Autism spectrum disorders (ASDs) comprise a heterogeneous set of neurodevelopmental and neuropsychiatric conditions that affect both social and nonsocial behavior. The laboratory mouse is currently the best mammalian system available for studying ‘genocopies’ of allelic variants found in humans. The use of such mouse genocopies in ASD research has come under criticism because of concerns that the behavioral assays used to phenotype these models are too crude and far removed from human behavior to be informative. These considerations have fueled a push for the use of non-human primate (NHP) models, such as the marmoset, in autism research. But such NHP models are expensive, laborious to generate, and present ethical problems. A complementary approach, therefore, is to develop more sensitive, objective and quantitative automated approaches to measuring ASD-related behavioral phenotypes in mice.

David Anderson proposes to use his recently developed tracking and annotation system¹ to analyze social behaviors during reciprocal dyadic interactions between freely behaving animals in genetic and circuit-level mouse models of autism. The high-throughput, automated nature of this system is uniquely suited to large-scale assessment of behavior across multiple mouse lines, and will allow for the identification of key points of difference in the behavioral phenotypes of different, existing, ASD genocopy lines. Anderson and his colleagues also plan to extend the applications of their system by adapting the hardware to extract more detailed parameters from existing behavioral assays, by integrating the hardware with a system for in vivo imaging of neuronal activity, and by incorporating unsupervised analysis for discovery of novel ASD-related behavioral phenotypes.

In addition to improving the characterization of existing genetic ASD models, Anderson and his colleagues plan to use this updated system to measure behavioral parameters in a novel ASD mouse ‘phenocopy’ system in which the function of neural circuits regulating social behavior are disrupted optogenetically or pharmacologically. The researchers anticipate that the application of this new automated technology will improve the efficiency, accuracy and dimensionality of behavioral phenotyping for the study of mouse genocopies of ASD. The proposed research will have a substantial impact on ASD research, and is expected to lead to important mechanistic insights into the genetic and neural circuitry of ASD.