One-Step Construction of Allogeneic CAR-NK Cells Preventing Rejection and Mediating Enhanced Anti-Tumor Responses

Chimeric antigen receptor (CAR)-armed immune cells have demonstrated promising efficacy in several hematologic malignancies. There is a great interest in developing allogeneic “off-the-shelf” cellular therapies to reduce manufacturing time and costs, allow multiple dosing, and increase accessibility. To mitigate the problem of rejection of allogeneic products, common approaches include abrogation of MHC I expression by deleting b2-microglobulin (B2M) or TAP genes to prevent host T cell killing, and expression of HLA-E, HLA-G or CD47 to inhibit host NK killing. Typically, B2M or TAP knockout is first carried out using CRISPR/CAS-9 and then, CAR is introduced by lentiviral transduction and thus requiring at least two steps of genetic manipulation.

Here, we describe the development of a single novel lentiviral gene construct incorporating a novel shRNA that selectively targets prevalent HLA-A, -B, -C (referred to as HLA-ABC) alleles while maintaining endogenous HLA-E and simultaneously expressing a CAR and a single chain HLA-E (SCE) or PD-L1. Our screening of shRNAs helped identify a novel shRNA (shRNA #1) targeting a unique sequence common to 121 prevalent HLA-ABC alleles but has two mismatches with HLA-E alleles. The data revealed that shRNA #1 universally decreased surface HLA-ABC in HEK293T, Jurkat T, and primary NK cells isolated from twelve unrelated donors. Furthermore, it did not decrease HLA-E expression in IFNγ pre-treated Jurkat T cells.

To investigate the effect of PD-1/PD-L1 interaction in the control of T cell responses to allogeneic NK cells. NK cells were transduced with a lentivector expressing PD-L1, and PD-L1⁺ NK cells were sorted and expanded. PD-L1⁺ NK cells were then co-cultured with HLA mismatched CD3⁺ T cells. The data revealed that PD-L1 could moderately inhibit allogeneic T cell responses, as indicated by lower percentages of CD69⁺, CD25⁺, and proliferated T cells after co-culture compared with the controls.

To inhibit allogeneic NK cell killing of HLA-ABC reduced cells, we tested three single-chain HLA-E (SCE) variants: SCE WT, SCE Y84A, and SCE Y84C. The SCE WT incorporates peptide derived from the HLA-G leader sequence (VMAPRTLFL), B2M, and HLA-E*01:03 heavy chain, whereas SCE Y84A and SCE Y84C have further mutations in HLA-E*01:03 heavy chain and/or the linker region. The data revealed that all three SCE variants supported surface HLA-E expression, with SCE Y84C showing the highest expression level. We also found that SCE Y84C was able to inhibit third party NK cells, as indicated by the lowest lysis of SCE Y84C⁺ target cells and decreased degranulation and IFNγ production upon coculture.

We show that combining HLA-ABC knockdown and PD-L1 or SCE expression conferred resistance to allogeneic immune cell killing in vitro and in vivo. The combination of HLA-ABC knockdown, CAR, and PD-L1 or SCE overexpression was achieved by one-step lentiviral transduction of NK cells with a novel CAR construct incorporating shRNA#1, CAR, and PDL-1 or SCE. To test the CAR-mediated target cell killing, we assayed the killing of Raji cells by CD19-CAR NK cells and OVCAR8 cells by MSLN-CAR NK cells in vitro. The CAR NK cells killed target cells efficiently, and unexpectedly, PD-L1 and SCE overexpression significantly enhanced CAR NK cell cytotoxicity, likely related to the upregulation of cytotoxicity-related genes and better proliferation fitness. In a xenograft model (NSG mice), the allogeneic CD19-CAR NK cells with HLA-ABC knockdown and PD-L1 or SCE overexpression escaped immune rejection (with third party PBMCs) while maintaining potent anti-tumor responses against Raji cells (as assessed by BLI).

In summary, we identified a potent shRNA that can specifically downregulate surface HLA-ABC while maintaining surface HLA-E expression, and showed that PD-L1 can moderately inhibit allogeneic T-cell responses, and SCE Y84C is potent in inhibiting host NK cell cytotoxicity. Transduction of NK cells with our gene construct that combines a CAR, shRNA#1, and PD-L1 or SCE expression allows one-step construction of allogeneic CAR-NK cells that escape host immune cell rejection in vitro and in vivo while mediating enhanced anti-tumor responses. Our approach could also be applied to other allogeneic immune cell-based therapies and therefore represents a significant advancement in enabling “off-the-shelf” allogeneic immunotherapies.