Factor XII C46T gene polymorphism and the risk of cerebral venous thrombosis.

BACKGROUND: The TT genotype of a functional factor XII (FXII) C46T gene polymorphism was shown to be a risk factor for peripheral venous thrombosis. We tested whether this genetic variant also increases the risk for cerebral venous thrombosis (CVT). METHODS: We performed a case-control study including 78 consecutive patients with proven CVT and 201 healthy population controls from South Germany. The FXII C46T genotype was assessed using a PCR technique. RESULTS: The TT genotype of the FXII C46T polymorphism was more common in patients (16.7%) than in controls (5.5%). A strong association of the TT genotype with CVT was found, which was independent of covariables (adjusted odds ratio 4.57; 95% CI 1.55 to 13.41; p = 0.006). CONCLUSION: The TT genotype of the functional factor XII C46T gene polymorphism may be a new independent risk factor for cerebral venous thrombosis (CVT). Our finding warrants confirmation in an independent study before this genetic variant should be added to the panel of established risk factors for CVT.

[Platelet aggregation disorders]. / Thrombozytenaggregationshemmer.

Irrespective of their mechanism of action, anticoagulants reduce the formation and action of thrombin. Thus they interfere with a final step in coagulation. Among platelet inhibitors only the GPIIb/IIIa antagonists inhibit the common pathway of aggregation, namely the formation of platelet-to-platelet bridges which are mediated by fibrinogen or von Willebrand factor. In contrast, acetylsalicylic acid (ASA), NSAIDs, clopidogrel (Plavix) or ticlopidine (Tyklid) inhibit platelet activation by abrogating the formation or action of a secondary platelet agonist, namely of thromboxane A(2) or ADP. They do not block platelet aggregation which is directly induced by thrombin. Therefore, ASA, clopidogrel, or ticlopidine are not associated with a significant risk of bleeding as long as other factors such as an extensive thrombocytopenia or anticoagulation are not involved. Therefore, in contrast to anticoagulants, therapeutic drug monitoring is not necessary with ASA, clopidogrel, nor ticlopidine. On the other hand, ASA has even to be applied at a dosage that almost completely inhibits thromboxane synthesis in order to act on platelet aggregation at all. Among the GPIIb/IIIa-antagonists only parenteral drugs have been approved for therapeutic use, e. g. abciximab (ReoPro), eptifibatide (Integrilin), and tirofiban (Aggrastat). Clinical studies revealed an increased risk of bleeding without a sufficient therapeutic benefit of oral GPIIb/IIIa antagonists. GPIIb/IIIa antagonists may induce thrombocytopenia that is attributed to an out-side-in signalling or immunological phenomena. A test system (Ultegra, Accumetrics) is available for a therapeutic drug monitoring of GPIIb/IIIa antagonists. However, estimation of the bleeding risk always requires an evaluation of all factors influencing the haemostatic system, especially when heparin or other inhibitors are applied additionally.

Platelet activation through triathlon competition in ultra-endurance trained athletes: impact of thrombin and plasmin generation and catecholamine release.

The aims of this study were to evaluate whether platelets are activated during strenuous exercise in healthy athletes. Also, to determine the impact of plasmin and thrombin activity and catecholamine release. Previous studies have shown activation of the hemostatic system after competitive exercise, but platelet activation was thought to be absent in trained athletes. The impact of thrombin and other potent platelet activators is still a matter for debate. We examined 30 healthy triathletes during a triathlon competition. Flow cytometric detection of CD62p (P-selectin) was used to measure in vivo activation of platelets. Platelet-leukocyte aggregates were also determined. Thrombin concentration was assessed by the thrombin-antithrombin III complex (TAT) and the fibrinolytic state was characterised by the plasmin-alpha2-antiplasmin complex (PAP). Catecholamines were measured by means of high-pressure liquid chromatography. CD62p rose from baseline (2.3%) to 3.4% and was still elevated after 2 hours (3.1%, p = 0.0133). Platelet-leukocyte aggregates were elevated 30 min after exercise (4.3 % vs 3.6%) and decreased significantly after 60 min (2.9 %, p = 0.008). TAT increased from 3.9 microg/l to 8.3 microg/l after competition and to 5.4 microg/l 2 hours later (p < 0.001). PAP increased 10-fold from 350 microg/l to 3,267 microg/l after the triathlon and was still elevated after 2 hours (1,074 microg/l, p<0.001). No linear correlation was found between the hemostatic markers, catecholamines and platelet activation. Platelets, coagulation and fibrinolysis are activated by competitive exercise in athletes, whereby fibrinolytic changes are pronounced. Mechanisms of platelet activation during exercise include phenomena other than plasmatic hemostatic factors and catecholamines.

