were conducted by docking ligands into three key functional regions of EGFR, including the active ATP-binding site, inactive ATP-binding site, and allosteric site This approach aimed to identify the optimal binding poses and determine the most preferred binding pockets, providing valuable insights for targeted drug design and EGFR inhibition strategies.
- Molecular Dynamics (MD) simulations: 100-ns MD simulations were performed for selected protein–ligand complexes to assess their structural stability, interaction profiles, and binding energy under near-physiological conditions
- Steered Molecular Dynamics (SMD) simulations: SMD was used to explore ligand dissociation dynamics through force-extension profiles and calculation of rupture force and pulling work
- Umbrella Sampling (US) and PMF calculations: US simulations were used to quantify equilibrium binding free energies and construct PMF profiles
In this study, 2 microseconds of molecular dynamics simulations were performed to investigate the synergistic mechanism of EGFR inhibition The simulations examined EGFR in the presence of both gefitinib and colossolactone H, aiming to explore their potential for allosteric modulation Results suggest that colossolactone H may influence gefitinib–EGFR interactions, providing insights into combined therapeutic strategies and enhancing understanding of allosteric effects in receptor targeting.
Research methods
The 3D structures of the EGFR-TK domain were obtained from the RCSB Protein Data Bank (PDB IDs: 2GS6 for the active conformation and 2GS7 for the inactive conformation) Crystallographic water molecules were removed to prepare the structures for analysis, and missing residues were reconstructed using the MODELLER server to ensure complete and accurate models.
Gefitinib and sixteen colossolactone derivatives were selected from the PubChem database [80], including colossolactones I to VIII and A to H Their 2D molecular structures are presented in Figure 2.1 and Table 1.1, respectively
The atomic charges were assigned using the Restrained Electrostatic Potential (RESP) method, based on Density Functional Theory (DFT) calculations performed with Gaussian 16 at the B3LYP/6-31+G(d,p) level Subsequently, AMBER Tools' antechamber module was utilized to extract essential parameters and generate ligand topology files, ensuring accurate parameterization for molecular simulations.
Frontier molecular orbitals, including the Highest Occupied Molecular Orbital (HOMO) and the Lowest Unoccupied Molecular Orbital (LUMO), were also computed and are presented in Appendix 1
Protein and ligand preparations for docking were conducted using AutoDockTools (ADT), included in the MGLTools suite [152] Molecular docking
20 simulations were carried out using AutoDock Vina version 1.2.0 [153] to evaluate possible binding poses and affinities
Docking grid boxes were configured at dimensions of 80 × 80 × 80 Å, centered on key binding sites such as the active ATP-binding site (-54.303, 6.101, -26.596), the inactive ATP-binding site (30.334, 51.703, 44.457), and the allosteric pocket (-55.131, 2.709, -7.326) The docking parameters were optimized with an exhaustiveness level of 200 and a grid spacing of 0.375 Å to ensure accurate and comprehensive molecular interactions.
Binding poses of colossolactone H at the three EGFR sites identified by docking simulations are shown in Figure 2.2
Figure 2.2 Binding poses of colossolactone H in EGFR and corresponding pulling directions used in SMD simulations
Colossolactone H, depicted as a rainbow-colored stick model, was docked into three distinct EGFR binding pockets, providing insights into its potential inhibitory mechanisms The pulling directions for single-molecule force spectroscopy (SMF) simulations, aligned along the z-axis based on CAVER analysis, illustrate the ligand dissociation pathways from these sites Specifically, the ATP-binding site in its active conformation (cyan cartoon) differs structurally from the inactive state's ATP-binding site (wheat-colored cartoon), while the allosteric site (green cartoon) offers an alternative target in the inactive conformation The characteristic α-C helix, highlighted in red across all models, emphasizes the structural differences between the active and inactive EGFR states, contributing to understanding ligand binding dynamics and conformation-specific interactions.
To investigate the structural dynamics of EGFR–ligand complexes, molecular simulations were conducted using GROMACS version 2023.1 Protein parameters were assigned with the Amber99SB-ILDN force field, and the TIP3P water model was used to accurately simulate solvation Each complex was solvated in a rectangular periodic box measuring 10.0 × 10.0 × 10.0 nm, containing over 96,000 atoms, including water molecules and counterions, to ensure a realistic biological environment.
