SEARCH

SEARCH BY CITATION

Keywords:

  • ELISA;
  • protein S;
  • protein S deficiency;
  • TFPI;
  • thrombin generation assay

Abstract

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

Summary. Background: Protein S acts as a cofactor for full-length tissue factor pathway inhibitor (TFPI) in the downregulation of thrombin formation. Objective: To develop a functional test to measure the activity of the TFPI–protein S system in plasma. Methods/Patients: Using calibrated automated thrombography, we quantified the activity of the TFPI–protein S system in plasma by measuring thrombin generation in the absence and presence of neutralizing antibodies against protein S or TFPI. Moreover, we designed an enzyme-linked immunosorbent assay (ELISA) to determine the level of full-length TFPI in plasma. The performance of these assays was examined in plasma from 85 normal individuals and from 35 members of protein S-deficient families. Results: The ratio of thrombin peaks determined in the absence and presence of anti-protein S antibodies (protein S ratio = 0.5 in normal plasma) is a measure of the TFPI cofactor activity of protein S, whereas the ratio of thrombin peaks determined in the absence and presence of anti-TFPI antibodies (TFPI ratio = 0.25 in normal plasma) is a measure of the overall activity of the TFPI–protein S system. Protein S and TFPI ratios were elevated in protein S-deficient individuals, indicating an impairment of the TFPI–protein S system. Both ratios correlated well with full-length TFPI levels, which were significantly lower in protein S-deficient patients than in normal family members. Conclusions: Functional assays for the TFPI–protein S system and an ELISA for full-length TFPI were developed. These assays show that the activity of the TFPI–protein S anticoagulant pathway is impaired in individuals with congenital protein S deficiency.


Introduction

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

Protein S is a vitamin K-dependent plasma protein (75 kDa) [1] that downregulates blood coagulation by multiple mechanisms. As a cofactor for activated protein C (APC), protein S accelerates APC-mediated inactivation of factor Va and FVIIIa [2,3]. Protein S also exhibits APC-independent anticoagulant activity [4–8], which was recently explained by its ability to stimulate the inactivation of FXa by tissue factor pathway inhibitor (TFPI) [9–13].

TFPI is a Kunitz-type inhibitor that circulates in plasma at a concentration of 2.5 nm [14]. The majority of plasma TFPI is truncated and bound to lipoproteins, and expresses reduced anticoagulant activity. Only 10% of total TFPI circulates free in a full-length 43-kDa form [15,16]. TFPI binds to and inhibits FXa through its Kunitz-2 domain, and subsequently inhibits FVIIa–tissue factor (TF) by forming an inactive FXa–TFPI–TF–FVIIa quaternary complex through its Kunitz-1 domain [17]. Protein S stimulates TFPI by enhancing the interaction between TFPI and FXa 10-fold [9].

Protein S deficiency [18] and low levels of TFPI [19] are risk factors for venous thrombosis. However, it is not known whether the TFPI cofactor activity of protein S is reduced in protein S-deficient individuals and, if so, whether this contributes to their hypercoagulable state. Therefore, we have designed thrombin generation-based tests to measure the TFPI cofactor activity of protein S and the overall activity of the TFPI–protein S system in plasma. These assays, which rely on the increase in thrombin generation after the addition of antibodies against protein S or TFPI, were validated in plasma of healthy and protein S-deficient individuals.

Materials and methods

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

Plasma samples

Normal pooled plasma (NPP) in 3.2% sodium citrate was prepared in-house by pooling plasma from 98 healthy volunteers. NPP in 3.8% sodium citrate was obtained from George King Biomedical (Overland Park, KS, USA).

Plasma samples were collected in 3.2% sodium citrate from 85 healthy employees of the University Hospital Maastricht. Blood was centrifuged at 2000 ×g for 15 min to separate plasma from blood cells, and again at 11 000 ×g for 5 min to obtain platelet-poor plasma (PPP). Moreover, 18 individuals with heterozygous type I protein S deficiency (total protein S levels, < 70%; free protein S levels, < 65%) belonging to 10 unrelated families were recruited at Padua Academic Hospital. Seventeen relatives without protein S deficiency acted as controls. Seven of 18 protein S-deficient individuals and two of 17 controls had experienced venous thrombosis. Blood was collected in 3.8% sodium citrate and centrifuged at 2000 ×g for 10 min to obtain PPP.

