The pathogenic and symptomatic heterogeneity of autism spectrum disorders (ASD) are two important factors limiting the development of new therapeutic strategies. Unravelling the molecular and cellular mechanisms underlying the different pathological behaviours represents a way to better understand the complexity of the disorders and develop new treatments. Recent studies on mouse models for ASD based on highly penetrant genetic signals in patients have begun to reveal aspects of ASD neuropathology, but our knowledge on causal links between neuronal networks and behaviours remains limited. We propose to address this question in several mouse models for ASD and identify general neuronal mechanisms underlying ASD-related behaviours. In this programme of work, we will define:The developmental trajectory of pathological behaviours.Windows of intervention to genetically rescue behavioural phenotypes.Causal relationships between neuronal circuits and pathological behaviours.As a short-term goal, in the scope of this application, we will use a well-characterised conditional mouse model where the adhesion synaptic protein Neuroligin 3 is lacking. As a long-term goal, we plan to use this knowledge to extend our approach to selected conditional mouse models carrying microdeletions frequently found in patients.