The regulation by fibrinogen and fibrin of tissue plasminogen activator kinetics and inhibition by plasminogen activator inhibitor 1

Authors


Colin Longstaff, National Institute for Biological Standards and Control, South Mimms, Herts EN6 3QG, UK.
Tel.: +44 1707 641253; fax: +44 1707 641050; e-mail: clongstaff@nibsc.ac.uk

Abstract

Summary.   Background:  Tissue plasminogen activator (tPA) is unusual in the coagulation and fibrinolysis cascades in that it is produced as an active single-chain enzyme (sctPA) rather than a zymogen. Two chain tPA (tctPA) is produced by plasmin but there are conflicting reports in the literature on the behaviour of sc- and tctPA and little work on inhibition by the specific inhibitor plasminogen activator inhibitor-1 (PAI-1) under physiological conditions.

Objectives:  To perform a systematic study on the kinetics of sctPA and tctPA as plasminogen activators and targets for PAI-1.

Methods:  Detailed kinetic studies were performed in solution and in the presence of template stimulators, fibrinogen and fibrin, including native fibrin and partially digested fibrin. Numerical simulation techniques were utilized to cope with the challenges of investigating kinetics of activation and inhibition in the presence of fibrin(ogen).

Results:  Enzyme efficiency (kcat/Km) was higher for tctPA than sctPA in solution with chromogenic substrate (3-fold) and plasminogen (7-fold) but in the presence of templates, such as fibrinogen and native or cleaved fibrin, the difference disappeared. sctPA was more susceptible to PAI-1 in buffer solution and in the presence of fibrinogen; however, in the presence of fibrin, PAI-1 inhibited more slowly and there was no difference between sc and tctPA.

Conclusions:  Fibrinogen and fibrin modulate the activity of tPA differently in regard to their activation of plasminogen and inhibition by PAI-1. Fibrinogen and fibrin stimulate tPA activity against plasminogen but fibrin protects tPA from PAI-1 to promote fibrinolysis.

Introduction

Hemostasis is a balance between coagulation and fibrinolysis. The coagulation pathway involves a sequence or cascade of reactions to amplify a signal in order to generate localized activation of proteolytic enzyme activity and formation of a fibrin clot [1]. Much is known about the regulation of the coagulation cascade and about the structural basis of zymogen activation in general [2]. However, tissue plasminogen activator (tPA), the main plasminogen activator of the fibrinolysis system [3], is unique in being secreted as an active single-chain enzyme (sctPA) which nevertheless can undergo cleavage at Arg 275 and conversion to a two-chain form (tctPA). Rather than regulation by limited proteolysis and zymogen activation, tPA activity is regulated by release from the endothelium, by binding to inhibitors and by binding to fibrin. However, it is thought that most stimulation is accounted for by co-localization effects on fibrin, rather than by allosteric changes resulting from fibrin binding [3,4]. As part of our efforts to construct a model of the kinetics of plasminogen activation by tPA in fibrin utilizing our method for measuring precise rates of plasmin production [5], we require information on sc- and tctPA kinetics with plasminogen substrate and as targets for the specific serpin inhibitor plasminogen activator inhibitor 1 (PAI-1). Consequently, we report a systematic study of the reactions of sc- and tctPA under a variety of conditions relevant to such a model. Our findings suggest little difference in sc- and tctPA as plasminogen activators in most circumstances that might be physiologically relevant. There are differences in the susceptibility to PAI-1 such that sctPA reacts more rapidly than tctPA in buffer and in the presence of fibrinogen. In the presence of fibrin, tPA is protected from PAI-1, any differences between sc- and tctPA disappear and the balance shifts significantly towards plasminogen activation rather than inhibition.

Materials and methods

Generation of tPA expression constructs

Full-length cDNA for tPA was purchased from the HGMP Resource Centre (IMAGE clone ID:3161005). The clone was used as a template for polymerase chain reaction (PCR) with primers 5′-CGGGATCCACCATGGATGCAATGAAGAGAGG-3′ and 5′-CCGCTCGAGTCACGGTCGCATGTTGTC-3′. The 1709-bp amplification product was ligated into the BamHI/XhoI site of pFastBac 1 (Invitrogen, Carlsbad, CA, USA) to create pFastBac-tPA. A derivative of pFastBac-tPA was generated in which codon 275 of tPA was converted from CGC (Arginine) to GAA (Glutamic acid). Primers 5′-ACAGCCAGCCTCAGTTTGAAATCAAAGGAGGGCTCTTCGC-3′ and 5′-GCGAAGAGCCCTCCTTTGATTTCAAACTGAGGCTGGCTGT-3′ were used for site-directed mutagenesis via ‘QuikChange’ (Stratagene, La Jolla, CA, USA) according to the manufacturer's protocols with pFastBac-tPA as template to generate pFastBac- sctPA:R275E. The generated constructs were checked by sequence analysis.

