A high throughput splicing assay to investigate the effect of variants of unknown significance on exon inclusion

PurposeInconclusive interpretation of pathogenicity of variants is a common problem in Mendelian disease diagnostics. We hypothesized that some variants of unknown significance (VUS) may lead to aberrant pre-mRNA splicing. To address this we have developed a high throughput splicing assay (HTSA) than can be utilized to test the effects of 1000s of variants on exon recognition. Methods2296 reference, control and variant sequences from 380 exons of 89 genes associated with inherited retinal degenerations (IRDs) were cloned as a pool into a split-GFP HTSA construct and expressed in landing pad RCA7 HEK293T cells. Exon inclusion led to disruption of GFP and exon skipping led to GFP reconstitution, enabling to separate GFP+ve and GFP-ve cells by fluorescence activated cell sorting. After deep sequencing-based quantification of studied sequences in each cell pool, exon inclusion index (EII) was determined, where EII = GFP-ve oligo count/total oligo count. ResultsHTSA showed high reproducibility when compared between different biological replicates (tetrachoric correlation coefficient r2 = 0.83). Reference exon sequences showed a high level of exon recognition (median EII = 0.88) which was significantly reduced by mutations to the essential splice sites (donor site variants: median EII=0.06; acceptor site variants: median EII=0.48). Of the 748 studied VUSs, 47 variants led to decreased exon inclusion ({Delta}EII [≤] -0.3) with 11 variants showing a strong effect ({Delta}EII [≤] -0,6). Using the HTSA data we were able to provide a likely genetic diagnosis to five IRD cases. ConclusionHTSA offers a robust method to study the effects of VUSs on exon recognition allowing to provide new diagnoses for patients with Mendelian disorders.