Incomplete inhibition of platelet aggregation and glycoprotein IIb-IIIa receptor blockade by abciximab: importance of internal pool of glycoprotein IIb-IIIa receptors.

Resting platelets contain a substantial internal pool of GPIIb-IIIa complexes that is exposed on the surface of activated platelets. Whether the exposure of internal GPIIb-IIIa complexes on the activated platelet surface affects therapy with GPIIb-IIIa antagonists is poorly understood. We addressed this issue in thirteen patients who underwent elective coronary stenting and received abciximab. Platelet aggregation, surface expression of GPIIb-IIIa and P-selectin, receptor blockade of GPIIb-IIIa, and platelet release in response to ADP and thrombin-receptor activating peptide (TRAP) were determined ex vivo by Lumi-aggregometry and flow cytometry before, during and after abciximab administration. We found that inhibition of aggregation and GPIIb-IIIa blockade of ADP-stimulated platelets was almost complete during abciximab administration. In contrast, when TRAP was used to stimulate platelets ex vivo aggregation was only partially inhibited, most likely due to release of internal pool of unblocked GPIIb-IIIa complexes. Using electron microscopy we found that 7E3-occupied GPIIb-IIIa complexes are internalized into the surface connected system (SCS) and the alpha-granules of washed platelets which was associated with a reduced degranulation of the alpha-granula membrane protein P-selectin. We conclude, that despite internalization of abciximab into the internal pool of GPIIb-IIIa, upon strong platelet activation with thrombin a significant amount of unblocked internal GPIIb-IIIa can be exposed on the platelet surface and mediate platelet aggregation. Incomplete blockade of the internal GPIIb-IIIa pool may limit clinical efficacy of abciximab.

Activation of blood platelets in response to maximal isometric exercise of the dominant arm.

Isometric exercise is a popular form of physical activity for many people. Only few studies exist on the effects of this type of exercise on the hemostatic system. Eleven male healthy subjects (21-42 years) of varying fitness levels were investigated before, immediately after and 10 min after strenuous isometric exercise of the dominant arm. Blood samples were drawn by repetitive puncture from both the exercising and the contralateral arm. The following variables were studied: Prothrombin time and partial thromboplastin time as group tests for the plasmatic coagulation system; platelet count as well as p-selectin expression for the platelet system; tissue plasminogen activator (t-PA) activity and antigen for the fibrinolytic system. The partial thromboplastin time was shortened immediately after maximal isometric exercise of the dominant arm, the prothrombin time remained unchanged. No change was found in the platelet count, but a marked p-selectin expression was observed immediately after maximal isometric exercise of the dominant arm (p < 0.05) and even in the resting contralateral arm. Values returned to baseline after 10 min. There was a slight increase of t-PA antigen concentration and white blood cell count at maximal isometric contraction which did not occur in the resting arm, although changes over the 3 time points were significant in both arms. Maximal isometric exercise leads to platelet activation in both arms, a slight aPTT decrease and t-PA antigen increase in local blood stream. As compensatory fibrinolytic changes do not occur, it is an open question whether isometric exercise increases the potential risk of thromboembolism.

[Grand mal series after Ecstasy abuse]. / Grand-mal-Serie nach Ecstasy-Einnahme.

The consumption of the jet set drug Ecstasy (3,4-methylenedioxymetamphetamine, abbreviated to MDMA) by young people is increasing markedly. Parallel to this development, there is a large number of reports on severe neurological, psychiatric and medical complications following the use of Ecstasy. Seizures are among the most common clinical complications of the CNS following the ingestion of Ecstasy. Our report presents the case of a 21-year-old patient, who had a series of grand mal seizures after taking 12 tablets of Ecstasy. 36 hours after ingestion the substance MDMA was demonstrated at a level of 300 ng/ml in the serum and CSF. Following treatment with Clonazepam and under an adequate level of carbamazepine, no further seizures occurred. The diagnosis was difficult because the patient initially denied the consumption of drugs and the routine toxicological screening does not contain the substance MDMA.

