We examined B10 cells in the peripheral blood of patients with active CD (N = 30; 16 men; 14 women; mean age: 32.5 ± 3.0 years; range: 14–52 years). Samples from healthy individuals (N = 50; 25 men; 25 women; mean age: 32.1 ± 2.5 years; range: 18–56 years) served as controls. Active CD was prospectively defined as a Crohn’s disease activity index (CDAI) >150. All the patients provided written informed consent to participate in the study, which was carried out in accordance with the principles of the Declaration of Helsinki and was approved by the ethics committee of Xinhua Hospital.

For intracellular cytokine staining, cells were stimulated with a cell stimulation cocktail plus protein transport inhibitors (eBioscience) for 5 h. Then, the cells were fixed and permeabilized with Cytofix/Cytoperm buffer, and intracellular cytokines were stained with antibodies against IL-10, IFNγ, and TNFα (eBioscience). Flow cytometric analysis was performed with a FACS Canto II instrument (BD Bioscience) and FlowJo software (TreeStar).

Interleukin-10 and TNFα levels were measured with ELISA kits according to the manufacturer’s instructions (R&D Systems).

Cells were directly lysed and subjected to 10% SDS-PAGE. Immunoblotting was performed after transferring the proteins onto nitrocellulose membranes (Schleicher & Schuell Microscience) with a Mini Trans-Blot apparatus (Bio-Rad). After 2 h of blocking, the membranes were incubated overnight at 4°C with the following specific primary antibodies: anti-Jarid2 (Abcam), anti-p-STAT3, anti-STAT3 (Cell Signaling), and anti-β-actin Abs (Sigma-Aldrich). After the membranes were washed, subsequent incubations with the appropriate HRP-conjugated secondary antibodies were conducted for 1 h at room temperature; after extensive washing, the signals were visualized with an ECL substrate (Pierce Chemical).

Expression levels of miR-155, IL-10, TNFα, and Jarid2 were measured by quantitative RT-PCR (qRT-PCR) using a SYBR green-based real-time quantitative PCR assay (Qiagen). The data were collected and quantitatively analyzed using an ABI Prism 7900 Sequence Detection System (Applied Biosystems). The relative level of miR-155 was normalized to that of small nuclear RNA U6, which is a ubiquitously expressed small nuclear RNA that has been widely used as an internal control. Primers for miR-155 and the small nuclear RNA U6 were obtained from Invitrogen. The sequences of the primers are as follows: miR-155, forward 5′-TTA ATG CTA ATC GTG ATA GGG G-3′ and reverse 5′-CGA ATT CTA GAG CTC GAG GCA GG-3′; U6, forward 5′-CTC GCT TCG GCA CA-3′ and reverse 5′-CGA ATT CTA GAG CTC GAG GCA GG-3′; IL-10, forward 5′-CTT CGA GAT CTC CGA GAT GCC TTC-3′ and reverse 5′-ATT CTT CAC CTG CTC CAC GGC CTT-3′; Jarid2, forward 5′-GCT TCC CAC CAG GAT GAC AG-3′ and reverse 5′-CCA AGG AGC CCA TTC ACA GT-3′; TNFα, forward 5′-CAC CAC TTC GAA ACC TGG GA-3′ and reverse 5′-AGG AAG GCC TAA GGT CCA CT-3′; and GAPDH, forward 5′-GCC ACC CAG AAG ACT GTG GAT GGC-3′ and reverse 5′-CAT GTA GGC CAT GAG GTC CA C CAC-3′. The data are presented according to the following equations: target miRNA expression = 2ΔCt, with ΔCt = (U6 Ct-target miRNA Ct), or target mRNA expression = 2ΔCt, with ΔCt = (GAPDH Ct-target mRNA Ct).

We added CFSE (Sigma-Aldrich) to B cell suspensions at a final concentration of 2 µM for 8 min at 37°C; then, we washed the cells three times with PBS and resuspended them in complete RPMI medium.

Cocultures of 2.0 × 10⁵ B cells and 2.0 × 10⁵ CD4⁺ CD25⁻ T cells stimulated for 72 h with plate-bound CD3 mAb (0.5 µg/mL) were activated with stimulation cocktail plus protein transport inhibitors (eBioscience) for the last 5 h. In addition, 2.0 × 10⁵ B cells and 2 × 10⁵ CD14⁺ monocytes cocultured for 24 h were stimulated with LPS (100 µg/mL) for the final 5 h. The cells were stained for surface markers, permeabilized, stained intracellularly for IFNγ or TNFα and Foxp3 and analyzed by flow cytometry.

