Biochemical and Pharmacological Properties of SANORG 32701
Comparison With the ‘Synthetic Pentasaccharide’ (SR 90107/ORG 31540) and Standard Heparin
SANORG 32701 is a new sulfated pentasaccharide obtained by total chemical synthesis. It is an analogue of the “synthetic pentasaccharide” (SR 90107/ORG 31540), which represents the antithrombin III (AT-III) binding site of heparin. Like SR 90107, it shows high affinity for human AT-III (Kd=3.7±0.7 nmol/L) and is a potent catalyst of its inhibitory effect with regard to factor Xa (1100±31 versus 850±27 anti-Xa U/mg for SR 90107). SANORG 32701 inhibited thrombin generation occurring via both the extrinsic and intrinsic pathways in vitro. After intravenous or subcutaneous administration to rabbits or rats, SANORG 32701 displayed prolonged anti–factor Xa activity and inhibition of thrombin generation ex vivo. SANORG 32701 was slowly eliminated, showing elimination half-lives between 2.8 and 4.9 hours with different doses. SANORG 32701 displayed antithrombotic activity by virtue of its potentiation of the anti–factor Xa activity of AT-III. It strongly inhibited thrombus formation in an experimental model of thromboplastin-induced venous thrombosis in rats (intravenously) and rabbits (subcutaneously) (ED50 values were 25.5±4.1 and 91±12.7 nmol/kg, respectively). SANORG 32701 inhibited the accretion of fibrinogen I 125 to a preformed thrombus in the rabbit jugular vein and significantly reduced thrombus growth occurring after electrical stimulation of the rabbit carotid artery. In the rabbit, intravenous injection of SANORG 32701 enhanced tissue plasminogen activator (TPA)–induced thrombolysis, suggesting that concomitant use of SANORG 32701 during TPA therapy may be helpful in preventing thrombus accretion, thus facilitating clot lysis. In the rat, SANORG 32701 potently inhibited thrombus formation induced on a silk thread in an arteriovenous shunt and in the vena cava. Compared with standard heparin, SANORG 32701 (1000 nmol/kg IV) caused only minimal bleeding enhancement and exhibited a favorable antithrombotic activity/bleeding risk ratio, therefore showing that it might be considered as a promising compound in the treatment and prevention of various thrombotic diseases.
In the past several years, considerable progress has been made in developing ideal antithrombotic drugs.1 2 Direct thrombin inhibitors such as argatroban and hirudin have demonstrated potent efficacy in various experimental thrombosis models and in several clinical situations.3 4 5 6 More recently, SR 90107, a new pentasaccharide obtained by total chemical synthesis,7 representing the minimal sequence on the heparin chains interacting with AT-III, has been shown to display antithrombotic activity by virtue of its potentiation of the anti–factor Xa activity of AT-III.7 8 This compound, which acts on free factor Xa and not against thrombin, has been shown to possess antithrombotic efficacy in vitro and in various thrombosis models without deleterious effects on hemostasis.9 10 11 12 13 14 Recently, safety, tolerance, and effect kinetics were investigated in healthy volunteers.15 In the present study, no problem was observed concerning clinical tolerance, showing that inhibition of factor Xa might be a promising approach for the treatment of thrombosis. Nevertheless, despite the fact that this compound appears as an interesting candidate, its complex structure requires numerous synthesis steps.7 To simplify the synthesis, analogues of SR 90107/ORG 31540 in which N-sulfate groups have been replaced by O-sulfates and hydroxyl groups by O-alkyl groups have been prepared.7 Surprisingly, these compounds have very similar biological properties, and their easier preparation makes them attractive drug candidates.
In the present study, we describe the in vitro and in vivo activity of SANORG 32701, a newly developed chemical analogue of SR 90107/ORG 31540. SANORG 32701 is an O-alkylated, O-sulfated pentasaccharide showing higher anti–factor Xa activity compared with SR 90107/ORG 31540.
Materials and Methods
Recombinant human tissue factor (Inovin), obtained from Dade (Baxter Diagnostics), was reconstituted with distilled water as indicated by the manufacturer and diluted with saline before use. The tissue factor solution was kept at 0°C for the entire length of the experiment. Tissue thromboplastin and standard heparin (162 IU/mg, sodium salt from pig intestinal mucosa; mean molecular weight, 15 000) were purchased from Sigma Chemical Co. Human fibrinogen I 125 (4.1 MBq/mg) was purchased from Amersham. Factor Xa from human and bovine plasma, purified human AT-III, HC-II, and chromogenic substrates S-2222 and S-2238 were from Kabivitrum. SR 90107/ORG 31540, a pentasaccharide that represents the minimal sequence on the heparin chains, interacting with AT-III and SANORG 32701 (Fig 1⇓) was from Sanofi Recherche. The pharmaceutical research and development on SR 90107/ORG 31540 and SANORG 32701 is being pursued within a partnership agreement between Organon and Sanofi. All compounds were solubilized in saline and administered as indicated. All other chemicals and solvents were reagent grade from Prolabo.
