Regulatory T cells can migrate to follicles upon T cell activation and suppress GC-Th cells and GC-Th cell-driven B cell responses

Abstract

How Tregs migrate to GCs, and whether they regulate the helper activity of the T cells in GCs (GC-Th cells) remains poorly understood. We found a T cell subset in human tonsils that displays potent suppressive activities toward GC-Th cell-dependent B cell responses. These Tregs with the surface phenotype of CD4+CD25+CD69- migrate well to CCL19, a chemokine expressed in the T cell zone, but poorly to CXCL13, a chemokine expressed in the B cell zone. This migration toward the T cell-rich zone rapidly changes to trafficking toward B cell follicles upon T cell activation. This change in chemotactic behavior upon activation of T cells is consistent with their switch in the expression of the 2 chemokine receptors CXCR5 and CCR7. CD4+CD25+CD69- Tregs suppress GC-Th cells and GC-Th cell-induced B cell responses such as Ig production, survival, and expression of activation-induced cytosine deaminase. Our results have identified a subset of Tregs that is physiologically relevant to GC-Th cell-dependent B cell responses and a potential regulation mechanism for the trafficking of these Tregs to GCs.

Figure 1

CD4⁺CD25⁺ T cell population and suppression of Ig production. (A) CD4⁺CD25⁺ T cell populations in tonsils. Tonsil mononuclear cells were stained with antibodies to CD4, CD57, CD25, and CD69. The graph shows the combined data (averages and SEM) from 5 different experiments. (B) CD4⁺CD25⁺CD69^– cells, not the total CD4⁺CD25⁺ T cell population, efficiently suppressed the CD57⁺ GC-Th cell_induced B cell production of Ig. CD4⁺CD25⁺ (CD57^–) T cell subsets were sorted and cultured with 10⁵ CD57⁺ GC-Th cells and 10⁵ tonsil CD19⁺ B cells (Total B) for 5 days, followed by ELISA to quantitate the total IgG secreted in the cultures. Representative data (averages and SD of triplicate ELISA measurements) from 5 independent experiments are shown.

Figure 2

Phenotype of CD4⁺CD25⁺CD69^– Tregs. (A) Freshly isolated tonsil T cell subsets were stained with antibodies to TGF-β1, CTLA-4, CD62 ligand (CD62L), and CD45RA. Stained cells were analyzed by a FACSCalibur. The histograms show representative data, and the graphs show combined data (averages and SEM, n = 3). Mean fluorescent intensity levels for the antigens expressed by the 2 populations are also shown in parentheses. (B) Expression of Foxp3 and GITR by CD4⁺CD25^–CD69⁺ cells, CD4⁺CD25⁺CD69⁺ cells, CD4⁺CD25⁺CD69^– Tregs, and CD4⁺CD25^–CD69^– T cells, determined by RT-PCR. (C) Foxp3 expression by T cell subsets measured by quantitative real-time RT-PCR. Foxp3 expression was normalized for β-actin levels. Representative data (averages and SD of triplicate measurements) from 4 independent experiments are shown.

Figure 3

Differential expression of chemokine receptors by CD4⁺CD25⁺CD69^– Tregs and CD4⁺CD25⁺CD69⁺ T cells. Fresh tonsil mononuclear cells were stained with antibodies to the indicated chemokine receptors and CD4, CD25, and CD69. (A) Representative dot plot data. (B) Averages and SEM of the data obtained from 3 independent experiments. *Significant differences between the 2 T cell subsets. The P values were 0.03 (CCR2), 0.007 (CCR4), 0.001 (CCR5), 0.039 (CCR6), 0.032 (CCR7), 0.001 (CXCR3), 0.002 (CXCR4), 0.001 (CXCR5), and 0.018 (CXCR6). Mean fluorescent intensity (MFI) levels for the chemokine receptors are also shown in parentheses in B. mIgG2b, mouse IgG2b.