Isolated B cells (3 × 10⁶ B cells, Miltenyi Biotec) in 100 µL of Amaxa mix were electroporated with 300 nM miR-155 mimic or inhibitor or control (Sigma) according to the manufacturer’s instructions. Six hours after transfection, we added 100 nM CpG oligonucleotides to the culture. Then, we harvested the cells for further analysis.

We performed RNA interference experiments using electroporation (Amaxa) according to the manufacturer’s protocol. Briefly, we mixed 300 nM Jarid2-specific siRNA or control siRNA (Sigma) with 3 × 10⁶ B cells in 100 µL of Amaxa mix and transfected the cells via electroporation according to the manufacturer’s instructions. Six hours after transfection, we added 100 nM CpG oligonucleotides to the culture. Then, we harvested the cells for further analysis.

We followed a previously described protocol to isolate cytokine-producing B cells (10, 29). First, B cells were pre-enriched via depletion of non-B cells (Miltenyi Biotec) and cultured for 2 days under stimulation with 100 nM CpG oligonucleotides. Second, stimulation cocktail (eBioscience) was added to the cultures for 3 h to induce IL-10 secretion. Third, the viable cytokine-producing B cells were specifically isolated using a cytokine secretion assay according to the manufacturer’s instructions. Briefly, the pre-enriched B cells were incubated with IL-10 and TNFα catch reagents. The cells were subsequently labeled with IL-10 detection antibody or TNFα detection antibody conjugated to PE or APC, respectively. The IL-10- and TNFα-secreting cells were then sorted by FACS (BD Aria II). The purity of the cells was further confirmed by measuring the expression of IL-10 and TNFα by q-RT-PCR.

ChIP assays were performed using a ChIP assay kit (Millipore) with modifications. Isolated B cells (5 × 10⁶ cells) were fixed in 1% formaldehyde, and the chromatin was sonicated and pre-cleared by incubation with Protein A/G agarose/salmon sperm DNA (Millipore). The precleared chromatin was immunoprecipitated with antibodies against H3K27me3 (Abcam) overnight at 4°C or mouse IgG monoclonal antibody followed by incubation with Protein A/G agarose/salmon sperm DNA for 1 h. The immunoprecipitates were denatured, and the DNA was purified. The amount of immunoprecipitated DNA was quantified by real-time PCR using SYBR Green and the ABI PRISM 7500 Sequence Detection System (Applied Biosystems). The primers used for the PCR analysis of the IL-10 promoter locus are as follows: forward: 5′ to 3′, CCA GGT AGA GCA ACA CTC; reverse: 5′ to 3′ CAG GCT CCT TTA CCC CGA TT.

We used one-way analysis of variance to initially determine whether an overall statistically significant difference existed before using Tukey’s post hoc test and the two-tailed paired or unpaired Student’s t-test to analyze the differences between two groups. The results are presented as the means ± SEM. A value of P < 0.05 was considered statistically significant.

First, we used two markers, CD24 and CD27, to determine levels of IL-10 production in different B cell subsets (10). The gating strategies and isotype control are shown in Figure S1 in Supplementary Material. As reported by other studies, we found that the CD24^hiCD27⁺ B cell subset produced higher levels of IL-10 after stimulation with CpG oligonucleotides for 48 h compared to those produced by other B cell subsets (Figure 1A). In addition, these data were supported by the TLR9 expression measured in different B cells subsets (Figure 1B); the data show that the CD24^hiCD27⁺ B cells, which are more sensitive to stimulation by CpG oligonucleotides, expressed higher levels of TLR9 than other B cell subsets. Isolated human CD19⁺ B cells and were stimulated with CPG or not for 48 h, the miRNA profiles of the cells were examined, and we found that miR-155 was significantly increased after CPG stimulation (Figure S2 in Supplementary Material). Next, we isolated different B cell subsets according to the CD24 and CD27 markers and analyzed the expression of IL-10 and miR-155 using Q-PCR. We observed that the expression of miR-155 and IL-10 increased over time with the stimulation of CD24^hiCD27⁺B cells by CpG oligonucleotides in culture; the expression reached its highest level at 24 h before gradually decreasing (Figure 1C). When cultured for 24 h, the CD24^hiCD27⁺ B cells expressed higher levels of IL-10 and miR-155 than other B cell subsets (Figures 1D,E). In addition, IL-10 expression levels were significantly correlated with those of miR-155 in the CD24^hiCD27⁺ B cell subset (R = 0.775, P < 0.01) (Figure 1F). These data suggest that the CD24^hiCD27⁺ B cell subset expresses high levels of miR-155, which are correlated with IL-10 production.

