Donor T cells primed on leukemia lysate-pulsed recipient APCs mediate strong graft-versus-leukemia effects across MHC barriers in full chimeras

Abstract

Antigen-presenting cells (APCs) of host origin drive graft-versus-leukemia (GVL) effects but can also trigger life-threatening graft-versus-host disease (GVHD) after hematopoietic cell transplantation (HCT) across major histocompatibility complex (MHC) barriers. We show that in vitro priming of donor lymphocytes can circumvent the need of recipient-derived APCs in vivo for mediating robust GVL effects and significantly diminishes the risk of severe GVHD. In vitro, generated and expanded T cells (ETCs) mediate anti-leukemia effects only when primed on recipient-derived APCs. Loading of APCs in vitro with leukemia cell lysate, chimerism status of the recipient, and timing of adoptive transfer after HCT are important factors determining the outcome. Delayed transfer of ETCs resulted in strong GVL effects in leukemia-bearing full chimera (FC) and mixed chimera (MC) recipients, which were comparable with the GVL/GVHD rates observed after the transfer of naive donor lymphocyte infusion (DLI). Upon early transfer, GVL effects were more pronounced with ETCs but at the expense of significant GVHD. The degree of GVHD was most severe in MCs after transfer of ETCs that had been in vitro primed either on nonpulsed recipient-derived APCs or with donor-derived APCs.

Figure 1

In vitro priming and expansion of donor-derived T cells result in ETCs reactive against C1498 only when they were primed with recipient-derived DCs_pulsed. Donor-derived ETCs primed with R- or D-DCs_pulsed were further characterized in vitro. (A) Expansion rates of ETCs either primed on R- DCs_pulsed (◆) or D-DCs_pulsed () were determined using the trypan blue exclusion method for viability assessment. Expansion rates of up to 10-fold were obtained within 4 days. Values are shown plus or minus SE (n = 3). Similar expansion rates were achieved after priming with R-or D-DCs_pulsed. (B) The phenotype of ETCs and naive donor splenocytes were characterized by flow cytometric analysis. In the uppermost panels, events were gated on live cells by forward-side scatter exclusion and at least 15 000 live events were acquired. In the bottom panels, data from events gated on live CD8⁺ cells stained with the indicated antibodies are depicted. (C) The ETCs obtained after priming with R-DCs_pulsed (black continuous line) or D-DCs_pulsed (gray continuous line) and donor splenocytes (black dotted line) were stained to study intracellular Granzyme B and surface CD95L expression. Data from events gated on CD8⁺ live cells of the indicated populations are depicted. (D) Left panel: ETCs primed on R-DCs_pulsed (◆) demonstrated robust anti-C1498 responses in vitro, whereas D-DCs_pulsed () failed to mediate significant cytotoxicity (P < .05). Right panel: Cytotoxicity assay performed after blocking target MHC-I (H2D^b, H2K^b) led to a complete loss of the response indicating a cytotoxic activity mediated by CD8⁺ T cells. Values are shown plus or minus SE.

Figure 2

Pulsing the recipient-derived DCs with tumor lysates before priming affects the response to stimulation and the phenotype of the ETCs. Donor-derived ETCs primed with R-DCs_pulsed (solid) or R-DCs_nonpulsed (hollow) were characterized in vitro. (A) The ETCs were stained to study expression of CXCR3 and α₄β₇. Mean fluorescent intensity of the expression on CD8+ live cells is shown. Values are shown plus or minus SE. (B) ETCs either primed on R-DCs_pulsed (◆) or R-DCs_nonpulsed (◇) were tested for their alloreactivity either in a MLR (left panel) at decreasing stimulator/responder ratios or in a cytotoxicity assay (right panel). Either irradiated allogeneic (B6 solid lines) or syngeneic splenocytes (B10.A broken lines) were used as stimulators. For the cytotoxicity assay ConA blasts of the respective strain were used as targets. Values are shown plus or minus SE, n = 3, P < .05. (C) Differences in the TCR Vβ repertoire of CD8 T cells primed on pulsed or nonpulsed DCs. CDR3 length profile of certain Vβ populations was determined by TCR spectratype analysis. cDNA of naive (top panels), R-DCs_pulsed, or R-DCs_nonpulsed ETCs (bottom panels) were amplified by PCR with 21 Vβ-specific primers (Vβ1, Vβ2, Vβ3, Vβ4, Vβ5.1, Vβ5.2, Vβ6, Vβ7, Vβ8.1, Vβ8.2, Vβ8.3, Vβ9, Vβ10, Vβ11, Vβ12, Vβ13, Vβ14, Vβ15, Vβ16, Vβ17, Vβ18) and 1 Cβ1-Cβ2 primer. FAM-labeled PCR products were analyzed by capillary electrophoresis. The Vβ1 is shown exemplarily for a nonaffected CDR3-profile (left panels), the Vβ5.1 and Vβ5.2 for skewed CDR3-profiles (middle and right panels). Representative results of 3 independent experiments are shown.

