Abstracts

Rationale: Developmental and epileptic encephalopathies (DEEs) are severe epilepsies characterized by uncontrolled seizures, frequent epileptiform activity and developmental slowing or regression. Sequencing approaches to identify causal genetic variation have been successful in DEE, with 30-50% of individuals now having a molecular genetic diagnosis. Of these genes, SCN1A is the most commonly mutated gene, accounting for 80% of all patients with the DEE subtype, Dravet syndrome, and at least 2% of DEE overall. These genetic studies highlight the prevailing genetic contribution to DEE, and also suggests that some of the undiagnosed cases may be due to genetic variants in the 99% of the genome not assessed in exome sequencing studies. Methods: To detect variants in the regions not normally targeted by exome sequencing we performed targeted resequencing of 11 SCN1A candidate regions representing putative enhancers, promoters and highly conserved regions in over 500 individuals with DEE. Results: We identified five variants in a highly conserved region in intron 20 of SCN1A. This region was shown previously in rat astrocytes to harbor a ‘poison exon’ (20N) that, when incorporated into the SCN1A transcript, leads to the introduction of a premature stop codon and a shortened protein. This truncated transcript is not normally present in neurons where the full length SCN1A protein is required for normal neuronal function. However, in a splice reporter assay we show that 3/5 of these patient-specific variants lead to increased inclusion of the 20N poison exon. Moreover, 4/5 patients have Dravet Syndrome and one a DEE. The variants occurred de novo in two individuals and segregated with genetic epilepsy with febrile seizures plus (GEFS+) in two additional families; segregation analysis is incomplete in the fifth family. Conclusions: We present compelling genetic, cellular and clinical evidence that deep intronic variants in SCN1A cause Dravet syndrome and GEFS+, and potentially a broader spectrum of DEEs. Our work highlights the role of poison exons as a novel pathogenic mechanism in epilepsy. More broadly we show that intronic regions harboring poison exons are prime targets for pathogenic variant discovery using whole genome sequencing studies. This is particularly important for genes that harbor poison exons and are already implicated in epilepsy, including encoding sodium and calcium channels, as well as chromatin remodelers. Understanding the mechanisms that govern poison exon splicing will be crucial to understand disease pathology. Importantly, this mechanism offers a critical opportunity for the design of novel RNA therapeutics. Funding: GLC is supported by NIH, NINDS (R00NS089858) and the American Epilepsy Society and Epilepsy Foundation Junior Investigator/ Kevin’s Fellows Award. HCM is supported by NIH, NINDS, NS069605.