In silico identification of five binding sites on the SARS-CoV-2 spike protein and selection of seven ligands for such sites

Abstract

To contribute to the development of products capable of complexing with the SARS-CoV-2 spike protein, and thus preventing the virus from entering the host cell, this work aimed at discovering binding sites in the whole protein structure, as well as selecting substances capable of binding efficiently to such sites. Initially, the three-dimensional structure of the protein, with all receptor binding domains in the closed state, underwent blind docking with 38 substances potentially capable of binding to this protein according to the literature. This allowed the identification of five binding sites. Then, those substances with more affinities for these sites underwent pharmacophoric search in the ZINC15 database. The 14,329 substances selected from ZINC15 were subjected to docking to the five selected sites of the spike protein. The ligands with more affinities for the protein sites, as well as the selected sites themselves, were used in the de novo design of new ligands that were also docked to the binding sites of the protein. The best ligands, regardless of their origins, were used to form complexes with the spike protein, which were subsequently used in molecular dynamics simulations and calculations of ligands affinities to the protein through the molecular mechanics/Poisson–Boltzmann surface area method (MMPBSA). Seven substances with good affinities to the spike protein (-12.9 to −20.6 kcal/mol), satisfactory druggability (Bioavailability score: 0.17 to 0.55), and low acute toxicity to mice (LD₅₀: 751 to 1421 mg/kg) were selected as potentially useful for the future development of new products to manage COVID-19 infections.

Communicated by Ramaswamy H. Sarma

Keywords:

• Spike protein
• SARS-CoV-2
• molecular docking
• molecular dynamics simulations
• MMPBSA
• ADMET
• COVID-19

Acknowledgements

This research was developed with the help of Centro Nacional de Processamento de Alto Desempenho em São Paulo (CENAPAD-SP; https://www.cenapad.unicamp.br/), to which the author is grateful.

Author contributions

Denilson Ferreira de Oliveira carried out all the work.

Disclosure statement

No potential conflict of interest was reported by the author.

Availability of data and materials

All data generated or analyzed during this study are included in this article and its supplementary materials.

Additional information

Notes on contributors

Denilson Ferreira de Oliveira

Denilson Ferreira de Oliveira carried out all the work.