Human adipose-tissue derived mesenchymal stem cells induce functional de-novo regulatory T cells with methylated FOXP3 gene DNA

Abstract

Due to their immunomodulatory properties, mesenchymal stem cells (MSC) are interesting candidates for cellular therapy for autoimmune disorders, graft-versus-host disease and allograft rejection. MSC inhibit the proliferation of effector T cells and induce T cells with a regulatory phenotype. So far it is unknown whether human MSC-induced CD4⁺CD25⁺CD127^–forkhead box P3 (FoxP3)⁺ T cells are functional and whether they originate from effector T cells or represent expanded natural regulatory T cells (nT_reg). Perirenal adipose-tissue derived MSC (ASC) obtained from kidney donors induced a 2·1-fold increase in the percentage of CD25⁺CD127^–FoxP3⁺ cells within the CD4⁺ T cell population from allostimulated CD25^–/dim cells. Interleukin (IL)-2 receptor blocking prevented this induction. The ASC-induced T cells (iT_reg) inhibited effector cell proliferation as effectively as nT_reg. The vast majority of cells within the iTreg fraction had a methylated FOXP3 gene Treg-specific demethylated region (TSDR) indicating that they were not of nT_reg origin. In conclusion, ASC induce T_reg from effector T cells. These iT_reg have immunosuppressive capacities comparable to those of nT_reg. Their induction is IL-2 pathway-dependent. The dual effect of MSC of inhibiting immune cell proliferation while generating de-novo immunosuppressive cells emphasizes their potential as cellular immunotherapeutic agent.

Introduction

Mesenchymal stem cells (MSC) can be isolated from an abundance of human tissue sites, including adipose tissue and bone marrow, and their expansion is accomplished easily 1–4. MSC possess immunosuppressive capacities and, as a consequence, over the past decennium MSC have been studied extensively as a prospective cellular therapeutic agent to prevent or treat autoimmune diseases, graft-versus-host disease (GVHD) and allograft rejection 5–12. Upon activation, MSC prevent the proliferation of various immune cells, in particular T cell proliferation 13–19. MSC mediate their suppressive effect through cell–cell contact and the secretion of various soluble factors such as transforming growth factor (TGF)-β, hepatocyte growth factor (HGF), interleukin (IL)-10, nitric oxide (NO), human leucocyte antigen G5 (HLA-G5), indoleamine 2,3-dioxygenase (IDO) and prostaglandins 20–28. While MSC strongly inhibit T cell proliferation via these mechanisms, they preserve the function of CD4⁺CD25⁺CD127^–forkhead box P3 (FoxP3)⁺ regulatory T cells (T_reg) 29. Beyond this, in-vitro studies and studies in animal models have indicated that MSC have the capacity to generate T_reg 9,30–34. Recent evidence was provided that this phenomenon might also occur in renal transplant patients undergoing MSC therapy 35. Intravenous administration of autologous MSC post-transplant led to a proportional increase of CD4⁺CD25⁺CD127^–FoxP3⁺ T cells. Despite their regulatory phenotype, it remains essential to investigate the characteristics and function of these cells. Further, due to the heterogeneity of the T_reg population these CD4⁺CD25⁺CD127^–FoxP3⁺ T cells could represent expanded natural T_reg (nT_reg) or newly induced T_reg (iT_reg).

nT_reg and iT_reg are distinct from each other with regard to their place of origin, the stability of their transcription factor FOXP3 expression and in their methylation pattern of the T_reg-specific demethylated region (TSDR) in the FOXP3 gene 36,37. nT_reg develop intrathymically, express FOXP3 constitutively and have a fully demethylated FOXP3 TSDR. In contrast, iT_reg development takes place in the periphery, their FOXP3 expression is inducible and their FOXP3 TSDR is fully methylated. The MSC-mediated generation of cells with an immunosuppressive function is of particular importance if one considers the fate of MSC after infusion; Eggenhofer et al. 38 showed recently that after intravenous administration into mice, MSC survive no longer than 24 h. This evident short lifespan of MSC in connection with their proven ability to prolong graft survival prompts further investigation to reveal how MSC accomplish long-term immunosuppression, and which mediators and mechanisms are involved in this phenomenon.

While we have studied previously the interaction between human adipose-tissue derived MSC (ASC) and natural T_reg 29, the aim of this study was to determine whether human ASC can generate functional de-novo iT_reg from CD25^-/dim effector T cells and to find evidence for the mechanisms involved in MSC-mediated T_reg induction.

Material and methods

Origin, isolation and culture of human ASC

Perirenal adipose tissue was removed surgically from living kidney donors and collected in minimum essential medium Eagle alpha modification (MEM-α) (Sigma-Aldrich, St Louis, MO, USA) supplemented with 2 mM L-glutamine (Lonza, Verviers, Belgium), 1% penicillin/streptomycin solution (P/S; 100 IU/ml penicillin, 100 IU/ml streptomycin; Lonza). Samples were obtained with written informed consent, as approved by the Medical Ethical Committee at Erasmus University Medical Center Rotterdam (protocol no. MEC-2006-190).

