• 2018-07
  • 2019-04
  • 2019-05
  • 2019-06
  • 2019-07
  • 2019-08
  • br Results This study was carried out with


    Results This study was carried out with a series of engineered extracellular matrix films of collagen assembled at hydrophobic surfaces, which allow for independent control of supramolecular structure, mechanical stiffness, and receptor recognition sites [17], [21]. Vascular smooth muscle Vicriviroc maleate chemical cultured on thin (less than a micrometer thick) films of fibrillar collagen exhibit spreading and proliferation identical to cells on conventional (millimeter thick) gels of collagen [17]. These films have robust handling properties, excellent optical properties, and allow high resolution analysis by atomic force microscopy and nanoindentation [18], [19]. The matrices are formed by self assembly of collagen fibrils of ∼200nm in diameter at hydrophobic surfaces such as monolayers of alkanethiols on gold [17] and untreated polystyrene dishes [21]. We have shown that collagen fibrils can be stiffened by simple dehydration, and cells respond to the stiffened matrix by increasing spreading and proliferation [18]. An additional level of control over collagen presentation is afforded by the fact that the extent of fibril formation can be decreased by lowering the concentration of the collagen solution exposed to the hydrophobic surface [17]. Collagen films formed from low concentrations of collagen in solution (10μg/mL) are monolayers of collagen and have few if any large fibrils (see Supplemental Fig. 1).
    Discussion Type I collagen is an important adhesive protein for use as scaffolding material for various tissue engineering applications [4], [5]. An understanding of how the various features of collagen matrices influence cell behavior is hence of great current interest [6]. Even for monotypic ligands, the context within which the adhesive ligands are presented to cells can have a dramatic effect on cellular phenotype [31], [32]. For naturally occurring extracellular matrix molecules such as collagen [33], [34] which present multiple adhesive interactions on the same molecule, the range of input cues sensed by cells may be expected to be even larger. Our study suggests that carefully engineered extracellular matrices that can be tailored to distinguish between the different features of the matrix can allow the dissection of cell responses with respect to matrix cues. For this study, we employed an engineered extracellular matrix of Type I collagen that was previously developed by our group, in order to dissect the role of collagen fibril mechanical stiffness and supramolecular structure on cell responses [17], [19], [21]. Direct observation of cells in contact with collagen fibrils indicates that collagen fibrils are mechanically flexible, and that vascular smooth muscle cells can move and rearrange the fibrils. Cells respond to this flexible matrix with poor spreading and proliferation equivalent to their response on collagen gels [18]. Passive dehydration makes the fibrils mechanically stiff; cells can no longer rearrange the fibrils and vascular smooth muscle cells spread and proliferate on these stiffer matrices [18]. In addition to increase the mechanical stiffness of the fibrils, we can also eliminate the presence of fibrils by decreasing the concentration of the collagen solution used to prepare the matrix. Low solution concentration of collagen results in a surface coating that appears to be composed of a monolayer of collagen devoid of large (∼200nm diameter) fibrils [17] (Supplementary Fig. 1). In this study, we show that when the collagen fibrils are treated with periodate, DDR2 activation is blocked. DDR2 activity has previously been shown to be dependent on recognition of glycosylation sites or other periodate-sensitive features [8]. We also show that cellular FAK is downregulated in vascular smooth muscle cells cultured on fibrillar Type I collagen but not monolayer Type I collagen. This downregulation appears to be independent of matrix stiffness of collagen fibrils. Downregulation of FAK also appears to be regulated by DDR2, as evidenced by the lack of DDR2 activation on a monolayer of collagen and the loss of FAK downregulation when cellular DDR2 was depleted by treating with DDR2 siRNA.