Human peripheral blood DNAM-1neg NK cells are a terminally differentiated subset with limited effector functions

Abstract

Natural killer (NK) cells are a heterogeneous population of innate lymphocytes whose potent anticancer properties make them ideal candidates for cellular therapeutic application. However, our lack of understanding of the role of NK cell diversity in antitumor responses has hindered advances in this area. In this study, we describe a new CD56^dim NK cell subset characterized by the lack of expression of DNAX accessory molecule-1 (DNAM-1). Compared with CD56^bright and CD56^dimDNAM-1^pos NK cell subsets, CD56^dimDNAM-1^neg NK cells displayed reduced motility, poor proliferation, lower production of interferon-γ, and limited killing capacities. Soluble factors secreted by CD56^dimDNAM-1^neg NK cells impaired CD56^dimDNAM-1^pos NK cell-mediated killing, indicating a potential inhibitory role for the CD56^dimDNAM-1^neg NK cell subset. Transcriptome analysis revealed that CD56^dimDNAM-1^neg NK cells constitute a new mature NK cell subset with a specific gene signature. Upon in vitro cytokine stimulation, CD56^dimDNAM-1^neg NK cells were found to differentiate from CD56^dimDNAM-1^pos NK cells. Finally, we report a dysregulation of NK cell subsets in the blood of patients diagnosed with Hodgkin lymphoma and diffuse large B-cell lymphoma, characterized by decreased CD56^dimDNAM-1^pos/CD56^dimDNAM-1^neg NK cell ratios and reduced cytotoxic activity of CD56^dimDNAM-1^pos NK cells. Altogether, our data offer a better understanding of human peripheral blood NK cell populations and have important clinical implications for the design of NK cell-targeting therapies.

Graphical abstract

Figure 1.

Low DNAM-1 expression defines a unique subset of CD56^dimNK cells displaying reduced expression of CD57, inhibitory KIRs, and LFA-1. (A) Representative fluorescence-activated cell sorter staining of healthy donor peripheral blood lymphocytes displaying the gating strategy used to identify NK cells (left) and CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cell subsets (right). Numbers show percentages of the gated populations. (B) Proportions of NK cell subsets defined as in panel A are shown as mean percentage ± standard deviation from 13 healthy donors. (C) Histograms showing DNAM-1 and CD16 expression on the 3 NK cell subsets defined as in panel A; CD56^brightDNAM-1^pos (red histogram), CD56^dimDNAM-1^pos (blue histogram), and CD56^dimDNAM-1^neg (green histogram). (D) Marker expression by NK cell subsets pre- and postovernight stimulation with IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL) is shown as percentages of positive cells or mean fluorescence intensity (MFI). Single data points represent the mean value of duplicate wells. Data are shown as mean ± standard error of the mean (SEM) from 4 to 9 healthy donor samples and were analyzed by using a 2-way ANOVA followed by a Tukey multiple comparisons post hoc test. *P < .05; **P < .01; ***P < .001; ****P < .0001.

Figure 2.

DNAM-1^posNK cell subsets display more dynamic and deliberate migration behavior than CD56^dimDNAM-1^negNK cells. CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cell subsets were fluorescence-activated cell sorter purified from healthy donor PBMCs. Cells were stimulated overnight with IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL). Time-lapse microscopy was performed by using an Olympus Xcellence IX81 microscope controlled by Xcellence RT software. Images were taken every 40 seconds for 40 to 90 minutes. Images were collected by using a 20× LUCPlanFLN lens with 0.45 NA and Olympus F-View II camera. (A) Representative images. (B) Single-cell tracking map of NK cell trajectories. Mean velocity (C), length of tracks (D), and displacement from origin (E) of individual NK cells were calculated. Each data point represents 1 single randomly selected NK cell tracked over 40 minutes. Results are displayed as the mean ± SEM of at least 6 representative time-lapse videos obtained from 3 independent experiments. Data were analyzed by using a 1-way ANOVA followed by a Tukey multiple comparisons post hoc test. ***P < .001; ****P < .0001.

Figure 3.

