NZ616746B2 - Hypoallergen - Google Patents

Abstract

Disclosed is a recombinant hypoallergenic birch pollen Bet v 1 polypeptide based on a wild type amino acid sequence template as depicted in SEQ ID NO: 3 or of any other Bet v 1 wild type isoform thereof, said polypeptide either naturally or by mutation comprising: (1) at least one first amino acid substitution at position E101, and (2) at least one second amino acid substitution at position N28. Also disclosed is a vaccine composition comprising at least one such hypoallergenic polypeptide and a pharmaceutically acceptable diluent or adjuvant. substitution at position E101, and (2) at least one second amino acid substitution at position N28. Also disclosed is a vaccine composition comprising at least one such hypoallergenic polypeptide and a pharmaceutically acceptable diluent or adjuvant.

Description

PCTIEP2012/057046 HYPOALLERGEN FIELD OF THE INVENTION The present invention relates to mutant polypeptides useful as hy— poallergens. More specifically the present invention relates to specific mutant Bet v 1 polypeptides and to the use of such polypeptides as hypoallergens for desensitizing against birch pollen allergies. rmore, the invention relates to vaccine formulations comprising such polypeptides; to the use of such for- mulations in vaccination; and to methods of vaccination against birch pollen allergy.

’IO BACKGROUND OF THE INVENTION ies are caused by the immune reaction to commonly harmless proteins, allergens. Allergic diseases are ng epidemic proportions all over the world. More than 25% of the population in industrialized countries suf- fer from type I allergy and the number is steadily increasing. Birch pollen aller- gy is a very common form of type I allergy. Bet v 1 is the major allergen of birch pollen. More information on the Bet v ‘I allergen, its ergens and var— iants, is found on the WHO website www.allergen.org.

Type | allergy is based on the formation of immunoglobulin E (lgE) antibodies. The symptoms occur when an allergen molecule binds to two lgE antibodies bound to receptors on a mast cell or basophile surface and induces cross-linking of the lgE-FcaRl complexes. This triggers the degranulation of biological mediators, such as histamine and lipid mediators, which cause inflammatory reactions and symptoms, such as allergic asthma, rhinitis, food and skin allergy, and even anaphylaxis.

The lgE is a large molecule that consists of two identical light and heavy chains. There are five domains in the heavy chain of lgE: VH, C51, 022, (323 and C54. The size of the complete lgE le is about 200 kDa. The crystal ures of the CeZ-Cs4 fragment bound to its chRl or and the C82-CE4 fragment have been determined (Garman et al., Nature 06):259—266, and Wan et al., Nature Immunology, 2002(3):681-686). in the last few years, the three—dimensional structures for a large number of different ens have been determined. Structurally, these aller- gens vary considerably, and no common structural motif that could explain the lity of allergens to cause the production of lgE dies has been iden- tified. However, there are studies implicating that allergenicity is restricted to W0 2012/143374 PCT/EPZO12/057046 only a few protein families, thus raising evidence that structural es of pro— teins could also have a role in allergenicity (Jenkins et a/., J y Clin. Im- munol. 2005(115):163—170; Raudauer et a/., J Allergy Clin Immunol. 2008(121):847—852; en et at, PloS ONE 2010(5):e9037).

The ial question when ng allergenicity involves the so— called B—cell epitope, the lgE antibody-binding site of an allergen. Unfortunate- ly, however, this B-cell epitope cannot be deduced ly from the three— dimensional structure of an allergen. Additionally, there are differences in the epitopes of a defined allergen recognized by individual patient’ lgE. Therefore, 1O B~cell epitopes have been sought using various techniques and various basis, such as by analyzing allergenic fragments or peptides, which react with poly- clonal lgE serum pools from allergic ts, site-directed mutagenesis of al— lergens, use of epitope mimics (mimotopes) and bioinformatics modeling stud— ies. However, as yet, no general maps of dominating epitopes exist for any al— lergen.

With regard to birch pollen, for instance, Holm et al. (The Journal of Immunology 2004 (173): 5258-5267) produced Bet v 1 mutants ning 4 and 9 point mutants with the aim to manipulate surface topology in "selected areas". The paper does not describe how this "selection" has been made. The mutants with four amino acid substitutions represented three different areas on the molecular surface and the mutants with nine amino acid substitutions rep- ed five different areas on the molecular surface. These Bet v 1 mutants had in some cases reduced capacity to bind human serum lgE and to trigger human basophile histamine release. They were also able to induce lgG anti- bodies against unmutated Bet v 1.

Further modified recombinant allergens have been reported: International patent publications WO 02/40676 and WO 03/096869 disclose numerous mutant forms of birch pollen allergen Bet v 1. These mu- tants were ed by introducing random mutations in the putative lgE bind— ing site, based on sequence analysis of conserved surface ures of the Bet v 1 polypeptide. WO 03/096869 discloses the use of four primary muta~ tions on different “small groups” on the allergen surface. international patent publication discloses a Bet v 1 polypeptide comprising three amino acid substitutions or deletions at amino acid sites 54, 115 or 123. There is no evidence that these mutants have re— duced histamine e capacity.

W0 2012/143374 International patent publication ses a Bet v 1 mutant having at least four mutations in the area amino acids 100-125. How- ever, due to the mutations the tree—dimensional structure of the ptide is lost, and there is no reported histamine release activity.

International patent publication discloses a meth- od of blocking the type I surface interaction of allergenic nces by modify- ing amino acid residues on non—continuous allergenic epitopes, i.e., on a pla- nar surface with an area of 600—900 A on the allergenic substance and sug— gests that hypoallergenic birch pollen proteins could be prepared accordingly. 1O Niemi at al., Structure 2007(15): 1413-21, disclose one approach in the search of specific allergen epitopes in the line with the disclosure of Laver et a/., Cell 1990(61):553-556, who state that the only rational method by which to determine the complete epitope of any allergen es measuring crystal structure of an allergen in x with an lgE antibody. Niemi et al. disclose the crystal structure of an lgE Fab fragment in complex with B-Iactoglobulin (BLG). They also show how two lgE/Fab molecules bind the dimeric BLG and that the lgE epitope is ent when compared to known lgG epitope struc- tures, being a “flat” e located in the [3 sheet region.

