M. Ieko, Department of Internal Medicine, Health Sciences University of Hokkaido, 1757-Kanazawa, Ishikari-Tobetsu, Hokkaido, 061-0293, Japan. Tel.: +81 1332 3 1211; fax: +81 1332 3 1534; e-mail: email@example.com
Summary. DX-9065a and JTV-803, synthetic selective inhibitors of activated factor X (FXa), have recently been demonstrated as strongly effective antithrombotic agents in animal thrombosis models, yet with a low risk of bleeding. The aim of the present study was to elucidate these characteristics. Using a chromogenic assay with purified coagulation factors, 73.9% of thrombin generation was suppressed by the addition of DX-9065a (0.20 µm) and 75.7% by JTV-803 (0.18 µm). Inhibition by argatroban (0.19 µm) was less (36.0%) and initial thrombin forming time (T50), the time required to generate 50% thrombin activity in vitro, which is considered important for platelet aggregation in hemostasis, was significantly prolonged by argatroban. In contrast, DX-9065a and JTV-803 had no apparent influence on T50, suggesting that initial thrombin was formed immediately, as in the control. We also investigated platelet aggregation in defibrinated plasma induced by tissue factor, to clarify whether initial thrombin contributes to hemostasis. Aggregation was not affected by the addition of either FXa inhibitor, whereas it was significantly reduced by argatroban. Our results suggest that initial thrombin, which is formed despite the presence of a FXa inhibitor, can activate platelets. We concluded that DX-9065a and JTV-803 are able to inhibit thrombin generation significantly without affecting the formation of initial thrombin for platelet activation, which may contribute to hemostasis through the preservation of normal bleeding time.
The goal of most current antithrombotic strategies is to control the action or generation of thrombin, an enzyme in the coagulation system that causes fibrin clotting. For the treatment and prevention of thromboembolic disorders, heparin, warfarin, and direct thrombin inhibitors are widely employed as antithrombotic agents. However, they are well known to lead to a bleeding tendency as a complication of their use.
Recently, a series of selective inhibitors of activated blood coagulation factor X (FXa) that act independently of antithrombin have been shown to be strongly effective antithrombotic agents in animal models. Of these, DX-9065a (2S-2-[4-[[(35S)-1-acetimidyl-3-pyrrolidinyl]oxy]phenyl]-3-[7-amidino-2-naphyl]propanoic acid hydrochloride pentahydrate) is a highly selective and competitive inhibitor of FXa [1,2], as its estimated dissociation constant (Ki) for FXa was reported to be 41 nM, while that for thrombin is > 2000 µM . This synthetic inhibitor has been found to exert effective protection against experimental tumor-induced disseminated intravascular coagulation (DIC) in rats and suggested to improve the hypercoagulable state induced by the progress of a solid tumor . Following DX-9065a, another synthetic FXa inhibitor, JTV-803 (4-[(2-amidino-1, 2, 3, 4-tetrahydroisoquinolin-7-yloxy) methyl]-1-(4-pyridinyl) piperidine-4-carboxylic acid monomethanesulfonate trihydrate), was developed. The Ki of JTV-803 for FXa was found to be 19 nM, with an effective dose ranging from approximately 0.1 to 0.5 µM . This agent has been shown to inhibit thrombus formation in an arteriovenous shunt model , and was also demonstrated to be effective for treating both lipopolysaccharide-induced and tissue factor (TF)-induced DIC in rat models .
DX-9065a is considered to be a new type of antithrombotic agent with few hemorrhagic effects that does not prolong bleeding time [7–9], probably because it does not inhibit platelet activation , though its mechanism with platelet function has not been clarified. To investigate their antithrombotic and hemorrhagic properties, we studied the effects of the FXa inhibitors DX-9065a and JTV-803 on thrombin generation by prothrombinase, which consists of factor (F)Va, FXa, and phospholipids, as well as on initial thrombin forming time, which is considered important for platelet aggregation in hemostasis.
