Combinatorial chemistry-driven In silico design and computational evaluation of covalent peptidomimetic SARS-CoV-2 main protease inhibitors via structure-based virtual screening and multivariate analysis

The COVID-19 pandemic has underscored the urgent need for specific pharmacological treatments beyond existing vaccines. One of the most attractive targets for antiviral therapies development is the SARS-CoV-2 Main Protease (M^PRO), a key enzyme in viral life. The lack of M^PRO human homologs and its conservation rate among coronaviruses make this enzyme strategically important. Considering its mechanism of action, the catalytic cysteine residue (Cys¹⁴⁵) presents a prime target for covalent inhibition. Electrophilic warhead inhibitors, designed to react with this catalytic site, mimic the amide peptide bonds of the viral polyproteins, thereby facilitating their binding and subsequent inactivation of the enzyme. Their activity is further potentiated when incorporated into peptidomimetic structures. In silico approaches are gaining increasing importance in the search for effective COVID-19 treatments. In this view, this study focuses on developing an innovative in silico protocol for identifying anti-SARS-CoV-2 agents covalently targeting M^PRO. To this purpose, a combinatorial library of 450 peptidomimetic compounds with aldehydic warheads was generated, refined to 388 compounds through docking studies, and further evaluated for covalent binding capabilities, revealing that compounds 9–14, 16, and 20 exhibited significantly higher affinities compared to known inhibitors, even during a 200 ns dynamics simulation, thus affirming the validity of the adopted design strategy. Furthermore, this work takes the advantages of our in-house ligand-based tool, the Biotarget Predictor Tool, available in DRUDIT, integrated with both structure-based techniques and, interestingly, multivariate statistical analysis.