NKG2D-IG fusion protein for cancer immunotherapy

Abstract

This application relates to NKG2D-IG fusion proteins for cancer immunotherapy. The present invention provides methods and compositions for cancer immunotherapy. The methods involve using chimeric molecules (such as fusion proteins) comprising a dimeric NKG2D portion and an Fc portion, which chimeric molecules bind to one or more NKG2D ligands. In some embodiments, the molecule further comprises a drug portion (such as an IL15 / Ra portion). The methods disclosed herein are applicable to treating cancers associated with the aberrant expression of one or more NKG2D ligands.

Description

[0001] This application is a divisional application of the invention patent application with the application date of November 11, 2016, application number “201680071013.5” and name “NKG2D-IG fusion protein for cancer immunotherapy”.

[0002] Related applications

[0003] This application claims priority under 35 U.S.C. §119(e) to U.S. Provisional Application No. 62 / 255,016, filed on November 13, 2015, the contents of which are incorporated herein by reference in their entirety. Technical Field

[0004] The present invention generally relates to NKG2D-IG fusion proteins for use in cancer immunotherapy. Background Art

[0005] NKG2D is a type II transmembrane glycoprotein with an extracellular lectin-like domain. This domain lacks the recognizable calcium-binding site found in true C-type lectins and binds protein rather than carbohydrate ligands. NKG2D is an activating receptor expressed on various immune cells. Human NKG2D is expressed on CD8+ αβT cells, γδT cells, NK cells, and NKT cells. In mouse systems, NKG2D is also found on macrophages. Human ligands for NKG2D include MHC class I chain-related molecules (MICA and MICB), UL16-binding proteins (ULBP1, ULBP2, ULBP3, and ULBP4), and RAET-1G; and mouse ligands for NKG2D include small amounts of histocompatibility antigen 60 (H60) and retinoic acid-inducible early transcript (RAE-1). Expression of NKG2D ligands also occurs in intestinal epithelial cells, tumor cells, and under conditions of stress or infection.

[0006] NKG2D exists as a disulfide-linked homodimer that transmits activation signals upon ligand binding. Signaling requires association with an adaptor protein. Alternative splicing of NKG2D mRNA produces isoforms with distinct cytoplasmic domains, which can associate with DAP12 to deliver the true activation signal or with DAP10 to generate a costimulatory signal. NKG2D has been implicated in immune surveillance and responses to viral infections. Furthermore, elevated levels of NKG2D ligands have been detected in proliferating cells and many types of cancer.

[0007] Certain NKG2D-Fc chimeras and their uses have been previously disclosed, for example, in published PCT application WO / 2010 / 080124, the entire contents of which are incorporated herein by reference. Summary of the Invention

[0008] In the present disclosure, novel compositions and methods for cancer therapy are provided. The invention is based, at least in part, on the surprising discovery that chimeric molecules comprising two NKG2D fragments and an Fc fragment (e.g., dimeric NKG2D-Fc chimeras) can bind to one or more NKG2D ligands and induce tumor cell death with improved efficacy compared to chimeric molecules comprising a single NKG2D fragment and an Fc fragment (e.g., monomeric NKG2D-Fc chimeras). In some embodiments, the dimeric NKG2D-Fc chimeras described herein bind to NKG2D ligands with increased affinity compared to monomeric NKG2D-Fc chimeras. In some embodiments, the affinity is increased by 2-fold, 5-fold, 10-fold, 100-fold, or 1000-fold.

[0009] Thus, in some aspects, the present disclosure provides a dimeric NKG2D-Fc chimera comprising: NKG2D1-NKG2D2-Fc, wherein NKG2D1 and NKG2D2 each comprise NKG2D or a fragment thereof and can bind to an NKG2D ligand; and the Fc comprises a fragment crystallizable region of an immunoglobulin (Fc). In some aspects, the present disclosure provides a composition comprising a dimeric NKG2D-Fc chimera as described herein and a pharmaceutically acceptable carrier.

[0010] In some embodiments, the dimeric NKG2D-Fc chimera further comprises a drug moiety. In some embodiments, the drug moiety is attached to the amino terminus or the carboxyl terminus of the chimera. In some embodiments, the drug moiety is attached to the carboxyl terminus of the chimera.

[0011] In some embodiments, the dimeric NKG2D-Fc chimera further comprises at least one linker molecule, wherein the at least one linker molecule is not a contiguous portion of NKG2D1, NKG2D2, Fc, or the drug moiety and covalently links: the amino acids of NKG2D1 to the amino acids of NKG2D2; the amino acids of NKG2D2 to the amino acids of Fc; or the amino acids of Fc to the drug moiety.

[0012] In some embodiments, at least one linking molecule is a peptide linker. In some embodiments, the length of the peptide linker ranges from about 2 to about 25 amino acids. In some embodiments, at least one linking molecule is a glycine-serine linker. In some embodiments, the glycine-serine linker is represented by the formula (GS) n wherein n is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12. In some embodiments, the glycine-serine linker is represented by the formula (GGGGS) n (SEQ ID NO: 2), wherein n is 1, 2, 3, 4 or 5.

[0013] In some embodiments, the chimera comprises three linker molecules, X1, X2, and X3, wherein X1 covalently links the amino acids of NKG2D1 to the amino acids of NKGD2; X2 covalently links the amino acids of NKG2D2 to the amino acids of Fc; and X3 covalently links the amino acids of Fc to the drug moiety. In some embodiments, X1 is (GS)3 (SEQ ID NO: 4) and X2, X3, and X4 are each (GGGGS)4 (SEQ ID NO: 3).

[0014] In some embodiments, the NKG2D fragment comprises an extracellular fragment of NKG2D. In some embodiments, the extracellular fragment of NKG2D is represented by SEQ ID NO: 1.

[0015] In some embodiments, the Fc comprises a fragment crystallizable region (Fc) of a human immunoglobulin (IgG). In some embodiments, the human immunoglobulin is IgG1.

[0016] In some aspects, the present disclosure provides methods for treating cancer comprising administering a dimeric NKG2D-Fc chimera as described in this document to an individual having an NKG2D ligand-expressing cancer in an amount effective to treat the cancer.

[0017] In some embodiments, the NKG2D ligand-expressing cancer is melanoma, lung cancer, plasma cell cancer, leukemia, lymphoma, ovarian cancer, colon cancer, pancreatic cancer, or prostate cancer. In some cases, one or more of these cancers may be present in an individual.

[0018] In some embodiments, the method further comprises treating the individual with other anti-cancer therapies. In some embodiments, the other anti-cancer therapies are selected from the group consisting of surgery, radiation therapy, chemotherapy, gene therapy, DNA therapy, viral therapy, RNA therapy, adjuvant therapy, and immunotherapy.

[0019] In some embodiments, the additional cancer therapy is a DNA damaging chemotherapy.

[0020] In some embodiments, the NKG2D ligand is MICA, MICB, ULBP1, ULBP2, ULBP3, ULBP4, ULBP5, or ULBP6.

[0021] Each of the limitations of the present invention can encompass various embodiments of the present invention. Therefore, it is contemplated that each of the limitations of the present invention relating to any one element or combination of elements can be included in each aspect of the present invention. The present invention is not limited to the details of the construction and arrangement of the components set forth in the following description or illustrated in the drawings in its application. The present invention can have other embodiments and be put into practice or carried out in various ways. In addition, the wording and terminology used herein are for descriptive purposes and should not be considered as restrictive. The use of "including," "comprising," or "having," "containing," "involving," and variations thereof herein is intended to encompass the items listed thereafter and their equivalents and additional items. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Schematics showing the dimeric NKG2D-Fc chimera with (right) and without (left) the drug moiety.

[0023] Figure 2 hNKG2Dx2-hIgG1-hIL15 / Ra was shown to be produced as a single fusion protein and purified by protein A.

[0024] Figure 3 hNKG2Dx2-hIgG1-IL15 / Ra promotes the proliferation of human NK cells similar to IL-15.

[0025] Figure 4 hNKG2Dx2-hIgG1-IL15 / Ra is shown to promote effective killing of multiple cell lines and outperform hNKG2Dx2-hIgG1 in cell lines with moderate ligand expression. Panel A shows that neither construct promoted killing of the B16 tumor cell line, which does not express NKG2D-L. Panel B shows that both constructs equally promoted killing of a synthetic B16 tumor cell line that expresses high levels of NKG2D ligand. Panel C shows that various tumors express varying levels of NKG2D ligand on their cell surfaces, as measured by NKG2D fusion protein binding. Panel D shows that hNKG2Dx2-hIgG1-IL15 / Ra kills cells expressing moderate NKG2D ligand more effectively than hNKG2Dx2-hIgG1.

[0026] Figure 5 Resting NK cells were shown to be activated by the fusion protein to produce IFN-γ, but maximal productivity required all three components: NKG2D, hIgG1 and IL- 15. N297Q is a mutation in hIgG1 that prevents CD16 (expressed on NK) from binding to hIgG1.

[0027] Figure 6 showed that preactivated NK cells required CD16 binding but not IL-15 to kill target cells.

[0028] Figure 7 demonstrated that optimal activation of resting NK cells and killing by resting NK cells required CD16 engagement and IL-15 activation.

[0029] Figure 8 Presented are ELISA analyses demonstrating that NKG2Dx2-hlgG1 binds to MICA*008 with increased affinity compared to hNKG2Dx1-hlgG1 which is a monomeric NKG2D-Fc chimera.

[0030] Figure 9 Presented are flow cytometry analyses demonstrating that hNKG2Dx2-hlgG1 binds to NKG2D ligand-expressing cells with increased avidity compared to hNKG2Dx1-hlgG1.

[0031] Figure 10 NKG2D-Fc drives NK cell killing of ligand-positive targets. Dimeric NKG2D-Fc chimeras mediate killing more efficiently than monomeric NKG2D-Fc chimeras. *depicts p<0.5 and **depicts p<0.01.

[0032] Figure 11 Dimeric NKG2D-Fc chimeras (eg, hNKG2Dx2-hlgG1) were shown to kill NKG2D ligand-expressing cells more efficiently than monomeric NKG2D-Fc chimeras (eg, hNKG2Dx1-hlgG1). **depicts p<0.01.

[0033] Figure 12 NKG2Dx2-hIgG1 exhibited superior neutralization of soluble MICA compared to NKG2Dx1-hIgG1. * depicts p < 0.05; ** depicts p < 0.01; *** depicts p < 0.005; and **** depicts p < 0.001.

[0034] Figure 13 A structural model of dimeric hNKG2D-hlgG1 in complex with human MICA (hMICA) is shown. (G4S)4 is SEQ ID NO: 3; GGSGGGSG is SEQ ID NO: 5. DETAILED DESCRIPTION

[0035] Disclosed herein are novel compositions and methods for cancer immunotherapy. The compositions and methods of the present invention are based, at least in part, on the surprising discovery that chimeric molecules comprising two NKG2D fragments and an Fc fragment (e.g., dimeric NKG2D-Fc chimeras) are capable of binding one or more NKG2D ligands and inducing tumor cell death with improved efficacy compared to chimeric molecules comprising a single NKG2D fragment and an Fc fragment (e.g., monomeric NKG2D-Fc chimeras).

[0036] Monomeric NKG2D-Fc chimeras described in the prior art (e.g., the constructs described in published PCT application WO / 2010 / 080124) exhibit low binding avidity (e.g., low binding avidity index) for NKG2D ligands. Compared to prior art monomeric constructs that provide multiple NKG2D receptors (or portions thereof) on the same molecule, the dimeric NKG2D-Fc constructs described herein provide increased binding avidity (e.g., an increased avidity index of at least 1.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more). Without wishing to be bound by any particular theory, it is believed that the presence of multiple NKG2D receptors on a single molecule increases the number and duration of NKG2D-NKG2D ligand binding interactions, thereby resulting in increased anti-tumor activity. In fact, as shown in the Examples section, the dimeric NKG2D-Fc chimeras exhibited up to 100-fold increased binding avidity compared to prior art monomeric NKG2D-Fc chimeras. However, the success of this approach is unpredictable because it is not known whether increasing the number of receptors (or portions thereof) in the chimeric construct will inhibit binding interactions (e.g., through steric hindrance) or cause aggregation of the chimera that could interfere with the stability of the molecule.

