Search

The direct binding of human plasminogen (hPg), via its kringle-2 domain (K2 hPg ), to streptococcal M-protein (PAM), largely contributes to the pathogenesis of Pattern D Group A Streptococcus pyogenes (GAS). However, the mechanism of complex formation is unknown. In a system consisting of a Class II PAM from Pattern D GAS isolate NS88.2 (PAM NS88.2 ), with one K2 hPg binding a-repeat in its A-domain, we employed biophysical techniques to analyze the mechanism of the K2 hPg /PAM NS88.2 interaction. We show that apo-PAM NS88.2 is a coiled-coil homodimer (M.Wt. ~80 kDa) at 4°C-25°C, and is monomeric (M.Wt. ~40 kDa) at 37°C, demonstrating a temperature-dependent dissociation of PAM NS88.2 over a narrow temperature range. PAM NS88.2 displayed a single tight binding site for K2 hPg at 4°C, which progressively increased at 25°C through 37°C. We isolated the K2 hPg /PAM NS88.2 complexes at 4°C, 25°C, and 37°C and found molecular weights of ~50 kDa at each temperature, corresponding to a 1:1 (m:m) K2 hPg /PAM NS88.2  monomer complex. hPg activation experiments by streptokinase demonstrated that the hPg/PAM NS88.2  monomer complexes are fully functional. The data show that PAM dimers dissociate into functional monomers at physiological temperatures or when presented with the active hPg module (K2 hPg ) showing that PAM is a functional monomer at 37°C., (© 2021 The Authors. MicrobiologyOpen published by John Wiley & Sons Ltd.)

To understand the role of substrate plasminogen kringles in its differential catalytic processing by the streptokinase - human plasmin (SK-HPN) activator enzyme, Fluorescence Resonance Energy Transfer (FRET) model was generated between the donor labeled activator enzyme and the acceptor labeled substrate plasminogen (for both kringle rich Lys plasminogen - LysPG, and kringle less microplasminogen - µPG as substrates). Different steps of plasminogen to plasmin catalysis i.e. substrate plasminogen docking to scissile peptide bond cleavage, chemical transformation into proteolytically active product, and the decoupling of the nascent product from the SK-HPN activator enzyme were segregated selectively using (1) FRET signal as a proximity sensor to score the interactions between the substrate and the activator during the cycle of catalysis, (2) active site titration studies and (3) kinetics of peptide bond cleavage in the substrate. Remarkably, active site titration studies and the kinetics of peptide bond cleavage have shown that post docking chemical transformation of the substrate into the product is independent of kringles adjacent to the catalytic domain (CD). Stopped-flow based rapid mixing experiments for kringle rich and kringle less substrate plasminogen derivatives under substrate saturating and single cycle turnover conditions have shown that the presence of kringle domains adjacent to the CD in the macromolecular substrate contributes by selectively speeding up the final step, namely the product release/expulsion step of catalysis by the streptokinase-plasmin(ogen) activator enzyme., (© 2020 The Author(s). Published by Portland Press Limited on behalf of the Biochemical Society.)

Hereditary angioedema (HAE) is a genetically heterogeneous disease that is characterized by recurrent skin swelling, abdominal pain attacks, and potentially life-threatening upper airway obstruction. The two classic types, HAE types I and II, are both caused by mutations in the complement C1 inhibitor (SERPING1) gene resulting either in a quantitative or a qualitative deficiency of C1 inhibitor. In so-called HAE type III, in contrast, patients show normal C1 inhibitor measurements in plasma ('HAE with normal C1 inhibitor'). As previously shown by us, one subgroup of 'HAE with normal C1 inhibitor' is caused by mutations of the coagulation factor XII (F12) gene. For the present study, following the exclusion of numerous candidate genes, we screened eight unrelated index patients representing eight 'HAE families with normal C1 inhibitor and no F12 mutation' for mutations in the plasminogen (PLG) gene. A rare non-conservative missense mutation was newly identified in exon 9 of the PLG gene. This mutation (c.1100A > G), encountered in three out of eight patients, predicts a lysine-to-glutamic acid substitution in position 311 of the mature protein (p.Lys311Glu). Using isoelectric focusing of plasma samples followed by an immunoblotting procedure we demonstrated that the presence of the mutation is associated with a dysplasminogenemia, namely the presence of an aberrant plasminogen protein. The predicted structural and functional impact of the mutation, its absence in 139 control individuals, and its co-segregation with the phenotype in three large families provide strong support that it causes disease. Extending a previously proposed gene-based alphabetic nomenclature for the various HAE types one may use the term 'HAE type C' for the HAE entity described here., (Copyright © 2017. Published by Elsevier Inc.)

Streptokinase (SK) is a protein co-factor with a potent capability for human plasminogen (HPG) activation. Our previous studies [1] have indicated a major role of long-range protein-protein contacts between the three domains (alpha, beta, and gamma) of SK and the multi-domain HPG substrate (K1-K5CD). To further explore this phenomenon, we prepared truncated derivatives of HPG with progressive removal of kringle domains, like K5CD, K4K5CD, K3-K5CD (K3K4K5CD), K2-K5CD (K2K3K4K5CD) and K1-K5CD (K1K2K3K4K5CD). While urokinase (uPA) cleaved the scissile peptide in the isolated catalytic domain (μPG) with nearly the same rate as with full-length HPG, SK-plasmin showed only 1-2% activity, revealing mutually distinct mechanisms of HPG catalysis between the eukaryotic and prokaryotic activators. Remarkably, with SK.HPN (plasmin), the 'addition' of both kringles 4 and 5 onto the catalytic domain showed catalytic rates comparable to full length HPG, thus identifying the dependency of the "long-range" enzyme-substrate interactions onto these two CD-proximal domains. Further, chimeric variants of K5CD were generated by swapping the kringle domains of HPG with those of uPA and TPA (tissue plasminogen activator), separately. Surprisingly, although native-like catalytic turnover rates were retained when either K1, K2 or K4 of HPG was substituted at the K5 position in K5CD, these were invariably lost once substituted with the evolutionarily more distant TPA- and uPA-derived kringles. The present results unveil a novel mechanism of SK.HPN action in which augmented catalysis occurs through enzyme-substrate interactions centered on regions in substrate HPG (kringles 4 and 5) that are spatially distant from the scissile peptide bond., (Copyright © 2011 Elsevier B.V. All rights reserved.)

Many proteins in the fibrinolysis pathway contain antiangiogenic kringle domains. Owing to the high degree of homology between kringle domains, there has been a safety concern that antiangiogenic kringles could interact with common kringle proteins during fibrinolysis leading to adverse effects in vivo. To address this issue, we investigated the effects of several antiangiogenic kringle proteins including angiostatin, apolipoprotein(a) kringles IV(9)-IV(10)-V (LK68), apolipoprotein(a) kringle V (rhLK8) and a derivative of rhLK8 mutated to produce a functional lysine-binding site (Lys-rhLK8) on the entire fibrinolytic process in vitro and analyzed the role of lysine binding. Angiostatin, LK68 and Lys-rhLK8 increased clot lysis time in a dose-dependent manner, inhibited tissue-type plasminogen activator-mediated plasminogen activation on a thrombin-modified fibrinogen (TMF) surface, showed binding to TMF and significantly decreased the amount of plasminogen bound to TMF. The inhibition of fibrinolysis by these proteins appears to be dependent on their functional lysine-binding sites. However, rhLK8 had no effect on these processes owing to an inability to bind lysine. Collectively, these results indicate that antiangiogenic kringles without lysine binding sites might be safer with respect to physiological fibrinolysis than lysine-binding antiangiogenic kringles. However, the clinical significance of these findings will require further validation in vivo.

Human plasminogen kringle 3 (hPgn K3) domain contains most elements of the canonical lysine-binding site (LBS) found in other Pgn kringles. However, it does not exhibit affinity for either lysine or structurally related zwitterionic ligands. It has been shown that lysine-binding activity can be engineered via a Lys57 --> Asp mutation [Burgin, J., and Schaller, J. (2009) Cell. Mol. Life Sci. 55, 135]. Using a recombinant construct expressed in Escherichia coli, the three-dimensional solution structure of hPgn K3 was determined via NMR spectroscopy [heavy atom averaged rmsd = 0.35 +/- 0.07 A (backbone) and 0.75 +/- 0.12 A (all)]. The (1)H/(15)N heteronuclear single-quantum correlated (HSQC) spectra for both wild-type K3 and mutated [r(K57D)K3] structures are essentially identical, implying that the two structures are effectively isomorphous. The affinity of r(K57D)K3 for the lysine analogue trans-(aminomethyl)cyclohexanecarboxylic acid (AMCHA) was investigated from ligand-induced NMR chemical shift perturbations, which enabled for mapping the binding site on the mutated domain surface. The equilibrium association constant, K(a), was determined to be approximately 5.23 +/- 0.03 mM(-1). Homology modeling combined with in silico docking of lysine-like zwitterionic ligands via AutoDock 4.0 supports functionality of the engineered (K57D)K3 LBS, whose electrostatic focal centers are defined by the Arg36/Arg71 cationic and Asp55/Asp57 anionic pairs. Comparison of K3-type sequences from different vertebrates, including kringles from hedgehog apolipoprotein(a) [Apo(a)] and Apo(a)-related (Arp) sequences, reveals that Lys57 is confined to the hPgn variant. Based on the likely phylogeny and ligand affinities of the homologous domains, it is suggested that the hPgn K3 is unique in that all other K3-type domains, including hedgehog Apo(a) and all Arp domains, except K3(1), are predicted to variously exhibit lysine-binding capability. In Arp K3(1) an Arg residue fills site 72, replacing the key aromatic residue found in other kringles, thus interfering with a requisite kringle-ligand hydrophobic interaction.

