The effects of phosphatidylserine-dependent antiprothrombin antibody on thrombin generation

Authors


Abstract

Objective

Antibodies to prothrombin (APTs) and to β2-glycoprotein I are the major autoantibodies responsible for lupus anticoagulant (LAC) activity. APTs comprise antibodies against prothrombin alone as well as antibodies against phosphatidylserine/prothrombin complex (anti-PS/PT), the latter being highly associated with the antiphospholipid syndrome (APS). The effect of anti-PS/PT on thrombin generation has not been elucidated, and the paradoxical effect of LAC (an anticoagulant in vitro, but a procoagulant in vivo) remains an enigma. The purpose of this study was to investigate the effects of anti-PS/PT on thrombin generation and to examine the LAC paradox.

Methods

We evaluated 36 anti-PS/PT–positive APS patients and 127 healthy subjects. Markers of in vivo thrombin/fibrin generation, including prothrombin fragment F1+2, thrombin–antithrombin III complex, soluble fibrin monomer, D-dimer, and fibrin degradation products, were measured. Mouse monoclonal anti-PS/PT antibody 231D was established, and its effects on in vitro thrombin generation were investigated by chromogenic assay.

Results

Significantly elevated levels of markers of thrombin/fibrin generation were observed in anti-PS/PT–positive patients, regardless of the presence or absence of anticardiolipin antibodies, as compared with healthy subjects. In the presence of low concentrations of human activated factor V (FVa), monoclonal antibody 231D increased thrombin generation in a dose-dependent manner. In contrast, when high concentrations of FVa were added, monoclonal antibody 231D decreased thrombin generation. Under a constant concentration of FVa, a high concentration of human FXa enhanced the effect of 231D.

Conclusion

The presence of anti-PS/PT greatly correlated with increased thrombin generation in APS patients. The in vitro effects of monoclonal antibody 231D on thrombin generation are “biaxial” according to the FVa/FXa balance. These data may serve as a clue to understanding the LAC paradox and the thrombogenic properties of anti-PS/PT.

Antiphospholipid antibodies (aPL) are immunoglobulins that are related to diverse clinical phenomena, such as arterial and venous thrombosis, complications of pregnancy, livedo reticularis, valvular disease, neurologic disorders, and thrombocytopenia. The term antiphospholipid syndrome (APS) is used to link thrombosis or pregnancy morbidity to the persistence of aPL as one of the most common causes of acquired thrombophilia (1).

It has been shown that despite their name, aPL are not directed against anionic phospholipids, as was previously thought, but are part of a large family of autoantibodies against phospholipid-binding plasma proteins or phospholipid–protein complexes (2). The most common and best characterized antigenic target of these antibodies is β2-glycoprotein I (β2GPI) (3–5), a phospholipid binding protein that has been extensively studied and has been shown to play a prominent role in the binding of aPL to phospholipid. Anticardiolipin antibodies (aCL), which are associated with APS, are not directed against cardiolipin alone, but require β2GPI as a cofactor for the binding of cardiolipin in enzyme-linked immunosorbent assay (ELISA) plates. Beta2-glycoprotein I bears the epitopes for aCL binding that are exposed when β2GPI binds to negatively charged phospholipids (6, 7).

Prothrombin, another main phospholipid binding protein, was first reported by Loeliger in 1959 (8) to be a probable cofactor for the lupus anticoagulant (LAC). Fleck et al (9) subsequently confirmed that antiprothrombin antibodies (APTs) are responsible for the LAC activity, and in 1991, Bevers et al (10) emphasized the importance of prothrombin in generating LAC activity. Some years later, the inhibitory effect of LAC on endothelial cell–mediated prothrombinase activity was reported, and it was also demonstrated that the IgG fraction containing LAC activity bound to the phospholipid–prothrombin complex (11). Therefore, prothrombin was recognized as another target for autoantibodies with LAC activity. Accordingly, it is widely accepted that APTs and anti-β2GPI antibodies are the 2 major autoantibodies responsible for LAC activity: APTs for prothrombin-dependent LAC and anti-β2GPI antibodies for β2GPI-dependent LAC.

