Homecage-Like Monitoring Reveals Domain-Specific Rescue of Central SPTAN1 R1098Q Phenotypes by Wild-Type Partner Co-Housing

Abstract Neurodegenerative disorders manifest with markedly different phenotypes, with motor- and memory impairments dominating both clinical and preclinical phenotyping efforts. However, neurodegeneration in unlikely to be confined to these two functional domains alone, and its broader behavioral consequences, including social alterations remain underexplored. Here, we present a non-invasive, high-content home-cage phenotyping framework to investigate social and exploratory dysfunction in SPTAN1 R1098Q mice, a model of early infantile epileptic encephalopathy (EIEE5). Our pipeline integrates three open-source AI engines: SLEAP.ai for markerless multi-animal pose estimation, DeepOF for automated classification of autonomous and social behaviors, and SAM2 for segmentation and quantification of nest-building material usage. Long-term monitoring revealed that SPTAN1 R1098Q mutants exhibit impaired spatial habituation in mutant-mutant (MUT - MUT) pairs. Mutants also displayed persistent, unreciprocated trailing (following) of cage mates and a pronounced reduction of nest-building behavior alongside. Notably, co-housing with a wild-type littermate (MUT - WT) attenuated mutant hyperactivity and restored more typical rest pattern, suggesting that social context can partially buffer these behavioral abnormalities. By revealing social and exploratory alterations that expand conventional phenotyping, this computational ethology pipeline advances 3R principles while revealing that intrinsic executive state-switching failures in neurodegenerative models can be functionally buffered by social dynamics.