Gene-disrupting genetic variants frequently lead to neurodevelopmental disorders, including autism spectrum disorder (ASD), when they occur in one of the several hundred genes associated with these conditions1. Many of these variants are de novo, observed in the affected child, but not in either parent, and capable of mediating substantial risk. Such variants alter the quantity or quality of the encoded proteins, through deletions, premature stop codons or missense variants.
In this project, Stephan Sanders and colleagues plan to focus on another category of gene-disrupting variants that act by modifying mRNA splicing and prevent a functional protein being produced. The two-nucleotide canonical splice sites either side of exons are critical for normal splicing processes to occur, with the ‘AG’ acceptor motif upstream and the ‘GT’ donor motif downstream. Mutations in these positions are a well-recognized contributor to ASD. However, many other nucleotides play an important role in splicing. Recent analytical advances, including the development and application of the deep neural network algorithm SpliceAI2, have enabled the reliable detection of mutations at these cryptic splice sites.
Sanders and his team plan to systematically identify cryptic splice site variants in exome and genome sequencing data of neurotypical individuals and those with ASD, including from the Simons Simplex Collection (SSC) and the SPARK cohort, to better define the contribution of splicing variation to ASD. High-depth RNA-seq will be used to validate and characterize the splicing disruption in patient-derived cells, starting with lymphoblastoid cell lines (readily available from SSC participants) and progressing to neurons derived from induced pluripotent stem cells, if available. Finally, in a subset of validated splice disrupting variants, antisense oligonucleotides will be designed and tested in vitro to assess if normal splicing behavior can be restored as a first step toward developing an individualized therapy for the affected individuals (e.g., 3).