The development of several organ systems through modeling and shaping of the tissue structure occurs from signaling through axon guidance molecules. The Slit family of ligands has been shown to regulate branching morphogenesis in mammary gland duct development and loss of Slit gene expression during this time leads to the formation of hyperplastic, disorganized lesions suggesting a potential role for Slits in cancer formation. Characterization of human pancreatic ductal adenocarcinoma cell lines showed a loss of SLIT2 expression in cells that contain activated Kras. Loss of SLIT2 expression was associated with DNA methylation of CpG sites within the SLIT2 core promoter and chromatin enrichment of repressive histone modifications at the SLIT2 transcriptional start site. Additionally, treatment of pancreatic ductal adenocarcinoma cell lines with demethylating agent 5-aza-2'-deoxycytidine led to SLIT2 re-expression while treatment with histone deacetylase inhibitor Trichostatin A did not. Mir-218-1 is an intronic microRNA encoded within intron 15 of the SLIT2 gene. Expression of mir-218-1 does not correlate with SLIT2 mRNA expression suggesting that it is transcribed from a promoter independent of the SLIT2 gene promoter. Pancreatic ductal adenocarcinoma cell lines showed a peak of H3K4me3 chromatin enrichment localized to a 1kb region within intron 4 of the SLIT2 gene denoting a candidate alternative promoter for mir-218-1. A concordant peak of H4ac chromatin enrichment overlapped the peak of H3K4me3 enrichment and transcriptional activity was measured from the 1kb region in all pancreatic ductal adenocarcinoma cell lines. A NF-κB binding site was also predicted to exist within the 1kb region. Transfection with two independent siRNAs to NF-κB led to an increase in both pre-mir-218-1 and mature mir-218-1 while treatment with an inhibitor to IκB kinase led to an increase in pre-mir-218-1 expression. Additionally, the p65 subunit of NF-κB was found to bind to the candidate mir-218-1 alternative promoter in pancreatic ductal adenocarcinoma cell lines that do not contain DNA CpG methylation at the predicted NF-κB binding site. It was discovered that miR-218 is a modulator of ARF6 expression suggesting a role in the inhibition of pancreatic ductal adenocarcinoma cell invasion through modulation of the actin cytoskeleton. Overexpression with a miR-218 precursor showed that miR-218 is an inhibitor of pancreatic ductal adenocarcinoma cell invasion in two dimensions. Additionally, it was found that while miR-218 does not have an affect on the ability of pancreatic ductal adenocarcinoma cells to form functional invadopodia, miR-218 is an inhibitor of the extracellular matrix degradation properties of mature invadopodia. Interestingly, the effect of miR-218 on pancreatic ductal adenocarcinoma cell invasion or extracellular matrix degradation is not reliant on the cell's dependency on Kras signaling for growth and survival. Collectively, these observations indicate that understanding the transcriptional regulation of SLIT2 and mir-218-1 expression as well as their signaling properties may provide a step toward the development of diagnostic tests and therapeutic treatments for patients with invasive or metastatic pancreatic ductal adenocarcinoma.

Le guidage axonal est un processus très important dans le développement du cerveau, permettant de lui donner sa structure et son organisation. La communauté scientifique des neurosciences lui porte un intérêt grandissant ces dernières années. Plusieurs outils appartenant au domaine des microtechnologies, que sont la microfluidique et le micropatterning, sont d’une aide importante pour étudier le guidage axonal in vitro. Ils permettent de confiner les neurones et leurs axones et de leur appliquer des gradients de molécules de guidage. Lors de ce travail de thèse, j’ai voulu développer un système pour étudier l’effet de gradients de molécules de guidage sur le guidage axonal. J’ai pour cela testé plusieurs configurations de dispositifs microfluidiques, de micromotifs (micropatterns) et de combinaisons de ces derniers.Nous avons d’abord utilisé deux approches pour isoler les axones de neurones dissociés de leurs somas afin de pouvoir étudier, à haut débit, l’effet de l’environnement moléculaire sur les cônes de croissance des neurones. La première approche consistait à faire pousser des neurones sur des motifs (patterns) de différentes protéines. Elle a permis de montrer leur capacité