Virtual Screening Identifies Inhibitors of SARS-CoV-2 Main Protease through Pharmacophore and Similarity Approaches.

Abstract

Introduction: The emergence of SARS-CoV-2 and the COVID-19 pandemic highlighted the urgent need for novel antiviral therapies. The main protease (M^pro) of SARS-CoV-2 is a key enzyme in viral replication and a promising therapeutic target.

Methods: This study employed virtual screening approaches to identify potential M^pro inhibitors, leveraging both structure- and ligand-based methods.

Results: Two optimum pharmacophore models were built from hundreds of crystallographic structures of M^pro, validated through ROC curve analysis and Dynophores dynamic simulations. These models captured ≈ 60K hits from six diverse compound libraries made of more than 3 million compounds. Additionally, a ligandbased similarity search using ROCS software identified 1024 potential hits based on shape and atom-based comparisons with co-crystallized ligands. Subsequent molecular docking and filtering based on physicochemical properties and structural diversity yielded 16 and 6 hits from structure- and ligand-based screening, respectively. Molecular dynamics simulations were conducted on the top-scoring hits to assess their binding stability within the M^pro active site. SCR00943 demonstrated stable binding, interacting favorably with key residues, including the catalytic dyad, resulting in a binding affinity of -61.2 kcal/mol.

Conclusion: This virtual screening campaign identified promising M^pro inhibitors, showcasing the potential of computational approaches to accelerate drug discovery efforts against COVID-19.