Endothelial cells produce a latent inhibitor of plasminogen activators that can be activated by denaturants

Conditioned medium from cultured bovine aortic endothelial cells contains an inactive plasminogen activator inhibitor (PAI). This latent PAI can be "activated" with denaturants. For example, less than 0.01 units/microliter of PAI activity was detected in untreated conditioned medium, but medium treated with sodium dodecyl sulfate (1.7 mM), guanidine HCl (4 M), urea (12 M) or KSCN (6 M) contained 0.9, 1.9, 0.8, and 0.5 units/microliter, respectively. This effect was dose-dependent with respect to the particular reagent used, and the same concentration of reagent which induced PAI activity also stimulated the ability of a component in conditioned medium to form sodium dodecyl sulfate-stable complexes with exogenously added plasminogen activators. Neither activity was stimulated by extensive dialysis or by treatment with NaCl (5 M), Na2SO4 (2.8 M), or dicetyl phosphate (0.1%). Analysis of treated and untreated conditioned medium by gel filtration revealed that the latent and active PAIs migrated with apparent Mr values of 30,000 and 50,000, respectively. Thus, "activation" is associated with an increase in the apparent Mr of the molecule. These observations suggest that activation does not result from the removal of either a small dialyzable component from the medium, or of a large Mr component that is bound to the latent PAI. Other possible mechanisms of activation are discussed. We recently isolated an active PAI from bovine endothelial cells (van Mourik, J.A., Lawrence, D.A., and Loskutoff, D.J. (1984) J. Biol. Chem. 259, 14914-14921). Monospecific antiserum to this active PAI selectivity immunoprecipitated the latent PAI from conditioned medium. These results indicate that the two PAIs are immunologically related and suggest that the latent form is converted into the active form by the sodium dodecyl sulfate present during the purification.     

Platelet plasminogen activator inhibitor: purification and characterization of interaction with plasminogen activators and activated protein C

Plasminogen activator inhibitor (PAI) was purified in active form from porcine platelets under nondenaturing conditions. The purified inhibitor (Mr 47,000) reacts with tissue-type plasminogen activator (t-PA), urokinase (UK), and activated protein C (APC) to yield both SDS-stable complexes and a modified PAI of slightly reduced molecular weight. The second-order rate constants for the inhibition of t-PA and UK by PAI are 3.5 X 10(7) and 3.4 X 10(7) M-1 s-1, respectively. Activated protein C reacts with PAI with a second-order rate constant of 1.1 X 10(4) M-1 s-1. This rate is not accelerated by protein S, phospholipid, and calcium, or heparin. It is concluded that (1) PAI can function as both inhibitor and substrate of its target proteases, (2) if APC promotes fibrinolysis via inactivation of PAI, then APC must be present in concentrations several orders of magnitude greater than t-PA, or the interaction of APC and PAI must be accelerated by presently unknown mechanisms, and (3) in the absence of heparin, platelet PAI is the most rapid inhibitor of APC yet described.     

Reversible interactions between plasminogen activators and plasminogen activator inhibitor-1.

We have shown that the urokinase (UK) kringle domain contains a high-affinity plasminogen activator inhibitor-1 (PAI-1) binding site, responsible for the 10-fold faster complex formation between UK and PAI-1 than between PAI-1 and low-molecular-weight urokinase (LMWUK). Complex formation between UK and PAI-1, but not between LMWUK and PAI-1, was suppressed 10-fold in the presence of peptide U-107 derived from the UK kringle domain. Peptide U-373 derived from the UK catalytic domain slowed complex formation between UK and PAI-1 and also LMWUK and PAI-1. Inactivation of tissue-type plasminogen activator (tPA) by PAI-1 was slowed 10-fold in the presence of peptides derived from the tPA finger and kringle-2 domains. DFP-inactivated (DIP) UK and both forms of DIP-tPA inhibited PAI-1 binding to U-107 and to U-373 whereas single-chain urokinase-type PA (scuPA) was unable to compete with either peptide for PAI-1 binding. These data suggest that the reversible PAI-1 binding site in the UK A-chain plays a role in the rapid association with PAI-1 as important as those that reside in the tPA A-chain and that reversible PAI-1 binding sites are expressed on the surface of UK upon conversion from scuPA, in contrast to tPA.     

