LEVERAGING THREE-DIMENSIONAL IN VITRO MODELS TO IDENTIFY EARLY NEURONAL VULNERABILITY AND TO TEST THERAPEUTIC STRATEGIES IN AMYOTROPHIC LATERAL SCLEROSIS

Amyotrophic lateral sclerosis (ALS) is a rare neurodegenerative disorder characterized by relentless and progressive loss of motor neurons (MNs), which still lacks an efficacious therapy. Main limitations are represented by the incomplete knowledge of pathogenic mechanisms underlying neurodegeneration and the absence of models that fully recapitulate human disease. Moreover, the lack of reliable biomarkers able to formulate early diagnosis and patient stratification and to predict disease progression along with the inaccurate estimation of the number of affected patients has hampered the success of clinical trials. In 2011, a hexanucleotide repeat expansion (HRE) in the noncoding region of the C9orf72 gene was associated with the most frequent genetic cause of both sporadic and familial ALS and frontotemporal dementia (FTD). In the present study, we investigate several facets of ALS, with a particular regard to those caused by the HRE in the C9orf72. In the first part of this project, we leveraged whole genome sequencing (WGS) data from patients recruited to the 100,000 Genomes Project (100k GP), one of the largest world genomic dataset, to model the prevalence of C9orf72-ALS/FTD. We found that C9orf72-ALS genetic prevalence was up to two times higher than the literature reported prevalence, likely due to an under ascertainment of mild and heterogenous cases, reduced penetrance and different distribution of the genetic defect among different ancestries. The second chapter of this work involves the measurement of neurofilament light chain (NfL), chitinase 1 (CHIT1) and microRNA 181-b (miR-181b) in the cerebrospinal fluid (CSF) of patients with ALS and controls, to investigate their diagnostic and prognostic utility. Levels of the three biomarkers were significantly higher in ALS patients compared to controls, however, diagnostic performance of CSF NfL was the most accurate. In line with data from the literature, we showed that, although all measures were correlated with disease duration, CSF NfL was the best independent predictor of disease progression and survival, outperforming CSF CHIT1 and miR-181b as diagnostic biomarkers. 6 In the third part of the thesis, we exploited three-dimensional patient-specific in vitro models able to reproduce cell-to-cell connections in the central nervous system (CNS) aiming at dissecting specific disease hallmarks and crucial MN associated vulnerability pathways. The core hypothesis is that early developmental defects in C9orf72-ALS could lead to late onset neurodegeneration in C9orf72. To address this theory, we generated induced Pluripotent Stem Cell (iPSC)-derived brain (BrOs) and spinal cord (ScOs) organoids of C9orf72-ALS patients and isogenic controls, using a free-floating, 3D-culture method, based on aggregation in embryo bodies, embedment in matrigel (in solution or as scaffold), and agitation in spinning bioreactor. Structural and functional characterization as well as single-cell transcriptomics were performed. Further, phenotypic consequences upon treatment with an antisense oligonucleotide (ASO) targeting the C9orf72-HRE, Morpholino B (MOB) were reported. BrOs and ScOs expressed pluripotency markers and mature neuronal markers in early and late stages, respectively, resembling human neurodevelopment. ALS organoids presented increased cell death and lower maturation compared to isogenic controls, suggesting that late onset neurodegenerative disorders may arise from a disruption in cellular processes already in early stages of development. Calcium imaging recording showed different susceptibility to glutamate stimulation in ALS and isogenic BrOs. ALS organoids recapitulated disease hallmarks, like RNA foci and dipeptide protein repeats (DPRs), and displayed a disruption of key cellular processes like DNA damage response and axonal elongation, that were rescued after treatment with MOB. These findings support the use of patient-specific iPSC-derived 3D CNS models as tools for disease modelling since they improve the characterization of C9orf72-ALS pathology and as a platform to test therapeutic strategies. Findings from this project improve our knowledge on mechanisms involved in C9orf72-ALS and pave the way for personalized molecular therapies.