US20100324829A1

US20100324829A1 - Method of selecting hla-dp4 ligands and the applications thereof

Info

Publication number: US20100324829A1
Application number: US12/719,712
Authority: US
Inventors: Bernard Maillere; Florence Castelli; Cécile Buhot; Bertrand Georges
Original assignee: Commissariat a lEnergie Atomique CEA; SEDAC-Therapeutics
Current assignee: SEDAC-Therapeutics; Commissariat a lEnergie Atomique et aux Energies Alternatives CEA
Priority date: 2001-10-17
Filing date: 2010-03-08
Publication date: 2010-12-23
Also published as: WO2003040299A2; US20050059107A1; AU2002360128A1; CA2463705A1; EP1436626A2; US7718575B2; WO2003040299A3; FR2830940B1; FR2830940A1

Abstract

A method of selecting a test molecule that binds to HLA-DP4 by (i) incubating HLA-DP4 with the test molecule and a labeled peptide of formula (I) Z₁X₁X₂X₃X₄X₅X₆X₇X₈X₉Z₂thereby forming respective complexes, wherein Z₁and Z₂, which may be identical or different, are each either zero or from 1 to 100 amino acids; X₁is an aromatic or hydrophobic amino acid, or S; X₆is an aromatic or hydrophobic amino acid, or C; X₉is an aromatic or hydrophobic amino acid, or C, D, Q, S, T, or E; and X₂, X₃, X₄, X₅, X₇and X₈are each an amino acid; (ii) separating the respective complexes formed; (iii) detecting the HLA-DP4/labeled peptide complexes; and (iv) selecting the test molecule that exhibits a binding activity of IC₅₀<1000 nM, which corresponds to the concentration of the test molecule that inhibits 50% of the competitive HLA-DP4 binding of the labeled peptide.

Description

The present invention relates to a method of selecting HLA-DP4 ligands and to the applications thereof.
CD4+ T lymphocytes are among the main cells which regulate the immune response. They are antigen-specific and are in fact capable of recognizing the presence of a pathogenic agent, of an allergen or of a tumor cell, and of triggering an immune response. Recognition of these antigens in fact results in the activation of the CD4+ T lymphocytes, which secrete most of the cytokines necessary for the recruitment of effector cells, namely cytotoxic CD8+ lymphocytes and antibody-producing B lymphocytes. CD4+ T lymphocytes are also involved in the activation of cells by cell contacts and, for example induce the activation, via the CD40 molecule, of antigen-presenting dendritic cells. CD4+ T lymphocyte activation is a favorable prognostic in infections with viruses such as the human immunodeficiency virus (HIV), human papilloma viruses (HPVs) or the hepatitis C virus (HCV), and these cells appear to be necessary for antitumor immunity. Their role is not, however, systematically beneficial for the organism. Autoimmune diseases very commonly result from uncontrolled activation of CD4+ T lymphocytes. This is the case of multiple sclerosis and of insulin-dependent diabetes. These lymphocytes also participate in the establishment of allergic diseases. IL4, which is mainly secreted by CD4+ T lymphocytes, is in fact the main factor which results in the production of IgE. Finally, the role of CD4+ T lymphocytes in the triggering of transplant rejection is well established. Thus, depending on the disease, treatments are aimed at triggering the activation of CD4+ T lymphocytes (immunization against a pathogenic agent or a tumor cell) or at decreasing the state of activation of CD4+T lymphocytes (desensibilization against the allergy, prevention of transplant rejection).
The activation of CD4+ T lymphocytes takes place under the effect of the presentation of antigenic peptides by the molecules of the major histocompatibility complex type II borne by the antigen-presenting cells (APCs); in humans, they are called HLA II molecules for human leukocyte Antigen type II. These antigenic peptides, called T epitopes, result from the proteolytic degradation of the antigens by the antigen-presenting cells. They are variable in length, generally from 13 to 25 amino acids, and have a sequence which makes them capable of binding to the HLA II molecules. It is well known that, just like the native antigen, a T-epitope peptide is capable of stimulating, in vitro, CD4+ T lymphocytes which are specific for them, or of recruiting them in vivo. It is therefore sufficient to induce a CD4+ T response. Interestingly, it is also known that, according to the modes of presentation (route of administration, doses, possible addition of an adjuvant), the recognition of these peptides leads either to activation of the CD4+ T lymphocytes (ALEXANDER et al., J. Immunol., 2000, 164, 1625-1633; DEL GUERCIO et al., Vaccine, 1997, 15, 441-448; FRANKE et al., Vaccine, 1999, 17, 1201-1205), or to anergy thereof (MULLER et al., J. Allergy Clin. Immunol., 1998, 101, 747-754; OLDFIELD et al., J. Immunol., 2001, 167, 1734-1739). These T-epitope peptides are therefore capable both of contributing to the composition of a vaccine or of serving to decrease the undesired activation of CD4+ T lymphocytes. They are also capable of contributing to a test for diagnosing the immune state of patients or of normal individuals, based on the direct detection (lymphocyte proliferation assay) or indirect detection (production of antibodies, of cytokines, etc.) of said activated CD4+ T lymphocytes. One of the major problems which limits the use of these peptides is the identification of these epitopes, given that their sequence varies from one individual to another due to the polymorphism of the HLA II molecules. In fact, the HLA II molecules are hetero dimers consisting of an alpha (α)-chain and of a beta (β)-chain which are polymorphic. Four types of HLA II molecules exist per individual (2 HLA-DR, 1 HLA-DQ and 1 HLA-DP). The HLA-DR molecule, the beta (β)-chain of which is encoded by the DRB1 gene, is the most commonly expressed. To date, the betα-chain encoded by the DRB1 gene is the most polymorphic and has 273 alleles. For the HLA-DQ and HLA-DP molecules, the two chains (α and β) of which they are formed are polymorphic but they have fewer alleles. There are in fact 21 DQA1 alleles (α-chain of HLA-DQ), 45 DQB1 alleles (β-chain of HLA-DQ), 19 DPA1 alleles (α-chain of HLA-DP) and 93 DPB1 alleles (β-chain of HLA-DP). However, combination between the two α- and β-chains encoded by these alleles gives rise to many HLA-DQ and HLA-DP molecules. Because of this polymorphism, these isoforms possess binding properties which are different from one another, which implies that they can bind different peptides of the same antigen. Thus, each individual recognizes in an antigen a set of peptides, the nature of which depends on the HLA II molecules which characterizes said individual. Since a large number of HLA II alleles exists, it may be assumed that there exists, in a given sequence, a considerable repertoire of T-epitope peptides having very different sequences, each one specific for a different allele.
However, this HLA II molecule diversity is not as great on the scale of each population as on a worldwide scale, as illustrated by table I below.

TABLE I

Gene frequencies for the HLA class II molecules that are most abundant in the Caucasian
population (Europe, USA)*, according to J. Colombani, 1993, HLA: immune functions and medical
applications, published by John Libbey Eurotext.

Molecules	α-chain	Frequency (%)	β-chain	Frequency (%)	Abundant molecules

HLA-DR	DRA*0101	100	DRB1*0101	9.3	DRA0101/DRB10101
			DRB1*1501	8.0	DRA0101/DRB11501
			DRB1*0301	10.9	DRA0101/DRB10301
			DRB1*0401	5.6	DRA0101/DRB10401
			DRB1*1101	9.2	DRA0101/DRB11101
			DRB1*1301	6.0	DRA0101/DRB11301
			DRB1*0701	14.0	DRA0101/DRB10701
			DRB3*0101	9.2	DRA0101/DRB30101
			DRB3*0202	12.0	DRA0101/DRB10202
			DRB4*0101	28.4	DRA0101/DRB40101
			DRB5*0101	7.9	DRA0101/DRB50101
HLA-DQ	DQA1*0101	17.0	DQB1*0501	14.9	DQA10101/DQB10501
	DQA1*0102	15.8	DQB1*0602	9.8	DQA10501/DQB10301
	DQA1*0201	12.4	DQB1*0603	5.8	DQA10501/DQB10201
	DQA1*0301	14.5	DQB1*0201	21.3	DQA10301/DQB10302
	DQA1*0501	20.9	DQB1*0301	12.0	DQA10102/DQB10602
			DQB1*0302	13.0	DQA10201/DQB10201
HLA-DP	*DPA10103**	78.2	DPB1*0101	7.1	DPA10201/DPB10101
	DPA1*0201	21.2	DPB1*0201	11.9	DPA10103/DPB10201
			DPB1*0301	9.1	DPA10103/DPB10301
			*DPB10401**	40.1	*DPA10103/DPB10401*
			*DPB10402**	11.0	*DPA10103/DPB10402*

*The gene frequencies for the 2 HLA-DP4 molecules are indicated in bold.

Thus, for the HLA-DR and HLA-DQ molecules, about ten alleles are sufficient to cover more than 60% of the gene frequency found in the Caucasian population and therefore to involve more than 85% of the Caucasian population.
For the HLA-DP molecules, the most common molecule is the DP4 molecule derived from the DPA1*0103 and DPB1*0401 alleles, which have gene frequencies of 78.2% and 40%, respectively. Three other DP molecules have frequencies which exceed 5%: a DP3 molecule (DPA1*0103/DPB1*0301), a DP2 molecule (DPA1*0103/DPB1*0201) and a DP4 molecule (DPA1*0103/DPB1*0402); the HLA II molecules are named as a function of the DPB1 allele encoding the β-chain which is the most polymorphic.
Thus, four HLA-DP molecules (1 DP3 molecule, 1 DP2 molecule and 2 DP4 molecules) are therefore sufficient to cover 71% of the gene frequency of the Caucasian population, the two DP4 molecules covering 51% by themselves. Each of these DP4 molecules comprises a variable α-chain encoded either by DPA1*0103, which is the most common (78.2%), or by DPA1*0201 (20.2%), the two α-chains differing only at the level of 3 amino acids (positions 31, 50 and 83; table II). The HLA-DP4 molecules which exhibit a high allelic frequency in the Caucasian population (of the order of 50% in Europe and 80% in North America) are also present at not insignificant frequencies in the other populations (allelic frequency of the order of 60% in South America, 60% in India, 25% in Africa and 15% in Japan, Colombani et al., mentioned above).

TABLE II

Polymorphism of the α- and β-chains of the HLA-DP molecules in the Caucasian population
(Europe, USA), according to J. Colombani, mentioned above.
Polymorphic positions*

	Alleles	Freq (%)	8	9	11	37	38	57	58	59	67	71	78	86	87	88	89

β-chain	401	40.1	L	F	G	F	A	A	A	E	I	K	M	G	G	P	M
	402	11.0	—	—	—	—	V	D	E	—	—	—	—	—	—	—	—
	201	11.9	—	—	—	—	V	D	E	—	—	E	—	—	—	—	—
	501	1.3	—	—	—	L	V	E	—	—	—	—	—	D	—	—	—
	101	7.1	V	Y	G	Y	—	—	—	—	—	—	—	D	E	A	V
	301	9.1	V	Y	L	—	V	D	E	D	L	—	V	D	E	A	V
	901	1.1	V	H	L	—	V	D	E	D	I	E	V	D	E	A	V

		Alleles	Freq (%)	31	50	83

	α-chain	103	78.2	M	Q	T
		201	21.2	Q	R	A

*The sequences, available on the internet site http://www.ebi.ac.uk/imgt/hla, are numbered according to the numbering of STERN et al. (Nature, 1994, 368, 215-221) for the HLA-DR molecule; as the residues at positions 23 and 24 are absent in the DPB sequences, the residues indicated at positions 37, 38, 57, 59, 67, 71, 78, 86, 87, 88 and 89 correspond respectively to positions 35, 36, 55, 56, 57, 65, 69, 76, 84, 85, 86 and 87 in the DPB sequence.

