SEARCH

SEARCH BY CITATION

Keywords:

  • anticoagulant;
  • factor Va;
  • factor VII;
  • factor Xa;
  • thrombin generation;
  • tissue factor

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Background

We observed that minute amounts of thrombin or the enzyme Russell's viper venom activating factor V (RVV-V) added to plasma strongly diminish the potential of that plasma to generate thrombin after being triggered by tissue factor.

Objective

To find the mechanism behind this phenomenon.

Methods and Results

Thrombin generation (TG) initiated by tissue factor (TF) is strongly and dose-dependently inhibited by addition of activated factor V (FVa) or by addition of a factor V activator (thrombin or RVV-V). No inhibition is seen when TG is triggered via the intrinsic pathway or by direct activation of factor X. The effect is independent of proteins C and S and tissue factor pathway inhibitor (TFPI). In factor VII-deficient plasma the effect is seen when it is spiked with recombinant factor VII (FVII) and to a much lesser extent when spiked with recombinant FVIIa. In a purified system, FVa also dose-dependently inhibits the activation of FX by TF/FVII(a). The inhibitory effect is neutralized by antibodies against the light chain of FVa but not by antibodies against the heavy chain.

Conclusions

Our observations can be explained by assuming that FVa, via its light chain, binds to the complex TF/FVII(a) and prevents it from activating FX. We assume that this mechanism reduces the possibility that thrombin and factor Xa escaping from a wound area into the circulation, together with blood-borne tissue factor, would trigger intravascular coagulation.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

The final common step in the different pathways that lead to thrombin formation is the activation of prothrombin by prothrombinase, a complex of activated factors V (FVa) and X (FXa) adsorbed onto a suitable phospholipid surface [1]. FVa increases the catalytic activity of FXa over a thousand-fold [2, 3]. For all practical purposes bulk thrombin formation is switched on by the activation of FV [4, 5] and switched off when activated protein C destroys activated factor V [6]. Indeed, lack of FV causes hemorrhagic syndrome [7] whereas its resistance to inactivation causes a prothrombotic tendency [8].

Human FV is a glycoprotein that circulates in plasma at a concentration of 20 nm, as a single 359 kD chain of three homologous A domains, two homologous C domains and a unique B domain (A1-A2-B-A3-C1-C2), [9-11]. Factor V is believed to have minor procoagulant activity [12] and is converted into FVa after proteolysis. Upon activation by thrombin, the B domain is removed and a dimeric structure of A1-A2 and A3-C1-C2 remains (168 kD) [4, 5, 11, 13].

It has previously been demonstrated that FV stimulates the anticoagulant activity of activated protein C (APC) by acting as a cofactor for APC inactivation of FVIIIa in the presence of protein S [14-16]. It has also been reported that congenital deficiency of FV is associated with decreased plasma levels of tissue factor pathway inhibitor (TFPI), suggesting a complex formation between TFPI and FV [17]. A recent report by Ndonwi et al. [18] has demonstrated that the C-terminus of the full-length TFPIα is required for its binding to both FV and FVa.

In this article we demonstrate an as yet unknown fourth regulatory function of FVa: when it is formed in plasma it has a strong inhibitory action on TF-induced thrombin generation (TG) because it inhibits the activation of FX by TF/FVII(a) independently of the presence of TFPI.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Materials

