How do membranes initiate Alzheimer's Disease? Formation of toxic amyloid fibrils by the amyloid β-protein on ganglioside clusters.

Alzheimer's disease (AD), a severe neurodegenerative disorder, causes more than half of dementia cases. According to the popular "Aβ hypothesis" to explain the mechanism of this disease, amyloid β-peptides (Aβ) of 39-43 amino acid residues aggregate and deposit onto neurons, igniting the neurotoxic cascade of the disease. Therefore, researchers studying AD would like to elucidate the mechanisms by which essentially water-soluble but hydrophobic Aβ aggregates under pathological conditions. Most researchers have investigated the aggregation of Aβ in aqueous solution, and they concluded that the final aggregation product, the amyloid fibrils, were less toxic than the component peptide oligomers. They consequently shifted their interests to more toxic "soluble oligomers", structures that form as intermediates or off-pathway products during the aggregation process. Some researchers have also investigated artificial oligomers prepared under nonphysiological conditions. In contrast to these "in solution" studies, we have focused on "membrane-mediated" amyloidogenesis. In an earlier study, other researchers identified a specific form of Aβ that was bound to monosialoganglioside GM1, a sugar lipid, in brains of patients who exhibited the early pathological changes associated with AD. This Account summarizes 15 years of our research on this topic. We have found that Aβ specifically binds to GM1 that occurs in clusters, but not when it is uniformly distributed. Clustering is facilitated by cholesterol. Upon binding, Aβ changes its conformation from a random coil to an α-helix-rich structure. A CH-π interaction between the aromatic side chains of Aβ and carbohydrate moieties appended to GM1 appears to be important for binding. In addition, as Aβ accumulates and reaches its first threshold concentration (Aβ/GM1 = ∼0.013), aggregated β-sheets of ∼15 molecules appear and coexist with the helical form. However, this β-structure is stable and does not form larger aggregates. When the disease progresses further and the Aβ/GM1 ratio exceeds ∼0.044, the β-structure converts to a second β-structure that can seed aggregates. The seed recruits monomers from the aqueous phase to form toxic amyloid fibrils that have larger surface hydrophobicity and can contain antiparallel β-sheets. In contrast, amyloid fibrils formed in aqueous solution are less toxic and have parallel β-sheets. The less polar environments of GM1 clusters play an important role in the formation of these toxic fibrils. Membranes that contain GM1 clusters not only accelerate the aggregation of Aβ by locally concentrating Aβ molecules but also generate amyloid fibrils with unique structures and significant cytotoxicity. The inhibition of this aggregation cascade could be a promising strategy for the development of AD-modulating therapies.