1 Introduction
Preeclampsia (PE) is a placenta-mediated pregnancy complication characterized by the onset of hypertension accompanied by proteinuria or other indications of maternal organ dysfunction at 20 weeks of gestation and beyond. PE is the leading cause of maternal and perinatal morbidity and mortality [
1,
2]. The only effective treatment is the delivery of the placenta; however, this treatment may inevitably lead to premature birth and neonatal death [
3]. Currently, low-dose aspirin is the only drug recommended for preventing PE; however, aspirin merely addresses early-onset preeclampsia (< 20% of all PE cases) and is ineffective for late-onset preeclampsia (> 80% of PE cases) [
4]. Therefore, methods for preventing PE are urgently needed.
Genetic, maternal, and immunological factors can cause placental dysfunction, which leads to the release of anti-angiogenic and inflammatory mediators into maternal circulation; these mediators activate leukocytes, endothelial cells, and platelets to promote systemic vascular dysfunction, inflammatory responses, and excessive thrombin generation [
5,
6]. Patients with PE are associated with a more serious inflammatory and hypercoagulable status as compared with healthy pregnant women [
6–
8]. Therefore, the effective control of thromboinflammatory responses might be the key to preventing PE incidence and progression.
Microparticles (MPs) are as reportedly involved in PE progression. MPs are membrane vesicles (diameter: 0.1–1 μm) shed from the cell membrane; they vary in phospholipid and protein contents and antigens, which reflect their cellular origin [
9]. An asymmetric distribution of phospholipid in the plasma membrane bilayer, with phosphatidylserine (PS) distributed in the inner layer, has been observed. Following stimulation, PS is oriented outward and exposed on the cell surface, which results in membrane blebbing and MP release [
10]. Our previous work and other studies reported elevated MP levels in patients with PE [
11–
13]. Additionally, Kohli
et al. [
14] reported that MPs promote thromboinflammatory response via NLRP3 inflammasome activation in trophoblasts, resulting in PE. MPs containing PS and tissue factor exhibit a high level of procoagulant activity [
15]. Moreover, MPs can interact with and activate target cells, such as monocytes, platelets, and endothelial cells, by increasing the expression of adhesion molecules and receptors on the cell surface and promoting endothelial dysfunction and inflammatory response [
16]. Therefore, the reduction of procoagulant and proinflammatory MPs may inhibit the thromboinflammatory response and prevent PE.
Recombinant protein diannexin is a homodimer of annexin A5, which is a potent anticoagulant protein highly expressed by human trophoblasts, vascular endothelial cells, and other cell types [
17]. Diannexin functions as a shield that coats PS by forming a 2D lattice structure, which prevents the activation of the clotting cascade [
18]. Additionally, annexin A5 can pinocytose PS and reverse the process of membrane blebbing [
19], which might reduce MP release. However, annexin A5 has a short half-life (< 20 min) because of its low molecular weight (35.7 kDa) [
20], which makes it unsuitable for clinical applications. Diannexin exhibits an approximately 10-fold stronger binding affinity for PS and longer half-life (6–7 h) relative to annexin A5 [
21]. Thus, it might effectively reduce MP release from activated cells. Notably, placental trophoblast cells from patients with PE demonstrate annexin A5 deficiency [
22,
23], which further suggests the possible utility of diannexin for preventing PE.
Therefore, this study explored the abilities of diannexin to reduce MP release and prevent PE-like symptoms in an animal model. The results of this study might offer a treatment strategy for preventing PE.
2 Materials and methods
2.1 Patients and control subjects
From December 2017 to December 2018, 24 pregnant women with PE and 14 healthy pregnant women were enrolled in Peking University Third Hospital, Beijing, China. PE diagnosis was performed in accordance with Hypertensive Disorders of Pregnancy: ISSHP Classification, Diagnosis, and Management Recommendations for International Practice [
1]. Patients with multiple pregnancies, secondary hypertension, proteinuria at < 20 weeks of gestation, chronic hypertension at the first prenatal visit, and the presence of fetal malformations were excluded. Blood samples were collected from all subjects, and placenta tissues were obtained at the time of caesarean delivery. Three healthy non-pregnant volunteers were also enrolled for
in vitro experiments. Each participant provided written informed consent before joining the study. The study was approved by the Hospital Research and Ethical Committee (Approval number: IRB00006761-2016055) in accordance with the
Declaration of Helsinki.
