Adaptation of plasminogen activator sequences to known protease structures

Streptokinase—a clinically useful thrombolytic agent

A failure of hemostasis and consequent formation of blood clots in the circulatory system can produce severe outcomes such as stroke and myocardial infraction. Pathological development of blood clots requires clinical intervention with fibrinolytic agents such as urokinase, tissue plasminogen activator and streptokinase. This review deals with streptokinase as a clinically important and cost-effective plasminogen activator. The aspects discussed include: the mode of action; the structure and structure-function relationships; the structural modifications for improving functionality; recombinant streptokinase; microbial production; and recovery of this protein from crude broths.

Complexation of the tissue plasminogen activator protease with benzamidine-type inhibitors: interference by the kringle 2 module

Well-resolved high-field 1H NMR signals between -0.1 and -0.7 ppm afford convenient probes to monitor the conformational state of the tissue plasminogen activator (tPA) protease, modulated by covalent inhibitor binding or activation cleavage [Hu, C.-K., Kohnert, U., Wilhelm, O., Fischer, S., & Llinas, M. (1994) Biochemistry 33, 11760-11766]. We have investigated recombinant BM 06.022 (a domain-deletion variant mutant from Escherichia coli comprising the kringle 2 and protease modules) and protease constructs of tPA in both single-chain (sc) and two-chain (tc) forms. The two proteins were studied when confronted with the noncovalent (i.e., reversible) active site inhibitors benzamidine and a series of bisbenzamidine derivatives: 2,5-bis(4-amidinobenzylidene)cyclopentanone, 2,6-bis(4-amidinobenzylidene)cyclohexanone, 2,7-bis(4-amidinobenzylidene)cycloheptanone, and 2,8-bis(4-amidino- benzylidene)cyclooctanone. At pH* 4.6, the 1H NMR spectrum is sensitive to complexation of the protease module with the various effectors. The amplitude of the inhibitor-shifted resonances is more pronounced for the tc-protease than for the sc-protease, suggesting that access of inhibitors to the protease catalytic site is facilitated upon conversion to the tc form. The effects detected by the NMR spectrum suggest a biphasic process, involving stronger (primary) and weaker (secondary) bindings to a single protease active site. Binding to the protease module in tc-BM 06.022 essentially generates the same spectral characteristics as detected upon binding to the isolated tc-protease construct. In contrast, a negligible perturbation by the inhibitors is observed on the (sc) BM 06.022. Hence, in the intact BM 06.022 the kringle 2 is structurally coupled to the protease module thus interfering with inhibitor molecules from accessing the protease active site. These domain-domain interactions relax upon conversion to the catalytically active tc form, thus decoupling the kringle 2 from the protease module in BM 06.022 while simultaneously exposing the active site to become accessible to effectors or substrates.

Knowledge-based protein modeling

Knowledge, both from the three-dimensional structures of homologous proteins and from the general analysis of protein structure, is of value in modeling a protein of known sequence but unknown structure. While many models are still constructed at least in part by manual methods on graphics devices, automated procedures have come into greater use. These procedures include those that assemble fragments of structure from other known structures and those that derive coordinates for the model from the satisfaction of restraints placed on atomic positions.

Physical characterization of recombinant tissue plasminogen activator

Electron microscopic and physical-chemical properties of one- and two-chain tissue plasminogen activator (t-PA) were studied. The molecular weight of one-chain t-PA obtained by both sedimentation equilibrium and SDS-PAGE was estimated to be about 65,000, while both chains in the reduced two-chain form were in the range of 35,000-40,000. Sedimentation coefficients were identical for both forms of t-PA (S(0)20,w = 4.12). The two forms of t-PA were indistinguishable by electron microscopic analysis, which confirmed the sedimentation results, and showed that they were ellipsoidal and relatively compact. The major and minor axes were approx. 13 nm and approx. 10 nm and f/f0 was 1.36. The individual domains of t-PA are relatively small and are folded within the molecule, so that the overall appearance is globular.