Prothrombin gene G20210-->A transition is a risk factor for cerebral venous thrombosis.

BACKGROUND AND PURPOSE: It has been recently reported that a G-->A transition at nucleotide position 20210 in the 3'-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increased risk of deep venous thrombosis. To date, it is unknown whether this polymorphism also represents a risk factor for cerebral venous thrombosis (CVT). METHODS: Venous blood samples were collected from 45 patients with CVT and from 354 healthy blood donors as controls. A second control group consisted of 131 subjects with acute ischemic stroke or transient ischemic attack (TIA). Genomic DNA was isolated from peripheral blood leukocytes. Amplification of DNA was performed by polymerase chain reaction (PCR). The G-->A transition at nucleotide position 20210 of the prothrombin gene was detected by allele-specific restriction digestion. RESULTS: The G20210-->A transition in the prothrombin gene was found in a heterozygous form in 4 of 45 patients with CVT (8.9%) and in 8 of 354 healthy control subjects (2.3%). This difference was statistically significant (P=0.010). The G20210-->A transition increased the relative risk for CVT approximately 5-fold (age-adjusted odds ratio 5.7; 95% CI 1.5 to 21.5). In contrast, in the group of patients with acute cerebral ischemia, only 3 of 131 subjects (2.3%) were heterozygous for the G20210-->A transition, which corresponded to the prevalence in the group of healthy blood donors. CONCLUSIONS: The recently described G20210-->A transition in the 3'-untranslated region of the prothrombin gene is an inherited risk factor for CVT but obviously not for acute ischemic stroke or TIA.

Increased fraction of circulating activated platelets in acute and previous cerebrovascular ischemia.

Determination of circulating activated platelets may be helpful to estimate the prognosis and to stratify therapies in arterial vascular disorders including stroke. We used flow cytometry and phase contrast microscopy to study whether the fraction of platelets expressing p-selectin and CD63 and the fraction of platelets with shape change are increased in patients with acute and previous cerebrovascular ischemia. The proportion of platelets expressing activation dependent antigens was higher in patients with acute (n = 24; p-selectin: 8.23 +/- 4.21%; CD63: 3.53 +/- 2.53%) and with previous cerebrovascular ischemia (n = 46; 3.86 +/- 1.98%; 2.80 +/- 1.79%) as compared to age- and sex-matched control subjects (n = 35; 2.17 +/- 0.96%; 1.79 +/- 0.75%; p < or = 0.005, respectively). In patients with previous ischemia, there was no difference between treatment with aspirin (n = 25) or phenprocoumon (n = 21). Hypertension, diabetes mellitus and smoking were not associated with increased antigen expression (analysis of variance). The fraction of discoid platelets and platelet counts were not significantly different between groups. Our results indicate increased expression of platelet neoantigens in acute and to a less degree in previous cerebrovascular ischemia. Ongoing platelet activation after cerebrovascular ischemia despite therapy with aspirin or phenprocoumon indicates that new anti-platelet drugs may be of benefit for these patients. Flow cytometry appears to be a useful tool to assess platelet function in cerebrovascular ischemia.

Characterization of the thromboxane synthase pathway product 12-oxoheptadeca-5(Z)-8(E)-10(E)-trienoic acid as a thromboxane A2 receptor antagonist with minimal intrinsic activity.