Streptokinase (SK) conformationally activates the central zymogen of the fibrinolytic system, plasminogen (Pg). The SK.Pg* catalytic complex binds Pg as a specific substrate and cleaves it into plasmin (Pm), which binds SK to form the SK.Pm complex that propagates Pm generation. Catalytic complex formation is dependent on lysine-binding site (LBS) interactions between a Pg/Pm kringle and the SK COOH-terminal Lys(414). Pg substrate recognition is also LBS-dependent, but the kringle and SK structural element(s) responsible have not been identified. SK mutants lacking Lys(414) with Ala substitutions of charged residues in the SK beta-domain 250-loop were evaluated in kinetic studies that resolved conformational and proteolytic Pg activation. Activation of [Lys]Pg and mini-Pg (containing only kringle 5 of Pg) by SK with Ala substitutions of Arg(253), Lys(256), and Lys(257) showed decreases in the bimolecular rate constant for Pm generation, with nearly total inhibition for the SK Lys(256)/Lys(257) double mutant. Binding of bovine Pg (BPg) to the SK.Pm complex containing fluorescently labeled Pm demonstrated LBS-dependent assembly of a SK.labeled Pm.BPg ternary complex, whereas BPg did not bind to the complex containing the SK Lys(256)/Lys(257) mutant. BPg was activated by SK.Pm with a K(m) indistinguishable from the K(D) for BPg binding to form the ternary complex, whereas the SK Lys(256)/Lys(257) mutant did not support BPg activation. We conclude that SK residues Arg(253), Lys(256), and Lys(257) mediate Pg substrate recognition through kringle 5 of the [Lys]Pg and mini-Pg substrates. A molecular model of the SK.kringle 5 complex identifies the putative interactions involved in LBS-dependent Pg substrate recognition.

Prion diseases are a group of fatal neurodegenerative disorders characterized by the accumulation of a misfolded form (PrP(Sc)) of the cellular prion protein (PrP(C)) in the brains of affected individuals. The conversion of PrP(C) to PrP(Sc) is thought to involve a change in protein conformation from a normal, primarily alpha-helical structure into a beta-sheet conformer. Few proteins have been identified that differentially interact with the two forms of PrP. It has been reported that plasminogen binds to PrP(Sc) from a variety of prion phenotypes. We have examined potential motifs within the kringle region that may be responsible for binding to PrP. We synthesized 12-15-mer peptides that contain small, repetitive stretches of amino acid residues found within the kringle domains of plasminogen. These synthetic peptides were found to capture PrP(Sc) from the brain homogenates of bovine spongiform encephalopathy affected cattle, chronic wasting disease affected elk, experimental scrapie of hamsters and that of subjects affected by Creutzfeldt-Jakob disease, without binding to PrP(C) in unaffected controls. Therefore, we have identified critical peptide motifs that may be important for protein-protein interactions in prion disease pathogenesis. The ability of these synthetic peptides to bind preferentially to PrP(Sc) suggests a potential application in the diagnosis of prion diseases.

VEK-30, a 30-amino acid internal peptide present within a streptococcal M-like plasminogen (Pg)-binding protein (PAM) from Gram-positive group-A streptococci (GAS), represents an epitope within PAM that shows high affinity for the lysine binding site (LBS) of the kringle-2 (K2) domain of human (h)Pg. VEK-30 does not interact with this same region of mouse (m)Pg, despite the high conservation of the mK2- and hK2-LBS. To identify the molecular basis for the species specificity of this interaction, hPg and mPg variants were generated, including an hPg chimera with the mK2 sequence and an mPg chimera containing the hK2 sequence. The binding of synthetic VEK-30 to these variants was studied by surface plasmon resonance. The data revealed that, in otherwise intact Pg, the species specificity of VEK-30 binding in these two cases is entirely dictated by two K2 residues that are different between hPg and mPg, namely, Arg-220 of hPg, which is a Gly in mPg, and Leu-222 of hPg, which is a Pro in mPg, neither of which are members of the canonical K2-LBS. Neither the activation of hPg, nor the enzymatic activity of its activated product, plasmin (hPm), are compromised by replacing these two amino acids by their murine counterparts. It is also demonstrated that hPg is more susceptible to activation to hPm after complexation with VEK-30 and that this property is greatly reduced as a result of the R220G and L222P replacements in hPg. These mechanisms for accumulation of protease activity on GAS likely contribute to the virulence of PAM(+)-GAS strains and identify targets for new therapeutic interventions.

Kringle1-5 (K1-5), a proteolytic fragment containing five kringle domains of human plasminogen generated by plasmin-mediated proteolysis, has been already identified by Cao et al. with relation to anti-angiogenesis and proliferation of endothelial cells. To investigate anti-angiogenesis activity of recombinant human K1-5 (rhK1-5) expressed in Escherichia coli BL21, the cDNA of human K1-5 obtained from cloning vector pUC57-K1-5 by PCR, was inserted into an expression vector pET30(+) to construct a prokaryotic expression vector pET-K1-5. Recombinant K1-5 efficiently expressed in E. coli BL21 after IPTG induction was monitored by SDS-PAGE and Western blotting with an anti-angiostatin monoclonal antibody and an anti-hexahistidine tag antibody. The expressed K1-5 accounted for approximately 32% of the total bacterial proteins as estimated by densitometry, and existed mainly as inclusion bodies. The inclusion bodies were washed, lysed, purified, and refolded to a purity of 96% as estimated by capillary electrophoresis and the final purification yield of K1-5 in E. coli system was approximately 5.8 mg/L. Purified K1-5 protein was tested on chicken embryo chorioallantoic membranes (CAMs), and a large number of newly formed blood vessels were significantly regressed. In the present study, we demonstrated that bacterial-expressed K1-5 effectively inhibited angiogenesis of the chicken embryo in a dose-dependent manner through CAM assay. In addition, the rhK1-5 potently inhibited endothelial cell proliferation but not non-endothelial cells. For the first time, these findings demonstrate that the rhK1-5 produced by a prokaryote expression system effectively inhibited angiogenesis of the chicken embryo in a dose-dependent manner and specially suppressed in vitro the proliferation of human umbilical vein endothelial cells. This fact derived from the present study further suggests the rhK1-5 can be used for anti-angiogenesis therapy of cancer.

Kringle 5, a proteolytic fragment of human plasminogen has been shown to potently inhibit angiogenesis. The tetrapeptide KLYD derived from kringle 5 has been shown to capture many activities of kringle 5 in vitro. Further simplification has been achieved by replacement of the two central amino acids with a 4-aminobenzoic acid spacer group. Molecules displaying the required recognition groups on this core show similar in vitro properties to kringle 5, and are able to displace radiolabeled protein from a high affinity binding site on endothelial cells.

Three peptide fragments of the kringle 5 region of plasminogen and their respective N- and C-terminus maleimido derivatives conjugated to Cys34 of human serum albumin were evaluated in vitro using a human umbilical vein endothelial cell (HUVEC) migration assay and a human plasma stability assay. The N-terminus maleimido derivative of the 64 to 74 segment of kringle 5 conjugated to human serum albumin possessed remarkable anti-migratory activity.

Objective: To obtain purified deletion mutant of plasminogen kringle 5 (K5) using gene mutation and genetic recombination methods and assess its anti-angiogenic activity in vitro., Methods: A deletion mutant of K5 was obtained by deleting 15 amino acids from K5 while retaining all the 3 disulfide bonds. This K5 mutant (Mut1) was expressed in E. coli and affinity purified. The inhibition effect of K5 Mut1 on primary retinal capillary endothelial cells and pericytes from the same origin was assessed by MTT assay., Results: The K5 Mut1 inhibited the proliferation of primary retinal capillary endothelial cells in a concentration-dependent manner, with an apparent half-inhibition concentration (EC(50)) of approximately 35 nmol/L, which was 2-fold more potent than intact K5. In the same concentration range, this peptide did not inhibit pericytes from the same origin, suggesting an endothelial cell-specific inhibition., Conclusion: This K5 deletion mutant is a more potent angiogenic inhibitor than K5 and may have therapeutic potential in the treatment of such disorders with abnormal neovascularization as diabetic retinopathy, age-related macular degeneration and solid tumor.