Tissue-type plasminogen activator mutants imitating urokinase in the peptide link between kringle and protease domains and at selected sites within the protease domain

Tissue-type plasminogen activator (tPA) mutants which, at selected amino acid positions, mimic urokinase-type plasminogen activator (uPA) were expressed in Chinese hamster ovary cells and examined for their catalytic properties. In one series of mutants, the dipeptide Ser262 Thr263 between kringle 2 and the protease domain of tPA was (a) replaced by an Ala residue, (b) lengthened by additional Ser and Ala residues, (c) exchanged for the 16-amino-acid link between kringle and protease domains of uPA and an additional Ala residue. The activities of the latter two mutants toward plasminogen were, in the absence of fibrin, 3-5-fold higher and, in the presence of fibrin, comparable to or lower than the activity of tPA. The kinetic data suggest a short interdomain peptide in tPA as most favorable for high fibrin stimulation of tPA activity. In a second series of mutant, selected amino acid residues of the tPA protease domain were replaced by residues of the homologous uPA domain. Positions chosen for exchange are either close to the active site or are part of a tPA-specific insertion in the variable region preceding the active-site Ser residue. Compared to authentic tPA, protease-domain mutants exhibited 7.3-424-fold lower activities toward plasminogen, mainly due to lower kcat values. Km values differed only moderately. A mutant containing an additional hydroxyl group at the S1 site, tPA A473S, had lost the preference of tPA for Arg over Lys as the P1 residue in peptide substrates.

Substrate specificity of tissue-type and urokinase-type plasminogen activators

Recent studies suggest that plasminogen activators not only hydrolyse a specific arginine-valine bond in plasminogen, but may also cleave other proteins such as fibronectin. We studied the substrate specificity, particularly the preference for arginyl over lysyl peptide bonds, of tissue-type plasminogen activator (t-PA) as well as of two-chain urokinase-type plasminogen activator (u-PA). The arginine/lysine preference was determined with three pairs of tripeptidyl-p-nitroanilide substrates having either arginine or lysine in the P1 position and varied from 5.2 to 14.1 for u-PA and from 55.6 to 99.8 for t-PA. It was concluded that both t-PA and u-PA preferred arginyl to lysyl peptide bonds. However, u-PA had a significantly lower arginine/lysine preference than t-PA, indicating that u-PA represents a less specific proteinase. This may point to functions of u-PA other than plasminogen activation, which involve cleavage of lysyl bonds.

Amino acid residues that affect interaction of tissue-type plasminogen activator with plasminogen activator inhibitor 1

Fibrinolysis is regulated in part by the interaction between tissue-type plasminogen activator (t-PA) and plasminogen activator inhibitor 1 (PAI-1, a serine protease inhibitor of the serpin family). It is known from our earlier work that deletion of a loop of amino acids (residues 296-302) from the serine protease domain of t-PA suppresses the interaction between the two proteins without altering the reactivity of t-PA towards its substrate, plasminogen. To define more precisely the role of individual residues within this loop, we have used site-directed mutagenesis to replace Lys-296, Arg-298, and Arg-299 with negatively charged glutamic residues. Replacement of all three positively charged amino acids generates a variant of t-PA that associates inefficiently with PAI-1 and is highly resistant to inhibition by the serpin. Two t-PAs with point mutations (Arg-298----Glu and Arg-299----Glu) are partially resistant to inhibition by PAI-1 and associate with the serpin at intermediate rates. Other point mutations (Lys-296----Glu, His-297----Glu, and Pro-301----Gly) do not detectably affect the interaction of t-PA with PAI-1. None of these substitutions has a significant effect on the rate of catalysis by t-PA or on the affinity of the enzyme for its substrate, plasminogen. On the basis of these results, we propose a model in which positively charged residues located in a surface loop near the active site of t-PA form ionic bonds with complementary negatively charged residues C-terminal to the reactive center of PAI-1.

