Mesenchymal Stem/Stromal Cells Inhibit the NLRP3 Inflammasome by Decreasing Mitochondrial Reactive Oxygen Species

Authors

  • Joo Youn Oh,

    Corresponding author
    1. Department of Ophthalmology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    2. Laboratory of Ocular Regenerative Medicine and Immunology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    • Correspondence: Joo Youn Oh, M.D, Ph.D., Department of Ophthalmology, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul 110–744, Korea. Telephone: +82-2-2072-0836; Fax: +82-2-741-3187; e-mail: jooyounoh77@gmail.com

    Search for more papers by this author
  • Jung Hwa Ko,

    1. Department of Ophthalmology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    2. Laboratory of Ocular Regenerative Medicine and Immunology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    Search for more papers by this author
  • Hyun Ju Lee,

    1. Department of Ophthalmology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    2. Laboratory of Ocular Regenerative Medicine and Immunology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    Search for more papers by this author
  • Ji Min Yu,

    1. Institute for Regenerative Medicine, Texas A&M Health Science Center College of Medicine at Scott & White, Texas, USA
    Search for more papers by this author
  • Hosoon Choi,

    1. Institute for Regenerative Medicine, Texas A&M Health Science Center College of Medicine at Scott & White, Texas, USA
    Search for more papers by this author
  • Mee Kum Kim,

    1. Department of Ophthalmology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    2. Laboratory of Ocular Regenerative Medicine and Immunology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    Search for more papers by this author
  • Won Ryang Wee,

    1. Department of Ophthalmology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    2. Laboratory of Ocular Regenerative Medicine and Immunology, Biomedical Research Institute, Seoul National University Hospital, Seoul, Korea
    Search for more papers by this author
  • Darwin J. Prockop

    1. Institute for Regenerative Medicine, Texas A&M Health Science Center College of Medicine at Scott & White, Texas, USA
    Search for more papers by this author

Abstract

Mesenchymal stem/stromal cells (MSCs) control excessive inflammatory responses by modulating a variety of immune cells including monocytes/macrophages. However, the mechanisms by which MSCs regulate monocytes/macrophages are unclear. Inflammasomes in macrophages are activated upon cellular “danger” signals and initiate inflammatory responses through the maturation and secretion of proinflammatory cytokines such as interleukin 1β (IL-1β). Here we demonstrate that human MSCs (hMSCs) negatively regulate NLRP3 inflammasome activation in human or mouse macrophages stimulated with LPS and ATP. Caspase-1 activation and subsequent IL-1β release were decreased in macrophages by direct or transwell coculture with hMSCs. Addition of hMSCs to macrophages either at a LPS priming or at a subsequent ATP step similarly inhibited the inflammasome activation. The hMSCs had no effect on NLRP3 and IL-1β expression at mRNA levels during LPS priming. However, MSCs markedly suppressed the generation of mitochondrial reactive oxygen species (ROS) in macrophages. Further analysis showed that NLRP3-activated macrophages stimulated hMSCs to increase the expression and secretion of stanniocalcin (STC)-1, an antiapoptotic protein. Addition of recombinant protein STC-1 reproduced the effects of hMSCs in inhibiting NLRP3 inflammasome activation and ROS production in macrophages. Conversely, the effects of hMSCs on macrophages were largely abrogated by an small interfering RNA (siRNA) knockdown of STC-1. Together, our results reveal that hMSCs inhibit NLRP3 inflammasome activation in macrophages primarily by secreting STC-1 in response to activated macrophages and thus by decreasing mitochondrial ROS. Stem Cells 2014;32:1553–1563

Introduction

Mesenchymal stem/stromal cells (MSCs) are potent modulators of the innate and adaptive immune responses [1-5]. Most of the studies investigating the mechanisms of MSCs so far have focused primarily on the interactions of MSCs with T lymphocytes, B lymphocytes, natural killer (NK) cells, or dendritic cells. Recently, a few reports suggest that MSCs also modulate the cells of monocytic lineages, especially macrophages [6-10]. However, the mechanisms by which MSCs regulate these cells are not fully elucidated.

