Effector memory CD4+ T cells mediate graft-versus-leukemia without inducing graft-versus-host disease

Abstract

Much of the efficacy of allogeneic hematopoietic stem cell transplantation (alloSCT) in curing hematologic malignancies is due to a graft-versus-leukemia (GVL) effect mediated by donor T cells that recognize recipient alloantigens on leukemic cells. Donor T cells are also important for reconstituting immunity in the recipient. Unfortunately, donor T cells can attack nonmalignant host tissues and cause graft-versus-host disease (GVHD). We previously reported that donor CD4(+) effector memory T cells (T(EMs)) do not cause GVHD but transfer functional T-cell memory. In the present work, we demonstrate in an MHC-mismatched model that CD4(+) T(EMs) (unprimed to recipient antigens) mediate GVL against clinically relevant mouse models of chronic phase and blast crisis chronic myelogenous leukemia, without causing GVHD. By creating gene-deficient leukemias and using perforin-deficient T cells, we demonstrate that direct cytolytic function is essential for T(EM)-mediated GVL, but that GVL is retained when killing via FasL, TNF-alpha, TRAIL, and perforin is individually impaired. However, T(EM)-mediated GVL was diminished when both FasL and perforin pathways were blocked. Taken together, our studies identify T(EMs) as a clinically applicable cell therapy for promoting GVL and immune reconstitution, particularly in MHC-mismatched haploidentical alloSCTs in which T cell-depleted allografts are commonly used to minimize GVHD.

Figure 1

CD4⁺ T_EMs do not induce GVHD in the B6^bm12→B6 strain pairing. B6 mice were irradiated and reconstituted with T-cell–depleted B6^bm12 BM, with no T cells, or with 5 × 10⁵ T_N or 10⁶ T_EM B6^bm12 CD4 cells. Data were combined from 2 independent experiments. (A) Sorting of T_N and T_EM CD4 cells from splenocytes. The first panel is gated on CD4⁺CD25⁻ cells, which were sorted into T_Ns and T_EMs based on the expression of CD62L and CD44. Numbers on plots are percentages of total cells within the rectangles. (B) Weight change (P < .007 at days 5, 7, and 9; P < .02 at days 34, 37, and 40 comparing recipients of T_Ns versus T_EMs). P values are not significant at any time point comparing BM versus T_EMs. (C) Pathology scores from day 43 after transplantation. P values comparing T_N and T_EM recipients: P < .001 for liver, small intestine, and colon; P = .055 for skin. P values are not significant comparing scores for recipients of BM alone versus CD4⁺ T_EMs. Horizontal lines represent mean values.

Figure 2

CD4⁺ T_EMs mediate GVL against mCP-CML without causing GVHD. B6 mice were irradiated and reconstituted with T cell–depleted B6^bm12 BM, with B6 mCP-CML with no T cells, or with 5 × 10⁵ T_N or 10⁶ T_EM B6^bm12 CD4 cells. (A) Shown is survival data combined from 2 independent experiments. P < .001 for recipients of T_EMs versus only T cell–depleted BM. (B) Representative serial flow cytometry of peripheral blood. Each row represents serial bleeds of individual mice. Data from 2 T_EM recipients are shown to illustrate the types of responses we observed. Numbers on plots are percentages of total cells within the rectangles. (C) Numbers of NGFR⁺ cells in the peripheral blood of mice that underwent transplantation taken at different time points. Each symbol represents an individual animal; solid lines are mean values. P < .002 comparing T_Ns versus T_EMs from day 15 onward.

Figure 3

GVL mediated by T_EMs was not due to contaminating CD4⁺ T_Ns. B6 mice were irradiated and reconstituted with T cell–depleted B6^bm12 BM, B6 mCP-CML with no T cells, 5 × 10⁵ T_Ns, 10⁶ T_EMs, or 2 × 10³ T_Ns (T_N control indicates the number of T_Ns contaminating the T_EM sorted cells) B6^bm12 CD4 cells. Shown are the survival data.

Figure 4

CD4⁺ T_EMs mediate GVL against mBC-CML. Lethally irradiated B6 mice were reconstituted with T cell–depleted B6^bm12 BM, 10⁴ B6 mBC-CML cells, with no T cells, or with 10⁶ CD4⁺ T_EMs or 5 × 10⁵ CD4⁺ T_Ns. Shown are results combined from 2 similar independent experiments. P = .024 comparing BM alone versus T_EM recipients.

