Molecular Docking and Pharmacokinetic Evaluation of Novel 1-(Substituted Phenyl)-N-(2-(2-(Pyridin-2-yloxy) ethoxy) ethyl) Methanimine Derivatives as Potential ?-Glucosidase Inhibitors

2.2.2 Protein Preparation

The three-dimensional crystal structure of the target protein α-glucosidase (PDB ID: 3A4A) was retrieved from the RCSB Protein Data Bank. The selected protein structure contained the protein backbone, with the water molecules, co-crystallized ligands, and other heteroatoms deleted. The atoms of the hydrogen were inserted, and the protein structure was optimized for good geometry for further molecular docking studies [15].

2.2.3 Molecular Docking

Molecular docking analysis was carried out using ArgusLab 4.0. It was concluded that the prepared protein is the receptor protein, while the optimized ligands are the input molecules. It was picked by using the active site residues or the position of the co-crystallized molecule in the binding site. Docking was done with the algorithm, and numerous poses were generated for each of the ligands [16]. Binding energy scores and interaction profiles were used to assess the docked conformations. The binding pose of each ligand was determined based on the lowest binding energy and the favorable interaction with the amino acids, including hydrogen bonds, hydrophobic interactions, and van der Waals forces.

2.2.4 Molecular visualization

Three-dimensional (3D) visualization of docked ligand-protein complexes was done using Discovery Studio Visualizer (BIOVIA). 3D images of the docked complexes were generated by the software, and the orientation of the ligands in the active binding pocket of the model protein was analyzed [17].

2.3 In Silico ADMET and Drug Likeness Prediction

Pharmacokinetic properties and drug-likeness profiles of the designed ligands (M1-M24) were analyzed utilizing online computational tools. The chemical structure of selected designed compounds was drawn using the software ChemDraw and converted to canonical SMILES format, entered as input for the ADME analysis. The Absorption, Distribution, Metabolism, and Excretion (ADME) properties of the ligands were predicted with the aid of the SwissADME web server. Physicochemical and pharmacokinetic properties such as molecular weight, LogP, aqueous solubility, gastrointestinal (GI) absorption, blood-brain barrier (BBB) permeation, bioavailability scores, and synthetic accessibility were investigated for the evaluation of drug-likeness of the compounds. The toxicity profiles of the designed ligands were predicted by using ProTox-II. The SMILES of every compound were submitted to determine such parameters as LD50 values and toxicity classes. The drug likeness of the compounds was evaluated using Molsoft. Drug-likeness scores and biological activity predictions were then performed based on the structural features using the Molsoft platform, giving insight into the potential interaction of the compounds with biological factors [18].

3. RESULTS AND DISCUSSION

3.1 Molecular Docking Study

Molecular docking was subsequently carried out to understand the binding affinity of the designed 1-(substituted phenyl)-N-(2-(2-(pyridin-2-yloxy) ethoxy) ethyl) methanimine derivatives (M1–M24) in the active site of α-glucosidase. Docking results are summarized and displayed as a table including the binding affinity, hydrogen bond interactions, bond length, and interacting amino acid residues (Table 1). The clinically-used α-glucosidase inhibitor, Miglitol (-7.55 kcal/mol), docked in this receptor and made interactions through six hydrogen bonds with the following residues: ILE272, SER298, GLU296, ASN259, HIS295, and ALA295, which are present in the receptor. All of the designed derivatives showed a docking score from -8.30 to -10.96 kcal/mol, with improved docking scores as compared to Miglitol. Among the evaluated compounds, M1 exhibited the highest binding affinity with a docking score of -10.96 kcal/mol, followed by M14 (-10.43 kcal/mol), M6 (-10.40 kcal/mol), M22 (-10.20 kcal/mol), M9 (-10.06 kcal/mol), M2 (-10.03 kcal/mol), M10 (-10.02 kcal/mol), and M24 (-10.01 kcal/mol). The obtained data indicate that the designed methanimine derivatives exhibit very good properties for binding into the catalytic pocket of α-glucosidase.

Analysis of hydrogen bonding interactions indicated that several polar interactions with important residues at the active site of the docked complexes were responsible for the stability of the complexes. The five H-bond interactions between compound M1 and GLN279, ARG315, ARG442, and ASN415, and the six H-bond interactions between compound M22 and HIS112, GLN182, ARG442, and ASP69 resulted in very stable complexes with the enzyme, suggesting that these would be excellent candidates for further development. The same was obtained for M9, as it interacted with five hydrogen bonds with the residues TYR347, ARG442, ASP69, and THR306, while M14 interacted through four hydrogen bonds with the residues ARG442, ARG213, GLN353, and THR306. Compound M12 was also observed to have six hydrogen bonds with GLU405, TYR416, and ASN417, but its docking score of (-9.03 kcal/mol) was not as high as the top-ranked compounds, indicating that binding affinity is dependent on both hydrogen bonding and hydrophobic interactions.

The hydrogen bond distances observed ranged from 2.22 to 3.00 Å, which were held to be strong and energetically favorable interactions between the active site of the enzyme and the interacting ligands. Several amino acid residues like ARG442, GLN279, THR306, ASP69, GLN353, ARG315, TYR347, and ASN415 repeatedly participated in the interactions with the ligand and were observed to play a crucial role in molecular recognition and stabilization of the docked complexes. The ARG442 residue was present in the highest number of residues interacting with the majority of high-scoring derivatives, indicating that ARG442 plays a vital role in the binding of the ligand in the active pocket of α-glucosidase. Interestingly, compound M8 yielded a docking score of -9.80 kcal/mol and did not showcase any hydrogen-bond interaction, suggesting that it primarily interacts by hydrophobic, van der Waals, and π-related interactions. The observations confirm that hydrogen bonding is not the only factor involved in binding affinity and that non-covalent complementary interactions play a major role in the overall stability of the protein–ligand association.