Expression and purification of wt tPA and sctPA:R275E

The Bac-to-Bac baculovirus expression system (Invitrogen) was used to obtain recombinant baculovirus. Sf9 cells (100 mL, 2 × 106 cells mL−1 in serum-free SF900II media) were infected with either wild-type (wt) tPA or sctPA:R275E baculovirus at a multiplicity of infection of 5.0 for 72 h at 28 °C. Sf9 cells were removed by centrifugation and secreted recombinant tPA was purified from the culture media using a heparin-agarose column equilibrated with 50 mm Tris pH 7.7. Protein eluting with 0.75 m NaCl was applied to a lysine-Sepharose column (Sigma, St Louis, MO, USA) and eluted with 0.2 m tranexamic acid (Sigma) then dialyzed against 10 mm sodium acetate buffer pH 4. wt tPA and sctPA:R275E had similar specific activities to the International Standard for recombinant tPA (98/714),

Proteins were visualized by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS–PAGE) for size and purity and quantitated using Coomassie Plus Protein Assay Reagent (Pierce, Rockford, IL, USA) with the commercial tPA product Actilyse (Boehringer, Ingelheim, Germany) as a standard.

Two-chain tPA (tctPA) was generated from wt tPA by incubation with plasmin-sepharose in 0.2 m Tris buffer pH 7.6; 0.15 m NaCl; 0.01%Tween-20; at ambient temperature with constant gentle mixing. Optimum conditions for treatment were derived by a time-course experiment visualized by SDS–PAGE. The derivative sctPA:R275E was verified as non-cleavable with identical treatment. Reteplase was the NIBSC reference reagent code 93/726 (NIBSC, South Mimms, UK).

Kinetic analysis

Amidolytic activity was measured using the chromogenic substrate S2288, H-D-Ile-Pro-Arg-pNA (Chromogenix, Milan, Italy). Activation of plasminogen was studied in the presence of buffer only, fibrinogen and in fibrin by tPA utilizing the transparent clot with chromogenic substrate detection of plasmin method as published previously [5]. Where appropriate, ranges of plasminogen substrate were included (approx. 10–2000 nm) to determine apparent Km and kcat values. These values are labeled apparent as they refer to the conditions used in the particular assay. Modified fibrin with cleaved C-terminal lysines was made by adding CPB (Calbiochem, San Diego, CA, USA) to fibrinogen (plasminogen-free; Sigma) immediately prior to clotting such that each clot contains approx. 1 U CPB. Modified fibrin with additional C-terminal lysines was made by pretreating fibrinogen with plasmin-Sepharose. Plasmin-Sepharose was made by coupling 1.5 mg glu-plasminogen (Chromogenix) to cyanogen bromide activated Sepharose 4B (Sigma), 0.3 g swelled to approx. 1 mL gel, at pH 8.3 for at least 2 h at room temperature with mixing. Remaining active groups were blocked with 1 m ethanolamine prior to washing. The plasminogen Sepharose was then treated with 0.4 μm streptokinase for 30 min at room temperature with mixing, washed and stored as a 50% slurry. The optimum conditions for pretreating fibrinogen were derived by time-course. Typically, the plasmin gel was added to fibrinogen for 20 min at room temperature with mixing at a 1:1000 v/v dilution. The plasmin-Sepharose was removed by filtration prior to clotting.

