SEARCH

SEARCH BY CITATION

Keywords:

  • Mesenchymal stem cells;
  • NK cells;
  • TLR;
  • NKG2D ligands;
  • Immunotolerance

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

Mesenchymal stem cells (MSCs) play a fundamental role in allograft rejection and graft-versus-host disease through their immunosuppressive abilities. Recently, Toll-like receptors (TLR) have been shown to modulate MSC functions. The aim of this study was to investigate the effects of several TLR ligands on the interaction between MSC and natural killer (NK) cells. Our results show that TLR-primed adult bone marrow and embryonic MSC are more resistant than unprimed MSC to IL-2-activated NK-induced killing. Such protection can be explained by the modulation of Natural Killer group 2D ligands major histocompatibility complex class I chain A and ULBP3 and DNAM-1 ligands by TLR-primed MSC. These results indicate that MSCs are able to adapt their immuno-behavior in an inflammatory context, decreasing their susceptibility to NK killing. In addition, TLR3 but not TLR4-primed MSC enhance their suppressive functions against NK cells. However, the efficiency of this response is heterogeneous, even if the phenotypes of different analyzed MSC are rather homogeneous. The consequences could be important in MSC-mediated cell therapy, since the heterogeneity of adult MSC responders may be explored in order to select the more efficient responders. Stem Cells 2014;32:290–300


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

At present, the definition of a mesenchymal stem cell (MSC) is based on the characteristics admitted by the International Society for Stem Cell Therapy (ISCT) [1]. Indeed, MSCs are characterized by multipotent differentiation abilities in several lineages, fast proliferation, and specific mesenchymal markers. These features, associated with their immunoregulatory properties, through cell-cell contact or secretion of soluble factors [2], may explain the role played by bone marrow-derived MSC (BM-MSC) in allograft rejection and graft-versus-host disease (GvDH) [3-5]. In addition, BM-MSCs express human leukocyte antigen (HLA) class I molecules that facilitate their escape to natural killer (NK) cells [6, 7]. In fact, the primary function of NK cells is the immunosurveillance and the absence or altered expression of major histocompatibility complex (MHC) class I molecules render target cells susceptible to NK cell attack [8-10]. Despite their HLA class I expression, BM-MSCs are recognized and killed by cytokine-activated NK cells, even if they can also strongly inhibit NK cell cytotoxicity [6, 7, 11]. The complex interaction between NK cells and BM-MSC is not completely understood, but it seems clear that the main mechanism of this interaction is mediated by the activating receptors on NK cells and activating ligands expressed by MSC. In particular, the natural killer group 2D (NKG2D) is not only critically involved in immune surveillance against malignant cells by recognizing stress-inducible MHC class I chain-related protein (MIC)-A and MICB and the UL16-binding proteins (ULBP)1-4 family expressed on tumor cells, but also on NK/MSC interaction [6, 12]. Recently, it has been reported that MSC derived from adult bone marrow also express functional Toll-like receptors (TLR) (in particular TLR-3 and TLR-4), that promote their survival and proinflammatory cytokine secretion [13-18]. TLR are also expressed on fibroblasts, macrophages, dendritic cells, epithelial cells, and lymphocytes. They recognize viral double-stranded RNA (dsRNA), single-stranded RNA (ssRNA), and bacterial products and they can be expressed on the cell surface (TLR1, 2, 4, 5, and 10) or require the internalization of the virus and its replication to release the viral RNA into endosomes (TLR3, 7, 8, and 9) [19, 20]. Beside TLR-3, TLR ligands activate the Myeloid differentiation primary response gene (88) (MyD88) and a complex network of adapter molecules that includes IL-1 receptor-associated kinase and the tumor necrosis factor-receptor-associated factor-6, which leads to the NF-kB nuclear translocation and the consequent inflammatory response. By modulating cell activity, TLRs play a critical role in immune response [21] and in allograft rejection [22, 23]. MSCs derived from bone marrow, cord blood, adipose tissues, and Wharton's jelly express several TLR that promote their viability, proliferation, and cytokine secretion [13, 15, 16, 24]. In addition, recently Waterman et al. [18] have demonstrated that TLR triggering polarizes MSC toward an immunosuppressive (MSC2) or a proinflammatory (MSC1) phenotype (TLR3 and TLR4, respectively). Noteworthy, MSC1 (TLR4) supports PBMC activation while unprimed MSC and MSC2 (TLR3) suppress it. Moreover, several papers report diverging effects of TLR on MSC regulatory functions [15, 16, 18, 25], but these different results are probably linked to different culture conditions and experimental procedures. Because MSC, TLR, and NK cells play a critical role in GvDH treatment [5, 22, 26, 27], this study was conducted in order to evaluate the effects of several TLR ligands on MSC/NK interaction. In addition, since TLR ligands apparently modulate sources of MSC other then those derived from bone marrow, we compared the effects of TLR ligands also on human embryonic stem cell lines (H1 and H9)-derived MSC (ES-MSC), which have already been demonstrated to display inhibitory properties against immune cells [28-30]. In conclusion, this study indicates that the proinflammatory environment induced by TLR ligands leads to the modulation of surface expression and secretion of MICA by primed MSC. Moreover, MSCs, behaving like tumor cells, exploit modulation of MICA expression/secretion MICA not only to protect themselves against activated NK cells but also to inhibit NK cytolytic functions.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

