WO2013041865A1

WO2013041865A1 - T cell receptors

Info

Publication number: WO2013041865A1
Application number: PCT/GB2012/052322
Authority: WO
Inventors: Bent Karsten Jakobsen; Nathaniel Ross Liddy; Annelise Brigitte Vuidepot; Peter Eamon Molloy
Original assignee: Immunocore Limited
Priority date: 2011-09-22
Filing date: 2012-09-20
Publication date: 2013-03-28

Abstract

The present invention relates to T cell receptors (TCRs) that the property of binding to EVDPIGHLY (SEQ ID No: 1) HLA-A1 complex and comprise a TCR alpha chain variable domain and a TCR beta chain variable domain. The EVDPIGHLY peptide is derived from the MAGE-3 protein which is expressed in many tumour types, including melanomas, and other solid tumours such as Head and Neck Squamous Cell, lung, bladder, gastric, prostate, colorectal and esophageal carcimomas. Also provided are nucleic acids encoding such TCRs, cells presenting such TCRs and pharmaceutical compositions comprising such TCRs.

Description

T Cell Receptors

The present invention relates to T cell receptors (TCRs) which bind the EVDPIGHLY peptide (derived from the MAGE-3 protein) presented as a peptide- HLA-A1 complex, the TCRs being mutated relative to the native MAGE-3 TCR alpha and/or beta variable domains. Certain preferred TCRs also bind the EVDPIGHVY peptide (derived from the MAGE-6 protein, and the the EVDPIRHYY peptide (derived from the MAGE-B18 protein) presented as a peptide-HLA-A1 complexes. The TCRs of the invention demonstrate excellent specificity profiles for those MAGE epitopes and have improved binding affinities for the complex, resulting in an enhanced ability to recognize the complex compared to the reference MAGE-3 TCR described below.

Background to the Invention

The EVDPIGHLY (SEQ ID No: 1 ) peptide corresponds to amino acid residue numbers 168-176 of the known MAGE-3 protein. The MAGE-3 protein is expressed in many tumour types, including melanomas, and other solid tumours such as Head and Neck Squamous Cell, lung, bladder, gastric, prostate, colorectal and esophageal carcimomas. The MAGE-3 peptide EVDPIGHLY is the best characterised MAGE-3 epitope. It is recognised by both HLA-A1 and HLA-B35 restricted T cells. It is able to elicit cytotoxic activity against peptide-pulsed, HLA-A1 positive target cells, and

MAGE-3-expressing HLA-A1 positive melanoma cell lines. The peptide, used as a vaccine, has been shown to induce tumour regression and elicit CTL responses in some of those patients. The MAGE-3 protein is a member of a family of MAGE proteins expressed in many tumour types. Two other members of the family are MAGE-6 and MAGE-B18. The MAGE-6 peptide epitope is EVDPIGHVY (SEQ ID No: 21 ) and the MAGE-B18 epitope is EVDPIRHYY (SEQ ID No: 22) and those epitopes are also recognised by HLA-A1 restricted T cells. The MAGE-6 and B18 epitopes have very similar sequences to that of the MAGE-3 epitope EVDPIGHLY.

Therefore, the EVDPIGHLY HLA-A1 complex, and the EVDPIGHVY HLA-A1 and EVDPIRHYY HLA-A1 complexes, provide cancer markers that the TCRs of the invention can target. TCRs of the invention may be transformed into T-cells, rendering them capable of destroying tumour cells presenting that HLA complex, for administration to a patient in the treatment process known as adoptive therapy. For this purpose, it would be desirable if the TCRs had a higher affinity and/or a slower off-rate for the peptide-HLA complex than native TCRs specific for that complex. Dramatic increases in affinity have been associated with a loss of antigen specificity in TCR gene-modified CD8 T cells, which could result in the nonspecific activation of these TCR-transfected CD8 cells, so TCRs having a somewhat higher affinity and/or a slower off-rate for the peptide-HLA complex than native TCRs specific for that complex, but not a dramatically higher affinity and/or dramatically slower off-rate for the peptide-HLA complex than native TCRs, would be preferred for adoptive therapy (see Zhao et ai, (2007) J Immunol. 179: 5845-54; Robbins et ai, (2008) J Immunol. 180: 61 16-31 ; and WO2008/038002). Some TCRs of the invention may be useful for the purpose of delivering cytotoxic or immune effector agents to the cancer cells. For this use it is desirable that the TCRs have a considerably higher affinity and/or a slower off-rate for the peptide-HLA complex than native TCRs specific for that complex. For example, the binding affinity may be at least double that of the reference MAGE-3 TCR described below.

TCRs are described using the International Immunogenetics (IMGT) TCR

nomenclature, and links to the IMGT public database of TCR sequences. Native alpha-beta heterodimeric TCRs have an alpha chain and a beta chain. Broadly, each chain comprises variable, joining and constant regions, and the beta chain also usually contains a short diversity region between the variable and joining regions, but this diversity region is often considered as part of the joining region. Each variable region comprises three CDRs (Complementarity Determining Regions) embedded in a framework sequence, one being the hypervariable region named CDR3. There are several types of alpha chain variable (Voc) regions and several types of beta chain variable (νβ) regions distinguished by their framework, CDR1 and CDR2 sequences, and by a partly defined CDR3 sequence. The Voc types are referred to in IMGT nomenclature by a unique TRAV number. Thus "TRAV21 " defines a TCR Voc region having unique framework and CDR1 and CDR2 sequences, and a CDR3 sequence which is partly defined by an amino acid sequence which is preserved from TCR to TCR but which also includes an amino acid sequence which varies from TCR to TCR. In the same way, "TRBV5-1 " defines a TCR νβ region having unique framework and CDR1 and CDR2 sequences, but with only a partly defined CDR3 sequence. The joining regions of the TCR are similarly defined by the unique IMGT TRAJ and TRBJ nomenclature, and the constant regions by the IMGT TRAC and TRBC nomenclature. The beta chain diversity region is referred to in IMGT nomenclature by the

abbreviation TRBD, and, as mentioned, the concatenated TRBD/TRBJ regions are often considered together as the joining region.

The a and β chains of αβ TCR's are generally regarded as each having two

"domains", namely variable and constant domains. The variable domain consists of a concatenation of variable region and joining region. In the present specification and claims, the term "TCR alpha variable domain" therefore refers to the

concatenation of TRAV and TRAJ regions, and the term "TCR alpha constant domain" refers to the extracellular TRAC region, or to a C-terminal truncated TRAC sequence. Likewise the term "TCR beta variable domain" refers to the concatenation of TRBV and TRBD/TRBJ regions, and the term "TCR beta constant domain" refers to the extracellular TRBC region, or to a C-terminal truncated TRBC sequence.

The unique sequences defined by the IMGT nomenclature are widely known and accessible to those working in the TCR field. For example, they can be found in the IMGT public database. The "T cell Receptor Factsbook", (2001 ) LeFranc and LeFranc, Academic Press, ISBN 0-12-441352-8 also discloses sequences defined by the IMGT nomenclature, but because of its publication date and consequent time-lag, the information therein sometimes needs to be confirmed by reference to the IMGT database.

We have further characterised a previously only partially-described native MAGE-3 TCR (Clone EB81 -103 from Dr. Pierre G. Coulie, Cellular Genetics Unit, University of Louvain, Avenue Hippocrate 74, UCL 7459, B-1200 Brussels, Belgium; see also Karanikas, et. al. (2003) "Monoclonal anti-MAGE-3 CTL responses in melanoma patients displaying tumor regression after vaccination with a recombinant canarypox virus." J. Immunol. 171 (9): 4898-904)) to ascertain that it has the following alpha chain and beta chain V, J and C gene usage: Alpha chain: - TRAV21 ^*01/TRAJ28/TRAC (the extracellular sequence of the native MAGE-3 TCR alpha chain is given in Figure 1 (SEQ ID No: 2). The CDRs are defined by amino acids 27-32 (CDR1 ), 49-52 (CDR2) and 95-103 (CDR3) of SEQ ID NO: 2.

Beta chain: - TRBV5-1 ^*01/TRBD1 /TRBJ2-7^*01 /TRBC2 (the extracellular sequence of the native MAGE-3 TCR beta chain is given in Figure 2 (SEQ ID

No: 3). . The CDRs are defined by amino acids 27-31 (CDR1 ), 50-54 (CDR2) and 91 -102 (CDR3) of SEQ ID NO: 3. (Note that the prior art only described the gene usage of this particular TCR to the broad level of TRBV5, whereas the applicants have ascertained that the actual gene usage is TRBV5-1 ). The TRBV5-1 sequence has 2 allelic variants, designated in

IMGT nomenclature as TRBV5-1 ^*01 and ^*02 respectively, and the native MAGE-3 TCR clone referred to above has the ^*01 variation. In the same way, the TRBJ2-7 sequence has two known variations and it is the ^*01 sequence which is present in the TCR clone referred to above. Note also that the absence of a"^*" qualifier means that only one allele is known for the relevant sequence.)

The terms "wild type TCR", "native TCR", "wild type MAGE-3 TCR", and "native MAGE-3 TCR" are used synonymously herein to refer to this naturally occurring TCR having the extracellular alpha and beta chain SEQ ID Nos: 2 and 3 respectively.

For the purpose of providing a reference TCR against which the binding profile of TCRs of the invention may be compared, it is convenient to use the soluble TCR having the extracellular sequence of the native MAGE-3 TCR alpha chain given in Figure 3 (SEQ ID No: 4) and the extracellular sequence of the native MAGE-3 TCR beta chain given in Figure 4 (SEQ ID No: 5). That TCR is referred to herein as the "the reference TCR" or "the reference MAGE-3 TCR". Note that SEQ ID No: 4 is the native alpha chain extracellular sequence ID No: 2 except that C162 has been substituted for T162 (i.e. T48 of TRAC). Likewise SEQ ID No: 5 is the native beta chain extracellular sequence ID No: 3 except that C169 has been substituted for S169 (i.e. S57 of TRBC2), A187 has been substituted for C187 and D201 has been substituted for N201 . These cysteine substitutions relative to the native alpha and beta chain extracellular sequences enable the formation of an interchain disulfide bond which stabilises the refolded soluble TCR, i.e. the TCR formed by refolding extracellular alpha and beta chains. Use of the stable disulfide linked soluble TCR as the reference TCR enables more convenient assessment of binding affinity and binding half life. Description of Figures

Figure 1 (SEQ ID No: 2) gives the amino acid sequence of the extracellular part of the alpha chain of a wild type MAGE-3-specific TCR with gene usage

TRAV21 ^*01/TRAJ28/TRAC.

