Virtual Combinatorial Library Screening of Quinadoline B Derivatives against SARS-CoV-2 RNA-Dependent RNA Polymerase

Simone Brogi,Jenina Beatriz Hernandez,Allan Patrick Macabeo,Jeremiah Gabriel Africa,Sophia Morgan Tan,Vincenzo Calderone,Mark Tristan Quimque,Kin Israel Notarte

doi:10.3390/computation10010007

Abstract

The unprecedented global health threat of SARS-CoV-2 has sparked a continued interest in discovering novel anti-COVID-19 agents. To this end, we present here a computer-based protocol for identifying potential compounds targeting RNA-dependent RNA polymerase (RdRp). Starting from our previous study wherein, using a virtual screening campaign, we identified a fumiquinazolinone alkaloid quinadoline B (Q3), an antiviral fungal metabolite with significant activity against SARS-CoV-2 RdRp, we applied in silico combinatorial methodologies for generating and screening a library of anti-SARS-CoV-2 candidates with strong in silico affinity for RdRp. For this study, the quinadoline pharmacophore was subjected to structural iteration, obtaining a Q3-focused library of over 900,000 unique structures. This chemical library was explored to identify binders of RdRp with greater affinity with respect to the starting compound Q3. Coupling this approach with the evaluation of physchem profile, we found 26 compounds with significant affinities for the RdRp binding site. Moreover, top-ranked compounds were submitted to molecular dynamics to evaluate the stability of the systems during a selected time, and to deeply investigate the binding mode of the most promising derivatives. Among the generated structures, five compounds, obtained by inserting nucleotide-like scaffolds (1, 2, and 5), heterocyclic thiazolyl benzamide moiety (compound 3), and a peptide residue (compound 4), exhibited enhanced binding affinity for SARS-CoV-2 RdRp, deserving further investigation as possible antiviral agents. Remarkably, the presented in silico procedure provides a useful computational procedure for hit-to-lead optimization, having implications in anti-SARS-CoV-2 drug discovery and in general in the drug optimization process.

Highlights

The continued rise in COVID-19 cases worldwide despite the availability of vaccines sustains the demand to discover treatment and prophylactic regimens, through natural products’ repurposing and design [1,2,3]
In COVID-19 drug discovery, several possible drug targets, comprising structural and non-structural proteins, have been exploited in searching novel chemical entities as antiSARS-CoV-2 agents [10,11,12,13]. Among these targets is the RNA-dependent RNA polymerase (RdRp), which is a multi-domain SARS-CoV-2 protein playing a crucial role in the viral life cycle
Different fungal derivatives, in particular quinaxoline alkaloids identified from the mangrove‐derived fungus Cladosporium sp., were identified alkaloids identified fromRdRp the mangrove-derived fungus

Summary

Introduction

The continued rise in COVID-19 cases worldwide despite the availability of vaccines sustains the demand to discover treatment and prophylactic regimens, through natural products’ repurposing and design [1,2,3]. Computational strategies play a crucial role in accelerating the discovery of effective anti-SARS-CoV-2 agents [4,5,6,7,8], as in silico experiments are vital in the screening of biologically active compounds, offering a rapid, low-cost, and effective adjunct to in vitro and in vivo experiments. In COVID-19 drug discovery, several possible drug targets, comprising structural and non-structural proteins, have been exploited in searching novel chemical entities as antiSARS-CoV-2 agents [10,11,12,13] Among these targets is the RNA-dependent RNA polymerase (RdRp), which is a multi-domain SARS-CoV-2 protein playing a crucial role in the viral life cycle. We combined several computational approaches for optimizing a previously described compound targeting

Methods

Results

Conclusion