Thrombin generation-based assays for measurement of the activity of the TFPI–protein S system

Thrombin generation was determined using calibrated automated thrombography. Plasma (68 μL) was incubated for 15 min at 37 °C with 4 μL of corn trypsin inhibitor (CTI; 33 μg mL−1 final concentration; Haematologic Technologies, Essex Junction VT, USA) and with 8 μL of either Hepes–NaCl buffer (25 mm Hepes, pH 7.4, 175 mm NaCl), polyclonal antibodies against protein S (2.80 μm; Zebra Bioscience, Enschede, The Netherlands), or monoclonal antibodies against TFPI (0.66 μm, MW1848; Sanquin, Amsterdam, The Netherlands). The anti-TFPI antibody was directed against the C-terminal domain, which is required for optimal inhibition of FXa [20,21] and stimulation by protein S [9]. Coagulation was initiated with 20 μL of a 7 : 1 mixture of the PPP low and PPP 5 pm reagents (Thrombinoscope B.V., Maastricht, The Netherlands), yielding approximately 1.5 pm TF and 4 μm phospholipids in a final volume of 120 μL. These concentrations are optimal for quantification of the activity of the TFPI–protein S system. After addition of 20 μL of CaCl2 and fluorogenic substrate (I-1140; Bachem, Bubendorf, Switzerland), giving final concentrations of 16 mm and 300 μm, respectively, substrate conversion by thrombin was followed in a Fluoroskan Ascent reader (Thermo Labsystems, Helsinki, Finland) with 390-nm excitation and 460-nm emission filter sets. Peak heights of thrombin generation were calculated using software obtained from Thrombinoscope B.V. [22]. The TFPI cofactor activity of protein S was expressed as the ratio of thrombin peaks determined in the absence and presence of anti-protein S antibodies (protein S ratio). The activity of the TFPI–protein S system as a whole was expressed as the ratio of thrombin peaks determined in the absence and presence of anti-TFPI antibodies (TFPI ratio). Peak height was used instead of endogenous thrombin potential (ETP) as a more reliable measure of thrombin generation [23]. In fact, at low TF concentrations, thrombin generation returns to zero relatively slowly after the peak, a phenomenon that is known as tailing. This can be clearly observed in Fig. 1B. Small changes in the tailing effect substantially change the ETP and affect its reproducibility, whereas peak heights are not influenced by differences in tailing.

image

Figure 1.  Effects of anti-protein S and anti-tissue factor pathway inhibitor (TFPI) antibodies on thrombin generation at a low tissue factor concentration (1.5 pm). Average of 12 separate thrombin generation curves (gray) representing the interassay variation in (A) in-house normal pooled plasma or in (B) George King normal pooled plasma without addition of antibodies (dotted line), with anti-protein S antibodies (dashed line) or with anti-TFPI antibodies (solid line).

Download figure to PowerPoint

Determination of factor levels in plasma

Protein S enzyme-linked immunosorbent assays (ELISAs) were performed as previously described [6,7], using NPP as a standard. Total and free protein S antigen levels were expressed as percentages of NPP.

Plasma free TFPI levels were measured with a commercial ELISA kit (Asserachrom Free TFPI; Diagnostica Stago, Asnières, France). Plasma TFPI levels were also determined with the Actichrome TFPI activity assay (American Diagnostica, Stamford, CT, USA), which quantifies total TFPI [24].

Full-length TFPI (TFPIFL) ELISA

Set-up  Microtiter plate wells were coated overnight at 4 °C with 50 μL of a solution containing 2 μg mL−1 of a monoclonal antibody against the C-terminal region of TFPI (MW1848; Sanquin) in 0.1 m Na2CO3 (pH 9.0). Further steps were performed at room temperature. After coating, the wells were blocked for 2 h with 200 μL of blocking buffer, consisting of Hepes-buffered saline (HBS: 25 mm Hepes, pH 7.7, 175 mm NaCl) with 30 mg mL−1 bovine serum albumin (BSA). Plasma samples were diluted 1 : 5 in blocking buffer, and 50 μL of the dilution was incubated in the wells for 1 h. After washing three times with 200 μL of washing buffer (HBS containing 0.03% Tween-20), 50 μL of a peroxidase-conjugated monoclonal anti-human TFPI antibody against Kunitz-2 (MW1845; Sanquin) diluted 1 : 1000 in HBS containing 5 mg mL−1 BSA was added (2 μg mL−1 final concentration) to the wells and incubated for 1 h. After washing five times with 200 μL of washing buffer, peroxidase activity was determined by chromogenic substrate conversion (TMB Enzymatic Kitl; Pierce, Rockford IL, USA), according to the manufacturer’s instructions. Serial dilutions of NPP were used as standards. Recombinant TFPIFL and truncated TFPI (TFPI1–161) were a kind gift from T. Lindhout from our institute.

