Excitatory and inhibitory neurons accumulate distinct Ab plaques in mouse models of APP overexpression

Amyloid plaque heterogeneity is a hallmark of Alzheimer's disease (AD), yet the mechanisms governing plaque diversity and its relationship to disease progression remain poorly understood. Although both diffuse and neuritic plaques are present in AD brain tissue, we do not know how these distinct conformations arise or influence pathological outcomes. Here, we developed transgenic mouse models expressing identical APP constructs in either glutamatergic or GABAergic neurons to investigate how cellular context shapes amyloid pathology. This approach revealed that the cellular source of APP influences both the biochemical composition and structural characteristics of resulting Aß deposits. APP expressed in GABAergic neurons generated exclusively diffuse plaques with elevated Aß42/Aß40 ratios that failed to induce gliosis, while glutamatergic expression produced primarily neuritic plaques surrounded by activated microglia. Differences in deposit characteristics appeared to arise from fundamental variations in APP processing between excitatory and inhibitory neurons. Moreover, only neuritic plaques generated by glutamatergic neurons caused behavioral deficits, while diffuse plaques from GABAergic neurons did not impair cognition despite substantial amyloid burden. These findings may provide new insight into the mechanisms underlying amyloid heterogeneity in AD and suggest that plaque conformation, rather than simply amyloid load, may influence cognitive trajectories. snRNA-seq revealed that microglia from the excitatory models showed many DAM genes were upregulated, which were absent in the inhibitory model. The data also revealed there were no changes in expression of key APP processing enzymes between excitatory or inhibitory neurons in each model. Overall design: Single nucleus RNA sequencing (snRNA-seq) of frontal cortexes from CAMKII-, Vglut1-, or Gad2-driven APP transgenic mice at 6 mo, 9 mo, and 24 mo, respectively. Animal ages were chosen based on amyloid load. Each sample contained 2 animals.