Antiphospholipid Antibody Syndrome. Rare Diseases of the Immune System

3. Antiphospholipid Antibody Mechanisms of Thrombosis

Pier Luigi Meroni1, 2, 3  , Chiara Crotti1, 2   and Cecilia Chighizola1, 2  


Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy


Istituto G. Pini and IRCCS Istituto Auxologico Italiano, Milan, Italy


Division of Rheumatology, Istituto G. Pini, Pzza. C. Ferrari, 1, Milan, 20122, Italy

Pier Luigi Meroni (Corresponding author)


Chiara Crotti


Cecilia Chighizola



Beta-2 glycoprotein IPathogenesisEndothelial cellMonocytesPlatelets

3.1 Introduction

Arterial and/or venous thrombosis in association with the persistent presence of antiphospholipid antibodies (aPL) is the hallmark of the antiphospholipid antibody syndrome (APS).

aPL can be detected by solid-phase assays (anticardiolipin [aCL] and/or anti-β2 glycoprotein I [β2GPI]) and/or by a functional assay (lupus anticoagulant [LA]) [1]. Besides their diagnostic value, aPL are also pathogenic autoantibodies that mediate the vascular manifestations in association with an additional second hit (two hits theory) [2].

3.2 Pathogenic aPL

aPL are mainly directed against two PL-binding proteins: β2GPI and prothrombin (PT). β2GPI-dependent aPL can be detected by both aCL and anti-β2GPI assay and are responsible for most of the LA positivities. There is a general agreement that β2GPI-dependent aPL are more predictive for the vascular (and the obstetric) manifestations of the syndrome [23].

aPL reacting with human β2GPI and cross-reacting with the animal molecule were shown to be pathogenic in all the in vivo experimental models. More importantly, the thrombogenic effect was reproduced by affinity-purified anti-β2GPI IgG and inhibited by specific absorption of the anti-β2GPI activity [2]. Accordingly, β2GPI-dependent aPL are generally thought to represent the main pathogenic antibody subpopulation for the APS thrombotic manifestations.

β2GPI is a large molecule belonging to the complement control protein (CCP) family and consists of 326 amino acid residues arranged in 5 CCP repeat domains, termed “sushi” domains. Hence, it is not surprising that autoantibodies can recognize several epitopes in the molecule. However, there is growing evidence that a cryptic epitope (Gly40-Arg43) of domain I (DI) may represent the immunodominant epitope that is recognized by most of the APS sera and closely related to the clinical manifestations of the syndrome [34].

We recently demonstrated that a human monoclonal antibody (moAb) specific for DI is pathogenic in a murine model of vascular thrombosis further supporting the hypothesis that such a subpopulation of anti-β2GPI may play a pathogenic role [5].

aPL are also known to react with other PL-binding proteins and in particular with PT. Although in vitro studies suggested that anti-PT may perturb endothelial cells (EC) by reacting with the molecule on the cell surface, still there is no clear evidence for anti-PT pathogenic activity in animal models and their possible pathogenic role is only supported by the epidemiological association with thrombosis [23]. The best association was reported for antibodies detected by the PS-PT assay, suggesting that the pathogenic antibodies may react with a conformational epitope(s) expressed by PT when complexed with anionic PL in the presence of calcium ions [6].

3.3 Why We Break the Tolerance Against Beta2GPI?

The mechanism(s) through which the tolerance towards β2GPI is lost is still matter of research, and several hypotheses have been suggested.

There is sound evidence that aPL can be detectable in the course of viral, bacterial, and parasitic infections as a transient phenomenon [78]. This is also true for anti-β2GPI antibodies suggesting a possible link between the production of β2GPI-dependent aPL antibodies and infectious processes or even an active immunization [9].

