The most important feature of B cells is the production of Abs upon activation; additionally, B cells produce pro- and anti-inflammatory cytokines in response to certain stimuli. IL-10–producing B cells represent a major subset of regulatory B cells (Bregs) that suppress autoimmune and inflammatory responses. B cells play a crucial role in the development and maintenance of the chronic inflammatory autoimmune disease rheumatoid arthritis (RA); however, controversial data are available on IL-10– producing Bregs in RA. Our aim was to identify the optimal conditions that induce IL-10+ Bregs and, furthermore, to shed light on the signaling pathways that are responsible for their expansion. The results show that dual stimulation by CpG and CD40L for 48 h is optimal for IL-10 induction, and this can be synergistically boosted by IL-21. We identified the CD19+CD27+ memory B cell population as the major source of IL-10+ Bregs. We detected significantly fewer CD19+CD27+IL-10+ cells in RA patients compared with healthy controls, and these were functionally defective in suppressing IFN-γ production by CD4+ T cells in coculture. IL-21 drastically increased the number of IL-10+ Bregs within the CD19+CD27+ and CD19+CD27− populations; furthermore, it induced the appearance of IL-10+Blimp-1+ plasmablasts. Monitoring the phosphorylation of key signaling molecules revealed that activation of ERK, p38, and CREB is indispensable for the induction of IL-10 production, whereas phosphorylation of STAT3 further enhances IL-10 expression in human Bregs. We conclude that CREB and STAT3 are the key transcription factors responsible for the expansion and differentiation of human IL-10–producing Bregs.
B cell subsets that negatively modify the immune response, termed regulatory B cells (Bregs), play important roles in inflammation and autoimmunity in mice and humans (1–4). Several Breg subsets were described that produce IL-10 (1, 2, 5, 6) or TGF-β (7, 8) or express CD25 (9, 10), TIM-1 (11), Foxp3 (8), or FasL (12, 13). The best-characterized population in mouse and humans is IL-10–producing Bregs (B10 cells) (2, 5, 6, 14–16).
Human IL-10+ Bregs have no uniform phenotype; they can be detected in the CD24hiCD38hi (17), CD24hiCD27+ (4, 14), CD5+CD1dhi (12, 18–20), and CD27hiCD38hi (21, 22) populations. In unstimulated samples, the frequency of IL-10+ Bregs is extremely low; however, stimulation of IL-10–competent cells with TLR9 and CD40 efficiently induces IL-10 synthesis and secretion (2, 14, 23). The optimal stimuli to selectively induce IL-10, the major source of B10 cells, and the molecular signals inducing human B10 cells have not been clarified.
IL-10+ Bregs exert their immunosuppressive effect by suppressing Th1 and Th17 responses, impairing APC function and proinflammatory cytokine release by monocytes, influencing inflammatory cytokine production, and inducing differentiation of IL-10– and TGF-β–producing regulatory T cells (Tregs) (1, 15, 24). Bregs can also contribute to Treg induction by secreting TGF-β (7, 9). The multiple Breg subsets that are induced by different environments and regulate a wide range of functions point to their intricate role during the immune response in health and disease (25).
Transfer of Bregs into animals with experimental autoimmune diseases was shown to improve disease symptoms, indicating that Bregs are involved in controlling autoimmunity (22, 26, 27). The function of Bregs in rheumatic diseases, such as rheumatoid arthritis (RA), has been studied intensively (14, 24, 28, 29). RA is a chronic inflammatory autoimmune disease that primarily affects the small joints, eventually leading to bone erosion and an inability to move (30). Autoantibodies specific for citrullinated proteins in the synovium, which are present in ∼70% of RA patients, may deposit in the joints and induce the release of inflammatory cytokines (31). Recently, Daien et al. (29) found that the number of B10 cells was lower in RA patients compared with healthy controls; moreover, the function of these cells was impaired. The investigators concluded that the loss of B10 cells and Breg function may contribute to the pathogenesis of disease (29). However, other investigators found that the proportion of B10 cells increased in some RA patients and in other autoimmune diseases compared with healthy controls (14).