ASC were isolated, cultured and characterized as described previously 29. In brief, perirenal adipose tissue was disrupted mechanically and digested enzymatically with collagenase type IV (Life Technologies, Paisley, UK). ASC were expanded using ASC culture medium consisting of MEM-α with 2 mM L-glutamine, 1% P/S and 15% fetal bovine serum (FBS; Lonza) in a humidified atmosphere with 5% CO₂ at 37°C. Culture medium was refreshed twice weekly. At subconfluency, ASC were removed from culture flasks using 0·05% trypsin-ethylenediamine tetraacetic acid (EDTA) (Life Technologies) and reseeded at 1000 cells/cm². ASC were characterized by means of immunophenotyping and by their ability to differentiate into adipocytes and osteoblasts. ASC cultured between two and six passages were used. ASC from these passages did not differ in their ability to differentiate or to exert their immunosuppressive functions.

Isolation of peripheral blood mononuclear cells

Peripheral blood mononuclear cells (PBMC) were isolated from buffy coats of healthy blood donors (Sanquin, Rotterdam, the Netherlands) by density gradient centrifugation using Ficoll-Paque PLUS (density 1·077 g/ml; GE Healthcare, Uppsala, Sweden). Cells were frozen at −150°C until further use in Roswell Park Memorial Institute (RPMI)-1640 medium with GlutaMAX™-I (Life Technologies) supplemented with 1% P/S, 10% human serum (Sanquin) and 10% dimethylsulphoxide (DMSO; Merck, Hohenbrunn, Germany).

Isolation of effector cells and nT_reg from PBMC

CD25^–/dim cells (effector cells) and CD25^bright cells (nT_reg) were separated by means of CD25 MicroBeads II (Miltenyi Biotec, Bergisch Gladbach, Germany) and magnetic cell sorting, as described previously 29. Cell fraction purity was determined by flow cytometry using monoclonal antibodies (mAbs) against CD3-AmCyan (clone SK7), CD4-Pacific Blue (RPA-TA), CD25-phycoerythrin (PE)-cyanin 7 (Cy7) (epitope B; M-A251), CD127-PE (HIL-7R-M21; all BD Biosciences, San Jose, CA, USA); and FoxP3-allophycocyanin (APC) (PCH101; eBioscience, San Diego, CA, USA). Intracellular FoxP3 staining was carried out following the manufacturer's instructions of the anti-human FoxP3 staining set APC (eBioscience). Flow cytometric analyses were performed using the BD FACSCanto II flow cytometer and BD FACSDiva software (both BD Biosciences).

Mixed lymphocyte reaction and suppression assay

Mixed lymphocyte reactions (MLR) consisted of 5 × 10⁴ CD25^–/dim effector cells stimulated with 5 × 10⁴ γ-irradiated (40 Gy) allogeneic PBMC in round-bottomed 96-well plates (Nunc, Roskilde, Denmark) using PBMC culture medium (PCM) consisting of MEM-α supplemented with 2 mM L-glutamine, 1% P/S and 10% heat-inactivated human serum. Effector–stimulator cell combinations were chosen on the basis of a minimum of four HLA mismatches. The immunomodulatory capacities of ASC (various concentrations), nT_reg (1:10), ASC-induced CD4⁺CD25⁺CD127^– T cells (1:10) and control CD4⁺CD25^– T cells (1:10) (all ratios: indicated cells/effector cells) on the MLR were determined in suppression assays. After an 8-h incubation period on day 7, [³H]-thymidine incorporation (0·25 μCi/well; PerkinElmer, Groningen, the Netherlands) was measured using the Wallac 1450 MicroBeta TriLux (PerkinElmer). When MLR were performed in microtitre plates with different well sizes, the number of cells was adjusted accordingly. When applicable, 50 μl cell-culture supernatant was harvested prior to the addition of [³H]-thymidine and frozen at −80°C until further use.

Induction of CD4⁺CD25⁺CD127^–FoxP3⁺ T cells by ASC

CD25^–/dim effector cells were labelled using the PKH67 Green Fluorescent Cell Linker Kit (Sigma-Aldrich). For discrimination, allogeneic stimulator PBMC were labelled with PKH26 (PKH26 Red Fluorescent Cell Linker Kit; Sigma-Aldrich), according to the manufacturer's instructions. MLR were performed for 7 days in the absence or presence of ASC (1:40; ASC/effector cells). The PKH-label dilution caused by proliferation was measured by flow cytometry. After a 7-day incubation period in the absence or presence of 4 μg/ml basiliximab (Novartis Pharma, Nürnberg, Germany), cells were stained with mAbs against CD3-AmCyan (clone SK7), CD4-Pacific Blue (RPA-TA), CD8-peridinin chlorophyll (PerCP) (SK1), CD25-APC-Cy7 (epitope B; M-A251) and CD127-PE-Cy7 (HIL-7R-M21; all BD Biosciences) and FoxP3-APC (PCH101; eBioscience). Allogeneic stimulator PBMC were excluded from the analysis based on their PKH26-label. To confirm that basiliximab does not interfere with the binding of the monoclonal anti-CD25 antibody on epitope B, a competition staining was performed. In the presence of basiliximab no weakening of the CD25 staining was observed.

