Full-Length Fibronectin Drives Fibroblast Accumulation at the Surface of Collagen Microtissues during Cell-Induced Tissue Morphogenesis

Generating and maintaining gradients of cell density and extracellular matrix (ECM) components is a prerequisite for the development of functionality of healthy tissue. Therefore, gaining insights into the drivers of spatial organization of cells and the role of ECM during tissue morphogenesis is vital. In a 3D model system of tissue morphogenesis, a fibronectin-FRET sensor recently revealed the existence of two separate fibronectin populations with different conformations in microtissues, i.e. ‘compact and adsorbed to collagen’ versus ‘extended and fibrillar’ fibronectin that does not colocalize with the collagen scaffold. Here we asked how the presence of fibronectin might drive this cell-induced tissue morphogenesis, more specifically the formation of gradients in cell density and ECM composition. Microtissues were engineered in a high-throughput model system containing rectangular microarrays of 12 posts, which constrained fibroblast-populated collagen gels, remodeled by the contractile cells into trampoline-shaped microtissues. Fibronectin’s contribution during the tissue maturation process was assessed using fibronectin-knockout mouse embryonic fibroblasts (Fn(-/-) MEFs) and floxed equivalents (Fn(f/f) MEFs), in fibronectin-depleted growth medium with and without exogenously added plasma fibronectin (full-length, or various fragments). In the absence of full-length fibronectin, Fn(-/-) MEFs remained homogenously distributed throughout the cell-contracted collagen gels. In contrast, in the presence of full-length fibronectin, both cell types produced shell-like tissues with a predominantly cell-free compacted collagen core and a peripheral surface layer rich in cells. Single cell assays then revealed that Fn(-/-) MEFs applied lower total strain energy on nanopillar arrays coated with either fibronectin or vitronectin when compared to Fn(f/f) MEFs, but that the presence of exogenously added plasma fibronectin rescued their contractility. While collagen decoration of single fibronectin fibers enhanced the non-persistent migration of both Fn(f/f) and Fn(-/-) MEFs, the migration speed was increased for Fn(-/-) MEFs on plasma fibronectin fibers compared to Fn(f/f) MEFs. In contrast, the average speed was the same for all cells on collagen-coated Fn fibers. A Fn-FRET sensor revealed that fibronectin on average was more extended on the microtissue surface compared to fibronectin in the core. Gradients of collagen-to-fibronectin ratios and of the fraction of collagen-adsorbed to stretched fibrillar fibronectin conformations might thereby provide critical cell migration cues. This study highlights a dominant role for fibronectin in tissue morphogenesis and the development of tissue heterogeneities.