Abstract
BRD4 (“Bromodomain-containing protein 4”), a recognized gene regulator, is an attractive target for therapeutic development, particularly for the management of neuroblastoma. An integrated pharmacoinformatic strategy for the development of new BRD4 inhibitors is examined in this research. Pharmacophores were used to digitally screen five databases, and the current study aims to determine the best binding modes by docking the screened hits to the BRD4 active site. Using the BRD4 protein co-crystal ligand (73B) (PDB ID: 4BJX) as a template, pharmacophore hypotheses were produced. Five databases were subjected to a pharmacophore-based virtual screening process, and 1089 hits that satisfied the screening requirements were selected for docking against the BRD4 receptor by using the SP module of the Glide tool. The top ten docked compounds with the highest binding affinities, ranging from − 9.623 to − 8.894 kcal/mol, were selected. Further, the biological activity and ADMET analysis revealed that the selected compounds have values that fall in the acceptable range. The protein-ligand complexes’ stability was verified by performing molecular dynamics (MD) simulations of the binding positions of the top two compounds against the BRD4 receptor. The stability and binding free energies of the compounds indicate that these compounds may function as lead compounds to affect the biological activity of BRD4 in the in vitro studies.
Introduction
Neuroblastoma is a cancer that arises from immature nerve cells in various parts of the body. It is mostly found in and around the adrenal glands, which are located above the kidneys. This cancer commonly affects children, particularly those under the age of five. The aggressiveness of neuroblastoma varies greatly, ranging from slow-growing tumors that may disappear on their own to highly aggressive forms that spread rapidly to other parts of the body1,2.
Bromodomain-containing protein 4 (BRD4) modifies chromatin structure to regulate gene expression. BRD4 belongs to the BET family. One of the most significant BET proteins is the epigenetic reader BRD4, which is implicated in angiogenesis and the emergence of cardiovascular, inflammatory, and central nervous system disorders as well as cancer, notably neuroblastoma3,4,5,6,7.
Like all BRD4 members, BD1 and BD2, two highly conserved N-terminal bromodomains (BDs), are present in BRD4. These BDs have the ability to detect acetylated lysine residues and non-histone proteins, which control transcriptionally gene expression, cell division, apoptosis, and cell cycle progression8,9. BD1 and BD2 sequences are extremely homologous and conserved. On the other hand, every BRD4 BD helps with protein-specific binding. The two BRD4 BDs may interact with distinct proteins, suggesting that they play distinct biological roles. Compared to BD2, BRD4 BD1 exhibits a greater affinity for the tetra-acetylated histone H4 peptide. Diacetylated histone H3 is recognized by BRD4 BD2, which then binds to it and recruits non-histone proteins.
BRD4 is regarded as a potentially effective therapeutic target for several human diseases10,11,12,13,14,15,16. According to research, BRD4 is frequently dysregulated in cancers such as neuroblastoma. Dysregulation of BRD4 can lead to abnormal cell growth and proliferation, making it an appealing target for cancer treatment. Therefore, BRD4 inhibitors must be found and developed, and this effort is receiving more and more attention2.
Clinical trials involving humans are presently investigating a number of BET inhibitors10. An important class of these tiny molecule inhibitors includes azaepine analogues, such (+)-JQ1. The pan-BET family member inhibitor (+)-JQ1, one of the first to be discovered, possessed IC50 values in the nanomolar range17. As a result, the pharmacological tool (+)-JQ1 has been extensively employed to explore the biological roles of the BET family in several disorders, including tumor angiogenesis, osteosarcoma, leukemia, and gallbladder cancer18,19. Regretfully, the limited selectivity of BET inhibitors within the BET family has limited their applications. According to research, BET family BDs represent exciting new therapeutic targets for the management of obesity, inflammation, cancer, and neurological disorders20,21. Thus, several distinct BET inhibitors, including MS436, RVX-208, Olinone, and MS765, have been created to target the BDs of BET family members, particularly. For example, RVX-208 and MS765 are less selective for BRD4 than other members of the BRD family, but they bind to BD2 of BET proteins more readily than BD1 of BET proteins22,23. Thus, it is essential to develop powerful and specific BRD4 BD1- and BRD4 BD2-selective compounds with lower affinity for other members of the BET family in order to get a better understanding of the distinct functions and biological activities of each BRD4 BD in a variety of human illnesses.
The purpose of our study is to predict novel BRD4 inhibitors via high-throughput virtual screening, molecular docking, ADMET analysis, and simulations, followed by the evaluation of their potential as therapeutic candidates based on stability and binding energy analysis. Molecular docking is a well-established computational technique used to predict the preferred orientation of a ligand as it binds to a protein receptor, offering insight into binding affinity and specificity24. Moreover, Molecular dynamics (MD) simulations assess the stability of the docked complexes under physiological conditions. When combined with ADMET analysis, which assesses the drug likeness, bioavailability, and toxicity, this approach provides a robust framework for in silico drug discovery25,26. Recent studies have successfully applied this combination to identify novel therapeutic candidates for various diseases, including cancer and infectious disorders27,28,29. These methods are particularly valuable in early-phase screening, where experimental validation is limited by time and resources.
The field of drug discovery has undergone a paradigm shift in recent years, with computational methods being crucial in the identification of new therapeutic agents30,31,32,33. This work proposes an integrative pharmacoinformatic approach to identify novel BRD4 inhibitors with potential relevance to neuroblastoma therapy. By combining pharmacophore-based virtual screening, molecular docking, ADMET prediction, and molecular dynamics (MD) simulations, we aim to prioritize compounds with favorable binding, pharmacokinetic, and stability profiles for further preclinical evaluation27,34.
Methodology
pharmacophore modelling
A chemical template containing the fundamental structural components of physiologically active molecules is known as a pharmacophore model. A pharmacophore model is generated based on the structural characteristics of an active drug and is subsequently processed to perform chemical database screening35. The pharmacophore model was created using the chemical structure of the co-crystal ligand and the interactions between the 73B ligand and the BRD4 protein’s (PDB ID: 4BJX) binding pocket via Pharmit web server36,37. The server provides a method for screening the chemical databases based on the derived pharmacophoric qualities.
Pharmacophore-based virtual screening
The virtual screening model was constructed using four pharmacophoric properties of the co-crystal ligand i.e., acceptor, hydrogen bond donor, hydrophobic, and aromatic. The parameters of the virtual screening were determined by applying Lipinski’s rule38: hydrogen bond acceptor (HBA) < 10, molecular weight < 500, hydrogen bond donor (HBD) < 5, and logP < 5. The following databases were examined for virtual screening: CHEMBL, Enamine, MCULE, Chemspace, and ChemDiv, with previously reported anticancer activity compounds and a broad range of scaffolds to increase the chances of novel hits.
