Peptide-MHC complex

Abstract

The present invention provides stabilized peptide-MHC (pMHC) complexes, such as peptide-HLA-E complexes, which have a non-native linkage, such as a disulfide bond, between the C-terminal anchor residue of the peptide and an amino acid residue within the F pocket of the MHC binding groove.

Description

Technical Field

[0001] The present invention relates to peptide-MHC (pMHC) complexes. In particular, the present invention relates to pMHC complexes that are stabilized and retain native-like TCR recognition.

Background Art

[0002] MHC molecules play an important role in immune surveillance by presenting endogenous and exogenous antigenic peptides to T cells. The MHC class I complex consists of a heavy chain composed of two subunits, three extracellular domains (α1, α2, and α3), a transmembrane domain, and a cytoplasmic tail, and a light chain called β2-microglobulin (β2M) (required for in vivo expression of all class I molecules). The α1 domain and α2 domain of the heavy chain are structurally arranged together to form a platform of eight anti-parallel β-strands sandwiched between two α-helices, which forms the peptide-binding groove. MHC class I molecules typically bind peptides of about 8 to 10 amino acids. The MHC class II complex has an overall structure similar to that of the class I complex, but the peptide-binding groove is formed from both subunits and has a more open configuration that accommodates longer peptides, typically peptides of 11 to 30 amino acids. The binding groove of the MHC complex can be considered to be divided into six pockets or subsites called pockets A to F. The pockets (A and F) at each end of the binding groove are highly conserved and are responsible for binding the N- and C-terminal anchor residues of the peptide via an extensive hydrogen bond network. The other pockets are polymorphic and thus play a role in determining the specificity of peptide interaction (Matsumura et al., Science. 1992 Aug 14;257(5072):927-34).

[0003] Isolated pMHC complexes are essential tools in immunological research and the development of various therapeutic modalities; however, in some cases, the peptide binds weakly to the MHC and thus dissociates rapidly. This can pose a problem for methodologies that rely on stable complexes (e.g., counting T cell responses using MHC multimers). The instability of MHC complexes seems to be particularly problematic for peptides that bind to the non-classical human MHC class I molecule HLA-E.

[0004] Unlike classical MHC molecules, HLA-E exists almost exclusively in only two allelic forms, E*01:01 and E*01:03, which differ by only 1 amino acid. For this reason, the peptides presented by HLA-E are attractive targets for immunotherapy because they can avoid the problems inherent in targeting highly polymorphic classical MHC molecules. Under normal conditions, HLA-E binds to leader sequence peptides cleaved from other HLA class I molecules and presents them to NK cells as a means of immune surveillance. Defects in the antigen processing machinery in tumor tissues or those caused by certain infectious agents result in targeted killing by NK cells and are associated with an increase in the HLA-E peptide repertoire (presumably enabling T cell immune surveillance). HLA-E can present peptides from bacteria and viruses (e.g., those derived from Mycobacterium tuberculosis or HIV), and these peptide-HLA-E complexes are CD8 +There is increasing evidence that T cells can be stimulated (van Hall et al., Microbes Infect. 2010 Nov;12(12-13):910-8; Joosten et al., PLoS Pathog. 2010 Feb 26;6(2):e1000782; Joosten et al., J Immunol Res. 2016;2016:2695396; Hansen et al., Science. 2016 Feb 12;351(6274):714-20; Nattermann et al., Antivir Ther. 2005;10(1):95-107; Nattermann et al., Am J Pathol. 2005 Feb;166(2):443-53)). The role of HLA-E appears to be conserved across a range of mammals, including primates (Wu et al., J Immunol. 2018 Jan 1;200(1):49-60) and mice (Oliveira et al., J Exp Med. 2010 Jan 18;207(1):207-21). Efforts to utilize immunotherapy approaches targeting peptide-HLA-E complexes have been hampered by the instability of isolated peptide-HLA-E complexes compared to classical class I complexes such as HLA-A * 02. For example, the instability of isolated peptide-HLA-E complexes can prevent the identification and subsequent development of T cell receptors (TCRs) or antibody-based therapeutic agents that specifically recognize the complex.

