Abstracts

Rationale: Five unrelated BFNS patients (seizure onset between 6-70 days) were negative for point mutations in the major genes, known to be responsible for BFNS, BFIS and BFNIS phenotypes, namely KCNQ2, PRRT2 and SCN2A. Due to the fact that they had very striking seizure similarities such as being triggered by feeding and characterised by apnea, movement arrest and cyanosis, temporal lobe origin, benign outcome with phenobarbital they were subjected to whole exome sequencing with the aim of identification of a common gene explaining the phenotype. Methods: Three patients belonged to consanguineous families and epilepsy was reported in both maternal and paternal ancestors in one patient. Family history was present in all except one.  Whole exomes of 5 patients were sequenced using NimbleGen EZ Whole Exome Enrichment Kit on Illumina HiSeq2000 platform with 100X coverage. Nonsense, nonsynonymous, splice-site variants, and variants rare in ExAc database were selected. Variants that did not have expressions in the brain and those shown to be benign by bioinformatis tools (SIFT, P2-HVAR, P2-HDIV and Mutation Taster) were eliminated.  Genes harboring these selected variants were examined for disease associations in OMIM. WES data was analysed by ANNOVAR and all plausible variants were validated by Sanger Sequencing and segregation analysis were performed if applicable. Results: More than one plausible variant was present in all patients.  Patient 1 had two heterozygous variants in the two ion channel genes SCN4A (e19:c.G3466A:p.A1156T) and HCN4 (e1:c.A76G:p.M26V) inherited paternally and maternally, respectively. SCN4A is dominantly expressed in skeletal musles and the variant was reported in ClinVar as patologic associated with Paramyotonia Congenita. HCN4 has AD inheritance, congenital onset and is associated with Brugada and sick synus syndrome. Patient 2, PCDHGC5 (e1:c.G2101T:p.V701F) variant inherited from the father is important in neuronal survival. SFTPC (e2:c.G115T:p.V39L) is inherited from the mother and is associated with a rare autosomal dominant disease associated with progressive respiratory insufficiency and lung disease. Patient 3 had FGFR3 (e16:c.G1966A:p.E656K) and CACNA1S (e12:c.G1817A:p.S606N) heterozygous variants both inherited from the affected mother. FGFR3 is associated with medial temporal lobe anomalies and plays a role in controlling nervous system development. CACNA1S encodes one of the five subunits of L-type voltage-dependent calcium channel in skeletal muscle cells. Mutations in this gene have been associated with hypokalemic periodic paralysis and thyrotoxic periodic paralysis. In patient 4, both LAMB1 (e23:c.G3124A:p.G1042S) and SCN3A (e3: c.C1619CT p.S540F) variants were homozygous. LAMB1 is associated with AR Lissencephaly. SCN3A is strongly expressed in the brain. Five families with reported mutations had variable onset focal epilepsies. Patient 5 was heterozygous for SCN1A (e10:c.T1436C:p.L479P), SCN5A (e2:c.C52T:p.R18W) and GLRA1 (e8:c.C832T:p.R278C) variants. Although epilepsy history was present in the maternal  side, none of the variants were inherited from the mother. Paternal DNA was not available suggesting the possibility of variants being de novo. SCN1A is the most relevant epilepsy gene associated with DS and GEFS+.  SCN5A is the main sodium channel of the heart, associated with LongQT and Brugada syndrome. GLRA1 is associated with apnea, myoclonus, nocturnal seizures, and episodic generalized skeletal muscle contractions. Conclusions: All patients had at least one variant in ion channel or epilepsy related genes, but none of the genes were common to any two or more patients. SCN1A variant was associated with BFNS phenotype for the first time in patient 23BF155. Patients also had variants in genes associated with neuromuscular phenotypes, neuronal survival and control of nervous system development, progressive respiratory insufficiency, lung disease, apnea, and heart disease that probably explains the common phenotype observed in all 5 patients. The study further reveals the value of WES analysis in delineating all plausable variants and making genonotype/phenotype correlations. Funding: Bogazici University Research Fund and TAM project.