Phospholipids (PL) consisted of 20 mole% phosphatidylethanolamine, 20 mole% phosphatidylserine and 60 mole% phosphatidylcholine (Avanti, Alabaster, AL, USA). Recombinant human tissue factor (rTF) was Innovin (Dade-Behring, Marburg, Germany). The fluorogenic substrate Z-Gly-Gly-Arg-aminomethylcoumarine (ZGGR-AMC) was purchased from Bachem (Basel, Switzerland). The chromogenic substrate S2238 was supplied by Chromogenix (Mölndal, Sweden). Thrombin calibrator (α2macroglobulin-thrombin) was prepared according to Hemker et al. [19]. Hepes buffers containing 5 mg mL−1 or 60 mg mL−1 bovine serum albumin (BSA5 and 60, respectively) were prepared as described previously [19]. Bovine FIXa was purified according to van Dieijen et al. [20], bovine FX(a) was prepared as reported in ref [21] and human FX(a) as previously documented [22]. Bovine α-thrombin was prepared and isolated as described previously [23]. Russell's viper venom enzymes activating FX and FV were a kind gift of Dr JW Govers-Riemslag (Department of Biochemistry, Maastricht, the Netherlands). Recombinant human TFPI was obtained from Sino Biological Inc. (Beijing, China). Human protein C-deficient plasma, human TFPI-depleted plasma, murine monoclonal antibodies against human TFPI (domains Kunitz-1 and Kunitz-2) and human FVII were obtained from American Diagnostica Inc. (Stamford, CT, USA). Mouse monoclonal antibodies against human TFPI (domain Kunitz-3 and C-terminus) were from Sanquin (Amsterdam, the Netherlands). Human FVII-deficient plasma was from Kordia (Leiden, the Netherlands). Recombinant human FVIIa was NovoSeven (Novo-Nordisk Inc., Bagsvaerd, Denmark). Human protein S-deficient plasma was obtained from Affinity Biological Inc. (Ancaster, Ontario, Canada). Human FIX- and FVIII-deficient plasmas were purchased from Dade-Behring GmbH (Marburg, Germany). As a thrombomodulin (TM) preparation we used recombinant human TM, a kind gift of Asahi Kasei Pharma (Japan). Human activated protein C (APC) was kindly provided by Dr V. Regnault (France). PPACK (D-phenylalanyl-L-prolyl-L-arginine chloromethylketone) was supplied by Calbiochem (San Diego, CA, USA).

Mouse monoclonal antibodies against FV/Va were purchased from Hematologic Technologies Inc. (Essex, VT, USA): ABV-5106 against the heavy chain region of bovine FV/Va, ABV-5107 against the light chain region of bovine FV/FVa and AHV-5108 against the light chain region of human FVa/FV (i.e. reactive to the epitope on the carboxyl-terminal segment, Mr = 74.000 fragment E).

Blood and plasma

Blood was acquired from 24 apparently healthy and consenting volunteers through antecubital venipuncture. Blood was collected into BD vacutainer tubes (1 volume trisodium citrate 0.105 m to 9 volumes blood) in the absence of a trypsin inhibitor (BD Vacutainer System, Roborough, Plymouth, UK). Normal pooled plasma (NPP) was prepared by centrifuging the blood at 2900 g during 10 min at room temperature. Plasma was aspirated and the procedure was repeated. Plasmas were pooled and further ultra-centrifugation was carried out. Aliquots of 1 mL were stored at −80 °C until use.

The calibrated automated thrombogram

The calibrated automated thrombogram (CAT) was measured [19] in a 96-well plate fluorometer (Ascent reader, Thermolabsystems OY, Helsinki, Finland) equipped with a 390/460 filter set (excitation/emission) and a dispenser. Immulon 2HB, round-bottom 96-well plates (Dynex) were used. Each experiment needs two sets of readings, one from a well in which thrombin generation takes place (TG well) and a second one from a well to which the calibrator has been added (CL well). All experiments were carried out at least three times. A dedicated software program (Thrombinoscope BV, Maastricht, the Netherlands) was used for the conversion of the fluorescent signal into the corresponding TG curve.

Thrombin generation was measured in a mixture of 80 μL plasma and 20 μL buffer, containing PL and a trigger as indicated in the individual experiments but without Ca2+. PL consisted of phosphatidylserine/phosphatidylcholine/phosphatidyl-ethanolamine vesicles (20, 60 and 20 mole%, respectively) in Hepes-buffered saline and were added at a final concentration of 4 μm. In the calibration experiments 20 μL of α2-macroglobulin-thrombin was added to 80 μL of plasma.

The plate was placed in the fluorometer and allowed to warm to 37 °C (minimally 10 min). The dispenser of the fluorometer was flushed with warm 100 mm CaCl2 solution, emptied, and then filled with warm FluCa (416 μm fluorogenic substrate ZGGR-AMC and 16.6 mm Ca2+). At the start of the experiment, the instrument dispensed 20 μL of FluCa to all the filled wells, registered this as zero time, and started reading after shaking for 10 s. During the measurement, the program compared the readings from the TG and the CL wells, calculated thrombin concentration and displayed the thrombin concentration in time.

