Summary. Heparin-induced thrombocytopenia (HIT) is a serious secondary event encountered in the clinical use of heparin. HIT results from the consumption of platelets that are immunologically activated by antibodies directed against complexes formed by platelet factor 4 (PF4) and sulfated polysaccharides that activate platelet aggregation, leading to paradoxical, life-threatening thrombosis. There is strong evidence that the ability of heparin and related compounds to induce HIT is closely linked to the structure of the polysaccharide, and particularly to its negative charge and to the length of the molecule. To test this hypothesis, we synthesized two sulfated oligosaccharides: SanOrg123781, a 16-mer, presenting two terminal charged domains separated by a 7-mer neutral linker, and SR121903, a highly sulfated 17-mer. Both of them displayed strong anti-factor (F) Xa and anti-FIIa activities but their affinities for PF4 were markedly different. SR121903 displaced PF4-bound heparin, whereas SanOrg123781 did not, underlining the importance of the charge of the molecule for the interaction with PF4. Platelet studies, in the presence of HIT serum, showed that SR121903 induced the secretion of platelet-dense granules (measured by the release of serotonin) whereas SanOrg123781 did not, a result in accordance with an absence of affinity of this molecule for PF4. These results were confirmed by measurements of platelet activation by flow cytometry (measured by annexin V binding, CD62 detection and activation of the GpIIb–IIIa complexes). In conclusion, we have demonstrated the importance of the charge of the polysaccharides in the HIT-induced platelet reactions measured by diverse methods, of which some are described for this purpose for the first time.
Heparin-induced thrombocytopenia (HIT) is a serious side-effect of the clinical use of heparin. About 5% of patients receiving unfractionated heparin after orthopedic surgery are affected by this life-threatening complication [1,2]. Platelet activation in HIT patients is known to be dependent on the presence of antibodies against heparin–platelet factor 4 (PF4) complexes. Studies have shown that the negatively charged polysaccharide chains of heparin bind strongly to PF4 in a non-specific manner, displacing it from endothelial glycosaminoglycans. The PF4–heparin complex can generate antibodies that in turn bind to the PF4–heparin complex to generate immune complexes, which are then recognized by the platelet FcγRIIA receptors, leading to platelet activation in an adenosine diphosphate (ADP)-dependent process as previously described . However, although many patients treated with heparin develop antibodies, only some of them are affected by HIT. The platelet activation associated with HIT seems to be dependent on a subset of these antibodies, and the qualitative evaluation of the activating properties is more relevant than a quantitative level . Several methods have been used for monitoring platelet activation. Platelet aggregation  and release of serotonin  are the methods currently used for determining the presence of HIT antibodies in sera, and these methods are said to provide relevant results for the diagnosis of HIT, whereas the levels of antibodies against the PF4–heparin complex are considered non-predictive. However, these methods are not easy to perform and demand a large amount of serum, which is not always possible in clinical practice.
There is strong evidence that the ability of heparin and related compounds to induce HIT is closely linked to the structure of the polysaccharide, and particularly to its negative charge and to the length of the molecule . In recent years, considerable progress has been made in the synthesis of oligosaccharides with the anticoagulant properties of heparin [7,8]. Among the new molecules available are two compounds of similar length but different charge profiles: SanOrg123781 and SR121903. SanOrg123781, a newly developed compound, is a 16-mer oligosaccharide designed to display specific antithrombin (AT)-mediated inhibition of factor (F) Xa and thrombin . SR121903, a 17-mer oligosaccharide, displays AT-mediated anti-FXa and antithrombin activities in the same range as SanOrg123781, but is totally sulfated .
In this article, we report a flow cytometry method allowing the evaluation of three different parameters of platelet activation, each of which describes a different step of platelet activation. Moreover, we have measured the comparative effects of two synthetic oligosaccharides on platelet activation induced by plasma from HIT patients.