We aimed to define the optimal experimental conditions to induce IL-10 expression by human B cells and to identify the dominant subpopulation producing IL-10 in the peripheral blood of healthy volunteers and RA patients. The results showed that CpG and CD40L double stimulation efficiently induced the development of B10 cells with a suppressive effect on CD4+ T cell IFN-γ production, primarily within the CD19+CD27+ memory B cell population. The addition of IL-21 synergistically induced the expansion of B10 cells, as well as the differentiation of IL-10– and B lymphocyte maturation protein 1 (Blimp-1)–expressing double-positive plasmablasts. Compared with healthy controls, RA patients had significantly fewer B10 cells; however, IL-21 increased their number in the CD19+CD27+ memory B cell compartment in healthy and RA samples. Phosphorylation of ERK, p38 MAPK, and CREB appears to be indispensable for the induction of Bregs, whereas IL-21–activated STAT3 further augments IL-10 secretion and induces the expansion and differentiation of B10 cells.
Materials and Methods
Human blood samples
Heparinized blood samples were collected from healthy blood donors and from patients with RA after written consent with ethical permission of the Scientific Research Ethics Committee of the Medical Scientific Board of Ministry of Human Resources (Medical Research Council BPR/021/00818-5/2014). The study was conducted in accordance with the Declaration of Helsinki. Table I summarizes the characteristics of RA patients. Disease diagnosis was based on the revised classification criteria of the American College of Rheumatology/European League against Rheumatism. The patients were treated with conventional disease-modifying antirheumatic drugs, such as methotrexate and/or sulfasalazine, and they did not receive biological therapy before blood samples were taken.
|.||RA Patients (n = 29) .||Normal Range .|
|Age (y; mean [range])||53.24 (18–80)|
|Disease duration (y; mean [range])||9 (1–29)|
|CRP (mg/l; mean [range])||5.6 (0.6–8.4)||0–5|
|DAS28 (mean [range])||3.87 (2.04–7.4)|
|CCP−||n = 3|
|CCP+ (mean [range])||976.59 (47–3010)|
|RF−||n = 3|
|RF+ (mean [range])||193 (20–378)|
|.||RA Patients (n = 29) .||Normal Range .|
|Age (y; mean [range])||53.24 (18–80)|
|Disease duration (y; mean [range])||9 (1–29)|
|CRP (mg/l; mean [range])||5.6 (0.6–8.4)||0–5|
|DAS28 (mean [range])||3.87 (2.04–7.4)|
|CCP−||n = 3|
|CCP+ (mean [range])||976.59 (47–3010)|
|RF−||n = 3|
|RF+ (mean [range])||193 (20–378)|
CCP, cyclic citrullinated peptide; CRP, C-reactive protein; RF, rheumatoid factor.
Separation and stimulation of cells
PBMCs were isolated by density gradient centrifugation on Ficoll-Paque PLUS (GE Healthcare). B lymphocytes were purified by negative selection using MACS, according to the manufacturer’s protocol (Miltenyi Biotec). B cells with purity > 98% were used.
Cells were stimulated for 48 h with the affinity-purified F(ab′)2 fragment of goat anti-human IgG+IgM (H+L) (10 μg/ml; Jackson ImmunoResearch Laboratories), recombinant human CD40L (1 μg/ml; ImmunoTools), recombinant human IL-21 (50 ng/ml; ImmunoTools), and phosphorothioated unmethylated CpG oligodeoxynucleotide (ODN-2006) (5′-TCGTCGTTTTGTCGTTTTGTCGT-3′) (Sigma-Aldrich) at two concentrations (2.5 μg/ml for intracellular cytokine detection, FlowCytomix assay, and ELISPOT assay and 5 μg/ml for phospho-kinase array, phospho-flow assay, and for kinase-inhibitor experiments).
