HYPOXIA-RESISTANT NATURAL KILLER CELLS

Abstract

NK cells and NK cell lines are modified so as to have a more cytotoxic phenotype, namely to have reduced function of Src homology 2 domain-containing protein tyrosine phosphatase 1 (SHP-1). Methods of making the modified cells and cell lines, compositions comprising the modified cells and cell lines, as well as uses of said cells, cell lines and compositions in therapy are also provided.

Claims

1-33. (canceled)

34. A method of treating cancer comprising administering to a patient a natural killer (NK) cell or NK cell line that has been modified to have reduced function of Src homology 2 domain-containing protein tyrosine phosphatase 1 (SHP-1).

35. The method according to claim 34, wherein the NK cell or NK cell line modification is a genetic modification.

36. The method according to claim 34, wherein the NK cell or NK cell line modification involves genetically knocking down SHP-1 expression.

37. The method according to claim 34, wherein the NK cell or NK cell line modification is a transient modification.

38. The method according to claim 37, wherein the transient modification is an RNAi knockdown of SHP-1 expression.

39. The method according to claim 37, wherein the transient modification involves expressing an inactive form of SHP-1.

40. The method according to claim 39, wherein the inactive form of SHP-1 is expressed from an extra-chromosomal nucleic acid.

41. The method according to claim 39, wherein the inactive form of SHP-1 is expressed via an mRNA-based transfection system.

42. The method according to claim 41, wherein the mRNA-based transfection system is the Maxcyte GT system.

43. The method according to claim 34, wherein the NK cell or NK cell line modification involves expressing a dominant negative form of SHP-1.

44. The method according to claim 43, wherein the dominant negative form of SHP-1 is inducible.

45. The method according to claim 44, wherein expression of the dominant negative form of SHP-1 is induced by an antibody or doxycycline.

46. The method according to claim 34, wherein SHP-1 function is reduced by a phosphatase inhibitor.

47. The method according to claim 46, wherein the phosphatase inhibitor is a reversible inhibitor.

48. The method according to claim 46, wherein the phosphatase inhibitor is selected from tyrosine phosphatase inhibitor 1 (TPI-1), sodium stibogluconate (SSG), sodium orthovanadate (SOV), and sodium fluoride (NaF).

49. The method according to claim 34, wherein the NK cell is selected from an iPSC-derived NK cell and an umbilical cord-derived NK cell.

50. The method according to claim 34, wherein at least some of the cancer resides in a hypoxic environment.

51. The method according to claim 34, wherein the cancer is a blood cancer or a solid cancer.

52. The method according to claim 51, wherein the blood cancer is selected from acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CIVIL), Hodgkin's lymphoma, non-Hodgkin's lymphoma, T-cell lymphoma, B-cell lymphoma, asymptomatic myeloma, multiple myeloma, active myeloma, and light chain myeloma.

53. A method of treating a solid cancer at least partially residing in a hypoxic environment, the method comprising administering to a patient a NK cell or NK cell line that has been modified to have reduced function of Src homology 2 domain-containing protein tyrosine phosphatase 1 (SHP-1).

Description

EXAMPLES

[0059] The present invention is now described in more and specific details in relation to the production of NK cells, modified to exhibit increased cytotoxic activity, wherein specific embodiments are illustrated with reference to the accompanying drawings in which:

[0060] FIG. 1a-1d show the effect of hypoxia on NK cell cytotoxicity;

[0061] FIG. 2a-1d show the effect of hypoxia on the expression of various cytotoxicity-related markers in NK cells;

[0062] FIG. 3a-3e show the effect of hypoxia on the phosphorylation levels of ERK and STAT3 in NK cells;

[0063] FIG. 4a-4e show the effect of hypoxia on SHP-1 and SHP-2 activation in NK cells;

[0064] FIG. 5a-5c show the effect of inhibiting SHP-2 on ERK signalling, STAT3 signalling and NK cell cytotoxicity;

[0065] FIG. 6a-6c show the effect of inhibiting SHP-1 on ERK signalling, STAT3 signalling and NK cell cytotoxicity;

[0066] FIG. 7 shows the baseline expression of TRAIL on KHYG-1 cells;

[0067] FIG. 8 shows the expression of TRAIL and TRAIL variant after transfection of KHYG-1 cells;

[0068] FIG. 9 shows the expression of CD107a after transfection of KHYG-1 cells;

[0069] FIG. 10 shows the effects of transfecting KHYG-1 cells with TRAIL and TRAIL variant on cell viability;

[0070] FIG. 11 shows the baseline expression of DR4, DR5, DcR1 and DcR2 on both KHYG-1 cells and NK-92 cells;

[0071] FIGS. 12-14 show the effects of expressing TRAIL or TRAIL variant in KHYG-1 cells on apoptosis of three target cell populations: K562, RPMI8226 and MM1.S, respectively; and

[0072] FIG. 15 shows mitigation of NK cell fratricide by knocking down DR5 expression.

[0073] DNA, RNA and amino acid sequences are referred to below, in which: [0074] SEQ ID NO: 1 is an example gRNA for DR5; [0075] SEQ ID NO: 2 is an example gRNA for DR4; and [0076] SEQ ID NO: 3 is a second example gRNA for DR4.

Examples 1-3: Detailed Figure Overview

[0077] FIG. 1

[0078] Hypoxic NK cells show lower cytotoxicity against tumour cells. (a-b) Flow cytometric analysis of KHYG-1 cells cytotoxicity against tumour cells. KHYG-1 cells were incubated with K562 (a) or MM.1S (b) tumour cells for 4 h at different E:T ratios after cultivation at normoxic (20% 02) and hypoxic (1% 02) conditions for 24 h. Left panel: a representation of results from three experiments; Right panel: statistical analysis showing the percentage of tumour cells killed by NK cells (n=3, * P<0.05). (c) Western blotting analysis of the effects of hypoxia on the expression of hypoxia marker HIF-1α. NK cells were cultured in 20% or 1% 02 for 24 h, and then western blotting analysis was performed. (d) Flow cytometric analysis of the effects of hypoxia on NK cell viability by performing Annexin V-FITC/7-AAD staining. NK cells were cultured in 20% or 1% 02 for 24 h, then flow cytometric staining was performed (n=3, ns=non-significant).

