Assessment of the Molecular Determinants Required for Dimerization of the Amyloid Precursor Protein Transmembrane Domain by a Combined Experimental and Computational Approach

Abstract

Removal of the ectodomain of the amyloid precursor protein (APP) by beta-secretase yields a carboxyl-terminal fragment (betaCTF) that is then degraded by the gamma-secretase to produce the neuropathogenic amyloid beta peptides (Abetas) involved in Alzheimer's disease (AD). Considerable evidence indicates that betaCTF is a transmembrane domain-mediated dimer and that dimer dissociation reduces the Abeta42/40 ratio, thus lowering the risk of AD. Little is known about the structural and thermodynamic features of betaCTF dimerization. Employing both coarse-grained and all-atom models coupled to metadynamics, we studied the free energy of dimerization of betaCTF transmembrane domain in an explicit dipalmitoyl phosphatidyl choline (DPPC) membrane bilayer. We show that the dimeric state consists of several stable configurations, featuring interfaces at different locations in the betaCTF transmembrane helix. The effect that point mutations along the transmembrane helix have on the stability of the dimer is assessed computationally and compared to experimental data. This information sets the stage to identify interface-derived small peptides as lead structures to guide the development of novel peptidomimetics that specifically target betaCTF dimerization and display therapeutic potential in AD.