STRUCTURE-BASED MOLECULAR DOCKING, RMSD VALIDATION, AND DRUG-LIKENESS PROFILING OF PLANT-DERIVED PHYTOCONSTITUENTS AS DIPEPTIDYL PEPTIDASE-4 INHIBITORS FOR TYPE 2 DIABETES MELLITUS

Abstract

Background: Dipeptidyl peptidase-4 (DPP-4) inhibition remains a validated strategy for glucose-dependent glycaemic control in type 2 diabetes mellitus (T2DM), yet the tolerability and cost limitations of synthetic gliptins sustain interest in plant-derived alternatives.

Objective: To evaluate the DPP-4 inhibitory potential of five phytoconstituents - andrographolide, berberine, gymnemic acid, charantin, and ursolic acid - through structure-based molecular docking against the DPP-4 catalytic domain, benchmarked against the enzyme's co-crystallised ligand (CCL), together with computed physicochemical and drug-likeness profiling.

Methods: Docking was performed against the DPP-4 crystal structure (PDB ID: 4A5S) following redocking of the co-crystallised ligand for protocol validation. Binding energies, hydrogen-bonding geometry, and non-covalent contact patterns (hydrophobic, pi-pi, pi-alkyl, and pi-donor interactions) were extracted for each complex. Physicochemical and drug-likeness parameters (Lipinski, Ghose, Veber, Egan, and Muegge filters; topological polar surface area; predicted gastrointestinal absorption and blood-brain barrier permeation) were computed in silico for each phytoconstituent and for sitagliptin as reference standard.

Results: Andrographolide returned the most favourable docking score among the phytoconstituents (-9.4 kcal/mol), approaching the co-crystallised reference ligand (-11.0 kcal/mol), followed by gymnemic acid (-8.9 kcal/mol), ursolic acid (-8.7 kcal/mol), and berberine (-8.6 kcal/mol); charantin showed the weakest affinity (-7.4 kcal/mol). Pose-reproducibility (RMSD lower-bound/upper-bound) was tightest for the co-crystallised ligand and charantin and widest for andrographolide and gymnemic acid. All compounds engaged catalytic-domain residues implicated in gliptin recognition, including Tyr547, Glu205/Glu206, Ser630, Phe357, and Trp629, via conventional and carbon hydrogen bonds together with pi-stacked, pi-alkyl, and pi-sigma contacts. Computed physicochemical profiling indicated full Lipinski compliance for andrographolide and berberine, one violation for ursolic acid, and two violations for the charantin glycoside component modelled in this study, attributable to elevated lipophilicity and molecular bulk.

Conclusion: Andrographolide and gymnemic acid emerge as the most promising phytoconstituent DPP-4 inhibitor candidates on binding-energy ranking and residue-level concordance with the catalytic pocket, though both show comparatively wide RMSD spreads warranting confirmatory redocking; berberine offers a more pose-stable, drug-likeness-compliant alternative, while ursolic acid and charantin warrant physicochemical optimisation. These computational findings provide a rationale for prioritising andrographolide, gymnemic acid, and berberine for in vitro DPP-4 inhibition assays and subsequent lead optimisation.