Research Project
10234717
Project Summary/Abstract Over one in one hundred infants born are affected by neurodevelopmental disorders (NDDs), which can cause a host of debilitating symptoms including early-onset seizures, developmental delay, and intellectual disability. Application of whole exome/genome sequencing to examine NDD has revealed many disease-causing variations, but in a substantial proportion of cases the genetic cause remains undetermined. The analysis of deep-intronic non-coding variants has provided diagnosis when coding variation cannot be implicated in disease. For example, intronic variants in the ?20N? poison exon of SCN1A results in loss of function, causing Dravet Syndrome, an NDD causing prolonged seizures in the first year of life. Poison exons are a mechanism of gene regulation whereby a premature termination codon caused by inclusion of the poison exon induces nonsense- mediated decay of the mRNA transcript. Preliminary analysis has shown that poison exon inclusion (PEI) in SCN1A and SCN8A decreases during embryonic brain development, inversely proportional to total RNA expression. The brain is a hotbed of alternative splicing during early neuronal development therefore it is plausible, if not probable that novel poison exons remain uncharacterized. The overall hypothesis of this proposal is that PEI is a mechanism to differentiallyregulate gene expression during earlyneuronal development, and intronic variants affecting PEI in genes of developmental importance contribute to disease phenotype. In Aim 1, the sodium and calcium voltage-gated channel alpha subunit families will be assessed to detect and characterize novel poison exons. Known and novel PEI will then be determined in the developing brain using publicly available RNA-seq experiments. In Aim 2 this analysis will be scaled up over 100-fold to detect and assess PEI in over 3000 genes associated with NDD and Autism Spectrum Disorder (ASD) in the developing brain and the driving forces of PEI evolution will be assessed. Comparative analysis of poison exons will be conducted using k-means hierarchical clustering and unsupervised random forest machine learning methodologies to determine prime biological factors driving PEI utilization. In Aim 3, intronic variants will be analyzed in probands of NDD and ASD in regions of known and hypothetical PEI and the affect these variants have on disease phenotype will be computationally assessed. These experiments will reveal new insights into alternative splicing in the developing brain and provide potential genetic explanation of currently unknown NDD disease phenotype. The research will be conducted at the HudsonAlpha Institute for Biotechnology, which allows developing scientists to hone skills in a dynamic, hands-on environment with supplemental training courses as necessary. Besides the research outlined in the proposal, training will include courses in conduct of responsible research, advanced biostatistics, machine learning, bioinformatics, and wet lab techniques next to personal development to improve oral and written communication and presentation skills as well as time management.
