Functional NK Cell Repertoires Are Maintained through IL-2Rα and Fas Ligand

Martin Felices, Todd R. Lenvik, Dave E. M. Ankarlo, Bree Foley, Julie Curtsinger, Xianghua Luo, Bruce R. Blazar, Stephen K. Anderson and Jeffrey S. Miller
J Immunol April 15, 2014, 192 (8) 3889-3897; DOI: https://doi.org/10.4049/jimmunol.1302601

Abstract

Acquisition of a functional NK cell repertoire, known as education or licensing, is a complex process mediated through inhibitory receptors that recognize self. We found that NK cells containing self-killer Ig-like receptors for cognate HLA ligand in vivo were less susceptible to apoptosis. In vitro IL-15 withdrawal showed that uneducated NK cells upregulated Bim and Fas. Conversely, educated NK cells upregulated Fas ligand (FasL) under these conditions. Induction of cell death and Bim expression on uneducated cells correlated with increased IL-2Rα expression. Overexpression and knockdown studies showed that higher IL-2Rα limits NK cell survival in a novel manner that is independent from the role of IL-2 in activation-induced cell death. To study the role of FasL in induction of IL-2Rαhi NK cell death, a coculture assay with FasL-blocking Abs was used. IL-15 withdrawal led to FasL-dependent killing of IL-2Rαhi NK cells by more educated IL-2Rαlo NK cells. Finally, CMV reactivation induces a potent long-lasting population of licensed NK cells with enhanced survival. These findings show that education-dependent NK cell survival advantages and killing of uneducated NK cells result in the maintenance of a functional repertoire, which may be manipulated to exploit NK cells for cancer immunotherapy.

Introduction

Natural killer cell–mediated immunotherapy is being tested in clinical trials (1, 2). Inhibitory killer Ig-like receptors (KIRs) on NK cells induce function through a process termed “education” or “licensing” (3–5). The success of NK cell therapy depends on how functionally competent cells are homeostatically maintained after adoptive transfer. Although NK cell homeostasis has been studied on bulk NK cells, questions remain as to how the repertoire of functionally competent NK cells is maintained to exploit its differentiation state and previous exposure history (6–9).

The balance between more mature KIR+ NK cells and less differentiated KIR− NK cells could be regulated by three mechanisms in the periphery: proliferation, differentiation, and survival. Higher proliferation in the KIR− NK cell subset (10) and poor proliferation of more mature CD57+ NK cells have been described (11). These data suggest that the differentiation status of NK cells correlates inversely with proliferative potential. Differentiation could explain the persistence of KIR+ NK cells that develop from KIR− NK cells. Alternatively, it was shown that KIR expression on T cells may inhibit activation-induced cell death (AICD) by inducing PI3K/Akt (12–15), suggesting that enhanced survival could control the balance between functional NK cells and their less-differentiated counterparts. In support of this, H2Dd-transgenic mice showed a dose-dependent enrichment of educated Ly49A+ NK cells with increased sensitivity to IL-15 and reduced apoptosis (16). Although these data suggest that education plays a role in NK cell survival, it is unknown whether KIRs mediate such signals in humans.

IL-15 and IL-2 have prominent and overlapping roles in proliferation, survival, and NK cell activation (17). Similarities in function stem from shared usage of the common γ-chain (CD132) and IL-2Rβ (CD122), with nearly identical downstream signaling components (18). Specificity is determined by selective binding to IL-2Rα (CD25) or IL-15Rα (CD215). Only IL-15 is critical for NK cell development and homeostasis (19–24), indicating divergence in function. Negative effects on survival, such as the well-described role of IL-2 in FAS-mediated AICD (25, 26), also were demonstrated. The role of NK cell education on cytokine-mediated survival remains unknown.

Materials and Methods

Cell isolation

Peripheral blood from healthy human donors was obtained from the Memorial Blood Center (Minneapolis, MN). A Histopaque gradient (Sigma-Aldrich) was used to obtain PBMCs. For NK cell–education studies, PBMCs were HLA and KIR typed, and cells were frozen down for later use. For experiments using NK cells alone, they were enriched using a MACS NK Cell Isolation Kit (Miltenyi Biotec), as per the manufacturer’s protocol. Where noted, cells were further sorted using a FACSAria II cell sorter (BD Biosciences) and processed for RNA or protein. Unless otherwise described, the following medium was used for experiments: complete DMEM (Cellgro) without exogenous cytokines (unless otherwise noted), supplemented with 10% human AB serum (Valley Biomedical), 30% Ham F-12 medium (Cellgro), 100 U/ml penicillin (Invitrogen), 100 U/ml streptomycin (Invitrogen), 24 μM 2-ME, 50 μM ethanolamine, 20 mg/l ascorbic acid, and 50 μg/l sodium selenate.

Mouse studies

NOD scid IL-2R γ-chain knockout (NSG) mice were purchased from The Jackson Laboratory and maintained at the University of Minnesota under specific pathogen–free conditions using protocols approved by the Institutional Animal Care and Use Committee. Mice were treated and assessed at different harvest times, as described.

Patients and samples

Adult donor hematopoietic cell transplant and umbilical cord blood allogeneic transplant samples were used from previously described studies (7, 8); a detailed description of patients is included in Supplemental Table I. Briefly, after transplant, patients were monitored weekly for CMV reactivation by quantitative PCR performed in the clinical virology laboratory, defined as CMV viremia (>100 copies). CMV reactivation occurred prior to day 100 after transplantation and was treated with an 8-wk course of ganciclovir. PBMCs were collected at the noted times for analysis based on the time after transplantation. Samples were obtained after informed consent and approval from the University of Minnesota Institutional Review Board according to the Declaration of Helsinki.

Flow cytometry

All fluorochrome-conjugated Abs were purchased from BioLegend, BD Biosciences, and Beckman Coulter. Staining, acquisition, and analysis were performed as previously described (7).

Proliferation and differentiation assays

PBMCs were labeled with CellTrace Violet Cell Proliferation Dye (Invitrogen), per the manufacturer’s protocol, and placed in culture medium supplemented with IL-15 (10 ng/ml; R&D Systems) for 7 d, followed by surface staining. For the differentiation assay, sorted CD56dimCD3−KIR− NK cells were placed in the same conditions and assessed for KIR acquisition.

Cell death assay

To evaluate viability, we used LIVE/DEAD Fixable Dead Cell Stain (Invitrogen), per the manufacturer’s protocol, followed by labeling with desired surface stains. Cell death was assessed on CD56dimCD3− cells. To study initiation of apoptosis in live cells, a similar approach was taken, but Annexin V Ab was added to the surface mix, and gating was done as defined.