Reagents

Monoclonal antibody anti-ULBP3, CD155 (PVR), CD112 (nectin2), and MICA were obtained from R&D Systems. Anti-CD107a and CD3 were purchased from Becton Dickinson, whereas anti-CD56, CD16, and NKp46 were purchased from Beckman Coulters. Lipopolysaccaride (LPS) (Escherichia coli, O55:B5) and the synthetic dsRNA analog polyinosinic:polycytidylic acid (poly(I:C)) were purchased from Sigma, whereas R848 was purchased from InVivoGen. IL-2 was purchased from Peprotech. The broad-spectrum matrix metalloproteinase (MMP) inhibitor GM6001 was a kind gift from Eric Rubinstein (Inserm U1004, Villejuif, France) and used at 5 μM working concentration. Monensin and Brefaldin A were purchased from Sigma. Endotoxin concentrations are routinely measured by the manufacturer's with the Limulus Amebocyte Lysate test, as mentioned in the notice. The same lot of medium culture and fetal calf serum (FCS) was used in this study.

MSC Generation and Priming

Bone marrow aspirates were drawn from the iliac crest of donors from CTSA Hospital Percy, Clamart (France). Informed written consent was obtained using an Institutional Review Board approved protocol from the Ethics committee. Briefly, bone marrow mononuclear cells were plated at 1.6 × 105 cells per square centimeter in α-MEM (Life Technologies, France) supplemented with 2 mM l-glutamine, 1% penicillin/streptomycin solution (all from Gibco, France), and 10% heat-inactivated FCS (PAA Laboratories, France) (complete medium) at 37°C in air with 5% CO2. After 3 days, nonadherent cells were discarded and the cells were plated at 4 × 103 cells per square centimeter in a α-MEM complete medium supplemented with basic fibroblast growth factor (1 ng/ml). The medium was changed twice a week. MSC were also generated from two ESC lines H1 and H9 (WiCell Research Institute, Madison, WI, http://www.wicell.org) (Authorization from French Biomedicine Agency N°: RE10–035 R/C), as previously described [28]. MSCs were characterized according to the ISCT criteria and as previously described [1]. Briefly, adherent cells after 3–4 weeks of culture were characterized by flow cytometry checking the expression of CD90, CD105, CD146, CD54 (ICAM-1), CD73, CD34, CD45, HLA class I (HLA-ABC), and class II (HLA-DR) and the pluripotent markers Oct4, Sox2, Nanog, and Lin28. Multipotency of MSC was tested for differentiation along the osteogenic, chondrogenic, and adipogenic lineages. In other experiments, adult and embryonic MSCs were primed with poly(I:C) (25 μg/ml, TLR3) and R848 (1 μg/ml, TLR7/8) (both from Invivogen) and LPS (1 μg/ml, TLR4) (Sigma) for 48 hours and subsequently analyzed for cell surface markers by flow cytometry and cytokine secretion by ELISA.

