Author contributions: V.B. and E.M.: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing, and final approval of manuscript; R.T., V.D.-N., and X.L.: collection and assembly of data; P.T., I.A., and B.V.: data analysis and interpretation; I.N. and P.N.: conception and design, data analysis and interpretation, manuscript writing, and final approval of manuscript. V.B., E.M., I.N., and P.N. contributed equally to this article.
Disclosure of potential conflicts of interest is found at the end of this article.
First published online in STEM CELLSEXPRESS August 7, 2012.
Besides their therapeutic benefit as cell source, neural stem/progenitor cells (NSPCs) exhibit immunosuppressive properties of great interest for modulating immune response in the central nervous system. To decipher the mechanisms of NSPC-mediated immunosuppression, activated T cells were exposed to NSPCs isolated from fetal rat brains. Analyses revealed that NSPCs inhibited T-cell proliferation and interferon-gamma production in a dose-dependent manner. A higher proportion of helper T cells (CD4+ T cells) was found in the presence of NSPCs, but analyses of FoxP3 population indicated that T-cell suppression was not secondary to an induction of suppressive regulatory T cells (FoxP3+ CD4+ CD25+). Conversely, induction of the high affinity interleukin-2 (IL-2) receptor (CD25) and the inability of IL-2 to rescue T-cell proliferation suggest that NSPCs display immunosuppressive activity without affecting T-cell activation. Cultures in Transwell chambers or addition of NSPC-conditioned medium to activated T cells indicated that part of the suppressive activity was not contact dependent. We therefore searched for soluble factors that mediate NSPC immunosuppression. We found that NSPCs express several immunosuppressive molecules, but the ability of these cells to inhibit T-cell proliferation was only counteracted by heme oxygenase (HO) inhibitors in association or not with nitric oxide synthase inhibitors. Taken together, our findings highlight a dynamic crosstalk between NSPCs and T lymphocytes and provide the first evidence of an implication of HO-1 in mediating the immunosuppressive effects of the NSPCs. STEM Cells2012;30:2342–2353
Neural stem/progenitor cells (NSPCs) present therapeutic potential because of their multipotency and their ability to be expanded in vitro, but recent evidence suggests that these cells display immune properties that would also be of great interest for restorative therapies. Indeed, mouse NSPCs implanted beneath the kidney capsule formed stable allografts that continued to thrive for at least 4 weeks, whereas neonatal mouse cerebellar grafts had already been destroyed at that time . Furthermore, porcine NSPCs transplanted into the striatum of immunocompetent rats showed long-term survival as compared to porcine neuroblasts [2, 3]. This long-term survival has been initially attributed to the low expression of molecules implicated in cell-mediated immune recognition such as the major histocompatibility complex or the costimulatory proteins CD40, CD80, and CD86 [1, 2, 4]. However, findings that multipotent cells such as mesenchymal stem cells (MSCs) are highly immunosuppressive [5–7] raise the possibility that NSPCs actively inhibit the host immune response. Indeed, like MSCs [5–8], NSPCs suppress T-cell proliferation in vitro  and attenuate experimental autoimmune encephalomyelitis (EAE) following intravenous injection [9–11]. MSC-mediated immunosuppression has been variously demonstrated to involve IL-10 [6, 12], TGFβ [6, 13], PGE2 , indoleamine 2,3-dioxygenase (IDO) , inducible nitric oxide synthase (iNOS) [16, 17], and the heme oxygenases (HOs) . In this article, we sought to clarify the mechanisms involved in the immunosuppressive effect of rat NSPCs (rNSPC). We demonstrate that the NSPCs modulate T lymphocytes so as to favor CD4+ helper T cells to the detriment of CD8+ cytotoxic T cells (also known as CTL or killer T cells). In vitro studies show that T-cell suppression by NSPCs is neither secondary to an induction of suppressive regulatory T cells (FoxP3+ CD4+ CD25+) nor secondary to a defect in T-cell activation (CD25+). Finally, we find that the HOs, alone or in combination with nitric oxide production, constitute a critical pathway used by NSPCs to actively modulate the immune response.
MATERIALS AND METHODS
All antibodies used in this study are presented in Table 1.
Table 1. Antibodies used in the study
Abbreviations: DSHB, Developmental Studies Hybridoma Bank; ECCC, European Collection of Cell Culture; GFAP, glial fibrillary acidic protein; iNOS, inducible nitric oxide synthase; nNOS, neuronal form of NOS; RIP.