Infectious agents may induce conformational changes in antigens, exposing hidden epitopes and leading to autoantibodies production. Accordingly, it has been suggested that the contact of β2GPI with an infectious agent or with endotoxin may favor the exposure of immunogenic epitopes. Beta2GPI is a plasma protein that may act as a scavenger molecule for lipopolysaccharide (LPS) through its C-terminal part [10]. Circulating plasma β2GPI exists in a circular form that, upon binding to suitable anionic surfaces as CL and other PL or to LPS, opens up to a J-shaped fishhook configuration [11]. Such a conformational change exposes DI that has been suggested to be the main antigenic target for anti-β2GPI antibodies [1213]. In particular, it has been recently demonstrated that the antibodies recognize the sequence of arginine 39-arginine 43, aspartic acid 8-aspartic acid 9, and possibly the interlinker region between DI and DII, with R39 being the most important residue [1415]. This epitope was reported to be a cryptic and conformation-dependent structure. In the β2GPI circular conformation, DI interacts with DV and the critical epitope is thus hidden. Conversely, upon opening to a J-configuration, the critical epitope arginine 39-glycine 43 is exposed, thus becoming available for antibody binding [16]. It has been suggested that upon the binding to LPS (or after the contact with an infectious agent) β2GPI may circulate in the open form, offering the DI immunodominant epitope to the afferent limb of the immune system and eventually triggering antibody production in susceptible individuals [17].

Besides exposing hidden epitopes, infections could also increase the immunogenicity of the molecule β2GPI by changing its redox potential. It has been reported, in fact, that the majority of circulating β2GPI exists in a reduced form and that oxidation increases immunogenicity. Accordingly, the generation of reactive oxidative and nitroxidative species by certain infectious agents may favor the generation of oxidized β2GPI and foster autoantibodies production [18].

Molecular mimicry is another suggested mechanism: nonself molecules (frequently infectious agents) may share similar amino acid sequences with β2GPI and may trigger the production of antibodies against the infectious antigen(s) cross-reacting with the β2GPI itself [19].

3.4 Pathogenesis of the Vascular Manifestations: Introduction

Traditionally, the clotting cascade is depicted as consisting of an intrinsic and extrinsic pathway (Fig. 3.1). The intrinsic pathway is initiated by the exposure of blood to a negatively charged surface. The extrinsic pathway is activated by tissue factor (TF) exposed at the site of injury or by TF-like material. Both pathways converge on the activation of factor X which, as a component of prothrombinase, converts prothrombin to thrombin, the final enzyme of the clotting cascade. Thrombin converts fibrinogen from a soluble plasma protein into an insoluble fibrin clot.


Fig. 3.1

Schematic representation of the coagulation cascade. HK high-molecular-weight kininogen, PK prekallikrein, PL phospholipid, PT prothrombin, TH thrombin coagulation cascade

Fluid-phase coagulation components are involved in propagation of the clotting process, in termination of clotting by antithrombotic control mechanisms, and in the removal of the clot by fibrinolysis. At the same time, endothelium and peripheral blood monocytes play a main role in expressing TF and supporting clot formation, while platelets are pivotal in the initiation and formation of the plug.

aPL procoagulant mechanism(s) has been initially related to the antibody’s binding to β2GPI or PT and the consequent interference with the natural anticoagulant system of protein C and with the reduction of fibrinolysis. However, two main reasons supported the search for additional pathogenic mechanisms:



As a consequence, a lot of investigations have been focused on aPL ability to react with PL-binding proteins expressed on the cell membrane of the cells involved in the coagulation cascade.

In this regard, β2GPI is expressed on the cell membranes at high antigenic density and so more easily recognized by low-avidity autoantibodies. In addition, it has been reported that circulating β2GPI is usually present in the circular-closed form, while it is opened only after binding to negative PL or after complexing with LPS. Since the autoantibodies were reported to react mainly with the opened form (i.e., reacting with the exposed cryptic epitope of DI), this finding may further explain the reason of the limited reactivity with the circulating form.

