Abstracts

Rationale: Dravet syndrome (DS) is a severe, infantile-onset developmental epileptic encephalopathy characterized by a distinct seizure pattern and neurocognitive deficits. Approximately 80% of cases of DS are associated with mutations in SCN1A, with 90% arising de novo. De novo mutations in STXBP1, PCDH19 or GABRA1 account for a small fraction of SCN1A-negative patients, however in most of the remaining 20% of patients a causative genetic mutation cannot be identified. We hypothesized that some of these cases could be caused by post-zygotic mosaic mutations in SCN1A or other DS-associated genes, where the mutation has not yet been identified because it is present at a low level in the tissue being tested. De novo mutations that arise post-zygotically in the developing fetus are only present in a subset of an individual's cells, and the fraction of cells harboring the mutation typically varies among different tissues. A disease-causing somatic mosaic mutation that is present in a large proportion of cells in the brain, and that results in epilepsy, may be present in only a small number of cells in the blood. Standard sequencing methods cannot routinely detect low-level mosaic mutations, so a more sensitive sequencing approach is required. Methods: We used highly sensitive small molecule molecular inversion probes (smMIPs) for targeted capture and sequencing of the coding region of SCN1A and seven other DS-associated genes (STXBP1, GABRA1, SCN2A, SCN8A, GABRG2, HCN1, and PCDH19) in a cohort of SCN1A mutation negative patients with DS. Where possible, we screened both blood and buccal samples from each patient to increase our ability to detect low level mosaicism. Protein changing mutations not present in the Exome Aggregation Consortium (ExAC) or inherited from a parent are considered de novo somatic mosaic mutations and potentially disease causing. Results: We screened 20 SCN1A mutation negative DS patients for low level mosaicism in SCN1A as well as 7 other DS-associated genes. Overall, 82% of the target was uniquely captured at least 50 times and sequenced at >500X coverage. We identified one patient with a low level mosaic p.Phe1798Leu SCN2A missense mutation. This variant was detected at approximately 2.9% in the blood of the patient ?" a level that would have been missed by traditional sequencing methods. The variant is not present in either parent, consistent with post-zygotic origin in the affected individual. The variant is also not in ExAC, and therefore, this mutation is likely to be the underlying cause of the patient's epileptic encephalopathy. Conclusions: We found that somatic mosaic mutations can cause DS. While these patients carry a clinical diagnosis of DS, establishing the genetic cause is important for understanding the underlying mechanism of disease, for counseling patients and families about prognosis and recurrence risk, and for optimizing medical management and outcome. Funding: DSF Postdoctoral Fellowship