Is plasminogen activator inhibitor-1 the molecular switch that governs urokinase receptor-mediated cell adhesion and release?

Induction of the urokinase type plasminogen activator receptor (uPAR) promotes cell adhesion through its interaction with vitronectin (VN) in the extracellular matrix, and facilitates cell migration and invasion by localizing uPA to the cell surface. We provide evidence that this balance between cell adhesion and cell detachment is governed by PA inhibitor-1 (PAI-1). First, we demonstrate that uPAR and PAI-1 bind to the same site in VN (i.e., the amino-terminal somatomedin B domain; SMB), and that PAI-1 competes with uPAR for binding to SMB. Domain swapping and mutagenesis studies indicate that the uPAR-binding sequence is located within the central region of the SMB domain, a region previously shown to contain the PAI-1-binding motif. Second, we show that PAI-1 dissociates bound VN from uPAR and detaches U937 cells from their VN substratum. This PAI-1 mediated release of cells from VN appears to occur independently of its ability to function as a protease inhibitor, and may help to explain why high PAI-1 levels indicate a poor prognosis for many cancers. Finally, we show that uPA can rapidly reverse this effect of PAI-1. Taken together, these results suggest a dynamic regulatory role for PAI-1 and uPA in uPAR-mediated cell adhesion and release.     

Characterization of a site on PAI-1 that binds to vitronectin outside of the somatomedin B domain

Vitronectin and plasminogen activator inhibitor-1 (PAI-1) are proteins that interact in the circulatory system and pericellular region to regulate fibrinolysis, cell adhesion, and migration. The interactions between the two proteins have been attributed primarily to binding of the somatomedin B (SMB) domain, which comprises the N-terminal 44 residues of vitronectin, to the flexible joint region of PAI-1, including residues Arg-103, Met-112, and Gln-125 of PAI-1. A strategy for deletion mutagenesis that removes the SMB domain demonstrates that this mutant form of vitronectin retains PAI-1 binding (Schar, C. R., Blouse, G. E., Minor, K. M., and Peterson, C. B. (2008) J. Biol. Chem. 283, 10297-10309). In the current study, the complementary binding site on PAI-1 was mapped by testing for the ability of a battery of PAI-1 mutants to bind to the engineered vitronectin lacking the SMB domain. This approach identified a second, separate site for interaction between vitronectin and PAI-1. The binding of PAI-1 to this site was defined by a set of mutations in PAI-1 distinct from the mutations that disrupt binding to the SMB domain. Using the mutations in PAI-1 to map the second site suggested interactions between alpha-helices D and E in PAI-1 and a site in vitronectin outside of the SMB domain. The affinity of this second interaction exhibited a K(D) value approximately 100-fold higher than that of the PAI-1-somatomedin B interaction. In contrast to the PAI-1-somatomedin B binding, the second interaction had almost the same affinity for active and latent PAI-1. We hypothesize that, together, the two sites form an extended binding area that may promote assembly of higher order vitronectin-PAI-1 complexes.     

Kinetic analysis of the interaction between vitronectin and the urokinase receptor.

Although the urokinase receptor (uPAR) binds to vitronectin (VN) and promotes the adhesion of cells to this matrix protein, the biochemical details of this interaction remain unclear. VN variants were employed in BIAcore experiments to examine the uPAR-VN interaction in detail and to compare it to the interaction of VN with other ligands. Heparin and plasminogen bound to VN fragments containing the heparin-binding domain, indicating that this domain was functionally active in the recombinant peptides. However, no significant binding was detected when uPAR was incubated with this domain, and neither heparin nor plasminogen competed with it for binding to VN. In fact, uPAR only bound to fragments containing the somatomedin B (SMB) domain, and monoclonal antibodies (mAbs) that bind to this domain competed with uPAR for binding to VN. Monoclonal antibody 8E6 also inhibited uPAR binding to VN, and this mAb was shown to recognize sulfated tyrosine residues 56 and 59 in the region adjacent to the SMB domain. Destruction of this site by acid treatment eliminated mAb 8E6 binding but had no effect on uPAR binding. Thus, there appears to be a single binding site for uPAR in VN, and it is located in the SMB domain and is distinct from the epitope recognized by mAb 8E6. Inhibition of uPAR binding to VN by mAb 8E6 probably results from steric hindrance.     

