Targeted il-15 construct delivery

Abstract

NK cells are modified to increase their cytotoxicity; the cells and compositions thereof have a use in the treatment of cancer. Production of modified NK cells and NK cell lines is via genetic modification to knock down expression of the CISH gene and to knock in expression of a gene encoding an IL-15-IL-15Rα fusion protein.

Description

[0001] TARGETED IL-15 CONSTRUCT DELIVERY

[0002] Introduction

[0003] The present invention relates to the modification of natural killer (NK) cells and NK cell lines to produce cells with a more cytotoxic phenotype. Furthermore, the present invention relates to methods of producing modified NK cells and NK cell lines, compositions containing the cells and cell lines and uses of said cells, lines and compositions in the treatment of cancer.

[0004] Background to the Invention

[0005] Typically, immune cells require a target cell to present antigen via major histocompatibility complex (MHC) before triggering an immune response resulting in the death of the target cell. This allows cancer cells not presenting MHC class I to evade the majority of immune responses.

[0006] NK cells, however, are able to recognize cancer cells in the absence of MHC class I expression. Hence, they perform a critical role in the body’s defense against cancer.

[0007] On the other hand, in certain circumstances, cancer cells demonstrate an ability to dampen the cytotoxic activity of NK cells, through expression of ligands that bind inhibitory receptors on the NK cell membrane. Resistance to cancer therapy can involve a balance between these and other factors.

[0008] Cytotoxicity, in this context, refers to the ability of immune effector cells, e.g. NK cells, to induce cancer cell death, e.g. by releasing cytolytic compounds or by binding receptors on cancer cell membranes and inducing apoptosis of said cancer cells. Cytotoxicity is affected not only by signals that induce release of cytolytic compounds but also by signals that inhibit their release. An increase in cytotoxicity will therefore lead to more efficient killing of cancer cells, with less chance of the cancer cells dampening the cytotoxic activity of the NK, as mentioned above.

[0009] Cytokine-inducible SH2-containing protein (CIS) expression is induced by certain growth cytokines and is a key negative regulator of interleukin-15 (IL-15) signaling in natural killer (NK) cells. CIS is encoded by the CISH gene. CIS expression is associated with limited cell expansion and decreased cytotoxic activity against multiple cancer cell lines when maintained at low cytokine concentrations. As such, knocking out CIS expression by deletion of the CISH gene in NK cells has recently been reported as beneficial for NK cell cytotoxicity against cancer cells (Zhu et al. 2020 “Metabolic Reprograming via Deletion of CISH in Human iPSC-Derived NK Cells Promotes In Vivo Persistence and Enhances Anti-tumor Activity” Cell: Vol. 27(2): pp. 224-237).

[0010] TGF beta receptor (TGFBR) signaling in NK cells is known to be immunosuppressive and affect the function of NK cells. TGF beta can alter the expression of effector molecules and of activating and chemokine receptors, influence metabolism, and induce the NK cell conversion toward the less cytotoxic ILC-1 phenotype (Regis et al. 2020 “NK Cell Function Regulation by TGF-[3-lnduced Epigenetic Mechanisms” Front Immunol. 11 : 311 ).

[0011] NK cells express the IL-15 receptor and are thus responsive to IL-15 in the cancer microenvironment, typically expressed by other cells of the immune system in both a soluble form and in a membrane-bound form.

[0012] Known art in this field includes Daher et al. 2021 “Targeting a cytokine checkpoint enhances the fitness of armored cord blood CAR-NK cells” Blood, American Society of Hematology: Vol. 137(5): pp. 624-636; US 2021 / 015859; Xu et al. 2021 “A novel multimeric IL15 / IL15R[alpha] - Fc complex to enhance cancer immunotherapy” Oncoimmunology: Vol.10(1); WO 2023 / 060136; WO 2022 / 263682; and WO 2022 / 120370.

[0013] Whilst the theoretical interplay between CISH expression and IL-15 signaling is known, it is not known whether any interplay can be exploited in order to provide new and improved cancer therapies.

[0014] There exists a need for alternative and preferably improved cell-based therapies with improved cytotoxicity and persistence, particularly against cancer cells.

[0015] An object of the present invention is to provide NK cells and NK cell lines that target cancer cells with a more cytotoxic and / or persistent phenotype. A further object is to provide methods for producing modified NK cells and NK cell lines, compositions containing the cells or cell lines and uses of such in therapy, specifically the treatment of cancers. More particular embodiments aim to provide treatments for identified cancers. Specific embodiments aim at combining two or more modifications of NK cells and NK cell lines to enhance the cytotoxicity and persistence of the modified cells.

[0016] Summary of the Invention

[0017] There are provided herein modified NK cells and NK cell lines with a more cytotoxic phenotype, and methods of making the cells and cell lines. Also provided are compositions of modified NK cells and NK cell lines, and uses of said cells I compositions for treating cancer. Cytotoxicity in this context, as above, refers to the killing of cancer cells. Improved cytotoxicity can be, for example, enhanced killing of target cells, killing of target cells over a shorter time period, or killing of different targets.

[0018] The invention provides NK cells and NK cell lines that exhibit (e.g. have been modified to have) reduced expression of CISH and to express an IL-15-IL-15Ra fusion protein, wherein the IL-15Ra preferably lacks the sushi domain.

[0019] According to the invention, there are further provided methods of treating cancer, e.g. solid tumors, using the modified NK cells and cell lines, both generally and specifically, wherein the modified NK cells and cell lines are engineered to have reduced expression of CISH and an IL-15-IL-15Ra fusion protein, wherein the IL- 15Ra preferably lacks the sushi domain, and, optionally, to express one or more chimeric antigen receptors.

[0020] Diseases particularly treatable according to the invention include cancers, specifically solid cancers, and more particularly ovarian or breast cancers. Tumors and cancers in humans in particular can be treated. References to tumors herein include references to neoplasms.

[0021] Details of the Invention

[0022] Accordingly, the present invention provides a natural killer (NK) cell or NK cell line that has been modified to have reduced expression of CISH and to express an IL-15- IL-15Ra fusion protein, wherein the IL-15Ra preferably lacks a functional sushi domain. An IL-15Ra subunit lacking the sushi domain is one in which the activity of the sushi domain is diminished by modification. It is preferred that the IL-15Ra modification is a genetic deletion of the sushi domain, as exemplified below. An alternative embodiment of the invention comprises a sushi domain, or a part thereof, which is modified so as to exhibit substantially no sushi domain function. The fusion protein is therefore without an active I functional sushi domain.

[0023] Unless the context suggests otherwise, the term ‘NK cell’ will henceforth be used to encompass both primary NK cells and NK cell lines, whether allogeneic or autologous.

[0024] The NK cell of the invention is preferably human. Optionally, the NK cell is derived from either human cord blood or human peripheral blood.

[0025] It is preferred that the NK cell has been further modified to have reduced expression of TGFBR2.

[0026] It is preferred that the NK cell is modified to have reduced expression of CISH and TGFBR2, as well as to express an IL-15-IL-15Ra fusion protein, wherein the IL-15Ra lacks the sushi domain.

[0027] Preferably, CISH expression is reduced by at least 50%, at least 75%, at least 90%, at least 95%, more preferably at least 99%, compared to the same NK cell or NK cell line without the modification (wildtype NK cell).

[0028] It is particularly preferred that NK cells according to the invention have CISH expression genetically knocked out.

[0029] Preferably, TGFBR2 expression is reduced by at least 50%, at least 75%, at least 90%, at least 95%, more preferably at least 99%, compared to the same NK cell or NK cell line without the modification (wildtype NK cell).

[0030] It is particularly preferred that NK cells according to the invention have TGFBR2 expression genetically knocked out. As such, it is preferred that CISH expression is genetically knocked out and the IL- 15-IL-15Ra fusion protein is genetically knocked in. Preferably, the IL-15-IL-15Ra fusion protein is knocked into the locus of the CISH gene. As already mentioned above, it is preferred that the IL-15Ra subunit lacks the sushi domain.

[0031] Advantageously, the inventors have found that, when the IL-15-IL-15Ra fusion protein (with the sushi domain deleted, i.e. non-functional) is expressed in NK cells with reduced CISH expression, the NK cells of the invention have a significantly enhanced ability to kill cancer cells. This is a truly surprising finding, because the sushi domain has long been seen as critical for IL-15 to exert its functional activity (Xiao-qing Wei et al. 2001 ; Mortier et al. 2006; Quemener et al. 2019).