[0079] FIG. 2

[0080] Hypoxia decreases the expression level of NK cell cytotoxicity related molecules. (a) Flow cytometric analysis of granzyme B and perforin expression in KHYG-1 (upper panel) and NK92 (lower panel) cells, respectively. KHYG-1 and NK92 were cultured in normoxic (20% 02) and hypoxic (1% 02) for 24 h, then intracellular staining was performed to analyze the expression of granzyme and perforin quantitatively. Left panel: Histogram overlays display representative examples of granzyme B, and perforin expression analyzed in normoxic and hypoxic cell samples compared to the fluorescence minus one (FMO) control; Right panel: statistical analysis of the flow cytometry data (n=3, * P<0.05, ** P<0.01). (b) Flow cytometric analysis of the intracellular level of IFN-γ in normoxic and hypoxic KHYG-1 and NK92 cells. Left panel: one of three representative flow cytometry results; Right panel: statistical analysis of the flow cytometry data (n=3, * P<0.05, ** P<0.01). (c) Flow cytometric analysis of the membrane staining of degranulation marker CD107a in normoxic and hypoxic KHYG-1 and NK92 cells. Left panel: one of three representative flow cytometry results; Right panel: statistical analysis of the flow cytometry data (n=3, *P<0.05, ** P<0.01). (d) Flow cytometric analysis of the membrane staining of activating receptor NKp30, NKp46, and NKG2D in normoxic and hypoxic KHYG-1 and NK92 cells. Left panel: one of three representative flow cytometry results; Right panel: statistical analysis of the flow cytometry data (n=3, * P<0.05, ** P<0.01).

[0081] FIG. 3

[0082] Hypoxia diminishes the phosphorylation level of ERK and STAT3. (a-c) Western blotting analysis shows the expression levels of the phosphorylated ERK and STAT3 in KHYG-1 (a), NK92 (b), and primary NK cells (c), respectively. (d) Inhibition of ERK and STAT3 decreases the expression of activating receptors on the NK cell surface. Representative flow cytometry results show the effects of STAT3 inhibitor cryptotanshinone (CPT) (upper panel) and ERK inhibitor U0126 (lower panel) on the expression of activating receptors on the NK cell surface. Left panel: one of three representative flow cytometry results; Right panel: statistical analysis of the flow cytometry data (n=3, * P<0.05, ** P<0.01). KHYG-1 and NK92 cells were treated with vehicle, 10 μM ERK inhibitor U0126, and 10 μM STAT3 inhibitor CPT for 24 h. (e) Inhibition of ERK and STAT3 decreases NK cell cytotoxicity. Statistical analysis showing the effects of ERK and STAT3 inhibition on NK cells cytotoxicity against K562 cells (n=3, * P<0.05, ** P<0.01). KHYG-1 cells were pretreated with 10 μM U0126 and 10 μM CPT for 6 h and then incubated with K562 at different E:T ratios for 4 h.

[0083] FIG. 4

[0084] Hypoxia activates SHP-1 and SHP-2 in NK cells. (a-c) Western blotting analysis shows SHP-1 and SHP-2 expression in normoxic (20% 02) and hypoxic (1% 02) KHYG-1 (a), NK92 (b) cells, and primary NK cells (c), respectively. (d) Western blotting analysis shows the effects of SHP-1 inhibitor TPI-1 on the phosphorylation of ERK and STAT3. Hypoxic KYHG-1 cells were pre-treated with 5 μM TPI-1 for 2 h, and then the cells were collected for western blotting analysis. (e) Flow cytometric analysis of the effects of TPI-1 on the NK cell cytotoxicity. Left panel: representative flow cytometry results of TPI-1 on the cytotoxicity of KHYG-1 cells. KHYG-1 cells were pretreated with 5 μM TPI-1 for 2 h, and then incubated with K562 cells at different E:T ratios for 4 h. Right panel: statistical analysis of the effects of TPI-1 on KHYG-1 cell cytotoxicity against K562 cells (n=3, * P<0.05).

[0085] FIG. 5

[0086] Inhibition of SHP-2 has no effect on NK cell cytotoxicity. (a) Western blotting shows the effects of SHP-2 inhibitor SHP099 on the phosphorylation of ERK and STAT3. Hypoxic KYHG-1 cells were pre-treated with 5 μM SHP099 for 2 h, and then the cells were collected for western blotting analysis. (b) Flow cytometric analysis shows the effect of SHP099 on the cytotoxicity of KHYG-1 cells. KHYG-1 cells were pretreated with 5 μM SHP099 for 2 h, then incubated with K562 cells at different E:T ratios for 4 h. (c) Statistical analysis of the effects of SHP099 on KHYG-1 cell cytotoxicity against K562 cells (n=3, ns=non-significant).

[0087] FIG. 6

[0088] The effects of gene silencing SHP-1 on ERK and STAT3 signaling as well as NK cell cytotoxicity. (a) Western blotting analysis of the SHP-1, ERK, and STAT3 expressions in siRNA-mediated knockdown of SHP-1 in KHYG-1 cells. (b) Statistical analysis of the effects of knocking down SHP-1 on NK cells cytotoxicity against K562 cells (n=3, * P<0.05). KHYG-1 cells were electroporated with 2 μg siRNA, and then cultured for 12-16 h in the RPMI 1640 growth medium containing IL-2. The electroporated cells were used for western blotting or killing assay as previously mentioned. (c) A schematic diagram shows how hypoxia impairs NK cell cytotoxicity in a SHP-1 dependent manner.

Examples 1-3: Materials and Methods

[0089] Antibodies and Reagents

[0090] Antibodies for Western blotting against phospho-Stat3 (#4113), Stat3 (#12640), Phospho-p44/42 MAPK (ERK1/2) (#9106), p44/p42 MAPK (ERK1/2) (#9102), Phospho-SHP-1 (#8849), Phospho-SHP-2 (#5431), SHP-1 (#3759), SHP-2 (#3397), HIF-1α (#14179) and β-actin (#58169) were bought from Cell Signaling. Peroxidase-conjugated goat anti-rabbit IgG (#111-035-003) or goat anti-mouse IgG (#115-005-003) were bought from Jackson ImmunoResearch. For flow cytometry analysis, Alexa Fluor 647-labeled anti-human perforin (#563576) was purchased from BD Biosciences. FITC-labeled Annexin V (#640945), PE-labeled anti-human IFN-γ (#506506), anti-human/mouse granzyme B (#372207), APC-labeled anti-human NKp46 (#137607), anti-human NKp30 (#325209), anti-human NKG2D (#320808), anti-human CD2 (#300214), anti-human CD107a (#12-1079-42) antibodies, were purchased from Biolegend. CD3-labeled CD3 (#560365), PE-labeled CD56 (#555516) were bought from BD Biosciences. Sytox® Green Dead Cell Stain was bought from Molecular Probes (#S34860). STAT3 inhibitor Cryptotanshinone was bought from Selleck Chemicals (#35825-57-1). SHP-1 inhibitor TPI-1 (#HY-100463), SHP-2 inhibitor SHP-099 (#HY-100388) and ERK inhibitor U0126 (#HY-12031) were bought from MedChem Express.

