Activation Status of Residual Host Dendritic Cells (DCs) but Not Their Number Governs Donor Lymphocyte Infusion (DLI)-Mediated Responses after Major Histocompatibility Complex (MHC)-Matched AlloBMT.

The role of host antigen presenting cells (APCs), particularly DCs in governing DLI-mediated alloresponses post MHC-matched alloBMT is unclear. To evaluate whether the amount of residual host-derived APCs determines the alloresponse of DLI we constructed MHC-matched B10.D2 (H2^d)→BALB/c (H2^d) chimeras with varying degrees of residual host APCs. After 8 weeks of rest, chimeras received DLI in the form of 2 x 10⁷ splenocytes from B10.D2 (Thy1.1⁺) mice that differ from BM donors by the expression of Thy 1.1⁺ marker on the lymphocytes. The differential expression of the Thy 1.1⁺ marker on DLI-derived T cells allowed us to precisely monitor their alloresponses in vivo. We found that in the MHC-matched model of alloBMT, contrary to the MHC-mismatched model, the amount of residual host APCs does not influence DLI-mediated T-cell alloresponses measured as the alloantigen-driven expansion of DLI-derived T cells and changes in donor chimerism. However, the early administration of DLI, 4 weeks after the conditioning resulted in statistically significant expansion of DLI-derived CD8⁺ and CD4⁺ T cells (p<0.01). Since these data suggest the importance of timing of DLI administration on the DLI-mediated alloresponses in MHC-matched setting, we next evaluated whether the augmented alloresponses may be reflective of kinetics of donor DC reconstitution, changes in DC populations or their qualitative characteristics at different time points post-transplant. To test these variables we used C3H.SW (H-2K^b, CD45.2⁺)→B6Ly5.2 (H-2K^b, CD45.1⁺) MHC-matched model that enabled us to dissociate the origin of DCs at different time points post-transplant based on the differential expression of CD45 marker on all host and donor hematopoietic tissues. We found that host to donor DC turnover is rapid, results in near full conversion to donor DC chimerism in the spleen (>99 %) and mesenteric lymph nodes (LN) (>99%), but not in the subcutaneous LNs ( 90%). Comprehensive phenotypic analysis showed that donor-derived DC populations in the spleen and LNs do not differ in the chimeras early or late after conditioning, however, residual host-derived DCs express significantly higher levels of costimulatory molecules, CD80, CD40 and especially CD86 at early time points after conditioning (3 weeks vs 8 weeks). These finding suggest that residual host DC activation status may be the critical factor influencing the potency of DLI-mediated responses. To further investigate the role of DC activation status, we co-administered DLI with ex vivo-matured DCs or systemically activated them with agonistic anti-CD40 antibody or CpG immunostimulatory oligonucleotides that stimulate Toll-like receptor (TLR)-9. While both approaches augmented the in vivo proliferation of DLI-derived T cells, only CpGs augmented effector function of DLI-derived T cells, particularly the production of interferon-γ (p<0.005) and conversion to full donor chimerism. Additionally, in the post-transplant C1498 leukemia relapse model survival of C3H.SW→C57BL/6 chimeras that received DLI and CpG immunostimulatory molecules was superior to the chimeras that received DLI only or DLI + ex vivo-matured host DCs (10/10 vs 2/10 vs 2/10; p<0.01). These data indicate that DLI-mediated alloresponses in the setting of MHC-matched alloBMT are critically influenced by the activation status of residual host DCs and can be further manipulated by in vivo administering TLR-9 ligands but not by administering ex vivo-matured host DCs.