Urokinase-catalyzed plasminogen activation at the monocyte/macrophage cell surface: a localized and regulated proteolytic system

Molecular imaging of the urokinase plasminogen activator receptor: opportunities beyond cancer

The urokinase plasminogen activator receptor (uPAR) plays a multifaceted role in almost any process where migration of cells and tissue-remodeling is involved such as inflammation, but also in diseases as arthritis and cancer. Normally, uPAR is absent in healthy tissues. By its carefully orchestrated interaction with the protease urokinase plasminogen activator and its inhibitor (plasminogen activator inhibitor-1), uPAR localizes a cascade of proteolytic activities, enabling (patho)physiologic cell migration. Moreover, via the interaction with a broad range of cell membrane proteins, like vitronectin and various integrins, uPAR plays a significant, but not yet completely understood, role in differentiation and proliferation of cells, affecting also disease progression. The implications of these processes, either for diagnostics or therapeutics, have received much attention in oncology, but only limited beyond. Nonetheless, the role of uPAR in different diseases provides ample opportunity to exploit new applications for targeting. Especially in the fields of oncology, cardiology, rheumatology, neurology, and infectious diseases, uPAR-targeted molecular imaging could offer insights for new directions in diagnosis, surveillance, or treatment options.

Co-localization of fibrinolytic activators and inhibitors with macrophages in atherosclerotic vessels

The plasmin system is involved in hemostasis and tissue remodelling. The accumulation of plasminogen activators and their inhibitors in atherosclerotic lesions may be due to invasion of inflammatory cells in the vessel wall. High concentrations of macrophages are associated with increased risk of atherosclerotic plaque rupture. By immunohistochemistry on circumferential serial sections of atherosclerotic and healthy vessels the morphological association of plasminogen activators and inhibitors with macrophages was studied. Urokinase plasminogen activator (u-PA), plasminogen activator inhibitor type 2 (PAI-2), and macrophages were mainly expressed within plaques while tissue plasminogen activator (t-PA) and plasminogen activator inhibitor type 1 (PAI-1) were also expressed outside plaque lesions. Computer assisted image analysis on diseased vessels showed that regulatory proteins of the fibrinolytic system were found more often in areas positive for macrophages than in other parts of the sections (p < 0.001). u-PA was significantly more defined to areas positive for macrophages than tissue plasminogen activator (t-PA) and plasminogen activator inhibitor type 1 (PAI-1) (p < 0.05). Similarly, PAI-2 expression was more associated with macrophage distribution than PAI-1 (p < 0.05). Tumor necrosis factor α (TNFα), an inflammatoric mediator of macrophages, had the same levels of co-localization with macrophages as u-PA and PAI-2. These results suggest that u-PA and PAI-2 might be key factors for inflammatory regulation of plasmin mediated proteolysis in the vessel walls.

The omptins of Yersinia pestis and Salmonella enterica cleave the reactive center loop of plasminogen activator inhibitor 1

Plasminogen activator inhibitor 1 (PAI-1) is a serine protease inhibitor (serpin) and a key molecule that regulates fibrinolysis by inactivating human plasminogen activators. Here we show that two important human pathogens, the plague bacterium Yersinia pestis and the enteropathogen Salmonella enterica serovar Typhimurium, inactivate PAI-1 by cleaving the R346-M347 bait peptide bond in the reactive center loop. No cleavage of PAI-1 was detected with Yersinia pseudotuberculosis, an oral/fecal pathogen from which Y. pestis has evolved, or with Escherichia coli. The cleavage and inactivation of PAI-1 were mediated by the outer membrane proteases plasminogen activator Pla of Y. pestis and PgtE protease of S. enterica, which belong to the omptin family of transmembrane endopeptidases identified in Gram-negative bacteria. Cleavage of PAI-1 was also detected with the omptins Epo of Erwinia pyrifoliae and Kop of Klebsiella pneumoniae, which both belong to the same omptin subfamily as Pla and PgtE, whereas no cleavage of PAI-1 was detected with omptins of Shigella flexneri or E. coli or the Yersinia chromosomal omptins, which belong to other omptin subfamilies. The results reveal a novel serpinolytic mechanism by which enterobacterial species expressing omptins of the Pla subfamily bypass normal control of host proteolysis.