Cell Lines

The human cervical carcinoma cell lines HeLa and human erythroleukemia K562 cells used in NK killing and degranulation ability were cultured at 37°C with 5% CO2 in complete RPMI-1640 medium (10% FCS, antibiotics, glutamine, and fungizone).

Isolation of Peripheral NK Cells

NK cells were isolated from peripheral blood mononuclear cells (PBMC) of healthy donors (n = 8) with the specific RosetteSep kit (StemCell Biotechnologies), according to the manufacturer's instructions. The resulting purified CD56+/CD3− (>95%, as checked by flow cytometry) NK cells were activated with IL-2 (100 U/ml) (0.75 × 106 cells per milliliter) for 48 hours and used against TLR-activated MSC in a 51Cr release assay. The HLA class I- K562 cell line was used as a target in the NK-cell cytotoxicity analysis in a 51Cr release assay.

Flow Cytometry and CD107a Degranulation

For surface-marker expression, cells were stained with saturating concentrations of the appropriate mAbs or isotype-matched control Ab, for 30 minutes at 4°C in the dark. NK cells were also tested for their cytolytic activity against NK susceptible target cells (ratio NK/target, 5:1) by quantifying surface expression of CD107a, as previously described [28]. Monensin (2 mM) was added for the last 4 hours to inhibit granules secretion. All samples were analyzed on a FACScalibur flow cytometer (BD), and data analysis was done using FlowJo software (TreeStar, Inc.).

51Chromium-Release Assay

MSCs were used as target cells in a conventional 51Cr-release assay. NK cells were mixed with 51Cr-labeled MSC (2,500 per well) at different effector-to-target (E/T) ratios. After 4 hours, the supernatants were harvested and transferred to LumaPlate 96-well plates (Perkin–Elmer), and were dried and counted on a Packard's Top Count NXT. The percentage of specific lysis was calculated as follows: % specific lysis =[counts per million (cpm) sample − cpm spontaneous/cpm max. − cpm spontaneous] × 100. Each point represents the average of triplicate values. Maximal release was obtained with 2% Triton X-100. NK-sensitive K562 cell line was used as positive control. For blocking experiments, the anti-MICA and anti-HLA-I (both from eBioscience), anti-HLA-E, NKG2D (both from Biolegend), DNAM-1, and ULBP3 (both from R&D Systems) were added (10 μg/ml) to primed MSC 30 minutes before the coincubation with NK cells.

ELISA

BM-MSC and ES-MSC were cultured for 48 hours with various TLR agonists (poly(I:C) at 25 μg/ml; LPS at 1 μg/ml and R848 at 1 μg/ml). The level of sMICA, as well as IL-1beta, MMP2, MMP3, and TGFβ in cell culture supernatants, was determined using the Human DuoSet ELISA Development Kit (R&D Systems), following the manufacturer's instructions. In addition, PGE2 (Arbor Assays), HLA-G (Exbio), MMP9 (Quantikine R&D Systems), IL-6, IL-8, and vascular endothelial growth factor (VEGF) (all eBioscience) productions by MSC were analyzed. The absorbance was measured at 450 nm. Results are shown as means ±SD of triplicates.

Statistical Analysis

Evaluation for ELISA and cellular cytotoxicity results was performed at least three times. Data are reported as mean ± SD. All calculations were carried out using Graph Pad Prism version 4.00 (Graph Pad Software, San Diego, CA). Comparisons between groups were made using paired t test. Values of p < .05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