[0037] NKG2D ligands are known to be expressed on cancer cells. Therefore, in some embodiments, the present disclosure provides methods for cancer therapy in an individual (e.g., a human individual), comprising administering a dimeric NKG2D-Fc chimera as described herein to an individual suffering from a cancer expressing an NKG2D ligand. Unlike immunotherapies that employ monoclonal antibodies against NKG2D ligands (e.g., MICA), the methods provided herein are believed to have broad effects on cancers, given that NKG2D binds to a variety of ligands.

[0038] Dimeric NKG2D-Fc chimeras can target any or all NKG2D ligands expressed on human tumor cells and are therefore capable of mediating tumor cell destruction through complement lysis and ADCC. NKG2D-Fc chimeras are also capable of opsonizing any tumor cell expressing at least one NKG2D ligand. NKG2D-Fc chimeras can promote efficient cross-presentation (e.g., activation) by dendritic cells, leading to the induction of effective T cell responses against tumors. Furthermore, such chimeras can bind to and sequester any "shed" (e.g., soluble or released) NKG2D ligands produced by tumor cells, thereby alleviating immunosuppression by downregulating NKG2D expression in response to tumor-derived soluble ligands.

[0039] NKG2D-Fc

[0040] In some aspects, the present disclosure provides a dimeric NKG2D-Fc chimera comprising: NKG2D1-NKG2D2-Fc, wherein NKG2D1 and NKG2D2 each comprise NKG2D or a fragment thereof and can bind to an NKG2D ligand; and the Fc comprises a fragment crystallizable region (Fc) of an immunoglobulin. In some embodiments, the NKG2D fragment comprises an extracellular fragment of NKG2D. In some embodiments, the extracellular fragment of NKG2D is represented by SEQ ID NO: 1.

[0041] As used herein, a "dimeric NKG2D-Fc chimera" is a chimeric molecule comprising two NKG2D ligand binding sites, wherein each ligand binding site comprises at least a portion or all of the NKG2D receptor and is capable of binding to a NKG2D ligand. The ligand binding site is fused to the Fc fragment. In the Examples section and the figures, the two NKG2D ligand binding sites of the dimeric NKG2D-Fc chimera are also collectively referred to as "NKG2Dx2". The monomeric NKG2D-Fc chimera described in the prior art may be referred to as "NKG2Dx1". The terms "chimera", "chimeric molecule", etc. generally refer to molecules composed of parts from multiple sources or origins. In some embodiments, the dimeric NKG2D-Fc is produced in the form of a recombinant chimeric fusion protein.

[0042] In some embodiments, the dimeric NKG2D-Fc chimeras described herein bind to the NKG2D ligand with increased avidity compared to the monomeric NKG2D-Fc chimera. As used herein, "avidity" refers to the combined strength of multiple affinities across the individual non-covalent binding interactions between the ligand and the receptor. Methods for measuring binding affinity are known in the art and include, for example, ELISA, surface plasmon resonance analysis, CD analysis, fluorescence quenching, size exclusion binding analysis, and isothermal titration calorimetry. For a brief description of these analyses, see, for example, Lengyel et al., 2007, J. Biol. Chem., 282:30658-666. In some embodiments, binding affinity is determined by measuring an avidity index. In some embodiments, the binding affinity of the dimeric NKG2D-Fc chimera for the NKG2D ligand increases by between about 2-fold and about 2000-fold compared to the monomeric NKG2D-Fc chimera. In some embodiments, the binding affinity increases by between about 2-fold and 1000-fold. In some embodiments, the binding affinity increases by between about 2-fold and 100-fold. In some embodiments, the binding affinity increases by between about 5-fold and 1000-fold. In some embodiments, the binding affinity increases by between about 5-fold and 200-fold. In some embodiments, the binding affinity increases by between about 2-fold and about 20-fold. In some embodiments, the binding affinity increases by 2-fold, 5-fold, 10-fold, 100-fold, or 1000-fold. In some embodiments, the binding affinity increases by at least 2-fold, at least 5-fold, at least 10-fold, at least 100-fold, or at least 1000-fold. In some embodiments, the dimeric NKG2D-Fc construct has an improved binding affinity index compared to the monomeric NKG2D-Fc chimera, for example, an improved affinity index of at least 1.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more.

[0043] NKG2D

[0044] In some aspects, the present disclosure provides dimeric NKG2D-Fc chimeras comprising two NKG2Ds or fragments thereof. NKG2D, also known as KLRK1; killer cell lectin-like receptor subfamily K, member 1; CD314; KLR; NKG2-D; FLJ17759; FLJ75772 or D12S2489E, is one of the primary triggering receptors for NK cells and is well known in the art. See, for example, Garrity et al. (2005). The portion of the NKG2D receptor used in the dimeric NKG2D-Fc is based on the known sequence of NKG2D (e.g., Accession No. NP_031386) or a derivative thereof that binds at least one ligand. Derivatives of NKG2D that can be used in the compositions and methods of the present invention include, but are not limited to, NKG2D sequences containing one or more mutations, such as point mutations, substitutions, deletion mutations, and / or insertion mutations. Suitable derivatives of NKG2D can be readily determined by one skilled in the art based on the teachings of this disclosure and the knowledge available in the art. At the cDNA level, such mutations may be silent mutations. Alternatively, the mutation may result in a change in the corresponding amino acid residue. In the latter case, the change may constitute a conservative change, whereby the amino acid residue is replaced by another amino acid residue with similar characteristics. However, in some cases, the mutation may produce a non-conservative substitution. Such mutations are acceptable if they result in a dimeric NKG2D-Fc chimera capable of binding to the NKG2D ligand.

[0045] In some embodiments, each NKG2D portion of the dimeric NKG2D-Fc chimera is a full-length NKG2D polypeptide. The full-length sequence of NKG2D has been described in the literature. See, for example, Reference Sequence Accession Number: NP_031386. Additionally, alternatively spliced ​​variants of NKG2D have been described. For the purposes of the present invention, any of these alternatively spliced ​​variants may be used, provided that the resulting polypeptide is capable of binding its ligand when constructed as a dimeric NKG2D-Fc chimera.

[0046] In some embodiments, each NKG2D portion of the dimeric NKG2D-Fc chimera is a partial sequence (i.e., a fragment) of an NKG2D receptor polypeptide, provided that the resulting polypeptide retains the ability to bind its ligand when constructed as a dimeric NKG2D-Fc chimera. For example, each NKG2D portion of the dimeric NKG2D-Fc construct can be shortened by one or more amino acid residues from either end of the NKG2D sequence. More specifically, the N-terminus of the NKG2D sequence can be deleted by 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, about 30, about 40, about 50, about 60, about 70, about 80, or more residues. Similarly, the C-terminus of the NKG2D sequence may be deleted by 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more residues. In some embodiments, both the N-terminus and the C-terminus may be shortened as described.

[0047] It has been shown that the extracellular portion of NKG2D contributes to the formation of homodimers and forms the ligand-binding site. Therefore, it is possible to delete some or all of the intracellular portion of NKG2D and still maintain the ability to bind its ligand. For example, the dimeric NKG2D-Fc chimeras described in this disclosure may primarily contain the extracellular fragment of the NKG2D receptor. Structural analysis has revealed that amino acid residues 78 to 216 of the human NKG2D sequence correspond to the extracellular portion of NKG2D that contains the ligand-binding site. For the mouse counterpart, the extracellular domain is amino acid residues 78-232, 94-232, or 92-232.

[0048] Thus, in some embodiments, each NKG2D in the dimeric NKG2D-Fc construct comprises an extracellular portion of the NKG2D sequence, such as amino acid residues 78-216 of human NKG2D; 78-232, 94-232, or 92-232 of mouse NKG2D. In some embodiments, the dimeric NKG2D-Fc construct comprises a portion of the extracellular domain. Thus, the extracellular domain of the dimeric NKG2D-Fc construct can be shortened at the N-terminus, the C-terminus, or both. For example, the N-terminus of the extracellular domain used to generate dimeric NKG2D-Fc can be shortened by one or more amino acid residues, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, about 30, about 40, about 50, about 60, etc., relative to the entire extracellular portion of the polypeptide. The C-terminus of the extracellular domain used to generate NKG2D-Fc can be shortened by one or more amino acid residues relative to the entire extracellular portion of the polypeptide, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, about 30, about 40, about 50, about 60, etc. Using human NKG2D as an example, the dimeric NKG2D-Fc construct can contain a fragment of the extracellular domain wherein the N-terminus of the domain begins at amino acid residue 79, 80, 81, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, about 110, about 120, about 130, about 140, or about 150. Similarly, the dimeric NKG2D-Fc construct may contain a fragment of the extracellular domain wherein the C-terminus of the domain ends at amino acid residues 231, 230, 229, 228, 227, 226, 225, 224, 223, 222, 221, 220, 219, 218, 217, 216, 215, 214, 213, 212, 211, 210, 209, 208, 207, 206, 205, etc. Such deletions at each end of the extracellular domain of the NKG2D sequence may be combined.

[0049] The skilled artisan will recognize that the dimeric NKG2D-Fc chimeras described herein may comprise two identical NKG2D fragments or two different NKG2D fragments. For example, in some embodiments, the dimeric NKG2D-Fc chimera comprises two NKG2D fragments corresponding to amino acid residues 78 to 216 of human NKG2D. In some other embodiments, the dimeric NKG2D-Fc chimera comprises two NKG2D fragments, wherein the first fragment corresponds to amino acid residues 78 to 216 of human NKG2D and the second fragment corresponds to a different portion of the NKG2D extracellular domain (e.g., amino acid positions 140 to 210 of human NKG2D).

[0050] Also contemplated are dimeric NKG2D-Fc derivatives comprising one or more mutations in the NKG2D portion of the construct located at the NKG2D-ligand binding interface. Specifically, certain mutations are known to affect the binding affinity between the NKG2D receptor and its ligand (e.g., MICA). See, for example, Lengyel et al., 2007, J. Biol. Chem., 282:30658-666. The three-dimensional structure of the complex between NKG2D and MICA has been described. Therefore, one skilled in the art can identify amino acid residues in NKG2D that contribute to interaction with its ligand and test the effects of mutations by systematically varying key residues. In any embodiment, the resulting dimeric NKG2D-Fc chimera is capable of binding to the ligand. For a comprehensive review of amino acid residues involved in receptor-ligand contacts, see, for example, Strong and McFarland, 2004, Advances in Protein Chemistry, 68:281-213. Based on published studies, key residues believed to be important for interaction with ligands have been mapped to amino acid residues approximately 150 to 207 in human NKG2D, which correspond to residues approximately 166 to 223 in mouse NKG2D. Thus, each NKG2D fragment of the dimeric NKG2D-Fc construct of the present invention preferably comprises a fragment spanning at least a majority of these residues (e.g., residues 150 to 207 in human NKG2D). Likewise, it will be appreciated that conservative substitutions, deletions, or mutations outside of these regions may be readily tolerated in many cases.

[0051] Several amino acid residues have been identified as being particularly important for mediating ligand binding. Specifically, residues in human NKG2D that are important for binding to MICA include Y152, Q185, K197, Y199, E201, and N207. Residues in human NKG2D that are important for binding to ULBP3 include I182, Y199, and Y152. Residues in mouse NKG2D that are important for binding to RAE-1β include K166, Y168, Y215, K213, E217, and N223. Therefore, in preferred embodiments, most or all of these residues (corresponding to the dimeric NKG2D construct) are maintained, without mutation or deletion, at positions where broad tolerance (e.g., specificity) for multiple ligands is desired. However, it is also possible to design dimeric NKG2D-Fc constructs that preferentially bind one ligand to another by strategically introducing mutations at one or more of these key residues that confer selective ligand recognition and binding. On the other hand, certain amino acid residues are involved in the binding of various ligands. For example, Y152 and Y199 in human NKG2D, which are equivalent to Y168 and Y215 in the mouse counterparts, respectively, contribute to binding to MICA and ULBP3. Therefore, in some embodiments, these residues are not modified in order to retain broad ligand specificity.

[0052] The examples provided below present representative dimeric NKG2D-Fc chimeras, wherein each NKG2D fragment corresponds to amino acid residues 78 to 216 of human NKG2D. However, it will be appreciated that the same approach can be applied to NKG2D sequences from any other species known to suffer from cancer. For example, the NKG2D fragments of the dimeric NKG2D-Fc may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid changes, such as deletions, insertions, and substitutions, as long as the dimeric NKG2D-Fc retains its ligand binding activity.