Preparation of synthetic TFPI modules for ELISA validation  The Kunitz-2 (amino acids 92–150; three disulfides), Kunitz-3 (amino acids 182–241; three disulfides) and C-terminal (amino acids 242–276) domains of TFPI were synthesized by tBoc solid-phase peptide synthesis as N-acetylated, C-amidated peptides, as previously described [25], except for the C-terminus, which contained a C-terminal carboxylic acid. Oxidative folding of the crude Kunitz modules was performed with 1 mm cystine/8 mm cysteine to yield internal disulfide-bridged Kunitz domains, which were purified by high-performance liquid chromatography and lyophilized. Electrospray-ionization quadrupole mass spectrometry revealed masses of 7008.8 ± 0.4 for Kunitz-2 (theoretical monoisotopic mass, 7006.1; average mass, 7010.8) and 6887.1 ± 0.8 for Kunitz-3 (theoretical monoisotopic mass, 6884.2; average mass, 6888.8), in both cases confirming a mass decrease of 6 units as compared with the reduced states of Kunitz domains, representing the loss of six protons owing to the formation of three disulfide bonds.

FXa inhibition by synthetic TFPI modules  FXa (0.2 nm final concentration) in HBS containing 3 mm CaCl2 and 30 mg mL−1 BSA were incubated with the synthetic TFPI modules in 150 μL at 37 °C. After 45 min, 50 μL of chromogenic substrate S2765 (Chromogenix, Malmö, Sweden) was added to a final concentration of 125 μm. Conversion of S2765 was measured in a 96-well plate reader. FXa activity was calculated from a linear fit of the first 4 min of substrate conversion.

Statistical analysis

Data are expressed as mean ± standard deviation. Population means were compared by Student’s t-test. Correlations are expressed as Pearson coefficients (r).

Results

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

Functional assays for the TFPI–protein S anticoagulant pathway

NPP (3.2% citrate) and George King normal plasma (GKP; 3.8% citrate) were incubated with CTI in the absence or presence of neutralizing antibodies against protein S or TFPI for 15 min at 37 °C. Thrombin generation was then initiated with a low TF concentration (1.5 pm).

The intra-assay and interassay variations of the peak heights of thrombin generation curves were determined by performing the assay 12 times in duplicate (Table 1). For the interassay variation, the average peak heights in the absence of antibodies were 36.0 nm for NPP and 45.7 nm for GKP (Fig. 1). Addition of antibodies to protein S increased the thrombin peaks to 73.1 nm for NPP and 100.9 nm for GKP, owing to inhibition of TFPI cofactor activity of protein S (Table 1; Fig. 1). Complete inhibition of the TFPI–protein S system was achieved through addition of inhibitory antibodies against TFPI, resulting in increases of peak heights to 149.5 nm for NPP and 153.8 nm for GKP (Table 1; Fig. 1). No additional increase in thrombin generation was observed when anti-protein S antibodies were added on top of anti-TFPI antibodies (data not shown). The anticoagulant activities of protein S and TFPI were expressed as ratio of peak heights obtained in the absence and presence of the respective antibodies. The protein S ratios were 0.49 and 0.45 for NPP and GKP, respectively, meaning that, under these conditions, protein S, via its TFPI cofactor activity, reduced thrombin generation by more than 50%. The TFPI ratios were 0.24 and 0.30, respectively, indicating that the TFPI–protein S system as a whole reduced thrombin generation by approximately 75%. The intra-assay and interassay coefficients of variation (CVs) of all parameters determined with these assays were ≤ 7% (Table 1).

Table 1.   Intra-assay and interassay variations of the protein S and tissue factor pathway inhibitor (TFPI) activity assays
 PlasmaPlasma + α-protein SPlasma + α-TFPIProtein S ratioTFPI ratio
  1. CV, coefficient of variation. Data are expressed as mean ± standard deviation.

Intra-assay variation (n = 12)
 Peak height (nm)  Normal pooled plasma30.4 ± 1.1 CV: 4%64.6 ± 3.0 CV: 5%143.4 ± 4.0 CV: 3%0.47 ± 0.03 CV: 6%0.21 ± 0.01 CV: 6%
 Peak height (nm)  George King plasma47.2 ± 2.5 CV: 5%105.2 ± 5.5 CV: 5%156.6 ± 3.9 CV: 2%0.45 ± 0.03 CV: 7%0.30 ± 0.01 CV: 5%
Interassay variation (n = 12)
 Peak height (nm)  Normal pooled plasma36.0 ± 2.0 CV: 6%73.1 ± 5.3 CV: 7%149.5 ± 5.2 CV: 3%0.49 ± 0.02 CV: 5%0.24 ± 0.01 CV: 6%
 Peak height (nm)  George King plasma45.7 ± 3.2 CV: 7%100.9 ± 7.4 CV: 7%153.8 ± 6.2 CV: 4%0.45 ± 0.02 CV: 5%0.30 ± 0.02 CV: 7%

