DESCRIPTION (provided by applicant): The association of proteins of known sequence is believed to be the principal cause of Alzheimer's and other neurodegenerative diseases. Upon aggregation, certain proteins form fibrillar structures that have been implicated as necessary pathogenic factors. The mechanisms of aggregation of polypeptide chains that lead to the formation of fibrillar structures is largely unknown. To date, experiments have provided only low resolution structures of the aggregated states of certain plaque forming peptides. In addition, the nature of conformational fluctuations leading to aggregation of polypeptide chains has not been fully elucidated. The long-term goal of this research is the elucidation of the fundamental principles responsible for polypeptide association and fibrillar formation. To achieve this goal, the PI's propose a multifaceted approach that includes the development and use of novel computational methods. They will employ all atom molecular dynamics simulations to probe the aggregation mechanism in Abeta-peptide (structured as monomeric peptide in water) and human amylin (structureless in the monomeric state). This will enable them to monitor in detail the nature of initiating (nucleating) structures responsible for fibril formation. The complete characterization of the peptide association pathways will be achieved through the direct simulation of protein-protein association, in conjunction with the application of reaction pathway algorithms to explore protein dimerization and fibril elongation. To discover the general principles governing amyloid formation, Dr. Straub will supplement the detailed molecular dynamics simulations with studies involving coarse grained models of polypeptides. This is necessary due to the prohibitively long times required for all atom simulations to examine a large number of peptide sequences. Using off-lattice and lattice models (with side chains), Dr. Straub proposes to examine the phase behavior, energetics and kinetics of peptide association. A further objective is to probe the role of folding intermediates in facilitating aggregation. Simplified models will be used to examine how the details of peptide sequence and initial conditions influence fibril formation. Preliminary results suggest that both atomistic and coarse-grained models can be effective tools for exploring the details of protein dynamics and equilibriums that arise in the aggregation process. The proposed combination of computational approaches will lead to a conceptual understanding of aggregation, at the molecular level, in polypeptide chains that is central to the understanding of amyloid diseases.