Glaucoma is the leading cause of irreversible blindness worldwide, and is a major cause of blindness in the United States. It affects approximately 5% of Caucasians and 10% of African- Americans over the age of 60 years. Elevated intraocular pressure (IOP) is currently the only modifiable risk factor for glaucoma. Impaired outflow of aqueous humor from the eye is thought to be the cause of pathologically elevated IOP. However, the etiology of outflow impairment is unknown. Anatomically, the aqueous humor drains into the iridocorneal angle of the eye, where the iris inserts at the transition between the cornea and sclera. In humans, approximately 80-90% of the aqueous traverses through the trabecular meshwork (TM), juxtacanalicular connective tissue (JCT), Schlemm’s canal, collector channels and empties into episcleral veins. Abnormalities at these sites are thought to cause impaired outflow. Abnormal accumulation of extracellular matrix (ECM) in the TM or JCT, abnormal endothelial function in Schlemm’s canal, or a combination of these components have been strongly implicated. Our laboratory has focused on the role of Secreted Protein Acidic and Rich in Cysteine (SPARC) in regulating outflow. SPARC is the prototypical matricellular protein that mediates ECM organization and turnover in numerous human tissues. Our lab was first to demonstrate that SPARC is highly expressed in the TM and JCT regions of the eye, and that the SPARC knockout (KO) mouse has a significant decrease in IOP of 15-20%. SPARC may affect the degree of segmental flow, a theory that states that variable aqueous outflow occurs around the circumference of the eye; only certain portions of the TM are thought to display active outflow at any particular moment. The cytokine transforming growth factor-ß2 (TGFß2) has been shown to modulate multiple ECM proteins, including SPARC. TGFß2 is significantly upregulated by 2 to 3-fold in the aqueous humor of glaucoma patients compared to controls. In addition, when TGFß2 is overexpressed in rodent eyes, increased ECM deposition is observed within the trabecular meshwork leading to IOP elevation. SPARC is one of the most highly upregulated proteins by TGFß2, and is downstream of TGFß2. We hypothesized that wild-type (WT) mice would demonstrate segmental flow, while SPARC KO mice would display a more continuous pattern of outflow around the eye. We also believed that IOP would be inversely correlated with outflow area. We also hypothesized that SPARC is essential to the process of TGFß2-mediated ocular hypertension, and that the lack of SPARC would impair IOP elevation. We conducted a tracer study utilizing fluorescent microbeads to determine the location of outflow circumferentially around the mouse TM. Microbeads were injected intracamerally into the eyes of WT and KO mice. After a 45-minute incubation period, the mice were euthanized and eyes were processed for confocal, light, and electron microscopy. During the second group of experiments, empty or TGFß2-containing adenovirus was injected intravitreally into WT and SPARC KO mice and IOP was measured for 2 weeks. Immunohistochemistry was completed on all tissues to assess for changes in major ECM proteins. Percentage effective filtration length (PEFL), or area of the TM labeled by tracer, was significantly increased in SPARC KO mice (70.61% ± 11.36%, p<0.005; N=11) compared to WT mice (54.68% ± 9.95%; N=11). In addition, the pressures between the two sets of eyes were significantly different with mean pressures of 16.3 mm Hg in WT mice and 12.6 mm Hg in KO mice (p<0.005, N=11 pairs). In addition, PEFL and IOP were inversely correlated with R2 = 0.72 (N=10 pairs); in eyes with higher IOP, PEFL was reduced. Electron microscopy demonstrated that high-tracer TM areas had a greater separation between trabecular beams. Collagen fibril diameter was found to be smaller in the KO (28.272 nm) compared to WT (34.961 nm; p<0.0005, N=3 pairs). These data provided structural correlations to the functional data regarding segmental flow. In the second set of experiments, IOP was found to be significantly elevated in TGFß2- injected WT mice compared to empty vector-injected WT mice during days 4-11 (p<0.05, N=8). However, IOP was not significantly elevated in TGFß2-injected KO mice compared to controls. Immunohistochemistry demonstrated that TGFß2 increased expression of collagen IV, fibronectin, plasminogen activator inhibitor-1 (PAI-1), connective tissue growth factor (CTGF), and SPARC within the TM of WT mice, but only PAI-1 and CTGF in KO mice (p<0.05, N=3 pairs). These data support our hypotheses, indicating that SPARC plays an integral role in the modulation of aqueous humor outflow. In addition, it appears as though SPARC is essential to the regulation of TGFß2-mediated ocular hypertension. Aside from providing further evidence of the importance of ECM in IOP regulation, our work presents the novel discovery of segmental flow in the mouse. Given the potential role of SPARC in TGFß2-mediated ocular hypertension, SPARC may not only play an integral role in ECM homeostasis within the trabecular meshwork, but may be a valuable target for pharmacologic therapy in treating primary open-angle glaucoma.

