Figure 3. In vivo model exhibits shared intrinsic regulators governing tumor infiltrating “exhausted” NK cells and those controlling NK cell “anergy”.
(A) Experimental protocol. Timeline of in vivo experiment depicting tumor engraftment, effector administration, tumor excision, and ex vivo analysis. NOD/SCID IL-2Rγnull (NRG) mice engrafted with 4 × 106 aggressive human pancreatic ductal carcinoma cells (PANC-1) received a single infusion of primary human NK cells (i.t.). They were allowed to grow until the tumor size plateaued, indicating decreased ability of the NK to control the tumors. Tumor burden was assessed daily. (B) Left panel: Graph illustrating progression of tumor growth (n = 6 mice). On day 14 these mice received single intratumor (i.t.) infusion of 1.1*107 human pNK from two healthy donors. The pink-shaded region on the graph corresponds to the period during which pNK cells were present within the tumor post injection (with exception of tumor-only group). The groups treated with pNK cells exhibited a significant reduction in tumor size, although this effect was short-lived, lasting only 4 days. This suggests a decrease in the tumor lysis capacity of the pNK cells after 4–5 days. Right panel: Graph based on data in left panel showing days 14–18 of the tumor growth and plateauing of the tumor size toward day 18 (experimental end point), suggesting the escape of the tumor cells from NK-mediated control. P values were calculated using one-way ANOVA with Tukeys post hoc multiple comparison test and are indicated within the graph. (C, D) FACS analysis: Tumors excised on day 18 were dissociated to single-cell suspensions. NK cells were subjected to flow cytometry and were differentiated by hCD45 expression. Representative graphs showing: Left panel: Percentage of pNK expressing CD107a and PD-1 (n = 3 mice) in naive vs TINK; Right panel: Relative MFI corresponding to the respective percentage values. (E, F) FACS analysis: Tumors excised on day 18 were dissociated to single-cell suspensions. NK cells were subjected to flow cytometry and were differentiated by hCD45 expression. Representative graphs showing: Left panel: Percentage of pNK expressing Egr2 and DGKα (n = 6 mice) in naive vs TINK; Right panel: Relative MFI corresponding to the respective percentage values. The P values were calculated using a two-tailed t test with pairing and are indicated within the graph (mean ± SEM) along with the number of experimental repeats. Each symbol represents an individual mouse. Source data are available online for this figure.