Antiphospholipid antibody‐activated NETs exacerbate trophoblast and endothelial cell injury in obstetric antiphospholipid syndrome

Abstract Despite the widespread use of antiplatelets and anticoagulants, women with antiphospholipid syndrome (APS) may face pregnancy complications associated with placental dysplasia. Neutrophil extracellular traps (NETs) are involved in the pathogenesis of many autoimmune diseases, including vascular APS; however, their role in obstetric APS is unclear. Herein, we investigated the role of NETs by quantifying cell‐free DNA and NET marker levels. Live‐cell imaging was used to visualize NET formation, and MAPK signalling pathway proteins were analysed. Cell migration, invasion and tube formation assays were performed to observe the effects of NETs on trophoblasts and human umbilical vein endothelial cells (HUVECs). The concentrations of cell‐free DNA and NETs in sera of pregnant patients with APS were elevated compared with that of healthy controls (HCs) matched to gestational week. APS neutrophils were predisposed to spontaneous NET release and IgG purified from the patients (APS‐IgG) induced neutrophils from HCs to release NETs. Additionally, APS‐IgG NET induction was abolished with inhibitors of reactive oxygen species, AKT, p38 MAPK and ERK1/2. Moreover, NETs were detrimental to trophoblasts and HUVECs. In summary, APS‐IgG‐induced NET formation deserves further investigation as a potential novel therapeutic target in obstetrical APS.


| INTRODUC TI ON
Antiphospholipid syndrome (APS) is an autoimmune disease characterized by persistently elevated titres of antiphospholipid antibodies (aPLs) that predispose individuals to arterial and/or venous thrombosis or to pregnancy complications 1,2 ; the latter includes recurrent spontaneous abortion, pre-mature birth caused by pre-eclampsia and foetal growth restriction. Although the pathogenesis of these pregnancy complications is still not fully understood, it has been suggested to be a consequence of intraplacental thrombosis 3,4 ; however, increasing studies have suggested that placental inflammation is a hallmark characteristic of adverse pregnancy outcome complicated by aPL. 4 Current treatments for APS focus on inhibiting coagulation rather than treating the potential pathophysiology. 5 Antiphospholipid antibodies are a group of heterogeneous autoantibodies that bind a complex antigen involving phospholipid-binding proteins, such as β2-glycoprotein I (β2GPI), and negatively charged phospholipids, such as cardiolipin. 6 Antiphospholipid antibodies induce thrombosis by activating endothelial cells and platelets and impede placentation by damaging trophoblasts directly. Furthermore, activated neutrophils in APS may play an important role in foetal loss. 7 Circulating neutrophils from patients with APS are in an activated state, as indicated by an associated increase in the production of reactive oxygen species (ROS). 8 In a mouse model of obstetric APS, Girardi and colleagues 9 observed a significant inflammatory response in decidual tissue with a large number of neutrophils present in vivo. In this mouse model, depleting mice of neutrophils or inhibiting complement factor C5a can eliminate the deleterious effects of aPLs, 9,10 indicating that both neutrophils and the complement system have important pathophysiological effects in APS. 9 Neutrophil extracellular traps (NETs) are structures made of chromatin scaffolds covered with histones, proteases, granules and cytosolic proteins. NETosis describes the process by which neutrophils produce and release NETs. 11 NET formation occurs both in infectious diseases and in non-pathogenic conditions, such as autoimmune diseases. Low-density granulocytes are an inflammatory subset of neutrophils in systemic lupus erythematosus (SLE), which demonstrate robust demethylation of interferon genes 12 and release excessive NETs. 13 NET-mediated endothelial cell toxicity has been observed in patients with lupus nephritis, 13 and NETs may induce endothelial cell apoptosis and promote vascular injury in murine lupus models. 14,15 Excessive NET formation and the incomplete degradation of NETs have been considered to contribute to the pathogenesis of anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV). 16 Moreover, the rheumatoid arthritis (RA) synovial microenvironment is highly conducive to the release of NETs with the potential to activate immune cells in the synovium; thus, targeting NET formation as a novel treatment strategy may improve the prognosis of RA patients. 17 On the other hand, in bladder cancer treatment with Bacillus Calmette-Guerin, NETs were found to exert cytotoxicity, induce apoptosis and cell-cycle arrest, and inhibit the migration of bladder tumour cells. 18 Yalavarthi et al 19 reported that incubation with aPL stimulates neutrophils from healthy volunteers to release NETs and that neutrophils in patients with APS are more prone to NETosis than are healthy volunteer neutrophils. As NETs stimulate platelet activation and thrombosis 20 and are present in placentas of patients with pre-eclampsia, 21 strong exposure to NETs may be related to the pathogenesis of APS. However, there are no studies reporting whether neutrophils from pregnant women with APS are more prone to NETosis, and, aside from the influence of NETs on thrombosis, whether they cause direct damage to umbilical vein endothelial cells and trophoblasts. Therefore, we investigated aPL/neutrophil interplay in pregnant women with APS and examined the effects of NETs on umbilical vein endothelial cells and trophoblasts, with the aim of improving the prognosis of pregnant women with APS.