Thromboxane synthase forms thromboxane (TX) A2 and 12(S)-hydroxyheptadeca-5(Z)-8(E)-10(E)-trienoic acid (HHT) at equimolar amounts. Twelve-oxoheptadeca-5(Z)-8(E)-10(E)-trienoic acid (Oxo-HT) is the primary metabolite of HHT and has been described to be an inhibitor of platelet aggregation. Functional studies, Schild analysis and competitive binding studies were performed to clarify its mode of action. Oxo-HT was prepared biosynthetically as well as chemosynthetically, purified and characterized by gas chromatography and mass spectrometry. Platelet activation was assessed by determination of shape change, aggregation, fibrinogen binding and P-selectin expression using optical aggregometry and flow cytometry. Oxo-HT 0.1 nM to 50 microM did not induce platelet activation. Furthermore, it had no effect on platelet activation induced by thrombin, ADP or PAF. In contrast, Oxo-HT inhibited platelet aggregation, fibrinogen binding and P-selectin expression induced by U46619 in a competitive manner. Schild analysis for U146619-induced fibrinogen binding and P-selectin expression revealed pA2 values of 6.1 and 6.6, respectively, which correspond to Kd values of approximately 0.8 microM and 0.3 microM, respectively. Oxo-HT also inhibited U46619 induced shape change (IC50 is approximately equal to 10 microM). However, Oxo-HT over a concentration range of 0.1-1 microM enhanced the partial shape change induced by low concentrations of U46619. Thus Oxo-HT seems to possess a minimal agonistic potential, which alone is not sufficient to trigger a platelet activation but can enhance low levels of platelet activation. Oxo-HT blocked the binding of [3H]SQ 29548 in a concentration-dependent manner, whereas HHT did not displace [3H]SQ 29548. The Kd of Oxo-HT determined from competition binding studies was 7.7 microM, about 10-25-fold higher than the apparent Kd determined by Schild analysis. This discrepancy might be due to a desensitization of the TXA2 receptor triggered by the minimal intrinsic activity of Oxo-HT. We conclude that Oxo-HT is a naturally occurring specific TXA2 receptor antagonist with minimal intrinsic activity. Oxo-HT may contribute to the regulation of TXA2-induced platelet activation in vivo.

In-vivo platelet activation correlates with red cell anionic phospholipid exposure in patients with beta-thalassaemia major.

Several clinical and laboratory findings suggest the presence of a chronic hypercoagulable state in patients with beta-thalassaemia major (TM). We have previously shown that isolated TM red blood cells (RBC) strongly enhance prothrombin activation, suggesting an increased membrane exposure of procoagulant phospholipids (i.e. phosphatidylserine). In this study we quantitated the procoagulant activity of RBC in TM and thalassaemia intermedia (TI) patients. We also determined the fraction of activated platelets expressing p-selectin (CD62p) or CD63 in these subjects. Both assays were performed by dual-colour flow cytometry. A significantly (P < 0.01) higher fraction of FITC-annexin V-labelled RBC was found in TM and TI patients, compared to the controls. A highly significant correlation (P < 0.001) was found in TM patients between the number of RBC-bound annexin V molecules and the fraction of CD62p (p-selectin) or CD63-positive platelets. This association between annexin V binding to TM RBC and the expression of platelet activation markers was also found in individual TM patients over time. Thus, the procoagulant surface of TM RBC may accelerate thrombin generation in vivo which, in turn, triggers platelet activation.

Discrimination between normal wildtype and carriers of coagulation factor V Leiden mutation by the activated protein C resistance test in the presence of factor V deficient plasma.

Blood samples from 104 patients with clinically suspected thrombophilia were analyzed for coagulation factor V Leiden mutation (1691, G-->A) by allele-specific polymerase chain reaction. In 86 individuals (82.7%), the mutation was not detectable, whereas 15 patients (14.4%) were heterozygous and three patients (2.9%) were homozygous for factor V Leiden mutation. Plasma samples from these individuals were also tested for functional resistance of coagulation factor V to activated protein C (activated protein C resistance). This test was performed on a Schnitger-Gross coagulometer using an activated partial thromboplastin time-based activated protein C resistance test modified by applying a 1 : 5 dilution with factor V-deficiency plasma. All the individuals negative for factor V Leiden mutation were also negative in the functional activated protein C resistance test. On the other hand, all patients carrying the mutation revealed pathologic results in the activated protein C resistance test. The cutoff value for the activated protein C resistance index (> or = 1.7 = negative) was determined by testing 31 male and female blood donors. One of them was heterozygous for factor V Leiden mutation and had an activated protein C resistance index of 1.4, whereas those without factor V Leiden mutation had an activated protein C resistance index of 1.9 +/- 0.1 (mean +/- SD). Patients with clinically suspected thrombophilia without factor V Leiden mutation had an activated protein C resistance index of 2.1 +/- 0.2 (mean +/- SD), whereas patients heterozygous for the mutation had an index of 1.5 +/- 0.1 (mean +/- SD). Within the group of patients carrying the mutation, the activated protein C resistance test even distinguished between heterozygous and three homozygous (activated protein C resistance 1.0 to 1.2) carriers. The data demonstrate that the activated protein C resistance test in the presence of factor V-deficiency plasma provides a clear-cut discrimination between normal wildtype and carriers of factor V Leiden mutation with a sensitivity and specificity of 100%. Verification of positive activated protein C resistance tests can be performed easily with a simple and reliable polymerase chain reaction protocol for the 1691, G-->A mutation.