Development of a TFPIFL ELISA

As protein S only stimulates TFPIFL, which is the molecular form of TFPI that is most active in downregulating the TF pathway, an ELISA was developed to measure the plasma levels of this form of TFPI. TFPIFL was quantified using a capturing monoclonal antibody against the C-terminus and a peroxidase-conjugated monoclonal antibody against the Kunitz-2 domain of TFPI for detection (Fig. 2A). TFPIFL was detectable in normal plasma but not in TFPI-depleted plasma (Fig. 2B). In addition, purified recombinant TFPIFL was detected by the ELISA, whereas purified truncated TFPI (TFPI1–161) lacking the C-terminal portion was not (Fig. 2C).

image

Figure 2.  Design of a full-length (FL) tissue factor pathway inhibitor (TFPI) enzyme-linked immunosorbent assay. (A) Full-length TFPI present in plasma was captured by monoclonal antibodies against the C-terminus (C-term) and detected by peroxidase-conjugated monoclonal antibodies against Kunitz-2 (K2) of TFPI. The arrow indicates the C-terminal residue (161) of truncated TFPI. (B) Calibration curve of normal pooled plasma (•) and TFPI-depleted plasma (○). (C) Calibration curve of TFPI-depleted plasma reconstituted with either purified full-length TPFI (•) or truncated TFPI1–161 (○). Averages of duplicate measurements are shown. OD, optical density.

Download figure to PowerPoint

The TFPIFL ELISA was validated in competition experiments with chemically synthesized TFPI domains (Kunitz-2, Kunitz-3 and the C-terminus). As an example, Kunitz-2 is shown (Fig. 3A); the observed mass of this after folding was 7008.8 ± 0.4 (Fig. 3B). To determine whether folded Kunitz-2 adopted a native-like conformation, its activity as an FXa inhibitor was determined. Kunitz-2 caused dose-dependent inhibition of FXa activity, giving total inhibition at concentrations > 5 μm, whereas Kunitz-3 and the C-terminus did not inhibit FXa (Fig. 3C).

image

Figure 3.  Validation of the full-length tissue factor pathway inhibitor (TFPI) enzyme-linked immunosorbent assay (ELISA) with synthetic TFPI modules. (A) Secondary structure of the Kunitz-2 domain of TFPI. (B) Electrospray-ionization mass spectrum of the purified folded synthetic Kunitz-2 domain of TFPI. The observed mass fits well between the theoretical monoisotopic mass of 7006.1 and the average mass of 7010.8. (C) Inhibition of FXa activity by the synthetic TFPI modules Kunitz-2 (bsl00066), Kunitz-3 (△), and C-terminus (•). (D) Competition between full-length TFPI and Kunitz-2 (bsl00066), Kunitz-3 (△) and C-terminal (•) domains added to normal pooled plasma (1 : 5 dilution) during the capture phase of the full-length TFPI ELISA. (E) Competition between full-length TFPI and Kunitz-2 (bsl00066), Kunitz-3 (△) and C-terminal (•) domains added together with the peroxidase-conjugated antibody during the detection phase of the full-length TFPI ELISA. Averages of duplicate measurements are shown (C, D, E). OD, optical density

Download figure to PowerPoint

Subsequently, we performed the TFPI ELISA on 1 : 5 diluted NPP supplemented with varying amounts of Kunitz-2, Kunitz-3 or the C-terminus of TFPI. Only the C-terminus was able to inhibit the binding of plasma TFPIFL to the capturing antibody, whereas Kunitz-2 and Kunitz-3 had no effect on the TFPIFL ELISA (Fig. 3D). When synthetic modules were added during the detection step, only Kunitz-2 inhibited detection of TFPIFL in plasma (Fig. 3E).

The intra-assay variation of the TFPIFL ELISA was 2.2% (n = 3 samples, equally divided over the plate) and the interassay variation was 4.6% (n = 11 plates).

