beta2-glycoprotein I-dependent lupus anticoagulant highly correlates with thrombosis in the antiphospholipid syndrome

The antiphospholipid syndrome is characterized by the presence of antiphospholipid antibodies in plasma of patients with thromboembolic complications. A major problem in defining the syndrome is that serologic assays to detect antiphospholipid antibodies have a low specificity. We recently published a method that specifically detects lupus anticoagulant (LAC) caused by anti-beta(2)-glycoprotein I antibodies. Here, we studied the clinical relevance of detecting beta(2)-glycoprotein I-dependent LAC. Plasma samples were collected from 198 patients with autoimmune diseases. In those samples with a positive partial thromboplastin time-lupus anticoagulant (PTT-LA), a modified activated partial thromboplastin time (aPTT)-based LAC test was performed with cardiolipin as confirming agent. Twenty-five of 58 patients with an aPTT-based LAC were dependent on the presence of anti-beta(2)-glycoprotein I antibodies. Presence of beta(2)-glycoprotein I-dependent LAC was almost completely associated with a history of thromboembolic complications (odds ratio, 42.3; 95% confidence interval, 194.3-9.9). An increased frequency of thrombosis was not found in 33 patients with LAC independent of anti-beta(2)-glycoprotein I antibodies (odds ratio, 1.6; 95% confidence interval, 3.9-0.8). The use of an LAC assay with cardiolipin as confirming agent strongly improves the detection of patients at risk of thrombosis. Our findings suggest that anti-beta(2)-glycoprotein I antibodies with LAC activity are antibodies that are responsible for the thromboembolic complications in the antiphospholipid syndrome.     

Some ''antiphospholipid antibodies'' bind to β2-glycoprotein I in the absence of phospholipid

Some antiphospholipid antibodies (aPL) only bind to anionic phospholipids in the presence of a serum cofactor, beta 2-glycoprotein I (beta 2GPI). Whether these aPL can bind to beta 2GPI in the absence of phospholipid is controversial. We have purified anticardiolipin antibodies (aCL) from the plasma of four patients and beta 2GPI from normal plasma by solid phase affinity methods. All four aCL bound to cardiolipin and phosphatidylserine in the presence of beta 2GPI but not in its absence. The binding of two of the antibodies to cardiolipin and phosphatidylserine at various concentrations of human beta 2GPI was compared with that obtained using 10% bovine serum. The two antibodies responded differently to increasing beta 2GPI concentrations, and binding to phosphatidylserine was relatively greater than to cardiolipin using human beta 2GPI. All four aCL bound to plastic plates coated with beta 2GPI in the absence of phospholipid, and beta 2GPI in the fluid phase had no effect on binding. Binding to beta 2GPI coated plates was increased equally when bovine serum or bovine albumin were used as the sample diluent in place of gelatine. These findings and those of others have important implications for the design of assays for antiphospholipid antibodies.     

Antiphospholipid syndrome: antibodies and antigens

Elucidation of the antibodies and antigens involved in the antiphospholipid syndrome has provided many new insights and research opportunities. The major autoantibodies associated with the syndrome and detected in clinical laboratory assays for antiphospholipid antibodies are directed against prothrombin and beta2-glycoprotein I beta2GPI), a phospholipid-binding plasma protein whose physiological function is unknown. Recent advances in our understanding of these antibodies and antigens include discovery of the crystal structure of beta2GPI, identification of a plasmin cleavage site in beta2GPI, genetic studies of beta2GPI polymorphisms, development of clinical laboratory assays using purified protein antigens, and the identification of antigen specific T cells.     

