Computational design of quinazolin-4(3H)-one derivatives as multitarget inhibitors against angiogenesis and metabolic reprogramming to help overcome cancer resistance

2.1. Data set

To build 3D-QSAR models for a set of 40 quinazolin-4(3H)-one derivative [27-32] (as indicated in Supplementary File 1 Table S1). The dataset was randomly divided into two subsets: one for training the model and the other for testing its performance.

2.2. Minimization and molecular alignment

The alignment technique utilized in comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) analysis, as emphasized by Xiao et al [32], poses a substantial difficulty. The SYBYL-X 2.0 molecular modeling program was used to model all the substances being considered. The module for the designed benzothiazole derivatives was optimized using the Tripos force field. The SYBYL program framework was employed to assign Gasteiger-Huckel charges, with convergence set at 0.01 kcal/mol. Figure 1 illustrates the alignment of the 40 benzothiazole derivative compounds to a common core, with molecule 39 (the most potent in the dataset) acting as the reference compound.

Close

Figure 1.

Alignment of 40 Quinazolin-4(3H)-one derivatives using molecule 39 as a template.

Export to PPT

2.3. Methodology for CoMFA and CoMSIA descriptor generation

The various 3D-QSAR models were constructed using the Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA) methods [33,34]. These methods were used to predict the anticancer activity (pIC50 Pred) of the studied quinazolin-4(3H)-one derivatives across various fields, including steric, electrostatic, hydrophobic, hydrogen bond donor, and acceptor interactions.

In the CoMFA method, all molecules were arranged on a grid lattice with a 2.0 Å spacing. An sp3 hybridized carbon atom with a charge +1 is used to model the electrostatic and steric field energies. The interactions’ cutoff energies have values of 30 kcal/mol, and the column filtering is set to a value of 2.0 kcal/mol [34].

The CoMSIA method computes different descriptors on a 2.0 Å grid. The cutoff energy for interactions is set around 30 kcal/mol, with an attenuation factor of 0.3. Column filtering is performed by evaluating molecular field energies, ensuring that the variation is less than 2.0 kcal/mol [35].

2.4. Partial least squares analysis

The 3D-QSAR models in this study are developed using the partial least squares (PLS) regression method, which establishes the correlation between two sets of variables [36]. In this investigation, the two variables are the CoMFA/CoMSIA descriptors and the pIC50 anticancer activity values. The calculation of the optimal number of components (N) and the determination of the cross-validation correlation coefficient (Q²) were performed via leave-one-out cross-validation [37]. Subsequently, non-cross-validation techniques were conducted to determine the determination coefficient (R²), the F-test value (F), and the standard error of estimation (SEE) [38].

Additionally, to assess the robustness of the developed models, they were externally validated using a set of four molecules [39]. The coefficient of determination (Rext²) for the test set was calculated using the following Eq. (1):

For the models to be considered acceptable, they must simultaneously meet the following criteria: Q² > 0.5, R² > 0.6, and Rext² > 0.6 [40]

2.5. In silico pharmacokinetics absorption, distribution, metabolism, excretion, and toxicity (ADMET)

ADMET studies are pivotal for evaluating potential drug candidates in drug research and development. ADMET refers to a set of pharmacokinetic parameters that can be evaluated through in silico approaches. These parameters include the absorption of a compound in the human intestine, its ability to cross the blood-brain barrier (BBB) and affect the central nervous system (CNS), the body’s biotransformation of the compound, as well as the drug’s overall clearance and potential toxicity levels. These factors are crucial in assessing the drug’s suitability for therapeutic use [21].