Rouvinen et 3]., Pics ONE 2010(5):e9037 investigated the role of dimeric structures of allergens using bioinformatics methods combined with native mass ometry. (Electrospray ionization Fourier Transform Ion Cy— clotron Mass Spectrometry, ESl FT—lCR—MS). The ESI-MS ements of the 55 known crystal structures of allergens showed that 80% of them exist in symmetric dimers or oligomers in crystals and that the majority is transient di— mers that are formed at high n concentrations. The possible relationship between dimeric structure and allergenicity was studied with a recombinant al— lergen from cow’s milk, B-Iactoglobulin (rBos d 5 B), which occurs as a dimer, and its mutant H146P, which occurs mainly as a monomer. A somewhat re- duced histamine releasing capacity was observed with the ric rBos d 5 B mutant H146P when compared to the native 808 d 5 B and recombinant 805 d 5 B. Although the authors conclude that dimerization could be a very com- mon and essential feature for allergens and generally t that the prepa- ration of purely monomeric variants of allergens could open up novel possibili— ties for specific immunotherapy, the ultimate role of structural features in aller— genicity remain n. From the in vitro crystal analysis direct conclusions as to how the allergen molecules behave in a human body, i.e. in vivo, cannot be drawn.

Today the trend in the treatment of all allergic symptoms is s an active induction of tolerance using allergen-specific desensitization instead of avoiding the al- lergen, which is often not possible, or merely treating the symptoms. Current desensiti- zation y is based on allergens purified from natural sources, wherein batch to batch variations may lead to problems related to finding and maintaining the right dos- age and efficiency of the treatment. These problems may lead to a ial risk of ana- phylactic side effects and sensitization to new allergens. 1O The use of recombinant allergens for desensitizing would remove the disad- vantages related to batch to batch variations, and the first recombinant allergens are in al trials (Valenta et a/., Annu Rev l 2010(28):211-41). The efficiency of such allergens in the clinic thus remains to be seen.

There is a recognized and large need for safe and efficient vaccines and therapy products to meet the increasing medical problem of allergy. At present the mar- ket for safe and efficient therapies of allergy is underdeveloped.

BRIEF DESCRIPTION OF THE INVENTION The present ion provides a recombinant hypoallergenic birch pollen Bet v 1 polypeptide based on a wild type amino acid sequence template as depicted in SEQ ID NO: 3 or of any other Bet v 1 wild type isoform thereof, said polypeptide either naturally or by mutation comprising (1) at least one first amino acid substitution at posi- tion E101, and (2) at least one second amino acid substitution at position N28.

The present invention generally relates to a recombinant hypoallergenic birch pollen Bet v 1 ptide based on a wild type amino acid sequence template as ed in SEQ ID NO: 3 or of any other Bet v 1 wild type isoform thereof, said poly- e either naturally or by mutation comprising (1) at least one first amino acid substitution at a position selected from the group consisting of amino acid residues E101, K80, N82, 884, 899, S117, and K119, (2) at least one second amino acid substitution at a position selected from the group ting of amino acid residues N28, D25, N43, G46, N47, and Y158.

The present invention r relates to a hypoallergenic polypeptide defined above for use as a vaccine for desensitizing against birch pollen.

The present invention further relates to a e composition comprising at least one hypoallergenic polypeptide defined above and at least one pharmaceutically acceptable adjuvant. In one embodiment of the present invention said vaccine is for sublingual administration.

W0 2012/143374 The present invention further relates to a method of vaccinating against birch pollen allergy, said method comprising administering to a subject suffering from birch pollen allergy a hypoallergenic polypeptide or a vaccine composition d above in an amount ive for desensitizing and for in— ducing the production of protective antibodies against birch pollen.

The hypoallergenic ptides according to the present invention have a histamine release capacity which is at least 20x reduced when com— pared to the histamine e capacity of the Bat v 1 wild type. In one embod- iment the polypeptides have a histamine release capacity which is reduced at 1O least 100x.

BRIEF DESCRIPTION OF THE FIGURES in the following the ion will be described in greater detail by means of preferred embodiments with nce to the attached drawings, in which Figure 1 illustrates the putative lgE epitope residues of Bet v1 (resi- dues in red) and putative residues involved in the dimerisation of Bet v 1 (resi- dues in violet) derived in Example 1; Figure 2 is a schematic presentation of the bacterial sion units for production of recombinant allergens, wherein Ptac is a promoter, PelB is the signal sequence linked to the coding region of recombinant allergens and the stars illustrate the amino acid substitution sites; Figure 3 is an amino acid sequence alignment of 36 isoforms of Bet v 1; Figure 4 shows the nucleic acid sequences of the Bet v 1 wild type polypeptide (A, SEQ lD N01) and the N28K—E101K polypeptide (B, SEQ lD NO:2) used in Example 2; Figure 5 shows the competitive inhibition of serum lgE binding to Bet v 1 with inant Bet v 1 and Bet v 1 N28K—E101K polypeptides; and Figure 6 shows the results of histamine release experiments with inant Bet v 1 and Bet v 1 N28K—E101K polypeptides.

Figure 7 shows the native ESl FT—lCR mass spectra of the recombi— nant Bet v 1 wild type and recombinant Bet v 1 mutant N28K—E101K at con— centration of 3 uM.

W0 2012/143374 PCTmP2012/057046 ED DESCRIPTION OF THE INVENTION In the ing ption, examples and claims both three-letter and one—letter codes are used for amino acids. See, for instance, IUPAC-IUB Joint Commission on Biochemical Nomenclature. Nomenclature and Symbol- ism for Amino Acids and Peptides. Eur. J. m. 138:9—37(1984).

The nation of amino acid sites in the polypeptides according to the present invention are exemplified as follows: N28 means that there is an asparagine residue at position 28, whereas N28K means that the asparagine residue at position 28 has been replaced by a lysine residue. Correspondingly, 1O E101 means that there is a glutamic acid residue at position 101, whereas E101 K means that the glutamic acid residue at position 101 has been replaced by a lysine residue, etc.

Birch pollen allergy is a very common form of allergy and pollen of the white birch (Bez‘u/a verrucosa) is one of the main causes of Type I allergy reactions in Europe and North a. It is estimated that about 10—15% of the population may suffer from birch pollen allergy. Furthermore, other aller— gens, such as apple allergens, cross-react with birch pollen specific lgE caus- ing allergic reactions even when the subject is not subjected to pollen.