Figure 2 (SEQ ID No: 3) gives the amino acid sequence of the extracellular part of the beta chain of a wild type MAGE-3-specific TCR TRBV5-1 ^*01/TRBD1/TRBJ2- 7^*01/TRBC2 beta chain amino acid sequence.

Figure 3 (SEQ ID No: 4) gives the amino acid sequence of the alpha chain of a soluble TCR (referred to herein as the "reference TCR"). The sequence is the same as that of Figure 1 except that a cysteine (bold and underlined) is substituted for T162 of SEQ ID No: 1 (i.e. T48 of the TRAC constant region).

Figure 4 (SEQ ID No: 5) gives the amino acid sequence of the beta chain of a soluble TCR (referred to herein as the "reference TCR"). The sequence is the same as that of Figure 2 except that a cysteine (bold and underlined) is substituted for S169 (i.e. S57 of the TRBC2 constant region) and A187 is substituted for C187 and D201 is substituted for N201 .

Figure 5 (SEQ ID No: 6) and Figure 6 (SEQ ID No: 7) give the amino acid sequence of the alpha chains which may be present in TCRs of the invention. The

subsequences forming the CDR regions, or substantial parts of the CDR regions are underlined. An introduced cysteine referred to in relation to Figure 3 is shown bold and underlined.

Figure 7 (SEQ ID No: 8), Figure 8 (SEQ ID No: 9), Figure 9 (SEQ ID No: 10), Figure 10 (SEQ ID No: 1 1 ), Figure 1 1 (SEQ ID No: 12) and Figure 12 (SEQ ID No: 13) give the amino acid sequence of the beta chains which may be present in TCRs of the invention. The subsequences forming the CDR regions, or substantial parts of the CDR regions are underlined. Relative to the wild type sequence shown in Figure 2, an introduced cysteine (referred to in relation to Figure 4) is shown bold and underlined and, also relative to the wild type sequence in Figure 2, C187 of has been mutated to A187 to eliminate an unpaired cysteine in any alpha-beta TCR having these beta chains. Figures 13A (SEQ ID No 18) and 13B (SEQ ID No 19) respectively give DNA sequences encoding the TCR alpha and beta chains for Figures 3 and 4 (introduced cysteines are shown in bold). Figure 14 (SEQ ID NO: 20) gives the amino acid sequence of an anti-CD3 scFv antibody fragment (bold type) fused via a linker namely GGGGS (underlined) at the N-terminus of a MAGE-3 TCR β chain. The MAGE-3 TCR β chain is that of Figure 10 (SEQ ID No: 12). Figures 15 (a)-(f) show the results of the tests described in Example 6 for the TCR- scFv antibody fusions prepared according to Example 5.

Figure 16 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having single point mutation in the CDR2 of the a chain to (ii) the corresponding binding of the reference TCR.

Figure 17 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having single point mutation in the CDR3 of the a chain to (ii) the corresponding binding of the reference TCR.

Figure 18 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having single point mutation in the CDR1 of the β chain to (ii) the corresponding binding of the reference TCR. Figure 19 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having single point mutation in the CDR2 of the β chain to (ii) the corresponding binding of the reference TCR.

Figure 20 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having single point mutation in the CDR3 of the β chain to (ii) the corresponding binding of the reference TCR.

Figure 21 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having multiple mutations in the CDR2 of the a chain to (ii) the

corresponding binding of the reference TCR. Figure 22 shows the ratio of binding to MAGE- A3 peptide:HLA for TCRs of the invention having multiple mutations in the CDR2 of the β chain to the corresponding binding of the reference TCR. Figure 23 shows the ratio of (i) binding to MAGE-A3 peptide:HLA for TCRs of the invention having multiple mutations in the CDR3 of the β chain to (ii) the

corresponding binding of the reference TCR.

Detailed Description of the Invention

According to a first aspect of the invention, there is provided a T cell receptor (TCR) having the property of binding to EVDPIGHLY (SEQ ID No: 1 ) HLA-A1 complex and comprising a TCR alpha chain variable domain and a TCR beta chain variable domain,

the alpha chain variable domain comprising an amino acid sequence that has at least 90% identity to the sequence of amino acid residues 1 -1 14 of SEQ ID No: 2, and

the beta chain variable domain comprising an amino acid sequence that has at least 90% identity to the sequence of amino acid residues 1 -1 12 of SEQ ID No: 3, wherein the alpha chain variable domain has at least one of the following mutations:

and/or the beta chain variable domain has at least one of the following mutations

Residue no.

52 M S G A H R C V Q

53 L I V E M Y A W C

54 L I V N G W A H C

94 G

95 L I F R Y K V E C T Q H S K A N W M

97 L

98 V I L

99 S T G

100 G R

102 Q with the proviso that the TCR is not a TCR comprising an alpha variable domain which has the amino acid sequence from K1 to P1 14 of SEQ ID NO: 2 except that at least one of the following mutations is present: 50V, 51 R, 52P and 53Y and/or a beta variable domain which has the amino acid sequence from K1 to T1 12 of SEQ ID NO: 3 except that at least one of the following mutations is present: 50F, 51 D, 52M, 53L and 54L.

In one embodiment, in the alpha chain variable domain

(i) the sequence of amino acid residues 1 -48 thereof (a) has at least 90% identity to the sequence of amino acid residues 1 -48 of SEQ ID No: 2 or (b) has one, two or three amino acid residues inserted or deleted relative to the sequence of (a);

(ii) amino acid residues 49-53 are as follows:

or another functionally silent amino acid residue;

(iii) the sequence of amino acid residues 54-94 thereof has (a) at least 90% identity to the sequence of amino acid residues 54-94 of SEQ ID NO: 2 or (b) has one, two or three amino acid residues inserted or deleted relative to the sequence of (a);

(iv) amino acid residues 95-98 are as follows:

Residue no.

96 G Q

97 A

98 G C

99 S P N R

100 Y D

101 Q F Y D M

103 L

103 T or another functionally silent amino acid residue;

(v) the sequence of amino acid residues 104-1 14 thereof is 100% identical to the sequence of amino acid residues 104-1 14 of SEQ ID No: 2 or has one, two or three insertions, deletions or substitutions relative thereto,

and in the beta chain variable domain

(i) the sequence of amino acid residues 1 -49 thereof has (a) at least 90% identity to the amino acid sequence of residues 1 -49 of SEQ ID No: 3 or (b) has one, two or three amino acid residues inserted or deleted relative to the sequence of (a);

(ii) amino acid residues 50-54 are:

or another functionally silent amino acid residue;

(iii) the sequence of amino acid residues 55-93 thereof has (a) at least 90% identity to the sequence of amino acid residues 55-93 of SEQ ID NO: 3 or (b) has one, two or three amino acid residues inserted or deleted relative to the sequence of (a);

(iv) amino acid residues 94-103 are as follows:

Residue no.

101 Q E

102 Y E

103 F Y or another functionally silent amino acid residue;

(v) the sequence of amino acid residues 104-1 12 thereof is 100% identical to the sequence of amino acid residues 104-1 12 of SEQ ID No: 3 or has one, two or three amino acid residues inserted, deleted or substituted relative thereto,

EXCEPT THAT at least one of the following mutations is present in the alpha chain variable domain:

and/or at least one of the following mutations is present in the beta chain variable domain:

In the first aspect of the invention, the amino acid sequence of the CDR1 of the alpha chain variable domain may be identical to the sequence of the CDR1 of the alpha chain of the wild type TCR. Additionally or alternatively, the amino acid sequence of the CDR1 of the beta chain variable domain may be identical to the sequence of the CDR1 of the beta chain of the wild type TCR.

TCRs of the invention may have more than one mutation present in the alpha chain variable domain and/or the beta chain variable domain. In certain embodiments, there are 2-20, 3-15, 4-12 or 4-10 mutations in one or both variable domains. There may be 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16 or 17 mutations in one or both variable domains. Mutations can be carried out using any appropriate method including, but not limited to, those based on polymerase chain reaction (PCR), restriction enzyme-based cloning, or ligation independent cloning (LIC) procedures. These methods are detailed in many of the of the standard molecular biology texts. For further details regarding polymerase chain reaction (PCR) and restriction enzyme-based cloning, see Sambrook & Russell, (2001 ) Molecular Cloning - A Laboratory Manual (3^rd Ed.) CSHL Press. Further information on ligation

independent cloning (LIC) procedures can be found in Rashtchian, (1995) Curr Opin Biotechnol S^ ) 30-6.

The inventors have found that certain amino acid residues may be changed without having any effect on binding of the TCR. These functionally silent changes provide several advantages to those skilled in the art, such as being employed to modify the biophysical or expression characteristics of the TCR, alter the docking characteristics of nearby residues and to allow one to ensure that there are no B or T cell epitopes in the TCR of the invention that could give an unwanted immune reaction when administered to a patient. The following table sets out examples of changes that may be made in the alpha chain variable domain without having any effect on the binding of the TCR (the wild type amino acids are shown in bold):

The following table sets out examples of changes that may be made in the beta chain variable domain without having any effect on the binding of the TCR (the wild type amino acids are shown in bold):

Accordingly, one or more of functionally silent changes may be included in addition to the or each mutation in the alpha and/or beta chain variable domain. In certain embodiments, there are 2-20, 3-15, 4-12 or 4-10 of these changes in one or both of the variable domains. There may be 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16 or 17 changes in one or both of the variable domains.

In some embodiments, TCRs of the invention have the following mutations in the alpha chain variable domain:

· 49Y, 50V, 51 R, 52P and 53Y, or

• 49Y, 50V, 51 R, 52P, 53Y, 99Y, 100V, 101 R, 102P and 103Y

and/or the following mutations in the beta chain variable domain:

• 50T, 51 D, 52F, 53L and 54L,

• 52F, 53L and 54L,

· 50T, 51 D and 52F,

• 95F, 99T and 100G, or • 94N, 95L, 97Q and 98V.