Microglia and the urokinase plasminogen activator receptor/uPA system in innate brain inflammation

The urokinase plasminogen activator (uPA) receptor (uPAR) is a GPI-linked cell surface protein that facilitates focused plasmin proteolytic activity at the cell surface. uPAR has been detected in macrophages infiltrating the central nervous system (CNS) and soluble uPAR has been detected in the cerebrospinal fluid during a number of CNS pathologies. However, its expression by resident microglial cells in vivo remains uncertain. In this work, we aimed to elucidate the murine CNS expression of uPAR and uPA as well as that of tissue plasminogen activator and plasminogen activator inhibitor 1 (PAI-1) during insults generating distinct and well-characterized inflammatory responses; acute intracerebral lipopolysaccharide (LPS), acute kainate-induced neurodegeneration, and chronic neurodegeneration induced by prion disease inoculation. All three insults induced marked expression of uPAR at both mRNA and protein level compared to controls (naïve, saline, or control inoculum-injected). uPAR expression was microglial in all cases. Conversely, uPA transcription and activity was only markedly increased during chronic neurodegeneration. Dissociation of uPA and uPAR levels in acute challenges is suggestive of additional proteolysis-independent roles for uPAR. PAI-1 was most highly expressed upon LPS challenge, whereas tissue plasminogen activator mRNA was constitutively present and less responsive to all insults studied. These data are novel and suggest much wider involvement of the uPAR/uPA system in CNS function and pathology than previously supposed.

VEGF-induced paracellular permeability in cultured endothelial cells involves urokinase and its receptor

Vascular endothelial growth factor/vascular permeability factor (VEGF) has been implicated in blood/tissue barrier dysfunctions associated with pathological angiogenesis, but the mechanisms of VEGF-induced permeability increase are poorly understood. Here, the role of VEGF-induced extracellular proteolytic activities on the endothelial cell permeability increase is evaluated. Confluent monolayers of bovine retinal microvascular endothelial (BRE) cells grown on porous membrane were treated with VEGF or urokinase plasminogen activator (uPA), and permeability changes were analyzed. uPA-induced permeability was rapid and sustained, but VEGF-induced permeability showed a biphasic pattern: a rapid and transient phase (1-2 h) followed by delayed and sustained phase (6-24 h). The delayed, but not the early phase of VEGF-induced permeability, was blocked by anti-uPA or anti-uPAR (uPA receptor) antibodies and was accompanied by reduced transendothelial electrical resistance, indicating the paracellular route of permeability. Confocal microscopy and Western blotting showed that VEGF treatment increased free cytosolic beta-catenin, which was followed by beta-catenin nuclear translocation, upregulation of uPAR, and downregulation of occludin. Membrane-bound occludin was released immediately after uPA treatment, but with a long delay after VEGF treatment, suggesting a requirement for uPAR gene expression. In conclusion, VEGF induces a sustained paracellular permeability in capillary endothelial cells that is mediated by activation of the uPA/uPAR system.

Urokinase-type plasminogen activator and its receptor synergize to promote pathogenic proteolysis

Urokinase-type plasminogen activator (uPA) is a potent catalyst of extracellular proteolysis, which also binds to a high-affinity plasma membrane receptor (uPAR). Binding of uPA may influence pericellular proteolysis and/or activate intracellular signal transduction. Transgenic mice overexpressing either uPA or uPAR in basal epidermis and hair follicles had no detectable cutaneous alterations. In contrast, bi-transgenic mice overexpressing both uPA and uPAR, obtained by crossing the two transgenic lines, developed extensive alopecia induced by involution of hair follicles, epidermal thickening and sub-epidermal blisters. The phenotype was due to uPA catalytic activity since combined overexpression of uPAR and uPAR-binding but catalytically inactive uPA in the same tissue was not detrimental in another bi-transgenic line. It was accompanied by increased plasmin-generating capacity, up-regulation and activation of matrix metalloproteinases type-2 and -9, and cleavage of uPAR. Thus, combined overexpression of uPA and uPAR acts in synergy to promote pathogenic extracellular proteolysis.

Topological localization of plasminogen activator inhibitor type 2

BACKGROUND

Plasminogen activator inhibitor type 2 (PAI-2) is a member of the serine protease inhibitor (SERPIN) superfamily and forms stable complexes with urokinase type plasminogen activator (uPA). uPA can be found on the cell surface attached to its specific receptor (uPAR), allowing for controlled degradation of the extracellular matrix by the activation of plasminogen into plasmin. The aim of this study was to evaluate if PAI-2 could also be detected on the cell surface, providing a means of regulating the activity of cell surface uPA.