[0032] For the avoidance of doubt, references herein to “modified” NK cells include the progeny of those cells, at least to the extent that the progeny carry the modification.

[0033] Genetic knockout of CISH and / or TGFBR2, as well as genetic knock-in of the IL-15- IL-15Ra fusion protein, is preferably by CRISPR gene editing.

[0034] There are many ways in which IL-15 can be administered to or with NK cells, e.g., in soluble form or complexed with the IL-15 receptor alpha chain. Additionally, NK cells may be modified to express IL-15 as a membrane-bound cytokine. However, modifying an NK cell to express an IL-15-IL-15Ra fusion protein lacking the sushi domain was not previously contemplated. Moreover, the improvement in cytotoxic potential observed in the modified NK cell, when combined with a CISH KO, was completely unexpected.

[0035] It is preferred that the IL-15-IL-15Ra fusion protein is at least partially secreted by the NK cell. As such, the IL-15 is preferably soluble I secreted IL-15 (sclL-15). Optionally, the sclL-15 comprises the N72D mutation. Optionally, the sclL-15 may be cleavable from the fusion protein, e.g., in vivo. Optionally, the IL-15R lacks a membranetethering portion. In other embodiments, the fusion protein is membrane-bound.

[0036] Accordingly, it is preferred that the IL-15-IL-15Ra fusion protein comprises a signal peptide. Optionally, the signal peptide is one that facilitates secretion of the fusion protein / sclL-15. Preferably, the signal peptide is an IgE signal peptide. CB-NK cells possess an immature phenotype, when compared to other known NK cells, e.g., peripheral blood NK cells, presenting in reduced cytotoxicity, with higher expression of classical inhibitory receptors, such as NKG2A. For this reason, along with the difficulties experienced in successfully isolating these cells from umbilical cord blood, CB-NK cells have hitherto been regarded as an unfavorable option for NK cells to be used in cell therapies for cancer. With that said, the present inventors have worked to identify and optimize CB-NK cell modifications that have promising value in cancer therapeutics.

[0037] NK cells of the invention are for use in therapy, especially in treating cancer in a patient. The cancer is suitably a solid tumor. Preferably, the cancer is ovarian cancer or breast cancer.

[0038] In a particularly preferred embodiment, the invention provides a human CB-NK cell, genetically modified to have CISH expression knocked out and to express an IL-15- IL-15Ra fusion protein, for use in solid tumor therapy.

[0039] The invention also provides a method of making a modified NK cell or NK cell line as described above or below, the method comprising the following steps: a) isolating an NK cell from human umbilical cord blood or human peripheral blood; b) knocking out expression of the CISH gene; and c) knocking in expression of a gene encoding an IL-15-IL-15Ra fusion protein, preferably at the locus of the CISH gene.

[0040] In embodiments of the invention, the respective knocking out and knocking in steps can be carried out in any order and may also be carried out simultaneously.

[0041] The IL-15-IL-15Ra fusion protein preferably has the sushi domain of the IL-15Ra deleted.

[0042] Compositions of the modified NK cells of the invention are also provided, as are methods of using the modified NK cells or compositions thereof in therapy, e.g., cancer therapy. Pharmaceutical compositions are preferred. In other embodiments of the invention, NK cells are provided that also express or overexpress a TRAIL ligand. Preferably, the TRAIL ligand is a mutant (variant) TRAIL ligand. The resulting NK cells exhibit increased binding to TRAIL receptors and, as a result, increased cytotoxicity against cancers, especially solid cancers, in particular ovarian, breast and colorectal cancers, and blood cancers, in particular leukemias. The NK cells with this combined activity may also be effective in reducing cancer metastases.

[0043] The TRAIL mutants I variants preferably have lower affinity (or in effect no affinity) for ‘decoy’ receptors, compared with the binding of wild type TRAIL to decoy receptors. Such decoy receptors represent a class of TRAIL receptors that bind TRAIL ligand but do not have the capacity to initiate cell death and, in some cases, act to antagonize the death signaling pathway. Mutant I variant TRAIL ligands may be prepared according to WO 2009 / 077857.

[0044] The mutants I variants may separately have increased affinity for TRAIL receptors, e.g. DR4 and DR5. Wildtype TRAIL is typically known to have a KD of >2 nM for DR4, >5 nM for DR5 and >20 nM for the decoy receptor DcR1 (WO 2009 / 077857; measured by surface plasmon resonance), or around 50 to 100 nM for DR4, 1 to 10 nM for DR5 and 175 to 225 nM for DcR1 (Truneh, A. et al. 2000; measured by isothermal titration calorimetry and ELISA). Therefore, an increased affinity for DR4 is suitably defined as a KD of <2 nM or <50 nM, respectively, whereas an increased affinity for DR5 is suitably defined as a KD of <5 nM or <1 nM, respectively. A reduced affinity for decoy receptor DcR1 is suitably defined as a KD of >50 nM or >225 nM, respectively. In any case, an increase or decrease in affinity exhibited by the TRAIL variant / mutant is relative to a baseline affinity exhibited by wildtype TRAIL. The affinity is preferably increased at least 10%, at least 25%, at least 50%, at least 100%, more preferably at least 1000%, compared with that exhibited by wildtype TRAIL.

[0045] The TRAIL variant preferably has an increased affinity for DR5 as compared with its affinity for DR4, DcR1 and DcR2. Preferably, the affinity is at least 1.5-fold, 2-fold, 5- fold, 10-fold, 100-fold, or even 1 ,000-fold or greater for DR5 than for one or more of DR4, DcR1 and DcR2. More preferably, the affinity is at least 1.5-fold, 2-fold, 5-fold, 10-fold, 100-fold, or even 1 ,000-fold or greater for DR5 than for at least two, and preferably all, of DR4, DcR1 and DcR2. The TRAIL variant preferably has an increased affinity for one or both of DR4 and DR5 as compared with its affinity for wildtype TRAIL. Preferably, the affinity is at least 1.5-fold, 2-fold, 5-fold, 10-fold, 100-fold, or even 1 ,000-fold or greater for DR4 and / or DR5 than for wildtype TRAIL.

[0046] Further specific embodiments comprise a NK cell expressing a mutant TRAIL ligand that has reduced or no affinity for TRAIL decoy receptors. Further specific embodiments comprise a NK cell expressing a mutant TRAIL ligand that has reduced or no affinity for TRAIL decoy receptors and increased affinity for DR4 and / or DR5.

[0047] Binding affinity may be measured according to any suitable method known in the art. Preferably, binding affinity is measured using surface plasmon resonance, isothermal titration calorimetry or ELISA.

[0048] In certain embodiments, the TRAIL variant comprises at least one amino acid substitution at a position selected from the group consisting of 131 , 149, 159, 160, 189, 191 , 193, 195, 199, 200, 201 , 203, 204, 212, 213, 214, 215, 218, 240, 251 , 261 , 264, 266, 267, 269, and 270.

[0049] In certain embodiments, the TRAIL variant comprises at least one substitution selected from the group consisting of G131 R, G131 K, R149I, R149M, R149N, R149K, S159R, G160E, Y189A, Y189Q, R191 K, Q193H, Q193K, Q193S, Q193R,

[0050] E195R, N199V, N199R, N199H, T200H, K201 R, K201 H, D203A, K204E, K204D,

[0051] K204Y, K212R, Y213W, T214R, S215D, S215E, S215H, S215K, S215D, D218H,

[0052] D218A, Y240A, K251 D, K251 E, K251 Q, T261 L, H264R, I266L, D267Q, D269A,

[0053] D269H, and H270D.

