Subcutaneous adipose tissue from healthy human donors that became available as a waste product during kidney donation procedures was collected after obtaining written informed consent as approved by the Medical Ethical Committee of the Erasmus University Medical Centre Rotterdam (protocol no. MEC-2006-190). The tissue was collected in minimum essential medium-α (MEM-α) (Sigma Aldrich, St. Louis, MO, USA) supplemented with penicillin (100 IU/ml), streptomycin (100 mg/ml) (1% P/S; Lonza, Verviers, Belgium), and 2 mM L-glutamine (Lonza) and stored at 4°C for 3–16 h. MSC were isolated as described previously (20). Briefly, adipose tissue was mechanically disrupted and digested enzymatically with 0.5 mg/mL collagenase type IV (Life Technologies, Paisley, UK) in RPMI 1640 Medium with glutaMAX (Life Technologies) for 30 min at 37°C under continuous shaking. Cultures were kept at 37°C, 5% CO₂, and 95% humidity and refreshed weekly with MEM-α with 1% P/S, and 15% heat-inactivated fetal bovine serum (FBS; Lonza). At 90% confluence, adherent cells were removed from culture flasks by incubation in 0.05% trypsin-EDTA (Life Technologies, Bleiswijk, The Netherlands) at 37°C and cells used for experiments or frozen at −150°C until further use. MSC were used for experiments between passages 2 and 5 and their phenotypic markers and osteogenic and adipogenic potential were tested as described before (21). MSC from 19 different donors were used in the experiments.

Mesenchymal stem or stromal cells were pretreated for 4 days with IFN-γ (50 ng/ml; Life technologies). For co-culture experiments, MSC were washed with phosphate buffered saline (PBS) and detached by incubation with 0.05% trypsin-EDTA before seeding them in 96 well-plates in Iscove’s Modified Dulbecco’s Medium (IMDM, Lonza) with 10% heat inactivated FBS. Phenotypical characteristics of MSC before and after IFN-γ were assessed measuring several markers on their surface: CD13-PeCy7 (clone L138), CD31-V450 (clone WM59), CD45-APC-H7 (clone 2D1), HLA-ABC-APC (clone G46-2.6), HLA-DR PerCP (clone L243) and CD73-PE (clone AD2; all BD Biosciences), CD90-APC (clone Thy-1A1), and CD105-FITC (clone 166707; all R&D Systems, Minneapolis, MN, USA) and PD-L1 PE (clone B7-H1; Biolegend, San Diego, CA, USA) by Flow Cytometry and optical microscopy morphology (Figure S1 in Supplementary Material).

The activity of IDO was determined by the measurement of L-kynurenine in the supernatant of four MSC cultures as described previously (22). Briefly, MSC were seeded at a density of 100,000 cells/well in a 6 wells plate and cultured for 4 days with or without 50 ng/mL IFN-γ. 30% trichloroacetic acid was added to the supernatant in a 1:3 ratio. Samples were incubated for 30 min at 50°C and spun down at 12,000 rpm for 5 min. Samples were plated in a 96 wells flat bottom plate and diluted 1:1 in Ehrlich reagent [200 mg 4-dimethylaminobenzaldehyde (Sigma-Aldrich, St. Louis, MO, USA) in 10 ml of glacial acetic acid]. Absorbance was read at 490 nm using a Wallac Victor2 1420 multilabel plate reader (Perkin Elmer, Waltham, MA, USA).

Spleens were obtained from post-mortal kidney donors (Erasmus MC Hospital, Rotterdam) and anonymously used for research purposes as described in article 13 of The Netherlands law of organ donation (Wet op Orgaandonatie, WOD). All samples and data were analyzed anonymously. Spleens were mechanically disrupted and filtered through a 70-µm cell strainer (Greiner Bio-one, Alphen a/d Rijn, The Netherlands) to obtain a single-cell suspension. Mononuclear cells, isolated using Ficoll-Paque (Amersham Pharmacia Biotech, Uppsala, Sweden) density gradient, were stored at −150°C until use. Upon thawing, quiescent B cells were isolated by negative selection using anti-CD43-magnetic beads (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) (23). Purity was determined by flow cytometry (FACS Canto II). Typically, cell suspensions consisted of >98% pure CD19⁺ B cells. B cells from spleens from 12 different donors were used it the experiments.