For inhibition studies, native PAI-1 (Sigma Chemical Company) or a recombinant stabilized mutant (Calbiochem [6]) were used. Conditions were first optimized to give a suitable range of rates with a fixed plasminogen concentration and range of PAI-1 concentrations giving a good span of plasminogen activation rates. Rates of plasmin production were calculated from transformed plots of absorbance of p-nitroaniline at 405 nm vs. time squared as described previously [5] using the chromogenic substrate, S2251, H-D-Val-Leu-Lys-pNA, (Chromogenix) calculating the rate of pmol L−1 plasmin s−1 generated in the conditions used for the initial phase of the reaction (to OD 0.1) where substrate depletion is < 10%. Where it was possible to calculate the apparent Km and kcat from hyperbolic plots of plasmin generation vs. plasminogen concentration this was performed and values of apparent kcat/Km were calculated. Where there was no hyperbolic plot, the ratio of apparent kcat/Km was given by the slope of the plot of plasmin production vs. plasminogen concentration (i.e. the initial linear portion of the hyperbolic plot). An alternative approach was to fit progress curves directly of absorbance vs. time using the numerical integration software package, Dynafit [7,8] (BioKin Ltd, Pullman, WA, USA). This approach allows complex mechanisms to be studied using raw data directly and permits use of all available data from progress curves while taking into account changing concentrations and depletion of reactants. Dynafit optimizes kinetic parameters to a scheme provided in a script file describing a mechanism. The script used for PAI-1 inhibition kinetics experiments was:

E + S = ESka, kb (forward and back, Km = kb/ka)
ES-> E + Pkc
G + A = GAkp, kq
GA-> E + Akr
A + I-> AIkon

Where E = [plasmin], S = [S2251], P = [p-nitroaniline], G =[Plasminogen], A = [tPA], I = [PAI-1]. Initial estimates for rate constants are provided along with known concentrations of reactants and an extinction value for 1.0 mol L−1 P under the conditions of the experiment. For plasminogen activation studies, in the absence of PAI-1, I = 0. In the presence of fibrinogen or fibrin, the apparent Km for plasmin on S2251 was modified to take into account competition by fibrin(ogen) [5]. Apparent values for kcat and Km were determined with a range of [plasminogen] in the absence of PAI-1 and used as constants in the data analysis with PAI-1 (fixed [plasminogen] and range of [PAI-1]) so the apparent first order rate constant for inhibition of tPA (kon) could be fitted to the raw data. Some alternative equilibrium binding and slow turnover mechanisms for inhibition of tPA were also explored by elaborating the reaction step for A+I shown in the script above. The script used for work in the presence of fibrin contained additional terms to deal with dilution of chromogenic substrate S2251. In this system, S2251 is added to the preformed clot as 40 μL of 0.6 mmol L−1 and during the course of the experiment will diffuse through the clot to a final concentration of 0.2 mmol L−1. Reaction of tPA and PAI-1 at the surface of the clot will encounter a changing S2251 concentration which is accounted for in the model. Independent experiments had shown this happens with a rate constant of 5 × 10−5 s−1. Additional terms in the reaction script were also included to investigate the significance of instability of PAI-1 and potential reaction of PAI-1 and plasmin. Terms accounting for instability of PAI-1 were investigated in the model using rate constants, kl = 1.33 × 10−6 s−1 and 9.625 × 10−5 s−1 (corresponding to half lives of 145 and 2 h [6]) for stabilized mutant PAI-1 and native PAI-1, respectively. To assess the significance of PAI-1-plasmin binding, the reaction E+I->EI was added with apparent rate constants, ki, of 0.085 and 0.29 mol−1 L s−1, in the presence of fibrin(ogen) or buffer only, respectively, based on a rate constant for inhibition of plasmin by PAI-1 in a pure system of 6.6 × 105 mol−1 L s−1 [9].

Results

Plasminogen activation

Wt insect cell tPA was expressed predominantly in the single chain form, but there was some batch-to-batch variation in the proportion of two chain material. In order to compare sc- and tctPA activities, wt tPA preparations were treated with plasmin-Sepharose to convert them to 100% tctPA and the non-cleavable sctPA:R275E (hereafter referred to as sctPA) was prepared. With the simple amidolytic substrate S2288 there was a clear difference in activity between these two forms of tPA reflected in both kcat and Km (see summary in Table 1). tctPA was 3-fold more active than sctPA. We term this ratio of activities (calculated as kcat/Km for tctPA/sctPA) as the zymogenicity, (Z) and hence the Z value for S2288 is 3.0. Moving on to the natural substrate, plasminogen, in the absence of any promoter, the difference in relative activities was maintained and the Z was calculated to be 7.1. Individual Km and kcat values could not be determined as our range of [plasminogen] was <Km, but the slopes of plots of activation rate vs. plasminogen concentration can be used to calculate kcat/Km.

