Induction of tolerance to bone marrow allografts by donor-derived host nonreactive ex vivo-induced central memory CD8 T cells

Abstract

Enabling engraftment of allogeneic T cell-depleted bone marrow (TDBM) under reduced-intensity conditioning represents a major challenge in bone marrow transplantation (BMT). Anti-third-party cytotoxic T lymphocytes (CTLs) were previously shown to be endowed with marked ability to delete host antidonor T cells in vitro, but were found to be less effective in vivo. This could result from diminished lymph node (LN) homing caused by the prolonged activation, which induces a CD44(+)CD62L(-) effector phenotype, and thereby prevents effective colocalization with, and neutralization of, alloreactive host T cells (HTCs). In the present study, LN homing, determined by imaging, was enhanced upon culture conditions that favor the acquisition of CD44(+)CD62L(+) central memory cell (Tcm) phenotype by anti-third-party CD8(+) cells. These Tcm-like cells displayed strong proliferation and prolonged persistence in BM transplant recipients. Importantly, adoptively transferred HTCs bearing a transgenic T-cell receptor (TCR) with antidonor specificity were efficiently deleted only by donor-type Tcms. All these attributes were found to be associated with improved efficacy in overcoming T cell-mediated rejection of TDBM, thereby enabling high survival rate and long-term donor chimerism, without causing graft-versus-host disease. In conclusion, anti-third-party Tcms, which home to recipient LNs and effectively delete antidonor T cells, could provide an effective and novel tool for overcoming rejection of BM allografts.

Figure 1

Anti–third-party CTLs are excluded from the LNs, and do not colocalize with HTCs in the early time points after BM transplantation. BALB/c mice were conditioned with 8 Gy of lethal TBI and infused with different cell mixtures as follows: (A) C57BL/6-NUDE BM (4 × 10⁶). (B) C57BL/6-NUDE BM + HTCs (1 × 10⁴) + DIR-labeled syngeneic naive HTCs (7 × 10⁶). (C) C57BL/6-NUDE BM + HTCs (1 × 10⁴) + DIR-labeled naive allogeneic CB6-derived CD8⁺ T cells (7 × 10⁶). (D) C57BL/6-NUDE BM + HTCs (1 × 10⁴) + DIR-labeled allogeneic CB6-derived anti–third-party CTLs (7 × 10⁶). Images were taken 36 hours after transplantation using IVIS. A representative mouse (of 6 mice in 2 independent experiments) is shown for each group.

Figure 2

In vitro induction of anti–third-party Tcms, and their homing to recipient LNs. (A) BALB/c or CB6 splenocytes were stimulated with irradiated FVB splenocytes for 60 hours or 6 days in the absence of cytokines. Subsequently, CD8⁺ T cells were positively selected and further cultured with rhIL-2 or rhIL-15 in an Ag-free environment. Alternatively, anti–third-party CTLs were grown by culturing the cells with rhIL-2 and reactivating them with FVB splenocytes, after the 6-day cytokine deprivation period. On day 15 of the culture (end of culture), the cells were evaluated for percentage of Tcms (CD44⁺CD62L⁺) using FACS analysis. Data represent average ± SD of at least 3 independent experiments for each group. ***P < .001 compared with the cells cultured with only IL-15 after 60 hours of cytokine deprivation. (B-E) Lethally irradiated (8 Gy) BALB/c mice received 4 × 10⁶ C57BL/6-NUDE BM cells and 1 × 10⁴ syngeneic T cells. Mice then received a transplant of 1 × 10⁷ DIR-labeled, CB6-derived, purified anti–third-party Tcms (B,D) or CTLs (C,E). After 2 (B-C) or 7 (D-E) days, selected recipients were killed and images were taken ex vivo using IVIS. A representative mouse, of 6 mice in 2 independent experiments, is displayed for each group.

