oa Stress-Fiber Remodeling in 3D: ‘Contact Guidance vs Stretch Avoidance’?

When engineering heart valves in vitro, matrix anisotropy is considered important for long-term in vivo functionality. However, it is not fully understood how to guide, maintain and control matrix anisotropy. Experiments suggest that collagen anisotropy is affected by actin-mediated cell traction and associated cellular orientation. Although cellular orientation in 2D can be manipulated via imposed uniaxial cyclic stretch, 3D data are lacking. We questioned how cyclic stretch influences actin and collagen orientation in 3D constructs. A novel micro-tissue platform system was designed, able to dynamically and biochemically load small-scale cell-populated fibrous tissues. Flexible membranes of Bioflex culture plates were provided with a rectangular array silicone posts. These silicone posts constrained a contracting gel mixture of human vena saphena cells (HVSC), collagen type I and matrigel. The constrained tissues were subjected to pure uniaxial cyclic stretch (10%, 0.5Hz, in the presence or absence of agents) on the Flexcell system. F-actin was taken as a measure for the cell traction direction, and the F-actin orientation was quantified throughout the complete tissue thickness (~ 300m) using fiber-tracking software, and was fitted using a bi-model distribution function. Uniaxial cyclic stretching for 3 days (preceded by 3 days of static constraint) resulted in stress-fibers that were oriented perpendicular to the stretching direction only at tissue surfaces, as generally observed in 2D. Strikingly, however, in the tissue core F-actin (and cell) and collagen orientation was biaxial. Immediate cyclic stretching, starting before polymerization of the collagen matrix, resulted in a strong stretch avoidance throughout the tissue, of both F-actin and collagen. We systematically investigated the effect of biochemical treatment, including MMP1 to perturb matrix integrity, MMP-1 + ROCK-inhibitor to counteract the possible MMP1-induced response, as well the effect of a lower and a higher initial collagen density. Experimental data suggests that F-actin stress-fibers avoid cyclic stretch in 3D, unless collagen contact guidance dictates otherwise.