Human CD56^bright NK Cells Acquire Potent Anti-Leukemia Functionality Following IL-15 Priming

Natural killer (NK) cells are innate lymphoid cells that mediate anti-tumor responses via cytotoxicity and effector cytokine production. Human NK cells are divided into two subsets based on relative expression of CD56 (CD56^bright and CD56^dim) that classically participate in distinct functions. Cytotoxic CD56^dim NK cells respond to tumor targets without prior stimulation, resulting in target cell death and transient secretion of effector cytokines (e.g. IFN-γ). In contrast, immunoregulatory CD56^bright NK cells secrete abundant IFN-γ and other cytokines in response to cytokine receptor stimulation, but respond minimally to tumor target-based triggering. As a result of this dichotomy, translational strategies to enhance NK cell function for cancer immunotherapy have focused exclusively on the CD56^dim subset.

Based upon studies in mouse NK cells, we hypothesized IL-15 priming would enhance CD56^bright anti-tumor functionality. Primary human NK cells from healthy donors were purified (>95% CD56⁺CD3^-), cultured overnight in medium alone (control) or medium with 5 ng/mL rhIL-15 (primed), washed, and assayed for anti-tumor responses. IL-15 priming significantly enhanced multiple CD56^bright NK cell functional responses to the prototypical AML target cell line K562 (CD107a+: control 20% vs. primed 59%, p<0.001; IFN-γ+: 3% vs. 27%, p<0.001; TNF: 3% vs. 20%, p<0.001), as well as primary AML blasts (N=3 unique AML sample: CD107a+,7% vs. 30%, p<0.001; IFN-γ 2% vs. 14%, p<0.001; TNF: 2% vs. 22%, p<0.001). IL-15-priming of CD56^bright NK cells was evident after 1 hour, and peaked after only 6 hours. In addition, flow-sorted IL-15-primed CD56^bright NK cells exhibited markedly increased killing of leukemia target cells. Similar results for functional comparisons were observed using primary NK cells from AML patients. Unexpectedly, IL-15-primed CD56^bright NK cell anti-leukemia responses significantly exceeded those of IL-15-primed CD56^dim NK cells. Multidimensional CyTOF analyses revealed that the maturity status of CD56^dim NK cells did not determine the extent to which they could be primed by IL-15.

In response to IL-15, we observed selective activation of the PI3K/Akt/mTOR (4.2 fold increase CD56^bright NK cells, 1.2 CD56^dim NK cells, p<0.001) and Ras/Raf/MEK/ERK (1.9 fold increase CD56^bright NK cells, 1.2 CD56^dim NK cells, p<0.001) pathways in CD56^bright NK cells. The Jak/STAT pathway was strongly activated in both CD56^bright and CD56^dim subsets. These data suggested a signaling mechanism for preferential CD56^bright NK cell priming by IL-15. Indeed, small molecule inhibitors of PI3K/Akt/mTOR and Ras/Raf/MEK/ERK pathways abrogated the anti-tumor responses of IL-15-primed CD56^bright NK cells, supporting this idea. Several NK cell effector mechanisms were enhanced in IL-15-primed CD56^bright NK cells, likely contributing to their augmented anti-tumor responsiveness. These included increased cytotoxic effector protein levels (granzyme B and perforin), as well as improved conjugate formation with tumor targets. Furthermore, blocking experiments demonstrated that IL-15-primed CD56^bright NK cell anti-tumor responses depended on LFA-1, CD2, and NKG2D receptor interactions with target cells. Finally, since IL-15-based therapeutics are being investigated in clinical trials, we tested whether the IL-15 super-agonist complex ALT-803 primes CD56^bright NK cells in vivo in cancer patients. There was an increase in leukemia target cell-triggered degranulation (CD107a+ unprimed 7% vs. primed 30%, p<0.001) and cytokine production (IFN-γ+ 6% vs. 31%, p<0.01; TNF+ 2% vs. 20%, p<0.05) by CD56^bright NK cells 24 hours post-ALT-803 administration to multiple myeloma patients, compared to unprimed, pre-therapy values.

Collectively, these results suggest that CD56^bright NK cells play an under-appreciated anti-tumor role in settings of abundant IL-15, such as following lymphodepleting chemotherapy, during preparation for stem cell transplantation, at sites of inflammation, or after exogenous IL-15 administration. Since CD56^bright NK cells have different in vivo tissue localization (secondary lymphoid organs), distinct inhibitory, activating, and chemokine receptor expression compared to CD56^dim NK cells, and are thought to be the most abundant NK cell subset when considering all human tissues, this study identifies a promising NK cell subset to harness for cancer immunotherapy.