Human buffy coats from healthy volunteers were obtained from the Center of Clinical Transfusion Medicine Tübingen. Primary leukemia cells were cryopreserved from untreated patients after informed consent was obtained. The study was approved by our institutional review board to be in accordance with the ethical standards and with the Helsinki Declaration of 1975, as revised in 2013 (IRB approvals 483/2015BO2 and 137/2017BO2).

iNKT cells were expanded from human peripheral blood mononuclear cells (PBMCs) in iNKT-cell culture medium consisting of RPMI 1640 GlutaMAX™ Medium (ThermoFisher Scientific, Waltham, MA, USA), 10% FBS (fetal bovine serum, Biochrom, Berlin, Germany), 100 IU/ml penicillin–streptomycin (Lonza, Basel, Switzerland), 5.5 µM 2-mercaptoethanol (Roth, Karlsruhe, Germany), 0.1 mM non-essential amino acids (NEAA) (Gibco, Grand Island, New York, NY, USA), 10 mM HEPES (Gibco, Grand Island, New York, NY, USA), and 1 mM sodium pyruvate (Gibco, Grand Island, New York, NY, USA). 2 × 10⁶ PBMCs/ml that contained >0.05% iNKT cells were co-incubated with 100 ng/ml KRN7000 (Sigma-Aldrich, St. Louis, MO, USA), PBS44, or PBS57 and 100 IU/ml recombinant human interleukin 2 (rhIL-2, Novartis, Basel, Switzerland) in a 24-well culture plate. KRN7000 was developed and based on the structures of glycolipids discovered in the marine sponge Agelas mauritianus (12). PBS57 was developed as an alternate α-galactosylceramide to KRN7000, and in some assays generated iNKT-cell responses at lower concentrations than KRN7000 (13). Both PBS44 and PBS57 contain an unsaturated acyl chain, which may improve solubility and loading into CD1d. After 7 and 14 days, 1 × 10⁶ cultured cells were re-stimulated with 2 × 10⁶ irradiated (30 Gy, cesium-137 irradiator Gammacell 1000, Atomic Energy of Canada Limited, Chalk River, Canada) and glycolipid-pulsed autologous PBMCs (responder to feeder ratio 1:2) together with rhIL-2 (100 IU/ml) and the respective glycolipid (100 ng/ml) in a 12-well (second week) and 6-well (third week) culture plate. To generate glycolipid-pulsed autologous PBMCs, cells were co-incubated with 100 ng/ml of the respective glycolipid antigen at 37°C for 4 h prior to autologous restimulation. After a total of 21 days, cell culture was completed.

PBS57-loaded and unloaded human CD1d tetramers were obtained from the National Institutes of Health Tetramer Core Facility (Atlanta, GA, USA). The following antibodies were purchased from BD Biosciences (Franklin Lakes, NJ, USA) or BioLegend (San Diego, CA, USA): anti-CD3 (HIT3a/OKT3), anti-CD4 (RPA-T4), anti-CD8 (HIT8a), anti-CD25 (BC96), anti-IFN-γ (4S.B3), anti-IL-4 (MP4-25D2), anti-IL-17 (BL168). Fluorescence minus one controls were used for proper gating. To stain dead cells, eBioscience Fixable Viability Dyes eFluor™ 506 and 780 (ThermoFisher Scientific, Waltham, MA, USA) were used. Data were acquired on a LSR Fortessa flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) and analyses were performed with FlowJo 10.2 (Tree Star, Ashland, OR, USA).

Culture-expanded human iNKT cells were stained with PBS57-CD1d tetramer phycoerythrin (PE) and enriched with anti-PE MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany). CD3+ T cells were isolated from human PBMCs with anti-CD3 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany). A QuadroMACS™ Separator (Miltenyi Biotec, Bergisch Gladbach, Germany) and LS Columns (Miltenyi Biotec, Bergisch Gladbach, Germany) were used according to the manufacturer’s instructions.

Culture-expanded human iNKT cells were stained with 4′,6-diamidino-2-phenylindole (DAPI, Merck, Darmstadt, Germany), CD3, CD4, CD8, and PBS57-CD1d tetramer and purified on a FACS Aria II cell sorter (BD Biosciences, Franklin Lakes, NJ, USA).

