Furthermore we have identified the

Furthermore, we have identified the tumour stroma, and pancreatic stellate oxygenase in particular, to be an abundant source of NODAL and ACTIVIN, thus creating a supportive niche for these highly plastic cells (Lonardo et al., 2012). Since pancreatic cancers are usually highly desmoplastic tumours which are characterized by extensive infiltration of stromal cells, paracrine NODAL/ACTIVIN secreted by stellate cells and TGFβ secreted by CSCs seem to be critical for the perpetuation of CSCs, and interference with this cross-stimulation seems to be highly relevant for successful therapy (Lonardo et al., 2011; Lonardo et al., 2012). Accordingly, a miR-19-72 cluster that has been shown to regulate TBX3 expression confers a cancer stem cell phenotype to non-CSCs (Cioffi et al., 2015).
Conflict of interest
Acknowledgement We are indebted to Claudia Ruhland and Claudia Längle their excellent technical support. We thank Daniel Hartmann for providing mRNA from normal pancreas and PDAC samples. This study was funded by the Deutsche Forschungsgemeinschaft (DFG, K.L. 2544/1-1 and 1-2), the Forschungskern SyStaR to A.K., BIU (Böhringer Ingelheim Ulm to A.K.), the Else-Kröner-Fresenius-Stiftung (2011_A200), a German Cancer Aid Max Eder Fellowship to P.C.H., a German Cancer Aid Grant to A.K. (111879), the Fritz-Thyssen Foundation to A.K. (2015-00363) and the Hector Foundation Cancer Research Fund to P.C.H. A.K. is indebted to the Baden-Württemberg Stiftung for the financial support of this research project by the Eliteprogramme for Postdocs. A.K. is also an Else-Kröner-Fresenius Memorial Fellow. This work was also supported by the Interdisciplinary Center for Clinical Research (IZKF Aachen), RWTH Aachen University Medical School, Aachen, Germany to I.G.C. The Else Kröner-Forschungskolleg Ulm supports L.P.M.H. receives funding from the Bausteinprogramm of Ulm University. We thank Dr. Ronan Russell for generating some of the Tbx3 gain of function cell lines and for supporting us with the CAM experiments and intellectual input.
Introduction The mammalian heart begins as a linear tube that is comprised to two cells types - myocardium, the muscle cells that drive mechanical function, and endocardium, the primary endothelial cell (EC) population (Harvey, 2002). In order to form a multi-chambered organ, the heart tube undergoes “looping” and endocardial cells undergo endothelial to mesenchymal transition (EndMT), migrate into the cardiac jelly and proliferate to form the cardiac valves and septum (von Gise and Pu, 2012). EndMT has long been thought to occur exclusively in the context of embryonic cardiogenesis, however, recent studies provide evidence for EndMT in adult tissues that is thought to contribute to the pathogenesis of many diseases, including pulmonary (Arciniegas et al., 2005), intestinal (Rieder et al., 2011), kidney (Zeisberg et al., 2008) and cardiac fibrosis (Zeisberg et al., 2007). As such, the molecular pathways that govern EndMT are not only relevant to basic studies of cardiogenesis, but also may enable targeted therapies that improve functional perfusion of diseased or injured tissues. Multiple signaling pathways have been linked to EndMT, including transforming growth factor beta (TGFβ) (Gonzalez and Medici, 2014), nitric oxide (NO) (Chang et al., 2011), and fibroblast growth factor (FGF) (Chen et al., 2012), however, Notch signaling in known to play a central role in both the demarcation of boundaries within the embryonic vascular network (Benedito et al., 2009) as well as EndMT (Niessen and Karsan, 2008). The Notch family is comprised of four transmembrane receptors (Notch1–4) that are activated by membrane-anchored ligands Delta-like (DLL) 1, 3 and 4 and Jagged (JAG) 1 and 2. Notch ligand on one cell can trans-activate Notch receptor on neighboring cells, resulting in release of the Notch intracellular domain (NICD) via proteolytic processing and translocation to the nucleus to activate transcription of downstream targets such as HES and HEY (Guruharsha et al., 2012). Yet Notch ligands are also capable of inhibiting receptor activation in the cell they are presented on (cis-inhibition) (de Celis and Bray, 1997; Klein et al., 1997; Micchelli et al., 1997; Miller et al., 2009; Sprinzak et al., 2010). Differential activity between DLL and JAG ligands adds further nuance to Notch signaling; in cells expressing “fringe” genes (lunatic fringe, manic fringe or radical fringe), Notch receptors are modified such that DLL ligand strongly activates, but JAG ligand inhibits, signaling (Benedito et al., 2009). This many-tiered regulatory apparatus enables the formation of stark boundaries that are essential for tissue morphogenesis. However, due to the complexity of Notch signal regulation, interpretation of phenotypes in transgenic mouse and in vitro gain/loss of function studies can be challenging.