Bone marrow myeloid-derived suppressor cells (MDSCs) inhibit graft-versus-host disease (GVHD) via an arginase-1-dependent mechanism that is up-regulated by interleukin-13

Abstract

Myeloid-derived suppressor cells (MDSCs) are a well-defined population of cells that accumulate in the tissue of tumor-bearing animals and are known to inhibit immune responses. Within 4 days, bone marrow cells cultured in granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor resulted in the generation of CD11b(+)Ly6G(lo)Ly6C(+) MDSCs, the majority of which are interleukin-4Rα (IL-4Rα(+)) and F4/80(+). Such MDSCs potently inhibited in vitro allogeneic T-cell responses. Suppression was dependent on L-arginine depletion by arginase-1 activity. Exogenous IL-13 produced an MDSC subset (MDSC-IL-13) that was more potently suppressive and resulted in arginase-1 up-regulation. Suppression was reversed with an arginase inhibitor or on the addition of excess L-arginine to the culture. Although both MDSCs and MDSC-IL-13 inhibited graft-versus-host disease (GVHD) lethality, MDSC-IL-13 were more effective. MDSC-IL-13 migrated to sites of allopriming. GVHD inhibition was associated with limited donor T-cell proliferation, activation, and proinflammatory cytokine production. GVHD inhibition was reduced when arginase-1-deficient MDSC-IL-13 were used. MDSC-IL-13 did not reduce the graft-versus-leukemia effect of donor T cells. In vivo administration of a pegylated form of human arginase-1 (PEG-arg1) resulted in L-arginine depletion and significant GVHD reduction. MDSC-IL-13 and pegylated form of human arginase-1 represent novel strategies to prevent GVHD that can be clinically translated.

Figure 1

Generation and phenotype of cultured MDSCs. (A) Fluorescence-activated cell sorter phenotype of B6 whole BM at day 4 of culture that has been left untreated, treated with G-CSF alone (100 ng/mL), treated with GM-CSF alone (250 U/mL), or treated with both G-CSF and GM-CSF. All gates based on isotype controls. Ly6G/Ly6C graphs were first gated on CD11b⁺ cells. Representative images of hematoxylin and eosin-stained cytospins of MDSCs with combined treatment (B) and combined treatment plus IL-13 addition on day 3 (C) Photographs were taken using 400× magnification with a RT-Spot camera mounted on an Olympus BX51 microscope (Olympus).

Figure 2

MDSCs inhibit T-cell alloresponses through the expression of arginase-1. (A) RT-PCR showing that MDSCs express arginase-1 and iNOS and up-regulate arginase-1 on stimulation with IL-13. Hypoxanthine phosphoribosyl transferase shown as endogenous control. M indicates marker. (B) Real-time RT-PCR showing quantitatively the extent of arginase-1 and iNOS up-regulation by IL-13-stimulated MDSCs. Standardized to glyceraldehyde-3-phosphate dehydrogenase endogenous control and relative expression compared with untreated BM. (C) Western blot showing arginase-1 and actin control at the protein level in untreated BM, MDSCs, and MDSC IL-13. (D) Arginase activity was determined by measuring the production of urea over time. MLR was performed by mixing B6 purified T cells with irradiated BALB/c stimulators (1:1 ratio) and MDSCs (1:10 ratio). Cultures were pulsed with ³H-thymidine on the indicated days and harvested after a 16-hour incubation. (E) MLR/MDSC cocultures were left untreated or were treated with arginase inhibitor, nor-NOHA (300μM), nitric oxide inhibitor, L-NMMA (300μM), or both. (F) Arginase inhibitor (nor-NOHA) was added to MLR/MDSC cocultures at increasing concentrations (30, 100, and 300μM). (G) B6 WT T cells or GCN2 KO T cells were used as responder cells, and MDSCs were added at 1:10 ratio. (H) Addition of excess L-arginine (5mM) or tryptophan (5mM) was added back to MLR/MDSC cocultures and assessed for T-cell proliferation.

Figure 3

MDSC-IL-13 migrate to sites of allopriming in a GVHD setting. (A) Fluorescence-activated cell sorter analysis of pretransplantation MDSC-IL-13 was gated on Gr1⁺CD11b⁺ and analyzed for the surface expression of molecules associated with homing and recruitment. Gates are based on isotype controls. (B) Lethally irradiated BALB/c recipients were given 10⁷ BM cells plus 2 × 10⁶ T cells with 6 × 10⁶ eGFP transgenic MDSC-IL-13. Control mice were given BM and T cells only. On days 7 and 14, LN and spleen were harvested and imaged using macroscopic fluorescent imaging. (C) The absolute number of eGFP⁺ cells that had migrated to the LNs and spleen was determined using flow cytometry. (D) Spleen was harvested from day 7 transplanted mice, and flow cytometry was performed to determine whether MDSC-IL-13 retained Gr1⁺CD11b⁺ phenotype. Cells were gated on donor (GFP⁺) CD11b⁺ cells.

