WO2004048410A2

WO2004048410A2 - T-cell receptors

Info

Publication number: WO2004048410A2
Application number: PCT/GB2003/005105
Authority: WO
Inventors: Edith Yvonne Jones; Guillaume B. E. Stewart-Jones
Original assignee: Isis Innovation Limited
Priority date: 2002-11-22
Filing date: 2003-11-24
Publication date: 2004-06-10
Also published as: AU2003302405A8; WO2004048410A3; GB0227351D0; AU2003302405A1

Abstract

The present invention relates to a T-cell receptor (TCR), nucleic acids encoding a TCR and a method of forming a TCR. The present invention also relates to the use of TCRs in diagnostic methods, for treating autoimmune disorders and graft rejection, as vaccines and for identifying agents that modulate TCR function.

Description

T-cell Receptors

The production of recombinant functional TCRs has encountered great difficulties in many laboratories. Attempts to engineer and produce TCRs in a variety of formats via different expression routes, in order to identify a general methodology for TCR synthesis, has led to the conclusion that no method has yet been of general utility.

TCRs are heterodimeric proteins found on the surface of CD4+ or CD8+ cytotoxic T-cells. They guide the T-cells to antigen-presenting cells (APCs) by their specific affinity to particular antigenic peptide-MHC (class I or class II) combinations presented on the surface of target cells. The association of an antigen-specific TCR with the cognate antigenic pMHC complex results in the formation of an immunological synapse that in the case of CD8+ cytotoxic T cells leads to lysis or apoptosis of the APC. The structure of the TCR is akin to the antibody Fab fragment; it has 4 immunoglobulin domains, two pairing constant and two pairing variable domains, and bears on the end of the variable domains the complementarity-determining regions (CDRs) that recognize the pMHC complex.

Eukaryotic insect expression systems have produced full-length glycosylated TCR in low yields for crystallographic studies (Garcia et al., PNAS USA, 94, 13838-13843, 1997), and mammalian cell expression systems have yielded some single chain (scFv) TCRs (VαVβ linked by a soluble ~16mer peptide) (Gregoire et al., Eur. J. Immunol., 26, 2410-2416, 1996). Some scFv TCRs have displayed instability and aggregation on storage and may therefore not be very robust (Pecorari et al, J. Mol. Biol., 285, 1831-43, 1999). Refolding studies on a bacterially expressed murine TCR (UZ3-4) lacking the transmembrane region has demonstrated that there was a narrow window of optimal redox balance and concentration for successful heterodimeric refolding (Pecorari et al, 1999 (supra)). However, the TCR αβ dimer displayed strong tendencies to dissociate into monomers. Subsequent fusion with jun/fos zippers significantly decreased the refolding yield and the presence of the native interchain disulfide bridge prevented refolding.

The use of carboxy terminal jun/fos fusion constructs by another group permitted JM22 and LCI 3 TCRs to be successfully refolded with high yields of relatively pure heterodimers being obtained (-10% refolding yield) from E. coli inclusion bodies (Willcox et al, Protein Science, 8, 2418-2423, 1999).

The instability of some scFv TCRs in solution and general difficulties in refolding stable dimerisation-motif free heterodimers may reflect the importance of the interchain disulfide bridge which may have evolved as an integral feature of the TCR structure for dimer stabilization of the diverse TCR VαVβ pairings derived from the germline repertoire.

The current range of methods for producing T-cell receptors via different formats and production systems are given in Table 1 below.

TABLE 1

see Hennecke et al., EMBO J., .19, 5611-5624, 2000. The αβ chain pair affinity of UZ3-4 TCR (Pecorari et al., 1999 (supra)) was measured as 1 μM indicating that there is rapid exchange of the α and β chains in solution. Accordingly, solution state macromolecular thermodynamics would predict that under more dilute conditions, the TCR heterodimer would dissociate more into the constituent monomers.

Also, it is demonstrated herein that on proteolytic removal of the zippers from proteolysis-engineered αβ LCI 3, the heterodimer was observed to monomerise very rapidly (in minutes), and the α chain also aggregated within several hours and even more rapidly at high protein concentrations suggesting intrinsic insolubility for some TCR α monomers.

This would therefore indicate that an αβ chain dimerisation and/or bridging motif is required to stabilize the refolded TCR heterodimer for long-term storage and to retain vaccine or affinity functionality when used under dilute conditions in vitro or in vivo, whilst an αβ dimerisation motif is required to direct refolding equilibrium kinetics in vitro towards the folded energetic minimum.

Li et al, Immunology, 88, 524-530, 1996, discloses a number of mutations of the constant region of the TCRβ chain. It is concluded that the interchain disulphide bond between the TCRα and TCRβ chains is not required for TCRαβ heterodimer formation.

Arnaud et al., International Immunology, 9, 615-626, 1997, describes a number of studies relating to TCRαβ interaction.

Chang et al, PNAS USA, 9_ 11408-11412, 1994, describes the use of a leucine zipper to improve the association between the extracellular domains of the TCRα and β chains. The native interchain disulphide bond of the chains was used to link the chains.

US Patent No. 6,147,203, discloses using cysteine residues to stabilise the interaction of antibody fragments and the α and β chains of a TCR. International patent application WO 03/020763 discloses using disulpide bonds to link together the constant domains of TCRα and TCRβ chains.

In vivo use of recombinant TCRs for targeting antigen-presenting cells may benefit from low immunogenic properties of the TCR complex for multiple administration purposes. Therefore reducing the dimerisation motif to a minimum variation from the native protein is an important consideration, and large motifs (e.g. zippers) may result in unforeseen metabolic or immunological side-effects. It appears that a dimerisation motif is necessary for the stability of the TCR dimer. Without a dimerisation motif data suggests that the TCR would rapidly disassemble, lose antigen binding functionality and bias antibody development towards the core TCR heterodimerisation interfaces. The αβ heterodimeric protein surface would be the target for antibody development and conformational integrity would seem central to the selection of specific antibodies to TCR α and/or β chains for the treatment of autoimmune disorders. It is therefore desirable to provide the α and β pair in a context most similar to the natural target components to generate specific antibodies to linear and/or non-linear TCR α and/or β epitopes. Basically in order to generate antibodies to a TCR it is preferable that the context of the α and β chains of the TCR is most similar to the natural target so that antibodies are generated against epitopes which are identical to epitopes displayed on the natural TCR.

As indicated above, a number of production methodologies have been applied to TCRs, with various emergent solutions to the problem of functional protein synthesis on large scale. However, the stability, quality, immunogenicity and functionality of these proteins have been on the whole highly variable. It would therefore be useful to those synthesizing TCRs to obtain a reliable technology for the production of highly pure robust heterodimeric TCR of low immunogenicity.

According to a first aspect of the present invention there is provided a recombinant T-cell receptor (TCR) comprising a TCR α chain and a TCR β chain, wherein the TCR α chain comprises a Vα domain, a Cα domain and a first dimerisation motif attached to the C-terminus of the Cα domain, and the TCR β chain comprises a Vβ domain, a Cβ domain and a second dimerisation motif attached to the C-terminus of the Cβ domain, wherein the first and second dimerisation motifs easily interact to form a covalent bond between an amino acid in the first dimerisation motif and an amino acid in the second dimerisation motif linking the TCR α chain and the TCR β chain together.

The TCR α chain can be any TCR α chain provided it comprises a Vα domain, a Cα domain and a first dimerisation motif. The TCR β chain can be any TCR β chain provided it comprises a Vβ domain, a Cβ domain and a second dimerisation motif. Preferably the TCR α chains and TCR β chains are derived from naturally occurring TCRs, especially TCRs that recognize antigens associated with infectious diseases (e.g. HIV), cancer and autoimmune diseases. The naturally occurring TCRs may recognize an antigenic peptide displayed in combination with an MHC class I molecule or an MHC class II molecule. Sequences of suitable TCR α chains and TCR β chains are described in the art (see, for example, the IMGT database and the "TCR FactsBook" by Lefranc M.P., 2001). The TCR chains derived from naturally occurring TCRs may be modified provided the chains retain substantially the same functions as the corresponding naturally occurring chains. Suitable modifications include making a small number (e.g. about 1 to 20) of amino acid substitutions, additions and/or deletions.

The terms "Vα domain" and "Cα domain" are standard terms used in the art to describe the distinct domains of a TCR α chain. The Vα domain comprises the complementarity determining regions (CDRs), which detenmine the specificity of the TCR α chain. Although there is some variation between the sequences of different TCR α chains, generally the Vα domain comprises amino acids 1 to 115 and the Cα domain generally comprises amino acids 116 to 206 of an α chain. The exact position of the Vα domain and Cα domain in a TCR α chain can be determined by comparing the sequence of the α chain with the sequence of α chains wherein the positions of the Vα domain and Cα domain are known.

The terms "Vβ domain" and "Cβ domain" are standard terms used in the art to describe the distinct domains of the TCR β chain. The Vβ domain comprises the complementarity determining regions (CDRs), which determine the specificity of the

TCR β chain. Although there is some variation between the sequences of different

TCR β chains, generally the Vβ domain comprises amino acids 1 to 118 and the Cβ domain generally comprises amino acids 119 to 244 of a β chain. The exact position of the Vβ domain and Cβ domain in a TCR β chain can be determined by comparing the sequence of the β chain with the sequence β chains wherein the positions of the Vβ domain and Cβ domain are known.