[0053] The present invention includes variants of dimeric NKG2D-Fc constructs containing one or more amino acid changes as described above, with respect to the binding of the dimeric NKG2D-Fc chimera to one or more of its natural ligands. To determine whether a dimeric NKG2D-Fc variant containing a particular mutation retains ligand binding activity, a binding assay can be performed in which the binding affinity and / or binding capacity of a particular dimeric NKG2D-Fc chimera to its ligand can be assessed. Numerous methods are known in the art by which receptor-ligand interactions can be measured. These methods for analyzing ligand binding include, but are not limited to, ELISA, surface plasmon resonance analysis, CD analysis, fluorescence quenching, size exclusion binding assays, and isothermal titration calorimetry. For a brief description of these assays, see, for example, Lengyel et al. (2007).

[0054] Fc fragment

[0055] In some embodiments, the dimeric NKG2D-Fc chimera comprises a fragment crystallizable region (Fc) of an immunoglobulin. The Fc region of an immunoglobulin plays an important role in mediating immune defense. FcγRs are widely expressed as transmembrane glycoproteins on many cell types, including macrophages, NK cells, dendritic cells, B cells, neutrophils, and mast cells. Fc-mediated activities include the recruitment of effector cells through Fc-FcγR interactions. There are two types of Fc receptors that can be functionally distinguished: activating Fc receptors and inhibitory Fc receptors. Activating Fc receptors include human FcγRIA, FcγRIIA, and FcγRIIIA, as well as their mouse orthologs, i.e., FcγRI, FcγRIII, and FcγRIV. Activated FcγRs mediate ADCC and ADCP, induce internalization of immune complexes that lead to antigen presentation, and contribute to the production and release of cytokines and proinflammatory cytokines. For a general review of IgG structure and mechanism of action, see Liu et al. (2008; Immunological Reviews, 222:9-27). As described in more detail herein, the Fc portion of the dimeric NKG2D-Fc is a domain that binds to activating Fc receptors, preferably activating Fc Ig domains, and includes a hinge region that allows dimerization.

[0056] The Fc portion of the dimeric NKG2D chimeras suitable for use in the present disclosure can be readily adapted to exhibit species specificity. For use in mouse systems, such as cells derived from mice, the Fc fragment used to generate the dimeric NKG2D-Fc is preferably an Fc fragment of mouse origin. In some embodiments, the Fc fragment of mouse IgG2a is preferred.

[0057] For use in human subjects, e.g., for cancer treatment, the Fc fragment used to produce dimeric NKG2D-Fc is preferably an Fc fragment of human origin. In particularly preferred embodiments, the NKG2D-Fc comprises an activating Fc Ig domain. Of the four human IgG isotypes, the activating Fc domain of IgG1 is preferably used to prepare dimeric NKG2D-Fc. Therefore, in some embodiments, the Fc comprises a fragment crystallizable region (Fc) of a human immunoglobulin (IgG). In some embodiments, the human immunoglobulin is IgG1. Experimental data related to chimeric constructs containing the Fc region of human IgG1 are provided in the Examples section.

[0058] It is known in the art that different antibody isotypes have different degrees of in vivo cytotoxic potential (see, e.g., Nimmerjahn F. and Ravetch JV., 2006, Immunity, 24:19-28). For example, mouse IgG2a and IgG2b isotypes are more efficient at clearing infections such as bacterial and viral infections and killing tumor cells than their IgG1 or IgG3 counterparts. This can be attributed, at least in part, to the differential ratio of activating FcRs to inhibitory FcRs present in vivo. Similarly, relative to human IgG isotypes, IgG1 and IgG3 interact more strongly with FcRs than with IgG2 or IgG4. In addition, certain polymorphic allotypes of a given isotype can affect the affinity of Fc receptors. In fact, there are allelic variants of activating FcRs that will significantly affect the affinity of certain antibody isotypes. For example, the FcγRIIIa receptor 158V isotype displays higher affinity for human IgG1 and increased antibody-dependent cellular cytotoxicity (Cartron G. et al., 2002, Blood, 99:754-758).

[0059] Without wishing to be bound by any particular theory, it is possible to optimize the interaction between the Fc portion of the dimeric NKG2D-Fc chimera and its corresponding Fc receptor by strategically selecting or modifying the Fc alleles used to prepare the dimeric NKG2D-Fc chimera. Thus, the present invention encompasses the use of mutants or isoforms of the Fc fragment. Many useful mutations within the Fc domain have been described that affect the interaction of the Fc with its receptors, the effector functions of the Fc, and the half-life of Fc-containing molecules. These include specific amino acid substitutions and / or modifications of the carbohydrate portion of the Fc. For reviews, see, for example, Liu et al., 2008, Immunological Reviews, 222:9-27; Niemoyan and Reifke, 2007, Curr. Opin. Immunol., 19(2):239-45.

[0060] The structure of the Fc fragment is generally known in the art. Briefly, the Fc region of a typical IgG molecule is a symmetrical homodimer of the carboxyl-terminal portion of the heavy chain and consists of C H 2 and C HThe IgG molecule is composed of three domains, which are separated from Fab by a flexible hinge region. The Fc region is stabilized by non-covalent interactions between the domains. The Fc region interacts with FcR to exert effector functions or regulate the catabolism of IgG. The heavy constant region (Cγ2 and Cγ3) and the hinge region between the variable domains and the constant region interact with C1q and Fc receptors (FcR). Therefore, the heavy constant region of the IgG molecule is responsible for its effector function because it includes binding sites for complement and FcR on different effector cells. Therefore, the recruitment of effector cells is mediated by Fc-FcγR interactions.

[0061] In general, the interaction of antibodies with complement initiates complement-dependent cytotoxicity (CDC), and FcγR interactions mediate antibody-dependent cellular toxicity (ADCC) and antibody-dependent cellular phagocytosis (ADCP). The typical activation pathway of CDC is triggered when the first component C1 of the pathway binds to the hinge-Fc portion of IgG in the antigen-antibody complex. Subsequent activation of the complement cascade ultimately induces the formation of the C5-C9 membrane attack complex that causes target cell death. On the other hand, ADCC relies on the ability of FcγR-carrying cells of the innate immune system (such as NK cells, monocytes, macrophages, and granulocytes) to recognize the Fc domain of antibodies bound to target cells. This recognition triggers effector cells to release cytoplasmic perforins, granulysins, and granzymes that induce apoptosis and lysis of target cells. The main effector cells in ADCC are NK cells, which express types of FcγRs that recognize IgG1 and IgG3 subclasses and trigger cytotoxic effects in vivo.

[0062] In the context of the present invention, as illustrated in the Examples, the dimeric NKG2D-Fc chimeras described herein, which bind to a dimeric NKG2D portion that broadly but specifically recognizes and binds its ligand, are able to mediate equivalent cellular effects by virtue of having a functional Fc portion.

[0063] As mentioned, there are activating receptors (FcγRI, FcγRIIA and FcγRIII) and inhibitory (FcγRIIB) receptors. In general, the interaction of IgG with activating FcγR triggers cell activation, while the interaction with FcγRIIB inhibits cell activation. Except for B cells and NK cells, activating and inhibitory FcγRs are co-expressed on the same effector cells, thereby generating a threshold for cell activation. B cells only express inhibitory FcγRIIB and therefore cannot be activated by endogenous IgG under physiological conditions. NK cells express activating FcγRIII so that they can kill target cells independently of pre-activation (or activation).

[0064] FcγRIIA and FcγRIII (CD16) have lower affinity for monomeric IgG and are thought to be critical for triggering effector functions, resulting in anti-tumor activity. Therefore, it is possible to design a dimeric NKG2D-Fc such that it is genetically engineered to have increased affinity for the activating FcγRIII and reduced affinity for the inhibitory FcγRIIB.

[0065] Thus, amino acid residues of the dimeric NKG2D-Fc molecule that contribute to its direct interaction with FcγRs can be modified, primarily in the lower hinge region adjacent to the Cγ2 region, and such variants are encompassed by the present invention. It has been demonstrated that binding to FcγRs requires a region corresponding to amino acid residues 234-237 of IgG. Additionally, other residues that have been shown to be important for IgG-FcγR interactions are located in the Cγ2 domain and include Asp265, Asp270, Ala327, Pro329, and Lys338.

[0066] Several strategies are envisioned for generating dimeric NKG2D-Fc chimeras with enhanced activity. At least two approaches are envisioned for engineering dimeric NKG2D-Fc with enhanced ADCC capacity. First, based on amino acid residues in IgG1 identified as critical for binding to activating and inhibitory FcγRs, the present invention provides variants of dimeric NKG2D-Fc chimeras that respectively increase or decrease affinity for these receptors. Thus, in one embodiment, a triple amino acid substitution, Ser298Ala / Glu333Ala / Lys334Ala, is provided, with the position of each residue based on IgG1. A dimeric NKG2D-Fc containing this triple mutation should exhibit higher affinity for FcγRIIIA but less affinity for FcγRIIB, thereby promoting ADCC. Similarly, in another embodiment, a dimeric NKG2D-Fc variant containing the double mutation, Ser239Asp / Ile332Glu, in the Fc region is expected to result in enhanced ADCC. Other mutations for improving ADCC include, but are not limited to, Ser239Asp / Ala330Leu / Ile332Glu and Ser239Asp / Ser298Ala / Ile332Ala. Similarly, in some embodiments, it is contemplated that a combination of mutations that increase binding to FcγRIIIA (e.g., activating receptors) and decrease binding to FcγRIIB are contemplated. Without limitation (residue position based on IgG1), examples of such Fc mutations include Phe243Leu / Arg292Pro / Tyr300Leu / Val305Ile / Pro396Leu.

[0067] Based on the observation that some modifications significantly affect the affinity of Fc for FcγR, the second approach involves modifying the carbohydrate moiety in Fc. It has been shown that the Fc domain contains two asparagine N-linked oligosaccharide sites (reviewed in Liu et al., 2008). ADCC requires the presence of certain oligosaccharides and depends on changes in the oligosaccharide structure. Specifically, previous studies have shown that removing the fucose moiety attached to the innermost GlcNAc of the biantennary complex oligosaccharide can significantly improve ADCC by improving the binding of Fc to FcγRIIIa without compromising CDC activity. Based on this observation, in one embodiment, the present invention provides a fucose-deficient dimeric NKG2D-Fc. In some embodiments, the chimera completely lacks the fucose moiety (i.e., non-fucosylated). In other embodiments, the chimera is fucosylated.

[0068] To produce dimeric NKG2D-Fc containing modified carbohydrates, host cells can be engineered to express an enzyme that catalyzes the desired modification. For example, host cells such as Chinese hamster ovary (CHO) cells can be transfected with the enzyme β-(1,4)-N-acetylglucosaminyltransferase III (GnT-III), which increases the amount of bisected non-fucosylated oligosaccharides. NKG2D-Fc products produced by these host cells can exhibit significantly enhanced ADCC activity. Additionally, in some embodiments, the fucose content in NKG2D-Fc can be manipulated using α-1,6-fucosyltransferase (FUT8), which lacks nuclear fucosyltransferase activity. Alternatively, small interfering RNA can be used to constitutively inhibit the expression of the FUT8 enzyme to achieve the same effect. In some embodiments, host cells deficient in guanosine diphosphate (GDP)-mannose 4,6-dehydratase (GMD) can be used to produce non-fucosylated NKG2D-Fc.

[0069] Next, various mutations in the Fc domain are envisioned to engineer dimeric NKG2D-Fc with enhanced complement activity. Generally speaking, complement can be activated through at least three pathways, leading to the formation of the membrane-attached complex C5b-9, which forms pores in the plasma membrane of target cells and causes their lysis. Binding of C1q to the Fc domain is a key step in this approach. Among human IgG subclasses, only IgG1 and IgG3 can initiate the complement cascade. In some embodiments, mutations are introduced into the Fc domain of dimeric NKG2D-Fc to promote C1q recruitment and C1q-Fc interactions. Fc residues targeted for such mutations include, but are not limited to, Asp270, Lys322, Pro329, and Pro331. These mutations involve replacing the corresponding residues with non-polar, neutral amino acids, such as Ala, Met, or Trp. In a specific embodiment, the dimeric NKG2D-Fc contains mutations Lys326Trp, Clu333Ser, or both.

[0070] To achieve increased C1q binding and improved CDC, some embodiments of the present invention involve introducing one or more mutations into certain residues in the hinge region of human IgG1. Non-limiting examples of such mutations include: Lys222Trp / Thr223Trp, Cys220Asp / Asp221Cys, Cys220Asp / Asp221Cys / Lys222Trp / Thr223Trp, Lys222Trp / Thr223Trp / His224Trp, and Asp221Trp / Lys222Trp.

