Inhibition of CD4+CD25+ regulatory T-cell function by calcineurin-dependent interleukin-2 production

Abstract

CD4+CD25+ regulatory T (Treg) cells control immunologic tolerance and antitumor immune responses. Therefore, in vivo modification of Treg function by immunosuppressant drugs has broad implications for transplantation biology, autoimmunity, and vaccination strategies. In vivo bioluminescence imaging demonstrated reduced early proliferation of donor-derived luciferase-labeled conventional T cells in animals treated with Treg cells after major histocompatibility complex mismatch bone marrow transplantation. Combining Treg cells with cyclosporine A (CSA), but not rapamycin (RAPA) or mycophenolate mofetil (MMF), suppressed Treg function assessed by increased T-cell proliferation, graft-versus-host disease (GVHD) severity, and reduced survival. Expansion of Treg and FoxP3 expression within this population was lowest in conjunction with CSA, suggesting that calcineurin-dependent interleukin 2 (IL-2) production is critically required for Treg cells in vivo. The functional defect of Treg cells after CSA exposure could be reversed by exogenous IL-2. Further, the Treg plus RAPA combination preserved graft-versus-tumor (GVT) effector function against leukemia cells. Our data indicate that RAPA and MMF rather than CSA preserve function of Treg cells in pathologic immune responses such as GVHD without weakening the GVT effect.

Figures

Figure 1.

In vitro effects of CSA, RAPA, and MPA on Treg cells. In vitro effects of CSA, RAPA, and MPA on allostimulated Treg cells with respect to suppressor function (A), FoxP3 expression (B,D), and IL-2 levels in the primary coculture (C). (A) Fresh isolated C57B/6 Treg cells (CD4+CD25highH-2kb+) were incubated with Balb/c APCs (CD11c H-2kd) with CSA (100 ng/mL), RAPA (10 ng/mL), or MPA (50 ng/mL). Treg cells were reisolated by sorting for CD4+CD25highH-2kb+ cells after 72 hours of primary coculture. Reisolated Treg cells were then incubated for 96 hours with γ-irradiated (30 Gy) APCs (CD11c+H-2kd+) and CFSE-labeled TCONV cells (CD4+CD25-H-2kb+Thy-1.2+), with each population 2 × 105 cells. Cell division was monitored by levels of CFSE dilution. The congenic markers Thy-1.1 and Thy-1.2 were used to distinguish between Treg and TCONV cells, respectively. Histograms show the FACS profile of CFSE+ TCONV cells. Numbers of events in each cell division (n) are indicated below the respective peak. The percentage of undivided CD4+CFSE+ cells in each culture condition is indicated next to the right upper corner of the respective histogram. One representative experiment of 4 is presented. Allostimulated Treg cells suppress alloantigen-driven expansion of CFSE+ TCONV (*) significantly more strongly than autostimulated Treg cells (+; P = .027). CSA-exposed Treg cells are still suppressive, but significantly less as compared to RAPA (#P = .031) and addition of IL-2 (50 IU/mL) to the activation culture restores Treg function partially (‡; P = .037). (B) Relative FoxP3 mRNA expression level in Treg cells (H-2kb+) exposed to allogeneic APCs (H-2kd+) alone or in conjunction with RAPA (10 ng/mL), MPA (50 ng/mL), CSA (100 ng/mL), or CSA (100 ng/mL) plus IL-2 or exposed to autologous APCs (H-2kd). The increase of FoxP3 expression in Treg cells during allostimulation is reduced by CSA addition and restored by IL-2. Data represent means ± SD of triplicates (**P = .004; ***P = .008; ****P = .007). (C) IL-2 concentrations measured by ELISA in supernatants from the following cocultures are depicted (y-axis): Treg cells (H-2kb) incubated with APCs (H-2kd) and CSA (▪), RAPA (▵), or MPA (•)at different concentrations (x-axis) after 72 hours. CSA but not RAPA or MPA reduces IL-2 levels in the Treg activation culture in a dose-dependent manner. (D) Histograms depict the intracellular expression profile of FoxP3 in live-gated Treg cells exposed to medium, CSA (100 ng/mL), RAPA (10 ng/mL), or MPA (50 ng/mL) for 72 hours (filled histogram, unstained cells; open histogram, FoxP3 staining). The mean fluorescence intensity (MFI) for FoxP3 expression is significantly lower in Treg cells reisolated from CSA (213 ± 24) as compared to cultures containing RAPA (412 ± 42) or MPA (392 ± 31; P < .05).

Figure 2.

