This invention relates to molecules which are suitable for inducing tolerance to Human Platelet Antigen la (HPA-la). Such molecules may be used in the treatment or prevention of neonatal alloimmune thrombocytopenia (NAIT) by immunisation of HPA-la negative women. The peptide-containing molecules of one aspect of the invention comprise amino acids 25 to 33 of integrin β3 in which residue 33 is modified to be isoleucine. Other amino acid modifications may also be present.

NAIT is a serious complication in foetal and neonatal development; it is also known as fetal and neonatal alloimmune thrombocytopenia (FNAIT). The condition is most commonly caused by maternal antibodies against foetal HPA-la which can be transferred over the placenta during pregnancy. The HPA-la platelet alloantigen is defined by a single amino acid difference in residue 33 of the 3-integrin (leucine versus proline). About 2% of Caucasians are homozygous for HP A- lb (P33).

Women with this phenotype may become immunized in connection with pregnancy, when the foetus has a paternally- inherited HPA-la allotype. The vast majority of HP A- la-immunized mothers carry the MHC allele HLA-DRB3*01:01 and the HPA- la peptide has been shown to bind to this MHC molecule. The allogeneic L33 residue has been assigned a role for anchoring the peptide, docking into the P9 pocket of DRA/DRB3*01 :01, by molecular modelling and crystallographic studies (Parry et al., J. Mol. Biol. (2007) 371 :435-446). Other residues close to L33 are predicted to contribute to MHC binding in pockets PI (W25), P4 (D28) and P6 (A30).

The present inventors have carried out a study to further characterize the T cell response in HPA-la alloimmunity associated with NAIT, in order to help direct future therapeutic strategies aimed at reducing haemorrhagic complications in thrombocytopenia. In the course of these studies they have discovered that certain modified HPA-la peptides have improved properties, including MHC binding and T cell activation. These peptides form the basis of the present invention, which may be used in the prophylaxis or treatment of NAIT, especially by immune tolerisation. Pharmaceutical compositions comprising HPA-la peptides have previously been suggested for use in the prevention or management of conditions caused by exposure to antithetical alleles of a platelet by tolerisation, e.g. by exposure to HPA- la. For example, US patent application No. 2007/042949 teaches the use of HPA-la (or peptides derived therefrom) in the prevention or management of NAIT. Specific 15-mer peptide fragments which span positions 19 to 39 of native integrin β3 and as such comprise the L33 residue are taught for such uses. However, there is no teaching to use modified versions of these peptides. Surprisingly, the present inventors have found that a mutation at position L33 results in a peptide which can activate HPA la-specific T cells more strongly than the corresponding human HPA- la peptide. Such peptides should be more suitable than native human HPA-la at tolerising antigen-specific T cell responses and can therefore be used in improved methods of immunisation of HPA-la negative women, e.g. using lower dosages, fewer doses, etc.

Thus, viewed from a first aspect the invention provides a molecule comprising (e.g. consisting of) a peptide of formula I (SEQ ID NO. 34):

Z ₁ -X ₁ X ₂ X ₃ DX ₄ X ₅ X ₆ PI-Z ₂ (I) wherein Z _\ and Z ₂ denote optional peptide extensions at the N- and C- termini, respectively. The iso leucine residue immediately to the N-terminal side of Z ₂ corresponds to position 33 of the native integrin β3.

When present, Z _\ denotes a hydrophobic amino acid residue or an analogue thereof, especially alanine. When present, Z ₂ preferably denotes -X ₇ X ₈ , in which X ₈ may be present or absent (i.e. Z ₂ is a di- or mono-peptide extension), wherein X ₇ denotes a hydrophobic amino acid, e.g. G, A, V, I, L, W, M, Y or F, preferably G, A, M, F or W, especially G, A or W, or an analogue thereof, especially G; and wherein X ₈ (when present) denotes a polar amino acid, e.g. S, T, C or Y, preferably S or Y, or an analogue thereof, especially S. In a preferred embodiment, Z ₂ denotes a di-peptide selected from GS, AS and WS, especially preferably GS or AS, e.g. GS. In one embodiment Z _\ is present and Z ₂ is absent; in another embodiment Z _\ is absent and Z ₂ is present; in a further embodiment Z _\ and Z ₂ are both absent; and in a yet further embodiment, Z _\ and Z ₂ are both present.

Xi denotes a hydrophobic amino acid, especially having an aromatic side- chain, e.g. selected from W and F, and analogues thereof. In a preferred

embodiment Xi denotes W.

X ₂ denotes a polar amino acid, e.g. selected from C, S and T, and analogues thereof. In a preferred embodiment, X ₂ denotes C or S, especially C.

X ₃ denotes a polar amino acid, e.g. selected from S, C and T, and analogues thereof. In a preferred embodiment, X ₃ denotes S or C, especially S.

X ₄ denotes an ionic amino acid having a negatively charged side chain, e.g. selected from E and D, and analogues thereof. In a preferred embodiment, X ₄ denotes E or D, especially E.

X ₅ denotes a polar or hydrophobic amino acid, e.g. selected from A and S, and analogues thereof. In a preferred embodiment, X ₅ denotes A or S, especially A.

X ₆ denotes a hydrophobic amino acid, e.g. selected from L, I and G, and analogues thereof. In one embodiment, X6 denotes L or I, preferably L.

The term "peptide" as used herein defines a chain of two or more amino acid and/or amino acid analogue residues which possesses the functional characteristics of an oligomer or polymer of amino acids, e.g. conformational flexibility. The amino acids or analogues thereof comprise side-chains (except in the case of glycine or glycine analogues) which are moieties that do not form part of the backbone of the chain and which can participate both in intra- and inter-molecular interactions such as disulphide bridges, ionic interactions (e.g. metal chelating), hydrogen bonding, aromatic interactions (e.g. pi-pi stacking), van der Waals interactions, etc.

By "amino acid" and "amino acid residue" is meant a moiety comprising the backbone structure -NH-CHR-C(O)- wherein R denotes one of the 20 standard (or "natural") amino acid side-chains. The conventional single-letter code nomenclature for standard amino acids is used herein. All of the natural amino acids, except for glycine, are L- isomers. For the purpose of the present invention a moiety having a backbone structure -NH-CHR-C(O)- wherein R is other than one of the 20 standard amino acid side-chains is considered to be an amino acid analogue. The chemical composition of the side chain is a major determinant in the nature of the amino acid or analogue thereof, e.g. in whether the residue can be classified as polar, charged or hydrophobic. A "polar" residue is typically a moiety having a side-chain which is partially charged under physiological conditions or which is capable of participating in hydrogen bonding, i.e. as a donor or an acceptor. Examples of polar amino acids include serine (S), threonine (T), cysteine (C), methionine (M), tyrosine (Y), tryptophan (W), asparagine (N) or glutamine (Q). A "charged" residue is typically a moiety having a side-chain which is charged or substantially charged under physiological conditions, e.g. aspartic acid (D), glutamic acid (E), histidine (H), lysine (K) and arginine (R). A "hydrophobic" residue is typically a moiety having a side-chain which is not substantially charged under physiological conditions, e.g. which cannot substantially interact with water molecules by hydrogen bonding or charge-charge interactions. Examples of hydrophobic amino acids include alanine (A), valine (V), leucine (L), isoleucine (I), methionine (M), proline (P), phenylalanine (F), tyrosine (Y) and tryptophan (W). For the purposes of the present invention, glycine (G) and its analogues are considered to be hydrophobic amino acids.

Amino acid analogues may be used in place of the amino acids mentioned herein the peptide part of the molecules of the invention. Such a peptide (sometimes called a peptidomimetic) will typically have similar polarity, three dimensional size and functionality to the corresponding peptide having all-standard amino acid residues but will possess certain improvements, e.g. increased resistance to hydrolysis, a more stable secondary structure and/or an increased biological activity. Examples of amino acid analogues include stereoisomers (e.g. D- isomers) of standard amino acids and beta-amino acids (e.g. β-alanine, β-phenylalanine, 3- aminoadipic acid, etc.).