The alpha chain variable domain may comprise an amino acid sequence that has at least 91 , 92, 93, 94, 95, 96, 97, 98 or 99% identity, or have 100% identity to the sequence of amino acid residues 1 -1 14 of SEQ ID No: 2, and/or the beta chain variable domain may comprise an amino acid sequence that has at least 91 , 92, 93, 94, 95, 96, 97, 98 or 99% identity, or have 100% identity to the sequence of amino acid residues 1 -1 12 of SEQ ID No: 3. The sequence of amino acid residues 1 -48, 54-94 and/or 104-1 14 of the alpha chain variable domain of the TCR of the invention may have at least 91 , 92, 93, 94, 95, 96, 97, 98 or 99% identity, or have 100% identity, to the sequence of amino acid residues 1 -48, 54-94 and/or 104-1 14 of SEQ ID No: 2, respectively. Similarly, the sequence of amino acid residues 1 -49, 55-93 and/or 103-1 12 of the beta chain variable domain of the TCR of the invention may have at least 91 , 92, 93, 94, 95, 96, 97, 98 or 99% identity, or have 100% identity, to the sequence of amino acid residues 1 -49, 55-93 and/or 103-1 12 of SEQ ID No: 3, respectively.

In a further aspect of the invention, there is provided an isolated, engineered or non- naturally occurring T cell receptor (TCR) comprising an alpha chain variable domain and a beta chain variable domain, wherein the TCR comprises:

(a) in the alpha chain variable domain, with reference to SEQ ID NO: 2:

(i) one or more mutations at positions 49, 50, 51 , 52, 53, 95, 96, 97, 98, 99, 100, 101 , 102 or 103, and

(ii) at positions 1 -48 there is at least 90% identity to residues 1 -48 of

SEQ ID NO:2, at positions 54-94 there is at least 90% identity to residues 54- 94 of SEQ ID NO:2, and at positions 104-1 14 there is at least 90% identity to residues 104-1 14 of SEQ ID NO:2, and

(b) in the beta chain variable domain, with reference to SEQ ID NO:3:

(i) one or more mutations at positions 31 , 50, 51 , 52, 53, 54, 93, 94,

95, 96, 97, 98, 99, 100, 101 , 102 or 103, and

(ii) at positions 55-90 there is at least 90% identity to residues 55-90 of SEQ ID NO:3, and at positions 104-1 12 there is at least 90% identity to positions 104-1 12 of SEQ ID NO:3,

and wherein the TCR binds to EVDPIGHLY (SEQ ID NO:1 ) HLA complex. The TCR of this aspect of the invention may not be a TCR comprising an alpha variable domain which has the amino acid sequence from K1 to P114 of SEQ ID NO: 2 except that at least one of the following mutations is present: 50V, 51 R, 52P and 53Y and/or a beta variable domain which has the amino acid sequence from K1 to T112 of SEQ ID NO: 3 except that at least one of the following mutations is present: 50F, 51 D, 52M, 53L and 54L.

In the alpha chain variable domain at positions 1-48, there may be at least 95% identity to residues 1-48 of SEQ ID NO:2, and/or at positions 54-94 there may be at least 95% identity to residues 54-94 of SEQ ID NO:2. In the beta chain variable domain at positions 55-90, there may be at least 95 % identity to residues 55-90 of SEQ ID NO:3. In the alpha chain variable domain at positions 104-114, there may be at least 95 % identity to residues 104-114 of SEQ ID NO:2, and/or in the beta chain variable domain at positions 104-112, there may be at least 95 % identity to positions 104-1 12 of SEQ ID NO:3.

In one aspect of the invention, there is provided a T cell receptor (TCR) having the property of binding to EVDPIGHLY (SEQ ID No: 1) HLA-A1 complex and comprising a TCR alpha chain variable domain and a TCR beta chain variable domain,

the said alpha chain variable domain having at least 95% sequence identity to the amino acid sequence from K1 to P114 of SEQ ID No: 6 or of SEQ ID No: 7; and the said beta chain variable domain having at least 95% sequence identity to the amino acid sequence from K1 to T112 of SEQ ID No: 8 or of SEQ ID No: 9 or of SEQ ID No: 10 or of SEQ ID No: 11 or of SEQ ID No: 12 or of SEQ ID No: 13;

PROVIDED THAT in the said alpha chain variable domain

in the case of SEQ ID No: 6, the subsequences from D27 to N32, and from

Y49 to P52 and from C90 to T103 are all invariant; and

in the case of SEQ ID No: 7, the subsequences from D27 to N32, and from

Y49 to P52 and from C90 to V103 are all invariant;

AND PROVIDED THAT in said beta chain variable domain

in the case of SEQ ID No: 8, the subsequences from S27 to S31 and from

T50 to L54 and from C91 to Y102 are all invariant; and

in the case of SEQ ID No: 9, the subsequences from S27 to S31 and from

F50 to L54 and from C91 to Y102 are all invariant; and

in the case of SEQ ID No: 10, the subsequences from S27 to S31 and from

T50 to Q54 and from C91 to Y102 are all invariant; and

RECTIFIED SHEET (RULE 91 ) ISA/EP in the case of SEQ ID No: 11 and SEQ ID No: 12 and SEQ ID No: 13, the subsequences from S27 to S31 and from F50 to Q54 and from C91 to Y102 are all invariant. The invariant subsequences of the TCR's of this aspect of the invention are, or constitute a substantial part of, the complementarity determining regions (CDRs) of the alpha and beta chain variable domains.

The TCRs of the invention have the property of binding the MAGE-3 EVDPIGHLY (SEQ ID No: 1 ) HLA-A1 complex. Certain TCRs of the invention also bind the

MAGE-6 EVDPIGHVY HLA-A1 and MAGE-B18 EVDPIRHYY HLA-A1 complexes which allows a greater percentage of patients to be treated with a single TCR of the invention. Certain TCRs of the invention have been found to be highly specific for those MAGE epitopes relative to other, irrelevant epitopes, and are thus particularly suitable as targeting vectors for delivery of therapeutic agents or detectable labels to cells and tissues displaying those epitopes. Specificity in the context of TCRs of the invention relates to their ability to recognise MAGE-3 antigen positive HLA-A1 positive target cells whilst having minimal ability to recognise MAGE-3 negative targets cells, particularly non-cancerous human cells. Specificity can be measured, for example, in cellular assays such as those described in Example 6. Achieving the required levels of antigen specificity proved significantly more difficult to achieve for the TCRs of the invention as it was found that the CD8 co-receptor expressed primarily on T cells was found to provide a stronger stabilising effect when binding HLA-A1 positive cells compared to when binding to HLA-A2 positive cells and that this enhanced stabilising effect enhanced any non-antigen specific binding of the TCR. Certain TCRs of the invention have been found to be highly suitable for use in adoptive therapy. Such TCRs may have a K_D for the complex of from about 6 μΜ to about 70 μΜ and/or have a binding half-life (T½) for the complex in the range of from about 1 to about s. Certain TCRs of the invention have been found to be highly suitable for use as therapeutics and/or diagnostics when coupled to a detectable label or therapeutic agent. Such TCRs may have a K_D for the complex in the range of from about 10 pM to about 100 nM.

The TCRs of the invention may be αβ heterodimers or may be in single chain format. Single chain formats include αβ TCR polypeptides of the Va-L-Vp, Ν β-L-V , Va-Ca- L-Ν β, or να-Ι--νβ-ΰβ types, wherein Va and \/β are TCR a and β variable regions

RECTIFIED SHEET (RULE 91 ) ISA/EP respectively, Coc and Οβ are TCR a and β constant regions respectively, and L is a linker sequence. For use as a targeting agent for delivering therapeutic agents to the antigen presenting cell the TCR may be in soluble form (i.e. having no

transmembrane or cytoplasmic domains). For stability, TCRs of the invention, and preferably soluble αβ heterodimeric TCRs, may have an introduced disulfide bond between residues of the respective constant domains, as described, for example, in WO 03/020763. TCRs of the invention may be isolated, engineered or non-naturally occurring. For use in adoptive therapy, an αβ heterodimeric TCR may, for example, be transfected as full length chains having both cytoplasmic and transmembrane domains.

In some embodiments, the alpha chain variable domain may have at least 96, 97, 98 or 99% sequence identity, or 100% sequence identity, to the amino acid sequence from K1 to P1 14 of SEQ ID No: 6 or of SEQ ID No: 7. The amino acids underlined in Figures 5 and 6 may be invariant.

In some embodiments, the beta chain variable domain may have at least 96, 97, 98 or 99% sequence identity, or 100% sequence identity, to the amino acid sequence from K1 to T1 12 of SEQ ID No: 8 or of SEQ ID No: 9 or of SEQ ID No: 10 or of SEQ ID No: 1 1 or of SEQ ID No: 12 or of SEQ ID No: 13. The amino acids underlined in Figures 7, 8, 9, 10, 1 1 and 12 may be invariant.

In one subclass of TCRs of the invention, the alpha chain variable domain may comprise K1 to P1 14 of SEQ ID No: 6; and the beta chain may comprise K1 to T1 12 of SEQ ID No: 8.

In another subclass of TCRs of the invention, the alpha chain variable domain may comprise K1 to P1 14 of SEQ ID No: 7 and the beta chain may comprise K1 to T1 12 of SEQ ID No: 9 or of SEQ ID No: 10 or of SEQ ID No: 1 1 or of SEQ ID No: 12 or of SEQ ID No: 13.

Alpha-beta heterodimeric TCRs of the invention usually comprise an alpha chain TRAC constant domain sequence and a beta chain TRBC1 or TRBC2 constant domain sequence. The alpha and beta chain constant domain sequences may be modified by truncation or substitution to delete the native disulfide bond between Cys4 of exon 2 of TRAC and Cys2 of exon 2 of TRBC1 or TRBC2. The alpha and beta chain constant domain sequences may also be modified by substitution of cysteine residues for Thr 48 of TRAC and Ser 57 of TRBC1 or TRBC2, the said cysteines forming a disulfide bond between the alpha and beta constant domains of the TCR.

