A monochromatic confocal micro-x-ray fluorescence (µXRF) spectrometer for the lab.

Confocal micro-x-ray fluorescence (µXRF) is a powerful tool to analyze the spatial distribution of major, minor, and trace elements in three dimensions. Typical (confocal) µXRF measurements in the lab use polychromatic excitation, complicating quantification and fundamental parameter-based corrections and furthermore deteriorating peak-to-background ratios due to scattered bremsstrahlung. The goal for the new setup was to remedy these problems, without sacrificing spatial resolution, and keep it flexible for different excitation energies and transportation to other sources. The source assembly consists of a water-cooled fine-focus x-ray diffraction tube and a parallel beam-mirror, which produces a quasi-parallel, monochromatic beam. The presented results were obtained using a 2 kW molybdenum tube and a mirror for Mo-Kα. The confocal setup itself consists of two polycapillary half-lenses, one for the source side and the other for the detector side, where a 50 mm2 silicon drift detector is mounted. Both polycapillaries have a focus size of ∼15 µm for Mo-Kα. The second polycapillary can also be exchanged for a custom-designed collimator in order to perform non-confocal µXRF. Details of the technical setup and results from technical and biological samples are presented. Detection limits for selected elements from Ca to Pb in the confocal and non-confocal mode were established (e.g., 1 µg/g non-confocal and 20 µg/g confocal for As) using the NIST standard reference materials (SRMs) 621 and 1412. Furthermore, the results of the measurements of SRM 621 were evaluated using the fundamental parameter based quantification software ATI-QUANT. The results are compared with the certified values and generally are in good agreement.

1-Methylnicotinamide (MNA), a primary metabolite of nicotinamide, exerts anti-thrombotic activity mediated by a cyclooxygenase-2/prostacyclin pathway.

BACKGROUND AND PURPOSE: 1-methylnicotinamide (MNA) has been considered to be an inactive metabolite of nicotinamide. Here we assessed the anti-thrombotic activity of MNA in vivo. EXPERIMENTAL APPROACH: Antithrombotic action of MNA was studied in normotensive rats with extracorporeal thrombus formation (thrombolysis), in renovascular hypertensive rats with intraarterial thrombus formation (arterial thrombosis) and in a venous thrombosis model in rats (venous thrombosis). KEY RESULTS: MNA (3-100 mg kg(-1)) induced a dose-dependent and sustained thrombolytic response, associated with a rise in 6-keto-PGF(1alpha) in blood. Various compounds structurally related to MNA were either inactive or weaker thrombolytics. Rofecoxib (0.01-1 mg kg(-1)), dose-dependently inhibited the thrombolytic response of MNA, indomethacin (5 mg kg(-1)) abolished it, while L-NAME (5 mg kg(-1)) were without effect. MNA (3-30 mg kg(-1)) also reduced arterial thrombosis and this effect was abrogated by indomethacin (2.5 mg kg(-1)) as well as by rofecoxib (1 mg kg(-1)). MNA, however, did not affect venous thrombosis. In vitro MNA did not modify platelet aggregation nor induce vasodilation. CONCLUSIONS AND IMPLICATIONS: MNA displayed a profile of anti-thrombotic activity in vivo that surpasses that of closely related compounds. MNA inhibited platelet-dependent thrombosis by a mechanism involving cyclooxygenase-2 and prostacyclin. Our findings suggest that endogenous MNA, produced in the liver by nicotinamide N-methyltransferase, could be an endogenous activator of prostacyclin production and thus may regulate thrombotic as well as inflammatory processes in the cardiovascular system.

In vivo endothelial interaction between ACE and COX inhibitors.

