There are multiple effectors of fetal injury downstream of complement activation, triggered by C5a–C5a receptor interactions. Tumor necrosis factor (TNF)-α is one mediator that links complement activation and pathogenic Antiphospholipid antibodies to fetal damage. Antiphospholipid antibodies, specifically targeted to decidual tissue, cause a rapid increase in decidual and systemic TNF-α levels. Complement 5-deficient mice were protected from fetal death and showed no increase in TNF-α levels identifying TNF-α as a critical intermediate that acts downstream of C5 activation [13]. That TNF-α is itself pathogenic is suggested by studies showing that miscarriage induced by aPL is less frequent in mice deficient in TNF-α or treated with TNF-α blockade [13]. Furthermore, in antibody-independent models of placental insufficiency and preeclampsia, complement activation at the maternal-fetal interface leads to elevation in local TNF-α levels, reduction of the essential angiogenic factor, vascular endothelial growth factor (VEGF), and, ultimately, abnormal placentation and fetal death. Blockade of complement activation or blockade of TNF-α improves spiral artery remodeling and rescues pregnancies, underscoring the relationship of these mediators and their importance [44].

Complement 5a also triggers fetal damage through induction of tissue factor (TF) expression. Treatment with aPL increases TF in neutrophils which enhances oxidative burst providing a mechanism for trophoblast injury and pregnancy loss triggered by these autoantibodies [19, 45]. Finally, complement activation products may cause an imbalance of angiogenic factors required for normal pregnancy. Satisfactory development of the fetomaternal vasculature is required for successful embryonic growth, and insufficient placental vascularization has been associated with early embryonic mortality, preeclampsia, and IUGR [46]. Normal placental development requires coordinated expression of angiogenic growth factors, and C5a–C5a receptor interactions trigger release of anti-angiogenic factors from leukocytes which can alter the balance of angiogenic factors in pregnancy and lead to the pregnancy complications associated with APS [16].

The placenta is a major target for aPL, in particular β₂GPI-dependent antibodies, which bind to human trophoblast. This may explain why pregnancy complications associated with placental development and function occur in women with APS. Indeed, the expression of β₂GPI on the placenta and, in particular, on trophoblast cell membranes is the prerequisite to explain aPL-placental tropism. While most cells will only bind β ₂ GPI on their cell surface under pathologic, stimulatory, or apoptotic conditions, when the inner negatively charged phospholipids become exposed onto the outer leaflet of the plasma membrane, the trophoblast is unusual in that it normally expresses these anionic phospholipids on its cell surface. This occurs as a result of the trophoblast’s high level of proliferation and differentiation that is associated with tissue remodeling during placentation [47, 48]. As a result, the positively charged plasma protein, β₂GPI, can bind to phosphatidylserine exposed on the external cell membranes of trophoblast undergoing syncytium formation, although additional receptors may also be involved [6, 7, 49]. Furthermore, the trophoblast synthesizes its own β ₂ GPI, and this protein translocates to the cell surface [50]. In vivo, there is evidence of β ₂ GPI localized to the surface of the extravillous trophoblast cells that invade the decidua and to the syncytiotrophoblast cells that are in direct contact with maternal blood [50, 51]. β ₂ -glycoprotein-I binds to the surface of the human trophoblast through the phospholipid-binding site in the fifth domain of the molecule, thus offering suitable epitopes for the maternal autoantibodies [6, 7, 49]. Hence, β₂GPI-dependent aPL appear to represent the main pathogenic autoantibodies in obstetrical APS. Accordingly, it has been hypothesized that most of them could be absorbed at the placental level (where β₂GPI is expressed) and not transferred to the fetus . This would explain why thrombotic events are rarely reported in babies born to aPL-positive mothers in spite of the high thrombophilic profile of neonates [52]. Although the syncytiotrophoblast and extravillous trophoblast populations both bind aPL recognizing β₂GPI, only the syncytiotrophoblast internalizes these antibodies via low-density lipoprotein receptors (LDLR) [53, 54]. The consequence of this aPL internalization is syncytiotrophoblast mitochondrial leak of cytochrome C [53] (Fig. 6.2B).

