HK1074451B - Therapeutic epitopes and uses thereof - Google Patents

Description

The invention relates to epitopes useful in the diagnosis and therapy of coeliac disease, including diagnostics, therapeutics, kits, and methods of using the foregoing.

An immune reaction to gliadin (a component of gluten) in the diet causes coeliac disease. It is known that immune responses in the intestinal tissue preferentially respond to gliadin which has been modified by an intestinal transglutaminase. Coeliac disease is diagnosed by detection of anti-endomysial antibodies, but this requires conformation by the finding of a lymphocytic inflammation in intestinal biopsies. The taking of such a biopsy is inconvenient for the patient.

Investigators have previously assumed that only intestinal T cell responses provide an accurate indication of the immune response against gliadins. Therefore they have concentrated on the investigation of T cell responses in intestinal tissue¹. Gliadin epitopes which require transglutaminase modification (before they are recognised by the immune system) are known².

The inventors have found the immunodominant T cell A-gliadin epitope recognised by the immune system in coeliac disease, and have shown that this is recognised by T cells in the peripheral blood of individuals with coeliac disease (see WO 01/25793 ). Such T cells were found to be present at high enough frequencies to be detectable without restimulation (i.e. a 'fresh response' detection system could be used). The epitope was identified using a non-T cell cloning based method which provided a more accurate reflection of the epitopes being recognised. The immunodominant epitope requires transglutaminase modification (causing substitution of a particular glutamine to glutamate) before immune system recognition.

Based on this work the inventors have developed a test which can be used to diagnose coeliac disease at an early stage. The test may be carried out on a sample from peripheral blood and therefore an intestinal biopsy is not required. The test is more sensitive than the antibody tests which are currently being used.

Through comprehensive mapping of wheat gliadin T cell epitopes (see Example 13), the inventors have also found epitopes bioactive in coeliac disease in HLA-DQ2+ patients in other wheat gliadins, having similar core sequences (e.g., SEQ ID NOS:18-22) and similar full length sequences (e.g., SEQ ID NOS:31-36), as well as in rye secalins and barley hordeins (e.g., SEQ ID NOS:39-41); see also Tables 20 and 21. Additionally, several epitopes bioactive in coeliac disease in HLA-DQ8+ patients have been identified (e.g., SEQ ID NOS:42-44,46). This comprehensive mapping thus provides the dominant epitopes recognized by T cells in coeliac patients. The methods of the invention described herein may be performed using the additional identified epitope comprising the sequence of SEQ ID NO:19 or a sequence obtainable by transglutaminase-deamidating the sequence of SED ID NO:19. That is, the agents of the invention involve this novel epitope.

The invention also provides use of the agent for the preparation of a diagnostic means for use in a method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual, said method comprising determining whether T cells of the individual recognise the agent, recognition by the T cells indicating that the individual has, or is susceptible to, coeliac disease.

The finding of an immunodominant epitope which is modified by transglutaminase (as well as the additional other epitopes defined herein) also allows diagnosis of coeliac disease based on determining whether other types of immune response to this epitope are present. Thus the invention also provides a method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual comprising determining the presence of an antibody that binds to the epitope in a sample from the individual, the presence of the antibody indicating that the individual has, or is susceptible to, coeliac disease.

The invention additionally provides the agent, optionally in association with a carrier, for use in a method of treating or preventing coeliac disease by tolerising T cells which recognise the agent. Additionally provided is the agent for use in a method of treating or preventing coeliac disease in an individual by tolerising the individual to prevent the production of such an antibody.

The invention provides a method of determining whether a composition is capable of causing coeliac disease comprising determining whether a protein capable of being modified by a transglutaminase to an oligopeptide sequence as defined above is present in the composition, the presence of the protein indicating that the composition is capable of causing coeliac disease.

SUMMARY OF THE INVENTION

The present invention provides an isolated peptide comprising at least one T cell epitope, wherein the peptide is not more than 50 amino acids in length and the T cell epitope comprises a sequence selected from FPQPQQPFP and a sequence obtainable by transglutaminase-deamidating the sequence FPQPQQPFP (the "peptide of the invention").

The present invention also provides a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, for use in a method of therapy.

The present invention also provides a composition comprising (a) a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, and (b) a second agent which is a peptide comprising at least one epitope comprising a sequence selected from PQPELPY and QLQPFPQPELPYPQPQS.

The present invention also provides a composition comprising a first HLA-DQ2-restricted agent and a second HLA-DQ8-restricted agent, wherein the composition comprises a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.

The present invention also provides a composition comprising an agent comprising a wheat epitope and an agent comprising a rye epitope, wherein the composition comprises a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.

The present invention also provides a composition comprising an agent comprising a wheat epitope and an agent comprising a barley epitope, wherein the composition comprises a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.

The present invention also provides a composition comprising an agent comprising a rye epitope and an agent comprising a barley epitope, wherein the composition comprises a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.

The present invention also provides a composition comprising an agent comprising a wheat epitope, an agent comprising a barley epitope, and an agent comprising a rye epitope, wherein the composition comprises a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.

The present invention still further provides a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, for use in tolerising an individual to a gliadin protein to suppress the production of a T cell response to the said peptide.

The present invention also provides a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, for use in treating or preventing coeliac disease.

Also provided by the present invention is a pharmaceutical composition comprising: (a) a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP; (b) a pharmaceutically acceptable carrier or diluent; and, optionally, (c) a peptide comprising at least one epitope comprising a sequence selected from PQPELPY and QLQPFPQPELPYPQPQS.

The present invention provides a method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual comprising:

1. (a) contacting a sample from the host with at least one peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, and optionally, in addition, a peptide comprising at least one epitope comprising a sequence selected from PQPELPY and QLQPFPQPELPYPQPQS;
2. (b) determining in vitro whether T cells in the sample recognise the peptide; recognition by the T cells indicating that the individual has, or is susceptible to, coeliac disease.

The present invention further provides the use of a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, for the preparation of a diagnostic means for use in a method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual, said method comprising determining whether T cells of the individual recognise the peptide, recognition by the T cells indicating that the individual has, or is susceptible to, coeliac disease.

The invention additionally provides a method for identifying an analogue, which analogue is a peptide of the invention, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, bound to an HLA molecule or a fragment of an HLA molecule capable of binding the peptide, which method comprises determining whether a candidate substance is recognised by a T cell receptor that recognises an epitope comprising a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, recognition of the substance indicating that the substance is an analogue.

The invention also provides a method of determining whether a composition is capable of causing coeliac disease comprising determining whether a protein comprising an oligopeptide sequence selected from FPQPQQPFP, QQPFPQPQQPFP and QQPFPQPQQPFP is present in the composition, the presence of the protein indicating that the composition is capable of causing coeliac disease.

Still further, the invention provides an antibody which specifically binds to a sequence selected from FPQPQQPFP,QQPFPQPQQPFP and QPFPQPQQPFPWQP.

Also provided by the invention is a kit for carrying out:

1. (i) the method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual of the invention; or
2. (ii) the use for the preparation of a diagnostic means of the invention;

the kit comprising a peptide of the invention wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, and a means to detect the recognition of the peptide by the T cell.

The invention also provides the use of a peptide of the invention to produce an antibody specific to the peptide.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is illustrated by the accompanying drawings in which:

• Figure 1 shows freshly isolated PBMC (peripheral blood mononuclear cell) IFNγ ELISPOT responses (vertical axis shows spot forming cells per 106 PBMC) to transglutaminase (tTG)-treated and untreated peptide pool 3 (each peptide 10 µg/ml) including five overlapping 15mers spanning A-gliadin 51-85 (see Table 1) and a-chymotrypsin-digested gliadin (40 µg/ml) in coeliac disease Subject 1, initially in remission following a gluten free diet then challenged with 200g bread daily for three days from day 1 (a). PBMC IFNγ ELISPOT responses by Subject 2 to tTG-treated A-gliadin peptide pools 1-10 spanning the complete A-gliadin protein during ten day bread challenge (b). The horizontal axis shows days after commencing bread.
• Figure 2 shows PBMC IFNγ ELISPOT responses to tTG-treated peptide pool 3 (spanning A-gliadin 51-85) in 7 individual coeliac disease subjects (vertical axis shows spot forming cells per 106 PBMC), initially in remission on gluten free diet, challenged with bread for three days (days 1 to 3). The horizontal axis shows days after commencing bread. (a). PBMC IFNγ Elispot responses to tTG-treated overlapping 15mer peptides included in pool 3; bars represent the mean (± SEM) response to individual peptides (10 µg/ml) in 6 Coeliac disease subjects on day 6 or 7(b). (In individual subjects, ELISPOT responses to peptides were calculated as a % of response elicited by peptide 12 - as shown by the vertical axis.)
• Figure 3 shows PBMC IFNγ ELISPOT responses to tTG-treated truncations of A-gliadin 56-75 (0.1 µM). Bars represent the mean (± SEM) in 5 Coeliac disease subjects. (In individual subjects, responses were calculated as the % of the maximal response elicited by any of the peptides tested.)
• Figure 4 shows how the minimal structure of the dominant A-gliadin epitope was mapped using tTG-treated 7-17mer A-gliadin peptides (0.1 µM) including the sequence, PQPQLPY (SEQ ID NO:4) (A-gliadin 62-68) (a), and the same peptides without tTG treatment but with the substitution Q→E65 (b). Each line represents PBMC IFNγ ELISPOT responses in each of three Coeliac disease subjects on day 6 or 7 after bread was ingested on days 1-3. (In individual subjects, ELISPOT responses were calculated as a % of the response elicited by the 17mer, A-gliadin 57-73.)
• Figure 5 shows the amino acids that were deamidated by tTG. A-gliadin 56-75 LQLQPFPQPQLPYPQPQSFP (SEQ ID NO:5) (0.1 µM) was incubated with tTG (50 µg/ml) at 37°C for 2 hours. A single product was identified and purified by reverse phase HPLC. Amino acid analysis allowed % deamidation (Q→E) of each Gln residue in A-gliadin 56-75 attributable to tTG to be calculated (vertical axis).
• Figure 6 shows the effect of substituting Q→E in A-gliadin 57-73 at other positions in addition to Q65 using the 17mers: ELQPFPQPELPYPQPQS (SEQ ID NO:6) (E57,65), QLQPFPQPELPYPQPES (SEQ ID NO:7) (E65,72), ELQPFPQPELPYPQPES (SEQ ID NO:8) (E57, 65, 72), and QLQPFPQPELPYPQPQS (SEQ ID NO:2) (E65) in three Coeliac disease subjects on day 6 or 7 after bread was ingested on days 1-3. Vertical axis shows % of the E65 response.
• Figure 7 shows that tTG treated A-gliadin 56-75 (0.1 µM) elicited IFN-g ELISPOT responses in (a) CD4 and CD8 magnetic bead depleted PBMC. (Bars represent CD4 depleted PBMC responses as a % of CD8 depleted PBMC responses; spot forming cells per million CD8 depleted PBMC were: Subject 4: 29, and Subject 6: 535). (b) PBMC IFNγ ELISPOT responses (spot forming cells/million PBMC) after incubation with monoclonal antibodies to HLA-DR (L243), -DQ (L2) and -DP (B7.21) (10 µg/ml) 1h prior to tTG-treated 56-75 (0.1 µM) in two coeliac disease subjects homozygous for HLA-DQ a1*0501, b1*0201.
• Figure 8 shows the effect of substituting Glu at position 65 for other amino acids in the immunodominant epitope. The vertical axis shows the % response in the 3 subjects in relation to the immunodominant epitope.
• Figure 9 shows the immunoreactivity of naturally occurring gliadin peptides (measuring responses from 3 subjects) which contain the sequence PQLPY (SEQ ID NO:12) with (shaded) and without (clear) transglutaminase treatment.
• Figure 10 shows CD8, CD4, β7, and αE -specific immunomagnetic bead depletion of peripheral blood mononuclear cells from two coeliac subjects 6 days after commencing gluten challenge followed by interferon gamma ELISpot. A-gliadin 57-73 QE65 (25mcg/ml), tTG-treated chymotrypsin-digested gliadin (100 mcg/ml) or PPD (10 mcg/ml) were used as antigen.
• Figure 11 shows responses in different patient groups.
• Figure 12 shows bioactivity of prolamin homologues of A-gliadin 57-73.
• Figure 13 shows, for healthy HLA-DQ2 subjects, the change in IFN-gamma ELISpot responses to tTG-deamidated gliadin peptide pools.
• Figure 14 shows, for coeliac HLA-DQ2 subjects, the change in IFN-gamma ELISpot responses to tTG-deamidated gliadin peptide pools.
• Figure 15 shows individual peptide contributions to "summed" gliadin peptide response.
• Figure 16 shows, for coeliac HLA-DQ2/8 subject C08, gluten challenge induced IFNγ ELISpot responses to tTG-deamidated gliadin peptide pools.
• Figure 17 shows, for coeliac HLA-DQ2/8 subject C07, gluten challenge induced IFNγ ELISpot responses to tTG-deamidated gliadin peptide pools.
• Figure 18 shows, for coeliac HLA-DQ8/7 subject C12, gluten challenge induced IFNγ ELISpot responses to tTG-deamidated gliadin peptide pools.
• Figure 19 shows, for coeliac HLA-DQ6/8 subject C11, gluten challenge induced IFNγ ELISpot responses to tTG-deamidated gliadin peptide pools.

Detailed Description of the Invention

The term "coeliac disease" encompasses a spectrum of conditions caused by varying degrees of gluten sensitivity, including a severe form characterised by a flat small intestinal mucosa (hyperplastic villous atrophy) and other forms characterised by milder symptoms.

The individual mentioned above (in the context of diagnosis or therapy) is human. They may have coeliac disease (symptomatic or asymptomatic) or be suspected of having it. They may be on a gluten free diet. They may be in an acute phase response (for example they may have coeliac disease, but have only ingested gluten in the last 24 hours before which they had been on a gluten free diet for 14 to 28 days).

The individual may be susceptible to coeliac disease, such as a genetic susceptibility (determined for example by the individual having relatives with coeliac disease or possessing genes which cause predisposition to coeliac disease).

The agent

The agent is a peptide of length up to 50 amino acids, such as 10 to 40, or 15 to 30 amino acids in length.

SEQ ID NO:1 is PQPELPY. SEQ ID NO:2 is QLQPFPQPELPYPQPQS. SEQ ID NO:3 is shown in Table 1 and is the sequence of a whole A-gliadin. The glutamate at position 4 of SEQ ID NO:1 (equivalent to position 9 of SEQ ID NO:2) is generated by transglutaminase treatment of A-gliadin. Transglutaminase is commercially available (e.g. Sigma T-5398).

An analogue of a peptide of the invention is capable of being recognised by a TCR which recognises the peptide of the invention.Therefore generally when the analogue is added to T cells in the presence of the peptide of the invention, typically also in the presence of an antigen presenting cell (APC) (such as any of the APCs mentioned herein), the analogue inhibits the recognition of the peptide of the invention, i.e. the analogue is able to compete with the peptide of the invention in such a system.

