William Fairbrother

William Fairbrother
Assistant Professor of Biology:
Molecular Biology, Cell Biology, & Biochemistry
Brown University

 

Screening Variants for Defects in Splicing

The Human Gene Mutation Database (HGMD) archives genomic variants that have been reported in the literature to cause hereditary disease(Krawczak et al. 2000).  About 85% of these entries are missense mutations. While most these mutations are thought to disrupt gene function by altering the protein sequence, a significant fraction have been hypothesized to disrupt the pre-mRNA processing by either  creating cryptic sites or disrupting exonic splicing control elements(Lim et al. ; Sterne-Weiler et al.) While biochemical analysis on isolated mutations have revealed numerous examples of missense mutations disrupting splicing, there has yet to be a global survey that describes the full extent of splicing mutations in human disease(Ars et al. 2000). Here, we report an unpublished survey from a dual invivo/in vitro massively parallel reporter assay that records the extent of miss-splicing phenotypes in missense mutations.  We describe over 5,000 wildtype mutant comparisons and correlate their splicing efficiency with a variety of other genomic and transcript features (e.g. conservation, predicted secondary structure, RBP binding motifs). Approximately 20% of missense mutations that cause disease significantly disrupt splicing in vivo and in vitro by a magnitude greater than 1.5 fold (Chi square statistic at FDR 5%). In addition we utilize these assays as a drug screening platform and demonstrate surprising effects of RNA binding pharmaceuticals and translational inhibitors on splicing.  In conclusion, massively parallel reporter assays offer an assay platform that is scaled to match sequencing based genotyping technologies. The results of these experiments can be readily mined for the determinants of splicing. Over 5,000 disease loci were compared to wild type in multiple, highly controlled contexts. We conclude that many mutations outside of canonical splice sites disrupt spicing and small compound screening targeting splicing defects offers a future avenue for therapy.