Preparation of NSPCs
Primary cultures of NSPCs were established from the whole brain of E15 Sprague-Dawley rat embryos as previously described [18, 19]. Briefly, tissues freed of meninges were incubated with 0.5 mg/ml trypsin for 15 minutes at 37°C. Following addition of 10% fetal calf serum (FCS, Sigma-Aldrich, Saint Louis, MO, http://www.sigmaaldrich.com), tissues were exposed to 0.1 mg/ml of DNase I prior to mechanical trituration. Aggregates were removed by decantation and cells were further purified from small debris by centrifugation. Cells were resuspended and plated for one night in medium composed of Dulbecco's modified Eagle's medium/Ham's F-12 (1/1, vol/vol), 33 mM glucose, 5 mM HEPES (pH 7.2), 100 U/ml penicillin, 0.1 mg/ml streptomycin, and 2 mM L-glutamine and supplemented with 10% FCS (serum-supplemented medium). The next day, the floating cells were recovered, washed, and resuspended in 10 ml of serum-free medium supplemented with N2 (Invitrogen, Life Technologies SAS, Saint Aubin, F, http://fr-fr.invitrogen.com). The cells were then plated in uncoated dishes and expanded as neurospheres for 10 days in the presence of 25 ng/ml basic fibroblast growth factor (b-FGF) with a complete change of the N2 medium after 5 days of culture and addition of b-FGF every 3 days. The neurospheres were then composed of 8.6% of glial fibrillary acidic protein (GFAP+) cells and 4% of RIP+ cells. No Tuj-1+ cell was detected. For adherent condition (Fig. 1F), the cells were plated at a concentration of 2 × 105 cells per square centimeter onto poly(L-ornithine) (50 μg/ml)/laminin (2 mg/ml)-coated culture dishes.
After 10 days in vitro, the neurospheres were collected, enzymatically dissociated with trypsin-EDTA, and submitted to mechanical trituration to give a single-cell suspension. Cells were plated at a concentration of 2 × 105 cells per square centimeter onto poly(L-ornithine)-coated glass coverslips in serum-supplemented medium for 2 hours. Cells were fixed with 4% paraformaldehyde (PFA) for 15 minutes at room temperature (RT), washed three times with phosphate-buffered saline (PBS), and incubated for 1 hour at RT in PBS/Triton X-100 medium (PBT)-donkey normal serum (DNS) (1× PBS, 4% bovine serum albumin, 0.1% Triton X-100, and 10% DNS). Expression of HO-1 and HO-2 in nestin-expressing NSPCs was detected by performing double immunocytofluorescence as followed. The cells were exposed overnight at 4°C to the monoclonal anti-nestin antibody together with the polyclonal anti-HO-1, anti-HO-2, neuronal form of NOS (nNOS), and iNOS antibodies. After washing, cells were incubated for 2 hours at RT with fluorescein (FITC)-conjugated anti-mouse IgG and Alexa Fluor 568-conjugated anti-rabbit diluted in PBT. Each primary antibody was tested individually with the corresponding secondary antibody. Specificity of the double staining was also controlled by omitting the two primary antibodies or by omitting one secondary antibody. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (1/3,000) before mounting the cells in an antifading medium (Dako France S.A.S., Trappes, F, http://www.dako.com). Analyses were performed using an Axioskop 2 plus microscope (Zeiss, Le Pecq, F, http://www.zeiss.fr), and pictures were acquired using a digital camera (AxioCam HRC, Zeiss) driven by AxioVision Release 4.2 software.
T-Cell Proliferation Assays
Part of this protocol has been extensively described in .
Preparation of T cells
T cells were purified as previously described . Briefly, T cells from Sprague-Dawley rat spleens were prepared by depletion of natural killer (NK) cells, B cells, and monocytes using anti-NKR-P1A, anti-CD45R, and anti-CD11b/c monoclonal antibodies (mAbs), respectively, followed by anti-mouse IgG-coated Dynabeads (Invitrogen). For polyclonal stimulation, T cells were incubated with 5 μg/ml anti-CD28 antibody in flat-bottomed 96-well plates previously coated with anti-CD3 (5 μg/ml; 2 hours at 37°C) except for adherent conditions (Fig. 1F) for which the anti-CD3 was directly added to the medium.
Polyclonally stimulated purified T cells (105 cells per well) together with irradiated (30 Gy) or nonirradiated rNSPCs (2.5 × 104; 5 × 104; 1 × 05; 2 × 105 cells per well) were seeded in triplicate in 96-well flat-bottomed culture plates and cocultured for 3 days in (1:1) N2 medium/Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin, 5% heat-inactivated FCS, 1% nonessential amino acids, 5 mM HEPES, 1 mM sodium pyruvate, and 1 μM 2-ME. After 3 days of cocultures, 9.12% (10.3% in case of cell irradiation) of the cells was GFAP+ and 2.75% (3.26% in case of cell irradiation) was RIP+. No Tuj-1-expressing cell was found. In one set of experiments, NSPCs were replaced by irradiated T cells to control cell specificity.