Certain TCRs of the invention have a binding affinity for, and/or a binding half-life for, the EVDPIGHLY-HLA-A1 complex substantially higher than that of the reference MAGE-3 TCR. Increasing the binding affinity of a native TCR often reduces the selectivity of the TCR for its peptide-MHC ligand, but the TCRs of the invention remain selective for the EVDPIGHLY-HLA-A1 complex, despite, in some

embodiments, having substantially higher binding affinity than the parent native TCR.

Binding affinity (inversely proportional to the equilibrium constant K_D) and binding half-life (expressed as T½) can be determined by any appropriate method. It will be appreciated that doubling the affinity of a TCR results in halving the K_D. T½ is calculated as In2 divided by the off-rate (k_off). So doubling of T½ results in a halving in k_off. K_D and k_off values for TCRs are usually measured for soluble forms of the TCR, i.e. those forms which are truncated to remove hydrophobic cytoplasmic and transmembrane domain residues. Therefore it is to be understood that a given TCR meets the requirement that it has a binding affinity for, and/or a binding half-life for, the EVDPIGHLY-HLA-A1 complex if a soluble form of that TCR meets that requirement. Preferably the binding affinity or binding half-life of a given TCR is measured several times, for example 3 or more times, using the same assay protocol, and an average of the results is taken. In a preferred embodiment these measurements are made using the Surface Plasmon Resonance (BIAcore) method of Example 3 herein. The reference MAGE-3 TCR has a K_D of approximately 250 μΜ as measured by that method, and its k_off is approximately 0.2 s^" (i.e T½ is

approximately 3 s). In a further aspect, the present invention provides nucleic acid encoding a TCR of the invention. The invention also provides a cell harbouring a TCR expression vector which comprises nucleic acid of the invention in a single open reading frame, or two distinct open reading frames encoding the alpha chain and the beta chain

respectively. Another aspect provides a cell harbouring a first expression vector which comprises nucleic acid encoding the alpha chain of a TCR of the invention, and a second expression vector which comprises nucleic acid encoding the beta chain of a TCR of the invention. Such cells are particularly useful in adoptive therapy.

Since the TCRs of the invention have utility in adoptive therapy, the invention includes an isolated or non-naturally occurring cell, especially a T-cell, presenting a TCR of the invention. There are a number of methods suitable for the transfection of T cells with nucleic acid (such as DNA or RNA) encoding the TCRs of the invention (see for example Robbins et ai, (2008) J Immunol. 180: 61 16-6131 ). T cells expressing the TCRs of the invention will be suitable for use in adoptive therapy- based treatment of MAGE-3⁺ HLA-A1 ⁺ cancers. As will be known to those skilled in the art, there are a number of suitable methods by which adoptive therapy can be carried out (see for example Rosenberg et ai, (2008) Nat Rev Cancer 8(4): 299- 308). Some soluble TCRs of the invention are useful for delivering detectable labels or therapeutic agents to the antigen presenting cells and tissues containing the antigen presenting cells. They may therefore be associated (covalently or otherwise) with a detectable label (for diagnostic purposes wherein the TCR is used to detect the presence of cells presenting the EVDPIGHLY-HLA-A1 complex); a therapeutic agent; or a PK modifying moiety (for example by PEGylation).

Detectable labels for diagnostic purposes include for instance, fluorescent labels, radiolabels, enzymes, nucleic acid probes and contrast reagents. Therapeutic agents which may be associated with the TCRs of the invention include immunomodulators, radioactive compounds, enzymes (perforin for example) or chemotherapeutic agents (cis-platin for example). To ensure that toxic effects are exercised in the desired location the toxin could be inside a liposome linked to TCR so that the compound is released slowly. This will prevent damaging effects during the transport in the body and ensure that the toxin has maximum effect after binding of the TCR to the relevant antigen presenting cells.

Other suitable therapeutic agents include:

• small molecule cytotoxic agents, i.e. compounds with the ability to kill

mammalian cells having a molecular weight of less than 700 Daltons. Such compounds could also contain toxic metals capable of having a cytotoxic effect. Furthermore, it is to be understood that these small molecule cytotoxic agents also include pro-drugs, i.e. compounds that decay or are converted under physiological conditions to release cytotoxic agents. Examples of such agents include cis-platin, maytansine derivatives, rachelmycin, calicheamicin, docetaxel, etoposide, gemcitabine, ifosfamide, irinotecan, melphalan, mitoxantrone, sorfimer sodiumphotofrin II, temozolomide, topotecan, trimetreate glucuronate, auristatin E vincristine and doxorubicin;

• peptide cytotoxins, i.e. proteins or fragments thereof with the ability to kill mammalian cells. For example, ricin, diphtheria toxin, pseudomonas bacterial exotoxin A, DNase and RNase;

• radio-nuclides, i.e. unstable isotopes of elements which decay with the

concurrent emission of one or more of a or β particles, or γ rays. For example, iodine 131 , rhenium 186, indium 1 1 1 , yttrium 90, bismuth 210 and 213, actinium 225 and astatine 213; chelating agents may be used to facilitate the association of these radio-nuclides to the high affinity TCRs, or multimers thereof;

• immuno-stimulants, i.e. immune effector molecules which stimulate immune response. For example, cytokines such as IL-2 and IFN-γ,

• Superantigens and mutants thereof;

• TCR-HLA fusions;

• chemokines such as IL-8, platelet factor 4, melanoma growth stimulatory

protein, etc;

• antibodies or fragments thereof, including anti-T cell or NK cell determinant antibodies (e.g. anti-CD3, anti-CD28 or anti-CD16);

• alternative protein scaffolds with antibody like binding characteristics

• complement activators;

• xenogeneic protein domains, allogeneic protein domains, viral/bacterial

protein domains, viral/bacterial peptides.

One preferred embodiment is provided by a TCR of the invention associated (usually by fusion to an N-or C-terminus of the alpha or beta chain) with an anti-CD3 antibody, or a functional fragment or variant of said anti-CD3 antibody. Antibody fragments and variants/analogues which are suitable for use in the compositions and methods described herein include minibodies, Fab fragments, F(ab')₂ fragments, dsFv and scFv fragments, Nanobodies™ (these constructs, marketed by Ablynx (Belgium), comprise synthetic single immunoglobulin variable heavy domain derived from a camelid (e.g. camel or llama) antibody) and Domain Antibodies (Domantis (Belgium), comprising an affinity matured single immunoglobulin variable heavy domain or immunoglobulin variable light domain) or alternative protein scaffolds that exhibit antibody like binding characteristics such as Affibodies (Affibody (Sweden), comprising engineered protein A scaffold) or Anticalins (Pieris (Germany)), comprising engineered anticalins) to name but a few.

Specific embodiments of anti-CD3-TCR fusion constructs of the invention include those which have alpha chain SEQ ID No: 6 or 7, and the anti CD3-TCR beta chain fusion SEQ ID No: 20, or the anti CD3-TCR beta chain fusion SEQ ID No. 20 in which the TCR beta chain sequence is replaced by SEQ ID No: 9, 10, 1 1 , or 13. In such constructs, the GGGGS linker sequence may replaced by an alternative linker sequence selected from GGGSG, GGSGG and GSGGG, and/or amino acid A1 of SEQ ID No: 20 by D, and/or by substituting amino acid K1 of SEQ ID Nos: 8-13 by A. For some purposes, the TCRs of the invention may be aggregated into a complex comprising several TCRs to form a multivalent TCR complex. There are a number of human proteins that contain a multimerisation domain that may be used in the production of multivalent TCR complexes. For example the tetramerisation domain of p53 which has been utilised to produce tetramers of scFv antibody fragments which exhibited increased serum persistence and significantly reduced off-rate compared to the monomeric scFv fragment (Willuda et al. (2001 ) J. Biol. Chem. 276 (17) 14385-14392). Haemoglobin also has a tetramerisation domain that could be used for this kind of application. A multivalent TCR complex of the invention may have enhanced binding capability for the EVDPIGHLY HLA-A1 complex compared to a non-multimeric wild-type or T cell receptor heterodimer of the invention. Thus, multivalent complexes of TCRs of the invention are also included within the invention. Such multivalent TCR complexes according to the invention are particularly useful for tracking or targeting cells presenting particular antigens in vitro or in vivo, and are also useful as intermediates for the production of further multivalent TCR complexes having such uses.

TCRs of the invention may be glycosylated when expressed by transfected cells. As is well known, the glycosylation pattern of transfected TCRs may be modified by mutations of the transfected gene.

For administration to patients, the TCRs of the invention (associated with a detectable label or therapeutic agent or expressed on a transfected T cell), may be provided in a pharmaceutical composition together with one or more pharmaceutically acceptable carriers or excipients. Therapeutic or imaging TCRs, or cells, in accordance with the invention will usually be supplied as part of a sterile, pharmaceutical composition which will normally include a pharmaceutically acceptable carrier. This pharmaceutical composition may be in any suitable form, (depending upon the desired method of administering it to a patient). It may be provided in unit dosage form, will generally be provided in a sealed container and may be provided as part of a kit. Such a kit would normally (although not necessarily) include instructions for use. It may include a plurality of said unit dosage forms.

The pharmaceutical composition may be adapted for administration by any appropriate route, preferably a parenteral (including subcutaneous, intramuscular, or preferably intravenous) route. Such compositions may be prepared by any method known in the art of pharmacy, for example by mixing the active ingredient with the carrier(s) or excipient(s) under sterile conditions.

Dosages of the substances of the present invention can vary between wide limits, depending upon the disease or disorder to be treated, the age and condition of the individual to be treated, etc. and a physician will ultimately determine appropriate dosages to be used.