Here we studied the mechanism of thrombolytic response (THR) induced by angiotensin converting enzyme (ACE-I) in vivo in anaesthetised Wistar rats with extracorporeal circulation. Intravenous injections of ACE-Is, i.e. perindopril or quinapril at non-hypotensive doses of 3-30 microg kg(-1) produced a dose-dependent thrombolysis that was associated with a parallel rise in arterial blood levels of 6-keto-PGF(1 alpha), but not those of TXB(2) or PGE(2). L-NAME at a dose of 5 mg kg(-1) affected significantly neither ACE-I-induced thrombolysis nor prostacyclinemia; however, the pre-treatment with icatibant (0.1-0.5 mg kg(-1)) abolished both effects. The selective COX-1 inhibitor, SC 560 (100-300 microg kg(-1) i.v.), or a would be selective COX-3 inhibitor--paracetamol (acetaminophen, 1-3 mg kg(-1)), both agents induced a transient thrombolysis and slightly potentiated thrombolysis by ACE-Is. In contrast, selective COX-2 inhibitors (rofecoxib>>celecoxib>nimesulide>NS 398) were thrombogenic, and abolished THR and rise in 6-keto-PGF(1 alpha) induced by ACE-Is. Summing up, in our in vivo bioassay system ACE-Is such as quinapril, perindopril or captopril at non-hypotensive doses evoke THR that is mediated by endogenous bradykinin and prostacyclin derived from endothelial COX-2.

Synthesis and thrombolytic activity of new thienopyrimidinone derivatives.

It has been observed that ticlopidine and clopidogrel show, apart from their delayed antiplatelet properties, an immediate and transient thrombolytic action related to the ability of these thienopyridines to stimulate the secretory function of vascular endothelium. With the objective to construct new molecules with identical thrombolytic potency but at a higher level, we carried out different structural modifications in the thienopyridine chemical molecule to conclude that the presence of a second N atom in the pyridine cycle (yielding pyrimidine moiety) and the presence of an additional cycle fused to the thienyl ring would lead to enhanced thrombolytic effects. Here we report the six-step synthesis of a series of new benzothienopyrimidinone derivatives characterized by this searched for potent thrombolytic activity. The pharmacological assay used anaesthetised Wistar rats with extracorporal circulation in which arterial blood superfused thrombi adhering to a strip of collagen. Weight of thrombi was continuously monitored. Six compounds of the series were much more potent thrombolytic agents than their thienopyridine references: the effective thrombolytic dose that produced 30% of maximum thrombolysis (ED30) was at a range of 8 to 170 microg kg(-1) as compared with ED30 values of 16000 to 20000 microg kg(-1) for clopidogrel and ticlopidine respectively. Especially with the most active compound, this difference in the threshold thrombolytic dose, giving an intensity of action higher by three orders of magnitude, was accompanied by a lengthening of the response. Apart from that these compounds have shown to be synthetic thrombolytics, they certainly deserve further studying.

Comparison of endothelial pleiotropic actions of angiotensin converting enzyme inhibitors and statins.

Two in vitro and one in vivo assay were performed to study the endothelial pleiotropic actions of "tissue type" angiotensin converting enzyme inhibitors (ACE-Is) such as perindopril and quinapril, their active forms, that is, quinaprilat and peridoprilat, or of statins belonging to natural (lovastatin), semisynthetic (simvastatin), and synthetic enantiomeric (atorvastatin, cerivastatin) classes. Cytoplasmic [Ca2+]i levels in cultured bovine aortic endothelial cells and endothelium-dependent nitric oxide-mediated coronary vasodilatation in the Langendorff preparation of guinea pig heart constituted our in vitro assays. The in vivo assay consisted of study of PGI2-mediated thrombolytic response in arterial blood of rats after intravenous administration of drugs. In this last assay, perindopril and quinapril proved to be, by two orders of magnitude, more potent PGI2-dependent thrombolytics than the most potent statin (atorvastatin). However, in both in vitro assays we found a higher endothelial efficacy of statins as compared to ACE-Is. In particular, those statins that contain the lactone ring in their molecules (lovastatin, simvastatin) were the most potent coronary vasodilators. In summary, the in vivo profile of action of ACE-Is and statins contrasted with their reversed order of potency in vitro. We hypothesize that the endocrine-like function of the pulmonary circulation [28-31] may be responsible for the in vivo bradykinin-triggered, PGI2-mediated thrombolysis by ACE-Is, whereas the pleiotropic action of statins, possibly involving inhibition of prenylation [14-19], is diffused throughout many vascular beds.

Thrombolysis by thienopyridines and their congeners.

