Shedding Light on Dark Chemical Matter: The Discovery of a SARS-CoV-2 Mpro Main Protease Inhibitor through Intensive Virtual Screening and In Vitro Evaluation.

The development of specific antiviral therapies targeting SARS-CoV-2 remains fundamental because of the continued high incidence of COVID-19 and limited accessibility to antivirals in some countries. In this context, dark chemical matter (DCM), a set of drug-like compounds with outstanding selectivity profiles that have never shown bioactivity despite being extensively assayed, appears to be an excellent starting point for drug development. Accordingly, in this study, we performed a high-throughput screening to identify inhibitors of the SARS-CoV-2 main protease (Mpro) using DCM compounds as ligands. Multiple receptors and two different docking scoring functions were employed to identify the best molecular docking poses. The selected structures were subjected to extensive conventional and Gaussian accelerated molecular dynamics. From the results, four compounds with the best molecular behavior and binding energy were selected for experimental testing, one of which presented inhibitory activity with a Ki value of 48 ± 5 µM. Through virtual screening, we identified a significant starting point for drug development, shedding new light on DCM compounds.

Fragment dissolved molecular dynamics: a systematic and efficient method to locate binding sites.

Diverse computational methods to support fragment-based drug discovery (FBDD) are available in the literature. Despite their demonstrated efficacy in supporting FBDD campaigns, they exhibit some drawbacks such as protein denaturation or ligand aggregation that have not yet been clearly overcome in the framework of biomolecular simulations. In the present work, we discuss a systematic semi-automatic novel computational procedure, designed to surpass these difficulties. The method, named fragment dissolved Molecular Dynamics (fdMD), utilizes simulation boxes of solvated small fragments, adding a repulsive Lennard-Jones potential term to avoid aggregation, which can be easily used to solvate the targets of interest. This method has the advantage of solvating the target with a low number of ligands, thus preventing the denaturation of the target, while simultaneously generating a database of ligand-solvated boxes that can be used in further studies. A number of scripts are made available to analyze the results and obtain the descriptors proposed as a means to trustfully discard spurious binding sites. To test our method, four test cases of different complexity have been solvated with ligand boxes and four molecular dynamics runs of 200 ns length have been run for each system, which have been extended up to 1 µs when needed. The reported results point out that the selected number of replicas are enough to identify the correct binding sites irrespective of the initial structure, even in the case of proteins having several close binding sites for the same ligand. We also propose a set of descriptors to analyze the results, among which the average MMGBSA and the average KDEEP energies have emerged as the most robust ones.

A multistep docking and scoring protocol for congeneric series: Implementation on kinase DFG-out type II inhibitors.

AIM: Rescoring of docking-binding poses can significantly improve molecular docking results. Our aim was to evaluate postprocessing docking protocols in order to determine the most suitable methodology for the study of the binding of congeneric compounds to protein kinases. MATERIALS & METHODS: Diverse ligand-receptor poses generated after docking were submitted to different relaxation protocols. The Molecular Mechanics Poisson-Boltzmann (Generalized Born) Surface Area approach was applied for the evaluation of the binding affinity of complexes obtained. The performance of various Molecular Mechanics Poisson-Boltzmann (Generalized Born) Surface Area methodologies was compared. RESULTS: The inclusion of a postprocessing protocol after docking enhances the quality of the results, although the best methodology is system dependent. CONCLUSION: An examination of the interactions established has allowed us to suggest useful modifications for the design of new type II inhibitors.