Le complexe transcriptionnel AP-1 est une famille ubiquitaire de facteurs de transcription dimériques. Ses composants les mieux étudiés sont les membres des familles multigéniques Fos et Jun. Les mécanismes transcriptionnels gouvernés par ce complexe sont encore mal caractérisés, en raison du grand nombre de dimères AP-1 possibles, de l’abondance et de l’activité finement régulées de ses constituants qui dépendent des contextes cellulaires et physiopathologiques. De plus, les limitations techniques ont longtemps donné l'impression que AP-1 régule l’expression de ses gènes cibles en se fixant principalement à proximité de leurs promoteurs. Fra-1 est la protéine de la famille Fos la plus souvent impliquée dans les cancers épithéliaux. En particulier, elle est surexprimée dans les cancers du sein triple négatifs (TNBCs) où elle contribue à la tumorigenèse et à l'agressivité tumorale par des effets pléiotropes. Dans ce contexte, l’objectif de mes travaux de thèse était d’aboutir à une meilleure compréhension des actions transcriptionnelles de Fra-1 au niveau du génome dans une lignée cellulaire TNBC de référence, la lignée MDA-MB-231. Pour ce faire, j'ai combiné des données transcriptomiques avec des données de ChIP-seq et de NG-Capture C (technique à haute résolution et à haut débit dérivée du 3C). J'ai également inclus dans ces études le membre Fra-2, de la famille Fos, qui présente la même spécificité de fixation à l’ADN et est également exprimé dans les TNBCs, bien qu'à un niveau beaucoup plus bas, où il contribue aussi au phénotype tumoral. En accord avec leurs effets pléiotropes, Fra-1 et Fra-2 activent ou répriment, soit individuellement soit de façon redondante ou complémentaire, l’expression de nombreux gènes associés à une large gamme de processus biologiques. Il est intéressant de noter que la régulation des gènes cibles est rarement due à la liaison de Fra-1 et/ou Fra-2 au niveau des régions promotrices de ces gènes mais fait intervenir leur liaison sur des enhancers distaux. Mes résultats de NG-Capture C suggèrent la présence d’interactions chromatiniennes à longue distance enhancer/promoteur, ainsi que des réseaux d’enhancers. Ces réseaux contiennent des enhancers liés par Fra-1 et d’autres indépendants de celui-ci. Aucune preuve d’un rôle de Fra-1 dans le contrôle des interactions chromatiniennes au niveau de ces réseaux n'a été trouvée en utilisant un panel de 35 gènes régulés par ce facteur. En parallèle, j'ai abordé les mécanismes de la répression transcriptionnelle médiée par Fra-1, mécanismes très rarement étudiés dans la littérature, en utilisant deux gènes modèles, TGFB2 et SMAD6. Ces études ont mis en évidence des mécanismes différents mis en jeu par Fra-1 pour la répression de ces deux gènes, ce qui montre la complexité des mécanismes de la régulation transcriptionnelle médiée par Fra-1. / The AP-1 transcription complex is a ubiquitous family of dimeric transcription factors. Its best-studied components are the members of the Fos and Jun multigene families. The mechanisms whereby AP-1 exerts its transcriptional actions are still ill-understood due to the wide number of possible AP-1 dimers and the exquisitely regulated abundance and activity of its constituents, that all depend on the cell types and physiopathological contexts. Moreover, technical limitations have long given the impression that AP-1 mostly operates in the vicinity of gene promoters. Fra-1 is the Fos family protein that is most often implicated in epithelial cancers. In particular, it is overexpressed in triple negative breast cancers (TNBCs) where it contributes to tumorigenesis and tumor aggressiveness through pleiotropic effects. Based on this, the aim of my thesis was to gain a more comprehensive view of Fra-1 transcriptional actions at the genome-wide level in the MDA-MB-231 reference TNBC cell line, . To this aim, I have combined transcriptomic data with ChIP-seq and NG-Capture C (high resolution, high throughput 3C-derived technique) data. I have also included in my studies its Fos family kin Fra-2, as it displays the same DNA binding specificity and is also expressed in TNBCs, albeit at a much lower level, where it also contributes to the tumor phenotype. Consistently with their pleiotropic effects, Fra-1 and Fra-2 were found to up- or down-regulate either individually, together or redundantly many genes associated with a wide range of biological processes. Interestingly, the regulation of target genes is rarely due to Fra-1 and Fra-2 binding at gene promoters, but involves their binding to distal enhancers. My NG-Capture C results imply the presence of long-range chromatin interactions in Fra-1 modes of action, as well as enhancer hubs containing Fra-1- and non-Fra-1-binding enhancers. No evidence for a role for Fra-1 in the control of chromatin looping was however found using a panel of 35 Fra-1-regulated genes. Moreover, I have addressed the mechanisms of transcriptional repression mediated by Fra-1, as these have practically never been studied, using two model genes, TGFB2 and SMAD6. These studies underlined different mechanisms employed by Fra-1 for the repression of these genes, embodying the complexity of Fra-1 transcriptional regulation mechanisms.