Cancer invasion and tissue remodeling: common themes in proteolytic matrix degradation.

Analysis of extracellular matrix degradation systems has led to the insight that in cancer invasion there is often crucial interplay between cancer cells and several types of surrounding non-neoplastic stromal cells. Likewise, in normal tissue remodeling processes, the synthesis of proteolytic components is often distributed between several cell types, and there are strong similarities between neoplastic and non-neoplastic processes in the same tissue. Thus, tissue remodeling events are excellent models for studies of extracellular proteolysis in cancer. This has become even clearer by recent analyses of genetically modified mice.

A plasma kallikrein-dependent plasminogen cascade required for adipocyte differentiation.

Here we show that plasma kallikrein (PKal) mediates a plasminogen (Plg) cascade in adipocyte differentiation. Ecotin, an inhibitor of serine proteases, inhibits cell-shape change, adipocyte-specific gene expression, and lipid accumulation during adipogenesis in culture. Deficiency of Plg, but not of urokinase or tissue-type plasminogen activator, suppresses adipogenesis during differentiation of 3T3-L1 cells and mammary-gland involution. PKal, which is inhibited by ecotin, is required for adipose conversion, Plg activation and 3T3-L1 differentiation. Human plasma lacking PKal does not support differentiation of 3T3-L1 cells. PKal is therefore a physiological regulator that acts in the Plg cascade during adipogenesis. We propose that the Plg cascade fosters adipocyte differentiation by degradation of the fibronectin-rich preadipocyte stromal matrix.

Impaired wound healing in mice with a disrupted plasminogen gene.

Activation of plasminogen (Plg) has been proposed to play a role in proteolytic degradation of extracellular matrices in tissue remodeling events, including wound healing. However, there has been no definitive proof of involvement of Plg in such processes. We now report that healing of skin wounds is severely impaired in mice made deficient in Plg by targeted gene disruption. The results demonstrate that Plg is required for normal repair of skin wounds in mice and support the assumption that it also plays a central role in other disease processes involving extracellular matrix degradation, such as cancer invasion.

Matrix metalloproteinase 9 protects mice from anti-glomerular basement membrane nephritis through its fibrinolytic activity.

Matrix metalloproteinase (MMP)9/gelatinase B is increased in various nephropathies. To investigate its role, we used a genetic approach. Adult MMP9-deficient (MMP9(-/)-) mice showed normal renal histology and function at 3 mo. We investigated the susceptibility of 3-mo-old mice to the accelerated model of anti-glomerular basement membrane nephritis, in which fibrin is an important mediator of glomerular injury and renal impairment. Unexpectedly, nephritis was more severe in MMP9(-/)- than in control mice, as attested by levels of serum creatinine and albuminuria, and the extent of crescents and fibrin deposits. Circulating or deposited immunoglobulin G, interleukin (IL)-1beta, or IL-10 were the same in MMP9(-/-) and MMP9(+/+) mice. However, we found that fibrin is a critical substrate for MMP9, and in its absence fibrin accumulated in the glomeruli. These data indicate that MMP9 is required for a novel protective effect on the development of fibrin-induced glomerular lesions.

Urokinase receptor (CD87) regulates leukocyte recruitment via beta 2 integrins in vivo.

The urokinase receptor (CD87; uPAR) is found in close association with beta 2 integrins on leukocytes. We studied the functional consequence of this association for leukocyte adhesion and migration. In vivo, the beta 2 integrin-dependent recruitment of leukocytes to the inflamed peritoneum of uPAR-deficient mice was significantly reduced as compared with wild-type animals. In vitro, beta 2 integrin-mediated adhesion of leukocytes to endothelium was lost upon removal of uPAR from the leukocyte surface by phosphatidyl-inositol-specific phospholipase C. Leukocyte adhesion was reconstituted when soluble intact uPAR, but not a truncated form lacking the uPA-binding domain, was allowed to reassociate with the cell surface. uPAR ligation with a monoclonal antibody induced adhesion of monocytic cells and neutrophils to vascular endothelium by six- to eightfold, whereas ligation with inactivated uPA significantly reduced cell-to-cell adhesion irrespective of the beta 2 integrin-stimulating pathway. These data indicate that beta 2 integrin-mediated leukocyte-endothelial cell interactions and recruitment to inflamed areas require the presence of uPAR and define a new phenotype for uPAR-deficient mice. Moreover, uPAR ligation differentially modulates leukocyte adhesion to endothelium and provides novel targets for therapeutic strategies in inflammation-related vascular pathologies.

