Researchers have used molecular docking software to screen 128 compounds, including 89 FDA-approved antiviral drugs and 39 plant-derived molecules, against three SARS-CoV-2 proteins: the nsp15 endoribonuclease, the main protease, and the nsp16-nsp10 complex. The computational study, published in the journal Green Technology, Resilience, and Sustainability, was designed as a proactive exercise in drug repurposing, aiming to build a pre-screened pool of candidates that could be rapidly evaluated if a similar viral outbreak were to recur. The work is entirely in silico, meaning the results reflect predicted binding interactions rather than measured antiviral activity in cells, animals, or patients.
The team, led by Triveni Singirisetty and colleagues, docked the compound libraries onto crystal structures obtained from the Protein Data Bank: 6VWW for nsp15, 6Y84 for the main protease, and 6W4H for the nsp16-nsp10 complex. Docking was performed with the licensed Schrödinger 2021_2 Glide Maestro tool, and the results were ranked by glide scores and expressed in binding energy values of kilocalories per mole. Before docking, the selected proteins were validated with Ramachandran plots, which assess whether the torsional angles of amino acid residues fall within favored or disfavored regions, and the binding sites were identified either through Glide’s sitemap tool or, for the nsp16-nsp10 complex, by anchoring the grid to the co-crystallized ligand S-adenosylmethionine, known as SAM.
Each target plays a distinct role in the viral life cycle. Nsp15 is a nidoviral uridylate-specific endoribonuclease whose hexamerized form binds RNA to execute its cleavage activity, and it requires metal ions such as manganese or magnesium as cofactors. The main protease is central to processing viral polyproteins, and the nsp16-nsp10 complex carries out 2′-O-methyltransferase activity, with nsp10 activating the methyltransferase function of nsp16. Because two-thirds of the coronavirus genome encodes non-structural proteins that drive replication, the authors argue that simultaneously targeting these three proteins offers a strategy to interfere with RNA processing, protein maturation, and RNA capping at different stages of the viral cycle.
Against nsp15, the top-ranked antiviral drugs by glide score were remdesivir at -7.228, ribavirin at -6.988, letermovir at -6.96, favipiravir at -6.823, and idoxuridine at -6.788. Remdesivir formed hydrogen bond interactions with the residues TYR 89, GLY 165, and SER 198, along with a hydrophobic interaction with LYS 90. When binding energy was considered instead, zidovudine stood out with a value of -72.11 kcal/mol, and compounds such as cobicistat, ritonavir, saquinavir, telaprevir, and daclatasvir also showed energies in the -60 kcal/mol range, leading the authors to suggest these deserve further in vitro study.
The plant-derived library produced notable results at the same target. Epigallocatechin achieved a glide score of -7.596, amentoflavone -7.417, and quercetin -7.408, all stronger than remdesivir’s -7.22 at nsp15. Ent-epiafzelechin followed at -6.944. The authors observed that amentoflavone and ent-epiafzelechin interact with distinct sets of amino acid residues, including GLU 203, ASP 268, THR 167, TYR 279, ASP 297, and VAL 295, and they speculate that such interaction patterns might interfere with the hexamerization that nsp15 requires for full endoribonuclease activity. Two natural compounds, saikosaponin A and vitamin B2, did not fit into the nsp15 active site and were excluded from that ranking.
Docking against the main protease yielded a different hierarchy. Adefovir led the antiviral drugs with a glide score of -7.527, followed by entecavir at -7.141, penciclovir at -7.07, ganciclovir at -6.852, and tenofovir at -6.624. Adefovir formed hydrogen bonds with MET 6, ASP 153, TYR 154, and ARG 298, plus hydrophobic contact with PHE 8. The study noted that glide energies at this target were generally lower than those observed for nsp15, though four drugs, cobicistat, telaprevir, remdesivir, and raltegravir, exceeded -50 kcal/mol in binding energy. Among natural compounds, ent-epiafzelechin scored -7.03, hippomannin A -6.71, epigallocatechin -6.425, quercetin -6.141, and proanthocyanidin -6.624. Six antiviral drugs and six natural compounds, including chebulagic acid and saikosaponin A, failed to dock into the main protease active site.