[0071] In addition, it should be noted that when fusion proteins with artificial sequences and activities are used as therapeutic agents, in some cases, patients treated with such fusion proteins trigger undesirable immune responses, such as the development of antibodies to the agent. Certain structural modifications of the Fc fragment have been shown to reduce the immunogenicity of therapeutic fusion proteins. See, for example, U.S. Patent No. 6,992,174B2 to Gillies et al., which is incorporated herein by reference; Liu et al., 2008, Immunological Reviews, 222:9-27. Such modifications may be applicable to the effective design of the dimeric NKG2D-Fc described in the present disclosure.

[0072] Linker

[0073] The dimeric NKG2D-Fc constructs used in the methods of the present disclosure can further comprise at least one linking moiety that connects the first NKG2D moiety (e.g., NKG2D1) to the second NKG2D moiety (NKG2D2), the NKG2D moiety (NKG2D1 or NKG2D2) to the Fc fragment, and / or connects the Fc fragment to the drug moiety. In some embodiments, the linking moiety (e.g., linker molecule) is referred to as X1, X2, or X3. In some cases, the hinge region of the Fc fusion protein molecule acts as a spacer between the Fc region and the fusion peptide (e.g., a soluble receptor), thereby allowing the two parts of the molecule to function separately (see, e.g., Ashkenazi et al., 1997).

[0074] In some embodiments, at least one linking moiety (e.g., a linking molecule) is not a continuous portion of NKG2D1, NKG2D2, Fc, or a drug moiety and covalently links: an amino acid of NKG2D1 to an amino acid of NKG2D2, an amino acid of NKG2D2 to an amino acid of Fc, or an amino acid of Fc to a drug moiety. As used herein, a linking molecule that is a "non-continuous portion" means that each NKG2D portion (e.g., NKG2D1 and NKG2D2), the NKG2D portion and the Fc portion, and / or the Fc portion and the drug moiety of the chimera is connected by another element that is not part of the NKG2D or immunoglobulin or drug moiety, which element is essentially continuous with the portion of the chimera to which it is connected and serves as a linker. Non-limiting examples of linking molecules that are not a continuous portion of the NKG2D, Fc, or drug moiety are described below.

[0075] The connecting molecule can be a peptide linker. In some embodiments, the length of the peptide linker is within the range of about 2 to about 25 amino acids. In some embodiments, the length of the peptide linker is 20 amino acids. In some embodiments, the length of the peptide linker is within the range of about 4 to about 16 amino acids. In some embodiments, the length of the peptide linker is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, or 25 amino acids. In some embodiments, the length of the peptide linker is greater than 25 amino acids. In the case where the linker is a peptide linker, conventional molecular biology / recombinant DNA methods can be used to produce the dimeric NKG2D-Fc chimera in the form of a single recombinant polypeptide.

[0076] In some embodiments, the peptide linker provides a protease-dependent cleavable site. Examples of protease-cleavable peptide linkers include, but are not limited to, the MMP-sensitive linker GGPLGLWAGG (SEQ ID NO: 6) and the factor Xa-sensitive linker IEGR (SEQ ID NO: 7). The art is familiar with various cleavable sequences that can be used in the methods provided herein, such as those disclosed in Chen et al., Adv. Drug Deliv. Rev. (2013), 65(10): 1357-69.

[0077] In some embodiments of the present invention, a flexible peptide connexon is used. The length of the flexible peptide connexon is preferably about 25 or less amino acids. In some embodiments, the length of the flexible peptide connexon is 20 amino acids. In some embodiments, the peptide connexon contains about 20 or less amino acid residues, such as 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 and 20. In some embodiments, the peptide connexon contains about 12 or less amino acid residues, such as 3, 4, 5, 6, 7, 8, 9, 10, 11 and 12. In some cases, the peptide connexon comprises two or more of the following amino acids: glycine, serine, alanine and threonine. In some embodiments, the flexible peptide connexon is a glycine-serine connexon.

[0078] In some embodiments, the glycine-serine linker is of formula (GS) n wherein n is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12. In some embodiments, the glycine-serine linker is represented by the formula (GGGGS) n(SEQ ID NO: 2), wherein n is 1, 2, 3, 4 or 5.

[0079] In some embodiments, the dimeric NKG2D-Fc chimera comprises three linker molecules, X1, X2, and X3, wherein X1 covalently links the amino acids of NKG2D1 to the amino acids of NKGD2; X2 covalently links the amino acids of NKG2D2 to the amino acids of Fc; and X3 covalently links the amino acids of Fc to the drug moiety. In some embodiments, X1 is (GS)3 (SEQ ID NO: 4) and X2, X3, and X4 are each (GGGGS)4 (SEQ ID NO: 3).

[0080] In some embodiments, the dimeric NKG2D-Fc chimera contains an IEGR (SEQ ID NO: 7) peptide linker.

[0081] Alternatively, the connecting molecule can be a non-peptide linker. As used herein, a "non-peptide linker" is a biocompatible polymer comprising two or more repeating units connected to each other. Examples of non-peptide polymers include, but are not limited to, polyethylene glycol (PEG), polypropylene glycol (PPG), co-poly (ethylene / propylene) glycol, polyoxyethylene (POE), polyurethane, polyphosphazene, polysaccharide, dextran, polyvinyl alcohol, polyvinyl pyrrolidone, polyethylene ethyl ether, polyacrylamide, polyacrylate, polycyanoacrylate, lipopolymer, chitin, hyaluronic acid, and heparin. For a more detailed description of non-peptide linkers suitable for Fc fusion molecules, see, for example, WO / 2006 / 107124, incorporated herein by reference. Typically, depending on the specific linker, such linkers will have a molecular weight range of about 1 kDa to 50 kDa. For example, typical PEG has a molecular weight of about 1 to 5 kDa, and polyethylene glycol has a molecular weight of about 5 kDa to 50 kDa, and more preferably about 10 kDa to 40 kDa.

[0082] Drug section

[0083] In some embodiments, the dimeric NKG2D-Fc chimera further comprises a drug moiety. As used herein, a "drug moiety" refers to a therapeutic agent intended for delivery to a target cell (e.g., a cancer cell). Generally, the drug moiety is conjugated (e.g., directly or indirectly covalently bound) to the carboxyl terminus of the dimeric NKG2D-Fc chimera. However, the skilled artisan recognizes that in some embodiments, the drug moiety is conjugated to the amino terminus of the dimeric NKG2D-Fc chimera. Examples of "drug moieties" include drugs (e.g., small molecules), toxins (e.g., molecules of the lymphotoxin family), radionuclides, enzymes, cytokines, chemokines, single-chain variable fragments of antibodies directed against activating compounds or those that inhibit angiogenesis, or essentially any anti-tumor compound.

[0084] In some embodiments, the drug moiety comprises a cytokine or a functional portion thereof. Cytokines are proteins and peptides that can regulate immune cell function. A "functional portion" of a cytokine is a cytokine fragment that retains the ability to regulate immune cell function (e.g., binding to one or more cytokine receptors). Examples of cytokines include, but are not limited to, interferon-α (IFN-α), interferon-β (IFN-β), and interferon-γ (IFN-γ), interleukins (e.g., IL-1 to IL-29, specifically IL-2, IL-5, IL-6, IL-7, IL-10, IL-12, IL-15, and IL-18), tumor necrosis factor (e.g., TNF-α and TNF-β), erythropoietin (EPO), MIP3a, monocyte chemotactic protein (MCP)-1, intracellular adhesion molecules (ICAMs), macrophage colony stimulating factor (M-CSF), granulocyte colony stimulating factor (G-CSF), and granulocyte-macrophage colony stimulating factor (GM-CSF). In some embodiments, the drug moiety comprises a cytokine selected from the group consisting of IL-2, IL-12, IL-15, IL-18, IL-21, and IFN-α.

[0085] In some embodiments, the drug moiety comprises a cytokine / cytokine receptor hybrid. Cytokine / cytokine receptor hybrids are known in the art and are described, for example, in Rowley et al., Eur J Immunol. 2009 Feb;39(2):491-506. In some embodiments, the dimeric NKG2D-Fc chimera comprises a drug moiety comprising an IL-15 (e.g., GeneBank AAX37025) / IL-15Ra (e.g., GeneBank AAP69528.1) hybrid. In some embodiments, the drug moiety comprises amino acids 31-107 of human IL-15 receptor alpha (hIL15Ra, GeneBank AAP69528.1) fused to amino acids 22-135 of IL-15 (GeneBank AAX37025). In some embodiments, IL-15 and IL-15Ra are separated by a linker, for example, a 20-amino acid (G4S)4 (SEQ ID NO: 3) linker. Dimeric NKG2D-Fc chimeras comprising IL-15 / IL-15Ra hybrids are further described in the Examples section. In some embodiments, the dimeric NKG2D-Fc chimeras comprise a drug moiety comprising a hybrid of IL-12p35 and IL-12p40. In some embodiments, IL-12p35 and IL-12p40 are separated by a linker, e.g., a 20-amino acid (G4S)4 (SEQ ID NO: 3) linker. In some embodiments, the dimeric NKG2D-Fc chimeras comprise a drug moiety comprising a hybrid of IL-23p19 and IL-23p40. In some embodiments, IL-23p19 and IL-23p40 are separated by a linker, e.g., a 20-amino acid (G4S)4 (SEQ ID NO: 3) linker. In some embodiments, the dimeric NKG2D-Fc chimeras comprise a drug moiety comprising a hybrid of IL-27p28 and EB1. In some embodiments, IL-27p28 and EB1 are separated by a linker, eg, a 20-amino acid (G4S)4 (SEQ ID NO: 3) linker. In some embodiments, each subunit of the cytokine / cytokine receptor hybrid is located on a different chain of the dimeric NKG2D-Fc chimera.

[0086] In some embodiments, the drug moiety is an antibody single chain variable fragment (single chain variable fragment; ScFv). As used herein, "antibody single chain variable fragment" refers to a fusion protein of the variable regions of the heavy chain (VH) and light chain (VL) of an immunoglobulin connected to a short connecting peptide. Despite the removal of the constant region and the introduction of a linker, the scFV protein maintains the specificity of the initial immunoglobulin. In some embodiments, ScFv binds to immune checkpoint proteins (e.g., PD1 or CTLA4). In some embodiments, ScFv hinders angiogenesis (e.g., binding to regulatory factors of angiogenesis, such as VEGF).

[0087] In some embodiments, the drug moiety is a chemokine. As used herein, "chemokine" refers to a low molecular weight protein that stimulates leukocyte recruitment. In general, chemokines are secondary proinflammatory mediators induced by primary proinflammatory mediators such as interleukin-1 (IL-1) or tumor necrosis factor (TNF). Chemokines can be classified into four families: CC chemokines (e.g., CCL1 to CCL-28), CXC (e.g., CXCL1 to CXCL17), C (e.g., XCL1, XCL2), and CX3C (CX3CL1).

[0088] In some embodiments, the drug moiety is a small molecule. As used herein, "small molecule" refers to a non-peptide, non-oligomeric organic compound synthesized in a laboratory or found in nature. Non-limiting examples of small molecule drugs include small molecule kinase inhibitors (e.g., everolimus, gefitinib, imatinib, etc.), bromodomain inhibitors (e.g., JQ1, I-BET 151, RVX-208, etc.), antibiotics (e.g., kanamycin, neomycin, ciprofloxacin, etc.), and antivirals (e.g., ribavirin, rimantadine, zidovudine, etc.). In some embodiments, the small molecule is an antitumor compound. Antitumor compounds are further discussed in detail elsewhere in this disclosure.

[0089] In some embodiments, the drug moiety is a radionuclide. As used herein, "radionuclide" refers to a medically useful radionuclide. Examples of radionuclides include 99m Tc, 188 Re、 186 Re、 153 Sm, 166 Ho, 90 Y. 89 Sr. 67 Ga,68 Ga, 111 In, 183 Gd, 59 Fe, 225 Ac, 212 Bi, 211 At 45 Ti, 60 Cu, 61 Cu, and 67 Cu.

[0090] Other parts

[0091] In some embodiments, the dimeric NKG2D-Fc chimeras suitable for use in the methods described herein may further comprise one or more auxiliary moieties, such as tag sequences and signal sequences. For example, the tag sequence can be used to detect and / or isolate the polypeptide. Examples of tags include, but are not limited to, HA, Flag, Myc, Glu, His, and maltobasic protein. The tag sequence can be located at the amino terminus, the carboxyl terminus, or somewhere within the dimeric NKG2D-Fc chimera (e.g., between modular peptide segments), provided that the presence of such a tag does not interfere with the function of the dimeric NKG2D-Fc molecule. In some cases, the tag sequence is cleavable.