[0091] Cell Lines and Cell Culture

[0092] The NK cell line KHYG-1 was cultured in RPMI-1640 (BasalMedium, #L210KJ) supplemented with 10% fetal bovine serum (FBS) (Biological Industries, #04-001-1ACS), 10 ng/mL human IL-2 (PeproTech, #200-02), 2 mM L-glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin. The NK cell line NK92 was cultured in Dulbecco's modified Eagle's medium (DMEM) (BasalMedium, #L110KJ) supplemented with 10% FBS, 10% horse serum, 10 ng/mL IL-2, 2 mM glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin. The human multiple myeloma cell line MM.1S and leukemia cell line K562 were grown in RPMI-1640 supplemented with 10% FBS, 2 mM glutamine, 100U/mL penicillin, and 100 μg/mL streptomycin. All cells were maintained at 37° C. in a humidified atmosphere containing 5% CO.sub.2. Normoxic or hypoxic cell culture conditions were obtained by culturing cells in a sealed incubator flushed with a mixture of 20% O.sub.2, 5% CO.sub.2, and 75% N.sub.2, or the mixture of 1% O.sub.2, 5% CO.sub.2, and 94% N.sub.2, respectively.

[0093] Human Primary NK Cell Enrichment and Activation

[0094] Primary NK cells were isolated from peripheral blood mononuclear cells of healthy human donors through an immunomagnetic negative selection strategy (EasySep Human NK cell Isolation Kit, Stemcell Technologies) according to the manufacturer's protocol. Purity of the purified NK cell populations was determined by flow cytometry using fluorochrome-conjugated antibodies against CD3, CD56. For short-term activation, purified NK cells (>90% pure) were resuspended in RPMI-1640 supplemented with 10% FBS, 5% human serum, 2 mM glutamine, 100U/mL penicillin and 100 μg/mL streptomycin at a density of 3×10.sup.6 cells/ml and cultured overnight in the presence of IL-2 (10 ng/ml) under normoxic or hypoxic conditions as described above.

[0095] Flow Cytometry

[0096] The expression of NK cell cytotoxicity effector molecules and activating receptors was analyzed by flow cytometry. For membrane staining, 5×10.sup.5 cells were collected and washed with staining buffer (PBS containing 0.1% NaN.sub.3 and 0.1% BSA) three times. The cells were then incubated for 30 min on ice, according to the instructions provided with the respective antibodies. After washing 3 times with cell stain buffer, the cells were resuspended in 300 μL staining buffer in the presence of Sytox Green or 7-AAD, which were used to gate out dead cells. Acquisition of 10,000 cells per reaction was performed using a CytoFLEX Cytometer (Beckman Coulter Life Sciences). Data were analyzed with Flowjo v7.6.2 (Tree Star). For intracellular staining, 5×10.sup.5 cells were collected and fixed with 1 mL 1% paraformaldehyde in PBS for 15 minutes at room temperature. After washing 3 times with cell stain buffer, the fixed cells were then resuspended in 2 mL permeabilization buffer (0.1% saponin in cell staining buffer) and incubated for 30 min at room temperature. The cells were collected again by centrifugation, stained with the antibody at an optimal working concentration in permeabilization buffer for 15 min on ice. After washing three times with permeabilization buffer, the cells were resuspended cells in 300 μL cell staining buffer for final flow cytometric analysis.

[0097] CD107a Degranulation Assay

[0098] Degranulation of cytotoxic contents from NK cells was measured by analysis of the degranulation marker CD107a by flow cytometry. Briefly, NK cells and tumour cells were individually pre-incubated for 14-16 h at 20% or 1% 02 and after that combined at 1:1 (E:T) ratio at either 20% or 1% 02 in 24-well plate. 5 μL of APC labeled anti-CD107a was added to the wells within 5-10 minutes after combining NK and tumour cells. Subsequently, Monensin and GolgiPlug (1:1000 dilution; BD Biosciences) were added. After a total incubation time of 4 h, the plate was placed on ice to stop the reaction. Cells were then harvested and analyzed using flow cytometry.

[0099] Flow Cytometric Cytotoxicity Assay

[0100] Prior to the assay, NK cells and tumour cells were individually pre-incubated for 24 h at 5% CO.sub.2 with 20% or 1% 02 first. NK and target cells were then incubated under comparable conditions in different E:T ratios in a 24 well plate. After the 4 h incubation, samples were harvested and washed followed by a combinational staining with CD2-APC, Annexin V-FITC as well as Sytox Green, in which CD2 was used to distinguish effector from target cells, and target cell death was detected with Annexin V-FITC and Sytox Green. A minimum of 10,000 target events were collected per sample, and the results were analyzed using Flowjo v7.6.2.

[0101] Western Blotting

[0102] For western blotting, treated and untreated KHYG-1 and NK92 cells were lysed in buffer containing 50 mM Tris, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS and protease inhibitors on ice for 30 min. Lysates were centrifuged at 12,000 rpm for 15 min, and supernatants were collected. Protein concentration was determined by the BCA protein assay kit (HEART Biotech, #WB003). Equal amounts of protein were loaded and separated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel, and transferred onto a PVDF membrane (Millipore, #IPVH00010). After blocking for 1 h with 5% non-fat milk in PBS with 0.1% Tween-20 at room temperature, the membrane was incubated with primary antibody at 4° C. overnight. Immunoblots were visualized using HRP-conjugated secondary antibodies and the ECL Western Blot Detection kit (Phygene Life Sciences, #PH0353).

