Neurodegenerative Disease–Associated Variants in TREM2 Destabilize the Apical Ligand-Binding Region of the Immunoglobulin Domain

Single nucleotide variations in Triggering Receptor Expressed on Myeloid Cells 2 (TREM2) have been linked to both late-onset Alzheimer's disease and behavioral variant frontotemporal dementia (FTD), the latter presenting either in isolation or with cystic bone lesions in a condition called Nasu-Hakola disease. Models of the extracellular domain of TREM2 show that Nasu-Hakola disease–associated mutations are grossly inactivating by truncation, frameshift, or unfolding, that Alzheimer's disease (AD)–associated variants localize to a putative ligand-interacting region (PLIR) on the extracellular surface, and that FTD-associated variants are found in the hydrophobic core. However, while these disease-associated residues are predicted to play some role in disrupting ligand binding to the extracellular domain of TREM2, how they ultimately lead to disease remains unknown. Here, we used in silico molecular modeling to investigate all-atom models of TREM2 and characterize the effects on conformation and dynamical motion of AD-associated R47H and R62H as well as FTD-associated T96K, D86V, and T66M variants compared to the benign N68K variant and the common variant. Our model, which is based on a published 2.2 Å resolution crystal structure of the TREM2 extracellular domain, finds that both AD- and FTD-associated variants cause localized instability in three loops adjacent to the PLIR that correspond to the complementarity-determining regions (CDRs) of antibodies. This instability ultimately disrupts tethering between these CDRs and the core of the immunoglobulin domain, exposing a group of otherwise-buried, negatively charged residues. This instability and exposure of negatively charged residues is most severe following introduction of the T66M variant that has been described as causing FTD even in the heterozygous state and is less severe following introduction of variants that are less strongly tied to FTD or of those associated with AD. Thus, our results provide further evidence that the proposed loss-of-function caused by neurodegenerative disease–associated variants may be driven by altered conformational stability of the ligand-interacting CDR and, ultimately, loss of affinity or specificity for TREM2 ligands.