[0092] In some embodiments, the dimeric NKG2D-Fc chimera may optionally include a signal sequence. A signal sequence is a short (usually about 3-60 amino acids long) peptide chain that directs the post-translational transport of a polypeptide, thereby allowing for greater polypeptide production. The amino acid sequence of the signal sequence directs the polypeptide (synthesized in the cytosol) to certain subcellular compartments, such as organelles. Signal sequences are also referred to as guidance signals, signal peptides, transit peptides, or localization signals. In some embodiments, after transport of the polypeptide, the signal sequence is cleaved from the polypeptide by a signal peptidase.

[0093] In some embodiments, the dimeric NKG2D chimera contains an N-terminal modified IL-2 signal sequence, which allows for optimal expression and secretion of the NKG2D-Fc construct. See, for example, Zhang et al., 2004, J. Gene Med., 7:354-65. In some embodiments, the dimeric NKG2D chimera contains a signal peptide derived from a CD33 polypeptide sequence. For example, the CD33 signal peptide may correspond to amino acid residues 1-16 of the CD33 polypeptide sequence. Those skilled in the art will appreciate that there are many other suitable signal peptide sequences that can be used to practice the methods provided herein. In addition, where a signal peptide is present in the NKG2D chimera, additional amino acid residues, such as spacers, may optionally be inserted between the N-terminal signal peptide and the Fc portion of the chimera. In some embodiments, for example, the signal sequence is followed by a Met-Asp dipeptide spacer.

[0094] Preparation of dimeric NKG 2D-Fc

[0095] The art is familiar with molecular biology and biochemistry techniques for preparing dimeric NKG2D-Fc chimeras with desired characteristics. Preferably, the dimeric NKG2D-Fc chimera construct is produced by conventional recombinant DNA methods. In a preferred embodiment, the dimeric NKG2D-Fc chimera is produced as a single (e.g., continuous) recombinant polypeptide. In other embodiments, two or more portions of the dimeric NKG2D-Fc are produced as separate fragments and subsequently linked together to produce the dimeric NKG2D-Fc molecule. For example, each NKG2D portion of the chimera (e.g., NKG2D1, NKG2D2) and the Fc portion of the dimeric NKG2D-Fc are each produced as separate recombinant polypeptides, which are then fused together by chemical ligation to produce the dimeric NKG2D-Fc. This production method may be particularly preferred in cases where a non-peptide linker is employed. Similarly, this method of production may also be preferred if the dimeric NKG2D-Fc chimera does not fold properly (eg, does not properly bind ligand) when produced as a single contiguous polypeptide.

[0096] For the production of recombinant polypeptides, various host organisms can be used. Suitable hosts include, but are not limited to, bacteria such as Escherichia coli (E. coli), yeast cells, insect cells, plant cells, and mammalian cells. The choice of a suitable host organism will depend on the specific application of the dimeric NKG2D-Fc chimera. A skilled artisan will understand how to consider certain criteria in selecting a suitable host for the production of recombinant polypeptides. Factors influencing the selection of a suitable host include, for example, post-translational modifications, such as phosphorylation and glycosylation patterns, as well as technical factors, such as overall expected yield and ease of purification. Host-specific post-translational modifications of the dimeric NKG2D-Fc to be used in vivo should be carefully considered, as certain post-specific modifications are known to be highly immunogenic (antigenic).

[0097] Once produced, the dimeric NKG2D-Fc can be purified by any suitable means, such as chromatography methods known to those skilled in the art. Examples of chromatography methods include gel filtration chromatography. See, for example, Caine et al., Protein Expr. Purif., 1996, 8:159-66. In some embodiments, the dimeric NKG2D-Fc is purified by protein A immunoaffinity chromatography.

[0098] As will be appreciated by those skilled in the art, the dimeric NKG2D chimera portions can also be prepared and isolated separately and joined by chemical synthesis.

[0099] NKG2D receptor ligands

[0100] In any of the embodiments described herein, the dimeric NKG2D-Fc chimera is capable of binding to endogenous ligands of the NKG2D receptor. Known NKG2D ligands in humans include MICA, MICB, RAET-1G, ULBP1, ULBP2, ULBP3, ULBP4, ULBP5, and ULBP6. Preferably, the dimeric NKG2D-Fc chimera of the present disclosure is capable of binding to more than one type of NKG2D receptor ligand.

[0101] In some embodiments, the dimeric NKG2D-Fc chimeric molecule is used with 10 -4 M or smaller, 10 -7 In some embodiments, the dimeric NKG2D-Fc molecule binds to its ligand with a high affinity of at least 5×10 6 Ka, at least 1×10 7 Ka, at least 2×10 7 Ka, at least 1×10 8 Ka or greater.

[0102] In some embodiments, NKG2D-Fc preferentially binds to (eg, with higher affinity for) a subset of NKG2D receptor ligands. 3D structural data combined with mutagenesis analysis have revealed that NKG2D permits recognition and binding of a diverse array of its endogenous ligands.

[0103] The ligands of NKG2D can be expressed on the cell surface. Alternatively, the ligands of NKG2D can be "shed" from the cell surface and exist in the form of soluble ligands. It is known that in certain cancers, NKG2D ligands such as MICA are overexpressed and in some cases released (e.g., shed) into the bloodstream or surrounding tissues, such as serum, in a soluble form. It is believed that this at least partially contributes to the pathogenesis and / or progression of cancer. Therefore, dimeric NKG2D-Fc is suitable for binding to such ligands present on the cell surface or in a released form, balancing the expression of ligands present at abnormally high levels by acting as a neutralizer.

[0104] In the event that an NKG2D ligand is expressed on the surface of a cancer cell in an individual, the dimeric NKG2D-Fc described herein binds to the cell surface ligand when administered to the individual. Binding of the dimeric NKG2D-Fc chimera to its ligand prevents activation of endogenous NKG2D receptors present on NK cells. In the event that an NKG2D ligand is "shed" from a cancer cell, for example, released into the individual's bloodstream, the dimeric NKG2D-Fc described herein binds to the soluble ligand, sequestering it from further action.

[0105] Therapeutic applications

[0106] Typically, expression of NKG2D ligands appears to be restricted to the gastrointestinal epithelium. Minimal expression is observed in quiescent epithelial cells, but higher levels occur in rapidly proliferating cells. NKG2D ligand expression is also upregulated in various transformed cells, particularly those of epithelial origin. Accordingly, provided herein are methods for treating cancer or cancer symptoms in an individual. The methods comprise administering to the individual a therapeutically effective amount of a dimeric NKG2D-Fc that binds to an NKG2D ligand in vivo.

[0107] The terms "treating", "treatment", "treat", and the like, in the context of cancer therapy, refer to administering a composition comprising a dimeric NKG2D-Fc as described herein to an individual suffering from cancer. The composition is administered to the individual in a therapeutically effective amount. As used herein, a therapeutically effective amount refers to an amount of treatment that is believed to achieve a statistically significant beneficial effect on an individual suffering from a disease or condition, such as certain types of cancer. Generally, a therapeutically effective amount is determined by administering a composition to a population of individuals suffering from a specified condition (such as the progression or stage of a disease) and evaluating the results in response. As used herein, therapeutic treatment should include, for example, completely preventing or eliminating symptoms of a disease, delaying the onset of symptoms of a disease, or reducing the severity of a disease.

[0108] cancer

[0109] Dimeric NKG2D-Fc chimeras are believed to be broadly suitable for immunotherapy of various cancers in which expression of one or more NKG2D ligands is elevated in an individual. Cancer broadly refers to proliferative diseases involving transformed cells, including both precancerous and malignant conditions. The present invention is suitable for treating individuals with cancer characterized by overexpression of one or more NKG2D ligands. In some embodiments, the cancer is characterized by overexpression of one (or predominantly one) ligand of the NKG2D receptor. In other embodiments, the cancer is characterized by overexpression of two or more NKG2D ligands.

[0110] The methods disclosed herein are useful therapies for treating precancerous conditions that carry with them the risk of progression to malignancy.

[0111] Examples of such conditions include, but are not limited to, dysplasia, hyperplasia, and plasma cell disorders, such as monoclonal gammopathy of undetermined significance (MGUS) and smoldering multiple myeloma (SMM). In some embodiments, the cancer is melanoma, lung cancer, breast cancer, kidney cancer, ovarian cancer, prostate cancer, pancreatic cancer, gastric cancer, and colon cancer, lymphoma, or leukemia. In some embodiments, the cancer is melanoma. In some embodiments, the cancer is a plasma cell malignancy, such as multiple myeloma (MM) or a precancerous condition of plasma cells. In some embodiments, the cancer is melanoma, lung cancer, plasma cell cancer, leukemia, lymphoma, ovarian cancer, colon cancer, pancreatic cancer, or prostate cancer. In some embodiments, the individual has been diagnosed with cancer or is susceptible to cancer. Therefore, the methods disclosed herein are also suitable for treating individuals who have metastasized and are therefore susceptible to relapse or recurrence. The methods are particularly useful in individuals at high risk, such as those with a family history of cancer or metastatic tumors, or those who exhibit a genetic predisposition to cancer metastasis. Specifically, the methods relate to treating cancers associated with expression of NKG2D ligands. In some embodiments, the NKG2D ligand is MICA. Thus, in some embodiments, the cancer results in a MICA-associated tumor.

[0112] Whether a particular individual (e.g., a patient) should receive a cancer therapy comprising NKG2D-Fc can be determined by testing for abnormal expression of one or more NKG2D ligands in the individual. "Abnormal expression of one or more NKG2D ligands" in an individual means overexpression of the ligand in a biological sample obtained from the individual. In some embodiments, the biological sample may include a biopsy sample taken from a tissue of an individual suspected of having cancer. For example, in some cases, a biological sample is collected from a solid tumor to test for malignancy. In other cases, the biological sample may constitute a blood sample, such as serum, stool sample, urine sample, etc. For the purpose of testing for a disease, such as the diagnosis or progression of cancer, the biological sample may be any cell or tissue sample collected from an individual.

[0113] Those skilled in the art are familiar with various laboratory techniques and protocols for analyzing the presence and content of one or more markers present in a biological sample. To determine whether an individual suffers from a cancer associated with overexpression of an NKG2D ligand, an immunoaffinity analysis is typically performed. In some cases, depending on the type of biological sample available, immunohistological or immunocytochemical analysis may be performed. Many antibodies are commercially available for performing these analyses. Commonly used methods for this purpose include, but are not limited to, ELISA, immunoblotting, and immunohistochemistry.

[0114] individual

[0115] The methods disclosed herein can be applied to a wide range of species known to suffer from cancer, such as humans, non-human primates (e.g., monkeys), horses, cattle, pigs, sheep, deer, elk, goats, dogs, cats, rabbits, guinea pigs, hamsters, rats, and mice. Thus, as used herein, an "individual" is a mammalian individual who suffers from a disease, or is at risk of suffering from a disease associated with abnormal expression of at least one NKG2D ligand, such as cancer. In a preferred embodiment, the individual is a human individual suffering from a cancer that exhibits elevated levels of one or more NKG2D ligands. In some embodiments, the NKG2D ligand comprises MICA.

[0116] If an individual already expresses elevated levels of one or more NKG2D ligands, then the individual can be treated with the methods described herein. In some cases, the individual has received or is currently receiving another cancer therapy. In some embodiments, the cancer may be in remission. In some cases, the individual is at risk of recurrence, such as metastasis. In some embodiments, overexpression of one or more NKG2D ligands is limited to cancer cells, such as tumors. In some embodiments, at least one of the NKG2D ligands expressed by cancer cells is shed into the bloodstream and is therefore detectable in the individual's serum.

[0117] Depending on the phenotype of a particular cancer, it may be possible to target one or more ligands that are overexpressed (by tumor cells) over other ligands whose expression is not significantly affected.

[0118] Mode of action

[0119] The present invention is based, at least in part, on the surprising discovery that chimeric molecules comprising two NKG2D fragments and an Fc fragment (e.g., dimeric NKG2D-Fc chimeras) are able to bind one or more NKG2D ligands to induce tumor cell death with improved efficacy compared to chimeric molecules comprising a single NKG2D fragment and an Fc fragment (e.g., monomeric NKG2D-Fc chimeras).

