Women with antiphospholipid syndrome (APS) are at risk for pregnancy complications. Antiphospholipid antibodies (aPL) alter trophoblast function by triggering an inflammatory cytokine response; modulating angiogenic factor secretion; and inhibiting migration. While patients with APS are often treated with hydroxychloroquine (HCQ), its effect on trophoblast function is poorly understood.
Method of study
A human first trimester trophoblast cell line was treated with or without antihuman β2GPI mAbs in the presence or absence of HCQ. Supernatants were analyzed by ELISA. Cell migration was measured using a colormetric assay.
Antiphospholipid antibodies-induced trophoblast IL-8, IL-1 β, PlGF, and sEndoglin secretion were not altered by HCQ. aPL-induced inhibition of trophoblast migration was partially reversed by HCQ, even though HCQ significantly increased secretion of pro-migratory IL-6 to greater than baseline. aPL-induced upregulation of TIMP2 appears to inhibit trophoblast migration; the inability of HCQ to prevent aPL-induced TIMP2 may explain why migration was only partially restored.
Hydroxychloroquine reversed the aPL-inhibition of trophoblast IL-6 secretion and partially limited aPL-inhibition of cell migration. Thus, some form of combination therapy that includes HCQ may be beneficial to pregnant APS patients.
Antiphospholipid syndrome (APS), an autoimmune systemic disease characterized by circulating antiphospholipid antibodies (aPL), places women at risk of adverse pregnancy outcomes, such as recurrent pregnancy loss and late gestational complications of pre-eclampsia, HELLP syndrome (hemolytic anemia, elevated liver enzymes, and low platelet counts), premature delivery, and intrauterine growth restriction. Complications such as pre-eclampsia are also known to lead to health risks for the mother later in life, including hypertension and cardiovascular disease. Methods of treatment for APS during pregnancy typically involve low molecular weight heparin (LMWH), either alone or in combination with aspirin.[3, 4] This treatment has been shown to significantly increase the live birth rate in patients with APS who have experienced multiple miscarriages. However, the incidence of severe late obstetric complications, including pre-eclampsia, placental insufficiency, intrauterine growth restriction and prematurity, remains high.[3, 4] In addition, clinical and experimental studies have produced contradictory results regarding the effectiveness of heparin in preventing adverse pregnancy outcomes in APS.[5-9] Therefore, the development of improved strategies for the prevention of pregnancy complications in women with APS is a pressing concern.
Despite the association of APS with thrombosis and the success of anticoagulants, such as heparin, in preventing pregnancy loss, intravascular or intravillous blood clots are not typically found in placenta samples from patients with APS. Rather, clinical and experimental studies suggest that adverse pregnancy events are primarily caused by inflammatory processes, including increased cytokine production, complement deposition, and immune cell activation.[11-15] Excessive inflammation causes placental insufficiency in APS-complicated pregnancies as evidenced by reduced trophoblast invasion and limited uterine spiral artery transformation.[10, 16, 17]
aPL specifically target the placenta by binding the phospholipid-binding protein, beta2-glycoprotein I (β2GPI), which is constitutively expressed on the trophoblast cell surface.[18-20] We have previously shown that aPL alter first trimester human trophoblast function through multiple mechanisms.[13, 17, 21] First trimester trophoblast cells exposed to aPL at a dose that has no effect on cell viability or proliferation produce elevated levels of pro-inflammatory cytokines (IL-8 and IL-1β,) through activation of the innate immune receptor, Toll-like receptor 4 (TLR4). Further, aPL inhibit trophoblast migration; and modulate the trophoblast angiogenic factor profile by increasing pro-angiogenic vascular endothelial growth factor (VEGF) and placenta growth factor (PlGF), and anti-angiogenic soluble endoglin secretion. Thus, the identification of therapeutic interventions that can prevent these aPL-mediated trophoblast responses may translate clinically to improved pregnancy outcomes in patients with APS. Indeed, we have used this in vitro model to assess the impact of heparin, aspirin, and pravastatin on trophoblast responses to aPL, and reported that only LMWH confers partially beneficial effects.[9, 13, 21, 22] Thus, there is a need to evaluate the efficacy of alternative therapeutics that could be used either alone or in combination with heparin.