Also provided by the invention are:

• a TCR which binds the EVDPIGHLY peptide (derived from the MAGE-3

protein) presented as a peptide- HLA-A1 complex, or a cell expressing and/or presenting such a TCR, for use in medicine, preferably in a method of treating cancer;

• the use of a TCR which binds the EVDPIGHLY peptide (derived from the

MAGE-3 protein) presented as a peptide-HLA-A1 complex, or a cell expressing and/or presenting such a TCR, in the manufacture of a medicament for treating cancer;

• a method of treating cancer in a patient, comprising administering to the patient a TCR which binds the EVDPIGHLY peptide (derived from the MAGE-3 protein) presented as a peptide- HLA-A1 complex, or a cell expressing and/or presenting such a TCR. It is preferred that the TCR which binds the EVDPIGHLY peptide (derived from the MAGE-3 protein) presented as a peptide-HLA-A1 complex is a TCR of the invention. The cancer to be treated may be a haematological cancer such as myeloma for instance, or solid tumours such as melanoma, Head and Neck Squamous Cell, lung, hepatocellular, bladder, gastric, prostate, colorectal and esophageal carcimomas for instance.

Preferred features of each aspect of the invention are as for each of the other aspects mutatis mutandis. The prior art documents mentioned herein are incorporated to the fullest extent permitted by law.

The invention is further described in the following examples.

Examples

Example 1

Cloning of the reference MAGE-3 TCR alpha and beta chain variable region sequences into pGMT7-based expression plasmids The reference MAGE-3 TCR variable alpha and TCR variable beta domains were PCR amplified from total cDNA isolated from a MAGE-3 T cell clone (Clone EB81 - 103 from Pierre Coulie University of Louvain, Belgium). In the case of the alpha chain, an alpha chain variable region sequence specific oligonucleotide A1

(ggaattccatatgaaacaagaagttactcaaattcc SEQ ID No: 14) which encodes the restriction site Ndel and an alpha chain constant region sequence specific oligonucleotide A2 (ttgtcagtcgacttagagtctctcagctggtacacg SEQ ID No: 15) which encodes the restriction site Sail are used to amplify the alpha chain variable domain. In the case of the beta chain, a beta chain variable region sequence specific oligonucleotide B1 (gaattccatatgaaagctggagttactcaaactccaag SEQ ID No: 16) which encodes the restriction site Ndel and a beta chain constant region sequence specific oligonucleotide B2 (tagaaaccggtggccaggcacaccagtgtggc SEQ ID No: 17) which encodes the restriction site Agel are used to amplify the beta chain variable domain.

The alpha and beta variable domains were cloned into pGMT7-based expression plasmids containing either Coc or C by standard methods described in (Molecular Cloning a Laboratory Manual Third edition by Sambrook and Russell). Plasmids were sequenced using an Applied Biosystems 3730x1 DNA Analyzer. The DNA sequences encoding the TCR alpha chain cut with Ndel and Sail were ligated into pGMT7 + Coc vector, which was cut with Ndel and Xhol. The DNA sequences encoding the TCR beta chain cut with Ndel and Agel was ligated into separate pGMT7 + C vector, which was also cut with Ndel and Agel.

Ligation

Ligated plasmids were transformed into competent E.coli strain XL1 -blue cells and plated out on LB/agar plates containing 100 μg/ml ampicillin. Following incubation overnight at 37^QC, single colonies are picked and grown in 10 ml LB containing 100 μg/ml ampicillin overnight at 37^QC with shaking. Cloned plasmids were purified using a Miniprep kit (Qiagen) and the plasmids were sequenced using an Applied

Biosystems 3730x1 DNA Analyzer. Figures 3 and 4 show respectively the reference MAGE-3 TCR a and β chain extracellular amino acid sequences (SEQ ID Nos: 4 and 5 respectively) produced from the DNA sequences of Figures 13A (SEQ ID No: 18) and 13B (SEQ ID No: 19) respectively. Note that, relative to the native TCR, cysteines were substituted in the constant regions of the alpha and beta chains to provide an artificial inter-chain disulphide bond on refolding to form the heterodimeric TCR. The introduced cysteines are shown in bold and underlined. The restriction enzyme recognition sequences in the DNA sequences of Figures 13A and 13B are underlined.

Example 2

Expression, refolding and purification of soluble reference MAGE-3 TCR

The expression plasmids containing the TCR oc-chain and β-chain respectively, as prepared in Example 1 , were transformed separately into E.coli strain BL21 pLysS, and single ampicillin-resistant colonies were grown at 37°Cin TYP (ampicillin 100 μg ml) medium to OD₆₀o of -0.6-0.8 before inducing protein expression with 0.5 mM IPTG. Cells were harvested three hours post-induction by centrifugation for 30 minutes at 4000rpm in a Beckman J-6B. Cell pellets were lysed with 25 ml Bug Buster (NovaGen) in the presence of MgCI₂ and DNasel. Inclusion body pellets were recovered by centrifugation for 30 minutes at 13000rpm in a Beckman J2-21 centrifuge. Three detergent washes were then carried out to remove cell debris and membrane components. Each time the inclusion body pellet was homogenised in a Triton buffer (50 mM Tris-HCI pH 8.0, 0.5% Triton-X100, 200 mM NaCI, 10 mM NaEDTA,) before being pelleted by centrifugation for 15 minutes at 13000rpm in a Beckman J2-21 . Detergent and salt was then removed by a similar wash in the following buffer: 50 mM Tris-HCI pH 8.0, 1 mM NaEDTA. Finally, the inclusion bodies were divided into 30 mg aliquots and frozen at -70 °G Inclusion body protein yield was quantified by solubilising with 6 M guanidine-HCI and an OD measurement was taken on a Hitachi U-2001 Spectrophotometer. The protein concentration was then calculated using the extinction coefficient.

Approximately 15mg of TCR β chain and 15mg of TCR a chain solubilised inclusion bodies were thawed from frozen stocks and diluted into 10ml of a guanidine solution (6 M Guanidine-hydrochloride, 50 mM Tris HCI pH 8.1 , 100 mM NaCI, 10 mM EDTA, 10 mM DTT), to ensure complete chain denaturation. The guanidine solution containing fully reduced and denatured TCR chains was then injected into 0.5 litre of the following refolding buffer: 100 mM Tris pH 8.1 , 400 mM L-Arginine, 2 mM EDTA, 5 M Urea. The redox couple (cysteamine hydrochloride and cystamine

dihydrochloride) to final concentrations of 6.6 mM and 3.7 mM respectively, were added approximately 5 minutes before addition of the denatured TCR chains. The solution was left for -30 minutes. The refolded TCR was dialysed in Spectrapor 1 membrane (Spectrum; Product No. 132670) against 10 L H₂0 for 18-20 hours. After this time, the dialysis buffer was changed twice to fresh 10 mM Tris pH 8.1 (10 L) and dialysis was continued at 5 °C± 3 °Cfor another ~8 hours. Soluble TCR was separated from degradation products and impurities by loading the dialysed refold onto a POROS 50HQ anion exchange column and eluting bound protein with a gradient of 0-500mM NaCI in 10 mM Tris pH 8.1 over 50 column volumes using an Akta purifier (GE Healthcare). Peak fractions were pooled and a cocktail of protease inhibitors (Calbiochem) were added. The pooled fractions were then stored at 4 °Cand analysed by Coomassie-stained SDS-PAGE before being pooled and concentrated. Finally, the soluble TCR was purified and characterised using a GE Healthcare Superdex 75HR gel filtration column pre-equilibrated in PBS buffer (Sigma). The peak eluting at a relative molecular weight of approximately 50 kDa was pooled and concentrated prior to characterisation by BIAcore surface plasmon resonance analysis. Example 3

Binding characterisation BIAcore Analysis

A surface plasmon resonance biosensor (BIAcore 3000™) can be used to analyse the binding of a soluble TCR to its peptide-MHC ligand. This is facilitated by producing soluble biotinylated peptide-HLA ("pHLA") complexes which can be immobilised to a streptavidin-coated binding surface (sensor chip). The sensor chips comprise four individual flow cells which enable simultaneous measurement of T-cell receptor binding to four different pHLA complexes. Manual injection of pHLA complex allows the precise level of immobilised class I molecules to be manipulated easily. Biotinylated class I HLA-A^*01 molecules were refolded in vitro from bacterially- expressed inclusion bodies containing the constituent subunit proteins and synthetic peptide, followed by purification and in vitro enzymatic biotinylation (O'Callaghan et al. (1999) Anal. Biochem. 266: 9-15). H LA- A^*01 -heavy chain was expressed with a C-terminal biotinylation tag which replaces the transmembrane and cytoplasmic domains of the protein in an appropriate construct. Inclusion body expression levels of -75 mg/litre bacterial culture were obtained. The MHC light-chain or β2- microglobulin was also expressed as inclusion bodies in E.coli from an appropriate construct, at a level of -500 mg/litre bacterial culture. E. coli cells were lysed and inclusion bodies are purified to approximately 80% purity. Protein from inclusion bodies was denatured in 6 M guanidine-HCI, 50 mM Tris pH 8.1 , 100 mM NaCI, 10 mM DTT, 10 mM EDTA, and was refolded at a concentration of 30 mg/litre heavy chain, 30 mg/litre β2ηι into 0.4 M L-Arginine, 100 mM Tris pH 8.1 , 3.7 mM cystamine dihydrochloride, 6.6 mM cysteamine hydrochloride, 4 mg/L of the MAGE-3 EVDPIGHLY peptide required to be loaded by the HLA-A^*01 molecule, by addition of a single pulse of denatured protein into refold buffer at < 5^QC.

Refolding was allowed to reach completion at 4^QC for at least 1 hour.

Buffer was exchanged by dialysis in 10 volumes of 10 mM Tris pH 8.1 . Two changes of buffer were necessary to reduce the ionic strength of the solution sufficiently. The protein solution was then filtered through a 1 .5μηι cellulose acetate filter and loaded onto a POROS 50HQ anion exchange column (8 ml bed volume). Protein was eluted with a linear 0-500 mM NaCI gradient in 10 mM Tris pH 8.1 using an Akta purifier (GE Healthcare). HLA-A^*01 -peptide complex eluted at approximately 250 mM NaCI, and peak fractions were collected, a cocktail of protease inhibitors (Calbiochem) was added and the fractions were chilled on ice. Biotinylation tagged pHLA molecules were buffer exchanged into 10 mM Tris pH 8.1 , 5 mM NaCI using a GE Healthcare fast desalting column equilibrated in the same buffer. Immediately upon elution, the protein-containing fractions were chilled on ice and protease inhibitor cocktail (Calbiochem) was added. Biotinylation reagents were then added: 1 mM biotin, 5 mM ATP (buffered to pH 8), 7.5 mM MgCI₂, and 5 μ9/ι ιΙ BirA enzyme (purified according to O'Callaghan et at. (1999) Anal. Biochem. 266: 9- 15). The mixture was then allowed to incubate at room temperature overnight.

