SEARCH

SEARCH BY CITATION

Keywords:

  • anticoagulant;
  • phospholipids;
  • protein S;
  • TF-FVIIa;
  • TFPI

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

Background

Tissue factor pathway inhibitor (TFPI) is a multi-Kunitz domain protease inhibitor that down-regulates the extrinsic coagulation pathway by inhibiting FXa and FVIIa.

Objectives

To investigate the role of the three Kunitz domains (KDs) of TFPI in FVIIa inhibition using full-length TFPI (TFPIfl) and truncated TFPI constructs.

Methods

Inhibition of FVIIa with/without relipidated tissue factor (TF) or soluble TF (sTF) by TFPIfl/TFPI constructs was quantified with a FVIIa-specific chromogenic substrate.

Results and Conclusions

TFPIfl inhibited TF-FVIIa via a monophasic reaction, which is rather slow at low TFPIfl concentrations (t½ ≈ 5 min at 2 nm TFPI) and has a Ki of 4.6 nm. In the presence of sTF and without TF, TFPIfl was a poor FVIIa inhibitor, with Ki values of 122 nm and 1118 nm, respectively. This indicates that phospholipids and TF significantly contribute to FVIIa inhibition by TFPIfl. TFPI constructs without the KD3-c-terminus (TFPI1–150 and KD1-KD2) were 7–10-fold less effective than TFPIfl in inhibiting TF-FVIIa and sTF-FVIIa, indicating that the KD3-C-terminus significantly contributes to direct inhibition of FVIIa by TFPI. Compared with KD1-KD2, KD1 was a poor TF-FVIIa inhibitor (Ki =434 nm), which shows that the KD2 domain of TFPI also contributes to FVIIa inhibition. Protein S stimulated TF-FVIIa inhibition by TFPIfl (Ki =0.7 nm). In the presence of FXa, a tight quaternary TF-FVIIa-TFPI-FXa complex is formed with TFPIfl, TFPI1–150 and KD1-KD2, with Ki values of < 0.15 nm, 0.5 nm and 0.8 nm, respectively, indicating the KD3-C-terminus is not a prerequisite for quaternary complex formation. Phospholipids and the Gla-domain of FXa are required for quaternary complex formation.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

After tissue damage, initiation of blood coagulation takes place upon the exposure of tissue factor (TF) [1]. TF acts as co-factor of factor (F) VIIa in the activation of the FX [2-4], a reaction that is greatly stimulated by the negatively charged phospholipids contributed by the damaged tissue. [5, 6].

Tissue factor pathway inhibitor (TFPI) is a major inhibitor of the extrinsic coagulation pathway, which inhibits both FXa and FVIIa [7]. TFPI is a multivalent Kunitz type inhibitor that consists of an acidic N-terminus followed by three tandem Kunitz domains (KDs) and a basic C-terminus [8]. Site-directed mutagenesis experiments showed that the Kunitz 1 domain (KD1) of TFPI inhibits FVIIa and that the Kunitz 2 domain (KD2) inhibits FXa [9]. The physiological importance of TFPI is demonstrated by premature mortality during embryogenesis in mice expressing a mutant TFPI lacking KD1 [10].

FXa inhibition by TFPI is a biphasic reaction that occurs via a so-called slow-tight binding mechanism [11]. In the first step, TFPI and FXa rapidly form a loose binary complex that slowly isomerises to a tight FXa-TFPI complex. FXa inhibition by TFPI is inhibited by Ca2+ ions and stimulated by phospholipids and protein S [12, 13]. Although KD2 of TFPI directly binds and inhibits FXa, the other domains of TFPI (KD1, KD3 and C-terminus) also contribute to FXa inhibition [14, 15]. In recent reports it was shown that KD1 is required for the transition of the loose to the tight FXa-TFPI complex and that KD3 promotes the formation of the loose FXa-TFPI complex [15] and is required for protein S stimulation of TFPI [16].

Much less is known about the inhibition of FVIIa by TFPI. Many reports showed that TF-FVIIa inhibition occurs via quaternary complex formation in which the KD2 domain of TFPI binds to the active site of FXa and the TF-FVIIa complex is subsequently inhibited by the interaction of KD1 of FXa-TFPI complex with the active site of FVIIa [17, 18]. TFPI can also inhibit TF-FVIIa in the absence of FXa [19, 20], but there is no information on the role of the different TFPI domains in FVIIa inhibition.

In the present study, we focused on the direct inhibition of FVIIa by TFPI and by various truncated TFPI constructs in order to investigate the role of the different domains of TFPI in the inhibition of TF-FVIIa. In addition, we tested the effect of relipidated TF, soluble TF, FXa and protein S on FVIIa inhibition by TFPIfl and the TFPI constructs. Our work provides new information on the contribution of the various TFPI domains, TF, phospholipid, FXa and protein S to the down-regulation of the extrinsic coagulation pathway by TFPI.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

Materials

The chromogenic substrates CS-11(65) (N-α-Benzyloxycarbonyl-D-arginyl-L-glycyl-L-arginine-p-nitroaniline-dihydrochloride) and Spectrozyme FVIIa (Methanesulphonyl-D-cyclohexylalanyl-butyl-arginine-paranitroaniline monoacetate salt) were purchased from Hyphen BioMed (Mason, OH, USA) and American Diagnostica (Stamford, CT, USA), respectively. Chromozyme t-PA (N-Methylsulfonyl-D-Phe-Gly-Arg-4-nitranilide acetate) was purchased from Roche Diagnostics (Mannheim, Germany). CaCl2, NaCl and EDTA were obtained from Merck Chemicals (Darmstadt, Germany). HEPES, Tris-hydrochloride, bovine serum albumin (BSA) and ovalbumin were obtained from Sigma Aldrich (Zwijndrecht, the Netherlands). Rivaroxaban was purchased from Bayer HealthCare (Leverkusen, Germany).