However, despite the high frequencies, each of these DP4 molecules has only been very partially studied; the most common studies in fact result from the isolation of HLA-DP4 molecule-restricted CD4+ T lymphocyte clones specific for the peptides given in table III below, derived from the following antigens:

- Antigens from pathogenic microorganisms, such as the tetanus toxin (WYSS CORAY et al., Eur J. Immunol., 1992, 22, 2295), the Blastomyces dermatitidis WI-1 antigen (CHANG et al., Inf. Immun., 2000, 68, 502), the Mycobacterium Bovis hsp 65 protein (GASTON et al., Int. Immunol., 1991, 3, 965-972), the hepatitis B virus S antigen (HBsAg; CELIS et al., J. Virol., 1989, 63, 747), the rabies virus non-structural phosphoprotein (RV-NS) or influenza B virus neuraminidase (IBV-Nm; CELIS et al., J. Immunol., 1990, 145, 305) and the herpes simplex virus type 1 UL21 protein (KOELLE et al., J. Virol., 2000, 74, 10930-10938),
- allergens such as the major allergen of Dermatophagoides pteronyssinus (HIGGINS, J. Allergy Clin. Immunol., 1992, 90, 749), and
- tumor antigens such as MAGE-A3 (SCHULTZ et al., Cancer Res., 2000, 60, 6272) and NY-ESO1 (ZENG et al., P.N.A.S., 2001, 98, 3964-3969).

TABLE III

Sequences of the peptides which are ligands for the HLA-DP4 molecules

Peptides	Sequences	Identification number	Reference

TT947-967	FNNFTVSFWLRVPKVSASHLE	SEQ ID NO: 1	Wyss Coray et al., 1992

PSN 265	DPYNCDWDPYHEKYDWDLWNK	SEQ ID NO: 2	Chang et al., 2000
	WCN

Sld (HBsAg 19-28)	FFLLTRILTI	SEQ ID NO: 3	Celis et al., 1989

NS-p2 (RV NS 101-120)	GVQIVRQIRSGERFLKIWSQ	SEQ ID NO: 4	Celis et al. 1990

FLU-p1 (IBV Nm 247-260)	GISKCRFLKIREGR	SEQ ID NO: 5	Celis et al., 1990

MT451-466	IAFNSGMEPGVVAEKV	SEQ ID NO: 6	Gaston et al., 1991

MT456-471	GMEPGVVAEKVRNLSV	SEQ ID NO: 7	Gaston et al., 1991

Derp 101-119	PNAQRFGISNYCQIYP	SEQ ID NO: 8	Higgins, 1992

MAG 247-258	TQHFVQENYLEY	SEQ ID NO: 9	Schultz et al., 2000

NY-ESO1 161-180	MWITQCFLPVFLAQPPSGQR	SEQ ID NO: 10	Zeng et al., 2001

NY-ESO1 156-175	QLSLLMWITQCFLPVFLAQPP	SEQ ID NO: 11	Zeng et al., 2001

UL21 283-293	RELWWVFYAGD	SEQ ID NO: 12	Koelle et al., 2000

IL3 127-146	GPGAPADVQYDLYLNVANRR	SEQ ID NO: 13	Falk et al., 1994

UNK1 (UNK 1-17)	EKKYFAATQFEPLAARL	SEQ ID NO: 14	Falk et al., 1994

UNK2 (UNK 2-17)	KKYFAATQFEPLAARL	SEQ ID NO: 15	Falk et al., 1994

UNK3 (UNK 1-13)	EKKYFAATQFEPL	SEQ ID NO: 16	Falk et al., 1994

The abovementioned studies which are based on the isolation of HLA-DP4 molecule-restricted CD4+ T lymphocyte clones use a functional assay (proliferation assay) which has not made it possible to demonstrate a DP4 molecule-binding unit shared by all these peptides. In addition, this assay is very laborious to implement due to the fact that it requires a large number of correctly sampled patients, so that they represent the diversity of HLA II molecules of the entire population. In addition, the epitopes defined are those used by the immune system of the patients during natural infection with an antigen; these epitopes are not necessarily the most effective for inducing an immune response against this same antigen.
Using another approach, namely the analysis of four peptides naturally present on DP4 molecules [peptide 127-146 of the α-chain of the IL3 receptor (IL3 127-146; SEQ ID NO: 13) and three peptides of unknown origin: UNK1 (SEQ ID NO:14), UNK2 (SEQ ID NO:15) and UNK3 (SEQ ID NO:16); table III], FALK et al. (Immunogenetics, 1994, 39, 230-242) have proposed a DP4 molecule-binding consensus sequence. This sequence comprises three anchoring residues, respectively at positions P1, P7 and P9/P10. P1 and P7 are hydrophobic or aromatic (Y, V, L, F, I, A, M, W) and P9/P10 are preferably aliphatic; however, a Y residue is tolerated at positions P9/P10, but not an F residue. The residues P1 and P7 are preceded by groups of charged residues (K, R, E, N, Q for P1 and N, K, E for P7), and small residues (A, V) are common at position P3 and position P9. Table III shows that the only other DP4-ligand peptides exhibiting this unit are the overlapping peptides NY-ESO1 161-180 and NY-ESO1 156-175. Consequently this unit proposed by FALK et al. does not make it possible to define a DP4 molecule-binding unit shared by all the peptides that are ligands for the DP4 molecules, identified by functional assays.
Although DP9 molecule-binding assays (DONG et al., J. Immunol. 1995, 154, 4536-4545) and DP2 molecule-binding assays (CHICZ et al., J. Immunol., 1997, 159, 4935-4942), derived from those developed from the DR molecules [MARSHALL et al., J. Immunol., 1994, 152, 4946 (HLA-DR1); patent FR 99 0879 and TEXIER et al., J. Immunol., 2000, 164, 3177 (HLA-DR1, -DR2, -DR3, -DR4, -DR7, -DR11 and -DR13)] make it possible to isolate peptides specific for, respectively, DP9 and DP2, these assays do not make it possible to isolate peptides specific for DP4 (CHICZ et al., mentioned above): the peptides isolated using the DP2-binding assay have a high affinity for DP2 (binding activity <10 nM), whereas peptides known to be DP4-restricted, such as the HBsAg 14-33 peptide, exhibit moderate activity (20 μM).
In addition, due to the considerable differences in the residues of the binding site, between the principal DP molecules, the binding assays developed for these molecules do not make it possible to identify DP4-restricted peptides.
Thus, the binding units shared by all the peptides capable of binding to HLA-DP4 molecules have not been identified, in particular due to the fact that there is no method of identifying such peptides which is simple to implement and suitable for the simultaneous screening of a large number of peptides, such as banks of overlapping peptides representing the antigen sequence of interest.
Nevertheless, given the frequency of the DP4 molecules, the peptides which bind to the DP4 molecules constitute candidate peptides for specific immunotherapy and immunization and could be used for diagnosing the immune state of patients or of normal individuals.
The inventors have therefore developed a method of selecting HLA-DP4 ligands, which has enabled them to isolate HLA-DP4-specific ligands, in particular peptides, and to specify the binding units shared by the HLA-DP4-ligand peptides, based on the peptides obtained.

BRIEF DESCRIPTION OF THE DRAWINGS

Besides the above arrangements, the invention also comprises other arrangements, which will emerge from the following description which refers to examples of implementation of the subject of the present invention, with references to the attached drawings in which:

FIG. 1 illustrates the activity of binding of peptides to the HLA-DP4 molecules encoded, respectively, by DPB1*0401 (A) and DPB1*0402 (B), determined according to the method in accordance with the invention with, as tracer peptide, the biotinylated UNK1 peptide (10 nM); the percentage of DP4 molecule-binding is expressed as a function of the molar concentration of the peptides. The maximum binding (100%) corresponds to the value obtained for the tracer peptide alone, in the absence of competitor peptide. UNK: UNK1 (SEQ ID NO:14), IL: IL3 127-146 (SEQ ID NO:13), MAG: MAG 245-258 (SEQ ID NO:19), NSP2: SEQ ID NO:4, TT: TT 947-963 (SEQ ID NO:20), DQB: DQB 43-57 (SEQ ID NO:23), HCI 46-63: SEQ ID NO:24) and HA: HA 306-318 (SEQ ID NO:21),

FIG. 2 illustrates the correlation between the HLA-DP4-binding score (estimated by the method of identifying HLA-DP4-ligand peptides in accordance with the invention) and the affinity for the HLA-DP4 molecules (determined by the IC₅₀value, measured using the HLA-DP4-binding assay defined in the method of selecting HLA-DP4-ligand peptides in accordance with the invention) analyzed on a set of 44 peptides.

Consequently, a subject of the present invention is a method of selecting HLA-DP4-ligand molecules, comprising the following steps:

- (i) incubation of purified HLA-DP4 with a tracer consisting of a prelabeled peptide and able to be detected by an appropriate signal, which tracer peptide is chosen from the group consisting of the peptides exhibiting a signal/background noise ratio greater than 5, at a concentration of 10 nM, in a direct HLA-DP4-binding assay, and corresponding to general formula (I) Z₁X₁X₂X₃X₄X₅X₆X₇X₈X₉Z₂in which:
  - Z₁and Z₂, which may be identical or different, are zero or each represent a peptide of 1 to 100 natural or synthetic amino acids, preferably of 1 to 30 amino acids, even more preferably of 1 to 10 amino acids;
  - X₆represents an aromatic or hydrophobic amino acid, or a cysteine (C);
  - X₁represents an aromatic or hydrophobic amino acid and/or X₉represents an aromatic or hydrophobic amino acid, or a cysteine (C), an aspartic acid (D), a glutamine (Q), a serine (S), a threonine (T) or a glutamic acid (E); and
  - X₂, X₃, X₄, X₅, X₇and X₈each represent a natural or synthetic amino acid, in the presence of various concentrations of test molecule(s);
- (ii) separation of the various complexes formed;
- (iii) detection of the HLA-DP4/tracer peptide complexes by measuring the signal associated with said tracer peptide; and
- (iv) selection of the ligand molecules which exhibit a binding activity IC₅₀<1000 nM, corresponding to the concentration of these molecules which inhibits 50% of the binding of the tracer peptide.

The residues X₁, X₆and X₉of general formula (I) as defined above, which constitute the residues for anchoring in the pockets of the HLA-DP4 molecule, are also called, respectively, residues P1, P6 and P9. Among these residues, the residues X1 (or P1) and X6 (or P6) are the residues which provide the main contribution to the binding to HLA-DP4. The residue X₉or P₉is less important and provides less contribution to the binding to HLA-DP4.
For the purpose of the present invention:

- the expression “natural or synthetic amino acids” is intended to mean the 20 natural α-amino acids commonly found in proteins (one-letter code: A, R, N, D, C, Q, E, G, H, I, L, K, M, F, P, S, T, W, Y and V), certain amino acids rarely encountered in proteins (hydroxyproline, hydroxylysine, methyllysine, dimethyl-lysine, etc.), amino acids which do not exist in proteins, such as β-alanine, γ-aminobutyric acid, homocysteine, ornithine, citrulline, canavanine, norleucine, cyclohexylalanine, etc., and the enantiomers and diastereoisomers of the above amino acids;
- the term “hydrophobic amino acid” is intended to mean an amino acid selected from (one-letter code): A, V, L, I, P, W, F and M;
- the term “aromatic amino acid” is intended to mean an amino acid selected from (one-letter code): F, W and Y.

