Invariant NKT (iNKT) cells are innate-like T lymphocytes that recognize and respond to glycolipid Ags such as α-galactosylceramide (α-GalCer). This unique property has been exploited in clinical trials for multiple malignancies. While investigating mouse iNKT cell responses to α-GalCer in vivo, we found a dramatically enlarged tissue-resident population surprisingly coexpressing select dendritic cell, NK cell, and B cell markers. Further phenotypic and functional analyses revealed the identity of this B220+CD11c+MHC class II+NK1.1+ population as precursors to mature NK (pre-mNK) cells, which also expressed high levels of proliferation and tissue retention markers but diminished sphingosine-1-phosphate receptor 1, a receptor that facilitates tissue trafficking. Accordingly, FTY720, a sphingosine-1-phosphate receptor 1 antagonist, failed to prevent pre-mNK cells’ intrahepatic accumulation. We found iNKT cell–driven expansion of pre-mNK cells to be dependent on IL-12 and IL-18. Although α-GalCer–transactivated pre-mNK cells lost their capacity to process a model tumor Ag, they selectively expressed granzyme A and directly lysed YAC-1 thymoma cells through granule exocytosis. They also contributed to β2 microglobulin–deficient target cell destruction in vivo. Therefore, α-GalCer treatment skewed pre-mNK cell responses away from an APC-like phenotype and toward killer cell–like functions. Finally, the ability of α-GalCer to reduce the pulmonary metastatic burden of B16-F10 mouse melanoma was partially reversed by in vivo depletion of pre-mNK cells. To our knowledge, our findings shed new light on iNKT cells’ mechanism of action and glycolipid-based immunotherapies. Therefore, we introduce pre-mNK cells as a novel downstream effector cell type whose anticancer properties may have been overlooked in previous investigations.

Invariant NKT (iNKT) cells comprise a specialized subset of innate-like T lymphocytes with remarkable immunomodulatory properties. They express a semi-invariant TCR with a distinctive Vα14-Jα18 configuration that uniquely recognizes naturally occurring and synthetic glycolipid Ags presented within the close cleft of CD1d (1, 2). Unlike their conventional counterparts, iNKT cells occur in a partially activated state and harbor preformed mRNA transcripts coding for several inflammatory cytokines, which they can release amply and speedily (3).

iNKT cells have been heralded for their roles in anticancer immune surveillance (4, 5). They can express perforin, granzymes (GZMs), Fas ligand, TNF-α, and TRAIL (68) and directly destroy glycolipid-pulsed mouse target cells in vivo (9) and human CD1d+ tumor cells in vitro (8, 10). Arguably, however, the most pronounced feature of iNKT cells is their ability to transactivate APCs (11, 12) and multiple effector cell types, including NK cells (13, 14), conventional CD8+ T cells (15), and γδ T cells (16).

CD1d-restricted agonists of iNKT cells are typified by α-galactosylceramide (α-GalCer). This α-anomeric glycolipid was initially extracted from a marine sponge, Agelas mauritanius, in a screen for anticancer compounds (2, 17) but also exists, in minute quantities, in mammalian cells (18). Human iNKT cells can recognize mouse CD1d and vice versa (19). Furthermore, both mouse and human iNKT cells are responsive to α-GalCer (4). Therefore, the findings of tumor models in which α-GalCer is therapeutically tested are potentially translatable to the clinic. In addition, the monomorphic nature of CD1d dictates that α-GalCer–based or –adjuvanted treatments should target iNKT cells in diverse human populations beyond the MHC restriction barrier.

In mice, α-GalCer has been employed against thymoma (20), melanoma (21), and pancreatic adenocarcinoma (22), among other cancers. The beneficial properties of α-GalCer, α-GalCer–pulsed dendritic cells (DCs), and α-GalCer–expanded iNKT cells have also been explored and exploited in clinical trials for multiple human neoplasms (4, 5). Despite encouraging results arising from such efforts, the full therapeutic potential of glycolipid-activated iNKT cells remains to be fully realized, and the range of downstream effectors they transactivate may need to be revisited or even revised.

While investigating in vivo responses to α-GalCer, we serendipitously found a dramatically expanded tissue-resident population with prominent anticancer functions, which turned out to be precursors to mature NK (mNK) (pre-mNK) cells. Pre-mNK cells are phenotypically defined as B220+CD11c+MHC class II (MHC-II)+NK1.1+ cells and comprise an intermediate in the differentiation pathway of conventional NK cells. They were initially called IFN-producing killer DCs (IKDCs), due ostensibly to their versatility in migrating to lymph nodes to present Ags, in synthesizing copious amounts of IL-12, type I IFN, and IFN-γ, and in exerting tumoricidal activities (23, 24). Subsequently, IKDCs were reported to belong to the NK lineage as evidenced by their strict developmental dependence on IL-15 (2527) and their propensity to acquire an mNK cell phenotype upon adoptive transfer (28). In fact, pre-mNK cells were demonstrated to be superior to mNK cells in terms of IFN-γ and TNF-α secretion (25).

In this study, we report that priming with α-GalCer expands, activates, and alters select tissue-resident pre-mNK cells through an IL-12/IL-18–dependent mechanism and as such adds a powerful new weapon to the immune system’s anticancer arsenal. To our knowledge, our findings define a novel communication axis involving two innate-like effector cell types, namely iNKT and pre-mNK cells, with clear implications for cancer immunotherapy. We propose that the roles fulfilled by pre-mNK cells in the context of α-GalCer–based treatments may have been overlooked because of their phenotypic resemblance to plasmacytoid DCs and mNK cells.

Wild-type (WT) C57BL/6 (B6) mice were purchased from Charles River Laboratories (St. Constant, QC, Canada) or bred in our institutional barrier facility. β2 microglobulin (β2M)−/− and GFP+ mice, on a B6 background, were provided by Drs. Anthony Jevnikar and Steven Kerfoot (Western University, London, ON, Canada), respectively. Age- and sex-matched adult mice were used in all experiments. Our animal use protocols (2010-241 and 2018-093) were reviewed and approved by the Western University Animal Use Subcommittee.

YAC-1 mouse lymphoma cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, GlutaMAX-I, 0.1 mM MEM nonessential amino acids, 1 mM sodium pyruvate, 120 U/ml penicillin, 100 μg/ml streptomycin, and 10 mM HEPES. The chicken OVA-expressing T cell lymphoma line EG7-OVA was supplied by Drs. Jack Bennink and Jonathan Yewdell (National Institute of Allergy and Infectious Diseases, National Institutes of Health) and maintained in RPM1 1640 medium containing 10% FBS and 400 μg/ml G418. B16-F10 mouse melanoma cells were grown in FBS-supplemented MEM-α medium. The B3Z hybridoma with specificity for SIINFEKL:H-2Kb was a gift from Dr. Sameh Basta (Queen’s University, Kingston, ON, Canada). B3Z cells were cultured in IMDM medium and 10% FBS in the presence of G418.

Mice were injected i.p. with 4 μg α-GalCer (Funakoshi) or with a vehicle containing 0.5% Tween 20, 56 mg/ml sucrose, and 7.5 mg/ml histidine. Where indicated and except in serum transfer experiments, a CD1d-blocking mAb (clone 20H2) or a rat IgG1 control (HRPN) was injected i.p. at 500 μg/dose 2 h before and 24 and 48 h after α-GalCer treatment. In a pilot experiment, an anti–common γ chain (γc) mAb (3E12) or a rat IgG2b control (LTF-2) was administered using a similar regimen except the experiment was ended 24 h postglycolipid treatment. To deplete B220+ or CD19+ cells, mice were given 200-μg i.p. injections of an anti-B220 mAb (RA3.3A1/6.1) or an anti-CD19 mAb (1D3) on days 1 and 3 post–α-GalCer priming. All the above mAbs and isotype controls were purchased from Bio X Cell (West Lebanon, NH).

FTY720 (Sigma-Aldrich) was reconstituted in water, diluted in PBS, and injected at 1 or 3 mg/kg i.p (29) 2 h prior to and 24 h following α-GalCer treatment.

Recombinant mouse IFN-γ, IL-12, and IL-18 (PeproTech) were diluted in sterile PBS and administered once via tail vein at final concentrations of 10, 4, and 200 ng, respectively.

Mouse peripheral blood was collected via the saphenous vein, and serum was isolated through centrifugation for 15 min at 17,000 × g. Samples were stored at −80°C until use. ELISA kits from Thermo Fisher Scientific were employed to quantitate the cytokine content of each sample.

α-GalCer– and vehicle-treated mice were sacrificed by cervical dislocation. Spleens and livers were homogenized in PBS using a glass plunger. Hepatic parenchymal cells were removed through centrifugation, without brake, at 700 × g for 12 min in a 33.75% Percoll PLUS solution (GE Healthcare). Lungs were cut into small pieces using sterile scissors. The resulting preparation was then treated for 1 h with 0.5 mg/ml collagenase IV (Sigma) in RPMI 1640 containing 10% FBS while being rotated inside a 37°C incubator. Splenic, hepatic, and pulmonary cell preparations were briefly exposed to ACK lysis buffer to eliminate erythrocytes, washed, and filtered through 70-μm pores of a cell strainer to remove debris. To block Fcγ receptors, splenic cells, hepatic mononuclear cells (HMNCs), and nonparenchymal lung mononuclear cells were incubated for 5 min at 4°C with 20 μl of the 2.4G2 hybridoma supernatant containing a CD16/CD32 mAb. Cells were then stained with fluorophore-conjugated mAbs to surface B220 (clone RA3-6B2), CD11b (M1/70), CD11c (N418), CD27 (LG.7F9), CD49b (DX5), CD69 (H1.2F3), CD107a (LAMP-1) (1D4B), CD122 (TM-b1), CD218a (P3TUNYA), FasL (MFL3), I-Ab (AF6-120.1), NK1.1 (PK136), TRAIL (N2B2), and/or intracellular GZM A (GzA-3G8.5), GZM B (NGZB), IFN-γ (XMG1.2), Ki67 (SolA15), perforin (eBioOMAK-D), and TNF-α (MP6-XT22). All the above mAbs and corresponding isotype controls were from Thermo Fisher Scientific.