[0005] Methods for stabilizing isolated pMHC complexes are known in the art (see, for example, US8992937; WO2013030620; Truscott J Immunol.2007 May 15;178(10):6280-9; Mitaksov et al., Chem Biol. 2007 Aug;14(8):909-22). SUMMARY OF THE INVENTION PROBLEMS TO BE SOLVED BY THE INVENTION

[0006] However, the inventors have found that such an approach is not suitable for stabilizing the pMHC complex (including the complex containing HLA-E) in a manner that retains native-like TCR binding. The present invention aims to provide a stabilized peptide-MHC (pMHC) complex that demonstrates native-like TCR recognition. **Means for Solving the Problems**

[0007] In a first aspect, the present invention provides a stabilized peptide-MHC (pMHC) complex comprising an unnatural linkage between the C-terminal anchor residue of the peptide and an amino acid residue within the F pocket of the MHC binding groove. The inventors have unexpectedly found that the pMHC complex can be stabilized by introducing an unnatural linkage between the C-terminal anchor amino acid residue of the peptide and an amino acid residue within the F pocket of the MHC binding groove. This linkage enables the pMHC complex to retain the three-dimensional structure of the native pMHC complex and be recognized by TCRs that recognize the native complex.

[0008] The pMHC complex of the present invention is stabilized in the sense that it has superior stability compared to the native complex that does not have an unnatural linkage between the C-terminal anchor residue of the peptide and the amino acid residue of the F pocket of the MHC binding groove. The stability can be evaluated by surface plasmon resonance (SPR) or biolayer interferometry (BLI) well-known to those skilled in the art, such as Biacore or Octet, respectively. The pMHC complex according to the present invention has a longer binding half-life than that of the native pMHC complex. The binding half-life of the pMHC complex of the present invention can be at least 2-fold, at least 3-fold, at least 4-fold, or at least 5-fold greater than that of the native pMHC complex. In the pMHC complex of the present invention, the peptide can have a binding half-life to MHC of at least 3 hours. Preferably, the binding half-life is at least 4 hours, at least 5 hours, at least 10 hours, at least 15 hours, or at least 20 hours. An alternative approach to determining the stability of the complex involves thermal denaturation.

[0009] The pMHC complexes of the present invention retain the native three-dimensional structure of native pMHC complexes. Thus, they can be recognized by peptide MHC-binding moieties (e.g., TCR or TCR-mimicking antibodies) that recognize native complexes. Recognition can be determined by SPR. The affinity of the binding moiety for the pMHC complexes of the present invention can be within a three-fold difference or less from the affinity of the binding moiety for the native complexes when measured under equivalent conditions. Those skilled in the art will understand that it may not be possible to compare the affinity of the binding moiety for the complexes of the present invention by measuring the affinity of the binding moiety for native complexes, since certain native pMHC complexes are very unstable under standard conditions (e.g., room temperature) (e.g., have a short binding half-life). In such cases, it may be possible to measure (and compare) the affinity by changing the conditions (e.g., lowering the temperature). The binding moiety has a K for the complex of at least 100 μM, at least 50 μM, at least 10 μM, at least 1 μM or less under standard conditions (e.g., the conditions shown in Example 2B). D It may have.

[0010] The pMHC complex of the present invention comprises an unnatural linkage between the C-terminal anchor residue of the MHC-binding peptide and an amino acid residue within the F pocket of the peptide-binding groove. This linkage stabilizes the peptide within the binding groove. The unnatural linkage does not disrupt the conformation of the peptide in the binding groove. This means that the peptide should be recognized in a native-like manner by a peptide-MHC-binding moiety (e.g., TCR). The conformation of the peptide can be determined by X-ray crystallographic analysis. The linkage can be a covalent bond. The covalent bond can be formed between amino acids substituted from the amino acid residues of the F pocket of the pMHC binding groove and / or the C-terminal anchor residue of the peptide, preferably from the amino acid residues of both the F pocket of the MHC binding groove and the C-terminal anchor residue of the peptide. At least one of the amino acids substituted from the amino acid residues of the F pocket of the MHC binding groove and / or the C-terminal anchor residue can be an unnatural amino acid. Preferably, the C-terminal anchor residue is an unnatural amino acid. Those skilled in the art are aware of which amino acid positions of the MHC molecule are located within the F pocket (see, for example, Table I of Matsumura et al., Science. 1992 Aug 14;257(5072):927-34). Preferably, the linkage is between the C-terminal anchor residue of the MHC-binding peptide and the amino acid residue at position 116 of the MHC heavy chain. Alternatively, a linkage between the C-terminal anchor residue of the MHC-binding peptide and the amino acid residue located at position 147 of the MHC heavy chain is also preferred. Other positions on the MHC heavy chain suitable for the linkage include positions 81, 124, and 143.

