Mesenchymal Stem Cells Inhibit the Differentiation of Dendritic Cells ...

This information is current as of August 8, 2007

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://www.StemCells.com/cgi/content/full/25/8/2025

STEM CELLS®, an international peer-reviewed journal, covers all aspects of stem cell research: embryonic stem cells; tissue-specific stem cells; cancer stem cells; the stem cell niche; stem cell genetics and genomics; translational and clinical research; technology development. STEM CELLS® is a monthly publication, it has been published continuously since 1983. The Journal is owned, published, and trademarked by AlphaMed Press, 318 Blackwell Street, Suite 260, Durham, North Carolina, 27701. © 2007 by AlphaMed Press, all rights reserved. Print ISSN: 1066-5099. Online ISSN: 1549-4918.

Downloaded from www.StemCells.com at Biblioteca Di Medicina Veterinaria on August 8, 2007

Mesenchymal Stem Cells Inhibit the Differentiation of Dendritic Cells Through an Interleukin-6-Dependent Mechanism Farida Djouad, Louis-Marie Charbonnier, Carine Bouffi, Pascale Louis-Plence, Claire Bony, Florence Apparailly, Céline Cantos, Christian Jorgensen and Danièle Noël Stem Cells 2007;25;2025-2032; originally published online May 17, 2007; DOI: 10.1634/stemcells.2006-0548

TISSUE-SPECIFIC STEM CELLS Mesenchymal Stem Cells Inhibit the Differentiation of Dendritic Cells Through an Interleukin-6-Dependent Mechanism FARIDA DJOUAD,a,b LOUIS-MARIE CHARBONNIER,a,b CARINE BOUFFI,a,b PASCALE LOUIS-PLENCE,a,b ´ CLAIRE BONY,a,b FLORENCE APPARAILLY,a,b CELINE CANTOS,a,b CHRISTIAN JORGENSEN,a,b,c DANIE` LE NOE¨ La,b a

Institut National de la Sante´ et de la Recherche Me´dicale, U 844, Montpellier, France; bUniversite´ Montpellier, Unite´ de Formation et de Recherche de Me´decine, Montpellier, France; cCentre Hospitalier Universitaire Montpellier, Hoˆpital Lapeyronie, Unite´ Clinique d’Immuno-Rhumatologie, Montpellier, France

ABSTRACT Mesenchymal stem cells (MSC) are of particular interest for their potential clinical use in tissue engineering as well as for their capacity to reduce the incidence and severity of graftversus-host disease in allogeneic transplantation. We have previously shown that MSC-mediated immune suppression acts via the secretion of soluble factor(s) induced upon stimulation. The aim of this study was to identify the molecule(s) involved and the underlying mechanism(s). We show that murine MSC secrete high levels of interleukin (IL)-6 and vascular endothelial growth factor, which are directly correlated to the inhibition of T-cell proliferation. The T-cell activation is partially restored upon addition of a neutralizing anti-IL-6 antibody or the prostaglandin E2 inhibitor

indomethacin. Interestingly, no indoleamine 2,3-dioxygenase activity was detected in our conditions. Instead, we show that MSC reduce the expression of major histocompatibility complex class II, CD40, and CD86 costimulatory molecules on mature dendritic cells (DC), which was responsible for a decrease in T-cell proliferation. Moreover, we show that the differentiation of bone marrow progenitors into DC cultured with conditioned supernatants from MSC was partly inhibited through the secretion of IL-6. Altogether, these data suggest that IL-6 is involved in the immunoregulatory mechanism mediated by MSC through a partial inhibition of DC differentiation but is probably not the main mechanism. STEM CELLS 2007;25:2025–2032

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION Bone marrow stroma includes progenitor cells, commonly termed mesenchymal stem cells (MSC), that can support hematopoiesis and differentiate along multiple mesenchymal lineages, such as osteocytes, chondrocytes, adipocytes, and myocytes [1]. MSC are mainly isolated from bone marrow but have also been identified in other tissues including umbilical cord blood, adipose tissue, muscle, and synovial membrane [2]. Due to their differentiation capacities, they have emerged as a promising tool for therapeutic applications in tissue engineering as well as cell and gene therapy. Animal studies have shown that implantation of MSC can repair critical bone fractures in a rat model of femoral segmental defect [3], and they localize to the site of experimentally induced fractures after systemic injection [4]. Pilot clinical studies have demonstrated the feasibility of allogeneic bone marrow transplantation in the treatment of osteogenesis imperfecta. In this case, MSC do engraft and are able to generate donor-derived osteoblasts, which contribute to the improvement of the clinical signs associated with the disease and the enhancement of the total body weight [5]. Besides their multilineage potential, they also display immunoregulatory properties that have prompted consideration of their potential use in bone marrow transplantation. Indeed, a recent case report describes the successful use of MSC to treat severe grade IV

acute graft-versus-host disease in a patient after allogeneic hematopoietic stem cell transplantation [6]. The precise mechanisms underlying the immunosuppressive effect of MSC still remain to be clarified. The suppression of proliferation of T cells stimulated by allogeneic lymphocytes, dendritic cells, and mitogen, such as phytohemagglutinin or concanavalin A, has been well documented (for a review, see [7]). Both inhibition via cell contact [8] as well as the activity of soluble factors were shown to be involved in this process [9, 10]. Although controversial results have been reported, more convergent data now suggest that the suppressive activity of MSC is not associated with the secretion of hepatocyte growth factor or transforming growth factor-␤1 [8, 9], and that immune suppression may rely, at least in part, on the generation of CD8⫹ regulatory T cells [11]. Another mechanism has also been postulated where the immunosuppressive effect of MSC is based on indoleamine 2,3-dioxygenase (IDO) activity, which leads to impaired protein synthesis by depletion of the essential amino acid tryptophan [12]. More recently, it has been reported that MSC inhibit T-cell proliferation through the induction of division arrest tolerance [13]. Another study has shown that MSC altered the cytokine secretion profile of the various cell components of the immune response to induce a more anti-inflammatory or tolerant phenotype, with an increase of interleukin (IL)-10 and IL-4 secretion and a decrease of tumor necrosis factor (TNF)-␣ and interferon (IFN)-␥ production [14]. Finally,