[0120] Without being limited by any particular theory, it is shown that the dimeric NKG2D-Fc chimera can act through two main components of the immune system: innate immunity and adaptive immunity. As used herein, innate immunity or the innate immune system refers to a nonspecific host defense mechanism against exogenous pathogens. Innate immunity includes physical barriers (such as skin, gastric acid, mucus or tears and cells) and active mechanisms, such as NK cells, phagocytes and the complement system. NK cells represent the main components of the innate immune system. NK cells have cytotoxicity, for example, they can attack cells infected by microorganisms, as well as some types of tumor cells. The cytotoxic activity of NK cells is mediated by cell surface receptors that recognize MHC class I alleles. Many receptor types are known in the art, including NKG2D as a receptor subtype. Phagocytes include neutrophils, monocytes, macrophages, basophils and eosinophils. The complement system is a biochemical cascade of the immune system that helps to remove pathogens from the host organism.

[0121] Generally speaking, adaptive immunity or the adaptive immune system refers to an antigen-specific antibody-mediated immune response. Adaptive immunity is generally mediated by specific antibody production by B lymphocytes and antigen-specific activity by T lymphocytes. The humoral response mediated by B lymphocytes primarily fights extracellular pathogens by producing circulating antibodies that mark foreign cells and molecules, thereby destroying them by other specialized cells and proteins. The cellular response mediated by T lymphocytes primarily fights intracellular pathogens and cancer cells by directly binding to and destroying diseased cells. According to the present disclosure, it is believed that dimeric NKG2D-Fc, as a non-antibody molecule, functionally mimics the functions of specific antibodies.

[0122] The present invention therefore encompasses methods for cancer treatment in which dimeric NKG2D-Fc directly binds to tumor cells expressing NKG2D ligands on their cell surfaces. In this mode of action, dimeric NKG2D-Fc can specifically identify and destroy tumor cells that overexpress NKG2D ligands, but not healthy cells that do not overexpress NKG2D ligands.

[0123] Dimeric NKG2D-Fc can target any or all NKG2D ligands expressed on human tumor cells in at least two ways. One mechanism of mediating tumor cell destruction is through the process of complement lysis (also known as complement-dependent lysis, complement-dependent cytotoxicity, or CDC). A second way of mediating tumor cell destruction is by triggering antibody-dependent cellular cytotoxicity (ADCC).

[0124] In some embodiments, dimeric NKG2D-Fc acts as an opsonizing agent. Opsonization is the process by which cells or particles are coated with molecules that bind to receptors on other cells (e.g., dendritic cells or phagocytes) to promote uptake. For antigen-presenting cells such as dendritic cells and macrophages, opsonization promotes efficient processing and presentation of antigens. Particularly useful are opsonizing agents that can specifically bind to both a target (e.g., a ligand) and a specific receptor on antigen-presenting cells (e.g., an FcR) that can mediate internalization and subsequent antigen processing.

[0125] Tumor cells that express one or more ligands for the NKG2D receptor on their cell surface can be opsonized, for example, by coating them with dimeric NKG2D-Fc molecules. For example, the NKG2D portion of the chimera can bind to the ligand on the tumor cell surface, exposing the Fc portion of the chimera. Dendritic cells possess FcγRs and can therefore bind and internalize tumor antigens (e.g., NKG2D ligands), leading to antigen presentation to cytotoxic T cells, also known as CD8+ T cells. This process is known as cross-activation. Similarly, opsonization results in the generation of MHC class II-restricted CD4+ T cell responses. Thus, through opsonization, the NKG2D-Fc chimera can promote efficient cross-presentation (e.g., activation) of dendritic cells, leading to the induction of effective T cell responses against the tumor.

[0126] Cancer patients often suffer from immunosuppression. In some cases, it is thought that immunosuppression may be caused, at least in part, by reduced NKG2D receptor signaling. For example, based on a popular model, shed MICA impairs host defense by inducing internalization of NKG2D receptor molecules on lymphocytes. Therefore, according to this model, tumor cells that shed MICA cause immunosuppression by downregulating NKG2D surface expression.

[0127] Therefore, the methods provided herein are suitable for counteracting or alleviating immunosuppression by administering a composition comprising dimeric NKG2D-Fc, particularly in cases where the patient exhibits elevated levels of soluble (i.e., shed) NKG2D ligands or ligands detectable in serum. The mode of action is that the NKG2D-Fc administered to the patient binds to (and thereby sequesters) excess soluble ligands of NKG2D shed from the tumor, thereby reversing the down-regulated expression of NKG2D receptors on the cell surface that leads to immunosuppression.

[0128] Thus, dimeric NKG2D-Fc chimeras may have multiple therapeutic functions, including: neutralizing soluble ligands shed by tumor cells; promoting ADCC and / or CDC in tumor cells expressing cell surface ligands; and mediating cross-presentation and activation of the adaptive immune system, including CD8 cytotoxic T lymphocytes (CTLs) and B cells that produce tumor-specific antibodies.

[0129] Application

[0130] The dimeric NKG2D-Fc composition can be administered directly to a subject. The subject is preferably a mammal. The terms "administration" and "administer" refer to the process of providing a pharmaceutical agent to a subject so that the pharmaceutical agent contacts its target cells (e.g., cancer cells) in vivo, i.e., within the subject's body. In some embodiments, the composition comprising NKG2D-Fc is administered systemically to the subject. In preferred embodiments, systemic administration is via intravenous injection. In some embodiments, the composition comprising dimeric NKG2D-Fc is administered topically. For example, in some cases, the composition can be delivered directly to a solid tumor or in close proximity to a solid tumor.

[0131] Pharmaceutically acceptable carrier

[0132] In some aspects, the present disclosure provides compositions comprising a dimeric NKG2D-Fc chimera as described herein and a pharmaceutically acceptable carrier. Generally, the composition comprising the dimeric NKG2D-Fc can be suspended in a pharmaceutically acceptable carrier (e.g., normal saline). Such carriers can include, but are not limited to, sterile aqueous or non-aqueous solutions, suspensions, and emulsions. Examples of non-aqueous solvents include, for example, mineral oil, propylene glycol, polyethylene glycol, vegetable oils, and injectable organic esters. Aqueous carriers include, but are not limited to, water, alcohol, saline, and buffered solutions. Preservatives, flavorings, and other additives, such as antimicrobials, antioxidants, chelating agents, and inert gases, may also be present. It will be understood that any material described herein that is to be administered to a mammal may contain one or more pharmaceutically acceptable carriers.

[0133] Route of administration

[0134] Any composition described herein can be applied to any part of the individual's body by various routes of administration. Can be applied by intravenous, intraperitoneal, intramuscular, subcutaneous, intramuscular, intrarectal, intravaginal, intrathecal, intratracheal, intradermal, or transcutaneous injection, by oral or nasal administration, by suction, or by progressive perfusion over time. Compositions can be delivered to specific tissues. For example, compositions can be delivered to, but are not limited to, mammalian joints, nasal mucosa, blood, lungs, intestines, muscle tissue, skin, or peritoneal cavity. In another example, an aerosol of the composition can be given to an individual by suction.

[0135] dose

[0136] The required dosage depends on the route of administration, the nature of the formulation, the nature of the patient's disease, the individual's height, weight, surface area, age and sex, other drugs administered and the judgment of the attending physician. Suitable dosages are generally in the range of 0.01-1,000 μg / kg. In view of the different efficiencies of the various dimeric NKG2D-Fc compositions and various routes of administration, it is expected that the required dosage will vary significantly. As is well known in the art, variations in these dosage amounts can be adjusted using standard empirical formulas to optimize them. Administration can be single or multiple (e.g., 2-fold, or 3-fold, 4-fold, 6-fold, 8-fold, 10-fold, 20-fold, 50-fold, 100-fold, 150-fold or more). Encapsulation of the composition in a suitable delivery vehicle (e.g., polymeric microparticles or implantable devices) can improve delivery efficiency.

[0137] Treatment options

[0138] The duration of treatment using any of the compositions provided herein can be any length of time, from as short as one day to as long as the lifespan of the subject (e.g., many years). For example, a dimeric NKG2D-Fc composition can be administered monthly for three months, or annually for a period of ten years. It should also be noted that the frequency of treatment can vary. For example, a dimeric NKG2D-Fc composition can be administered daily, weekly, monthly, or annually (or twice, three times, etc.). A dimeric NKG2D-Fc composition can be administered, for example, at the same time point or sequentially with one or more other cancer therapies. For example, a patient can receive an autologous tumor cell vaccine, followed by an anti-CTL4 antibody, and then a dimeric NKG2D-Fc therapy, separated by intervals of hours, days, months, or years.

[0139] effective dose

[0140] Any composition described herein can be administered to an individual in an effective amount. As used herein, the term "effective" refers to inducing a desired therapeutic effect, such as an immune response, without inducing any amount of significant toxicity in an individual. Such amounts can be determined by assessing the biological response (e.g., immune response) and symptom improvement of an individual after administering a known amount of a specific composition. In addition, toxicity (if present) levels can be determined by assessing the clinical symptoms of an individual before and after administering a known amount of a specific composition. It should be noted that the effective amount of the specific composition administered to an individual can be adjusted according to desired results and subject response and toxicity levels. Significant toxicity can be changed for each specific subject and depends on a variety of factors, including but not limited to individual disease conditions, age, and pain tolerance.

[0141] Combination therapy

[0142] In some embodiments, individuals in need of cancer treatment are treated with the dimeric NKG2D-Fc compositions described herein in combination with other cancer therapies. In some embodiments, the other cancer therapies include cytotoxic and / or non-cytotoxic agents. "Cytotoxic agent" refers to a substance that inhibits or prevents cell function and / or causes cell destruction. The term is intended to include radioactive isotopes (e.g., 131 I. 125 I. 90 Y and 186 Re), chemotherapeutic agents and toxins, such as enzymatically active toxins of bacterial, fungal, plant or animal origin, or synthetic toxins, or fragments thereof. Non-cytotoxic agents refer to substances that do not inhibit or prevent cellular function and / or do not cause cell damage. "Non-cytotoxic agents" may include agents that can be activated to become cytotoxic. Non-cytotoxic agents may include beads, liposomes, matrices or particles (see, for example, U.S. Patent Publications 2003 / 0028071 and 2003 / 0032995, which are incorporated herein by reference). Such agents may be conjugated, coupled, linked or associated with the dimeric NKG2D-Fc compositions described herein.

[0143] In some embodiments, conventional cancer drugs are administered together with the compositions described herein. In some cases, individuals in need of cancer treatment are treated with the dimeric NKG2D-Fc compositions described herein in combination with one or more additional agents for targeting cancer cells. Highly suitable agents include those that promote DNA damage in cancer cells (e.g., double-strand breaks in cellular DNA). Any form of DNA damaging agent known to those skilled in the art can be used. DNA damage can typically be produced by radiotherapy and / or chemotherapy. DNA damaging agents are also referred to as genotoxic agents. As used herein, "in combination with" should mean administering dimeric NKG2D-Fc to an individual in combination with one or more additional therapies (simultaneously or separately but very closely), before or after administering one or more additional therapies.

[0144] Examples of radiation therapy include, but are not limited to, external beam radiation therapy and internal beam radiation therapy (also known as brachytherapy). Energy sources for external beam radiation therapy include x-rays, gamma rays, and particle beams; energy sources for internal beam radiation include radioactive iodine (iodine 125 or iodine 131 ),strontium 89 , or a radioactive isotope of phosphorus, palladium, cesium, indium, phosphate, or cobalt. Methods of administering radiation therapy are well known to those skilled in the art.

[0145] Examples of DNA damaging chemotherapeutic agents include, but are not limited to, busulfan (Myleran), carboplatin (Paraplatin), carmustine (BCNU), chlorambucil (Leukeran), cisplatin (Platinol), cyclophosphamide (Cytoxan, Neosar), dacarbazine (DTIC-Dome), ifosfamide (Ifex), lomustine (CCNU), mechlorethamine (nitrogen mustard, Mustarden), melphalan (Alkeran), and procarbazine (Matulane).

