In a targeted proteomic screen to identify polarity protein complexes, a number of Scribble (Scrib) -associating proteins were identified; of particular interest was the Nitric Oxide Synthase 1 Adaptor Protein (NOS1AP). NOS1AP contains an N-terminal phosphotyrosine binding (PTB) domain and a C-terminal PSD-95/Dlg homology/ZO-1 (PDZ) binding motif that associates with neuronal NOS (NOS1). We show that the PTB domain of NOS1AP associates with the fourth PDZ domain of Scrib. We identified NOS1AP binding partners including three key regulators of dendritic spine formation, beta-Pix, Git1, and PAK, which require Scrib to associate with NOS1AP. Overexpression of NOS1AP in cultured hippocampal neurons increases dendritic protrusions, a process dependent on the PTB domain. The increase in dendritic protrusions can be blocked by the co-expression of a dominant negative Rac construct. NOS1AP, and the PTB domain of NOS1AP influence Rac activity. Together these data suggest that Scrib and NOS1AP function as important scaffolding proteins in the mammalian synapse and that NOS1AP functions in the dendritic spine by influencing Rho GTPase activity.

The current study characterizes a novel isoform of the Nitric Oxide Synthase 1 Adaptor Protein (NOS1AP), herein NOS1APc. NOS1APc was identified in a proteomic screen for Scribble interacting proteins. Northern blot analysis revealed a 7kb transcript that was present in a number of different tissues and cell lines. Highest levels were detected in the cerebellum. In situ hybridization studies revealed NOS1APc mRNA throughout the cortex and hippocampus. In addition, cerebellar Purkinje cells and different brainstem nuclei also contained NOS1APc mRNA. Antibodies directed against NOS1APc revealed a 100 kDa protein, while immunohistochemical staining revealed high levels of this protein within the molecular layer of the cerebellum. Finally, immunocytochemical studies using isolated primary astrocytes revealed a subset of BrdU positive nuclei that co-expressed NOS1APc, suggesting that this protein may function in some capacity in cell-cycle progression.

Au cours de la tumorigénèse, les cellules tumorales accumulent plusieurs mutations. Chez les modèles animaux comme la drosophile ou la souris, il a été montré que la coopération d’au moins deux altérations génétiques permet de reproduire une croissance néoplasique dans des cellules épithéliales : la sur-activation d’une voie de signalisation comme Ras or Notch, et des mutations altérant l’adhésion cellulaires ou la polarité. Bien que cette coopération soit très décrite, ses mécanismes sous-jacents ne sont que peu étudiés. Deux principaux modèles peuvent être proposés : soit les altérations génétiques fonctionnent indépendamment (modèle additif), ou ils s’influencent l’un l’autre pour faire émerger de nouveaux comportements (modèle synergique). En utilisant un paradigme généré par Notch de croissance hyperplasique et néoplasique dans des disques larvaires d’aile de Drosophila melanogaster, nous cherchons à confirmer l’un de ces deux modèles.En combinant des analyses de RNAseq et de ChIP, nous avons pu identifier des ensembles de gènes directement activés par Notch dans des disques hyperplasiques (Notch activé) et néoplasiques (Notch activé et perte de polarité). Bien que des cibles directes de la voie Notch soient partagées par les deux types de croissances, nos analyses ont montré qu’une grande partie de ces cibles sont spécifiques de chacune des conditions ce qui suggère que l’état polarisé ou non d’une cellule influence le contrôle transcriptionnel de la voie Notch. Enfin en réalisant un screen centré sur les gènes les plus affectés en néoplasie, nous avons identifié un ensemble de gènes qui contrôlent la transition entre l’hyperplasie et la néoplasie. / During tumourigenesis, tumour cells accumulate many mutations. In animal models such as Drosophila or mouse, the cooperation of at least two genetic insults has been shown to recapitulate neoplastic transformation of epithelial cells: an overactive signalling pathway such as Ras or Notch, and mutations affecting cell adhesion or polarity. While this cooperation is well documented, the mechanisms underlying it are still poorly understood. Two main models could be proposed: either the genetic alterations act independently (additive model), or they influence each other to allow the emergence of new behaviours (synergistic model). Using Notch driven paradigms of hyperplastic and neoplastic growth in Drosophila wing discs, we sought to distinguish between these two models.Combining RNAseq and ChIP analysis, we were able to identify the repertoires of genes directly activated by Notch in hyperplastic discs (Notch activation) and in neoplastic discs (Notch activation and polarity loss). While some Notch direct targets are shared between the two types of growth, our analysis revealed that the majority of Notch targets are actually specific to each condition, suggesting that a correct polarity has a major influence on the transcriptional output of the Notch signalling pathway, and strongly supports the synergistic model. Our studies further revealed that histone composition changes and polycomb mark changes are amongst the mechanisms that redirect the Notch pathway. Finally, using a small scale functional screen centred on the most affected genes in neoplasia, we identified a core set of genes controlling the transition from hyperplastic to neoplastic growth.