An ELISA for the detection of APTs using prothrombin alone as the antigen coated onto irradiated plates (APT-alone assay) was described in 1995 (12). Since then, a number of studies have investigated the clinical relevance of testing APT alone; nevertheless, the association between APT alone and clinical manifestation of APS is still a subject of controversy (13). In 1996, antibodies against the phosphatidylserine/prothrombin complex (anti-PS/PT; or phosphatidylserine-dependent APTs) were described in LAC-positive patients (14). Moreover, the ELISA using phosphatidylserine-bound prothrombin as antigen was reported to be more sensitive for detecting the presence of APTs than the ELISA using prothrombin alone as antigen (15). Our group assessed the anti-PS/PT ELISA in a large population of patients with autoimmune diseases and found that IgG anti-PS/PT were highly prevalent in patients with APS as compared with patients with other diseases (16). We also showed that the detection of anti-PS/PT strongly correlated with the clinical manifestations of APS and with the presence of LAC.

In APS patients, the LAC paradox, that is, the behavior of LAC as an anticoagulant in vitro but a procoagulant in vivo, remains unresolved. In addition, the effects of anti-PS/PT on thrombin generation, whether in vitro or in vivo, have not been clarified. In order to investigate the effects of anti-PS/PT on thrombin generation, we evaluated markers of thrombin generation and fibrinolytic turnover in plasma samples from APS patients with anti-PS/PT antibodies. Furthermore, we established a mouse monoclonal anti-PS/PT antibody (231D) and used this monoclonal antibody to analyze thrombin generation in vitro.

PATIENTS AND METHODS

Patients.

Plasma and serum samples were obtained from 36 APS patients with IgG and/or IgM anti-PS/PT antibodies (32 women and 4 men with a mean age of 46 years [range 22–74 years]) who fulfilled the revised Sapporo criteria for APS (1). Fifteen patients were diagnosed as having primary APS, and 21 patients had APS in association with other connective tissue diseases. Twenty-six patients (72%) had experienced arterial thrombotic events, such as stroke, myocardial infarction, and iliac artery occlusion, as confirmed by computed tomography scanning, magnetic resonance imaging, or conventional angiography. Deep vein thrombosis and pulmonary thrombosis were defined as venous thrombosis (12 of 36 patients [33%]) and were confirmed by angiography or scintigraphy. Thirteen women (36%) had pregnancy morbidity as defined by the APS criteria. Anti-PS/PT antibodies of IgG, IgM, and both isotypes were detected in 47%, 22%, and 31% of patients, respectively.

None of the patients had thrombotic events or pregnancy complications within 3 months before blood collection. Signs of acute thrombosis were not detected in any patient at the time blood was drawn. The time since the latest manifestation of APS varied from 4 months to 6 years. Therefore, our data correspond to the baseline of thrombin generation in anti-PS/PT–positive patients. When blood was drawn for this study, no patients were receiving heparin; some were taking warfarin, but there had been no modification of any medications within the 3 previous months. None of the patients had a tendency toward bleeding.

Blood samples were also collected from 127 apparently healthy subjects who consented to join the study. There were a total of 51 women and 76 men with a mean age of 34 years (range 18–65 years).

The study was performed in accordance with the Declaration of Helsinki and the Principles of Good Clinical Practice. Approval was obtained from the Local Ethics Committee, and informed consent was obtained from each study subject before enrollment.

Plasma samples.

Venous blood was collected into tubes containing a one-tenth volume of 0.105M sodium citrate and was centrifuged immediately at 4°C. Plasma samples were depleted of platelets by filtration then stored at –80°C until they were used in the experiments.

ELISA for the detection of anti-PS/PT.

Anti-PS/PT antibodies were detected by ELISA, as previously described (16). Briefly, nonirradiated microtiter plates (Sumilon Type S; Sumitomo Bakelite, Tokyo, Japan) were coated with 30 μl of a 50 μg/ml preparation of phosphatidylserine (Sigma, St. Louis, MO) and dried overnight at 4°C. To avoid nonspecific binding of proteins, the wells were blocked with 150 μl of Tris buffered saline (TBS) containing 1% fatty acid–free bovine serum albumin (BSA) (catalog no. A6003; Sigma) and 5 mM CaCl2 (BSA–CaCl2). After 3 washes in TBS containing 0.05% Tween 20 (Sigma) and 5 mM CaCl2 (TBS–Tween–CaCl2), 50 μl of a 10 μg/ml preparation of human prothrombin (Diagnostica Stago, Asnières-sur-Seine, France) in BSA–CaCl2 was added to half of the wells in the plates, and the same volume of BSA–CaCl2 alone (as sample blank) was added to the other half.

After 1 hour of incubation at 37°C, the plates were washed, and 50 μl of serum diluted 1:100 in BSA–CaCl2 was added to duplicate wells. Plates were incubated for 1 hour at room temperature, and alkaline phosphatase–conjugated goat anti-human IgG or IgM and substrate were added. The optical density of wells coated with phosphatidylserine alone was subtracted from that of wells containing phosphatidylserine/prothrombin. The anti-PS/PT antibody titer of each sample was derived from the standard curve according to dilutions of the positive control.