The nsp16-nsp10 complex was screened at the SAM binding site, using the co-crystallized SAM as a reference standard. SAM itself recorded the highest docking score of -10.06 with a glide energy of -76.49 kcal/mol. Vidarabine emerged as the closest antiviral lead at -9.399, followed by tenofovir at -8.654, telbivudine at -7.839, fosamprenavir at -7.808, and ribavirin at -7.654. Zidovudine, however, displayed the lowest binding energy of the screen at -80.07 kcal/mol, below even SAM’s -76.49, and formed an ionic bond with ASP 6897. The authors interpret this as indicating a more stable predicted zidovudine-protein complex than the SAM-protein complex, and they also flagged ritonavir, cobicistat, saquinavir, and emricasan as having comparable energies. Among natural molecules, apigenin scored -7.58, quercetin -7.316, silymarin -7.252, amentoflavone -7.182, and ent-epiafzelechin -7.142, with several sharing common residue interactions with SAM, including ASP 6912, CYS 6913, and ASP 6928.
The authors place their findings in the context of a broader repurposing literature. They cite reports that remdesivir showed an EC90 of 1.76 micromolar against the virus in Vero E6 cells and that it acts on the RNA-dependent RNA polymerase by delaying chain termination. They also reference in vitro work showing that epigallocatechin gallate, a major green tea component, blocks the interaction between the viral spike protein and the ACE2 receptor, and computational studies in which flavonoids such as hesperidin and rutin showed strong affinity for viral proteins. The paper notes that repurposing can shorten development timelines, citing figures that 30 percent of repurposed drugs reach Phase II trials, that timelines shrink from 10 to 15 years to 3 to 6 years, and that costs fall by 40 to 60 percent, though these are literature-derived estimates rather than findings of the present study.
The limitations of the approach are significant and acknowledged by the authors themselves. Docking scores are computational predictions of binding affinity, not proof of antiviral effect, and the paper explicitly states that the identified compounds warrant further investigation through in vitro and in vivo validation studies before any therapeutic conclusions can be drawn. The study also did not perform molecular dynamics simulations or free energy calculations, which the authors recommend as future refinements to improve binding predictions. The compound library was assembled from earlier reports rather than an exhaustive chemical space survey, and a handful of molecules failed to dock at certain targets, illustrating how ligand preparation and active site geometry can shape outcomes.
As a preparedness exercise, the study offers a ranked shortlist rather than a treatment recommendation. Its practical value lies in narrowing hundreds of possibilities to a manageable set of leads, remdesivir, ribavirin, and letermovir at nsp15; adefovir, entecavir, and penciclovir at the main protease; vidarabine, tenofovir, fosamprenavir, zidovudine, and saquinavir at the SAM site; and plant molecules such as apigenin, quercetin, silymarin, amentoflavone, hippomannin A, epigallocatechin, ent-epiafzelechin, and chebulagic acid, that laboratories could prioritize for experimental testing. The authors argue that maintaining such pre-validated leads, particularly among drugs with established safety profiles, could accelerate the response to future coronavirus outbreaks, provided the computational predictions survive the rigors of laboratory and clinical validation.
Subject of Research: In silico drug repurposing screening of antiviral drugs and plant-derived compounds against SARS-CoV-2 nsp15, main protease, and nsp16-nsp10 complex
Article Title: Drug repurposing in silico screening of antiviral drugs and evidence-based plant-derived anti-oxidant molecules targeting nsp15, main protease, and nsp16-nsp10 complex of SARS-CoV-2: strategic insights for managing future recurrence
Article References: Singirisetty, T., Chilamakuru, N. B., Bodapati, A., Kallam, S. D. M., Thangaswamy, S., Nelson, V. K., & Suryadevara, P. R. (2025). Drug repurposing in silico screening of antiviral drugs and evidence-based plant-derived anti-oxidant molecules targeting nsp15, main protease, and nsp16-nsp10 complex of SARS-CoV-2: strategic insights for managing future recurrence. Green Technology, Resilience, and Sustainability, 5(1), Article 2. https://doi.org/10.1007/s44173-025-00017-5
Image Credits: AI Generated
DOI: 10.1007/s44173-025-00017-5
Keywords: SARS-CoV-2, drug repurposing, molecular docking, nsp15, main protease, nsp16-nsp10 complex, antiviral drugs, phytocompounds, virtual screening, in silico, zidovudine, remdesivir
