Author:

Keywords:

in vitro angiogenesis model, cytoskeleton, displacement fields, collagen, sprouts, endothelial cells, extracellular matrix, inhibitors, cytochalasin D, blebbistatin, actin, non muscle myosin II

Abstract:

Introduction: During sprouting angiogenesis rearrangements of acto-myosin stress fibers and focal adhesion complexes are indispensable. These processes allow the endothelial cells to generate cellular forces required for protruding into the surrounding matrix. At the cell-matrix interface the traction forces exerted by the sprout can be investigated indirectly by mapping the resulting matrix deformations. In this study the roles of actin filaments and non-muscle myosin II (NMMII) for sprouting angiogenesis are assessed by mapping the matrix deformations of growing sprouts under chemically defined culture conditions blocking acto-myosin force generation. Materials and Methods: Human umbilical vein endothelial cells (HUVEC) were transduced with a LifeAct adenoviral vector to visualize the actin cytoskeleton during live sprouting into a collagen type I gel containing the pro-angiogenic factor sphingosine-1-phosphate. The sprouts were grown in endothelial cell growth medium (EGM2) supplemented with or without blebbistatin (inhibition of NMMII ATPase activity) or cytochalasin D (CytoD; inhibition of actin polymerization) and for varying concentrations and incubation periods. Second harmonic generation (SHG) and laser-scanning confocal microscopy (CLSM) were used to acquire Z-stacks of label-free collagen fibers and 200 nm fluorescent beads dispersed in the collagen, respectively, during the live imaging before and after chemically induced relaxation of the cells. The calculation of the matrix deformations was formulated as a B-spline-based 3D non-rigid image registration process that warps the image of the stressed gel to match the image after relaxation, either from beads or from fibers. Results: For a concentration of 4µM CytoD from the onset of the sprouting, sprout formation is blocked entirely, while 50 µM of blebbistatin still allows sprout formation. For the same concentrations, mature HUVEC sprouts (grown overnight) show reduced to entirely blocked filopodia dynamics within 90 minutes when treated in situ with CytoD, while the effect of blebbistatin is again less prominent. Fiber and bead displacements were of the order of magnitude of 5 to 10 µm. Conclusion: Our methodology allows mapping traction-induced matrix deformations of 3D collagen matrix around angiogenic sprouts. Ultimately, this study will allow estimating the forces exerted by sprouting endothelial cells within a range of chemically defined culturing conditions, in this way shedding light on mechano-chemical feedback mechanisms important for sprouting angiogenesis.