Correspondence: Danie`le Noe¨l, Ph.D., Institut National de la Sante´ et de la Recherche Me´dicale, U 844, Hoˆpital Saint-Eloi, Baˆt INM, 80 Avenue Augustin Fliche, F-34091 Montpellier, France. Telephone: 33 (0) 4 99 63 60 26; Fax: 33 (0) 4 99 63 60 20; e-mail: [email protected] Received August 31, 2006; accepted for publication May 9, 2007; first published online in STEM CELLS EXPRESS May 17, 2007. ©AlphaMed Press 1066-5099/2007/$30.00/0 doi: 10.1634/stemcells.2006-0548

STEM CELLS 2007;25:2025–2032 www.StemCells.com


Key Words. Stromal cells • Tolerance/suppression/anergy • Cytokines

Interleukin-6-Mediated Immunosuppression by MSC

2026

MATERIALS

AND

METHODS

Cell Culture The murine C3H10T1/2 (C3) MSC line and NIH-3T3 fibroblasts were grown in complete Dulbecco’s modified Eagle’s medium (DMEM) (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) supplemented with 10% fetal calf serum (FCS) (HyClone, Logan, UT, http://www.hyclone.com), 2 mM glutamine, 100 U/ml penicillin, and 100 ␮g/ml streptomycin (Invitrogen, Carlsbad, CA, http://www. invitrogen.com). BMC9 cells were maintained in complete minimal essential medium (MEM) consisting in MEM␣ (Invitrogen) supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin, 100 ␮g/ml streptomycin. Murine primary cells (mMSC) were obtained from bone marrow from DBA1 mice and cultured in DMEM containing FCS, 2 mM glutamine, and 1 ng/ml basic fibroblast growth factor (bFGF) (R&D Systems Inc., Minneapolis, http:// www.rndsystems.com). Human primary MSCs (hMSC) were obtained from bone marrow aspirates and expanded at low density in complete MEM supplemented with 1 ng/ml bFGF, as previously described [11]. For macroarray analysis, human MSC were cocultured for 3 days in the presence of splenocytes from DBA1 and BALB/c mice (ratio 1:1:1) in the medium used for MLR (see below). When cultures reached near confluence, cells were detached with 0.05% trypsin and 0.53 mM ethylene diamine tetraacetic acid (EDTA) and subsequently replated at the density of 1,000 cells per cm2. Human peripheral blood mononuclear cells (hPBMC) were isolated from heparinized blood by centrifugation on a FicollHypaque cushion.

value 150. Analysis was performed using the Cluster and TreeView hierarchical clustering software developed by Eisen et al. [16]. Two filters have been used: one aimed at retaining only genes expressed above the median value and the second retaining genes for which the difference between maximum and minimum values was above twice the median value. Data were log transformed (log base 2) and genes were median centered and clustered by correlation average linkage clustering. The hierarchical clustering was visualized with TreeView.

Mixed Lymphocyte Reaction MLR were performed using mouse splenocytes obtained as already described [11]. Stimulator splenocytes (107 cells per milliliter) were treated with 50 ␮g/ml mitomycin C (Sigma) at 37°C for 45 minutes followed by five extensive washes with FCS-containing RPMI 1640 medium (Invitrogen). Responder splenocytes from BALB/c mice and stimulator splenocytes from DBA/1 mice were resuspended in RPMI 1640 medium containing 10% FCS, 2 mM glutamine, 100 U/ml penicillin and 100 ␮g/ml streptomycin, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 20 mM HEPES, and 5 ⫻ 10⫺5 M 2-mercaptoethanol (Invitrogen). Splenocytes were seeded in triplicates at the concentration of 105 cells per 100 microliters per well in 96-well round bottom plates. In some experiments, responder T lymphocytes were used after purification using the Mouse T-cell Negative Isolation Kit according to the recommendations of the supplier (Dynal, Compie`gne, France, http://www.invitrogen.com/ dynal). MSC (5 ⫻ 104 cells unless otherwise mentioned) were added to the MLR to obtain a 300-␮l final volume. The neutralizing BE-8 monoclonal antibody specific for human IL-6 [17] was a kind gift from B. Klein’s team (Institut National de la Sante´ et de la Recherche Me´dicale U 475, Montpellier) and used at the concentration of 9 ␮g/ml. When tested, two concentrations (5 and 125 ␮M) of indomethacin (Sigma) were added to the wells. After 3 days of incubation, 1 ␮Ci/well 3H-thymidine was added overnight, and thymidine incorporation was measured using a ␤ scintillation counter. The proliferative alloresponse, corresponding to the mean cpm of the allogeneic responder splenocytes, was attributed a 100% value. All experiments were performed in triplicates and repeated at least twice.

Cytokine Quantification by Enzyme-Linked Immunosorbent Assay Secretion of murine IL-1␤, IL-2, IL-4, IL-6, IL-10, IFN-␤, TNF-␣, and VEGF was determined in culture supernatants by specific enzyme-linked immunosorbent assays (ELISAs) (BD Biosciences). Prostaglandin E2 (PGE2) was quantified using a specific ELISA (R&D Systems). Human IL-6 produced in cell supernatant was quantified using the human IL-6 ELI-PAIRS kit according to the manufacturer’s recommendations (Diaclone, Besanc¸on, France, http://www.diaclone.com). Quantification was performed at least twice.

