L’adhérence, l’activation et l’agrégation des plaquettes sanguines sont essentielles à l’hémostase mais peuvent également conduire à la thrombose artérielle sur plaque d’athérosclérose, aujourd’hui première cause de mortalité dans le monde. Les anti-thrombotiques actuels, dirigés contre l’activation et l’agrégation plaquettaires, ont une efficacité reconnue mais ont pour inconvénient d’augmenter le risque de saignement. L’objectif de cette thèse a été d’explorer de nouvelles stratégies réduisant la thrombose tout en préservant l’hémostase. L’utilisation de souris modifiées génétiquement a mis en évidence que l’intégrine alpha6 beta1, impliquée dans l’adhérence des plaquettes aux laminines, joue un rôle critique en thrombose expérimentale mais pas en hémostase. De plus, nous avons montré dans un système de perfusion de sang qu’une protéine préférentiellement exprimée dans les plaques d’athérosclérose, la ténascine-C, permet l’adhérence et l’activation des plaquettes. En revanche, la beta-arrestine-1, une protéine de signalisation, ne contribue que modestement aux fonctions plaquettaires et à la thrombose. En conclusion, ce travail a permis de dégager deux nouvelles pistes anti-thrombotiques potentiellement capables de préserver l’hémostase, basées sur le ciblage de l’intégrine alpha6 beta1 ou de l’interaction plaquette/ténascine-C. / Following vascular injury, blood platelet adhesion, activation and aggregation are essential for hemostasis but can also lead to arterial thrombosis, which is a leading cause of death worldwide. Current antithrombotic drugs impede platelet activation and aggregation, thereby considerably reducing cardiovascular mortality, but their use is linked to an increased bleeding risk. This thesis aimed to explore more selective strategies causing minimal perturbation of hemostasis. The use of genetically-modified mice has revealed an unsuspected important contribution of integrin alpha6 beta1, which mediates platelet adhesion to laminins, to experimental arterial thrombosis but not hemostasis. In addition, we showed that tenascin-C, an extracellular matrix protein overexpressed in atherosclerotic plaques, can support platelet adhesion and activation under flow. In contrast, the signaling protein beta-arrestin-1 does not play a major role in platelet function, hemostasis and thrombosis. In conclusion, this work provides two interesting candidates, namely integrin alpha6 beta1 and tenascin-C, to put into practice the concept of targeting thrombosis while minimally impairing hemostasis.

Integrins have the ability to impact major aspects of epithelial biology including adhesion, migration, invasion, signaling and differentiation, as well as the formation and progression of cancer (Hynes 2002; Srichai and Zent 2010; Anderson et al. 2014). This thesis focuses on how integrins are regulated and function in the context of mammary epithelial biology and breast cancer with a specific focus on the α6 integrin heterodimers (α6β1 and α6β4). These integrins function primarily as receptors for the laminin family of extracellular matrix (ECM) proteins and they have been implicated in mammary gland biology and breast cancer (Friedrichs et al. 1995; Wewer et al. 1997; Mercurio et al. 2001; Margadant and Sonnenberg 2010; Muschler and Streuli 2010; Nistico et al. 2014). The first project investigates how alternative splicing of the α6 subunit impacts the genesis and function of breast cancer stem cells (CSCs). This work revealed that the α6Bβ1 splice variant, but not α6Aβ1, is necessary for the function of breast CSCs because it activates the Hippo transducer TAZ (Zhao et al. 2008a), which is known to be essential for breast CSCs (Cordenonsi et al. 2011). My work also led to the discovery that laminin (LM) 511 is the specific ligand for α6Bβ1 and that autocrine LM511, which is mediated by TAZ, is needed to sustain breast CSCs by functioning as a ‘ECM niche’. An important aspect of this study is the finding that surface-bound LM511 characterizes a small population of cells in human breast tumors with CSC properties. The second project of my thesis concentrated on identifying transcription factors that regulate expression of the β4 subunit. The expression of the α6β4 integrin is repressed during the epithelial-mesenchymal transition (EMT) (Yang et al. 2009) but the contribution of specific transcription factors to this repression is poorly understood. This study revealed that Snai1 is a transcriptional repressor of β4, which is responsible for establishing the PRC2 (Polycomb complex 2)- associated repressive histone mark H3K27Me3. However, I also found that the ability of Snai1 to repress transcription is abrogated by its interaction with Id2. Specifically, I identified the biochemical mechanism for how Id2 regulates Snai1. Id2 binds the SNAG domain of Snai1 that is the docking site for several corepressors (Peinado et al. 2004; Lin et al. 2010b; Dong et al. 2012a). One important consequence of Id2 interacting with Snai1 on the β4 promoter is that it prevents repressive epigenetic modifications. This finding may explain why some epithelial cells express Snai1 and β4 because they also express Id2 (Vincent et al. 2009; Bastea et al. 2012). The repression of the α6β4 integrin during the EMT is consistent with data indicating that this integrin is not expressed in CSCs (Mani et al. 2008; Goel et al. 2012; Goel et al. 2013; Goel et al. 2014). An important question going forward is to understand how the α6β4 integrin contributes to tumor formation. In summary, my thesis provides novel insights into the biology of the α6 integrins that has important implications for the function of these integrins in mammary gland biology and breast cancer, especially our understanding of breast CSCs.