We propose that anti-platelet thienopyridines, such as ticlopidine or clopidogrel, are thrombolytic owing to endothelial release of prostacyclin (PGI2) and tissue plasminogen activator (t-PA). In this study we used anaesthetised Wistar rats with extracorporal circulation in which arterial blood superfused thrombi which adhered to a strip of collagen. Weight of thrombi was continuously monitored. When administered intravenously, clopidogrel or its R enantiomer deprived of anti-platelet action, both at doses of 3 mg x kg(-1), produced lost in weight of thrombi by 14.1 +/- 1.3% or 16.0 +/- 1.4% (n = 9), and at doses 10 mg x kg(-1) by 28.3 +/- 2.3% or 30.4 +/- 1.9% (n = 8), respectively. Maximum of thrombolysis occurred 30-45 min following the drug administration. Ticlopidine at a dose of 30 mg x kg(-1) reduced weight of thrombi by 33.7 +/- 1.7% (n = 32). Thrombolytic action of ticlopidine was accompanied by a rise in 6!keto-PGF1alpha blood levels from 0.42 +/- 0.10 to 1.58 +/- 0.29 ng x ml(-1) and t-PA antigen plasma levels from 4.70 +/- 1.00 to 12.90 +/- 1.15 ng x ml(-1) (n = 7). Five out of eleven tested thienopyridine congeners with pyrimidine or pyrimidinone instead of pyridine rings had thrombolytic potencies similar to that of clopidogrel (ED30s at a range of 6.2-11.4 mg x kg(-1)). A substantial increase in thrombolytic potency (ED30s at a range of 0.3-2.1 mg x kg(-1)) was observed for congeners in which thienyl ring was condensed with an additional cyclopentyl, cyclohexyl or cycloheptyl structures or in which thienopyridine complex was replaced for a pyridopyrimidine one. We claim that thienopyridines, independently of their delayed anti-platelet action, do produce immediate thrombolysis in vivo. This new activity emulates capacity of their native, non-metabolised molecules to release prostacyclin and tissue plasminogen activator. We have also shown that structural changes in molecules of thienopyridines may intensify their thrombolytic potency.

Effect of ticlopidine on streptokinase-induced thrombolysis in rats.

Using our original assay system we found that ticlopidine (TP, 30 mg/kg i.v.) had produced a prompt thrombolysis of preformed clots in extracorporeal circulation of anaesthetized rats. In contrast with ticlopidine streptokinase (SK, 30,000 U/kg i.v.) proved thrombogenic for an initial period of 10-30 min, followed only later by the expected thrombolytic action. In ticlopidine pretreated rats the thrombogenic effect of streptokinase was eliminated, and its thrombolytic potency intensified. This is the first experimental support for a conception that a combined therapy with ticlopidine and streptokinase may restrict the "early hazard" associated with streptokinase therapy.

Thrombolytic and antiplatelet action of xanthinol nicotinate (Sadamin): possible mechanisms.

Here we report on thrombolytic and hypotensive actions of Xanthinol nicotinate (Sadamin) in rats and on its anti-platelet and fibrinolytic effects in patients with peripheral arterial obliterative disease (PAOD). Special consideration was given to a proposal of new mechanisms of anti-platelet and thrombolytic actions of Sadamin. We conclude that the mechanism of anti-platelet and thrombolytic activity of Sadamin partly consists of a simultaneous release of endogenous prostacyclin and nitric oxide by the nicotinate component of Sadamin, whereas the theophylline component is responsible for enhancement of physiological actions of these endothelial mediators at the level of cyclic nucleotides which are their second messengers.

Thrombolytic activity of beta-adrenolytic drug, sotalol.

Sotalol is a beta-adrenoreceptor blocking drug, the clinical efficacy of which has been linked up to its negative chrono- and inotropic effects and its hypotensive action. In addition, beta-adrenolytic drugs are known to inhibit platelet aggregation in vitro possibly through lowering of calcium ions level. Here, we report that in rats sotalol at a dose of 10-20 mg/kg i.v., apart from hypotension, evokes instantaneous thrombolytic effect. This is associated with an increase in plasma level of tissue plasminogen activator (t-PA). In vitro, sotalol at a concentration of 1-100 microM inhibits thrombogenesis on surface of rabbit aorta endothelium superfused with blood. Sotalol also has a weak anti-aggregatory activity (IC50 approximately 500-1000 microM) in human platelet rich plasma (PRP). Since the thrombolytic and fibrinolytic but not hypotensive effects of sotalol were inhibited by cyclooxygenase inhibitor, indomethacin, while its hypotensive but not thrombolytic potency was dimished by an inhibitor of nitric oxide synthase, NG-nitro-L-arginine (L-NNA), we have linked up the sotalol-induced effects in vivo with the release of prostacyclin and nitric oxide. Our data point out to a possibility that prostacyclin and nitric oxide concomitantly released from endothelium and/or from other blood cells after administration of sotalol, may play different roles: prostacyclin may be responsible for fibrinolytic, thrombolytic and antithrombotic properties, while nitric oxide may take part in the mechanism of sotalol-induced hypotension.