Bet v 1 is the major allergen of birch pollen and it is responsible for the lgE g in more than 95% of birch pollen allergic subjects. Bet v 1 is a protein having a molecular weight of 17 kD. The amino acid ce of wild type Bet v 1 is given in SEQ ID NO: 3. The WHO allergen website (wwwallergenorg) lists thirty-six (36) isoforms of Bet v 1, which have been se— quence aligned in Figure 3. The alignment shows that Bet v 1 is highly con- served. The template of the m used as a wild—type Bet v 1 in the present invention is from isoform Bet v 1a (Bet v 1.0101), but any one of these iso~ allergens may be used, as appropriate, to provide a hypoallergenic polypeptide variant according to the present invention.

The amino acid ces of all 36 Bet v 1 isoforms are disclosed in the sequence listing, as follows: 1.0101 (SEQ ID NO: 3), 1.0102 (SEQ ID NO: 4), 1.0103 (SEQ ID N025), 1.2501 (SEQ ID N016), 1.1501 (SEQ ID NO: 7), 1.1502 (SEQ ID NO: 8), 1.2801 (SEQ ID NO: 9), 1.3001 (SEQ ID NO: 1 0) I 1.2901 (SEQ ID NO: 11), 1.2301 (SEQ ID NO: 12), 1.0501 (SEQ ID NO: 1 3), 1.0601 (SEQ ID NO: 14), 1.0602 (SEQ ID NO: 15), 1.0801 (SEQ ID NO: 1 07 ) 1.1701 (SEQ ID NO: 17), 1.0401 (SEQ ID NO: 18), 1.0402 (SEQ ID NO: 1 9) 1.0701 (SEQ ID NO: 20), 1.1001 (SEQ ID NO: 21), 1.2401 (SEQ ID NO: 22) W0 2012/143374 ZO12/057046 1.2601 1.0201 SEQ ID NO: 26), 1.0901 (SEQ ID NO: 27), 1.0301 (SEQ ID NO: 28), 1.1401 AAAA SEQ ID NO: 23), 1.2701 (SEQ ID NO: 24), 1.2201 (SEQ ID NO: 25), SEQ ID NO: 29), 1.1402 (SEQ ID NO: 30), 1.1901 (SEQ ID NO: 31), 1.2001 SEQ ID NO: 32), 1.1801 (SEQ ID NO: 33), 1.1101 (SEQ ID NO: 34), 1.1201 (SEQ ID NO: 35), 1.1601 (SEQ ID NO: 36), 1.2101 (SEQ ID NO: 37), and 1.1301 (SEQ ID NO: 38), respectively.

The isoforms of Bet v 1 include variants which have different aller— genic potential. The isoforms of Bet v 1 are at least 94% identical to Bet v 1 wild type amino acid sequence of SEQ ID NO: 3. For instance, isoforms Bet v 1O 1.0401 with 96% amino acid e identity and Bet v 1.1001 with 94% resi- due identity to Bet v 1.0101 have been fied as natural hypoallergens, be— cause they were poor inducers of a mediator release. As compared to Bet v 1.0101, Bet v 1.1001 contains, e.g., the N28K mutation. They have also been regarded to fulfill the criteria to represent excellent vaccine candidates. (Wag~ ner et a/., J. Allergy Clin Immunol 2008;121:725—735). However, no scientific, experimental or clinical data is available.

The present invention provides mutated hypoallergenic polypeptide variants of Bet v 1, which are useful as es for immunizing subjects in need thereof and thus preventing and/or ating allergy and desensitizing subjects suffering from allergy against birch pollen.

The recombinant birch pollen Bet v 1 polypeptides according to the present ion have a wild type amino acid sequence, but they contain mu- tations at selected amino acid positions to reduce or fully diminish their ability to induce the production of lgE but to retain their capacity to induce the produc— tion of protective lgG dies, i.e., they are hypoallergenic.

Specifically, the present invention relates to a recombinant hypoal— lergenic birch pollen Bet v 1 polypeptide based on a wild type amino acid se~ quence template as depicted in SEQ ID NO: 3 or of any other Bet v 1 wild type isoform thereof, said polypeptide either naturally or by mutation comprising (1) at least one first amino acid tution at a position selected from the group consisting of amino acid residues E101, K80, N82, 884, 899, S117, and K119, and (2) at least one second amino acid substitution at a position select- ed from the group consisting of amino acid residues N28, D25, N43, G46, N47, and Y158.

In one embodiment, the polypeptide of the present invention has the template of a wild type amino acid sequence as depicted in SEQ ID NO: 3. In another embodiment, the polypeptide of the present invention has the template of a wild type amino acid sequence selected from the group consisting of SEQ ID NO: 4—38.

In one embodiment of the invention, the recombinant hypoallergenic birch pollen Bet v 1 polypeptide is ented by the amino acid sequence depicted in SEQ ID NO:39. SEQ ID NO: 39 ses Bet v 1 polypeptide se- quence with ons for amino acid substitutions (amino acid positions 25, 28, 43, 46, 47, 80, 82, 84, 99, 101, 117, 119, 158) and also discloses wild type amino acids at these positions.

Any combination of first and second substitutions at any ed positions results in a hypoallergenic polypeptide. In one embodiment of the in- vention, the polypeptide ses an amino acid sequence selected from the group consisting of SEQ ID NO: 41—47, and at least one second amino acid substitution at a position ed from the group consisting of amino acid resi— dues N28, D25, N43, G46, N47, and Y158, or isoform thereof. Preferably, the polypeptide has an amino acid sequence selected from the group consisting of SEQ ID NO: 41—47, and at least one second amino acid substitution at a posi— tion selected from the group consisting of amino acid residues N28, D25, N43, G46, N47, and Y158, or isoform thereof. In another ment of the inven— tion, the polypeptide comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 41, wherein amino acid 101 is K, SEQ ID NO: 42, wherein amino acid 80 is Y, SEQ ID NO: 43, wherein amino acid 82 is K, SEQ ID NO: 44, wherein amino acid 84 is K, SEQ ID NO: 45, wherein amino acid 99 is K, SEQ ID NO: 46, wherein amino acid 117 is K, and SEQ ID NO: 47, wherein amino acid 119 is E; and at least one second amino acid substitu- tion at a position ed from the group consisting of amino acid residues N28, D25, N43, G46, N47, and Y158. Preferably the polypeptide has an amino acid ce selected from the group consisting of SEQ ID NO: 41, wherein amino acid 101 is K, SEQ ID NO: 42, wherein amino acid 80 is Y, SEQ ID NO: 43, n amino acid 82 is K, SEQ ID NO: 44, wherein amino acid 84 is K, SEQ ID NO: 45, wherein amino acid 99 is K, SEQ ID NO: 46, wherein amino acid 117 is K, and SEQ ID NO: 47, wherein amino acid 119 is E; and at least one second amino acid substitution at a position selected from the group con- sisting of amino acid residues N28, D25, N43, G46, N47, and Y158. SEQ ID W0 43374 PCTfEP2012/057046 NO: 41 discloses Bet v 1 polypeptide sequence, n amino acid at position 101 is a substitution and not wild type amino acid E. rly SEQ ID NOs: 42- 47 show Bet v 1 polypeptides, wherein amino acids at positions 80, 82, 84, 99, 117 and 119, respectively, are substitutions and not wild type amino acids.

