Trường Đại học Khoa học, Đại học Huế
COMPUTATIONAL SCREENING OF NUCLEOSIDES FROM CORDYCEPS CICADAE AS a-GLUCOSIDASE INHIBITORS FOR DIABETES
https://doi.org/10.64302/joshusc.v35n2.1432
Nguyen Cong Kinh, Nguyen Dai Chau, Pham Van Vinh, Nguyen Vinh Phu, Phan Tu Quy, Tran Quang Huy, Nguyen Thi Thanh Hai, Nguyen Thi Ai Nhung
Email: ntanhung@hueuni.edu.vn
α-glucosidase inhibition slows intestinal carbohydrate digestion and remains a practical strategy for controlling postprandial hyperglycemia in diabetes. Five nucleosides and nucleoside derivatives from Cordyceps cicadae, namely, N6-(2-hydroxyethyl)-adenosine (C1), thymidine (C2), uridine (C3), 2′-deoxyinosine (C4), and 2′-deoxyguanosine (C5) were evaluated against α-glucosidase (PDB ID: 3W37) by molecular docking, 100 ns molecular dynamics simulation, drug-likeness assessment, and ADMET prediction, as the reference inhibitor (D). Binding-site prescreening across four candidate sites determined the preferred docking position for each compound. Docking scores ranked C3 (-12.0 kcal.mol-1) > C2 (-11.9 kcal.mol-1) > C4 (-11.6 kcal.mol-1) > C5 (-10.8 kcal.mol-1) > C1 (-10.0 kcal.mol-1), with C3 showing the most favorable predicted binding and twelve van der Waals interactions at Site 2. Backbone RMSD of the C3-3W37 complex restabilized near 0.28 nm after a transient excursion at 60-80 ns, during which the three to four hydrogen bonds maintained over most of the trajectory briefly fell to one or two, while Rg and SASA remained consistent. All five compounds satisfied Lipinski's Rule of Five, and ADMET profiling indicated favorable pharmacokinetic profiles, with C3 presenting the most acceptable predicted safety margins. Within the limits of these in silico predictions, C3 emerges as a prioritized candidate for subsequent experimental evaluation.
mucluc.pdf
