Background
Chimeric antigen receptor (CAR) expressing NK cells induce robust antitumor responses in patients with hematologic malignancies but have limited persistent without cytokines supplement. IL-15 and IL-21 promote NK cell expansion, survival, and possess both overlapping and distinctive functions. CAR NK cells engineered to express IL-15 have shown long-term persistence and potent antitumor activity, however, the pleiotropic effects of IL-21 on CAR NK cell-mediated tumor rejection are lacking. Therefore, we compared the therapeutic potency of CD19-specific human NK cells that constitutively express IL-15 or IL-21.
Methods
The codon-optimized minigene encoding cytokine IL-21 or IL-15, linked with CD19-CAR containing 4-1BB costimulatory domain by a P2A sequence were synthesized and sub-cloned into the pSFG retroviral vector yielding the BBz.IL21 or BBz.IL15 retroviral construct. To produce CD19-CAR retrovirus, 293T/17 cells were transfected with a combination of plasmids containing CAR constructs, RDF plasmid and PegPam3 plasmid. Human peripheral blood NK cells isolated from healthy volunteer donors were stimulated with CD16/CD137 mAb and expanded with 500 IU/ml IL-2 (PeproTech) for 5 days before transduced with CAR retrovirus in plates coated with RetroNectin. The expression of CAR, phenotype, degranulation, cytokine production, and cytotoxicity of CAR NK cells were analyzed on day 3 post transduction. To measure IL-15 and IL-21 production, supernatants were collected from CAR NK cells cultured in the absence of IL-2 for 24h and evaluated by ELISA.
Results
After transduction, all constructs were stably expressed by NK cells, with construct containing IL-21 (71.23 ± 4.05 %, n = 3) demonstrating higher transduction efficiency compared with the BBz.IL15 (61.13 ± 3.43 %), but slightly lower than the BBz construct (79.63 ± 3.83 %). High levels of IL-21 were found in the supernatant of BBz.IL21 NK cells after resting 24h (3464 ± 199 pg/ml), while no detectable IL-21 was observed on BBz and BBz.IL15 NK cells. NK cells expressing BBz.IL15 produced significantly more IL-15 than BBz.IL21 and BBz NK cells as baseline (BBz.IL15: 213 ± 14 pg/ml, BBz.IL21: 61.86± 4 pg/ml, BBz: 60.56 ± 11 pg/ml). Given that IL-15 or IL-21 could potentially induce antigen-independent NK cell proliferation, we compared the ability of each NK-cell group to maintain growth in the presence of IL-2. The expansion of CAR NK cells expressing IL-21 was significantly higher than that of CAR NK cells expressing IL-15, while the number of BBz.IL15 NK cells and BBz NK cells were comparable (fold expansion in 5 days: BBz.IL21: 26.6 fold, BBz.IL15: 15.9 fold, BBz: 17.2 fold), suggesting a synergistic effect of IL-21 and IL-2 on CAR NK cell expansion. BBz.IL21 NK cells shared a similar immunophenotypic profile to BBz.IL15 NK cells, as the expression of inhibitory receptors PD1, LAG3, TIM3 and NKG2A, as well as activation receptors NKG2D and CD69 were comparable. CAR engagement by CD19-postive Raji cells specifically induced cytokines production by CD19-CAR NK cells but not by control NK cells. IL-21 coexpression significantly enhanced IFN-γ and TNF-α production compared with CAR NK cells as well as CAR NK cells expressing IL-15 (BBz.IL21 vs. BBz.IL15, P < 0.05; BBz.IL21 vs. BBz, P < 0.05; BBz.IL15 vs. BBz, P > 0.05). Moreover, a higher percentage of degranulated NK cells was observed on BBz.IL21 NK cells compared to BBz.IL15 NK cells and BBz NK cells. To confirm this observation, NK cells were cocultured with Raji cells at various E:T ratios for 6h. CAR NK cells expressing IL-21 showed superior cytotoxicity, compared to CAR NK cells expressing IL-15.
Conclusion
CD19-CAR NK cells engineered to express IL-21 gene produced high levels of IL-21 during cultivation and enabled greater ex vivo expansion of CAR NK cells compared with CAR NK cells expressing IL-15. IL-21 coexpression induced a significant increase in IFN-γ, TNF-α production and degranulation compared with IL-15 coexpression, resulting in enhanced antitumor activity of CD19-CAR NK cells against CD19-postive lymphoma.
Disclosures
No relevant conflicts of interest to declare.