In one embodiment of the invention, said at least one first amino ac— id substitution is at a position selected from the group consisting of amino acid es E101 and 899, and said at least one second amino acid substitution is at a position selected from the group consisting of amino acid substitutions N28 and D25. 1O In yet another preferred embodiment of the invention said at least one first amino acid tution is at position E101, and said at least one se- cond amino acid substitution is at position N28. In one preferred embodiment of the invention, the recombinant hypoallergenic birch pollen Bet v 1 polypep- tide is represented by the amino acid sequence ed in SEQ ID NO:40.

SEQ ID NO: 40 shows Bet v 1 polypeptide sequence with substitutions at posi— tions 28 and 101, amino acids at these positions being lysines (K).

The hypoallergenic polypeptide variants of the present invention comprise also variants, which contain more than one first and second tu- tion. The terms “at least one first substitution” and “at least one second substi- tution” mean that said first and second substitutions may each se one, two, three or four or more substitutions, and any combinations of first and se- cond substitutions are possible, as long as the ability of the hypoallergenic Bet v 1 polypeptide to induce protective antibody response retains. Thus in one preferred embodiment of the invention the first amino acid substitution com— prises substitution at position E101, and said second amino acid substitutions are at ons N28 and D25 or at positions N28, D25 and Y158. ln another preferred embodiment of the invention first amino acid substitution comprises substitutions at positions E101 and S99 and said second amino acid substitu- tion is at position N28. In yet another preferred embodiment of the invention said first amino acid substitutions are at positions E101 and 899, and said se- cond amino acid substitutions are at positions N28 and D25.

In a further preferred embodiment of the ion, the substitutions of Bet v1 are at least E101 and N28, E101 and D25, E101 and N43, E101 and G46, E101 and N47, E101 and Y158, K80 and N28, K80 and D25, K80 and N43, K80 and G46, K80 and N47, K80 and Y158, N82 and N28, N82 and D25, N82 and N43, N82 and G46, N82 and N47, N82 and Y158, 884 and N28, S84 2012/057046 and 025, S84 and N43, S84 and G46, 884 and N47, S84 and Y158, S99 and N28, 899 and D25, S99 and N43, S99 and G46, S99 and N47, 899 and Y158, S117 and N28, 8117 and D25, S117 and N43, 8117 and G46, 8117 and N47, 8117 and Y158, K119 and N28, K119 and D25, K119 and N43, K119 and G46, K119 and N47 or K119 and Y158. in a further preferred embodiment of the invention, the substitutions are one of these combinations. ln one preferred embodiment of the invention, there are altogether at least two, three or four amino acid substitutions. Preferably, the polypeptide of the invention has two, three, four, five, six, seven, eight, nine or ten substitu- tions. More preferably, the polypeptide has two, three or four substitutions.

One of the advantages of the invention is that only a small number of substitu— tions (at least two) are needed for the desired effects.

The polypeptides according to the present invention are hypoaller- genic, and t a histamine release ty which is at least 20x, preferably 100x, reduced when compared to the histamine release capacity of the corre— sponding unmutated Bet v 1 wild type.

The hypoallergenic ptides according to the present invention are useful as vaccines against allergy, especially birch pollen allergy. Vaccines comprising ptides according to the t invention are formulated ac— cording to standard pharmaceutical procedures known to skilled persons in the art. Vaccines according to the t invention are especially suited for sub— lingual administration. lergenic variants according to the present invention are ob- tained by mutating chosen specific amino acid residues, e.g., residues with bulky side chains, located on the epitope surfaces of Bet v 1. The selected amino acid residues are those, whose side chains point outside towards the solvent. Mutating such residues cause minimal change to the basic 3— dimensional structure of the allergen. Preferably, however, the nesis modifies the surface of the epitope to such an extent that the binding and 3O cross-linking of lgE antibodies on the mast cell surface is prevented or ly reduced, while the over-all structure of the variant is still very similar to that of the wild type allergen. Such a on favors the induction of lgG and other protective antibodies, having the ability of binding both to the wild-type allergen and to the mutated t allergen. The effect of the mutation is determined as a lower affinity of the allergen specific lgE antibody towards the modified Bet v 1 allergen. Preferably the mutation decreases the affinity of the specific lgE an— W0 2012i143374 tibody at least tenfold, preferably at least 20-fold, and more preferably 20- to 100-fold, and most preferably more than 100-fold. The resulting modified Bet v 1 allergen can be used to evoke tolerance against birch pollen in allergic pa- tients.

The hypoallergenic variant polypeptides according to the present in- vention, useful in allergen—specific desensitization, possess two features: 1) the ability to strongly reduce an lgE—mediated reaction; and 2) a retained ype 3D g, and thus the capability of inducing the production of lgG~antibodies capable to bind wild type allergen. 1O The knowledge of the structure of the lgE g epitope would greatly simplify the design of lergenic variants, as mentioned above.

However, the structure of Bet v 1 complexed with IgE antibody is unfortunately not ble. It is not known how much there are differences in the epitopes of a defined en recognized by individual patient’ lgE. The use of peptides in the epitope ng is also unreliable and actually useful only when scanning linear epitopes (Niemi et a/., Structure 2007(15): 1413—21). The conformation as well as the physical properties, e.g., solubility, of a single peptide may differ markedly from those of ponding portion of a ptide chain forming part of a native protein structure. Therefore, the design of the mutant Bet v 1 allergens was based on molecular surface analysis using molecular graphics programs, such as PyMOL, to elucidate the structure of the e and to test ial hits by preparing and testing the mutants. In addition, the ability of dimerization of Bet v 1 was taken into account in the design.

The crystal structure of Bet v 1 (protein data bank code 1BV1) was used to define the quaternary structure of Bet v 1. The PDBePlSA internet server was used for creating coordinates for the symmetric dimer of Bet v 1. It has been estimated that the distance between lgE antibodies in the cluster on the mast cell e is about 5 nm (Knol, EF; Mol.Nutr.Res. 50(2006):620). By ng the molecular surface of the Bet v 1 dimer around the two-fold sym- metry axis within a distance of 2.5 nm from the symmetry axis, two putative epitopes were identified on the molecular surface of Bet v 1 (Fig. 1).

