<p>Basement membranes are a dense, sheet-like form of extracellular matrix that underlie epithelia and endothelia, and surround muscle, fat and Schwann cells. Basement membranes separate tissues and protect them from mechanical stresses. Although traditionally thought of as a static support structure, a growing body of evidence suggests that dynamic basement membrane deposition and modification instruct cell behavior and morphogenetic processes. In this thesis, I discuss how changes to basement membrane affect anchor cell (AC) invasion during C. elegans uterine vulval attachment. During AC invasion, the uterine AC breaches two juxtaposed basement membranes to contact the underlying vulval epithelium. Using live-cell imaging, genetics, molecular biology and electron microscopy I identify three modifications to the BM that affect AC invasion. In Chapter 2, I describe a system for linking juxtaposed basement membranes to stably align or connect adjacent tissues. This adhesion system promotes rapid AC invasion and also regulates a more long-term connection between the uterine tissue and the hypodermal seam cell in the adult worm. Chapter 3 elucidates how the BM component SPARC promotes cell invasion. As SPARC overexpression is correlated with cancer metastasis, this aims to understand how SPARC overexpression promote invasion in a pathological situation. In Chapter 4, I discuss preliminary data showing that the AC actively secretes laminin into the basement membrane targeted for invasion. I outline how future studies could elucidate the mechanism by which AC-derived laminin might promote cell invasion. Finally, Chapter 5 discusses conclusions and future directions for these studies.</p> / Dissertation

<p>Basement membrane (BM) is the thin, dense, highly cross-linked form of extracellular matrix that underlies all epithelia and endothelia, as well as surrounds muscle, nerve and fat. These sheet-like networks function as physiological barriers to maintain tissue homeostasis. During normal developmental processes and immune surveillance, cells invade through BM to establish tissues and fight infection. Similarly, metastatic cancer cells are thought to co-opt normal programs for BM transmigration as they spread from primary tumors and colonize distant tissues. The difficulty of visualizing cell-BM interactions during invasion in vivo has left the cellular and molecular mechanisms used to breach BM undefined. Specialized F-actin-rich matrix-degrading membrane protrusions, termed invadosomes, have been described in cultured invasive cell lines for more 30 years. Invadosomes are hypothesized to mediate BM penetration during cancer metastasis. Despite promising advances in intravital imaging technologies, however, invadosomes have yet to be observed in cells transmigrating BM in vivo, leaving their physiological relevance unclear. Anchor cell invasion in C. elegans is a simple in vivo model of cell invasion that allows for combined visual and genetic analysis of BM transmigration. In this dissertation I develop high-resolution time-lapse imaging approaches to understand the dynamic interactions that occur at the AC-BM interface during invasion. Through the course of this work we identify an integrin-based mechanism that polarizes the AC towards the BM. We further discover protrusive F-actin-based invadosome structures that mediate BM breach during anchor cell (AC) invasion. We find that in most cases only one or two invadosomes penetrate the BM and then transform into an invasive protrusion that guides the AC through a single BM gap. Using genetics and quantitative single-cell image analysis we characterize several molecular regulators of invadosome formation in vivo. Our findings establish an essential role for invadosomes during BM transmigration in vivo, and support the idea that these structures are a core, conserved element of a normal invasive cellular strategy activated during cancer metastasis.</p> / Dissertation

La membrane basale (MB) est une feuille dense de matrice extracellulaire spécialisée qui sépare l'épithélium du tissu sous-adjacent. La pénétration des cellules à travers les barrières des MB, c’est appelée «invasion», est un processus important pour le développement normal des tissus et dans la métastase du cancer. Beaucoup a été compris sur la génétique et la signalisation sur comment les trous sont formés dans le MB pendant de l'invasion ont été compris. Cependant, les forces physiques impliquées sont moins comprises: comment la contractilité de la myosine participe à l'élimination de la MB et comment les différents facteurs de polymérisation de l’actine et les protéines de réticulation contribuent au processus invasif. Pour répondre à ces questions, nous avons étudié un événement d'invasion dans un processus de développement, l’invasion de la cellule anchre (AC) chez Caenorhabditis elegans. La rupture de la MB par l’AC est connue pour dépendre d'une protrusion riche en actine