The biotinylated pHLA-A^*01 molecules were purified using gel filtration

chromatography. A GE Healthcare Superdex 75 HR 10/30 column was pre- equilibrated with filtered PBS and 1 ml of the biotinylation reaction mixture was loaded and the column was developed with PBS at 0.5 ml/min using an Akta purifier (GE Healthcare). Biotinylated pHLA-A^*01 molecules eluted as a single peak at approximately 15 ml. Fractions containing protein were pooled, chilled on ice, and protease inhibitor cocktail was added. Protein concentration was determined using a Coomassie-binding assay (PerBio) and aliquots of biotinylated pHLA-A^*01 molecules were stored frozen at -20^QC.

Such immobilised complexes are capable of binding both T-cell receptors and the coreceptor CD8ococ, both of which may be injected in the soluble phase. The pHLA binding properties of soluble TCRs are observed to be qualitatively and quantitatively similar if the TCR is used either in the soluble or immobilised phase. This is an important control for partial activity of soluble species and also suggests that biotinylated pHLA complexes are biologically as active as non-biotinylated complexes.

The BIAcore 3000™ surface plasmon resonance (SPR) biosensor measures changes in refractive index expressed in response units (RU) near a sensor surface within a small flow cell, a principle that can be used to detect receptor ligand interactions and to analyse their affinity and kinetic parameters. The BIAcore experiments were performed at a temperature of 25 °Q using PBS buffer (Sigma, pH 7.1 -7.5) as the running buffer and in preparing dilutions of protein samples.

Streptavidin was immobilised to the flow cells by standard amine coupling methods. The pHLA complexes were immobilised via the biotin tag. The assay was then performed by passing soluble TCR over the surfaces of the different flow cells at a constant flow rate, measuring the SPR response in doing so.

Equilibrium binding constant

The above BIAcore analysis methods were used to determine equilibrium binding constants. Serial dilutions of the disulfide linked soluble heterodimeric form of the reference MAGE-3 TCR were prepared and injected at constant flow rate of 5 μΙ min^"1 over two different flow cells; one coated with ~1000 RU of specific EVDPIGHLY HLA- A^*01 complex, the second coated with -1000 RU of non-specific HLA-A2 -peptide (KIFGSLAFL (SEQ ID No: 23)) complex. Response was normalised for each concentration using the measurement from the control cell. Normalised data response was plotted versus concentration of TCR sample and fitted to a non-linear curve fitting model in order to calculate the equilibrium binding constant, K_D (Price & Dwek, Principles and Problems in Physical Chemistry for Biochemists (2^nd Edition) 1979, Clarendon Press, Oxford). The disulfide linked soluble form of the reference MAGE-3 TCR (Example 2) demonstrated a K_D of approximately 250 μΜ. From the same BIAcore data the T½ was approximately 3 s. Kinetic Parameters

The above BIAcore analysis methods were also used to determine equilibrium binding constants and off-rates. For high affinity TCRs (see Example 4 below) K_D was determined by experimentally measuring the dissociation rate constant, k_off, and the association rate constant, k_on. The equilibrium constant K_D was calculated as k_off/k_on.

TCR was injected over two different cells one coated with ~1000 RU of specific EVDPIGHLY HLA-A^*01 complex, the second coated with ~1000 RU of non-specific HLA-A1 -peptide complex. Flow rate was set at 50 μΙ/min. Typically 250 μΙ of TCR at ~ 1 μΜ concentration was injected. Buffer was then flowed over until the response had returned to baseline or >2 hours had elapsed. Kinetic parameters were calculated using BIAevaluation software. The dissociation phase was fitted to a single exponential decay equation enabling calculation of half-life. Example 4

Preparation of TCRs of the invention

Expression plasmids containing the TCR oc-chain and β-chain respectively for the following TCRs of the invention were prepared as in Example 1 :

The plasmids were transformed separately into E.coli strain BL21 pLysS, and single ampicillin-resistant colonies grown at 37°Cin TYP (ampicillin 100 μg ml) medium to OD₆₀₀ of -0.6-0.8 before inducing protein expression with 0.5 mM IPTG. Cells were harvested three hours post-induction by centrifugation for 30 minutes at 4000rpm in a Beckman J-6B. Cell pellets were lysed with 25 ml Bug Buster (Novagen) in the presence of MgCI₂ and DNasel. Inclusion body pellets were recovered by centrifugation for 30 minutes at 13000rpm in a Beckman J2-21 centrifuge. Three detergent washes were then carried out to remove cell debris and membrane components. Each time the inclusion body pellet was homogenised in a Triton buffer (50 mM Tris-HCI pH 8.0, 0.5% Triton-X100, 200 mM NaCI, 10 mM NaEDTA,) before being pelleted by centrifugation for 15 minutes at 13000rpm in a Beckman J2-21 . Detergent and salt was then removed by a similar wash in the following buffer: 50 mM Tris-HCI pH 8.0, 1 mM NaEDTA. Finally, the inclusion bodies were divided into 30 mg aliquots and frozen at -70 °C. Inclusion body protein yield was quantified by solubilising with 6 M guanidine-HCI and an OD measurement was taken on a Hitachi U-2001 Spectrophotometer. The protein concentration was then calculated using the extinction coefficient.

Approximately 10mg of TCR β chain and 10mg of TCR a chain solubilised inclusion bodies for each TCR of the invention were diluted into 10ml of a guanidine solution (6 M Guanidine-hydrochloride, 50 mM Tris HCI pH 8.1 , 100 mM NaCI, 10 mM EDTA, 10 mM DTT), to ensure complete chain denaturation. The guanidine solution containing fully reduced and denatured TCR chains was then injected into 0.5 litre of the following refolding buffer: 100 mM Tris pH 8.1 , 400 mM L-Arginine, 2 mM EDTA, 5 M Urea. The redox couple (cysteamine hydrochloride and cystamine dihydrochloride) to final concentrations of 6.6 mM and 3.7 mM respectively, were added

approximately 5 minutes before addition of the denatured TCR chains. The solution was left for -30 minutes. The refolded TCR was dialysed in Spectrapor 1 membrane (Spectrum; Product No. 132670) against 10 L H₂0 for 18-20 hours. After this time, the dialysis buffer was changed twice to fresh 10 mM Tris pH 8.1 (10 L) and dialysis was continued at 5 °C± 3 °Cfor another ~8 hours.

Soluble TCR was separated from degradation products and impurities by loading the dialysed refold onto a POROS 50HQ anion exchange column and eluting bound protein with a gradient of 0-500mM NaCI in 10 mM Tris pH 8.1 over 15 column volumes using an Akta purifier (GE Healthcare). The pooled fractions were then stored at 4 °Cand analysed by Coomassie-stained SDS-PAGE before being pooled and concentrated. Finally, the soluble TCRs were purified and characterised using a GE Healthcare Superdex 75HR gel filtration column pre-equilibrated in PBS buffer (Sigma). The peak eluting at a relative molecular weight of approximately 50 kDa was pooled and concentrated prior to characterisation by BIAcore surface plasmon resonance analysis.

The affinity profiles of the thus-prepared TCRs for the MAGE-3 epitope were assessed using the method of Example 3, and compared with the reference TCR. The results are set forth in the following table:

The affinity profiles of the a3b3 and a24b53 TCRs for the MAGE-6 EVDPIGHVY HLA-A1 and MAGE-B18 EVDPIRHYY HLA-A1 complexes were also assessed, and those TCRs were also shown to bind the MAGE-6 epitope with similar affinity and kinetics compared to the MAGE-3 epitope, but with weaker affinity to MAGE-B18.

Example 5

Expression, refolding and purification of soluble anti-CD3 scFv-MAGE-3 TCR fusion

SEQ ID No: 20 (Figure 14) is the amino acid sequence of an anti-CD3 scFv antibody fragment (bold type) fused via a linker namely GGGGS (underlined) at the N- terminus of a MAGE-3 TCR β chain. The MAGE-3 TCR β chain is SEQ ID No: 12.

The construct above was prepared as follows:

Ligation

Synthetic genes encoding (a) the TCR Voc chain SEQ ID No: 7 and (b) the fusion sequence SEQ ID No: 20, were separately ligated into pGMT7 + Coc vector and pGMT7-based expression plasmid respectively, which contain the T7 promoter for high level expression in E.coli strain BL21 -DE3(pLysS) (Pan et al., Biotechniques (2000) 29 (6): 1234-8).

Expression

The expression plasmids were transformed separately into E.coli strain BL21 (DE3) Rosetta pLysS, and single ampicillin-resistant colonies were grown at 37°Cin TYP (ampicillin 100μg/ml) medium to OD₆₀o of -0.6-0.8 before inducing protein expression with 0.5mM IPTG. Cells were harvested three hours post-induction by centrifugation for 30 minutes at 4000rpm in a Beckman J-6B. Cell pellets were lysed with 25ml Bug Buster (NovaGen) in the presence of MgCI₂ and DNase. Inclusion body pellets were recovered by centrifugation for 30 minutes at 13000rpm in a Beckman J2-21 centrifuge. Three detergent washes were then carried out to remove cell debris and membrane components. Each time the inclusion body pellet was homogenised in a Triton buffer (50mM Tris-HCI pH 8.0, 0.5% Triton-X100, 200mM NaCI, 10mM NaEDTA,) before being pelleted by centrifugation for 15 minutes at 13000rpm in a Beckman J2-21 . Detergent and salt was then removed by a similar wash in the following buffer: 50mM Tris-HCI pH 8.0, 1 mM NaEDTA. Finally, the inclusion bodies were divided into 30 mg aliquots and frozen at -70 °C.

