Microglial cells, the resident immune cells of the central nervous system, have particularly dynamic processes. Various studies suggested that beyond a possible role in surveillance, this dynamism could also allow physical communication with neighboring cells, including neurons. Thus, this cellular dynamism could be related to synaptic mechanisms, under the control of neuronal activity. However, the signaling pathways modulating neuronal control of microglial motility remain largely unknown.
Among many communication pathways between microglial cells and neurons, we decided to focus on one that involves a neuronal chemokine, fractalkine, and its microglial receptor, Cx3cr1. The interest of this signaling pathway, besides granting direct and specific communication between these two cell types, is its impact on chemotaxis, as well as on synaptic plasticity.
Thus, in order to study the impact of neuronal activity on microglial cells and their possible involvement in mechanisms of plasticity, we decided to study microglial morphodynamics during sleep-wake cycles. This approach allowed us to perform non-pathological modulations of neuronal activity, but also to study the possible function of microglial cells in a context known to modulate synaptic plasticity, namely sleep. Then, we studied the consequences of microglial Cx3cr1 receptor invalidation on these morphodynamic regulations to assess the involvement of the fractalkine pathway.
Morphodynamic changes were assessed by in vivo transcranial imaging using two-photon microscopy, while electroencephalogram and electromyogram were recorded to define vigilant states. We then evaluated the impact of sleep-wake cycles and fractalkine receptor invalidation to understand their role in microglial dynamics.
Our results indicate a decrease in microglial morphodynamics during slow wave sleep in the early inactive phase. In addition, we found that depletion of the fractalkine receptor abolished these morphodynamic changes, indicating that fractalkine may be involved in the microglia’s ability to detect and/or respond to changes in neuronal activity.
In conclusion, this work identifies fractalkine as a potential new modulator of microglial dynamics in response to changes in neuronal activity during sleep.

Microglia are the resident immune cells of the central nervous system. They widely contribute to neuroinflammatory processes, but their exact function in the intact adult brain remains elusive. In non- pathological conditions, microglia are characterized by highly ramified and dynamic processes that constantly probe the parenchyma. It has been shown recently that microglia contact neuronal elements among which synapses. However, the exact purpose of microglia-synapse contacts and the mechanisms controlling these contacts remain to be determined.
Some studies suggest that microglia could contribute to synaptic plasticity, and recent data indicate that variations in neuronal activity and states of consciousness correlate with changes in microglial morphodynamics. Adenosine, the byproduct of ATP, appears as a potential regulator mediating the relationship between neuron and microglia. In addition, adenosine is a key molecule in sleep regulation and adenosine receptors, expressed by microglia, could be involved in microglia chemotaxis. The present study aimed first at better characterizing the link between synaptic activity and microglial dynamics. The second objective was to determine if adenosine signaling pathways are involved in microglia-neuron communication.
To that end, in vivo imaging was performed on a transgenic mouse model to monitor both microglial morphodynamics and calcium activity of dendritic spines. Whisker stimulations (WS) were used to physiologically manipulate neuronal activity while imaging in the dendritic L2-3 layer of the barrel cortex. IP administrations of agonist and antagonist of adenosine receptors were used to evaluate the contribution of the adenosine pathways to this communication.
Our results indicated that the increase of spine activity triggered by WS induced a shortening of the distance between the spine and the closest microglial process, an increased probability of contact, and a longer duration of contact in average. Conversely, the decrease of spine activity triggered by WS induced a shortening of the contact duration and remote location of microglial processes in respect to their initial distance. In addition, we observed that the systemic administration of caffeine and CGS- 21680 had an impact on neuronal activity. Looking at microglial dynamics towards dendritic spines, we showed that both compounds blocked this activity-driven microglia-neuron coupling. More interestingly, we observed a decrease of microglial contacts in presence of the A2AR agonist CGS- 21680. Those results indicate that adenosine could be an important signaling molecule contributing to microglia-neuron interactions.
Overall, the present work shows that microglial processes’ proximity with spines is directly controlled by change of synaptic activity. In this bi-directional communication, the adenosine signaling pathway seems to play a major role. Further experiments will be necessary to better understand the role of adenosine in this communication, and more generally to overcome the function of this neuromodulator and microglia in the control of synaptic homeostasis.