The analogue may be one which is capable of binding the TCR which recognises the peptide of the invention. Such binding can be tested by standard techniques. Such TCRs can be isolated from T cells which have been shown to recognise the peptide of the invention (e.g. using the method of the invention). Demonstration of the binding of the analogue to the TCRs can then shown by determining whether the TCRs inhibit the binding of the analogue to a substance that binds the analogue, e.g. an antibody to the analogue. Typically the analogue is bound to a class II MHC molecule (e.g. HLA-DQ2) in such an inhibition of binding assay.

Typically the analogue inhibits the binding of the peptide of the invention to a TCR. In this case the amount of the peptide of the invention which can bind the TCR in the presence of the analogue is decreased. This is because the analogue is able to bind the TCR and therefore competes with the peptide of the invention for binding to the TCR. T cells for use in the above binding experiments can be isolated from patients with coeliac disease, for example with the aid of the method of the invention. Other binding characteristics of the analogue may also be the same as the peptide of the invention, and thus typically the analogue binds to the same MHC class II molecule to which the peptide binds (HLA-DQ2 or -DQ8). The analogue typically binds to antibodies specific for the peptide of the invention, and thus inhibits binding of the peptide of the invention to such antibodies.

The analogue is typically a peptide. It may have homology with the peptide of the invention, typically at least 70% homology, preferably at least 80, 90%, 95%, 97% or 99% homology with the peptide of the invention, for example over a region of at least 15 more (such as the entire length of the analogue and/or the peptide of the invention, or across the region which contacts the TCR or binds the MHC molecule) contiguous amino acids. Methods of measuring protein homology are well known in the art and it will be understood by those of skill in the art that in the present context, homology is calculated on the basis of amino acid identity (sometimes referred to as "hard homology").

For example the UWGCG Package provides the BESTFIT program which can be used to calculate homology (for example used on its default settings) (Devereux et al (1984) Nucleic Acids Research 12, p387-395). The PILEUP and BLAST algorithms can be used to calculate homology or line up sequences (typically on their default settings), for example as described in Altschul S. F. (1993) J Mol Evol 36:290-300; Altschul, S, F et al (1990) J Mol Biol 215:403-10.

Software for performing BLAST analyses is publicly available through the National Center for Biotechnology Information on the world wide web through the internet at, for example, "www.ncbi.nlm.nih.gov/". This algorithm involves first identifying high scoring sequence pair (HSPs) by identifying short words of length W in the query sequence that either match or satisfy some positive-valued threshold score T when aligned with a word of the same length in a database sequence. T is referred to as the neighbourhood word score threshold (Altschul et al, supra). These initial neighbourhood word hits act as seeds for initiating searches to find HSPs containing them. The word hits are extended in both directions along each sequence for as far as the cumulative alignment score can be increased. Extensions for the word hits in each direction are halted when: the cumulative alignment score falls off by the quantity X from its maximum achieved value; the cumulative score goes to zero or below, due to the accumulation of one or more negative-scoring residue alignments; or the end of either sequence is reached. The BLAST algorithm parameters W, T and X determine the sensitivity and speed of the alignment. The BLAST program uses as defaults a word length (W) of 11, the BLOSUM62 scoring matrix (see Henikoff and Henikoff (1992) Proc. Natl. Acad. Sci. USA 89: 10915-10919) alignments (B) of 50, expectation (E) of 10, M=5, N=4, and a comparison of both strands.

The BLAST algorithm performs a statistical analysis of the similarity between two sequences; see e.g., Karlin and Altschul (1993) Proc. Natl. Acad. Sci. USA 90: 5873-5787. One measure of similarity provided by the BLAST algorithm is the smallest sum probability (P(N)), which provides an indication of the probability by which a match between two nucleotide or amino acid sequences would occur by chance. For example, a sequence is considered similar to another sequence if the smallest sum probability in comparison of the first sequence to the second sequence is less than about 1, preferably less than about 0.1, more preferably less than about 0.01, and most preferably less than about 0.001.

The homologous peptide analogues typically differ from the peptide of the invention by 1, 2, 3, 4, 5, 6, 7, 8 or more mutations (which may be substitutions, deletions or insertions). These mutations may be measured across any of the regions mentioned above in relation to calculating homology. The substitutions are preferably 'conservative'. These are defined according to the following Table. Amino acids in the same block in the second column and preferably in the same line in the third column may be substituted for each other:

ALIPHATIC	Non-polar	G A P
		I L V
	Polar - uncharged	C S T M
		N Q
	Polar - charged	D E
		K R
AROMATIC	H F W Y

Typically the amino acids in the analogue at the equivalent positions to amino acids in the peptide of the invention that contribute to binding the MHC molecule or are responsible for the recognition by the TCR, are the same or are conserved.

Typically the analogue peptide comprises one or more modifications, which may be natural post-translation modifications or artificial modifications. The modification may provide a chemical moiety (typically by substitution of a hydrogen, e.g. of a C-H bond), such as an amino, acetyl, hydroxy or halogen (e.g. fluorine) group or carbohydrate group. Typically the modification is present on the N or C terminus.

The analogue may comprise one or more non-natural amino acids, for example amino acids with a side chain different from natural amino acids. Generally, the non-natural amino acid will have an N terminus and/or a C terminus. The non-natural amino acid may be an L- or a D- amino acid.

The analogue typically has a shape, size, flexibility or electronic configuration that is substantially similar to the peptide of the invention. It is typically a derivative of the peptide of the invention. In one embodiment the analogue is a fusion protein comprising the sequence of SEQ ID NO: 1 or 2, or any of the other peptides mentioned herein; and non-gliadin sequence.

In one embodiment the analogue is or mimics the peptide of the invention bound to a MHC class II molecule. 2,3,4 or more of such complexes may be associated or bound to each other, for example using a biotin/streptavidin based system, in which typically 2, 3 or 4 biotin labelled MHC molecules bind to a streptavidin moiety. This analogue typically inhibits the binding of the peptide of the invention /MHC Class II complex to a TCR or antibody which is specific for the complex.

The analogue is typically an antibody or a fragment of an antibody, such as a Fab or (Fab)₂ fragment. The analogue may be immobilised on a solid support, particularly an analogue that mimics peptide bound to a MHC molecule.

The analogue is typically designed by computational means and then synthesised using methods known in the art. Alternatively the analogue can be selected from a library of compounds. The library may be a combinatorial library or a display library, such as a phage display library. The library of compounds may be expressed in the display library in the form of being bound to a MHC class II molecule, such as HLA-DQ2 or -DQ8. Analogues are generally selected from the library based on their ability to mimic the binding characteristics of the peptide of the invention. Thus they may be selected based on ability to bind a TCR or antibody which recognises the peptide of the invention.

Typically analogues will be recognised by T cells to at least the same extent as any of the agents the peptide of the invention, for example at least to the same extent as the equivalent epitope and preferably to the same extent as the peptide represented by SEQ ID NO:2, as recognised in any of the assays described herein, typically using T cells from coeliac disease patients. Analogues may be recognised to these extents in vivo and thus may be able to induce coeliac disease symptoms to at least the same extent as any of the agents mentioned herein (e.g. in a human patient or animal model).

Analogues may be identified in a method comprising determining whether a candidate substance is recognised by a T cell receptor that recognises an epitope of the invention, recognition of the substance indicating that the substance is an analogue. Such TCRs may be any of the TCRs mentioned herein, and may be present on T cells. Any suitable assay mentioned herein can be used to identify the analogue. In one embodiment this method is carried out in vivo. As mentioned above preferred analogues are recognised to at least the same extent as the peptide SEQ ID NO:2, and so the method may be used to identify analogues which are recognised to this extent.

In one embodiment the method comprises determining whether a candidate substance is able to inhibit the recognition of an epitope of the invention, inhibition of recognition indicating that the substance is an analogue.

The agent may be a product comprising at least 2, 5, 10 or 20 peptides of the invention. Typically the composition comprises epitopes from different gliadins, such as any of the species or variety of or types of gliadin mentioned herein. Compositions comprise at least one epitope of the invention. Preferred compositions comprise at least one epitope, or equivalent analogue, from all of the gliadins present in any of the species or variety mentioned herein, or from 2, 3, 4 or more of the species mentioned herein (such as from the panel of species consisting of wheat, rye, barley, oats and triticale). Thus, the agent may be monovalent or multivalent.

Diagnosis

The invention provides a method of diagnosis as defined herein. More generally, though, methods of diagnosis making use of the peptides of the invention may be based on the detection of T cells that bind the agent or on the detection of antibodies that recognise the agent.

The T cells that recognise the agent in the method (which includes the use mentioned above) are generally T cells that have been pre-sensitised in vivo to gliadin. As mentioned above such antigen-experienced T cells have been found to be present in the peripheral blood.

In the method the T cells can be contacted with the agent in vitro or in vivo, and determining whether the T cells recognise the agent can be performed in vitro or in vivo. Thus the invention provides the agent for use in a method of diagnosis practiced on the human body. Different agents are provided for simultaneous, separate or sequential use in such a method.

The in vitro method is typically carried out in aqueous solution into which the agent is added. The solution will also comprise the T cells (and in certain embodiments the APCs discussed below). The term 'contacting' as used herein includes adding the particular substance to the solution.

Determination of whether the T cells recognise the agent is generally accomplished by detecting a change in the state of the T cells in the presence of the agent or determining whether the T cells bind the agent. The change in state is generally caused by antigen specific functional activity of the T cell after the TCR binds the agent. The change of state may be measured inside (e.g. change in intracellular expression of proteins) or outside (e.g. detection of secreted substances) the T cells.

The change in state of the T cell may be the start of or increase in secretion of a substance from the T cell, such as a cytokine, especially IFN-γ, IL-2 or TNF-α. Determination of IFN-γ secretion is particularly preferred. The substance can typically be detected by allowing it to bind to a specific binding agent and then measuring the presence of the specific binding agent/substance complex. The specific binding agent is typically an antibody, such as polyclonal or monoclonal antibodies. Antibodies to cytokines are commercially available, or can be made using standard techniques.

Typically the specific binding agent is immobilised on a solid support. After the substance is allowed to bind the solid support can optionally be washed to remove material which is not specifically bound to the agent. The agent/substance complex may be detected by using a second binding agent that will bind the complex. Typically the second agent binds the substance at a site which is different from the site which binds the first agent. The second agent is preferably an antibody and is labelled directly or indirectly by a detectable label.

Thus the second agent may be detected by a third agent that is typically labelled directly or indirectly by a detectable label. For example the second agent may comprise a biotin moiety, allowing detection by a third agent which comprises a streptavidin moiety and typically alkaline phosphatase as a detectable label.

In one embodiment the detection system which is used is the ex-vivo ELISPOT assay described in WO 98/23960 . In that assay IFN-γ secreted from the T cell is bound by a first IFN-γ specific antibody that is immobilised on a solid support. The bound IFN-γ is then detected using a second IFN-γ specific antibody which is labelled with a detectable label. Such a labelled antibody can be obtained from MABTECH (Stockholm, Sweden). Other detectable labels which can be used are discussed below.

The change in state of the T cell that can be measured may be the increase in the uptake of substances by the T cell, such as the uptake of thymidine. The change in state may be an increase in the size of the T cells, or proliferation of the T cells, or a change in cell surface markers on the T cell.

In one embodiment the change of state is detected by measuring the change in the intracellular expression of proteins, for example the increase in intracellular expression of any of the cytokines mentioned above. Such intracellular changes may be detected by contacting the inside of the T cell with a moiety that binds the expressed proteins in a specific manner and which allows sorting of the T cells by flow cytometry.

In one embodiment when binding the TCR the agent is bound to an MHC class II molecule (typically HLA-DQ2 or -DQ8), which is typically present on the surface of an antigen presenting cell (APC). However as mentioned herein other agents can bind a TCR without the need to also bind an MHC molecule.

Generally the T cells which are contacted in the method are taken from the individual in a blood sample, although other types of samples which contain T cells can be used. The sample may be added directly to the assay or may be processed first. Typically the processing may comprise diluting of the sample, for example with water or buffer. Typically the sample is diluted from 1.5 to 100 fold, for example 2 to 50 or 5 to 10 fold.

The processing may comprise separation of components of the sample. Typically mononuclear cells (MCs) are separated from the samples. The MCs will comprise the T cells and APCs. Thus in the method the APCs present in the separated MCs can present the peptide to the T cells. In another embodiment only T cells, such as only CD4 T cells, can be purified from the sample. PBMCs, MCs and T cells can be separated from the sample using techniques known in the art, such as those described in Lalvani et al (1997) J. Exp. Med. 186, p859-865.

In one embodiment, the T cells used in the assay are in the form of unprocessed or diluted samples, or are freshly isolated T cells (such as in the form of freshly isolated MCs or PBMCs) which are used directly ex vivo, i.e. they are not cultured before being used in the method. Thus the T cells have not been restimulated in an antigen specific manner in vitro. However the T cells can be cultured before use, for example in the presence of one or more of the agents, and generally also exogenous growth promoting cytokines. During culturing the agent(s) are typically present on the surface of APCs, such as the APC used in the method. Pre-culturing of the T cells may lead to an increase in the sensitivity of the method. Thus the T cells can be converted into cell lines, such as short term cell lines (for example as described in Ota et al (1990) Nature 346, p183-187).

The APC that is typically present in the method may be from the same individual as the T cell or from a different host. The APC may be a naturally occurring APC or an artificial APC. The APC is a cell that is capable of presenting the peptide to a T cell. It is typically a B cell, dendritic cell or macrophage. It is typically separated from the same sample as the T cell and is typically co-purified with the T cell. Thus the APC may be present in MCs or PBMCs. The APC is typically a freshly isolated ex vivo cell or a cultured cell. It may be in the form of a cell line, such as a short term or immortalised cell line. The APC may express empty MHC class II molecules on its surface.

In the method one or more (different) agents may be used. Typically the T cells derived from the sample can be placed into an assay with all the agents which it is intended to test or the T cells can be divided and placed into separate assays each of which contain one or more of the agents.

The peptides of the invention, such as two or more of the agents mentioned herein (e.g. the combinations of agents which are present in the composition agent discussed above), can also be applied to simultaneous separate or sequential use (eg. to in vivo use).

In one embodiment agent per se is added directly to an assay comprising T cells and APCs. As discussed above the T cells and APCs in such an assay could be in the form of MCs. When agents that can be recognised by the T cell without the need for presentation by APCs are used then APCs are not required. Analogues which mimic the original peptide of the invention bound to a MHC molecule are an example of such an agent.

In one embodiment the agent is provided to the APC in the absence of the T cell. The APC is then provided to the T cell, typically after being allowed to present the agent on its surface. The peptide may have been taken up inside the APC and presented, or simply be taken up onto the surface without entering inside the APC.

The duration for which the agent is contacted with the T cells will vary depending on the method used for determining recognition of the peptide. Typically 10⁵ to 10⁷, preferably 5x10⁵ to 10⁶ PBMCs are added to each assay. In the case where agent is added directly to the assay its concentration is from 10^-1 to 10³µg/ml, preferably 0.5 to 50µg/ml or 1 to 10µg/ml.