B cells were co-cultured in IMDM-10%FBS with a cocktail to mimic antigen and T cell help: 10 mg/ml F(ab)2 anti-IgM (Jackson, ImmunoResearch laboratories, Inc., West Grove, PA, USA), 10³ IU IL-2 (Proleukin, Prometheus laboratories Inc., San Diego, CA, USA), and 5 mg/ml anti-CD40 agonistic monoclonal antibody (Bioceros, Utrecht, The Netherlands). In some of the experiments, 200 µM tryptophan (TRP, l-tryptophan, Sigma-Aldrich) was added to the stimulation cocktail to counteract the activity of IDO.

24-wells plates with 0.4 μm pore polycarbonate membrane inserts (Costar, Corning, Kennebunk, ME, USA) were used for the TW cultures. MSC were seeded on the membrane of the inserts and B cells were added to the lower chamber at a ratio MSC:B cells 1:5. After 7 days, inserts were removed; and B cells from the lower chamber were collected for further analysis and B cell subsets characterization.

To study the effect of cell surface molecules but not the secreted factors, MSC were inactivated as previously described (24). Shortly, MSC were heated in suspension in PBS in parafilm-sealed tubes by 30 min incubation at 50°C in a temperature-regulated water bath. The inactivated cells were then washed and counted and used for further experiments.

B cells were labeled by incubation with 5,6-carboxy-succinimidyl-fluoresceine-ester (CFSE) (Molecular Probes Invitrogen, Karlsruhe, Germany) for 10 min at 37°C.

After 7 days, B cells were collected and processed for flow cytometric analysis (FACS Canto II, Diva Software, BD Biosciences, San Jose, CA, USA), and supernatants were stored at −80°C for cytokine and Immunoglobulin-G (IgG) determination. The antibodies used for flow cytometry phenotyping were as follows: CD27-PE-Cy7 (clone 0323), CD38-PE (clone HB7), CD19-BV512 (clone HIB19) and CD24-APC (clone SN3 A5-2H1D) (eBioscience, San Diego, CA, USA), IL-10-Bv421 (Clone Jes3-9D7, Biolegend), and Via Probe for determination of cell viability (BD Biosciences, San Jose, CA, USA). After 7 days, proliferation of B cells was assessed by measuring CFSE dye dilution on a FACSCanto II flow cytometer (BD Biosciences). 12 h before harvesting the cells, Monensin (Golgi Stop, BD Biosciences) was added to the wells and the intracellular staining was performed without restimulation using Intrastain kit (Dako, Denmark).

Supernatants from MSC-B cell co-cultures kept at −80°C were thawed and used for measurement of cytokine levels. IL-10 was quantified using a Milliplex kit (Merck Millipore, Amsterdam, The Netherlands) according to manufacturer’s instructions. Human cytokine standards were provided by the kit and a standard curve was prepared from 10,000 to 3.2 pg/ml. Samples and standards mixed with antibody-coated magnetic beads were incubated overnight in a 96-well plate at 4°C under continuous agitation. Plates were washed and incubated with detection antibodies for 1 h. Finally, plates were washed and incubated with streptavidin-phycoerythrin for 30 min. The samples were measured on a Luminex 100/200 cytometer (Luminex, Austin, TX, USA) using Xponent software.

Plates were coated with goat anti human Ig-UNLB (Southern Biotechnology Associates; Birmingham, AL, USA). Plates were washed with PBS 0.05% Tween and blocked with PBS 5% FBS for 2 h. Diluted samples and standard IgG (Sigma-Aldrich) were added to the plate and incubated for 90 min. IgG-HRP (My Biosource; San Diego, CA, USA) was used as a conjugate and 3,3,5,5-tetramethylbenzidine (TMB) was used to visualize bound IgG. Absorbance was read at 595 nm using a Wallac Victor2 1420 multilabel plate reader (Perkin Elmer, Waltham, MA, USA).