Bovine plasminogen activator inhibitor 1: specificity determinations and comparison of the active, latent, and guanidine-activated forms

The plasminogen activator inhibitor 1 (PAI-1) synthesized and released by cultured bovine aortic endothelial cells is present in conditioned medium in a latent form that can be activated by guanidine hydrochloride [Hekman, C. M., & Loskutoff, D. J. (1985) J. Biol. Chem. 260, 11581-11587]. The purified, guanidine-activated PAI-1 was shown to inhibit both plasmin and trypsin in a dose- and time-dependent manner. Second-order rate constants for these interactions were calculated to be 6.6 X 10(5) and 7.0 X 10(6) M-1 s-1 for plasmin and trypsin, respectively. Experiments were conducted to compare the inherently active and the guanidine-activated forms of PAI-1. The two active forms had similar kinetic parameters for interaction with urokinase (Kd, 0.3 pM; kassoc, 1.5 X 10(8) M-1 s-1) and were both inactivated upon treatment with acid or base and by incubation at 37 degrees C. The latent form was relatively stable when incubated under similar conditions. The decrease in PAI-1 activity upon incubation at 37 degrees C was partially restored by a second treatment with guanidine hydrochloride. However, the degree of recovery decreased as a function of incubation time at 37 degrees C. These data suggest that active and guanidine-activated PAI-1 represent a single form of PAI-1. Incubation of this form at 37 degrees C yields two distinct populations of inactive PAI-1, one capable of reactivation and another that appears to be irreversibly inactivated.     

Purification and characterization of active and latent forms of recombinant plasminogen activator inhibitor 1 produced in Escherichia coli.

Plasminogen activator inhibitor 1 (PAI-1), the principal physiological inhibitor of tissue plasminogen activator (tPA), is a protein of 379 amino acids and belongs to the SERPIN family of serine protease inhibitors. We have previously described methods to express [Sisk et al. (1990) Gene 96, 305-309] and purify [Reilly et al. (1990) J. Biol. Chem. 265, 9570-9574] a highly active form of the protein in substantial amounts, from Escherichia coli. Further analyses of this material showed the presence of small but significant amounts of latent rPAI-1. The present paper describes for the first time purification of latent and active forms of rPAI-1 from a single preparation, as well as the functional and structural characteristics of the two forms. Latent rPAI-1, which has properties similar to the latent forms described by other groups, was separated from active rPAI-1 by high-resolution ion-exchange chromatography or by affinity chromatography using immobilized anhydrotrypsin. It had low intrinsic activity (< 5% of active rPAI-1) and was partially reactivated by guanidine hydrochloride treatment or by incubation with vitronectin. Conversion of the active rPAI-1 to the latent form was influenced by temperature and additives including sucrose, EDTA, and arginine. Active and latent rPAI-1 did not show any obvious differences in their primary structures and displayed remarkably similar secondary structures as determined by circular dichroism spectral analyses. However, they did exhibit differences in tryptophan fluorescence, suggesting tertiary structural differences between the two forms.     

Mechanisms of conversion of plasminogen activator inhibitor 1 from a suicide inhibitor to a substrate by monoclonal antibodies