[0011] Preferably, the unnatural linkage is a disulfide bond. This bond can be formed between amino acids (e.g., cysteine or its derivatives) substituted from the amino acid residues of the F pocket of the MHC binding groove and / or the C-terminal anchor residue of the peptide, preferably from the amino acid residues of both the F pocket of the MHC binding groove and the C-terminal anchor residue of the peptide, which can form a disulfide bond. The following table shows the identity and position of preferred cysteine or its derivative mutations in various MHC class I molecules. The numbering refers to the position on the MHC heavy chain.

[0012]

Table 1

[0013] The most preferred mutations are cysteine substitutions or cysteine derivative substitutions at position 116 or 147. Particularly preferred mutations include cysteine substitutions or cysteine derivative substitutions at position F116 or S147 in HLA-E. Alternative positions for cysteine substitutions or cysteine derivative substitutions in HLA-E are L81, S143 or L124.

[0014] The pMHC complex can comprise two MHC subunit heavy and light chains. The MHC subunits associate with a peptide ligand, which binds to the binding groove formed by one or both subunits. Preferably, the MHC complex is a class I MHC complex. Alternatively, the MHC complex is a class II MHC complex. The MHC complex can be soluble. Methods for producing soluble complexes are known in the art; for example, the heavy chain of a class I complex can be shortened to remove the transmembrane and cytoplasmic regions. Preferably, the MHC complex is of human origin and can be referred to as a human leukocyte antigen (HLA) complex instead of an MHC. Alternatively, the MHC complex can be of another species (e.g., mouse or non-human primate). The MHC complex can be classical or non-classical. Classical human MHC complexes include multimeric HLA-A, HLA-B and HLA-C. Non-classical MHC complexes that present peptide ligands include HLA-E, HLA-F and HLA-G. Preferably, the MHC complex is non-classical HLA-E.

[0015] The pMHC complex may contain one or more mutations within the MHC subunit relative to the native MHC complex. Mutations include substitutions, insertions, and deletions. Preferably, the mutations are made at one or more positions within the F pocket of the peptide-binding groove. Alternatively or in addition, mutations (including insertions, substitutions, or deletions) may be made at other positions within the MHC, provided that they do not interfere with the recognition by the stability or binding portion of the isolated complex. Mutations within the F pocket of the binding groove include substitutions of one or more amino acids to cysteine.

[0016] The pMHC complex comprises a peptide ligand, which may be referred to as an MHC-binding peptide. The MHC-binding peptide can be 8 to 30 amino acids in length. The peptide can be 8, 9, 10, 11, 12, 13, or 14 amino acids in length or 15 amino acids in length or more. Preferably, the MHC-binding peptide is 9 amino acids in length. The MHC-binding peptide may have an amino acid sequence corresponding to that of the native MHC-binding peptide sequence. Alternatively, the MHC-binding peptide may contain one or more mutations relative to the amino acid sequence of the native MHC-binding peptide. Mutations may include substitutions, insertions, and deletions. Preferably, the MHC-binding peptide is an HLA-E-binding peptide. Alternatively or in addition, the MHC-binding peptide may bind to other MHC complexes. There are many examples of MHC-binding peptides known in the art. The MHC-binding peptide may be obtained from a foreign protein, including a viral or bacterial protein, or from an endogenous self-protein. Methods for identifying MHC-binding peptides are known in the art, for example, in silico prediction (e.g., SYFPEITHI, (Rammensee et al., Immunogenetics (1999) 50: 213-219) and NetMHCpan (Jurtz et al., J Immunol. 2017 Nov 1;199(9):3360-3368)) and / or mass spectrometry to identify peptide-MHC complexes eluted from the cell surface.

[0017] MHC-binding peptides may contain mutations at one or more positions involved in binding to MHC. As is known to those skilled in the art, MHC-binding peptides contain anchor residues that are involved in stabilizing the interaction between the peptide and the MHC binding groove. The binding groove of MHC class I is closed at both ends by conserved tyrosine residues, which usually limits the size of the bound peptide to 8-10 residues, and the C-terminal portion of the peptide docks into the F pocket of MHC. The positions and identities of the anchor residues are known (Falk et al., Nature. 1991 May 23;351(6324):290-6). For example, the anchor residues of peptides that bind to HLA-A2 are located at positions 2 and 9. Similar positions of anchor residues have also been found for HLA-E-binding peptides (Lampen et al., Mol Immunol. 2013 Jan;53(1-2):126-31). Typically, the identity of the amino acids at the anchor positions is fixed or shows limited variation. For all MHC-binding peptides, the C-terminal anchor residue is hydrophobic, and its side chain is accommodated within the deep hydrophobic F pocket of the MHC binding groove. Preferably in the present invention, the MHC-binding peptide is mutated at the C-terminal anchor residue (designated as P9 or PΩ). When the MHC-binding peptide binds to HLA-E, the C-terminal anchor residue can be at position 9 of the peptide. Peptides that bind to HLA-E have a strong preference for leucine at P9. The mutation can be a substitution to an amino acid that can form a disulfide bond (e.g., the natural amino acid cysteine), or the substitution can be to a non-natural amino acid that can form a disulfide bond.