Determination of aCL and LAC.

IgG and IgM aCL were measured according to a standard aCL ELISA, as described elsewhere (17).

Two clotting tests were performed for LAC determination, using a semiautomated hemostasis analyzer (STart 4; Diagnostica Stago) according to the guidelines recommended by the Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibody of the Scientific and Standardisation Committee of the International Society on Thrombosis and Haemostasis (18). For measurement of the activated partial thromboplastin time (APTT), a sensitive reagent with a low phospholipid concentration (test PTT-LAC; Diagnostica Stago) was used for screening, and the results were confirmed with the use of a StaClot LAC kit (Diagnostica Stago). The dilute Russell's viper venom time (dRVVT) was screened for and confirmed by use of a Gradipore LAC test (Sydney, New South Wales, Australia). LAC was considered positive when at least 1 of these tests confirmed its presence.

Assessment of markers of thrombin and plasmin generation in vivo.

Plasma levels of soluble fibrin antigen (Mitsubishi Kagaku Iatron, Tokyo, Japan), prothrombin fragment F1+2 (Enzygnost F1+2 Micro; Dade-Behring, Marburg, Germany) and thrombin–antithrombin III complex (TAT test Kokusai F; International Reagent Corporation, Kobe, Japan) were assessed as markers of thrombin generation. Among them, F1+2 was not measured in patients receiving warfarin. We also evaluated D-dimer (D-dimer test-F; International Reagent Corporation) and fibrin/fibrinogen degradation products (Nonapia p-FDP; Daiichi Kagaku, Tokyo, Japan) as markers of fibrinolytic turnover.

Establishment of a mouse monoclonal anti-PS/PT antibody using prothrombin as antigen.

Eight-week-old female BALB/c mice were immunized intraperitoneally and were given 2 booster injections with 50 μg of human prothrombin (Enzyme Research Laboratories, Swansea, UK) emulsified with Freund's complete adjuvant and with Freund's incomplete adjuvant (Difco, Detroit, MI), respectively. The spleens were excised from the mice, and spleen cells were fused with P3U1 mouse myeloma cells (19). Cells producing antibodies against the phosphoserine/prothrombin complex were screened by anti-PS/PT ELISA. Antibody-producing hybridomas were cloned by serial limiting dilution and injected intraperitoneally into pristane-pretreated BALB/c nude mice to obtain ascitic fluid. Monoclonal antibody 231D was sequentially purified by protein G–Sepharose affinity chromatography (MabTrap-TMGII; Amersham Pharmacia Biotech, Uppsala, Sweden).

Establishment of a mouse monoclonal anti-PS/PT antibody using prethrombin 1 as antigen.

Prothrombin (1 mg/ml in TBS) was digested for 3 hours at 37°C with 10 units of bovine thrombin (Sigma). The reaction was stopped by the addition of 1 mM p-ABSF and p-APMSF. Prethrombin 1, which lacks the prothrombin domain 1 that comprises the phospholipid-binding site (Gla-domain), was purified from the solution by ion-exchange chromatography. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis of prethrombin 1 fractions revealed a single band at 50 kd under nonreducing conditions.

To obtain a monoclonal APT that binds to prothrombin but does not interact with the phospholipid-binding site of prothrombin, a BALB/c mouse was immunized with prethrombin 1. Hybridomas were screened using an anti-PS/PT ELISA, and monoclonal antibody 51A6 was established and purified in the same manner as described for 231D.

APT-alone assay for activity of the monoclonal anti-PS/PT antibody.

An APT-alone assay was performed as described previously (20), with some modifications. Briefly, either irradiated microtiter plates (MaxiSorp; Nunc, Roskilde, Denmark) or nonirradiated plates (Sumilon Type S) were coated overnight at 4°C with 10 μg/ml of purified human prothrombin in TBS containing 5 mM CaCl2. Wells were blocked for 1 hour at 37°C with 0.5% gelatin. After 3 washes with TBS–Tween–CaCl2, 50 μl of sample (monoclonal antibodies, control mouse IgG, or serum from mouse immunized with human prothrombin), diluted in BSA–CaCl2 as appropriate, was added to duplicate wells. Plates were incubated for 1 hour at room temperature, followed by the addition of alkaline phosphatase–conjugated goat anti-mouse IgG and substrate. Optical density at 405 nm was then measured.

Detection of LAC activity in normal plasma using monoclonal anti-PS/PT antibody.

