Binding of the NG2 proteoglycan to kringle domains modulates the functional properties of angiostatin and plasmin(ogen).

Interactions of the developmentally regulated chondroitin sulfate proteoglycan NG2 with human plasminogen and kringle domain-containing plasminogen fragments have been analyzed by solid-phase immunoassays and by surface plasmon resonance. In immunoassays, the core protein of NG2 binds specifically and saturably to plasminogen, which consists of five kringle domains and a serine protease domain, and to angiostatin, which contains plasminogen kringle domains 1-3. Apparent dissociation constants for these interactions range from 12 to 75 nm. Additional evidence for NG2 interaction with kringle domains comes from its binding to plasminogen kringle domain 4 and to miniplasminogen (kringle domain 5 plus the protease domain) with apparent dissociation constants in the 18-71 nm range. Inhibition of plasminogen and angiostatin binding to NG2 by 6-aminohexanoic acid suggests that lysine binding sites are involved in kringle interaction with NG2. The interaction of NG2 with plasminogen and angiostatin has very interesting functional consequences. 1) Soluble NG2 significantly enhances the activation of plasminogen by urokinase type plasminogen activator. 2) The antagonistic effect of angiostatin on endothelial cell proliferation is inhibited by soluble NG2. Both of these effects of NG2 should make the proteoglycan a positive regulator of the cell migration and proliferation required for angiogenesis.

Urokinase-type plasminogen activator binding to its receptor stimulates tumor cell migration by enhancing integrin-mediated signal transduction.

Urokinase-type plasminogen activator (uPA) and its receptor (uPAR) participate in matrix degradation and cell migration by focusing proteolysis and functioning as a signaling ligand/receptor complex. uPAR, anchored by a lipid moiety in the membrane, is thought to require a transmembrane adapter to transduce signals into the cytoplasm. To study uPAR signaling, we transfected the prostate carcinoma cell line LNCaP, which does not express endogenous uPA or uPAR, with a uPAR encoding cDNA, resulting in high-level surface expression. We studied migration of these cells on fibronectin, which is mediated by the integrin alpha5beta1. Ligation of uPAR with uPA or its amino-terminal fragment enhanced haptotactic migration to fibronectin. In cells on fibronectin, but not on poly-l-lysine, ligation of uPAR also resulted in tyrosine phosphorylation of several proteins, including two proteins involved in integrin signaling, focal adhesion kinase and the crk-associated substrate p130(Cas). Furthermore, after uPAR ligation, uPAR was co-immunoprecipitated with beta1 integrins from the detergent-insoluble fraction of cell lysates. Thus, our data suggest that uPAR occupancy results in an interaction between uPAR and integrins and a potentiation of integrin-mediated signaling, which leads to enhanced cell migration.

High-affinity binding of basic fibroblast growth factor and platelet-derived growth factor-AA to the core protein of the NG2 proteoglycan.

NG2 is a transmembrane chondroitin sulfate proteoglycan that is expressed by immature progenitor cells in several developmental lineages and by some types of malignant cells. In vitro studies have suggested that NG2 participates in growth factor activation of the platelet-derived growth factor-alpha receptor. In this study the ability of recombinant NG2 core protein to interact with several different growth factors (epidermal growth factor (EGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor (PDGF)-AA, PDGF-BB, vascular endothelial growth factor (VEGF)165 and transforming growth factor (TGF)-beta1) was investigated using two different assay systems: enzyme-linked immunosorbent assay-type solid-phase binding and an optical biosensor (BIAcore) system. High-affinity binding of bFGF and PDGF-AA to the core protein of NG2 could be demonstrated with both types of assays. Using both the BIAcore software analysis program and nonlinear regression analysis of the solid phase binding data, KD values in the low nanomolar range were obtained for binding of each of these growth factors to NG2. The results further indicate that NG2 contains at least two binding sites for each of these two growth factors. PDGF-BB, TGF-beta1, VEGF, and EGF exhibited little or no binding to NG2 in either type of assay. These data suggest that NG2 can have an important role in organizing and presenting some types of mitogenic growth factors at the cell surface.

