Computational screening of Thiazole-based natural compounds against Xanthine Oxidase: Consensus screening, molecular dynamics, and DFT

Background: Xanthine oxidase (XO) is an established molecular target for controlling uric acid production; however, identification of diverse natural compounds, particularly thiazole-based compounds capable of interacting with XO favorably remains of interest. Thiazole-based have received comparatively limited computational evaluation against XO despite the pharmacological versatility of the thiazole scaffold. Objective: Therefore, this study employed an integrated in silico workflow to evaluate thiazole-containing natural compounds as XO-binding candidates. Methods: A total of 10,586 thiazole-containing natural compounds were retrieved from the COCONUT database. These were filtered based on the Lipinski rule of five, and PAINS filtering using druglikeFilter 1.0. The consensus virtual screening of natural library against XO receptor was performed using Glide, and rescoring using Prime MM-GBSA. The best hits were subjected for molecular dynamics (MD) simulations using Desmond v2021.4 for up to 200 ns. The trajectories profiles of RMSD, RMSF, Rg, 2D-PCA, and 3D-FEL were analyzed. Results: From the comprehensive screening, four compounds—CNP0259475, CNP0472606, CNP0403609, and CNP0486839 were showed superior binding affinity than the co-crystal Quercetin and the standard Febuxostat. The corresponding MM-GBSA binding free energies were -66.66 kcal/mol, -65.13 kcal/mol, -64.38 kcal/mol, -59.86 kcal/mol, -59.76 kcal/mol and -56.19 kcal/mol for CNP0259475, CNP0472606, CNP0403609, CNP0486839, Quercetin and Febuxostat, respectively. The MD simulations of all systems achieved convergence and stabilized achieving RMSD within 2.0 Å. The predictive toxicity revealed notable toxicity liabilities which warrants further lead optimization. Conclusion: This study identified thiazole-based natural chemotypes which serve as a starting point for developing analogs. However, further experimental validation is warranted to confirm these in silico findings.