The putative epitope is composed of amino acid residues V2—E6; R70-D75; N78-884; E96-K103; and K115-H121, whereas r-monomer interface of Bet v 1 dimer is composed of amino acid residues K20-K32; 839— P50; V74—H76; and D156—N159.

W0 20 12/143374 These putative epitopes were carefully ed, in order to identify amino acid residues, which could serve as mutation points, Preferred muta— tions points should have the ability to decrease the binding of the allergen to lgE antibodies but still maintain the three—dimensional structure of the wild-type allergen. The ve e includes amino acid residues K80, N82, S84, S99, E101, S117, and K119. The monomer—monomer interface es resi— dues D25, N28, N43, G46, N47, and Y158. The residues E101 and N28 were considered as the most interesting mutation points, as they are located in the center of the putative epitope and in a strategic on on the monomer- 1O monomer interface, respectively. In the present invention the first amino acid substitution(s) is(are) in the e area and the second amino acid substitu~ tion(s) is(are) in the monomer—monomer—interface.

These two residues are highly conserved in all 36 isoforms of Bet v 1. No natural variation exists in residue 101. The natural variations are found in residue 28, which exists as asparagine (in 31 isoforms), lysine (in 2 isoforms) or tyrosine (in 3 isoforms). Residues D25, N43, G46, N47, K80, S84, K119, and Y158 are conserved, whereas residues N82, S99, and 8117 vary ly, as shown in Figure 3.

The next step was to select an appropriate mutation for each resi- due. As an example, 899 is a small hydrophilic and neutral amino acid residue.

The mutation which would interfere lgE g would thus be of "opposite" na- ture, i.e., large and/or charged, for example Ser to Lys, Arg, Asp, Tyr, and Val.

Similarly, E101 and N28 can be replaced by a residue with the opposite charge (Lys, Arg) or with a hydrophobic residue (Tyr, lie, or Trp). Substitutions at resi— dues D25, N48, G46, N47, K80, N82, 884, S117, and K119 could be designed correspondingly. Table 1 lists potential substitutions, which would yield Bet v 1 hypoallergenic mutants according to the present invention set forth in SEQ ID NO: 39.

PC17EP2012/057046 Table 1. Mutants of Bet v 1 wt EPITOPE MUTANTS Bet v 1 wt mutant 1 mutant 3 mutant 5 899 small hy- K large charged R large d Y large hydro- V large hydro- drophilic phobic phobic E101 charged K large opposite- R large opposite~ Y large hydro- l large hydro- Wlarge hydro- ly charged ly charged phobic phobic phobic K80 charged Y large hydro- E oppositely W large hydro- | large hydro- L large hydro- phobic charged phobic phobic phobic 884 small hy- K large d R large charged D charged E charged Y large hydro- lic phobic N82 hydrophilic K large charged R large charged Y large hydro- E charged L hydroph obic phobic 8117 small K large charged R large charged D charged Y large hydro- L large hydro- hydrophilic phobic phobic K119 charged E oppositely Y hydrophobic ‘l L hydrophobic W hobic | hydrophobic charged Segments: V2—E6; R70-D75; N78-S84; 03; K115-H121 MONOMER MUTANTS Bet v 1 wt 1 mutant 1 mutant 2 mutant 3 mutant 4 mutant 5 025 d K large opposite- R large opposite— Y large hydro— H large charged L large hydro— iy charged ly charged phobic phobic N28 hydrophilic K large charged R large charged Y large hydro- | large hydro- W large hydro~ phobic phobic hobic N43 hydrophilic Y large hydro- H large charged | large hydro- L large hydro- phobic phobic phobic G46 no side P main chain V hydrophobic D charged T hydrophilic L hydrophobic chain N47 hydrophilic E large charged L hyd rophobic l hydrophobic Y large hydro- P main chain ahobic Y158 large hy- D charged E charged L hydrophobic P main chain l hydrophobic drophobic The modified Bet v 1 hypoallergens ing to the present inven- tion are useful as vaccines. Conventional allergy vaccination is typically carried out as multiple subcutaneous immunizations over an extended time period, e.g., one to two years. In order to minimize the risk of anaphylactic reactions, the immunization scheme is applied in two phases, an initial up—dosing phase and a maintenance phase. The up~dosing phase starts with minute doses, which are then slowly increased, typically over a 16—week period until the maintenance dose is d. The maintenance phase typically comprises in- jections every sixth week. Such a vaccination regime is s for the patient, requiring a long-term commitment. Moreover, it puts high impact on the stabil- ity of the vaccine, in terms of safety and reproducibility. The patients need to be strictly monitored, often hospitalized, after each injection.

As the histamine release capacity of the hypoallergens according to the present invention is substantially reduced, the dosing—up phase could be significantly shorter than that of a conventional allergy ation, or at best no dosing-up scheme could be needed. Modified, recombinant hypoallergens ing to the present ion do not present any batch—to—batch variation.

Thus, close monitoring of the dose—response and possible eactions is not Thus, the present invention further relates to a use of a hypoaller- genic Bet v 1 polypeptide described in detail above as a vaccine and to a vac— cine composition comprising at least one recombinant hypoallergenic Bet v 1 polypeptide of the invention and at least one pharmaceutically acceptable dilu- ent or adjuvant, such as saline, buffer, aluminum hydroxide and like. The pre- sent invention further relates to a method of ating against birch pollen allergy, said method comprising administering to a t ing from birch pollen allergy a hypoallergenic polypeptide or a vaccine composition as de— fined above in an amount and using a vaccination schedule effective for induc— ing the production of protective antibodies against birch pollen.

A “subject” of vaccination is a human (adult, child or adolescent) or an animal. Preferably, the animal is any ic animal such as a dog, cat, horse, cow, sheep or pig.

For instance, a hypoallergen according to the present invention is formulated as conventional e formulations, such as aluminum hydroxide— adsorbed vaccines, using methods well known in the art rberger et al., PNAS, 101(2):14677-82, 2004). Alternatively and ably, however, the hy— poallergens according to the present invention may be administered by other suitable vaccination routes and schemes, such as oromucosal or sublingual W0 20 12/143374 administration, using methods and formulations known in the art. See, e.g., Eu- ropean Patent publication EP 1812059.