Figure 3

ETCs primed with B6-DCs_C1498 in vitro mediate strong GVL effects in leukemia-bearing MC and FC recipients after delayed transfer. Hematopoietic chimeras were established with a mixture of 15 × 10⁶ TCD B10.A BM plus 5 × 10⁶ TCD B6 BM for MCs or with 15 × 10⁶ TCD B10.A BM alone for FCs. The recipients were injected with 0.6 × 10⁶ C1498 cells intravenously on day 55. On day 56, the recipients were treated with 40 × 10⁶ effector cells. Survival was monitored in MC and FC recipients for 100 days after treatment. As effectors either (A) ETCs:R-DCs_pulsed (-♦-, n = 9-MC, 18-FC), (B) ETCs:D-DCs_pulsed (-♦-, n = 9), or (C) ETCs:R-DCs_nonpulsed (-◇-, n = 9) were given. Naive DLI (-■-) and PBS (-▴-) were given as controls. P greater than or equal to .05 was considered not significant (NS).

Figure 4

Transfer of ETCs that had been primed on recipient-derived DCs mediated a shift from mixed to full hematopoietic chimerism. No GVHD was observed. (A) GVHD was monitored on a clinical 10-point scale weekly for each animal. Each symbol represents the severest degree of GVHD that was observed in a single animal during the observation period. Groups were either treated with ETCs:R-DCs_pulsed (-♦-, n = 18-MC, 18-FC), ETCs:D-DCs_pulsed (-♦-,n = 9-MC and FC), ETCs:R-DCs_nonpulsed (-◇-,n = 9-MC and FC), naive DLI (-■-, n = 9 MC and FC), or PBS (-▴-, n = 9-MC, 18-FC). (B) MC were injected with 0.6 × 10⁶ C1498 cells intravenously on day 55 and 40 × 10⁶ effector cells on day 56. Chimerism was calculated by determining the percentage of CD45⁺H2D^d+ cells among the CD45⁺ cells. The values shown represent the mean chimerism of all mice per treatment group plus or minus SE.

Figure 5

Early adoptive transfer of ETCs is associated with improved leukemia-free survival at the expense of severe GVHD. Mixed (MC) and full chimeras (FC) were challenged with 0.6 × 10⁶ C1498 cells intravenously on day 20. On day 21, the recipients were treated with 40 × 10⁶ effector cells. As effectors either (A) ETCs:R-DCs_pulsed (-♦-, n = 9), (B) ETCs:D-DCs_pulsed (-♦-, n = 9), or (C) ETCs:R-DCs_nonpulsed (-◇-, n = 9) were adoptively transferred by tail vein injection. Naive DLI (-■-, n = 8) and PBS (-▴-, n = 9) were used as controls. Survival was monitored in MC and FC recipients for 100 days after treatment. Deaths due to GVHD are indicated by an asterisk. Pooled data from 2 independent experiments are shown. P values are added to the figure. P less than .05 was considered to be significant.

Figure 6

Early adoptive transfer of ETCs triggers less severe GVHD in FCs than in MCs. ETCs:R-DCs_pulsed (-♦- n = 18-MC, n = 9-FC), ETCs:R-DCs_nonpulsed (-◇-n = 18-MC, n = 9-FC), ETCs:D-DCs_pulsed (-♦-, n = 18-MC, n = 9-FC), naive DLI (-■-, n = 18-MC, n = 9-FC), or PBS (-▴-, n = 18-MC, n = 9-FC) were given to either MCs (A,B) or FCs (C,D). Weight changes show more weight loss in MC animals that in FC animals. Mean weights plus or minus SE from each cohort are presented. GVHD was monitored weekly for each animal on a clinical 10-point scale. Each symbol represents the severest degree of GVHD that was observed in a single animal during the post transplant course. Mice treated with ETCs:R-DCs_pulsed developed less severe GVHD scores compared with all other treatment groups.

Figure 7

ETCs persist longer in secondary lymphoid organs of FCs as compared with MCs. Hematopoietic chimeras received late adoptive transfer (AT) with 40 × 10⁶ ETCs:R-DCs_pulsed (ETCs:R-DCp), ETCs:R-DCs_nonpulsed (ETC:R-DCnp), or naive DLI. The effectors were labeled with 5 μM CFSE. Spleens of the transplanted animals were harvested 3 or 7 days later, single cell suspensions were made and analyzed for CFSE intensity. (A) Representative day 7 results from flow cytometric analysis of harvested splenocytes from MCs and FCs are shown. A CFSE high peak (nondivided ETCs) was detected in FCs only (the marker labels the CFSE high/low region). (B) Adoptively transferred ETCs clear faster from MCs than from FCs. Spleens from MCs and FCs were harvested on days 3 and 7 after delayed adoptive transfer. Data obtained from animals that received either ETCs:R-DCsp or ETC:R-DCsnp are presented separately. A DLI group was added as a control. The fraction of CFSE^hi cells (indicating nondivided ETCs) and CFSE^lo cells (indicating proliferated ETCs) was determined by flow cytomertic analysis. (C) Both, in MCs and FCs more vigorous proliferation of ETCs is seen when adoptive transfer is performed early after transplantation (day 21). Spleens from MCs and FCs were harvested on day 7 after early or delayed AT. The absolute number of CFSE^hi and CFSE^lo cells was calculated. Mean values are presented plus or minus SE.