1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

Clinical importance of the antiphospholipid antibody syndrome (APS).

APS is a condition in which clinical features such as vascular thromboses, recurrent fetal loss, and thrombocytopenia (1, 2) occur in association with the presence of antiphospholipid antibodies (aPL) in the blood. APS is now the most common cause of acquired hypercoagulability in the general population (3) and a major cause of morbidity in pregnancy (4). APS may be considered primary if there is no coexistent autoimmune rheumatic disease (ARD) (5) or secondary when in the presence of another ARD (6).

The risk of thrombosis and recurrent fetal loss in APS is high. A multicenter study of 1,000 patients with APS (4) showed that 39% of these patients had a deep vein thrombosis, 20% had strokes, 11% had transient ischemic attacks, and pulmonary embolism occurred in 14%, while <50% of 1,580 pregnancies continued to term. In both primary and secondary APS, recurrence rates of up to 29% for thrombosis and a mortality of up to 10% over a 10-year followup period (7) have been reported. The only treatment proven to reduce the risk of thrombosis in APS is long-term anticoagulation (8), which may have severe side effects. It is therefore important to develop new treatments that are both more effective and more accurately targeted to the disease process in APS. In particular, it may be advantageous to block or manipulate interactions between aPL and their major epitopes, since considerable evidence exists that aPL play a major role in the clinical features of APS (9–13).

Importance of β2-glycoprotein I (β2GPI) in APS.

The aPL that are found in patients with APS generally target negatively charged phospholipids and require serum cofactors for their binding to phospholipids. These aPL are, in fact, directed against an array of protein cofactors, such as β2GPI, protein C, protein S, and prothrombin, which then bind to anionic phospholipids. β2GPI is the most extensively studied (14–17) of these protein cofactors and appears to be one of the most relevant clinically (18–20). In contrast, nonpathogenic aPL, which are not associated with the development of clinical features of APS, do not recognize β2GPI and bind both neutral and negative phospholipids without any cofactor dependence (21). Binding to β2GPI, although undoubtedly important, is not an absolute requirement in determining the pathogenicity of aPL, since human monoclonal IgG aPL that bind phospholipids in the absence of β2GPI have been isolated and shown to be highly pathogenic in vivo (13, 22).

Although β2GPI is a major target antigen in the pathogenesis of APS, the exact nature of the aPL–β2GPI interaction remains a matter of some debate. Opinion is divided as to whether pathogenic aPL are directed against the β2GPI–phospholipid complex (23) or against a cryptic epitope revealed on β2GPI by binding to phospholipids (or certain synthetic surfaces) (24), or whether they bind directly to an increased density of β2GPI immobilized on phospholipids or γ-irradiated plates (25) (see Figure 1). The location of the major epitopes on β2GPI is also unclear, and published evidence supports the existence of epitopes in various different regions of β2GPI.

thumbnail image

Figure 1. Schematic representation of various hypotheses regarding the interaction between antiphospholipid antibodies (aPL) and β2-glycoprotein I (β2GPI). A, Pathogenic aPL bind to the phospholipid (PL)–β2GPI complex. B, Pathogenic aPL only bind to a conformationally altered, thus cryptic, epitope on β2GPI. C, Increased density of β2GPI on anionic phospholipids is required for bivalent aPL binding.

Download figure to PowerPoint

The purpose of this article is to review the evidence from experiments that attempt to define the nature and location of epitopes on β2GPI that are bound by aPL. Recent data describing the crystal (26, 27) and solution (28) structure of β2GPI are also reviewed to elucidate the aPL–β2GPI–phospholipid interaction, which appears to play an important role in the pathogenesis of APS. An understanding of these processes is potentially important to enable the design of drugs that would inhibit this interaction. No single drug or epitope, however, is likely to be suitable for interaction with all aPL, since these antibodies are heterogeneous. Thus, to develop new therapeutic strategies, it is necessary to identify those epitopes on β2GPI that interact with aPL from the majority of patients with APS. This review will assess the extent to which this can be done, on the basis of the evidence currently available.

Molecular structure and function of β2GPI

  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

A detailed account of the molecular structure of β2GPI is beyond the scope of this report (for a detailed review, see refs.20 and29). Briefly, β2GPI is a 50-kd single-chain polypeptide (30) with 5 oligosaccharide attachment points (31, 32). It is composed of 5 common, structurally related repeats known as complement control protein domains, each of which is ∼60 amino acid residues in length, in addition to domain V, which contains 82 amino acids and a long carboxyl-terminal tail.

The precise physiologic function of human β2GPI remains unclear. In vitro, β2GPI binds to negatively charged phospholipids (such as cardiolipin [CL] and phosphatidylserine [PS]) as well as inhibiting contact activation of the intrinsic coagulation pathway (33), platelet prothrombinase activity (34), and ADP-mediated platelet aggregation (35). It also aids clearance of oxidized low-density lipoproteins (LDL) (36). In vivo, β2GPI has been shown to be involved in the clearance of apoptotic bodies (37) and liposomes (38) as well as interfering with the protein C pathway (39). More recently, β2GPI was found to adhere to resting human endothelial cells. Subsequent binding of circulating aPL led to endothelial cell activation (40) (for review, see ref.41). Thus, β2GPI appears to promote clearance of certain products of oxidation and prevent thrombus formation (for review, see ref.29). Binding of aPL to epitopes on β2GPI may interfere with these functions and thus promote thrombosis.

β2GPI may also have an effect on the development of atherosclerosis, which has been associated with aPL in several studies (42, 43). In the presence of aPL, β2GPI increases the uptake of oxidized LDL by macrophages, an important step in the formation of an atherosclerotic plaque (for review, see ref.44). β2GPI-dependent aPL may cause atherosclerosis by direct activation of the vascular endothelium (40, 45, 46), increased clearance of oxidized LDL (36), or reduced paraoxonase activity, leading to increased oxidation of LDL (47).

What is the nature of the epitope for binding of aPL?

  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

Evidence in favor of a cryptic epitope on β2GPI.

Matsuura et al (48) demonstrated that polyclonal human aPL and monoclonal murine aPL bound directly to β2GPI coated on an irradiated polystyrene surface in the absence of CL. Binding of aPL was inhibited by the simultaneous addition of CL-coated latex beads and β2GPI, but not by the addition of β2GPI alone. These aPL did not bind directly to β2GPI in either the fluid or solid phase on plain (nonirradiated) plates. The authors concluded that the specificity demonstrated by aPL in these experiments was directed against a cryptic epitope on β2GPI, exposed by a conformational change following interaction either with an irradiated surface or with phospholipids (48).

Chamley et al (49) also found that polyclonal aPL isolated from the sera of APS patients bound to β2GPI when it was immobilized on irradiated, but not plain, polystyrene enzyme-linked immunosorbent assay (ELISA) plates, even when the concentration of β2GPI on the plain plates was higher. Thus, the authors postulated that a conformational change in β2GPI was required to explain the increased binding of aPL to irradiated plates.

Indirect methods using infrared spectroscopy (50), spectrophotometry to produce a circular dichroism spectrum (51), and microcalorimetry (52) have all been used to infer an alteration in the structure of β2GPI upon binding to phospholipids. These experiments, however, do not prove that such a conformational change is essential for binding of aPL to β2GPI. Indeed, the results obtained by Subang et al (51) suggest quite the opposite. These authors demonstrated a conformational change in β2GPI upon binding to CL, but not upon binding to other anionic phospholipids such as phosphatidylglycerol (PG) and PS. Despite this finding, β2GPI complexes with PG and PS were still found to be immunogenic, leading to the production of aPL and lupus anticoagulant (LAC) in BALB/c mice.