The modified Bet v 1 hypoallergens could be used in concentrations of, e.g., 0.5 pg/ml, 5 pg/ml or 50 pg/ml. Exemplary doses may vary between 0.05 pg and 2pg during a possible dosing—up phase, and between 3—15 pg during the maintenance phase, preferably 5—15 pg, most preferably about 10 pg, depending on the severity of the allergy, the age and medical history of the patient. A suitable dose is easily decided by a clinician familiar with treating and preventing allergy. 1O International patent publication WOO4/O47794 discloses a solid fast- dispersing dosage form for sublingual stration of an allergy vaccine, and US patent application 2009/0297564 ses a liquid vaccine formulation for oromucosal administration.

The ed Bet v ”i hypoallergens according to the present inven- tion are particularly suitable for sublingual administration using sublingual drops. For this purpose the hypoallergenic ptides are provided in saline.

A safe and effective dose range for administration of the polypeptides, as well as the dosing regimen capable of eliciting a desired immune response is de- termined during clinical development of the e candidates according to the present invention, using methods and schemes known in the art.

A maximum tolerated single dose of a hypoallergen according to the present invention is determined in a study in allergic male and female subjects, which are exposed to increasing sublingual doses. When the maximal tolerat— ed dose of predefined dose is reached, the study is adapted to a dose g study with daily dosing, where the dose levels differ by a factor of 2 to 4. The initial dose is in the range of 10-400 pg, and the study provides the maximal tolerated sublingual dose, which may be as high as 20 mg.

Thereafter dose escalation and dose ranging over a wide dose range administered daily or weekly are studied. The safety of the vaccination dose range is inary tested with a Skin Prick Test prior to administering multiple doses. These studies provide ily immunological ters, and arily, eventual efficacy after challenge by birch pollen.

The lergenic polypeptide vaccines according to the present invention should elicit a T—cell response detectable as a shift from TH2- to THt-type. Production of lgG antibodies should be demonstrable before enter- ing enic challenge testing.

WO 43374 Finally, a study in allergic ts is performed, as a double blind, randomized placebo controlled itization study in allergic male and fe— male subjects exposed to a number of sublingual doses during 3-6 months, with a follow up for 12 months initially. The subjects will be challenged by aller— gen prior to the start of the study as well as every six months thereafter in a double blind manner.

The study will show a statistically and clinically significant difference between the groups receiving placebo and a hypoallergen vaccine according to the present invention, when they are challenged to the native allergen. 1O EXAMPLES The following examples are given to further illustrate embodiments of the present invention, but are not intended to limit the scope of the inven— tion. It will be obvious to a person skilled in the art, as technology advances, that the inventive concept can be implemented in various ways. The invention and its embodiments are thus not limited to the es described herein, but may vary within the scope of the claims.

Example 1. Design of the Betv 1 ons The goal in the hypoallergen design is to achieve a mutant allergen whose y to bind and link lgE—antibodies on the mast—cell surface is strongly d but which still ins a very similar structure as the wild type allergen. This would favor the induction of lgG and other antibodies which would have ability to bind both to ype allergen and mutant allergen.

The knowledge of the lgE epitope would greatly simplify design.

However, there is no structure of Bet v 1 complexed with the lgE antibody available. The use of peptides in the epitope scanning is also unreliable (Niemi et al., Structure (15):1413~21, 2007). The only method to suggest an epitope is to study the molecular surface of Bet v 1 en and test the possible hit by preparing mutants. Firstly, we identified a putative epitope (Fig. 1) on the mo- r surface of Bet v 1. Secondly, we selected such residues on this ve epitope which as mutated would maintain a three-dimensional structure similar to the wild-type allergen and still have the ability to decrease binding to lgE an- tibodies. The putative epitope includes amino acid residues K80, N82, 884, 899, E101, S117, and K119.

The third step is to select mutation for each residue. As an example, 899 is a small hydrophilic and neutral residue, serine. The mutation which W0 43374 PCTIEP2012/057046 would interfere lgE binding would thus be "opposite", i.e., large and/or charged, for example 899K (serlne to lysine), 899R (serine to arginine), S99D e to ic acid), SQQY (serine to ne), and 899V (serine to ). in the case of E101, mutations interfering with the lgE binding could include using residues with an opposite charge (Lys, Arg) or using hydrophobic resi- dues (Tyr, lle, Trp).

Finally, to introduce the feature of prohibiting dimerisation into the polypeptide variants of the invention monomer a on design, based on crystal structure of Bet v 1 (PDB code 1BV1), was used. The model for Bet v 1 1O dimer was created with the PlSA . The monomer—monomer interface was studied by using molecular graphics program. The mutants were designed on this interface using the same principle as used in the epitope mutants. As an example, N28 on the monomer~monomer interface is a medium—sized hy— drophilic residue. The mutation, which interferes the dimer ion, would thus be a very large charged (lysine, arginine) or hydrophobic residue (tyro- sine, isoleusine, tryptophane).

Example 2. Cloning of the recombinant Bet v 1 molecules To produce the wild type (wt) and the mutant of the recombinant Bet v 1 molecules (rBet v 1) the cDNAs encoding these particular proteins were cloned into a bacterial expression plasmid (Fig. 2). First, the rBet v 1 cDNAs designed in Example 1, with the codon optimization for ichia coli pro— duction in vector pUC57 (wt and N28K-E101K) were ordered from GenScrlpt Corporation (USA). The cDNAs contained Ncol restriction site at the send and Hindlll at the 3'end. The cDNAs were cloned as indlll fragments into bacterial expression vector pKKtac encoding the Ervinia carotovora’s pectate lyase (pelB) signal sequence (Takkinen et al., Protein Eng. (4): 837-841, 1991) and sion plasmids were transformed into E. coli XL~1 Blue strain. The DNA ces of the rBet v 1 and the N28K-E101K mutant were verified by DNA sequencing (ABI 3100 Genetic Analyzer, Applied Biosystems), and are herein depicted as SEQ ID N025 1—2.

Example 3. Production of the recombinant Bet v 1 molecules The sion vector of wild type rBet v 1 was transformed into E. co/i BL21 DE strain and the expression vector Bet v 1 N28K-E101 K mutant into E. coli RV308 strain for bacterial expression. Single colonies of rBet v 1 wt and Bet v 1 N28K-E101K mutant were inoculated into 5 ml SB or TB, 100pg/ml W0 2012(143374 PCTfEP2012/057046 ampicillin and 1% glucose, respectively, and ated for 16 h at +37 °C with 220 rpm shaking. Cultivations were 1:50 d into 3 x 300 ml SB or TB with 100pg/ml ampicillin and cultivated at +37°C until the ODGOO reached 4. Protein expression was d by the on of lPTG to a final concentration of 1mM, and the cells were cultivated for 16 h at RT with 170 rpm shaking. Cells were harvested by centrifugation for 15 min at 5000g at +4°C, and the peri— plasmic fraction of the cells was isolated by an osmosis—shock method de— scribed by Boer et al. (Protein Expression & Purification, 1): 216-226).