The use of a tracer peptide, as defined in step (i), makes it possible to effectively select ligands specific for HLA-DP4, i.e. molecules, in particular peptides, which exhibit good affinity for HLA-DP4, i.e. a binding activity <1000 nM.
A tracer peptide in accordance with the invention is selected by carrying out a direct HLA-DP4-binding assay, for example by following steps (i), (ii) and (iii) of the protocol defined above, but in the absence of competitor, corresponding to the test molecule. The appropriate signal detected (fluorescence, etc.) reveals the HLA-DP4/tracer peptide complexes [step (iii)] and the background noise represents the corresponding signal, obtained in the absence of HLA-DP4.
Preferably:

- X₆is selected from L, I, W, F, M, Y and C, and

X₁is selected from A, V, L, I, W, F, M and Y, and/or X₉is selected from A, V, L, I, P, W, F, M, Y, C, D, Q, S, T and E.
Tracer peptides in accordance with the invention are represented by the peptides NS-p2 (SEQ ID NO:4), MAG 247-258 (SEQ ID NO:9), UL21 283-293 (SEQ ID NO:12), IL3 127-146 (SEQ ID NO:13), UNK1 (SEQ ID NO:14), UL21 283-302 (SEQ ID NO:18) and MAG 245-258 (SEQ ID NO:19).
According to an advantageous embodiment of said method, the tracer peptide is chosen from the group consisting of biotinylated peptides, radiolabeled peptides and peptides coupled to a fluorochrome.
In step (ii), the separation of the formed complexes from the unbound peptides is carried out, for example, by transfer of the formed complexes onto a microtitration plate precoated with an HLA-DP-specific antibody, by chromatography on a gel filtration column or by centrifugation.
When the tracer peptide is radiolabeled or coupled to a fluorochrome, in particular to europium, the HLA-DP4/tracer peptide complexes are detected directly by measuring the radioactivity or the fluorescence emitted by said complexes.
When the tracer peptide is biotinylated, the HLA-DP4/tracer peptide complexes are detected indirectly using conjugated streptavidin, for example by immuno-enzymatic detection using streptavidin conjugated to an enzyme such as alkaline phosphatase, and a substrate for alkaline phosphatase such as 4-methylumbelliferyl phosphate (MUP).
According to another advantageous embodiment of said method, the tracer peptide is used at a concentration <200 nM, preferably less than 20 nM, even more preferably the tracer is used at a concentration of 10 nM.
According to yet another advantageous embodiment of said method, said HLA-DP4 in step (i) is chosen from the group consisting of the molecules encoded by the DPA1*103/DPB1*0401 and DPA1*103/DPB1*0402 alleles.
The method according to the invention advantageously makes it possible to select any HLA-DP4 ligand; this involves both mineral or organic molecules such as peptides and pseudopeptides.
According to another advantageous embodiment of said method, said test molecules represent a library of overlapping peptides covering the sequence of an antigen.
A subject of the present invention is also HLA-DP4 ligands which can be obtained by the method of selection as defined above, corresponding to a mineral or organic, natural or synthetic molecule exhibiting an HLA-DP4-binding activity of less than 1000 nM.
The HLA-DP4-binding activity of a ligand molecule, as defined above, corresponds to the concentration of said ligand molecule which inhibits 50% of the HLA-DP4-binding of a labeled tracer peptide, in a competition assay such as the method of selecting HLA-DP4 ligands defined above.
Among these ligand molecules, mention may in particular be made of peptides and modified peptides such as glycopeptides, lipopeptides, and peptides comprising D-amino acids, pseudopeptide bonds (pseudopeptides) or modifications of the C- or N-terminal ends.
The lipid portion of the ligand lipopeptide is in particular obtained by addition of a lipid unit to an α-amino function of said peptides or to a reactive function of the side chain of an amino acid of the peptide portion; it may comprise one or more chains, derived from C₄-C₂₀fatty acids, which are optionally branched or unsaturated (palmitic acid, oleic acid, linoleic acid, linolenic acid, 2-aminohexadecanoic acid, pimelautide, trimexautide) or derived from a steroid. The method of preparing such lipopeptides is in particular described in international applications WO 99/40113 and WO 99/51630. The preferred lipid portion is in particular represented by an N^α-acetyl-lysine N^ε (palmitoyl) group, also referred to as Ac-K (Pam).
According to an advantageous embodiment of said HLA-DP4-ligand peptide as defined above, its peptide sequence corresponds to general formula (I) Z₁X₁X₂X₃X₄X₅X₆X₇X₈X₉Z₂, in which:

- Z₁and Z₂, which may be identical or different, are zero or each represent a peptide of 1 to 10 amino acids as defined above, preferably of 1 to 30 amino acids, even more preferably of 1 to 100 amino acids;
- X₆represents an aromatic or hydrophobic amino acid, or a cysteine (C);
- X₁represents an aromatic or hydrophobic amino acid and/or X₉represents an aromatic or hydrophobic amino acid, or else a cysteine (C), an aspartic acid (D), a glutamine (Q), a serine (S), a threonine (T) or a glutamic acid (E); and
- X₂, X₃, X₄, X₅, X₇and X₈each represent a natural or synthetic amino acid, on condition that said HLA-DP4-ligand peptide of general formula (I) does not correspond to any of the sequences SEQ ID NOS: 1 to 17.

According to an advantageous embodiment of said ligand peptide:

- X₆is selected from L, I, W, F, M, Y and C, and
- X₁is selected from A, V, L, I, W, F, M and Y, and/or X₉is selected from A, V, L, I, P, W, F, M, Y, C, D, Q, S, T and E.

According to an advantageous arrangement of this embodiment, said ligand peptide binds specifically to DPB1*0401 (DPB1*0401-binding activity at least two times greater than the DPB1*0402-binding activity) and X₆is different from C, and/or X₁is different from A and from V, and/or X₉represents W or Y or X₉is different from E and C, and/or X₄is different from K and R.
According to another advantageous arrangement of this embodiment, said ligand peptide binds specifically to DPB1*0402 (DPB1*0402-binding activity at least two times greater than the DPB1*0401-binding activity) and X₆represents C, and/or X₁represents A or V, and/or X₉represents E or C or X₉is different from Y and W, and/or X₄represents K or R.
According to another advantageous embodiment of said ligand peptide, Z₁and Z₂are chosen from the group consisting of:

- the sequences of the antigen which are adjacent to the HLA-DP4-restricted CD4+ epitope as defined above, and/or
- one or more CD8+ T epitopes, and/or
- multiple CD4+ epitopes, such as peptide 830-846 of the tetanus toxin TT (O′SULLIVAN et al., J. Immunol., 1991, 147, 2663-2669), peptide 307-319 of influenza virus hemagglutinin HA (O′SULLIVAN et al., mentioned above), the Pan DR or PADRE epitope (ALEXANDER et al., DEL GUERCIO et al., FRANKE et al., mentioned above) and peptides derived from the antigens of Plasmodium falciparum, such as the CS.T3 peptide (SINIGAGLIA et al., Nature, 1988, 336, 778-780) and the CSP, SSP2, LSA-1 and EXP-1 peptides (DOOLAN et al., J. Immunol., 2000, 165, 1123-1137) and/or
- one or more B epitopes, for example a peptide or a glycopeptide in which said B epitope consists of a sugar (ALEXANDER et al., mentioned above).

Such sequences advantageously make it possible to trigger or to modulate an immune response in an appropriate manner.
According to yet another advantageous embodiment of said ligand peptide, it has the sequence SEQ ID NO:84 corresponding to the NY-ESO1 87-111 peptide.
The present invention also encompasses the ligand peptides as defined above, which have been polymerized.
A subject of the present invention is also a method of identifying HLA-DP4-ligand peptides as defined above, based on an amino acid sequence, characterized in that it comprises at least the following steps:

- a) establishing an HLA-DP4-binding matrix by calculating—for all the mutants of a tracer peptide as defined above, representing all the substitutions of the residues at position 1, 4, 6 or 9 of said tracer peptide with the other 19 natural amino acids—the ratio of the IC₅₀values for said mutant peptides and for said tracer peptide, using the HLA-DP4-binding assay as defined in the method above,
- b) evaluating the HLA-DP4-binding of peptides of at least 9 amino acids included in said amino acid sequence, by calculating, for each fragment of 9 amino acids of said peptide, the sum of the scores for HLA-DP4-binding of the residues at positions 1, 4, 6 and 9 of said fragment, using the binding matrix established in a), and
- c) identifying the HLA-DP4-ligand peptides corresponding to those to which the loss of HLA-DP4-binding, relative to said tracer peptide, is smallest, i.e. those for which the sum of the binding scores has the lowest value expressed as a logarithm (log); preferably less than 2, preferably less than 1, even more preferably close to 0.

This HLA-DP4-binding matrix which is illustrated for the reference peptide UNK 3-15 (SEQ ID NO:28), in table XIV of example 6, makes it possible to estimate the HLA-DP4-binding activity of any peptide of at least 9 amino acids; peptides having binding scores of, respectively, 0, 1, 2, 3 and 4 correspond to peptides exhibiting a 1-fold, 10-fold, 100-fold, 1000-fold and 10 000-fold loss of binding relative to the UNK 3-15 peptide (IC ₅₀10 nM), i.e. exhibiting an estimated binding activity of, respectively, 10 nM, 100 nM, 1000 nM, 10 μM and 100 μM.
For example, the tumor antigen NY-ESO1 comprises a peptide WITQCFLPV having a DP*0401- and DP*0402-binding score of 0+0.3+0+0.3=0.6 corresponding to an estimated binding activity (using the HLA-DP4-binding matrix) of 39 nM and a calculated activity (using the DP*0401- and DP*0402-binding assay) of, respectively, 20 nM and 67 nM.
The method of identifying HLA-DP4-ligand peptides according to the invention, which is easy to implement and can be automated, makes it possible, in particular using appropriate software, to predict the sequence of the HLA-DP4-ligand peptides present in all proteins representing antigens of interest. The peptide sequences thus identified can then be verified using an HLA-DP4-binding assay, defined in the method of selecting HLA-DP4 ligands according to the invention.
A subject of the present invention is also a nucleic acid molecule, characterized in that it encodes a ligand peptide as defined above.
The subject of the invention also encompasses the recombinant nucleic acid molecules comprising at least one nucleic acid molecule in accordance with the invention, linked to at least one heterologous sequence.
For the purpose of the present invention, the expression “sequence which is heterologous relative to a nucleic acid sequence encoding a ligand peptide” is intended to mean any nucleic acid sequence other than those which, naturally, are immediately adjacent to said nucleic acid sequence encoding a peptide.
The subject of the present invention encompasses in particular:

- expression cassettes comprising at least one nucleic acid molecule in accordance with the invention and the sequences required for controlling the transcription and the translation of said nucleic acid molecule (promoter, intron, initiation codon (ATG), stop codon, polyadenylation signal), and
- recombinant vectors comprising an insert consisting of a nucleic acid molecule in accordance with the invention. Advantageously, these expression vectors comprise at least one expression cassette as defined above.

