To unite or disconnect : AmotL2 in tubulogenesis and tumor invasion

For an epithelial- or endothelial cell sheet or ‐tissue to form, individual cells must come together unite and connect. Contacts to adjacent cells and underlying matrix must be initiated, and the junctional proteins mediating the contacts must be further linked to intracellular cytoskeletal networks. By connecting neighboring cells into one unit, contractile actomyosin filaments allow for a closely interlinked, but yet dynamic tissue. Eventhough essential for developmental processes such as organ formation, the exact mechanism of how the cell‐cell junctions connect to the contractile actomyosin network has not yet been completely revealed. In the papers of this thesis, we identify the protein Angiomotin Like 2 (AmotL2 p100) as a linker between the contractile radial actin filaments and VE‐cadherin at the adherens junctions. Furthermore, AmotL2 p100 enables controlled‐ and synchronized morphological alterations of individual cells, which can further result in the creation of new tissue level structures, for example through tubulogenesis. We show the radial AmotL2‐mediated actin filaments to be crucial for force‐generation during morphological transformation and further aortic lumen expansion. Morphological transformations usually require organization and collaboration of several processes, such as formation/disassembly of cell‐cell and cell‐ECM contacts, establishment/disruption of apical-basal polarity and polymerization/disassembly of cytoskeletal filaments. Just like proper regulation of those intra‐ and inter cellular processes and –signals can result in complex structures with diverse morphologies and functionalities, deregulation of the same signals might cause devastating consequences for the function of an organ and the entire organism. In Paper IV of this thesis, we could identify a shorter-, hypoxia regulated AmotL2 p60 isoform. We show the actions of AmotL2 p60 to cause retainment of apical polarity proteins in cytoplasmic vesicles, hence preventing the establishment of apical-basal polarity. Furthermore, we could show AmotL2 p60 to weaken cell‐cell junctions and sensitize cells to growth factor stimuli. The combined actions of AmotL2 p60 cause tumor cell invasion into the surrounding extracellular matrix (ECM). In conclusion, we here provide data showing the two AmotL2 isoforms to possess entirely distinct functions, uniting cells into a multicellular structure, and disconnecting cells during tumor invasion, respectively