An integrated, next-generation approach to identify new genes and new pathways in hereditary ataxias

The Hereditary ataxias (HAs) are a group of heterogenous neurological disorders associated with multiple genetic etiologies and encompassing a wide spectrum of phenotypes, where ataxia is the prominent feature. HAs are characterized by degeneration of Purkinje cell and/or spinocerebellar connections, often associated with defects in additional brain structures, and all patterns of inheritance may occur. Similar to other fields of medical genetics, Next Generation Sequencing (NGS) has entered the HA scenario widening our genetic and clinical knowledge of this condition, but routine NGS applications still miss genetic diagnosis in about two third of patients. In this doctoral study, we applied multi-gene panels to define the molecular basis in 259 patients with a clinical diagnosis of HA and negative to tests for pathological expansion in SCA1, 2, 3, 6, 7, 8, 12, 17 and FXN. We found a positive molecular diagnosis in 25% of patients, whereas a similar number of patients had an uncertain diagnosis due to the presence of either variants of uncertain significance or lack of biological samples to determine segregation among family members. Hence despite a higher positive diagnostic rate compared to similar studies described in literature, a half of patients lacked any indication of the genetic cause of their disease. Using exome sequencing as a second-tier approach in some families, refractory to multi-gene panel analysis, did not significantly improved our diagnostic yield. On the other hand, NGS analysis in our cohort indicated that familial cases were more easily diagnosed rather than sporadic cases, and also that combining massive sequencing with detailed clinical information and family studies increases the likelihood to reach a molecular diagnosis. Among positive patients, we could expand clinical and allelic information in a subgroup of genes offering original description of new mutations and corroborating genetic findings with functional investigations that took advantage of different in vitro or in vivo platforms. In particular, through functional studies in SPG7 knock-down models of Drosophila melanogaster, we remarked that SPG7, whose mutations cause spastic paraplegia type 7, has a critical role in neurons more than in skeletal muscle. The high frequency of p.Ala510Val mutation in SPG7 observed in our cohort as well in similar studies performed elsewhere moved us to develop a humanized knock-in fruit fly model harboring that specific mutation and prepare preliminary characterizations. Similar studies in fruit fly were performed silencing AFG3L2, the gene causing SPAX5 in a child in association with an unusual, relatively milder phenotype. Furthermore, combination of skin fibroblasts and Saccharomyces cerevisiae as models was employed in the genetic characterization of new mutations in a novel recessive HARS-related phenotype whereas primary human cells, yeast and Danio rerio models were used to functionally characterize new HA-related mutations in COQ4. Finally, we could expand the clinical presentation of rare causes of HAs describing new dominant mutations in STUB1 and biallelic variants in RFN216, COQ8A, and ATP13A2. Altogether, studies performed during this doctoral work further underlined the usefulness of NGS in HAs and highlighted how NGS technologies rely on the integrated use of family and clinical studies and different in vitro/in vivo platforms to substantiate molecular findings. The latter platform will be also a tool for future investigations to dissect pathogenesis and to improve therapies.