Potency Prediction\nof Covalent Inhibitors against\nSARS-CoV‑2 3CL-like Protease and Multiple Mutants by Multiscale\nSimulations

3-Chymotrypsin-like protease (3CL<sup>pro</sup>) is a\nprominent\ntarget against pathogenic coronaviruses. Expert knowledge of the cysteine-targeted\ncovalent reaction mechanism is crucial to predict the inhibitory potency\nof approved inhibitors against 3CL<sup>pro</sup>s of SARS-CoV-2 variants\nand perform structure-based drug design against newly emerging coronaviruses.\nWe carried out an extensive array of classical and hybrid QM/MM molecular\ndynamics simulations to explore covalent inhibition mechanisms of\nfive well-characterized inhibitors toward SARS-CoV-2 3CL<sup>pro</sup> and its mutants. The calculated binding affinity and reactivity\nof the inhibitors are highly consistent with experimental data, and\nthe predicted inhibitory potency of the inhibitors against 3CL<sup>pro</sup> with L167F, E166V, or T21I/E166V mutant is in full agreement\nwith IC<sub>50</sub>s determined by the accompanying enzymatic assays.\nThe explored mechanisms unveil the impact of residue mutagenesis on\nstructural dynamics that communicates to change not only noncovalent\nbinding strength but also covalent reaction free energy. Such a change\nis inhibitor dependent, corresponding to varied levels of drug resistance\nof these 3CL<sup>pro</sup> mutants against nirmatrelvir and simnotrelvir\nand no resistance to the <b>11a</b> compound. These results\ntogether suggest that the present simulations with a suitable protocol\ncan efficiently evaluate the reactivity and potency of covalent inhibitors\nalong with the elucidated molecular mechanisms of covalent inhibition.