Most of the pathogenic mechanisms potentially responsible for thrombus formation have been demonstrated using in vitro models. However, three different in vivo models of thrombosis induced respectively by mechanical, chemical, or photochemical trauma have confirmed the pathogenic effect of aPL [2]. Two of these experimental models showed an increase of the thrombus size already triggered respectively by the mechanical or the chemical stimulus and a delayed dissolution. On the other hand, the third model showed that the passive infusion of human aPL IgG together with a small amount of LPS is able by itself to start clotting in the rat mesenteric arterial microcirculation [21].

Table 3.1 and Fig. 3.2 summarize the main pathogenic mechanisms that have been reported in the literature [222].

Table 3.1

Antiphospholipid antibody-mediated pathogenic mechanisms for thrombosis

Interference with fluid-phase coagulation components:

Inhibition of natural anticoagulants

Inhibition of protein C activationa

Disruption of the annexin V shielda

Inhibition of fibrinolysis

Interference with cells of the coagulation cascade:

Endothelial cell perturbationa

Monocyte activation (TF expression)a

Platelet activationa

Complement activationa

aβ2GPI/anti-β2GPI antibody involvement


Fig. 3.2

Pathogenic clotting mechanisms mediated by aPL. aPL actions favor clot formation through several routes. 1 aPL interact with endothelial cells, primarily through binding of β2GPI on the cell surface, and induce a procoagulant and proinflammatory endothelial phenotype. 2 aPL upregulate tissue factor expression on endothelial cells and blood monocytes, and promote endothelial leukocyte adhesion, cytokine secretion, and PGE2synthesis. 3 aPL recognize phospholipid-binding proteins expressed on platelets—aPL binding potentiates platelet aggregation induced by another agonist. 4 aPL interfere with plasma components of the coagulation cascade, by inhibiting anticoagulant activity, by affecting fibrinolysis, and by displacing the binding of the natural anticoagulant annexin A5 to anionic structures. These mechanisms all contribute to a procoagulant state that is necessary but not sufficient for clotting. Clot formation seems to require two steps: the presence of aPL provides the ‘first hit’, which produces clotting when accompanied by another procoagulant condition, a ‘second hit’. Complement activation seems to be necessary for clot formation in vivo. Abbreviations: aPL anti-phospholipid autoantibodies, β2GPI β2 glycoprotein I, PGE2 prostaglandin E2 (Permission from Meroni et al [2])

Still open is the question why some aPL-positive patients develop both arterial and venous events and why others can display arterial or venous thrombosis only.

3.5 Interference of aPL with Fluid-Phase Coagulation Cascade

The evidence of aPL interference with fluid-phase components of coagulation has been provided mostly by in vitro models [222].

aPL can bind some members of the serine protease (SP) family, which enlists proteins involved in hemostasis (procoagulant factors such as thrombin, PT, FVIIa, FIXa, and FXa and natural anticoagulants as protein C) and in fibrinolysis (plasmin and tissue plasminogen activator (tPA)). The fact that β2GPI and SP enzymatic domain share conformational epitopes was suggested to be the reason for the cross-reactivity.

aPL interact with thrombin and FXa, interfering with the formation of thrombin-antithrombin (AT) and FXa-AT complexes, thus affecting the AT inactivation of thrombin and FXa. Moreover, aPL are reported to decrease activated protein C (APC) activity by competing with APC for PL binding. Accordingly, an increased APC resistance has been reported in APS. Several groups have also found decreased levels of both proteins C and S and aPL with affinity for either protein S or C. In addition, some aPL were shown to inhibit plasmin-mediated fibrinolysis, eventually leading to an impairment of fibrin dissolution by plasmin. Some aPL subpopulations were also found to bind to tPA, inhibiting tPA-mediated conversion of plasminogen to plasmin. Anti-tPA antibodies were shown to inversely correlate with plasma tPA activity in APS.

Besides the inhibitory effects on the natural anticoagulant systems, aPL may also increase the enzymatic activity of procoagulants. In fact some aPL subpopulations were shown to induce a gain of function of PT leading to increased fibrin production. In addition, aPL have been reported to disrupt the crystallization of annexin A5 on EC monolayer. The “shield” of annexin A5 is thought to represent a potent anticoagulant barrier that prevents PL bioavailability for coagulation enzymes [222].