Patients with lupus and APS are often treated with the anti-malarial drug hydroxychloroquine (HCQ),[23-26] and it is useful for treating lupus flares during pregnancy. In addition to having antithrombotic effects by preventing platelet aggregation and arachidonic acid release, HCQ has anti-inflammatory and immunomodulatory properties.[23, 24, 26, 28, 29] Although HCQ is safe to use during pregnancy,[25, 26] little is known about its effects on aPL-induced adverse pregnancy outcome, and its effect on trophoblast function is poorly understood. However, recent studies have reported that HCQ reduces the binding of aPL-β2GPI complexes to phospholipid bilayers and protects the anticoagulant protein annexin A5 from disruption by aPL in term trophoblast, again by reduced antibody binding.[30-32] Since circulating aPL are present in women with APS at the time of implantation, the objective of this study was to investigate the impact of HCQ on first trimester human trophoblast cell exposed to aPL.
Materials and methods
Trophoblast Cell Line
The human first trimester trophoblast cell line, SVneo transformed HTR8, was used in all experiments. Cells were cultured at 37°C/5% CO2 in RPMI-1640 media (GIBCO; Grand Island, NY, USA), supplemented with 10% fetal bovine serum (Hyclone; South Logan, UT, USA), 10 mm Hepes, 0.1 mm MEM non-essential amino acids, 1 mm sodium pyruvate, and 100 nm penicillin/streptomycin (GIBCO).
The antiphospholipid antibodies used in all experiments were the mouse IgG1 anti-human β2GPI mAbs, ID2 and IIC5. These antibodies were produced by one of us (LWC). ID2 and IIC5 bind β2GPI in the same manner as human aPL when immobilized on a suitable negatively charged surface, such as the phospholipids, cardiolipin or phosphatidyl serine, or irradiated polystyrene. ID2 and IIC5 have been shown to alter trophoblast function similarly to patient-derived polyclonal aPL upon binding to first trimester trophoblast cells.[13, 21, 36] In order to further characterize the epitope specificity of the mAbs, dot blots were conducted to compare their reactivity against recombinant β2GPI which lacked domain I or domain V. Native β2GPI was purified from human plasma by precipitation, heparin–Sepharose chromatography and size exclusion chromatography as previously described and was used as a positive control. Domain deletion mutants of β2GPI were produced recombinantly with N-terminal hexahistidine tags in E. coli and were purified by nickel affinity chromatography by GenScript (Table 1). Both native β2GPI and recombinant β2GPI lacking domains I or V were at least 85% pure as judged by Coomassie Blue stained SDS-PAGE. Plasma-purified whole β2GPI and the recombinant β2GPI lacking domains I or V were blotted onto duplicate nitrocellulose membranes (GE Pharmacia, Auckland, New Zealand) in a doubling-dilution series starting at 15 μg/well using a Bio-Dot apparatus (Bio-Rad, Auckland, New Zealand). One membrane was used for immunoblotting as follows. Non-specific binding was blocked by incubation with 5% non-fat milk powder in phosphate buffered saline pH 7.4 containing 0.1% Tween-20 (PBST). The blots were probed with IIC5 or ID2 in 5% non-fat milk powder/PBST for 1 hr room temperature. Bound antibodies were detected using horseradish peroxidase-coupled antimouse antibody (GE Pharmacia) and ECL prime (GE Pharmacia). The second membrane was used to detect total membrane-bound protein. After application of the β2GPI this membrane was washed with PBST for 1 hr room temperature then stained with Coomassie Blue protein stain (0.5 g/L Brilliant blue G, 50% methanol, 40% H2O, 10% acetic acid). Membranes stained with Coomassie Blue or probed with ID2/IIC5 were imaged using a LAS3000 scanner (Fujifilm, Auckland, New Zealand).