Preparation of hit’s pool for docking purposes
The Schrödinger Maestro LigPrep tool was used to optimize, minimize, and prepare the 1089 hits39. Conformers were constructed to optimize the shape of each ligand. By adjusting the ligands’ geometry using the OPLS_2005 forcefield, the conformations that are energetically advantageous were guaranteed. The popular scientific forcefield for determining and visualizing the energy of atom-to-atom interactions in molecules is the OPLS_2005 forcefield40. Protonation states were generated at pH 7 using Epik. All unfavorable interactions and stretched geometries were eliminated by reducing the compounds’ energy using the conjugate gradient algorithm. The optimized hits with correct ionization states, proper stereochemistry, and minimized geometries were subjected to docking studies.
Structure-based molecular docking
Molecular docking was performed on the prepared hits against the BRD4 receptor. The BRD4 protein’s crystal structure (PDB ID: 4BJX) was obtained from the PDB database with 1.59 Å resolution, which is the most suitable experimental model for the target protein. The Protein Preparation Wizard was used to prepare the receptor for docking41. There were several stages to the receptor’s preparation process. Disulfide bonds were created, metal bonds with zero order were allocated, and bond orders were set. Polar hydrogens were also introduced into the protein structure. Any extra ligands or water molecules were also removed from the crystal structure, and the PROPKA program was used to calculate ionizable groups and the pKa values. At pH 7.0, the hydrogen bonding in the protein was optimized. Finally, the protein structure’s energy was reduced by applying the OPLS_2005 force field42. Following protein preparation, a 3D grid was constructed at X, Y, Z coordinates 14.09, 0.72, 9.68 for site-specific docking. The ligands were docked in SP (Standard Precision) mode using the Glide docking module, and 09 poses were generated for each compound43. The compounds were examined and selected based on Glide scores.
Biological activity and drug-likeness analysis
The biological activity, bioavailability scores, and druglike properties of selected hits were predicted via Molinspiration web server (https://www.molinspiration.com/cgi/properties). SwissAdme44 and DataWarrior software45 respectively. These parameters help to choose which compound can be used for further analysis to get potential leads.
ADME and toxicity profile analysis
We looked deeper into the ADMET (absorption, distribution, metabolism, excretion, and toxicity) characteristics of the selected docked ligands. These characteristics were forecasted with the QikProp tool from Maestro46. Several attributes are estimated by the QikProp tool based on the ligands’ molecular structures. Hydrogen bond acceptors, molecular weight, hydrogen bond donors, QPlogPo/w, QPlogHERG, QPPCaco, QPlogBB, and QPlogKhsa are typical expected features. The logarithm of the octanol-water partition coefficient, which gauges the hydrophobicity and membrane permeability of a compound, is anticipated by QPlogPo/w. QPlogHERG evaluates the likelihood that a ligand will block the hERG potassium channel, which has been associated with potential cardiac toxicity. Using the monolayer of Caco-2 cells as a model for intestinal absorption, QPPCaco calculates a compound’s permeability through the cells. QPlogBB can be used to predict the partition coefficient of the blood-brain barrier, which measures a substance’s ability to penetrate the blood-brain barrier and reach the central nervous system. A vital protein that affects drug distribution and binding in the bloodstream is human serum albumin. The logarithm of this protein’s binding affinity is computed by QPlogKhsa.
Binding free energy profile of complexes
A computational approach, MM-GBSA, was carried out to investigate binding free energy (BFE) of docked ligands in contact with the targeted receptor47. Their efficient binding helps to determine the potential of compounds, i.e., do they function as druglike molecules? For this, MMGBSA module of Prime was employed to estimate the BFE of all protein-ligand complexes. The BFE profile of complexes is a combination of specific components i.e., Coulombic, H-bond, covalent, van der Waals, lipophilic, and solvation interactions. MMGBSA values of complexes were calculated via the Prime tool with the following equations48.
Where ΔGbind refers to the total binding free energy of complexes in kcal/mol, which is a combination of molecular mechanics gas phase energy (ΔEgas), and solvation free energy (ΔGsol) as displayed in Eq. (2).
MD simulation study
For the two best compounds, 200 ns Molecular Dynamics simulations were performed using Desmond49. Molecular Dynamics simulations were performed to evaluate the ligand and protein complexes’ stability. The stability of complexes was assessed through MD simulations, following a series of steps, including optimization, preprocessing, and minimization. The minimizing procedure was carried out using the OPLS_2005 force field40. The compounds underwent solubility in a 10 Å periodic box that contained water molecules of TIP3P50. 0.15 M NaCl salt and counter ions were added as needed to neutralize the systems and replicate physiological conditions. The temperature and pressure values for the NPT ensemble were 300 K and 1 atm, respectively. Prior to the simulation initiation, the systems underwent a relaxation phase. Trajectories were recorded and saved at 40 ps intervals during the simulation, enabling subsequent analysis of the results.
Principal component analysis of potential hits
The principal component analysis (PCA) is a statistical technique utilized to analyze collective and correlated molecular movements of studied complexes. The motions of biological systems were computed and represented in the form of eigenvalues, eigenvectors, and covariance matrices, utilizing proteins’ backbone atoms. These compact motions allow us to investigate conformational changes of the protein in contact with ligand. For this, covariance matrices were formed based on atomic coordinate data of carbon alpha atoms and utilized to obtain corresponding eigenvectors and eigenvalues51.
Results
Pharmacophore modelling and virtual screening
The pharmacophore query model was created using the pharmacophoric characteristics of the 73B ligand involved in the molecular interactions with the BRD4 protein. Five features (02 hydrophobic, 02 aromatic, and 01 hydrogen bond acceptor) were used to generate the query model (Fig. 1). The X, Y, and Z coordinates of the features are shown in Table 1.
The pharmacophore query model generated by the Pharmit server. Green spheres show the hydrophobic group, purple shows the aromatic ring, while the orange sphere shows the hydrogen bond acceptor.
Based on these features, pharmacophore-based virtual screening of five databases was conducted, and the hits meeting the screening criteria were selected (Table 2). A total of 1089 hits were collectively obtained from all screened databases. Among these hits, the MCULE database produced the highest number of hits.
Structure-based molecular docking
The 1089 hit compounds were selected after virtual screening and docked to the BRD4 receptor to predict the binding affinities by using the standard precision mode of the Glide tool. Based on the binding affinities, the top ten compounds were selected for further analysis. The chosen compounds’ binding affinities fell within − 9.623 to -8.894 kcal/mol range. The binding affinities for the selected compounds indicated that these have the probability of inhibiting the function of BRD4. The BIOVIA Discovery Studio tool was utilized to analyze the molecular interactions between the chosen hits and the binding pocket of the BRD4 receptor. The observed interactions mainly involved: hydrogen bonds, van der Waal interactions, Pi-Pi Stacked, Pi-Sulfur, Pi-Sigma, and Alkyl interactions. The binding affinities and docking scores of each of the best candidate compounds are largely determined by these interactions. Notably, the overall strength of the resulting complex is significantly influenced by the creation of hydrogen bonds between the ligand and the amino acid inside the active regions of molecules. Consequently, these interactions consistently enhance the docking results52.