Blood samples from 4 healthy donors were collected in precooled tubes containing a one-tenth volume of 0.105M sodium citrate and were immediately centrifuged at 2,000g for 15 minutes. Platelets were removed by filtration, and the platelet-free plasma was stored at –80°C. Different concentrations of monoclonal antibodies (50–3.1 μg/ml) were spiked into the pooled normal plasma, and clotting times were determined using the Start 4 system. Measurements of dRVVT and APTT were performed as described above. In addition, the kaolin clotting time (KCT) was measured with a kaolin solution (Dade-Behring) according to standard protocols.

Competitive ELISAs.

IgG from 9 APS patients with high titers of anti-PS/PT antibodies was purified using protein G–Sepharose affinity chromatography (MabTrap-TMGII). Monoclonal antibody 231D or 51A6 (200 or 20 ng/ml) was added to plates that had been coated with PS/PT complex, and the plates were incubated for 1 hour at room temperature. Purified IgG (1 mg/ml) was added to the wells, and binding to PS/PT complex was determined by anti-PS/PT ELISA. The inhibition of IgG binding by monoclonal antibodies was calculated by comparing the optical density values with the values for IgG binding in the absence of monoclonal antibodies.

An additional competitive ELISA was performed in which 200 ng/ml of either 231D or 51A6 was coincubated with several concentrations (200, 50, 12.5, and 3.1 μg/ml) of 2 representative purified IgG from APS patients.

Measurement of in vitro thrombin generation.

The effects of anti-PS/PT antibodies on thrombin generation were evaluated with a chromogenic assay, using the prothrombinase complex phospholipid, CaCl2, purified human activated factor V (FVa; Haematologic Technologies, Essex Junction, VT), and FXa (Enzyme Research Laboratories). The thrombin generation assays used in this study were based on our previous analyses (21). Thrombin generation was measured by using a specific substrate for thrombin, D-Phe-pipecolyl-Arg-paranitroaniline (S-2238; Chromogenix Instrumentation Laboratory, Milan, Italy). Cephalin (PTT-Reagent RD; Roche Diagnostics, Basel, Switzerland), a phospholipid from rabbit brain extract, was used as the source of phospholipid. Cephalin was used at a dilution of 1:63 in assay buffer (1% BSA, 0.1 mM CaCl2, TBS).

Ten microliters of a 10 μg/ml preparation of purified human prothrombin diluted in assay buffer, 10 μl of diluted phospholipid, and 40 μl of 231D at various concentrations was transferred into each well of a 96-well microtiter plate and then incubated at 37°C for 20 minutes. Ten microliters of FVa (0–1 ng/ml) and FXa (0.5–5 μg/ml) was added to the preincubated mixture, and the plate was left at room temperature for 2 minutes. The coagulation reaction was initiated by adding 25 μl of a 50 mM concentration of CaCl2, followed by 25 μl of 2 mM concentration of S-2238. After incubation at 37°C, the absorbance of the mixture was measured at 405 nm with a Multiscan Ascent plate reader (Thermo Electron Corporation, Waltham, MA).

Statistical analysis.

Statistical evaluation was performed by Mann-Whitney U test, Fisher's exact test, or Student's t-test, as appropriate. P values less than 0.05 were considered significant.

RESULTS

Plasma levels of thrombin generation and markers of fibrinolytic turnover.

Levels of all markers of thrombin generation and fibrinolytic turnover were higher in APS patients with anti-PS/PT antibodies as compared with those in healthy control subjects. The distribution of representative markers, soluble fibrin antigen, and D-dimer are displayed in Figure 1. Plasma levels of soluble fibrin antigen and D-dimer were higher in both aCL subgroups of anti-PS/PT–positive patients as compared with those in healthy controls.

Figure 1.

Distribution of plasma levels of soluble fibrin antigen (SF) and D-dimer in patients with antiphospholipid syndrome (APS) and in healthy individuals. Plasma levels of A, soluble fibrin antigen and B,D-dimer were measured in healthy controls, in all APS patients with anti–phosphatidylserine/prothrombin complex (anti-PS/PT) antibody, and in anti-PS/PT antibody–positive APS patients with or without anticardiolipin antibody (aCL). Horizontal line shows the cutoff level of positivity, which was defined as the mean ± 2SD of the level in control subjects. Data are shown as individual results as well as box plots, where each box represents the 25th to 75th percentiles. Lines inside the boxes represent the median. Lines outside the boxes represent the 10th and the 90th percentiles. Values across the bottom are the number of subjects positive/total number tested, as well as the percentages.