3.6 Interference of aPL with Cells Involved in the Coagulation Cascade

More recently, research has focused on aPL interaction with cells involved in the hemostatic balance as platelets, monocytes, and EC [2].

3.6.1 Platelet Involvement

The occurrence of mild thrombocytopenia is a frequent clinical finding among APS patients and strongly suggested the possible interaction between platelets and aPL.

aPL reactivity with platelets has been shown only after their pre-activation by agonists such as thrombin or collagen. The reason for this was thought to be related to the exposure of phosphatidylserine (PS) on the cell outer membrane after platelet activation and the consequent binding of the cationic β2GPI. However, β2GPI can also bind two surface receptors on platelets. Apolipoprotein E receptor 2’ (ApoER2’) is a member of low-density lipoprotein (LDL) receptor family expressed by platelets; its LDL-binding domain I recognizes the PL-binding site of β2GPI domain V. Lastly β2GPI has been shown to bind directly GPIba, a subunit of the GPIb-IX-V platelet receptor [23] and to platelet factor 4 [24].

Altogether these findings suggest that the main antigenic target on platelets is the β2GPI itself which can be recognized by specific antibodies.

There is evidence that such interaction may be responsible for several biological effects. In general, the binding of the antibodies with their own antigenic target can trigger platelet activation. In addition, aPL have been shown to neutralize β2GPI interaction with von Willebrand factor (vWF). β2GPI acts as a biologically relevant inhibitor of vWF function, thus interfering with vWF-dependent platelet adhesion. This action could be hampered by aPL so favoring thrombosis and may explain the consumptive thrombocytopenia frequently observed in aPL carriers. Once bound, aPL have been shown to enhance the expression of platelet membrane glycoproteins (GPs) IIb/IIIa and IIIa. GPs are fibrinogen receptors that mediate platelet aggregation, whose role in APS pathogenesis is further supported by in vivo findings [25]. In fact, aPL did not affect thrombus formation in GPIIb/IIIa-deficient mice and pretreatment with a monoclonal anti-GPIIb/IIIa antibody inhibited aPL-mediated reduced thrombus formation. In vivo evidence of platelet activation by aPL has been gained in other models: (i) aPL produced a platelet-rich thrombus in rats, after treatment with low concentration of adenosine diphosphate (ADP), and (ii) platelet involvement was reported in the thrombus formation by photochemical injury in the rat. Platelet activation by aPL has been supported by ex vivo studies since elevated levels of thromboxane (TX, the major eicosanoid metabolic breakdown product) have been found in the urine of APS patients [26].

3.6.2 Involvement of Monocytes

aPL upregulate TF expression on peripheral blood monocytes. In particular, monocyte TF expression is increased in patients with APS and correlates with the expression of β2GPI. Accordingly, cell-surface TF on monocytes is higher in APS patients than in subjects without thrombotic events. Interestingly, vascular endothelial growth factor (VEGF) stimulates TF expression in monocytes through its tyrosine-kinase receptor Flt-1, and expression of both VEGF and Flt-1 is increased in monocytes from APS patients [27]. On monocytes, aPL have been suggested to interact with β2GPI in association with annexin A2 [28] and toll-like receptor (TLR) 4 within lipid rafts [29]. Other authors provided indirect evidence of TLR2 involvement in mediating aPL-induced monocyte activation [2].

3.6.3 Involvement of Endothelial Cells

Endothelium has been suggested to represent a cell type deeply involved in APS pathogenesis since the earliest pathogenic studies. The milestone was the demonstration that aPL could react with EC mainly recognizing β2GPI present on the cell membrane [3031]. Once bound the β2GPI-dependent aPL can trigger EC perturbation with the induction of a proinflammatory and procoagulant cell phenotype [22226].

