Nearly 5 million people in the US alone are afflicted with Alzheimer's disease, and there is currently no means to prevent or treat the disease. Though the molecular basis of the disease is unclear, the deposition of amyloid plaques in the brain is a key hallmark of the disease. These plaques are widely believed to be pathogenic, and tremendous efforts have focused on the mechanism governing their formation. Amyloid plaques are composed of the 42-residue amyloid peptide (A), which is generated through proteolytic cleavage of the amyloid precursor protein (APP) by -secretase. This enzyme is localized within cholesterol-rich lipid raft membrane domains, while the APP substrate exists in both the fluid-phase and lipid raft domains of cellular membranes. Therefore, the efficiency with which the A peptide is generated may be governed by the distribution of APP between the raft and non-raft membrane domains. Recent work in the Sanders lab on the 99-residue C-terminal fragment of APP (C99) has revealed a cholesterol binding pocket within the TM domain (present in APP). These studies have confirmed that C99 binds cholesterol with high affinity in bilayers, which suggests that the localization of C99 (and APP) may depend on the distribution of cholesterol in biological membranes. We recently tested this hypothesis by characterizing the localization of C99 within phase-separated giant unilamellar vesicles (GUVs), which contain both fluid phase (L?) and liquid-ordered (Lo) domains. The results show that C99 is specifically localized within raft-like Lo domains. However, C99 variants carrying mutations that abolish cholesterol binding strongly prefer the non-raft L? phase. This confirms cholesterol binding directly affects the distribution of C99 within the membrane. The most intuitive explanation for this phenomenon is that C99 is driven into the Lo domain due to the high concentration of the cholesterol ligand, which leads to favorable binding energetics. However, thermodynamic evaluations of this partitioning suggest that binding energetics cannot account for the observed differences. Thus, the physical mechanism for this coupled binding and partitioning remains unclear. In the following, I propose a series of experiments aimed at dissecting the energetic contributions of both the membrane and the protein in the coupled binding and partitioning of C99. I will first use EPR spectroscopy to assess the binding energetics of cholesterol in Lo and L? like membranes. The results will suggest whether the cholesterol binding energetics are sensitive to changes in the bilayer. Next, I will use protein engineering and confocal fluorescence microscopy to determine how differences in the length and rigidity of the TM domain affect its partitioning. These studies will reveal the structural features of C99 that are critical for its sorting within te membrane. Finally, I will examine the structural dynamics of free and cholesterol-bound C99s using solution NMR in both Lo and L? like bicelles in order to determine how bilayers affect its binding mode. Together, the results will provide novel insights into the molecular basis of Alzheimer's disease and elucidate the molecular determinants of protein sorting within the membrane.
The distribution of the amyloid precursor protein (APP) in cellular membranes has been implicated in the pathogenesis of Alzheimer's disease. To gain insight into physical determinants of the partitioning of APP between membrane domains, we will investigate how the interplay between the structure of the membrane and the structure of the TM domain influence its partitioning into lipid raft membrane domains. The results will provide insights into the molecular basis of Alzheimer's disease as well as basic insights into the mechanisms for protein sorting within the membrane.