Table 1. Protein sequences of recombinant β2GPI domain deletion mutants lacking domain 1 or domain V
Domain deletion mutant
β2GPI lacking domain 1 (Dll-V)
β2GPI lacking domain V (DI-IV)
Cell Viability Assay
Trophoblast cell viability was determined using the CellTiter 96 viability assay (Promega, Madison, WI, USA), as previously described. Cells (1 × 104) were seeded into wells of a 96-well plate in growth media and incubated overnight. The media was then replaced with serum-free OptiMEM (Invitrogen; Carlsbad, CA, USA), and cells incubated for another 4 hrs. Cells were then treated with HCQ (Sigma Aldrich) at 0, 0.01, 0.1, 1, 10, and 100 μg/mL. HCQ was reconstituted in endotoxin-free water and subsequent dilutions made using OptiMEM prior to treatment. After 48 hrs, the CellTiter 96 substrate (Promega) was added to all wells and incubated for 2 hrs at 37°C. Optical densities were read at 490 nm. All samples were assayed in triplicate, and cell viability was presented as a percentage of the untreated control (0 μg/mL).
Cytokine, Angiogenic Factor, and TIMP Secretion
Trophoblast cells (1 × 105) were seeded into 35 mm tissue culture dishes in growth media and allowed to adhere overnight. The media was then replaced with serum-free OptiMEM (Invitrogen), and cells incubated for another 4 hrs. Cells were then treated with or without aPL (ID2 or IIC5) at 20 μg/mL in OptiMEM (Invitrogen) in the presence and absence of HCQ at 1 μg/mL. After 72 hrs, cell-free supernatants were collected by centrifugation at 400 g for 10 minutes and stored at −80°C until analysis was performed as previously described.[13, 21, 36] The concentrations of interleukin 1 beta (IL-1β), interleukin 8 (IL-8), interleukin 6 (IL-6), vascular endothelial growth factor (VEGF), placenta growth factor (PlGF), soluble fms-like tyrosine kinase-1 (sFlt-1), soluble endoglin (sEndoglin), tissue inhibitor of metalloproteinase 1 (TIMP1) and tissue inhibitor of metalloproteinase 2 (TIMP2) were evaluated by ELISA (R & D Systems; Minneapolis, MN, USA).
To assess trophoblast migration, a two chamber colorimetric assay was used, as previously described. An 8-μm pore size cell culture insert served as the top chamber (BD Biosciences, Franklin Lakes, NJ, USA), while the lower chambers were wells of a 24-well tissue culture plate (BD Falcon, Franklin Lakes, NJ, USA). The lower chamber was filled with 800 μL of OptiMEM while the top chamber was seeded with 1 × 105 cells HTR8 cells suspended in 200 μL of treatments. Following a 48-hr incubation, the 8 μm-inserts were removed and trophoblast migration across the membrane was determined using the QCM 24-well colorimetric cell migration assay according to the manufacturer's instructions (Chemicon International, Temecula, CA, USA). Briefly, migrated cells were stained, collected, and lysed. The resulting colored mixture was transferred to a 96-well plate and optical densities read in triplicate at 560 nm. A 100% migration control consisted of the starting number of cells (1 × 105). Observed optical density values were measured using a BioRad plate reader (Hercules, CA, USA), and were expressed as a percentage of the starting total number of cells.
All experiments were performed at least three times. Data are expressed as mean ± SEM. Statistical significance was set at P <0.05 and was determined by one-way ANOVA using Prism software (GraphPad Software, Inc., La Jolla, CA).
The Antiphospholipid Antibodies IIC5 and ID2 Recognize Epitopes in Domain V of β2 Glycoprotein I
Dot blotting was used to determine the domain(s) of β2GPI that contain the epitopes for the anti- β2GPI mAbs used in this study, ID2 and IIC5. Both ID2 and IIC5 recognized β2GPI purified from human plasma as well as recombinant β2GPI that lacked domain I. Neither ID2 nor IIC5 bound to recombinant β2GPI in which domain V was not expressed. Despite probing equal amounts of full-length β2GPI and deletion mutants, we consistently found that more recombinant β2GPI, lacking domain V was retained on the membrane, reinforcing the lack of reactivity between ID2 or IIC5 and this deletion mutant (Fig. 1).