Table 1.   Kinetic parameters for the activity of sc- and tctPA against chromogenic substrate or plasminogen in the absence of presence of promoters
SubstratePromotersctPAtctPAZ
Km*kcatkcat/KmKmkcatkcat/Km
  1. *Apparent kinetic parameters are shown with units of μmol L−1 for Km, s−1 for kcat and μmol−1 L s−1 for kcat/Km. Z is a ratio of kcat/Km and is dimensionless. Zymogenicity of tissue plasminogen activator (tPA) calculated as the ratio of kcat/Km values for tctPA divided by sctPA. NA indicates the value is not determined as the maximum observed rate was << Vmax. §Mean values from four independent determinations. The Z values ranged from 0.6 to 1.4 but overall there was no significant difference between sc- and tctPA kinetics in the presence of fibrinogen. The three types of fibrin used are N, normal, using fibrinogen with no special treatment; C, CPB fibrin, for fibrin with CPB incorporated into the clots to continually remove C-terminal lysines; P, plasmin-treated, where fibrinogen was first treated with plasmin-Sepharose to produce fibrin with optimized levels of C-terminal lysines.

S2288None14607.30.00569410.70.0153.0
PlasminogenNoneNANA0.004NANA0.0277.1
PlasminogenFibrinogen§NANA0.16NANA0.140.9
PlasminogenFibrin N0.0870.0230.270.1090.0270.250.9
PlasminogenFibrin C0.3410.0240.070.3620.0260.071.0
PlasminogenFibrin P0.0860.0230.270.1190.0280.240.9

However, as tPA is regulated largely by template molecules such as fibrin and cellular receptors, we determined Z in the presence of these promoters. Fibrinogen was chosen first as there is known to be some stimulation by fibrinogen before conversion to fibrin. In order to identify the optimum concentration of fibrinogen, tPA activation of plasminogen (500 nmol L−1) was measured in the presence of a range of fibrinogen concentrations. Commercial fibrinogen from two suppliers was investigated with very similar results, as shown in Fig. 1. There was clearly stimulation of activity by fibrinogen but there is no saturation. Rather the profile obtained is typical of a template where there is an optimum concentration of fibrinogen needed to bind and bring together both enzyme and substrate. The optimum concentration of fibrinogen was around 1.0–1.5 mg mL−1, and this was not dependent on the source. With both fibrinogens, the displacement of the sc- and tctPA curves suggests a tighter interaction of sctPA with fibrinogen than tctPA. Plasminogen activation by sc- and tctPA was then investigated under these optimized conditions of 1.5 mg mL−1 fibrinogen, over a range of plasminogen concentrations in an attempt to determine Km and kcat values. As was the situation for plasminogen activation in the absence of fibrinogen, we could only calculate the ratio kcat/Km from slopes of activation rate vs. plasminogen concentration, or from direct fittings of plasminogen activation using Dynafit. The kinetic parameters obtained are summarized in Table 1, where we see the differences between sc- and tctPA disappear in the presence of fibrinogen over this range of plasminogen. The next stage was to study plasminogen activation by sc- and tctPA in the presence of fibrin. This was investigated using our previously published assay using clear fibrin and in the presence of a chromogenic substrate [5]. Hence, the tPA added to the clot will bind to fibrin and will be concentrated in a zone at the top of the clot rather than be distributed throughout the reaction mixture as in the case with fibrinogen. In the presence of fibrin, the Km values fall and it becomes possible to determine values for apparent Km, as shown in Table 1. A significant conclusion from the data shown in Table 1 is that the activities of sc- and tctPA are little different in the presence of fibrin and Z is again close to 1.

Figure 1.

 Stimulation of tissue plasminogen activator (tPA) activation of plasminogen by fibrinogen. Fibrinogen was from two commercial sources, Sigma (panel A) and Calbiochem (panel B). In both cases, activation rates were determined using sctPA:R275E (open symbols and solid lines) and tctPA (closed symbols and dashed lines). In all cases, the tPA concentration was 0.8 nmol L−1, the plasminogen concentration was 0.5 μmol L−1 and S2251 was 0.15 mmol L−1.

