22119
Characterization of Four ASD Mouse Models Reveals Common Behavioral Phenotypes and Transcriptional Networks in the Striatum
In recent years, multiple mouse models have been produced to study autism. In order to gain high-impact information from these mice models, we propose that the parallel behavioral and molecular phenotyping of several mouse models will help to identify behaviors and molecular mechanisms that are in common, and are therefore more likely to be directly involved in autistic behavior.
Objectives:
In our current study, we looked at locomotor function, anxiety, and risk assessment behaviors in four well-established autism mouse models. In order to discover common molecular pathways that are dysregulated in all mouse models, we did whole transcriptome sequencing (RNA-seq) of the striatum of these mice models. By comparing the behavior of these multiple models and the transcriptome dysregulation in the striatum, a brain area highly involved in these behaviors, we can determine specific molecular mechanisms that are directly responsible for the relevant behavioral dysfunctions.
Methods:
We performed motor-related and risk- assessment related behavioral and molecular experimentation on four mouse models of ASD: Shank3 KO, CNTNAP2 KO, Chr16p11.2del, and BTBR mice. We performed Open Field (OF), rotorod, Dark Light (DL), and Elevated Plus Maze (EPM) on all mouse models. We used this data to quantify locomotor function, anxiety-levels, and risk assessment behaviors. RNA was extracted from the striatum of all four mouse models, and their controls, and performed whole throughput RNA sequencing (RNA-seq) on all samples. RNA-seq data was analyzed by both differential expression and WGCNA (Weighted Gene Correlation Network Analysis) to identify genes that are dysregulated in all models and that correlate with the behavioral phenotypes.
Results:
We found that all four mouse models showed either hyperactive or hypoactive behaviors (Shank3 KO was hypoactive, and the others are hyperactive), and all mouse models showed differential behavior in the anxiety/risk assessment tests. Of interest, we found that all genetic mouse models displayed a decrease in risk assessment behavior. Striatal gene expression analysis found that three of the four models share 31 genes that are commonly upregulated, including IGF2, IGFBP2, and Sema3b. WGCNA analysis, followed by protein-protein interaction networks, found that specific gene expression networks, such as networks including HDAC genes, could be correlated to the dysregulated risk assessment behavior in the autism mouse models.
Conclusions:
Common dysfunctions in risk assessment and locomotor behaviors are found in multiple autism mouse models. In addition, we can correlate these behaviors to distinct gene expression networks. This study provides evidence that we can discover novel molecular pathways involved in the autism-related behavior through the parallel molecular and behavioral phenotyping of multiple autism mouse models.