Figure 4

Cultured MDSC-IL-13 enhance GVHD survival in an arginase-1-dependent fashion. Lethally irradiated BALB/c recipients were given 10⁷ B6 BM cells plus 2 × 10⁶ purified CD25-depleted T cells plus either 2 × 10⁶ MDSC-IL-13 or 6 × 10⁶ MDSC-IL-13, both IL-13-treated. Kaplan-Meier survival curve of transplanted mice (A). Data represent 2 pooled experiments (BM only, N = 18; BM + T, N = 18; BM + T + 2 × 10⁶ MDSC IL-13-tx, N = 22; BM + T + 6 × 10⁶ MDSC- IL-13, N = 22; BM + T vs 2 × 10⁶ MDSC IL-13, P < .001; BM + T vs 6 × 10⁶ MDSC IL-13, P < .001; 2 × 10⁶ MDSC IL-13 vs 6 × 10⁶ MDSC- IL-13, P = .06). (B) Corresponding weights. (C) Kaplan-Meier survival curve of BM transplantation using 6 × 10⁶ MDSC IL-13 generated from either WT mice or arginase-1 KO mice. Data represent one experiment (BM only, N = 8; BM + T, N = 12; BM + T + 6 × 10⁶ WT MDSC-IL-13, N = 12; BM + T + 6 × 10⁶ arginase-I KO MDSC-IL-13, N = 12; BM + T vs 6 × 10⁶ WT MDSCs, P = .05; BM + T vs 6 × 10⁶ arginase-I KO MDSC-IL-13, P = .08; arginase-1 KO MDSC-IL-13 vs WT MDSC-IL-13, P = .1). (D) Corresponding weights.

Figure 5

MDSC-IL-13 negatively impact proliferation, activation, and effector function of donor T cells. Lethally irradiated BALB/c mice were transplanted with 15 × 10⁶ CFSE-labeled CD25-depleted B6 Ly5.1-expressing T cells alone or with 10 × 10⁶ CD11b purified MDSC-IL-13. (A) Day 4 splenocytes were analyzed via flow cytometry for CFSE dilution on CD4 and CD8 T cells. (B) Flow cytometry was used to detect the amount of CD3ζ present on CD4 and CD8 splenic T cells. (C) Graph of common activation markers (CD25 and CD62L) on CD4 and CD8 splenic T cells. (D) Intracellular cytokine staining was performed on LN CD4 and CD8 T cells to determine the percentage of cells producing IFN-γ.

Figure 6

MDSCs preserve GVL effect of allogeneic donor T cells. Lethally irradiated BALB/c mice were given 10 × 10⁶ B6 T cell–depleted BM cells alone or with 2 × 10⁶ CD25-depleted B6 T cells. Cohorts of mice also received 3 × 10⁵ A20 cells and/or 6 × 10⁶ B6 MDSC-IL-13. All cells were given intravenously on day 0 of transplantation. (A) Kaplan-Meier survival curve of mice receiving BM only (N = 10), BM + T (N = 10), or BM + T + MDSC-IL-13 (N = 10). (B) Corresponding weights from these mice. (C) Kaplan-Meier survival curve of mice receiving BM + A20 (N = 10), BM + T + A20 (N = 10), or BM + T + A20 + MDSC-IL-13 (N = 10). (D) Corresponding weights. (E) Mice were monitored on days 14, 18, and 50 using bioluminescent imaging to detect Renilla luciferase-expressing A20 cells. BM + T versus BM + T + MDSCs, P < .001; BM + A20 versus BM + T + A20, P < .001; BM + T + A20 versus BM + T + A20 + MDSCs, P < .001. Data represent one experiment.

Figure 7

PEG-arg1 has similar protective effects as MDSC IL-13-tx. Lethally irradiated BALB/c mice were given 10 × 10⁶ B6 BM cells alone or with 2 × 10⁶ CD25-depleted T cells. In addition to this, cohorts were also given 6 × 10⁶ MDSC IL-13 on day 0 or PEG-arg1 at 1 mg/mouse, 2 times weekly. (A) Kaplan-Meier survival curve. (B) Corresponding weights. P values for survival are BM + T versus 6 × 10⁶ MDSCs, P = .03; BM + T versus PEG-arg1, P = .003. (C) L-Arginine was quantified using high performance liquid chromatography on day 14 from the peripheral blood of transplanted mice. WT B6 mice were used as controls. Data are representative of 2 replicate experiments with similar results.