[0054] In certain embodiments, the TRAIL variant comprises at least two substitutions selected from the group consisting of G131 R, G131 K, R149I, R149M, R149N, R149K, S159R, G160E, Y189A, Y189Q, R191 K, Q193H, Q193K, Q193S, Q193R,

[0055] E195R, N199V, N199R, N199H, T200H, K201 R, K201 H, D203A, K204E, K204D,

[0056] K204Y, K212R, Y213W, T214R, S215D, S215E, S215H, S215K, S215D, D218H,

[0057] D218A, Y240A, K251 D, K251 E, K251 Q, T261 L, H264R, I266L, D267Q, D269A,

[0058] D269H, and H270D. In certain embodiments, the TRAIL variant comprises at least three substitutions selected from the group consisting of G131R, G131K, R149I, R149M, R149N, R149K, S159R, G160E, Y189A, Y189Q, R191K, Q193H, Q193K, Q193S, Q193R,

[0059] E195R, N199V, N199R, N199H, T200H, K201R, K201H, D203A, K204E, K204D,

[0060] K204Y, K212R, Y213W, T214R, S215D, S215E, S215H, S215K, S215D, D218H,

[0061] D218A, Y240A, K251D, K251E, K251Q, T261L, H264R, I266L, D267Q, D269A,

[0062] D269H, and H270D.

[0063] In certain embodiments, amino acid substitution of the TRAIL variant is selected from the group consisting of G131R, G131K, R149I, R149M, R149N, R149K, S159R, G160E, Y189A, Y189Q, R191K, Q193H, Q193K, Q193S, Q193R, E195R, N199V,

[0064] N199R, N199H, T200H, K201R, K201H, D203A, K204E, K204D, K204Y, K212R,

[0065] Y213W, T214R, S215D, S215E, S215H, S215K, S215D, D218H, D218A, Y240A,

[0066] K251D, K251E, K251Q, T261L, H264R, I266L, D267Q, D269A, D269H, H270D,

[0067] T214R / E195R, T214R / D269H, Y189A / Q193S / N199V / K201R / Y213W / S215D, Y213W / S215D, N199R / K201H, N199H / K201R, G131R / N199R / K201H, G131R / N199R / K201H / R149I / S159R / S215D, G131R / R149I / S159R / S215D, G131R / N199R / K201H / R149I / S159R / S215D, G131R / D218H, Y189Q / R191K / Q193R / H264R / I266L / D267Q, T261 L / G160E, T261 L / H270D, T261 L / G160E / H270D, and T261L / G160E / H270D / T200H (use of 7” indicates multiple amino acid substitutions).

[0068] In certain embodiments, amino acid substitution of the TRAIL variant is selected based on the variant having an increased affinity for DR5; a substitution of this kind may be selected from the group consisting of D269H, E195R, T214R, D269H I E195R, T214R / E195R, T214R / D269H, N199V, Y189A / Q193S / N199V / K201 R / Y213W / S215D, Y213W / S215D, D269A and Y240A.

[0069] In certain embodiments, amino acid substitution of the TRAIL variant is selected based on the variant having an increased affinity for DR4; a substitution of this kind may be selected from the group consisting of G131R, G131K, R149I, R149M, R149N, R149K, S159R, Q193H, W193K, N199R, N199R / K201H, N199H / K201R, G131R / N199R / K201H, G131 R / N199R / K201 H, G131R / N199R / K201H / R149I / S159R / S215D, G131 R / R149I / S159R / S215D, G131R / D218H, K201R, K201H, K204E, K204D, K204L, K204Y, K212R, S215E, S215H, S215K, S215D, D218H, K251 D, K251 E, K251 Q and Y189Q / R191 K / Q193R / H264R / I266L / D267Q. In certain embodiments, amino acid substitution of the TRAIL variant is selected based on the variant having a decreased affinity for TRAIL decoy receptors; a substitution of this kind may be selected from the group consisting of T261 L, H270D, T200H, T261 L / G160E, T261 L / H270D, T261 L / G160E / H270D, T261 L / G160E / H270D / T200H, D203A and D218A.

[0070] Treatment of a cancer using modified NK cells expressing TRAIL or a TRAIL variant is optionally enhanced by administering to a patient an agent capable of upregulating expression of TRAIL death receptors on cancer cells. This agent may be administered prior to, in combination with or subsequently to administration of the modified NK cells. It is preferable, however, that the agent is administered prior to administering the modified NK cells. The agent upregulates expression of DR5 on cancer cells. The agent may optionally be a chemotherapeutic medication, e.g. a proteasome inhibitor, one of which is Bortezomib, and administered in a low dose capable of upregulating DR5 expression on the cancer. Other examples of DR5- inducing agents include Gefitinib, Piperlongumine, Doxorubicin, Alpha-tocopheryl succinate and HDAC inhibitors.

[0071] In certain embodiments of the invention, NK cells are provided that are further modified so as to have reduced or absent function of a checkpoint inhibitory receptor. NK cells may be produced that have one or more checkpoint inhibitory receptor genes knocked down / out. Preferably, these receptors are specific checkpoint inhibitory receptors. Preferably still, these checkpoint inhibitory receptors are one or more or all of TIGIT, CD96 (TACTILE), CD152 (CTLA4), CD223 (LAG-3), CD279 (PD-1 ), CD328 (SIGLEC7), SIGLEC9 and / or TIM-3. In other embodiments, NK cells are provided in which one or more inhibitory receptor signaling pathways are knocked out or exhibit reduced function - the result again being reduced or absent inhibitory receptor function. For example, signaling pathways mediated by SHP-1 , SHP-2 and / or SHIP are knocked out by genetic modification of the cells.

[0072] It is preferred to reduce function of checkpoint inhibitory receptors over other inhibitory receptors, due to the expression of the former following NK cell activation. The normal or ‘classical’ inhibitory receptors, such as the majority of the KIR family, NKG2A and LIR-2, bind MHC class I and are therefore primarily involved in reducing the problem of self-targeting. Preferably, therefore, checkpoint inhibitory receptors are knocked out. Reduced or absent function of these receptors according to the invention prevents cancer cells from suppressing immune effector function (which might otherwise occur if the receptors were fully functional). Thus, a key advantage of these embodiments of the invention lies in NK cells that are less susceptible to suppression of their cytotoxic activities by cancer cells; as a result, they are useful in cancer treatment.

[0073] As used herein, references to inhibitory receptors generally refer to a receptor expressed on the plasma membrane of an immune effector cell, e.g. a NK cell, whereupon binding its complementary ligand resulting intracellular signals are responsible for reducing the cytotoxicity of said immune effector cell. These inhibitory receptors are expressed during both Testing’ and ‘activated’ states of the immune effector cell and are often associated with providing the immune system with a ‘selftolerance’ mechanism that inhibits cytotoxic responses against cells and tissues of the body. An example is the inhibitory receptor family ‘KIR’ which are expressed on NK cells and recognize MHC class I expressed on healthy cells of the body.

[0074] Also as used herein, checkpoint inhibitory receptors are usually regarded as a subset of the inhibitory receptors above. Unlike other inhibitory receptors, however, checkpoint inhibitory receptors are expressed at higher levels during prolonged activation and cytotoxicity of an immune effector cell, e.g. a NK cell. This phenomenon is useful for dampening chronic cytotoxicity at, for example, sites of inflammation. Examples include the checkpoint inhibitory receptors PD-1 , CTLA-4 and CD96, all of which are expressed on NK cells.

[0075] It is preferred in the invention not to reduce the function of inhibitory receptors that bind MHC class I.

[0076] The modified cells of the invention may also express a chimeric antigen receptor (CAR). The membrane-bound CAR typically comprises a targeting sequence (commonly an antibody-derived single-chain fragment (scFv)) and usually a hinge (to overcome steric hindrance issues), a spacer, a membrane-spanning element and a signalling endodomain.

[0077] It is preferred that the gene encoding the CAR is inserted with the IL-15-IL-15Ra fusion protein gene into a bicistronic vector. Preferably, this bicistronic vector is used to modify the NK cells of the invention to express the exogenous genes at the CISH gene locus, consequently knocking out expression of CISH.

[0078] The CAR preferably comprises a targeting region that binds an antigen selected from MUC-1 , HER2, VEGF, EpCAM, EGFR, CD38, CD96, CLL-1 , SLAMF7 and CD19. Preferably, the CAR binds an antigen selected from HER2, VEGF, EpCAM and EGFR.

[0079] In a preferred embodiment, the CAR-NK cell binds the CD38 antigen.

[0080] In a preferred embodiment, the CAR-NK cell binds the HER2 antigen.

[0081] In a preferred embodiment, the CAR-NK cell binds the VEGF antigen.

[0082] In a preferred embodiment, the CAR-NK cell binds the EpCAM antigen.

[0083] In a preferred embodiment, the CAR-NK cell binds the EGFR antigen.

[0084] The CAR may be bispecific and bind two of the above-mentioned antigens.