Proteins

Recombinant human full-length TFPI (TFPIfl) (1–276), produced in a bacterial expression system (Escherichia coli), was kindly provided by T. Lindhout (Cardiovascular Research Institute Maastricht). Purification and characterization of TFPIfl was described earlier [21]. TFPIfl stocks were diluted to 380 nm in 25 mm HEPES, 175 mm NaCl, pH 7.5 and 5 mg mL−1 BSA (HNBSA) and stored at −80 °C. TFPI1–150 (1–150), Kunitz 1 linker Kunitz 2 (22–150, KD1-KD2), N-terminus Kunitz 1 (1–83, N-KD1), Kunitz 1 (22–79, KD1) and Kunitz 2 (93–150, KD2) were expressed in a bacterial expression system (E. coli) BL21(DE3)pLysS (Merck, Darmstadt, Germany) using the expression vector pET19b (Merck, Darmstadt, Germany). The proteins were purified from inclusion bodies. Cell pellets were solubilised in 8 m urea, 20 mm DTT, 50 mm Tris-HCl, pH 8.0. All the constructs were folded in 50 mm Tris-HCl, pH 10.0, 1.1 mm oxidized glutathione by rapid dilution followed by dialysis against 20 mm Tris/HCl, pH 7.0. Purification was performed by a two-step procedure, a Q-Sepharose column followed by a Streptavidin affinity column with an immobilized peptide specific towards TFPI. TFPIfl and TFPI constructs were stored in HNBSA at −80 °C. Endogenous glycosylated full-length (TFPIfl) secreted from the human adenocarcinoma SK-Hep1 cell line (ATCC) was purified using a two-step affinity chromatography protocol. First, total TFPI was captured by an anti-TFPI peptide (Baxter Innovations GmbH, Vienna, Austria) followed by a monoclonal antibody affinity chromatography directed against the TFPI C-terminus (MW1848, Sanquin, Amsterdam, the Netherlands). The final purification product appears on SDS-PAGE as a single band of 43 kDa with a purity of > 98%. Glycosylated TFPIfl stocks were diluted to 750 nm in 25 mm HEPES, 175 mm NaCl, pH 7.5 and 5 mg mL−1 BSA (HNBSA) and stored at −80 °C.

TF (Innovin) was purchased from Siemens Healthcare (Marburg, Germany). FVIIa was purchased from Novo Nordisk (Bagsvaerd, Denmark). Recombinant Soluble TF (sTF) was purchased from Creative BioMart (New York, NY, USA). Human FXa (FXa) was purchased from Kordia (Leiden, the Netherlands) and bovine FXa (bFXa) was purified from bovine plasma [22]. Gla-domainless human FXa (GD-FXa) and protein S were from Haematologic Technologies (Essex Junction, VT, USA). Sheep polyclonal anti-protein S antibodies were purchased from Enzyme Research Laboratories (South Bend, IN, USA).

Concentration determination of TFPI modules

The concentrations of TFPIfl, TFPI1–150 and KD1-KD2 were determined by titrating a known amount of bFXa with TFPIfl and TFPI constructs [23]. bFXa (1 nm) was incubated with increasing concentrations of TFPIfl or TFPI constructs in HNBSA for 60 min at room temperature. As TFPI forms a 1:1 stoichiometric complex with FXa [9, 17], the x-axis intercept of a plot of residual activity of FXa determined with CS-11(65) as a function of the amount of TFPI allows calculation of the functional TFPI concentration. The concentrations of N-KD1, KD1 and KD2 were determined by titration of a known amount of trypsin and calculation of the functional concentrations of constructs as described above for TFPI.

Concentration determination of FVIIa

The concentration of FVIIa was determined by titrating FVIIa with a known amount of FXa (0–1 nm) in the presence of 5 nm TF, 3 mm CaCl2 and an excess of TFPI (1 nm). After incubating the reaction mixture for 15 min, residual FVIIa was quantified by adding 0.5 mm Spectrozyme FVIIa and 40 ug mL−1 Rivaroxaban to inhibit free FXa. As FXa and TFPI form a tight stoichiometric complex with FVIIa, the x-axis intercept of a plot of the residual FVIIa activity as a function of the FXa concentration is equal to the functional FVIIa concentration in the reaction mixture.

Inhibition of relipidated TF-FVIIa by TFPIfl

A reaction mixture containing 0.3 nm FVIIa, 5 nm relipidated TF (a concentration supporting maximal stimulation of FVIIa) and 3 mm CaCl2 in HNBSA buffer was incubated for 15 min at 37 °C. Subsequently, different amounts of TFPI were added and incubated for 15 min at 37 °C, after which pre-warmed 0.5 mm Spectrozyme FVIIa was added. Chromogenic substrate conversion, which is a measure for non-inhibited FVIIa, was followed at 405 nm in an Ultra Microplate Reader (Bio-Tek, Burlington, VT, USA) for 45 min. The final concentrations in the well were: 0.3 nm FVIIa, 5 nm TF, 0.5 mm Spectrozyme FVIIa and 0–18 nm TFPIfl. To understand the mechanism of inhibition of relipidated TF-FVIIa by TFPIfl (monophasic or biphasic inhibition), we also performed experiments in which substrate was added simultaneously with TFPIfl (Fig. 2A).

Inhibition of sTF-FVIIa and FVIIa by TFPI

A reaction mixture containing 0.6 nm FVIIa, 50 nm sTF (a concentration supporting maximal stimulation of FVIIa) and 3 mm CaCl2 in HNBSA buffer was incubated for 15 min at 37 °C. Different amounts of TFPI were then added and the mixture was incubated for an additional 15 min at 37 °C. Subsequently, pre-warmed 0.5 mm Spectrozyme FVIIa was added and chromogenic substrate conversion was followed at 405 nm in an Ultra Microplate Reader (Bio-Tek) for 45 min. The final concentrations in the well were: 0.6 nm FVIIa, 50 nm sTF, 0.5 mm Spectrozyme FVIIa and 0–1600 nm TFPIfl. In experiments without TF, 6 nm of FVIIa was used instead of 0.6 nm and 1 mm Chromozyme t-PA substrate was used instead of 0.5 mm Spectrozyme FVIIa, and chromogenic substrate conversion was followed as described above.

Inhibition of relipidated TF-FVIIa and sTF-FVIIa by TFPI constructs

Titrations of TF and sTF with TFPI constructs were performed as described above at 0.3 nm FVIIa, 5 nm relipidated TF, 3 mm CaCl2, 0.5 mm Spectrozyme FVIIa and 0–250 nm TFPI1–150, 0–250 nm KD1-KD2, 0–2000 nm N-KD1, 0–2000 nm KD1 or 0–800 nm KD2 (final concentrations in the well). For sTF-FVIIa inhibition the final concentrations in the well were: 0.6 nm FVIIa, 50 nm sTF, 3 mm CaCl2, 0.5 mm Spectrozyme FVIIa and 0–250 nm TFPI1–150, 0–250 nm KD1-KD2, 0–2000 nm N-KD1, 0–2000 nm KD1 or 0–800 nm KD2.

