Peyer's patches-derived CD11b+ B cells recruit regulatory T cells through CXCL9 in dextran sulphate sodium-induced colitis

Abstract

Regulatory T (Treg) cells play an essential role in the maintenance of intestinal homeostasis. In Peyer's patches (PPs), which comprise the most important IgA induction site in the gut-associated lymphoid tissue, Treg cells promote IgA isotype switching. However, the mechanisms underlying their entry into PPs and isotype switching facilitation in activated B cells remain unknown. This study, based on the dextran sulphate sodium (DSS)-induced colitis model, revealed that Treg cells are significantly increased in PPs, along with CD11b⁺ B-cell induction. Immunofluorescence staining showed that infiltrated Treg cells were located around CD11b⁺ B cells and produced transforming growth factor-β, thereby inducing IgA⁺ B cells. Furthermore, in vivo and in vitro studies revealed that CD11b⁺ B cells in PPs had the capacity to recruit Treg cells into PPs rather than promoting their proliferation. Finally, we found that Treg cell recruitment was mediated by the chemokine CXCL9 derived from CD11b⁺ B cells in PPs. These findings demonstrate that CD11b⁺ B cells induced in PPs during colitis actively recruit Treg cells to accomplish IgA isotype switch in a CXCL9-dependent manner.

Keywords: CD11b+ B cells; colitis; regulatory T cells.

Figure 1

Both regulatory T (Treg) cells and CD11b⁺ B cells increased in Peyer's patches (PPs) during dextran sulphate sodium (DSS) ‐induced colitis. (a, b) Flow cytometry was used to determine the percentage and absolute number of Treg cells and CD11b⁺ B cells in PPs of wild‐type (WT) mice on days 0, 4, 7 and 10 after colitis induction. Treg cell expressions (CD4⁺ Foxp3⁺) within the CD4‐gated population is shown in (a). CD11b⁺ B‐cell expressions within the CD19‐gated population are shown in (b). *P < 0·05; ***P < 0·001. Data shown are the mean ± SEM from one experiment with eight mice, which was repeated at least three times with similar results. (c) Immunofluorescence microscopic analysis was performed to examine the distribution of cells expressing CD19 (red), CD11b (cyan) (CD19 and CD11b double‐positive, yellow) and Foxp3 (green) in the representative PP sections obtained from WT mice on days 0, 4, 7 and 10 after colitis induction. Bars indicate 50 μm.

Figure 2

Induced CD11b⁺ B cells may increase regulatory T (Treg) cells in Peyer's patches (PPs) in vivo. (a) μMT mice were intravenously (i.v.) administered with PP‐derived B cells from wild‐type (WT) mice and CD11b‐deficient mice, respectively, and administered with dextran sulphate sodium (DSS) 2 days later after adoptive transfer. Treg cell percentages (Foxp3⁺) in PPs within the CD4‐ or CD45‐gated population were detected on day 7 after colitis induction. (b) Purified CD19⁺ CD11b⁺ B cells and CD19⁺ CD11b⁻ B cells in PPs were sorted from colitis mice on day 7 after DSS administration and were adoptively transferred into WT mice. Treg cell percentages (Foxp3⁺) in PPs within the CD4‐ or CD45‐gated population were detected 4 days later. The number of transferred cells were at 5 × 10⁶ cells suspended in 200 μl sterile phosphate‐buffered saline (PBS). The control group mice receive PBS instead. *P < 0·05; data are representative of six independent experiments.

Figure 3

Induced CD11b⁺ B cells, but not CD11b⁻ B cells, could increase regulatory T (Treg) cells though recruiting route rather than proliferation in vitro. (a) CD11b⁺ B cells of Peyer's patches (PPs) from day 7 in mice with colitis were sorted with a MoFlo high‐speed cell sorter (Beckman Coulter, Pasadena, CA). Treg cells (CD4⁺ CD25⁺; 2 × 10⁵ cells/well) were purified from naive wild‐type (WT) mice with MicroBeads (Miltenyi Biotec) and were cultured alone or with CD11b⁺ B cells (4 × 10⁵ cells/well) in 48‐well plates for 72 hr in complete RPMI‐1640. The percentage of Ki‐67⁺ Treg cells was determined by FACS. CD11b⁺ B cells and Treg cells were prepared from four mice. (b) The supernatant samples were obtained either from purified CD11b⁺ B cells or CD11b⁻ B cells (3 × 10⁵ cells/well) sorted from PPs on days 4, 7 and 10, from mice with colitis and then cultured in RPMI‐1640 (without fetal bovine serum) for 24 hr. Total PP lymphocyte suspensions (2 × 10⁶ cells/well) were prepared from the respective period (days 4, 7 and 10) of mice with colitis. Supernatants and cell suspensions were separated by Transwell in a 24‐well plate for 4 hr. Percentage and absolute number of Treg cells were determined by FACS. *P < 0·05; **P < 0·01. Results presented as the means of four independent experiments.

Figure 4

The induced CD11b⁺ B cells initially produce CXCL9 to increase regulatory T (Treg) cells. (a) Real‐time PCR analysis was performed to examine the chemokine expressions of Peyer's patches (PPs) CD11b⁺ B cells purified from dextran sulphate sodium (DSS) ‐treated mice. The chemokine profile was developed using the hemi software. Samples of every phage in colitis contain four mice. (b) CXCL9⁺ expression within the CD11b⁺ or CD11b⁻‐gated population was obtained in PPs on days 0, 4, 7 and 10 after colitis induction. (c) PP‐derived CD11b⁺ B cells from mice with colitis were sorted and cultured in complete RPMI‐1640 for 24 hr. The cell supernatant was harvested, and the level of chemokine CXCL9 was determined by enzyme‐linked immunosorbent assay. (d) CXCR3 expressions gated on Foxp3⁺ T cells from PPs during the process of colitis were detected using flow cytometry. (e) Total PP‐derived lymphocyte suspension (described as in Fig. 3b, 2 × 10⁶ cells/well) was either cultured alone or with CXCL9 (50 ng/ml), CD11b⁺ B‐cell supernatant (described as in Fig. 3b) CD11b⁺ B cells supernatant and CXCL9 or CD11b⁺ B‐cell supernatant and anti‐CXCL9 antibody (1 μg/ml), which were separated in a 24‐well Transwell plate for 4 hr. Percentage and absolute number of Treg cells were determined by FACS. (f; i) Illustration of the protocol of anti‐CXCL9 antibody (4·5 μg/g) administration in DSS‐induced mice. (f; ii) Frequencies and absolute numbers of PP‐derived Treg cells on days 7 and 10 after CXCL9 neutralization were analysed by flow cytometry. *P < 0·05; *P < 0·05; **P < 0·01. Results presented as the mean ± SEM of one experiment, which was repeated at least three times yielding similar results.