Analysis of IL-10 production
A total of 2 × 105 purified B cells or 2 × 106 PBMCs was cultured in medium (RPMI 1640 with 10% FCS; Sigma-Aldrich) for 48 h; PMA (50 ng/ml; Sigma-Aldrich), ionomycin (1 μg/ml; Sigma-Aldrich), and brefeldin A (1 μl/ml; BD Biosciences) were added for the last 5 h. For intracellular protein detection, Fc receptors were blocked using 10% mouse serum, and dead cells were excluded by staining with 7-aminoactinomycin D (Life Technologies). Cells were labeled with anti-CD19–PE and anti-CD27–FITC (ImmunoTools), fixed and permeabilized using a Cytofix/Cytoperm Kit (BD Biosciences), according to the manufacturer’s instructions, and stained with anti-IL-10–allophycocyanin and anti-TNF–Alexa Fluor 488 or anti-IL-6–FITC (BD Biosciences) or anti-Blimp-1–Alexa Fluor 488 conjugate (R&D Systems) and the appropriate isotype controls: allophycocyanin Rat IgG2a κ Isotype Control, Alexa Fluor 488 Mouse IgG1 κ Isotype Control, FITC Rat IgG2a κ Isotype Control (all from BD Biosciences), and Mouse IgG1 Alexa Fluor 488–conjugated Ab (R&D Systems). The intracellular expression of cytokines and Blimp-1 was analyzed by flow cytometry (FACSCalibur; BD Biosciences). The separated B cells were labeled with anti-CD38–FITC (ImmunoTools), anti-CD24–PE (BD Biosciences), anti-CD27–PerCp–Cy5.5 (BioLegend), and anti-IL-10–allophycocyanin (BD Biosciences) or with the appropriate isotype controls: Mouse IgG1 Isotype control FITC-conjugated (ImmunoTools), PE Mouse IgG2a, κ Isotype Control (BD Biosciences), PerCp/Cy5.5 Mouse IgG1, κ Isotype Ctrl Ab (BioLegend), and allophycocyanin Rat IgG2a κ Isotype Control (BD Biosciences) to further analyze the phenotype of IL-10+ B cells.
Secretion of IL-6, TNF, IL-10, IFN-γ, and IL-17 was measured from the supernatants of 2 × 105 purified B cells cultured under various conditions. The supernatants were collected after 43 h (before the addition of PMA, ionomycin, and brefeldin A [PIB]) and stored at −70°C until being assayed with a FlowCytomix bead array (Bender MedSystems), according to the manufacturer’s instructions, or by ELISA.
PBMCs were labeled with anti-CD4–allophycocyanin and anti-CD19–PE (ImmunoTools), and CD4+ Th cells and CD19+ B cells were sorted with a BD FACSAria III. T cells were stimulated with plate-bound anti-CD3 (10 μg/ml) and anti-CD28 (1 μg/ml) (ImmunoTools) for 48 h. B cells were cultured with CpG (2.5 μg/ml) and CD40L (1 μg/ml) in the presence or absence of IL-21 (50 ng/ml) for 48 h. Unstimulated or stimulated B cells were mixed with the prestimulated T cells at a 1:1 ratio for an additional 24 h; cells were treated with PIB for the last 5 h. T cells were intracellularly stained with anti-IFN-γ–Alexa Fluor 488 conjugates (BD Biosciences). Relative inhibition of IFN-γ production was calculated by comparing the numbers of IFN-γ+CD4+ T cells in cocultures with unstimulated and stimulated B cells.
Cells were cultured for 4 d with the appropriate stimuli, and the frequencies of IgG- and IgM- secreting cells were evaluated by ELISPOT assay. Briefly, 96-well nitrocellulose plates were coated for 2 h at room temperature with 15 μg/ml mouse anti-human IgG or 10 μg/ml mouse anti-human IgM (both from Mabtech). Plates were washed with sterile PBS and blocked with 10% FCS RPMI medium. A total of 105 stimulated PBMC was added to each well and incubated at 37°C. Cells were aspirated 20 h later, and plates were washed with PBS. Biotin-conjugated mouse anti-human IgG or IgM Ab (1 μg/ml; Mabtech) was added for 2 h. Streptavidin-HRP conjugate (1:1000; Mabtech) was added after washing, and the plates were incubated at room temperature. Spots were developed with TMB (Sigma-Aldrich) solution and counted using a CTL ImmunoSpot Reader (Cellular Technology).
We performed a phospho-kinase array (R&D Systems) for the initial screening of the phosphorylation status of a variety of signaling molecules. A total of 106 B cells was treated with the appropriate stimuli for 30 min at 37°C in serum-free medium, and the experiments were carried out according to the manufacturer’s instructions.