Matrix metalloproteinase 9 and vascular endothelial growth factor are essential for osteoclast recruitment into developing long bones.

Bone development requires the recruitment of osteoclast precursors from surrounding mesenchyme, thereby allowing the key events of bone growth such as marrow cavity formation, capillary invasion, and matrix remodeling. We demonstrate that mice deficient in gelatinase B/matrix metalloproteinase (MMP)-9 exhibit a delay in osteoclast recruitment. Histological analysis and specialized invasion and bone resorption models show that MMP-9 is specifically required for the invasion of osteoclasts and endothelial cells into the discontinuously mineralized hypertrophic cartilage that fills the core of the diaphysis. However, MMPs other than MMP-9 are required for the passage of the cells through unmineralized type I collagen of the nascent bone collar, and play a role in resorption of mineralized matrix. MMP-9 stimulates the solubilization of unmineralized cartilage by MMP-13, a collagenase highly expressed in hypertrophic cartilage before osteoclast invasion. Hypertrophic cartilage also expresses vascular endothelial growth factor (VEGF), which binds to extracellular matrix and is made bioavailable by MMP-9 (Bergers, G., R. Brekken, G. McMahon, T.H. Vu, T. Itoh, K. Tamaki, K. Tanzawa, P. Thorpe, S. Itohara, Z. Werb, and D. Hanahan. 2000. Nat. Cell Biol. 2:737-744). We show that VEGF is a chemoattractant for osteoclasts. Moreover, invasion of osteoclasts into the hypertrophic cartilage requires VEGF because it is inhibited by blocking VEGF function. These observations identify specific actions of MMP-9 and VEGF that are critical for early bone development.

Metastatic behavior of human melanoma cell lines in nude mice correlates with urokinase-type plasminogen activator, its type-1 inhibitor, and urokinase-mediated matrix degradation.

Five out of six human melanoma cell lines tested were able to degrade in vitro a smooth muscle cell extracellular matrix in a plasmin-dependent way. In three of these five cell lines, this process was mediated by tissue-type plasminogen activator (t-PA) and in the other two cell lines by urokinase-type plasminogen activator (u-PA). All melanoma cell lines produced t-PA mRNA and protein, whereas only the two cell lines showing u-PA-mediated matrix degradation produced u-PA mRNA and protein. These latter cell lines also produced plasminogen activator inhibitor type-1 (PAI-1) and type-2 (PAI-2) mRNA and protein. u-PA receptor (u-PA-R) mRNA and binding of radiolabeled u-PA was found in all melanoma cell lines. The metastatic capacity of these cell lines was studied in nude mice. All cell lines were able to develop primary tumors at the subcutaneous inoculation site. The production of plasminogen activators, their inhibitors and urokinase receptor by subcutaneous tumors corresponded with the production by the parental cell lines in vitro. The two u-PA and PAI-1 producing cell lines showed the highest frequency to form spontaneous lung metastases after subcutaneous inoculation, whereas five of the six cell lines formed lung colonies after intravenous inoculation. In conclusion, u-PA mediated matrix degradation in vitro and production of u-PA and PAI-1 by human melanoma cell lines correlated with their ability to form spontaneous lung metastasis in nude mice. No correlation was found with the ability to form lung colonies after intravenous injection. These findings suggest a role for u-PA and PAI-1 in a relatively early stage of melanoma metastasis.