The reduced, denatured somatomedin B domain of vitronectin refolds into a stable, biologically active molecule

The high-affinity binding site in human vitronectin (VN) for plasminogen activator inhibitor-1 (PAI-1) has been localized to the NH(2)-terminal cysteine-rich somatomedin B (SMB) domain (residues 1-44). A number of published structural and biochemical studies show conflicting results for the disulfide bonding pattern and the overall fold of the SMB domain, possibly because this domain may undergo disulfide shuffling and/or conformational changes during handling. Here we show that bacterially expressed recombinant SMB (rSMB) can be refolded to a single form that shows maximal activity in binding to PAI-1 and to a conformation-dependent monoclonal antibody (mAb 153). The oxidative refolding pathway of rSMB can be followed in the presence of glutathione redox buffers. This approach allowed the isolation and analysis of a number of intermediate folding species and of the final stably folded species at equilibrium. Competitive surface plasmon resonance analysis demonstrated that the stably refolded rSMB regained biological activity since it bound efficiently to PAI-1 and to mAb 153. In contrast, none of the folding intermediates bound to PAI-1 or to mAb 153. We also show by NMR analysis that the stably refolded rSMB is identical to the material used for the solution structure determination [Kamikubo et al. (2004) Biochemistry 43, 6519] and that it binds specifically to mAb 153 via an interface that includes the three aromatic side chains previously implicated in binding to PAI-1.     

Identification of a PAI‐1‐binding site within an intrinsically disordered region of vitronectin

The serine protease inhibitor, plasminogen activator inhibitor Type‐1 (PAI‐1) is a metastable protein that undergoes an unusual transition to an inactive conformation with a short half‐life of only 1–2 hr. Circulating PAI‐1 is bound to a cofactor vitronectin, which stabilizes PAI‐1 by slowing this latency conversion. A well‐characterized PAI‐1‐binding site on vitronectin is located within the somatomedin B (SMB) domain, corresponding to the first 44 residues of the protein. Another PAI‐1 recognition site has been identified with an engineered form of vitronectin lacking the SMB domain, yet retaining PAI‐1 binding capacity (Schar, Blouse, Minor, Peterson. J Biol Chem. 2008;283:28487–28496). This additional binding site is hypothesized to lie within an intrinsically disordered domain (IDD) of vitronectin. To localize the putative binding site, we constructed a truncated form of vitronectin containing 71 amino acids from the N‐terminus, including the SMB domain and an additional 24 amino acids from the IDD region. This portion of the IDD is rich in acidic amino acids, which are hypothesized to be complementary to several basic residues identified within an extensive vitronectin‐binding site mapped on PAI‐1 (Schar, Jensen, Christensen, Blouse, Andreasen, Peterson. J Biol Chem. 2008;283:10297–10309). Steady‐state and stopped‐flow fluorescence measurements demonstrate that the truncated form of vitronectin exhibits the same rapid biphasic association as full‐length vitronectin and that the IDD hosts the elusive second PAI‐1 binding site that lies external to the SMB domain of vitronectin.     

Interaction of plasminogen activator inhibitor type-1 (PAI-1) with vitronectin (Vn): mapping the binding sites on PAI-1 and Vn

