Analyzing the binding affinity of potential homodimers produced by the C55 portion of the amyloid precursor protein, a molecular dynamics study.

Abstract

Alzheimer's disease is a neurodegenerative disease that affects millions of people in the United States alone. Agglomeration of amyloid beta (Aβ) peptides is a potential cause of the disease and the loss of mental acuity it induces. Aβ is created via the cleavage of the amyloid precursor protein (APP). When APP is cleaved, part of it—approximately the C99 fragment—remains in the lipid membrane, with its N-terminus being the Aβ peptide. The dimerization of C99 is theorized to limit its further cleavage and thereby limit the production of Aβ. However, this has yet to be proven experimentally. The first 55 residues of the C99 fragment (called C55) contains both the transmembrane domain and Aβ. In this work, we explored multiple potential conformations of C55 dimers (starting from the PDB ID 2LP1) and calculated their binding affinity to better understand the viability of homodimerization. The homodimer conformations were explored using both implicit and explicit water Molecular Dynamics simulations. We used an Extended Adaptive Biasing Force method to calculate the binding affinity of these dimers in a POPC lipid membrane. This methodology is characterized by the use of a continuously updated biasing force that cancels all force along a specified reaction coordinate and, in this way, enables uniform sampling on a flat free-energy surface. Applying this methodology using collective variables for root-mean-square deviation, spin, tilt, and Euclidean distance for both the bulk (monomer) and bound (dimer) states, we compute a binding affinity of roughly −4 kcal/mol for one of our dimer conformations. This lends in silica credence to the hypothesis that the dimerization of C55 limits the formation of Aβ, as the dimer appears energetically viable.