The plasminogen activator inhibitor PAI-1 controls in vivo tumor vascularization by interaction with proteases, not vitronectin. Implications for antiangiogenic strategies.

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.

Systemic administration of anti-urokinase plasminogen activator receptor monoclonal antibodies induces hepatic fibrin deposition in tissue-type plasminogen activator deficient mice.

BACKGROUND: Degradation of extracellular matrix proteins, such as fibrin, is pivotal to tumor invasion. Inhibition of the interaction between urokinase plasminogen activator (u-PA) and its receptor (u-PAR), and hence pro-u-PA activation, is an attractive approach to anti-invasive cancer therapy. A number of inhibitors exist for the human system, but because of species specificity none of these are efficient in mice. We have recently generated an inhibitory monoclonal antibody (mAb) against mouse u-PAR (mR1) by immunization of u-PAR-deficient mice. OBJECTIVES: To evaluate the effect of mR1 in vivo in a physiological setting sensitive to deregulated fibrinolysis, we have administered mR1 systemically and quantitated the effect on liver fibrin accumulation. METHODS: Wild-type and tissue-type plasminogen activator (t-PA) deficient mice were administered with mR1, or control antibody, during 6 weeks. Thereafter, the livers were retrieved and the amount of liver fibrin measured by unbiased morphometrical analysis of immunofluorescence signal. RESULTS: Systemic administration of mR1 caused significantly increased fibrin signal in anti-u-PAR treated t-PA-deficient mice compared to mock-treated, which mimics the phenotype of u-PAR;t-PA double-deficient mice. Fibrin and fibronectin accumulated within the sinusoidal space and was infiltrated by inflammatory cells. Analysis of small and rare hepatic fibrin plaques observed in t-PA-deficient mice showed infiltrating macrophages that, contrary to surrounding Kuppfer cells, expressed u-PAR. CONCLUSION: We show that u-PAR-expressing macrophages are involved in cell-mediated fibrinolysis of liver fibrin deposits, and that the antimouse-u-PAR mAb is effective in vivo and thus suited for studies of the effect of targeting the u-PA/u-PAR interaction in mouse cancer models.

Transforming growth factor beta 1-responsive element: closely associated binding sites for USF and CCAAT-binding transcription factor-nuclear factor I in the type 1 plasminogen activator inhibitor gene.

Transforming growth factor beta (TGF-beta) is the name of a group of closely related polypeptides characterized by a multiplicity of effects, including regulation of extracellular proteolysis and turnover of the extracellular matrix. Its cellular mechanism of action is largely unknown. TGF-beta 1 is a strong and fast inducer of type 1 plasminogen activator inhibitor gene transcription. We have identified a TGF-beta 1-responsive element in the 5'-flanking region of the human type 1 plasminogen activator inhibitor gene and shown that it is functional both in its natural context and when fused to a heterologous nonresponsive promoter. Footprinting and gel retardation experiments showed that two different nuclear factors, present in extracts from both TGF-beta 1-treated and nontreated cells, bind to adjacent sequences contained in the responsive unit. A palindromic sequence binds a trans-acting factor(s) of the CCAAT-binding transcription factor-nuclear factor I family. A partially overlapping dyad symmetry interacts with a second protein that much evidence indicates to be USF. USF is a transactivator belonging to the basic helix-loop-helix family of transcription factors. Mutations which abolish the binding of either CCAAT-binding transcription factor-nuclear factor I or USF result in reduction of transcriptional activation upon exposure to TGF-beta 1, thus showing that both elements of the unit are necessary for the TGF-beta 1 response. We discuss the possible relationship of these findings to the complexity of the TGF-beta action.

Plasminogen activator inhibitor type 1 biosynthesis and mRNA level are increased by dexamethasone in human fibrosarcoma cells.

