Placental multipotent mesenchymal stromal cell-derived Slit2 may regulate macrophage motility during placental infection

Abstract

Slit proteins have been reported to act as axonal repellents in Drosophila; however, their role in the placental microenvironment has not been explored. In this study, we found that human placental multipotent mesenchymal stromal cells (hPMSCs) constitutively express Slit2. Therefore, we hypothesized that Slit2 expressed by hPMSCs could be involved in macrophage migration during placental inflammation through membrane cognate Roundabout (Robo) receptor signaling. In order to develop a preclinical in vitro mouse model of hPMSCs in treatment of perinatal infection, RAW 264.7 cells were used in this study. Slit2 interacted with Robo4 that was highly expressed in RAW 264.7 macrophages: their interaction increased the adhesive ability of RAW 264.7 cells and inhibited migration. Lipopolysaccharide (LPS)-induced CD11bCD18 expression could be inhibited by Slit2 and by hPMSC-conditioned medium (CM). LPS-induced activation of p38 and Rap1 was also attenuated by Slit2 and by hPMSC-CM. Noticeably, these inhibitory effects of hPMSC-CM decreased after depletion of Slit2 from the CM. Furthermore, we found that p38 siRNA inhibited LPS-induced Rap1 expression in RAW 264.7 cells, indicating that Rap1 functions downstream of p38 signaling. p38 siRNA increased cell adhesion and inhibited migration through reducing LPS-stimulated CD11bCD18 expression in RAW 264.7 cells. Thus, hPMSC-derived Slit2 may inhibit LPS-induced CD11bCD18 expression to decrease cell migration and increase adhesion through modulating the activity and motility of inflammatory macrophages in placenta. This may represent a novel mechanism for LPS-induced placental infection.

cell–cell interactions, placenta, pregnancy

Introduction

Slit secretory proteins were originally identified as axonal repellents, which regulated the migration of neurons and axons in Drosophila by binding to membrane cognate Roundabout (Robo) receptors (Dickson and Gilestro, 2006). The secreted Slit proteins have three isoforms (Slit1–3), and Robo proteins have four isoforms (Robo1–4). Immune cells, including monocytes, macrophages and monocyte-derived dendritic cells, express Robo receptors (Guan et al., 2003; Prasad et al., 2007; Geutskens et al., 2010). However, Slit/Robo signaling functions in various biological processes, such as kidney development (Grieshammer et al., 2004), myogenesis (Kramer et al., 2001), leukocyte chemotaxis (Prasad et al., 2007; Geutskens et al., 2010), angiogenesis in endothelial cells (Jones et al., 2008), metastasis of carcinoma cells (Dallol et al., 2002) and trophoblast migration (Li et al., 2017). Therefore, the roles of Slit/Robo signaling in placental infection and the involved mechanisms must be investigated.

Escherichia coli is the main pathogen involved in placental infection, which leads to preterm births (Stoll et al., 2011). Lipopolysaccharide (LPS) of gram-negative bacteria is one of the main inflammatory pathogens, and its effects are primarily mediated by cells of the immune system, such as macrophages (Morrison and Ryan, 1979). LPS affects macrophage spreading (Wells et al., 2004). In subacute or chronic chorioamnionitis, the amniochorial placental tissue and intervillous space are characterized by a mixed infiltration of macrophages and neutrophils (Redline, 2012). Cellular and molecular mechanisms that modulate macrophage migration and activation at the feto–maternal interface during perinatal infections are unknown. Multipotent mesenchymal stromal cells (MSCs) express various growth factors and cytokines, including hepatic growth hormone, basic fibroblast growth factor, vascular endothelial growth factor, monocyte chemoattractant protein-1 and interleukin (IL)-6 (Kinnaird et al., 2004, Potapova et al., 2007). Human placental MSCs (hPMSCs) were found in the villous stroma, but their role in placental infection is still unknown. We demonstrated that hPMSCs produced IL-8 and IL-6, which attracted neutrophils from circulation to the placental tissue and thus prolonged their life span during infection (Chen et al., 2014). Here, we further hypothesize that hPMSCs produce Slit2, which modulates macrophage migration through Robo receptor signaling.

In placenta, Slit2 exerts anti-inflammatory effects through the inhibition of IL-6 expression and secretion (Lim and Lappas, 2015). During placental inflammation, leukocyte recruitment involves the coordination of leukocyte tethering, rolling, adhesion, transendothelial migration and chemotaxis (Herter and Zarbock, 2013). These processes involve integrin activation (Herter and Zarbock, 2013). The leukocyte-specific β2 integrin (CD18) subfamily, including αMβ2 (CD11bCD18), plays a pivotal role in mediating leukocyte adhesion and spreading in response to stimulation from bacterial products, cytokines or chemokines (Herter and Zarbock, 2013).

