In vivo-activated CD103+CD4+ regulatory T cells ameliorate ongoing chronic graft-versus-host disease

Abstract

CD103 (alphaEbeta7) has been shown to be an excellent marker for identifying in vivo-activated FoxP3(+)CD4(+) regulatory T (Treg) cells. It is unknown whether reinfusion of in vivo-activated donor-type CD103(+) Treg cells from recipient can ameliorate ongoing chronic graft-versus-host disease (GVHD). Here, we showed that, in a chronic GVHD model of DBA/2 (H-2(d)) donor to BALB/c (H-2(d)) recipient, donor-type CD103(+) Treg cells from recipients were much more potent than CD25(hi) natural Treg cells from donors in reversing clinical signs of GVHD and tissue damage. Furthermore, in contrast to CD25(hi) natural Treg cells, CD103(+) Treg cells expressed high levels of CCR5 but low levels of CD62L and directly migrated to GVHD target tissues. In addition, the CD103(+) Treg cells strongly suppressed donor CD4(+) T-cell proliferation; they also induced apoptosis of in vivo-activated CD4(+) T and B cells and significantly reduced pathogenic T and B cells in GVHD target tissues. These results indicate that CD103(+) Treg cells from chronic GVHD recipients are functional, and reinfusion of the CD103(+) Treg cells can shift the balance between Treg cells and pathogenic T cells in chronic GVHD recipients and ameliorate ongoing disease.

Figure 1

Differential percentage of donor-type CD103⁺FoxP3⁺ Treg cells in chimeric recipients with or without chronic GVHD. (A,B) Sublethally irradiated BALB/c recipients were transplanted with 6.25 or 100 × 10⁶ DBA/2 donor spleen cells. The recipients were monitored for proteinuria and survival weekly. There were 12 mice in each group combined from 3 replicate experiments. (C,D) Spleen cells from additional recipients with or without clinical chronic GVHD 20 days after transplantation and control DBA/2 donor mice were stained with anti-CD5.1 (donor T cell marker), CD4, CD103, and FoxP3. A representative staining pattern of 4 recipients and the percentage (mean ± SE) of CD103⁺FoxP3⁺ and CD103⁻FoxP3⁺ cells among total CD5.1⁺CD4⁺ cells are shown.

Figure 2

Differential surface receptors, tissue migration, and suppressor function of CD103⁺ Treg and CD25^hi natural Treg cells. Naive CD25^hi natural Treg cells were freshly isolated from spleen of DBA/2 donor mice. The in vitro-activated and expanded CD25^hi natural Treg cells were from sorted naive CD25^hi natural Treg cells after being cultured with anti-CD3/CD28 beads and IL-2 for 7 days. The CD103⁺ Treg cells were from spleen of day 20 chronic GVHD recipients. (A) The viability as judged by 4,6 diamidino-2-phenylindole (DAPI) staining, the expression levels of FoxP3, CD25, CD103, CD62L, and CCR5 by freshly isolated CD25^hiCD4⁺ T cells, in vitro-activated CD25^hiCD4⁺ T cells, and in vivo–activated CD103⁺CD4⁺ T cells. One representative is shown of 4 replicated experiments. (B) CFSE-labeled in vitro–activated CD25^hi natural Treg cells and in vivo–activated CD103⁺ Treg cells (10⁶ each) were injected intravenously into chronic GVHD recipients, respectively. Twelve hours after injection, the percentage of CFSE⁺FoxP3⁺ cells among total FoxP3⁺ cells in the MLN, spleen, and liver was compared. One representative of 4 examined recipients in each group is shown. The means (± SE) of CD25^hi natural Treg cells versus CD103⁺ Treg cells in different tissues are as follows: 10.8% (± 1.2%) versus 4.9% (± 1.3%) in MLN, 0.3% (± 0.2%) versus 0.5% (± 0.3%) in spleen, and 0.4% (± 0.2%) versus 3.5% (± 0.8%) in liver. (C) Suppression of host DC-induced donor CD4⁺CD25⁻ T proliferation by freshly isolated and in vitro–activated CD25^hi natural Treg cells as well as CD103⁺ Treg cells. (D) CD103⁺ Treg cell suppression of donor CD4⁺CD25⁻ T proliferation induced by host, MHC-matched third-party DCs, or MHC-mismatched third-party DCs. (E) In vitro–activated CD25^hi natural Treg cell suppression of donor CD4⁺CD25⁻ T proliferation induced by host or MHC-mismatched third-party DCs. (C-E) Means (± SE) of 4 replicated experiments.

