Two for one: targeting BCMA and CD19 in B-cell malignancies with off-the-shelf dual-CAR NK-92 cells

Abstract

Background: Chimeric antigen receptor (CAR) T-cell therapy has proven to be a valuable new treatment option for patients with B-cell malignancies. However, by applying selective pressure, outgrowth of antigen-negative tumor cells can occur, eventually resulting in relapse. Subsequent rescue by administration of CAR-T cells with different antigen-specificity indicates that those tumor cells are still sensitive to CAR-T treatment and points towards a multi-target strategy. Due to their natural tumor sensitivity and highly cytotoxic nature, natural killer (NK) cells are a compelling alternative to T cells, especially considering the availability of an off-the-shelf unlimited supply in the form of the clinically validated NK-92 cell line.

Methods: Given our goal to develop a flexible system whereby the CAR expression repertoire of the effector cells can be rapidly adapted to the changing antigen expression profile of the target cells, electrotransfection with CD19-/BCMA-CAR mRNA was chosen as CAR loading method in this study. We evaluated the functionality of mRNA-engineered dual-CAR NK-92 against tumor B-cell lines and primary patient samples. In order to test the clinical applicability of the proposed cell therapy product, the effect of irradiation on the proliferative rate and functionality of dual-CAR NK-92 cells was investigated.

Results: Co-electroporation of CD19 and BMCA CAR mRNA was highly efficient, resulting in 88.1% dual-CAR NK-92 cells. In terms of CD107a degranulation, and secretion of interferon (IFN)-γ and granzyme B, dual-CAR NK-92 significantly outperformed single-CAR NK-92. More importantly, the killing capacity of dual-CAR NK-92 exceeded 60% of single and dual antigen-expressing cell lines, as well as primary tumor cells, in a 4h co-culture assay at low effector to target ratios, matching that of single-CAR counterparts. Furthermore, our results confirm that dual-CAR NK-92 irradiated with 10 Gy cease to proliferate and are gradually cleared while maintaining their killing capacity.

Conclusions: Here, using the clinically validated NK-92 cell line as a therapeutic cell source, we established a readily accessible and flexible platform for the generation of highly functional dual-targeted CAR-NK cells.

Keywords: Bispecific; Chimeric antigen receptor; Dual; Leukemia; Lymphoma; Myeloma; NK-92; Off-the-shelf.

Fig. 1

Generation and functional validation of dual-CAR NK-92. A Structural composition of the BCMA- and CD19-specific CARs used for the dual-CAR approach. B High CAR expression in transfected NK-92 cells, 24 h after electroporation with 50 µg/mL CAR-encoding mRNA (N = 26–28). C CD19 and BCMA-CAR surface expression kinetics in NK-92 over four days (N = 1). D Only target cells expressing cognate antigen are lysed, confirming CAR specificity. Statistical analysis was performed using ANOVA with Tukey’s correction for multiple comparisons (N = 3). E CAR expression of NK-92 transfected with mRNA encoding one (CD19-CAR NK-92 and BCMA-CAR NK-92) or both CARs (dual-CAR NK-92). Expression for dual-CAR NK-92 represents cells positive for both CARs (N = 19). F Dual-CAR NK-92 lyse BCMA⁺CD19⁺ (Daudi), as well as single BCMA⁺ (BCMA-K562) or CD19⁺ (CD19-K562) tumor cells (N = 3). Statistical analysis was performed using ANOVA with Dunnett’s correction for multiple comparisons relative to the dual-CAR NK-92 condition. *p < 0.05. **p < 0.01. ****p < 0.0001. BCMA: B-cell maturation antigen; CAR: chimeric antigen receptor; CS: co-stimulatory domain; H: hinge domain; scFv: single-chain variable fragment; SD: signaling domain; SP: signal peptide; TM: transmembrane domain

Fig. 2

Degranulation and activation of dual-CAR NK-92. A CD107a expression demonstrated antigen-specific degranulation of CAR-engineered NK-92 during 5 h of co-culture at a 1:2 E:T ratio (N = 3). B High granzyme B secretion by dual-CAR NK-92 in the supernatant of 4 h co-cultures (1:1 E:T ratio; N = 3–6). C Dual-CAR NK-92 also significantly secrete IFN-γ upon activation (16 h co-culture at 1:1 E:T ratio; N = 3). Statistical analysis was performed using ANOVA with Dunnett’s correction for multiple comparisons with dual-CAR NK-92 as a reference. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. BCMA: B-cell maturation antigen; CAR: chimeric antigen receptor

Fig. 3

The effect of irradiation on proliferation, viability, CAR expression and functionality of dual-CAR NK-92. A Proliferation of NK-92 cells (top) was successfully inhibited after irradiation and viability (bottom) gradually declined over the course of a week (N = 3). Follow-up on day 5 and 6 was not performed and, therefore, not shown in the graph. B Peak CAR expression (24 h post electroporation) is maintained following irradiation (N = 3). C Despite irradiation, cytotoxic activity of NK-92 towards Daudi, Namalwa and U266 cells was largely preserved (4 h co-culture at 1:1 E:T ratio). Statistical analysis in B and C between non-irradiated and irradiated conditions was performed using an unpaired, two-tailed student t test. *p < 0.05. BCMA: B-cell maturation antigen; CAR: chimeric antigen receptor

Fig. 4

Lysis of primary tumor samples by CAR NK-92. High cytotoxicity of CAR-engineered NK-92 towards two primary B-ALL (N = 2) and one primary MM (N = 1) tumor samples after a 4 h co-culture at different E:T ratios. ALL#, B-ALL sample number. MM#, MM sample number