Isolation and function test of ASC-induced CD4⁺CD25⁺CD127^– T cells

ASC-induced CD4⁺CD25⁺CD127^– T cells were generated in primary MLR consisting of CD25^–/dim effector cells and allogeneic stimulator PBMC in the presence or absence of ASC (1:40; ASC/effector cells). Allogeneic stimulator PBMC were labelled with PKH67 (Sigma-Aldrich). After 7 days, cells were stained with mAbs against CD3-AmCyan (clone SK7), CD4-Pacific Blue (RPA-TA), CD25-PE-Cy7 (epitope B; M-A251), CD127-PE (HIL-7R-M21) and BD Via-Probe™ (7-AAD-PerCP) (all BD Biosciences). ASC-induced T_reg were defined as PKH67^–7-AAD^–CD3⁺CD4⁺CD25⁺CD127^– cells. Cell sorting was performed using the BD FACSAria II cell sorter (BD Biosciences).

Sorted PKH67^–7-AAD^–CD3⁺CD4⁺CD25⁺CD127^– cells (1:10) were reseeded into secondary MLR. PKH67^–7-AAD^–CD3⁺CD4⁺CD25^– cells (1:10) were used as negative control; nT_reg (1:10) obtained from PBMC by magnetic cell separation served as positive control (all ratios: indicated cells/effector cells). After an 8-h incubation period on day 7, [³H]-thymidine incorporation (0·25 μCi/well; PerkinElmer) was measured. Alternatively, sorted cell samples were washed twice with phosphate-buffered saline-diethylpyrocarbonate (PBS-DEPC) and frozen at −80°C until further use.

Quantitative DNA methylation analysis of the FOXP3 gene TSDR

To quantify DNA methylation of the FOXP3 gene TSDR, the EZ DNA Methylation-Direct™ Kit (Zymo Research, Irvine, CA, USA) was used according to the manufacturer's instructions. Cell pellets were digested with proteinase K prior to bisulphite conversion. During DNA bisulphite treatment, unmethylated cytosines are converted into uracils while methylated cytosines remain unmodified. After bisulphite treatment, the TSDR of the FOXP3 gene was amplified by quantitative real-time PCR (qPCR) using the StepOnePlus™ real-time PCR system and the TaqMan® Genotyping Master Mix (all Applied Biosystems). Methylation-specific and demethylation-specific amplification primers and probes were chosen as suggested by Wieczorek et al. 39. The percentage of cells within a cell fraction with a methylated TSDR was calculated using the ratio of amplified methylated TSDR copies and the sum of amplified methylated and unmethylated TSDR copies. To correct for the X-linked nature of the FOXP3 gene, results obtained from female PBMC donors were multiplied by 2. Intra- and interassay variations were determined by negative controls and positive reference samples.

Flow cytometric characterization of CD4⁺CD25⁺CD127^–FoxP3⁺ T cells and CD4⁺CD25⁺FoxP3^– T cells

The induction of CD4⁺CD25⁺CD127^–FoxP3⁺ T cells by ASC (1:40; ASC/effector cells) was initiated as described above. Allogeneic stimulator PBMC were labelled with either PKH67 or BD Horizon Violet Cell Proliferation Dye 450 (VPD450; BD Biosciences). After 7 days, cells were stained with mAbs against CD3-AmCyan (clone SK7), CD4-PerCP (SK3), CD25-APC-Cy7 (epitope B; M-A251), CD127-PE-Cy7 (HIL-7R-M21), cytotoxic T lymphocyte antigen-4 (CTLA-4)-APC (BNI3; all BD Biosciences), glucocorticoid-induced TNFR-related protein (GITR)-fluorescein isothiocyanate (FITC) (110416; R&D Systems Europe Ltd, Abingdon, UK), Helios-Pacific Blue (22F6; BioLegend, San Diego, CA, USA) and FoxP3-PE or FoxP3-APC (both PCH101; eBioscience). Fluorescence minus one (FMO) controls were used to determine negative expression.

