Myocardial infarction (MI) results in scarring that is essential for mechanical reinforcement of the damaged wall. At later times however, progressive fibrosis of adjoining tissue contributes to subsequent heart failure. The cardiac fibroblasts (CFs) are the main cell type that mediates fibrosis via conversion to myofibroblasts. This change in cell fate is driven mainly by TGFÎ², which induces SMAD2/3 signalling and myofibroblast gene expression. Previous studies suggested that in healthy hearts, strains are primarily uniaxial, whereas in injured hearts, increases in multi-axial components are predictive of fibrosis. To study this problem in vitro, CFs were suspended in fibrin gels, resulting in replacement of the fibrin with cell-derived matrix over about a week. Gels were anchored either uniaxially so that force direction mimicked normal heart tissue, or biaxially to introduce a perpendicular component of mechanical stress. Myofibroblast conversion was strikingly higher in biaxial matrix constructs, with higher expression of Î±-SMA and production of fibrillar collagens. Despite similar levels of active TGFÎ² in both settings, SMAD2 activation was strikingly higher in biaxial constructs, demonstrating than TGFÎ² sensitivity rather than availability mediates the difference. These effects correlated with higher levels of TGFÎ²R1 in biaxial matrix constructs. Investigation of mechanism behind decreased TGFÎ²R1 levels ruled out mRNA levels and degradation rates as potential mechanisms. Decreased TGFÎ²R1 translation would appear to be the remaining explanation, though this point remains to be directly demonstrated. Analysis of exogenous TGFÎ²R1 vectors suggests that the 5âUTR contains a positive element that overcomes a negative signal to promote TGFÎ²R1 translation in biaxial setting. Applying strains to the 3D constructs showed that TGFÎ²R1 expression and matrix phenotype were strongly affected by polarity but only weakly by strain magnitude. To validate these results in vivo, we examined mouse heart tissue after MI due to ligation of the left anterior descending coronary artery. In the infarct zone at 14 days post-MI, depolarization of the scar adjoining tissue correlated with high SMAD2 activation, consistent with our in vitro model. Together, these data indicate that cell polarity is a major factor in progressive cardiac fibrosis after MI, through modulation of TGFÎ² sensitivity by regulation of TGFÎ²R1 protein levels. We also analysed miRNAs in this process. Biaxial constructs showed decreased abundance of mature microRNAs (miRs). This effect coincided with an increase in total Drosha levels but it was relocalized to the cytoplasm in biaxial matrix constructs. Drosha accumulation was decreased by inhibition of p38 and FAK but their effects on Drosha localization remains to be established. Drosha levels were controlled by cellular tension and unaffected by changes in polarity. These findings suggest a hypothetical pathway in which higher tension in the biaxial setting results in activation of FAK, which activates p38, which phosphorylates Drosha to induce its nuclear export and stabilization. Loss of Drosha in the nucleus may impair processing of the majority of pri-miRs and decrease mature miRs. Decreased miRNAs-mediated suppression of pro-fibrotic genes then promotes production of fibrotic matrix.