Thrombin Contributes to Anti-myeloperoxidase Antibody Positive IgG-Mediated Glomerular Endothelial Cells Activation Through SphK1-S1P-S1PR3 Signaling

Background: Activation of coagulation system plays an important role in antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) pathogenesis. Thrombin, generated during coagulation could disrupt endothelial barrier integrity through protease-activated receptor 1 (PAR1). Our previous study found that sphingosine-1-phosphate (S1P) contributed to myeloperoxidase (MPO)-ANCA-positive IgG-induced glomerular endothelial cell (GEnC) activation through a S1P receptor (S1PR)-dependent route. In recent years, S1P signaling was reported to be involved in thrombin effects on endothelial cells. This current study investigated whether the interaction between thrombin-PAR and S1P-S1PR signaling contributed to MPO-ANCA-positive IgG-induced GEnC dysfunction. Methods: The effect of thrombin on GEnC activation was analyzed from three aspects. First, morphological alteration of GEnCs was observed. Second, permeability assay was performed to determine GEnC monolayer activation quantitatively. Third, endothelin-1 (ET-1) levels were measured. Expression levels of sphingosine kinases (SphKs) and S1PRs were detected. In addition, antagonists of PAR1 and S1PR3 were employed to determine their roles. Eventually, PAR1 and tissue factor (TF) expression levels as well as TF procoagulant activity were analyzed. Results: Thrombin induced further damage of tight junction, increase in endothelial monolayer permeability as well as upregulation of ET-1 levels in GEnCs stimulated with MPO-ANCA-positive IgG. Blocking PAR1 downregulated ET-1 levels in the supernatants of GEnCs treated by thrombin plus MPO-ANCA-positive IgG. Expression levels of SphK1, S1PR3 increased significantly in GEnCs treated with thrombin plus MPO-ANCA-positive IgG. S1P upregulated PAR1 and TF expression, and enhanced procoagulant activity of TF in MPO-ANCA-positive IgG-stimulated GEnCs. Conclusion: Thrombin synergized with SphK1-S1P-S1PR3 signaling pathway to enhance MPO-ANCA-positive IgG-mediated GEnC activation.

Sphingosine-1-phosphate (S1P) is a bioactive sphingolipid metabolite and produced by phosphorylation of sphingosine by sphingosine kinases (SphKs). S1P is the ligand for five Gprotein-coupled receptors (GPCRs) named S1PR1-5 (8). S1P and S1PRs participate in the pathogenesis of a variety of vascular inflammatory conditions including ischemia-reperfusion injury, atherosclerosis and sepsis (9)(10)(11). In recent years, clinical trials that targeted S1PRs for autoimmune diseases have attracted wide interest. Of note, FTY720 (Fingolimod, Gilenya, Novartis), a functional antagonist of S1PR1, 3, 4, and 5, has already been approved and used in treating multiple sclerosis (12)(13)(14). Moreover, cumulating evidences supported a vital role of FTY720 in endothelial barrier enhancement both in vivo and in vitro (15)(16)(17). In our previous studies, we found that the circulating levels of S1P and the renal expression of S1PRs correlated with renal involvement and disease activity of AAV. In addition, it was found that S1P enhanced MPO-ANCA-positive IgG-induced GEnC activation through S1PR2-5 and RhoA signaling pathway (18)(19)(20). All these studies indicated a pathogenic role of S1P in AAV.
Although the pathogenesis of AAV is not yet fully clear, the interaction among ANCA, neutrophils and complement activation is of vital importance in the development of this disease [reviewed by Chen et al. (21)]. In recent years, more and more evidence has suggested that activation of coagulation system may also play an important role. Patients with AAV are in a hypercoagulable state, with an increased risk of developing venous thromboembolic events (22,23). Moreover, the interaction between coagulation and complement system also contributes to the pathogenesis of glomerular capillary tuft infarction and to the increased frequency of thromboembolic events in AAV. Some serine proteases from the coagulation cascade, in particular plasmin and thrombin, can directly activate C3 and C5, independent of the traditional C3/C5 convertase (24,25). C5a-primed neutrophils produce tissue-factor-expressing microparticles and neutrophil extracellular traps (NETs) after stimulation with ANCAs, which subsequently activate the coagulation system (26). Platelets are activated via thrombin-PARs pathway and can activate the alternative complement pathway in AAV (27).
The coagulation system is initiated in two distinct mechanisms: the contact pathway and the tissue factor (TF) pathway. Both pathways result in the generation of thrombin, the best-characterized activator of protease-activated receptors (PARs) (28). PARs are a family of G protein-coupled receptors including 4 members named PAR1-4. PAR1 is the major effector of thrombin signaling in most cell types including endothelial cells. Thrombin activates PAR1 by catalyzing the cleavage of the Arg41-Ser42 peptide bond on the N-terminal extracellular domain of the receptor (29). It was reported that thrombin-activated PAR1 could induce disruption of endothelial barrier integrity (30).
Thrombin effects in endothelial cells involve S1P signaling. According to Tauseef et al. SphK1-S1P-S1PR1 signaling could counteract the detrimental effect of thrombin-PAR1 signaling on endothelial barrier function. On the one hand, thrombinactivated-PAR1 interrupts endothelial barrier integrity via Rho signaling pathway; on the other hand, thrombin also induces expression of SphK1 and increases S1P generation, which in turn transactivates S1PR1 leading to the activation of Rac1 signaling pathway. This effect improves endothelial integrity to counteract and limit thrombin-induced endothelial damage and vascular leakage (31). However, some other studies revealed a synergistic effect of S1P on thrombin-induced endothelial dysfunction, including enhanced NF-κB binding activity and TF expression in endothelial cells (32,33). Given the potential effect of thrombin-PAR and SphK-S1P-S1PR signaling on regulating endothelial barrier function, our current study aimed to investigate whether the interaction between thrombin-PAR and SphK-S1P-S1PR signaling participated in MPO-ANCA-positive IgG-induced GEnC dysfunction.