Cell competition leads to elimination of a viable cell population, by fitter cells. Despite over forty years of research, the molecular mechanisms of competition in mammals are poorly understood. During my PhD I have investigated the mechanisms of competition by exploring an established mammalian cell culture system, in which wild-type MDCK cells eliminate scribble-deficient cells, and I have also developed a novel cell culture system to model mammalian competition. My work contributed to the discovery that scribble-deficient cells are eliminated not by biochemical exchange among cells, but by mechanical compaction. We termed this phenomenon mechanical competition. I employed transcriptional profiling to determine the molecular signature of mechanical losers, and identified activation of p53 signalling as their hallmark. My colleagues and I then demonstrated that elevation of p53 is both necessary and sufficient to trigger mechanical competition. In further investigating the mechanisms of mechanical competition, I found that compaction activates ROCK in scribble-deficient cells, and that this is required for their elimination. Inhibition of Src signalling in mechanical losers also protected them form out-competition, and integrin signalling is another pathway likely involved in mechanical competition. While investigating p53 competition, we observed that p53-high and p53-low cells engage in directional migration, with p53-high cells always at the migrating front. As a side-project, I investigated the role of p53 in directional migration, by exploring an established model with a single leader cell and multiple followers. We established a method to generate multinucleated leaders on demand. By creating leaders from p53-deficient cells, I established that p53 signalling is required for some, but not all multinucleated cells to trigger collective migration, thus implicating p53 signalling in a type of migration involved in wound healing. Finally, I successfully modelled p53-driven mechanical competition in a differentiated primary tracheal epithelial cell culture, thereby establishing a novel system to study mammalian competition, and also proving that p53 competition is conserved between different mammalian epithelia. Considering the involvement of p53, mechanical competition may play a major role in cancer.

Pathological changes to tumor extracellular matrix (ECM) composition, mechanics, and architecture promote cancer progression and metastasis. Exploration of tumor-ECM interactions using in vitro matrix-mimetic culture systems has largely been restricted to naturally-derived matrix materials that permit limited experimental control. Such study of a novel lung adenocarcinoma model in Matrigel™ (MG) has suggested key matrix cues that mediate epithelial-mesenchymal transition (EMT) and metastasis. In this thesis work, synthetic hydrogel scaffolds based on poly(ethylene glycol) (PEG) featuring high experimental control and modular bioactivity were used to study matrix influences on the EMT-prone model line 344SQ. Encapsulation of 344SQ cells in PEG hydrogels modified for cell adhesivity and cell-mediated enzymatic degradability induced formation of lumenized, polarized spheres mimicking the epithelial phenotype observed in three-dimensional MG. Tuning matrix stiffness, adhesive ligand concentration, and ligand spatial presentation altered epithelial morphogenesis. Exploration of the EMT phenotype of PEG-encapsulated 344SQ cells revealed TGFβ-initiated changes in morphology, polarity, expression levels of EMT marker genes and their epigenetic controller, and the organization of cell-secreted ECM. Notably, a potent role for adhesive ligand was illuminated as matrices with low PEG-RGDS concentration even in the absence of TGFβ induced formation of spheres with a post-EMT phenotype by several of these measures. A matrix-invasive phenotype was also revealed by altering matrix structural parameters and tuned with incorporation of an alternative protease-cleavable sequence. Finally, the influence of cell-cell contacts was explored by covalent incorporation of cadherin proteins into the matrix. Matrix-tethered E- and -N-cadherin affected 344SQ sphere development in otherwise non-cell-adhesive matrices and modulated polarity and the degree of TGFβ response. Further, in 344SQ with a knockdown of the essential polarity-determining protein Scribble, matrix-tethered cadherin influenced the formation of a phenotype with partially normalized epithelial polarity with corresponding differences in membrane localization of cell-expressed E-cadherin. Overall, this thesis demonstrates the utility of the more experimentally controllable PEG system in studying ECM influences on cancer progression with findings providing greater insight into stromal biomechanical, biochemical, and cell-cell factors that mediate lung adenocarcinoma epithelial morphogenesis and EMT. These contributions help advance the state of the field towards a goal of developing new metastasis-targeting cancer therapeutics.