Quantitative RT-PCR and Western blot analysis

For evaluation of transcripts and protein, pure populations of CD56dimCD3−KIR− or CD56dimCD3−KIR+ NK cells were isolated, as stated. RNA was processed and analyzed as previously described (7). For immunoblots, all procedures were done as previously described (27). Bcl-xL Ab was purchased from Cell Signaling.

Transfection studies

Magnetically enriched NK cells were placed in medium with IL-15 (1 ng/ml) for 24 h and transduced with the Amaxa Human Macrophage Nucleofector Kit (Lonza) in a Nucleofector II (Amaxa Biosystems). For overexpression studies, the GFP of the pmax vector was replaced with the PCR product of IL-2RA. For knockdown studies, we used 300 pmol (total) FlexiTube small interfering RNAs (siRNAs) specific for IL2RA, FAS, AllStars Negative Control siRNA, and Alexa Fluor 488–labeled AllStars Negative Control siRNA (QIAGEN). Five hours posttransduction, the cells were replated with media plus IL-15 (1 ng/ml). After 24 h (48 h for FAS studies), cells were spun and resuspended in RPMI 1640 with 10% serum to simulate IL-15 withdrawal. Forty-eight hours later, cells were stained (or incubated with killers in the case of FAS studies), and cell numbers and cell death were assessed by flow cytometry.

Statistical analysis

Paired t tests were used for comparisons that maintained internal pairing within each donor sample. For comparisons across different donors or the same donor harvested at different time points, an unpaired t test with the Welch correction was used. Multiple comparisons were adjusted using the method of Hommel (28). All tests were two sided. Statistical analyses were carried out with Prism software and SAS 9.3 (SAS Institute, Cary, NC). On all graphs, bars represent the mean ± SEM.

Results

The homeostatic balance between KIR+ and KIR− NK cells is not fully explained by proliferation or continuous differentiation

NK cell function is determined by the repertoire of individual NK cells displaying a variety of receptors. We chose to focus on the more mature CD56dim cells throughout this study. Our first aim was to distinguish among proliferation, differentiation, and survival as possibilities to maintain the functional NK cell repertoire. Proliferation was studied first by labeling PBMCs from healthy donors with a proliferation dye, followed by culture with 10 ng/ml IL-15 for 7 d. Similar to previous studies (10), KIR− NK cells proliferate better than KIR+ NK cells (Fig. 1A). Approximately twice as many KIR− NK cells proliferated past cell division 3 compared with KIR+ NK cell subsets (Fig. 1B). Although KIR− NK cells had increased significantly after 7 d in culture with IL-15 compared with day 0 (Fig. 1C), the increase in KIR− frequencies was lower than expected based on the proliferation data. This discrepancy may be a result of differentiation from KIR− NK cells into KIR+ NK cells. To test this possibility, KIR− NK cells were sorted (Supplemental Fig. 1A) and placed into the same culture conditions for 7 d. Although a small amount of KIR acquisition was observed (Fig. 1D, 1E), this amount does not account for the frequency of KIR+ NK cells after IL-15 culture, given the marked proliferative advantage of KIR− NK cells. To determine whether the in vitro proliferation discrepancy impacted KIR balance in a more physiological setting, a xenogeneic mouse model was used to track proportions of KIR− and KIR+ NK cells after transfer (Fig. 1F). In this setting, proportions of KIR− NK cells did not increase compared with the starting population (Fig. 1G), despite the enhanced proliferation seen with IL-15 in vitro. This inconsistency between proliferation in vitro and homeostasis of KIR proportions in vivo indicated that a mechanism other than proliferation is operant in maintaining the balance between the KIR+ and KIR− populations.

FIGURE 1.
FIGURE 1.

In vitro proliferation and xenogeneic transplantation studies show a discrepancy in the proportion of KIRs expressed by NK cells. (A) Representative graphs of KIR subset proliferation (measured by CellTrace dilution) on NK cells (PBMCs were gated on CD56+CD3− NK cells and the indicated KIR subset) after 7 d in culture with 10 ng/ml IL-15. (B) Aggregate data showing the proliferation (past division 3) of NK cells (CD56+CD3−) stratified by KIR and CD57 expression as indicated (n = 17). (C) Proportion of KIR+ and KIR− NK cells at day 0 (○) and day 7 (●) after in vitro incubation of PBMCs with 10 ng/ml of IL-15 (n = 10). (D) KIR− NK cells were sorted using a FACSAria II sorter and placed in culture with 10 ng/ml IL-15. After 7 d of differentiation, KIR+ NK cells were assessed by flow cytometry. (E) Aggregate data for the differentiation assay results (n = 4). (F) Schematic of transplant methodology. NSG mice were irradiated (275 cGy) prior to i.v. injection of 1 × 106 human NK cells. Mice were injected i.p. on Monday, Wednesday, and Friday with 5 μg rIL-15/mouse (six doses total) over 2 wk. No treatment (NT) was given in the final 2 wk. Blood was collected at days 7 and 28 for flow cytometry analysis of KIR expression (n = 4). (G) KIR expression is shown on human CD56dimCD3− NK cells on day 0 prior to infusion (d0; one human donor), as well as after infusion (n = 4 NSG mice) on day 7 (d7) and day 28 (d28). **p < 0.01, ***p < 0.001.

NK cell education determines NK cell survival

NK cell survival was studied using an in vitro serum-starvation assay. A marked difference in cell death was observed between KIR+ NK cells (Fig. 2A, 2B, 8.9 ± 1.2%, 6.7 ± 0.5%, and 6.2 ± 0.9%) and KIR− NK cells (Fig. 2A, 2B, 33 ± 2.4%). To exclude possible confounding issues with receptor downregulation in dead cells, NK cells were evaluated for early indications of apoptosis (CD56dimCD3−LIVE/DEAD dye−Annexin V+ cells, Supplemental Fig. 1B). We found that the majority of live NK cells entering apoptosis belonged to the KIR− subset (Fig. 2C, 73.9 ± 1.6%). These data indicate that KIR− NK cells are more prone to undergo cell death under the stress of serum starvation.

FIGURE 2.
FIGURE 2.

