AGGREGATION MECHANISMS OF TDP-43 PROTEIN IN RESPONSE TO STRESS IN AMYOTROPHIC LATERAL SCLEROSIS AND THERAPEUTIC APPROACHES

Amyotrophic Lateral Sclerosis (ALS) is a rare neurodegenerative disorder affecting upper and/or lower motor neurons for which no effective cure is available to date. The neuropathological hallmark of the disease, present in 97% of ALS cases, consists in TDP- 43 proteinopathy characterized by the accumulation of ubiquitinated and phosphorylated TAR DNA-binding protein 43 (TDP-43) in the cytoplasm, accompanied by the concomitant loss of TDP-43 splicing activity in the nucleus of affected neurons. The scarcity of experimental models of TDP-43 proteinopathy capable of recapitulating both gain and loss of TDP-43 function has hindered the discovery of novel therapeutic approaches for the treatment of ALS. Aim of this work was therefore to generate human 2D and 3D in vitro models of TDP-43 proteinopathy, recapitulating TDP-43 cytoplasmic mislocalization and loss of nuclear function, to be used for drug screening approaches. We induced a chronic and mild oxidative insult by sodium arsenite (ARS) in human neuroblastoma SK-N-BE cells and ALS patient-derived primary fibroblasts, induced pluripotent stem cells (iPSCs)-motor neurons and brain organoids. As a readout of TDP-43 functionality, we tested its subcellular localization and its splicing activity towards the target genes UNC13A and POLDIP3. We demonstrated that chronic oxidative stress in these cellular models induced TDP-43 cytoplasmic mislocalization and defective TDP-43 splicing activity, thus recapitulating both TDP-43 proteinopathy features. Since we observed dysregulated gene expression of autophagy and senescence markers in ARS-treated human neuroblastoma cells, we preliminarily tested in this experimental model three drugs involved in promoting autophagy, namely rapamycin, lithium carbonate and metformin. Only rapamycin was capable of significantly rescuing ARS-induced loss of TDP-43 splicing activity on UNC13A. We therefore tested rapamycin in primary fibroblasts obtained from ALS patients with mutations in the C9orf72 gene, where we demonstrated that it reduced ARS-induced formation of phosphorylated TDP-43 (P-TDP-43) aggregates and stress granules (SGs). We also tested rapamycin in iPSCs-motor neurons from C9orf72-mutated patients, where we confirmed its efficacy in attenuating P-TDP-43 aggregates, SGs and loss of TDP-43 splicing activity on UNC13A induced by ARS treatment. We then employed a more complex 3D neuroglial model, namely iPSC-brain organoids, obtained from C9orf72 ALS patients and exposed to chronic ARS, where we observed that rapamycin rescued loss of TDP-43 splicing activity on UNC13A gene. In conclusion, we generated and characterized different human cell models of TDP-43 proteinopathy in which rapamycin was able to rescue chronic oxidative stress-induced TDP-43 cytoplasmic mislocalization and defects in its splicing activity. We therefore propose that human SK-N-BE cells and ALS patient-derived 2D and 3D iPSC models chronically treated with ARS can be exploited as in vitro models to be used for future drug screening approaches.