Tissue-type plasminogen activator and its substrate Glu-plasminogen share common binding sites in limited plasmin-digested fibrin

The enzyme tissue-type plasminogen activator (t-PA) and its substrate Glu-plasminogen can both bind to fibrin. The assembly of these three components results in about a 1000-fold acceleration of the conversion of Glu-plasminogen into plasmin. Fibrin binding of t-PA is mediated both by its finger (F) domain and its kringle-2 domain. Fibrin binding of Glu-plasminogen involves its kringle structures (K1-K5). It has been suggested that particular kringles contain lysine-binding sites and/or aminohexyl-binding sites, exhibiting affinity for specific carboxyl-terminal lysines and intrachain lysines, respectively. We investigated the possibility that t-PA and Glu-plasminogen kringles share common binding sites in fibrin, limitedly digested with plasmin. For that purpose we performed competition experiments, using conditions that exclude plasmin formation, with Glu-plasminogen and either t-PA or two deletion mutants, lacking the F domain (t-PA del.F) or lacking the K2 domain (t-PA del.K2). Our data show that fibrin binding of t-PA, mediated by the F domain, is independent of Glu-plasminogen binding. In contrast, partial inhibition by Glu-plasminogen of t-PA K2 domain-mediated fibrin binding is observed that is dependent on carboxyl-terminal lysines, exposed in fibrin upon limited plasmin digestion. Half-maximal competition of fibrin binding of both t-PA and t-PA del.F is obtained at 3.3 microM Glu-plasminogen. The difference between this value and the apparent dissociation constant of Glu-plasminogen binding to limitedly digested fibrin (12.1 microM) under these conditions is attributed to multiple, simultaneous interactions, each having a separate affinity. It is concluded that t-PA and Glu-plasminogen can bind to the same carboxyl-terminal lysines in limitedly digested fibrin, whereas binding sites composed of intrachain lysines are unique both for the K2 domain of t-PA and the Glu-plasminogen kringles.     

The binding of plasminogen fragments to cultured human umbilical vein endothelial cells.

Glu-plasminogen, kringle 1-5, kringle 1-3, and miniplasminogen exhibited strong binding to human umbilical vein endothelial cells (HUVEC). On the other hand, no significant binding was obtained with microplasminogen and kringle 4. Kringle 1-5 and miniplasminogen, which both contained kringle 5, specifically inhibited the binding of plasminogen to HUVEC while kringle 1-3 did not. The results implied plasminogen molecule contained at least two binding sites, with which it interacted HUVEC. The stronger binding site was located in kringle 5 and the weaker one was in kringle 1-3. Kringle 4 and the active site domain exhibited no significant binding to HUVEC. The interaction of plasminogen with HUVEC is mainly through binding site on kringle 5.     

Isolation, purification and 1H-NMR characterization of a kringle 5 domain fragment from human plasminogen

A scheme is proposed for generating the intact Val-448-Phe-545 polypeptide of human plasminogen which contains the fifth kringle domain of the plasmin heavy chain. The procedure is based on a pepsin fragmentation of miniplasminogen and involves the purification of the kringle 5-containing fragment by gel filtration and ion-exchange chromatography. The final product is characterized by amino acid analysis, N- and C-terminal analyses, and high-resolution 1H-NMR spectroscopy at both 300 MHz and 611 MHz. We detect a (40:60%) Asp/Asn heterogeneity at site 452 of the Glu-plasminogen molecule. In the conventional kringle numbering system, the kringle 5 domain extends from Cys-1 to Cys-80, which corresponds to Cys-461 to Cys-540 in plasminogen. A preliminary 1H-NMR characterization of kringle 5 focuses on the global conformational features of the polypeptide. Assignments are given for a number of resonances, including the Tyr-72, the His imidazoles' and the Trp indoles' spin systems. Comparison with human plasminogen kringles 1 and 4 shows that the kringle 5 conformation is highly structured and very similar to that of the homologous domains. This conservancy is particularly striking in the environment surrounding Leu-46 and in the overall features of the aromatic spectrum. There are some differences, particularly in the buried His-33 imidazole group, whose H2 resonance is shifted to 9.67 ppm. A preliminary study of benzamidine-binding shows that the ligand interacts weakly (Ka approximately equal to 1.7 mM -1) mainly through the amidino functional group. Trp-62 and Tyr-72 are significantly perturbed by benzamidine, suggesting that these residues are part of the ligand-binding site.     