Typically the length of time for which the T cells are incubated with the agent is from 4 to 24 hours, preferably 6 to 16 hours. When using ex vivo PBMCs it has been found that 0.3x10⁶ PBMCs can be incubated in 10µg/ml of peptide for 12 hours at 37°C.

The determination of the recognition of the agent by the T cells may be done by measuring the binding of the agent to the T cells (this can be carried out using any suitable binding assay format discussed herein). Typically T cells which bind the agent can be sorted based on this binding, for example using a FACS machine. The presence of T cells that recognise the agent will be deemed to occur if the frequency of cells sorted using the agent is above a "control" value. The frequency of antigen-experienced T cells is generally 1 in 10⁶ to 1 in 10³, and therefore whether or not the sorted cells are antigen-experienced T cells can be determined.

The determination of the recognition of the agent by the T cells may be measured in vivo. Typically the agent is administered to the host and then a response which indicates recognition of the agent may be measured. The agent is typically administered intradermally or epidermally. The agent is typically administered by contacting with the outside of the skin, and may be retained at the site with the aid of a plaster or dressing. Alternatively the agent may be administered by needle, such as by injection, but can also be administered by other methods such as ballistics (e.g. the ballistics techniques which have been used to deliver nucleic acids). EP-A-0693119 describes techniques that can typically be used to administer the agent. Typically from 0.001 to 1000 µg, for example from 0.01 to 100 µg or 0.1 to 10 µg of agent is administered.

In one embodiment a product can be administered which is capable of providing the agent in vivo. Thus a polynucleotide capable of expressing the agent can be administered, typically in any of the ways described above for the administration of the agent. The polynucleotide typically has any of the characteristics of the polynucleotide provided by the invention which is discussed below. The agent is expressed from the polynucleotide in vivo. Typically from 0.001 to 1000 µg, for example from 0.01 to 100 µg or 0.1 to 10 µg of polynucleotide is administered.

Recognition of the agent administered to the skin is typically indicated by the occurrence of inflammation (e.g. induration, erythema or oedema) at the site of administration. This is generally measured by visual examination of the site.

The method of diagnosis based on the detection of an antibody that binds the agent is typically carried out by contacting a sample from the individual (such as any of the samples mentioned here, optionally processed in any manner mentioned herein) with the agent and determining whether an antibody in the sample binds the agent, such a binding indicating that the individual has, or is susceptible to coeliac disease. Any suitable format of binding assay may be used, such as any such format mentioned herein.

Therapy

The identification of the immunodominant epitope and other epitopes described herein allows therapeutic products to be made which target the T cells which recognise this epitope (such T cells being ones which participate in the immune response against gliadin). These findings also allow the prevention or treatment of coeliac disease by suppressing (by tolerisation) an antibody or T cell response to the epitope(s).

Certain agents of the invention bind the TCR that recognises the epitope of the invention (as measured using any of the binding assays discussed above) and cause tolerisation of the T cell that carries the TCR. Such agents, optionally in association with a carrier, can therefore be used to prevent or treat coeliac disease.

Generally tolerisation can be caused by the same peptides which can (after being recognised by the TCR) cause antigen specific functional activity of the T cell (such as any such activity mentioned herein, e.g. secretion of cytokines). Such agents cause tolerisation when they are presented to the immune system in a 'tolerising' context.

Tolerisation leads to a decrease in the recognition of a T cell or antibody epitope by the immune system. In the case of a T cell epitope this can be caused by the deletion or anergising of T cells that recognise the epitope. Thus T cell activity (for example as measured in suitable assays mentioned herein) in response to the epitope is decreased. Tolerisation of an antibody response means that a decreased amount of specific antibody to the epitope is produced when the epitope is administered.

Methods of presenting antigens to the immune system in such a context are known and are described for example in Yoshida et al. Clin. Immunol. Immunopathol. 82,207-215 (1997), Thurau et al. Clin. Exp. Immunol. 109, 370-6 (1997), and Weiner et al. Res. Immunol. 148, 528-33 (1997). In particular certain routes of administration can cause tolerisation, such as oral, nasal or intraperitoneal. Tolerisation may also be accomplished via dendritic cells and tetramers presenting peptide. Particular products which cause tolerisation may be administered (e.g. in a composition that also comprises the agent) to the individual. Such products include cytokines, such as cytokines that favour a Th2 response (e.g. IL-4, TGF-β or IL-10). Products or agent may be administered at a dose that causes tolerisation.

The invention provides a protein that comprises a sequence able to act as an antagonist of the T cell (which T cell recognises the agent). Such proteins and such antagonists can also be used to prevent or treat coeliac disease. The antagonist will cause a decrease in the T cell response. In one embodiment, the antagonist binds the TCR of the T cell (generally in the form of a complex with HLA-DQ2 or -DQ8) but instead of causing normal functional activation causing an abnormal signal to be passed through the TCR intracellular signalling cascade, which causes the T cell to have decreased functional activity (e.g. in response to recognition of an epitope, typically as measured by any suitable assay mentioned herein).

In one embodiment the antagonist competes with epitope to bind a component of MHC processing and presentation pathway, such as an MHC molecule (typically HLA-DQ2 or -DQ8). Thus the antagonist may bind HLA-DQ2 or -DQ8 (and thus be a peptide presented by this MHC molecule).

Methods of causing antagonism are known in the art.

Since the T cell immune response to the epitope of the invention in an individual is polyclonal, more than one antagonist may need to be administered to cause antagonism of T cells of the response which have different TCRs. Therefore the antagonists may be administered in a composition which comprises at least 2, 4, 6 or more different antagonists, which each antagonise different T cells.

In one embodiment, the antagonists (including combinations of antagonists to a particular epitope) or tolerising (T cell and antibody tolerising) agents are present in a composition comprising at least 2, 4, 6 or more antagonists or agents which antagonise or tolerise to different epitopes of the invention, for example to the combinations of epitopes discussed above in relation to the agents which are a product comprising more than one substance.

Testing whether a composition is capable of causing coeliac disease

As mentioned above the invention provides a method of determining whether a composition is capable of causing coeliac disease comprising detecting the presence of a particular protein sequence (which is capable of being modified by a transglutaminase; such transglutaminase activity may be a human intestinal transglutaminase activity). Typically this is performed by using a binding assay in which a moiety which binds to the sequence in a specific manner is contacted with the composition and the formation of sequence/moiety complex is detected and used to ascertain the presence of the agent. Such a moiety may be any suitable substance (or type of substance) mentioned herein, and is typically a specific antibody. Any suitable format of binding assay can be used (such as those mentioned herein).

In one embodiment, the composition is contacted with at least 2, 5, 10 or more antibodies which are specific for epitopes from different gliadins, for example a panel of antibodies capable of recognising the combinations of epitopes discussed above in relation to agents of the invention which are a product comprising more than one substance.

The composition typically comprises material from a plant that expresses a gliadin which is capable of causing coeliac disease (for example any of the gliadins or plants mentioned herein). Such material may be a plant part, such as a harvested product (e.g. seed). The material may be processed products of the plant material (e.g. any such product mentioned herein), such as a flour or food that comprises the gliadin. The processing of food material and testing in suitable binding assays is routine, for example as mentioned in Kricka LJ, J. Biolumin. Chemilumin. 13, 189-93 (1998).

Binding assays

The determination of binding between any two substances mentioned herein may be done by measuring a characteristic of either or both substances that changes upon binding, such as a spectroscopic change.

The binding assay format may be a 'band shift' system. This involves determining whether the presence of one substance (such as a candidate substance) advances or retards the progress of the other substance during gel electrophoresis.

The format may be a competitive binding method which determines whether the one substance is able to inhibit the binding of the other substance to an agent which is known to bind the other substance, such as a specific antibody.

Kits

The invention also provides a kit for carrying out the method comprising one or more agents and a means to detect the recognition of the agent by the T cell. Typically the different agents are provided for simultaneous, separate or sequential use. Typically the means to detect recognition allows or aids detection based on the techniques discussed above.

Thus the means may allow detection of a substance secreted by the T cells after recognition. The kit may thus additionally include a specific binding moiety for the substance, such as an antibody. The moiety is typically specific for IFN-γ. The moiety is typically immobilised on a solid support. This means that after binding the moiety the substance will remain in the vicinity of the T cell which secreted it. Thus "spots" of substance/moiety complex are formed on the support, each spot representing a T cell which is secreting the substance. Quantifying the spots, and typically comparing against a control, allows determination of recognition of the agent.

The kit may also comprise a means to detect the substance/moiety complex. A detectable change may occur in the moiety itself after binding the substance, such as a colour change. Alternatively a second moiety directly or indirectly labelled for detection may be allowed to bind the substance/moiety complex to allow the determination of the spots. As discussed above the second moiety may be specific for the substance, but binds a different site on the substance than the first moiety.

The immobilised support may be a plate with wells, such as a microtitre plate. Each assay can therefore be carried out in a separate well in the plate.

The kit may additionally comprise medium for the T cells, detection moieties or washing buffers to be used in the detection steps. The kit may additionally comprise reagents suitable for the separation from the sample, such as the separation of PBMCs or T cells from the sample. The kit may be designed to allow detection of the T cells directly in me sample without requiring any separation of the components of the sample.

The kit may comprise an instrument which allows administration of the agent, such as intradermal or epidermal administration. Typically such an instrument comprises plaster, dressing or one or more needles. The instrument may allow ballistic delivery of the agent. The agent in the kit may be in the form of a pharmaceutical composition.

The kit may also comprise controls, such as positive or negative controls. The positive control may allow the detection system to be tested. Thus the positive control typically mimics recognition of the agent in any of the above methods. Typically in the kits designed to determine recognition in vitro the positive control is a cytokine. In the kit designed to detect in vivo recognition of the agent the positive ontrol may be antigen to which most individuals should response.

The kit may also comprise a means to take a sample containing T cells from the host, such as a blood sample. The kit may comprise a means to separate mononuclear cells or T cells from a sample from the host.

Antibodies

The invention also provides monoclonal or polyclonal antibodies which specifically recognise the agents (such as any of the epitopes of the invention) and use of the agents in methods of making such antibodies. Antibodies of the invention bind specifically to these substances of the invention.

For the purposes of this invention, the term "antibody" includes antibody fragments such as Fv, F(ab) and F(ab)₂ fragments, as well as single-chain antibodies.

A method for producing a polyclonal antibody comprises immunising a suitable host animal, for example an experimental animal, with the immunogen and isolating immunoglobulins from the serum. The animal may therefore be inoculated with the immunogen, blood subsequently removed from the animal and the IgG fraction purified. A method for producing a monoclonal antibody comprises immortalising cells which produce the desired antibody. Hybridoma cells may be produced by fusing spleen cells from an inoculated experimental animal with tumour cells (Kohler and Milstein (1975) Nature 256, 495-497).

An immortalized cell producing the desired antibody may be selected by a conventional procedure. The hybridomas may be grown in culture or injected intraperitoneally for formation of ascites fluid or into the blood stream of an allogenic host or immunocompromised host. Human antibody may be prepared by in vitro immunisation of human lymphocytes, followed by transformation of the lymphocytes with Epstein-Barr virus.

For the production of both monoclonal and polyclonal antibodies, the experimental animal is suitably a goat, rabbit, rat or mouse. If desired, the immunogen may be administered as a conjugate in which the immunogen is coupled, for example via a side chain of one of the amino acid residues, to a suitable carrier. The carrier molecule is typically a physiologically acceptable carrier. The antibody obtained may be isolated and, if desired, purified.

The agent or antibody of the invention may carry a detectable label. Detectable labels which allow detection of the secreted substance by visual inspection, optionally with the aid of an optical magnifying means, are preferred. Such a system is typically based on an enzyme label which causes colour change in a substrate, for example alkaline phosphatase causing a colour change in a substrate. Such substrates are commercially available, e.g. from BioRad. Other suitable labels include other enzymes such as peroxidase, or protein labels, such as biotin; or radioisotopes, such as ³²P or ³⁵S. The above labels may be detected using known techniques.

Agents or antibodies of the invention may be in substantially purified form. They may be in substantially isolated form, in which case they will generally comprise at least 80% e.g. at least 90, 95, 97 or 99% of the polynucleotide, peptide, antibody, cells or dry mass in the preparation. The agent or antibody is typically substantially free of other cellular components. The agent or antibody may be used in such a substantially isolated, purified or free form in the method or be present in such forms in the kit.

The invention provides therapeutic (including prophylactic) agents or diagnostic substances (the agents of the invention). These substances are formulated for clinical administration by mixing them with a pharmaceutically acceptable carrier or diluent. For example they can be formulated for topical, parenteral, intravenous, intramuscular, subcutaneous, intraocular, intradermal, epidermal or transdermal administration. The substances may be mixed with any vehicle which is pharmaceutically acceptable and appropriate for the desired route of administration. The pharmaceutically carrier or diluent for injection may be, for example, a sterile or isotonic solution such as Water for Injection or physiological saline, or a carrier particle for ballistic delivery.

The dose of the substances may be adjusted according to various parameters, especially according to the agent used; the age, weight and condition of the patient to be treated; the mode of administration used; the severity of the condition to be treated; and the required clinical regimen. As a guide, the amount of substance administered by injection is suitably from 0.01 mg/kg to 30 mg/kg, preferably from 0.1 mg/kg to 10 mg/kg.

The routes of administration and dosages described are intended only as a guide since a skilled practitioner will be able to determine readily the optimum route of administration and dosage for any particular patient and condition.

The agent of the invention can be made using standard synthetic chemistry techniques, such as by use of an automated synthesizer. The agent may be made from a longer polypeptide e.g. a fusion protein, which polypeptide typically comprises the sequence of the peptide. The peptide may be derived from the polypeptide by for example hydrolysing the polypeptide, such as using a protease; or by physically breaking the polypeptide. The polynucleotide of the invention can be made using standard techniques, such as by using a synthesiser.

The invention is illustrated by the following nonlimiting Examples:

Example 1

We carried out epitope mapping in Coeliac disease by using a set of 51 synthetic 15-mer peptides that span the complete sequence of a fully characterized a-gliadin, "A-gliadin" (see Table 1). A-Gliadin peptides were also individually treated with tTG to generate products that might mimic those produced in vivo³. We also sought to study Coeliac disease patients at the point of initiation of disease relapse to avoid the possibility that epitope "spreading" or "exhaustion" may have occurred, as described in experimental infectious and autoimmune diseases.

Clinical and A-gliadin specific T cell responses with 3 and 10 day bread challenge

In a pilot study, two subjects with Coeliac disease in remission, defined by absence of serum anti-endomysial antibody (EMA), on a gluten free diet were fed four slices of standard gluten-containing white bread daily in addition to their usual gluten free diet. Subject 1 ceased bread because of abdominal pain, mouth ulcers and mild diarrhoea after three days, but Subject 2 continued for 10 days with only mild nausea at one week. The EMA became positive in Subject 2 one week after the bread challenge, indicating the bread used had caused a relapse of Coeliac disease. But in Subject 1, EMA remained negative up to two months after bread challenge. In both subjects, symptoms that appeared with bread challenge resolved within two days after returning to gluten free diet.