Rap1 is a small GTPase, which regulates integrin-dependent adhesion (Bos, 2005). p38 mitogen-activated protein kinase (MAPK) is a crucial component in stress-induced signal transduction pathways, which lead to inflammatory cytokine production (Tibbles and Woodgett, 1999). It is unclear whether Rap1 functions downstream or upstream of p38; moreover, the role of p38 signaling in Rap1 activation in macrophages treated with Slit2 is unknown.

The molecular link between hPMSCs and inflammation in placental infection has not been well established. We hypothesize that Slit2 produced by hPMSCs inhibits the expression of integrin adhesion molecules and the migration of macrophages after LPS stimulation. The signaling pathways involved in the expression of CD11b and CD18, namely p38 MAPK and Rap1 activation in RAW 264.7 cells, were studied.

Materials and methods

Cell culture and treatments

Placental tissue was obtained from the pregnant women after informed consent, and all experiments were approved by the Institutional Review Board of MacKay Memorial Hospital, Taipei. hPMSCs were isolated from clinically normal human term placentas (37–40 weeks of gestation) collected after cesarean section, as we described in a previous study (Chen et al., 2015). Briefly, we minced villous tissue from central placental cotyledons after trimming off the decidual layer and chorionic plate. Approximately 0.3 g of tissue was explanted onto a six-well culture plate (Greiner Bio-One, Kremsmünster, Austria) and cultured in Minimum Essential Medium-α (Hyclone, Logan, UT, USA) supplemented with 20% fetal bovine serum (FBS; Hyclone), basic fibroblast growth factor (4 ng/ml; Peprotech, Rocky Hill, NJ, USA) and penicillin–streptomycin (100 U/ml and 100 mg/ml, respectively) at 37°C in a humidified atmosphere with 5% CO₂. The medium was changed every 3 days during culturing. Plates were left undisturbed for 7–10 days to allow cell migration from the explants. Passaging began when cells reached 70% confluence. We used hPMSCs at passage 4–5. Five batches of hPMSCs were isolated from five different placentas in this study. The characteristics of these cells did not show significant differences (Chen et al., 2015).

The human monocytic THP-1 cells (TIB-202, American Type Culture Collection [ATCC]) were cultured in RPMI 1640 (Gibco, Grand Island, NY, USA) with 2 mM L-glutamine containing 1.5 g/l sodium bicarbonate, 4.5 g/l glucose, 10 mM HEPES and 1.0 mM sodium pyruvate supplemented with 10% FBS (Biological Industries, Kibbutz Beit–Haemek, Israel) and 0.05 mM 2-mercaptoethanol (Sigma-Aldrich, St. Louis, MO, USA). The cells were differentiated with 100 nM PMA (phorbol 12-myristate 13-acetate; Sigma-Aldrich) for 3 days before experiments. Mouse macrophage RAW 264.7 (TIB-71, ATCC) cells were incubated in DMEM (Gibco) supplemented with 10% FBS (Biological Industries), 100 U/ml penicillin and 100 mg/ml streptomycin at 37°C under 5% CO₂. RAW 264.7 cells were treated with LPS (0–1000 ng/ml, Sigma-Aldrich) and Slit2 (0–2 μg/ml, Peprotech) for the indicated times.

Preparation of the conditioned medium

hPMSCs were grown in DMEM with 10% FBS until 80% confluence. The culture medium was removed, and cell layers were washed and incubated with serum-free DMEM for 2 days. Conditioned medium (CM) was collected, centrifuged at 2300×g for 5 min at 4°C, passed through a 0.22-μm filter and stored at −80°C for use in subsequent experiments.

To remove Slit2 from CM, CM was incubated with 2.4 µg/ml anti-Slit2 polyclonal antibody (GeneTex, Hsinchu, Taiwan) or isotype IgG (Dako, Glostrup, Denmark) overnight at 4°C with rotation. Then, 100 µl of 50% protein G-agarose bead slurry (Thermo Fisher Scientific, Rockford, IL, USA) was added and the mixture was incubated for 1 h at 4°C with rotation. The CM containing agarose beads was centrifuged at 2300×g for 5 min at 4°C, and the supernatant was collected and used immediately. To verify that Slit2 was removed from CM, an aliquot of CM stripped of Slit2 by immunoprecipitation was tested for Slit2 concentration using ELISA.

RT-PCR

Total RNA (1 µg) from isolated RAW 264.7 cells was extracted using the TRIzol reagent (Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s protocol, and cDNA was synthesized using oligodeoxythymidine (Promega Corporation, Madison, WI, USA) and a Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA, USA). The cDNA was amplified by PCR with 35 cycles at 95°C for 30 s, 60°C for 30 s and 72°C for 30 s. We used 18s rRNA as an internal control. PCR products were analyzed on 2% agarose gels (Amresco, Solon, OH, USA) by electrophoresis. The primers used are the same as those used in a previous report (Liao et al., 2012).