2.2 Mice
Wild-type C57BL/6 mice (6–8 weeks) were purchased from the Laboratory Animal Science Department of Peking University Health Science Center (Production certificate number: Beijing SCXK 2011 0012). The mice were raised under specific pathogen-free conditions with free access to water and a commercial diet. All animal experiments were conducted in the Department of Laboratory Animal Science of Peking University Health Science Center. All efforts were made to minimize the number and suffering of animals, and all animal experiments were followed according to the standards and procedures approved by the Local Animal Care and Use Committee.
2.3 Reagents
FITC–annexin V (BD Biosciences, #556419, San Jose, CA, USA), anti-human HLA-G antibody (Biolegend, #335906, San Diego, CA, USA), anti-mouse CD31 antibody (Biolegend, #160204), and LEGENDplex™ mouse inflammation panel (Biolegend, #740446) were used in flow cytometry. Anti-NLRP3 antibody (1:1000, Abcam, #ab210491, Cambridge, MA, USA), anti-caspase 1 antibody (1:1000, Abcam, #ab207802), anti-IL-1β antibody (1:500, Abcam, #ab216995), anti-IL-18 antibody (1:1000, Abcam, #ab243091), anti-β-actin antibody (1:3000, Applygen, #C1313, Beijing, China), anti-rabbit IgG, horseradish peroxidase (HRP)-linked antibody (1:3000, CST, #7074, Danvers, MA, USA), anti-mouse IgG, HRP-linked antibody (1:3000, CST, #7076), and Mouse Reactive Inflammasome Antibody Sampler Kit (CST, #20836) were used in Western blot analysis. Anti-NLRP3 antibody (1:200, Abcam, #ab214185), anti-caspase 1 antibody (1:500, Abcam, #ab74279), and anti-IL-1β antibody (1:500, Abcam, #ab2105) were used in immunohistochemistry detection.
The following reagents were purchased from Sigma-Aldrich (Taufkirchen, Germany): calcimycin A23187, lipopolysaccharide (LPS), 1,2-diacyl-sn-glycero-3-phospho-L-serine (PS), and L-α-phosphatidylcholine (PC). Dichloromethane and chloroform were from Tong Guan, Inc. (Beijing, China). Fetal bovine serum (FBS), DMEM Ham’s F-12K medium, and trypsin–EDTA (Gibco; Thermo Fisher Scientific, Germany) were used for cell culture.
Diannexin was synthesized and supplied by State Key Laboratory of Proteomics, Beijing Institute of Lifeomics according to a nucleotide sequence from Alavita Pharmaceuticals, Inc. The nucleotide sequence and protein information are provided in the Supplementary Materials.
2.4 Cell culture and MP generation
Citrated blood from the healthy non-pregnant donors was centrifuged at 200× g for 8 min to obtain platelet-rich plasma, which was then separated into two parts. One part was stimulated with 10 μmol/L calcimycin A23187 at 37 °C for 10 min, and the other was pre-incubated with diannexin at a final concentration of 20 or 40 nmol/L at 25 °C for 30 min and then treated with 10 μmol/L calcimycin A23187. After both treatments, platelet-rich plasma was collected and centrifuged at 2500× g for 20 min at 25 °C to acquire the supernatant for the flow cytometric analysis of platelet microparticles (PMPs).
The human-derived umbilical vein endothelial cell line, EA.Hy-926 (ATCC, Manassas, VA, USA), was cultured in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin at 37 °C in 5% CO2 atmosphere. The EA.Hy-926 cells were then stimulated with 20 μg/mL LPS for 6 h at 37 °C or pre-incubated with diannexin for 1 h at a final concentration of 20 or 40 nmol/L, followed by stimulation. The culture supernatant was collected and then centrifuged at 2500× g for 20 min for flow cytometry analysis. Next human- and mouse-derived endothelial microparticles (EMPs) were prepared. EA.Hy-926 cells and mouse-derived lymphoid endothelial cell line (SVEC4-10, ATCC) were separately stimulated with 20 μg/mL LPS. The culture supernatant was collected, centrifuged at 2500× g for 10 min to remove apoptotic bodies, and then centrifuged at high speed (20 000× g) for 50 min to pellet pure EMPs for in vivo intervention (with SVEC4-10-derived EMPs) and in vitro trophoblast cell treatment (with EA.Hy-926-derived EMPs).