Next, we explored the effect of miR-155 on the regulation of IL-10 production. We isolated CD19⁺ B cells from healthy controls (HCs) and overexpressed miR-155 mimic or control using electroporation and stimulated the cells with CpG oligonucleotides for an additional 48 h. The survival of B cells was not significantly changed after transfection with the miR-155 mimic or control (Figure S2A in Supplementary Material). We found that the miR-155 expression level was significantly increased in the mimic-transfected group compared to that in the control group (18.33 ± 2.23 and 5.18 ± 0.77%, respectively), as well as the IL-10 production level (Figures 2A–C). As STAT3 phosphorylation is important for IL-10 production (11), we measured the phosphorylation of STAT3 in B cells overexpressing miR-155. By Western blot, we observed that after overexpression of miR-155 mimic in B cells, STAT3 phosphorylation and IL-10 expression were significantly increased (Figure 2D). We also conducted functional assays and found that when B cells had overexpression of miR-155 and were cocultured with purified CD4⁺T cells, INFγ production by T cells was significantly reduced (Figure 2E). In contrast, when we sorted CD24^hiCD27⁺ B cells from the HCs and electroporated them with miR-155 inhibitor, the survival of B cells was not significantly changed after transfection with the miR-155 inhibitor or control (Figure S2B in Supplementary Material). We found that the expression level of miR-155 was decreased (14.11 ± 1.050 and 2.318 ± 0.6336%, respectively) and that the production level of IL-10 was significantly reduced compared to the corresponding levels in the control group (Figures 2F–H). Also, the inhibition of INFγ production by T cells was significantly decreased when CD24^hiCD27⁺ B cells were transfected with miR-155 inhibitor (Figure 2I). These data demonstrate that miR-155 regulates IL-10 production in B cells, as well as B cell function.

Furthermore, we wanted to explore the molecular mechanisms by which miR-155 regulates IL-10 expression in B cells. It has been reported that DNA-binding protein Jarid2 recruits polycomb repressive complex 2 (PRC2) to chromatin and inhibits gene expression in miR-155-deficient cells (28); however, this molecular mechanism has not been examined in B cells. Therefore, we speculated that miR-155 can activate IL-10 production by regulating Jarid2 and relieving the H3K27 histone methylation repression of IL-10. First, we sorted the B cell subsets according to their expression of CD24 and CD27 using FACS and then measured Jarid2 gene expression by Q-PCR; we found that CD24^hiCD27⁺ B cells express much lower levels of Jarid2 than other B cell subsets (Figure 3A). We also analyzed Jarid2 expression in CD24^hiCD27⁺ B cells in response to stimulation with CpG oligonucleotides over a time course and found that Jarid2 expression decreased with increasing stimulation time (Figure 3B). In addition, Jarid2 expression was negatively correlated with IL-10 gene expression in the CD24^hiCD27⁺ B cells (Figure 3C). Furthermore, when the CD24^hiCD27⁺ B cells were electroporated with miR-155 inhibitor, Jarid2 gene and protein levels were significantly increased compared to those in the control group (Figures 3D,E). These data indicate that miR-155 can modulate Jarid2 expression. We further sorted the IL-10⁺ and IL-10⁻ B cells using beads combined with FACS as previously described (29). We found that the IL-10⁺ B cells expressed much lower levels of Jarid2 than the IL-10⁻ B cells (Figure 3F). When Jarid2 expression was knocked down in the B cells and the cells were stimulated with CpG oligonucleotides for 48 h, the B cells produced much higher levels of IL-10 and TNFα than the control group (Figures 3G–I). It has been reported that Jarid2 can silence the transcription of its target genes through H3K27me3 (28). Therefore, we used ChIP-qPCR to determine whether Jarid2 directly targets H3K27me3 and regulates IL-10 production in B10 cells. As the data showed when the B cells were electroporated with Jarid2-specific siRNA and H3K27me3 was pulled down, the binding of H3K27me3 to the promoter of IL10 was significantly reduced compared to that in the control group (Figure 3J). These data demonstrate that miR-155 activates IL-10 gene expression in B cells by regulating the Jarid2 target H3K27me3 and mediating repression.