Effect of HCQ Dose on Trophoblast Viability
The first objective of this study was to determine the range of HCQ concentrations that lacks trophoblast cytotoxicity. Figure 2 shows that at the two highest doses tested, 10 μg/mL and 100 μg/mL, HCQ significantly reduced trophoblast cell viability when compared to the untreated control. There was no significant effect on cell viability at all other concentrations tested (Fig. 2). Therefore, for all subsequent experiments, the highest non-cytotoxic dose of 1 μg/mL HCQ was used, which is consistent with previous reports.[30-32]
HCQ has no Effect on aPL-induced Pro-Inflammatory Trophoblast Response
Having established the drug concentration, we first sought to determine the effect of HCQ on aPL-induced trophoblast inflammation.[9, 13, 22] Treatment of trophoblast cells with ID2 or IIC5 significantly increased IL-8 secretion when compared to the no treatment (NT) control. HCQ had no significant effect on aPL-induced trophoblast IL-8 production (Fig. 3a). Similarly, treatment of trophoblast cells with ID2 induced a significant increase in IL-1β production, irrespective of the presence of HCQ (Fig. 3b). Further, HCQ alone had no effect on basal IL-8 or IL-1β secretion by trophoblast (Fig. 3).
Effect of HCQ on aPL Modulation of Trophoblast Angiogenic Factor Production
We next investigated the effect of HCQ on aPL-induced modulation of trophoblast angiogenic factor secretion.[9, 21, 22] HCQ alone had no effect on basal secretion of VEGF by trophoblasts (Fig. 4a), PlGF (Fig. 4b), sFlt-1 (Fig. 4c), or sEndoglin (Fig. 4d). Treatment of trophoblasts with ID2 or IIC5 induced a significant increase in the secretion of VEGF (Fig. 4a). The presence of HCQ slightly, yet significantly, lowered the ID2-induced VEGF secretion, but had no effect on IIC5-induced VEGF (Fig. 4a). Treatment of trophoblast cells with ID2 or IIC5 significantly increased secretion of PlGF by trophoblasts, and this was unaffected by HCQ (Fig. 4b). Neither ID2 nor IIC5 in the presence or absence of HCQ affected sFlt-1 secretion by trophoblast (Fig. 4c). ID2 significantly increased secretion of sEndoglin but this was unaffected by HCQ (Fig. 4c).
HCQ Attenuates the Detrimental Effect of aPL on Trophoblast Migration
We previously demonstrated that IL-6 produced by first trimester trophoblast drives chemokinesis through activation of STAT3. Treatment of trophoblast cells with ID2 or IIC5 reduced IL-6 secretion, although not to the level of significance. However, in the presence of HCQ, IL-6 secretion was significantly increased when compared with the NT control or treatment with ID2 or IIC5 alone (Fig. 5a). HCQ had no significant effect on basal IL-6 secretion, although an increase was noted. Treatment of trophoblast cells with ID2 or IIC5 significantly reduced trophoblast migration by 65.9 ± 32.0% and 65.6 ± 18.2%, respectively, when compared to the NT control. The presence of HCQ had no effect on basal trophoblast migration. However, HCQ slightly, but significantly, reversed the ID2 and IIC5-induced inhibition of trophoblast migration (Fig. 5b).
Effect of HCQ on aPL-Induced Trophoblast TIMP1 and TIMP2 Secretion
Trophoblast migration and invasion are partially governed by the balance between production of the matrix metalloproteinases (MMPs), MMP2, and MMP9, and the tissue inhibitors of matrix metalloproteinases (TIMPs), TIMP1, and TIMP2. While treatment of trophoblasts with aPL had no effect on the expression levels of MMP2 or MMP9 (data not shown), ID2 as well as IIC5 significantly increased secretion of TIMP1 and TIMP2 when compared with the NT and IgG controls (Fig. 6a). Furthermore, treatment of trophoblasts with recombinant (r) TIMP2, alone or combined with rTIMP1, significantly reduced trophoblast migration (Fig. 6b). Treatment with ID2 or IIC5 significantly increased trophoblast secretion of TIMP1 and TIMP2 when compared with NT control (Fig. 7a, b, respectively). The presence of HCQ modestly but significantly, inhibited ID2-induced secretion of TIMP1 (Fig. 7a). By contrast, a modest but significant, increase in TIMP2 secretion response to IIC5 was observed in cells treated with HCQ (Fig. 7b).