The cutoff level of each marker was defined as the mean ± 2SD of the levels in control subjects. A higher prevalence of elevation in the levels of markers of thrombin/plasmin generation (F1+2, thrombin–antithrombin III complex, soluble fibrin antigen, D-dimer, and fibrin/fibrinogen degradation products) was found in all anti-PS/PT–positive patients, in anti-PS/PT–positive patients with aCL, and in anti-PS/PT–positive patients without aCL as compared with the levels in healthy subjects (P < 0.05 for each comparison) (Table 1).

Table 1. Prevalence of markers of increased thrombin/plasmin generation in patients and healthy controls*
 All patientsAnti-PS/PT+ patientsHealthy controls
aCL+aCL−
  • *

    Values are the number positive/total number tested (%). All values were statistically significant as compared with those in the controls (P < 0.05). Anti-PS/PT = anti–phosphatidylserine/prothrombin complex; aCL = anticardiolipin antibody.

Prothrombin fragment F1+210/28 (36)1/7 (14)9/21 (43)3/60 (5)
Thrombin–antithrombin III complex10/36 (28)2/14 (14)8/22 (36)7/73 (10)
Soluble fibrin antigen16/36 (44)8/14 (57)8/22 (36)6/127 (5)
D-dimer18/36 (50)8/14 (57)10/22 (45)1/73 (1)
Fibrin/fibrinogen degradation products10/36 (28)6/14 (43)4/22 (18)3/74 (4)

Binding activity of mouse monoclonal anti-PS/PT antibody.

Two anti-PS/PT antibody clones, 231D and 51A6, were obtained. The 231D antibody was established from a mouse that had been immunized with human prothrombin, and the 51A6 antibody was established from a mouse that had been immunized with human prethrombin 1. Both clones bound strongly to the PS/PT complex, but not to phosphatidylserine alone (Figures 2A and B). Both murine monoclonal antibodies are of IgG1 isotype.

Figure 2.

Binding activity of mouse monoclonal anti–phosphatidylserine/prothrombin complex (anti-PS/PT) antibodies 231D and 51A6. A and B, Binding activity of 231D and 51A6 for the PS/PT complex (A) and for phosphatidylserine alone (B) was determined by enzyme-linked immunosorbent assay (ELISA) using phosphatidylserine-coated plates. C and D, Binding activity of 231D and 51A6 for antiprothrombin antibody (APT), using prothrombin alone as the antigen coated onto either irradiated (C) or nonirradiated (D) plates, was determined by ELISA. Monoclonal antibody 51A6 bound to prothrombin coated on both irradiated and nonirradiated plates, whereas 231D showed little binding to prothrombin under both conditions. In all experiments, control IgG, consisting of purified mouse IgG from pooled normal mouse serum, was used at the indicated concentrations (ng/ml), and antisera, consisting of sera from mice that had been immunized with human prothrombin, were used at the indicated dilutions. Values are from a representative experiment.

APT-alone activity was also investigated in the monoclonal anti-PS/PT antibodies. We found that 51A6 bound to prothrombin coated onto both irradiated and nonirradiated ELISA plates, but 231D showed a lower level of binding to prothrombin under both conditions (Figures 2C and D). Normal mouse IgG and pooled sera obtained from mice that had been immunized with human prothrombin were used as the negative control and the positive control, respectively.

LAC activity of monoclonal anti-PS/PT antibody.

Purified 231D or 51A6 monoclonal antibody was added to normal plasma, and the dRVVT was measured in the monoclonal anti-PS/PT antibody–spiked plasma (Figure 3). With reagent 1 of the dRVVT test, which has a low phospholipid concentration, the clotting time of 231D-spiked plasma was prolonged in a dose-dependent manner. The clotting time was largely more prolonged with reagent 1 of the dRVVT test than with reagent 2, which contains a high concentration of phospholipids. Similar results were obtained with the 51A6-spiked plasma, but the clotting time was not as prolonged as with the 231D-spiked plasma.

Figure 3.

Lupus anticoagulant (LAC) activity of mouse monoclonal anti–phosphatidylserine/prothrombin complex antibodies 231D and 51A6, as determined by A, the dilute Russell's viper venom time (dRVVT), B, the activated partial thromboplastin time (APTT), and C, the kaolin clotting time (KCT). For measurement of dRVVT, purified 231D or 51A6 monoclonal antibody was added to normal plasma, and the dRVVT was determined. Reagent 1 (R1) contains a low concentration of phospholipid, and reagent 2 (R2) contains a high concentration of phospholipid. The APTT and KCT were determined in plasma that had been spiked with either 231D or 51A6. In all experiments, control IgG consisted of purified mouse IgG from pooled normal mouse serum. Numbers across the bottom are the concentration (in μg/ml) of 231D, 51A6, and control IgG tested. Values are from a representative experiment.