Since aPL bind to the trophoblast, it seems likely that the pathogenesis of pregnancy failure/complications in patients with APS is initiated at the placenta. Consequently, a number of studies have sought to evaluate the effects of aPL on trophoblast cells in vitro and have found that these autoantibodies affect several cell functions. Studies using human term trophoblast or choriocarcinoma cells show that aPL inhibit proliferation and syncytia formation [55–57], alter adhesion molecule expression [58], reduce invasiveness [56, 59–62], and decrease human chorionic gonadotropin [56, 61]. However, women with APS and pregnancy failure have circulating aPL at the time of implantation, and the most frequent clinical outcome is early pregnancy loss. Thus, more recent studies have shifted their focus to understanding the effects of aPL on the first trimester trophoblast. First trimester placental explants exposed to aPL have also been shown to produce less human chorionic gonadotropin (hCG) [63]. First and third trimester placental explants exposed to aPL generate distinct metabolic markers, ceramide, and diacylglycerols that may be involved in trophoblast responses to aPL [64]. Other studies using first trimester placental explants have found that the serum of patients with SLE/APS and recurrent pregnancy loss, anticoagulant-containing sera, or patient-derived aPL cause increased trophoblast apoptosis [63, 65, 66]. Anti-β₂GPI antibodies also augment the non-apoptotic shedding of trophoblastic material from first trimester placental explant cultures [67]. This is an important observation since during normal pregnancy, the placenta constitutively releases trophoblast microparticles, mononuclear trophoblast, and trophoblast syncytial knots from its outer syncytiotrophoblast layer into the maternal circulation [68], and in preeclampsia, a common outcome in APS-complicated pregnancies, shedding, or deportation of this material is significantly increased [69] (Fig. 6.2B).

Using in vitro cultures of human first trimester trophoblast cell lines and primary cells, mouse anti-β₂GPI monoclonal antibodies and purified patient-derived polyclonal aPL with β₂GPI reactivity have been found to enhance cytokine and chemokine secretion, specifically interleukin (IL)-8, IL-1β, growth-regulated protein (GRO)-α, and monocyte chemotactic protein (MCP)-1, which might explain the immune cell infiltration seen at the maternal-fetal interface. This aPL-induced inflammatory response is mediated by the Toll-like receptor 4 (TLR4)/MyD88 pathway [70], most likely because of molecular mimicry between β₂GPI and bacterial components, like lipopolysaccharide (LPS) [71, 72]. Indeed, in vivo studies have shown that animals immunized with microbial components develop aβ₂GPI antibodies and APS-like symptoms [73, 74]. Downstream of TLR4, trophoblast secretion of IL-1β secretion is mediated by the induction of endogenous uric acid, which in turn activates the Nod-like receptor (NLR), Nalp3, leading to Nalp3/ASC/caspase-1 inflammasome activation, and subsequent IL-1β [70, 75]. In parallel, IL-8 secretion is mediated by the induction of miR-146a-3p, which in turn activates the RNA sensor, TLR8 [76]. Thus, aPL-induced miR-146a-3p and uric acid act as endogenous secondary signals for trophoblast TLR8 and Nalp3 inflammasome activation, to drive trophoblast inflammation (Fig. 6.2C). The aPL-associated upregulation of trophoblast-derived uric acid and miR-146a-3p was found to be mirrored in the serum of women with aPL and adverse pregnancy outcomes [75, 76].

In parallel to the aPL-induced inflammatory response, aPL diminish the trophoblast’s ability to migrate, independently of the TLR4 signaling pathway, by inhibiting the cell’s constitutive production of IL-6, which in turn leads to decreased STAT3 activity [77], a critical mediator of trophoblast invasiveness [78]. This reduction in trophoblast migration was also found to involve aPL-induced upregulation of TIMP2 [79]. In a follow-up study, the surface-expressed receptor, apolipoprotein E receptor 2 (apoER2), was found to mediate this reduced migration [80]. However a recent study reported that patient-derived aPL reduce trophoblast invasion in a TLR4-depedent manner [81] (Fig. 6.2D).

Lastly, aβ₂GPI disrupts the basal trophoblast angiogenic factor balance, by inducing the secretion VEGF, PlGF, and sEndoglin levels (Fig. 6.2D). Although TLR4 is not involved in this response, functional MyD88 was required for the aPL-induced upregulation of PlGF, suggesting that receptors other than TLR4 that utilize MyD88, such as IL-1R or TLR8 [75, 76], are involved. All these aPL-mediated effects may play a role in causing a defective placentation [6, 7, 49]. Indeed, using an in vitro model of spiral artery transformation, aPL and sera from APS patients with pregnancy morbidity were shown to disrupt the normal trophoblast-endothelial cell interactions leading to reduced trophoblast-endothelial tube stability [82].

