Journal of Thrombosis and Thrombolysis

, Volume 30, Issue 4, pp 412–418

Differential inhibitory effect of fondaparinux on the procoagulant potential of intact monocytes and monocyte-derived microparticles


  • Sonia Ben-Hadj-Khalifa-Kechiche
    • Laboratoire d’HématologieCHU Robert Debré
    • Faculté de PharmacieUnité de Recherche des Maladies Hématologiques et Auto-Immunes
  • Nathalie Hezard
    • Laboratoire d’HématologieCHU Robert Debré
    • EA-3801, Faculté de MédecineUniversité Reims-Champagne Ardenne
  • Stephane Poitevin
    • Faculté de PharmacieINSERM UMR-S 608, Physiopathologie de l’Endothélium
  • Marie-Geneviève Remy
    • Laboratoire d’HématologieCHU Robert Debré
  • Bernadette Florent
    • Laboratoire d’HématologieCHU Robert Debré
  • Touhami Mahjoub
    • Faculté de PharmacieUnité de Recherche des Maladies Hématologiques et Auto-Immunes
    • Laboratoire d’HématologieCHU Robert Debré
    • EA-3801, Faculté de MédecineUniversité Reims-Champagne Ardenne

DOI: 10.1007/s11239-010-0490-4

Cite this article as:
Ben-Hadj-Khalifa-Kechiche, S., Hezard, N., Poitevin, S. et al. J Thromb Thrombolysis (2010) 30: 412. doi:10.1007/s11239-010-0490-4


Monocytes and monocyte-derived microparticles (MMPs) play a major role in acute coronary syndrome (ASC). Activated monocytes (ac-M) and MMPs support thrombin generation via tissue factor (TF). The aim of this study was to evaluate the inhibitory effect of fondaparinux, a selective Xa inhibitor, on thrombin generation supported by activated monocytes and MMPs. Monocytes were purified by elutriation. They were activated by LPS, allowing to obtain both ac-M and MMPs. Thrombin generation was performed using Fluoroscan® in these two cell models, in comparison with a cell-free model (TF 5 pM final). Two concentrations of ac-M (0.2 × 106 and 1 × 106/well) and four concentrations of MMPs (40,000; 80,000; 120,000 and 160,000/well) were tested. TGT was evaluated for increasing fondaparinux concentrations (0, 0.1, 0.4, 0.7 and 1.2 μg/ml). Without fondaparinux, 0.2 × 106 ac-M and 160,000 MMPs induced comparable results. Fondaparinux inhibited thrombin generation in the three models. Inhibition was fondaparinux concentration dependent. Rate index was the most sensitive parameter, compared to lag-time, peak and endogenous thrombin potential. The rate index IC50 were 0.69 ± 0.03 μg/ml for ac-M, 0.20 ± 0.03 μg/ml for MMPs, and 0.22 ± 0.02 μg/ml for cell-free model. Fondaparinux exerted an inhibitory effect at all concentrations, including the lowest (0.1 μg/ml). The extend of inhibition was similar between MMPs and cell-free models, and stronger than ac-M model. We assume that the efficacy of fondaparinux 2.5 mg once daily in ACS patients may be in part attributed to its inhibitory effect on MMPs.


FondaparinuxThrombin generationMonocyteMonocyte-derived microparticleTissue factor


Monocytes are actively involved in the physio-pathological process related to acute coronary syndrome (ACS) [1]. Thus, circulating monocytes are significantly increased in such patients, and they are recruited after activation directly in the arterial wall [1, 2]. They are involved in different steps of atherosclerosic plaque formation, until its disruption [1]. Activated monocytes express and/or release mediators involved in plaque destabilization, i.e., cytokines, metalloproteases, reactive oxygen species, and tissue factor (TF) [1, 3, 4].