Low-density-lipoprotein-receptor-related protein (LRP) interacts with a GTP-binding protein.

The low-density-lipoprotein-receptor-related protein (LRP) binds and internalizes numerous ligands, including lipoproteins, proteinase-inhibitor complexes and others. We have shown previously that LRP-mediated ligand internalization is dependent on cAMP-dependent protein kinase (PKA) activity. Here, we investigated whether ligation of LRP increases the intracellular cAMP level and PKA activity via a stimulatory GTP-binding protein. Treatment of LRP-expressing cell lines with the LRP ligands lactoferrin or urokinase-type plasminogen activator caused a significant elevation in cAMP and stimulated PKA activity in a dose-dependent manner. Addition of the 39 kDa receptor-associated protein (RAP), an antagonist for ligand interactions with LRP, blocked the lactoferrin-induced increase in PKA activity, demonstrating a requirement for ligand binding to LRP. Incubation of cell membrane fractions with lactoferrin increased GTPase activity in a time- and dose-dependent manner, and treatment with LRP ligands suppressed cholera-toxin-mediated ADP-ribosylation of the Gsalpha subunit of a heterotrimeric G-protein. Affinity precipitation of LRP with RAP resulted in co-precipitation of two isoforms of Gsalpha from detergent extracts. We thus conclude that LRP is a signalling receptor that associates directly with a stimulatory heterotrimeric G-protein and activates a downstream PKA-dependent pathway.

Receptor-mediated endocytosis of urokinase-type plasminogen activator is regulated by cAMP-dependent protein kinase.

Internalization of the urokinase-type plasminogen activator (uPA) requires two receptors, the uPA receptor (uPAR) and the low density lipoprotein receptor-related protein (LRP)/alpha2-macroglobulin (alpha2M) receptor. Here, we address whether protein kinases are involved in the internalization of uPA by human melanoma cells. Initially, we found that the internalization of uPA was significantly inhibited by the serine/threonine protein kinase inhibitors staurosporine, K-252a and H-89, but not by the tyrosine kinase inhibitors, genistein and lavendustin A. Internalization of uPA was also inhibited by a pseudosubstrate peptide for cAMP-dependent protein kinase (PKA), but not by a pseudosubstrate peptide for protein kinase C. We confirmed a requirement for PKA-activity and implicated a specific isoform by using an antisense oligonucleotide against the regulatory subunit RI alpha of PKA which suppresses PKA-I activity. Exposure of cells to this oligonucleotide led to a specific, dose-dependent decrease in RI alpha protein and to a significant inhibition in the rate of uPA internalization. We further demonstrate that treatment of melanoma cells with either H-89 or PKA RI alpha antisense oligonucleotides also resulted in a decreased internalization of two other ligands of LRP, activated alpha2M and lactoferrin, indicating that PKA activity is associated with LRP. Finally, we demonstrate that PKA activity is also required for the internalization of transferrin, but not for the internalization of the epidermal growth factor or adenovirus 2, suggesting that in melanoma cells, PKA activity is not generally required for clathrin-mediated endocytosis, but is rather associated with specific internalization receptors.

Inhibition of the interaction of urokinase-type plasminogen activator (uPA) with its receptor (uPAR) by synthetic peptides.

Focusing of the serine protease urokinase-type plasminogen activator (uPA) to the cell surface via interaction with its specific receptor (uPAR, CD87) is an important step for tumor cell invasion and metastasis. The ability of a synthetic peptide derived from the uPAR-binding region of uPA (comprising amino acids 16-32 of uPA; uPA(16-32)) to inhibit binding of fluorescently labeled uPA to uPAR on human promyeloid U937 cells was assessed by quantitative flow cytofluorometric analysis (FACS) and compared to the inhibitory capacities of other synthetic peptides known to interfere with uPA/uPAR-interaction. An about 3000-fold molar excess of uPA(16-32) resulted in 50% inhibition of pro-uPA binding to cell surface-associated uPAR. Using a solid-phase uPA-ligand binding assay employing recombinant soluble uPAR coated to microtiter plates, the minimal binding region of wild-type uPA was determined. The linear peptide uPA(19-31) and its more stable disulfide-bridged cyclic form (cyclo(19,31)uPA(19-31)) displayed uPAR-binding activity whereas other peptides such as uPA(18-30), uPA(20-32) or uPA(20-30) did not react with uPAR. Cyclic peptide derivatives of cyclo(19,31)uPA(19-31) in which certain amino acids were deleted and/or replaced by other amino acids as well as uPAR-derived wild-type peptides did also not inhibit uPA/uPAR-interaction. Therefore, the present investigations identified cyclo(19,31)uPA(19-31) as a potential lead structure for the development of uPA-peptide analogues to block uPA/uPAR-interaction.

