Autism can be thought of as a disconnection syndrome, in which the normal interactions between areas of the cerebral cortex are disrupted. Information flows between cortical areas along a hierarchy. Of particular interest is the influence of cortical areas that are higher in the hierarchy on early stages of sensory processing.  This hierarchy is referred to as top-down control. Such interactions facilitate our interpretation of our external environment, by segmenting visual scenes into objects and their backgrounds. They play a role in object recognition, and they provide the necessary selection of scene elements that are relevant to perceptual tasks as opposed to those that are irrelevant.

When top-down interactions are not operating properly, we lose the ability to highlight relevant components of the environment and suppress irrelevant ones. Under these circumstances, the cacophony of mutually interfering sensory inputs leads to avoidance of novel input and social withdrawal.

Charles Gilbert and his team at Rockefeller University in New York plan to develop a behavioral mouse model involving engagement of top-down influences, where animals are trained to do visual discrimination tasks.

Once animals are trained, the researchers plan to use a high-resolution two-photon microscope to observe activity of neurons and axons, or neuronal projections, in the brain while animals perform the tasks. This will enable Gilbert and his colleagues to view the brain circuitry responsible for top-down influences and to observe the influences at the single-neuron and axon level.

The researchers plan to compare the physiological mechanisms and behaviors observed in wild mice with those of mice bearing mutations in autism-linked genes. This study will therefore help them to establish the link, at the level of the circuitry in the cerebral cortex, between the genetics of autism and the functional deficits seen in individuals with the disorder.