Purification and activation of FV

Factor V was isolated from bovine or human plasma with a slight modification of the procedure of Lindhout et al. [24]. Citrated plasma was treated with BaSO4 in order to remove vitamin K-dependent clotting factors. The plasma supernatant was stirred with 30 g Swollen QAE-Sephadex in 100 mm NaCl, 20 mm Tris–HCl (pH 7.5). After washing, FV was eluted using a salt-gradient. From the FV containing fractions the fraction precipitating between 30% and 60% saturated ammonium sulfate was collected and dissolved in 100 mm NaCl, 20 mm MES, 200 nm PPACK (pH 6.5). Ten milliliters of this fraction was applied to a Sephacryl S300-HR-column (20 cm² × 96 cm), which was equilibrated with MES-buffer without PPACK. At a rate of 0.7 mL min−1, FV eluted in a sharp peak before most contaminating proteins. As a last step FV was purified on a Mono-S column using 20 mm Hepes, 5 mm CaCl2, 100–800 mm NH4Cl gradient (pH 7.4).

The final purity of the protein was ascertained by SDS-PAGE to be > 90%.

Factor Va was prepared by activation of FV with thrombin; 50 nm FIIa was added to FV in 100 mm NaCl, 40 mm Hepes, 5 mm CaCl2 (pH 7.5). After 30 min incubation at 37 °C, 100 nm PPACK was added. The solution was applied to the Mono-S column (1 mL) and a 30 mL gradient of 50–1000 mm NH4Cl in 20 mm Hepes, 5 mm CaCl2 (pH 7.4) was applied. Two FVa (Va1 and Va2) bands eluted from the column as reported previously [25].

Determination of FVa activity

The activity of FVa was tested in a two-stage procedure. The FVa preparation was added to a mixture containing FXa (1 nm), PL vesicles (20 μm) and CaCl2 (5 mm). At zero time prothrombin was added to a final concentration of 2 μm. After 2, 4 and 6 min the reaction was stopped with 10 mm EDTA in buffer and the thrombin formed was tested with a chromogenic substrate (S2238, 100 μm final). All reactions were conducted at 37 °C.

Gel electrophoresis

The indicated amounts of proteins were mixed 1:1 with 10 μL of Laemmli buffer (Bio-Rad Laboratories, Veenendal, the Netherlands). Samples were incubated for 5 min at 97 °C before being separated on a 4–15% precast SDS gel (Bio-Rad). A protein marker (Bio-Rad) with known molecular weight was run with the samples in the same gel. The run time was approximately 45 min at 200 volts. Proteins in the gel were stained with Coomassie Brilliant Blue G-250 (Bio-Rad) for 60 min while shaking gently. Gels were rinsed twice for 30 min with distilled water and analyzed with the Gel Doc EZ gel Imager (Bio-Rad) R.TF.

Western blotting

Gel electrophoresis was performed as described above. Immediately after the SDS page, proteins were transferred to a nitrocellulose membrane using the Trans-Blot Turbo system, according to the recommendations of the manufacturer (Bio-Rad). The membrane was washed twice with TBST (140 mm NaCl, 50 mm Tris, 0.05% Tween, pH 7.4) and blocked with TBST containing 3% BSA (Sigma Aldrich, Zwijndrecht, the Netherlands) for 1 h at RT. Membranes were washed four times with TBST for 5 min and incubated with a Kunitz-2 domain-specific or C-terminus domain-specific monoclonal anti-TFPI antibody (2 μg mL−1 in blocking buffer) for 1 h at RT. After four washes of 5 min each, the membranes were incubated with a 1/500 dilution of rat anti-mouse IgG-HRP (Invitrogen, Bleiswijk, the Netherlands) in blocking buffer for 1 h at RT. After four washes with TBST, immunoreactive bands were detected using the DAB substrate kit (Thermo Scientific, Landsmeer, the Netherlands) according to the recommendations of the manufacturer. After 10 min, membranes were washed twice with distilled water, dried and photographs were taken.

