Abstract
The formation of amyloid filaments is characteristic of various degenerative diseases. Recent breakthroughs in electron cryo-microscopy (cryo-EM) have led to atomic structure determination of multiple amyloid filaments, both of filaments assembled in vitro from recombinant proteins, and of filaments extracted from diseased tissue. These observations revealed that a single protein may adopt multiple different amyloid folds, and that in vitro assembly does not necessarily lead to the same filaments as those observed in disease. In order to develop relevant model systems for disease, and ultimately to better understand the molecular mechanisms of disease, it will be important to determine which factors determine the formation of distinct amyloid folds. High-throughput cryo-EM, in which structure determination becomes a tool rather than a project in itself, will facilitate the screening of large numbers of in vitro assembly conditions. To this end, we describe a new filament picking algorithm based on the Topaz approach, and we outline image processing strategies in Relion that enable atomic structure determination of amyloids within days.
Highlights
Amyloids are lamentous, helical aggregates of proteins that are characterised by a cross-b-sheet quaternary structure
O en the helical symmetry of amyloids is incompatible with crystallisation, and their size precludes solution-state nuclear magnetic resonance (NMR)
Before the advent of atomic structure determination by cryo-EM, amyloids were mainly studied by solid-state NMR, which requires large amounts of 13C 15N labelled protein.[2,3]
Summary
Amyloids are lamentous, helical aggregates of proteins that are characterised by a cross-b-sheet quaternary structure. Paper so ware.[4,5] Application of the same technique to amyloids wasn’t successful until the implementation of helical symmetry in the Relion program.[6] Using this approach, the rst cryo-EM structure of an amyloid to sufficient resolution for atomic modelling was reported for laments of the protein tau that were extracted from the brain of an individual with Alzheimer’s disease.[7] Typically of lower resolution, larger 2D class averages facilitate identi cation of distinct lament types, can help measuring cross-over distances, and reduce the number of variables to optimise in the relion_helix_inimodel2d program for initial model generation.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