PBMC responses in IFNγ ELISPOT assays to A-gliadin peptides were not found before or during bread challenge. But from the day after bread withdrawal (Day 4) in Subject 1 a single pool of 5 overlapping peptides spanning A-gliadin 51-85 (Pool 3) treated with tTG showed potent IFNγ responses (see Figure 1a). In Subject 1, the PBMC IFNγ response to A-gliadin peptide remained targeted to Pool 3 alone and was maximal on Day 8. The dynamics and magnitude of the response to Pool 3 was similar to that elicited by α-chymotrypsin digested gliadin. PBMC IFNγ responses to tTG-treated Pool 3 were consistently 5 to 12-fold greater than Pool 3 not treated with tTG, and responses to α-chymotrypsin digested gliadin were 3 to 10-fold greater if treated with tTG. In Subject 2, Pool 3 treated with tTG was also the only immunogenic set of A-gliadin peptides on Day 8, but this response was weaker than Subject 1, was not seen on Day 4 and by Day 11 the response to Pool 3 had diminished and other tTG-treated pools of A-gliadin peptides elicited stronger IFNα responses (see Figure 1b).

The pilot study indicated that the initial T cell response in these Coeliac disease subjects was against a single tTG-treated A-gliadin pool of five peptides and was readily measured in peripheral blood. But if antigen exposure is continued for ten days instead of three, T cell responses to other A-gliadin peptides appear, consistent with epitope spreading.

Coeliac disease-specific IFN-g induction by tTG-treated A-gliadin peptides

In five out of six further Coeliac disease subjects on gluten free diet (see Table 1), bread challenge for three days identified tTG-treated peptides in Pool 3, and in particular, peptides corresponding to 56-70 (12) and 60-75 (13) as the sole A-gliadin components eliciting IFNγ from PBMC (see Figure 2). IL-10 ELISPOT assays run in parallel to IFNγ ELISPOT showed no IL-10 response to tTG-treated peptides 12 or 13. In one subject, there were no IFNγ responses to any A-gliadin peptide or α-chymotrypsin digested gliadin before, during or up to four days after bread challenge. In none of these Coeliac disease subjects did EMA status change from baseline when measured for up to two months after bread challenge.

PBMC from four healthy, EMA-negative subjects with the HLA-DQ alleles α1 *0501, β1 *0201 (ages 28-52, 2 females) who had been challenged for three days with bread after following a gluten free diet for one month, showed no IFNγ responses above the negative control to any of the A-gliadin peptides with or without tTG treatment. Thus, induction of IFNγ in PBMC to tTG-treated Pool 3 and A-gliadin peptides 56-70 (12) and 60-75 (13) were Coeliac disease specific (7/8 vs. 0/4, p<0.01 by Chi-squared analysis).

Fine mapping of the minimal A-gliadin T cell epitope

tTG-treated peptides representing truncations of A-gliadin 56-75 revealed that the same core peptide sequence QPQLP (SEQ ID NO:9) was essential for antigenicity in all of the five Coeliac disease subjects assessed (see Figure 3). PBMC IFNγ responses to tTG-treated peptides spanning this core sequence beginning with the 7-mer PQPQLPY (SEQ ID NO:4) and increasing in length, indicated that the tTG-treated 17-mer QLQPFPQPQLPYPQPQS (SEQ ID NO:10) (A-gliadin 57-73) possessed optimal activity in the IFNγ ELISPOT (see Figure 4).

Deamidation of Q65 by tTG generates the immunodominant T cell epitope in A-gliadin

HPLC analysis demonstrated that tTG treatment of A-gliadin 56-75 generated a single product that eluted marginally later than the parent peptide. Amino acid sequencing indicated that out of the six glutamine (Q) residues contained in A-gliadin 56-75, Q65 was preferentially deamidated by tTG (see Figure 5). Bioactivity of peptides corresponding to serial expansions from the core A-gliadin 62-68 sequence in which glutamate (E) replaced Q65, was equivalent to the same peptides with Q65 after tTG-treatment (see Figure 4a). Replacement of Q57 and Q72 by E together or alone, with E65 did not enhance antigenicity of the 17-mer in the three Coeliac disease subjects studied (see Figure 6). Q57 and Q72 were investigated because glutamine residues followed by proline in gliadin peptides are not deamidated by tTG in vitro (W. Vader et al, Proceedings 8th International Symposium Coeliac Disease). Therefore, the immunodominant T cell epitope was defined as QLQPFPQPELPYPQPQS (SEQ ID NO:2).

immunodominant T cell epitope response is DQ2-restricted and CD4 dependent

In two Coeliac disease subjects homozygous for HLA-DQ α1 *0501, β1*0201, anti-DQ monoclonal antibody blocked the ELISPOT IFNγ response to tTG-treated A-gliadin 56-75, but anti-DP and -DR antibody did not (see Figure 7). Anti-CD4 and anti-CD8 magnetic bead depletion of PBMC from two Coeliac disease subjects indicated the IFNγ response to tTG-treated A-gliadin 56-75 is CD4 T cell-mediated.

Discussion

In this study we describe a rather simple dietary antigen challenge using standard white bread to elicit a transient population of CD4 T cells in peripheral blood of Coeliac disease subjects responsive to a tTG-treated A-gliadin 17-mer with the sequence: QLQPFPQPELPYPQPQS (SEQ ID NO:2) (residues 57-73). The immune response to A-gliadin 56-75 (Q→E65) is restricted to the Coeliac disease-associated HLA allele, DQ α1*0501, β1*0201. Tissue transglutaminase action in vitro selectively deamidates Q65. Elicited peripheral blood IFNg responses to synthetic A-gliadin peptides with the substitution Q→E65 is equivalent to tTG-treated Q65 A-gliadin peptides; both stimulate up to 10-fold more T cells in the IFNg ELISPOT than unmodified Q65 A-gliadin peptides.

We have deliberately defined this Coeliac disease-specific T cell epitope using in vivo antigen challenge and short-term ex vivo immune assays to avoid the possibility of methodological artifacts that may occur with the use of T cell clones in epitope mapping. Our findings indicate that peripheral blood T cell responses to ingestion of gluten are rapid but short-lived and can be utilized for epitope mapping. In vivo antigen challenge has also shown there is a temporal hierarchy of immune responses to A-gliadin peptides; A-gliadin 57-73 modified by tTG not only elicits the strongest IFNg response in PBMC but it is also the first IFNg response to appear.

Because we have assessed only peptides spanning A-gliadin, there may be other epitopes in other gliadins of equal or greater importance in the pathogenesis of Coeliac disease. Indeed, the peptide sequence at the core of the epitope in A-gliadin that we have identified PQPQLPY (SEQ ID NO:4) is shared by several other gliadins (SwissProt and Trembl accession numbers: P02863, Q41528, Q41531, Q41533, Q9ZP09, P04722, P04724, P18573). However, A-gliadin peptides that have previously been shown to possess bioactivity in biopsy challenge and in vivo studies (for example: 31-43, 44-55, and 206-217)^4,5 did not elicit IFNg responses in PBMC following three day bread challenge in Coeliac disease subjects. These peptides may be "secondary". T cell epitopes that arise with spreading of the immune response.

Example 2 The effect on T cell recognition of substitutions in the immunodominant epitope

The effect of substituting the glutamate at position 65 in the 57-73 A-gliadin epitope was determined by measuring peripheral blood responses against the substituted epitopes in an IFNγ ELISPOT assay using synthetic peptides (at 50 µg/ml). The responses were measured in 3 Coeliac disease subjects 6 days after commencing gluten challenge (4 slices bread daily for 3 days). Results are shown in table 3 and Figure 8. As can be seen substitution of the glutamate to histidine, tyrosine, tryptophan, lysine, proline or arginine stimulated a response whose magnitude was less than 10% of the magnitude of the response to the immunodominant epitope. Thus mutation of A-gliadin at this position could be used to produce a mutant gliadin with reduce or absent immunoreactivity.

Example 3 Testing the immunoreactivity of equivalent peptides from other naturally occurring gliadins

The immunoreactivity of equivalent peptides form other naturally occurring wheat gliadins was assessed using synthetic peptides corresponding to the naturally occurring sequences which were then treated with transglutaminase. These peptides were tested in an ELISPOT in the same manner and with PBMCs from the same subjects as described in Example 2. At least five of the peptides show immunoreactivity comparable to the A-gliadin 57-73 E65 peptide (after transglutaminase treatment) indicating that other gliadin proteins in wheat are also likely to induce this Coeliac disease-specific immune response (Table 4 and Figure 9).

Methods

Subjects: Patients used in the study attended a Coeliac Clinic in Oxford, United Kingdom. Coeliac disease was diagnosed on the basis of typical small intestinal histology, and normalization of symptoms and small intestinal histology with gluten free diet.

Tissue typing: Tissue typing was performed using DNA extracted from EDTA-anticoagulated peripheral blood. HLA-DQA and DQB genotyping was performed by PCR using sequence-specific primer mixes^6-8.

Anti-endomysial antibody assay: EMA were detected by indirect immunofluorescence using patient serum diluted 1:5 with monkey oesophagus, followed by FITC-conjugated goat anti-human IgA. IgA was quantitated prior to EMA, none of the subjects were IgA deficient.

Antigen Challenge: Coeliac disease subjects following a gluten free diet, consumed 4 slices of gluten-containing bread (50g/slice, Sainsbury's "standard white sandwich bread") daily for 3 or 10 days. EMA was assessed the week before and up to two months after commencing the bread challenge. Healthy subjects who had followed a gluten free diet for four weeks, consumed their usual diet including four slices of gluten-containing bread for three days, then returned to gluten free diet for a further six days.

IFNγ and IL-10 ELISPOT: PBMC were prepared from 50-100 ml of venous blood by Ficoll-Hypaque density centrifugation. After three washes, PBMC were resuspended in complete RPMI containing 10% heat inactivated human AB serum. ELISPOT assays for single cell secretion of IFNγ and IL-10 were performed using commercial kits (Mabtech; Stockholm, Sweden) with 96-well plates (MAIP-S-45; Millipore, Bedford, MA) according to the manufacturers instructions (as described elsewhere⁹) with 2-5x10⁵ (IFNγ) or 0.4-1x10⁵ (IL-10) PBMC in each well. Peptides were assessed in duplicate wells, and Mycobacterium tuberculosis purified protein derivative (PPD RT49) (Serum Institute; Copenhagen, Denmark) (20 µg/ml) was included as a positive control in all assays.

Peptides: Synthetic peptides were purchased from Research Genetics (Huntsville, Alabama) Mass-spectroscopy and HPLC verified peptides' authenticity and >70% purity. Digestion of gliadin (Sigma; G-3375) (100 mg/ml) with α-chymotrypsin (Sigma; C-3142) 200:1 (w/w) was performed at room temperature in 0.1 M NH₄HCO₃ with 2M urea and was halted after 24 h by heating to 98°C for 10 minutes. After centrifugation (13,000g, 10 minutes), the gliadin digest supernatant was filter-sterilized (0.2 mm). Digestion of gliadin was verified by SDS-PAGE and protein concentration assessed. α-Chymotrypsin-digested gliadin (640 µg/ml) and synthetic gliadin peptides (15-mers: 160 µg/ml, other peptides: 0.1 mM) were individually treated with tTG (Sigma; T-5398) (50 µg/ml) in PBS + CaCl₂ 1 mM for 2 h at 37°C. Peptides and peptide pools were aliquotted into sterile 96-well plates and stored frozen at -20°C until use.

Amino acid sequencing of peptides: Reverse phase HPLC was used to purify the peptide resulting from tTG treatment of A-gliadin 56-75. A single product was identified and subjected to amino acid sequencing (automated sequencer Model 494A, Applied Biosystems, Foster City, California). The sequence of unmodified G56-75 was confirmed as: LQLQPFPQPQLPYPQPQSFP (SEQ ID NO:5), and tTG treated G56-75 was identified as: LQLQPFPQPELPYPQPQSFP (SEQ ID NO:11). Deamidation of glutamyl residues was defined as the amount (pmol) of glutamate recovered expressed as a percent of the combined amount of glutamine and glutamate recovered in cycles 2, 4, 8, 10, 15 and 17 of the amino acid sequencing. Deamidation attributable to tTG was defined as (% deamidation of glutamine in the tTG treated peptide - % deamidation in the untreated peptide) / (100 - % deamidation in the untreated peptide).

CD4/CD8 and HLA Class II Restriction: Anti-CD4 or anti-CD8 coated magnetic beads (Dynal, Oslo, Norway) were washed four times with RPMI then incubated with PBMC in complete RPMI containing 10% heat inactivated human AB serum (5x10⁶ cells/ml) for 30 minutes on ice. Beads were removed using a magnet and cells remaining counted. In vivo HLA-class II restriction of the immune response to tTG-treated A-gliadin 56-75 was established by incubating PBMC (5x10⁶ cells/ml) with anti-HLA-DR (L243), -DQ (L2), and -DP (B7.21) monoclonal antibodies (10 µg/ml) at room temperature for one hour prior to the addition of peptide.

Example 12 Development of interferon gamma ELISpot using PBMC and A-gliadin 57-73 QE65 and P04724 84-100 QE92 as a diagnostic for coeliac disease: Definition of immune-responsiveness in newly diagnosed coeliac disease

Induction of responsiveness to the dominant A-gliadin T cell epitope in PBMC measured in the interferon gamma ELISpot follows gluten challenge in almost all DQ2+ coeliac subjects following a long term strict gluten free diet (GFD) but not in healthy DQ2+ subjects after 4 weeks following a strict GFD. A-gliadin 57-73 QE65 responses are not measurable in PBMC of coeliac subjects before gluten challenge and pilot data have suggested these responses could not be measured in PBMC of untreated coeliacs. These data suggest that in coeliac disease immune-responsiveness to A-gliadin 57-73 QE65 is restored following antigen exclusion (GFD). If a diagnostic test is to be developed using the ELISpot assay and PBMC, it is desirable to define the duration of GFD required before gluten challenge is capable of inducing responses to A-gliadin 57-73 QE65 and other immunoreactive gliadin peptides in blood.

Newly diagnosed DQ2+ coeliac subjects were recruited from the gastroenterology outpatient service. PBMC were prepared and tested in interferon gamma ELISpot assays before subjects commenced GFD, and at one or two weeks after commencing GFD. In addition, gluten challenge (3 days consuming 4 slices standard white bread, 200g/day) was performed at one or two weeks after starting GFD. PBMC were prepared and assayed on day six are after commencing gluten challenge. A-gliadin 57-73 QE65 (A), P04724 84-100 QE92 (B) (alone and combined) and A-gliadin 57-73 QP65 (P65) (non-bioactive variant, see above) (all 25 mcg/ml) were assessed.