Measurement of FXa generation

Factor Xa generation was measured in a reconstituted system using a two-step cascade method analogous to the determination of FVIIIa developed earlier [26]. The reaction mixture consisted of 50 nm human or bovine FX, 5 mm Ca2+ and 400 μm of the fluorogenic substrate ZGGR-AMC plus or minus non-membrane-bound FVa as indicated. The reaction was started with 20 nm human or bovine FVII pre-incubated with 5 pm rTF. No procoagulant PL vesicles were added. The course of fluorescence was kinetically measured at 37 °C for 40 min in a 96-well plate fluorometer (Spectra Max, Molecular Devices, Sunnyvale, CA, USA) at wave length of 390/460 (excitation/emission).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Factor Va has an inhibitory effect on TG

We observed that amounts of thrombin, small enough so as not to make the plasma clot (1–2 nm), when added to citrated plasma, strongly diminish the capacity of that plasma to form thrombin after recalcification in the presence of tissue factor (Fig. 1A). Because thrombin in plasma has a half-life of around 1 min and the phenomenon persisted unaltered for longer than 10 min, a product of the enzymatic action of thrombin was the likely cause. Further experiments showed that this product probably was activated FV. In the first place the addition of activated FV (both human and bovine) also caused inhibition whereas FV that was not activated did not (Fig. 1B). The inhibitory effect of either of these measures remained unchanged for at least 2 h (Fig. S1).

image

Figure 1. Inhibition of thrombin formation after pre-incubation with RVV-V, traces of α-thrombin or factor (F) Va. Thrombin generation was measured at 5 pm rTF and 4 μm phospholipids (PL) in normal plasma. Frame A: plasma was pre-incubated with saline or RVV-FV 1:400 dilution or with 2 nm α-thrombin. Frame B: plasma was spiked with saline or 5 nm FV or increasing amounts of activated FV (0–40 nm). Representative curves are shown (n = 3).

Download figure to PowerPoint

In the second place the effect of thrombin could be mimicked by the FV activating enzyme from Russell's viper venom (RVV-V) pre-incubated with normal plasma for 15 min (Fig. 1A). In the presence of very high TF concentrations (≥0.2 nm), some residual TG (< 10% of normal) can be provoked in FV-deficient plasma [17]. Pre-incubation with RVV-V of the FV-deficient plasma at 0.2 nm TF brought about marginal changes in the residual activity that could not be interpreted as an alteration of TG (data not shown).

The two forms of activated FV (FVa1 and FVa2) diminished TG similarly while FV1 and FV2 showed no effect (Fig. S2).

Factor Va inhibits only TF-triggered TG

The inhibition is only seen when TG is triggered with rTF (≥ 3 pm) and not when it is started by contact activation, by isolated FIXa or by the RVV fraction that activates FX (RVV-X) (Fig. 2, frames A, B and C, respectively). The effect is present in FVIII and FIX-deficient plasmas (Fig. S3, frames A and B, respectively). This makes it unlikely that the mechanism of inhibition involves the intrinsic pathway (factors XII, XI, IX and VIII, high-molecular-weight kininogen or prekalikrein) or prothrombin conversion itself.

image

Figure 2. Trigger-dependency of the inhibitory effect of factor (F) Va. Thrombin generation was measured in normal plasma with 4 μm phospholipids (PL). Black lines: no FVa added. Red lines: 2 nm FVa added. Further additions: A, Kaolin 0.5 mg mL−1 (pre-incubated); B, 2 nm FIXa; C, RVV-X (1:1000 dilution). Representative curves are shown (n = 3).

Download figure to PowerPoint

The inhibitory effect of FVa is independent of APC

The inhibitory effect is also observed in plasma with a congenital deficiency of proteins C or S (Fig. 3, frame A and B, respectively). Also when normal plasma is spiked with TM or APC, the inhibiting action of FVa persists and additive to the inhibition caused by these agents (Fig. S4).

image

Figure 3. The inhibitory effect of factor (F) Va is independent of the activated protein C (APC) system. Thrombin generation was measured at 5 pm rTF and 4 μm phospholipids (PL) in plasma deficient in protein C (frame A) or in protein S (frame B). Black lines: absence of FVa. Red lines: 2 nm FVa. Representative curves are shown (n = 3).

