REVIEW ARTICLE: Guilty As Charged: All Available Evidence Implicates Complement’s Role in Fetal Demise

Recurrent miscarriages are fairly common. On average, one in five pregnancies will end in a miscarriage and some research shows that there are up to 800,000 miscarriages a year in the USA. Women who experience repeated miscarriages may undergo expensive and lengthy tests to try to identify a cause, but often get no answers. In 80% of the unexplained abortions, an immune mechanism has been proposed. Mammalian mothers are faced with a problem. The genome of the fetus they carry within their wombs is half maternal and half paternal. Fetuses express paternal antigens early in development Thus, antigens presented by the fetus that are paternal in origin would be considered foreign by the mother’s immune system. The mechanisms of protection of the allogeneic fetus from the maternal immune response during pregnancy remain mysterious more than 50 years after the paradox of maternal tolerance was first raised by Peter Medawar. How does the fetus evade the maternal immune system and survive for 9 months in an environment that considers it foreign? Experimental observations suggest a relationship between pregnancy complications and an abnormal maternal immune response.^1,2

Complement activation and tissue injury

Complement is a phylogenetically ancient system and an essential component of the innate immune system. The complement system, composed of over 30 proteins that act in concert to protect the host against invading organisms, initiates inflammation and tissue injury (Fig. 1).^3,4 Complement activation promotes chemotaxis of inflammatory cells and generates proteolytic fragments that enhance phagocytosis by neutrophils and monocytes.^5,6 The classical pathway is activated when natural or elicited antibodies bind to antigen and unleash potent effectors associated with humoral responses in immune-mediated tissue damage. Activation of the classical pathway by natural antibodies plays a major role in the response to neoepitopes unmasked on ischemic endothelium, and thus may be involved in reperfusion injury.⁷ The mannose-binding lectin (MBL) pathway is activated by MBL recognition of carbohydrates (often on infectious agents) and MBL-associated serine protease-2, which activates and cleaves C2 and C4. Alternative pathway activation differs from classical and MBL activation because it is initiated directly by spontaneous deposition of complement on cell surfaces. Under normal physiologic conditions, C3 undergoes low-grade spontaneous hydrolysis and deposits on target surfaces, allowing binding and activation of factor B, formation of the alternative pathway C3 convertase, and further amplification of C3 cleavage. This pathway is antibody-independent and is triggered by the activity of factor B, factor D and properdin.^8,9 By means of these recognition and activation mechanisms, the complement system identifies and responds to foreign antigens, pathogens, tissue injury, ischemia, apoptosis and necrosis.¹⁰ This capacity places the complement system at the center of many clinically important responses to pathogens and, relevant to this review, to fetal injury mediated by cellular or humoral immune mechanisms.

The convergence of three complement activation pathways on the C3 protein results in a common pathway of effector functions. The initial step is generation of the fragments C3a and C3b. C3a, an anaphylotoxin that binds to receptors on leukocytes and other cells, causes activation and release of inflammatory mediators.¹¹ C3b and its further sequential cleavage fragments are ligands for complement receptors 1 and 2 and the β2 integrins present on a variety of inflammatory and immune accessory cells.^12,13 C3b attaches covalently to targets and allows the assembly of C5 convertase, which cleaves C5 to C5a and C5b. C5a is a potent soluble inflammatory, anaphylatoxic and chemotactic molecule that promotes recruitment and activation of neutrophils and monocytes and mediates endothelial cell activation through its receptor. Expression of C5aR, initially thought to be limited to neutrophils, monocytes and eosinophils, has recently been shown to be more widespread and to include endothelial cells, vascular smooth muscle cells and mast cells among others.¹⁴

