Biophysical Studies of Amyloid Fiber Stability

Abstract The doctoral work presented here deals with amyloid fiber stability. The 42 residues amyloid beta-peptide, Abeta42, which forms fibers in the brains of Alzheimer's disease patients, was chosen as a preferred model system for amyloidosis due to the clear biomedical implications and the wealth of literature available on its amyloid formation. Other fiber forming peptides, such as the 37 amino acids Human Islet Amyloid Polypeptide (involved in type II diabetes) and the hexapeptides STVIIE (synthetic), KVQIIE (from Tau protein sequence) and ISFLF (from a prion protein sequence) were also studied. All peptides used are highly amyloidogenic and are stable as fibers in aqueous solution under near physiologic conditions (neutral pH, low salt). Nevertheless, biophysical assays revealed that the amyloid fibers generated from these peptides could be disassembled into protofibrils and smaller oligomers by exposure to phospholipidic vesicles. Since lipid induced fiber disassembly occurred with several different amyloid sequences, this phenomenon might be a general property of lipid-amyloid interactions and bear general importance to amyloid-related disease. Abeta42 fibers interaction with lipid vesicles was studied in detail. Initial results indicated that lipid induced protofibrils were cytotoxic. This finding led to a collaboration with Prof. Bart De Strooper's laboratory (KUL, Leuven, Belgium), where the toxicity of lipid-exposed fibre samples was confirmed both in vivo and in vitro. Given the relevance of these findings, the effect of the lipid vesicle composition on the Abeta42 fiber stability was studied. It was found that vesicles enriched in brain phospholipids, such as sphyngomyelin and monoganglioside, induced more fiber disassembly and neurotoxicity, suggesting a possible link with the cognitive decline and neurodegeneration that is typical of Alzheimer's disease. Since Abeta42 amyloid fibers may not be as stable as anticipated, additional experiments on their stability were conducted. More specifically, as the amyloid plaques typical of Alzheimer's disease are mostly constituted of Abeta40 and Abeta42 peptides (respectively, 40 and 42 residues long), fiber structure in co-incubation experiments of both species was analyzed. It was found that, while a physiological ratio of 9/1 Abeta40/Abeta42 produces typical amyloid fibers, a pathological ratio of 7/3 Abeta40/Abeta42 results in the presence of protofibrils, possibly explaining the earlier disease onset associated to this ratio. As it became evident that fiber stability is a clear issue in Abeta42 toxicity, the key regions were mapped through a combined computational and experimental approach. It was found that the 18-24, the 30-35 and the 40-42 regions are key for fiber stability. These are therefore optimal target areas for research aiming at the inhibition of A? toxicity in Alzheimer's disease. In summary, it is clear that the role of amyloid fiber stability in disease should be considered with care. These original findings are of general impact for the field of protein aggregation and amyloidosis, paving the way to further research on the details of fiber stability in Alzheimer's Disease and in other related amyloid diseases.