Preferential survival of educated NK cells. (A) Representative graphs of cell death (measure by LIVE/DEAD dye incorporation) in CD56dimCD3− KIR subsets after overnight serum starvation (RPMI-0%). Aggregate cell death (from A) data (B) and early apoptosis (annexin V+LIVE/DEAD−) data (C) in CD56dim NK cells stratified by KIR expression after serum starvation (n = 9 for both). (D) CD56dimCD3− NK cells were broken down into subsets based on their expression of KIR and NKG2A, and cell death was measured after serum starvation (n = 9). (E) Donor cells were KIR and HLA typed to determine the NK cell educational status. NKG2A−single KIR+ NK cells were considered educated when corresponding cognate HLA ligand was present for self-KIRs, whereas those lacking cognate ligand are uneducated (nonself KIRs). Educated and uneducated NK cells were analyzed for cell death after serum starvation (n = 30). (F) Proportions of CD57+ and CD57− NK cells undergoing cell death after serum starvation (n = 9). Proportion (G) and intensity (H) of CD57 on self-KIR educated versus nonself-KIR uneducated NK cells (n = 20). For education experiments, to achieve internal comparison per donor all samples selected had at least one uneducated KIR. ***p < 0.001.

The most functionally differentiated NK cells are those that express an inhibitory receptor with specificity for self-HLA molecules (29–31). Therefore, cell death was measured after overnight serum starvation in NK cell subsets based on the presence or absence of educating signals. In addition to education through KIRs, NK cells can acquire function through CD94/NKG2A heterodimers (3). The KIR−NKG2A− NK cells, which lack all education signals, consistently exhibited more cell death compared with the KIR−NKG2A+, KIR+NKG2A+, and KIR+NKG2A− subsets (Fig. 2D). It was reported that additive educating signals yield higher functionality (4). Consistent with that finding, cells educated through both NKG2A and KIRs have better survival than those that receive signals through NKG2A or KIRs alone. Although differences in survival were seen based on NKG2A and KIR staining in the absence of ligand determination, education needs to be determined based on the presence or absence of self-ligand in the environment. Therefore, donors were HLA and KIR typed to define educated (NKG2A− NK cells containing a single KIR corresponding to self-ligand [self-KIR]) and uneducated cells (NKG2A− NK cells containing a single KIR without ligand [nonself-KIR]), and survival was assessed after overnight serum starvation. Educated NK cells survived better than uneducated NK cells (Fig. 2E), indicating that signaling through inhibitory KIRs is necessary for the enhanced survival. More differentiated CD57+ NK cells also survived better than did CD57− NK cells after serum starvation (Fig. 2F). As previously reported, there was a higher frequency of CD57 expression on KIR+ NK cells (Supplemental Fig. 2C). Educated NK cells displayed a higher proportion of CD57+ cells (Fig. 2G) and more CD57 on a per-cell basis (Fig. 2H). Taken together, these data show that NK cell education enhances survival and CD57 expression.

KIR+ NK cells express more antiapoptotic molecules

To understand differences in survival, an apoptosis PCR array was used to look at survival-related molecules in sorted CD56dimCD3−KIR+ and CD56dimCD3−KIR− NK subsets (Supplemental Fig. 1D). Although no expression differences were seen in proapoptotic genes, the antiapoptotic genes BCL2 and BCL2L1 [BCLXL] differed (data not shown). The increases in BCL2 and BCL2L1 transcripts in the KIR+ population compared with the KIR− population were independently confirmed with standard quantitative real-time-PCR (Fig. 3A). Increased Bcl-xL protein expression was observed in KIR+ NK cells compared with KIR− NK cells (Fig. 3B). Bcl-2 protein expression was next analyzed by median fluorescence intensity (MFI) on NK subsets under different settings. Although no significant difference was observed under basal conditions, serum starvation or treatment with 10 ng/ml IL-15 overnight (Fig. 3C, data not shown) induced a small increase in Bcl-2 expression on KIR+ NK cells. A differential increase in Bcl-2 also was seen on educated NK cells under basal, serum-starving, or IL-15 treatment conditions overnight (Supplemental Fig. 2B–D). When evaluating proteins involved in cell death under basal conditions, Fas (CD95) was expressed at higher levels on KIR− NK cells than on KIR+ NK cells (Fig. 3D), and there were no differences for Fas ligand (FasL) (CD178) or Bim expression. These results indicate that KIR expression and education equip NK cells, under basal conditions, for survival, with increased antiapoptotic mechanisms and decreased expression of Fas.

FIGURE 3.
FIGURE 3.

KIR+ NK cells express more antiapoptotic molecules and less Fas. NK cells were enriched using the Miltenyi Human NK Cell Isolation Kit and sorted into CD56dimKIR− and CD56dimKIR+ populations. (A) mRNA was isolated, and expression of different antiapoptotic genes (BCL2 and BCL2L1 [BCLXL]) was normalized to 18s rRNA. The ratio of KIR+/KIR− expression is shown (n = 5). (B) Representative Western blot of Bcl-xL protein expression on KIR− versus KIR+ NK cells (n = 6). (C) Bcl-2 protein expression (shown as Bcl-2 MFI) on NK cells was analyzed after PBMCs were incubated overnight with 10 ng/ml IL-15 (n = 9). (D) Fas expression (MFI) was analyzed on different subsets of NK cells under basal conditions (n = 9). **p < 0.01, ***p < 0.001.

Cytokine receptor levels correlate with KIR expression and education

To investigate differences in cytokine receptor signaling between the KIR+ and KIR− subsets, we examined the expression of IL-15Rα (IL15RA [CD215]), IL-2Rα (IL2RA [CD25]), and common signaling components used in both pathways: the common γ-chain (IL2RG [CD132]) and IL-2Rβ (IL2RB [CD122]). Small and variable differences were found in all components (Supplemental Fig. 2A). No difference in IL-15Rα or IL-2Rβ protein expression was found between KIR+ and KIR− subsets at basal levels (Fig. 4A, 4B). KIR+ NK cells expressed more common γ-chain (Fig. 4C). In contrast, decreased IL-2Rα protein was observed on KIR+ NK cells compared with KIR− NK cells (Fig. 4D). Educated NK cells expressed more common γ-chain, but no differences were seen for IL-2Rα under basal conditions (Supplemental Fig. 2F, data not shown).

FIGURE 4.
FIGURE 4.

Differential IL-2 cytokine receptor component expression on educated NK cells. Healthy donor PBMCs were gated on CD56dimCD3− NK cells, and KIR subsets were analyzed for protein expression (MFI) of IL-15Rα (n = 9) (A), IL-2Rβ (n = 7) (B), the common γ-chain (n = 15) (C), and IL-2Rα (n = 7) (D). IL-2Rα MFI on NK cells stratified by KIR expression (n = 9) (E) or NK cell education (n = 18) (F) after 24-h (left panels) or 72-h (right panels) treatment with IL-15 (1 ng/ml). For education experiments, to achieve internal comparison per donor all samples selected had at least one uneducated KIR. *p < 0.05, **p < 0.01, ***p < 0.001.