Materials and methods
DX-9065a was obtained from Daiichi Pharmaceutical Co. Ltd. (Tokyo, Japan) and JTV-803 came from Central Pharmaceutical Research Institute, Japan Tobacco (Takatsuki, Japan). Low-molecular-weight heparin (LMWH) (dalteparin sodium; Kissei, Matsumoto, Japan) and argatroban, a selective thrombin inhibitor (Daiichi Pharmaceutical), were also used. Purified human prothrombin, coagulation FV, and FXa were obtained from Calbiochem (La Jolla, CA, USA). Reptilase was obtained from Zeria Pharmaceutical (Tokyo, Japan) and purified human antithrombin (Anthrobin P) was obtained from Aventis Pharma (Tokyo, Japan). Platelin (Organon Teknika, Durham, NC, USA), a phospholipid substitute for platelets, was used as the phospholipid source.
Preparation of defibrinated plasma
Pooled plasma from healthy donors who had not taken any medications within the previous 2 weeks was mixed with a one-tenth volume of reptilase solution (1 Klobusitzky unit). The mixture was incubated at 37 °C for 10 min, followed by 10 min at 4 °C. Fibrin clots were discarded by centrifugation at 1800 × g for 15 min at 4 °C and the defibrinated plasma was stored at − 80 °C until use in the experiments.
Preparation of washed platelets
Washed platelets were prepared from platelet-rich plasma (PRP) according to the procedure described by Huang . Briefly, fresh whole blood was obtained from healthy donors, anticoagulated with 3.8% sodium citrate, and centrifuged at 800 × g for 10 min to prepare PRP. Platelets were separated from PRP by centrifugation at 1500 × g for 10 min in the presence of 2 mm EDTA, then washed in Ca2+-free Tyrode's buffer (137 mm NaCl, 2.7 mm KCl, 0.5 mm NaH2PO4, 12 mm NaHCO3, 1 mm MgCl2, 5.6 mm glucose) containing 0.2 mm EDTA, and collected again by centrifugation. Washed platelets were finally suspended in Ca2+-free Tyrode's buffer at a platelet count of 40 × 104 µL−1.
Prothrombin time (PT) and activated partial thromboplastin time (APTT) were measured with a coagulometer (ST-4; Diagnostica Stago, Asniere, France). PT was determined by adding 100 µL of PT reagent (Thrombocheck PT; International Reagent Corp., Kobe, Japan) to the sample, which consisted of 40 µL of pooled plasma from the healthy donors and 10 µL of either saline or a saline solution with an inhibitor (DX-9065a, JTV-803, or argatroban at a final concentration of 0.05–0.5 µm). For the measurement of APTT, 50 µL of the mixture of pooled plasma (4 volumes) and either saline or the inhibitor (1 volume) were added to 50 µL of APTT reagent (Thrombocheck aPTT; International Reagent Corp.) in a plastic tube. A coagulation reaction was started by the addition of 50 µL of 20 mm CaCl2. Each experiment was repeated four times.
Thrombin generation assay
To investigate the influence of each antithrombotic agent on the total activity of formed thrombin, a thrombin generation assay was performed according to our previously published procedure . Briefly, 20 µL of diluted phospholipid [platelin diluted 1 : 15 in assay buffer (20 mm Tris–HCl pH 7.3, with 150 mm NaCl, 1 mm CaCl2, 0.1% bovine serum albumin)], 20 µL of the antithrombotic agent, 20 µL of purified human FXa [1.4 nm (final concentration)], 20 µL of FV (2.4 × 10−2 nm), 20 µL of prothrombin (1.2 µm), and 40 µL of assay buffer were transferred into each well of a microtiter plate, followed by 50 µL of 50 mm CaCl2 to initiate thrombin generation. DX-9065a was used at 0.20–2.00 µm (final concentration), JTV-803 at 0.04–0.36 µm (final concentration), argatroban at 0.04–0.94 µm (final concentration), and LMWH at 0.10–1.00 U mL−1 (final concentration), with the presence of 5 mU mL−1 of antithrombin. After incubation at 37 °C for 10 min, 50 µL of 100 mm EDTA were added to each well. A portion (100 µL) of the reaction mixture was then transferred to another well and 50 µL of 1.8 mm synthetic substrate S-2238 (Chromogenics AB, Möndal, Sweden) were added. After a final incubation at 37 °C for 5 min, absorbance was measured at 405 nm with a plate reader (MTP-120; Corona Electric, Katsuta, Japan). Total activity of the formed thrombin in each sample was expressed as a percentage of the activity in the control. The inhibition activity of generated thrombin by each inhibitor was calculated by subtracting the percent of the sample from that of the control (100%). Experiments with each inhibitor were performed with triplicate samples and repeated four times.