A total of 106 PBMCs per sample were stimulated with CpG and CD40L, in the presence or absence of IL-21, at 37°C for the indicated times. After treatment, cells were fixed with 2% paraformaldehyde for 10 min at 37°C and washed twice with PBS. The fixed PBMCs were stained with anti-CD19–PE and then permeabilized on ice for 30 min using 500 μl of 90% MeOH. Cells were washed and stained with Alexa Fluor 488–conjugated phospho-p38– or phospho-ERK–specific Abs (BD Biosciences) or anti-phospho–STAT3–Alexa Fluor 488 conjugate (Cell Signaling Technology). The relative expression (%) was calculated based on the change in phosphorylated protein expression in resting and activated cells (mean fluorescence intensity [MFI] of stimulated sample/MFI of resting sample).
Purified B cells were activated in the presence of the specific inhibitors of CREB (25 μM, KG-501; Sigma-Aldrich), p38 (10 μM, SB203580; Cell Signaling Technology), MEK (20 μM, PD98059; Sigma-Aldrich), PI3K (25 μM, LY294002; Sigma-Aldrich), and ERK (10 μM, UO126; Promega). Supernatants were collected 48 h later, and IL-10 concentrations were measured by IL-10 ELISA (ImmunoTools).
All data are shown as mean ± SEM. Significant differences between samples were determined using the Student t test. The p values <0.05 were considered significant.
Optimization of stimuli inducing human B10 cells
B cells negatively selected from the blood of healthy donors were stimulated with anti-IgG+IgM F(ab)2, CpG, CD40L, or with combinations of these for 48 h. Cells were treated with PIB for the last 5 h to stop cytokine secretion, and intracellular IL-10+ cells were detected by flow cytometry. Without stimulation, a negligible amount (<0.1%) of B10 cells was detected, whereas stimulation with BCR, CD40, or their combination increased IL-10 positivity. In contrast, B cells stimulated with CpG contained a significantly higher percentage of B10 cells, and this was increased further by CD40L (5–7%) (Fig. 1A, 1B). Interestingly, the inclusion of BCR-mediated stimuli reduced the percentage of B10 cells, confirming earlier results (14, 29). Based on this, we used the combination of CpG and CD40L to induce human B10 cells in subsequent experiments.
Next, we used PBMCs to study the B cell subpopulations affected by CpG plus CD40L double stimulation. CD19 and CD27 markers were used to differentiate between CD19+CD27− naive and CD19+CD27+ memory B cells (32). CD19 is expressed on B cells from the pre-B stage to the plasmablast stage (33). Without stimulation, CD19+CD27− naive B cells were negative for IL-10, and the CD27+ subset showed very low expression of IL-10 (<1%). CpG+CD40L stimulation induced IL-10 in <3% of CD27− B cells, whereas close to 10% of the C19+CD27+ subset was IL-10+ (Fig. 1C).
B10 cells appear to be important in controlling autoimmune and inflammatory diseases (6, 16, 18, 34); therefore, we compared the levels of B10 cells in the blood of healthy donors and RA patients(Table 1). Unstimulated control cells did not exhibit any differences between healthy and RA samples in the CD27− or CD27+ subset, and we detected an increased percentage of B10 cells within CD27+ subsets. In contrast, in the CpG+CD40L-stimulated samples, significantly fewer B10 cells were detected within the CD27+ B cell population of RA patients compared with healthy controls, whereas the proportions of B10 cells within the CD27− subsets were similar (Fig. 1D). These data indicate that the induction of B10 cells is impaired in RA.
By further monitoring of the surface phenotype of IL-10–producing B cells, we found that the majority of IL-10+CD27+ cells were CD24+CD38− in control and CpG+CD40L-stimulated samples, indicating that CD24+CD27+ memory B cells are the major source of B10 cells in humans (Fig. 1E).