The binding of aPL to β2GPI and β2GPI to CL is significantly reduced by plasmin cleavage of β2GPI between the residues Lys-317 and Thr-318 in domain V (53, 54). Matsuura et al (55) showed no binding of polyclonal or monoclonal human aPL or monoclonal murine aPL to plasmin-cleaved β2GPI on polyoxygenated plates. Binding of β2GPI to the plates was unaffected by plasmin cleavage as measured by binding to murine anti-human β2GPI. Since this experiment measured binding of aPL to β2GPI in the absence of phospholipids, these results were taken as evidence that proteolytic cleavage of β2GPI between Lys-317 and Thr-318 does not simply reduce binding to phospholipids, but also prevents exposure of cryptic epitopes for aPL.

Evidence against a cryptic epitope.

Roubey et al (25) proposed that the increased aPL binding to β2GPI on irradiated plates may be due to an increased density of the antigen, rather than being the result of exposure to a cryptic epitope. At coating concentrations higher than 0.5 μg/ml, the amount of β2GPI bound to the plain plates reached a plateau, whereas amounts bound to the γ-irradiated plates increased. Binding of aPL, however, was detected only when the amount of β2GPI coated on the high-binding plates exceeded that which could be coated on the plain plates. Inhibition experiments demonstrated little detectable binding of aPL to physiologic concentrations of fluid-phase β2GPI or a low density of immobilized β2GPI. Binding of monovalent Fab′ fragments of aPL to β2GPI on irradiated plates was substantially less than that of bivalent F(ab′)2 fragments (25). Roubey et al thus argued that aPL that bind to β2GPI are of low affinity, requiring bivalent binding, which only occurs when β2GPI is “clustered” on an irradiated plate or anionic phospholipid surface.

The striking differences between the results obtained by Roubey et al (25) and those by Chamley et al (49) may be explained in a number of ways. First, the 2 groups used different techniques to quantify the amount of β2GPI bound to the plates, and carried out their experiments at significantly different coating concentrations. Second, different antibodies were tested in the 2 experiments. It is possible that some aPL require the presence of a cryptic epitope for binding to β2GPI, whereas others do not.

Reddel et al (56) have confirmed, using an ELISA with irradiated plates, that an antigen-coating concentration threshold exists for the detection of polyclonal anti-β2GPI antibodies isolated from APS patients, a pattern consistent with that expected for bivalent binding of the antibodies studied. Iverson et al (57) studied the reactivity of aPL with various β2GPI domain-deleted mutants (DM) expressed in insect cells. The expression vectors were designed such that expression products contained carboxyl-terminal Gly/His-6 tags for binding to nickel chelate plates. This binding involves the tag, but not the protein itself, and should therefore not induce a conformational change in the DM molecules. The fact that these expression products (including the full-length β2GPI product) bind aPL despite undergoing no conformational change on binding to the plate is further evidence that exposure of a cryptic epitope is not always required for binding.

The insect expression system was also used to produce a variant of β2GPI that dimerizes spontaneously (58). This variant bound polyclonal human aPL better than wild-type β2GPI on nonirradiated plates, wild-type β2GPI on irradiated plates, and β2GPI in the fluid phase. Bivalent F(ab′)2 fragments, but not monovalent Fab′ fragments, could bind the dimerizing variant (58). These experiments again suggest the importance of bivalent binding of low-affinity aPL to clustered β2GPI. It is possible, however, that the process of dimerization itself might lead to exposure of a cryptic epitope.

Using surface plasmon resonance, the interactions of 5 monoclonal murine aPL with β2GPI have been studied (59). Molar ratios of the monoclonal aPL binding to β2GPI immobilized on sensor chips were ∼0.5, indicating bivalent binding.

If soluble native β2GPI were able to inhibit the anti-β2GPI binding activity of aPL, this would present a strong case for the existence of a noncryptic epitope on native β2GPI. The investigators who have been unable to demonstrate such inhibitory activity of β2GPI in solution have generally used very low concentrations (e.g., 80 μg/ml [60] and 8 μg/ml [61]) of fluid-phase β2GPI. Guerin et al (62), however, failed to absorb the antibody reactivity in APS sera when using up to 640 μg/ml of β2GPI in solution. In contrast, many other authors have found that sufficiently high concentrations of β2GPI (∼400 μg/ml) in free solution significantly reduce aPL binding to β2GPI on ELISA plates, in a dose-dependent manner (25, 58, 63–67). The evidence that native β2GPI may be recognized by aPL in Western blotting experiments is conflicting; some authors demonstrate binding (68, 69), but others do not (62, 70, 71).

Conclusion: Is the epitope cryptic?

It is important to determine whether the epitope is cryptic if we hope to design an effective drug to treat APS that works by blocking the aPL–β2GPI interaction. If the major epitopes are cryptic, then any drug developed on the basis of its ability to block the interaction between aPL and native β2GPI would be highly likely to fail.

A summary of the conflicting data both for and against the presence of cryptic epitopes is shown in Tables 1 and 2. The binding of aPL to β2GPI on certain synthetic surfaces in the absence of phospholipids can be interpreted both for and against a cryptic epitope for aPL on β2GPI. Increased binding of aPL to mutant forms of “dimerized” β2GPI as well as a lack of binding of monovalent aPL Fab′ fragments confirm that many aPL are intrinsically low-affinity antibodies for which bivalent binding is required on clustered β2GPI. Evidence of aPL binding to fluid-phase native β2GPI strongly supports the argument against the existence of a cryptic epitope.

Table 1. Experimental evidence supporting a cryptic epitope on β2-glycoprotein I (β2GPI)*
Author (reference)Origin of aPLBinding to fluid-phase β2GPIResult/conclusion
  • *

    aPL = antiphospholipid antibodies; SAPS = secondary antiphospholipid antibody syndrome; CL = cardiolipin; APS = antiphospholipid antibody syndrome; ND = not determined; PL = phospholipids; CD = circular dichroism; PG = phosphatidylglycerol; PS = phosphatidylserine; PAPS = primary antiphospholipid antibody syndrome; SLE = systemic lupus erythematosus; ARD = autoimmune rheumatic disease; LAC = lupus anticoagulant.

  • For determination of SAPS (secondary to SLE), the APS classification criteria were not applied.

  • The APS classification criteria were applied, with no distinction made between primary or secondary APS.

  • §

    The APS classification criteria were not applied.

  • Hypercoagulable disorders, with possible APS in some patients.