Many vectors into which a nucleic acid molecule of interest can be inserted in order to introduce it into and to maintain it in a eukaryotic or prokaryotic host cell are known in themselves; the choice of a suitable vector depends on the use envisioned for this vector (for example, replication of the sequence of interest, expression of this sequence, maintenance of this sequence in extrachromosomal form, or else integration into the host's chromosomal material), and also on the nature of the host cell.
For example, use may be made, inter alia, of viral vectors such as adenoviruses, retroviruses, lentiviruses and AAVs, into which the sequence of interest has been inserted beforehand; said sequence (isolated or inserted into a plasmid vector) can also be combined with a substance which allows it to cross the host cell membrane, for example a preparation of liposomes, of lipids or of cationic polymers, or it can be injected directly into the host cell, in the form of naked DNA.
A subject of the invention is also prokaryotic or eukaryotic cells transformed with at least one nucleic acid molecule in accordance with the invention.
Transformed cells in accordance with the invention can be obtained by any means, known in themselves, making it possible to introduce a nucleic acid molecule into a host cell. For example, in the case of animal cells, use may be made of the vectors or the lipid preparations as defined above.
A subject of the present invention is also an immunomodulatory composition, characterized in that it comprises at least one HLA-DP4 ligand or a nucleic acid molecule encoding an HLA-DP4-ligand peptide as defined above, and a pharmaceutically acceptable vehicle.
Advantageously, said nucleic acid molecule is included in a vector as defined above.
Depending on the choice of the HLA-DP4 ligand and of its mode of presentation (route of administration, dose, possible addition of adjuvant) such an immunomodulatory composition results in either activation of T lymphocytes, or the anergy thereof.
In fact, it has been shown that a single subcutaneous injection of a small amount of peptide results in anergy (OLDFIELD et al., J. Immunol., 2001, 167, 1734-1739). On the other hand, it is known that repeated injections of large amounts of peptide in the presence of adjuvant result in T lymphocyte activation. In addition, culmination with Z₁and Z₂peptides as defined above also makes it possible to increase the antigen-specific immune response (ALEXANDER et al., mentioned above).
Consequently, said composition is used both for immunization against a pathogenic agent or a tumor cell and for the treatment of autoimmune diseases (multiple sclerosis, insulin-dependent diabetes), of allergy or of transplant rejection.
A subject of the present invention is also a reagent for diagnosing the immune state of an individual, characterized in that it comprises at least one HLA-DP4 ligand as defined above, optionally labeled or complexed, in particular complexed with labeled (biotinylated) HLA-DP4 molecules, in the form of multimeric complexes such as tetramers.
A subject of the present invention is also the use of an HLA-DP4 ligand or of a nucleic acid molecule encoding an HLA-DP4-ligand peptide as defined above, for preparing an immunomodulatory medicinal product or a reagent for diagnosing the immune state of an individual.
For the purpose of the present invention, the expression “diagnosing the immune state of an individual” is intended to mean detecting the presence, in said individual, of CD4+ T lymphocytes specific for an antigen derived from a pathogenic agent or from a tumor cell, for an allergen, for an alloantigen or for an autoantigen.
The reagent in accordance with the invention which is capable of detecting the presence of CD4+ T lymphocytes specific for an antigen is used for detecting: an infection with a pathogenic agent, a cancer, an autoimmune disease, an allergy or a transplant rejection, based on a biological sample from a patient.
A subject of the present invention is also a method of diagnosing the immune state of an individual, comprising the steps of:

- bringing a biological sample from said individual into contact with a diagnostic reagent as defined above, and
- detecting the CD4+ T lymphocytes specific for an antigen by any suitable means.

A subject of the present invention is also a kit for detecting the immune state of an individual, characterized in that it comprises at least one reagent as defined above, combined with a means for detecting CD4+ T lymphocytes specific for an antigen.
The CD4+ T lymphocytes specific for an antigen are detected by any means known in themselves. For example, use may be made of direct means such as lymphocyte proliferation assays or flow cytometry in the presence of multimeric complexes as defined above, or else indirect means, for instance the assaying of cytokines such as IL2, IL4, IL5 or γIFN, in particular by immunoenzymatic techniques (ELISA, RIA, ELISPOT).
More precisely:

- as regards the proliferation assay:

A suspension of cells (PBMCs, CD8+ cell-depleted PBMCs, T lymphocytes pre-enriched by a step consisting in culturing in vitro with the peptides as defined above, or cloned T lymphocytes) is cultured for 3 to 5 days in the presence of said HLA-DP4 ligands and, as needed, of appropriate presenting cells, such as dendritic cells, autologous or heterologous PBMCs, lymphoblastoid cells such as those obtained after infection with the EBV virus or genetically modified cells. The proliferation of the cells is measured by incorporation of tritiated thymidine into the DNA of the cells. The peptides as defined above make it possible to detect in the initial suspension the presence of cells specific for these peptides.

- as regards the ELISPOT assay:

The ELISPOT assay makes it possible to detect the presence of γIFN-secreting T cells specific for a peptide as defined above.
More precisely, the T cells are detected by measuring the secretion of γIFN after incubation of the PBMCs from patients with said peptides in accordance with the method described in International Application WO 99/51630 or Gahéry-Ségard et al., (J. Virol., 2000, 74, 1964).

- as regards the use of multimeric complexes and in particular of tetrameric complexes:
  - a biological sample, preferably peripheral blood mononuclear cells (PBMCs), is brought into contact with tetrameric complexes as defined above, and
  - labeled cells are analyzed by flow cytometry.

Advantageously, prior to bringing the biological sample into contact with said complex, it is enriched in CD4+ T cells by bringing it into contact with anti-CD4 antibodies so as to enrich said sample.
The tetramers are prepared as specified, for example, in NOVAK et al. (J. Clin. Investig., 1999, 104, R63-R67) or in KURODA et al. (J. Virol., 2000, 74, 18, 8751-8756).
Briefly, the tetramers are produced by incubating, for 72 hours at 37° C. and in a 10 mM citrate phosphate buffer containing 0.15 M NaCl, at a pH of between 4.5 and 7, soluble and biotinylated HLA II molecules with a 10-fold excess of HLA-DP4 ligands as defined above.
The tetramerized form is obtained by adding streptavidin labeled with a fluorochrome to the preparation, in an amount four times less (mole for mole) than the amount of HLA II molecules. The entire mixture is incubated overnight at ambient temperature.
To use these tetramers, a suspension of cells (PBMCs, CD8+ cell-depleted PBMCs, T lymphocytes pre-enriched by a step consisting in culturing in vitro with the HLA-DP4 ligands as defined above, or cloned T lymphocytes) is brought into contact with one or more tetramers (10 to 20 mg/ml) for 1 to 3 hours. After washing, the suspension is analyzed by flow cytometry: the labeling of the cells with the tetramers is visualized by virtue of the fact that these constructs are fluorescent.
The flow cytometry makes it possible to separate the cells labeled with the tetramers from the nonlabeled cells and to thus perform cell sorting.
A subject of the present invention is thus also a method of sorting CD4+ T lymphocytes specific for an antigen, characterized in that it comprises at least the following steps:

- bringing a cell sample into contact with tetramers labeled with a fluorochrome, prepared from complexes between HLA-DP4 ligands as defined above and soluble HLA-DP4 molecules, and
- sorting these cells bound to said tetramers, by flow cytometry.

- FIG. 1 illustrates the activity of binding of peptides to the HLA-DP4 molecules encoded, respectively, by DPB1*0401 (A) and DPB1*0402 (B), determined according to the method in accordance with the invention with, as tracer peptide, the biotinylated UNK1 peptide (10 nM); the percentage of DP4 molecule-binding is expressed as a function of the molar concentration of the peptides. The maximum binding (100%) corresponds to the value obtained for the tracer peptide alone, in the absence of competitor peptide. UNK: UNK1 (SEQ ID NO:14), IL: IL3 127-146 (SEQ ID NO:13), MAG: MAG 245-258 (SEQ ID NO:19), NSP2: SEQ ID NO:4, TT: TT 947-963 (SEQ ID NO:20), DQB: DQB 43-57 (SEQ ID NO:23), HCI 46-63: SEQ ID NO:24) and HA: HA 306-318 (SEQ ID NO:21),
- FIG. 2 illustrates the correlation between the HLA-DP4-binding score (estimated by the method of identifying HLA-DP4-ligand peptides in accordance with the invention) and the affinity for the HLA-DP4 molecules (determined by the IC₅₀value, measured using the HLA-DP4-binding assay defined in the method of selecting HLA-DP4-ligand peptides in accordance with the invention) analyzed on a set of 44 peptides.

EXAMPLE 1

Principle of the HLA-DP4/Peptide Binding Assay

1) Peptide Preparation

All the peptides were synthesized according to the Fmoc strategy in parallel solid-phase synthesis, purified by HPLC and controlled by mass spectrometry (ES-MS).
The peptides are biotinylated on their NH₂-terminal residue, according to the protocol as described in Texier et al., mentioned above.

2) Antibody Preparation

The HLA-DP molecule-specific antibodies, such as the antibody B7/21 (WATSON, et al., Nature, 1983, 304, 358-361), are purified from the culture supernatant of the corresponding hybridomas, on protein A-sepharose columns. These antibodies are then coupled onto sepharose 4B or protein A-sepharose columns for purification of the HLA-DP4 molecules.
More precisely, after centrifugation at 1100 g, the culture supernatant of the B7/21 antibody-producing cells is filtered through 0.22 μm and its pH is adjusted to 7-8 with 0.1 M Tris-HCI buffer, pH 8. This supernatant is then applied to a 10 ml protein A-sepharose 4 fast flow column prewashed with 100 ml of 0.1 M Tris-HCI buffer, pH 8. The column is then washed with 100 ml of 0.1 M Tris-HCI buffer, pH 8, and 100 ml of 0.01 M Tris-HCI buffer, pH 8. The antibodies are eluted with 0.1 M glycine-HCI buffer, pH 3. The column is rinsed with 100 ml of 0.1 M Tris-HCI buffer, pH 8. The eluted fraction which contains the B7/21 antibody is immediately neutralized with 1 M Tris-HCI buffer, pH 8, before being thoroughly dialyzed against 0.1 M borate buffer, pH 8.2. The amount of antibodies obtained is determined based on the optical density (OD) at 278 nm.
The affinity columns intended for purification of the HLA-DP4 molecules are prepared in the following way: 0.75 g of protein A-sepharose 4B (3 ml of final gel) is swollen in water and then in 0.1 M borate buffer pH 8.2. 15 mg of monoclonal antibody, such as B7/21, in 0.1 M borate buffer, pH 8.2, are added to the gel, centrifuged beforehand. The coupling is carried out for two hours at ambient temperature and then controlled by absorbence of the supernatant at 278 nm. The gel is then washed successively with 100 ml of 0.1 M borate buffer, pH 8.2, 120 ml of 0.2 M triethanolamine buffer, pH 8.2, 120 ml of 20 mM dimethylpyrimidate buffer, 0.2 M triethanolamine, pH 8.2 and 150 ml of 0.2 M ethanolamine buffer, pH 8.2. After the gel has been poured, it is rinsed with 150 ml of 0.1 M borate buffer, pH 8.2. The final control for the coupling is performed by an elution in 0.1 M glycine buffer, pH 2.5, containing 0.5 M NaCl; the absorbence at 278 nm of the 1 ml fractions should be less than 0.1. The column is immediately rinsed with 50 ml of 0.1 M borate buffer, pH 8.2, and 20 ml of 0.1 M borate buffer containing 0.02% NaN₃, pH 8.2. It is stored in this buffer at 4° C. until use.

3) Purification of the HLA-DP4 Molecules

The HLA-DP4 molecules are purified from various human B lymphocyte lines transformed with the Epstein Barr virus (EBV), which are homozygotes for DP, by immunoaffinity using monoclonal antibodies specific for all DP molecules. The origin of the lines and the alleles which characterize them are indicated in table IV.

TABLE IV

	DP	DPA1	DPB1
Lines	specificity	allele	allele	Reference

HOM2	DP4	DPA1*0103	DPB1*0401
BOLETH	DP4	DPA1*0103	DPB1*0401
PITOUT	DP4	DPA1*0103	DPB1*0401	SOUTHWOOD et
				al., mentioned
				above
HHKB	DP4	DPA1*0103	DPB1*0401	DAVENPORTH et
				al., P.N.A.S.,
				1995, 92, 6567
SHU	DP4	DPA1*0103	DPB1*0402
MLF	DP4	DPA1*0103	DPB1*0402
BM92	DP4	DPA1*0103	DPB1*0402

the origin of the lines is described on the internet site of the European cell culture collection (http://fuseiv.co.uk/camr/).