Production of proinflammatory cytokines by stimulated B cells of healthy individuals and RA patients
B cells may secrete proinflammatory cytokines in response to various stimuli (35, 36). Because most RA patients have high serum levels of inflammatory cytokines, especially TNF, which is produced primarily by inflammatory T cells, we examined whether B cells exposed to CpG and CD40L are able to produce TNF or other proinflammatory cytokines. We detected IL-10 and TNF by double fluorescent labeling. The results of a representative experiment (Fig. 2A) clearly show that CpG and CD40L induce ∼15% of B cells to produce TNF. The majority of these cells are single positive; only ∼1% are double positive, producing IL-10 and TNF. Thus, the same stimuli may induce both IL-10 and TNF production, but typically in different B cell subsets. Fig. 2B shows the frequency of CD27− naive and CD27+ memory B cells within the IL-10+TNF+ double-positive population in CpG+CD40L-stimulated samples from healthy individuals. IL-10+TNF+ double-positive cells were detected primarily in the CD27− population.
TNF and IL-10 production was also compared in B cells from healthy donors and RA patients stimulated by various means. The most remarkable differences were that the level of TNF-producing B cells and the ratio of TNF/IL-10–producing B cells were significantly higher in RA patients compared with healthy controls. The percentage of IL-10 and TNF double-positive cells within the IL-10+ population was also increased in RA (Fig. 2C).
Secreted cytokines were measured in the supernatant of B cells from healthy individuals treated with different stimuli for 48 h. Stimulation with BCR induced a high level of IFN-γ and IL-17, whereas the combined BCR and TLR9 signals synergistically increased IL-6 and TNF secretion. In the CpG+CD40L-stimulated samples, we detected a significantly higher level of secreted IL-10 compared with most proinflammatory cytokines (with the exception of IL-6) (Fig. 3A).
Simultaneous intracellular staining for IL-6 and IL-10 showed that CpG and the combined stimuli induced IL-6 production in 30–60% of B cells; most of the cells were single positive, whereas most IL-10+ cells also produced IL-6 (Fig. 3B).
IL-21 synergistically expands the B10 cell population in healthy individuals and RA patients
IL-21 induces human B cell differentiation into Ig-secreting plasma cells (37–39). IL-21 also was shown to be required for the maturation of B10 cells into IL-10–secreting effector cells that inhibit in vivo autoimmune disease in mice (23). However, the impact of IL-21 on human B10 cells had not been investigated. Therefore, we stimulated human B cells from healthy donors and RA patients with CpG, CD40L, and IL-21, alone and in combination, and assessed the percentage and number of B10 cells 48 h later. IL-21 alone had very little effect; however, in combination with CpG and CD40L, it synergistically increased IL-10 expression in the CD27+ and CD27− B cell subsets from healthy and diseased individuals (Fig. 4A). In the presence of IL-21, the number of B10 cells was ∼4–5-fold higher in the CD19+CD27+ memory B cell population of healthy blood donors, whereas the number of IL-10–producing CD27+ cells increased 10-fold in samples from RA patients, resulting in an equally high number of B10 cells in healthy and RA samples (Fig. 4B).
To test whether B10 cells and the IL-21–expanded B10 population have a suppressive function, we performed coculture experiments with autologous CD4+ T cells and assessed IFN-γ production in T cells by intracellular staining. Comparing the relative inhibitory effects of stimulated B cells showed that CpG+CD40L-stimulated B10 cells from RA patients had a significantly lower suppressive effect compared with those from healthy subjects, whereas the addition of IL-21 to the cell cultures significantly increased the inhibitory effect on IFN-γ production by T cells in healthy and RA subjects, correlating with the number of B10 cells detected in the same samples (Fig. 4C).
IL-21 synergistically stimulates IL-10 and Ab production by regulatory plasmablasts/plasma cells
Next, we tested Ab secretion by B cells treated with CpG, CD40L, and IL-21 after 5 d of culture. IgG and IgM Ab-forming cells were detected at equally high numbers in the presence of IL-21 (Fig. 5A). A high number of B10 cells was also detected in the same samples (Fig. 5C), indicating that the emerging B10 cells and IL-10 production had no effect on IgG/IgM secretion.