The cDNA encoding Kringle 1-5 domains of human plasminogen (designated as K1-5), obtained from HepG2 by RT-PCR, was cloned into expression vector pHIL-S1. The recombinant plasmid pHIL-K1-5 was transformed into Pichia pastoris GS115 and the recombinant yeast was induced by methanol to express the recombinant protein. The expressed protein was purified by lysine affinity chromatography. The recombinant K1-5 inhibited the growth of bovine aortic endothelial cells (BAEC) stimulated by the basic fibroblast growth factor (bFGF), in a dosage-dependent manner, with a half maximal concentration of 14 mg/L. And rhK1-5 inhibited 47% of the BAEC migration stimulated by bFGF at the concentration of 50 mg/L. rhK1-5 also affected the cell cycle of BAEC and caused G(0)-G(1) arrest at the concentration of 14 mg/L.

The cDNA encoding Kringle 1-4 and part of Kringle 5 domains of human plasminogen (K1-4.5), obtained from HepG2 by RT-PCR, was cloned into expression vector pHIL-S1. The recombinant plasmid pHIL-K1-4.5 was transformed into Pichi pastoris GS115 and the recombinant yeast was induced to express the recombinant proteins by methanol. The expressed proteins were purified by lysine affinity chromatography to a purity of 95%. The recombinant K1-4.5 inhibited the growth of bovine capillary endothelial cells (BAEC) stimulated by the basic fibroblast growth factor (bFGF), in a dosage-dependent manner with a half maximal concentration of 2 mg/L. rhK1-4.5 also inhibited 40% of the BAEC migration stimulated by bFGF in the concentration of 1 mg/L.

Neovascular diseases of the retina include age-related macular degeneration and diabetic retinopathy, and together they comprise the leading causes of adult-onset blindness in developed countries. Current surgical, pharmaceutical, and laser therapies for age-related macular degeneration (AMD) rarely result in improved vision, do not significantly prevent neovascularization (NV), and often result in at least some vision loss. To address this therapeutic gap, we determined the efficacy of recombinant adeno-associated viral (rAAV) serotype-2-mediated expression of pigment epithelium-derived factor (PEDF) or Kringle domains 1-3 of angiostatin (K1K3) in reducing aberrant vessel formation in a mouse model of ischemia-induced retinal NV. Both PEDF and K1K3 are potent inhibitors of NV when injected directly, hence expression of these therapeutic factors from rAAV may provide long-term protection from neovascular eye disease. rAAV vectors expressing the therapeutic gene were injected into one eye of postnatal day 0 (P0) newborn mouse pups. Retinal NV was induced in P7 mice by exposure to elevated oxygen for 5 days followed by room air for another five days. Retinal NV was quantified by the number of vascular-endothelial-cell nuclei above the inner-limiting membrane in P17 eyes. The number of such vascular endothelial cell nuclei in eyes treated with rAAV-PEDF or rAAV-K1K3 was significantly reduced (both P < 0.0000002) compared with control eyes. Ocular protein levels detected by ELISA correlate well with the reduction in NV and confirm that expression of antineovascular agents from rAAV vectors may be a therapeutically useful treatment of retinal or choroidal neovascular disease.

Monoclonal antibodies directed against recombinant apolipoprotein (a) (r-apo(a)) lacking plasminogen-like KIV-2 repeats were used to identify structurally related conformational epitopes in various members of the plasminogen-prothrombin gene family. A number of procedures including a fibrin-binding inhibition immunoassay and surface plasmon resonance studies were used. Two antibodies (A10.1 and A10.4) recognised common conformational structures in r-apo(a), prothrombin, factor XII, plasminogen and its tissue-type and urokinase-type activators. In contrast, two other antibodies recognised specifically an epitope comprising residues of the lysine-binding site (A10.2) or close to it (A10.5) and inhibited the fibrin-binding function of r-apo(a) (IC(50)=36 pmol/l and 9.76 nmol/l, respectively). Interestingly, these antibodies distinctly recognised the elastase-derived fragments of plasminogen K4 (A10.2) and K1+2+3 (A10.5) without affecting plasminogen binding to fibrin. These results suggest that highly conserved conformational regions are common to various proteins of the plasminogen-prothrombin gene family and are in agreement with the concept that these proteins constitute a monophyletic group derived from an ancestral gene.

The X-ray crystal structure of a complex of a modified recombinant kringle-2 domain of human plasminogen, K2Pg[C4G/E56D/L72Y] (mK2Pg), containing an upregulated lysine-binding site, bound to a functional 30 residue internal peptide (VEK-30) from an M-type protein of a group A Streptococcus surface protein, has been determined by molecular replacement methods using K4Pg as a model, and refined at 2.7 A resolution to a R-factor of 19.5 %. The X-ray crystal structure shows that VEK-30 exists as a nearly end-to-end alpha-helix in the complex with mK2Pg. The final structure also revealed that Arg17 and His18 of VEK-30 served as cationic loci for Asp54 and Asp56 of the consensus lysine-binding site of mK2Pg, while Glu20 of VEK-30 coordinates with Arg69 of the cationic binding site of mK2Pg. The hydrophobic ligand-binding pocket in mK2Pg, consisting primarily of Trp60 and Trp70, situated between the positive and negative centers of the lysine-binding site, is utilized in a novel manner in stabilizing the interaction with VEK-30 by forming a cation-pi-electron-mediated association with the positive side-chain of Arg17 of this peptide. Additional lysine-binding sites, as well as exosite electrostatic and hydrogen bonding interactions involving Glu9 and Lys14 of VEK-30, were observed in the structural model. The importance of these interactions were tested in solution by investigating the binding constants of synthetic variants of VEK-30 to mK2Pg, and it was found that, Lys14, Arg17, His18, and Glu20 of VEK-30 were the most critical amino acid binding determinants. With regard to the solution studies, circular dichroism analysis of the titration of VEK-30 with mK2Pg demonstrated that the peptidic alpha-helical structure increased substantially when bound to the kringle module, in agreement with the X-ray results. This investigation is the first to delineate structurally the mode of interaction of the lysine-binding site of a kringle with an internal pseudo-lysine residue of a peptide or protein that functionally interacts with a kringle module, and serves as a paradigm for this important class of interactions., (Copyright 2001 Academic Press.)

The DNA sequence coding for plasminogen kringle 5 (pK5), an inhibitor of angiogenesis, was fused with that coding for interferon gamma and over-produced in the form of inactive inclusion bodies in E. coli. The amount of fusion protein was about 40% of total protein produced. The fusion protein contained in the inclusion bodies was solubilized in 8 M urea and purified by anion-exchange chromatography. We employed the orthogonal experimental design L16(4(5)) (5 factors, 4 levels, 16 experiments) procedure for researching the influence of denaturant, aggregation suppressor L-arginine, NaCl, pH, and glycine on the refolding procedure. Our results suggest that the presence of appropriate L-arginine, NaCl, and denaturant in the refolding buffer inhibits the aggregation of the fusion protein and increases the yield of renatured protein with biological activity. The refolded fusion protein, gammaIFN/pk5, has in vitro anti-endothelial cell proliferation activity.

Hepatocyte growth factor (HGF), also known as scatter factor, is a mesenchymal or stromal-derived mediator with angiogenic activity. There are four kringle domains in its amino terminus. They display considerable sequence similarity with those of angiostatin, an angiogenesis inhibitor. We now describe that the recombinant kringle1 of HGF (HGFK1) inhibits bovine aortic endothelial (BAE) cell proliferation stimulated by basic fibroblast growth factor in a dose-dependent manner, with an ED(50) of approximately 0.7 microg/ml, while ED(50) of angiostatin is 3 microg/ml. Treatment of BAE cell with HGFK1 caused cell apoptosis. This report thus constitutes the first demonstration that kringle1 of HGF is a selective inhibitor for BAE cell proliferation stimulated by bFGF., (Copyright 2000 Academic Press.)

The internal motions of the backbone nitrogen atoms of the kringle 1 domain of human plasminogen (K1(Pg)) were examined in the absence and presence of the ligand, epsilon-aminocaproic acid. These dynamic properties were determined from (15)N NMR relaxation data in terms of the extended model-free parameters. The model of isotropic reorientation was found sufficient to account for overall molecular tumbling for both apo and EACA-bound K1(Pg). The global rotational correlation time (tau(m)) for apo-K1(Pg) was 5.87(+/-0.01) ns, while the tau(m) for ligand-bound K1(Pg) was 5.20(+/-0.01) ns, suggesting that perhaps some small degree of aggregation occurred in the apo form of the kringle module. Complexation of K1(Pg) with ligand mainly reduced those internal motions that occurred on a 100 ps to 5 ns time-scale. The magnitude of the chemical exchange was also attenuated upon ligand binding. These data are consistent with studies employing other approaches that suggest that the binding pocket is preformed in K1(Pg)., (Copyright 2000 Academic Press.)