Dose-dependent prolongations of the clotting time in both the APTT and the KCT tests were also found in 231D-spiked plasma and in 51A6-spiked plasma. Plasma containing 231D showed stronger anticoagulant properties than did plasma containing 51A6.

Findings of competitive ELISAs.

The binding of purified IgG from anti-PS/PT–positive APS patients to the PS/PT complex was inhibited by 231D (35–70%). In contrast, there was no significant effect of 51A6 on the binding of IgG fractions to PS/PT complex (Figure 4A). Coincubation of 231D with purified IgG from APS patients also produced dose-dependent inhibition (Figure 4B).

Figure 4.

Competitive enzyme-linked immunosorbent assay (ELISA). IgG was purified from serum samples obtained from 9 patients with antiphospholipid syndrome (APS) who had anti–phosphatidylserine/prothrombin complex (anti-PS/PT) antibodies. A, Percentage inhibition of IgG binding in the presence of monoclonal antibody 231D, 51A6, or mouse IgG. Monoclonal antibody 231D, 51A6, or purified mouse IgG from pooled normal mouse serum (20 or 200 ng/ml) was preincubated on plates that had been coated with PS/PT complex, and 1 mg/ml of purified IgG from 7 of the APS patients was added. The percentage inhibition of IgG binding was calculated by comparing the optical density values in the presence of 231D, 51A6, or mouse IgG with the optical density values in the absence of 231D, 51A6, or mouse IgG, respectively. B, Inhibition curves following coincubation of monoclonal antibody 231D or 51A6 with the indicated concentrations of purified IgG from APS patients 8 and 9. As controls, purified IgG from the 2 APS patients was also incubated without monoclonal antibody.

Effects of monoclonal anti-PS/PT antibody on thrombin generation.

The effect of monoclonal anti-PS/PT antibody on thrombin generation in vitro was evaluated by chromogenic assay using purified human clotting factors (Figures 5A–C). In the absence or in the presence of a very low concentration of FVa (0.1 ng/ml), the 231D monoclonal antibody increased thrombin generation by as much as 87% and in a dose-dependent manner. In contrast, when a high concentration of FVa (1 ng/ml) was added, 231D decreased thrombin generation by as much as 35%. The 51A6 monoclonal antibody displayed a lower level of inhibition of thrombin generation regardless of the concentration of FVa.

Figure 5.

Evaluation of the effects of monoclonal anti–phosphatidylserine/prothrombin complex (anti-PS/PT) antibodies on thrombin generation in vitro, as determined by chromogenic assay. Purified human clotting factors were used in these experiments. A–C, Thrombin generation was measured in the absence of activated factor V (FVa) (A), in the presence of 0.1 ng/ml of FVa (B), and in the presence of 1 ng/ml of FVa (C), using a constant concentration of 1.25 μg/ml of FXa. D, Thrombin generation was measured in the absence of FVa (left) and in the presence of 1 ng/ml of FVa (right), using 0.5, 1.25, and 5 μg/ml of FXa. A constant concentration of 2.5 μg/ml of 231D or control IgG (purified mouse IgG from pooled normal mouse serum) was used. Values are the mean ± SEM difference in thrombin generation, as determined by the optical density at 405 nm (OD405) value minus the OD405 value in the absence of monoclonal antibodies.

We also examined whether various concentrations of FXa altered the effects of 231D on thrombin generation (Figure 5D). Under 2 different constant concentrations of FVa, the effects of 231D on thrombin generation were increased in the presence of increasing concentrations of FXa. Again, we found that the 51A6 monoclonal antibody exhibited little inhibition of thrombin generation under any condition examined.

DISCUSSION

In this study, we demonstrated that the plasma levels of markers of thrombin generation/fibrinolysis turnover were elevated in patients with anti-PS/PT antibody, regardless of the coexistence of aCL. The mouse monoclonal anti-PS/PT antibody 231D, which has binding properties similar to those of anti-PS/PT found in patients with APS, showed “bipolar” effects on thrombin generation triggered by FXa.

Despite the proposal by some investigators of a possible correlation between APT and thrombosis, no clinical data have reported a link between increased thrombin/plasmin generation and antibodies against prothrombin. This study is the first to show the up-regulation of thrombin/plasmin generation in patients with anti-PS/PT antibody regardless of the presence of aCL. We also tested anti-β2GPI antibodies in this study (data not shown), and the results were almost identical to those found in aCL. (None of our patients were positive for anti-β2GPI antibodies but negative for aCL.)

