CD44v6 specific CAR-NK cells for targeted immunotherapy of head and neck squamous cell carcinoma

Abstract

Head and neck squamous cell carcinoma (HNSCC) is a major challenge for current therapies. CAR-T cells have shown promising results in blood cancers, however, their effectiveness against solid tumors remains a hurdle. Recently, CD44v6-directed CAR-T cells demonstrated efficacy in controlling tumor growth in multiple myeloma and solid tumors such as HNSCC, lung and ovarian adenocarcinomas. Apart from CAR-T cells, CAR-NK cells offer a safe and allogenic alternative to autologous CAR-T cell therapy. In this paper, we investigated the capacity of CAR-NK cells redirected against CD44v6 to execute cytotoxicity against HNSCC. Anti-CD44v6 CAR-NK cells were generated from healthy donor peripheral blood-derived NK cells using gamma retroviral vectors (gRVs). The NK cell transduction was optimized by exploring virus envelope proteins derived from the baboon endogenous virus envelope (BaEV), feline leukemia virus (FeLV, termed RD114-TR) and gibbon ape leukemia virus (GaLV), respectively. BaEV pseudotyped gRVs induced the highest transduction rate compared to RD114-TR and GaLV envelopes as measured by EGFP and surface CAR expression of transduced NK cells. CAR-NK cells showed a two- to threefold increase in killing efficacy against various HNSCC cell lines compared to unmodified, cytokine-expanded primary NK cells. Anti-CD44v6 CAR-NK cells were effective in eliminating tumor cell lines with high and low CD44v6 expression levels. Overall, the improved cytotoxicity of CAR-NK cells holds promise for a therapeutic option for the treatment of HNSCC. However, further preclinical trials are necessary to test in vivo efficacy and safety, as well to optimize the treatment regimen of anti-CD44v6 CAR-NK cells against solid tumors.

Keywords: CAR-NK cells; CD44v6; NK cells; chimeric antigen receptor; head and neck cancer; immune cell therapy; solid tumor.

Figure 1

Optimization of the anti-CD44v6 retroviral vector CAR delivery and expression in primary NK cells. (A) Schematic representation of the anti-CD44v6 CAR construct used to generate CAR-NK cells. The anti-CD44v6 scFv (aCD44v6 scFv) derives from the BIWA8 clone of the anti-CD44v6 humanized antibody bivatuzumab. A human IgG1 hinge domain (huIgG1 hinge) links the scFv to a CD8 transmembrane domain (CD8a TMD), CD28 costimulatory domain (CD28) and CD3ζ signaling domain (CD3ζ). The genes for the enhanced GFP (EGFP) reporter and the CAR are connected through a P2A self-cleaving peptide sequence. The constitutive expression of the CAR gene in primary NK is controlled by the cytomegalovirus (CMV) promoter. The expression cassette is cloned into a gamma retroviral vector (gRV). (B) Example of gating strategy and comparison between different MOIs of the gamma retroviral vectors produced in HEK293-T cells and used to transduce primary NK cells. The two-parameter dot plots represent EGFP+, EGFP+CAR+ and CAR+ expression in transduced primary NKs with gRV pseudotyped with BaEV on day 10 post transduction. CAR surface expression was detected by staining with an APC IgG1 monoclonal antibody. (C) NK-cell transduction and percentage of total transduced cells (EGFP+, EGFP+CAR+, CAR+) at days 3, 7, 10 and 14 post transductions with gRV pseudotyped with BaEV, RD114 and GaLV. Unconcentrated viral vectors (UNC) and three MOIs of 0.5, 1 and 5 concentrated viral vectors were used to transduce NK cells. Expanded NK cells (EXP pNKs) and the process control NK cells (PC pNKs) serve as controls. Data of five independent experiments using five different healthy NK cells donors is presented as mean and standard deviation.

Figure 2

Surface marker expression on healthy donor NK and genetically modified NK cells was analyzed using flow cytometry. The histograms (A–H) illustrate exemplary data showcasing surface marker expression on different cell populations: unstimulated pNK cells on day 0 (Day 0 pNKs – white), expanded NK cells (EXP pNKs - black), no-vector transduced, process control NK cells (PC pNKs - grey) and viral vector transduced NK cells on day 10 post transduction. For the viral vector transduced pNKs, the following populations are highlighted: CD56+EGFP+ subpopulation from pNK cells transduced with BaEV MOI 5 (dark orange), CD56+EGFP- subpopulation from pNK cells transduced with BaEV MOI 5 (light orange), CD56+EGFP+ subpopulation from pNK cells transduced with RD114 MOI 5 (dark blue), and CD56+EGFP- subpopulation from pNK cells transduced with RD114 MOI 5 (light blue). The column dot plots display the mean and individual data of the investigated markers (n=3 NK donors for activation/inhibitory markers, n=4 NK donors for check point inhibitors). Percentages of the transduced pNK cell subpopulation (CD56+EGFP+) and untransduced subpopulation (CD56+EGFP-) expressing the activation markers NKG2D (A), NKp44 (B), NKp46 (C), CD69 (D), inhibitory markers such as DNAM-1 (E), NKG2A (F), as well as the check point inhibitors TIGIT (G) and PD-1 (H).

Figure 3

Killing efficacy of anti-CD44v6 CAR-NK cells against different HNSCC cell lines after 4, 6, 8 and 24 h of co-culture. (A) Schematic representation of the kinetic killing assay in which target cells genetically modified to express firefly luciferase (F.luc) are co-cultured with effector NK cells. Killing efficacy is monitored by recording living cells through the use of D-luciferin as substrate. This substrate is reduced to oxyluciferin by living cells. The bioluminescent signal is quantified and normalized to indicate cell killing. (B–D) CD44v6 positive UT-SCC-14-F.luc; UT-SCC-42B-F.luc and SCC-25-F.luc cells are set in co-culture with EXP pNKs; PC pNKs and CD44v6 CAR-NKs at 2.5:1 and 1:1 effector to target ratios. Killing efficacy is quantified after 4, 6 and 8 h of co-culture (n=5 donors). (E–G) Killing efficacy against the same target cell lines was checked after 24 h using 2.5:1, 1:1 and 0.1:1 effector to target ratios (n=3 donors). Data is presented as mean and standard deviation of 5 or 3 independent experiments. Descriptive statistics were calculated using two-way ANOVA and Turkey’s multiple comparison test (* p<0.05; **p<0.01; ***p<0.001; ****p<0.0001).