Myeloid Suppressive Cells Mobilized by GM-CSF in Non-Tumor Bearing Mice Are Dependent On Interferon Gamma for Function

Abstract 832

Myeloid-derived-suppressor cells (MDSCs) are enriched in tumors, and exist to a lesser extent in the blood, spleen and bone marrow of tumor bearing mice. Monocytic MDSCs (monoMDSCs) suppress CD8+ T-cells via expression of Arginase 1 (ARG1) and inducible nitric oxide synthase (iNOS). Tumor derived factors are critical to the maintenance of MDSCs and preventing differentiation to mature macrophages and dendritic cells. GM-CSF is a hematopoietic cytokine that can be secreted by tumors and promotes MDSC generation. Cells phenotypically similar to MDSCs can be isolated from blood of normal individuals but lack suppressive function. Hematopoietic peripheral blood stem cell mobilization with G-CSF and GM-CSF enriches for cells phenotypically similar to MDSCs. There is limited data on the role and function of these cells isolated from non-tumor bearing, normal individuals. Recent evidence suggests that graft-versus-host disease (GvHD) can be abrogated in mice by ex vivo expanded, bone marrow derived, MDSCs generated in the presence of GM-CSF, G-CSF and IL-13 (Highfill et al. Blood 2010 116:5738). It remains to be shown whether phenotypic MDSCs identified in non-tumor bearing mice are capable of immune suppression; and, the mechanism by which an immature myeloid cell becomes a functional MDSC remains unknown. We have observed an increase (up to 8 fold) in a population of cells phenotypically resembling monoMDSCs (CD11b+/Ly6C+/Ly6G-) in the spleens and blood of mice mobilized with pegylated-murine-GM-CSF (peg-mGM-CSF). We hypothesized that this population of cells would have suppressive function similar to MDSCs in vitro and in vivo, and may have the potential to abrogate graft-versus-host disease (GvHD). To investigate the function of MDSCs found in spleens of C57/Bl6 (B6) mice treated with peg-mGM-CSF we performed CFSE based anti-CD3/CD28 antibody stimulated T-cell proliferation assays, mixed leukocyte reactions and transwell assays. We observed that CD11b+Ly6C+Ly6G- cells isolated from spleens of mice treated with peg-mGM-CSF have potent suppressive function in vitro that is contact dependent and abrogated by blocking ARG1 or iNOS. This suppressive effect was lost in APC stimulated MLRs using B6 T-cells and Balb/C stimulators (confirmed in two separate experiments). Furthermore, the in vivo potential of these putative MDSCs to abrogate murine GvHD was investigated using a B6 to Balb/C donor leukocyte infusion GvHD model. Adoptive transfer of purified splenic CD11b+Ly6C+Ly6G- cells from peg-mGM-CSF mobilized B6 donors along with an equivalent number of congenic T-cells failed to abrogate GvHD. We investigated timing of MDSC infusion in the B6 to Balb/C GvHD model and found no improvement in weight loss, GvHD score or survival in mice receiving 5×10⁵ monoMDSCs IV on day 1, 6 or 10 after transplant compared to T-cells alone control (n = 5 – 10/group, Log rank, p= NS). To address in vivo function further in a bioluminescent imaging (BLI) tumor model. Balb/C recipients were injected SC with A20 cells mixed +/− monoMDSCs at a 1:10 ratio after lethal irradiation and T-cell deplete bone marrow on day 0. Donor T-cells were infused at day +11. The rate of tumor growth measured by photon flux was the same between subcutaneous tumors either with or without monoMDSCs. (two separate experiments, 5 mice/group). This in vivo data suggested that a critical factor present in vitro might be lacking or insufficient in vivo. To investigate the critical factor(s) present in vitro we performed T-cell proliferation assays in the presence of blocking antibodies against IFNy, TNFalpha, IL-10, GM-CSF and CD154. Only neutralization of IFNy resulted in negation of the suppressive effects of these cells. To investigate the source of IFNy production we used transgenic IFNy knockout mice as T-cell and MDSC donors. Proliferation of IFNy deficient T-cells was suppressed efficiently by wild-type (WT) MDSCs, and, neutralizing IFNy using a blocking antibody negated suppression. This suggested IFNy production by a cell within the putative MDSC sorted population might be critical for MDSC function. IFNy deficient peg-mGM-CSF mobilized CD11b+Ly6C+Ly6G- spleen cells failed to suppress WT or IFNy deficient T-cell proliferation. These results suggest a critical role for IFNy production by CD11b+Ly6C+Ly6G- myeloid cells in maintaining their suppressive phenotype in vitro and perhaps in vivo.