Support Vector Machine Identification of Small Molecule Binders to an Understudied Allosteric Site of SARS-CoV-2 Mpro for Next-Generation PROTAC-Based Therapeutics #MMPMID41388974
Fassi EMA; Mekni N; Albani M; Maehrlein S; Weldert AC; Schirmeister T; Langer T; Razioso G
Arch Pharm (Weinheim) 2025[Dec]; 358 (12): e70169 PMID41388974show ga
The emergence of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has underscored the urgent need for novel antiviral strategies. One of the primary targets of interest is the SARS-CoV-2 main protease (Mpro), which plays a crucial role in viral replication. Building on our prior work involving machine learning (ML)-based virtual screening for potential Mpro inhibitors, we sought to experimentally validate top-ranked candidates. Microscale thermophoresis (MST) was used to assess the binding affinity, leading to the identification of three promising hits from a library of 180 compounds. Notably, one compound demonstrated high-affinity binding to SARS-CoV-2 Mpro (K(d) = 2.8 +/- 0.9 microM). However, enzymatic assays revealed that none of the hit compounds inhibited the activity of the protease, suggesting a non-competitive binding. Docking and molecular dynamics (MD) simulations allowed to identify an accessory site in which the compounds exhibited stable interactions. These findings suggest that the identified compounds may serve as a starting point for the rational design of degradation-inducing strategies, such as proteolysis-targeting chimeras (PROTACs), targeting SARS-CoV-2 Mpro, and highlight the value of integrating ML-driven discovery with biophysical and computational validation in antiviral drug development.
Arch+Pharm+(Weinheim) 2025 ; 358 (12): e70169