[0103] siRNA-Mediated Gene Silencing in NK Cells

[0104] Prior to siRNA transfection, KHYG-1 cells were washed in pre-warmed Opti-MEM medium (Life Technologies, Carlsbad, Calif., USA) and resuspended in the same medium. Then, 10.sup.6 cells were electroporated with 2 μg of siRNA in 100 μL Opti-MEM medium in 0.2 cm cuvettes with an electroporator CUY21EDIT (BEX Co. Ltd, Japan). The electroporation program was set as follows: PpV=200V, Pp on 10 ms, Pp off 10 ms, PdV=25V, Pd on 50 ms, Pd off 50 ms; Pd N=10, capacity=940 μF, and exponential decay wave type. Following electroporation, cells were resuspended in 2 mL complete media and cultured in hypoxic condition (1% O.sub.2). 16-24 h after electroporation, the cells were used for western blotting or killing assay. Transfection efficiency and viability were analyzed by flow cytometry 2-6 h after electroporation by quantitatively measuring the expression of fluorescein isothiocyanate (FITC)-labeled siRNA or 7-AAD. SHP-1 mRNA was silenced by using a gene-specific siRNA pool from GenePharma.

[0105] Statistical Analysis

[0106] Statistical analyses were performed using the Prism software package 5.0 (GraphPad Software, San Diego, Calif., USA). Data are expressed as the mean±SEM of at least three independent experiments. Statistical significance was evaluated by two-tailed paired Student's t-test. P<0.05 (*), P<0.01(**), or P<0.001(***) was considered statistically significant.

Example 1—Reduced Cytotoxicity of Hypoxic NK Cells

[0107] The effect of hypoxia on NK cell-mediated lysis of tumour cells was investigated. KHYG-1 cells were cultured in the presence of IL-2 under hypoxic (1% O.sub.2) or normoxic (20% O.sub.2) conditions for 24 hours, and subsequently the cells were incubated with cancer cell lines K562 or MM.1S at different effector:target (E:T) ratios for 4 hours, in order to evaluate cytotoxicity by flow cytometry. As shown in FIG. 1a and FIG. 1b, NK cell cytotoxicity was significantly decreased in hypoxic conditions.

[0108] Western blotting revealed a significant accumulation of the hypoxia marker HIF-1α in hypoxic NK cells (KHYG-1 and NK-92), whereas HIF-1a was only weakly expressed in normoxic NK cells—see FIG. 1c.

[0109] The possibility that the decreased cytotoxicity of hypoxic NK cells was caused by reduced NK cell viability was eliminated, as there was no significant difference in NK cell death between the hypoxic and normoxic samples (FIG. 1d).

[0110] To further investigate how hypoxia reduces NK cell cytotoxicity, the expression levels of the cytotoxic effectors granzyme B and perforin were measured. As shown in FIG. 2a, hypoxia led to decreased secretion of both granzyme B and perforin. Additionally, a reduced expression of the cytokine IFN-γ was observed in hypoxic NK cells, when compared to normoxic NK cells (FIG. 2b). The degranulation marker CD107a was also diminished by hypoxia (FIG. 2c).

[0111] Surface expression of the activating receptors NKp46, NKp30 and NKG2D was measured by flow cytometry on both normoxic NK cells and hypoxic NK cells. As shown in FIG. 2d, hypoxic conditions decreased the expression of these activating receptors on the NK cell surface.

Example 2—Hypoxic Attenuation of ERK and STAT3-Mediated NK Cell Activation

[0112] It is known that intracellular signals activating NK cell cytotoxicity are propagated primarily through protein phosphorylation of extracellular signal-regulated kinase (ERK) and signal transducer and activator of transcription 3 (STAT3). The role of hypoxia in the activation of ERK and STAT3 was therefore investigated.

[0113] It was revealed that hypoxia markedly decreased the phosphorylation level at the tyrosine sites of ERK and STAT3 in KHYG-1 cells, NK-92 cells and primary NK cells (FIGS. 3a, 3b and 3c).

[0114] To further validate the effects of ERK and STAT3 phophorylation on NK cell cytotoxicity under hypoxic conditions, specific small molecule inhibitors U0126 and cryptotanshinone were used to block ERK and STAT3 signaling, respectively. As shown in FIG. 3d, inhibition of ERK and STAT3 significantly reduced the expression of activating receptors NKp30, NKp46 and NKG2D (KHYG-1) and NKp46 and NKG2D (NK-92). Moreover, inhibition of ERK or STAT3 resulted in significantly impaired NK cell cytotoxicity against cancer cells (FIG. 3e).

[0115] Cell surface receptors harboring intracytoplasmic tyrosine-based activation motifs (ITAMs) or intracytoplasmic tyrosine-based inhibitory motifs (ITIMs) are often phosphorylated by Src family protein tyrosine kinase (PTK), which in turn creates docking sites for the protein tyrosine phosphatases SHP-1 and SHP-2. The involvement of SHP-1 and SHP-2 in the decrease of ERK and STAT3 phosphorylation by hypoxia was therefore investigated. As shown in FIGS. 4a, 4b and 4c, hypoxia induced a significant increase in the phosphorylation of SHP-1 and SHP-2 in the two NK cell lines and in primary NK cells. When using a specific SHP-1 inhibitor, TPI-1, it was observed that TPI-1 could reverse the decrease of the phosphorylation of both ERK and STAT3 (FIG. 4d). Moreover, it was also observed that pretreatment with the p-SHP1 inhibitor TPI-1 could restore NK cell cytotoxicity under hypoxia (FIG. 4e). The same effects were not observed when using the specific SHP-2 inhibitor SHP099, which had no effect on the phosphorylation levels of ERK and STAT3, or NK cell cytotoxicity (FIGS. 5a, 5b and 5c).

[0116] It was thus concluded that a hypoxia-induced decrease in the phosphorylation level of STAT3 and ERK was mediated by the activation of the protein tyrosine phosphatase SHP-1, as opposed to SHP-2

Example 3—Knockdown of SHP-1 Restores NK Cell Cytotoxicity in Hypoxia

[0117] To further validate the role of SHP-1 in regulating NK cell cytotoxicity, SHP-1 gene expression was silenced in KHYG-1 cells. It was confirmed that knockdown of SHP-1 could increase the phosphorylation level of ERK and STAT3 under hypoxia (FIG. 6a). Moreover, it was confirmed that NK cells with SHP-1 silenced showed greater cytotoxicity against K562 cells than control NK cells under hypoxic conditions (FIG. 6b).