Kringle domains are found in a number of proteins where they govern protein-protein interactions. These interactions are often sensitive to lysine and lysine analogues, and the kringle-lysine interaction has been used as a model system for investigating kringle-protein interactions. In this study, we analyze the interaction of wild-type and six single-residue mutants of recombinant plasminogen kringle 4 expressed in Escherichia coli with the recombinant C-type lectin domain of tetranectin and trans-aminomethyl-cyclohexanoic acid (t-AMCHA) using isothermal titration calorimetry. We find that all amino acid residues of plasminogen kringle 4 found to be involved in t-AMCHA binding are also involved in binding tetranectin. Notably, one amino acid residue of plasminogen kringle 4, Arg 32, not involved in binding t-AMCHA, is critical for binding tetranectin. We also find that Asp 57 and Asp 55 of plasminogen kringle 4, which both were found to interact with the low molecular weight ligand with an almost identical geometry in the crystal of the complex, are not of equal functional importance in t-AMCHA binding. Mutating Asp 57 to an Asn totally eliminates binding, whereas the Asp 55 to Asn, like the Arg 71 to Gln mutation, was found only to decrease affinity.

Lysine side chains induce conformational changes in plasminogen (Pg) that regulate the process of fibrinolysis or blood clot dissolution. A lysine side-chain mimic, epsilon amino caproic acid (EACA), enhances the activation of Pg by urinary-type and tissue-type Pg activators but inhibits Pg activation induced by streptokinase (SK). Our studies of the mechanism of this inhibition revealed that EACA (IC(50) 10 microM) also potently blocked amidolytic activity by SK and Pg at doses nearly 10000-fold lower than that required to inhibit the amidolytic activity of plasmin. Different Pg fragments were used to assess the role of the kringles in mediating the inhibitory effects of EACA: mini-Pg which lacks kringles 1-4 of Glu-Pg and micro-Pg which lacks all kringles and contains only the catalytic domain. SK bound with similar affinities to Glu-Pg (K(A) = 2.3 x 10(9) M(-1)) and to mini-Pg (K(A) = 3.8 x 10(9) M(-)(1)) but with significantly lower affinity to micro-Pg (K(A) = 6 x 10(7) M(-)(1)). EACA potently inhibited the binding of Glu-Pg to SK (K(i) = 5.7 microM), but was less potent (K(i) = 81.1 microM) for inhibiting the binding of mini-Pg to SK and had no significant inhibitory effects on the binding of micro-Pg and SK. In assays simulating substrate binding, EACA also potently inhibited the binding of Glu-Pg to the SK-Glu-Pg activator complex, but had negligible effects on micro-Pg binding. Taken together, these studies indicate that EACA inhibits Pg activation by blocking activator complex formation and substrate binding, through a kringle-dependent mechanism. Thus, in addition to interactions between SK and the protease domain, interactions between SK and the kringle domain(s) play a key role in Pg activation.

Kringle 1-3 of human plasminogen is a potent inhibitor of endothelial cell proliferation. To understand a possible role for the unique cystine bridge between kringle 2 and kringle 3, we disrupted the interkringle disulfide bond by mutating Cys(169) and Cys(297) to serine residues. The yield of the mutant during the refolding process was decreased significantly. Anti-endothelial cell proliferative activity of the mutant was similar to that of the wild type. There was no significant difference in in vivo antiangiogenic activity between the wild type and the mutant in chorioallantoic membrane assay. However, in the mutant, the weak lysine binding capability of kringle 2 was not detected and its mobility in nonreducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis is different from that of the wild type. These results support the notion that the overall antiangiogenic function of angiostatin is mediated by individual kringles, and suggest that the lysine binding capability of kringle 2 is likely not important for the antiangiogenic activity of kringle 1-3., (Copyright 2000 Academic Press.)

Phylogenetic analysis of protease domains of the vertebrate plasminogen-prothrombin family revealed two major subfamilies: (1) a subfamily containing macrophage-stimulating protein (MSP), hepatocyte growth factor (HGF), plasminogen, and apolipoprotein(a) (APOA); and (2) a subfamily containing prothrombin, HGF activator, and plasminogen activators. There was evidence that these two subfamilies diverged prior to the divergence of amphibians and amniotes. The phylogeny indicated a close relationship of APOA from the European hedgehog, rhesus monkey, and human with plasminogen. Phylogenetic analysis of repeated kringle domains supported the hypothesis that APOA evolved independently in hedgehog and primates through numerous duplications of different kringle domains of the ancestral plasminogen. Phylogenies of kringle domains revealed two modes of evolution: (1) a conservative mode, whereby duplication of kringle domains occurred prior to cladogenesis and the same kringle structure has been maintained in different lineages (exemplified by plasminogen and prothrombin); and (2) a concerted mode, whereby kringle domains have duplicated since cladogenesis and thus orthologous relationships do not exist between kringles of different lineages (exemplified by APOA)., (Copyright 2000 Academic Press.)

The kringle 2 (K2) module of human plasminogen (Pgn) binds L-lysine and analogous zwitterionic compounds, such as the antifibronolytic agent trans-(aminomethyl)cyclohexanecarboxylic acid (AMCHA). Far-UV CD and NMR spectra reveal little conformational change in K2 upon ligand binding. However, retarded (1)H-(2)H isotope exchange kinetics induced by AMCHA indicate stabilization of the K2 conformation by the ligand. Assessment of secondary structure content from CD spectra yields approximately 26% beta-STRAND, approximately 13% beta-TURN, approximately 15% 3(1)-HELIX, and approximately 6% 3(10)-HELIX. The NMR solution conformation of the K2 domain complexed to AMCHA has been determined [heavy atom rmsd = 0.49 +/- 0.09A (BACKBONE) AND 1.02+/- 0.08 (ALL)]. The K2 molecule has overall dimensions of approximately 34.5A times approximately 33.4A times approximately 22.7A . Analogous with the polypeptide outline of homologous domains, K2 contains three short antiparallel beta-sheets (paired strands 15-16/20-21, 24-25/48-49, and 62-64/72-74) and four defined beta-turns (residues 6-9, 16-19, 53-56, AND 67-70). Consistent with the CD analysis, albeit novel in the context of kringle folding, the NMR structure reveals an unpaired beta-strand structured by residues 30-32, a turn of 3(10)-helix compromising residues 38-41, and a 3(1)-helix for residues 21-24 and 74-79. We also identify alignable 3(1)-helices in previously reported homologous kringle structures. Rather high order parameter S(2) values (= approximately 0.85 +/- 0.04) characterize the K2 backbone dynamics. The lowest flexibility is observed for the two inner loop segments of residues 51-63 AND 63-75 (= approximately 0.86-0.87 +/- 0.03). Overhauser connectivities reveal close hydrophobic contacts of the ligand ring with side chains of Tyr(36), Trp(62), Phe(64), Trp(72), AND Leu(74). In most K2 structures, the N atom of AMCHA places itself approximately 3.9 and 4.4A from the anionic groups of Glu(57) and Asp(55), respectively, while its carboxylate group, H-bonded to the Tyr(36) side chain OH(eta), ion-pairs the Arg(71) guanidinium group. Consistent with the preference of K2 for binding 5-aminopentanoic acid over 6-aminohexanoic acid, the positions of the ionic centers within the K2 binding site approach each other approximately 1A closer relative to what is observed in lysine binding sites of homologous Pgn modules.

Although the therapeutic value of angiostatin, a proteolytic fragment of plasminogen, has been recognized for the treatment of cancer, the production of bioactive angiostatin remains a difficult task. Here we report that expression of a cDNA encoding a secreted, four-kringle human angiostatin inhibited tumor growth of B16F10 melanoma cells in mice but did not suppress tumor cell growth in culture. After transfection and selection, stable expression of the angiostatin cDNA was demonstrated in several B16F10 clones by quantitative mRNA analysis using the Taqman method. Cells that expressed angiostatin at either a low, medium, or high level were injected into C57BL/6 mice. s.c. Growth of B16F10 tumors was diminished by the angiostatin transgene, and the inhibition was directly proportional to the expression level of angiostatin in the transfected cells. However, suppression of s.c. tumor growth was transient, and eventually, tumors emerged with a strongly decreased expression of the transgene. Angiostatin expression also reduced lung metastasis from i.v.-injected B16F10 cells. Our data indicate that a cDNA encoding bioactive human angiostatin is potentially useful for gene therapy of human cancers, but the delivery of the transgene may require repeated dosing to achieve sustained dormancy of primary tumors and cancer metastases.

Angiostatin protein, which comprises the first four kringle domains of plasminogen, is an endogenous inhibitor of angiogenesis that inhibits the growth of experimental primary and metastatic tumors. Truncation of Angiostatin K1-4 to K1-3 retained the activity of Angiostatin. We recombinantly expressed full-length human Angiostatin protein corresponding to the first four kringle domains of human plasminogen and a truncated form of the Angiostatin protein, kringles 1-3. Purified recombinant Angiostatin K1-3 and K1-4 proteins inhibited the formation of experimental B16-BL6 lung metastases by greater than 80% when administered at 30 nmol/kg/day. We demonstrate for the first time that Angiostatin protein, consisting of the first three kringle domains of human plasminogen, has in vivo biological activity in this assay indistinguishable from that of the full-length Angiostatin K1-4 protein and that the fourth kringle of plasminogen, when linked in sequence to K1-3, plays no direct role in the antitumor activity of Angiostatin., (Copyright 1999 Academic Press.)