Cell Culture
Primary human glomerular endothelial cells (GEnC; ScienCell, San Diego, CA, USA) were cultured in endothelial cell basal medium (ECM) (ScienCell San Diego, CA, USA) supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin and 1% endothelial cell growth factor. Cultures were grown in an atmosphere of 5% CO 2 at 37 • C. After starving in ECM with additional 0.5% FBS for 8 h, GEnC in selected wells were washed with phosphate buffered saline (PBS) and then stimulated with thrombin (Sigma, Darmstadt, Germany), MPO-ANCApositive IgG, normal IgG or 2 µmol/L S1P (Sigma, Darmstadt, Germany), which was comparable to the levels of circulating S1P in AAV patients at active stage, as demonstrated by our previous study (18).

Preparation of Immunoglobulin (Ig)Gs
Preparation of IgGs was performed according to the methods described previously (34). MPO-ANCA-positive IgGs and normal IgGs and were prepared from plasma exchange liquid of eight patients with active MPO-ANCA-positive primary small vessel vasculitis and plasma of six healthy donors, respectively. Then we further screened the prepared IgGs for the presence of anti-endothelial cell antibody (AECA) through an ELISA method described previously (35), and AECA-positive IgGs were excluded in our following experiments. Eventually, normal IgGs from plasma of five healthy donors and MPO-ANCA-positive IgGs from plasma exchange liquid of five AAV patients were included, respectively. Our research was in compliance with the Declaration of Helsinki and approved by the clinical research ethics committee of the Peking University First Hospital.

