Next generation sequencing approaches in rare diseases: the study of four different families

The main purpose of this PhD project was to study the molecular bases of rare Mendelian diseases through Next Generation Sequencing (NGS), finding the most appropriate NGS technology and data analysis approach. To this aim, we enrolled at Umberto I General Hospital and Sapienza University of Rome four different families with a phenotype due to a supposed genetic cause, in order to find the causative gene/genes. The selection of the experimental strategy, the number of subjects to sequence (the most distant family members, trio or singleton) and the data analysis approach were dictated by considerations on the diagnostic potential of each sequencing strategy and its feasibility and cost: the diagnostic rate, the possibility to re-evaluate the NGS data periodically, the management of NGS data, the functional interpretation of coding and non coding variants and the number of secondary findings were some of the criteria driving the choice of the NGS test. The choice was also influenced by specific features of each case, e.g. the supposed mode of inheritance, the available samples and the information about the phenotype. Whole exome sequencing (WES) and clinical exome sequencing (CES) experiments were performed in our laboratory or by outer companies. We analysed sequencing data through a dedicated bioinformatic pipeline and we filtered and prioritized the variants according to several parameters, specific for each case. Then, we validated the selected variant/variants through Sanger sequencing on the proband and the other family members, to study their segregation in the family, and we investigated the functional link between the candidate variant/variants and the phenotype. To study the molecular bases of the complex phenotypes regarding canine agenesis and eruption anomalies in the family A, we performed a WES approach on three first degree cousins. Different data analyses, based on different shared genetic causes, allowed us to identify several candidate variants: two missense variants in EDARADD and COL5A1, previously associated with tooth agenesis and a syndromic phenotype including dental anomalies, respectively; three missense variants in RSPO4, T and NELL1, genes functionally related to tooth morphogenesis. The segregation analyses pointed to two different signaling pathways as responsible for the phenotypes, one of them (i.e. EDA) for the canine agenesis, and the other (i.e. WNT) for the less severe canine eruption anomalies. To find the cause of the isolated brachydactyly observed in family B, we used a WES approach on the proband and his grandfather. We identified a shared frameshift variant in the GDF5 gene, encoding for a secreted ligand of TGF-β and predominantly expressed in long bones during embryonic development. This was important for genetic counselling as it is causative of a mild phenotype in heterozygous state, but also of a very severe phenotype in homozygous state. To find the cause of the corpus callosum anomaly observed in the proband of family C, we chose a trio approach. We performed a CES, using an enrichment kit that included 171 of 180 genes reported in literature as causative of corpus callosum malformations. We identified in the proband a supposedly de novo nonsense variant in the ARX gene, critical for early development and formation of a normal brain. Segregation analysis disclosed the presence of the same variant also in the fetus of a previous pregnancy, suggesting a gonadal or gonosomal mosaicism in one of the parents. The identification of this variant was important for genetic counselling as there is an increased recurrence risk for the couple to have a child with the same disorder. It was also important for the proband’s clinical prognosis and to properly calculate the risk to transmit the mutation, which is associated with different clinical outcomes depending on the sex. To investigate the molecular bases of the recurrent Dandy- Walker malformation observed in the family D, we performed WES only of the proband. We identified a homozygous missense variant in FKTN gene, encoding a glycosyltransferase with a role in brain development. In order to test the pathogenicity of the variant, we also performed a structural modeling of FKTN. This result allowed to properly redefine the clinical diagnosis as a Muscular Dystrophy-Dystroglycanopathy Type A, with implications on recurrence risk for the couple and on reproductive choices. The adopted experimental and data analysis strategies allowed us to identify the molecular causes of phenotypes involving different systems and belonging to different clinical pictures, with significant impact on diagnosis, prognosis and genetic counselling. These results show how NGS is revolutionizing medical genetics, accelerating the research about rare-genetic diseases, facilitating clinical diagnosis and leading us to the personalized medicine.