2.6. Molecular docking

The 3D structures of all 18 protein targets, as outlined in Supplementary File 1 Table S2, were retrieved from the Protein Data Bank (RCSB) via https://www.rcsb.org/(accessed in May 2024). These structures were visualized using UCSF Chimera [41] and prepared with MGLtools (version 1.5.6, The Scripps Research Institute, La Jolla, CA, USA) [41]. Preprocessing involved removing water molecules, heteroatoms (hetatm), and co-crystallized ligands. Subsequently, polar hydrogen atoms were added, Gasteiger charges were assigned, and the structures were converted to pdbqt format for further analysis [42]. The grid box spacing was set to 0.375 Å, centered on the regions where co-crystallized ligands interact with active site residues. Docking simulations were conducted for all 18 protein targets, generating nine poses per protein-ligand complex based on docking affinity. The docking outcomes were visualized and analyzed using Discovery Studio Viewer to identify critical interactions between ligands and protein binding sites. For each protein target, the conformation with the lowest binding affinity (as indicated by docking scores) and the highest number of bonds was chosen as the initial binding mode for subsequent molecular dynamics simulations.

Next, we conducted redocking, a widely accepted technique for validating the molecular docking process. This method entails removing the crystallized ligand from the protein and conducting a docking analysis within the same binding site [42]. By comparing the alignment of the docked ligand with the original crystallized ligand, we can accurately assess the docking accuracy through the analysis of the root mean square deviation (RMSD) parameter using PyMOL [43].

2.7. Molecular dynamics simulations (DFT)

Molecular dynamics simulations (MDs) were carried out to examine the conformations of each complex with the lowest binding affinities, utilizing GROMACS version 2019.3 [44] on a high-performance computing cluster. Protein topologies were generated using the pdb2gmx module in GROMACS with the CHARMM27 force field [45], while ligand topologies were created using SwissParam to derive the necessary force field parameters [46]. Each complex was positioned within a dodecahedral box (1.0 nm) filled with TIP3P water molecules and neutralized with counter ions. Energy minimization was performed using the steepest descent method, ensuring that the maximum force (Fmax) remained below 1000 kJ/mol/nm [47]. The system was equilibrated at 300 Kelvin and 1 bar pressure through two consecutive 100 ps simulations employing the canonical NVT and isobaric NPT ensembles. Temperature and pressure were controlled using the Berendsen thermostat and the Parrinello-Rahman barostat. The MD simulations were conducted over 100 ns, totaling 50,000,000 steps, with coordinates recorded every 2 fs. The output trajectories were generated, and the corresponding data files were analyzed to gain a deeper understanding of the protein’s behavior.

2.8. Density functional theory

All calculations were performed within the DFT framework using the Gaussian software [48]. The electronic and structural characteristics of the selected compounds, including molecule 39, were examined at the B3LYP/6-311++G(d,p) level of theory. Various molecular parameters were computed, including the highest occupied molecular orbital (HOMO), the lowest unoccupied molecular orbital (LUMO), their respective energies (E_HOMO and E_LUMO), the energy gap (ΔEgap), ionization potential (I), electron affinity (A), electronegativity (χ), chemical hardness (η), chemical potential (μ), chemical softness (S), electrophilicity index (ω), dipole moment (D), and molecular electrostatic potential (MEP) maps.

3.1. Alignment of quinazolin-4(3H)-one derivatives and 3D-QSAR models generation

Molecule 39, which demonstrates the highest inhibitor activity was selected to construct a robust 3D-QSAR model, serve as a template for data alignment (Table S1), and facilitate the visualization of the contour maps in the CoMFA and CoMSIA analyses (Figure 1).

The primary goal of this step is to develop robust CoMFA and CoMSIA models based on the observed and predicted pIC50 values for the training and test sets. For the CoMFA model, steric and electrostatic fields were combined (Supplementary File 2 Table S1). In contrast, a total of 31 combinations of steric, electrostatic, hydrophobic, H-bond donor, and H-bond acceptor were employed in developing CoMSIA (Supplementary File 2 Tables S2-S3). The optimal model was determined by selecting the one with the highest coefficient of determination values for both non-cross-validation (R²) and cross-validation (Q²), along with the lowest standard error of estimate (SEE), the fewest principal components (N), and the most significant F-test value [49]. Table 1 lists the best CoMFA and CoMSIA models.