Inflammasomes are cytosolic multiprotein complexes that are present mainly in myeloid cells including macrophages and dendritic cells. They proteolytically activate caspase-1 in response to cellular “danger” signals [11-13]. Activated caspase-1, in turn, processes proinflammatory cytokines including pro-interleukin (IL)−1β into the biologically active and secreted forms. IL-1β acts on neighboring nonimmune and immune cells to initiate and orchestrate inflammatory responses. Considering the strong proinflammatory activity of IL-1β, the activation of inflammasomes must be tightly regulated.

Here we explored the possibility that MSCs might act as regulators of the NLRP3 inflammasome, the most well-characterized inflammasome. We found that human MSCs (hMSCs) negatively regulated NLRP3 inflammasome activation in human and mouse macrophages that were exposed to lipopolysaccharide (LPS) and ATP. The hMSCs did not affect NLRP3 and IL-1β expression at mRNA levels but markedly suppressed mitochondrial reactive oxygen species (ROS) production that is critical for NLRP3 inflammasome activation [14-16]. Of note, hMSCs were activated by LPS/ATP-treated macrophages to increase the expression and secretion of stanniocalcin (STC)-1, an antiapoptotic protein. The hMSCs with an siRNA knockdown of STC-1 did not inhibit NLRP3 inflammasome activation nor ROS production, while an addition of recombinant protein STC-1 suppressed NLRP3 inflammasome activation.

Materials and Methods

Cells, Reagents, and Stimulation

Human peripheral blood-derived monocytes were purchased from Cell bank in Seoul National University Hospital (Seoul, Korea) and were differentiated into macrophages as previously described [17]. Mouse macrophages (RAW264.7 cells) were purchased from American Type Culture Collection (ATCC, Rockville, MD, http://www.atcc.org), and peritoneal macrophages were acquired from C57BL/6 mice as previously described [7]. Mouse macrophages were cultured in high glucose Dulbecco's modified Eagle's medium (DMEM) (Welgene, Daegu, Korea) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco, Grand Island, NY, http://www.invitrogen.com) and 1% penicillin-streptomycin (PS, Lonza, http://www.lonza.com, Basel, Switzerland) at 37°C in 5% CO2. For NLRP3 inflammasome stimulation, macrophages were seeded at 50,000 per cm2 in six-well plates. Two hours later, the cells were treated with 2 µg/mL LPS (Ultra-pure LPS, InvivoGen, San Diego, CA) in high glucose DMEM (Welgene) including 2% heat-inactivated FBS (Gibco) for 4 hours. Then, the cells were washed with PBS three times and incubated in 5 mM ATP (InvivoGen, http://www.invivogen.com) for 45 minutes. After PBS washing, the cells were cultured in DMEM (Welgene, http://www.welgene.com) with 2% heat-inactivated FBS (Gibco) and 1% PS (Lonza) at 37°C in 5% CO2 until further assays.

Human bone marrow-derived MSCs (No. 8004L) were obtained from the Center for the Preparation and Distribution of Adult Stem Cells (http://medicine.tamhsc.edu/irm/msc-distribution.html) that supplies standardized preparations of MSCs enriched for early progenitor cells to more than 300 laboratories under the auspices of an NIH/NCRR grant (P40 RR 17447-06). The cells were cultured at 37°C in 5% CO2 in complete culture medium (CCM) that consisted of α-minimal essential medium (α-MEM), 17% FBS (Gibco), 1% PS (Lonza), and 2 mM l-glutamine (Gibco). For coculture experiments, hMSCs were seeded at 10,000 per cm2 in six-well plates or in six-well transwell and cultured in CCM for 24 hours before coculture with macrophages. Just prior to coculture, hMSCs were washed with PBS three times and cocultured with macrophages in the macrophage culture medium.

For siRNA experiments, hMSCs (50,000 cells per well in six-well plate) were cultured in CCM without 1% PS for 24 hours. The cells were then transfected with siRNA for STC-1 (sc-44126, Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com) or scrambled siRNA (sc-37007, Stealth RNAi siRNA Negative Control, Invitrogen, Carlsbad, CA, http://www.invitrogen.com) with a commercial kit (Lipofectamine RNAiMAX reagent, Invitrogen), and cultured in serum-free Opti-MEM (Invitrogen) for 4 hours at 37°C. To confirm successful knockdown of STC-1 expression, RNA was extracted from the cells at 24 hours after the start of transfection (RNeasy Mini kit, Qiagen, Valencia, CA, http://www1.qiagen. com) and assayed for STC-1 by real-time reverse transcriptase polymerase chain reaction (RT-PCR). The knockdown efficiency of STC-1 in hMSCs was approximately 95%. At the same time point, STC-1-siRNA or scrambled siRNA MSCs were cocultured with macrophages (Supporting Information Fig. 1).