The final differentiation of B lymphocytes into Ab-secreting plasma cells depends on the expression of Blimp-1 (40). Next, we tested how B cell differentiation into plasma cells influences IL-10 production. B cells stimulated by CpG+CD40L in the presence of IL-21 were intracellularly labeled for IL-10 and Blimp-1 after 2 and 5 d of culture. In triple-stimulated B cells, significant Blimp-1 expression was already observed after 2 d (Fig. 5B), and it increased further by day 5 (Fig. 5C). Approximately one quarter of Blimp-1+ cells expressed intracellular IL-10; IL-10 single-positive cells also were detected. IL-21 significantly increased the frequency of single- and double-positive cells after 2 d compared with the CpG+CD40L-stimulated samples (Fig. 5C). The number of Blimp-1+ cells increased with the time of stimulation, indicating the progression of differentiation. A higher level of IL-10 secretion was also observed in the same triple-stimulated samples (Fig. 6B).
Activation of signaling pathways leading to the synergistic expression of IL-10
To assess the most important signaling molecules regulating IL-10 expression, we first analyzed the activation of purified B cells by a phospho-kinase protein array. Of the 43 spots on the membrane, only a few provided signals in response to CpG+CD40L stimulation; the phosphorylation levels of the transcription factors CREB, p38, ERK, and p27 were increased considerably in the stimulated samples (Fig. 6A). To validate the results of the phospho-kinase array, secretion of IL-10 in response to the same stimuli was compared in the presence and absence of various kinase inhibitors. Specific inhibitors of CREB, p38, ERK, and PI3K significantly inhibited IL-10 secretion in all samples, whereas the MEK inhibitor had no significant effect (Fig. 6B).
Furthermore, we confirmed and expanded our results by monitoring phosphorylation of key signaling molecules using the phospho-flow method. PBMCs were used in these experiments; the gating strategy and a representative experiment are shown in Fig. 6C. Phosphorylation of p38, ERK, and the IL-21 target STAT3 was followed within CD19+ B cells. CpG+CD40L stimulation significantly enhanced p38 and ERK phosphorylation, whereas it had no effect on the phosphorylation of STAT3. In contrast, IL-21+CpG+CD40L induced STAT3 phosphorylation but did not modify the activity of p38 and ERK (Fig. 6D). IL-21 alone did not activate STAT3 (data not shown) and did not induce IL-10 in B cells.
Activation of signaling molecules was compared in healthy donors and RA patients. We observed a lower level of activation-induced phosphorylation of all molecules tested in RA patients compared with healthy controls. The phosphorylation level of p38 and Erk was significantly lower in RA patients in the double- and triple-stimulated samples, whereas the difference in STAT3 phosphorylation was not significant (Fig. 6E).
B cells as effectors of the humoral immune response secrete Abs that induce Ag clearance; depending on the stimuli, they may also produce pro- and anti-inflammatory cytokines, thus having roles in initiating and suppressing inflammation and autoimmunity. B10 cells can be functionally characterized by their capacity to produce IL-10. They represent a very small subset (<1%) of human peripheral blood B cells that expands after certain stimuli. In this study, we optimized the conditions to define the main source of B10 cells in human blood and compared the function of these cells in healthy individuals and RA patients.
Data from the literature show some discrepancies with regard to the signals that induce human B10 cells; BCR, CD40, and TLR–mediated stimuli were applied at various doses, for various lengths of time (14, 15, 29, 41). We confirm that dual stimulation with CpG and CD40L for 48 h is optimal for the induction of human IL-10–producing B cells. This combination of stimuli typically induced IL-10 production in ∼10% of CD27+ memory B cells and at a much lower level (typically 1–3%) in CD19+CD27− naive B cells. Detailed phenotypic analyses showed that the majority of IL-10–producing CD27+ cells are CD24+ but CD38− (40–60%), and only a small proportion (10–20%) expresses CD38. We also confirmed that BCR-mediated signals decreased the level of TLR9- and CD40-induced B10 cells. Additionally, we showed that mimicking Ag stimulation through BCR and CD40 induced IL-10 in very few B cells. This suggests that B10 cells might evolve in vivo when the BCR signal is low (e.g., upon autoimmune response, when autoreactive BCR bind self-antigens with low affinity).