Quenching of the amidolytic activity of one-chain tissue-type plasminogen activator by mutation of lysine-416

In contrast to most other serine proteases, tissue-type plasminogen activator (t-PA) possesses enzymatic activity as the one-chain zymogen form. The hypothesis that lysine residues 277 or 416 may be involved in stabilization of an active conformation of one-chain t-PA via salt-bridge formation with aspartic acid residue 477 was tested by site-directed mutagenesis. Four recombinant t-PA mutants were constructed. The amidolytic activities of these analogues were compared to that of authentic t-PA. Substitution of arginine-275 provided an analogue [( R275G]t-PA) resistant to plasmin cleavage. The amidolytic activity of [R275G]t-PA was comparable to that of authentic one-chain t-PA, and so was the activity of [R275L,K277L]t-PA, in which additional substitution of lysine residue 277 was carried out. This suggested that its presence was nonessential for obtaining one-chain t-PA activity. In contrast, substitution of lysine residue 416 to obtain [K416S]t-PA and [K416S,H417T]t-PA resulted in substantial quenching of amidolytic one-chain activity. As expected, the amidolytic activities of the two-chain forms were less affected by the substitution. Involvement of lysine residue 416 in one-chain t-PA activity was also indicated by decreased activities of [K416S]t-PA and [K416S,H417T]t-PA with plasminogen as the substrate. The one-chain activity of the lysine residue 416 substitution analogues was partially restored in the presence of fibrin. This could indicate that strong ligands such as fibrin might provide an alternative stabilization of the active conformation of one-chain t-PA.

Serpin-resistant mutants of human tissue-type plasminogen activator

Tissue-type plasminogen activator (t-PA) converts the inactive zymogen, plasminogen, into the powerful protease, plasmin, which then degrades the fibrin meshwork of thrombi. To prevent systemic activation of plasminogen, plasma contains several inhibitors of t-PA, the most important of which is plasminogen activator inhibitor-1 (PAI-1), a member of the serpin superfamily. As the ability to produce serpin-resistant variants of t-PA could increase the potential of this enzyme as a thrombolytic agent, we have used the known three-dimensional structure of the complex between trypsin and bovine pancreatic trypsin inhibitor (BPTI) to model the interactions between the active site of human t-PA and PAI-1. On the basis of this model we then altered by site-directed mutagenesis those amino acids of t-PA predicted to make contact with PAI-1 but not with the substrate plasminogen. We report here that although the resulting mutants have enzymatic properties similar to those of wild-type t-PA, they display significant resistance to inhibition by PAI-1. For example, following incubation with an amount of the serpin that completely inhibits the wild-type enzyme, one variant retains 95% of its initial activity. This mutant is also resistant to inhibition by the complex mixture of serpins present in human plasma.

Prediction of the three-dimensional structure of the enzymatic domain of t-PA

Tissue plasminogen activator (t-PA), an enzyme of the fibrinolytic system, is responsible for lysis of fibrin via activation of plasminogen, and therefore for degradation of blood clots. There are currently no X-ray crystal structure data of the t-PA molecule available either in whole or in part. We therefore predicted the three-dimensional structure of the protease domain by means of computer-graphical methods*. The model obtained forms a basis for understanding the binding of plasminogen to the active site of t-PA. In addition, the interactions of various inhibitors with t-PA were studied by modeling them into the active site. The model also yields an explanation for the observed amidolytic activity of t-PA in the single chain form.

The effect of polymerised fibrin on the catalytic activities of one-chain tissue-type plasminogen activator as revealed by an analogue resistant to plasmin cleavage

A one-chain recombinant tissue-type plasminogen activator (EC 2.4.31.-) (tPA) analogue was constructed in which Arg-275 of the activation site was changed to Gly by site-directed mutagenesis. This analogue, tPA-Gly275, was very resistant to plasmin (EC 2.4.21.5) cleavage. It has been used to gain information about the activity of the uncleaved one-chain tPA form, also when plasmin is generated as a result of a plasminogen activation reaction. The amidolytic activity of tPA-Gly275 with less than Glu-Gly-Arg-pNA was investigated and compared to that of one-chain and two-chain wild-type recombinant tPA. A small but significant intrinsic amidolytic activity was observed with the analogue as well as the wild-type one-chain tPA form. However, it was much lower than that of two-chain tPA. Polymerised fibrin enhanced the amidolytic activity of both one-chain tPA forms but not of two-chain tPA. Measurements of the plasminogen activation kinetics in the absence of fibrin revealed that tPA-Gly275 possessed a significant intrinsic activity. However, it was 30-fold lower than that of two-chain tPA. Addition of polymerised fibrin profoundly enhanced the plasminogen activation rate of both tPA-Gly275 and wild-type one- and two-chain tPA to approximately the same maximal level. The results were interpreted to mean that fibrin binding can induce an activated state of the intact tPA one-chain form.