LGI1 is a neuronal secreted protein expressed in the central nervous system, mainly in the hippocampus. Dysfunctions of LGI1 protein have been identified in two pathologies: an inherited form of autosomal dominant temporal lobe epilepsy (ADLTE) caused by mutations in LGI1 gene, and in a subtype of auto-immune limbic encephalitis (LE) characterized by neurological disorders due to the presence of LGI1 auto-antibodies in the serum and the cerebrospinal fluid of patients. To understand how the alteration of LGI1 protein leads to epileptic seizures, studies investigated LGI1 function in the regulation of the neuronal network. It was shown that LGI1 protein acts at excitatory synapses by forming a trans-synaptic complex with its transmembrane partners ADAM22 and ADAM23. Through this complex, LGI1 protein regulate the expression of AMPA-receptors (AMPA-R) involved in synaptic transmission. Surprisingly, it was shown that the impairment of LGI1 protein reduces the expression of AMPA-R at the surface of excitatory synapses. Nevertheless, this result is in contradiction with the epileptic seizures generated because of a hyperexcitability of the neuronal network. Thus, the goal of my thesis project was to understand how a reduction of AMPA-R expression leads to epileptic seizures when LGI1 protein in disturb. To explain this, we suggested that the reduction of the expression of AMPA-R would mainly occur at excitatory synapses on inhibitory neurons. This would decrease the inhibitory neuronal activity in favor of an hyperexcitability of the neuronal network. To investigate this, we studied the effects of a blockage of LGI1 protein by LGI1 antibodies (abs) purified from the serum of patients with LGI1 LE. Using an infused mouse model with LGI1 abs, we observed, by super-resolution microscopy STORM, that LGI1 abs reduced the expression of AMPA-R on excitatory and inhibitory neurons in the dentate gyrus of the hippocampus. By electrophysiology, we show that the blockage of LGI1 protein increases the hyperexcitability of the neuronal network. Moreover, we observed, by pharmacology, that LGI1 abs altered the inhibitory network activity that failed to control the neuronal excitability. Altogether, our results indicate that the reduction of AMPA-R expression in the dentate gyrus of the hippocampus, by LGI1 abs, reduce more drastically the activity of the inhibitory network. Thus, my results suggest that the alteration of the inhibitory network would be a mechanism involved in the trigger of epileptic seizures in patients with LGI1 LE.

Skeletal muscles are formed of multinucleated fibers. Muscle fusion is a crucial process that starts during embryogenesis and continues throughout lifetime during growth and repair. Fusion is tightly regulated in time and space. In fact, over-fusion leads to aberrant muscle masses whereas defect in fusion induces abnormal formation and/or repair of the muscles. The fusion process has been studied for years and many genes are known to be involved in this process. Most of them are required for fusion: during cell-cell recognition, adhesion, actin rearrangement or for transcriptional regulation. This field of research is still in expansion, recently new crucial pro-fusion genes have been identified, such as Myomaker and Myomerger. Even if the cellular events of fusion are well understood, the cellular and molecular mechanisms regulating fusion are more obscure. The goal of my work was to understand the dynamics and molecular mechanisms underlying this process during early muscle formation, particularly to identify inhibitors of fusion. A C2C12 esiRNA screen performed in our lab identified a number of genes involved in fusion regulation. Among them, the TGF-beta superfamily was identified as inhibitor of fusion without impacting on muscle differentiation. During my thesis I used in situ hybridization and in ovo electroporation in the chicken embryo to decipher the role of TGF-beta signaling in vivo during myogenesis. I show that TGF-beta regulates the pace of fusion by acting as a molecular brake of muscle fusion in vivo during the first step of myogenesis.

The spatial organization of the neurochemical synapse components conditions the efficiency of information transfer. In particular, the location of receptors close to neurotransmitter release sites is a determining factor in the postsynaptic response. This distribution most often depends on cytoplasmic protein scaffolding that anchors the receptors to the synapse. Nevertheless, extracellular matrix proteins (ECM) are also actively involved in the organization of the synapse. In C. elegans, MADD-4 (or Ce-punctin) is a ECM protein secreted by monotoneurons that acts as an organizer of neuromuscular junctions and controls the cholinergic excitatory or GABAergic inhibitory identity of the postsynaptic domains. During my thesis, I showed that MADD-4 controls the synaptic localization of syndecan, a MEC protein belonging to the heparan sulfate proteoglycan (HSPG) family. HSPGs are involved in cell and axonal migration in the mammalian central nervous system. They are also present at synapses, but their roles are poorly characterized. I have been interested in the role of syndecan in controlling the localization of acetylcholine and GABA receptors at neuromuscular junctions. Syndecan (sdn-1) is a type I transmembrane protein consisting of an extracellular domain bearing glycosaminoglycan chains, and an intracellular domain whose last amino acids (EFYA) allow binding to proteins containing a PDZ domain. In the absence of syndecan, axonal guidance defects have been observed in previous studies. I have shown that the syndecan is enriched at the neuromuscular junctions of C. elegans and is localized at both types of synapses. Insertion of a DEGRON label allowed tissue-specific degradation of the syndecan protein, demonstrating that its presence at the synapse is largely dependent on muscle tissue. I also studied the domains of SDN-1 necessary for its synaptic localization. The extracellular domain is necessary, but glycosaminoglycan chains do not seem to be required for its localization at the synapse. In addition, the presence of MADD-4 appears to be essential for the recruitment of SDN-1 at synapses. The absence of syndecan leads to a decrease in the quantity of receptors at synapses, the most drastic of which was observed for cholinergic receptors composed by the ACR-16 subunit, which are sensitive to nicotine. The PDZ binding domain of syndecan is essential for the localization of these receptors. This study positions syndecan as a major player in the synaptic localization of nicotine-sensitive cholinergic receptors in C. elegans. These results allow new hypotheses on the function of syndecans in the CNS of vertebrates.