Cell surface staining was conducted at 4°C for 30 min. Intracellular staining of freshly isolated cells was performed using an eBioscience Fixation and Permeabilization Buffer Set. To detect cytolytic molecules, splenic cells or HMNCs were first cocultured with YAC-1 tumor cells in the presence of 10 μg/ml brefeldin A (BFA) (Sigma-Aldrich) for 4 h at 37°C and 6% CO2. Cells were then washed, stained for pre-mNK cell surface markers, fixed, permeabilized, stained for indicated intracellular molecules, washed again, and analyzed. For CD107a staining, anti-CD107a mAb was present in cocultures that also contained 2 μM monensin (BioLegend) in addition to BFA.

Surface staining for sphingosine-1-phosphate receptor 1 (S1PR1) was conducted in two steps. First, cells were incubated with an unconjugated anti-S1PR1 mAb (clone no. 713412; R&D Systems) for 30 min at 4°C. This was followed by the addition of allophycocyanin-labeled goat anti-rat IgG F(ab′)2 fragments (R&D Systems) for 30 min at 4°C in the dark. Control samples were incubated first with a control rat IgG2a (clone 2A3) and subsequently with the secondary reagent.

Stained cells were interrogated using a FACSCanto II cytometer, and data were analyzed using FlowJo software (Tree Star).

Two days after α-GalCer or vehicle administration, hepatic B220+CD11c+NK1.1+ cells were purified using a BD FACSAria III cytometer after live gating based on forward and side scatter profiles and doublet exclusion. The purity of sorted pre-mNK cells was always >95%. A Purelink RNA Mini Kit (Thermo Fisher Scientific) was used to extract RNA, which was then converted to cDNA using the Invitrogen SuperScript VILO Master Mix. Quantitative PCR was carried out in Custom TaqMan Array 96-Well Fast Plates (Thermo Fisher Scientific) using a StepOnePlus Real-Time PCR System. The ΔΔ cycle threshold method was employed to assess changes in gene expression.

Sorted hepatic pre-mNK cells were pulsed with 1 μM SIINFEKL for 45 min at 37°C. They were then washed and coincubated at a 1:1 ratio with B3Z cells for 24 h at 37°C and 6% CO2. After washing in PBS, cells were treated with a lysis buffer containing 0.125% Nonidet P-40, 9 mM MgCl2, and 0.1 mM 2-ME in the presence of 5 mM ortho-nitrophenyl-β-d-galactopyranoside as the substrate. Four hours later, the OD of supernatants was determined at 415 nm.

To detect Ag cross-presentation in vivo, B6 mice were injected i.p. with α-GalCer or vehicle 24 h before they received 1 × 106 EG7-OVA tumor cells i.v. Four days later, the frequency of SIINFEKL-presenting pre-mNK cells among HMNCs was determined by flow cytometry using a PE-conjugated mAb that recognizes SIINFEKL:H-2Kb complexes (clone 25-D1.16) (30).

Five days after α-GalCer injection into GFP+ mice, animals were euthanized, and hepatic B220+CD11c+NK1.1+ and B220CD11cNK1.1+ cells were FACS-purified. Sorted populations, which were always >95% pure, were injected i.v. into WT B6 recipients, which were subsequently sacrificed for their liver on days 5 and 15 posttransfer. Nonparenchymal HMNCs were isolated and analyzed for their pre-mNK cell content.

For adoptive serum transfer experiments, α-GalCer–injected donor WT B6 mice were terminally bled at 2-, 6-, 12- and 24-h time points. Two hundred microliters of pooled sera collected at each individual time point were transferred via tail vein into a recipient WT B6 mouse. In a separate setting, 60 μl of serum from the 6-h time point were combined with 200 μl from the 12- or the 24-h time point, as indicated, and injected into a B6 mouse. In both scenarios, recipient animals were given a solitary 500-μg i.p. dose of an anti-CD1d mAb (20H2) 2 h before they received serum.

In several experiments, IFN-γ, IL-12, and/or IL-18 were magnetically removed from serum samples. Briefly, sera were incubated for 10 min at room temperature with 10 μg/ml of biotinylated mAbs to IFN-γ (clone DB1; Thermo Fisher Scientific), IL-12 (C17.8; Thermo Fisher Scientific), and/or IL-18 (93-10C; Medical and Biological Laboratories). Streptavidin-conjugated RapidSpheres (STEMCELL Technologies) were then added to each mixture followed by magnetic separation of the particles as per manufacturer’s instructions. A small aliquot of each cytokine-replete and -depleted sample was stored at −80°C for cytokine measurements to confirm the procedure’s efficacy, and the remainder was adoptively transferred.

YAC-1 cells were labeled with 100 μCi of Na251CrO4 for 90 min at 37°C, washed with PBS, and used as target cells. Bulk HMNCs or purified hepatic pre-mNK cells from α-GalCer– and vehicle-treated mice were employed at indicated effector/target ratios against 51Cr-labeled YAC-1 cells in 96-well microplates. After a 4-h incubation period at 37°C and 6% CO2, plates were spun before a 100-μl supernatant aliquot was collected from each well and read by a gamma counter. In several experiments, a combination of EGTA (3 mM) and MgCl2 (2 mM) was added to cocultures to block granule exocytosis (31). Alternatively, 100 nM concanamycin A (CMA) was used to pretreat effector cells for 1 h and was also present during 4-h cocultures (32). Experimental release values were determined in supernatant samples from wells containing both effector and target cells. Spontaneous release and total release were obtained from wells in which target cells were suspended in medium only or in 1% Triton X-100, respectively. Specific killing of YAC-1 cells was calculated using the following formula: % specific lysis = [(experimental release − spontaneous release)/(total release − spontaneous release)] × 100.

Erythrocyte-depleted syngeneic target splenocytes were prepared from WT and β2M−/− B6 mice and labeled with 0.2 and 2 μM CFSE, respectively. Target cells were washed, mixed in equal numbers, and injected at 1 × 107 cells in 200 μl PBS into the tail vein of recipient B6 mice that had been treated 5 d earlier with α-GalCer or vehicle. In several experiments, B220+ or CD19+ cells were depleted. Recipient mice were euthanized 3 h later, and CFSE-labeled target cells were detected among HMNCs by flow cytometry. A constant number of CFSElow events, typically 2 × 103, were acquired for each sample, and percentage of specific killing of target cell populations was calculated using the following equation: % specific cytotoxicity = {1 − [(CFSEhigh event number in recipient mouse/CFSElow event number in recipient mouse)/(CFSEhigh event number among mixed target cells before injection/CFSElow event number among mixed target cells before injection)]} × 100.

B6 mice received α-GalCer or vehicle 24 h prior to an i.v. injection of 5 × 105 B16-F10 melanoma cells. In a limited number of experiments, α-GalCer–treated animals also received an anti-B220 mAb or an anti-CD19 mAb as described above. Fourteen days after melanoma cell injections, α-GalCer– and/or vehicle-treated mice were euthanized, and pulmonary metastatic nodules were enumerated.

Statistical comparisons were made using Graphpad Prism 5 software. We used unpaired Student t tests and ANOVA as appropriate. Statistically significant differences were reported as p < 0.05, p < 0.01, and p < 0.001, which were denoted by *, **, and ***, respectively.

While investigating hepatocellular responses to α-GalCer, we found a prominent cell population with forward and side scatter profiles consistent with those of lymphocytes (Fig. 1A). At first glance, this was not surprising because α-GalCer induces an early and robust proliferative burst in the iNKT cell compartment (33). Intriguingly, however, a substantial proportion of the enlarged population expressed CD11c, a classic mouse DC marker that is not typically expressed by iNKT cells. This prompted us to further their characterization. We found them to also express B220 and NK1.1 (Fig. 1A), which are usually considered B cell/plasmacytoid DCs and NK cell markers, respectively. The only cell type that concomitantly expresses CD11c, B220, and NK1.1 is the pre-mNK cell (23, 24) whose tissue presence has been reported in the bone marrow, spleen, and lymph nodes (23, 24, 26, 34, 35) but never in the liver before.

FIGURE 1.

In vivo stimulation of iNKT cells with α-GalCer enlarges a hepatocellular population phenotypically resembling pre-mNK cells. (A) B6 mice (n = 3) were injected i.p. with either vehicle or α-GalCer. Five days later, HMNCs were isolated and analyzed cytofluorimetrically for their forward and side scatter characteristics and for their reactivity with mAbs against CD11c, B220, and NK1.1. Representative plots are illustrated. (B) CD3CD19 HMNCs from α-GalCer–treated mice (n = 3) were further immunophenotyped using mAbs to B220, CD11c, CD49b, CD122, and I-A/E. (C) B6 mice (n = 4/group) were pretreated with a CD1d-blocking mAb or isotype control 2 h before and once every 24 h after α-GalCer administration as applicable. Animals were bled, and serum IL-4 and IFN-γ concentrations were quantified at 2 and 6 h post–α-GalCer injection, respectively. On day 3, B220+CD11c+NK1.1+ cell frequencies among bulk HMNCs were determined by flow cytometry. Error bars represent mean ± SEM values. Filled histograms in (A) and (B) represent staining with appropriate isotype controls. **p < 0.01, unpaired Student t test.

FIGURE 1.

In vivo stimulation of iNKT cells with α-GalCer enlarges a hepatocellular population phenotypically resembling pre-mNK cells. (A) B6 mice (n = 3) were injected i.p. with either vehicle or α-GalCer. Five days later, HMNCs were isolated and analyzed cytofluorimetrically for their forward and side scatter characteristics and for their reactivity with mAbs against CD11c, B220, and NK1.1. Representative plots are illustrated. (B) CD3CD19 HMNCs from α-GalCer–treated mice (n = 3) were further immunophenotyped using mAbs to B220, CD11c, CD49b, CD122, and I-A/E. (C) B6 mice (n = 4/group) were pretreated with a CD1d-blocking mAb or isotype control 2 h before and once every 24 h after α-GalCer administration as applicable. Animals were bled, and serum IL-4 and IFN-γ concentrations were quantified at 2 and 6 h post–α-GalCer injection, respectively. On day 3, B220+CD11c+NK1.1+ cell frequencies among bulk HMNCs were determined by flow cytometry. Error bars represent mean ± SEM values. Filled histograms in (A) and (B) represent staining with appropriate isotype controls. **p < 0.01, unpaired Student t test.