Principle component analysis of the selected models indicated that the one with the highest Q² (0.717) and R² (0.995), the lowest SEE (0.101), and the most significant F-value (946.847) was the most robust and reliable. This model employed steric, hydrophobic, and hydrogen bond acceptor (SHA) fields and demonstrated the highest predictive power (R²_Pred = 0.832). Therefore, the CoMSIA/SHA model exhibited good stability and predictability, as indicated in Table 2. The model’s predictive ability was validated according to Golbraikh and Tropsha’s statistical criteria [50-52], as shown in Table 3. The proposed CoMSIA/ steric, hydrophobic, acceptor (SHA) model was further validated based on these criteria.

3.2. Graphical interpretation of the best COMSIA/SHA model

The validated 3D-QSAR model was visualized using CoMSIA/SHA contour maps, using molecule 39 as a template. The contour maps aid in an understanding of the structural characteristics of the compounds and provide insights into how molecular features contribute to their biological activities. Figure 2 illustrates the critical fields: steric (a), hydrophobic (b), and hydrogen bond acceptor (c). In the steric field (a), green contours represent regions where larger substituents are favorable, indicating that bulkier groups in these areas enhance activity, while yellow contours highlight regions where minimal steric bulk is preferred, suggesting that excessively bulky groups could negatively impact activity. The hydrophobic field (b) is represented by blue and white contours, with blue areas favoring positive hydrophobicity, indicating that more hydrophobic substituents in these regions could boost activity, whereas white contours signify regions where hydrophilic properties are more favorable, suggesting that hydrophobic groups should be minimized. The hydrogen bond acceptor field (c) shows magenta contours in regions that favor hydrogen bond acceptor groups, highlighting the importance of such functionalities for improving biological interactions. Conversely, red contours mark areas where hydrogen bond acceptor properties are disfavored, suggesting that such features could reduce activity if placed there. In summary, the structure–activity relationship (SAR) of potent compounds highlights the importance of steric and hydrophobic contributions and the hydrogen bond acceptor field. These contour maps collectively offer a strategic framework for designing more potent VEGFR-2 inhibitors.

Close

Figure 2.

Contour maps (CoMSIA/SHA) of molecule 39 with 2 Å grid spacing; (a) Steric fields, (b) Hydrophobic fields, and (c) H-bond Acceptor fields.

Export to PPT

The significant values of the CoMSIA field fractions outlined in Table 4 suggest that enhancing hydrogen bonding, steric, and hydrophobic interactions can significantly improve the efficacy of these compounds against tumor angiogenesis.

3.3. Design new inhibitor candidates against VEGFR-2 kinase based on the selected molecule 39

To design new VEGFR-2 kinase inhibitors by investigating the different structural features extracted from the selected CoMSIA/SHA models (Figure 3). We substituted the sulfonic amide group (–H₂NO₂S) with a 1-hydrosulfonyl-1H-1,2,3-triazole (–C₂H₂N₃O₂S) group, hydrosulfonylbenzene (–C₆H₅O₂S), or 2-cyclopentylpropan-2-yl hydrosulfonylmethanoate (–C₉H₁₅O₄S) at the R₁ position. We added the proper substituents in the R₂, R₃, R₄, R’₄, R₅, R’₅, R₆, R₇, R₈, and R₉ positions according to the contour maps (Figure 3 and Supplementary File 1 Table S3). As a result, fifty-nine new Kinase inhibitors were proposed, and their VEGFR-2 kinase activity was computed using the proposed CoMFA and CoMSIA/SHA models (Table 5).

Close

Figure 3.

CoMSIA/SHA SAR of molecule 39.