[0085] Examples of sequences that are known to bind aberrantly glycosylated MUC-1 can be used as the targeting sequence, such sequences including 5E5, SM3 and HMFG2, and are suitably incorporated in a CAR of the invention; preferably the CAR comprises the HMFG2 sequence. Nevertheless, further sequences for targeting aberrantly glycosylated MUC-1 may be identified through screening methods known in the prior art, wherein high affinity sequences can be used to produce CAR-NK cells targeting MUC-1 .

[0086] Examples of sequences that are known to bind CD38 can be used as the targeting sequence, such sequences including daratumumab, isatuximab and those disclosed in WO 2018 / 104562, and are suitably incorporated in a CAR of the invention. It is preferred that the CD38 CAR has reduced affinity for CD38, compared to the affinity of daratumumab for CD38. Preferably, this reduced affinity is by at least 10%, at least 25%, more preferably at least 45%, compared to the affinity of daratumumab. Preferably, also, this reduced affinity is no more than 90%, no more than 75%, more preferably no more than 55%, compared to the affinity of daratumumab. The CARs used in the NK cells of the invention may comprise or be linked to one or more NK cell costimulatory domains, e.g. CD28, CD134 I 0X40, 4-1 BB I CD137, CD3zeta I CD247, DAP 12 or DAP 10. Binding of the CAR to its antigen on a target cell thus promotes cytotoxic signals in the modified NK cell.

[0087] NK cells according to the invention may also be treated or pre-treated to render them incapable of division. This results in further reduced lifetime in circulation in the patient, e.g. in comparison with T cells, further mitigating the risks above, and also with reduced or absent propensity to form tumors in a patient.

[0088] NK cells of the invention are for use in therapy, especially in treating cancer in a patient. The cancer is suitably a solid tumor. Preferably, the cancer is ovarian cancer or breast cancer.

[0089] NK cell resistant cancers in general are well-known in the art (Pardoll, D.M. Immunity (2015) 42:605-606). Sensitivity of cancer cells to NK cell-mediated killing is determined by a number of factors. There exists a balance of positive and negative signals, largely delivered through membrane receptors on NK cells interacting with ligands on cancer cells. It is often the balance of expression of the ligands for these receptors that determines whether a cancer is sensitive or resistant to killing by NK cells (Yokoyama, W.M. Immunol Res (2005) 32:317-325).

[0090] Cancer sensitivity to NK cell-mediated cytotoxicity is generally understood as falling into one of the following categories: highly resistant, resistant, sensitive and highly sensitive. In the laboratory, cancer cells can be screened for their sensitivity to NK cell-mediated cytotoxicity through the use of cytotoxicity assays. Each category is then understood as corresponding to the percentage of cancer cells that are killed during exposure to NK cells at a specific effector: target (E:T) ratio and for a specific amount of time.

[0091] In examples of the invention, a cancer is said to be highly resistant to NK cell- mediated killing if < 25% of the cancer cells are killed after incubation with NK cells for up to 15 hours at an E:T ratio of up to 5: 1 . In examples of the invention, a cancer is said to be resistant to NK cell-mediated killing if < 50% of the cancer cells are killed after incubation with NK cells for up to 15 hours at an E:T ratio of up to 5: 1 .

[0092] In examples of the invention, a cancer is said to be sensitive to NK cell-mediated killing if > 50% of the cancer cells are killed after incubation with NK cells for up to 15 hours at an E:T ratio of up to 5: 1 .

[0093] In examples of the invention, a cancer is said to be highly sensitive to NK cell- mediated killing if > 75% of the cancer cells are killed after incubation with NK cells for up to 15 hours at an E:T ratio of up to 5: 1 .

[0094] Optional features of the invention include providing further modifications to the NK cells and NK cell lines described above, wherein, for example, a Fc receptor (which can be CD16, CD32 or CD64, including subtypes and derivatives) is expressed on the surface of the cell. In use, these cells can show increased recognition of antibody-coated cancer cells and improved activation of the cytotoxic response.

[0095] Modified NK cells and compositions thereof described herein, above and below, are suitable for treatment of cancer, in particular cancer in humans, e.g. for treatment of solid cancers. The NK cells and derivatives thereof are preferably human NK cells. For human therapy, human NK cells are preferably used. The invention also provides methods of treating cancer in humans comprising administering an effective amount of NK cells or compositions comprising the same.

[0096] Various routes of administration will be known to the skilled person to deliver active agents and combinations thereof to a patient in need. Administration of the modified NK cells can be systemic or localized, for example via the intraperitoneal route. In other embodiments, the active agent is administered more directly. Thus, administration can be directly intratumorally, suitable especially for solid tumors.

[0097] Examples

[0098] The invention is now illustrated in specific embodiments with reference to the accompanying drawings in which:

[0099] Fig. 1 shows that CISH KO and sclL-15 knock-in in NK cells synergistically enhanced the cytotoxic profile against human breast cancer cells; Fig. 2 shows that CISH KO and sclL-15 knock-in in NK cells synergistically enhanced the cytotoxic profile against human ovarian cancer cells;

[0100] Fig. 3 shows that NK cells with CISH and TGFBRII double KO (DKO) and sclL-15 knock-in were highly effective in cancer cell killing;

[0101] Fig. 4 shows that sclL-15 genomic insertion at the CISH locus increased NK cell expansion;

[0102] Fig. 5 shows that sclL-15 genomic insertion at the CISH locus increased IL-15 expression in NK cells, compared to those cells expressing the sclL-15 gene extra- genomically; and

[0103] Fig. 6 shows that CISH KO NK cells with the IL-15-IL-15Ra fusion protein knocked in were highly effective in cancer cell killing.

[0104] DNA, RNA and amino acid sequences are referred to below, in which:

[0105] SEQ ID NO: 1 refers to the DNA sequence encoding the full-length sclL-15 peptide;

[0106] SEQ ID NO: 2 refers to the amino acid sequence of the full-length sclL-15 peptide;

[0107] SEQ ID NO: 3 refers to the DNA sequence encoding the IL-15-IL-15Ra fusion protein without the sushi domain and including the IgE signal peptide; and

[0108] SEQ ID NO: 4 refers to the amino acid sequence of the IL-15-IL-15Ra fusion protein without the sushi domain and including the IgE signal peptide.

[0109] Example 1 - Design Protocol for Knockout of CISH in NK Cells

[0110] NK cells were prepared as follows, having CIS function removed. gRNA constructs were designed and prepared to target the CISH gene in NK cells. CRISPR / Cas9 genome editing was used to knock out CISH expression.

[0111] A total of 3 gRNA candidates were selected for the CISH gene and their cleavage efficacies in primary expanded NK cells were determined. The cells were then electroporated with the gRNA:Cas9 ribonucleoprotein (RNP) complex using Maxcyte® GT and subsequently knockout of CISH was analysed using flow cytometry. The cleavage activity of the gRNA was also determined using an in vitro mismatch detection assay. T7E1 endonuclease recognises and cleaves non-perfectly matched DNA, allowing the parental CISH gene to be compared to the mutated gene following CRISPR / Cas9 transfection and non-homologous end joining (NHEJ). The gRNA with highest KO efficiency was selected for further experiments to knockout CISH in NK cells. Knockout of CISH was determined by flow cytometry based assays.

[0112] Example 2 - Knockout of CISH in NK Cells

[0113] Materials

[0114] 1. Enriched cord blood-derived NK cells.

[0115] 2. NK-MACS medium, Premium grade IL-15, anti-CD3 VioBlue, Anti-CD56VB515, Inside Stain Kit, MACSquant 16 (Miltenyi Biosciences).

[0116] 3. Human Serum Albumin (Sigma).

[0117] 4. Gene knockout Kit V2 for CISH and Cas9 recombinant protein (Synthego).

[0118] 5. Cloudz™ Cell Activation Kits (R&D systems).

[0119] 6. Anti-CISH antibody (D4D9) (Cell Signaling Technology).

[0120] 7. Electroporation buffer (EB buffer), OC-100X2 processing assembly (PA), MaxCyte ATx electroporation system (Maxcyte).

[0121] 8. Solu-Cortef (Hydrocortisone) (Pfizer).

[0122] 9. Ficoll (Cytiva).