Close modal

Wright–Giemsa staining of sorted B220+CD11c+NK1.1+ cells revealed mononuclear cells with a basophilic cytoplasm (Supplemental Fig. 1). Upon further cytofluorimetric analyses, these were T/B lineage-negative (CD3/CD19) cells that also stained positively with mAbs against CD49b, CD122 (IL-2Rβ), and MHC-II molecules (Fig. 1B), which established their identity as pre-mNK cells at least by immunophenotypic standards (36).

We confirmed that the observed pre-mNK cell accumulation was secondary to iNKT cell activation because it could be significantly reversed by a CD1d-blocking mAb (clone 20H2) (Fig. 1C). This mAb was efficient in preventing iNKT cell responses because IL-4 and IFN-γ were virtually undetectable in the serum 2 and 6 h after α-GalCer injection, respectively (Fig. 1C).

Our kinetic experiments demonstrated gradual accumulation of pre-mNK cells after glycolipid priming, which became evident after 24 h and reached a plateau between 3 and 5 d (Fig. 2A). It was most pronounced in the liver but could also be observed within the spleen and the lungs (Fig. 2B).

FIGURE 2.

α-GalCer–induced pre-mNK cell accumulation is rapid and relatively widespread. (A) The kinetics of hepatic pre-mNK cell accumulation was studied in α-GalCer–primed mice (n = 3). (B) Five days after vehicle or α-GalCer administration (n = 4/group), the frequencies of B220+CD11c+NK1.1+ pre-mNK cells were determined among bulk splenocytes and nonparenchymal hepatic or pulmonary mononuclear cells. Representative plots from two independent experiments are shown.

FIGURE 2.

α-GalCer–induced pre-mNK cell accumulation is rapid and relatively widespread. (A) The kinetics of hepatic pre-mNK cell accumulation was studied in α-GalCer–primed mice (n = 3). (B) Five days after vehicle or α-GalCer administration (n = 4/group), the frequencies of B220+CD11c+NK1.1+ pre-mNK cells were determined among bulk splenocytes and nonparenchymal hepatic or pulmonary mononuclear cells. Representative plots from two independent experiments are shown.

Close modal

The functional maturity of cells belonging to the NK lineage has been linked to their differential expression of CD27 and CD11b (37). We found CD27highCD11blow cells, which are among the most immature subsets of NK lineage cells (36, 37), to comprise a sizeable proportion of our population of interest (Supplemental Fig. 2). It was therefore pertinent to determine whether and how quickly α-GalCer–transactivated hepatic pre-mNK cells may mature. This is experimentally achieved via adoptive cell transfer into another animal and monitoring for a loss of B220 (28). To this end, α-GalCer was administered to GFP-transgenic (GFP+) B6 mice followed 5 d later by cytofluorimetric isolation of hepatic B220+CD11c+NK1.1+ (pre-mNK) and B220NK1.1+ (control conventional NK) cells, which were then transferred into separate cohorts of WT B6 mice and tracked. As hypothesized, pre-mNK cells lost their B220 dramatically on day 5 and almost completely on day 15 posttransfer (Fig. 3). We also noticed a gradual disappearance of CD11c+ cells, albeit to a much lesser extent. The loss of cell surface receptor expression was not a global effect because NK1.1 and CD122 levels were maintained in both pre-mNK and conventional NK cells (Fig. 3).

FIGURE 3.

α-GalCer–transactivated pre-mNK cells lose B220 expression upon adoptive transfer. Five days after α-GalCer administration to GFP+ mice, hepatic B220+CD11c+NK1.1+ and B220NK1.1+ cell populations were FACS-sorted and transferred into separate cohorts of WT B6 recipients (n = 2 per group). Mice were sacrificed on day 5 or day 15 postadoptive transfer for their liver in which GFP+ cells were traced and examined for their expression levels of B220, CD11c, NK1.1, and CD122. Representative FACS plots are illustrated.

FIGURE 3.

α-GalCer–transactivated pre-mNK cells lose B220 expression upon adoptive transfer. Five days after α-GalCer administration to GFP+ mice, hepatic B220+CD11c+NK1.1+ and B220NK1.1+ cell populations were FACS-sorted and transferred into separate cohorts of WT B6 recipients (n = 2 per group). Mice were sacrificed on day 5 or day 15 postadoptive transfer for their liver in which GFP+ cells were traced and examined for their expression levels of B220, CD11c, NK1.1, and CD122. Representative FACS plots are illustrated.

Close modal

Next, we set out to determine whether intrahepatic pre-mNK cells had expanded locally or emigrated from other sites. We performed a comprehensive gene expression analysis of pre-mNK cells sorted from α-GalCer– (n = 3) and vehicle-treated mice (n = 12). The proliferation marker Ki67 was the most upregulated gene, with a 16-fold increase in mRNA transcript levels in pre-mNK cells from α-GalCer–treated mice relative to control (Fig. 4A). Proliferating cell nuclear Ag (Pcna) was also increased 4-fold (Fig. 4A). In contrast, there was a 4-fold drop in the expression of S1pr1 (Fig. 4A), which encodes a G protein–coupled receptor that mediates lymphocyte egress from lymphoid organs and their trafficking into the periphery (38). These results suggested that pre-mNK cells had undergone rigorous proliferation following α-GalCer administration as opposed to infiltrating the liver.

FIGURE 4.

Tissue accumulation of pre-mNK cells following α-GalCer treatment is due to their in situ expansion. (A) B6 mice were injected i.p. with vehicle (n = 12) or α-GalCer (n = 3). Two days later, pre-mNK cells were FACS-sorted and pooled before RNA was extracted for quantitative RT-PCR. A heat map was generated to demonstrate fold changes in transcript levels of indicated genes in pre-mNK cells isolated from α-GalCer–primed mice relative to those sorted from vehicle-treated controls. (B) Two days after treatment with vehicle or α-GalCer, HMNCs were stained for surface B220, CD11c, and NK1.1 as well as intracellular Ki67, and the percentages of Ki67+ cells among pre-mNK cells were determined by flow cytometry. Each symbol represents an individual mouse. (C) Surface expression of CD69 and S1PR1 by hepatic pre-mNK cells was assessed at indicated time points after α-GalCer administration (n = 4/time point). Error bars represent SEM. (D) Mice were injected with indicated doses of FTY720 or with PBS 2 h before and 24 h after they received α-GalCer. Five days later, HMNCs and splenocytes were harvested and subjected to staining with anti-CD3, -B220, -CD11c, and -NK1.1 mAbs. B220+CD11c+NK1.1+ pre-mNK cells and CD3+NK1.1 T cells were then enumerated by flow cytometry. Error bars represent SEM for six mice per group from two independent experiments. Statistical comparisons were made using unpaired Student t test (B) or a two-way ANOVA with Bonferroni post hoc test (D). *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 4.

Tissue accumulation of pre-mNK cells following α-GalCer treatment is due to their in situ expansion. (A) B6 mice were injected i.p. with vehicle (n = 12) or α-GalCer (n = 3). Two days later, pre-mNK cells were FACS-sorted and pooled before RNA was extracted for quantitative RT-PCR. A heat map was generated to demonstrate fold changes in transcript levels of indicated genes in pre-mNK cells isolated from α-GalCer–primed mice relative to those sorted from vehicle-treated controls. (B) Two days after treatment with vehicle or α-GalCer, HMNCs were stained for surface B220, CD11c, and NK1.1 as well as intracellular Ki67, and the percentages of Ki67+ cells among pre-mNK cells were determined by flow cytometry. Each symbol represents an individual mouse. (C) Surface expression of CD69 and S1PR1 by hepatic pre-mNK cells was assessed at indicated time points after α-GalCer administration (n = 4/time point). Error bars represent SEM. (D) Mice were injected with indicated doses of FTY720 or with PBS 2 h before and 24 h after they received α-GalCer. Five days later, HMNCs and splenocytes were harvested and subjected to staining with anti-CD3, -B220, -CD11c, and -NK1.1 mAbs. B220+CD11c+NK1.1+ pre-mNK cells and CD3+NK1.1 T cells were then enumerated by flow cytometry. Error bars represent SEM for six mice per group from two independent experiments. Statistical comparisons were made using unpaired Student t test (B) or a two-way ANOVA with Bonferroni post hoc test (D). *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal

We validated the observed changes in Ki67 and S1pr1 expression at the protein level (Fig. 4B, 4C). In addition, we examined the surface expression of CD69 on hepatic pre-mNK cells because CD69 can directly suppress S1PR1 and, in doing so, serves as a tissue retention molecule (39, 40). We found an inverse correlation between S1PR1+ and CD69+ pre-mNK cell frequencies in α-GalCer–treated animals. Accordingly, a relatively sharp decline in the former population was reciprocally accompanied by an equally sharp rise in the latter on day 1 postglycolipid injection (Fig. 4C).

To more definitively rule out the possibility that α-GalCer mobilizes pre-mNK cells to amass in select tissues, we employed two separate doses of FTY720 (also known as fingolimod), an immunomodulatory compound that causes S1PR1 downregulation and inhibits lymphocytes’ emigration from lymphoid organs (38). FTY720 administration 2 h before and 24 h after α-GalCer injection significantly lowered T cell numbers, but not pre-mNK cell numbers in the liver and in the spleen (Fig. 4D).

Collectively, the above findings indicate that tissue-resident pre-mNK cells downregulate their expression of S1PR1, upregulate CD69, and proliferate locally in response to systemic α-GalCer treatment.

We sought to uncover the mechanism underlying pre-mNK cell expansion in our system. α-GalCer stimulation of iNKT cells induces the production of a myriad of cytokines that modify the biological behavior of multiple downstream effector cell types. Our transcript analyses pointed to altered expression of several cytokine receptors in transactivated pre-mNK cells (Fig. 4A), suggesting that inflammatory cytokines may be involved. IL-15 is known to promote pre-mNK cell expansion (41). However, α-GalCer treatment did not change the expression levels of two IL-15R subunits, namely Cd132c, which is also shared by receptors for IL-2, -4, -7, -9, and -21) and Cd122 (which constitutes a component of both IL-2R and IL-15R) (Fig. 4A). We did not include Il-15rα in our array simply because pre-mNK cells are reportedly devoid of this molecule (41). Il-15rα is found instead on the surface of APCs that trans-present IL-15 to other immunocytes (42).