d’adhésion spécifique sur ces motifs. La seconde consistait à ensemencer des neurones dans un dispositif microfluidique dans lequel, lors de leur pousse, les axones sont séparés des somas par des microcanaux. Nous avons ensuite étudié l’effet, sur les axones, de gradients de molécules de guidage. Pour commencer, nous avons mesuré l’effet de deux molécules de guidage : l’éphrine et la sémaphorine, en cultivant des neurones en présence de gradients patternés de ces deux molécules. Par la suite, nous avons étudié un autre modèle où les neurones sont plus proches de leur environnement in vivo, des explants poussant sur des motifs de laminine contenant un gradient. Pour aider au positionnement de l’explant, nous avons polymérisé des hydrogels. Ensuite, nous avons mis des explants à côté de gradients patternés d’éphrine. Enfin, nous avons cherché à obtenir un gradient soluble de molécules de guidage entretenu sur des temps longs, plus proche des gradients existant in vivo. Dans ce but, nous avons voulu fabriquer un dispositif microfluidique permettant d’appliquer un gradient soluble de molécules de guidage sur des neurones. Pour obtenir un gradient stable dans le temps, nous avons aussi cultivé des neurones à côté de cellules exprimant la nétrine, une autre molécule de guidage. Pour finir, nous avons cultivé des neurones et des glies dissociés pour étudier leurs interactions.L’ensemble de ces recherches n’a pas permis d’obtenir un dispositif fiable pour étudier l’effet de molécules sur la pousse et le guidage des axones. Néanmoins, la configuration consistant en une coculture de neurones à proximité de cellules relargant de la nétrine nous a permis d’obtenir des premiers résultats encourageants. Nous avons ainsi mis au point un ensemble de méthodes qui pourront nous permettre de finaliser le développement d’un système pour étudier le guidage axonal, fonctionnel et efficace. / Axonal guidance is a very important process during brain development, allowing to give it its structure and organization. The neuroscience scientific community has a growing interest in it during the last years. Several tools belonging to the field of microtechnologies, microfluidics and micropatterning are of important help to study axonal guidance in vitro. They allow to confine neurons and their axons and to apply gradients of guidance molecules. During this thesis, my goal was to develop a system to study the effect of guidance molecules gradients on axonal guidance. For that, I tested several configurations of microfluidic devices, micropatterns and combinations of both.First, we used two approaches to isolate dissociated neurons axons from their somas. Our goal was to study the effect of the molecular environment on neurons growth cones, with a high throughput. The first approach consisted in growing neurons on different proteins patterns. It also allowed to show their capacity to adhere on these patterns. The second one consisted in seeding neurons in a microfluidic device in which, during their growth, axons are separated from somas by microchannels. Then we studied the effect, on the axons, of guidance molecules gradients. To begin, we measured the effect of two guidance molecules: ephrin and semaphorin, by culturing neurons in the presence of patterned gradients of these two molecules. After that, we studied another model where neurons are closer from their environment in vivo, explants growing on laminin patterns containing a gradient. To help the explant positioning, we polymerized hydrogels. Then, we put explants next to patterned gradients of ephrin. Finally, we tried to obtain a soluble gradient of guidance molecules, over a long period of time (days), closer to existing gradients in vivo. In that goal, we wanted to build a microfluidic device enabling the application of a soluble gradient of guidance molecules on neurons. To obtain a constant gradient, we also cultured neurons next to cells expressing netrin, another guidance molecule. Finally, we cultured dissociated neurons and glial cells to study their interactions.All these experiments did not allow to obtain a reliable device to study the effect of molecules on axons growth and guidance. Nevertheless, the configuration consisting in a coculture of neurons next to cells releasing netrin allows us to obtain promising preliminary results. We thus drew up a group of methods that will enable us to finalize the development of a system to study axonal guidance, functional and efficient.