Human trophoblast-like cell lines (BeWo and SWAN-71, ATCC) were used to investigate the role of diannexin in inhibiting trophoblast cell activation. The BeWo cells were cultured in F-12K medium supplemented with 10% FBS and 1% penicillin–streptomycin at 37 °C in 5% CO2 atmosphere. The SWAN-71 cells were cultured in DMEM with 10% FBS and 1% penicillin–streptomycin at 37 °C in 5% CO2 atmosphere. The two cell lines were then incubated with diannexin (10, 20, 30, or 40 nmol/L) for 1 h at 37 °C prior to the addition of EMPs (200/μL final concentration). After the cell lines were cultured for 24 h, the supernatant was collected and centrifuged at 2500× g for 10 min for the detection of trophoblast MPs and IL-1β. Finally, the cells were harvested for the Western blot analysis of the expression of NLRP3 inflammasome markers.
2.5 Flow cytometry for MP count
The MP count was measured by flow cytometry according to our previous study [
11]. Briefly, 50 μL of platelet-poor plasma or culture supernatant was added to Trucount Absolute Counting Tubes (#340334, BD Biosciences) containing a specific number of fluorescent beads for quantifying MPs and then labeled with 5 μL of FITC–annexin V or PE anti-human HLA-G antibody. The samples were incubated for 30 min at room temperature in the dark and then analyzed using a BD FACSCanto II (BD Biosciences) equipped with BD FACSDiva
TM software. The number of MPs per microliter was calculated using the formula: MPs (/μL) = (number of bead counts per tube × number of MP events acquired) / (number of bead counts collected × sample volume).
2.6 Construction and detection of PS microcapsules and PC microcapsules
A total of 2 mg PS or PC and 0.1 g polylactide were dissolved in a solution containing 4 mL of dichloromethane and 500 μL of chloroform. The solution was then placed in a 20 °C thermostat bath, followed by emulsification with 20 mL of water using a disperser at 6500 rpm for 8 min. Organic solvents were evaporated via magnetic stirring for 3 h at 300 rpm. After the evaporation was completed, the microcapsules were washed, resuspended in water, and adjusted to 2 mg/mL. For morphological observation, 10 μL of PS or PC microcapsules were mounted on 200-mesh Formvar carbon-coated copper grids, negatively stained with phosphotungstic acid, examined under a JEM-1400 PLUS (Jeol, Tokyo, Japan) transmission electron microscope, and captured using a QUEMESA CCD camera system.
2.7 Confocal microscopy
The PS and PC microcapsules were incubated with diannexin (20 μg/mL) for 30 min, followed by anti-His-Alexa 488 (1 μg/mL; MBL, Tokyo, Japan) in the dark, to determine the binding capability of diannexin to PS. Afterward, the samples were washed and tested using a Leica SP8 confocal microscope. Platelet-rich plasma stimulated with A23187 or PBS (control) was centrifuged at 500× g for 10 min to pellet the platelets, which were washed and incubated with diannexin (20 μg/mL) for 30 min and labeled with anti-His-Alexa 488 for detection to test the binding of diannexin to PS exposed at the cell surface.
2.8 Animal study
Paired female mice were housed with males (female:male = 2:1) for one night, and the following morning, the females were examined for the presence of a vaginal plug. The day a vaginal plug was first noted was designated as day 0.5 post-coitum (p.c.). The pregnant mice were then randomly categorized into three equivalent groups: control group, PE-model group, and diannexin group. The pregnant mice were injected with 200 μL of mouse-derived EMPs into the tail vein on days 8.5 and 10.5 p.c. to establish a PE model. An equal volume of PBS was injected for the control. Mice were simultaneously injected with diannexin (500 mg/kg) intravenously from days 8.5 to 11.5 p.c. daily to explore the potential role of diannexin. On days 8.5 and 10.5 p.c., diannexin was injected 1 h prior to the injection of EMPs, and PBS as the control was injected 1 h prior to the EMP injection in the PE model group. The treatment procedure was displayed in Fig. S3.