Antibody specificity and related clinical features in antiphospholipid syndrome

The concept of antiphospholipid syndrome (APS) has been widely accepted. Antiphospholipid antibodies originally included anticardiolipin antibodies and lupus anticoagulants. However, recent advances have shown that most pathogenic antiphospholipid antibodies are directed to phospholipid binding proteins such as β₂-glycoprotein I (β₂-GPI) and prothrombin. Other candidates for antigens of so-called antiphospholipid antibodies are protein C, protein S, factor V, factor X, annexin V, high and low molecular weight kininogens, and complex with β₂-GPI and oxidized low-density lipoprotein (LDL). These autoantibodies directed to different phospholipid binding proteins are present in the blood stream, and contribute to triggering procoagulant effects on endothelial cells and platelets, leading to thrombosis. The heterogeneity of these antiphospholipid antibodies appears to explain various clinical manifestations in patients with APS. The preliminary classification criteria for definite APS and a general management policy have been proposed, although successful treatment of patients with antiphospholipid antibodies have only been shown by retrospective studies. Further prospective investigations are required to confirm the diagnostic and therapeutic criteria for patients with APS.     

Primary antiphospholipid antibody syndrome with mutations in the phospholipid binding domain of beta(2)-glycoprotein I

beta(2)-Glycoprotein I, an anionic phospholipid-binding 50-kDa plasma protein, circulates in the plasma at a concentration of 30-200 microg/ml. Its physiological role remains uncertain, but an important clue to this role is suggested by the finding that antibodies to this protein are frequently found in patients with antiphospholipid antibodies and thrombosis. beta(2)-Glycoprotein I belongs to the complement control protein (CCP) superfamily with five CCP domains. The fifth CCP domain of beta(2)-glycoprotein I has a unique structure and contains a stretch of positively charged amino acids that mediates the binding to phospholipids. This interaction may mediate the clearance of anionic phospholipid-containing surfaces from the circulation. Mutations in this domain affect its binding to phospholipids. We have identified a patient with primary antiphospholipid syndrome who is a compound heterozygous for two mutations in the fifth CCP. One mutation is located in exon 7 (codon 306), and the second mutation is in exon 8 (codon 316). The mutant beta(2)-glycoprotein I was present in normal quantities in his plasma but did not bind to cardiolipin. He had recurrent deep vein thrombosis and pulmonary embolism at age 28 and a thrombotic stroke at age 35, with no other identifiable risk factor for a hypercoagulable state. This report offers some insight into the mechanism of formation of antiphospholipid antibodies and suggests the possible role of the deficiency of beta(2)-glycoprotein I in the pathogenesis of thrombosis.     

Immunologic and hematologic properties of antibodies to prothrombin and plasminogen in a mouse model

Antibodies to prothrombin have been associated with venous and arterial thrombosis, and they cross-react with a structurally closely related protein plasminogen. We immunised 16 mice with human prothrombin and 15 mice with human plasminogen. Mice immunised with prothrombin developed cross-reactive antibodies to plasminogen (12/16), beta2-glycoprotein I (4/16), tissue-type plasminogen activator (6/16) and cardiolipin (11/16). Mice immunised with plasminogen developed cross-reactive antibodies to prothrombin (8/15), tissue-type plasminogen activator (2/12) and cardiolipin (5/12). Functional effects of antibodies were examined. Immunisation with prothrombin induced lupus anticoagulant activity in 9/14 mice. In mice immunised with plasminogen, radial fibrinolysis was inhibited in 8/10 and plasminogen activation in the chromogenic assay was inhibited in 9/11. No cross-functionality was observed. In conclusion, antibodies to prothrombin and plasminogen cross-react in vivo. Antibodies to prothrombin and plasminogen have different functional profiles, immunisation with prothrombin leads to prolonged blood clotting time, and immunisation with plasminogen induces antibodies interfering with fibrinolysis.     

Antiphospholipid antibodies require beta 2-glycoprotein I (apolipoprotein H) as cofactor.

IgG was isolated from 9 patients with high levels of anticardiolipin antibodies (aCL). The standard ELISA was modified, using gelatin to postcoat ELISA plates and as diluent. Under these conditions, 8/9 samples of IgG bound to cardiolipin only in the presence of a cofactor, found in fetal calf serum and normal human serum (NHS). beta 2-glycoprotein I (apolipoprotein H) was at least as effective as NHS as a cofactor for 7/9 IgG samples. Our studies confirm that for some aCL detected in conventional assays, the antibody will react with cardiolipin only if beta 2-glycoprotein I is present.     