Die Semaphorine Sema3A und Sema3C sowie Netrin-1 und deren Rezeptoren, die Neuropiline und DCC wurden in der vorliegenden Arbeit hinsichtlich ihrer Expression und auf ihre funktionellen Eigenschaften bezüglich des Auswachsens von Axonen, die die intrinsischen und afferenten hippokampalen Projektionen bilden, untersucht. Während die Expressionsmuster von Sema3A schon gut bekannt waren, wurde in der hier vorliegenden Arbeit die Expression des Rezeptors von Sema3A, Neuropilin-1 (NP-1), untersucht. NP-1 wird von Embryonaltag E17 an im entorhinalen Kortex, dem Subiculum und der hippokampalen Anlage exprimiert. Es konnte eine starke postnatale Expression von NP-1 in der CA3-Region und eine schwächere Expression in der CA1-Region, dem Gyrus dentatus und dem entorhinalen Kortex gezeigt werden. Außerdem wurden in dieser Arbeit die Expressionsmuster von Sema3C und Neuropilin-2 (NP-2) genauer analysiert. Etwa zum Zeitpunkt der Geburt (P0) wurde Sema3C im Gyrus dentatus und in der Cornu ammonis Region exprimiert. Der Sema3C-Rezeptor Neuropilin-2 wurde zu diesem Zeitpunkt ebenso im Gyrus dentatus und CA3-Region, schwächer auch in der CA1-Region exprimiert. Es wurde keine Expression dieser beiden Faktoren im entorhinalen Kortex detektiert. In Kokulturstudien zwischen mit Sema3A bzw. Sema3C transfizierten Zellaggregaten und Explantaten aus den hippokampalen Subregionen wurden für spezifische Explantate ein funktioneller Zusammenhang zwischen der Sekretion der Semaphorine und dem Auswachsen der jeweiligen Explantate in einer drei-dimensionalen Kollagenmatrix deutlich. Sema3A besitzt repulsive Eigenschaften auf Explantate vom Gyrus dentatus, der CA1- und der CA3-Region sowie dem entorhinalen Kortex. Die Interaktion zwischen Sema3A und NP1 beeinflußt das Einwachsen bzw. die Terminierung entorhinaler Fasern in der Molekularschicht des Gyrus dentatus, indem Sema3A eine repulsive Barriere für einwachsende Fasern und Moosfasern, die in Richtung der CA3-Region auswachsen, darstellt. Sema3C besitzt repulsive Eigenschaften auf Fasern des medialen Septums und beeinflußt dadurch das Einwachsen dieser Fasern entlang der Cornu ammonis Region in den Hippokampus. Weiterhin wurden in dieser Arbeit die Expressionsmuster von Netrin-1 und DCC im Hippokampus sowie die funktionellen Eigenschaften von Netrin-1 untersucht. Netrin-1 wird bereits zum Zeitpunkt E17 im Neokortex exprimiert, konnte im Hippokampus jedoch erst ab dem postnatalen Entwicklungsstadium P1 detektiert werden. Während im Gyrus dentatus nur ein schwaches und im entorhinalen Kortex kein Signal gefunden werden konnte, wird Netrin-1 stark in der Cornu ammonis Region exprimiert. Der Rezeptor DCC wird dagegen schon früher in der Embryonalentwicklung, ab E15, diffus in der hippokampalen Anlage exprimiert. Ab P1 lassen sich diese Signale im Gyrus dentatus und in den CA1-CA3-Regionen unterscheiden. Von den untersuchten Explantaten der hippokampalen Region zeigte Netrin-1 nur auf die Fasern von Gyrus dentatus und CA3, welche die hippokampale Kommissur bilden, einen attraktiven Wachstumseffekt. Dies bestätigt Befunde aus Netrin-1- und DCC-defizienten Tieren, in denen die hippokampale Kommissur aufgrund des fehlenden axonalen Leitmoleküls bzw. seines Rezeptors nicht ausgebildet wird. / In this work, the semaphorins Sema3A and Sema3C as well as Netrin-1 and their receptors, the neuropilins and DCC, were investigated regarding their expression and functional properties on outgrowing axons, which are forming the intrinsic and afferent hippocampal projections. Because of the already well known expression patterns of Sema3A, this work focused on the expression of the receptor of Sema3A, NP-1. From embryonic stage E17 on, NP-1 is expressed in the entorhinal cortex, the subiculum and the hippocampal Anlage. A strong postnatal expression of NP-1 in the CA3-region could be detected, while the expression pattern in the CA1-region, the dentate gyrus and the entorhinal cortex was weaker. Additionally, the expression patterns of Sema3C and NP-2 were investigated in greater detail. At birth (P0), Sema3C was expressed in the dentate gyrus and the cornu ammonis region. The expression of its receptor NP-2 could be detected at the same timepoint P0 in the dentate gyrus and the CA3-region and, less pronounced, in the CA1-region.There could not be detected any expression of Sema3C or NP-2 in the entorhinal cortex. In functional coculture studies between with Sema3A or Sema3C transfected cell clusters and neuronal explants from subregions of the hippocampal formation, these factors were investigated for their influence on axonal outgrowth within a three-dimensional collagen gel matrix. Sema3A has repulsive properties on explants from the dentate gyrus, the CA1- and CA3- regions and the entorhinal cortex. I the resulting model, the interaction between Sema3A and NP-1 influences the ingrowth and/or the termination of entorhinal fibers into the molecular layer of the dentate gyrus by a repulsive barrier formed by Sema3A. The same barrier also acts on mossy fibers to allow them to grow only in direction of the CA3-region. Sema3C has repulsive properties on fibers from the medial septum and shapes the ingrowth of these fibers along the cornu ammonis region into the hippocampus. Additionally, the expression patterns of Netrin-1 and DCC and their functional properties in the hippocampus were investigated. Netrin-1 is already expressed in the cortex at E17, although the onset of expression in the hippocampus is at P1. In the dentate gyrus, a weak signal could be detected, but no signal was found in the entorhinal cortex. In the cornu ammonis region, however, Netrin-1 showed a strong expression signal. The Netrin-1 receptor DCC could be detected as early as E15 with a diffuse distribution in the hippocampal Anlage. From P1 on, these signals could be distinguished in the dentate gyrus and the CA1-CA3-regions. Netrin-1 showed attractive properties only on fibers from explants of the dentate gyrus and the CA3-region, which form the hippocampal commissure. These results confirm previous findings from Netrin-1 and DCC deficient animals in which the absence of the hippocampal commissure was described.

