Generation of Natural Killer Cells with Enhanced Function from a CRISPR/Cas12a-Edited Induced Pluripotent Stem Cell Line

Adoptive cell therapy using T cells to treat cancer is efficacious in a number of hematologic malignancies. Recently, natural killer (NK) cells have emerged as an alternative cell type for clinical utility given the low propensity for graft-versus-host disease, thereby making NK cells a potential off-the-shelf cell therapy. NK cells distinguish tumor from healthy tissue via multiple mechanisms, including recognition of stress ligands and loss of MHC class I expression. For instance, KIR mismatching enables allogenic NK cells to kill MHC-positive tumor cells similar to MHC-negative tumor cells. Effector function of allogeneic NK cells are typically diminished by limited functional persistence, as well as tumor-intrinsic immunosuppressive mechanisms, such as TGF-β, a pleiotropic cytokine that inhibits immune effector function. Gene editing, however, can overcome these biological limitations. We hypothesized that knockout of CISH, a negative regulator of IL-2/IL-15 signaling, would improve NK cell effector function, while knockout of the TGF-β receptor gene 2, TGFBR2, would render NK cells resistant to TGF-β mediated suppression. NK cells are typically isolated from either cord blood or peripheral blood of healthy donors but recent advances with induced pluripotent stem cells (iPSCs) allows a nearly unlimited supply of iPSC-derived natural killer cells (iNK).

In this study, we used CRISPR/Cas12a to generate edited iPSC lines that were differentiated into CD56+ iNK cells. Specifically, we generated TGFβR2^-/-, CISH^-/-, and TGFβR2^-/-/CISH^-/- iPSC clones with bi-allelic gene disruption confirmed by next generation sequencing. Importantly, we also confirmed that the edited clones were pluripotent. In particular, a minimum of 3 clones from each genotype were differentiated to CD56+ iNK cells. After differentiation, >90% of the cells expressed CD56 for all genotypes. Additionally, we observed the expression of canonical natural killer cell markers such as CD16, NKG2A, KIRs, NKp46, NKp44, and NKp30 within this CD56+ population.

We tested the effector function of TGFβR2^-/-, CISH^-/-, and TGFβR2^-/-/CISH^-/- iNKs in a variety of molecular and functional assays, including a spheroid killing assay and an in vitro serial killing assay. For example, we utilized a SK-OV-3 spheroid killing assay to determine the intrinsic ability for the iNK cells to kill tumor targets following the differentiation process. TGFβR2-/-, CISH-/-, and TGFβR2-/-/CISH-/- iNKs reduce the size of SK-OV-3 ovarian tumor spheroids more effectively than control iNK cells in the presence of exogenous TGF-β.

In conclusion, we have established an iPSC editing platform that can generate a near infinite supply of natural killer cells with enhanced tumor killing function, paving the way for future off-the-shelf cell therapies for application to broad oncology indications.