Moreover, monocytes are able to generate microparticles. These monocyte-derived microparticles (MMPs) are elevated in acute clinical setting compared to stable angina patients or healthy volunteers [5]. MMPs represent with lymphocyte-derived microparticles the major shed membrane apoptotic microparticles produced in human atherosclerosis [6, 7]. MMPs express a high level of TF, and it is well known that TF is crucial in atherosclerosis [6]. Indeed, MMPs may be considered as the major key of the procoagulant activity of atherosclerotic plaque. At the time of plaque rupture, these MMPs will get in contact with plasma coagulation factors, trigger thrombin generation, thrombus formation and accelerate vessel occlusion that could lead to acute infarction [6, 8]. Taken together, an effective inhibition of thrombin generation supported by MMPs and/or by activated monocytes may be of interest in ACS. Different antithrombotic drugs are indicated in ACS, including the recently developed fondaparinux. Fondaparinux is a synthetic pentasaccharide that selectively inhibits factor Xa [9]. Its efficacy in ACS has been reported in clinical trials for 2.5 mg once daily [10, 11]. The aim of this study was to evaluate in in vitro models the inhibitory effect of fondaparinux on thrombin generation supported by activated monocytes and MMPs.

Materials and methods

Platelet poor plasma preparation

Venous blood samples were obtained from 5 healthy volunteers (3 men and 2 women; mean age 25 ± 2 years). Volunteers were laboratory staff members; they had been free of any medication for the last 2 weeks. Blood was withdrawn by antecubital venipuncture into Monovette® tubes (0.106 M citrate, Sarstedt, Germany). A three-step centrifugation procedure was used: 10 min at 190 g, 10 min at 1,750 g, and 30 min at 13,000 g. The five platelet poor plasma (PPP) supernatants were pooled. Aliquots of PPP (1 ml) were stored at -80°C. PPP was thawed at 37°C for 5 min before use.

Monocyte purification

Cytapheresis residues were obtained from 6 healthy volunteers. Volunteers were admitted for platelet donation in the blood transfusion unit of CHU Robert Debré. Informed consents were obtained from all of them. Monocytes were purified from cytapheresis residues by elutriation as previously described [12]. Purity was assessed by staining CD14-positive cells (>95%). Cell viability was measured by trypan blue exclusion principle (>98%). Purified monocytes obtained from each donor were used for preparation of both activated monocytes (ac-M) and MMPs.

Monocyte activation

Purified monocytes were washed in RPMI-1640 medium (Invitrogen, Cergy Pontoise, France), and then adjusted to 1.5 × 106 cells/ml in RPMI-1640 containing 5% of heat-inactivated foetal calf serum (Invitrogen, Illkirsch, France), 2 mM l-glutamine (Sigma, Saint Quentin Fallavier, France) and 100 ng/ml of lipopolysaccharide (LPS) (Escherichia coli serotype O55:B5, Sigma). Monocytes were incubated for 5 h at 37°C in a 5% CO2 humidified atmosphere. Supernatant were removed after a centrifugation of 400 g for 5 min. The pellet corresponding to ac-M was resuspended in 150 μl of phosphate-buffered saline (PBS) (BioMérieux®, Crayonne, France).

MMPs preparation and quantification

Monocytes were incubated as described above, with an incubation duration of 18 h. Supernatants were then collected after a centrifugation of 2,200 g for 5 min. MMPs were obtained using an additional centrifugation of 17,000 g for 30 min, as previously described [13]. MMPs were quantified using flow cytometry. MMPs (100 μl) were incubated for 15 min in the dark with 1 μl of annexin V-FITC (AV-FITC, Biovision, Clinisciences, Montrouge, France), according to the manufacturer’s recommendations. Then, 400 μl of annexin V binding buffer (Biovision) and 100 μl of flow-count Fluorospheres (Beckman Coulter, France) were added. Fluorescence was acquired for 60 s on a Epics XL-MCL® flow cytometer (Beckman Coulter, France) using System 2 software. MMPs were quantified using the flow-count Fluorospheres, and expressed as the number of MMPs per μl.