Adult and Embryonic MSC Express Functional TLRs

Human bone marrow MSC (n = 10), cultured between passage 2 and 4, and human embryonic MSC, cultured between passage 10 and 15 (n = 2) (BM-MSC and ES-MSC, respectively), were analyzed for their TLR expression by flow cytometry. Both MSCs expressed intracellular levels of TLR-3 and surface levels of TLR-4, while intracellular expression of TLR-7 was almost undetectable (Fig. 1A). The functionality of these TLRs was evaluated using TLR ligands (poly(I:C), LPS, and R848 for TLR-3, −4, and −7, respectively) which promoted, after 48 hours of stimulation, MSC viability (MTT assay) (Fig. 1B), and proliferation (Fig. 1C, 1D). Interestingly, LPS appeared to be the best stimulator on both MSCs proliferation.

image

Figure 1. Expression of TLR in MSC and effect of TLR ligands on MSC proliferation. (a) MSC derived from adult bone marrow and embryonic cells were checked for TLR expression. In particular, flow cytometry analysis revealed surface expression of TLR4 and intracellular expression of TLR3. The gray curves indicate the corresponding negative mouse IgG1 or IgG2a antibodies. One independent representative experiment is shown. For intracellular staining, MSC were fixed in PFA 2% and permeabilized with PBS/saponin 0.1%. In addition, we evaluated the effects of TLR ligands (48h) in their viability (MTT assay) (b) and proliferation (c-d). Time course of cell growth in culture of BM-MSC and ES-MSC: 5x104 MSC were seeded in 24 wells plate, and cell count of MSC was performed after 24h and 48h. Averages of three independent experiments are shown. Poly (I:C), LPS and R848 were used at 25mg/ml, 1mg/ml and 1mg/ml, respectively.

Download figure to PowerPoint

Proinflammatory Factor Secretion in TLR-Primed MSC

Since a common mechanism used by BM-MSC to inhibit immune responses is the secretion of soluble factors [2], we investigated whether TLR ligands could modulate this mechanism. Our results show that poly(I:C) (TLR-3) and LPS (TLR-4), and R848 (TLR7/8) although at a very lesser extent, induced proinflammatory and proangiogenic cytokines (IL-6, IL-8, and VEGF) secretion by both MSCs (Fig. 2). In addition, an increase of IL-1-β and PGE2 in MSC supernatants was observed. Noteworthy stimulation of MSCs with TLR ligands did not increase their secretion of important immunoregulatory factors such as HLA-G or TGFβ whose secretion was even decreased following TLR priming.

image

Figure 2. TLR ligands induce pro-inflammatory secretion by MSC. Results are expressed as mean of a) IL-6, IL-8, IL-1b, VEGF, PGE2, HLA-G and TGF-b. Soluble factors levels were measured by ELISA assay in independent experiments performed using different MSC populations. Data are expressed in pg/ml (IL6, IL8, VEGF, PGE2 and TGF-b) or U/ml (HLA-G), respectively. * denote a statistically significant difference between TLR-primed MSC when compared to untreated MSC. ***, p<0.0001; **, p<0.001 and *, p<0.05.

Download figure to PowerPoint

TLR Ligands Protect MSC from NK Attack

Since adult and embryonic MSC can be susceptible to IL-2-activated NK killing [6, 7, 11, 28, 30] and TLR ligands (in particular TLR-3 and −4) might modulate BM-MSC immunosuppressive properties [15, 25], we investigated the effect of these ligands on MSC/NK cell interactions. Our results show that TLR-3 render ES-MSC more resistant to IL-2-activated NK killing, as compared to untreated cells. In contrast, no significant protection from activated NK cells was observed after TLR-4 and TLR-7/8 priming (Fig. 3A). Comparable results were obtained with BM-MSC. Moreover, lysosomal-associated membrane protein-1 (CD107a) degranulation by activated NK cells was significantly decreased after contact with TLR-stimulated MSC (in particular TLR-3), in comparison with untreated MSC (Fig. 3B). Leukemic K562 and HeLa carcinoma cell lines were used as positive controls in the CD107a induction assay. Nonactivated (resting) NK cells, which were almost unable to express CD107a, were used as negative control (Fig. 3C).