We have delineated two different reaction mechanisms of monoclonal antibodies (mAbs), MA-8H9D4 and either MA-55F4C12 or MA-33H1F7, that convert plasminogen activator inhibitor 1 (PAI-1) to a substrate for tissue (tPA)- and urokinase plasminogen activators. MA-8H9D4 almost completely (98-99%) shifts the reaction to the substrate pathway by preventing disordering of the proteinase active site. MA-8H9D4 does not affect the rate-limiting constants (k(lim)) for the insertion of the reactive center loop cleaved by tPA (3.5 s(-1)) but decreases k(lim) for urokinase plasminogen activator from 25 to 4.0 s(-1). MA-8H9D4 does not cause deacylation of preformed PAI-1/proteinase complexes and probably acts prior to the formation of the final inhibitory complex, interfering with displacement of the acylated serine from the proteinase active site. MA-55F4C12 and MA-33H1F7 (50-80% substrate reaction) do not interfere with initial PAI-1/proteinase complex formation but retard the inhibitory pathway by decreasing k(lim) (>10-fold for tPA). Interaction of two mAbs with the same molecule of PAI-1 has been directly demonstrated for pairs MA-8H9D4/MA-55F4C12 and MA-8H9D4/MA-33H1F7 but not for MA-55F4C12/MA-33H1F7. The strong functional additivity observed for MA-8H9D4 and MA-55F4C12 demonstrates that these mAbs interact independently and affect different steps of the PAI-1 reaction mechanism.     

The use of monoclonal antibodies to distinguish several chemically modified forms of human alpha 1-proteinase inhibitor.

The purpose of our investigation was to obtain monoclonal antibodies that could distinguish three forms of alpha 1-proteinase inhibitor (alpha 1-PI): native alpha 1-PI, N-chlorosuccinimide-oxidized alpha 1-PI (Ox-alpha 1-PI) and proteolytically modified alpha 1-PI (alpha 1-PI). Three specific monoclonal antibodies were characterized as to their binding properties. By using the Bio-Dot assay, it was found that all three forms of alpha 1-PI were capable of binding to antibody 6D4-6-18, that only Ox-alpha 1-PI, but not native alpha 1-PI or alpha 1-PI, could bind to antibody 6C7-5, and that alpha 1-PI and a complex between alpha 1-PI and trypsin uniquely were not able to bind to antibody 5C12-8-7. Thus it was concluded that it is possible to use monoclonal antibodies with different epitopic specificities to distinguish two chemically modified forms of alpha 1-PI from the native protein.     

Conformational studies of plasminogen activator inhibitor type 1 by fluorescence spectroscopy. Analysis of the reactive centre of inhibitory and substrate forms, and of their respective reactive-centre cleaved forms.

The inhibitors that belong to the serpin family are suicide inhibitors that control the major proteolytic cascades in eucaryotes. Recent data suggest that serpin inhibition involves reactive centre cleavage followed by loop insertion, whereby the covalently linked protease is translocated away from the initial docking site. However under certain circumstances, serpins can also be cleaved like a substrate by target proteases. In this report we have studied the conformation of the reactive centre of plasminogen activator inhibitor type 1 (PAI-1) mutants with inhibitory and substrate properties. The polarized steady-state and time-resolved fluorescence anisotropies were determined for BODIPY(R) probes attached to the P1' and P3 positions of the substrate and active forms of PAI-1. The fluorescence data suggest an extended orientational freedom of the probe in the reactive centre of the substrate form as compared to the active form, revealing that the conformation of the reactive centres differ. The intramolecular distance between the P1' and P3 residues in reactive centre cleaved inhibitory and substrate mutants of PAI-1, were determined by using the donor-donor energy migration (DDEM) method. The distances found were 57+/-4 A and 63+/-3 A, respectively, which is comparable to the distance obtained between the same residues when PAI-1 is in complex with urokinase-type plasminogen activator (uPA). Following reactive centre cleavage, our data suggest that the core of the inhibitory and substrate forms possesses an inherited ability of fully inserting the reactive centre loop into beta-sheet A. In the inhibitory forms of PAI-1 forming serpin-protease complexes, this ability leads to a translocation of the cognate protease from one pole of the inhibitor to the opposite one.     

Vitronectin governs the interaction between plasminogen activator inhibitor 1 and tissue-type plasminogen activator.