Other examples of amino acid analogues include moieties which are modified chemically in the backbone -NH-CHR-C(O)- (see e.g. Vagner et al., Curr Opin Chem Biol. (2008) 12(3):292-296; and Chapter 14 of "Drug Design and Development", Krogsgaard, Larsen, Liljefors and Madsen (Eds) 1996, Horwood Acad. Pub, the contents of each of which are incorporated herein in their entirety). Backbone modifications include replacement of the -NH- with -N(OH)-, -O- or -S-, or alkylation of the nitrogen atom (see e.g. Schmidt et al., Int. J. Peptide Protein Res. (1995) 46(l):47-55). Analogues can also be made by replacement of the carboxyl oxygen with sulphur or replacement of the -C(O)- with -PO(OH)- or the like (see e.g. Sherman et al., J. Am. Chem. Soc. (1990) 112(1):433-441; and Hoffman et al., J. Org. Chem. (1995) 60(16):5107-5113). Analogues may also be generated by inclusion of additional atoms, e.g. carbon atoms, in the backbone, for example to yield ψ-carba, -alkene or -alkyne derivatives (e.g. the

-NH-CHR-C≡C-C(0)- alkyne). Other atoms may also be included in the backbone, e.g. oxygen, nitrogen and sulphur atoms, as well as functional groups such as hydrazide and amideoxy groups. Other analogues include hydroxymethylene and fluorovinyl derivatives (see e.g. Allmendinger et al., Tetrahedron Lett. (1990) 31(50):7297-7300), and sulfonamido derivatives (see e.g. Luisi et al., Tetrahedron Lett. (1993) 34(14):2391-2392). Further analogues include reduced peptides where the amide bond has been reduced to a methylene amine by treatment with a reducing agent, e.g. borane or a hydride reagent such as lithium aluminium hydride. Such a reduction typically has the added advantage of increasing the overall cationicity of the molecule. Preferred peptidomimetic backbones include polyesters, polyamines and derivatives thereof as well as substituted alkanes and alkenes. Yet further examples of amino acid analogues include N- or O-substituted amino acids, especially N-substituted glycine. These analogues form peptidomimetics known as peptoids and may be formed, for example, by the stepwise synthesis of amide- functionalised polyglycines (see e.g. Ostresh et al., Proc. Natl. Acad. Sci. USA ( 1994) 91 : 11138-11142) . Strongly basic conditions will favour N-methylation over O-methylation and result in methylation of some or all of the nitrogen atoms in the peptide bonds and the N-terminal nitrogen. The substituent may be one of the standard or non-standard amino acid side-chains described herein.

Further examples of amino acid analogues are those which comprise nonstandard side-chains. These may be present in the context of the natural peptide backbone or attached to backbone moieties which are chemically modified as defined herein. Examples of such compounds include natural amino acids with other moieties attached, e.g. modified lysine, cysteine, aspartate or glutamate residues, which may be modified by attachment of carbohydrates (e.g. oligosaccharides such as heparin), lipophilic substances (e.g. fatty acids such as lauric, myristic, palmitic, stearic, behenic, palmitoleic, oleic, linoleic or arachidonic acids), or small molecules (e.g. amino acids or formyl or acetyl groups). Other examples of non-standard amino acid side-chains include those found in carnitine, hydroxyproline, selenomethionine, ornithine, citrulline, lanthionine, hypusine, dehydroalanine, etc. (also see Ashton Crop et ah, Trends in Genetics (2004) 20(12): 625 -630, the contents of which is incorporated herein in its entirety). By way of example, analogues of hydrophobic amino acids (e.g. alanine, glycine, valine, and leucine) include a-methoxyglycine, a-phenylglycine, a-methyl-leucine, β-chloro-alanine, 3-fluoro-valine, 4,4,4-trifluoro-valine and a-t-butylglycine.

Analogues of basic amino acids (e.g. lysine and arginine) include ornithine, a- methyl-ornithine, citrulline and N-trimethyl- lysine. Analogues of polar amino acids (e.g. serine and threonine) include 2-amino-3-hydroxypentanoic acid, a- methylserine and 2-amino-3-hydroxy-3-methylbutanoic acid. Analogues of other standard amino acids would be apparent to the skilled person.

As would be appreciated by the skilled person, an amino acid analogue may have more than one "modification" relative to a standard amino acid. For example, it may comprise both a non-standard side-chain and also a non-standard backbone, or it may comprise both a non-standard side-chain and also a non-natural stereochemistry (e.g. D-isomerism). All compatible combinations of the types of analogue discussed herein are contemplated as forming part of the present invention.

Amino acid analogues may be obtained from commercial sources (e.g. the "unnatural" amino acids of Sigma- Aldrich) or synthesised according to methods known in the art.

Thus, as discussed above, peptidomimetics may comprise readily identifiable amino acid analogue sub-units; peptidomimetics may also mimic a peptide sequence in terms of the 3D spatial arrangement and presentation of functional groups of the peptide overall, while no longer comprising simple amino acid analogues (Kharb et al, J. Chem. Pharm. Res., 2011, 3(6): 173-186). Such peptidomimetics of the peptides of the invention described herein constitute a further aspect of the invention. In a preferred embodiment the molecule of the invention comprises (e.g. consists of) a peptide of formula II (SEQ ID NO. 35):

Z1-WX2X3DX4X5X6PI-Z2 (II) wherein Z _{l s} Z ₂ and X ₂ to Xe are as defined herein. Z _{l s} when present, preferably denotes A. X ₂ preferably denotes C, S or T (preferably C). X3 preferably denotes S, C or T (preferably S). X4 preferably denotes E or D (preferably E). X ₅ preferably denotes A or S (preferably A). X ₆ preferably denotes L, I or G

(preferably L). Z ₂ , when present, preferably denotes AS, WS or GS.

In a further preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula III (SEQ ID NO. 36):

Zi-XiX ₂ X3DEX ₅ X6PI-Z2 (III) wherein Z _{l s} Z ₂ , Xi to X ₃ , X ₅ and X ₆ are as defined herein. Z _{l s} when present, preferably denotes A. Xi preferably denotes W or F (preferably W). X ₂ preferably denotes C, S or T (preferably C). X3 preferably denotes S, C or T (preferably S). X ₅ preferably denotes A or S (preferably A). X ₆ preferably denotes L, I or G

(preferably L). Z ₂ , when present, preferably denotes AS, WS or GS.

In a further preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula IV (SEQ ID NO. 37):

Zi-WX ₂ SDEAX ₆ PI-Z ₂ (IV) wherein Z _{l s} Z ₂ , X ₂ and X ₆ are as defined herein.

In another preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula V (SEQ ID NO. 38): Zi-WX ₂ X3DEX ₅ LPI-Z ₂ (V) wherein Z _{l s} Z ₂ , and X ₂ to X ₅ are as defined herein. Preferably, Z _{l s} when present, denotes A. X ₂ preferably denotes C. X ₃ preferably denotes S. X ₄ preferably denotes E or D (preferably E). X5 preferably denotes A. Z ₂ , when present, preferably denotes -X ₇ X ₈ as herein defined, especially wherein X ₇ denotes G, A, M or W (preferably G or A) and X ₈ denotes S or Y.

Especially preferably, the molecule of the invention comprises (e.g. consists of) a peptide of formula VI (SEQ ID NO. 39):

Zi-WCSDEX ₅ LPI-Z ₂ (VI) wherein Z _{l s} Z ₂ and X ₅ are as defined herein. Preferably, Z _{l s} when present, denotes A. X ₅ preferably denotes A or S (especially A). Z ₂ , when present, preferably denotes GS, AS or WS (especially GS or AS).

In a particularly preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula VII (SEQ ID NO. 40):

Zi-WCSDEALPI-Z ₂ (VII) wherein Z _\ and Z ₂ are as defined herein. Preferably Z _{l s} when present, denotes A; and/or Z ₂ , when present, denotes GS, AS or WS (preferably GS or AS).

In especially preferred embodiments, the molecule of the invention comprises (e.g. consists of) a peptide having a sequence of any one of SEQ ID NOs 1 to 14:

Amino Acid Sequence SEQ ID NO:

WCSDEALPI 1

AWCSDEALPI 2

WCSDEALPIG 3

AWCSDEALPIG 4

WCSDEALPIGS 5

AWCSDEALPIGS 6

WCSDEALPIA 7

AWCSDEALPIA 8

WCSDEALPIAS 9

AWCSDEALPIAS 10 WCSDEALPIW 11

AWCSDEALPIW 12

WCSDEALPIWS 13

AWCSDEALPIWS 14

Especially preferred molecules comprise (e.g. consist of) a peptide having an amino acid sequence selected from SEQ ID NOs: 1, 6, 7, 10, 11 and 14, e.g. SEQ ID NO: 1, 6, 10 or 14. Molecules comprising (e.g. consisting of) a peptide having the amino acid sequence of SEQ ID NO: 6 are particularly preferred.

The peptide-containing molecules of the invention have a high binding affinity for their cognate MHC molecule and are capable of activating HPA-la specific T cells. Typically, molecules of the invention have a higher binding affinity for HLA-DRA/DRB3*01 :01 and/or are capable of activating T cells to a greater extent than their corresponding L33 analogues, e.g. molecules comprising a peptide having the amino acid sequence WCSDEALPL (SEQ ID NO: 15). Methods for determining the strength of peptide binding and the extent of T cell activation are known generally in the art and are also described in the following Examples.

The results of Figure 4 and Table 5 of the Examples taken together also indicate that modification of HPA-la peptides by altering G34 to a larger, hydrophobic residue such as alanine or tryptophan leads to significantly improved binding and T cell activation. The Examples indicate that, surprisingly, peptides having a larger hydrophobic residue at position 34 (relative to the glycine of integrin β3) bind more strongly to the MHC and can elicit a stronger immune response. Such peptides can be expected to be more efficient than the native G34 peptide in inducing tolerance to HPA-la in vivo.

Thus, in a further aspect, the present invention provides a molecule comprising (e.g. consisting of) a peptide of formula X (SEQ ID NO. 41): Zi-XiX ₂ X3DX4X5X6PX'7X'8-Z' ₂ (X) wherein and Z' ₂ denote optional peptide extensions at the N- and C- termini, respectively. When present, Z _\ denotes a hydrophobic amino acid residue or an analogue thereof, especially A. When present, Z' ₂ denotes a polar amino acid residue or an analogue thereof, e.g. S, T, Y or C, preferably S or Y, and especially S. In one embodiment Z _\ is present and Z' ₂ is absent; in another embodiment Z _\ is absent and Z' ₂ is present; in a further embodiment Z _\ and Z' ₂ are both absent; and in a yet further embodiment, Z _\ and Z' ₂ are both present.