image

Figure 3. TLR3 ligands triggering protect adult and embryonic MSC from NK killing. a) Purified NK cells were cultured for 48h in IL2 (100U/ml) and then co-cultured for 4h with TLR-primed MSC in the chromium release assay (NK/MSC ratio, 30:1). b) In other experiments, NK cells were cultured with adult and embryonic MSC (ratio 5:1) for 4h to evaluate their ability to release cytotoxic granules. For CD107a degranulation, an anti-CD107a antibody and monensin (5mg/ml) were added at the beginning of the co-culture. Then cells where washed in PBS and stained with an anti-CD56 antibody. Non-activated (resting) NK cells were used as negative control, whereas K562 and HeLa cells were used as positive control for CD107a degranulation, respectively. All experiments were carried out at least six times. Horizontal scale bars show the median value.

Download figure to PowerPoint

MHC Class I Molecules Expressed on MSC Do Not Play a Major Role in TLR-Induced MSC Protection

To investigate the mechanism associated with the TLR-induced MSC resistance against NK cells, we examined the expression of activating and inhibitory ligands of NK cells expressed on MSC surface. NK cell responses are mainly mediated by direct cytolysis of target cells and soluble cytokines production (such as IFN-γ), and their targets recognition is regulated through a balance of activating or inhibitory signals [31, 32]. Usually, MHC class I molecules function as inhibitory signals for NK cells [10]. In this context, it has been shown that NK-mediated lysis was inhibited when IFN-γ-exposed BM-MSCs were used as target cells. This inhibition was a consequence of the upregulation of MHC class I molecules at the MSC surface [7]. Herein, we observed that TLR3 and TLR4 but not TLR7/8 stimulation (Fig. 4A) resulted in an increased surface expression of HLA-I molecules on both types of MSC, compared to control cells. Interestingly, as shown in Figure 4B, incubation with anti-HLA-I mAbs did not significantly increase cytotoxic efficiency of NK cells against unstimulated or LPS-stimulated MSC, while poly(I:C)-mediated class I molecules upregulation was involved (p = .0410 and p = .0611 for TLR3 and TLR4 in BM-MSC, and p = .0210 and p = .0750 in ES-MSC, respectively) in MSC protection. The fact that LPS triggering increased HLA-I expression but did not protect MSC confirms that the absence of biological response could be due to a poor HLA modulation, which in contrast is found after strong triggering as TLR3 and IFN-γ, as reported by Spaggiari et al. [7]. These results suggest also that other factors or molecules might be involved in MSC protection.

image

Figure 4. Role of HLA class I molecules in NK cell-mediated lysis of adult and embryonic MSC. a) FACS analysis of human leukocyte antigen (HLA) class I molecule expression on BM-MSC and ES-MSC treated with TLR ligands for 48h. Results are expressed as Mean Fluorescence Intensity (MFI) values and an average of five independent experiments is shown. b) MSC was used in a 51Cr release assay with activated NK cells. To determine the impact of HLA class I molecules in MSC killing, these latter were incubated with an anti-HLA class I blocking mAb or with an isotype matched control mAb (10mg/ml), 30 min prior to co-incubation with activated NK cells. Results are shown as the mean±SD of triplicate samples.