The term "dimerisation motifs" refers to the regions attached to the C-terminus of the Cα domain and the Cβ domain, which have been specifically constructed to interact and form a covalent bond. Each dimerisation motif comprises an amino acid that forms a covalent bond with a corresponding amino acid in the other dimerisation motif. Each dimerisation motif may comprise at least some of the sequence attached to the C-terminus of the Cα or Cβ domain present in the native TCRα or TCRβ chain, provided at least one, preferably both, dimerisation motifs differ from the native sequence by the addition and/or substitution of at least one amino acid which induces the interaction of the dimerisation motifs.

The term "easily interact" as used herein means that at least one, preferably both, dimerisation motifs comprise amino acids that induce the dimerisation motifs to interact and form an interchain covalent bond more readily than the corresponding sequences attached to the C-terminus of the constant domains in the native TCR chains. In other words at least one, preferably both, of the dimerisation motifs have been modified to increase the interaction of the dimerisation motifs.

The first dimerisation motif comprises an amino acid that forms a covalent bond with an amino acid in the second dimerisation motif to link the TCR α chain and the TCR β chain together. The amino acids capable of forming a covalent bond may be natural amino acids or non-natural amino acids. The amino acids may also be modified post-translationally so that they can form a covalent bond, e.g. hydroxylysines. The amino acids are preferably capable of forming a disulfide bond. Preferably the amino acids capable of forming the disulfide bond are cysteines. The terni "amino acid" as used herein refers to natural, non-natural and post-translationally modified amino acids. The amino acids may be the L form or the D form.

The dimerisation motifs may be constructed to easily interact by comprising one or more of the following: amino acids that lead to electrostatic interactions between the dimerisation motifs; amino acids that increase the flexibility of the dimerisation motifs; amino acids that position at least one of the covalent bond forming amino acids away from the TCR constant domain; hydrophobic amino acids leading to hydrophobic interactions; and hydrophilic amino acids leading to hydrophilic interactions. Other types of interactions between the dimerisation motifs can also be induced by making suitable modifications.

According to a first particularly preferred embodiment of the present invention, the dimerisation motifs interact electrostatically.

The term "interacting electrostatically" refers to the situation wherein the first and second dimerisation motifs comprise oppositely charged amino acids at corresponding positions so that the dimerisation motifs are brought together via the electrostatic interactions of the oppositely charged amino acids. By ensuring that the oppositely charged amino acids are at corresponding positions on the dimerisation motifs, it is possible to ensure that correspondingly positioned amino acids capable of forming a covalent bond are brought into close proximity thereby encouraging the formation of a covalent bond between the TCR α chain and the TCR β chain. Preferably the oppositely charged residues are at corresponding positions distributed evenly along the dimerisation motifs. It is preferred that there are at least 2 pairs of oppositely charged amino acids.

The charged amino acids may be natural, non-natural, or post-translationally modified amino acids having a charge. The amino acids may be the L form or the D form. Polar amino acids such as glutamine, serine, threonine and asparagine are considered herein to be charged amino acids. Preferred charged amino acids include aspartic acid, arginine, lysine, histidine and glutamic acid.

It is particularly preferred that the C-terminal amino acid of the first dimerisation motif and the C-terminal amino acid of the second dimerisation motif are oppositely charged amino acids capable of interacting electrostatically.

It is also preferred that the amino acid forming the covalent bond in the first dimerisation motif is adjacent a charged amino acid and that the amino acid forming the covalent bond in the second dimerisation motif is adjacent a charged amino acid, wherein the charged amino acids interact electrostatically.

It has been found that by positioning oppositely charged amino acids at the C-terminus of the TCRα and TCR β chains, and having the amino acids capable of forming the covalent bond adjacent to oppositely charged amino acids, that the dimerisation motifs easily interact to form a covalent bond.

According to a second particularly preferred embodiment of the present invention, the dimerisation motifs are flexible.

The term "flexible" as used herein means that the dimerisation motifs do not form any secondary structure such as α helixes or β structures and are more flexible than the corresponding regions of native TCRα and β chains. It is particularly preferred that at least one, preferably both, dimerisation motifs comprise one or more amino acids that allow flexibility. Particular amino acids that allow flexibility include glycine, serine and alanine. Most preferably the amino acids that allow flexibility are glycine. Preferably at least one, more preferably both dimerisation motifs comprise at least 3 amino acids that allow flexibility.

It has been found that the ability of the dimerisation motifs to interact and form an interchain covalent bond is increased by making them flexible. According to a third particularly preferred embodiment of the present invention, the first dimerisation motif is constructed so that the amino acid that forms the interchain covalent bond is separated from the Cα domain by at least 5 amino acids.

Preferably the covalent bond forming amino acid is separated from the Cα domain by between 6 and 20 amino acids, more preferably between 8 to 15 amino acids, most preferably 8 + 1 amino acids.

Alternatively, or in addition to the third preferred embodiment of the present invention, the second dimerisation motif of the TCR of the present invention is constructed so that the amino acid that forms the interchain covalent bond is separated from the Cβ domain by at least 3 amino acids.

Preferably, the covalent bond forming amino acid is separated from the Cβ domain by between 4 and 20 amino acids, more preferably between 5 and 15 amino acids, most preferably by 9 + 1 amino acids.

Preferably the amino acid in the first dimerisation motif that forms the interchain covalent bond is separated from the Cα domain by at least 5 amino acids and the amino acid in the second dimerisation motif that forms the interchain covalent bond is separated from the Cβ domain by at least 3 amino acids.

By distancing at least one, preferably both, amino acids that form the interchain covalent bond from the constant domains of the TCR chains it has been found that the covalent bond forming amino acids interact more easily.

Preferably, the covalent bond forming amino acids are separated from the constant domains by the same number of amino acids in the TCRα and β chains.

It is also preferred that the dimerisation motifs of the TCR according to the third preferred embodiment of the present invention, and the alternative third preferred embodiment, are flexible. Flexible dimerisation motifs are described above. Preferably, dimerisation motifs of the TCR according to the first prefeιτed embodiment of the present invention are also specifically constructed to be flexible. Such flexible dimerisation motifs are defined with respect to the second preferred embodiment of the present invention.

Preferably, the dimerisation motifs of the TCR according to the third preferred embodiment of the present invention, and the alternative third preferred embodiment, are also specifically constructed so that they interact electrostatically. Such dimerisation motifs are as defined with respect to the first preferred embodiment of the present invention. Preferably, the dimerisation motifs are also flexible as defined with respect to the second preferred embodiment of the present invention.

Preferably, the first dimerisation motif is from 6 to 21, more preferably 9 to 16, and most preferably 9 + 1 amino acids in length.

Preferably, the second dimerisation motif is from 4 to 21, more preferably from 6 to 16, more preferably 10 + 1 amino acids in length.

The TCR according to any embodiment of the present invention may additionally comprise a tag. The tag may be attached at any suitable position. The tag may be attached to the TCR α chain or the TCR β chain. Preferably the tag is attached to the first and/or second dimerisation motif.

The tag may be any detectable tag allowing the identification and/or isolation of the TCR. Suitable tags are well known to those skilled in the art. In particular, the tag may be a polyhistidine peptide (e.g. 6 Histidines), a biotinylation sequence, a fluorescent label, metal ion, FLAG tag, GST, T7 (Promega), etc.

Preferably the tag is proteolytically removable allowing the tag to be removed by suitable treatment. In particular, if the tag is attached to the free carboxy end of the first or second dimerisation motif, it can be removed using a carboxypeptidase provided a lysine residue is present to restrict the progression of the carboxypeptidase.

In a preferred embodiment, the first dimerisation motif of the TCR according to the present invention has the formula:

-(Y)m-(Zl)p-(Y)m-(F)q-(Y)m-B-(Z2)p (I) wherein,

Y is any amino acid except an amino acid that is capable of forming a disulfide bond; m is independently 0 to 10; Zl is a charged amino acid; p is independently 1 to 40;

F is an amino acid that increases flexibility of the dimerisation motif region; q is 1 to 10;

B is an amino acid capable of forming a disulfide bond; and Z2 is a charged amino acid having the opposite charge to Zl , wherein the dimerisation motif is at least 6 amino acids in length.

In a preferred embodiment, the second dimerisation motif of the TCR according to the present invention has the formula: -(Y)m-(Z2)p-(Y)m-(F)q-(Y)m-B-(Zl)p (II) wherein,

Y is any amino acid except an amino acid that is capable of forming a disulfide bond; m is independently 0 to 10;

Zl is a charged amino acid; p is independently 1 to 40;

B is an amino acid capable of forming a disulfide bond; and

Z2 is a charged amino acid having the opposite charge to Zl, wherein the dimerisation motif is at least 4 amino acids in length.

In formula (I) and formula (II) preferably m is independently 1 to 3. In formula (I) and formula (II) it is preferred that p is independently 1 to 3.

In formula (I) and formula (II) preferably p is 1.

In formula (I) and formula (II) preferably q is 3.

Preferably the dimerisation motif represented by formula (I) is between 7 and 21 amino acids in length, more preferably between 9 and 16 amino acids, most preferably 9 + 1 amino acids in length.

Preferably the dimerisation motif represented by formula (II) is between 5 and 21 amino acids in length, more preferably between 6 and 16 amino acids in length, most preferably 10 + 1 amino acids in length.

In a particularly preferred embodiment, the first dimerisation motif of the TCR according to the present invention has the sequence: PENDGGGCK

In a particularly preferred embodiment, the second dimerisation motif of the TCR according to the present invention has the sequence: ADQDRGGGCD

As indicated above, the dimerisation motifs may comprise a tag. It is particularly preferred that the first dimerisation motif comprises a 6 histidine tag attached to the C-terminus.