Xi to X6 are as defined above.

X' ₇ denotes a hydrophobic amino acid, e.g. selected from L, A, I, V, G, M, Y and F, and analogues thereof. In one embodiment, X' ₇ is selected from Y and F, and analogues thereof, especially Y or F. In another embodiment, X' ₇ is selected from L, A, I, V and G, and analogues thereof. Preferably, X' ₇ is other than P. In one embodiment, X' ₇ denotes L, I or V, e.g. L or V, preferably L. In another

embodiment, X' ₇ does not denote I.

X's denotes a hydrophobic amino acid selected from A, I, L, V, W, M, Y and F (preferably from A, I, L, V, W and F), and analogues thereof, especially A or W.

In short, in this aspect, the peptides of the invention comprise amino acids 25 to 34 of integrin β3, in which residue 34 is modified, other amino acid modifications may also be present. The residue at position 34 (G34) has not been identified in binding sturdios as having a particular role in the interaction between the peptide and the binding pocket of the HHC.

In a preferred embodiment of this aspect of the invention the molecule of the invention comprises (e.g. consists of) a peptide of formula XI (SEQ ID NO. 42): wherein Z _{l s} Z' ₂ , Xi to X ₃ and X ₅ to X's are as defined above. Preferably, Z _{l s} when present, denotes A. Xi preferably denotes W or F (preferably W). X ₂ preferably denotes C, S or T (preferably C or S). X ₃ preferably denotes S, C or T (preferably S or C). X ₅ preferably denotes A or S (preferably A). X ₆ preferably denotes L, I or G (preferably L). X' ₇ preferably denotes L, I or V (preferably L). X's preferably denotes A, V, W or F (preferably A or W). Z' ₂ , when present, preferably denotes S or Y (preferably S). In a further preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula XII (SEQ ID NO. 43):

Zi-WX ₂ X3DEX ₅ X ₆ PX'7X' ₈ -Z'2 (III) wherein Z _{l s} Z' ₂ , X ₂ to X ₃ and X ₅ to X'g are as defined herein. Preferably, Z _{l s} when present, denotes A. X ₂ preferably denotes C, S or T (preferably C or S). X ₃ preferably denotes S, C or T (preferably S or C). X ₄ preferably denotes E or D (preferably E). X ₅ preferably denotes A or S (preferably A). X ₆ preferably denotes L, I or G (preferably L). X' ₇ preferably denotes L, I or V (preferably L). X' ₈ preferably denotes A, V, W or F (preferably A or W). Z' ₂ , when present, preferably denotes S or Y (preferably S).

In a further preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula XIII (SEQ ID NO. 44):

Zi-WX ₂ SDEAX ₆ PX'7X' ₈ -Z' ₂ (XIII) wherein Z _{l s} Z' ₂ , X ₂ and X ₆ to X' ₈ are as defined herein. Preferably, X' ₇ denotes L and X' ₈ denotes A or W.

In another preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula XIV (SEQ ID NO. 45):

Zi-WX ₂ X ₃ DX ₄ X ₅ LPLX' ₈ -Z' ₂ (XIV) wherein Z _{l s} Z' ₂ , X ₂ to X ₅ and X' ₈ are as defined herein. Preferably, Z _{l s} when present, denotes A. X ₂ preferably denotes C or S. X ₃ preferably denotes S or C. X preferably denotes E or D (preferably E). X ₅ preferably denotes A or S. X' ₈ preferably denotes A or W. Z' ₂ , when present, preferably denotes S or Y (preferably S).

Especially preferably, the molecule of the invention comprises (e.g. consists of) a peptide of formula XV (SEQ ID NO. 46): Zi-WCSDEX ₅ LPLX' ₈ -Z' ₂ (XV) wherein Z _{l s} Z' ₂ , X ₅ and X's are as defined herein. Preferably, Z _{l s} when present, denotes A. X ₅ preferably denotes A or S. X's preferably denotes A or W. Z' ₂ , when present, preferably denotes S.

In a particularly preferred embodiment, the molecule of the invention comprises (e.g. consists of) a peptide of formula XVI (SEQ ID NO 47):

Zi-WCSDEALPLX's-Z'2 (XVI) wherein Z _{l s} X's and Z' ₂ are as defined herein. Preferably Z _{l s} when present, denotes A; X's denotes A or W; and Z' ₂ , when present, denotes S.

In relation to all formulae describing peptides of the invention, X's is particularly preferably A or W, especially A.

In especially preferred embodiments, the molecule of the invention comprises (e.g. consists of) a peptide having a sequence of any one of SEQ ID NOs 17 to 32:

Amino Acid Sequence SEQ ID NO:

WCSDEALPLA 17

AWC SDEALPL A 18

WCSDEALPLAS 19

AWCSDEALPLAS 20

WCSDEALPLW 21

AWC SDEALPL W 22

WCSDEALPLWS 23

AWCSDEALPLWS 24

WCSDEAVPLA 25

AWC SDEVLPL A 26

WCSDEVLPLAS 27

AWCSDEVLPLAS 28

WCSDEAVPLW 29

AWCSDEVLPLW 30

WCSDEVLPLWS 31

AWCSDEVLPLWS 32

Especially preferred molecules comprise (e.g. consist of) a peptide having an amino acid sequence selected from SEQ ID NOs: 17 to 24, particularly from SEQ ID NOs: 17, 20, 21 and 24. Molecules comprising (e.g. consisting of) a peptide having the amino acid sequence of SEQ ID NO: 20 or SEQ ID NO: 24 are particularly preferred, with SEQ ID NO: 20 being especially preferred.

The peptide-containing molecules of the invention have a high binding affinity for their cognate MHC molecule and are capable of activating HPA-la specific T cells. Typically, molecules of this aspect of the invention have a higher binding affinity for HLA-DRA/DRB3*01 :01 and/or are capable of activating T cells to a greater extent than their corresponding G34 analogues, e.g. molecules comprising a peptide having the native amino acid sequence WCSDEALPLG (SEQ ID NO: 33). Methods for determining the strength of peptide binding and the extent of T cell activation are known generally in the art and are also described in the following Examples.

The peptide part of the molecules of the invention, which may make up the entire molecule or just an element thereof, comprises an epitopic fragment (e.g. as defined in the sequences listed above). Preferably, the molecules of the invention are peptides, i.e. they consist entirely of amino acids or analogues thereof. The epitopic fragment is typically of 9 or more amino acids in length, especially greater than 10, 11, 12, 13, 14 or 15 amino acids in length, e.g. at least 20, 25, 30, 50 or 100 amino acids in length. Preferably, the epitopic fragment is between 9 and 25, more preferably 9-16 amino acids in length, e.g. around 12 or 15 amino acids in length, i.e. 11 to 16 amino acids. The molecules of the invention may consist only of such an epitopic fragment.

In another embodiment, the epitopic fragment is extended at its N- and/or C- terminus to include other functionality, e.g. a linker, a tag, a immunological carrier, etc., or to cap the sequence, e.g. using an N-acetyl group or a C-terminal alcohol to provide an ester. This extension may itself be peptidic, e.g. a linker may comprise one or more amino acids having side-chains that are amenable to chemical or biological cross-linking (e.g. a cysteine or lysine residue). Alternatively, the linker may be present to provide spacing or conformational flexibility to allow attachment to another moiety or to allow the epitopic fragment to form the appropriate structure. For example, a linker may be used to attach the epitopic fragment to a tag such as a hexa-histidine tag or to biotin. One example of a suitable linker is a peptide having the amino acid sequence KSGGGSGGGSGGGSGG (SEQ ID NO: 16) which may be attached to the N-terminus of the epitopic fragment by a peptide bond via the carboxyl group on the terminal glycine residue of SEQ ID NO: 16. The linked moiety, such as biotin, may then be attached to the N-terminus of the linker, e.g. via a peptide bond.

The epitopic fragment may be attached (directly or indirectly) to an immunological carrier, which may be a protein, peptide, or the like. The purpose of attachment to the carrier is to enhance the immunological response of the epitopic fragment, i.e. to enhance the immune response and the efficacy of treatment with the molecule of the invention. Where the carrier is covalently attached to the peptide, for example where the carrier is directly bonded via a linker and/or other spacer (e.g. using peptide bonds), the molecule of the invention will comprise the carrier.

Where the carrier is not covalently attached to the peptide, for example where the carrier is associated via an interaction with a component of the molecule of the invention (e.g. via a biotin/streptavidin interaction), the molecule of the invention will form a complex with the carrier. This complex may be used in the medical methods described hereinafter.

Thus, in one embodiment then invention provides a molecule comprising a peptide as defined herein attached to an immunological carrier. In an alternative embodiment, the invention provides a complex comprising a molecule as defined herein and an immunological carrier, wherein said molecule and said immunological carrier are associated via non-covalent bonds.

Examples of immunological carriers include ovalbumin, bovine serum albumin (BSA), Tetanus toxoid, keyhole limpet hemocyanin (KLH) and the B subunit of cholera toxin.

