No description available.

La maladie d’Alzheimer (MA) est la forme la plus commune de démence chez les personnes âgées. La maladie se caractérise par des pertes de fonctions cognitives et plusieurs études ont montré une étroite relation entre la neuroinflammation, les déficits synaptiques et la perte des fonctions cognitives dans la MA. L'importance de la composante neuroinflammatoire a été démontrée essentiellement grâce à des données épidémiologiques rapportant une prévalence diminuée de 40 à 70% chez des patients traités chroniquement par des anti-inflammatoires non stéroïdiens (AINS) pour d'autres pathologies. Les AINS sont des inhibiteurs des enzymes de synthèse des prostaglandines. Les prostaglandines sont des métabolites de l’acide arachidonique. Parmi ces prostaglandines, la PGE2 est connue pour moduler la transmission et les plasticités synaptiques dans l’hippocampe et son expression est fortement augmentée dans la maladie d’Alzheimer. De nombreux travaux rapportent l'existence de déficits synaptiques dans la MA, notamment dans l'hippocampe, siège de la mémoire et de l’apprentissage. Ces travaux se sont focalisés sur les déficits postsynaptiques à la synapse archétypique formée entre les cellules pyramidales CA3 et CA1. A l'inverse, la synapse formée entre les fibres moussues et les cellules pyramidales CA3 (FM-CA3) exprime des plasticités présynaptiques atypiques, à court et à long terme, indépendantes des récepteurs NMDA et il n'existe que très peu d'études concernant ces plasticités dans le contexte de MA. L’objectif de cette étude a été de montrer l’implication de PGE2 dans les déficits synaptiques à la synapse FM-CA3 dans un modèle murin de la MA, la souris double transgénique APPswe/PS1ΔE9 (APP/PS1). Nos résultats montrent que l’application exogène de PGE2 chez des souris sauvages entraîne un déficit de plasticité uniquement sur la potentialisation à long terme (PLT) exprimée présynaptiquement via l’activation spécifique du récepteur EP3. Nous montrons aussi que dans la souris APP/PS1, seule cette PLT présynaptique est impactée à partir de 12 mois. Enfin, ce déficit de la PLT présynaptique pour la souris APP/PS1 est réversé par un inhibiteur spécifique des récepteurs EP3 montrant ainsi un rôle clé pour la signalisation PGE2 - EP3 dans les déficits synaptiques hippocampaux de ce modèle murin de la maladie d’Alzheimer. / Alzheimer’s disease (AD) is the most common form of dementia in elder people characterized by a loss of cognitive function linked to synaptic deficits. There is considerable evidence that neuroinflammation and AD are intimately linked. The key role of neuroinflammation in the course of the disease was figured out by epidemiological studies reporting a reduced prevalence to develop AD for patients chronically treated with Non-Steroidal Anti-Inflammatory Drugs (NSAIDs). Prostaglandins are lipidic mediators derived from arachidonic acid and their synthesis is inhibited by NSAIDs. Among prostaglandins, PGE2 is known to modulate synaptic transmission and plasticity in the hippocampus and its concentration is higher in brains from AD patients. Numerous studies have reported synaptic deficits in the course of AD, mainly in the hippocampus which is essential for cognitive functions like learning or memory formation. The vast majority of these studies were focused on postsynaptic deficits at the canonical CA3-CA1 synapse. On the opposite, the synapse between mossy fiber and CA3 pyramidal cell (Mf-CA3) that express presynaptic short-term and long-term plasticity, was poorly studied in the context of AD. The aim of my project was to decipher the involvement of PGE2 in synaptic deficits in a mouse model of AD, the APPswe/PS1ΔE9 (APP/PS1). Our results show that acute application of PGE2 on wild type young mice impairs only presynaptic long term potentiation (LTP) at the Mf-CA3 synapse via the specific activation of EP3 receptor. In APP/PS1 mice, we demonstrate that the sole deficit at the Mf-CA3 synapse is an impairment of the presynaptic LTP at 12 months of age. Finally we demonstrate that the impaired presynaptic LTP in APP/PS1 mice can be rescued by the acute application of a specific EP3 receptor antagonist, pointing out the key role of PGE2 - EP3 signaling pathway in synaptic deficits in hippocampus in a mouse model of AD.

No description available.