The HLA-DP4 molecules are purified from a pellet of these EBV-transformed human cells, according to a protocol derived from those used for the HLA-DR molecules (GORGA et al., J. Biol. Chem., 1987, 262, 16087; TEXIER et al., mentioned above).
More precisely, 5 to 6×10⁹cells are lysed at a concentration of approximately 10⁸cells/ml in lysis buffer (0.01 M Tris, 0.15 M NaCl, 0.02% NaN₃, pH 7, 1% NP40, 10 μg/ml aprotinin, 5 mM EDTA, 10 μM PMSF) in ice for 30 minutes. The large cell debris is removed from the lysis medium by centrifugation at 1100 g for 10 minutes at 4° C. The supernatant is then ultracentrifuged at 100 000 g at 4° C. for 1 hour. The rest of the purification takes place in a cold room at 4° C. The lysate is passed successively over a sepharose 4B column (10 ml of gel prepared in 1×PBS), a protein A-sepharose 4B column (5 ml of gel prepared in 1×²PBS) and then over the anti-DP affinity column. The columns are then rinsed with 250 ml of lysis buffer. The sepharose 4B column is discarded. The protein A-sepharose 4B column is rinsed with 25 ml of 1×TBS buffer (0.01 M Tris, 0.15 M NaCl, 0.02% NaN₃, pH 7), 50 ml of (0.1 M glycine; 0.5 M NaCl, pH 2.5) buffer and 200 ml of 1×PBS buffer, before being stored at 4° C. The anti-DP column is rinsed with 250 ml of TBS buffer containing 1 mM of dodecyl maltoside (DM). It is then eluted individually with the elution buffer (100 mM Na₂CO₃, 500 mM NaCl, 0.02% NaN₃, 1.1 mM DM, pH 11.5) in 15 fractions of 3 ml. The eluate is immediately neutralized with 10% of buffer (2 M Tris-HCI, pH 6.8), and then thoroughly dialyzed at 4° C. against 1×PBS buffer containing 1 mM of DM.

4) HLA-DP4-Peptide Binding Assay

The assay for binding of the peptides to the HLA-DP4 molecules is a competition assay with immunoenzymatic detection, derived from that developed for HLA-DR molecules (HLA-DR1: MARSHALL et al., mentioned above), HLA-DR1, -DR2, -DR3, -DR4, -DR7, -DR11 and -DR13: patent FR 99 0879 and TEXIER et al., mentioned above). It is carried out in 96-well plates, which makes it possible to study many samples in the same experiment. Briefly, the purified HLA-DP4 molecules are incubated with a biotinylated peptide which serves as a tracer, and various concentrations of the test peptide. The biotinylated peptide is a DP4-ligand peptide; it is a peptide recognized by DP4-specific CD4+ T lymphocytes, such as those specified in table III above, or a new peptide isolated using the present DP4-binding assay. Among these peptides, mention may be made, for example, of the UNK1 peptide or the IL3 127-146 peptide. The incubation is carried out in a buffer, the pH of which can vary. It is generally 5, and the incubation generally lasts 24 h. After incubation, the samples are neutralized, and then 100 μl of each sample is transferred onto an ELISA plate pre-coated with an anti-DP antibody, such as B7/21. The HLA-DP molecule/biotinylated peptide complexes attached to the bottom of the plate via the antibody are revealed by means of streptavidin phosphatase and a fluorescent substrate. The activity of each peptide is characterized by the concentration which inhibits 50% of the binding of the biotinylated peptide (IC₅₀).

EXAMPLE 2

Determination of the Binding Assay Parameters

1) Tracer Peptide

a) Materials and Methods

The binding of peptides to the two DP4 molecules (DP401 encoded by the DPA1*0103/DPB1*0401 allele and DP402 encoded by the DPA1*0103/DPB1*0402 allele) was analyzed by ELISA using the following direct binding assay:
The HLA-DP4 molecules purified according to the protocol described in example 1 are diluted 10 times in 10 mM phosphate buffer (1/10 dilution). They are then incubated with various concentrations of a biotinylated peptide (10⁻⁶M, 10⁻⁷M and 10⁻⁸M) in 10 mM phosphate buffer containing 150 mM NaCl, 1 mM DM, 10 mM citrate and 0.003% thimerosal, pH 5, in 96-well polypropylene plates, for 24 h at 37° C. Samples without DP4 molecules are used as a control. At the end of the incubation, the samples are neutralized with 50 μl of 450 mM Tris-HCI buffer, pH 7.5, containing 0.003% thimerosal, 0.3% BSA and 1 mM DM. They are then transferred onto 96-well maxisorp ELISA plates onto which the anti-DP antibodies have been pre-adsorbed. Specifically, 10 μg/ml of anti-DP antibodies were incubated overnight at 4° C. (100 μl/well) and the plates were then saturated with the 100 mM Tris-HCI buffer, pH 7.5, containing 0.3% BSA and 0.003% thimerosal overnight at 4° C. The incubation of the samples on these plates is carried out for two hours at ambient temperature, like the remainder of the assay, and then extensive washing is performed, between each step, in 0.1 M Tris-HCI buffer, pH 7.5, containing 0.05% Tween-20. The biotinylated peptide bound to the HLA-DP molecules is detected by adding 100 μl/well of the streptavidin-alkaline phosphatase conjugate (45 minutes) diluted 1/2000 in the 10 mM Tris buffer, pH 7, containing 0.15 M NaCl, 0.05% Tween 20, 0.2% BSA and 0.003% thimerosal, and then by adding 200 μl/well of 100 μM MUP substrate in 0.05 M NaHCO₃buffer, pH 9.8, containing 1 mM MgCl₂. The emission of fluorescence by the product of the enzymatic reaction is measured at 450 nm after excitation at 365 nm, and the ratio of the values obtained with or without DP4 is determined (R_F=fluorescence value in the presence of DP4/fluorescence value in the absence of DP4).

b) Results

The DP401 molecule-binding of peptides derived from the DP4-ligand peptides described above (table III) was assayed:

- peptide bUL21 283-302 (RELWWVFYAGDRALEEPHAE; SEQ ID NO: 18)
- peptide bIL3 127-146 (SEQ ID NO:13)
- peptide bMAG 245-258 (LLTQHFVQENYLEY; SEQ ID NO:19)
- peptide bMT 451-466 (SEQ ID NO:6)
- peptide bNS-p2 (SEQ ID NO:4)
- peptide bTT 947-963 (FNNFTVSFWLRVPKVSA; SEQ ID NO:20)
- peptide bUNK1 (SEQ ID NO:14)

Two peptides specific, respectively, for DR1 and for DR7 were used as control for the DP4-binding specificity:

- peptide bHA 306-318 (PKYVKQNTLKLAT; SEQ ID NO:21) described by HILL et al., J. Immunol., 1994, 152, 2890, and
- peptide bYKL (AAYAAAKAAALAA; SEQ ID NO:22) described by MARSHALL et al., mentioned above.

The results are given in table V.

TABLE V

Selection of the tracers using a direct HLA-
DP4-binding assay

	R_F= fluorescence value in
	the presence of DP4/
	fluorescence value in
	the absence of DP4

	peptide	10⁻⁶M	10⁻⁷M	10⁻⁸M

bUL21 283-302	8.3	13.3	10.1
bIL3 127-146	10	20	17.3
bMAG 245-258	1.5	6.7	5.4
bMT 451-466	5.5	4	2
bNS-p2	1	4.2	7.8
bTT 947-963	1.7	3.6	2.7
bUNK1	21.6	21.7	20
bHA 306-318	4.9	4.7	2.7
bYKL	4.8	3.2	3

The results show that:

- the bUNK1 and bIL3 127-146 peptides have a high DP4-binding capacity,
- the bTT 947-963 and bMT 451-466 peptides have a low binding capacity, and
- the bNS-p2, pUL21 283-302 and bMAG 245-258 peptides have an intermediate binding capacity.

The peptides exhibiting an R_F>5 at the concentration of 10⁻⁸M are considered to be good tracers which can be used in the binding assay.
The UNK1 peptide, which gave the best results, was used to optimize the binding assay.

2) Time, pH, Concentration of the Tracer Peptide

In order to have a sensitive and DP4-specific assay, the concentration of HLA-DP4 molecules, the concentration of the biotinylated peptide, the pH and the peptide/HLA-DP4 molecule incubation time were optimized with the bUNK1 peptide.

a) Materials and Methods

The HLA-DP4 molecules purified according to the protocol described in example 1 were diluted 1/10, 1/20, 1/40 and 1/80 in 10 mM phosphate buffer containing 150 mM of NaCl, 1 mM DM, 10 mM citrate and 0.003% thimerosal, at various pH values (pH 4, 5, 5.5, 6, 6.5 and 7), with the bUNK1 peptide at various concentrations and several concentrations of competitor peptides, in 96-well polypropylene plates. At the end of the incubation at 37° C., the samples were neutralized with 50 μl of 450 mM Tris-HCI buffer, pH 7.5, containing 0.003% thimerosal, 0.3% BSA and 1 mM DM. They were then transferred onto 96-well maxisorp ELISA plates onto which the anti-DP antibodies had been pre-adsorbed. Very specifically, 10 μg/ml of anti-DP antibodies were incubated overnight at 4° C. (100 μl/well) and the plates were then saturated with the 100 mM Tris-HCI buffer, pH 7.5, containing 0.3% BSA and 0.003% thimerosal overnight at 4° C. The incubation of the samples on these plates was carried out for two hours at ambient temperature, like the remainder of the assay, and then thorough washing was performed, between each step, in 0.1 M Tris-HCI buffer, pH 7.5, containing 0.05% Tween-20. The biotinylated peptide bound to the HLA-DP molecules was detected by adding 100 μg/well of the streptavidin-alkaline phosphatase conjugate (45 minutes) diluted 1/2000 in the 10 mM Tris buffer, pH 7, containing 0.15 M NaCl, 0.05% Tween 20, 0.2% BSA and 0.003% thimerosal, and then by adding 200 μl/well of 100 μM MUP substrate in 0.05 M NaHCO₃buffer, pH 9.8, containing 1 mM MgCl₂. The emission of fluorescence by the product of the enzymatic reaction was measured at 450 nm after excitation at 365 nm. The maximum binding was determined by incubating the biotinylated peptide with the MHC II molecule in the absence of competitor peptide. The binding specificity was controlled by adding an excess of nonbiotinylated peptide. The background noise obtained does not differ significantly from that obtained by incubating the biotinylated peptide in the absence of the MHC II molecules.
The results are expressed in the form of the concentration of competitor peptide which inhibits 50% of the maximum binding of the labeled peptide (IC₅₀).

b) Results

The optimum conditions are given in table VI:

TABLE VI

Conditions for the DP4 molecule-binding assay

		DPA1*0103/	DPA1*0103/
	Molecules	DPB1*0401	DPB1*0402

	Biotinylated	bUNK 1-17	bUNK 1-17
	peptide
	Concentration
	10 nM	10 nM
	pH	5	5
	Incubation	24 h	24 h
	time
	Concentration		Dilution
	of HLA-DP4		1/20 to 1/4

he dilutions indicated are carried out using the preparation of purified HLA-DP4 molecules obtained in example 1.

EXAMPLE 3

Sensitivity and Specificity of the DP4 Molecule-Binding Assay

a) Specificity

The results illustrated in FIG. 1 show that the binding activity measured in the assay is specific for HLA-DP4 insofar as:

- peptides known to be ligands for DP4 molecules effectively bind the HLA-DP*0401 and 0402 molecules. They are the IL3 127-146 peptide (naturally present on a DP4 molecule), DP4-restricted CD4+ T lymphocyte-specific peptides: NS-p2 (or NSP2), TT 947-963 and, to a lesser extent, the MAG 245-258 peptide, and
- peptides known to bind other HLA II molecules do not bind to the HLA-DP*0401 and 0402 molecules. They are the peptides: DQB 43-57 (DVEVYRAVTPLGPPD, SEQ ID NO:23), HCI 46-63 (EPRAPWIEQEGPEYWDQE, SEQ ID NO:24) and HA 306-318 (SEQ ID NO:21) which are known to bind, respectively, to HLA-DQ3, HLA-DQ2 and HLA-DR molecules (MARSHALL et al., mentioned above, JOHANSEN et al., Immunogenetics, 1996, 45, 142).

