CD8⁺DCs are required for optimal GVT effect. Host CD8⁺ DCs are able to increase GVT effect in allo-HCT. β2m^−/− animals received 10 Gy on day −1 and 0.5 × 10⁶ CD8⁺ T cells along with 5 × 10⁶ TCD-BM from allogeneic C3H.sw donors; syngeneic 2 × 10⁴ MBL-2 tumor cells and 1 × 10⁶ splenic CD8⁺ DCs or CD8⁻ DCs were added at the time of BMT. MBL-2–specific CD8⁺ T cells were analyzed by using gag-tetramer staining. (A) The representative histogram of gag expression in CD229.1⁺CD8⁺ on day 14 after allo-HCT. The percentage of gag positivity in allogeneic β2MG^−/− with host-type CD8⁺ DCs was 1.8. (B) The absolute number of donor-derived gag-tetramer⁺CD8⁺ T cells (n = 6 per group). (C) Absolute number of infiltrated GFP⁺MBL-2 tumor cells in spleen (n = 6 per group). (D) Impaired GVL response is the intrinsic effect of the absence of Batf3 on DCs. β2m^−/− animals received 10 Gy on day −1 and 0.5 × 10⁶ CD8⁺ T cells along with 5 × 10⁶ TCD-BM from either syngeneic B6 or allogeneic C3H.sw donors; syngeneic MBL-2 tumor cells and 2 × 10⁶ splenic DCs from either B6 WT or Batf3^−/− mice were added at the time of BMT. Survival (n = 7 to 11 per group, pooled from 2 experiments). (E) Splenic DC phenotype. The frequency of CD8⁺ DCs and the expression of costimulation molecules of CD11c⁺ cells were analyzed from WT B6 and the naive Batf3^−/− animals that were not transplanted (n = 3 to 5 per group, pooled from 2 experiments). The bars represent mean ± standard deviation. (F) Tumor mortality data of 2 × 10⁴ MBL-2 cells (n = 9 to 22 per group, pooled from 3 experiments) in the 10× generations Batf3^−/− hosts.