Bregs were shown to regulate autoimmune response in various mice models (20, 26, 42). Very little data are available on the function of human B10 cells in autoimmune diseases (3). The B10 population was shown to be increased in some patients with systemic or organ-specific autoimmune diseases (14), whereas other investigators found that RA patients with active disease have a reduced number of CD24hiCD38hi Bregs compared with healthy controls (24). Also, Bregs are functionally impaired in systemic lupus erythematosus (17). Recently, a detailed analysis showed that the level of CD24hiCD27+ B10 cells is lower in RA patients compared with healthy controls, especially during the early phase of the disease, and these cells have an impaired regulatory function (29). In concert with this finding, we found a significantly lower number of CD19+CD27+IL-10+ B cells in blood samples from RA patients with long-standing (>1 y) moderately active/active disease compared with healthy controls. Furthermore, we also showed that B10 cells in RA patients have a significantly lower ability to inhibit IFN-γ production of autologous CD4+ Th1 cells compared with healthy subjects, pointing to the role that B10 cells might play in RA.
However, addition of IL-21 significantly increased the number and the suppressive capacity of B10 cells in healthy and RA samples.
Stimulation of B cells also may result in the production of an array of proinflammatory cytokines (35, 36). B cells stimulated via BCR and CD40, a signal combination that mimics T-dependent Ag stimulation, proliferate and secrete TNF, IL-6, and other cytokines that may amplify the ongoing immune response (36). In contrast, CD40 stimulation in combination with CpG mimics the inflammatory conditions induced by bacterial DNA or DNA released from damaged cells that, in the presence of bystander T cell help (CD40L signal), would induce the production of IL-10 to suppress inappropriate immune responses. Such a mechanism may protect the organism against autoimmune diseases.
We found that the same CpG+CD40L stimuli also induced the production of TNF; however, IL-10 and TNF were produced mostly by different subsets of B cells, and very few (<1%) double-positive cells were detected. Comparing different stimuli, the combination of CpG+CD40L induced the highest level of IL-10–producing B cells and the lowest level of TNF-producing B cells in samples from healthy blood donors, shifting the net outcome toward suppression. However, a significantly higher TNF production was detected in B cell samples from RA patients in response to all stimuli, including CpG and CD40L, and the percentage of IL-10 and TNF double-positive cells was also increased, which might contribute to the impaired regulatory function of B10 cells in RA patients.
Measuring the secreted cytokines from the supernatants of B cells from healthy individuals stimulated in different ways confirmed that the proinflammatory cytokines IL-6, TNF, IL-17, and IFN-γ were induced mainly by combined signals involving BCR. However, CpG+CD40L dual stimulation induced significantly higher IL-10 secretion than proinflammatory cytokine secretion, with the exception of IL-6, which was at about the same level as IL-10. Assessment of IL-10 and IL-6 double-positive cells in healthy individuals clearly shows that almost all IL-10–producing B cells also produce IL-6 in response to CpG+CD40L stimuli. Thus, because IL-6 inhibits Th1 polarization and, thus, IFN-γ production while promoting Th2 differentiation (43) and, furthermore, upregulates IL-21 in naive CD4+ T cells (44), we suppose that, altogether, these features may contribute to the suppressive function of human B10 cells.
IL-21 plays a central role in the differentiation of plasma cells during T-dependent B cell responses (37, 39). CD40 and IL-21 receptor–mediated signals dramatically enhance B10 cell development in mice and generate B10 cells that inhibit disease symptoms when transferred into mice with autoimmune disease, raising the possibility that IL-21 might have therapeutic potential (23, 45). However, the effect of IL-21 on human B10 cell generation and function had not been studied. We show for the first time, to our knowledge, that CpG+CD40L- mediated stimuli in the presence of IL-21 induce a significantly higher number of B10 cells in the CD19+CD27+, as well as the CD19+CD27−, population. Additionally, we could not detect any difference between the number and suppressive function of B10 cells in the CD27+ B cell subset of healthy and RA samples in the presence of IL-21; both inhibited IFN-γ production of CD4+ T cells.