The majority of the studies have been performed in vitro showing an upregulation of endothelial cellular adhesion molecules (such as ICAM-1, VCAM-1, and E-selectin) and the synthesis of proinflammatory cytokines (as interleukin (IL)-1β, IL-6, and IL-8). Accordingly, in vivo studies have shown that passive infusion of aPL induced an increase in the rate of leukocytes adhering to the endothelium. Lastly, aPL may modulate the vascular tone by inhibiting endothelial nitric oxide synthase and altering prostaglandin metabolism [226].

Ex vivo studies in APS subjects reported raised levels of soluble cellular adhesion molecules such as s-ICAM-1, VCAM-1, and P-selectin although with contrasting results among the different groups. More importantly, APS patients were found to display endothelial perturbation, as suggested by the impaired brachial artery flow-mediated vasodilation responses and the significant increase in the numbers of circulating ECs and in tPA and vWF titers [32].

aPL reactivity with EC is mainly mediated by autoantibodies to β2GPI expressed on the endothelial cell membrane. Since endothelial synthesis of the molecule has not been confirmed, it has been suggested that β2GPI is adhering to EC membrane. The adhesion is apparently mediated by several different receptors. Annexin A2, a receptor for tPA and plasminogen, has been demonstrated to directly bind β2GPI. As annexin A2 lacks an intracytoplasmic tail, a coreceptor is required to trigger signaling cascade. TLRs 2 and 4, heparan sulfate, and ApoER2’ have all been indirectly shown to bind β2GPI on endothelial surface. TLRs 2 and 4 have been found to mediate also intracellular aPL signaling in ECs; to note, TLR2 is expressed by ECs only upon cell activation, while TLR4 is constitutively expressed [2]. A direct interaction between β2GPI and TLR4 has been recently reported [33]. More recently, it has been suggested that a multiprotein complex including TLR4 and annexin A2 is involved in aPL-induced EC activation [34].

Annexin A2, TLR4, and ApoER2’ have also been investigated in in vivo models. Animals deficient in any one of these molecules are only partially protected against aPL thrombogenic effects, suggesting redundancy in the receptor function and signaling cascade [2].

There is general agreement that nuclear factor κ B (NFκB) and p38 mitogen-activated protein kinase (MAPK) are the main downstream signaling pathways engaged by aPL in EC and monocyte activation [2].

3.7 The “Two-Hit Hypothesis”

To explain the clinical observation that the thrombotic events occur only occasionally in spite of the persistent presence of aPL a two-hit hypothesis has been suggested [2].

The antibody (first hit) induces a thrombophilic condition, but clotting takes place in the presence of another thrombophilic condition (second hit). Accordingly, the administration of small amount of LPS was required for human β2GPI-specific aPL IgG to display their thrombogenic effect in the mesenteric microcirculation [521]. Such animal models fits very well with the clinical observation that infectious processes may frequently precede the full-blown picture of the syndrome and may be the initiator of the catastrophic subtype [35]. To explain the relationship between infections and APS the potential involvement of pattern recognition receptors (i.e., TLRs) in sensing microbes and triggering an inflammatory response was suggested. Since TLRs 2 and 4 have been reported to contribute to EC and/or monocyte activation by β2GPI-dependent aPL, it has been suggested that the infection plus TLR perturbation mediated by the autoantibodies overcomes the threshold for triggering thrombosis [2]. Alternatively, infection/inflammation may increase the expression of the target antigen for aPL or the expression of antigenic epitopes that are hidden in resting conditions as shown in murine tissues after LPS injection [36].

Additional second hits have been reported in two other different models: (i) mechanical trauma on the mouse femoral vein [37] and (ii) a photochemical injury [38].

3.8 Complement Activation

APS animal models showed that complement activation is a necessary requirement since animals deficient in complement components or complement receptors or treated with inhibitors of complement activation were protected from the aPL thrombogenic effects [25].

On the other hand, sera from the majority of APS patients were found to fix complement in vitro.