Download figure to PowerPoint

The independence of the inhibitory phenomenon from both FVIII and the APC-system indicates that it is not associated with the cofactor role of FV in the APC-mediated inactivation of FVIIIa [14-16].

FVa effect is independent of TFPI

The inter-activation of FVII and X is known to be inhibited by a bimolecular complex consisting of TFPI and FXa [27], an effect very similar to the inhibition described here. We counted with the possibility of complex formation between TFPI and FV(a) as suggested by Duckers et al. [17] and of subsequent co-purification (i.e. of contamination of our FVa preparation with TFPI). However, the inhibitory effect of FVa on TG persisted in the presence of different anti-TFPI antibodies (against the domains Kunitz-1, Kunitz-2, Kunitz 3 and C-terminus) as well as in TFPI-depleted plasma (Fig. 4, frames A and B, respectively).

image

Figure 4. The inhibitory action of factor (F) Va is independent of tissue factor pathway inhibitor (TFPI). Thrombin generation was measured at 5 pm rTF and 4 μm phospholipids (PL) in normal plasma pre-incubated with excess (550 nm) of anti-human TFPI antibodies (against domains K1, K2 and K3 and the C-terminus) (frame A) or in TFPI-depleted plasma (frame B) in the absence (black lines) and presence (red lines) of 2 nm FVa. Representative curves are shown (n = 3). Furthermore, gel electrophoretic analysis and Western blotting were applied in order to assure that TFPI was not co-purified with FV(a) preparations that were used in all assays. Full-length TFPIα was used here as a control sample. Frame C: gel electrophoretic analysis of the purified FV. The indicated amounts of FV (lane 1–2) or TFPI (lane 4–5) were mixed with Laemmli sample buffer, loaded on a 4–15% precast gel and subjected to SDS-page. Lane 3 contains molecular weight markers. Proteins on the gel were stained with Coomassie Brilliant Blue G-250. Frames D and E: Western blot analysis of FVa (lane 1; 2 μg), FV (lane 2; 1.5 μg) and TFPI (lane 4; 0.8 μg) preparations. As in C, samples were mixed with Laemmli sample buffer and loaded on a 4–15% precast gel. Immediately after the SDS page, proteins were transferred to a nitrocellulose membrane using a Trans-Blot Turbo system. Upon washing and blocking the membrane, TFPI was detected using a K2 domain-specific (frame D) or C-terminus domain-specific monoclonal anti-TFPI antibody (frame E), as described in Materials and Methods.

Download figure to PowerPoint

A recent report by Ndonwi et al. [18] has shown that the C-terminus of the full-length TFPIα is responsible for its interaction with FV(a). Therefore, gel electrophoresis and western blotting were applied on our FV(a) preparations and compared with a standard TFPI preparation using antibodies against Kunitz-2 (responsible for the anti-FXa activity) or against the C-terminus [responsible for the interaction with FV(a)]. Our data clearly showed that our FV and its activated form were not associated (or co-purified) with TFPI (Fig. 4C,D and D).

Together these results suggest that the inhibition is located at the level of the mutual activation of FX and VII in the presence of tissue factor.

The most direct demonstration that FVa interferes with this mutual activation is the observation that the full inhibitory effect (> 70% of the peak height) remains in FVII-deficient plasma that is spiked with recombinant FVII (Fig. 5 frame A) whereas when it is spiked with recombinant activated FVII (Fig. 5 frame B) the inhibition is much less (< 20%).

image

Figure 5. Effect of factor (F) Va on thrombin generation in FVII-deficient plasma. Thrombin generation was measured at 5 pm rTF and 4 μm phospholipids (PL) in the absence (black lines) or presence (red lines) of 2 nm FVa in (A) FVII-deficient plasma (inset) or FVII-deficient plasma spiked with 2 nm of human FVII and (B) FVII-deficient plasma spiked with 2 nm of recombinant FVIIa. Representative curves are shown (n = 3).