Drugs and reagents
Heparin (sodium heparinate) was purchased from Sigma (l'Ile d'Abeau, France). The synthetic oligosaccharides SanOrg123781 and SR121903 (structures on Table 1) were synthetized by Sanofi-Synthelabo [7,8] (Toulouse, France). Sera from patients diagnosed with heparin-induced thrombocytopenia (HIT) were kindly provided by Dr J. Fareed (Loyola University Medical Center, Maywood, IL, USA). 14C-serotonin (10 µCi mg−1) was purchased from Amersham (Les Ulis, France). FITC-labeled annexin V and labeled monoclonal antibodies (mAbs): FITC-PAC1 mAb and PE-anti-CD62 mAb were purchased from Becton Dickinson (San José, CA, USA). Labeled annexin V and antibodies were diluted in a buffer containing 20 mmol L−1 HEPES, 150 mmol L−1 NaCl and 2.2 mmol L−1 CaCl2, pH 7.3. Purified α-thrombin and PF4 were provided by the Centre de Transfusion Sanguine (Strasbourg, France). Human FXa, human AT and chromogenic substrates S2222 and S2238 (used in accordance with manufacturer instructions), were purchased from Chromogenix (Mölndahl, Sweden). [3H(G)] heparin sodium salt (0.5 µCi µg−1) was obtained from Dupont NEN (Boston, MA, USA).
Table 1. Main characteristics of the oligosaccharides. Results are given as mean ± SD (n = 3). the anti-FXa and anti-FIIa activities were measured in previous studies [8-10]
Anti-FXa Activity (IC50, ng mL−1)
Anti-FIIa activity (IC50, ng ml−1)
120 ± 10
3.3 ± 0.5
80 ± 2
5.3 ± 0.2
77 ± 5
44 ± 03
The effects of standard heparin and synthetic oligosaccharides on FXa were determined by a modification of the Teien and Lie  procedure as already described . Human FXa (2.4 nkat mL−1) was incubated for 2 min with purified human AT (0.17 U mL−1) at 37 °C in the presence of various concentrations of the two compounds in Tris/maleate 20 mmol L−1 buffer pH 7.4, NaCl 150 mmol L−1 (volume: 0.2 mL). To measure the residual FXa activity, S2222 substrate (dissolved in 50 mmol L−1 Tris/HCl buffer, pH 8.4, NaCl 175 mmol L−1, EDTA 27.5 mmol L−1) was added (0.25 mmol L−1). The reaction was stopped 2 min later by the addition of a 50% aqueous acetic acid solution, and absorbance at 405 nm was read on a spectrophotometer. The percentage of inhibition was then calculated as inhibition percentage = 100 × [(OD blank–OD sample)/OD blank] and the IC50 calculated with RS/1 software (Sanofi-Synthelabo, Toulouse, France).
Anti FIIa activitiy
Purified human alpha thrombin (2 IU mL−1) was incubated for 1 min with AT (0.25 U mL−1) at 37 °C in the presence of the compounds. To measure the residual activity a 1-mmol L−1 solution of S2238 chromogenic substrate was added (volume: 0.2 mL). The reaction was stopped 1 min later by the addition of a 50% aqueous acetic acid solution, and absorbance at 405 nm was read. The percentage of inhibition was calculated as inhibition percentage = 100 × [(OD blank – OD sample)/OD blank] and the IC50 calculated with RS/1 software.
Competition with heparin binding to PF4
Experiments were performed according to Stringer and Gallagher . Briefly, 10 µL of 3H-radiolabeled heparin (200 nmol L−1 determined from saturation experiment; 0.5 µCi µg−1) was incubated with10 µL of human PF4 (1 µg) in the presence of various concentrations of heparin, SanOrg123781 or SR121903 for 10 min at 37 °C in 10 µL of Tris buffer (130 mmol L−1 NaCl, 50 mmol L−1 Tris HCl, pH 7.3). The volume was then made up to 300 µL by the addition of Tris buffer and the samples were drawn through buffer-equilibrated cellulose nitrate filters on a vacuum manifold. The filters were washed with 2 × 4 mL of buffer, and bound material eluted with 2 × 5 mL of NaCl 50 mmol L−1 Tris-HCl. The radioactivity of the eluate was counted by liquid scintillation. The competition was evaluated by the percentage of reduction of radioactivity in the presence of the tested compounds, compared with the control sample.