Matsuura et al (48)SAPS sera; murine monoclonal aPLNoaPL only bind to β2GPI on irradiated plate and CL-coated latex beads
Chamley et al (49)APS seraNoaPL bind to 1.25 μg/ml β2GPI on irradiated plate and >40 μg/ml β2GPI on plain plate
Borchman et al (50)NoneNDConformational change of β2GPI on binding PL, by infrared spectroscopy
Subang et al (51)Murine monoclonal aPLNDConformational change of β2GPI on binding CL, by CD analysis; β2GPI–CL, β2GPI–PG, and β2GPI–PS immunogenic in BALB/c mice
Maatsura et al (55)PAPS and SAPS sera; murine monoclonal aPL§NDNo binding of aPL to plasmin-cleaved β2GPI
Keeling et al (60)SLE and PAPS sera§NoaPL bind to β2GPI in absence of PL with no inhibition by soluble β2GPI
Martinuzzo et al (61)PAPS, SAPS, ARD, and other disease sera§NoSoluble β2GPI up to 8 μg/ml unable to inhibit aPL
Guerin et al (62)APS seraNoβ2GPI only recognized when bound to high-affinity support
Pierangeli et al (70)APS seraNoβ2GPI enhanced binding of aPL to PL, but no aPL binding to β2GPI alone; no inhibition of aPL binding with soluble β2GPI up to 200 μg/ml
Permpikul et al (71)LAC-positive seraNoNo aPL binding to β2GPI on Western blot
Table 2. Experimental evidence supporting a noncryptic epitope on β2GPI*
Author (reference)Origin of aPLBinding to fluid-phase β2GPIResult/conclusion
  • *

    SPR = surface plasmon resonance; ELISA = enzyme-linked immunosorbent assay (see Table 1 for other definitions).

  • The APS classification criteria were not applied. SAPS was secondary to SLE.

  • The APS classification criteria were not applied, and no distinction was made between primary or secondary APS.

  • §

    The APS classification criteria were applied. SAPS was secondary to SLE.

Roubey et al (25)PAPS and SAPS seraYesaPL only bind to β2GPI on irradiated plate and CL-coated latex beads; binding of aPL to Fab << F(ab)2
Iverson et al (57)PAPS and SAPS seraYesaPL binding to domain I of β2GPI immobilized via Gly/His-6 tag
Sheng et al (58)APS seraYes, dimerized > native β2GPIaPL binding to dimerized β2GPI > native β2GPI, in fluid and solid phase
Regnault et al (59)Murine monoclonal aPLYesBivalent interaction with immobilized β2GPI >>> soluble β2GPI, using SPR
Tincani et al (67)PAPS and SAPS seraYesaPL low-affinity binding to solid-phase β2GPI >> fluid-phase β2GPI in absence of PL
Viard et al (64)SLE and SAPS seraYesSoluble β2GPI inhibits the binding of aPL in a dose-dependent manner; clinical association of anti-β2GPI antibodies with APS
Balestrieri et al (63)SLE, SAPS, and PAPS seraYesSoluble, high concentration of β2GPI inhibits binding of IgG, but not IgM aPL
Forastiero et al (66)SLE, PAPS, SAPS, and aPL-positive seraYesaPL bind to β2GPI on dot-blot, correlating with anti-β2GPI on ELISA
Cabiedes et al (68)SLE, SAPS sera§NDaPL bind to β2GPI on Western blot, correlating with anti-β2GPI on ELISA
Sorice et al (69)PAPS, SAPS, SLE sera§NDaPL bind to β2GPI on Western blot

It is important to recognize that the ideas of cryptic epitopes and high-density binding are not mutually exclusive. A particular aPL may bind preferentially to a cryptic epitope exposed when β2GPI binds to an irradiated plate, but may also be capable of binding to native β2GPI when presented at sufficiently high density. Overall, however, the evidence in favor of the importance of binding to noncryptic epitopes currently seems to be the most persuasive, because it has been obtained using a wider range of techniques and antibodies.

Which domains of β2GPI contain major epitopes for aPL?

  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

It is important to know which of the 5 domains of β2GPI contains epitopes most likely to bind aPL in the majority of patients with APS, since drugs directed against these interactions are most likely to be useful. During the last 8 years, the evidence has demonstrated the specificity of aPL for epitopes on all 5 domains of β2GPI. The basis of each of these claims is summarized in Table 3 and will be discussed herein in light of the recently discovered 3-dimensional crystal structure of β2GPI. As much as possible, the origin of each aPL studied has been localized to specific disease groups or subsets. The use of any diagnostic or classification criteria (2, 72–74) to define APS has also been noted. Those aPL derived from patients with primary APS may have different binding properties compared with those derived from patients with secondary APS or from those with other diseases. Lack of patient stratification and lack of control groups as well as limited statistical power are commonplace problems among these studies, making it difficult to decide whether the conclusions are likely to be applicable to the majority of patients with APS.

Table 3. Summary of experimental data linking various domains of β2GPI with aPL binding*
Author (reference)MethodOrigin of aPLOriginal conclusionReassessment
  • *

    DM = domain-deleted mutant; dRVVT = dilute Russell viper venom test; KCT = kaolin clotting time (see Tables 1 and 2 for other definitions).

Hunt et al (53)Plasmin-cleaved β2GPIPAPS seraDomain V: PL binding and cofactor activityCrystal structure reveals domain V to be intimately involved in PL binding; domain V DMs do bind aPL
Hunt and Krilis (54)Domain V peptidesSLE and PAPS seraDomain V: PL binding (CKNKEKKC) and aPL binding (intact Lys-317–Thr-318)
Wang et al (79)Domain V peptidesPAPS monoclonal aPLDomain V: PL binding (CKNKEKKC) and aPL binding (? exact site) 
Sheng et al (75)Mutagenesis Lys-284, -286, -287APS seraLys-284, -286, -287 critical for PL binding 
Igarashi et al (80)DMPAPS monoclonal aPL and murine monoclonal aPLDomain V: PL binding; domains I–IV: favored domain IV aPL bindingCrystal structure reveals domains III and IV to be heavily glycosylated; domain IV produced from insect cells may be less glycosylated than human β2GPI
George et al (81)DMPAPS and SAPS sera, murine monoclonal aPLDomain V: PL binding; domains I–IV: aPL binding
Blank et al (84)Hexapeptide phage displayPAPS and normal seraDomains I–II interlinker, domain III and domain IV contain epitopes for aPL binding 
Iverson et al (57)DM: Gly/His-6 tagPAPS and SAPS seraDomain I binds aPLCrystal structure reveals domains I and II to be the most accessible domains for aPL binding
McNeeley et al (87)SPR, DMAPS seraDomain I binding in 80% of sera 
Guerin et al (62)Elastase and protease cleavage of β2GPIAPS seraDomain V cleavage: lost aPL binding in 91% of sera; domain I cleavage: 50% reduction in aPL binding in 81% of seraMost persuasive evidence for domain I binding; pathogenic aPL contain clusters of positively charged residues; in domain I there is a surface-exposed pocket of negative charge
Sheng et al (88)LAC activity domain–specific aPLMurine monoclonal aPLDomain I binding of aPL prolongs dRVVT and KCT; domain II binding of aPL prolongs KCT 
Reddel et al (86)Mutagenesis domain I β2GPIAPS and aPL-positive sera, murine monoclonal aPLDomain I binds aPL 

Domain V.

It is now generally accepted that domain V contains the major site of binding to phospholipids, but it is much less clear whether that domain interacts with aPL. The lysine-rich sequence CKNKEKKC (Cys-281–Lys-Asn-Lys-Glu-Lys-Lys–Cys-288) is particularly crucial in binding to phospholipids. β2GPI loses its ability to bind phospholipids or to act as a cofactor for aPL/phospholipid binding when the number of lysines in this region is reduced by mutagenesis (75). Peptides containing the CKNKEKKC sequence inhibit binding of purified aPL in a CL ELISA. The importance of domain V in binding to phospholipids is underlined by the finding that a preparation of β2GPI cleaved between residues Lys-317 and Thr-318 in domain V completely lost its ability to bind negatively charged phospholipids and to act as a cofactor for affinity-purified aPL from “autoimmune” patients (53). It was suggested that this cleavage also reduced the binding of β2GPI to polyclonal human aPL (54), but mutations in the CKNKEKKC sequence did not affect this binding (75). These data gave conflicting messages as to whether domain V is important in the interaction between β2GPI and aPL.