ELISA

ELISA was used to assess the Slit2 concentration in the hPMSC supernatant as per the manufacturer’s instructions (MyBioSource, San Diego, CA, USA).

Western blotting

Total protein (20 μg) was separated by electrophoresis on 10% sodium dodecyl sulfate-polyacrylamide gels and transferred to an immobilon polyvinylidene difluoride membrane (Amersham Biosciences, Piscataway, NJ, USA). We conducted immunoblot analysis using antibodies against Slit2 (1:1000), Robo4 (1:1000; Abcam, Cambridge, UK), p38 (40 kDa; 1:1000; Santa Cruz, Dallas, TX, USA), phosphorylated p38 (pp38; 1:1000; Thr180/Tyr182; Cell Signaling) or Rap1 (1:1000; Millipore, Temecula, CA, USA) and the loading control, α-tubulin (1:5000; Millipore) overnight at 4°C, followed by incubation with appropriate horse-radish peroxidase-conjugated secondary antibodies for 1 h at room temperature. The proteins were visualized using a chemiluminescence detection kit (Millipore).

Co-immunoprecipitation assay

To demonstrate the interaction between Slit2 and Robo4, RAW 264.7 cells treated with or without 2 μg/ml Slit2 recombinant protein for 10–15 min were extracted in 1× CHAPS lysis buffer (20 mM Tris, 136.8 mM NaCl [pH 7.5], 1% CHAPS) and centrifuged at 10 000×g for 15 min at 4°C to obtain cell lysates. Cell lysates (150 μg) were precleared with 40 μl protein G bead slurry (Thermo Fisher Scientific) for 30 min at 4°C and subsequently centrifuged at 4000×g for 5 min at 4°C to discard protein G beads. Cell lysates were incubated with anti-Robo4 antibody (R&D, Minneapolis, MN, USA) overnight at 4°C, followed by the addition of 40 μl protein G beads for 3 h at 4°C. Protein G beads were obtained through centrifugation at 3000×g for 5 min at 4°C, and the immunoprecipitates were washed twice with 1× CHAPS lysis buffer and once with PBS. The protein complexes were subjected to immunoblotting with specific primary anti-Slit2 (1:1000), Robo4 (1:1000; Abcam) and α-tubulin (1:5000) and were subsequently hybridized with the respective secondary antibodies.

Flow cytometry

RAW 264.7 cells were blocked with 0.5% bovine serum albumin (BSA) in 1× PBS at room temperature for 10 min and centrifuged at 2300×g for 3 min, and the pellets were resuspended in PBS containing 10 μl anti-Robo1, Robo4 (R&D), CD11b (BioLegend, San Diego, CA, USA) or CD18 (Integrin β2 active form; Bio-Rad, Hercules, CA, USA) conjugated with phycoerythrin at room temperature for 1 h. Subsequently, cells were washed with blocking buffer twice and PBS once. These markers were quantified through flow cytometric analysis. The unstained cells were used for the negative control in the study of CD11b and CD18. The cells only stained with phycoerythrin-conjugated secondary antibodies without primary antibodies were used for the negative control in the study of Robo1 and Robo4. Data were collected and analyzed using a FACScan (Becton Dickinson, San Jose, CA, USA) equipped with CellQuest software. Mean fluorescence intensity is used to measure the shift in fluorescence intensity of RAW 264.7 cells to demonstrate an increase in expression of a specific antigen marker per cell under the same antibody concentrations and flow cytometry settings. As CD11b is highly expressed in RAW 264.7 cells, alteration of CD11b abundance on the cell surface was evaluated through fluorescence intensity, an index of the surface concentration of integrin per cell (Chen et al., 2013).

Transwell migration assay

RAW 264.7 cells (1.5 × 10⁴ cells/well) treated with Slit2 (0–2 μg/ml) were added to the upper chamber of a Transwell device (Costar, Corning Incorporated, Kennebunk, ME, USA) and were allowed to migrate for 24 h in DMEM with 2% FBS added to the lower chamber. Transmembrane migrated cells were stained with DAPI (Sigma-Aldrich) and quantified under a microscope (magnification 50×; Axiovert 200; Carl Zeiss MicroImaging) equipped with Image-Pro Plus software (Media Cybernetics).

