Abstract

Abstract 241

Background:

Myeloid-derived suppressor cells (MDSC) are a well-defined population of cells that accumulate in the tissue of tumor-bearing animals and are known to inhibit immune responses. Here, we have developed a novel method by which we can generate MDSC from the BM of wild type mice. Our goals were to determine the effectiveness of these cells in inhibiting an allogeneic T-cell reaction, the mechanism by which this occurs, and the impact on the graft-versus-tumor (GVT) activity of donor T-cells.

Results:

Our results show that the incubation of WT BM with G-CSF and GM-CSF for a period of 4 days results in a population of cells predominately co-expressing CD11b and Gr1 (50%). Most of these cells have a monocytic phenotype of Ly6GloLy6C+ (73%) and also express other markers associated with MDSCs such as IL4Ra (64%), F4/80 (63%) and CD115 (43%). Such MDSC potently inhibited in vitro allogeneic T-cell responses and the addition of IL-13 to the MDSC culture enhanced their suppressive capacity (60% suppression by MDSC at the peak of the response, day 5 vs. 75% suppression by MDSC-IL13, P<0.001). Suppression was dependent on L-arginine depletion because only the addition excess L-arginine, but not tryptophan, resulted in a significant reduction in suppression (60% reduction at day 4, P=0.001). Adding IL-13 to the MDSC cultures resulted in arginase-1 upregulation (20 fold increase in expression over control BM for MDSC vs. 350 fold increase in MDSC-IL13, P<0.001). The arginase inhibitor, nor-NOHA, significantly reduced MDSC suppression in vitro. GNC2 kinase knockout T cells were resistant to suppression, consistent with their inability to sense L-arginine depletion. Although iNOS was upregulated 5-fold in MDSCs vs BM, IL13 addition did not further upregulate iNOS. Suppression was mostly contact-independent. In vivo, MDSC-IL13 derived from GFP transgenic BM migrated to sites of allopriming. MDSC-IL13 from luciferase transgenic donors increased over a 3 week period of time as assessed by bioluminescent imaging. Although both MDSC and MDSC-IL13 inhibited GVHD lethality, MDSC-IL13 were more effective at enhancing survival (MDSC vs. MDSC-IL13, P=0.001). When compared with untreated mice, GVHD inhibition in MDSC-IL13 treated mice was associated with: 1) Limited donor T-cell proliferation (CD4+CFSE, 40% vs. 17%, P=0.01; CD8+CFSE, 51% vs. 16%, P=0.01); 2) Decreased donor T-cell activation (CD4+CD62L, 74% vs. 50%, P<0.001; CD8+CD62L, 77% vs. 41%, P<0.001); 3) Decreased proinflammatory cytokine production (CD4+IFNg+, 37% vs. 24%, P=0.04; CD8+IFNg+, 30% vs. 15%, P=0.007); 4) Decreased expression of intracellular CD3z chain (CD4+CD3z+, 86% vs. 56%, P<0.001; CD8+CD3z+, 85% vs. 68%, P<0.001). On day 14, donor MDSC-IL13 cells, distinguishable from donor BM and the host by a CD45 congenic marker, were re-isolated from the spleen of murine GVHD recipients. MDSC-IL13 cells retained their initial phenotype and were potent ex vivo suppressors of an allogeneic T cell response. Arginase-1 knockout MDSC-IL13 did not have a significant effect on reducing GVHD, indicating that arginase-1 expression was vital to the in vivo suppressive effect of MDSCs. MDSC-IL13 did not abrogate the graft-versus-tumor effect of donor T-cells. In MDSC-IL13 treated mice, donor T-cells retained their ability to eliminate A20 lymphoma cells, and at the same time, had improved survival when compared to mice receiving A20 cells plus T-cells all of which succumbed to GVHD (BM+T+A20 vs. BM+T+A20+MDSC-IL13, P<0.001). Since arginase-1 expression was critical for suppression, we exploited this mechanism of action by administering a pegylated form of human arginase-1 (PEG-arg1) to systemically deplete L-arginine. We find that GVHD survival is significantly improved (BM+T vs. BM+T+PEG-arg1, P=0.003).

Conclusions:

We can conclude that BM-derived MDSC-IL13 have the ability to dampen GVHD and enhance survival when using a fully-mismatched murine model of BMT. Arginase-1 expression was found to play a critical role in MDSC-mediated suppression. A GVT effect was not ablated by MDSC-IL13. In vivo administration of PEG-arg1 resulted in L-arginine depletion and significant GVHD reduction. Both MDSC infusion and PEG-arg1 administration are promising strategies that warrant further preclinical studies to prevent GVHD that may be readily translatable in the clinic.

Disclosures:

Highfill:UIniversity of Minnesota: Patents & Royalties; Louisiana State University: Patents & Royalties; Medical College of Georgia: Patents & Royalties. Rodriguez:University of Minnesota: Patents & Royalties; Louisiana State University: Patents & Royalties; Medical College of Georgia: Patents & Royalties. Ochoa:University of Minnesota: Patents & Royalties; Louisiana State University: Patents & Royalties; Medical College of Georgia: Patents & Royalties. Blazar:University of Minnesota: Patents & Royalties; Louisiana State University: Patents & Royalties; Medical College of Georgia: Patents & Royalties.

Author notes

*

Asterisk with author names denotes non-ASH members.