In Vitro Studies
Determination of the Inhibitory Activity of SANORG 32701 for Factor Xa and Thrombin
Anti–factor Xa activities were determined by an amidolytic method (modification of the procedure of Teien and Lie16 ). Bovine or human factor Xa (2.4 nkat/mL) was incubated for 2 minutes with purified human AT-III (0.17 U/mL) at 37°C in the presence of various concentrations of the oligosaccharides in 20 mmol/L Tris-maleate buffer, pH 7.4, and 150 mmol/L NaCl. To measure the residual factor Xa activity, S-2222 substrate (dissolved in [mmol/L] Tris-HCl buffer 50, pH 8.4, NaCl 175, and EDTA 27.5) was added (final, 0.25 mmol/L). The reaction was stopped 2 minutes later by the addition of a 50% aqueous acetic acid solution, and the absorbance at 405 nm was read on a spectrophotometer. The percentage of inhibition was then calculated as follows: inhibition %=100×(OD blank−OD sample)/OD blank. The activity per milligram of the compounds was determined by comparison with a calibrated standard using Excel 4.0 Software (Microsoft).
Anti-Xa Activities Against Rabbit and Rat Factor Xas
Rat and rabbit plasmas were treated by Russel viper venom (50 and 5 μg/mL, respectively) in the presence of CaCl2 (50 mmol/L) to activate factor X into factor Xa. One hundred microliters of the obtained mixture containing factor Xa was then incubated with various concentrations of the oligosaccharides in 20 mmol/L Tris-maleate buffer, pH 7.4, and 150 mmol/L NaCl. To measure the residual factor Xa, S-2222 substrate (dissolved in [mmol/L] Tris-HCl buffer 50, pH 8.4, NaCl 175, and EDTA 27.5) was added (final, 0.25 mmol/L), and the absorbance at 405 nm was read on a spectrophotometer. The percentage of inhibition was calculated using the following formula: inhibition %=100×(OD blank−OD sample)/OD blank. IC50 values were determined using the four-parameter logistic model with a confidence interval of 95%. The adjustment was obtained by nonlinear regression using the Levenberg-Marquard algorithm in RS/1 software.
Purified human α-thrombin (3000 IU/mg, Centre de Transfusion Sanguine) (2 UI/mL) was incubated for 1 minute with HC-II (0.1 U/mL) or AT-III (0.25 U/mL) at 37°C in the presence of the oligosaccharides. To measure the residual thrombin activity, a 1 mmol/L solution of S-2238 chromogenic substrate was added (0.1 mL). The reaction was stopped 1 minute later by the addition of a 50% aqueous acetic acid solution, and the absorbance at 405 nm was read.
Determination of the Anticoagulant Activity of SANORG 32701
APTT was measured with Actin FS (Dade-Baxter) in a Amelung-Baxter KC10 coagulometer. The TG methodology was adapted from Hemker et al,17 as previously reported by us.18 TG was triggered either by kaolin (final, 5 μg/mL) (intrinsic TG) or by rabbit brain thromboplastin (extrinsic TG) in defibrinated platelet-poor plasma supplemented with 2 μmol/L cephalin. Aliquots were taken every 15 seconds, and the concentration of active thrombin was measured using the chromogenic substrate S-2238. The total amount of thrombin generated was quantified by computing AUC. IC50 values were determined using the four-parameter logistic model with a confidence interval of 95%. The adjustment was obtained by nonlinear regression using the Levenberg-Marquard algorithm in RS/1 software (BBN).
Affinity for Human, Rat, and Rabbit AT-III
AT-III from human, rabbit, and rat plasmas was purified by affinity chromatography on a heparin-agarose column and passage through a Q Sepharose fast-flow column.19 Affinity of SANORG 32701 for AT-III of the various animals species was determined by fluorescence, as described by Atha et al,20 using a Perkin-Elmer LS-50 type spectrofluorimeter at an excitation wavelength of 280 nm and an emission wavelength of 338 nm at 37°C, under continuous stirring. Oligosaccharides were added into 2 mL of 0.01 mmol/L Tris-HCl buffer, pH 7.0, containing 0.15 mol/L NaCl and 5 to 60 nmol/L AT-III. The ratio and the concentrations of the AT-III–oligosaccharide complex were calculated, and dissociation constants (Kds) were determined from Scatchard plots using RS/1 software.
In Vivo Studies
Ex Vivo Pharmacological Activities
Male New Zealand rabbits (2.5 to 3 kg, Lago, Vonnas, France) or Sprague-Dawley rats (250 to 300 g, Iffa Credo, L'Arbresle, France) were anesthetized by an intravenous injection of sodium pentobarbital (30 mg/kg) and treated by subcutaneous or intravenous injections of SANORG 32701 or SR 90107. Arterial blood samples (2 mL) were withdrawn in a 3.8% trisodium citrate solution (1/9 [vol/vol]). Each blood sample was centrifuged (1800g, 10 minutes), and the platelet-poor plasma was stored at −20°C. Anti-Xa activity of SANORG 32701 and its effects on TG and APTT were determined as described above.