Two-pore domain (K2P) potassium channels belong to a large family of ion channels implicated in determining and maintaining the resting cell membrane potential. K2P channels are proteins extensively conserved throughout evolution, being present in almost all animal cells. In the nematode Caenorhabditis elegans, 47 genes code K2P channels sub-units, but only three of them have been characterized and reported in the literature. By tagging a certain number of them with fluorescent proteins (CRISPR/Cas9), we have found that nine channels are co-expressed in body wall muscle, showing a highly specific sub-cellular distribution. The most fascinating distribution was the one of TWK-28, which exhibits a polarized comet-like pattern that occupy only the anterior tip of each body wall muscle cell. In order to elucidate the cellular mechanisms underlying this particular distribution, we performed a genetic screen on the novel TWK-28 gain-of-function strain. We revealed that genes belonging to Dystrophin-Associated Protein Complex (DAPC) are involved in determining the amount of this channel at the muscle cell surface. DAPC is composed of at least 10 intra and extracellular proteins and plays a key role in physically connecting the extracellular matrix to the actin cytoskeleton. Interestingly, when tagging multiple components of DAPC with fluorescent proteins by CRISPR/Cas9 gene editing, we found that most of the dystrophin-associated proteins, such as syntrophin/STN-1, dystrobrevin/DYB-1 or even sarcoglycans (SGCA-1 and SGCB-1), show a particularly asymmetric distribution in muscle. We also revealed the, to date excluded, presence of dystroglycan/DGN-1 in body wall muscle of C. elegans. Finally, the asymmetric distribution of TWK-28 along the antero-posterior axis on a cellular and tissue scale, suggests that the Planar Cell Polarity pathways might be implicated. By gene candidate approach of the WNT pathway, we showed that proteins such as Disheveled, ROR/CAM-1 or WNT ligand/EGL-20 can modify the localization of TWK-28 by driving it into a new posterior sub-complex in the muscle cells.

Antibodies against CASPR2 (Contactin-2 Associated Protein), a neuroglial cell-adhesion protein, have been described in at least three neurological syndromes: autoimmune limbic encephalitis, acquired neuromyotonia (or Isaacs' syndrome) and Morvan syndrome. However, the clinical phenotype associated with anti-CASPR2 antibodies is not yet completely understood. In addition, some authors consider that instead of specific syndromes, anti-CASPR2 antibodies associate with a set of core symptoms that combine randomly in the patients. Last, the pathophysiologic factors underpinning clinical variability in the anti-CASPR2 antibodies patients are unknown. In this PhD project, we use a nationwide, retrospective cohort of anti-CASPR2 antibodies patients, in order to address the issue of the clinical characterization of anti-CASPR2 antibodies patients. We aimed at describing the clinical presentation of CASPR2 encephalitis and Morvan syndrome, studying the outcomes of CASPR2 encephalitis, and analyzing the repartition of the patients' symptoms in order to assess if the symptoms are distributed randomly or if instead they form distinct clinical patterns. The present PhD project is divided into three studies. In the first study, we analyze clinical presentations and outcomes of anti-CASPR2 antibodies patients with limbic encephalitis. We observe that most patients were males from 50 to 75 years old, and frequently had extra-limbic symptoms, such as cerebellar ataxia. In addition, response to immunotherapy was good, even though 25% of the patients did not return to baseline and were left with residual symptoms, including cognitive disturbances, epilepsy, and cerebellar ataxia. In the second study, we describe the first reported cases of autoimmune episodic ataxia, a novel symptom that so far has been found only in anti-CASPR2 antibody associated autoimmune limbic encephalitis patients. It consists in transient episodes of paroxysmal ataxia, and is reminiscent of hereditary episodic ataxia. Interestingly, we found in two patients rare variants of CACNA1A and KCNA1, two genes involved in the main types of hereditary episodic ataxia. While the impact of these variants on ion channel functions is unknown, it raises the question of the role of the genetic background in phenotype determination in anti-CASPR2 antibodies patients. In the third study, we use a statistical cluster analysis to assess anti-CASPR2 antibodies patients' symptoms combinations. We found that the symptoms do not form random combinations, but that instead clinical patterns can be identified, which correspond to patients with limbic encephalitis, Morvan syndrome, and neuromyotonia. In addition, we confirm the expansive clinical presentation of limbic encephalitis, since more than a third of the patients had non-limbic symptoms such as cerebellar ataxia, dysautonomia, weight loss, and movement disorders. Notably, less than ten percent of the patients had a combination of neuromyotonia and limbic symptoms. Finally, the Morvan syndrome patients had severe peripheral nerve hyperexcitability features, severe dysautonomia, severe insomnia, weight loss, and frequently had a malignant thymoma. This clinical classification into three specific syndromes is supported by differences in term of autoantibody specificities, as limbic encephalitis patients tended to have higher anti-CASPR2 antibodies levels and were more frequently cerebrospinal fluid-positive, and by the genetic background, since the Morvan syndrome patients did not have the HLA DRB1*1101 association that is found in limbic encephalitis patients. In conclusion, the present PhD project supports the view that anti-CASPR2 antibodies patients can be classified into three specific syndromes, autoimmune limbic encephalitis, neuromyotonia, and Morvan syndrome. Differences in etiopathogeny likely account for the clinical variability observed in anti-CASPR2 antibodies patients.