The serpin plasminogen activator inhibitor type-1 (PAI-1), as the primary physiological inhibitor of both urokinase-type (uPA) and tissue-type (tPA) plasminogen activator, plays an important role in the regulation of the fibrinolytic system as well as in extracellular remodeling in both physiological and pathophysiological processes. In plasma as well as in the extracellular matrix PAI-1 binds to vitronectin (Vn), an interaction that affects the function of both proteins. As PAl-1/Vn interaction has a significant regulatory function in fibrinolysis, thrombolysis, and cell adhesion in cancer spread, there is a strong interest in defining the binding sites on PAI-1 and Vn as the basis of a rational design of novel drugs that may modulate PAI-1/Vn-mediated effects. In this minireview, we give an overview on the approaches to define the Vn binding site of PAI-1 and vice versa. Although in the case of PAI-1 the region around alpha-helix E and alpha-helix F of PAI-1 has been demonstrated to be important for its interaction with Vn, the precise location of the Vn-binding region has not completely been resolved. The major high-affinity PAI-1 binding region of Vn is localized within the N-terminal somatomedin B (SMB) domain of Vn. There are indications for at least one other low-affinity PAI-1 binding site in the C-terminal region of Vn, which seems to be involved in the formation of larger PAI-1/Vn complexes.     

Functional structure of the somatomedin B domain of vitronectin

The N-terminal somatomedin B domain (SMB) of vitronectin binds PAI-1 and the urokinase receptor with high affinity and regulates tumor cell adhesion and migration. We have shown previously in the crystal structure of the PAI-1/SMB complex that SMB, a peptide of 51 residues, is folded as a compact cysteine knot of four pairs of crossed disulfide bonds. However, the physiological significance of this structure was questioned by other groups, who disputed the disulfide bonding shown in the crystal structure (Cys5-Cys21, Cys9-Cys39, Cys19-Cys32, Cys25-Cys31), notably claiming that the first disulfide is Cys5-Cys9 rather than the Cys5-Cys21 bonding shown in the structure. To test if the claimed Cys5-Cys9 bond does exist in the SMB domain of plasma vitronectin, we purified mouse and rat plasma vitronectin that have a Met (hence cleavable by cyanogen bromide) at residue 14, and also prepared recombinant human SMB variants from insect cells with residues Asn14 or Leu24 mutated to Met. HPLC and mass spectrometry analysis showed that, after cyanogen bromide digestion, all the fragments of the SMB derived from mouse or rat vitronectin or the recombinant SMB mutants are still linked together by disulfides, and the N-terminal peptide (residue 1-14 or 1-24) can only be released when the disulfide bonds are broken. This clearly demonstrates that Cys5 and Cys9 of SMB do not form a disulfide bond in vivo, and together with other structural evidence confirms that the only functional structure of the SMB domain of plasma vitronectin is that seen in its crystallographic complex with PAI-1.     

Mapping of binding sites for heparin, plasminogen activator inhibitor-1, and plasminogen to vitronectin's heparin-binding region reveals a novel vitronectin-dependent feedback mechanism for the control of plasmin formation.

Vitronectin (VN) has been implicated as a major matrix-associated regulator component of plasminogen activation by serving as a potent stabilizing cofactor of plasminogen activator inhibitor-1 (PAI-1). The direct binding of heparin, plasminogen as well as PAI-1 in its latent and active form to immobilized VN was studied in the absence or presence of competitors. Monoclonal antibodies against the carboxyl-terminal portion of VN inhibited both PAI-1 and plasminogen binding, whereas heparin, heparan sulfate with a high degree of sulfation, or dextran sulfate interfered with PAI-1 binding (KD = 20 nM) only. Utilizing synthetic peptides encompassing overlapping sequences of the heparin-binding domain of VN, adjacent heparin and PAI-1-binding sites were localized within the sequence 348-370 of VN. Although a number of other serine protease inhibitors which do not form binary complexes with VN contain a reactive-site Ser at their P1'-position, a reactive-site P1' mutant of PAI-1 (Met----Ser) showed comparable if not increased binding to VN. Binding of Lys-plasminogen and active-site-blocked plasmin was at least 10-fold higher in affinity (KD = 85-100 nM) compared to Glu-plasminogen (KD approximately 1 microM) and could be inhibited by lysine analogs but not by glycosaminoglycans or PAI-1, indicating that heteropolar plasmin(ogen) binding of VN occurs to an adjacent segment upstream to the heparin and PAI-1-binding sites. This contention was further supported in binding studies with plasmin-modified VN which lost both heparin and PAI-1 binding but exhibited 2-3-fold higher capacity to bind plasminogen. The essential plasmin(ogen)-binding site was mapped by ligand blot analysis to the carboxyl-terminal portion of proteolytically trimmed VN (M(r) = 61,000). Moreover, treatment of the extracellular matrix of human umbilical vein endothelial cells with plasmin resulted in partial degradation of matrix-associated VN and concomitant release of PAI-1, but increased the ability of the matrix by about 2-fold to bind plasminogen. These results are indicative of differential interactions of VN with components of the plasminogen activation system, whereby plasmin itself may provoke the switch of VN from an anti-fibrinolytic into a pro-fibrinolytic cofactor. This process reflects a novel role for the adhesive protein and its degradation product(s) in the possible feedback regulation of localized plasmin formation at extracellular sites.     