After 7 days of co-culture, B cells were recovered, pelleted in PBS–DEPC and snap frozen. RNA was isolated and 500 ng was used for cDNA synthesis as described previously (25). Gene expression was determined by real-time RT-PCR using universal PCR master mix (Life Technologies) and an assay-on-demand for IL-10 (Hs00174086.m1) (Applied Biosystems, Foster City, CA, USA) and analyzed on an ABI PRISM 7700 sequence detector (Applied Biosystems). Data are expressed as relative copy number of the PCR products with respect to the housekeeping gene GAPDH.

Data are expressed as means ± SEM. Significant differences within groups were calculated using repeated measures non-parametric analysis of variance (ANOVA: Friedman test) with Dunnett’s posttest performed by GraphPad Prism 5 software (GraphPad Software, San Diego, CA, USA). P values were indicated as * for P < 0.05; ** for P < 0.01; and *** for P < 0.001.

Previously, we showed that the inhibition of B cell proliferation by MSC was dependent on the presence of T cells (11). We hypothesized that MSC needed to be activated by IFN-γ secreted by T cells to mediate their anti-proliferative effects on B cells.

Here, we analyzed the anti-proliferative capacity of MSC and IFN-γ-pretreated MSC (MSC–IFN-γ) on anti-CD40, anti-IgM, and IL-2 stimulated B cells using flow cytometry. After 7 days of co-culturing with MSC or MSC–IFN-γ, viable naïve and memory B cells were distinguished based on intensity of CD27 as shown in Figure 1A.

Co-culture of B cells with MSC significantly increased the proliferation of the total B cell population (Figures 1B,C). MSC that were pretreated with IFN-γ did not increase the proliferation of B cells but, by contrast, inhibited B cell proliferation from 65 to 40%. Naïve and memory B cell subsets (CD27⁻ and CD27⁺, respectively) showed similar increases in proliferation when co-cultured with MSC and inhibition of proliferation when co-cultured with MSC–IFN-γ.

These results show that MSC need to be pre-activated with IFN-γ to bring about their anti-proliferative effect on B cells.

Mesenchymal stem or stromal cells reduced IgG production by activated B cells (Figure 2A). Pretreatment of MSC with IFN-γ significantly enhanced the inhibitory effect of MSC on IgG production. In accordance with the reduced proliferation, an even stronger reduction of IgG levels was measured in the supernatant of B cells co-cultured with MSC–IFN-γ.

To investigate whether MSC-induced B cells with a regulatory phenotype, frequencies of CD19⁺ CD38^high CD24^high transitional B cells (Bregs), and IL-10 production were measured. After 7 days of co-culturing MSC and MSC–IFN-γ with T cell-like stimulated B cells, the percentage of Bregs was measured using flow cytometry as shown in Figure 2B. MSC significantly induced an increase of this subset. By contrast, IFN-γ-pretreated MSC were not able to induce an increase in Bregs (Figures 2B,C). In accordance with this, the absolute number of Bregs was significantly increased when MSC were co-cultured with B cells (Figure 2D).

To analyze whether the induced cells had regulatory potential, the anti-inflammatory cytokine IL-10 gene expression was analyzed. MSC induced a higher trend in IL-10 gene expression, while MSC–IFN-γ did so only to a very low extent (Figure 2E). In accordance with this, IL-10 protein levels were significantly increased in the B cell and MSC co-cultures supernatants, whereas no increase in IL-10 levels was found in the supernatant of B cell and MSC–IFN-γ (Figure 2F). The proportion of IL-10-producing B cells was also analyzed in the different conditions by intracellular staining and accordingly we identified a higher percentage in the co-culture with MSC (Figure 2G).

To further analyze the phenotype of the IL-10-producing B cells induced by MSC we performed intracellular IL-10 staining. The transitional CD38^high CD24^high subset showed the highest percentage of IL-10⁺ cells, although also within the naïve CD38^int CD24^int B cell subset significant numbers of IL-10-producing cells were found (data not shown). In absolute numbers, the largest proportion of IL-10-producing B cells was observed in the CD38^int CD24^int subset, which corresponds to the mature naïve subset. We observed that all IL-10⁺ B cells produced in the presence of MSC were CD27⁻ (Figure 2H).