Characterization of the high-affinity interaction between human plasminogen and pro-urokinase

Activation of human Glu-plasminogen, Lys-plasminogen and low-Mr plasminogen (lacking lysine-binding sites) by pro-urokinase (pro-UK), obtained from a human lung adenocarcinoma cell line (Calu-3, ATCC), obeys Michaelis-Menten kinetics. Activation occurs with a comparable affinity (Km 0.40-0.77 microM), while the catalytic rate constant (kcat) is comparable for Glu-plasminogen (0.0022s-1) and low-Mr plasminogen (0.0034 s-1), but is somewhat higher for Lys-plasminogen (0.0106 s-1). The rate of activation of plasminogen by pro-UK is not significantly influenced by the presence of 6-aminohexanoic acid, purified fragments LBS I or LBS II or histidine-rich glycoprotein, indicating that the high affinity of pro-UK for plasminogen is not mediated via the high-affinity lysine-binding site of plasminogen located in kringles 1-3 (LBS I) nor via the low-affinity lysine-binding site comprised within kringle 4 (LBS II). The site(s) in plasminogen involved in the high-affinity interaction with pro-UK thus appear to be located within the low-Mr plasminogen moiety.     

The X-ray crystal structure of full-length human plasminogen

Plasminogen is the proenzyme precursor of the primary fibrinolytic protease plasmin. Circulating plasminogen, which comprises a Pan-apple (PAp) domain, five kringle domains (KR1-5), and a serine protease (SP) domain, adopts a closed, activation-resistant conformation. The kringle domains mediate interactions with fibrin clots and cell-surface receptors. These interactions trigger plasminogen to adopt an open form that can be cleaved and converted to plasmin by tissue-type and urokinase-type plasminogen activators. Here, the structure of closed plasminogen reveals that the PAp and SP domains, together with chloride ions, maintain the closed conformation through interactions with the kringle array. Differences in glycosylation alter the position of KR3, although in all structures the loop cleaved by plasminogen activators is inaccessible. The ligand-binding site of KR1 is exposed and likely governs proenzyme recruitment to targets. Furthermore, analysis of our structure suggests that KR5 peeling away from the PAp domain may initiate plasminogen conformational change.     

Deletion of kringle domains or the N-terminal hairpin structure in hepatocyte growth factor results in marked decreases in related biological activities.

To determine the essential domain for biological activity in the hepatocyte growth factor (HGF) molecule, we prepared various mutated recombinant HGFs using site-directed mutagenesis, and examined the effects on DNA synthesis in hepatocytes, scattering of MDCK cells and the antiproliferative activity on HepG2 hepatoma cells. Native HGF and mutant HGFs, in which Gln534 and/or Tyr673 were respectively substituted for His and Ser to coincide with the catalytic triad amino acids in plasmin, markedly stimulated DNA synthesis of hepatocytes and scattering of MDCK cells but inhibited DNA synthesis of HepG2 cells. The mutant HGF deleted with the third or fourth kringle domain resulted in marked decrease of all three biological activities, while deletion of the N-terminal hairpin structure or the first or second kringle domain almost completely inactivated biological activities. We propose that the N-terminal hairpin structure and the first and second kringle domains are essential for biological activities of HGF and possibly for binding to its receptor.     

Features of the interaction of Glu- and Lys-forms of plasminogen with native and partially hydrolyzed fibrin]

Glu- and Lys-plasminogen interaction with native and desAABB-fibrin obtained from fibrinogen partially hydrolyzed by plasmin was studied. It was found that native fibrin adsorbs 6 times more Lys-plasminogen as compared to the native form of the proenzyme. The range of the Lys-plasminogen binding does not change, if part of the fibrinogen molecules hydrolyze down to X-fragments. At the same time, the appearance in the system of 1% Xi-fragments leads to a 6-fold increase in the Glu-plasminogen binding. The amount of adsorbed Glu-plasminogen reaches the level of Lys-plasminogen adsorption both in the native and partially hydrolyzed fibrin. It was found that kringle K 1-3 or 6-aminohexanoic acid at saturating for high-affinity lysine-binding sites concentrations do not influence the Glu-plasminogen binding to native fibrin but inhibit it when the partially purified form is used. It is assumed that the manyfold increase of the Glu-plasminogen binding to partially hydrolyzed fibrin is due to the alteration of the proenzyme conformation at the initial steps of fibrin hydrolysis during the formation of Xi fragments.     

