A Homozygous Deep Intronic SNX14 Variant Activates Pseudo-Exon Inclusion in a Patient with SCAR20

Background: The contribution of intronic variants to the etiology of Mendelian diseases is still underrecognized, impacting the diagnostic yield. Whole genome sequencing (WGS) detects intronic variants, but besides canonical splice-sites, intronic variants are frequently excluded from the interpretation step or are classified as variants of uncertain significance (VUS). In fact, assessing their clinical significance often requires validation via RNA-sequencing (RNA-seq) or in vitro studies. Methods: We studied a 31-year-old patient with spinocerebellar ataxia who lacked a molecular diagnosis after WGS analysis. We applied the Detection of RNA Outliers Pipeline (DROP) to analyze RNA-seq data from patient fibroblasts. DROP integrates OUTRIDER and FRASER 2.0 algorithms designed to identify aberrant gene expression and splicing, respectively. Results: DROP identified differential expression and aberrant splicing of SNX14. Retrospective WGS data analysis revealed a homozygous NM_153816.6(SNX14): c.867+288A>G deep intronic variant, which caused pseudo-exon activation and reduced transcript levels. Biallelic loss-of-function variants in SNX14 cause autosomal recessive spinocerebellar ataxia type 20 (SCAR20; OMIM 616354), consistent with the clinical presentation of this case. Conclusions: We identify a deep intronic SNX14 variant as the genetic basis of SCAR20. We demonstrate the utility of RNA-seq to increase the diagnostic yield by identifying and resolving the pathogenicity of deep intronic variants. Defining aberrant splicing events is therapeutically relevant, as these mechanisms are targets for antisense oligonucleotide (ASO) based interventions.