Figure 1.

MSCs inhibited NLRP3 inflammasome-mediated caspase-1 activation and IL-1β secretion in macrophages. Human blood monocyte-derived macrophages or mouse macrophages (peritoneal macrophages or RAW264.7 cells) were primed with LPS (2 µg/mL) for 4 hours, followed by incubation with ATP (5 mM) for 45 minutes. Human MSCs, in direct or transwell coculture, were added to macrophages at LPS priming step (A–F) or at ATP stimulation step (G–L). At 18 hours, caspase-1 activation was analyzed in lysates of macrophages by Western blotting (B, H), and the release of active caspase-1 was determined in culture supernatants by measuring caspase-1 activity (C, I). Also, IL-1β secretion in the supernatants was quantified by enzyme-linked immunosorbent assay (D–F, J–L). Data on caspase-1 activation are representative of three independent experiments, and data on IL-1β secretion represents six or more independent experiments. All data are presented as mean ± SE. Abbreviations: IL, interleukin; LPS, lipopolysaccharide; MSC, mesenchymal stem cell.

Recombinant human STC-1 was purchased from BioVendor, http://www.biovendor.com (Brno, Czech Republic). Recombinant human chemokine (C-X-C motif) ligand 12 (CXCL12) was obtained from (R&D Systems, Minneapolis, MN, http://www.rndsystems.com).

Western Blot Analysis

For protein extraction, the cells were sonicated on ice in PRO-PREP Protein Extraction Solution (Intron Biotechnology, http://www.intronbio.com, Seongnam, Korea) containing a protease inhibitor cocktail (Roche, Indianapolis, IN, http://www.roche-applied-science.com). After centrifugation at 12,000 rpm at 4°C for 20 minutes, clear cell lysates were measured for protein concentration by Bradford assay. A total of 20 µg protein was fractionated by SDS-PAGE on 10% bis-tris gel (Invitrogen), transferred to nitrocellulose membrane (Invitrogen), and then blotted with antibodies against caspase-1 (sc-1597, Santa Cruz Biotechnology) or β-actin (Santa Cruz Biotechnology).

Enzyme-Linked Immunosorbent Assay

The cell-free supernatants were collected from cell cultures and assayed for concentrations of human or mouse IL-1β, STC-1, CXCL12, an thrombospondin (TSP-1) with ELISA kits (R&D Systems) or for active caspase-1 with Colorimetric Caspase 1 Assay Kit (Abcam, Eugene, OR, http://www.abcam.com). Human and mouse IL-18 levels were measured using ELISA kits purchased from eBioscience (San Diego, CA) and R&D Systems, respectively. Specifically, for coculture experiments of mouse macrophages and hMSCs, mouse-specific enzyme-linked immunosorbent assay (ELISA) was used to measure the concentration of IL-1β and IL-18.

Microarrays

For RNA extraction, the cells were lysed in RNA isolation reagent (RNA Bee, Tel-Test Inc., http://www.tel-test.com, Friendswood, TX), and total RNA was extracted using RNeasy Mini kit (Qiagen). About 250 ng of total RNA in each sample was loaded for GeneChip Human Genome U133 plus 2.0 Arrays (Affymetrix, Santa Clara, CA, http://www.affymetrix.com). The data were analyzed with Microarray Suite version 5.0 (MAS 5.0) using Affymetrix default analysis settings and global scaling as normalization method. Hierarchical clustering data were clustered that behave similarly across experiments using GeneSpring GX 11.5.1 (Agilent Technologies, Santa Clara, CA, http://www.agilent.com).

Quantitative Real Time RT-PCR

Total RNA was extracted as described above, and double-stranded cDNA was synthesized by reverse transcription (High Capacity RNA-to-cDNA kit, Applied Biosystems, Carlsbad, CA, http://www.appliedbiosystems.com). Real-time amplification was performed (Taqman Universal PCR Master Mix, Applied Biosystems) and analyzed on an automated instrument (7500 Real-Time PCR System, Applied Biosystems). Human or mouse-specific PCR probe sets were commercially purchased (Taqman Gene Expression Assay Kits, Applied Biosystems). Values were normalized to 18s RNA and expressed as fold changes relative to controls.