Isolation of Total RNA and cDNA Hybridization

Indoleamine 2,3-Dioxygenase Activity Measurement

Total RNAs from four separate primary cultures of human MSC and four cocultures of MSC and splenocytes were extracted using the RNeasy mini kit (Qiagen, Hilden, Germany, http:// www1.qiagen.com) according to manufacturer’s instructions. Radiolabeled cDNA was prepared from each RNA sample with the Atlas array kit (Clontech, Palo Alto, CA, http://www.clontech.com) by a reverse transcription step in the presence of ␣-[32P]dATP. The radiolabeled samples were hybridized to the Human Cytokine/ Receptor Atlas Nylon cDNA Expression Array (BD Biosciences, San Diego, http://www.bdbiosciences.com). After stringent washes, membranes were scanned using a PhosphorImager (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com).

Human and C3 MSC were stimulated with IFN-␥ (1,000 U/ml) for 48 hours in complete DMEM medium supplemented with L-tryptophan (100 ␮g/ml). IDO enzyme activity was measured by tryptophan-to-kynurenine conversion with photometric determination of kynurenine concentration in the supernatant as the readout as previously reported [18]. Briefly, 160 ␮l of cell supernatant was transferred to a 96-well culture plate, and 10 ␮l of 30% trichloroacetic acid was added for 30 minutes at 50°C. After centrifugation, 100 ␮l of supernatant was mixed to 100 ␮l of freshly prepared Ehrlich’s solution, and absorbance was read with a microplate reader at 450 nm.

Generation of DC and Coculture with C3 Gene Array Analysis Quantification was performed using the AtlasImage software (BD Biosciences). Data from each array were normalized by the median value to eliminate the variability due to the sample labeling or exposure duration. The normalized median was arbitrarily given the

The DC were obtained as previously described [11]. Briefly, DC were obtained from bone marrow progenitors after 6 days in culture with 1,000 U/ml recombinant granulocyte macrophage– colonystimulating factor and 1,000 U/ml recombinant (r)IL-4 (R&D Systems). The maturation was induced by addition of 1 ␮g/ml lipo-


it has been postulated that MSC may act by suppressing the differentiation of monocytes into mature dendritic cells (DC) impairing the stimulation of T cells [8, 15]. Most of these proposed mechanisms are actually not necessarily exclusive. We have previously shown that MSC-mediated immune suppression could act via the secretion of a soluble factor induced upon stimulation [11]. The aim of this study was to identify the molecule(s) involved in this process by analyzing the gene expression profile of MSC cocultured with splenocytes using the Atlas Human Cytokine/Receptor Array. We observed that MSC secrete higher levels of IL-6 and vascular endothelial growth factor (VEGF) when cultured in mixed lymphocyte reactions (MLR) and that T-cell activation is partially restored upon anti-IL-6 neutralizing antibody addition to the cultures. Moreover, we report that IL-6 is involved in the reversion of the maturation of DC to a less mature phenotype and in the partial inhibition of bone marrow progenitors to DC. Altogether, our data suggest that IL-6 is involved in the immunoregulatory mechanism mediated by MSC.

Djouad, Charbonnier, Bouffi et al.

2027

polysaccharides (LPS) (Sigma) for another day. The phenotype of DC was analyzed by flow cytometry using a FACScan cytofluorometer (BD Biosciences). When tested, MSC were added to the mature DC at the ratio 1:1 for 3 other days. Recombinant IL-6 (from 0.1–50 ng/ml) was added to the culture medium either on day 0 and every other day or on day 6. Conditioned supernatants were obtained after 48 hours of culture of C3 MSC, primary mMSC, or NIH-3T3 cells in the presence of murine splenocytes and filtration through a 22-␮M porous membrane (BD Biosciences). The anti-IL-6 antibody was used at the concentration of 9 ␮g/ml. In some experiments, 105 mature DC (mDC) were cultured with 3 ⫻ 105 allogeneic T lymphocytes and/or 2 ⫻ 105 C3 for 3 other days in six-well plates.

Phenotypic Characterization

Statistical Analysis Statistics were done with Student’s t test or a nonparametric MannWhitney test to compare data for statistical significance. All data were analyzed by the program Instat (GraphPad, San Diego, http:// www.graphpad.com).

RESULTS IL-6 and VEGF Are Upregulated by MSC Upon Coculture with Splenocytes We previously showed that the soluble factor(s), responsible for the immunosuppression induced by MSC, were secreted only when the latter cells were activated by allogeneic splenocytes [11]. In order to identify this/these molecule(s), we compared the transcriptional profile of primary human MSC cultured in the absence or in the presence of splenocytes using the Human Cytokine/Receptor Atlas Nylon cDNA Expression Array. Using this assay, which specifically detects human sequences, murine splenocytes were used as stimulator cells instead of human hematopoietic cells to avoid any contamination by irrelevant RNA. Using the hierarchical clustering analysis, three out of four samples of MSC cultured in the presence of splenocytes were clustered, indicating similarities in their pattern of gene expression. Interestingly, among genes that were overexpressed in these samples, expression of transcripts corresponding to the IL-6 precursor, the VEGF B precursor, and the IL-2 receptor ␥ were upregulated (Fig. 1A). The secretion of IL-6 by human MSC was confirmed at the protein level, and IL-6 concentration was increased 10-fold when these cells were cultured in the presence of murine splenocytes or hPBMC (Fig. 1B). Similarly, low levels of IL-6 and VEGF were secreted in the culture supernatant of the murine C3 MSC line (Fig. 1C), and these levels were increased by a 10-fold factor in a MLR. To determine whether the cytokines were produced by the MSC or the immune cells present in the culture, we recovered the immune cells and kept the adherent MSC in the culture wells. After thorough washes, cells were independently cultured for another www.StemCells.com

Figure 1. IL-6 secretion by MSC after stimulation. (A): Transcriptome analysis of human MSC cultured with/without splenocytes. (B): Secretion of IL-6 by human MSC cultured alone or in the presence of splenocytes or human peripheral blood mononuclear cells and phytohemagglutinin. (C): Secretion of IL-6 by C3 MSC or allogeneic splenocytes before and after stimulation in a mixed lymphocyte reaction and another independent culture for 48 hours. Abbreviations: BMP, bone morphogenetic protein; C3, C3H10T1/2; ERBB2, v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2; IL, interleukin; PBMC, peripheral blood mononuclear cell; PHA, phytohemagglutinin; TNF, tumor necrosis factor; VEGF, vascular endothelial growth factor.