Measurements of GEnC Activation and Injury
Immunofluorescence Staining of Zonula Occludens-1 (ZO-1) and Vascular Endothelial (VE)-Cadherin As important markers for endothelial barrier function, the distribution of the tight junction scaffolding protein ZO-1 and adherens junction protein VE-cadherin were observed (36). After relevant treatment, GEnCs were washed in PBS and fixed with 4% formaldehyde for 30 min. Next, the GEnCs were permeabilized with 0.5% Triton X-100, washed and blocked with 5% BSA for 1 h at room temperature. After incubation with primary antibodies (ZO-1, dilution 1/100, Life, Carlsbad, CA, USA; VE-cadherin, dilution 1/200, Abcam, Cambridge, MA, USA) at 4 • C overnight and a thorough wash in PBS, the GEnCs were incubated with fluorescein isothiocyanate (FITC)-conjugated secondary antibodies (for the detection of ZO-1, dilution 1/200, Jackson ImmunoResearch, West Grove, PA, USA; for the detection of VEcadherin, dilution 1/500, Abcam, Cambridge, MA, USA) at 37 • C for 1 h. Eventually, the specimens were stained with 10 µg/ml 4' ,6diamidino-2-phenylindole (DAPI) and mounted with Mowiol. The immunofluorescence staining was photographed by a fluorescence microscope (Nikon Eclipse 90i, Nikon Instruments Inc., Tokyo, Japan). At least 10 visual fields per slide of GEnCs at ×400 were observed blindly. Image J software (National Institutes of Health, Bethesda, MD, USA) was used to evaluate the immunofluorescence staining of ZO-1 and VE-cadherin. Positive signals were quantified as signal intensity.

Permeability Assay
The permeability of GEnC monolayers was determined using Costar Transwell plate with 0.5-µm porous filters and FITClabeled BSA (Sigma-Aldrich, Darmstadt, Hessen, Germany), as described previously (37). GEnCs were grown on the upper chamber of Costar Transwell until confluent. The tracer protein FITC-albumin was added to the upper chamber after relevant stimulation. After incubation at 37 • C for 30 min, samples from both the upper and lower chambers were collected for fluorometric analysis. Fluorescent intensity (FI) was measured using a microplate fluorescence reader (Tristar TM LB941, Berthold, Germany) with filter settings of 485 nm (excitation) and 538 nm (emission). Eventually, these fluorescence readings were used for calculation of the permeability coefficient, which is indicative of vascular barrier disruption. The permeability coefficient was calculated according to the following formula: Permeability coefficient = FI (lower chamber) × 100% / (FI (upper chamber) + FI (lower chamber)).

Evaluation of Endothelium Activation by Endothelin-1 (ET-1) Quantification
As a biomarker of endothelial cell activation and injury (38), ET-1 levels in GEnC supernatants were measured using commercial ELISA kits (R&D, Minneapolis, MN, USA).

Detection of SphK1 by Western Blot
Samples were incubated for 10 min at 95 • C in loading buffer. Samples were then subjected to electrophoresis on 10% SDSpolyacrylamide gels and transferred to nitrocellulose membranes. The membranes were incubated with primary antibodies (for the detection of SphK1, dilution 1/1000, Abcam, Cambridge, MA, USA; for the detection of β-actin, dilution 1/1000, Santa Cruz, Dallas, TX, USA) followed by horseradish peroxidaseconjugated secondary antibodies (each diluted 1:2000; both from Proteintech, Chicago, IL, USA). Proteins were visualized on autoradiographic film using an ECL Plus Western blot detection system (GE Healthcare).

TF Procoagulant Activity Assay
To analyze TF procoagulant activity, a Cell Tissue Factor Assay Kit (Genmed Scientifics Inc, Wilmington, DE, USA) was used following manufacturer's instructions. GEnCs were lysed and 50 µg proteins and ∼2 × 10 6 cells of each sample were used. Samples were incubated with prothrombin complex (including Factor II, VII, IX, X) and CaCl 2 . Reaction was terminated by adding EDTA buffer. Eventually, we added a chromogenic substrate (Spectrozyme factor Xa) and measured the absorbance at 405 nm.

Statistical Analysis
SPSS version 13.0 (SPSS Inc., Chicago, IL, USA) was used to perform data analysis. Normality of the data was evaluated by kurtosis and skewness (both the absolute values were <3). Data was generally presented as mean ± standard deviation (SD) and compared by ANOVA followed by Bonferroni correction for multiple testing. P < 0.05 were considered statistically significant.

