Les macrophages primitifs envahissent la moelle épinière dès E11.5-12.5 chez l'embryon de souris. Ces cellules interagissent avec les neurites spinaux en croissance des neurones sensoriels des ganglions de la racine dorsale dès E12.5. La microglie est connue pour réguler plusieurs processus développementaux dans le système nerveux central, mais leur rôle dans le développement du système nerveux périphérique reste inconnu. Nous montrons que les macrophages primitifs régulent le développement des neurones sensoriels dans l'embryon de souris. Nous avons d'abord découvert que les macrophages primitifs interagissent avec les neurites périphériques des neurones sensoriels dès E11.5. Nous avons ensuite démontré que l'absence de macrophage chez les souris PU.1KO ou leur ablation immunopharmacologique, entraînait une réduction initiale des neuronaux TrkB+ et TrkC+ à E11.5, suivie d'une augmentation transitoire de leur nombre à E12.5. Cette augmentation est associée à une augmentation transitoire de leur mort développementale ce qui expliquerait pourquoi leur nombre redevient normal à partir de E13.5. Chez les embryons dépourvus de macrophage la mort augmente à nouveau à E15.5 pour les neurones TrkC+ et dès E14.5 pour les neurones TrkB+. Concernant les neurones TrkA+, leur nombre reste déficitaire entre E12.5 et E15.5, bien que leur mort développementale ne soit pas affectée. Nous avons enfin montré que la prolifération cellulaire était diminuée à E12.5 dans les ganglions des souris PU.1KO. Ces résultats sont la première démonstration que les macrophages primitifs et/ou les microglies immatures peuvent aussi réguler le développement embryonnaire du système nerveux périphérique. / The primitive macrophages invade the spinal cord as early as E11.5-12.5 in the mouse embryo. These cells interact with growing spinal neurites of the dorsal root ganglia sensory neurons as early as E12.5. Microglia is known to regulate several developmental processes in the central nervous system, but their role in the development of the peripheral nervous system remains unknown. We show that primitive macrophages regulate the development of sensory neurons in the mouse embryo. We first discovered that primitive macrophages interact with peripheral neurites of sensory neurons as early as E11.5. We then demonstrated that the absence of macrophage in PU.1KO mice or their immunopharmacological ablation resulted in an initial reduction of TrkB+ and TrkC+ neurons at E11.5, followed by a transient increase in their number at E12.5. This increase is associated with a transient increase in their developmental death which would explain why their numbers become normal again from E13.5. In macrophage-free embryos, death increases again to E15.5 for TrkC+ neurons and as early as E14.5 for TrkB+ neurons. Concerning TrkA+ neurons, their number remains deficient between E12.5 and E15.5, although their developmental death is not affected. We have finally shown that cell proliferation was decreased at E12.5 in ganglions of PU.1KO mice. These results are the first demonstration that primitive macrophages and / or immature microglia can also regulate the embryonic development of the peripheral nervous system.

A hemorragia intracerebral (HIC) é uma das principais causas de mortalidade em todo mundo, desencadeando eventos adversos e déficits neurológicos graves. A partir disso, essa tese teve como objetivo avaliar as respostas celulares e suas interações envolvidas na HIC e investigar os possíveis mecanismos que contribuem para os déficits sensório-motores relacionados, principalmente, à habilidade dos membros anteriores dos animais. Para isso foram realizados dois experimentos. Em ambos os animais de 90 dias foram submetidos à HIC induzida por colagenase no estriado dorsolateral. No primeiro experimento, após 24h, 72h e 7 dias os animais foram avaliados funcionalmente, e histologicamente quanto ao volume da lesão. Ainda, durante 6h, 24h, 72h e 7 dias, o estriado e o córtex foram avaliados quanto ao número de células neuronais, gliais, e em apoptose; a expressão de GFAP e S100B, no estriado e no córtex, a secreção de S100B, no soro e líquido cérebro-espinhal (LCE). A HIC provocou déficits sensório-motores e perda tecidual; induziu morte neuronal e gliose reativa no estriado. O córtex obteve uma resposta astrocitária similar, mas quanto à microglia houve um aumento apenas em 6h. A S100B correlacionou-se positivamente com o número de células gliais; e negativamente com a força e função dos membros anteriores dos animais lesados. No segundo experimento, foram avaliados o estado oxidativo, o nível de TNF-α, a atividade da glutamina sintetase (GS) e a captação de glutamato nos tempos de 6h, 24h, 72h e 7 dias pós-lesão. Após 6h da HIC houve aumento no nível de TNF-α e excitotoxicidade glutamatérgica; entretanto, a função de clearence astrocitário foi restabelecida 7 dias após o dano. A HIC gerou estresse oxidativo e aumento de antioxidantes endógenos, além disso, as espécies reativas de oxigênio (ERO) inibiram a atividade da GS em 24h e 7 dias após a lesão. A partir dos resultados obtidos, podemos concluir que a S100B atua na lesão hemorrágica produzindo respostas distintas, região dependente, contribuindo para o aumento das células gliais no estriado e córtex, correlacionando-se positivamente com os déficits funcionais. Este é o primeiro relato descrevendo a relação da S100B com as alterações celulares e funcionais durante a progressão da HIC experimental. Além disso, os mecanismos envolvidos na HIC demonstraram padrões e respostas envolvendo excitotoxicidade glutamatérgica e inflamação, que colaboraram para o dano oxidativo; em resposta ao estresse houve o aumento de antioxidantes endógenos que podem ter contribuído para o reestabelecimento da função astrocitária. / Intracerebral haemorrhage (ICH) is a major cause of mortality worldwide, triggering adverse events and serious neurological deficits. From this standpoint, this thesis aimed to evaluate the cellular responses and interactions involved in ICH and investigate the possible mechanisms that contribute to sensory-motor deficits, mainly related to the ability of the forelimbs of animals. For this, two experiments were conducted. In both, animals of 90 days of age were subjected to ICH induced by collagenase in dorsolateral striatum. In the first experiment, after 24h, 72h and 7 days, the animals were evaluated functionally, and histologically about the lesion volume. Also, at 6h, 24h, 72h and 7 days, the striatum and the cortex was evaluated as the number of neuronal, glial and apoptotic cells; the expression of GFAP and S100B, the striatum and cortex, S100B secretion in serum and cerebrospinal fluid (CSF). ICH caused sensorimotor deficits and tissue loss; induced neuronal death and reactive gliosis in the striatum. The cortex astrocyte obtained a similar response, but about the microglia there was an increase only at 6h. S100B was positively correlated with the number of glial cells; and negatively correlated with the strength and function of the forelimbs of injured animals. In the second experiment, were evaluated the oxidative state, the level of TNF-α, the activity of glutamine synthetase (GS) and glutamate uptake at 6h, 24h, 72h and 7 days post-injury. After 6h from ICH, there was an increase in TNF-α levels and glutamatergic excitotoxicity; however, the astrocyte clearance function was established 7 days after the damage. HIC generated oxidative stress and increase of endogenous antioxidants, in addition, reactive oxygen species (ROS) inhibited the activity of GS in 24h and 7 days after injury. Based on the results obtained, we can conclude that S100B acts in hemorrhagic lesions producing different responses, region-dependent, contributing to the increase in glial cells in the striatum and cortex, correlating positively with functional deficits. This is the first report describing the relation of S100B with cellular and functional changes during the progression of experimental ICH. Furthermore, the mechanisms involved in ICH displayed patterns and responses involving glutamatergic excitotoxicity and inflammation, that contributed to oxidative damage; in response to stress there was an increase of endogenous antioxidants which may have contributed to the re-establishment of astrocyte function.