Download figure to PowerPoint

NKG2D Ligands Expressed on MSC Are Modulated by TLR Ligand Stimulation

Modulation of MICA and ULBP families represents an important mechanism used by tumor cells to escape from immunosurveillance [9, 33-36]. MICA is also expressed on adult MSC [6], but its expression on embryonic MSC has not been investigated. In our experiments, we detected a relevant surface expression of MICA not only in adult but also in embryonic MSC. TLR-3 treatment induced MICA downregulation in embryonic MSC, while an extreme heterogeneity was observed among the BM-MSC since in four out of eight we observed a decreased expression of MICA, while in three of eight samples MICA expression was not modified and in one of eight BM-MSC cultures TLR ligands even upregulated MICA expression (Fig. 5A). In addition, since tumor cells evade immune responses by shedding surface MICA, which results in impaired NKG2D functions on NK cells [37, 38], we determined the levels of soluble MICA (sMICA) protein in the culture supernatants of TLR-primed MSC. Although MSC (in particular adult MSC) showed certain heterogeneity in MICA secretion, 48 hours treatment with TLR ligands (TLR-3 in particular) significantly increased the levels of sMICA in the supernatant of embryonic MSC and in the supernatants of 4/8 BM-MSC (Fig. 5B). The involvement of MICA in MSC protection from NK killing, prompted us to investigate the putative mechanism involved in MICA shedding. Results depicted in Figure 1A, 1B Supporting Information show a pronounced reduction of sMICA release by TLR-primed MSC cultured with GM6001 (an inhibitor of MMP), which was associated with a significant restoration of MICA surface expression. Interestingly, MMPs-dependent MICA shedding was associated with increased detection of MMP3 and MMP9 in TLR-primed supernatants. By contrast, MMP2 was not significantly modulated by TLR stimulation (Fig. 2 Supporting Information).

image

Figure 5. NKG2D ligands modulation by TLR ligands in adult and embryonic MSC. a) FACS analysis of NKG2D ligands MICA expression on MSC treated with TLR ligands for 48h. This graph is shown for both MSCs to explain the strong heterogeneity between different donors, in particular in adult MSC. b) Results are expressed as dot plots for soluble MICA levels measured in independent experiments performed using different MSC populations. Each dot represents a different value for sMICA secreted by MSC. Data are expressed in pg/ml. Horizontal scale bars show the median value.

Download figure to PowerPoint

MICA Expressed on MSC Plays a Critical Role in MSC Protection

Subsequently, we studied the role of MICA in MSC/NK interaction. Our data show that, unprimed ES-MSC and BM-MSC were efficiently killed by IL-2-activated NK cells and that use of anti-MICA neutralizing mAbs counteracted this effect. In addition, after TLR priming, the residual surface expression of MICA still protected MSCs by NK cells, which were restored by the anti-MICA neutralizing. These results show that TLR-primed MSCs enhance their capacity to escape NK surveillance, through MICA modulation (Fig. 6). Moreover, the correlation between MICA expression and % of MSC lysis was statistically not significant, confirming that other ligands expressed on MSC (or the respective counterpart expressed by NK cells) participates to the NK/MSC interaction.

image

Figure 6. Role of MICA in MSC protection. a-c) To determine the impact of MICA in MSC killing, the latter were incubated with an anti-MICA blocking mAb or with an isotype matched control mAb (10mg/ml), 30 min prior to co-incubation with activated NK cells. Results are shown as the mean±SD of triplicate samples. b-d) Spearman correlation was used to show the correlation between MICA expression and % of MSC lysis. Each dot represents a different value for MSC lysis. Horizontal scale bars show the median value.

Download figure to PowerPoint

Activating and Inhibitory Ligands Expressed by MSC Participate to the MSC Protection

Since MICA was only partially involved in TLR-induced MSC protection by NK cells, we investigated the possible role of other activating and/or inhibitory ligands expressed in NK cells and MSC as already reported by others [6, 7]. Our results show that NKG2D expressed by NK cells plays an important role in the NK/MSC interaction, and inhibition by specific mAbs of DNAM-1 and ULBP3 (a NKG2D ligand) significantly modified NK-induced MSC lysis, whereas inhibition of anti-HLA-E (a NKG2A ligand) did not modified the efficiency of NK killing (Fig. 7). Interestingly, after TLR triggering we did not observed a further role of this ligands in MSC protection. In addition, CD112 (a DNAM-1 ligand) and ULBP3, but not CD155 (another DNAM-1 ligand), were slightly increased by TLR3 treatment (Fig. 3, Supporting Information).

image

Figure 7. Role of other activating and inhibitory receptors on NK/MSC interaction. To determine the impact of other molecules in MSC killing, NK cells were incubated with an anti-NKG2D and an anti-DNAM1, while MSC were incubated with an anti-HLA-E and anti-ULBP3 blocking mAb or with an isotype matched control mAb (10mg/ml), 30 min prior the 51Cr release assay (ratio NK/MSC, 25:1).

