(323c) Exploration of Interresidual Contacts in Combined Coarse Graining and Experiment Directed Simulations of A?42

Experiment Directed Simulations (EDS) and coarse-grained molecular dynamics (MD) were combined to improve the simulation of Aβ42 peptide fragments and to match their computed chemical shifts to experimental results. EDS is a maximum entropy biasing technique that minimally changes (biases) the potential energy of a simulation so that some constraint is satisfied. Constraints are usually an agreement between average position dependent variables from experiments and simulations. EDS has been found to indirectly improve dynamical and structural properties in different systems like peptides, water models, proteins, and lithium ions. EDS has been used in ab-initio, atomistic MD, and coarse-grained MD simulations. Coarse-grained EDS simulations improved the accuracy of actin monomer simulations by reproducing the fluctuations present in the larger actin filaments. In this work, EDS and coarse-graining are used to understand intramolecular and intermolecular interactions in Aβ42. Aβ42 is a 42-residue long amyloid peptide that is associated with Alzheimer’s disease. Atomistic MD simulations of multiple chains of 21-30 fragment of amyloid peptide with EDS resulted in their self-assembly previously. Knowing the process of self-assembly in whole Aβ42, including the dynamics and governing interactions will help find therapeutics that halt these processes. In order to study these interactions in Aβ42, we divided Aβ42 into 5 fragments (each fragment consists of 10 residues) with overlapping ends and simulated all of the possible combinations of these 5 fragments in separate simulations. We used chemical shifts from nuclear magnetic resonance spectroscopy results to bias the fragments with chemical shifts from the whole Aβ42 peptide in EDS. All-atom MD simulations were performed with EDS and enhanced sampling in GROMACS simulation engine. We used force matching bottom-up approach to map all-atom forces to a mapped coarse-grained potential. We classified which fragments and which residues interact the most with each other. When EDS bias had been applied, we saw behavior that was predicted in the past experimental literature. During coarse-graining, the interactions were observed between specific fragments of Aβ42 and not with others, which could lead to self-assembly in the full Aβ42.