Modeling the Aggregation Propensity and Toxicity of Amyloid-β Variants.

Protein aggregation is a hallmark of many neurodegenerative disorders. Alzheimer's disease (AD) is directly linked to deposits of amyloid-β (Aβ) derived from the amyloid-β protein precursor (AβPP), and multiple experimental studies have investigated the aggregation behavior of these amyloids. The present paper reports modeling of the aggregation propensities and cell toxicities of genetic variants of Aβ known to increase disease risk. From correlation to experimental data, and using four distinct experimental structures to test structural sensitivity, we find that the Spatial Aggregation Propensity (SAP) formalism can describe the relative experimental aggregation propensities of Aβ 42 variants (R2 = 0.49 and 0.70, p∼0.02 and 0.002, for 1IYT and 1Z0Q conformations using a probe radius of 10 Å). Our analysis finds correlation between the reduction in hydrophilic surface and experimental aggregation propensities. Finally, we show that experimental cell toxicities of Aβ variants are well described by computed SAP values, suggesting direct interplay between aggregation propensity and cell toxicity and providing a step toward a first computational estimator of Aβ toxicity. The present study contributes to our understanding of amyloid aggregation and suggests a method to predict aggregation propensity and toxicity of Aβ variants, and potentially to reduce aggregation propensities of amyloids by molecular intervention directed toward specific conformations of the peptides.