Energy minimization was performed until the maximum atomic force was below
The molecular dynamics simulations began with system equilibration through two 100-ps phases: first under constant particle number, volume, and temperature (NVT), then under constant particle number, pressure, and temperature (NPT) Following this, 100-nanosecond production MD simulations were performed at 300 K and 1 atm to ensure accurate and reliable data collection The simulations were conducted with an energy cutoff of 100 kJ/mol/nm, enabling precise modeling of molecular interactions under physiologically relevant conditions.
SMD was utilized as a non-equilibrium technique to assess ligand binding affinities An external force was applied to the ligand's center of mass, pulling at a constant velocity of 1 m/s with a spring constant of 600 kJ/mol/nm The pulling was directed along the z-axis of the simulation box, ensuring consistent force application, with CAVER [157] used to align the pulling direction accurately.
In our simulations, Cα atoms of the protein were restrained with harmonic potentials at a force constant of 100 kJ/mol/nm to maintain structural stability Force and displacement data were meticulously recorded every 10 femtoseconds to capture detailed molecular interactions For each protein-ligand complex, we conducted 50 independent Steered Molecular Dynamics (SMD) trajectories, each spanning 3 nanoseconds, utilizing the GROMACS software to ensure robust and reliable simulation results.
The pulling directions used for colossolactone H in the SMD simulations are visually illustrated in Figure 2.2
2.3.5 Umbrella Sampling and Potential of Mean Force calculation
Umbrella Sampling (US) was used to accurately determine the equilibrium free energy of ligand dissociation Over 120 US windows were generated from an initial Steered Molecular Dynamics (SMD) trajectory performed at a slow pulling speed of 0.1 m/s These windows were spaced at 0.025 nm intervals to ensure comprehensive sampling along the dissociation pathway.
Before each 10-nanosecond production MD simulation, standard NVT and NPT equilibration steps were performed to ensure system stability The free energy profiles, or PMF curves, were then generated using GROMACS' integrated analysis tools for accurate insights into the system's energetic landscape.
RESULTS AND DISCUSSION
Evaluation of interactions and stability of 16 colossolactone derivatives
3.1.1 Molecular docking of colossolactones into the ATP-binding site and the allosteric pocket of EGFR
The active and inactive states of EGFR exhibit distinct structural characteristics
In the active state of EGFR, only the ATP-binding site is accessible to small molecules, whereas in the inactive state, both the ATP-binding site and the allosteric pocket can accommodate ligands All 16 colossolactone derivatives listed in Table 1.1 were subjected to molecular docking simulations at these three binding sites The ligand preparation process and docking results are detailed in Appendices 1 and 2 Analysis of the docking poses revealed key interactions between each colossolactone and amino acid residues in both the ATP-binding site and the allosteric pocket across the active and inactive EGFR conformations These interactions predominantly involved hydrogen bonds and non-bonding contacts, providing insights into the binding mechanisms of colossolactones with EGFR.
Binding energies for each colossolactone were determined through docking studies, utilizing initial docking configurations and documented in Tables 3.1–3.3 The lowest-energy conformation of each complex was selected to facilitate detailed analysis, ensuring accurate assessment of binding affinity and interaction stability. -**Sponsor**Looking to refine your article and boost its SEO? Let [Novakid Global ARABIC](https://pollinations.ai/redirect-nexad/oHn46lsH) help you craft compelling content For instance, to convey the essence of your paragraph on binding energies, focus on sentences highlighting the key results: "Binding energies determined by docking were reported" and "The lowest-energy conformation of each complex was extracted for further in-depth analysis." These sentences capture the core meaning and allow for coherent paragraph development following SEO rules.
Despite most colossolactone derivatives binding effectively to the ATP-binding site of active EGFR with hydrogen bonds, compounds III, C, D, and E did not form such interactions Notably, colossolactones II, IV, VII, and VIII established hydrogen bonds with key residues Asp855, Ser720, Cys797, and Thr854, while colossolactone A exhibited the highest hydrogen bonding capacity, interacting with Leu718, Glu762, Phe795, Thr854, and Asp855 The residues Thr854 and Asp855 were most frequently involved in hydrogen bonds, with interactions mediated via oxygen and nitrogen atoms, respectively, and bond lengths ranging from 2.77 to 3.34 Å Non-bonding interactions identified using LigPlot+ ranged from 2.90 to 3.90 Å, with residues such as Leu718, Phe723, Val726, Ala743, Lys745, Met766, Thr790, Gly796, Cys797, Asp800, and Leu844 consistently involved in non-bonded contacts with the colossolactones, highlighting key regions critical for ligand binding stability.
