Notch Activation Rescues Exhaustion in CISH-Deleted Human Ipsc-Derived Natural Killer Cells to Promote in Vivo Persistence and Enhance Anti-Tumor Activity

Huang Zhu,Robert Blum,Zhengming Wu,Andres Bahena,Hanna Julie Hoel,Eivind Heggernes Ask,Kun-Liang Guan,Karl-Johan Malmberg,Dan S. Kaufman
DOI: https://doi.org/10.1182/blood-2018-99-112791
IF: 20.3
2018-01-01
Blood
Abstract:Cytokine-inducible SH2-containing protein (CIS) is a critical negative regulator of IL-15 signaling in natural killer (NK) cells and is encoded by CISH gene in human. Recent studies in a murine model demonstrate that CIS is a potent inhibitory checkpoint in NK cell-mediated tumor immunity. However, it is unclear if CIS similarly regulates human NK cell-mediated anti-tumor activity. Unlike CTLA-4 and PD1 that are expressed at the cell surface and can be blocked by antibody-mediated therapy, CIS is expressed intracellularly. Therefore, we deleted CISH gene in human induced pluripotent stem cells (iPSCS) using CRISPR/Cas9 technology and characterized the CISH-knockout iPSC lines. Our study demonstrates that CISH does not regulate maintenance of undifferentiated human pluripotent stem cells. We then derived NK cells from these gene-modified iPSCs using a two-stage in vitro hematopoietic-lymphoid differentiation protocol developed by our group. Initial hematopoietic differentiation of the iPSCs was unaffected by CISH deletion. However, deletion of CISH in iPSCs markedly delayed the second stage of in vitro NK cell differentiation. Specifically, whereas NK cell differentiation was typically fully complete with >90% NK cells after 4 weeks using WT iPSCs, the CISH-/- iPSC- cells only produced ~10% CD45+CD56+ NK cells at 4 weeks, though by 5 weeks these cultures were >80% NK cells. The CISH-/- iPSC-derived NK cells demonstrated typical NK surface maker expression, including CD94, CD16, NKG2D, NKp44, NKp46, FasL, and KIRs. Initial studies demonstrated that CISH-/- human iPSC-derived NK cells had significantly reduced ability to expand in vitro with evidence of NK cell exhaustion, including increased TIM-3 expression, decreased IFN-γ production and decreased cytotoxicity. RNA-seq analysis also confirmed that expression of exhaustion-related genes, including TIM3, CTLA4, and 2B4, were all significantly increased in CISH-/- NK cells. Interestingly, mass cytometric (CyTOF) analysis of CISH-/- NK cells with a panel of 36 phenotypic and functional NK cell markers identified an exhausted sub-population (increased expression of exhaustion marker TIM3 and inhibitory receptors such as ILT2 and Siglec7, as well as decreased proliferation marker Ki67). Importantly, this exhaustion could be rescued by co-culturing CISH-/- NK cells with Notch ligand-expressing OP9-DL4 stromal cells, leading to production of fully functionally mature NK cells. The CISH-/- NK cells stimulated by Notch maintained better expansion and improved cytotoxic function with low concentrations of either IL15 or IL2 compared to WT-iPSC-derived NK cells. CISH-/- NK cells demonstrate increased cytotoxic activity against leukemia cell lines K562 and MOLM-13 cells in vitro. Moreover, single-cell cytokine response analysis of CISH-/- NK cells after Notch stimulation showed >10 fold enrichment of polyfunctional cell subsets with effector cytokine production (Granzyme B, IFN-γ, MIP-1α, Perforin, TNF-α) compared with WT NK cells or NK cells from peripheral blood. In a MOLM-13 xenograft model, CISH-/- NK cells displayed significantly increased persistence in peripheral blood in comparison with WT NK cells (CISH-/- NK cells 6.81 ± 0.85 % vs WT NK cells 2.05 ± 0.25 N=5). More importantly, CISH-/- NK cells show significantly better control of tumor progression in the MOLM-13 xenograft model compared with WT NK cells. Together, these studies demonstrate CISH plays a key role to regulate NK cell activation-induced exhaustion and that Notch activation prevents this exhaustion to enable production of functionally hyperactive NK cells.
What problem does this paper attempt to address?