[0018] At least one of the amino acids substituted from the amino acids of the F pocket of the MHC binding groove and / or the C-terminal anchor residue of the peptide can be a non-natural amino acid, which can be a non-natural amino acid capable of forming a disulfide bond. It is preferred that the C-terminal anchor residue of the peptide be so substituted. Examples of preferred unnatural amino acids that can form disulfide bonds include homocysteine, and homocysteine analogs having an extended carbon side chain with additional (e.g., 1 or 2) methyl groups incorporated. Preferred examples include 2-amino-5-sulfanyl-pentanoic acid (referred to herein as "h3C") (e.g., those provided by Chem-Impex International Inc., Cat No 29777; or those provided by Iris Biotech GmbH, Cat. No. # 917883-62-6) and 2-amino-6-sulfanylhexanoic acid (referred to herein as "h4C") (which is available, for example, as a custom synthesis from Creative Chemistry Solutions). Unnatural amino acids (including h3C and h4C) can be in the D or L isomeric configuration.

[0019]

Chemical formula

[0020] The chemical structures of homocysteine, h3C, and h4C are shown below:

Chemical formula

[0021] The extended carbon side chains of h3C and h4c mean that the length of the disulfide bridge formed between this and a cysteine residue increases. The following is a schematic diagram showing that the length of the disulfide bridge formed between H3C or H4C and cys is increased relative to cys-cys.

Chemical formula

[0022] The C-terminal anchor residue of the peptide is replaced with one of these unnatural amino acids (preferably, h3C or h4C) capable of forming a disulfide bond, and the amino acid (preferably, residue 116 or 147) in the F pocket of the MHC binding groove is preferably replaced with cysteine. Taking this information into account, one of ordinary skill in the art can determine which unnatural amino acid (h3C or h4C) should be substituted for the C-terminal anchor residue of the peptide and which residue (116 or 147) in the F pocket of the MKC binding groove should be substituted with cysteine. The disulfide bond is preferably formed between h3C at the C-terminal anchor position of the peptide and cysteine at position 147 of HLA-E, or between h4C at the C-terminal anchor position of the peptide and cysteine at position 116 of HLA-E. Alternatively, the disulfide bond is formed between h3C at the C-terminal anchor position of the peptide and cysteine at position 116 of HLA-E, or between h4C at the C-terminal anchor position of the peptide and cysteine at position 147 of HLA-E.

[0023] Methods for producing pMHC complexes are known in the art. Typically, MHC complexes are recombinantly produced in a bacterial expression system and refolded in vitro with synthetic peptides. Suitable methods are provided by Garboczi et al. (Proc Natl Acad Sci U S A. 1992 Apr 15;89(8):3429-33) and are further described in Example 1 of this specification. MHC-binding peptides may be produced chemically, which means they are chemically synthesized. Methods for producing synthetic peptides are known in the art, particularly solid-phase peptide synthesis (SPPS), also known as Merrifield synthesis. The complex of the present invention can be in an isolated and / or substantially pure form. For example, the complex can be provided in a form that is substantially free of other polypeptides or proteins. The pMHC complex can be in a soluble form, which means that the MHC complex can be shortened such that its transmembrane and cytoplasmic regions are removed. The pMHC complex of the present invention may be further modified. For example, the pMHC complex may be fused to a therapeutic moiety and / or attached to a solid support and / or fused to a tag such as a biotin tag and / or may be in a multimeric form. The tag may be at the C-terminus.

[0024] The pMHC complex of the present invention may be (covalently or otherwise) linked to a moiety that can induce a therapeutic effect. Such a moiety can be a carrier protein known to be immunogenic. KLH (keyhole limpet hemocyanin) and bovine serum albumin are examples of suitable carrier proteins used in vaccine compositions. Alternatively, the polypeptide and / or polypeptide-MHC complex of the present invention may be linked to a fusion partner. The fusion partner can be used for detection purposes or to attach the complex to a solid support or for oligomerization of pMHC. The pMHC complex may incorporate a biotinylation site to which biotin can be added, for example, using the BirA enzyme (O'Callaghan et al., 1999 Jan 1;266(1):9-15). Other suitable fusion partners include, but are not limited to, fluorescent or luminescent labels, radiolabels, nucleic acid probes and contrast reagents, antibodies, or enzymes that generate detectable substances. Detection methods include flow cytometry, microscopy, electrophoresis or scintillation counting. Examples of fusion partners include cytokines such as interleukin 2, interferon α and granulocyte-macrophage colony-stimulating factor.