Figure 3

Tcms, in contrast to CTLs, populate the recipient LNs and proliferate extensively shortly after the BMT. Lethally irradiated (8 Gy) BALB/c (H-2^d) mice received 4 × 10⁶ C57BL/6-NUDE (H-2^b) BM cells and 1 × 10⁴ syngeneic HTCs. Mice then received a transplant of 10⁷ DIR-labeled, CB6 (H-2^bd)–derived, purified anti–third-party Tcms or CTLs. After 2 or 7 days, selected recipients were killed; the peripheral LNs (pLNs), mesenteric LNs (mLNs), spleen, liver, and BM were extracted; cells were purified from the different organs; and the purified cells were analyzed by FACS. To obtain absolute values of cells, samples were suspended in constant volume and flow cytometric counts for each sample were obtained during a constant, predetermined period of time and were compared with flow cytometric counts obtained from control samples that were set up with fixed volume and fixed numbers of input cells. (A) Representative FACS analysis demonstrating the presence of CD8⁺ and alive (7AAD⁻) CB6-derived cells (Tcms or CTLs) in the peripheral LNs of the recipients 2 days after transplantation. (B-C) Quantification of the FACS analysis as described in panel A, demonstrating the distribution of CB6-derived cells in various organs at 2 (B) and 7 (C) days after transplantation. (D) The sum of the total number of Tcms or CTLs, harvested from all organs tested, at 2 or 7 days after transplantation. In panels B-D, results shown represent average ± SD of pooled data from 6 animals from each group in 2 independent experiments. *P < .05.

Figure 4

Tcms display low veto activity in vitro, but upon reactivation acquire an effector phenotype, which is associated with potent and specific veto activity. (A) 2c splenocytes were stimulated with irradiated BALB/c (H-2^d) splenocytes in the presence or absence of CB6 (H-2^bd)–derived purified Tcms, CTLs, or purified Tcms reactivated in vitro with their cognate third-party FVB stimulators for 60 hours (R′ Tcms). The veto cells were added at the indicated veto-effector ratios. Veto activity was analyzed by FACS analysis 3 days after the initiation of the mixed lymphocyte reaction (MLR), to monitor the inhibition of CD8⁺1B2⁺ 2c cell expansion. Results are presented as mean ± SD of percentage inhibition from 5 independent experiments. (B) FACS analysis of annexin V levels on living (7AAD⁻) CD8⁺1B2⁺ 2c cells, at the end of the MLR, plated with or without veto cells at a veto-effector ratio of 0.02. Results are presented as mean ± SD of percentage annexin V in 7 independent experiments. (C) MLR was established as in panel A. The inhibition of 2c cell expansion was evaluated when veto cells derived from specific CB6 (H-2^bd, spe′) or nonspecific C3B6F1 (H-2^bk, nonspe′) mice were added at a 0.02 veto-effector ratio. Results are presented as mean ± SD of percentage inhibition from 4 independent experiments. The 2c cells were also analyzed for annexin V levels (D). (E) Lethally irradiated (8 Gy) BALB/c mice received 4 × 10⁶ C57BL/6-NUDE BM cells and 1.25 × 10⁴ syngeneic T cells. Mice then received a transplant of 2 to 10 × 10⁶ purified CB6 CD8⁺ Tcms, which were analyzed for CD62L phenotype before the adoptive transfer (left panel, Tcms). Mice were killed 4 days after adoptive transfer, and LNs were harvested and mashed; CD62L expression on CB6 CD8⁺ T cells, isolated from the LNs, was analyzed by FACS (right panel, LN Tcms). Representative result of Tcms isolated from LNs of 1 of 20 mice tested in 4 independent experiments is displayed. (A-E) ***P < .001.