Gharavi et al (76) demonstrated that immunization of mice with purified whole human or bovine β2GPI induced production of aPL. Subsequent work by the same group has implicated the domain V CKNKEKKC sequence in this process. In experiments with mice, aPL were induced by immunization with a synthetic peptide (called GDKV) spanning Gly-274–Cys-288 of domain V in β2GPI. Mice immunized with the peptide and carrier protein complexes produced significantly higher levels of aPL than those mice immunized with either the peptide or carrier protein alone. The GDKV-induced monoclonal aPL were shown to enhance thrombus formation and increase leukocyte adherence to endothelial cells in a mouse model (77). Furthermore, Gharavi et al (78) recently have shown that pathogenic aPL can be generated by immunizing mice with a peptide derived from human cytomegalovirus, which resembles the GDKV peptide.

However, consideration of the crystal structure of β2GPI reveals that the sequence Ser-311–Lys-317 forms a hydrophobic loop acting as a membrane stabilizer for the positively charged CKNKEKKC phospholipid-binding region. This region is so closely apposed to the phospholipid membrane that it is unlikely to be a major epitope for binding to aPL under physiologic conditions. Although previous studies suggested that regions of domain V outside CKNKEKKC might be responsible for binding of aPL (54, 79), this now seems unlikely. Those regions would be sterically hindered from interaction with aPL, because they are not surface-exposed or because they are in close proximity to the phospholipid membrane bound to domain V. In light of the crystal structure, it would seem more likely for the aPL binding site to be contained in another domain.

Domain IV.

Igarashi et al (80) and George et al (81) studied mutant forms of β2GPI in which various domains were deleted and expressed in insect cells. Both groups confirmed that domain V was essential for binding to CL, whereas binding to human or murine monoclonal aPL took place in the absence of domain V. The mutant domains I–IV bound the murine antibody WB-CAL-1 better than either whole β2GPI or mutant domains I–III. Of note, WB-CAL-1 was able to bind to domains I–IV on plain polystyrene plates, so this mutant was thought to have undergone a conformational change following the deletion of domain V, thus exposing an epitope for aPL. It was concluded that domain IV contains a major epitope for binding of these aPL. Interestingly, the authors did not demonstrate binding of WB-CAL-1 to any DM lacking domain I and, conversely, DM peptides containing only domain I or domains I–II were not made or tested. These authors then showed that human aPL could competitively inhibit the binding to β2GPI of a murine monoclonal antibody Cof 21, which is known to target an epitope on domain IV, but not the binding of other murine monoclonal antibodies directed against a different epitope on domains IV or V. They concluded that autoimmune aPL recognize an epitope in domain IV of β2GPI (81). This conclusion might have been more compelling if domains II–V, domains III–V, and domains IV–V had been shown to support aPL binding.

Pursuing this hypothesis, Koike et al (82) produced 4 mutant forms of β2GPI in which point mutations were made in domain IV. The binding of human monoclonal aPL and polyclonal sera from 30 anti-β2GPI antibody–positive APS serum samples to these mutant forms of β2GPI immobilized on microtiter plates was greatly reduced in comparison with that of wild-type β2GPI. However, it has subsequently been shown (83) that these mutant forms of β2GPI do bind aPL in the fluid phase. Therefore, the mutations in domain IV alter the interaction of β2GPI with particular ELISA plates, rather than affecting an epitope for binding of aPL. Thus, as with domain V, experimental evidence currently does not support the idea that domain IV contains the major epitopes for the interaction between β2GPI and aPL.

Domain I–II linker region and domain III.

Using a phage-display library, Blank et al (84) identified 3 peptides that react specifically with the 3 human IgM β2GPI-dependent monoclonal aPL, ILA-1, ILA-3, and H3. These monoclonal aPL were isolated from cells derived from a patient with APS, and had previously been shown to cause endothelial cell activation and induce experimental APS in pregnant BALB/c mice (85). The peptides bore sequence homology to regions of β2GPI in the interlinker of domains I-II, domain IV, or domain III. All 3 peptides specifically inhibited the in vitro and in vivo biologic functions of the corresponding monoclonal aPL. Exposure of endothelial cells to ILA-1, ILA-3, and H3 in the presence of their corresponding peptides inhibited, in vitro, endothelial cell activation. In vivo, the ability of each antibody to induce features of APS in pregnant BALB/c mice was reduced if the corresponding specific peptide was administered after the monoclonal antibodies (84).

When affinity-purified polyclonal β2GPI-dependent aPL isolated from APS patients were examined, different fractions of antibodies were identified that specifically recognize each peptide. Sera from only 10 of 43 patients, however, were found to contain antibodies to 1, 2, or all 3 of the peptides. The amounts of the peptide-specific anti-β2GPI fractions varied between 0.4% and 43% of the total number of anti-β2GPI antibodies identified in each of these 10 patients. The majority of the human anti-β2GPI, therefore, must bind to epitopes other than these peptides.

The crystal structure shows that domains III and IV are both heavily glycosylated in native β2GPI. Insect cells, however, do not glycosylate proteins in the same way as mammalian cells. Recombinant whole β2GPI expressed in insect cells has an estimated molecular mass of 43 kd, as opposed to 50 kd for native β2GPI purified from human serum, and this is presumably due to differences in the oligosaccharide chains attached by glycosylation (80, 81). Therefore, domain IV in β2GPI produced from insect cells may be more accessible for binding to aPL than domain IV in human β2GPI in vivo. This is a caveat in interpreting data on domains III and IV derived from experiments using the insect expression system.

Domain I.

Iverson et al (57) studied binding of DMs to affinity-purified polyclonal human aPL from 11 patients (9 of whom had APS). All 11 aPL bound strongly to DM domains I–IV, with significant, but more variable, binding to other DMs containing domain I. In contrast, the DMs in which domain I had been deleted showed little or no specific binding to aPL. In the fluid phase, binding of aPL from all 11 patients to β2GPI-coated plates was inhibited by DMs containing domain I, but not by DMs that lacked domain I. These results support the presence of an epitope for aPL residing in a surface-exposed region of domain I, which does not need to be revealed by a conformational change.

Iverson et al (57), Igarashi et al (80), and George et al (81) all studied similar DMs produced from the same insect expression system. Each showed that the domains I–IV DM bound aPL better than did any other DM containing domain IV. Binding to the domains I–III DM was demonstrated by Iverson et al (57) and Igarashi et al (80), but not by George et al (81). If domain IV does contain the major epitope for aPL, then the DMs for domains II–V, III–V, and IV–V would be expected to bind aPL to variable degrees. These mutants were either not examined (80, 81) or found to be unable to bind aPL (57). Thus, the data from Iverson et al complements and extends the findings from the other two groups, and clearly implicates domain I in the binding of aPL, particularly since they found that the domains I–II and domain I DMs both bound aPL (57).

