Abstracts

Rationale: Despite decades of study, the genetic causes of the Juvenile Myoclonic Epilepsies (JMEs) are yet to be determined in the majority of cases because they are highly heterogeneous. We sought the epilepsy-causing gene of a multigenerational family from Honduras ascertained through a proband with JME and adolescent onset pyknoleptic absence. We followed the NHGRI and ACMG guidelines for implicating and assigning disease causality to sequence variants in human disease.Methods: Exhibiting an autosomal dominant mode of inheritance, this 32 member family had 6 non-proband members with epilepsy and 3 asymptomatic members with EEG spike and polyspike wave complexes. Genome-wide genotyping with 5185 SNPs of 32 members was followed by genetic linkage and haplotype analyses. Whole exomes of five affected individuals including the proband, 2 married-in controls, and 1 individual with an “unknown” phenotype were sequenced. Candidate variants within linkage and haplotype regions were confirmed by finding co-segregation of variant with affected members during Sanger sequencing, and by screening for variant in ancestry matched controls. Once confirmed, other probands with IGE diagnoses were screened for additional rare and potentially damaging variants within the same gene.Results: Whole genome haplotype analysis of SNP genotypes revealed a single 44 cM co-segregating region from 13q13.3-q31.2, coinciding with Zmax of 1.90 in 13q21.31 from linkage analysis. Within the co-segregating region, whole exome sequencing identified an exonic variant – p.L660F - in PROSER1, or proline and serine rich 1 gene. Variant p.L660F had a total and Latino allele frequency in public databases of 0.0023, was absent in 181 Honduran and 671 Mexican controls, and co-segregated with all 10 affected family members during Sanger sequencing confirmation. An additional 187 probands of Hispanic (n=134), Caucasian (n=52), and African (n=2) descent from Honduras (n=24), Mexico (n=96), Brazil (n=9), and the USA (n=58) with diagnoses of JME (n=87), CAE (n=39), CAE evolving to JME (n=46), and IGE (n=15) were screened for mutations in the exons, 5’ and 3’ untranslated regions (UTR), and transcription factor binding sites of PROSER1. The original variant, p.L660F was replicated in another proband. Three other rare exonic variants (p.I95L, p.R154C, p.A848T), 1 novel exonic variant (p.S341L), a novel homozygous variant in 5’UTR and a novel variant in 3’UTR were identified in 10 probands. Three of the five exonic variants are predicted to be possibly damaging in 2 or more pathogenicity predicting algorithms. All five exonic variants are predicted to be highly conserved in at least 1 conservation predicting algorithm.Conclusions: Using NHGRI and ACMG guidelines, 5 exonic variants and 2 variants in 5’UTR and 3’UTR of PROSER1 are implicated as possibly disease causing in JME. Further statistical genetic and experimental tests are underway to assess pathogenicity of the 7 variants.