Adhesion assay

RAW 264.7 cells (1.5 × 10⁴ cells/well) were seeded into 96-well plates and incubated under 5% CO₂ at 37°C for 1 h. Subsequently, the supernatant was discarded, and cells were washed with PBS twice. Then 100 μl substrate solution (40 mg phosphatase substrate; Sigma-Aldrich) dissolved in 8 ml lysis buffer (50 mM sodium acetate [pH 5.0] + 0.1% Triton-X-100) was added to each well in 96-well plates. Cells were incubated under 5% CO₂ at 37°C for 2 h in the dark. Finally, 100 μl stop solution (1N NaOH) was added to each well; the optical density values were then examined to determine absorbance at 405 nm (phosphatase activity) on an ELISA reader (Tecan, Männedorf, Switzerland).

Rap1 activation assays

According to the manufacturer’s instructions, Rap1 activation assays were performed using a commercial Rap1-activity assay kit (Thermo Fisher Scientific). Briefly, RAW 264.7 cells were detached using 2 mM EDTA and resuspended in serum-free DMEM containing 0.05% BSA (Sigma-Aldrich). RAW 264.7 cells were stimulated with LPS (1 µg/ml), Slit2 (2 µg/ml), p38 siRNA (Cell Signaling) or hPMSC-CM for the indicated times and then lysed in Rap1 activation lysis buffer. Lysates were clarified through centrifugation, and then 200 µl lysates were incubated with GST-tagged RBD of RalGDS (residues 788-884 of human Ral GDS-Rap binding domain) fusion protein precoupled to glutathione beads (Millipore) to specifically pull down the GTP-bound form of Rap1. Samples were incubated for 1 h at 4°C with rotation. Beads were washed three times with lysis buffer. Rap1 was detected through western blotting with anti-Rap1 antibody.

Transfection with p38 siRNA

RAW 264.7 cells (5 × 10⁵) were seeded into a 6-cm dish with DMEM supplemented with 10% FBS overnight. RAW 264.7 cells were then transfected with p38 siRNA (Cell Signaling) and non-silencing control siRNA (Invitrogen) using a Lipofectamine RNAiMAX (Invitrogen) transfection reagent. After 48 h of transfection, the p38 level was confirmed through western blotting analysis.

Statistics

All values are reported in mean ± SD. At least three independent tests were conducted for each experiment. Differences were assessed using the Student’s t test or Fisher’s exact test, as appropriate. A P-value of <0.05 was considered significant. Data were analyzed using the Statistical Package for the Social Sciences Version 20.0 (IBM Inc., Armonk, NY, USA).

Results

Slit2 interacts with Robo4, decreases the cell migration and increases adhesion in macrophages

hPMSCs were isolated from term placentas as in our previous reports (data not shown) (Chen et al., 2013, 2014, 2015). Furthermore, we observed that hPMSCs secreted Slit2 (Fig. 1A). However, the role of Slit2 in the placental villous microenvironment is unknown.

Figure 1.

Determination of Slit2 expression in human placental multipotent mesenchymal stromal cells, and the expression of Robo1 and Robo4 molecules stimulated by lipopolysaccharide and Slit2 in RAW 264.7 cells. (A) Slit2 levels in the CM of hPMSC before and after immunoprecipitation with isotype IgG (CM/IgG) or Slit2 depletion with anti-Slit2 polyclonal antibody (CM/dS) was assessed through ELISA (N = 6). (B) The mRNA of the Robo1–4 receptor in RAW 264.7 cells and THP-1 cells were examined 24 h after cell plating by using RT-PCR. (C) The expression of Robo1 and Robo4 in RAW 264.7 cells was examined through flow cytometry after stimulation with LPS (1 μg/ml) or Slit2 (2 μg/ml) for 24 h. A representative flow cytometry figure is shown. The expression of Robo1 (D) and Robo4 (E) was not altered by LPS and Slit2. (D, E) Four independent experiments were quantified. (CTL: cells without treatment). The uncropped agarose gels can be seen in Supplementary Fig. S1. CM, conditioned medium; hPMSC, human placental multipotent mesenchymal stromal cells; LPS, lipopolysaccharide.

Robo receptors are molecular ligands for Slit2 (Dickson and Gilestro, 2006). The mRNA encoding Robo1–4 of RAW 264.7 cells and THP-1 cells was studied, and the mRNAs of Robo1 and Robo4 were found in RAW 264.7 cells and THP-1 cells. RAW 264.7 cells expressed high level of Robo4 mRNA and very low level of Robo1 mRNA (Fig. 1B). In order to establish the in vitro cell model between the hPMSCs and macrophages before the preclinical animal study for perinatal infection, the RAW 264.7 cells were used for the following experiments. We analyzed the effect of LPS on the Slit2/Robo signaling pathway. We examined whether Robo1 and Robo4 levels are altered with LPS or Slit2 treatment. RAW 264.7 cells were stimulated with LPS (1 µg/ml) or Slit2 (2 μg/ml) for 24 h; flow cytometry showed that Robo1 and Robo4 were not influenced by LPS or Slit2 (Fig. 1C–E). Moreover, Slit2 did not alter the expression of Robo1 or Robo4 in LPS-stimulated RAW 264.7 cells. Slit2 could interact with cell surface Robo4 in RAW 264.7 cells with Slit2 recombinant protein treatment for 10–15 min, as shown by co-immunoprecipitation (Fig. 2A).