We then shifted our focus onto other cytokines that activate NK cells and other innate-like lymphocytes. Of note, Il-21r and Il-12r2 transcripts were each elevated over 4-fold (Fig. 4A). In vivo blockade of CD132 before and after α-GalCer administration failed to prevent pre-mNK cell expansion, thus ruling out a role for γc chain cytokines, including IL-15 and IL-21 (data not shown). We then zeroed in on IL-12 and cytokines with which it synergizes. We injected mice with rIL-12, IL-18, and IFN-γ in several combinations in lieu of α-GalCer. The only combination that induced hepatic pre-mNK cell accumulation was IL-12 plus IL-18 (Fig. 5A). We also observed a moderate reduction in Il-18r mRNA levels 2 d after α-GalCer treatment (Fig. 4A), a finding we validated by flow cytometry at a slightly later time point. Three days after the injection of α-GalCer or vehicle, 36.5 ± 2.4% and 62.3 ± 1.4% of hepatic pre-mNK cells expressed CD218a (IL-18Ra), respectively (n = 3/group). In addition, we found a marginal decrease in the geometric mean fluorescence intensity (gMFI) of CD218a in α-GalCer-primed animals (1180 ± 4) compared with controls (1317 ± 19). These changes likely reflect activation-induced IL-18R downregulation at later time points following α-GalCer treatment.

FIGURE 5.

Hepatic pre-mNK cell expansion in response to α-GalCer is dependent on IL-12 and IL-18. (A) Naive B6 mice (n = 3–4 per group) were injected with indicated recombinant cytokines followed 3 d later by the determination of pre-mNK cell frequencies in the liver by flow cytometry. (B) B6 mice were primed with α-GalCer or injected with vehicle before they were bled at indicated time points. Sera were isolated, pooled as indicated, and adoptively transferred into B6 mice (n = 3/group) that had received a CD1d-blocking mAb 2 h earlier. Three days after serum transfer, hepatic pre-mNK cell frequencies were determined by flow cytometry. (C) Sera collected at 6 and 12 h post–α-GalCer injection were pooled. An aliquot was depleted of IL-12 and IL-18. Cytokine-replete and -depleted serum samples were then transferred into separate cohorts of anti-CD1d–pretreated mice (n = 3), and pre-mNK cell percentages were determined. Error bars represent SEM. Statistical analyses were performed using a one-way ANOVA with Tukey post hoc analysis (A) or unpaired Student t test (C). *p < 0.05, ***p < 0.001.

FIGURE 5.

Hepatic pre-mNK cell expansion in response to α-GalCer is dependent on IL-12 and IL-18. (A) Naive B6 mice (n = 3–4 per group) were injected with indicated recombinant cytokines followed 3 d later by the determination of pre-mNK cell frequencies in the liver by flow cytometry. (B) B6 mice were primed with α-GalCer or injected with vehicle before they were bled at indicated time points. Sera were isolated, pooled as indicated, and adoptively transferred into B6 mice (n = 3/group) that had received a CD1d-blocking mAb 2 h earlier. Three days after serum transfer, hepatic pre-mNK cell frequencies were determined by flow cytometry. (C) Sera collected at 6 and 12 h post–α-GalCer injection were pooled. An aliquot was depleted of IL-12 and IL-18. Cytokine-replete and -depleted serum samples were then transferred into separate cohorts of anti-CD1d–pretreated mice (n = 3), and pre-mNK cell percentages were determined. Error bars represent SEM. Statistical analyses were performed using a one-way ANOVA with Tukey post hoc analysis (A) or unpaired Student t test (C). *p < 0.05, ***p < 0.001.

Close modal

Previous studies have shown that serum IL-12 reaches its peak level at 6 h post–α-GalCer injection (43, 44). In addition, IL-18 typically follows IL-12 and plateaus between 12 and 24 h after in vivo priming with select inflammatory stimuli (45). To first demonstrate that circulating cytokines mediate the accumulation of pre-mNK cells, we obtained serum samples at several time points after α-GalCer treatment, which were then pooled and transferred into naive B6 mice (Fig. 5B). This was followed by pre-mNK cell enumeration in the liver. Transferring pooled sera prepared at 2-, 6-, or 12-h time points did not result in pre-mNK cell expansion and neither did combined sera collected at 6- and 24-h time points (Fig. 5B). Strikingly, however, a combination of samples from 6- and 12-h time points worked synergistically to recapitulate the effect of α-GalCer (Fig. 5B). Of note, in these experiments, we injected a CD1d-blocking mAb into the recipients prior to serum transfer. This was to avoid false positive results due to the presence of free-floating α-GalCer in transferred sera.

The above time points are consistent with the peaks of serum IL-12 and IL-18 among other soluble mediators. To confirm the roles of IL-12 and IL-18, we used a previously established protocol to remove cytokines from serum samples. Unlike cytokine-replete sera, IL-12/IL-18–depleted samples failed to expand hepatic pre-mNK cells (Fig. 5C). Therefore, α-GalCer–triggered pre-mNK cell expansion is dependent on the synergistic functions of IL-12 and IL-18.

Splenic and bone marrow–derived pre-mNK cells are capable of cross-presenting peptide Ags to T cells, thus engaging the adaptive arm of anti-tumor immunity (35, 46, 47). We explored whether α-GalCer–transactivated hepatic pre-mNK cells can similarly serve as APCs. When pulsed with SIINFEKL, a synthetic peptide corresponding to the immunodominant epitope of chicken OVA (OVA257–264), purified hepatic pre-mNK cells were able to activate B3Z cells (Fig. 6A), a CD8+ hybridoma line that recognizes SIINFEKL in the context of H-2Kb. Therefore, hepatic pre-mNK cells had retained their MHC class I (MHC-I) expression and peptide presentation capacity. It was, however, more important to determine whether pre-mNK cells’ in vivo interactions with malignant cells could result in tumor Ag-derived peptide cross-presentation. To this end, we inoculated vehicle- and α-GalCer–pretreated B6 mice with EG7-OVA lymphoma cells via tail vein. Four days later, nonparenchymal HMNCs were isolated, and the percentage of SIINFEKL-presenting pre-mNK cells was determined using an mAb against SIINFEKL:H-2Kb complexes (clone 25-D1.16) (30). Interestingly, a “25-D1.16+” fraction was clearly detectable among pre-mNK cells sorted from vehicle-treated mice and at a lowered frequency among those isolated from glycolipid-treated animals (Fig. 6B). Therefore, α-GalCer administration impairs Ag cross-presentation by pre-mNK cells, a finding that is also consistent with decreased I-a/e, Cd40, and Cd80 transcript levels (Fig. 4A).

FIGURE 6.

α-GalCer–transactivated hepatic pre-mNK cells are unable to cross-present tumor Ags. (A) B6 mice (n = 4 from two independent experiments) were euthanized 5 d after they were injected with α-GalCer. Hepatic pre-mNK cells were FACS-sorted and pulsed with SIINFEKL before they were cocultured with B3Z cells. After 24 h, cells were lysed in the presence of ortho-nitrophenyl-β-d-galactopyranoside, and the OD415 nm of supernatants was determined 4 h later. (B) EG7-OVA tumor cells were injected i.v. into B6 mice that had received α-GalCer or vehicle 24 h earlier (n = 3 per group). Four days after the metastatic tumor challenge, SIINFEKL:H-2Kb complexes were detected via staining with 25-D1.16. Histograms representing two independent experiments yielding similar results are depicted. Error bars represent SEM. An unpaired Student t test (A) and one-way ANOVA with Tukey post hoc analysis (B) were used for statistical comparisons. *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 6.

α-GalCer–transactivated hepatic pre-mNK cells are unable to cross-present tumor Ags. (A) B6 mice (n = 4 from two independent experiments) were euthanized 5 d after they were injected with α-GalCer. Hepatic pre-mNK cells were FACS-sorted and pulsed with SIINFEKL before they were cocultured with B3Z cells. After 24 h, cells were lysed in the presence of ortho-nitrophenyl-β-d-galactopyranoside, and the OD415 nm of supernatants was determined 4 h later. (B) EG7-OVA tumor cells were injected i.v. into B6 mice that had received α-GalCer or vehicle 24 h earlier (n = 3 per group). Four days after the metastatic tumor challenge, SIINFEKL:H-2Kb complexes were detected via staining with 25-D1.16. Histograms representing two independent experiments yielding similar results are depicted. Error bars represent SEM. An unpaired Student t test (A) and one-way ANOVA with Tukey post hoc analysis (B) were used for statistical comparisons. *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal

Pre-mNK cells reportedly express TRAIL and kill tumor targets (23). It was therefore critical to test whether the expanded hepatic pre-mNK cells retained their oncolytic arsenal and functions. We found unfractionated HMNCs from α-GalCer–treated mice to readily and dose-dependently destroy YAC-1 lymphoma cells (Supplemental Fig. 3), a classic mouse NK cell target. To directly assess pre-mNK cell–mediated cytotoxicity, we also employed purified hepatic pre-mNK cells from glycolipid-primed animals against YAC-1 cells. In certain experiments, CMA or a combination of EGTA and MgCl2 was added to cocultures. CMA increases the pH of lytic granules to accelerate the degradation of perforin (48). EGTA/MgCl2 chelates extracellular Ca++, which is required at several steps during the perforin/GZM pathway (49, 50). Purified pre-mNK cells from α-GalCer–treated mice could efficiently lyse YAC-1 cells (Fig. 7A). However, this response was completely abolished in the presence of either EGTA/MgCl2 or CMA in cocultures (Fig. 7A), clearly implicating the granule exocytosis pathway in our system. This notion was supported by the increased expression of CD107a, a degranulation marker, among hepatic pre-mNK cells after they had established an immunological synapse with YAC-1 cells (Fig. 7B). To identify the cytotoxic effector molecules pre-mNK cells used to kill tumor cells, we determined the frequencies of FasL+, TRAIL+, TNF-α+, GZM A+, and GZM B+ pre-mNK cells from α-GalCer–treated mice after they had engaged YAC-1 cells. Surprisingly, GZM A, but not TRAIL (or any other mediators for that matter), was upregulated (Fig. 7C). Also interestingly, α-GalCer treatment alone had resulted in elevated levels of GZM A on a per cell basis as judged by the gMFI of its staining (Fig. 7C).