METHODS

Intact or permeabilized cell lines or human peripheral blood leukocytes were assayed by flow cytometry for cell surface uPA or PAI-2. Plasma membrane-enriched preparations prepared from Jurkat, HaCaT, THP-1, U937, or MM6 cells were assayed by enzyme-linked immunosorbent assay (ELISA) or Western blotting for PAI-2 antigen.

RESULTS

By flow cytometry, cell surface PAI-2 was not detected on monocytes from human peripheral blood, MM6, or HaCaT cells. Cell surface PAI-2 was only detected very weakly on the surface of U937 cells. In contrast, PAI-2 could be detected in all of these cells when fixed and permeabilized. By ELISA, PAI-2 was very abundant in the cytosol-enriched preparations of U937, MM6, and HaCaT cells, but was present in lower amounts in the plasma membrane-enriched preparations. By Western blotting, monomeric nonglycosylated PAI-2, but not uPA/PAI-2 complexes, could be detected in the cytosol and plasma membrane-enriched preparations.

CONCLUSIONS

These results indicate that PAI-2 cannot be detected on the surface of PAI-2-expressing cells, and confirm that PAI-2 is predominantly a cytosolic protein.

Activation of rat alveolar macrophage-derived latent transforming growth factor beta-1 by plasmin requires interaction with thrombospondin-1 and its cell surface receptor, CD36

Transforming growth factor-beta-1 (TGF-beta1) is secreted by cells in a latent form (L-TGF-beta1) noncovalently bound to a latency-associated peptide. Activated alveolar macrophages obtained from rat lungs after bleomycin-induced pulmonary injury released increased amounts of active TGF-beta1 as well as plasmin, a protease, and thrombospondin-1 (TSP-1), a trimeric glycoprotein. Previously we had demonstrated that plasmin was critical to the activation of L-TGF- beta1. In the present study we demonstrated that TSP-1 is also important for the activation of L-TGF- beta1 because the activation can be inhibited by anti-TSP-1 monoclonal antibody. Proteins obtained from alveolar macrophage cell lysates immunoprecipitated with antibodies specific for TSP-1 were identified on immunoblots as LAP and TGF-beta1, indicating that TSP-1/L-TGF-beta1 complexes are present on alveolar macrophages. However, in the presence of plasmin both latency-associated peptide and TGF-beta1 were decreased in the same cell lysates, indicating that L-TGF-beta1 associated with TSP-1 is released by plasmin. Using immunofluorescence and antibodies to TGF-beta1 and CD36, a receptor for TSP-1, there was colocalization of TGF-beta1 with CD36. Because TSP-1 but not TGF-beta1 is a natural ligand for CD36, these findings suggest that the L-TGF-beta1 in a complex with TSP-1 localizes to the macrophage cell surface when TSP-1 interacts with its receptor, CD36. Furthermore, the association of TSP-1/L-TGF-beta1 complex with CD36 is necessary to the activation of L-TGF-beta1 because antibodies to CD36 prevent the colocalization of TGF-beta1 with CD36 as observed by immunofluorescence and inhibit activation of the L-TGF-beta1 by explanted alveolar macrophages. These findings suggest that activation of L-TGF-beta1 by plasmin occurs at the cell surface of activated alveolar macrophages and requires a TSP-1/CD36 interaction.

Heterogeneous distribution of macrophages, tumour necrosis factor alpha, tissue factor and fibrinolytic regulators in atherosclerotic vessels

OBJECTIVE

To assess the representativity of small vascular samples with regard to the expression of six important factors in atherosclerotic vessels.

MATERIALS

Circumferential frozen sections of iliac and femoral arteries (n = 9) from patients undergoing reconstructive bypass surgery.

METHODS

Immunohistochemistry with antibodies against macrophages (CD68), tumour necrosis factor alpha, tissue factor, tissue plasminogen activator, urokinase plasminogen activator and plasminogen activator inhibitor type 1. The distribution of these antigens was characterised by computer assisted image analysis. Antigen positive area in randomly chosen samples of varying size was compared with antigen positive area in the whole vessel transections.

RESULTS

Marked heterogeneity was found with respect to the expression of these factors in atherosclerotic vessels. The representativity of samples was highly dependent on the size of the samples. PAI-1 was more evenly expressed compared to the other antigens.

CONCLUSIONS

Estimates of the expression of these factors based on small samples from atherosclerotic arteries are unreliable.