Diagnosis and Management of Patients with the Antiphospholipid Syndrome

The antiphospholipid syndrome is an autoimmune disorder characterized by venous or arterial thrombosis, recurrent pregnancy loss, and thrombocytopenia combined with laboratory tests that indicate the presence of antibodies against phospholipid-binding proteins. The antibodies are directed against a complex of phospholipid with a protein such as 2-glycoprotein I (2-GPI) or prothrombin and are detected by means of phospholipid-dependent coagulation assays (known as assays for lupus anticoagulants) and by ELISAs that contain 2-GPI and a phospholipid (e.g., cardiolipin). The antiphospholipid syndrome can be associated with other connective tissue disorders such as systemic lupus erythematosus or may be the only manifestation of an autoimmune disorder. Management of patients with this disorder usually includes anticoagulation, which has been found to reduce the rate of recurrence of venous and arterial thrombosis as well as the rate of fetal loss.     

Antiphospholipid syndrome: clinical and diagnostic utility of laboratory tests

Antiphospholipid antibodies are a wide and heterogeneous group of immunoglobulins that include, among others, lupus anticoagulants and anticardiolipin antibodies. Their presence in patients with arterial and venous thrombosis, and/or obstetrical complications, defines the antiphospholipid syndrome. Antiphospholipid antibodies do not recognize anionic phospholipids, but recognize plasma proteins bound to suitable anionic surfaces: beta2-glycoprotein I and prothrombin are the most common and most frequently investigated antigens. We systematically reviewed published articles on the antiphospholipid syndrome to investigate the association between thrombosis and some antiphospholipid antibodies. Lupus anticoagulants were a clear risk factor for thrombosis, irrespective of the site and type of thrombosis, the presence of systemic lupus erythematosus, and the methods used to detect them. Anticardiolipin and anti-beta2-glycoprotein I antibodies were possible risk factors of thrombosis, at least in some selected situations. The measurement of antiprothrombin antibodies is not helpful to define the patient's risk of thrombosis. These results are mainly due to the still far from optimal standardization of the methods to detect antiphospholipid antibodies; the lack of standardized reference materials; the heterogeneity in reagents, calibrators, and assay conditions; and the methods used to calculate the results. Many efforts are currently being made to improve assay standardization and harmonization, which should help to clarify the clinical relevance of antiphospholipid antibodies.     

Accelerated atheroma, antiphospholipid antibodies, and the antiphospholipid syndrome

Indirect data coming from animal studies and in vitro observations support the contention that the mere presence of antiphospholipid antibodies may be sufficient to increase atheroma development, regardless of other predisposing factors. It seems that humoral and cellular immune responses to beta 2-glycoprotein I can play an important role in mediating the increased propensity to atherosclerosis.     

Autoantibodies Directed Against Domain I of Beta2-Glycoprotein I

Patients diagnosed with the antiphospholipid syndrome typically suffer from vascular thrombosis, pregnancy morbidity, or a combination of the two. Due to the high prevalence of these clinical symptoms, the diagnosis of antiphospholipid syndrome is almost completely dependent on the detection of antiphospholipid antibodies in patient plasma. However, not every individual with antiphospholipid antibodies in his or her plasma suffers from thrombosis and/or pregnancy morbidity, which suggests the existence of different populations of antiphospholipid antibodies. Although many antigens have been identified in relation to the antiphospholipid syndrome, β2-glycoprotein I is regarded as clinically most significant. During the past decade, evidence has accumulated to suggest the presence of a dominant epitope on the first domain of β2-glycoprotein I. Several studies have detected a specific population of antibodies recognizing a cryptic epitope on domain I, at least comprising arginine 39 to arginine 43. In contrast to antibodies recognizing other domains of β2-glycoprotein I, anti-domain I antibodies are found to be highly associated with clinical symptoms. This review discusses several studies that have investigated a role for domain I within the antiphospholipid syndrome on a predominantly diagnostic level.     