Export to PPT

3.4. Drug-likeness assessment and ADMET predictions

Several developed drugs fail in the clinical development stage because of problems like inadequate blood–brain barrier (BBB) permeability, toxicity, or insufficient efficacy. Consequently, predicting and optimizing the ADMET properties of new chemical compounds is crucial to avoid complications in clinical trials [53,54]. This study evaluated the ADMET pharmacokinetic parameters of the 59 newly designed compounds using pkCSM [55]. The ADMET properties and safety endpoints show the compounds’ high intestinal absorption rates, ranging from 38.533% to 100%, with compound Pred49 showing the highest value (Supplementary File 1 Table S4).

A Volume of Distribution at Steady State (VDss) value is classified as low when it falls below 0.71 L/kg, corresponding to a logarithmic value (log VDss) of less than -0.15. Conversely, a VDss value is considered high when it exceeds 2.81 L/kg, equating to a logarithmic value (log VDss) greater than 0.45 [36-58]. Based on these criteria, all proposed compounds, molecule 39, respect the defined barrier for low VDss values, suggesting limited tissue distribution for these molecules.

For BBB and the CNS permeability, compounds exhibiting a LogBB value below -1 indicate poor diffusion into the brain. In contrast, those with a LogBB above 0.3 are deemed capable of breaching the BBB. Furthermore, compounds with a LogPS value over -2 are considered effective in penetrating the CNS; however, those with a LogPS less than -3 are viewed as having difficulties in CNS barrier penetration [59-61]. Our analysis of the permeability results shows that the 59 newly designed compounds are predicted to fail to traverse these barriers.

Among 17 cytochrome P450 (CYP) families, 57 CYP genes were identified in humans, suggesting that CYP enzymes are the main ones responsible for the biotransformation of most foreign substances, including 70-80% of all drugs in clinical use [62-64]. Thus, most of the administered drugs undergo the initial phase of metabolism with these enzyme families. CYP1A2, 2C19, 2D6, 3A4, and 2C9 enzymes are used for our analysis. Therefore, all proposed ligands were identified as either inhibitors or substrates for CYP2D6 (Supplementary File 1 Table S4). Additionally, none of the newly designed compounds exhibited issues with clearance, which measures the rate of drug elimination from the body [65].

In the early stages of drug development, assessing the toxicity of proposed compounds is essential to ensure their safety and efficacy. Therefore, each newly designed compound examined in this study is subjected to a toxicity assessment using the AMES test [66]. This test is widely used due to its simplicity, cost-effectiveness, and ability to provide rapid results regarding a compound’s genotoxicity. As shown in Supplementary File 1 Table S4, the designed ligands were non-toxic. However, molecule 39, Pred11, Pred30, and Pred51 displayed some toxicity. These findings suggest that most of the compounds are promising for further development. Nevertheless, certain compounds need alterations to improve their non-toxicity. The high absorption, distribution, and permeability properties observed make them strong candidates for clinical evaluation.

3.5. Molecular docking (MD) study

Molecular docking and dynamics are central to designing and screening new bioactive molecules [67]. Molecular docking of the 18 protein targets with all newly designed molecules was performed to identify the optimal binding modes, allowing inhibition of the target proteins. This was identified by virtual screening to select the molecules with the highest binding affinity scores. The molecular docking procedures were initially validated as outlined in the methodology section. The Root Mean Square Deviation (RMSD) scores obtained during the validation of the chosen targets ranged from 0.2 to 1.7 Å. An RMSD value below 2.0 Å indicates a reliable predictive protocol for assessing protein-ligand interactions [68]. This indicates that the protocols used were appropriate for docking the proposed compound.

Additionally, parameters such as binding affinity and specific amino acid residues, along with the dimensions and coordinates of the grid box, were observed during validation. Lower Binding affinity values indicate better interaction between the ligand and the target [69]. Supplementary File 2 Table S4 presents the binding affinity values for the most favorable interaction poses of all 59 designed molecules with the selected receptor tyrosine kinase (RTK) targets and PFKFB3. Molecule 39 and the FDA-approved drug for each target were used as a control for comparison. This analysis shows that Pred3, Pred4, Pred34, and Pred49, exhibit better binding affinities than the reference compounds and molecule 39 (Table 6).