Figure 3

CD103⁺ Treg cells potently ameliorated ongoing chronic GVHD. Sublethally irradiated BALB/c recipients were given 100 × 10⁶ DBA/2 donor spleen cells to induce chronic GVHD with proteinuria. Twenty days after HCT, the chronic GVHD recipients were infused with either CD103⁺CD4⁺ Treg cells from other recipients or freshly isolated or in vitro–activated CD25^hi natural Treg cells from DBA/2 donors (1-5 × 10⁶ each). The control recipients were given PBS only. The recipients were monitored daily for proteinuria and survival. (A) The percentage of recipients with reversal of proteinuria and percentage of survival. There were 6 to 12 recipients in each group combined from 2 or 3 replicate experiments. (B) The serum levels of anti-dsDNA IgG 20 days after infusion of CD103⁺ Treg cells (10⁶), in vitro–activated natural Treg cells (5 × 10⁶), or PBS. (C) Histopathology score (mean ± SE, n = 6). (D) Histology of skin, liver, and kidney of representative recipients before and after treatment with CD103⁺ Treg cells or in vitro–activated natural Treg cells. Tissues of BALB/c mice without HCT are shown as normal control. Slides were examined at 400× magnification. The samples were visualized with an Olympus BX51 fluorescent microscope equipped with Olympus 20×/0.70 NA and 40×/0.90 PlanApo objectives (Olympus America, Melville, NY) and a Pixera (600CL) cooled CCD camera (Pixera, Los Gatos, CA). Fluorescent images relative to each marker were collected using a corresponding filter set and Pixera Viewfinder acquisition software 3.0. Color composite images were generated using Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA).

Figure 4

CD103⁺ Treg cells induced apoptosis of autoantibody-secreting plasma cells. (A) Twenty days after infusion of CD103⁺ Treg cells, the spleen cells of the untreated and treated recipients were stained for B220 versus syndecan. The percentage of syndecan⁺ plasma cells and syndecan⁻B220⁺ B cells is shown beside the gating box. One representative of 4 examined recipients is shown. The means (± SE) of syndecan⁺ cells versus syndecan⁻B220⁺ cells in the untreated and treated recipients are 4.1% (± 0.4%) versus 1.0% (± 0.2%) and 34.5% (± 4.5%) versus 33.9% (± 3.8%), respectively. (B) Spleen cells (10⁶) of the untreated recipients, including whole spleen (SPL), syndecan⁺ cell–depleted spleen (Syn⁻-SPL), sorted syndecan⁺ cells (Syn⁺), sorted syndecan⁻B220⁺ cells (Syn⁻), were cocultured with or without sorted CD103⁺ or CD103⁻CD4⁺ T cells (0.2 × 10⁶) from the spleen of chronic GVHD recipients for 5 days. Anti-dsDNA IgG in the culture supernatant was measure by enzyme-linked immunosorbent assay. The mean (±SE) of 4 replicated experiments is shown. N/A indicates an antibody concentration below detectable levels. (C) Sorted activated syndecan⁺ and syndecan⁻B220⁺ B cells from chronic GVHD recipients and B220⁺ B cells from normal donors were cocultured with CD103⁺ or CD103⁻CD4⁺ T cells from chronic GVHD recipients for 8 hours. Thereafter, the cells were stained for annexin V versus DAPI. The percentage of annexin V⁺DAPI⁺ cells among Syn⁺ or sydecan⁻B220⁺ cells is shown beside the gating box. One representative of 4 replicated experiments is shown. The means (± SE) of the percentage of annexin V⁺DAPI⁺ cells among recipient syndecan⁺ cells are 5.4% (± 1.3%), 28.6% (± 3.6%), and 6.1% (± 1.7%), respectively, when cocultured alone or cocultured with CD103⁺ or CD103⁻CD4⁺ T cells. Similarly, the means (± SE) percentage among recipient syndecan⁻B220⁺ cells are 6.5% (± 1.9%), 15.4% (± 2.5%), and 6.8% (± 1.3%), and the mean (± SE) percentage among donor B220⁺ B cells are 5.5% (± 1.3%), 6.3% (± 1.5%), and 7.7% (± 1.4%).