Figure 4

Chemotaxis assay. (A) Chemotactic behaviors of CD4⁺CD25⁺CD69^– Tregs and CD4⁺CD25⁺CD69⁺ T cells. (B) CD4⁺CD25⁺CD69^– Tregs poorly migrate to CCL1. Fresh tonsil mononuclear cells were used as input cells for chemotaxis assays. Indicated chemokines were first titrated to determine optimal concentrations: CXCL13 (4,000 ng/ml), CXCL12 (100 ng/ml), CXCL10 (1,000 ng/ml), CCL19 (2,000 ng/ml), CCL17 (200 ng/ml), CCL4 (100 ng/ml), and CCL1 (500 ng/ml). Cells were allowed to migrate for 3 hours. The migrated cells and input cells were harvested, stained for CD4, CD25, and CD69, and counted by a FACSCalibur. Specific migration after subtraction of the background migration is shown. The background migration rates (percent averages and SEM, 3 experiments) for the 4 subsets were 25 ± 2.6 (CD4⁺CD25^–CD69⁺), 13.4 ± 4.3 (CD4⁺CD25⁺CD69⁺), 6.1 ± 0.4 (CD4⁺CD25⁺CD69^– Treg), and 7.2 ± 2 (CD4⁺CD25^–CD69^–). The averages and SEM of the data obtained from 3 (A) and 4 (B) independent experiments are shown. *Significant differences between the 2 subsets (A) or from CD4⁺CD25⁺CD69^– Tregs (B). The P values were 0.048 (CXCL13) and 0.015 (CCL19) in A; and 0.046 (CD25⁺CD69^– Treg vs. CD25⁺CD69⁺), 0.007 (CD25⁺CD69^– Treg vs. CD25^–CD69⁺), and 0.03 (CD25⁺CD69^– Treg vs. CD25^–CD69^–) in B.

Figure 5

CD4⁺CD25⁺CD69^– Tregs switch their expression pattern of chemokine receptors and chemotactic responsiveness upon T cell activation. Tonsil mononuclear cells, freshly isolated or briefly activated for 10_14 hours with phytohemagglutinin (PHA) or for 5_10 hours with anti-CD3 and anti-CD28, were examined for their expression of chemokine receptors (A, B, and D) or for chemotactic responsiveness (C and E). In D, the chemokine receptors were examined by 1- or 2-step staining methods to avoid cross-reaction with the anti-CD3 and anti-CD28 antibodies used to activate the cells; these methods are less sensitive than the 3-step method used for A and B. The data obtained from 3 independent experiments were combined, and averages (and SEM in A) are shown. The background percent migration rates (averages and SEM, 3 experiments) for the 4 cell subsets were 23 ± 4.4 (CD4⁺CD25^–CD69⁺), 16.3 ± 3.2 (CD4⁺CD25⁺CD69⁺), 23 ± 1.0 (CD4⁺CD25⁺CD69^– Treg), and 11 ± 3.2 (CD4⁺CD25^–CD69^–). *Significant differences between the 2 samples. The P values were 0.024 (a), 0.042 (b), 0.032 (c), 0.019 (d), 0.045 (e), 0.033 (f), 0.041 (g), 0.002 (h), 0.049 (i), 0.021 (j), 0.023 (k), 0.009 (l), 0.004 (m), and 0.006 (n).

Figure 6

CD4⁺CD25⁺CD69^– Tregs are found in the IFAs and GCs of tonsils. To measure distribution of CD4⁺CD25⁺CD69^– Tregs, frozen tonsil sections were stained for CD4, CD25, and CD69 (A and B). Secondary follicles with GCs were identified by staining of serial sections with antibodies for CD57, IgD, and CD4 (not shown). Stained sections were analyzed with a confocal microscopy system (Bio-Rad Laboratories Inc. MRC-1024UV microscope, and Nikon Inc. Diaphot 300 microscope). A representative set of data from 5 independent experiments using 3 different tonsil specimens is shown. Arrows indicate CD4⁺CD25⁺CD69^– Tregs, and the circled cells are CD4⁺CD25⁺CD69⁺ T cells.

