Extrinsic factors within the host environment are crucial in determining recruitment of graft-versus-host (GVH)-reactive T cells to peripheral tissues and the capacity of these cells to induce graft-versus-host disease (GVHD). In this study, we have examined how the host environment influences graft-versus-leukemia (GVL) activity. Transfer of small numbers of allogeneic T cells to freshly irradiated (TBI allo) mice induces both GVL and GVHD, whereas transfer of much higher numbers of T cells to established (>8 weeks) mixed chimeras (MC) can induce GVL without GVHD. Using an EL4 tumor protection assay and low doses of B10.A splenocytes (3 × 10e6, a dose 1 log lower than normally transferred to MC to induce GVL), we observed that tumor-free survival of recipient mice was greater following transfer to TBI allo B6 recipients than in B6 + B10.A → B6 MC. To determine the mechanisms for this disparity, we used a BALB/c recipient/B6 donor strain combination, in which we could track the distribution of donor T cells within secondary lymphoid organs and bone marrow (BM) following transfer to TBI allo or MC recipients. Despite similar expansions within the spleen, there was a significant delay in the accumulation of polyclonal donor T cells (B6 CD45.1) and transgenic 2C GVH-reactive CD8+ T-cells (bearing TCR specific for recipient antigen) in the BM of MC compared to TBI allo mice. Moreover, in vivo cytotoxicity of host B cell targets occurred rapidly and was virtually complete in TBI allo recipients, but was absent in MC even at late time points. To evaluate the acquisition of effector functions in a clonal GVH-reactive CD8+ T cell population, we sorted 2C T cells at intervals following transfer and performed quantitative RT-PCR of molecules linked to effector differentiation. Strikingly, transcription of IFN-γ, granzyme B and TNF-α was significantly higher in cells derived from TBI allo hosts compared to MC. Similar differences in IFN-γ and granzyme B protein expression were confirmed in the polyclonal donor CD8+ T cell population. Since, full GVL activity might also depend upon the survival of GVH-reactive CTL, we also examined the viability of donor T cells during the initial response in both environments. We observed higher rates of sustained 2C CD8+ T cell apoptosis (as indicated by annexin V staining) following T cell transfer to MC. Furthermore, we also detected lower expression of common γ chain cytokine receptors that mediate responsiveness to IL-2, IL-7 and IL-15, upon donor T cells from MC. However, following secondary co-transfer to syngeneic recipients for 21 days, memory phenotype polyclonal donor CD8+ T cells derived from established MC (CD45.1+) were recovered to a greater extent than T cells initially derived from TBI allo mice (Thy1.1+), arguing against any intrinsic defect in the viability of GVH-reactive T cell populations emerging in the former setting. Indeed, when co-cultured in the presence of individual cytokines, MC-derived CD8+ T cells maintained viability to a greater (IL-2, IL-7) or equivalent (IL-15) extent as TBI allo-derived cells. Taken together, these data suggest that disparity in GVL activity following to TBI allo recipients and MC can be explained by differences in the effector function and survival of anti-host CTL in quiescent MC environment. Reduced GVL activity of donor T cells on a per-cell basis in MC can be compensated for by transferring greater numbers of cells.

Author notes

Disclosure: No relevant conflicts of interest to declare.