Thrombin Induces GEnC Morphological Alteration of GEnC Monolayers
Immunofluorescence staining of ZO-1 and VE-cadherin were performed to observe the structure of the tight junction and adherens junction in GEnCs, respectively. We found that compared with untreated cells, the application of thrombin or MPO-ANCA-positive IgG alone could disrupt tight junction and adherens junction structures ( (Figure 1). These data revealed that thrombin synergized with MPO-ANCA-positive IgG to exert damage effects on endothelial barrier integrity.

PAR1 Mediates the Thrombin-Induced Endothelial Dysfunction
GEnCs were pre-incubated with PAR1 antagonist RWJ for 15 min before stimulation with thrombin and MPO-ANCApositive IgG, and the ET-1 levels in the supernatants were measured. We found that the ET-1 levels reduced from 545.39 ± 15.06 pg/ml in the supernatants of GEnCs stimulated by thrombin and MPO-ANCA-positive IgG to 319.86 ± 19.07 pg/ml, upon pre-incubation with PAR1 antagonist RWJ (compared with that without the antagonist, P < 0.001, with the inhibition rate of 41.35 ± 3.50%) ( Figure 2C). These data revealed that PAR1 mediated ET-1 upregulation in thrombin and MPO-ANCA-positive IgG-treated GEnCs.

S1PR3 Mediates the Thrombin-Induced Endothelial Dysfunction
GEnCs were pre-incubated with S1PR3 antagonist TY for 15 min before stimulation with thrombin and MPO-ANCA-positive IgG, and the ET-1 levels in the supernatants were measured. We found that pre-incubation of GEnCs with TY significantly decreased ET-1 level in the supernatants of GEnCs stimulated by thrombin and MPO-ANCA-positive IgG (545.39 ± 15.06 pg/ml vs. 217.52 ± 18.99 pg/ml, P < 0.001, with the inhibition rate of 60.12 ± 3.48%) ( Figure 3B). These data revealed that S1PR3 activation was involved in thrombin-induced ET-1 upregulation in GEnCs in the presence of MPO-ANCA-positive IgG.
We also pre-incubated GEnCs with FTY720 before stimulated with thrombin and MPO-ANCA-positive IgG, and the ET-1 levels in the supernatants were measured. We found that the ET-1 levels reduced from 552.69 ± 20.46 pg/ml in the supernatants of GEnCs stimulated by thrombin and MPO-ANCA-positive IgG to 241.53 ± 21.22pg/ml, upon pre-incubation with FTY720 (compared with those without FTY720, P < 0.001, with the inhibition rate of 43.70 ± 3.84%) (Figure 3C).

PAR1 Expression Levels Are Elevated in GEnCs Upon Stimulation by S1P
GEnCs were stimulated with MPO-ANCA-positive IgG plus 2 µmol/L S1P, which was comparable to the levels of circulating S1P in AAV patients at active stage, as demonstrated by our previous study (12), and PAR1 expression levels in GEnCs were measured by qRT-PCR. It was found that compared with GEnCs stimulated by MPO-ANCA-positive IgG alone, the expression levels of PAR1 in GEnCs treated with S1P plus MPO-ANCApositive IgG increased significantly (1.50 ± 0.27 vs. 1.01 ± 0.06, P < 0.01) (Figure 4).