Polyclonally stimulated purified T cells (4 × 105) were placed in the lower chamber of a 24-well cell culture insert companion plate (Falcon, BD Biosciences, San Diego, CA, http://www.bdbiosciences.com) and cultured for 3 days with irradiated (30 Gy) or nonirradiated NSPCs (105; 2 × 105; 4 × 105; 8 × 105 cells per well) seeded in the upper chamber.
Effect of NSPC-Conditioned Medium
NSPCs were cultured in N2 medium as described above. The supernatants were collected everyday from day 0 to day 5. Fifty microliters of supernatant sample was added to 50 μl of polyclonally stimulated purified T cells (105 cells per well). Stimulated T cells were plated and cultured for 3 days in 96-well flat-bottomed culture plate.
Tritiated Thymidine Incorporation Assays
Proliferation was determined by addition of [3H]thymidine (0.625 μCi per well, PerkinElmer, Boston, MA, http://www.perkinelmer.com) for the last 12 hours of culture. Cells were harvested on fiberglass filters using a harvester (Tomtec, Hamden, CT, http://www.tomtec.com) and radioactivity was measured by standard scintillation technique.
Purified T cells were labeled with carboxyfluorescein succinimidyl ester (CFSE) 5 μM (3 minutes at RT). The cells were washed three times with serum-supplemented medium and cultured in direct contact (see NSPC/T-cell cocultures) or in Transwell (see Transwell assays) with NSPCs. Proliferation of CFSE-labeled T cells was investigated by flow cytometry. A Canto cytometer (BD Biosciences, San Diego, CA, http://www.bdbiosciences.com) was used to measure fluorescence and data were analyzed using with FlowJo software (Tree Star, Inc., Ashland, OR, http://www.treestar.com/).
Lymphocyte Inhibition Assay
The mechanisms underlying the immunosuppressive effects of NSPCs were characterized by treating the cocultures (one T cell for two irradiated or nonirradiated NSPCs) for 3 days with recombinant human interleukin-2 (IL-2, 100 U/ml), human neuropeptide Y (NPY) or human peptide YY (PYY) (10−8 M, Sigma-Aldrich), anti-TGFβ blocking mAb 2G7 (50 μg/ml, ECCC, European Collection of Cell Culture, Salisbury, UK, http://www.hpacultures.org.uk), 1-methyl-D,L-tryptophan (1-MT) (200 μM, Sigma-Aldrich), indomethacin (10 μM, Sigma-Aldrich), ketoprofen (50 μM, Sigma-Aldrich), or L-NG-monomethyl-L-arginine (L-NMMA, 1.25 mM, Sigma-Aldrich). The implication of HO was determined by pretreating the NSPCs with 50 μM of tin protoporphyrin (SnPP; Frontier Scientific, Logan, UT, http://www.frontiersci.com) for 2 hours before the setting of cocultures. [3H]thymidine incorporation was assessed at day 3.
Cytokine Production Assay
The release of IFN-γ in the supernatants was assessed by ELISA using a kit from R&D Systems (catalog number: DY585; R&D Systems Europe, Lille, F, http://www.rndsystems.com) according to the manufacturer's instructions. Culture conditions are described in the subsection NSPC/T-cell cocultures.
RNA Extraction and Retrotranscription
Stimulated T cells, NSPCs cultured for 10 days as neurospheres, or stimulated T cells cocultured with NSPCs for 3 days were disrupted in Trizol reagent (Invitrogen) and homogenized using syringe and needle according to the manufacturer's specifications. Potential genomic DNA contamination was removed by treatment with Turbo DNase (Ambion Inc., Austin, TX). RNA was quantified using ND-1000 UV-Vis Spectrophotometer (Nanodrop Technologies, Wilmington, DE, http://www.nanodrop.com) and RNA integrity was determined on agarose gel. cDNA was synthesized from 5 μg of total RNA using the Moloney Murine Leukemia Virus reverse-transcriptase kit (Invitrogen) as previously described  and diluted to a final concentration of 100 ng cDNA/μl.
Analyses of the transcripts were performed with a GenAmp 7700 sequence detection system (AB, Life Technologies SAS, Saint Aubin, F, http://fr-fr.invitrogen.com) using SYBR Green PCR core reagents (AB). Oligonucleotide sequences are listed in Table 2. The PCR method and the 2 quantification method, after normalization to HPRT values, have been previously described . The mRNA expression level is defined as the fold change in mRNA levels in a given sample relative to levels in a calibrator—in this case, RNA from splenocytes and brain (50:50). The mRNA expression level is calculated as follows: mRNA expression level = 2, where ΔΔCt = (Ct(Target) − Ct(HPRT))sample − (Ct(Target) − Ct(HPRT))CB. Specific amplifications were checked by amplicon melting curves (Table 2).