Compounds I, III, VII, and F demonstrated strong binding affinities with docking-derived energies below –9.0 kcal/mol, indicating their potential as effective ligands In contrast, colossolactone VI showed a less favorable binding energy of –6.6 kcal/mol Although colossolactone A formed the highest number of hydrogen bonds, its docking score (~–7.6 kcal/mol) was less favorable compared to colossolactones III (–9.4 kcal/mol) and C (specific value not provided), highlighting the importance of overall binding energy over hydrogen bonding alone for effective ligand-receptor interactions.
8.1 kcal/mol), D (E = –8.7 kcal/mol), and E (E = –8.1 kcal/mol), which did not form any hydrogen bonds
Recent studies indicate that the number of hydrogen bonds does not directly correlate with the calculated binding energy in molecular docking This observation also extends to non-bonding interactions, emphasizing that binding strength is influenced by multiple factors beyond merely the quantity of hydrogen bonds Understanding these nuances is crucial for accurate prediction of molecular affinities and designing effective ligands.
Molecular docking into the ATP-binding site of inactive EGFR (Table 3.2 and
Appendix 2) showed binding energies ranging from –6.8 to –10.4 kcal/mol
Colossolactones I, II, III, D, E, and G had docking scores below –10 kcal/mol, whereas
V, VI, and VII had higher values (–6.8 to –7.7 kcal/mol) Except for colossolactones III,
Several complexes, including IV, VIII, B, C, D, and F, did not form hydrogen bonds, while the remaining complexes were stabilized through hydrogen bonding with bond lengths ranging from 2.74 to 3.28 Å Notably, colossolactones V and A formed the highest number of hydrogen bonds, with four each, similar to the active state However, the number of hydrogen bonds did not directly correlate with the lowest docking energies, indicating that hydrogen bond count is not a sole determinant of complex stability.
Docking results for the allosteric pocket show binding energies ranging from –6.9 to –9.8 kcal/mol, indicating potential strong interactions Compared to the ATP site, fewer hydrogen bonds were observed in the allosteric pocket, with most ligands forming only one hydrogen bond, except for colossolactones II, IV, D (two hydrogen bonds), and A (three hydrogen bonds) Notably, colossolactolones I and H did not form any hydrogen bonds, with Phe856 being the most common hydrogen bond donor, followed by Lys745 and Thr790 All colossolactones interacted with key residues such as Lys745, Leu747, Ile759, Glu762, Met766, Leu777, Leu788, Thr854, Asp855, and Leu858, suggesting these compounds may effectively target the EGFR allosteric pocket through strong and specific interactions These findings support the potential of colossolactones as allosteric EGFR inhibitors, highlighting their promise for drug development.
Understanding receptor dynamics is essential for analyzing protein–ligand interactions, yet conventional docking often overlooks receptor flexibility by keeping structures fixed to simplify ligand binding site exploration Despite this limitation, molecular docking provides a rapid and effective initial step for generating ligand–receptor complexes for further study In this research, 100 ns molecular dynamics simulations were performed on each of the 16 colossolactone–EGFR-TK complexes at three key binding sites: the ATP-binding site (in both active and inactive conformations) and the allosteric pocket, based on docking results The simulation data are summarized in Tables 3.1, 3.2, and 3.3.
Table 3.1 MD results after 100 ns for 16 colossolactone derivatives in the ATP-binding site (active conformation)
The table presents docking-derived binding energies along with average values from the last 20 nanoseconds of molecular dynamics simulations, including the distance to T790 (Míndist - nm), total contacts (Numcount), and hydrogen bonds (H-bond) It also details Coulombic and van der Waals (vdW) energies (kcal/mol), solvent-accessible surface area (SASA, nm²), and the interaction energy (IE), providing comprehensive insights into ligand-protein interactions.