The cell pellet equivalent of 900 ml of the culture was re—suspended in 300ml, 1O 30mM Tris/HCI, 20 % sucrose, pH 8.0, and 1mM EDTA, and incubated for 20 min under shaking on ice. The suspension was centrifuged for 20min at 8000g at 4°C. After this the pellet was re-suspended in 75ml of ice—cold 5mM MgSO4 and shaken for 20min at 4°C on ice, and the osmotic shock fluid was harvest~ ed by centrifugation at 8000g for 20 min at 4°C.

Example 4. Purification of the rBet v 1 molecules asmic fractions of the wild type rBet v 1 and the N28K-E101K mutant were supplemented with 1M NaCl and the first chromatographic purifi- cation step for both expressed ns was carried out by a phenyl—Sepharose column (GE Healthcare) with 20mM NaH2P04, 1M NaCl, pH 5.0 buffer using the flow rate 2 ml/min. The elution was performed with a linear gradient of 20mM Tris—HCl, pH 9.3, with 7.5% isopropanol. Fractions containing the re- combinant Bet v 1 wild type or N28K—E101K mutant were pooled and concen— trated. The wild type Bet v 1 was further purified by a Bio-Gel P60 size exclu- sion chromatography column with a bed height of 460 mm and 1x PBS buffer with a flow rate of 0.3 ml/min. In the case of the rBet v 1 N28K—E101K mutant an additional amylose resin tography step was required to get rid of Eco/i maltose g protein contamination before the size exclusion chroma— tography.

The protein concentration of pooled rBet v 1 ons was deter— mined at 280 nm.

Example 5. Analysis of rBet v 1 and the N28K-E101K mutant by Mass Spectrometry Mass-spectrometric experiments were performed with a 4.7 T Bruker BioAPEX—ll ESl FT—lCR mass spectrometer (Bruker Daltonics, Billerica, Massachusetts, USA) equipped with a conventional ESI source (Apollo-llTM).

W0 2012/143374 Native mass spectra: desalted allergen samples at concentration of 3 uM in 10 mM ammonium acetate buffer (pH 6.9) were directly infused at a flow rate of 1.5 mL/min with dry nitrogen serving as the drying , 6 mbar) and nebu— lizing gas. All instrumental parameters were zed to in non- covalent interactions in the gas-phase and to maximize ion transmission at m/z 2000—3000. The same instrumental parameter settings were employed through— out to avoid any bias between different samples. lly, 500—1000 co— added 128—kword time—domain transients were recorded and processed to ord data prior to fast Fourier transform and magnitude calculation. Mass 1O calibration was done externally with respect to the ions of an ES Tuning Mix (Agilent Technologies, Santa Clara, CA, USA). Denaturated spectra were typi— cally measured in acetonitrile/water/acetic acid solution. All data were acquired and sed with the use of Bruker XMASS 7.0.8 software. The native ESl FT—ICR mass spectra shows that the recombinant Bet v 1 mutant N28K—E101 K folds similarly as the recombinant Bet v 1 wild type (Figure 7).

Example 6. Inhibition of serum lgE binding to recombinant Bet v 1 and Bet v 1 N28K-E101K polypeptides analyzed by a competitive ELISA The g of an lgE serum sample ofa birch pollen allergic person (E8) to biotinylated rBet v 1 immobilized on streptavidin wells was inhibited by 2O sing s of the rBet v 1 and rBet v 1 N28K-E101K mutants. First, commercially available rBet v 1 (wild type, Biomay) was biotinylated using Sul- fo-NHS—LC—biotin (Pierce) according to manufacturer’s protocol. The biotinylat- ed rBet v 1 (0.5 pg/well) was immobilized onto streptavidin (SA) wells (Roche Diagnostics Gmbh) ed by a washing step and the addition of E3 serum (1:6 dilution). After a 2-hour incubation at RT in a shaker and a washing step different amounts (4, 1, 0.25, 0.0625, 0.0156, and 0.0039 pg) of rBet v 1 were added and incubated for 2 h at RT in a shaker. After a washing step, the detec- tion of bound lgE molecules was performed using a 1:1000 dilution of an AFOS—conjugated anti—human lgE antibody (Southern Biotech Associates inc.) with tion for 1 h at RT in a shaker. Finally the ate solution, p— nitrophenylphosphate ), was added and the absorbance values at 405 nm were measured (Varioscan, Thermo Electron Corporation).

The result of serum lgE binding to r Bet v 1 polypeptides analysed by a competitive ELISA is shown in Fig. 5. The rBet v 1 polypeptides, wt and N28-E101 K mutant, were used for competing the binding of serum of a Bet v 1 allergic person (E3) to immobilized rBet v 1 (Biomay). Both rBet v 1 wild type W0 2012(143374 molecules (a commercial one from Biomay and own product) inhibited the lgE binding to the lized Bet v 1. The Bet v 1 mutant N28K—E101K showed reduced inhibition when compared to the rBet v 1 controls, ting that the mutations E101 K and N28K affect the lgE epitope and dimerization of Bet v 1. e 7. Histamine Release Assay The biological activity of the ed inant Bet v 1 polypep— tides was ed by the method of passive ization of stripped baso— phils and a subsequent challenge with the allergen molecules. The histamine release assay was performed as an outsourced e at RefLab ApS, Co- ‘ penhagen, Denmark, having an accredited histamine release assay method.

The induction of the in vitro release of histamine from basophilic leukocytes by a commercial recombinant Bet v 1 (Biomay, Austria) and the two recombinant Bet v 1 proteins, wt and N28K—E101K, was measured. Each of the three aller- gens was tested in the passive transfer test as a dose response study with the concentration range of: 20—006 ng/ in duplicates with the serum of a Bet v 1 allergic person (E3).

The result of the histamine release assay is shown in Figure 6. The rBet v 1 N28k—E101 K was 100 times less biologically active compared to the reference, commercial recombinant Bet v 1 (Biomay), and rBet v 1 wild type It will be obvious to a person skilled in the art that, as the technology es, the inventive concept can be implemented in various ways. The in— vention and its embodiments are not limited to the examples described above but may vary within the scope of the claims.