Cellular and Mitochondrial ROS Analysis

Cellular or mitochondrial ROS was measured in cells using CellROX Deep Red Reagent (Invitrogen) and MitoTracker Green FM Dye (Invitrogen) [18, 19]. Briefly, 2 hours after LPS-primed cells were treated with ATP for 45 minutes, the cells were loaded with CellROX dye (5 µM) and MitoTracker Green dye (100 nM) at 37°C for 30 minutes and analyzed by flow cytometry (FACSCanto flow cytometer, BD Biosciences, San Jose, CA, http://www.bdbiosciences.com). The data were analyzed using Flowjo software (Tree Star, Inc., http://www.treestar.com Ashland, OR).

Results

MSCs Suppressed Caspase-1 Activation and IL-1β Secretion in Activated Macrophages

To investigate the effects of MSCs on NLRP3 inflammasome activation, human blood monocyte-derived macrophages or mouse macrophages (peritoneal macrophages or RAW264.7 cells) were primed with LPS for 4 hours to induce the expression of NLRP3 and pro-IL-1β and then stimulated with ATP for 45 minutes to activate the NLRP3 inflammasome. The hMSCs were added to macrophages in direct or transwell coculture system either simultaneously with LPS priming (Fig. 1A–1F) or concomitantly with ATP licensing (Fig. 1G–1L). Caspase-1 activation was markedly suppressed in macrophages by transwell coculture with hMSCs as assessed by Western blot in cell lysates (Fig. 1B, 1H). The release of active caspase-1 in culture supernatants, another indicator of caspase-1 activation, was also decreased by direct or transwell coculture with hMSCs (Fig. 1C, 1I). Consistent with the reduction in caspase-1 activity, the levels of IL-1β and IL-18 in the supernatants were significantly reduced by direct or transwell coculture with hMSCs as measured by ELISA (Fig. 1D–1F, 1J–1L; Supporting Information Fig. 2). Also, the effects of hMSCs on IL-1β and IL-18 secretion were dependent on the number of cocultured hMSCs (Supporting Information Fig. 3). To rule out the possibility that IL-1β or IL-18 might be secreted from hMSCs, we stimulated cultures of hMSCs alone with LPS and ATP but found no detectable amount of IL-1β or IL-18 in the supernatants indicating that the NLRP3 inflammasome was not activated in hMSCs (Supporting Information Fig. 4). Together, the data indicate that hMSCs inhibited NLRP3 inflammasome activation in macrophages, and the effects were at least in part mediated by soluble factor(s).

MSCs Did Not Inhibit NLRP3 and Pro-IL-1β Expression at Transcription Level

To explore the mechanism(s) of the inhibitory effects of hMSCs on NLRP3 inflammasome activation, we first evaluated whether hMSCs might increase transcription of the genes for the inflammasome components, NLRP3, ASC, and pro-IL-1β at transcription level during LPS priming. As expected, LPS/ATP stimulation induced a robust increase in IL-1β mRNA and a modest increase in NLRP3 mRNA in human and mouse macrophages as measured by real time RT-PCR (Fig. 2A-2D). However, ASC mRNA level remained unchanged by LPS/ATP (data not shown). Coculture with hMSCs did not affect the transcript levels of NLRP3 and IL-1β in macrophages (Fig. 2A–2D), while TNF-α or IL-6 transcripts were significantly reduced by hMSCs (Fig. 2E, 2F). Thus, data suggest that hMSCs did not modulate the NLRP3 inflammasome via inhibition of LPS priming.

Figure 2.

MSCs did not decrease NLRP3 and IL-1β expression at mRNA levels. Human or mouse macrophages were incubated in LPS (2 µg/mL) for 4 hours and then stimulated with ATP (5 mM) for 45 minutes. Human MSCs (hMSCs) were cocultured with macrophages simultaneously with LPS priming. Six hours after LPS priming, macrophages were evaluated for expression of NLRP3, IL-1β, TNF-α, or IL-6 at mRNA levels by real time reverse transcriptase polymerase chain reaction. Human macrophages that were cocultured in transwell with hMSCs were analyzed using human specific probes (A, B). Mouse macrophages with direct or transwell coculture of hMSCs were analyzed using mouse-specific probes (C–F). The fold changes relative to unstimulated macrophages were calculated by the 2−ΔΔCT method. Data shown are representative of three independent experiments and presented as mean ± SE. Abbreviations: IL, interleukin; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; TNF-α, Tumor Necrosis Factor-α.