Serotonin release study
The experiment was performed according to Sheridan et al. . Platelet-rich plasma (PRP) was obtained by centrifugation (100 g, 5 min) of citrated human venous blood, provided by the Centre d'Investigation Clinique (Hôpital Purpan, Toulouse, France). PRP was incubated with 14C-serotonin for 45 min at 20 °C (1 µCi 14C-serotonin per ml of PRP). The platelets were washed twice with buffer (134 mmol L−1 NaCl, 12 mmol L−1 NaHCO3, 0.34 mmol L−1 Na2HPO4, 2.9 mmol L−1 KCl, 1 mmol L−1 MgCl2, 5 mmol L−1 glucose, 5 mmol L−1 Hepes, pH 7.14), and resuspended in the same volume of plasma. 75 µL of 14C-serotonin-loaded platelets, 20 µL of HIT serum and 5 µL of 0.9% NaCl containing various concentrations of the tested compounds (heparin, SanOrg123781A or SR121903A) were added in the tubes and incubated at room temperature under constant stirring for 60 min. Control samples contained 5 µL of saline in place of the tested compounds. To stop the reaction, 100 µL of 0.5% EDTA in 0.15 mol L−1 NaCl was added to each tube. The platelets were centrifuged at 2000 g for 10 min, and 14C radioactivity was measured in 50 µL of the supernatant.
Flow cytometry studies
PRP was acidified with a buffer containing: trisodium citrate dihydrate 22 mg mL−1, citric acid monohydrate 8 mg mL−1, dextrose 25 mg mL−1 (180 µL for 1 mL of PRP). PGI2 was added (12 ng mL−1 final) and platelets were isolated by centrifugation 500 × g, 10 min. The platelet pellet was then resuspended at 0.3 × 106 pl µL−1 in the same buffer as above, with the addition of 0.35% BSA, (pH 7.4).
Platelet suspension samples (140 µL) were incubated for 30 min at 28 °C, with 20 µL of the tested compounds in a 0.9% NaCl solution (which was used as the vehicle for the control samples) and 40 µL of HIT sera. Blank samples were prepared in the same way but with the tested compounds and HIT sera replaced by the same volume of buffer.
Aliquots (10 µL) were then pipeted into three different tubes containing either 4 µL of FITC-annexin V diluted in 86 µL of Hepes/Ca2+ buffer (A), 20 µL of FITC-PAC1 mAb diluted in 70 µL of Hepes/Ca2+ buffer (B), or 20 µL of PE-anti CD62 mAb diluted in 70 µL of Hepes/Ca2+ buffer (C). After 15 min of incubation (room temperature, in the dark), samples were diluted in 400 µL of fixative solution (CellFIX Becton Dickinson diluted 10-fold in water) and analyzed by flow cytometry. Analysis was performed on 5000 events, with a Becton Dickinson FacsCalibur flow cytometer. Fluorescence signal was analyzed with the CellQuest program and a threshold of positive events was set up in order to measure less than 2% of positive events in the blank samples. The results were expressed as a percentage of positive events, defined as events above the threshold. With this method, control samples, where HIT sera were replaced by sera from healthy human volunteers, did not show any activation, either in the presence of heparin or synthetic oligosaccharides (n = 3, data not shown).
Anti-FXa and anti-FIIa activities
The effect of the oligosaccharides on the coagulation factors FXa and FIIa were measured in vitro by a chromogenic test with purified human enzymes, and the results compared with those of standard heparin. The activities, expressed as the concentration required to inhibit 50% of the enzyme activity, were quite similar for all the compounds with both enzymes (Table 1).
