Development of a CAR-NK therapy regulated by the KLRC1 locus as a new platform for selective control of acute myeloid leukaemia

Abstract In the era of CAR-T cell therapies, growing efforts are underway to adapt these strategies to Acute Myeloid Leukaemia (AML). Natural killer (NK) cells offer a promising alternative due to their role in post-transplant tumour control and the association between their dysfunction and AML. NK cells recognize tumour cells through a balance of activating and inhibitory receptor signals. One such inhibitory receptor, NKG2A (encoded by the KLRC1 gene), forms a heterodimer with CD94 and recognizes HLA-E. The expression of these molecules has been implicated in mechanisms of AML relapse. Knockout (KO) of KLRC1 has enhanced tumour control in several cancer models, including AML. Moreover, KLRC1 demonstrates inducible expression in response to cellular activation, allogeneic AML recognition, and cytokines such as IL-2 and IL-12. Its induction by TGF-β in T cells has also been described, suggesting a similar regulatory mechanism may exist in NK cells. These characteristics make this locus an ideal target for CAR insertion, offering two key advantages: elimination of an inhibitory pathway and CAR expression limited to cellular activation and tumour recognition; particularly relevant in allogeneic transplant settings, where this approach may improve the safety profile compared to constitutive systems like EF1α. Objectives: To develop a “2-in-1” Chimeric Antigen Receptor-Natural Killer (CAR-NK) cell therapy platform by knocking out KLRC1 and inserting an anti-CD33 CAR into its locus, in-frame, to enhance both the efficacy and safety of AML-targeted therapy. Methods: CRISPR-Cas9 was used to target exon 2 of KLRC1, together with a single-stranded AAV (ssAAV) vector carrying the anti-CD33 CAR flanked by homology arms for integration via homology-directed repair (HDR). Successful recombination was confirmed by sequencing and flow cytometry. NK cells were isolated from leukapheresis samples of healthy donors. In vitro cytotoxicity was measured by bioluminescence or flow cytometry against AML cell lines and primary AML samples. For in vivo testing, NSG mice were engrafted with MOLM14-Luc/GFP or patient-derived AML samples, and CAR-NK cells were administered post-engraftment. Tumour burden and bone marrow were analysed to assess response. Results: Accurate, in-frame insertion of the CAR into the KLRC1 locus was confirmed by Sanger sequencing. Following a 7-day expansion with K562 feeder cells, approximately 50% of primary NK cells expressed the CAR, as assessed by flow cytometry. Genomic disruption of KLRC1 was confirmed by Inference of CRISPR Edits (ICE), showing 80% indel formation. CAR-NK cells showed significantly enhanced cytotoxicity against AML cell lines and primary AML samples across multiple effector-to-target ratios (p < 0.05). In vivo, a single low dose (2.5×10⁶) of CAR-KO-NK cells significantly reduced intramedullary leukemic burden compared to KO-only or mock NK cell treatments. However, KLRC1-KO NK cells exhibited reduced expansion, consistent with activation-induced cell death mechanisms previously described (Chijioke et al., Sci Rep, 2023). To address this, we developed a secondary strategy involving cis-targeted IL-2, which binds to a truncated EGFR reporter to selectively expand edited cells. This approach restored NK cell numbers both in vitro and in vivo, outperforming regular IL-2 treatment while maintaining specificity for KLRC1-expressing (i.e., successfully edited) NK cells. Reporter-positive NK cells increased from 50% to 80–90% after two weeks of cytokine treatment, with improved persistence compared to mock or KLRC1-KO NK cells. Enhanced early tumour control (day 15 post-infusion) was observed in both bone marrow and peripheral blood. In a longer patient-derived AML xenograft model, persistence of edited cells was observed up to day 45, while cells in control conditions (mock or regular IL-2) were undetectable by then. Conclusion: Targeted integration of a CAR into the inducible KLRC1 locus enhances NK cell anti-leukemic activity while limiting CAR expression to activation contexts. The cis-targeted IL-2 strategy further enables selective expansion and maintenance of edited NK cells, improving both efficacy and durability. While challenges remain in achieving long-term tumour control, this dual-editing and selection strategy represents a next-generation platform for safer, more effective AML immunotherapy.