Stasis-Induced Venous Thrombosis in the Rabbit
Rabbits were anesthetized by an intravenous injection of sodium pentobarbital (30 mg/kg). Stasis-induced venous thrombosis was induced according to Buchanan et al21 with slight modifications. Each jugular vein was isolated, and two loose sutures were placed 2 cm apart. Test compounds or placebos were administered intravenously through a marginal ear vein 5 minutes or orally by gavage 2 hours before ligation of the jugular veins. Recombinant human tissue factor (1 ng/kg) was injected 5 minutes before the induction of stasis. Both jugular vein segments were occluded by the distal and proximal sutures, and stasis was maintained for 15 minutes. The veins were opened longitudinally, and the thrombus, if apparent, was removed, blotted on filter paper, and weighed. Wet weights of thrombi were averaged for left and right jugular veins. Test compounds or the vehicle was administered subcutaneously 1 hour before the intravenous injection of tissue factor.
Stasis-Induced Thrombosis in the Rat
Thrombus formation by a combination of stasis and hypercoagulability was induced as described by Vogel et al.9 Male Sprague-Dawley rats (250 to 300 g) were anesthetized with sodium pentobarbital (30 mg/kg IP). The abdomen of each animal was surgically opened, and after careful dissection, the vena cava was exposed and dissected free from surrounding tissue. SANORG 32701 or SR 90107 was administered intravenously 5 minutes before thrombosis induction. Two loose sutures were prepared 0.7 cm apart on the inferior vena cava, and all collateral veins were ligated. Human tissue factor (1 ng/kg IV) was injected into the dorsal vein of the penis. Ten seconds after the end of the injection, stasis was established by tightening the two sutures, first the proximal and then the distal. The abdominal cavity was provisionally closed, and stasis was maintained for 20 minutes. The cavity was then reopened, the ligated segment was opened longitudinally, and the thrombus formed was removed, rinsed, blotted on filter paper, dried overnight at 60°C, and weighed. Under these experimental conditions, control thrombus weight was 6.1±0.2 mg (n=20).
Rat Vena Cava Model
Male Wistar rats (250 to 350 g, Harlan, Zeist, Netherlands) were anesthetized intraperitoneally with sodium pentobarbital (Nembutal, 60 mg/kg, Sanofi), and body temperature was maintained between 36.5°C and 38.5°C. A silk thread was inserted into a cannula (9-cm PE-50, Clay Adams) to be used for guidance at insertion. The femoral vein was isolated from the artery, 2 cm distal from the vena cava and up to the iliac vein, and was closed just proximal to the superficial epigastric vein. All side branches of the femoral vein were thermically closed, and the cannula containing the silk thread was inserted and led via the femoral and iliac veins up to 5 cm in the vena cava. Immediately thereafter, the silk thread was pushed forward an additional 7 cm. Then, the whole cannula plus thread was withdrawn, leaving 5 cm of thread in the vena cava, and the femoral vein and the artery were closed. Compounds were administered by intravenous bolus (1 mL/kg) in the penis vein 1 minute before insertion of the silk thread. After 15 minutes, blood (0.7 mL) was sampled in EDTA from the left common carotid artery for platelet counting, and the silk thread with the thrombus was carefully pulled out and rinsed, and the 5-cm piece that had been positioned in the vena cava was cut off and weighed. Control thrombus weights amounted to 30.5±2.2 mg (mean±SEM, n=12).
Fibrinogen I 125 Accretion on a Preformed Thrombus
The antithrombotic effect of SANORG 32701 was assessed by measuring its ability to inhibit the accretion of fibrinogen I 125 onto autologous nonradioactive venous thrombi preformed in the jugular veins of rabbits, as described by Boneu et al.22 New Zealand rabbits (2.5 to 3 kg) were anesthetized with sodium pentobarbital (30 mg/kg IV). Both jugular veins were exposed, and a 2-cm segment was isolated on either side. Each jugular vein segment was emptied of blood, and blood flow was temporally occluded by proximal and distal clamps. One milliliter of blood was collected from a carotid cannula and mixed with 50 μL thrombin (20 U/mL), and 150 μL of clotting blood was immediately injected into the isolated segment. A 10-cm length of silk thread was passed longitudinally through the forming thrombus and the vessel wall to keep the thrombus in place. Then flow was restored. Fifteen minutes after the thrombus was formed, each animal was injected with 20 μCi of fibrinogen I 125. Five minutes later, each rabbit was injected with a bolus dose of SANORG 32701 or SR 90107 or an equivolume of saline, followed by a continuous infusion of the pentasaccharides for 4 hours. At the end of the 4-hour infusion, the venous segments containing the thrombi were tied off and slit open longitudinally, and the remaining thrombi were removed. The specific activity of the whole-blood fibrinogen was estimated from the mean of blood samples taken at hourly intervals throughout the infusion. The ratio of radioactivity of the thrombus to circulating fibrinogen radioactivity was used to quantify thrombus size.