Figure 2.

Interaction between Slit2 and Robo4, and the Slit2 effect on migration and adhesion abilities of RAW 264.7 cells. (A) Interaction between Slit2 and Robo4 with or without 2 μg/ml Slit2 treatment for 10–15 min was examined through co-IP in RAW 264.7 cell lysate. RAW 264.7 cell lysate, Robo4 was immunoprecipitated on to protein-G beads from RAW 264.7 cell lysate and probed using antibodies against Slit2 and Robo4. The WB revealed that Slit2 co-immunoprecipitates with Robo4. The lower panel shows the input of Slit2, Robo4 and tubulin. (B, C) Adhesion (B) and migration (C) abilities were examined after stimulation with various doses of Slit2 (0–2 μg/ml). Slit2 increased RAW 264.7 cell adhesion (B) and decreased RAW 264.7 cell migration (C) capacities significantly. The data of at least three independent experiments are shown. The uncropped WBs can be seen in Supplementary Fig. S1. IP, immunoprecipitation; WB, western blot.

Cell migration and adhesion are crucial processes in physiological functions, such as embryogenesis, tissue remodeling, wound healing and leukocyte homing (Marhaba and Zoller, 2004). To determine whether the migration and adhesion abilities are altered by Slit2, we treated RAW 264.7 cells with 0–2 μg/ml Slit2, which decreased the migration of RAW 264.7 cells and considerably increased their adhesion (Fig. 2B and C).

Slit2 inhibits the LPS-induced CD11bCD18 activation in macrophages

Cell adhesion and migration are regulated by adhesion molecules (Van Der Laan et al., 2001). Integrins are heterodimeric cell surface glycoproteins consisting of α and β subunits. CD11b (integrin alpha M) and CD18 (integrin β2) form the heterodimeric integrin α_Mβ₂ molecule, which is expressed on the surfaces of leukocytes that are involved in the innate immune system, including monocytes, natural killer cells, granulocytes and macrophages (Solovjov et al., 2005). To examine whether these adhesion molecules are influenced by LPS and Slit2, RAW 264.7 cells were stimulated with 1 µg/ml LPS with or without 2 μg/ml Slit2 pretreatment. Flow cytometry revealed that CD11b and CD18 (active form) were upregulated with LPS, but the effect was inhibited by Slit2 pretreatment for 1 h (Fig. 3A–D, LPS vs Slit2 + LPS). Furthermore, LPS affected RAW 264.7 cell adhesion and migration. Slit2 considerably increased LPS-induced RAW 264.7 cell adhesion but reduced cell migration (Fig. 3E and F). Next, we examined whether Slit2 secreted by hPMSCs influences macrophage motility. A similar effect was observed in RAW 264.7 cells stimulated with 1 µg/ml LPS with or without hPMSC-CM pretreatment. The inhibition effect of hPMSC-CM on the expression of CD11b and CD18 decreased with Slit2 depletion (Fig. 4A–D). LPS increased RAW 264.7 cell adhesion and migration. hPMSC-CM significantly increased LPS-induced RAW 264.7 cell adhesion but reduced cell migration (Fig. 4E and F). These effects were abrogated after Slit2 depletion from CM (Fig. 4E and F).

Figure 3.

The effect of Slit2 on the expression of CD11b and CD18 (active form) in RAW 264.7 cells. Expression of CD11b and CD18 (active form) was determined in RAW 264.7 cells stimulated using 1 µg/ml LPS with or without 2 μg/ml Slit2 pretreatment. (A) CD11b expression was examined through flow cytometry after stimulation of RAW 264.7 cells with LPS for 24 h with or without Slit2 pretreatment. CD11b was upregulated by LPS, but the effect was attenuated by Slit2 pretreatment for 1 h (LPS vs Slit2 + LPS) in representative flow cytometry data. (B) Quantification of CD11b expression through peak fluorescence intensity of three independent experiments is shown. (C) CD18 (active form) expression was examined through flow cytometry after stimulation of RAW 264.7 cells with LPS for 24 h with or without Slit2 pretreatment. CD18 (active form) was upregulated by LPS, but the effect was attenuated by Slit2 pretreatment for 1 h (LPS vs Slit2 + LPS) in the representative flow cytometry data. (D) Quantification of CD18 (active form) expression through flow cytometry of four independent experiments is shown. (E) Adhesion of RAW 264.7 cells was examined after stimulation with LPS with or without Slit2 treatment. (F) Migration of RAW 264.7 cells was examined after stimulation with LPS with or without Slit2 treatment. Slit2 significantly increased RAW 264.7 cell adhesion induced by LPS but reduced cell migration. Five independent experiments are quantified in (E) and (F). (Culture conditions: RAW 264.7 cells without treatment in DMEM supplemented with 10% fetal bovine serum [CTL] or containing 2 μg/ml Slit2 [Slit2]).