Le risque de chute est une problématique bien présente chez les personnes âgées ou ayant une atteinte neurologique et reflète un déficit des mécanismes neuronaux assurant l’équilibre. De précédentes études démontrent que l’intégration des informations sensorielles est essentielle au contrôle de l’équilibre et que l’inhibition présynaptique (IP) serait un mécanisme important dans le contrôle de la transmission sensorielle. Ainsi, le but de cette étude était d’identifier la contribution du mécanisme d’IP à l’induction de réponses posturales efficaces suite à une perturbation d’équilibre. Notre hypothèse est qu’une diminution d’IP contribuerait à l’induction des ces réponses, en augmentant l’influence de la rétroaction sensorielle sur les réseaux de neurones spinaux. Afin de démontrer cette hypothèse, nous avons d’abord évalué l’excitabilité spinale pendant les perturbations vers l’avant ou vers l’arrière, à l’aide du réflexe H. L’excitabilité spinale était modulée selon la direction de la perturbation et cette modulation survenait dès 75 ou 100 ms (p<0.05), soit avant l’induction des réactions posturales. Puis, à l’aide de techniques plus précises de convergence spinale, nous avons démontré que l’IP était diminuée dès 75 et 100 ms dans les deux directions, suggérant que la transmission des informations sensorielles vers la moelle épinière est accrue juste avant le déclenchement de la réponse posturale. Cette étude met en évidence un mécanisme-clé permettant d’augmenter la rétroaction des informations sensorielles nécessaires à l’induction de réponses posturales appropriées. L’évaluation de ce mécanisme pourrait mener à une meilleure identification des individus à risque de chute. / Falls are a significant problem among the elderly or persons with a neurological impairment, and reflect a deficit in the nervous mechanisms underlying postural control. Previous research shows that the integration of sensory feedback is a crucial component of postural control and that presynaptic inhibition (PSI) plays an important role in controlling the sensory processing of information. The aim of this study was to identify the contribution of PSI to the induction of effective postural responses following an unexpected balance perturbation. We hypothesized that a decrease in PSI would contribute to the induction of these responses by increasing the influence of sensory feedback onto spinal networks during the perturbation. First we assessed the level of spinal excitability during perturbations, using the soleus H-reflex. Results show that spinal excitability is modulated according to the direction of the perturbation (forward and backward tilts) and that this modulation occurs 75 and 100 ms after tilt-onset in all subjects (p<0.05). To further estimate changes in PSI, spatial facilitation techniques were used. PSI was shown to decrease in both perturbation directions shortly after tilt onset at 75 and 100 ms (p<0.05), suggesting an increase in sensory transmission in the spinal cord. These observations suggest that sensory feedback is critical for the induction of effective postural responses and that impaired sensory transmission or integration, due to CNS lesions or ageing, may lead to certain balance deficits. Identifying patients with such impairments may improve fall risk-assessment and prevention.

Introdução:As adaptações neurais ao treinamento físico vêm sendo amplamente estudadas e a medula espinhal é um dos locais de possível adaptação. No entanto nenhuma avaliação longitudinal havia sido feita diretamente sobre as circuitarias inibitórias medulares. Até o presente momento as alterações eram somente suposições. O presente trabalho verificou as circuitarias medulares responsáveis pela inibição recíproca (IR) e inibição pré-sináptica (IPS) em sujeitos submetidos a diferentes treinamentos. Materiais e Métodos: Para o treino aeróbico (resistência) foram avaliados 25 soldados submetidos ao treinamento militar do Exército Militar Brasileiro. Foram feitas 3 avaliações uma pré-treino e outras duas com 3 e 9 meses após o inicio das atividades no ano de 2006. Outros 29 sujeitos foram divididos em 3 grupos: controle (permaneceram 8 semanas sem atividades de reinamento), grupo de treino de força máxima e treino de potência. Eles foram submetidos a 8 semanas de treino, realizado com séries de agachamento livre com peso. Para avaliação das circuitarias medulares foi utilizado o reflexo H do sóleus