48 hours. We observed that IL-6 was mainly produced by the C3 cells, since no significant increase was detected in the culture supernatants from splenocytes. In parallel, we found that VEGF was exclusively secreted by the MSC (Fig. 1C). No expression of the IL-2 receptor ␥ chain was observed on C3 MSC by fluorescence-activated cell sorting analysis, neither before nor after stimulation with splenocytes (data not shown). These results suggest that, in MLR, the MSC upregulate their secretion of IL-6 no matter what their species of origin, and the C3H10T1/2 cells were thus used for further experiments. We then investigated the possible role of other cytokines in the suppressive mechanism. IL-10 was never detected in the supernatants of MLR (data not shown). Among the tested cytokines, IL-1␤ and IL-4 levels were consistently detected in the MLR supernatants, either in the absence or the presence of C3 (Fig. 2A). In contrast, the production of the immunostimulatory cytokines TNF-␣, IFN-␥, and IL-2 induced by the coculture with allogeneic splenocytes was repressed when MSC were added. As we previously reported that the inhibition of T-cell proliferation was proportional to the number of MSC added in


For flow cytometry, MSC were harvested by treatment with 0.05% trypsin and 0.53 mM EDTA and resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin and 0.01% sodium azide. Cell aliquots (105 to 5 ⫻ 105 cells per 100 microliters) were incubated on ice with conjugated monoclonal antibodies against CD14, CD34, CD44, CD45, CD73, CD90, and CD105 (BD Pharmingen, San Diego, http://www.bdbiosciences.com/index_us.shtml) or conjugated isotypic controls. DC were recovered by centrifugation (300g, 5 minutes). The antibodies specific for DC (fluorescein isothiocyanate-labeled anti-major histocompatibility complex [MHC] class II and phycoerythrin [PE]-labeled anti-CD11c) and costimulatory molecules (PE-labeled anti-CD40, anti-CD80, anti-CD86) were purchased from BD Biosciences. Flow cytometry was performed on a fluorescence-activated cell sorter (FACScan) and data analyzed with the CellQuest software (BD Biosciences).

2028


secretion of VEGF was detected only when C3 MSC were present during the MLR and in stable amounts irrespective to the addition of the antibody. Figure 2. Cytokine expression in mixed lymphocyte reaction supernatants. (A): Cytokine profiles of murine C3 MSC, allogeneic splenocytes, or mixed cultures. (B): Proliferative response of allogeneic splenocytes cultured with increasing amounts of C3. (C): Cytokine profiles of cells cultured in (B) (results are the mean ⫾ SEM of two experiments in triplicate). Abbreviations: C3, C3H10T1/2; IFN, interferon; IL, interleukin; TNF, tumor necrosis factor; VEGF, vascular endothelial growth factor.

the well [11], we compared the amounts of secreted cytokines related to the number of MSC (Fig. 2B). Although IFN-␥ levels were high in the MLR, low amounts of IFN-␥ and high levels of IL-6 were detected when 105 C3 MSC were added (Fig. 2C). The decrease of IL-6 and VEGF was proportional to the decreasing number of MSC, whereas, in parallel, the presence of IFN-␥ increased in the culture supernatants. Together, these data show that the secretion of immunostimulatory cytokines is inversely correlated with the presence of MSC as well as with the degree of the inhibition of the T-cell response, in contrast to that of IL-6, the production of which is strongly upregulated.

IL-6 Is Part of the Immunomodulatory Effect Mediated by the MSC To investigate whether the overexpression of IL-6 was related to the suppressive activity mediated by the MSC, a neutralizing anti-IL-6 antibody was added to the MLR in the presence of C3 cells. As shown in Figure 3A, neutralization of IL-6 induced a significant increase of the allogenic T-cell proliferation. However, the T-cell proliferation did not completely recover, since it was still significantly lower than the allogeneic response observed in the presence of the anti-IL-6 antibody but in the absence of C3 cells. No increase of the lymphocyte proliferation was observed using an isotypic control (data not shown). In parallel, the secretion of IFN-␥ increased upon the addition of the anti-IL-6 antibody but to lower levels than those measured in an MLR in the absence of C3 MSC, confirming that the reversion of the T-cell inhibition was only partial (Fig. 3B). The