Effects of different doses of CSA, RAPA, and MMF on GVHD-related mortality. Data from 3 independent experiments are combined. (A) Percentage survival of animals receiving CSA 10 mg/kg (□, n = 12), CSA 50 mg/kg (▪, n = 12), or PBS (⋄, n = 12). The higher CSA dose improves survival (▪ versus ⋄, P = 0.032). (B) Percentage survival of animals receiving RAPA 0.5 mg/kg (○, n = 12), RAPA 1.5 mg/kg (•, n = 12), or PBS (⋄, n = 12). The higher RAPA dose improves survival (• versus ⋄, P = 0.029). (C) Percentage survival of animals receiving MMF 90 mg/kg (▵, n = 12), MMF 200 mg/kg (▴, n = 12), or PBS (⋄, n = 12), not significant (NS). Control animals received TCD-BM only (*). (D) Distribution of donor derived luc+ TCONV cells in Balb/c recipients at day 5 after BMT with coadministration of different doses of CSA (top row), RAPA (middle row), and MMFl (bottom row). At this time point the major TCONV expansion is found in cLNs, GIT, and SPL. Although the localization pattern of TCONV is conserved, proliferation is gradually reduced when different doses of immunosuppressants are administered.

Figure 3.

CSA but not RAPA or MMF interferes with Treg suppressor function in vivo. Expansion of luc+ donor TCONV cells in animals receiving TCONV cells alone or TCONV and Treg cells or TCONV and Treg cells in conjunction with CSA (10 mg/kg), RAPA (0.5 mg/kg), or MMF (90 mg/kg) as shown for 3 representative animals at days 4, 6, 8, and 14 after BMT (A) and as quantified by emitted photons over total body area at serial time points after BMT (B). GVHD was induced as described in “Materials and methods.” Data from 3 independent experiments are combined. (A) Proliferation of luc+ TCONV (first column from left) is reduced by cotransfer of Treg cells (second column from left). Addition of CSA reduces the suppressive effect of Treg cells (third column from left). In contrast RAPA allows for Treg function (fourth column from left) and MMF has minimal impact (fifth column from left). (B) TCD-BM (▴,n = 15), with TCONV cells (•,n = 15), with TCONV and Treg cells (▾,n = 10), with TCONV and Treg cells and CSA (▵,n = 10), with TCONV and Treg cells and RAPA (□,n = 10), or with TCONV and Treg cells and MMF (○,n = 10). Signal intensity is significantly higher in animals receiving TCONV and Treg cells and CSA as compared to RAPA (▵ versus □, P = .001, day 14) and MMF (▵ versus ○, P = .009, day 14). (C) Percentage survival of Balb/c recipients is significantly lower when combining Treg cells with CSA as compared to RAPA (▵ versus □, P = .001) or MMF (▵ versus ○, P = .007). (D) Expansion of luc+ TCONV cells at day 10 after BMT in the GIT. (Di) Background signal in the GIT of animals having received only TCD-BM without any luc+ cells. Proliferation of TCONV cells in mLNs and the GIT (Dii) is reduced by cotransfer of Treg cells (Diii). Addition of CSA reduces the protective effect of Treg cells (Div). RAPA (Dv) and MMF (Dvi) do not abrogate Treg suppression of TCONV proliferation in vivo.

Figure 4.

Serum IFN-γ and TNF-α levels. Serum was collected on day 7 after transplantation of Balb/c recipients as indicated for each individual bar (each 5 animals). Data are derived from one of 2 independent experiments. (A) Animals receiving Treg cells in conjunction with RAPA have a significantly lower IFN-γ serum level as compared with Treg cells plus CSA, *P = .02. (B) TNF-α levels are lowest in animals receiving TCD-BM. Differences between the groups that receive TCONV/Treg cells in conjunction with different immunosuppressants are not statistically significantly different. (C) The histopathologic score for the different groups is shown. Analysis included the large bowel, small bowel, and liver from recipients on days 5 and 15. Coded tissue samples were scored by a pathologist blinded to the treatment groups as described in “Materials and methods.”

Figure 5.

CSA administration is associated with reduced numbers of CD4+FoxP3+ cells and has only minor impact on the expansion of donor-derived CD4+CD25+ cells. Expansion of luc+ donor Treg cells in Balb/c mice (H-2kd) receiving TCONV and Treg cells (both H-2kq) alone or TCONV and Treg cells along with CSA (10 mg/kg), RAPA (0.5 mg/kg), or MMF (90 mg/kg) as shown for representative animals on day 7 after BMT (A) and as quantified by emitted photons over total body area at serial time points after BMT (B). (A) Expansion of luc+ Treg (first column from left) is slightly reduced by conjunction of CSA (second column from left) and to a lesser extent by MMF (first column from right). Addition of RAPA (third column from left) has no impact on Treg expansion. (B) TCD-BM (*, n = 10), with TCONV and Treg cells (□,n = 10), with TCONV and Treg cells and CSA (▵,n = 10), with TCONV and Treg cells and RAPA (□, n = 10), or with TCONV and Treg cells and MMF (○, n = 10). Signal intensity is decreased in animals receiving TCONV and Treg cells and CSA or MMF as compared to RAPA (▵ versus □, and ○ versus □, NS). (C) Splenic CD4 donor T cells (H-2kq) were investigated for FoxP3 expression in Balb/c recipients on day 7 after BMT, having received TCONV and Treg cells alone (first column from left) in conjunction with CSA (second column from left), RAPA (third column from left), or MMF (fourth column from left). The percentage of FoxP3+ Treg cells is significantly reduced in the presence of CSA as compared to PBS injection (P = .005) or RAPA (P = .004).