Measurement of initial thrombin forming time
We also investigated the effects of the synthetic FXa inhibitors on thrombin forming time, which was the time taken to generate 50% thrombin activity in vitro, using defibrinated plasma and APTT reagent (Thrombocheck APTT). A thrombin forming time assay was performed according to a modified method of a previous report . Briefly, 40 µL of defibrinated plasma were mixed with 150 µL of 0.15 m saline and 10 µL of either the inhibitor [0.20–2.00 µm (final concentration) of DX-9065a, 0.04–0.36 µm (final concentration) of JTV-803, 0.04–0.38 µm (final concentration) of argatroban, 0.10–1.00 U mL−1 (final concentration) of LMWH] or saline, followed by the addition of 100 µL of APTT reagent and 100 µL of buffer A (0.05 m Tris–HCl pH 7.4, containing 0.1 m NaCl and 0.1% bovine serum albumin). After the mixture was incubated at 37 °C for 10 min, a coagulation reaction was initiated by the addition of 25 mm CaCl2 (time 0). At 15-s intervals from time 0, a 50-µL aliquot of the mixture was added to 465 µL of buffer A containing 20 mm EDTA to stop the coagulation reaction. Next, 25 µL of 0.5 mm S-2238 were added, followed by incubation for 2 min at 37 °C. After adding 300 µL of a 50% acetate solution to stop the chromogenic reaction, thrombin activity in each sample was measured at 405 nm with an autoreader. Each experiment with each inhibitor was performed using triplicate samples and repeated 10 times. The results are expressed as the time taken to reach 50% of maximal thrombin activity, which was determined as maximal optical density at 405 nm (T50). The thrombin forming time of each sample is expressed as a ratio of the T50 result of the sample to that of the control.
Effects of inhibitors on platelet aggregation induced by tissue factor
The effects of the antithrombotic agents on platelet aggregation, which was induced by thrombin formed in defibrinated plasma by TF, were investigated. Each sample, composed of defibrinated plasma (100 µL), washed platelets (40 × 104 platelets µL−1, 80 µL), and 20 µL of either Tris buffer saline (0.02 m Tris–HCl pH 7.4, 0.15 m NaCl) or the inhibitor, was preincubated at 37 °C for 2 min, followed by the addition of 20 µL of TF to induce platelet aggregation. The antithrombotic agents used in this experiment were 0.07–0.71 µm (final concentration) of DX-9065a, 0.13–1.26 µm (final concentration) of JTV-803, and 0.13–0.50 µm (final concentration) of argatroban. As a source of TF, thromboplastin from rabbit brain samples concentrated four times with PT-reagent (Thrombocheck PT) was used as a trigger to generate thrombin in defibrinated plasma, which was considered the minimum concentration to cause aggregation of the platelets. Platelet aggregation was observed for 10 min using an aggregometer (Haema tracer 601; Niko Bioscience, Tokyo, Japan), with maximum aggregation expressed as a percentage. Absorbance of the mixture, composed of washed platelets, defibrinated plasma, and Tris buffer saline, was taken as 0% aggregation, and absorbance of the mixture without washed platelets was taken as 100% aggregation. Each experiment was repeated six times.