TFPI–protein S activity in plasma from normal individuals

Thrombin generation and protein S and TFPI ratios were determined in 85 normal individuals, comprising 48 males and 37 females (Fig. 4; Table 2). The average peak height of thrombin generation in the absence of antibodies was 38.3 ± 22.1 nm, and was different between males and females (males, 33.2 ± 11.3 nm; females, 45.0 ± 29.9 nm; P = 0.014). In the presence of anti-protein S antibodies, the average peak height increased to 77.1 ± 28.3 nm and was no longer significantly different between males and females (72.2 ± 20.7 and 83.4 ± 35.2 nm, respectively, P = 0.07). In the presence of anti-TFPI antibodies, the peak height increased to 150.5 ± 31.9 nm and was the same for males and females (149.8 ± 26.8 and 152.6 ± 37.8 nm, respectively, P = 0.7). Therefore, the differences between the protein S and TFPI ratios of males and females mostly originated from differences in thrombin generation determined in the absence of antibodies, in which females have a significantly higher peak height than males (Table 2). The protein S ratio was 0.45 ± 0.08 in males and 0.51 ± 0.11 in females (P = 0.009). The TFPI ratio was 0.22 ± 0.05 in males and 0.28 ± 0.12 in females (P = 0.006).

image

Figure 4.  Functional tissue factor pathway inhibitor (TFPI)–protein S assays in normal individuals. (A) Peak height of thrombin generation in the presence or absence of antibodies against protein S (αPS) or TFPI (αTFPI) in plasma from 85 healthy individuals, comprising 48 males (•) and 37 females (○). (B) Protein S ratio and TFPI ratio in plasma from the same individuals.

Download figure to PowerPoint

Table 2.   Tissue factor pathway inhibitor (TFPI)–protein S parameters in normal individuals
 Males (n = 48)Females (n = 37)
  1. Data are expressed as mean ± standard deviation. *P < 0.05; **P < 0.001.

Peak height (nm)33.2 ± 11.345.0 ± 29.9*
Peak height + α-protein S (nm)72.2 ± 20.783.4 ± 35.2
Peak height + α-TFPI (nm)149.8 ± 26.8152.6 ± 37.8
Protein S ratio0.45 ± 0.080.51 ± 0.11*
TFPI ratio0.22 ± 0.050.28 ± 0.12*
Free protein S (%)102.2 ± 16.284.0 ± 17.3**
Total protein S (%)98.3 ± 14.384.1 ± 17.0**
Free TFPI (%)95.6 ± 20.286.3 ± 18.4*
Full-length TFPI (%)100.0 ± 17.383.9 ± 22.8**
Total TFPI (%)114.9 ± 40.695.7 ± 37.2*

Plasma protein S, TFPIFL, free TFPI and total TFPI (which comprises all forms of TFPI that contain Kunitz-2) were also measured in the 85 healthy individuals (Table 2). Total and free protein S levels were significantly lower in females than in males. In addition, females had significantly lower levels of TFPIFL, free TFPI and total TFPI than males (Table 2).

TFPIFL levels were highly correlated with free TFPI levels (r = 0.834, Fig. 5A), but not with total TFPI levels (Fig. 5B). In addition, a strong correlation was found between the levels of TFPIFL and those of total protein S (r = 0.529, P < 0.001) (Fig. 5C) and free protein S (r = 0.597, P < 0.001) (data not shown). Interestingly, no correlations were found between the TFPI cofactor activity of protein S (protein S ratio) and free (r = 0.201) or total protein S (r = 0.105) levels (Fig. 5D). However, the protein S ratio showed an inverse correlation (r = − 0.479, P < 0.001) with TFPIFL levels (Fig. 5E), although not with total TFPI levels (data not shown). Finally, the TFPI ratio showed a significant correlation with TFPIFL levels (r = − 0.663, P < 0.001) (Fig. 5E), indicating that the plasma level of TFPIFL is an important determinant of the activity of the TFPI–protein S system.

image

Figure 5.  Correlations between coagulation parameters. Correlations between free tissue factor pathway inhibitor (TFPI) and full-length TFPI levels (A), total TFPI levels (determined with the commercial chromogenic assay) and full-length TFPI levels (B), total protein S and full-length TFPI levels (C), protein S ratio and total protein S levels (D), protein S ratio and full-length TFPI levels (E) and TFPI ratio and full-length TFPI levels (F) in plasma from 85 healthy individuals, comprising 48 males (•) and 37 females (○).

