Engineering of potent CAR NK cells using non-viral Sleeping Beauty transposition from minimalistic DNA vectors

Abstract

Natural killer (NK) cells have high intrinsic cytotoxic capacity, and clinical trials have demonstrated their safety and efficacy for adoptive cancer therapy. Expression of chimeric antigen receptors (CARs) enhances NK cell target specificity, with these cells applicable as off-the-shelf products generated from allogeneic donors. Here, we present for the first time an innovative approach for CAR NK cell engineering employing a non-viral Sleeping Beauty (SB) transposon/transposase-based system and minimized DNA vectors termed minicircles. SB-modified peripheral blood-derived primary NK cells displayed high and stable CAR expression and more frequent vector integration into genomic safe harbors than lentiviral vectors. Importantly, SB-generated CAR NK cells demonstrated enhanced cytotoxicity compared with non-transfected NK cells. A strong antileukemic potential was confirmed using established acute lymphocytic leukemia cells and patient-derived primary acute B cell leukemia and lymphoma samples as targets in vitro and in vivo in a xenograft leukemia mouse model. Our data suggest that the SB-transposon system is an efficient, safe, and cost-effective approach to non-viral engineering of highly functional CAR NK cells, which may be suitable for cancer immunotherapy of leukemia as well as many other malignancies.

Keywords: CAR; CD19; NK cells; Sleeping Beauty; acute lymphoblastic leukemia; allogenic; non-viral; transposon.

Graphical abstract

Figure 1

Development of virus-free Sleeping Beauty transposon-engineered primary CAR NK cells (A) Graphic representation of the workflow for stable non-viral CAR NK cell generation. Primary NK cells were isolated from healthy human donors (Buffy coats), genetically modified using nucleofection-based electroporation, followed by expansion. (B) Schematic illustration of the cut-and-paste DNA transposition via the SB transposase resulting in stable CAR transgene integration into the genome of NK cells: 1–4: SB transposase enzymes (orange circles) translated from mRNA molecules (orange wavy line) bind to the ITRs (purple) of the transposon cassette encoding the GOI (red) and excise the transposable element from the donor minicircle (MC). The excised transposon integrates into a genomic location in the host genome. (C) Frequency of stably expressing tEGFR-positive CD19-CAR NK cells between day 4 and day 23 after nucleofection (not analyzed for all shown donors at every time point; n = 4–10). (D) Viability (gated on FSC/SSC) of CD19-CAR NK cells vs. NT NK cells during long-term cultivation using feeder-cell free IL-15 stimulation (n = 7–8). (E) Representative fluorescence-activated cell sorting (FACS) plots of CD19-CAR NK cells and NK cells showing the CD19-CAR/tEGFR expression in CD56⁺ cells (gated on viable and CD56⁺/CD3⁻ cells). (F) Analyses of NK cell receptor expression measured by flow cytometry after long-term cultivation (n = 1–4). (G) Representative FACS plots showing the percentages of NK-cell specific markers in CD19-CAR and NT NK cells. Percentages showing the expression of depicted NK cell marker gated on single cells, FSC/SSC, CD3-CD19-CD14-7AAD^neg cells, CD45^pos, CD56^pos cells (data from one representative donor). Mean ± SD, two-way ANOVA multiple comparison, paired t test (F) (n ≥ 3); ∗∗∗∗p ≤ 0.0001, ns p > 0.05. If not indicated, no statistical analyses were performed (n < 3).