A novel antithrombotic role for high molecular weight kininogen as inhibitor of plasminogen activator inhibitor-1 function

The adhesive glycoprotein vitronectin (VN) forms a function-stabilizing complex with plasminogen activator inhibitor-1 (PAI-1), the major fibrinolysis inhibitor in both plasma and vessel wall connective tissue. VN also interacts with two-chain high molecular weight kininogen (HKa), particularly its His-Gly-Lys-rich domain 5, and both HKa and PAI-1 are antiadhesive factors that have been shown to compete for binding to VN. In this study the influence of HKa and domain 5 on the antifibrinolytic function of PAI-1 was investigated. In a purified system, HKa and particularly domain 5 inhibited the binding of PAI-1 to VN and promoted PAI-1 displacement from both isolated VN as well as subendothelial extracellular matrix-associated VN. The sequence Gly(486)-Lys(502) of HKa domain 5 was identified as responsible for this inhibition. Although having no direct effect on PAI-1 activity itself, HKa domain 5 or the peptide Gly(486)-Lys(502) markedly destabilized the VN.PAI-1 complex interaction, resulting in a significant reduction of PAI-1 inhibitory function on plasminogen activators, resembling the effect of VN antibodies that prevent stabilization of PAI-1. Furthermore, high affinity fibrin binding of PAI-1 in the presence of VN as well as the VN-dependent fibrin clot stabilization by the inhibitor were abrogated in the presence of the kininogen forms mentioned. Taken together, our data indicate that the peptide Gly(486)-Lys(502) derived from domain 5 of HKa serves to interfere with PAI-1 function. Based on these observations potential low molecular weight PAI-1 inhibitors could be designed for the use in therapeutic interventions against thromboembolic complications.     

Structural requirements for the extracellular interaction of plasminogen activator inhibitor 1 with endothelial cell matrix-associated vitronectin.

The interaction of plasminogen activator inhibitor-1 (PAI-1) with its binding protein vitronectin (VN) (Declerck, P. J., De Mol, M., Alessi, M.-C., Baudner, S., Paques, E.-P., Preissner, K. T., Müller-Berghaus, G., and Collen, D. (1988) J. Biol. Chem. 263, 15454-15461) in the extracellular matrix (ECM) of cultured human endothelial cells (HUVEC) was studied. Like PAI-1, VN was found associated with the ECM as evidenced by direct antibody binding, by Western blot analysis as well as by diffuse immunofluorescence staining in permeabilized HUVEC. The specific interaction of VN with confluent monolayers of HUVEC was found to be saturable within 2-4 h at 37 degrees C only with respect to binding to the cells, while no saturable binding to the underlying ECM was observed, indicating that the majority if not all ECM-associated VN was derived from the culture medium. In contrast to PAI-1, ECM-associated VN was resistant toward glycine (pH 2.3), guanidine or urokinase treatment, suggesting that VN was tightly associated with the ECM network. Binding of recombinant PAI-1 (rPAI-1) was largely blocked by anti-VN IgG and only partly by anti-collagen IgG but not by antibodies against other ECM components, indicating that VN constitutes the primary binding protein for ECM-associated PAI-1. This contention was supported by ligand blotting experiments in which rPAI-1 was reacted with nitrocellulose replicas of electrophoretically separated ECM components. Protein band(s) (Mr = 63,000-67,000), comigrating with bovine VN (i.e. medium-derived VN) rather than with human VN were identified as major binding component(s). Moreover, binding studies with purified components revealed that PAI-1 did not show any affinity for collagen (type I/III) alone, whereas VN collagen coating was a much better template for PAI-1 binding than VN alone and that conformationally extended VN provides maximal PAI-1 binding capacity. Binding of rPAI-1 to surface-coated VN was saturable and revealed that (unlike urokinase) heparin or the synthetic peptide Gly-Arg-Gly-Asp-Ser did not inhibit PAI-1 binding. Ligand binding of rPAI-1 to nitrocellulose replicas from sodium dodecyl sulfate-polyacrylamide gels containing electrophoretically separated peptides from VN digests documented the association of PAI-1 with Mr = 10,000-20,000 fragments originating from the heparin-binding domain of VN. These results indicate that the exposure of the glycosaminoglycan-binding domain in VN may allow the concomitant binding of PAI-1 and heparin-like molecules to this region of the VN molecule.(ABSTRACT TRUNCATED AT 400 WORDS)     