Download figure to PowerPoint

TFPI–protein S activity in plasma from protein S-deficient individuals

Protein S and TFPI ratios were also determined in 18 protein S-deficient individuals (total protein S, 53.4% ± 10.4%; free protein S, 38.5% ± 9.7%), of whom seven had experienced thrombosis, and in 17 normal family members (total protein S, 91.8% ± 13.8%; free protein S, 86.0% ± 15.5%), of whom two had experienced thrombosis (Table 3). Peak heights of thrombin generation without antibodies were significantly higher in protein S-deficient individuals (119.4 ± 33.6 nm) than in normal family members (67.5 ± 28.3 nm) (Fig. 6A). When anti-protein S antibodies were added, the peak heights increased both in protein S-deficient individuals (149.5 ± 30.7 nm) and in normal family members (131.2 ± 27.9 nm) (Fig. 6B). Although the average of thrombin peak heights was still higher in the protein S-deficient individuals than in normal family members, the difference between the groups was no longer significant, suggesting that protein S levels explain at least part of the difference in thrombin generation between protein S-deficient and normal individuals (Fig. 6A). After the addition of anti-TFPI antibodies, peak heights increased to 188.5 ± 31.3 nm and 193.2 ± 18.5 nm for protein S-deficient patients and normal family members, respectively (Fig. 6C). The fact that, in the presence of anti-TFPI antibodies, the thrombin peak heights became virtually identical suggested that TFPIFL levels might be different between protein S-deficient individuals and their normal family members. Indeed, it was found that free TFPI and TFPIFL levels were significantly lower in protein S-deficient individuals (free TFPI, 55.9% ± 17.3%; TFPIFL, 67.1% ± 19.3%) than in their normal family members (free TFPI, 89.6% ± 23.4%; TFPIFL, 104.3% ± 32.0%) (Table 3).

Table 3.   Tissue factor pathway inhibitor (TFPI)–protein S parameters in families with protein S deficiency
 Normal family members (n = 17)Protein S-deficient individuals (n = 18)
  1. Data are expressed as mean ± standard deviation. **P < 0.001.

Peak height (nm)67.5 ± 28.3119.4 ± 33.6**
Peak height + α-protein S (nm)131.2 ± 27.9149.5 ± 30.7
Peak height + α0TFPI (nm)193.2 ± 18.5188.5 ± 31.3
Protein S ratio0.50 ± 0.120.79 ± 0.11**
TFPI ratio0.35 ± 0.130.63 ± 0.12**
Free protein S (%)86.0 ± 15.538.5 ± 9.7**
Total protein S (%)91.8 ± 13.853.4 ± 10.4**
Free TFPI (%)89.6 ± 23.455.9 ± 17.3**
Full-length TFPI (%)104.3 ± 32.067.1 ± 19.3**
image

Figure 6.  Thrombin generation in protein S-deficient individuals and normal family members. Peak heights of thrombin generation determined in the absence of antibodies (A), in the presence of anti-protein S antibodies (αPS) (B), or in the presence of anti-tissue factor pathway inhibitor antibodies (αTFPI) (C) in plasma from 18 protein S-deficient (PS-def.) individuals (○) and 17 normal family members (•).

Download figure to PowerPoint

Both the protein S ratio and TFPI ratio were higher in the protein S-deficient individuals than in their normal family members (Table 3). These differences are probably caused by a combination of decreased levels of protein S and TFPI in the protein S-deficient individuals (Table 3). Owing to the small population size, no conclusive relationships between protein S and TFPI ratios and thrombotic events could be established.

In contrast to the population of normal individuals, and to the population of normal family members, a correlation between the protein S ratio and protein S levels was observed (r = − 0.527) in the protein S-deficient individuals (Fig. 7A). Interestingly, TFPIFL levels were significantly decreased in plasma of protein S-deficient patients and correlated with protein S levels in both protein S-deficient individuals and their family members (Fig. 7B).

image

Figure 7.  Protein S and full-length tissue factor pathway inhibitor (TFPI) levels in protein S-deficient individuals and normal family members. (A) Correlation between total protein S levels and the thrombin generation-based protein S ratio in plasma from 18 protein S-deficient individuals (△) and 17 normal family members (•). (B) Correlation between total protein S and full-length TFPI levels in plasma from 18 protein S-deficient individuals (△) and 17 normal family members (•).

Download figure to PowerPoint

Discussion

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

Using commercially available reagents, we developed thrombin generation-based assays to quantify the activity of the TFPI–protein S system. Thrombin generation was initiated with a low trigger (1.5 pm TF), because the ability of the TFPI–protein S system to inhibit the extrinsic coagulation pathway depends on the TF concentration [8,9]. The TFPI cofactor activity of protein S and the overall anticoagulant activity of the TFPI–protein S system were expressed as ratios of peak heights of thrombin generation curves determined without and with antibodies against protein S or TFPI. In normal plasma, the protein S ratio was approximately 0.5, meaning that, under the experimental conditions used, protein S reduced thrombin generation by 50% via its ability to act as a cofactor of TFPI. The TFPI ratio was approximately 0.25, meaning that the TFPI–protein S pathway reduced thrombin generation by approximately 75%. Thrombin generation parameters and derived ratios showed intra-assay and interassay CVs of 7% or lower. The differences between NPP and GKP in the absolute peak heights of thrombin generation are most likely due to differences in citrate concentration (3.2% and 3.8%, respectively), but can also depend on the number of individuals contributing to the pool (n = 98 and n = 20, respectively) and/or to plasma handling. Nevertheless, protein S and TFPI–protein S activities, expressed as ratios of peak heights, showed only minor differences between NPP and GKP, making the described assay suitable for different plasma preparations.