Furthermore, the 2D binding interaction of the compounds (Pred3, Pred4, Pred34, and Pred49), molecule 39, and the reference drug for each protein target showed a similar interaction in the binding pocket of all the targets. This is due to several amino acids participating in the same interactions compared to the food and drug administration (FDA) drugs and molecule 39 (Supplementary File 1 Table S5). Additionally, the proposed compounds demonstrate various interactions with different residues, especially in the regions where fragments have been attached to the core structure across all targeted proteins. These results emphasize the strong potential of the proposed compounds as multi-target drugs. Research indicates that multi-target compounds are particularly effective in treating complex diseases [70].

3.6. Molecular dynamics (MD)

Molecular dynamics simulations were conducted to assess the impact of the selected lead compounds (Pred3, Pred4, Pred34, and Pred49), molecule 39, and the reference drug for each target on their respective structures and stability within the binding pockets of each protein target. The optimal pose, which exhibited the most significant binding affinity for each target, was employed as the initial structure for the 100 ns molecular dynamics simulation [71]. The resultant trajectory data files were analyzed for RMSD, RMSF, Rg, and hydrogen bond.

During the simulation, the RMSD was calculated to determine the overall stability of the selected complexes [72] and taken as the primary indicator of the system’s convergence. A lower RMSD value signifies a higher level of system stability and vice versa [72]. Upon examining the RMSD graph shown in Figure 4, it is evident that all the protein-ligand complexes formed by the proposed compounds exhibited reduced fluctuation in their spectrum compared to the complexes formed by the reference drug and molecule 39, which indicates a minimal disruption in their conformational dynamics during the whole simulation. Pred49 shows a deviation between 65 and 80 ns for FGFR3, reaching approximately 0.37 nm before stabilizing. Similarly, for FGFR4, Pred49 displays a slight increase after 80 ns but stabilizes at around 0.35 nm until the end of the simulation. Meanwhile, in the case of KIT proto-oncogene receptor tyrosine kinase, a slight deviation was observed within the first 50 ns, where the RMSD briefly increased to 0.38 nm before stabilizing and maintaining a consistent value near 0.25 nm until 100 ns. These observations suggest that while minor fluctuations occurred, they were brief and quickly resolved, demonstrating that the proposed compounds exhibit greater stability than molecule 39 and reference drugs. Interestingly, this finding is consistent with the docking results, where the proposed compounds exhibited better binding affinities (Table 6), showing their strong potential to serve as promising candidates for inhibiting the selected RTKs and PFKFB3.

Close

Figure 4.

RMSD profiles of the designed compounds, molecule 39, and the reference drug for the selected RTKs and PFKFB3.

Export to PPT

RMSF plots during 100 ns were generated to investigate the binding affinity effects of the selected compounds on the dynamic structure of proteins and the behavior of their essential amino acids [73]. An increase in RMSF values indicates that the protein structure is more likely to have greater flexibility and lower residual stability [74].

RMSF data analysis of the selected complexes generally showed minimal fluctuations in the residues (<4 nm) during the simulations (Figure 5). Compared with the reference drugs and molecule 39, Pred49 exhibited significant fluctuations in residues 550-570 (FGFR3) and 860-870 (HER3). Additionally, Pred4 displayed higher fluctuations in residues 560-580 (FGFR2). Similarly, Pred3, Pred4, and Pred34 are associated with increased fluctuations of residues 760-780 within EPHB4. The fluctuations in these proposed compounds during the simulations could suggest that they assume different conformations to reach the most energetically stable state within the binding cavity of the targets [69]. Notably, compounds Pred3, Pred4, Pred34, and Pred49 displayed the lowest average RMSF values, suggesting they could be more effective inhibitors than the reference drugs.

Close

Figure 5.