PGE2 Is Involved in the Suppressive Mechanism Induced by the MSC PGE2 has been shown to modulate a variety of immune functions in vitro and to upregulate IL-6 and VEGF secretion [19]. Because it has recently been reported to be expressed by hMSC and to increase upon coculture with hPBMC [14], we quantified the secretion of PGE2 by MSC in culture conditions or in MLR and examined the effect of the PGE2 inhibitor indomethacin. We could detect PGE2 in both situations with a slight increase in the MLR and, as expected, the addition of indomethacin decreased the PGE2 secretion (Fig. 4A). However, the addition of anti-IL-6 antibody has no effect on the secretion of PGE2 (data not shown). In the MLR, in absence of C3 MSC, the T-cell proliferation was stable or slightly enhanced in the presence of indomethacin. The addition of MSC resulted in the inhibition of 90% of the T-cell proliferation, which was statistically reversed (approximately fivefold) with the highest dose of indomethacin (Fig. 4B). This thereby demonstrates that secretion of PGE2 may be involved in the immunosuppressive mechanism induced by MSC. The addition of the neutralizing anti-IL-6 antibody did not further increase the inhibitory effects of indomethacin on MSC-mediated immune suppression, suggesting that there is no clear additive role of IL-6 and PGE2 for the suppression of the T-lymphocyte activation. We confirmed that the two concentrations of indomethacin used in this study were able to repress by half the secretion of IL-6, suggesting that PGE2 may act through the induction of IL-6 secretion (Fig. 4C). In contrast, the secretion of VEGF did not change upon addition of indomethacin and/or anti-IL-6 antibody (Fig. 4D), suggesting that VEGF is poorly involved in the suppressive effect of MSC. The production of IFN-␥, which reflects the proliferative activity of the T-cell population, decreased in the MLR with increasing doses of indomethacin and enhanced in the presence of MSC, concomitantly with the reversion of the immunosuppression (Fig. 4B, 4E). Altogether, these data indicate that PGE2 might be


Figure 3. Reversion of MSC-mediated immunosuppression by an antibody specific for IL-6. (A): Proliferative response of allogeneic splenocytes cultured with C3 MSC in the presence or absence of anti-IL-6 antibody. (B): IFN-␥ and VEGF secretion in supernatants from (A) (results are the mean ⫾ SEM of three experiments in triplicate). Abbreviations: C3, C3H10T1/2; IFN, interferon; IL, interleukin; VEGF, vascular endothelial growth factor.


2029

MSC line BMC9 or the human MSC-like cells F-Stro1A, whereas it was clearly measured in the supernatants of primary human MSC (data not shown), confirming previous data [12]. Consistent with the absence of IDO activity upon IFN stimulation, no activity was measured when splenocytes were stimulated by concanavalin A or allogeneic splenocytes, although the addition of MSC inhibited the proliferative response. Because C3 MSC do not display any IDO activity in MLR and still are able to inhibit the T-cell response, this enzymatic activity is likely not involved in the suppressive mechanism mediated by MSC, at least in the murine MSC line tested here.

Figure 4. Effect of the PGE2 inhibitor indomethacin on MSC-mediated immunosuppression. (A): Quantification of PGE2 in culture supernatants. (B): Proliferation of allogeneic splenocytes cultured with/without C3 MSC, indomethacin, or anti-IL-6 antibody (results are the mean ⫾ SEM of three experiments in triplicate; ⴱ, p ⬍ .05). Measure of IL-6 (C), VEGF (D), and IFN-␥ (E) levels in mixed lymphocyte reaction supernatants from (B). Abbreviations: C3, C3H10T1/2; IFN, interferon; IL, interleukin; ND, not determined; PGE2, prostaglandin E2; VEGF, vascular endothelial growth factor.

involved, at least partly, in the suppressive role mediated by MSC.

C3 MSC Do Not Express IDO Activity A recent study has reported that human MSC exhibit IDO activity that could be responsible for the T-cell unresponsiveness in MLR [12]. We thus investigated whether the C3 MSC line displays such enzymatic activity. Upon stimulation with IFN-␥, no IDO activity could be detected in the supernatants of C3 cells cultured for 48 hours in medium alone. Similarly, no IDO activity was measured in the supernatants of the murine www.StemCells.com

Recent data suggest that human MSC may suppress T-cell activation and proliferation by inhibiting differentiation of monocytes into DC and by downregulating the expression of costimulatory molecules on mDC [15]. To determine whether a similar mechanism may occur with the murine counterpart, we first investigated the effect of C3 on the expression of costimulatory molecules on mDC when both cells (ratio 1:1) were cocultured for 3 days. A slight but significant and consistent downregulation of the expression of CD40 and CD86 on MHC class II positive mDC was observed under these experimental conditions in the presence of MSC (94.74 ⫾ 1.27 vs. 78.35 ⫾ 4.99 with MSC and 96.45 ⫾ 2.87 vs. 87.97 ⫾ 1.27 with MSC, respectively). No significant changes in CD14 or CD11c expression were detected (0.78 ⫾ 0.81 vs. 2.17 ⫾ 5.06 with C3 and 59.25 ⫾ 28.59 vs. 50.42 ⫾ 11.83 with C3, respectively). The results suggest that the C3 MSC downregulate the expression of costimulatory molecules on mDC. This downregulation of costimulatory molecules was associated with a less efficient stimulation of lymphocyte proliferation (data not shown). We then wanted to determine whether IL-6 secreted by MSC may account for the lower levels of costimulatory molecules on mDC. Indeed, in humans, IL-6 has been proposed to inhibit the differentiation of monocytes to DC, rendering them functionally impaired [20, 21]. In the first step, we investigated whether various concentrations of recombinant IL-6 may have an impact on the generation of murine mDC. Indeed, the number of MHC II⫹/ CD11c⫹ mDC decreases with increasing doses of IL-6 (Fig. 5A, 5B). Because the dose of 50 ng/ml IL-6 was necessary to show a significant reduction of mDC number, we used this concentration in the following experiments. We then tested the addition of rIL-6 on DC generated in vitro from murine bone marrow progenitors, either during the early stages of DC generation or during maturation. To this aim, the cytokine was added on day 0, at the beginning of the culture, or on day 6, where maturation of DC is induced by addition of LPS. As shown in Figure 5C, the addition of IL-6 during the process of maturation did not change the phenotypic profile of DC. On the contrary, when IL-6 was added during the overall period of culture, a significant reduction of the DC markers, CD11c/MHC II, together with the costimulatory molecules CD40, CD80, and CD86 was consistently observed. In parallel, the level of IL-12 secreted by control mDC reached 3,521 pg/ml, whereas it falls down to 839 pg/ml when 50 ng/ml IL-6 was added to the culture. These data suggest that IL-6 interferes with the generation of DC by reducing the expression of costimulatory molecules but has little impact on the decrease of these molecules on mDC. In the second step, we investigated whether the secretion of IL-6 by the MSC may impair the generation of DC. In this experiment, to avoid the possible effect of culture density of the MSC/DC coculture, we used the conditioned supernatant obtained after coculture of MSC and murine splenocytes for 48 hours. The bone marrow progenitors were cultured for 6 days and then ana-