Submitted by Daniella Sodre (daniella.sodre@ufpe.br) on 2015-04-08T18:22:35Z No. of bitstreams: 2 TESE Geórgia de Sousa Ferreira Soares.pdf: 2287209 bytes, checksum: 9762ebb4675d95dd929fcb3db11523e7 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Made available in DSpace on 2015-04-08T18:22:35Z (GMT). No. of bitstreams: 2 TESE Geórgia de Sousa Ferreira Soares.pdf: 2287209 bytes, checksum: 9762ebb4675d95dd929fcb3db11523e7 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Previous issue date: 2014-02-21 / Capes; CNPq / A lectina Concanavalina A (ConA), obtida das sementes de Canavalia ensiformis, é uma proteína que se liga especificamente à manose e glicose (e estruturas mais complexas, como receptores celulares, que contenham resíduos expostos desses carboidratos). Essa lectina possui várias atividades biológicas, incluindo a modulação de propriedades imunológicas e electrofisiológicos do cérebro. No presente estudo, foi caracterizada a ação da ConA sobre o fenômeno dependente da excitabilidade cerebral conhecido como Depressão Alastrante Cortical (DAC); estudou-se também a imunomarcação da microglia com um anticorpo policlonal contra a proteína Iba1 (Ionized calcium binding adaptor molecule 1). Ratos Wistar machos (n = 89) foram amamentados em condições favoráveis ou desfavoráveis de lactação, representadas respectivamente por ninhadas com 6-7 filhotes (grupo N6) ou 12-14 filhotes (grupos N12). Do 5º ao 24º dia pós-natal, foram tratados por via intraperitoneal com 1 mg/kg ou 10 mg/kg de ConA (grupos L1 e L10, respectivamente), ou com solução salina NaCl 0,9% (grupo Sal), ou sem nenhum tratamento (grupo Ingênuo). Aos 90-120 dias de idade, a DAC foi quimicamente induzida no córtex frontal e registrada em dois pontos da região parietal durante 4 h, e seus parâmetros de velocidade de propagação, amplitude e duração das ondas foram mensurados. Após o registro da DAC, os cérebros foram perfundidos e fixados, seguidos de secções cerebrais para reação com anticorpos anti-Iba1 para quantificação da imunomarcação da microglia. Os grupos N12 apresentaram maior velocidade de propagação da DAC que os grupos N6. Em ambas as condições de lactação, o tratamento sistêmico com a ConA resultou em diminuição significativa, dose-dependente, da velocidade de propagação da DAC (p < 0,05), em comparação aos grupos controle (Salina e Ingênuo). A amplitude e duração do componente negativo da variação lenta de voltagem da DAC permaneceram inalterados. Em um grupo adicional de ratos adultos, a aplicação tópica de ConA à superfície cortical também reduziu reversivelmente a velocidade de propagação da DAC. A imunorreatividade da microglia, avaliada no córtex e hipocampo, foi menor nos grupos tratados com ConA, em comparação com os respectivos controles. O hemisfério cerebral no qual a DAC foi repetidamente induzida apresentou maior imunorreatividade em comparação com o hemisfério oposto, principalmente no córtex. Podemos concluir que a atenuação na propagação da DAC e a diminuição da imunorreatividade na microglia em ratos adultos em decorrência da aplicação da lectina ConA (1 mg/kg e 10 mg/kg) durante o período de lactação (fase de grande plasticidade e neurogênese), indicam que essa lectina pode influenciar o desenvolvimento cerebral com ação protetora duradoura no cérebro de ratos. Essa ação não foi dificultada pela condição desfavorável de lactação. Desse modo, a lectina ConA apresentou-se como uma proteína com potencial ação protetora no cérebro, a julgar pelos efeitos sobre a DAC e sobre a reação microglial.

Sepse e suas complicações (sepse grave e choque séptico) ainda são a principal causa de morte nas unidades de terapia intensiva em todo o mundo. Estudos clínicos e experimentais têm demonstrado que na fase inicial da sepse a concentração plasmática de arginina vasopressina (AVP) está elevada. No entanto, durante o processo fisiopatológico os níveis plasmáticos da mesma permanecem inadequadamente baixos, apesar de haver hipotensão persistente. Uma das hipóteses sugeridas para essa deficiência relativa de AVP é a apoptose de neurônios vasopressinérgicos. Nosso objetivo foi identificar elementos envolvidos na morte celular hipotalâmica, além de avaliar o comportamento de células gliais e da barreira hematoencefálica (BHE) durante a sepse. Ratos Wistar foram submetidos à sepse por ligadura e punção cecal (CLP) ou não manipulados (naive) como controle e então divididos em dois grupos. No primeiro, foram perfundidos e os cérebros coletados para imunohistoquímica. Outro grupo foi decapitado para a retirada de sangue para dosagem de interferon- gama (IFN-?) e encéfalo para análise da expressão de proteínas no hipotálamo ou nos núcleos