TGF-β Enhances Immunosuppression of Myeloid-Derived Suppressor Cells to Induce Transplant Immune Tolerance Through Affecting Arg-1 Expression

Abstract

Myeloid-derived suppressor cells (MDSCs) are a class of heterogeneous myeloid cells, which play an important role in immunosuppression. We intended to find an effective method that can produce MDSCs with significantly better efficiency and promote immune tolerance for transplant rejection through cell therapy. It has been reported that granulocyte and macrophage colony-stimulating factor (GM-CSF) could induce MDSCs in vitro to cause immunosuppression. In the present study, transforming growth factor β (TGF-β) was added to the induction system, and flow cytometry analysis was used to detect the phenotypes of induced MDSCs. Their potential immunosuppressive function and mechanisms were determined by co-culturing MDSCs with stimulated T cells in vitro and transferring MDSCs to the skin grafted C57BL/6J mouse models in vivo. It was found that the addition of TGF-β could effectively cause bone marrow cells to differentiate into a group of cells with stronger immunosuppressive functions, thereby inhibiting the proliferation of stimulated T cells. The population of CD11b⁺Gr-1⁺ MDSCs also increased significantly as compared with GM-CSF alone treatment. While detecting for immunosuppressive effectors, we found that expression of arginase 1 (Arg-1) was significantly upregulated in these MDSCs, and inhibitor of Arg-1 significantly suppressed their immunosuppressive capabilities. Moreover, an adoptive transfer of these cells significantly prolonged survival of allo-skin and improved immune tolerance in vivo. These findings indicated that TGF-β + GM-CSF could serve as an effective and feasible method to induce powerful immunosuppressive MDSCs in vitro. Thus, TGF-β + GM-CSF-induced MDSCs may have a promising role in prevention of the graft rejection.

Keywords: ARG-1; Immune rejection; TGF-β; immune tolerance; myeloid derived suppressor cells (MDSCs); transplant.

Figure 1

The characteristics of TGF-β + GM-CSF–induced MDSCs. (A) The precursor cells from the bone marrow cells were cultured in GM-CSF alone or TGF-β + GM-CSF as described in Materials and Methods for 4 days. The percentages of CD11b⁺Gr-1⁺ MDSCs have been shown by FCM analysis. (B) The proportions and numbers of CD11b⁺Gr-1⁺ MDSCs that were induced by either GM-CSF alone or TGF-β + GM-CSF. (C) FCM analysis showed the typical phenotypes of the cells stained by FITC-, PE-, PE-Cy5–, or APC-labeled anti-mouse-CD11b, Gr-1, Ly6C, Ly6G, F4/80, CD11c, CD80, CD86, I-Ab, CD31, CD40, CD62L, CD115, CD124, and CD274 Abs. (D) The proportions of the indicated markers have been summarized. The data have been shown as mean ± SEM (n = 3) and were analyzed by an unpaired two-tailed Student’s t-test, which were collected from three independent experiments. P < 0.05 was considered as statistically significant between the groups. *P < 0.05, **P < 0.01, and ***P < 0.001 upon comparison between the two groups.

Figure 2

TGF-β + GM-CSF–induced MDSCs exhibited significant immunosuppressive functions. GM-CSF and TGF-β + GM-CSF–induced MDSCs were added to the stimulated T cells proliferation system at different concentrations for 3 days. (A) Typical samples of FCM analysis for T-cell proliferation of CD4⁺ (T cells: MDSCs = 1:1/2). (B) Percentage of CD4⁺ T-cell proliferation was measured by CFSE dye dilution co-cultured with induced MDSCs at indicated ratios. (C) Typical samples of FCM analysis for T-cell proliferation of CD8⁺ (T cells: MDSCs = 1:1/2). (D) Percentage of CD4⁺ T-cell proliferation was measured by CFSE dye dilution co-cultured with induced MDSCs at indicated ratios. (E) The levels of inflammatory cytokines, IL-10, IL-4, TNF-α, and IFN-γ in the co-culture supernatant were detected by ELISA assay (T cells: MDSCs = 1:1/2). Data have been shown as mean ± SEM (n = 3) and were analyzed by an unpaired two-tailed Student’s t-test and one-way ANOVA analysis, which were collected from three independent experiments. *P < 0.05, **P < 0.01, and ***P < 0.001 upon comparison between the two groups.