Preferably the TCR according to the present invention additionally comprises an active molecule. The active molecule may be attached to the TCR at any position provided it does not interfere with the activity of the TCR. The active molecule can be any molecule having a desired activity. For example, the active molecule may be a radioisotope, a prodrug converting enzyme, a toxin or a diagnostic label. In Hennecke et al, (supra) it is indicated that an antigenic peptide attached to the N-terminus of a TCR α chain or a TCR β chain increases the stability of the interaction between the α chain and the β chain when in contact with the corresponding MHC. Accordingly , the active molecule may be an antigenic peptide which functions to increase the stability of the TCR.

It is also preferred that the TCR of the present invention is a soluble TCR.

The present invention also provides the TCR α chain of the TCR according to the first aspect of the present invention. The TCR α chain comprises a Vα domain, a Cα domain and a first dimerisation motif as defined above. The TCR α chain may additionally comprise a tag. The tag is as defined above.

The present invention also provides the TCR β chain of the TCR according to the first aspect of the present invention. The TCR β chain comprises a Vβ domain, a Cβ domain and a second dimerisation motif as defined above. The TCR β chain may additionally comprise a tag. The tag is as defined above.

If the TCR α chain and the TCR β chain both comprise a tag, it is preferred that they comprise different tags so that the TCR α chain and the TCR β chain can be distinguished and/or separately isolated.

The present invention also provides a nucleic acid molecule encoding a TCR according to the first aspect of the present invention.

The present invention, also provides a nucleic acid molecule encoding the TCR α chain according to the present invention.

The present invention, also provides a nucleic acid molecule encoding the TCR β chain according to the present invention. The nucleic acid molecule may be any nucleic acid, such as DNA or RNA. Preferably the nucleic acid is DNA. The nucleic acid is preferably in the form of a vector and comprises the necessary regulatory elements enabling expression of the nucleic acid, such as promoters, transcription termination sequences, etc. Suitable vectors for the expression nucleic acids are well known to those skilled in the art. The sequence of the nucleic acid encoding the TCR, TCR α chain or TCR β chain may be varied depending on the optimal codon usage of the host cell in which the nucleic acid is to be expressed. In other words the nucleic acid sequence may be optimized for host cell codon usage.

The present invention also provides a method of producing a TCR α chain according to the present invention, comprising expressing a nucleic acid molecule encoding the TCR α chain in a cell under suitable conditions and isolating the TCR α chain.

The present invention also provides a method of producing a TCR β chain according to the present invention, comprising expressing a nucleic acid molecule encoding the TCR β chain in a cell under suitable conditions and isolating the TCR β chain.

Methods for expressing nucleic acids in a cell are well known to those skilled in the art. In particular, the nucleic acid may be expressed in a eukaryotic cell or a prokaryotic cell. The advantage of expressing the nucleic acid in a eukaryotic cell is that the encoded TCR α chain or TCR β chain will be glycosylated. However, if the nucleic acid is expressed in a prokaryotic cell, the TCR α chain or TCR β chain will not be glycosylated. When preparing crystals of the TCR α chain or TCR β chain for crystallographic analysis it is desirable that the protein is not glycosylated. Preferably the nucleic acid of the present invention is expressed in E. coli. It is also preferred that the sequence of the nucleic acid is optimized for E. coli codon usage.

The present invention also provides a method of forming the TCR according to the first aspect of the present invention, comprising simultaneously adding the TCR α chain according to the present invention and the TCR β chain according to the present invention to a refolding buffer, allowing the TCR α chain and the TCR β chain to refold and heterodimerise, and isolating the correctly folded TCR.

Preferably the redox potential of the refolding buffer allows intrachain domain assembly in addition to the formation of the covalent bond between the first and second dimerisation motifs. Preferably the buffer contains an oxidised/reduced glutathione couple. In a particularly preferred embodiment the redox ratio of the refolding buffer is 90% reduced glutathione, 10% oxidised glutathione.

It is also preferred that the refolding buffer comprises urea at a concentration which is optimal for controlled formation of the TCR domains. Preferably the urea is at a concentration of about 3 M.

It has also be found that by adding the α chain and the β chain to the refolding buffer simultaneously the efficiency of producing correctly folded TCR is increased.

The present invention also provides a multivalent TCR complex comprising a plurality of the TCRs according to the first aspect of the present invention. The multivalent TCR complex may be formed by linking the plurality of TCRs together via a suitable linker molecule. For example the TCR molecules may be biotinylated and then linked together using avidin, streptavidin, etc.

The present invention also provides the TCR according to the first aspect of the present invention for use in therapy.

The present invention, also provides the use of the TCR according to the first aspect of the present invention in the manufacture of a composition for treating an immunological disorder.

The present invention, also provides a method for treating an immunological disorder comprising administering to a patient in need of such a treatment an effective dose of the TCR according to the first aspect of the present invention. The term "immunological disorder" as used herein refers to any disorder involving the immune system, including any pathogenic infections, such as viral infections (e.g. HIV, EBV, CMV, HBV, HCV, etc); malarial plasmodium infections, autoimmune diseases such as asthma, arthritis, diabetes, graft rejection, etc.; and cancers.

The present invention, also provides a pharmaceutical composition comprising the TCR according to the first aspect of the present invention in combination with a pharmaceutically acceptable excipient.

Suitable pharmaceutically acceptable excipients are well known to those skilled in the art. Pharmaceutically acceptable excipients that may be used in the pharmaceutical composition of this invention include, but are not limited to, ion exchangers, alumina, aluminum stearate, lecithin, serum proteins, such as human serum albumin, buffer substances such as phosphates, glycine, sorbic acid, potassium sorbate, partial glyceride mixtures of saturated vegetable fatty acids, water, salts or electrolytes, such as protamine sulfate, disodium hydrogen phosphate, potassium hydrogen phosphate, sodium chloride, zinc salts, colloidal silica, magnesium trisilicate, polyvinyl pyrrolidone, cellulose-based substances, polyethylene glycol, sodium carboxymethylcellulose, polyacrylates, waxes, polyethylene- polyoxypropylene-block polymers, polyethylene glycol and wool fat.

The pharmaceutical composition of this invention may be administered orally, parenterally, by inhalation spray, topically, rectally, nasally, buccally, vaginally or via an implanted reservoir. Preferably the pharmaceutical compostion is administered orally or by injection. The pharmaceutical composition of this invention may contain any conventional non-toxic pharmaceutically-acceptable carriers, adjuvants or vehicles. The term parenteral as used herein includes subcutaneous, intracutaneous, intravenous, intramuscular, intra-articular, intrasynovial, intrasternal, intrathecal, intralesional and intracranial injection or infusion techniques. The pharmaceutical composition may be in the form of a sterile injectable preparation, for example, as a sterile injectable aqueous or oleaginous suspension. This suspension may be formulated according to techniques known in the art using suitable dispersing or wetting agents (such as, for example, Tween 80) and suspending agents. The sterile injectable preparation may also be a sterile injectable solution or suspension in a non-toxic parenterally-acceptable diluent or solvent, for example, as a solution in 1,3-butanediol. Among the acceptable vehicles and solvents that may be employed are mannitol, water, Ringer's solution and isotonic sodium chloride solution. In addition, sterile, fixed oils are conventionally employed as a solvent or suspending medium. For this purpose, any bland fixed oil may be employed including synthetic mono- or diglycerides. Fatty acids, such as oleic acid and its glyceride derivatives are useful in the preparation of injectables, as are natural pharmaceutically-acceptable oils, such as olive oil or castor oil, especially in their polyoxyethylated versions. These oil solutions or suspensions may also contain a long-chain alcohol diluent or dispersant such as Ph. Helv or a similar alcohol.

The pharmaceutical composition of this invention may be orally administered in any orally acceptable dosage form including, but not limited to, capsules, tablets, and aqueous suspensions and solutions. In the case of tablets for oral use, carriers which are commonly used include lactose and corn starch. Lubricating agents, such as magnesium stearate, are also typically added. For oral administration in a capsule form, useful diluents include lactose and dried corn starch. When aqueous suspensions are administered orally, the active ingredient is combined with emulsifying and suspending agents. If desired, certain sweetening and/or flavoring and/or coloring agents may be added.

The pharmaceutical composition of this invention may be administered by nasal aerosol or inhalation. Such compositions are prepared according to techniques well-known in the art of pharmaceutical formulation and may be prepared as solutions in saline, employing benzyl alcohol or other suitable preservatives, absorption promoters to enhance bioavailability, fluorocarbons, and/or other solubilizing or dispersing agents known in the art. The present invention, also provides the use of the TCR according to the first aspect of the present invention as an antigen to raise an antibody molecule. Methods for producing antibody molecules having affinity for a particular antigen are well known to those skilled in the art.

The present invention also provides an antibody molecule having affinity for the TCR according to the first aspect of the present invention. The antibody molecule may be a polyclonal or monoclonal antibody, or an antigen binding fragment thereof, such as Fv, Fab, F(ab')₂ fragments and single chain Fv fragments.

The TCR according to the present invention may be used in a variety of different ways. In particular, the TCR may be used to screen for agents that interact, e.g. specfically bind, to the TCR. Preferably the TCR is used to screen for agents that modify TCR function.