[0025] An isolated peptide-MHC complex can be immobilized by attaching it to a suitable solid support. Examples of solid supports include, but are not limited to, beads, membranes, sepharose, magnetic beads, plates, tubes, columns. The peptide-MHC complex can be attached to an ELISA plate, magnetic beads or a surface plasmon resonance biosensor chip. Methods for attaching a peptide-MHC complex to a solid support are known to those skilled in the art and include, for example, using an affinity binding pair such as biotin and streptavidin or an antibody and an antigen. In a preferred embodiment, the peptide-MHC complex is labeled with biotin and attached to a surface coated with streptavidin.

[0026] In a second aspect, the invention provides multimers of the complexes of the first aspect. The pMHC complexes of the invention can be in multimeric forms, such as dimers or tetramers or pentamers or octamers or higher multimers. Examples of suitable methods for the production of multimeric peptide-MHC complexes are described in Greten et al. (Clin. Diagn. Lab. Immunol. 2002 Mar;9(2):216-20) and references therein and Wooldridge et al. (Immunology (2009) 126(2):147-64). Generally, peptide-MHC multimers can be produced using peptide-MHC tagged with biotin residues and complexed via fluorescently labeled streptavidin. Alternatively, multimeric polypeptide-MHC complexes can be formed by using an immunoglobulin as a molecular scaffold. In this system, the extracellular domain of the MHC molecule is fused to the constant region of the immunoglobulin heavy chain, separated by a short amino acid linker. Polypeptide-MHC multimers have been produced using a carrier molecule (e.g., dextran) (WO02072631). Multimeric peptide-MHC complexes can be useful for improving the detection of binding moieties (e.g., T cell receptors) that bind the complex due to the avidity effect.

[0027] In a third aspect, the present invention provides a method for producing a stabilized pMHC complex of the first aspect, which includes forming an unnatural linkage between a C-terminal amino acid anchor residue of a peptide and an amino acid residue within the F pocket of the MHC binding groove. A convenient method for producing MHC useful in the present invention is to express the nucleic acid encoding the MHC in an expression system. The present invention further provides an isolated nucleic acid encoding a stabilized MHC useful in the present invention. Nucleic acids include DNA and RNA. Those skilled in the art can determine substitutions, deletions and / or additions to the nucleic acids that still provide a stabilized MHC useful in the present invention.

[0028] The present invention also provides constructs in the form of plasmids, vectors, transcription or expression cassettes, which contain at least one of the above nucleic acids. The present invention also provides recombinant host cells containing one or more of the above constructs. As described above, the nucleic acid encoding a stabilized MHC useful in the present invention constitutes one aspect of the present invention, similar to the method for producing a stabilized MHC including the expression from the nucleic acid encoding it. Expression can be readily achieved by culturing recombinant host cells containing the nucleic acid under appropriate conditions. After production by expression, the stabilized MHC can be appropriately used after being isolated and / or purified using any appropriate technique. Systems for cloning and expressing polypeptides in various host cells are well known. Suitable host cells include bacteria, mammalian cells, yeast and baculovirus systems. Mammalian cell lines available in the art for the expression of heterologous polypeptides include Chinese hamster ovary cells, HeLa cells, baby hamster kidney cells, NSO mouse melanoma cells and many other cells. A common and preferred bacterial host is E. coli. Expression in prokaryotic cells (e.g., E. coli) is well established in the art.

[0029] Appropriate regulatory sequences (including promoter sequences, terminator sequences, polyadenylation sequences, enhancer sequences, marker genes) and other necessary sequences can be appropriately selected or constructed. The vector can be, if necessary, a plasmid, viral, for example, phage or phagemid. For further details, see, for example, Sambrook et al., 1989. Molecular Cloning: a laboratory manual (Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory, 1989). Many known techniques and protocols for nucleic acid manipulation (e.g., preparation of nucleic acid constructs, mutagenesis, sequencing, introduction of DNA into cells and in gene expression) and protein analysis are described in detail in Ausubel et al., Short protocols in molecular biology: a compendium of methods from Current protocols in molecular biology (Brooklyn, NY: Green Pub. Associates: New York, NY: Wiley).