Axon guidance receptors often bind several ligands, themselves mediating their effects via different receptors, which complexifies the understanding of functional outcomes downstream of each ligand-receptor couple. Likewise, during commissural axon navigation across the spinal cord midline in the floor plate (FP), PlexinA1 (PlxnA1) contributes as a receptor for two guidance cues, SlitC and Sema3B when complexed with Neuropilin2 (Nrp2). SlitN, produced with SlitC by full-length Slit processing bind other receptors, the Roundabouts (Robo1/2). To decipher the selective contributions of PlxnA1/SlitC signaling to the FP navigation, we generated a mouse strain baring PlxnA1 Y1815F mutation, previously found in vitro to abolish SlitC but not Sema3B responses. We found that this mutation selectively recapitulates one of the phenotypes of PlxnA1-/- embryos. This outcome is not triggered by specific spatial arrangement of SlitC in the FP, as using various fluorescent reporter tools we found SlitC has a pattern closely resembling that of SlitN, suggesting SlitC outcome might rather be set at the receptor levels. To get further insights, we explored the nature of the process requiring the Y1815 residue. Through imaging of pHLuo-tagged PlxnA1 receptor in various ex vivo paradigms and mass spectrometry analysis, we observed that Y1815 alters the surface dynamics of PlxnA1 and its network of interactions, revealing this is a crucial parameter for SlitC functions. More generally, our study provides the first dissection of ligand-specific outcomes of a guidance receptor during commissural axon navigation.

Long distance navigation of axons is marked by choice points, instructing highly stereotyped directional changes of axon trajectories. In this stepwise model, each step is thought to be essential for the next one, but intriguingly, examples suggest that pathway experience can be dispensable for axons to reach their final destination. We investigated pathway-independent ability of axons to locate their target, using two populations of spinal cord neurons having drastically different target location in the organism: the dorsal interneurons, which target the central nervous system and ventral motoneurons, which target muscles. After grafting these neurons at ectopic positions in the chicken embryo, both neuron-types were observed to form axons which, remarkably, oriented towards and reached appropriate targets. This suggests that, in the embryo, an overall guidance information might exist that enables the axons to locate positions over large scales. Beside well-studied chemical cues, bioelectric signals are attractive candidates for this function. Electric Fields (EF) were detected in the embryo and reported to encode spatial information. Thus, using in vitro set-ups, we investigated whether EFs in the range of the ones measured in the embryo could influence the navigation of chick motor and dorsal interneuron axons. We found that both axon subsets orient parallel to EFs. Yet, they significantly exhibited different sensitivities, which could contribute to elicit different trajectory choices in vivo. Next, we found that Concanavalin A (ConA) could block axon response to EF, supporting a role of cell surface receptors known to bind to ConA. Thus, we performed a pharmacological screening on ion channels and pumps that bind ConA and identified Na+/K+ ATPases as promising candidates. Preliminary knock-down experiments targeting Na+/K+ ATPases subunits suggest their contribution to CE response and axon navigation in vitro and in vivo. Collectively, our findings should provide novel insights into the mechanisms ensuring axon guidance fidelity and resilience and reveal unknown contributions of bioelectric signals and Na+/K+ ATPases during neuronal development.

