Developing a functionally valid model of the TREM2-APOE complex to better understand its role in Alzheimer's disease.

Abstract

Triggering Receptor Expressed on Myeloid Cells 2 (TREM2) and Apolipoprotein E (ApoE) are the most potent genetic risk factors for late-onset Alzheimer's disease (AD). ApoE can act as a ligand of TREM2, activating an immune response around AD-associated pathology in the brain; however, without a robust model of the binding between the two proteins, it is difficult to study the mechanism of this interaction or to target it for drug development. To define the structural mechanism by which AD-associated variants impaired TREM2's function as an innate immune receptor, we used a combination of sequence-based binding site prediction, in silico molecular modelling, and biophysical validation with biolayer interferometry (BLI) to develop an initial all-atom model for the TREM2-ApoE complex. Sequence-based binding site prediction and BLI both identify the apical hydrophobic site of TREM2 and hinge region of ApoE as the likely binding sites for the complex. Early models from in silico docking and molecular simulations at these sites suggest formation of a stable complex model and reveal TREM2 variant-dependent changes in the conformation and dynamics of ApoE. Extended simulations using site-specific mutations of TREM2 in models with restraints optimized based on BLI results are also in progress, which will be further validated by comparing predicted binding affinities from the simulated complex to experimental affinities measured with BLI. Our work linking AD-associated variants to instability in TREM2's hydrophobic site, where we find ApoE to bind, provides further evidence that the loss-of-function caused by these variants is likely driven by impaired binding of ligands like ApoE. This work also provides further justification for targeting the TREM2-ApoE interaction as a potential therapeutic for AD, while simultaneously providing a model that may be useful for developing and refining such therapeutics.