Download figure to PowerPoint

Factor Va inhibits the activation of FX

We measured the influence of non-membrane-bound FVa on the generation of FXa directly. Figure 6 shows that FX activation by FVII(a)/TF is strongly and dose-dependently inhibited in the presence of FVa. This inhibitory effect is seen at different FX concentrations both with FVII and its activated form (Fig. S5).

image

Figure 6. Dependency of factor (F) Xa generation on FVa activity. Bovine FX (5 nm) was activated with the complex FVII (20 nm)/rTF (5 pm) in the presence of 5 mm Ca2+ at increasing concentrations of FVa (0–40 nm). The activity of FXa generated was assessed by the substrate ZGGR-AMC (400 μm). Representative curves are shown (n = 3).

Download figure to PowerPoint

To rule out any possible contamination of FV(a) with TFPI we added an excess amount of antibodies (550 nm) against human TFPI (i.e. against Kunitz 1, 2, 3 and the C-terminus) and found that the inhibitory effect of FVa persisted (Fig. S6).

In the reconstituted system the inhibitory effect of the same dose of FVa is higher than in plasma (Figs 1B and 6). This could be due to the higher concentration of FX in plasma (approximately 120 nm) than in the reconstituted system experiment (5 nm). Indeed the addition of 2 nm FVa to FX-deficient plasma, which was supplemented with 12 nm human FX, resulted in a much stronger inhibitory effect than was observed in normal plasma (Fig. S7).

No enzymatic activity of the complex FVII(a)/TF or FVIIa alone was observed on the substrate ZGGR-AMC (data not shown), and the enzymatic activity of FXa (at 5 or 50 nm) on ZGGR-AMC was not boosted by the presence of FVa (data not shown).

Taken together, these results indicate that the activation of FX by TF/FVII(a) is inhibited, most likely because FVa binds to the TF-FVII(a) complex and prevents FX activation.

The light chain region of FVa is accountable for the inhibitory effect

In order to determine in which part of the FVa molecule the inhibitory capacity is located we determined the effect of antibodies (Abs) against the heavy- (A1-A2) or the light-chain (A3-C1-C2) of both the human and the bovine protein. Figure 7(A) shows that FVa treated with Abs against the heavy chain inhibited TG as effectively as untreated FVa, whereas FVa treated with Abs against the light chain was hardly effective. This indicates that the inhibitory effect of FVa is localized in the light region of FVa (i.e. domains A3-C1-C2).

image

Figure 7. Effect of anti-FVa antibodies on the thrombin and factor (F) Xa generation. Frame A: thrombin generation was measured in normal plasma at 5 pm rTF and 4 μm phospholipids (PL) in the absence (black line) and presence of 2 nm bovine FVa that was treated with saline (red line) or with antibodies against the heavy (green line) or light (blue line) chain region of bovine FVa. Frame B: bovine FX (50 nm) was activated with the complex FVII (20 nm)/rTF (5 pm) and 5 mm Ca2+ in the absence and presence of 2 nm FVa that was already treated with saline, with antibodies against the heavy or the light chain region of FVa. The activity of FXa was kinetically measured using 400 μm ZGGR-AMC (the same legend as frame A). Representative curves are shown (n = 3).

Download figure to PowerPoint

We also studied the effect of these antibodies on FXa generation. As in TG, FVa treated with Abs against the heavy chain maintained its inhibitory effect whereas FVa treated with Abs against the light chain lost its inhibitory capacity (Fig. 7B). Similar to the data seen in plasma, light chain antibodies did not affect the activation of FX by TF/FVIIa in the absence of FVa (data not shown).

These data strongly suggest that the light chain region of FVa is responsible for the inhibitory effect seen on FIIa and FXa generation.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

We report in this study that activated FV strongly inhibits tissue factor-induced TG, either when added to plasma or when formed in situ by thrombin or RVV-V. The intrinsic and common pathways are not inhibited. The effect is not mediated through the APC system because it persists in plasmas lacking protein C or S. The action of FVa on TG is also present in hemophilia A plasma, which indicates that the phenomenon cannot be associated with the cofactor role of FV in the APC-mediated inactivation of FVIIIa [14-16]. Neither is it dependent upon the activation of FIX by TF/VIIa because it also is present in hemophilia B plasma.