Les commissures forment un ensemble de connexions nerveuses assurant la communication entre les neurones de chaque hémi partie du système nerveux central des bilatériens. Au cours du développement embryonnaire, les axones des neurones commissuraux sont guidés au travers de la ligne médiane délimitant ces deux parties. Plusieurs sources de signaux de guidage attractifs et répulsifs agissent de concert pour organiser les trajectoires de ces axones. Dans la moelle épinière, les axones commissuraux traversent la ligne médiane dans un territoire ventral, la plaque du plancher (PP). Au cours de la traversée de la PP, ils acquièrent une sensibilité à des signaux répulsifs exprimés par ce territoire qui leur empêchent de rebrousser le chemin et qui les poussent hors de la PP. Plusieurs couples ligands/récepteurs médient ces forces répulsives mais les mécanismes qui sous-tendent l'acquisition de la sensibilité aux signaux répulsifs restent encore peu connus. Par exemple on ignore si les axons se sensibilisent à tous les signaux répulsifs en même temps, quand précisément ce switch de réponse se fait, et les contributions précises de chacun de ces signaux. Une spécificité fonctionnelle est suggérée par l'analyse des phénotypes d'invalidation des gènes codant pour ces récepteurs chez la souris ou encore par des manipulations d'expression chez l'embryon de poulet. L'objectif de mes travaux de thèse a été de tester l'hypothèse selon laquelle la génération de spécificités fonctionnelles pourrait résulter de contrôles précis et distincts de la dynamique spatiale et temporelle des récepteurs de guidage à la surface du cône de croissance. J'ai tout d'abord développé un dispositif de vidéomicroscopie adapté à l'enregistrement de cônes de croissance accomplissant la traversée de la PP, sur des moelles épinières en configuration de «livre ouvert». Afin de visualiser l'adressage à la surface du cône de croissance, j'ai exploité une forme de GFP sensible au pH, dont les propriétés de fluorescence à pH neutre permettent un suivi spécifique du pool de surface des protéines (Nawabi et al., 2010; Delloye-Bourgeois et al, 2014). J'ai utilisé ce paradigme pour comparer la dynamique temporelle de 4 récepteurs médiant les réponses aux divers signaux répulsifs de la PP: Nrp2, Robo1, Robo2 et PlxnA1. Les vecteurs d'expression de ces formes pHLuo de récepteurs ont été introduits dans les neurones commissuraux de la moelle épinière d'embryon de poulet par électroporation in ovo. Par des approches de microscopie à super-résolution sur les livres-ouverts, j'ai aussi étudié la distribution spatiale des récepteurs répulsifs à la surface des cônes de croissances au cours de la traversée. L'ensemble de ces expériences a pu démontrer que les récepteurs sont adressés à la membrane à différents temps de la navigation de la PP et occupent, de plus, des domaines distincts du cône de croissance. J'ai ensuite adapté la technique d'électroporation à la moelle épinière d'embryon de souris. Ces expériences ont montré que les séquences temporelles observées chez le poulet sont conservées chez la souris. J'ai également réintroduit le récepteur Robo1 dans une lignée de souris présentant une invalidation des récepteurs Robo1/2 et montré que l'altération de la traversée de la PP caractéristique de cette lignée est abolie dans la population d'axones capables d'adresser le récepteur Robo1 à la membrane. Au final, mes résultats démontrent que les axones commissuraux ne sont pas sensibilisés aux signaux répulsifs par la mise en œuvre d'un programme général. Au contraire, les récepteurs de guidage possèdent des profils de dynamiques temporelles spécifiques, et des domaines de distribution distincts dans le cône de croissance. Le contrôle de la dynamique d'adressage représente ainsi un mécanisme permettant de discriminer des signaux concomitants, en les fonctionnalisant à différents temps de la navigation de la moelle épinière

Two-pore domain potassium channels (K2P) are major regulators of cell excitability, playing a central role in the establishment and maintenance of the resting potential of animal cells. Despite their fundamental role, little is known about the cellular processes that control K2P channels function in vivo. In particular, we know only few factors that directly control thenumber, activity, and localization of K2P on the cell surface.During my thesis, I used state-of-the art genome engineering technologies combined with genetic approaches to characterize the C. elegans potassium channel EGL-23. For this, I realized a phenotypic suppressor screen of the egg-laying defective mutant egl-23(n601) and a visual screen on an egl-23 translational fluorescent reporter. Using whole genome sequencing, I was able to clone for new genes involved in EGL-23 regulation

Two-pore potassium channels (K2P) control cell excitability and play a central role in the establishment and the maintenance of the resting membrane potential of almost all animal cells. Since their identification in the late 90s, these channels have been implicated in a large number of functions ranging from neuronal and cardiac activity to kidney physiology. Despite the crucial functions of these channels, comparatively little is known about the cellular processes controlling their function in vivo. During my PhD, I used a wide range of strategies including genetics, microscopy and electrophysiology to understand how the expression, the distribution and the activity of the K2P channel UNC-58 are controlled in the model nematode C. elegans. I have first performed a genetic suppressor screen targeting the locomotion phenotype of the gain of function mutant unc-58(e665). The screen yielded 133 mutants, displaying a wide range of suppression level, suggesting that several genes may be implicated in the channel regulation process. By using whole genome sequencing technologies, I’ve been able to clone six new genes required for the function of UNC-58.Then, I’ve characterized in detail the role of unc-44/ankyrin in the trafficking of UNC 58. This project led to 4 main conclusions : (1) UNC-58, despite its potassium channel structure, has an altered ionic selectivity, allowing preferably sodium ions to pass through the channel (2) the addition of a fluorescent protein to UNC-58 by CRISPR/Cas9 approaches allowed us for the first time to directly observe the addressing of the UNC-58 channel to the muscle surface and axons of ALM mechanosensory neurons. This function involves a lipid binding pocket located within the Zu5N-Zu5C-UPA module of UNC-44, (4) this mechanism is highly selective since it is not required for the addressing of 6 other muscular channels.My screen also identified a genetic interaction between unc-70/ß-spectrin and unc-44/ankyrin. However, the exact molecular nature of this interaction remains to be elucidated.