mRNA expression analysis

ASC were cultured alone or co-cultured with MLR (1:5; ASC/effector cells) for 7 days. In co-cultures, MLR were separated from ASC by cell culture inserts with permeable membrane supports (0·4 μm pore size; Greiner Bio-One, Alphen a/d Rijn, the Netherlands). ASC were harvested and washed twice with PBS-DEPC. Cells were either handled immediately or snap-frozen in liquid nitrogen and stored at −80°C. Total RNA was purified using the High Pure RNA Isolation Kit (Roche Diagnostics), according to the manufacturer's instructions. Complementary DNA (cDNA) was synthesized by reverse transcription using random primers. qPCR was performed using 500 ng cDNA, the StepOnePlus™ real-time PCR system, TaqMan Universal PCR Master Mix and the assay-on-demand primer/probes for IL-2 (Hs00174114.m1) (Applied Biosystems, Foster City, CA, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA served as endogenous control for normalization (Hs99999905.m1; Applied Biosystems). Intra- and interassay variations were determined by negative controls and positive reference samples. Changes in target gene expression relative to the GAPDH gene were quantified using the comparative C_T method 40. Fold changes fewer than three were considered insignificant.

Cytometric bead array (CBA)

Supernatants were obtained from ASC cultures, MLR and ASC-MLR co-cultures (1:40; ASC/effector cells) in the absence or presence of 4 μg/ml basiliximab (Novartis Pharma) after a 7-day incubation period. They were frozen until further use at −80°C. IL-2 concentrations were measured using the BD cytometric bead array human T helper type 1 (Th1)/Th2/Th17 Cytokine Kit (BD Biosciences), according to the manufacturer's instructions.

Flow cytometric analysis of IL-2 expression by CD4⁺CD25⁺CD127^–FoxP3⁺ T cells and CD4⁺CD25⁺FoxP3^– T cells

CD4⁺CD25⁺CD127^–FoxP3⁺ T cells were induced by ASC (1:40; ASC/effector cells) as described above. CD25^–/dim effector cells were labelled with VPD450 (BD Biosciences). For discrimination, allogeneic stimulator PBMC were labelled with PKH67 (Sigma-Aldrich). MLR were performed in the absence or presence of ASC and 4 μg/ml basiliximab (Novartis Pharma). After 7 days, cells were stimulated with 50 ng/ml phorbol 12-myristate 13-acetate (PMA) and 1 μg/ml calcium ionomycin (both Sigma-Aldrich) in the presence of BD GolgiStop (BD Biosciences) for 4 h. Subsequently cells were stained with mAbs against CD3-AmCyan (clone SK7), CD4-PerCP (SK3), CD25-APC-Cy7 (epitope B; SK1), CD127-PE-Cy7 (HIL-7R-M21), IL-2-APC (5344.111; all BD Biosciences) and FoxP3-PE (PCH101; eBioscience). FMO controls were used to determine negative expression.

Statistical analysis

Statistical analyses were performed by means of one-way analysis of variance (anova), Bonferroni's multiple comparison tests and (un)paired t-tests using GraphPad Prism 5 software (GraphPad Software, San Diego, CA, USA). A P-value lower than 0·05 was considered statistically significant. Two-tailed P-values are stated.

Results

ASC inhibit the proliferation of CD25^–/dim effector cells dose-dependently and induce CD4⁺CD25⁺CD127^–FoxP3⁺ T cells

The immunosuppressive effect of ASC on the proliferation of CD25^–/dim effector cells was examined by means of [³H]-thymidine incorporation. CD25^–/dim effector cells were stimulated with γ-irradiated allogeneic PBMC for 7 days resulting in a strong proliferative activity of these cells (Fig. 1a). Co-culture with third-party ASC suppressed the effector cell proliferation in a dose-dependent manner (one-way anova, P < 0·0001), confirming our previously published data 29. ASC reduced significantly the proliferation from a ratio of 1:80 (ASC/effector cells). ASC suppressed the proliferation of CD4⁺ T cells and CD8⁺ T cells (Fig. 1b).

Fig. 1.

Effect of adipose-tissue derived mesenchymal stem cells (ASC) on the proliferation of alloactivated CD25^–/dim cells. (a) Mixed lymphocyte reaction (MLR) (white bar) consisted of CD25^–/dim effector cells stimulated with γ-irradiated allogeneic peripheral blood mononuclear cells (PBMC) with a minimum of four human leucocyte antigen (HLA)-mismatches. MLR were performed in the presence of various ASC concentrations for 7 days (n = 9, mean ± standard error of the mean, one-way anova: P < 0·0001). The Bonferroni multiple comparison test was used to compare the different conditions with the MLR condition. *P < 0·05; **P < 0·01; ***P < 0·001. (b) Proliferation of CD4⁺ T cells and CD8⁺ T cells in a 7-day MLR in the absence (dotted histograms) and presence of ASC (1:40 ratio, ASC/effector cells; solid histograms) was tracked using PKH67-labelling.

The effect of ASC on the generation of CD4⁺CD25⁺CD127^–FoxP3⁺ T cells from CD25^–/dim cells was determined by flow cytometry. At a 1:40 ratio, ASC reduced the proliferation of CD25^–/dim effector cells by 59% (Fig. 1a) and mediated a 2·1-fold increase in the percentage of CD25⁺CD127^–FoxP3⁺ cells within the CD4⁺ T cell population (Fig. 2).

