Coagulation and fibrinolytic activities in 2 siblings with β₂-glycoprotein I deficiency

β₂-Glycoprotein I (β₂GPI), formed by 5 short consensus repeat domains, is a 50-kd phospholipid-binding protein with a plasma concentration of about 200 μg/mL.1 It bears the epitope(s) for anticardiolipin antibodies (anti-β₂GPI antibodies) that are associated with the antiphospholipid syndrome.2 Despite studies on the function of β₂GPI in vitro, the physiologic role of β₂GPI has remained uncertain. β₂GPI inhibits adenosine diphosphate (ADP)-induced platelet aggregation,3 factor XII activation,4prothrombinase activity,5 factor Va degradation by activated protein C,6 and factor Xa generation.7 Despite these multiple regulatory functions, Bancsi et al8 reported that a β₂GPI-deficient family was apparently not at risk for thrombosis.

Recently we encountered 2 Japanese individuals (siblings) with homozygous β₂GPI deficiency. To determine the significance of the physiologic roles of β₂GPI, we examined the profile and behavior of coagulation and fibrinolysis indices in these 2 individuals.

Two individuals with congenital homozygous β₂GPI deficiency were identified. Person 1 was a 36-year-old woman and person 2 was her sibling, a 34-year-old man. Genomic analysis of these persons showed that a thymine corresponding to position 379 of the β₂GPI cDNA9 in exon 4 was deleted; hence, a frame shift would occur and this would make the gene code for an amino acid sequence unrelated to β₂GPI beyond this position.10 These individuals had no history of thrombosis or abnormal bleeding.

Serum levels of mutated β₂GPI were investigated by sandwich enzyme-linked immunosorbent assay, Western blotting, immunoelectrophoresis with rabbit polyclonal anti-human serum antibodies, and Ouchterlony double immunodiffusion with rabbit polyclonal anti-human β₂GPI antibody (Diagnostica Stago, Asnieres, France).

In vitro platelet aggregability was investigated with the plasma from the homozygous β₂GPI-deficient patients in the presence of thrombin or ADP.

Prothrombin time (PT), activated partial thromboplastin time (aPTT), plasma fibrinogen, antithrombin III (ATIII) activity and antigen, plasminogen, and α₂-plasmin inhibitor (α₂PI) were measured as hemostatic and fibrinolytic indices. Thrombin generation and fibrinolytic turnover in vivo were evaluated by measuring plasma levels of prothrombin fragment 1 + 2 (Enzygnost F1 + 2 micro; DADE Behring, USA), thrombin–antithrombin complex (TAT) (TAT test Kokusai F; Kokusai Shiyaku, Japan), plasminogen–plasmin inhibitor complex (PIC test Kokusai F; Kokusai Shiyaku), and D-dimer antigen (LPIA200 D-dimer; Dia-atron). Plasma protein C (Chromogenix, Sweden), protein S antigen (Asserachrom Total Protein S; Diagnostica Stago), and free protein S antigen (Asserachrom Free Protein S, Diagnostica Stago) were quantified. The effect of exogenous activated protein C (APC) in the plasma was tested with a clotting assay (Activated protein C resistance kit; Chromogenix).

To investigate the intensity of thrombin formation, we assessed the rate of thrombin generation by measuring the time taken to reach 50% maximal activity (T1/2).11 In addition, we evaluated PT using highly diluted tissue factor reagent (Dil-PT) to investigate the hypercoagulable state in a patient with β₂GPI deficiency.

Fibrinolytic potential of the euglobulin fraction was also evaluated by fibrin-plate lysing assay in the presence or absence of 2.5 mg/mL of kaolin.

The effect of exogenous β₂GPI on Russell viper venom time (RVVT) was tested using DVV-screen (American Diagnostica Inc, Greenwich, CT). β₂GPI was purified from normal pooled plasma as described.12 To investigate the effect of exogenous β₂GPI on RVVT, we added 25 μL of purified human β₂GPI (0-1250 μg/mL final concentration) to 50 μL of the plasma sample from person 1 or to pooled healthy plasma. After incubation at 37°C for 120 seconds, RVVT was measured.

Plasma mutated β₂GPI was not detected by any of the methods described in “Study design.” Platelets were normally aggregated in β₂GPI-deficient plasma by thrombin or ADP.

The results of hemostatic variables are shown in Table1. PT, aPTT (not done in person 2), fibrinogen, ATIII, plasminogen, and α₂PI were normal in both individuals. No increased thrombin generation or fibrinolysis was indicated because levels of F1 + 2, TAT, PIC, and D-dimer were normal. Plasma levels of protein C activity were slightly increased in both individuals, but those of total protein S antigen and free protein S antigen were within normal values. The effects of exogenous APC in both β₂GPI-deficient plasma samples did not differ from findings in plasma samples from healthy controls. There were no significant differences in T1/2 and Dil-PT between the β₂GPI-deficient plasma and plasma samples of normal subjects. Therefore, no increased in vitro thrombin generation was suggested by these assays. Fibrinolytic activity in the euglobulin from the patients' plasma, as induced by the addition of kaolin, was similar to that in plasma from normal controls.

Shortened RVVT was found in both β₂GPI-deficient persons as compared with controls (Figure1A). To investigate whether the shortened RVVT was due to the β₂GPI deficiency, we added a series of concentrations of exogenous β₂GPI to one of the β₂GPI-deficient plasma samples and measured RVVT. The prolongation of the RVVT by exogenous β₂GPI was dose dependent in the β₂GPI-deficient plasma as well as a normal control plasma (Figure 1B). However, even an excess of exogenous β₂GPI (1250 μg/mL) did not normalize the shortened RVVT. Moreover, Russell viper venom is a strong activator of factor X,13 but such potent substances have not been detected in the human coagulation system in vivo. Therefore, the relevance of the phenomenon remains obscure.

This is the first documentation that most hemostatic and fibrinolytic markers are normal and that there is no increased thrombin generation in individuals with congenital β₂GPI deficiency. In physiologic circumstances, β₂GPI is not a high-affinity phospholipid-binding protein14 and therefore may not interfere with the coagulation/fibrinolysis system. Our data partly support the observation of Bancsi et al8 that congenital β₂GPI deficiency clinically may not be a risk for thrombosis and also show that congenital β₂GPI deficiency is not associated with increased or decreased thrombin formation (ie, “subclinical” thrombotic or bleeding tendency) either in vivo or in vitro. Therefore, congenital β₂GPI deficiency is not a risk factor for either thrombosis or a bleeding tendency.