COVID-19 has resulted in ~2 million deaths and devastated economies around the world. While vaccines are being rolled out, viral inhibitors are still critical for treating moderate to severe diseases. Herein, we target the SARS-CoV-2 non-structural protein 15 (nsp15), a uracil-dependent endonuclease, which is believed to cleave viral RNA and evade host immune responses. As part of the National Virtual Biotechnology Laboratory (https://science.osti.gov/nvbl), a consortium of DOE National Laboratories focused on response to COVID19, with the later funding provided by the Coronavirus CARES Act, we searched compound libraries containing registered drugs and those that passed initial phases of clinical trials. Positive hits against nsp15 can be potentially repurposed for COVID-19 with minimal development time. A high throughput fluorescence resonance energy transfer (FRET) assay was developed to screen the inhibitory effect of candidate compounds, which narrowed down the selection to 15 compounds.

We then used native mass spectrometry to verify the binding. Because nsp15 is a relatively large homo-hexameric complex of 240 kDa, we applied surface induced dissociation (SID) and collision induced dissociation (CID) to release monomers from the hexamers with bound compounds, to better mass resolve the stoichiometry. Interestingly, most compounds only showed weak association to the nsp15 hexamer. Only Exebry-1 showed strong binding with a K_d of around 12 µM, and multiple binding per monomer was detected as compound concentration was increased. Docking simulation also showed multiple binding sites and poses for Exebry-1, including one region outside the catalytic site, deep into the pore of the hexamer. The data suggest potential allosteric mechanisms for inhibiting nsp15 that may not be fully appreciated.

An AmplexRed assay, that can identify some Pan-Assay Interference Molecules (PAINS), deprioritized probable false positives due to redox activity, resulting in 3 compounds for validation. A gel-based RNA cleavage assay and an in-cell viral inhibition assay further supported the selection of Exebryl-1. Although in vivo delivery is currently a bottleneck to achieve the required concentration for an antiviral effect, Exebryl-1 is a potential template for further optimization via structure-guided drug discovery.

Justin Craig

Appearances