To generate dendritic cells (DCs), plastic-adherent monocytes isolated from PBMCs were cultured for 6 days in RPMI 1640 GlutaMAX™ Medium (ThermoFisher Scientific, Waltham, MA, USA), 10% FBS (Biochrom, Berlin, Germany), 100 IU/ml penicillin–streptomycin (Lonza, Basel, Switzerland), 11.4 µM 2-mercaptoethanol (Roth, Karlsruhe, Germany), 0.1 mM NEAA (Gibco, Grand Island, New York, NY, USA), and 1 mM sodium pyruvate (Gibco, Grand Island, New York, NY, USA) supplemented with 50 ng/ml IL-4 and 100 ng/ml GM-CSF (Miltenyi Biotec, Bergisch Gladbach, Germany) every other day. Major-mismatched DCs (stimulators) were plated together with allogeneic CD3+ T cells (responders) at a 1:1 ratio and different doses of culture-expanded MACS or FACS purified donor iNKT cells. Cells were analyzed by flow cytometry for activation markers and proliferation after 1, 3 and 7 days, respectively.

Cells were stimulated with 1× eBioscience Cell Stimulation Cocktail (ThermoFisher Scientific, Waltham, MA, USA) for 4 h at 37°C in iNKT-cell culture medium. After staining surface antigens, cells were fixed and permeabilized (ThermoFisher Scientific, Waltham, MA, USA) prior to staining of intracellular and intranuclear antigens. Stained cells were measured using a LSR Fortessa flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) and analyses were performed with FlowJo 10.2 (Tree Star, Ashland, OR, USA).

CD3+ MACS purified T cells were resuspended in phosphate-buffered saline (PBS, Gibco, Grand Island, New York, NY, USA) and stained with CellTrace CFSE cell proliferation kit (BioLegend, San Diego, CA, USA) for 5 min at room temperature. Immediately after staining, cells were washed in pure FBS then two times in PBS supplemented with 5% FBS and finally resuspended in iNKT-cell culture medium. CFSE-labeled CD3+ T cells were tested in a MLR. The percentage of proliferating CD3+CD4+ and CD3+CD8+ T cells was determined by flow cytometric analysis.

Culture-expanded iNKT cells were co-incubated at increasing ratios with 4 × 10⁴ Jurkat cells (Clone E6-1, ATCC, Manassas, VA, USA), MOLT-4 cells (ATCC, Manassas, VA, USA), or primary leukemia cells in iNKT-cell culture medium for 16 h at 37°C in a 96-well round bottom plate. Leukemia cell lysis through iNKT cells was measured on a LSR Fortessa flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) as described previously (14). To identify tumor cells, PBS57-CD1d tetramer+ iNKT cells were excluded. Only tumor cell samples with a purity ≥90% were used for these experiments. In certain experiments, iNKT cells were culture-expanded from donor cells and co-incubated with primary leukemia cells from patients that received HLA-matched hematopoietic stem cell grafts from their respective donors (10/10-antigen/allele matched unrelated donors). Supernatants were analyzed with a Legendplex Human CD8/NK Panel (BioLegend, San Diego, CA, USA) according to the manufacturer’s instructions.

Student’s t test and analysis of variance were used for statistical analysis of two and three groups, respectively. P < 0.05 was considered statistically significant. Data were analyzed with Prism 7.01 (GraphPad Software, La Jolla, CA, USA). All experiments were performed in triplicates and repeated independently at least three times.

In our study cohort of 59 healthy volunteers, iNKT cells constitute 0.09% of PBMCs (range, 0.003–0.63). Although our previous studies showed that iNKT-cell/T-cell ratios of less than 1:20 are sufficient to protect from lethal GVHD in a murine model of allogeneic HCT (3), prior ex vivo expansion is required to obtain enough iNKT cells for cytotherapy in humans. We therefore, studied iNKT-cell expansion using the glycolipid KRN7000 and its derivatives PBS44 and PBS57. Figure 1A depicts the protocol of our 21-day cell culture with autologous restimulation. Cell numbers and immunophenotypes were assessed every 7 days by trypan blue vital stain and flow cytometry, respectively. Three weeks of cell culture resulted in a 7,130-fold (range, 1,852–16,886) expansion of iNKT cells with a purity of 30.4% (range, 8.4–62.1). However, stimulation with different glycolipids had no impact on iNKT-cell expansion capacity (Figures 1B,C). In addition, using the PBS57-loaded CD1d tetramer, iNKT cells could be purified to 95.7% (range, 78.4–99.9) by MACS or 94.2% (range, 78.7–100) by FACS following iNKT-cell culture (Figure 1D).