Role of urokinase in the fibrogenic response of the lung to mineral particles

The lung plasminogen activator (PA) response was examined in four different models of particle-induced pulmonary lesions in NMRI mice (single intratracheal administration, 0.75 to 5 mg/mouse). Sequential changes in cellular (total and differential counts) and biochemical markers of alveolitis (lactate dehydrogenase [LDH], total proteins) were monitored in bronchoalveolar fluid (BALF) and the fibrotic lung response was assessed histologically. An intense but spontaneously resolving alveolitis was produced by manganese dioxide (MnO2) and a fibrosing alveolitis was elicited by crystalline silica (DQ12). Minimal and noninflammatory responses were obtained after instillation of titanium dioxide (TiO2) and tungsten carbide (WC), respectively. The comparison between the resolving and the fibrosing alveolitis model was especially taken into consideration in an attempt to identify fibrinolytic changes associated with the development of fibrosis. At the alveolitis stage, similarly increased BALF PA activities were measured in both the resolving and the fibrosing alveolitis models whereas only slight and no PA modifications were noted after administration of TiO2 and WC, respectively. Persistently (up to 120 d) increased BALF PA activity was selectively associated with the progression to fibrosis (DQ12), suggesting that PA is involved in the fibrotic process. ELISA measurements demonstrated that the changes in BALF PA activity were exclusively related to changes in urokinase (uPA), not tissue-type PA. A rapid and persisting (up to Day 30) upregulation of cell-associated PA activity occurred after DQ12, MnO2, and TiO2 treatment only. Cellular PA activity was however significantly higher in fibrogenic inflammatory cells recovered from DQ12 than from MnO2-treated mice suggesting that the intensity of cellular PA upregulation may represent an early indicator of the progression to fibrosis. The implication of urokinase in the pathogenesis of silica-induced fibrosis was demonstrated by the use of a uPA knockout mice. The acceleration of the fibrotic process in uPA-deficient compared with the wild type animals demonstrated the contribution of uPA to limit the fibrotic process.

The human ENO1 gene product (recombinant human alpha-enolase) displays characteristics required for a plasminogen binding protein

Plasminogen binds with low affinity in a lysine-dependent manner to many cell types. Previously, a 54 kDa plasminogen receptor found on the surface of U-937 cells was identified as an alpha-enolase-like molecule. The aims of this study were to determine whether recombinant alpha-enolase (r-alpha-enolase), encoded by ENO1, was a plasminogen binding protein and to generate polyclonal antibodies against this antigen. Plasminogen specifically bound r-alpha-enolase with a Kd 1.9 microM and approached saturation at 10 microM. Lysine-dependent plasminogen binding to r-alpha-enolase was demonstrated by a greater than 80% inhibition of binding by the lysine analogues epsilon-amino caproic acid and tranexamic acid, whilst only 14% inhibition occurred with the arginine analogue benzamidine. Removal of the C-terminal lysine residue of r-alpha-enolase with carboxy-peptidase B significantly reduced its plasminogen binding capacity, suggesting that binding required C-terminal lysine residue of r-alpha-enolase. Binding to r-alpha-enolase enhanced the activation rate of plasminogen by urokinase but prevented alpha 2-antiplasmin from binding plasminogen. Taken together, these data suggest that the gene product of human ENO1 encodes an authentic plasminogen binding protein.

Urokinase-type plasminogen activator is effective in fibrin clearance in the absence of its receptor or tissue-type plasminogen activator

The availability of gene-targeted mice deficient in the urokinase-type plasminogen activator (uPA), urokinase receptor (uPAR), tissue-type plasminogen activator (tPA), and plasminogen permits a critical, genetic-based analysis of the physiological and pathological roles of the two mammalian plasminogen activators. We report a comparative study of animals with individual and combined deficits in uPAR and tPA and show that these proteins are complementary fibrinolytic factors in mice. Sinusoidal fibrin deposits are found within the livers of nearly all adult mice examined with a dual deficiency in uPAR and tPA, whereas fibrin deposits are never found in livers collected from animals lacking uPAR and rarely detected in animals lacking tPA alone. This is the first demonstration that uPAR has a physiological role in fibrinolysis. However, uPAR-/-/tPA-/- mice do not develop the pervasive, multi-organ fibrin deposits, severe tissue damage, reduced fertility, and high morbidity and mortality observed in mice with a combined deficiency in tPA and the uPAR ligand, uPA. Furthermore, uPAR-/-/tPA-/- mice do not exhibit the profound impairment in wound repair seen in uPA-/-/tPA-/- mice when they are challenged with a full-thickness skin incision. These results indicate that plasminogen activation focused at the cell surface by uPAR is important in fibrin surveillance in the liver, but that uPA supplies sufficient fibrinolytic potential to clear fibrin deposits from most tissues and support wound healing without the benefit of either uPAR or tPA.