2.9 Ex vivo analysis
Blood pressure was measured at least five times per mouse on day 12.5 p.c. using a noninvasive mouse tail-cuff blood pressure analysis system (Softron, Softron Biotechnology, Japan). Before the experiment, the mice were trained for 1 week to acclimate to the tail-cuff procedure. Systolic, diastolic, and mean blood pressure were recorded. Spot urine samples were collected on day 12.5 p.c., and the protein-to-creatinine ratio was analyzed using an automatic biochemical analyzer (AU5800, Bechman Coulter). After urine was collected, the mice were anesthetized with pentobarbital sodium (50 mg/kg, intraperitoneal injection). Blood was collected by cardiac puncture, followed by organ collection. Embryo viability was assessed using a visible heart beat or pulsatile blow flow from the umbilical cord. The developmental stage was evaluated according to the Theiler Staging Criteria. Placentae were separated from the embryos. Half of the placental tissues were fixed in 4% buffered formalin, and the other half was frozen in liquid nitrogen.
2.10 Enzyme-linked immunosorbent assay (ELISA)
The procoagulant activity of plasma MPs was tested using the Zymuphen MP Activity ELISA Kit (#521096, Hypen BioMed, France). Mouse plasma soluble fms-like tyrosine kinase 1 (sFlt-1) concentration was measured using a mouse VEGF R1/Flt-1 ELISA kit (R&D Systems, USA). The IL-1β protein level in mouse placenta was determined using a mouse IL-1β ELISA kit (KE10003, Proteintech Group, Inc., China), whereas IL-1β in the culture supernatant of BeWo or SWAN-71 cells was determined using a QuantiCyto Human IL-1β ELISA kit (NeoBioscience, China). All ELISA analyses were performed according to the manufacturers’ instructions.
2.11 Western blot
Tissues or cells were homogenized and lysed with radioimmunoprecipitaion assay lysis buffer (Beyotime, China) in the presence of a proteinase inhibitor cocktail (Beyotime). The protein samples were loaded by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and electroblotted onto polyvinylidene difluoride membranes. The membranes were blocked with 5% bovine serum albumin for 1 h at room temperature, probed with primary antibodies, and then incubated with HRP-labeled secondary antibodies. The immunoblots were enhanced by Immobilon Western Chemiluminescent HRP Substrate (Merck Millipore, USA), visualized using a Tanon 5200 Multi Automatic Chemiluminescence System (Tanon, China), and analyzed using ImageJ software.
2.12 Immunohistochemistry
The formalin-fixed tissues of human and mice placentae were detected by immunohistochemistry. Briefly, endogenous peroxidase activity was quenched by 10 min of incubation in a 3% hydrogen peroxide/methanol buffer, and antigen retrieval was conducted by incubating the slides in sodium citrate buffer (pH 6.0) at 100 °C for 2 min. The slides were blocked with normal goat serum and incubated with primary antibodies overnight at 4 °C. The sections were incubated with secondary antibodies for 30 min at 37 °C. Color was developed using 3,3′-diaminobenzidine tetrahydrochloride, followed by nuclear counterstaining with hematoxylin. Positive controls were established according to the antibody instructions, and the blank control was established by replacing primary antibody with PBS. A semiquantitative scoring system was used to evaluate protein expression by two independent analysts. The final immunostaining score of each section was determined by multiplying the intensity score (0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining) by the percentage of stained trophoblast cells (< 10% = 0; 10%–25% = 1; 26%–50% = 2; 51%–75% = 3; and 76%–100% = 4).
2.13 Measurement of mouse serum cytokine content
Serum cytokines were measured and quantified using the LEGENDplex™ mouse inflammation panel according to the manufacturer’s instructions (Biolegend, #740446). All data were collected on a BD FACSCanto II and analyzed using the LEGENDplex™ software (BioLegend).
2.14 Apoptosis detection and cell counting kit 8 (CCK8) detection
Apoptosis assays and CCK8 detection were performed to assess the potential side effects of diannexin on BeWo cells. Cell apoptosis was detected via flow cytometry with an annexin V–FITC/propidium iodide (PI) apoptosis assay kit (Zoman Biotechnology Co., Ltd., Beijing, China). Briefly, the cells treated with diannexin (40 nmol/L) or an equal volume of PBS were harvested, resuspended at a density of 1×105 cells/tube, and centrifuged. Then, the samples were resuspended with 500 μL of 1× binding buffer, added with 5 μL of annexin V–FITC and 10 μL of PI, and incubated in the dark for 15 min at room temperature. The cells positive for annexin V (Q2 and Q4) were considered to be undergoing early and late apoptosis. In all cases, 10 000 events/tube were acquired. For determining the cell viability, BeWo cells were seeded onto 96-well plates in triplicate at a density of 1×104 cells/well, incubated for 24 h, and then 10–80 nmol/L diannexin was added to the wells. After secondary incubation for 24 h, 10 μL of CCK8 solution was added and further incubated for 2 h according to the CCK8 kit’s instructions (Dojindo, China). Optical density (OD) at 450 nm was measured, and cell viability was calculated as follows: (OD value of experimental groups − OD value of blank groups)/(OD value of control groups − OD value of the blank groups).