In order to identify agents that modify TCR function, i.e. enhance or inhibit TCR function, the TCR can be used to screen libraries of compounds in a variety of drug screening techniques. Suitable TCR functions that can be modified include recognition of MHC class I, MHC class II, CD1 or other ligands. Such functions can be measured by ligand binding or cellular signalling assays. In particular, SPR/Biacore measurements of pMHC ligand binding can be made. Cr⁵² cell killing assays using CD8⁺ CTL can also be used to determine TCR function. Candidate agents may be isolated from, for example, cells, cell-free preparations, chemical libraries, or natural product mixtures. These candidate agents may be natural or modified substrates, ligands, enzymes, receptors or structural or functional mimetics. For a suitable review of such screening techniques, see Coligan et al, Current Protocols in Immunology l(2):Chapter 5 (1991).

The TCR according to the present invention employed in such a screening technique may be free in solution or affixed to a solid support. The adherence of a candidate agent to a surface bearing the TCR can be detected by means of a label directly or indirectly associated with the candidate agent or in an assay involving competition with a labelled competitor.

Another technique for drug screening which may be used provides for high throughput screening of compounds having suitable binding affinity to the TCR (for example, see international patent application WO84/03564). In this method, large numbers of different small test compounds are synthesised on a solid substrate, which may then be reacted with the TCR. Bound TCR may then be detected using methods that are well known in the art, including using biophysical techniques such as surface plasmon resonance and spectroscopy. Other screening methods for identifying agents that modify the function of a target protein (i.e. the TCR) are well known to those skilled in the art.

In a particularly preferred embodiment of the present invention, the TCR according to the first aspect of the present invention is in the form of a crystal structure alone or in combination with MHCI or MHCII Methods for producing crystal structures of proteins are well known to those skilled in the art and are described in volume F of the International Tables of Crystallography.

The crystal structure of the TCR protein according to the first aspect of the present invention may be used to help design improved TCRs or agents that modify TCR function or bind to a TCR.

According to a second aspect of the present invention there is provided a recombinant T-cell receptor (TCR) comprising a TCRα chain and a TCRβ chain, each chain comprising a variable domain and a constant domain, wherein the chains are linked together via a disulphide bond formed between cysteine residues substituted for Pro89 ofexon 1 of TRAC*01 and Alal9 ofexon 1 of TRBCl *01 or TRBC2*01.

It has been found that the TCR having the specific cysteine residues indicated above is very efficient at refolding to form a functional TCR. In particular, it has been found that the specifically claimed TCR is approximately 11% more efficient at refolding than the best soluble TCR disclosed in International patent application WO 03/020763. The TCR according to the second aspect of the present invention has a unique disulphide bridging position which will be of great potential use for the production of a range of different soluble TCR molecules with a range of specificities and uses. By ensuring that the cysteine residues are as indicated above, successful in vitro production of TCRs which refold can be achieved.

The numbering of TCR amino acids used herein follows the IMGT system. This system is described in the "TCR factsbook" by Lefranc M.P., 2001. In accordance with the IMGT system, the α chain constant domain has the notation: TRA*01. "TR" indicates T-cell receptor genes, "A" indicates that it is the α chain gene, "C" indicates that it is a constant region, and "*01" indicates that it is allele 1. With respect to the β chain constant domain, the notation: TRBC1*01 or TRBC201 can be used in view of the fact that there are two possible constant region genes.

The TCR chains used are preferably derived from naturally occurring TCRs but may be modified provided that the chains retain substantially the same functions as the corresponding naturally occurring chains. Suitable modifications include making a small number (e.g. about 1 to 20) of amino acid substitutions additions and/or deletions. Particularly preferred modifications include removing any naturally occurring cysteine residues which form the native interchain disulphide bond. Furthermore, the TCR may comprise additional covalent bonds formed between the constant domains of the TCR chains. For example, additional suitable disulphide bonds are specified in Interntional patent application WO 03/020763.

The TCR according to the second aspect of the present invention may comprise the first and second dimerisation motifs referred to with respect to the TCR defined in the first aspect of the present invention. Furthermore, the TCR according to the second aspect of the present invention may comprise dimerisation motifs which correspond to those used in the TCR according to the first aspect of the present invention but which interact electrostatically and do not form a covalent bond. Such dimerisation motifs may be seen to correspond to the dimerisation motifs used in the TCR according to the first aspect of the present invention except that they do not comprise residues capable of forming a covalent bond, such as a disulphide bond.

As indicated above with respect to the TCR according to the first aspect of the present invention, the TCR according to the second aspect of the present invention may comprise a tag or an active molecule attached at any suitable position.

It is specifically preferred that the TCR according to the second aspect of the present invention is a soluble TCR. The present invention also provides a nucleic acid molecule encoding the TCR according to the second aspect of the present invention or the TCR α chain or TCR β chain of the TCR according to the second aspect of the present invention.

The present invention also provides the method for producing the TCR according to the second aspect of the present invention. Suitable methods are described above with respect to the TCR according to the first aspect of the present invention.

The present invention also provides the TCR according to the second aspect of the present invention for use in therapy.

The present invention also provides the use of the TCR according to the second aspect of the present invention in the manufacture of a composition for treating an immunological disorder.

The present invention also provides a method for treating an immunological disorder comprising administering to a patient in need of such a treatment an effective dose of the TCR according to the second aspect of the present invention.

The present invention also provides a pharmaceutical composition comprising the TCR according to the second aspect of the present invention in combination with a pharmaceutically acceptable excipient. Suitable pharmaceutically acceptable excipients are defined above. The TCR according to the second aspect of the present invention may be used in a variety of different ways. In particular, the TCR may be used to screen for agents that interact, e.g. specfically bind, to the TCR. Preferably the TCR is used to screen for agents that modify TCR function. The TCR according to the second aspect of the present invention can be used in the same way as the TCR according to the first aspect of the present invention.

In a particularly preferred embodiment of the present invention, the TCR according to the second aspect of the present invention is in the form of a crystal structure alone or in combination with MHCI or MHCII. Methods for producing crystal structures of proteins are well known to those skilled in the art and are described in volume F of the International Tables of Crystallography.

The crystal structure of the TCR protein according to the second aspect of the present invention may be used to help design improved TCRs or agents that modify TCR function or bind to a TCR.

The present invention is now described, by way of example only, with reference to the following Figures.

Figure 1 shows schematically the structure of the TCR of the present invention.

Figure 2 shows the SDS PAGE gel of JM22αβ TCR under reduced (R) and non-reduced (N/R) loading conditions. The respective bands are labelled α, β and αβ. Under reduced conditions the TCR migrates much slower because the αβ disulphide bridged heterodimer is a larger molecule than the individual component chains. Equimolar ratios of the α and β chains are observed under reducing conditions.

Figure 3 shows the SDS PAGE gel of LC13αβ TCR under reduced (R) and non-reduced (N/R) loading conditions. The respective bands are labelled α, β and αβ. Under reduced conditions the TCR migrates much slower because the αβ disulphide bridged heterodimer is a larger molecule than the individual component chains. Equimolar ratios of the α and β chains are observed under reducing conditions.

Figure 4 shows crystals of a variety of complexes.

Figure 5 shows a soluble TCR protein according to the second aspect of the present invention run on a 12% SDS PAGE gel under reducing conditions (R) and non-reducing conditions (NR) with Invitrogen Benchmark Protein Ladder molecular weight markers (M), demonstrating the integrity of the interchain disulphide bridge.

Figure 6 shows the elution profile with NaCl gradient of the refolded TCR according to the second aspect of the present invention from a Q ion exchange column.

Figure 7 shows the superdex gel filtration column elution profile indicating refolded protein material at the correct expected molecular weight (65kDa) for a soluble TCR heterodimer according to the second aspect of the present invention.

Figure 8 shows the elution profiles for TRAC* 01 Cys89/TRBC2*01 Cysl9 JM22α β heterodimer (line A) and TRAC*01 Cys50/TRBC2*01 Cys57 JM22 αβ heterodimer (line B).

EXAMPLES

MATERIALS AND METHODS CLONING:

1) Build two 50μl PCR reactions each containing (α and β PCR reactions): lμl lOuMdNTPs (Stratagene), 5μl 10X Pfu buffer (Stratagene), lμl cloned TCR α or β plamid (from miniprep or about 50-100ng), 38μl MilliQ water, 2μl lOpmol/μl forward primer, 2μl lOpmol/μl reverse primer and lμl Pfu polymerase (Stratagene). Cycle this in the thermal cycler (PCR machine) with the following parameters: 1) 95°C 30sec; 2) 95°C 25sec; 3) 50°C 25sec; 4) 72°C lmin30sec; cycles 2)-4) repeated 25 times; and 5) 72°C 5min.

Forward primers must contain an Ndel site (CATATG) at the predicted leader peptide cleavage site for initiation of translation, alternatively blunt-end cloning should be applied.

Reverse primers for TCR chains are as follows:

TCRAcysEND:

CCCGCTCGAGTTTACAACCACCACCGTCGTTTTCTGGGCTGGGGAAGAAGG TGTCTTCTGG

TCRBcysEND:

CCGGGATCCTCGAGCTAGTCACAACCACCACCCCTGTCCTGGTCTGCTCTAC CCCAGGCCTCGGCGCTGACG

2) Add 5μl 10X DNA electrophoresis loading buffer (80% glycerol, 0.0001% coomassie blue (Sigma), 9.9999% water, 10% 1 M Tris pH 7.2) to each PCR reaction.

3) Cast a 1.2% agasose gel containing ethidium bromide with a ~55μl well capacity comb. Immerse gel in IX TBE with a small amount of ethidium bromide into the electrophoresis equipment.