In vivo degradation of human fibrinogen A alpha: detection of cleavage sites and release of antithrombotic peptides.

Several degradation products of fibrinogen have been shown to possess regulatory functions. Using peptide extracts from human blood filtrate, a large number of fibrinogen A alpha fragments was identified. These fragments are generated at known plasmin attack sites and at several novel cleavage sites especially at hydrophobic and basic amino acid residues. One fragment containing the cell attachment site (RGD sequence) of fibrinogen A alpha efficiently inhibits fibrinogen binding and platelet aggregation (IC50:20-50 microM) in vitro. We conclude that in vivo degradation of fibrinogen A alpha results in generation of endogenous antithrombotic peptides with local importance in fibrinolysis and platelet aggregation.

Platelet-induced neutrophil activation: platelet-expressed fibrinogen induces the oxidative burst in neutrophils by an interaction with CD11C/CD18.

Mutual contacts and platelet-expressed fibrinogen seem to be required for the stimulation of neutrophils by activated platelets. The beta 2-integrins CD11b/CD18 and CD11c/CD18 are potential receptors for fibrinogen on neutrophils. In order to investigate whether binding of fibrinogen to these integrins is involved, monoclonal antibodies (MoAbs) and Gly-Pro-Arg-Pro (GPRP) peptide that inhibits fibrinogen binding to CD11c/CD18 were checked for their effects on the interaction of activated platelets and neutrophils. The luminol-amplified chemiluminescence (CL) as a measure for the oxidative burst of neutrophils was recorded simultaneously to the platelet aggregation in mixed cell suspensions. The adhesion of platelets and neutrophils was determined microscopically. The thromboxane A2 mimetic U46619 was used as a potent platelet agonist but that does not stimulate neutrophils. aggregation and a strong CL of neutrophils. The platelet-induced activation of neutrophils required added fibrinogen which fibronectin or thrombospondin could not substitute for. Cytochalasin D (Cyto D) that blocks actin polymerization totally abrogated the platelet-induced Cl of neutrophils. The MoAb OKM1 against CD11b, which blocks fibrinogen binding to CD11b/CD18 as well as the MoAbs IOT16 and IOT18 directed against CD11a and CD18, respectively, had no effect. In contrast, the MoAb LeuM5 which inhibits the binding of fibrinogen to CD11c/CD18 revealed a strong inhibition. Furthermore, GPRP peptide which CD11c/CD18 recognizes on the A alpha-chain of fibrinogen also strongly inhibited the platelet-induced CL of neutrophils, whereas control peptides such as Gly-His-Arg-Pro (GHRP) or Gly-Pro-Gly-Gly (GPGG) had no effect. In contrast to the platelet-induced CL of neutrophils, Cyto D, MoAb against CD11c and GPRP peptide did not inhibit the CL induced by FMLP and PAF in pure neutrophil suspensions. They also did not affect U46619-induced platelet aggregation. The adhesion of platelets and neutrophils was neither dependent on added fibrinogen nor inhibited by Cyto D, MoAb against CD11c and the GPRP-peptide. Therefore fibrinogen and actin polymerization seem not to be required for the adhesion of neutrophils to platelets. However, the activation of neutrophils depends on the interaction of CD11c/CD18 with the A alpha-chain of platelet-expressed fibrinogen and the contractile system of neutrophils.

Flow cytometric detection of activated platelets: comparison of determining shape change, fibrinogen binding, and P-selectin expression.