Table 2. Oligonucleotide sequences and amplicon sizes
Abbreviations: HO, heme oxygenase; HPRT, hypoxanthine-guanine phosphoribosyltransferase; iNOS, inducible nitric oxide synthase; nNOS, neuronal form of NOS; NPY, neuropeptide Y.
Flow Cytometry Analysis
T-cell phenotype was determined by incubating the cells with OX35 (CD4)-PECy7, Ox39 (CD25)-PE, and R73 (panT)-Alexa Fluor 488 or OX8(CD8)-PE, OX39(CD25)-Alexa Fluor 647, and R73 (pan T)-Alexa Fluor 488 for 20 minutes at 4°C. Cells were then fixed for 15 minutes at 4°C with 2% PFA and washed with PBS/FCS/sodium azide medium (PFN; 1× PBS, 2% FCS, and 0.1% sodium azide). Membrane permeabilization was induced by incubating the cells with PBS/FCS/saponin medium (PFS) (1× PBS, 0.2% FCS, and 0.5% saponin) for 15 minutes at 4°C prior to the incubation with anti-Foxp3-APC for 1 hour. After washing the cells three times with PFN, fluorescent labeling was measured using a FACS LSR II (BD Biosciences) and analyzed with FlowJo software.
Data are expressed as mean values ± SEM of at least three independent experiments performed in duplicate or triplicate. Statistical analyses were performed using the one-way analysis of variance followed by Bonferroni post hoc tests. *, p <.05; **, p <.001; ***, p <.0001.
NSPCs Inhibit T-Cell Proliferation Through Soluble Factors
To decipher the immunosuppressive effects of rNSPCs, their impact on T-cell proliferation was tested. As simultaneous stimulation of the T-cell receptor complexes (TCR/CD3) and the costimulatory receptor CD28 activates T cells, purified rat T cells derived from spleens were stimulated with anti-CD3/anti-CD28 mAbs and cocultured for 3 days with increasing number of irradiated (30 Gy) NSPCs. Absence of NSPC proliferation was checked by [3H]thymidine uptake (Fig. 1A). [3H]thymidine was added for the last 12 hours of culture. Stimulating effect of anti-CD3/anti-CD28 mAbs was confirmed by a strong increase of [3H]thymidine incorporation in treated T cells as compared to untreated T cells (Fig. 1B). Culture of activated T cells with NSPCs inhibited significantly [3H]thymidine uptake since 27%, 64%, and 93% decrease was observed at NSPC/T-cell ratios of 1:2, 1:1, and 2:1, respectively. This effect was not due to the coverage of the bottom plate by NSPCs since our culture conditions favor the formation of neurosphere-like clusters that are mostly floating (Fig. 1C). Similar dose-dependent inhibition was obtained when Concanavalin-A, a non-TCR-mediated mitogenic stimulus, was used (data not shown). As T-cell activation also induces the release of cytokines, we investigated by ELISA the concentration of IFN-γ released in the media after 3 days of culture in presence or absence of NSPCs. As shown in Figure 1D, the production of IFN-γ was strongly impaired by the presence of NSPCs and the inhibition was closely dependent on the NSPC/T-cell ratio. The immunosuppressive effect of NSPCs on T cells was specific since irradiated T lymphocytes did not inhibit the proliferation of activated T cells (Fig. 1E). To evaluate the impact of culture conditions on NSPC properties, cells obtained from the same preparation were grown as neurospheres or as adherent cells and placed with activated T lymphocytes. Both adherent and nonadherent NSPCs were capable of inhibiting the proliferation of stimulated T cells. As presented in the Figure 1F, the effect increased with the number of adherent NSPCs and statistical differences were observed from a ratio of one NSPC per 40 stimulated T cells. A dose-dependent effect was also observed for nonadherent NSPCs; however, statistical differences were only observed at the ratio of one NSPC for one stimulated T cells, probably due to the large variations in thymidine uptake observed in this particular condition. Importantly, for a given ratio, we did not found statistical differences between adherent and nonadherent NSPCs. The next step was to determine whether NSPC-mediated immunosuppression required cell-cell contact. Transwell experiments showed that even separated by a porous membrane of 0.4 μm, NSPCs inhibited T-cell proliferation (Fig. 1G). This observation indicates that a part of the inhibitory effect of NSPCs on T-cell proliferation is mediated by soluble factors.