Because IL-15 signaling on NK cells can alter IL-2Rα levels (32), we set out to investigate whether it could differentially regulate IL-2Rα expression on NK cell populations determined by KIR expression and education. PBMCs were treated for 24 or 72 h with IL-15 and then assessed for IL-2Rα levels on different NK cell subsets. KIR− NK cells had 1.26-fold more IL-2Rα than KIR+ NK cells after 24 h of treatment, and the difference increased to 1.45-fold after 72 h (Fig. 4E). At 24 h, no significant differences in IL-2Rα levels could be noted based on education, but a 3-fold increase was seen on uneducated NK cells after 72 h of treatment, whereas no increase occurred on educated NK cells (Fig. 4F). These results indicate that NK cell education yields differential regulation of cytokine receptor genes that might be involved in survival.

IL-15 withdrawal promotes cell death on IL-2Rαhi uneducated NK cells through differential Bim and FasL expression

Our results suggest that NK cell education influences IL-2Rα expression, but the role of IL-2Rα on survival remains unclear. To address this point in a more physiologic setting, we used a cytokine-withdrawal assay to model the contraction phase after virus infection in which Ag depletion results in decreased inflammation. NK cells were treated for 48 h with 1 ng/ml IL-15 and then IL-15 was withdrawn for 48 h to look at cell death. IL-15 treatment followed by withdrawal yielded a 1.45-fold difference in IL-2Rα expression on uneducated NK cells compared with their educated counterparts (Fig. 5A). The amount of cell death paralleled the fold difference seen in IL-2Rα expression (Fig. 5B, 1.44-fold increase in cell death on uneducated NK cells). To test whether IL-2Rα expression itself directly correlated with cell death, we gated on IL-2Rαlo and IL-2Rαhi NK cell populations and evaluated cell death post–IL-15 withdrawal. IL-2Rαhi NK cell death was 1.7-fold higher than IL-2Rαlo NK cell death (Fig. 5C), establishing a definitive link between IL-2Rα expression and cell death.

FIGURE 5.
FIGURE 5.

IL-15 withdrawal leads to a disparity in survival proteins on educated versus uneducated NK cells that correlates with IL-2Rα expression. Healthy donor PBMCs were cultured with 1 ng/ml (A, B, C, D, E, F, and G) or 10 ng/ml (H, I, J, and K) IL-15 for 48 h, and cytokine was washed away. Subsequently, PBMCs were cultured in RPMI-10% (serum) without cytokine for another 48 h before analysis. IL-2Rα expression (A) and cell death (B) on self-KIR educated versus nonself-KIR uneducated NK cell subsets after 1 ng/ml IL-15 withdrawal (n = 18). (C) Percentage of NK cell death post–1 ng/ml IL-15 withdrawal based on IL-2Rα (IL-2RA) expression (n = 10). (D and H) Bim expression based on NK cell education after 1 and 10 ng/ml IL-15 withdrawal (n = 18). (E and I) Bim expression based on NK cell IL-2Rα expression after 1 and 10 ng/ml IL-15 withdrawal (n = 10). (F and J) Fas expression based on NK cell education after 1 and 10 ng/ml IL-15 withdrawal (n = 18). (G and K) FasL expression based on NK cell education after 1 and 10 ng/ml IL-15 withdrawal (n = 18). For education experiments, to achieve internal comparison per donor all samples selected had at least one uneducated KIR. *p < 0.05, **p < 0.01, ***p < 0.001.

Next, we asked what molecules could be involved in this process. Although Bcl-2 is found at slightly higher levels on educated NK cells early on (Supplemental Fig. 2B–D), no change was seen under IL-15–withdrawal conditions (Supplemental Fig. 2E). Previous publications (16, 33) indicated that NK cell withdrawal from cytokines results in proapoptotic Bim upregulation. Withdrawal from a 1-ng/ml IL-15 treatment yields a 1.2-fold increase in Bim on uneducated NK cells (Fig. 5D) and IL-2Rαhi NK cells (Fig. 5E). Withdrawal from a 10-ng/ml IL-15 treatment augments the Bim differential to 1.4- and 1.3-fold increases on uneducated (Fig. 5H) and IL-2Rαhi (Fig. 5I) NK cells, respectively. Differences can be augmented further by longer initial IL-15–incubation periods, followed by withdrawal from both IL-15 and serum (Supplemental Fig. 3A–C). Although withdrawal from 1 ng/ml IL-15 resulted in higher Fas on uneducated NK cells (Fig. 5F), initial incubation with 10 ng/ml IL-15 negated this difference (Fig. 5J). Surprisingly, educated NK cells expressed more FasL when IL-15 was withdrawn (Fig. 5G, 5K, 1.4- and 1.5-fold, respectively). These data indicate that, when cytokine is limiting, NK cell education and absence of IL-2Rα expression are advantageous for cell survival through decreased Bim and Fas expression. Furthermore, educated NK cells might actively kill uneducated NK cells through FasL when IL-15 is scarce, perhaps to compete for cytokine.

IL-2Rα directly impacts survival through FasL mediated by NK–NK interactions

To test whether IL-2Rα had a direct role in NK cell survival, NK cells were purified, and IL-2Rα was knocked down or overexpressed in the IL-15–withdrawal setting (Fig. 6A). Although IL-2 signaling can enhance AICD, there was no IL-2 or cross-linking signals present in these cultures. Therefore, cell death in this setting is independent of IL-2 signaling and must be attributed to another mechanism. The majority of the cells expressed siRNAs on the day harvested (Supplemental Fig. 3D). siRNA knockdown of IL-2Rα (siIL2RA) led to a 1.5-fold increase in NK cell numbers (Fig. 6B) compared with the siRNA control (siC). Similarly, the control had 1.5-fold more cells than did the samples in which IL-2Rα was overexpressed (oIL2RA). To ensure that the changes were due to differential cell death (and not proliferation), the proportion of dying cells was evaluated. A 1.4-fold decrease in cell death was observed in the IL-2Rα knockdown compared with the control (Fig. 6C). The data indicate that the mechanism for differential survival of NK cells is controlled, at least in part, by the amount of IL-2Rα expressed, independent of IL-2 signaling and AICD.