MSCs Reduced Mitochondrial ROS in Macrophages

Based on previous observations that hMSCs inhibited NLRP3 inflammasome activation in LPS-primed macrophages (Fig. 1G–1L), we next tested whether hMSCs might affect ATP licensing step by modulating mitochondrial ROS production. Mitochondrial ROS are known to be induced in macrophages by extracellular ATP and extricably linked to caspase-1 activation by the NLRP3 inflammasome [14-16]. We evaluated cellular ROS by measuring fluorescence of the cells labeled with CellROX Deep Red Reagent, a novel cell-permeant dye that fluoresces (near-infrared) upon oxidation by ROS [18, 19]. Also, we assessed mitochondrial ROS by correlating cellular ROS with the distribution of mitochondrial mass after incubating the cells in CellROX and MitoTracker Green, a probe that stains total mitochondria regardless of mitochondrial membrane potential [18, 19]. As shown in Figure 3, LPS/ATP stimulation greatly increased both cellular and mitochondrial ROS in human macrophages. Direct or transwell coculture of hMSCs markedly decreased both cellular and mitochondrial ROS in LPS/ATP-treated macrophages (Fig. 3). Similar results were obtained when hMSCs were added to cultures of macrophages at the LPS priming step (Fig. 3A–3D) or at the ATP licensing step (Fig. 3E–3H). These data collectively suggest that hMSCs suppressed mitochondrial ROS production in macrophages.

Figure 3.

MSCs decreased cellular and mitochondrial ROS in macrophages. Human macrophages were treated with LPS (2 µg/mL, 4 hours) and ATP (5 mM, 45 minutes). Two hours after ATP stimulation, macrophages were stained with CellROX Deep Red and MitoTracker Green and analyzed by flow cytometry. The cells with increased mitochondrial ROS were presented as the percentage of cells that were both positive for CellROX Deep Red and MitoTracker Green. Human MSCs were cocultured with macrophages directly or in transwell from LPS priming (A–D) or from ATP licensing (E–H). Experiments were repeated independently six times with consistent results. Data are presented as mean ± SE. Abbreviations: IL, interleukin; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; ROS, reactive oxygen species.

MSCs Increased Expression and Secretion of STC-1 in Response to LPS/ATP-Stimulated Macrophages

To search for candidate factor(s) that were secreted from hMSCs and suppressed NRLP3 activation in macrophages, we performed microarrays and analyzed the transcriptome of (a) hMSCs cultured alone, (b) hMSCs cocultured in transwell with unstimulated human macrophages, and (c) hMSCs cocultured in transwell with LPS/ATP-stimulated human macrophages. The results demonstrated that hMSCs cocultured with LPS/ATP-stimulated macrophages upregulated expression of 372 gene transcripts and downregulated 471 gene transcripts by twofold or more, compared to hMSCs cultured alone or hMSCs cocultured with unstimulated macrophages (Fig. 4A). We further screened the upregulated genes for candidate molecules responsible for the MSC action and found eight genes encoding secreted proteins (Supporting Information Table 1). Of special interest were STC-1, an antiapoptotic protein and CXCL12, stromal cell-derived factor 1, because their transcripts were highly increased in hMSCs by LPS/ATP-treated macrophages (73.6-fold increase in STC-1 and 11.6-fold increase in CXCL12 compared to hMSCs cultured alone as detected by microarrays) (Supporting Information Table 1). We further confirmed the increases of STC-1 and CXCL-12 in hMSCs at mRNA levels by real time RT-PCR and in culture supernatants at protein levels by ELISA (Fig. 5). Notably, STC-1 transcript was increased 7,458-fold in hMSCs in response to incubation with LPS/ATP-stimulated macrophages. Cultures of macrophages rarely expressed STC-1 (Fig. 5A). Consistent with increase in mRNA for STC-1 in the MSCs was the observation that the secretion of STC-1 protein was increased to levels as high as 5,134–10,096 pg/mL in the supernatants of hMSCs cocultured with LPS/ATP-stimulated macrophages (Fig. 5B, 5C). In contrast, although CXCL12 transcripts were 300-fold increased (Fig. 5A), the levels of secreted CXCL-12 were relatively low (160–213 pg/mL) in the supernatants of hMSCs cocultured with LPS/ATP-treated macrophages (Fig. 5B, 5C). Another observation of interest was that the expression of a large number of proinflammatory cytokines including IL-1β and chemokines were significantly downregulated in hMSCs by coculture with LPS/ATP-stimulated macrophages compared to hMSCs cocultured with unstimulated macrophages (Supporting Information Table 2). The results therefore suggest that proinflammatory microenvironments cause downregulation of proinflammatory and immune responses in MSCs.