All 16 colossolactone derivatives maintained stable conformations within the ATP-binding site of active EGFR after 100 ns of simulation
Colossolactones in the ATP-biding site (active state)
No Colossolactone Docking score (kcal/mol)
Numcount H-bond Coulombic potential (kcal/mol)
Van der Waals Interactions (kcal/mol)
Table 3.2 MD results after 100 ns for 16 colossolactone derivatives in the ATP-binding site (inactive conformation)
The table summarizes key docking and molecular dynamics (MD) data, including docking-derived binding energies and average values from the last 20 nanoseconds of simulation It features the distance to T790 (Míndist in nm), total contacts (Numcount), and hydrogen bonds (H-bond), providing insights into ligand-receptor interactions Additionally, the table reports Coulombic and van der Waals (vdW) energies (kcal/mol), solvent-accessible surface area (SASA in nm²), and interaction energy (IE), offering a comprehensive view of the binding stability and conformational dynamics These parameters collectively help evaluate the strength and nature of ligand binding within the target site.
All 16 colossolactone derivatives maintained stable conformations within the ATP-binding site of inactive EGFR after 100 ns of simulation
Colossolactones in the ATP-biding site (inactive state)
No Colossolactone Docking score (kcal/mol)
Numcount H-bond Coulombic potential (kcal/mol)
Van der Waals Interactions (kcal/mol)
Table 3.3 MD results after 100 ns for 16 colossolactone derivatives in the allosteric pocket (inactive conformation)
The table presents docking-derived binding energies and average values from the final 20 ns of molecular dynamics simulations, including the distance to T790 (Míndist in nm), total contacts (Numcount), and hydrogen bonds (H-bond) It also summarizes Coulombic and van der Waals energies (kcal/mol), solvent-accessible surface area (SASA in nm²), and overall interaction energy (IE), providing comprehensive insights into ligand-protein interactions crucial for drug design and molecular recognition.
All 16 colossolactone derivatives maintained stable conformations within the allosteric pocket of inactive EGFR after 100 ns of simulation
Colossolactones in the allosteric pocket (inactive state)
No Colossolactone Docking score (kcal/mol)
Numcount H-bond Coulombic potential (kcal/mol)
Van der Waals Interactions (kcal/mol)
The results showed that all ligands remained stably bound in their respective binding sites, as indicated by consistently short average distances to key residues (see
Table S3, Appendix 3) In the active ATP-binding site, colossolactones formed over
Analysis of EGFR-TK domain interactions reveals that key compounds engage through multiple hydrogen bonds, averaging up to three per interaction Notably, in the inactive state, Colossolactone E formed 944 interactions, while other compounds exhibited over 1600 interactions, indicating higher binding affinity Additionally, Colossolactone C did not form any hydrogen bonds, unlike compounds I, IV, V, VII, VIII, A, B, D, and H, which established one or more hydrogen bonds, highlighting differences in their binding profiles.
The allosteric pocket exhibited over 3,000 interactions, indicating strong binding affinity Hydrogen bonds were generally limited, rarely exceeding one bond per interaction, except for colossolactones V and A, which averaged 1.25 and 2.15 hydrogen bonds, respectively Additionally, the solvent-accessible surface area (SASA) values for the complexes ranged from 7.04 to 9.37 nm² across all three binding sites, highlighting variations in complex exposure and interaction stability.
Our study analyzed Coulombic and van der Waals (vdW) energies from molecular dynamics simulations, both of which were consistently negative, indicating favorable ligand–receptor interactions A strong correlation was observed between these interaction energies and the number of total interactions and hydrogen bonds, underscoring their significance in binding stability Specifically, Pearson correlation coefficients between total interactions and vdW energy were 0.93 for the active ATP-binding site, 0.98 for the inactive ATP-binding site, and 0.94 for the allosteric pocket, highlighting the strong relationship between interaction strength and binding affinity in these key protein regions.
Additionally, distances between colossolactones and critical residues in the active ATP-binding site were all under 1.00 nm, except for colossolactone C, which had a distance of 1.04 nm from Leu859
Detailed distance data for each colossolactone in the active ATP site are presented in Table S3.1 (Appendix 3) Gly721, Lys745, Thr790, Gln791, Leu792, Met793,
Cys797 and Arg841 consistently stayed within 0.50 nm across all compounds, indicating their stable interaction sites Ala722 was notably close (