BPS2026 – Ahead of the membrane curve: In silico insights into amyloid-β aggregation

Amyloid-β (Aβ) oligomers are implicated as neurotoxic agents in Alzheimer’s disease (AD). Despite extensive efforts, the structural heterogeneity and transient nature of these oligomers continue to impede rational drug design. It has been demonstrated that lipid membranes accelerate Aβ aggregation by promoting β-sheet-rich conformations. However, these investigations have predominantly employed planar lipids bilayers, neglecting the role of membrane curvature, which is important for protein-membrane interactions and a key feature in many cellular processes. To elucidate the influence of curvature on Aβ aggregation, we performed all-atom molecular dynamics simulations using the CHARMM36m force field on the special-purpose supercomputer anton 3. Systems included either two planar bilayers or a spherical synaptic vesicle model, each simulated with and without Aβ42 monomers for 18 μs. The membrane compositions mimicked neuronal lipid composition. Comparative analysis of the trajectories revealed curvature-dependent modulation of Aβ conformational dynamics, membrane perturbation profiles, and peptide-membrane binding affinities. Specifically, vesicular curvature enhanced Aβ-membrane affinity and promoted localized membrane lipid re-arrangement. These findings suggest that membrane curvature not only facilitates Aβ surface nucleation but may also contribute to the spatial heterogeneity of aggregation observed in vivo. This work provides mechanistic insight into curvature-mediated Aβ aggregation and underscores the importance of incorporating realistic membrane geometries in computational models of amyloidogenesis.