[0030] Thus, a further aspect of the present invention provides a host cell comprising a nucleic acid disclosed herein. Yet another aspect provides a method comprising introducing the nucleic acid into a host cell. The introduction may employ any available technique. For eukaryotic cells, suitable techniques include calcium phosphate transfection, DEAE-dextran, electroporation, liposome-mediated transfection, and transduction using retroviruses or other viruses (e.g., baculovirus for vaccinia or insect cells). For bacterial cells, suitable techniques include calcium chloride transformation, electroporation, and transfection using bacteriophage. After introduction, expression from the nucleic acid may be induced or enabled, for example, by culturing the host cell under conditions of gene expression. The nucleic acid of the present invention may be integrated into the genome (e.g., chromosome) of the host cell. Integration may be facilitated by including sequences that promote recombination with the genome according to standard techniques. The present invention also provides a method comprising expressing the above-described stabilized MHC using the above construct in an expression system.

[0031] The present invention also provides the following: combining a complex of the first aspect with a T cell receptor (TCR), a TCR-mimicking antibody, or a population of T cells; and identifying a TCR, a TCR-mimicking antibody, or a T cell that binds to the complex and provides a screening method. Preferred features of each aspect of the present invention are the same above for each of the other aspects, with the necessary modifications. Prior art documents referred to herein are incorporated to the maximum extent permitted by law. BRIEF DESCRIPTION OF THE DRAWINGS

[0032]

Figure 1

Figure 2

Figure 3

Example

[0033] Example Here, the present invention will be described with reference to the following non-limiting examples and drawings. Figure 1 shows the stability of the indicated pMHC complex determined by the loss of ILT2 binding over time. The native pHLA-E complex exhibits limited stability. Figure 2 shows the difference in TCR recognition between a native pMHC complex and an equivalent pMHC complex stabilized using existing methodologies. Figure 3 shows the difference in TCR recognition between a native pMHC complex and an equivalent stabilized pMHC complex of the present invention.

[0034] Example 1 - Isolated peptide HLA-E complexes have limited stability This example demonstrates that isolated peptide HLA-E complexes have a short half-life, meaning they are not stable enough to be used for the identification and characterization of binding substances (such as TCRs and antibodies). For this purpose, a half-life of at least 4 hours is typically preferred, and a half-life substantially exceeding this is desirable. Stability is evaluated using several peptides known to be presented by HLA-E. These peptides include the MTB peptide and the HIV peptide described by Joosten et al. (PLoS Pathog. 2010 Feb 26;6(2):e1000782) and Hansen et al. (Science. 2016 Feb 12;351(6274):714 - 20), respectively, and two self-peptides corresponding to the leader peptides from HLA-A * 02 and HLA-Cw3.

[0035] Method Peptide The peptide was obtained by chemical synthesis from Peptide Protein Research Ltd and solubilized to a concentration of 4 mg / ml in DMSO before use. HLA-E * 01:01 and HLA-E * Preparation of 01:03 peptide complex The HLA-E heavy chain and β2-microglobulin (β2m) were separately expressed as inclusion bodies in E. coli and then refolded and purified using the method described previously (Garboczi et al., Proc Natl Acad Sci U S A. 1992 Apr 15;89(8):3429-33). The HLA-E heavy chain contained a C-terminal biotinylation tag (AviTag TM GLNDIFEAQKIEWHE), but the transmembrane and cytoplasmic domains were removed. Briefly, the HLA-E heavy chain and β2m were mixed with the peptide of interest at a ratio of 30:5:2 and refolded. The refolded soluble pHLA was then purified using a two-step protocol incorporating anion exchange followed by size exclusion chromatography (SEC). To generate the biotinylated complex, the biotinylation step was included after anion exchange and before SEC using biotin-protein ligase (BirA) as described by O'Callaghan et al., Anal Biochem. 1999 Jan 1;266(1):9-15.

[0036]

Chemical formula

[0037] AviTag TM Underline the sequences of AviTag and its GSGG linker, and show F116 and S147 in bold and underlined.

Chemical formula

[0038] Evaluation of the stability of peptide-HLA complexes The stability of the peptide-HLA-E complex was evaluated by surface plasmon resonance (SPR) using a BIAcore T200 instrument. Purified biotinylated peptide-HLA-E monomers of approximately 500 - 1000 response units (RU) were captured on a streptavidin-coupled CM-5 series S sensor chip. Soluble form of Ig-like transcript 2 receptor (ILT2) with enhanced affinity was flowed over the chip surface at a flow rate of 10 μl / min for 60 seconds at a concentration of 1 μM. Since ILT2 binds to class I HLA molecules in a conformation-dependent manner, it was used as an indicator of the stability of the complex. The binding of ILT2 to the pHLA-E complex was measured at regular intervals over 5 hours, and then the response was normalized for ILT2 binding by subtracting the response of the bulk buffer in a control flow cell without peptide-HLA. The binding half-life (T 1 / 2 ) was calculated by plotting the % activity against time using BIA T200 evaluation software and GraphPad Prism 8.