The "serpin" plasminogen activator inhibitor 1 (PAI-1) is the fast acting inhibitor of plasminogen activators (tissue-type (t-PA) and urokinase type-PA) and is an essential regulatory protein of the fibrinolytic system. Its P1-P1' reactive center (R346 M347) acts as a "bait" for tight binding to t-PA/urokinase-type PA. In vivo, PAI-1 is encountered in complex with vitronectin, an interaction known to stabilize its activity but not to affect the second-order association rate constant (k1) between PAI-1 and t-PA. Nevertheless, by using PAI-1 reactive site variants (R346M, M347S, and R346M M347S), we show that the binding of vitronectin to the PAI-1 mutant proteins improves plasminogen activator inhibition. In the absence of vitronectin the PAI-1 R346M mutants are virtually inactive toward t-PA (k1 less than 1 x 10(3) M-1 s-1). In contrast, in the presence of vitronectin the rate of association increases about 1,000-fold (k1 of 6-8 x 10(5) M-1 s-1). This inhibition coincides with the formation of serpin-typical, sodium dodecyl sulfide-stable t-PA.PAI-1 R346M (R346M M347S) complexes. As evidenced by amino acid sequence analysis, the newly created M346-M/S347 peptide bond is susceptible to attack by t-PA, similar to the wild-type R346-M347 peptide bond, indicating that in the presence of vitronectin M346 functions as an efficient P1 residue. In addition, we show that the inhibition of t-PA and urokinase-type PA by PAI-1 mutant proteins is accelerated by the presence of the nonprotease A chains of the plasminogen activators.     

Expression of human plasminogen activator inhibitor type-1 (PAI-1) in Escherichia coli as a soluble protein comprised of active and latent forms. Isolation and crystallization of latent PAI-1

Expression of human recombinant plasminogen activator inhibitor type-1 (PAI-1) in Escherichia coli has led to crystallization of ‘latent’ PAI-1. Cleavage with restriction endonucleases of a cDNA clone encoding PAI-1 yielded an 1127 base pair fragment encoding residues 2–376 of the 379 amino acid serpin. Synthetic DNA linkers were ligated to the 5′ and 3′ ends of the subclone to add an initiation codon and restore the full coding sequence, and the resulting semisynthetic gene was incorporated into an expression plasmid, pPAIST-7, under the control of the E. coli trp promoter. Transformation of E. coli GE81 with pPAIST-7 led to expression of unglycosylated PAI-1. Lysates of expression cultures contained PAI-1 activity and PAI-1 protein with the predicted M_r. Unglycosylated PAI-1 from E. coli exhibited characteristic properties of authentic PAI-1: (1) it was recovered in both active and inactive (latent) forms; (2) its activity declined during incubation at 37°C; (3) latent PAI-1 was activated by treatment with 4 M guanidine hydrochloride; (4) reactivated PAI-1 formed a detergent-stable complex with tissue plasminogen activator. Latent PAI-1 accounted for more than 85% of PAI-1 in cell lysates and was purified by ammonium sulfate fractionation, anion-exchange chromatography and hydrophobic interaction chromatography. The purified latent PAI-1 was crystallized.     

Conjugation of plasminogen activators and fibrin-specific antibodies to improve thrombolytic therapeutic agents.

Here we have reviewed chemical and recombinant approaches to the construction of hybrid molecules that combine a "targeting" antibody and an "effector" enzyme activity. There are advantages and disadvantages to both chemical and recombinant methods, and one goal of this review has been to elucidate these so that the appropriate method can be used by those interested in using hybrid molecules to study questions of basic or therapeutic importance. The system studied in greatest detail has as its goal the targeting of a plasminogen activator to an occlusive intravascular thrombus. We have, therefore, used this system as an example of currently available approaches. Now that these methodologies have been studied and put into use, it is anticipated that this principle will be generalized both to other therapeutic applications, as well as to the design and construction of molecules that will allow more basic questions to be addressed.     

Epitope mapping for four monoclonal antibodies against human plasminogen activator inhibitor type-1: implications for antibody-mediated PAI-1-neutralization and vitronectin-binding.