The Immunosuppressive Effect of MSC Is Associated with the Generation of Less Mature Dendritic Cells

2030


lyzed for the expression of costimulatory molecules and their capacity to induce the proliferation of allogeneic T cells. In the presence of the conditioned supernatant and compared with control bone marrow progenitors, the progenitors failed to proliferate and form clumps of cells (Fig. 5, upper panel). This was also observed when recombinant IL-6 was added to the culture wells. Interestingly, the addition of the anti-IL-6 antibody together with the conditioned supernatant reversed the inhibition of clump formation, confirming the role of IL-6 secreted by MSC in this process. These observations were associated with the significant decrease of the CD11c/MHC II double positive DC and the costimulatory molecules at the surface of the DC when IL-6 was added to the culture (Fig. 5, middle panel). The decrease of these markers was also observed when the conditioned supernatant was added, although it was significant only for the CD11c marker. Interestingly, the proliferation of T lymphocytes induced by addition of allogeneic DC (ratio DC:T lymphocytes of 1:100 or 1:300) was significantly inhibited when DC were generated in presence of IL-6 or conditioned supernatant (Fig. 5, lower panel). No inhibition of T-cell proliferation was observed by addition of the anti-IL-6 antibody to the culture of MSC supernatant-modulated DC. In order to determine whether primary MSC may display the same effect on DC generation as C3 MSC, we performed another similar experiment and added NIH-3T3 fibroblasts as control, since these cells are not immunosuppressive. The levels of IL-6 secreted by the primary MSC were shown to be 35.8 ⫾ 5.3 pg/ml in basal conditions and increased by a sevenfold factor in MLR stimulated conditions. As already reported [11], we checked that the conditioned supernatant from activated primary mMSC reduced the proliferation of splenocytes in the MLR (data not shown). In addition, the conditioned supernatant from activated primary mMSC was able to decrease the percentage of CD11c⫹/MHC II⫹ DC with little impact on the percentage of costimulatory molecules

(Fig. 6B). Surprisingly, control NIH-3T3 display the same effect on DC. Altogether, these results suggest that IL-6 secreted by cells from mesenchymal origin may reduce the number of mDC generated from bone marrow mononuclear cells.

DISCUSSION A number of data have recently emerged on the comprehension of the immunoregulatory mechanisms induced by MSC. Among the various hypotheses, variation of cytokine profiles inducing a tolerant phenotype [14], IDO-mediated T-cell inhibition [12], T-cell division arrest anergy [13], inhibition of monocyte differentiation to DC, which will reduce antigen presentation [8, 15], and induction of regulatory T cells [11] have been suggested. Convergent data now suggest that MSC-induced immunosuppression acts through the secretion of a soluble factor. In this report, we addressed the soluble factor(s) secreted by MSC, which could be at the initiation step of the regulation of the immune response, and investigated its role in the suppressive mechanism. By gene expression analysis, we were able to discriminate the MSC samples and to separate the control samples from those that have been cocultured with allogeneic splenocytes suggesting a specific activation profile. Among the RNAs that allowed the discrimination, we found that VEGF and IL-6 were highly upregulated in MSC after stimulation. This is in agreement with recent data reporting the upregulation of VEGF and IL-6 in the supernatants of cocultures of human primary MSC and peripheral blood mononuclear cells [14]. VEGF has been postulated to contribute to immune suppression by blocking the differentiation and/or the emigration of the lymphoid progenitors residing in the bone marrow [22]. We show here that, among the various


Figure 5. Phenotypic analysis of bone marrow-derived dendritic cells. Progenitor cells were cultured with recombinant granulocyte macrophage– colony-stimulating factor and recombinant IL-4 ⫾ IL-6 for 6 days and matured for 24 hours. Results are expressed as the percentage of CD11c⫹/MHC II⫹ cells (A) or CD11c⫹cells expressing MHC II⫹ molecules (B) or double positive cells for costimulatory molecules and MHCII⫹ cultured with IL-6 from d0 or d6 (C) (p value). Abbreviations: d0, day 0; d6, day 6; DC, dendritic cells; IL, interleukin; MHC, major histocompatibility complex; ND, not determined; NS, not significant.


2031

cell components of the MLR, VEGF is produced uniquely by the MSC, and its secretion is increased in the MLR. As both VEGF and IL-6 levels are upregulated by PGE2 [19], which is also enhanced in the MLR (our data and [14]), we used the PGE2 inhibitor, indomethacin, to discriminate between the respective role of VEGF and IL-6 in immunoregulation. Indeed, the levels of VEGF were not affected by the PGE2 inhibitor, suggesting that VEGF is unlikely involved in the suppressive mechanism. On the contrary, in the presence of PGE2 inhibitor, a partial reversion of the immunosuppressive effect of MSC is associated with a decrease of IL-6 levels. These data are in agreement with the recent study of Aggarwal et al. suggesting that PGE2 plays a key role in the immunoregulation mediated by MSC [14]. Indeed, we propose that IL-6 is an important cytokine playing a role in the immune regulatory mechanism mediated by MSC. First, we show that MSC, among all cell types present in the MLR, were responsible for the increased secretion of IL-6. Second, IL-6 levels were proportional to the number of MSC added in the MLR and inversely correlated with the production of IFN-␥, IL-2, and TNF-␣. Third, the production of IL-6 associated to the immune suppression was affected by the PGE2 inhibitor indomethacin, suggesting that PGE2 may act through the increase of IL-6. Finally, the addition of a neutralizing anti-IL-6 antibody partly restored the proliferation of T cells in the MLR. However, because the reversion of immunosuppression was only partial, IL-6 does not account for the overall immunoregulatory effect mediated by the MSC. The expression and the functional activity of the IDO protein by MSC have been proposed to take part in their immunosuppressive effect [12]. Upon IFN-␥ activation, the enzyme depletes the tryptophan from the cellular environment, resulting www.StemCells.com