Highly sulfated glycosaminoglycans augment the cross-linking of vitronectin by guinea pig liver transglutaminase. Functional studies of the cross-linked vitronectin multimers.

Vitronectin (VN) is an adhesive glycoprotein with roles in the complement, coagulation, and immune systems. Many of the functions of VN are mediated by a glycosaminoglycan binding site, near its carboxyl-terminal end. In this paper, we show that the highly sulfated glycosaminoglycans (GAGs), dextran sulfate, pentosan polysulfate, and fucoidan effectively augment [14C]putrescine incorporation into VN and cross-linking of VN into high molecular multimers by guinea pig liver transglutaminase (TG). Other GAGs including heparin, low molecular weight heparin, dermatan sulfate, keratan sulfate, and the nonsulfated dextrans were ineffective in accelerating these reactions. Dextran sulfate of average molecular mass 500 kDa was more effective than dextran sulfate of average molecular mass 5 kDa, supporting a template mechanism of action of the GAGs, in which VN molecules align on the GAG in a conformation suitable for cross-linking. The VN multimers catalyzed by TG retained functional activity in binding [3H]heparin, platelets, and plasminogen activator inhibitor type-1 (PAI-1). [3H]Heparin bound selectively to the 65-kDa monomeric band of VN and to the multimers derived from this band. PAI-1, however, bound equally to both the 75- and 65-kDa monomeric forms of VN, suggesting that the PAI-1 binding site on VN is distinct from the GAG binding site. The interaction of GAGs with the TG-catalyzed cross-linking of VN may facilitate studies of VN structure-function relationships.     

The solution structure of the N-terminal domain of human vitronectin: proximal sites that regulate fibrinolysis and cell migration

The three-dimensional structure of an N-terminal fragment comprising the first 51 amino acids from human plasma vitronectin, the somatomedin B (SMB) domain, has been determined by two-dimensional NMR approaches. An average structure was calculated, representing the overall fold from a set of 20 minimized structures. The core residues (18-41) overlay with a root mean square deviation of 2.29 +/- 0.62 A. The N- and C-terminal segments exhibit higher root mean square deviations, reflecting more flexibility in solution and/or fewer long-range NOEs for these regions. Residues 26-30 form a unique single-turn alpha-helix, the locus where plasminogen activator inhibitor type-1 (PAI-1) is bound. This structure of this helix is highly homologous with that of a recombinant SMB domain solved in a co-crystal with PAI-1 (Zhou, A., Huntington, J. A., Pannu, N. S., Carrell, R. W., and Read, R. J. (2003) Nat. Struct. Biol. 10, 541-544), although the remainder of the structure differs. Significantly, the pattern of disulfide cross-links observed in this material isolated from human plasma is altogether different from the disulfides proposed for recombinant forms. The NMR structure reveals the relative orientation of binding sites for cell surface receptors, including an integrin-binding site at residues 45-47, which was disordered and did not diffract in the co-crystal, and a site for the urokinase receptor, which overlaps with the PAI-1-binding site.     