Fig. 2.

Effect of adipose-tissue derived mesenchymal stem cells (ASC) on the induction of CD4⁺CD25⁺CD127^–forkhead box P3 (FoxP3)⁺ T cells. Mixed lymphocyte reaction (MLR) (white bar) consisted of CD25^–/dim effector cells stimulated with γ-irradiated allogeneic peripheral blood mononuclear cells (PBMC) and were performed for 7 days. In MLR co-cultures ASC were used at a 1:40 ratio (ASC/effector cells). The percentage of CD25⁺CD127^–FoxP3⁺ cells within the CD4⁺ T cell population was determined by flow cytometry. (a) Representative examples are shown. (b) Data of multiple experiments are shown. n = 6, mean ± standard error of the mean; paired t-test; **P < 0·01.

ASC-induced CD4⁺CD25⁺CD127^– T cells are immunosuppressive

The suppressive capacity of ASC-induced CD4⁺CD25⁺CD127^– T cells was determined by means of [³H]-thymidine incorporation and compared to the effect of nT_reg. Sorted ASC-induced CD4⁺CD25⁺CD127^– T cells (median purity: 98·7%; range 95·3%–99·7%) were added at a 1:10 ratio to secondary MLR. They inhibited the proliferation of CD25^–/dim effector cells as effectively as nT_reg (1:10; 61% suppression versus 48%, respectively; P = 0·402; Fig. 3). Hence, ASC induce functional regulatory T cells. MLR-induced CD4⁺CD25⁺CD127^– T cells (median purity: 99·5%; range 98·5–99·8%) also suppressed proliferation (1:10; 87%). Sorted CD4⁺CD25^– T cells from an MLR or MLR–ASC co-culture (1:10; median purities 99·8 and 99·7%; ranges 99·5–99·8% and 99·3–99·9%), as negative controls, did not inhibit the proliferative activity of the effector cells.

Fig. 3.

Effect of adipose-tissue derived mesenchymal stem cells (ASC)-induced CD4⁺CD25⁺CD127^– T cells on the proliferation of alloactivated CD25^–/dim effector cells. Primary mixed lymphocyte reaction (MLR) consisted of CD25^–/dim effector cells stimulated with γ-irradiated allogeneic peripheral blood mononuclear cells (PBMC). For the induction of CD4⁺CD25⁺CD127^– T cells, ASC (1:40; ASC/effector cells) were added to primary MLR. After an incubation period of 7 days, CD4⁺CD25⁺CD127^– T cells and CD4⁺CD25^– T cells were separated from the total lymphocyte populations from MLR or ASC–MLR co-cultures by cell sorting and added to secondary MLR at a 1:10 ratio (indicated cells/effector cells). CD25^bright natural regulatory T cells (nT_reg) (1:10) were incubated with the same MLR to compare the function of the sorted fractions to the suppressive function of nT_reg. For statistical analyses, the MLR condition was used as reference. n = 4, mean ± standard error of the mean; paired t-test; **P < 0·01.

Functional ASC-induced CD4⁺CD25⁺CD127^– T_reg are de-novo cells

To determine the origin of the ASC-induced T_reg, i.e. iT_reg or nT_reg, the methylation status of the TSDR was investigated. The percentage of cells with a methylated FOXP3 gene TSDR present in different cell fractions was determined by means of quantitative DNA methylation analysis. The CD25^–/dim fraction, obtained from PBMC by magnetic cell separation, consisted almost entirely (98·1%) of cells with a methylated FOXP3 TSDR (Fig. 4a). In contrast, the nT_reg fraction consisted to 67·3% of cells with a demethylated FOXP3 TSDR, confirming their thymic origin. After 7-day MLR in the presence of ASC (1:40; ASC/effector cells) using allostimulated CD25^–/dim cells as effector cells, sorted fractions of CD4⁺CD25⁺CD127^– iT_reg and CD4⁺CD25^– T cells contained 83·7 and 99·9% cells with a methylated FOXP3 TSDR, respectively. The small percentage of cells (16·3%) with a demethylated FOXP3 TSDR in the CD4⁺CD25⁺CD127^– iT_reg fraction probably represents proliferated nT_reg which were present in the initial CD25^-/dim fraction (1·9%). This demonstrates that ASC expand nT_reg, but that the majority of the CD4⁺CD25⁺CD127^– T_reg found after 7-day MLR–ASC co-culture are induced from CD25^–/dim cells.

Fig. 4.