4) Carefully add the α and β PCR reactions into individual wells alongside a lkb molecular weight marker.

5) Set the electrophoresis at 50V for 20 mins.

6) Analyse the gel under low frequency UV light and excise bands corresponding to α TCR DNA, expected at 0.8Kb and β TCR DNA at 1Kb. 7) Qia spin purify the DNA from the gel fragments according to the manufacturer's instructions and elute into 30μl EB.

8) Add 3.5μl NEB buffer 2 and 0.35μl 100X BSA and mix by stirring with pipette tip.

9) Add lμl Xhol (NEB) at lOOOOU/ml and lμl Ndel (NEB) at lOOOOU/ml and mix.

10) Incubate restriction digests from step 9 at 37°C for 2 hrs.

1 1) Gel purify the DNA fragments as before, elute into 30μl water, and keep on ice.

12) Digest 600ng gel purified Pett22b+ (Novagen) vector with Ndel and Xhol in IX NEB buffer 2 with IX BSA in a 35μl volume using lμl (10U) each enzyme for 4 hrs and gel purify with IX DNA loading buffer, eluting into 40μl water.

13) Analyse via agarose gel electrophoresis the relative molar concentrations of DNA in each purified digested sample by loading 2μl of each miniprep onto a gel and analyzing the relative intensities of the bands under UV light.

14) Mix a 3:1 molar ratio of digested TCR DNA α or β to open plasmid to a final volume of 8μl in a tube containing lμl of 10X DNA ligase buffer (NEB). Add lμl (2U) T4 DNA ligase (NEB) and incubate at 16°C for about 20 hrs. Set up controls with only vector DNA at the same working concentration as non-controls.

15) Transform DH5α E. coli (BD Biosciences Clontech.) with l-2μl ligation reaction, either via electroporation or heat shock and plate onto LB-agar containing lOOμg/ml ampicillin. Incubate plate at 37°C for 16-20 hrs.

16) Pick -5 colonies from each non-control plate and place into separate 50ml flacon tubes containing 10ml sterile LB with lOOug/ml carbenicillin. Incubate with shaking at 37°C for 12-16 hrs.

17) Harvest cells by centrifugation at 3500G for 15 mins, discard the supernatant, and QiaSpin miniprep the plasmid DNA from the cells according to the manufacturers instructions, eluting into 50μl EB. 18) Digest 5μl of the plasmids with Ndel/Xhol (NEB) as in steps 8 and 9 in a 20μl final volume, electrophorese as in steps 3 and 5 and analyse digests under UV light to identify constructs containing the α or β TCR insert.

19) Sequence recombinant inserts using appropriate vector primers (T7 forward/reverse sequencing primers) to verify correct TCR sequence.

20) For blunt end cloning, design forward PCR primer to carry a terminal TG sequence followed by the next codon. Follow protocol as above, except omit Ndel digestion from step 9. Use Xbal/Xhol double digest for the recombinant plasmid identification screen stage, or an appropriate double digest based on primer design. For blunt vector production, digest Pett22b+ as before with Ndel only. Qiaspin purify DNA and elute into 43μl MQ water. Add 5μl 10X Pfu buffer, lμl lOuM dNTPs and lμl Pfu enzyme and incubate at 65°C for 10 mins, then re-column purify, eluting into 30μl water, and then proceed with steps 8 and 9 with a single Xhol digest. Likewise, only digest the PCR product with Xhol, and ligate as normal.

EXPRESSION:

1) Transform BLR (Stratagene) expression E. coli with α and β TCR constructs and plate onto LB-agar ampicillin for 16 hrs.

2) Pick a single colony and resuspend in 2ml LB

3) Place 1 ml of the colony of step 2 into 2x1 litre of autoclaved LB containing 1 OOmg/1 carbenicillin and shake at 37°C at 200 RPM until OD₆₀₀ ~0.7.

4) Add 0.5 ml 1M IPTG and shake at 37°C at 200 RPM for 4-5 hrs.

5) Harvest the cells by centrifugation at 5000G for 15 mins at 4°C.

6) Resuspend the cells in 20ml final volume 50mM NaCl, 50mM Tris pH 8.0, and add 25μl lOOmg/ml lysozyme mix and incubate at 37°C for 15 mins.

7) Top the volume up to 40 ml with Triton 2X and mix well. 8) Sonicate cells with 12mm tip probe (Sonics Vibra-Cell) by pulsing for 9.9 sec on, 9.9 sec off for a total of 16 mins at 60% power on ice.

9) Centrifuge cells at 17.5KRPM in a 30.50 rotor in Beckman Oakridge tubes for 15 mins, and discard supernatant and cell debris. Resuspend in Triton X and top up to ~40ml. Sonicate for 10 mins pulsing 9.9sec on 9.9sec off on ice at 60% power.

10) Centrifuge cells at 17.5KRPM in a 30.50 rotor in Beckman Oakridge tubes for 15 mins, and discard supernatant.

1 1) Resuspend into 40ml Triton X and repeat step 10 twice.

12) Resolubilise inclusion bodies in 15ml 8M urea, lOOmM Tris pH 8.0, 2mM DTT (resolubilisation buffer) for max 12 hrs at 4°C on a shaking table.

13) Centrifuge at 17500RPM in a 30.50 rotor in Beckman Oakridge tubes for 15 mins and aliquot supernatant into 1ml fractions and store in -80 °C.

14) Analyse 2μl of TCR α and β chains by coomassie staining a SDSPAGE gel with reducing loading buffer in a final loading volume of 20μl, run at 150V for approx 45 mins.

15) Determine the protein concentration by the Bradford assay.

REFOLDING (on lOmg scalel:

1) Prepare 1 litre refolding buffer: 3M urea, 200mM L-Arginine, 150mM Tris pH 8.0, 1.5mM reduced glutathione, and 0.15mM oxidized glutathione, 0.5mM PMSF and equilibrate at 4°C.

2) Mix 30mg of TCR α chain and 30mg β chain to a final volume of 5-7ml in TCR refolding buffer.

3) Inject TCR chain mixture into vortex of stirring refolding buffer over a 60s period and allow to stir for 1 hr. 4) Pour the refolding mixture into a large diameter lOKDa dialysis tubing (Sigma) and secure the ends with knots or clips.

5) Dialyse the refold against 10 litres of dialysis buffer (lOmM Tris pH 8.5 in deionised water) with stirring for 36hrs.

6) Replace the 10 litres of dialysis buffer and continue dialysis for 48 hrs.

7) Centrifuge the refold at 6000G for 30 mins, with slow deceleration.

8) Decant the supernatant, filter through a 0.22um filter and load onto a 5ml Hitrap Q HP column (Amersham Pharmacia Biotech) at 7ml/min.

PURIFICATION (on lOmg scale):

1) On an FPLC machine, equilibrate the column with buffer A (lOmM NaCl, lOmM Tris pH8.5)

2) Elute the refolded TCR with a 0-33% gradient of buffer B (1M NaCl) over 90ml, followed by a column cleanup gradient of 33%>-100%> B over 30ml ail at a flow rate of 5ml/min, collecting 3ml fractions throughout.

3) Re-equilibrate column with buffer A at a flow rate of 5ml/min.

4) SDS PAGE analyse the elution profile, and pool the peak corresponding to the refolded TCR.

5) Apply TCR pool to 2ml Ni²⁺ immobilised metal Chelating Sepharose Fast Flow Beads (Amersham Pharmacia Biotech) packed into a Poly-Prep chromatography column (BioRad) and equilibrated with Ni²⁺ wash buffer (lOmM immidazole, 50mM NaCl, lOOmM Tris pH 7.5).

6) Wash column with 10ml Ni²⁺ wash buffer.

7) Step elute protein with 8ml Ni²⁺ elution buffer (500mM immidazole, 50mM NaCl, lOOmM Tris pH 7.5), collecting 1ml fractions in eppendorph tubes. 8) SDS PAGE analyse fraction contents and pool TCR-containing fractions.

9) Concentrate the TCR using the lOOOODa molecular weight cutoff 6ml concentrator (Sartorius) by spinning at 3000G to 300μl volume and load onto a SD200 column (Amersham Pharmacia Biotech) pre-equilibrated with 150mM NaCl, 50mM Tris pH7.5.

10) Elute the protein (65kDa) into 1ml fractions at a 0.5ml flowrate with 150mM NaCl, 50mM Tris pH7.5.

11) SDS PAGE analyse the fractions, pool those containing the TCR, and measure the protein concentration and adjust the concentration to 2.2mg/ml.

12) Aliquot the protein pool into 1ml fractions in eppendorphs.

13) Prepare 1.2U carboxypeptidase immobilized beads from 80μl slurry (Sigma cat. C1261) by washing twice with lOOmM Tris pH7.5, 150mM NaCl in 1.5ml eppendorph tubes.

14) Add the washed 1.2U carboxypeptidase beads to protein solution, close and parafilm the lid of the tube and shake 800RPM at 20°C for 16hrs to remove the his₆ tag.

15) Centrifuge out beads and concentrate the TCR using the lOOOODa molecular weight cutoff 6ml concentrators (Sartorius) and centrifugation at 3000G to 300μl volume and load onto a SD200 column pre-equilibrated with lOmM NaCl, lOmM Tris pH7.0.

16) Elute the protein (65kDa) into 1ml fractions at a 0.5ml flowrate with lOmM NaCl, 1 OmM Tris pH7.0 buffer.