The specificity of the assay results from the use:

- of an HLA-DP molecule-specific antibody for the purification and for the adsorption of the ELISA plates, and
- of the biotinylated peptide which binds with high affinity; FIG. 1 shows that the UNK peptide, which is the nonbiotinylated counterpart of the bUNK tracer peptide, completely inhibits the binding of the tracer both to the HLA-DPB1*0401 molecule and to the DPB1*0402 molecule.

b) Sensitivity

The sensitivity of the assay is reflected by the IC₅₀observed with the nonbiotinylated peptide (UNK1) corresponding to the tracer (bUNK1). FIG. 1 indicates that the values of, respectively, 8 and 9 nM for DPB1*0401 and DPB1*0402 reflect good sensitivity.
FIG. 1 also shows that the DPB1*401 molecule-binding and DPB1*402 molecule-binding activities of the peptides, although comparable, are different. These results confirm that the DPB1*401 and DPB1*402 molecules exhibit differences which can be detected by this binding assay.

EXAMPLE 4

Screening for HLA-DP4-Ligand Peptides Using a Peptide Library

1) Materials and Methods

The overlapping peptides carrying the complete sequence of the major allergen of bee venom (Api ml), described in patent FR 99 00879, were synthesized and subjected to the DP401 molecule-binding and DP402 molecule-binding assays, under the conditions defined in table VI.

2) Results

The results are given in table VII below:

TABLE VII

Binding of the Api m1 peptides to the HLA-
DPB10401 and DPB10402 molecules

IC₅₀(nM)

Peptide	DP0401	DP0402

1-18	>10 000	>10 000
5-22	>10 000	>10 000
9-26	>10 000	>10 000
13-30	>10 000	>10 000
17-34	>10 000	>10 000
21-38	>10 000	>10 000
25-42	>10 000	>10 000
29-46	>10 000	>10 000
33-50	>10 000	>10 000
37-54	>10 000	>10 000
41-58	>10 000	>10 000
45-62	>10 000	>10 000
49-66	>10 000	>10 000
53-70	>10 000	>10 000
57-74	>10 000	>10 000
61-78	>10 000	>10 000
65-82	50 000	6500
69-86	>10 000	>10 000
73-90	>10 000	>10 000
77-94	450	175
81-98	2250	1225
85-102	>10 000	>10 000
89-106	>10 000	>10 000
93-110	>10 000	>10 000
97-114	>10 000	>10 000
101-118	>10 000	>10 000
105-122	>10 000	>10 000
109-126	>10 000	>10 000
113-130	>10 000	>10 000
117-134	>10 000	>10 000

These results show that peptide 77-94 of the major antigen of bee venom 94 (Api ml 77-94: TISSYFVGKMYFNLIDTK, SEQ ID NO:17) is a peptide which is a ligand for the DPB1*0401 and DPB1*0402 molecules.

EXAMPLE 5

Determination of the DP4 Molecule-Binding Units

1) Materials and Methods

a) Determination of a Minimum Peptide Derived from UNK1 Capable of Binding to the DP4 Molecules
Peptides derived from the UNK1 peptide (UNK 1-17, SEQ ID NO:14) comprising increasing deletions at one of the NH₂or COOH ends, or at both ends, were synthesized. The sequence of these peptides (UNK 1-11, 1-12, 1-13, 2-14, 3-15, 4-16, 5-17, 6-17, 7-17, 3-17 and 1-15) corresponding, respectively, to SEQ ID NOS:25-26, 16 and 27-34 is given in table VIII. The binding activity of the peptides, expressed by the IC₅₀value, was determined using the competition binding assay, under the conditions defined in table VI.
b) Determination of the Residues of the UNK1 Peptide which are Involved in Binding to the DP4 Molecules
Mutants of the UNK 3-15 peptide (KYFAATQFEPLAA; SEQ ID NO:28) were synthesized; each mutant contains only one of the residues Y4, F5, T8, Q9, F10, E11, P12 and L13 substituted with alanine or with lysine, the residue K3 substituted with alanine, or else one of the residues A6, A7, A14 and A15 substituted with lysine.
The sequence of these peptides (UNK K3A, Y4A, FSA, TBA, Q9A, F10A, E11A, P12A, L13A, Y4K, FSK, A6K, A7K, T8K, Q9K, F10K, E11K, P12K, L13K, A14K and A15K) corresponding, respectively, to the sequences SEQ ID NOS:35 to 55 is given in table IX.
The binding activity of the peptides, expressed by the IC₅₀value, was determined using the competition binding assay, under the conditions defined in table VI. The loss of binding of the mutant peptides is expressed by the ratio of the IC₅₀values for the mutant peptide and for the UNK1 peptide.

c) Determination of the DP401 Molecule-Binding and DP402 Molecule-Binding Units

The residues F in P1, T in P4, F in P6 or L in P9 of the UNK1 peptide were substituted with one of the following residues:

- P1: Y, L, E, N, T, D, G, H, I, M, P, Q, R, S, V or W,
- P4: F, L, E, D, N, Y, R, S, G, H or P
- P6: Y, L, W, E, N, T, D, G, H, I, M, P, Q, R, S, V, W or C,
- P9: F, Y, E, D, N, R, V, G, H, I, P, Q, S, T or W.

The sequence of these peptides, corresponding to the sequences SEQ ID NOS: 56 to 81 and SEQ ID NOS:96 to 129, is given, respectively, in tables Xa and Xb.
The loss of DP4 molecule-binding of the mutant peptides was determined using the assay described for the alanine and lysine mutants.

2) Results

a) Determination of a Minimum UNK1 Peptide

The results are given in table VIII below:

TABLE VIII

Binding to DP401 and DP402 of the peptides derived from UNK1 (UNK 1-17)

Position of the sequences

IC₅₀ (nM)

Peptides	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	401	402

UNK1-17

E

K

Y

F

A

T

Q

F

E

P

L

A

R

L

11

14

UNK1-11

E

K

Y

F

A

T

Q

F

E

55

95

UNK1-12

E

K

Y

F

A

T

Q

F

E

P

93

100

UNK1-13

E

K

Y

F

A

T

Q

F

E

P

L

14

28

UNK2-14

K

Y

F

A

T

Q

F

E

P

L

A

15

23

UNK3-15

K

Y

F

A

T

Q

F

E

P

L

A

19

23

UNK4-16

Y

F

A

T

Q

F

E

P

L

A

R

38

25

UNK5-17

F

A

T

Q

F

E

P

L

A

R

L

65

38

UNK6-17

A

T

Q

F

E

P

L

A

R

L

15000

6000

UNK7-17

A

T

Q

F

E

P

L

A

R

L

15000

12500

UNK3-17

K

Y

F

A

T

Q

F

E

P

L

A

R

L

18

14

UNK1-15

E

K

Y

F

A

T

Q

F

E

P

L

A

18

17

The results obtained show that:

- the loss of binding observed with the UNK 1-11, UNK 1-12, UNK 4-16 and UNK 5-17 peptides suggests that the minimum peptide is 13 to 15 amino acids in size,
- the loss of binding observed with the UNK 6-17 and UNK 7-17 peptides suggests that the F residue at position 5 is the first residue for anchoring of the peptides in the binding site of the DP4 molecules (residue P1).
  b) Determination of the Residues of the UNK1 Peptide which are Involved in Binding to the DP4 Molecules

The results are given in table IX below:

TABLE IX

Loss of binding to DP401 and DP402 of the alanine and lysine mutants

PEPTIDE	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	DP*0401	DP*0402

UNK1

E

K

Y

F

A

T

Q

F

E

P

L

A

R

L

1

1.00

UNK 3-15

K

Y

F

A

T

Q

F

E

P

L

A

1.5

1.7

P1

P4

P6

P9

UNK K3A

A

Y

F

A

T

Q

F

E

P

L

A

0.9

1

UNK Y4A

K

A

F

A

T

Q

F

E

P

L

A

1.2

0.9

UNK F5A

K

Y

A

T

Q

F

E

P

L

A

41

5.7

UNK T8A

K

Y

F

A

Q

F

E

P

L

A

0.8

UNK Q9A

K

Y

F

A

T

A

F

E

P

L

A

0.8

0.6

UNK F10A

K

Y

F

A

T

Q

A

E

P

L

A

336

62

UNK E11A

K

Y

F

A

T

Q

F

A

P

L

A

0.5

UNK P12A

K

Y

F

A

T

Q

F

E

A

L

A

1.2

0.9

UNK L13A

K

Y

F

A

T

Q

F

E

P

A

3.1

2.3

UNK Y4K

K

F

A

T

Q

F

E

P

L

A

1.2

1

UNK F5K

K

Y

K

A

T

Q

F

E

P

L

A

569

297

UNK A6K

K

Y

F

K

A

T

Q

F

E

P

L

A

1.6

1

UNK A7K

K

Y

F

A

K

T

Q

F

E

P

L

A

2

1.2

UNK T8K

K

Y

F

A

K

Q

F

E

P

L

A

31

3.6

UNK Q9K

K

Y

F

A

T

K

F

E

P

L

A

1.6

0.9

UNK F10K

K

Y

F

A

T

Q

K

E

P

L

A

420

350

UNK E11K

K

Y

F

A

T

Q

F

K

P

L

A

1.3

0.7

UNK P12K

K

Y

F

A

T

Q

F

E

K

L

A

1.8

1.4

UNK L13K

K

Y

F

A

T

Q

F

E

P

K

A

81

70

UNK A13K

K

Y

F

A

T

Q

F

E

P

L

K

A

1.7

1.4

UNK A15K

K

Y

F

A

T

Q

F

E

P

L

A

K

1.7

1.4

The results show a loss of binding for F5A, F10A, F5K, F10K and L13K, which strongly suggests that the residues for anchoring of the peptides in the binding site of the DP4 molecules are, respectively, F5 at position P1, F10 at position P6 and L13 at position P9.

c) Determination of the DP401 Molecule-Binding and DP402 Molecule-Binding Units

The results are given in tables Xa and Xb and table XI below:

TABLE Xa

Loss of binding to DP401 and DP40 of the mutants at P1, P4,
P6 and P9 (SEQ ID NOS: 56 to 81)

PEPTIDE	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	DP*0401	DP*0402

UNK1

E

K

Y

F

A

T

Q

F

E

P

L

A

R

L

1.00

UNK 3-15

K

Y

F

A

T

Q

F

E

P

L

A

1.4

P1

P4

P6

P9

UNK F5Y

K

Y

A

T

Q

F

E

P

L

A

3

3.9

UNK F5L

K

Y

L

A

T

Q

F

E

P

L

A

3.6

6

UNK F5E

K

Y

E

A

T

Q

F

E

P

L

A

117

123

UNK F5N

K

Y

N

A

T

Q

F

E

P

L

A

398

76

UNK F5T

K

Y

T

A

T

Q

F

E

P

L

A

234

126

UNK T8F

K

Y

F

A

F

Q

F

E

P

L

A

2.7

1.7

UNK T8L

K

Y

F

A

L

Q

F

E

P

L

A

1.9

2.6

UNK T8E

K

Y

F

A

E

Q

F

E

P

L

A

2.7

3.7

UNK T8D

K

Y

F

A

D

Q

F

E

P

L

A

2

2.3

UNK T8N

K

Y

F

A

N

Q

F

E

P

L

A

4.1

4.8

UNK T8Y

K

Y

F

A

Y

Q

F

E

P

L

A

4.1

2.4

UNK T8R

K

Y

F

A

R

Q

F

E

P

L

A

17

6.7

UNK T8S

K

Y

F

A

S

Q

F

E

P

L

A

2.7

3.4

UNK F10Y

K

Y

F

A

T

Q

Y

E

P

L

A

2.4

3.4

UNK F10L

K

Y

F

A

T

Q

L

E

P

L

A

7.2

6.7

UNK F10W

K

Y

F

A

T

Q

W

E

P

L

A

2.2

2.5

UNK F10E

K

Y

F

A

T

Q

E

P

L

A

1405

1190

UNK F10N

K

Y

F

A

T

Q

N

E

P

L

A

2460

1809

UNK F10T

K

Y

F

A

T

Q

T

E

P

L

A

1171

476

UNK L13F

K

Y

F

A

T

Q

F

E

P

F

A

1.7

4

UNK L13Y

K

Y

F

A

T

Q

F

E

P

Y

A

2

15

UNK L13E

K

Y

F

A

T

Q

F

E

P

E

A

11

8

UNK L13D

K

Y

F

A

T

Q

F

E

P

D

A

9.4

9

UNK L13N

K

Y

F

A

T

Q

F

E

P

N

A

17

10

UNK L13R

K

Y

F

A

T

Q

F

E

P

R

A

64

43

UNK L13V

K

Y

F

A

T

Q

F

E

P

V

A

2.5

1.9

he values greater than 10 are indicated in bold

TABLE Xb

Loss of binding to DP401 and DP40 of the mutants at P1,
P4, P6 and P9 (SEQ ID NOS: 96 to 129)