Characterization of adipose-tissue derived mesenchymal stem cells (ASC)-induced CD4⁺CD25⁺CD127^– regulatory T cells (T_reg). (a) Methylation analysis of the FOXP3 gene T_reg-specific demethylated region (TSDR). CD25^–/dim effector cells and CD25^bright natural T_reg (nT_reg) were obtained from peripheral blood mononuclear cells (PBMC) by means of cell separation. CD4⁺CD25^– T cells and ASC-induced CD4⁺CD25⁺CD127^– T_reg were isolated from lymphocyte populations by cell sorting after 7-day mixed lymphocyte reaction (MLR)–ASC co-cultures; co-cultures consisted of CD25^–/dim effector cells, γ-irradiated allogeneic PBMC and ASC (1:40; ASC/effector cells). The methylation status of the FOXP3 TSDR in the different cell populations was determined. The percentage of cells with a methylated TSDR was calculated using the ratio of amplified methylated TSDR copies and total TSDR copies. Results obtained from female PBMC donors were multiplied by 2; n = 7; six male, one female; mean ± standard error of the mean (s.e.m.). (b) Gating strategy to discriminate CD4⁺CD25⁺CD127^–forkhead box P3 (FoxP3)⁺ T cells from CD4⁺CD25⁺FoxP3^– T cells. A representative example for MLR–ASC co-culture (1:40; ASC/effector cells) is shown. (c) Expression of glucocorticoid-induced tumour necrosis factor receptor (TNFR)-related protein (GITR), cytotoxic T-lymphocyte antigen-4 (CTLA-4) and Helios. Flow cytometric analyses of protein expression by CD4⁺CD25⁺FoxP3^– T cells (white histogram) and CD4⁺CD25⁺CD127^–FoxP3⁺ T cells (grey histogram) are shown. Fluorescence minus one (FMO) controls are depicted as dotted histogram. Representative examples for MLR–ASC co-culture (1:40; ASC/effector cells) are shown. Data of multiple experiments are shown in d. (d). Expression of GITR, CTLA-4 and Helios by CD4⁺CD25⁺FoxP3^– T cells (white bars) and CD4⁺CD25⁺CD127^–FoxP3⁺ T cells (checkered bars). Protein expression is shown as fold changes of median fluorescence intensity (MFI) between CD4⁺CD25⁺FoxP3^– T cells and CD4⁺CD25⁺CD127^–FoxP3⁺ T cells. Expression levels of CD4⁺CD25⁺FoxP3^– T cells were used as reference. n = 3, mean ± s.e.m.; paired t-test; **P < 0·01.

To characterize further the ASC-induced CD4⁺CD25⁺CD127^–FoxP3⁺ T_reg (Fig. 4b), their expression of GITR, CTLA-4 and Helios was investigated (Fig. 4c,d). While the expression levels of GITR were similar for iT_reg and CD4⁺CD25⁺FoxP3^– T cells, iT_reg showed the tendency to have higher Helios levels and expressed 3·2-fold more CTLA-4 than FoxP3^– T cells (P = 0·003).

The induction of CD4⁺CD25⁺CD127^–FoxP3⁺ T_reg by ASC coincides with increased IL-2 levels and is IL-2 pathway-dependent

IL-2 is known to be required for the expansion and function of nT_reg. Therefore, its involvement in the ASC-mediated induction of CD4⁺CD25⁺CD127^–FoxP3⁺ T_reg was investigated (Fig. 5). Seven-day MLR were cultured in the absence or presence of ASC (1:40; ASC/effector cells) and concentrations of secreted IL-2 in the supernatant were analysed. IL-2, 20·1 pg/ml, was detected in the supernatant of an MLR of CD25^–/dim effector cells stimulated with allogeneic PBMC. Non-activated ASC and MLR-activated ASC do not express IL-2 (data not shown). However, activated ASC mediated a 15·1-fold increase in IL-2 levels by effector cells during MLR suppression and induction of CD4⁺CD25⁺CD127^–FoxP3⁺ T_reg. The monoclonal anti-IL-2 receptor antibody basiliximab (4 μg/ml) effectively inhibited the IL-2 consumption in MLR and MLR–ASC co-cultures (1:40; ASC/effector cells; Fig. 5a); in the presence of basiliximab IL-2 concentrations in MLR and MLR–ASC co-cultures accumulated to 217 and 525 pg/ml, respectively.

Fig. 5.

Effect of mixed lymphocyte reaction (MLR) adipose-tissue-derived mesenchymal stem cells (ASC) co-culture and basiliximab on interleukin (IL)-2 expression. (a) IL-2 levels in cell culture supernatants. MLR (white bars) consisted of CD25^–/dim effector cells stimulated with γ-irradiated allogeneic peripheral blood mononuclear cells (PBMC) and were performed for 7 days in the presence and absence of basiliximab. In co-cultures with MLR, ASC were used at a 1:40 ratio (ASC/effector cells; black bars). The IL-2 concentration in the cell culture supernatants was determined by cytometric bead array (CBA). n = 6, mean ± standard error of the mean (s.e.m.); paired t-test; ** P < 0·01. (b) Percentage of IL-2-expressing cells within the CD4⁺CD25⁺ forkhead box P3 (FoxP3)^– T cells and CD4⁺CD25⁺CD127^–FoxP3⁺ T cells after 7-day MLR–ASC co-culture (representative examples are shown). (c) Data of multiple experiments showing the percentage of IL-2-expressing cells within the CD4⁺CD25⁺FoxP3^– T cells (white bars) and CD4⁺CD25⁺CD127^–FoxP3⁺ T cells (chequered bars); n = 5; mean ± s.e.m.; paired t-test; ***P < 0·001. (d) Levels of expressed IL-2. Median fluorescence intensity (MFI); n = 5.