17) Check protein purity and integrity of interchain disulfide bond by running reduced and non-reduced loaded samples on an SDS PAGE gel, followed by coomassie staining.

Buffers Required:

11 Buffer A - 1 OmM NaCl, 1 OmM Tris pH8.5 500ml Buffer B - 1M NaCl

500ml Ni²⁺ wash Buffer - lOmM immidazole, 50mM NaCl, lOOmM Tris pH 7.5

500ml Ni²⁺ elution Buffer - 500mM immidazole, 50mM NaCl, lOOmM Tris pH 7.5

500ml final Buffer - 1 OmM NaCl, 1 OmM Tris pH7.0

100ml Inclusion body resolubilisation buffer - 15ml 8M urea, lOOmM Tris pH 8.0, 2mM DTT

EXAMPLE 1

A diagram of the engineered refolded dimer is shown in Figure 1. The invention relies on the carboxy terminal flanking sequences on both α and β chains. Each flanking sequence was specifically designed to the structural surface and charge distribution at the top of the human TCR constant domains based on the A6 human TCR-pMHC complex (PDB ID: 1QSF) (Ding et al, Immunity, I I, 45, 1999), and avoiding any similar sequence to the regions around the native internal immunoglobulin domain cysteines. The flanking sequences were designed to provide sufficient length for the cysteines to pair and form a disulfide bridge between polypeptide chains judged from the crystal structure 1QSF. The refolding buffer was designed to allow rapid intrachain domain assembly, to promote the natural tendency for αβ pairing followed by the formation of the carboxy-terminal disulfide bridge through glutathione oxidation. Aspartic acid was selected as the terminal β-chain residue to maximise the negative charge on the tip of the chain, with the aim of promoting scamiing interactions with the hexahistidine tag and the adjacent lysine side-chain and induce carboxy-terminal chain interaction. The DNA sequences were also optimized for E. coli codon usage.

The flanking protein sequences are as follows (native TCRseq in brackets):

α [PEDTFFPSP]ENDGGGCKHHHHHH-

β [VSAEAWGRAD]QDRGGGCD- The ability to express, refold and purify stable TCRs according to the present invention has been demonstrated for two different TCRs: LC13αβ and JM22αβ. For both TCRs, the α and β protein chains expressed similarly in both cases: 105mg of resolubilised α chain per litre LB and 30mg resolubilised β chain per litre LB. The counterpart TCR chains were refolded as described, and subjected to the sequence of purification steps.

The refolded protein was eluted in response to the increasing salt gradient. Fractions containing the refolded protein were loaded directly onto a Ni²⁺ binding column. The α his₆ tagged heterodimer was then step-eluted and analysed by SDS PAGE with reducing and non-reducing loading buffers. The protein is already very pure at this stage, and the heterodimeric quality is good, as the his-tagged alpha chain has demonstrably carried the β chain in a stoichiometric ratio. After gel-filtration and the 16hr treatment with carboxypeptidase A, the sample was passaged through a Ni²⁺ column. All the protein was washed out and no protein could be detected in the elution demonstrating that the his-tag had been entirely removed. Following the final gel filtration, the TCR was mixed with pre-refolded and purified pMHC for crystallization trials, and the complex was analysed on SDS PAGE.

Figure 2 shows the SDS PAGE analysis of the LC13αβ and JM22αβ TCR complexes in reduced and non-reduced loading buffers and the analysis of the individual components. The spread of reducing agent in the early stages of electrophoresis caused the sides of the protein bands in the non-reduced lanes to become reduced and thence migrate as reduced. The disulfide-bridged αβ dimers can be clearly seen in non-reduced conditions as they migrate much slower in the gel due to their higher molecular weight.

Figure 3 is the gel filtration elution profile for LC13αβ dimer and analysis of the protein purity and integrity of the interchain disulfide bridge by reducing and non-reducing SDS PAGE. All samples were boiled prior to loading, and it was observed that the intermolecular disulfide bridge dissociation during this step was proportional to the length of time of boiling. In the lower gel, the LC13αβ TCR was incubated for two weeks at room temperature to demonstrate the maintained high purity of the TCR and the robust nature of the intermolecular disulfide bridge. No evidence during this period indicated any protein aggregation as with scFv TCRs and JM22 could be reproducibly crystallized after several weeks of incubation at room temperature.

The protein was sufficiently pure and robust to crystallize as large X-ray diffraction quality crystals. The JM22αβ TCR dimer crystals (P6ι space group) were up to 1.5mm in length and diffracted to 1.98A resolution at the ESRF. The JM22αβ TCR dimers also crystallized as a series of crystal forms in complex with the cognate pMHC (space groups P2ι and C2). These crystals were of X-ray diffraction quality and surpassed the highest published pMHC-TCR complex resolution (2.4A) with X-ray diffraction to 1.4A resolution (see Figure 4).

Two recombinant TCRs according to the present invention have been obtained and the stability of the TCRs has been demonstrated.

The problem of TCR production and isolation as a stable heterodimer has been overcome. The speed of production (2 weeks from clone to pure protein) is a great benefit combined with the high yields of protein rapidly obtained from E. coli and the efficient yields on refolding and purification. TCR production has been a major problem in a number of laboratories and this technology could be used to expand the breadth of research stemming from the requirement of TCRs in primary research, diagnostics and in therapy.

The present invention can be used to generate any TCR. For example, human TCRs against HIV and cancer for MHC class I receptors, and TCRs recognizing an MHC class II molecule for autoimmune diseases. The TCRs of the present invention are also applicable to the emergent field of TCR recognition of MHC like molecules, such as CD1.

As indicated above, the TCRs of the present invention can be utilized in a number of ways, including:

1) Raising antibodies (human/murine/phage display) to the VαVβ domains for treatment of autoimmune disease; 2) Cancer therapy via conjugation of TCR monomers/multimers with radioisotopes/prodrug converting enzyme (similar to ADEPT) or otherwise treating disease via specific antigen presenting cell targeting;

3) Technicium-99 coupling (via hisβ tag coupling) for diagnostic MRI imaging of locality of antigenic tissue;

4) Engineering in silico a TCR of hightened affinity or design for inhibiting pMHC-TCR complex formation from crystal structure determination; and

5) Screening of drugs for autoimmune disease or graft rejection (surface plasmon resonnance), NMR, etc.

EXAMPLE 2

The constant domain interface of a TCR has since been systematically and experimentally searched for unique cysteine pairing positions, one on either α or β constant domains. The different positions were selected on the basis of special proximity and suitability of Cα-Cβ bond orientation, and their capacity of forming an inter-chain disulphide bridge assessed using the same refolding procedure as described above. Protein was expressed from E.coli as inclusion bodies and the engineered TCRs were variations of the JM22 α and β TCR chain constructs utilised previously. Each candidate position for introduction of a disulphide bridge in the interface was identified by careful examination of the crystal structures of the JM22 in isolation (1.9A resolution) and in complex with HLA-A2-flu (1.4A resolution) in addition to comparisons with other published TCR structures (PDB accession codes 1QSF and 1BD2).

A number of positions between the TCR constant domains that on cysteine engineering allow disulphide bridge formation to occur have been identified (see International patent application WO 03/20763).

All different interchain disulphide bridge positions in the constant regions result in different levels of yield of refolded protein, ranging from 0% to 55% efficiency of conversion of denatured protein chains into refolded soluble heterodimers after Q-column anion exchange, Ni++ affinity purification and gel filtration. This indicated differential efficiency of refolding engineered α and β chains depending on the position of the engineered disulphide bridge in the constant domains.

A very efficient interchain disulphide bridge position (TRAC*01 Cys89/TRBC2*01 Cysl9) has been identified and is approximately l l%o more efficient at refolding the JM22 TCR than the next best disulphide bridging position (TRAC*01 Cys50/TRBC2*01 Cys57 JM22). The new unique disulphide bridging position here identified could be of great potential use for the production of a range of different soluble TCR molecules with a range of specificities and uses. This technology may allow successful in vitro production of TCRs that have failed to refold with other engineered constant domain forms.

The unique pair of cysteine positions identified are mutations as follows:

Native TCR α constant domain sequence:

....PETFFPS... (Amino acids 84-90 of exon 1 of the TRAC*01 gene)

Engineered α constant domain sequence: ...PETFFCSKHHHHHH[stop]

Native TCR β constant domain sequence:

....PSEAEISHT... (Amino acids 16-24 of exon 1 TRBC1*01 & TRBC2*01 genes)

Engineered β constant domain sequence:

....PSECEISHT...

Constructs were generated with the Pett22b+ expression vector by Ndel/Xhol ligation and protein expressed in the BLR strain of E.coli (Novagen).

The mutations can be classified as residues Proline 200 from the α constant domain and Alanine 185 from the β constant domain based on the residues from the published crystal structure of the JM22-A2flu complex (PDB accession code 10GA) (see attached sequences). A hexahistidine tag has been engineered onto the C terminus of the α chain to enable additional purification, as with the TCR purification protocol described above, however the β chain C-terminal stop codon has been engineered at the end of the β constant domain as in the following sequence: ...EWGRAD[stop]

The following primers have been utilised to introduce the E. coli optimised cysteine codons:

G3AR gggccgctcgagtttgctgcagaagaaggtgtcttctgg

G3BF gccatcagagtgcgagatctcccacaccc G3BR gggagatctcgcactctgatggctcaaac Bend gggccgctcgagctagtctgctctaccccaggcctcgg

G3 AR was used to amplify the TCR α chain with the T7 forward sequencing primer (Novagen) and subsequently cloned Ndel/Xhol into Pett22b+.

