Multi-target Computational Profiling of Caffeine and Vernonia amygdalina Phytoconstituents Reveals Potential Neuroprotective Modulators of Diabetes-Associated Neurodegeneration

Abstract

Diabetes-associated neurodegeneration involves multiple pathological processes, including oxidative stress, impaired neurotrophic signalling, neurotransmitter dysregulation, altered metabolic signalling and progressive neuronal dysfunction. This study evaluated the in silico interaction profiles of caffeine and selected phytoconstituents identified from the hydro-methanolic leaf extract of Vernonia amygdalina against protein targets implicated in diabetes-related neurodegenerative processes. Compounds identified by GC-MS profiling were assessed, together with caffeine, for their interactions with tropomyosin receptor kinase B, N-methyl-D-aspartate receptor GluN1, dopamine D2 receptor, muscarinic M1 receptor, glucagon-like peptide-1 receptor and the Nrf2-MafG complex. Ligands and proteins were prepared using standard molecular modelling protocols, followed by Glide docking, MM-GBSA binding free energy estimation and in silico prediction of selected pharmacokinetic and drug-likeness properties. The GC-MS profile showed caryophyllene as the most abundant constituent, followed by a 1,3-cyclohexadiene derivative, alloaromadendrene, humulene, benzene-1-(1,5-dimethyl-4-hexenyl)-4-methyl, caryophyllene oxide, bornyl acetate and caffeine. Docking results indicated target-dependent interaction patterns among the tested compounds. Bornyl acetate showed notable docking scores against TrkB and GLP-1 receptor, while alloaromadendrene, humulene, caryophyllene and caryophyllene oxide showed broader interaction profiles across selected receptors. MM-GBSA analysis supported some docking trends but also indicated that binding stability varied across targets and ligands. Predicted ADMET parameters suggested favourable drug-likeness or blood-brain barrier permeability for selected compounds. Overall, the findings provide computational evidence that caffeine and selected V. amygdalina phytoconstituents may interact with molecular targets relevant to diabetes-associated neurodegeneration. These results remain predictive and require biochemical, cellular and in vivo validation.