supraópticos (SON) e paraventriculares (PVN). Um terceiro foi separado para investigação da permeabilidade da BHE. Apesar de aumento da imunomarcação de CD8 e MHC-I no SON dos animais sépticos, não encontramos indícios de morte celular mediada por células imunes. No SON e PVN de animais sépticos, a expressão de fatores envolvidos na ativação da via extrínseca de apoptose (tBID, caspase-8 clivada) se manteve inalterada, enquanto fatores anti-apoptóticos relacionados à via intrínseca (BCL-2, BCL-xL) estavam diminuídos no hipotálamo. No SON destes animais a micróglia assumiu uma morfologia associada à sua ativação, concomitante com o aumento plasmático de IFN-?. Houve rompimento transitório da BHE no hipotálamo após 6 horas do CLP. Os resultados indicam que a via intrínseca de apoptose parece ser a responsável pela morte celular que é observada nos núcleos vasopressinérgicos e essa condição está temporalmente associada à ativação microglial e rompimento da BHE / Sepsis and its complications (severe sepsis and septic shock) remain as the main cause of death in intensive care units worldwide. Clinical and experimental studies have shown that in the early phase of sepsis the plasma concentration of arginine vasopressin (AVP) is increased. However, during the pathophysiological process the plasma levels remain inadequately low, despite of persistent hypotension. One of the hypothesis suggested for this relative deficiency is the apoptosis of vasopressinergic neurons. Our objective was to identify elements involved in the hypothalamic cellular death and evaluate the modifications of glial cells and blood-brain-barrier (BBB) during sepsis. Wistar rats were submitted to sepsis by cecal ligation and puncture (CLP) or non-manipulated (naïve), as control and then divided in two groups. In the first one, they were perfused and brains were collected for immunohistochemistry. In another one they were decapitated for blood collection and further plasma interferongama (IFN-?) analysis by ELISA. Brain was also collected for apoptosis-related proteins expression analysis in the hypothalamus or in the supraoptic (SON) and paraventricular (PVN) nuclei. A third set was separated for the investigation of BBB permeability. Despite of increased immunostaining for CD8 and MHC-I in the SON of septic animals, we did not find evidence of cell death mediated by immune cells. In the SON and PVN of septic animals, the expression of proteins involved in the activation of the extrinsic apoptosis pathway (tBID, cleaved caspase-8) was not altered, whereas anti-apoptotic factors related to the intrinsic pathway (BCL-2, BCLxL) were decreased. In the SON of these animals, microglia assumed a morphology related to its activation, associated with the increase of plasma IFN-?. There was a transitory breakdown of BBB in hypothalamus after 6 hours following CLP. The results indicate that the intrinsic apoptosis pathway seems to be responsible for the cell death observed in vasopressinergic nuclei and this condition is temporally associated with microglial activation and BBB leaking

Au cours de ma thèse, j'ai observé que la microglie est nécessaire à la maturation des circuits hippocampaux par la formation de boutons multi-synaptiques. J'ai également étudié la mécanique d'élimination des synapses, et observé que la microglie n'élimine pas directement les compartiments post-synaptiques. En revanche, elle contacte spécifiquement et rapidement certaines épines, en induisant un étirement de la tête de l'épine. Les petites épines sont préférentiellement contactées, et leur proximité avec les compartiments phagocytiques de la microglie suggère qu'elles pourraient être digérées sans être détachées du dendrite auquel elles appartiennent. Enfin, le système du complément n'est pas requis pour la reconnaissance ni les interactions entre microglie et épines, mais semble nécessaire à leur maturation. / During my thesis, I found that microglia is necessary for the maturation of hippocampalcircuits through the formation of multiple synapse boutons. I investigated how microgliacould mechanistically eliminate synapses, and found that microglia do not eliminate entirepost-synaptic spines but instead make fast and specific contacts that often result in spinehead stretching. Small, immature spines are preferentially targeted by microglia, and theirproximity to phagocytic compartment suggests that microglia could subtly erode themwithout to challenge their attachment to the dendritic shaft. Last, the complement system isnot necessary for recognition and interaction of microglia with spines, however seemsnecessary for proper maturation of post-synaptic spines.

No description available.