Refolding

Approximately 20mg of TCR a chain and 40mg of scFv-TCR β chain solubilised inclusion bodies were thawed from frozen stocks, diluted into 20ml of a guanidine solution (6 M Guanidine-hydrochloride, 50mM Tris HCI pH 8.1 , 100m NaCI, 10mM EDTA, 10mM DTT), and incubated in a 37°Cwater bath for 30min-1 hr to ensure complete chain de-naturation. The guanidine solution containing fully reduced and denatured TCR chains was then injected into 1 litre of the following refolding buffer: 100mM Tris pH 8.1 , 400mM L-Arginine, 2mM EDTA, 5M Urea. The redox couple (cysteamine hydrochloride and cystamine dihydrochloride (to final concentrations of 10mM and 2.5mM, respectively)) were added approximately 5 minutes before addition of the denatured TCR a and scFv-TCR β chains. The solution was left for ~30minutes. The refolded scFv-TCR was dialysed in dialysis tubing cellulose membrane (Sigma-Aldrich; Product No. D9402) against 10 L H₂0 for 18-20 hours. After this time, the dialysis buffer was changed twice to fresh 10 mM Tris pH 8.1 (10 L) and dialysis was continued at 5°C± 3 °Cfor another ~8 hours. Soluble and correctly folded scFv-TCR was separated from degradation products and impurities by a 3-step purification method as described below.

First purification step

The dialysed refold (in 10mM Tris pH8.1 ) was loaded onto a POROS 50HQ anion exchange column and the bound protein eluted with a gradient of 0-500mM NaCI over 6 column volumes using an Akta purifier (GE Healthcare). Peak fractions (eluting at a conductivity ~20mS/cm) were stored at 4°C. Peak fractions were analysed by Instant Blue Stain (Novexin) stained SDS-PAGE before being pooled.

Second purification step

Cation exchange purification

The anion exchange pooled fractions were buffer exchanged by dilution with 20mM MES pH6-6.5, depending on the pi of the scFv-TCR fusion. The soluble and correctly folded scFv-TCR was separated from degradation products and impurities by loading the diluted pooled fractions (in 20mM MES pH6-6.5) onto a POROS 50HS cation exchange column and eluting bound protein with a gradient of 0-500mM NaCI over 6 column volumes using an Akta purifier (GE Healthcare). Peak fractions (eluting at a conductivity ~10mS/cm) were stored at 4°C. Final purification step

Peak fractions from second purification step were analysed by Instant Blue Stain (Novexin) stained SDS-PAGE before being pooled. The pooled fractions were then concentrated for the final purification step, when the soluble scFv-TCR was purified and characterised using a Superdex S200 gel filtration column (GE Healthcare) pre- equilibrated in PBS buffer (Sigma). The peak eluting at a relative molecular weight of approximately 78 kDa was analysed by Instant Blue Stain (Novexin) stained SDS- PAGE before being pooled.

In a similar way, corresponding anti-CD3 scFv-MAGE-3 TCR fusion constructs were prepared wherein the MAGE-3 TCR β chain part of SEQ ID No: 20 is replaced by the beta chain SEQ ID Nos 9, 10, 1 1 and 13. In a similar way an anti-CD3 scFv-MAGE-3 TCR fusion construct was prepared in which the alpha chain is SEQ ID No: 6 and the beta chain part of SEQ ID No: 20 is replaced by the beta chain SEQ ID No 8.

Thus the full array of anti CD3 scFv-MAGE-3 fusion constructs prepared is summarised in the following Table:

The constructs prepared according to the above Example 5 may be varied by substituting the GGGGS linker sequence with an alternative linker sequence selected from GGGSG, GGSGG and GSGGG, and/or by substituting amino acid D1 of SEQ ID No: 20 by A, and/or by substituting amino acid K1 of SEQ ID Nos: 8-13 by A. EXAMPLE 6

Redirection of T cells by anti-CD3 scFv-MAGE-3 high affinity TCR fusions against the cell-lines 375 and HCT116 ELISPOT protocol

The following assays were carried out to demonstrate activation of cytotoxic T lymphocytes (CTLs) by anti-CD3 scFv- MAGE-3 TCR fusions (MAG-ICs 0-5) via specific peptide-MHC recognition. IFN-γ or GrB production, measured by the ELISPOT assay, was used as a read-out for CTL activation and enabled evaluation of the potency of the anti-CD3 scFv portion of the fusion proteins.

Reagents

Assay media: 10% FCS (Heat Inactivated, Sera Laboratories International, cat# EU- 000-FI), 88% RPMI 1640 (Invitrogen, cat# 42401018), 1 % glutamine (Invitrogen, cat# 25030024) and 1 % penicillin/streptomycin (Invitrogen, cat#15070063).

Wash buffer: 1 xPBS sachet (Sigma, cat# P3813), containing 0.05% Tween-20, made up in deionised water

PBS (Invitrogen, cat# 10010015)

Dilution Buffer: PBS and 10% FCS (Heat Inactivated)

The Human IFNy ELISPOT PVDF-Enzymatic kit (BD Biosciences, cat# 551849) and the Human GrB ELISPOT PVDF-Enzymatic kit (BD Biosciences, cat# 552572) contain all other reagents required (capture and detection antibodies, streptavidin- HRP and substrate solution as the Human IFN-γ or GrB PVDF ELISPOT 96 well plates)

Method

Target cell preparation

Target cells were characterised for MAGE antigen expression by quantitative RT- PCR using standard procedures and primers specific for the antigen. Positive target cells used in the assay were cell-lines shown to express MAGE; the A375 melanoma cell-line and the HCT1 16 colorectal cell-line are both HLA-A1 ⁺ and MAGE⁺. The negative target cells were HLA-Α cells that do not contain detectable MAGE antigen by qRT-PCR; the colorectal cell-line Colo205, and the normal primary cells CIL-1 and HEP2 are all HLA-ΑΓ MAGE^" (see Kim, K. H., J. S. Choi, et al. (2006). "Promoter hypomethylation and reactivation of MAGE-A1 and MAGE-3 genes in colorectal cancer cell lines and cancer tissues." World J Gastroenterol 12(35): 5651 - 7). Sufficient target cells (to allow for 50,000 cells/well in the assay) were washed by centrifugation three times at 1200 rpm for 5 min in a Megafuge 1 .0 (Heraeus). Cells were then re-suspended in assay media at a density of 10⁶ cells/ml. Effector Cell Preparation

The effector cells (T cells) used in this method was either peripheral blood mononuclear cells (PBMC) or CD8⁺ T cells. PBMCs were isolated from blood using standard procedures utilising Lymphoprep (Axis-Shields, cat# NYC-1 1 14547) and Leucosep tubes (Greiner, cat# 227290). CD8⁺T cells were obtained by negative selection (using the CD8 Negative Isolation Kit, Invitrogen, cat# 1 13.48D) from

PBMCs. Effector cells were defrosted and placed in assay media prior to washing by centrifugation at 1300 rpm for 10 min in a Megafuge 1 .0 (Heraeus). Cells were then re-suspended in assay media at 4x the final required concentration. Reagent/Test Compound Preparation

Varying concentrations of the anti-CD3 scFv- MAGE-3 TCR fusion proteins (from 10 nM to 0.1 pM) were prepared by dilution into assay media to give 4x the final concentration. ELISPOTs

Plates were prepared as follows: 100 μΙ anti-IFN-γ or anti-GrB capture antibody was diluted in 10 ml sterile PBS per plate. 100 μΙ of the diluted capture antibody was then dispensed into each well. The plates were then incubated overnight at 4°C.

Following incubation the plates were washed (programme 1 , plate type 2, Ultrawash Plus 96-well plate washer; Dynex) to remove the capture antibody. Plates were then blocked by adding 200 μΙ of assay media to each well and incubated at room temperature for two hours. The assay media was then washed from the plates (programme 1 , plate type 2, Ultrawash Plus 96-well plate washer, Dynex) and any remaining media was removed by flicking and tapping the ELISPOT plates on a paper towel.

The constituents of the assay were then added to the ELISPOT plate in the following order: 50 μΙ of target cells 10⁶ cells/ml (giving a total of 50,000 target cells/well)

50 μΙ of reagent (the anti-CD3 scFv-TCR fusions; varying concentrations)

50 μΙ media (assay media) 50 μΙ effector cells (20,000 or 30,000 PBMC cells/well; 2000 CD8⁺ cells/well).

The plates were then incubated overnight (37°C / 5%C0₂). The next day the plates were washed three times (programme 1 , plate type 2, Ultrawash Plus 96-well plate washer, Dynex) with wash buffer and tapped dry on paper towel to remove excess wash buffer. 100 μΙ of primary detection antibody was then added to each well. The primary detection antibody was diluted into 10ml of dilution buffer (the volume required for a single plate) using the dilution specified in the manufacturer's instructions. Plates were then incubated at room temperature for at least 2 hours prior to being washed three times (programme 1 , plate type 2, Ultrawash Plus 96-well plate washer, Dynex) with wash buffer; excess wash buffer was removed by tapping the plate on a paper towel.

Secondary detection was performed by adding 100 μΙ of diluted streptavidin-HRP to each well and incubating the plate at room temperature for 1 hour. The streptavidin- HRP was diluted into 10ml dilution buffer (the volume required for a single plate), using the dilution specified in the manufacturer's instructions. The plates were then washed three times (programme 1 , plate type 2, Ultrawash Plus 96-well plate washer, Dynex) with wash buffer and tapped on paper towel to remove excess wash buffer. Plates were then washed twice with PBS by adding 200 μΙ to each well, flicking the buffer off and tapping on a paper towel to remove excess buffer. No more than 15 mins prior to use, one drop (20 ul) of AEC chromogen was added to each 1 ml of AEC substrate and mixed. 10ml of this solution was prepared for each plate; 100 μΙ was added per well. The plate was then protected from light using foil, and spot development monitored regularly, usually occurring within 5 - 20 mins. The plates were washed in tap water to terminate the development reaction, and shaken dry prior to their disassembly into three constituent parts. The plates were then allowed to dry at room temperature for at least 2 hours prior to counting the spots using an Immunospot Plate reader (CTL; Cellular Technology Limited).

