Fibrin-specific and effective clot lysis requires both plasminogen activators and for them to be in a sequential rather than simultaneous combination.

Thrombolysis with tissue plasminogen activator (tPA) has been a disappointment and has now been replaced by an endovascular procedure whenever possible. Nevertheless, thrombolysis remains the only means by which circulation in a thrombosed artery can be restored rapidly. In contrast to tPA monotherapy, endogenous fibrinolysis uses both tPA and urokinase plasminogen activator (uPA), whose native form is a proenzyme, prouPA. This combination is remarkably effective as evidenced by the fibrin degradation product, D-dimer, which is invariably present in plasma. The two activators have complementary mechanisms of plasminogen activation and are synergistic in combination. Since tPA initiates fibrinolysis when released from the vessel wall and prouPA is in the blood, they induce fibrinolysis sequentially. It was postulated that this may be more effective and fibrin-specific. The hypothesis was tested in a model of clot lysis in plasma in which a clot was first exposed to tPA for 5 min, washed and incubated with prouPA. Lysis was compared with that of clots incubated with both activators simultaneously. The sequential combination was almost twice as effective and caused less fibrinogenolysis than the simultaneous combination (p < 0.0001) despite having significantly less tPA, as a result of the wash. A mechanism is described by which this phenomenon can be explained. The findings are believed to have significant therapeutic implications.

C1-inhibitor prevents non-specific plasminogen activation by a prourokinase mutant without impeding fibrin-specific fibrinolysis.

BACKGROUND: Prourokinase (prouPA) is unstable in plasma at therapeutic concentrations. A mutant form, M5, made more stable by reducing its intrinsic activity was therefore developed. Activation to two-chain M5 (tcM5) induced a higher catalytic activity than two-chain urokinase plasminogen activator (tcuPA), implicating an active site functional difference. Consistent with this, an unusual tcM5 complex with plasma C1-inhibitor was recently described in dog and human plasma. The effect of C1-inhibitor on fibrinolysis and fibrinogenolysis by M5 is the subject of this study. METHODS AND RESULTS: Zymograms of tcM5 and tcuPA incubated in plasma revealed prominent tcM5-C1-inhibitor complexes, which formed within 5 min. The inhibition rate by purified human C1-inhibitor (250 microg mL(-1)) was about 7-fold faster for tcM5 than it was for tcuPA (10 microg mL(-1)). The effect of the inhibitor on the stability of M5 and prouPA was determined by incubating them in plasma at high concentrations (10-20 microg mL(-1)) +/- C1-inhibitor supplementation. Above 10 microg mL(-1), depletion of all plasma plasminogen occurred, indicating plasmin generation and tcM5/tcuPA formation. With supplemental C1-inhibitor, M5 stability was restored but not prouPA stability. Clot lysis by M5 +/- supplemental C1-inhibitor showed no attenuation of the rate of fibrinolysis, whereas fibrinogenolysis was prevented by C1-inhibitor. Moreover, because of higher dose-tolerance, the rate of fibrin-specific lysis reached that achievable by non-specific fibrinolysis without inhibitor. CONCLUSIONS: Plasma C1-inhibitor stabilized M5 in its proenzyme configuration in plasma by inhibiting tcM5 and thereby non-specific plasminogen activation. At the same time, fibrin-specific plasminogen activation remained unimpaired. This unusual dissociation of effects has significant implications for improving the safety and efficacy of fibrinolysis.

Thrombolysis vs. bleeding from hemostatic sites by a prourokinase mutant compared with tissue plasminogen activator.