The molecules of the invention may be generated using recombinant DNA technology, especially where they consist of amino acids, but may also be prepared using chemical synthesis, e.g. where the peptide part of the molecule is generated by stepwise elongation, one amino acid at a time before other parts of the molecule are added by chemical or biological reactions. Methods for synthesising peptides and peptidomimetics are well known in the art. The present invention also provides nucleic acid molecules encoding the peptides of the invention. Where the molecule comprises a peptidic carrier, linker and/or tag which is attached to the peptide by a conventional amide bond, the nucleic acid may encode the peptide and the carrier, linker and/or tag. The nucleic acid is preferably a ribonucleic acid (R A) or a deoxyribonucleic acid (DNA), especially a DNA, and is preferably an isolated nucleic acid molecule.

The present invention also extends to a vector (e.g. an expression vector) comprising the nucleic acid of the invention and to a host cell containing said nucleic acid or said vector. Suitable cloning and expression vectors and host cells are known in the art and include bacterial cells (e.g. E. coli), mammalian cells (e.g. CHO cells) and human cells (e.g. HeLa cells), and vectors for use therein, e.g. pET vectors.

The invention further provides a process for preparing a molecule according to the invention, the process comprising the expression of a nucleic acid of the invention and optionally coupling a peptide product of said expression with a carrier. Alternatively a molecule of the invention may be prepared by de novo peptide synthesis with optional coupling of the resulting peptide product to a carrier.

The present invention extends to pharmaceutical compositions comprising a molecule or complex of the invention and one or more physiologically-tolerable carriers, diluents or excipients. In a preferred embodiment the composition further comprises an adjuvant.

The composition may be any formulation suitable for parenteral or mucosal administration, e.g. in the form of a tablet, powder, aerosol, solution, suspension, dispersion, emulsion, gel, paste, syrup, cream, ointment, implant, suppository, re-dissolvable powder, granulate or lyophilisate. Formulations for mucosal administration may be preferred.

In certain embodiments, and to encourage tolerisation, the molecules of the invention are preferably administered in soluble form and preferably are

administered subcutaneously or orally.

Parenteral administration includes intravenous, intramuscular, subcutaneous, intradermal, epicutaneous/transdermal, intraperitoneal and intralymphatic routes.

Preferred parenteral compositions are in injectable, e.g. subcutaneously injectable, forms. Suitable formulations would be apparent to the skilled person and include solutions in sterile saline or other isotonic (or slightly hypertonic) media, such as Ringer's solutions (e.g. lactated Ringer's solution) or Hartmann's solution.

Parenteral formulations may alternatively be in the form of a re-dissolvable lyophilisate.

Suitable routes for mucosal administration include sublingual, buccal, nasal, inhalable, oral, vaginal and rectal routes. Preferred compositions are in an inhalable or orally administrable form. Suitable formulations would be apparent to the skilled person and include inhalable powder and tablet, or capsule, forms.

Examples of suitable carriers, excipients, and diluents include lactose, dextrose, sucrose, sorbitol, mannitol, starches, gum acacia, calcium phosphate, inert alginates, tragacanth, gelatine, calcium silicate, microcrystalline cellulose, polyvinylpyrrolidone, cellulose, water syrup, water, water/ethanol, water/ glycol, water/polyethylene, hypertonic salt water, glycol, propylene glycol, methyl cellulose, methylhydroxybenzoates, propyl hydroxybenzoates, talc, magnesium stearate, mineral oil or fatty substances such as hard fat or suitable mixtures thereof. Preferred excipients and diluents are mannitol and hypertonic salt water (saline).

The compositions may additionally include lubricating agents, wetting agents, emulsifying agents, suspending agents, preserving agents, sweetening agents, flavouring agents, and the like.

Adjuvants which may be incorporated in the pharmaceutical composition may be any conventional immunological adjuvant, including oxygen-containing metal salts (especially aluminium hydroxide, aluminium phosphate, aluminium acetate, calcium phosphate, calcium tartrate or zinc sulphate), heat-labile enterotoxin (LT), cholera toxin (CT), cholera toxin B subunit (CTB), polymerised liposomes, mutant toxins, e.g. LTK63 and LTR72, interleukins (e.g. IL-1 [beta], IL-2, IL-7, IL- 12, INF[gamma]), GM-CSF, MDF derivatives, CpG oligonucleotides, LPS, MPL, phosphophazenes, sterilised aluminium phosphate wet gels (e.g. Adju-Phos ^® ), glucan, liposomes, DDE, DHEA, DMPC, DMPG, DOC/Alum Complex, Freund's incomplete adjuvant, immunostimulating complexes (e.g. ISCOMs ^® ), LT Oral Adjuvant, muramyl dipeptide, monophosphoryl lipid A and

phospatidylethanolamine. Suitable oxygen-containing metal salts and methods for mucosal administration using these salts as adjuvants are described in WO00/45847, the content of which is hereby incorporated by reference in its entirety. The concentration of the adjuvant in the formulation is preferably between 0.01 and 100 mg/ml, especially between 0.1 and 10 mg/ml, e.g. between 0.5 and 5 mg/ml.

In a further aspect, the invention provides a molecule, complex or composition as defined herein for use in medicine (therapy). More specifically, the molecule, complex or composition is for use in the prophylactic treatment, treatment or prevention of NAIT. For convenience herein, and unless otherwise clear from the context, 'treatment' is used to refer to treatment in the sense of treating an existing condition, as well as prophylaxis and prevention. Thus, it is not necessary to establish whether a fetus has any signs of FNAIT or whether the mother is generating anti-HPAla antibodies (although these investigations may be undertaken) in order to perform a 'treatment' in accordance with the present invention.

Since anti-HPA la antibody formation and NAIT are dependent on HPA la- specific T cell responses, antigen-specific inhibition of these responses should inhibit or prevent antibody production and thus lessen or abrogate NAIT.

In a related aspect, the invention provides a method for the prophylaxis or treatment of NAIT, characterised in that a molecule, complex or composition as defined herein is administered to a subject. The subject will be a woman, in particular a woman identified as being in need of or potentially benefiting from such treatment, for example because she is homozygous for HPA- lb and, optionally, is pregnant or suspected of being pregnant or has carried a foetus which is or was HPA- la-positive; optionally the woman has been shown to have anti-HPA- la antibodies in her serum.

In another related aspect, the invention provides the use of a molecule, complex or composition as defined herein in the preparation of a medicament for use in the prophylaxis or treatment of NAIT.

Treatment typically comprises immunisation (i.e. vaccination) of a woman who is susceptible to carrying a child that could suffer from NAIT. In the context of the present invention, the form of immunisation which takes place is tolerization, the antigenic molecules of the invention stimulate the immune system in a way which renders it unresponsive to subsequent antigen exposure, as might happen if a woman is carrying an HPA- la-positive fetus. Treatment in accordance with the invention can tolerise HPA la-specific T cells in the woman to inhibit or prevent subsequent formation of anti-HPA la antibodies.

The subject to be treated is preferably a woman who is negative for HPA la (i.e. HPA lbb), preferably one who carries at least one DRB3*01:01 allele.

Alternatively viewed, the subject to be treated is the foetus or child. In one embodiment, the woman is treated before her immune system has started to produce antibodies against HPA la, i.e. who has an essentially undetectable level of anti- HPA la antibodies in her plasma. In another embodiment, the woman is treated after an immune response to HPA la has been established, i.e. when there is a detectable level of anti-HPA la antibodies in her serum. This treatment may be able to reverse the established immune response.

Treatment of N AIT may improve mortality rates or reduce the likelihood, severity or number of haemorrhages.

The molecules, complexes and compositions of the invention may, more generally, be used to prevent or manage conditions caused by exposure to HPA- la, e.g. following transfusion or during pregnancy. The treatment will typically be by tolerisation of the subject to HPA- la. Examples of such conditions are given in US patent application No. 2007/0042949 (the contents of which are hereby incorporated by reference in their entirety) including the prevention or management of posttransfusion purpura due to anti-HPA la and platelet refractoriness. 'Management' includes any amelioration in symptoms or reduction in clinical indicators of the condition. 'Prevention' includes partially effective preventative measures.

Accordingly, the invention provides a molecule, complex or composition as defined herein for use in the prevention or management of a condition caused by exposure to HPA- la. Such conditions would include post-transfusion purpura due to anti-HPA la and/or platelet refractoriness in a subject.

In a related aspect, the invention provides a method for the prevention or management of a condition caused by exposure to HPA- la (e.g. post-transfusion purpura due to anti-HPA la and/or platelet refractoriness), characterised in that a molecule, complex or composition as defined herein is administered to a subject. In such aspects, the subject may be male or female. In another related aspect, the invention provides the use of a molecule, complex or composition as defined herein in the preparation of a medicament for use in the prevention or management of a condition caused by exposure to HPA-la (e.g. post-transfusion purpura due to anti-HPA la and/or platelet refractoriness).