CD56^dimDNAM-1^negNK cells proliferate poorly and produce limited amount of IFN-γ and granulocyte-macrophage colony-stimulating factor (GM-CSF) in response to cytokine stimulation. (A) CD56^brightDNAM-1^pos (red), CD56^dimDNAM-1^pos (blue), and CD56^dimDNAM-1^neg (green) NK cell subsets were purified from healthy donor PBMCs and stained with carboxyfluorescein diacetate succinimidyl ester (CFSE). The proliferation of NK cell subsets was assessed by measuring CFSE dilution after 5 days of culture with IL-2 (500 IU) and IL-15 (10 ng/mL). One representative histogram from 4 individual donors is displayed. (B) CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cell subsets were purified from healthy donor PBMCs. Intracellular IFN-γ production was assessed after overnight stimulation with IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL). Representative fluorescence-activated cell sorting plots and cumulative data are shown. Each data point represents the mean percentage of IFN-γ⁺ NK cells obtained from culture duplicates; data are shown as mean ± SEM from 4 donor samples. (C) Healthy donor NK cells were purified into 6 subsets based on their expression of CD16 (+ or –), CD56 (bright [b] or dim [d]) and DNAM-1 (+ or –) and were cultured overnight with (striped fill pattern) or without (no fill pattern) stimulation with IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL). Concentrations of indicated cytokines in the culture supernatant were analyzed by cytometric bead array. Data were normalized to reflect the relative cytokine/chemokine production per 10 000 cells and are presented as mean ± SEM of duplicate wells from 20 individual donors over 10 independent experiments. Data were analyzed with a 1-way (B) or 2-way (C) ANOVA followed by a Tukey multiple comparison post hoc test. *P <. 05; **P < .01; ***P < .001; ****P <.0001. No significant difference between subsets was found in the nonstimulated samples in panel C.

Figure 4.

CD56^dimDNAM-1^negNK cells present poor killing capabilities and downregulate the killing activity of CD56^dimDNAM-1^posNK cells. (A) NK cells were enriched from PBMCs by Ficoll density gradient centrifugation followed by magnetic bead negative cell selection. K562 target cells were added to wells containing total NK cells in a 10:1 effector:target ratio. After 4 hours, cells were stained and analyzed for degranulation by quantifying the percentages of CD107a-positive cells after gating on CD56^brightDNAM-1^pos (red), CD56^dimDNAM-1^pos (blue), or CD56^dimDNAM-1^neg (green) NK cell subsets. Results are shown as percentages of positive cells ± SEM from 8 individual healthy donors, pooled from 2 independent experiments. (B-E) Fluorescence-activated cell sorting NK cell subsets were stimulated overnight in IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL) and added to wells containing K562 target cells in a 10:1 effector:target ratio. (B) After 4 hours of culture, cytotoxicity of NK cell subsets against K562 target cells was quantified by flow cytometry with dead target cells identified as carboxyfluorescein diacetate succinimidyl ester (CFSE)^neg Annexin V/propidium iodide (PI)^pos. Target cells alone were used as control (–). Data are presented as mean ± SEM of percentages of Annexin V⁺ and/or PI⁺ target cells from 8 individual donors run in duplicate. Each dot represents one technical replicate. Data were pooled from 5 independent experiments. (C) The conjugation time between NK cells and target cells leading to target cell death was analyzed by using time-lapse microscopy. Time to membrane ruffle corresponds to the time (in seconds) between initial cell contact between the NK cell and the target cell and the first sign of target cell membrane ruffle. Results are shown as mean ± SEM from 20 to 30 individual NK cell:target cell conjugations pooled from at least 6 time-lapse videos obtained from 3 independent experiments. (D-E) Cytotoxicity of NK cell subsets was assessed as in panel B, but cells were cultured with (+) or without (–) supernatant from CD56^dimDNAM-1^neg (D^neg), heat-inactivated CD56^dimDNAM-1^neg (HI D^neg), or CD56^brightDNAM-1^pos (CD56^bright) NK cells that had been stimulated overnight in IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL). Each graph shows the mean ± SEM of percentages Annexin V⁺ and/or PI⁺ target cells from 4 individual donors pooled from 2 independent experiments. Individual dots represent the mean value of duplicate wells for 1 donor. Data obtained from 1 individual donor are depicted with the same symbol. Data were analyzed by 2-way (A) or 1-way ANOVA without pairing (B-C) or a 1-way ANOVA with pairing (D-E) followed by a Tukey multiple comparison post hoc test. *P < .05; **P<.01; ***P < .001; ****P < .0001.