Domain-deleted mutants may be subject to misfolding; therefore, failure of an aPL to bind to a DM may occur because an epitope has been hidden, rather than deleted. Single-point mutations of specific nucleotides in the DNA encoding β2GPI may be less likely to alter protein structure. Reddel et al (86) have studied DMs and whole β2GPI with single-point mutations in domain I. They demonstrated that monoclonal murine β2GPI-dependent aPL and polyclonal aPL from 18 of 21 human sera did not bind to domain I DMs. In addition, charge-altering mutations of Gly to Glu at position 40 and Arg to Gly at position 43 in domain I both led to a decreased binding to the majority of sera tested. The charge-conserving mutation Thr to Ala at position 50 did not show a consistent pattern of altered binding to these sera.

Iverson et al (83) subsequently showed that the Arg-to-Gly mutation at position 43 led to a reduction in binding to monoclonal aPL derived from patients with APS. These results were reproduced in the solid and fluid phase regardless of the type of ELISA plate used, implying that this mutation is really altering an epitope bound by these monoclonal aPL. In contrast, reduced binding with domain IV point mutants was only seen when they were coated on certain types of plate.

Serum samples from a large cohort of 106 patients with APS have been tested for binding to β2GPI, by surface plasmon resonance. These samples revealed a significantly greater interaction with immobilized native β2GPI in comparison with a DM for domains II–V that lacked domain I. Eighty percent of the patient samples exhibited a marked preference for whole β2GPI containing domain I, with almost half of the total patient samples showing an almost complete lack of binding to the DM for domains II–V (87).

Guerin et al (62) showed that incubation of soluble β2GPI with staphylococcal V8 protease resulted in a 50% reduction in antibody binding to sera from 81% of the 15 APS patients tested. The main cleavage site for this enzyme was in domain I, while the remaining structure of the molecule was left intact.

The LAC activity of murine monoclonal β2GPI-dependent IgG aPL was found by Sheng et al to be domain specific (88). Monoclonal aPL that bound to domain I caused considerable prolongation of the dilute Russell viper venom test (dRVVT) and the kaolin clotting time, whereas aPL binding to domain II caused a lesser prolongation, with a trend to affecting only the dRVVT. The aPL that bound to other domains did not have LAC activity.

Overall, evidence that domain I contains the major epitopes is more robust and more convincing than that for the other domains. This evidence is supported by consideration of the crystal structure.

Crystal structure of β2GPI

  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

The crystal structure of β2GPI (26, 27) shows a J-shaped molecule with domain V interacting with phospholipids. The solution structure of β2GPI (28), however, reveals an S-shaped model of human β2GPI in which additional carbohydrate residues missing from the crystal structure are modeled and the angle between domain I and domain II is rotated. The corrections evident in the solution structure do not alter the conclusions described below.

Domain V is seen to be intimately involved in binding anionic phospholipids or certain synthetic surfaces, and domains IV and III are both heavily glycosylated so even surface-exposed residues are shielded by the glycan moieties. Thus, the most easily accessible domains for aPL binding would appear to be domains I and II. A detailed examination of surface-exposed residues in domain I reveals a region that is rich in negatively charged residues. This region contains 2 Asp residues at positions 8 and 9 and 3 Glu residues at positions 23, 26, and 27, which are in close approximation to each other on the surface of the molecule, and only 1 salt bridge is formed from Glu-26 to a Lys at position 19. Such a pocket of negative charge may be of significance because pathogenic aPL are known to contain clusters of positively charged arginine and lysine residues in their antigen-binding regions (89).

Implications for the future

  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

Identification of the major aPL-binding region on β2GPI has important therapeutic implications in patients with APS. Blank et al (84) have shown that the pathogenic effects of monoclonal anti-β2GPI antibodies in mice could be neutralized with synthetic peptides that bound to domains I–II, domain III, and domain IV. To be therapeutically useful, however, this approach requires the ability to identify a broadly cross-reactive peptide capable of binding with aPL in the majority of patients with APS. The peptides manufactured by Blank et al (84) did not meet these criteria.

An alternative approach would be to develop a drug containing the epitope for aPL that would be able to tolerize aPL-specific B cells, leading to their anergy or deletion. A drug that targets and clears B cells producing anti–double-stranded DNA antibodies is currently undergoing clinical trials in patients with systemic lupus erythematosus (90, 91). The evidence reviewed in the present report suggests that a similar approach to the treatment of APS should target B cells that make antibodies which bind domain I of β2GPI, because this is the site of the most promising epitope for therapeutic purposes. The development of a drug to tolerize domain I–specific B cells as an approach to reduce anti-β2GPI antibody titers is already in progress (92, 93).


  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary

The debate as to the nature and sites of the major epitopes on β2GPI is likely to continue. Since aPL are heterogeneous, it is not surprising that different results have been obtained by different groups, and it is likely that different antibodies recognize epitopes on different domains. The balance of the most persuasive evidence from crystallography, clinical studies, and studies of domain-deleted mutants of β2GPI points to the existence of a major noncryptic epitope on domain I. This epitope may represent the best target for a drug designed to inhibit the aPL–β2GPI interaction which would thus ameliorate APS.