An important feature of this assay is the TF concentration used for initiation. High TF concentrations lead to high FXa concentrations that are beyond regulation by the TFPI–protein S anticoagulant pathway. In our hands, a TF concentration of approximately 1.5 pm was optimal for detection of the activity of the TFPI–protein S system. As no commercial source with this TF concentration was available, we mixed two commercial reagents to obtain a final concentration of 1.5 pm TF and 4 μm of phospholipids. As thrombin generation assays will vary slightly in different laboratory settings, the best strategy for obtaining optimal sensitivity for the TFPI–protein S system is to use a TF concentration that yields a peak height of thrombin generation between 30 and 40 nm in the absence of antibodies. Under these conditions, the assay is optimally sensitive to the TFPI–protein S anticoagulant pathway. As thrombin generation is relatively low and slow under these conditions, it is recommended to add CTI to avoid any unwanted contribution of contact activation [26].

The thrombin generation-based TFPI–protein S activity assays were applied to plasma from 85 healthy individuals, and the findings were compared with the antigen levels of protein S and TFPI. In normal individuals, the TFPI cofactor activity of protein S (protein S ratio) correlated with TFPIFL levels, but not with protein S levels. The absence of a correlation between protein S ratio and protein S level in normal individuals might be due to the fact that normal plasma protein S levels are not limiting for the TFPI cofactor activity.

The activity of the TFPI–protein S system was also measured in plasma from individuals with heterozygous protein S deficiency and their normal family members. The average peak heights of thrombin generation in the plasma samples from Italy were, like those in GKP, higher than those observed in NPP and the normal population, which may be attributed to their higher citrate concentration (3.8%). However, as with GKP, the protein S and TFPI ratios in the normal family members were similar to those of NPP (approximately 0.5 and approximately 0.35, respectively). Protein S-deficient individuals showed reduced activity of the TFPI–protein S system (higher protein S and TFPI ratios) when compared to the population of normal family members. The difference between the two groups largely disappeared upon addition of anti-protein S antibodies, indicating that the low protein S levels were responsible for the impairment of the TFPI–protein S pathway in the protein S-deficient individuals. The remaining difference in peak height between protein S-deficient and normal individuals determined in the presence of anti-protein S antibodies was explained by TFPIFL levels, which were lower in protein S-deficient individuals than in their normal family members. (Free) TFPI levels are known to covary with protein S levels (Fig. 5D) [27,28], and a decrease in TFPI levels in conjunction with protein S deficiency exacerbates the hypercoagulable phenotype [28].

In conclusion, we have developed functional assays for the TFPI–protein S system as well as an ELISA for TFPIFL, the form of TFPI that is active in the TFPI–protein S anticoagulant pathway. These assays pave the way for a better characterization of the role of protein S in the TFPI anticoagulant pathway in the near future. Our data suggest that the activity of the TFPI–protein S anticoagulant system is impaired in protein S deficiency, and this probably contributes to the increased risk of venous thrombosis of protein S-deficient individuals [18,29].

Acknowledgements

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

This work was supported by VIDI grants (no. 917-36-372 to T. M. Hackeng and no. 917-76-312 to E. Castoldi) from the Dutch Organization for Scientific Research (NWO).