Competition with heparin for binding to PF4
Tritiated heparin bound to platelet F4 was totally displaced by unlabeled heparin (Fig. 1). In the same way, SR121903 competed with radioactive heparin with an efficacy similar to that of standard heparin. The IC50 values of both compounds were near to 2 µg mL−1. By contrast, SanOrg123781 did not affect the binding of radioactive heparin to PF4, indicating that it did not bind to the heparin binding sites on PF4. This result is in accordance with previously published information on this compound .
HIT-induced serotonin release
When tested in the presence of serum from HIT patients, standard heparin stimulated the release of 14C-serotonin from human platelets (Fig. 2). The concentration/effect relationship displayed a bell-shaped curve, as already described with this test , peaking between 0.3 and 3 µg mL−1 of heparin. Under the same experimental conditions, SR121903 was found to behave like heparin and induced the release of serotonin from platelets in the presence of serum from HIT patients, with a comparable efficacy. However, SanOrg12378A failed to stimulate platelets in the presence of HIT serum, as already described .
In parallel to this ‘classical’ method of measurement, we tested a new cytometric method, indicative for platelet procoagulant activity stimulation, release reaction and Gp IIb–IIIa activation.
HIT-induced PC–PS exchange
Annexin V is currently used as a probe for measuring the exposure of electronegative phospholipids on membrane during cell apoptosis , microparticles production or platelet aggregation . Annexin V is a positively charged protein, which binds to electronegative phospholipids such as phosphatidylserine. When platelets are stimulated, the flip-flop reaction exchanges phosphatidylcholine with phosphatidylserine from the inside, allowing the platelet surface to trigger thrombin generation. Negatively charged phospholipids exposure can therefore be detected by the binding of FITC-labeled annexin V.
In the presence of HIT serum, standard heparin induced the binding of annexin V to platelets. A similar bell-shaped dose-related curve to that observed in the test of 14C-serotonin release was found (Fig. 3A). SR121903 strongly stimulated the labeling of platelets with FITC-annexin V, its most active concentration being about 1 µg mL−1. Under the same conditions, no binding of annexin V was observed with SanOrg123781A, for all of the tested concentrations.
HIT-induced CD62 secretion
CD62 is an intraplatelet protein that is not present on the surface of resting platelets. It is stored in the membrane of the α-granules and is exposed when these granules fuse with the external membrane . It can therefore be considered as a marker of platelet release. CD62 exposure has been shown to be usually dependent on a strong challenge of platelets such as the one observed with thrombin . After incubation of platelets with heparin, in the presence of HIT serum, platelets presented a strong labeling with anti-CD62 antibodies, indicative of a massive secretion of the platelet α-granule content (Fig. 3B). Similarly, an important labeling was observed after exposure of platelets to SR121903. The active concentration was always around 1 µg mL−1, which seems to be less than the active concentration of standard heparin. In contrast, SanOrg123781A did not demonstrate any activation of platelet release by this method.
HIT-induced GpIIb–IIIa activation
PAC1 is a monoclonal antibody that recognizes the active conformation of human GpIIb–IIIa  present at the platelet surface. It has been used for evaluating the activity of several drugs, including anti-GpIIb–IIIa  and ADP receptor (anti-P2Y12) antagonists such as thienopyridines . The activation of GpIIb–IIIa is clearly correlated with the ability of the complex to bind soluble fibrinogen and result in platelet aggregation. These methods have been compared with the ‘classical’14C-serotonin release method.
The incubation of platelets in sera from HIT patients produced an obvious activation of the GpIIb–IIIa complex in the presence of standard heparin, as well as SR121903 (Fig. 3C). The active concentrations were similar to those that showed an effect in the other tests. Once again, SanOrg123781 failed to activate platelets, and GpIIb–IIIa activation did not occur.