2.15 Statistical analysis
Data were analyzed using GraphPad Prism 5 software (GraphPad Software, Inc., San Diego, CA, USA) and SPSS (v.26.0 IBM, USA). Chi-square test was used for comparing categorical variables, and Student’s t-test, paired t-test, Spearman’s correlation, and ANOVA were performed as appropriate. P < 0.05 was considered significant.
3 Results
3.1 Elevated circulating MPs and increased NLRP3 inflammasome activation in patients with PE
The clinical characteristics of the participants are shown in Table S1. The number of circulating MPs and their procoagulant activity were significantly higher in patients with PE relative to the levels observed in healthy pregnant women (P < 0.05, Fig.1 and 1B). Additionally, the correlation analysis between the number of MPs and their procoagulant activity in healthy pregnant women and patients with PE revealed a positive correlation between MP number and their procoagulant activity. Moreover, compared with healthy pregnant women, the procoagulant activity of MPs increased more substantially as their number increased in women with PE; this result suggests a stronger procoagulant activity of MPs in women with PE (Fig. S2). Furthermore, the evaluation of NLRP3 inflammasome activation in placental tissues indicated a higher expression of NLRP3, cleaved caspase 1, and cleaved IL-1β in the placentae of patients with PE than in healthy pregnant women (P < 0.05, Fig.1 and 1D), which reflects inflammasome activation in the placentae of patients with PE. However, no remarkable differences in pro-caspase 1 and pro-IL-1β levels were observed. Immunohistochemical analysis showed that NLRP3, caspase 1, and IL-1β were expressed by syncytiotrophoblasts, and the immunostaining intensities of these targets were significantly higher in the placentae of patients with PE than in healthy pregnant women (P < 0.05, Fig.1 and 1F).
3.2 Recombinant protein diannexin exclusively binds to PS
The synthesized recombinant protein diannexin has a molecular weight of 61.54 kDa and features a His-tag (Fig. S1). PS microcapsules, PS-exposing cells, and PS-exposing MPs were constructed to assess the binding of diannexin to PS. The PS microcapsules and PC microcapsules were ~500 nm in diameter according to transmission electron microscopy evaluation and had no differences in shape (Fig.2). The PS microcapsules (not PC microcapsules) could be visualized by green fluorescence at the surface after co-incubation with diannexin and a green fluorescence-labeled anti-His antibody. This result demonstrates that diannexin bound specifically to PS (Fig.2). When cells are activated, PS evaginates and is exposed at the outer surface [
10]. The inner leaflet of platelets is rich in PS; therefore, platelets were stimulated to construct PS-exposing cells. Confocal microscopy revealed that the stimulated platelets showed a stronger mean fluorescence intensity compared with the unstimulated platelets (
P < 0.01, Fig.2 and 2D), which indicates that diannexin could bind to the PS exposed on the outer leaflet of the membrane. Next, the platelets and EA.Hy-926 cells were activated using agonists to generate PMPs and EMPs, respectively. The MPs were treated with FITC–annexin V only or additionally with diannexin to test the interaction of diannexin and PS-exposing MPs. Flow cytometric analysis indicated that the number of MPs was significantly reduced in the diannexin-treated group (
P < 0.05, Fig.2), which suggests that diannexin successfully competed with annexin V to bind PS.
3.3 Diannexin can inhibit MP generation from activated cells
Platelets and EA.Hy-926 cells were pretreated with different diannexin doses, and then with stimulators to determine whether diannexin can inhibit MP release from activated cells. The number of MPs present in the supernatant was measured. Compared with the control, the number of PMPs and EMPs considerably increased after stimulation. However, the number of PMPs generated by platelets or EMPs generated by EA.Hy-926 cells pretreated with diannexin were reduced in a dose-dependent manner (P < 0.05, Fig.2 and 2G).