Example 4—Knock-In of CD19 CARs in Primary NK Cells

[0118] The anti-CD19-ζ, anti-CD19-BB-ζ, and anti-CD19-truncated (control) plasmids used have been described previously (Imai et al. 2004. Leukemia. 18(4):676-84). The cDNA encoding the intracellular domains of human DAP10 and 4-1BB ligand (4-1BBL), and interleukin-15 (IL-15) with long signal peptide were sub-cloned by polymerase chain reaction (PCR) with a human spleen cDNA library used as a template. An anti-CD19-DAP10 plasmid was constructed by replacing the sequence encoding CD3 with that encoding DAP10, using the splicing by overlapping extension by PCR (SOE-PCR) method. The cDNA encoding the signal peptide of CD8α, the mature peptide of IL-15 and the transmembrane domain of CD8α were assembled by SOE-PCR to encode a “membrane-bound” form of IL-15. The resulting expression cassettes were sub-cloned into EcoRI and Xhol sites of murine stem-cell virus-internal ribosome entry site-green fluorescent protein (MSCV-IRES-GFP).

[0119] The RD114-pseudotyped retrovirus was generated as described previously (Imai et al. 2004. Leukemia. 18(4):676-84). A calcium phosphate DNA precipitation was used to transfect 293T cells with anti-CD19-ζ, anti-CD19-DAP10, anti-CD19-BB-ζ, or anti-CD19-truncated; pEQ-PAM3(-E); and pRDF. Conditioned medium containing retrovirus was harvested at 48 hours and 72 hours after transfection, immediately frozen in dry ice, and stored at −80° C. until use.

[0120] K562 cells were transduced with the construct encoding the “membrane-bound” form of IL-15. Cells were cloned by limiting dilution, and a single-cell clone with high expression of GFP and surface IL-15 (K562-mb15) was expanded. This clone was subsequently transduced with human 4-1BBL (K562-mb15-41BBL). K562 cells expressing wildtype IL-15 (K562-wt15) or 4-1BBL (K562-41BBL) were produced by a similar procedure. Peripheral blood mononuclear cells (1.5×10.sup.6) were incubated in a 24-well tissue-culture plate with or without 10.sup.6 K562-derivative stimulator cells in the presence of 10 IU/mL human IL-2 in RPMI 1640 and 10% FCS.

[0121] Mononuclear cells stimulated with K562-mb15-41BBL were transduced with retroviruses, as described previously (Imai et al. 2004. Leukemia. 18(4):676-84). Briefly, 14 mL polypropylene centrifuge tubes were coated with human fibronectin (100 μg/mL) or RetroNectin (50 μg/mL). Pre-stimulated cells (2×10.sup.5) were re-suspended in the tubes in 2-3 mL virus-conditioned medium with Polybrene (4 μg/mL) and centrifuged at 2400 g for 2 hours (centrifugation was omitted when RetroNectin was used). The multiplicity of infection (4-6) was identical in each experiment comparing the activity of different CARs. After centrifugation, cells were left undisturbed for 24 hours in a humidified incubator at 37° C., 5% CO.sub.2. The transduction procedure was repeated on 2 successive days. After a second transduction, the cells were re-stimulated with K562-mb15-41BBL in the presence of 10 IU/mL IL-2. Cells were maintained in RPMI 1640, 10% FCS and 10 IU/mL IL-2.

[0122] Transduced NK cells were stained with goat anti-mouse (Fab).sub.2 polyclonal antibody conjugated with biotin followed by streptavidin conjugated to peridinin chlorophyll protein. For Western blotting, cells were lysed in RIPA buffer (PBS, 1% Triton-X100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS]) containing 3 μg/mL pepstatin, 3 μg/mL leupeptin, 1 mM phenylmethylsulfonyl fluoride (PMSF), 2 mM ethylenediaminetetraacetic acid (EDTA) and 5 μg/mL aprotinin. Centrifuged lysate supernatants were boiled with an equal volume of loading buffer, with or without 0.1 M dithiothreitol (DTT), and then separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) on a precast 10-20% gradient acrylamide gel. The proteins were transferred to a polyvinylidene fluoride (PVDF) membrane which was incubated with primary mouse anti-human CD3 monoclonal antibody (clone 8D3). Membranes were then washed, incubated with a goat anti-mouse IgG horseradish peroxidase-conjugated second antibody and developed using an enhanced chemiluminescence system.

[0123] The following antibodies were used for immunophenotypic characterization of expanded and transduced cells: anti-CD3 conjugated to fluorescein isothiocyanate (FITC), to PerCP or to energy-coupled dye (ECD); anti-CD10 conjugated to phycoerythrin (PE); anti-CD19 PE; anti-CD22 PE; anti-CD56 FITC, PE or allophycocyanin (APC); anti-CD16 CyChrome; and anti-CD25 PE. Surface expression of KIR and NK activation molecules was determined with specific antibodies conjugated to FITC or PE, as described previously (Leung et al. 2004. Journal of Immunology. 172:644-650). Antibody staining was detected with a FACScan or an LSR II flow cytometer.

Example 5—Knock-In of TRAIL/TRAIL Variant in NK Cells

[0124] KHYG-1 cells were transfected with both TRAIL and TRAIL variant, in order to assess their viability and ability to kill cancer cells following transfection.

[0125] The TRAIL variant used is that described in WO 2009/077857. It is encoded by the wildtype TRAIL gene containing the D269H/E195R mutation. This mutation significantly increases the affinity of the TRAIL variant for DR5, whilst reducing the affinity for both decoy receptors (DcR1 and DcR2).

[0126] Baseline TRAIL Expression

[0127] Baseline TRAIL (CD253) expression in KHYG-1 cells was assayed using flow cytometry.

[0128] Mouse anti-human CD253-APC (Biolegend catalog number: 308210) and isotype control (Biolegend catalog number: 400122) were used to stain cell samples and were analyzed on a BD FACS Canto II flow cytometer.

[0129] KHYG-1 cells were cultured in RPMI 1640 medium containing 10% FBS, 2 mM L-glutamine, penicillin (100 U/mL)/streptomycin (100 mg/mL) and IL-2 (10 ng/mL). 0.5-1.0×10.sup.6 cells/test were collected by centrifugation (1500 rpm×5 minutes) and the supernatant was aspirated. The cells (single cell suspension) were washed with 4 mL ice cold FACS Buffer (PBS, 0.5-1% BSA, 0.1% NaN.sub.3 sodium azide). The cells were re-suspended in 100 μL ice cold FACS Buffer, add 5 uL antibody was added to each tube and incubated for 30 minutes on ice. The cells were washed 3 times by centrifugation at 1500 rpm for 5 minutes. The cells were then re-suspended in 500 μL ice cold FACS Buffer and temporarily kept in the dark on ice.