Binding of C5b to the target initiates the assembly of the C5b-9 membrane attack complex (MAC). MAC, a pore-forming lipophilic complex, may cause tissue necrosis by lysing cells. However, a majority of cells are resistant to lysis and non-lethal effects of the MAC, which trigger cell activation, are likely to be of more relevance to pathology. MAC can activate pro-inflammatory signaling pathways through the interaction of membrane-associated MAC proteins with heterotrimeric G proteins.^15,16 As a consequence of these signaling events, pro-inflammatory and tissue damaging phenotypes develop characterized by cell proliferation, release of reactive oxygen species (ROS), leukotrienes, thromboxane and platelet-derived growth factor and increased procoagulant activity.^17–20 MAC plays a major role in human disease. MAC is present at sites of tissue damage in rheumatoid arthritis and glomerulonephritis.^4,7 Indeed, deficiency of CD59, the complement regulatory protein that inhibits MAC formation, seen in paroxysmal nocturnal hemoglobinuria and diabetes, respectively, is associated with thrombosis and vasculopathy.²¹

Because complement fragments bind and damage self-tissues, autologous cells need to be protected from the deleterious effects of complement. To this point, most human and murine cells express soluble and membrane-bound molecules that limit the activation of various complement components.^3,22

Complement activation is at core of a long list of disease pathologies. Altered complement regulation causes and may perpetuate complications of pregnancy.^23,24 As fetal tissues are semi-allogeneic and alloantibodies commonly develop in the mother, the placenta is potentially subject to complement-mediated immune attack at the maternal–fetal interface.²⁵ Uncontrolled complement activation is prevented in successful pregnancy by three regulatory proteins.^26,27 Two of these membrane-bound proteins, DAF and MCP, regulate the activation of C3 and C4 on the surface of cells.^28,29 DAF inhibits the assembly and accelerates the decay of the C3 convertase enzymes that activate C3 and amplify the classical and alternative complement pathways. MCP is a cofactor for factor I-mediated degradation and inactivation of C3b and C4b.²⁷ The third protein, CD59, is also membrane anchored and prevents assembly of C5b-9 MAC, blocking the terminal effector functions of complement.²⁸ Several studies have shown that the levels of activated complement components present at a given site depends on the relative effects of complement activators and their inhibitors, DAF, MCP and CD59.²⁵ All three proteins are expressed on the trophoblast in contact with maternal blood and tissues, providing a mechanism to protect the fetus from damage due to complement pathway activation by alloantibodies. Indeed, on the basis of their characteristic distribution patterns, it is likely that these proteins are strategically positioned for this purpose.

Complement activation and pregnancy loss

Studies by Molina et al. underscore the importance of complement regulation in fetal control of maternal processes that mediate tissue damage. In mice, Crry (complement receptor-1 related gene/protein Y) is a membrane-bound intrinsic complement inhibitory protein. Crry has DAF-like activity, accelerating decay of C3 convertases, and MCP-like activity, serving as a factor I cofactor to accelerate degradation of activated C3 and C4.³⁰ Thus, Crry acts as an inhibitor of the classical and alternative pathways. Crry, key part of innate immunity, is necessary for the embryo to survive pregnancy in mice. That appropriate complement inhibition is an absolute requirement for normal pregnancy has been demonstrated by the finding that Crry deficiency in utero leads to embryonic death.³¹ Crry−/− embryos, generated by gene targeting, die and no live births have been found.³¹ Embryos die surrounded by activated C3 fragments and polymorphonuclear cells within the developing placenta. Importantly, when the Crry+/− mice are intercrossed with C3−/− mice to generate C3−/−, Crry−/− embryos, there is complete rescue from this lethal outcome.³¹ This indicates that the Crry−/− embryos die in utero because of their inability to inhibit spontaneous complement activation and tissue damage mediated by C3. Lack of complement inhibitor Crry on the trophoblasts leads to complement activation that attracts and activates inflammatory cells that destroy the embryo.

Women with recurrent pregnancy loss develop hypocomplementemia, in the absence of autoantibodies, suggesting a role of the complement system in the pathogenesis of fetal injury.³² Other human studies also show that it is possible that complement activation is also involved in miscarriages. A cross-sectional study showed that plasma concentrations of complement component C5a are higher in patients with fetal death than in normal pregnant women.³³ Increased complement deposition was observed in placentas from patients with anti-phospholipid antibodies (aPL), compared with control patients.³⁴

Taken together, the murine and human findings suggest that there is recognition of fetal tissues by immune mechanisms that trigger complement fixation and that in the presence of excessive complement activation or in the absence of complement inhibitors the fetus is at risk for injury.

