Fibrotic Extracellular Matrix Preferentially Induces a Partial Epithelial-Mesenchymal Transition Phenotype in a 3-D Agent Based Model of Fibrosis

Abstract

AbstractOne of the main drivers of fibrotic diseases is epithelial-mesenchymal transition (EMT): a transdifferentiation process in which cells undergo a phenotypic change from an epithelial state to a pro-migratory state. The cytokine transforming growth factor-β1 (TGF-β1) has been previously shown to regulate EMT. TGF-β1 binds to fibronectin (FN) fibrils, which are the primary extracellular matrix (ECM) component in renal fibrosis. We have previously demonstrated experimentally that inhibition of FN fibrillogenesis and/or TGF-β1 tethering to FN inhibits EMT. However, these studies have only been conducted on 2-D cell monolayers, and the role of TGF-β1-FN tethering in 3-D cellular environments is not clear. As such, we sought to develop a 3-D computational model of epithelial spheroids that captures both EMT signaling dynamics and TGF-β1-FN tethering dynamics. We have incorporated the bi-stable EMT switch model developed by Tian et al. (2013) into a 3D multicellular model, capturing both temporal and spatial TGF-β1 signaling dynamics. We show that the addition of TGF-β1 affects cell proliferation, EMT progression, and cell migration. We then incorporate TGF-β1-FN fibril tethering by locally reducing the TGF-β1 diffusion coefficient as a function of EMT to simulate the reduced movement of TGF-β1 when tethered to FN fibrils during fibrosis. We show that incorporation of TGF-β1 tethering to FN fibrils promotes a partial EMT state, independent of exogenous TGF-β1 concentration, indicating a mechanism by which fibrotic ECM can promote a partial EMT state.Author summaryEpithelial-mesenchymal transition (EMT) is a key cellular process where epithelial cells transform and become mesenchymal. The EMT states are not binary, but instead exhibit a spectrum of partial states where epithelial cells express a combination of epithelial and mesenchymal markers, along with several markers distinct to the partial state. In diseases such as fibrosis and cancer, growing evidence supports the finding that diseased epithelial cells exist primarily in a partial EMT state. However, the mechanisms and signaling factors that drive this partial EMT state in fibrotic diseases is unclear. We use an agent-based model that looks at EMT progression in a population of cells embedded in an ECM environment with controllable fibrotic properties to provide a more systematic approach at studying spatial and temporal changes in the microenvironment that could drive EMT progression and maintain specific EMT phenotypes.