Figure 4.

Heat map of microarray assays of MSCs and macrophages. Human macrophages were activated with LPS (2 µg/mL, 4 hours) and ATP (5 mM, 45 minutes). Human MSCs (hMSCs) in transwell were added to macrophages simultaneously with ATP and cocultured for 4 hours. (A): Total RNA was extracted from separate cultures of (a) hMSCs alone, (b) hMSCs cocultured with unstimulated macrophages, and (c) hMSCs cocultured with LPS/ATP-unstimulated macrophages and analyzed by microarrays. Transcriptome of hMSCs cocultured with unstimulated or LPS/ATP-stimulated macrophages were compared to that of hMSCs cultured alone. Expression changes in hMSCs cocultured with LPS/ATP-stimulated macrophages relative to hMSCs cocultured with unstimulated macrophages were presented as a ratio. (B): Also, macrophages were analyzed on microarrays, and transcriptome of LPS/ATP-stimulated macrophages with hMSCs was compared to that of LPS/ATP-stimulated macrophages without hMSCs. Data were normalized to a value of one and variance of 3 SD (+3, red; −3, green). Gene ontology categories and the number of genes with expression differences are depicted in the boxes. Abbreviations: LPS, lipopolysaccharide; MSC, mesenchymal stem cell.

Figure 5.

MSCs were activated to express STC-1 and CXCL12 by LPS/ATP-stimulated macrophages. Human macrophages were stimulated with LPS (2 µg/mL, 4 hours) followed by ATP (5 mM, 45 minutes) and cocultured with human MSCs (hMSCs). Macrophages or hMSCs were assayed for STC-1 and CXCL12 expression by real time reverse transcriptase polymerase chain reaction at 6 hours (A). Secretion of STC-1 and CXCL12 was analyzed in culture supernatants by enzyme-linked immunosorbent assay at 18 hours (B). Data represent three independent experiments and are presented as mean ± SE. Abbreviations: CXCL12, chemokine (C-X-C motif) ligand 12; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; STC-1, stanniocalcin-1.

In addition, we assayed the macrophage transcriptome for changes after LPS/ATP stimulation and upon coculture with hMSCs by microarrays (Fig. 4B). The results revealed that genes involved in inflammatory or immune responses, cell death, or apoptosis were generally downregulated in LPS/ATP-treated macrophages upon coculture with hMSCs. In contrast, genes related to cell homeostasis, survival, or growth were upregulated in macrophages by hMSCs.

In search for signals from the activated macrophages that might drive hMSCs to secrete STC-1, we next screened genes for secreted factors that were upregulated in macrophages that were treated with LPS/ATP and cocultured with hMSCs. Of the upregulated genes, the highest increase was observed in TSP-1 (Supporting Information Table 3). The level of TSP-1 transcript was increased 223.3-fold in LPS/ATP-stimulated macrophages by hMSCs compared to unstimulated macrophages and was 109.4-fold higher compared to LPS/ATP-stimulated macrophages that were not cocultured with hMSCs. The increase of TSP-1 in macrophages by hMSCs was further confirmed by real time RT-PCR and ELISA (Supporting Information Fig. 5). However, addition of recombinant TSP-1 did not induce STC-1 expression in hMSCs, indicating that other signals than TSP-1 were involved in STC-1 upregulation in hMSCs (Supporting Information Fig. 6).