DISCUSSION
In our current study, we demonstrated that thrombin could enhance MPO-ANCA-positive IgG-induced GEnC activation via PAR1, and thrombin could activate SphK1-S1P-S1PR3 axis in GEnCs in the presence of MPO-ANCA-positive IgG. At the same time, S1P, at pathophysiological concentration in active AAV patients, could induce PAR1 expression as well as enhance both expression level and procoagulant activity of TF in MPO-ANCA-positive IgG-treated GEnCs, which may further activate the coagulation system, thus forming a vicious loop (Figure 6).
Anti-MPO antibody could cause activation of GEnCs by recognizing moesin even though MPO is not expressed in endothelial cells (41). Moesin, whose full name is membraneorganizing extension spike protein, shares certain similar sequences with those on the N-terminal region of the MPO heavy chain (7). Binding of anti-MPO antibody to moesin was able to increase permeability and to up-regulate adhesion molecules of human GEnCs (42). Recently, it was reported that thrombin was able to induce phosphorylation of moesin within seconds (43). Likewise, S1P could also cause acute and potent moesin activation (44,45). Therefore, we speculate that moesin recognized by MPO-ANCA could be further activated by thrombin or S1P, which might cause enhanced GEnC activation in vitro. However, anti-PR3 antibody might induce endothelial cells dysfunction through different mechanisms. According to the study by Le Roux S et al., anti-PR3 antibodies could induce a potent inhibitor of vascular endothelial growth factor named soluble Flt1 to release from monocytes rather than endothelial cells, therefore leading to an anti-angiogenic state that hinders endothelial repair in AAV (46).
In our current study, we found that thrombin could activate SphK1-S1P-S1PR3 axis, thrombin induced upregulation of SphK1 expression levels in GEnCs in the presence of MPO-ANCA-positive IgG (confirmed by both Western blot and PCR), therefore promoting the generation of S1P. However, the exact involvement of PAR1 during this process remains to be determined. According to the study by Parker et al. thrombin could cause activation of the small GTPase RhoA in vivo (47). This is of particular interest, because small GTPases are confirmed to play critical roles in mediating signaling responses of the S1PR (48), and our previous work also demonstrated that RhoA activated by S1PR2-5 dominated the S1P-induced MPO-ANCA-positive IgG-mediated GEnC activation (20). Activation of RhoA signaling induces endothelial barrier disruption by remodeling cytoskeleton and enhancing the formation of contractile stress fibers which are connected to junctions and generate pulling forces within neighboring cells, thus destabilizing cell contact and forming "discontinuous" adherens junctions and tight junctions (49). All these evidences not only suggest a mutual interaction of S1PR-initiated signaling and regulation of S1P synthesis, but also provide clues to the synergistic effect of thrombin and SphK-S1P-S1PR3 signaling on endothelial barrier dysfunction.
Under homeostatic conditions, high levels of S1P in circulation (∼1 µM) are a result of its release from endothelial cells and red blood cells, while platelets may only release large amounts of S1P upon platelet activation when endothelial cells are damaged. Thromboxane plays a crucial role in S1P release from human platelets. The coagulation factors thrombin and FXa interact with local S1P availability and its cellular effects at multiple levels (8). A recent study by Campos et al. demonstrated that in rodent models of stroke, the functional S1P receptor antagonist fingolimod could enhance blood-brain barrier integrity and reduce infarct size, indicating S1P as a potential link between coagulation and inflammation system (50). Our previous studies illustrated that the renal expression of S1PRs correlated with both inflammatory and coagulation parameters among AAV patients, and S1P contributed to MPO-ANCA-positive IgG induced GEnC activation through S1PR2-5 and RhoA signaling pathway (18)(19)(20). In our current study, we found that thrombin-PAR could interact with SphK-S1P-S1PR signaling pathway to induce GEnC dysfunction in the presence of MPO-ANCA-positive IgG, and S1P could enhance both expression level and activity of TF in MPO-ANCA-positive IgGtreated GEnCs, thus further activating the coagulation system. Therefore, we speculate that in AAV, S1P might act as a mutual link between inflammation and coagulation system. Blockade of this Sphk1-S1P-S1PR3 signaling pathway may be critical for attenuating the pathological processes associated with overactivation of both coagulation system and inflammation system in AAV.

CONCLUSIONS
In conclusion, thrombin is able to enhance MPO-ANCA-positive IgG-mediated GEnC activation via Sphk1-S1P-S1PR3 signaling pathway. These findings are helpful to figure out the linking role of S1P between coagulation and inflammation in AAV, thus provide potential clues for intervention strategies.

ETHICS STATEMENT
Informed consent from each participant was obtained. The study was conducted in line with the Declaration of Helsinki and was approved by the ethics committees of Peking University First Hospital.

AUTHOR CONTRIBUTIONS
X-JS conducted the experiments, analyzed the data, and drafted the manuscript. M-HZ involved in its design, assisted with interpretation of data, and provided suggestion for revising the manuscript. MC conceived of the study, participated in the revision of the manuscript, and provided final approval of the version of the submitted manuscript. All authors read and approved the manuscript. Frontiers in Immunology | www.frontiersin.org