Inhibition of relipidated TF-FVIIa and sTF-FVIIa by TFPIfl and TFPI constructs in the presence of FXa

A reaction mixture containing 0.3 nm FVIIa, 5 nm relipidated TF, 3 mm CaCl2 and 1 nm FXa in HNBSA buffer was pre-incubated at 37 °C for 15 min. Subsequently, different concentrations of TFPIfl or TFPI constructs were added and incubated for 15 min at 37 °C. As Spectrozyme FVIIa is also a good substrate for FXa, prior to quantification of FVIIa, 10 ug mL−1 rivaroxaban was added to inhibit any free FXa present in the reaction mixture. As a control, 10 ug mL−1 rivaroxaban was added before the addition of FXa to prevent the quaternary complex formation and to test whether inhibition of FVIIa by TFPI (constructs) is affected by rivaroxaban, which did not appear to be the case. Rivaroxaban is a direct FXa inhibitor, which has been used in experiments with FXa to prevent quaternary complex formation. To keep the conditions similar, rivaroxaban was present in all the experiments. Rivaroxaban, which is a highly specific FXa inhibitor [24, 25], did not cause dissociation of pre-formed quaternary TFPI-TF-FVIIa-FXa complexes. The final concentrations in the well were: 0.3 nm FVIIa, 5 nm relipidated TF/50 nm sTF, 0.5 mm Spectrozyme FVIIa, and 0–5 nm TFPIfl, 0–50 nm TFPI1–150 or 0–50 nm KD1-KD2 and 1 nm FXa. Similar experiments were also performed with 1 nm GD-FXa.

Inhibition of relipidated TF-FVIIa by TFPI/TFPI constructs in the presence of protein S

A reaction mixture containing 0.3 nm FVIIa, 5 nm TF, 3 mm CaCl2, 10 μg mL−1 rivaroxaban and 100 nm protein S was incubated in HNBSA buffer for 15 min at 37 °C. As the protein S preparation contained traces of FX, 10 μg mL−1 rivaroxaban was present in the reaction mixtures. Different concentrations of TFPI/TFPI constructs were then added and the reaction mixture was incubated for an additional 15 min at 37 °C. Subsequently, 0.5 mm pre-warmed Spectrozyme FVIIa was added and chromogenic substrate conversion, which is a measure for non-inhibited FVIIa, was followed at 405 nm in an Ultra Microplate Reader (Bio-Tek) for 45 min. The final concentrations in the well were: 0.3 nm FVIIa, 5 nm relipidated TF/50 nm sTF, 10 μg mL−1 rivaroxaban, 100 nm protein S, 0.5 mm Spectrozyme FVIIa, and 0–40 nm TFPIfl, 0–250 nm TFPI1–150 or 0–250 nm KD1-KD2. A similar experiment was performed in the presence of 0.8 μm sheep polyclonal anti-protein S antibody.

Calculation of the Ki values of inhibition of FVIIa by TFPI and TFPI constructs

From the rates of chromogenic substrate conversion determined in the microplate reader the percentage inhibition of FVIIa by TFPI or TFPI constructs was calculated. The percentage FVIIa inhibition was plotted as a function of the TFPI (construct) concentration [TFPI] and fitted to the equation for the Langmuir binding isotherm:

  • display math(1)

where, I is the % of inhibition of FVIIa, Imax is the maximal inhibition (%) of FVIIa and IC50 is the inhibition constant (i.e. the concentration of TFPI or TFPI construct at which 50% of the maximal inhibition of FVIIa is observed).

The presence of chromogenic substrate may affect the IC50 of TFPI inhibition through competition with TFPI for the active site of FVIIa. The extent of competition depends on the concentration of chromogenic substrate relative to its Km for FVIIa. The Km values of Spectrozyme FVIIa for FVIIa at the reaction conditions in our experiments were 0.9 mm ( + TF), 2.6 mm ( + sTF) and 6.1 mm (no TF) (data not shown). As the substrate concentration used in our experiments (0.5 mm) was below the Km, Spectrozyme FVIIa had a small effect on the IC50 and Ki values for inhibition of TF-FVIIa by TFPI and TFPI constructs were calculated from the IC50 using the equation inline image [26].

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

TFPI constructs used in the study

The following TFPI constructs (Fig. 1) have been used in this study: TFPIfl (1–276 a.a.), TFPI1–150 (1–150 a.a.), which lacks KD3 and the C-terminus, KD1-KD2 (22–150 a.a.), lacking KD3 and the C-terminus and acidic N-terminus of TFPI, N-KD1 (1–83 a.a.), lacking KD2, KD3 and the C-terminus, KD1 (22–79 a.a.), similar to N-KD1 without the acidic N-terminus, and KD2 (93–150 a.a.).

image

Figure 1. TFPI constructs used in the study. (A) Primary structure of TFPI. Black amino acids indicate charged residues, grey diamonds and circles indicate potential N-linked and O-linked carbohydrates, grey numbered boxes indicate exons. P1 residues in the Kunitz domains are indicated by arrows. (B) List of TFPI constructs used in the current study.

Download figure to PowerPoint

Because all TFPI constructs used in our study were derived from E. coli and therefore not glycosylated we have compared them with Ecoli-derived full-length TFPI, which is also not glycosylated. A glycosylated form of TFPI expressed in human SK-Hep1 cells was used in a few control experiments (see below) to test whether glycosylation affected FVIIa inhibition by TFPI.

Direct inhibition of FVIIa by TFPIfl

In order to understand the mechanism of direct inhibition of FVIIa by TFPIfl, experiments were performed at three different conditions (i.e. FVIIa in the presence of relipidated TF and soluble TF (sTF) and in the absence of TF). Titrations of FVIIa with TFPIfl and TFPI constructs were carried out in the presence of 3 mm CaCl2. In the first experiment, TFPIfl was not pre-incubated with relipidated TF-FVIIa, but was added simultaneously with chromogenic substrate. The progress curves of chromogenic substrate conversion bend off at low TFPI concentrations (Fig. 2A). This is indicative of time-dependent inhibition of FVIIa at low TFPI concentrations, which can be either due to a biphasic slow-tight binding mechanism or slow monophasic binding [27] (see Discussion). The percentage of inhibition of FVIIa (calculated from the amidolytic activity) was plotted as a function of the TFPI concentration and fitted to a hyperbola yielding a Ki value of 4.6 nm (Fig. 2B, Table 1). Similar Ki values were obtained when FVIIa inhibition was quantified with chromogenic substrate after 15 min of pre-incubation of relipidated TF-FVIIa with TFPIfl. A glycosylated form of TFPI expressed in human SK-Hep1 cells inhibited TF-FVIIa with a Ki that was ~ 2-fold higher than the Ki of non-glycosylated TFPI (data not shown), which demonstrates that glycosylation has a small effect on the TF-FVIIa inhibitory activity of TFPI.