However, a clear complement consumption has not been described in patients, with only two studies reporting mild hypocomplementemia in primary APS [2]. So, if definite conclusions about the involvement of complement in APS-associated thrombosis cannot currently be drawn, the potential role of complement in aPL-mediated thrombosis should not be neglected.



Miyakis S et al (2006) International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS). J Thromb Haemost 4:295–306PubMedCrossRef


Meroni PL et al (2011) Pathogenesis of antiphospholipid syndrome: understanding the antibodies. Nat Rev Rheumatol 7:330–339PubMedCrossRef


Meroni PL, Chighizola C, Rovelli F, Gerosa M (2014) Antiphospholipid syndrome 2013: more than autoantibodies only. Arthr Res Ther 16:209


Chighizola CB et al (2014) New tests to detect antiphospholipid antibodies: anti-domain I beta-2-glycoprotein-I antibodies. Curr Rheumatol Rep 16:402–407PubMedCrossRef


Agostinis C et al (2014) A non-complement-fixing antibody to β2 glycoprotein I as a novel therapy for antiphospholipid syndrome. Blood 123:3478–3487PubMedCrossRef


Sciascia S et al (2014) New tests to detect antiphospholipid antibodies: antiprothrombin (aPT) and anti-phosphatidylserine/prothrombin (aPS/PT) antibodies. Curr Rheumatol Rep 16:415PubMedCrossRef


S’ene D et al (2009) Antiphospholipid antibodies, antiphospholipid syndrome and viral infections. Rev Med Interne 30:135–141CrossRef


Ben-Chetrit E et al (2013) Antiphospholipid antibodies during infectious mononucleosis and their long term clinical significance. J Clin Virol 56:312–315PubMedCrossRef


Cruz-Tapias P et al (2012) Infections and vaccines in the etiology of antiphospholipid syndrome. Curr Opin Rheumatol 24:389–393PubMedCrossRef


Agar C et al (2011) Beta2-glycoprotein I: a novel component of innate immunity. Blood 117:6939–6947PubMedCrossRef


De Laat B et al (2011) Autoantibodies directed against domain I of beta2-glycoprotein I. Curr Rheumatol Rep 13:70–76PubMedCentralPubMedCrossRef


Iverson GM et al (1998) Anti-β2 glycoprotein I (β2GPI) autoantibodies recognize an epitope on the first domain of β 2GPI. Proc Natl Acad Sci U S A 95:15542–15546PubMedCentralPubMedCrossRef


McNeely PA et al (2001) β 2-glycoprotein I-dependent anticardiolipin antibodies preferentially bind the amino terminal domain of β 2-glycoprotein I. Thromb Haemost 86:590–595


Iverson GM et al (2002) Use of single point mutations in domain I of β2-glycoprotein I to determine fine antibody specificity of antiphospholipid autoantibodies. J Immunol 169:7097–7103PubMedCrossRef


Ioannou Y et al (2007) Binding of antiphospholipid antibodies to discontinuous epitopes on domain I of human beta(2)-glycoprotein I: mutation studies including residues R39 to R43. Arthritis Rheum 56:280–290PubMedCentralPubMedCrossRef


De Laat B et al (2006) Pathogenic anti-beta2 glycoprotein I antibodies recognize domain I of beta-2 glycoprotein I only after a conformational change. Blood 107:1916–1924PubMedCrossRef


De Laat B et al (2011) Immune responses against domain I of β2-glycoprotein I are driven by conformational changes: domain I of β2-glycoprotein I harbors a cryptic immunogenic epitope. Arthritis Rheum 63:3960–3968PubMedCrossRef


Passam FH et al (2011) Molecular pathophysiology of the antiphospholipid syndrome: the role of oxidative post-translational modification of beta 2 glycoprotein I. J Thromb Haemost 9(Suppl 1):275–282PubMedCrossRef


