Icaritin and its N-Boc-amino acid ester derivatives: Chemistry, structure-activity relationships, and preliminary...

Type 2 diabetes increasingly calls for multi-target therapies that act along the intestinal-hepatic axis rather than at a single node, yet few chemical strategies have been designed to pursue both nodes from a single natural scaffold. Icaritin (ICT), the principal bioactive metabolite of the Epimedium flavonoid icariin, has long been used in traditional Chinese medicine to relieve metabolic wasting-thirst disturbances resembling diabetes and shows promising hypoglycemic activity. However, its free phenolic hydroxyl groups confer extensive hydrogen bonding with water, a structural feature expected to limit aqueous solubility and membrane permeability. Here, ICT was prepared from Epimedium leaves via a green enzyme-assisted hydrolysis route and converted, in synthetically useful yields (75-90%), into five N-Boc-amino acid ester derivatives (M1-M5) designed to mask its reactive phenolic hydroxyls while adding tunable lipophilic bulk. The N-Boc-l-proline ester M1 was the most potent α-glucosidase inhibitor (IC₅₀ = 5.5 ± 0.5 nM), 1.5-fold and 2.0-fold more active than its parent ICT (8.3 ± 0.5 nM) and acarbose (11.0 ± 0.6 nM), and remained intact in simulated gastric and intestinal fluids. Docking and 100 ns molecular dynamics simulations provided preliminary structural insights into the possible interaction of M1 with α-glucosidase, with the rigid N-Boc-pyrrolidine group positioned in a hydrophobic cleft near the active site. In palmitate-induced insulin-resistant HepG2 cells, M1 (20 μM) increased glucose consumption by 58.1% over the model group, lowered intracellular ROS to approximately 2.6-fold of control, and reduced lipid accumulation to a level comparable to metformin. These effects were accompanied by downregulation of glucose-6-phosphatase and upregulation of catalase, without significantly altering basal glucose consumption in normal HepG2 cells. Together, these findings establish a scalable route to ICT and demonstrate that esterification of its phenolic hydroxyl groups can modify α-glucosidase inhibition and cellular metabolic phenotypes, identifying M1 as a lead compound meriting in vivo validation.