All data are expressed as mean ± SD. Grouped data were analyzed for significance by comparative analysis using a two-tailed Student's t-test. P-values < 0.05 were considered to be statistically significant.
Each inhibitor (DX-9065a, JTV-803, and argatroban, in concentrations from 0.05 to 0.5 µm) prolonged PT in a concentration-dependent manner. Clotting times by the FXa inhibitors at concentrations more than 0.10 µm were slightly prolonged compared with those by argatroban, though the differences were not statistically significant (Fig. 1A). Each FXa inhibitor also prolonged APTT in a concentration-dependent manner; however, argatroban at concentrations of 0.1–0.5 µm prolonged APTT significantly more than DX-9065a and JTV-803 at the same concentrations (Fig. 1B).
Influences of inhibitors on thrombin generation
The influence of each antithrombotic agent on thrombin generation was investigated using a chromogenic assay with purified human coagulation factors. As shown in Table 1, DX-9065a at final concentrations of 0.10–2.00 µm inhibited 61.6–95.1% of the thrombin activity generated in the control, JTV-803 at 0.04–1.78 µm inhibited 47.9–92.7% of the thrombin activity generated in the control, and LMWH at 0.05–1.00 U mL−1 inhibited 36.6–80.3% of the thrombin activity generated in the control. In contrast, argatroban at 0.04–1.88 µm inhibited 19.7–74.6% of thrombin activity generated in the control, which was significantly lower than the other inhibitors.
Table 1. Influence of antithrombotic agents on thrombin generation and initial thrombin forming time
Inhibition of thrombin generation (%)
Initial thrombin forming time (T50 ratio)
Mean ± SD. *P < 0.001 vs. control; **P < 0.002; ***P < 0.05. LMWH, Low-molecular-weight heparin.
Influences of inhibitors on initial thrombin forming time
The influence of the antithrombotic agents on initial thrombin forming time (T50), which was the time to generate 50% thrombin activity in vitro, was also investigated. The ratio of T50 with 0.10–1.00 µm of DX-9065a was from 1.05 to 1.12, which was not significantly different from the control (1.01 ± 0.10), whereas 2.00 µm of DX-9065a caused a significant increase in T50 ratio (1.35 ± 0.12, P < 0.001). When JTV-803 was added at concentrations of 0.04 and 0.18 µm, the T50 ratio was 1.04 ± 0.09 and 1.10 ± 0.04, respectively, which was not significantly different from the control, whereas JTV-803 at concentrations > 0.36 µm resulted in a significant increase of T50 ratio. LMWH at 0.10 U mL−1 with 0.5 U mL−1 antithrombin did not affect the T50 ratio (1.07 ± 0.10); however, the ratio was significantly increased by the addition of 0.50 and 1.00 U mL−1 of LMWH. In contrast, when argatroban was added at concentrations from 0.19 to 1.88 µm, the T50 ratio was significantly increased (T50 ratio 1.33 ± 0.04–2.26 ± 0.14, P < 0.002 vs. control) (Table 1).
Effects of inhibitors on platelet aggregation induced by tissue factor
Next, platelet aggregation was induced by TF in order to determine whether thrombin formed by a minimum concentration of TF could activate platelets in the presence of each inhibitor. When TF was added to a sample without any inhibitor (control), maximum platelet aggregation within 10 min was 62.4 ± 24.0%. The addition of 0.13, 0.26, and 0.50 µm of argatroban significantly reduced maximum aggregation to 33.7 ± 35.1%, 11.8 ± 19.4%, and 0.3 ± 0.6%, respectively. In contrast, the addition of DX-9065a (0.07–0.50 µm) or JTV-803 (0.13–0.50 µm) did not have a significant effect on aggregation (DX-9065a, 66.6–57.7%; JTV-803, 63.7–40.7%) (Table 2).