Electrical Stimulation of the Carotid Artery
Thrombus formation was induced by electrical stimulation of the carotid artery according to the method described by Hladovec.23 Male New Zealand rabbits (2.5 to 3 kg) were anesthetized with sodium pentobarbital (30 mg/kg IV). A segment of the left carotid artery (≈10 mm long) was exposed and dissected free of surrounding tissue. A small piece of insulating film (Parafilm M) and two stainless steel electrodes were inserted under the artery. Using a constant DC power supply (Apelex 3500), the artery was stimulated at 2.5 mA for 3 minutes. The thrombotic occlusion was monitored by measuring the blood flow cranially (downstream) from the site of electrical stimulation. The blood flow was measured at 5-minute intervals over a 45-minute period. SANORG 32701 and SR 90107 were administered intravenously 5 minutes before electrical stimulation of the carotid artery. The antithrombotic effect of the various compounds was determined 45 minutes after administration of the drugs.
Thrombosis on a Silk Thread in an Arteriovenous Shunt in the Rat
The antithrombotic activity of SANORG 32701 was determined in an arteriovenous shunt as described by Umetsu and Sanai.24 Two 12-cm polyethylene tubings (0.6- and 0.9-mm inner and outer diameter, respectively) linked to a central part (6 cm long, 0.9-mm inner diameter) containing a 5-cm silk thread and filled with a heparin saline solution (50 IU/mL) were placed between the right carotid artery and the left jugular vein of pentobarbital-anesthetized rats 5 minutes after an intravenous administration of SANORG 32701 or SR 90107. The central part of the shunt was removed after 20 minutes of blood circulation, and the silk thread carrying the thrombus was pulled out. The wet weight of the thrombus was determined. Under these experimental conditions, control thrombus weight was 36.1±1.9 mg (n=10).
The thrombolytic effect of TPA, in association with SANORG 32701, was evaluated by the lysis of standard-sized preformed fibrinogen I 125–labeled thrombi produced in the external jugular vein of rabbits.25 The jugular vein of a pentobarbital-anesthetized rabbit was uncovered, and all tributaries were ligated at a distance of 4 cm from the main bifurcation of the external jugular and facial veins. A silk thread was then inserted through the vessel in order to anchor the thrombus and avoid embolization. After ligation of the vein with two vessel clamps, the enclosed blood was removed and exchanged with 200 μL of citrated rabbit blood mixed with 0.5 μCi human fibrinogen I 125 and 10 IU bovine thrombin. After clot formation (30 minutes), the surgical clamps were removed, and the blood flow was restored. The infusion was given via a contralateral marginal ear vein over time periods as indicated. At the end of the TPA infusion period (4 hours), the thrombus was carefully removed from the vein and weighed. The amount of radioactivity remaining in the clot was determined with a gamma counter (1261 Multigamma counter, Wallac). The extent of fibrinogen I 125 lysis was calculated at the end of the infusion as the difference between the radioactivity originally incorporated in the thrombus and that remaining in the residual thrombus. Percent decrease of the thrombus-associated radioactivity was calculated with reference to the initial fibrinogen I 125 added into the initial clot.
Bleeding in the Rabbit
Rabbits were anesthetized with sodium pentobarbital (30 mg/kg IV). SANORG 32701, SR 90107, or standard heparin was administered intravenously, and five standardized incisions were made through the ear with a scalpel blade (No. 21, Swann-Norton). Care was taken to avoid any macroscopically visible vessel. The ear was immersed in a 500-mL saline bath at 37°C with continuous stirring. Blood loss was determined 10 minutes later by measurement of the hemoglobin content using a spectrophotometric method. The determination of the hemoglobin content of the water bath was performed after the addition of hemolizing reagent (Zapoglobin, Coultronics). The volume of blood contained in the sample was deduced from a standard curve, establishing the relation between blood volume and the corresponding absorbance at OD of 540 nm.
Bleeding Time in the Rat
Bleeding time was determined by transection of the tail, 5 mm from the tip, of pentobarbital-anesthetized rats (30 mg/kg IP). SANORG 32701, SR 90107, or standard heparin was injected intravenously at the indicated doses 15 seconds before tail transection. Blood was carefully blotted every 15 seconds on a filter paper. Hemostasis was considered to be achieved when no blood stain was observed over 1 minute.
The results shown are arithmetic means. The doses of drugs that inhibited 50% of the thrombus formation were calculated from dose-response curves. Grouped data were analyzed for significance by comparison with the vehicle-treated group using the Mann-Whitney test with Holm-Bonferroni adjustment, with P<.05 indicating a significant difference.