It is known that fibrin binding is a major regulator of tPA activity but also that fibrin itself is not a simple entity. C-terminal lysines are generated in fibrin and these have the ability to bind both tPA and plasminogen and stimulate activation rates. Thus fibrin may be thought of as being in two different conditions: (i) native and (ii) partially cleaved, before more complete digestion occurs. In very simple terms, it is believed that tPA binds initially to native fibrin via the finger domain and subsequently binding to C-terminal lysines in cleaved fibrin is regulated by the kringle 2 domain. In order to investigate whether these two forms of fibrin interact differently with sc- and tctPA, we used normal fibrin and fibrin formed in two ways: either using fibrinogen that has been pretreated with plasmin-Sepharose (stock suspension of around 500 nm active plasmin per mL of Sepharose and a final concentration with fibrinogen of approx 5 nmol L−1 for 20 min at 37 °C); or formed with normal fibrin but the clots have incorporated CPB to eliminate forming C-terminal lysines. These two forms of fibrin mimic the native form initially encountered by tPA and the cleaved form present during fibrinolysis. In order to optimize our preparation of these two forms of fibrin, Reteplase was used as a marker for the presence or absence of C-terminal lysines, as they are required for binding of Reteplase which has no finger domain and binds primarily via the kringle 2 domain. The data shown in Fig. 2 for Reteplase activity validate our fibrin preparations as being significantly different. However, the activity of sc- and tctPA are similar to each other in both fibrin forms and also between fibrin forms.

Figure 2.

 Michaelis–Menten plots of plasminogen activation in the presence of fibrin. Plasminogen activation over a range of plasminogen concentrations incorporated into fibrin clots was measure using sc-, and tctPA and the truncated variant Reteplase. Two types of fibrin were investigated, panel A has clots incorporating carboxypeptidase B to mimic the initial phase of fibrinolysis before generation of C-terminal lysines and panel B has clots made from fibrinogen pretreated with plasmin-Sepharose to model partially digested fibrin. Symbols are A, circles tissue plasminogen activator (tPA) (open sc-, closed tctPA), triangles Reteplase; and B, squares tPA (open sc-, closed tctPA), triangles Reteplase.

Inhibition by PAI-1

Another significant regulatory mechanism of tPA activity is the interaction with the specific serpin inhibitor, PAI-1. The kinetics of tPA (and uPA) interaction with PAI-1 have been studied in great detail; however, these data are for the reaction in a purified system in solution, in the absence of template promoters such as fibrinogen and fibrin that might well be expected to affect the interaction. Clearly the presence of these templates is much more relevant to the behaviour of PAI-1 and tPA in vivo. Furthermore, most studies have used tctPA and not examined any differences between sc- and tctPA in detail. Consequently, we attempted to assess the effectiveness of PAI-1 as an inhibitor of sc- and tctPA activation of plasminogen without a template (in buffer) and in the presence of fibrinogen and fibrin. Our assay system utilized the natural substrate of tPA, plasminogen, in order to get a better impression of the effectiveness of PAI-1 in a more physiologically relevant system. Such experiments are technically demanding but data analysis is made tractable by the use of numerical integration programs such as Dynafit. Initially a stabilized variant of PAI-1 was used to minimize complications arising from formation of latent, inactive PAI-1 during kinetic experiments. A range of parallel experiments were performed under identical conditions to compare the activation of plasminogen by sc- and tctPA in the presence of no template, fibrinogen and fibrin, with the inclusion of a range of PAI-1 concentrations. A number of variations of the simplest model (see Materials and methods) for the inhibition of tPA by PAI-1 were explored, but more extensive work would be needed to fully characterize detailed binding mechanisms in the presence of fibrin(ogen), which was not within the scope of the current work. For the purposes of the present study, a quasi-irreversible mechanism was assumed giving a simple second order rate constant, kon, derived by global fitting to a set of curves over an optimized range of PAI-1 for each assay system and each tPA for comparative purposes. Thus it would be possible to compare the effectiveness of PAI-1 inhibition between systems and also compare changes in the apparent second order rate constants for PAI-1 inhibition with the second order rate constants for plasminogen activation. An example set of data and fitting is shown in Fig. 3 for the inhibition of sctPA by PAI-1 in the presence of fibrinogen. A parallel set of data was collected using tctPA (not shown). From simple visual inspection, tctPA appeared to be less susceptible to inhibition by PAI-1 and this was confirmed by the fitting of the global second order rate constant kon. Values derived from global fitting were 6.8 (3.8%) and 2.14 (9.5%) μmol−1 L s−1 (% error of fitting) for sc- and tctPA, respectively. There was a general trend that tctPA displayed slightly poorer fitting, suggesting a different mechanism of binding between sc- and tctPA. Alternative models were investigated including equilibrium between A+I and slow turnover of inhibitor which showed some reduction in sum of squares for fitting, but these were not explored in detail and the global kon for inhibition is used here for comparative purposes.