Figure 5

Tcms specifically delete antidonor T cells in vivo. (A-B) Lethally irradiated (10 Gy) C57BL/6 mice received 1 × 10⁵ purified CD8⁺ 2c cells and 5 × 10⁵ irradiated BALB/c splenocytes. The following day, the mice received a transplant of 1 × 10⁶ C57BL/6-NUDE BM cells or received, in addition, 5 × 10⁶ specific, derived from CB6, or nonspecific, derived from C57BL/6, purified Tcms. Recipients were killed 8 days after transplantation, their spleens were harvested, and the deletion of 2c T cells was monitored by FACS. (A) Representative result demonstrating the level of surviving (7AAD⁻) 2c cells in the absence (left panel, 2c alone) or presence (right panel, 2c + specific Tcms) of specific Tcms. (B) Quantification of results measuring the inhibition of the 2c cells by specific and nonspecific Tcms. Data represent average ± SD of percentage inhibition from at least 10 animals in each group, pooled from 2 independent experiments. (C-D) Syngeneic BMT model was established as in panels A and B, but 5 × 10⁵ purified CD8⁺ 2c cells and 2.5 × 10⁶ irradiated BALB/c splenocytes were administrated. Recipients were killed 8 days after transplantation, their spleens were harvested, and FACS analysis of annexin V levels on living (7AAD⁻) CD8⁺1B2⁺ 2c cells was conducted. (C) Representative result demonstrating annexin V levels on 2c cells in the absence (left panel, 2c alone) or presence (right panel, 2c + specific Tcms) of specific Tcms. (D) Quantification of results measuring annexin V levels on the 2c cells after interactions with specific or nonspecific Tcms. Data present average ± SD of percentage annexin V levels in at least 4 animals from each group, in 1 representative experiment of 3 performed. **P < .01; ***P < .001.

Figure 6

Tcms are endowed with marked tolerance induction capabilities, and persist in vivo at least 1 year after BMT. (A) Lethally irradiated (10 Gy) C3H (H-2^K) mice received 1.25 × 10⁴ syngeneic HTCs. Mice then received a transplant of 3 × 10⁶ BALB/c-NUDE BM cells (H-2^d,BA-NU BM) in the presence or absence of different doses of (C3H × BALB/c)F1 (H-2^Kd, C3BF1) purified CD8⁺ Tcms. Data were pooled from 6 independent experiments. (B) Graft rejection model was established as in panel A. Mice received a transplant of 3 × 10⁶ BALB/c-NUDE BM cells (H-2^d, BA-NU BM) in the presence or absence of 5 × 10⁶ (C3H × BALB/c)F1 (H-2^Kd, C3BF1) purified Tcms or 1 × 10⁷ (C3H × BALB/c)F1 CTLs. Data were pooled from 5 independent experiments. (C) Peripheral blood levels of Tcms were analyzed 1 year after BMT by FACS measuring H2K^kH2D^d double-positive cells in the CD8⁺ gate. The figure shows representative mice of 7 mice that received a transplant of BM only (BM alone) or 7 mice that received a transplant of BM + HTCs + 5 × 10⁶ (C3H × BALB/c)F1 CD8⁺ Tcms (BM + Tcms).

Figure 7

Fully allogeneic anti–third-party Tcms are depleted of GVH reactivity and support engraftment of TDBM allografts. (A-B) Supralethally irradiated (11 Gy) C3H mice were radioprotected with 5 × 10⁶ BALB/c-NUDE BM cells in the presence or absence of 5 × 10⁶ or 2 × 10⁶ BALB/c-derived CD8⁺ purified Tcms (BM + Tcm 5 or BM + Tcm 2, respectively) or naive cells (BM + naive 5 or BM + naive 2, respectively). The GVH reactivity of the Tcms or the naive cells was reflected by survival percentage (A) or by average weight change (B) during 100 days after transplantation. Data represent averages of 3 independent experiments, with at least 5 mice for each group, in each experiment. **P < .01 compared with the group of mice that received only BM. (C) Lethally irradiated (10 Gy) C3H mice received 1.25 × 10⁴ syngeneic HTCs. Mice then received a transplant of 3 × 10⁶ BALB/c-NUDE BM cells (BA-NU BM) in the presence or absence of 5 × 10⁶ BALB/c CD8⁺ purified Tcms or 1 × 10⁷ BALB/c CTLs. Data were pooled from 3 independent experiments.