Inhibition of the process of Glu-plasminogen activation by tissue activator with fibrin, DDE-complex, and D-dimer using alpha-2-antiplasmin

The alpha-2-antiplasmin influence on the Glu-plasminogen activation by tissue activator both on fibrin and fibrin(ogen) fragments was investigated. The kinetics of activation was studied and velocity of this process in the absence and presence of the inhibitor was calculated. It was established that alpha-2-antiplasmin decreased the velocity of Glu-plasminogen activation on desAABBfibrin, DDE-complex and DD-dimer and did no influence upon proenzyme activation on fibrinogen fragment--Ho1-DSK. In the presence of fibrin plasminogen activation linear related to the amount added tissue activator in limit concentration from 5 before 50 units/ml. It was shown that alpha-2-antiplasmin reduced the activation velocity with used concentration of tissue activator. Fibrin hydrolysis by plasmin, forming on its surface during the plasminogen activation by tissue activator, was also inhibited with alpha-2-antiplasmin. The obtained results are explained by the influence of the inhibitor on formation of the triple complex between plasminogen, tissue activator and fibrin, and competition of the alpha-2-antiplasmin for lysine-binding sites of tissue activator kringle 2 or for binding sites of the activator on fibrin.     

Variants of human tissue-type plasminogen activator that lack specific structural domains of the heavy chain.

The heavy chain of tissue plasminogen activator (t-PA) consists of four domains [finger, epidermal-growth-factor (EGF)-like, kringle 1 and kringle 2] that are homologous to similar domains present in other proteins. To assess the contribution of each of the domains to the biological properties of the enzyme, site-directed mutagenesis was used to generate a set of mutants lacking sequences corresponding to the axons encoding the individual structural domains. The mutant proteins were assayed for their ability to hydrolyze artificial and natural substrates in the presence and absence of fibrin, to bind to lysine-Sepharose and to be inhibited by plasminogen activator inhibitor-1. All the deletion mutants exhibit levels of basal enzymatic activity very similar to that of wild-type t-PA assayed in the absence of fibrin. A mutant protein lacking the finger domain has a 2-fold higher affinity for plasminogen than wild-type t-PA, while the mutant that lacks both finger and EGF-like domains is less active at low concentrations of plasminogen. Mutants lacking both kringles neither bind to lysine-Sepharose nor are stimulated by fibrin. However, mutants containing only one kringle (either kringle 1 or kringle 2) behave indistinguishably from one another and from the wild-type protein. We conclude that kringle 1 and kringle 2 are equivalent in their ability to mediate stimulation of catalytic activity by fibrin.     

Studies on the lysine-binding sites of human plasminogen. The effect of ligand structure on the binding of lysine analogs to plasminogen

A method is described for measuring relative binding constants of lysine and analogs of lysine to plasminogen and plasminogen 'kringle' fragments. Plasminogen or kringle fragments adsorbed to lysine-Sepharose are eluted with increasing concentrations of lysine or other ligands, the concentration of ligand required to elute 50% of the protein being taken as a measure of the binding constant. The method is simple and is not dependent on monitoring conformational changes. We confirm earlier reports that the best ligands for the lysine binding sites of plasminogen are omega-amino acids containing five or six carbons. We show further that both Glu-plasminogen (the native form with N-terminal glutamic acid) and Lys-plasminogen (a degraded form with N-terminal lysine), as well as the heavy chain fragments, kringle 4 and kringle 1+2+3, have very similar properties with regard to binding specificity for omega-amino acids. For all species optimal binding is observed when the distance between the amino and carboxyl carbon is about 0.68 nm. The finding of ligands is decreased by the presence of polar atoms on the alpha and beta positions of the carbon chain of amino acids. Arginine binds relatively weakly at the lysine site and there does not appear to be a separate arginine binding site in plasminogen.     

Fragment E-2 from fibrin substantially enhances pro-urokinase-induced Glu-plasminogen activation. A kinetic study using the plasmin-resistant mutant pro-urokinase Ala-158-rpro-UK.