In Vitro Studies
Anti–Factor Xa Activity of SANORG 32701 In Vitro
Using a synthetic oligopeptide substrate (S-2222), SANORG 32701, in the presence of AT-III inhibited, in a dose-dependent manner, the activity of human, rabbit, and rat factor Xas. The IC50 values (concentrations that inhibit 50% of the enzyme activity) were 36±4, 28±4, and 26±2 nmol/L (n=6) with regard to human, rabbit, and rat factor Xas, respectively (Table 1⇓). Under the same experimental conditions, SR 90107 showed a lower inhibitory activity (Table 1⇓). It exhibited a specific activity of 850±27 versus 1100±31 anti-Xa U/mg for SANORG 32701. Affinity constants (Kds) of SANORG 32701 for human, rat, and rabbit AT-IIIs were in the same concentration range, with affinities of SANORG 32701 for AT-III of the various species being 10- to 20-fold higher than that measured for SR 90107 (Table 1⇓). In the presence of AT-III or HC-II, neither SANORG 32701 nor SR 90107 exhibited antithrombin activity (Table 1⇓).
Effect of SANORG 32701 on TG
As shown in Fig 2⇓, SANORG 32701 inhibited TG in human plasma. As already shown for SR 90107,18 SANORG 32701 impaired TG occurring through the extrinsic pathway more efficiently than that occurring through the intrinsic pathway, showing respective IC50 values of 0.3±0.05 and 2.2±0.3 μmol/L (n=6). A similar effect of SANORG 32701 on TG occurred in factor VII– or factor IX–depleted human plasma (not shown), therefore confirming that it is a selective factor Xa inhibitor.
In Vivo Studies
Anti–Factor Xa and Anticoagulant Activities of SANORG 32701 Ex Vivo
The anti–factor Xa and anticoagulant activities of SANORG 32701 were examined after subcutaneous injection into rabbits. After subcutaneous administration, anti–factor Xa activity in rabbit plasma increased in a dose-dependent manner (Fig 3A⇓). The mean Cmax, t1/2, and AUC were related to the dose injected and progressively increased as a consequence of a distribution phase, becoming more apparent at the higher doses (Table 2⇓). After subcutaneous injection, SANORG 32701 slightly but significantly prolonged APTT even after the injection of the lowest dose (30 nmol/kg SC) (Fig 3B⇓). This effect, which occurred in a dose-dependent manner, lasted for at least 12 hours at the highest dose. As already observed in vitro, SANORG 32701 inhibited TG ex vivo (Fig 3C⇓). After the subcutaneous administration of 100 nmol/kg, TG was significantly inhibited for at least 6 hours. On all of the parameters studied, SR 90107 administered at a dose of 100 nmol/kg was less efficient than SANORG 32701 administered at the same dose (Fig 3⇓). The results obtained after intravenous bolus injection to rats and rabbits are summarized in Table 2⇓. They are consistent with the results obtained after subcutaneous injection in rabbits. They also show that in the rat, SANORG 32701 is significantly more potent than SR 90107, but this is probably the result of a lower elimination rate in this species (Fig 4⇓). Moreover, these results show that the subcutaneous bioavailability of SANORG 32701 is near 100%. After nephrectomy, the elimination rates after intravenous administration in rats increased to a level close to that of endogenous AT-III (11.8 hours), showing that the plasma elimination of SANORG 32701 was largely due to renal clearance (not shown). This confirmed earlier observations26 showing that binding to AT-III regulates renal clearance; this role of AT-III becomes more important if renal clearance is prevented.
Stasis-Induced Thrombosis After Injection of Tissue Factor in the Rabbit
Using a combination of a thrombogenic challenge (1.0 ng/kg of tissue factor) and stasis, SANORG 32701 was tested for its ability to affect thrombus formation in a venous thrombosis model in the rabbit. SANORG 32701 administered subcutaneously 2 hours before thrombosis induction displayed a significant dose-dependent antithrombotic effect, with maximum inhibition of thrombosis being observed at a dose of 300 nmol/kg (Fig 5⇓). The ED50 value was 91±12.7 nmol/kg (n=10). In this experimental model, SR 90107 and standard heparin were as active as SANORG 32701, showing ED50 values of 74±6.1 and 83±10.2 nmol/kg, respectively (n=10). Higher doses of standard heparin, however, were able to totally inhibit thrombus growth, whereas the pentasaccharides were not.
Stasis-Induced Thrombosis After Injection of Tissue Factor in the Rat
After intravenous injection, SANORG 32701 prevented stasis-induced thrombosis in a dose-dependent manner (Fig 6⇓). The ED50 value was 25.5±4.1 nmol/kg. Under the same experimental conditions, SR 90107 and standard heparin inhibited tissue factor+stasis–induced thrombosis with ED50 values of 39.6±9.2 nmol/kg (Fig 6⇓) and 8.8±2.1 nmol/kg.
