Rett syndrome (RTT) is a severe neurological disorder that affects approximately 1:10000 girls. Classical RTT is defined by a developmental regression phase and subsequent stabilisation of diagnostic criteria, which include partial or complete loss of spoken language, dyspraxic gait and stereotypic hand movements such as hand mouthing. RTT is a monogenic disorder, with the majority of cases being due to loss-of-function mutations in MeCP2 (methyl-CpG binding protein 2). Due to this clear genotype-phenotype link multiple RTT mouse models have been used to elucidate the molecular details, and consequent neuropathogenesis, of this complex neurological disease, as well as for the development of potential therapeutics for RTT. However, as the molecular details become clearer, the need for a simpler model system becomes evident. Human induced pluripotent stem cells (hiPSCs) generated from RTT patient fibroblasts are an option; however the handling of these cells is laborious, time-consuming and expensive and they often differentiate into a heterogeneous population of cells. To explore an alternative human model system I have been genetically engineering and experimenting with the human dopaminergic LUHMES cell line. LUHMES cells are an immortalised pre-neuronal cell line derived from an 8-week old, female foetus and can readily be differentiated into a homogeneous population of mature, electrically active neurons in just one week. In this thesis I have assessed the phenotypic properties of the wild-type cell line, demonstrated the ease of genetic manipulation of LUHMES cells by CRISPR/Cas9 approaches, generated seven mutant MECP2 LUHMES cell lines and explored the potential of protein therapy as a therapeutic approach for RTT. The LUHMES cell line proves to be extremely easy to handle and robust and has yielded novel molecular insights into the function of MeCP2 in human neurons. In particular, MeCP2-null cells show a striking relationship between the level of gene body methylation and the extent of transcriptional upregulation when compared to wild-type neurons. In contrast neurons that express a form of MeCP2 that can bind to DNA but cannot recruit a transcriptional corepressor complex (the R306C mutant) do not exhibit substantial gene expression alterations, yet do display a consistent decrease in total RNA amount. This decrease in total RNA is recapitulated in MeCP2-null LUHMES-derived neurons and in brain regions from MeCP2-R306C mice. The requirement for functional DNA binding for normal gene-body methylation dependent gene repression is demonstrated by assessing LUHMES cells that overexpress MeCP2-R111G, a protein that cannot bind to DNA. Furthermore, overexpression of the MeCP2-R306C protein highlights the importance of NCoR binding for normal gene repression, but also demonstrates that MeCP2-R306C protein retains some gene repression activity. Thinking more broadly, this cell line also has applications as a model system for a variety of other neurological disorders; as a simplified model system to elucidate molecular and neurological phenotypes, and as a relevant human system that can be cultured in a high-throughput manner for testing therapeutic strategies.

Dans de nombreuses maladies neurodégénératives dont la Maladie de Parkinson, mais également dans le processus de vieillissement normal, il a été observé une accumulation excessive des niveaux de fer associée à une production accrue d‘espèces réactives de l’oxygène (ROS).Il est bien connu que le fer (i) participe à la réaction de Fenton pour produire le ROS le plus toxique, le radical hydroxyle, (ii) accélère l’auto-oxidation de la dopamine, augmentant le stress oxydant et (iii) augmente la peroxydation lipidique qui va déclencher l'agrégation des protéines, dont l’α-synucléine, une caractéristique de la maladie de Parkinson. Tous ces phénomènes rendent les neurones dopaminergiques très sensible au stress oxydant.Récemment, une nouvelle forme de mort cellulaire a été découverte sur des cellules cancéreuses. Cette mort cellulaire est appelée « ferroptose » car elle dépend essentiellement du fer intracellulaire. De plus, elle est liée à une très grande augmentation du stress lipidique.Le principal objectif de ce travail de thèse a donc porté sur la possible implication de la ferroptose dans la mort des cellules dopaminergiques.Dans un premier temps, nous avons montré que la lignée de neurones dopaminergiques (LUHMES) est un modèle particulièrement adapté à l’étude de mort cellulaire induite par le fer comparativement à d’autres modèles plus classique.La seconde partie du travail a donc été consacrée à l'étude de l’implication de la ferroptose dans les neurones dopaminergiques. D’abord par une approche cellulaire avec les cellules LUHMES, nous avons pu observer (i) que l’inducteur spécifique de cette mort, l’érastine, est très efficace pour induire la mort des cellules dopaminergiques, (ii) que l'inhibition de la ferroptose protégeait de la mort induite par des agents oxydants classiquement utilisés dans les études sur la maladie de Parkinson comme le MPP+ et (iii) que la PKC joue un rôle majeur dans la mort par ferroptose. Tous ces résultats ont ensuite été confirmés dans un modèle de culture organotypique, à l’interface entre la culture cellulaire et le modèle animal.Enfin, nous avons montré que l’utilisation des inhibiteurs de la ferroptose ainsi que celui de la PKC, confère une protection des neurones dopaminergiques dans un modèle de souris MPTP, modèle d’étude de la maladie de Parkinson.En conclusion ce travail de thèse montre pour la première fois l'implication de la ferroptose dans une maladie neurodégénérative comme la maladie de Parkinson et suggèrent que le développement d'inhibiteurs spécifiques de cette mort cellulaire pourrait être des futures cibles thérapeutiques possibles. / In many neurodegenerative diseases including Parkinson's disease, as well as in the regular aging process, an excessive accumulation of iron levels associated with increased production of oxygen species (ROS) have been observed.It is well known that iron (i) participates in the Fenton reaction to produce the most toxic ROS, the hydroxyl radical, (ii) accelerates auto-oxidation of dopamine, increasing the oxidative stress and (iii) increases lipid peroxidation that will trigger the aggregation of proteins, including α-synuclein, a hallmark of Parkinson's disease. All theses factors make dopamine neurons very sensitive to oxidative stress.Recently, a new form of cell death was found on cancer cells. This cell death is called ferroptosis because it essentially depends on the intracellular iron. In addition, it is linked to a very large increase in lipid stress.The main objective of this work has focused on the possible involvement of ferroptosis in the death of dopamine cells.First, we showed that the line of dopaminergic neurons (LUHMES) is a model particularly suited to the study of cell death induced by iron compared to other more conventional models.The second part of the work has been devoted to the study of the involvement of ferroptose in dopaminergic neurons. First by a cellular approach with LUHMES cells, we observed that (i) the specific inducer of this death, erastin, is very effective to induce the death of dopaminergic cells, (ii) inhibition of ferroptosis protected from death induced by oxidizing agents conventionally used in studies of Parkinson's disease like MPP+ and (iii) that PKC plays a major role in death by ferroptosis. These results were then confirmed in an organotypic culture model, at the interface between cell culture and animal models.Finally, we have shown that the use of ferroptosis inhibitors and PKC inhibitor, provides protection of dopaminergic neurons in MPTP mice model, animal model for studying of Parkinson's disease.In conclusion this study demonstrates for the first time the involvement of ferroptosis in a neurodegenerative disease such as Parkinson's disease and suggest that the development of specific inhibitors of this cell death could be possible future therapeutic targets.

Parkinson’s Disease (PD) is the second most common neurodegenerative disease (NDD) affecting approximately 1 - 2% of the population older than 65 and it is characterized by both motor and non-motor symptoms such as rest tremors, stooping posture, and rigidity. The neuromolecular basis of PD is quite complex and that is why there is a need for in vitro systems that can be utilized for studies of PD and NDDs in general. Human-derived cell lines are a good candidate for in vitro systems since they are easy to manipulate and are a less costly alternative to post-mortem human tissue sections or animal models. In this study, I optimize the Lund human mesencephalic (LUHMES) cell line differentiation protocol by determining that the optimal seeding density of cells is 37 500 cells/ml and that the differentiation media can contain quadruple the recommended concentration of tetracycline hydrochloride. Additionally, I use the differentiated LUHMES cells to conduct an exploratory study by treating the cells with cerebrospinal fluid (CSF) from PD patients and CSF from healthy individuals to investigate the neuromodulatory effects of the CSF on the neuronal cell culture. Cell viability assay showed neurotoxicity 24 hours post-treatment for the control CSF and 48 hours post-treatment for both control and PD CSF. Immunohistochemistry showed differential expression of proteins of interest that reflect hallmarks of neurodegenerative diseases. Further studies are needed to reach conclusive results.