Paucity of Amyloid Nuclei Defy Isolation and Toxicity Evaluation

Abstract

The molecular rearrangement of soluble proteins into fibers is a common attribute of amyloid diseases: Alzheimer's disease, Parkinson's disease, spongiform encephalopathies (including mad cow disease), and other prion diseases. In the past century considerable progress has been made in characterizing amyloid diseases, but the connection between amyloidosis and the disease is still unclear: Contradictory reports suggest that the fibrils and/or the oligomer precursors cause toxicity. The debate is still open. Here, we offer the first attempt to “reverse-estimate” the concentrations of nuclei, starting from a distribution of fibril lengths. Assuming the nucleation model is valid, with a few reasonable assumptions, a fibril length distribution and a set of seeding experiments, we estimated the in vitro concentration of nuclei for the model hormone, recombinant human insulin, to be in the picomolar range. Fibril lengths are measured with an atomic force microscope and seeding shows that fibrils of different lengths exhibit similar growth rates. Because of their propensity to form aggregates (non-ordered) and fibrils (ordered), this very low concentration could explain the difficulty in fractionating, isolating and blocking nuclei toxicity. Moreover, this theoretical approach, based on our measurements and a structural fibril model recently published by David Eisenberg's group at UCLA, is general and could be used for other amyloid proteins.