Drug Development.

Background: To improve clinical translatability of non-clinical in-vivo Alzheimer's disease (AD) models, a humanized APP knock-in mouse model (APP) was recently created (Xia, D. et al., 2022). The genetic modifications lead to increased Aβ42/40 ratios in AD relevant tissues, resulting in an age-dependent amyloid deposition. Here we assess the value of this model for non-clinical efficacy studies of experimental treatments with diverse mechanisms of action to facilitate the development of novel AD therapeutics.

Method: APP mice and WT controls were aged and sacrificed at various time points. In addition, APP mice were treated for 3 months from the age of 3 months onward with an NLRP3-inhibitor, an in-house developed calcium homeostasis modulator or vehicle. The major pathological hallmarks were investigated with biochemical and immunohistological assays. Additionally a preliminary screen for AD biomarkers was conducted in CSF samples. Since APP mice lack an obvious behavioral phenotype, synaptic plasticity was investigated to serve as a functional readout.

Result: Cortical Aβ40, Aβ42 and pyroglutamate modified Aβ42 (N3pE-42) levels were shown to be affected with age. In addition, cortical amyloid plaques, surrounded by activated microglia, were shown to be present already at an age of 3 months, albeit at a low level. With age, the number of cortical plaques increased. In addition, we were able to confirm the presence of endogenous phospho-Tau positive dystrophic neurites within these plaques, clearly demonstrating the translational relevance of this model. Interestingly, several of the observed hallmarks were counteracted by compound intervention, including phospho-Tau pathology. Finally, we show that several CNS biomarkers can be detected in mouse CSF.

Conclusion: These data lead us to conclude that this knock-in model is a broadly applicable tool to investigate efficacy of disease-modifying drugs with diverse mechanisms of action even without directly targeting protein aggregates.