Abstract 4112 Introduction:Naturally occurring regulatory T cells (nTregs) may prevent graft-versus-host disease (GVHD) while preserving graft-versus-leukemia (GVL) activity. However, clinical application of nTregs has been severely hampered by their scarce availability and non-selective suppression. To overcome these limitations, we took an alternative approach to generate antigen-specific induced Tregs (iTregs), and tested their efficacy and selectivity in the prevention of GVHD in pre-clinical models of bone marrow transplantation (BMT). Methods:We selected HY as target antigen, because it is a naturally processed and ubiquitously expressed minor histocompatibility antigen (miHAg) with a proven role in GVHD and GVL effect. To generate HY-specific iTregs, naïve CD4+CD25− cells were isolated from MHC II-restricted HY-specific transgenic mice, and were stimulated with HY peptide and APCs, in the presence of TGFb and retinoic acid. Using similar protocol, we also generated polyclonal iTregs from CD4+CD25− cells of normal C57BL/6 (B6) mice with allogeneic dendritic cells (DCs). iTregs were isolated by positively selecting CD4+CD25hi cells 6–7 days after generation. Frequency of T cells in recognizing alloantigens was measured using a limited dilution assay. Various MHC-mismatched or matched murine BMT models were used, where polyclonal T cells (Teffs hereafter) were transplanted with donor bone marrow to induce GVHD in myeloablative allogeneic recipients. Results:We first assessed the effect of HY-specific iTregs on GVHD using an MHC II-mismatched B6 ® (B6 × bm12)F1 BMT model, and found that HY-specific iTregs prevented GVHD mortality in male (HY+) but not female (HY−) recipients. On the per-cell basis, HY-specific iTregs were significantly more potent than polyclonal Tregs in the prevention of GVHD. By measuring iTregs and Teffs in spleen and liver of the recipients, we found that HY-specific iTregs expanded extensively and significantly suppressed expansion, activation and infiltration of Teffs in male but not female recipients. To exclude the possibility the observation was model specific, we evaluated the efficacy of HY-specific iTregs and found that those iTregs were highly effective in the prevention of GVHD in two additional BMT models, including one MHC-matched but miHAg-mismatched B6 ® BALB.b model and one haplo-mismatched B6 ® B6D2F1 model. To increase translational potential of our approach, we extended our studies to alloantigen-specific polyclonal iTregs. After 7-day culture in iTreg- generating condition with BALB/c DCs, the frequency of B6 iTregs in recognizing BALB/c alloantigens was increased for 16-fold than that of naïve CD4+CD25− T cells. Furthermore, these iTregs were 64–128 fold more suppressive than nTregs to conventional T-cell response against BALB/c alloantigens. Their highly suppressive activity was antigen-specific, because the same alloreactive iTregs had significant lower suppressive activity to T-cell response against the third-party alloantigens. In vivo using an MHC-mismatched B6 ® BALB/c BMT model, we found that these alloreactive iTregs could effectively prevent GVHD at 2:1 ratio of Teff:Treg, at which polyclonal nTregs had a minimal effect. Conclusion:Using monoclonal HY-specific or polyclonal alloantigen-specific iTregs, these studies demonstrate that antigen-specific iTregs were highly effective in controlling GVHD in an activation-dependent manner. Because iTregs specific for a given miHAg (e.g. HY) can control polyclonal Teffs in response to multiple alloantigens and prevent GVHD in allogeneic recipient, these data also indicate that Tregs may control GVHD through a linked-suppression on Teffs in vivo. In conclusion, the current study presents a promising strategy to generate alloantigen-specific iTregs for effective GVHD prevention in human allogeneic hematopoietic cell transplantation. Disclosures:No relevant conflicts of interest to declare.