[0130] The cells were subsequently analyzed on the flow cytometer (BD FACS Canto II) and the generated data were processed using FlowJo 7.6.2 software.

[0131] As can be seen in FIG. 7, FACS analysis showed weak baseline expression of TRAIL on the KHYG-1 cell surface.

[0132] TRAIL/TRAIL Variant Knock-In by Electroporation

[0133] Wildtype TRAIL mRNA and TRAIL variant (D269H/195R) mRNA was synthesized by TriLink BioTechnologies, aliquoted and stored as −80° C. Mouse anti-human CD253-APC (Biolegend catalog number: 308210) and isotype control (Biolegend catalog number: 400122), and Mouse anti-human CD107a-PE (eBioscience catalog number: 12-1079-42) and isotype control (eBioscience catalog number: 12-4714) antibodies were used to stain cell samples and were analyzed on a BD FACS Canto II flow cytometer. DNA dye SYTOX-Green (Life Technologies catalog number: S7020; 5 mM Solution in DMSO) was used. To achieve transfection efficiencies of up to 90% with high cell viability in KHYG-1 cells with the Nucleofector™ Device (Nucleofector II, Lonza), a Nucleofector™ Kit T from Lonza was used. Antibiotics-free RPMI 1640 containing 10% FBS, L-glutamine (2 mM) and IL-2 (10 ng/mL) was used for post-Nucleofection culture.

[0134] KHYG-1 and NK-92 cells were passaged one or two days before Nucleofection, as the cells must be in the logarithmic growth phase. The Nucleofector solution was pre-warmed to room temperature (100 μl per sample), along with an aliquot of culture medium containing serum and supplements at 37° C. in a 50 mL tube. 6-well plates were prepared by filling with 1.5 mL culture medium containing serum and supplements and pre-incubated in a humidified 37° C./5% CO.sub.2 incubator. An aliquot of cell culture was prepared and the cells counted to determine the cell density. The required number of cells was centrifuged at 1500 rpm for 5 min, before discarding the supernatant completely. The cell pellet was re-suspended in room temperature Nucleofector Solution to a final concentration of 2×10.sup.6 cells/100p1 (maximum time in suspension=20 minutes). 100 μl cell suspension was mixed with 10 μg mRNA (volume of RNA <10 μL). The sample was transferred into an Amaxa-certified cuvette (making sure the sample covered the bottom of the cuvette and avoiding air bubbles). The appropriate Nucleofector program was selected (i.e. U-001 for KHYG-1 cells). The cuvettes were then inserted into the cuvette holder. 500 μl pre-warmed culture medium was added to the cuvette and the sample transferred into a prepared 6-well plate immediately after the program had finished, in order to avoid damage to the cells. The cells were incubated in a humidified 37° C./5% CO.sub.2 incubator. Flow cytometric analysis and cytotoxicity assays were performed 12-16 hours after electroporation. Flow cytometry staining was carried out as above.

[0135] As can be seen in FIGS. 8 and 9, expression of TRAIL/TRAIL variant and CD107a (NK activation marker) increased post-transfection, confirming the successful knock-in of the TRAIL genes into KHYG-1 cells.

[0136] FIG. 10 provides evidence of KHYG-1 cell viability before and after transfection via electroporation. It can be seen that no statistically significant differences in cell viability are observed following transfection of the cells with TRAIL/TRAIL variant, confirming that the expression of wildtype or variant TRAIL is not toxic to the cells. This observation contradicts corresponding findings in NK-92 cells, which suggest the TRAIL variant gene knock-in is toxic to the cells (data not shown). Nevertheless, this is likely explained by the relatively high expression levels of TRAIL receptors DR4 and DR5 on the NK-92 cell surface (see FIG. 11).

[0137] Effects of TRAIL/TRAIL Variant on KHYG-1 Cell Cytotoxicity

[0138] Mouse anti-human CD2-APC antibody (BD Pharmingen catalog number: 560642) was used. Annexin V-FITC antibody (ImmunoTools catalog number: 31490013) was used. DNA dye SYTOX-Green (Life Technologies catalog number: S7020) was used. A 24-well cell culture plate (SARSTEDT AG catalog number: 83.3922) was used. Myelogenous leukemia cell line K562, multiple myeloma cell line RPMI8226 and MM1.S were used as target cells. K562, RPMI8226, MM1.S were cultured in RPMI 1640 medium containing 10% FBS, 2 mM L-glutamine and penicillin (100 U/mL)/streptomycin (100 mg/mL).

[0139] As explained above, KHYG-1 cells were transfected with TRAIL/TRAIL variant.

[0140] The target cells were washed and pelleted via centrifugation at 1500 rpm for 5 minutes. Transfected KHYG-1 cells were diluted to 0.5×10.sup.6/mL. The target cell density was then adjusted in pre-warmed RPMI 1640 medium, in order to produce effector:target (E:T) ratios of 1:1.

[0141] 0.5 mL KHYG-1 cells and 0.5 mL target cells were then mixed in a 24-well culture plate and placed in a humidified 37° C./5% CO.sub.2 incubator for 12 hours. Flow cytometric analysis was then used to assay KHYG-1 cell cytotoxicity; co-cultured cells (at different time points) were washed and then stained with CD2-APC antibody (5 μL/test), Annexin V-FITC (5 μL/test) and SYTOX-Green (5 μL/test) using Annexin V binding buffer.

[0142] Data were further analyzed using FlowJo 7.6.2 software. CD2-positive and CD2-negative gates were set, which represent KHYG-1 cell and target cell populations, respectively. The Annexin V-FITC and SYTOX-Green positive cells in the CD2-negative population were then analyzed for TRAIL-induced apoptosis.

[0143] FIGS. 12, 13 and 14 show the effects of both KHYG-1 cells expressing TRAIL or TRAIL variant on apoptosis for the three target cell lines: K562, RPMI8226 and MM1.S, respectively. It is apparent for all target cell populations that TRAIL expression on KHYG-1 cells increased the level of apoptosis, when compared to normal KHYG-1 cells (not transfected with TRAIL). Moreover, TRAIL variant expression on KHYG-1 cells further increased apoptosis in all target cell lines, when compared to KHYG-1 cells transfected with wildtype TRAIL.