Here, we examine the evidence for the involvement of complement activation in two mouse models of pregnancy loss: an antibody-dependent model induced by aPL and an antibody independent model [DBA/2-mated female CBA/J mice (CBA/J x DBA/2)].

Antibody-dependent mouse model of pregnancy loss: aPL-induced fetal death

The anti-phospholipid syndrome (APS) is characterized by recurrent fetal loss, vascular thrombosis and thrombocytopenia occurring in the presence of aPL antibodies. We extensively studied the role of complement activation in the pathogenesis of fetal loss in a mouse model of APS. We hypothesized that aPL antibodies activate complement in the placenta, generating split products that mediate placental injury and lead to fetal loss and growth restriction.^35,36 To demonstrate our hypothesis, we used a murine model of APS in which pregnant mice were injected with human IgG containing aPL antibodies or human aPL monoclonal antibodies.³⁵ We found that blocking complement activation with the C3 convertase inhibitor Crry-IgG or genetic deletion of C3 (C3−/− mice) protected mice from aPL-induced fetal loss and growth restriction (Fig. 2).³⁵ To determine the role of specific complement components in aPL-induced pregnancy loss, we focused on C3, a pivotal element in the complement cascade, which when cleaved leads to the generation of C3a (anaphylotoxin) and C3b (opsonin). We studied the role of C3a by performing experiments in C3aR deficient mice (C3aR−/−) (Fig. 2). C3aR−/− mice treated with aPL-IgG had fetal resorption frequencies comparable to those of C3aR sufficient mice treated with aPL-IgG, indicating that C3a-C3aR interaction is not required for aPL-induced pregnancy loss.³⁷ On the other hand, C5 deficient mice (C5−/−) and mice treated with anti-C5 monoclonal antibody were protected from aPL-induced pregnancy loss (Fig. 2), indicating that C5 activation is required for aPL-induced pregnancy complications. Therefore, we considered the possibility that C5 cleavage products might induce fetal injury. Cleavage of C5 generates two complement fragments: anaphylotoxin C5a that promotes inflammation by recruiting and activating inflammatory cells and C5b, which initiates the assembly of the MAC (C5b-9) that can destroy cells by permeabilization of the membranes and can also trigger cell activation. To identify the C5 cleavage product responsible for embryonic death in aPL-treated mice, we first focused on the MAC. To investigate whether pregnancy loss is caused by MAC formation, we studied C6 deficient mice (C6−/−). C6−/− mice treated with aPL-IgG were not protected from fetal death and showed injured embryos³⁷ (Fig. 2). Pregnancy outcomes in aPL-treated C6−/− mice were similar to those observed in C6+/+ mice, indicating that MAC formation is not required for fetal injury.

To study the role of the C5a-C5aR pathway in mediating fetal injury, we used C5aR deficient mice (C5aR−/−) and we blocked C5aR in wild type mice with a highly specific peptide antagonist of C5aR, AcPhe(l-ornithine-Pro-d-cyclohexylalanine-Trp-Arg), which possesses potent in vivo anti-inflammatory activity in mouse models of endotoxic shock.³⁸ Both strategies prevented pregnancy loss induced by aPL antibodies, indicating that C5a-C5aR interaction is crucial for fetal death in aPL-treated mice (Fig. 2). However, we were unable to detect C5aR expression on trophoblasts by fluorescence-activated cell sorter analysis or in mouse decidua by immunoblotting and polymerase chain reaction analysis. Therefore, C5a-C5aR interaction must take place on cells other than the decidual trophoblasts. C5a is one of the most potent inflammatory peptides. Monocytes and neutrophils, the major phagocytic leukocytes, migrate to inflammatory sites by sensing chemoattractants, such as anaphylatoxin C5a. The ligand–receptor interaction of these chemoattractants induces not only chemotactic movement but also the increase in cytoplasmic Ca²⁺ concentration, the release of granular components, and the generation of ROS. Given that neutrophil infiltration was observed in deciduas from aPL-treated mice³⁶ and that neutrophils express C5aR, we tested the hypothesis that neutrophils are important effectors of tissue damage in this model of antibody dependent-pregnancy loss. Indeed, neutrophil depletion rescued embryos from the lethal effects of aPL, demonstrating that neutrophils are crucial effectors of fetal injury in APS (Fig. 2). Figure 2 summarizes the different strategies we used to identify the complement components and cell subtypes involved in aPL-induced pregnancy loss.