The invention will now be further described with reference to the following non-limiting Examples and Figures, in which:

Figure 1 shows T cell activation and the role of the L33 residue in the HPA- la peptide. CFSE labelled HPA-la- specific T cell clones (CFSE labelled) were stimulated with DRA/DRB3 * 01 : 01 -positive APCs pulsed with different peptides . T cell activation was measured by intracellular cytokine staining (IFNy) by flow cytometry after 18 hours incubated in presence of Brefeldin A (Upper panels - 1A). T cell activation measured was also by proliferation (reduction in CFSE) by flow cytometry after 7 days (Lower panels - IB).

Figure 2 shows the binding of different peptides to DRA/DRB3 * 01 : 01 positive APCs. Cells were pulsed with biotinylated peptides in the presence of AdEtOH for 4 hours, stained with Streptavidin-PE and analysed in flow cytometry. Substitutions of anchor residues W25→A and D28→A strongly reduced binding, as did L33→E/R (<40%). However, significant binding (>70%)was detected by the substitution L33→I. The x-axis is a logarithmic scale showing peptide binding (mean PE Intensity).

Figure 3 shows a comparison of peptide binding to different B-LCLs.

Efficient peptide binding of L33 is seen only in the DRB3*01 :01 positive cell lines D4BL4 and STEINLIN, and not in control cell lines DUCAF and EMJ.

Figure 4 shows peptide binding to DRA/DRB3 *01 :01 -positive cell line

STEINLIN. Binding is measured as binding ratio compared to L33 peptide (defined as 100 %). Ratios are based on PE Mean Fluorescence Intensity (MFI) with live gate on B-LCLs.

Figure 5 shows that high concentrations of P33 peptide can also activate T cell clones. Binding of peptide to HLA-DRB3*0101 positive antigen-presenting cells, measured in flow cytometry after peptide-pulsing with AdEtOH as MLE. Figure 6 shows T cell activation by P33 peptide at high concentrations. Corresponding T cell activation (to those shown in Figure 5) measured by TNFa secretion with clone D48T12 upon stimulation with APCs pulsed with increasing peptide concentrations.

Figure 7 shows peptide binding to DRA/DRB3 *01 :01 -positive cells for peptides having P, L, V and I at position 33 (Example 5).

Figure 8 is a bar plot showing the activation of T-cells measured by the percentage of IFNy secreting cells.

Figure 9 is a bar plot showing mean PE fluorescence intensity representing peptide binding of the different peptides to the surface of the antigen-presenting cells at 0.5 μΜ and 5 μΜ concentration.

Figure 10 is a bar plot showing activation of T-cells after stimulation presented as mean fluorescence intensity of PerCP-Cy5.5 (TNFa) and APC (IFNy). Bars represent the mean values, whiskers indicate the range (duplicates).

Examples

Example 1 - Production of peptides Peptide synthesis

Peptides were synthesised by the Biotechnology Centre of Oslo, University of Oslo, Norway. All peptides were dissolved in 20μί DMSO, and diluted to 1 mL in sterile water, and further diluted to 88μΜ stock solutions in 0.2% PBSA buffer and stored at -20°C. The peptides used for T-cell stimulation assays are listed in Table 1 below. Identical peptides with an N-terminal biotinylated linker (biotin- KSGGGSGGGSGGGSGG-peptide) were used for peptide-binding assays.

Table 1 - List of peptide sequences

Antigen Peptide Amino acid sequence * SEQ ID No.

L33

HPA-la (Integrin β3 CSPMCAWCSDEALPLGSPRC 48

19-38)

P33

HP A- lb (Integrin β3 CSPMCAWCSDEALPPGSPRC 49

19-38)

LolPl (196- control VWRIDTPDKLTG 50

208)

L33

HPA-la (Integrin β3 AWCSDEALPLGS 51

24-35)

P33

HP A- lb (Integrin β3 AWCSDEALPPGS 52

24-35)

A25 AACSDEALPLGS 53

125 AICSDEALPLGS 54

A26 AWASDEALPLGS 55

A27 AWCADEALPLGS 56

A28 AWCSAEALPLGS 57

N28 AWCSNEALPLGS 58

E28 AWCSEEALPLGS 59

R28 AWCSREALPLGS 60

V28 AWCSVEALPLGS 61

D29 AWCSDDALPLGS 62

A29 AWCSDAALPLGS 63

Q29 AWCSDQALPLGS 64 L29 AWCSDLALPLGS 65

T30 AWCSDETLPLGS 66

R30 AWCSDERLPLGS 67

V30 AWCSDEVLPLGS 68

A31 AWCSDEAAPLGS 69

V31 AWCSDEAVPLGS 70

A32 AWCSDEALALGS 71

V32 AWCSDEALVLGS 72

133 AWCSDEALPIGS 6

V33 AWCSDEALPVGS 73

E33 AWCSDEALPEGS 74

R33 AWCSDEALPRGS 75

A34 A WC SDE ALPL AS 76

W34 AWCSDEALPLWS 77

Y35 A WC SDE ALPLG Y 78

* Anchor residues for DRA-DRB3*01 :01 binding are underlined, and substituted amino acids are in bold.

Example 2 - Isolation of different HP A- la-specific T cell clones

The study was approved by the Regional Committee for Medical Research Ethics, North Norway. Blood samples were drawn from donors after written informed consent was obtained in accordance with the Declaration of Helsinki. Human donors

Peripheral blood was drawn from HPA-la immunized women. Donor #8 is a

54-year-old female, who gave birth to a severely thrombocytopenic child in 1980.

Subsequent investigations revealed that the mother is HPA-lbb, and her affected daughter is HPA-la positive. Anti-HPA-la IgG antibodies was detected in maternal plasma. The anti-HPA-la level was ~30 IU/mL at recent measurements (2008 and

2010). The donor carries one DRB3*01 :01 allele, in addition to DRB1 *01 and

DRB1 * 13.

Donor #48 is a HP A- la-negative 41 -year-old female, who gave birth to a severely thrombocytopenic, HPA-1 incompatible child in 1999, and had a maternal anti-HPA-la IgG level of 37 IU/mL at the time of delivery, and another

thrombocytopenic male infant in 2003 where maternal antibody level at delivery was 17 IU/mL. The level was 108 IU/mL in the blood sample drawn in 2008, which was used for isolation of cells in this study. The donor carries two DRB3*01 :01 alleles, and is otherwise homozygous for DRB 1 *03 :01 , DQA1 *05 :01 and

DQB1 *02:01. Handling of blood samples

Blood samples were centrifuged; plasma was collected, and the PBMC isolated by Lymphoprep density reagent (Ahlen et ah , Blood (2009) 113:3838- 3844). Cells were cryo -preserved in freezing media (10% DMSO, 90%> heat- inactivated FBS), and stored in liquid nitrogen until use. Frozen cells were rapidly thawed at 50°C, and washed immediately in cold complete media (CM) consisting of Iscove modified Dulbecco medium (IMDM), 10% FBS, 4% serum from an HPA- lbb donor and penicillin/streptomycin. All thawed cells were rested overnight at standard cell culturing conditions; 37°C in a 7.5% C0 ₂ humidified atmosphere, and live cells were collected using Lymphoprep. In one experiment, fresh donor PBMCs were used.

Antibody reagents

Anti-human CD4 mAb (PE-Alexa-610), anti-human CD3 mAb (APC), anti- human CD8 (FITC), anti-human CD 19 (FITC) and anti-human CD45RA (FITC) and anti-human IFNy mAb (APC) were all from Invitrogen (Carlsbad, CA, USA). Anti-human CD25 mAb (PE), anti-human CD14 mAb (APC-Alexa-750), and anti-human TNFa mAb (PE) were from BD Biosciences (San Jose, CA, USA).

Isolation of HP A- la specific T cell clones

Clonal HP A- la-specific T cells were isolated by different methods:

Method 1 : CFSE (Invitrogen) proliferation assay of Donor #8 PBMCs (1 x 10 ⁶ cells per reaction; cryo -preserved) was performed by one round of stimulation using synthetic L33 peptide (12-mer), or control peptide LolPl (20-mer), in CM, supplemented with interleukin (IL)-2 (30U/mL; Peprotech, London, UK) on day 2. On day 8, cells (200 μί) were stained with anti-CD4 mAb (Ι μί) and anti- CD14 mAb (2\iL) and proliferated (CFSE low) CD 14 ^" CD4 ⁺ cells from the L33 stimulated culture were single-cell sorted on FACS Aria (BD, Franklin Lakes, NJ, USA);

Method 2: Initial staining of cryo -preserved PBMCs from the two alloimmunized donors, Donor #8 and Donor #48, with anti-CD4 mAb (Ιμί) and anti-CD25 mAb (2μί) in 200 μί, was followed by a pre-sort of CD4 ⁺ CD25 ^dim cells on FACS Aria (BD). The CD4 ⁺ CD25 ^dim cells were labelled with CFSE and stimulated with synthetic L33 peptide, or control peptide LolPl, in CM

supplemented with IL-2 (30U/mL; Peprotech). After 8 days, proliferating cells were single cell sorted;

Method 3 : 15 10 ⁶ fresh PBMCs from Donor #8 were rested in CM (10%

Human Serum, no FBS) for a week, live cells were collected by standard

lymphoprep method, and stimulated with L33 peptide or control peptide LolPl, in CM (2μΜ, 20-mer peptides) in the presence of anti-human TNFa (0.2μί) and ΙΟμΜ TAPI-0 in a total volume of 200μΙ. at 37°C. After 4.5 hour, anti-CD4 mAb (Ιμί), and FITC-labelled (dumping channel for exclusionary gating) antibodies;

anti- human CD8 mAb (2μί), anti- human CD 19 mAb (2μί) and anti- human CD45RA mAb (2μ1) were added for 10 minutes at room temperature, washed and CD4 ⁺ TNFa-producing cells were single-cell sorted; and

Method 4: 5x 10 ⁶ cryopreserved PBMCs from Donor #8, were labelled with CFSE, and stimulated with 10 ⁷ HP A- la-positive platelets for 14 days. Cells were stained with anti-CD4 mAb (Ιμί), and proliferated (CFSE low) CD4 ⁺ cells were single sorted by FACS.

