Target protein retrieval and preparation

The crystal structure of the PTP1B (PDB ID: 2CNG), refined at a 1.90 Å resolution, was downloaded from the Protein Data Bank. Following that, the co-crystal ligand, IZE, also known as BDBM13474, which is a PTP1B inhibitor, was removed from the protein, and the clean protein structure was minimized using the discovery studio visualizer 2020 and saved in .pdb format.

Natural compound library preparation and virtual screening

The ZINC database was used to obtain a collection of 2700 naturally occurring compounds in ‘.sdf’ file format. These compounds' structures were minimized by employing a Universal Force Field (UFF) and prepared for further VS. VS is a drug development computer approach used to identify prospective compounds with promising pharmacological activity from huge compound libraries. It applies drug design theoretical principles as well as computer technology and specific software applications [13]. The major goal is to systematically filter novel lead chemicals from large molecular databases. Using the PyRx 0.8 tool [14], the prepared compound library was screened against the PTP1B.

Physiochemical and ADMET properties

The physicochemical and drug-likeness characteristics of the top five screened compounds were investigated using the SwissADME web tool. With the help of this efficient tool, it was possible to thoroughly assess and estimate the different properties connected to these compounds, which gave important details about their chemistry and potential as treatment candidates [15].

Results

Virtual screening of natural compounds

To identify a potential natural inhibitor of PTP1B, a VS methodology was utilized. A library of 2700 natural compounds, obtained from the ZINC database, was subjected to this screening process. Here in this study, co-crystal ligand (BDBM13474), and two well-known PTP1B inhibitors, namely Erythribyssin A [16] and Formylchromone [17] were used as positive controls. The XYZ coordinates, specifically 2.124500, 9.759750, and 46.315444 were derived from the co-crystal ligand.

Targeting the active site residues of PTP1B, 2700 natural compounds were computationally screened, and several compounds with higher binding energies than the positive controls used in this study were identified as possible candidates. The top 15 compounds that interacted with important PTP1B residues and had binding energies greater than the control are listed in Table 1.

The binding energies of all 15 compounds were found to be lower than those of the control compounds, indicating a stronger affinity for the target protein PTP1B. Furthermore, these compounds exhibited interactions with a significant number of the key residues of PTP1B. Following a thorough analysis and visualization of the interactions between the docked complexes, the top five hits (Figure 1) in this study, which demonstrated more effective binding by interacting with crucial PTP1B residues, are thoroughly discussed.

The co-crystal ligand (BDBM13474), Erythribyssin A, and Formylchromone, which were used as positive controls, have been shown to interact with key PTP1B residues like (Asp48, Ile219, Gln262, Gly220, Arg221, Cys215, Phe182, Ser216, Asp181, and Tyr46) (Figure 2).

The analysis focused on the interaction profiles of the top 5 hits, specifically ZINC899884, ZINC56981, ZINC252509722, ZINC1843029, and ZINC21789 within the active site residues of PTP1B. This analysis included both 3D interactions (Figure 3) and 2D interactions (Figure 4).

ZINC899884 interacted with Asp48, Phe182, Ala217, Gln262, Ile219, Gly220, Gly218, Ser216, Cys215, Arg221, Asp181, Tyr46, Lys120, and Arg47 residues of PTP1B protein. Gly220, Ser216, Arg221, and Tyr46 residues were involved in H-bonding with ZINC899884 (Figure 4A). ZINC56981 interacted with Gly218, Cys215, Ser216, Arg221, Asp181, Gln266, Phe182, Ala217, Val49, Gln262, Asp48, Tyr46, Ile219, and Gly220 residues of PTP1B protein. Among them, Gly218, Ser216, Arg221, and Gln262 H-bonded with ZINC56981 (Figure 4B). ZINC252509722 binds with the Glu115, Asp181, Lys120, Tyr46, Phe182, Gly220, Ser216, Ala217, Ile219, Arg221, Cys215, Gly218, Val49, Gln262, and Asp48 residues of PTP1B protein. Gln262, Ala217, Ser216, Tyr46, Lys120, and Asp181 residues were involved in H-bonding with ZINC252509722 (Figure 4C). ZINC1843029 was found to interact with Ile219, Val49, Gln262, Ser216, Gly218, Ala217, Arg221, Cys215, Asp181, Gln266, Gly183, Phe182, and Gly220 residues of PTP1B protein. Gln266, Phe182, Ala217, and Ser216 residues make the H-bond with ZINC1843029 (Figure 4D). Further, ZINC21789 interacted with Tyr46, Val49, Gln262, Phe182, Ile219, Gln266, Asp181, Gly220, Cys215, Arg221, Gly218, Ser216, Ala217, and Lys120 residues of PTP1B protein. Among them, Gln262, Gly220, Arg221, Ser216, and Ala217 residues were H-bonded with ZINC21789 (Figure 4E).   

Physicochemical, pharmacokinetics and drug likeness assessment

The colored zone represents the physiochemical space that is deemed suitable for achieving oral bioavailability. Compounds possessing physiochemical properties falling within this zone demonstrate oral activity. Figure 5 illustrates the physiochemical properties as bioavailability radar of the five selected hits. ZINC252509722 and ZINC21789 exhibited minimal deviation from the designated pink region, within which the molecular radar plot must entirely reside in order to be classified as drug-like (Figure 5).

The lipophilicity of all five selected molecules is below 5, suggesting that these molecules have the potential to be investigated as orally active compounds. Table 2 provides a summary of the physicochemical properties and various lipophilicity scores.

Water-soluble molecules are highly advantageous in numerous drug development endeavors, primarily due to their convenient handling and formulation properties. The selected 5 compounds exhibited favorable water solubility across all parameter classes, namely ESOL, Ali, and Silicos-IT. The pharmacokinetic properties of the selected compounds were also predicted. The study revealed that the gastrointestinal absorption rates of ZINC899884, ZINC56981, and ZINC21789 were observed to be significantly high, whereas ZINC252509722 and ZINC1843029 exhibited notably low absorption rates. These findings suggest that ZINC899884, ZINC56981, and ZINC21789 are characterized by a favorable ability to be absorbed efficiently from the intestinal tract. The compounds ZINC899884 and ZINC56981 have demonstrated the ability to cross the blood-brain barrier (BBB). Further pharmacokinetic characteristics of the compounds, including their capacity to inhibit cytochrome P450 enzymes (CYPs), suggest that they do not exhibit any interaction with any specific isoform of cytochrome P450. This implies that these isoforms do not play a role in the biotransformation process of the selected molecule (Table 3).

Lipinski's rule, which meets the requirements of an ideal drug molecule, is the generally recognized physicochemical property guideline. The selected molecules' bioavailability Score was 0.55, except ZINC252509722 (0.11), which violated one parameter of the Lipinsky rule, indicating that these molecules can be synthesized and advanced to the next stage of the drug development pipeline. The findings indicate that all of the compounds exhibit a favorable drug-likeness score, demonstrating minimal or no violations of the drug-likeness criteria under investigation (Table 4).

Figures & Tables

Discussion