Thrombolytic action of ticlopidine: possible mechanisms.

Ticlopidine (Ticlide), an anti-platelet drug with a broad scope of clinical applications, is claimed to be an antagonist of adenosine diphosphate on platelet receptors. In vitro this antagonism cannot be demonstrated. Ex vivo it is detectable many hours after oral administration of the drug, perhaps subsequently to its biotransformation to an unknown metabolite. Here, we report for the first time that in patients with peripheral arterial disease and in cats with extracorporal circulation ticlopidine evokes instantaneous thrombolytic or fibrinolytic effects which are not associated with inhibition of platelet aggregation. Shortening of euglobulin clot lysis time and increase in plasma levels of tissue plasminogen activator were observed 1-2 h after oral ingestion of ticlopidine at a single dose of 500 mg. In cats ticlopidine produced instantaneous anti-thrombotic and thrombolytic effects at doses of 0.3-1 mg/kg and 10-15 mg/kg i.v., respectively. Thrombolysis by ticlopidine (10 mg/kg i.v.) was comparable to that by prostacyclin at a dose of 0.3 microgram/kg i.v. Ticlopidine at a concentration of 100 microM increased endothelial thromboresistance in vitro. The drug did not inhibit the activity of cyclooxygenase-1 or 12-lipoxygenase while it inhibited lipid autooxidation (IC50 = 18 microM) in rat liver microsomes. Our data point to a possibility that the therapeutic efficacy of ticlopidine might be associated not only with its delayed anti-platelet effects but also with its immediate thrombolytic action which is likely to be mediated by endothelial prostacyclin and tissue plasminogen activator rather than by platelet mechanisms.

Prostacyclin, nitric oxide, and atherosclerosis.

Disorders in arterial production of PGI2 and NO occur in atherosclerosis. Exogenous PGI2 and NO are capable of interacting pharmacologically. We claim that no such direct interactions occur between endogenous endothelial PGI2 and NO. Studying mechanisms of cardiac reactive hyperemia in guinea pigs and of thrombolysis in cats, we surmise that in vivo vascular intima releases PGI2 intraluminally while NO is secreted abluminally and thus these two ephemeral mediators do not see each other. Hence, in any disease, the disturbances in endothelial generation of PGI2 or NO have to be scrutinized separately. It may well be that endogenous PGI2 maintains endothelial thromboresistance while NO controls arterial myocytes and tissues in which microcirculation is embedded. These responsibilities remain unshared. Interactions between PGI2 and NO are confined to pharmacological domains.

The mechanism of anti-thrombotic, thrombolytic and fibrinolytic actions of camonagrel--a new thromboxane synthase inhibitor.

So far pharmacological consequences of inhibition of thromboxane A2 (TXA2) synthase by imidazole derivatives (e.g., camonagrel or dazoxiben) were linked to suppression of platelet activity. Here we report that in patients with peripheral atherosclerosis or in cats with extracorporeal thrombogenesis treatment with camonagrel is associated with activation of fibrinolysis or thrombolysis. These phenomena seem to be related to the camonagrel-induced shift in metabolism of prostaglandin endoperoxides from TXA2 to prostacyclin (PGI2), although in an in vitro model the involvement of the L-arginine/nitric oxide pathway cannot be excluded. In cats camonagrel (10 mg/kg i.v.) produced not only a fall in TXB2 but also a rise in 6-keto-PGF1 alpha and no change in cyclic-GMP plasma levels. This points to PGI2 rather than to nitric oxide as an in vivo mediator of camonagrel-induced thrombolysis. The crucial role of endogenous PGI2 in the thrombolytic response to camonagrel in cats was evidenced by its blockade following pretreatment of animals with a megadose of aspirin (50 mg/kg i.v.) and lack of any effect on pretreatment with L-NAME (100 micrograms/kg/min, i.v.). Obviously TXA2 synthase inhibitors (e.g., camonagrel) and cyclo-oxygenase inhibitors (e.g., aspirin) antagonize each other in their anti-thrombotic actions and must not be administered at the same time. Furthermore, in patients camonagrel (800 mg orally) suppressed TXA2 generation by 99.5% and doubled the plasma level of 6-keto-PGF1 alpha.(ABSTRACT TRUNCATED AT 250 WORDS)

Permissive effect of molsidomine towards cardioprotective action of iloprost in myocardial ischemia in cats.

