The connectome of developing circuits in mouse models of autism

The underlying causes of ASD remain poorly understood, hindering the development of effective treatments. That’s partly because, until recently, mapping brains at the nanoscale was hard, and most of the action of synapse formation and of the missteps in ASD happen at the nanoscale. Thus, synaptic development is studied with diverse approaches that have provided important insights but also have inherent limitations as proxies for connectivity:

• Optical reconstructions of axonal arbors over development: offer limited sampling (<1 percent of neurons) and rely on inferences about connectivity.
• Trans-synaptic labeling: indirect measures of connectivity, unknown false positive and negative rates.
• Electrophysiological recordings: focus on connected pairs, not overall patterns of synaptic organization.
• Gene expression studies: correlational, not direct measures of connectivity.

Since it is hard to compare across approaches, we still lack definitive answers to fundamental questions about canonical models of circuit formation – do axons initially project broadly with little specificity refine back due to patterns of activity? How do cellular changes, for example, the arborization of dendrites, correlate with synaptic and circuit changes? Are the rules different for inhibitory versus excitatory circuits, long distance versus local? A detailed understanding of the normal rules of synaptic development is critical to pinpoint when those programs go awry in ASD, a prerequisite to more effective therapies and cures.

This project is significant for autism research since it will use large volume multi-scale imaging with nano x-rays (nXCT) and volume EM to provide a unifying framework for an unbiased and complete detailing of how neurons and circuits change in ASD. The researchers focus on the development of long-distance connections from thalamus and local inhibitory circuits since developmental defects in both are implicated in the pathogenesis of ASD. The investigators target primary sensory cortices because both mouse models show sensory processing deficits. More broadly these foundational maps, which work in any species and any brain region, will enable future experiments, looking at other brain regions, comparing mouse models to changes in human brains, and the effects of interventions.