Table 1. Ki values with standard deviation for FVIIa inhibition by TFPI and TFPI constructs
 Ki (nm) (Relipidated TF)Ki (nm) sTFKi (nm) No TF
−PS + PS + FXa+FXa (−GD)−PS + PS + FXa−PL + PL + FXa
TFPI (FL)4.6 ± 0.80.7 ± 0.3< 0.1512 ± 2122 ± 18187 ± 28152 ± 491118 ± 213957 ± 238561 ± 23
TFPI (1–150)44 ± 676 ± 60.5 621 ± 63  >2500  
KD1-KD236 ± 844 ± 140.8 984 ± 247  >2500  
N-KD1380 ± 26 287 ± 42 1471 ± 503  >2500  
KD1434 ± 34 554 ± 95 1786 ± 305  >2500  
KD2459 ± 37 948 ± 406 1080 ± 319  >2500  
image

Figure 2. Direct inhibition of (s)TF-FVIIa by TFPIfl. (A) Progress curves of 500 μm Spectrozyme FVIIa substrate conversion by relipidated TF-FVIIa at different TFPIfl concentrations. Inhibition of 0.3 nm FVIIa, 5 nm TF by 0–100 nm TFPIfl in the presence of 3 mm CaCl2 without pre-incubation. Concentrations of TFPIfl are 0 nm, 1 nm, 2 nm, 5 nm, 10 nm, 20 nm, 50 nm and 100 nm. (B) Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex inhibited as a function of TFPIfl (0–100 nm). Solid lines represent fits for hyperbolic function (equation (1), Materials and methods). (C) Percentage of 0.6 nm FVIIa and 50 nm sTF complex (●) and 6 nm FVIIa (■) inhibited as a function of TFPIfl (0–1600 nm). Solid lines represent fits for hyperbolic function.

Download figure to PowerPoint

FVIIa was also titrated with TFPIfl in the presence of sTF and in the absence of TF. sTF (1–219 a.a.), which lacks the transmembrane and intracellular domain, does not contain lipid and does not bind to phospholipids [28]. Compared with relipidated TF-FVIIa, sTF-FVIIa was inhibited by TFPIfl with a ~ 26-fold higher Ki value (Ki = 122 nm, Table 1, Fig. 2C). In the absence of (s)TF, FVIIa was even more poorly inhibited by TFPIfl (Fig. 2C) and the Ki value was more than 200-fold higher (Ki ~ 1.118 nm) than that obtained in the presence of relipidated TF (Table 1).

Direct inhibition of relipidated and soluble TF-FVIIa by TFPI1–150 and KD1-KD2

To investigate the importance of the KD3-C-terminus of TFPI for the inhibition of FVIIa, two truncated constructs were used: TFPI1–150 and KD1-KD2. Compared with TFPI1–150, KD1-KD2 lacks the acidic N-terminus. The Ki values obtained for inhibition of the relipidated TF-FVIIa by both constructs of TFPI were similar, but were ~ 7–10-fold higher than those of TFPIfl (Fig. 3, Table 1). In presence of sTF, inhibition was even less efficient and the Ki values were 15–30-fold higher then those obtained with relipidated TF. When titrations of FVIIa with TFPI1–150 and KD1-KD2 were performed in the absence of TF, Ki values were too high to be accurately determined (> 2.500 nm).

image

Figure 3. Direct inhibition of (s)TF-FVIIa by TFPI1-150 and KD1-KD2. Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex or 0.6 nm FVIIa and 50 nm sTF complex inhibited as a function of TFPI1–150 and KD1-KD2 (0–250 nm). Solid lines represent fits for hyperbolic function (equation (1)). (●) Relipidated TF with TFPI1–150; (■) relipidated TF with KD1-KD2; (○) sTF with TFPI1–150; (□) sTF with KD1-KD2.

Download figure to PowerPoint

Direct inhibition of relipidated and soluble TF-FVIIa by N-KD1, KD1 and KD2

It is generally accepted that KD1 of TFPI inhibits FVIIa and that KD2 inhibits FXa [9]. Therefore, it was of interest to determine the inhibition of FVIIa by the isolated Kunitz domains, N-KD1, KD1 and KD2. High concentrations of these constructs were required for inhibition of relipidated TF-FVIIa and sTF-FVIIa. Surprisingly, FVIIa was not only inhibited by N-KD1 and KD1 but also by KD2, with similar Ki values (Fig. 4). However, the Ki values were much higher than those determined for (s)TF-FVIIa inhibition by TFPIfl, TFPI1–150 and KD1-KD2 (Table 1), which indicates that the other Kunitz domains (KD2 and KD3) are required for efficient inhibition of TF-FVIIa by TFPIfl.

image

Figure 4. Direct inhibition of (s)TF-FVIIa by N-KD1 and KD1 and KD2. Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex or 0.6 nm FVIIa and 50 nm sTF complex inhibited as a function of N-KD1 and KD1 (0–2000 nM); KD2 (0–800 nm). Solid lines represent fits for hyperbolic function (equation (1)). (●) Relipidated TF with N-KD1; (■) relipidated TF with KD1; (▲) relipidated TF with KD2; (○) sTF with N-KD1; (□) sTF with KD1; (∆) sTF with KD2.