Comparison of RMSF profiles between the designed compounds, molecule 39, and the reference drug for each RTK target and PFKFB3.

Export to PPT

The plots of radius of gyration (Rg) are calculated to assess how compact the protein changes upon binding to various compounds throughout the entire simulation [75]. A stable Rg value usually indicates a well-folded structure, while variations in the Rg value during the simulation suggest that the protein structure may be unfolding [76]. The plot shows that, except for Pred4, which has a larger EGFR Rg value, all the remaining proposed compound complexes have a lower Rg value, making them relatively stable compared to the reference drugs and molecule 39 complexes (Figure 6). Notably, the four proposed compounds display lower Rg values than reference compounds, suggesting they are more compact.

Close

Figure 6.

Comparison of Radius of gyration profiles between the designed compounds, molecule 39, and the reference drug for each RTK target and PFKFB3.

Export to PPT

Finally, hydrogen bonds play a crucial role in determining binding affinity and significantly contribute to the interaction between ligands and proteins. The quality and positioning of hydrogen bonds are equally important, as they directly influence the strength and stability of the ligand-protein complex. Furthermore, a greater number of hydrogen bonds generally enhances the overall stability and binding affinity of the complex [77]. To confirm the stability of the docked complexes, we assessed the hydrogen bonds between each protein target and selected compounds, including the reference drug and molecule 39, in a solvent environment during MD simulations using the Gromacs gmx H-bond module (Supplementary File 3). The comparison of Pred3, Pred4, Pred34, and Pred49 with reference drugs and molecule 39 across 18 protein targets demonstrates consistent and improved interaction profiles, as evidenced by H-bond analysis. Across all target proteins, these compounds exhibit a comparable or higher number of H-bonds than the reference drugs, indicating enhanced binding affinity and stability within the binding pockets throughout the simulation. Pred3 and Pred4 consistently displayed the highest H-bond numbers among the designed compounds, particularly for key targets such as FGFR1, FGFR3, KIT, PFKFB3, and VEGFR2, which are critical for their inhibitory activity. Pred34 and Pred49 also demonstrated strong H-bond interactions, with numbers comparable to Pred3 and Pred4 for several targets, including FGFR2, EGFR, and IGFR. This consistent performance across multiple targets highlights the versatility and effectiveness of the designed compounds in forming stable ligand-protein complexes, reinforcing their potential as selective modulators of protein activity.

3.7. DFT analysis

DFT studies were conducted on the best-selected compounds, including molecule 39, with optimized ground-state geometries calculated using the DFT-B3LYP/6-31G++(d, p) basis set [78]. Molecular orbital energies explain a molecule’s electrical and optical properties. Frontier molecular orbitals (FMOs), particularly the HOMO and the LUMO, are essential in demonstrating the active sites of compounds during molecular interactions [79]. The HOMO value indicates the region with the greatest energy and highly reactive electrons, while LUMO identifies where the atoms are most likely to accept electrons [80].

The HOMO/LUMO diagrams for the selected compounds have been shown in Figure 7; the red and green regions correspond to the positive and negative phases of the molecular orbitals, respectively. Instead of being located over the same atomic sites, the FMOs lie in different sites of each compound, indicating their polar nature and the feasibility of intramolecular interaction with multiple sites in the targeted binding pocket. The Pred3, Pred4, Pred34, and Pred49 compounds have energy gaps of 3.80, 3.53, 3.80, and 4.80 eV, respectively. The energy gap between HOMO and LUMO reflects the compound’s stability and reactivity. As a result, less energy is needed to move electrons from the ground state HOMO to the excited state LUMO (Figure 7). Notably, Pred3 and Pred34, with an energy gap of 3.80, are comparable to molecule 39, suggesting similar reactivity and stability (Figure 7). However, Pred4, with a lower energy gap of 3.53, is expected to be more reactive, while Pred49, with a higher energy gap of 4.80, indicates lower reactivity and higher stability. Numerous studies have demonstrated that a smaller HOMO–LUMO gap usually indicates increased reactivity and a greater potential for chemical interactions with biological molecular targets [81].