All but one newly diagnosed coeliac patient was DQ2+ (one was DQ8+) (n=11). PBMC from newly diagnosed coeliacs that were untreated, or after 1 or 2 weeks following GFD did not show responses to A-gliadin 57-73 QE65 and P04724 84-100 QE92 (alone or combined) that were not significantly different from blank or A-gliadin 57-73 QP65 (n=9) (see Figure 11). Gluten challenge in coeliacs who had followed GFD for only one week did not substantially enhance responses to A-gliadin 57-73 QE65 or P04724 84-100 QE92 (alone or combined). But gluten challenge 2 weeks after commencing GFD did induce responses to A-gliadin 57-73 QE65 and P04724 84-100 QE92 (alone or combined) that were significantly greater than the non-bioactive variant A-gliadin 57-73 QP65 and blank. Although these responses after gluten challenge at 2 weeks were substantial they appear to be less than in subjects >2 months after commencing GFD. Responses to A-gliadin 57-73 QE65 alone were equivalent or greater than responses to P04724 84-100 QE92 alone or when mixed with A-gliadin 57-73 QE65. None of the subjects experienced troubling symptoms with gluten challenge.

Immune responsiveness (as measured in PBMC after gluten challenge) to A-gliadin is partially restored 2 weeks after commencing GFD, implying that "immune unresponsiveness" to this dominant T cell epitope prevails in untreated coeliac disease and for at least one week after starting GFD. The optimal timing of a diagnostic test for coeliac disease using gluten challenge and measurement of responses to A-gliadin 57-73 QE65 in the ELISpot assay is at least 2 weeks after commencing a GFD.

Interferon gamma-secreting T cells specific to A-gliadin 57-73 QE65 cannot be measured in the peripheral blood in untreated coeliacs, and can only be induced by gluten challenge after at least 2 weeks GFD (antigen exclusion). Therefore, timing of a diagnostic test using this methodology is crucial and further studies are needed for its optimization. These finding are consistent with functional anergy of T cells specific for the dominant epitope, A-gliadin 57-73 QE65, reversed by antigen exclusion (GFD). This phenomenon has not been previously demonstrated in a human disease, and supports the possibility that T cell anergy may be inducible with peptide therapy in coeliac disease.

Example 13 Comprehensive Mapping of Wheat Gliadin T Cell Epitopes

Antigen challenge induces antigen-specific T cells in peripheral blood. In coeliac disease, gluten is the antigen that maintains this immune-mediated disease. Gluten challenge in coeliac disease being treated with a gluten free diet leads to the appearance of gluten-specific T cells in peripheral blood, so enabling determination of the molecular specificity of gluten T cell epitopes. As described above, we have identified a single dominant T cell epitope in a model gluten protein, A-gliadin (57-73 deamidated at Q65). In this Example, gluten challenge in coeliac patients was used to test all potential 12 amino acid sequences in every known wheat gliadin protein derived from 111 entries in Genbank. In total, 652 20mer peptides were tested in HLA-DQ2 and HLA-DQ8 associated coeliac disease. Seven of the 9 coeliac subjects with the classical HLA-DQ2 complex (HLA-DQA1 *05, HLA-DQB1*02) present in over 90% of coeliacs had an inducible A-gliadin 57-73 QE65- and gliadin-specific T cell response in peripheral blood. A-gliadin 57-73 was the only significant α-gliadin T cell epitope, as well as the most potent gliadin T cell epitope, in HLA-DQ2-associated coeliac disease. In addition, there were as many as 5 families of structurally related peptides that were between 10 and 70% as potent as A-gliadin 57-73 in the interferon-γ ELISpot assay. These new T cell epitopes were derived from γ- and ω-gliadins and included common sequences that were structurally very similar, but not identical to the core sequence of A-gliadin 57-73 (core sequence: FPQPQLPYP (SEQ ID NO:18)), for example: FPQPQQPFP (SEQ ID NO:19) and PQQPQQPFP (SEQ ID NO:20). Although no homologues of A-gliadin 57-73 have been found in rye or barley, the other two cereals toxic in coeliac disease, the newly defined T cell epitopes in γ- and ω-gliadins have exact matches in rye and barley storage proteins (secalins and hordeins, respectively).

Coeliac disease not associated with HLA-DQ2 is almost always associated with HLA-DQ8. None of the seven HLA-DQ8+ coeliac subjects had inducible A-gliadin 57-73-specific T cell responses following gluten challenge, unless they also possessed the complete HLA-DQ2 complex. Two of 4 HLA-DQ8+ coeliac subjects who did not possess the complete HLA-DQ2 complex, had inducible gliadin peptide-specific T cell responses following gluten challenge. In one HLA-DQ8 subject, a novel dominant T cell epitope was identified with the core sequence LQPQNPSQQQPQ (SEQ ID NO:21). The transglutaminase-deamidated version of this peptide was more potent than the non-deamidated peptide. Previous studies suggest that the transglutaminase-deamidated peptide would have the sequence LQPENPSQEQPE (SEQ ID NO:22); but further studies are required to confirm this sequence. Amongst the healthy HLA-DQ2 (10) and HLA-DQ8 (1) subjects who followed a gluten free diet for a month, gliadin peptide-specific T cell responses were uncommon, seldom changed with gluten challenge, and were never potent T cell epitopes revealed with gluten challenge in coeliac subjects. In conclusion, there are unlikely to be more than six important T cell epitopes in HLA-DQ2-associated coeliac disease, of which A-gliadin 57-73 is the most potent. HLA-DQ2- and HLA-DQ8-associated coeliac disease do not share the same T cell specificity.

We have shown that short-term gluten challenge of individuals with coeliac disease following a gluten free diet induces gliadin-specific T cells in peripheral blood. The frequency of these T cells is maximal in peripheral blood on day 6 and then rapidly wanes over the following week. Peripheral blood gliadin-specific T cells express the integrin α4β7 that is associated with homing to the gut lamina propria. We exploited this human antigen-challenge design to map T cell epitopes relevant to coeliac disease in the archetypal gluten α-gliadin protein, A-gliadin. Using 15mer peptides overlapping by 10 amino acids with and without deamidation by transglutaminase (tTG), we demonstrated that T cells induced in peripheral blood initially target only one A-gliadin peptide, residues 57-73 in which glutamine at position 65 is deamidated. The epitope is HLA-DQ2-restricted, consistent with the intimate association of coeliac disease with HLA-DQ2.

Coeliac disease is reactivated by wheat, rye and barley exposure. The α/β-gliadin fraction of wheat gluten is consistently toxic in coeliac disease, and most studies have focused on these proteins. The gene cluster coding for α/β-gliadins is located on wheat chromosome 6C. There are no homologues of α/β-gliadins in rye or barley. However, all three of the wheat gliadin subtypes (α/β, γ, and co) are toxic in coeliac disease. The γ- and ω-gliadin genes are located on chromosome 1A in wheat, and are homologous to the secalins and hordeins in rye and barley.

There are now genes identified for 61 α-gliadins in wheat (Triticum aestivum). The α-gliadin sequences are closely homologous, but the dominant epitope in A-gliadin derives from the most polymorphic region in the α-gliadin sequence. Anderson et al (1997) have estimated that there are a total of about 150 distinct α-gliadin genes in T. aestivum, but many are psuedogenes. Hence, it is unlikely that T-cell epitopes relevant to coeliac disease are not included within known α-gliadin sequences.

Our work has identified a group of deamidated α-gliadin peptides almost identical to A-gliadin 57-73 as potent T cell epitopes specific to coeliac disease. Over 90% of coeliac patients are HLA-DQ2+, and so far, we have only assessed HLA-DQ2+ coeliac subjects after gluten challenge. However, coeliac patients who do not express HLA-DQ2 nearly all carry HLA-DQ8. Hence, it is critical to know whether A-gliadin 57-73 and its homologues in other wheat, rye and barley gluten proteins are the only T-cell epitopes recognized by T cells induced by gluten challenge in both HLA-DQ2+ and HLA-DQ8+ coeliac disease. If this were the case, design of peptide therapeutics for coeliac disease might only require one peptide.

Homologues of A-gliadin 57-73 as T-cell epitopes

Initial searches of SwissProt and Trembl gene databases for cereal genes coding for the core sequence of A-gliadin 57-73 (PQLPY <SEQ ID NO:12>) only revealed α/β-gliadins. However, our fine-mapping studies of the A-gliadin 57-73 QE65 epitope revealed a limited number of permissive point substitutions in the core region (PQLP) (note Q65 is actually deamidated in the epitope). Hence, we extended our search to genes in SwissProt or Trembl databases encoding for peptides with the sequence XXXXXXXPQ[ILMP][PST]XXXXXX (SEQ ID NO:23). Homologues were identified amongst γ-gliadins, glutenins, hordeins and secalins (see Table 12). A further homologue was identified in ω-gliadin by visual search of the three ω-gliadin entries in Genbank.

These homologues of A-gliadin 57-73 were assessed after deamidation by tTG (or synthesis of the glutamate(QE)-substituted variant in four close homologues) using the IFNγ ELISpot assay with peripheral blood mononuclear cells after gluten challenge in coeliac subjects. The m-gliadin sequence (AAG17702 141-157) was the only bioactive peptide, approximately half as potent as A-gliadin 57-73 (see Table 12, and Figure 12). Hence, searches for homologues of the dominant A-gliadin epitope failed to account for the toxicity of γ-gliadin, secalins, and hordeins.

Methods Design of a set ofpeptides spanning all possible wheat gliadin T-cell epitopes

In order to identify all possible T cell epitopes coded by the known wheat (Triticum aestivum) gliadin genes or gene fragments (61 α/β-, 47 γ-, and 3 ω-gliadin entries in Genbank), gene-derived protein sequences were aligned using the CustalW software (MegAlign) and arranged into phylogenetic groupings (see Table 22). Many entries represented truncations of longer sequences, and many gene segments were identical except for the length of polyglutamine repeats or rare substitutions. Hence, it was possible to rationalize all potential unique 12 amino acid sequences encoded by known wheat genes to be included in a set of 652 20mer peptides. (Signal peptide sequences were not included). Peptide sequences are listed in Table 23.

Comprehensive epitope mapping

Healthy controls (HLA-DQ2+ n=10, and HLA-DQ8+ n=1) who had followed a gluten free diet for 4 weeks, and coeliac subjects (six HLA-DQ2, four complex heterozygotes HLA-DQ2/8, and three HLA-DQ8/X) (see Table 13) following long-term gluten free diet were studied before and on day 6 and 7 after 3-day gluten challenge (four 50g slices of standard white bread - Sainsbury's sandwich bread, each day). Peripheral blood (a total of 300ml over seven days) was collected and peripheral blood mononuclear cells (PBMC) were separated by Lymphoprep density gradient. PBMC were incubated with pools of 6 or 8 20mer peptides, or single peptides with or without deamidation by tTG in overnight interferon gamma (IFNγ) ELISpot assays.

Peptides were synthesized in batches of 96 as Pepsets (Mimotopes Inc., Melbourne Australia). Approximately 0.6 micromole of each of 652 20mers was provided. Two marker 20mer peptides were included in each set of 96 (VLQQHNIAHGSSQVLQESTY - peptide 161 (SEQ ID NO:24), and IKDFHVYFRESRDALWKGPG (SEQ ID NO:25)) and were characterized by reverse phase-HPLC and amino acid sequence analysis. Average purities of these marker peptides were 50% and 19%, respectively. Peptides were initially dissolved in acetonitrile (10%) and Hepes 100mM to 10mg/ml.

The final concentration of individual peptides in pools (or alone) incubated with PBMC for the IFNγ ELISpot assays was 20 µg/ml. Five-times concentrated solutions of peptides and pools in PBS with calcium chloride 1mM were aliquotted and stored in 96-well plates according to the template later used in ELISpot assays. Deamidated peptides and pools of peptides were prepared by incubation with guinea pig tissue tTG (Sigma T5398) in the ratio 100:32 µg/ml for two hours at 37°C. Peptides solutions were stored at -20°C and freshly thawed prior to use.

Gliadin (Sigma G3375) (100 mg/ml) in endotoxin-free water and 2M urea was boiled for 10 minutes, cooled to room temperature and incubated with filter (0.2 µm)-sterilised pepsin (Sigma P6887) (2 mg/ml) in HCl 0.02M or chymotrypsin (C3142) (4mg/ml) in ammonium bicarbonate (0.2M). After incubation for 4 hours, pepsin-digested gliadin was neutralized with sodium hydroxide, and then both pepsin- and chymotrypsin-digested gliadin were boiled for 15 minutes. Identical incubations with protease in which gliadin was omitted were also performed. Samples were centrifuged at 15 000g, then protein concentrations were estimated in supernatants by the BCA method (Pierce, USA). Before final use in IFNγ ELISpot assays, aliquots of gliadin-protease were incubated with tTG in the ratio 2500:64 µg/ml.

IFNγ ELISpot assays (Mabtech, Sweden) were performed in 96-well plates (MAIP S-45, Millipore) in which each well contained 25µl of peptide solution and 100µl of PBMC (2-8x10⁵/well) in RPMI containing 10% heat inactivated human AB serum. Deamidated peptide pools were assessed in one 96-well ELISpot plate, and peptides pools without deamidation in a second plate (with an identical layout) on both day 0 and day 6. All wells in the plate containing deamidated peptides included tTG (64 µg/ml). In each ELISpot plate there were 83 wells with peptide pools (one unique pool in each well), and a series of wells for "control" peptides (peptides all >90% purity, characterized by MS and HPLC, Research Genetics): P04722 77-93 (QLQPFPQPQLPYPQPQP (SEQ ID NO:26)), P04722 77-93 QE85 (in duplicate) (QLQPFPQPELPYPQPQP (SEQ ID NO:27)), P02863 77-93 (QLQPFPQPQLPYSQPQP (SEQ ID NO:28)), P02863 77-93 QE85 (QLQPFPQPELPYSQPQP (SEQ ID NO:29)), and chymotrypsin-digested gliadin (500 µg/ml), pepsin-digested gliadin (500 µg/ml), chymotrypsin (20 µg/ml) alone, pepsin (10 µg/ml) alone, and blank (PBS+/-tTG) (in triplicate).

After development and drying, IFNγ ELISpot plates were assessed using the MAIP automated ELISpot plate counter. In HLA-DQ2 healthy and coeliac subjects, induction of spot forming cells (sfc) by peptide pools in the IFNγ ELISpot assay was tested using a one-tailed Wilcoxon Matched-Pairs Signed-Ranks test (using SPSS software) applied to spot forming cells (sfc) per million PBMC minus blank on day 6 versus day 0 ("net response"). Significant induction of an IFNγ response to peptide pools in PBMC by in vivo gluten challenge was defined as a median "net response" of at least 10 sfc/million PBMC and p<0.05 level of significance. Significant response to a particular pool of peptides on day 6 was followed by assessment of individual peptides within each pool using PBMC drawn the same day or on day 7.

For IFNγ ELISpot assays of individual peptides, bioactivity was expressed as a percent of response to P04722 77-93 QE85 assessed in the same ELISpot plate. Median response to blank (PBS alone) was 0.2 (range 0-5) sfc per well, and the positive control (P04722 77-93 QE85) 76.5 (range: 25-282) sfc per well using a median of 0.36 million (range: 0.3-0.72) PBMC. Hence, median response to blank expressed as a percentage of P04722 77-93 QE65 was 0.2% (range: 0-6.7). Individual peptides with mean bioactivity greater than 10% that of P04722 QE85 were analyzed for common structural motifs.

