Fig 2: Active NK cells with a unique transcriptome profile. a Top two enriched gene sets (ranked by normalized enrichment score) of five different datasets from GSEA of the “Inflamed NK” cluster compared to the rest of the cells were plotted. b The expression of CD69 in the BM sample was shown as a violin plot. The y-axis represents log-normalized expression value. c Module score was calculated using up-regulated DEGs of “Active NK” (left) or “Inflamed NK” (right) cluster from BM sample and plotted via violin plots. d Up-regulated IEGs from “Active NK” cluster were plotted using heatmap of the BM sample. e The expression of CXCR4 in the BM sample was shown as a violin plot. The y-axis represents log-normalized expression value. f Percentage of CXCR4+ NK cells (gated on Lin-CD56+ cells) was evaluated via flow cytometry. g The expression of CXCR4 in CD57+/-, CD62L+/-, or NKG2A+/- CD56dim NK populations from BM was assessed via flow cytometry (top). Percentage of CXCR4+ cells within each population were quantified (bottom). n = 6 from two to five independent experiments. Paired Student’s t test was used for the statistical analysis. *P < 0.05; **p < 0.01; ***p < 0.001; n.s. stands for “not significant.” Source data for f and g are provided as a Source Data file. See also Supplementary Fig. 5

Fig 3: Engagement of NKG2A during exhaustion rescues defects in cytotoxicity and cytokine production.(A) NK cells were incubated with both positive (anti-NKp46 plus MICA/B) and inhibitory (anti-NKG2A) stimuli or with positive stimuli plus isotype IgG. NK cells plated on isotype-bound cells served as a control for stimulation through the low-affinity Fc receptor CD16. NK cells were incubated with K-562 target cells (E/T 2:1), and live cell killing assays were performed using the IncuCyte platform as previously described. (B) Percentage of live targets was quantified for each group of cells at multiple time points (n = 6). One-way ANOVA with Dunnett’s multiple comparisons was used for statistical analysis (with each time point compared independently). (C and D) Expression of granzyme B (C) (n = 5) and TRAIL (D) (n = 4) was assessed via flow cytometry. One-way ANOVA with multiple comparisons was used for statistical analysis. (E) Irradiated P815 cells were coated with either anti-NKp30 and anti-NKG2A or anti-NKp30 and isotype IgG. P815 coated in isotype alone served as a negative control. NK cells were incubated with P815 target cells (E/T 2:1) for 5 days and subsequently stimulated with PMA/ionomycin for 4 hours. Following stimulation, cytokine production was assessed via flow cytometry (n = 5). One-way ANOVA with multiple comparisons was used for statistical analysis (with IFN-? and TNF-a compared independently). *P = 0.05; **P < 0.01.

Fig 4: CD11b, NKp46, CD62 L, CD127, and NKG2A of IL-12/15/18 preactivated CD3+ CD56+ NKT-like cells show evident and consistent change trends. (a) MFI shows the expression profile of the cell surface antigens. n = 8. none P > 0.05, **P < 0.01,***P < 0.001 (error bars, mean ± SEM). (b) Overlaid histograms and IFN-? production by cytokines trained cells restimulated with the same combinations of cytokines on day 1 and day 7. ***P < 0.001(error bars, mean ± SEM).

Fig 5: Intrinsic proliferation ability of NK cells correlated with EZH2 expression. (A) PBMCs were labeled with CFSE and cultured in vitro under control or preactivation conditions. The upper panel shows overlaid histograms of one representative experiment of the CFSE dilution profile in all 4 NK cell subsets on day 3. The lower panel shows representative overlaid histograms of the expression profile of EZH2 in the same cells as in the upper panel. (B) Upper panel: representative histogram of 3 generations of CFSE-labeled NK cells in PBMCs cultured in vitro on day 5 post preactivation with IL-12/18; lower panel: representative overlaid histogram of the expression of CD57, NKG2A, and EZH2 in the first 3 generations of NK cells. (C, D) Summary bar graph showing the expression levels of EZH2 in all 4 indicated NK cell subsets (C) and in the first 3 generations of NK cells (D) on day 3 and day 5 after preactivation with IL-12/18. (E) CFSE-labeled PBMCs were cultured in vitro under control or preactivation conditions, restimulated with IL-12/18 on day 5, and assayed for cell division and IFN-? production. The MFI of IFN-? in each generation of NK cells from the control and preactivation groups is shown as a summary bar graph. For all experiments, n = 5, *P < 0.05, **P < 0.01, and ***P < 0.001 (error bars, mean ± SEM). Data are representative of at least 3 independent experiments.