Primary CD33-targeting CAR-NK cells for the treatment of acute myeloid leukemia.

Acute myeloid leukemia (AML) is a malignant disorder derived from neoplastic myeloid progenitor cells characterized by abnormal proliferation and differentiation. Although novel therapeutics have recently been introduced, AML remains a therapeutic challenge with insufficient cure rates. In the last years, immune-directed therapies such as chimeric antigen receptor (CAR)-T cells were introduced, which showed outstanding clinical activity against B-cell malignancies including acute lymphoblastic leukemia (ALL). However, the application of CAR-T cells appears to be challenging due to the enormous molecular heterogeneity of the disease and potential long-term suppression of hematopoiesis. Here we report on the generation of CD33-targeted CAR-modified natural killer (NK) cells by transduction of blood-derived primary NK cells using baboon envelope pseudotyped lentiviral vectors (BaEV-LVs). Transduced cells displayed stable CAR-expression, unimpeded proliferation, and increased cytotoxic activity against CD33-positive OCI-AML2 and primary AML cells in vitro. Furthermore, CD33-CAR-NK cells strongly reduced leukemic burden and prevented bone marrow engraftment of leukemic cells in OCI-AML2 xenograft mouse models without observable side effects.

CD33 Delineates Two Functionally Distinct NK Cell Populations Divergent in Cytokine Production and Antibody-Mediated Cellular Cytotoxicity.

The generation and expansion of functionally competent NK cells in vitro is of great interest for their application in immunotherapy of cancer. Since CD33 constitutes a promising target for immunotherapy of myeloid malignancies, NK cells expressing a CD33-specific chimeric antigen receptor (CAR) were generated. Unexpectedly, we noted that CD33-CAR NK cells could not be efficiently expanded in vitro due to a fratricide-like process in which CD33-CAR NK cells killed other CD33-CAR NK cells that had upregulated CD33 in culture. This upregulation was dependent on the stimulation protocol and encompassed up to 50% of NK cells including CD56dim NK cells that do generally not express CD33 in vivo. RNAseq analysis revealed that upregulation of CD33+ NK cells was accompanied by a unique transcriptional signature combining features of canonical CD56bright (CD117high, CD16low) and CD56dim NK cells (high expression of granzyme B and perforin). CD33+ NK cells exhibited significantly higher mobilization of cytotoxic granula and comparable levels of cytotoxicity against different leukemic target cells compared to the CD33- subset. Moreover, CD33+ NK cells showed superior production of IFNγ and TNFα, whereas CD33- NK cells exerted increased antibody-dependent cellular cytotoxicity (ADCC). In summary, the study delineates a novel functional divergence between NK cell subsets upon in vitro stimulation that is marked by CD33 expression. By choosing suitable stimulation protocols, it is possible to preferentially generate CD33+ NK cells combining efficient target cell killing and cytokine production, or alternatively CD33- NK cells, which produce less cytokines but are more efficient in antibody-dependent applications.