FIGURE 6.
FIGURE 6.

IL-2RA expression determines NK cell survival and renders IL-2RAhi NK cells sensitive to FasL-mediated “attack” from KIR+ NK cells. NK cells were enriched by bead separation and placed in 1 ng/ml IL-15 overnight. The next day, they were transduced with a negative control siRNA containing an Alexa Fluor 488 tag (siC), a set of four siRNAs specific for human IL-2Rα (siIL2RA), or a plasmid to overexpress IL-2Rα (oIL2RA). After transduction, the cells were kept in 1 ng/ml IL-15 for another 24 h. IL-15 was then withdrawn for the next 48 h, and cells were analyzed using flow cytometry. (A) Representative graphs of transduction efficiency showing IL-2Rα levels in the control (siC), the knockdown (siIL2RA), and the overexpression vector (oIL2RA). (B) Events were collected for 60 s at the same rate for each sample, and the total number of NK cells collected during this period was calculated by gating on live CD56+CD3− NK cells. Aggregate data are presented for each of the three treatment groups (n = 4). (C) The proportion of dead NK cells, 3 d posttransduction, was measured by looking at LIVE/DEAD dye incorporation in CD56+CD3− NK cells. Aggregate data for each of the three treatment groups are shown (n = 4). (D) Representative dot plots showing IL-2Rα expression in mock-transduced (left panel) or IL-2Rα–transduced (center panel) KIR+ and KIR− NK cells. The dot plot (right panel) shows eGFP levels in a representative control eGFP transduction in KIR+ versus KIR− NK cells. Large bolded numbers show proportion of expression (IL-2Rα [left and center panels] or eGFP [right panel]) in the positive population of KIR+ or KIR− NK cells. (E) Quantification of differential expression of overexpressed proteins in KIR+ versus KIR− NK cells. The graph shows the ratio (KIR+/KIR−) of protein expression after transduction, where a value of 1 indicates equal protein expression in both populations (n = 4). (F) Schematic diagram for FasL NK cell–NK cell killing experiment. NK cells were enriched from healthy donor PBMCs; for each donor, cells were either frozen for use as “targets” or bead separated into KIR+ and KIR− NK cells for use as “killers.” Killers were treated with 10 ng/ml IL-15 for 72 h (in RPMI-10% serum) and then IL-15 was withdrawn for 48 h by changing media to RPMI-10% to drive FasL expression on killers. Killers were then labeled with CellTrace and cocultured with NK targets from the same donor in the presence or absence of FasL-blocking Ab (NOK-1; BioLegend). After 24 h of coculture, cells were harvested and stained. For analysis, killers (CellTrace+ NK cells) were excluded, and the proportion of live cells entering apoptosis (LIVE/DEAD dye−/annexin V+) was evaluated on IL-2RAlo and IL-2RAhi targets by flow cytometry. (G) Analysis of annexin V expression on targets after 24 h incubation with killers, with or without FasL-blocking Ab (n = 4). *p < 0.05, **p < 0.01.

Overexpression of IL-2Rα was not as efficient in the KIR+ NK cells as it was in the KIR− NK cells (Fig. 6D, center panel). To ensure that this was not due to a differential transduction of KIR+ NK cells versus KIR− NK cells, we transduced cells with the same construct expressing eGFP alone. There were no significant differences in transduction efficiency between KIR+ and KIR− cells, excluding this possibility (Fig. 6D, right panel) and leading to the conclusion that there is an active mechanism for modulation of IL-2Rα expression in KIR+ NK cells, despite forced expression (Fig. 6E).

To connect the role of IL-2Rα in sensitizing NK cells to death with the finding that educated NK cells upregulate FasL when IL-15 is limiting, KIR+-containing educated NK cells and KIR−-containing uneducated NK cells were purified to test their ability to kill NK cells (Fig. 6F). Donor NK cells were saved as “targets” before separation into KIR+ or KIR− “killers.” Killers were subjected to IL-15 withdrawal (to induce FasL expression), labeled (to allow for exclusion during analysis), and incubated overnight with NK targets from the same donor, with or without FasL Ab blockade. As expected, IL-2Rαhi targets were more sensitive to apoptosis than IL-2Rαlo targets (Fig. 6G). Death in IL-2Rαlo targets remained low, regardless of treatment. In contrast, IL-2Rαhi targets were more sensitive (by ∼1.65-fold) to apoptosis when incubated with KIR+-containing educated NK cell killers than with KIR− NK cells. This fold difference is similar to the difference in apoptosis in IL-2Rαhi targets when comparing NK targets alone versus NK targets incubated with KIR+ killers overnight (Supplemental Fig. 3E). The IL-2Rαhi target sensitivity to apoptosis was decreased when FasL was blocked on KIR+ killers, but no changes were seen when FasL was blocked on KIR− killers. Knockdown of Fas on NK cell targets (Supplemental Fig. 3F) similarly decreased the ability of KIR+ killers to induce apoptosis on targets (Supplemental Fig. 3G), confirming involvement of this pathway in NK cell fratricide. These results indicate that populations enriched with educated NK cells induce cell death autologously on IL-2Rαhi NK cells when IL-15 is limiting, partially through FasL signaling.

NK cell survival is enhanced after CMV reactivation

We wanted to investigate whether survival of educated NK cells is enhanced under pathologic settings known to activate NK cells. We showed that CMV reactivation after transplant results in a clonal expansion of educated NK cells that is maintained long-term after transplant, suggesting NK cell memory (7, 8). In an adult donor allogeneic hematopoietic transplantation setting, expression of KIR (Fig. 7A) or CD57 (Fig. 7B) on NK cells, both markers of maturity, resulted in diminished cell death at all time points after transplant. We hypothesized that CMV reactivation would result in selection of educated NK cells through better survival. In an adult allogeneic transplantation setting in which 50% of normal adult donors may have had prior exposure capable of passing on memory NK cells, CMV reactivation resulted in better NK cell survival, which was statistically significant at 6 mo after transplantation (Fig. 7C). To differentiate between primary and secondary CMV activation, we used umbilical cord blood transplantation in which donor units are CMV naive, and activation represents primary infection in the recipient of the newly developing donor NK cells. CMV activation before day 100 after umbilical cord transplantation protected day-100 posttransplantation NK cells from cell death (Fig. 7D). These results indicate that enhanced survival of more-differentiated NK cells after primary and secondary CMV activation promotes the long-lived repertoire of functional NK cells seen in this setting (7, 8).

FIGURE 7.
FIGURE 7.