Example 8. Skin prick test Skin prick tests (SPTs) with three voluntaries with two diagnosed birch pollen allergy and with one non-atopic person were performed with recombinant Bet v 1 polypeptides and relevant controls after the approval of the ethical committee of Helsinki University Central Hospital. The endotoxins of the recombinant Bet v 1 polypeptide preparations were removed by the Detoxi—Gel Endotoxin Removing Gel (Thermo: CatiNo. 20344) where after endotoxin con- tent was analysed by ToxinSensor Endotoxin Detection System (GenScript: Cat. No. LOO35OC). The recombinant Bet v 1 polypeptide ations were fil- ter sterilised by Costar SPIN—X (Cat.No. 8160) and stored in aliquots at 20°C.

SPT was carried out using inant Bet v 1a wt (Biomay) and N28K—E101 K mutant at the trations of 50 and 5 pg/ml and a commercial W0 2012/143374 birch pollen extract ello). Sodium chloride (0.9%) and histamine dihy— drochloride (AlkAbelIo) served as negative and positive controls, respectively.

Before ng the skin, s were set in the tubes containing the skin prick reagents. The responses were measured after 15 minutes and after 6 and 20 hours. The er of the skin response for histamine dihydrochloride in each tested individual was 5 mm after 15 min and this value was selected as posi- tive (+) response (Table 2). The immediate skin reactions induced by the Bet v 1 mutant N28K-E101K with the concentration of 50 ug/ml were milder when compared to the reactions induced by the Bet v 1 wt at the same concentra- 1O tion. The skin reactions induced by the Bet v 1 wt and N28K—E101K mutant with the concentration of 5 ug/ml were comparable. Remarkably in the case of both ic patients the skin reactions induced by the Bet v 1 N28K-E101K mutant disappeared within a notably shorter time than compared to the Bet v 1 W0 2012!143374 PCT/EPZO121057046 Table 2. Results of the skin prick test Patient 1 E 15 min 6 h 20 h Histamine dihydroohloride (10 mg/ml) + + - Birch pollen extract 10HEP (AlkAbello) E + Eget v1 wt (50 pg/ml) + - ++ ++ rBet v1 wt (5 ug/ml) + + rBet v1 N28K+E101K (50 pg/ml) ++ - rBet v1 N28K+E101K (5 ug/ml) + - Patient 2 _ Histamine dihydrochloride (10 mg/ml) + + - Birch pollen t 10HEP(AIkAbello) + + _ rBet v1 wt (50 pg/ml) +++ +++ ++ rBet v1 wt (5 pgiml) + + + rBet v1 N28K+E101K (50 pg/ml) 1- ~ - rBet v 1 N28K+E101K (5 pg/ml) + - _ Non-afopic person E - Histamine dihydrochlon'de (10 mg/ml) + - - Birch pollen extract 10HEP (AlkAbello) E - - rBet v1 wt (50 pg/ml) - - - rBet v1 wt (5 pg/ml) - - - rBet v1 N28K+E101K (50 pg/ml) E - - - rBet v1 N28K+E101K (5 pg/mi) E - - - + diameter of the skin se 2 5mm ++ diameter of the skin response 2 8mm +++ diameter ofthe skin response 2 11mm

Claims (16)

WHAT IS CLAIMED IS

1. A recombinant hypoallergenic birch pollen Bet v 1 polypeptide based on a wild type amino acid sequence template as depicted in SEQ lD NO: 3 or of any other Bet v 1 wild type isoform thereof, said polypeptide either naturally or by on comprising (1) at least one first amino acid substitution at position E101, and (2) at least one second amino acid substitution at position N28.

2. The polypeptide according to claim 1, n the template of a wild type amino acid sequence is as depicted in SEQ lD NO: 3. 10

3. The polypeptide according to claim 1 or 2, wherein the polypeptide has an amino acid sequence as depicted in SEQ lD NO: 39.

4. The polypeptide according to claim 1, wherein the template of a wild type amino acid sequence is selected from the group consisting of SEQ lD NO: 4—38. 15

5. The polypeptide according to any one of claims 1—4, wherein the polypeptide comprises an amino acid sequence ed from SEQ ID NO: 41, and at least one second amino acid substitution at position N28.

6. The ptide according to any one of claims 1—5, wherein the polypeptide comprises 20 an amino acid sequence selected from the group consisting of SEQ ID NO: 41, n amino acid 101 is K, and at least one second amino acid substitution at a position selected from the group consisting of amino acid residues N28.

7. The polypeptide according to claim 8, wherein the polypeptide 25 has an amino acid sequence as depicted in SEQ lD NO: 40.

8. The polypeptide according to any one of claims 1—7, having a ine release capacity which is at least 20x reduced when compared to the histamine release capacity of the wild type Bet v 1.

9. The polypeptide according to claim 8, having a histamine release 30 capacity which is at least 100x reduced when compared to the histamine release capacity of the wild type Bet v 1.

10. The polypeptide according to any one ,of previous claims, wherein there are ther at least two, three or four amino acid substitutions. 35

11. A vaccine composition sing at least one hypoallergenic ptide according to any one of claims 1~10 and a pharmaceutically acceptable diluent or adjuvant.

12. The e composition ing to claim 11, wherein the vaccine is formulated for sublingual administration.

13. Use of a recombinant birch pollen Bet v 1 polypeptide based on a wild type amino acid ce template as depicted in SEQ ID NO: 3 or of any other Bet v 1 wild type isoform thereof, said polypeptide either lly or by mutation comprising (1) at least one first amino acid substitution at position E101, and (2) at least one second amino acid substitution at position N28, in the manufacture of a vaccine for vaccinating against birch pollen allergy. 10

14. Use of a hypoallergenic polypeptide according to any one of claims 1 to 10 or a vaccine composition according to claim 11 or 12 in an amount ive for desensitizing and for inducing the production of protective antibodies against birch pollen in the manufacture of a medicament for vaccinating against birch pollen allergy.

15 15. A method of vaccinating against birch pollen allergy, said meth- od comprising administering to a non~human subject suffering from birch pollen allergy a hypoallergenic polypeptide according to any one of claims 1 to 10 or a vaccine composition according to claim 11 or 12 in an amount effective for itizing and for inducing the production of protective antibodies against 20 birch pollen.

16. The polypeptide of claim 1, substantially as herein described with reference to any one of the Examples and/or