Figure 4.

The effect of human placental multipotent mesenchymal stromal cell-conditioned medium on the expression of CD11b and CD18 (active form) in RAW 264.7 cells. Expression of CD11b and CD18 (active form) was determined in LPS-stimulated (1 μg/ml) RAW 264.7 cells with or without hPMSC-CM pretreatment. (A) CD11b expression was examined through flow cytometry after stimulation of RAW 264.7 cells with LPS for 24 h with or without hPMSC-CM pretreatment. CD11b was upregulated by LPS (CTL vs LPS), but the effect was attenuated by CM pretreatment for 1 h (LPS vs CM + LPS) in representative flow cytometry data. The suppression effect was reduced after the Slit2 was depleted from CM (CM + LPS vs CM/dS + LPS). (B) CD11b expression quantified through peak fluorescence intensity of five independent experiments is shown. (C) CD18 (active form) expression increased after stimulation with LPS (CTL vs LPS). CM significantly inhibited LPS-induced CD18 expression (LPS vs CM + LPS), but the inhibition was abrogated with Slit2 depletion from CM (CM + LPS vs CM/dS + LPS) in representative flow cytometry data. (D) CD18 (active form) expression quantified through flow cytometry of five independent experiments is shown. (E) Adhesion of RAW 264.7 cells was examined after stimulation with LPS with or without CM treatment. (F) Migration of RAW 264.7 cells was examined after stimulation with LPS with or without CM treatment. CM significantly increased RAW 264.7 cell adhesion induced by LPS but reduced cell migration. These effects were reduced after Slit2 depletion from CM. Five independent experiments are quantified in (E) and (F). (Culture conditions: RAW 264.7 cells without treatment in DMEM supplemented with 10% fetal bovine serum (FBS) [CTL]; CTL mixed with CM containing 10% FBS [1:1 ratio; CM]; Slit2 depletion from the conditioned medium [CTL mixed with CM with 10% FBS after Slit2 depletion, 1:1 ratio; CM/dS]).

Slit2 inhibits the LPS-induced CD11bCD18 activation in macrophages via p38-Rap1 signaling

p38 MAPK was involved in stress-induced signal transduction and inflammatory cytokine production (Tibbles and Woodgett, 1999). The present data show that LPS-induced p38 phosphorylation in RAW 264.7 cells, which reached the maximum level at 30 min: this effect was suppressed by Slit2 treatment (Fig. 5A). This inhibition effect was further observed in RAW 264.7 cells pretreated with hPMSC-CM. The basal level of phosphorylated p38 varied in these cells; however, the suppression effect on p38 phosphorylation at 15 min decreased after Slit2 depletion (Fig. 5B). Rap1 may regulate integrin activity and affinity (Bos, 2005). We further found that LPS-induced Rap1 activation in RAW 264.7 cells, which was inhibited by Slit2 pretreatment (left panel, Fig. 5C). A similar effect was observed in RAW 264.7 cells stimulated with 1 µg/ml LPS with hPMSC-CM pretreatment. The inhibition effect of hPMSC-CM on Rap1 activation in RAW 264.7 cells decreased with Slit2 depletion (right panel, Fig. 5C). Next, we analyzed whether Rap1 functions upstream or downstream of p38 for controlling macrophage motility. The expression of total p38 and Rap1 decreased in RAW 264.7 cells after p38 siRNA treatment without LPS stimulation (left panel, Fig. 5D). In LPS-treated RAW 264.7 cells, Rap1 expression significantly decreased after p38 siRNA treatment compared with after control-siRNA treatment. Thus, Rap1 functions downstream of p38 signaling (right panel, Fig. 5D).

Figure 5.