La maladie de Parkinson (MP) est caractérisée par une dégénérescence des neurones dopaminergiques de la voie nigrostriée. La thérapie cellulaire, par transplantation intranigrale de cellules fœtales issues de mésencéphale ventral (MV), assure un rétablissement anatomique et fonctionnel de cette voie. Des molécules de guidage axonal (MGA) joueraient ainsi un rôle dans la reconnexion axonale des cellules transplantées. Pour tester cette hypothèse, nous avons étudié l'expression de MGA dans le cerveau adulte intact et dans des cellules destinées à la transplantation, ainsi que dans le cerveau adulte d'un modèle murin de la MP après transplantation. Dans le tissu intact, nous avons montré que semaphorin7A (Sema7A) et Sema3A et leurs récepteurs, plexinC1 et neuropilin1, conservent leur expression protéique. De plus, grâce à l'utilisation de puces à ADN, nous avons montré que les récepteurs Robo2, neuropilin1, neuropilin2, EphA5 et DCC sont exprimés de manière différentielle dans les deux populations cellulaires utilisées pour la transplantation. Ceci suggère que ces molécules seraient impliquées dans la restauration fonctionnelle observée. Enfin, dans le tissu lésé, nous avons observé, par RT-qPCR, des variations d'expression de l'ARNm de ces MGA après transplantation intranigrale des cellules fœtales du MV, suggérant plus particulièrement l'implication de Sema3A, Sema3F et Sema7A dans la reconstruction de la voie. Ce travail met en lumière l'action de sémaphorines dans le guidage axonal des cellules transplantées. L'intégration de ces MGA dans les procédures de transplantation pourrait aider à optimiser les procédures de thérapie cellulaire dans la MP. / Parkinson's disease (PD) is characterised by the degeneration of the dopaminergic nigrostriatal pathway. Cell therapy using intranigral transplantation of foetal ventral mesencephalon (VM) cells in a mouse model of PD results in anatomical and functional reconstruction of the pathway. This suggests a role for axon guidance molecules (GMs) in reconnecting transplanted cells to their striatal target. To test this hypothesis, we studied the expression of axon GMs in the intact adult brain, on cells used for transplantation and in a mouse model of PD after cell therapy. In the intact brain, we showed that GMs as semaphorin7A (Sema7A) and Sema3A and their corresponding receptors, plexinC1 and neuropilin1, retain an expression at the protein level, therefore showing a possible role for these guidance cues in the adult brain. Moreover, using microarray, we studied GM receptor expression profiles in two types of cells used for transplantation and exhibiting different functional ameliorations. Robo2, neuropilin1, neuropilin2, EphA5 and DCC receptors showed differential expression between the two cellular populations, indicating their possible contribution to the different functional outcomes observed. In the lesioned mouse brain, we observed, using RT-qPCR, variations of mRNA expression of these axon GMs after intranigral transplantation of foetal VM derived cells, thus suggesting the implication of Sema3A, Sema3F, and Sema7A in the reconstruction of the pathway. Overall, this work highlights particular importance of semaphorins in the nigrostriatal pathway reconstruction. Integrating these cues in transplantation procedures can possibly optimize cell therapy for PD patients.