[0123] Protocol

[0124] Cord Blood Mononuclear Cells (CBMCs) were isolated by the Ficoll method and enriched for 15 days in NK-MACS medium with 10% human serum albumin, CD2 / NKp46 microsphere (Cloudz™), anti-CD3, anti-CD16, IL-15 and hydrocortisone.

[0125] NK cell enrichment of over 90% was confirmed by FACS (MACSquant 16) analysis by staining with anti-CD3 and anti-CD56 antibodies. CB-NK cells were washed twice with Maxcyte electroporation buffer (EB buffer) 300xg for 10m ins and 5 million cells were taken in 75pl of EB buffer. sgRNA for CISH was dissolved in nuclease-free TE buffer to get a stock of 200pM. 1200pmol of sgRNA was mixed with 160pmol Cas9 at a 7.5: 1 ratio. The volume was adjusted with EB buffer to 25pl and incubated at room temperature for 10 minutes for formation of RNP complex. 25pl RNP complex was mixed with 75pl CB-NK cells and transferred to the OC- 100X2 processing assembly. The cells were then electroporated using the NK5 protocol and left to rest at room temperature for 15 minutes.

[0126] Electroporated cells were then transferred to the Grex 6 well plate in NK-MACS medium with 10% AB and 20ng / mL IL-15 and incubated at 37°C in a 5% CO2 incubator for 13 days. The cells were analysed for CISH knockout every 3-4 days. Intracellular expression of CISH was analysed using Inside Stain Kit and an anti- CISH antibody.

[0127] Example 3 - Design Protocol for Knockout of TGFBRII in NK Cells

[0128] NK cells were prepared as follows, having TGFBRII function removed. gRNA constructs were designed and prepared to target the TGFBRII gene in NK cells. CRISPR / Cas9 genome editing was then used to knock out TGFBRII expression.

[0129] A total of 3 gRNA candidates were selected for the TGFBRII gene and their cleavage efficacies in primary expanded NK cells were determined. The cells were electroporated with the gRNA:Cas9 ribonucleoprotein (RNP) complex using Maxcyte® GT and subsequently knockout of TGFBRII was analysed by flow cytometry. The cleavage activity of the gRNA was also determined using an in vitro mismatch detection assay. T7E1 endonuclease recognises and cleaves non-perfectly matched DNA, allowing the parental TGFBRII gene to be compared to the mutated gene following CRISPR / Cas9 transfection and non-homologous end joining (NHEJ).

[0130] The gRNA with highest KO efficiency was selected for further experiments to knockout TGFBRII in NK cells. Knockout of TGFBRII was then determined by flow cytometry based assays.

[0131] Example 4 - Knockout of TGFBRII in NK Cells

[0132] Materials

[0133] 1. Enriched cord blood-derived NK cells.

[0134] 2. NK-MACS medium, Premium grade IL-15, anti-CD3 VioBlue, Anti-CD56VB515, Inside Stain Kit, MACSquant 16 (Miltenyi Biosciences).

[0135] 3. Human Serum Albumin (Sigma). 4. Gene knockout Kit for TGFBRII and Cas9 recombinant protein.

[0136] 5. Cloudz™ Cell Activation Kits (R&D systems).

[0137] 6. pSMAD2 antibody.

[0138] 7. Electroporation buffer (EB buffer), OC-100X2 processing assembly (PA), MaxCyte ATx electroporation system (Maxcyte).

[0139] 8. Solu-Cortef (Hydrocortisone) (Pfizer).

[0140] 9. Ficoll (Cytiva).

[0141] Protocol

[0142] Cord Blood Mononuclear Cells (CBMCs) were isolated by the Ficoll method and enriched for 15 days in NK-MACS medium with 10% human serum albumin, CD2 / NKp46 microsphere (Cloudz™), anti-CD3, anti-CD16, IL-15 and hydrocortisone.

[0143] NK cell enrichment of over 90% was confirmed by FACS (MACSquant 16) analysis by staining with anti-CD3 and anti-CD56 antibodies. CB-NK cells were washed twice with Maxcyte electroporation buffer (EB buffer) 300xg for 10 minutes and 5 million cells were taken in 75pl of EB buffer. sgRNA for TGFBRII was dissolved in nuclease-free TE buffer to get a stock of 200pM. 1200pmol of sgRNA was mixed with 160pmol Cas9 at a 7.5: 1 ratio. The volume was adjusted with EB buffer to 25pl and incubated at room temperature for 10 minutes for formation of RNP complex.

[0144] 25pl RNP complex was mixed with 75pl CB-NK cells and transferred to the OC- 100X2 processing assembly. The cells were then electroporated using the NK5 protocol and left to rest at room temperature for 15 minutes.

[0145] Electroporated cells were subsequently transferred to the Grex 6 well plate in NK- MACS medium with 10% AB and 20ng / mL IL-15 and incubated at 37°C in a 5% CO2 incubator for 13 days. The cells were analysed for TGFBRII knockout every 3-4 days. Successful TGFBRII knockout was determined by pSMAD2 expression using an anti- pSMAD2 antibody following stimulation with 50ng / ml of TGFB1 for 15 minutes.

[0146] Example 5 - CISH KO NK cells with sclL-15 KI demonstrated synergy in enhancing NK cell cytotoxicity against cancer cells NK cell expansion

[0147] Fresh cord blood was processed using Lymphoprep followed by ACK lysis and CD3 depletion. Derivative CD3 negative CBMCs were seeded in NK MACS medium at 500,000 / ml with 250IU / ml human IL-2, and EBV-LCL feeders at CBMC: feeder ratio 1 :2.5 and cultured in 37°C incubator with 5% CO2. IL-2 was spiked in the medium at day 3 of culture. CISH gene was knocked out using LNP on day 5, and the sclL-15 gene (SEQ ID NO: 1 ) was knocked-in by electroporation on day 7 using Tc Buster Transposon nanoplasmid and Transposase enzyme. Cells were then restimulated with EBV-LCL feeders at 4:1 EBV:CBMC ratio at day 9, along with 200nM methotrexate for selection and 250IU / ml IL-2. Cells were then expanded in NK MACS for total of 19 days after which cells were harvested and used for tumor killing assays.

[0148] CISH knockout by CRISPR / Cas9

[0149] CBMC from day 5 of culture were washed with CTS-Optmizer and resuspended at 10 million cells / ml CTS-Optmizer. CISH gene knockout was done by incubating CBMC cells with 91 mer CISH guide encapsulated in a LNP at a concentration of 500,000 cells / ml with 1.25ug / ml CISH LNP and 2.5ug / ml ApoE3 in CTS-OpTmizer containing 250IU / ml human IL-2 and 5ng / ml human IL-15. sclL-15 knock-in using Tc Buster Transposase

[0150] CBMC from day 7 of culture were washed with Opti-MEM medium. 20 million cells were resuspended in 95ul cold Opti-MEM, and 2.5ug of sclL-15 transposon plasmid and 2.5ug of TcBuster transposase were added to the cells. Total 10Oul reaction were added to OC100x2 cuvette and electroporated using MaxCyte NK-6 program. The cells were then recovered in NK MACS medium with 250IU / ml IL-2 and 5ng / ml IL-15.

[0151] NK cell killing assay using 3D spheroid model

[0152] Cord blood-derived NK cells (CB-NKs) were isolated for use in wildtype form or for use in the presence of a CISH gene knockout and / or sclL-15 gene knock-in.

[0153] The cytotoxicity of wildtype (WT) or CISH KO NK cells alone, or with the addition of sclL-15 knock-in, against ZR-75 GFP breast cancer cells or SKOV-3 GFP ovarian cancer cells was evaluated in the following experimental setup. Human breast cancer or ovarian cancer target cells were plated in a 96 well eSight plate and allowed to adhere for 3 days before the addition of fresh NK cells at 10:1 or 3:1 effector : target (E:T) ratio for 120 hours. The change in tumour growth was estimated by GFP expression using the Incucyte Spheroid module. The experiments were conducted using 1 ) Cancer target cells alone (ZR-75 GFP or SKOV-3 GFP), 2) WT NK cells alone, 3) WT NK cells + sclL-15 KI, 4) CISH KO NK cells alone and 5) CISH KO NK cells + sclL-15 KI.