Thrombus Formation in the Rat Vena Cava
Fig 7⇓ shows that both SR 90107 and SANORG 32701 inhibited, in a dose-dependent manner, thrombus formation and platelet count reduction, which occurred simultaneously as a consequence of platelet accretion to the forming thrombus. The ED50 values with regard to thrombus formation were 43±8 and 31±8.5 nmol/kg for SR 90107 and SANORG 32701, respectively. Although not significantly different, the difference observed for the two compounds reflected mean values in specific anti–factor Xa activity per mole. In this particular model, standard heparin was less potent than both pentasaccharides.
Fibrinogen I 125 Accretion on a Preformed Thrombus
The effects of SANORG 32701, SR 90107, and heparin on inhibition of fibrinogen I 125 accretion onto the preformed thrombi are shown in Fig 8⇓. In saline-treated animals, 5.5±0.8 μg (n=9) of fibrinogen I 125 was accreted onto the preformed thrombi after 4 hours. SANORG 32701 and SR 90107 tested at the same doses were equipotent in inhibiting fibrinogen I 125 accretion onto the preformed thrombus. Under the same experimental conditions, standard heparin was more potent than either SR 90107 or SANORG 32701 (Fig 8⇓).
Thrombus Formation After Electrical Stimulation of the Rabbit Carotid Artery
Electrical stimulation of the carotid artery in rabbits treated with the vehicle resulted in the rapid accumulation of thrombotic material at the injured surface, leading to partial occlusion of the luminal surface of the artery within 45 minutes. This was evidenced by a progressive fall in blood flow, measured downstream from the site of electrical stimulation. In 80% of the animals, drop of blood flow was maximal 20 to 30 minutes after the stimulation and reached a maximum at 45 minutes. Indeed, within this time period, the blood flow fell 80±15% (mean±SD) (n=10). In unstimulated arteries, no variation of blood flow could be detected, thus showing that no thrombus formation occurred (not shown). The thrombus formation was strongly reduced by SANORG 32701, administered intravenously 5 minutes before the electrical stimulation of the carotid artery (Fig 9⇓). SANORG 32701 (300 nmol/kg IV) inhibited thrombus formation by 68.7±16% (P<.001, n=6). SR 90107, administered intravenously 5 minutes before the electrical stimulation of the carotid artery, significantly reduced the incidence of thrombotic occlusion at a dose of 600 nmol/kg (32±15% inhibition, P<.05) (Fig 9⇓). Under the same experimental conditions, standard heparin also inhibited thrombus formation and was as potent as SANORG 32701.
Arteriovenous Shunt in the Rat
The intravenous administration of SANORG 32701 and SR 90107, 5 minutes before thrombosis induction, reduced in a dose-dependent manner the thrombus formation in an arteriovenous shunt (Fig 10⇓). Under the same experimental conditions, standard heparin was much more potent compared with both SR 90107 and SANORG 32701 and, at the highest doses, inhibited thrombus growth to a greater extent.
Effect of SANORG 32701 on TPA-Induced Thrombolysis
The extent of thrombolysis expressed as the isotope recovery balance under continuous infusion of TPA, with or without bolus injections of SANORG 32701, is shown in Fig 11A⇓. In the control group, infused with saline for 4 hours, the average degree of thrombolysis was 18.9±2.0% (mean±SEM, n=10), as a result of spontaneous lysis, washout of unclotted radiolabeled material, or both. This effect was similar to that observed in the absence of TPA after a single bolus injection of SANORG 32701 (300 nmol/kg IV). Systemic infusion of TPA (0.5 mg/kg IV) resulted in a 43.1±9.5% (n=5) fibrinogen I 125 lysis, which was significantly enhanced (P<.05) by a bolus intravenous injection of SANORG 32701 (30 nmol/kg IV). Better efficacy of SANORG 32701 was observed in the decrease of the thrombus weight. Indeed, thrombus weights fell significantly from a control value of 134.8±9.1 to 58.7±8.6 mg (n=8) in the TPA-treated group, whereas concomitant treatment with SANORG 32701 reduced thrombus weights in a dose-dependent manner (Fig 11B⇓).
Hemorrhagic Effect of SANORG 32701 in the Rabbit and Rat
Hemorrhagic risk associated with a treatment with SANORG 32701 was determined using two different experimental models of bleeding in the rabbit and rat. In the rabbit, the intravenous administration of standard heparin significantly increased blood loss, whereas SANORG 32701 or SR 90107, even when administered at a high dose (1000 nmol/kg IV), did not affect bleeding (Fig 12A⇓). Similar results were obtained after transection of the tail of rats (Fig 12B⇓).