Heparin is one of the standard anticoagulant treatments used to prevent obstetric complications in patients with aPL antibodies.³⁹ Yet, heparin is used in sub anti-coagulant doses (unfractionated 5000 U; s.c. twice daily) in these patients. As early as 1929, heparin was shown to inhibit complement activation. Indeed, sub-anticoagulant doses of heparin have therapeutic benefits in other conditions associated with complement activation, including reperfusion injury and coronary artery disease.^40,41 We hypothesized that heparin protects pregnant women with APS from complications through inhibition of complement. Treatment with heparin [UFH or low molecular weight (LMWH)] prevented complement activation in vivo and in vitro and protected mice from pregnancy complications induced by aPL antibodies.⁴² Even in the absence of a detectable anticoagulant effect, as in mice treated with UFH 10 U, heparins protected mice from aPL antibody-induced pregnancy loss. Fondaparinux, a synthetic pentasaccharide composed of the minimal antithrombin-binding subunit of heparin, is a specific inhibitor of factor Xa, and lacks N- and O-sulfation, which is required for the anti-complement activity associated with heparin.⁴³ Hirudin is a direct inhibitor of thrombin that acts independently of antithrombin and other plasma proteins and has no known effects on complement.⁴⁴ Neither fondaparinux nor hirudin inhibited the generation of complement split products or prevented pregnancy loss, demonstrating that anticoagulation is insufficient therapy for APS-associated miscarriage. Our data suggest that heparins prevent obstetrical complications in women with APS because they block activation of complement induced by aPL antibodies targeted to decidual tissues, rather than by their anticoagulant effects.

Anti-phospholipid syndrome is considered a thrombophilic disorder. However, animal studies from our laboratory have shown the importance of inflammation in APS.^36,44 That neutrophils are important mediators of fetal death in APS and that anticoagulation therapy is not enough to prevent miscarriages highlights the important role of inflammation in the pathogenesis of aPL-induced pregnancy loss. Recently, human studies showed that inflammation in the placenta may contribute to APS pregnancy complications, reinforcing this new concept of the APS as an inflammatory disorder.⁴⁵ Thrombosis and inflammation are linked in many clinical conditions. Tissue factor (TF), the major cellular initiator of the coagulation protease cascade, plays important roles in both thrombosis and inflammation.⁴⁶ We studied whether TF contributes to aPL-induced fetal loss in mice. Blockade of TF activity with a specific monoclonal antibody⁴⁷ and experiments in low TF expressing mice⁴⁸ showed less inflammation (fewer neutrophils and less C3 deposition) in deciduas and embryos survival in aPL-treated mice, indicating that TF is a proinflammatory molecule in this model. Experiments performed in TF^{floxed/floxed}/LysMCre that do not express TF on myeloid cells, allowed us to distinguish the role of trophoblasts TF from that of myeloid cells TF. The protection from aPL-induced pregnancy loss observed in these mice emphasizes the crucial role of TF in maternal myeloid cells.³⁷ Moreover, knowing that monocytes are not required for aPL-induced pregnancy loss and that neutrophils from aPL-treated TF^{floxed/floxed}/LysM-Cre mice do not express TF allow us to conclude that TF expression on maternal neutrophils plays a causative and crucial role in aPL-induced fetal injury. TF ^{floxed/floxed}/LysM-Cre mice treated with aPL showed normal pregnancies and diminished decidual inflammation, suggesting that TF expression on neutrophils modulates neutrophils tissue-injuring ability. Indeed, neutrophils from TF^{floxed/floxed}/LysM-Cre mice treated with aPL showed a lower generation of oxidants and less free-radical-mediated lipid peroxidation in deciduas, when compared with TF^{floxed/floxed} mice, that express TF, suggesting that TF is a modulator of oxidative burst in neutrophils.