Samples for sorting were washed in 0.2% PBSA and diluted in CM with 40%) FBS, and sorted by single cell sorting into 96 well round-bottom plates with ΙΟΟμί of CM, by FACS on FACSAria (BD) using low-pressure stream and flow rate no higher than 5 (FACS Diva Software, BD). All single-sorted cells were expanded using anti-CD3, IL-15 and feeder cells (PBMC and B-LCLs), and after two weeks, successfully expanded clones were specificity- tested by ELISPOT assays as previously described (Ahlen et al., Id.). All cell cultures were kept at 37°C in a 7.5% C0 ₂ humidified atmosphere. All T cell clones were maintained by monthly expansion, and supplemented with IL-2 (50 ng/niL) and IL-15 (5 ng/mL) every 3-5 days. T cell receptor gene analyses

Total RNA was isolated from ~1 x 10 ⁶ clonal T cells using RNeasy Mini Kit (QIAGEN, Hilden, Germany) eluted in H ₂ 0. cDNA was synthesized from 500ng total RNA by Superscript Reverse Transcriptase III (Invitrogen) and random primers (Promega, Madison, WI) in 30-μί reactions. For determination of the TRBV, a set of 35 forward primers and a common reverse primer were used, as previously described (Ahlen et al, Id.; and Lee et al, J Immunol (1998) 161:4183-4194). In cases where no amplification was seen with these primer sets, but with successful amplification of the TRBC internal control, the TRBV primer sets reported by

Akatsuka was used (Akatsuka et al, Tissue Antigens (1999) 53: 122-134). TRAV determination was performed, by the same strategy and cycling programs, using a set of 34 forward primers and a common reverse primers (Yao et al., Cell Mol Immunol (2007) 4:215-220). All assays were performed using QuantiTect SYBR Green PCR kit (QIAGEN), on ABI-Prism 7900-HT (Applied Biosystems, Foster

City, CA). Results were analysed using Sequence Detection Systems 2.2.2 software (Applied Biosystems). For TRAV, occasionally two V-region-primers will give positive amplification signal, and in such cases both are used for sequencing, as a given clone can have more than one rearranged transcript for TCRa, however in general only one should have a productive rearrangement. When the AV and BV primers were determined for a given clone, PCR with these primers was performed, and sequencing as previously described (Ahlen et al, Id.).

Antigen-presenting cells

HLA class II defined EBV-transformed B-LCL, were obtained from

International Histocompatibility Working Group (IHWG; Seattle, WA, USA); EMJ (IHW#9097), STEINLIN (IHW#9087) and DUCAF (IHW#9019). All

lymphoblastoid cells were cultured in IMDM, 10% FBS and penicillin- streptomycin. In addition, an EBV-transformed B-lymphoid cell line from a HPA- lbb donor was used (D4BL4). All APC cell lines are listed in Table 2 below. Table 2 - HLA class II defined B-LCLs as antigen-presenting cells

HLA class II

Cell line DRB3 DRB1 DQAl DQB1 DPA1 DPB1

STEINLIN (IHW#9087) ^{† ♦} 01 :01 *03:01 *05:01 ^♦ 02:01 ^♦ 01 *03:01,*04:01

DUCAF (IHW9019) ^† *02:02 *03:01 *05:01 ^♦ 02:01 ^♦ 01 ^♦ 02:02

EMJ (IHW9097) ^† *03:01 * 13:02 ^♦ 01 :02 ^♦ 06:04 ^♦ 01 *03:01,*04:01

D4BL4 *01 :01,*02:02 *03:01,* 14:54 ^♦ 05:01, ND *02:01,*05:03 ND ^♦ 04:01, *32:01

† Cell lines were obtained from 10 Workshop B-LCL Reference Panel (IHWG).

ND: not determined. Peptide pulsing of B-LCLs

T cell clones were labelled with CFSE (Invitrogen). B-LCLs were pulsed with 0.5-15 μΜ peptide for 90 or 180 minutes at 37°C in presence of 2.5 μΜ

Adamantane Ethanol (AdEtOH; Sigma- Aldrich Inc, St. Louis, MO) as an MHC- Loading Enhancer (MLE), and then washed twice in PBSA (0.2%) and once in

TCM to ensure minimal rest of unbound peptide. For peptide binding assays

biotinylated peptides were used, and the cells were stained with Streptavidin-PE, washed and analysed in flow cytometry on FACS CantoII (BD). Data were

analysed using FlowJo7.5 software (TreeStar, Ashland, OR). Peptide binding assays were performed using several peptide concentrations between 0.5 and 15μΜ.

Intracellular cytokine staining

Production of IFNy was measured after co-culture of T cells and peptide- pulsed B-LCLs for -18 hours. Brefeldin A (Sigma) was added after 90 min of incubation to a final concentration of Cells were washed and stained with APC-labelled anti-human IFNy mAb (Invitrogen). Experiments were analysed in flow cytometry (FACS Canto II;BD).

TNFa secretion

The production of TNFa by T cells was measured after 4.5 hours of

stimulation in the presence of anti-human TNFa (PE) mAb (BD Biosciences) and

10μΜ TAPI-0 (Merck KGaA, Darmstadt, Germany) and analysed using flow

cytometry (FACS Canto II;BD). TAPI-0 (TNFa Protease Inhibitor) efficiently inhibits the cleavage of the membrane-bound form of TNFa produced upon cell activation, to soluble TNF by TACE (TNFa Convertase; ADAM 17).

Proliferation assay

T cell clones labelled with CFSE were stimulated with peptide-pulsed B-LCLs. Cultures were supplemented with IL-2 (30U/mL) and IL-15 (5 ng/mL) on day 2. Cells were analyzed on FACS CantoII (BD Biosciences) after 7 days.

Results

For detailed and representative analysis of antigen recognition by HP A- la- specific T cells, sixteen different HP A- la-specific clonal T cell lines were established in long-term cultures. Two of these (D7T1 and D7T4) have been described previously (Ahlen et ah, Id.).

Six unique clones were directly isolated by detection of TNF -secreting cells following antigen stimulation of 15 l0 ⁶ PBMCs (Donor #8; D8T102, D8T103, D8T57, D8T106, D8T107, D8T108). From CFSE proliferation culture experiments; 3 unique clones were identified from l lO ⁶ PBMC (Donor #8; D8T7, D8T37, D8T57), 2 unique clones were identified from l .l x lO ⁶ pre-sorted CD4 T cells (Donor #8; D8T57, D8T64) and 3 unique clones were identified from 1.8>< 10 ⁵ pre-sorted CD4 T cells (Donor #48; D48T1, D48T10, D48T12). An additional 2 unique clones (Donor #8; D8T48, D8T114) were identified by the platelet- stimulated Donor #8 cultures, and several previously identified clones from this donor were also re-isolated with this method.

The CDR3s of the rearranged TCRa and TCR-β are crucial for interaction with the MHC-presented antigenic peptide, thus defining the specific recognition of each T cell clone. All the 12 clones described above have unique TCRa and TCRP chains. The TCRA and TCRB rearrangements for the HPA-la specific T cell clones are listed in Tables 3 and 4 below (using IMGT nomenclature). Table 3 - The TCRA rearrangements in the T cell clones

TCRA rearrangements CDR3 SEQ ID NO.

Clone TRA V TRAJ Length Sequence

D7T1 not identified

D7T4 17*01 57*01 12 CATFAQGGSEKLVF 79

D8T7 26-1 *01 or *02 23*01 14 CIVRLSYYNQGGKLIF 80

D8T37 AV36/DV7*02 or 7*01 9 CAVGGNNRLAF

81

*03 or *04

D8T57 2*01 or *02 8*01 11 CAVMNTGFQKLVF 82

D8T64 8-3*01 or *02 23*01 14 CAVSNNQGGKLIF 83

D8T48 not identified

D8T102 8-4*01/*04 37*01 12 CAVSEIGNTGKLIF 84

D8T103 9- 42*01 11 CALGVGSQGNLIF

85

2*01/*02/*03/*04 (+ one unproductive)

D8T106 8-4*01/8-4*04 8*01 12 CVVSEGTGFQKLVF 86

D8T107 25*01 44*01 11 CAVYTGTASKLTF 87

D8T108 8-6*02 29*01 12 CAVRHNSGNTPLVF

88

(+ one unproductive)

D8T114 AV14/DV4*02/*03 42*01 10 CAGGGSQGNLIF 89

D8T114 12-3*01/*02 52*01 13 CAMSAGGTSYGKLTF 90

D48T1 17*01 8*01 11 CATSFTGFQKLVF 91

D48T10 26-l *01/*02 43*01 11 CIVRAYNNNDMRF

92 (+ one unproductive)

D48T12 8-4*03 37*01 12 CAVSAQGNTGKLIF 93

Table 4 - The TCRB rearrangements in the T cell clones

SEQ ID

TCRB rearrangements CDR3 NO.