Fluorogenic measurement of thrombin generation

Thrombin generation test (TGT) was performed according to the assay described by Hemker, and modified by Poitevin [13, 14]. We defined ac-M and MMPs as cell models. Thrombin generation was performed for these two cell models, in comparison with a cell-free model. PPP was used for all experimental conditions, and was systematically supplemented with aprotinin (Sigma) at the final concentration of 200 kallikrein inhibitory units (KUI)/Ml. Moreover, PPP was spiked with increasing concentrations of fondaparinux: 0; 0.1; 0.4; 0.7 and 1.2 μg/ml. For cell models, 20 μl of each concentration of ac-M or MMPs were mixed with 80 μl of PPP in each microplate well. Two concentrations of ac-M (0.2 × 106 and 1.0 × 106 ac-M/well) and 4 concentrations of MMPs (40,000; 80,000; 120,000 and 160,000 MMPs/well) were tested. For the cell-free model, 20 μl of the mixture containing 5 pM of TF and 4 μM of synthetic phospholipids (PPP-low, Biodis, France) were mixed with 80 μl of PPP.

The fluorometric measurements of thrombin generation were performed on Fluoroskan® Ascent plate reader (ThermoLabsystems, Helsinki, Finland), following automated addition of the fluorogenic substrate Z-Gly-Gly-Arg-AMC (Bachem, Bubendorph, Switzerland). ThrombinoscopeTM software was used for the calculation of thrombin generation (Synapse BV, Maastricht, The Netherlands). Four parameters of the thrombin generation curve were analyzed: (i) lag-time (min), (ii) thrombin peak (peak, nM), (iii) endogenous thrombin potential (ETP, nM min) and the rate index of propagation phase calculated by the formula peak/(time to peak − lag-time) (rate index, nM min−1).

Statistical analysis

Data distribution was analyzed with Shapiro–Wilk test. For statistical comparisons, Student’s t test and Mann–Whitney test were performed. P value < 0.05 were considered as statistically significant.


Thrombin generation induced in cell and cell-free models

In the ac-M model, two concentrations were tested: 0.2 × 106 and 1 × 106 ac-M per well. Thrombin generation was observed in a ac-M concentration-dependent manner. At the concentration of 0.2 × 106 versus 1 × 106 ac-M, lag-time values were 2.1 ± 0.1 min versus 1.1 ± 0.2 min. Peak values were 181 ± 6 nM versus 311 ± 11 nM. ETP was 1,624 ± 84 versus 1,695 ± 49 nM min and the rate index was 59 ± 4 versus 159 ± 7 nM min−1.

In the MMPs model, 4 concentrations were compared: 40,000; 80,000; 120,000 and 160,000 MMPs per well. As ac-M, MMPs supported thrombin generation in a concentration-dependent manner. Referring to these concentrations, lag-time values were 5.1 ± 0.6, 4.0 ± 0.5, 3.4 ± 0.4 and 2.7 ± 0.3 min, peak values were 98 ± 10, 150 ± 14, 186 ± 16 and 195 ± 19 nM, ETP values were 1507 ± 90, 1865 ± 116, 1912 ± 138 and 1851 ± 149 nM min, and rate index values were 20 ± 4, 39 ± 8, 53 ± 12 and 56 ± 14 nM min−1.

Comparison between ac-M and MMPs showed statistically similar values for the four parameters between 0.2 × 106 ac-M per well and 160,000 MMPs per well. These two concentrations were chosen for further fondaparinux assays.

At last, the cell-free model showed lag-time, peak, ETP and rate index of 2.5 ± 0.1 min, 362 ± 21 nM, 1902 ± 91 nM min and 165 ± 18 nM min−1, respectively.

Effect of fondaparinux on thrombin generation supported by ac-M

The effect of fondaparinux was studied only for the concentration of 0.2 × 106 ac-M per well. At each tested concentration (ranging from 0.1 to 1.2 μg/ml), fondaparinux significantly modified all thrombin generation parameters (expect ETP at 0.1 μg/ml of fondaparinux) (Fig. 1b). Moreover, fondaparinux-induced inhibition was observed in a concentration-dependent manner (Fig. 1). The maximal percentages of thrombin generation inhibition displayed from 10 to 55%, with a maximal magnitude of inhibition for the rate index (Fig. 1b).
Fig. 1

Effect of increasing concentrations of fondaparinux (0, 0.1, 0.4, 0.7, 1.2 mg/ml) on thrombin generation induced by 0.2 × 10acM. a A representative thrombogram. b Data obtained in 6 experiments represented as means ± SEM