Data also suggest that aPL can cause abnormalities at the maternal side of the placenta. Impaired endometrial differentiation and lower expression of complement regulatory proteins (DAF/CD55) were found in endometrial biopsies from APS patients. These alterations before conception may compromise implantation and predispose to complement-mediated pregnancy failure [83, 84]. In addition, β₂GPI-dependent aPL are able to react with human stromal decidual cells in vitro and induce a pro-inflammatory phenotype [85]. These findings do suggest that APS-associated pregnancy complications can be mediated by several distinct pathogenic events.

At variance with the vascular manifestations of the syndrome, the two-hit hypothesis may not fit well with the APS obstetrical manifestations [6]. Passive infusion of IgG fractions with aPL activity induces fetal loss in naive pregnant mice and does not apparently require a second hit. β₂-glycoprotein-I is largely expressed in placental tissues even in physiological conditions [51, 86], and binding of labeled exogenous β₂GPI infused into naïve pregnant mice to trophoblast and endothelial in the labyrinth was recently documented in vivo by eXplore Optix™ imager [87]. Thus, the large availability of the target antigen for pathogenic aPL at the placental level is in strong contrast with the lack of a comparable expression in other tissues of naïve mice and even in highly vascularized human tissues such as the kidney [7]. It is possible to speculate that a high expression of β₂GPI at the placental level, together with the hormonal and blood flow modifications linked to the pregnancy, is sufficient to favor the pathogenic activity of the autoantibodies without any additional factor.

Preeclampsia, a hypertensive disorder unique to human pregnancy, is one of the obstetric conditions often associated with aPL, and aPL have been reported to increase a woman’s risk for developing preeclampsia almost tenfold [88–95]. While the pathogenic processes leading to preeclampsia are not fully understood, it is apparent that a toxin or toxins released from the placenta trigger maternal endothelial cell dysfunction . This maternal endothelial cell dysfunction is thought to be key to development of the maternal disease and may precede the onset of the clinical signs of preeclampsia by several weeks.

The human placenta is covered by a single multinucleated cell, the syncytiotrophoblast, which is bathed in maternal blood. Most cells produce subcellular lipid enclosed packages called extracellular vesicles that are released into the extracellular environment and are important in intercellular signaling. Extracellular vesicles from mononuclear cells are limited to microvesicles and nano-vesicles, some of which are exosomes. However due to its multinucleated structure, the syncytiotrophoblast also produces a unique type of multinucleated macro-vesicles called syncytial nuclear aggregates (SNAs), which are thought to be the end of the life cycle of the syncytiotrophoblast and therefore may be somewhat equivalent to apoptotic blebs produced by mononuclear cells. These multinucleated vesicles are larger than most mononuclear cells, being on average 72 μm in length and often have a tear-dropped shape [96, 97]. Syncytial nuclear aggregates are extruded from the surface of the syncytiotrophoblast directly into the maternal blood, which deports them to the maternal lungs where they become trapped in the pulmonary capillaries (the site where they were first identified in association with eclampsia over 120 years ago)[98]. While about 100,000 SNAs are deported daily from the placenta in all normal pregnancies [99], there is 20-fold increase in the number of SNAs deported in preeclampsia [96, 100, 101]. Syncytial nuclear aggregates are cleared from the maternal pulmonary capillaries on average in 3–4 days, and there is evidence that this clearance is due to phagocytosis of the SNAs by the maternal pulmonary endothelial cells against which they are trapped [102–104]. It has been suggested that SNAs from preeclamptic placentae may be one of the placental toxins that trigger preeclampsia [104, 105].

It has long been known that aPL can interact with the syncytiotrophoblast since, as described in the preceding section, aPL can reduce production of hCG, reduce formation of syncytiotrophoblast, and have been shown to disrupt the annexin V anticoagulant shield on the surface of the syncytiotrophoblast [10].

More recently it was reported that aPL, when incubated with first trimester placental explants, increases production of syncytial nuclear aggregates from the syncytiotrophoblast [103, 106]. So how do aPL induce the production of SNAs? Using a combination of murine monoclonal and patient-derived aPL, it was shown that aPL rapidly penetrate the syncytiotrophoblast with this process requiring as little as 2 min [53]. The internalization of the antibodies was receptor dependent. While the exact identity of the receptor remains unclear, transport of aPL into the syncytiotrophoblast was inhibited by receptor-associated protein (RAP) , a low-density lipoprotein receptor (LDLR) family-binding protein (and by chloroquine) suggesting that aPL enter the syncytiotrophoblast via LDLR-mediated endocytosis [53] (Fig. 6.2B).

