Author:

Abstract:

Early-life social behaviors impact an organisms further physical, socio-emotional and cognitive development profoundly. In all mammalian species the mother is a central interaction partner. Disruptions in early-life mother-infant interactions have been associated with an increased risk for physical, socio-emotional and cognitive development. Therefore, preclinical research is necessary to better understand underlying neural mechanisms involved in healthy and abnormal socio-communicative development. However, establishing valid and reproducible research findings requires precise and accurate assessment tools. The Pup Retrieval test (PRT) is the leading behavioral assay to assess the maternal behavioral response to infant isolation distress in laboratory rodents. The PRT quantifies maternal behavior of a dam by the time it needs to approach, find and carry back an isolated pup. Despite the effectiveness and usefulness of the PRT, sampling procedures are currently lab and even researcher dependent, and test output is rather limited. In this project, we developed novel phenotypes for assessment of early-life bidirectional development based on the PRT. First, we focused on extending the level of behavioral analysis and improving test accuracy, reliability and reproducibility. Using open-access deep neural networks and machine learning technology we developed an automated procedure which is able to accurately and reliably estimate the classic PRT parameters automatically. Additionally, we extended this to quantify ethologically relevant components of maternal retrieval behavior such as maternal approach and carrying. Second, the automated PRT was used as basis for a novel behavioral test assay: BAMBI. BAMBI is the first behavioral test assay to assess early-life bidirectionality in mother-pup dyads in mice. In contrast to existing behavioral paradigms, BAMBI simultaneously records pup isolation vocalizations and maternal behaviors to investigate the action-reaction dynamic between them. This is particularly important in preclinical research of rodent models of disorders with early-life communication deficits such as autism spectrum disorders since infant and/or maternal factors might underlie and ameliorate infant social deficits. In our experiments, for example, we showed a significant association between number of ultrasonic vocalizations of an infant during PRT and retrieval success. Finally, we applied this early-life methodology to the prenatal valproate exposure mouse model, an established model for autism spectrum disorder (ASD). Here, we administered THIQ, a potent and selective melanocortin-4 receptor agonist, in both male and female neonates to profile its impact on early-life and adult ASD-like behaviors. We were able to show differences between treatment groups in early-life social interactions with their mother as well as long lasting behavioral alterations.