APS poses a substantial risk of pregnancy complications, even when patients are treated with heparin and/or aspirin. Furthermore, several studies have disputed the effects of heparin on trophoblasts and have called its efficacy into question.[3-7, 9, 39] Thus, further research into the potential management of APS-related pregnancy complications is critical. While HCQ has been used to treat APS and related autoimmune disorders such as lupus, its effect on trophoblast function remains poorly understood.[23-26] Therefore, using an in vitro system, we sought to determine the ability of HCQ to alter aPL-mediated modulation of first trimester trophoblast function, specifically, cell migration and the secretion of inflammatory chemokines and angiogenic factors. In this study, we report that HCQ partially antagonizes aPL-induced inhibition of trophoblast migration, possibly through the modulation of IL-6 production.
Recent studies have shown that the immunomodulatory properties of HCQ may be explained by an inhibition of TLR signaling through its interference with downstream mitogen-activated protein kinase (MAPK) activation and by dampening the expression of TLRs and their associated signaling molecules.[24, 28, 29, 40] Based on this, we postulated that HCQ might prevent the TLR4-depedent trophoblast inflammatory cytokine response induced by aPL. Moreover, recent studies have demonstrated that HCQ is able to reduce the binding of aPL-β2GPI complexes to phospholipid bilayers and protect the anticoagulant protein annexin A5 from disruption by aPL in term trophoblast cells.[31, 32] In our initial experiments, it was found that high doses of HCQ (10 μg/mL and 100 μg/mL) negatively affected cell viability. Therefore, a dose of 1 μg/mL was selected for use in further experiments, which correlated with concentrations previously used.[30-32]
We found that HCQ alone had no significant effect on basal trophoblast secretion of inflammatory cytokines or angiogenic factors. Moreover, HCQ had no effect on aPL-induced IL-8 and IL-1β production, suggesting that the ability of HCQ to inhibit TLR4 signaling may not be applicable to the trophoblast. Alternatively, it could be a reflection of the short period of exposure of the trophoblast to HCQ. While a similar in vitro exposure to HCQ can inhibit activation of the endosomal TLRs,[28, 29] the ability of HCQ to down-modulate the TLR4 signaling pathway has only been shown in vivo after a long-term exposure of 6 months. Our data also suggests that HCQ does not prevent aPL binding to first trimester trophoblasts. This is in contrast to what has been reported for the effects of HCQ on aPL-β2GPI complexes binding to lipid bilayers and for aPL binding to term trophoblast,[30-32] even though for these studies as well as our own, aPL and HCQ were delivered to the cells at the same time.
The monoclonal antibodies used in this study have been previously shown to have similar reactivity to human aPL, and here we demonstrate that they react specifically with epitopes within domain V of β2GPI. β2GPI consists of five domains of approximately 60 amino acids referred to as sushi domains or short consensus repeats. aPL are known to react with epitopes in all five domains of β2GPI, although there is particular interest in antibodies directed against domains I and V.[44, 45] We sought here to determine the domain specificity of the monoclonal aPL, IIC5, and ID2, by testing their reactivity with two domain deletion mutants of β2GPI. Both monoclonal antibodies bound to recombinant β2GPI lacking domain I. In contrast, they did not bind to recombinant β2GPI lacking domain V, suggesting that they recognize epitopes within domain V which is deleted in this mutant. Since the density of immobilized β2GPI may be important to allow divalent binding of aPL, we sought to confirm that the lack of reactivity between the mAbs and β2GPI lacking domain V was not due to inadequate density of this domain deletion mutant. Coomassie Blue staining demonstrated that considerably more of the recombinant β2GPI lacking domain V was immobilized on the membrane than either whole β2GPI or the mutant lacking domain I despite the native and recombinant β2GPI preparations being applied to the membrane at the same concentration. This makes it highly unlikely that failure of the antibodies to bind to this variant was due to inadequate antigen density.