TRB

Clone TRBV D TRB J TRBC Length Sequence

D7T1 29-1 1 *01 1-2*01 1 9 CSGGDGRGYT 94 *01/*03 F

D7T4 12-3*01/12- 1 *01 2-1 *01 2 12 CASRGTSLYN

95 4*01 EQFF

D8T7 9*01 1 *01 1-3*01 1 11 CASAQGFGNTl

96

YF

D8T37 11-2*01 1 *01 1-2*01 1 12 CASSLVQGGY

97

GYTF

D8T57 12-3*01/12- 2*01 2-6*01 2 13 CASRTSGREA

4*01 /2*0 NVLTF 98 z

D8T64 11-2*01 1 *01 2-5*01 2 11 CASSFRHGET

99 QYF

D8T48 7-2*01/*04 1 *01 1-2*01 1 11 CASSSDILRGY

100 TF

D8T102 2*01/2*02 2*01 2-3*01 2 11 CASSSRATDT

101 QYF

D8T103 9*01 - 2-2*01 2 11 CASSVVVTGE

102 LFF

D8T106 20-1 * 1 *01 1-2*01 1 11 CSARVPQVYG

01/02/04/05 YTF 103 /06

D8T107 11-1 *01 2*01 2-1 *01 2 13 CASSFRPRGV

104 NEQFF

D8T108 20- 2*02 2-2*01 2 17 CSARGSQGLA

105 l */04/05/06 GTHTGELFF

D8T114 6-5*01 1 *01 1-3*01 1 16 CASSPSPPVPG

106 SGNTIYF

D48T1 3-1 *01 2*02 2-1 *01 2 15 CASSQGVGGG

107

IYENEQF

D48T10 not identified

D48T12 12-3*01 1 *01 1-6*02 1 15 CASSLGQTTIF 108 LKPLHF

The isolation of several different HP A- la-specific T cell clones is indicative of a diverse response. Example 3 - Assessing binding affinity and ability to activate T cells of various modified HPA-la derived peptides

The Leu33 residue defines the HPA-la alloantigen in the β3 integrin polypeptide and is by definition a part of the antigen recognized by various anti- HPA-la antibodies. The β3 integrin-derived peptide has also been shown to fit into the MHC class II molecule DRA/DRB3*01 :01 , where the Leu33 residue is reported to be important for anchoring the peptide, by docking into the P9 pocket of

DRA/DRB3*01 :01 (Parry et ah, Id.). If Leu33 serves as an anchor residue, its role in T cell recognition may be due more to binding of the peptide epitope that is recognized by HP A- la-specific T cells to the DRA/DRB3*01 :01 molecule and less to direct recognition of this residue by the T cell receptor. Also, the side chain of the anchor residue is predicted to be buried in a pocket of the peptide-binding groove of the MHC molecule and therefore not likely to be directly contacted by the T cell receptor. To test for the role of the Leu33 residue in the activation of HPA- la- specific T cells, HPA-la and HP A- lb peptides and variants with substitutions of residue 33 were tested for their capacities for both binding to DRA/DRB3*01 :01- positive APCs and for T cell activation using methods described in Example 2. Results are shown in Figures 1 to 6.

HPA-la (L33) peptide both bound efficiently to the APCs, and activated HP A- la-specific T cell clones, while HP A- lb (P33) peptide bound poorly and was poor at activating T cell clones. Notably, the binding of the P33 peptide was not completely absent at the relatively low peptide concentrations employed in these experiments. Similarly, the peptides R33 and E33, with charged and bulky side chains, thus predicted not to fit into the hydrophobic P9 pocket of

DRA/DRB3 *01 :01 , bound poorly to the APCs and did not activate the T cell clones. However, the two peptides with substitutions at residue 33 with small, hydrophobic side chains, 133 and V33, bound efficiently to the APCs and efficiently activated the T cell clones (with one exception), measured by proliferation and secretion of cytokines (IFNy ) shown in Figure 1 (peptide sequences are listed in Table 1). The 133 peptide bound with the highest efficiency.

This result suggested that the hydrophobic Leu33 residue of HP A- la is not directly contacted by the TCR of HP A- la-specific T cells and that this residue confers T cell activation by virtue of anchoring the HPA-la-peptide to the MHC molecule, thus rendering it detectable by T cells. However, TCR contact of a hydrophobic residue at position P9 could possibly be required for T cell activation. As noted above, the P33 peptide displayed poor but detectable binding at relatively low peptide concentrations. This could be due to the presence of the anchor residues at position PI and P4 providing partial binding affinity to MHC. Indeed, alanine substitutions of the anchor residues W25 and D28 clearly showed reduced binding (Figure 2), demonstrating the effect of these residues for binding of the peptide to the DRA/DRB3*01 :01 molecule. Neither T cell activation (data not shown) nor significant peptide binding (Figure 3) were seen in experiments with control peptides L33 and P33 in the MHC class II matched DRB3*01 :01 -negative cell lines DUCAF and EMJ. Peptide-binding could indeed be detected without any MLEs, however the signals were weak. AdEtOH was used as an MLE in the peptide binding assays.

The binding of each modified peptide to DRA/DRB3*01 :01-positive APCs were compared to peptides L33, LolPl and P33, and each peptide had a binding ratio calculated in percent of L33 mean fluorescent intensity. In general, the binding ratio for a given peptide was comparable for different peptide concentrations. The 40% and 70%> limits could discriminate between the peptides, as only very few peptides showed binding ratios between 40%> and 70%> (Figures 2 and 4). Peptides were grouped according to their peptide-binding ratio at a 5μΜ peptide

concentration, as high-binding (more than 70% binding compared to L33) and low- binding (less than 40%). As all of the T cell clones were stimulated efficiently with APCs pulsed with L33 peptide within the range of 2.5-5 μΜ, it was considered that a given peptide with a binding ratio of 70%> should be bound sufficiently to stimulate T cells - if recognized by the T-cell receptor. It was reasoned that if a hydrophobic residue at position P9 is not required as a TCR contact residue, then the entire T cell epitope contacted by the TCR might indeed be present on the P33 peptide. Whether higher concentrations of P33 could result in increased peptide binding to MHC and subsequent T cell activation was therefore assessed. The peptide binding assays showed that increasing

concentrations of peptide increased peptide binding to the same extent for L33 and LolPl (binding control) peptides, whereas no increase was seen for the linker control, ensuring that the increase seen is due to binding of the peptide, not the biotin-linker. Indeed, for peptide P33, higher peptide concentrations, also resulted in detectable peptide binding in a dose-dependent manner (Figure 5). Moreover, increased P33 peptide concentration also resulted in T cell activation (Figure 6). Thus, although non-immunogenic at physiological concentrations, the P33 peptide binds with relatively low affinity to DRA/DRB3*01 :01 -positive APCs and high concentrations of this peptide can activate HP A- la-specific T cell clones. No T cell activation was seen with the LolPl peptide control at these high peptide

concentrations, although very high binding was demonstrated, supporting that the integrity of T cell specificity is not affected by the increased peptide binding.

Peptide binding assays were performed with the DRB3*01 :01 positive B- LCLs STEINLIN and D4BL4, in addition to the control cell lines (negative for DRB3*01 :01, otherwise matching the class II variants on STEINLIN), and expressing the DRB3*02:02 or DRB3*03:01 variants (DUCAF and EMJ, respectively). STEINLIN is HPA-la positive and expresses endogenous β3 integrin, presented by DRA/DRB3*01 :01, resulting in background stimulation of high- responsive T cell clones. However, whereas STEINLIN is homozygous for

DRB3*01 :01 (two copies), D4BL4 is hemizygous for DRB3*01 :01 (one copy). Comparable peptide binding was seen, however somewhat higher intensities (5-30%) were repeatedly detected with STEINLIN compared to D4BL4 (Figure 3), likely due to increased expression of DRA/DRB3*01 :01 on the surface of the homozygous cells.

The results of Example 3 indicate that L33 is not directly recognised by the

TCR of most HPA-la specific T cell clones. This is in accordance with the current model for the binding of the HPA-la peptide to MHC. However, it was surprisingly found that peptides having isoleucine in place of L33 bind with greater affinity and activate T cells to a greater extent.

Example 4 - Recognition of the peptide :MHC complex by HPA-la specific clones

The impact that individual amino acid substitutions have on the MHC- binding efficiency of the peptide was then investigated using methods as described in Example 2. Without this information, a lack of T cell activation may erroneously be interpreted as lack of direct T cell recognition, while it may indeed be due to poor or reduced binding of the peptide to MHC (it may be noted that all peptide bindings are shown in Figure 4 as a representative experiment with triplicates). Results of these experiments are shown in Table 5, below, in which responsiveness of T cell clones when stimulated with DRA/DRB3*01 :01 positive APCs pulsed with modified peptides was measured by TNFa secretion with "-" designating <5 % (no response), "+" 5-40 %, "++" 40-80%, and "+++" >80%. "(+)" designates a weak response in some experiments.