Systematic mutational analysis of the receptor-binding region of the human urokinase-type plasminogen activator.

The amino-terminal fragment of human uPA (ATF; amino acids 1-135), which contains the binding site for the uPA receptor (uPAR, CD87) was expressed in the yeast Saccharomyces cerevisiae. Recombinant yeast ATF, modified and extended by an amino-terminal in-frame insertion of a His6 tract, was purified from total protein extracts by nickel chelate affinity chromatography and shown to be functionally active since it efficiently competes with uPA for binding to cell-surface-associated uPAR. The ATF expression plasmid served as a template for the construction of a series of site-directed mutants in order to define those amino acids that are important for binding to uPAR. All mutant ATF proteins but one (deletion of Ser26) were expressed in a stable form (about 20-30 ng/mg total protein) and the binding capacity of each mutant was tested by a uPA-ligand binding assay employing recombinant uPAR immobilized to a microtiter plate. Each of the 11 amino acids of loop B of the binding region of uPA (amino acids 20-30) were individually substituted with alanine. Lys23, Tyr24, Phe25, IIe28, and Trp30 were important determinants for uPAR binding. A systematic alanine scan was also performed with chemically synthesized linear peptides spanning amino acids 14-32 of ATF. Comparable results to those with the yeast ATF mutants were obtained. In a different set of experiments, those amino acids of the uPAR-binding region of uPA that are only conserved between man and baboon but not in other species were altered: whereas substitution of Thr18 by alanine or Asn32 by serine had hardly any effect, replacement of Asn22 by tyrosine and Trp30 by arginine (both positions are strictly conserved in other mammals) led to ATF variants incapable of interacting with human uPAR. Deletion of either Val20, Ser21, Lys23, His29 or Val20 plus Ser21, respectively, also generated non-reactive ATF mutants. Finally, Lys23 in ATF was substituted with certain amino acids: whereas the replacement of Lys23 by alanine, histidine or glutamine generated ATF variants with moderate uPAR-binding activity, the introduction of a negatively charged amino acid (exchange of Lys23 by glutamic acid) completely abolished uPAR-binding activity. The results presented for the ATF mutants and uPA-derived peptides may provide clues necessary to establish the nature of the physical interaction of uPA with its receptor and may help to develop uPA-derived peptide analogues as potential therapeutic agents to block tumor cell-associated uPA/uPAR interaction.

A competitive chromogenic assay to study the functional interaction of urokinase-type plasminogen activator with its receptor.

Urokinase-type plasminogen activator (uPA) converts plasminogen to plasmin which degrades various extracellular matrix components. uPA is focused to the cell surface via binding to a specific receptor (uPAR, also termed CD87). uPAR-bound uPA mediates pericellular proteolysis in a variety of biological processes, e.g. cell migration, tissue remodeling and tumor invasion. We have developed a competitive microtiter plate-based chromogenic assay which allows the analysis of uPA/uPAR interaction. The plates are coated with recombinant uPAR expressed in Chinese hamster ovary (CHO) cells. Proteolytically active uPA (HMW-uPA) is added to the microtiter plate-attached uPAR. The amount of receptor-bound uPA is then determined indirectly via addition of plasminogen, which is activated to plasmin, followed by cleavage of a plasmin-specific chromogenic substrate. Substances interfering with binding of HMW-uPA to uPAR diminish the generation of plasmin, as indicated by a reduction of cleaved chromogenic substrate. This assay was used to analyze the inhibitory capacity of a variety of proteins and peptides, respectively, on the uPA/uPAR interaction: i) uPAR and uPAR-variants expressed in CHO cells, yeast or E. coli, ii) the aminoterminal fragment (ATF) of human uPA or yeast recombinant pro-uPA, iii) synthetic peptides derived from the sequence of the uPAR-binding region of uPA, and iv) antibodies directed against uPAR. This assay may be helpful in identifying uPA and uPAR analogues or antagonists which efficiently block uPA/uPAR interaction.