Localization of the high molecular weight kininogen binding site in the heavy chain of human factor XI to amino acids phenylalanine 56 through serine 86

We have previously demonstrated that a monoclonal antibody (5F7) directed against the heavy chain region of factor XI inhibits the binding of factor XI to high molecular weight kininogen (high Mr kininogen) and the surface-mediated proteolytic activation of factor XI by factor XIIa in the presence of high Mr kininogen. In order to identify the structural domain of factor XI that binds high Mr kininogen, CNBr-digested factor XI was passed over a 5F7 antibody affinity column. One of two CNBr peptides that bound to this 5F7 affinity column inhibited binding of 125I-factor XI to high Mr kininogen, as did intact factor XI. Polyacrylamide gel electrophoresis in sodium dodecyl sulfate of an inhibitory peptide purified by high performance liquid chromatography revealed an Mr of 10,000-15,000. Gas-phase sequencing of this peptide revealed the following amino-terminal sequence: X-X-Val-Thr-Gln-Leu-Leu-Lys-Asp-Thr. These data together with the amino acid composition of the isolated peptide indicate that both the epitope recognized by antibody 5F7 and at least a portion of the high Mr kininogen binding site are contained within the amino-terminal portion of factor XI comprising residues Glu-1 through Met-102. Further cleavage of this peptide with o-iodosobenzoic acid at a tryptophanyl peptide bond revealed that an Mr 5,000 peptide (with the amino-terminal sequence Trp-Phe-Thr-Cys-Val-Leu) bound to a high Mr kininogen affinity column and inhibited binding of 125I-factor XI to high Mr kininogen. Finally, a synthetic peptide comprising residues Phe-56 through Ser-86 inhibited 125I-factor XI binding to high Mr kininogen. These experiments strongly suggest that the high Mr kininogen binding site is contained within the domain in the heavy chain region of factor XI comprising residues Phe-56 through Ser-86.     

Plasminogen activator inhibitor-1 regulates cell adhesion by binding to the somatomedin B domain of vitronectin

Plasminogen activator inhibitor-1 (PAI-1) binds to the somatomedin B (SMB) domain of vitronectin. It inhibits the adhesion of U937 cells to vitronectin by competing with the urokinase receptor (uPAR; CD87) on these cells for binding to the same domain. Although the inhibitor also blocks integrin-mediated cell adhesion, the molecular basis of this effect is unclear. In this study, the effect of the inhibitor on the adhesion of a variety of cells (e.g., U937, MCF7, HT-1080, and HeLa) to vitronectin was assessed, and the importance of the SMB domain in these interactions was determined. Although PAI-1 blocked the adhesion of all of these cells to vitronectin-coated wells, it did not block adhesion to a variant of vitronectin which lacked the SMB domain. Interestingly, HT-1080 and U937 cells attached avidly to microtiter wells coated with purified recombinant SMB (which does not contain the RGD sequence), and this adhesion was again blocked by the inhibitor. These results affirm that PAI-1 can inhibit both uPAR- and integrin-mediated cell adhesion, and demonstrate that the SMB domain of vitronectin is required for these effects. They also show that multiple cell types can employ uPAR as an adhesion receptor. The use of purified recombinant SMB should help to further define this novel adhesive pathway, and to delineate its relationship with integrin-mediated adhesive events.     

Identification of regulatory sequences in the type 1 plasminogen activator inhibitor gene responsive to transforming growth factor beta.