Conformational exploration of binding poses
MCULE-3,895,595,807 formed four hydrogen bonds with Gln85, Pro82, Val87, Asn140, and seven alkyl interactions with Leu94, Tyr139, Leu92, Met132, Phe83, Cys136, Ile146 residues (Fig. 2a). MCULE-8,812,795,806 formed five hydrogen bonds with Tyr97, Asn140, Asp88, Val87, Cys136, and five alkyl interactions with Ile146, Phe83, Pro82, Leu92, Lys91 (Fig. 2b). CSC101378549 formed four hydrogen bonds with Pro82, Gln85, Tyr97, Cys136 and the three alkyl interactions with the Leu92, Ile146, Val87 (Fig. 2c). MCULE-6,216,741,487 formed three hydrogen bonds with Gln85, Tyr97, Cys136, one Pi-Sigma interaction with Pro82, and five alkyl interactions with Lys91, Leu92, Trp81, Pro82, Val87, Ile146 (Fig. 2d). CSC101661920 formed one hydrogen bond with Asn140, one Pi-Sigma interaction with Pro82, and nine alkyl interactions with Ile146, Trp81, Val87, Phe83, Cys136, Tyr97, Tyr139, Leu92, Leu94 (Fig. 2e). MCULE-3,872,683,406 formed four hydrogen bonds with Pro82, Asn140, Ile146, Asp145, and four alkyl interactions with Leu92, Val87, Phe83, Cys136 (Fig. 2f). MCULE-2,532,030,345 formed four hydrogen bonds with Gln85, Tyr97, Asn140, Cys136, and five alkyl interactions with Phe83, Val87, Ile146, Leu92, Pro82 (Fig. 2g). CHEMBL4216188 formed two hydrogen bonds with Lys91, Asn149, and six alkyl interactions with Ile46, Pro82, Val87, Leu92, Tyr139, Leu194 (Fig. 2h). MCULE-6,796,614,739 formed five hydrogen bonds with Gln85, Pro82, Asn140, Ile146, Asp145, and four alkyl interactions with Val87, Phe83, Cys136, Leu92 (Fig. 2i). MCULE-1,485,100,159 formed seven hydrogen bonds with Lys91, Asp88, Val87, Met105, Met132, Cys136, Pro82, and three alkyl interactions with Leu92, Ile146, Phe83 (Fig. 2j). The docked compounds’ binding modes are shown in Fig. 3. The docking scores, the hydrophobic interaction, and hydrogen bond lengths are shown in Table 3. Upon comparison with the known pharmaceutical agents, such as RVX-208 and MS436, which depict similar H-bond interactions with Asn140, Asn135, and Pro82 residues. Similar interacting residues in the active site exhibit analogous binding strengths. Besides this, the selected compounds show lower binding energy than known pharmaceutical agents, which indicates stronger protein-ligand interactions.
The molecular interactions of the compounds. (a) MCULE-3,895,595,807, (b) MCULE-8,812,795,806, (c) CSC101378549, (d) MCULE-6,216,741,487, (e) CSC101661920, (f) MCULE-3,872,683,406, (g) MCULE-2,532,030,345, (h) CHEMBL4216188, (i) MCULE-6,796,614,739, (j) MCULE-1,485,100,159. Green spheres show hydrogen bonds, purple shows Pi-Sigma interaction, and magenta shows alkyl interactions.
The Binding modes of the docked compounds. (a) MCULE-3,895,595,807, (b) MCULE-8,812,795,806, (c) CSC101378549, (d) MCULE-6,216,741,487, (e) CSC101661920, (f) MCULE-3,872,683,406, (g) MCULE-2,532,030,345, (h) CHEMBL4216188, (i) MCULE-6,796,614,739, (j) MCULE-1,485,100,159.
Prediction of biological activity, availability, and drug-likeness of top ten hits
Biological activity of top hits was determined against human receptors that can act as drug targets. In Table 4, bioactivity scores of chemicals were calculated for certain parameters i.e., GPCR ligand, enzyme inhibition, kinase, and protease inhibition. The biological activity scores considered significant above 0.0, moderate ranged between − 5 and 0.0, while below − 5 were considered inactive53. The top 10 chemicals’ bioactivity scores, ranging from 0.68 to − 0.48, clearly reveal that they have medicinal potential. MCULE-2,532,030,345 and CSC101378549 hits displayed significant results within the active range, depicting a low risk of undesirable behavior i.e., mutagenicity or poor absorption, and implied potential drug-like behavior54.
All hits lie within the range of drug-likeness (− 10 to + 10) criteria, while MCULE-3,895,595,807, MCULE-8,812,795,806, CSC101661920, MCULE-2,532,030,345, and CHEMBL4216188 hits are considered as orally active drugs compared to others. They have lower chances of rejection in the drug development process and possess better pharmacokinetic properties. All compounds possess 0.55 bioavailability scores, have significant absorption, and are preferred for oral delivery55. Some of the compounds depict mutagenicity, tumorigenicity, and irritant effects, while none show reproductive effects, as shown in Table 5.
ADME and toxicological properties of top hits
The preferred compounds’ ADMET (absorption, distribution, metabolism, excretion, and toxicity) characteristics were examined. The compounds’ ADMET characteristics range within permissible bounds. “QPlogPo/w” (− 2.0 to 6.5), “QPlogHERG” (<− 5), “QPlogBB” (− 3.0 to 1.2), “QPPCaco” (< 25 poor, > 500 great), and “QPlogKhsa” (− 1.5 to 1.5) were the cutoff values for the ADMET parameters. The Compounds’ ADMET and physicochemical properties are shown in Table 6. Upon analysis of the ADMET properties, it was observed that only CSC101378549 and MCULE-2,532,030,345 compounds met the selection criteria, so these two compounds were subjected to binding pose analysis and MD simulations for stability analysis.
“QPlogHERG” (<− 5), “QPlogPo/w” (− 2.0 to 6.5), “QPlogBB” (− 3.0 to 1.2), “QPPCaco” (< 25 poor, > 500 great), and “QPlogKhsa” (− 1.5 to 1.5).