Immunosuppressive Effect of NSPCs is Not Due to the Inhibition of T-Cell Activation or to the Induction of Regulatory T Cells
To decipher the mechanisms responsible for NSPC-mediated immunosuppression, T lymphocytes were characterized by flow cytometry using specific mAbs. T cells were identified with R73, an antibody directed against the TCR, whereas T-cell subsets such as helper and cytotoxic/suppressor T cells were characterized using antibodies directed against the cell surface antigens CD4 and CD8, respectively. T-cell activation was monitored by assessing the expression of the IL-2 receptor alpha subunit (CD25).
Flow cytometry analyses indicate efficient T-cell activation as 100% of the cells was CD25+ after 3 days of treatment with anti-CD3/CD28 mAbs (Fig. 2A). No significant difference in the percentage of CD25-expressing T cells was observed in absence or presence of NSPCs indicating that NSPCs did not inhibit CD25 induction (Fig. 2A). This observation suggests that T-cell activation is not a direct target of NSPC-mediated immunosuppression. The second point was to control potential changes in the proportion of helper (CD4+) and cytotoxic (CD8+) T cells. In absence of NSPCs, only 37% of the T cells was CD4+ while in presence of two NSPCs for one T cell, 70% of the T cells was CD4+ cells (Fig. 2B). This increment paralleled a diminution in the percentage of CD8+ cells (Fig. 2C) and led to an increase in the CD4/CD8 ratio from 0.63 (stimT cells) to 1.63 (two NSPCs for one stimT cell). Variations in the proportion of CD4+ and CD8+ cells among the CD25+ population were similar (Fig. 2E, 2F). This result was not surprising since nearly all T cells treated with anti-CD3/CD28 mAbs expressed this T-cell activation marker even in the presence of NSPCs (Fig. 2A). The higher proportion of CD4+ T cells in presence of NSPCs prompted us to control for a possible involvement of suppressive regulatory T cells (also known as FoxP3+ CD4+ CD25+ Treg) in NSPC-mediated suppressive activity. To this end, expression of the Treg marker FoxP3 was assessed by flow cytometry. Analyses did not reveal significant differences in the percentage of FoxP3+ cells among the CD4+ T-cell subset in the presence or absence of NSPCs (Fig. 2D, 2G). These data suggest that the suppressive effect of NSPCs is not mediated through the induction of regulatory T cells.
The HO Is Implicated in the NSPCs-Mediated Immunosuppression
Since maximal inhibition of T-cell proliferation was obtained at a ratio of two NSPCs for one stimulated T cell, this ratio was used for all subsequent in vitro experiments, unless specified. Before analyzing the implication of various molecules of interest in T-cell suppression, we sought to verify whether the inhibition of T-cell proliferation could be reversed by the administration of IL-2. For this purpose, stimulated T cells cultured in the presence of NSPCs were incubated for 3 days with 100 U/ml of recombinant human IL-2. As indicated by [3H]thymidine incorporation, IL-2 failed to rescue T-cell proliferation (Fig. 3A), suggesting that NSPC-treated T cells were not in a state of anergy. The next step was to determine which molecules expressed by NSPCs might interfere with the proliferation of stimulated T cells. We first checked the impact of NPY, a neuropeptide that regulates innate and adaptive immune response . NPY and the related peptide PYY mediate their function through four different receptors, called Y1, Y2, Y3, and Y5. All of them are cloned except Y3 receptor that preferentially binds NPY over PYY. So, to determine their implication in the suppressive effect of NSPCs, stimulated T cells cultured in absence or presence of NSPCs were treated with NPY (10−8 M) or PYY (10−8 M). As shown in Figures 3B and 3C, NPY neither inhibited nor rescued T-cell proliferation. The same result was obtained with PYY. This lack of effect could be explained by the absence of the main NPY receptors in T cells. Indeed, none of the transcripts coding for Y1, Y2, and Y5 were detected in naïve or polyclonally stimulated rat T cells although NPY mRNA was clearly expressed by these cells in both conditions (data not shown). To assess a potential role of TGFβ-1 in the immunosuppressive effect of NSPCs, cocultures were treated with the blocking mAb 2G7. An increase in [3H]thymidine incorporation was observed following the addition of the mAb 2G7, but the difference between treated and untreated cells was not statistically significant (Fig. 3D). Similarly, treatment of the cells with 1-MT, a specific inhibitor of the IDO, did not reverse the inhibitory effect of NSPCs on T-cell proliferation (Fig. 3E), indicating that full activity of this enzyme is not essential for the immunosuppressive properties of NSPCs. We also tested the potential implication of prostaglandins using nonselective inhibitors of the cyclooxygenases 1 and 2 (COX1 and COX2). However, neither indomethacin (Fig. 3F) nor ketoprofen (data not shown) was able to rescue T-cell proliferation. Finally, we examined the potential involvement of NOS and HO, which have both been implicated in several mechanisms of immunosuppression. For this purpose, the cells were treated with L-NMMA, a nonselective inhibitor of NOS, or with SnPP that irreversibly binds and inactivates HO enzymatic activity . Pretreatment of NSPCs with SnPP significantly altered their immunosuppressive effect