3.4 Diannexin ameliorates the PE-like phenotype induced by MPs in pregnant mice
Whether diannexin can prevent PE by lowering MP level was explored. Proinflammatory and procoagulant MPs can trigger NLRP3 inflammasome activation within trophoblast cells, which is required and sufficient for developing a PE-like phenotype in pregnant mice [
14]. The results showed that the MP-injected pregnant mice developed the characteristic hallmarks of PE defined by elevated blood pressure, proteinuria, and increased plasma sFlt-1 levels on day 12.5 p.c. (
P < 0.05, Fig.3–3C). In the PE model group, the surviving embryos were smaller, less developed (impaired forelimbs and reduced retinal pigmentation as shown in Fig. S3), and had a reduced placental diameter (
P < 0.05, Fig.3–3G). By contrast, prophylactic diannexin administration protected normal pregnant mice from MP-induced PE as indicated by the substantially reduced blood pressure and plasma sFlt-1 levels and improved pregnancy outcomes, such as improved embryo and placental development.
3.5 Diannexin lowers the levels of circulating MPs and placental NLRP3 inflammasome activation in PE-like symptom mice
The levels of MPs and placental NLRP3 inflammasome activation are increased in patients with PE, and similar symptoms were observed in the PE model mice. Therefore, we explored whether diannexin can prevent the PE-like phenotype by inhibiting the expression of proinflammatory factors. The results demonstrated that the diannexin-treated group showed significantly lower levels of circulating MPs and EMPs compared with the PE model mice, which suggests that diannexin inhibited the release of MPs (P < 0.05, Fig.4 and 4B). Additionally, diannexin treatment lowered the serum concentration of IFN-γ, IL-1β, and IL-6 relative to those observed in the PE model group (P < 0.05, Fig.4–4E). Changes in other cytokines are shown in Fig. S4.
Changes in NLRP3 inflammasome activation in the placentae were also evaluated. Although the expression of pro-caspase 1 and pro-IL-1β did not significantly decrease, we observed a decrease in the expression of inflammasome markers (NLRP3, cleaved caspase 1, and cleaved IL-1β) in placental tissues from the diannexin-treated mice (P < 0.05, Fig.4 and 4G). Furthermore, the immunohistochemical analysis of NLRP3 and caspase 1 showed that these proteins were expressed in trophoblast cells from the placentae. Compared with the PE model group, diannexin treatment lowered the intensity of NLRP3 expression (P < 0.05), although caspase 1 levels showed no significant difference (Fig.4–K). Moreover, the ELISA analysis of IL-1β levels in the placentae revealed a lower expression following diannexin treatment (P < 0.05, Fig.4), which suggests that diannexin inhibited NLRP3 inflammasome activation to prevent PE.
3.6 Diannexin can reduce NLRP3 inflammasome activation induced by MPs in trophoblast cells
BeWo and SWAN-71 cells were exposed to 200 EMPs/μL to further assess the role of diannexin in inhibiting placental inflammasome activation. The levels of NLRP3 and cleaved caspase 1 were significantly higher in the group not treated with diannexin, whereas the trophoblast cells pre-incubated with diannexin showed decreased inflammasome activation in a dose-dependent manner (P < 0.05, Fig.5, 5B, 5E, and 5F) with a similar result observed for IL-1β levels in the culture supernatant (P < 0.05, Fig.5 and 5G). Additionally, a dose-dependent diannexin-mediated reduction in trophoblast MPs in the supernatant was observed (P < 0.05, Fig.5 and 5H), which suggests that diannexin inhibited trophoblast activation according to reduced trophoblast MP release and inflammasome activation. Before assessing diannexin function, we assessed its effect on cell viability and apoptosis. CCK8 assays revealed no remarkable effect on cell viability up to a concentration of 70 nmol/L, at which point we observed an 8.5% decrease in viability (Fig. S5). Furthermore, apoptosis detection showed that 40 nmol/L diannexin did not induce cell apoptosis (Fig. S6). These results suggested that diannexin did not impact the normal biological processes of the cells.