condicionado com estímulos no nervo fibular comum (NFC) - que inerva o músculo tibial anterior (TA). O intervalo entre o estímulo condicionante e o estímulo teste determinou a avaliação da IR, da inibição D1 e da inibição D2 (IPS). Outras variáveis também foram calculadas como: contração voluntária máxima isométrica (CVM) do sóleus e TA e seus respectivos eletromiogramas (EMG), relação elétrica e mecânica entre Hmax/Mmax e condicionamento do EMG do sóleus por estímulos no NFC. Foram feitas análises pareadas com teste t-student para o grupo militar e ANOVA two-way para comparação dos grupos de força máxima e potência com o grupo controle. Principais Resultados: O grupo do exército apresentou aumento na força do sóleus e do TA, juntamente com aumento no RMS do EMG do sóleus e do torque gerado pela onda Mmax, sem alterações nos relações Hmax/Mmax. O treinamento militar reduziu significativamente a inibição D1 e mostrou tendências a aumento da IPS. O grupo de força máxima não mostrou aumento de força isométrica, no entanto apresentou aumento na relação elétrica Hmax/Mmax, com concomitante redução da IR e aumento da IPS. O grupo de potência mostrou ganho na força máxima isométrica somente do sóleus. A capacidade de gerar torque reflexamente também aumentou neste grupo, com aumento significativo na relação mecânica Hmax/Mmax. Esta melhora na utilização do arco reflexo também foi verificada com redução da IPS e aumento da IR neste grupo.Conclusões: Estes resultados mostraram que a medula espinal sofre plasticidade nas vias inibitórias IR, inibição D1 e D2, e que esta plasticidade é dependente do tipo de tarefa realizada. / Introduction: Neural adaptations with physical training have been widely studied. The spinal cord is a possible locus of adaptation. However, longitudinal studies that evaluate directly the spinal cord pathways have not been found in the literature. Therefore, all reports from the literature justify changes found in measured responses to exercise by hypotheses on spinal cord mechanisms. This study had the objective of measuring features of specific spinal cord pathways to check if they change according to the type of physical training. The pathways related to reciprocal inhibition (RI) and pre-synaptic inhibition (PSI) were investigated in subjects undergoing different trainings. Materials and Methods: For endurance training 25 soldiers were subjected to military training of the Brazilian Army. Evaluations were made three times, one previous to the beginning of the activity and twice post-training (within 3 and 9 months). Other 29 subjects were divided into: control group (with no training), maximal strength group and power group. They were subjected to 8 weeks of training with series of squat movements. The soleus H reflex conditioning with stimuli in the common peroneal nerve (CPN) was used to evaluate the spinal cord pathways. The interval between the conditioning and the test stimulus determine the assessment of RI, D1 inhibition and D2 inhibition (PSI). Other variables were also calculated: maximum voluntary isometric contraction from soleus and tibialis anterior and their electromyograms (EMG), electrical and mechanical Hmax/Mmax ratio and 3 inhibitions over the soleus EMG conditioned by stimuli to the CPN. The results were analyzed with paired t-student test for the military group and with two-way ANOVA to compare the maximal strength and power groups with the control group. Main Results: The military group had increased strength of the soleus and the TA muscles, with an increase in the RMS of the soleus EMG. This group also increased the torque generated by the Mmax-wave, without changes in Hmax/Mmax ratio. The military training significantly reduced D1 inhibition and showed tendencies to increase the PSI. The maximal strength group showed no differences in isometric strength, but had increased Hmax/Mmax ratio with concomitant reduction of RI and increased PSI. The power group increased isometric strength only for the soleus muscle. This group also improved the ability to generate torque by reflex pathways, with significant increase in the mechanical Hmax/Mmax ratio, with a reduction of PSI and increase of RI. Conclusions: These results show that spinal cord plasticity occurs in the inhibitory pathways of reciprocal inhibition, D1 inhibition and D2 inhibition (pre-synaptic inhibition), and that plasticity is dependent on the type of trained movement.