Lipoprotein(a) [Lp(a)] is associated with atherosclerosis and with disease processes involving thrombosis. Lp(a) contains apoprotein (a) [apo(a)], which has a sequence highly homologous to plasminogen. Hence, Lp(a) binds directly to extracellular matrix, cellular plasminogen receptors and fibrin(ogen) and competes for the binding of plasminogen to these regulatory surfaces. These interactions may contribute to the proatherothrombogenic consequences of high Lp(a) levels. These interactions are mediated by lysine binding sites (LBS). Therefore, we examined the role of apo(a) kringle IV-10 [the only apo(a) kringle demonstrated to exhibit lysine binding activity in the intact lipoprotein] in the interaction of Lp(a) with these regulatory molecules. We have compared directly apo(a) KIV-10 with plasminogen K4 to examine whether these highly structurally homologous kringle modules are also functionally homologous. Futhermore, because the plasminogen K5-protease domain (K5-PD) binds directly to fibrin, we have also examined the ability of this plasminogen fragment to inhibit the interaction of Lp(a) with these regulatory molecules and with extracellular matrix. Apo(a) KIV-10 competed effectively for the binding of 125I-Lp(a) to these surfaces but was less effective than either intact Lp(a), plasminogen K4 or plasminogen. Plasminogen KS-PD was a better competitor than apo(a) KIV-10 for 125I-Lp(a) binding to the representative extracellular matrix, Matrigel, and to plasmin-treated fibrinogen. In contrast, plasminogen K5-PD did not compete for the interaction of Lp(a) with cells, although it effectively competed for plasminogen binding. These results suggest that Lp(a) recognizes sites in all of the regulatory molecules that are also recognized by apo(a) KIV-10 and that Lp(a) recognizes sites in extracellular matrix and in plasmin-modified fibrinogen that also are recognized by plasminogen K5-PD. Thus, the interaction of Lp(a) with cells is clearly distinct from that with extracellular matrix and with plasmin-treated fibrinogen and the recognition sites within Lp(a) and plasminogen for these regulatory molecules are not identical.

Interactions between the kringle 4 (K4) domain of human plasminogen (Pgn) and segments of the N-terminal Glu1-Lys77 peptide (NTP) have been investigated via 1H-NMR at 500 MHz. NTP peptide stretches devoid of Lys residues but carrying an internal Arg residue show negligible affinity toward K4 (equilibrium association constant Ka < 0.05 mM(-1)). In contrast, while most fragments containing an internal Lys residue exhibit affinities comparable to that shown by the blocked Lys derivative Nalpha-acetyl-L-lysine-methyl ester (Ka approximately 0.2 mM(-1), peptides encompassing Lys50O consistently show higher Ka values. Among the investigated linear peptides, Nalpha-acetyl-Ala-Phe-Tyr-His-Ser-Ser-Lys5O-Glu-Gln-NH2 (AcAFYHSK5OEQ-NH2) exhibits the strongest interaction with K4 (Ka approximately 1.4 mM(-1)), followed by AcYHSK50EQ-NH2 (Ka approximately 0.9 mM(-1)). Relative to the wild-type sequence, mutated hexapeptides exhibit lesser affinity for K4. When a Lys50 --> Ser mutation was introduced (==> AcYHSS50EQ-NH2), binding was abolished. The Ile27-lle56 construct (L-NTP) contains the Lys50 site within a loop constrained by two cystine bridges. The propensity of recombinant Pgn K1 (rK1) and K2 (rK2) modules, and of Pgn fragments encompassing the intact K4 and K5 domains, for binding L-NTP, was investigated. We find that L-NTP interacts with rK1, rK2, K4, and K5-all lysine-binding kringles-in a fashion that closely mimics what has been observed for the Glul-HSer57 N-terminal fragment of Pgn (CB-NTP). Thus, both the constellation of kringle lysine binding site (LBS) aromatic residues that are perturbed upon complexation of L-NTP and magnitudes of kringle-L-NTP binding affinities (rK1, Ka approximately 4.3 mM(-1); rK2, Ka approximately 3.7 mM(-1; K4, Ka approximately 6.4 mM(1); and K5, Ka approximately 2.1 mM(-1)) are essentially the same as for the corresponding kringle-CB-NTP pairs. Molecular modeling studies suggest that the Glu39-Lys50 stretch in NTP generates an area that complements, both topologically and electrostatically, the solvent-exposed kringle LBS surface.

The Glu1-Val79 N-terminal peptide (NTP) domain of human plasminogen (Pgn) is followed by a tandem array of five kringle (K) structures of approximately 9 kDa each. K1, K2, K4, and K5 contain each a lysine-binding site (LBS). Pgn was cleaved with CNBr and the Glul-HSer57 N-terminal fragment (CB-NTP) isolated. In addition, the Ile27-Ile56 peptide (L-NTP) that spans the doubly S-S bridged loop segment of NTP was synthesized. Pgn kringles were generated either by proteolytic fragmentation of Pgn (K4, K5) or via recombinant gene expression (rK1, rK2, and rK3). Interactions of CB-NTP with each of the Pgn kringles were monitored by 1H-NMR at 500 MHz and values for the equilibrium association constants (Ka) determined: rK1, Ka approximately 4.6 mM(-1); rK2, Ka approximately 3.3 mM(-1); K4, Ka approximately 6.2 mM-'; K5, K, 2.3 mM(-1). Thus, the lysine-binding kringles interact with CB-NTP more strongly than with Nalpha-acetyl-L-lysine methyl ester (Ka < 0.6 mM(-l), which reveals specificity for the NTP. In contrast, CB-NTP does not measurably interact with rK3. which is devoid of a LBS. CB-NTP and L-NTP 1H-NMR spectra were assigned and interproton distances estimated from 1H-1H Overhauser (NOESY) experiments. Structures of L-NTP and the Glul-Ile27 segment of CB-NTP were computed via restrained dynamic simulated annealing/energy minimization (SA/EM) protocols. Conformational models of CB-NTP were generated by joining the two (sub)structures followed by a round of constrained SA/EM. Helical turns are indicated for segments 6-9, 12-16, 28-30, and 45-48. Within the Cys34-Cys42 loop of L-NTP, the structure of the Glu-Glu-Asp-Glu-Glu39 segment appears to be relatively less defined, as is the case for the stretch containing Lys5O within the Cys42-Cys54 segment, consistent with the latter possibly interacting with kringle domains in intact Glul-Pgn. Overall, the CB-NTP and L-NTP fragments are of low regular secondary structure content-as indicated by UV-CD spectra- and exhibit fast amide 1H-2H exchange in 2H2O, suggestive of high flexibility.

Human Glu-plasminogen adopts at least three conformations that provide a means for regulating the specificity of its activation in vivo. It has been proposed previously that the closed (alpha) conformation of human Glu-plasminogen is maintained through physical interaction of the kringle 5 domain and a lysine residue within the N-terminal peptide (NTP). To examine this hypothesis, site-directed mutagenesis was used to generate variant proteins containing substitutions either for aspartic acid residues within the anionic centre of the kringle 5 domain or for conserved lysine residues within the NTP. Size-exclusion HPLC and rates of plasminogen activation by urokinase-type plasminogen activator were used to determine the conformational states of these variants. Variants with substitutions within the kringle 5 lysine-binding site demonstrated extended conformations, as did variants with alanine substitutions for Lys50 and Lys62. In contrast, molecules in which NTP residues Lys20 or Lys33 were replaced were shown to adopt closed conformations. We conclude that the lysine-binding site of kringle 5 is involved in maintaining the closed conformation of human Glu-plasminogen via an interaction with the NTP, probably through Lys50 and/or Lys62. These conclusions advance the current model for the initial stages of fibrinolysis during which fibrin is thought to compete with the NTP for the kringle 5 lysine-binding site.