in the inhibition of T-cell proliferation due to the lack of this essential amino acid. Although we detected IDO activity in the supernatants of primary human MSC after IFN-␥ induction (data not shown), we were unable to demonstrate the IDO activity among the various MSC lines from murine or human origin used in this study. One possible explanation could be the loss of the enzymatic function during the immortalization process of these cell lines. Indeed, the expression of IDO is confined to a limited range of cell types, and the transcription of the gene has to be stringently controlled to respond to specific inflammatory mediators because IDO would probably be toxic if constitutively active. Our results are in agreement with the study of Tse et al., who excluded the tryptophan depletion in the culture medium of hPBMC as a possible mechanism for T-cell suppression mediated by MSC [9]. Consistently, the consequences of tryptophan deprivation are the induction of T-cellcycle arrest or the production of toxic metabolites from tryptophan, such as quinolinic acid and 3-hydroxy-anthranillic acid, both of which have been shown to induce apoptosis [23]. However, the role of MSC as possible inducers of T-cell apoptosis has been excluded [9, 10, 14]. Since the C3 cell line used in this study displays an immunosuppressive function but is devoid of IDO activity, we can exclude a major role for IDO in a general suppressive mechanism, although it may participate in the process, at least in human MSC. The suppressive effect of MSC has also been postulated to act through the modulation of DC differentiation and function either by inhibiting the differentiation of monocytes into DC or by downregulating the costimulatory molecules on mature DC [15]. Consistent with this study, we observed a slight but significant decrease in the percentage of mDC expressing the costimulatory molecules when cultured in the presence of MSC.


Figure 6. Effect of MSC on the differentiation of dendritic cells. Bone marrow progenitors (BMP) were cultured with IL-6 or conditioned supernatants from MSC/splenocyte cocultures ⫾ anti-IL-6 antibody. (A): Upper panel: Microscopic analysis of BMP on day 6. Middle panel: Immunophenotype of DC. Lower panel: Proliferation of T lymphocytes stimulated by DC (ⴱ, p ⬍ .05; ⫹, p ⬍ .1). (B): One representative experiment of BMP cultured with conditioned supernatants from NIH-3T3 and mMSC. Abbreviations: C3, C3H10T1/2; DC, dendritic cells; IL, interleukin; mDC, mature dendritic cells; MHC, major histocompatibility complex; mMSC, murine MSC; NIH, NIH-3T3.


2032

REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–147. Jorgensen C, Noel D, Gross G. Could inflammatory arthritis be triggered by progenitor cells in the joints? Ann Rheum Dis 2002;61:6 –9. Tsuchida H, Hashimoto J, Crawford E et al. Engineered allogeneic mesenchymal stem cells repair femoral segmental defect in rats. J Orthop Res 2003;21:44 –53. Devine MJ, Mierisch CM, Jang E et al. Transplanted bone marrow cells localize to fracture callus in a mouse model. J Orthop Res 2002;20:1232–1239. Horwitz EM, Prockop DJ, Fitzpatrick LA et al. Transplantability and therapeutic effects of bone marrow-derived mesenchymal cells in children with osteogenesis imperfecta. Nat Med 1999;5:309 –313. Le Blanc K, Rasmusson I, Sundberg B et al. Treatment of severe acute graft-versus-host disease with third party haploidentical mesenchymal stem cells. Lancet 2004;363:1439 –1441. Barry FP, Murphy JM. Mesenchymal stem cells: Clinical applications and biological characterization. Int J Biochem Cell Biol 2004;36: 568 –584. Beyth S, Borovsky Z, Mevorach D et al. Human mesenchymal stem cells alter antigen-presenting cell maturation and induce T-cell unresponsiveness. Blood 2005;105:2214 –2219. Tse WT, Pendleton JD, Beyer WM et al. Suppression of allogeneic T-cell proliferation by human marrow stromal cells: Implications in transplantation. Transplantation 2003;75:389 –397. Di Nicola M, Carlo-Stella C, Magni M et al. Human bone marrow stromal cells suppress T-lymphocyte proliferation induced by cellular or nonspecific mitogenic stimuli. Blood 2002;99:3838 –3843. Djouad F, Plence P, Bony C et al. Immunosuppressive effect of mesenchymal stem cells favors tumor growth in allogeneic animals. Blood 2003;102:3837–3844. Meisel R, Zibert A, Laryea M et al. Human bone marrow stromal cells inhibit allogeneic T-cell responses by indoleamine 2,3-dioxygenasemediated tryptophan degradation. Blood 2004;103:4619 – 4621. Glennie S, Soeiro I, Dyson PJ et al. Bone marrow mesenchymal stem cells induce division arrest anergy of activated T cells. Blood 2005;105: 2821–2827. Aggarwal S, Pittenger MF. Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood 2005;105:1815–1822.