Blank M et al (2002) Bacterial induction of autoantibodies to beta2-glycoprotein-I accounts for the infectious etiology of antiphospholipid syndrome. J Clin Invest 109:797–804PubMedCentralPubMedCrossRef


Tincani A et al (1996) 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 157:5732–5738PubMed


Fischetti F et al (2005) Thrombus formation induced by antibodies to beta2-glycoprotein I is complement dependent and requires a priming factor. Blood 106:2340–2346PubMedCrossRef


Giannakopoulos B et al (2013) The pathogenesis of the antiphospholipid syndrome. N Engl J Med 368:1033–1044PubMedCrossRef


Pennings MT et al (2006) Interaction of beta2-glycoprotein I with members of the low density lipoprotein receptor family. J Thromb Haemost 4:1680–1690PubMedCrossRef


Sikara MP et al (2010) {beta}2 Glycoprotein I ({beta}2GPI) binds platelet factor 4 (PF4): implications for the pathogenesis of antiphospholipid syndrome. Blood 115(3):713–723PubMedCrossRef


Shi T et al (2006) Anti-beta2-glycoprotein I antibodies in complex with beta2-glycoprotein I can activate platelets in a dysregulated manner via glycoprotein Ib-IX-V. Arthritis Rheum 54:2558–2567PubMedCrossRef


Pierangeli SS et al (2008) Antiphospholipid antibodies and the antiphospholipid syndrome: pathogenic mechanisms. Semin Thromb Hemost 34:236–250PubMedCrossRef


Boles J et al (2010) Role of tissue factor in thrombosis in antiphospholipid antibody syndrome. Lupus 19:370–378PubMedCentralPubMedCrossRef


Zhang J et al (2005) Annexin A2 mediates endothelial cell activation by antiphospholipid/anti-beta2 glycoprotein I antibodies. Blood 105:1964–1969PubMedCrossRef


Sorice M et al (2007) Anti-beta2-glycoprotein I antibodies induce monocyte release of tumor necrosis factor alpha and tissue factor by signal transduction pathways involving lipid rafts. Arthritis Rheum 56:2687–2697PubMedCrossRef


Vismara A et al (1988) Relationship between anti-cardiolipin and anti-endothelial cell antibodies in systemic lupus erythematosus. Clin Exp Immunol 74:247–253PubMedCentralPubMed


Del Papa N et al (1995) Relationship between anti-phospholipid and anti-endothelial cell antibodies III: beta 2 glycoprotein I mediates the antibody binding to endothelial membranes and induces the expression of adhesion molecules. Clin Exp Rheumatol 13:179–185PubMed


Cugno M et al (2010) Patients with antiphospholipid syndrome display endothelial perturbation. J Autoimmun 34:105–110PubMedCrossRef


Raschi E et al. (2014) β2-glycoprotein I, lipopolysaccharide and endothelial TLR4: three players in the two hit theory for anti-phospholipid-mediated thrombosis. J Autoimmun 1–9. doi: 10.​1016/​j.​jaut.​2014.​03.​001


Allen KL et al (2012) A novel pathway for human endothelial cell activation by antiphospholipid/anti-β2 glycoprotein I antibodies. Blood 119:884–893PubMedCentralPubMedCrossRef


Shoenfeld Y et al (2006) Infectious origin of the antiphospholipid syndrome. Ann Rheum Dis 65:2–6PubMedCentralPubMedCrossRef


Agostinis C et al (2011) In vivo distribution of β2 glycoprotein I under various pathophysiologic conditions. Blood 118:4231–4238PubMedCrossRef


Pierangeli SS et al (1996) Thrombogenic properties of murine anti-cardiolipin antibodies induced by beta 2 glycoprotein 1 and human immunoglobulin G antiphospholipid antibodies. Circulation 94:1746–1751PubMedCrossRef


Jankowski M et al (2003) Thrombogenicity of beta 2-glycoprotein I-dependent antiphospholipid antibodies in a photochemically induced thrombosis model in the hamster. Blood 101:157–162PubMedCrossRef