Example 6—Knock-In of CD19 CARs and TRAIL Variants in Primary NK Cells

[0144] Anti-CD19-CD28(TM)-CD3ζ, anti-CD19-41BB(TM)-CD3ζ, and anti-CD19-truncated (control) plasmids were used. The cDNA encoding the CD19 scFv, with transmembrane domains of human CD3, CD28 or 4-1 BB ligand (4-1 BBL), and with intracellular domains of CD3 were used as a template mRNA. The gene cassette containing the combination of CD19 CAR and TRAIL variant was synthesized as mRNA. CD19 CAR and high affinity TRAIL DR5 variant was delivered to the NK cells as two separate in vitro synthesized mRNAs at the same time.

[0145] The TRAIL variant used is that described in WO 2009/077857. It is encoded by the wildtype TRAIL gene containing the D269H/E195R mutation. This mutation significantly increases the affinity of the TRAIL variant for DR5, whilst reducing the affinity for both decoy receptors (DcR1 and DcR2).

[0146] Electroporated NK cells were stained with goat anti-mouse (Fab).sub.2 polyclonal antibody conjugated with biotin followed by streptavidin conjugated to PE or FITC flurophore. For Western blotting, cells were lysed in RIPA buffer (PBS, 1% Triton-X100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS]) containing 3 μg/mL pepstatin, 3 μg/mL leupeptin, 1 mM phenylmethylsulfonyl fluoride (PMSF), 2 mM ethylenediaminetetraacetic acid (EDTA) and 5 μg/mL aprotinin. Centrifuged lysate supernatants were boiled with an equal volume of loading buffer, with or without 0.1 M dithiothreitol (DTT), and then separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) on a precast 10-20% gradient acrylamide gel. The proteins were transferred to a polyvinylidene fluoride (PVDF) membrane which was incubated with primary mouse anti-human CD3 monoclonal antibody (clone 8D3). Membranes were then washed, incubated with a goat anti-mouse IgG horseradish peroxidase-conjugated second antibody and developed using an enhanced chemiluminescence system.

[0147] The following antibodies were used for immunophenotypic characterization of expanded and transduced cells: anti-CD3 conjugated to fluorescein isothiocyanate (FITC), to PerCP or to energy-coupled dye (ECD); anti-CD10 conjugated to phycoerythrin (PE); anti-CD19 PE; anti-CD22 PE; anti-CD56 FITC, PE or allophycocyanin (APC); anti-CD16 CyChrome; and anti-CD25 PE. Surface expression of KIR and NK activation molecules was determined with specific antibodies conjugated to FITC or PE, as described previously (Leung et al. 2004. Journal of Immunology. 172:644-650). Antibody staining was detected with a FACS Canto II flow cytometer.

[0148] Baseline TRAIL (CD253) expression in naive or expanded NK cells was assayed using flow cytometry.

[0149] Mouse anti-human CD253-APC (Biolegend catalog number: 308210) and isotype control (Biolegend catalog number: 400122) were used to stain cell samples and were analyzed on a BD FACS Canto II flow cytometer.

[0150] Primary NK cells were cultured in Miltenyi's NK cell expansion medium containing 10% human AB serum, penicillin (100 U/mL)/streptomycin (100 mg/mL) and IL-2 (500U/ml). 0.5-1.0×10.sup.6 cells/test were collected by centrifugation (1500 rpm×5 minutes) and the supernatant was aspirated. The cells (single cell suspension) were washed with 4 mL ice cold FACS Buffer (PBS, 0.5-1% BSA, 0.1% NaN.sub.3 sodium azide). The cells were re-suspended in 100 μL ice cold FACS Buffer and 5 μL antibody was added to each tube and incubated for 30 minutes on ice. The cells were washed 3 times by centrifugation at 1500 rpm for 5 minutes. The cells were then re-suspended in 500 μL ice cold FACS Buffer and temporarily kept in the dark on ice.

[0151] The cells were subsequently analyzed by flow cytometer (BD FACS Canto II) and the generated data were processed using FlowJo 7.6.2 software.

[0152] FACS analysis showed weak baseline expression of TRAIL on the NK cell surface.

[0153] TRAIL variant (D269H/195R) mRNA was synthesized by TriLink BioTechnologies, aliquoted and stored as −80° C. Mouse anti-human CD253-APC (Biolegend catalog number: 308210) and isotype control (Biolegend catalog number: 400122), antibodies were used to stain cell samples and were analyzed on a BD FACS Canto II flow cytometer. Propidium Iodide was used for cell viability. In order to achieve transfection efficiencies of up to 90% with high NK cell viability, an electroporation based technique was implemented using Maxcyte GT. Cells were processed in Maxcyte buffer prior to electroporation and cells were electroporated with 5-10 μg/ml of each individual mRNA (i.e. 5 μg/ml of CD19 CAR and 5 μg/ml TRAIL variant)

[0154] Naive or expanded NK cells were passaged one or two days before electroporation, as the cells must be in the logarithmic growth phase. 6-well plates were prepared by filling with 1.5 mL culture medium containing serum and supplements and pre-incubated in a humidified 37° C./5% CO.sub.2 incubator. An aliquot of cell culture was prepared and the cells counted to determine the cell density. The required number of cells was centrifuged at 1500 rpm for 5 min, before discarding the supernatant completely. The cell pellet was re-suspended in room temperature Maxcyte buffer to a final concentration of 2×10.sup.6 cells/100p1 (maximum time in suspension=20 minutes). 100 μl cell suspension was mixed with 5 μg mRNA. The samples were transferred into Maxcyte-certified cuvettes OC-100×2 (making sure the samples covered the bottom of the cuvettes and avoiding air bubbles). The appropriate electroporation program was selected (i.e. NK-4). The cuvettes were then inserted into the cuvette holder. Immediately after the program had finished, in order to avoid damage to the cells, the cells were transferred to 6-well plates and incubated for 20 mins at 37° C. Flow cytometric analysis and cytotoxicity assays were performed 20-24 hours after electroporation. Flow cytometry staining was carried out as above.

[0155] Expression of CD19 CAR and TRAIL variant was shown to increase post-transfection, confirming the successful knock-in of the CD19 CAR and TRAIL variant genes into primary NK cells.

[0156] There was evidence of NK cell viability before and after transfection via electroporation. It can be seen that no statistically significant differences in cell viability are observed following transfection of the cells with TRAIL variant, confirming that the expression of variant TRAIL is not toxic to the cells.