In a previous study, it was shown that fibrin fragment E-2 selectively promotes the activation of plasminogen by pro-urokinase (pro-UK) [Liu, J., & Gurewich, V. (1991) J. Clin. Invest. 88, 2012-2017]. In this study, the kinetics of this promotion by fragment E-2 was studied. Alanine-158-rpro-UK (A-pro-UK), a recombinant plasmin-resistant mutant, was used in order to avoid interference by UK generation during the reaction. In some experiments, pro-UK was substituted in order to validate the mutant as a surrogate. In the presence of a range of concentrations (0-20 microM) of fragment E-2, a linear promotion of the catalytic efficiency of A-pro-UK against native Glu-plasminogen was seen which was 245.5-fold at the highest concentration of fragment E-2 and 450-fold at the highest ratio of E-2/plasminogen used. The promotion was largely a function of an increase in kcat, since fragment E-2 induced a less than 10-fold reduction in KM (8.50-1.40 microM). In contrast to this ligand, epsilon-aminocaproic acid (EACA) induced a biphasic promotion of the activation of Glu-plasminogen which was only 18-fold at maximum. Fragment E-2 did not promote the activation of Lys-plasminogen, but the catalytic efficiency of A-pro-UK was 19.7-fold greater against the open Lys-form than against the closed Glu- form of plasminogen. Fragment E-2 had no effect on the amidolytic activity of A-pro-UK or pro-UK, suggesting that the promotion of their activities was indirect and related to a fragment E-2-induced conformational change in Glu-plasminogen.(ABSTRACT TRUNCATED AT 250 WORDS)     

Cross-linking of two beta subunits in the closed conformation in F1-ATPase

In the crystal structure of mitochondrial F1-ATPase, two beta subunits with a bound Mg-nucleotide are in "closed" conformations, whereas the third beta subunit without bound nucleotide is in an "open" conformation. In this "CCO" (beta-closed beta-closed beta-open) conformational state, Ile-390s of the two closed beta subunits, even though they are separated by an intervening alpha subunit, have a direct contact. We replaced the equivalent Ile of the alpha3beta3gamma subcomplex of thermophilic F1-ATPase with Cys and observed the formation of the beta-beta cross-link through a disulfide bond. The analysis of conditions required for the cross-link formation indicates that: (i) F1-ATPase takes the CCO conformation when two catalytic sites are filled with Mg-nucleotide, (ii) intermediate(s) with the CCO conformation are generated during catalytic cycle, (iii) the Mg-ADP inhibited form is in the CCO conformation, and (iv) F1-ATPase dwells in conformational state(s) other than CCO when only one (or none) of catalytic sites is filled by Mg-nucleotide or when catalytic sites are filled by Mg2+-free nucleotide. The alpha3beta3gamma subcomplex containing the beta-beta cross-link retained the activity of uni-site catalysis but lost that of multiple catalytic turnover, suggesting that open-closed transition of beta subunits is required for the rotation of gamma subunit but not for hydrolysis of a single ATP.     

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

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

Apolipoprotein(a): A Puzzling Evolutionary Story

Human apolipoprotein(a), a risk factor for heart disease, has over 80% sequence identity to plasminogen. Plasminogen contains five distinct kringle domains plus a catalytic protease subunit. Human apo(a) consists of multiple copies (the number varies in individuals) of a domain resembling kringle 4, a single copy of a domain resembling kringle 5, and a protease-like domain. The recently cloned hedgehog version of apolipoprotein(a), which contains 31 nearly identical copies of plasminogen kringle 3 and lacks a protease domain, has prompted us to investigate the evolutionary history of the apolipoprotein (a) gene in mammals. Our analysis supports the nonfunctionality of the human apolipoprotein(a) protease domain, and a single (or multiple) duplication of plasminogen gene before mammal radiation, which originated apolipoprotein(a) in mammals. Received: 26 February 1996 / Accepted: 6 August 1996     

Topography of the high-affinity lysine binding site of plasminogen as defined with a specific antibody probe