[0154] As can be seen from Figure 1 , CISH KO NK cells or NK cells with sclL-15 KI alone exhibited limited cytotoxicity in the ZR-75 human breast cancer model, with a steady increase in cancer cell GFP signal detected in both groups. At both 10:1 and 3:1 E:T ratios, the CISH KO + sclL-15 KI NK cells demonstrated potent cytotoxicity against breast cancer, when compared to CISH KO NK cells or sclL-15 KI NK cells alone.

[0155] As seen from Figure 2, a similar effect was observed in the SKOV-3 human ovarian breast cancer model. No difference in cytotoxic effect was detected between WT NK cells, NK cells with sclL-15 KI, or CISH KO NK cells alone. On the other hand, the CISH KO + sclL-15 KI NK cells demonstrated potent cytotoxicity.

[0156] These data demonstrate that a genetic knockout of CISH works synergistically with a genetic knock-in of sclL-15, resulting in an enhanced cytotoxic phenotype against cancer cells.

[0157] Example 6 - CISH and TGFBRII Double KO with sclL-15 KI NK cells are highly effective in mediating cancer cell killing

[0158] NK cell expansion

[0159] Fresh cord blood was processed using Lymphoprep followed by ACK lysis and CD3 depletion. Derivative CD3 negative CBMCs were seeded in NK MACS medium at 500,000 / ml with 250IU / ml human IL-2, and EBV-LCL feeders at CBMC: feeder ratio 1 :2.5, and cultured in 37°C incubator with 5% CO2. IL-2 was spiked in the medium at day 3 of culture. CISH gene was knocked out using LNP on day 5, TGF|3RII gene was knocked out using LNP on day 6, and the sclL-15 gene was knocked-in by electroporation on day 7 using Tc Buster Transposon nanoplasmid and Transposase enzyme. Cells were then restimulated with EBV-LCL feeders at 4:1 EBV:CBMC ratio at day 9, along with 200nM methotrexate for selection and 250IU / ml IL-2. Cells were then expanded in NK MACS for total of 19 days which cells were harvested and used for tumor killing assays

[0160] CISH and TGFBRII knockout by CRISPR / Cas9

[0161] CBMC from day 5 of culture were washed with CTS-Optmizer and resuspended at 10 million cells / ml CTS-Optmizer. CISH gene knockout was done by incubating CBMC cells with 91 mer CISH guide encapsulated in an LNP at a concentration of 500,000 cells / ml with 1.25ug / ml CISH LNP and 2.5ug / ml ApoE3 in CTS-OpTmizer containing 250IU / ml human IL-2 and 5ng / ml human IL-15.

[0162] On day 6, TGF|3RII gene knockout was achieved by incubating CBMC cells with TGF[3RII LNP and ApoE3 at concentration of 500,000 cells / ml with 1.25ug / ml TGFBRII LNP and 2.5ug / ml ApoE3 in CTS-OpTmizer containing 250IU / ml human IL- 2 and 5ng / ml human IL-15. sclL-15 knock-in using Tc Buster Transposase system

[0163] CBMC from day 7 of culture were washed with Opti-MEM medium. 20 million cells were resuspended in 95ul cold Opti-MEM, and 2.5ug of sclL-15 transposon transposon plasmid and 2.5ug of TcBuster transposase were added to the cells. Total 100ul reaction were added to OC100x2 cuvette and electroporated using MaxCyte NK-6 program. The cells were then recovered in NK MACs medium with 250IU / ml IL-2 and 5ng / ml IL-15.

[0164] NK cell killing assay using 3D spheroid model

[0165] Cord blood-derived NK cells (CB-NK cells) were isolated for use in wildtype form or for use in the presence of a CISH and / or TGFBRII knockout with sclL-15 knock-in.

[0166] The cytotoxicity of wildtype (WT), CISH KO NK cells, TFGBRII KO NK cells, and CISH + TGFBRII double KO (DKO) NK cells with the addition of sclL-15 knock-in against SKOV-3 GFP ovarian cancer target cells was evaluated in the following experimental setup. Human ovarian cancer target cells were plated in a 96 well eSight plate and allowed to adhere for 3 days before the addition of fresh NK cells at 10:1 or 3:1 effector : target (E:T) ratios for 120 hours in the presence of TGFB1. The change in tumour growth was estimated by GFP expression using the Incucyte Spheroid module. The experiments were conducted using 1 ) Cancer target cells alone (SKOV-3 GFP), 2) WT NK cells with sclL-15 KI, 3) CISH KO NK cells + sclL-15 KI, 4) TGFBRII KO NK cells with sclL-15 KI, and 5) CISH + TGFBRII DKO NK cells with sclL-15 KI.

[0167] As can be seen from Figure 3, NK cells with the CISH + TGFBRII DKO and the scIL- 15 KI were highly effective in mediating SKOV-3 cancer cell killing, in comparison to WT NK with sclL-15 KI, as well as both CISH or TGFBRII KO NK cells with sclL-15 KI at both 10:1 and 3:1 ratios.

[0168] These data demonstrate CISH + TGFBRII DKO CB-NK cells with sclL-15 knocked in result in an enhanced cytotoxic phenotype against cancer cells.

[0169] Example 7 - Insertion of sclL-15 gene into CISH gene locus increases NK cell expansion

[0170] NK cell expansion

[0171] Fresh cord blood was processed using Lymphoprep followed by ACK lysis and CD3 depletion. Derivative CD3 negative CBMCs were seeded in NK MACs medium at 500,000 / ml with 250IU / ml human IL-2, and EBV-LCL feeders at CBMC: feeder ratio 1 :2.5, and cultured in 37°C incubator with 5% CO2. IL-2 was spiked in the medium at day 3 of culture. CISH gene knockout and sclL-15 knock-in using GenWand DNA was done in one step on day 5. The peptide sequence of sclL-15 corresponded to amino acids 49-162 of the full-length sequence (SEQ ID NO: 2), i.e., having the signal peptide removed. Cells were then restimulated with EBV-LCL feeders at 4:1 EBV:CBMC ratio at day 9, along with 200nM methotrexate for selection and 250IU / ml IL-2. Cells were then expanded in NK MACs for a total of 18 days.

[0172] CISH knockout and sclL-15 knock-in using GenWand

[0173] 10 million CBMC from day 5 of culture were washed with Opti-MEM and resuspended in 95ul Opti-MEM medium. 21.3ug 91 mer CISH gRNA, 20ug Cas9, and 10ug sclL-15 GenWand DNA were added to the cells, Total 10Oul reaction were added to OC 100x2 cuvette and electroporated using MaxCyte NK-6 program. The cells were then recovered in NK MACS medium with 250IU / ml IL-2 and 5ng / ml IL-15.

[0174] NK cell expansion assay

[0175] Cord blood-derived NK cells were isolated for use with an sclL-15 knock-in or for use in the presence of a CISH gene knockout with sclL-15 knock-in. The expansion of the wildtype NK cells with extra-genomic sclL-15 (WT_GW_sclL- 15) or CISH KO NK cells with sclL-15 genomically inserted into the CISH locus (CISHK0_GW_sclL15) was evaluated. The cells were collected at different days through the expansion and the total cell count was counted using the NC-3000.

[0176] As can be seen from Figure 4, the CISH KO NK cells with sclL-15 genomic insertion at the CISH locus resulted in significantly higher numbers of total cell count in comparison to the wildtype NK cells expressing extra-genomic sclL-15.

[0177] These data demonstrate that the genetic knock-in of the sclL-15 gene at the CISH gene locus is significantly more effective in increasing NK cell expansion than the use of extra-genomic expression of the gene.

[0178] Example 8 - Insertion of sclL-15 gene into CISH gene locus increases IL-15 Levels

[0179] NK cell expansion

[0180] Fresh cord blood was processed using Lymphoprep followed by ACK lysis and CD3 depletion. Derivative CD3 negative CBMCs were seeded in NK MACs medium at 500,000 / ml with 250IU / ml human IL-2, and EBV-LCL feeders at CBMC: feeder ratio 1 :2.5, and cultured in 37°C incubator with 5% CO2. IL-2 was spiked in the medium at day 3 of culture. CISH gene knockout and sclL-15 knock-in using GenWand DNA was done in one step on day 5. Cells were then restimulated with EBV-LCL feeders at 4:1 EBV:CBMC ratio at day 9, along with 200nM methotrexate for selection and 250IU / ml IL-2. Cells were then expanded in NK MACs for a total of 18 days.