et de l'activité des métalloprotéases (MMP), similaire à l'invasion de cellules cancéreuses. Inactivation génique par RNAi de différents activateurs et nucléateurs de la polymérisation d'actine, et l'expression spécifiquement dans l'AC d'une forme négative dominante d'un activateur du complexe Arp2/3 a montré que l'invasion de l’AC dépendes fortement des filaments ramifiés formés via l'activation WASP / WSP-1 du complexe Arp2/3. La microscopie à haute résolution a indiqué que la protrusion invasive de l’AC était densément assemblée, en accord avec l'idée que la protrusion invasive était fortement ramifiée. Nous avons également montré qu'un autre activateur du complexe Arp2/3, WAVE / WVE-1, pouvait permettre une invasion lorsque WASP / WSP-1 était absent. Les formines semblaient ne pas de jouer un rôle majeur et les protéines de réticulation d'actine étaient également dispensables pour l’invasion de l’AC. Dans les vers de type normaux, nous avons observé que l'activité de la myosine n'était pas nécessaire pour l'invasion. Cependant, il a été rapporté que les cellules cancéreuses augmentent la contractilité de la myosine pour envahir en l'absence de protéases, nous avons donc utilisé un ver sans les cinq principales MMPs du génome du ver pour tester le rôle de la myosine dans ce contexte. L'invasion de l’AC a eu lieu dans en l’absence des MMPs, mais avec un retard. L’inactivation génique par RNAi de différents composants lies à l’activité de la myosine n'a pas amélioré le défaut d'invasion. En plus, la visualisation du cytosquelette d'actine dans les vers sans MMPs a révélé que l'actine était concentrée dans la protrusion de l’AC et à peine détectable dans le cortex, ce qui rendait improbable que la contraction de la myosine du cortex aiderait la compression cellulaire à travers de la MB comme il a été reporté dans les cellules cancéreuses en l'absence de protéases. Tous ces résultats ensemble, ont montré que la cellule invasive a adapté sa polymérisation de filaments d'actine ramifiée pour maintenir l'invasion dans différents contextes biochimiques et environnementaux. Cette plasticité est un point crucial qui doit être mieux compris pour développer de futurs traitements visant l'invasion de cellules cancéreuses. / Basement membrane (BM) is a dense sheet of specialized extracellular matrix that separates epithelia from underlying tissue. The penetration of cells through BM barriers, called “invasion”, is an important process during normal tissue development and in cancer metastasis. Much has been understood concerning the genetics and signaling of how holes are formed in the BM during invasion. However less is clear about the physical forces involved: how myosin contractility participates in BM removal and how different actin polymerization factors and crosslinkers contribute to the invasive process. To address these questions, we studied an invasion event in a developmental process, anchor cell (AC) invasion in Caenorhabditis elegans. AC breaching of the BM is known to depend on an actin-rich protrusion and the activity of matrix metalloproteases (MMPs), similar to cancer cell invasion. RNAi knockdown of different actin polymerization activators and nucleators, and expression of a dominant negative form of an Arp2/3 complex activator specifically in the AC showed that AC invasion depended strongly on branched filaments formed via WASP/WSP-1 activation of the Arp2/3 complex. Super-resolution microscopy indicated that the AC invasive protrusion was densely packed with filaments, in keeping with the idea that the invasive protrusion was highly branched. We further showed that another Arp2/3 complex activator, WAVE/WVE-1, could enable invasion when WASP/WSP-1 was absent. Formins appeared not to play a major role and actin cross-linking proteins were likewise dispensable for AC invasion. In wild type worms, we observed that myosin activity was not needed for invasion. However it has been reported that cancer cells upregulate myosin contractility to invade in the absence of proteases, so we used a worm deleted for the five main MMPs of the worm genome to test the role of myosin in this context. AC invasion took place in MMP- worms, but with a time delay. RNAi knockdown of different components of the myosin machinery gave no enhancement of the invasion defect. In addition visualization of the actin cytoskeleton in MMP- worms revealed that actin was concentrated in the AC protrusion and barely detectable in the cortex, making it unlikely that myosin contraction of the cortex was helping the cell squeeze through the BM as reported in cancer cells in the absence of proteases. All together these results showed that the invasive cell adapted its branched actin filament polymerization to maintain invasion in different biochemical and environmental contexts. This plasticity is a crucial point that needs to be better understood in order to develop future treatments targeting cancer cell invasion.