Throughout nervous system development, activity of the post-synaptic targets can regulate the connectivity of neural networks, affecting both the number and strength of synapses. Using the neuromuscular junction of Caenorhabditis elegans as a model system, we studied two processes displaying such plasticity. First, we show that the number of receptors present at the neuromuscular synapse is regulated by muscle activity: an increase in synaptic activity can lead to a differential regulation of the three types of receptors present at the neuromuscular junction. Second, we studied the activity-dependent morphological changes of one type of motor neurons in the worm’s head, called the SAB neurons. A decrease of muscle activity during a critical developmental phase leads to SAB axonal overgrowth. Using several approaches, we were able to observe suppression of SAB axonal overgrowth in mutants with a disruption of neuropeptides biosynthesis. Finally, we give evidence that axonal overgrowth also occurs following more general disruptions of cell physiology, such as a heat-shock or transgene overexpression, which suggest that the SAB system is plastic and sensitive during development.

The epithelio-mesenchymal transition (EMT) is a well-known mechanism by which epithelial cells lose their adherent connections and gain migratory properties, associated with a gain of a mesenchymal phenotype. This EMT is required in numerous processes as gastrulation, organogenesis, fibrosis and cancers. Various molecular pathways orchestrate the EMT depending on the EMT biological context. Recently, our laboratory highlighted the implication of the cytoplasmic Notch pathway in the dorso-medial lip (DML) EMT. In the DML tissue, theEMT is synchronized with differentiation pathways, to generate cells forming the primary myotome. Our laboratory showed that neural crests cells expressing DLL1 activate NOTCH receptor of the DML cells, via a “kiss and run” model. This leads to NOTCH cleavage, releasing an activated intra-cytoplasmic NOTCH domain (NICD). In the cytoplasm, NICD inhibits the GSK3ß kinase, leading to the stabilization of SNAIL and the free cytoplasmic ßcatenin. These molecules translocate into the nucleus and lead to the activation of MRF as Myf5 (ß-catenin) and to the repression of adherent genes (SNAIL). Therefore, Notch cytoplasmic pathway allows a synergized induction of both, the EMT and myogenic programs. This pathway remains controversial and the precise mechanism how NICD inhibits GSK3ß needs to be elucidated. Therefore, the aim of my thesis project was to clarify how NICD inhibits GSK3ß activity. First, I confirmed that NICD and GSK3ß physically interact by CoIP. Moreover, I demonstrated that the serin-threonin kinase AKT, known to inhibit GSK3ß by phosphorylation and also to mediate EMT in cancer, can physically interact with NICD in the cytoplasm. I have also shown that AKT mediates the induction of the myogenic program through the inhibitory phosphorylation of GSK3ß and that SNAIL is downstream of AKT. Together, these experiments indicate that AKT mediates, through phosphorylation, the cytoplasmic NICD inhibition of GSK3ß leading to myogenesis. A comparison of the chicken NICD1 and the 4 isoforms of mouse NICD highlighted that these 5 proteins induce EMT and myogenesis similarly. The dissection of the different conserved domains in the 5 different NICD proteins demonstrated that the RAM domain, known to activate transcription by binding to RBPJ, is necessary and sufficient for GSK3ß inhibition. A second axis of the thesis has been to test the involvment of the cytoplasmic Notch pathway in other EMT contexts. First, I highlighted that this pathway induces myogenesis, showing that NICD inhibits GSK3ß activity in the ventro-lateral lip. I further demonstrated that the cytoplasmic Notch pathway is implicated in the EMT and differentiation of the neural crests cells delaminating from the dorsal neural tube. Particularly, I have shown a co-activation of the Wnt and Notch pathway in premigratory and migratory neural crests. Moreover, I demonstrated a cytoplasmic inhibition of the kinase activity of GSK3ß by NICD, as well as the induction of the differentiation by cytoplasmic ß-catenin or SNAIL2. In a third axis of my thesis, I tried to clarify the regulatory mechanism involved in Notch activation. Previously it has been demonstrated that in all the DML cells Notch can be activated by an overexpression of DLL1 and that an ectopic expression of NICD in the DML cells induce a massive differentiation and depletion of the progenitor pool. To determine if the regulation of this initiation of the myogenic program occurs before or after Notch activation, I designed a plasmid to visualize Notch activation in vivo. In order to be able to follow the DLM cells and Notch activation in vivo, I initiated a collaboration with an ILM team to create a vertical SPIM biphoton microscope. In the future, this microscope will allow us to follow cells in living chicken embryos.