Immunofluor-gold electron microscopy was used to show that once internalized into the syncytiotrophoblast, the aPL were bound to mitochondria and that there was an increase in the number of mitochondria with swollen morphology associated with the aPL [53]. Further, once internalized, aPL disrupted mitochondrial function resulting in decreased oxidative phosphorylation though complex IV and the translocation of cytochrome C from the mitochondria to the cytoplasm of the syncytiotrophoblast [53]. This movement of cytochrome C has the potential to increase cell death in the syncytiotrophoblast by promoting the formation of the apoptosome, while the disruption of mitochondrial function may lead to a more necrotic type of death process in the syncytiotrophoblast. It is likely that, via these effects on the mitochondria, aPL increase the production of SNAs (Fig. 6.2B).

Not only do aPL increase the number of SNAs extruded from the placenta but importantly they change the nature of SNAs. It has been shown that SNAs from normal first trimester placenta, when phagocytosed by endothelial cells, render those endothelial cells resistant to activation, and this is important to the normal maternal adaptation to pregnancy [102]. In stark contrast, the SNAs extruded from placentae exposed to aPL become “dangerous” and induce endothelial cell activation similar to that seen in preeclampsia [104] (Fig. 6.3B). Providing further evidence that aPL induce the syncytiotrophoblast to extrude dangerous SNAs as a result of an interaction with the mitochondria, a proteomic comparison of SNAs from aPL or control antibody-treated placentae showed that of 72 regulated proteins 13 were involved in mitochondrial function [106].

It is not yet entirely clear what changes aPL induce in SNAs to make them dangerous, but it has been shown that the alarmin, high mobility group box 1 (HMGB1), is increased in the cytoplasm of SNAs from aPL-treated placentae. Similarly, the amount of “free” mitochondrial DNA, which is also an inflammatory danger signal, is increased in the smaller micro- and nano-vesicles that are released from aPL-treated placentae. Thus, aPL rapidly enter the syncytiotrophoblast via a receptor-mediated mechanism and disrupt mitochondrial function, resulting in the release of dangerous SNAs into the maternal blood. Maternal endothelial cells become activated in response to these dangerous SNAs in the same manner as occurs in preeclampsia. This provides a potential mechanism to explain why aPL are such a potent risk factor that predispose women to develop preeclampsia.

Apolipoprotein E receptor 2 (apoER2) , also known as LRP8, is a member of the low-density lipoprotein (LDL) receptor family, which includes LDL receptor, LRP1, megalin, and very low-density lipoprotein receptor. When human apoER2 cDNA was first cloned, abundant expression of mRNA of the receptor was detected in the brain and in the placenta [107]. Subsequently, a critical role of the receptor was discovered in neuronal development as a receptor for reelin [108–110]. In the neuronal cells, reelin binding to apoER2 initiates activation of a series of kinases to regulate cellular function required for normal brain development [109–112]. More recently, it has been shown that the receptor is highly expressed in vascular cells, including platelets, monocytes/macrophages, vascular smooth muscle cells, and endothelial cells [113–120]. In 2003 de Groot’s group reported for the first time that that apoER2, a splice variant of apoER2 expressed in platelets , can interact with dimerized β₂GPI in platelets [121]. Because aβ₂GPI-β₂GPI complex plays a key role in pathogenesis of APS, a series of studies ensued in in vitro experiment, in cell culture, and in mouse models to determine the role of apoER2 in the actions of aPL [119, 122–126].

