Microfluidic Devices for Angiogenesis - Mechanical and Chemical Signaling in Angiogenesis - page 101

Although the surfaces of these devices had always been treated with some form of
adhesion factors (e.g., collagen, fibronectin, etc.), an important step toward greater
realism occurred when these devices began to incorporate compliant substrates,
either fabricated from elastomeric materials or incorporating regions of hydrogel
that the cells could be plated onto, thereby relaxing the constraint of using an
unrealistic substrate stiffness. Hydroglels such as collagen, fibrin, Matrigel TM , and
others (gelatin, laminin, alginate, and chitosan) came into common usage [ 69 ].
These became widely referred to as three-dimensional (3D) substrates, even though
the cells generally remained on the 2D surface of the gel. Most of these hydrogels
naturally contained or were functionalized with motifs that could bind to integrin
receptors (e.g., the RGD sequence), giving rise to cell adhesion, but also to the
signaling associated with integrin activation. Results from these compliant systems
differed from those on rigid substrates in numerous ways, but most noticeably in
terms of the nature of the focal adhesions formed, being more focal in nature on rigid
substrates, but more diffuse and ill-defined on 3D gels [ 70 ].
Studies were also conducted on or in rigid substrates in 1D, 2D, and even 3D
[ 71 , 72 ] with the aim of developing microvascular networks for applications such
as tissue engineering. While these could sustain a flow, they lacked endothelial
cells, and were often not biocompatible. They did, however, demonstrate the
capability of fabricating complex vascular-like structures using micro- and nano-
fabrication methods.
A second major advance occurred with the introduction of soft lithographic
methods as a means of producing microfluidic channels and systems, typically
from poly(dimethylsiloxane) (PDMS), a biocompatible elastomer [ 26 ]. Numerous
advantages accrued from this new approach. One was that, since the systems were
fabricated by means of casting the PDMS onto silicon wafers, the wealth of
knowledge gained from the use of silicon technology in the semiconductor
industry could immediately be applied to the fabrication of these miniaturized
devices for cell culture. Ultimately, the inherent low Young's modulus of PDMS,
although far in excess of biological tissues, could be used to allow substrate
stretch, simulating the deformations of the vascular wall, or to measure contractile
forces of cells using systems of micropillars upon which the cells could be cultured
[ 73 ]. Finally, the high permeability of PDMS to gas and liquid [ 74 ] proved a mixed
blessing. On one hand, gas exchange in an incubator helped to maintain desired
levels of O 2 and CO 2 . Given the reduced volumes of these miniaturized systems
(larger surface-to-volume ratio), this no doubt helps limit the need for frequent
media changes. On the other hand, it also led to higher rates of evaporation, a
limitation in some experimental situations. Finally, the tendency for PDMS to
adsorb small hydrophobic molecules also has a negative influence in some
experiments, especially in the context of drug screening [ 75 ].
Despite the obvious advantages of microfluidics in studying vascular biology,
its use for investigations of angiogenesis is a relatively recent development. Early
attempts to produce microvascular networks focused not on angiogenic responses,
but rather on the fabrication of preformed vascular network patterns in PDMS or
stiff gels, and subsequent seeding with endothelial cells. This work was motivated