Binding free energy calculations of docked complexes
The binding energy of ligands to protein molecules is commonly measured using the MM/GBSA technique56. The BRD4 complex’s binding free energy and the impact of extra non-bonded interaction energies were assessed (Table 7). The binding free energy of MCULE-2,532,030,345 was − 86.300 kcal/mol, whereas that of CSC101378549 to BRD4 was − 82.851 kcal/mol, indicating stronger binders than others. \(\:\varDelta\:\)Gbind is governed by non-bonded interactions such as \(\:\varDelta\:\)Gbind Coulomb, \(\:\varDelta\:\)Gbind Packing, \(\:\varDelta\:\)Gbind Hbond, \(\:\varDelta\:\)GbindLipo, and \(\:\varDelta\:\)GbindvdW. The average binding energy was most affected by the \(\:\varDelta\:\)GbindLipo, \(\:\varDelta\:\)GbindvdW, and \(\:\varDelta\:\)GbindCoulomb energies among all interaction types, representing polar interactions and a better hydrophobic fit. The hydrophobicity of compounds is due to the nonpolar solvation interactions, which depict good binding of chemicals with the targeted receptor32,33. Conversely, the final average binding energies were least affected by the \(\:\varDelta\:\)GbindSolv_GB, \(\:\varDelta\:\)GbindCovalent, and strain energies, demonstrating less conformational distortion. Furthermore, BRD4 complexes showed stable hydrogen bonds with amino acid residues based on their GbindHbond interaction values. Thus, the binding energy obtained from the docking data was well-supported by the MM-GBSA estimations57.
Binding pose analysis of potent complexes
The two chemicals chosen for further examination were positioned onto the co-crystal ligand to explore their potential binding mechanisms after examining their molecular interactions. The analysis revealed that the docked compounds were positioned onto the co-crystal ligand in the same active site and showed a similar binding mode (Fig. 4). Thus, the possible binding mechanisms of the selected compounds were subjected to stability analysis by employing the MD Simulation study.
The plausible binding modes of the best two compounds aligned on the co-crystal ligand (cyan sticks). (A) CSC101378549 (magenta sticks), (B) MCULE-2,532,030,345 (yellow sticks).
Structure stability analysis of potential complexes
RMSD
The stability of the chosen compounds against the BRD4 receptor was verified using 200 ns molecular dynamics (MD) simulations. The root mean square deviation (RMSD) of the carbon alpha atoms was calculated to investigate the complexes’ overall structural variations and alterations over the simulation58. The RMSD values for the CSC101378549 progressively elevated to ~ 3.2 Å till 25 ns and then showed a decrease in RMSD value of about 2.1 Å up to 40 ns. After 40 ns, it increased again and reached ~ 3.0 Å. Following that, it stabilized in this range until the simulation’s end, while the ligand’s RMSD completely lined up with the protein’s RMSD. (Fig. 5a). However, the RMSD of MCULE-2,532,030,345 was lower than the RMSD of the protein till half of the simulation time and depicts more fluctuations than CSC101378549 within the acceptable range. It started at a range of ~ 1.6 Å at 5 ns, and stayed there until 75 ns with minor escalation, which increased to 2.0 Å. However, at 80 ns, it reached stability at 2.8 Å towards the end of the simulation (Fig. 5b). After equilibrium, no clear fluctuations were observed, which indicates that both compounds remained bound in the active site of the targeted receptor.
The RMSD of BRD4 complexes calculated during the 200 ns simulation. (A) CSC101378549, (B) MCULE-2,532,030,345.
RMSF
Root mean square fluctuations (RMSF) values were computed in order to study the dynamic behavior of the proteins while they were bound to the ligands59. For each protein residue over the simulation period, RMSF values give detailed information on the residue’s mobility and flexibility. Based on the expected RMSF values, most protein residues changed very slightly during the simulation, which was less than 2Å. This suggests that these residues maintained relative stability and stiffness while the ligands were present. However, the RMSF values were higher in the protein’s loop regions, which were around 4Å60. According to the RMSF analysis, most residues in the protein-ligand complexes held onto their rigid forms, causing the complex to stay stable. RMSF values of loop regions were higher, indicating that these areas displayed larger variations than others and could have engaged in dynamic interactions with the ligands (Fig. 6). Most protein residues showed relatively slight changes, but loop parts showed significantly higher levels of flexibility (Fig. 7). The blue regions revealed the presence of alpha helices, while the orange color indicates the beta sheets. The loops were exhibited in white. During the simulation, it was estimated that the secondary structures did not show fluctuations and remained stable upon binding the ligands.
The residual fluctuations of the BRD4 receptor upon binding of the selected compounds. (a) CSC101378549, (b) MCULE-2,532,030,345.
The percentage of secondary structure elements of BRD4 receptor upon binding of the selected compounds. (a) CSC101378549, (b) MCULE-2,532,030,345.
Protein-Ligand contacts
Based on the MD Simulation study, the most important types of interactions between the ligands and the protein were identified as hydrophobic, hydrogen, and ionic bonds. The functional properties of the protein-ligand complexes are stabilized and regulated by these interactions34. CSC101378549 participated in the hydrogen bonding with the residues Pro82, Gln85, Tyr97, and Asn135. Met105, and Tyr139 residues show water bridges while Phe83, Val87, Leu94, Met134, Cys136, Ile146 depicts hydrophobic interactions (Fig. 8a). In the complex of MCULE-2,532,030,345, residues that form bonds with hydrogen were Gln85, Tyr97, Asn135, Cys136, Tyr139, and Asn140 while Phe83, Val87, Leu92, Leu94 and Tyr139 form hydrophobic interactions (Fig. 8b). The interactions between hydrogen bonds that were seen in the MD simulations not only highlighted the specific residues that were crucial for stabilizing the protein-ligand complexes, but they also provided insight into the crucial interactions that maintain the complexes’ general stability and binding affinity61.
Protein-ligand interaction during MDS. (a) CSC101378549 and (b) MCULE-2,532,030,345. Large, stacked bars represent the interacting residues. Green represents H-bonds, grey for hydrophobic interactions, blue for water bridges, and pink for ionic interactions.
hydrogen bonding analysis
Hydrogen bonding is essential for the protein-ligand combination to remain stable. Consequently, the ligand hydrogen bonds and active site residues were calculated. The hydrogen bonding plots indicate that CSC101378549 made at least 02 hydrogen bond with the protein. The number of hydrogen bonds exceeded 03 and 04 at some frames (Fig. 9a). On the other hand, MCULE-2,532,030,345 made up to 03 hydrogen bonds till 50 ns, and then the number of hydrogen bonds reduced to 02 in the later part of the simulation. Once again, 03 H-bonds can be seen at the end of the simulation (Fig. 9b).
The number of hydrogen bonds between BRD4 and selected ligands (CSC101378549 and MCULE-2532030345) calculated during a 200 ns simulation.
