Poster T19 - Directed Induction of the Epicardial to Mesenchymal Transition

Introduction: During heart development, epicardial cells (EpiCs) play a key role in supporting cardiomyocytes and creating functional stromal cell populations. EpiCs provide paracrine cues and undergo the epicardial-to-mesenchymal transition (EMT), where they migrate and differentiate into mesenchymal cell types including cardiac fibroblasts, smooth muscle cells, and endothelial cells [1]. EMT is heavily regulated by the TGF-β pathway; when inhibited, it causes EpiCs in vitro to maintain their epithelial state. Typically, EpiCs become dormant following development, however, there is evidence that following ischemic injuries like myocardial infarction EpiCs become active again and contribute to the wound healing process [2]. This process is limited, and full regeneration would require additional engineered tissue or other external catalysts to promote the healing process. EpiCs can be incorporated into tissue-engineered heart constructs to improve the healing process, but initially, the process of EMT needs to be further explored. In this study, EpiCs underwent EMT induction via modulation of the TGF-β pathway; with both TGF-β and FGF2 known to play roles in heart development, using them to direct EMT in vitro should result in homogenous populations of smooth muscle cells and fibroblasts, respectively, without impeding migration typically observed in EMT. Initially, a combinatorial assay was performed to determine what growth factor concentration optimized EMT induction, followed by a 3D microtissue expansion study to observe how these factors affect migration during EMT. Finally, a 2D study at the same concentration was performed in a further effort to determine the phenotype of the mesenchymal cells derived from EMT induction.

Materials and

Methods: Human induced pluripotent stem cells (iPSCs) were differentiated into cardiac progenitor cells and then into EpiCs. To differentiate into EpiCs, the Wnt pathway was modulated following a standard protocol [3]. At D11, EpiCs were singularized with Accumax and replated at 30,000 cell/cm2 onto 0.1% Gelatin (combinatorial assay and 2D phenotype experiments) or formed into 10,000 cell microtissues following a protocol developed previously [4] (3D expansion experiments). For the combinatorial assay, cells were expanded until confluency and then treated with EMT-inducing conditions for 7 days; these conditions were 0-50 ng/mL FGF2 and/or 0-10 ng/mL TGF-β, with 0.5 μM A83-01 (TGF-β inhibition) acting as the control. In both 3D expansion and 2D induction, EpiCs were treated with the same conditions as identified in literature for 7 days: 0.5 μM A83-01, no treatment, 5 ng/mL TGF-β, 10 ng/mL FGF2, or 3 days with 5 ng/mL TGF-β followed by 3 days with 10 ng/mL FGF2 (referred to as T0F). After 7 days, cells were fixed with 4% paraformaldehyde and treated for immunofluorescent imaging. Phalloidin was used to assess the morphology of cells, WT1 to identify the EpiC population, and Vimentin and Calponin to identify the mesenchymal population (fibroblasts and smooth muscle cells, respectively); Hoechst stained all nuclei. The total expansion area (mm2) was compared between D1 and D7 and measured manually using ImageJ, and statistical analysis was performed across and within treatments.

Results, Conclusions, and Discussions: Following 7 days of treatment in the combinatorial assay, all groups containing A83-01 retained their epithelial state, and without it had heavy loss of WT1 expression. Phalloidin revealed that groups without A83-01 had prominent stress fibers representative of cell migration and more spindled morphology indicative of mesenchyme cell types. Vimentin had its highest expression with 5 ng/mL TGF-β and 50 ng/mL FGF2, while Calponin had its highest expression with 10 ng/mL FGF2 and 5 ng/mL TGF-β. In 3D expansion, all groups exhibited significant expansion across 7 days of culture (Fig. 1C), though there were varying amounts across groups (Fig. 1B). The A83 group exhibited the most expansion, consistent with EpiC behavior, with the FGF group following, indicative of fibroblast migration. Directed-EMT groups all displayed high expression of Vimentin and Calponin, with the FGF displaying the highest Vimentin expression, the TGF and FGF displaying similarly high Calponin expression, and the T0F having reduced expression of both (Fig. 1A). Under the same conditions in 2D, WT1 expression was lost following 7 days in EMT groups, with phalloidin showing spindled morphology in those groups (Fig. 2). Vimentin expression was reduced within the FGF group compared to 3D expansion, indicating less fibroblast specificity. Additionally, stress fibers were prominent within the A83 group with increased expression of Vimentin and decreased WT1 expression. This could be a sign of premature EMT, with overproliferation of EpiCs causing migratory behavior to create an area for expansion, and the loss of WT1 in select areas combined with increased Vimentin expression identifying cells that could have led that process. Loss of WT1 expression and morphological changes under EMT-inducing conditions shows the importance of A83-01 in regulating EpiC phenotypes; overproliferation could have caused premature EMT and migration of EpiCs, warranting further investigation. Vimentin expression is inconsistent between the same conditions in 3D and 2D, but a concentration of 50 ng/mL FGF combined with 2 ng/mL TGF-β shows the most favorable condition for fibroblast differentiation. Ongoing studies including bulk RNA sequencing will be used to assess the homogeneity and efficacy of each differentiation condition directing EMT.

Acknowledgements (Optional): Funding for this study was provided by the National Science Foundation (CAREER EBMS 2237898) and National Institutes of Health (1R15HL140745).