We have compared SanOrg123781 and SR121903 (Table 1) for their affinity for PF4 and their ability to activate platelets in the presence of HIT serum. The former is a 16-mer oligosaccharide, presenting two terminal charged domains, separated by a 7-mer neutral linker. The later is a highly sulfated 17-mer oligosaccharide. Both of them display a strong anti-FXa and anti-FIIa activity (Table 1) that can be attributed to the presence of an anti-FXa pentasaccharidic sequence and a thrombin-binding domain in their extremities.
However, their affinities for PF4 were markedly different, since SR121903 could displace bound heparin, whereas SanOrg123781 could not (Fig. 1). This result underlines the importance of the charge for the interaction with PF4. SR121903 has a stronger negative charge, similar to that of heparin, due to a long sulfated sequence between the thrombin and the antithrombin binding domains.
Platelet studies, in the presence of HIT serum, showed that heparin could activate secretion of platelet-dense granules (measured by the release of serotonin), as already shown in various studies [4,5,8,18]. SR121903 demonstrated a similar releasing effect, indicating its ability to stimulate platelets by a HIT-dependent mechanism, in accordance with its binding capacity for PF4. On the contrary, SanOrg123781 did not activate the release of serotonin. This result was in accordance with absence of affinity of this molecule for PF4.
In order to confirm these results, we measured platelet activation by flow cytometry, with markers specific for different activation steps. The exchange of phospholipids (measured by annexin V binding), the exposure of alpha-granule membrane proteins (measured by CD62 detection) and the activation of GpIIb–IIIa complexes (measured by PAC1 labeling), were all stimulated when platelets were incubated with heparin in the presence of HIT serum. As expected, SR121903, the charged oligosaccharide, behaved in a way comparable with that of heparin, and activated the platelets with a comparable intensity. In contrast, when platelets were incubated with SanOrg123781 in the presence of HIT serum, none of these activation markers was expressed, confirming the absence of immuno-induced response with a less charged and better designed oligosaccharide.
These activations did correlate each other regarding their effect–concentration relationships. This observation clearly demonstrates that all these activation parameters are dependent on the same primary event. Therefore, when serotonin is released from the alpha granules, exposure of negative phospholipids, dense granule secretion and GpIIb–IIIa activation occur.
We know that the inside-out exchange of negative phospholipids in the platelet membrane generates a template surface for the coagulation enzymes, which in turn, strongly participate in the thrombin generation, blood coagulation and thrombus formation.
Activation of the GpIIb–IIIa complex increases its affinity for fibrinogen, and results in the ability of platelets to rapidly aggregate. These effects are highly relevant for the clinical expression of HIT, in which a thrombotic situation may occur. The secretion of dense granules is an activation step occurring with the strongest platelet agonists only. The HIT-dependent activation can be considered as a potent challenge for platelets. The consequence of this secretion is thought to be of dramatic importance since significant amounts of PF4 are released, which promote the immunological response. Moreover, platelet-released inflammatory mediators such as CD40 ligand (CD154) and growth factors such as platelet-derived growth factor cause vessel wall injury by inducing apoptosis of endothelial cells, and stimulating the proliferation of the smooth muscle cells. Exposed CD62 is also known to mediate platelet–neutrophil interaction and to favor an inflammatory response. Finally, adhesive proteins such as fibrinogen, fibronectin, thrombospondin and von Willebrand factor are also released during the secretion of platelet-dense granules, and participate in the platelet aggregation process.
In conclusion, we have demonstrated the importance of the charge of the polysaccharides in the HIT-induced platelet reactions measured by diverse methods, some of them being described here for this purpose for the first time. By comparing two well-defined synthetic oligosaccharides with comparable length, and similar antithrombin and anti-Xa activities, we have found that the affinity for PF4 and the resulting HIT-mediated platelet activation are strongly linked to the number of sulfated functions and therefore to the electronegativity of the molecules.
The authors wish to thank P. Barron for manuscript revision.