Abstracts

Rationale: Elucidation of genetic generalised epilepsies and other common epilepsy syndromes, where the inheritance is often complex due to the contribution of multiple genes, is challenging. To unravel the complex genetics we are using discordant monozygous twins to exploit new sequencing technologies and identify somatic mutations that explain the discordance. In a previous study we identified a somatic mutation causing Dravet syndrome in one pair of discordant monozygous twins, demonstrating proof-of-principle. Here we report interrogation of eight discordant monozygous twin pairs where only one twin has a genetic generalised epilepsy subtype by whole exome sequencing of lymphocyte DNA.Methods: We sequenced twin pairs at a high depth of coverage (minimum seventy-fold) to detect somatic mutations that may occur at low abundance during the first few cleavages but not at later stages of development. To minimise experimental variation we sequenced all twin samples in a single pool on an Illumina sequencer and searched for variants present only in the affected twin of each pair. For the analysis we focussed on variants in coding, splice site or untranslated regions covered by at least ten sequence reads and with minor allele frequencies less than one percent according to the Exome Variant Server database. Results: Using these criteria we identified thirty-one potential discordant variants (approximately four per pair) that we then validated by conventional Sanger sequencing. Of these variants, three were real but present in both twins, and the remaining twenty-eight were false positive, most likely due to sequencing errors. Conclusions: This study highlights the challenges of identifying discordant mutations in identical twins on a genome wide basis: (1) they may be present at low abundance and (2) they may show variable expression due to tissue mosaicism. To address these issues we will increase sequence coverage to over 100-fold and source DNA from neural tissue or cells. It may turn out that somatic mutation is a relatively rare cause of genetic generalised epilepsies but this will only be determined by interrogating a larger cohort of twins. That we have previously shown early SCN1A mutation leads to discordance in identical twins for Dravet syndrome is reason to believe it may be a more general phenomenon. Although we have established a framework for tracking down such mutations, the limitations of current sequencing technology mean we will have to investigate many candidate mutations to find truly discordant ones.