Collectively, our study demonstrates that maternal TF is essential in the pathogenesis of aPL-induced fetal loss and reveals a functional linkage between complement components, TF and neutrophils. After aPL-IgG binding to trophoblasts, complement activation occurs leading to generation of the anaphylatoxin C5a, which attracts and activates neutrophils through C5aR (Fig. 3). C5a-C5aR interaction on neutrophils results in TF expression. Activated neutrophils release ROS and proteolytic enzymes leading to decidual damage. TF expression on neutrophils leads to the formation of complexes that amplify inflammation, cell injury and fetal death.

Antibody independent mouse model of pregnancy loss: DBA/2-mated female CBA/J mice (CBA/J × DBA/2)

A second murine model of pregnancy loss underscores the importance of complement as a mediator of fetal injury. DBA/2-mated female CBA/J mice (CBA/J × DBA/2) are a well-studied model of immunologically mediated pregnancy loss that shares features with human recurrent miscarriage.^49,50 In this murine model of fetal allograft rejection, maternal T cells specific for paternal antigens trigger complement activation and extensive deposition of C3 at the maternal–fetal interface despite adequate Crry expression.⁵¹ Thirty percent of the embryos of DBA/2-mated female CBA/J mice are resorbed and increased complement deposition and macrophage infiltration is observed surrounding the dead embryos. The surviving embryos show intrauterine growth restriction (IUGR).⁵¹ Complement inhibition rescued embryos of DBA/2-mated CBA/J mice providing direct evidence that complement activation is a critical mediator rather than a consequence of embryo injury in this model of spontaneous pregnancy loss. Blockade of C3 activation with Crry-IgG that prevents all downstream complement effector activity prevented pregnancy loss in DBA/2-mated CBA/J mice, indicating that complement is required for fetal loss. Crry-Ig also protected surviving fetuses from IUGR, indicating that complement activation also contributes to growth failure in developing embryos. To identify the complement pathway that mediates fetal injury and placental dysfunction, we studied the role of C5 by treating DBA/2-mated CBA/J pregnant mice with an anti-C5 mAb that effectively prevents C5 activation in aPL-induced pregnancy loss.³⁶ Blockade of C5 cleavage with anti-C5 mAb averted pregnancy loss and fetal growth restriction in CBA/J × DBA/2 matings. We then treated pregnant CBA/J × DBA/2 and control CBA/J × BALB/c matings with a specific peptide antagonist of C5aR, which showed potent anti-inflammatory activity in the mouse model of APS.³⁶ Administration of C5aR antagonist peptide prevented both fetal resorption and IUGR in CBA/J × DBA/2 matings. Protection conferred by the C5aR antagonist indicates that C5a-C5aR interactions are important mediators in pregnancy damage in CBA/J × DBA/2 matings.

Immunohistological analysis of deciduas from CBA/J × DBA/2 mice treated with C5aR antagonist peptide showed minimal C3 deposition surrounding normal appearing fetuses and no evidence of inflammation,⁵¹ suggesting that recruited inflammatory cells amplify complement activation on trophoblasts and within deciduas. Inflammatory cells recruited and activated by C5a-C5aR interaction can amplify complement activation by releasing alternative pathway components. The role of alternative pathway activation was later confirmed using an inhibitory mAb to factor B. Fetal rejection and growth failure in CBA/J × DBA/2 matings was prevented by anti-factor B mAb, indicating that the alternative pathway plays a central role in initiating and/or amplifying injury.

