John Kremer, PhD 2001

Thesis Title: Interactions between beta-amyloid peptide and model neuronal membranes

β-amyloid (Aβ) peptide is the primary proteinaceous component of senile plaques, one of the hallmark characteristics of Alzheimer's Disease (AD). Aggregation of Aβ into fibrils is toxic to neurons, but the mechanism of toxicity remains unproven.

In this project, we investigated the interactions between Aβ aggregates and model neuronal membranes, specifically single- and multi-component phospholipid liposomes. We proposed that the aggregation of Aβ leads to a non-specific, but direct, association between Aβ aggregates and neuronal membranes.

To investigate this hypothesis, we characterized Aβ aggregate size and hydrophobicity, and correlated changes in the aggregation state of Aβ with its ability to bind to and penetrate into the phospholipid bilayer interior. Aβ aggregation at physiological pH generated soluble fibrils that elongated for several days and possessed surface-exposed hydrophobic area. At endosomal pH, Aβ aggregation proceeded very rapidly and produced insoluble, hydrophobic, and amorphous aggregates.

Aβ-membrane binding was strongly dependent on Aβ aggregation state. Specifically, small Aβ species (monomers and dimers) did not bind to membranes, while aggregated Aβ readily attached to liposomes. Aβ aggregated at acidic pH bound to liposomes even more readily. We propose that the principal driving force of Aβ-membrane interactions is the relative insolubility of Aβ aggregates in aqueous solutions.

We further explored the consequences of Aβ-membrane interactions with respect to changes in membrane structure. Aβ aggregates considerably decreased the fluidity in the lipid bilayer center in a manner that correlated to the amount of surface-exposed hydrophobic area on the Aβ aggregates. Aβ-membrane binding was necessary, but not sufficient, to induce these membrane structural changes. We hypothesize that Aβ-membrane interactions are mediated by hydrophobic association between surface-exposed hydrophobic areas on Aβ aggregates and the lipid bilayer interior.

Taken together, our investigation into Aβ-membrane interactions strongly endorses our original hypothesis that Aβ aggregation leads to a non-specific, but direct, interaction between Aβ aggregates and model neuronal membranes. Further, these interactions are principally the result of the marginal solubility of Aβ aggregates, and Aβ-membrane interactions are mediated primarily by hydrophobic association. Our results suggest that Aβ-induced changes in membrane structure may be associated with the pathology seen in AD.