Binding of tissue plasminogen activator to cultured human endothelial cells

Tissue plasminogen activator (t-PA) and urokinase (u-PA), the major activators of plasminogen, are synthesized and released from endothelial cells. We previously demonstrated specific and functional binding of plasminogen to cultured human umbilical vein endothelial cells (HUVEC). In the present study we found that t-PA could bind to HUVEC. Binding of t-PA to HUVEC was specific, saturable, plasminogen-independent, and did not require lysine binding sites. The t-PA bound in a rapid and reversible manner, involving binding sites of both high (Kd, 28.7 +/- 10.8 pM; Bmax, 3,700 +/- 300) and low (Kd, 18.1 +/- 3.8 nM; Bmax 815,000 +/- 146,000) affinity. t-PA binding was 70% inhibited by a 100-fold molar excess of u-PA. When t-PA was bound to HUVEC, its apparent catalytic efficiency increased by three- or fourfold as measured by plasminogen activation. HUVEC-bound t-PA was active site-protected from its rapidly acting inhibitor: plasminogen activator inhibitor. These results demonstrate that t-PA specifically binds to HUVEC and that such binding preserves catalytic efficiency with respect to plasminogen activation. Therefore, endothelial cells can modulate hemostatic and thrombotic events at the cell surface by providing specific binding sites for activation of plasminogen.

The use of folding patterns in the prediction of protein topologies

We proved previously that the distribution of formation energies which may be associated with the predicted secondary structures (or nuclei) is specific of the folding process (Busetta, B. 1986, Biochim. Biophys. Acta 870, 327-338). We developed a new predictive algorithm for protein topologies, based on the search of standard 'folding patterns'. In another manner, the strongest predicted nuclei are used to propose a fast sequence-alignment process which is efficient for distantly related proteins.

Characterization of primary amino acid sequence of human complement control protein factor I from an analysis of cDNA clones

A cDNA clone of the mRNA coding for the human complement system control protein Factor I has been isolated. The predicted amino acid sequence obtained from the DNA sequence demonstrates a protein consisting of a heavy chain (Mr 35,400) linked to a light chain (Mr 27,600), both of which contain three sites for N-linked glycosylation. The light chain has clear homology with other serine proteinases, most notably in the region of the catalytically active and structurally important amino acids and shares some of the features characteristic of the plasminogen activators. The heavy chain has a clear 'mosaic' nature typical of the plasma serine proteinases; in particular it contains class A and class B LDL (low-density lipoprotein) receptor repeats with conserved cysteine residues similar to those found in other complement proteins.

Proenzyme to urokinase-type plasminogen activator in human colon cancer: in vitro inhibition by monocyte minactivin after proteolytic activation

Marked increases of plasminogen activator activity were observed in human colon cancer tissue, compared to corresponding normal tissues. This increase was attributable to urokinase-type activator (HPA52), with no increase evident in the level of the tissue-type plasminogen activator (HPA66). Human monocyte minactivin specifically inhibited HPA52 activity in cancer tissue homogenates and in colon cancer cell supernatants, an effect that was greatly enhanced by preincubation with plasminogen, indicating that the predominant form of HPA52 in tissue and the form that is secreted in vitro is the proenzyme. Inactivation of HPA52 by minactivin was shown to be dependent on proteolytic activation of HPA52 proenzyme. Utilization of HPA52 activity by tumors in vivo could therefore be dependent upon a protease, such as plasmin, to generate the extracellular proteolytic activity necessary to digest the intercellular matrix and permit invasion of normal tissue structures by colon cancer cells.