SASA and radius of gyration
Solvent accessible surface area (SASA) demonstrates protein-ligand complex accessibility to solvent and interactions with solvent. SASA also gives insights regarding protein folding and stability events, along with conformational changes associated with ligand binding62. Moreover, SASA impacts drug absorption and metabolism. The average SASA values for CSC101378549-BRD4 complex were 8200 ± 200 Å2, and for MCULE-2,532,030,345-BRD4 system lie within the acceptable range of 7800 to 8400 Å2, which indicates that the ligand stays stable inside the target protein’s binding pocket and protein remains compact over the simulation time (Fig. 10).
Solvent accessible surface area (SASA) for 200ns molecular dynamic simulations. (A) CSC101378549 and (B) MCULE-2,532,030,345.
The radius of gyration (Rg), used to track compactness and structural arrangement (folding) of protein63. Figure 11 demonstrates fluctuations in Rg of CSC101378549-BRD4 and MCULE-2,532,030,345-BRD4 systems over the simulation time. The average Rg values were 15 ± 0.5 Å2 and 15 ± 0.3 Å2 for CSC101378549-BRD4 and MCULE-2,532,030,345-BRD4 systems, respectively. CSC101378549-BRD4 complex depicts no sharp and sudden peaks, while MCULE-2,532,030,345-BRD4 system demonstrates inevitable fluctuation at 40ns. According to these findings, Rg values of both complexes varied within the permitted range, and the receptor remained compact during the simulation time.
Radius of Gyration (rGyr.) for 200ns MD Simulation. (A) CSC101378549 and (B) MCULE-2,532,030,345.
Principal component analysis
Principal component analysis (PCA) was performed to calculate the variance percentage within protein clusters. Moreover, it helps identify significant collective motions of complexes. The dominant movements were observed in the first five eigenvectors in both complexes. The eigenvalues in CSC101378549 complex were 21.6, 36.1, 46.4, 52.4, and 66% in the first five eigenvectors, respectively. The plotted PCs depict < 50% of the total variations. The highest fluctuations were observed in PC1 (21.61%), while PC3 exhibited minimal variation of 10.25% compared to the others (Fig. 12a). Similarly, the eigenvalues in the first five eigenvectors of MCULE-2,532,030,345 were 39.2%, 56.5%, 63.5%, 68%, and 77.4%, respectively. The highest variation was observed in PC1, which recorded 39.18% fluctuations during simulation. Contrastingly, 7.04% variation was observed in PC3, depicting stabilized binding (Fig. 12b). The PCA-based 2D energy surface was also generated to calculate the configurations with stable thermodynamic values. The energy surface calculated the fluctuation direction of energy in two PCs (PC1 and PC2) for carbon alpha atoms by employing the Boltzmann relation64. Most of the clusters lie in the local minima region depicted in purple color, indicating a stable transition from one configuration to another (Fig. 13).
The principal component analysis indicates fluctuations in different hyperspaces. (a) CSC101378549 and (b) MCULE-2,532,030,345. The Blue regions showed the most significant movement, the white regions showed intermediate movement, while the red regions showed less flexibility.
PCA-based free energy surface of the complex calculated during simulation. (a) CSC101378549 and (b) MCULE-2,532,030,345.
Maps of positive and negative correlation motions
The dynamic cross-correlation matrix (DCCM) provides insights regarding positively and negatively correlated motions over time within both CSC101378549-BRD4 and MCULE-2,532,030,345-BRD4 systems at residual levels (Fig. 14). The blue and skin (lighter) color intensity in this heatmap represents correlated and anticorrelated motions, respectively, and assists in analyzing collective motions. According to current findings, the cutoff criteria for positively correlated residues lie above 0.8, while those <-0.4 for negatively correlated residues65. The blue diagonal line indicates a highly positive correlation of residues with themselves, displaying receptor rigidity too66. The residues range from 20 to 45, and 80 to 110 depict ligand-induced positively correlated motions. Moreover, the DCC matrix displays correlation with neighboring residues in both complexes. Resultantly, the matrix demonstrates stability in both complexes, while active site residues are involved in binding.
Dynamic Cross Correlation Matrix (DCCM) for (A) CSC101378549 and (B) MCULE-2,532,030,345 systems from 200ns MD simulations. The extents of correlated and anti-correlated motions are color-coded from blue to skin tone.
Discussion
BRD4, also known as bromodomain-containing protein 4, is involved in various biological processes, including transcriptional regulation and the progression of the cell cycle. The correlation of BRD4 dysregulation with various diseases, such as neuroblastoma, has underscored its potential as a therapeutic target. The integrative pharmacoinformatic approach discussed here aims to speed up the discovery of novel compounds that can inhibit BRD410,67.
Pharmacophores were used to virtually screen the five databases (Enamine, MCULE, Chemspace, ChemDiv, and CHEMBL). In order to determine the optimal binding modes, the screened hits were then docked to the BRD4 active site. Pharmacophore modeling allows for the identification of critical chemical features required for effective ligand binding. Using this method, compounds with specific pharmacophoric properties in connection to BRD4 inhibition are found by searching through large chemical databases. The virtual screening step reduces the potential inhibitor count in the pool by functioning as an initial filter68.
Using the chemical characteristics of a BRD4 protein co-crystal ligand (73B) (PDB ID: 4BJX), a pharmacophore model was created. The virtual screening model was developed using the four pharmacophoric properties of co-crystal ligands: hydrogen bond acceptor, hydrogen bond donor, aromatic, and hydrophobic. Based on these features, a pharmacophore-based virtual screening of five databases was carried out, and 1089 hits that satisfied the screening requirements were selected.
The PDB was used to obtain the crystal structure of the BRD4 protein. Using data from the PDB database, the crystal structure of the BRD4 protein (PDB ID: 4BJX) was retrieved and prepared for docking purposes. The hit compounds discovered during virtual screening were docked to the prepared BRD4 receptor using the Glide tool’s standard precision mode to ascertain binding affinities. This stage makes it easier to forecast whether the chemicals will be able to inhibit BRD4’s enzymatic activity24,69. The top ten compounds were chosen for further investigation based on their binding affinities, interactions, distances, and RMSD scores. The chosen chemicals’ binding affinities varied from − 9.623 to -8.894 kcal/mol, RMSD ≤ 1 Å, and the distance among interacting residues and ligands was ≤ 3 Å. The selected compounds’ binding affinities indicated that they would hinder the BRD4 protein’s ability to function. The known pharmaceutical agents, RVX-208 and MS436, serve as benchmarks in molecular docking and depict that selected compounds (CSC101378549 and MCULE-2532030345) display potential BRD4 inhibition capability.
We examined the molecular interactions between the selected hits and the binding pocket of the BRD4 receptor. Alkyl, carbon, conventional, van der Waal, pi-pi stacking, pi-sulfur, and pi-sigma interactions were the most often found interactions. These interactions have a significant impact on each of the top candidate compounds’ binding affinities and docking scores. Nonpolar, uncharged hydrophobic Pro82, polar uncharged Tyr97, and Asn140 residues depict significant interactions with the targeted receptor.