The immunoregulatory properties of iNKT cells largely depend on the release of effector molecules and cytokines. It has been shown previously that a Th2 immune bias of T cells induced by CD4+ iNKT cells is critical for tolerance induction and prevention of GVHD (3, 4, 7, 8). We therefore, studied the immune polarization of culture-expanded iNKT-cell subsets after stimulation with different glycolipids. Remarkably, we observed a preferential expansion of CD4+ iNKT cells from 40% (range, 9.9–85.4) to 63% (range, 12.0–98.8) (Figures 2A,B). In contrast, percentage of CD8+ iNKT cells did not change significantly whereas CD4−CD8− iNKT cells showed a relative decline after 21 days of cell culture (Figures 2A,B). However, no relevant difference of preferential CD4+ iNKT-cell expansion was noted when KRN7000, PBS44, or PBS57 were used for iNKT-cell stimulation (Figure S1A in Supplementary Material). Subsequently, we performed intracellular cytokine staining of surrogate cytokines IFN-γ, IL-4, and IL-17 to evaluate cytokine profiles of culture-expanded iNKT cells (Figure 2C). All iNKT-cell subsets were functional as assessed by effective cytokine production after 21 days of cell culture. The choice of glycolipid for cell culture did not affect immune polarization of culture-expanded iNKT cells (Figure S1B in Supplementary Material). As already indicated by an increasing CD4+ iNKT-cell population, we observed a significant shift to Th2-biased iNKT cells when compared with unstimulated iNKT cells prior to iNKT-cell expansion (Figure 2D). Detailed cytokine profiles of iNKT-cell subsets before and after cell culture are shown in Figure S1C in Supplementary Material. Our data indicate a preferential expansion of highly tolerogenic iNKT cells irrespective of the glycolipid used for cell culture.

GVHD is mediated by alloreactive donor T cells that encounter antigen-presenting cells in secondary lymphoid organs of the host prior to destruction of target tissues. Preclinical animal studies demonstrate that adoptively transferred iNKT cells inhibit activation and proliferation of such alloreactive donor T cells. Therefore, we tested culture-expanded human iNKT cells for their capacity to control alloreactivity in a MLR. Donor T cells were analyzed by flow cytometry to determine cellular activation. We found a significantly decreased expression of early activation marker CD69 at day 1 (40.0 versus 70.8%, p = 0.03) and IL-2 receptor α-chain CD25 at day 3 (Figures 3A,B) on CD3+ T cells when culture-expanded iNKT cells were added. Comparable results were found for CD4+ and CD8+ T-cell subsets (Figure 3B). Fluorescence-activated cell sorted CD4+ iNKT cells and double negative (DN) iNKT cells (Figure S4A in Supplementary Material) were both potent suppressors of T-cell activation (Figure S4B in Supplementary Material). To study proliferation, donor T cells were labeled with CFSE prior to the MLR. At day 7, CFSE-labeled CD3+, CD4+, and CD8+ T cells were analyzed by flow cytometry. The percentage of proliferating cells was decreased in a dose-dependent manner when culture-expanded iNKT cells were added to the MLR (Figures 3C,D). Using the KRN7000 analogs PBS44 and PBS57 for iNKT-cell expansion resulted in comparable suppression of donor T-cell activation and proliferation (Figure S2 in Supplementary Material). Our studies indicate control of alloimmunity through culture-expanded human iNKT cells.

In addition of regulating immune responses by secretion of various cytokines, iNKT cells are capable of killing malignant cells through granzyme B and perforin or Fas–Fas ligand pathways (15–17). Moreover, rapid transactivation of other effector cells such as natural killer cells has been observed (18). Therefore, we studied the direct cytotoxic properties of culture-expanded human iNKT cells against various leukemia cell lines. Dose-dependent cell lysis of MOLT-4 and Jurkat cells was observed when co-cultured with iNKT cells for 16 h (Figure 4A). Comparable results were found for the KRN7000 analogs PBS44 and PBS57 (Figures S3A,B in Supplementary Material). Co-incubation of culture-expanded iNKT cells with Jurkat cells induced the release of perforin, granzyme A, and granzyme B (Figure S3C in Supplementary Material). In addition, fluorescence-activated cell sorted DN iNKT cells more efficiently killed Jurkat cells compared with CD4+ iNKT cells (Figure S4C in Supplementary Material). Furthermore, we investigated whether culture-expanded human iNKT cells are capable of lysing patient leukemia cells. We found an efficient dose-dependent killing of AML M1 and AML M5 cells when co-incubated with iNKT cells for 16 h (Figure 4B). Finally, we expanded donor iNKT cells and co-incubated them with primary leukemia cells from hitherto untreated patients that subsequently received the respective HLA-matched hematopoietic stem cell graft to simulate GVL activity ex vivo. We could detect potent dose-dependent antileukemia effects although iNKT cells and patient leukemia cells were HLA-matched (Figure 4C). In conclusion, our data indicate that culture-expanded human iNKT cells exert potent cytotoxic functions while controlling alloreactive immune responses.