Slit2 attenuated LPS-induced p38 and Rap1 activation in RAW 264.7 cells. (A) The phosphorylation of p38 (pp38) was examined in RAW 264.7 cells after LPS treatment (1 µg/ml) in the presence of (A) recombinant Slit2 protein (2 μg/ml) and (B) hPMSC-CM with or without Slit2 depletion (CM/dS). (C) Active Rap1 expression was examined in RAW 264.7 cells after treatment with LPS (1 µg/ml) in the presence of recombinant Slit2 protein (2 μg/ml) or CM with or without Slit2 depletion (CM/dS). (D) Total p38 and Rap1 expression was examined in RAW 264.7 cells treated with p38 siRNA. LPS (1 µg/ml) induced Rap1 expression in RAW 264.7 cells, which was inhibited by p38 siRNA treatment, compared with cells treated with CTL (scramble)-siRNA. This indicates that Rap1 functions downstream of p38 activation. (Culture conditions: RAW 264.7 cells without treatment in DMEM supplemented with 10% fetal bovine serum (FBS) [CTL] or containing 2 μg/ml Slit2 [Slit2]; CTL mixed with a conditioned medium containing 10% FBS [1:1 ratio; CM]; Slit2 depletion from conditioned medium [CTL mixed with conditioned medium with 10% FBS after Slit2 depletion, 1:1 ratio; CM/dS]); CTL-siRNA: scramble siRNA. The uncropped western blots can be seen in Supplementary Fig. S2.

Flow cytometry revealed that the upregulation of CD11b and CD18 (active form) induced by LPS in RAW 264.7 cells was inhibited by p38 siRNA treatment (Fig. 6A–D). LPS increased RAW 264.7 cell adhesion and migration. Similar to the hPMSC-CM significantly increased LPS-induced RAW 264.7 cell adhesion but reduced cell migration (Fig. 4E and F), the LPS-treated RAW 264.7 cell adhesion was enhanced and migration was inhibited by p38 siRNA treatment compared with control-siRNA treatment (Fig. 6E and F). Thus, hPMSC-derived Slit2 regulated LPS-induced RAW 264.7 cell motility, which involved p38-Rap1 signaling.

Figure 6.

The role of p38 in the expression of CD11b and CD18 (active form) of RAW 264.7 cells. Expression of CD11b and CD18 (active form) was determined in RAW 264.7 cells stimulated with 1 µg/ml LPS with or without p38 siRNA pretreatment. (A) CD11b expression was examined through flow cytometry after stimulation of RAW 264.7 cells with LPS for 24 h with or without p38 siRNA pretreatment. CD11b was upregulated by LPS, but the effect was attenuated by p38 siRNA pretreatment (CTL-siRNA LPS vs p38 siRNA LPS) in representative flow cytometry data. (B) CD11b expression quantified through peak fluorescence intensity of five independent experiments is shown. (C) CD18 (active form) expression was examined through flow cytometry after stimulation of RAW 264.7 cells with LPS for 24 h with or without p38 siRNA pretreatment. CD18 (active form) was upregulated by LPS, but the effect was attenuated by Slit2 pretreatment (CTL-siRNA LPS vs p38 siRNA LPS) in representative flow cytometry data. (D) CD18 (active form) expression quantified through flow cytometry of five independent experiments is shown. (E) Adhesion of RAW 264.7 cells was examined after stimulation with LPS with or without p38 siRNA treatment. (F) Migration of RAW 264.7 cells was examined after stimulation with LPS with or without p38 siRNA treatment. p38 siRNA treatment significantly increased LPS-induced RAW 264.7 cell adhesion but reduced cell migration. Five independent experiments are quantified in (E) and (F). CTL-siRNA: scramble siRNA. RAW 264.7 cells without LPS treatment in DMEM supplemented with 10% fetal bovine serum (CTL).

Discussion

LPS-induced inflammation is a critical pathological event in placental infection and preterm births. In this study, we discovered for the first time that hPMSCs expressed Slit2. Macrophage RAW 264.7 cells expressed a high level of Robo4. Slit2 or hPMSC-CM inhibited macrophage migration. LPS-induced CD11bCD18 (integrin αMβ2) expression, which was inhibited by Slit2 or hPMSC-CM. p38-Rap1 signaling was involved in CD11b and CD18 activation in RAW 264.7 cells. These findings are in agreement with previous reports that CM derived from mesenchymal stem cells exerts immunomodulatory effects in various preclinical models. For example, the CM of mesenchymal stem cells obtained from human periodontal ligament tissue showed anti-inflammatory properties and reduced the expression of TLR4 (toll like receptor 4) and NFκB (nuclear factor NF-kappa-B) in LPS-stimulated mouse motoneurons, and this CM significantly inhibited the LPS-induced production of TNF-α and IL-1β proinflammatory cytokines in human monocytic and oligodendrocytic cell lines (Rajan et al., 2016; Ballerini et al., 2017).

The mechanisms underlying hPMSCs-mediated immunomodulation remain unclear. In experimental animal models of ischemic hind limb, stroke and spinal cord injury, hPMSCs have been shown to produce an effect through the expression of neurotrophic factors or cytokines (Kranz et al., 2010; Kim et al., 2013; Li et al., 2014; He et al., 2017;, Mattar and Bieback, 2015). In order to develop a preclinical mouse model of hPMSCs in treatment of perinatal infection, RAW 264.7 cells were used in this study. The in vitro model enables specific evaluation of the Slit2-mediated crosstalk between hPMSCs and macrophages.

