Pathogenicity prediction for novel genetic variants related to hearing loss in a cohort of patients from Argentina

Abstract

Hearing loss (HL) is the most common disorder affecting 1:500 newborn children. Identification of causative mutations is demanding due to the more than 100 genes involved. Whole-exome sequencing (WES) has become a cost-effective approach for molecular diagnosis. However, the follow-up of novel variants, in particular missense changes, which can lead to a spectrum of phenotypes and unequivocal genotype-to-phenotype correlations, is not always straightforward. In this study, we investigated the genetic cause of sensorineural hearing loss in severe/profound deafness patients. After the exclusion of frequent GJB2-GJB6 mutations by Sanger Sequencing, we performed WES in 32 unrelated Argentinean families. Mutations were detected in 16 known deafness genes in 20 patients: ACTG1, ADGRV1 (GPR98), CDH23, COL4A3, COL4A5, DFNA5 (GSDDE), EYA4, LARS2, LOXHD1, MITF, MYO6, MYO7A, TECTA, TMPRSS3, USH2A and WSF1. Notably, 11 variants affecting 9 different non-GJB2 genes resulted novel: c.12829C>T, p.(Arg4277*) in ADGRV1; c.337del, p.(Asp109*) and c.3352del, p.(Gly1118Alafs*7) in CDH23; c.3500G>A, p.(Gly1167Glu) in COL4A3; c.1183C>T, p.(Pro395Ser) and c.1759C>T, p.(Pro587Ser) in COL4A5; c.580+2T>C in EYA4; c.1481dup, p.(Leu495Profs*31) in LARS2; c.1939T>C, p.(Phe647Leu) in MYO6; c.733C>T, p.(Gln245*) in MYO7A and c.242C>G, p.(Ser81*) in TMPRSS3 genes. To predict the effect of these variants, novel protein modeling and protein stability analysis were employed. These results highlight the value of WES to identify candidate variants, as well as bioinformatic strategies to infer their pathogenicity.