Increased survival properties of NK cells can be seen posttransplantation, and NK cell survival is accentuated by CMV reactivation. NK cell survival was measured based on KIR (A) or CD57 (B) expression on NK cells from adult donor allogeneic transplant patients at day 28 (n = 28), day 100 (n = 25), and 6 mo (n = 27) posttransplant. (C) NK cell survival was segregated by CMV reactivation (CMVR) in adult donor allogeneic transplantation patients (no reactivation = CMVR− [●]: day 28 [n = 19], day 100 [n = 18], and 6 mo [n = 19]; CMV reactivation = CMVR+ [○]: day 28 [n = 9], day 100 or 8 wk postviral diagnosis [n = 7], and 6 mo [n = 8]). (D) NK cell survival at 100 d (CMV−, n = 12) or 8 wk (CMV+, n = 11) postviral diagnosis after umbilical cord allogeneic transplant. *p < 0.05, ***p < 0.001.

Discussion

We propose a model of long-term NK cell homeostasis determined largely by NK cell survival mechanisms. In this model, proliferation and differentiation form the functional NK cell repertoire, whereas enhanced survival by interaction with self-MHC maintains these educated NK cells. This process is tunable, and additive signals through NKG2A and KIR enhance survival, much like what is seen in terms of function (4, 29–31). Sensitivity to cell death is linked to reduced Bim expression, in agreement with another report (16), as well as modest increases in antiapoptotic proteins Bcl-2 and Bcl-xL. Survival is negatively modulated by IL-2Rα expression, and educated NK cells are less susceptible to Fas-driven AICD. Educated NK cells have the capacity to dominate over their hyporesponsive counterparts by inducing death of IL-2Rαhi NK cells, which are enriched in uneducated NK cell populations, through FasL when IL-15 is limiting. These data show that educated NK cells persist longer and control the survival of uneducated cells, a mechanism that we refer to as NK cell probation. The rapidly proliferating uneducated NK cells are in a “probationary” phase marked by high expression of IL-2Rα and Fas, whereas educated NK cells expressing FasL are the “probation officers” that remove hyporesponsive NK cells. In this context, NK cell education is required for acquisition of function, whereas NK cell probation shapes the repertoire by preferentially maintaining more functional cells and allowing clonal dominance by direct NK–NK interactions.

NK cell probation could be explained by activation of the PI3K/Akt pathway through cross-linking of KIRs on NK3.3 cells (12). However, these data are at odds with data on murine SHIP−/−, which is activated downstream of Ly49 in mice and analogous to SHP-1 in humans, showing that SHIP opposes PI3K-mediated Akt activation in mice (34, 35). A more recent study (16) showed that dose-dependent escalation of cognate MHC signals yielded increased selection of educated NK cells, consistent with the probation hypothesis.

Cytokine receptor balance plays a key role in probation, with IL-2Rα expression leading to increased cell death, whereas the common γ-chain is associated with enhanced survival. The lack of change observed in IL-15Rα on NK cells might be explained by the fact that this protein is thought to be trans-presented (36, 37). Given these results, we hypothesize that IL-2Rα might directly influence survival by competing for common signaling components with trans-presented IL-15Rα, whose signals yield greater survival (18, 19, 38). This hypothesis is supported by the fact that cell death and Bim expression correlate well with IL-2Rα levels after withdrawal and the finding that manipulation of IL-2Rα directly influences survival in an IL-2–free system; however, more experiments are needed to fully prove the mechanism. The concept is that IL-2Rαhi NK cells are sensitized to proapoptotic signals, particularly FasL, presented by more educated NK cells during times of stress.

These findings have important clinical implications in NK cell–mediated immunotherapies (1, 2). In both T cell– and NK cell–mediated cell therapy, efficacy seems to correlate best with the persistence of lymphocytes after adoptive transfer (39, 40). For NK cells, survival could be modulated by targeting expression of IL-2Rα through lentiviral knockdown in the NK cell–infusion product or manipulating IL-2–signaling negative-feedback loops (32) by treating with IL-2, perhaps following pharmacologic “inflammation” using IL-15, to reduce the effect of IL-2Rα on NK cells in both infusion and transplant patients. In addition, blocking Fas/FasL interactions during the NK cell expansion stage during transplants could lead to increased NK cell reconstitution. Conversely, maximal NK cell functionality in infusion products also could be achieved by treating NK cells with multiple rounds of IL-15, followed by IL-15 withdrawal, resulting in an enriched population of educated NK cells. Finally, the concept of NK cell memory has relevance for the longevity of cell-based immunotherapy, as well as the observation that memory cells exhibit enhanced function (7–9). We show in this study that CMV reactivation results in better survival of NK cells. In some settings, this correlates with protection from leukemia relapse (41, 42). Gaining a better understanding of how survival is mediated will help to enhance future use of NK memory products for immunotherapy. In conclusion, this study presents novel findings showing that NK cell differentiation and education have an important role in NK cell survival through the process of education/probation, which is mediated by IL-2Rα expression, Bim, minor changes in antiapoptotic molecules, and “predatory” FasL expression when cytokine is limiting.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank the Tolar laboratory for assistance with the mouse experiments.

Footnotes

  • This work was supported in part by the National Cancer Institute, National Institutes of Health, under contracts HHSN261200800001E, 2T32HL007062-36, CA111412, and CA65493. This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Center for Cancer Research.

  • The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    AICD
    activation-induced cell death
    FasL
    Fas ligand
    KIR
    killer Ig-like receptor
    MFI
    median fluorescence intensity
    NSG
    NOD scid IL-2R γ-chain knockout
    siC
    small interfering RNA control
    siRNA
    small interfering RNA.

  • Received September 26, 2013.
  • Accepted February 12, 2014.