Prognostic value of the cysteine proteases cathepsins B and cathepsin L in human breast cancer.

The lysosomal cysteine proteases cathepsin B and cathepsin L have been implicated in tumor spread and metastasis. To evaluate the prognostic impact of these proteases for disease-free survival and overall survival in breast cancer, the antigen content of cathepsin B and cathepsin L was determined using ELISA in tumor cytosol fractions of 167 breast cancer patients and in cytosols of 29 benign breast tissue specimens. Median values of 856 ng versus 76 ng cathepsin B/mg protein and of 428 ng versus 56 ng cathepsin L/mg protein were found in tumor versus benign cytosol fractions. A positive correlation between cathepsin B and cathepsin L (r = 0.32, P = 0.0000, Spearman test) was found. Cathepsin L was inversely correlated to hormone receptor status (P = 0.0014, Mann-Whitney U test) and to the presence of tumor necrosis (P = 0.009, Mann-Whitney U test). There were no correlations of cathepsin B or cathepsin L to tumor size, axillary lymph node status, age, menopausal status, tumor grading, and vessel invasion. To perform univariate analyses of disease-free survival, optimal cutoff points were determined by isotonic regression and classification and regression trees analysis. Patients with a high content of cathepsin B (>1092 ng/mg protein) or cathepsin L (>376 ng/mg protein) in their primary tumors had a statistically significantly higher risk of recurrence than patients with a low content of cathepsin B or cathepsin L (5-year disease-free survival: cathepsin B, 70% versus 52%, P = 0.04; cathepsin L, 83% versus 52%, P = 0.0002). Median follow-up was 39 (range, 6-73) months. Multivariate analysis for disease-free survival showed that cathepsin L is a strong and independent prognostic factor with a prognostic impact comparable to that of axillary lymph node status and grading. We conclude that both cathepsin B and cathepsin L may serve as prognostic factors for tumor recurrence in human breast cancer. These data underline the significance of tumor-associated proteolysis for invasion and metastasis.

Expression of the human urokinase-type plasminogen activator receptor in E. coli and Chinese hamster ovary cells: purification of the recombinant proteins and generation of polyclonal antibodies in chicken.

The receptor for urokinase-type plasminogen activator (uPAR) may contribute to the invasive and metastatic capacity of tumor cells by focusing the serine protease urokinase-type plasminogen activator (uPA) to the cell surface. uPA activates plasminogen to plasmin which in turn degrades extracellular matrix proteins or activates other proteases. Mature uPAR is a heavily glycosylated protein of about 284 amino acids attached to the plasma membrane via a glycosyl-phosphatidylinositol (GPI) anchor. A set of different polyclonal uPAR antibodies has been generated in order to investigate the role of uPAR in tumor spreading in more detail. For this purpose, uPAR (lacking the GPI anchor) was expressed in E. coli and Chinese hamster ovary (CHO) cells. Recombinant uPAR from E. coli (corresponding to amino acids 1-284 of human uPAR) was expressed with an N-terminal histidine-tag insertion and purified by nickel chelate affinity chromatography. Soluble uPAR, synthesized by CHO cells (corresponding to amino acids 1-277 of human uPAR), was isolated by ligand (uPA) affinity chromatography. Expression in E. coli led to a nonglycosylated form of uPAR, whereas uPAR produced by CHO cells seemed to be glycosylated to a similar extent as the naturally occurring human form of uPAR (as analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis). Prior to immunization the N-termini of the recombinant uPAR variants were determined by amino acid sequence analysis. Polyclonal antibodies were generated in chickens and purified from egg yolk. The reaction patterns of these antibodies were analyzed by Western blot analyses and flow cytofluorometry.