Disclosure of Conflict of Interests

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

The authors state that they have no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  • 1
    DiScipio RG, Davie EW. Characterization of protein S, a gamma-carboxyglutamic acid containing protein from bovine and human plasma. Biochemistry 1979; 18: 899904.
  • 2
    Walker FJ. Regulation of activated protein C by a new protein. A possible function for bovine protein S. J Biol Chem 1980; 255: 55214.
  • 3
    Koedam JA, Meijers JC, Sixma JJ, Bouma BN. Inactivation of human factor VIII by activated protein C. Cofactor activity of protein S and protective effect of von Willebrand factor. J Clin Invest 1988; 82: 123643.
  • 4
    Heeb MJ, Mesters RM, Tans G, Rosing J, Griffin JH. Binding of protein S to factor Va associated with inhibition of prothrombinase that is independent of activated protein C. J Biol Chem 1993; 268: 28727.
  • 5
    Hackeng TM, Van’t Veer C, Meijers JC, Bouma BN. Human protein S inhibits prothrombinase complex activity on endothelial cells and platelets via direct interactions with factors Va and Xa. J Biol Chem 1994; 269: 210518.
  • 6
    Seré KM, Janssen MP, Willems GM, Tans G, Rosing J, Hackeng TM. Purified protein S contains multimeric forms with increased APC-independent anticoagulant activity. Biochemistry 2001; 40: 885260.
  • 7
    Koenen RR, Christella M, Thomassen LG, Tans G, Rosing J, Hackeng TM. Effect of oral contraceptives on the anticoagulant activity of protein S in plasma. Thromb Haemost 2005; 93: 8539.
  • 8
    Seré KM, Rosing J, Hackeng TM. Inhibition of thrombin generation by protein S at low procoagulant stimuli: implications for maintenance of the hemostatic balance. Blood 2004; 104: 362430.
  • 9
    Hackeng TM, Seré 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.
  • 10
    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.
  • 11
    Castoldi E, Hackeng TM. Regulation of coagulation by protein S. Curr Opin Hematol 2008; 15: 52936.
  • 12
    Hackeng TM, Maurissen LF, Castoldi E, Rosing J. Regulation of TFPI function by protein S. J Thromb Haemost 2009; 7(Suppl. 1): 1658.
  • 13
    Hackeng TM, Rosing J. Protein S as cofactor for TFPI. Arterioscler Thromb Vasc Biol 2009; 29: 201520.
  • 14
    Novotny WF, Brown SG, Miletich JP, Rader DJ, Broze GJ Jr. Plasma antigen levels of the lipoprotein-associated coagulation inhibitor in patient samples. Blood 1991; 78: 38793.
  • 15
    Novotny WF, Girard TJ, Miletich JP, Broze GJ Jr. Purification and characterization of the lipoprotein-associated coagulation inhibitor from human plasma. J Biol Chem 1989; 264: 188327.
  • 16
    Broze GJ Jr, Lange GW, Duffin KL, MacPhail L. Heterogeneity of plasma tissue factor pathway inhibitor. Blood Coagul Fibrinolysis 1994; 5: 5519.
  • 17
    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.
  • 18
    Schwarz HP, Fischer M, Hopmeier P, Batard MA, Griffin JH. Plasma protein S deficiency in familial thrombotic disease. Blood 1984; 64: 1297300.
  • 19
    Dahm A, Van Hylckama Vlieg A, Bendz B, Rosendaal F, Bertina RM, Sandset PM. Low levels of tissue factor pathway inhibitor (TFPI) increase the risk of venous thrombosis. Blood 2003; 101: 438792.
  • 20
    Nordfang O, Bjorn SE, Valentin S, Nielsen LS, Wildgoose P, Beck TC, Hedner U. The C-terminus of tissue factor pathway inhibitor is essential to its anticoagulant activity. Biochemistry 1991; 30: 103716.
  • 21
    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.
  • 22
    Hemker HC, Giesen P, AlDieri R, Regnault V, De Smed E, Wagenvoord R, Lecompte T, Béguin S. The calibrated automated thrombogram (CAT): a universal routine test for hyper- and hypocoagulability. Pathophysiol Haemost Thromb 2002; 32: 24953.
  • 23
    Duchemin J, Pan-Petesch B, Arnaud B, Blouch MT, Abgrall JF. Influ-ence of coagulation factors and tissue factor concentration on the throm-bin generation test in plasma. Thromb Haemost 2008; 99: 76773.
  • 24
    Bendz B, Andersen TO, Sandset PM. A new sensitive chromogenic substrate assay of tissue factor pathway inhibitor type 1. Thromb Res 2000; 97: 46372.
  • 25
    Schnolzer M, Alewood P, Jones A, Alewood D, Kent SB. In situ neutralization in Boc-chemistry solid phase peptide synthesis. Rapid, high yield assembly of difficult sequences. Int J Pept Protein Res 1992; 40: 18093.
  • 26
    Mann KG, Whelihan MF, Butenas S, Orfeo T. Citrate anticoagulation and the dynamics of thrombin generation. J Thromb Haemost 2007; 5: 205561.
  • 27
    Dahm AE, Sandset PM, Rosendaal FR. The association between protein S levels and anticoagulant activity of tissue factor pathway inhibitor type 1. J Thromb Haemost 2008; 6: 3935.
  • 28
    Castoldi E, Simioni P, Tormene D, Rosing J, Hackeng TM. Hereditary and acquired protein S deficiencies are associated with low TFPI levels in plasma. J Thromb Haemost 2010; 8: 294300.
  • 29
    Makris M, Leach M, Beauchamp NJ, Daly ME, Cooper PC, Hampton KK, Bayliss P, Peake IR, Miller GJ, Preston FE. Genetic analysis, phenotypic diagnosis, and risk of venous thrombosis in families with inherited deficiencies of protein S. Blood 2000; 95: 193541.