Abstracts

Rationale: De novo mutations have been increasingly recognized as causative for epileptic encephalopathy, and mutations in over 50 genes have now been described. However, a genetic diagnosis can still only be made in ~30-50% of patients. Most causative mutations reported to date occur in protein-coding regions; few studies have focused on the role of non-coding regions such as enhancers and promoters in the pathogenesis of epilepsy. We hypothesize that de novo mutations in non-coding regions that regulate the expression of known 'epilepsy genes' can lead to loss of transcription from one allele, and thus haploinsufficiency, in the same way that a coding mutation may cause loss of function of that gene. Methods: We selected 132 putative enhancer and promoter regions around 34 known epileptic encephalopathy genes for targeted resequencing in a cohort of 732 patients with epileptic encephalopathy. The regions selected were predicted from the Roadmap Epigenomics Project data to be promoters and enhancers that are active in neuronal cells or tissue. We performed targeted resequencing of these regions and identified all single nucleotide and indel variants. Moreover, by analysis of read depth, we identified copy number variants (CNVs) across these regions; validation of predicted CNVs by array-CGH is ongoing. We are performing segregation analysis for all variants not present in the general population to identify de novo variants for further analysis. Results: To date we have identified five candidate de novo single nucleotide or indel non-coding variants within the predicted promoter and enhancer regions of SCN1A, TBL1XR1, MEF2C, FOXG1 and CDKL5. In addition, we have identified at least one deletion upstream of SCN1A in a patient with Dravet syndrome in whom no SCN1A mutation was previously detected. This deletion, confirmed by array-CGH, is not present in the unaffected mother or unaffected dizygotic twin. The deletion spans 100kb and deletes a predicted enhancer and promoter, but does not encompass any SCN1A coding sequence. However, RNA-seq data from early neuronal subtypes suggests that the deleted non-coding region may be actively transcribed; the nature of this putative novel transcript is under further investigation. Conclusions: We have identified at least six candidate pathogenic variants in non-coding regions around known epileptic encephalopathy genes. Validation and segregation analysis of single nucleotide, indel and CNVs is ongoing. We will assess the downstream effects of all candidate de novo pathogenic variants on target gene transcription. We hypothesize that mutations within promoter and enhancer regions may lead to a failure of transcription from one allele and haploinsufficiency. Our early investigations of SCN1A suggest that this approach may also be used to identify novel transcripts. In Dravet syndrome, for which 10% of cases lack a molecular diagnosis, it is an attractive hypothesis that the remaining cases may be attributed to mutation in an alternate transcript and that this finding may also apply to epileptic encephalopathy more broadly. Funding: Citizens United for Research in Epilepsy (CURE) Taking Flight Award and NIH (NINDS) Pathway to Independence Award (K99/R00) (1K99NS089858-01) to GLC; NIH (NINDS 5R01NS069605) to HCM and National Health and Medical Research Council of Australia (NHMRC) Program Grant and Practitioner Fellowship to IES.