Regulation of the human type 1 plasminogen activator inhibitor (PAI-1) promoter by transforming growth factor-beta (TGF beta) was studied. An 800-base pair fragment from the PAI-1 promoter and 5'-flanking region was fused to the firefly luciferase reporter gene and transfected into Hep3B human hepatoma cells. Treatment of the cells with TGF beta induced luciferase activity by more than 50-fold. Transfection studies using constructs with 5' or 3' deletions through this region revealed that two sequences were important in the TGF beta response. The first sequence was located in the proximal promoter (-49 to -87) and mediated an 11-fold induction with TGF beta, while the second more distal region (-636 to -740) contained two sequences which together mediated a 50-fold or greater response. Sequence comparison indicated that both of the responsive regions contained sequences with high homology to the AP-1 consensus binding site. Moreover, gel retardation analysis experiments demonstrated that both sequences bound a common nuclear protein, and that an oligonucleotide containing a consensus AP-1 sequence was able to compete for the binding of this common protein. Thus, the response of the PAI-1 gene to TGF beta is mediated by at least two separate regions, and both of these regions contain DNA sequences homologous to the AP-1 binding site.     

Kinetic analysis of the interaction between type 1 plasminogen activator inhibitor and vitronectin and evidence that the bovine inhibitor binds to a thrombin-derived amino-terminal fragment of bovine vitronectin.

The interaction between guanidine-activated bovine type 1 plasminogen activator inhibitor (PAI-1) and bovine vitronectin was investigated. Activated PAI-1 bound to vitronectin in a dose- and time-dependent manner, and binding was saturable. The dissociation constant (Kd) for this interaction was estimated to be 3.10(-10) mol/l by Scatchard analysis. Complexes of activated PAI-1 and vitronectin were relatively stable at 4 degrees C (T1/2 greater than 24 h), but dissociated with a T1/2 of 4 h at 37 degrees C. The half-life of PAI-1 activity was increased from 2.5 to 4.5 h upon binding to immobilized vitronectin. In order to identify the binding domain(s) in vitronectin for activated PAI-1, the ability of PAI-1 to bind to vitronectin fragments was assessed. Vitronectin was cleaved by thrombin in a dose- and time-dependent manner, generating fragments of Mr 60,000, 54,000 and 38,000. The PAI-1 binding domain(s) were not destroyed by this treatment, since the digested vitronectin competed with immobilized vitronectin for PAI-1 binding to the same extent as uncleaved vitronectin. The thrombin digested vitronectin fragments were fractionated by SDS-PAGE and analyzed by PAI-1 ligand binding. The smallest fragment (Mr 38,000) retained PAI-1 binding function, and sequence analysis demonstrated that this fragment contained the NH2-terminus of bovine vitronectin. These results suggest that the high-affinity binding site for activated PAI-1 is located in the NH2-terminal region of the bovine vitronectin molecule.     

Identification of the disulfide bonds in the recombinant somatomedin B domain of human vitronectin

The NH(2)-terminal somatomedin B (SMB) domain (residues 1-44) of human vitronectin contains eight Cys residues organized into four disulfide bonds and is required for the binding of type 1 plasminogen activator inhibitor (PAI-1). In the present study, we map the four disulfide bonds in recombinant SMB (rSMB) and evaluate their functional importance. Active rSMB was purified from transformed Escherichia coli by immunoaffinity chromatography using a monoclonal antibody that recognizes a conformational epitope in SMB (monoclonal antibody 153). Plasmon surface resonance (BIAcore) and competitive enzyme-linked immunosorbent assays demonstrate that the purified rSMB domain and intact urea-activated vitronectin have similar PAI-1 binding activities. The individual disulfide linkages present in active rSMB were investigated by CNBr cleavage, partial reduction and S-alkylation, mass spectrometry, and protein sequencing. Two pairs of disulfide bonds at the NH(2)-terminal portion of active rSMB were identified as Cys(5)-Cys(9) and Cys(19)-Cys(21). Selective reduction/S-alkylation of these two disulfide linkages caused the complete loss of PAI-1 binding activity. The other two pairs of disulfide bonds in the COOH-terminal portion of rSMB were identified as Cys(25)-Cys(31) and Cys(32)-Cys(39) by protease-generated peptide mapping of partially reduced and S-alkylated rSMB. These results suggest a linear uncrossed pattern for the disulfide bond topology of rSMB that is distinct from the crossed pattern present in most small disulfide bond-rich proteins.     

