Several neurodegenerative diseases are characterized by the aggregation and deposition of abnormally folded proteins, which has led to their being called protein conformational diseases. Included among these diseases are Alzheimer's disease, Parkinson's disease, polyglutamine diseases, and prion diseases such as Creutzfeld-Jakob disease (humans), scrapie (sheep) and bovine spongiform encephalopathy (mad cow disease). Although the proteins that are implicated in pathogenesis differ among these diseases, a common feature is the conversion of alpha-helical domains to beta-sheet domains. Of these protein conformational diseases, Alzheimer's disease is the most prevalent in the human population and it is the focus of this research proposal. However, the knowledge gained from the fundamental studies proposed herein will expand our understanding of other conformational diseases such as scrapie and mad cow disease.<P>
Specific objectives are: <ol><li> To examine the effect of the length of the Abeta peptide and depth of insertion on interactions with lipid bilayers. The rationale for this aim is that processing of APP results in Abeta fragments of different lengths, which have different tendencies to aggregate and to cause toxicity. <li> To determine the effect of pH on the interaction of Abeta peptides with lipid bilayers. The pH affects the ionization state of several amino acids in Abeta, and studies have reported conformational shifts in Abeta depending on pH. <li> To examine the effect of membrane composition on Abeta and its interaction with membranes. Experimental studies suggest that specific types of lipids, the charge associated with the bilayer, and cholesterol can all affect the interaction of Abeta with membranes. <li> To define interactions among multiple Abeta peptides within phospholipid bilayers. A proposed mechanism of toxicity is based on aggregation of 4 to 6 Abeta molecules to form ion channels in the plasma membrane. <li> To examine the effect of oxidized Met-35 (methionine sulfoxide) in Abeta on structure and aggregation of Abeta as well as its interaction with lipid bilayers. Oxidative stress is implicated in the etiology of Alzheimer's disease, and one manifestation of oxidative stress is oxidation of Met-35 to the sulfoxide. </ol>
NON-TECHNICAL SUMMARY: Alzheimers disease is a debilitating neurodegenerative disorder that affects millions of people. Other neurodegenerative diseases with similar characteristics include Parkinson's disease, polyglutamine diseases, Creutzfeld-Jakob disease (humans), scrapie (sheep) and bovine spongiform encephalopathy (mad cow disease). This project applies computer simulation methods to study several factors that may contribute to the biochemical changes that lead to the development of Alzheimers disease and expand our understanding of other devastating neurodegenerative diseases that affect humans and animals.
APPROACH: Molecular dynamics (MD) simulations will be applied in addressing the aims of this project. MD is routinely applied to simulations of the structure and dynamics of proteins and nucleic acids, and also is being used increasingly in simulating phospholipid bilayers and proteins in association with bilayers as a way to mimic protein interactions with membranes. Notable examples of MD simulations of proteins in phospholipid bilayers include simulations involving melittin, dynorphin peptides, and ion channels. A recognized limitation of MD simulations is the force field used to describe intramolecular and intermolecular interactions, though improvements in the quality of force fields have resulted in better agreement between experimental data and computed results. However, simulations must be based on experimental data, whether for establishing a starting point for a simulation, an ending target of the simulation, or both. Force fields for phospholipid bilayers are available, and the parameters have been incorporated into some molecular modeling packages. Of the commonly used packages, GROMACS has been used in many studies of lipid bilayers and will be the package used in the proposed research. A method within GROMACS to limit consideration of bond vibrations will be used so that a time step of 2 fs can be applied. Our molecular systems will include Abeta in a bilayer, surrounded by water molecules. The Particle Mesh Ewald (PME) method for calculating electrostatic interactions will be applied. The length of each simulation may vary depending on the progress of the simulation, but based on our initial study described above, simulations of at least 100-ns are not unreasonable. For example, the 100-ns simulation described above required approximately 3 weeks to run on 20 processors of the System X supercomputer; we are able to run multiple simulations simultaneously. Variables to be considered are the length of the Abeta peptide, depth of insertion in the bilayer, effect of ionization state, effect of membrane composition, and the effect of Abeta oxidation. Models of Abeta will be taken from the Protein Data Bank. Most pertinent for this project are Abeta(1-40) in micelles (PDB code 1BA4), Abeta(1-40) with Met-35 oxidized in micelles (1BA6), and Abeta(1-42) in hexafluoroisopropanol (1IYT). Peptides will be inserted into the bilayers using an in-house script we have developed. A hole is created in the bilayer that allows insertion of the peptide without generating unfavorable van der Waals contacts. After the peptide is inserted, the system is equilibrated to allow the lipids in the bilayer to relax around the peptide. A final general consideration regarding the simulations relates to the analyses that will be carried out. In particular, we will consider conformations of Abeta, interactions between Abeta and the lipid bilayer, and order parameters related to the phospholipids in the bilayer. These structural features are of interest because they relate to effects of the membrane on Abeta structure and vice-versa.