Download figure to PowerPoint

The effect of protein S on the inhibition of FVIIa by TFPI and TFPI constructs

It has been shown that protein S is a co-factor of TFPIfl in FXa inhibition [13]. To study whether protein S also acts as co-factor for TFPIfl in relipidated TF-FVIIa inhibition, experiments were carried out in the presence of 100 nm protein S. In the presence of protein S there was a 6.5-fold decrease in the Ki value, which shows that protein S also promotes the inhibition of relipidated TF-FVIIa by TFPI (Fig. 5A). In order to confirm that it was not an artefact, the protein S titration was also performed in the presence of an anti-sheep polyclonal anti-protein S antibody. In the presence of the anti-protein S antibody, the co-factor activity of protein S was completely abrogated and the Ki value for TFPI was similar to the one obtained in the absence of protein S. Titration of TFPI1–150 and KD1-KD2 has also been performed in the presence of protein S. Both truncated forms of TFPI lack the KD3-C-terminus of TFPI and were not stimulated by protein S (Fig. 5B). This observation is in line with the recent observation that KD3 is required for the binding of TFPI to protein S [16, 29]. To study the role of phospholipids in the protein S co-factor activity for TFPIfl, sTF was used instead of relipidated TF. Protein S did not have any stimulatory effect on the inhibition of sTF-FVIIa by TFPIfl (Fig. 5C), indicating that not only the KD3-C-terminus, but also phospholipids, are required for protein S co-factor activity in FVIIa inhibition by TFPI.

image

Figure 5. Direct inhibition of (s)TF-FVIIa by TFPI and TFPI constructs in the presence and absence of protein S. (A) Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex inhibited as a function of TFPIfl (0–40 nm) in the absence of protein S represented by (●); in the presence of 100 nm protein S represented by (■); in the presence of 100 nm protein S and 0.8 μm sheep polyclonal anti-protein S antibody represented by (▲). (B) Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex inhibited as a function of TFPI1–150 and KD1-KD2 (0–250 nm) in the absence of protein S represented by (●) and (▲), respectively; in the presence of 100 nm protein S (○) and (∆), respectively. (C) Percentage of 0.6 nm FVIIa and 50 nm sTF complex inhibited as a function of TFPIfl (0–1600 nm) in the absence of protein S (●) and in the presence of 100 nm protein S (○). Solid lines represent fits for hyperbolic function (equation 1).

Download figure to PowerPoint

As in the absence of protein S, glycosylated TFPI isolated from human SK-Hep1 cells inhibited TF-FVIIa with a Ki that was ~ 2-fold higher than that determined for non-glycosylated TFPI (data not shown).

Effect of FXa on relipidated and soluble TF-FVIIa inhibition by TFPI and TFPI constructs

It has been shown that inhibition of relipidated TF-FVIIa by TFPI is greatly enhanced by FXa, which enables the formation of a tight quaternary TF-FVIIa-TFPI-FXa complex [19]. In order to further study the role of the TFPI domains in quaternary complex formation, we have tested the effect of FXa on the inhibition of TF-FVIIa by TFPIfl and TFPI constructs.

TFPIfl formed a 1:1 stoichiometric quaternary complex (relipidated TF-FVIIa-TFPI-FXa) with a Ki value that cannot be reliably determined and is < 0.15 nm (i.e. < 50% of the concentration of FVIIa present in the reaction mixture) (Fig. 6A). As the Gla-domain of FXa is essential for the Ca2+-dependent binding of FXa to phospholipids, GD-FXa was also used in the study. No quaternary complex was formed when GD-FXa was used instead of FXa (Fig. 6A). The phospholipid requirement for quaternary complex formation was studied by performing a similar experiment in which relipidated TF was replaced by sTF. sTF-FVIIa did not form a tight quaternary complex with TFPI and FXa (Fig. 6C).

image

Figure 6. TF-FVIIa inhibition by TFPI and TFPI constructs in the presence of FXa. (A) Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex inhibited as a function of TFPIfl (0–5 nm) in the absence of FXa (●);  + 1 nm FXa (■); + 10 ug mL−1 Rivaroxaban + 1 nm FXa (▲); + 1 nm GD-FXa (♦). (B) Percentage of 0.3 nm FVIIa and 5 nm relipidated TF complex inhibited as a function of TFPI1–150 and KD1-KD2 (0–50 nm) in the absence of FXa represented by (●) and (■), respectively; in the presence of 1 nm FXa (○) and (□), respectively. (C) Percentage of 0.6 nm FVIIa and 50 nm sTF complex inhibited as a function of TFPIfl (0–1600 nm) in the absence of FXa (●) and in the presence of 1 nm FXa (○). Solid lines represent fits for hyperbolic function (equation (1), Materials and methods).

Download figure to PowerPoint

In the absence of FXa, TFPI1–150 and KD1-KD2 had 7–10-fold higher Ki values than TFPIfl for inhibition of the relipidated TF-FVIIa complex (Table 1). Therefore, it was of interest to study the role of these two TFPI constructs in the inhibition of TF-FVIIa in the presence of FXa. Both TFPI1–150 and KD1-KD2 readily formed a quaternary complex, with Ki values of 0.5 nm and 0.8 nm, respectively (Fig. 6B). The observations with the truncated forms of TFPI show that the KD3-C-terminus is not absolutely required for quaternary complex formation.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

Although direct inhibition of TF-FVIIa by TFPI was described by Pederson et al. [19], most of the subsequent studies on FVIIa inhibition by TFPI were performed in the presence of FXa and/or cell surfaces. It is generally accepted that the KD1 domain of TFPI inhibits FVIIa and that KD2 binds and inhibits FXa [9, 17]. For inhibition of FXa by TFPI there are several reports that show other Kunitz domains (KD1 and KD3) and the C-terminus of TFPI contribute to FXa inhibition [14, 30, 31]. As limited information is available on the role of the different Kunitz domains of TFPI in the direct inhibition of FVIIa, we have performed a detailed kinetic analysis of FVIIa inhibition by TFPIfl and TFPI constructs lacking Kunitz domains and C- and N-termini. To establish whether phospholipids and TF also contribute to FVIIa inhibition by TFPI, experiments were performed at three different conditions (i.e. in the presence of relipidated TF and sTF and in the absence of TF). Full-length TF, a membrane integral glycoprotein (46 kDa), is a 263 a.a. single-chain polypeptide that consists of a 219 a.a. extracellular N-terminus and a 23 a.a. transmembrane domain followed by an intracellular 21 a.a. C-terminus [32-34]. The extracellular region of TF contains the FVII/FVIIa binding domains. sTF lacks both the transmembrane and the intracellular domain. The transmembrane domain of TF plays a crucial role in TF-FVIIa anchorage to the cell surface [35-37], in the activation of FX by TF-FVIIa [38] and in the autoactivation of FVII [39]. Hence, experiments were performed in the presence and absence of relipidated full-length TF and sTF to obtain information on the role of phospholipids and TF in the inhibition of FVIIa by TFPI and TFPI constructs.