Insight into the structural features of human lipoprotein(a) [Lp(a)] which underlie its functional implication in fibrinolysis may be gained from comparative studies of apo(a). Indeed, cloning of rhesus monkey apo(a) has shown that a Trp72 --> Arg mutation in the lysine-binding site (LBS) of KIV-10 leads to loss of lysine-binding properties of the rhesus Lp(a) particle. Consequently, comparative studies of apo(a) sequences in different Old World monkey species should further our understanding of the molecular role of Lp(a) in the fibrinolytic process. In contrast to other Old World monkeys, including rhesus monkey, cynomolgus, and baboon, the chimpanzee exhibits an elevated level of Lp(a) and a distinct isoform distribution as compared to humans [Doucet et al. J. Lipid Res. (1994) 35, 263-270]. Clearly then, the chimpanzee is an interesting animal model for study of the structure, function, and potential pathophysiological roles of Lp(a). We have cloned and sequenced the region of chimpanzee apo(a) cDNA spanning KIV-3 to the stop codon. The global organization of this region is similar to that of human apo(a) with the presence of KV, which is absent in rhesus monkey apo(a). Nucleotide sequence comparison indicates a variation of 1.4% between chimpanzee and man and 5.1% between chimpanzee and rhesus monkey. The differences concerned single base changes. An Asp57 --> Asn mutation was detected in KIV-10; this residue is critical to the LBS of KIV-10 in human apo(a). To verify that the Asp57 --> Asn substitution was specific to apo(a), we have also cloned the cDNA-encoding plasminogen, which exhibited an Asp at the corresponding position in kringle IV. Using an in vitro binding assay, we have demonstrated that chimpanzee Lp(a) exhibits poor lysine-specific interaction with both intact and plasmin-degraded fibrin as compared to its human counterpart. We propose that the Asn57 substitution in KIV-10 of chimpanzee apo(a) is responsible for this property. Chimpanzee Lp(a) therefore represents an appropriate particle with which to explore the potential effects of Lp(a) on the fibrinolytic system, such as the inhibition of plasminogen activation or inhibition of t-PA activity.

The interaction of various small aliphatic and aromatic ionic ligands with the human plasminogen (HPg) recombinant kringle 2 (r-K2) domain has been investigated by 1H-NMR spectroscopy at 500 MHz. The results are compared against ligand-binding properties of the homologous, lysine-binding HPg kringle 1 (K1), kringle 4 (K4), and kringle 5 (K5). The investigated ligands include the omega-aminocarboxylic acids 4-aminobutyric acid (4-ABA), 5-aminopentanoic acid (5-APA), 6-aminohexanoic acid (6-AHA), 7-aminoheptanoic acid (7-AHA), lysine and arginine derivatives with free and blocked alpha-amino and/or carboxylate groups, and a number of cyclic analogs, zwitterions of similar size such as trans-(aminomethyl)cyclohexanecarboxylic acid (AMCHA) and p-benzylaminesulfonic acid (BASA), and the nonzwitterions benzylamine and benzamidine. Equilibrium association constant (Ka) values were determined from 1H-NMR ligand titration profiles. Among the aliphatic linear ligands, 5-APA (Ka approximately 3.4 mM-1) shows the strongest interaction with r-K2 followed by 6-AHA (Ka approximately 2.3 mM-1), 7-AHA (Ka approximately 0.45 mM-1), and 4-ABA (Ka approximately 0.22 mM-1). In contrast, r-K1, K4, and K5 exhibit a preference for 6-AHA (Ka approximately 74.2, 21.0, and 10.6 mM-1, respectively), a ligand approximately 1.14 A longer than 5-APA. Mutations R220G and E221D increase the affinity of r-K2 for these ligands but leave the selectivity profile essentially unaffected: 5-APA > 6-AHA > 7-AHA > 4-ABA (Ka approximately 6.5, 3.9, 1.8, and 0.74 mM-1, respectively). We find that, while r-K2 definitely interacts with Nalpha-acetyl-L-lysine and L-lysine (Ka approximately 0.96 and 0.68 mM-1, respectively), the affinity for analogs carrying a blocked carboxylate group is relatively weak (Ka approximately 0.1 mM-1). We also investigated the interaction of r-K2 with L-arginine (Ka approximately 0.31 mM-1) and its derivatives Nalpha-acetyl-L-arginine (Ka approximately 0.55 mM-1), Nalpha-acetyl-L-arginine methyl ester (Ka approximately 0.07 mM-1), and L-arginine methyl ester (Ka approximately 0.03 mM-1). Zwitterionic gamma-guanidinobutyric acid, containing one less methylene group than arginine, exhibits a Ka of approximately 0.28 mM-1. The affinity of r-K2 for lysine and arginine derivatives suggests that K2 could play a role in intermolecular as well as intramolecular interactions of HPg. As is the case for the HPg K1, K4, and K5, among the tested ligands, AMCHA is the one which interacts most firmly with r-K2 (Ka approximately 7.3 mM-1) while the aromatic ligands BASA, benzylamine, and benzamidine exhibit Ka values of approximately 4.0, approximately 0.04, and approximately 0.03 mM-1, respectively. The relative stability of these interactions indicates a strict requirement for both cationic and anionic polar groups in the ligand, whereas the presence of a lipophilic aromatic group seems to be of lesser consequence. Ligand-induced shifts of r-K2 (1)H-NMR signals and two-dimensional nuclear Overhauser effect (NOESY) experiments in the presence of 6-AHA reveal direct involvement of residues Tyr36, Trp62, Phe64, and Trp72 (kringle residue numbering convention) in ligand binding. Starting from the X-ray crystallographic structure of HPg K4 and the intermolecular 1H-NMR NOE data, two models of the K2 lysine binding site complexed to 6-AHA have been derived which differ mainly in the extent of electrostatic pairing between the K2 Arg56 and Glu57 side chains. Competition between these two conformations in equilibrium may account for the relatively lesser affinity of the K2 domain for zwitterionic lysine-type ligands.

Angiostatin is a potent angiogenesis inhibitor which has been identified as an internal fragment of plasminogen that includes its first four kringle modules. We have recently demonstrated that the anti-endothelial cell proliferative activity of angiostatin is also displayed by the first three kringle structures of plasminogen and marginally so by kringle 4 (Cao, Y., Ji, R.-W., Davidson, D., Schaller, J., Marti, D., Sohndel, S., McCance, S. G., O'Reilly, M. S. , Llinás, M., and Folkman, J. (1996) J. Biol. Chem. 271, 29461-29467). We now report that the kringle 5 fragment of human plasminogen is a specific inhibitor for endothelial cell proliferation. Kringle 5 obtained as a proteolytic fragment of human plasminogen displays potent inhibitory effect on bovine capillary endothelial cells with a half-maximal concentration (ED50) of approximately 50 nM. Thus, kringle 5 would appear to be more potent than angiostatin on inhibition of basic fibroblast growth factor-stimulated capillary endothelial cell proliferation. Appropriately folded recombinant mouse kringle 5 protein, expressed in Escherichia coli, exhibits a comparable inhibitory effect as the proteolytic kringle 5 fragment. Thus, kringle 5 domain of human plasminogen is a novel endothelial inhibitor that is sufficiently potent to block the growth factor-stimulated endothelial cell growth.

Elevated plasma levels of lipoprotein(a) [Lp(a)] represent a significant independent risk factor for the development of atherosclerosis. Interindividual levels of apo(a) vary over 1000-fold and are mainly due to inheritance that is linked to the locus of the apolipoprotein(a) [apo(a)] gene. The apo(a) gene encodes multiple repeats of a sequence exhibiting up to 85% DNA sequence homology with plasminogen kringle IV (K.IV), a lysine binding domain. In our search for sequence polymorphisms in the K.IV coding domain, we identified a polymorphism predicting a Thr-->Pro substitution located at amino acid position 12 of kringle IV type 8 of apo(a). The functional and clinical significance of this polymorphism was analysed in a case-control study and by comparing the in vitro lysine binding characteristics of the two Lp(a) subtypes. The case-control study (involving 153 subjects having symptomatic atherosclerosis and 153 age and gender matched normolipidemic controls) revealed a overall allele frequency for the Thr12-->Pro substitution in kringle IV type 8 of 14% and a negative association between presence of the Pro12-subtype and symptomatic atherosclerosis (p < 0.03). The in vitro lysine binding studies, using Lp(a) isolated from subjects homozygous for either Thr12 or Pro12 in K.IV type 8, revealed comparable lysine-Sepharose binding fractions for the two subtypes. The binding affinity (Kd) for immobilised plasmin degraded des-AA-fibrin (Desafib-X) was also comparable for the two subtypes, however a decreased maximal attainable binding (Bmax) for immobilised desafib-X was observed for the Pro12-subtype Lp(a).

The mature form of the zymogen, human plasminogen (HPlg), contains 791 amino acids present in a single polypeptide chain. The fibrinolytic enzyme, human plasmin (HPlm), is formed from HPlg as a result of activator-catalysed cleavage of the Arg561-Val562 peptide bond in HPlg. The resulting HPlm contains a heavy chain of 561 amino acid residues, originating from the N-terminus of HPlg, doubly disulfide-linked to a light chain of 230 amino acid residues. This latter region, containing the C-terminus of HPlg, is homologous to serine proteases such as trypsin and elastase. The heavy chain of HPlm consists of five repeating triple-disulfide-linked peptide regions, c. 80 amino acid residues in length, termed kringles (K), that are responsible for interactions of HPlg and HPlm with substrates, inhibitors and regulators of HPlg activation. Important among the ligands of the kringles are positive activation effectors, typified by lysine and its analogues, and negative activation effectors, such as Cl-. The kringle domains of HPlg that participate in these binding interactions are K1, K4 and K5, and perhaps K2. These modules appear to function as independent domains. The amino acid residues important in these kringle/ligand binding interactions have been proposed by structural determinations, and their relative importance quantified by site-directed mutagenesis experimentation.