Figure 7

CD4⁺CD25⁺CD69^– Tregs suppress GC-Th cell_induced Ig production by naive, GC, and memory B cells. (A) CD4⁺CD25⁺CD69^– Tregs and CD57⁺ GC-Th cells were cocultured with naive, GC, or memory B cells. CD4⁺CD25⁺CD69^– Tregs (10⁵) were added to the cultures of 10⁵ CD57⁺ GC-Th cells and 10⁵ B cells. *Less than 1% of the B plus GC-Th levels. (B) Various numbers of CD4⁺CD25⁺CD69^– Tregs (0.5 × 10⁵ to 2 × 10⁵) were cultured with CD57⁺ GC-Th cells and GC-B cells. (C) Neutralizing antibodies to CTLA-4 and TGF-β1 or isotype control antibodies were added to the cultures of CD57⁺ GC-Th cells, GC-B cells, and Tregs. Concentrations of IgM, IgG, IgA, and IgE were determined by ELISA after 5 days in culture. ELISA was performed in triplicate (A and C) or duplicate (B), and the averages and differences are shown. Representative data from 3 independent experiments are shown in A and B. For C, the data obtained from 3 independent experiments were combined, and averages and SEM are shown. *Significant differences from GC-B + GC-Th + Treg + isotype Ab culture. The P values were 0.0001 (a), 0.0001 (b), 0.0001 (c), 0.0001 (d), 0.001 (e), 0.006 (f), 0.01 (g), 0.03 (h), 0.004 (i), and 0.003 (j).

Figure 8

CD4⁺CD25⁺CD69^– Tregs suppress CXCL13 production by CD57⁺ GC-Th cells. Equal numbers (10⁵ cells) of tonsil CD19⁺ B cells, CD57⁺ GC-Th cells, and/or CD4⁺CD25⁺CD69^– Tregs were cultured together for 5 days in the presence of staphylococcal enterotoxin B. The concentrations of CXCL13 protein were measured by ELISA. One set of representative data from 4 independent experiments is shown. UD, undetectable.

Figure 9

CD4⁺CD25⁺CD69^– Tregs suppress GC-B cell survival and GC-Th cell_induced AID expression. (A) CD4⁺CD25⁺CD69^– Tregs suppress CD57⁺ GC-Th cell_dependent GC-B cell survival. CD4⁺CD25⁺CD69^– Tregs were cultured along with 10⁵ (×1) CD57⁺ GC-Th cells and GC-B cells for 5 days. Live B cells (CD19⁺CD4^– cells), acquired by a FACSCalibur for 100 seconds after culture, are shown as dot plots. One set of representative dot plot data from 3 independent experiments is shown. *Significant differences from the cultures of GC-B and GC-Th cells in 3 independent experiments. (B and C) CD4⁺CD25⁺CD69^– Tregs suppress the expression of AID induced by CD57⁺ GC-Th cells. Indicated numbers of CD4⁺CD25⁺CD69^– Tregs were cultured for 5 days with 0.5 × 10⁶ (×1) CD57⁺ GC-Th cells and 0.5 × 10⁶ naive (NV) B cells. Cultured cells were harvested and examined for the expression of AID and β-actin by RT-PCR. A representative set of data from 3 independent experiments is shown in B, and combined data (averages and SEM) in graph form are shown in C.

Figure 10

CD4⁺CD25⁺CD69^– Tregs need to contact target cells for effective suppression. Equal numbers (10⁵ cells) of CD19⁺ B cells, CD57⁺ GC-Th cells, and/or CD4⁺CD25⁺CD69^– Tregs were cultured in upper and/or lower Transwell chambers in the presence of staphylococcal enterotoxin B. Cells were cultured for 5 days, and the 4 subsets of Ig in the lower chambers were quantitated by ELISA. Relative Ig responses (percent) to that of the cultures of B cells and CD57⁺ GC-Th cells are shown. The data from 5 independent experiments were combined, and the averages and SEM are shown.