We show that relipidated TF-FVIIa is directly inhibited by TFPIfl with a Ki value of 4.6 nm (Table 1). Progress curves of TF-FVIIa inhibition determined at increasing TFPI bend off at low TFPI concentrations (< 2 nm). This is indicative of time-dependent inhibition of FVIIa at low TFPI concentrations, which can be either due to a biphasic slow-tight binding mechanism or slow monophasic binding [27]. As this bending-off is only observed at TFPI concentrations < 2 nm (Fig. 2A), it is likely that the time-dependent inhibition at low TFPI concentrations is due to slow monophasic inhibition (t½ ≈ 5 min at 2 nm TFPI) and not to a slow-tight binding mechanism as observed for FXa inhibition by TFPIfl or TFPI1–150 [12-15]. Inhibition of FVIIa by TFPI was much less efficient for sTF (Ki = 122 nm) and in the absence of tissue factor (Ki = 1118 nm). These differences can be either attributed to binding of TFPI to phospholipids and/or TF promoting the interaction between TFPI and FVIIa or to effects of TF and/or PL on the conformation of FVIIa, exposing the active site of FVIIa and making it more accessible to inhibition by TFPI [40, 41].

A possible role of the KD3-C-terminus in the inhibition of relipidated TF-FVIIa was studied using truncated forms of TFPI (TFPI1–150 and KD1-KD2). The KD3-C-terminus significantly contributed to the inhibition of relipidated TF-FVIIa by TFPI as the Ki values of TFPI1–150 (Ki = 44 nm) and KD1-KD2 (Ki = 36 nm) were some 7–10-fold higher than the Ki determined for TFPIfl (Ki = 4.6 nm). As the C-terminus of TFPI is known to interact with the phospholipids [42], the absence of the KD3-C-terminus might affect the interaction of TFPI with TF and/or FVIIa, which are both bound to the phospholipid surface. However, TFPI1–150 and KD1-KD2 also had 5–8-fold higher Ki values than TFPIfl for inhibition of sTF-FVIIa, which indicates that the KD3-C-terminus is also important for direct interaction with and inhibition of sTF-FVIIa, independent of phospholipid.

KD2 of TFPI also contributes towards the direct inhibition of FVIIa by TFPIfl or TFPI1–150 as removal of KD2 results in TFPI constructs (KD1 or N-KD1) that are very poor inhibitors of relipidated TF-FVIIa with Ki values (380–434 nm) that are more than 10–12-fold higher than the Ki of TFPI1–150 and KD1-KD2 (Table 1). In the presence of sTF the Ki values for the isolated Kunitz domains were some 3–4-fold increased and in the absence of (s)TF inhibition was not detectable. The (s)TF requirement for FVIIa inhibition by isolated Kunitz domains suggests that FVIIa inhibition by isolated Kunitz domains is a specific phenomenon. Surprisingly, FVIIa was inhibited by (N-)KD1 and by KD2 with similar Ki values. This shows that the presence of the other Kunitz domains promotes specific inhibition of TF-FVIIa by KD1, a phenomenon that is similar to that observed for FXa inhibition by KD2 [14, 15].

Protein S is a vitamin-K-dependent protein with two distinct anticoagulant functions. Protein S acts as a co-factor of APC in FVa and FVIIIa inactivation [43] and enhances FXa inhibition by TFPI [13]. The current study shows that protein S also decreases the Ki value of relipidated TF-FVIIa inhibition by TFPI, which in the presence of protein S (Ki = 0.7 nm) is some 6–7-fold lower than the Ki observed in the absence of protein S (Ki = 4.6 nm). Protein S did not act as a co-factor of TFPI in FVIIa inhibition in the presence of anti-protein S antibodies, which shows that the protein S effect is specific and not due to contaminations in the protein preparations. Protein S did not stimulate relipidated TF-FVIIa inhibition by truncated forms of TFPI (TFPI1–150 and KD1-KD2) and also had no effect on the inhibition of sTF-FVIIa by TFPIfl. This indicates that both the KD3-C-terminus and phospholipids are required for protein S co-factor activity in FVIIa inhibition by TFPI.

Similar observations were made for the co-factor activity of protein S in FXa inhibition by TFPI, which is also only observed with TFPIfl in the presence of phospholipids and not with truncated forms of TFPI or in the absence of phospholipids [44]. The KD3-C-terminus requirement for the expression of protein S co-factor activity is supported by the observations of Ndonwi et al. and Ahnstrom et al. [16, 29], who showed that the KD3 domain is essential for protein S binding to TFPI and for the co-factor activity of protein S in FXa inhibition by TFPI. The possible physiological importance of the TFPI co-factor activity of protein S is discussed below.

It has been reported that efficient inhibition of TF-FVIIa requires the presence of FXa, which promotes binding of TFPI to TF-FVIIa, and the fast formation of a tight quaternary TF-FVIIa-TFPI-FXa complex in which both FXa and FVIIa are inhibited. We performed experiments in the presence of FXa to further investigate the mechanism of quaternary complex formation. In the presence of FXa, TFPIfl rapidly formed a tight 1:1 stoichiometric quaternary complex with TF and FVIIa with a Ki value < 0.15 nm (Fig. 6). FXa was also able to form a quaternary complex with TF-FVIIa and TFPI1–150 and KD1-KD2, but not with KD1 or KD2, which indicates that the presence of both the KD1 and KD2 domains is essential for quaternary complex formation, but that the KD3-C-terminus is not an absolute requirement. Quaternary complex formation appeared to be also strictly dependent on the presence of phospholipids and on the presence of a Gla-domain in FXa because FXa did not stimulate the inhibition of sTF-FVIIa by TFPI and GD-FXa had no effect on relipidated TF-FVIIa inhibition by TFPI. Together these data indicate that actually a pentenary (quintenary) TF-FVIIa-TFPI-FXa-PL complex rather than a quaternary complex is the basis of physiological TF-FVIIa inhibition by TFPI.

The formation of the quaternary TF-FVIIa-TFPI-FXa complex is fast and efficient. We were not able to follow the quaternary complex formation in time and also did not succeed in measuring the Ki or Kd of the quaternary complex because at FXa concentrations in the range of the FVIIa concentration a tight stoichiometric complex was formed. This suggests that in the presence of FXa, inhibition of TF-FVIIa by TFPI is so efficient that it may not require protein S. Indeed Ndonwi and Broze reported that in a model system containing purified proteins, protein S does not enhance TFPI inhibition of TF-FVIIa but promotes FXa inhibition by TFPI [45]. Experiments in more complex reaction systems (e.g. plasma) are required to appreciate the role of TFPI as an anticoagulant protein in down-regulating thrombin formation and to quantify the contribution of TF-FVIIa and FXa inhibition by TFPI and the effects of protein S.