  1. Top of page
  2. Introduction
  3. Molecular structure and function of β2GPI
  4. What is the nature of the epitope for binding of aPL?
  5. Which domains of β2GPI contain major epitopes for aPL?
  6. Crystal structure of β2GPI
  7. Implications for the future
  8. Summary
  • 1
    Hughes GR. Thrombosis, abortion, cerebral disease, and the lupus anticoagulant. Br Med J 1983; 287: 10889.
  • 2
    Hughes GR, Harris NN, Gharavi AE. The anticardiolipin syndrome. J Rheumatol 1986; 13: 4869.
  • 3
    Petri M. In: The antiphospholipid syndrome. II: Autoimmune thrombosis. AshersonRA, CerveraR, PietteJ, ShoenfeldY, editors. 2nd ed. Amsterdam: Elsevier Science; 2002. p. 1120.
  • 4
    Cervera R, Piette JC, Font J, Khamashta MA, Shoenfeld Y, Camps MT, et al. Antiphospholipid syndrome: clinical and immunologic manifestations and patterns of disease expression in a cohort of 1,000 patients. Arthritis Rheum 2002; 46: 101927.
  • 5
    Asherson RA, Khamashta MA, Ordi-Ros J, Derksen RH, Machin SJ, Barquinero J, et al. The “primary” antiphospholipid syndrome: major clinical and serological features. Medicine (Baltimore) 1989; 68: 36674.
  • 6
    Morrow WJW, Nelson L, Watts R, Isenberg DA. Autoimmune rheumatic disease. 2nd ed. Oxford: Oxford University Press; 1999.
  • 7
    Shah NM, Khamashta MA, Atsumi T, Hughes GR. Outcome of patients with anticardiolipin antibodies: a 10 year follow-up of 52 patients. Lupus 1998; 7: 36.
  • 8
    Khamashta MA, Cuadrado MJ, Mujic F, Taub NA, Hunt BJ, Hughes GR. The management of thrombosis in the antiphospholipid-antibody syndrome. N Engl J Med 1995; 332: 9937.
  • 9
    Alarcon-Segovia D, Deleze M, Oria CV, Sanchez-Guerrero J, Gomez-Pacheco L, Cabiedes J, et al. Antiphospholipid antibodies and the antiphospholipid syndrome in systemic lupus erythematosus: a prospective analysis of 500 consecutive patients. Medicine (Baltimore) 1989; 68: 35365.
  • 10
    Lynch A, Marlar R, Murphy J, Davila G, Santos M, Rutledge J, et al. Antiphospholipid antibodies in predicting adverse pregnancy outcome: a prospective study. Ann Intern Med 1994; 120: 4705.
  • 11
    Pierangeli SS, Harris EN. Antiphospholipid antibodies in an in vivo thrombosis model in mice. Lupus 1994; 3: 24751.
  • 12
    Blank M, Cohen J, Toder V, Shoenfeld Y. Induction of anti-phospholipid syndrome in naive mice with mouse lupus monoclonal and human polyclonal anti-cardiolipin antibodies. Proc Natl Acad Sci U S A 1991; 88: 306973.
  • 13
    Ikematsu W, Luan FL, La Rosa L, Beltrami B, Nicoletti F, Buyon JP, et al. Human anticardiolipin monoclonal autoantibodies cause placental necrosis and fetal loss in BALB/c mice. Arthritis Rheum 1998; 41: 102639.
  • 14
    Galli M, Comfurius P, Maassen C, Hemker HC, de Baets MH, van Breda-Vriesman PJ, et al. Anticardiolipin antibodies (ACA) directed not to cardiolipin but to a plasma protein cofactor. Lancet 1990; 335: 15447.
  • 15
    McNeil HP, Simpson RJ, Chesterman CN, Krilis SA. Anti-phospholipid antibodies are directed against a complex antigen that includes a lipid-binding inhibitor of coagulation: beta 2-glycoprotein I (apolipoprotein H). Proc Natl Acad Sci U S A 1990; 87: 41204.
  • 16
    Matsuura E, Igarashi Y, Fujimoto M, Ichikawa K, Koike T. Anticardiolipin cofactor(s) and differential diagnosis of autoimmune disease. Lancet 1990; 336: 1778.
  • 17
    Ordi J, Selva A, Monegal F, Porcel JM, Martinez-Costa X, Vilardell M. Anticardiolipin antibodies and dependence of a serum cofactor: a mechanism of thrombosis. J Rheumatol 1993; 20: 13214.
  • 18
    Tsutsumi A, Matsuura E, Ichikawa K, Fujisaku A, Mukai M, Kobayashi S, et al. Antibodies to β2-glycoprotein I and clinical manifestations in patients with systemic lupus erythematosus. Arthritis Rheum 1996; 39: 146674.
  • 19
    McNally T, Mackie IJ, Machin SJ, Isenberg DA. Increased levels of beta 2 glycoprotein-I antigen and beta 2 glycoprotein-I binding antibodies are associated with a history of thromboembolic complications in patients with SLE and primary antiphospholipid syndrome. Br J Rheumatol 1995; 34: 10316.
  • 20
    Kandiah DA, Sali A, Sheng Y, Victoria EJ, Marquis DM, Coutts SM, et al. Current insights into the “antiphospholipid” syndrome: clinical, immunological, and molecular aspects. Adv Immunol 1998; 70: 50763.
  • 21
    Loizou S, Mackworth-Young CG, Cofiner C, Walport MJ. Heterogeneity of binding reactivity to different phospholipids of antibodies from patients with systemic lupus erythematosus (SLE) and with syphilis. Clin Exp Immunol 1990; 80: 1716.
  • 22
    Di Simone N, Meroni PL, de Papa N, Raschi E, Caliandro D, De Carolis CS, et al. Antiphospholipid antibodies affect trophoblast gonadotropin secretion and invasiveness by binding directly and through adhered β2-glycoprotein I. Arthritis Rheum 2000; 43: 14050.
  • 23
    Shoenfeld Y, Gharavi A, Koike T. β2GP-I in the anti phospholipid (Hughes') syndrome—from a cofactor to an autoantigen—from induction to prevention of antiphospholipid syndrome. Lupus 1998; 7: 5036.
  • 24
    Koike T, Ichikawa K, Kasahara H, Atsumi T, Tsutsumi A, Matsuura E. Epitopes on β2-GPI recognized by anticardiolipin antibodies. Lupus 1998; 7: S147.
  • 25
    Roubey RA, Eisenberg RA, Harper MF, Winfield JB. “Anticardiolipin” autoantibodies recognize beta 2-glycoprotein I in the absence of phospholipid: importance of Ag density and bivalent binding. J Immunol 1995; 154: 95460.
  • 26
    Bouma B, de Groot PG, van den Elsen JM, Ravelli RB, Schouten A, Simmelink MJ, et al. Adhesion mechanism of human beta(2)-glycoprotein I to phospholipids based on its crystal structure. EMBO J 1999; 18: 516674.
  • 27
    Schwarzenbacher R, Zeth K, Diederichs K, Gries A, Kostner GM, Laggner P, et al. Crystal structure of human beta2-glycoprotein I: implications for phospholipid binding and the antiphospholipid syndrome. EMBO J 1999; 18: 622839.
  • 28
    Hammel M, Kriechbaum M, Gries A, Kostner GM, Laggner P, Prassl R. Solution structure of human and bovine beta(2)-glycoprotein I revealed by small-angle X-ray scattering. J Mol Biol 2002; 321: 8597.
  • 29
    Inanc M, Radway-Bright EL, Isenberg DA. β2-glycoprotein I and anti-β2-glycoprotein I antibodies: where are we now? Br J Rheumatol 1997; 36: 124757.
  • 30
    Schultze HE, Heide K, Haupt H. Über ein bisher unbekanntes niedermolekulares Beta2-Globulin des Humanserums. Naturwissenschaften 1961; 48: 719.
  • 31
    Lozier J, Takahashi N, Putnam FW. Complete amino acid sequence of human plasma beta 2-glycoprotein I. Proc Natl Acad Sci U S A 1984; 81: 36404.
  • 32
    Steinkasserer A, Estaller C, Weiss EH, Sim RB, Day AJ. Complete nucleotide and deduced amino acid sequence of human beta 2- glycoprotein I. Biochem J 1991; 277: 38791.
  • 33