These results indicate that the inhibitory action must be localized at the point of interaction of tissue factor with the factors VII(a) and X.

The proenzyme FVII, when bound to TF, obtains a certain proteolytic activity and activates FX [27]. Once activated, FXa converts the proenzyme into the more active FVIIa [28]. This mutual activation of FVII and FX is known to be inhibited by a bimolecular complex consisting of TFPI and FXa [29]. Our results resemble this effect but a role of TFPI appears unlikely because neither the addition of different antibodies against TFPI (i.e. against the domains; Kunitz-1, Kunitz-2, Kunitz-3 and C-terminus) nor depletion of TFPI from plasma influenced the effect. We counted with the possibility of complex formation between TFPI and FV(a) as suggested by Duckers et al. [17] and of subsequent co-purification (i.e. of contamination of our FVa preparations with TFPI), although it would be strange that such TFPI would only show activity after activation of FV. Prolonged pre-incubation of our FV(a) preparation with different antibodies against the Kunitz-1, Kunitz-2 and Kunitz-3 domains and the C-terminus of TFPI did not influence the inhibitory action of FVa. Finally, the TFPI independency was confirmed with Western blotting as TFPI was absent in our FV(a) preparations.

Very suggestive of an inhibitory effect of FVa on the reciprocal activation of FX and FVII is the observation that FVa does inhibit TG in FVII-deficient plasma that is reconstituted with recombinant FVII and to a lesser extent after reconstitution with FVIIa.

Our conclusions are corroborated by observations on the generation of FXa by TF-FVII(a) in buffer. A dose-dependent inhibitory action of FVa could be demonstrated that was not influenced by antibodies against TFPI.

Antibodies against the light chain of FVa, both in plasma and in the purified system, abolished its inhibitory capacity, whereas antibodies against the heavy chain did not.

This clearly demonstrates that the inhibitory effect of FVa on the coagulation and specifically on the activation of FX is mediated through the light chain region of FVa (i.e. the sequence A3-C1-C2). The C1 and C2 domains of FVa contribute both to phospho-lipids binding and subsequently the ability of FVa to bind FXa with high affinity [30].

Normally FV is activated by meizothrombin, an intermediate of prothrombin activation that has thrombin activity but remains membrane bound [31]. When membrane-bound FVa binds to membrane-bound FXa it forms the prothrombin-converting complex. FVa and FXa can also interact when no phospholipid surface is present and then form a complex that has little prothrombin-converting activity [24]. Therefore, when FVa is not formed on a surface or is added (or originates) in free solution, as in the present experiments, it seems that it can interact with the TF-FVII(a)-FX(a) complex and interfere with the mutual activation of these factors.

This anticoagulant function of FVa may reduce the possibility that thrombin and FXa escaping from a wound area into the circulation, together with blood-borne tissue factor, would trigger intravascular coagulation [32].

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

The authors are grateful for the learned discussion with the anonymous colleagues that have reviewed this manuscript, which led to significant improvements.

Disclosure of Conflict of Interests

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Authors are employees of Synapse BV and have no conflict of interests.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
jth12126-sup-0001-Legends.docWord document23KFigure S1. Effect of pre-incubation of FVa in plasma.
jth12126-sup-0002-Fig1.tifimage/tif2028KFigure S2. Effect of FV(a)1 and FV(a)2 on thrombin generation.
jth12126-sup-0003-Fig2.tifimage/tif2028KFigure S3. The inhibitory effect of FVa is independent of the clotting factors VIII and IX.
jth12126-sup-0004-Fig3.tifimage/tif2028KFigure S4. FVa inhibitory effect is maintained in the presence of soluble TM or APC.
jth12126-sup-0005-Fig4.tifimage/tif2028KFigure S5. Effect of FVa on FX activation by FVII or FVIIa at varying FX concentrations.
jth12126-sup-0006-Fig5.tifimage/tif2028KFigure S6. Inhibition of FX activation by FVa is independent of TFPI.
jth12126-sup-0007-Fig6.tifimage/tif2028KFigure S7. Effect of FVa on the thrombin generation measured in FX-deficient plasma.
jth12126-sup-0008-Fig7.tifimage/tif2028K 

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.