Recently we have identified angiostatin, an endogenous angiogenesis inhibitor of 38 kDa which specifically blocks the growth of endothelial cells (O'Reilly, M. S., Holmgren, L., Shing, Y., Chen, C. , Rosenthal, R. A., Moses, M., Lane, W. S., Cao, Y., Sage, E. H., and Folkman, J. (1994) Cell 79, 315-328; Folkman, J. (1995) Nat. Med. 1, 27-31). Angiostatin was shown to represent an internal fragment of plasminogen containing the first four kringle structures. We now report on the inhibitory effects of individual or combined kringle structures of angiostatin on capillary endothelial cell proliferation. Recombinant kringle 1 and kringle 3 exhibit potent inhibitory activity with half-maximal concentrations (ED50) of 320 nM and 460 nM, respectively. Also, recombinant kringle 2 displays a significant inhibition, although decreased compared with both kringle 1 and kringle 3. In contrast, kringle 4 is an ineffective inhibitor of basic fibroblast growth factor-stimulated endothelial cell proliferation. Among the tandem kringle arrays, the recombinant kringle 2-3 fragment exerts inhibitory activity similar to kringle 2 alone. However, relative to kringle 2-3, a marked enhancement in inhibition is observed when individual kringle 2 and kringle 3 are added together to endothelial cells. This implies that it is necessary to open the cystine bridge between kringle 2 and kringle 3 to obtain the maximal inhibitory effect of kringle 2-3. An increased (<2-fold) inhibitory activity is observed for the kringle 1-3 fragment (ED50 = 70 nM) compared with kringle 1-4 (ED50 = 135 nM). These data indicate that the anti-proliferative activity of angiostatin on endothelial cells is shared by kringle 1, kringle 2, and kringle 3, but probably not by kringle 4 and that more potent inhibition results when kringle 4 is removed from angiostatin. Thus, in view of the variable lysine affinity of the homologous domains, it would appear that lysine binding capability does not correlate with the relative inhibitory effects of the kringle-containing constructs. However, as we also demonstrate, appropriate folding of kringle structures is essential for angiostatin to maintain its full anti-endothelial activity.

Kringles are protein modules found within a wide variety of fibrinolytic and coagulation-related proteins that show binding affinity for lysine, lysine analogs and for fibrin. We report here the crystal structures of apolipoprotein(a) kringle IV37 (apo(a) K4(37)) in its free state and in separate complexes with two omega-amino acids, 6-aminohexanoic acid (6AHA) and p-aminomethylbenzoic acid (PAMBA). The structures of the unliganded form and of both complexes have been determined and refined by restrained least-squares methods to about 2.0 angstrom. The overall kringle architecture is essentially identical with that determined in other kringles but it shows some small significant structural changes in the lysine binding site. Ther is virtually no difference in conformation between the unliganded and complexed forms, suggesting that apo(a) K4(37) does not undergo any conformational rearrangement upon binding. The 6AHA molecule binds to apo(a) K4(37) in a completely different way from that observed with the kringle 4 of plasminogen (PGK4). Its amino group makes an ion pair interaction with the two aspartate residues (Asp55/Asp57) of the anionic center and its carboxylate group faces out into the solvent making water-mediated contacts with the protein. The mode of binding of PAMBA resembles more that decribed for 6AHA when bound to PGK4. The PAMBA molecule is bound by ion pair interactions with the two aspartate residues (Asp55/Asp57) and with Arg71 from the cationic center and by van der Waals contacts. The relative importance of the cationic center from kringles for binding zwitterionic ligands is discussed.

The plasminogen kringle 2 (K2HPg) and kringle 3 (K3HPg) modules occur in tandem within the polypeptide segment that affords the heavy chain of plasmin. The K2HPg and K3HPg are unique among the plasminogen kringle domains in that they also are linked to each other via the Cys169-Cys297 (Cys4 of K2HPg to Cys43 of K3HPg, kringle numbering convention) disulfide bridge, thus generating a K2HPg-K3HPg "supermodule". The kringle (2 + 3) sequence of human plasminogen (r-EE[K2HPgK3HPg]DS) was expressed in Escherichia coli, using an expression vector containing the phage T5 promoter/operator N250PSN250P29 and the codons for an N-terminal hexahistidine tag to ensure the isolation of the recombinant protein by affinity chromatography on Ni(2+)-nitrilotriacetic acid/agarose under denaturing and reducing conditions. Kringle (2 + 3) was refolded in the presence of glutathione redox buffer. By taking advantage of the lysine affinity of kringle 2, the protein was purified by affinity chromatography on lysine-Bio-Gel. Recombinant kringle (2 + 3) was identified by amino acid composition, N-terminal sequence and mass determination. The 1H NMR spectrum shows that the intact r-K2HPgK3HPg is properly folded. By reference to spectra of the individual kringles, r-K2HPg and r-K3HPg, resonances of the K2HPg and K3HPg components in the spectrum of the intact r-K2HPgK3HPg can be readily distinguished. The strictly conserved Leu46 residue (kringle residue number convention) yields delta-methyl signals that are characteristic for K2HPg and K3HPg, exhibiting chemical shifts of -0.87 and -0.94 ppm, respectively, which are distinct from those of K1HPg, K4HPg, and K5HPg, (-1.04 to -1.05 ppm). Thus, the high-field Leu46 signals from K2HPg and K3HPg are well resolved from those of other kringles and can be identified unambiguously in spectra of the K1HPgK2HPgK3HPg elastolytic fragment of plasminogen as well as in spectra of Glu-plasminogen. Overall, r-K2HPgK3HPg exhibits broader resonance line widths than does the K1HPg component, consistent with a lesser mobility of the K2HPgK3HPg segment within the K1HPgK2HPgK3HPg fragment, a reflection of the extra structural constraint imposed by the disulfide bridge linking K2HPg to K3HPg. The ligand 6-aminohexanoic acid (6-AHA), which is known to interact with r-K2HPg but not with r-K3HPg, selectively perturbs K2 aromatic signals in the intact r-K2HPgK3HPg spectrum while leaving K3 resonances largely unaffected. Association constant (K(a)) values for 6-AHA determined from 1H NMR ligand titration experiments yield K(a) approximately 2.2 +/- 0.3 mM(-1) for the intact r-K2HPgK3HPg, comparable to K(a) approximately 2.3 +/- 0.2 mM(-1) determined for the isolated r-K2HPg, which demonstrates that the interactions of 6-AHA with the K2HPg ligand-binding site are not significantly affected by the neighboring K3HPg domain within the intact r-K2HPgK3HPg supermodule.

The roles of each of the three omega-amino acid-binding kringles (K) of Glu1-Pg, viz., [K1Pg], [K4Pg], and [K5Pg], in engendering the Cl(-)-induced alteration to its tight (T) conformation and in effecting the epsilon-aminocaproic acid (EACA)-mediated change to the relaxed (R) protein conformation have been investigated by mutagenesis strategies wherein the omega-amino acid ligand-binding energies in the individual kringles in recombinant (r)-Glu1-Pg were greatly reduced. This was accomplished in the most conservative manner possible by altering a critical Asp residue in each relevant kringle to Asn. The particular mutations chosen were r-[D139N]Glu1-Pg, r-[D413N]Glu1-Pg, and r-[D518N]Glu1-Pg, in which a conserved Asp residue at a homologous sequence position in each of the three kringle domains is eliminated. These changes also lead to a great reduction of the EACA-binding strength of [K1Pg], [K4Pg], and [K5Pg], respectively. The s0(20,w) of wild-type (wt) r-Glu1-Pg in the presence of levels of Cl(-)-sufficient to fully occupy its binding sites on this protein was 5.9 S, a value reduced to 4.9 S as a result of addition of saturating concentrations of EACA to the Cl-/Glu1-Pg complex. Neither Cl- nor EACA substantially altered the s0(20,w) value of 5.2 S for r-[D139N]Glu1-Pg (4.8 S) or r-[D413N]Glu1-Pg (4.5 S). On the other hand, the s0(20,w) value of 5.2 S for r-[D518N]Glu1-Pg at saturating levels of Cl- is slightly reduced to 4.8 S upon addition of binding maximal concentrations of EACA.(ABSTRACT TRUNCATED AT 250 WORDS)

The interaction between the plasminogen kringle 4 module and a synthetic peptide corresponding to the tryptic heptapeptide fragment Ala-Phe-Gln-Tyr-His-Ser-Lys (AFQYHSK), segment 44-50 of the plasminogen N-terminal peptide (Wiman and Wallén, Eur J Biochem 1975; 50:489-494), has been investigated by 1H-NMR spectroscopy. AFQYHSK, as well as the shorter fragments thereof, FQYHSK, QYHSK and YHSK, all bound to kringle 4 with equilibrium association constant (Ka) values ranging between 2.5 and 8.5 mM-1. The NMR evidence also indicates that binding is mediated by the canonical kringle lysine binding site and involves the C-terminal Lys residue of the ligand peptide. The results (a) support a potential interaction between plasminogen Lys-binding kringles and the N-terminal activation peptide, and (b) unambiguously demonstrate the capability of such kringles to bind polypeptides ending with C-terminal lysine.