To test whether soluble factors or cell contact-dependent mechanisms are involved in the effects of MSC and MSC–IFN-γ on B cell proliferation and Breg induction, activated B cells were co-cultured with MSC and MSC–IFN-γ in a TW system to prevent cell–cell contact as shown in Figure 3A. By preventing direct cell–cell contact, the stimulatory effect of MSC on B cell proliferation was abolished (Figure 3B). Moreover, a small decrease in memory B cell proliferation was measured when B cells were co-cultured with MSC in a TW setting. Interestingly, prevention of direct cell–cell contact also abolished the anti-proliferative capacity of MSC–IFN-γ both in the total B cell population and in the naïve and memory B cell populations (Figure 3B). In accordance with the lack of proliferation inhibition in co-cultures of B cells with MSC–IFN-γ in a TW system, levels of IgG were not affected by MSC and MSC–IFN-γ (Figure 3C). No Bregs were induced when B cells were co-cultured with MSC and MSC–IFN-γ in TW system (Figure 3D), and, in correspondence, no increase in IL-10 levels was found (Figure 3E). These results indicate that the inhibition of B cell proliferation, inhibition of IgG production, and induction of IL-10 production by MSC is dependent on cell contact or at least close proximity of MSC and B cells.

To examine whether the inhibition of B cell proliferation by MSC–IFN-γ requires merely interaction via membrane proteins or requires metabolically active MSC–IFN-γ, activated B cells were co-cultured with heat-inactivated MSC (HI-MSC) (Figure 4A). HI-MSC are immunophenotypically intact but release no soluble factors, as previously described (24). Culturing B cells with HI-MSC abolishes the stimulatory effect of MSC on B cell proliferation (Figure 4B) and furthermore the proliferation of B cells was not significantly inhibited by HI-MSC–IFN-γ (Figure 4B). HI-MSC and HI-MSC–IFN-γ induced an increase in Bregs but this increase was not linked to an increase in IL-10 production (Figures 4C,D). These data indicate that the inhibition of B cell proliferation is dependent on metabolic activity of MSC–IFN-γ. Furthermore, the induction of Bregs cannot be recuperated by inactivating MSC–IFN-γ but requires metabolically active MSC. Activating MSC with IFN-γ appears to overrule the Breg inducing capacity of MSC.

We hypothesized that the inhibition of B cell proliferation by MSC is mediated by IFN-γ triggered IDO induction, leading to degradation and depletion of TRP. When MSC were cultured for 4 days with IFN-γ high levels of l-kynurenine, the breakdown product of TRP, were detected (Figure 5A). When 200 µM TRP was added to B cell and MSC–IFN-γ co-cultures to counteract the effect of IDO activity, B cell proliferation increased from 17 to 48% in the total B cell population, from 19 to 52% in the case of naïve B cell proliferation, and from 16 to 36% in the case of memory B cell proliferation (Figure 5B). TRP supplementation, furthermore, reversed the effect of MSC–IFN-γ on IgG production by B cells (Figure 5C).

We showed that IFN-γ-pretreated MSC were not able to induce an increase in Bregs. TRP supplementation to MSC–IFN-γ and B cell co-cultures showed a trend toward increased frequencies of CD38^high CD24^high Bregs (Figure 5D). In accordance with this, the levels of IL-10 in the supernatant of the MSC–IFN-γ cultures were significantly increased when B cell proliferation was rescued with TRP supplementation (Figure 5E). Stimulation of MSC and MSC–IFN-γ with T cell-like stimulation and TRP did not induce IL-10 secretion by MSC, eliminating the possibility that the IL-10 in the stimulated cultures is secreted by MSC–IFN-γ (data not shown). These data indicate that the incapability of MSC–IFN-γ to induce Bregs is caused by TRP depletion mediated by IFN-γ triggered IDO activity in MSC.