Figure 3.

 Plasminogen activation by sctPA in the presence of fibrinogen and S2251 and a range of plasminogen activator inhibitor-1 (PAI-1) concentrations. Pairs of curves are shown for 0.8 nmol L−1 tissue plasminogen activator (tPA) with 200 nmol L−1 plasminogen, 0.15 mmol L−1 S2251 and 1.5 mg mL−1 fibrinogen in the presence of 0 (plus), 0.2 (circles), 0.4 (triangles), 0.6 (squares), 0.8 (crosses), 1.6 (diamonds), 2.4 (downward triangles) and 3.2 nmol L−1 (stars) PAI-1. Only every 5th data point is shown for clarity although fitting was performed using all data. The fitted data represent a global kon for PAI-1 with tPA of 6.82 μmol−1 L s−1.

Similar experiments were also performed in the same way in the absence of any template and also in the presence of fibrin and the global kon constants determined in the same way. A reduced effectiveness of PAI-1 was noted in the fibrin assay system and an increased range of inhibitor concentrations was required. This was reflected in the apparent global kon values which were reduced significantly to 0.37 and 0.45 μmol−1 L s−1 for sc- and tctPA, respectively (reduced eighteen and 5-fold). These studies are summarized in Fig. 4 which presents the apparent kinetic efficiency (kcat/Km) of sc- and tctPA as plasminogen activators and the global kon for inhibition of tPA in the presence of buffer, fibrinogen and fibrin. Values for kcat/Km are taken from Table 1. All units are μmol−1 L s−1 thus permitting comparison of the relative enzyme efficiency and susceptibility to inhibition. However, caution is needed when comparing the fibrin assay system with the others because of the heterogeneity of the fibrin system, but trends are obvious. Several points may be noted from Fig. 4 and Table 1. First of all, fibrinogen stimulates both the activity of sc- and tctPA against plasminogen and in reaction with PAI-1. In the presence of buffer and fibrinogen, reaction with the inhibitor is very much favoured over plasminogen activation. The situation changes markedly in the presence of fibrin because enzyme activity is stimulated and the effectiveness of PAI-1 is reduced. Rate constants for plasminogen and PAI-1 are similar and there is no difference between sc- and tctPA.

Figure 4.

 The effect of templates on the apparent second order rate constants for the activation of plasminogen and inhibition by plasminogen activator inhibitor-1 (PAI-1). The three sets of data are for Buffer (no template), the presence of 1.5 mg mL−1 fibrinogen and solid fibrin. The bars represent the apparent second order rate constants for the activation of plasminogen by sctPA (light shading) and tctPA (black) and the kon for the inactivation by PAI-1 of sctPA (white) and tctPA (grey).

All experiments following plasminogen activation by sc- and tctPA, in the presence of buffer only, fibrinogen and fibrin were repeated using native PAI-1 in place of the stabilized variant [6] used above. Results for both PAI-1 molecules were similar. In the presence of buffer, fibrinogen and fibrin the ratio of kon values for sc/tctPA were 4.2, 3.8 and 0.7, respectively for native PAI-1. The same ratios for the stabilized variant were 3.9, 3.2 and 0.8. Thus, looking at inhibition of plasminogen activation, sctPA is more rapidly inhibited than tctPA in buffer or in the presence of fibrinogen. In fibrin the rates are similar. Overall, the absolute kon values were consistently higher for native PAI-1. The improvement in kon values ranged from 1.5-fold in buffer to 2.8-fold in the presence of fibrin. An increase in kon values up to tenfold for native vs. stabilized PAI-1 was observed in comparative studies in a purified system with tPA or 1.8-fold when the target was uPA [6]. When a reaction for the formation of latent PAI-1 was added to the Dynafit script file using published half lives in free solution [6], kon values were found to be unchanged for the stabilized PAI-1 variant and increased only on the order of 5% for native PAI-1. Thus in our systems the numbers are only marginally affected by stability considerations and do not affect the findings for the relative behaviour of sc- and tctPA. Similarly, we investigated the significance of PAI-1 reacting with plasmin generated during the course of our assays as PAI-1 is a reasonable inhibitor of plasmin [9]. Overall, the effect of adding the E+I->EI reaction to the scheme shown in Materials and methods was small, resulting in a 10% or 20% reduction in the calculated kon value for the tPA-PAI-1 reaction in the presence of fibrinogen or fibrin, respectively, with no significant changes in the sc-/tctPA ratio. In the absence of fibrin(ogen), the formation of EI resulted in a small, up to 10%, reduction in kon for sctPA-PAI-1 binding but a more marked reduction of around 40% for the slower tctPA-PAI-1. Thus the ratio of kon values for sc-/tctPA is exaggerated, but plasminogen activation by sctPA is still more effectively inhibited than with tctPA.