Saturation of tumour cell surface receptors for urokinase-type plasminogen activator by amino-terminal fragment and subsequent effect on reconstituted basement membranes invasion.

Single-chain urokinase-type plasminogen activator (pro-uPA) is bound to a specific surface receptor on ovarian cancer HOC-I cells that is incompletely saturated. Saturation of uncovered receptors by uPA polypeptides with intact amino-terminal fragment (ATF) derived from pro-uPA by limited proteolysis (human leucocyte elastase [HLE] or V8 protease) has been studied. HOC-I cells preferentially invaded reconstituted basement membranes in a time- and plasminogen-dependent manner. This process was inhibitable by preincubation with uPA polypeptides in the medium at levels which suggested that complete saturation of cell surface uPA receptors occurred. This result indicates that occupation of uPA receptors by enzymatically inactive uPA fragments or prevention of rebinding of pro-uPA synthesised by tumour cells to the receptors specifically reduces the invasion of the tumour cells through basement membranes in vitro.

Effective activation of the proenzyme form of the urokinase-type plasminogen activator (pro-uPA) by the cysteine protease cathepsin L.

Increased levels of both the cysteine protease, cathepsin L, and the serine protease, uPA (urokinase-type plasminogen activator), are present in solid tumors and are correlated with malignancy. uPA is released by tumor cells as an inactive single-chain proenzyme (pro-uPA) which has to be activated by proteolytic cleavage. We analyzed in detail the action of the cysteine protease, cathepsin L, on recombinant human pro-uPA. Enzymatic assays, SDS-PAGE and Western blot analysis revealed that cathepsin L is a potent activator of pro-uPA. As determined by N-terminal amino acid sequence analysis, activation of pro-uPA by cathepsin L is achieved by cleavage of the Lys158-Ile159 peptide bond, a common activation site of serine proteases such as plasmin and kallikrein. Similar to cathepsin B (Kobayashi et al., J. Biol. Chem. (1991) 266, 5147-5152) cleavage of pro-uPA by cathepsin L was most effective at acidic pH (molar ratio of cathepsin L to pro-uPA of 1:2,000). Nevertheless, even at pH 7.0, pro-uPA was activated by cathepsin L, although a 10-fold higher concentration of cathepsin L was required. As tumor cells may produce both pro-uPA and cathepsin L, implications for the activation of tumor cell-derived pro-uPA by cathepsin L may be considered. Different pathways of activation of pro-uPA in tumor tissues may coexist: (i) autocatalytic intrinsic activation of pro-uPA; (ii) activation by serine proteases (plasmin, kallikrein, Factor XIIa); and (iii) activation by cysteine proteases (cathepsin B and L).

Fluorescent probes as tools to assess the receptor for the urokinase-type plasminogen activator on tumor cells.

Flow cytofluorometric protocols (FACScan) are described for the rapid and quantitative real-time analysis of binding of FITC-pro-u-PA to cell surface receptors (u-PAR) on living, resting, and also on PMA-stimulated human monocytic U937 cells. Binding of pro-u-PA was visualized by CLSM. This fairly new technique is superior over conventional fluorescence microscopy and is an alternative to electron microscopic approaches. Both flow cytofluorometry and confocal laser scanning microscopy allow the analysis quantitatively and with high-sensitivity binding of FITC-pro-u-PA to single suspended or adherent cells. By CLSM u-PA/u-PAR were found to be located in heterogeneously distributed discrete patches at the cell surface on U937 and not inside the cell. This is in agreement with previous studies by Hansen et al, who applied radioiodinated u-PA and electron microscopy to locate u-PAR on microvilli of fixed U937 cells. By flow cytofluorometry, it was possible to quantify the time-dependent and temperature-dependent binding of FITC-pro-u-PA to living single U937. Apparent saturation of u-PAR was achieved at 5 nM FITC-pro-u-PA for both nonstimulated and PMA-stimulated U937 cells. Half saturation of u-PAR was also determined. Nonstimulated U937 was 0.7 nM, and PMA-stimulated U937 was 1.1 nM of FITC-pro-u-PA. This increase in half-saturation concentration in PMA-stimulated cells is paralleled by a steep increase in binding sites (3.6-fold). The use of fluoresceinated reference beads is recommended to verify changes in affinity and binding sites. Using CLSM or flow cytofluorometry, it is also possible to study the structure relationship of u-PA/u-PAR in the presence of competitive binding analogues or inhibitors. Fluorescence techniques will also permit the identification of u-PAR-positive cells in blood, ascitic fluid, or biopsies obtained from cancer patients.