β2-glycoprotein 1 (β2GP1) enhances cardiolipin binding activity but is not the antigen for antiphospholipid antibodies

Some investigators have reported that a serum protein, beta 2-glycoprotein 1 (beta 2GP1), either alone or in combination with negatively charged phospholipid, may be the antigen for anticardiolipin (aCL) antibodies. To examine these reports further, ELISA tests, inhibition experiments, Ouchterlony and Western blot techniques were used to examine anticardiolipin binding to beta 2GP1. Sera from patients with the antiphospholipid syndrome (APS) and syphilis were studied, as well as whole IgG immunoglobulin and affinity purified (a.p.) IgG aCL antibodies. Results showed no binding of aCL antibodies to beta 2GP1 in the absence of cardiolipin. beta 2GP1 caused enhanced binding of aCL antibodies to cardiolipin, but this enhancement was not observed in inhibition experiments. Binding to cardiolipin occurred in the absence of beta 2GP1. Enhancement of cardiolipin binding activity by beta 2GP1 was observed for APS, but not for syphilis. We conclude that beta 2GP1 is not the antigen for aCL antibodies, nor is it likely that the antibody recognizes shared beta 2GP1-cardiolipin epitopes. Instead, this protein may make cardiolipin more available for aCL binding on solid surfaces by some yet undefined mechanism. This effect may not extend to aqueous suspensions.     

The nature of antiphospholipid antibodies.

Despite the striking clinical manifestations associated with antiphospholipid antibodies (aPL) the role of these autoantibodies in disease and the nature of their true "inducing" and "target" antigens remain elusive. To address these issues, we investigated the immunogenic potential of phospholipid structures. To date, phospholipid immunogens have included hexagonal (II) forms of phosphatidylethanolamine and mixtures of apolipoprotein H (beta 2-glycoprotein I) with cardiolipin. Both hexagonal (II) phosphatidylethanolamine and the cardiolipin/apolipoprotein H mixture were capable of inducing aPL with lupus anticoagulant activity. Bilayer phosphatidylethanolamine and cardiolipin in the absence of apolipoprotein H were nonimmunogenic. Our data support our views that specific phospholipid structures are recognized by the immune system and that such structures serve as inducing and/or target antigens in the pathogenesis of aPL in vivo.     

T cells that are autoreactive to beta2-glycoprotein I in patients with antiphospholipid syndrome and healthy individuals

OBJECTIVE: To identify the T cells responsive to beta2-glycoprotein I (beta2GPI) that mediate antiphospholipid antibody production in patients with antiphospholipid syndrome (APS). METHODS: In vitro proliferative responses and anti-beta2GPI antibody production induced by beta2GPI were examined in peripheral blood mononuclear cell (PBMC) cultures from 12 APS patients, 13 systemic lupus erythematosus patients without APS, and 12 healthy donors. RESULTS: Peripheral blood T cells from all subjects failed to respond to beta2GPI in its native form. In contrast, reduced beta2GPI was able to stimulate T cells not only from all 12 patients with anti-beta2GPI antibodies, but also from 10 of 25 individuals without anti-beta2GPI antibodies. The specificity of the responses to beta2GPI was confirmed by activation of the reduced beta2GPI-primed T cells by recombinant beta2GPI in secondary cultures. Characterization of the T cell response induced by beta2GPI revealed that the response was associated with the presence of the DR53-associated alleles, the responding T cells were CD4+ and restricted by HLA class II, and antigenic peptides were located in domains IV and/or V. Anti-beta2GPI antibody production was induced specifically in anti-beta2GPI antibody-positive patients, in PBMC cultures with reduced beta2GPI. Anti-beta2GPI antibodies produced in vitro recognized beta2GPI immobilized with cardiolipin or beta2GPI coated on "high-binding" polystyrene plates. CONCLUSION: These results strongly suggest that CD4+ and HLA class II-restricted T cells responsive to beta2GPI are involved in the production of antiphospholipid antibodies in APS patients.     

Atherosclerosis-related markers in systemic lupus erythematosus patients: the role of humoral immunity in enhanced atherogenesis.