FIGURE 7.

Transactivated pre-mNK cells are oncolytic and partially protective against metastatic B16 melanoma. (A) Hepatic pre-mNK cells were FACS-sorted from α-GalCer–primed mice (n = 4) and used as effector cells against 51Cr-labeled YAC-1 target cells in 4 h cocultures containing or lacking a combination of EGTA and MgCl2 or CMA. The 51Cr activity of culture supernatants was quantitated using a gamma counter. (B) Bulk HMNCs were isolated from α-GalCer–treated mice and coincubated with YAC-1 cells in the presence of BFA and monensin. Four hours later, the surface expression of CD107a on pre-mNK cells was determined by flow cytometry. Representative plots and cumulative data from two independent experiments (n = 5 per group) are shown. (C) HMNCs from α-GalCer–primed mice (n = 3) were coincubated with YAC-1 cells in the presence of BFA before surface or intracellular staining for indicated effector molecules. The expression of each molecule by pre-mNK cells was analyzed by flow cytometry. In addition, the gMFI of GZM A staining in pre-mNK cells is shown shortly after HMNC isolation (n = 3/group). (D) CFSElow WT splenocytes (control target cells) and CFSEhigh β2M−/− splenocytes (MHC-I–deficient target cells) were mixed in equal numbers and injected i.v. into naive mice or mice that had been primed with α-GalCer and also injected with an anti-B220 mAb, an anti-CD19 mAb, or PBS as described in 2Materials and Methods. Three hours later, target cells were identified in the liver using their differential CFSE fluorescence, and the percentage of in vivo cytotoxicity against each target population was calculated using a formula that is also provided in the 2Materials and Methods. Representative contour plots demonstrate the efficacy of anti-B220 in depleting pre-mNK cells. For in vivo killing assays, representative histograms and cumulative data from two independent experiments are shown. (E) Five hundred thousand B16-F10 melanoma cells were injected i.v. into vehicle-treated mice or α-GalCer–primed animals that had previously received anti-B220, anti-CD19, or PBS. Two weeks later, metastatic nodules in the lungs were enumerated. Results are depicted as fold change in nodule numbers relative to vehicle-treated mice in three independent experiments. All error bars represent SEM. Statistical analyses were performed using Student t tests (B and C) or one-way ANOVA with Newman–Keuls post hoc test (D and E). *p < 0.05.

FIGURE 7.

Transactivated pre-mNK cells are oncolytic and partially protective against metastatic B16 melanoma. (A) Hepatic pre-mNK cells were FACS-sorted from α-GalCer–primed mice (n = 4) and used as effector cells against 51Cr-labeled YAC-1 target cells in 4 h cocultures containing or lacking a combination of EGTA and MgCl2 or CMA. The 51Cr activity of culture supernatants was quantitated using a gamma counter. (B) Bulk HMNCs were isolated from α-GalCer–treated mice and coincubated with YAC-1 cells in the presence of BFA and monensin. Four hours later, the surface expression of CD107a on pre-mNK cells was determined by flow cytometry. Representative plots and cumulative data from two independent experiments (n = 5 per group) are shown. (C) HMNCs from α-GalCer–primed mice (n = 3) were coincubated with YAC-1 cells in the presence of BFA before surface or intracellular staining for indicated effector molecules. The expression of each molecule by pre-mNK cells was analyzed by flow cytometry. In addition, the gMFI of GZM A staining in pre-mNK cells is shown shortly after HMNC isolation (n = 3/group). (D) CFSElow WT splenocytes (control target cells) and CFSEhigh β2M−/− splenocytes (MHC-I–deficient target cells) were mixed in equal numbers and injected i.v. into naive mice or mice that had been primed with α-GalCer and also injected with an anti-B220 mAb, an anti-CD19 mAb, or PBS as described in 2Materials and Methods. Three hours later, target cells were identified in the liver using their differential CFSE fluorescence, and the percentage of in vivo cytotoxicity against each target population was calculated using a formula that is also provided in the 2Materials and Methods. Representative contour plots demonstrate the efficacy of anti-B220 in depleting pre-mNK cells. For in vivo killing assays, representative histograms and cumulative data from two independent experiments are shown. (E) Five hundred thousand B16-F10 melanoma cells were injected i.v. into vehicle-treated mice or α-GalCer–primed animals that had previously received anti-B220, anti-CD19, or PBS. Two weeks later, metastatic nodules in the lungs were enumerated. Results are depicted as fold change in nodule numbers relative to vehicle-treated mice in three independent experiments. All error bars represent SEM. Statistical analyses were performed using Student t tests (B and C) or one-way ANOVA with Newman–Keuls post hoc test (D and E). *p < 0.05.

Close modal

To measure pre-mNK cell–mediated cytotoxicity in vivo, we modified and used a previously described protocol (51). We coinjected CFSElow WT B6 splenocytes (control target cells) and CFSEhigh β2M−/− splenocytes (NK-sensitive target cells) into α-GalCer–treated mice. Labeled targets were tracked using their differential CFSE fluorescence. The peak corresponding to β2M−/− target cells was always smaller, indicating that they had been eliminated (Fig. 7D). To determine the partial contribution of pre-mNK and conventional NK cells, we used an anti-B220 mAb that depletes the former but not the latter. This approach resulted in significantly reduced cytotoxicity (Fig. 7D), indicating that pre-mNK cells were partially responsible for the elimination of β2M−/− target cells. To rule out a role for B220+ B cells in this model, we used an anti-CD19 mAb in parallel. This mAb should remove B cells but not pre-mNK cells. As anticipated, anti-CD19 administration failed to diminish the lysis of β2M−/− cells (Fig. 7D).

Finally, we tested the antimetastatic capacity of transactivated pre-mNK cells using the B16-F10 melanoma model. This model was chosen because it is responsive to α-GalCer treatment (52), which we first verified (Fig. 7E, Supplemental Fig. 4). In addition, pre-mNK cells reportedly fulfill a protective role against metastatic B16-F10 melanoma (23). We found the depletion of B220+ cells, but not CD19+ cells, to partially repress α-GalCer’s antimetastatic activity and almost double the number of pulmonary metastatic nodules (Fig. 7E). Taken together, the above results demonstrate that α-GalCer selectively boosts the expression of GZM A by hepatic pre-mNK cells and imparts a primarily oncolytic and antimetastatic phenotype to these unique cytotoxic lymphocytes.

Although pre-mNK cells were discovered more than a decade ago, only a precious little is known about immunoregulatory mechanisms that control or modify their abundance and activities. In this work, we have identified and characterized a novel population of pre-mNK cells that amass in select tissues, especially in the liver, following α-GalCer administration and that are endowed with potent cytolytic and antimetastatic properties.

Previous investigations have found pre-mNK cells in the spleen, lymph nodes, and bone marrow at low frequencies (23, 24, 26, 34, 35). We now describe a liver- and lung-resident population that can be dramatically enlarged through α-GalCer priming, likely because of their local expansion as opposed to their recruitment from other sites. This is in agreement with the notion that pre-mNK cells are rapidly recycling cells by nature (28). We found increased levels of Ki67 and CD69, decreased S1PR1 expression, and the failure of FTY720 to prevent the observed accumulation. We are cognizant of previous reports that S1PR1 and S1PR5 can both mediate NK cell trafficking (53, 54). However, we focused our efforts on FTY720-sensitive S1PR1 because our gene array analyses showed only a modest change in S1pr5 transcript levels in pre-mNK cells (Fig. 4A). Interestingly, Walzer et al. (53) found that NK cells from S1PR5−/− mice were unable to home to several tissue compartments with the notable exception of the liver, the prominent site of pre-mNK cell accumulation following α-GalCer treatment in our model. Although tissue recruitment was not responsible for the observed numerical rise in hepatic pre-mNK cells, it will be important to explore the migratory properties of these cells in the future.

Pre-mNK cell differentiation is typically dependent on IL-15R signaling (41), and their proliferation can be triggered by a combination of IL-15 and IL-18 (27). However, α-GalCer treatment did not alter IL-15R levels in hepatic pre-mNK cells, and our mechanistic experiments revealed a previously unappreciated role for IL-12 and IL-18 in pre-mNK cell expansion. We also found a higher percentage of IL-18R+ pre-mNK cells in vehicle-treated mice when compared with their α-GalCer–transactivated counterparts. IL-18R downregulation in α-GalCer–treated mice may have followed an early burst of IL-12, which may have initially elevated the expression level of IL-18R. IL-12–induced IL-18R upregulation has been previously reported in other cell types (55). IL-12 and IL-18 are known for their ability to activate innate and innate-like lymphocytes, and our findings widen their range of functions by introducing pre-mNK cells as one of their cellular targets.

The cross-talk between iNKT and pre-mNK cells represents a new intercellular communication cascade following treatment with α-GalCer. The importance of this cascade is severalfold. First, pre-mNK cells can now be viewed as a downstream effector cell type mediating some of the beneficial effects of α-GalCer. Given pre-mNK cells’ phenotypic and functional similarities to several other cell types (e.g., DCs and mNK cells), they may have been given a mistaken identity in the past. Many investigations have taken advantage of an anti-NK1.1 mAb to delineate the roles of α-GalCer–transactivated NK cells in destroying tumor cells in vitro (56, 57) and in lowering metastatic tumor burden in vivo (58). Although this approach is widely accepted and still valid, it may ignore pre-mNK cells’ potential participation in such responses. In this study, pre-mNK cells were indeed partially responsible for oncolytic and antimetastatic activities of α-GalCer. By the same token, Fujii et al. (12) described a splenic CD3CD11c+ population capable of producing IFN-γ 2 h after an i.v. injection of α-GalCer. This was before IKDCs/pre-mNK cells found their way into the literature. Therefore, the possibility that such IFN-γ–producing DCs may have contained a pre-mNK cell component is not far-fetched.

Pre-mNK cells can serve as APCs in adaptive immunity (24, 35, 46, 47). They typically express MHC-I and MHC-II, which should enable them to activate CD8+ and CD4+ T cells. We were able to detect SIINFEKL:H-2Kb complexes on the surface of hepatic pre-mNK cells after i.v. inoculation of an OVA-expressing lymphoma cell line. This provides direct evidence in support of a role for pre-mNK cells in tumor Ag cross-presentation in vivo. Interestingly, however, whereas α-GalCer–transactivated pre-mNK cells maintained their expression of MHC-I, they lost their ability to cross-present SIINFEKL. This may be due, at least partially, to the expansion of the CD27+CD11b fraction of hepatic pre-mNK cells by α-GalCer (Supplemental Fig. 2) because Terme et al. (46) reported that pre-mNK cells’ cross-presentation capacity resides within their CD11b+ subset.