Our results showed that Slit2 or hPMSC-derived Slit2 reduced macrophage migration. Slit2 can interact with the transmembrane receptor Robo4 in macrophages, as shown by co-immunoprecipitation, indicating that Slit2/Robo4 signaling may regulate macrophage motility during placental infection. Slit2 inhibited the chemokine-induced chemotaxis of leukocytes in vitro (Ye et al., 2010). Moreover, Slit2 inhibited monocyte tethering and adhesion to endothelial cells and the ability of monocytes to resist detachment due to circulation shear forces (Mukovozov et al., 2015). Slit2 reduced inflammatory cell infiltration and acute tubular injury after ischemia–reperfusion injury in mice (Chaturvedi et al., 2013). However, Slit2 did not impair the protective innate immune functions of neutrophils, such as phagocytosis and superoxide production, and did not prevent neutrophils from killing extracellular pathogens (Chaturvedi et al., 2013).

In this study, we observed that LPS increased the adhesion and migration in RAW 264.7 cells, whereas Slit2 increased cell adhesion but inhibited cell migration. Thus, hPMSC-CM, which contained Slit2, inhibited LPS-induced cell migration but not cell adhesion in RAW 264.7 cells. LPS-induced CD11bCD18 activation in RAW 264.7 cells, which was inhibited by Slit2 recombinant protein or Slit2 derived from hPMSC-CM. β2 integrins are heterodimeric adhesion receptors expressed on leukocytes, and they are essential for monocyte migration and phagocytosis. Furthermore, β2 integrins are involved in leukocyte recruitment to inflammatory sites (Herter and Zarbock, 2013). αMβ2 (Mac-1, CD11bCD18) is expressed predominantly on myeloid cells and mediates the adhesive reactions of leukocytes during the inflammatory response. αMβ2 contributes to the firm adhesion of neutrophils to endothelial cells, promotes their diapedesis, and participates in neutrophil migration to inflammation sites (Ding et al., 1999).

Chemo-attractants or inflammatory mediators can cause inside-out signaling through Rap1 to regulate integrin activation (Shimonaka et al., 2003; Bos, 2005). In this study, Rap1 and p38 were activated by LPS. On LPS challenge, the levels of activated Rap1 and p38 phosphorylation increased, indicating that p38 and Rap1 activation is involved in integrin activation in response to LPS-induced macrophage cell motility. The increased active Rap1 expression was inhibited by Slit2 or hPMSC-CM. These findings are supported by reports indicating that LPS-induced monocyte modulation and neutrophil chemotaxis, which were dependent on p38 MAPK signaling (Shimonaka et al., 2003; Arefieva et al., 2005). Rap1 translates chemokine signals into leukocyte integrin upregulation and promotes subsequent leukocyte migration (Shimonaka et al., 2003). p38 siRNA inhibited p38 phosphorylation and Rap1 activation. Furthermore, we observed that the Rap1 expression was inhibited by p38 siRNA. This finding revealed that Rap1 functions downstream of p38 in LPS-stimulated RAW 264.7 cells.

In summary, Slit2/Robo4 signaling is crucial for regulating LPS-induced macrophage motility. hPMSC-derived Slit2 represses the LPS-induced activation of CD11bCD18 for regulating macrophage adhesion and migration. Slit2 is a potential paracrine factor enabling communication between hPMSCs and macrophages in the placental villus during placental infection. The present in vitro model provides the opportunity to define how hPMSCs shape a placental microenvironment, and to evaluate whether suppression of the LPS-mediated activation in macrophages by hPMSCs prevents inflammation progression at maternal-fetal interface. These findings may further provide useful information to design a pre-clinical mouse model, such as hPMSC dosing, method of hPMSC administration and treatment efficacy.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Supplementary data

Supplementary data are available at Molecular Human Reproduction online.

Authors’ roles

C.P.C. conceived, designed the experiments. C.P.C. wrote the paper. L.K.W., Chia-Yu Chen and Y.H.W. performed the experiments. C.P.C., Y.H.W. and Chen-Yu Chen analyzed the data. C.P.C. contributed reagents/materials/analysis tools. All authors have read and agree to the published version of the manuscript.

Funding

This work was supported by grant from the Ministry of Science and Technology of Taiwan (MOST 104-2314-B-195-010-MY3 to C.-P.C.) and MacKay Memorial Hospital (MMH-E 105001 to C.-P.C.).

Conflict of interest

The authors declare no potential conflicts of interest.