The inhibitory mechanism of serine proteinase inhibitors of the serpin family is based on their unique conformational flexibility. The formation of a stable proteinase-serpin complex implies insertion of the reactive centre loop of the serpin into the large central beta-sheet A and a shift in the relative positions of two groups of secondary structure elements, the smaller one including alpha-helix F. In order to elucidate this mechanism, we have used phage-display and alanine scanning mutagenesis to map the epitopes for four monoclonal antibodies against alpha-helix F and its flanking region in the serpin plasminogen activator inhibitor-1 (PAI-1). One of these is known to inhibit the reaction between PAI-1 and its target proteinases, an effect that is potentiated by vitronectin, a physiological carrier protein for PAI-1. When combined with the effects these antibodies have on PAI-1 activity, our epitope mapping points to the mobility of amino-acid residues in alpha-helix F and the loop connecting alpha-helix F and beta-strand 3A as being important for the inhibitory function of PAI-1. Although all antibodies reduced the affinity of PAI-1 for vitronectin, the potentiating effect of vitronectin on antibody-induced PAI-1 neutralization is based on formation of a ternary complex between antibody, PAI-1 and vitronectin, in which PAI-1 is maintained in a state behaving as a substrate for plasminogen activators. These results thus provide new details about serpin conformational changes and the regulation of PAI-1 by vitronectin and contribute to the necessary basis for rational design of drugs neutralizing PAI-1 in cancer and cardiovascular diseases.     

Intracellular forms of Drosophila topoisomerase II detected with monoclonal antibodies.

We developed monoclonal antibodies against Drosophila topoisomerase II and studied the intracellular forms and the in vivo and in vitro proteolytic degradation of the enzyme. In purified enzyme preparations polyclonal sera and monoclonal antibodies recognized several polypeptides in the 170-132 kD molecular weight range. In vivo, however, the pattern was much simpler. In Drosophila embryos, pupae, fly heads and Schneider S3 tissue culture cells topoisomerase II appeared as a single 166 kD polypeptide. In Drosophila embryos, with two monoclonal antibodies topoisomerase II appeared as a doublet composed of the 166 kD canonical form and a slightly higher molecular weight polypeptide. Topoisomerase II was shown to be present also in fly heads which are composed entirely of nonproliferative tissues.     

Structural differences between active forms of plasminogen activator inhibitor type 1 revealed by conformationally sensitive ligands

Plasminogen activator inhibitor type 1 (PAI-1) is a serine protease inhibitor (serpin) in which the reactive center loop (RCL) spontaneously inserts into a central beta-sheet, beta-sheet A, resulting in inactive inhibitor. Available x-ray crystallographic studies of PAI-1 in an active conformation relied on the use of stabilizing mutations. Recently it has become evident that these structural models do not adequately explain the behavior of wild-type PAI-1 (wtPAI-1) in solution. To probe the structure of native wtPAI-1, we used three conformationally sensitive ligands: the physiologic cofactor, vitronectin; a monoclonal antibody, 33B8, that binds preferentially to RCL-inserted forms of PAI-1; and RCL-mimicking peptides that insert into beta-sheet A. From patterns of interaction with wtPAI-1 and the stable mutant, 14-1B, we propose a model of the native conformation of wtPAI-1 in which the bottom of the central sheet is closed, whereas the top of the beta-sheet A is open to allow partial insertion of the RCL. Because the incorporation of RCL-mimicking peptides into wtPAI-1 is accelerated by vitronectin, we further propose that vitronectin alters the conformation of the RCL to allow increased accessibility to beta-sheet A, yielding a structural hypothesis that is contradictory to the current structural model of PAI-1 in solution and its interaction with vitronectin.     