Anti-NMDA receptor (NMDAR) encephalitis is a rare but severe neuropsychiatric disorder initially described by J. Dalmau and colleagues in 2007. Anti-NMDAR encephalitis is defined by a clinical picture of encephalitis associated with the presence of IgG directed against the GluN1 subunit of NMDAR (NMDAR-Abs) in the cerebrospinal fluid (CSF) of patients. This disorder predominantly affects young women. Clinical presentation usually includes psychiatric symptoms and/or neurological symptoms often accompanied by decreased responsiveness and autonomic instabilities during the course of the disease . Despite the severity of the disease, 81% of patients recover fully or with mild sequels . 38% of patients had an underlying neoplasm, 94% of which were ovarian teratomas , indicating a role for this tumor in the immunopathogenesis of the disease. Studies in vitro and on animal models have demonstrated the pathogenicity of NMDAR-Abs but more studies are required to decipher the pathological role of anti-NMDAR antibodies. Two main research focuses have emerged in our group: understanding the events leading to the immune dysregulation in the ovary teratoma and identifying the pathological element(s) and how they act at the molecular and cellular levels to cause the broad neurological spectrum of symptoms observed in patients. My PhD was especially focused on 1) understanding the involvement of the underlying ovary teratoma in the triggering of the immune response during anti-NMDAR encephalitis and 2) studying the impact of prolonged exposure of the neuronal network to patients’ NMDAR-Abs and the potential involvement of microglial cells in the physiopathology of the disease.

Neuroblastoma is a pediatric cancer diagnosed in very young children. In order to study the impact of embryonic signaling on the development of the disease in vivo, we have developed a new model for the study of neuroblastoma in the avian embryo. This new model allows us to study in vivo the contribution of different signaling to neuroblastic tumorigenesis. We have thus highlighted the role of Sema3C and HDGF in neuroblastic tumorigenesis.

Anti-NMDAR autoimmune encephalitis is a newly described pathology, characterized by the presence of IgG antibodies (Abs) directed against NMDA receptor (NMDAR). Glutamatergic synapses are the main component of excitatory transmission in the adult brain and the ionotropic glutamate receptor NMDAR has a key role in synaptic plasticity. Synaptic plasticity is defined by the synapses property to modify their transmission strength and seems to be the cellular correlate of learning and memory. In vitro and in vivo studies on anti-NMDAR Abs effects showed an altered dynamic at the membrane followed by an internalization of NMDAR. Thus, Abs seem to have a pathogenic effect, able to explain clinical symptoms. During my thesis, I wanted to deepen the understanding of this pathology on the clinical and mechanistic level. To this end, I studied the clinical and histological features of patients with an associated tumor. Results obtained allow me to establish management recommendations considering the high mortality rate in these patients and a higher frequency of immature ovarian teratomas than in the general population. Simultaneously, I studied the effect of anti-NMDAR Abs on synaptic transmission and plasticity in a murine model allowing chronic Abs infusion. These results obtained by electrophysiology on acute slices allowed me to demonstrate an effect of CSF Abs on synaptic plasticity but of less amplitude that the in vitro studies had suggested. Results also highlighted the importance of the duration to antibodies exposure.

Cilia and flagella are highly conserved organelles among eukaryotes species. They are composed of a microtubular cytoskeleton and play essential functions during development and in numerous physiological processes. As a result, in humans, cilia dysfunction leads to a wide range of pathologies, called ciliopathies.

Cilia and flagella are highly conserved organelles among eukaryotes species. They are composed of a microtubular cytoskeleton and play essential functions during development and in numerous physiological processes. As a result, in humans, cilia dysfunction leads to a wide range of pathologies, called ciliopathies.At the ciliary base, the transition zone (TZ), a complex structure, is required for proper cilia assembly and regulates the traffic of ciliary components in and out cilia. Defects in TZ proteins lead to severe ciliopathies. The TZ is composed of 3 protein complexes, MKS, NPHP et CEP290 that closely interact. Additional proteins, like CBY, conserved between mammals and Drosophila, have been described at the TZ but their precise role and relationships with the other TZ complexes are unknown. Two modes of cilia assembly have been described: compartmentalized and cytosolic ciliogenesis. Whereas the function of the TZ in compartmentalized ciliogenesis is well documented, its role in cytosolic ciliogenesis remains poorly characterized. During my PhD, I characterized new TZ proteins conserved in mammals and Drosophila and analyzed their function during cilia assembly in Drosophila. First, I performed a proteomic screen in murine IMCD3 cells and characterized the CBY module composed of CBY, FAM92A1 and DZIP1L. This complex is conserved in Drosophila and locates at the TZ. Moreover, I showed that this module is necessary for TZ assembly and centriolar docking to the plasma membrane and hence required for cilia and flagella assembly. In absence of these proteins, Drosophila show severe ciliogenesis defects both in sperm cells and in sensory neurons.In conclusion, this work brings new insights into the understanding of TZ assembly and of the mechanisms, that control ciliogenesis.

