Abstract
AIMS: Pistacia atlantica Desf. (Anacardiaceae) is traditionally used in Mediterranean medicine, with previous studies showing antidiabetic potential in its fruits and leaves. This study evaluates the antidiabetic activity of P. atlantica galls (PAG) extracts using in vitro, chemometric, and in silico approaches. METHOD: The antidiabetic activity of the samples were studied by measuring their half-maximal inhibitory concentrations (IC(50)s) concentrations according to the in vitro enzyme inhibition assays and modelled as a function of the LC fingerprints using the partial least squares technique. Crystal structures of the human pancreatic α-amylase (HPA) and the α-glucosidase homologue isomaltase were obtained from the Protein Data Bank website (http://www.rcsb.org/pdb). Docking simulations and calculations were carried out using AutoDock Vina. RESULTS: PAG extracts inhibited HPA (IC(50)s ranging from 1.85 to 2.92 mg/mL) and α-glucosidase (IC(50)s ranging from 34 to 49 µg/mL) activities, with galls collected from male plants showing higher activity than those from female plants. UPLC fingerprinting, linked to chemometric analysis using a partial least squares regression model, putatively identified five compounds (quinic acid, methyl gallate, digalloyl quinic acid, methyl digallate, and valoneic acid dilactone) responsible for this antidiabetic effect. Molecular docking using AutoDock Vina revealed that the identified compounds interacted with key amino acid residues of HPA and α-glucosidase. CONCLUSIONS: By employing UPLC fingerprinting combined with chemometric analysis and molecular docking simulations, quinic acid and digalloyl quinic acid were identified from P. atlantica gall extract as the most promising ligands for further investigation into their antidiabetic potential.