Effect of fondaparinux on thrombin generation supported by MMPs

The effect of fondaparinux was studied only for the concentration of 160,000 MMPs per well. As for ac-M, fondaparinux was significantly effective for all concentrations, including the lowest (0.1 μg/ml), on all thrombin generation parameters (expect ETP at 0.1 μg/ml of fondaparinux). Fondaparinux modified also thrombin generation in a concentration-dependent manner (except lag-time) (Fig. 2a). The maximal inhibitory effect of fondaparinux on MMPs-induced thrombin generation ranged from 55 to 85%. The rate index was the most sensitive parameter (Fig. 2b).
Fig. 2

Effect of increasing concentrations of fondaparinux (0, 0.1, 0.4, 0.7, 1.2 mg/ml) on thrombin generation induced by 160,000 MMPs. a A representative thrombogram. b Data obtained in 6 experiments represented as means ± SEM

Effect of fondaparinux in a cell-free thrombin generation assay

Fondaparinux affected significantly all thrombin generation parameters from the lowest to the highest tested concentrations (from 0.1 to 1.2 μg/ml). All parameters were modified in a concentration-dependent manner (Fig. 3a). Maximal fondaparinux inhibition of thrombin generation parameters displayed from 25 to 90%. The rate index was the most affected by fondaparinux (Fig. 3b).
Fig. 3

Effect of increasing concentrations of fondaparinux (0, 0.1, 0.4, 0.7, 1.2 mg/ml) on thrombin generation in cell-free model induced by TF 5 pM final. a A representative thrombogram. b Data obtained in 6 experiments represented as means ± SEM

Calculation of IC50

The inhibitory effect of fondaparinux on each thrombin generation parameter was evaluated by IC50 calculation. For peak, ETP and rate index, the IC50 were considered as the fondaparinux concentration allowing a decrease of 50%. For the lag-time, IC50 was calculated as the fondaparinux concentration leading to double this parameter.

The rate index was the only parameter for which IC50 was calculated in the three models. For MMPs and cell-free models, IC50 values were statistically similar (0.20 ± 0.03 and 0.22 ± 0.02 μg/ml, respectively) and 3 times higher compared to rate index IC50 values obtained for ac-M (0.69 ± 0.03 μg/ml) (Fig. 4).
Fig. 4

IC50 obtained in the three models. Histograms represent mean values ± SEM of six experiments. * P < 0.05, NS non significant, ND not determined

On the contrary, IC50 was not systematically calculated for the other parameters. The lag-time IC50 was not reached in any model. ETP IC50 could be calculated only in the MMP model (0.63 ± 0.16 μg/ml). Peak IC50 could be calculated only for both MMPs and cell-free models with statistically comparable values (0.47 ± 0.1 and 0.62 ± 0.1 μg/ml, respectively) (Fig. 4).


Thrombin generation test (TGT) has been proposed to study behaviour and mechanism of action of new anticoagulants, including fondaparinux, a selective FXa inhibitor acting via antithrombin [1517]. In these in vitro studies, TGT was realized in platelet poor plasma (PPP) model, or in platelet rich plasma (PRP) model as the source of phospholipids [1517]. Compared to the current tests such as anti-Xa assay and activated partial thromboplastin time (APTT), TGT is more physiological since it reflects the overall function of the blood clotting process including TF-triggering, amplification and antithrombin-dependent inhibition of the thrombin generation [18]. Moreover, unlike other assays which are performed in cell-free plasma, TGT can be analyzed in the presence of cells and cell-derived MPs [13, 16].

The aim of our study was to evaluate the inhibitory effect of fondaparinux on thrombin generation induced by human monocytes and their microparticles, both playing major role in the thrombogenicity in ACS.

Here, monocyte activation and microparticle generation were induced by LPS, which has been reported extensively in the literature as the major inducer of tissue factor expression and microvesiculation by monocytes [13, 19]. Moreover, LPS plays a role in pathophysiology, including cardiovascular diseases. Indeed, it was recently reported that patients presenting with ACS were LPS-seropositive, and LPS may be the strongest inducer of procalcitonin (PCT) release, a biomarker of ACS [20].