During development, the orientation of cell division is crucial to correctly organize andshape tissues and organs and also to generate cellular diversity. As cell mitosis proceeds, thesegregation of chromatids and cytoplasmic material occurs along a division axis. Itsorientation largely determines the relative position of daughter cells and the partition ofmother cell subcellular domain between them. The orientation of the cell division isprefigured by the position of a complex microtubule-based scaffold, the mitotic spindle.Until now, the intrinsic molecular machinery positioning the mitotic spindle and its couplingto cell polarities have been study in details. In contrast, the contribution of extracellularsignals to cell division orientation is less characterised. My research shows that these signalsin the CSF contribute to the orientation of cell division in neural progenitors. Removal theCSF cues by opening the neural tube or by genetic engineering affects the proportion ofplanar and oblique divisions. We identified Semaphorin 3B (Sema3B), released from thefloor plate and the nascent choroid plexus, as an important actor in this extrinsic control ofprogenitor division. Knockout of Sema3B phenocopies the loss of progenitor access to CSF.Delivery of exogenous Sema3B to progenitors in living embryos compensates this deficiency.We showed that Sema3B binds to Neuropilin receptors at the apical surface of mitoticprogenitors and exerts its effect through GSK3b activation and subsequent inhibition of themicrotubule stabilizer CRMP2. Thus extrinsic signaling mediated by Semaphorins directs theorientation of progenitor division in neurogenic zones.In order to identify new factors implicated in Sema3B-dependant mitotic spindleposition, we performed a transcriptomic analysis of Sema3B -/- neural progenitors. Severalderegulated candidate genes were considered. In the second part of my thesis, I focus onone of this, Norbin/Neurochondrin. Interestingly, the invalidation of Norbin/Neurochondrinalters the orientation of the mitotic spindle in HeLa cells.My PhD work reveals the contribution of a large family of topographic cues known tofunction in axon guidance has a novel role in the orientation of cell division.

Neurite formation is a crucial step of neuronal differentiation. However, the mechanisms that determine how and at which position neurites emerge in the soma are still poorly understood. We postulated that a molecular polarity could prefigure the morphological differentiation, with some molecules that could accumulate at the future site of axon initiation. Interestingly, such molecular polarity has been evidenced in the contest of yeast budding, with bud forming at specific position relatively to the previous bud site. Genome-wide screen identified hundreds of proteins that control bud site location. Among the vertebrate molecules homologous to those involved in budding site selection, we selected the Septins as promising candidates. These GTP-ases form filaments that act as diffusion barriers and molecular scaffolds. We investigated the contribution of Septins to axon initiation using the chick dorsal root ganglion (DRG) neurons as a model. Monitoring of cell morphology in nascent ganglia indicates that DRG neurons form a single axon at the ventral pole and a second one at the dorsal pole and that these axons seem to emerge directly after their last division. This suggests that two initiation sites are selected at opposite pole of the soma.We found that Septins homologous with those controlling budding are expressed in the early DRG developmental stages. My analyses by time-lapse video-microscopy showed that Septin7 accumulate at the site of axon emergence, just before or during its formation.We observed that a pharmacological inhibitor and a dominant-negative construct block axon formation both in vitro and in vivo respectively. Furthermore, blocking Septin function leads to the appearance of uncommon round or sea urchin-like neurons. Thus, Septins appear to regulate early step of morphological differentiation of DRG neurons, possibly by controlling axon initiation site selection.

Les cils et les flagelles sont des organites présents à la surface cellulaire. Ils sont conservés chez les eucaryotes chez lesquels ils jouent un rôle essentiel dans la régulation de nombreux processus physiologiques. La zone de transition (ZT) est une structure complexe, localisée à la base des cils, qui assure une fonction importante dans l'assemblage et la régulation du trafic des constituants ciliaires. Trois complexes protéiques ont été identifiés à la ZT : MKS-JBTS, NPHP1-4-8 et NPHP5-CEP290. D'autres protéines sont également situées à la ZT telles que CBY et AZI1 mais leur interaction avec ces trois modules reste encore peu connue. Chez l'Homme, des mutations de gènes codant des protéines de la ZT sont associées à des maladies génétiques rares, les ciliopathies. Deux modes d'assemblage des cils ont été décrits : la ciliogenèse compartimentée et la ciliogenèse cytosolique. Alors que la fonction de la ZT au cours de la ciliogenèse compartimentée a été bien étudiée, son rôle dans la ciliogenèse cytosolique reste peu connu. La Drosophile possède deux sortes de cellules ciliées, les neurones sensoriels et les flagelles de spermatozoides dont les cils s'assemblent selon ces deux modes d'assemblage. Au cours de ma thèse, j'ai utilisé ce modèle pour analyser la fonction des protéines de la ZT dans ces deux types cellulaires. Mes résultats montrent que les protéines MKS ne jouent pas un rôle essentiel dans l'assemblage de la ZT dans ces deux types cellulaires. J'ai aussi révélé que CBY et AZI1, coopèrent pour assembler la ZT et qu'elle est nécessaire à l'ancrage du corps basal à la membrane plasmique. De plus, mes travaux ont démontré que KLP59D, une kinésine dépolymérisante des microtubules, est indispensable à la régulation de l'élongation de l'axonème au cours de la ciliogenèse cytosolique. En conclusion, ce travail apporte de nouvelles connaissances sur la dynamique d'assemblage de la ZT des cils et sur les mécanismes qui contrôlent l'élongation de l'axonème.