Primer G3BR was used to amplify the TCR β chain with the T7 forward sequencing primer (Novagen), and the G3BF primer used to amplify the TCR β chain with the primer Bend. These two PCR products were gel purified (Qiaquick gel purification, according to manufacturer's instructions), mixed in stoichiometric ratio and subjected to another round of amplification using the T7 forward sequencing primer (Novagen) and the Bend primer. The fusion product was gel purified (Qiaquick gel purification, according to manufacturer's instructions), and cloned Ndel/Xhol into Pett22b+.

Constructs were sequenced to verify cloning accuruacy prior to E.coli expression.

This interchain disulphide bridge position pair is to be introduced into the constant domain of the TCR constructs described in Example 1 above. It is speculated that the dimerisation motifs, consisting of residues that induce interchain interaction would enhance the efficiency of heterodimer refolding and interchain disulphide bridge formation. In this case the TCR would be cross-linked in two positions: one disulphide bridge in the constant domains and another disulphide bridge in the dimerisation motifs.

Altered forms of the dimerisation motifs with one or both of the original dimerisation motifs cysteine residues mutated to serine, would allow interchain affinity via the electrostatic interaction of the dimerisation motifs and thus promote disulphide bridge formation at the engineered constant domain cysteine pair highlighted above.

These engineered forms of the TCR nucleotide sequences can be utilised in mammalian or insect cell tissue culture for expression of soluble TCR molecules.

The two JM22 engineered TCR sequences are shown below as open reading frames (bold, italic and underlined sequences correspond to the new engineered cysteine mutant codons):

Cys-mutant JM22alpha nucleotide sequence:

ATGCAACTACTAGAACAAAGTCCTCAGTTTCTAAGCATCCAAGAGGGAGAAAATCTCACTGTGTACTGCA ACTCCTCAAGTGTTTTTTCCAGCTTACAATGGTACAGACAGGAGCCTGGGGAAGGTCCTGTCCTCCTGGT GACAGTAGTTACGGGTGGAGAAGTGAAGAAGCTGAAGAGACTAACCTTTCAGTTTGGTGATGCAAGAAAG GACAGTTCTCTCCACATCACTGCGGCCCAGCCTGGTGATACAGGCCTCTACCTCTGTGCAGGAGCGGGAA GCCAAGGAAATCTCATCTTTGGAAAAGGCACTAAACTCTCTGTTAAACCAAATATCCAGAACCCTGACCC TGCCGTGTACCAGCTGAGAGACTCTAAATCCAGTGACAAGTCTGTCTGCCTATTCACCGATTTTGATTCT CAAACAAATGTGTCACAAAGTAAGGATTCTGATGTGTATATCACAGACAAAACTGTGCTAGACATGAGGT CTATGGACTTCAAGAGCAACAGTGCTGTGGCCTGGAGCAACAAATCTGACTTTGCATGTGCAAACGCCTT CAACAACAGCATTATTCCAGAAGACACCTTCTTCTGCAGCAAACTCGAGCACCACCACCACCACCACTGA

Cys-mutant JM22beta nucleotide sequence:

ATGGTGGATGGTGGAATCACTCAGTCCCCAAAGTACCTGTTCAGAAAGGAAGGACAGAATGTGACCCTGA GTTGTGAACAGAATTTGAACCACGATGCCATGTACTGGTACCGACAGGACCCAGGGCAAGGGCTGAGATT GATCTACTACTCACAGATAGTAAATGACTTTCAGAAAGGAGATATAGCTGAAGGGTACAGCGTCTCTCGG GAGAAGAAGGAATCCTTTCCTCTCACTGTGACATCGGCCCAAAAGAACCCGACAGCTTTCTATCTCTGTG CCAGTAGTTCGAGGAGCTCCTACGAGCAGTACTTCGGGCCGGGCACCAGGCTCACGGTCACAGAGGACCT GAAAAACGTGTTCCCACCCGAGGTCGCTGTGTTTGAGCCATCAGAG

ΓGCGAGATCTCCCACACCCAAAAGGCCACACTGGTGTGCCTGGCCACAGGCTTCTACCCCGACCACGTGG AGCTGAGCTGGTGGGTGAATGGGAAGGAGGTGCACAGTGGGGTCAGCACAGACCCGCAGCCCCTCAAGGA GCAGCCCGCCCTCAATGACTCCAGATACAGCCTGAGCAGCCGCCTGAGGGTCTCGGCCACCTTCTGGCAG AACCCCCGCAACCACTTCCGCTGTCAAGTCCAGTTCTACGGGCTCTCGGAGAATGACGAGTGGACCCAGG ATAGGGCCAAACCTGTCACCCAGATCGTCAGCGCCGAGGCCTGGGGTAGAGCAGACTAG

Protein sequences are as shown below (bold, italic and underlined sequences correspond to the new engineered cysteine mutant amino acids):

Cys-mutant JM22alpha protein sequence:

MQLLEQSPQFLSIQEGENLTVYCNSSSVFSSLQ YRQEPGEGPVLLVTVVTGGEVKKLKR LTFQFGDARKDSSLHITAAQPGDTGLYLCAGAGSQGNLIFGKGTKLSVKPNIQNPDPAVY QLRDSKSSDKSVCLFTDFDSQTNVSQSKDSDVYITDKTVLDMRSMDFKSNSAVA SNKSD FACANAFNNSIIPEDTFFCSKLEHHHHHH-

Cys-mutant JM22beta protein sequence:

1MVDGGITQSPKYLFRKEGQNVTLSCEQNLNHDAMYWYRQDPGQGLRLIYYSQIVNDFQKG DIAEGYSVSREKKESFPLTVTSAQKNPTAFYLCASSSRSSYEQYFGPGTRLTVTEDLKNV FPPEVAVFEPSECEISHTQKATLVCLATGFYPDHVELSW VNGKEVHSGVSTDPQPLKEQ PALNDSRYSLSSRLRVSATFWQNPRNHFRCQVQFYGLSENDE TQDRAKPVTQIVSAEAW GRAD-

Definition of TCR α and β constant domains:

The numbering of TCR amino acids used to define exons 1 of TRAC*01, TRBC1*01 and TRBC2*01 genes used herein follows the IMTG system described in The T Cell Receptor Factsbook, 2001, LeFranc & LeFranc, Academic Press.

Amino acids of exon 1 of the TRAC*01 gene (residues 1 - 91):

NIQNPDPAVYQLRDSKSSDKSVCLFTDFDSQTNVSQSKDSDVYITDKTVLDMRSMDFK SNSAVAWSNKSDFACANAFNNSIIPEDTFFPSK

Amino acids of exon 1 TRBC1*01 gene (residues 1 - 130): EDLNKVFPPEVAVFEPSEAEISHTQKATLVCLATGFFPDHVELSWWVNGKEVHSGVST DPQPLKEQPALNDSRYSLSSRLRVSATFWQNPRNHFRCQVQFYGLSENDE TQDRAKP VTQIVSAEAWGRAD

Amino acids of exon 1 TRBC2*01 gene (residues 1 - 130):

EDLKNVFPPEVAVFEPSEAEISHTQKATLVCLATGFYPDHVELSWWVNGKEVHSGVST DPQPLKEQPALNDSRYSLSSRLRVSATF QNPRNHFRCQVQFYGLSENDEWTQDRAKP VTQIVSAEAWGRAD

The novel interchain disulphide bridge position is generated from mutations conferring changes at TRAC*01 Pro89 to Cys89 and TRBC1*01/TRBC2*01 Alal9 to Cysl9.

Alternatively, the mutations can be classified as residues Proline 200 from the α constant domain and Alanine 185 from the β constant domain based on the residues from the published crystal structure of the JM22- A2flu complex (PDB accession code 1OGA).

Figure 5 illustrates the soluble TCR protein run on a 12%> SDS PAGE gel under reducing conditions (R) and non-reducing conditions (NR) with Invitrogen Benchmark Protein Ladder molecular weight markers (M), demonstrating the integrity of the interchain disulphide bridge.

Figure 6 illustrates the elution profile with NaCl gradient of the refolded TCR (blue) from the Q ion exchange column.

Figure 7 illustrates the superdex gel filtration column elution profile indicating refolded protein material at the correct expected molecular weight (65kDa) for a soluble TCR heterodimer.

Figure 8 shows a superimposition of the elution profiles for the TRAC*01 Cys89/ TRBC2*01 Cysl9 JM22αβ heterodimer and the TRAC*01 Cys50/ TRBC2*01 Cys57 JM22αβ heterodimer disclosed in WO 03/20763. The calculated surface areas under the main peaks (peaks A and B respectively) were 3612.7mAU*ml and 3245.2mAU*ml resectively. Comparison of the yields can be made from comparison of the surface areas under the main peak (which would be proportional to the quantity of soluble TCR), given equivalent levels of purity at this stage (confirmed by SDSPAGE). The TRAC*01 Cys89/ TRBC2*01 Cysl9 JM22αβ refolding capacity was 11.3% higher than the TRAC*01 Cys50/ TRBC2*01 Cys57 JM22αβ heterodimer refolding capacity based on this analysis.

Refolding was performed with 15mg (carefully measured and evaluated) of both denatured chains (30mg total), mixed, and injected into 500ml aliquots of the refolding buffer dispensed from a larger stock volume.