References

  1. ↵
    1. Murphy W. J.,
    2. P. Parham,
    3. J. S. Miller
    . 2012. NK cells–from bench to clinic. Biol. Blood Marrow Transplant. 18 (1 Suppl.): S2–7.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Thielens A.,
    2. E. Vivier,
    3. F. Romagné
    . 2012. NK cell MHC class I specific receptors (KIR): from biology to clinical intervention. Curr. Opin. Immunol. 24: 239–245.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Anfossi N.,
    2. P. André,
    3. S. Guia,
    4. C. S. Falk,
    5. S. Roetynck,
    6. C. A. Stewart,
    7. V. Breso,
    8. C. Frassati,
    9. D. Reviron,
    10. D. Middleton,
    11. et al
    . 2006. Human NK cell education by inhibitory receptors for MHC class I. Immunity 25: 331–342.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Brodin P.,
    2. T. Lakshmikanth,
    3. S. Johansson,
    4. K. Kärre,
    5. P. Höglund
    . 2009. The strength of inhibitory input during education quantitatively tunes the functional responsiveness of individual natural killer cells. Blood 113: 2434–2441.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Kim S.,
    2. J. Poursine-Laurent,
    3. S. M. Truscott,
    4. L. Lybarger,
    5. Y. J. Song,
    6. L. Yang,
    7. A. R. French,
    8. J. B. Sunwoo,
    9. S. Lemieux,
    10. T. H. Hansen,
    11. W. M. Yokoyama
    . 2005. Licensing of natural killer cells by host major histocompatibility complex class I molecules. Nature 436: 709–713.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Sun J. C.,
    2. L. L. Lanier
    . 2011. NK cell development, homeostasis and function: parallels with CD8⁺ T cells. Nat. Rev. Immunol. 11: 645–657.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Foley B.,
    2. S. Cooley,
    3. M. R. Verneris,
    4. M. Pitt,
    5. J. Curtsinger,
    6. X. Luo,
    7. S. Lopez-Vergès,
    8. L. L. Lanier,
    9. D. Weisdorf,
    10. J. S. Miller
    . 2012. Cytomegalovirus reactivation after allogeneic transplantation promotes a lasting increase in educated NKG2C+ natural killer cells with potent function. Blood 119: 2665–2674.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Foley B.,
    2. S. Cooley,
    3. M. R. Verneris,
    4. J. Curtsinger,
    5. X. Luo,
    6. E. K. Waller,
    7. C. Anasetti,
    8. D. Weisdorf,
    9. J. S. Miller
    . 2012. Human cytomegalovirus (CMV)-induced memory-like NKG2C(+) NK cells are transplantable and expand in vivo in response to recipient CMV antigen. J. Immunol. 189: 5082–5088.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Sun J. C.,
    2. J. N. Beilke,
    3. L. L. Lanier
    . 2009. Adaptive immune features of natural killer cells. Nature 457: 557–561.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Björkström N. K.,
    2. P. Riese,
    3. F. Heuts,
    4. S. Andersson,
    5. C. Fauriat,
    6. M. A. Ivarsson,
    7. A. T. Björklund,
    8. M. Flodström-Tullberg,
    9. J. Michaëlsson,
    10. M. E. Rottenberg,
    11. et al
    . 2010. Expression patterns of NKG2A, KIR, and CD57 define a process of CD56dim NK-cell differentiation uncoupled from NK-cell education. Blood 116: 3853–3864.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Lopez-Vergès S.,
    2. J. M. Milush,
    3. S. Pandey,
    4. V. A. York,
    5. J. Arakawa-Hoyt,
    6. H. Pircher,
    7. P. J. Norris,
    8. D. F. Nixon,
    9. L. L. Lanier
    . 2010. CD57 defines a functionally distinct population of mature NK cells in the human CD56dimCD16+ NK-cell subset. Blood 116: 3865–3874.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Marti F.,
    2. C. W. Xu,
    3. A. Selvakumar,
    4. R. Brent,
    5. B. Dupont,
    6. P. D. King
    . 1998. LCK-phosphorylated human killer cell-inhibitory receptors recruit and activate phosphatidylinositol 3-kinase. Proc. Natl. Acad. Sci. USA 95: 11810–11815.
    OpenUrlAbstract/FREE Full Text
    1. Ugolini S.,
    2. C. Arpin,
    3. N. Anfossi,
    4. T. Walzer,
    5. A. Cambiaggi,
    6. R. Förster,
    7. M. Lipp,
    8. R. E. Toes,
    9. C. J. Melief,
    10. J. Marvel,
    11. E. Vivier
    . 2001. Involvement of inhibitory NKRs in the survival of a subset of memory-phenotype CD8+ T cells. Nat. Immunol. 2: 430–435.
    OpenUrlPubMed
    1. Vivier E.,
    2. N. Anfossi
    . 2004. Inhibitory NK-cell receptors on T cells: witness of the past, actors of the future. Nat. Rev. Immunol. 4: 190–198.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Young N. T.,
    2. M. Uhrberg,
    3. J. H. Phillips,
    4. L. L. Lanier,
    5. P. Parham
    . 2001. Differential expression of leukocyte receptor complex-encoded Ig-like receptors correlates with the transition from effector to memory CTL. J. Immunol. 166: 3933–3941.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Brodin P.,
    2. T. Lakshmikanth,
    3. K. Kärre,
    4. P. Höglund
    . 2012. Skewing of the NK cell repertoire by MHC class I via quantitatively controlled enrichment and contraction of specific Ly49 subsets. J. Immunol. 188: 2218–2226.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Meazza R.,
    2. B. Azzarone,
    3. A. M. Orengo,
    4. S. Ferrini
    . 2011. Role of common-gamma chain cytokines in NK cell development and function: perspectives for immunotherapy. J. Biomed. Biotechnol. 2011: 861920.
    OpenUrlPubMed
  18. ↵
    1. Ring A. M.,
    2. J. X. Lin,
    3. D. Feng,
    4. S. Mitra,
    5. M. Rickert,
    6. G. R. Bowman,
    7. V. S. Pande,
    8. P. Li,
    9. I. Moraga,
    10. R. Spolski,
    11. et al
    . 2012. Mechanistic and structural insight into the functional dichotomy between IL-2 and IL-15. Nat. Immunol. 13: 1187–1195.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Carson W. E.,
    2. T. A. Fehniger,
    3. S. Haldar,
    4. K. Eckhert,
    5. M. J. Lindemann,
    6. C. F. Lai,
    7. C. M. Croce,
    8. H. Baumann,
    9. M. A. Caligiuri
    . 1997. A potential role for interleukin-15 in the regulation of human natural killer cell survival. J. Clin. Invest. 99: 937–943.
    OpenUrlCrossRefPubMed
    1. Kennedy M. K.,
    2. M. Glaccum,
    3. S. N. Brown,
    4. E. A. Butz,
    5. J. L. Viney,
    6. M. Embers,
    7. N. Matsuki,
    8. K. Charrier,
    9. L. Sedger,
    10. C. R. Willis,
    11. et al