[0181] CISH knockout and IL-15 knockin using GenWand

[0182] 10 million CBMC from day 5 of culture were washed with Opti-MEM and resuspended in 95ul Opti-MEM medium. 21.3ug 91 mer CISH gRNA, 20ug Cas9, and 10ug sclL-15 GenWand DNA were added to the cells, Total 10Oul reaction were added to OC 100x2 cuvette and electroporated using MaxCyte NK-6 program. The cells were then recovered in NK MACS medium with 250IU / ml IL-2 and 5ng / ml IL-15.

[0183] NK cell IL-15 expression assay Cord blood-derived NK cells were isolated for use in wildtype form or for use in the presence of a CISH gene knockout with or without sclL-15 knock-in.

[0184] The expression of sclL-15 in wildtype (WT) NK cells, CISH KO NK cells alone, or CISH KO NK cells with the addition of sclL-15 knock-in was evaluated in the following experimental setup. The supernatant was collected on Day 18 of NK cell expansion from the donor cord blood samples S-CB87 and S-CB88. The sclL-15 expression was measured using the ELISA MAX DELUXE Human IL-15 set and the concentration was calculated based on the standard curve provided by the kit. The experiments were conducted using 1 ) WT control NK cells, 2) CISH KO NK cells, 3) WT NK cells with extra-genomic sclL-15, 4) CISH KO NK cells with sclL-15 genomic insertion into the CISH locus.

[0185] As can be seen from Figure 5, NK cells with CISH KO alone failed to increase IL-15 levels when compared to the WT NK cells. In addition, WT NK cells with extra- genomic sclL-15 only mildly increased levels of IL-15 when compared with the WT NK cell. The CISH KO NK cells with sclL-15 insertion into the CISH locus greatly increased IL-15 levels, with both examples expressing more than 200 pg / ml of IL-15.

[0186] Interestingly, although CISH is a negative regulator of IL-15, NK cells with CISH KO alone did not show increased IL-15 levels compared to WT NK cells. These data also show that extra-genomic expression of sclL-15 only has mild effects on IL-15 levels, while the genetic insertion of sclL-15 into the CISH locus results in significantly enhanced IL-15 levels. Taken together, these data demonstrate that CISH KO alone is not sufficient to increase IL-15 levels significantly and the importance of having sclL-15 genomically inserted into the CISH locus.

[0187] Example 9 - CISH KO NK cells expressing an IL-15-IL-15Ra fusion protein are highly effective in mediating cancer cell killing

[0188] NK cell expansion

[0189] Fresh cord blood was processed using Lymphoprep followed by ACK lysis and CD3 depletion. Derivative CD3 negative CBMCs were seeded in NK MACS medium at 500,000 / ml with 250IU / ml human IL-2, and EBV-LCL feeders at CBMC: feeder ratio 1 :2.5 and cultured in 37°C incubator with 5% CO2. IL-2 was spiked in the medium at day 3 of culture. CISH gene was knocked out using LNP on day 5, and the various (sample-dependent) sclL-15 or IL-15-IL-15Ra genes were knocked-in by electroporation on day 7 using To Buster Transposon nanoplasmid and Transposase enzyme. Cells were then restimulated with EBV-LCL feeders at 4:1 EBV:CBMC ratio at day 9, along with 200nM methotrexate for selection and 250IU / ml IL-2. Cells were then expanded in NK MACS for a total of 19 days after which cells were harvested and used for tumor killing assays.

[0190] CISH knockout by CRISPR / Cas9

[0191] CBMC from day 5 of culture were washed with CTS-Optmizer and resuspended at 10 million cells / ml CTS-Optmizer. CISH gene knockout was done by incubating CBMC cells with 91 mer CISH guide encapsulated in a LNP at a concentration of 500,000 cells / ml with 1.25ug / ml CISH LNP and 2.5ug / ml ApoE3 in CTS-OpTmizer containing 250IU / ml human IL-2 and 5ng / ml human IL-15. sclL-15 or IL-15-IL-15Ra knock-in using Tc Buster Transposase

[0192] CBMC from day 7 of culture were washed with Opti-MEM medium. 20 million cells were resuspended in 95ul cold Opti-MEM, and 2.5ug of transposon plasmid and 2.5ug of TcBuster transposase were added to the cells. Total 10Oul reaction were added to OC 100x2 cuvette and electroporated using MaxCyte NK-6 program. The cells were then recovered in NK MACS medium with 250IU / ml IL-2 and 5ng / ml IL-15. The transposon plasmids used for the various CISH KO NK cell samples were as follows:

[0193] • MND-IL15-DHFR (sclL-15);

[0194] • EF1 a-lgE-l L15-DHFR (sclL-15);

[0195] • EF1a-lgE-IL15(N72D)-DHFR (sclL-15 comprising the N72D mutation);

[0196] • EF1a-lgE-IL15-IL15Ra-SUSHI-DHFR (IL-15-IL-15Ra fusion protein comprising the sushi domain); and

[0197] • EF1a-lgE-IL15-IL15Ra-DHFR (IL-15-IL-15Ra fusion protein lacking the sushi domain).

[0198] The nucleotide sequence encoding the IL-15-IL-15Ra fusion protein lacking the sushi domain is provided as SEQ ID NO: 3, with the amino acid sequence provided as SEQ ID NO: 4. These sequences include an IgE signal peptide (nucleotides 1-54; amino acids 1-18).

[0199] After expansion, supernatants were collected on day 15 for IL-15 detection. Samples were assayed at a 1 :1 dilution as per the instructions from ELISA MAX Deluxe Set Human IL-15, Lot: B390308. Normalization formulation: (IL-15 pg / m Unoriginal volume) / (groups of 10 million cells = total cell number / 10 million). For all CISH KO NK cell samples including a sclL-15 or IL-15-IL-15Ra knock-in, significantly higher levels of IL-15 were detected in the supernatant, when compared to the levels detected in wildtype (WT) NK cell and CISH KO NK cell sample supernatants.

[0200] NK cell killing assay using 3D spheroid model

[0201] Cord blood-derived NK cells (CB-NKs) were isolated for use in the following experimental setup, including samples as follows:

[0202] 1 . SKOV-3 cancer cells (no effector NK cells);

[0203] 2. Wildtype NK cells + SKOV-3 cancer cells;

[0204] 3. CISH KO NK cells + SKOV-3 cancer cells;

[0205] 4. CISH KO / sclL-15 KI NK cells + SKOV-3 cancer cells;

[0206] 5. CISH KO / sclL-15(N72D)KI NK cells + SKOV-3 cancer cells;

[0207] 6. CISH KO / IL-15-IL15Ra(sushi) KI NK cells + SKOV-3 cancer cells; and

[0208] 7. CISH KO / IL-15-IL15Ra(no sushi) KI NK cells + SKOV-3 cancer cells.

[0209] The cytotoxicity of each sample against SKOV-3 GFP ovarian cancer cells was evaluated. Human ovarian cancer target cells were plated in a 96 well eSight plate and allowed to adhere for 3 days before the addition of fresh NK cells at 10:1 or 3:1 effector: target (E:T) ratio for 168 hours. The change in tumour growth was estimated by GFP expression using the Incucyte Spheroid module.

[0210] As can be seen from Figure 6, wildtype NK cells or CISH KO NK cells alone exhibited limited cytotoxicity in the SKOV-3 human ovarian cancer model, when compared to the control containing SKOV-3 cancer cells alone (no effector NK cells). Meanwhile, at both 10:1 and 3:1 E:T ratios, the other samples (samples 4-7 above) demonstrated potent cytotoxicity against ovarian cancer, when compared to their CISH KO NK cell control; in all cases substantially all cancer cells being killed within the 168-hour period.

[0211] Perhaps most surprising was the consistently potent cytotoxicity observed for sample 7 (CISH KO NK cells expressing the IL-15-IL-15Ra fusion protein lacking the sushi domain), with all cancer cells being killed in as little as 24 hours (see e.g., S-CB-111 ; E:T 10:1 in Figure 6). In all four experiments illustrated in Figure 6, these modified NK cells were consistently more potent in killing cancer cells than any of the other modified NK cells tested, including CISH KO NK cells expressing the IL-15-IL-15Ra fusion protein with the sushi domain (sample 6). This was completely unexpected, considering the general teaching in the field that the sushi domain is of fundamental importance for achieving the advantages of IL-15 signaling in NK cells.