In endothelial cells, our group (Chieko et al.) found that monoclonal aβ₂GPI or polyclonal human aPL attenuate activity of endothelial nitric oxide (eNOS) through dephosphorylation of the enzyme at Ser1177, the critical phosphorylation site for the activity of the enzyme, leading to impaired production of nitric oxide (NO) [119] (Fig. 6.3A), which is a key signaling molecule for maintenance of normal vascular function. Nitric oxide produced by eNOS stimulates vascular relaxation, prevents endothelial inflammation, attenuates platelet activation, and reduces smooth muscle cell proliferation and migration [127, 128]. In cultured endothelial cells, we further demonstrated that inhibition of eNOS by aPL contributes to attenuated cell migration and increased adhesion to monocytes and that these aPL actions were abrogated with apoER2 knockdown by siRNA [119]. Mirroring the findings in culture, aPL administration in wild-type mice decreased eNOS-mediated vascular relaxation, impaired carotid artery reendothelialization after thermal injury, and increased leukocyte adhesion to the vascular endothelial layer [119, 129]. In contrast, apoER2-deficient mice were protected from these effects by aPL. A requirement for apoER2 in aPL-induced thrombus formation has also been determined in mouse models of APS by two laboratories independently [119, 126]. Our group has shown that aPL administration enhances thrombus formation in the mesenteric arterioles in wild-type mice, whereas it does not do so in apoER2 knockout mice [119]. Using eNOS knockout mice, we further demonstrated that eNOS antagonism is likely an underlying cause of aPL-induced thrombosis. The other group has also demonstrated that apoER2-deficient mice are protected from aPL- or dimerized β₂GPI-induced thrombosis in femoral vein [126]. The latter study also used soluble form of the apoER2 binding domain 1 (sBD1), a peptide mimicking the first ligand binding domain which blocks interactions with β₂GPI, to further verify that aPL-β₂GPI interaction with apoER2 mediates aPL actions in vivo. In addition to propensity for developing thrombosis, APS patients have elevated risk for non-thrombotic vascular occlusion [130–133] and greater stenosis of the celiac, intracranial, mesenteric, and renal arteries compared to non-APS individuals [134–140]. We have recently reported that when medial hypertrophy and neointima formation are invoked in mice by carotid artery endothelial denudation, aPL administration induces exaggerated neointima formation compared to control human IgG [129]. We further demonstrated that the neointima hyperplasia induced by aPL is related to impaired reendothelialization after injury. In contrast to the wild-type animals, apoER2-deficient mice were protected from the adverse effects of aPL on reendothelialization. Inhibitory effect of aPL on endothelial repair was prevented by concurrent administration of molsidomine, exogenous NO donor. These results indicate that aPL exacerbate both thrombosis and non-thrombotic neointima hyperplasia in the mouse models of APS and that the effect of aPL is caused by endothelial dysfunction mediated by aPL-β₂GPI interaction with apoER2 (Fig. 6.3A).

In the realm of pregnancy complications in the APS, numerous studies in trophoblast-derived cell lines, human primary trophoblasts, or placental explants have established that aPL treatment attenuates cell proliferation and migration, increases apoptosis, and impairs differentiation by altering the expression or activation of key proteins such as matrix metalloproteinases, interleukins, and chorionic gonadotropin [141]. Abundant expression of apoER2 mRNA was found in the placenta, and our recent study has indicated that both human and mouse trophoblasts express apoER2 protein [80, 107]. The study has further determined for the first time that apoER2 is required for aPL-induced trophoblast dysfunction in culture and fetal loss and IUGR in mice [80]. In cultured trophoblasts, the receptor was required for aPL inhibition of epidermal growth factor (EGF)-induced Akt phosphorylation, cell migration, and proliferation. In mice, pregnant apoER2-deficient mice injected with human aPL showed attenuated fetal loss and IUGR compared to the wild-type mice. In addition to impaired trophoblast function, endothelial cell dysfunction in the placenta also influences pregnancy outcome through insufficient development of endometrial angiogenesis [142, 143]. It has been reported in cultured endometrial endothelial cells and in a mouse angiogenesis model that aPL impair endothelial cell migration and neovascularization [144]. The inhibitory action of aPL was associated with decreased expression of VEGF and matrix metalloproteinases, and it was abrogated by the synthetic peptide TIFI, a competitive blocker of aPL binding to endothelium [144, 145] (Fig. 6.1B). Our work showed in aortic or carotid artery endothelium that aPL inhibit endothelial cell migration, which is dependent on the presence of apoER2 [129]. The requirement for apoER2 in aPL-induced pregnancy loss has been shown in the mice with global apoER2 deficiency, and further studies are warranted to determine whether apoER2 in trophoblast, endothelium, or platelets (or in all three cell types) is specifically required for aPL-induced pregnancy morbidity. Interestingly, in humans polymorphisms in apoER2 gene have been associated with fetal growth restriction [146], although the role of the polymorphism in the pregnancy phenotypes in APS has not been explored.

In summary, aPL binding to β₂GPI activates apoER2 to disturb normal cellular functions in endothelial cells and trophoblasts, contributing to enhanced thrombosis and non-thrombotic vascular occlusion as well as adverse fetal outcome in mice. We have recently developed a monoclonal antibody, which inhibits aPL-induced formation of β₂GPI-apoER2 complex [147] (Fig. 6.1A). We have shown that the monoclonal antibody decreases thrombus formation and fetal loss in the mouse model of APS. Thus, apoER2 and its downstream effector molecules may provide a new mode of therapeutic interventions to combat APS.

There is sound evidence that β₂GPI is present both at the maternal (decidual and uterine endothelial cells) and fetal side (trophoblast cells) acting as an antigenic target for β₂GPI-dependent pathogenic aPL in obstetric APS [6, 167].