Drug development also requires a critical evaluation of ADMET features. The pharmacokinetic and toxicological properties of the discovered BRD4 inhibitors are assessed computationally to offer information about their possible safety and effectiveness. Drug likeness, bioavailability, and toxicity analysis were also used to display the safety profile of compounds. The predicted biological activity of CSC101378549 and MCULE-2,532,030,345 compounds demonstrates greater inhibitory potential than other studied compounds. Early in the drug discovery process, these crucial steps eliminate potentially harmful pharmacological traits, saving time and money70. All of the compounds’ values fell within acceptable ranges when the expected ADMET properties of the chosen compounds were examined.
Following the molecular interaction analysis, the potential binding processes of each were investigated by matching the top two chemicals (CSC101378549 and MCULE-2532030345) on the co-crystal ligand. The docked molecules showed a comparable binding mechanism and were positioned suitably on the co-crystal ligand, according to the study. This analysis supports the accuracy and reliability of the docking protocol. Additionally, it assists in confirming that hits interact with active site residues.
With Molecular Dynamics (MD) simulations, the dynamic behavior of BRD4-inhibitor complexes across time is also examined. This computational method offers useful insights into structural alterations that might impact the inhibitor’s effectiveness, allowing researchers to explore the stability and flexibility of binding interactions71. Flexibility and rigidity of ligands play a significant role in influencing their binding interactions with the BRD4 receptor. Flexible ligands have more rotatable bonds, which enable them to occupy several conformations and better adjust in the binding pocket of the targeted receptor. On the other hand, rigid ligands may fit more accurately into well-defined binding pockets but with more specific interactions and lack flexible binding sites too72. In the current study, both compounds depict moderate flexibility and rigidity (as reflected in RMSD and RMSF profiles during MD simulations) following the stable interactions with the receptor. CSC101378549 and MCULE-2,532,030,345 continued to be potent inhibitors inside the protein binding area, according to molecular dynamics simulations. Both of the compounds show stability and mobility with RMSD and RMSF values of ≤ 3 Å and 1.6 Å, respectively, strengthening the reliability of the docking results. Moreover, they depict H-bonds with Gln85, Tyr97, Asn135, Tyr139 residues, water bridges with Tyr97, Asn135, Tyr139 residues, and hydrophobic interactions with Phe83, Val87, Leu94 residues mainly. Some of the residues show more than one type of interaction, i.e., Tyr97 and Asn135; such residues play a significant inhibitory role in the active site of BRD4 receptor. On the other hand, compactness and protein folding were assessed through analyses such as radius of gyration and SASA, whose values lie within an acceptable range. Additionally, collective, correlated, and anti-correlated motions were analyzed via PCA and dynamic cross-correlation analysis34. According to all of these results, the selected hit compounds could function as lead compounds, which would reduce the biological activity of BRD4.
Conclusion
This study presents an integrated pharmacoinformatic strategy for the identification of novel BRD4 inhibitors with potential application in neuroblastoma treatment. By combining pharmacophore-based virtual screening, molecular docking, ADMET prediction, and molecular dynamics simulations, we effectively narrowed down a large chemical space to a shortlist of promising candidates. Among 1089 hits initially identified, ten compounds exhibited strong binding affinity to the BRD4 active site, with docking scores ranging from − 9.623 to − 8.894 kcal/mol. ADMET profiling indicated favorable pharmacokinetic and toxicity profiles, while 100 ns molecular dynamics simulations confirmed the conformational stability of the top ligand-receptor complexes.
Beyond the identification of potential lead compounds, this study demonstrates the power of In silico workflows to accelerate early-phase drug discovery and reduce the cost and time associated with experimental screening. The findings support the continued exploration of BRD4 as a therapeutic target in oncology and highlight a rational pipeline that can be adapted for other epigenetic targets. Future experimental validation of the lead compounds will be essential to advance them toward preclinical development.
Data availability
Data is provided within the manuscript.
References
Fetahu, I. S. & Taschner-Mandl, S. Neuroblastoma and the epigenome. Cancer Metastasis Rev. 40 (1), 173–189 (2021).
Gao, Y. et al. Synergistic anti-tumor effect of combining selective CDK7 and BRD4 Inhibition in neuroblastoma. Front. Oncol. 11, 773186 (2022).
Tan, Y. et al. Inhibition of BRD4 suppresses tumor growth in prostate cancer via the enhancement of FOXO1 expression. Int. J. Oncol. 53 (6), 2503–2517 (2018).
Wu, X. et al. Inhibition of BRD4 suppresses cell proliferation and induces apoptosis in renal cell carcinoma. Cell. Physiol. Biochem. 41 (5), 1947–1956 (2017).
Zhao, Y. et al. Pharmacoproteomics reveal novel protective activity of bromodomain containing 4 inhibitors on vascular homeostasis in TLR3-mediated airway remodeling. J. Proteom. 205, 103415 (2019).
Tian, B. et al. Efficacy of novel highly specific bromodomain-containing protein 4 inhibitors in innate inflammation–driven airway remodeling. Am. J. Respir. Cell Mol. Biol. 60 (1), 68–83 (2019).
Tian, B. et al. BRD4 mediates NF-κB-dependent epithelial-mesenchymal transition and pulmonary fibrosis via transcriptional elongation. Am. J. Physiology-Lung Cell. Mol. Physiol. 311 (6), L1183–L1201 (2016).
Dey, A. et al. Brd4 marks select genes on mitotic chromatin and directs postmitotic transcription. Mol. Biol. Cell. 20 (23), 4899–4909 (2009).
Dey, A. et al. A bromodomain protein, MCAP, associates with mitotic chromosomes and affects G2-to-M transition. Mol. Cell. Biol. 20 (17), 6537–6549 (2000).
Liu, Z. et al. Drug discovery targeting bromodomain-containing protein 4. J. Med. Chem. 60 (11), 4533–4558 (2017).
Shi, J. et al. Disrupting the interaction of BRD4 with diacetylated twist suppresses tumorigenesis in basal-like breast cancer. Cancer cell. 25 (2), 210–225 (2014).
Brasier, A. R. & Zhou, J. Validation of the epigenetic reader bromodomain-containing protein 4 (BRD4) as a therapeutic target for treatment of airway remodeling. Drug Discovery Today. 25 (1), 126–132 (2020).
Alamer, E. et al. Epigenetic suppression of HIV in myeloid cells by the BRD4-selective small molecule modulator ZL0580. J. Virol. 94 (11), 01880–01819. https://doi.org/10.1128/jvi (2020).