Le cerveau est très riche en acide docosahexaénoique (DHA, acide gras polyinsaturé (AGPI) n-3) et en acide arachidonique (AGPI n-6) qui sont de puissants immunomodulateurs. Ils pourraient être impliqués dans le contrôle de la neuroinflammation via leur conversion en dérivés lipidiques bioactifs. Dans ce contexte, notre objectif était de définir le rôle des médiateurs lipidiques dérivés des AGPI n-3 possédant des propriétés anti-inflammatoires et pro-résolutives dans la régulation de l’inflammation au niveau du cerveau. Nous avons d’abord caractérisé l’impact d’une modulation nutritionnelle en AGPI n-3 sur la composition lipidique cérébrale. Dans un modèle d’inflammation cérébrale, la consommation d’AGPI n-3 induit 1) une augmentation des métabolites lipidiques dérivés des AGPI n-3, 2) une diminution des métabolites lipidiques dérivés des AGPI n-6, et 3) une diminution de l’inflammation dans l’hippocampe. De plus, un apport en AGPI n-3 au cours de la période périnatale n’affecte pas la composition lipidique des cellules immunitaires du cerveau, les cellules microgliales. Ensuite, nous avons choisi une approche thérapeutique afin de démontrer in vitro dans un modèle de cellules microgliales que la RvD1, dérivée du DHA, en se fixant à son récepteur ALX/Fpr2, atténue l’inflammation via la régulation de la voie NFκB et de microARN. In vivo, l’injection i.c.v. de la RvD1 et du DHA atténue l’inflammation dans l’hippocampe par des processus différents. Les dérivés lipidiques bioactifs issus des AGPI n-3 pourraient être les médiateurs par lesquels les AGPI n-3 exercent leur effet bénéfique sur la régulation de l’inflammation au niveau du cerveau, la RvD1 étant fortement impliquée. / The brain is highly enriched in docosahexaenoic acid (DHA, n-3 polyunsaturated fatty acid, PUFA) and in arachidonic acid (n-6 PUFA) that are strong immunomodulators. They could be involved in the regulation of neuroinflammation through their conversion into bioactive lipid derivatives. Then, our objective was to define the role of n-3 PUFA derived lipid mediators that have anti-inflammatory and pro-resolutive properties in the regulation of brain inflammation. First, we characterized the impact of dietary n-3 PUFA modulation on brain lipid composition. In a central inflammatory model, n-3 PUFA intake induced 1) an increase in n-3 PUFA-derived lipid mediators, 2) a decrease in n-6 PUFA-derived lipid mediators, and 3) a decrease in inflammation in the hippocampus. Moreover, n-3 PUFA intake during the perinatal period did not affect lipid composition of brain immune microglial cells. Then, we chose a therapeutic approach to demonstrate in vitro in microglial cells that RvD1 derived from DHA, through the binding on its receptor ALX-Fpr2, attenuated inflammation via the regulation of NFκB pathway and microRNA expressions. In vivo, intracerebral injection of RvD1 and DHA reduced inflammation in the hippocampus by different pathways. Thus, the bioactive lipid derivatives from n-3 PUFA could be the mediators by which n-3 PUFA exert their beneficial effects on neuroinflammation, RvD1 playing a crucial role in this regulation.

Les cellules microgliales constituent la population de macrophages résidents du système nerveux central. De par leur appartenance au système immunitaire, elles furent longtemps considérées actives uniquement en conditions pathologiques. Au contraire, ces dernières décennies, elles sont apparues comme physiologiquement actives, notamment au cours de la période critique de formation du système nerveux central. Au cours du développement embryonnaire et postnatal, les neurones nouvellement générés migrent vers leur position définitive avant de développer leur arbre dendritique et axonal afin de former les connexions synaptiques à la base des réseaux nécessaires aux fonctions cérébrales. L'étude des microglies au cours de la période postnatale, a montré l'implication d'un mode de communication spécifique entre les neurones et la microglie, la voie Fractalkine/CX3CR1, dans la mise en place des cellules microgliales d'une part et dans le développement synaptique glutamatergique d'autre part. Cependant, l'importance de cette communication neurone-microglie pour le développement du système inhibiteur GABAergique est peu connue. Au cours de mon travail de thèse, je me suis intéressée au rôle de la voie de communication FractalKine/CX3CR1 dans la distribution des cellules microgliales et le développement postnatal du réseau GABAergique de l'Hippocampe. Nous avons ainsi montré que la suppression du récepteur microglial CX3CR1 induit une diminution du nombre de microglies dans la région CA3 de l'Hippocampe, dans une fenêtre temporelle précise entre 7 et 2 jours après la naissance. Cette diminution du nombre de microglies est corrélée avec une altération de l'activité de réseau au niveau de cette région. En effet, la fréquence des GDPs (Giant Depolarizing Potentials), une activité de