Dexamethasone increases type 1 plasminogen activator inhibitor (PAI-1) activity released from the human fibrosarcoma cell line HT-1080. We demonstrated that dexamethasone caused about 10-fold increases in the intracellular and extracellular levels of PAI-1 protein, as measured by an enzyme-linked immunosorbent assay, in the rate of PAI-1 biosynthesis, and in the PAI-1 mRNA level. The effects on PAI-1 biosynthesis and mRNA level were detectable within 4 h and were maximal 16 to 24 h after the addition of dexamethasone. Cycloheximide did not inhibit the dexamethasone-induced increases in the capacity of the cells to synthesize PAI-1 and in the PAI-1 mRNA level.

Collagenase-3 expression in breast myofibroblasts as a molecular marker of transition of ductal carcinoma in situ lesions to invasive ductal carcinomas.

Collagenase-3 (matrix metalloproteinase 13; MMP-13), a protease originally identified in breast carcinoma, is characterized by a potent degrading activity against a wide spectrum of extracellular matrix proteins. The aims of this study were to localize and identify the MMP-13-expressing cells in invasive human breast carcinoma and to evaluate the role of MMP-13 in transition to invasive lesions by studying ductal carcinoma in situ (DCIS). We found expression of MMP-13 in stromal fibroblast-like cells in all 21 invasive ductal carcinomas studied and in 4 of 9 invasive lobular carcinomas. In most carcinomas, expression of MMP-13 was limited to small stromal foci in the tumor area. Combined in situ hybridization and immunohistochemistry showed coexpression of alpha-smooth muscle actin immunoreactivity and MMP-13 mRNA in myofibroblasts. In contrast, cytokeratin-positive cancer cells, alpha-smooth muscle actin-positive vascular smooth muscle cells, CD68-positive macrophages, and CD31-positive endothelial cells were all MMP-13 mRNA negative. In situ hybridization for MMP-13 in 17 DCIS lesions revealed expression in 10 cases. Immunohistochemical analysis of all DCIS cases identified microinvasion in 8 of the 17 lesions. Seven of the eight lesions with microinvasion were MMP-13 positive. Further analysis showed that MMP-13 expression was often associated with the microinvasive events. This particular expression pattern was unique for MMP-13 among other MMPs analyzed, including MMP-2, -11, and -14. We conclude that MMP-13 is primarily expressed by myofibroblasts in human breast carcinoma and that expression in DCIS lesions often is associated with microinvasive events. On the basis of these data, we propose that MMP-13 may play an essential role during transition of DCIS lesions to invasive ductal carcinomas.

Reduced metastasis of Polyoma virus middle T antigen-induced mammary cancer in plasminogen-deficient mice.

To investigate the role of plasmin(ogen) in mammary tumor development and progression, plasminogen-deficient mice were crossed with transgenic mice expressing Polyoma middle T antigen under the control of the mouse mammary tumor virus long terminal repeat. Virgin females carrying the Polyoma middle T antigen uniformly developed multiple, bilateral mammary tumors, regardless of the presence or absence of circulating plasminogen. Both the age at which these tumors became palpable and subsequent tumor growth were indistinguishable between plasminogen-deficient mice and plasminogen-expressing littermates. However, plasminogen was found to greatly modify the metastatic potential in this model system; lung metastasis in plasminogen-deficient mice was significantly reduced as compared to littermate controls with respect to frequency of occurrence, total number of metastases, and total metastatic tumor burden. Plasminogen activators, as well as other key factors that govern the conversion of plasminogen to plasmin, were expressed within the mammary tumors, suggesting that the plasminogen/plasmin system may promote metastasis by contributing to tumor-associated extracellular proteolysis. The data provide direct evidence that plasmin(ogen) is a tumor progression factor in PymT-induced mammary cancer, and support the hypothesis that hemostatic factors play an important role in tumor biology.

Urokinase binds to a plasminogen activator inhibitor type-2-like molecule in placental microvillous membranes.