Figure 6.

CFSE-based proliferation analysis. (A) In vivo expansion of CFSE-labeled C57B/6 Treg (H-2kb, Thy-1.1) in Balb/c recipients (H-2kd) on day 6 after BMT is depicted. The congenic markers Thy-1.1 and Thy-1.2 were used to distinguish between Treg and TCONV cells, respectively. All Balb/c recipients (H-2kd) were given 5 × 106 C57B6 TCD-BM cells (H-2kb) after lethal irradiation with 800 cGy. Additionally 5 × 105 Treg (day 0) plus 1 × 106 CD4+/CD8+ T cells (day 2) were given (both H-2kb). Histograms show the FACS profile of CFSE+ Treg cells (H-2kbThy1.1). Numbers of events in each cell division (n) are indicated below the respective peak. Percentages of cells having undergone either 0 or 1 cell division are indicated in each histogram. Proliferation of Treg cells (first column from left) is slightly reduced in the presence of CSA (10 mg/kg; second column from left) and to a lesser extent by MMF (90 mg/kg; fourth column from left). Addition of RAPA (0.5 mg/kg; third column from left) has no impact on Treg expansion. Cells are gated on the donor marker H-2kb and the congenic marker Thy 1.1. (B) Intracellular expression of Foxp3 in donor-derived Thy1.1+H-2kb+CD4+CD25high+ T cells derived from Balb/c recipients having received TCONV (Thy-1.2) and Treg (Thy-1.1) cells alone (first column from left) in conjunction with CSA (second column from left), RAPA (third column from left), or MMF (fourth column from left). Cells were sorted, fixed onto glass slides, and analyzed by immunofluorescence. FoxP3 (Alexa 546), red; DNA (DAPI), blue; double positive, purple; magnification × 200. Percentage of FoxP3+ cells within the donor CD4+CD25+ population (average of 3 independent experiments): 95% ± 3% no immunosuppressant, 57% ± 3.2% CSA (second column from left), 94% ± 3.5% RAPA (third column from left), 89% ± 1.7% MMF (fourth column from left).

Figure 7.

GVT effector function of TCONV cells is preserved in the presence of Treg cells and RAPA. Tumor growth and elimination are visualized by bioluminescence imaging (A), FACS analysis (B), histology (C), emitted photons over time (D), and survival (E). All Balb/c recipients (H-2kd) were given 1 × 105 A20 leukemia cells (H-2kd) and BMT was performed as described in “Materials and methods.” (A) luc+A20 tumor-cell localization at 5 days (top row), 10 days (middle row), and 15 days (bottom row) after BMT in representative animals given transplants with TCD-BM alone (first column from left), TCD-BM plus TCONV cells (second column from left), TCD-BM plus TCONV and Treg cells (third column from left), and TCD-BM plus TCONV and Treg cells and RAPA (fourth column from left). ST indicates sternum; HU, humerus; FE, femur; SPL, spleen. (B) Bone marrow samples with yfp+ A20 cells of recipients on day 12 after transplantation with TCD-BM alone (first column from left), TCD-BM plus TCONV cells (second column from left), TCD-BM plus TCONV and Treg cells (third column from left), and TCD-BM plus TCONV and Treg cells and rapamycin (fourth column from left). (C) H&E stains of representative spleen samples gathered on day 12 from animals receiving TCD-BM only (Ci), TCD-BM plus TCONV cells (Cii), TCD-BM plus TCONV and Treg cells (Ciii), and TCD-BM plus TCONV and Treg cells and RAPA (Civ). A homogenous leukemic infiltration is found only in animals receiving TCD-BM only (Ci) but not in the other groups. Microscopic evaluation was performed on a Nikon Eclipse TE 300 microscope equipped with 20 ×/0.45 and 40 ×/0.60 objective lenses (Nikon, Melville, NY). Photographs were captured by using a Spot digital camera (Diagnostic Instruments, Sterling Heights, MI). Digital images were saved as TIFF files, which were inserted into and processed with PowerPoint software (Microsoft, Redmond, WA). (D) Expansion of luc+A20 tumor cell as measured in photons over total body area (photons/second/mouse). Animals rejecting the A20 tumor demonstrate a rapid signal loss within the first 15 days after BMT. (E) Survival of BALB/c mice receiving TCD-BM alone (▵, n = 15), TCD-BM plus A20 leukemia cells (□, n = 15), TCD-BM plus A20 leukemia cells plus TCONV cells (•,n = 15), TCD-BM plus A20 leukemia cells plus TCONV and Treg cells (⋄,n = 12), or TCD-BM plus A20 leukemia cells plus TCONV and Treg cells and RAPA 0.5 mg/kg; ○,n = 12). Data from 3 independent experiments are combined.