réseau impliquée dans la formation et la maturation des synapses et spécifiquement générée en CA3, est diminuée à la fin de la première semaine postnatale. De plus, malgré l'absence de modification majeure de l'activité synaptique glutamatergique et GABAergique, les évènements postsynaptiques GABAergiques présentent une sous population d'évènements plus amples et des cinétiques légèrement plus rapides, pouvant suggérer une modification de la population d'interneurones mis en jeu. L'ensemble de mon travail de thèse met en évidence l'impact de la communication neurone-microglie par la voie Fractalkine/CX3CR1 sur le développement postnatal de l'Hippocampe Son absence affecte d'une part, la colonisation microgliale, et d'autre part, une activité de réseau caractéristique de l'Hippocampe, dans une fenêtre temporelle critique pour la mise en place des connexions synaptiques et la formation des réseaux neuronaux . / Microglial cells, the resident macrophages of the central nervous system, were mainly studied for their role in pathological conditions, but they recently appeared to be involved in synaptic development and circuits formation during postnatal period. During this critical period, microglial cells colonize the central nervous system and interact with other cell types, including neurons. A specific way of communication between neurons and microglia involves neuronal released fractalkine (CX3CL1) and its specific microglial receptor CX3CR1. CX3CR1 KO mice contributed to unclose microglial role during development. Indeed, CX3CR1 ablation alters microglia distribution in the brain, and it affects glutamatergic transmission and synapse maturation. However, these effects seem to be transient and brain region specific and their mechanisms are poorly understood. Furthermore, some effects observed in juvenile or adult mice may have origin during development, when neuronal connections are established. GABA plays a fundamental role in this process since it is excitatory The influence of neuron.microglia interaction on neuronal activity in the hippocampus during this period is poorly understood. In particular, nothing is known on GABAergic activity, known to be synaptogenic during this period My PhD project aimed at investigating how the signaling fractalkine pathway impacts microglial coloniation of the hippocampus and neuronal activity during the first two postnatal weeks. Our results indicate that in CX3XR1KO mice there is a reduction in the density of microglial cells at P7-P9 in the CA3 hippocampal area, accompanied at P7 by a significant reduction of frequency of Giant Depolarizing Potentials (GDPs), a network activity involved in hippocampal synapse formation and maturation Furthermore, despite no overall difference in glutamatergic or GABAergic synaptic activity, GABAergic events display a subpopulation of larger events, and the kinetics was slightly faster. Thus, the disruption of the specific neuronal.microglia signaling pathway on one hand impacts the microglia coloniation of the hippocampus and on the other hands affects specifically neuronal network activity during a time window critical for the establishment of neuronal connections.

La formation des circuits cérébraux, dont l’altération est associée à divers troubles neurologiques et psychiatriques, commence pendant l’embryogenèse. Les microglies sont les macrophages du cerveau et répondent à une inflammation ou une infection par une production de facteurs et une phagocytose active. En plus de leurs fonctions immunitaires, ces cellules ont été récemment impliquées dans l’émergence de troubles associés à des maladies neurodéveloppementales. Elles agissent également sur le remodelage post-natal des circuits neuronaux en phagocytant des synapses et des neurones immatures. Étant donné que les microglies colonisent précocement le cerveau embryonnaire, elles pourraient également participer au développement cérébral précoce. Mes travaux de thèse ont montré que les microglies embryonnaires, par le biais d’une localisation hétérogène, modulent la formation du cerveau antérieur. En effet, l’étude phénotypique comparative in vivo de modèles murins dans lesquels les microglies sont absentes ou leurs fonctions perturbées montrent des défauts de positionnement d’interneurones corticaux et de progression des axones dopaminergiques. De plus, les microglies sont impliquées dans la progression d’un autre faisceau axonal, la capsule externe. Après avoir observé une association très forte entre les microglies et des axones spécifiques, j’ai utilisé deux approches ex vivo et in vivo afin de déterminer si les microglies pourraient agir par phagocytose. Mes résultats montrent que malgré la présence de caractéristiques morphologiques phagocytaires, les microglies ne phagocytent pas les axones de manière intense. Ces travaux suggèrent ainsi que la phagocytose des axones à elle seule ne peut pas expliquer les rôles des