Download figure to PowerPoint

TLR-Primed MSCs Display Inhibitory Properties Against NK Cells

Finally, we evaluated the role of sMICA on NK activity, culturing NK cells with adult and embryonic MSC by cell-cell contact, transwell (TW), and TLR-primed MSC supernatants (SN) (50%) for 6 days. Medium was supplemented with IL-2 and an anti-MICA mAb (10 μg/ml) at the start of the culture. Our results show that upon cell-cell contact unprimed MSC strongly inhibited NK cytotoxicity against K562 cells, and that sMICA was significantly involved for ES-MSC but not for BM-MSC (p = .0321 and p = .1212 for ES-MSC and BM-MSC, respectively). By contrast, TLR3-primed MSC enhanced their immunosuppressive functions, and sMICA was significantly involved (p = .0133). TLR4 priming did not modulate MSC immunosuppression. Similar results were obtained in the TW system. Interestingly, TLR-3-primed MSC SN, but not control SN and TLR-4-primed MSC SN, were able to inhibit NK by 10%–15% cytolysis against K562 target cells, while in the presence of anti-MICA neutralizing mAbs, we noticed a partial recovery of NK cytolytic activity against K562 (Fig. 8). Our results highlight that TLR ligands induce not only significant amounts of sMICA in primed MSC but also that this sMICA plays some role in the inhibition of NK cell functions.

image

Figure 8. TLR3-primed MSC display enhanced inhibition against NK cells. Purified NK cells were cultured in IL2 (100U/ml) in presence of MSC by cell-cell contact, Transwell (TW) or with MSC supernatants (SN). In the last case, SN of primed MSC was harvested and MSC were cultured for further 24h with fresh RPMI complete medium. 50% of MSC SN was added to NK cells. After 6 days, NK cells were used in a 51Cr release assay against K562 cells (25:1). To determine the impact of soluble MICA molecules in K562 killing, an anti-MICA neutralizing mAb (10mg/ml) was added at the beginning of the culture. Results are expressed as MFI values and an average of six independent experiments is shown.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