Cathepsin B efficiently activates the soluble and the tumor cell receptor-bound form of the proenzyme urokinase-type plasminogen activator (Pro-uPA).

Action of purified human cathepsin B on recombinant single-chain urokinase-type plasminogen activator (pro-uPA) generated enzymatically active two-chain uPA (HMW-uPA), which was indistinguishable by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blot from plasmin-generated HMW-uPA and from elastase- or thrombin-generated inactive two-chain urokinase-type plasminogen activator. Preincubation of cathepsin B with E-64 (transepoxysuccinyl-L-leucylamino- (4-guanidino)butane, a potent inhibitor for cathepsin B) prior to the addition of pro-uPA prevented the activation of pro-uPA. The cleavage site within the cathepsin B-treated urokinase-type plasminogen activator (uPA) molecule, determined by N-terminal amino acid sequence analysis, is located between Lys158 and Ile159. Pro-uPA is cleaved by cathepsin B at the same peptide bond that is cleaved by plasmin or kallikrein. Binding of cathepsin B-activated pro-uPA to the uPA receptor on U937 cells did not differ from that of enzymatically inactive pro-uPA, indicating an intact receptor-binding region within the growth factor-like domain of the cathepsin B-treated uPA molecule. Not only soluble but also tumor cell receptor-bound pro-uPA could be efficiently cleaved by cathepsin B to generate enzymatically active two-chain uPA. Thus, cathepsin B can substitute for plasmin in the proteolytic activation of pro-uPA to enzymatically active HMW-uPA. In contrast, no significant activation of pro-uPA by cathepsin D was observed. As tumor cells may produce both pro-uPA and cathepsin B, implications for the activation of tumor cell-derived pro-uPA by cellular proteases may be considered.

Biological and clinical relevance of the urokinase-type plasminogen activator (uPA) in breast cancer.

Tumor cell invasion and metastasis is a multifactorial process, which at each step may require the action of proteolytic enzymes such as collagenases, cathepsins, plasmin, or plasminogen activators. An enzymatically inactive proenzyme form of the urokinase-type plasminogen activator (pro-uPA) is secreted by tumor cells which may be converted to an enzymatically active two-chain uPA-molecule (HMW-uPA) by plasmin-like enzymes. Action of proteases on pro-uPA may generate the enzymatically active or inactive high-molecular-weight form of uPA (HMW-uPA). Some proteases (plasmin, cathepsin B and L, kallikrein, trypsin or thermolysin) activate pro-uPA by cleaving the peptide bond Lys158 and IIe159. Other proteases (elastase, thrombin) cleave pro-uPA at different positions to yield enzymatically inactive HMW-uPA. HMW-uPA may be split into the enzymatically active LMW-uPA and the enzymatically inactive ATF (amino terminal fragment). ATF may be cleaved between peptide sequence 20 and 40 within the receptor binding domain of uPA (GFD). Such impaired ATF does not bind to uPA-receptors. Action of the bacterial endoproteinase Asp-N from Pseudomonas fragi mutant on pro-uPA or HMW-uPA, however, generates intact ATF which efficiently competes for binding of HMW-uPA or pro-uPA to receptors on tumor cells. High uPA-antigen content (pro-uPA, HMW-uPA, or LMW-uPA) in breast cancer tissue (not in plasma) indicates an elevated risk for the patient of recurrences and shorter overall survival. Thus pro-uPA/uPA-antigen content in breast cancer tissue serves as an independent prognostic parameter for the outcome of the disease. Cathepsin D is also an independent prognostic factor for recurrences and overall survival. High content of cathepsin D in breast cancer tumors is, however, not correlated with elevated levels of pro-uPA/uPA indicating that synthesis and release of cathepsin D and pro-uPA/uPA are independent events.