Neural development involves a sequence of temporally overlapping processes that ultimately build a functional adult circuit. We posit that most ASD gene mutations cause both an initial, direct perturbation in early cell or circuit development, and also trigger a cascade of secondary changes building up through later development and into adulthood. Many of these secondary effects may be compensatory or homeostatic, and thus partially ameliorate the initial insult. We hypothesize that late-emerging compensatory developmental processes are a common contributor to circuit pathophysiology in ASD. Supporting this idea, multiple ASD mouse models show synaptic hallmarks of homeostasis, and/or physiological and behavioral phenotypes that appear transiently during development but then lessen or normalize in adults. But other models show signs of impaired or saturated homeostasis, including loss of synaptic scaling. We propose a ‘dual-hit’ model in which highly penetrant ASD gene mutations cause both a primary disruption of early circuit development, and also impair or saturate homeostatic processes to prevent effective compensation.

To test these hypotheses, we study the development of corticostriatal circuits from somatosensory (S1) and medial prefrontal cortex (mPFC) into the dorsal striatum. Corticostriatal circuit dysfunction is likely to underlie restricted, repetitive and inflexible behaviors in ASD. S1, mPFC and striatal circuits are dysfunctional in a wide variety of ASD models, but the developmental origins of these changes are not known. We focus on the Scn2a+/- and Syngap1+/- models of ASD, because these represent two distinct, well-defined first-order biological insults, which will likely engage secondary compensatory mechanisms. For each model, we will track circuit phenotypes through development to identify early-emerging (putative primary) and late-emerging (putative secondary) effects. We will experimentally determine which secondary effects are compensatory by testing whether they emerge after transiently manipulating early circuit activity, and whether they spread across cell types and brain areas downstream of the direct genetic insult. Finally, we will directly test whether these ASD models have impaired cellular and circuit homeostasis, as predicted by the dual-hit model. Comparing Scn2a and Syngap1, across three linked brain areas, will provide a powerful test for the role of developmental compensation in ASD circuit pathogenesis, and may reveal common principles that would suggest effective interventions.

This project brings together the labs of Helen Bateup, Dan Feldman and Linda Wilbrecht labs at the University of California, Berkeley. Each PI has a long record of studying ASD mouse models, with expertise in synaptic development of corticostriatal and striatal circuits (Bateup), synapse, circuit and coding dysfunction in S1 (Feldman), and synapse, circuit and behavioral dysfunction in PFC and striatum in developing rodents (Wilbrecht). The scale of the proposed project, across mouse models, brain areas and levels of analysis, is only possible with a multi-laboratory collaboration.