An antibody population that reacted with the high-affinity lysine binding site of human plasminogen was elicited by immunizing rabbits with an elastase degradation product containing kringles 1-3 (EDP I). This antibody was immunopurified by affinity chromatography on plasminogen-Sepharose and elution with 0.2 M 6-aminohexanoic acid. The eluted antibodies bound [125I]EDP I, [125I]Glu-plasminogen, and [125I]Lys-plasminogen in radioimmunoassays, and binding of each ligand was at least 99% inhibited by 0.2 M 6-aminohexanoic acid. The concentrations for 50% inhibition of [125I]EDP I binding by tranexamic acid, 6-aminohexanoic acid, and lysine were 2.6, 46, and 1730 microM, respectively. Similar values were obtained with plasminogen and suggested that an unoccupied high-affinity lysine binding site was required for antibody recognition. The antiserum reacted exclusively with plasminogen derivatives containing the EDP I region (EDP I, Glu-plasminogen, Lys-plasminogen, and the plasmin heavy chain) and did not react with those lacking an EDP I region [miniplasminogen, the plasmin light chain or EDP II (kringle 4)] or with tissue plasminogen activator or prothrombin, which also contain kringles. By immunoblotting analyses, a chymotryptic degradation product of Mr 20,000 was derived from EDP I that retained reactivity with the antibody. The high-affinity lysine binding site was equally available to the antibody probe in Glu- and Lys-plasminogen and also appeared to be unoccupied in the plasmin-alpha 2-antiplasmin complex. alpha 2-Antiplasmin inhibited the binding of radiolabeled EDP I, Glu-plasminogen, or Lys-plasminogen by the antiserum, suggesting that the recognized site is involved in the noncovalent interaction of the inhibitor with plasminogen.(ABSTRACT TRUNCATED AT 250 WORDS)     

Identification of sequences in apolipoprotein(a) that maintain its closed conformation: a novel role for apo(a) isoform size in determining the efficiency of covalent Lp(a) formation

We have previously demonstrated that, in the presence of the lysine analogue epsilon-aminocaproic acid, apolipoprotein(a) [apo(a)] undergoes a conformational change from a closed to an open structure that is characterized by a change in tryptophan fluorescence, an increase in the radius of gyration, an alteration of domain stability, and an enhancement in the efficiency of covalent lipoprotein(a) [Lp(a)] formation. In the present study, to identify sequences within apo(a) that maintain its closed conformation, we used epsilon-aminocaproic acid to probe the conformational status of a variety of recombinant apo(a) isoforms using analytical ultracentrifugation, differential scanning calorimetry, intrinsic fluorescence, and in vitro covalent Lp(a) formation assays. We observed that the closed conformation of apo(a) is maintained by intramolecular interaction(s) between sequences within the amino- and carboxyl-terminal halves of the molecule. Using site-directed mutagenesis, we have identified the strong lysine-binding site present within apo(a) kringle IV type 10 as an important site within the C-terminal half of the molecule, which is involved in maintaining the closed conformation of apo(a). Apo(a) exhibits marked isoform size heterogeneity because of the presence of varying numbers of copies of the kringle IV type-2 domain located within the amino-terminal half of the molecule. Using recombinant apo(a) species containing either 1, 3, or 8 copies of kringle IV type 2, we observed that, while apo(a) isoform size does not alter the affinity of apo(a) for low-density lipoprotein, it affects the conformational status of the protein and therefore influences the efficiency of covalent Lp(a) assembly. The inverse relationship between apo(a) isoform size and the efficiency of covalent Lp(a) formation that we report in vitro may contribute to the inverse relationship between apo(a) isoform size and plasma Lp(a) concentrations that has been observed in vivo.     

Apolipoprotein(a) and plasminogen interactions with fibrin: a study with recombinant apolipoprotein(a) and isolated plasminogen fragments.

