Analysis of Phenotypes in Rodent Models Based on High-Risk ASD Genetic and Environmental Factors

Background:  With identification of risk genes associated with autism spectrum disorder (ASD), rodent models provide access to the unique and common phenotypic outcomes for their perturbations. Increasing evidence also presents strong corroboration towards the involvement of environmental factors, such as maternal immune activation (MIA) during fetal development, as a risk factor for development of ASD. Again, rodent models have been successfully used to provide insights into the regulators of perturbation to pivotal pathways in the developing fetus in MIA, e.g., the role of proinflammatory cytokines, like IL-6 and IL-17a. However, given the heterogeneity of ASD features in the human population, as well as the relatively tenuous parallels between face validity in animal models and human behavior in the manifestation of the complex, multifactorial, neurodevelopmental disorders like ASD—i.e., in social behavior, communication and repetitive behavior—we look into the phenotypic characteristics of animal models based on strong genetic or environmental risk factors of ASD.

Objectives:  Our aim is to assess the face validity of the entire set of models developed for genetic and environmental factors associated with ASD based on multiple lines of evidence. We examine the underlying causes in this set, something that has not been dealt in a detailed manner to the best of our knowledge.

Methods:  Using the data in the Animal Module of AutDB (also known as SFARI Gene), we look into the detailed phenotypic data on the models based on high-confidence and strong-candidate genes to uncover similarities in complex behaviors as well as in possible underlying neurophysiological or neuroanatomical substrates. Additionally, we also correlate the phenotypic outcomes seen in high-risk genes to those of maternal immune activation in various rodent models using several immune activators: influenza virus infection, viral infection mimicry using polyinosinic–polycytidylic acid (poly(I:C)), bacterial infection, its mimicry using lipopolysaccharide (LPS) and modulation of cytokine levels. Our analysis looks into changes in phenotypes beginning during embryonic development to identify any common anomalies seen in brain development, followed by distinctive patterns in neuroanatomy and neurophysiology, by age: early postnatal, juvenile stages, through adulthood.

Results:  Loss of function of high-confidence genes in mice causes increased embryonic lethality, which is not surprising as it is likely that these genes perform important developmental roles that have not been highlighted previously. Data from heterozygous and intricate conditional knockout mice of high-risk genes confirm that mouse models do recapitulate ASD features, reliably preserving face validity. From the data on genetic models combined with MIA models, there is some indication of common underlying changes in brain structure and function that are related to the observed impairments in social and repetitive behavior, even if deficiencies in communication are not that saliently reflected in these models.

Conclusions:  Our analysis provides support for looking at large datasets of rodent models showing face validity of some features of ASD in order to elucidate common underlying mechanisms. We believe that, while striking observations are noted by studies conducted on one or two important ASD genes, a combined approach can bring out similarities not otherwise highlighted.