microglies sur la croissance axonale. Ce travail, qui montre un rôle anténatal des microglies, révèle une interaction nouvelle entre le développement des systèmes nerveux et immunitaires, et ouvre des perspectives nouvelles pour l’étude sur les mécanismes contrôlant la formation du cerveau en condition physiologique et pathologique. / Brain functioning relies on complex neural circuits that are built during embryogenesis and defects in this process can lead to neurologic or psychiatric disorders. Microglia are the brain resident macrophages, which respond to inflammation and infection by active phagocytosis and secretion of various molecules. In addition to these immune-related functions, microglia were recently involved in the onset of several neuropsychiatric disorders, as well as in postnatal neuronal plasticity, notably through the phagocytosis of developing neurons and synapses. Since microglia colonize the brain during early embryogenesis, they might also participate to the elaboration of neural networks. Here, we reveal that embryonic microglia, via a heterogeneous localization, are modulators of forebrain wiring. Using a cross-comparative analysis of phenotypes induced by either an absence or a perturbation of microglia activity, we showed that microglia directly regulate interneuron positioning as well as dopaminergic axons outgrowth. Moreover, microglia are involved in the formation of another tract, the external capsule. Having observed a strong association between microglia and axons, we combined ex vivo and in vivo approaches in order to decipher whether microglia could act by means of phagocytosis. Despite showing some clear morphological features of intense phagocytosis, our results suggest that microglia do not perform active phagocytosis of axons in vivo. Thus, the phagocytic activity of embryonic microglia by itself is unlikely to explain their role on axonal progression. This study shows for the first time an early prenatal role for microglia and reveal a novel interplay between the central and nervous systems during embryonic development.

Alors que les synapses sont des structures relativement stables, les éléments qui la composent sont, eux, en permanent échange dans le temps et dans l'espace. Les composants des densités postsynaptiques sont renouvelés avec des cinétiques caractéristiques de chaque molécule et de chaque sous compartiment synaptique. La compatibilité entre le comportement dynamique des composants de la synapse et son maintien structural et fonctionnel à long terme implique une conception de ces assemblages multimoléculaires en équilibre dynamique. De nombreux paramètres peuvent influencer la dynamique des récepteurs aux neurotransmetteurs (RNT) dans les synapses, y compris l'activité synaptique et les protéines de la matrice extracellulaire. Cependant, le rôle des cellules gliales dans ce mécanisme est inconnu. Mon travail de thèse a porté sur l'exploration d'une possible contribution de la microglie, les cellules immunitaires du système nerveux central, à la stabilité des RNT et à l'efficacité des synapses inhibitrices de la moelle épinière. Mon travail de thèse démontre pour la première fois comment et en quoi la microglie est un partenaire clé de l'équilibre dynamique qui régit la structure et la fonction de la synapse inhibitrice dans la moelle. Par conséquent, il donne un éclairage nouveau sur la façon de concevoir l'efficacité synaptique et sa régulation de façon non neurone autonome. / Whereas synapses are relatively stable structures, their molecular constituents are continuously recycled and exchanged in time and space. Each of the molecules that contribute to build synaptic structures is renewed with specific kinetics, depending on their organisation in the postsynaptic densities. The compatibility between a dynamic behaviour and a long-term maintenance of synapses implies to think synapses as multi-molecular assemblies in a dynamic equilibrium. Several parameters can influence the dynamics of receptors to neurotransmitters(RNT) at synaptic sites, including neuronal activity and extracellular matrix proteins. However,the role of glial cells in this mechanism is unknown. During my thesis work, I explored the roleof microglia, the resident immune cells of the central nervous system, on the lateral diffusion ofRNT and synaptic efficacy at spinal cord inhibitory postsynaptic densities. My work demonstrates for the first time a partnership between microglia and synapses. It shows that immune cells can take part to the regulation of synaptic strength very rapidly but also at basal state, by regulating RNT dynamics. Furthermore it identifies microglia as a key partner for a heterocellular stabilization of synaptic receptors. This work raises the intriguing possibility that the general regulation of network activity may also be explained by a fine modulation of receptors stability at the synapse controlled by microglia.