The matricellular extracellular matrix protein thrombospondin-1 (TSP1) stimulates focal adhesion disassembly through a sequence (known as the hep I peptide) in its heparin-binding domain. This mediates signaling through a receptor co-complex involving calreticulin and low-density lipoprotein (LDL) receptor-related protein (LRP). We postulate that this transition to an intermediate adhesive state enhances cellular responses to dynamic environmental conditions. Since cell adhesion dynamics affect cell motility, we asked whether TSP1/hep I-induced intermediate adhesion alters cell migration. Using both transwell and Dunn chamber assays, we demonstrate that TSP1 and hep I gradients stimulate endothelial cell chemotaxis. Treatment with focal adhesion-labilizing concentrations of TSP1/hep I in the absence of a gradient enhances endothelial cell random migration, or chemokinesis, associated with an increase in cells migrating, migration speed, and total cellular displacement. Calreticulin-null and LRP-null fibroblasts do not migrate in response to TSP1/hep I, nor do endothelial cells treated with the LRP inhibitor receptor-associated protein (RAP). Furthermore, TSP1/hep I-induced focal adhesion disassembly is associated with reduced chemotaxis to basic fibroblast growth factor (bFGF) but enhanced chemotaxis to acidic (a)FGF, suggesting differential modulation of growth factor-induced migration. Thus, TSP1/hep I stimulation of intermediate adhesion regulates the migratory phenotype of endothelial cells and fibroblasts, suggesting a role for TSP1 in remodeling responses.
http://jcs.biologists.org/content/116/14/2917.long

v
3 integrins in different cellular migration. Using our newly developed micro-volume chemotaxis assay, we developed an improved quantitative method to measure in vitro chemotaxis of smooth muscle or endothelial cells toward different extracellular matrix proteins. The convenience in setup and counting of migrated cells using this method allows for large capacity screening and for various research applications with other cells as well. The signal to noise ratios were in the range of 10/1, along with about 10–20% intra- or inter-assay variabilities. Using this method, we have determined that either vitronectin at 0.4 µg/well or osteopontin at 0.4 µg/well are selective