There are several reports showing enhanced thrombin generation and fibrinolytic turnover in APS patients with aCL (22, 23). In addition, de Laat et al (24) showed that β2GPI-dependent LAC is highly correlated with thrombosis in patients with APS. Those reports clearly indicated that antibodies against β2GPI, represented by aCL, anti-β2GPI antibodies, or β2GPI-dependent LAC, are correlated with high levels of thrombin generation. In contrast, there has been no report showing a correlation between antibodies against prothrombin, represented by prothrombin dependent LAC or anti-PS/PT. De Laat et al (24) failed to demonstrate increased thrombin generation in patients with β2GPI-independent LAC, but such LAC would comprise antibodies against prothrombin and antibodies against nonspecific (or undetermined) proteins. Our data revealed that in the absence of antibodies against β2GPI, levels of anti-PS/PT antibody correlated with elevated levels of markers of thrombin generation in APS patients, providing evidence that the increased thrombin/fibrin generation in these patients is related to anti-PS/PT itself.

The plasma samples were collected at least 3 months after the last thrombotic event, suggesting that patients with anti-PS/PT antibody are basically in a thrombophilic state. Some prothrombotic triggers may alter the balance between thrombin generation and regulators of thrombin generation, eventually leading to thrombosis.

The antibody responsible for prothrombin-dependent LAC activity is closely related to APTs detected by anti-PS/PT assay. In the setting of autoimmune disease, both anti-PS/PT and APT alone have been shown to be correlated with the presence of LAC, but anti-PS/PT had a markedly stronger relative risk for the presence of LAC than did APT alone (16). Many patients in that study had both anti-PS/PT and APT alone, but no correlation of their titers was found, even though some patients had very high levels of anti-PS/PT antibody in the absence of APT alone and vice versa.

To clarify the characteristics and properties of anti-PS/PT in thrombin generation or in the prothrombotic state observed in patients with anti-PS/PT, we successfully established 2 monoclonal antibodies. The 231D monoclonal antibody, which was obtained by immunizing mice with whole prothrombin, showed strong anti-PS/PT activity. Prothrombin was digested with thrombin, and the prethrombin 1 fraction, which lacks the phospholipid-binding domain of prothrombin, was used as immunogen to establish monoclonal APTs with phosphatidylserine-independent binding activity. We established monoclonal antibody 51A6, which as we expected, had strong APT-alone activity but lower anti-PS/PT activity. Monoclonal antibody 231D had minor APT-alone activity as compared with the 51A6 monoclonal antibody. The presence of calcium did not affect APT-alone activity in either of these monoclonal antibodies. The affinity constant (Ka) of 51A6 to prothrombin, as determined by liquid-phase inhibition ELISA, was 5.49 × 10−9M (data not shown), suggesting that 51A6 had moderate or strong affinity to the prothrombin molecule. The binding affinity of 231D to the PS/PT complex may be comparable to that of 51A6 according to the results of the anti-PS/PT ELISA.

Competitive ELISA revealed that 231D partially inhibited the binding to the PS/PT complex of the autoimmune anti-PS/PT antibody derived from patients with APS, implying that 231D shared the epitope(s) on phosphatidylserine-bound prothrombin with autoimmune anti-PS/PT. In contrast, 51A6 did not display any interaction in the binding between autoimmune anti-PS/PT and the PS/PT complex; thus, the 51A6 epitope on prothrombin is independent of those of autoimmune anti-PS/PT antibody.

LAC activity of monoclonal APTs alone has previously been reported (25). However, our data showed that the 231D monoclonal antibody had a stronger inhibitory effect in the APTT test than did the 51A6 monoclonal antibody, suggesting that 231D represents immunologic and hematologic properties of autoimmune anti-PS/PT antibody found in patients with APS.

Prothrombin is a single-chain glycoprotein composed of 3 structural regions as follows: fragment 1, which contains the Gla domain and kringle 1 domain, fragment 2, which mainly contains the kringle 2 domain, and a serine protease precursor domain (26, 27). Prothrombin is activated and cleaved into α-thrombin in a membrane-dependent process that includes the actions of FXa, its cofactor FVa, and divalent calcium ions assembled into a complex on the membrane.

