Advances in modelling of epithelial to mesenchymal transition
2013-07-11T15:14:00Z (GMT) by
Epithelial to Mesenchymal Transition (EMT) is a cellular transformation process that is employed repeatedly and ubiquitously during vertebrate morphogenesis to build complex tissues and organs. Cellular transformations that occur during cancer cell invasion are phenotypically similar to developmental EMT, and involve the same molecular signalling pathways. EMT processes are diverse, but are characterised by: a loss of cell-cell adhesion; a gain in cell-matrix adhesion; an increase in cell motility; the secretion of proteases that degrade basement membrane proteins; an increased resistance to apoptosis; a loss of polarisation; increased production of extracellular matrix components; a change from a rounded to a fibroblastic morphology; and an invasive phenotype. This thesis focuses explicitly on endocardial EMT, which is the EMT that occurs during vertebrate embryonic heart development. The embryonic heart initially forms as a tube, with myocardium externally, endocardium internally, with these tissue layers separated by a thick extracellular matrix termed the cardiac jelly. Some of the endocardial cells in specific regions of the embryonic heart tube undergo EMT and invade the cardiac jelly. This causes cellularised swellings inside the embryonic heart tube termed the endocardial cushions. The emergence of the four chambered double pump heart of mammals involves a complex remodelling that the endocardial cushions play an active role in. Even while heart remodelling is taking place, the heart tube is operating as a single-circulation pump, and the endocardial cushions are performing a valve-like function that is critical to the survival of the embryo (Nomura-Kitabayashi et al. 2009). As the endocardial cushions grow and remodel, they become the valve leaflets of the foetal heart. The endocardial cushions also contribute tissue to the septa (walls) of the heart. Their correct formation is thus essential to the development of a fully functional, fully divided, double-pump system. It has been shown that genetic mutations that cause impaired endocardial EMT lead to the development of a range of congenital heart defects (Fischer et al. 2007). An extensive review is conducted of existing experimental investigations into endocardial EMT. The information extracted from this review is used to develop a multiscale conceptual model of endocardial EMT, including the major protein signalling pathways involved, and the cellular phenotypes that they induce or inhibit. After considering the requirements for computational simulations of EMT, and reviewing the various techniques and simulation packages available for multi-cell modelling, cellular Potts modelling is selected as having the most appropriate combination of features. The open source simulation platform Compucell3D is selected for model development, due to the flexibility, range of features provided and an existing implementation of multiscale models; that include subcellular models of reaction pathways. Based on the conceptual model of endocardial EMT, abstract computational simulations of key aspects are developed, in order to investigate qualitative behaviour under different simulated conditions. The abstract simulations include a 2D multiscale model of Notch signalling lateral induction, which is the mechanism by which the embryonic heart tube is patterned into cushion and non-cushion forming regions. Additionally, a 3D simulation is used to investigate the possible role of contact-inhibited mitosis, upregulated by the VEGF protein, in maintaining an epithelial phenotype. One particular in vitro investigation of endocardial EMT (Luna-Zurita et al. 2010) is used to develop quantitative simulations. The quantitative data used for fitting the simulations consist of cell shape metrics that are derived from simple processing of the imaging results. Single cell simulations are used to investigate the relationship between cell motility and cell shape in the cellular Potts model. The findings are then implemented in multi-cell models, in order to investigate the relationship between cell-cell adhesion, cell-matrix adhesion, cell motility and cell shape during EMT.