A Novel Method for Highly Efficient Ex Vivo Expansion and Genetic Engineering of Human Natural Killer Cells for Cancer Therapy
Nobuhiro Kubo,Yudai Murayama,Yasushi Kasahara,Chansu Shin,Minori Baba,Ryosuke Hosokai,Haruko Iwabuchi,Imamura Masaru,Akihiko Saitoh,Chihaya Imai
DOI: https://doi.org/10.1182/blood-2021-152356
IF: 20.3
2021-11-05
Blood
Abstract:Abstract Cancer immunotherapy with adoptive transfer of human leukocyte antigen-mismatched, CD19-targetd chimeric antigen receptor (CAR)-transduced natural killer (NK) cells has attracted attention because of its efficacy and safety when infused in patients with refractory and relapsed B-cell lymphomas. However, generating clinical doses of CAR-NK cells is still a challenge. The methods for ex vivo expansion and genetic modification of primary human NK cells usually rely on the use of irradiated feeder cell lines, which has been restrictive due to high costs, scale-up difficulties, and licensing restrictions. Hence, novel strategies that do not require feeder cells will be beneficial in standardizing these types of cell therapies. In this study, we show the effectiveness of a novel feeder-free culture system in expanding NK cells ex vivo and generating CAR-NK cells. Unsorted peripheral blood mononuclear cells (PBMCs) collected from healthy donors were cultured with a reagent containing dissolvable microspheres that are conjugated with anti-CD2/NKp46 antibodies (Cloudz TM Human NK Cell Expansion Kit) and a combination of multiple cytokines, including interleukin (IL)-2, IL-12, IL-18, and IL-21 in medium supplemented with 10% fetal bovine serum. The activated NK cells were transduced using the RD114-pseudotyped retrovirus vector. To test whether the reagent promoted ex vivo NK cell expansion, we cultured PBMCs from 13 donors. The percentage of NK (CD56 + CD3 -) cells of initially isolated PBMCs was 15.3±7.5%. In the presence of multiple cytokine combinations, NK cell purity gradually increased and reached 91.6±7.6% by day 21. The NK cells expanded to 75.6±59.2-fold at day 10, 334±217-fold at day 14 and 1,542±913-fold at day 21. The expanded NK cells degranulated and produced intracellular cytokines upon exposure to K562 myeloid leukemia cells. The NK cells efficiently killed myeloid leukemia cells, such as K562, THP1, and KG1. The expression pattern of killer cell immunoglobulin-like receptors on NK cells remained unchanged. The expression of activating NK cell receptors, including NKp30 and NKp44, increased after 21 days of culture. Thereafter, a gene transfer to the primary human NK cells was conducted. We tested transduction efficiency and yields of modified cells on 7 days after the procedure by empty-vector transduction into NK cells expanded ex vivo for 3, 6, and 10 days (n=3, each). The results were presented as means ± standard deviation; 55.6±11.6%, 61.6±14.1%, and 73.6%±6.2% for GFP positivity in NK cells and 5.7±1.5 folds, 56.4±42.2 folds, and 12.7±5.8 folds for yields of modified cells. We selected the condition in which the transduction was carried out using NK cells expanded ex vivo for 6 days, although the differences were not statistically significant. Next, anti-CD19 CAR with a 4-1BB costimulatory and CD3z domain was transduced into NK cells. We confirmed high transduction efficiency (59.8% ±20.5%, n=3) and high CAR protein expression on the cell surface, while NK cells maintained their purity and minimal T cell outgrowth was observed. CAR-NK cells maintained their proliferative status and further expanded 15.2±6.4-fold after 7 days of the procedure. To determine whether the generated anti-CD19 CAR-NK cells had a specific effect on B cell malignancies, a CD107a mobilization assay, intracellular cytokine assay, and a flow cytometry-based cytotoxicity assay was employed. We found that CAR transduction could render NK cells to generate specific and powerful responses against CD19-positive, NK-resistant leukemia and lymphoma cell lines, such as BCR-ABL-positive acute lymphoblastic leukemia (ALL) OP-1, Burkitt lymphoma Raji, and KMT2A-rearranged ALL RS4;11, at various effector: target (ET) ratios. For example, in a 4 hour -assay, the cytotoxic effects of anti-CD19 CAR-NK cells showed 86.9±0.2% cytotoxicity against OP-1, while mock NK cells showed 17.4±2.9% cytotoxicity (ET ratio 1:1). In conclusion, this study revealed highly efficient functions of the novel feeder-free culture system, including highly efficient ex vivo expansion of primary human NK cells and generation of genetically modified NK cells for cancer immunotherapy. In future studies, we will investigate large-scale cultures using specialized flasks and GMP-grade reagents for clinical translation and the in vivo activities of the cell products in mouse xenograft models. Disclosures Imai: Juno Therapeutics: Patents & Royalties: chimeric receptor with 4-1BB signaling domain.
hematology