BACKGROUND: A single site mutant (M5) of prourokinase (proUK) was developed to make proUK less vulnerable to spontaneous activation in plasma. This was a problem that seriously compromised proUK in clinical trials, as it precluded proUK-mediated fibrinolysis at therapeutic concentrations. METHODS AND RESULTS: After completing dose-finding studies, 12 anesthetized dogs with femoral artery thrombosis were given either M5 (2.0 mg kg(-1)) or tissue plasminogen activator (t-PA) (1.4 mg kg(-1)) by i.v. infusion over 60 min (20% administered as a bolus). Two pairs of standardized injuries were inflicted at which hemostasis was completed prior to drug administration. Blood loss was quantified by measuring the hemoglobin in blood absorbed from these sites. Thrombolysis was evaluated at 90 min and was comparably effective by both activators. Rethrombosis developed in one t-PA dog. The principal difference found was that blood loss was 10-fold higher with t-PA (mean approximately 40 mL) than with M5 (mean approximately 4 mL) (P = 0.026) and occurred at more multiple sites (mean 2.7 vs. 1.2). This effect was postulated to be related to differences in the mechanism of plasminogen activation by t-PA and M5 in which the latter is promoted by degraded rather than intact (hemostatic) fibrin. In addition, two-chain M5 was efficiently inactivated by plasma C1 inactivator, an exceptional property which helped contain its non-specific proteolytic effect. CONCLUSIONS: Intravascular thrombolysis by M5 was accompanied by significantly less bleeding from hemostatic sites than by t-PA. This was attributed to the proUK paradigm of fibrinolysis being retained at therapeutic concentrations by the mutation.

Platelet associated u-PA up-regulates u-PA synthesis by endothelial cells.

Adhesion of platelets to endothelium has been shown to induce important changes in endothelial properties. In this study, we examined the effect of platelet-endothelial cell interactions on the expression of urokinase-type plasminogen activator (u-PA) by human microvascular endothelial cells. After incubation of endothelial cells with platelets, a dose-dependent increase in the expression of u-PA Ag was observed and reached a plateau for a ratio of 300 platelets per endothelial cells. The u-PA Ag upregulation resulted from an increase in u-PA mRNA that originated from a synthesis by endothelial cells since no u-PA mRNA was detected in platelets. The platelet-induced u-PA synthesis was inhibited when the endothelial cells were pre-treated with phospholipase C to remove the u-PA receptor, or when the platelets were incubated with an antibody that blocks the binding of u-PA to u-PAR. Taken together, these data indicate that u-PA present on the platelet surface interacts with u-PAR on the endothelial cells and induces the u-PA synthesis. This mechanism may represent a physiological control of platelet-mediated intravascular fibrin deposition.

Urokinase-type plasminogen activator up-regulates its own expression by endothelial cells and monocytes via the u-PAR pathway.

Signal transduction by urokinase-type plasminogen activator (u-PA) bound to its cell receptor has been well established. In the present study, we found, for the first time to our knowledge, that u-PA promotes its own synthesis by endothelial cells and monocytes. This phenomenon was characterized and shown to involve the u-PA receptor (u-PAR) pathway. The finding may be of general importance, since most cells that express u-PAR also produce u-PA. Human umbilical vein endothelial cells (HUVECs), U937 monocytes, and human peripheral blood monocytes (PFMCs) were incubated with diisopropylfluorophosphate (DFP)-pretreated u-PA, the amino-terminal fragment (ATF) of u-PA, or the kringle domain. A threefold up-regulation of u-PA secretion and synthesis by u-PA or ATF was found. The predominant effect was expressed in HUVECs, in which u-PA mRNA was also up-regulated. The u-PA kringle domain had no effect on u-PA synthesis, leading to the conclusion that the EGF domain was responsible. This was also consistent with the additional finding that the u-PAR, to which the EGF domain binds, was necessary for the up-regulation. The results indicate that u-PA up-regulates itself via its EGF domain and u-PAR. The possibilities that the results were related to displacement of receptor-bound u-PA or the blocking of u-PA incorporation into the cells were excluded. A modest up-regulation of u-PAR was also associated with this phenomenon.

Fibrinolysis: an unfinished agenda.