Results Healthy HLA-DQ2 subjects

None of the healthy HLA-DQ2+ subjects following a gluten free diet for a month had IFNγ ELISpot responses to homologues of A-gliadin 57-73 before or after gluten challenge. However, in 9/10 healthy subjects, gluten challenge was associated with a significant increase in IFNγ responses to both peptic- and chymotryptic-digests of gliadin, from a median of 0-4 sfc/million on day 0 to a median of 16-29 sfc/million (see Table 14). Gliadin responses in healthy subjects were unaffected by deamidation (see Table 15). Amongst healthy subjects, there was no consistent induction of IFNγ responses to specific gliadin peptide pools with gluten challenge (see Figure 13, and Table 16). IFNγ ELISpot responses were occasionally found, but these were weak, and not altered by deamidation. Many of the strongest responses to pools were also present on day 0 (see Table 17, subjects H2, H8 and H9). Four healthy subjects did show definite responses to pool 50,and the two with strongest responses on day 6 also had responses on day 0. In both subjects, the post-challenge responses to pool 50 responses were due to peptide 390 (QQTYPQRPQQPFPQTQQPQQ (SEQ ID NO:30)).

HLA-DQ2 coeliac subjects

Following gluten challenge in HLA-DQ2+ coeliac subjects, median IFNγ ELISpot responses to P04722 77-93 E85 rose from a median of 0 to 133 sfc/million (see Table 4). One of the six coeliac subjects (C06) did not respond to P04722 77-93 QE85 (2 sfc/million) and had only weak responses to gliadin peptide pools (maximum: Pool 50+tTG 27 sfc/million). Consistent with earlier work, bioactivity of wild-type P04722 increased 6.5 times with deamidation by tTG (see Table 15). Interferon-gamma responses to gliadin-digests were present at baseline, but were substantially increased by gluten challenge from a median of 20 up to 92 sfc/million for chymotryptic-gliadin, and from 44 up to 176 sfc/million for peptide-gliadin. Deamidation of gliadin increased bioactivity by a median of 3.2 times for chymotryptic-gliadin and 1.9 times for peptic-gliadin (see Table 15). (Note that the acidity required for digestion by pepsin is likely to result in partial deamidation of gliadin.)

In contrast to healthy subjects, gluten challenge induced IFNγ ELISpot responses to 22 of the 83 tTG-treated pools including peptides from α-, γ- and ω-gliadins (see Figure 14, and Table 17). Bioactivity of pools was highly consistent between subjects (see Table 18). IFNγ ELISpot responses elicited by peptide pools were almost always increased by deamidation (see Table 17). But enhancement of bioactivity of pools by deamidation was not as marked as for P04722 77-73 Q85, even for pools including homologues of A-gliadin 57-73. This suggests that Pepset peptides were partially deamidated during synthesis or in preparation, for example the Pepset peptides are delivered as salts of trifluoracetic acid (TFA) after lyophilisation from a TFA solution.

One hundred and seventy individual tTG-deamidated peptides from 21 of the most bioactive pools were separately assessed. Seventy-two deamidated peptides were greater than 10% as bioactive as P04722 77-93 QE85 at an equivalent concentration (20 µg/ml) (see Table 19). The five most potent peptides (85-94% bioactivity of P04722 QE85) were previously identified α-gliadin homologues A-gliadin 57-73. Fifty of the bioactive peptides were not homologues of A-gliadin 57-73, but could be divided into six families of structurally related sequences (see Table 20). The most bioactive sequence of each of the peptide families were: PQQPQQPQQPFPQPQPFPW (SEQ ID NO:31) (peptide 626, median 72% bioactivity of P04722 QE85), QQPQQPFPOPQQPQLPFPQQ (SEQ ID NO:32) (343, 34%), QAFPQPQQTFPHQPQQQFPQ (SEQ ID NO:33) (355, 27%), TQQPQQPFPQQPQQPFPQTQ (SEQ ID NO:34) (396, 23%), PIQPQQPFPQQPQQPQQPFP (SEQ ID NO:35) (625, 22%), PQQSFSYQQQPFPQQPYPQQ (SEQ ID NO:36) (618, 18%) (core sequences are underlined). All of these sequences include glutamine residues predicted to be susceptible to deamidation by transglutaminase (e.g. QXP, QXPF (SEQ ID NO:37), QXX[FY] (SEQ ID NO:38)) (see Vader et al 2002). Some bioactive peptides contain two core sequences from different families.

Consistent with the possibility that different T-cell populations respond to peptides with distinct core sequences, bioactivity of peptides from different families appear to be additive. For example, median bioactivity of tTG-treated Pool 81 was 141% of P04722 QE85, while bioactivity of individual peptides was in rank order: Peptide 631 (homologue of A-gliadin 57-73) 61%, 636 (homologue of 626) 51%, and 635 19%, 629 16%, and 634 13% (all homologues of 396).

Although likely to be an oversimplification, the contribution of each "peptide family" to the summed IFNγ ELISpot response to gliadin peptides was compared in the HLA-DQ2+ coeliac subjects (see Figure 15. Accordingly, the contribution of P04722 77-73 E85 to the summed response to gliadin peptides is between 1/5 and 2/3.

Using the peptide homology search programme, WWW PepPepSearch, which can be accessed through the world wide web of the internet at, for example, "cbrg.inf.ethz.ch/subsection3_1_5.html.", and by direct comparison with Genbank sequences for rye secalins, exact matches were found for the core sequences QQPFPQPQQPFP (SEQ ID NO:39) in barley hordeins (HOR8) and rye secalins (A23277, CAA26449, AAG35598), QQPFPQQPQQPFP (SEQ ID NO:40) in barley hordeins (HOG1 and HOR8), and for PIQPQQPFPQQP (SEQ ID NO:41) also in barley hordeins (HOR8).

HLA-DQ8-associated coeliac disease

Seven HLA-DQ8+ coeliac subjects were studied before and after gluten challenge. Five of these HLA-DQ8+ (HLA-DQA0*0301-3, HLA-DQB0*0302) subjects also carried one or both of the coeliac disease-associated HLA-DQ2 complex (DQA0*05, DQB0*02). Two of the three subjects with both coeliac-associated HLA-DQ complexes had potent responses to gliadin peptide pools (and individual peptides including P04722 77-93 E85) that were qualitatively and quantitatively identical to HLA-DQ2 coeliac subjects (see Figures 16 and 17, and Table 18). Deamidated peptide pool 74 was bioactive in both HLA-DQ2/8 subjects, but only in one of the 6 HLA-DQ2/X subjects. Pretreatment of pool 74 with tTG enhances bioactivity between 3.8 and 22-times, and bioactivity of tTG-treated pool 74 in the three responders is equivalent to between 78% and 350% the bioactivity of P04722 77-93 E85. Currently, it is not known which peptides are bioactive in Pool 74 in subject C02, C07, and C08.

Two of the four HLA-DQ8 coeliac subjects that lacked both or one of the HLA-DQ2 alleles associated with coeliac disease showed very weak IFNγ ELISpot responses to gliadin peptide pools, but the other two did respond to both protease-digested gliadin and specific peptide pools. Subject C12 (HLA-DQ7/8) responded vigorously to deamidated Pools 1-3 (see Figure 18). Assessment of individual peptides in these pools identified a series of closely related bioactive peptides including the core sequence LQPQNPSQQQPQ (SEQ ID NO:42) (see Table 20). Previous work (by us) has demonstrated that three glutamine residues in this sequence are susceptible to tTG-mediated deamidation (underlined). Homology searches using WWW PepPepSearch have identified close matches to LQPQNPSQQQPQ (SEQ ID NO:43) only in wheat α-gliadins.

The fourth HLA-DQ8 subject (C11) had inducible IFNγ ELISpot responses to tTG-treated Pool 33 (see Figure 19). Pools 32 and 33 include polymorphisms of a previously defined HLA-DQ8 restricted gliadin epitope (QQYPSGQGSFQPSQQNPQ (SEQ ID NO:44)) active after deamidation by tTG (underlined Gln are deamidated and convey bioactivity) (van der Wal et al 1998). Currently, it is not known which peptides are bioactive in Pool 33 in subject C11.

Comprehensive T cell epitope mapping in HLA-DQ2-associated coeliac disease using in vivo gluten challenge and a set of 652 peptides spanning all known 12 amino acid sequences in wheat gliadin has thus identified at least 72 peptides at 10% as bioactive as the known α-gliadin epitope, A-gliadin 57-73 E65. However, these bioactive peptides can be reduced to a set of perhaps as few as 5 distinct but closely related families of peptides. Almost all these peptides are rich in proline, glutamine, phenylalanine, and/or tyrosine and include the sequence PQ(QL)P(FY)P (SEQ ID NO:45). This sequence facilitates deamidation of Q in position 2 by tTG. By analogy with deamidation of A-gliadin 57-68 (Arentz-Hansen 2000), the enhanced bioactivity of these peptides generally found with deamidation by tTG may be due to increased affinity of binding for HLA-DQ2.

Cross-reactivity amongst T cells in vivo recognizing more than one of these bioactive gliadin peptides is possible. However, if each set of related peptides does activate a distinct T cell population in vivo, the epitope corresponding to A-gliadin 57-73 E65 is the most potent and is generally recognized by at least 40% of the peripheral blood T cells that secrete IFNγ in response to gliadin after gluten challenge.

No gliadin-peptide specific responses were found in HLA-DQ2/8 coeliac disease that differed qualitatively from those in HLA-DQ2/X-associated coeliac disease. However, peripheral blood T cells in HLA-DQ8+ coeliac subjects without both HLA-DQ2 alleles did not recognize A-gliadin 57-73 E65 homologues. Two different epitopes were dominant in two HLA-DQ8+ coeliacs. The dominant epitope in one of these HLA-DQ8+ individuals has not been identified previously (LQPQNPSQQQPQ (SEQ ID NO:46)).

Given the teaching herein, design of an immunotherapy for coeliac disease utilizing all the commonly recognised T cell epitopes is practical and may include fewer than six distinct peptides. Epitopes in wheat γ- and ω-gliadins are also present in barley hordeins and rye secalins.

Example 14

Several ELISpot assays were performed as previously described and yielded the following results and/or conclusions:

Examination of multiple α-gliadin polymorphisms with PQLPY

Potent agonists of A-gliadin 57-73QE (G01) include identity with the G01 sequence in the particular position.

Gluten challenge induces A-gliadin 57-73 QE65 T cells only after two weeks of gluten-free diet in newly diagnosed coeliac disease

Additional analyses indicated that tTG-deamidated gliadin responses change after two weeks of gluten-free diet in newly diagnosed coeliac disease. Other analyses indicated that deamidated gliadin-specific T cells are CD4⁺α₄β₇ ⁺ HLA-DQ2 restricted.

Optimal epitope (clones versus gluten challenge)

A "dominant" epitope is defined by γIFN ELISpot after gluten challenge. QLQPFPQPELPYPQPQS (100% ELISpot response). Epitopes defined by intestinal T cell clones: QLQPFPQPELPY (27%), PQPELPYPQPELPY (52%), and QQLPQPEQPQQSFPEQERPF (9%).

Dominance among individual peptide responses

Dominance depends on wheat or rye. For wheat, dominant peptides include peptide numbers 89, 90 and 91 (referring to sequence numbers in Table 23). For rye, dominant peptides include peptide numbers 368, 369, 370, 371, and 372 (referring to sequence numbers in Table 23). Some peptides, including 635 and 636 (referring to sequence numbers in Table 23) showed activity in both rye and wheat.

In vivo gluten challenge allows T cell epitope hierarchy to be defined for coeliac disease

The epitope hierarchy is consistent among HLA-DQ2+ coeliacs but different for HLA-DQ8⁺ coeliacs. The hierarchy depends on what cereal is consumed. Deamidation generates almost all gliadin epitopes. HLA-DQ2, DQ8, and DR4 present deamidated peptides. HLA-DQ2/8-associated coeliac disease preferentially present DQ2-associated gliadin epitopes. Gliadin epitopes are sufficiently restricted to justify development of epitope-based therapeutics.

Other analyses indicated the following: HLA-DR3-DQ2 (85-95%) and HLA-DR4-DQ8 (5-15%).

Other analyses indicated the following:

	HLA-DQ	HLA-DQA1 allele	HLA-DQB1 allele	Duodenal histology	Gluten free	EMA on gluten (on GFD)
C01	2,6	102/6, 501	201, 602	SVA	1 yr	+(-)
C02	2,2	501	201	SVA	1 yr	+(-)
C03	2,5	101/4/5, 501	201, 501	PVA	1 yr	+(-)
C04	2,5	101/4/5, 501	201, 501	SVA	7 yr	+(-)
C05	2,2	201, 501	201, 202	SVA	4 mo	+(ND)
C06	2,2	201, 501	201, 202	SVA	2 yr	+(-)
C07	2,8	301-3, 501	201, 302	SVA	1 yr	+(-)
C08	2,8	301-3, 501	201, 302/8	SVA	11 yr	ND (-)
C09	2,8	301-3, 501	201, 302	SVA	29 yr	+(-)
C10	2,8	201, 301-3	202, 302	IEL	1 yr	+(-)
C11	6,8	102/6, 301-3	602/15, 302/8	IEL	9 mo	- (ND)
C12	8,7	301-3, 505	302, 301/9-10	SVA	2 yr	-(-)
C13	8,8	301	302	SVA	1 yr	+(+)

Another analysis was carried out to determine the bioactivity of individual tTG-deamidated peptides in pools 1-3 in subject C12. The results are as follows (sequence numbers refer to the peptides listed in Table 23): Sequence 8 (100%), Sequence 5 (85%), Sequence 6 (82%), Sequence 3 (77%), Sequence 1 (67%), Sequence 2 (59%), Sequence 9 (49%), Sequence 7 (49%), Sequence 10 (33%), Sequence 4 (15%), Sequence 12 (8%), Sequence 11 (0%), Sequence 23 (26%), Sequence 14 (18%), Sequence 15 (18%), Sequence 17 (18%), Sequence 16 (13%), Sequence 14 (8%), Sequence 22 (5%), Sequence 18 (3%), Sequence 19 (3%), Sequence 20 (0%), Sequence 21 (0%). The predicted deamidated sequence is LQPENPSQEQPE.

Individual ELISpot responses by PBMC (Spot forming cells determined by ELISpot Reader)

Peptide (see Table 23)	C01	C02	C03	C04	C05
65	16	2	1	2	3
66	32	6	13	0	6
67	16	3	4	0	4
68	25	8	4	2	2
69	4	0	0	0	0
70	2	1	0	0	0
71	1	1	0	0	1
72	0	0	0	0	0
73	95	21	42	31	31
74	122	15	29	21	28
75	5	1	2	2	5
76	108	13	28	16	22
77	3	0	1	0	1
78	21	2	3	5	3
79	20	0	2	0	2
80	5	2	0	0	3
81	4	1	2	3	1
82	3	3	5	2	2
83	14	2	0	0	1
84	3	0	0	0	0
85	14	1	2	1	2
86	11	0	2	0	2

Cross-reactivity

To deal with data from 652 peptides in 29 subjects, or to determine when a particular response is a true positive peptide-specific T-cell response, or to determine when a response to a peptide is due to cross-reactivity with another structurally related peptide, expression of a particular peptide response can be as a percentage of a "dominant" peptide response. Alternately, the expression can be a "relatedness" as correlation coefficients between peptide responses, or via bioinformatics.