α-GalCer–transactivated hepatic pre-mNK cells exhibited cytotoxicity against NK targets, namely YAC-1 thymoma and β2M−/− splenocytes. They relied on granule exocytosis for their oncolytic function and expressed high levels of GZM A, which was evident even before conjugate formation with YAC-1 cells. The tumoricidal activity of pre-mNK cells was initially reported to be TRAIL-dependent (23). However, α-GalCer–transactivated pre-mNK cells lack TRAIL and likely employ GZM A to kill tumor targets. Of note, we also noticed a ∼16-fold increase in GZM K expression by these cells at the mRNA level (Fig. 4A). However, a reliable anti-mouse GZM K mAb is not available at this point to enable validation of this finding at the protein level.

The above results, together with the intact antimetastatic activity of α-GalCer–transactivated pre-mNK cells, indicate that exposure to this clinically relevant glycolipid skews pre-mNK cell responses away from an Ag-presenting phenotype in favor of NK-like anti-tumor behaviors. This may be a unique property of α-GalCer and potentially other glycolipid agonists of iNKT cells. In fact, stimulation with CpG oligodeoxynucleotides skews pre-mNK cells in the opposite direction and results in their loss of NK-like functions, upregulation of MHC-II and costimulatory molecules, and acquisition of DC-like Ag-presenting activity (24).

Taieb et al. (23) found IKDCs/pre-mNK cells to produce more IFN-γ in comparison with mNK cells and to exert prolific cytotoxicity against B16-F10 melanoma cells. These cells proliferated in response to a combination of imatinib mesylate and IL-2 and infiltrated pulmonary metastases of B16-F10. Taieb et al. also isolated B220+NK1.1+ pre-mNK cells from mice that had been treated with imatinib plus IL-2 and adoptively transferred them into melanoma-bearing Rag2−/−Il2rg−/− mice. They found a lower tumor burden in animals that had received pre-mNK cells but not in recipients of mNK cells. In our work, we used a WT nontransgenic mouse model and cell-depleting mAbs to establish a similarly protective role for α-GalCer–transactivated pre-mNK cells. In contrast, Wilson et al. (59) suggested that pre-mNK cell depletion augments the protective effects of adoptively transferred melanoma-specific CD4+ T cells in melanoma-bearing RAG−/− mice. It is noteworthy that to deplete pre-mNK cells, the authors used an mAb (clone HB220) that targets the CD45RB isoform of B220. In our hands, this mAb was inefficient in appreciably removing pre-mNK cells (data not shown). In addition, the experimental systems used in the two studies are fundamentally different. We examined the efficacy of α-GalCer and α-GalCer–transactivated pre-mNK cells in a pulmonary metastasis model in WT mice. By comparison, Wilson et al. addressed the role of pre-mNK cells in the regulation of CD4+ T cell–mediated immunity against primary melanoma established in immunodeficient mice. Such stark differences likely account for the different conclusions reached.

Although pre-mNK cells may express higher levels of certain inflammatory and effector molecules in comparison with mNK cells on a per cell basis (23, 25), their therapeutic value is shadowed by their relative paucity. However, this limitation may be remedied by treatments that expand pre-mNK cells without hampering their anticancer potentials. This may be particularly important for tumors that arise from the liver or metastasize to this organ (for instance, through the portal vein circulation) because hepatic pre-mNK cell expansion can be profound. We posit that iNKT cell–based glycolipid immunotherapies should be regarded as examples of such treatments.

CD56bright and HLA-DR+ subsets of human NK cells have been proposed to be the mouse pre-mNK cell counterparts (36). In contrast to “classical” CD56dim NK cells, which are often considered the main executors of NK cell–mediated anti-tumor responses, the CD56bright population has been viewed as an immunomodulatory subset. However, this paradigm was shifted by a recent report that upon priming with IL-15, CD56bright NK cells exhibit markedly enhanced degranulation, cytotoxicity, and cytokine production (60). In a phase I clinical trial, imatinib mesylate plus IL-2 expanded the HLA-DR+ NK cell population and improved the progression-free survival and overall survival of patients with refractory solid tumors (61). Furthermore, α-GalCer–based therapies have shown promise in several clinical trials for cancer (4). It is tempting to speculate an additive or synergistic effect through combining imatinib, IL-2, and α-GalCer.

Finally, it will be interesting to determine whether the anti-tumor activities of pre-mNK cells can be further amplified through treatment with TH1-polarizing α-GalCer analogues such as α–C-GalCer (62) or α-GalCer/α–C-GalCer–pulsed DCs that may be superior to free-floating glycolipids (63). Much work lies ahead in characterizing circulating and tissue-resident pre-mNK cell subsets in mice and humans and in deciphering the regulatory mechanisms and therapeutic modalities that dictate or alter their functional attributes.

We thank Dr. Howard Young (National Cancer Institute, Frederick, MD) for valuable intellectual input into this work. We also thank Tunyalux Langsub for expert assistance with figures and other members of the Haeryfar Laboratory for helpful discussions.

This work was supported by Canadian Institutes of Health Research Grants PJT-156295 and PJT-376303 to S.M.M.H. and S.L., respectively. J.C. is a recipient of a Queen Elizabeth II Graduate Scholarship in Science and Technology from the Ontario Ministry of Training, Colleges and Universities. P.T.R. is a recipient of an Alexander Graham Bell Canada Graduate Scholarship from the Natural Sciences and Engineering Research Council of Canada.

The online version of this article contains supplemental material.

Abbreviations used in this article:

B6

C57BL/6

BFA

brefeldin A

γc

common γ chain

CMA

concanamycin A

DC

dendritic cell

α-GalCer

α-galactosylceramide

gMFI

geometric mean fluorescence intensity

GZM

granzyme

HMNC

hepatic mononuclear cell

IKDC

IFN-producing killer DC

iNKT

invariant NKT

β2M

β2 microglobulin

MHC-I

MHC class I

MHC-II

MHC class II

mNK

mature NK

pre-mNK

precursor to mNK

S1PR1

sphingosine-1-phosphate receptor 1

WT

wild-type.

