Host Environment Determines the Development of Gvhd but Not the Extent of Activation and Expansion of Gvh-Reactive CD4 T Cells

Transfer of small numbers of naïve allogeneic T cells into freshly irradiated (FI) mice leads to severe, lethal GVHD. In contrast, when large numbers of naïve donor T cells are transferred as delayed donor lymphocyte infusions (DLI) given to established mixed hematopoietic chimeras (MC), a GVH reaction limited to the lymphohematopoietic system (LHGVHR) occurs and does not lead to GVHD. A similar phenomenon can occur when DLI are given to patients in whom mixed chimerism is established with non-myeloablative conditioning. LHGVHR without GVHD across MHC barriers provides an approach to separating GVL and GVHD and an understanding of this phenomenon will be essential to optimize its clinical application. We have previously shown that directly GVH-alloreactive CD8 T cells injected into FI mice or in delayed DLI given to MC show rapid proliferation and accumulation in lymphoid organs and similar modification of homing receptors, but with different kinetics in each type of recipient. However, GVH-reactive CD8 T cells traffic to and accumulate in the epithelial GVHD target tissues only in FI mice and not in MC. In order to determine the influence of the host environment on CD4-mediated GVH responses, we compared the expansion and activation of GVH-reactive CD4 T cells specific for recipient minor (mHA) and MHC antigens in the presence of a full haplotype mismatch (B6→B6D2F1) in MC and FI mice. With a non-myeloablative protocol, we established B6→B6D2F1 MC in transgenic B6D2F1 recipients expressing membrane-bound chicken ovalbumin (OVA-B6D2F1). Six weeks later, a group of MC and a group of OVA-B6D2F1 mice lethally irradiated on the same day received 10⁶ OTII CD4 T cells (specific for OVA peptide 323-339/IA^b, representing a mHA) along with 10⁷ polyclonal B6 splenocytes to preserve the strong polyclonal response. FI mice also received B6 bone marrow cells (BMC) to prevent death from aplasia. Activation of OTII CD4 T cells was similar in both MC and FI mice, with no statistically significant differences in the increase in CD44 (average median fluorescence intensity [MFI] 245 and 235, respectively), %CD69+ (average 52% in MC and 36% in FI) and %CD25+ cells (average 18% in MC and 20% in FI) and downregulation of CD62L (average 18% positive in MC and 10% in FI). These changes occurred as early as day 3 in FI mice and were delayed to day 6 in MC. On day 6, OTII CD4 T cells accumulated to a 5-fold greater extent in spleens of MC compared to FI mice (mean±SD of 1.47±0.12 × 10⁶in MC vs 0.30±0.03×10⁶in FI; p<0.001). FI mice died of GVHD by Day 9, whereas GVHD did not develop in MC. Similar studies were performed with directly class II MHC alloreactive CD4 T cells. Six weeks after establishing B6→B6D2F1 MC, a DLI consisting of 10⁶ OTII CD4 T cells (non-specific T cells in this model because they do not recognize any recipient antigen), 10⁶ 4C TCR transgenic CD4 T cells (directly alloreactive for recipient IA^d class-II antigen) and 10⁷ polyclonal B6 spleen cells was injected into MC and FI B6D2F1 mice that also received B6 BMC. On Days 6 and 9, greater accumulation of 4C CD4 T cells compared to OTII CD4 T cells (p<0.05) was detected in the spleens of both groups. Expression of activation markers on 4C cells occurred in both FI mice and MC, with similar MFI for CD44 (mean 120 in FI mice and 198 in MC; NS), similar %CD69+ (average 54.9% positive in FI and 60.2% in MC; NS) and %CD25+ cells (average 19.6% positive in FI mice and 11.2% in MC; NS). Downregulation of CD62L was also similar but slightly delayed in MC. By day 16, FI mice developed significant weight loss and GVHD, whereas MC did not develop GVHD. These data confirm the influence of the host environment in the development of GVHD and show that, similar to GVH-reactive CD8 T cells, marked antigen-driven activation and proliferation of GVH-reactive CD4 T cells occurs in both FI mice and MC, although with different kinetics.