Since thrombin inhibitors have been shown to significantly increase the risk of abnormal bleeding,27 an alternative strategy to prevent thrombin formation is the inhibition of factor Xa, and it is now well admitted that this represents an attractive approach for clinical intervention in various thrombotic disorders.28 Results obtained in the present study confirm and extend these latter observations but also clearly demonstrate that synthetic analogues of the unique AT-III–binding sequence present on the heparin chains8 9 10 11 12 13 14 represent a new class of interesting antithrombotic agents. SANORG 32701 interacts with human, rat, and rabbit AT-IIIs with a dissociation constant in the nanomolar range (the affinity for human AT-III being >10 times higher than that observed for SR 90107). Binding of the pentasaccharides to AT-III induces a conformational change, which results in an accelerated inactivation of factor Xa.29 Thus, SANORG 32701 is a potent and selective AT-III–dependent inhibitor of human, rat, and rabbit factor Xas, showing IC50 values that compared favorably with that of the “original” pentasaccharide, SR 90107. Because the Kd values of SR 90107 or SR 32701 are much smaller than the concentration of AT-III in the assay mixture or in plasma, the concentration of the binary AT-III–pentasaccharide complex formed in these conditions is practically equal to the pentasaccharide concentrations. Therefore, in spite of its much higher affinity for AT-III, the anti–factor Xa and the resulting overall anticoagulant effect elicited by SANORG 32701 appeared to be only slightly enhanced compared with that elicited by SR 90107. SANORG 32701 efficiently impaired TG in human plasma, with IC50 values in the micromolar range. Most interestingly, the extrinsic pathway appeared to be more sensitive to SANORG 32701 than the intrinsic one. Similar results were obtained with SR 90107.18 Since factor Xa has been shown to be involved in the activation of factor VII into VIIa,30 one possible origin of this difference could be the indirect impairment of VIIa formation by factor Xa–inactivating compounds such as SR 90107 or SANORG 32701. Alternatively, the recently discovered AT-III–mediated inhibition of the factor VIIa–tissue factor complex by the pentasaccharides31 might also account for the observed greater susceptibility of the coagulation extrinsic pathway to inhibition by SANORG 32701.
Since these in vitro studies suggest that SANORG 32701 fulfills the requirements for a disease-modifying factor Xa inhibitor, we investigated its activity in vivo. We first evaluated the pharmacokinetics of a subcutaneous or intravenous administration of SANORG 32701 by means of its anti-Xa activity in plasma. SANORG 32701 exhibited a strong dose-dependent anti-Xa activity after subcutaneous and intravenous administration in rabbits and rats. This effect, which occurred in a dose-dependent manner, almost completely paralleled APTT prolongation and inhibition of TG. These data are consistent with the data of others involving similar compounds,26 32 which showed that there was a close relationship between the plasma half-lives of anti–factor Xa pentasaccharides and their respective affinities for AT-III. The antithrombotic potency of this compound was then compared with that of SR 90107. The role of factor Xa bound to fibrin and platelets and its subsequent effect on prothrombin activation have been emphasized, showing that these complex phenomena ensure the progressive growth of the thrombus over several hours.27 33 SANORG 32701 displayed marked dose-dependent antithrombotic activity in different experimental models of venous and arterial thrombosis in rats and rabbits. These were (1) stasis-induced venous thrombosis in rats and rabbits in which thrombus formation occurred because of a combination of high-grade stenosis and thrombogenic challenge, (2) fibrinogen accretion to a preformed thrombus in the rabbit jugular vein, (3) thrombotic occlusion of the rabbit carotid artery following injury, (4) thrombosis on a silk thread in an arteriovenous shunt or in the vena cava in rats in which thrombus formation occurred on a foreign surface, and (5) potentiation of TPA-induced thrombolysis of a thrombus implanted in the rabbit jugular vein. Despite the fact that these various models were either arterial or venous (ie, models in which platelets or coagulation played preponderant roles), the effective antithrombotic doses were in the same range. Moreover, the curves describing the relationship between the doses, the subsequent plasma concentrations, and the resultant antithrombotic effect were parallel. This finding suggests that only one mechanism is involved in the antithrombotic property of SANORG 32701, namely, its pure anti–factor Xa effect. One important observation was that compared with standard heparin, the relative efficacy of SANORG 32701 varied depending on the experimental model used. Such an effect has already been reported for small antithrombotic oligosaccharides such as SR 90107,9 10 34 but the reasons for such differences in efficacy remain obscure. In a recent in vitro study, Prager et al35 provided evidence that the procoagulant activity of clots is primarily dependent on de novo activation of prothrombin mediated by clot-associated factor Xa. These authors postulated that this could be the molecular mechanism for the observed efficacy of factor Xa inhibitors in preventing occlusion. Since it is now generally admitted that because of steric hindrance, heparin cannot inhibit clot-bound thrombin or factor Xa, one can also imagine that because of their smaller size, anti–factor Xa pentasaccharides will exhibit higher efficacy with regard to fibrin-bound factor Xa. Such an assumption is now under investigation in our laboratory. Another hypothesis is that such small oligosaccharides might not be affected by platelet proteins released after platelet activation, which occurs during thrombus formation. These proteins, such as platelet factor 4, are known to neutralize some of the effect of standard heparin, whereas they do not interfere with SANORG 32701 and SR 90107 (not shown).
