Testing Affordable Strategies for the Computational Study of Reactivity in Cysteine Proteases: The Case of SARS-CoV-2 3CL Protease Inhibition

Abstract

Cysteine proteases are an important target for the development of inhibitors that could be used as drugs to regulate the activity of these kinds of enzymes involved in many diseases, including COVID-19. For this reason, it is important to have methodological tools that allow a detailed study of their activity and inhibition, combining computational efficiency and accuracy. We here explore the performance of different quantum mechanics/molecular mechanics methods to explore the inhibition reaction mechanism of the SARS-CoV-2 3CL protease with a hydroxymethyl ketone derivative. We selected two density functional theory (DFT) functionals (B3LYP and M06-2X), two semiempirical Hamiltonians (AM1d and PM6), and two tight-binding DFT methods (DFTB3 and GFN2-xTB) to explore the free energy landscape associated with this reaction. We show that it is possible to obtain an accurate description combining molecular dynamics simulations performed using tight-binding DFT methods and single-point energy corrections at a higher QM description. The use of a computational strategy that provides reliable results at a reasonable computational cost could assist the in silico screening of possible candidates during the design of new drugs directed against cysteine proteases.