| Patient information
Given the established links between NETs and autoimmune diseases and that are some differences between obstetric APS and vascular APS, we focused on pregnant women with APS for the main experi-

| Purification of patient immunoglobulin G (IgG)
IgG was purified from APS or control sera with a NAb™ Protein A Plus Spin Kit (Thermo Fisher Scientific, Waltham, MA) according to manufacturer's instructions and as previously described. 24 Briefly, sera were passed through a Protein A Agarose Column at least three times. IgG was then eluted with 0.1 mol/L glycine and neutralized with 1 mol/L Tris. IgG purified from APS sera was termed APS-IgG. IgG purified from control sera was termed HC-IgG. IgG concentrations were determined by a BCA protein assay (Solarbio, Beijing, China) according to manufacturer's instructions. IgG purity was verified with Coomassie staining. All IgG samples were determined to contain no detectable endotoxins using a Chromogenic Endotoxin Quantitation Kit (Thermo Fisher Scientific) according to manufacturer's instructions.

| Cell-free NET purification
To purify cell-free NETs, 2 × 10 6 cells were added into 6-well plates, incubated and stimulated as described above. Following stimulation for 4 hours, cells were gently washed after the medium was removed. After addition of 500 µL RPMI (phenol red-free) to the adherent film and vigorous agitation, the samples were centrifuged at 2000 × g for 5 minutes and the supernatant collected as previously described. [25][26][27] Cell-free DNA and protein levels were quantified using the Quant-iT PicoGreen dsDNA Assay Kit and a BCA protein assay, respectively, according to manufacturers' instructions.

| Isolation of human umbilical vein endothelial cells
We

| Cell culture
The first-trimester human extravillous trophoblast cell line HTR-8/ SVneo was purchased from the American Type Culture Collection (ATCC; Manassas, VA). The cells were maintained in RPMI 1640 medium supplemented with 10% FBS and 100 nmol/L penicillin/ streptomycin (Gibco). Neutrophils and HUVECs were isolated as mentioned above. All cells were maintained in a humidified chamber at 37°C and 5% CO 2 .

| Cell migration and invasion
Cell migration and invasion were measured using a two-chamber Transwell migration assay, as previously described. 28 The lower chamber (24-well plate) was filled with 600 μL of complete me-

| Scratch wound-healing assay
HTR-8/SVneo cells (5 × 10 5 ) were seeded in a 6-well plate and cultured under a humidified atmosphere of 5% CO 2 at 37°C. A clean scratch across the centre of the cell layer was generated with a p1000 sterile pipette tip after the cells had formed a confluent monolayer.
Floating cells were gently washed away with PBS. Subsequently, cells were treated with NETs or medium and then imaged after 24 hours (4× objective). Cell migration area was estimated using ImageJ.

| Tube formation
Matrigel was diluted with ECM at a 1:1 ratio and added to 96-well plates. The plates were placed in a 37°C incubation room for 30 minutes to allow the Matrigel to polymerize. HUVECs were seeded onto the Matrigel-coated wells at 1 × 10 4 cells/well and incubated for 6 hours at 37°C in growth supplement-free medium in the presence of the indicated test compounds. Samples were visualized using an inverted microscope, and ImageJ was used to quantify tube formation.

| Statistical analysis
Data are expressed as means ± SEM of three independent experiments. Statistical differences between samples were assessed by Student's t test or one-way ANOVA. Statistical analysis was performed with GraphPad Prism (ver. 7; GraphPad Software Inc, La Jolla, CA). P values < 0.05 were considered statistically significant.

| Sera from pregnant women with APS show increased cell-free DNA and NETs
Cell-free DNA and NETs are increased in APS patients with thrombosis 19 and other disease conditions, such as sepsis, 29

| Neutrophils from pregnant patients with APS are primed for NETosis
Without specific in vitro stimulation, APS neutrophils displayed intensified spontaneous NET release compared with that of HC neutrophils (Figure 2A-C). Moreover, cell-free DNA in APS neutrophils was much higher than that in HCs (263.9 ± 8.44 vs 197.6 ± 7.85 ng/ mL, P < 0.01; Figure 2A). To determine whether elevated cell-free DNA results from NET release, the cell-impermeable DNA dye SYTOX Green was used to monitor extracellular release of DNA from stimulated neutrophils, whereas the cell-permeable DNA dye Hoechst 33342 was used to label intracellular DNA, followed by direct live-cell imaging. The per cent of APS neutrophils forming NETs was significantly higher than that of HC neutrophils (26.75% ± 1.03% vs 21.25% ± 1.55%, P < 0.05; Figure 2B,C). The prolonged interval between peripheral blood collection and neutrophil isolation may affect cell-free DNA levels ( Figure S3). Therefore, we suggest that the blood should be processed within two hours of drawing from the donor.
Therefore, APS-IgG significantly stimulates NET release compared with that of HC-IgG.

