A Computational Model of Cell-Generated Traction Forces and Fibronectin Assembly

Abstract

The extracellular matrix (ECM) is an assembly of proteins that surround cells and serves as the cell substrate in vivo. A primary component of newly synthesized ECM is fibronectin (FN), critical for embryonic development and wound healing. Despite years of research, the mechanism of FN assembly is still not understood. We hypothesize that FN assembly occurs through the revelation of buried FN-FN binding sites within any of the15 Type III FN domains, and that these binding sites are exposed by stretching of FN-III beta strands, allowing attachment of additional FN molecules via a beta strand addition mechanism. To investigate this hypothesis, we developed a biophysical computational model of cell-FN-substrate biomechanical-chemical interactions. In the model, FN-III domains are represented by Hookean springs with distinct stiffnesses; thus, each FN dimer is represented by 30 springs in series. Integrin binding/unbinding is represented by a stochastic first-order reversible chemical reaction with force-dependent off-rate. Model results are validated using in vitro tools: FN fibril growth and traction forces applied over time are measured using microfabricated pillar arrays in a human fibroblast cell line. Beta-strand exposure is probed using Thioflavin T, a fluorescent probe that binds to free beta strands. Results indicate that the computational model recapitulates three unique features that are observed in experimental measurements of FN fibrils: 1) FN fibrils are highly elastic, but have a maximal 4-fold elongation of their resting length; 2) FN fibril length and force are not strongly correlated; and 3) FN fibrils have discrete stable lengths, suggesting local minima of force balance within growing fibrils. These results will be discussed and compared with simulations in which FN fibril assembly is mediated by either a single cryptic binding site or a subset of cryptic binding sites.