One may question whether direct inhibition of TF-FVIIa by TFPI in the presence of phospholipids and protein S (Ki ~ 0.7 nm) is physiologically important (plasma full-length TFPI concentration ~ 0.25 nm). However, the local TFPI concentration at the site of thrombus formation may be substantially higher than 0.25 nm (e.g. through the release of TFPI from platelets at the growing thrombus). Under such conditions, local high concentrations of TFPI can directly inhibit TF-FVIIa, thereby contributing to the down-regulation of thrombin formation. An important role for TFPI released from platelets in regulating thrombus formation is indicated by a study of Maroney et al. [46], who demonstrated that in hemophilic mice with TFPI−/− hematopoietic cells blood loss on tail transections was significantly decreased.

In addition, it should also be emphasized that in a normal population the plasma TFPI concentration shows a large variation and individuals may have TFPI levels that are two to three times higher than the population average (0.25 nm). Levels of TFPI also considerably increase during heparin treatment [47] and are significantly elevated in pathological conditions such as cancer or coronary heart disease [48]. In these situations direct inhibition of TF-FVIIa by TFPI may contribute to the inhibition of coagulation.

Addendum

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

S. Peraramelli designed and performed the experiments, analyzed the data and wrote the manuscript; S. Thomassen and A. Heinzmann assisted in performing the experiments; J. Rosing supervised the project, designed the experiments, analyzed data and revised the manuscript; T. Hackeng supervised the project and revised the manuscript; R. Hartmann provided TFPI constructs and revised the manuscript; M. Dockal provided TFPI constructs, and contributed to data analysis and revising the manuscript; F. Scheiflinger critically revised the manuscript.

