Circadian Network and Autism: Unusual Alternative Splicing Pattern of the JARID1 Genes

Background: The circadian clock coordinates diverse cellular processes and functional outcomes, including behavior and cognition. Abnormalities in the clock genes may have a role in autism, but the underlying mechanism remains unknown. Members of the JARID1 gene family (JARID1a-d), histone demethylases, have been shown to be involved in the circadian molecular machinery in a recent work of our collaborator.  JARID1a activates CLOCK-BMAL1, whereas JARID1b and JARID1c act as repressors. These opposing roles suggest that the optimal ratio among JARID1 isoforms is vital in maintaining the proper function of the circadian system. Imbalance in these isoform ratios may contribute to the etiology of diseases. Furthermore, miR132 orchestrates translational control of the circadian clock by targeting chromatin remodeling genes, including MECP2 and JARID1a. The splicing profiles of JARID1, and factors that regulate the expression of such isoforms, are not fully known. Our hypothesis is that JARID1 mis-splicing may present in at least a subset of subjects with autism. Recent findings have shown cross links between DNA methylation and gene regulatory processes, including alternative splicing and miRNA. Therefore, we used data from a DNA methylation marker for subject stratifications.

Objectives: To investigate the role of circadian genes and their potential mis-splicing in autism.

Methods: We evaluated the expression level of multiple JARID1 alternative splicing transcripts in autistic subjects stratified based on the absence or presence of a given DNA methylation-related marker (DM) in lymphoblastoid cell line-derived RNAs, using  Exon array profiling, TaqMan assays, followed by DNA sequencing. Expression levels of the identified isoforms were also investigated in the brain samples from control and a Jarid1 mice model at different circadian timing.

Results:  A distinct pattern was detected in the expression level of alternatively spliced JARID1 isoforms for autistic subjects with DM compared to those without DM and controls. Additional experiments, including miRNA mimics, are underway to further characterize the role of miR132 in regulating JARID1a by finding which isoforms(s) show miR132-dependent expression.

Conclusions:  This is the first study to evaluate a clock gene in autism, at the alternative splicing level in conjunction with DM markers.  Our data indicates an unusual splicing process for the X-linked member of this circadian gene family (JARID1c), resulting in unusually high level of intron retained isoforms, as well as the potential role of the long noncoding RNAs and circular RNAs in the regulation of this circadian gene expression.