Evaluation of Presumably Disease Causing SCN1A Variants in a Cohort of Common Epilepsy Syndromes

Abstract

ObjectiveThe SCN1A gene, coding for the voltage-gated Na+ channel alpha subunit NaV1.1, is the clinically most relevant epilepsy gene. With the advent of high-throughput next-generation sequencing, clinical laboratories are generating an ever-increasing catalogue of SCN1A variants. Variants are more likely to be classified as pathogenic if they have already been identified previously in a patient with epilepsy. Here, we critically re-evaluate the pathogenicity of this class of variants in a cohort of patients with common epilepsy syndromes and subsequently ask whether a significant fraction of benign variants have been misclassified as pathogenic.MethodsWe screened a discovery cohort of 448 patients with a broad range of common genetic epilepsies and 734 controls for previously reported SCN1A mutations that were assumed to be disease causing. We re-evaluated the evidence for pathogenicity of the identified variants using in silico predictions, segregation, original reports, available functional data and assessment of allele frequencies in healthy individuals as well as in a follow up cohort of 777 patients.Results and InterpretationWe identified 8 known missense mutations, previously reported as pathogenic, in a total of 17 unrelated epilepsy patients (17/448; 3.80%). Our re-evaluation indicates that 7 out of these 8 variants (p.R27T; p.R28C; p.R542Q; p.R604H; p.T1250M; p.E1308D; p.R1928G; NP_001159435.1) are not pathogenic. Only the p.T1174S mutation may be considered as a genetic risk factor for epilepsy of small effect size based on the enrichment in patients (P = 6.60 x 10−4; OR = 0.32, fishers exact test), previous functional studies but incomplete penetrance. Thus, incorporation of previous studies in genetic counseling of SCN1A sequencing results is challenging and may produce incorrect conclusions.

Mutant Plasticity Related Gene 1 (PRG1) acts as a potential modifier in SCN1A related epilepsy

“…It was found only once in the heterozygous state amongst 245,604 alleles from the gnomAD database, a large gene mutation databases of non-epileptic individuals. With respect to pathogenicity of a SCN1A variant we are aware that a number of SCN1A variants in the HGMD database would not be classifiable as "clearly pathogenic", and that, as recently pointed out, a significant fraction of patients identified with SCN1A mutations may actually not carry any SCN1A variant relevant for epileptogenesis (Lal et al, 2016). We agree with these authors that the role of SCN1A missense variants in the pathogenesis of common epilepsies should not be overstated.…”

“…Based on the functional data, we assume that a combined haploinsuffiency of PRG1 and SCN1A might be potent enough to evoke a transient severe seizure phenotype in our patient. The p.N541S variant is located in the large cytoplasmic loop between the first and second transmembrane segments of the SCN1A channel, a region known to be sometimes tolerant to substitutions, even if the substitution affects an evolutionarily conserved residue (Lal et al, 2016). Hence this variant might only have a modest effect on channel function, illustrated by the fact the single SCN1A heterozygous individuals (Fig.…”

Expanding spectrum of SCN1A-related phenotype with novel mutations

“…In a recent study, the authors reported that p.Arg542Gln mutation is not a pathogenic variant, based on the original reports, the high frequency in control individuals and missing or negative segregation and functional result. 25 But the same mutation was reported in another study in a 5 year-old girl who was operated due to cleft palate and also suffered absence and atonic seizures. In this study, the authors suggested that the mutation disrupts the predicted tyrosine kinase site and destroys channel function.…”

Alterations in the α2δ ligand, thrombospondin‐1, in a rat model of spontaneous absence epilepsy and in patients with idiopathic/genetic generalized epilepsies

“…24,25 Only samples with a minimum of 90% of all bases in the coding region of the THBS1 and CACNA2D1 being covered by at least 11 reads were used and the Human Gene Mutation Database (HGMD) variants identified in the patients were validated by Sanger sequencing. Their clinical information, including data on EEG and antiepileptic therapy, were recorded on data collection forms.…”

“…Genomic DNA isolation and genetic analysis was carried out with the Nimblegen-SeqCapEZ-V244M enrichment kit on the Illumina HiSeq2000 system as described. 24,25 Only samples with a minimum of 90% of all bases in the coding region of the THBS1 and CACNA2D1 being covered by at least 11 reads were used and the Human Gene Mutation Database (HGMD) variants identified in the patients were validated by Sanger sequencing. Variant annotation was performed using the HGMD Professional 2013.…”