1
Lantz
,
O.
,
A.
Bendelac
.
1994
.
An invariant T cell receptor alpha chain is used by a unique subset of major histocompatibility complex class I-specific CD4+ and CD4-8- T cells in mice and humans.
J. Exp. Med.
180
:
1097
1106
.
2
Kawano
,
T.
,
J.
Cui
,
Y.
Koezuka
,
I.
Toura
,
Y.
Kaneko
,
K.
Motoki
,
H.
Ueno
,
R.
Nakagawa
,
H.
Sato
,
E.
Kondo
, et al
.
1997
.
CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides.
Science
278
:
1626
1629
.
3
Stetson
,
D. B.
,
M.
Mohrs
,
R. L.
Reinhardt
,
J. L.
Baron
,
Z. E.
Wang
,
L.
Gapin
,
M.
Kronenberg
,
R. M.
Locksley
.
2003
.
Constitutive cytokine mRNAs mark natural killer (NK) and NK T cells poised for rapid effector function.
J. Exp. Med.
198
:
1069
1076
.
4
van den Heuvel
,
M. J.
,
N.
Garg
,
L.
Van Kaer
,
S. M.
Haeryfar
.
2011
.
NKT cell costimulation: experimental progress and therapeutic promise.
Trends Mol. Med.
17
:
65
77
.
5
King
,
L. A.
,
R.
Lameris
,
T. D.
de Gruijl
,
H. J.
van der Vliet
.
2018
.
CD1d-invariant natural killer T cell-based cancer immunotherapy: α-galactosylceramide and beyond. [Published erratum appears in 2018 Front. Immunol. 9: 2606.]
Front. Immunol.
9
:
1519
.
6
Kawano
,
T.
,
J.
Cui
,
Y.
Koezuka
,
I.
Toura
,
Y.
Kaneko
,
H.
Sato
,
E.
Kondo
,
M.
Harada
,
H.
Koseki
,
T.
Nakayama
, et al
.
1998
.
Natural killer-like nonspecific tumor cell lysis mediated by specific ligand-activated Valpha14 NKT cells.
Proc. Natl. Acad. Sci. USA
95
:
5690
5693
.
7
Takahashi
,
T.
,
K.
Haraguchi
,
S.
Chiba
,
M.
Yasukawa
,
Y.
Shibata
,
H.
Hirai
.
2003
.
Valpha24+ natural killer T-cell responses against T-acute lymphoblastic leukaemia cells: implications for immunotherapy.
Br. J. Haematol.
122
:
231
239
.
8
Metelitsa
,
L. S.
,
K. I.
Weinberg
,
P. D.
Emanuel
,
R. C.
Seeger
.
2003
.
Expression of CD1d by myelomonocytic leukemias provides a target for cytotoxic NKT cells.
Leukemia
17
:
1068
1077
.
9
Wingender
,
G.
,
P.
Krebs
,
B.
Beutler
,
M.
Kronenberg
.
2010
.
Antigen-specific cytotoxicity by invariant NKT cells in vivo is CD95/CD178-dependent and is correlated with antigenic potency.
J. Immunol.
185
:
2721
2729
.
10
Metelitsa
,
L. S.
,
O. V.
Naidenko
,
A.
Kant
,
H. W.
Wu
,
M. J.
Loza
,
B.
Perussia
,
M.
Kronenberg
,
R. C.
Seeger
.
2001
.
Human NKT cells mediate antitumor cytotoxicity directly by recognizing target cell CD1d with bound ligand or indirectly by producing IL-2 to activate NK cells.
J. Immunol.
167
:
3114
3122
.
11
Kitamura
,
H.
,
K.
Iwakabe
,
T.
Yahata
,
S.
Nishimura
,
A.
Ohta
,
Y.
Ohmi
,
M.
Sato
,
K.
Takeda
,
K.
Okumura
,
L.
Van Kaer
, et al
.
1999
.
The natural killer T (NKT) cell ligand alpha-galactosylceramide demonstrates its immunopotentiating effect by inducing interleukin (IL)-12 production by dendritic cells and IL-12 receptor expression on NKT cells.
J. Exp. Med.
189
:
1121
1128
.
12
Fujii
,
S.
,
K.
Liu
,
C.
Smith
,
A. J.
Bonito
,
R. M.
Steinman
.
2004
.
The linkage of innate to adaptive immunity via maturing dendritic cells in vivo requires CD40 ligation in addition to antigen presentation and CD80/86 costimulation.
J. Exp. Med.
199
:
1607
1618
.
13
Hayakawa
,
Y.
,
K.
Takeda
,
H.
Yagita
,
S.
Kakuta
,
Y.
Iwakura
,
L.
Van Kaer
,
I.
Saiki
,
K.
Okumura
.
2001
.
Critical contribution of IFN-gamma and NK cells, but not perforin-mediated cytotoxicity, to anti-metastatic effect of alpha-galactosylceramide.
Eur. J. Immunol.
31
:
1720
1727
.
14
Nishio
,
S.
,
N.
Yamada
,
H.
Ohyama
,
K.
Yamanegi
,
K.
Nakasho
,
M.
Hata
,
Y.
Nakamura
,
S.
Fukunaga
,
H.
Futani
,
S.
Yoshiya
, et al
.
2008
.
Enhanced suppression of pulmonary metastasis of malignant melanoma cells by combined administration of alpha-galactosylceramide and interleukin-18.
Cancer Sci.
99
:
113
120
.
15
Nakagawa
,
R.
,
T.
Inui
,
I.
Nagafune
,
Y.
Tazunoki
,
K.
Motoki
,
A.
Yamauchi
,
M.
Hirashima
,
Y.
Habu
,
H.
Nakashima
,
S.
Seki
.
2004
.
Essential role of bystander cytotoxic CD122+CD8+ T cells for the antitumor immunity induced in the liver of mice by alpha-galactosylceramide.
J. Immunol.
172
:
6550
6557
.
16
Paget
,
C.
,
M. T.
Chow
,
H.
Duret
,
S. R.
Mattarollo
,
M. J.
Smyth
.
2012
.
Role of γδ T cells in α-galactosylceramide-mediated immunity.
J. Immunol.
188
:
3928
3939
.
17
Morita
,
M.
,
K.
Motoki
,
K.
Akimoto
,
T.
Natori
,
T.
Sakai
,
E.
Sawa
,
K.
Yamaji
,
Y.
Koezuka
,
E.
Kobayashi
,
H.
Fukushima
.
1995
.
Structure-activity relationship of alpha-galactosylceramides against B16-bearing mice.
J. Med. Chem.
38
:
2176
2187
.
18
Kain
,
L.
,
B.
Webb
,
B. L.
Anderson
,
S.
Deng
,
M.
Holt
,
A.
Costanzo
,
M.
Zhao
,
K.
Self
,
A.
Teyton
,
C.
Everett
, et al
.
2014
.
The identification of the endogenous ligands of natural killer T cells reveals the presence of mammalian α-linked glycosylceramides. [Published erratum appears in 2014 Immunity 41: 867.]
Immunity
41
:
543
554
.
19
Brossay
,
L.
,
M.
Chioda
,
N.
Burdin
,
Y.
Koezuka
,
G.
Casorati
,
P.
Dellabona
,
M.
Kronenberg
.
1998
.
CD1d-mediated recognition of an alpha-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution.
J. Exp. Med.
188
:
1521
1528
.
20
Nakagawa
,
R.
,
K.
Motoki
,
H.
Nakamura
,
H.
Ueno
,
R.
Iijima
,
A.
Yamauchi
,
S.
Tsuyuki
,
T.
Inamoto
,
Y.
Koezuka
.
1998
.
Antitumor activity of alpha-galactosylceramide, KRN7000, in mice with EL-4 hepatic metastasis and its cytokine production.
Oncol. Res.
10
:
561
568
.
21
Nakagawa
,
R.
,
I.
Serizawa
,
K.
Motoki
,
M.
Sato
,
H.
Ueno
,
R.
Iijima
,
H.
Nakamura
,
A.
Shimosaka
,
Y.
Koezuka
.
2000
.
Antitumor activity of alpha-galactosylceramide, KRN7000, in mice with the melanoma B16 hepatic metastasis and immunohistological study of tumor infiltrating cells.
Oncol. Res.
12
:
51
58
.
22
Nagaraj
,
S.
,
C.
Ziske
,
J.
Strehl
,
D.
Messmer
,
T.
Sauerbruch
,
I. G.
Schmidt-Wolf
.
2006
.
Dendritic cells pulsed with alpha-galactosylceramide induce anti-tumor immunity against pancreatic cancer in vivo.
Int. Immunol.
18
:
1279
1283
.
23
Taieb
,
J.
,
N.
Chaput
,
C.
Ménard
,
L.
Apetoh
,
E.
Ullrich
,
M.
Bonmort
,
M.
Péquignot
,
N.
Casares
,
M.
Terme
,
C.
Flament
, et al
.
2006
.
A novel dendritic cell subset involved in tumor immunosurveillance.
Nat. Med.
12
:
214
219
.
24
Chan
,
C. W.
,
E.
Crafton
,
H. N.
Fan
,
J.
Flook
,
K.
Yoshimura
,
M.
Skarica
,
D.
Brockstedt
,
T. W.
Dubensky
,
M. F.
Stins
,
L. L.
Lanier
, et al
.
2006
.
Interferon-producing killer dendritic cells provide a link between innate and adaptive immunity.
Nat. Med.
12
:
207
213
.
25
Blasius
,
A. L.
,
W.
Barchet
,
M.
Cella
,
M.
Colonna
.
2007
.
Development and function of murine B220+CD11c+NK1.1+ cells identify them as a subset of NK cells.
J. Exp. Med.
204
:
2561
2568
.
26
Vosshenrich
,
C. A.
,
S.
Lesjean-Pottier
,
M.
Hasan
,
O.
Richard-Le Goff
,
E.
Corcuff
,
O.
Mandelboim
,
J. P.
Di Santo
.
2007
.
CD11cloB220+ interferon-producing killer dendritic cells are activated natural killer cells.
J. Exp. Med.
204
:
2569
2578
.
27
Caminschi
,
I.
,
F.
Ahmet
,
K.
Heger
,
J.
Brady
,
S. L.
Nutt
,
D.
Vremec
,
S.
Pietersz
,
M. H.
Lahoud
,
L.
Schofield
,
D. S.
Hansen
, et al
.
2007
.
Putative IKDCs are functionally and developmentally similar to natural killer cells, but not to dendritic cells.
J. Exp. Med.
204
:
2579
2590
.
28
Guimont-Desrochers
,
F.
,
G.
Boucher
,
Z.
Dong
,
M.
Dupuis
,
A.
Veillette
,
S.
Lesage
.
2012
.
Redefining interferon-producing killer dendritic cells as a novel intermediate in NK-cell differentiation.
Blood
119
:
4349
4357
.
29
Allende
,
M. L.
,
T.
Sasaki
,
H.
Kawai
,
A.
Olivera
,
Y.
Mi
,
G.
van Echten-Deckert
,
R.
Hajdu
,
M.
Rosenbach
,
C. A.
Keohane
,
S.
Mandala
, et al
.
2004
.
Mice deficient in sphingosine kinase 1 are rendered lymphopenic by FTY720.
J. Biol. Chem.
279
:
52487
52492
.
30
Porgador
,
A.
,
J. W.
Yewdell
,
Y.
Deng
,
J. R.
Bennink
,
R. N.
Germain
.
1997
.
Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody.
Immunity
6
:
715
726
.
31
Welsh
,
R. M.
,
W. K.
Nishioka
,
R.
Antia
,
P. L.
Dundon
.
1990
.
Mechanism of killing by virus-induced cytotoxic T lymphocytes elicited in vivo.
J. Virol.
64
:
3726
3733
.
32
Kataoka
,
T.
,
K.
Takaku
,
J.
Magae
,
N.
Shinohara
,
H.
Takayama
,
S.
Kondo
,
K.
Nagai
.
1994
.
Acidification is essential for maintaining the structure and function of lytic granules of CTL. Effect of concanamycin A, an inhibitor of vacuolar type H(+)-ATPase, on CTL-mediated cytotoxicity.
