Circulating CD56 Bright Natural Killer Cells after Allogeneic Stem Cell Trasnplantation Express Functional TRAIL; Implication for a Novel Target of Immunotherapy.

Relapse of malignancies remains to be one of the major problems after allogeneic stem cell transplantation (allo-SCT). It is well-recognized that natural killer cells (NK-cells) are predominated in early phase of immune reconstitution after allo-SCT, and several studies demonstrated that CD56 bright CD16 negative (CD56++CD16−) NK-cells, which account for only a few percentage of peripheral NK-cells in healthy individuals, constitute a large subset of NK-cells at this phase. Although CD56++CD16− NK-cells possess unique ability to proliferate and produce proinflammatory cytokines in response to monokines or IL-2, they have been regarded to be less cytotoxic and unfavorable for graft-versus leukemia effects. To verify this issue, we compared the frequency of peripheral CD56++ NK-cells among total NK-cells with subsequent relapse in 25 allo-SCT recipients. Although the ratio of CD56++ NK-cells was gradually decreased as the increased duration between phlebotomy and allo-SCT, we could divide these patients into two groups. Group 1 was consisted of patients who showed consistently elevated ratio of CD56++ NK-cells, and the remainder was categorized into group 2. The relapse after allo-SCT was seen in 1 out of 8 patients in group 1, whereas it was documented in 5 out of 17 patients in group 2. This finding suggested that CD56++ NK-cells might also have a role in preventing relapse. We have found that peripheral CD56++CD16− NK-cells from patients after allo-SCT consistently expressed TNF-related apoptosis-inducing ligand (TRAIL), although its expression was faintly detectable on circulating NK-cells from healthy volunteers. As reported, stimulation with IL-2 or IL-15 resulted in the increased expression of TRAIL on NK-cells from healthy volunteers as well as the recipients of allo-SCT. However, its expression was always stronger in the CD16- subset than CD16+ in both groups. Cultivation of purified NK-cells from healthy volunteers with 0.5 nM of IL-2 for more than 2 weeks resulted in the expansion of both NK-cell subsets, and after sorting into CD16− and CD16+ NK-cells, cytotoxic assays against Jurkat were performed in the presence or absence of concanamycin A, neutralizing anti-Fas antibody, and neutralizing anti-TRAIL antibody. Cytotoxicity was more prominent in the CD16− subset than CD16+, and blocking study revealed that TRAIL expressed on CD16− NK-cells was strongly involved in the killing of Jurkat. We could not detect TRAIL-mediated cytotoxicity in the CD16+ subset, because the expression of TRAIL was much lower in the CD16+ subset than CD16−. Next, NK-cells purified from allo-SCT recipients and healthy volunteers were overnight cultured with 0.5 nM of IL-2 and their cytotoxicity against Jurkat was examined. NK-cells from patients who received allo-SCT within 3 months and those from healthy volunteers showed equivalent cytotoxicity. In patients who showed increased ratio of CD56++CD16− NK-cells, TRAIL was strongly expressed on overnight cultured CD56++CD16− NK cells, and TRAIL-mediated cytotoxicity was also detected. In murine models, TRAIL has been reported to exert strong graft-versus-tumor effects without causing GVHD. As CD56++CD16− NK cells readily express functional TRAIL on cytokine stimulation, and they usually reconstituted shortly early after allo-SCT, they may become promising targets for immunological intervention.