Close

Figure 7.

FMOs (a) HOMO, LUMO, and HOMO-LUMO gap with corresponding energies, and density of states (DOS) spectra of (b) molecule 39, (c) Pred3, (d) Pred4, (e) Pred34, and (f) Pred49, respectively.

Export to PPT

In addition to FMO, the global descriptive parameters of the compounds were calculated, including global softness, hardness, ionization energy, and electron affinity (Supplementary File 2, Table S4). These values were obtained from HOMO and LUMO energy. Global softness and hardness are indicators of stability that increase with hardness and decrease with softness. The reactivity of the ligands decreases with hardness and increases with softness under study [82]. All the studied compounds were found to be stable within permissible limits. Among them, Pred3 and Pred34 exhibit a comparable value to that of molecule 39. Pred49 is the most stable but least reactive, and Pred4 is the most reactive but least stable. The same description applies to global hardness, the opposite of global softness. The ligands’ ionization potential (IP) and electron affinity (EA) were identified. IP refers to the amount of energy needed to remove an electron from an atom of a compound in its ground state. Chemically, a compound with a lower IP is generally more stable but less reactive, while a higher IP indicates less stability but greater reactivity [83]. Furthermore, EA, which indicates how likely an electron is to gain a negative charge when attached to a neutral atom, was also evaluated, according to stable, FDA-approved, or commercialized drugs, which typically exhibit IP values between 4 and 15 eV and EA values between -3 and 7 eV [84]. The proposed compounds fall within these ranges, indicating that they possess acceptable electronic properties, making them promising candidates for drug design.

3.8. The molecular electrostatic potential

Molecular electrostatic potential (MEP) analysis of compounds Pred3, Pred4, Pred34, Pred49, and molecule 39 offers important information about their chemical reactivity and potential for molecular interactions (Figure 8). The analyses show that the proposed compounds exhibit notable electrostatic characteristics, primarily influenced by the positive charge of nitrogen and the negative charge of the oxygen atoms, which play critical roles in their interaction potential.

Close

Figure 8.

MEP analyses of (a) Pred3, (b) Pred4, (c) Pred34, (d) Pred49, and (e) molecule 39.

Export to PPT

In the designed compounds Pred3, Pred4, Pred34, and Pred49, the nitrogen atoms consistently contribute to regions of negative electrostatic potential, indicating their nucleophilic properties and allowing enhanced interactions with protein targets. These nucleophilic regions facilitate strong binding by forming electrostatic interactions and hydrogen bonds with the protein’s positively charged or electrophilic regions, such as active site residues of the targets, such as FGFRs and VEGFR2 (Figures 8a-e). The oxygen atoms of the four compounds exhibit regions of negative electrostatic potential, emphasizing their role as hydrogen bond acceptors and contributing to their interaction potential. Notably, the MEP profiles of Pred3, Pred4, Pred34, and Pred49 are identical, indicating a common electrostatic potential distribution that enhances their binding affinity and supports their potential as selective modulators of protein activity. Molecule 39 shows a similar MEP profile with comparable negative potentials near the nitrogen atoms and positive potentials near the oxygen atoms, supporting its nucleophilic activity and ability to accept hydrogen bonds.

Finally, the results of the DFT and MEP analyses offer a comprehensive view of the electronic, energetic, and electrostatic properties of the designed compounds Pred3, Pred4, Pred34, Pred49, and molecule 39. The DFT analysis provides insights into total and binding energies, HOMO-LUMO gaps, and other key parameters. In addition, the MEP analysis identifies regions prone to nucleophilic and electrophilic interactions. Together, these findings thoroughly understand the compounds’ chemical behavior. The distinct electronic configurations and reactivity profiles highlight the interest of the designed compounds Pred3, Pred4, Pred34, and Pred49 in cancer treatment.