Figure 5

CD4⁺ T_Ns and T_EMs require cognate interactions with MHCII⁺ mCP-CML cells to mediate GVL. B6 mice were irradiated and reconstituted with T cell–depleted B6^bm12 BM, wt B6, or IA^bβ^−/− mCP-CML, with no T cells or with 5 × 10⁵ T_N or 10⁶ T_EM B6^bm12 CD4 cells. (A) Survival data from 1 of 2 similar experiments; P = .002 for recipients of IA^bβ^−/− mCP-CML and T_EMs versus wt B6 mCP-CML and T_EMs; P = .004 comparing recipients of IA^bβ^−/− mCP-CML and T_Ns versus wt B6 mCP-CML and T_Ns. (B) Numbers of NGFR⁺ cells in the peripheral blood of mice that underwent transplantation taken at different time points. At day 9, peripheral blood was harvested from only 2 of 8 recipients of wt mCP-CML and T_Ns and 2 of 4 recipients of IA^bβ^−/− mCP-CML and T_Ns, as we were concerned that the remainder might not tolerate the procedure due to the severity of GVHD. Comparing recipients of wt mCP-CML and T_Ns versus T_EMs, P < .04 on days 15, 22, and 39 and P > .3 on days 37 and 54. (C) Spleen weight of each recipient after death or killing at day + 50. P > .4 comparing spleen weights of recipients of IA^bβ^−/− mCP-CML and either T_Ns or T_EMs versus no T cells; P < .008 comparing spleen weights of recipients of wt mCP-CML and either T_Ns or T_EMs versus no T cells. Horizontal bars represent mean values.

Figure 6

GVL mediated by CD4⁺ T_EMs is intact when killing by FasL, TNF, TRAIL, and perforin is individually blocked, but is reduced when killing by both FasL and perforin is prevented. B6 mice were irradiated and reconstituted with T cell–depleted B6^bm12 BM, mCP-CML derived from wt, Fas^lpr, or TNFR1R2^−/− B6 mice (A,C,D) or TRAILR^−/− or TRAILR^+/+ littermates (B; data from 1 of 2 similar experiments), with no T cells, CD4⁺ T_Ns, or CD4⁺ T_EMs from wt or perforin^−/− B6^bm12 mice. Survival curves are plotted (A-E). The strains from which donor T cells and mCP-CML were derived are noted above each graph. When individually blocked, killing via TNFR1/R2 (A; P > .21 for recipients of wt T_EMs or T_Ns and TNFR1R2^−/− mCP-CML versus wt mCP-CML), TRAILR (B; P > .12 for recipients of wt T_EMs or T_Ns and TRAILR^−/− mCP-CML versus TRAILR^+/+ mCP-CML), Fas (C; P > .63 for recipients of wt T_EMs or T_Ns and Fas^lpr mCP-CML versus wt mCP-CML), or perforin (D; P > .47 for recipients of wt mCP-CML and Prf1^−/− versus wt T_Ns or T_EMs) is not required for GVL by either CD4⁺ T_Ns or T_EMs. However, when killing via both perforin and Fas was prevented (E), GVL by T_EMs but not T_Ns was diminished (P = .03 comparing recipients of CD4⁺ Prf1^−/− T_EMs and wt mCP-CML versus CD4⁺ Prf1^−/− T_EMs and Fas^lpr mCP-CML; P = .01 for recipients of Fas^lpr mCP-CML and no T cells versus Fas^lpr mCP-CML and CD4⁺ Prf1^−/− T_EMs; P = .71 for recipients of CD4⁺ Prf1^−/− T_Ns and wt mCP-CML versus CD4⁺ Prf1^−/− T_Ns and Fas^lpr mCP-CML). Survival plots in panels C-E are from data combined from 2 independent experiments.

Figure 7

T_EMs are defective in their ability to produce cytokines and induce lower levels of TNF-α and IFN-γ in recipients. B6 CD45.1 mice were irradiated and reconstituted with TCD B6^bm12 BM with either 10⁶ B6^bm12 CD4⁺ T_EMs or 5 × 10⁵ CD4⁺ T_Ns. (A) Representative staining for intracellular IFN-γ and IL-2 in donor-derived CD45.1⁻CD4⁺ cells in spleens harvested day +7 after BM transplantation. Numbers on plots are the percentages of total cells in the rectangles. (B) Total number of donor CD4⁺Thy1.2⁺ cells in spleen on day +7 (left panel; P = .002) and the percentage of these that stain for intracellular IFN-γ (right panel; P = .001). Each symbol represents data from an individual mouse; the horizontal line is the mean. (C) Mice were bled on days +5, +7, and +10, and serum levels of IFN-γ and TNF-α were measured by Luminex technology. Recipients of T_Ns (relative to T_EMs) had higher serum levels of IFN-γ (day +7; P < .001) and TNF-α (days +7 and +10; P = .002 and P = .015, respectively). Error bars represent SD.