Pharmacology of thrombolytic drugs

Streptokinase and urokinase have proved to be useful in a limited number of clinical conditions. Mainly because of the risk and unpredictability of bleeding with this first generation of thrombolytic agents, thrombolysis has not been ingrained in medical practice. In the interim, more fibrin-specific thrombolytic agents have been developed such as acylated streptokinase-human plasminogen complex, tissue-type plasminogen activator (t-PA) and single chain urokinase-type plasminogen activator (scu-PA or pro-urokinase). Only the latter two drugs do not induce major systemic fibrinogenolysis at thrombolytic effective doses. These two agents, obtained by recombinant techniques, as well as acylated streptokinase-plasminogen complex are available for clinical investigations. The first results of systemic administration of recombinant tissue-type plasminogen activation (t-PA) in patients with acute myocardial infarction were published and are promising. Continued experimentation with t-PA and pro-urokinase in evolving myocardial infarction and other thrombotic disorders is essential to better delineate their therapeutic index.

The fibrinolytic system in man

The fibrinolytic system comprises a proenzyme, plasminogen, which can be activated to the active enzyme plasmin, that will degrade fibrin by different types of plasminogen activators. Inhibition of fibrinolysis may occur at the level of plasmin or at the level of the activators. Fibrinolysis in human blood seems to be regulated by specific molecular interactions between these components. In plasma, normally no systemic plasminogen activation occurs. When fibrin is formed, small amounts of plasminogen activator and plasminogen adsorb to the fibrin, and plasmin is generated in situ. The formed plasmin, which remains transiently complexed to fibrin, is only slowly inactivated by alpha 2-antiplasmin, while plasmin, which is released from digested fibrin, is rapidly and irreversibly neutralized. The fibrinolytic process, thus, seems to be triggered by and confined to fibrin. Thrombus formation may occur as the result of insufficient activation of the fibrinolytic system and (or) the presence of excess inhibitors, while excessive activation and/or deficiency of inhibitors might cause excessive plasmin formation and a bleeding tendency. Evidence obtained in animal models suggests that tissue-type plasminogen activator, obtained by recombinant DNA technology, may constitute a specific clot-selective thrombolytic agent with higher specific activity and fewer side effects than those currently in use.

[Plasminogen activators: general aspects and recent developments]

Considerable interest in plasminogen activators as human thrombolytic drugs has stimulated rapid biotechnologic progresses. These enzymes have been classified in two immunochemically distinct groups: "urokinase-like" activators or u-PA which do not interact with fibrin and "tissue activator-like" activators or t-PA which interact with fibrin. Plasminogen activators are widely distributed in normal and malignant tissues and they are implicated in various physiological and pathological processes. They maintain the functional integrity of the vascular system and their presence may be of importance in tissue remodeling and cell migration. Urokinase and streptokinase are used in human thrombolytic therapy. However, the properties displayed by t-PA suggest that this enzyme may be a superior fibrinolytic agent. The primary structures of urokinase and t-PA are known; both enzymes have been synthesized by DNA technology. In order to produce t-PA in large quantities by gene cloning, intensive studies are conducted by pharmaceutical industries. Clinical trials using t-PA for dissolving thrombi in coronary heart disease, strokes and pulmonary embolism are in progress. This review presents the molecular and structural properties of plasminogen activators, as well as related physiological, pathological and therapeutic aspects.

The human urokinase-plasminogen activator gene and its promoter

The urokinase type of plasminogen activator (uPA) is subject to regulation by hormones, phorbol esters and oncogenic transformation. This enzyme has been suggested to play a key role in processes involving cell migration and tissue remodeling, and to be essential for tumor metastasis. In order to study these processes, we have isolated the human uPA gene, and have determined its entire nucleotide sequence. The gene is organized in 11 exons and is 6.4 kb long. The 5' end of uPA mRNA has been determined by both S1 mapping and primer extension experiments. A fragment of 800 bp containing the entire 5' flanking region shows promoter activity when introduced upstream of a bacterial chloramphenicol acetyltransferase gene and introduced into human cells. The hexanucleotide sequence GGCGGG, previously found at similar regions in several viral and eukaryotic promoters and shown to be essential for promoter activity (McKnight et al. (1984) Cell, 37, 253-262), is repeated three times between the CAAT and the TATA boxes.

