Lead generation of UPPS inhibitors targeting MRSA: Using 3D-QSAR pharmacophore modeling, virtual screening, molecular docking, and molecular dynamic simulations

Abstract

Undecaprenyl Pyrophosphate Synthase (UPPS) is a vital target enzyme in the early stages of bacterial cell wall biosynthesis. UPPS inhibitors have antibacterial activity against resistant strains such as MRSA and VRE. In this study, we used several consecutive computer-based protocols to identify novel UPPS inhibitors. The 3D QSAR pharmacophore model generation (HypoGen algorithm) protocol was used to generate a valid predictive pharmacophore model using a set of UPPS inhibitors with known reported activity. The developed model consists of four pharmacophoric features: one hydrogen bond acceptor, two hydrophobic, and one aromatic ring. It had a correlation coefficient of 0.86 and a null cost difference of 191.39, reflecting its high predictive power. Hypo1 was proven to be statistically significant using Fischer's randomization at a 95% confidence level. The validated pharmacophore model was used for the virtual screening of several databases. The resulting hits were filtered using SMART and Lipinski filters. The hits were docked into the binding site of the UPPS protein, affording 70 hits with higher docking affinities than the reference compound (6TC, - 21.17 kcal/mol). The top five hits were selected through extensive docking analysis and visual inspection based on docking affinities, fit values, and key residue interactions with the UPPS receptor. Moreover, molecular dynamic simulations of the top hits were performed to confirm the stability of the protein-ligand complexes, yielding five promising novel UPPS inhibitors.

Keywords: 3D QSAR pharmacophore; Antibacterial; Bacterial resistance; Drug repurposing; Dynamic simulations; HypoGen algorithm; Methicillin-resistant Staphylococcus aureus; Molecular docking; UPPS.

Fig. 1

Biosynthetic pathway of peptidoglycan bacterial cell wall including sites of action of UPPS inhibitors, methicillin, and vancomycin

Fig. 2

Training set ligands along with their IC₅₀ values ranging from 0.04 to 35 μM

Fig. 3

Test set ligands along with their IC₅₀ values ranging from 0.5 μM to 58 μM

Fig. 4

Spatial arrangement of the valid pharmacophore model with the distances and angles displayed

Fig. 5

The difference in total cost values of hypotheses between a Hypo1 spreadsheet and 19 random spreadsheets

Fig. 6

The difference in correlation values of hypotheses between a Hypo1 spreadsheet and 19 random spreadsheets

Fig. 7

A. The 2D structure of the reference drug with reported interactions highlighted together with the chemical features. B. The mapping of the reference drug into HYPO1 with a fit value of 7.57 and an estimated IC₅₀ of 0.018 µM

Fig. 8

Schematic representation of the virtual screening process used to identify potential hits

Fig. 9

The 2D and 3D representation of the docking of the reference ligand (6TC) inside the binding site of UPPS (PDB ID: 5KH5)

Fig. 10

Root Mean Square Deviations (RMSD) of protein 5KH5 alone, the docked reference ( 6TC) and the selected top five hits

Fig. 11

Radius of Gyration of protein 5KH5 alone, the docked reference 6TC and the selected top five hits

Fig. 12

Root mean square deviation (RMSF) of protein 5KH5 alone, the docked reference 6TC and the selected top five hits