Abstract
A variety of species express the amyloid β-protein (Aβ (the term "Aβ" refers both to Aβ40 and Aβ42, whereas "Aβ40" and "Aβ42" refer to each isoform specifically). Those species expressing Aβ with primary structure identical to that expressed in humans have been found to develop amyloid deposits and Alzheimer's disease-like neuropathology. In contrast, the Aβ sequence in mice and rats contains three amino acid substitutions, Arg5Gly, His13Arg, and Tyr10Phe, which apparently prevent the development of AD-like neuropathology. Interestingly, the brush-tailed rat, Octodon degus, expresses Aβ containing only one of these substitutions, His13Arg, and does develop AD-like pathology. We investigate here the biophysical and biological properties of Aβ peptides from humans, mice (Mus musculus), and rats (Octodon degus). We find that each peptide displays statistical coil → β-sheet secondary structure transitions, transitory formation of hydrophobic surfaces, oligomerization, formation of annuli, protofibrils, and fibrils, and an inverse correlation between rate of aggregation and aggregate size (faster aggregation produced smaller aggregates). The rank order of assembly rate was mouse > rat > Aβ42. The rank order of neurotoxicity of assemblies formed by each peptide immediately after preparation was Aβ42 > mouse ≈ rat. These data do not support long-standing hypotheses that the primary factor controlling development of AD-like neuropathology in rodents is Aβ sequence. Instead, the data support a hypothesis that assembly quaternary structure and organismal responses to toxic peptide assemblies mediate neuropathogenetic effects. The implication of this hypothesis is that a valid understanding of disease causation within a given system (organism, tissue, etc.) requires the coevaluation of both biophysical and cell biological properties of that system.