Abstract
Wolfram Syndrome (WS) is a neurodegenerative disease that combines diabetes mellitus, diabetes insipidus, optic atrophy, sensorineural hearing loss and various neurological symptoms. Patients die early, most often from respiratory distress or miscarriage, and to date no effective treatment is available. WS is caused by mutations in the WFS1 gene that encodes an endoplasmic reticulum (ER) transmembrane protein, Wolframine. Mutations generally reduce the stability of the protein, altering its homeostasis, decreasing calcium transfer, increasing the activation of the Unfolded Protein Response (UPR) and leading to mitochondrial dysfunction and cell death. The objective of my thesis was to explore original therapeutic approaches in this indication. For this purpose, several models were used and characterized: patient fibroblasts in vitro, and in vivo, wfs1aC825X, wfs1bW493X and wfs1abKO mutant zebrafish lines, and Wfs1∆Exon8 mice. We showed that: (1) These different models develop cellular alterations, such as mitochondrial deficits, impaired mitochondrial calcium transfer, activation of UPR pathways, and autophagy, associated with behavioral deficits, such as visual, locomotor, anxiety, and cognitive impairments. (2) Overexpression of NCS1, a partner of wolframine, in vivo restored the behavioral and cellular deficits developed by the zebrafish line wfs1abKO confirming the in vitro study previously conducted on patient fibroblasts. (3) The chaperone protein sigma-1 (S1R) was a relevant target to restore functional ER-mitochondria communication in WS. Our in vitro and in vivo studies showed that activation of S1R by an agonist or by overexpression of its mRNA reversed behavioral and cellular alterations. (4) Based on the motor response of wfs1abKO zebrafish larvae, we conducted a phenotypic screening of 371 molecules potentially affine to S1R which led to the identification of 8 hits allowing to fully restore the locomotor deficit of mutant larvae. Some of them will be valorised by an optimization approach, but among the repositionable molecules, we have identified MED0092 which acts as a positive modulator of S1R and whose beneficial effects have been shown on behavioral deficits in two mouse models of neurodegenerative diseases, SW and Alzheimer's disease. To conclude, this thesis work has not only allowed the phenotyping of new mouse and zebrafish models of WS, but also the validation of two therapeutic targets allowing a functional recovery of MAM: NCS1, which can be directly overexpressed by gene therapy, and S1R, which can be targeted by small molecules and for which new very promising chemical series of synthetic or natural origin have been highlighted.