4 Discussion
The thromoboinflammatory response is involved in PE development and progress. Therefore, restraining excessive inflammation and a hypercoagulable state should be an effective approach to prevent PE occurrence and development. MPs participate in PE pathogenesis by amplifying the thromboinflammatory response. In this study, we presented a strategy for preventing PE by reducing MP-induced thromboinflammatory response. The recombinant diannexin reduced the generation of MPs from activated cells, and diannexin ameliorated the PE-like phenotype induced by MPs in pregnant mice. Additionally, decreased MP levels and reduced NLRP3 inflammasome activation were observed in the diannexin-treated mice, as well as in trophoblast cells. The results indicate that diannexin could play a role in preventing PE by reducing the generation of MPs and inhibiting the inflammasome activation in the placenta.
Previous studies reported changes in the concentration and cargo of MPs during PE development through the enhancement of proinflammatory and procoagulatory activities. Other studies revealed that compared with healthy pregnant women, patients with PE show a higher level of MPs, which agreed with the observations in the present study [
11–
13]. However, some reports suggested that circulating MP levels do not substantially differ between patients with PE and healthy pregnant women, which might be attributed to inconsistencies in testing standards or the use of different targets (e.g., targeting annexin A5-negative MPs instead of annexin A5-positive MPs or PS-exposing MPs) [
24,
25]. Currently, the pathogenic role of annexin A5-negative MPs remains unclear; however, Omatsu
et al. [
26] found that injecting lipid microvesicles containing PS promoted a PE-like phenotype in pregnant mice, which suggests that PS exposed on the MP surface might be involved in PE occurrence. Furthermore, in the study, the increased PS-exposing MPs in patients with PE demonstrated higher levels of procoagulant activity. In addition to the change in their quantity, MP cargo in patients with PE also changed. Proteomics analysis revealed the involvement of PE-related differentially-expressed proteins related to immune response, coagulation, oxidative stress, and apoptosis [
27]. Additionally, MPs carry multiple receptors and adhesion molecules that originate from mother cells; mediate cell–cell interactions; and interact with and activate endothelial cells, platelets, and other cells to release more MPs to amplify the thromboinflammatory response [
28]. Shomer
et al. [
24] found that compared with those from healthy pregnant women, the MPs extracted from the patients with PE affected endothelial cell migration and tubule formation and increased endothelial cell apoptosis. Moreover, Tong
et al. [
29] reported that MPs from the placentae of patients with PE carried sFlt-1, which could bind to vascular endothelial growth factor and contribute to endothelial cell dysfunction. Other studies showed that MPs from patients with PE can activate monocytes and promote the release of proinflammatory cytokines, such as TNF-α, IL-1β, and IL-6, to enhance the maternal inflammatory response [
30,
31]. Kohli
et al. [
14] injected MPs into pregnant mice and found that EMPs or PMPs promote PE by accumulating activated maternal platelets within the placenta and activating NLRP3 inflammasome in trophoblast cells. Furthermore, Han
et al. [
32] revealed that placenta-derived MPs induce a PE-like phenotype in pregnant mice. Therefore, in the present study, we injected MPs into pregnant mice to establish a PE-like model.
When cells are in the resting state, the membrane bilayers maintain an asymmetric distribution of phospholipids, where the outer layer is enriched with PC and sphingomyelin, and the inner layer is predominantly enriched with PS. Upon stimulation, the level of cytosolic Ca
2+ increases, which promotes PS evagination to expose the outer layer and lead to the destruction of membrane asymmetry [
9]. Membrane instability caused by negative charges at the surface and a broken cytoskeleton promotes membrane blebbing [
9,
10]. Therefore, covering the PS might reduce membrane instability and inhibit the release of MPs. Indeed, Kenis
et al. [
19] found that annexin A5 can form a 2D lattice structure to cover PS and bend the membrane patch into the cell to inhibit membrane blebbing. Additionally, Ravassa
et al. [
33] found that the binding of annexin A5 to PS can internalize PS-expressing membrane patches and tissue factor and inhibit coagulation. Therefore, covering PS might inhibit membrane blebbing and prevent MP release. However, annexin A5 (35.7 kDa) has a short half-life (< 20 min) owing to its low molecular weight [
34], which makes it unsuitable for clinical applications. Recombinant protein diannexin, a homodimer of annexin A5, has a higher molecular weight and longer half-life
in vivo (~6–7 h) relative to annexin A5 [
35]. In the present study, we showed that the synthesized diannexin demonstrated high affinity for PS and bound exclusively to PS microcapsules, as well as to the PS exposed on the surface of activated platelets. Furthermore, diannexin competed with annexin A5 to bind to PS-exposing MPs. Rand
et al. [
21] and Kuypers
et al. [
35] reported that diannexin can compete with annexin A5 to bind to PS on the cell surface and inhibit the activity of prothombinase complexes. In the present study, we found that diannexin treatment reduced MP release by covering exposed PS. Combes
et al. [
36] found that fluorescence-labeled diannexin bound to endothelial cells activated by TNF-α. We observed that diannexin decreased MP release considerably after an overnight incubation, which suggests that diannexin might inhibit endothelial vesiculation by binding to exposed PS.