There has been a recent decline in interest in fibrinolysis, suggesting that its physiological basis is sufficiently understood and that therapeutic thrombolysis has reached its limit. The importance of the subject has not diminished since cardiovascular disease is now a leading health problem even in developing countries. Certain highlights and inconsistencies are reviewed. The clinical trials of tissue plasminogen activator (t-PA) revealed a major discrepancy between its fibrinolytic efficacy and its clinical benefit (the 't-PA paradox') that is unexplained. Dose-finding studies also showed that the fibrinolytic efficacy of t-PA required significant nonspecific plasminogen activation. Furthermore, the longstanding belief that t-PA is responsible for physiological fibrinolysis and urokinase-type PA (u-PA) for pericellular plasminogen activation is belied by extensive experimental animal data, but these findings have had little impact on traditional thinking. As a result, the mechanisms responsible for the u-PA paradigm of fibrinolysis have received little attention. Clinical experience with pro-u-PA remains limited and most clinical trials have used infusion rates at which pro-u-PA is largely converted systemically to urokinase. This is due to the unanticipated instability of pro-u-PA in plasma at pharmacological concentrations. Insufficient understanding of basic mechanisms of fibrinolysis has handicapped the design of chimeric or mutant activators. It is submitted that physiological fibrinolysis remains to be better defined, and that it is premature to conclude that therapeutic thrombolysis will be inevitably accompanied by side effects that undermine this method of inducing reperfusion.

Catalytic and fibrinolytic properties of recombinant urokinase plasminogen activator from E. coli, mammalian, and yeast cells.

The enzymatic and fibrinolytic properties of glycosylated and nonglycosylated recombinant human pro-urokinase (pro-UK) produced in yeast Pichia pastoris were characterized and compared with those of Escherichia coli and mammalian cell-derived pro-UK. Among the five different forms of pro-UK, the yeast glycosylated pro-UK was activated by plasmin with the lowest catalytic efficiency (kcat/Km). The yeast glycosylated urokinase (UK) also had the highest Km in its activation of Glu-plasminogen, and had a substantially lower fibrinolytic activity than the other four forms. These findings suggest that the poly-mannose on Asn-302 of yeast glycosylated pro-UK interfered with its activation by plasmin and its binding interaction with plasminogen. By contrast to plasminogen, the activation of the small synthetic substrate, S2444, was comparable for all five forms of recombinant UK. It is concluded that the glycosyl residue on pro-UK/UK is functionally important and modulates its activatability and its catalytic efficiency against its natural substrate. Therefore, pro-UK from different expression systems cannot be assumed to have comparable fibrinolytic activities.

Evidence for the expression of urokinase-type plasminogen activator by human venous endothelial cells in vivo.

Endothelial cells (ECs) in culture synthesize and secrete urokinase-type plasminogen activator (u-PA), but the normal vascular endothelium is believed to synthesize only tissue plasminogen activator (t-PA), which is thought to be responsible for intravascular fibrinolysis. More recently, animal studies have shown that the biological role of u-PA in fibrinolysis has been underestimated, prompting a re-examination of its synthesis by the endothelium. In this study, we investigated whether u-PA was synthesized by non-atherosclerotic endothelial cells in vivo by testing ECs dislodged by venipuncture from 12 normal volunteers and 17 patients admitted for plasmapheresis. The ECs were isolated with an anti-endothelial monoclonal antibody coupled to immunomagnetic beads and characterized by morphology and by labelling for vWF, CD31, and UEA-1 binding. U-PA antigen was found in 50% of the ECs from the normal subjects and in 60% of those from patients. U-PA enzymatic activity on zymograms was detected in 50% of the normal samples and 60% of the patient samples, with the latter being more frequently and more strongly positive. U-PA mRNA was found in all the normal and patient samples tested. The results indicate that u-PA is synthesized by the venous endothelium in vivo but that its expression is highly variable.

Analysis of the forces which stabilize the active conformation of urokinase-type plasminogen activator.