Table 2. Effects of antithrombotic agents on tissue factor-induced platelet aggregation in defibrinated plasma
An immediate aggregation of platelets was observed in the presence of 0.50 µm of DX-9065a and 0.50 µm of JTV-803, which was similar to the control. However, aggregation in the presence of 0.25 µm of argatroban was significantly delayed. Typical platelet aggregation patterns from these experiments are shown in Fig. 2.
The generation of thrombin is a crucial step in the process of blood coagulation, and thrombosis results from a series of proteolytic activating reactions that are initiated via intrinsic and extrinsic pathways of blood coagulation cascades. FXa is a serine protease positioned at the convergence of those pathways, and DX-9065a and JTV-803 have been newly developed as synthetic inhibitors of FXa.
In the present study, the synthetic FXa inhibitors did not exert a serious influence on intrinsic coagulation, though argatroban, a selective thrombin inhibitor, significantly prolonged clotting time in APTT (Fig. 1B). The influence of these Xa inhibitors on PT and APTT may be different from our results if another reagent of APTT or PT is used. However, our findings suggested that a small amount of thrombin remained in the bloodstream after the FXa inhibitor was applied. Most of the existing FXa is inhibited when an FXa inhibitor is administered, but the small amount of FXa that is unaffected may bring about the generation of thrombin, which can consequently activate factor (F)XI independently of factor XII [13,14] and accelerate the intrinsic coagulation reaction. Further, there is a small amount of thrombin in blood that seems to be generated by the FXI that is activated automatically [13,14] and/or by factor VII activated through an unknown mechanism . Thrombin can stimulate the intrinsic coagulation reaction, therefore the administration of a FXa inhibitor may exert a weak inhibitory influence on clotting time in APTT. On the other hand, when argatroban is added, it may completely inhibit not only the thrombin formed through coagulation cascades but also that remaining in blood. Thus, argatroban apparently prolongs clotting time in APTT, because no acceleration of the intrinsic coagulation pathway is caused by thrombin. In contrast, during administration of an FXa inhibitor, the small amount of thrombin generated by the remaining FXa may not directly affect the extrinsic coagulation pathway. Therefore, the present clotting time results in the PT assay following administration of the FXa inhibitors were not significantly different from those for argatroban.
The purpose of the present study was to investigate the efficiency of FXa inhibitors on thrombin generation. For analysis of whole thrombin generation, including endogenous thrombin potential, the method established by Hemker [16,17] is known to be reliable. However, we performed two different kinds of experiments to measure total thrombin formed for a definite term without influence by the innate inhibitor, antithrombin, and to determine precisely the time necessary for forming a definite amount of thrombin during initial thrombin generation, termed initial thrombin forming time.
We first established a thrombin generation assay using a chromogenic assay with prothrombinase consisting of purified human coagulation factors and phospholipid, which can measure the genuine inhibitory effect of an inhibitor on fibrin formation. Total thrombin generated for 10 min was significantly decreased by the presence of DX-9065a or JTV-803, as even low concentrations inhibited > 60% of the thrombin generated in their absence, while argatroban inhibited thrombin generation to a lesser degree than either of the FXa inhibitors. Therefore, we concluded that selective inhibitors of FXa were extremely effective for thrombosis, which was supported by our results of inhibiting FXa with LMWH, a coenzyme of antithrombin. The good effect of synthetic FXa inhibitors for prevention of thrombosis has also been shown in animal models with DIC [3,5,6].
To clarify whether immediate formation of thrombin actually occurred, we also investigated the influence of the antithrombotic agents on T50, which represents initial thrombin forming time and is considered important for platelet aggregation in hemostasis, but not in the formation of fibrin clotting. This experiment was performed using defibrinated plasma, which contained an innate inhibitor both to thrombin and to FXa, antithrombin, and we detected a small amount of thrombin present under conditions similar to those seen in vivo. We consider that the formation of initial thrombin for platelet aggregation can be reflected in the determination of T50 and such an examination is useful for investigation of hemorrhage tendency, because the results correlate with coagulable potency, as shown by Ibbotson .
