Microfluidic Devices for Quantifying the Role of Soluble Gradients in Early Angiogenesis - Mechanical and Chemical Signaling in Angiogenesis - page 53

4 Recent Results and Quantitative Studies
Our laboratory recently reported the development of microfluidic devices that
enable the culture of cells in a low-shear, stable soluble gradient environment for
mechanistic studies of sprouting morphogenesis. These devices allow decoupling
of the role of soluble gradients from other environmental cues, such as signals
from the ECM or biomechanical forces. To design such devices, we first simulated
several configurations in silico [ 30 ]. For each simulation, the diffusion and flows
were modeled using the Navier-Stokes equations, solved numerically by a finite
element method (Fig. 4 b, c). Depending on the diffusivity of the gradient mole-
cules (which is related to the molecular weight of the diffusing solute and the
medium present in the cell culture chamber), devices with either two or four inlets
were selected (Fig. 4 a). Soluble factors diffuse through micro-capillaries that
connect the reagent channels to the cell culture chamber and inhibit convective
transport into the culture chamber. In order for the gradient to remain stable, the
characteristic diffusion time across the culture chamber must be less than the
characteristic convection time through the source and sink channels. A flow rate of
8 nL/min was found to be suitable for a soluble gradient of VEGF diffusing
through liquid culture medium. The distribution of a fluorescently labeled marker
(e.g., 20 kDa FITC-dextran) can be quantified to validate the stability and linearity
of the gradient (Fig. 4 d, e) [ 30 ]. This model molecule formed a quantitatively
predictable gradient at steady-state after 40 min. At this flow rate, convective
forces in the culture chamber are predicted to generate a maximal shear stress on
the cell-surface of 10 -5 dyn/cm 2 (using a Stoke's drag approximation), which is
orders of magnitude less than the amount of shear known to induce cellular
alignment of ECs [ 65 ]. In contrast, devices that use flows within the cell culture
chamber to generate steady-state gradients expose cells to shear on the order of
10 -2 dyn/cm 2 [ 66 ].
Use of this microfluidic device has demonstrated that (1) the magnitude of the
VEGF gradient must meet a minimum threshold to induce EC chemotaxis, and
(2) chemotaxis is independent of concentration across the VEGF range of 18-32 ng/
mL [ 30 ]. Chemotaxis in response to identical gradient steepness but different
average concentrations was quantified by counting net accumulation or depletion of
cells after 6 h in each quadrant of a device, (Fig. 5 a, b). By this measure, human
umbilical vein endothelial cell (HUVEC) chemotaxis was induced by a VEGF
gradient of magnitude 14 ng/mL/mm (Fig. 5 c), whereas a magnitude of 2 ng/mL/
mm was insufficient to induce chemotaxis (Fig. 5 d, e). This threshold is a conse-
quence of stochasticity in factor diffusion and receptor binding. Although the
probability that a gradient imparts its direction to a cell via spatially biased receptor
potentiation increases linearly with the magnitude of the gradient, cells have
intrinsic randomness in motility that must be overcome before chemotaxis can
happen [ 67 ]. In addition to reductively assessing gradient magnitude, this experi-
ment decoupled the effects of the average VEGF concentration from the VEGF
gradient steepness. During analysis, the cell culture chamber was divided into four