A hidden structural variation in a known IRD gene: a cautionary tale of two new disease candidate genes.

Rod-cone dystrophy (RCD), also known as retinitis pigmentosa, is an inherited condition leading to vision loss, affecting 1 in 3500 people. More than 270 genes are known to be implicated in the inherited retinal degenerations (IRDs), yet genetic diagnosis for ∼30% of IRD of patients remains elusive despite advances in sequencing technologies. The goal of this study was to determine the genetic causality in a family with RCD. Family members were given a full ophthalmic exam at the Retinal Service at Massachusetts Eye and Ear and consented to genetic testing. Whole-exome sequencing (WES) was performed and variants of interest were Sanger-validated. Functional assays were conducted in zebrafish along with splicing assays in relevant cell lines to determine the impact on retinal function. WES identified variants in two potential candidate genes that segregated with disease: GNL3 (G Protein Nucleolar 3) c.1187 + 3A > C and c.1568-8C > A; and PDE4DIP (Phosphodiester 4D Interacting Protein) c.3868G > A (p.Glu1290Lys) and c.4603G > A (p.Ala1535Thr). Both genes were promising candidates based on their retinal involvement (development and interactions with IRD-associated proteins); however, the functional assays did not validate either gene. Subsequent WES reanalysis with an updated bioinformatics pipeline and widened search parameters led to the detection of a 94-bp duplication in PRPF31 (pre-mRNA Processing Factor 31) c.73_266dup (p.Asp56GlyfsTer33) as the causal variant. Our study demonstrates the importance of thorough functional characterization of new disease candidate genes and the value of reanalyzing next-generation sequencing sequence data, which in our case led to identification of a hidden pathogenic variant in a known IRD gene.