Argatroban significantly prolonged T50 in a concentration-dependent manner, probably because it immediately inhibited all of the initial formed thrombin. On the other hand, there was no significant difference in T50 results between the FXa inhibitors and the control (Table 1). Morishima  found that DX-9065a dose-dependently inhibited thrombus formation; however, it did not inhibit the elevation in plasma thrombin–antithrombin complex (TAT) levels in AV shunt model rats, whereas argatroban inhibited both thrombus formation and TAT elevation. Since the elevation of TAT indicates the formation of thrombin, those results suggested that thrombin may be formed during DX-9065a administration.
Our next area of investigation was whether the small amount of thrombin formed during administration of an FXa inhibitor could activate platelets, resulting in effective platelet aggregation for hemostasis. Accordingly, we tested the effects of the inhibitors on platelet aggregation induced by TF, because TF is an important factor for early hemostasis in humans. With the addition of argatroban, platelet aggregation in defibrinated plasma was remarkably delayed and the maximum percent of aggregation significantly decreased, compared with the control. On the other hand, the addition of either FXa inhibitor, even at a concentration able to inhibit thrombin generation, did not inhibit platelet aggregation, suggesting the presence of residual FXa that was able to form adequate thrombin to activate platelets despite the addition of an FXa inhibitor. These results support previous findings that the addition of a synthetic FXa inhibitor did not exert an influence on bleeding time. Yokoyama  reported that DX-9065a inhibited the formation of venous-type fibrin-rich thrombus by inactivating bound and soluble FXa without impairing platelet hemostatic function in baboon models treated with an arteriovenous shunt. DX-9065a has also been reported to have no effect on bleeding time, while it effectively reduced the severity in acute DIC rat models induced by lipopolysaccharide and thromboplastin [7,8]. Further, Murayama  reported that bleeding time was not prolonged when DX-9065a was intravenously administered to healthy male volunteers, even at the highest plasma concentration of 1640 ng mL−1 (2.87 µm). Intravenous infusion of JTV-803 at 1–10 mg kg−1 h−1 also had less of an effect on bleeding time in rats in another study . Therefore, our results regarding the effects of synthetic FXa inhibitors on T50 and platelet aggregation induced by TF may explain the phenomena described in previous reports.
The process by which these FXa inhibitors do not prolong bleeding time in spite of a strong inhibition of thrombin formation remains poorly understood. We consider that the initial immediate formation of a small yet adequate amount of thrombin to activate platelets is important for hemostasis. Since the affinity of thrombin for platelets is hundreds of times higher than that for fibrinogen [21,22], a minimal amount of thrombin, though insufficient to convert fibrinogen to fibrin, may effectively activate platelets to induce primary hemostasis. Namely, in patients given an FXa inhibitor, a small amount of thrombin might immediately be formed when any trigger for bleeding is stimulated, similar to that in healthy subjects. The formed thrombin may be enough to activate platelets for early hemostasis, though not adequate to cause the insoluble fibrin to be related with thrombus. Tanabe  also speculated that the competitive and reversible inhibition of FXa by DX-9065a might result in thrombin generation sufficient to induce hemostatic plug formation, thought it would be insufficient to facilitate thrombus formation.
In the present study, we showed the formation of initial thrombin with and without the presence of a FXa inhibitor by measuring T50, and also found that the immediate aggregation of platelets induced by the initial thrombin was not disturbed by its presence. The fact that primary hemostasis is conserved may not mitigate hemorrhaging seen, for example, in a serious surgical situation. However, we are convinced that the tested FXa inhibitors may have a role with initial thrombin forming time by preserving local hemostasis often seen during the treatment of thrombosis. Therefore, DX-9065a and JTV-803 are considered unique antithrombotic agents without hemorrhagic effects.
We thank Miyako Arakawa for her technical assistance, and Chizuru Imamura and Kaori Saito for their help in preparation of this manuscript.