IL-6, which will be concentrated at the vicinity of DC and therefore rapidly active. Another possible explanation may be the induction of other suppressive mediators by IL-6, further suggested by a partial reversion of lymphocyte proliferation using specific antibodies. Although we can not exclude that the generation of immature DC might induce the generation of CD8⫹ T suppressive cells as demonstrated with tolerogenic DC [20, 26], a lack of T-cell activation by less functional immature DC may be the underlying mechanism. This mechanism may occur in vivo in different tissues as suggested by very recent data reporting that a “pre-cDC” precursor population including cells “precommitted” to form conventional DC is present in the spleen [27]. Altogether, these data suggest that, through IL-6 secretion, MSC possibly act by decreasing the percentage of presentation and costimulatory molecules at the surface of DC. As a result, the activation and proliferation of T lymphocytes is reduced in the MLR. This is also supported by the fact that NIH-3T3 fibroblasts were able to decrease the percentage of mDC, although they have no suppressive property but may secrete IL-6 upon activation [28]. Indeed, the effect of IL-6 on DC maturation is certainly one mechanism by which MSC display suppressive activity but probably is not the main mechanism.

DISCLOSURE

OF

POTENTIAL CONFLICTS

OF INTEREST The authors indicate no potential conflicts of interest.

15 Jiang XX, Zhang Y, Liu B et al. Human mesenchymal stem cells inhibit differentiation and function of monocyte-derived dendritic cells. Blood 2005;105:4120 – 4126. 16 Eisen MB, Spellman PT, Brown PO et al. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 1998;95: 14863–14868. 17 Bataille R, Barlogie B, Lu ZY et al. Biologic effects of anti-interleukin-6 murine monoclonal antibody in advanced multiple myeloma. Blood 1995;86:685– 691. 18 Daubener W, Hucke C, Seidel K et al. Interleukin-1 inhibits gamma interferon-induced bacteriostasis in human uroepithelial cells. Infect Immun 1999;67:5615–5620. 19 Inoue H, Takamori M, Shimoyama Y et al. Regulation by PGE2 of the production of interleukin-6, macrophage colony stimulating factor, and vascular endothelial growth factor in human synovial fibroblasts. Br J Pharmacol 2002;136:287–295. 20 Filaci G, Fravega M, Fenoglio D et al. Non-antigen specific CD8⫹ T suppressor lymphocytes. Clin Exp Med 2004;4:86 –92. 21 Hegde S, Pahne J, Smola-Hess S. Novel immunosuppressive properties of interleukin-6 in dendritic cells: Inhibition of NF-kappaB binding activity and CCR7 expression. FASEB J 2004;18:1439 –1441. 22 Ohm JE, Gabrilovich DI, Sempowski GD et al. VEGF inhibits T-cell development and may contribute to tumor-induced immune suppression. Blood 2003;101:4878 – 4886. 23 Mellor AL, Munn DH. IDO expression by dendritic cells: Tolerance and tryptophan catabolism. Nat Rev Immunol 2004;4:762–774. 24 Mitani H, Katayama N, Araki H et al. Activity of interleukin 6 in the differentiation of monocytes to macrophages and dendritic cells. Br J Haematol 2000;109:288 –295. 25 Chomarat P, Banchereau J, Davoust J et al. IL-6 switches the differentiation of monocytes from dendritic cells to macrophages. Nat Immunol 2000;1:510 –514. 26 Suciu-Foca N, Manavalan JS, Scotto L et al. Molecular characterization of allospecific T suppressor and tolerogenic dendritic cells: Review. Int Immunopharmacol 2005;5:7–11. 27 Naik SH, Metcalf D, van Nieuwenhuijze A et al. Intrasplenic steady-state dendritic cell precursors that are distinct from monocytes. Nat Immunol 2006;7:663– 671. 28 Fima E, Shahaf G, Hershko T et al. Expression of PKCeta in NIH-3T3 cells promotes production of the pro-inflammatory cytokine interleukin-6. Eur Cytokine Netw 1999;10:491–500.


The downregulation of the costimulatory signals was associated with the decrease of T-cell activation and subsequent proliferation, suggesting that, in our conditions, MSC are likely to reverse the maturation of DC. Furthermore, we show that bone marrow progenitors cultured in the presence of conditioned supernatants from MSC were partly inhibited to differentiate to functional mDC, and this effect was associated, at least in part, with the secretion of IL-6 by the MSC. This cytokine was proven to skew the differentiation of monocytes into DC being functionally impaired [21]. This last study also argues for the immunosuppressive activity of IL-6 via the decreased production of proinflammatory mediators, the inhibition of CCR7 expression on DC, and the inhibition of T-cell proliferation. It was also shown to inhibit the differentiation of monocytes to DC by promoting their differentiation toward macrophages [24]. The demonstration that IL-6 is an essential factor in the molecular control of antigen-presenting cell development was presented in the study of Chomarat et al. [25]. They report that human fibroblasts, on contact with monocytes, release IL-6, which upregulates the expression of macrophage– colony-stimulating factor (M-CSF) receptors at the surface of monocytes. In turn, monocytes consume autocrine M-CSF, leading to the emergence of macrophages at the expense of DC. In our study, we show that MSC, after activation in coculture with immune cells, secrete high levels of IL-6, which interfere with the generation of DC, leading to less mature DC. In our study, the concentration of IL-6 produced by MSC is low as compared with that of rIL-6. The reason for this discrepancy is not fully understood, but we may speculate that, in close contact with the DC in the MLR, MSC will secrete low but continuous levels of

Mesenchymal Stem Cells Inhibit the Differentiation of Dendritic Cells Through an Interleukin-6-Dependent Mechanism Farida Djouad, Louis-Marie Charbonnier, Carine Bouffi, Pascale Louis-Plence, Claire Bony, Florence Apparailly, Céline Cantos, Christian Jorgensen and Danièle Noël Stem Cells 2007;25;2025-2032; originally published online May 17, 2007; DOI: 10.1634/stemcells.2006-0548 This information is current as of August 8, 2007 Updated Information & Services

including high-resolution figures, can be found at: http://www.StemCells.com/cgi/content/full/25/8/2025 Downloaded from www.StemCells.com at Biblioteca Di Medicina Veterinaria on August 8, 2007