Placental microvillous membranes exhibited saturable binding of urokinase-type plasminogen activator with plateau achieved by 30 min at 4 degrees C and 10 min at 37 degrees C. The binding was essentially irreversible. The capacity was about 8 pmol urokinase per mg membrane protein. Half-maximal displacement of 125I-labelled urokinase was achieved with about 1.0 nM unlabelled urokinase when using 75 micrograms membrane protein/ml. 125I-labelled urokinase did not bind when treated with diisopropylfluorophosphate to block the catalytic activity. Single-chain urokinase (prourokinase), devoid of catalytic activity, did not bind. Catalytically active tissue-type plasminogen activator did compete with 125I-labelled urokinase for binding although less efficiently than urokinase. Binding activity remained in the 100,000 x g pellet after treatment of the membranes with 3 M KCl, alkaline stripping at pH 12 or extraction by the detergent Triton X-100. The binding was essentially blocked by antibodies against plasminogen activator inhibitor-type-2 (PAI-2). Sodium dodecyl sulfate polyacrylamide gel electrophoresis of solubilized membranes with bound 125I-labelled urokinase showed that the urokinase-PAI-2 complexes largely migrated in fractions corresponding to a very large Mr although no clearly defined peaks were observed. It is suggested that PAI-2 occurs in a form anchored to syncytiotrophoblast microvilli, possibly to the cytoskeleton.

Neutrophils express tumor necrosis factor-alpha during mouse skin wound healing.

The expression pattern of tumor necrosis factor-alpha (TNF-alpha) mRNA and protein was examined in vivo in experimental mouse skin wounds by in situ hybridization and immunohistochemistry. TNF-alpha mRNA and protein is detected in a distinct layer of mainly neutrophils subadjacent to the would clot. The layer of TNF-alpha-positive cells extends from the margin of the advancing epithelial outgrowth to the opposing one. By in situ hybridization the TNF-alpha mRNA is detectable 12 h after wounding; the signal peaks after 72 h and remains visible up to at least 120 h after wounding. TNF-alpha mRNA could not be detected in the normal skin or in 5-hour-old wounds. Immunohistochemical staining for TNF-alpha and macrophages on adjacent sections confirms that the main part of TNF-alpha-positive cells are polymorphonuclear neutrophils and shows that most of the cells located just beneath the layer of TNF-alpha-positive neutrophils are macrophages with weak TNF-alpha immunoreactivity. The data reported here show that neutrophils serve as an important source of TNF-alpha during healing of mouse skin wounds. We suggest that this specific expression of TNF-alpha is related to the process of re-epithelialization.

Differential expression of urokinase-type plasminogen activator, its receptor, and inhibitors in mouse skin after exposure to a tumor-promoting phorbol ester.

The cellular distribution of mRNAS for urokinase-type plasminogen activator (uPA), its specific receptor (uPAR), and its inhibitors (PAI-1 and -2) in mouse skin was analyzed by in situ hybridization after topical application of the tumor promoter phorbol 12-myristate 13-acetate. In the epidermis, strong signals for uPA and PAI-1 mRNA were detected 24 h after treatment in the basal and suprabasal epidermal keratinocytes in areas with pronounced hyperproliferation and increased terminal differentiation, and in some hair follicle keratinocytes. After 48 h, both uPAR and PAI-2 mRNAs were expressed in the epidermal layers from the suprabasal keratinocytes up to the differentiating cells beneath the cornified layer and in hair follicle keratinocytes. Induction of PAI-2 mRNA was detected in epidermis as early as 3 h after treatment and remained stable for up to 7 days. In the dermis, 5 h after application of phorbol 12-myristate 13-acetate to the skin, uPA mRNA was detected in fibroblast-like cells below and around the skin muscle, and PAI-1 mRNA was detected in stromal cells located above the skin muscle. After longer exposure to phorbol 12-myristate 13-acetate, the PAI-1 mRNA-expressing stromal cells were located more superficially, apparently moving toward the epidermal layer. After 9 h, most of the PAI-1 mRNA-positive cells were identified as endothelial cells. Up to 24 h after the application of phorbol 12-myristate 13-acetate, the intensity of the signal for both uPA and PAI-1 increased, followed by a gradual decrease for up to 7 days. These results show that in mouse skin treated with a tumor-promoting phorbol ester, the various components of the plasminogen activation system are expressed by both epithelial and stromal cell types, which in dermis and subcutis are located in different places, depending on the time of exposure to the phorbol ester. Our results suggest that urokinase-mediated extracellular proteolysis has diverse functional roles during the early steps of tumor promotion.