Zentralnervöse Lernvorgänge hängen wesentlich von der Plastizität der synaptischen Übertragung ab. Synapsen verändern ihre Effizienz sowohl durch Änderungen in der Freisetzungswahrscheinlichkeit von Neurotransmittern als auch durch Variabilität der postsynaptischen Rezeptorausstattung. Darüberhinaus gibt es seit einiger Zeit Hinweise, dass auch die Konzentration des Neurotransmitters in synaptischen Terminalen variieren kann und dadurch die Freisetzungswahrscheinlichkeit von Vesikeln verändert wird. Um den Zusammenhang zwischen der Füllung synaptischer Vesikel und ihrer Dynamik im Vesikelzyklus besser zu verstehen, habe ich zunächst ein Computermodell entwickelt. Ich habe gefunden, dass nur eine Modifikation des Nachschubs von Vesikeln aus der Reservepopulation in die Population unmittelbar freisetzbarer Vesikel die experimentell gemessene Abhängigkeit der Freisetzungswahrscheinlichkeit von der präsynaptischen Transmitterkonzentration reproduzieren kann, nicht jedoch ein direkter Effekt auf die Freisetzungsrate. Einer der im Modell simulierten Mechanismen für den beobachteten Effekt des vesikulären Füllungszustandes auf die Vesikelfreisetzung besteht in einer Rückwirkung des freigesetzten Transmitters auf ionotrope Autorezeptoren. Diesen Mechanismus habe ich anschliessend an GABAergen Synapsen in der CA3-Region des Hippocampus untersucht. Tatsächlich habe ich durch patch-clamp Messungen herausgefunden, dass sowohl die Antwortwahrscheinlichkeit auf extrazelluläre Stimulation als auch die Frequenz von spontanen Vesikelfreisetzungen nach Applikation des GABAA Rezeptor-Agonisten Muscimol signifikant verringert ist. Diese Befunde weisen darauf hin, dass GABA seine eigene Freisetzung über ionotrope Autorezeptoren hemmt. Direkte Messungen der Vesikelfreisetzungen sind an GABAergen Synapsen nur durch bildgebende Verfahren möglich. Mit Hilfe von Zwei-Photonen Mikroskopie ist es mir erstmalig gelungen, im Hirnschnitt die spontane Fusion von Vesikeln zu untersuchen. Diese Methode könnte für eine Reihe von Fragestellungen relevant sein. Dabei hat sich gezeigt, dass nach Applikation von Muscimol die spontane Abnahme der Fluoreszenz vorher angefärbter synaptischer Vesikel ausschliesslich in perisomatischen (überwiegend GABAergen) synaptischen Boutons signifikant verringert ist. Zusammenfassend habe ich bei der experimentellen Prüfung von Vorhersagen eines Computermodells durch patch-clamp Untersuchungen und durch eine neue Methode mit funktioneller Bildgebung Hinweise auf funktionelle präsynaptische GABAA Rezeptoren an GABAergen Terminalen der CA3-Region des Hippocampus gefunden, die eine negative Rückwirkung auf die Freisetzung von Vesikeln ausüben. / Learning processes depend on synaptic plasticity. Synaptic efficacy depends both on the probability of transmitter release and on the variability of postsynaptic receptors. Furthermore, recent data suggest that the transmitter concentration in presynaptic terminals may be an additional variable influencing the release probability of synaptic vesicles. To better understand the relationship between vesicular filling and vesicular dynamics, I first developed a computer model. I found that only a modification of the replenishment of the readily releasable pool from a reserve pool reproduces the experimentally observed dependence of vesicular release on the presynaptic transmitter concentration, but not a direct effect on the release rate. One of the simulated mechanisms consists in a feedback of released transmitter on ionotropic autoreceptors. I investigated this mechanism in the CA3 region of the hippocampus. Indeed, using patch-clamp recordings I observed that application of the GABAA receptor agonist muscimol decreases significantly the response probability to extracellular stimulation as well as the frequency of spontaneous transmitter release. These results suggest that GABA inhibits its own release via ionotropic autoreceptors. Direct measurements of GABAergic vesicular release are only possible by imaging techniques. Using two-photon microscopy, I was (to my knowledge) the first one to investigate spontaneous vesicle fusion in brain slices. This method could be relevant to a variety of topics. I found out that application of muscimol leads to a significant decrease of fusion-associated fluorescence decay in perisomatic (mostly GABAergic) synaptic boutons. Taken together, I used patch-clamp recordings and a new application of two-photon microscopy to verify predictions of a computer model. The results suggest a negative feedback of GABA release via presynaptic GABAA receptors in the CA3 region of the hippocampus.

