Acquisition of invasive/metastatic potential through protease expression is an essential event in tumor progression. High levels of components of the plasminogen activation system, including urokinase, but paradoxically also its inhibitor, plasminogen activator inhibitor 1 (PAI1), have been correlated with a poor prognosis for some cancers. We report here that deficient PAI1 expression in host mice prevented local invasion and tumor vascularization of transplanted malignant keratinocytes. When this PAI1 deficiency was circumvented by intravenous injection of a replication-defective adenoviral vector expressing human PAI1, invasion and associated angio-genesis were restored. This experimental evidence demonstrates that host-produced PAI is essential for cancer cell invasion and angiogenesis.Tumor cell invasion and metastatic processes require the coordinated and temporal regulation of a series of adhesive, proteolytic and migratory events 1 . The plasminogen activator (PA)-plasmin proteolytic system has been implicated in these processes. Urokinase-type (uPA) and tissue-type (tPA) plasminogen activators are serine proteases that catalyze the conversion of inactive plasminogen into plasmin, a broadly acting enzyme able to degrade a variety of extracellular matrix proteins and to activate metalloproteinases and growth factors 2,3 . Plasminogen and uPA bind to their specific receptors directing plasmin activity to the migrating tumor cell surface. The activities of PA are directly controlled by specific inhibitors, the PA inhibitors 1 and 2 (PAI1 and PAI2) (ref. 4).Many studies have focused on the role of uPA in cellular invasion and metastasis. Much of the data supporting the role of uPA in these events derives from in vitro and in vivo experiments demonstrating a correlation between uPA expression and cell invasion and metastasis as well as reduction of metastatic potential by using natural or synthetic serine protease inhibitors, neutralizing antibodies to uPA or antisense oligonucleotides 5,6 . PAI1 may also be directly involved in cancer progression. Both tumor cells and capillary endothelial cells express higher levels of PAI1 than other cell types [7][8][9] . Surprisingly, this inhibitor is necessary for optimal invasion of cultured lung cancer cells 10 , and an increasing number of clinical studies have demonstrated that high PAI1 levels indicate a poor prognosis for the survival of patients suffering from a variety of cancers [11][12][13] . However, as PAI1 is an acute-phase reactant 14 , it remains undetermined whether the increased PAI1 levels causally contribute to, or rather are the consequence of, the malignancy.Various observations indicate that the PA system may provide both surface-associated protease activity and an adhesion mechanism for cells through interaction with vitronectin. Deng et al. suggested that the balance between cell adhesion and cell detachment is governed by PAI1 (ref. 15). The PAI1-mediated release of cells attached to vitronectin seems to occur independently of the abili...

The plasminogen (Plg)/plasminogen activator (PA) system plays a key role in cancer progression, presumably via mediating extracellular matrix degradation and tumor cell migration. Consequently, urokinase-type PA (uPA)/plasmin antagonists are currently being developed for suppression of tumor growth and angiogenesis. Paradoxically, however, high levels of PA inhibitor 1 (PAI-1) are predictive of a poor prognosis for survival of patients with cancer. We demonstrated previously that PAI-1 promoted tumor angiogenesis, but by an unresolved mechanism. We anticipated that PAI-1 facilitated endothelial cell migration via its known interaction with vitronectin (VN) and integrins. However, using adenoviral gene transfer of PAI-1 mutants, we observed that PAI-1 promoted tumor angiogenesis, not by interacting with VN, but rather by inhibiting proteolytic activity, suggesting that excessive plasmin proteolysis prevents assembly of tumor vessels. Single deficiency of uPA, tissue-type PA (tPA), uPA receptor, or VN, as well as combined deficiencies of uPA and tPA did not impair tumor angiogenesis, whereas lack of Plg reduced it. Overall, these data indicate that plasmin proteolysis, even though essential, must be tightly controlled during tumor angiogenesis, probably to allow vessel stabilization and maturation. These data provide insights into the clinical paradox whereby PAI-1 promotes tumor progression and warrant against the uncontrolled use of uPA/plasmin antagonists as tumor angiogenesis inhibitors.

Plasminogen activator inhibitor 1 (PAI-1) is believed to control proteolytic activity and cell migration during angiogenesis. We previously demonstrated in vivo that this inhibitor is necessary for optimal tumor invasion and vascularization. We also showed that PAI-1 angiogenic activity is associated with its control of plasminogen activation but not with the regulation of cell-matrix interaction. To dissect the role of the various components of the plasminogen activation system during angiogenesis, we have adapted the aortic ring assay to use vessels from gene-inactivated mice. The single deficiency of tPA, uPA, or uPAR, as well as combined deficiencies of uPA and tPA, did not dramatically affect microvessel formation. Deficiency of plasminogen delayed microvessel outgrowth. Lack of PAI-1 completely abolished angiogenesis, demonstrating its importance in the control of plasmin-mediated proteolysis. Microvessel outgrowth from PAI-1-/- aortic rings could be restored by adding exogenous PAI-1 (wild-type serum or purified recombinant PAI-1). Addition of recombinant PAI-1 led to a bell-shaped angiogenic response clearly showing that PAI-1 is proangiogenic at physiological concentrations and antiangiogenic at higher levels. Using specific PAI-1 mutants, we could demonstrate that PAI-1 promotes angiogenesis at physiological (nanomolar) concentrations through its antiproteolytic activity rather than by interacting with vitronectin.

The matrix metalloproteinases (MMPs) play a key role in normal and pathological angiogenesis by mediating extracellular matrix degradation and/or controlling the biological activity of growth factors, chemokines, and/or cytokines. Specific functions of individual MMPs as anti-or proangiogenic mediators remain to be elucidated. In the present study, we assessed the impact of single or combined MMP deficiencies in in vivo and in vitro models of angiogenesis (malignant keratinocyte transplantation and the aortic ring assay, respectively). MMP-9 was predominantly expressed by neutrophils in tumor transplants, whereas MMP-2 and MMP-3 were stromal. Neither the single deficiency of MMP-2, MMP-3, or MMP-9, nor the combined absence of MMP-9 and MMP-3 did impair tumor invasion and vascularization in vivo. However, there was a striking cooperative effect in double MMP-2:MMP-9-deficient mice as demonstrated by the absence of tumor vascularization and invasion. In contrast, the combined lack of MMP-2 and MMP-9 did not impair the in vitro capillary outgrowth from aortic rings. These results point to the importance of a cross talk between several host cells for the in vivo tumor promoting and angiogenic effects of MMP-2 and MMP-9. Our data demonstrate for the first time in an experimental model that MMP-2 and MMP-9 cooperate in promoting the in vivo invasive and angiogenic phenotype of malignant keratinocytes. Keywordsangiogenesis; tumor invasion; proteolysis; gelatinases; stromal MMP Matrix metalloproteinases (MMPs) are a family of structurally related zinc-and calciumdependent endopeptidases that can degrade extracellular matrix (ECM) components (1,2).

Angiogenesis, a key step in many physiological and pathological processes, involves proteolysis of the extracellular matrix. To study the role of two enzymatic families, serineproteases and matrix metalloproteases in angiogenesis, we have adapted to the mouse, the aortic ring assay initially developed in the rat. The use of deficient mice allowed us to demonstrate that PAI-1 is essential for angiogenesis while the absence of an MMP, MMP-11, did not affect vessel sprouting. We report here that this model is attractive to elucidate the cellular and molecular mechanisms of angiogenesis, to identify, characterise or screen "pro-or anti-angiogenic agents that could be used for the treatment of angiogenesis-dependent diseases. Approaches include using recombinant proteins, synthetic molecules and adenovirus-mediated gene transfer.