[0157] The effects of the TRAIL variant on NK cell cytotoxicity were also measured. Mouse anti-human CD2-APC antibody (BD Pharmingen catalog number: 560642) was used. Annexin V-FITC antibody (ImmunoTools catalog number: 31490013) was used. A 96-well cell culture plate (SARSTEDT AG catalog number: 83.3922) was used. B-cell lymphoma cell lines OCI-LY10, RIVA, and SU-DHL6 were cultured in RPMI 1640 medium containing 10% FBS, 2 mM L-glutamine and penicillin (100 U/mL)/streptomycin (100 mg/mL).

[0158] The target cells were washed and pelleted via centrifugation at 1500 rpm for 5 minutes. Transfected NK cells were diluted to achieve a concentration of 2×10.sup.6/mL cells. The target cell density was then adjusted in pre-warmed RPMI 1640 medium, in order to produce effectortarget (E:T) ratios of 5:1, 2.5, and 1.25:1.

[0159] 0.1 mL electroporated NK cells and 0.1 mL target cells were then mixed in a 96-well culture plate and placed in a humidified 37° C./5% CO.sub.2 incubator for 16 hours. Flow cytometric analysis was then used to assay NK cell cytotoxicity; co-cultured cells (at different time points) were washed and then stained with CD2-APC antibody (5 μL/test), and cell viability was assessed using propidium iodide.

[0160] Data were further analyzed using FlowJo 7.6.2 software. CD2-positive and CD2-negative gates were set, which represented NK cell and target cell populations, respectively. The PI cells in the CD2-negative population were then analyzed for TRAIL-induced apoptosis.

[0161] The effects of NK cells expressing TRAIL variant on apoptosis were assessed for the three target cell lines: OCI-LY10, RIVA, and SU-DHL6, respectively. It is apparent for all target cell populations that TRAIL variant expression on NK cells increased the level of apoptosis, when compared to normal NK cells (not transfected with TRAIL variant).

[0162] Cells of the invention, expressing both the CD19 CAR and the TRAIL variant, offer a significant advantage in cancer therapy, due to their ability to specifically target cancer cells with high affinity and then kill those cells via the death receptor DR5. When challenged by the cells of the invention, cancer cells are prevented from developing defensive strategies to circumvent death via a single pathway. For example, cancers cannot effectively circumvent TRAIL-induced cell death by upregulating TRAIL decoy receptors, as cells of the invention are modified so that they remain cytotoxic in those circumstances.

Example 7—Knockout of NK Cell TRAIL Receptors DR4 and DR5

[0163] NK cells are prepared as follows, having death receptor 5 (DR5) and/or death receptor 4 (DR4) function removed.

[0164] gRNA constructs targeting TRAIL-R2 (DR5) and TRAIL-R1 (DR4) are designed (e.g. [0165] SEQ ID NO:1: CCCAUCUUGAACAUACCAG (DR5), [0166] SEQ ID NO:2: AACCGGUGCACAGAGGGUGU (DR4) and [0167] SEQ ID NO:3: AUUUACAAGCUGUACAUGGG (DR4))
and prepared to target endogenous genes encoding DR5 and DR4 gene(s) in NK cells. CRISPR/Cas9 genome editing is then used to knock out the DR5 and/or DR4 target genes.

[0168] A total of 3 gRNA candidates are selected for the DR5 gene and their cleavage efficacies in RPMI8226 cells determined. A total of 3 gRNA candidates are selected for the DR4 gene and their cleavage efficacies in HL60 cells determined. RPMI8226 cells and HL60 are electroporated with the gRNA:Cas9 ribonucleoprotein (RNP) complex using Maxcyte® GT and subsequently knockout of DR5 and/or is analyzed by flowcytometry. The cleavage activity of the gRNA is also determined using an in vitro mismatch detection assay. T7E1 endonuclease recognizes and cleaves non-perfectly matched DNA, allowing the parental DR5 gene/DR4 gene to be compared to the mutated gene following CRISPR/Cas9 transfection and non-homologous end joining (NHEJ).

[0169] The gRNA with highest KO efficiency is selected to further experiments to knockout DR5/DR4 in primary NK cells, NK cell lines or CD34+ progenitors (for subsequent differentiation and expansion to NK cells). Knockout of DR4/DR5 is determined by flowcytometry based assays.

Example 8—Knockdown of TRAIL Receptors DR4 and DR5 in NK Cells

[0170] siRNA knockdown of DR4 and/or DR5 in NK-92 cells, KHYG-1 cells and primary NK cells was performed by electroporation. siRNA based delivery was performed using the Maxcyte GT system.

[0171] The cells were then incubated in a humidified 37° C./5% CO.sub.2 incubator until DR4 and/or DR5 knockdown analysis was performed. Flow cytometry analysis was performed 72 hours after electroporation, and (optionally) just prior to electroporation of TRAIL variant (e.g. E195R/D269H) mRNA in order to measure DR4 and/or DR5 expression levels. This electroporation protocol was found to reliably result in DR4 and DR5 knockdown in KHYG-1 cells, NK-92 cells and primary NK cells.

Example 9—NK Cell Fratricide Resistance

[0172] As illustrated in FIG. 15, NK cell fratricide in the following cells was assessed: (1) primary NK cells, otherwise referred to as mock or wildtype NK cells, (2) primary NK cells expressing high affinity membrane-bound TRAIL ligand DR5.sup.E195R;D269H, (3) primary NK cells with a DR5.sup.KD via siRNA and (4) primary NK cells with a DR5.sup.KD via siRNA and also expressing high affinity membrane-bound TRAIL ligand DR5.sup.E195R;D269H.

[0173] Primary NK cells receiving the DR5 knockdown were electroporated with the DR5 siRNA on day 9 of the expansion, whereas primary NK cells receiving the DR5 TRAIL variant were electroporated with the variant mRNA on day 12 of the expansion.

[0174] It was observed that after prolonged expansion of the primary NK cells in IL-2 containing growth media that DR5 expression became upregulated, leading to increased fratricide when the DR5 TRAIL variant was expressed.

[0175] The data clearly indicate that knocking down DR5 expression using siRNA protects primary expanded NK cells from fratricide, regardless of whether those NK cells express wildtype TRAIL or the high-affinity DR5 TRAIL variant.

[0176] The invention thus provides highly cytotoxic NK cells for use in therapy.