Fig. 3: Unmanipulated NK cells from CBMCs of Opt-Cs and Sub-Cs display distinct signatures.
Phenotype of unmanipulated NK cells in cryopreserved CBMCs from each of the cords used to manufacture the clinical CAR19/IL-15 NK cell product (a–c). For experiments in d–g, we used an independent cohort of CBUs. a, Phenotype of unmanipulated NK cells in CBMCs of Sub-Cs (n = 18 donors) versus Opt-Cs (n = 13 donors) by CyTOF. Only samples with viable CBMCs > 1,500 cells were analyzed. Frequencies of each cluster (1–4) are indicated; size and color of nodes within each cluster represent numbers of clustered cells. b, Bar graph shows the NK cell percentage within cluster 1 for Sub-Cs versus Opt-Cs. c, Heat map of marker expression within the main subclusters of clusters 1–4. Each column represents a major node within the SPADE tree clusters. The major nodes are representative of the majority of cells across all conditions. The expression level for each marker is represented from blue (low) to red (high). d, Heat map of differentially expressed genes (adjusted P value < 0.1 and absolute log2 fold change (FC) > 1.5) in unmanipulated NK cells from CBMCs of Opt-Cs (n = 18 samples) versus Sub-Cs (n = 14 samples). TPM, transcript per million. e, GSEA enrichment plots show differentially regulated pathways for NK cells from Opt-Cs versus Sub-Cs. f, Volcano plot (top) showing the log2 fold change (log2FC) in TF activity levels between NK cells from Sub-Cs (n = 9 samples, black) and Opt-Cs (n = 8 samples, yellow). Log plots (bottom) showing the top TF-binding motifs enriched in the preferentially open chromatin regions of unmanipulated NK cells from Opt-Cs and Sub-Cs (by HOMER). g, ATAC-seq tracks for selected genes in NK cells from Sub-Cs (n = 8 samples; top, black) versus Opt-Cs (n = 8 samples; bottom, yellow). Box plots comparing the gene-level accessibility score between the two groups. P values were determined by two-tailed Student’s t-test in b and g, a two-tailed Wilcoxon rank-sum test in the volcano plot and a one-tailed binomial test for motif enrichment analysis in f. q values were determined by two-tailed two-sample t-test with false discovery rate correction for multiple testing in e. Data are shown as the mean + s.e.m.
