Structural Model for Self-Limiting β-strand Arrangement Within an Alzheimer's Amyloid-β Oligomer

Abstract

Previous reports revealed that sodium dodecyl sulfate near its critical micelle concentration can drive the assembly of A{beta}42 along an oligomeric pathway. This pathway produces a 150 kDa peptide oligomer (approximately 32 peptide molecules or protomers) that does not aggregate further into amyloid fibrils. Solid-state nuclear magnetic resonance (NMR) spectroscopy revealed structural features distinguishing the 150 kDa oligomer from fibrils. A puzzling feature was the coexistence of parallel and antiparallel {beta}-sheets within the oligomer structure. Here we present new atomic-level structural constraints obtained via solid-state NMR spectroscopy, benefitting from improved resolution via sample concentration by ultracentrifugation. In addition, two-dimensional cryo-electron microscopy (cryo-EM) reconstruction revealed a 4-fold symmetric shape. We propose a structural model to rationalize the solid-sate NMR- and cryo-EM-derived structural constraints. This model has a hollow square cylinder shape, with antiparallel {beta}-sheets formed by residues 33-39 lining the inner walls and parallel {beta}-sheets formed by residues 11-22 lining the outer walls. Within successive layers, the outer {beta}-strands on each side of the square cylinder alternate between two forms: one within a U-shaped protomer and another within L-shaped protomer. Molecular dynamics simulations show that, when the oligomer model is embedded in a lipid membrane, ions permeate through the central pore, with cation selectivity. The model further motivates an assembly pathway-based interpretation that may explain why the 150 kDa oligomer does not undergo further aggregation into amyloid fibrils. Significance StatementA{beta} oligomers are thought to be the most toxic species in Alzheimers disease. Their sizes range from 2 to [~]50 protomers. Most published experimental data on A{beta} oligomers indicate that they, like fibrils, are composed of {beta}-sheets, but it is a mystery why any {beta}-sheet aggregate would exist as a stable oligomer without undergoing further aggregation into fibrils. Here, structural constraints from solid-state NMR and cryo-EM led us to an oligomer model with a hollow square cylinder shape capable of conducting ions when embedded in a lipid membrane. Based on the model, we argue that geometric frustration may distinguish the assembly pathway that produces this oligomer from fibril-forming assembly pathways.