A Layered Cloaking Strategy to Generate Allogeneic iPSC-Derived CD8 T-Cells That Evade NK Clearance

Gene-editing approaches have paved the way to manipulate allogeneic CAR-T cells for longer persistence by escaping host immune cell mediated rejection. While the design concepts for evading patient T-cell mediated rejection have been widely agreed upon, a lack of consensus remains around the optimal design for avoiding missing-self natural killer (NK) responses resulting from HLA Class I knockout. The most common approach for NK cell evasion includes knock in of HLA-E, a non-classical class I molecule with limited polymorphism that inhibits NK cells via binding to an inhibitory receptor NKG2A. HLA-E can also stimulate NK subsets expressing the activating receptor NKG2C, leading to accelerated allogeneic clearance of HLA-E engineered cells. The prevalence of NKG2C+ adaptive NK cells can increase after cytomegalovirus reactivation in patients who have undergone hematopoietic stem cell transplantation and lymphodepletion, common features of CAR-T cell eligible patients. Taking cues from the biology of adaptive NK cells, we designed a layered approach to cloak allogeneic iPSC-derived CD8 T-cells for protection against a broader NK cell phenotype. It has been long known that CD2 and CD58 interactions are pivotal for the activation and function of adaptive NK cells via formation of a stable immune synapse. We demonstrate that NKG2C+ adaptive NK cells express higher levels of the co-activating receptor CD2, and we also observe that CD58 is upregulated on iPSC-derived CD8 T-cells. To avoid rejection of class I deficient iPSC CD8 T cells by NK cells, we evaluated use of HLA-E in combination with disruption of NK-iPSC CD8-T cell synapse. To ensure effective inhibition of NK cells by engagement of NKG2A while minimizing activating NK cells via NKG2C, HLA-E expression levels were optimized by regulating expression of HLA-E by site-specific integration under the control endogenous and exogenous promoters. We show that surface expression level of HLA-E is critical to balance engagement of NKG2A while minimizing activation of NK cells via NKG2C. We also show that additional layering of synapse disruption by knockout of adhesion molecules such as CD58 and ICAM3 (CD50) leads to best-in-class NK cell protection against NKG2C+ NK donors. Our in vitro clearance assays demonstrate that HLA-E plus NK-T cell synapse disruption together are required for comprehensive NK evasion in iPSC derived CD8 T cells. Individually these approaches are ineffective for providing protection against all NK donors. We also show that incorporation of these allogeneic edits does not interfere with the ability of these iPSCs to differentiate into CD8 T-cells and function against disease relevant tumor cells in vitro. In summary, rejection by host NK cells is a critical consideration for allogeneic cell therapies incorporating HLA class I disruption. Our layered approach of incorporating HLA-E and synapse disruption allows for universal protection against broader subsets of NK cells. After decades of development, the promise of off-the-shelf allogeneic CAR-T cells from iPSCs is closer to becoming a clinical reality. With our scalable manufacturing platform, we are making best-in-class cytotoxic T-cells in stirred-tank reactors and with gene-edited starting material incorporating a cloaking strategy we will be able to rapidly bring affordable allogeneic T-cells for clinical application against a wide range of disease indications.