Molsidomine, a donor of nitric oxide, is a drug used in the treatment of ischaemic heart disease. Iloprost, a stable analogue of prostacyclin, is a cardioprotective agent in dogs, cats and rats but not in men. We have studied an interaction between molsidomine and iloprost in protecting against consequences of the "no-reperfusion" myocardial ischaemia. In ten control open-chest cats the left descending coronary artery (LDCA) was ligated at a site of its branching. This procedure caused 80% of mortality and the survival time was 40.9 +/- 8.6 min. The death of cats was preceded by continuous premature ventricular contractions (PVC) which appeared 2.1 +/- 0.3 min after LDCA and occurred with frequency of 7.3 +/- 0.6 per min. Molsidomine at a dose of 20 micrograms/kg i.v. given to ten cats before LDCA was neither cardioprotective nor it influenced the rate of mortality while iloprost at a dose of 2 micrograms/kg i.v. opposed the outcome of LDCA as alluded by the elongation of the survival time to 66.6 +/- 7.6 min and the delay of the onset of PVC to 9.1 +/- 1.9 min; also the frequency of PVC fell to 3.6 +/- 0.4 per min, however, the LDCA-induced mortality (60%) was not significantly different from that in control animals (80%). On the other hand, in ten cats with LDCA which were pretreated with a mixture of molsidomine and iloprost there was observed a significant reduction of the LDCA-induced mortality (down to 20%) and a two fold increase in the survival time. Thereby, we conclude that molsidomine permitted enhancing the cardioprotective potency of iloprost.(ABSTRACT TRUNCATED AT 250 WORDS)

Kinins and thrombolysis.

In cats with extracorporeal circulation arterial blood pressure and thrombolysis were assayed. In this model apart from their hypotensive properties kallikrein (3-10 units/kg, i.v.) and captopril (> 200 micrograms/kg, i.v.) dissipated blood clots which were preformed on superfused collagen strips. Captopril at a lower dose of 50 micrograms/kg i.v. potentiated the thrombolytic effect of kallikrein while aprotinin (100,000 unit/kg, i.v.) abolished it. Thrombolysis by kallikrein was mediated by an unstable principle which was decomposed by blood during 15 min of incubation at 37 degrees C. Generation of this principle was inhibited by pretreatment of animals with aspirin (50 mg/kg, i.v.). The above analysis points to prostacyclin which owing to its platelet-suppressant and fibrinolytic properties induces thrombolysis when released by kinins.

Cardioprotection by molsidomine and iloprost in myocardial ischemia in anaesthetized cats.

In ten anaesthetized open chest cats the ligation of left descending coronary artery (LLDCA) resulted in 80 percent of mortality with survival time of 40.9 +/- 8.6 min. The death of animals was preceded by premature ventricular contractions which occurred with frequency 7.3 +/- 0.6/min after 2.1 +/- 0.3 min following LLDCA. Molsidomine as a dose of 20 micrograms/kg given i.v. to 10 cats 15 min before LLDCA affected none of the above characteristics of acute myocardial ischemia. The pretreatment with iloprost (2 micrograms/kg given i.v. to 10 cats) diminished frequency of premature ventricular contractions and elongated a period of time required for their occurrence after LLDCA, however, iloprost was also unable to diminish mortality of LLDCA cats. Only a combined administration of molsidomine and iloprost significantly diminished mortality among LLDCA cats down to 20 percent and doubled their survival time as compared to the controls. It is concluded that NO-donors (molsidomine) exert permissive effect towards the cardioprotective action of prostacyclin analogues (iloprost) in cats with myocardial ischemia. LLDCA cats pretreated with molsidomine + iloprost enjoyed a more complete cardioprotection than those pretreated with a high dose (3000 units/kg i.v.) of superoxide dismutase.

Occlusion and reperfusion-induced arrhythmias in rats: involvement of platelets and effects of calcium antagonists.

