From High-Throughput Analysis of Genetic Variants to the Experimental Validation of Putative Protein Function

The state-of-the-art approach for the genetic molecular cause research relies on massively parallel gene sequencing, which represents a challenge both in data handling and variant prioritization. The univocal assignment of disease pathogenicity to the sequence variants is often difficult, and requires the integration of different lines of evidence for a comprehensive interpretation. During my thesis, I contributed to the development of novel approaches to evaluate rare variant contribution to the clinical phenotype. These methods were presented and evaluated at the Critical Assessment of Genome Interpretation, ranking among top programs considering either performance or the number of correct assigned disease predictions. A similar strategy was employed for identification of disease genes linked to neurodevelopmental disorders (NDDs) comorbidity. In this case, computational methods were applied to select the most promising candidate genes for the design of diagnostic panel, which is currently used for patient screening at Pediatrics Clinic of the University of Padova. The variants found within the panel genes have been selected according frequency, pathogenicity prediction and variant segregation analysis within the family. Furthermore, I took advantage of different computational tools to investigate the mutated gene function, and used this information to demonstrate the impact of likely pathogenic variant on clinical phenotype. In several cases, likely pathogenic mutations mapped to intrinsically disordered regions (IDRs), which lack a fixed three-dimensional structure. Coherently, several studies demonstrate that mutations in IDRs are often associated with the pathogenesis of various human diseases. Thus, IDRs classification could represent a critical step for understanding the impact of possibly disease-causative variants mapping in these regions. Due to the influence of intrinsically disordered proteins (IDPs) in diseases, I participated to the manual curation and update of entries in the DisProt database, the primary repository of disorder-related data on sequence. Interestingly, increasing evidence from literature highlights the IDPs involvement in neuronal signal transduction. Among the proteins encoded by diagnostic panel genes, TANC2 especially emerged as intrinsically disordered protein with a possible role in synaptic signal transduction. As TANC2 and its protein family function was poorly characterized, I performed an in silico analysis to characterize the TANC protein activity, and the implicated biological processes. The functional hypothesis emerged from the bioinformatics analysis was used to drive further experimental investigations. In vitro validation of predicted TANC2-CDKL5 interaction highlighted the relevance of the IDRs in regulating degradation of CDKL5, whose mutations are associated with a heterogeneous set of NDD phenotypes. Furthermore, I demonstrated that TANC2 contributes to downregulate CDKL5 expression levels. For this reason, TANC2 protein could represent a novel therapeutic target to design new drugs for the treatment of CDKL5 over-expression associated diseases