as a 3.7-fold increase in T-cell proliferation was observed following treatment with this HO inhibitor (Fig. 3G). Interestingly, this increase was enhanced by exposing the cells to L-NMMA. Indeed, a 6.9-fold increase in T-cell proliferation was observed when the cocultures were treated by both L-NMMA and SnPP (Fig. 3G). To confirm the expression and regulation of NOS and HO in our cultures, total RNA from NSPCs, stimT, or NSPC/stimT cocultures were collected and analyzed by quantitative PCR (Q-PCR). As illustrated in Figure 4A, the inducible (HO-1) and constitutive (HO-2) forms of HO were both expressed in NSPCs and stimulated T cells, but only the inducible form showed a dramatic increase when both cell types were mixed. Indeed, HO-1 mRNA levels were 15.2- and 10-fold higher in cocultures as compared to monocultures of NSPCs or stimulated T cells, respectively. NOS showed also differential expression according to the culture conditions. While the nNOS was clearly expressed by NSPCs and not by stimulated T cells, a 4.7-fold decrease in nNOS mRNA levels was observed when both cell types were mixed. Conversely, the inducible form of NOS (iNOS) that showed a very low expression in NSPCs was strongly upregulated in NSPCs/stimT cocultures (22.2-fold increase). These results suggest a dynamic crosstalk between NSPCs and T lymphocytes that favors the transcription of HO-1 and iNOS. Expression of HO-1 and HO-2, nNOS, and iNOS by NSPCs was confirmed at the protein level by performing double immunocytofluorescence using anti-HO-1, HO-2, nNOS, or iNOS antibodies, together with an antibody raised against nestin, an intermediate filament expressed by NSPCs (Fig. 4B).
Finally, impact of the irradiation on NOS and HO expression was evaluated by Q-PCR. We found no significant difference in the levels of iNOS, nNOS, and HO-2 transcripts between nonirradiated and irradiated NSPCs, but an upregulation of HO1 mRNA (1.26 ± 0.54 vs. 32.02 ± 8.15, 2, p <.05, n = 3) was observed 24 hours after irradiation. This observation prompted us to control the effect of nonirradiated NSPCs on the proliferation of stimulated T cells. In Figure 5A, we show that nonirradiated NSPCs cocultured with stimulated T cells were able to inhibit their proliferation in a dose-dependent manner. We also demonstrate that medium conditioned by nonirradiated cells inhibited efficiently the proliferation of stimulated T cells. The effect was correlated to the length of the conditioning period (Fig. 5B). The last point was to confirm involvement of the HO pathway in mediating the immunosuppressive effect of nonirradiated NSPCs. For this purpose, T cells were labeled with CFSE and cocultured with nonirradiated NSPCs. As shown in Figure 5C, a partial restoration of T-cell proliferation (43% of stimT) was observed after treatment of NSPCs with SnPP. This effect was even stronger in Transwell coculture systems (100% of stimT) (Fig. 5D). This high efficiency might be partially explained by the lower expression of HO-1 in nonirradiated NSPCs as compared to irradiated NSPCs. Regarding NOS, the main point is little (Contact, Fig. 5C) or absence (Transwell, Fig. 5D) of effect of L-NMMA/SnPP cotreatment on nonirradiated NSPCs which contrasts with their complementary effect on irradiated cells. This discrepancy may be due to the strong efficiency of SnPP treatment on nonirradiated NSPCs (Fig. 5C, 5D).
Initially thought to replace lost neuronal population in case of neurodegenerative diseases, NSPCs are now considered for their immunosuppressive properties, exhibiting long-term survival in the brain of xenogenic host [2, 3] and improving the course of autoimmune diseases [9-11, 25]. In an attempt to decipher the mechanisms underlying their immunomodulatory effects, the impact of rNSPCs was evaluated on anti-CD3/CD28-stimulated rat T lymphocytes. In this study, we show that rNSPCs inhibit T-cell immune response in a dose-dependent manner, as reflected by impairment of IFN-γ secretion and T-cell proliferation. The immunosuppressive activity is specific and not due to cell irradiation since irradiated T cells did not produce such an effect and nonirradiated NSPCs also inhibited T-cell proliferation. Furthermore, the efficiency of NSPCs to inhibit the proliferation of concanavalin-A-stimulated lymphocytes indicated that NSPCs could also suppress T-cell response to TCR-independent polyclonal stimuli. We also asked whether inhibition of T-cell proliferation by NSPCs might result from their interference on T-cell activation. However, no inhibition of the T-cell activation marker CD25 was detected and administration of IL-2 did not rescue T-cell proliferation. Such an observation is in agreement with the fact that other multipotent cells such as MSCs display immunosuppressive activity without affecting T-cell activation . Conversely, as MSCs have been shown to exert part of their immunoregulatory effects through expansion of regulatory T cells [27–29] and as increased percentage of CD4+/CD25+ T cells was observed in presence of NSPCs, we searched for a potential increase in the number of CD4+/CD25+/FoxP3+ T cells in NSPC/T-cell cocultures. Flow cytometry analyses did not reveal any significant increase in the percentage of CD25+/FoxP3+ cells among the CD4-positive T-cell subset, indicating that the suppressive effect of NSPCs was not due to the induction of regulatory T cells.
