Biophysical studies on aggregation processes and amyloid fibrils with focus on Alzheimer's disease

Abstract

Alzheimer's disease (AD) is a fatal neurodegenerative disease, and is the most common form of dementia, affecting 1-2% of the population in industrialized countries. It is believed that intermediate species formed during amyloid Bpeptide (AB) polymerization play a key role in the pathogenesis of AD, but it is not known how such toxic effects are exerted. A detailed model of the initial aggregation of AB, as well as the structure of amyloid fibrils, is a significant goal for the understanding of the molecular mechanisms behind Alzheimer's disease.

The aim of this thesis was to I) study AB polymerization and interaction with cell membranes using fluorescence correlation spectroscopy (FCS). II)Utilizing the high sensitivity of FCS for developing a method for detecting and counting single AB aggregates in cerebrospinal fluid (CSF) from AD patients and controls. III)Delineating the minimum structural requirements for amyloid fibril formation in general and assessing the alignment of strands within AB-fibrils using model peptides.

This thesis is based on six papers. In paper I fibrin polymerization was used as a model system. The spatially confined detection volume used in FCS allowed the detection of diffusing species within the gel network, both byproducts of the activation of fibrinogen (fibrinopeptide A) and fibrin oligomers. The diffusion coefficient of fibrinogen found was consistent with previously reported values. In paper II the interaction between AO and the cell membrane of erythrocytes was studied. It was demonstrated that AB was inserted into, or bound to, the cell membrane. Unlike a fluorescent lipid tracer, the membrane-inserted AB displayed two characteristic diffusion times, one similar to the lipid tracer and one slower. This could be explained either as polymerization of AB within the membrane, or as partition of AB into different domains of the membrane.

In paper III a modified FCS setup combined with a microfluidic system was developed for detecting single AB aggregates in CSF from patients and controls. No significant difference between cases arid controls was found, however. In paper IV several of the possible alignments of the peptide strands in AB fibrils were evaluated by constructing conformationally constrained dimers, each with a fixed alignment between the two constituent strands. The results demonstrated that different alignments could exist in fibrils. The results also indicated that both hydrophobic and electrostatic interactions are important in fibril formation.

In paper V amyloid fibrils were formed from peptide as short as four residues, but not shorter sequences. By modifying the sequence it was shown that sidechain charge attraction and high B-propensity of the amino acids were required for fibril formation. Based on this, peptides that formed fibrils in mixtures but not on their own were designed. In paper VI fibrilization competent tetrapeptides were fused to form open or beta-hairpin like structures. It was shown that the closed conformation was unable to form fibrils, instead it formed stable oligomers. Fibril formation thus depends on the structural context of the arnyloidogenic sequence.