Filippakopoulos, P. & Knapp, S. Targeting bromodomains: epigenetic readers of lysine acetylation. Nat. Rev. Drug Discovery. 13 (5), 337–356 (2014).
Cochran, A. G., Conery, A. R. & Sims, R. J. Bromodomains: a new target class for drug development. Nat. Rev. Drug Discovery. 18 (8), 609–628 (2019).
Cully, M. BET inhibitor attenuates heart failure. Nat. Rev. Drug Discovery. 16 (7), 453–453 (2017).
Filippakopoulos, P. et al. Selective Inhibition of BET bromodomains. Nature 468 (7327), 1067–1073 (2010).
Hao, J. et al. Downregulation of BRD4 inhibits gallbladder cancer proliferation and metastasis and induces apoptosis via PI3K/AKT pathway. Int. J. Oncol. 51 (3), 823–831 (2017).
Bid, H. K. et al. The bromodomain BET inhibitor JQ1 suppresses tumor angiogenesis in models of childhood sarcoma. Mol. Cancer Ther. 15 (5), 1018–1028 (2016).
Padmanabhan, B. et al. Bromodomain and extra-terminal (BET) family proteins: new therapeutic targets in major diseases. J. Biosci. 41, 295–311 (2016).
Belkina, A. C. & Denis, G. V. BET domain co-regulators in obesity, inflammation and cancer. Nat. Rev. Cancer. 12 (7), 465–477 (2012).
Chiang, C. M. Nonequivalent response to bromodomain-targeting BET inhibitors in oligodendrocyte cell fate decision. Chem. Biol. 21 (7), 804–806 (2014).
Liu, Z. et al. Discovery of potent and selective BRD4 inhibitors capable of blocking TLR3-induced acute airway inflammation. Eur. J. Med. Chem. 151, 450–461 (2018).
Er-Rajy, M. et al. Design of novel anti-cancer drugs targeting TRKs inhibitors based 3D QSAR, molecular Docking and molecular dynamics simulation. J. Biomol. Struct. Dynamics. 41 (21), 11657–11670 (2023).
Aloui, M. et al. QSAR modelling, molecular docking, molecular dynamic and ADMET prediction of pyrrolopyrimidine derivatives as novel bruton’s tyrosine kinase (BTK) inhibitors. Saudi Pharm. J. 32 (1), 101911 (2024).
Aloui, M. et al. QSAR modelling, molecular docking, molecular dynamic and ADMET prediction of pyrrolopyrimidine derivatives as novel bruton’s tyrosine kinase (BTK) inhibitors. Saudi Pharm. J. ; 32: 101911. (2024).
Er-rajy, M. et al. 2D-QSAR modeling, drug-likeness studies, ADMET prediction, and molecular Docking for anti-lung cancer activity of 3-substituted-5-(phenylamino) Indolone derivatives. Struct. Chem. 33 (3), 973–986 (2022).
Er-rajy, M. et al. Design of novel anti-cancer agents targeting COX-2 inhibitors based on computational studies. Arab. J. Chem. 16 (10), 105193 (2023).
Er-Rajy, M. et al. Design of potential anti-cancer agents as COX-2 inhibitors, using 3D-QSAR modeling, molecular docking, oral bioavailability proprieties, and molecular dynamics simulation. Anti-Cancer Drugs. 35 (2), 117–128 (2024).
Rudrapal, M., Khairnar, S. J. & Jadhav, A. G. Drug repurposing (DR): an emerging approach in drug discovery. Drug repurposing-hypothesis, molecular aspects and therapeutic applications, 10. (2020).
Ugbe, F. A. et al. Computational evaluation of the inhibitory potential of some urea, thiourea, and Selenourea derivatives of diselenides against leishmaniasis: 2D-QSAR, pharmacokinetics, molecular docking, and molecular dynamics simulation. J. Mol. Struct. 1302, 137473 (2024).
Edache, E. I. et al. Drug-like screening, molecular docking, molecular dynamics simulations, and binding free energies on the interaction of pyrazole derivatives as inhibitors of lysosomal storage disorders and anticancer activity. Discover Chem. 1 (1), 22 (2024).
Edache, E. I. et al. DFT studies on structure, electronics, bonding nature, NBO analysis, thermodynamic properties, molecular docking, and MM-GBSA evaluation of 4-methyl-3-[2-(4-nitrophenyl)-1, 3-dioxo-2, 3-dihydro-1 H-isoindole-5-amido] benzoic acid: a potent inhibitor of Graves’ disease. J. Umm Al-Qura Univ. Appl. Sci. 1–19. (2024).
Edache, E. I. et al. Evaluation of novel Anti-SARS-CoV-2 compounds by targeting nucleoprotein and envelope protein through homology modeling, Docking simulations, ADMET, and molecular dynamic simulations with the MM/GBSA calculation. Intell. Pharm. 2 (3), 346–366 (2024).
Naqvi, A. A. et al. Advancements in Docking and molecular dynamics simulations towards ligand-receptor interactions and structure-function relationships. Curr. Top. Med. Chem. 18(20): pp. 1755–1768. (2018).
Sunseri, J. D.R.J.N.a.r. Koes. Pharmit: Interact. Explor. Chem. Space. 44 (W1), W442–W448 (2016).
Prabitha, P. et al. Multi-conformational frame from molecular dynamics as a structure-based pharmacophore model for mapping, screening and identifying ligands against PPAR-γ: a new protocol to develop promising candidates. J. Biomol. Struct. Dyn. 40(6) 2663–2673. (2022).
Oduselu, G. O., Ajani, O. O. & Ajamma, Y. U. Structure-Based drug design in discovering target specific drugs against plasmodium falciparum adenylosuccinate lyase. (2021).
LigPrep, L. P. Schrödinger, LLC. (2018).
Shivakumar, D. et al. Improving the prediction of absolute solvation free energies using the next generation OPLS force field. J. Chem. Theory Comput. 8(8): pp. 2553–2558. (2012).
Schrödinger, L. J. S. S. Schrödinger, LLC; New York, NY: 2017. 2: pp. 2017–1. (2017).
Kim, M. O. et al. Effects of histidine protonation and rotameric States on virtual screening of M. tuberculosis RmlC. J. Comput.-aided Mol. Des. 27(3): pp. 235–246. (2013).
Friesner, R. A. et al. Glide: a new approach for rapid, accurate Docking and scoring. 1. Method and assessment of Docking accuracy. J. Med. Chem. 47(7): pp. 1739–1749. (2004).
Daina, A., Michielin, O. & Zoete, V. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci. Rep. 7 (1), 42717 (2017).
Sander, T. et al. DataWarrior: an open-source program for chemistry aware data visualization and analysis. J. Chem. Inf. Model. 55 (2), 460–473 (2015).
