A Multistage In Silico Study of Natural Potential Inhibitors Targeting SARS-CoV-2 Main Protease

Abstract

Among a group of 310 natural antiviral natural metabolites, our team identified three compounds as the most potent natural inhibitors against the SARS-CoV-2 main protease (PDB ID: 5R84), M^pro. The identified compounds are sattazolin and caprolactin A and B. A validated multistage in silico study was conducted using several techniques. First, the molecular structures of the selected metabolites were compared with that of GWS, the co-crystallized ligand of M^pro, in a structural similarity study. The aim of this study was to determine the thirty most similar metabolites (10%) that may bind to the M^pro similar to GWS. Then, molecular docking against M^pro and pharmacophore studies led to the choice of five metabolites that exhibited good binding modes against the M^pro and good fit values against the generated pharmacophore model. Among them, three metabolites were chosen according to ADMET studies. The most promising M^pro inhibitor was determined by toxicity and DFT studies to be caprolactin A (292). Finally, molecular dynamics (MD) simulation studies were performed for caprolactin A to confirm the obtained results and understand the thermodynamic characteristics of the binding. It is hoped that the accomplished results could represent a positive step in the battle against COVID-19 through further in vitro and in vivo studies on the selected compounds.

Keywords: ADMET; DFT; MD simulations; SARS-CoV-2; docking; main protease; pharmacophoric; structural similarity.

Scheme 1

In silico filtration protocol.

Figure 1

Structural similarity between GWS and the examined metabolites. Green = GWS, red = similar metabolites, blue = dissimilar metabolites. (A) Metabolites 1–50; (B) metabolites 51–100l; (C) metabolites 101–150; (D) metabolites 151–200; (E) metabolites 201–250; (F) metabolites 251–310.

Figure 1

Structural similarity between GWS and the examined metabolites. Green = GWS, red = similar metabolites, blue = dissimilar metabolites. (A) Metabolites 1–50; (B) metabolites 51–100l; (C) metabolites 101–150; (D) metabolites 151–200; (E) metabolites 201–250; (F) metabolites 251–310.

Figure 2

Metabolites with molecular similarity with GWS.

Figure 3

(A) 2D interaction of the tested compounds in the active site of M^pro; (A) GWS, (B) compound 291, (C) compound 292, (D) compound 293.

Figure 4

(A) Generated 3D-pharmacophore geometry with three features: one H-bond donor (pink color) and two hydrophobic centers (blue). (B) Mapping of the cocrystallized ligand on the generated pharmacophore (fit value = 2.804).

Figure 5

Mapping of the tested metabolites on the generated pharmacophore. (A) Metabolite 15 (fit value = 1.982), (B) metabolite 30 (fit value = 2.465), (C) metabolite 55 (fit value = 2.285), (D) metabolite 109 (fit value = 2.552), (E) metabolite 112 (fit value = 2.471), and (F) metabolite 234 (fit value = 1.418).

Figure 6

Mapping of the tested metabolites on the generated pharmacophore. (A) metabolite 236 (fit value = 0.523), (B) metabolite 291 (fit value 2.887), (C) metabolite 292 (fit value = 2.890), (D) metabolite 293 (fit value = 2.587), (E) metabolite 303 (fit value = 2.867), and (F) metabolite 305 (fit value = 2.792).

Figure 7

Expected ADMET study.

Figure 8

Spatial distribution of molecular orbitals for (A) GWS, as well as metabolites (B) 292 and (C) 293.

Figure 9

Molecular electrostatic potential map of (A) GWS, (B) metabolite 292, and (C) metabolite 293.

Figure 9

Molecular electrostatic potential map of (A) GWS, (B) metabolite 292, and (C) metabolite 293.

Figure 10

Molecular dynamics simulations of caprolactin A-M^pro complex. (A) RMSD analysis of caprolactin A bindings with M^pro. (B) RMSF plot of VGL residues. (C) Radius of gyration for M^pro within 150 ns.

Figure 10

Molecular dynamics simulations of caprolactin A-M^pro complex. (A) RMSD analysis of caprolactin A bindings with M^pro. (B) RMSF plot of VGL residues. (C) Radius of gyration for M^pro within 150 ns.