Introdução:As adaptações neurais ao treinamento físico vêm sendo amplamente estudadas e a medula espinhal é um dos locais de possível adaptação. No entanto nenhuma avaliação longitudinal havia sido feita diretamente sobre as circuitarias inibitórias medulares. Até o presente momento as alterações eram somente suposições. O presente trabalho verificou as circuitarias medulares responsáveis pela inibição recíproca (IR) e inibição pré-sináptica (IPS) em sujeitos submetidos a diferentes treinamentos. Materiais e Métodos: Para o treino aeróbico (resistência) foram avaliados 25 soldados submetidos ao treinamento militar do Exército Militar Brasileiro. Foram feitas 3 avaliações uma pré-treino e outras duas com 3 e 9 meses após o inicio das atividades no ano de 2006. Outros 29 sujeitos foram divididos em 3 grupos: controle (permaneceram 8 semanas sem atividades de reinamento), grupo de treino de força máxima e treino de potência. Eles foram submetidos a 8 semanas de treino, realizado com séries de agachamento livre com peso. Para avaliação das circuitarias medulares foi utilizado o reflexo H do sóleus condicionado com estímulos no nervo fibular comum (NFC) - que inerva o músculo tibial anterior (TA). O intervalo entre o estímulo condicionante e o estímulo teste determinou a avaliação da IR, da inibição D1 e da inibição D2 (IPS). Outras variáveis também foram calculadas como: contração voluntária máxima isométrica (CVM) do sóleus e TA e seus respectivos eletromiogramas (EMG), relação elétrica e mecânica entre Hmax/Mmax e condicionamento do EMG do sóleus por estímulos no NFC. Foram feitas análises pareadas com teste t-student para o grupo militar e ANOVA two-way para comparação dos grupos de força máxima e potência com o grupo controle. Principais Resultados: O grupo do exército apresentou aumento na força do sóleus e do TA, juntamente com aumento no RMS do EMG do sóleus e do torque gerado pela onda Mmax, sem alterações nos relações Hmax/Mmax. O treinamento militar reduziu significativamente a inibição D1 e mostrou tendências a aumento da IPS. O grupo de força máxima não mostrou aumento de força isométrica, no entanto apresentou aumento na relação elétrica Hmax/Mmax, com concomitante redução da IR e aumento da IPS. O grupo de potência mostrou ganho na força máxima isométrica somente do sóleus. A capacidade de gerar torque reflexamente também aumentou neste grupo, com aumento significativo na relação mecânica Hmax/Mmax. Esta melhora na utilização do arco reflexo também foi verificada com redução da IPS e aumento da IR neste grupo.Conclusões: Estes resultados mostraram que a medula espinal sofre plasticidade nas vias inibitórias IR, inibição D1 e D2, e que esta plasticidade é dependente do tipo de tarefa realizada. / Introduction: Neural adaptations with physical training have been widely studied. The spinal cord is a possible locus of adaptation. However, longitudinal studies that evaluate directly the spinal cord pathways have not been found in the literature. Therefore, all reports from the literature justify changes found in measured responses to exercise by hypotheses on spinal cord mechanisms. This study had the objective of measuring features of specific spinal cord pathways to check if they change according to the type of physical training. The pathways related to reciprocal inhibition (RI) and pre-synaptic inhibition (PSI) were investigated in subjects undergoing different trainings. Materials and Methods: For endurance training 25 soldiers were subjected to military training of the Brazilian Army. Evaluations were made three times, one previous to the beginning of the activity and twice post-training (within 3 and 9 months). Other 29 subjects were divided into: control group (with no training), maximal strength group and power group. They were subjected to 8 weeks of training with series of squat movements. The soleus H reflex conditioning with stimuli in the common peroneal nerve (CPN) was used to evaluate the spinal cord pathways. The interval between the conditioning and the test stimulus determine the assessment of RI, D1 inhibition and D2 inhibition (PSI). Other variables were also calculated: maximum voluntary isometric contraction from soleus and tibialis anterior and their electromyograms (EMG), electrical and mechanical Hmax/Mmax ratio and 3 inhibitions over the soleus EMG conditioned by stimuli to the CPN. The results were analyzed with paired t-student test for the military group and with two-way ANOVA to compare the maximal strength and power groups with the control group. Main Results: The military group had increased strength of the soleus and the TA muscles, with an increase in the RMS of the soleus EMG. This group also increased the torque generated by the Mmax-wave, without changes in Hmax/Mmax ratio. The military training significantly reduced D1 inhibition and showed tendencies to increase the PSI. The maximal strength group showed no differences in isometric strength, but had increased Hmax/Mmax ratio with concomitant reduction of RI and increased PSI. The power group increased isometric strength only for the soleus muscle. This group also improved the ability to generate torque by reflex pathways, with significant increase in the mechanical Hmax/Mmax ratio, with a reduction of PSI and increase of RI. Conclusions: These results show that spinal cord plasticity occurs in the inhibitory pathways of reciprocal inhibition, D1 inhibition and D2 inhibition (pre-synaptic inhibition), and that plasticity is dependent on the type of trained movement.

