Engineering Human Pluripotent Stem Cell-Derived Natural Killer Cells to Prevent CD16a Shedding for Enhanced Anti-Tumor Killing

Natural killer (NK) cells are lymphocytes able to recognize and lyse transformed cells without HLA-restriction and without prior sensitization, mediated through a repertoire of cell surface receptors. Solid tumors are typically more refractory to NK cell-mediated killing compared to hematologic malignancies. To improve NK cell-mediated killing of solid tumors, we have engineered NK cells to express a more stable form of CD16a. The IgG Fc receptor CD16a represents one of the most potent activators of NK cell anti-tumor activity. This receptor binds antibodies attached to target cells to initiate antibody-dependent cell-mediated cytotoxicity (ADCC). Importantly, CD16a undergoes a rapid down-regulation in expression by ADAM17 (A Disintegrin And Metalloprotease-17) mediated proteolytic cleavage upon cell activation. This loss of CD16a expression leads to reduced ADCC activity and decreased anti-tumor killing. Previous studies by our group identified cleavage sites in CD16a that are targeted by ADAM17 activity. Substitution of the serine at position 197 for a proline (S197P) in the middle of the cleavage region effectively blocks ADAM17-mediated CD16a cleavage. To evaluate the ability of this cleavage-resistant CD16a(S197P) to improve ADCC, we have used a system previously developed by our group to efficiently produce NK cells from both human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs). These hESC/iPSC-derived NK cells have a phenotype and function similar to NK cells isolated from peripheral blood (PB-NK cells). However, unlike PB-NK cells, it is relatively easy to genetically-modify hESC/iPSC-derived NK cells by engineering at the stem cell level. Here, we use two strategies to generate iPSC-derived NK cells with stabilized CD16a expression. First, we used the Sleeping Beautytransposon system to stably express the cleavage-resistant CD16a(S197P) in iPSCs. Second, we have used the CRISPR/Cas9 system to delete the ADAM17 gene. This strategy allows us to compare specifically engineering stabilized CD16a expression to the broader effect of ADAM17-deletion which cleaves multiple cell-surface proteins.

NK cells derived from the CD16a(S197P) iPSCs are resistant to CD16a shedding upon activation with IL-12 and IL-18 as 97.5% of iPSC-NK cells are CD16+ before stimulation and 95.2% of cells are CD16+ after stimulation. In contrast, PB-NK cells demonstrate a decrease from 85% CD16+ cells before stimulation to 67% CD16+ cells after stimulation. Next, we demonstrate improved ADCC and NK cell degranulation (measured by CD107a expression) in this engineered iPSC-derived NK cell compared to non-engineered cells upon addition of the anti-Her2 antibody (ab) Herceptin to ovarian cancer cells with differing levels of Her2 expression. Specifically, CD16a(S197P) iPSC-NK cells stimulated with the Her2-high SKOV3 ovarian carcinoma cells were 26.7% CD107a+ cells with addition of anti-Her2 and 10.8% CD107a+ cells without ab. In contrast, unmodified iPSC-NK cells were 9.6% vs. 6.9% CD107a+ with and without anti-Her2; and PB-NK cells were 12.4% vs. 9.4% CD107a+ with and without anti-Her2. However, when co-cultured with the Her2-low MA-148 ovarian carcinoma none of the three NK cell populations demonstrated increased CD107a+ cells with anti-Her2 treatment compared to no ab controls. Similar results were obtained when testing the ADCC of these cells using the Cr-51 release cytotoxicity assay. Killing of SKOV3 cells by CD16a(S197P) iPSC-NK cells was markedly increased upon addition of anti-Her2 ab from 32.2% cytoxicity with ab vs. 11% cytotoxicity without ab. In contrast, unmodified iPSC-NK cells had less killing of the SKOV3 cells (8.1% vs. 3.7% with or without anti-Her2 ab) and PB-NK cell killing of SKOV3 cells was 20.5% vs. 9.2% with or without anti-Her2 ab. Currently, we are comparing anti-tumor activity of the CD16a(S197P) iPSC-NK cells to the ADAM17-deleted iPSC-NK cells, both as resting NK cells and IL12+IL18-stimulated NK cells. Overall, this strategy will allow us to improve the ability of readily available anti-tumor antibodies to improve NK cell-mediated killing of solid tumors and use engineered iPSC-derived NK cells as an "off-the-shelf" approach for allogeneic cell-based immunotherapies.