Peptides	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17		Loss
UNK	E	K	K	Y	F	A	A	T	Q	F	E	P	L	A	A	R	L	DP*0401	DP*0402

UNK F5D

K

Y

D

A

T

Q

F

E

P

L

A

150

57

UNK F5G

K

Y

G

A

T

Q

F

E

P

L

A

160

55

UNK F5H

K

Y

H

A

T

Q

F

E

P

L

A

150

36

UNK F5I

K

Y

I

A

T

Q

F

E

P

L

A

3

2

UNK F5M

K

Y

M

A

T

Q

F

E

P

L

A

3

UNK F5P

K

Y

P

A

T

Q

F

E

P

L

A

100

48

UNK F5Q

K

Y

Q

A

T

Q

F

E

P

L

A

260

97

UNK F5R

K

Y

R

A

T

Q

F

E

P

L

A

140

54

UNK F5S

K

Y

S

A

T

Q

F

E

P

L

A

160

42

UNK F5V

K

Y

V

A

T

Q

F

E

P

L

A

11

6

UNK F5W

K

Y

W

A

T

Q

F

E

P

L

A

0.5

1

UNK T8G

K

Y

F

A

G

Q

F

E

P

L

A

1

2

UNK T8P

K

Y

F

A

P

Q

F

E

P

L

A

3

2

UNK T8H

K

Y

F

A

H

Q

F

E

P

L

A

6

2

UNK F10D

K

Y

F

A

T

Q

D

E

P

L

A

2700

120

UNK F10G

K

Y

F

A

T

Q

G

E

P

L

A

>7100

3700

UNK F10H

K

Y

F

A

T

Q

H

E

P

L

A

2400

220

UNK F10I

K

Y

F

A

T

Q

I

E

P

L

A

7

6

UNK F10M

K

Y

F

A

T

Q

M

E

P

L

A

4

1

UNK F10P

K

Y

F

A

T

Q

P

E

P

L

A

210

12

UNK F10Q

K

Y

F

A

T

Q

E

P

L

A

>7100

230

UNK F10R

K

Y

F

A

T

Q

R

E

P

L

A

>7100

370

UNK F10S

K

Y

F

A

T

Q

S

E

P

L

A

500

33

UNK F10V

K

Y

F

A

T

Q

V

E

P

L

A

42

22

UNK F10W

K

Y

F

A

T

Q

W

E

P

L

A

1

UNK F10C

K

Y

F

A

T

Q

C

E

P

L

A

17

2

UNK L13G

K

Y

F

A

T

Q

F

E

P

G

A

15

12

UNK L13H

K

Y

F

A

T

Q

F

E

P

H

A

13

35

UNK L13I

K

Y

F

A

T

Q

F

E

P

I

A

1

3

UNK L13P

K

Y

F

A

T

Q

F

E

P

A

9

3

UNK L13Q

K

Y

F

A

T

Q

F

E

P

Q

A

8

2

UNK L13S

K

Y

F

A

T

Q

F

E

P

S

A

9

5

UNK L13T

K

Y

F

A

T

Q

F

E

P

T

A

5

3

UNK L13W

K

Y

F

A

T

Q

F

E

P

W

A

1

13

he values greater than 10 are indicated in bold

TABLE XI

Specificity of the DP4 molecule bindin

P1

P4

P6

P9

	+	−	+	−	+	−	+	−

DP*	F, Y,	H, T,	F, L,	K, R	F, Y,	H, V,	F, Y ,	H, G,
0401	L, I,	S, A,	Y, H,		L, W,	A, G,	V, A,	K, R,
	M, W	G, P,	T, A,		I, M	T, P,	I, P,	E, N, C
		K, R,	S, G,			S, K,	W , L,
		D, E,	P, I,			R, E,	M, Q,
		N, Q,	M, Q,			D, N,	S, T,
		C, V	E, D,			Q, C	D
			N, V,
			W, C
DP*	F, Y,	H, T,	F, L,		F, Y,	H, V,	F, V,	Y, H,
0402	L, A ,	S, G,	Y, H,		L, W,	A, G,	A, I,	W, G,
	I, M,	P, K,	T, A,		I, M, C	T, P,	P, L,	K, R, N
	V , W	R, D,	S, G,			S, K,	M, C ,
		E, N,	P, I,			R, E,	Q, S,
		Q, C	MQ,			D, N,	T, D,
			E, D,			Q	E
			N, V,
			K, W,
			C, R

he amino acids which caused, respectively, a loss of binding by a factor greater than 10 and less than 10 are indicated in the column (−) and the column (+); the anchoring residues are represented in bold and the differences between DPB10401 and DPB10402 are underlined.

The results obtained show that:

- pockets P1, P6 and P9 of the binding site of the DP4 molecules accept aromatic or hydrophobic residues,
- pocket P4 of the binding site of the DP4 molecules is very permissive, as is, to a lesser extent, pocket P9, which also accepts Q, S, T and D residues, and
- the DPB1*0401 allele accepts a tyrosine or tryptophan residue at P9 but does not accept cysteine or glutamic acid residues at this position, nor alanine or valine residues at P1, nor lysine or arginine residues at P4, nor a cysteine residue at P6; the DPB1*0402 allele accepts an alanine or valine residue at P1, a lysine or arginine residue at P4, a cysteine residue at P6 and a cysteine or glutamic acid residue at P9, but does not accept a tyrosine or tryptophan residue at this position.

d) Identification of a Binding Unit in the Sequence of DP4-Ligand Peptides

The binding activity of various DP4-ligand peptides was measured using the competition binding assay, under the conditions defined in table VI. The results are expressed by the IC₅₀values (table XIII) or by the value of the ratio of the IC₅₀values for the ligand peptide and for the UNK1 peptide (table XII).
The peptides tested, corresponding respectively to the following sequence SEQ ID NOS, are given in tables XII and XIII:

- TT 947-963 (SEQ ID NO:20)
- UNK1 9 (SEQ ID NO:14)
- NY-ESO1 87-11, 119-143, 158-180, 166-180, (SEQ ID NO:94, 83, 82, 95),
- UL21 283-302 (SEQ ID NO:18)
- IL3 127-146 (SEQ ID NO:13)
- NS-P2 (SEQ ID NO:4)
- Api-ml 65-72, 69-86, 73-90, 77-94, 81-98 (SEQ ID NO: 85, 88, 89, 17, 86)
- MAG 245-258 (SEQ ID NO:19)
- MART1 1-20, 41-60, 51-73, 62-72, 103-118 (SEQ ID NO:90, 91, 87, 92, 93).

In parallel, the sequences of these DP4-ligand peptides were aligned and the presence of a DP4-binding unit as defined in table XI was sought (tables XII and XIII).

TABLE XII

Alignment of the sequences of the
DP4-ligand peptides

Peptides	Sequences	040	040

	P1 P6 P9
TT947-963	FNN F TVSF W LRVPKVSA	0.57	0.75

UNK1	EKKY F AATQ F EPLAARL	1	1

NY-ESO1	LLM W ITQC F LPVFLAQPPSGQRR	2	7
158-180

UL21 283-302	RE L WWVF Y AG D RALEEPHAE	2.86	2.78

IL3 127-146	GPGAPAD V QYDL Y LN V ANRR	3.98	3.06

NS-p2	GVQIVRQIRSGERF L KI W SQ	4.69	3.89

NY-ESO1	PGVLLKEFTVSGN I LTIR L TA A DHR	12	12
119-143

NY-ESO1	PGVL L KEFT V SGNILTIRLTAADHR	12	12
119-143

NY-ESO1	PG V LLKE F TV S GNILTIRLTAADHR	12	12
119-143

NY-ES01	LLEF Y LAMP F ATPMEAELARRSLAQ	20	9
87-111

Api m1 77-94	TISSY F VGKM Y FN L IDTK	30	10

Api m1 77-94	TISSYF V GKMY F NL I DTK	30	10

Api m1 65-72	DKFYDCLKNSADTSSYF	NT	371.4

MAG 245-258	L L TQH F VQEN Y LE Y	30.2	38.89

MAG 245-258	L L TQHF V QENYLEY	30.2	38.89

Api m1 81-98	YFVGKMYFNLIDTKCYKL	150	70

MART1 51-73	RNGYRA L MDKS L HVGTQCALTRR	857	414

401: relative DPB1*0401-binding activity of the
peptides compared to the UNK1 peptide (binding activity
equal to 1)
402: relative DPB1*0402-binding activity of the
peptides compared to the UNK1 peptide (binding activity
equal to 1)

TABLE XIII

Binding of the Api m1, NY-ESO1 and MART1
peptides to DPB10401 and DPB040

IC50 (nM)

peptides	Sequences	401	402

UNK1	EKKY F AATQ F EP L AARL	12	13

Api m1 65-72	DKFYDCLKNSADTISSYF	50000	6500

Api m1 69-86	DCLKNSADTISSYFVGKM	>10000	>10000

Api m1 73-90	NSADTISSY F VGKM Y FN L	>10000	>10000

Api m1 77-94	TISSY FV GKMYFNLIDTK	450	175

Api m1 81-98	Y FV GKM YF N LI DTKCYKL	2250	1225

MART1 1-20	MPREDAHFIYGYPKKGHGHS	>10000	>10000

MART1 41-60	LLIGCWYCRRRNGYRALMDK	>10000	>10000

MART1 51-7	RRNGYRALMDKSLHVGTQCA	8000	4000
	LTRR

MART1 62-72	LMDKSLHVGTQCALTRRCPQ	>10000	>10000

MART1 103-118	AYEKLSAEQSPPPYSP	>10000	>10000

NY-ESO1	LLEFYLAMPFATPMEAELAR	183	83
87-111	RSLAQ

NY-ESO1	PGVLLKEFTVSGN I LTIR L T	110	117
119-14	A A DHR

NY-ESO1	LLM W ITQC F LP V FLAQPPSGQRR	20	67
158-180

NY-ESO1	FLPVFLAQPPSGQRR	>10000	>10000
166-180

he amino acids compatible with the unit for binding
to the DPB10401 and DPB10402 alleles are indicated in
bold and the complete binding units are underlined.
he sequence of this peptide is described in Zarour et
al., PNAS, 2000, 97, 400-405.
he sequence of this peptide is described in Zarour
et al., Cancer Res., 2000, 60, 4646-4952.

The results show that, for most of the DP4-ligand peptides, a strong correlation exists between the presence of at least two residues P1 and P6 or P6 and P9 as defined in table XI and a high affinity for the DP4 molecules (IC₅₀<1000 nM). These results also show that the most important residues in the binding to HLA-DP4 are the aromatic or hydrophobic residues at position P1 and P6; these same residues at position P9 are less important.

