Pathogenesis
Anti-β2GPI Domain I Antibodies
Antiphospholipid antibodies do not provide merely serum APS biomarkers, but rather exert a direct pathogenic role in both vascular and obstetric events. Antibodies reacting against β2GPI, in particular those targeting the immuno-dominant epitope in the N-terminal domain I of the molecule, have been identified as the aPL subset mediating aetio-pathogenic mechanisms. A direct demonstration of the pathogenic effect of anti-domain I antibodies has been recently obtained using anti-beta 2 glycoprotein I minibody (MBB2) – a human monoclonal IgG antibody targeting β2GPI–domain I. Its infusion induced fetal losses in pregnant mice and blood clots in rat mesenteric microcirculation after lipopolysaccharide (LPS) priming. A parent monoclonal antibody – MBB2[DELTA]CH2 – displays the same antigen specificity of MBB2, but lacks the CH2 domain, being unable to activate complement. This might have relevant therapeutic implications: MBB2[DELTA]CH2 has been shown to prevent aPL pro-coagulant and pro-abortive effects in vivo by competing with autoantibodies for binding to β2GPI. MBB2[DELTA]CH2 might be exploited as a therapeutic tool even in obstetric APS as it does not affect the expression of placental β2GPI, which exerts a physiological role in embryo implantation and placental morphogenesis. Another pioneer therapeutic option exploiting the pathogenicity of anti-domain I antibodies envisages the use of tolerogenic dendritic cells (tDCs) pulsed with both the whole β2GPI molecule and β2GPI–domain I. Infusion of tDCs to β2GPI-immunized BALB/c mice resulted in an efficient reduction of the fetal loss rate by induction of T-regulatory cells, decreased anti-β2GPI antibody titres and raised expression of anti-inflammatory cytokines. Interestingly, a better beneficial effect was obtained with dendritic cells pulsed with domain I than with the whole molecule.
In line with the hypothesis of anti-domain I antibodies as the main aPL pathogenic subset, patients at higher clinical risk, namely those with triple aPL-positivity, have been shown to display higher anti-β2GPI–domain I antibody titres compared with double/single aPL-positive patients. Consistently, almost all anti-β2GPI–domain I IgG positivities were confirmed after 12 weeks, at variance of innocent aPL transiently occurring as immune epiphenomenon during infectious diseases. It is thus not surprising if the most recent studies have further confirmed the role of anti-domain I antibody positivity as a strong predictor of thrombotic events and, to a lesser extent, of pregnancy complications. Noteworthy, up to one-third of patients carrying anti-β2GPI antibodies were negative for anti-domain I IgG; the observation that a consistent rate of APS patients display autoantibodies reacting with β2GPI epitopes other than domain I was confirmed in a multi-centre study using several peptides spanning the different domains of the molecule.
Therefore, the replacement of the test using the whole molecule with anti-domain I assay could not yet be advocated because of possible misdiagnosis; on the contrary, anti-domain I antibodies provide a promising second-line test for a better risk-stratification of clinical events.
Cell Receptors for β2GPI/Anti-β2GPI Antibodies
In-vitro studies focusing on the interaction of circulating aPLs with target cells convincingly identified AnnexinA2 and Toll-like receptor (TLR)4 as receptors for whole β2GPI. Novel insights allow further characterization of β2GPI interaction with TLR4: β2GPI was found to bind to cells expressing TLR4 as the only membrane receptor, being then internalized into the cytoplasm. In addition, TLR4 emerged as the key player in driving endothelial perturbation: tlr4 but not annexinA2-silencing prevented the up-regulation of adhesion molecules. Conversely, β2GPI binding to endothelial cells was reduced in both conditions – a paradox explained by the fact that AnnexinA2 is not a trans-membrane protein. To note, co-silencing for both receptors did not elicit a complete inhibition of anti-β2GPI antibody binding, a finding that fits well with the possibility that additional surface molecules may be involved in β2GPI binding. Such hypothesis has been explored in few studies: TLR1, TLR2 and TLR6 have indeed been appointed as candidate β2GPI co-receptors in addition to the ones previously reported. The role of TLR2 and TLR4 has been confirmed in a recent ex-vivo study: an increased mRNA expression of these innate immunity receptors and a markedly raised phosphorylation level of interleukin-1 receptor-associated kinase 1 (IRAK-1) – a major mediator in TLR transduction pathway – were observed in peripheral blood mononuclear cells (PBMCs) from APS patients. Consistently, TLR1, TLR2 and TLR6 have been shown to co-localize with aPL IgG; in addition, antibodies blocking TLR1, TLR2 and TLR6 decreased aPL-mediated up-regulation of tumour necrosis factor (TNF) and tissue factor (TF) in human monocytes. The lack of evidence in support of the TLR4 role emerged in this study might be ascribable to a cell-specific orchestra of receptors, co-receptors and accessory molecules deputed to aPL binding. Since β2GPI specifically binds to LPS via its domain V, it has been proposed that such a binding might account for TLR4 engagement. To rule out this hypothesis, in-vitro experiments were conducted in the presence of increasing amounts of exogenous LPS. Only high endothelial cell-activating LPS concentrations affected β2GPI–TLR4 interaction. This in-vitro evidence helps in understanding why LPS is required both to enhance β2GPI distribution in vascular tissues and to induce clotting in animal models. It can be speculated that the amount of β2GPI on the resting endothelium is not enough to allow sufficient aPL binding to trigger clotting, whereas LPS may increase its vascular distribution up-regulating TLR2 and, to a greater extent, TLR4.