Mali, SN. & Chaudhari, H.K. Computational studies on imidazo [1, 2-a] pyridine-3-carboxamide analogues as antimycobacterial agents: Common pharmacophore generation, atom-based 3D-QSAR, molecular dynamics simulation, QikProp, molecular docking and prime MMGBSA approaches. Open Pharma. Sci. J. 5(1). (2018).
Genheden, S. & Ryde, U. The MM/PBSA and MM/GBSA methods to estimate ligand-binding affinities. Expert Opin. Drug Discov. 10 (5), 449–461 (2015).
Ali, I. et al. Computational exploration of novel ROCK2 inhibitors for cardiovascular disease management; insights from high-throughput virtual screening, molecular docking, DFT and MD simulation. Plos One. 18 (11), e0294511 (2023).
Bowers, K. J. et al. Scalable algorithms for molecular dynamics simulations on commodity clusters. in Proceedings of the ACM/IEEE Conference on Supercomputing (2006).
Price, D. J. & C.L.J .T.J.o.c.p. Brooks III, A modified TIP3P water potential for simulation with Ewald summation. J. Chem. Phys. 121(20), 10096–10103. (2004).
Mir, I. H. et al. Elucidation of Escitalopram oxalate and related antidepressants as putative inhibitors of PTP4A3/PRL-3 protein in hepatocellular carcinoma: A multi-computational investigation. Comput. Biol. Chem. 108039. (2024).
Thillainayagam, M. et al. In-Silico molecular docking and simulation studies on novel chalcone and flavone hybrid derivatives with 1, 2, 3-triazole linkage as vital inhibitors of Plasmodium falciparum dihydroorotate dehydrogenase. J. Biomol. Struct. Dyn. 36(15), 3993–4009. (2018).
Lata, S. et al. The study of structure and effects of two new proanthocyanidins from anogeissus pendula leaves on rumen enzyme activities. Front. Veterinary Sci. 10, 1163197 (2023).
Khan, T. et al. Molecular docking, PASS analysis, bioactivity score prediction, synthesis, characterization and biological activity evaluation of a functionalized 2-butanone thiosemicarbazone ligand and its complexes. J. Chem. Biol. 10, 91–104 (2017).
Chow, S. C. Bioavailability and bioequivalence in drug development. Wiley Interdisciplinary Reviews: Comput. Stat. 6 (4), 304–312 (2014).
Godschalk, F. et al. Comparison of MM/GBSA calculations based on explicit and implicit solvent simulations. Phys. Chem. Chem. Phys. 15(20): pp. 7731–7739. (2013).
Decherchi, S. & Cavalli, A. J. C. R. Thermodyn. Kinetics drug-target Binding Mol. Simul. 120(23): 12788–12833. (2020).
Sargsyan, K. et al. How molecular size impacts RMSD applications in molecular dynamics simulations. J. Chem. Theory Comput. 13(4): pp. 1518–1524. (2017).
Martínez, L. J. P. Automatic identification of mobile and rigid substructures in molecular dynamics simulations and fractional structural fluctuation analysis.PloS one 10(3): p. e0119264. (2015).
Ugbe, F. A. et al. Cheminformatics-based discovery of new organoselenium compounds with potential for the treatment of cutaneous and visceral leishmaniasis. J. Biomol. Struct. Dynamics. 42 (24), 13830–13853 (2024).
Ugbe, F. A. et al. Cheminformatic evaluation of the multi-protein binding potential of some diselenide derivatives: A plausible drug discovery approach for leishmaniasis. Discover Chem. 1 (1), 25 (2024).
Ullah, S. et al. Identification of lead compound screened from the natural products atlas to treat renal inflammasomes using molecular Docking and dynamics simulation. J. Biomol. Struct. Dynamics, 1–11. (2023).
Thirumal Kumar, D. et al. A molecular Docking and dynamics approach to screen potent inhibitors against fosfomycin resistant enzyme in clinical Klebsiella pneumoniae. J. Cell. Biochem. 118 (11), 4088–4094 (2017).
Dalal, V. et al. Structure-based identification of potential drugs against FmtA of Staphylococcus aureus: virtual screening, molecular dynamics, MM-GBSA, and QM/MM. Protein. J. 40, 148–165 (2021).
Ali, I. et al. Identification of natural lead compounds against hemagglutinin-esterase surface glycoprotein in human coronaviruses investigated via MD simulation, principal component analysis, cross-correlation, H-bond plot and MMGBSA. Biomedicines 11 (3), 793 (2023).
Ashraf, N. et al. Combined 3D-QSAR, molecular Docking and dynamics simulations studies to model and design TTK inhibitors. Front. Chem. 10, 1003816 (2022).
Liang, D., Yu, Y. & Ma, Z. Novel strategies targeting bromodomain-containing protein 4 (BRD4) for cancer drug discovery. Eur. J. Med. Chem. 200, 112426 (2020).
Muhammed, M. T. & Esin, A. Y. Pharmacophore modeling in drug discovery: methodology and current status. J. Turkish Chem. Soc. Sect. A: Chem. 8 (3), 749–762 (2021).
Jakhar, R. et al. Relevance of molecular Docking studies in drug designing. Curr. Bioinform. 15 (4), 270–278 (2020).
Jia, C. Y. et al. A drug-likeness toolbox facilitates ADMET study in drug discovery. Drug Discovery Today. 25 (1), 248–258 (2020).
Salo-Ahen, O. M. et al. Molecular dynamics simulations in drug discovery and pharmaceutical development. Processes 9 (1), 71 (2020).
Bobrovnik, S. The influence of rigid or flexible linkage between two ligands on the effective affinity and avidity for reversible interactions with bivalent receptors. J. Mol. Recognition: Interdisciplinary J. 20 (4), 253–262 (2007).
Acknowledgements
The authors express their gratitude to the Department of Biosciences, COMSATS University Islamabad, Islamabad Campus, 45550, Islamabad, Pakistan, for the support. The authors acknowledge the support of the publication fee by the CSIC Open Access Publication Support Initiative through its Unit of Information Resources for Research (URICI), and the Deanship of Scientific Research, King Faisal University, through the project number: KFU253329.
Author information
Authors and Affiliations
Contributions
I.A.: Conceptualization, Investigation, writing-original draft, methodology, designed and conceived the study. F.A.: MD Simulation. M.A., N.S.Y.: Formal analysis, review, and editing. M.A., N.S.Y., and A.A.K.: Resources. G.Y.: Supervision. All authors have read and agreed to the published version of the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Ali, I., Almostafa, M., Abbas, F. et al. Computational discovery of BRD4 inhibitors for neuroblastoma therapy using pharmacophore screening and molecular simulations. Sci Rep 15, 36850 (2025). https://doi.org/10.1038/s41598-025-20714-2
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41598-025-20714-2