As the relative reactivity of sc- and tctPA seems to be different from that observed in purified systems where kon for tctPA is generally higher, we investigated the kinetics of our sc- and tctPA preparations with native PAI-1 in a purified system monitoring tPA activity with the chromogenic substrate S2288 and using a modified Dynafit script without plasminogen activation. In agreement with many previous studies, we also observed a more rapid inhibition of tctPA, with kon values of 1.74 and 19.4 μmol−1 L s−1 for sc- and tctPA, respectively. These values are within the expected range of around 107 mol−1 L s−1.

Discussion

tPA is widely recognized as a unique enzyme in hemostasis that undergoes single-chain to two-chain conversion but this is not a zymogen activation reaction, a central regulatory feature of the coagulation cascade, for example [10]. Rather, tPA activity is regulated more subtly by binding to promoter species such as fibrin and also by inhibition by inhibitors, specifically PAI-1. However, a systematic investigation using precise kinetic methods of sc- and tctPA has not been performed to fully understand how these molecules are affected as activators of plasminogen and targets of PAI-1 by the presence of physiological promoters. Such a study is needed if we are to build a kinetic model of fibrinolysis in vivo and was undertaken here as a step towards constructing such a model. As in some previous studies, we have also utilized a site-directed variant of tPA (R275E) which is non-cleavable and compared this with native tctPA enzyme. In line with many previous reports, we observed that there is a small increase in amidolytic activity of <10-fold with the conversion of sc- to tctPA [11]. There is general consensus that the activities of sc- and tctPA are quite similar in the presence of fibrin (e.g. reviewed [12]), but these studies are complicated by different techniques and promoters. In the present study, we have used precise kinetic methods to determine enzyme efficiencies (kcat/Km) of plasminogen activation in fibrinogen and fibrin. Furthermore we have included fibrin before and after plasmin cleavage which is known to introduce new binding sites for both tPA and plasminogen and alter the mode of binding and the domains interacting with fibrin. Despite some early controversy about the zymogen nature of sctPA a consensus emerged that sctPA did have significant activity against plasminogen before conversion to the tc form and in the absence of any promoter molecules. However, the published ratio of activities of tc-/sctPA is variable and ranged from approximately equal [13] to small at around fifteenfold [14] to more significant at thirty- to one hundredfold [15]. We obtained a ratio of 7 which was the largest difference we found between the sc- and tc-enzyme forms in all the conditions we used.

Stimulation of tPA activity by fibrinogen is a slightly contentious issue and questioned in some quarters, but is also widely observed. Comparing levels of stimulation is not easy as a result of different conditions, but can vary from values such as 5-fold [16] to twenty-fivefold [17]. Poorer stimulation is associated with tctPA and the higher stimulation factors with sctPA, which is in line with what we observed (Table 1), where stimulation by fibrinogen is around 5-fold for tctPA and fifty-threefold for sctPA. The result is to bring plasminogen activation rates by sc- and tctPA close together. Similar activities have been reported elsewhere for sc- and tctPA as plasminogen activators but we have now extended these studies to include different forms of fibrin and to investigate PAI-1 inhibition. Investigating PAI-1 inhibition kinetics in the presence of fibrino(ogen) imposes more technical demands than studies performed in pure systems, as is usually the case. However, our methods have enabled us to investigate kinetics of inhibition and our studies have been helped by using numerical integration approaches to data analysis using Dynafit in this instance [7]. Although this approach is ideally suited to the study of reaction mechanisms (e.g. [8]), it was not our intention to investigate the reaction kinetics of PAI-1 and tPA in detail. Recent studies have proposed multi-step reversible mechanisms prior to a rate-limiting reversible step [18]. However, our analysis was limited to differentiating between inhibition by sc- and tc- forms of tPA in the presence of fibrinogen and fibrin and we expressed inhibition as a global rate constant, kon, simply describing the loss of enzyme activity. Further detailed studies would be needed to elaborate detailed binding mechanisms in fibrin(ogen). We did observe that tPA is much more sensitive to PAI-1 in fibrinogen than it is in fibrin, in other words fibrin does offer some protection for tPA. Because the fibrin system is of necessity a heterogeneous system rather than a pure solution, as in the case of fibrinogen, direct comparisons are not straightforward, but an impression of the protective effect of fibrin can be gained and suggests that the kon for the reaction is decreased around seven-to twenty-fivefold (depending on the tPA form) in going from fibrinogen to fibrin. Using a semi-quantitative approach, Carr et al. [19] suggested a PAI-1 was of the order of 5-fold more effective in plasma than in the presence of fibrin, a trend in agreement with our findings. The reaction in the presence of fibrinogen is rapid and in line with rates reported in free solution at around 8–10 μmol−1 L s−1 [20,21]. However, most published work uses tctPA to investigate PAI-1 kinetics and there is little work on comparing sc and tctPA in the same systems.