Additional epitopes

A representative result is as follows: Combination of peptides with P04722E (all 20mcg/ml) (n=4)

	Alone	P04722E+
Pep 626	60	135
P04722E	100	110
HLAa	0	85

(expressed as percent P04722E) 626+tT: PQQPQQPQQPFPQPQQPFPW P04724E: QLQPFPQPELPYPQPQL TTG-deamidation of peptide 626 (n=12) No tTG = 100% TTG = 170% Substitution at particular positions Substitution of Peptide 626 PQQP[Q1]QP[Q2]QPFPQP[Q3]QPFPW(n=12)

	Glu	Arg
Q1	95	90
Q2	145	80
Q3	155	10

(expressed as percent wild-type peptide) Bioactivity of tTG-treated 15mers spanning Peptide 626/627 (PQQPQQPQQPFPQPQQPFPWQP) (n=8)

P1-15	5
P2-16	4
P3-17	3
P4-18	38
P5-19	65
P6-20	95
P7-21	65
P8-22	90

(expressed as percent of maximal 15mer response)

Multiple epitopes:

• P04724E: QLQPFPQPQLPYPQPQL
• 626+tTG: PQQPQQPQQPFPQPQQPFPW
• Minimal epitope: QPQQPFPQPQQPFPW

Immunomagnetic depletion of PBMC by beads coated with anti-CD4 and by anti-integrin β₇ depleted IFNγ ELISpot responses, while immunomagnetic depletion of PBMC by beads coated with anti-CD8 or anti-alphaE integrin. Thus, the PBMC secreting IFNγ are CD4+ and α₄β₇+, associated with homing to the lamina propria in the gut.

Blocked by anti-DQ antibody but not by anti-DR antibody in heterozygotes and homozygotes for HLA-DQ2. This may imply multiple epitopes within one sequence.

T cell epitopes in coeliac disease

Other investigators have characterized certain intestinal T cell clone epitopes. See, e.g., Vader et al., Gastroenterology 2002, 122:1729-37; Arentz-Hansen et al., Gastroenterology 2002, 123:803-809. These are examples of epitopes whose relevance is at best unclear because of the in vitro techniques used to clone T cells.

Intestinal versus peripheral blood clones

Intestinal: 1) intestinal biopsies, 2) T cell clones raised against peptic-tryptic digest of gluten, 3) all HLA-DQ2 restricted, 4) clones respond to gliadin deamidated by transglutaminase. Peripheral blood: 1) T cell clones raised against gluten are HLA-DR, DQ and DP restricted. Result: Intestinal T cell clones can be exclusively used to map coeliac disease associated epitopes

GDA_9 Wheat 307 aa Definition Alpha/Beta-Gliadin MM1 Precursor (Prolamin) Accession P18573 -- Genbank (which is incorporated herein by reference in its entirety)

Intestinal T cell clone epitopes

A definition of intestinal T cell clone epitopes can be found in; for example, Arentz-Hansen et al., J Exp Med. 2000,191:603-12. Also disclosed therein are gliadin epitopes for intestinal T cell clones. Deamidated QLQPFPQPQLPY is an epitope, with a deamidated sequence of QLQPFPQPELPY. There is an HLA-DQ2 restriction..A homology search shows other bioactive rAlpha-gliadins include PQPQLPY singly or duplicated. A majority of T cell clones respond to either/or DQ2-αI: QLQPFPQPELPY DQ2-αII: PQPELPYPQPELPY

Dominant gliadin T cell epitopes

All deamidated by transglutaminase. Peripheral blood day 6 after gluten challenge: A-gliadin 57-73: QLQPFPQPELPYPQPQS Intestinal T cell clones: DQ2-αI: QLQPFPQPELPY DQ2-αII: PQPELPYPQPELPY

Intestinal T-cell Clone Epitope Mapping

α-Gliadins	A1	PFPQPQLPY
A2	A2	PQPQLPYPQ
A3	PYPQPQLPY
Glia-20	PQQPYPQPQPQ
Γ-Gliadins	G1	PQQSFPQQQ
G2	IIPQQPAQ
G3	FPQQPQQPYPQQP
G4	FSQPQQQFPQPQ
G5	LQPQQPFPQQPQQPYPQQPQ
Glu-21	QSEQSQQPFPQQF
Glu-5	Q(IL)PQQPQQF
Glutenin	Glt-156	PFSQQQQSPF
Glt-17	PFSQQQQQ

Gluten exposure and induction of IFNγ-secreting A-Gliadin 57-73QE65-specific T cells in peripheral blood

Untreated coeliac disease, followed by gluten free diet for 1, 2, or 8 weeks, followed by gluten exposure (3 days bread 200g/day), followed by gluten free diet

Result 1: Duration of gluten free diet and IFNγ ELISpot responses on day 0 and day 6 of gluten challenge: A-gliadin 57-73 QE65 (results expressed as IFNγ specific spots/million PPBMC)

• Day 0: none (5), 1 week (1), 2 weeks (2), 8 weeks (1)
• Day 6: none (0), 1 week (4), 2 weeks (28), 8 weeks (48)

Result 2: Duration of gluten free diet and IFNγ ELISpot responses on day 0 and day 6 of gluten challenge: tTG-gliadin (results expressed as IFNγ specific spots/million PPBMC)

• Day 0: none (45), 1 week (62), 2 weeks (5), 8 weeks (5)
• Day 6: none (0), 1 week (67), 2 weeks (40), 8 weeks (60)

Result 3: Duration of gluten free diet and IFNγ ELISpot responses on day 0 and day 6 of gluten challenge: A-gliadin 57-73 P65 (results expressed as IFNγ specific spots/million PPBMC)

• Day 0: none (1), 1 week (2), 2 weeks (1), 8 weeks (1)
• Day 6: none (0), 1 week (0), 2 weeks (0), 8 weeks (0)

Result 4: Duration of gluten free diet and IFNγ ELISpot responses on day 0 and day 6 of gluten challenge: PPD (results expressed as IFNγ specific spots/million PPBMC)

• Day 0: none (90), 1 week (88), 2 weeks (210), 8 weeks (150)
• Day 6: none (0), 1 week (100), 2 weeks (210), 8 weeks (100)

Result 5: Duration of gluten free diet and IFNγ ELISpot responses on day 0 and day 6 of gluten challenge: tTG (results expressed as IFNγ specific spots/million PPBMC)

• Day 0: none (5), 1 week (4), 2 weeks (3), 8 weeks (2)
• Day 6: none (0), 1 week (4), 2 weeks (1), 8 weeks (2)

Gluten challenge in HLA-DQ2 coeliac disease on long term gluten

Characterization of anti-gliadin T cell response was carried out in peripheral blood on day 6-8 after 3-day gluten challenge.

Result 1: PBMC Day 6 Long-term gluten free diet (preincubation with anti-HLA-DR and -DQ antibody) (expressed as % inhibition)

• DR-: tTG-gliadin 100 mcg/ml (105), A-gliadin 57-73 QE65 50 mcg/ml (90), PPD 5 mcg/ml (30)
• DQ-: tTG-gliadin 100 mcg/ml (5), A-gliadin 57-73 QE65 50 mcg/ml (22), PPD 5 mcg/ml (78).

Result 2: PBMC Day 6 Long-term gluten free diet (expressed as % CD8-depleted PBMC response)

• B7 depletion: tTG-gliadin n=6 (7), A-gliadin 57-73 n=9 (6), PPD n=8 (62)
• AE depletion: tTG-gliadin n=6 (120), A-gliadin 57-73 n=9 (80), PPD n=8 (110).
• CD4 depletion: tTG-gliadin n=6 (10), A-gliadin 57-73 n=9 (9), PPD n=8 (10).

Therapeutic peptides include, but are not limited to

• QLQPFPQPQLPYPQPQS (AG01)
• QLQPFPQPQLPYPQPQP (AG02)
• QLQPFPQPQLPYPQPQL (AG03)
• QLQPFPQPQLPYLQPQP (AG04)
• QLQPFPRPQLPYPQPQP (AG05)
• QLQPFPQPQLPYSQPQP (AG06)
• QLQPFLQPQLPYSQPQP (AG07)
• QLQPFSQPQLPYSQPQP (AG08)
• QLQPFPQPQLSYSQPQP (AG09)
• PQLPYPQPQLPYPQPQP (AG10)
• PQLPYPQPQLPYPQPQL (AG11)
• PQPQPFLPQLPYPQPQS (AG12)
• PQPQPFPPQLPYPQPQS (AG13)
• PQPQPFPPQLPYPQYQP (AG14)
• PQPQPFPPQLPYPQPPP (AG015)

Briefly after oral antigen challenge, specificities of peripheral blood T cells reflect those of intestinal T cell clones. In peripheral blood, epitopes of intestinal T cell clones are sub-optimal compared to A-gliadin 57-73 QE65, which is an optimal α-gliadin epitope.

Example 17

Table 24 shows the results of analyses examining the 652 peptides with several patients challenged with wheat or rye.

References

1. 1. Molberg O, et al. Nature Med. 4, 713-717 (1998).
2. 2. Quarsten H, et al. Eur. J. Immunol. 29, 2506-2514 (1999).
3. 3. Greenberg CS et al. FASEB 5, 3071-3077 (1991).
4. 4. Mantzaris G, Jewell D. Scand. J. Gastroenterol. 26, 392-398 (1991).
5. 5. Mauri L, et al. Scand. J. Gastroenterol. 31, 247-253 (1996).
6. 6. Bunce M, et al. Tissue Antigens 46, 355-367 (1995).
7. 7. Olerup O, et al. Tissue antigens 41, 119-134 (1993).
8. 8. Mullighan CG, et al. Tissue-Antigens. 50, 688-92 (1997).
9. 9. Plebanski M et al. Eur. J. Immunol. 28,4345-4355 (1998).
10. 10. Anderson DO, Greene FC. The alpha-gliadin gene family. II. DNA and protein sequence variation, subfamily structure, and origins of pseudogenes. Theor Appl Genet (1997) 95:59-65.
11. 11. Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Van der Wal Y, Kooy YMC, Lundin KEA, Koning F, Roepstorff P, Sollid LM, McAdam SN. The intestinal T cell response to alpha-gliadin in adult celiac disease is focused on a single deamidated glutamine targeted by tissue transglutaminase. J Exp Med. 2000; 191:603-12.
12. 12. Vader LW, de Ru A, van der Wal, Kooy YMC, Benckhuijsen W, Mearin ML, Drijfhout JW, van Veelen P, Koning F. Specificity of tissue transglutaminase explains cereal toxicity in celiac disease. J Exp Med 2002; 195:643-649.
13. 13. van der Wal Y, Kooy Y, van Veelan P, Pena S, Mearin L, Papadopoulos G, Koning F. Selective deamidation by tissue transglutaminase strongly enhances gliadin-specific T cell reactivity. J Immunol. 1998; 161:1585-8.
14. 14. van der Wal Y, Kooy Y, van Veelan P, Pena S, Mearin L, Molberg O, Lundin KEA, Sollid L, Mutis T, Benckhuijsen WE, Drijfhout JW, Koning F. Proc Natl Acad Sci USA 1998; 95:10050-10054.
15. 15. Vader W, Kooy Y, Van Veelen P et al. The gluten response in children with celiac disease is directed toward multiple gliadin and glutenin peptides. Gastroenterology 2002, 122:1729-37
16. 16. Arentz-Hansen H, McAdam SN, Molberg O, et al. Celiac lesion T cells recognize epitopes that cluster in regions of gliadin rich in proline residues. Gastroenterology 2002, 123:803-809.

Coeliac disease subjects studied

1	64 f	14 yr	Homozygote	3 days	Abdominal pain, lethargy, mouth ulcers, diarrhoea
2	57 m	1 yr	Heterozygote	10 days	Lethargy, nausea
3	35 f	7 yr	Heterozygote	3 days	Nausea
4	36 m	6 wk	Homozygote	3 days	Abdominal pain, mouth ulcers, diarrhoea
5	26 m	19 yr	Heterozygote	3 days	None
6	58 m	35 yr	Heterozygote	3 days	None
7	55 m	1 yr	Heterozygote	3 days	Diarrhoea
8	48 f	15 yr	Homozygote	3 days	Abdominal pain, diarrhoea

Aminoacid at position 65	Range	Mean
Glutamate	(100)	100%
Asparagine	(50-84)	70%
Aspartate	(50-94)	65%
Alanine	(44-76)	64%
Cysteine	(45-83)	62%
Serine	(45-75)	62%
Valine	(24-79)	56%
Threonine	(46-66)	55%
Glycine	(34-47)	40%
Leucine	(8-46)	33%
Glutamine	(16-21)	19%
Isoleucine	(3-25)	14%
Methionine	(3-32)	14%
Phenylalanine	(0-33)	12%
Histidine	(0-13)	8%
Tyrosine	(0-17)	8%
Tryptophan	(0-17)	8%
Lysine	(0-11)	4%
Proline	(0-4)	2%
Arginine	(0-2)	1%

o TG	TG
(1-13)	QLQPFPQPQLPYPQPQS	57-73	α-Gliadin (T. aestivum) Q41545
100 (100)	QLQPFPQPELPYPQPQS	57-73	α-Gliadin (T. aestivum) Q41545
(1-7)	53(44-67)	QLQPFPQPQLPYSQPQP	77-93	α/β-Gliadin precursor (Tricetum. sestivum) P02863
76-92	α-Gliadin (T. aestivum) Q41528
77-93	α-Gliadin storage protein (T. aestivum) Q41531
57-73	α-Gliadin mature peptide (T. aestivum) Q41533
77-93	α-Gliadin precursor (T. spelta) Q9ZP09
2 (0-20)	83 (61-113)	QLQPFPQPQLPYPQPQP	77-93	α/β-Gliadin A-II precursor (T. aestivum) P0472
9 (0-33)	83 (74-97)	QLQPFPQPQLPYPQPQL	77-93	α/β-Gliadin A-IV precursor (T. aestivum) P04724
77-93	α/β-Gliadin MM1 precursor (T. aestivum) P18573
(0-7)	109 (41-152)	PQLPYPQPQLPYPQPQP	84-100	α/β-Gliadin A-IV precursor (T. aestivum) P04724
ID	PQLPYPQPQLPYPQPQL	84-100	α/β-Gliadin MM1 precursor (T. aestivum) P18573
(0-1)	3(0-7)	QLQPFLQPQLPYSQPQP	77-93	α/β-Gliadin A-I precursor (T. aestivum) P04721
77-93	α-Gliadin (T. aestivum) Q41509
(0-0)	2 (0-7)	QLQPFSQPQLPYSQPQP	77-93	α-Gliadin storage protein (T. aestivum) Q41530
ID	PQPQPFPPQLPYPQTQP	77-93	α/β-Gliadin A-III precursor (T. aestivum) P04723
7 (0-40)	24 (11-43)	PQPQPFPPQLPYPQPQS	82-98	α/β-Gliadin A-V precursor (T. aestivum) P04725
0 (0-30)	19 (11-33)	PQPQPFPPQLPYPQPPP	82-98	α/β-Gliadin clone PW1215 precursor (T. aestivum) P04726
82-98	α/β-Gliadin (T. urartu) Q41632
0 (0-30)	21 (11-33)	PQPQPFLPQLPYPQPQS	79-95	α/β-Gliadin clone PW8142 precursor (T. aestivum) P04726
79-95	α-Gliadin (T. aestivum) Q41529
79-95	α/β-Gliadin precursor (T. aestivum) Q41546