Isolation of tissue-type plasminogen activator-inhibitor complexes from human plasma. Evidence for a rapid plasminogen activator inhibitor

Plasminogen activator-inhibitor complexes were analyzed by SDS-polyacrylamide gel electrophoresis and enzymography. The complexes appeared as fibrinolytically active bands in the fibrin-indicator gel. A high-molecular-weight t-PA form comigrating with a t-PA-inhibitor complex (Mr 95 000-135 000) from cultured human endothelial cells was purified from plasma by immunoadsorption on anti-t-PA-Sepharose followed by gel filtration on Sephadex G-150. The high-molecular-weight t-PA form was fibrinolytically inactive when assayed by the fibrin-plate method. It was converted to a form with the same electrophoretic mobility as t-PA (Mr 72 000) when treated with 1.5 M NH4OH/39 mM SDS. These observations suggested that the plasma high-molecular-weight t-PA form was an enzyme-inhibitor complex. The complex did not show immunological cross-reactivity with a number of known plasma serine proteinase inhibitors. Both t-PA and u-PA rapidly formed complexes with an inhibitor which was present in plasma in pmolar concentrations. p-Aminobenzamidine blocked the reaction, indicating that the active center of the activator was indeed implicated in complex formation. The complex between the plasma inhibitor and t-PA and the high-molecular-weight t-PA had the same electrophoretic mobilities. The rapid plasminogen activator inhibitor in plasma showed remarkable similarity to a plasminogen activator inhibitor from cultured human endothelial cells. In addition to the high-molecular-weight t-PA form described above, three other t-PA forms were isolated from plasma. Our results indicated that they represented free t-PA and t-PA in complex with respectively C1-esterase inhibitor and alpha 2-antiplasmin.

New approaches to thrombolytic therapy

Tissue-type plasminogen activator (t-PA), purified from the culture fluid of a stable human melanoma cell line, is a serine protease, different from urokinase, with a molecular weight of about 70,000. It is composed of one polypeptide chain, which is converted to a two-chain molecule by limited plasmic action. Activation of plasminogen to plasmin occurs by cleavage of the Arg 560-Val 561 peptide bond. Kinetic analysis has shown that the activation obeys Michaelis-Menten kinetics and that the presence of fibrin strikingly enhances the activation rate by increasing the affinity of plasminogen for fibrin-bound t-PA. The directed action of plasmin toward fibrin in vivo, might be explained by the low Michaelis constant in the presence of fibrin (0.16 microM), which allows efficient plasminogen activation on a fibrin clot, while its high value in the absence of fibrin (65 microM) prevents efficient activation in plasma. Plasmin formed on the fibrin surface would then be protected from rapid inactivation by alpha 2-antiplasmin. An important consequence of this molecular model for physiological fibrinolysis is that specific thrombolysis is only expected with the use of a specific plasminogen activator, which confines activation to the fibrin surface. Studies on the thrombolytic properties of purified t-PA in various animal species and in humans have revealed a higher specific thrombolytic activity than urokinase. Thrombolysis could be achieved without causing significant plasminogen activation, alpha 2-antiplasmin consumption, or fibrinogen breakdown. Alternatively, pro-urokinase, the zymogen precursor of urokinase, also displays a certain degree of fibrin specificity. Its mechanism of action and potential therapeutic value remain to be established.

The structure of the human tissue-type plasminogen activator gene: correlation of intron and exon structures to functional and structural domains

A genomic clone carrying the human tissue-type plasminogen activator (t-PA) gene was isolated from a cosmid library, and the gene structure was elucidated by restriction mapping, Southern blotting, and DNA sequencing. The cosmid contained all the coding parts of the mRNA, except for the first 58 bases in the 5' end of the mRNA, and had a total length of greater than 20 kilobases. It was separated into at least 14 exons by at least 13 introns, and the exons seemed to code for structural or functional domains. Thus, the signal peptide, the propeptide, and the domains of the heavy chain, including the regions homologous to growth factors, and to the "finger" structure of fibronectin, are all encoded by separate exons. In addition, the two kringle regions of t-PA were both coded for by two exons and were cleaved by introns at identical positions. The region coding for the light chain, comprising the serine protease part of the molecule was split by four introns, revealing a gene organization similar to other serine proteases.