Since the main source of LPS in healthy individuals is the gut, it might be envisaged that the resident microbiota might affect LPS uptake. Gut microbiota is indeed increasingly recognized as a major player in the development of autoimmunity; commensal bacteria might contribute to APS pathogenesis inducing autoreactive CD4 T cells and anti-β2GPI antibody production, via molecular mimicry mechanisms or favouring conformational changes in β2GPI, ultimately inducing autoantibody production. Consistent evidence comes from in-vivo models: depletion of gut microbiome with broad-spectrum antibiotics in APS-prone animals markedly prevented thrombotic events, increased survival and reduced anti-β2GPI IgG titres.
Intracellular Signalling Mediators Engaged by Anti-β2GPI Antibodies
Upon engagement of β2GPI receptors on target cells, anti-β2GPI antibody binding results in the activation of intracellular mediators such as nuclear factor kappa B (NF-κB) and p38 mitogen-activated protein kinase (p38MAPK). The cellular steps leading to NF-κB translocation in monocytes have been recently clarified: aPLs engaged NF-κB via a clathrin-dependent endocytic pathway, a mechanism that requires CD14 (TLR4 co-receptor) and AnnexinA2. NF-κB might also provide a potential future pharmacological target: in BALB/c mice, treatment with a specific NF-κB inhibitor, Dehydroxymethylepoxyquinomicin, prevented aPL-mediated thrombus formation. The above discussed intracellular mediators recruited by anti-β2GPI antibodies are schematically represented in Fig. 1.
(Enlarge Image)
Figure 1.
A schematic overview of the most recent insights into the intracellular pathways engaged by anti-β2GPI antibodies in target cells. β2GPI, β2 glycoprotein I; anti-β2GPI, anti-β2GPI antibodies; DI, domain I; IRAK-1, interleukin-1 receptor-associated kinase 1; mTOR, mammalian target of rapamycin; MyD88, myeloid differentiation primary response 88; NF-κB, nuclear factor kappa B; p38MAPK, p38 mitogen-activated protein kinase; PI3K, phosphatidylinositol 3-kinase; S6RP, S6 ribosomal protein; TF, tissue factor; TLR, Toll-like receptor; TNF, tumour necrosis factor
The heterogeneity in the clinical spectrum of APS, with some thrombotic APS women experiencing uneventful pregnancies and obstetric APS patients never developing vascular events, suggests that IgG from thrombotic versus obstetric patients might elicit different biological effects. In-vitro support for this intriguing theory has been raised using monocytes and, most recently, first-trimester trophoblast cells. IgG from neither obstetric nor thrombotic APS patients altered the mRNA levels of myeloid differentiation primary response 88 (MyD88), the expression of interleukin (IL)-6 and IL-8 or the phosphorylation rate of p38MAPK, NF-κB and extracellular signal-regulated kinase. Nevertheless, aPLs obtained from women with pure pregnancy manifestations, at variance of those purified from thrombotic patients, significantly reduced trophoblast invasion, an effect that appeared to be mediated by TLR4.
The phosphatidylinositol 3-kinase (PI3K)–AKT pathway has been recently identified as an additional signalling cascade engaged by aPLs. It culminates in the recruitment of mammalian target of rapamycin (mTOR), a kinase modulating cellular growth, proliferation and apoptosis. In human microvascular endothelial cells, stimulation with aPL IgG resulted in PI3K-mediated activation of two components of the mTOR pathway: S6 ribosomal protein (S6RP) and AKT. These findings were confirmed at immunostaining: the phosphorylation rate of both S6RP and AKT was markedly increased in vascular endothelial cells in renal specimens from APS patients with aPL nephropathy and in the carotid and left anterior descending arteries of catastrophic APS [catastrophic antiphospholipid syndrome (CAPS)] patients. mTOR pathway might therefore mediate the intimal hyperplasia leading to chronic vasculopathy characteristic of both conditions. Given the in-vitro effectiveness of two mTOR inhibitors, PP242 and sirolimus, in preventing the recruitment of S6RP and AKT, sirolimus was prescribed to patients with APS nephropathy undergoing kidney transplantation who developed no vascular lesion recurrence and, compared with patients in the standard regimen arm, displayed a higher rate of functioning allograft at 144 months and a decreased vascular proliferation on biopsy. A potential caveat of such approach is provided by the pro-thrombogenicity of mTOR inhibitors emerged in some studies; the therapeutic relevance of sirolimus warrants confirmation in APS patients with the commonest clinical manifestations.
The Role of Complement in Antiphospholipid Syndrome Pathogenesis
Apart from anti-β2GPI antibodies, a key role in APS pathogenesis is played by complement. The involvement of the complement cascade is confirmed by the failure of the monoclonal anti-domain I MBB2 to induce vascular thrombosis in C6-deficient rats and fetal loss in C5-depleted mice. Indirect evidence comes also from the in-vivo effectiveness of pharmacological tools acting on C5: the non-complement-fixing anti-domain I monoclonal MBB2[DELTA]CH2 was shown to inhibit in-vivo effects induced by aPL infusion, the complement C5-inhibitor rEV576 covers in inhibited aPL-mediated venous thrombosis and TF production in a preclinical mouse model.