Exploring the non-coding regions of the genome: the contribution of cryptic splicing variants to the onset of Joubert syndrome

Joubert syndrome (JS) is a mendelian disease that falls into the category of cerebellar and brainstem congenital defects. It represents a complex syndrome with predominantly autosomal recessive inheritance caused by causative biallelic variants affecting about 40 known genes, involved in the biogenesis and correct functioning of the primary cilium. JS is characterized by a congenital malformation of the brain stem and agenesis or hypoplasia of the cerebellar vermis. The distinctive neuroradiological hallmark of JS is the molar tooth sign (MTS), detectable on encephalic MRI. The main clinical signs are hypotonia, respiratory abnormalities, ataxia, and delayed motor and intellectual development. Although the progresses accomplished in the field of neuroimaging and Next-Generation Sequencing (NGS) molecular diagnostics are continuous, to date, about 30-40% of patients clinically and neuroradiologically diagnosed with JS remain without a genetic diagnosis, thus living the so called “diagnostic odyssey” and leading to increased socioeconomic burden. A subset of these missing diagnoses is likely due to cryptic intronic or structural variants, that, with current criteria, are classified as uncertain clinical significance. This thesis aims to explore the hypothesis that a subset of mutations is cryptic and escape conventional diagnosis. To address this, 32 JS patients from Professor Valente’s cohort (Department of Molecular Medicine, University of Pavia) who were heterozygous for a deleterious or possibly deleterious variant in a JS major gene, whose reported phenotypes were consistent with the patients’, and who did not carry other obvious variants in other genes were selected. Whole Exome Sequencing (WES) was performed on all patients whose genomic DNA was available and the intronic regions included in the sequencing were not filtered out during the bioinformatic analysis. After this process, 7 different cryptic splicing variants were identified and their in silico predicted effects were validated through direct RNA studies, if available, or minigene assays if patient lymphoblastoid cell lines or blood RNA samples were not available. Therefore, it was possible to identify and validate in vitro the second cryptic causative splicing variant in 9 patients, thus ending their diagnostic odyssey. Eventually, the selected probands, who remained without a genetic diagnosis after this screening, represent a very promising cohort for further investigations aimed to unravel other classes of cryptic variants, using innovative technologies capable of sequencing the deepest regions of human genome.