    . 2000. Reversible defects in natural killer and memory CD8 T cell lineages in interleukin 15-deficient mice. J. Exp. Med. 191: 771–780.
    OpenUrlAbstract/FREE Full Text
    1. Lodolce J. P.,
    2. D. L. Boone,
    3. S. Chai,
    4. R. E. Swain,
    5. T. Dassopoulos,
    6. S. Trettin,
    7. A. Ma
    . 1998. IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation. Immunity 9: 669–676.
    OpenUrlCrossRefPubMed
    1. Puel A.,
    2. S. F. Ziegler,
    3. R. H. Buckley,
    4. W. J. Leonard
    . 1998. Defective IL7R expression in T(−)B(+)NK(+) severe combined immunodeficiency. Nat. Genet. 20: 394–397.
    OpenUrlCrossRefPubMed
    1. Roifman C. M.,
    2. J. Zhang,
    3. D. Chitayat,
    4. N. Sharfe
    . 2000. A partial deficiency of interleukin-7R alpha is sufficient to abrogate T-cell development and cause severe combined immunodeficiency. Blood 96: 2803–2807.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Willerford D. M.,
    2. J. Chen,
    3. J. A. Ferry,
    4. L. Davidson,
    5. A. Ma,
    6. F. W. Alt
    . 1995. Interleukin-2 receptor alpha chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3: 521–530.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Nelson B. H.,
    2. D. M. Willerford
    . 1998. Biology of the interleukin-2 receptor. Adv. Immunol. 70: 1–81.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Ortaldo J. R.,
    2. R. T. Winkler-Pickett,
    3. S. Nagata,
    4. C. F. Ware
    . 1997. Fas involvement in human NK cell apoptosis: lack of a requirement for CD16-mediated events. J. Leukoc. Biol. 61: 209–215.
    OpenUrlAbstract
  27. ↵
    1. Gleason M. K.,
    2. T. R. Lenvik,
    3. V. McCullar,
    4. M. Felices,
    5. M. S. O’Brien,
    6. S. A. Cooley,
    7. M. R. Verneris,
    8. F. Cichocki,
    9. C. J. Holman,
    10. A. Panoskaltsis-Mortari,
    11. et al
    . 2012. Tim-3 is an inducible human natural killer cell receptor that enhances interferon gamma production in response to galectin-9. Blood 119: 3064–3072.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Hommel G.
    1988. A stagewise rejective multiple test procedure based on a modified Bonferroni test. Biometrika 75: 383–386.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Elliott J. M.,
    2. W. M. Yokoyama
    . 2011. Unifying concepts of MHC-dependent natural killer cell education. Trends Immunol. 32: 364–372.
    OpenUrlCrossRefPubMed
    1. Höglund P.,
    2. P. Brodin
    . 2010. Current perspectives of natural killer cell education by MHC class I molecules. Nat. Rev. Immunol. 10: 724–734.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Orr M. T.,
    2. L. L. Lanier
    . 2010. Natural killer cell education and tolerance. Cell 142: 847–856.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Pillet A. H.,
    2. J. Thèze,
    3. T. Rose
    . 2011. Interleukin (IL)-2 and IL-15 have different effects on human natural killer lymphocytes. Hum. Immunol. 72: 1013–1017.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Huntington N. D.,
    2. C. A. Vosshenrich,
    3. J. P. Di Santo
    . 2007. Developmental pathways that generate natural-killer-cell diversity in mice and humans. Nat. Rev. Immunol. 7: 703–714.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Young N. T.,
    2. M. Uhrberg
    . 2002. KIR expression shapes cytotoxic repertoires: a developmental program of survival. Trends Immunol. 23: 71–75.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Wang J. W.,
    2. J. M. Howson,
    3. T. Ghansah,
    4. C. Desponts,
    5. J. M. Ninos,
    6. S. L. May,
    7. K. H. Nguyen,
    8. N. Toyama-Sorimachi,
    9. W. G. Kerr
    . 2002. Influence of SHIP on the NK repertoire and allogeneic bone marrow transplantation. Science 295: 2094–2097.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Huntington N. D.,
    2. N. Legrand,
    3. N. L. Alves,
    4. B. Jaron,
    5. K. Weijer,
    6. A. Plet,
    7. E. Corcuff,
    8. E. Mortier,
    9. Y. Jacques,
    10. H. Spits,
    11. J. P. Di Santo
    . 2009. IL-15 trans-presentation promotes human NK cell development and differentiation in vivo. J. Exp. Med. 206: 25–34.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Bergamaschi C.,
    2. J. Bear,
    3. M. Rosati,
    4. R. K. Beach,
    5. C. Alicea,
    6. R. Sowder,
    7. E. Chertova,
    8. S. A. Rosenberg,
    9. B. K. Felber,
    10. G. N. Pavlakis
    . 2012. Circulating IL-15 exists as heterodimeric complex with soluble IL-15Rα in human and mouse serum. Blood 120: e1–e8.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Waldmann T. A.
    2006. The biology of interleukin-2 and interleukin-15: implications for cancer therapy and vaccine design. Nat. Rev. Immunol. 6: 595–601.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Grupp S. A.,
    2. M. Kalos,
    3. D. Barrett,
    4. R. Aplenc,
    5. D. L. Porter,
    6. S. R. Rheingold,
    7. D. T. Teachey,
    8. A. Chew,
    9. B. Hauck,
    10. J. F. Wright,
    11. et al
    . 2013. Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N. Engl. J. Med. 368: 1509–1518.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Porter D. L.,
    2. B. L. Levine,
    3. M. Kalos,
    4. A. Bagg,
    5. C. H. June
    . 2011. Chimeric antigen receptor-modified T cells in chronic lymphoid leukemia. N. Engl. J. Med. 365: 725–733.
    OpenUrlCrossRefPubMed
  41. ↵
    1. Elmaagacli A. H.,
    2. N. K. Steckel,
    3. M. Koldehoff,
    4. Y. Hegerfeldt,
    5. R. Trenschel,
    6. M. Ditschkowski,
    7. S. Christoph,
    8. T. Gromke,
    9. L. Kordelas,
    10. H. D. Ottinger,
    11. et al
    . 2011. Early human cytomegalovirus replication after transplantation is associated with a decreased relapse risk: evidence for a putative virus-versus-leukemia effect in acute myeloid leukemia patients. Blood 118: 1402–1412.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Ito S.,
    2. P. Pophali,
    3. W. Co,
    4. E. K. Koklanaris,
    5. J. Superata,
    6. G. A. Fahle,
    7. R. Childs,
    8. M. Battiwalla,
    9. A. J. Barrett
    . 2013. CMV reactivation is associated with a lower incidence of relapse after allo-SCT for CML. Bone Marrow Transplant. 48: 1313–1316.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section