STC-1 Suppressed NLRP3 Inflammasome Activation and Mitochondrial ROS

To further identify the molecule(s) mediating the inhibitory action of hMSCs on NLRP3 inflammasome activation, we directly added recombinant human (rh) STC-1 or CXCL12 to human macrophages either at LPS priming step (Fig. 6A–6D) or at ATP licensing step (Fig. 6E–6H). Recombinant STC-1 dose dependently decreased caspase-1 activation as well as IL-1β and IL-18 secretion (Fig. 6A, 6C, 6E, 6G; Supporting Information Fig. 7A). In contrast, rhCXCL12 had no effect on IL-1β or IL-18 secretion in macrophages (Fig. 6B, 6F; Supporting Information Fig. 8). Cellular ROS levels were significantly reduced by rhSTC-1 in a dose-dependent manner (Fig. 6D, 6H). Also, STC-1 was similarly effective in inhibiting IL-1β or IL-18 secretion and ROS production when added to LPS-primed macrophages (Fig. 6E–6H), suggesting that STC-1 inhibited the NLRP3 inflammasome by modulating the ATP licensing step.

Figure 6.

STC-1 inhibited NLRP3 inflammasome activation and reactive oxygen species (ROS) production in macrophages. Human macrophages stimulated by LPS (2 µg/mL, 4 hours) and ATP (5 mM, 45 minutes) were treated with recombinant STC-1 (10 pg/mL to 100 ng/mL) or CXCL-12 (10 ng/mL to 500 ng/mL) from LPS priming step (A–D) or from ATP licensing step (E–H). At 18 hours, caspase-1 activation was assayed in lysates of macrophages by Western blotting (A, E), and IL-1β secretion was measured in the supernatants by ELISA (B, C, F, G). In addition, at 6 hours, macrophages were stained with CellROX Deep Red and analyzed for total cellular ROS by flow cytometry (D, H). Data are representative of three independent experiments and presented as mean ± SE. Abbreviations: CXCL-12, chemokine (C-X-C motif) ligand 12; IL, interleukin; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; STC-1, stanniocalcin-1.

MSCs with an siRNA Knockdown of the STC-1 Gene Had No Effect on NLRP3 Inflammasome Activation

To further verify that STC-1 might mediate the effects of hMSCs, hMSCs were transfected with an siRNA for STC-1. The hMSCs with an siRNA knockdown of STC-1 suppress neither caspase-1 activation nor IL-1β/IL-18 secretion in macrophages after LPS/ATP stimulation, while hMSCs transfected with the scrambled siRNA significantly inhibited NLRP3 inflammasome activation (Fig. 7A–7C, 7E–7G; Supporting Information Fig. 7B). Similarly, hMSCs with STC-1 siRNA were less effective in decreasing mitochondrial ROS than hMSCs transfected with the scrambled siRNA (Fig. 7D, 7H).

Figure 7.

MSCs with STC-1 siRNA knockdown did not inhibit NLRP3 inflammasome activation. Human MSCs were pretreated with siRNA against STC-1 or with scrambled siRNA and cocultured with LPS/ATP-stimulated macrophages from LPS priming step (A–D) or from ATP addition step (E–H). IL-1β levels were assayed in culture supernatants by ELISA at 18 hours (B, F), and caspase-1 cleavage was evaluated in cell lysates by Western blot at 18 hours (C, G). Mitochondrial ROS were measured in macrophages by flow cytometry at 6 hours (D, H). Data are presented as mean ± SE and represent three independent experiments. Abbreviations: IL, interleukin; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; ROS, reactive oxygen species; SCR, Scrambled; STC-1, stanniocalcin-1.

Discussion

Monocytes and macrophages are among the first responders to tissue injury that detect “danger” signals and initiate the inflammatory process to defend against injury or infection [20-24]. Thus, a tight regulation of monocyte/macrophage activation is essential for protecting host from excessive inflammation and collateral damage. MSCs are good candidates for regulators of monocyte/macrophage activation, because MSCs respond to inflammatory microenvironments and modulate exacerbated immune responses [1-5]. In support for this, several studies recently demonstrated that MSCs decreased the production of TNF-α, IL-1, and IL-6 and increased IL-10 production in macrophages by secreting soluble factors such as Prostaglandin E2 (PGE2) or TSG-6 [6-10]. In this study, we further elucidated the interactions between MSCs and macrophages by investigating the effects of MSCs on the inflammasome of macrophages, a signaling pathway causing macrophages to release cytokines that trigger innate immune responses against infection and injury [11-13]. We found that hMSCs efficiently inhibited NLRP3 inflammasome activation in macrophages, and these effects were mediated by hMSCs being activated to secrete STC-1 and thereby decreasing mitochondrial ROS in macrophages.