In sum, our results show that factor B, C3, C5 and C5aR are required for pregnancy loss and growth restriction in this antibody-independent model of spontaneous miscarriage.³⁵

Normal placental development is required for successful embryonic growth and requires co-ordinated expression of angiogenic growth factors, vascular endothelial growth factor (VEGF) and placenta growth factor (PlGF), as well as expression of their respective receptors on invasive trophoblasts.⁵² sVEGFR-1 is a potent anti-angiogenic molecule that sequesters circulating VEGF and PlGF and inhibits placental cytotrophoblast differentiation and vessels formation leading to abnormal placentation.⁵³ The average weight of placentas from CBA/J × DBA/2 matings was 24% lower than that from CBA/J × BALB/c pairs, consistent with abnormal development. Because defects in trophoblast differentiation and placental development limit embryonic growth, we performed histologic analyses of mid-trimester placentas from abortion-prone and control matings. Trophoblast giant cells are required for implantation and invasion of the conceptus into maternal decidua, and reduction in their number or abnormal differentiation can cause placental defects and compromise pregnancies. A relative deficiency of trophoblast giant cells was observed in placentas from CBA/J × DBA/2 matings compared with CBA/J × BALB/c matings suggesting abnormal trophoblast differentiation. Impaired placental vascularization (fewer fetal vessels in the labyrinthine layer) was also observed in CBA/J × DBA/2 matings.⁵¹

Because proteins and receptors of the VEGF family are important for adequate placental development and fetuses from DBA/2-mated CBA/J mice die or show inadequate growth in utero, we considered the possibility that complement proteins trigger dysregulation of angiogenic factors and produce placental dysfunction. We hypothesized that complement activation prevents VEGF signaling leading to pregnancy complications in CBA/J × DBA/2 matings. In CBA/J × BALB/c matings, free VEGF levels increased throughout pregnancy, whereas in DBA/2-mated CBA/J pregnancies free VEGF did not increase above pre-pregnancy levels. Plasma levels of sVEGFR-1, that antagonizes VEGF, were higher in CBA/J × DBA/2 matings than in controls. These findings demonstrate a dysregulation of angiogenic factors in abortion-prone matings that correlate with the defective placental development we observed.

Does complement activation contribute to angiogenic factor dysregulation? To assess the effect of complement activation on levels of angiogenic factors, we measured free VEGF and sVEGFR-1 levels in CBA/J × DBA/2 matings treated with the complement inhibitors that protected pregnancies. We did not observe a sustained elevation of plasma sVEGFR-1 in mice treated with blockers of complement, suggesting that sVEGFR-1 is a critical effector of fetal injury acting downstream of complement activation. Inhibition of complement activation at the levels of C3 (Crry-Ig), C5 (anti-C5 mAb) or C5a-C5aR interactions (C5aR antagonist peptide) prevented the increase in sVEGFR-1 and associated decrease in free VEGF in CBA/J × DBA/2 matings.⁵¹ Complement inhibition also prevented placentas abnormalities and pregnancies were protected. Taken together, our data indicate that blockade of complement activation prevents the increase in sVEGFR-1; permits free VEGF signaling in trophoblasts averting the fetal resorptions and growth restriction characteristic of CBA/J × DBA/2 pregnancies.

C3 deposition on decidual tissue and trophoblasts was associated with intense infiltration of monocytes. Mononuclear phagocytes, that can express sVEGFR-1, are essential effectors of damage in this model. Can complement split products trigger the production of sFlt-1 in mononuclear cells and inhibit VEGF activity? Incubation of mouse spleen macrophages with synthetic C5a induced sVEGFR-1 production, while C5aR-deficient macrophages did not increase sVEGFR-1 release after stimulation with immune complexes in the presence of serum. Furthermore, we found that C5a induced sVEGFR-1 release directly, independently of TNF-α.⁵¹ In conclusion, complement split product C5a triggers monocytes to secrete sVEGFR-1 which sequesters VEGF. Low levels of free VEGF leads to abnormal placental development, inadequate placental perfusion and causes fetal growth restriction or death (Fig. 4).

We showed that complement activation is a required intermediary event in the pathogenesis of fetal injury in this antibody-independent model of spontaneous miscarriage and IUGR. By implicating inflammatory cells (recruited and activated by complement split products) as a source of inhibitors of angiogenesis, we have identified a new mechanism by which immune activation, an important contributor to pregnancy failure, may cause recurrent abortion and fetal growth restriction.

Our studies in antibody-dependent and antibody-independent models of pregnancy complications identified complement activation as the key innate immune effectors that mediate poor pregnancy outcomes and will allow development of new therapies to prevent fetal injury and growth restriction in patients with APS.