The allure and pitfalls of the prion-like aggregation in neurodegeneration

Prion diseases are fatal neurodegenerative disorders where the formation of amyloids is thought to be infectious by templating their conformation on to natively-folded counterparts. Postulated nearly four decades ago, the search for the mechanism behind the conformational templating has proceeded to no avail. Here, we extend the thermodynamic hypothesis of protein folding (Anfinsen's dogma) to the amyloid phenomenon and illustrate that the amyloid conformation (cross-β) is one of two conformational states that are thermodynamically accessible to any protein sequence depending on concentration. A protein spontaneously assumes its native conformation below supersaturation and the amyloid cross-β conformation above supersaturation. The information to adopt the native conformation and the amyloid conformation is present in the primary sequence and the backbone of the protein, respectively, and does not require templating. The rate-limiting step for proteins to adopt the cross-β conformation of amyloid is termed nucleation, which can be catalyzed by surfaces (heterogeneous nucleation) or preformed amyloid fragments (seeding). Irrespective of the nucleation pathway, once triggered, amyloid formation proceeds spontaneously in fractal-like fashion, where the surfaces of the growing fibrils act as heterogeneous nucleation catalysts for new fibrils, a phenomenon known as secondary nucleation. This pattern is in contrast to the linear growth assumptions that the prion hypothesis necessitates for faithful prion strain replication. Additionally, the cross-β conformation buries the majority of the protein side chain inside the fibrils, making the fibrils inert, generic, and extremely stable. As such, the source of toxicity in prion disorders may come to a greater extent from the loss of proteins in their normal, soluble, and therefore functional state rather than from their transformation into stable, insoluble, nonfunctioning amyloids.