These studies therefore clearly demonstrate that SANORG 32701 and also SR 90107 are efficacious antithrombotic drugs after intravenous or subcutaneous administration. Indeed, both of these drugs inhibited several types of experimental venous and arterial thromboses and potentiated TPA-induced thrombolysis where factor Xa generation seemed important. Additionally, SANORG 32701 and SR 90107 significantly reduced thrombosis without affecting bleeding. Since thrombin not only converts fibrinogen into fibrin to form clots but also strongly induces platelet aggregation, direct thrombin inhibitors impair hemostasis to some extent at or beyond the effective dose in thrombotic models,5 36 37 38 which may be related to their antiplatelet activities. In contrast to thrombin inhibitors, SANORG 32701 and SR 90107, which did not affect platelet aggregation (data not shown), would not compromise the hemostatic response of platelets, thereby indicating the possibility of using them as conjunctive agents for thrombolytic therapy or adjunctive agents with antiplatelet drugs without the risk of abnormal bleeding.
Thus, these compounds appear to be promising antithrombotic agents that result in a reduced bleeding tendency, but further clinical evaluation is necessary.
Selected Abbreviations and Acronyms
|APTT||=||activated partial thromboplastin time|
|AUC||=||area under TG curve|
|HC-II||=||heparin cofactor II|
|TPA||=||tissue plasminogen activator|
The authors thank T. Barzu for her expert assistance in defining the affinity of SANORG 32701 and SR 90107 for AT-III.
- Received January 23, 1996.
- Accepted May 22, 1996.
Van Boeckel CAA, Petitou M. The unique antithrombin binding domain of heparin: a lead to new synthetic antithrombotics. Angew Chem Ind (Ed Engl). 1993;32:1671-1690.
Boneu B, Necciari J, Magnani H, Kieffer G, Mant T, Morrison P, Singer B, Lamond G, Dickinson J. Pharmacokinetics of the natural pentasaccharide (SR 90107/ORG 31540) with high affinity for antithrombin III following subcutaneous administration in man. Thromb Haemost. 1993;69:406. Abstract.
Buchanan MR, Boneu B, Ofosu FA, Hirsh J. The relative importance of thrombin and factor Xa to the antithrombotic effect of heparin. Blood. 1985;65:198-201.
Collen D, Stassen JM, Verstraete M. Thrombolysis with human extrinsic (tissue-type) plasminogen activator in rabbits with experimental jugular vein thrombosis: effect of molecular form and dose of activator, age of the thrombus, and route of administration. J Clin Invest. 1983;71:368-376.
Van Amsterdam RGM, Vogel GMT, Visser A, Kop WJ, Buiting MT, Meuleman DG. Synthetic analogues of the antithrombin III-binding pentasaccharide sequence of heparin: prediction of the in vivo residence times. Arterioscler Thromb Vasc Biol. 1995;15:495-503.
Maffrand JP. Direct thrombin inhibitors. Nouv Rev Fr Hematol. 1992;34:405-419.
Olson ST, Bjo¨rk I, Sheffer R, Craig PA, Shore JD, Choay J. Role of the antithrombin-binding pentasaccharide in heparin acceleration of antithrombin-proteinase reactions. J Biol Chem. 1992;267:12528-12538.
Masys DR, Bajaj SP, Rapaport SI. Activation of human factor VII by activated factors IX and X. Blood. 1982;60:1143-1150.
Lormeau JC, Herault JP, Herbert JM. The synthetic pentasaccharide representing the heparin binding site to antithrombin III inhibits factor VIIa-tissue factor complex. Thromb Haemost. In press.
Carrie D, Caranobe C, Saivin S, Houin G, Petitou M, Lormeau JC, Van Boeckel C, Meuleman D, Boneu B. Pharmacokinetic and antithrombotic properties of two pentasaccharides with high affinity to antithrombin III in the rabbit: comparison with CY 216. Blood. 1994;84:2571-2577.
Eisenberg PR, Jeffrey ES, Abendschein DR, Miletich JP. Importance of factor Xa in determining the procoagulant activity of whole-blood clots. J Clin Invest. 1993;91:1877-1883.
Prager NA, Abendschein DR, McKenzie CR, Eisenberg PR. Role of thrombin compared with factor Xa in the procoagulant activity of whole blood clots. Circulation. 1995;92:962-967.
Just M, Tripier D, Seiffge D. Antithrombotic effects of recombinant hirudin in different animal models. Haemostasis. 1991;21:80-87.
Jackson CV, Crowe G, Frank JD, Wilson HC, Coffman WJ, Utterback BG, Jakubowshi JA, Smith GF. Pharmacological assessment of antithrombotic activity of the peptide inhibitor, D-methyl-phenylalanyl-prolyl-arginyl (GYKI-14766), in a canine model of coronary artery thrombosis. J Pharmacol Exp Ther. 1992;261:546-552.