In vitro analyses rather support a direct impact of NSPCs on T-cell proliferation through the release of soluble factors. Indeed, the suppressing effect has been also observed using Transwell coculture system or after the addition of NSPC-conditioned media to stimulated T cells. In 2003, Klassen et al. reported an expression of TGFβ-1 in rNSPCs grown as a monolayer in laminin-coated flasks . This study confirmed the presence of TGFβ transcripts in rNSPCs grown as floating neurospheres. Since TGFβ is a multifunctional cytokine with immunosuppressive effects on T cells in vitro , blocking anti-TGFβ antibodies were added to the cocultures. However, inefficiency of the treatment to reverse T-cell inhibition suggested a minor role of TGFβ in NSPC-mediated immunosuppression. As NPY suppresses EAE in mice  and modulates various immune responses via Y1 receptors , a potential involvement of this neuropeptide in mediating the suppressive effect of NSPCs was considered. Such an implication was, then again, unlikely since NPY did not affect naïve or polyclonally stimulated rat T cells, probably due to the lack of at least three NPY receptors in these cells. An effect through IDO was another alternative as this enzyme, which displays immunosuppressive activity through the depletion of tryptophan coupled with the production of kynuric metabolites, has been identified as a key mediator of MSC-mediated immunosuppression [15, 34]. Induction of IDO in NSPC/T-cell cocultures came in support of such a hypothesis but treatment with 1-MT, a specific IDO inhibitor, was unable to restore T-cell proliferation. Similarly, indomethacin or ketoprofen did not rescue T-cell proliferative response indicating that PGE2 are most probably not involved in the immunosuppressive activity of rNSPCs in contrast to what has been reported with the NSPC-like cell line C17.2 .
In fact, our data point out a role for HO in mediating the immunosuppressive activity of rNSPCs as inhibition of the activity of these enzymes with SnPP restored T-cell proliferation. A role for HO in mediating the immunosuppressive activity of rNSPCs is in agreement with growing evidence of anti-inflammatory and immunosuppressive impact of HO-1 [36, 37]. Indeed, the long-term survival of cardiac xenograft in concordant models has been associated with the local production of HO-1 by endothelial and smooth muscle cells [38, 39] while HO-1 overexpression in allogenic models of transplantation was shown to inhibit chronic and acute rejection by reducing leukocyte infiltration [40, 41]. Finding that NSPCs inhibit T-cell proliferation and function through HO-1 activity extends the immunosuppressive role of this enzyme to neural stem cells and raises the possibility that HO-1-expressing NSPCs actively modulate adaptive immune response in case of intracerebral transplantation. The long-term survival of porcine NSPCs in the rat striatum [2, 3] supports such a hypothesis and interestingly, HO-1 was recently reported to mediate the tolerogenic properties of dendritic cells . The mechanism by which HO-1 mediates the immunosuppressive properties of NSPCs remains to be fully determined, but it is believed that heme derivatives such as carbon monoxide (CO) or biliverdin/bilirubin mediate part of HO-1 activity in vitro and in vivo [37, 43-46].
