BPS2026 – Structure-activity of antioxidants in lipid membranes

Antioxidants that block lipid peroxidation are central to managing oxidative stress in conditions such as inflammation, aging, and neurodegeneration. Yet, highly effective radical-trapping agents (RTAs) remain elusive because reactivity within lipid membranes cannot be predicted from solution studies. To address this challenge, we developed a fluorescence-based high-throughput assay to screen compound libraries for membrane-specific antioxidant activity. Our current work focuses on phenolic metabolites, naturally occurring plant compounds that quench radicals through electron or hydrogen atom transfer. Preliminary results reveal broad variability in activity, underscoring the complexity of structure-activity relationships in bilayer environments. To define the molecular features governing potency, we integrate machine learning models with experimental data. This combined approach aims to reveal design rules for next-generation antioxidants with enhanced efficacy, ultimately surpassing natural agents such as α-tocopherol and informing therapeutic strategies against oxidative stress-driven diseases.