Figure 2

Equivalent efficiency but safer genomic insertions for non-viral SB technology compared with traditional LV-generated CAR NK cells (A–D) Functionality of LV-transduced CAR NK cells was compared with non-viral SB-generated CAR NK cells in luciferase-mediated bioluminescence cytotoxicity assays for different E:T ratios (n = 3). The target cell lines NALM-6 and SUP-B15 were engineered to express eGFP and firefly luciferase (luc⁺/GFP⁺). (A and B) Target cell lysis (SUP-B15, luc⁺/GFP⁺), indicated by reduction of luciferase signal in target cells, was analyzed after co-incubation with LV- and SB-engineered CAR NK cells of the same donor after 4 h (A) and 8 h (B). (C and D) Target cell lysis (NALM-6, luc⁺/GFP⁺), indicated by reduction of luciferase signal in target cells, was analyzed after co-incubation with LV- and SB-engineered CAR NK cells of the same donor after 4 h (C) and 8 h (D). (E and F) Integration frequencies of LV and non-viral SB transposons in genomic safe harbours (GSH) of NK cells compared with in silico computer-generated, random positions in the genome. GSH are defined as regions of human chromosomes that fulfill the following five criteria shown of the x axis: not ultraconserved elements (UCEs); >300 kb away from miRNA genes; >50 kb away from transcriptional start sites (TSSs) up- and downstream; >300 kb away from genes involved in cancer; >300 kb away from genes outside transcription units (F). (E) Percentages of insertions fulfilling all five criteria. (G) Base composition of LV and SB target sites in the chromosomes of human NK cells. The degree of base conservation is depicted by the height of the letters. Shown are data from different NK cell donors after nucleofection with SB100X/MC CD19-CAR and transduction with LV vectors expressing CD19-CAR and cultured for three weeks in vitro (n = 3–5). (H) Determination of the numbers of integrations per genome as vector copy number (VCN) for SB-CD19-CAR bulk NK cells cultured for 3–4 weeks in vitro (n = 4). Mean ± SD, two-way ANOVA multiple comparison, (n ≥ 3); ∗ p ≤ 0.05, ∗∗ p ≤ 0.01, ∗∗∗ p ≤ 0.001, ∗∗∗∗ p ≤ 0.0001, ns p > 0.05. If not indicated, no statistical analyses were performed.

Figure 3

Potent effector function of SB-engineered CD19-CAR NK cells in vitro and in vivo (A) Specific tumor cell lyses of the CD19⁺ SUP-B15 ALL cell line analyzed by flow cytometry after 4 h of co-culture with NT and CD19-CAR NK cells at different E:T ratios (n = 5–8). (B) Representative fluorescence-activated cell sorting (FACS) plots showing the percentage of dead (DAPI⁺) SUP-B15 tumor cells (CFSE⁺) after 4 h of co-culture with NT or CAR NK cells (E:T 10:1). (C) Specific cytolytic activity of NT NK cells or CD19-CAR NK cells within 4 h of co-culture with CD19⁺ NALM-6 target cells at different E:T ratios (n = 3). (D) Specific cytolytic activity of NT NK cells or CD19-CAR NK cells within 4 h of co-culture with CD19⁻ K562 control cells at different E:T ratios (n = 3). (E) Schematic representation of experimental layout for in vivo assessment of CD19-CAR NK cells in NALM-6 (that express luciferase [luc⁺]) engrafted xenograft NSG mice. After inoculation of 0.5 × 10⁶ NALM-6/luc leukemia cells, engraftment was confirmed in every mouse by BLI analysis. On day 3, mice were injected with a single dose of 1 × 10⁷ NT or CD19-CAR NK cells i.v., followed by daily subcutaneous administration of IL-2 and BLI imaging every 3–4 days. (F) Representative BLI images of PBS injected UT mice, NT, and CD19-CAR NK cells after NALM-6/luc engrafted NSG mice 7, 11, and 14 days after leukemia engraftment are shown (n = 4 per group). (G) Total flux analysis of BLI analyses after leukemia engraftment (n = 4) (area under the curve). (H) Total flux analysis of femur or tibiae and spleen for all mice group treated with PBS, NT, and CD19-CAR NK cells at day 14 (the endpoint of the experiment) (n = 4). (I) Serum cytokine (INF-γ, GM-CSF) analyses from peripheral blood of all mice treated with PBS, NT and CD19-CAR NK cells at the endpoint of the experiment (n = 4). Data are representative for results obtained in at least two independent mice experiments. BLI. Mean ± SD, two-way ANOVA multiple comparison, are under the curve (G) (n ≥ 3); ∗ p ≤ 0.05, ∗∗ p ≤ 0.01, ∗∗∗ p ≤ 0.001, ∗∗∗∗ p ≤ 0.0001, ns p > 0.05.