Wheat: α-gliadin	A-gliadin (57-73)	100 (0)
Wheat: ω-gliadin	AAG17702 (141-157)	PQ...........................F......QSE	32 (6.4)
Barley: C-hordein	Q40055 (166-182)	...QPFPL...............F............Q	2.3 (2.0)
Wheat γ-gliadin	P21292 (96-112)	...QTFPQ...............F......QPQ	2.1 (4.2)
Rye: secalin	Q43639 (335-351)	...QPSPQ...............F............Q	1.6 (1.4)
Barley: γ-hordein	P80198 (52-68)	...QPFPQ...............HQHQFP	-1.0 (1.8)
Wheat: LMW glutenin	P16315 (67-83)	LQ...QPIL............FS...Q...Q	-0.9 (1.0)
Wheat: HMW glutenin	P08489 (718-734)	HGYYPTS.........SGQGQRP	6.4 (4.0)
Wheat γ-gliadin	P04730 (120-136)	...QCCQQL......I...QQSRYQ	0.7 (0.9)
Wheat: LMW glutenin	P10386 (183-199)	...QCCQQL......I...QQSRYE	-0.7 (0.5)
Wheat: LMW glutenin	049958 (214-230)	...QCCRQL......I...EQSRYD	-1.1 (0.3)
Barley: B1-hordein	P06470 (176-192)	...QCCQQL......I...EQFRHE	1.8 (1.4)
Barley: B-hordein	Q40026 (176-192)	...QCCQQL......ISEQFRHE	0.5 (0.9)

2, 6	102/6, 501	201, 602	SVA	1 yr	+ (-)
2, 2	501	201	SVA	1 yr	+ (-)
2, 5	101/4/5, 501	201, 501	PVA	1 yr	+ (-)
2, 5	101/4-5, 501	201, 501	SVA	7 yr	+ (-)
2, 2	201, 501	201, 202	SVA	4 mo	+ (ND)
2, 2	201, 501	201, 202	SVA	2 yr	+(-)
2, 8	301-3, 501	201, 302	SVA	1 yr	+ (-)
2, 8	301-3, 501	201, 302/8	SVA	11 yr	ND (-)
2, 8	301-3, 501	201, 302	SVA	29 yr	+ (-)
2, 8	201, 301-3	202, 302	IEL	1 yr	+ (-)
6, 8	102/6, 301-3	602/15, 302/8	IEL	9 mo	- (ND)
8, 7	301-3, 505	302, 301/9-10	SVA	2 yr	- (-)
8, 8	301	302	SVA	1 yr	+ (+)

SVA subtotal villous atrophy, PVA partial villous atrophy, IEL increased intra-epithelial atrophy, GFD gluten-free diet, ND not done.

P04722 77-93	0 (-4 to 17)	0 (-5 to 9)	-2 (-3 to 0)
P04722 77-93 + tTG	0 (-5 to 4)	0 (-9 to 3)	0 (-4 to 11)
P04722 77-93 QE85	0 (-5 to 5)	0 (-3 to 4)	0 (-6 to 14)
P02863 77-93	0 (-4 to 13)	2 (-3 to 5)	-2 (-3 to 2)
P02863 77-93 + tTG	-1 (-5 to 4)	-1 (-4 to 11)	1 (-4 to 6)
P02863 77-93 QE85	0 (-4 to 13)	0 (-4 to 14)	-1 (-4 to 6)
Gliadin chymotrypsin	2 (-5 to 20)	20 (11 to 145)	92 (50 to 154)
Gliadin chymotrypsin + tTG	0 (-1 to 28)	55 (29 to 248)
Chymotrypsin	0 (-4 to 5)	1 (-4 to 11)	-2 (-5 to 5)	1 (-4 to 8)
Chymotrypsin + tTG	0 (-5 to 8)	6 (0 to 29)	-2 (-3 to 11)
Gliadin pepsin	4 (-4 to 28)	44 (10 to 221)
Gliadin pepsin + tTG	2 (-3 to 80)	61 (8 to 172)
Pepsin	0 (-4 to 10)	0 (-3 to 12)	0 (-2 to 3)	2 (-2 to 8)
Pepsin + tTG	0 (-3 to 8)	0 (-5 to 9)	1 (-6 to 3)	0 (-3 to 14)
PBS alone	4 (0 to 6)	2 (0 to 6)	4 (1 to 12)	4 (0 to 4)
PBS + tTG	3 (0 to 8)	3 (0 to 11)	4 (2 to 10)	4 (2 to 11)

Day 6 vs. Day 0: *P<0.05 **P,0.02, ***P<0.01 by one-tailed Wilcoxon Matched-Pairs Signed-Ranks test

Gliadin chymotrypsin	0.94 (0.4-9-0)	2.1 (0.8-6.8)*	3.2 (1.8 -4.2)**
Gliadin pepsin	1.4 (0.5-1.4)	1.4 (0.8-4.0)*	1.9 (1.1-4.4)**
P04722 77-93 Q85	6.5 (2.3-12)**
P04722 77-93 E85	0.7 (0.6-1.1)
P02863 77-93 Q85	7.5 (3.9-19.9)**
P02863 77-93 E85	1.0 (0.8-1.2)

TTG>no tTG: *P<0.05 **P,0.02, ***P<0.01 by one-tailed WilcoxonMatched-Pairs Signed-Ranks test

	59***	1.0		46***	1.4
	116**	1.7		50***	4.6
	24***	2.5		40***	1.7
	133***	1.1		30***	3.1
	26**	2.1		27**	1.4
	30**	1.2		17***	1.1
	32***	1.3		20***	0.9
	24***	1.5		83***	1
	10***	1.1		141***	1.1
	12***	2.1		22***	1.5
	17***	1.4		16**	1.8

Day 6 vs. Day 0 **P<0.02, ***P<0.01 by one-tailed Wilcoxon Matched-Pairs Signed-Ranks test

A1a1	AAA96525, EEWTA, P02863	A1b13	B22364, P04271
A1a2	CAB76963	A2a1	AAB23109, CAA35238, P18573, S10015
A1a3	AAA96276	A2a2	CAB76964
A1a4	CAA26384, S07923	A2b1	P04724, T06500, AAA348282
A1a5	AAA34280	A2b2	D22364
A1a6	P04728	A2b3	P04722, T06498, AAA34276
A1b1	CAB76962	A2b4	C22364
A1b2	CAB76961	A2b5	CAB76956
A1b3	BAA12318	A3a1	AAA34277, CAA26383, P04726, S07361
A1b4	CAB76960	A3a2	1307187B, A27319, S13333
A1b5	CAB76958	A3b1	AAA96522
A1b6	CAB76959	A3b2i	AAA34279, P04727,
A1b7	CAB76955	A3b2ii	CAA26385, S07924
A1b8	AAA96524	A3b3	A22364, AAA34278, AAB23108, C61218, P04725
A1b9	CAA10257	A4a	P04723, AAA34283, T06504
A1b10	AAA96523, T06282	A4b	E22364
A1b11	AAA17741, S52124	A4c	CAB76957
A1b12	AAA34281	A4d	CAB76954

			Gamma-gliadins
GI1a	P08079, AAA34288, PS0094, CAC11079, AAD30556, CAC11057, CAC11065, CAC11056	GI5a	AAK84774, AAK84772
GI1b	CAC11089, CAC11064, CAC11080, CAC11078, AAD30440	GI5b	AAK84773
GI1c	CAC11087	GI5c	AAK84776
GI1d	CAC11088	GI6a	JA0153, P21292, AAA34272, 1507333A
Glle	CAC11055	GI6b	AAK84777
GI2a	JS0402, P08453, AAA34289	GI6c	1802407A, AAK84775, AAK84780
GI2b	AAF42989, AAK84779, AAK84779	GI7	AAB31090
GI3a	AAK84778	GIIa	AAA34287, P04730, S07398
GI3b	CAB75404	GIIb	1209306A
GI3c	BAA11251	GIII1a	P04729
GI4	EEWTG, P06659, AAA34274	GIII1b	AAA34286

O1a	AAG17702
O1b	P02865
O1c	A59156

Claims (42)

1. An isolated peptide comprising at least one T cell epitope, wherein the peptide is not more than 50 amino acids in length and the T cell epitope comprises a sequence selected from FPQPQQPFP and a sequence obtainable by transglutaminase-deamidating the sequence FPQPQQPFP.
2. The peptide according to claim 1, which comprises the sequence FPQPEQPFP.
3. The peptide according to claim 1 or 2, wherein the peptide is 10-50 amino acids in length.
4. The peptide according to claim 1, 2 or 3, which comprises a sequence selected from QQPFPQPQQPFP and a sequence obtainable by transglutaminase-deamidating QQPFPQPQQPFP, and wherein optionally a modification is present on the N terminus.
5. The peptide according to any one of claims 1-4, which consists of a sequence selected from QQPFPQPQQPFPWQP or a sequence obtainable by transglutaminase-deamidating QQPFPQPQQPFPWQP, and wherein optionally a modification is present on the N terminus.
6. The peptide according to any one of claims 1-5, wherein a modification is present on the N terminus.
7. The peptide according to any one of claims 1-6, wherein a modification is present on the C terminus.
8. The peptide according to any one of claims 4-7, wherein said modifications are natural post-translational modifications.
9. The peptide according to any preceding claim, which is HLA-DQ2-restricted.
10. The peptide according to any one of claims 1-8, which is HLA-DQ8-restricted.
11. The peptide according to any preceding claim, comprising a wheat epitope.
12. The peptide according to any preceding claim, comprising a wheat epitope, a barley epitope, and a rye epitope.
13. A peptide as defined in any one of the preceding claims, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, for use in a method of therapy.
14. The peptide for use according to claim 13, which comprises the sequence FPQPEQPFP.
15. A composition comprising (a) a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, and (b) a second agent which is a peptide comprising at least one epitope comprising a sequence selected from PQPELPY and QLQPFPQPELPYPQPQS.
16. A composition comprising a first HLA-DQ2-restricted agent and a second HLA-DQ8-restricted agent, wherein the composition comprises a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.
17. A composition comprising an agent comprising a wheat epitope and an agent comprising a rye epitope, wherein the composition comprises a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.
18. A composition comprising an agent comprising a wheat epitope and an agent comprising a barley epitope, wherein the composition comprises a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.
19. A composition comprising an agent comprising a rye epitope and an agent comprising a barley epitope, wherein the composition comprises a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.
20. A composition comprising an agent comprising a wheat epitope, an agent comprising a barley epitope, and an agent comprising a rye epitope, wherein the composition comprises a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP.
21. The peptide as defined in claim 13 for use in tolerising an individual to a gliadin protein to suppress the production of a T cell response to a peptide as defined in claim 13.
22. The peptide as defined in claim 13 for use in treating or preventing coeliac disease.
23. A pharmaceutical composition comprising: (a) a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP; (b) a pharmaceutically acceptable carrier or diluent; and, optionally, (c) a peptide comprising at least one epitope comprising a sequence selected from PQPELPY and QLQPFPQPELPYPQPQS.
24. A method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual comprising:

    (a) contacting a sample from the host with at least one peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, and optionally, in addition, a peptide comprising at least one epitope comprising a sequence selected from PQPELPY and QLQPFPQPELPYPQPQS;

    (b) determining in vitro whether T cells in the sample recognise the peptide; recognition by the T cells indicating that the individual has, or is susceptible to, coeliac disease.
25. Use of a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, for the preparation of a diagnostic means for use in a method of diagnosing coeliac disease, or susceptibility to coeliac disease, in an individual, said method comprising determining whether T cells of the individual recognise the peptide, recognition by the T cells indicating that the individual has, or is susceptible to, coeliac disease.
26. A method or use according to claim 24 or 25 wherein the peptide is bound to (a) an HLA molecule, or (b) a fragment of an HLA molecule capable of binding the peptide.
27. A method or use according to claim 26 wherein the HLA molecule or fragment is in a complex comprising four HLA molecules or fragments of HLA molecules.
28. Use according to claim 25, 26 or 27 wherein the method comprises administering the peptide to the skin of an individual and detecting the presence of inflammation at the site of administration, the detection of inflammation indicating that the T cells of the individual recognise the peptide.
29. A method according to claim 24, 26 or 27 wherein the sample is a blood sample.
30. A method according to claim 24, 26, 27 or 29 wherein the T cells are not restimulated in antigen specific manner in vitro before the said determining.
31. A method or use according to any one of claims 24-30 in which the recognition of the peptide by the T cells is determined by detecting the secretion of a cytokine from the T cells.
32. A method or use according to claim 31 in which the cytokine is IFN-γ.
33. A method or use according to claim 31 or claim 32 in which the cytokine is detected by allowing the cytokine to bind to an immobilised antibody specific to the cytokine and then detecting the presence of the antibody/cytokine complex.
34. A method or use according to any one of claims 24 to 30 wherein said determining is done by measuring whether the peptide binds the T cell receptor.
35. A method for identifying an analogue, which analogue is a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, bound to an HLA molecule or fragment as defined in claim 26 or 27, which method comprises determining whether a candidate substance is recognised by a T cell receptor that recognises an epitope comprising a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, recognition of the substance indicating that the substance is an analogue.
36. A method of determining whether a composition is capable of causing coeliac disease comprising determining whether a protein comprising an oligopeptide sequence selected from FPQPQQPFP, QQPFPQPQQPFP and QQPFPQPQQPFP is present in the composition, the presence of the protein indicating that the composition is capable of causing coeliac disease.
37. A method according to claim 36 wherein the said determining is done by contacting the composition with an antibody specific for the said oligopeptide sequence, binding of the antibody to a protein in the composition indicating the composition is capable of causing coeliac disease.
38. An antibody which specifically binds to a sequence selected from FPQPQQPFP, QQPFPQPQQPFP and QQPFPQPQQPFPWQP.
39. A kit for carrying out a method or use according to any one of claims 24 to 34 comprising a peptide as defined in any one of claims 1 to 12, wherein the peptide comprises a sequence obtainable by transglutaminase-deamidating FPQPQQPFP, and a means to detect the recognition of the peptide by the T cell.
40. A kit according to claim 39 wherein the means to detect recognition comprises an antibody to IFN-γ.
41. A kit according to claim 40 wherein the antibody is immobilised on a solid support and optionally the kit also comprises a means to detect the antibody/IFN-γ complex.
42. Use of a peptide as defined in any one of claims 1-12 to produce an antibody specific to the peptide.