Hybrid Epithelial–Mesenchymal Phenotypes Are Controlled by Microenvironmental Factors

Abstract

Epithelial-to-mesenchymal transition (EMT) has been associated with cancer cell heterogeneity, plasticity, and metastasis. However, the extrinsic signals supervising these phenotypic transitions remain elusive. To assess how selected microenvironmental signals control cancer-associated phenotypes along the EMT continuum, we defined a logical model of the EMT cellular network that yields qualitative degrees of cell adhesions by adherens junctions and focal adhesions, two features affected during EMT. The model attractors recovered epithelial, mesenchymal, and hybrid phenotypes. Simulations showed that hybrid phenotypes may arise through independent molecular paths involving stringent extrinsic signals. Of particular interest, model predictions and their experimental validations indicated that: (i) stiffening of the extracellular matrix was a prerequisite for cells overactivating FAK_SRC to upregulate SNAIL and acquire a mesenchymal phenotype and (ii) FAK_SRC inhibition of cell-cell contacts through the receptor-type tyrosine-protein phosphatases kappa led to acquisition of a full mesenchymal, rather than a hybrid, phenotype. Altogether, these computational and experimental approaches allow assessment of critical microenvironmental signals controlling hybrid EMT phenotypes and indicate that EMT involves multiple molecular programs.Significance: A multidisciplinary study sheds light on microenvironmental signals controlling cancer cell plasticity along EMT and suggests that hybrid and mesenchymal phenotypes arise through independent molecular paths.

In silicological modelling to uncover cooperative interactions in cancer

“…In agreement with experimental data reported in the literature, this model indicates that cellular inflammation simulated by the constant overactivation of NFKB, results in a higher likelihood of reaching a mesenchymal stem-like state. Still focusing on EMT, and using cell adhesion properties as read-outs for the acquisition of EMT phenotypes, we recently published a model, which accounts for the epithelial, hybrid and mesenchymal phenotypes acquired by cancer cells [41]. Main outcomes of this model, and novel analyses that lead to the identification of cooperative mechanisms underlying the EMT process are described in the next section.…”

“…Using a logical model of the EMT cellular network, we have identified cooperative signals controlling cancer-associated phenotypes amid the EMT continuum [41]. This in silico model encompasses 39 intracellular nodes, including EMT transcription factors, epithelial (ECad and miR200) and mesenchymal (SNAIL, SLUG, ZEB, TCF/LEF, BCat) markers and known EMT signalling pathways (RAS, NOTCH, WNT, TGFB, JAK/STAT, Hippo, Integrins and AKT).…”

“…Experimental validations using the MCF10A and MDCK cell lines with conditional SRC activation, grown on collagen gels of different Young’s moduli or in which the expression of RPTP-kappa was affected, confirmed the model predictions. Thus, our model permits to uncover cooperative mechanisms leading to the acquisition of EMT phenotypes [41].…”

<p>Role of αVβ3 in Prostate Cancer: Metastasis Initiator and Important Therapeutic Target</p>

“…Metastasis is a process that involves interactions between the cell and extracellular matrix (ECM) as well as cell and target organ. [9][10][11] At the beginning of metastasis, tumor cells must destroy the ECM as this structure prevents cancer cells from leaving their primary site. Cells must survive in the vasculature and finally be deposited in the corresponding tissue.…”

Semi-automatic model revision of Boolean regulatory networks: confronting time-series observations with (a)synchronous dynamics

“…Different formalisms have been used by the community to model biological regulatory networks, such as Ordinary Differential Equations (ODE), Piecewise-Linear Differential Equations (PLDE), Logical Formalism, Sign Consistency Model (SCM), among others (De Jong, 2002;Karlebach and Shamir, 2008;Siegel et al, 2006). Here, we consider the (Boolean) logical formalism, which has proven useful to study and analyse dynamical behaviour of biological models (Thomas, 1973;Naldi et al, 2015;Abou-Jaoudé et al, 2016;Selvaggio et al, 2020).…”