To gain knowledge regarding which cell subset contributed most to the IL-2 concentrations detected in the supernatant, the percentage of IL-2-expressing cells within the CD4⁺CD25⁺CD127^–FoxP3⁺ iT_reg and CD4⁺CD25⁺FoxP3^– T cells was investigated. The majority of the IL-2-expressing cells were CD4⁺CD25⁺FoxP3^– T cells (Fig. 5b,c). The IL-2 levels produced were similar between the IL-2 expressing CD4⁺CD25⁺CD127^–FoxP3⁺ iT_reg and CD4⁺CD25⁺FoxP3^– T cells (Fig. 5d).

To examine the effect of the IL-2 pathway on the ASC-mediated generation of functional CD4⁺CD25⁺CD127^–FoxP3⁺ iT_reg from CD25^–/dim cells, IL-2 binding to its receptor was blocked with basiliximab. MLR were performed for 7 days in the absence and presence of ASC (1:40; ASC/effector cells). Basiliximab reduced the percentage of CD25⁺CD127^–FoxP3⁺ cells within the CD4⁺ T cell population (Fig. 6). In the presence of basiliximab, ASC did not induce a proportional increase of CD25⁺CD127^–FoxP3⁺ cells within the CD4⁺ T cell population.

Fig. 6.

Effect of basiliximab on the induction of CD4⁺CD25⁺CD127^– forkhead box P3 (FoxP3)⁺ regulatory T cells (T_reg). Mixed lymphocyte reaction (MLR) (white bars) consisted of CD25^–/dim effector cells stimulated with γ-irradiated allogeneic peripheral blood mononuclear cells (PBMC) and were performed for 7 days. In MLR co-cultures adipose tissue-derived mesenchymal stem cells (ASC) were used at a 1:40 ratio (ASC/effector cells; black bars). In the absence and presence of basiliximab, the percentage of CD25⁺CD127^–FoxP3⁺ cells within the CD4⁺ T cell population was determined by flow cytometry; n = 6, mean ± standard error of the mean; paired t-test; *P < 0·05; **P < 0·01; ***P < 0·001.

Discussion

Knowledge about the underlying mechanisms of how ASC contribute to a reduced responsiveness of immune effector cells is scarce. The present study provides the first evidence that human ASC mediate their immunosuppressive effect via the formation of functional de-novo iT_reg. Our results are in line with earlier in-vivo studies with bone-marrow MSC in animal models and a recent case report of MSC administration to renal transplant patients; these studies also observed proportional increases of T_reg 9,35,41,42. However, as these groups did not investigate the T_reg origin and T_reg functionality, it is not clear whether the observed rises in T_reg percentages are a consequence of the de-novo formation of T_reg or a result of the expansion of existing nT_reg. We found that ASC mediate the generation of iT_reg from effector T cells and that these newly formed cells have immunosuppressive capacities. iT_reg also formed in MLR without ASC, but at lower numbers. While ASC-induced T_reg showed similar suppressive capacities to nT_reg, MLR-induced T_reg had a stronger inhibitory effect. The induction of a more potent immunosuppressive phenotype in these cells might be due to their generation under highly proinflammatory conditions. In the presence of ASC effector cell proliferation is reduced. Because of this milder environment higher numbers of less effective iT_reg are generated. Therefore, the suppressive strength of iT_reg might be proportional to the stimulus under which they are induced. Further, due to the presence of a few nT_reg in our initial effector T cell population we were able to infer that ASC also mediate an increase in nT_reg.

Although the first evidence of extrathymic conversion of conventional T cells into iT_reg was found almost a decade ago, no iT_reg-specific marker has yet been identified 43–45. The only tool currently available to distinguish iT_reg from nT_reg is the determination of the methylation status of the TSDR. While nT_reg have a demethylated TSDR, this specific region of the FOXP3 gene is methylated in iT_reg 37. The high percentage of cells with a methylated TSDR in the ASC-induced T_reg fraction indicates that the vast majority of these cells originated from CD25^–/dim effector cells. To characterize further the ASC-induced T_reg we investigated their expression of GITR, CTLA-4 and Helios. GITR expression was similar in both iT_reg and CD4⁺CD25⁺FoxP3^– T cells. In contrast, iT_reg showed a tendency towards a higher Helios expression and expressed significantly higher CTLA-4 levels than FoxP3^– T cells. It has been described that CTLA-4 expression is up-regulated upon binding of FoxP3 to the promoter of the CTLA-4 gene 46–50. Therefore, the observed difference in CTLA-4 expression between iT_reg and FoxP3^– T cells can be attributed to their differing FoxP3 expression. Whether Helios is a potential marker for human iT_reg is an ongoing debate in the field, as conflicting results have been reported 51–53.