IL-21 is required for the differentiation of plasma cells and Bregs (37–39). It induces the expression of Blimp-1, a key mediator of B cell differentiation into plasma cells (37). Monitoring Blimp-1 and IL-10 expression in B cells stimulated with CpG+CD40L in the presence of IL-21 showed that a significantly higher percentage of Blimp-1+ cells had already developed after 2 d, and some cells were double positive. After 5 d, the percentage of Blimp-1+ cells increased further, the percentage of IL-10–expressing cells decreased, and the percentage of double-positive cells remained the same. These data indicate that, upon activation in the presence of IL-21, expansion of IL-10–producing B10 cells eventually results in the differentiation into Blimp-1+ plasma cells. The IL-10/Blimp-1 double-positive cells may represent IL-10–secreting regulatory plasmablasts/plasma cells (21, 46). In agreement with previous results in mice, our data also show that human B10 cells, while inhibiting IFN-γ production of CD4+ T cells, have no effect on IgG and IgM production (21, 47), indicating that the main target of Bregs/plasmablasts is the inflammatory, and not the Ab, response.
Signaling pathways resulting in the induction of IL-10 in human B cells are not completely clarified; ERK, p38 MAPK, and STAT3 were shown to play important roles (41, 48–50). Thus, we monitored the phosphorylation of key molecules in response to stimulation via TLR9 and CD40. Activation of TLR9 results in the activation of ERK, p38, and NF-κB pathways and the expression of IL-10 in macrophages (51). Additionally, it was reported that, in mouse splenic B cells, CD40-induced activation of the PI3K–AKT pathway controls downregulation of p27, which is an inhibitor of cyclin-dependent kinases and cell proliferation (52). In this article, we show that, in addition to p38 and ERK kinases activating the transcription factor CREB (51), p27 is phosphorylated in the double-stimulated samples. Phosphorylation of p27 may trigger its downregulation, resulting in the increase in the size of the B cell population producing IL-10.
Studying the effect of specific inhibitors and checking the phosphorylation level by phospho-flow analysis confirmed that, similarly to dendritic cells (53) and macrophages (51), activation of p38, ERK, and CREB is indispensable for IL-10 expression in human B cells. IL-10 production was not significantly decreased by MEK inhibitor, because it inhibits only MEK1 (and not MEK2) at the concentration used.
STAT3 plays a crucial role in generating effector B cells from naive precursors in humans (48). In contrast to a previous study (50), CpG+CD40L-mediated dual signals phosphorylated CREB but did not activate STAT3, probably because of the doses and timing that we applied. Time-dependent STAT3 phosphorylation was observed only after triple stimulation: CpG+CD40L in the presence of IL-21.
Activation of the key signaling molecules was also monitored in B cells from RA patients. It revealed significantly lower activation-induced phosphorylation levels of p38 and Erk compared with healthy controls, which corresponded to the lower level of B10 cells in patients; however, the difference in the phosphorylation level of STAT3 was not significant.
The synergistic effect of CpG, CD40L, and IL-21–mediated signals on the induction of IL-10 transcription might be explained by the collaboration between CREB and STAT3. We suppose that CREB binding protein (p300), which is a coactivator of STAT3 and stabilizes its nucleic expression and enhances its transcriptional activity (54), might play a role in the synergism. In addition, STAT3 is also involved in IL-10R signaling, suggesting a positive-feedback mechanism on IL-10 transcription.
Taken altogether, we conclude that the TLR9, CD40, and IL-21–mediated combined stimuli are highly efficient at inducing human B10 cells, which primarily originate from the CD19+CD27+ subset. Although we found a reduced number of B10 cells in the CpG+CD40L-stimulated samples from RA patients, we did not detect a significant difference between B10 cells in the CD19+CD27+ subsets of RA patients and healthy subjects when IL-21 was also added. TLR9- and CD40-activated CREB and IL-21–activated STAT3 play a synergistic role in signaling pathways that induce B10 cells. IL-21 expanded the B10 population and induced the differentiation of CD27+ B10 cells into Blimp-1+ regulatory plasmablasts, which may further differentiate into Ig-secreting plasma cells.
This work was supported by the National Research Development and Innovation Office (OTKA NK 104846 and OTKA-NN111023 [to G.N.]). The European Union and the European Social Fund provided financial support to the project under Grant Agreement TAMOP 4.2.1./B-09/1/KMR-2010-0003.
Abbreviations used in this article:
- B10 cell
B lymphocyte maturation protein 1
regulatory B cell
mean fluorescence intensity
PMA, ionomycin, and brefeldin A
regulatory T cell.
The authors have no financial conflicts of interest.