[0039] Results Table 1 below provides the half-life (T * ) of each of the indicated peptides complexed with HLA-E 1 / 2 as determined by ILT2 binding. All of the indicated complexes have a half-life of 5 hours or less, and some complexes have a half-life of 1 hour or less. A representative example of the binding data is provided in Figure 1 (left panel).

[0040]

Table 2

[0041] Example 2 - Peptide HLA-E complexes can be stabilized by the cys trap method but show disrupted TCR binding A) In this example, the HLA-E heavy chain was modified such that a cysteine mutation was incorporated at position Y84; the peptide was modified such that an additional 3 amino acids (Gly-Cys-Gly) were included at the C-terminus. This approach, generally referred to as "Cys trap", has been used successfully to improve the stability of various HLA complexes by "trapping" the peptide in the binding groove (Truscott J Immunol 2007 May 15;178(10):6280-9; Mitaksov et al., Chem Biol. 2007 Aug;14(8):909-22).

[0042] Method The same experimental method as described in Example 1 was used. Results Table 2 below provides the half-life of each of the indicated cys-captured peptide-HLA-E complexes as determined by ILT2 binding. All of the indicated complexes have a substantially extended half-life compared to the unmodified complex shown in Example 1, with most exceeding 20 hours. A representative example of the binding data is provided in Figure 1 (right panel).

[0043]

Table 3

[0044] B) Subsequently, the Cys-captured stabilized peptide-HLA-E complex containing the MTB peptide RLPAKAPLL+GCG was tested for recognition by antigen-specific TCRs and compared to the unmodified complex. This peptide was selected because the unmodified native peptide-HLA-E complex has a relatively long half-life and is thus suitable for the evaluation of TCR binding.

[0045] Method Evaluation of TCR binding to peptide-HLA-E complexes Four TCRs that recognize the MTB peptide RLPAKAPLL - HLA - E complex were isolated from a naive phage library and prepared as soluble αβ heterodimers as previously described (Boulter et al., Protein Eng. 2003 Sep;16(9):707 - 11).

[0046] Determination of binding characteristics Binding analysis of the purified soluble TCR to the peptide - HLA complex was performed by surface plasmon resonance (SPR) using a BIAcore T200 instrument. Biotinylated pHLA - E molecules were refolded with the peptide of interest as described in Example 1 above. All measurements were performed at 25 °C in Dulbecco's PBS buffer supplemented with 0.005% surfactant P20. The biotinylated peptide - HLA - E monomer was immobilized on a streptavidin - binding CM - 5 series S sensor chip. The equilibrium binding constant was determined using serial dilutions of the soluble TCR injected at a constant flow rate of 10 - 30 μl / min onto a flow cell coated with approximately 1000 response units (RU) of the peptide - HLA - E * 01:03 complex. The equilibrium response was normalized for each TCR concentration by subtracting the response of the bulk buffer in a control flow cell without peptide - HLA. K D values were obtained by non - linear curve fitting using GraphPad Prism 8 software and the Langmuir binding isotherm Binding = C×Max / (C + KD) (where "Binding" is the equilibrium binding (in RU) at the injected TCR concentration C and Max is the maximum binding).

[0047] For high-affinity interactions, the binding parameters were determined by single-cycle kinetics analysis. Five different concentrations of soluble TCR were injected onto a flow cell coated with a peptide-HLA complex of approximately 50-200 RU at a flow rate of 50-60 μl / min. Typically, 60-200 μl of soluble TCR was injected at the highest concentration of 100-1000 nM, and two-fold serial dilutions were used for the other four injections. The lowest concentration was injected first. To measure the dissociation phase, buffer was injected until more than 10% dissociation occurred (typically after 1-3 hours). The kinetic parameters were calculated using BIAevaluation® or BIAcore T200 evaluation software. The dissociation phase was fitted to a monoexponential decay equation to calculate the half-life. The equilibrium constant K D is k off / k on calculated from.

[0048] Results The binding affinities shown in Table 2 below, together with the binding curves shown in Figure 2, indicate that while antigen-specific TCR can recognize the native non-stabilized peptide HLA-E complex, the recognition of the cys-trap stabilized complex is substantially reduced and in some cases below the detection level.