Localization of an integrin binding site to the C terminus of talin

Talin, consisting of a 47-kDa N-terminal head domain (residues 1-433) and a 190-kDa C-terminal rod domain (residues 434-2541), links integrins to the actin cytoskeleton. We previously reported that the binding stoichiometry of integrin alpha(IIb)beta(3):talin is approximately 2:1. More recently, an integrin binding site has been localized to the talin head domain. In the present study, we identified another integrin binding site at the C-terminal region of the talin rod domain. In a solid phase binding assay, RGD affinity-purified alpha(IIb)beta(3) bound in a dose-dependent manner to microtiter wells coated with the isolated 190-kDa proteolytic fragment of the talin rod domain. Additionally, alpha(IIb)beta(3) also bound to the talin rod domain captured by 8d4, an anti-talin monoclonal antibody. Polyclonal antibodies raised against a recombinant protein fragment corresponding to the entire talin rod domain (anti-talin-R) inhibited alpha(IIb)beta(3) binding to intact talin by approximately 50% but completely blocked alpha(IIb)beta(3) binding to the talin rod domain. To localize the integrin binding site, we examined alpha(IIb)beta(3) binding to recombinant polypeptide fragments corresponding to partial sequences of the talin rod domain. Whereas alpha(IIb)beta(3) bound effectively to talin-(1075-2541) and talin-(1984-2541), it failed to bind to talin-(434-1076) and talin-(434-1975). Furthermore, the binding of alpha(IIb)beta(3) to talin-(1984-2541) was inhibited by anti-talin-R. These results indicate that an integrin binding site is located within residues 1984-2541 of the talin rod domain. Thus, talin contains two integrin binding sites, one in the homologous FERM (band four-point-one, ezrin, radixin, moesin) domain and another near its C terminus. Because talin exists as an anti-parallel homodimer in focal adhesions, the two integrin binding sites in the adjacent talin molecules would be in close proximity with each other.     

The plasminogen activator inhibitor-1 binding site in the kringle-2 domain of tissue-type plasminogen activator

We have shown that synthetic peptides containing the amino acid sequence Asn-Arg-Arg-Leu, derived from the amino acid sequence of the inner loop of the kringle-2 domain of tissue-type plasminogen activator (tPA), inhibited complex formation between two chain tPA and plasminogen activator inhibitor-1 (PAI-1) by binding to PAI-1. This binding was reversible and was inhibited by not only tPA but also by enzymatically inactive tPA. Quantitative analyses of the interaction of PAI-1 with the peptide containing the Asn-Arg-Arg-Leu sequence indicated that the PAI-1 binding site residues in the inner loop of the kringle-2 domain and is preferentially expressed in two chain tPA.     

Binding of plasminogen activator inhibitor type-1 to extracellular matrix of Hep G2 cells. Evidence that the binding protein is vitronectin.

Catabolism of plasminogen activators by Hep G2 cells is mediated by a specific receptor which recognizes complexes of these serine proteases with their physiological inhibitor, plasminogen activator inhibitor type-1 (PAI-1). This catabolic process is initiated by interaction of exogenous plasminogen activators with bioactive PAI-1, which is secreted and localizes in an active form to the extracellular matrix (ECM) of Hep G2 cells. We now report that vitronectin (VN) mediates the specific binding of PAI-1 to the ECM of these cells. Purified bovine or human VN competes for specific binding of PAI-1 to Hep G2 ECM, and ligand blotting reveals specific binding of PAI-1 to ECM-associated VN. Hep G2 cells secrete both VN and PAI-1, and pulse-chase studies strongly suggest that these proteins associate only following secretion. Although Hep G2 cell-derived VN does not significantly bind to ECM in vitro, 30-40% of endogenous PAI-1 binds to the ECM, even in the presence of human serum, suggesting that ECM-associated VN is entirely derived from bovine serum. PAI-1 was localized by indirect immunofluorescence to ECM beneath cells and at cell margins, whereas VN exhibited a uniform distribution throughout the growth substratum. VN associated with the ECM may confer retention and bioactivity to PAI-1, potentially facilitating both pericellular regulation of plasmin generation and the rapid hepatic clearance of plasminogen activators.