Development of an automated approach for investigating social learning in mice

Mice have been shown to learn from each other through social interactions. However, the extent and strategies of social learning in mice remain largely unknown, beyond spatially and temporally limited tests of social memory retention. Here, we present a method that integrates (1) the IntelliCage, a commercially available tool for automated behavioral testing, and (2) the Live Mouse Tracker (LMT), an open-source solution for 24/7 live animal tracking. This approach allows for the investigation of learning behavior in semi-naturalistic group settings while minimizing experimenter interference. In this study, we report on the development of the method, evaluate its accuracy, and identify current limitations. While automated presentation of spatial learning tasks and behavior annotation were effective, identifying individual animals proved unreliable in a highly enriched environment. In response, we provide a rationale for identifying the reliable portion of tracking data, to which we confine the exemplary behavioral analysis. We acknowledge imperfect animal identification as a clear limitation of the method in its current configuration. However, we outline a path to mitigate this and are confident in presenting a promising tool that, after straightforward optimization, may prove useful for various research questions, including the investigation of social learning behavior in mice. The proof-of-principle experiment in this study did not indicate that place learning is facilitated by co-learning over individual learning, and we could not establish a clear association between social interactions and learning performance. While we observed some sporadic differences in interaction rates between co-learning and individually learning animals, we emphasize that these results do not support any conclusions about the mechanisms of social learning in mice. Rather, we present a tool for simultaneously studying learning and tracking behavior with minimal experimenter interference, which, after refinement, may aid future studies of social learning in mice alongside other research questions.