Diannexin can reduce the release of MPs from activated cells. Therefore, the effect of diannexin was evaluated in an MP-induced PE-like mouse model. Compared with the PE-model group, the diannexin-treated group showed alleviated symptoms and remarkably lower levels of MPs relative to the PE-model group. NLRP3 inflammasome activation in the placentae was reduced after diannexin intervention; thus, we speculated that diannexin might prevent PE by reducing MP levels and the MP-induced inflammatory response. Interestingly, Zhou
et al. [
37] found that the microvesicle-scavenging factor, lactadherin, can promote microvesicle clearance and prevent coagulopathy; therefore, it can improve the survival of cases with severe traumatic brain injury. Additionally, Han
et al. [
32] employed lactadherin to reduce MP levels, which prevented a PE-like condition. These findings suggest that lower MP levels might inhibit the amplification of thromboinflammatory response and therefore prevent PE. Moreover, diannexin reportedly protected against organ transplant ischemia–reperfusion injury [
20,
34,
38–
42]. Hashimoto
et al. [
41] showed that the diannexin covering PS in endothelial cells reduced ischemia–reperfusion injury and increased the survival rate of transplanted organs by inhibiting platelet and leukocyte adhesion, tissue inflammation, and coagulation. Teoh
et al. [
20,
42] showed that hepatocytes exposed to MPs isolated from the circulation during ischemia–reperfusion injury increase mitochondrial membrane permeability, activate platelets, and induce neutrophil migration to result in secondary microcirculatory inflammation and coagulation. However, all these processes were blocked by coating the MPs with diannexin. Considering that the placentae also have ischemia–reperfusion injury related to PE pathogenesis, diannexin might alleviate the ischemia–reperfusion injury of the placentae, inhibit oxidative stress and inflammation, and therefore prevent PE.
Consistent with other studies, we found that NLRP3 inflammasome was remarkably activated in patients with PE along with increased expression of NLRP3, cleaved caspase 1, and cleaved IL-1β in the placentae [
43,
44]. The inhibition of NLRP3 inflammasome activation can reduce the release of mature IL-1β [
44]. In the present study, we observed that MP-induced PE in pregnant mice increased NLRP3 inflammasome activation and mature IL-1β levels. However, diannexin intervention improved PE symptoms and reduced NLRP3 inflammasome activation, which suggests that decreasing NLRP3 inflammasome activation might prevent PE. Indeed, Kohli
et al. [
14] found that
Nlrp3-knockout mice did not present a PE-like phenotype. This finding suggests NLRP3 inflammasome activation plays an important role in PE progression. Additionally, Qiu
et al. [
45] found that MPs can cause intracellular NLRP3 inflammasome activation via the Toll-like receptor 4-mediated PI3K/Akt signaling pathway. Our results showed that increased NLRP3 inflammasome activation and IL-1β levels were observed in mice without diannexin treatment, whereas trophoblast cells preincubated with diannexin showed a dose-dependent reduction in inflammasome activation.
In vitro, the supernatant from diannexin-treated trophoblasts showed a lower number of trophoblast-derived MPs, which suggests a reduced trophoblast activation.
5 Conclusions
MPs play an important role in PE pathogenesis. Here, a possible strategy for preventing PE was identified by reducing MPs and MP-induced thromboinflammatory response. The results demonstrated that diannexin indeed reduced MP release by binding to PS exposed on the cell surface. Furthermore, diannexin prevented PE-like symptoms in a pregnant mouse model by inhibiting the MP-induced inflammasome activation in the placentae. These findings suggest that diannexin might represent an effective therapeutic strategy for preventing PE.
PDF
(4841KB)