RESULTS

The anti-CD3 scFv- MAGE-3 TCR fusions were tested by ELISPOT Assay (as described above). The number of ELISPOT spots observed in each well was plotted against the concentration of the fusion construct using Prism (Graph Pad) (see Figure 1 ). From these dose-response curves, the EC₅₀ values were determined (EC₅₀ are determined at the concentration of anti-CD3 scFv- MAGE-3 TCR fusion that induces 50% of the maximum response).

The graphs in Figure 15 show the specific activation of T cells by the different anti- CD3 scFv- MAGE-3 high affinity TCRs in the presence of MAGE-3 presenting cells (A375 or HCT1 16 cells). The data is representative of at least two separate assays in each case, except for (e) which was performed once in this assay format.

The data from Figure 15 yields the following EC₅₀ values; 21 pM for MAG-ICO, 14pM for MAG-IC1 , 83pM for MAG-IC2, 83pM for MAG-IC3, 63pM for MAG-IC4 and 238pM for MAG-IC5.

Example 7

Preparation of further TCRs of the invention

Method

Expression plasmids encoding single point and multiple mutants of the MAGE-A3 TCR alpha and beta chains were produced using standard molecular biology techniques. Plasmids were transformed into Rosetta DE3 chemically-competent cells and grown overnight at 37°C. Protein expression was induced by the addition of Isopropyl β-D-l -thiogalactopyranoside (IPTG) to 1 mM and bacteria were grown for a further 3 hours at 37°C. Bacteria were harvested by centrifugation at 4000 x g for 15 minutes and lysed in a protein extraction reagent containing DNAse. Lysis proceeded for 1 hour at room temperature with agitation before inclusion bodies were harvested by centrifugation at 10000 x g for 5 minutes. Pellets were washed twice with a detergent buffer containing 1 % Triton X100 and resuspended in a buffered saline solution.

Soluble TCRs were prepared by dissolving alpha and beta inclusion bodies in 6M guanidine-HCI containing 10 mM dithiothreitol and incubating at 37°Cfor 30 minutes. Samples were diluted into 50 ml urea folding buffer (5 M urea; 0.4 M L-arginine; 0.1 M Tris-CI, pH 8.1 ; 2 mM EDTA; 6.5 mM β-mercapthoethylamine; 1 .9 mM cystamine) and dialysed against eight volumes of water overnight at 4°C, followed by dialysis for a further 24 hours in eight volumes of 10 mM Tris (8.1 ), with one buffer change. Dialysate (30 ml) was concentrated to 1 ml using a vivaspin 15 concentrator with a 10 kDa filter. Concentrated protein was diluted to 5 ml in phosphate-buffered saline and concentrated to 0.5 ml. The purity of each mutant was analysed by gel electrophoresis using non-reduced SDS-polyacrylamide gels. Purity levels were comparable for all mutants.

Binding of mutants to MAGE-A3 peptide:HLA was analysed by surface plasmon resonance using a Biacore 3000. Concentrated samples were flowed (20 μΙ/min) over a surface containing immobilised MAGE-A3 peptide:HLA, with an irrelevant peptide:HLA complex used as a background subtraction. The measured level of binding in response units (RU) was divided by the concentration of protein, calculated by absorbance at 280 nM. The concentration-adjusted level of binding for each mutant was compared to that of the wild-type protein, which had been prepared at the same time under identical conditions. Mutants that gave a higher level of binding than the wild-type protein were assumed to have an increased affinity.

Results

The results are shown in the attached Figures as follows:

• Figure 16 - single point mutations in the CDR2 of the a chain

• Figure 17 - single point mutations in the CDR3 of the a chain

• Figure 18 - single point mutations in the CDR1 of the β chain

• Figure 19 - single point mutations in the CDR2 of the β chain

· Figure 20 - single point mutations in the CDR3 of the β chain

The following table summarises the single point mutations that resulted in a higher level of binding than the wild-type protein

In addition, mutations in the CDRs of the a and β chains were found that had no effect on the binding of the TCR. These mutations are summarised in the following table, and one or more can be included in TCRs of the present invention. Beta chain

TCRs were produced with combinations of different mutations which result in increased binding/higher affinity, and combinations of mutations which result in increased binding/higher affinity together with mutations which have no effect on binding/affinity. The following tables summarise the mutants that were tested: Alpha chain

The results obtained for these multiple mutants are shown in Figures follows

• Figure 21 - multiple mutations in the CDR2 of the a chain

• Figure 22 - multiple mutations in the CDR2 of the β chain

• Figure 23 - multiple mutations in the CDR3 of the β chain

Claims

1 . A T cell receptor (TCR) having the property of binding to EVDPIGHLY (SEQ ID No: 1 ) HLA-A1 complex and comprising a TCR alpha chain variable domain and a TCR beta chain variable domain,

with the proviso that the TCR is not a TCR comprising an alpha variable domain which has the amino acid sequence from K1 to P1 14 of SEQ ID NO: 2 except that at least one of the following mutations is present: 50V, 51 R, 52P and 53Y and/or a beta variable domain which has the amino acid sequence from K1 to T1 12 of SEQ ID NO: 3 except that at least one of the following mutations is present: 50F, 51 D, 52M, 53L and 54L.

2. A TCR as claimed in claim 1 , wherein

in the alpha chain variable domain

(ii) amino acid residues 49-53 are as follows:

or another functionally silent amino acid residue;

(iv) amino acid residues 95-98 are as follows:

or another functionally silent amino acid residue;

(iv) the sequence of amino acid residues 104-1 14 thereof is the same as the sequence of amino acid residues 104-1 14 of SEQ ID No: 2 or has one, two or three insertions, deletions or substitutions relative thereto, and in the beta chain variable domain

(ii) amino acid residues 50-54 are:

or another functionally silent amino acid residue;

(iv) amino acid residues 94-103 are as follows:

or another functionally silent amino acid residue;

(iv) the sequence of amino acid residues 104-1 12 thereof is the same as the sequence of amino acid residues 104-1 12 of SEQ ID No: 3 or has one, two or three amino acid residues inserted, deleted or substituted relative thereto,

Residue no.

50 V L F M

51 R K L T A V I M W

53 Y F W H A Q

97 T V S

100 F

101 A N

102 V

103 A K R and/or at least one of the following mutations is present in the beta chain variable domain:

3. A TCR as claimed in claim 1 or claim 2 wherein the alpha chain variable domain comprises K1 to P1 14 of SEQ ID No: 6; and the beta chain comprises K1 T1 12 of SEQ ID No: 8.

4. A TCR as claimed in claim 1 , 2 or 3 wherein the beta chain variable domain comprises K1 to P1 14 of SEQ ID No: 7; and the beta chain comprises K1 to T1 12 of SEQ ID No: 9 or of SEQ ID No: 10 or of SEQ ID No: 1 1 or of SEQ ID No: 12 or of SEQ ID No: 13.

5. A TCR as claimed in any one of claims 1 to 4 which is an alpha-beta heterodimer, having an alpha chain TRAC constant domain sequence and a beta chain TRBC1 or TRBC2 constant domain sequence.

6. A TCR as claimed in claim 5 wherein the alpha and beta chain constant domain sequences are modified by truncation or substitution to delete the native disulfide bond between Cys4 of exon 2 of TRAC and Cys2 of exon 2 of TRBC1 or TRBC2.

7. A TCR as claimed in claim 5 or claim 6 wherein the alpha and beta chain constant domain sequences are modified by substitution of cysteine residues for Thr 48 of TRAC and Ser 57 of TRBC1 or TRBC2, the said cysteines forming a disulfide bond between the alpha and beta constant domains of the TCR.

8. A TCR as claimed in any of claims 1 to 4 which is in single chain format of the type Voc-L-νβ, νβ-L-Voc, Voc-Coc-L-νβ, or Voc-L-Vp-Cp, wherein Va and νβ are TCR a and β variable regions respectively, Coc and C are TCR a and β constant regions respectively, and L is a linker sequence.

9. A TCR as claimed in any one of the preceding claims which is associated with a detectable label, a therapeutic agent or a PK modifying moiety.

10. A TCR as claimed in any one of claims 1 to 8 which is associated with an anti-CD3 antibody covalently linked to the C- or N-terminus of the alpha or beta chain of the TCR. 1 1 . A TCR as claimed in claim 10 which has alpha chain SEQ ID No: 6 or 7, and the anti CD3-TCR beta chain fusion SEQ ID No: 20, or the anti CD3-TCR beta chain fusion SEQ ID No. 20 in which the TCR beta chain sequence is replaced by SEQ ID No: 9, 10,

1 1 , or 13.

12. A TCR as claimed in claim 1 1 in which the GGGGS linker sequence is replaced by an alternative linker sequence selected from GGGSG, GGSGG and GSGGG, and/or by substituting amino acid A1 of SEQ ID No: 20 by D, and/or by substituting amino acid K1 of SEQ ID Nos: 8-13 by A.

13. Nucleic acid encoding a TCR as claimed in any one of the preceding claims.

14. An isolated or non-naturally occurring cell, especially a T-cell, presenting a TCR as claimed in any one of claims 1 to 12.

15. A cell harbouring

(a) a TCR expression vector which comprises nucleic acid as claimed in claim 12 in a single open reading frame, or two distinct open reading frames encoding the alpha chain and the beta chain respectively; or

(b) a first expression vector which comprises nucleic acid encoding the alpha chain of a TCR as claimed in any of claims 1 to 12, and a second expression vector which comprises nucleic acid encoding the beta chain of a TCR as claimed in any of claims 1 to 12,

16. A pharmaceutical composition comprising a TCR as claimed in any one of claims 1 to 12 or a cell as claimed in claim 14 or claim 15, together with one or more pharmaceutically acceptable carriers or excipients.

17. A TCR which binds the EVDPIGHLY peptide (derived from the MAGE-3 protein) presented as a peptide- HLA-A1 complex, or a cell expressing and/or presenting such a TCR, for use in medicine

18. The TCR or cell for use of claim 17, for use in a method of treating cancer.

19. The TCR or cell for use of claim 17 or claim 18, wherein the TCR is as claimed in any one of claims 1 to 12 and/or wherein the cell is as claimed in claim 14 or claim 15.