EXAMPLE 6

Prediction of the Sequence of HLA-DP4-Ligand Peptides Based on an Amino Acid Sequence

1) Materials and Methods

An HLA-DP4 molecule-binding matrix was established based on the binding activities (IC₅₀) of the mutants of the UNK 3-15 peptide, measured using the assay for binding to the DP4 molecules encoded by the DPB1*0401 and DPB1*0402 alleles, as defined above, using the UNK 3-15 peptide as tracer peptide (example 5 and tables IX, Xa, Xb and XI). The contribution of each of the amino acids at positions P1, P4, P6 and P9 of said mutant peptides to the binding to the DP*0401 and DP*0402 molecules is evaluated by a binding score corresponding to the logarithm (log) of the ratio of the IC₅₀values for the mutant peptide and for the UNK 3-15 peptide. The score for binding to the DP*0401 and DP*0402 molecules, obtained for each of the amino acids at positions P1, P4, P6 and P9 of said mutant peptides, constitutes the HLA-DP4-binding matrix shown in table XIV below:

TABLE XIV

HLA-DP4-binding matrix

Amino acid

DP4	Position	A	D	E	F	G	H	I	K	L	M

DP*0401	P1	1.6	1.7	1.85	0	2.2	2.17	0.48	2.6	0.48	0.48
	P4	0	0.3	X0.3	0.3	0	0.78	0.3	1.48	0.3	0.3
	P6	2.38	3.32	2.9	0	3.85	3.4	0.85	2.43	0.78	0.6
	P9	0.6	0.85	1.04	0	1.18	1.11	0	1.85	0	0
DP*0402	P1	0.85	1.76	1.85	0	1.76	1.56	0.3	2.28	0.48	0.48
	P4	0	0.3	0.48	0	0	0.3	0.3	0.7	0.3	0.3
	P6	1.88	3.85	2.9	0	3.57	2.33	0.78	2.28	0.7	0
	P9	0.6	0.9	0.9	0.7	1.08	1.54	0.48	1.9	0	0.3

Amino acid

DP4	Position	N	P	Q	R	S	T	V	W	Y	C

DP*0401	P1	2.32	1.98	2.41	2.14	2.2	2.15	1.04	0	0.48	2.32
	P4	0.6	0.48	0.3	1.5	0.3	0	0.15	0.3	0.6	0.6
	P6	3.15	2.32	3.85	3.85	2.7	2.85	1.62	0.3	0.3	1.23
	P9	1.15	0.95	0.9	1.7	0.95	0.6	0.3	0	0.3	1.15
DP*0402	P1	1.78	1.68	1.99	1.72	1.62	1.85	0.78	0	0.48	1.78
	P4	0.6	0.3	0.3	0.7	0.48	0	0.15	0	0	0.6
	P6	3.08	1.08	2.36	2.57	1.52	2.48	1.34	0.3	0.48	0.3
	P9	0.85	0.48	0.3	1.6	0.7	0.6	0.3	1.11	1.18	0.85

Starting with an amino acid sequence, the binding of the peptides of at least 9 amino acids included in said sequence is calculated, based on the matrix above, by adding, for each fragment of 9 amino acids of said peptide, for example overlapping fragments of 9 amino acids covering this entire sequence, the binding scores for the residues at positions 1, 4, 6 and 9 of said fragment.
Peptides having binding scores of, respectively, 0, 1, 2, 3 and 4 correspond to peptides which exhibit a 1-, 10-, 100-, 1000- and 10 000-fold loss of binding relative to the UNK 3-15 peptide (IC ₅₀10 nM), i.e. exhibiting an estimated binding activity of, respectively, 10 nM, 100 nM, 1000 nM, 10 μM and 100 μM.
The peptides having the lowest binding scores, preferably less than 2, preferably less than 1, even more preferably close to 0, are selected; these peptides correspond to those for which the HLA-DP4-binding activity, estimated based on the binding matrix as defined above, is the highest.

2) Results

The correlation between the binding score for the peptides (estimated using the HLA-DP4-binding matrix, table XIV), and their affinity for the HLA-DP4 molecules (residues in the binding assay as defined above) was analyzed for 44 peptides studied in examples 4 and 5.
The results are given in table XV and FIG. 2.

TABLE XV

Binding score and HLA-DP4-binding activity of
various peptides

DP*0401

DP*0402

		Binding		Binding
Peptide	IC₅₀(nM)	score	IC₅₀(nM)	score

TT947-963	4.61	0.90	7	1.08
UNK 1	9.47	0.00	10.00	0.00
NY-ESO1 158-180	30.00	0.60	70.00	0.60
UL21 283-302	22.89	1.77	20.00	2.01
IL3 127-146	30.37	1.94	30.00	1.86
NSP2	40.00	2.44	40.00	2.69
NY-ESO1 119-143	100.00	2.16	100.00	1.90
NY-ESO1 87-11	186.12	1.73	87.00	1.26
MAG 245-258	250.00	0.90	400.00	1.68
MART1 51-73	8000.00	3.76	4000.00	2.45
Api-m1 1-18	10000.00	1.89	10000.00	2.08
Api-m1 5-22	10000.00	4.88	10000.00	3.33
Api-m1 9-26	10000.00	3.58	10000.00	3.33
Api-m1 13-30	10000.00	2.92	10000.00	2.38
Api-m1 17-34	10000.00	2.92	10000.00	2.38
Api-m1 21-38	10000.00	4.08	10000.00	3.40
Api-m1 25-42	10000.00	3.97	10000.00	2.64
Api-m1 29-46	10000.00	2.56	10000.00	1.86
Api-m1 33-50	10000.00	2.56	10000.00	1.86
Api-m1 37-54	10000.00	4.00	10000.00	3.07
Api-m1 41-58	10000.00	4.00	10000.00	3.07
Api-m1 45-62	10000.00	3.23	10000.00	2.45
Api-m1 49-66	10000.00	3.23	10000.00	2.45
Api-m1 53-70	10000.00	3.15	10000.00	2.45
Api-m1 57-74	10000.00	3.15	10000.00	2.90
Api-m1 61-78	10000.00	3.15	10000.00	2.90
Api-m1 65-82	50000.00	3.64	6500.00	2.90
Api-m1 69-86	10000.00	2.70	10000.00	1.94
Api-m1 73-90	10000.00	1.78	10000.00	1.18
Api-m1 77-94	450	1.34	175.00	1.18
Api-m1 81-98	2250	1.34	1225.00	0.78
Api-m1 85-102	10000.00	1.71	10000.00	0.78
Api-m1 89-106	10000.00	1.71	10000.00	0.78
Api-m1 93-110	10000.00	4.27	10000.00	3.02
Api-m1 97-114	10000.00	4.98	10000.00	3.46
Api-m1 101-118	10000.00	3.67	10000.00	3.21
Api-m1 105-122	10000.00	3.65	10000.00	3.21
Api-m1 109-126	10000.00	3.65	10000.00	3.44
Api-m1 113-130	10000.00	3.96	10000.00	3.50
Api-m1 117-134	10000.00	1.86	10000.00	1.98
MART1 1-20	10000.00	3.16	10000.00	2.16
MART1 41-60	10000.00	2.48	10000.00	2.20
MART1 62-72	10000.00	4.00	10000.00	2.46
MART1 103-118	10000.00	4.11	10000.00	2.95

These results show that a strong correlation exists between the binding score and the affinity of the peptides for HLA-DP4. Specifically, taking as activity threshold a binding score <2 and a binding activity IC₅₀<1000 nM:

- 80% of the peptides having a binding score <2 for the HLA-DP4 molecules (DP*0401 and DP*0402) exhibit a high affinity for these molecules (IC₅₀<1000 nM, true positive peptides), and
- 82% and 76% of the peptides having a binding score ≧2 for the HLA-DP4 molecules (respectively DP*0401 and DP*0402) exhibit a low affinity for these molecules (IC₅₀≦1000 nM, true negative peptides).

Consequently, the HLA-DP4-binding matrix can be used to predict the sequence of HLA-DP4-ligand peptides from any amino acid sequence, in particular a sequence representing an antigen of interest.

Claims

1-26. (canceled)

27. A method of identifying an HLA-DP4-ligand peptide or a modified HLA-DP40-ligand peptide, which comprises at least one of the following steps:

(a) establishing an HLA-DP4-binding matrix by calculating for all the mutants of a tracer peptide, representing all the substitutions of the residues at positions 1, 4, 6 or 9 of said tracer peptide with the other 19 natural amino acids, the ratio of the IC₅₀values for said mutant peptides and for said tracer peptide, wherein said ratios correspond to the concentration of these molecules which inhibit 50% of the binding of the tracer peptide;

(b) evaluating the HLA-DP4-binding of peptides of at least 9 amino acids included in said amino acid sequence, by calculating, for each fragment of 9 amino acids of said peptide, the sum of the scores for HLA-DP4-binding of the residues at positions 1, 4, 6 and 9 of said fragment, using the binding matrix established in (a), and/or

(c) identifying the HLA-DP4-ligand peptides corresponding to those to which the loss of HLA-DP4-binding, relative to said tracer peptide, is smallest, i.e. those for which the sum of the binding scores has the lowest value expressed as a logarithm (log);

wherein said tracer peptide is at least one peptide selected from the group of consisting of the peptides exhibiting a signal/background noise ratio greater than 5, at a concentration of 10 nM, in a direct HLA-DP4-binding assay, and corresponding to a formula (I) Z₁X₁X₂X₃X₄X₅X₆X₇X₈X₉Z₂; wherein

Z₁and Z₂, which may be identical or different, are zero or each represent a peptide of 1 to 100 natural or synthetic amino acids;

X₆represents an aromatic or hydrophobic amino acid, or a cysteine;

X₁represents an aromatic or hydrophobic amino acid and/or X₉represents an aromatic or hydrophobic amino acid, or a cysteine (C), an aspartic acid (D), a glutamine (Q), a serine (S), a threonine (T) or a glutamic acid (E); and

X₂, X₃, X₄, X₅, X₇and X₈each represent a natural or synthetic amino acid.

28. The method of claim 27, which comprises steps (a), (b), and (c).

29. The method of claim 27, comprising selecting an HLA-DP4 peptide or modified peptide that exhibits an HLA-DP4-binding activity of less than 1000 nM.

30. The method of claim 27, which identifies an HLA-DP4 ligand peptide.

31. The method of claim 27, which identities a modified HLA-DP4 ligand peptide.

32. The method of claim 27, wherein the HLA-DP4 peptide or modified peptide sequence corresponds to formula (I) Z₁X₁X₂X₃X₄X₅X₆X₇X₈X₉Z₂, wherein

X₆represents an aromatic or hydrophobic amino acid, or a cysteine;

X₁represents an aromatic or hydrophobic amino acid; and/or

X₉represents an aromatic or hydrophobic amino acid, or else a cysteine (C), an aspartic acid (D), a glutamine (Q), a serine (S), a threonine (T) or a glutamic acid (E); and

X₂, X₃, X₄, X₅, X₇and X₈each represent a natural or synthetic amino acid,

on condition that said HLA-DP4-ligand peptide of general formula (I) does not correspond to any of the sequences SEQ ID NOS: 1 to 17.

33. The method of claim 32, wherein

X₆is selected from the group consisting of L, I, W, F, M, Y and C, and

X₁is selected from the group consisting of A, V, L, I, W, F, M and Y; and/or

X₉is selected from the group consisting of A, V, L, I, P, W, F, M, Y, C, D, Q, S, T and E.

34. The method of claim 33, wherein the peptide or modified peptide hinds specifically to the HLA-DPB1*0401 molecule and

X₆is different from C, and/or

X₁is different from A and from V, and/or

X₉represents W or Y or

X₉is different from E and from C, and/or

X₄is different from K and from R.

35. The method of claim 33, wherein the peptide or modified peptide hinds specifically to the DPB1*0402 molecule and X₆represents C, and/or X₁represents A or V, and/or X₉represents E or C or X₉is different from Y and W, and/or X₄represents K or R.