Platelet activation plays an important role in the pathomechanisms of arterial vascular disorders including stroke, peripheral arterial disease (PAD), and myocardial infarction. Circulating activated platelets may be useful markers of local thrombotic events occurring in these diseases. Using flow cytometry circulating activated platelets can be detected by determining: 1. the platelets' shape change on the basis of the different light scatter properties of discocytes and spherocytes, 2. the expression of platelet bound fibrinogen or conformation specific neoantigens on fibrinogen and on its platelet receptor, and 3. the exposure of granule membrane proteins such as P-selectin as a result of platelet secretion. The concentration-effect relationships were determined for the ADP and U46619 induced shape change, fibrinogen binding, and expression of P-selectin. The EC50 for the shape change was 4 times lower than the EC50 for the fibrinogen binding and 29 times lower than the EC50 for the expression of P-selectin. These data clearly demonstrate that the shape change is a more sensitive indicator of platelet activation in vitro than fibrinogen binding or P-selectin expression. Both the shape change and fibrinogen binding were reversible, whereas the expression of P-selectin was irreversible upon stimulation. Reversibility of the shape change may be responsible for the fact that in patients with stroke or PAD the fraction of discocytes did not differ from controls, whereas more than 80% of them revealed a significantly higher fraction of P-selectin positive platelets. Thus the determination of the P-selectin expression reveals a higher diagnostic sensitivity for detecting a platelet activation in vivo than the determination of the shape change.

Contact-induced neutrophil activation by platelets in human cell suspensions and whole blood.

Platelet-dependent activation of polymorphonuclear neutrophils (PMNL) was investigated with a lumi-aggregometer in heparinized whole blood and platelet-PMNL suspensions. The lumi-aggregometer allowed us to simultaneously monitor increases in impedance or light transmission as consequences of platelet aggregation and luminol-enhanced chemiluminescence (CL) as a measure of the oxidative burst in PMNL. Aggregation and platelet-PMNL contacts were also checked by light and electron microscopy. In whole blood, adenosine diphosphate (ADP) and the thromboxane A2 mimetic U 46619 induced the aggregation (increase in impedance) and the CL, which were both suppressed by EDTA, arginyl-glycyl-aspartyl-serine (RGDS) peptide, and the absence of stirring. In contrast, FMLP caused only CL that was unaffected by EDTA, RGDS peptide, and nonstirring. Similar observations were obtained with mixed suspensions containing washed platelets and PMNL at their physiologic concentrations. ADP, U 46619, and thrombin induced both aggregation (increase in light transmission) and CL, whereas FMLP caused CL but only very weak aggregation. Exogenous fibrinogen strongly enhanced the effects of ADP and U 46619. Iloprost, EDTA, RGDS peptide, red blood cell (RBC) ghosts, and nonstirring inhibited the effects induced by the platelet agonists, but were ineffective on the CL induced by FMLP. Treatment of platelets with aspirin did not affect the CL of PMNL induced by platelets. Microscopic examination, the requirements of stirring, Ca2+, and fibrinogen, and the inhibitory effects of RGDS peptide and RBC ghosts show that stimulated platelets activate PMNL in a contact-dependent manner that depends on fibrinogen binding. This was confirmed by the immunochemical demonstration of fibrinogen (but not of fibronectin) in the contact spaces between activated platelets and PMNL. Because supernatants and lysates of resting or thrombin-stimulated platelets did not induce the CL of PMNL, soluble agonists did not appear to be involved. Nonstimulated washed platelets also caused CL of PMNL that required stirring and Ca2+ and was inhibited by RBC ghosts. No CL occurred in unstimulated stirred whole blood, suggesting that a preactivation of platelets during the preparation may be responsible for the effects of unstimulated washed platelets. The results show that platelets provide a strong stimulus for PMNL that requires intercellular contact. Fibrinogen exposure on the platelet surface seems to be necessary for the activation of PMNL by stimulated platelets.

Transport of anti-glycoprotein IIb/IIIa-antibodies into the alpha-granules of unstimulated human blood platelets.

The redistribution of the antibody-glycoprotein (GP) IIb/IIIa complex was investigated with the immuno-gold labeling technique in order to trace its transport in resting platelets. Washed platelets were incubated in the presence of aspirin and a prostacyclin analogue (iloprost) with three different monoclonal antibodies (Gi3, J15 and P2) against GPIIb/IIIa. The examination of ultrathin serial sections showed that the surface labeling was internalized into the surface connected membrane system (SCS). Labels were found within the alpha-granules after 40 min and the number of labels increased during longer incubation periods (max. 120 min). The transport possibly involved coated membranes. The alpha-granules were neither found to be altered during this process nor were any morphological signs of platelet activation detectable. The anti-GPIIb/IIIa complex remained membrane-associated during the transfer. These observations indicate that the membrane-GPIIb/IIIa complex was stable and transported from the surface into the alpha-granules of resting platelets. Since this transport was not influenced by iloprost or by aspirin it may be interpreted as constitutive endocytosis.