[0146] Many other chemotherapeutic agents can also be used alone or in combination in the methods described herein. These agents include: methotrexate, vincristine, adriamycin, cisplatin, non-sugar-containing chloroethyl nitrosourea, 5-fluorouracil, mitomycin C, bleomycin, cranberry, dacarbazine, paclitaxel, fragyline, meglumine GLA, valrubicin, carmustine and polyphenylpropion, MMI270, BAY 12-9566, RAS farnesyltransferase inhibitors, farnesyltransferase inhibitors, MMP, MTA / LY231514, LY264618 / Lometexol, Glamolec, CI-994, TNP-470, Hycamtin / Topotecan, PKC412, Valspodar / PSC833, Novantrone / Mitoxantrone, Metaret / Suramin, Batimastat, E7070, BCH-4556, CS-682, 9-AC, AG3340, AG3433, Incel / VX-710, VX-853, ZD0101, ISI641, ODN 698, TA 2516 / Marmistat, BB2516 / Marmistat, CDP 845, D2163, PD183805, DX8951f, Lemonal DP2202, FK 317, Picibanil / OK-432, AD 32 / Valrubicin, Metastron / Strontium derivative, Temodal / Temozolomide, Evacet / Liposomal Doxorubicin, Yewtaxan / Pacific Taxol, Paclitaxel / Pacific Taxol, Xeload / Capecitabine, Furtulon / Doxifluridine, Cyclopax / Oral Pacific Taxol, Oral Taxanes, SPU-077 / Cisplatin, HMR1275 / Flavopiridol, CP-358(774) / EGFR, CP-609(754) / RAS oncogene inhibitor, BMS-182751 / oral platinum, UFT (Uracil), Ergamisol / levamisole, Enuracil / 776C85 / 5FU booster, Campto / levamisole, Camptosar / irinotecan, Tumodex / Ralitrexed, Leustatin / Cladribine, Paxex / paclitaxel, Doxitaxel / liposomal paclitaxel, Caelyx / liposomal paclitaxel, Fludara / fludarabine, Pharmarubicin / epibricin, DepoCyt, ZD1839, LU 79553 / Bis-naphthamide, LU 103793 / Dolastain, Caetyx / Liposomal Berry, Gemzar / Gemcitabine, ZD 0473 / Anormed, YM 116, Rodin Seed, CDK4 and CDK2 Inhibitors, PARP Inhibitors, D4809 / Desifrinamide, Ifes / Mesnex / Ifosamide, Vumon / Teniposide, Paraplatin / Carboplatin, Plantinol / Cisplatin, Vepeside / Etoposide, ZD9331, Taxotere / docetaxel, prodrugs of guanine arabinoside, taxane analogs, nitrosoureas, alkylating agents such as melphalan and cyclophosphamide, aminoglutethimide, asparaginase, busulfan, carboplatin, chlorambucil, cisplatin, cytarabine HCl, actinomycin, daunorubicin HCl, estramustine sodium phosphate, etoposide (VP16-213), floxuridine, 5-fluorouracil (5-FU), flutamide, hydroxyurea (hydroxyurea), ifosfamide, interferon alpha-2a, alpha-2b, leuprolide acetate (LHRH-releasing factor analog), lomustine (CCNU), nitrogen mustard H Cl (nitrogen mustard), mercaptopurine, mesna, mitotane (op'-DDD), mitoxantrone HCl, octreotide, plicamycin, procarbazine HCl, streptozotocin, tamoxifen citrate, thioguanine, thiotepa, vinblastine sulfate, amsacrine (m-AMSA), azacitidine, erythropoietin, hexamethylmelamine (HMM), interleukin 2, propionylguanidine (methyl-GAG; methylglyoxal bis-guanylhydrazone; MGBG), pentostatin (2'deoxycoformycin), semustine (methyl-CCNU), teniposide (VM-26) and vindesine sulfate, but are not so limited.

[0147] In addition, the following agents may also be suitable for use in the present invention: alkylating agents, such as carboplatin and cisplatin; nitrogen mustard alkylating agents; nitrosourea alkylating agents, such as carmustine (BCNU); antimetabolites, such as methotrexate, folinic acid, purine analog antimetabolites, mercaptopurine, pyrimidine analog antimetabolites, such as fluorouracil (5-FU) and gemcitabine. Hormonal anticancer drugs, such as goserelin, leuprolide, and tamoxifen; natural anticancer drugs, such as aldesleukin, interleukin-2, docetaxel, etoposide (VP-16), interferon α, and paclitaxel and retinoic acid (ATRA); antibiotic natural antitumor drugs such as bleomycin, actinomycin D, daunorubicin, daunorubicin, daunomycin and mitomycins including mitomycin C; and vinca alkaloid natural antitumor drugs such as vinblastine, vincristine, vindesine, hydroxyurea, acetone, adriamycin, ifosfamide, enocitabine, cyclothiocarb, aclarubicin, ancitabine, nimustine, procarbazine hydrochloride, carboquinone, carboplatin, carmofur, chromomycin A3, antitumor polysaccharides, antitumor platelet factor, cyclophosphamide Schizosaccharide, cytarabine (cytosine arabinoside), dacarbazine, thioinosine, thiotepa, fumaridine, dolastatin, dolastatin analogs such as auristatin, CPT-11 (irinotecan), mitozantrone, vinorelbine, teniposide, aminopterin, carbomycin, esperamicin (see, e.g., U.S. Pat. No. 4,675,187, incorporated herein by reference), new tumor suppressor protein, OK 432, bleomycin, flutolone, broxundine, busulfan, tetrasodium diethylstilbestrol diphosphate, peplomycin, bestatin Interferon-beta, melastane, mitobromtol, melphalan, laminin peptide, lentinan, Coriolus versicolor extract, fluazifop / uracil, estramustine (estrogen / nitrogen mustard), thalidomide, and lenalidomide

[0148] Other suitable chemotherapeutic agents include proteasome inhibitors. Proteasome inhibitors hinder the action of the proteasome, a cellular complex that degrades proteins, especially short-lived proteins involved in cell maintenance, growth, division, and cell death. Examples of proteasome inhibitors include bortezomib Lactacystin (AG Sciences, San Diego, CA), MG132 (Biomol International, Plymouth, ME), PS-519, eponemycin, epoxomicin, aclarubicin A, dipeptide benzamide, CVT-63417, and a vinyl sulfone tripeptide proteasome inhibitor.

[0149] In some embodiments, the methods described herein are used in combination with one or more other cancer treatments, including cancer immunotherapy. Cancer immunotherapy harnesses the immune system to fight cancer. The primary premise is to stimulate the individual's immune system to attack the tumor cells that cause the disease. This can be done by immunizing the individual, in which case the individual's own immune system is primed to recognize tumor cells as targets for destruction, or by administering a therapy, such as an antibody as a drug, in which case the therapeutic agent restores the individual's immune system to destroy tumor cells. Cancer immunotherapy includes antibody-based therapies and cytokine-based therapies.

[0150] Many therapeutic monoclonal antibodies have been approved by the FDA for use in humans, and more are in progress. FDA-approved monoclonal antibodies for cancer immunotherapy include antibodies to CD52, CD33, CD20, ErbB2, vascular endothelial growth factor, and epidermal growth factor receptor. These and other antibodies targeting one or more cancer-associated antigens are therefore suitable for combination therapies that will be administered in conjunction with dimeric NKG2D-Fc. Examples of FDA-approved monoclonal antibodies for cancer therapy include, but are not limited to, rituximab (Rituximab) (available as Rituxan TM Purchased), Trastuzumab (Herceptin TM Alemtuzumab (available from Campath-IH TM Cetuximab (available as Erbitux TM Bevacizumab (available from Avastin TM Panitumumab (available from Vectibix TM Gemtuzumab ozogamicin (available from Mylotarg TM Ibritumomab tiuxetan (available from Zevalin TM Tositumomab (available from Bexxar) TM Examples of monoclonal antibodies currently undergoing human clinical testing for cancer therapy in the United States include, but are not limited to: WX-G250 (available from Rencarex TMavailable), Ipilimumab (available as MDX-010), Zanolimumab (available as HuMax-CD4), Ofatumumab (available as HuMax-CD20), ch14.18, Zalutumumab (available as HuMax-EGFr), Oregovomab (available as B43.13, OvalRex TM purchased), Edrecolomab (available from IGN-101, Panorex TM Purchased), 131 I-chTNT-I / B (Can Cotara TM Pemtumomab (available from Theragyn TM available), Lintuzumab (available as SGN-33), Labetuzumab (available as hMN14, CEAcide TM Catumoximab (available from Removab TM available), CNTO 328 (available as cCLB8), 3F8, 177Lu-J591, Nimotuzumab, SGN-30, Ticilimumab (available as CP-675206), Daclizumab (available as Zenapax TM Epratuzumab (available from hLL2, LymphoCide TM Purchased), 90 Y-Epratuzumab, Galiximab (available as IDEC-114), MDX-060, CT-011, CS-1008, SGN-40, Mapatumumab (available as TRM-1), Apolizumab (available as HuID10, Remitogen TM and Volociximab (available as M200).

[0151] Cancer immunotherapy also includes cytokine-based therapies. Cytokine-based cancer therapies utilize one or more cytokines that regulate an individual's immune response. Non-limiting examples of cytokines suitable for cancer treatment include interferon-α (IFN-α), interleukin-2 (IL-2), granulocyte-macrophage colony stimulating factor (GM-CSF), and interleukin-12 (IL-12).

[0152] All references cited throughout this application (including literature references, issued patents, published patent applications, and co-pending patent applications) are hereby expressly incorporated by reference in their entirety.

[0153] Examples

[0154] Structure

[0155] hIgG1

[0156] This construct is the hIgG1 portion from the parental pFUSE-hIgG1 vector (Invitrogen).

[0157] hNKG2Dx2-hIgG1

[0158] Two copies of human NKG2D (F78-V216) with an amino acid spacer between them were cloned by unrestricted cloning 5' of the pFUSE-hIgG1 vector (Invitrogen). A schematic diagram of this construct is depicted in Figure 1 Center, left.

[0159] hNKG2D, reference sequence NP_031386.2, amino acids 78-216:

[0160] FLNSLFNQEVQIPLTESYCGPCPKNWICYKNNCYQFFDESKNWYESQASCMSQNASLLKVYSKEDQDLLKLVKSYHWMGLVHIPTNGSWQWEDGSILSPNLLTIIEMQKGDCALYASSFKGYIENCSTPNTYICMQRTV(SEQID NO:1)

[0161] hIgG1-X or hNKG2Dx2-hIgG1-X

[0162] Using the hIgG1 vector or the hNKG2Dx2-hIgG1 parent vector described above, various constructs were cloned 3' to the hIgG1 segment (indicated by "X" above). These constructs are described below.

[0163] hIL15 / hIL15Ra

[0164] A codon-optimized version of a portion of the human IL-15 receptor alpha (hIL15Ra, GenBank AAP69528.1, amino acids 31-107) was fused to a codon-optimized version of IL-15 (IL15, GenBank AAX37025, amino acids 22-135). hIL15Ra and hIL-15 are separated by twenty amino acids (G4S)4 (SEQ ID NO: 3). The amino acid sequence is shown below.

[0165] hIL15Ra, GenBank AAP69528.1, amino acids 31-107:

[0166] ITCPPPMSVEHADIWVKSYSLYSRERYICNSGFKRKAGTSSLTECVLNKATNVAHWTTPSLKCIRDPALVHQRPAPP(SEQ ID NO:8)

[0167] IL15, GenBank AAX37025, amino acids 22-135:

[0168] NWVNVISDLKKIEDLIQSMHIDATLYTESDVHPSCKVTAMKCFLLELQVISLESGDASIHDTVENLIILANNSLSSNGNVTESGCKECEELEEKNIKEFLQSFVHIVQMFINTS(SEQ ID NO:9)

[0169] aPD1

[0170] ScFv against mouse PD1 was cloned 3' of the hIgG1 segment in hIgG1 vector or hNKG2Dx2-hIgG1 vector.

[0171] aCTLA4

[0172] ScFv against mouse CTLA4 was cloned 3' of the hIgG1 segment in hIgG1 vector or hNKG2Dx2-hIgG1 vector.

[0173] Heterologous expression and purification

[0174] The indicated fusion constructs were generated in 293FT cells by calcium phosphate transfection of plasmids encoding the constructs. Supernatants were collected and the fusion constructs were purified by protein A chromatography. 1 μg of each construct (reduced with 2-mercaptoethanol and heated to 70° C. for 10 minutes or not) was loaded onto an 8-10% SDS-PAGE gel at 100 V for 60-120 minutes and proteins were visualized using Coomassie Blue staining. Figure 2 hNKG2Dx2-hIgG1-hIL15 / Ra was shown to be produced as a single fusion protein and was easily purified by Protein A.

[0175] Proliferation analysis

[0176] Human NK cells were isolated from normal donors using RosetteSep (StemCell Technologies). NK cells were labeled with 5 μM carboxyfluorescein succinimidyl ester (CFSE) (Invitrogen). The cells were then cultured for 4 days in RPMI+10% FBS with various dilutions of IL-15 constructs (hIgG1-hIL15 / Ra or hNKG2Dx2-20AA-hIgG1-hIL15 / Ra) or IL-15, and the CFSE dilutions were measured using FACSCanto (BD). Figure 3 Results are provided showing that incubation with hNKG2Dx2-hIgG1-IL15 / Ra promotes proliferation of human NK cells similar to incubation with IL-15.

[0177] Killing analysis

[0178] Human NK cells were isolated from normal donors using RosetteSep (StemCell Technologies) and frozen in BamBanker (Wako). NK cells were thawed and recovered overnight in RPMI + 10% FBS + 200 IU / mL hIL-2. The cells were then washed three times in PBS and added to various tumor targets previously labeled with CFSE (Invitrogen). The cells were centrifuged at 1000 rpm for 1 minute and co-cultured at 37 degrees in a humidified CO2 incubator for 5 hours. After 4 hours, 7-aminoactinomycin D (7-AAD) was added and tumor target cell death was analyzed.

[0179] Figure 4 hNKG2Dx2-hIgG1-IL15 / Ra was shown to promote efficient killing of multiple cell lines and was superior to hNKG2Dx2-hIgG1 in cell lines with moderate ligand expression. Panel A shows that the construct did not promote killing of the B16 tumor cell line that does not express NKG2D-L. Both hNKG2Dx2-hIgG1 and hNKG2Dx2-hIgG1-IL15 / Ra constructs promoted killing of cells expressing high levels of NKG2D ligand ( Figure 4 , Panel B). Various tumors express different levels of NKG2D ligands on their cell surfaces, as measured by NKG2D fusion protein binding ( Figure 4, Panel C). hNKG2Dx2-hIgG1 and hNKG2Dx2-hIgG1-IL15 / Ra constructs also promoted killing of tumors naturally expressing high levels of NKG2D ligands (K562), but the hNKG2Dx2-hIgG1-IL15 / Ra construct drove superior killing in a gradient manner that was inversely correlated with NKG2D ligand expression ( Figure 4 , Figure D).

[0180] IFNγ ELISA

[0181] Human NK cells were isolated, frozen, and added to tumor targets as described above, except that there was no overnight recovery period for NK cells (e.g., RPMI + 10% FBS + 200 IU / mL hIL-2) and the tumor targets were not labeled with CFSE. After 24 hours of co-culture, the plates were pelleted by centrifugation and the supernatant was aspirated for analysis by IFN-γ (Becton Dickinson).

[0182] Figure 5 Resting NK cells were shown to be activated by the fusion protein to produce IFN-γ, but maximal productivity required all three components: NKG2D, hIgG1 and IL- 15. It should be noted that N297Q is a mutation in hIgG1 that prevents CD16 (expressed on NK) from binding to hIgG1.

[0183] CD16 and IL-15 activation

[0184] Figure 6 showed that pre-activated NK cells (eg, cultured overnight with hIL-2) require CD16 binding for target cell killing, but not IL-15.

[0185] Human NK cells were isolated, frozen, and added to tumor targets as described above. There was no overnight recovery period for the NK cells. The tumor targets were labeled with CFSE. Figure 7 demonstrated that optimal activation of resting NK cells and killing by resting NK cells required CD16 engagement and IL-15 activation.

[0186] Improved features of dimeric NKG2D-Fc constructs

[0187] A protein model showing NKG2Dx2-hIgG1 in complex with the NKG2D ligand MICA is shown in Figure 13 Binding of NKG2Dx2-hIgG1 and hNKG2Dx1-hIgG1 constructs to MICA*008 was analyzed by ELISA and flow cytometry. Figure 8Presented are ELISA data demonstrating that NKG2Dx2-hlgG1 binds to MICA*008 with increased affinity compared to hNKG2Dx1-hlgG1. Figure 9 Depicted are flow cytometry data showing that hNKG2Dx2-hlgG1 binds to NKG2D ligand-expressing cells with increased avidity compared to hNKG2Dx1-hlgG1.

[0188] Figure 10 NKG2D-Fc was shown to drive NK cell killing of ligand-positive (e.g., NKG2D ligand-expressing) targets. Specifically, hNKG2Dx2-20AA-hlgG1 mediated significantly higher killing of B16 cells than hNKG2Dx1-20AA-hlgG1.

[0189] Figure 11 Demonstrated that hNKG2Dx2-hlgG1 enhances killing of NKG2D ligand expressing cells compared to hNKG2Dx1-hlgG1.B16 - ULBP2OE and HeyA8 cells tested.

[0190] Tumors can shed their NKG2D ligands, which further impairs immune responses. One possible solution to this problem is to "suck off" soluble NKG2D ligands (such as soluble MICA) to restore immune system function. Figure 12 Neutralization of NKG2Dx2-hIgG1 by displaying soluble MICA was superior to that of NKG2Dx1-hIgG1.

[0191] Equivalent

[0192] Those skilled in the art will recognize or be able to ascertain using no more than routine experimentation many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims. Sequence Listing <110> Dana-Farber Cancer Institute, Inc. <120> NKG2D-IG fusion proteins for cancer immunotherapy <130> D0504.70102WO00 <140> Not Yet Assigned <141> Concurrently Herewith <150> US 62 / 255,016 <151> 2015-11-13 <160> 9 <170> PatentIn version 3.5 <210> 1 <211> 139 <212> PRT <213> Homo sapiens <400> 1 Phe Leu Asn Ser Leu Phe Asn Gln Glu Val Gln Ile Pro Leu Thr Glu 1 5 10 15 Ser Tyr Cys Gly Pro Cys Pro Lys Asn Trp Ile Cys Tyr Lys Asn Asn 20 25 30 Cys Tyr Gln Phe Phe Asp Glu Ser Lys Asn Trp Tyr Glu Ser Gln Ala 35 40 45 Ser Cys Met Ser Gln Asn Ala Ser Leu Leu Lys Val Tyr Ser Lys Glu 50 55 60 Asp Gln Asp Leu Leu Lys Leu Val Lys Ser Tyr His Trp Met Gly Leu 65 70 75 80 Val His Ile Pro Thr Asn Gly Ser Trp Gln Trp Glu Asp Gly Ser Ile 85 90 95 Leu Ser Pro Asn Leu Leu Thr Ile Ile Glu Met Gln Lys Gly Asp Cys 100 105 110 Ala Leu Tyr Ala Ser Ser Phe Lys Gly Tyr Ile Glu Asn Cys Ser Thr 115 120 125 Pro Asn Thr Tyr Ile Cys Met Gln Arg Thr Val 130 135 <210> 2 <211> 5 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 2 Gly Gly Gly Gly Ser 1 5 <210> 3 <211> 20 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 3 Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly 1 5 10 15 Gly Gly Gly Ser 20 <210> 4 <211> 6 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 4 Gly Ser Gly Ser Gly Ser 1 5 <210> 5 <211> 8 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 5 Gly Gly Ser Gly Gly Gly Ser Gly 1 5 <210> 6 <211> 10 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 6 Gly Gly Pro Leu Gly Leu Trp Ala Gly Gly 1 5 10 <210> 7 <211> 4 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 7 Ile Glu Gly Arg 1 <210> 8 <211> 77 <212> PRT <213> Homo sapiens <400> 8 Ile Thr Cys Pro Pro Pro Met Ser Val Glu His Ala Asp Ile Trp Val 1 5 10 15 Lys Ser Tyr Ser Leu Tyr Ser Arg Glu Arg Tyr Ile Cys Asn Ser Gly 20 25 30 Phe Lys Arg Lys Ala Gly Thr Ser Ser Ser Leu Thr Glu Cys Val Leu Asn 35 40 45 Lys Ala Thr Asn Val Ala His Trp Thr Thr Pro Ser Leu Lys Cys Ile 50 55 60 Arg Asp Pro Ala Leu Val His Gln Arg Pro Ala Pro Pro 65 70 75 <210> 9 <211> 114 <212> PRT <213> Artificial sequence <220> <223> synthetic peptides <400> 9 Asn Trp Val Asn Val Ile Ser Asp Leu Lys Lys Ile Glu Asp Leu Ile 1 5 10 15 Gln Ser Met His Ile Asp Ala Thr Leu Tyr Thr Glu Ser Asp Val His 20 25 30 Pro Ser Cys Lys Val Thr Ala Met Lys Cys Phe Leu Leu Glu Leu Gln 35 40 45 Val Ile Ser Leu Glu Ser Gly Asp Ala Ser Ile His Asp Thr Val Glu 50 55 60 Asn Leu Ile Ile Leu Ala Asn Asn Ser Leu Ser Ser Asn Gly Asn Val 65 70 75 80 Thr Glu Ser Gly Cys Lys Glu Cys Glu Glu Leu Glu Glu Lys Asn Ile 85 90 95 Lys Glu Phe Leu Gln Ser Phe Val His Ile Val Gln Met Phe Ile Asn 100 105 110 Thr Ser

Claims

1. A recombinant DNA molecule encoding a recombinant polypeptide, the recombinant polypeptide sequentially comprising: (a) A first NKG2D fragment that retains the ability to bind its ligand; (b) A second NKG2D fragment that retains the ability to bind its ligand; and (c) The Fc domain of an immunoglobulin, the Fc domain being configured to allow dimerization and comprising: (i) A CH2 domain; and (ii) A CH3 domain.

2. The recombinant DNA molecule according to claim 1, wherein each of the first NKG2D fragment and the second NKG2D fragment comprises an extracellular fragment of the NKG2D receptor.

3. The recombinant DNA molecule according to claim 2, wherein the first NKG2D fragment and the second NKG2D fragment are identical.

4. The recombinant DNA molecule according to claim 2, wherein the first NKG2D fragment and the second NKG2D fragment are different.

5. The recombinant DNA molecule according to claim 1, wherein the NKG2D is human NKG2D.

6. The recombinant DNA molecule according to claim 2, wherein the NKG2D is human NKG2D.

7. The recombinant DNA molecule according to claim 3, wherein the NKG2D is human NKG2D.

8. The recombinant DNA molecule according to claim 4, wherein the NKG2D is human NKG2D.

9. The recombinant DNA molecule according to claim 1, wherein the immunoglobulin is a human immunoglobulin.

10. The recombinant DNA molecule according to claim 1, wherein the immunoglobulin is IgG1.

11. The recombinant DNA molecule according to claim 9, wherein the immunoglobulin is IgG1.

12. The recombinant DNA molecule according to claim 1, wherein the recombinant polypeptide comprises a linker between the first ligand-binding fragment and the second ligand-binding fragment.

13. The recombinant DNA molecule according to claim 12, wherein the linker is a flexible peptide linker.

14. The recombinant DNA molecule according to claim 13, wherein the linker comprises two or more of the following amino acids: glycine, serine, alanine, and threonine.

15. The recombinant DNA molecule according to claim 13, wherein the linker has a length of 25 amino acids or fewer.

16. The recombinant DNA molecule according to claim 13, wherein the linker has a length of 20 amino acids or fewer.

17. The recombinant DNA molecule according to claim 13, wherein the linker has a length of 12 amino acids or fewer.

18. The recombinant DNA molecule according to claim 1, wherein the Fc domain of the recombinant polypeptide further comprises a hinge region.

19. The recombinant DNA molecule according to claim 1, wherein the recombinant polypeptide does not comprise a cytokine.

20. A host cell comprising the recombinant DNA molecule according to any one of claims 1-19, wherein the host cell is a bacterium, a yeast cell, an insect cell or a mammalian cell.

21. The host cell according to claim 20, which is a mammalian cell.

22. A method for producing a recombinant polypeptide, comprising culturing the host cell according to claim 20 or 21.

23. The method according to claim 22, further comprising purifying the recombinant polypeptide.

24. A composition comprising the recombinant polypeptide obtained or obtainable by the method according to claim 22 or 23.

25. A recombinant polypeptide, the sequence of which comprises: (a) a first NKG2D fragment that retains the ability to bind its ligand; (b) a second NKG2D fragment that retains the ability to bind its ligand; and (c) an Fc domain of an immunoglobulin, the Fc domain being configured to allow dimerization and comprising: (i) a CH2 domain; and (ii) a CH3 domain.

Images