Abstract / Summary
Variants affecting minor (U12-type) intron splice sites are underreported, and their consequences are poorly understood, which makes predicting their pathogenicity challenging. We investigated the impact on splicing of an SCN5A minor intron donor variant in an individual with cardiac arrhythmias using patient-specific induced pluripotent stem cell cardiomyocytes (iPSC-CMs). Genetic testing in the proband identified an SCN5A c.392+3A>G variant of uncertain significance in a minor intron donor splice site, which segregated in affected family members with cardiac phenotypes, including Brugada syndrome. Transcriptome sequencing of proband-specific iPSC-CMs, treated with and without inhibition of nonsense-mediated decay, revealed aberrant SCN5A splicing. Four variant-associated isoforms were absent from control iPSC-CMs and control myectomy samples, and one skipped Exon 3 with a 96 bp extension of Exon 4. All alternatively spliced SCN5A transcripts were T/A cloned and confirmed by Sanger sequencing. These alternatively spliced transcripts used novel combinations of cryptic minor and major splice sites and introduced premature termination codons or an in-frame deletion, supporting reclassification of the variant as likely pathogenic. A genome-wide analysis of SpliceAI scores across minor intron splice sites retrieved from the Minor Intron Database revealed clear patterns of variants with high predicted splicing impact. This analysis also highlighted distinct splice site sequence motifs between minor and major splice sites that are not currently considered in ACMG/AMP variant interpretation guidelines. We further demonstrate that some cryptic donor-acceptor motif pairings are likely incompatible with splicing, which may inform future improvements in in silico splice site prediction tools. Integrating functional RNA sequencing with SpliceAI-based predictions at minor introns offers a pathway towards refining variant interpretation frameworks in clinical genomics.