Elucidation of potent mammalian enzymes inhibitors targeting anti-diabetes drug from Castanea mollissima blume, 1851: an edible herbal collected in Vietnam via experimental and computation approaching

This study aimed to explore the extracts of Castanea mollissima Blume, 1851 as rich sources of enzyme inhibitors targeting type 2 diabetic (T2D) drugs. It was proved that the enzyme inhibitor with moderate effect against α-amylase and high inhibition against α-glucosidase may be a potential candidate for T2D management. Among the parts used of C. mollissima, the trunk bark extract demonstrated moderate pancreatic α-amylase inhibitory activity and a high inhibitory effect against α-glucosidase from rat with the IC₅₀ inhibition values of 371.23 and 212.79 μg/mL, respectively. Acarbose—a commercial antidiabetic compound with testing IC₅₀ of 24.01 and 192.35 μg/mL against α-amylase and α-glucosidase, respectively. Notably, the potent mammalian α-glucosidase inhibitory effect of the C. mollissima trunk bark extract collected in Vietnam was reported for the first time. Nineteen compounds (named 1–19) were detected and identified from the C. mollissima trunk bark extract using GCMS and HPLC analyses. Based on GCMS analysis, 9 volatiles were identified (1–9), of these, compounds 1, 4, and 9 were found as major volatiles with areas of 11.44, 13.5 and 45.27%, respectively. The utilization of HPLC techniques resulted in the detection and identification of 3 phenolic (10, 12, 15) and 7 flavonoid compounds (11, 1 3, 14, 16, 17, 18, 19). Of these, compounds 11, 12, 13, 14, 15, and 17 were found to contain in the C. mollissima trunk bark extract with high contents of 120.734, 40.373, 79.826, 43.845, 57.933, and 19.145 μg/mg of dried extract, respectively. The docking result indicated that most major compounds (1, 4, 9, 11, 12, 14, and 17) showed effective binding to α-glucosidase with good binding energy and acceptable RMSD values. In addition, almost all the identified compounds possessed drug properties and showed non-toxic for human use via Lipinski’s rule of five and ADMET analyses.