Figure 5

CD103⁺ Treg cells induced apoptosis of activated CD4⁺ T cells from chronic GVHD recipients. (A) Twenty days after infusion of CD103⁺ Treg cells, the spleen cells of the untreated and treated recipients as well as control donor DBA/2 mice were stained for CD5.1 versus CD4. The CD5.1⁺CD4⁺ donor-type T cells were gated. One representative of 4 recipients is shown. The mean (± SE) of the percentage of CD5.1⁺CD4⁺ T cells among spleen mononuclear cells of untreated and treated recipients are 16.8% (± 3.6%) and 8.5% (± 1.6%), respectively. (B) Gated FoxP3⁻CD4⁺ T cells were shown in CD44 versus CD62L. The activated CD44^hiCD62L^loCD4⁺ T cells were gated, and the percentage of the cells among total FoxP3⁻CD4⁺ T cells was shown above the gated box. One representative of 4 recipients is shown. The means (± SE) of the percentage of CD44^hiCD62L^loCD4⁺ T cells among total FoxP3⁻CD4⁺ T cells of untreated and treated recipients are 71.0% (± 4.7%) and 38.5% (± 6.7%), respectively. (C) CD103⁻CD4⁺ T cells from chronic GVHD recipients or normal DBA/2 donors (10⁶ each) were cocultured with CD103⁺ Treg cells from chronic GVHD recipients for 8 hours. Thereafter, the cells were stained for annexin V versus DAPI. The percentage of annexin V⁺DAPI⁺ cells among CD103⁻CD4⁺ T cells is shown beside the gating box. One representative of 4 experiments is shown. The means (± SE) of annexin V⁺DAPI⁺ cell among CD103⁻CD4⁺ T cells from chronic GVHD recipients with or without CD103⁺ Treg cells in coculture are 25.3% (± 3.3%) versus 4.9% (± 1.4%). The means (± SE) of annexin V⁺DAPI⁺ cells among donor CD103⁻CD4⁺ T cells with or without CD103⁺ Treg cells in coculture are 4.7% (± 1.1%) versus 4.2% (± 0.8%).

Figure 6

CD103⁺ Treg cells reduced pathogenic CD4⁺ T cells in GVHD target tissues. Twenty days after infusion of CD103⁺ Treg cells, mononuclear cells from kidneys and livers of untreated and treated recipients were stained for CD4, CD103, IL-17, IL-10, and IFN-γ. (A) The gated CD4⁺ T cells were shown in FoxP3 versus CD103. (B) The gated CD4⁺ T cells were shown in IL-17 vs CD4. (C) Gated CD4⁺ T cells were shown in IL-10 vs IFN-γ. One representative of 4 replicated experiments is shown. The means (± SE) of percentage of FoxP3⁺ cells among CD4⁺ T cells in kidney and liver of untreated and treated recipients are 12.6% (± 2.5%) versus 31.7% (± 3.6%) and 16.8% (± 2.5%) versus 26.9% (± 4.5%), respectively. The means (± SE) of percentage of IL-17⁺ cells among CD4⁺ T cells in kidney and liver of untreated and treated recipients is 3.1% (± 1.7%) versus 1.6% (± 0.2%) and13.6% (± 2.1%) versus 7.4% (± 0.7%), respectively. The mean (± SE) percentages of IL-10⁺IFN-γ⁺ cells among CD4⁺ T cells in kidney and liver of untreated and treated recipients are 15.8% (± 3.5%) versus 1.8% (± 0.4%) and 0.8% (± 0.3%) versus 0.3% (± 0.2%), respectively. The mean (± SE) percentages of IFN-γ⁺ cells among CD4⁺ T cells in kidney and liver of untreated and treated recipients are 7.9% (± 0.9%) versus 3.8% (± 0.7%) and 19.4% (± 2.8%) versus 8.1% (± 1.2%), respectively.