It was recently proposed that hydrophobic interactions control the active conformation of serine proteases in the trypsin family (Hedstrom et al. (1996) Biochemistry 35, 4515-23) rather than a charge interaction with Asp next to the active site Ser, as formerly believed. In the present study, certain site-directed mutants of the serine protease zymogen pro-urokinase (pro-UK) and its two-chain enzymatic derivative urokinase (UK) were characterized. The results provide information on the structure-function of the catalytic domain of pro-UK/UK, which is relevant to this controversy. Mutations at Asp355(c194), which eliminated its charge, induced a 6250-fold reduction in the catalytic activity of UK. By contrast, reducing the hydrophobicity at the neoterminal Ile159(c16) of UK had relatively little effect. However, when both the hydrophobicity and the size of the side chain were reduced by a glycine substitution at this position, a major reduction (9090-fold) in the catalytic efficiency of UK occurred. This effect was related to the smaller side chain increasing the cavity and the flexibility of the N-terminus and thereby interfering with its charge interaction with Asp355(c194). A similar mechanism, rather than a change in hydrophobicity, is believed also to explain the reduction in the stabilization energy of the activation domain observed in a trypsin mutant by Hedstrom et al. (1996). Although hydrophobic interaction facilitated the charge interaction with Asp355(c194), the latter was the primary force which stabilized the active conformation of UK. The charge interaction with Asp355(c194) was also found to be the principal determinant of the intrinsic catalytic activity of single-chain pro-UK. Additionally, the findings confirmed that the KM of pro-UK for its natural substrate was significantly lower than that of UK. Since this same phenomenon was also seen with each of the mutants, the substrate binding pocket of these single-chain zymogens was better formed than that of their two-chain, enzymatic derivatives.

Demonstration of covalent binding of lipoprotein(a) [Lp(a)] to fibrin and endothelial cells.

It has been well documented that Lp(a) binds noncovalently to fibrin or human umbilical vein endothelial cells. This binding is to lysines and is inhibited by lysine analogues such as epsilon-aminocaproic acid (EACA). In the present study, Lp(a) (0.006-0.6 microM) binding to immobilized fibrin and endothelial cells was evaluated by ELISA with an anti-Lp(a) antibody. A significant portion (approximately 65%) of the Lp(a) was found to resist dissociation by EACA (0.2 M). The EACA resistant binding of Lp(a) was time and concentration dependent. The addition of EDTA to the incubation mixture had no effect, thereby excluding cross-linking by transglutaminase as a mechanism. This portion of Lp(a) was also resistant to dissociation by acid (0.1 N HCl), 0.1% SDS, 1 M benzamidine, Tris-HCl (1 M, pH 12), or DTT (5 mM), but it was washed off by 0.1 N NaOH (which did not remove the immobilized fibrin). This suggested that the Lp(a) was covalently linked by an ester bond. Covalent binding was inhibited when Lp(a) was mildly oxidized by BioRad Enzymobeads, which may explain why it escaped recognition in experiments with radiolabeled Lp(a). Covalent binding was attenuated when Lp(a) was pretreated with DFP suggesting that the serine residue in the pseudo active site of Lp(a) was involved. Lp(a) also bound covalently to immobilized BSA, indicating some nonspecificity. However, binding to BSA was almost 3-fold less than to fibrin, suggesting that lysine binding may facilitate covalent binding. A similar proportion of EACA resistant binding of Lp(a) was found with endothelial cells. In conclusion, the findings demonstrate a novel, covalent binding by Lp(a) which is kringle independent and is postulated to involve the pseudo protease domain of Lp(a). This property may contribute to the deposition of Lp(a) on endothelial surfaces and its colocalization with fibrin in atheromas.

Thrombin stimulation of platelets induces plasminogen activation mediated by endogenous urokinase-type plasminogen activator.