The central nervous system is largely responsible for receiving sensory information from the environment and determining motor output. Yet, centrally-derived behavior and sensation depends on the optimal maintenance of the cells, tissues, and organs that feed and support these functions. Most of visceral regulation occurs without conscious oversight, making the spinal cord a key site for integration and control. How the spinal cord modulates output to our organs, or sensory information from them, is poorly understood. The overall aim of this dissertation was to better understand spinal processing of both visceral sensory information to and sympathetic output from the spinal cord. I first established and validated a HB9-GFP transgenic mouse model that unambiguously identified sympathetic preganglionic neurons (SPNs), the spinal output neurons for the sympathetic nervous system. Using this model, I investigated the electrophysiological similarities and diversity of SPNs, and compared their active and passive membrane properties to those in other animal models. My results indicate that while many of the same characteristics are shared, SPNs are a heterogeneous group that can be differentiated based on their electrophysiological properties. Since descending monoaminergic pathways have particularly dense projections to sympathetic regions of the spinal cord, I next examined the modulatory role that the monoamines have on spinal sympathetic output. While each neuromodulator tested had a unique signature of action, serotonin and norepinephrine appeared to increase the excitability of individual SPNs, while dopamine had more mixed actions. Since many autonomic reflexes are integrated by the spinal cord, I also questioned whether these reflexes would be similarly modulated. I therefore developed a novel in vitro spinal cord and sympathetic chain preparation, which allowed for the investigation of visceral afferent-mediated reflexes and their neuromodulation by monoamines. This preparation exposed a dichotomy of action, where sympathetic and somatic motor output is generally enhanced by the monoamines, but reflexes mediated by visceral input are depressed. Utilizing the spinal cord and sympathetic chain preparation, I also investigated how the spinal cord modulates visceral sensory information. One of the most powerful means of selectively inhibiting afferent information from reaching the spinal cord is presynaptic inhibition. I hypothesized that both spinal visceral afferents and descending monoaminergic systems would depress transmission of visceral afferents to the spinal cord. My results demonstrated that activity in spinal visceral afferents can lead to spinally generated presynaptic inhibition, and that in addition to depressing synaptic transmission to the spinal cord, the monoamines also depress the intrinsic circuitry that generates this activity-dependent presynaptic inhibition. Taken together, my results indicate that descending monoaminergic pathways act to limit the amount of visceral sensory information reaching the central nervous system and increase sympathetic output, resulting in an uncoupling of output from visceral sensory input and transitioning to a feed-forward, sympathetically dominant control strategy. This combination offers complex modulatory strategies for descending systems.

No description available.

Emerging evidence in both vertebrates and invertebrates is redefining glia as active players in the development and integrity of the nervous system. The formation of functional neuronal circuits requires the precise addition of new synapses. Mounting evidence implicates glial function in synapse remodeling and formation. However, the precise molecular mechanisms governing these functions are poorly understood. My thesis work begins to define the molecular mechanisms by which glia communicate with neurons at the Drosophila neuromuscular junction (NMJ). During development glia play a critical role in remodeling neuronal circuits in the CNS. In order to understand how glia remodel synapses, I manipulated a key component of the glial engulfment machinery, Draper. I found that during normal NMJ growth presynaptic boutons constantly shed membranes or debris. However, a loss of Draper resulted in an accumulation of debris and ghost boutons, which inhibited synaptic growth. I found that glia use the Draper pathway to engulf these excess membranes to sculpt synapses. Surprisingly, I found that muscle cells function as phagocytic cells as well by eliminating immature synaptic ghost boutons. This demonstrates that the combined efforts of glia and muscle are required for the addition of synapses and proper growth. My work establishes that glia play a crucial role in synapse development at the NMJ and suggests that there are other glial-derived molecules that regulate synapse function. I identified one glial derived molecule critical for the development of the NMJ, a TGF-β ligand called Maverick. Presynaptically, Maverick regulates the activation of BMP pathway confirmed by reducing the transcription of the known target gene Trio. Postsynaptically, it regulates the transcription of Glass bottom boat (Gbb) in the muscle suggesting that glia modulate the function of Gbb and consequently the activation of the BMP retrograde pathway at NMJ. Surprisingly, I also found that glial Maverick regulates the transcription of Shaker potassium channel, suggesting that glia potentially could regulate muscle excitability and consequently modulate synaptic transmission. Future work will elucidate such hypothesis. My work has demonstrated two novel roles for glia at the NMJ. First is that glia engulfing activity is important for proper synaptic growth. Second is that the secretion of glial-derived molecules are required to orchestrate synaptic development. This further supports that glia are critical active players in maintaining a functional nervous system.