Characterization of the knock-in App SAA mouse model as a powerful tool for Alzheimer’s disease preclinical research

BACKGROUND Preclinical Alzheimer’s disease (AD) research requires models with age-dependent, physiologically relevant amyloid pathology, unlike overexpression in transgenic lines. We characterized the humanized App SAA knock-in model to define translationally relevant therapeutic windows. METHODS Homozygous App SAA and control mice were evaluated from 4 to 19 months using whole-brain 3D plaque imaging, cerebral amyloid angiopathy (CAA) quantification, multiplex neuropathology, brain transcriptomics, and plasma biomarkers. Hippocampal synaptic analysis and context-dependent memory was performed at 12 months. RESULTS Plaques, CAA, plaque-associated neuroinflammation, and dystrophic neurites accumulated with age. By 12 months, brain transcriptomes aligned with human AMP-AD co-expression modules, and plasma pTau-217, GFAP, and SNAP25 rose with age. App SAA mice retained associative memory but showed memory interference and CA1 dendritic-spine loss that was exacerbated in proximity to plaques. CONCLUSIONS App SAA mice recapitulate age-dependent, human-relevant AD phenotypes with defined therapeutic windows for preclinical research. The model is available without restriction from The Jackson Laboratory.