With the aim of elucidating the evolutionary processes of the kringle and protease domains in serine proteases which are involved with the system of blood coagulation and fibrinolysis, we constructed phylogenetic trees for the kringle and protease domains, separately, by use of amino acid sequence data. The phylogenetic trees constructed clearly showed that the topologies were different between the kringle and protease domains. Because both domains are coded by single peptides of serine proteases, this strongly suggests that the kringle and protease domains must have undergone different evolutionary processes. Thus, these observations imply that serine proteases evolve in a way such that each domain is a unit of evolution, exemplifying a typical mode of domain evolution. A possible relationship between the domain evolution and the exon shuffling theory is also discussed from the viewpoint of gene evolution.

In order to identify the individual contributions of the kringle (K) domains of human plasminogen (Pg) to the epsilon-aminocaproic acid (EACA) induced stimulation of Pg activation by low-molecular-weight urokinase-type plasminogen activator (LMW-uPA) and inhibition of this same activation by Cl-, we constructed the most conservative recombinant- (r-) Pg mutants possible that would greatly reduce the strength of the EACA binding site in the omega-amino acid binding kringles, [K1Pg] ([D139-->N]r-Pg), [K4Pg] ([D413-->N]r-Pg), and [K5Pg] ([D515--N]r-Pg). In each case, this involved mutation of a critical Asp (to Asn) within these three kringle domains in intact Pg. The three r-mutants were expressed in r-baculovirus-infected lepidopteran insect (Trichoplusia ni) cells. In the presence of Cl-, the positive activation effector, EACA, first stimulated and then inhibited the LMW-uPA-catalyzed initial activation of wild-type (wt) r-[Glu1]Pg and, to a lesser extent, the [K5Pg] mutant, [D518-->N/Glu1]r-Pg. The concentration of EACA that produced 50% stimulation of activation (C50) occurred at 3.3 mM for wtr-[Glu1]Pg and at 0.7 mM for [D518-->N/Glu1]r-Pg. Subsequent inhibition by EACA occurred with a C50 of approximately 15 mM and is likely due to inhibition of the amidolytic activity of plasmin generated during the activation. Similar initial activation rates of both [D139-->N]r-Pg and [D413N]r-Pg did not display this initial EACA-mediated stimulatory phase but did undergo ultimate inhibition with a C50 for this process that was similar to wtr-[Glu1]Pg and [D518-->N/Glu1]r-Pg.(ABSTRACT TRUNCATED AT 250 WORDS)

Regions of the human plasminogen (Pg) cDNA containing its kringle 4 (K4) and K5 domains have been expressed in Escherichia coli, and binding constants of omega-amino acid ligands for recombinant (r)-[K4Pg] and r-[K5Pg] have been obtained. In each case, the results showed that of a series of aliphatic alpha, omega-amino acid analogues, 6-aminohexanoic acid showed maximal affinity for these modules, and all ligands interacted more strongly with r-[K4Pg] than with r-[K5Pg]. Site-directed mutagenesis investigations demonstrated that the major amino acid side chain contributors to ligand binding were similar for each of these kringles. Ligand binding was stabilized by charged groups at Asp56 of r-[K4Pg] and Asp57 of r-[K5Pg] as well as by Arg69 of both r-[K4Pg] and r-[K5Pg]. Some hydrophobic amino acids that contributed significantly to the binding strength of the ligands were identified. These were provided by homologous residues in each of the domains, viz. Trp60 and Trp70 of r-[K4Pg] and Trp62 and Tyr72 of r-[K5Pg]. Tyr74 of r-[K5Pg] also substantially contributed to its ligand binding energy.

Plasminogen-related growth factors, a new family of polypeptide growth factors with the basic domain organization and mechanism of activation of the blood proteinase plasminogen, include hepatocyte growth factor/scatter factor (HGF/SF), a potent effector of the growth, movement, and differentiation of epithelia and endothelia, and hepatocyte growth factor-like/macrophage stimulating protein (HGF1/MSP), an effector of macrophage chemotaxis and phagocytosis. Phylogeny of the serine proteinase domains and analysis of intron-exon boundaries and kringle sequences indicate that HGF/SF, HGF1/MSP, plasminogen, and apolipoprotein (a) have evolved from a common ancestral gene that consisted of an N-terminal domain corresponding to plasminogen activation peptide (PAP), 3 copies of the kringle domain, and a serine proteinase domain. Models of the N domains of HGF/SF, HGF1/MSP, and plasminogen, characterized by the presence of 4 conserved Cys residues forming a loop in a loop, have been modeled based on disulfide-bond constraints. There is a distinct pattern of charged and hydrophobic residues in the helix-strand-helix motif proposed for the PAP domain of HGF/SF; these may be important for receptor interaction. Three-dimensional structures of the 4 kringle and the serine proteinase domains of HGF/SF were constructed by comparative modeling using the suite of programs COMPOSER and were energy minimized. Docking of a lysine analogue indicates a putative lysine-binding pocket within kringle 2 (and possibly another in kringle 4). The models suggest a mechanism for the formation of a noncovalent HGF/SF homodimer that may be responsible for the activation of the Met receptor. These data provide evidence for the divergent evolution and structural similarity of plasminogen, HGF/SF, and HGF1/MSP, and highlight a new strategy for growth factor evolution, namely the adaptation of a proteolytic enzyme to a role in receptor activation.

The crystal structure of a monoclinic form of human plasminogen kringle 4 (PGK4) has been solved by molecular replacement using the orthorthombic structure as a model and it has been refined by restrained least-squares methods to an R factor of 16.4% at 2.25 A resolution. The X-PLOR structure of kringle 2 of tissue plasminogen activator (t-PAK2) has been refined further using PROFFT (R = 14.5% at 2.38 A resolution). The PGK4 structure has 2 and t-PAK2 has 3 independent molecules in the asymmetric unit. There are 5 different noncrystallographic symmetry "dimers" in PGK4. Three make extensive kringle-kringle interactions related by noncrystallographic 2(1) screw axes without blocking the lysine binding site. Such associations may occur in multikringle structures such as prothrombin, hepatocyte growth factor, plasminogen (PG), and apolipoprotein [a]. The t-PAK2 structure also has noncrystallographic screw symmetry (3(1)) and mimics fibrin binding mode by having lysine of one molecule interacting electrostatically with the lysine binding site of another kringle. This ligand-like binding interaction may be important in kringle-kringle interactions involving non-lysine binding kringles with lysine or pseudo-lysine binding sites. Electrostatic intermolecular interactions involving the lysine binding site are also found in the crystal structures of PGK1 and orthorhombic PGK4. Anions associate with the cationic centers of these and t-PAK2 that appear to be more than occasional components of lysine binding site regions.

The solution structure of the human plasminogen kringle 1 domain complexed to the antifibrinolytic drug 6-aminohexanoic acid (epsilon Ahx) was obtained on the basis of 1H-NMR spectroscopic data and dynamical simulated annealing calculations. Two sets of structures were derived starting from (a) random coil conformations and (b) the (mutated) crystallographic structure of the homologous prothrombin kringle 1. The two sets display essentially the same backbone folding (pairwise root-mean-square deviation, 0.15 nm) indicating that, regardless of the initial structure, the data is sufficient to locate a conformation corresponding to an essentially unique energy minimum. The conformations of residues connected to prolines were localized to energetically preferred regions of the Ramachandran map. The Pro30 peptide bond is proposed to be cis. The ligand-binding site of the kringle 1 is a shallow cavity composed of Pro33, Phe36, Trp62, Tyr64, Tyr72 and Tyr74. Doubly charged anionic and cationic centers configured by the side chains of Asp55 and Asp57, and Arg34 and Arg71, respectively, contribute to anchoring the zwitterionic epsilon Ahx molecule at the binding site. The ligand exhibits closer contacts with the kringle anionic centers (approximately 0.35 nm average O...H distance between the Asp55/Asp57 carboxylate and ligand amino groups) than with the cationic ones (approximately 0.52 nm closest O...H distances between the ligand carboxylate and the Arg34/Arg71 guanidino groups). The epsilon Ahx hydrocarbon chain rests flanked by Pro33, Tyr64, Tyr72 and Tyr74 on one side and Phe36 on the other. Dipolar (Overhauser) connectivities indicate that the ligand aliphatic moiety establishes close contacts with the Phe36 and Trp62 aromatic rings. The computed structure suggests that the epsilon Ahx molecule adopts a kinked conformation when complexed to kringle 1, effectively shortening its dipole length to approximately 0.65 nm.