Detection of cathepsin B, plasminogen activators and plasminogen activator inhibitor in human non-small lung cancer cell lines

Human non-small lung cancer cell lines HS-24 (established from a primary squamous cell carcinoma) and SB-3 (established from a metastasis of a primary adenocarcinoma of the lung into the adrenal gland) were analysed for the proteinases tissue-type plasminogen activator (tPA), urokinase-type plasminogen activator inhibitor (PAI-1). The proteinases were characterized by activity measurements, inhibition studies, enzyme-linked immunosorbent assay (ELISA), and Western blot analysis. Cell-associated proteinases were determined in cell lysates, secreted proteinases in cell conditioned culture media. Both cell lines were found to secrete uPA and PAI-1, whereas tPA could be detected only in HS-24 conditioned media. No cathepsin B activity could be detected in media of both cell lines. However, activation experiments and western blot analysis showed, that at least HS-24 secrete an inactive precursor. Cell lysates of HS-24 and SB-3 show PA activity, but on a low level. Cathepsin B activity was also found to be low in HS-24 lysates. However, SB-3 lysates show high cathepsin B activity. Further characterization of the proteinases by their sensitivity against several inhibitors suggests that they are similar to the corresponding proteinases of normal, nonmalignant cells.     

Isolation and characterization of monoclonal antibodies reactive with endothelial cells

Monoclonal antibodies were generated to antigens on cultured human umbilical vein endothelial cells. Spleen cells from BALB/c mice, immunized with low passage cultures of human umbilical vein endothelial cells, were fused with the non-secretory myeloma line, P3 x 63Ag 8.653. Hybridoma supernatants were screened for the desired immunological reactivity using ELISA binding assays. Hybridomas secreting antibodies reacting with the immunizing endothelial cells, but not with peripheral blood mononuclear cells, were cloned by limiting dilution and three stable clones were chosen for study. Further testing by ELISA revealed that each antibody displayed a unique pattern of reactivity. One antibody, 14E5, reacted with the macrophage-like cell line DHL-2, cultured macrophages derived from peripheral blood monocytes, and macrophages derived from malignant effusions. The antibody failed to react with fibroblasts or bovine endothelial cells. The second antibody, 12C6, reacted with human and primate fibroblasts and endothelial cells derived from bovine arteries, but not with mature macrophages. The third clone, 10B9, reacted only with the immunizing endothelial cells and the immature-macrophage line U-937. All three antibodies failed to react with long-term human B or T lymphoblastoid cell lines, leukemic cell lines, or murine macrophage lines. None of the antibodies reacted with a battery of human epithelial derived cell lines or primary cultures of human epithelial cells. Indirect immunofluorescence assays revealed that the antigens were expressed on the cell surface. These antibodies should prove useful as differentiation markers of human endothelial cells and in studies of endothelial cell function.     

Tryptophan properties in fluorescence and functional stability of plasminogen activator inhibitor 1

Plasminogen activator inhibitor 1 harbors four tryptophan residues at positions 86, 139, 175, and 262. To investigate the contribution of each tryptophan residue to the total fluorescence and to reveal the mutual interactions of the tryptophan residues and interactions with the other amino acids, 15 mutants in which tryptophan residues have been replaced by phenylalanines were constructed, purified, and characterized. Conformational distribution analysis revealed that the tryptophan mutants have a similar conformational distribution pattern as wild-type plasminogen activator inhibitor 1. Mutants in which tryptophan residue 175 was replaced by a phenylalanine displayed an increased functional half-life of the active conformation, whereas the functional half-life of mutants in which tryptophan residue 262 was replaced by a phenylalanine was substantially decreased. Comparative analysis of the fluorescence lifetimes, the extinction coefficients, and the quantum yields of the individual tryptophan residues demonstrates that tryptophan residue 262 gives the highest contribution to the total fluorescence. The other tryptophan residues have a very low quantum yield. In the wild-type protein, the fluorescence of all tryptophan residues is partially quenched as compared to the mutants that contain single tryptophan residues, due to conformational effects. The fluorescence of tryptophan residue 262 is very likely also partially quenched by energy transfer to tryptophan residue 175.