While HCQ had no affect on aPL-induced PlGF or sEndoglin secretion, we did observe a slight, but significant, reduction in VEGF secretion induced by one of the anti-β2GPI mAbs, ID2. This is in keeping with a study showing that that chloroquine reduces epidermal VEGF expression. While HCQ did not have an overt effect on the angiogenic factor profile, nor did it prevent aPL-induced trophoblast inflammation, we did observe an effect on the aPL-mediated reduction in IL-6 secretion and cell migration. Conflicting reports of the effects of HCQ on cell migration exist, but its effects on migration of trophoblasts have not been studied.[47, 48] We had previously reported that treatment of trophoblasts with aPL reduced basal IL-6 secretion, STAT3 phosphorylation, and trophoblast migration. Moreover, by blocking the IL-6-IL-6 receptor interaction, we could also inhibit trophoblast migration, suggesting this to be the mechanism by which aPL limit cell migration. In this current study, we found that HCQ completely reversed the effect of aPL on IL-6 production and even increased secretion when compared to basal levels. However, the aPL-induced reduction in cell migration was only partially reversed by the presence of HCQ. This finding suggests that while HCQ might have a beneficial effect on aPL-induced inhibition of trophoblast migration, another factor in addition to IL-6 may be involved in its regulation, and this factor is not altered by HCQ.
Trophoblast migration and invasion can be regulated by the balance between the cell's production of the matrix metalloproteinases, MMP2 and MMP9, and the tissue inhibitors of matrix metalloproteinases, TIMP1 and TIMP2. Therefore, we sought to determine whether aPL disrupted this balance of MMPs and TIMPs in the trophoblast, and whether the presence of HCQ provided any protection. Unlike previous reports that have shown aPL to reduce MMP2 and MMP9 expression in trophoblast cells,[49, 50] we found MMP2 and MMP9 expression to be unchanged after exposure to either of the anti- β2GPI mAbs. Instead, our data show that aPL upregulate both TIMP1 and TIMP2 secretion, and that the presence of recombinant TIMP2 either alone, or in combination with TIMP1, can limit trophoblast migration, similarly to treatment with aPL or disruption of IL-6 signaling. This is in keeping with studies in other cellular systems showing the regulation of cellular migration by TIMP1 and TIMP2.[51-53] However, the presence of HCQ was only able to slightly prevent ID2-induced TIMP1 secretion and had no protective effect on aPL-induced TIMP2 secretion. Our findings are supported by previous data that chloroquine promotes endothelial cell migration and contradicts previous reports that chloroquine increases TIMP1 in serum levels of patients with lupus. Our studies also indicate that the beneficial effects of HCQ on IL-6 and TIMP1 secretion, but its inability to affect TIMP2 secretion, may explain why HCQ was able to improve but not wholly reverse the inhibition of trophoblast migration by aPL.
Together, our data indicate that HCQ may confer beneficial effects on the placenta in patients with APS, although this response seems limited trophoblast migration. In addition, these in vitro studies may not perform as a comprehensive model of the in vivo condition of APS during pregnancy. As HCQ has been shown to be safe to use during pregnancy and is often used to treat APS and lupus as an anti-inflammatory and immunomodulatory drug, it may be useful to examine whether HCQ produce more clinical benefits when combined with other drugs, such as heparin. However, further experimentation is necessary to determine the optimal combination of drug therapy in this complex condition.
CRA, WJS, CAV, LWC, and VMA participated in the study design, analysis and manuscript drafting. CRA, WJS, CAV and MM executed the study. VMA, LWC, and JJB contributed to manuscript revision and critical discussion.
This work was supported by grants from the Lupus Foundation of America (to CRA) and the March of Dimes (to VMA). CAV is the recipient of a University of Auckland Doctoral Scholarship.