Activation of the NLRP3 inflammasome can occur through several pathways and usually requires two signals. The first is a priming signal that can be provided by microbial components such as LPS or endogenous molecules that engage Nuclear Factor-κB (NF-κB) pathway to induce the expression of pro-IL-1β and NLRP3 [25, 26]. The second is a licensing signal that can be provided by extracellular ATP to directly activate NLRP3 and that can also involve cellular signals such as potassium efflux, pore formation in cell membranes, lysosomal damage, or the elevation of ROS and mitochondrial damage [26-32]. Mitochondrial production of ROS is especially important as a central and common upstream of cellular signals for NLRP3 inflammasome activation. The importance of ROS is reflected by the observations that most NLRP3 agonists induced ROS, and ROS blockade by chemical scavengers suppressed inflammasome activation [14-16]. Also, NLRP3 inflammasome were activated by the release of oxidized mitochondrial DNA into the cytosol or by the blockade of mitophagy/autophagy that caused accumulation of damaged, ROS-generating mitochondria [30, 31]. In addition, activation of the NLRP3 inflammasome was suppressed by autophagy proteins that inhibited the release of mitochondrial DNA [32]. Considering the critical role of mitochondrial ROS in NLRP3 activation, suppression of mitochondrial ROS by MSCs might account for their negative regulation of the NLRP3 inflammasome that was observed in our study.

STC-1 is a 247 amino acid glycoprotein that was originally identified as a calcium/phosphate regulatory protein in fish [33]. Although the mechanism of STC-1 has not been clarified, STC-1 in mammalian has been shown to have multiple effects including protection against ischemia, [34, 35] suppression of inflammatory responses [36, 37] or reduction of ROS [37-40], and the subsequent apoptosis in alveolar epithelial cancer cells and retinal ganglion cells [40, 41]. Of note, STC-1 is secreted by hMSCs in response to signals from apoptotic cells and mediates an antiapoptotic action of hMSCs [41]. Additionally, we here found that hMSCs produced a large amount of STC-1 in response to LPS/ATP-treated macrophages, and STC-1 mediated the function of hMSCs in suppressing mitochondrial ROS and NLRP3 inflammasome activation in macrophages. These results are also consistent with the emerging consensus that MSCs modulate immune responses predominantly through the production of trophic factors in response to microenvironmental cues [1-5].

One of the observations made here was that inflammasome-activated macrophages cocultured with MSCs expressed high levels of TSP-1. TSP-1 is a multifaceted protein involved in modulation of inflammation/angiogenesis and enhancement of inflammation resolution/tissue homeostasis [42]. Further investigation of the role of TSP-1 induced in macrophages by MSCs would help elucidate the mechanisms of MSCs in modulating macrophages and innate immune system.

As cellular regulators of inflammasomes, previous reports demonstrated that effector and memory CD4+ T cells inhibited inflammasome-mediated caspase-1 activation and IL-1β release through ligands of the Tumor Necrosis Factor (TNF) superfamily or by Interferon-γ (IFN-γ) signaling [43, 44]. These studies indicate that T cells or adaptive immune system suppress potentially damaging inflammation by acting as a feedback regulator of inflammasome responses in macrophages. Our study suggests that MSCs are one of cellular regulators that keep the inflammasome under control to prevent excessive inflammatory responses. To further explore the clinical relevance of our findings in diseases, in vivo study would be necessary.

Conclusion

In conclusion, our data demonstrate that hMSCs negatively regulate the NLRP3 inflammasome in macrophages primarily by secreting STC-1 in response to crosstalk with activated macrophages and thus by decreasing mitochondrial ROS. Our observations not only provide a novel mechanism of MSCs by which the cells modulate macrophages but also suggest a role of MSCs as regulators that maintain innate tissue homeostasis.

Acknowledgments

This work was supported by a grant of the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A111609).

Author Contributions

J.Y.O.: conception and design, financial support, provision of study material, collection and assembly of data, data analysis and interpretation, manuscript writing, and final approval of manuscript; J.H.K.: provision of study material, collection and assembly of data, and data analysis and interpretation; H.J.L.: and J.M.Y.: provision of study material and collection and assembly of data; H.C., M.K.K., and W.R.W.: conception and design; D.J.P.: conception and design, data analysis and interpretation, manuscript writing, and final approval of manuscript.

Disclosure of Potential Conflict of Interests

The authors indicate no potential conflict of interests.

Ancillary