To investigate the direct effect of anti-PS/PT on thrombin generation in vitro, we prepared a chromogenic thrombin generation assay, and the effect of the 231D monoclonal antibody was explored in the presence of different concentrations of FVa and FXa. In the presence of a low concentration of FVa, 231D increased thrombin generation in a dose-dependent manner. In contrast, when a high concentration of FVa was added, 231D diminished thrombin generation. In the second set of experiments, a high concentration of FXa was found to enhance the effect of 231D in the presence of a constant concentration of FVa. When FXa was added at a high concentration, the relative FVa concentration was low, resulting in increased thrombin generation by 231D. Taken together, the balance of FVa/FXa was the determinant of the behavior of 231D with regard to the generation of thrombin, antithrombin, or prothrombin.

In the presence of sufficient amounts of FVa, the 231D monoclonal antibody decreased thrombin generation, and the 51A6 monoclonal antibody showed a similar effect, although its potential was lower. A previous study has also shown inhibitory effects of APTs on thrombin generation. Church et al (28) produced 5 monoclonal antibodies to prothrombin kringle 2, and 2 of them inhibited FVa-dependent prothrombin activation. In terms of the phospholipid-dependency of LAC-like activity shown by the phospholipid-neutralizing test in the LAC assay, the interpretation of the behavior of monoclonal APTs may be as follows: under in vitro conditions, the higher the amount of phospholipid the more prothrombinase and/or prothrombin are available, leading to the acceleration of thrombin generation in the presence of APTs.

The in vitro effects of the 231D monoclonal antibody on thrombin generation, on the other hand, are different according to the balance of FVa and FXa. Zhao et al (29) generated and characterized a human monoclonal antiprothrombin antibody with strong LAC activity that enhanced prothrombin binding to phospholipid and shortened the plasma coagulation times (29). Their data provide an explanation for the LAC paradox, showing that a single, highly purified aPL can behave as LAC and can paradoxically increase coagulation in endothelial cell–based coagulation assays. Our current findings support the hypothesis stated above, showing that the mouse monoclonal anti-PS/PT antibody 231D, which carries strong LAC activity, increased thrombin generation in the absence of FVa as well as in the presence of very low concentrations of FVa. The 231D monoclonal antibody may allow prothrombin to bind more firmly to phospholipid, assembling prothrombin on the phospholipid and subsequently increasing thrombin generation. When abundant FXa is present, FXa is able to act on prothrombin to generate thrombin. This may not be the situation when large amounts of FVa are present, since the coenzyme activity of FVa is sufficiently potent to overcome the antithrombin generation effects of 231D. The 51A6 monoclonal antibody, needing no phospholipid involvement for its binding to prothrombin, would not play a role in the augmentation of thrombin generation in this mechanism.

This is the first study to show that monoclonal antibodies against prothrombin do not exclusively have LAC-like thrombin reduction potential, but are also able to increase thrombin generation. However, the phenomena we observed are evidently not the only mechanism of thrombosis in patients with APS. Recently, great interest has arisen with regard to the binding of aPL to procoagulant cells and how this binding mediates cell activation related to the clinical manifestations of APS. Within the last few years, studies examining the mechanism of signal transduction implicated in the induction of procoagulant substances by aPL have been performed. There is now clear evidence that the p38 MAPK pathway of cell activation plays an important role in anti-β2GPI antibody–mediated cell activation (30–32). Considering the similarities of the properties of anti-β2GPI and APT, procoagulant cell activation may be a major event in the generation of thrombosis in APS patients who have anti-PS/PT antibody. Thrombin may serve as a trigger for cells to express phosphatidylserine on their surface via protease-activated receptors, which are present on many types of procoagulant cells, and with glycoprotein Ib–IX–V complex on the surface of platelets, leading to platelet aggregation and activation (33).

Thrombin is a key enzyme in hemostasis and is a multipotential enzyme in the coagulation/inflammation system. The direct involvement of anti-PS/PT antibody in thrombin generation may be a clue to the pluripathologic process that occurs in patients with APS (34). Although further clarification of the roles of anti-PS/PT antibodies in APS is essential, we believe that the findings of this study contribute to the understanding of the pathophysiology of thrombophilia in patients with APS.

AUTHOR CONTRIBUTIONS

Dr. Atsumi had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Atsumi, Ieko, Koike.

Acquisition of data. Sakai, Atsumi, Ieko, Amengual, Furukawa, Furusaki, Bohgaki, Kataoka, Horita, Yasuda, Koike.

Analysis and interpretation of data. Sakai, Atsumi, Ikeo, Amengual, Koike.

Manuscript preparation. Sakai, Atsumi, Amengual, Koike.

Statistical analysis. Sakai, Atsumi.

Ancillary