Lipoprotein(a) [Lp(a)], but not low-density lipoprotein (LDL), was previously shown to impair the generation of fibrin-bound plasmin [Rouy et al. (1991) Arterioscler. Thromb. 11, 629-638] by a mechanism involving binding of Lp(a) to fibrin. It was therefore suggested that the binding was mediated by apolipoprotein(a) [apo(a)], a glycoprotein absent from LDL which has a high degree of homology with plasminogen, the precursor of the fibrinolytic enzyme plasmin. Here we have evaluated this hypothesis by performing comparative fibrin binding studies using a recombinant form of apo(a) containing 17 copies of the apo(a) domain resembling kringle 4 of plasminogen, native Lp(a), and Glu-plasminogen (Glu1-Asn791). Attempts were also made to identify the kringle domains involved in such interactions using isolated elastase-derived plasminogen fragments. The binding experiments were performed using a well-characterized model of an intact and of a plasmin-digested fibrin surface as described by Fleury and Anglés-Cano [(1991) Biochemistry 30, 7630-7638]. Binding of r-apo(a) to the fibrin surfaces was of high affinity (Kd = 26 +/- 8.4 nM for intact fibrin and 7.7 +/- 4.6 nM for plasmin-degraded fibrin) and obeyed the Langmuir equation for adsorption at interfaces. The binding to both surfaces was inhibited by the lysine analogue AMCHA and was completely abolished upon treatment of the degraded surface with carboxypeptidase B, indicating that r-apo(a) binds to both the intrachain lysines of intact fibrin and the carboxy-terminal lysines of degraded fibrin. As expected from these results, both r-apo(a) and native Lp(a) inhibited the binding of Glu-plasminogen to the fibrin surfaces.(ABSTRACT TRUNCATED AT 250 WORDS)     

Enhanced plasminogen activation by staphylokinase in the presence of streptokinase beta/betagamma domains: plasminogen kringles play a role

Presence of isolated beta or betagamma domains of streptokinase (SK) increased the catalytic activity of staphylokinase (SAK)-plasmin (Pm) complex up to 60%. In contrast, fusion of SK beta or betagamma domains with the C-terminal end of SAK drastically reduced the catalytic activity of the activator complex. The enhancement effect mediated by beta or betagamma domain on Pg activator activity of SAK-Pm complex was reduced greatly (45%) in the presence of isolated kringles of Pg, whereas, kringles did not change cofactor activity of SAK fusion proteins (carrying beta or betagamma domains) significantly. When catalytic activity of SAK-microPm (catalytic domain of Pm lacking kringle domains) complex was examined in the presence of isolated beta and betagamma domains, no enhancement effect on Pg activation was observed, whereas, enzyme complex formed between microplasmin and SAK fusion proteins (SAKbeta and SAKbetagamma) displayed 50-70% reduction in their catalytic activity. The present study, thus, suggests that the exogenously present beta and betagamma interact with Pg/Pm via kringle domains and elevate catalytic activity of SAK-Pm activator complex resulting in enhanced substrate Pg activation. Fusion of beta or betagamma domains with SAK might alter these intermolecular interactions resulting in attenuated functional activity of SAK.     

The aromatic 1H-NMR spectrum of plasminogen kringle 4. A comparative study of human, porcine and bovine homologs

The isolated kringle 4 domain of human plasminogen has been compared with homologous structures from bovine and porcine sources, both free and in the presence of the ligand 6-aminohexanoic acid, by two-dimensional 1H-NMR spectroscopies at 300 MHz and 600 MHz. The chemical-shift-correlated, spin-echo-correlated, and double-quantum-correlated aromatic spectra of the three proteins reveal that the globular conformation of the fourth kringle is closely maintained throughout the set of homologs. Direct comparison shows that the three conserved Trp residues (at sites 25, 62 and 72) which exhibit highly non-degenerate subspectra, find themselves in similar intramolecular environments. In particular, proton Overhauser experiments reveal that the close steric interaction between the Trp-II (Trp62 or Trp25) indole group and the aromatic ring at site 74 (Tyr74 or Phe74) is strictly preserved. This feature forces the kringle inner loop, closed by the Cys51-Cys75 link, to fold back onto itself so as to place the site 74 residue proximal to the Cys22-Cys63 bridge. Single-residue substitutions enable unambiguous assignments of His-I to His3, Tyr-III to Tyr41 and Tyr-IV to Tyr74. From this direct evidence, comparison with the kringle 1 spectrum, and the previously reported chemical modification of Tyr-II (Tyr50) [Trexler M., Bányai L., Patthy L., Pluck N. D. & Williams R. J. P. (1985) Eur. J. Biochem. 152, 439-446], Tyr-I and Tyr-V (the latter, an immobile ring on the 600-MHz time scale) could be assigned to Tyr2 and Tyr9, respectively. Since Trp-III has previously been assigned to Trp72 at the lysine-binding site, the present study completes the assignment of 10 out of 12 aromatic spin systems in the kringle 4 1H-NMR spectrum; the only ambiguity which remains concerns the Trp-I and Trp-II indole spin systems, which are totally identified but as yet only tentatively assigned to Trp25 and Trp62, respectively.