Figure 4

Prolonged overall survival upon administration with two low-doses of SB CAR NK cells in vivo (A) Schematic representation of experimental layout for in vivo assessment of CD19-CAR NK cells in NALM-6 (that express luciferase (luc⁺)) engrafted xenograft NSG mice. After inoculation of 0.5 × 10⁶ NALM-6/luc leukemia cells, engraftment was confirmed in every mouse by BLI analysis. On day 3 and day 7: mice were injected with one single dose of 0.5 × 10⁷ NT or non-viral CD19-CAR NK cells i.v., followed by daily subcutaneous administration of IL-2 and BLI imaging every 3–4 days until day 27. (B) Representative BLI images of PBS injected UT mice, NT and CD19-CAR NK cells after NALM-6/luc engrafted NSG mice 6, 10, 13, 17, 21, 24, and 27 days after leukemia engraftment are shown (n = 5 for UT, n = 6 for NT and n = 5 for CAR NK cell treated mice group). (C) Total flux analysis of BLI analyses after leukemia engraftment for every time point of analyses (n = 5–6). (D) Kaplan-Meier plots showing the probability of survival for the two groups of mice treated with NT versus non-viral CD19-CAR NK cells described in B. Mean ± SD, two-way ANOVA multiple comparison, Kaplan-Meier (D), (n ≥ 3); ∗ p ≤ 0.05, ∗∗ p ≤ 0.01, ∗∗∗ p ≤ 0.001, ∗∗∗∗ p ≤ 0.0001, ns p > 0.05.

Figure 5

Increased lyses of B-ALL patient derived bone marrow cells by non-viral generated CD19-CAR NK cells (A) Workflow for SB-generated CAR-NK cell efficiency analyses in a close to clinic setting: Bone marrow cells of B-ALL pre-treated patients (four different patients) have been isolated and co-incubated ex vivo with CAR NK cells preparations from three different healthy donors (donors A, B, and C) and measured in fluorescence-activated cell sorting-based cytotoxic assay (B–D). Specific tumor cell lyses of thawed primary bone marrow derived cells from four differently pre-treated B-ALL patients (Pat) (Pat 1, Pat 2, Pat 3, Pat 4) after 2 h co-culture with NT and CD19-CAR NK cells derived from allogenic donor A (B) and donor B (C) and donor C (D) at different E:T rations (Pat 1–3: E:T 5:1, 2.5:1, 1.25:1, Pat 4: E:T 5:1 due to limited patient materiel). (E–G) Corresponding FACS-plots of CD19-CAR NK donor preparations used in (B), (C), and (D): Purity of NK cells were analyzed by gating on CD56⁺/CD3⁻ single cells followed by viability (7AAD⁻) and CAR expression (tEGFR). (H) Cytokine concentrations (pg/mL) in supernatants collected after 2 h of co-culture of four primary B-ALL patient-derived cells alone (Pat) with NT and CD19-CAR NK cells (CAR) derived from three different healthy NK cell donors (A, B, C) at an E:T 5:1. Shown are concentrations of TNF-α and MIP-1α. The other analyzed cytokines (GM-CSF, IFN-γ, IL-6 IL-10) were below threshold value of the assay (<30 pg/mL). Mean ± SD, two-way ANOVA multiple comparison, (n ≥ 3); ∗ p ≤ 0.05, ∗∗ p ≤ 0.01, ∗∗∗ p ≤ 0.001, ∗∗∗∗ p ≤ 0.0001, ns p > 0.05. If not indicated no statistical analyses were performed.