Figure 3

TGF-β + GM-CSF–induced MDSCs can effectively suppress cytokine production in the stimulated T cells and cause an increase in the number of Tregs. GM-CSF–induced and TGF-β + GM-CSF–induced MDSCs co-cultured with the stimulated T cell (stimulated cell–to–MDSC ratio = 1:1/2) for 3 days followed by adding PMA and ionomycin for stimulation. Typical samples of FCM analysis conducted for levels of the cytokine produced by the T cells in (A) and (B). The percentages of CD4⁺and CD8⁺ T cells producing IL-2 and IFN-γ (C) and TNF-α and IFN-γ (D), respectively, with or without induced MDSCs in the culture system have been summarized. (E) Typical samples of FCM analysis for CD4⁺Foxp3⁺ Tregs in the culture system with or without the induced MDSCs. (F, G) The percentages and numbers of Tregs in the co-culture have been summarized. The data have been shown as mean ± SEM (n = 3) and were analyzed by one-way ANOVA analysis, which were collected from three independent experiments. *P < 0.05, **P < 0.01, and ***P < 0.001 upon comparison between the different groups. ns, no significance.

Figure 4

TGF-β + GM-CSF–induced MDSCs exert immunosuppressive functions through the Arg-1 pathway. (A) The mRNA expression levels of iNOS, Arg-1, OH-1, IDO, IL-10, TGF-β, COX2, and NOX2 in GM-CSF and TGF-β + GM-CSF–induced MDSCs were determined by real-time qPCR assay. Typical dot plot of FCM analysis conducted for levels of iNOS (B) and Arg-1 (C) produced by the induced MDSCs. The percentages of CD11b⁺iNOS⁺ and CD11b⁺Arg-1⁺ cells in the GM-CSF– and TGF-β + GM-CSF–induced system have been summarized. (D) GM-CSF– and TGF-β + GM-CSF–induced MDSCs in different concentrations co-cultured with stimulated T cells and the NO levels in the culture medium were detected. (E) The activity of arginine in the co-culture system (MDSC–to–stimulated cell ratio = 1/2:1 and 1/8:1) was detected. Administration of L-arginine (F), iNOS inhibitor L-NMMA (G), and Arg-1 inhibitor nor-NOHA (H), respectively, in co-culture system to analyze the potential inhibitory effects of GM-CSF+TGF-β + GM-CSF–induced MDSCs on the stimulated T-cell proliferation (MDSC–to–stimulated cell ratio = 1/2:1). The data have been shown as mean ± SEM (n = 3) and were analyzed by an unpaired two-tailed Student’s t-test, which were collected from three independent experiments. *P < 0.05, **P < 0.01, and ***P < 0.001 compared between the different groups.

Figure 5

Adoptive transfer of TGF-β + GM-CSF–induced MDSCs could effectively prolong the survival of allo-skin. The recipient female B6 mice were injected intravenously (i.v.) with induced MDSCs (5 × 10⁶) on day 1 before operation and day 7 after operation. The mice in the control group were injected i.v. with vehicle PBS. In addition, TGF-β + GM-CSF–induced MDSCs and Arg-1 inhibitor, nor-NOHA, was administered to another group. (A) The photographs of the graft rejection have been shown in the four different groups. (B) Representative HE staining of skin grafts on day 21 has been shown (400×). (C) Quantification of H&E staining corresponding to the four groups on day 21. (D) The survival time of the graft after transplantation have been summarized (n = 6). P values were determined by log-rank tests. *P < 0.05, **P < 0.01, and ***P < 0.001 compared between the different groups. (E) The levels of the donor-specific antibodies (DSAs), IgG1, IgG2a, and IgG2b were detected by FCM analysis. The data have been shown as mean ± SEM (n = 3) and were analyzed by one-way ANOVA analysis. *P < 0.05, **P < 0.01, and ***P < 0.001 compared between the different groups.

Figure 6

Adoptive transfer of TGF-β + GM-CSF–induced MDSCs could significantly promote the expansion of MDSCs and Tregs in the transplant recipients. Allo-skin mice were divided into four different groups and then treated with PBS, GM-CSF–induced MDSCs, TGF-β + GM-CSF–induced MDSCs, and TGF-β + GM-CSF–induced MDSCs with Arg-1 inhibitor, respectively. Thereafter, allografts, peripheral blood, and spleen of transplant recipients were obtained on day 21 after operation for FCM assay. Typical examples of FCM analysis and proportion of CD11b⁺Gr-1⁺ MDSCs in allografts (A), peripheral blood (B), and spleen (C) of the transplanted mice. Typical examples of flow cytometry analysis and percentages of CD4⁺Foxp3⁺ Tregs in allografts (D), peripheral blood (E), and spleen (F) of the transplant mice. The data have been shown as mean ± SEM (n = 3) and were analyzed by one-way ANOVA analysis. *P < 0.05, **P < 0.01, and ***P < 0.001 compared between the different groups.