The core set of behavioral symptoms that define autism spectrum disorders can arise from a wide array of genetic and environmental causes. This many-to-one correspondence between genetic causes and behavioral phenotypes suggests that understanding and eventually treating autism may depend on identifying common intermediate biomarkers  at the level of neural circuits or biochemical pathways.

Functional magnetic resonance imaging studies show that the reliability, or consistency across trials, of sensory-evoked brain responses is reduced in people with autism. Degraded response reliability may be related to the sensory processing impairments and repetitive, stereotyped behaviors commonly seen in autism.

Nathan Urban and his colleagues at Carnegie Mellon University in Pittsburgh aim to determine whether the reliability of sensory-evoked responses is reduced in several genetic mouse models of autism and to understand the mechanisms of the reduction.  Previous experimental and computational work from the Urban lab on the mechanisms and consequences of altered neuronal reliability provides important background for these studies.

Starting with three different autism mouse models, Urban and his colleagues plan to measure the reliability of sensory-evoked responses by recording from primary sensory areas of the mouse brains, using electrophysiological and optical methods. These experiments may allow for a close comparison between human data that show specific changes in reliability of stimulus-evoked responses without an alteration in spontaneous background activity.

Such changes in reliability could be caused by altered properties of single neurons, altered properties of synapses (neuronal junctions) or altered circuit properties. Urban’s group plans to investigate these possibilities in parallel using in vitro brain slice preparations and computational metrics for neuronal reliability. Different mouse models may show the same reliability, even with different underlying biophysical and synaptic mechanisms.