    Schousboe I. β2-glycoprotein I: a plasma inhibitor of the contact activation of the intrinsic blood coagulation pathway. Blood 1985; 66: 108691.
  • 34
    Nimpf J, Bevers EM, Bomans PH, Till U, Wurm H, Kostner GM, et al. Prothrombinase activity of human platelets is inhibited by beta 2-glycoprotein-I. Biochim Biophys Acta 1986; 884: 1429.
  • 35
    Nimpf J, Wurm H, Kostner GM. Beta 2-glycoprotein-I (apo-H) inhibits the release reaction of human platelets during ADP-induced aggregation. Atherosclerosis 1987; 63: 10914.
  • 36
    Hasunuma Y, Matsuura E, Makita Z, Katahira T, Nishi S, Koike T. Involvement of beta 2-glycoprotein I and anticardiolipin antibodies in oxidatively modified low-density lipoprotein uptake by macrophages. Clin Exp Immunol 1997; 107: 56973.
  • 37
    Pittoni V, Ravirajan CT, Donohoe S, Maclin SJ, Lydyard PM, Isenberg DA. Human monoclonal anti-phospholipid antibodies selectively bind to membrane phospholipid and beta2-glycoprotein I (beta2-GPI) on apoptotic cells. Clin Exp Immunol 2000; 119: 53343.
  • 38
    Chonn A, Semple SC, Cullis PR. Beta 2 glycoprotein I is a major protein associated with very rapidly cleared liposomes in vivo, suggesting a significant role in the immune clearance of “non-self” particles. J Biol Chem 1995; 270: 258459.
  • 39
    Roubey RA. Immunology of the antiphospholipid antibody syndrome. Arthritis Rheum 1996; 39: 144454.
  • 40
    Del Papa N, Guidali L, Sala A, Buccellati C, Khamashta MA, Ichikawa K, et al. Endothelial cells as target for antiphospholipid antibodies: human polyclonal and monoclonal anti–β2-glycoprotein I antibodies react in vitro with endothelial cells through adherent β2-glycoprotein I and induce endothelial activation. Arthritis Rheum 1997; 40: 55161.
  • 41
    Meroni PL, Del Papa N, Raschi E, Panzeri P, Borghi MO, Tincani A, et al. Beta2-glycoprotein I as a ‘cofactor’ for anti-phospholipid reactivity with endothelial cells. Lupus 1998; 7: S447.
  • 42
    Roman MJ, Salmon JE, Sobel R, Lockshin MD, Sammaritano L, Schwartz JE, et al. Prevalence and relation to risk factors of carotid atherosclerosis and left ventricular hypertrophy in systemic lupus erythematosus and antiphospholipid antibody syndrome. Am J Cardiol 2001; 87: 6636.
  • 43
    Ames PR, Margarita A, Delgado Alves J, Tommasino C, Iannaccone L, Brancaccio V. Anticardiolipin antibody titre and plasma homocysteine level independently predict intima media thickness of carotid arteries in subjects with idiopathic antiphospholipid antibodies. Lupus 2002; 11: 20814.
  • 44
    Petri M. 1998 update on antiphospholipid antibodies. Curr Opin Rheumatol 1998; 10: 42630.
  • 45
    Simantov R, LaSala JM, Lo SK, Gharavi AE, Sammaritano LR, Salmon JE, et al. Activation of cultured vascular endothelial cells by antiphospholipid antibodies. J Clin Invest 1995; 96: 22119.
  • 46
    Pierangeli SS, Colden-Stanfield M, Liu X, Barker JH, Anderson GL, Harris EN. Antiphospholipid antibodies from antiphospholipid syndrome patients activate endothelial cells in vitro and in vivo. Circulation 1999; 99: 19972002.
  • 47
    Delgado Alves J, Ames PR, Donohue S, Stanyer L, Nourooz-Zadeh J, Ravirajan C, et al. Antibodies to high-density lipoprotein and β2-glycoprotein I are inversely correlated with paraoxonase activity in systemic lupus erythematosus and primary antiphospholipid syndrome. Arthritis Rheum 2002; 46: 268694.
  • 48
    Matsuura E, Igarashi Y, Yasuda T, Triplett DA, Koike T. Anticardiolipin antibodies recognize beta 2-glycoprotein I structure altered by interacting with an oxygen modified solid phase surface. J Exp Med 1994; 179: 45762.
  • 49
    Chamley LW, Duncalf AM, Konarkowska B, Mitchell MD, Johnson PM. Conformationally altered beta 2-glycoprotein I is the antigen for anti-cardiolipin autoantibodies. Clin Exp Immunol 1999; 115: 5716.
  • 50
    Borchman D, Harris EN, Pierangeli SS, Lamba OP. Interactions and molecular structure of cardiolipin and beta 2-glycoprotein 1 (beta 2-GP1). Clin Exp Immunol 1995; 102: 3738.
  • 51
    Subang R, Levine JS, Janoff AS, Davidson SM, Taraschi TF, Koike T, et al. Phospholipid-bound beta 2-glycoprotein I induces the production of anti-phospholipid antibodies. J Autoimmun 2000; 15: 2132.
  • 52
    Hammel M, Schwarzenbacher R, Gries A, Kostner GM, Laggner P, Prassl R. Mechanism of the interaction of beta(2)-glycoprotein I with negatively charged phospholipid membranes. Biochemistry 2001; 40: 1417381.
  • 53
    Hunt JE, Simpson RJ, Krilis SA. Identification of a region of beta 2-glycoprotein I critical for lipid binding and anti-cardiolipin antibody cofactor activity. Proc Natl Acad Sci U S A 1993; 90: 21415.
  • 54
    Hunt J, Krilis S. The fifth domain of beta 2-glycoprotein I contains a phospholipid binding site (Cys281-Cys288) and a region recognized by anticardiolipin antibodies. J Immunol 1994; 152: 6539.
  • 55
    Matsuura E, Inagaki J, Kasahara H, Yamamoto D, Atsumi T, Kobayashi K, et al. Proteolytic cleavage of beta(2)-glycoprotein I: reduction of antigenicity and the structural relationship. Int Immunol 2000; 12: 118392.
  • 56
    Reddel SW, Wang YX, Krilis SA. Anti-beta2-glycoprotein I autoantibodies require an antigen density threshold, consistent with divalent binding. Lupus 2003; 12: 3745.
  • 57
    Iverson GM, Victoria EJ, Marquis DM. Anti-beta2 glycoprotein I (beta2GPI) autoantibodies recognize an epitope on the first domain of beta2GPI. Proc Natl Acad Sci U S A 1998; 95: 155426.
  • 58
    Sheng Y, Kandiah DA, Krilis SA. Anti-beta 2-glycoprotein I autoantibodies from patients with the “antiphospholipid” syndrome bind to beta 2-glycoprotein I with low affinity: dimerization of beta 2-glycoprotein I induces a significant increase in anti-beta 2-glycoprotein I antibody affinity. J Immunol 1998; 161: 203843.
  • 59
    Regnault V, Arvieux J, Vallar L, Lecompte T. Both kinetic data and epitope mapping provide clues for understanding the anti-coagulant effect of five murine monoclonal antibodies to human beta2-glycoprotein I. Immunology 1999; 97: 4007.
  • 60
    Keeling DM, Wilson AJ, Mackie IJ, Machin SJ, Isenberg DA. Some ‘antiphospholipid antibodies’ bind to beta 2-glycoprotein I in the absence of phospholipid. Br J Haematol 1992; 82: 5714.
  • 61
    Martinuzzo ME, Forastiero RR, Carreras LO. Anti beta 2 glycoprotein I antibodies: detection and association with thrombosis. Br J Haematol 1995; 89: 397402.
  • 62
    Guerin J, Sim R, Yu BB, Ferluga J, Feighery C, Jackson J. Heterogeneous recognition of beta 2-glycoprotein I by antibodies from antiphospholipid syndrome patients. Thromb Haemost 2000; 84: 37480.
  • 63
    Balestrieri G, Tincani A, Spatola L, Allegri F, Prati E, Cattaneo R, et al. Anti-beta 2-glycoprotein I antibodies: a marker of antiphospholipid syndrome? Lupus 1995; 4: 12230.
  • 64