[0212] The invention thus provides CISH KO NK cells with an IL-15-IL-15Ra fusion protein knock-in, as well as uses of these NK cells in cancer therapy.

[0213] Sequence Listing

[0214] SEQ ID NO: 1 (DNA sequence encoding full-length sclL-15 peptide)

[0215] ATGCGAATCTCAAAGCCCCACCTGAGAAGCATCTCCATCCAGTGCTACCTGTGC

[0216] CTGCTGCTGAATAGCCACTTTCTGACAGAGGCCGGCATCCATGTGTTCATCCTG

[0217] GGATGTTTTAGCGCCGGACTGCCTAAGACCGAGGCCAACTGGGTGAACGTGAT

[0218] TAGTGATCTGAAGAAGATCGAGGACCTGATCCAGAGCATGCACATCGACGCTAC

[0219] ACTGTACACCGAATCTGATGTGCACCCTTCTTGTAAAGTGACAGCCATGAAGTG

[0220] CTTCCTGCTCGAGCTGCAAGTGATCAGCCTGGAAAGCGGCGACGCCTCTATCC

[0221] ACGACACCGTCGAAAACCTGATCATCCTGGCCAACAACAGCCTGTCTAGCAACG

[0222] GCAATGTGACTGAATCCGGATGCAAGGAGTGCGAGGAACTGGAAGAGAAAAAT

[0223] ATCAAGGAATTCCTTCAAAGCTTCGTGCATATCGTGCAGATGTTCATCAACACCA GC

[0224] SEQ ID NO: 2 (Amino acid seguence of full-length sclL-15 peptide)

[0225] MRISKPHLRSISIQCYLCLLLNSHFLTEAGIHVFILGCFSAGLPKTEANWVNVISDLKKI

[0226] EDLIQSMHIDATLYTESDVHPSCKVTAMKCFLLELQVISLESGDASIHDTVENLIILAN

[0227] NSLSSNGNVTESGCKECEELEEKNIKEFLQSFVHIVQMFINTS

[0228] SEQ ID NO: 3 (DNA seguence encoding IL-15-IL-15Ra fusion protein without the sushi domain and including the IgE signal peptide)

[0229] ATGGACTGGACCTGGATCCTGTTTCTGGTGGCCGCTGCTACAAGAGTGCATTCT

[0230] AACTGGGTGAACGTGATCAGCGATCTGAAAAAGATCGAGGACCTGATCCAGTC

[0231] CATGCACATCGATGCCACGCTGTACACCGAGTCGGACGTGCATCCTAGCTGCA

[0232] AGGTGACAGCCATGAAGTGCTTCCTGCTGGAACTGCAGGTGATTTCTCTGGAAA

[0233] GCGGCGACGCCAGCATCCACGACACCGTGGAAAACCTGATCATCCTGGCTAAT

[0234] AACAGCCTGAGCTCCAATGGCAACGTGACCGAATCTGGTTGTAAAGAGTGCGA

[0235] GGAGCTGGAGGAGAAAAACATCAAGGAGTTCCTTCAGAGCTTCGTGCACATCG

[0236] TGCAGATGTTCATCAATACATCTAGCGGAGGAGGCTCCGGCGGCGGCGGCAGC

[0237] GGCGGCGGAGGTTCCGGCGGAGGCGGCAGCGGCGGAGGATCCCTGCAAATC

[0238] ACCCGGGACCCTGCTCTAGTGCACCAGCGGCCTGCCCCTCCAAGCACCGTCAC

[0239] AACCGCCGGCGTGACCCCACAACCCGAAAGCCTGAGCCCCTCCGGAAAGGAA

[0240] CCCGCCGCTTCCAGCCCTTCGAGCAACAACACCGCCGCTACAACCGCCGCCAT

[0241] CGTCCCCGGCAGCCAGCTGATGCCTAGCAAAAGCCCGAGCACAGGCACCACC

[0242] GAGATCTCTAGCCACGAGAGCTCTCACGGCACACCTAGCCAGACAACCGCCAA GAACTGGGAGCTGACAGCCAGCGCCAGCCACCAGCCTCCAGGAGTGTACCCT

[0243] CAGGGCCACAGCGATACAACAGTGGCCATCAGCACCAGCACCGTGCTGCTGTG

[0244] CGGCCTGAGCGCCGTGAGCCTCCTGGCATGCTACCTCAAGTCTCGCCAGACCC

[0245] CTCCTCTGGCCAGCGTCGAGATGGAAGCCATGGAAGCCCTGCCTGTGACCTGG GGCACCTCCAGCAGGGACGAGGATCTGGAAAACTGTAGCCACCACCTG

[0246] SEQ ID NO: 4 (Amino acid sequence of IL-15-I L-15Ra fusion protein without the sushi domain and including the IqE signal peptide)

[0247] MDWTWILFLVAAATRVHSNVWNVISDLKKIEDLIQSMHIDATLYTESDVHPSCKVTA MKCFLLELQVISLESGDASIHDTVENLIILANNSLSSNGNVTESGCKECEELEEKNIKE

[0248] FLQSFVHIVQMFINTSSGGGSGGGGSGGGGSGGGGSGGGSLQITRDPALVHQRP

[0249] APPSTVTTAGVTPQPESLSPSGKEPAASSPSSNNTAATTAAIVPGSQLMPSKSPST

[0250] GTTEISSHESSHGTPSQTTAKNWELTASASHQPPGVYPQGHSDTTVAISTSTVLLC

[0251] GLSAVSLLACYLKSRQTPPLASVEMEAMEALPVTWGTSSRDEDLENCSHHL

Claims

Claims1. A natural killer (NK) cell or NK cell line that has been modified to have expression of CISH knocked out and to express an IL-15-IL-15Ra fusion protein, wherein the IL-15Ra lacks a functional sushi domain.

2. An NK cell or NK cell line according to claim 1 , wherein the NK cell or NK cell line is human.

3. An NK cell or NK cell line according to either claim 1 or claim 2, wherein the NK cell or NK cell line is derived from either human cord blood or human peripheral blood.

4. An NK cell or NK cell line according to any preceding claim, wherein the NK cell or NK cell line is derived from human cord blood.

5. An NK cell or NK cell line according to any preceding claim, wherein the NK cell or NK cell line is derived from human peripheral blood.

6. An NK cell or NK cell line according to any preceding claim, wherein the fusion protein is at least partially secreted by the NK cell or NK cell line.

7. An NK cell or NK cell line according to any preceding claim, further modified to have reduced expression of TGFBR2.

8. An NK cell or NK cell line according to any preceding claim, wherein the IL-15- IL-15Ra fusion protein comprises an IgE signal peptide.

9. An NK cell or NK cell line according to claim 8, wherein the IgE signal peptide is cleavable.

10. An NK cell or NK cell line according to any preceding claim, wherein the IL-15- IL-15Ra fusion protein comprises a N72D mutation.

11. An NK cell or NK cell line according to any preceding claim, wherein expression of TGFBR2 is knocked out.

12. An NK cell or NK cell line according to any preceding claim, further modified to express a chimeric antigen receptor (CAR) that binds a target selected from CD38, HER2, VEGF, EpCAM and EGFR.

13. An NK cell or NK cell line according to claim 12, wherein the CAR binds CD38.

14. An NK cell or NK cell line according to claim 12, wherein the CAR binds HER2.

15. An NK cell or NK cell line according to claim 12, wherein the CAR binds VEGF.

16. An NK cell or NK cell line according to claim 12, wherein the CAR binds EpCAM.

17. An NK cell or NK cell line according to claim 12, wherein the CAR bindsEGFR.

18. An NK cell or NK cell line according to any preceding claim, wherein a gene encoding the IL-15-IL-15Ra fusion protein is expressed at the CISH locus, thereby knocking out expression of CISH.

19. An NK cell or NK cell line according to any preceding claim, wherein the IL-15 portion of the fusion protein is cleavable, such that soluble IL-15 is secreted.

20. An NK cell or NK cell line according to any preceding claim, for use in treating cancer.21 . An NK cell or NK cell line for use according to claim 20, wherein the cancer is a solid cancer.

22. An NK cell or NK cell line for use according to claim 21 , wherein the solid cancer is breast cancer.

23. An NK cell or NK cell line for use according to claim 21 , wherein the solid cancer is ovarian cancer.

Images