Disclosure of Conflict of Interests

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References

JR receives research support from, and acts as consultant for, Baxter Innovations GmbH. MD, RH and FS are full-time employees of Baxter Innovations GmbH. None of the other authors declares any conflict of interest.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Disclosure of Conflict of Interests
  9. References
  • 1
    Wilcox JN, Smith KM, Schwartz SM, Gordon D. Localization of tissue factor in the normal vessel wall and in the atherosclerotic plaque. Proc Natl Acad Sci USA 1989; 86: 283943.
  • 2
    Morrissey JH, Macik BG, Neuenschwander PF, Comp PC. Quantitation of activated factor VII levels in plasma using a tissue factor mutant selectively deficient in promoting factor VII activation. Blood 1993; 81: 73444.
  • 3
    Nemerson Y. The reaction between bovine brain tissue factor and factors VII and X. Biochemistry 1966; 5: 6018.
  • 4
    Osterud B, Rapaport SI. Activation of factor IX by the reaction product of tissue factor and factor VII: additional pathway for initiating blood coagulation. Proc Natl Acad Sci USA 1977; 74: 52604.
  • 5
    Bom VJ, Bertina RM. The contributions of Ca2+, phospholipids and tissue-factor apoprotein to the activation of human blood-coagulation factor X by activated factor VII. Biochem J 1990; 265: 32736.
  • 6
    Salemink I, Franssen J, Willems GM, Hemker HC, Lindhout T. Inhibition of tissue factor-factor VIIa-catalyzed factor X activation by factor Xa-tissue factor pathway inhibitor. A rotating disc study on the effect of phospholipid membrane composition. J Biol Chem 1999; 274: 2822532.
  • 7
    Sanders NL, Bajaj SP, Zivelin A, Rapaport SI. Inhibition of tissue factor/factor VIIa activity in plasma requires factor X and an additional plasma component. Blood 1985; 66: 20412.
  • 8
    Wun TC, Kretzmer KK, Girard TJ, Miletich JP, Broze GJ Jr. Cloning and characterization of a cDNA coding for the lipoprotein-associated coagulation inhibitor shows that it consists of three tandem Kunitz-type inhibitory domains. J Biol Chem 1988; 263: 60014.
  • 9
    Girard TJ, Warren LA, Novotny WF, Likert KM, Brown SG, Miletich JP, Broze GJ Jr. Functional significance of the Kunitz-type inhibitory domains of lipoprotein-associated coagulation inhibitor. Nature 1989; 338: 51820.
  • 10
    Chan JC. Gene targeting in hemostasis. tissue factor pathway inhibitor. Front Biosci 2001; 6: D21621.
  • 11
    Broze GJ Jr, Girard TJ, Novotny WF. Regulation of coagulation by a multivalent Kunitz-type inhibitor. Biochemistry 1990; 29: 753946.
  • 12
    Huang ZF, Wun TC, Broze GJ Jr. Kinetics of factor Xa inhibition by tissue factor pathway inhibitor. J Biol Chem 1993; 268: 269505.
  • 13
    Hackeng TM, Sere KM, Tans G, Rosing J. Protein S stimulates inhibition of the tissue factor pathway by tissue factor pathway inhibitor. Proc Natl Acad Sci USA 2006; 103: 310611.
  • 14
    Higuchi DA, Wun TC, Likert KM, Broze GJ Jr. The effect of leukocyte elastase on tissue factor pathway inhibitor. Blood 1992; 79: 17129.
  • 15
    Peraramelli S, Suylen DP, Rosing J, Hackeng TM. The Kunitz 1 and Kunitz 3 domains of tissue factor pathway inhibitor are required for efficient inhibition of factor Xa. Thromb Haemost 2012; 108: 26676.
  • 16
    Ndonwi M, Tuley EA, Broze GJ Jr. The Kunitz-3 domain of TFPI-alpha is required for protein S-dependent enhancement of factor Xa inhibition. Blood 2010; 116: 134451.
  • 17
    Broze GJ Jr, Warren LA, Novotny WF, Higuchi DA, Girard JJ, Miletich JP. The lipoprotein-associated coagulation inhibitor that inhibits the factor VII-tissue factor complex also inhibits factor Xa: insight into its possible mechanism of action. Blood 1988; 71: 33543.
  • 18
    Warn-Cramer BJ, Rao LV, Maki SL, Rapaport SI. Modifications of extrinsic pathway inhibitor (EPI) and factor Xa that affect their ability to interact and to inhibit factor VIIa/tissue factor: evidence for a two-step model of inhibition. Thromb Haemost 1988; 60: 4536.
  • 19
    Pedersen AH, Nordfang O, Norris F, Wiberg FC, Christensen PM, Moeller KB, Meidahl-Pedersen J, Beck TC, Norris K, Hedner U, et al. Recombinant human extrinsic pathway inhibitor. Production, isolation, and characterization of its inhibitory activity on tissue factor-initiated coagulation reactions. J Biol Chem 1990; 265: 1678693.
  • 20
    Hamik A, Setiadi H, Bu G, McEver RP, Morrissey JH. Down-regulation of monocyte tissue factor mediated by tissue factor pathway inhibitor and the low density lipoprotein receptor-related protein. J Biol Chem 1999; 274: 49629.
  • 21
    Diaz-Collier JA, Palmier MO, Kretzmer KK, Bishop BF, Combs RG, Obukowicz MG, Frazier RB, Bild GS, Joy WD, Hill SR, et al. Refold and characterization of recombinant tissue factor pathway inhibitor expressed in Escherichia coli. Thromb Haemost 1994; 71: 33946.
  • 22
    Esnouf MP, Lloyd PH, Jesty J. A method for the simultaneous isolation of factor X and prothrombin from bovine plasma. Biochem J 1973; 131: 7819.
  • 23
    Girard TJ, Broze GJ Jr. Tissue factor pathway inhibitor. Methods Enzymol 1993; 222: 195209.
  • 24
    Verma AK, Brighton TA. The direct factor Xa inhibitor rivaroxaban. Med J Aust 2009; 190: 37983.
  • 25
    Samama MM. The mechanism of action of rivaroxaban–an oral, direct Factor Xa inhibitor–compared with other anticoagulants. Thromb Res 2011; 127: 497504.
  • 26
    Lazareno S, Birdsall NJ. Estimation of competitive antagonist affinity from functional inhibition curves using the Gaddum, Schild and Cheng-Prusoff equations. Br J Pharmacol 1993; 109: 11109.
  • 27
    Morrison JF, Walsh CT. The behavior and significance of slow-binding enzyme inhibitors. Adv Enzymol Relat Areas Mol Biol 1988; 61: 201301.
  • 28
    Mackman N. Alternatively spliced tissue factor - one cut too many? Thromb Haemost 2007; 97: 58.
  • 29
    Ahnstrom J, Andersson HM, Hockey V, Meng Y, McKinnon TA, Hamuro T, Crawley JT, Lane DA. Identification of functionally important residues in TFPI Kunitz domain 3 required for the enhancement of its activity by protein S. Blood 2012; 120: 505962.
  • 30
    Piro O, Broze GJ Jr. Role for the Kunitz-3 domain of tissue factor pathway inhibitor-alpha in cell surface binding. Circulation 2004; 110: 356772.
  • 31
    Wesselschmidt R, Likert K, Girard T, Wun TC, Broze GJ Jr. Tissue factor pathway inhibitor: the carboxy-terminus is required for optimal inhibition of factor Xa. Blood 1992; 79: 200410.
  • 32
    Petersen LC, Valentin S, Hedner U. Regulation of the extrinsic pathway system in health and disease: the role of factor VIIa and tissue factor pathway inhibitor. Thromb Res 1995; 79: 147.
  • 33
    Chu AJ. Tissue factor mediates inflammation. Arch Biochem Biophys 2005; 440: 12332.
  • 34
    Nemerson Y. Tissue factor and hemostasis. Blood 1988; 71: 18.
  • 35
    Ruf W, Rehemtulla A, Morrissey JH, Edgington TS. Phospholipid-independent and -dependent interactions required for tissue factor receptor and cofactor function. J Biol Chem 1991; 266: 16256.
  • 36
    Neuenschwander PF, Morrissey JH. Deletion of the membrane anchoring region of tissue factor abolishes autoactivation of factor VII but not cofactor function. Analysis of a mutant with a selective deficiency in activity. J Biol Chem 1992; 267: 1447782.
  • 37
    Paborsky LR, Caras IW, Fisher KL, Gorman CM. Lipid association, but not the transmembrane domain, is required for tissue factor activity. Substitution of the transmembrane domain with a phosphatidylinositol anchor. J Biol Chem 1991; 266: 219116.
  • 38
    Waxman E, Ross JB, Laue TM, Guha A, Thiruvikraman SV, Lin TC, Konigsberg WH, Nemerson Y. Tissue factor and its extracellular soluble domain: the relationship between intermolecular association with factor VIIa and enzymatic activity of the complex. Biochemistry 1992; 31: 39984003.
  • 39
    Fiore MM, Neuenschwander PF, Morrissey JH. The biochemical basis for the apparent defect of soluble mutant tissue factor in enhancing the proteolytic activities of factor VIIa. J Biol Chem 1994; 269: 1439.
  • 40
    Lawson JH, Butenas S, Ribarik N, Mann KG. Complex-dependent inhibition of factor VIIa by antithrombin III and heparin. J Biol Chem 1993; 268: 76770.
  • 41
    Rand KD, Jorgensen TJ, Olsen OH, Persson E, Jensen ON, Stennicke HR, Andersen MD. Allosteric activation of coagulation factor VIIa visualized by hydrogen exchange. J Biol Chem 2006; 281: 2301824.
  • 42
    Valentin S, Schousboe I. Factor Xa enhances the binding of tissue factor pathway inhibitor to acidic phospholipids. Thromb Haemost 1996; 75: 796800.
  • 43
    Dahlback B, Protein S. and C4b-binding protein: components involved in the regulation of the protein C anticoagulant system. Thromb Haemost 1991; 66: 4961.
  • 44
    van Wijnen M, Stam JG, van't Veer C, Meijers JC, Reitsma PH, Bertina RM, Bouma BN. The interaction of protein S with the phospholipid surface is essential for the activated protein C-independent activity of protein S. Thromb Haemost 1996; 76: 397403.
  • 45
    Ndonwi M, Broze G Jr. Protein S enhances the tissue factor pathway inhibitor inhibition of factor Xa but not its inhibition of factor VIIa-tissue factor. J Thromb Haemost 2008; 6: 10446.
  • 46
    Maroney SA, Cooley BC, Ferrel JP, Bonesho CE, Nielsen LV, Johansen PB, Hermit MB, Petersen LC, Mast AE. Absence of hematopoietic tissue factor pathway inhibitor mitigates bleeding in mice with hemophilia. Proc Natl Acad Sci USA 2012; 109: 392731.
  • 47
    Sandset PM, Abildgaard U, Larsen ML. Heparin induces release of extrinsic coagulation pathway inhibitor (EPI). Thromb Res 1988; 50: 80313.
  • 48
    Lindahl AK, Sandset PM, Abildgaard U. The present status of tissue factor pathway inhibitor. Blood Coagul Fibrinolysis 1992; 3: 43949.