No description available.

No description available.

No description available.

Planar cell polarity (PCP) is a developmental signaling mechanism that establishes a polarity within the plane of an epithelium. PCP has been shown to play a role in guiding numerous neurodevelopmental processes such as convergent extension, neuron migration, and axon pathfinding. Certain commissural neurons in the dorsal spinal cord make a series of guidance decisions en route to the brain: first, a ventral projection along the D-V axis, followed by a midline crossing, and after exiting the floorplate, a dorso-anterior turn along the A-P axis. Here, we provide in vivo evidence that the axons of the Commissural Primary Ascending (CoPAs) neurons in zebrafish require the PCP genes fzd3a, vangl2, and scribble for rostral pathfinding both before and after crossing the midline. Dorsoventral guidance of CoPA axons is unaltered in fzd3a, vangl2, and scribble mutants, suggesting that the PCP signaling pathway only controls A-P guidance of CoPAs. Our results have provided evidence for two potential non- mutually exclusive models: (i) A-P axon guidance is achieved by cell-autonomous Wnt-Frizzled signaling or that (ii) A-P axon guidance is achieved by non-cell-autonomous PCP signaling in the neuroepithelial environment. The single-cell nature of the CoPA axon system allows for simple genetic manipulation and visualization, which will potentially elucidate the validity of either model. Scribble (Scrib), a member of the LAP family, plays a critical role in establishing and regulating cell polarization in epithelia and during cell migration. In zebrafish, Scrib mutants have defects in convergent extension (CE) cell movements and facial branchiomotor neuron (FBMN) migration. Despite our understanding of Scrib’s genetic role in neurodevelopment, little is known about the subcellular localization of endogenous Scrib in vivo during CE and FBMN migration. We have generated a monoclonal antibody against the C-terminus of zebrafish Scrib and have shown that this antibody is specific against endogenous Scrib in both western blot and immunocytochemical applications. Confocal microscopy of Scrib immunocytochemistry shows that at various developmental stages, Scrib distinctly localizes to basolateral membranes of non polarized epithelium, to the membrane in mesodermal cells undergoing CE, and to the membrane of migrating FBMNs. Furthermore, the distribution of Scrib puncta along membranes of FBMN- FBMN contact is significantly altered in the PCP mutant pk1b. Further application of our newly generated Scrib antibody will potentially lead to new insight on Scrib’s role in neurodevelopment.

We worked on finding a new kinase regulator to develop basic data to be used for cancer prevention. Our work found a link between three previously unrelated proteins involved in cancer, ERK2 and ERK6 and Scribble. The MAP kinase cascade is involved in cell proliferation, which is highly deregulated in cancer. Through the screening of ERK6 associating molecules, we found the cell polarity, and cell cycle related molecule Scribble. Furthermore, we found that Scribble was a dual-specific kinase regulator. We clearly demonstrated that these ERKs interact with Scribble through the LRR and the PDZ domains of Scribble. We hypothesize that Scribble may function as a scaffolding protein for ERK2 and ERK6, since Scribble has been found to down regulate the kinase activity of these ERKs.