The urokinase receptor (CD87) facilitates CD11b/CD18-mediated adhesion of human monocytes

Urokinase receptors (uPAR; CD87) from complexes with complement receptor 3 (CR3) (CD11b/CD18), a beta2 integrin. In this study, we sought to determine if this association modulates the adhesive function of CR3. Both CR3 and uPAR concentrate at the ventral surface of fibrinogen-adherent human monocytes, and CR3-uPAR coupling increases substantially upon adhesion to fibrinogen. Pretreatment with anti-uPAR monoclonal antibody reduced adhesion to CR3 counterligands (fibrinogen and keyhole limpet hemocyanin) by 50%, but did not affect adhesion to fibronectin, a beta1 integrin counterligand. Antisense (AS) oligonucleotides were used to determine if selectively suppressing uPAR expression also modulates CR3 adhesive function. AS-uPAR oligo reduced CR3-dependent adhesion by 43+/-9% (P<0.01), but did not affect CR3-independent adhesion. To determine if the effects of uPAR are mediated through its ligand, monocytes were pre-treated with AS oligo to block uPA expression. Unlike the effects of blocking uPAR expression, AS-uPA oligo increased adhesion by 46% (P<0.005), and exogenous intact uPA, but not uPA fragments, reversed this effect. We conclude that complex formation with uPAR facilitates the adhesive functions of CR3. This function of uPAR is not dependent upon its occupancy with uPA, which negatively influences adhesion.

The urokinase/urokinase-receptor system and cancer invasion

u-PA binds with high affinity to its specific GPI-anchored receptor on the cell surface. The binding has at least two important consequences: (1) it enhances the rate of plasminogen activation on the cell surface; and (2) it focuses the u-PA proteolytic activity at the leading front of migrating cells. Several recent findings suggest that surface-bound u-PA is essential for the invasive ability of tumour cells, even if a picture is emerging indicating a concerted action with other proteases, like collagenases and cathepsin B (Kobayashi et al, 1992; Ossowski, 1992; Schmitt et al, 1992; (Danø et al, 1994). In some tumours, e.g. colon, breast and skin cancer, in situ hybridization studies have given an insight into the u-PA/u-PAR tumour biology showing a complex interplay between stromal and cancer cells Danø et al, 1994). u-PA, u-PAR, and PAI-1 tumour content are now well established prognostic factor in breast cancer. This body of knowledge could be used for theurapeutic purposes. For example, a large study with 671 patients has allowed the identification of node-negative patients which, according to their u-PA levels, would need adjuvant therapy (Foekens et al, 1992). Many other tumours, especially colorectal cancer, expect a direct clinical evaluation of u-PA, u-PAR and serpins as prognostic factors.

Urokinase plasminogen activator receptor, beta 2-integrins, and Src-kinases within a single receptor complex of human monocytes

The glycosylphosphatidylinositol (GPI)-anchored membrane protein urokinase plasminogen activator-receptor (uPA-R; CD87) is one of the key molecules involved in migration of leukocytes and tumor cells. uPA bound to uPA-R provides the cell proteolytic potential used for degradation of extracellular matrix. uPA-R is also involved in induction of cell adhesion and chemotaxis. Here, we provide a molecular explanation for these uPA-R-related cellular events. By size fractionation of monocyte lysate and affinity isolation on its natural ligand uPA, we demonstrate uPA-R as a component of a receptor complex of relatively large size. Reprecipitation and immunoblotting techniques allowed us to detect the protein tyrosine kinases (PTKs) p60fyn, p53/56lyn, p58/64hck, and p59fgr as components of this "uPA-R complex". Activation of monocytes even with enzymatically inactivated uPA resulted in induction of tyrosine phosphorylation, suggesting modulation of uPA-R-associated PTKs upon ligand binding. In spite of their presence in large complexes, we did not find the GPI-linked proteins CD14, CD58, and CD59 in the uPA-R complex, which indicates the presence of different receptor domains containing GPI-linked proteins in monocytes. However, we identified the leukocyte integrins LFA-1 and CR3 as components of the uPA-R complex as indicated by coisolation of these molecules, as well as by cocapping and comodulation of uPA-R and leukocyte integrins on the monocyte surface. The assemblage of uPA-R, PTKs and membrane spanning beta 2-integrins in one receptor complex indicates functional cooperation. In regard to the involvement of these molecules in pericellular proteolysis, signal transduction, as well as adhesion and chemotactic movement, we suggest uPA-R complex as a potential cellular device for cell migration.

