D. BRUNNER, T. HANANIA, M. MAZZELLA, H. HAIN, E. SABATH, V. ALEXANDROV, J. BERGER, P. KABITZKE, K. COX, M. WINDISCH
Three popular models for Alzheimer’s disease were investigated using behavioral high-throughput phenotyping technologies: : the single transgenic Tg 2576 (mutant APP) mouse model, the double transgenic model from the cross between the Tg 2576 and a mutant presenilin (PS1) line both expressing increased amyloid- beta 40 and 42 levels and amyloid pathology, and the rTg4510 mice, overexpressing human tau with the P301L mutation under control of a tetracycline responsive transacting element, showing cognitive impairment, motor deficits, tau hyperphosphorylation, neurofibrillary tangles, and neuronal loss in the forebrain. We exploited the option to switch off tau expression by doxycycline in the Tg4510 model, as a calibration for maximal effect of a putative treatment. PhenoCube® NeuroCube® and SmartCube® are high-throughput platforms that assess circadian, cognitive, motor behavior, anxiety, gait, and other domains using PGI’s proprietary Computer Vision automated scoring system and machine learning algorithms to define phenotypic signatures. APP/PS1 mice were hyperactive as early as 12 weeks of age as assessed in our high-throughput phenotypic platforms. Tg2576 mice showed a milder yet similar signature. The deficits in the Tg4510 were age-dependent unlike those of the tTa control, which showed a significant phenotype at the earlier ages. Doxycycline partially reversed the Tau behavioral signature. The possibility to evaluate behavioral deficits with the novel high-throughput technology early in life forms a basis to test therapeutic interventions at early stages of developing brain pathology.