[0049]

Table 4

[0050] Example 3 - Preparation of a Stable Peptide HLA-E Complex with Minimal Change in TCR Recognition A) In this example, the peptide-HLA-E complex was modified such that a novel engineered disulfide bond was incorporated between the peptide-binding groove of the HLA-E heavy chain and the C-terminal anchor residue of the peptide. To create the novel disulfide bond, the P9 anchor residue of the RLPAKAPLL peptide, an MTB peptide, was modified to the unnatural amino acid L-3-C-homocysteine (2-amino-5-sulfanyl-pentanoic acid) (RLPAKAPL-h3C), and either the F116 or S147 position of the HLA-E heavy chain was mutated to cysteine.

[0051] Method As described in Example 1, peptide-HLA-E complexes were prepared and evaluated for stability. TCR binding was evaluated as described in Example 2. Results Table 4 below demonstrates that the novel disulfide bond resulted in a substantial improvement in stability, as indicated by increasing the half-life of the complex compared to the native complex.

[0052] [Table 5]

[0053] To demonstrate that the novel stabilized peptide-HLA-E complex retains native-like TCR recognition, binding was evaluated for nine different TCRs isolated from a phage library with respect to the improved recognition of the peptide (RLPAKAPLL)-HLA-E complex. In each case, the kinetics of TCR binding to the stabilized complex were compared to those of the native complex. Tables 5 and 6 show that TCR binding to the stabilized complexes (with cysteine mutations at F116 or S147, respectively) is preserved in all cases. For each TCR, only a slight difference in binding was observed between the stabilized and native complexes, indicating that the peptide adopts a near-native-like conformation. Figure 3 shows the binding curves for four TCRs from Table 5.

[0054]

Table 6

[0055]

Table 7

[0056] B) In a further embodiment, the P9 anchor residue of the RLPAKAPLL peptide, which is an MTB peptide, was modified to the unnatural amino acid L-4-C homocysteine (2-amino-6-sulfanylhexanoic acid) (RLPAKAPL-h4C), and the F116 or S147 position of the HLA-E heavy chain was mutated to cysteine. The stability of the complex and TCR binding were evaluated as described in part A.

[0057] Results The binding half-life of the resulting complex was 24.47 hours. This demonstrates that the novel disulfide has brought about a substantial improvement in stability compared to the native complex (shown in Table 4). TCR binding was evaluated for six TCRs. In all six cases, TCR binding to the stabilized complex was preserved. The difference in binding compared to the native complex was in the range of 1.53 to 3.24-fold for the disulfide with F116 and in the range of 1.11 to 2.83-fold for the disulfide with S147.

[0058]

Table 8

Claims

1. (a) one amino acid residue at the position of the C-terminal anchor residue of the peptide; (b) the amino acid residue at position 116 or 147 in the F pocket of the MHC binding groove containing an unnatural covalent bond therebetween, and at least one of the amino acid residues (a) and (b) being substituted, A stabilized peptide-MHC (pMHC) complex.

2. The complex according to claim 1, wherein both of the amino acid residues (a) and (b) are substituted.

3. The complex according to claim 1 or 2, wherein the unnatural covalent bond is a disulfide bond.

4. The complex according to claim 3, wherein the amino acid residue (b) is substituted with cysteine.

5. The complex according to any one of claims 1 to 4, wherein the amino acid residue (a) is an unnatural amino acid.

6. The complex according to claim 5, wherein the amino acid residue (a) is substituted with a homocysteine analog that differs from homocysteine only in having an extended carbon side chain.

7. The complex according to claim 6, wherein the homocysteine analog is 2-amino-5-sulfanyl-pentanoic acid or 2-amino-6-sulfanylhexanoic acid.

8. The complex according to any one of claims 1 to 7, which is soluble.

9. The complex according to any one of claims 1 to 8, wherein the MHC contains a biotinylation tag, which may or may not be at the C-terminus.

10. The complex according to any one of claims 1 to 9, wherein the MHC is HLA-E.

11. The multimer of the complex according to any one of claims 1 to 10.

12. A method for producing a peptide-MHC complex according to any one of claims 1 to 10, comprising forming a covalent bond between the amino acid residue at position 116 or 147 in the F pocket of the MHC binding groove and one amino acid residue located at the position of the C-terminal amino acid anchor residue of the peptide.

13. Combining the complex according to any one of claims 1 to 10 with a T cell receptor (TCR), a TCR-mimicking antibody or a population of T cells; and Identifying a TCR, a TCR-mimicking antibody or a T cell that binds to the complex A screening method comprising the above.

Images