Cancer cell expression of urokinase-type plasminogen activator receptor mRNA in squamous cell carcinomas of the skin.

In this study we have used in situ hybridization with radiolabeled antisense RNA probes to examine the expression of mRNA for urokinase-type plasminogen activator and its receptor in histologic samples of squamous cell (n = 7) and basal cell (n = 7) carcinomas of the skin. Messenger RNA for both urokinase-type plasminogen activator and its receptor were expressed in all of the squamous cell carcinomas, but could not be detected in the basal cell carcinomas. In all of the seven squamous cell carcinomas a signal for urokinase-type plasminogen activator receptor mRNA was detected focally in well-differentiated cancer cells surrounding keratinized pearls, and in four specimens urokinase-type plasminogen activator receptor mRNA was in addition expressed by cancer cells at the edge of invasively growing strands of tumor. Urokinase-type plasminogen activator mRNA expression was found in virtually all the cancer cells of the squamous cell carcinomas, and importantly we found, by hybridizations for urokinase-type plasminogen activator and its receptor mRNA on adjacent sections of squamous cell carcinomas, that it was exactly the invading cancer cells that simultaneously expressed both these components required for plasmin-mediated proteolysis at the cell surface. We have previously shown that both urokinase-type plasminogen activator and its receptor mRNA are expressed by the leading-edge keratinocytes in regenerating epidermis during mouse skin wound healing, and that wound healing is impaired in mice made deficient in plasminogen by targeted gene disruption. We propose that there are similarities between the mechanisms of generation and regulation of extracellular proteolysis during skin re-epithelialization and squamous cell carcinoma invasion. The ability of the squamous carcinoma cells to mimic the "invasive" phenotype of re-epithelializing keratinocytes may be one of the factors that make squamous cell carcinomas more aggressive tumors than basal cell carcinomas.

Urokinase- and tissue-type plasminogen activators in keratinocytes during wound reepithelialization in vivo.

Urokinase- and tissue-type plasminogen activators (u-PA and t-PA) were identified immunohistochemically during reepithelialization of mouse and human skin wounds, by means of polyclonal and monoclonal antibodies. In incised mouse skin wounds u-PA immunoreactivity was found in keratinocytes at the edge of the wound after 12 h, and at days 2 to 10 after wounding it was found in virtually all keratinocytes of the epithelial outgrowth that gradually covered the wound. At day 14, the epidermis appeared normal and no u-PA immunoreactivity was detected. t-PA immunoreactivity was found from day 5 to day 10 in some keratinocytes located superficially in the epidermal outgrowths near the edge of the mouse wounds. In 3- and 5-day old human skin wounds, u-PA immunoreactivity was found in keratinocytes in the epithelial outgrowths, whereas no t-PA immunoreactivity was detected. No u-PA and no t-PA immunoreactivity was found in normal mouse and human epidermis. The specificity of the staining was supported by a variety of controls, including absorption of the polyclonal antibodies with highly purified u-PA and t-PA preparations and zymographic analysis of extracts of wound tissue. The function of the plasminogen activators during reepithelialization is discussed and it is suggested that the keratinocytes use plasmin activated by u-PA for dissecting their way through the provisional matrix in the upper part of the granulation tissue.