The role of blood platelets in ischemia- and reperfusion-induced arrhythmias and the efficacy of three calcium blocking drugs (verapamil, diltiazem, and nicardipine) in preventing the arrhythmias were investigated. Using anesthetized rats, we measured platelet count (Pc) continuously in vivo with a Technicon autocounter. Thromboxane B2 (TxB2) and 6-keto-PGF1 alpha levels in blood from coronary sinus were determined by radioimmunoassay (RIA). Myocardial ischemia and arrhythmias were monitored from lead I ECG during and after occlusion of the left anterior descending coronary artery (LAD) for 7 min. Ischemia-induced arrhythmias were mainly ventricular ectopic contractions (VECs), whereas reperfusion produced VECs, ventricular tachycardia (VT), and reversible and irreversible ventricular fibrillation (VF). Both ischemia and reperfusion decreased platelet count and increased TxB2 level in blood from the coronary sinus. The effects of the CEBs were determined at two dose levels (0.1 and 0.3 mg/kg). Each calcium entry blocker (CEB), at both dose levels, significantly inhibited ischemia-induced arrhythmias. Verapamil and diltiazem significantly reduced reperfusion-induced VECs, prevented VT and irreversible VF, and reduced the number of animals with reversible VF. Nicardipine in preventing arrhythmias was not very effective at either dose. The CEBs also inhibited both ischemia- and reperfusion-induced decreases in PC with a moderate increase (up to 7%) as compared with levels in sham-operated controls. The CEBs also significantly reduced TxB2 levels in blood from the coronary sinus. These results indicate that ischemia and postischemic reperfusion both induce platelet aggregation in rats. Aggregating platelets release biologically active substances including thromboxane A2 (TxA2) which exacerbates existing ischemia and facilitates generation of arrhythmias. CEBs inhibit platelet aggregation and TxA2 release and enhance PGI2 synthesis, thereby preventing arrhythmias.

Kallikrein-thrombolytic and hypotensive action in cats--preliminary results.

An intravenous injection of kallikrein produced hypotensive and thrombolytic effects in anesthetized cats, using the blood superfused tendon technique. The thrombolytic action of kallikrein was mediated by an unstable substance. The generation of this substance was abolished by either acetylsalicylic acid (ASA) or aprotonin and enhanced by captopril. The hypotensive action of kallikrein was only partially inhibited by ASA. It is proposed that both these pharmacological effects of kallikrein are mediated by bradykinin which in turn releases prostacyclin from the endothelium. However, in contrast to the thrombolytic effect of kallikrein which is totally mediated by prostacyclin the hypotensive action of kallikrein depends not only on prostacyclin but also on another endothelium-derived vasorelaxant, e.g. EDRF.

Prostacyclin and the mechanism of action of defibrotide.

Defibrotide (DEF) is a product of the controlled hydrolysis of DNA from mammalian lungs. In vitro DEF (1-2000 micrograms/ml) neither inhibited the aggregation of human, rabbit and cat platelets nor released PGI2 from cultured porcine aortic endothelial cells. Preincubated with platelet preparations, DEF (10 micrograms/ml) lowered the IC50 for the anti-aggregatory action of PGI2 from 2.1-6.5 nM to 0.6-2.4 nM. In vivo DEF (2-20 mg/kg i.v. or 0.1-10 micrograms/ml locally o.t.) dispersed platelet thrombi in the extracorporeal circulation of anaesthetized heparinized cats. In contrast to an immediate and reversible action of PGI2 (0.3-3 micrograms/kg i.v. or 1-3 ng/ml locally o.t.), the delayed and irreversible dispersion of thrombi by DEF in vivo was inhibited by aspirin (50 mg/kg i.v.), indomethacin (10 mg/kg i.v.) or dexamethasone (2 mg/kg i.v.). After pretreatment with these drugs DEF, although deprived of its own effect on thrombi, still enhanced their dispersion by exogenous and endogenous PGI2. Similarly, at its subthreshold effective concentration (10 ng/ml, o.t.) DEF potentiated the thrombolytic action of PGI2 (1 microgram/kg, i.v.), dazoxiben (1 mg/kg, i.v.) and nicotinamide (100 mg/kg i.v.). It is concluded that DEF potentiates the action of PGI2 on platelets. In vivo DEF may facilitate an interaction between platelets and endothelium, leading to augmented generation of PGI2 or of its biologically active products, which are endowed with fibrinolytic properties.