In silico screening of potential anti-inflammatory compounds from Sulawesi-endemic Begonia medicinalis

Abstract

Chronic inflammation contributes to the pathogenesis of many degenerative diseases. Long-term use of conventional anti-inflammatory drugs may be associated with serious adverse effects, prompting the search for new natural candidates. Begonia medicinalis is a plant endemic to Central Sulawesi that is traditionally used to treat fever and joint pain, but its phytochemical profile and molecular mechanisms underlying its potential anti-inflammatory activity have not been scientifically reported. The aim of this study was to identify the secondary metabolite profile of the ethanol extract of B. medicinalis leaves and to assess its molecular interactions with pivotal pro-inflammatory targets, namely cyclooxygenase-2 (COX-2), inducible nitric oxide synthase (iNOS), tumor necrosis factor-alpha (TNF-α), and interleukin-1 beta (IL-1β), using a computational approach. The study combined a laboratory-based phytochemical analysis with in silico analyses, including structure-activity relationship (SAR), absorption, distribution, metabolism, and excretion (ADME), toxicity prediction, molecular docking, and molecular dynamics simulations. B. medicinalis leaves were extracted by maceration using absolute ethanol, and the metabolite profiles were analyzed using gas chromatography-mass spectrometry (GC-MS). Biological activity prediction was performed using PASS Online, pharmacokinetic profiling using SwissADME, toxicity evaluation using ProTox 3.0, and protein-protein interaction analysis using STRING v12.0. Molecular docking was performed using PyRx with AutoDock Vina and visualized using BIOVIA Discovery Studio 2025. GC-MS analysis identified 21 compounds, including palmitic acid, dihomo-γ-linolenic acid, and stigmasterol, which showed predicted anti-inflammatory potential, favorable safety profiles, and acceptable drug-likeness characteristics. Among these compounds, stigmasterol showed favorable predicted binding affinities toward COX-2, iNOS, TNF-α, and IL-1β, while molecular dynamics simulations supported the stability of the resulting complexes over 100 ns. These computational findings suggest that B. medicinalis contains diverse bioactive compounds with potential anti-inflammatory properties. Stigmasterol emerged as the most promising lead candidate against the predicted anti-inflammatory targets, providing a theoretical molecular foundation for the future exploration of B. medicinalis therapeutic potential.