Our findings show that ac-M and MMPs are able to induce thrombin generation. This observation is in line with previous results showing that ac-M and MMPs supported thrombin generation via the expression of high level of TF [13]. Here, we report that the magnitude of induced thrombin generation was directly bound to the amount of cells or microparticles added in the system. Interestingly, we observed that 0.20 × 106 ac-M per test led to a similar amount of thrombin generation than 160,000 MMPs (0.16 × 106) per test, despite the considerable difference of size between intact monocytes (~30 μm) and MMPs (defined as below 1 μm). This suggests that MMP surface has a higher procoagulant activity than that of intact ac-M. We could suppose that the density of TF at the surface of MMPs is higher than on intact activated-monocytes [21, 22]. We also could assume that the mechanism of decryption is related with MMP formation and associated with an increase of procoagulant activity [23]. On the other hand, we could propose that MMPs by expressing a higher density of phosphatidyl serine (PS) are more procoagulant than ac-M. Indeed it was described that the complex TF-VIIa is more active when the local PS content is higher [24]. Moreover, higher level of PS, by facilitating assembly of the prothrombinase complex, contributes to a higher level of thrombin generation [25].

Our results show that the inhibitory effect of fondaparinux is different on ac-M-induced thrombin generation compared to MMPs-induced thrombin generation. We first compared the different quantitative parameters measured from the analysis of the TGT curves. The rate index was the most sensitive to fondaparinux in cell-free model. This is in agreement with results obtained in PPP or PRP showing that fondaparinux acts mainly by limiting the propagation phase via decreasing the rate index [1517]. Moreover, our results show a calculated rate index IC50 of 0.22 ± 0.02 μg/ml. This value is comparable to previous reported data by Robert et al. [17] (IC50 of 140 nM, corresponding to 0.23 μg/ml). Considering the rate index IC50, fondaparinux induced a stronger thrombin generation inhibition in MMPs model compared to ac-M model. Additionally, the extent of MMPs-thrombin generation inhibition was comparable to the extent of inhibition observed in cell-free model.

Our findings suggest that (1) the efficacy of the low fondaparinux dosage (2.5 mg daily) in ACS may be at least in part mediated by the inhibition of MMPs procoagulant activity, and (2) the differential effect of fondaparinux between ac-M and MMPs may be attributed to a different membrane environment. Considering the first point, we showed the efficacy of low fondaparinux concentration in both MMPs- and cell-free models. The phase II PENTUA clinical trial reported two unexpected data: the efficacy of the lowest dosage of 2.5 mg of fondaparinux, and no increase of clinical efficacy despite increased fondaparinux dosages (displayed from 2.5 to 12 mg once daily) [26]. These results are intriguing, because (1) in ACS, fondaparinux was compared to enoxaparin high dose (1 mg/kg daily), and (2) fondaparinux 2.5 mg once a day is the dosage corresponding to prophylaxis in deep venous thrombosis (DVT), the active treatment of DVT requiring a higher dosage (7.5 mg once a day) [2628]. Moreover, the activity anti-Xa substudy of OASIS-5 showed that fondaparinux 2.5 mg daily compared with enoxaparin 1 mg/kg twice daily produced a lower anticoagulant intensity level [29]. Anti-Xa levels were 0.52 IU ml−1 in patients assigned to fondaparinux versus 1.2 IU ml−1 in patients assigned with enoxaparin [29]. Our second observation is the differential effect of fondaparinux in our two cell models. We assume that membrane environment, in which thrombin is generated, may be involved and may affect the accessibility of factor Xa to antithrombin, via, as previously discussed (1) TF density, (2) TF decryption, and (3) PS density.

In conclusion, fondaparinux exerts in vitro a thrombin generation inhibitory effect in a monocyte models. Moreover, a superior inhibitory effect was observed for monocyte-derived microparticles compared with intact monocytes. Interestingly, the inhibition of thrombin generation induced by MMPs was observed at low concentration of fondaparinux.


The authors are grateful for the excellent laboratory assistance from G. Simon. We also thank C. Mace for her expertise in statistics.

Disclosure and conflict of interests statement

The authors state they have no conflict of interest.

Copyright information

© Springer Science+Business Media, LLC 2010