Systemic Lupus Erythematosus (SLE) patients experience premature atherosclerosis. A deranged lipid metabolism and use of immunosuppressive medications accounts partially for the accelerated process. The role of autoimmunity in atherosclerosis has recently been highlighted. Autoantigenic determinants thought to play a role in the development of atherosclerosis include: modified lipoproteins, heat shock proteins and beta2-glycoprotein I (a target of 'autoimmune' anticardiolipin antibodies). In this present work we determined autoimmune markers which may be associated with premature atherosclerotic process found in SLE patients. We have found that antibodies to oxLDL were raised in the sera of lupus patients and cross-reacted with cardiolipin and with beta2GPI. OxLDL containing immune-complexes of the IgG and IgM isotypes were both elevated in the SLE patients as compared with healthy controls. Patients with high Lipoprotein (a) concentrations (>30 mg/dl) had higher levels of IgM oxLDL-containing immune-complexes. IgM but not IgG anti-HSP-65 antibodies were elevated in the lupus patients and levels of oxLDL containing immune-complexes correlated positively with the presence of anti-HSP 65 antibodies. Lysophosphatidylcholine (LPC) is a peroxide-derivative formed during LDL oxidation, was shown to evoke a humoral response in healthy subjects. Antibodies to lysophosphatidylcholine of the IgG but not the IgM isotype were reduced in SLE patients compared with controls, suggesting it may be 'consumed' into oxLDL containing immune complexes. Therefore, SLE patients exhibit a humoral autoimmune response towards the antigenic candidates incriminated in the progression of atherosclerosis. These findings may help identify factors that are involved in accelerating atherogenesis in SLE patients.     

Beta(2)-glycoprotein I promotes the binding of anionic phospholipid vesicles by macrophages

Beta(2)-Glycoprotein I is a single-chain 50-kDa protein that circulates in plasma at a concentration of approximately 200 microg/mL. Its physiological role remains uncertain, but an important clue is the frequent presence of antibodies to this protein in patients with recurrent thrombosis. We have isolated beta(2)-glycoprotein I and examined its effect on the binding of phosphatidylserine (PS) vesicles by human monocyte-derived macrophages and by phorbol ester-stimulated THP-1 cells. beta(2)-Glycoprotein I stimulated the binding of PS vesicles by these cells in a concentration-dependent manner. Vesicles containing other anionic phospholipids, such as cardiolipin, phosphatidic acid, or cardiolipin, inhibited the binding, whereas PC vesicles had no effect. Platelet-derived microvesicles, which contain anionic phospholipid on the outer leaflet of their phospholipid bilayer, also inhibited beta(2)-glycoprotein I-dependent binding of anionic phospholipid vesicles. The binding is associated with incorporation of phospholipid in the cell membrane and internalization of beta(2)-glycoprotein I. These findings suggest a physiological function for beta(2)-glycoprotein I in the clearance of procoagulant anionic phospholipid-containing cell surfaces from the circulation.     

Characterization of β2-glycoprotein I-dependent and -independent “antiphospholipid” antibodies from lupus-prone NZW/BXSB F1 hybrid male mice

Male (NZW × BXSB)F1 (W/BF1) mice develop a systemic lupus-like syndrome characterized by thrombocytopenia, coronary vascular disease, nephritis, and anticardiolipin antibodies. Three stable hybridoma cell lines secreting monoclonal anticardiolipin antibodies were developed from these mice by fusing their splenic lymphocytes with nonsecreting myeloma cell line, NS-1. Monoclonal antibody A1.17 reacted with cardiolipin in a β₂-Glycoprotein I-dependent manner. The epitope for this antibody consisted of β₂-glycoprotein I bound to cardiolipin or immobilized on plastic plates. Other anionic phospholipid-binding proteins, such as prothrombin or annexin V, had no significant effect in the reactivity of these antibodies. The specificity is similar to the autoimmune anticardiolipin antibodies described in patients with systemic lupus erythematosus and other infectious diseases. In contrast, monoclonal antibodies A1.72 and A1.84 reacted with cardiolipin in the absence of β₂-glycoprotein I. β₂-glycoprotein I, either in the fluid phase or bound to cardiolipin, inhibited the binding of these antibodies. The specificity of the latter two antibodies was similar to that described in patients with syphilis and allied disorders. Both types of antibodies had lupus anticoagulant properties. Thus lupus-prone male (NZW × BXSB)F1 (W/BF1) mice develop both β₂-glycoprotein I-dependent and β₂-glycoprotein I-independent anticardiolipin antibodies. Am. J. Hematol. 56:86–92, 1997. © 1997 Wiley-Liss, Inc.     

