An Approach for Novel Antiviral Compound Discovery
The study, “AI-guided Designing of SARS-CoV-2 Main Protease (Mpro) Inhibitors: An Approach for Novel Antiviral Compound Discovery,” was carried out by Anand Kumar Pandey and Soumya Rathore in Current Biological Sciences.
Researchers in India have used an AI-based molecule generation platform to design a novel compound that computationally blocks one of the most conserved and functionally essential enzymes of SARS-CoV-2 — the main protease — with drug-like properties, no predicted toxicity, and a binding profile that compares favourably with clinically studied inhibitors, making it a candidate worth pursuing for further laboratory validation.
Why the Main Protease Remains One of the Best Targets Against SARS-CoV-2
Despite the development of vaccines and antiviral drugs since the emergence of COVID-19, the continued evolution of SARS-CoV-2 into new variant strains keeps the search for broadly effective treatments urgent. Many of the virus’s surface proteins mutate frequently enough to reduce the effectiveness of drugs and antibodies designed against earlier strains. The main protease of SARS-CoV-2, known as Mpro, is different. It is an enzyme the virus cannot do without: it cleaves the large polyproteins the virus produces into the functional components needed to assemble new viral particles, cutting at 11 specific sites using a pair of amino acids — Cysteine 145 and Histidine 41 — that form its catalytic core. Crucially, these catalytic residues have remained conserved not only across all known SARS-CoV-2 variants but also across other members of the coronavirus family, and no equivalent enzyme exists in humans, which means a drug targeting Mpro is unlikely to interfere with normal human biology. These properties make Mpro one of the most attractive drug targets in antiviral research. Several Mpro inhibitors have reached clinical use, including Pfizer’s nirmatrelvir, but side effects and the need for broader-spectrum coverage have kept the search for improved inhibitors active.
If you want to publish your article please visit : https://bit.ly/4de0DRi
How the Molecule Was Designed and What It Does
Rather than screening existing compound libraries, the research team at Bundelkhand University, Jhansi, used an AI-based generative platform called MolProphet to design new molecules directly from the three-dimensional structure of the Mpro enzyme’s active site. Working from the crystal structure of a known Mpro-inhibitor complex (PDB ID: 7GEF), the model generated 200 candidate compounds by scanning the structural features of the binding pocket and building molecules predicted to fit it tightly and interact with its key residues. From this pool, one compound, named Mpro_inhi_1 and chemically identified as 6-(2-anilinoethoxy)-3,4-dihydroquinolin-2(1H)-one, emerged as the strongest candidate based on its predicted inhibitory activity and binding interaction profile. The molecule has a molecular weight of 282.34 Da, satisfies all standard drug-likeness rules including Lipinski’s rule of five with no violations, and has a mild lipophilicity value that supports effective absorption and distribution in the body. Toxicity screening using the ProTox-3.0 server showed zero similarity to any known toxic receptor ligand across 16 safety targets present in the human body, with hepatotoxicity, nephrotoxicity, cardiotoxicity, mutagenicity, carcinogenicity, and cytotoxicity all predicted to be insignificant. Quantum mechanical analysis using Density Functional Theory confirmed the molecule’s chemical stability through a large HOMO-LUMO energy gap of 9.289 eV, indicating low susceptibility to unwanted chemical degradation. Molecular docking at the Mpro active site showed a binding energy of -5.8 kcal/mol, with hydrogen bonds formed to seven active site residues including the catalytic dyad residues Cys145 and His41, and hydrophobic contacts with Thr25 and Leu27. The compound occupied both the S1 and S2 subpockets of the enzyme, the same binding regions targeted by the clinically studied Pfizer inhibitors PF-00835231 and PF-07304814, which showed binding energies of -5.8 and -5.1 kcal/mol respectively at the same site. A 100-nanosecond molecular dynamics simulation confirmed that the compound remained stably bound throughout, with the protein-inhibitor complex showing lower structural deviation than the free enzyme, and free binding energy calculations yielded an average total delta G of -12.92 kcal/mol, further validating the strength and consistency of the interaction. Importantly, the binding of Mpro_inhi_1 induced significant conformational changes in the enzyme architecture that are consistent with driving it toward an inactive form, reducing its ability to cleave viral polyproteins and thereby disrupting viral replication.
Promising Computational Results — and the Work That Remains
The authors are clear that while the computational evidence supporting Mpro_inhi_1 as a lead compound is comprehensive and encouraging, it remains entirely in silico. The predicted inhibitory activity (IC50 of approximately 164 µM) is in a range comparable to other computationally identified Mpro inhibitors reported in the literature, but this figure must be confirmed in biochemical assays using purified Mpro enzyme, followed by cell-based antiviral activity studies and ultimately in vivo pharmacological testing, before any clinical relevance can be established. The study also notes that MD simulation replicates were not conducted due to computational resource limitations, which is a constraint on the statistical robustness of the dynamics results. To support the practical path forward, the MolProphet platform was also used to predict a chemically feasible synthesis pathway for Mpro_inhi_1, which the authors provide as a starting point for laboratory chemists aiming to produce the compound for wet-lab validation. The research was led by corresponding author Dr. Anand Kumar Pandey of the Department of Biotechnology Engineering, Institute of Engineering and Technology, Bundelkhand University, Jhansi, 284128, Uttar Pradesh, India, alongside co-author Soumya Rathore of the Department of Food Technology, Raja Balwant Singh Engineering Technical Campus, Bichpuri, Agra, Uttar Pradesh, India.
Read the published article here: https://bit.ly/4yPBhCW
Article title: AI-guided Designing of SARS-CoV-2 Main Protease (Mpro) Inhibitors: An Approach for Novel Antiviral Compound Discovery
DOI: 10.2174/0129496616498683260910203520
Current Biological Sciences
10.2174/0129496616498683260910203520
Computational simulation/modeling
https://www.benthamscience.com/article/158574?utm_source=bentham_pr&utm_medium=pressrelease&utm_campaign=cbios26&utm_id=cbios&utm_term=organic&utm_content=pressrelease26