ABSTRACT
Small molecules that bind the SARS-CoV-2 nonstructural protein 3 Mac1 domain in place of ADP-ribose could be useful as molecular probes or scaffolds for COVID-19 antiviral drug discovery because Mac1 has been linked to the ability of coronaviruses to evade cellular detection. A high-throughput assay based on differential scanning fluorimetry (DSF) was therefore optimized and used to identify possible Mac1 ligands in small libraries of drugs and drug-like compounds. Numerous promising compounds included nucleotides, steroids, ß-lactams, and benzimidazoles. The main drawback to this approach was that a high percentage of compounds in some libraries were found to influence the observed Mac1 melting temperature. To prioritize DSF screening hits, the shapes of the observed melting curves and initial assay fluorescence were examined, and the results were compared with virtual screens performed using AutoDock Vina. The molecular basis for alternate ligand binding was also examined by determining a structure of one of the hits, cyclic adenosine monophosphate, with atomic resolution.
Subject(s)
Coronavirus Papain-Like Proteases/chemistry , Coronavirus Papain-Like Proteases/metabolism , High-Throughput Screening Assays/methods , SARS-CoV-2/chemistry , Antiviral Agents/chemistry , Antiviral Agents/pharmacology , Binding Sites , Coronavirus Papain-Like Proteases/genetics , Cyclic AMP/chemistry , Cyclic AMP/metabolism , Ligands , Models, Molecular , Molecular Docking Simulation , Protein Conformation , Protein Domains , SARS-CoV-2/drug effectsABSTRACT
Small molecules that bind the SARS-CoV-2 non-structural protein 3 Mac1 domain in place of ADP-ribose could be useful as molecular probes or scaffolds for COVID-19 antiviral drug discovery because Mac1 has been linked to coronavirus' ability to evade cellular detection. A high-throughput assay based on differential scanning fluorimetry (DSF) was therefore optimized and used to identify possible Mac1 ligands in small libraries of drugs and drug-like compounds. Numerous promising compounds included nucleotides, steroids, beta-lactams, and benzimidazoles. The main drawback to this approach was that a high percentage of compounds in some libraries were found to influence the observed Mac1 melting temperature. To prioritize DSF screening hits, the shapes of the observed melting curves and initial assay fluorescence were examined, and the results were compared with virtual screens performed using Autodock VINA. The molecular basis for alternate ligand binding was also examined by determining a structure of one of the hits, cyclic adenosine monophosphate, with atomic resolution.