J. Immunol.
153
:
3938
3947
.
33
Parekh
,
V. V.
,
A. K.
Singh
,
M. T.
Wilson
,
D.
Olivares-Villagómez
,
J. S.
Bezbradica
,
H.
Inazawa
,
H.
Ehara
,
T.
Sakai
,
I.
Serizawa
,
L.
Wu
, et al
.
2004
.
Quantitative and qualitative differences in the in vivo response of NKT cells to distinct alpha- and beta-anomeric glycolipids.
J. Immunol.
173
:
3693
3706
.
34
Himoudi
,
N.
,
S.
Nabarro
,
J.
Buddle
,
A.
Eddaoudi
,
A. J.
Thrasher
,
J.
Anderson
.
2008
.
Bone marrow-derived IFN-producing killer dendritic cells account for the tumoricidal activity of unpulsed dendritic cells.
J. Immunol.
181
:
6654
6663
.
35
Himoudi
,
N.
,
M.
Yan
,
G.
Bouma
,
D.
Morgenstern
,
R.
Wallace
,
B.
Seddon
,
J.
Buddle
,
A.
Eddaoudi
,
S. J.
Howe
,
N.
Cooper
,
J.
Anderson
.
2009
.
Migratory and antigen presentation functions of IFN-producing killer dendritic cells.
Cancer Res.
69
:
6598
6606
.
36
Guimont-Desrochers
,
F.
,
S.
Lesage
.
2013
.
Revisiting the prominent anti-tumoral potential of pre-mNK cells.
Front. Immunol.
4
:
446
.
37
Hayakawa
,
Y.
,
M. J.
Smyth
.
2006
.
CD27 dissects mature NK cells into two subsets with distinct responsiveness and migratory capacity.
J. Immunol.
176
:
1517
1524
.
38
Matloubian
,
M.
,
C. G.
Lo
,
G.
Cinamon
,
M. J.
Lesneski
,
Y.
Xu
,
V.
Brinkmann
,
M. L.
Allende
,
R. L.
Proia
,
J. G.
Cyster
.
2004
.
Lymphocyte egress from thymus and peripheral lymphoid organs is dependent on S1P receptor 1.
Nature
427
:
355
360
.
39
Feng
,
C.
,
K. J.
Woodside
,
B. A.
Vance
,
D.
El-Khoury
,
M.
Canelles
,
J.
Lee
,
R.
Gress
,
B. J.
Fowlkes
,
E. W.
Shores
,
P. E.
Love
.
2002
.
A potential role for CD69 in thymocyte emigration.
Int. Immunol.
14
:
535
544
.
40
Bankovich
,
A. J.
,
L. R.
Shiow
,
J. G.
Cyster
.
2010
.
CD69 suppresses sphingosine 1-phosophate receptor-1 (S1P1) function through interaction with membrane helix 4.
J. Biol. Chem.
285
:
22328
22337
.
41
Ullrich
,
E.
,
M.
Bonmort
,
G.
Mignot
,
B.
Jacobs
,
D.
Bosisio
,
S.
Sozzani
,
A.
Jalil
,
F.
Louache
,
E.
Bulanova
,
F.
Geissman
, et al
.
2008
.
Trans-presentation of IL-15 dictates IFN-producing killer dendritic cells effector functions.
J. Immunol.
180
:
7887
7897
.
42
Steel
,
J. C.
,
T. A.
Waldmann
,
J. C.
Morris
.
2012
.
Interleukin-15 biology and its therapeutic implications in cancer.
Trends Pharmacol. Sci.
33
:
35
41
.
43
Tomura
,
M.
,
W. G.
Yu
,
H. J.
Ahn
,
M.
Yamashita
,
Y. F.
Yang
,
S.
Ono
,
T.
Hamaoka
,
T.
Kawano
,
M.
Taniguchi
,
Y.
Koezuka
,
H.
Fujiwara
.
1999
.
A novel function of Valpha14+CD4+NKT cells: stimulation of IL-12 production by antigen-presenting cells in the innate immune system.
J. Immunol.
163
:
93
101
.
44
Yang
,
Y. F.
,
M.
Tomura
,
S.
Ono
,
T.
Hamaoka
,
H.
Fujiwara
.
2000
.
Requirement for IFN-gamma in IL-12 production induced by collaboration between v(alpha)14(+) NKT cells and antigen-presenting cells.
Int. Immunol.
12
:
1669
1675
.
45
Tsuji
,
H.
,
N.
Mukaida
,
A.
Harada
,
S.
Kaneko
,
E.
Matsushita
,
Y.
Nakanuma
,
H.
Tsutsui
,
H.
Okamura
,
K.
Nakanishi
,
Y.
Tagawa
, et al
.
1999
.
Alleviation of lipopolysaccharide-induced acute liver injury in Propionibacterium acnes-primed IFN-gamma-deficient mice by a concomitant reduction of TNF-alpha, IL-12, and IL-18 production.
J. Immunol.
162
:
1049
1055
.
46
Terme
,
M.
,
G.
Mignot
,
E.
Ullrich
,
M.
Bonmort
,
V.
Minard-Colin
,
A.
Jacquet
,
J. L.
Schultze
,
G.
Kroemer
,
C.
Leclerc
,
N.
Chaput
,
L.
Zitvogel
.
2009
.
The dendritic cell-like functions of IFN-producing killer dendritic cells reside in the CD11b+ subset and are licensed by tumor cells.
Cancer Res.
69
:
6590
6597
.
47
Pletneva
,
M.
,
H.
Fan
,
J. J.
Park
,
V.
Radojcic
,
C.
Jie
,
Y.
Yu
,
C.
Chan
,
A.
Redwood
,
D.
Pardoll
,
F.
Housseau
.
2009
.
IFN-producing killer dendritic cells are antigen-presenting cells endowed with T-cell cross-priming capacity.
Cancer Res.
69
:
6607
6614
.
48
Kataoka
,
T.
,
N.
Shinohara
,
H.
Takayama
,
K.
Takaku
,
S.
Kondo
,
S.
Yonehara
,
K.
Nagai
.
1996
.
Concanamycin A, a powerful tool for characterization and estimation of contribution of perforin- and Fas-based lytic pathways in cell-mediated cytotoxicity.
J. Immunol.
156
:
3678
3686
.
49
Hiserodt
,
J. C.
,
L. J.
Britvan
,
S. R.
Targan
.
1982
.
Characterization of the cytolytic reaction mechanism of the human natural killer (NK) lymphocyte: resolution into binding, programming, and killer cell-independent steps.
J. Immunol.
129
:
1782
1787
.
50
Haeryfar
,
S. M.
,
E.
Nagy
,
E.
Baral
,
G.
Krepart
,
R.
Lotocki
,
I.
Berczi
.
2000
.
Antiestrogens affect both pathways of killer cell-mediated oncolysis.
Anticancer Res.
20
(
3A
):
1849
1853
.
51
Oberg
,
L.
,
S.
Johansson
,
J.
Michaëlsson
,
E.
Tomasello
,
E.
Vivier
,
K.
Kärre
,
P.
Höglund
.
2004
.
Loss or mismatch of MHC class I is sufficient to trigger NK cell-mediated rejection of resting lymphocytes in vivo - role of KARAP/DAP12-dependent and -independent pathways.
Eur. J. Immunol.
34
:
1646
1653
.
52
Kobayashi
,
E.
,
K.
Motoki
,
T.
Uchida
,
H.
Fukushima
,
Y.
Koezuka
.
1995
.
KRN7000, a novel immunomodulator, and its antitumor activities.
Oncol. Res.
7
:
529
534
.
53
Walzer
,
T.
,
L.
Chiossone
,
J.
Chaix
,
A.
Calver
,
C.
Carozzo
,
L.
Garrigue-Antar
,
Y.
Jacques
,
M.
Baratin
,
E.
Tomasello
,
E.
Vivier
.
2007
.
Natural killer cell trafficking in vivo requires a dedicated sphingosine 1-phosphate receptor.
Nat. Immunol.
8
:
1337
1344
.
54
Jenne
,
C. N.
,
A.
Enders
,
R.
Rivera
,
S. R.
Watson
,
A. J.
Bankovich
,
J. P.
Pereira
,
Y.
Xu
,
C. M.
Roots
,
J. N.
Beilke
,
A.
Banerjee
, et al
.
2009
.
T-bet-dependent S1P5 expression in NK cells promotes egress from lymph nodes and bone marrow.
J. Exp. Med.
206
:
2469
2481
.
55
Yoshimoto
,
T.
,
K.
Takeda
,
T.
Tanaka
,
K.
Ohkusu
,
S.
Kashiwamura
,
H.
Okamura
,
S.
Akira
,
K.
Nakanishi
.
1998
.
IL-12 up-regulates IL-18 receptor expression on T cells, Th1 cells, and B cells: synergism with IL-18 for IFN-gamma production.
J. Immunol.
161
:
3400
3407
.
56
Nakagawa
,
R.
,
I.
Nagafune
,
Y.
Tazunoki
,
H.
Ehara
,
H.
Tomura
,
R.
Iijima
,
K.
Motoki
,
M.
Kamishohara
,
S.
Seki
.
2001
.
Mechanisms of the antimetastatic effect in the liver and of the hepatocyte injury induced by alpha-galactosylceramide in mice.
J. Immunol.
166
:
6578
6584
.
57
Hasegawa
,
H.
,
K.
Yamashita
,
D.
Otubo
,
S.
Fujii
,
T.
Kamigaki
,
D.
Kuroda
,
Y.
Kakeji
.
2014
.
Allogeneic DCG promote lung NK cell activation and antitumor effect after invariant NKT cell activation.
Anticancer Res.
34
:
3411
3417
.
58
Chiodoni
,
C.
,
A.
Stoppacciaro
,
S.
Sangaletti
,
G.
Gri
,
B.
Cappetti
,
Y.
Koezuka
,
M. P.
Colombo
.
2001
.
Different requirements for alpha-galactosylceramide and recombinant IL-12 antitumor activity in the treatment of C-26 colon carcinoma hepatic metastases.
Eur. J. Immunol.
31
:
3101
3110
.
59
Wilson
,
K. A.
,
S. R.
Goding
,
H. R.
Neely
,
K. M.
Harris
,
P. A.
Antony
.
2015
.
Depletion of B220+NK1.1+ cells enhances the rejection of established melanoma by tumor-specific CD4+ T cells.
OncoImmunology
4
: e1019196.
60
Wagner
,
J. A.
,
M.
Rosario
,
R.
Romee
,
M. M.
Berrien-Elliott
,
S. E.
Schneider
,
J. W.
Leong
,
R. P.
Sullivan
,
B. A.
Jewell
,
M.
Becker-Hapak
,
T.
Schappe
, et al
.
2017
.
CD56bright NK cells exhibit potent antitumor responses following IL-15 priming.
J. Clin. Invest.
127
:
4042
4058
.
61
Chaput
,
N.
,
C.
Flament
,
C.
Locher
,
M.
Desbois
,
A.
Rey
,
S.
Rusakiewicz
,
V.
Poirier-Colame
,
P.
Pautier
,
A.
Le Cesne
,
J. C.
Soria
, et al
.
2013
.
Phase I clinical trial combining imatinib mesylate and IL-2: HLA-DR+ NK cell levels correlate with disease outcome.
Oncoimmunology
2
: e23080.
62
Fujii
,
S.
,
K.
Shimizu
,
H.
Hemmi
,
M.
Fukui
,
A. J.
Bonito
,
G.
Chen
,
R. W.
Franck
,
M.
Tsuji
,
R. M.
Steinman
.
2006
.
Glycolipid alpha-C-galactosylceramide is a distinct inducer of dendritic cell function during innate and adaptive immune responses of mice.
Proc. Natl. Acad. Sci. USA
103
:
11252
11257
.
63
Toura
,
I.
,
T.
Kawano
,
Y.
Akutsu
,
T.
Nakayama
,
T.
Ochiai
,
M.
Taniguchi
.
1999
.
Cutting edge: inhibition of experimental tumor metastasis by dendritic cells pulsed with alpha-galactosylceramide.
J. Immunol.
163
:
2387
2391
.

The authors have no financial conflicts of interest.

Supplementary data