The overall docking results suggest that the designed 1-(substituted phenyl)-N-(2-(2-(pyridin-2-yloxy) ethoxy) ethyl) methanimine derivatives have very good binding affinity towards α-glucosidase with a higher dock score compared to the reference compound Miglitol. The compounds M1, M14, M6, M22, M9, M2, M10, and M24 were found to be the most promising candidates, exhibiting promising binding energies and interactions with the key catalytic residues. The three-dimensional molecular visualization of the docked protein-ligand complexes, as shown in Figure 1, shows the binding orientation of the selected ligands in the active site of α-glucosidase, emphasizing their good orientation within the catalytic pocket. The complexes show that the stabilizing interactions of the ligands involve mostly hydrogen bonds, hydrophobic contacts, π-π stacking, and van der Waals interactions between the ligand and key residues in the active site. These interactions facilitate the excellent docking scores found for the lead compounds and reveal structural details of how they might inhibit α-glucosidase. Based on these results, the authors believe that the derivatives are promising candidates for further studies by molecular dynamics simulation, chemical synthesis, and biological testing in order to confirm their α-glucosidase inhibitory bioactivity as possible antidiabetic agents.

Table 1. Molecular docking simulation of the novel designed 1-(substituted phenyl)-N-(2-(2-(pyridin-2-yloxy) ethoxy) ethyl) methanimine derivatives.

Figure 1. 3D Three-dimensional molecular docking visualization of the protein-ligand complexes of Miglitol and the potent designed derivatives bound to the active site of α-glucosidase (PDB ID: 3A4A).

3.2 In Silico Pharmacokinetic Study and Drug-Likeness Evaluation

The pharmacokinetic characteristics and drug-likeness of the designed derivatives (M1–M24) were evaluated using in silico ADMET prediction tools to assess their suitability as potential drug candidates. All the calculated physicochemical and pharmacokinetic parameters (molecular weight (MW), number of heavy atoms (nHA), number of aromatic heavy atoms (nAHA), number of rotatable bonds (nRB), number of hydrogen bond acceptor (nHBA), gastrointestinal absorption (GIT), blood–brain barrier (BBB) permeability, number of hydrogen bond donor (nHBD), drug-likeness score, and predicted acute oral toxicity (LD50)) are summarized in Table 2.

Molecular weights of designed derivatives were in the range of 270.33 – 430.67 g/mol, which is below the generally accepted upper limit of 500 g/mol advised by Lipinski's rule of five. The M1 had the lowest molecular weight of 270.33 g/mol, while the M7 had the highest molecular weight of 430.67 g/mol. Counties of 20 to 25 heavy atoms were observed between the derivatives, and the heavy atoms in the aromatic ring were levied at 12 atoms for each derivative, thus maintaining a common framework for the aromatic system throughout the designed series. The number of rotatable bonds of structural flexibility was varied between eight and 10, indicating moderate flexibility of the molecule, which is desirable for good receptor binding and not causing molecular instability. Hydrogen bond acceptors varied from 4 to 7, while hydrogen bond donors showed ranges between 0 and 2, which supports all the derivatives having suitable properties for favorable oral bioavailability. Twenty-four derivatives were predicted with gastrointestinal absorption, and it was found that all of them have high GIT absorption, suggesting their potential to be highly absorbed by the oral route and thus negligible side effects in the nervous system. system targets or were avoided for peripherally acting drugs to avoid neurological side effects. Predicted drug/likeness scores ranged from -0.59 to 0.45. The results tended to be more positive for the drug likeness for most of the compounds, suggesting a good physicochemical property for drug development. The compound with the highest drug-likeness score was M7 (0.45), followed by M13 (0.43) and M9 (0.42), indicating an optimal combination of molecular features and pharmacokinetic properties. However, compounds M5 (-0.59), M11 (-0.24), M1 (-0.14), M19 (-0.07), and M8 (-0.03) had slightly negative scores and may also not preclude further consideration for biological evaluation, especially in the presence of encouraging biological activity. Acute oral toxicity prediction revealed good safety profiles of the molecules designed. The acute toxicity estimates ranged from 1250 to 2000 mg/kg estimated LD₅₀, indicating relatively low toxicity.

The highest predicted LD50 values were obtained for compounds M3, M9, M11, M12, M14-M17, and M23, with 2000 mg/kg, which are comparatively less toxic than derivatives with 1250 mg/kg predicted LD₅₀ values. In conclusion, the in silico pharmacokinetic evaluation showed that the engineered derivatives have drug-like attributes like low toxicity in general, good gastrointestinal absorption, acceptable toxicity, molecular flexibility, adequate hydrogen-bonding ability, and acceptable molecular weight. Most compounds also presented good drug-likeness scores, while showing BBB permeability, suggesting they could be orally active lead compounds. M7, M9, M12, and M13 showed a very well-balanced pharmacokinetic profile – these derivatives absorbed best into the gut, had good BBB permeability, good drug-likeness scores, and acceptable toxicity predictions. All this, along with their biological activity and molecular docking results confirm that these compounds could be used for experimental investigation of pharmacological aspects in the further steps of their optimization.

MW: Molecular weight, nHA: No. of heavy atom, nAHA: No. of Aromatic heavy atoms, nRB: No. of Rotatable bonds nHBA:  No of H-bond acceptors, nHBD: No. of H-bond donor, LD: Lethal dose.

Table 2. In silico ADMET and physicochemical profiles of the designed 1-(substituted phenyl)-N-(2-(2-(pyridin-2-yloxy) ethoxy) ethyl) methanimine derivatives.