Additional information:

Native JM22 TCR α sequence:

QLLEQSPQFLSIQEGENLTVYCNSSSVFSSLQ YRQEPGEGPVLLVTVVTGGEVKKLKRLTFQFGDARKD SSLHITAAQPGDTGLYLCAGAGSQGNLIFGKGTKLSVKPNIQNPDPAVYQLRDSKSSDKSVCLFTDFDSQ TNVSQSKDSDVYITDKTVLDMRSMDFKSNSAVAWSNKSDFACANAFNNSIIPEDTFFPSK

Native JM22 TCR β sequence:

VDGGITQSPKYLFRKEGQNVTLSCEQNLNHDAMYWYRQDPGQGLRLIYYSQIVNDFQKGDIAEGYSVSRE KKESFPLTVTSAQKNPTAFYLCASSSRSSYEQYFGPGTRLTVTEDLKNVFPPEVAVFEPSEAEISHTQKA TLVCLATGFYPDHVELS VNGKEVHSGVSTDPQPLKEQPALNDSRYSLSSRLRVSATFWQNPRNHFRCQ VQFYGLSENDEWTQDRAKPVTQIVSAEAWGRAD

All documents cited above are incorporated herein by reference.

Claims

1. A recombinant T-cell receptor (TCR) comprising a TCR α chain and a TCR β chain, wherein the TCR α chain comprises a Vα domain, a Cα domain and a first dimerisation motif attached to the C-terminus of the Cα domain, and the TCR β chain comprises a Vβ domain, a Cβ domain and a second dimerisation motif attached to the C-terminus of the Cβ domain, wherein the first and second dimerisation motifs easily interact to form a covalent bond between an amino acid in the first dimerisation motif and an amino acid in the second dimerisation motif linking the TCR α chain and the TCR β chain together.

2. The TCR according to claim 1, wherein the dimerisation motifs interact electrostatically.

3. The TCR according to claim 2, wherein the first and second dimerisation motifs comprise oppositely charged amino acids at corresponding positions.

4. The TCR according to claim 2 or claim 3, wherein the C-terminal amino acid of the first dimerisation motif and the C-terminal amino acid of the second dimerisation motif are oppositely charged amino acids capable of interacting electrostatically.

5. The TCR according to any one of the preceding claims, wherein the amino acid forming the covalent bond in the first dimerisation motif is adjacent a charged amino acid and that the amino acid forming the covalent bond in the second dimerisation motif is adjacent a charged amino acid, wherein the charged amino acids interact electrostatically.

6. The TCR according to any one of the preceding claims, wherein the dimerisation motifs are flexible.

7. The TCR according to claim 6, wherein the dimerisation motifs comprise amino acids that allow flexibility.

8. The TCR according to claim 7, wherein the amino acids that allow flexibility are selected from glycine, serine and alanine.

9. The TCR according to any one of the preceding claims, wherein the first dimerisation motif is constructed so that the amino acid that forms the interchain covalent bond is separated from the Cα domain by at least 5 amino acids.

10. The TCR according to claim 9, wherein the covalent bond forming amino acid is separated from the Cα domain by between 6 and 20 amino acids.

11. The TCR according to any one of the preceding claims, wherein the second dimerisation motif is constructed so that the amino acid that forms the interchain covalent bond is separated from the Cβ domain by at least 3 amino acids.

12. The TCR according to claim 11, wherein the covalent bond forming amino acid is separated from the Cβ domain by between 4 and 20 amino acids.

13. The TCR according to any one of the preceding claims, which additionally comprises a tag.

14. The TCR according to any one of the preceding claims, wherein the first dimerisation motif has the formula: -(Y)m-(Zl)p-(Y)m-(F)q-(Y)m-B-(Z2)p (I) wherein,

Zl is a charged amino acid; p is independently 1 to 40; F is an amino acid that increases flexibility of the dimerisation motif region; q is 1 to 10;

B is an amino acid capable of forming a disulfide bond; and Z2 is a charged amino acid having the opposite charge to Zl, wherein the dimerisation motif is at least 6 amino acids in length.

15. The TCR according to any one of the preceding claims, wherein the second dimerisation motif has the formula:

-(Y)m-(Z2)p-(Y)m-(F)q-(Y)m-B-(Zl)p (II) wherein,

B is an amino acid capable of forming a disulfide bond; and Z2 is a charged amino acid having the opposite charge to Zl , wherein the dimerisation motif is at least 4 amino acids in length.

16. The TCR according to any one of claims 14 to 15, wherein m is independently l to 3.

17. The TCR according to any one of claims 14 to 16, wherein p is 1.

18. The TCR according to any one of claims 14 to 17, wherein q is 3.

19. The TCR according to any one of the preceding claims, wherein the first dimerisation motif has the sequence: -PENDGGGCK.

20. The TCR according to any one of the preceding claims, wherein the second dimerisation motif has the sequence: -ADQDRGGGCD.

21. The TCR according to any one of the preceding claims, which additionally comprises an active molecule.

22. The recombinant T-cell receptor (TCR) α chain of the TCR according to any one of the preceding claims.

23. The recombinant T-cell receptor (TCR) β chain of the TCR according to any one of claims 1 to 21.

24. A nucleic acid molecule encoding the TCR according to any one of claims 1 to 21.

25. A nucleic acid molecule encoding the T-cell receptor (TCR) α chain according to claim 22.

26. A nucleic acid molecule encoding the T-cell receptor (TCR) β chain according to claim 23.

27. A method of producing a TCR α chain according to claim 22 comprising expressing the nucleic acid molecule according to claim 25 in a cell under suitable conditions and isolating the TCR α chain.

28. A method of producing a TCR β chain according to claim 23 comprising expressing the nucleic acid molecule according to claim 26 in a cell under suitable conditions and isolating the TCR β chain.

29. A method of forming the TCR according to any one of claims 1 to 21, comprising simultaneously adding the TCR α chain and the TCR β chain to refolding buffer, allowing the TCR α chain and the TCR β chain to refold and pair, and isolating the correctly folded TCR.

30. The method according to claim 29, wherein the redox potential of the refolding buffer allows intrachain domain assembly in addition to the formation of the covalent bond between the dimerisation motifs.

31. The TCR according to any one of claims 1 to 21 for use in therapy.

32. Use of the TCR according to any one of claims 1 to 21 in the manufacture of a composition for treating immunological disorder.

33. Use of the TCR according to any one of claims 1 to 21 as an antigen to raise antibody molecules.

34. An antibody molecule having affinity for the TCR according to any one of claims 1 to 21.

35. Use of the TCR according to any one of claims 1 to 21 to screen for agents that interact with the TCR.

36. Use of the TCR according to any one of claims 1 to 21 to screen for agents that modify TCR function.

37. The TCR according to any one of claims 1 to 21 as a crystal structure.

38. Use of the crystal structure according to claim 37 to design improved TCRs or agents that modify TCR function or agents that bind to a TCR.

39. A recombinant T-cell receptor (TCR) comprising a TCRα chain and a TCRβ chain, each chain comprising a variable domain and a constant domain, wherein the chains are linked together via a disulphide bond formed between cysteine residues substituted for Pro89 of exon 1 of TRAC*01 and Alal9 of exon 1 of TRBC1 *01 or TRBC2*01.

40. The TCR according to claim 39, which comprises one or more additional covalent bonds formed between the constant domains of the TCR chains.

41. The TCR according to claim 39 or claim 40, wherein a dimerisation motif is attached to the constant domain of each chain, and wherein the dimerisation motifs interact electrostatically.

42. The TCR according to claim 41, wherein the dimerisation motifs are flexible.

43. The TCR according to any one of the preceding claims, which additionally comprises a tag.

44. The TCR according to any one of claims 38 to 43, which additionally comprises an active molecule.

45. The recombinant T-cell receptor (TCR) α chain of the TCR according to any one of claims 38 to 44.

46. The recombinant T-cell receptor (TCR) β chain of the TCR according to any one of claims 38 to 44.

47. A nucleic acid molecule encoding the TCR according to any one of claims 38 to 44.

48. A nucleic acid molecule encoding the T-cell receptor (TCR) α chain according to claim 45.

49. A nucleic acid molecule encoding the T-cell receptor (TCR) β chain according to claim 46.

50. A method of producing a TCR α chain according to claim 45 comprising expressing the nucleic acid molecule according to claim 48 in a cell under suitable conditions and isolating the TCR α chain.

51. A method of producing a TCR β chain according to claim 46 comprising expressing the nucleic acid molecule according to claim 49 in a cell under suitable conditions and isolating the TCR β chain.

52. A method of forming the TCR according to any one of claims 38 to 44, comprising simultaneously adding the TCR α chain and the TCR β chain to refolding buffer, allowing the TCR a chain and the TCR β chain to refold and pair, and isolating the correctly folded TCR.

53. The method according to claim 52, wherein the redox potential of the refolding buffer allows intrachain domain assembly in addition to the formation of the covalent bond between the constant domains.

54. The TCR according to any one of claims 38 to 44 for use in therapy.

55. Use of the TCR according to any one of claims 38 to 44 in the manufacture of a composition for treating immunological disorder.

56. Use of the TCR according to any one of claims 38 to 44 as an antigen to raise antibody molecules.

57. Use of the TCR according to any one of claims 38 to 44 to screen for agents that interact with the TCR.

58. Use of the TCR according to any one of claims 38 to 44 to screen for agents that modify TCR function.

59. The TCR according to any one of claims 38 to 44 as a crystal structure.

0. Use of the crystal structure according to claim 59 to design improved TCRs or agents that modify TCR function or agents that bind to a TCR.