    Viard JP, Amoura Z, Bach JF. Association of anti-beta 2 glycoprotein I antibodies with lupus-type circulating anticoagulant and thrombosis in systemic lupus erythematosus. Am J Med 1992; 93: 1816.
  • 65
    Arvieux J, Roussel B, Jacob MC, Colomb MG. Measurement of anti-phospholipid antibodies by ELISA using beta 2-glycoprotein I as an antigen. J Immunol Methods 1991; 143: 2239.
  • 66
    Forastiero RR, Martinuzzo ME, Carreras LO. Binding properties of antibodies to prothrombin and beta2-glycoprotein I (beta2-GPI) assayed by ELISA and dot blot. Clin Exp Immunol 1999; 118: 4806.
  • 67
    Tincani A, Spatola L, Prati E, Allegri F, Ferremi P, Cattaneo R, et al. The anti-beta2-glycoprotein I activity in human anti-phospholipid syndrome sera is due to monoreactive low-affinity autoantibodies directed to epitopes located on native beta2-glycoprotein I and preserved during species' evolution. J Immunol 1996; 157: 57328.
  • 68
    Cabiedes J, Cabral AR, Alarcon-Segovia D. Clinical manifestations of the antiphospholipid syndrome in patients with systemic lupus erythematosus associate more strongly with anti-beta 2-glycoprotein-I than with antiphospholipid antibodies. J Rheumatol 1995; 22: 1899906.
  • 69
    Sorice M, Circella A, Griggi T, Garofalo T, Nicodemo G, Pittoni V, et al. Anticardiolipin and anti-beta 2-GPI are two distinct populations of autoantibodies. Thromb Haemost 1996; 75: 3038.
  • 70
    Pierangeli SS, Harris EN, Davis SA, DeLorenzo G. Beta 2-glycoprotein 1 (beta 2GP1) enhances cardiolipin binding activity but is not the antigen for antiphospholipid antibodies. Br J Haematol 1992; 82: 56570.
  • 71
    Permpikul P, Rao LV, Rapaport SI. Functional and binding studies of the roles of prothrombin and beta 2-glycoprotein I in the expression of lupus anticoagulant activity. Blood 1994; 83: 287892.
  • 72
    Harris EN. Syndrome of the black swan. Br J Rheumatol 1987; 26: 3246.
  • 73
    Alarcon-Segovia D, Cabral AR. Antiphospholipid antibodies. Where do they come from? Where do they go?. J Rheumatol 1994; 21: 9829.
  • 74
    Wilson WA, Gharavi AE, Koike T, Lockshin MD, Branch DW, Piette JC, et al. International consensus statement on preliminary classification criteria for definite antiphospholipid syndrome: report of an international workshop. Arthritis Rheum 1999; 42: 130911.
  • 75
    Sheng Y, Sali A, Herzog H, Lahnstein J, Krilis SA. Site-directed mutagenesis of recombinant human beta 2-glycoprotein I identifies a cluster of lysine residues that are critical for phospholipid binding and anti-cardiolipin antibody activity. J Immunol 1996; 157: 374451.
  • 76
    Gharavi AE, Sammaritano LR, Wen J, Elkon KB. Induction of antiphospholipid autoantibodies by immunization with beta 2 glycoprotein I (apolipoprotein H). J Clin Invest 1992; 90: 11059.
  • 77
    Gharavi AE, Pierangeli SS, Colden-Stanfield M, Liu XW, Espinola RG, Harris EN. GDKV-induced antiphospholipid antibodies enhance thrombosis and activate endothelial cells in vivo and in vitro. J Immunol 1999; 163: 29227.
  • 78
    Gharavi AE, Pierangeli SS, Espinola RG, Liu X, Colden-Stanfield M, Harris EN. Antiphospholipid antibodies induced in mice by immunization with a cytomegalovirus-derived peptide cause thrombosis and activation of endothelial cells in vivo. Arthritis Rheum 2002; 46: 54552.
  • 79
    Wang MX, Kandiah DA, Ichikawa K, Khamashta M, Hughes G, Koike T, et al. Epitope specificity of monoclonal anti-beta 2-glycoprotein I antibodies derived from patients with the antiphospholipid syndrome. J Immunol 1995; 155: 162936.
  • 80
    Igarashi M, Matsuura E, Igarashi Y, Nagae H, Ichikawa K, Triplett DA, et al. Human beta2-glycoprotein I as an anticardiolipin cofactor determined using mutants expressed by a baculovirus system. Blood 1996; 87: 326270.
  • 81
    George J, Gilburd B, Hojnik M, Levy Y, Langevitz P, Matsuura E, et al. Target recognition of beta2-glycoprotein I (beta2GPI)-dependent anticardiolipin antibodies: evidence for involvement of the fourth domain of beta2GPI in antibody binding. J Immunol 1998; 160: 391723.
  • 82
    Koike T, Ichikawa K, Atsumi T, Kasahara H, Matsuura E. β2-glycoprotein I–anti-β2-glycoprotein I interaction. J Autoimmun 2000; 15: 97100.
  • 83
    Iverson GM, Matsuura E, Victoria EJ, Cockerill KA, Linnik MD. The orientation of beta2GPI on the plate is important for the binding of anti-beta2GPI autoantibodies by ELISA. J Autoimmun 2002; 18: 28997.
  • 84
    Blank M, Shoenfeld Y, Cabilly S, Heldman Y, Fridkin M, Katchalski-Katzir E. Prevention of experimental antiphospholipid syndrome and endothelial cell activation by synthetic peptides. Proc Natl Acad Sci U S A 1999; 96: 51648.
  • 85
    George J, Blank M, Levy Y, Meroni P, Damianovich M, Tincani A, et al. Differential effects of anti-beta2-glycoprotein I antibodies on endothelial cells and on the manifestations of experimental antiphospholipid syndrome. Circulation 1998; 97: 9006.
  • 86
    Reddel SW, Wang YX, Sheng YH, Krilis SA. Epitope studies with anti-beta 2-glycoprotein I antibodies from autoantibody and immunized sources. J Autoimmun 2000; 15: 916.
  • 87
    McNeeley PA, Dlott JS, Furie RA, Jack RM, Ortel TL, Triplett DA, et al. Beta2-glycoprotein I-dependent anticardiolipin antibodies preferentially bind the amino terminal domain of beta2-glycoprotein I. Thromb Haemost 2001; 86: 5905.
  • 88
    Sheng Y, Hanly JG, Reddel SW, Kouts S, Guerin J, Koike T, et al. Detection of ‘antiphospholipid’ antibodies: a single chromogenic assay of thrombin generation sensitively detects lupus anticoagulants, anticardiolipin antibodies, plus antibodies binding beta(2)- glycoprotein I and prothrombin. Clin Exp Immunol 2001; 124: 5028.
  • 89
    Giles IP, Haley JD, Nagl S, Isenberg DA, Latchman DS, Rahman A. A systematic analysis of sequences of human antiphospholipid and anti-beta2-glycoprotein I antibodies: the importance of somatic mutations and certain sequence motifs. Semin Arthritis Rheum 2003; 32: 24665.
  • 90
    Furie RA, Cash JM, Cronin ME, Katz RS, Weisman MH, Aranow C, et al. Treatment of systemic lupus erythematosus with LJP 394. J Rheumatol 2001; 28: 25765.
  • 91
    Alarcon-Segovia D, Tumlin JA, Furie RA, McKay JD, Cardiel MH, Strand V, et al. LJP 394 for the prevention of renal flare in patients with systemic lupus erythematosus: results from a randomized, double-blind, placebo-controlled study. Arthritis Rheum 2003; 48: 44254.
  • 92
    Iverson GM, Jones DS, Marquis D, Linnik MD, Victoria EJ. A chemically defined, toleragen-based approach for targeting anti-beta2-glycoprotein I antibodies. Lupus 1998; 7: S1669.
  • 93
    Jones DS, Cockerill KA, Gamino CA, Hammaker JR, Hayag MS, Iverson GM, et al. Synthesis of LJP 993, a multivalent conjugate of the N-terminal domain of beta2GPI and suppression of an anti-beta2GPI immune response. Bioconjug Chem 2001; 12: 101220.