Anti β2-glycoprotein I antibodies in a patient with catastrophic antiphospholipid syndrome

Summary Autoimmune antiphospholipid antibodies are a hallmark of patients with antiphospholipid syndrome, and require a protein cofactor, 2-glycoprotein I, to bind anionic phospholipids. In these same patients, moreover, IgG directly binding 2-glycoprotein I are described. We found high plasma titres of both IgM and IgG anti 2-glycoprotein I antibodies in a patient with catastrophic antiphospholipid syndrome. After passing plasma through a Sephacryl S-300 column, an identical distribution pattern between anti 2-glycoprotein I and anticardiolipin antibodies was observed. Moreover, when IgG immunocomplexes were isolated from a high molecular fraction, IgM anti-2-glycoprotein I and anticardiolipin antibodies were detected. Thus IgG, IgM and IgG-IgM complexes of anti-2-glycoprotein I antibodies are present at the same time in a patient with antiphospholipid syndrome.     

Antibodies to factor XII and recurrent fetal loss in patients with the anti-phospholipid syndrome

Forty female patients with either primary anti-phospholipid syndrome (n = 26) or systemic lupus erythematosus (anti-phospholipid syndrome positive) (n = 14) were investigated for levels of factor XII, the presence of lupus anticoagulant and antibodies to cardiolipin, beta 2-glycoprotein I and factor XII. Twenty-one patients had a history of recurrent fetal loss (> 2, mean = 2.6). Lupus anticoagulant positivity showed a weak association with recurrent fetal loss (odds ratio = 1.1). While there was no association between the presence of antibodies to cardiolipin or beta 2-glycoprotein I with recurrent fetal loss, antibodies to factor XII showed a strong and statistically significant association (odds ratio = 5.4, P = 0.025).     

Heterogeneity of antibodies to beta2-glycoprotein 1 from patients with systemic lupus erythematosus

We studied antibodies to beta2-glycoprotein 1 (anti-beta2GP1) from 72 patients with systemic lupus erythematosus (SLE) with or without antiphospholipid syndrome (APS) or with or without anticardiolipin antibodies (aCL). Fifteen patients had APS and positive antiphospholipid antibodies [clinical APS(+)/aPL(+)], 12 patients had APS, negative serum IgG and IgM aCL, antiphosphatidylethanolamine, anti-phosphatidylserine and no lupus anticoagulant [clinical APS(+)/ aPL(-)]. A third group included 16 patients without APS but high aCL levels [clinical APS(-)/ aPL(+)]. In a fourth group we studied 29 patients without clinical manifestations of APS or aCL [clinical APS(-)/aPL(-)]. One hundred anticardiolipin and VDRL-negative normal sera were studied as controls. IgG antibodies to cardiolipin proper in a bovine beta2GP-free system, to human beta2GP1 immobilized on cardiolipin or to human beta2GP1 alone were detected in all sera by ELISA using irradiated and nonirradiated plates from two manufacturers. Sera from APS(+)/aPL(+) patients showed IgG binding to CL, CL + beta2GP1 and beta2GP1 in irradiated and nonirradiated plates. APS(+)/ aPL(-) sera had more significant IgG binding to beta2GP1 than normal controls when studied in both irradiated or nonirradiated plates (P = 0.001). This binding was inhibited by solid-phase cardiolipin in a dose-dependent manner. Sera from the APS(-)/aPL(+) subgroup had comparable IgG activity in both the CL and CL + beta2GP1 assays, while no anti-beta2GP1 activity was detected in these sera. Sera from the clinical APS(-)/aPL(-) patients were negative in the three ELISA systems. Antibodies to human beta2GP1 from SLE patients recognize various epitopes. Those from APS(+)/ aPL(+) patients appear to react with an epitope boosted by cardiolipin in addition to another one present in the native protein. In contrast, anti-beta2GP1 from patients with APS(+)/aPL(-) are blocked by cardiolipin, suggesting that their epitope is the phospholipid-binding site.