| APS-IgG mediates NET release via ROS production
To determine how APS-IgG promotes NET release, we analysed

| AKT, ERK1/2 and p38 MAPK pathways participate in NET release
As the AKT, ERK and p38 MAPK pathways are involved in PMAinduced, ROS-dependent NET release, phosphorylation of these molecules was assessed by Western blot analysis. We found that AKT, p38 MAPK and ERK1/2 phosphorylation was significantly increased in cells treated with APS-IgG and PMA ( Figure 5A-C).
We next treated neutrophils cultured in the presence of APS-IgG with specific AKT, ERK1/2 and p38 MAPK pathway inhibitors to assess their role in the induction of APS-IgG-mediated NET release. As shown in Figure 5D-F, incubation with the relevant inhibitors for 30 minutes was sufficient to inhibit AKT, p38 MAPK and ERK1/2 phosphorylation. As expected, MK2206, SCH772984 and SB203580 decreased p-AKT, p-ERK and p-p38 MAPK levels, respectively, whereas total AKT, ERK and p38 MAPK levels remained unaltered ( Figure 5D-F). MK2206, SCH772984 and SB203580 treatment had no effect on cell-free DNA levels ( Figure 6A). APS-IgG combined with MK2206, SCH772984 and SB203580 treatment decreased cell-free DNA levels ( Figure 6A). Direct microscopic observation also yielded similar results ( Figure 6B,C). These findings suggest that AKT, ERK1/2 and p38 MAPK pathway activation is required for NET release. When treated with HC-IgG-induced NETs, trophoblast migration and invasion capacity did not show the same changes as when treated with APS-IgG-induced NETs ( Figure S5A-D).

| APS-IgG-induced NETs influence the migration and tube formation ability of HUVECs
Endothelial cells play a decisive role in maintaining homeostasis during the first trimester, which is necessary for sustaining normal pregnancy; thus, we examined the effects of NETs on HUVECs and found that NETs inhibited HUVEC migration (53.33 ± 13.54 vs 173.00 ± 5.51, P < 0.01; Figure 7E,F), as demonstrated by Transwell assays. We also evaluated the effects of APS-IgG-induced NETs on the angiogenic ability of HUVECs by performing a tube formation assay. Exposure to NETs reduced tube formation, as demonstrated by the decreased number of tube branches, junctions, meshes and mesh areas. These inhibitory effects were reversed, in part, when DNase I was added ( Figure 7G,H). The effects of DNase I on PMAinduced NETs were similar to its effects on APS-IgG-induced NETs ( Figure S4E-H). Moreover, when treated with HC-IgG-induced NETs, HUVEC migration and tube formation abilities did not show the same changes as when treated with APS-IgG-induced NETs ( Figure   S5E-H). were also shown to induce NETosis in RA cases. 37 Moreover, enhanced NET formation was observed in patients with SLE, which is connected to both autophagy and defective clearance of NETs. 38 ily. 46 In our study, APS-IgG was found to activate ERK1/2 and p38

| D ISCUSS I ON
MAPK, in addition to promoting phosphorylation of the AKT pathway. Interestingly, AKT, ERK1/2 and p38 MAPK inhibitors only partly inhibited APS-IgG-induced NETosis, demonstrating that although AKT, ERK1/2 and p38 MAPK were necessary for aPL-induced ROS production and NET formation, there are other pathways taking part in aPL-induced NET formation. The signalling pathway involved in pathogenesis is likely to resemble that of PMA-induced NET release, instead of lipopolysaccharide-induced NET release. 47 The relationship between AKT, ERK1/2 and p38 MAPK in aPL-induced NET formation and other pathways needs further investigation.
NET components, such as DNA and histones, promote thrombosis in different diseases. 48 For instance, the DNA backbones of NETs are capable activators of factor XII because of its negatively charged surface. 49 Meanwhile, histones may accelerate coagulation via promotion of platelet activation, aggregation and thrombin generation. 50,51 Besides thrombosis, NETs have been shown to trigger endothelial and epithelial cytotoxicity and induce endothelial dysfunction by decreasing cell proliferation and increasing cell apoptosis. 14,25,52 Recently, NETs were found to play an important role in bladder cancer treatment by attacking tumour cells. 18 In the present study, we further found that NETs decrease the invasive and migratory abilities of trophoblasts and have a negative effect on the migration and tube formation ability of HUVECs. Therefore, we suggest that NETs contribute to pregnancy morbidity by di-

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.