We then determined the number of B10 cells in CD patients. Interestingly, there was no significant difference in IL-10 production and Jarid2 mRNA expression in the B cells from the HC and CD groups (Figure 4A and Figure S3 in Supplementary Material). Next, we determined the percentages of B cell subsets using the markers CD24 and CD27 and found that the number of CD24^hiCD27⁺ B cells were markedly reduced in the CD patients (12.43 ± 0.758 and 5.49 ± 0.912%, respectively) (Figure 4B). Additionally, the CD24^hiCD27⁺ B cells from CD patients produce lower levels of IL-10 than those from the HCs (Figure 4C). Therefore, we isolated the CD24^hiCD27⁺ B cells by FACS and conducted functional assays. We observed that when the CD24^hiCD27⁺ B cells isolated from CD patients and HCs were stimulated with CpG oligonucleotides for 24 h, washed and cocultured with CD14⁺ monocytes, the CD24^hiCD27⁺ B cells from the HCs could significantly inhibit TNFα production by the monocytes, while the CD24^hiCD27⁺ B cells from the CD patients could not (Figure 4D). Additionally, when activated B cells were cocultured with purified CD4⁺ T cells, IFNγ production by the T cells was significantly reduced by the B cells from HCs but not by those from the CD patients (Figure 4E). Therefore, in the CD patients, not only was the number of CD24^hiCD27⁺ B cells reduced, but their function was also impaired.

We next wanted to explore the reason underlying the impaired function of CD24^hiCD27⁺ B cells in CD patients. It has been reported that B cells can produce the inflammatory cytokine IL-6 and contribute to the pathogenesis of multiple sclerosis (30). In our study, we did not find an increase in IL-6 level in the B cells of CD patients (data not shown); however, interestingly, we found that CD24^hiCD27⁺ B cells from CD patients could produce higher levels of TNFα than the cells from HCs. Additionally, the percentage of IL-10⁺/TNFα⁺ B cells was much higher, and the ratio of IL-10⁺/TNFα⁺ B cells to IL-10⁺/TNFα⁻ B cells was significantly increased in the CD patients (Figures 5A–D). We then sorted the IL-10⁺/TNFα⁺ and IL-10⁺/TNFα⁻ B cells using beads combined with FACS and determined their function. As shown by the data, the expression level of IL-10 mRNA was very high in both subsets, but that of TNFα mRNA was very low in the IL-10⁺/TNFα⁻ B cells (Figure 5E). Furthermore, we determined the proliferative ability of the two subsets and found that the IL-10⁺/TNFα⁻ B cells had a much higher proliferative ability than the IL-10⁺/TNFα⁻ B cells (Figure 5F). In addition, when the isolated subsets were cocultured with purified CD4⁺CD25⁻ T cells and stimulated with anti-CD3 for 3 days, the Foxp3 and IFNγ expression levels were analyzed, and the IL-10⁺/TNFα⁻ B cells demonstrated a greater ability to induce Foxp3 expression and a greater ability to inhibit IFNγ production by the CD4⁺CD25⁻ T cells (Figures 5G,H). Also, IL-10⁺/TNFα⁻ B cells could inhibit monocyte production of TNFα better than other B cell subsets (Figure 5I). These data suggest that higher numbers of IL-10^+/TNFα⁺ B cells may contribute to the functional impairment of CD24^hiCD27⁺ B cells in CD patients.

To study the role of miR-155 in the function of CD19⁺CD24^hiCD27⁺ B cells from CD patients, first, we determined the miR-155 expression in CD19⁺CD24^hiCD27⁺ B cells. There was no significant difference between the HCs and CD patients (Figure 6A). We then transfected miR-155 mimic or inhibitor into the CD19⁺CD24^hiCD27⁺ B cells from the HCs and CD patients, stimulated them with CpG oligonucleotides for 2 days and analyzed their IL-10 and TNFα production. The data showed that after the transduction of miR-155 mimic into the CD19⁺CD24^hiCD27⁺ B cells, the production of both IL-10 and TNFα was increased in the HCs and CD patients, but the ratio of the IL-10⁺/TNFα⁺ B cells to IL-10⁺/TNFα^-B cells in the CD patients was much higher than that in the control group (Figures 6B–E). When the CD19⁺CD24^hiCD27⁺ B cells were transfected with the miR-155 inhibitor, the IL-10 and TNFα expression levels were decreased in both HC and CD groups. Interestingly, the ratio of IL-10⁺/TNFα⁺ B cells to IL-10⁺/TNFα⁻ B cells in the CD patients was still higher than that in the control group (Figures 6B–E). Therefore, in the CD patients, miR-155 could induce B cells to simultaneously secrete both IL-10 and a much higher level of TNFα, which may have reduced the function of CD19⁺CD24^hiCD27⁺ B cells.