In search of possible underlying mechanisms involved in MSC-mediated iT_reg generation, it has been reported recently that programmed death ligand-1 (PD-L1), a protein expressed by MSC, promotes the differentiation of Th1 cells into T_reg 22,54. Other molecules that were suggested to play a role in the induction of T_reg by MSC are IDO, prostaglandin E₂ (PGE₂), transforming growth factor (TGF)-β and haem oxygenase-1 (HO-1) 41–58. In this study we focused on the role of IL-2 in ASC-mediated T_reg induction, as IL-2 drives T cell proliferation and, paradoxically, is also essential for tolerance induction by regulating nT_reg function 59. We observed that in the presence of ASC, the IL-2 concentration in the cell culture supernatant increased. Because IL-2 is not expressed by ASC, the IL-2 surplus originated from the alloactivated effector cells. Upon T cell-receptor engagement and co-stimulation, activated effector T cells consume IL-2. During the inhibition of effector T cell proliferation by ASC, IL-2 concentrations in the cell culture supernatant rose. This indicates that although their proliferation is suppressed, effector T cells remain activated and continue to secrete IL-2. The diminished proliferation rate of effector T cells causes a reduced IL-2 consumption and hence a surplus of IL-2. High levels of IL-2 are instrumental for T_reg induction. In the presence of basiliximab, we found further accumulation of IL-2 in the supernatant. In its function as anti-rejection therapy in kidney transplant patients, basiliximab prevents the binding of IL-2 to its receptor and the subsequent uptake of IL-2. As a result, basiliximab inhibits T cell proliferation but also iT_reg generation, indicating that T_reg induction by ASC is IL-2 pathway-dependent. Surprisingly, IL-2 was expressed not only by CD4⁺CD25⁺FoxP3^– T cells but also by a smaller percentage of immunosuppressive ASC-induced T_reg. Phenotypic adaptation of T_reg in response to the cytokines present in their environment is a known phenomenon 60. The inhibition of T cell proliferation is not an exclusive effect of MSC, as other stromal cells share this characteristic 61,62 Hence, it is possible that these cells can also create an environment that favours iT_reg generation.

The graft-supporting effects of MSC in animal models and the positive results of MSC treatment for graft-versus-host disease (GVHD) in patients who underwent haematopoietic stem cell transplantation (HSCT) are strongly convincing 8,63. The reported short lifespan of MSC after infusion, however, indicates that MSC are only initiators of these effects. Long-term, MSC do not actively promote immunosuppression themselves. Nevertheless, they are able to pass on their immunosuppressive capabilities through the induction of functional de-novo T_reg, the expansion of nT_reg or possibly through other immune cells which remain to be identified.

Despite continuous efforts, the translation of the experimental success of T_reg-mediated immune regulation to its application in clinical settings proves difficult. In animal models, ex-vivo- or in-vivo-generated T_reg prevent type 1 diabetes, reduce the severity of experimental autoimmune encephalomyelitis (EAE), control the acute and chronic rejection of allografts and attenuate or prevent GVHD 64–70. In humans, thus far, ex-vivo-expanded T_reg have been used only for the treatment of patients undergoing HSCT. In these few clinical studies, T_reg were able to reduce the incidence of GVHD or to prevent GVHD 71–73. Due to the heterogeneity of the T_reg population and the lack of a specific marker for human T_reg, ex-vivo T_reg expansion bears the risk of contamination with effector T cells. Therefore, in-vivo induction of functional T_reg mediated by MSC represents a beneficial effect of MSC therapy in addition to its immunosuppressive effect on effector T cells. Once the functionality of MSC-induced T_reg has been confirmed in autoimmune disease patients and allograft recipients, strategies to enhance the MSC-mediated in-vivo generation of iT_reg may be considered. Possible approaches could be MSC treatment combined with low-dose IL-2 therapy or the use of rapamycin and rabbit anti-thymocyte globulin (rATG), anti-rejection drugs which were found to advance nT_reg expansion and to allow T_reg induction, respectively 70,74–76.

In conclusion, our study demonstrates that human adipose tissue-derived MSC induce T_reg from effector T cells and that these de-novo T_reg are immunosuppressive. In conjunction with the well-known MSC function of preventing immune cell proliferation, our findings encourage the advancement of MSC therapy into clinical development for autoimmunity, GVHD and allograft rejection.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose.