WO2022118043A1

WO2022118043A1 - Novel protein-ligand complex

Info

Publication number: WO2022118043A1
Application number: PCT/GB2021/053176
Authority: WO
Inventors: Houman Ashrafian; Magdalena VON ESSEN
Original assignee: Enara Bio Limited
Priority date: 2020-12-03
Filing date: 2021-12-03
Publication date: 2022-06-09

Abstract

There is provided inter alia an isolated MR1-ligand complex comprising (a) an MR1 protein and (b) a cyclic dinucleotide as ligand for the MR1 protein which is capable of being specifically bound by a T-cell receptor.

Description

Novel protein-ligand complex Field of the invention This invention relates inter alia to an isolated MR1-ligand complex, its use in raising and discovering antibodies, T-cell receptors and T-cells (and other killer cells), and use of these and related therapy modes (such as BiTEs and CAR-T-cells) in the treatment or prevention of cancer. Products for the treatment or prevention of autoimmune diseases are also provided. Background WO2019/081902 (Sewell et al) and Crowther et al (2020) describe a new class of T- cell (called “K43A sensitive MR1-T-cells”) effective for treating cancer, which recognize cancer cells through population-invariant major histocompatibility complex class related protein MR1. The identification of this new type of T-cell stemmed from experiments searching for T-cells recognising cancer cells without the requirement for a specific Human Leukocyte Antigen (HLA). The HLA locus is highly variable with over 17,000 different alleles having been described today. As such, any therapeutic approach that works via an HLA, can only be effective in a minority of patients. In contrast, the entire human population expresses MR1 which means that MR1 targeted TCRs should be effective across most of the population. The main type of MR1-restricted T-cells that are known are called mucosal-associated invariant T-cells (MAITs). MAITs are known to recognise intermediates of mycobacterial riboflavin biosynthesis. Ligands of MAIT cells are described in WO2015/149130 (Corbett et al) and include, for example, the substance 5-OP-RU. K43A sensitive MR1-T-cells, such as clone MC.7.G5 (disclosed in WO2019/081902), have target specificity via MR1, but the T-cell receptors (TCRs) of such T-cells do not bind to MR1 per se or to MR1 loaded with known bacterial ligands, rather, they apparently recognise a cancer-specific ligand within the MR1 binding groove; MR1 presents a cancer-specific, or cancer-upregulated, ligand to the TCR. A feature of clone MC.7.G5 as described in the Sewell/Crowther disclosures is the apparent importance of the K43 residue for TCR binding since MC.7.G5 did not significantly recognise target cells expressing the MR1 protein containing the K43A mutation (but not the wild type MR1). Different MR1-T-cells which do tolerate the K43A mutation have been described (Vacchini et al (2020)). The K43A-sensitive MR1-T-cells were found to target a wide range of cancer cell lines implying that the cancer-specific, or cancer-upregulated, ligand is common to a wide range of different types of cancers. There is great interest in identifying the ligand for the TCR of MR1-T-cells including K43A-sensitive MR1-T-cells. Knowledge of the identity of this ligand would enable further TCRs and antibodies to be raised, as well as other therapeutic modalities, all of which could be useful e.g. in the treatment or prevention of cancer. The work of the present inventors has identified such a ligand. Summary of the invention The inventors have identified that cyclic dinucleotides are ligands for MR1 and the MR1-dinucleotide complex is recognised by K43A sensitive MR1-T-cells and K43A sensitive MR1-T-cells bind to the complex. Structural modelling studies described in the Examples section shows that cyclic dinucleotides bind in the ligand binding pocket formed in the groove between the α1 and α2 domains of the heavy chain of MR1. More particularly, cyclic dinucleotides can adopt an “open” and a “closed” conformation, as illustrated for 2’,3’-cGAMP, and the cyclic dinucleotides can bind in the ligand binding pocket formed in the groove when in the extended conformation but not in the closed conformation (Table 1). Without being limited by theory it is believed that cancers either constitutionally, through genomic instability (e.g. DNA damage response, radiotherapy, chemotherapy, immunotherapy or otherwise), release double-stranded DNA (“dsDNA”) and this dsDNA aberrantly accumulates in the cytosol of tumor cells. Enzymes such as cyclic GMP-AMP synthase (cGAS) bind DNA in the cytosol and generate cyclic dinucleotides such as 2’,3’-cGAMP, a soluble second-messenger that activates STING (stimulator of interferon genes) resulting in the downstream production of type 1 interferons and other cytokines. STING, Stimulation of Interferon Genes (STING), is an ER-resident protein encoded by TMEM173. As a canonical cyclic dinucleotide sensor, on exposure to its ligand, STING multimerises (e.g. dimerises) and is translocated to different cellular compartments, whereupon it activates downstream pathways (e.g., TANK-binding kinase 1 (TBK1) which further phosphorylates interferon regulatory transcription factor 3 (IRF3); NF-κB is also liberated). STING is an important mediator of the innate immune response to pathogens and in cancer. Cyclic dinucleotides such as 2’,3’-cGAMP are widely distributed intra- and extracellularly, to neighbouring cells via gap junctions or extracellularly through vesicles or through a variety of transporters (e.g. LRRC8 and SLC19A1). MR1 has the capacity to load a variety of potential ligands, mainly but not limited to metabolites, either by binding to de novo synthesized MR1 or through binding to previously synthesized forms which continuously recycle to the cell surface, presenting ligands. MR1 has broad low-level expression across multiple cancer and tissue types. These cyclic dinucleotides can access MR1 rich compartments and in a steric conformation that has a stabilising affinity through binding in the MR1 ligand- binding pocket; stabilise the MR1 structure leading it to traffic to the plasma membrane and present the antigen; thus, cyclic dinucleotides are ligands for MR1 and are recognised by K43A sensitive MR1-T-cells such as MC.7.G5. Thus, according to the invention, there is provided an isolated MR1-ligand complex comprising (a) an MR1 protein and (b) a cyclic dinucleotide as ligand for the MR1 protein which is capable of being specifically bound by a T-cell receptor. Brief description of the figures Figure 1: 2’,3’-cGAMP in semi-extended conformation (black stick) docked to MR1 structure (grey ribbon, PDB ID 4pj9). Model of the ligand created in PyMol Version 2.4.1. by scrambling unconstrained coordinates option of the Sculpt function Figure 2: 3’,3’-cGAMP in extended conformation (black stick, PDB ID 6ael) docked to MR1 structure (grey ribbon, PDB ID 4pj9) Figure 3: 5-OP-RU (black stick) bound to MR1 (grey ribbon) in original structure (PDB ID 4pj9) Figure 4: Detail of MR1 binding pocket with crucial interactions between MR1 residues and docked ligand 2’,3’-cGAMP in semi-extended conformation (modelled by PyMol by scrambling unconstrained coordinates) (A); 3’,3’-cGAMP in extended conformation (PDB ID 6ael) (B); 5-OP-RU as in its original crystal structure bound to MR1 (PDB ID 4pj9) (C) Figure 5: Structures of 3’,3’-cGAMP (A) and 2’,3’-cGAMP (B) Figure 6: 2’,3’-cGAMP (A) and 3’,3’-cGAMP (B) closest distance to K43 (side chain N atom) within MR1 binding pocket, values in Å Figure 7: 2’,3’-cGAMP (grey stick) and 3’,3’-cGAMP (black stick) representation within MR1 binding pocket. Figure 8. Comparative IFNγ production in C1R.MR1 cells post T cell activation assay with cyclic dinucleotide. Figure 9. Fold change relative to untreated C1R.MR1 cells in IFNγ production in C1R.MR1 cells + T cell control for post T cell activation assay. Figure 10. Effect of 2,3-cGAMP +/- MEG on MR1 surface expression; graph of fold change for MR1 expression in C1R.MR1 cells (fold change of the geometric mean compared to vehicle). Detailed description of the invention Definitions Suitably, the polypeptides and polynucleotides used in the present invention are isolated. An “isolated” polypeptide or polynucleotide is one that is removed from its original environment. For example, a naturally occurring polypeptide or polynucleotide is isolated if it is separated from some or all of the coexisting materials in the natural system. A polynucleotide is considered to be isolated if, for example, it is cloned into a vector that is not a part of its natural environment. A polypeptide is isolated if it is, for example, a recombinant polypeptide. "Naturally occurring", “natural” or “native”, which terms are interchangeable, when used with reference to a polypeptide or polynucleotide sequence means a sequence found in nature and not synthetically modified. The term “artificial” when used with reference to a polypeptide or polynucleotide sequence means a sequence not found in nature i.e. not natural, which is, for example, a synthetic polypeptide or protein or polynucleotide, a synthetic modification of a natural sequence, or contains an unnatural polypeptide or polynucleotide sequence, or for example comprises modified subunits such as modified or derivative forms of amino acids or nucleotides, such as modified bases, modified elements or modified linkages between such elements, amino acids or nucleotides. The term “engineered” when used with reference to a polypeptide (such as a TCR) or cell means a polypeptide or cell not found in nature which is, for example, a synthetic polypeptide or protein or polynucleotide, a synthetic modification of a natural polypeptide or cell, for example, because it contains or expresses foreign residues or elements and/or lacks natural residues or elements. The term “heterologous” or “exogenous” when used with reference to the relationship of one polynucleotide or polypeptide to another polynucleotide or polypeptide indicates that the two or more sequences are not found in the same relationship to each other in nature. The term “heterologous” when used with reference to the relationship of one polynucleotide or polypeptide sequence to a cell means a sequence which is not isolated from, derived from, expressed by, associated with or based upon a naturally occurring polynucleotide or polypeptide sequence found in, or endogenous to, the said cell. As used herein, the term “chimeric” means, in the context of a polypeptide or polynucleotide, an artificial polypeptide or polynucleotide that is engineered to contain elements (e.g. sequences of amino acids or nucleotides) of more than one origin. The term “domain”, when used with reference to a TCR or T cell receptor protein as used herein, is generally used to refer to a part of the TCR or T cell receptor protein formed of the corresponding region of the two chains comprising said domain. For example, the transmembrane regions of the α and β chains of αβTCRs form the transmembrane domain. The term “T-cell effector function domain” means a domain associated with the effector function of the T-cell as opposed to a domain associated with the target binding function of the T-cell and includes, for example, a CD3zeta intracellular signalling domain or a co-stimulatory domain. Other types of immune cell receptor (e.g. for NK cells or NKT cells) have analogous immune cell effector function domains. The term “intracellular” domain or region is used interchangeably with the term “cytoplasmic” domain or region and in the literature this is sometimes referred to as the “cytosolic” domain or region. As used herein, the term “polynucleotide” means a polymeric macromolecule made from nucleotide monomers particularly deoxyribonucleotide or ribonucleotide monomers in natural form or in the form of analogues comprising one or more unnatural backbone residues, linkages or bases. Typically, a polynucleotide which is a DNA comprises units composed of deoxyribose, phosphate and bases selected from guanine, adenine, cytosine and thymidine. Typically, a polynucleotide which is an RNA comprises units composed of ribose, phosphate and bases selected from guanine, adenine, cytosine and uracil. As used herein, the term “dinucleotide” means a nucleotide consisting of two units, each unit composed of a phosphate, a ribose and a nitrogen base. A “cyclic dinucleotide” is a dinucleotide in which the ribose groups of the two units contain two linkages via phosphate groups to form a cyclic structure. As used herein, the term “MR1 protein” means the protein MR1 (UniProt accession no. Q95460) from human or other mammal (such as mouse, rat or bovine), particularly from human, as well as derivatives and fragments thereof which are capable of binding ligand as heavy chain when non-covalently associated with a β2- microglobulin protein as light chain and thus comprising, in particular, the α1, α2 and α3 domains of the heavy chain. Other variants of non- β2-microglobulin MR1s are also described and may present cyclic dinucleotides (Lion et al (2013). The polypeptide sequence of human MR1 is provided herein as SEQ ID NO: 1: MGELMAFLLP LIIVLMVKHS DSRTHSLRYF RLGVSDPIHG VPEFISVGYV DSHPITTYDS VTRQKEPRAP WMAENLAPDH WERYTQLLRG WQQMFKVELK RLQRHYNHSG SHTYQRMIGC ELLEDGSTTG FLQYAYDGQD FLIFNKDTLS WLAVDNVAHT IKQAWEANQH ELLYQKNWLE EECIAWLKRF LEYGKDTLQR TEPPLVRVNR KETFPGVTAL FCKAHGFYPP EIYMTWMKNG EEIVQEIDYG DILPSGDGTY QAWASIELDP QSSNLYSCHV EHCGVHMVLQ VPQESETIPL VMKAVSGSIV LVIVLAGVGV LVWRRRPREQ NGAIYLPTPD R [SEQ ID NO;1] As used herein, the term “β2 microglobulin protein” means the protein β2 microglobulin (UniProt accession no. P61769) from human or other mammal (such as mouse, rat or bovine), particularly from human, as well as derivatives and fragments thereof which are capable of binding ligand, as light chain when non- covalently associated with an MR1 protein as heavy chain. The polypeptide sequence of human β2 microglobulin protein is provided herein as SEQ ID NO: 2: MSRSVALAVL ALLSLSGLEA IQRTPKIQVY SRHPAENGKS NFLNCYVSGF HPSDIEVDLL KNGERIEKVE HSDLSFSKDW SFYLLYYTEF TPTEKDEYAC RVNHVTLSQP KIVKWDRDM [SEQ ID NO: 2] As used herein, the term “ligand” means a binding molecule. A ligand for MR1 is a molecule capable of binding to MR1. In the context of the present invention, a ligand for MR1 may be associated with, or capable of associating with, a binding pocket on the MR1 molecule. Such a binding pocket will typically have a size and shape and contain appropriate residues for the binding of the ligand. The ligand may be a cyclic dinucleotide as herein described. In the context of the present invention the term “binding” will typically mean non-covalent binding i.e. means binding by means of ionic, hydrophobic or Van-de-Waals interactions however binding covalently e.g. via Schiff base linkages is not excluded, for example with residue K43 of MR1. As used herein, the term “MR1-ligand complex” means a complex formed by the binding of a ligand for MR1 to MR1, specifically a binding in the binding pocket for said ligand. Binding of ligand to or within the binding groove or pocket of MR1 may constitute specific binding for example such that the ligand is presented for recognition by a further binding partner, e.g. by an immune cell receptor In this form, the MR1-ligand complex presents the ligand as an antigen that is capable of being recognised by immune cell receptors, such as T cells, T cell receptor proteins or T cell TCRs. As used herein, the term “immune cell receptor” means a receptor that is capable of being expressed on the surface of an immune cell and that has a binding function i.e. the function of binding to a target antigen epitope on a target cell and an effector function i.e. the function of eliciting a functional behaviour of the immune cell, such as cell killing or recruitment of other cells, such as other cells of the immune system in response to the binding event. When an immune cell receptor binds to a target antigen epitope, for example an MR1 presented ligand, on a target cell (e.g. a cancerous cell) and the effector function results, this may be referred to herein as “recognition” i.e. the immune cell receptor recognises the target antigen epitope on the target cell. The immune cell may be a T-cell in which case the immune cell receptor is a T-cell receptor (TCR). As used herein, the term “T-cell receptor protein” means a protein or complex of proteins which is T-cell receptor formed of two chains, typically an α chain and a β chain in the case of αβ T-cells or a γ chain and a δ chain in the case of γδ T-cells, or a fragment of said receptor, capable of recognising a target antigen epitope. The chains of a T-cell receptor protein typically comprise a variable region and a constant region. The variable region of a chain which, when the two chains are paired binds to the target antigen epitope, typically contains 3 CDRs and 4 framework regions. The constant region typically comprises an extracellular region, a connecting peptide region, a transmembrane region and an intracellular region. Depending on the context, the receptor protein may be useful in soluble form i.e. in the form of a fragment of a T-cell receptor in which the constant region comprises an extracellular region but lacks a transmembrane region and an intracellular region. The soluble form of a T-cell receptor protein lacks an effector function. Single chain formats where the two chains or fragments thereof are linked to form a single polypeptide and which are capable of recognising a target antigen epitope are also embraced by the term “T-cell receptor protein”. As used herein, the term “immune cell” includes T-cell, NK-cell and NKT-cell. For example, according to the present invention the immune cells can be cells of the lymphoid lineage, comprising B, T or natural killer (NK) cells. The immune cells may be cells of the lymphoid lineage including T cells, Natural Killer T (NKT) cells, γδ T- cells and precursors thereof including embryonic stem cells, and pluripotent stem cells (e.g, those from which lymphoid cells may be differentiated). T cells can be lymphocytes that mature in the thymus and are chiefly responsible for cell-mediated immunity and also involved in the adaptive immune system. According to the present invention the T cells can include, but are not limited to, helper T cells, cytotoxic T cells, memory T cells (including central memory T cells, stem-cell-like memory T cells (or stem-like memory T cells), and two types of effector memory T cells: e.g. , TEM cells and TEMRA cells, Regulatory T cells (also known as suppressor T cells), Natural killer T cells, Mucosal associated invariant T cells, and gamma-delta T cells. Cytotoxic T cells (CTL or killer T cells) are a subset of T-lymphocytes capable of inducing the death of infected somatic or tumour cells. A subject’s own T cells may be genetically modified to target specific antigens through the introduction of a heterologous TCR. Preferably, the modified immune cell is a T cell optionally a CD4⁺T cell or a CD8⁺T cell. Accordingly, the immune cells may be T-cells, optionally CD4+ T cells or CD8+ T cells, or the immune cells may be a population of modified T-cells, optionally CD4+ T cells; or CD8+ T cells, or a mixed population of CD4+ T cells and CD8+ T cells. As used herein, the term “T-cell” includes CD4+ T-cells and CD8+ T-cells and especially includes cytotoxic T-cells and γδ T-cells. As used herein, the term “NK-cell” includes umbilical cord blood NK cells, iPSC derived NK cells, PMBC derived NK cells and NK cells from NK92, YTS and NKL cell lines. As used herein, the term “NKT-cell” includes cells that have properties of both T cells and NK cells. These cells are traditionally thought to recognise the non- polymorphic CD1d molecule presenting self and foreign lipids and glycolipids. They constitute only approximately 1% of all peripheral blood T cells. As used herein, the term “immune cell engaging protein” means an artificial bispecific molecule being capable of binding (e.g. specifically binding) (i) an antigen and (i) a molecule expressed on the surface of an immune cell. Typically, an immune cell engaging protein binds an extracellular part of an immune cell receptor protein expressed on the surface of an immune cell, for example to any one or more of CD3, CD5, CD16, CD16a, CD16b, NKp30, NKp46, NKG2D and DNAX Accessory Molecule-1 (DNAM-1) on immune cells for example on T-cells, T-lymphocytes or NK cells, As used herein, the term “chimeric immune cell receptor protein” means an artificial molecule comprising the effector domains of an immune cell receptor and having the binding region of an antibody protein. Typically, a chimeric immune cell receptor protein comprises the effector domains of an immune cell receptor (including the transmembrane and intracellular domains of an immune cell receptor) fused to an scFv, said scFv being capable of specifically binding an antigen on a target cell. As used herein, the term “MR1-T-cell” means a T-cell expressing on its surface a T- cell receptor which binds to and/or recognises an MR1-ligand complex. The term “K43A sensitive MR1-T-cell” means an MR1-T-cell which, when its T-cell receptor comprises the K43A mutation, is incapable of (or substantially impaired in) binding to and/or recognising the MR1-ligand complex normally bound and/or recognised by an MR1-T-cell. As used herein, the term “specifically binding” in relation to the binding of A to B means that A binds to B, for example at or within a respective specific binding site, domain or pocket, with an affinity typically associated with the binding of ligands to receptors or typically associated with molecules of the immune system, such as antibodies and T-cell receptors, optionally of the binding affinity level of micromolar or nanomolar affinity, such that the affinity of binding of A to B greatly exceeds that of the binding of A to other molecules not intended to be targeted by A. The term “being specifically bound” is to be interpreted in a similar sense. As used herein, the term “antibody protein” means an antibody, an antibody fragment, an antibody conjugated to an active moiety, a fusion protein comprising one or more antibody fragments, or a derivative of any of the aforementioned. Antibody fragments are most suitably antigen binding fragments of antibodies. Examples of derivatives include conjugated derivatives e.g. an antibody or antibody fragment conjugated to another moiety. Such moieties include chemically inert polymers such as PEG. Antibodies may include monoclonal antibodies and polyclonal antibodies, preferably monoclonal antibodies. The monoclonal antibodies can be, for example, mammalian (e.g. murine) or avian, chimeric or reverse chimeric, for example, human/mouse or human/primate chimeras, humanized antibodies or fully human antibodies. Antibodies may be produced in a non-human species (e.g., rodent) genetically modified to have elements of a human immune system. Suitable antibodies include an immunoglobulin, such as IgG, including IgG₁, IgG₂, IgG₃ or IgG₄, IgM, IgA, such as IgA₁ or IgA₂, IgD, IgE or IgY. Suitable antibodies also include single chain antibodies. Also included are antibody fragments including Fab, Fab2, scFv fragments and the like, scFv-Fc, single domain antibody, diabody, dsFv, Fab', minibody, diabody, single-chain antibody molecules especially such fragments which bind antigens. Also embraced are single domain antibodies and heavy chain only antibodies derived from Camelids (e.g. llamas) and sharks and fragments thereof such as the variable portion (VHH). As used herein, the term “4-chain monoclonal antibody” means a monoclonal antibody of conventional type having two heavy chains and two light chains, said heavy and light chains being paired to form at their extremities two variable regions which each constitute different antigen-binding sites. As used herein, the term “scFv” means a single chain variable fragment that may be engineered to substitute for one of the antigen-binding domains of a 4-chain monoclonal antibody in which the variable region of the heavy chain (V_H) is connected by a linker to the variable region of the light chain (VL). As used herein, the term “cancerous” means, in relation to a cell, a cell of malignant character typically associated with the behaviour of uncontrolled or dysregulated proliferation. Cyclic dinucleotides Exemplary bases of dinucleotides include cytosine, adenine, guanine and thymidine as well as derivatives thereof. Typically, one linkage of the two units of a dinucleotide is 3’ to 5’ (i.e. a 3’ phosphate of the first unit is bound to a hydroxyl of the second unit. Typically, a second linkage of the two units is 2’ or 3’ to 5’ i.e. between either the 2’ or 3’ phosphate of the second unit and a 5’ hydroxyl of the first unit. Preferably the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP-AMP (cGAMP), optionally the linkage between the nucleosides or nucleotide units is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’, optionally the cyclic dinucleotide is in extended or semi-extended conformation as described herein. MR1-ligand complex As noted above, the invention provides an isolated MR1-ligand complex comprising (a) an MR1 protein and (b) a cyclic dinucleotide as ligand for the MR1 protein which is capable of being specifically bound by a T-cell receptor (herein after the “isolated complex of the invention”). Suitable MR1 in the complex comprises the heavy chain of MR1 and thus, in particular, comprises the α1, α2 and α3 domains of MR1, optionally wherein the MR1 has an amino acid sequence which is any one of 90%,91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NO:1. Suitably the heavy chain of MR1 in the complex is non-covalently associated with a β2 microglobulin protein as light chain. Alternatively, an artificial single-chain construct may be created in which the heavy chain of an MR1 protein in the complex is covalently linked to a β2 microglobulin protein for example via a linker peptide, optionally wherein the β2 microglobulin protein has an amino acid sequence which is any one of 90%,91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NO:2. The cyclic dinucleotide can be, for example, a cyclic dinucleotide which comprises nucleotides comprising a base selected from A and G. In particular, the cyclic dinucleotide is a cyclic GMP-AMP (cGAMP). In an embodiment the cyclic dinucleotide is 2’,3’-cGAMP. In another embodiment the cyclic dinucleotide is 3,3- cGAMP. The structures of 2’,3’-cGAMP and 3’,3’-cGAMP are shown in Figure 5. For use in anti-cancer applications described below (i.e. targeting said complex with an immune cell receptor protein or immune cell presenting such receptor protein, an immune cell engaging protein, or an antibody) suitably the cyclic dinucleotide is a cyclic dinucleotide as herein described, preferably it is 2’,3’-cGAMP. Optionally the cyclic dinucleotide as herein defined is capable of combining with, for example to become bonded to or form a bond between, for example by condensation, an alpha-dicarbonyl compound, for example methylglyoxal or glyoxal, to bind to MR1 or MR1 binding pocket, for example through a covalent bond with Lys43 of the MR1 binding pocket, preferably to form a Schiffs base. Hence the MR1-ligand complex according to the invention may comprise (a) an MR1 protein and (b) a cyclic dinucleotide combined with or in combination with or bound to an alpha-dicarbonyl compound as ligand for the MR1 protein which is capable of being bound and/or specifically bound by a T-cell receptor, optionally wherein the T-cell receptor is an MR1 specific T-cell receptor, preferably wherein the cyclic dinucleotide is as herein defined (e.g. selected from any one of cyclic GMP-GMP, cyclic AMP- AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP-AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’), preferably either 2’,3’-cGAMP or 3’,3’-cGAMP, preferably 2’,3’-cGAMP, optionally wherein the alpha-dicarbonyl compound is methylglyoxal or glyoxal. Antibody proteins The invention also provides an antibody protein which specifically binds to the isolated complex of the invention (herein after an “antibody protein of the invention”). In one embodiment the antibody protein is a 4-chain monoclonal antibody. In another embodiment the antibody protein is an scFv. In another embodiment the antibody protein is a construct which comprises two or more scFvs, for example, two or more scFvs linked in series. Antibody proteins of the invention may be raised by conventional methods, e.g. phage/yeast display or involving immunisation an experimental animal (such as rabbit, mouse, rat, guinea pig, hamster, llama etc.) with an MR1-ligand complex as immunogen, optionally together with an immunostimulant (e.g. specol), and obtaining antibodies or antibody producing cells from said animal (e.g. from PBMCs or from spleen). Monoclonal antibodies may be obtained by fusing said antibody producing cells with immortal cells to generate corresponding antibody producing hybridomas therefrom. Antibody proteins can be selected and cloned using conventional phage or yeast display technology and can be modified (e.g. to humanise or to introduce stability conferring mutations) and produced by conventional molecular biology and genetic engineering technology. Antibodies which are partially or completely human can also be produced in experimental animals which have genes of the human immune system. Immune cell engaging proteins The invention also provides an immune cell engaging protein which is capable of targeting a cell expressing a cell presenting on its surface an MR1-ligand complex comprising (a) a cell targeting portion comprising an antibody protein of the invention and (b) an immune cell engaging portion. The immune cell may be as herein described and for example may be a T-cell, an NK-cell or an NKT cell and in particular is a T-cell. Thus, an embodiment, the immune cell engaging protein is a bispecific T-cell engaging protein (BiTE). The immune cell engaging portion suitably comprises an antibody protein which is capable of binding and/or specifically binding to any one or more of CD3, CD5, CD16, CD16a, CD16b, NKp30, NKp46, NKG2D and DNAX Accessory Molecule-1 (DNAM-1) on immune cells for example on immune cells, for example T-cells, T-lymphocytes or NK cells, preferably CD3 on immune cells. Anti-CD3 monoclonal antibodies are known in the art e.g. muromonab-CD3, otelixizumab, teplizumab and visilizumab. The antibody protein which is capable of specifically binding to CD3 on immune cells may for example be an scFv e.g. a scFv derived from one of the aforementioned anti-CD3 monoclonal antibodies. The antibody portion of the cell targeting portion comprising an antibody protein of the invention may for example be an scFv. Preferably the immune cell engaging protein is a bispecific T-cell engaging protein The bispecific T-cell engaging protein may be of a formula: X-Ll '-Y or a formula: Y-Ll '-X, wherein: X comprises an antibody or antibody fragment recognising the immune cell (immune cell engaging portion ); LI ' comprises the one or more linkers; and Y comprises a second antibody or antibody fragment recognising a targeted MR1-ligand complex, particularly the MR1 ligand complex according to the invention, particularly as expressed and/or presented on the surface of a cell, for example a cancer, cancerous or tumour cell. X and/or Y may comprise a human, human engineered, humanized, chimeric antibody or fragment. X and/or Y may comprise a human engineered antibody or antibody fragment. X may comprise a fully human antibody or fully human antibody fragment. X and/or Y may comprise any one of; one or more Fv, one or more Fc, one or more Fab, one or more (Fab')2, one or more single chain Fv (scFv), one or more diabodies, or more triabodies, one or more tetrabodies, one or more bifunctional hybrid antibodies, one or more CDR1, one or more CDR2, one or more CDR3, one or more combinations of CDR's, one or more variable regions. X and/or Y may comprise one or more framework regions. X and/or Y may comprise one or more constant regions. X and/or Y may comprise one or more heavy chains. X and/or Y may comprise one or more light chains. X and/or Y may comprise one or more and variable regions. X and/or Y may comprise one or more alternative scaffold non-antibody molecules. X and/or Y may comprise a combination of any of Fv, Fc, Fab, (Fab')2, single chain Fv (scFv), diabodies, triabodies, tetrabodies, bifunctional hybrid antibodies, CDR1, CDR2, CDR3, combinations of CDR's, variable regions, framework regions, constant regions, heavy chains, light chains, and variable regions, alternative scaffold non-antibody molecules. Preferably X and/ or Y may comprise a Fab fragment or single chain Fv (scFv), preferably single chain Fv (scFv). Y may comprise at least a portion of an antibody, preferably scFv or Fab fragment. Y may comprise at least a portion of an antibody or antibody fragment that binds to an antigen on a lymphocyte or to an antigen on a B- cell or B-cell progenitor or to an antigen on a cancerous, cancer or tumour cell, particularly wherein said cell or cells present the MR1-ligand complex according to the present invention. X may comprise at least a portion of an antibody or antibody fragment that binds to an antigen selected from any of CD3, CD5, CD16, CD16a, CD16b, NKp30, NKp46, NKG2D and DNAX Accessory Molecule-1 (DNAM-1) on immune cells for example on T-cells, T-lymphocytes or NK cells, preferably CD3 on immune cells Chimeric immune cell receptor protein The invention also provides a chimeric immune cell receptor protein which comprises (a) a cell targeting portion comprising an antibody protein according to the invention and (b) a portion comprising immune cell effector function domains (hereinafter a “chimeric T-cell receptor protein of the invention”). The cell targeting portion is extracellular when the receptor protein is expressed at the surface of an immune cell and the portion comprising immune cell effector function domains is intracellular when the receptor protein is expressed at the surface of an immune cell. Typically, the chimeric immune cell receptor protein (e.g. chimeric antigen receptor (CAR) proteins) will comprise a transmembrane portion e.g. derived from CD8, CD16 or CD28. The transmembrane portion is transmembrane when the receptor protein is expressed at the surface of an immune cell. Chimeric NK receptor proteins may alternatively comprise a transmembrane portion e.g. derived from DAP12, 2B4, NKp44, NKp46 or NKG2D. The immune cell may, for example be a T-cell, an NK-cell or an NKT cell and in particular is a T-cell. Suitably the portion comprising an immune cell effector function domain comprises a CD3zeta intracellular signalling domain. Suitably the portion comprising an immune cell effector function domain also comprises one or more co-stimulatory domains. Co-stimulatory domains may for example be or be derived from the intracellular portions of proteins selected from CD28, OX40, 4-1BB, DAP12, DAP10 and 2B4, especially selected from the intracellular portions of CD28, OX40 and 4-1BB, For chimeric antigen receptor proteins designed for use in T cells, said co- stimulatory domains may for example be or be derived from the intracellular portions of proteins selected from CD28, OX40 or 4-1BB (such as a combination of the intracellular portions of CD28 and 4-1BB or CD28 and OX40). For chimeric antigen receptor proteins designed for use in NK cells, said co-stimulatory domains are for example selected or derived from the intracellular portions of CD28, 4-1BB, DAP12, DAP10 and 2B4. For chimeric antigen receptor proteins designed for use in NKT cells, said co-stimulatory domains may for example be or be derived from the intracellular portions of proteins selected from CD28 and 4-1BB (such as a combination of CD28 and 4-1BB). Chimeric immune cell receptor proteins may also be expressed by immune cells in conjunction with cytokines e.g. IL-2, IL-5 or IL-12 or other costimulatory ligands. Further details of the domain structure and generation of chimeric immune cell receptor proteins may be gleaned from Xie et al (2020). T-cell receptor proteins The invention also provides an isolated T-cell receptor protein which is capable of specifically binding to the isolated complex of the invention which T-cell receptor protein is not a T-cell receptor protein described in WO2019/081902 (University College Cardiff Consultants Ltd) such as MC.7.G5, specifically an antibody protein having an alpha and a beta chain having the sequences recited in Figure 3 thereof (hereinafter a “T-cell receptor protein of the invention”). Suitably it is not a T-cell receptor protein described in European patent application no.20192986.6 (University College Cardiff Consultants Ltd) filed on 20 August 2020. The T-cell receptor protein is also not a T-cell receptor protein described in WO2018/162563 (Universität Basel) since the T-cell receptor proteins described in this patent application are believed not to be T-cell receptors of K43A sensitive MR1-T-cells. The differences include that the Basel MR1-T-cells are: (1) not cancer- specific; (2) they activate in response to monocyte-derived DC (see B in Figure 8 of Lepore et al (2017)) and (3) do not require Lysine 43 for target recognition (see Figure 4B). These MR1-T-cells are common and supposedly occur at a frequency between 1:2500 and 1:5000 whereas K43A sensitive MR1-T-cells are rarer. T-cell receptor proteins of the invention may be raised and selected by priming and stimulating primary T-cells, particularly CD8+ T-cells, with antigen presenting cells (APCs) loaded with the MR1 ligand, for example a cyclic dinucleotide as herein described, or naturally expressing the MR1-ligand complex, and selecting T-cells that respond specifically to the APCs. APCs include dendritic cells, B-cells, and immortalized cell lines. Such T-cell cloning methods have been described. Alternatively, multimerization of the soluble, refolded, MR1-ligand complex could be used to directly label T-cells with such T-cell receptor proteins, for example using tetramers or dextramers. Additionally, soluble T-cell receptor protein libraries could be generated, using methods such as phage display, and the soluble, refolded MR1- ligand monomers could be used to pan these libraries to select and enrich for specific T-cell receptor proteins. T-cell receptor proteins of the invention may also be raised and selected by identification of T cell clones from human donors that display activity against (including killing of) cancer cells that are of different HLA types, and then confirming that the reactivity of such cells is dependent on MR1 expression, but not the HLA types expressed by the tumor cells (see Crowther et al. (2020)). Polynucleotides The invention provides isolated polynucleotide which encode proteins of the invention described herein (hereinafter “a polynucleotide of the invention”). For example, the invention provides an isolated polynucleotide which encodes a chimeric immune cell receptor of the invention. The invention also provides an isolated polynucleotide which encodes a T-cell receptor protein according to the invention. Vectors The invention provides a vector comprising a polynucleotide of the invention (hereinafter a “vector of the invention”). Suitably the vector is a viral vector, such as a lentiviral vector or a retroviral vector (e.g. ^-retrovirus). Other examples of viral vectors include vectors derived from adenovirus, adeno-associated virus (AAV), alphavirus, herpes virus, arenavirus, measles virus, poxvirus or rhabdovirus. DNA molecules, for example transposons, may also be suitable vectors to transduce T- cells with TCR genes. Vectors should suitably comprise such elements as are necessary for permitting transcription of a translationally active RNA molecule in the host cell, e.g. T-cell, such as a promoter and/or other transcription control elements such as an internal ribosome entry site (IRES) or a termination signal. A “translationally active RNA molecule” is an RNA molecule capable of being translated into a protein by the host cell’s translation apparatus. Engineered immune cells The invention also provides an engineered immune cell which expresses on its surface a chimeric immune cell receptor protein of the invention or T-cell receptor protein of the invention. Suitably the engineered immune cell is an engineered T- cell and the chimeric immune cell receptor protein is a chimeric antigen receptor protein. The invention also provides an engineered immune cell which expresses on its surface a T-cell receptor protein of the invention. Thus, suitably the engineered cell is a CAR-T-cell or TCR T-cell. The engineered immune cell may, for example be an engineered T-cell, NK-cell or NKT cell and in particular is an engineered T-cell, preferably an engineered T-cell engineered to express on its surface a TCR or chimeric immune cell receptor protein. For example, the invention provides an engineered immune cell (such as an engineered T-cell) which is transduced with a vector of the invention. This vector may be, for example a lentiviral vector encoding the T-cell receptor protein or derivative thereof or chimeric immune cell receptor protein or derivative thereof, as a transgene. Non-viral delivery vectors may also be used for example, transposons. Enriched T-cells may be activated with specific reagents such as antibodies to CD3 and CD28 or other known activation markers and transduced with the vector encoding the T-cell receptor protein or derivative thereof or chimeric immune cell receptor protein or derivative thereof, using established methods. Treatment or prevention of cancer The invention provides a method of treatment or prevention of cancer which comprises administering to a patient in need thereof an immune cell engaging protein of the invention. Suitably the immune cell engaging protein is a T-cell engaging protein. The invention also provides a method of treatment or prevention of cancer which comprises administering to a patient in need thereof an engineered immune cell of the invention. Suitably the engineered immune cell is an engineered autologous immune cell. Suitably the immune cell is a T-cell such as an engineered autologous T-cell. The invention also provides an immune cell engaging protein of the invention for use in the treatment or prevention of cancer. It also provides use of an immune cell engaging protein of the invention in the manufacture of a medicament for the treatment or prevention of cancer. The invention also provides an engineered immune cell of the invention for use in the treatment or prevention of cancer. It also provides use of an engineered immune cell of the invention in the manufacture of a medicament for the treatment or prevention of cancer. Thus, in a further aspect of the present invention, there is provided an ex vivo process comprising (i) obtaining immune cells from a patient, (ii) optionally expanding the immune cells (iii) introducing a heterologous polynucleotide of the invention or a vector of the invention into the immune cells to produce modified immune cells which express an T-cell receptor protein or a chimeric immune cell receptor protein of the invention; and (iii) reintroducing said modified immune cells into the patient. In yet a further aspect of the present invention, there is provided a method of treatment or prevention of cancer comprising administering to a patient in need thereof transduced immune cells wherein the immune cells are immune cells that have been obtained from said patient and a heterologous polynucleotide of the invention or a vector of the invention has been introduced into the immune cells, such that the immune cells express a T-cell receptor (TCR) protein or a chimeric immune cell receptor protein of the invention. Suitably the immune cells are T-cells. Suitably the chimeric immune cell receptor protein is a chimeric T-cell receptor protein. The invention also provides said transduced immune cells for use in the treatment or prevention of cancer. It also provides use of said transduced immune cells in the manufacture of a medicament for the treatment or prevention of cancer. According to this aspect of the invention, immune cells such as T-cells are first obtained from the patient or can be generated allogeneically. One suitable method is to collect PBMCs by apheresis. Alternatively, PBMCs may be isolated by density centrifugation. T-cells may thereafter be enriched from the PBMC fraction using positive or negative selection. An example of positive selection would be anti-CD3 or anti-CD4 or anti-CD8 microbeads. Negative selection could be achieved using antibody coated microbeads that specifically bind non-T-cell immune cells present in the PBMC fraction. Other methods of enrichment besides microbeads could be used, for example fluorescent activated sorting. T-cells may be expanded by ex vivo by culturing with media containing standard stimulatory cytokines such as IL-2, IL-7, IL-15, IL-21 and mixtures thereof. The invention also provides a method of treatment or prevention of cancer which comprises administering to a patient in need thereof a cyclic dinucleotide, preferably a cyclic dinucleotide as defined herein. The invention also provides a method of treatment or prevention of cancer which comprises administering to a patient in need thereof an agent which mimics the cyclic dinucleotide or stimulates the production of a cyclic dinucleotide in vivo. The invention also provides a cyclic dinucleotide as herein described for use in the treatment or prevention of cancer. It also provides use of a cyclic dinucleotide in the manufacture of a medicament for the treatment or prevention of cancer. The invention also provides an agent which mimics the cyclic dinucleotide or stimulates the production of a cyclic dinucleotide in vivo for use in the treatment or prevention of cancer. It also provides use of an agent which mimics the cyclic dinucleotide or stimulates the production of a cyclic dinucleotide in vivo in the manufacture of a medicament for the treatment or prevention of cancer. An agent which mimics the cyclic dinucleotide includes synthetic cyclic dinucleotides such as cyclic dinucleotides with thiophosphatediester bonds, phosphorothioate analogs of a cyclic dinucleotide, R,R-stereoisomers of a cyclic dinucleotide, the R,R- stereoisomers of the phosphorothioate analogs cyclic dinucleotides; examples include RR-CDG, the R,R-stereoisomer of the phosphorothioate analog c-di-GMP- S2 or RR-c-di-GMP-S2; RR-cyclic-di-guanine (CDG), R(P),R(P) dithio-c-di-GMP, ADU-S100, the R,R-stereoisomer of the phorphorothioate analog 2030 -cAAMP-S2 or RR-2030 -cAAMP or RR-CDA; c-di-GMP, and cGAMP-NPs, 2’3’-cyclic dinucleotide ADU-S100 with thiophosphatediester bonds and 2’AL3’TL-cGAMP, a cGAMP analog with one amide and one triazole linkage. The cyclic dinucleotide to be administered or produced in vivo can be, for example, a cyclic dinucleotide which comprises nucleotides comprising a base selected from A and G as herein defined. Preferably the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP-AMP (cGAMP), optionally the linkage between the nucleosides or nucleotide units is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’- 3’, optionally the cyclic dinucleotide is in extended or semi-extended conformation as described herein. In particular, the cyclic dinucleotide is a cyclic GMP-AMP (cGAMP). In an embodiment the cyclic dinucleotide is 2’,3’-cGAMP. The agent which stimulates the production of a cyclic dinucleotide in vivo can be, for example, an agonist of cyclic GMP-AMP synthase (cGAS). Example agonists of cGAS include G3-YSD, or cDNA, ds DNA, RNA or fragments thereof, DNA exonuclease I / III inhibitors, topoisomerase I inhibitors, topotecan, campothecin, doxorubicin. The patient is suitably a human patient. In the context of a method of treatment of cancer, a patient is a subject suffering from cancer. In the context of a method of prevention of cancer, a patient is a subject in whom cancer is to be prevented. Treatment or prevention of other diseases The invention also provides a method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof an agent which inhibits the production of a cyclic dinucleotide in vivo. The invention also provides an agent which inhibits the production of a cyclic dinucleotide in vivo for use in the treatment or prevention of an autoimmune disease. It also provides use of an agent which inhibits the production of a cyclic dinucleotide in vivo in the manufacture of a medicament for the treatment or prevention of an autoimmune disease. The agent which inhibits the production of a cyclic dinucleotide in vivo can be, for example, an inhibitor of cyclic GMP-AMP synthase (cGAS) activity. Example inhibitors of cGAS activity include RU.521, IRAK4-IN-4 and PF-06928215 as well as G140 and G150 (Lama et al (2019)). The patient is suitably a human patient. In the context of a method of treatment of an autoimmune disease, a patient is a subject suffering from an autoimmune disease In the context of a method of prevention of an autoimmune disease, a patient is a subject in whom autoimmune disease is to be prevented. Methods of raising T-cells The invention provides an in vitro method of obtaining T-cells capable of specifically binding to cells which present an MR1-ligand complex on their surface, preferably the complex or MR1-ligand complex is an MR1-ligand complex according to the invention, which comprises selecting T-cells obtained from a donor, particularly a human donor, e.g. by apheresis, which recognize an MR1-ligand complex e.g. by testing if said T-cells recognise (i) an MR1-ligand complex presented on a cell (such as a dendritic cell or other suitable antigen presenting cell) exogenously loaded with the ligand or the MR1-ligand complex or (ii) said complex in monomer or multimeric form, preferably the complex or MR1-ligand complex is an MR1-ligand complex according to the invention. Suitably the complex includes β2-microglobulin. Such recognition can be determined by conventional methods such as measuring T cell activation (for example via interferon gamma release assays) or T cell mediated killing (for example via measuring target cell death using flow cytometry assays with known markers of cell death such as 7-aminoactinomycin D) in response to the MR1-ligand complex. Said multimeric forms include tetramers, dextramers and streptavidin based multimers. T-cells could be directly isolated using labelled multimers, for example fluorescently labelled multimers that would stain the MR1- ligand-specific TCR on the surface of the T-cell clone, or T-cells capable of specifically binding to cells which present an MR1-ligand complex on their surface, and thus allow subsequent cell sorting based on fluorescence to allow for expansion of the T-cell. T-cells so selected can be amplified by well-known methods. The invention also provides a T-cell obtainable by or obtained by the aforesaid method. Methods of raising antibodies The invention provides an in vitro method of raising antibodies capable of specifically binding to cells which present an MR1-ligand complex on their surface which comprises immunising an experimental animal (such as rabbit, mouse, rat, guinea pig, hamster, llama etc.) with an MR1-ligand complex as immunogen, optionally together with an immunostimulant (e.g. specol), and obtaining antibodies or antibody producing cells from said animal (e.g. from PBMCs or from spleen). See further details in the “Antibody proteins” section above. The invention also provides an antibody obtainable by or obtained by the aforesaid method. Agents and uses The invention also provides a method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof an agent which is not a natural cyclic dinucleotide and which binds MR1 and prevents the formation of a MR1-ligand complex on the surface of cells and hence prevents the binding of T-cells to said cells. In one embodiment, the agent competes with a natural cyclic dinucleotide for binding to MR1 in the same binding site. The invention also provides said agent which is not a natural cyclic dinucleotide and which binds MR1 and prevents the formation of a MR1-ligand complex on the surface of cells and hence prevents the binding of T-cells to said cells for use in the treatment or prevention of an autoimmune disease. It also provides use of said agent which is not a natural cyclic dinucleotide and which binds MR1 and prevents the formation of a MR1-ligand complex on the surface of cells and hence prevents the binding of T-cells to said cells in the manufacture of a medicament for the treatment or prevention of an autoimmune disease. The invention also provides a method of treatment or prevention of cancer which comprises administering to a patient in need thereof an agent which is not a natural cyclic dinucleotide and which binds MR1 to form a MR1-agent complex on the surface of cells and which complex is capable of specifically binding a T-cell. In one embodiment, the agent competes with a natural cyclic dinucleotide for binding to MR1 in the same binding site, for example a derivative of 2’,3’-cGAMP, 3’,3’-cGAMP or 2’,5’- cGAMP, 3’,5’-cGAMP or phosphothioate or bisphosphothioate analogues thereof. The invention also provides said agent which is not a natural cyclic dinucleotide and which binds MR1 to form a MR1-agent complex on the surface of cells and which complex is capable of specifically binding a T-cell for use in the treatment or prevention of cancer. It also provides use of said agent which is not a natural cyclic dinucleotide and which binds MR1 to form a MR1-agent complex on the surface of cells and which complex is capable of specifically binding a T-cell in the manufacture of a medicament for the treatment or prevention of cancer. Cancer Exemplary cancers which may be treated according to the invention include solid tumours and blood cancers, for example cancers selected from lung, melanoma (e.g. skin melanoma), bone, breast, blood (e.g. leukemia), prostate, kidney, bladder, cervical, ovarian and colorectal cancers, and in particular selected from lung, melanoma (e.g. skin melanoma), bone and breast cancers. Cancer may be primary cancer or metastatic cancer. Autoimmune diseases Exemplary autoimmune diseases which may be treated according to the invention include rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, systemic lupus erythematosus, multiple sclerosis, psoriasis, Crohn’s disease, ulcerative colitis, uveitis, cryopyrin-associated periodic syndromes, Muckle-Wells syndrome, juvenile idiopathic arthritis, chronic obstructive pulmonary disease and Aicardi-Goutieres syndrome. Combination therapy In the treatment or prevention of diseases, different products described herein may be used in combination and one or more products described herein may be used in combination with other treatments (or preventions) for the given disease. For example, in the treatment or prevention of cancer one or more of the following substances can be combined: (a) an agent which stimulates the production of a cyclic dinucleotide in vivo; (b) a T-cell engaging protein of the invention and (c) and engineered T-cell of the invention, particularly (a) plus (b) or (a) plus (c). Any of the products for treatment or prevention of cancer described herein may be used in combination with another anti-cancer drug e.g. selected from alkylating agents (e.g. nitrogen mustard analogues, nitrosoureas, alkyl sulfonates, platinum containing compounds, ethylemines, and imidazotetrazines), cytotoxic antibiotics (e.g. anthracyclines, actinomycins), plant alkaloids and other natural products (e.g. campthotecin derivatives, epipodophyllotoxins, taxanes, and vinca alkaloids), antimetabolites (e.g. cytidine analogues, folic acid analogues, purine analogues, pyrimidine analogues, urea derivatives) and drugs for targeted therapy (e.g. kinase inhibitors, monoclonal antibodies and other immunotherapies) or radiotherapy. Immunotherapies include check point inhibitors such as anti-PD1, anti-PD-L1 and anti-CTLA-4 antibodies such as ipilumamab, nivolumab, pembrolizumab and atezolizumab. Diagnosis The invention also provides a method of determining whether a cell is cancerous which comprises determining whether said cell expresses on its surface an MR1- ligand complex comprising (a) MR1 and (b) a cyclic dinucleotide as ligand for MR1. The step of determining whether said cell expresses on its surface an MR1-ligand complex suitably includes the step of detecting the binding of an antibody protein to said MR1-ligand complex. The antibody protein may, in particular, be an antibody protein of the invention linked to a detectable label such as a fluorescent label. Production of antibody proteins, immune cell engaging proteins, chimeric immune cell receptor proteins, T-cell receptor proteins and other polypeptides of the invention Antibody proteins, immune cell engaging proteins, chimeric immune cell receptor proteins, T-cell receptor proteins and other polypeptides of the invention described herein can be obtained and manipulated using the techniques disclosed for example in Green and Sambrook 2012 Molecular Cloning: A Laboratory Manual 4th Edition Cold Spring Harbour Laboratory Press. In particular, artificial gene synthesis may be used to produce polynucleotides (Nambiar et al. (1984), Sakamar and Khorana, (1988), Wells et al. (1985) and Grundstrom et al. (1985)) followed by expression in a suitable organism to produce polypeptides. A gene encoding a polypeptide of the invention can be synthetically produced by, for example, solid-phase DNA synthesis. Entire genes may be synthesized de novo, without the need for precursor template DNA. To obtain the desired oligonucleotide, the building blocks are sequentially coupled to the growing oligonucleotide chain in the order required by the sequence of the product. Upon the completion of the chain assembly, the product is released from the solid phase to solution, deprotected, and collected. Products can be isolated by high-performance liquid chromatography (HPLC) to obtain the desired oligonucleotides in high purity (Verma and Eckstein (1998)). These relatively short segments are readily assembled by using a variety of gene amplification methods (Methods Mol Biol., 2012; 834:93-109) into longer DNA molecules, suitable for use in innumerable recombinant DNA-based expression systems. In the context of this invention one skilled in the art would understand that the polynucleotide sequences encoding the TCRs and fragments thereof described in this invention could be readily used in a variety of protein production systems, including, for example, viral vectors. For the purposes of production of polypeptides of the invention in a microbiological host (e.g., bacterial such as E coli or fungal such as yeast), polynucleotides of the invention will comprise suitable regulatory and control sequences (including promoters, termination signals etc) and sequences to promote polypeptide secretion suitable for protein production in the host. Similarly, polypeptides of the invention could be produced by transducing cultures of eukaryotic cells (e.g., Chinese hamster ovary cells or drosophila S2 cells) with polynucleotides of the invention which have been combined with suitable regulatory and control sequences (including promoters, termination signals etc) and sequences to promote polypeptide secretion suitable for protein production in these cells. Improved isolation of the polypeptides of the invention produced by recombinant means may optionally be facilitated through the addition of a purification tag at one end of the polypeptide. An example purification tag is a stretch of histidine residues (e.g.6-10 His residues), commonly known as a His-tag. Further aspects of the invention are defined by the following clauses: 1. An isolated MR1-ligand complex comprising (a) an MR1 protein and (b) a cyclic dinucleotide as ligand for the MR1 protein which is capable of being bound and/or specifically bound by a T-cell receptor, optionally wherein the T- cell receptor is an MR1 specific T-cell receptor. 2. The isolated complex according to clause 1 comprising the heavy chain of MR1, optionally wherein the MR1 has an amino acid sequence which is any one of 90%,91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NO:1 3. The isolated complex according to clause 2 wherein the heavy chain is non- covalently associated with a β2 microglobulin protein as light chain or is covalently linked to a β2 microglobulin protein in an artificial single-chain construct, optionally wherein the β2 microglobulin protein has an amino acid sequence which is any one of 90%,91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NO:2 4. The isolated complex according to any one of clauses 1 to 3 wherein the cyclic dinucleotide comprises nucleotides comprising a base selected from (Adenine) A and (Guanine) G. 5. The isolated complex according to clause 4 wherein the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP- GMP, cyclic GMP-AMP, preferably a cyclic GMP-AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’. 6. The isolated complex according to clause 5 wherein the cyclic dinucleotide is either 2’,3’-cGAMP or 3’,3’-cGAMP, preferably 2’,3’-cGAMP, optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation. 7. An antibody protein which specifically binds to the isolated complex according to any one of clauses 1 to 6. 8. The antibody protein according to clause 7 which is a 4-chain monoclonal antibody or fragment thereof, optionally any one of Fab (fragment antigen binding), scFv (single chain fragment variable), scFv-Fc, single domain antibody, diabody, dsFv, Fab', (Fab’)2, minibody, diabody, single-chain antibody molecule. 9. The antibody protein according to clause 7 which is an scFv. 10. An immune cell engaging protein which is capable of targeting a cell presenting on its surface an MR1-ligand complex comprising (a) a cell targeting portion comprising an antibody protein according to any one of clauses 8 to 10 and (b) an immune cell engaging portion. 11. The immune- cell engaging protein according to clause 10 wherein the immune cell is selected from any one of a T-cell, natural killer (NK) cell, Natural Killer T (NKT) cell, Tumour Infiltrating Lymphocyte (TIL), optionally wherein the immune cell may be a CD4+ T cell or CD8+ T cell, or may be a population of T-cells, optionally CD4+ T cells; or CD8+ T cells, or a mixed population of CD4+ T cells and CD8+ T cells, preferably the immune cell is a T-cell. 12. The immune cell engaging protein according to clause 10 wherein the immune cell engaging portion that is a T-cell engaging portion comprises an antibody protein which is capable of specifically binding to CD3 on T-cells. 13. The immune cell engaging protein according to clause 12 wherein the antibody protein which is capable of specifically binding to CD3 on T-cells is an scFv. 14. A chimeric immune cell receptor protein which comprises (a) a cell targeting portion comprising an antibody protein according to any one of clauses 7 to 9 and (b) a portion comprising immune cell effector function domains. 15. A chimeric immune cell receptor protein according to clause 14 wherein the immune cell is a T-cell. 16. The chimeric immune cell receptor protein according to clause 15 wherein the portion comprising T-cell effector function domains comprises a CD3zeta intracellular signalling domain. 17. The chimeric immune cell receptor protein according to clause 15 or clause 16 wherein the immune cell is a T-cell and the portion comprising T-cell effector function domains comprises one or more co-stimulatory domains. 18. An isolated T-cell receptor protein which is capable of specifically binding to the isolated complex according to any one of clauses 1 to 6 which T-cell receptor protein is not a T-cell receptor protein described in WO2019/081902. 19. An engineered immune cell which expresses on its surface a chimeric immune cell receptor protein according to any one of clauses 14 to 17 or a T-cell receptor protein according to clause 18. 20. An engineered immune cell according to clause 19 wherein the immune cell is selected from any one of a T-cell, natural killer (NK) cell, Natural Killer T (NKT) cell, Tumour Infiltrating Lymphocyte (TIL), or the immune cell may be optionally a CD4+ T cell or CD8+ T cell, or may be a population of T-cells, optionally CD4+ T cells; or CD8+ T cells, or a mixed population of CD4+ T cells and CD8+ T cells, preferably the immune cell is a T-cell. 21. An isolated polynucleotide which encodes a chimeric immune cell receptor protein according to any one of clauses 14 to 17 or a T-cell receptor protein according to clause 18. 22. A vector comprising a polynucleotide according to clause 21. 23. The vector according to clause 22 which is a viral vector, for example a lentiviral vector or a retroviral vector (e.g. ^-retrovirus) or a vector derived from adenovirus, adeno-associated virus (AAV), alphavirus, herpes virus, arenavirus, measles virus, poxvirus or rhabdovirus, preferably the vector is a lentiviral vector. 24. An engineered immune cell which comprises or is transduced with a vector according to clause 22 or clause 23 or which comprises or is transduced with a polynucleotide according to clause 21. 25. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an immune cell engaging protein according to any one of clauses 10 to 12. 26. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an engineered immune cell according to any one of clauses 19, 20 or 24. 27. A method according to clause 26 wherein the engineered immune cell is an engineered autologous immune cell, optionally wherein the immune cell is selected from any one of a T-cell, natural killer (NK) cell, Natural Killer T (NKT) cell, Tumour Infiltrating Lymphocyte (TIL), optionally wherein the immune cell may be a CD4+ T cell or CD8+ T cell, or may be a population of T-cells, optionally CD4+ T cells; or CD8+ T cells, or a mixed population of CD4+ T cells and CD8+ T cells, preferably the immune cell is a T-cell. 28. A method of obtaining T-cells capable of specifically binding to cells which present an MR1-ligand complex on their surface which comprises selecting T- cells obtained from a donor which recognize an MR1-ligand complex e.g. by testing if said T-cells recognise (i) an MR1-ligand complex presented on a cell exogenously loaded with the MR1-ligand complex or (ii) said complex in monomer or multimeric form, optionally wherein the MR1-ligand complex is the MR1-ligand complex according to any one of clauses 1-6. 29. A method of obtaining T-cells capable of specifically binding to cells which present an MR1-ligand complex which comprises priming and stimulating T- cells, particularly primary T-cells, particularly CD8+ T-cells, with antigen presenting cells (APCs) loaded with ligand and selecting T-cells that specifically bind to the APCs, optionally wherein the MR1-ligand complex is the MR1-ligand complex according to any one of clauses 1-6. 30. A method for identification of a T-cell reactive to an MR1-ligand complex in a T-cell containing preparation which comprises providing a T-cell containing preparation, contacting T-cells of said T-cell containing preparation with said MR1-ligand complex and isolating a T-cell that is reactive to said MR1-ligand complex, optionally wherein the T-cell is reactive as judged by proliferation response or cytokine production for example by production of interferon- gamma or interleukin (IL)-10 or granzyme B or TNFα, for example as assayed by tetramer staining, intracellular staining or ELISPOT (enzyme-linked immunospot assay), optionally wherein the MR1-ligand complex is the MR1- ligand complex according to any one of clauses 1-6. 31. A T-cell obtainable by or obtained by the method of any one of clauses 28 to 30. 32. A method of obtaining a T-cell receptor protein capable of specifically binding to cells which present an MR1-ligand complex which comprises obtaining T-cells according to the method of clause 28 or clause 29 or isolating T-cells according to clause 30 and obtaining the T-cell receptor protein from said T-cells. 33. An isolated T-cell receptor protein obtainable by or obtained by the method of clause 32, optionally wherein the T-cell receptor protein is a TCR (T cell receptor). 34. A method of raising antibodies capable of specifically binding to cells which present an MR1-ligand complex on their surface which comprises immunising an experimental animal with an MR1-ligand complex according to any one of clauses 1 to 6 as immunogen, optionally together with an immunostimulant, and obtaining antibodies or antibody producing cells from said animal. 35. An antibody obtainable by or obtained by the method of clause 34. 36. A method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof an agent which is not a natural cyclic dinucleotide and which binds MR1 and prevents the formation of a MR1-ligand complex on the surface of cells and hence prevents the binding of T-cells to said cells. 37. The method according to clause 36 wherein the agent competes with a natural cyclic dinucleotide for binding to MR1 in the same binding site, preferably wherein the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP- AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’, optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation. 38. A method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof a cyclic dinucleotide, preferably wherein the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP-AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’, optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation. 39. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an agent which is not a natural cyclic dinucleotide and which binds MR1 to form a MR1-agent complex on the surface of cells and which complex is capable of specifically binding a T-cell. 40. The method according to clause 39 wherein the agent competes with a natural cyclic dinucleotide for binding to MR1 in the same binding site, preferably wherein the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP- AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’, optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation. 41. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof a cyclic dinucleotide. 42. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an agent which stimulates the production of a cyclic dinucleotide in vivo, preferably wherein the agent is an agonist of an enzyme responsible for cyclic dinucleotide synthesis for example an agonist of cGAS or is an intermediate of cyclic dinucleotide synthesis. 43. A method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof an agent which inhibits the production of a cyclic dinucleotide in vivo, preferably wherein the agent is an inhibitor of an enzyme responsible for cyclic dinucleotide synthesis or of an enzyme responsible for synthesis of an intermediate of cyclic dinucleotide synthesis, for example an inhibitor of cGAS. 44. The method according to any one of clauses 41 to 43 wherein the cyclic dinucleotide comprises nucleotides comprising a base selected from A and G. 45. The method according to clause 39 wherein the cyclic dinucleotide preferably wherein the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP- AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’, optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation, preferably wherein the cyclic dinucleotide is a cyclic GMP-AMP (cGAMP). 46. The isolated complex according to clause 45 wherein the cyclic dinucleotide is 2’,3’-cGAMP or 3’,3’-cGAMP optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation, preferably wherein the cyclic dinucleotide is 2’,3’-cGAMP. 47. A method of determining whether a cell is cancerous which comprises determining whether said cell expresses on its surface an MR1-ligand complex comprising (a) MR1 and (b) a cyclic dinucleotide as ligand for MR1, optionally wherein the MR1-ligand complex is the MR1-ligand complex according to any one of clauses 1-6. 48. A method according to clause 47 wherein the step of determining whether said cell expresses on its surface an MR1-ligand complex includes the step of detecting the binding of an antibody protein to said MR1-ligand complex, optionally wherein the antibody is the antibody according to any one of clauses 7 to 9. 49. A method of clause 48 wherein the antibody protein is an antibody protein according to any one of clauses 7 to 9 optionally linked to a detectable label.

Examples Example 1 - Docking studies suggest cyclic dinucleotides could be ligands for MR1 To determine whether cyclic dinucleotides are cancer-specific MR1 ligands, a docking experiment was performed using the macromolecular structure of MR1 and utilising CCDC GOLD docking software with Hermes 2020.2 visualizer (Jones et al (1997)). The MR1 structure for docking was extracted from RCSB PDB Protein Data Bank (PDB ID 4pj9 – the structure of human MR1-5-OP-RU in complex with human MAIT TRAJ20 TCR, Sidonia et al (2014)). This structure is shown in Figure 3. For the cyclic dinucleotide molecular docking analysis, the following steps were taken: Firstly, the structures of the MAIT TCR and known MR1 ligand 5-OP-RU were removed from the overall macromolecular structure file, leaving only the MR1 protein structure. All missing hydrogen atoms were then added to the MR1 protein structure to ensure an explicit representation of the protein. The inventors observed in this MR1 structure, as well as other high-resolution structures of MR1, that four water molecules are consistently bound to the same regions of the MR1 protein. In addition, in the PDB ID 4pj9 MR1 structure, two more water molecules were found presented within 3Å of the site where the 5-OP-RU ligand is known to be bound. All of these water molecules were assumed to contribute to the stabilization of the MR1 binding pocket as well as to the increase in the stability of the protein-ligand complex. To validate this assumption, the effect of these water molecules was confirmed during the GOLD docking parameter setting test runs, where the known MR1 ligand 5-OP-RU was docked into the MR1 protein structure. The location and orientation of the docked 5-OP-RU ligand was compared to the experimentally derived location and orientation from the original crystal structure.4 water molecules (identifiers HOH441, HOH443, HOH445 and HOH458) were identified as crucial for the success of the docking of the known 5-OP-RU ligand and therefore were included in the docking run for the cyclic dinucleotides (using the toggle function in GOLD). The likely binding pocket for the cyclic dinucleotides to MR1 was estimated using the location and orientation of 5-OP-RU ligand in the original crystal structure file, with the pocket encompassing residues of the MR1 protein within 6Å of 5-OP- RU. The Configuration Template in CCDC GOLD was left in default settings. CHEMPLP Docking Scoring Function under default parameters was used to assess the goodness of the ligand fit in the MR1 pocket. Cyclic dinucleotides ligands tested in the docking runs were extracted from available structure files in RCSB PDB; 3’,3’-cGAMP in ‘closed’ conformation (PDB ID 4yaz, Ren et al (2015)), 3’,3’-cGAMP in ‘open” conformation (PDB ID 6ael, Kato et al (2018)) and 2’,3’-cGAMP in closed conformation (PDB ID 4ksy, Zhang et al (2013)). The quality of the docking of the various cyclic dinucleotides to the MR1 protein structure was assessed via the CHEMPLP Fitness score defined by the CCDC GOLD software, as well as visual inspection of the fit of the molecules into the known MR1 pocket. The CHEMPLP Fitness score for the known MR1 ligand 5-OP-RU was re-generated and used as a reference (3D reference conformer was downloaded from National Center for Biotechnology Information (NCBI) PubChem database). The following scores were observed for the different docking tests: Table 1: CHEMPLP Fitness Scores from docking studies.

The results of the docking tests suggested that the extended conformation of 3’-3’- cGAMP could be an excellent ligand for MR1, with the CHEMPLP Fitness Score for this putative ligand being higher than for the known MR1 ligand 5-OP-RU. Conversely the docking tests for the closed conformations of both 2’,3’-cGAMP and 3’,3’-cGAMP suggest these would be poor ligands for MR1. The docking of the extended conformation of 3’,3’-cGAMP to the MR1 structure is shown in Figure 2. The strong docking results for the extended conformation of 3’,3’-cGAMP led the inventors to consider whether a similarly extended conformation of 2’,3’-cGAMP could also be an excellent candidate for an MR1 ligand. 2’,3’-cGAMP in an extended conformation was not available in RCSB PDB database and so a potential structure for the ligand was built using PyMOL Version 2.4.1.builder and its Sculpt function to scramble unconstrained coordinates, with the structure quality assessed by GOLD tool MOGUL. As evidenced in Figure 1, this structure of 2’,3’-cGAMP, whilst not fully extended, fills a similar space in the MR1 binding pocket as the experimentally observed extended conformation of 3’,3’-cGAMP, in terms of ligand location and orientation with the GMP moiety of the ligand deeper inside MR1 pocket. Figure 7 provides more evidence of these similarities with close superimposition of both ligands docked into the MR1 structure. Docking of the GMP subunits of both ligands is to a similar space in the binding pocket. This location also corresponds to the location of the uracil aromatic ring of 5-OP-RU ligand (see Figure 4C). It is conceivable that a fully extended conformation of 2’,3’-cGAMP exists in mammals but is yet to be experimentally observed and this may result in further optimisation of MR1-ligand binding. Further analysis of these putative ligands in the MR1 binding site suggest that the same amino acid side chains of MR1 would form similar non- covalent interactions with both 2’,3’-cGAMP and 3’,3’-cGAMP (Figure 4A and 4B). Amino acid side chains that are known to interact with 5-OP-RU (Figure 4C) are also predicted to be involved in 2’,3’-cGAMP and 3’,3’-cGAMP binding. As noted above, there is a class of T-cells, referred to as K43A sensitive MR1-T- cells, which have activity against a range of cancer cell lines. The susceptibility of certain cancer cell lines to these T cells is significantly, but not entirely, diminished if the lysine-43 residue of the MR1 protein is mutated to alanine. The inventors explored the structures of MR1 docked with the extended conformations of 2’,3’- cGAMP and 3’,3’-cGAMP with a view to rationalising a link between cyclic dinucleotides as cancer-specific ligands of MR1 and the observed importance of lysine-43 of MR1 in modulating K43A sensitive MR1-T-cell activity. The inventors found that lysine-43 is well positioned, within approximately 4Å, to make a significant non-covalent interaction with the guanosine ring of both 2’,3’-cGAMP and 3’,3’- cGAMP deep in the binding pocket (Figure 6A and 6B). It is of note that this interaction differs from the covalent Schiff base bond that is known to be formed between lysine-43 and 5-OP-RU. These observations are consistent with the observed importance of lysine-43 for the anti-cancer activity of K43A sensitive MR1- T-cells. Overall, these macromolecular docking studies provide compelling evidence that the extended conformations of cyclic dinucleotides such as 2,3-cGAMP and 3,3-cGAMP could be ligands for MR1. Example 2 - Demonstrating that cyclic dinucleotides are ligands of MR1 Known MR1 ligands (e.g. acetyl-6-FP) have been shown to stabilise intracellular pools of ligand-free (‘empty’) MR1 when co-incubated with MR1-expressing cells at varying concentrations over time. This stabilisation then leads to trafficking of the MR1-ligand complex to the cell surface which can be detected via antibody detection methods such as flow cytometry, using antibodies to MR1. Ac-6-FP (and other synthetic molecules) were important in discovering bacterially produced ligands that mediate the anti-bacterial activity of MR1-restricted mucosal invariant T cells (MAITs). Importantly, although Ac-6-FP can upregulate cell-surface expression of MR1, the Ac-6-FP-liganded MR1 is not a target for MAITs (Kjer-Nielsen et al., Nature, 2012). Rather, biologically relevant ligands for MAIT cells appear to be the product of condensation reactions of riboflavin precursors with small molecules (Corbett et al. (2014); Awad et al. (2020)). To demonstrate a role of cyclic dinucleotides in liganding MR1, MR1-expressing cells can be pulsed with titrating amounts of cyclic dinucleotides (e.g.2’,3’-cGAMP) over a time course (1 – 48 hr) and surface expression of MR1 can be assessed by flow cytometry. Cancer cell lines known to express MR1 and beta-2-microglobulin with detectable MR1 at the surface would be tested first. Known MR1 ligands (e.g. Ac-6- FP) would be run in parallel as a control. Cells that are thought to have little-to-no surface expression such as normal primary cells would then also be tested. An increase in MR1 at the surface observed at any point or concentration of cyclic dinucleotides would indicate the cyclic dinucleotide was a potential ligand for MR1. Accordingly, C1R.MR1 cells were plated in a 96 well plate (50,000 cells/well) and incubated with 2,3-cGAMP ± 10µM MEG at five doses / concentrations in triplicate for up to 24 hours. Cells were washed then stained with 5 µg/ml anti-human MR1 for 30 min at 4°C. After 24 hours co-incubation, 2,3-cGAMP +/- MEG treatment resulted in increased MR1 expression in C1R.MR1 cells at all tested doses (0.1-1000 µM), see Figure 10, which shows fold change of the geometric mean compared to the respective vehicle (0.2% H₂O). The data indicate that 2,3-cGAMP is an MR1 ligand. To further validate the association of the mammalian cyclic dinucleotide pathway with MR1-ligand cell surface presentation, key components of the cyclic dinucleotide pathway can be modulated. cGAS is responsible for generating 2’,3’-cGAMP from GTP and ATP in the presence of cytosolic dsDNA (Sun et al.2013), therefore agonism of cGAS could lead to an increase in MR1- cyclic dinucleotide. Cancer cell lines would be treated with a cGAS agonist such as G3-YSD and MR1 surface levels assessed using methods such as flow cytometry. An increase in surface MR1 in treated cells would be consistent with increase in 2’,3’-cGAMP production via cGAS where 2’,3’-cGAMP is a ligand of MR1. Genetic modification of cell lines would also be used to demonstrate cyclic dinucleotides as MR1-binding ligands. Overexpression of cGAS would be performed (preferably with an inducible, e.g., tet-responsive promoter) and influence of cGAS expression on increased surface MR1 would be assessed, with or without exogenous supply of cyclic dinucleotides (pulsing) or cGAS agonists. Further, genetic manipulation of MR1 to introduce wild type MR1 to be overexpressed in MR1-negative cells and assessed for surface presentation following cyclic dinucleotide pulsing or cGAS agonism would be performed. In addition, mutant forms of MR1, such as the well-described K43A mutation, could also be introduced into MR1-negative cell lines to demonstrate the importance of certain amino acids known to be involved in ligand binding in the MR1 binding groove are important in liganding of cyclic dinucleotides. The MR1-β2-microglobulin complex can be generated in a soluble form and refolded with MR1 ligands.2’,3’-cGAMP would be provided in excess with refolding of MR1- β2-microglobulin soluble protein, and the interaction could be stabilised through UV irradiation, which should cross-link the nucleotide to the protein. This refolded cyclic dinucleotide-MR1- β2-microglobulin soluble complex would be assessed by mass spectrometry and/or grown as crystals and subject to X-ray crystallography to confirm that 2’,3’-cGAMP is a bona fide ligand of MR1. Example 3 - Demonstrating the function of MR1-bound cyclic dinucleotides as targets for cancer-reactive T cells To demonstrate cyclic dinucleotides can bind MR1 and are ligands for MR1-T-cells reactive to cancer cells, functional assays with known cancer-reactive MR1-T-cells would be used with optimal concentrations and time-points as determined in the previous set of experiments. Cancer-reactive MR1-T-cell clones (or primary T cells transduced to express the TCR from such clones) would be co-cultured with cancer cell lines that had been pulsed with 2’,3’-cGAMP (or other cyclic dinucleotide) as well as untreated target cells. Following co-culture for 24 – 48 hr, killing of the target cell lines would be assessed (Flow cytometry and/or xCELLigence methods), and activation of the MR1-T-cells would be assessed by evaluation of T cell activation markers by using assays such as production of IFNγ (e.g., by ELISpot) and/or TNFα (TAPI-0 flow cytometry-based assay). Increased target cell death and/or MR1-T-cell activation in the presence of targets incubated with cyclic dinucleotides over untreated targets would demonstrate MR1-cyclic dinucleotide is an antigen for MR1- T-cells. Assessment of TCR-T cell activation was carried out by measuring IFNγ production following co-incubation of TCR- T cells (T cells transduced with the TCR from clone MC.7.G5) with C1R.MR1 cells (human B-cell lymphoblastoid line transduced with MR1 for ligand presentation) treated with 2,3-cGAMP (+/- Methylglyoxal, MeG). C1R.MR1 cells (250,000 cells/well) were incubated with 2,3-cGAMP at the highest viable concentration (in triplicate) for 24 hours, either with or without the presence of 5mM Methylglyoxal (MeG), control cells were left untreated (no 2,3-cGAMP) for 24 hours. Viable target cells were counted and plated in presence of MR1 specific TCR-T cells in 1:1 E:T ratio (25,000 C1R.MR1 cells/well each). After 16h incubation, supernatants were collected and assessed for IFNγ content by TR-FRET assays. Figures 8 and 9 show the resulting raw and comparative IFNγ production in C1R.MR1 cells post the T cell activation assay. Fold change of IFNγ expression was assessed using untreated C1R.MR1 incubated with TCR-T cells as a baseline (grey line at 1.0 fold change IFNγ expression in Figure 9. All samples were tested neat as shown, and no IFNγ was detected in C1R.MR1 or T cells alone. Statistical significance between the untreated and treatment groups was determined using one- way ANOVA with Dunnett’s post-hoc test for multiple comparisons. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 and **** p ≤ 0.0001. MeG is an alpha-dicarbonyl compound (common intermediates in mammalian glycolysis) capable of condensation with free amines of putative ligands of MR1, the resulting adduct may be captured and stabilized through a covalent bond with Lys43 of the MR1 binding pocket to form a schiffs base. However, potential ligands may also form non-covalent interactions within the binding pocket. The data (Figures 8 and 9) showed a negative fold change of IFNγ expression with respect to untreated C1R.MR1 incubated with TCR-T cells. The measured effect was determined to be an inhibitory effect of 2,3-cGAMP, seen both with and without MeG, on the IFNγ expression by the MR1 specific TCR-T cells, indicative of competition of 2,3-cGAMP with the natural ligand of the test MR1 specific TCR. Similar to the approach described above (Example 2) for modulating cGAS activity, the expectation of these studies is that inhibition of cGAS should lead to a loss in MR1- cyclic dinucleotide, and therefore reduce interactions with anti-cancer MR1-T cells. Target cancer cell lines would be treated with a cGAS inhibitor, such as RU.521, and a dose-dependent inhibition of cytotoxicity and/or T cell activation would be observed when treated cells are co-incubated with anti-cancer MR1-T-cells in functional assays. Likewise, treating cancer cell lines with a G3-YSD (or other cGAS agonist) would increase the activity of anti-cancer MR1-T-cells when incubated with the treated cells, as measured by target cell cytotoxicity and/or T cell activation markers. Genetic manipulation of target cell lines would also be used to demonstrate the hypothesis (see Example 2). Briefly, modulation, genetic knock-down, or knock-out of cGAS in cell lines would be performed which would be expected to lead to a reduction in target cell death and/or T cell activation of anti-cancer MR1-T-cells due to a loss of cyclic dinucleotide as a ligand for MR1. Conversely, overexpression of cGAS (assuming dsDNA substrate is not limiting) in cancer cell lines would lead to an increase in target cell death and/or T cell activation with anti-cancer MR1-T-cells. Soluble cyclic dinucleotides-MR1- β2-microglobulin complexes generated would be made into multimeric structures (e.g. tetramer, dextramer, etc). These multimers would be used to stain anti-cancer MR1-T-cells to demonstrate TCR-recognition of cyclic dinucleotides in complex with MR1. In addition, these multimers (or monomers) would be titrated into co-culture assays of MR1+ cancer cell lines and anti-cancer MR1-T-cells to demonstrate a dose-dependent blocking of functional activity (IFNγ and/or TNFα release) or cytotoxicity. Example 4 – Preparation and characterization of a MR1-cDN ligand complex MR1 is unable to assemble with β2 microglobulin in the absence of a ligand, however, the subunits obtained from inclusion bodies of E.coli expressing the individual MR1 heavy chain and β2 microglobulin can be mixed together and refolded with ligands as described by Corbett et al (2014). Briefly, the inclusion bodies containing these 2 proteins are mixed with ligand in a refolding solution (for example: 0.1 M Tris, pH 8.5, 2 mM EDTA, 0.4 M arginine, 0.5 mM oxidized glutathione and 5 mM reduced glutathione) in the presence of various cDNs. The refolded MR1-cDN antigen complexes can then be purified by sequential application of DEAE anion exchange, gel filtration, and MonoQ anion exchange chromatography (Corbett et al. (2014)). Refolding can also be accomplished in the presence of the riboflavin intermediate 5- A-RU and methylglyoxal or glyoxal to produce the MR1 ligated to the MAIT-ligand, 5- OE-RU, described by Corbett et al. (2014) to confirm the conditions used for folding and subsequent purification. Confirmation that the cDN-ligated MR1 molecules contain the co-incubated cDN can be accomplished by using mass spectroscopy under conditions that resolve the cDN molecules incubated with the exclusion products, or the 5-OE-RU (Corbett et al. (2014)) in the case of 5-OE-RU-liganded MR1 complex. Side-by-side evaluation of purified cDNs can be used to confirm the identity of the spectra assigned to the ligated cDN molecules/5-OE-RU molecule. Purified cDN-ligated MR1 molecules can be tested for their ability to activate cancer- reactive MR1-T-cell clones (or primary T cells transduced to express the TCR from such clones), by methods that include TNFα (TAPI-0 flow cytometry-based assay). Alternatively, intracellular cytokine staining can be used to demonstrate that the cancer-reactive MR1 T cells are activated by treatment with the MR1-cDN complexes. Lack of activation of MAIT cells (or primary cells transduced to express the TCR of MAIT cells) can also be evaluated. Both types of T cells (cancer-reactive and MAIT) can be tested on 5-OE-RU-liganded MR1, cDN-liganded MR1, and the non-activating MR1-Ac-6-FP complexes, to confirm the specificity of MR1 ligand activation. Example 5 - Identification of T-cells that recognise MR1-cDN ligand complexes T cells that recognize the MR1-cDN complex can be identified by multiple methods using blood-derived T cells from normal subjects and cancer patients. For example, naïve of patient T cells can be incubated with MR1-cDN complexes and then specifically activated T cells can be identified by using flow-cytometry based methods such as detection of activation-induced TNFα (TAPI-0-based assay). Alternatively, intracellular cytokine staining can be used to identify if T cells incubated with the MR1-cDN complex are activated. Appropriate controls for these studies can include the non-activating MR1-Ac-6FP complex which is incapable of activating T cells specific for MR1 ligands (Kjers-Nielsen et al., Nature, 2012). Studies conducted side-by-side with 5-OE-RU-liganded MR1 can also be conducted. In these studies, co-staining of activated cells with antibodies specific for TRAV1-2 could be used to demonstrate that MAIT cells are only activated with the 5-OE-RU- liganded MR1, and that the cells reactive to the MR1-cDN complex are not MAIT- like. Example 6 - Preparation of antibodies against MR1-cDN ligand complex. Antibodies that specifically recognize the MR1-cDN ligand complex could be generated by immunisation of an experimental animal with an MR1-cDN ligand complex. Alternatively, antibodies can be generated through the use of phage- or yeast-display technology (Sheehan & Marasco, 2015), which are based on large libraries of antibody-like reactive molecules that can be screened with the MR1-cDN complex to discover antibody complementary determining paratopes that react with MR1. Briefly, individual clones of phage or yeast found in large libraries displaying diverse scFv or Fab fusions can be selected for their ability to bind the MR1-cDN complexes, and the selected antibody fragments can be reconstructed into functional antibodies for further use. Counter screening of the positive phage/yeast clones with MR1 (K43A) mutant refolded without ligand and MR1 folded with various MAIT ligands can be used to eliminate MR1-cDN complex-reactive phage/yeast clones that are not specific for the MR1-cDN complex. References Awad et al (2020) Nature Immunology 21:400-411 Crowther et al (2020) Nature Immunology, 21, 178-185 Grundstrom et al. (1985) Nucl. Acids Res., 13, 3305-3316) Jones et al (1997) J. Mol. Biol., 267, 727-748 Lion et al (2013) Eur. Journ Immunology 43(5) 1363-1373 Lepore et al (2017) eLife 6, e24476 Nambiar et al. (1984) Science, 223, 1299-1301 Ren, et al. (2015) Cell Reports, 11 (1) 1 – 12 Sakamar and Khorana (1988) Nucl. Acids Res., 14, 6361-6372 Vacchini et al (2020) Frontiers in Immunology, 11, 1-8 Verma and Eckstein (1998) Annu. Rev. Biochem., 67, 99-134 Wells et al. (1985) Gene, 34, 315-323 Xie et al (2020) EBioMedicine, 59, 102975 Corbett et al. (2014) Nature 509, 361-365 Kato, K. et al (2018) Nat Commun 9, 4424 Kjers-Nielsen et al. (2012) Nature 491, 717-723 Lama et al (2019) Nature Communications 10, 2261 Sun et al. (2013) Science 339, 786-791 Sidonia B.G. et al (2014) J Exp Med 211 (8): 1585–1600 Sheehan & Marasco (2015) Microbiol Spectrum 3(1):AID-0028-2014. doi:10.1128 /microbiolspec.AID-0028-2014. Zhang X et al (2013) Molecular Cell 51(2):226-235 Throughout the specification and the claims which follow, unless the context requires otherwise, the word ‘comprise’, and variations such as ‘comprises’ and ‘comprising’, will be understood to imply the inclusion of a stated integer, step, group of integers or group of steps but not to the exclusion of any other integer, step, group of integers or group of steps. All patents, patent applications and references mentioned throughout the specification of the present invention are herein incorporated in their entirety by reference. The invention embraces all combinations of preferred and more preferred groups and suitable and more suitable groups and embodiments of groups recited above.

Claims

Claims 1. An isolated MR1-ligand complex comprising (a) an MR1 protein and (b) a cyclic dinucleotide as ligand for the MR1 protein which is capable of being specifically bound by a T-cell receptor.

2. The isolated complex according to claim 1 comprising the heavy chain of MR1.

3. The isolated complex according to claim 2 wherein the heavy chain is non- covalently associated with a β2 microglobulin protein as light chain or is covalently linked to a β2 microglobulin protein in an artificial single-chain construct.

4. The isolated complex according to any one of claims 1 to 3 wherein the cyclic dinucleotide is a cyclic GMP-AMP (cGAMP).

5. The isolated complex according to claim 4 wherein the cyclic dinucleotide is 2’,3’-cGAMP.

6. An antibody protein which specifically binds to the isolated complex according to any one of claims 1 to 5.

7. An immune cell engaging protein which is capable of targeting a cell presenting on its surface an MR1-ligand complex comprising (a) a cell targeting portion comprising an antibody protein according to claim 6 and (b) an immune cell engaging portion.

8. A chimeric immune cell receptor protein which comprises (a) a cell targeting portion comprising an antibody protein according to claim 6 and (b) a portion comprising immune cell effector function domains.

9. An isolated T-cell receptor protein which is capable of specifically binding to the isolated complex according to any one of claims 1 to 5 which T-cell receptor protein is not a T-cell receptor protein described in WO2019/081902.

10. An engineered immune cell which expresses on its surface a chimeric immune cell receptor protein according to any claim 8 or a T-cell receptor protein according to claim 9.

11. An isolated polynucleotide which encodes a chimeric immune cell receptor protein according to claim 8 or a T-cell receptor protein according to claim 9.

12. A vector comprising a polynucleotide according to claim 11.

13. An engineered immune cell which comprises or is transduced with a vector according to claim 12 or which comprises or is transduced with a polynucleotide of claim 11.

14. An immune cell engaging protein according to claim 7 or an engineered immune cell according to claim 10 for use in the prevention or treatment of cancer.

15. A method of obtaining T-cells capable of specifically binding to cells which present an MR1-ligand complex on their surface which comprises selecting T- cells obtained from a donor which recognize an MR1-ligand complex e.g. by testing if said T-cells recognise (i) an MR1-ligand complex presented on a cell exogenously loaded with the MR1-ligand complex or (ii) said complex in monomer or multimeric form.

16. A method of obtaining T-cells capable of specifically binding to cells which present an MR1-ligand complex which comprises priming and stimulating T- cells, particularly primary T-cells, particularly CD8+ T-cells, with antigen presenting cells (APCs) loaded with ligand and selecting T-cells that specifically bind to the APCs, wherein the MR1-ligand complex is the MR1- ligand complex according to any one of claims 1-4.

17. A method for identification of a T-cell reactive to an MR1-ligand complex in a T-cell containing preparation which comprises providing a T-cell containing preparation, contacting T-cells of said T-cell containing preparation with said MR1-ligand complex and isolating a T-cell that is reactive to said MR1-ligand complex, wherein the MR1-ligand complex is the MR1-ligand complex according to any one of clauses 1-4.

18. A T-cell obtainable by or obtained by the method of any one of claims15 to 17.

19. A method of obtaining a T-cell receptor protein capable of specifically binding to cells which present an MR1-ligand complex which comprises obtaining T-cells according to the method of claim 15 or claim 16 or isolating T-cells according to claim 17 and obtaining the T-cell receptor protein from said T-cells.

20. An isolated TCR obtainable by or obtained by the method of claim 19.

21. A method of raising antibodies capable of specifically binding to cells which present an MR1-ligand complex on their surface which comprises immunising an experimental animal with an MR1-ligand complex according to any one of claims 1 to 5 as immunogen, optionally together with an immunostimulant, and obtaining antibodies or antibody producing cells from said animal.

22. An antibody obtainable by or obtained by the method of claim 21.

23. A method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof an agent which is not a natural cyclic dinucleotide and which binds MR1 and prevents the formation of a MR1-ligand complex on the surface of cells and hence prevents the binding of T-cells to said cells.

24. A method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof a cyclic dinucleotide, preferably wherein the cyclic dinucleotide is selected from any one of cyclic GMP-GMP, cyclic AMP-AMP, cyclic AMP-GMP, cyclic GMP-AMP, preferably a cyclic GMP-AMP (cGAMP), further optionally wherein the linkage between the nucleosides is selected from 2’-3’, 3’-3’, 2’-5’, 3’-5’, preferably 2’-3’, optionally wherein the cyclic dinucleotide is in extended or semi-extended conformation.

25. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an agent which is not a natural cyclic dinucleotide and which binds MR1 to form a MR1-agent complex on the surface of cells and which complex is capable of specifically binding a T-cell.

26. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof a cyclic dinucleotide.

27. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an agent which stimulates the production of a cyclic dinucleotide in vivo, preferably wherein the agent is an agonist of an enzyme responsible for cyclic dinucleotide synthesis for example an agonist of cGAS or is an intermediate of cyclic dinucleotide synthesis.

28. A method of treatment or prevention of an autoimmune disease which comprises administering to a patient in need thereof an agent which inhibits the production of a cyclic dinucleotide in vivo, preferably wherein the agent is an inhibitor of an enzyme responsible for cyclic dinucleotide synthesis or of an enzyme responsible for synthesis of an intermediate of cyclic dinucleotide synthesis, for example an inhibitor of cGAS.

29. A method of determining whether a cell is cancerous which comprises determining whether said cell expresses on its surface an MR1-ligand complex comprising (a) MR1 and (b) a cyclic dinucleotide as ligand for MR1, optionally wherein the MR1-ligand complex is the MR1-ligand complex according to any one of claims 1 to 5.

30. A method of treatment or prevention of cancer which comprises administering to a patient in need thereof an immune cell engaging protein according to claim 7 or an engineered immune cell according to claim 10.