Figure 5.

CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^negNK cell populations display distinct gene expression profiles. CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cell subsets were purified from healthy donor PBMCs, and their gene expression profiles were analyzed by using RNAseq. Four independent samples, each one consisting of NK cell mRNAs pooled from 10 to 20 donors, were analyzed for each population. (A) Relative expression of Ncam1 and Cd226 mRNAs in reads per kilobase million (RPKM). Data were analyzed with a 1-way ANOVA followed by a Tukey multiple comparison post hoc test. (B) Unsupervised principal component analysis revealed 3 clusters corresponding to the 3 NK cell populations (Benjamini-Hochberg corrected P value, q = 0.05). (C) Heat map and hierarchical clustering of the most differentially expressed genes between the 3 populations (q = 0.05). Genes can be divided into 6 groups (A-F) according to their pattern of expression. (D) Heat map displaying selected NK cell-related genes. (E) Heat map displaying selected genes related to inflammation and immunosuppression. ***P < .001; ****P < .00001.

Figure 6.

CD56^dimDNAM-1^negNK cells are terminally differentiated. (A-B) CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cell subsets were purified from healthy donor PBMCs and cultured for 4 days in the presence of IL-2 (600 IU) (A) or IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL) (B). At days 1, 2, and 4, NK cells were analyzed by using flow cytometry for surface expression of CD56 and DNAM-1. Results are displayed as the percentage of cells falling within CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, or CD56^dimDNAM-1^neg gates. Data are shown as the mean ± SEM of duplicate wells from 5 individual donors pooled from 3 independent experiments. (C) Telomere length was determined by hybridization of a fluorescein isothiocyanate–labeled peptide nucleic acid probe to telomeric repeats in the DNA of NK cell subsets. For each NK cell subset, mean fluorescence intensity (MFI) values of the incorporated probe was normalized to K562 control cell fluorescence. Data are shown as mean ± SEM values obtained for 6 individual donors and are pooled from 3 independent experiments, with each symbol representing one individual donor. No significant difference was found by using a 1-way ANOVA on paired values. (D-E) Purified CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cells were cultured overnight in the absence (D) or presence (E) of IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL). Apoptosis was determined by measuring Annexin V uptake by using flow cytometry. Data are shown as mean ± SEM from 5 individual donors. Individual dots represent the mean value of duplicate wells, with each symbol representing one individual donor. Data were analyzed by using a 1-way ANOVA on paired values followed by a Tukey multiple comparison post hoc test. *P < .05; **P < .01.

Figure 7.

CD56^dimDNAM-1^negNK cells are enriched in the peripheral blood of patients with hematologic malignancies. (A-B) NK cell proportions in peripheral blood of patients with HL or DLBCL compared with that of age- and sex-matched healthy control donors was determined via flow cytometry by gating on live CD3^negCD56^pos NK cells. Representative fluorescence-activated cell sorter plots (A) and the mean ± SEM of duplicate wells from 7 to 10 individual donors (B) from each group are shown. (C) NK cell subset distribution was determined by flow cytometry from healthy donor and HL and DLBCL patient samples. Results are shown as the mean ± SEM of subset frequencies within the whole lymphocyte population (left) or the NK cell population (right). (D) Graph shows the ratio of CD56^dimDNAM-1^pos over CD56^dimDNAM-1^neg NK cell as mean ± SEM. (E) CD56^brightDNAM-1^pos, CD56^dimDNAM-1^pos, and CD56^dimDNAM-1^neg NK cell subsets were purified from healthy donor and HL or DLBCL PBMCs and stimulated overnight in IL-12 (10 ng/mL), IL-15 (100 ng/mL), and IL-18 (50 ng/mL). The following day, cells were added to wells containing K562 target cells in a 10:1 effector:target ratio. After 4 hours of culture, cytotoxicity of NK cell subsets against K562 target cells was measured by fluorescence-activated cell sorter staining for Annexin V/PI. Results are shown as mean ± SEM from 4 individual donors run in duplicate and pooled from 2 independent experiments. Individual dots represent the mean value of duplicate wells. Data were analyzed with a 1-way ANOVA (B,D) or a 2-way ANOVA (C,E) followed by a Tukey multiple comparison post hoc test. *P < .05; **P < .01; ***P < .001; ****P < .0001.