As HO-1 inhibitor only partially reversed the immunosuppressive effect of rNSPCs, we searched for other mechanisms. In particular, we were interested in a potential role of NOS since cocultures of NSPC with stimulated T cells triggered an upregulation of iNOS. A possible involvement of NOS in NSPC-mediated immunosuppression was supported by the fact that NO mediates part of the immunosuppressive activity of macrophages [47, 48], but also by the observation that induction of iNOS in macrophages by T-cell-derived IFN-γ is a basal mechanism that limits the proportion of lymphocytes in inflammatory infiltrates . We did find such an induction as well as a downregulation of nNOS in NSPC/T-cell cocultures. These findings are of great interest since an upregulation of iNOS mRNA by cytokines has been previously associated with a primary inhibition of basal NO production and a downregulation of nNOS [49–51]. As NO displays an inhibitory effect on the activation of NFκB, a key regulator of iNOS gene, the downregulation of nNOS, and the subsequent decrease in the basal level of NO are thought to be critical for the induction of iNOS gene, constituting an emergency system to guarantee a strong increase of NO production. Another interesting point relative to NO induction is the crosstalk between NO and CO pathway . Indeed, NO donors induce the upregulation of HO-1 in cells such as endothelial or smooth muscle cells [53–55]. The possibility that NO contribute to T-cell inhibition through HO-1 is one possibility, and interestingly, we observed more proliferating T cells when NSPCs were treated with both HO and NO inhibitors and placed in direct contact with activated T cells. This observation suggests an implication of NOS pathway in the immunosuppressive effect of NSPCs, even though its specific contribution remains to be clearly defined. Indeed, sole treatment with NO inhibitor had no significant effect, and addition of L-NMMA to SnPP-treated NSPCs did not provoke further increase in the number of proliferating T cells in Transwell cocultures. This absence of effect is probably due to the strong efficiency of HO inhibitor in Transwell cocultures as NSPC-mediated immunosuppression was completely reverted by SnPP pretreatment in such conditions. Such a strong effect emphasizes the critical role of HO and indicates that heme degradation byproducts (i.e., CO and bilirubin/biliverdin) are key players on the immunosuppressive activity induced by soluble factors. Conversely, the partial effect of SnPP treatment in contact conditions suggests the involvement of other molecules such as membrane bound proteins, in the NSPC-mediated immunosuppression.
While many believe that stem cells will find the most use to replace lost cells and escape host immune response because of their low immunogenicity, our present data provide support for an active role of NSPCs in modulating the immune response . These observations go along with growing evidence showing that other stem cells such as MSCs and embryonic stem cells exert immunosuppressive effects and favor the induction of tolerance [26, 57-59]. Here, we show that immunosuppression triggered by NSPCs occurs for a minimum of one NSPC for two activated T cells, which ratio is quite high as compared to MSCs. The difference may be due to the fact that NSPCs constitute a much more heterogeneous population than MSCs. Indeed, proliferation of fetal brain cells with b-FGF generates a mixed population of neural stem cells and progenitors, and only part of them may exhibit immunosuppressive properties. The difference between NSPCs and MSCs may also be due to the culture conditions. Indeed, MSCs are plated as a monolayer, whereas NSPCs are cultured as neurospheres which may limit the release of immunosuppressive factors. This hypothesis is supported by the fact that a significant immunosuppressive effect is observed for a ratio of one NSPC for 40 activated T cells when NSPCs are plated as a monolayer. The quantity of rNSPCs necessary to induce an immunosuppression in the brain remains to be determined but interestingly, long-term survival of xenogenic NSPCs was observed for an amount of 400,000 porcine NSPCs transplanted in the brain of nonimmunosuppressed rats [2, 3]. This observation is of great interest as porcine and rNSPCs show similar immunosuppressive effect in vitro . An immunosuppressive effect of NSPCs in vivo is also supported by the therapeutic benefit of NSPCs transplanted into a mouse model of chronic EAE [9-11, 25]. It is believed that the beneficial effect of NSPCs is favored by the restricted traffic of T cells in the CNS, the possibility of a local accumulation of NSPC-derived factors, and the ability of NSPCs to migrate to inflammatory areas. Relevance of these observations in human remains to be demonstrated but local immunosuppression with NSPCs could be of great interest in regenerative medicine since immune response has been observed following allotransplantation of fetal neural transplants in patients with Huntington's disease .
Whether immunoregulatory activity is a common property of stem cells and whether these cells act through similar mechanisms remain to be clarified, but in this study, we show that HO is an important mediator of NSPC immunosuppressive activity. Although other pathways are probably involved, our findings provide the first evidence that expression of HO is a major mechanism for NSPCs to evade the host immune response and exert their immunosuppressive activity.
We are very grateful to Dr. P. Brachet and Pr. J.-P. Soulillou for their support. We express special thanks to “Etablissement Français du Sang” (EFS, Nantes) that kindly irradiated the NSPCs. Also, we thank Dr. Gareth R. John for critical reading of the manuscript. The nestin antibody obtained from the Developmental Studies Hybridoma Bank was developed by Susan Hockfield under the auspices of the NICHD and maintained by the University of Iowa, Department of Biological Sciences, Iowa City, IA. The work was supported by the “Association Française contre les Myopathies”, the “Fédération des Groupements de Parkinsoniens”, and Progreffe. V. Bonnamain was supported by a fellowship from the french « Ministère de l'Enseignement Supérieur et de la Recherche » and Progreffe Foundation. E. Mathieux and X. Lévêque were supported by CECAP, Progreffe, and Centaure Foundations.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.