Proteins Do Not Replicate, They Precipitate: Phase Transition and Loss of Function Toxicity in Amyloid Pathologies

Simple Summary Amyloid aggregation of proteins in disease has been known for over a hundred years; however, effective therapeutics for amyloid pathologies such as Alzheimer's disease and Parkinson's disease are still lacking. This review divides the amyloid phenomenon into four major questions: What are amyloids? How do amyloids form? Can proteins replicate? How do amyloids cause toxicity? The aim is to answer these questions within a unified physicochemical framework that links the structural biology of amyloids to the thermodynamics of amyloid formation and the pathophysiology of amyloid aggregates in different diseases. We illustrate that the thermodynamics of protein aggregation does not support the prion protein-only replication hypothesis, and how the structural biology of amyloids makes them largely domainless, generic, and inert. The implications of this understanding for the etiology, pathogenesis and potential therapeutics of amyloid diseases are briefly discussed. Protein aggregation into amyloid fibrils affects many proteins in a variety of diseases, including neurodegenerative disorders, diabetes, and cancer. Physicochemically, amyloid formation is a phase transition process, where soluble proteins are transformed into solid fibrils with the characteristic cross-beta conformation responsible for their fibrillar morphology. This phase transition proceeds via an initial, rate-limiting nucleation step followed by rapid growth. Several well-defined nucleation pathways exist, including homogenous nucleation (HON), which proceeds spontaneously; heterogeneous nucleation (HEN), which is catalyzed by surfaces; and seeding via preformed nuclei. It has been hypothesized that amyloid aggregation represents a protein-only (nucleic-acid free) replication mechanism that involves transmission of structural information via conformational templating (the prion hypothesis). While the prion hypothesis still lacks mechanistic support, it is also incompatible with the fact that proteins can be induced to form amyloids in the absence of a proteinaceous species acting as a conformational template as in the case of HEN, which can be induced by lipid membranes (including viral envelopes) or polysaccharides. Additionally, while amyloids can be formed from any protein sequence and via different nucleation pathways, they invariably adopt the universal cross-beta conformation; suggesting that such conformational change is a spontaneous folding event that is thermodynamically favorable under the conditions of supersaturation and phase transition and not a templated replication process. Finally, as the high stability of amyloids renders them relatively inert, toxicity in some amyloid pathologies might be more dependent on the loss of function from protein sequestration in the amyloid state rather than direct toxicity from the amyloid plaques themselves.