The regulation of tPA activity has long been known to be under much more subtle control than other enzymes in hemostasis, for example, the zymogen activation steps that lead to a rapid generation and amplification of enzyme activity seen in the coagulation cascade. Binding of tPA and substrate plasminogen to promoter templates such as fibrin plays a major role in stimulating and regulating enzyme activity in fibrinolysis. More subtle intramolecular interactions between tPA binding domains and the serine protease domain have also been identified by calorimetry between the finger and/or growth factor domain and the protease domain [22]; and by NMR observations of binding between the kringle 2 and protease domains [23]. Hence fibrin binding by both finger and kringle domains may affect the protease domain and these interdomain interactions may be different in sc- and tctPA [23]. Because the binding to native and cleaved fibrin is known to primarily involve finger and kringle 2 domains, respectively [24], we considered it important to investigate the kinetics of plasminogen activation in fibrin that mimics native fibrin (containing CPB to remove C-terminal lysines) and in fibrin containing C-terminal lysines (pretreated with plasmin). Furthermore, complex exocite interactions have been identified between tPA and PAI-1 especially in the region of residues 296–299 [20,25–27] and these may be involved in subtle regulation of the interaction kinetics of tPA and its specific inhibitor. Our approaches enabled us to investigate PAI-1 inhibition kinetics in the presence of fibrin(ogen) and plasminogen, all of which may modulate kinetics. Little detailed work is available on inhibition kinetics in fibrinogen or fibrin and we now report that there is a difference in sensitivity of sc- and tctPA to PAI-1 in fibrinogen but not in fibrin, which may well be attributed to these additional interactions, not present in most other serpin-protease interactions. Another possible explanation for the action of fibrin in reducing the effectiveness of PAI-1 is that fibrin binds PAI-1 and may remove it from the pool of PAI-1 able to react with tPA (e.g. see [28]).

We can conclude that tPA has evolved to respond in a different way to fulfill its particular role in fibrinolysis compared with enzymes in the coagulation cascade. Physiologically, fibrinolysis is generally needed to clear fibrin from the sites of repair, whereas coagulation is initiated during acute situations to prevent potentially fatal blood loss. Thus, coagulation requires an explosion of activity so thrombin is generated at the end of an amplifying cascade of zymogen activation reactions. Initially, coagulation is also enhanced by the release of PAI-1 from activated platelets to inhibit fibrin destruction. Fibrinolysis on the other hand has more of a housekeeping function and regulation involves controlled secretion of tPA and PAI-1 into the circulation, modulation of activity by binding and assembly of enzyme and substrate on templates, and inhibition by PAI-1 under different circumstances. As PAI-1 effectiveness declines (maybe diffusion from the clot, or binding and isolation on fibrin or depletion by forming complexes with tPA from circulation or formation of latent PAI-1), fibrinolysis can assume a larger role. Ironically, fibrinolysis initiated during thrombolytic therapy requires an explosion of activity to clear potentially fatal arterial clots and this is a role that tPA may not have evolved to perform optimally. Understanding the detailed regulation of tPA activity using kinetic models that include activation and inhibition in the presence of fibrin and fibrinogen will help to improve thrombolytic therapy by aiding the design of better drugs and treatment regimes [29].

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.

Ancillary