Gene knockout mice studies indicate that urokinase-type plasminogen activator (u-PA) is importantly involved in fibrinolysis, but its physiologic mechanism of action remains poorly understood. We postulated that platelets may be involved in this mechanism, as they carry a novel receptor for u-PA and a portion of the single-chain u-PA (scu-PA) intrinsic to blood is tightly associated with platelets. Therefore, plasminogen activation by platelet-associated u-PA was studied. When washed platelets were incubated with plasminogen, no plasmin was generated as detected by plasmin synthetic substrate (S2403) hydrolysis; however, after the addition of thrombin, but not other agonists, platelet-dependent plasminogen activation occurred. Plasminogen activation was surface-related, being inhibited by blocking platelet fibrinogen receptors or by preventing plasminogen binding to the thrombin-activated platelet surface. U-PA was identified as the only plasminogen activator responsible and enrichment of platelets with exogenous scu-PA significantly augmented plasminogen activation. These findings appeared paradoxical because thrombin inactivates scu-PA. Indeed, zymograms showed inactivation of scu-PA during the first hour of incubation with even the lowest dose of thrombin used (1 u/mL). However, this was followed by a thrombin dose-dependent (1 to 10 u/mL) partial return of u-PA activity. Reactivation of u-PA was not due to the direct action of thrombin, but required platelets and was found to be related to a platelet lysosomal thiol protease, consistent with cathepsin C. In conclusion, a new pathway of plasminogen activation by platelet-associated endogenous or exogenous scu-PA was demonstrated, which is specifically triggered by thrombin activation of platelets. These findings may help explain u-PA-mediated physiological fibrinolysis and have implications for therapeutic thrombolysis with scu-PA.

Identification of a flexible loop region (297-313) of urokinase-type plasminogen activator, which helps determine its catalytic activity.

Pro-urokinase has a much higher intrinsic catalytic activity than other zymogens of the serine protease family. Lys300(c143) in an apparent "flexible loop" region (297-313) was previously shown to be an important determinant of this intrinsic catalytic activity. This was related to the loop allowing the positive charge of Lys300(c143) to transiently interact with Asp355(c194), thereby inducing an active conformation of the protease domain (Liu, J. N., Tang, W., Sun, Z., Kung, W., Pannell, R., Sarmientos, P., and Gurewich, V. (1996) Biochemistry 35, 14070-14076). To further test this hypothesis, the charge at position 300(c143) and the flexibility of the loop were altered using site-directed mutagenesis designed according to a computer model to affect the interaction between Lys300(c143) and Asp355(c194). When the charge at Lys300(c143) but not Lys313(c156) was reduced, a significant reduction in the intrinsic catalytic activity occurred. Similarly, when the flexibility (wobbliness) of the loop was enhanced reducing the size of side chain, the intrinsic catalytic activity was also reduced. By contrast, when the loop was made less flexible, the intrinsic catalytic activity was increased. These findings were consistent with the hypothesis. The effects of these mutations on two-chain activity were less and often discordant with the intrinsic catalytic activity, indicating that they can be modulated independently. This structure-function disparity can be exploited to create a more zymogenic pro-urokinase (lower intrinsic catalytic activity) with a high catalytic activity, as exemplified by two of the mutants. The changes in intrinsic catalytic activity and two-chain activity induced by the mutations were due to changes in kcat rather than Km. Some significant structure-function differences between pro-urokinase and its highly homologous counterpart, tissue plasminogen activator, were also found.

An efficient system for production of recombinant urokinase-type plasminogen activator.

A system was developed to produce recombinant urokinase-type plasminogen activator in Escherichia coli. The urokinase-type plasminogen activator was produced with a 6 x His-tag at the C-terminus which was shown to have the same activity, after refolding, as the wild-type protein. Purification of the recombinant protein to homogeneity (95%) was possible by one-step affinity chromatography under denaturing conditions. As a result, proteolysis by bacterial proteases during purification was avoided. A higher refolding efficiency (40%) and a higher total recovery yield (25%) of the recombinant protein were obtained by this method.