Plant compounds show strong binding to diabetes drug targets in computer simulations

Two plant-derived molecules have emerged as the strongest performers in a computational screen for new type 2 diabetes drug candidates. N-methyl-1-adamantaneacetamide, traced to the herb Asystasia gangetica, and 3′,4′-di-O-methyl ellagic acid-4α-L-rhamnopyranoside, isolated from the plant Synadenium glaucescens pax, showed favourable predicted pharmacokinetics, strong binding energies and stable behaviour in molecular dynamics simulations against key receptors involved in blood sugar regulation. The findings, published in Discover Pharmaceutical Sciences, are entirely computational and the authors stress that laboratory validation in cells and living organisms is still required before any therapeutic claims can be made.

The research team, comprising Nawazi M. Kagimbo, Christopher J. Mwankuna and Faith P. Mabiki, set out to evaluate secondary metabolites from the two plants against four protein targets central to type 2 diabetes management: the peroxisome proliferator-activated receptor gamma (PPARγ), 11β-hydroxysteroid dehydrogenase 1 (11β-HS1), dipeptidyl peptidase-4 (DPP4) and the sodium-glucose cotransporter 2 (SGLT2).

The choice of plants rests on ethnopharmacological records. Asystasia gangetica leaves are used in East African folk medicine for glycemic control, while ellagic acid derivatives from Synadenium glaucescens exhibit anti-inflammatory and antioxidant activity relevant to type 2 diabetes pathogenesis.

Using Swiss Target Prediction and the ChEMBL database, the team screened roughly 135 candidate compounds and narrowed the field to 15 that met the criteria for potential interaction with the selected proteins. Thirteen of these, plus the ellagic acid derivative, came from A. gangetica; compound 14, the rhamnopyranoside, originated from S. glaucescens pax. Standard drugs including empagliflozin, dapagliflozin, linagliptin, saxagliptin, pioglitazone and rosiglitazone served as reference ligands for comparison.

Protein crystal structures were retrieved from the RCSB Protein Data Bank under the codes 8HEZ for SGLT2, 1J2E for DPP4, 1BHS for 11β-HS1 and 1PRG for PPARγ, then prepared by removing water molecules, adding polar hydrogens and assigning Kollman charges. Docking was carried out with Chimera and AutoDock Vina, and reliability was checked by redocking each co-crystallised ligand, which returned a root-mean-square deviation of 2 Å or less from the experimental pose. Absorption, distribution, metabolism and excretion properties were predicted with SwissADME, covering lipophilicity, aqueous solubility, gastrointestinal absorption, blood–brain barrier permeation, P-glycoprotein substrate status and cytochrome P450 inhibition.

However, the authors note that these two high scorers carried poor ADME profiles, suggesting limited oral bioavailability, and they propose nanoformulation or structural optimisation as possible routes to improve their therapeutic potential.

Yet the same ligand paid a heavy desolvation price: its polar solvation energy of 256.3 ± 19.38 kJ/mol, a positive term that opposes binding, dragged its total binding energy down to –39.48 ± 15.85 kJ/mol. N-methyl-1-adamantaneacetamide, by contrast, combined a van der Waals contribution of –115.83 ± 8.75 kJ/mol with a modest polar solvation penalty of 50.166 ± 9.21 kJ/mol, yielding the most favourable total binding energy in the PPAR system at –96.03 ± 9.75 kJ/mol.

The pattern reversed at SGLT2. There, the ellagic acid derivative recorded van der Waals and electrostatic contributions of –152.75 ± 15.19 kJ/mol and –24.86 ± 8.64 kJ/mol respectively, and its polar solvation penalty of 102.24 ± 15.07 kJ/mol was far smaller than the 228.42 ± 17.55 kJ/mol incurred by benzyl β-D-glucopyranoside. The authors attribute the adamantane compound’s hydrophobic strength to its adamantane core, which maximises contact with hydrophobic pockets, and the ellagic acid derivative’s van der Waals dominance to its multiple fused aromatic rings and methyl groups.

Molecular dynamics simulations run for 100 nanoseconds in GROMACS added a stability dimension to the picture. Root mean squared deviation values for the PPAR receptor stayed between 0.2 and 0.35 nm, with benzyl β-D-glucopyranoside averaging the lowest at 0.20 nm and N-methyl-1-adamantaneacetamide holding steady at 0.25 nm. Radius of gyration analysis showed the ellagic acid derivative and the adamantane compound preserved the PPAR receptor’s compact structure at 1.95 nm, while at SGLT2 the ellagic acid derivative and benzyl β-D-glucopyranoside maintained 2.40 nm.

The authors conclude that N-methyl-1-adamantaneacetamide and 3′,4′-di-O-methyl ellagic acid-4α-L-rhamnopyranoside are promising multi-target candidates for blood sugar regulation, with the former favoured at PPARγ and the latter at SGLT2 on the basis of binding properties and stability. They are careful to frame the work as preliminary and based solely on computational predictions, noting that enzyme inhibition assays and glucose uptake experiments in cell models are planned as the next step. Until such in vitro and in vivo tests confirm efficacy, safety and mechanisms of action, the compounds remain lead candidates for future drug development rather than proven therapies.

Subject of Research: Computational evaluation of plant secondary metabolites as potential antidiabetic agents against type 2 diabetes protein targets

Article Title: In silico evaluation of N-methyl-1-adamantaneacetamide and 3′,4′-di-O-methyl ellagic acid-4α-La-rhamnopyranoside from Asystasia gangetica and Synadenium glaucescens pax as potential antidiabetic agents

Article References: Kagimbo, N. M., Mwankuna, C. J., & Mabiki, F. P. (2026). In silico evaluation of N-methyl-1-adamantaneacetamide and 3′,4′-di-O-methyl ellagic acid-4α-La-rhamnopyranoside from Asystasia gangetica and Synadenium glaucescens pax as potential antidiabetic agents. Discover Pharmaceutical Sciences, 2(1), Article 3. https://doi.org/10.1007/s44395-025-00036-y

Image Credits: AI Generated

DOI: 10.1007/s44395-025-00036-y

Keywords: type 2 diabetes, molecular docking, molecular dynamics simulation, PPARγ, SGLT2, DPP4, 11β-HS1, Asystasia gangetica, Synadenium glaucescens, ADME prediction, MM/PBSA, drug discovery