MSCs play a critical role in GvDH prevention and graft rejection through their immunosuppressive properties and their capacity to adapt to inflammatory conditions [2]. Recently, interest has risen in TLRs, an important “system alert” which controls immune responses [21]. Although it is stated that TLR triggering promotes several MSC functions [14], the effects of TLR ligands on MSC immunoregulatory functions are still confusing [15, 16, 18, 25]. In this study, we show for the first time that MSC isolated from different sources (bone marrow and embryonic lines) became resistant to activated NK cells after TLR priming (in particular TLR-3). To explain this TLR-induced protection, we first evaluated the role of MHC class I molecules on MSC/NK interactions. While MHC class I molecules seem to play an important role in the protection from NK killing of IFN-γ-primed MSC [7], our results suggest that TLR3-induced HLA-I upregulation expression on MSC does not play a major role in MSC protection against NK cells. Interesting is the fact that LPS triggering increased HLA-I expression but without protecting MSC, by suggesting that a poor HLA modulation is not enough to protect MSC from NK cells. These results led us to investigate other molecules known to actively participate to the adult MSC/NK interaction: namely MICA and other NKG2D ligands [6]. Using anti-MICA neutralizing mAbs, we observed that NK cytolysis against unprimed adult and embryonic MSC was reduced. Interestingly, TLR-primed MSCs display an opposite behavior in comparison with monocytes and macrophages. In fact, nonactivated monocytes and macrophages are resistant to autologous NK cytolysis, but they became susceptible to NK killing upon TLR triggering. The explanation seems to be linked to the upregulation of MICA expression in both type of cells [39-42]. In MSC, our data indicate that their TLR priming induced a heterogeneous modulation of MICA expression that seems to be donor-dependent, especially concerning BM-MSC. Indeed, in half of adult MSCs and in all ES-MSC used in this study, TLR priming induced shedding of MICA molecules associated to the detection of sMICA in their supernatants. These effects are controlled by metalloproteases activation (MMP) induced by TLR ligands. In several adult MSC, TLR priming induced an MMP-dependent decreased surface expression of MICA that was not associated to the detection of sMICA. Moreover, in 3/8 adult MSC, we did not detect, after TLR priming, decreased surface expression of MICA and induction of sMICA. Finally, in 1/8 BM-MSC, we even observed increased surface expression of MICA. Adult MSC resistant to TLR induced MICA down-modulation behaved as unprimed MSC, that is, they were highly prone to activated NK cells killing. Intriguingly, the residual MICA present on the surface of both types of MSC after MMP-induced shedding, acquired immunoregulatory properties opposite to those exhibited by the “constitutive” MICA expressed on unprimed MSC, protecting TLR-primed MSC by NK cytolytic activity. By acting as soluble antagonist, sMICA recalls a similar mechanism described in several cancer cells [43-45]. These data highlight the importance of MICA as a central player in the suppression of NK activity displayed by MSC that likely represents an additional possible mechanism of MSC to escape immune responses. In conclusion, this work shows a beneficial effect of TLR priming, by increasing MSC proliferation, viability, and cytokine secretion. In particular, we investigated the well-known “classical” soluble factors secreted by MSC involved in immune responses inhibition [11, 46]. In this context, we observed an increased production by both MSCs of proinflammatory and angiogenic factors after TLR priming, while among the so-called immunoregulatory soluble factors only the secretion of PGE2, known to inhibit NK cell functions [7, 28] was significantly increased. Moreover, our data indicate that MICA is part, with DNAM-1 and ULBP3 ligands (together with other immunoregulatory factors such as PGE2, MMPs and IL-1β) of a network, which contributes to assure immunoprotection of MSC against NK cytolytic activity in a context of a proinflammatory microenvironment mediated by TLR ligands. Despite their important role in GvDH treatment and in organ transplantation [3, 5, 26, 27], adult MSC frequently do not assure long-lasting immunoregulatory functions both in vitro and in vivo. In this context, our results show that embryonic MSCs of two donors have same efficient immunoregulatory properties [28], and abilities to adapt themselves in inflammatory conditions. However, heterogeneity of MSC behavior and differential response to TLR ligation should be taken in account. For this purpose, it will be important to define and select different types of MSC and type of priming, suitable for the clinical context and for cell therapy objectives.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

We describe in this work that TLR3-primed adult bone marrow and embryonic MSC are more resistant than unprimed MSC to IL-2-activated NK-induced killing. NKG2D ligand expressed on MSC (in particular MICA) and MMP is involved in this protection. These results indicate that MSCs are able to adapt their immuno-behavior in an inflammatory context, by decreasing their susceptibility to NK killing.

Acknowledgments

This work was supported by grants from INSERM, Association Vaincre le Cancer-NRB (Nouvelles Recherches Biomédicales), DIM STEM POLE, Région Ile de France (MEDICEN IngeCell programm), and the Université Paris Sud-11. This work was supported in part by grants from the Association pour la Recherche sur le Cancer (ARC) (no. 3272), Ligue contre le Cancer, as well as the Institut National du Cancer (INCa), and Agence Nationale de la Recherche (ANR). M.G. is recipient of post doctoral SANOFI fellowships. Authors thank Alessandro Poggi for helpful scientific discussion and Kyle Lund for English writing assistance.

Author Contributions

M.G. and A.B.G.: designed and performed experiments, analyzed data, and wrote the manuscript; A.N. and N.O.: performed experiments and analyzed data; S.C., B.A., A.D., and J.J.L.: designed research, analyzed and interpreted data, and wrote the manuscript. M.G., A.B.-G., A.D., and J.-J.L. contributed equally to this article.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Disclosure of Potential Conflicts of Interest
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
stem1563-sup-0001-suppfig1.tif758KSupporting Information Figure 1
stem1563-sup-0002-suppfig2.tif692KSupporting Information Figure 2
stem1563-sup-0003-suppfig3.tif554KSupporting Information Figure 3

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.