Defining the Neuropathological Aggresome across in Silico, in Vitro, and ex Vivo Experiments

Abstract

The loss of proteostasis over the life course is associated with a wide range of debilitating degenerative diseases and is a central hallmark of human aging. When left unchecked, proteins that are intrinsically disordered can pathologically aggregate into highly ordered fibrils, plaques, and tangles (termed amyloids), which are associated with countless disorders such as Alzheimer's disease, Parkinson's disease, type II diabetes, cancer, and even certain viral infections. However, despite significant advances in protein folding and solution biophysics techniques, determining the molecular cause of these conditions in humans has remained elusive. This has been due, in part, to recent discoveries showing that soluble protein oligomers, not insoluble fibrils or plaques, drive the majority of pathological processes. This has subsequently led researchers to focus instead on heterogeneous and often promiscuous protein oligomers. Unfortunately, significant gaps remain in how to prepare, model, experimentally corroborate, and extract amyloid oligomers relevant to human disease in a systematic manner. This Review will report on each of these techniques and their successes and shortcomings in an attempt to standardize comparisons between protein oligomers across disciplines, especially in the context of neurodegeneration. By standardizing multiple techniques and identifying their common overlap, a clearer picture of the soluble neuropathological aggresome can be constructed and used as a baseline for studying human disease and aging.

Figure 1.

A workflow depicting the 3D profile method, which (left) reads a protein sequence six residues-at-a-time and threads the amino acids onto an experimentally determined NNQQNY steric zipper backbone, (middle) adjusts XYZ in order to create a grid of relative orientations, and (right) utilizes RosettaDesign to energetically score each complex, take the minimum value, and compare this to a heuristic threshold that dictates amyloidogenicity. Reprinted with permission from Thompson et al. Copyright 2006 National Academy of Sciences.

Figure 2.

Microsecond molecular dynamics simulations reveal a number of structurally distinct Aβ42 tetramers on the surface of zwitterionic lipid bilayers, indicating that amyloid oligomers remain heterogeneous on the surface of plasma membranes. Spheres represent N- and C-termini, which do not have a fixed relative-position across multiple simulation replicates. Adapted with permission from Brown and Bevan. Copyright 2016 Elsevier.

Figure 3.

Size exclusion chromatography (A) and SDS-PAGE (B) under ideal and non-ideal experimental circumstances. (A1) Ideally, retention is determined solely by steric properties such that the larger particles elute first. (A2) Under non-ideal circumstances, retention is based on their interactions with the column media, which in this illustration are similar for the differently sized particles. (B1) Ideally, molecules migrate according to their size (proportional to molecular weight), as SDS binding erases information about charge and shape. Under non-ideal circumstances, SDS binding could promote disaggregation (B2) or aggregation (B3) leading to artifactually lower and higher molecular weight aggregates, respectively.

Figure 4.

Standard biophysical assays reveal structural characteristics of soluble Aβ aggregates in forward (monomer-grown) and backward (fibril-degraded) oligomers. While TEM, DLS, CD, and FTIR are useful for discerning oligomer morphology, size, and secondary structure, tools such as SEC or SDS-PAGE can report smaller structures than what is observed in TEM. Adapted with permission from Martins et al. Copyright 2007 John Wiley and Sons.

Figure 5.

Fluorescence microscopy data for Aβ42 monomers and oligomers (tagged with Alexa 647) over four timepoints on (A) glass or (B) lipid (POPC) surfaces. Single (C) and double (D) photobleaching events correspond to Aβ monomers (circle in A) and dimers (square in A), while exponential fluorescence decays (E) correspond to Aβ oligomers, which form preferentially on lipid surfaces. Reprinted with permission from Drews et al. Copyright 2016 Springer Nature.

Figure 6.

An improved technique for isolating and purifying soluble aβ aggregates from human AD brain segments. Most notable is the substitution of CHAPS for SDS (which may nucleate protein oligomerization) and the use of a flotation assay (addition of a sucrose cushion) during ultracentrifugation. Reprinted with permission from Esparza et al. Copyright 2016 Springer Nature.

Figure 7.

Solid-State NMR structures of ex vivo Aβ40 fibrils (A-D) reveal key differences from in vitro fibrils (F). These include a hollow triangular core with three-fold symmetry (E) compared to two-fold symmetric fibrils observed in vitro (F). Reprinted with permission from Lu et al. Copyright 2013 Cell Press.