Autism spectrum disorders (ASDs) are complex, multigene conditions which ultimately converge to cause changes in the electrical activity in the brain. There is substantial evidence that one of the major manifestations of ASDs is an imbalance of synaptic excitation and inhibition, favoring excitation, which in turn causes disruptions in electrical signaling in the brain.

Interpreting these data in light of the wide range of symptoms in ASDs has been difficult for a variety of reasons, three of which stand out as particular conceptual hurdles: 1) The relationship between the measured imbalance in synaptic excitation and inhibition and the actual ongoing electrical activity of the brain is neither simple nor clear, partly because of compensatory mechanisms that come into play during development to mitigate the expected disruptions of activity; 2) Disruptions in ongoing electrical activity have very different effects depending on the stage of development at which they occur. In the mature brain, such changes give rise to acute behavioral symptoms, whereas in the immature brain they may cause permanent effects by disrupting activity-dependent processes of neural circuit development; 3) Imbalances between excitation and inhibition are likely to have very different effects on activity in different regions of the brain.

William Moody and his colleagues propose to use newly developed experimental and computational techniques to assess how cortical activity, arising from both sensory stimulation and from intrinsic mechanisms, is disrupted at critical neonatal stages in mouse models of ASD. Using a cross between two genetic mouse models of ASD (Cntnap2 and Shank1) and mice expressing the calcium indicator GCaMP6s in cortical glutamatergic neurons, they plan to use calcium imaging to measure brain activity over the entire surface of one hemisphere through the intact skull in unanesthetized animals undergoing spontaneous sleep-wake cycles.

Pilot experiments have shown that this method can be used to measure cortical activity for more than one hour at high temporal and spatial resolution. A dimensionality-reduction algorithm called Non-Negative Matrix Factorization (NNMF) was used to extract patterns of activity from within the large, complex data sets that result from long-term recordings over the entire surface of the cortex. Preliminary results suggest that one of the most characteristic forms of spontaneous activity in the neonatal brain that propagate over the entire cortex — pan-cortical waves — are almost completely segregated into sleep cycles. Wake cycles, on the other hand, appear more optimized to mediate more localized forms of activity. This temporal segregation of developmentally relevant activity over different spatial domains into sleep and wake stages may be a mechanism by which both types of activity can carry out their developmental functions without interference from the other.

Pan-cortical waves serve to help establish long-range synaptic circuitry in the cortex; such circuitry is known to be disrupted in ASD. Because of this, and because these waves occur during sleep cycles, which are also disrupted in ASD, Moody proposes that ASD mutations cause changes in pan-cortical waves early in development, thus compromising cortical circuit formation.

By using the above methods in mouse models of ASD, Moody’s group will be able to detect and quantify how these mutations disrupt both sensory-driven and spontaneous activity in the cortex, including pan-cortical waves, in both sleep and wake cycles, at stages when such disruptions are likely to have profound effects on brain development. The results will provide a quantitative picture of how the primary synaptic effects of ASD mutations interact with activity-dependent developmental processes to produce core ASD symptoms.