Angiogenesis, the formation of new blood vessels, is an essential process for tumour progression and is an area of significant therapeutic interest. Different in vitro systems and more complex in vivo systems have been described for the study of tumour angiogenesis. However, there are few human 3D in vitro systems described to date which mimic the cellular heterogeneity and complexity of angiogenesis within the tumour microenvironment. In this study we describe the Minitumour model – a 3 dimensional human spheroid-based system consisting of endothelial cells and fibroblasts in co-culture with the breast cancer cell line MDA-MB-231, for the study of tumour angiogenesis in vitro. After implantation in collagen-I gels, Minitumour spheroids form quantifiable endothelial capillary-like structures. The endothelial cell pre-capillary sprouts are supported by the fibroblasts, which act as mural cells, and their growth is increased by the presence of cancer cells. Characterisation of the Minitumour model using small molecule inhibitors and inhibitory antibodies show that endothelial sprout formation is dependent on growth factors and cytokines known to be important for tumour angiogenesis. The model also shows a response to anti-angiogenic agents similar to previously described in vivo data. We demonstrate that independent manipulation of the different cell types is possible, using common molecular techniques, before incorporation into the model. This aspect of Minitumour spheroid analysis makes this model ideal for high content studies of gene function in individual cell types, allowing for the dissection of their roles in cell-cell interactions. Finally, using this technique, we were able to show the requirement of the metalloproteinase MT1-MMP in endothelial cells and fibroblasts, but not cancer cells, for sprouting angiogenesis.

The migration of leukocytes through the endothelium to sites of infection or inflammation is a key process for the maintenance of physiological defense mechanisms. When this process is dis-regulated and becomes chronic, inflammatory diseases such as arteriosclerosis and arthritis manifest. The steps in leukocyte transmigration (TM) 5 are initiated though activation of the endothelial cells (ECs) by cytokines such as TNF-␣, interleukin-1, and interleukin-6. The selectins initiate the rolling and tethering of leukocytes to the endothelium. This step permits the engagement between ␤ 2 and ␤ 1 integrins with intercellular cell adhesion molecule (ICAM-1) and vascular cell adhesion molecule (VCAM-1) to allow firm adhesion and spreading of leukocytes. ECs express low levels of ICAM-1 and VCAM-1, but cytokine stimulation elevates the expression of these receptors on ECs. The migration of leukocytes through the EC barrier involves platelet endothelial cell adhesion molecule-1 and junctional adhesion molecules. Finally, proteolytic degradation of the basement membrane extracellular matrix by metalloproteinases (MPs) in particular is required to promote extravasation (reviewed in Refs. 1 and 2).Although ICAM-1, VCAM-1, and platelet endothelial cell adhesion molecule-1 play an important role in in vivo and in vitro experimental models of TM (1-4), there is little data on the mechanistic role of the individual adhesion molecules. ICAM-1 comprises of five immunoglobulin-like motifs on the extracellular surface, followed by a stem, a transmembrane domain, and a short cytoplasmic tail (5). The cytoplasmic tail contains three tyrosine residues, two of which become phosphorylated at positions 485 and 474 upon ligation to modulate ICAM-1 function (6, 7). The role of these residues in leukocyteendothelial migration has not yet been defined.Proteolysis is an important step during and after transmigration (8, 9), as degradation of the basement membrane and matrix of the media (of larger vessels) or stroma is required. The zinc-dependent MPs belonging to the metzincin family possess a highly conserved catalytic domain, yet have an enormous * This work was supported in part by National Institutes of Health Grant PO1ES011860, the Jewish Hospital Foundation, Louisville (to S. E. D.), and an American Heart Association Postdoctoral Award (to S. D. S.) from the Ohio Valley Affiliate. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

p300 is a transcriptional cofactor and prototype histone acetyltransferase involved in regulating multiple cellular processes. We generated p300 deficient (p300 À ) cells from the colon carcinoma cell line HCT116 by gene targeting. Comparison of epithelial and mesenchymal proteins in p300 À with parental HCT116 cells showed that a number of genes involved in cell and extracellular matrix interactions, typical of 'epithelial to mesenchyme transition' were differentially regulated at both the RNA and protein level. p300 À cells were found to have aggressive 'cancer' phenotypes, with loss of cell -cell adhesion, defects in cell -matrix adhesion and increased migration through collagen and matrigel. Although migration was shown to be metalloproteinase mediated, these cells actually showed a downregulation or no change in the level of key metalloproteinases, indicating that changes in cellular adhesion properties can be critical for cellular mobility.

In our study, we examined the mechanism by which granulocytemacrophage colony stimulating factor (GM-CSF) regulates angiogenesis using in vitro models. GM-CSF significantly increased precapillary sprout-like formation from endothelial cell spheroids seeded in type-I collagen gels and tubule formation on coculture of endothelial cells with fibroblasts. In both cases, sprout and tubule formation was highly dependent on metalloproteinase activity. Tissue Inhibitor of metalloproteinase (TIMP) profiling in the spheroid and coculture models showed inhibition by TIMP-2 but not by TIMP-1, indicative of activity of membrane-type matrix metalloproteinases (MT-MMPs). GM-CSF induced sprout formation in spheroids was found to be potently inhibited by siRNA specific for MT1-MMP. Subsequent analysis showed that GM-CSF transiently increased MT1-MMP mRNA in endothelial cells in a MEK-dependent mechanism, which led to increased surface levels of MT1-MMP. This was accompanied by an increase in MT1-MMP-dependent degradation of DQ-collagen by lysates of GM-CSF stimulated endothelial cells. GM-CSF did not increase MT1-MMP levels in fibroblasts. The effect of GM-CSF on endothelial cell sprout formation could be mimicked by adenoviral transduction of intact spheroids with virus expressing MT1-MMP, but not by transduction of endothelial cells before spheroid formation, suggesting that upregulation of MT1-MMP must only occur in cells directly involved in tubule formation. ' 2007 Wiley-Liss, Inc.