- Responsiveness of T cell clones

Peptide Donor #7 Donor #8 Donor #48

T4 T7 T48 T64 T102 T103 T105 T106 T107 T108 T114 Tl T10 T12

None - - - - - - - - - - - - - -

LolPl - - - - - - - - - - - - - -

L33 +++ ++ +++ + ++ + +++ +++ +++ +++ +++ + +++ ++

P33 - - - - - - - ++ - - + - - (+)

A25 - - - - + - - + ++ - - - - -

125 + + +++ - ++ ++ +++ +++ +++ +++ - + (+) ++

A26 - + +++ ++ - ++ +++ - +++ +++ - ++ - -

A27 + + + ++ ++ ++ +++ +++ - - - + (+) ++

A29 - - - - + - - + - - - - - -

D29 - - - - ++ - - +++ - - - - - -

Q29 - - - - + - - - - - - - - -

L29 - - - - - - - - - - - - - -

T30 ++ - +++ - ++ - ++ +++ +++ - - ++ - ++

V30 - - - - +++ - ++ + - - - - - +

V31 - + +++ - ++ - ++ +++ - - + - - -

A31 - - - - ++ - - - - - - - - -

A32 + - - - - - + +++ - - - + - -

V32 - - - - - - + ++ - - - + - -

133 ++ ++ +++ + ++ + ++ +++ ++ +++ - + +++ +

V33 ++ ++ ++ + ++ + ++ +++ ++ +++ - + +++ +

A34 + ++ +++ + ++ ++ +++ +++ +++ +++ + ++ ++ ++ W34 ++ ++ ++ ++ +++ + +++ +++ +++ +++ - ++ (+) ++

Y35 ++ ++ +++ + +++ ++ +++ +++ +++ +++ - ++ (+) ++

Several modified peptides with particular substitutions comprising PI, P4, P6 and P9 anchor residue substitutions (β3 integrin residues 25, 28, 30 and 33) clearly demonstrated reduced binding (binding ratio < 40 % compared to L33 peptide) to DRA/DRB3*01 :01 APCs. These substitutions typically involved amino acid changes, with fundamentally different properties compared to the original residue. However, substitutions to more similar amino acids in these anchor positions could mimic the effect of the original residue in some cases.

The recognition by HP A- la-specific T cells were tested for peptides with demonstrated efficient binding to DRA/DRB3*01 :01 -positive APCs. Peptides with (PI and P9) anchor residue substitution allowing efficient binding, also generally stimulated T cells to secrete TNFa. No substitutions in "non-anchor" residues reduced binding to DRA/DRB3*01 :01 -positive APCs (STEINLIN and D4BL4); as they all demonstrated binding ratios >70 %. These residues; 26, 27, 29, 31 and 32, are all potentially in contact with TCR, as they are predicted not to be hidden in the peptide-binding groove of the MHC molecule. Residues E29, L31 and P32 constitute specific parts recognized by most clones, as Glu29→Val/Leu/ Asp/Gin, and Leu31→Ala/Val and Pro32→ Ala/Val substitutions disrupted the recognition, despite that both peptides demonstrated efficient binding to MHC. It is also worth mentioning, that the Glu29→Asp substitution allowed strong recognition by a couple of clones, considering that the side chain of Asp has comparable chemical properties to Glu, however being sterically slightly smaller. The T cell

responsiveness to the different peptides was consistent in repeated experiments, however some exceptions in responsiveness were in fact seen, probably due to the status of the T cell clone upon testing (ie. time since last expansion, cell confluence etc) or status of the APCs (time since last lymphoprep, cell confluency). Overall, the results demonstrate that the T cell recognition of the peptide:MHC by the T cell clones was heterogeneous. Example 5 - Binding of 133 peptide to APCs expressing DRA/DRB3*01 :01

B-lymphoblasts (APCs) expressing the DRA/DRB3*01 :01 MHC molecule, were pulsed with the HPA-la peptide (L33) extended with a biotinylated linker, or single residue variants of the L33 with linker: P33, V33 or 133 peptides, at concentrations ranging from 0.1 to 10 μΜ as described above. Peptide binding was measured by flow cytometry after reacting the peptide-pulsed APCs with PE- conjugated streptavidin.

Results are shown in Figure 7 and are representative of several similar experiments. The 133 peptide binds more strongly to DRA/DRB3*01 :01 positive cells than the native L33 peptide or the V33 peptide across all concentrations.

The results of the Examples presented above indicate that peptides having a leucine to isoleucine substitution at P9 (corresponding to position 33 of integrin β3) bind more strongly to the MHC and can elicit a stronger immune response. Certain other mutations in the core epitopic sequence of HPA-la may also increase binding efficiency and T cell activation. Overall, this suggests that peptides having the 133 substitution (as well as other mutations as defined herein) will be more efficient than the native L33 peptide in inducing tolerance to HPA-la in vivo.

Example 6 - T cell activation studies

For T cell activation studies we employed an in- house HPA-lbb HLA-

DRB3*01 :01-positive EBV-transformed B-lymphoblast cell line D4BL4 as antigen- presenting cells (DRB 1 *03:01, DRB1 * 14:54, DRB3*01 :01, DRB3*02:02,

DQA1 *05:01, DQB1 *02:01 DQB1 *05:03, DPA1 *01, DPB1 *04:01, DPB1 *32:01).

D4BL4 cells were peptide-pulsed with 5μΜ peptide 12-mer peptides for 4 hours, without any MHC loading enhancer.

(P33) AWCSDEALPPGS (SEQ ID NO. 52)

(L33) AWCSDEALPLGS (SEQ ID NO. 51)

(133) AWCSDEALPIGS (SEQ ID NO. 6)

After peptide pulsing, the D4BL4 cells were washed in PBSA (0.2%). HPA-la-specific T cell clones (D48T10, D8T37, D8T57) 30.000 cells were stimulated with 30.000 peptide-pulsed D4BL4 cells in a total reaction of 200 μΐ with 5μg/mL Brefeldin-A and incubated for 10 hours at 37°C. After incubation, cells were surface stained with anti- human CD4 mAb, fixed in ΙΟΟμί fixation buffer

(PBS 4% Paraformaldehyde) for 10 minutes, washed once in ΙΟΟμΙ. PBSA (0.2%) and permeabilized with 30 0.1% saponin buffer for 10 minutes. Cells were stained with APC-labelled anti-human IFNy mAb, and analysed in flow cytometry (FACS Canto). Data was analysed with FloJo Software.

Data for the three HP A- la-specific T cell clones D48T10, D8T37 and D8T57 indicate that 133 has potential to stimulate HPA- la-specific T cell clones at least as well as L33 (Figure 8). L33 is the native HPA-la peptide, P33 the HPA- lb peptide, and 133 the designed peptide for potential tolerization.

Example 7 - Further T cell activation studies

We performed new T cell activation experiments with the HPA- la-specific T cell clone D48T10 using a new batch of 12-mer peptides with biotin- linker

(synthesized in Tromso), with >95% purity (HPLC/UPLC)

(P33) Biotin- KSGGGSGGGSGGGSGGG-AWCSDEALPPGS

(L33) Biotin- KSGGGSGGGSGGGSGGG-AWCSDEALPLGS

(133) Biotin- KSGGGSGGGSGGGSGGG-AWCSDEALPIGS B-LCL D4BL4 were again employed as antigen presenting cells. Peptide- pulsing conditions: 0.5 μΜ and 5 μΜ peptides for 2.5 hours in the presence of the MHC loading enhancer 1-adamantane ethanol (125 μΜ). After peptide pulsing, the D4BL4 cells were washed in PBSA (0.2%). A sample of the D4BL4 was stained with Streptavidin-PE to measure the peptide binding of the cells in the experiment. These were then washed in PBSA (0.2%>) 0.02%> Azide and analysed in flow cytometry (FACS Canto). 55.000 D48T10 cells were stimulated with 40.000 peptide-pulsed D4BL4 cells in a total reaction of 200 μΐ with 5μg/mL Brefeldin-A and incubated for 6.5 hours at 37°C. After incubation, cells were surface stained with PE-labelled anti- human CD3 mAb, fixed in 100 μΐ _^ fixation buffer (PBS 4% Paraformaldehyde) for 10 minutes, washed in 200 μΙ_, PBSA (0.2%) and permeabilized with 30 μΙ_, 0.1 % saponin buffer for 10 min. Cells were stained with APC-labelled anti-human IFNy mAb, and PerCP-Cyanine5.5-labelled anti-human TNFa mAb, and analysed in flow cytometry (FACS Canto). Samples were run in duplicates. Data was analysed with FloJo Software.

The peptide binding to the D4BL4 cells in the experiment is shown in Figure 9. In this experiment the 133 binds 120% compared to L33 (defined as 100%) at both 0.5 μΜ and 5 μΜ concentrations.

T cell activation data for the HP A- la-specific T cell clone D48T10 demonstrate that 133 stimulates HP A- la- specific T cell clones better than L33 (Figure 10) and confirm previous experiments. We measured the co-secretion of IFNy and TNFa upon stimulation. Samples were run in duplicate.