The role of an enolase-related molecule in plasminogen binding to cells

The alpha isoform of enolase is a candidate plasminogen receptor on U937 monocytoid cells [Miles, L. A., Dahlberg, C. L., Plescia, J., Felez, J., Kato, K. & Plow, E. F. (1991) Biochemistry 30, 1682-1691]. In the present study, an enolase-related molecule was detected on the surfaces of peripheral blood monocytes and neutrophils by fluorescence-activated cell sorting. A mRNA transcript encoding a unique membrane form of an enolase-related molecule was not detected by Northern-blotting and primer-extension analyses, consistent with the cell-surface protein being authentic alpha-enolase. Both the alpha and beta isoforms of purified enolase, bound plasminogen with an affinity similar to that of the cell surface. Moreover, immunopurified alpha-enolase enhanced plasminogen activation by tissue plasminogen activator and blocked the binding of plasminogen to alpha 2-antiplasmin, mimicking functions arising from the association of plasminogen with cells. The interaction of the enolase isoforms with plasminogen was dependent upon recognition of the C-terminal lysyl residue of the enolases by the lysine-binding sites of plasminogen, as the interaction was blocked by (a) peptides with C-terminal lysine residues and (b) an antibody to the C-terminal aspect of enolase. A monoclonal antibody was developed, characterized and utilized to quantify the enolase molecules present on the surface of U937 cells. A substantial number of molecules, 1.8 x 10(6)/cell, was present, accounting for approximately 10% of the plasminogen-binding capacity of these cells. These studies clearly establish the role of enolase as a cell-surface plasminogen-binding site with profibrinolytic functions.

Cleaved intracellular plasminogen activator inhibitor 2 in human myeloleukaemia cells is a marker of apoptosis

The proteolytic modification of plasminogen activator inhibitor 2 (PAI-2) was studied during apoptosis in the human promyelocytic leukaemic NB4 cell line during treatment with the phosphatase inhibitors okadaic acid and calyculin A as well as the protein synthesis inhibitor cycloheximide. The apoptic type of cell death was ascertained by morphological and biochemical criteria. In cell homogenates PAI-2 was probed by [125I]urokinase plasminogen activator (uPA) and detected as a sodium dodecyl sulphate-stable M(r) 80,000 complex after reducing sodium dodecyl sulphate-polyacrylamide gel electrophoresis and autoradiography. During apoptosis a smaller (M(r) 70,000) uPA-PAI-2 complex was consistently detected. The modification was in the PAI-2 moiety, as the [125I]uPA tracer could be extracted in its intact form from the complex. Thus the cleaved PAI-2 isoform is a biochemical marker of apoptosis in the promyelocytic NB4 cell line. The modified PAI-2 isoform was also detected in homogenates made from purified human mononuclear leukaemic cells aspirated from the bone marrow of patients suffering from acute and chronic myeloid leukaemia.

High-affinity urokinase receptor antagonists identified with bacteriophage peptide display

Affinity selection of a 15-mer random peptide library displayed on bacteriophage M13 has been used to identify potent ligands for the human urokinase receptor, a key mediator of tumor cell invasion. A family of receptor binding bacteriophage ligands was obtained by sequentially and alternately selecting the peptide library on COS-7 monkey kidney cells and baculovirus-infected Sf9 insect cells overexpressing the human urokinase receptor. Nineteen peptides encoded by the random DNA regions of the selected bacteriophage were synthesized and tested in a urokinase receptor binding assay, where they competed with the labeled N-terminal fragment of urokinase with IC50 values ranging from 10 nM to 10 microM. All of the isolated peptides were linear and showed two relatively short conserved subsequences: LWXXAr (Ar = Y, W, F, or H) and XFXXYLW, neither of which is found in urokinase or its receptor. Competition experiments demonstrated that the most potent peptide, clone 20, prevented binding of bacteriophage displaying the urokinase receptor binding sequence (urokinase residues 13-32). In addition, this peptide blocked other apparently unrelated receptor binding bacteriophage, suggesting overlapping receptor interaction sites for all of these sequences. These results provide a demonstration of bacteriophage display identifying peptide ligands for a receptor expressed on cells and yield leads for the development of urokinase receptor antagonists.