GB2503883A - Variant IgE with reduced affinity to CD23 - Google Patents

Abstract

The invention relates to a variant IgE Fc region, or an antibody comprising that region, comprising at least one amino acid modification in the Ce3 domain relative to a wild-type IgE Fc region, wherein the amino acid modification is part of a CD23 binding site, leading to a reduced binding affinity to CD23 compared to wild-type IgE Fc region. The antibody may be used in treating or preventing a CD23-mediated side effect, for example during the treatment of cancer, autoimmune disease or infectious disease.

Description

ANTIBODY

The present invention relates to antibodies, and in particular to immunoglobulin E (IgE) antibodies. The invention is primarily concerned with variant Fc regions of IgE antibodies and their use in the creation of novel engineered or modified IgE antibodies. The invention extends to methods for modifying antibodies and to pharmaceutical compositions comprising such antibodies. The invention also covers uses of the modified antibodies in a wide range of therapeutic applications.

Immunoglobulin E (IgE) antibodies play a key role in the mechanisms of allergic /0 disease, not only recognizing allergens through their Fab regions, but also interacting via their Fc regions with two very different cdl surface receptors. FcERI, the receptor found on mast cells and basophils, binds IgE with high affinity (KA = 1010_loll M1) and is responsible for allergic sensitization and the immediate (Type I) hypersensitivity reaction in which minute amounts of allergen cross-link receptor- /5 bound IgE and trigger cell degranulation. The IgE-binding a-chain of FceRI consists of two extraceflular Ig-Bke domains (sFcaRla).

In contrast, FcERII (also known as CD23), expressed on B cells, consists of three C- type lectin "head" domains connected to the membrane by a trimeric a-helical coiled-coil "stalk". A single head domain binds to IgE-Fc with lower affinity (KA = 1O51o6 M-') than FcrRI, although avidity of the trimer can substantiafly enhance this interaction. Membrane CD23 (mCD23) is deaved from the cefi surface by endogenous proteases such as ADMIio to yield soluble trimeric and monomeric forms (5CD23), which have been implicated in both positive and negative feedback mechanisms for the regulation of IgE synthesis by B cells that have switched to IgE production. Both FcERI and CD23 are also expressed on a range of antigen-presenting cells (APC) where they play similar roles in trapping IgE-allergen complexes and promoting the allergic response, but the functional interplay (i.e. cooperation or competition) between these two receptors in the context of APCs is Jo not well understood.

CD23 expressed on B cells also has the potential to contribute to the clinically serious phenomenon of the spreading of allergic reactivity to unrelated allergens, through its ability to internalize IgE-allergen complexes irrespective of the allergen, in contrast to mIgE-mediated allergen-specific presentation through the B cell receptor. CD23 expressed on gastrointestinal epithelial cells also contributes to IgE-allergen

--

transport across the gut epithelia harrier to trigger food &lergenic reactions and similarly on respiratory tract epithelial cells to contribute to airway allergic inflammation. Understanding the IgE-CD23 interaction thus has implications for many aspects of allergic disease.

As described in the Examples, using x-ray crystallography, Ni\'IR spectroscopy and site-directed mutagenesis, the inventors have now accurately defined the CD23 binding site on the Fe region of IgE, and have cleaHy demonstrated that the binding site is located in the Cs3 domain of IgE. Based on these results, the inventors have /0 created a series of variant IgE Fc regions, which may be used in the creation of novel and clinicafly safer IgE-based antibody therapies.

Therefore, in a first aspect of the invention, there is provided a variant IgE Fc region comprising at least one amino acid modification in the Cs3 domain relative to a wild- / 5 type 1gB Fc region, wherein the at least one amino acid in the wfld-type IgE Pc region which is modified is part of a CD23 binding site, and wherein the variant IgE Fe region exhibits a reduced binding affinity to CD23 compared to the wfld-type IgE Fc region.

Advantageously, and preferab'y, the inventors have found that the variant 1gB Fc region of the first aspect exhibits reduced, and preferably aNated, binding to CD23 (FcERII), but is capable of binding to FcsRl. Thus, the variant IgE Fc region may still bind to antigens and tngger mast cells through its interaction with FcRI, but, because it cannot bind (or only poorly binds) to CD23, it avoids the problematic side effects that are commonly associated with CD23 binding, including epitope spreading. Therefore, the Fc region of the invention has considerable utility in IgE antibody engineering. The inventors have surprising'y shown that the CD23 and FcRI binding sites occur on opposite ends of the Cc3 domain of IgE, and that allosteric inhibition prohibits simultaneous binding of these two receptors, thereby Jo preventing engagement and cross-linking of IgE bound to mast cells by soluble CD23.

It will be appreciated that immunoglobulin E (IgE) is a class of antibodies found only in mammals. IgE is a dimeric antibody with four Ig-like domains (Csi-Cs4). The ability to mediate q4otoxic and phagocytic effector functions are potent mechanisms by which antibodies destroy targeted cells.

The Fc region links the recognition domain of antibodies to these effector functions through an interaction with Fc receptors and ligands. Manipulation of these effector functions by alteration of the Fc region has important implications in the treatment of numerous medical conditions, e.g. cancer, autoimmune disease and infectious diseases. The Fc region comprises the constant region of an antibody excluding the first constant region immunoglobulin domain. Wild type Fc region refers to the amino acid or nucleotide sequences that are found in nature, including allelic variations. Variant Fc region refers to an Fe sequence that differs from the wild type sequence by virtue of at least one amino acid modification. An Fc variant may only /0 encompass an Fc region, or it may exist in the context of an antibody, Fc ftision, isolated Fc, Fe fragment, or other polypeptide that is substantially encoded by Fc. Fc variant may refer to the Fc polypeptide itself, compositions comprising the Fc variant polypeptide or the amino acid sequence.

/5 The DNA sequence encoding wild-type human IgE-Fc is provided herein as SEQ ID No:i, as follows: (J(iuijAC!1IC(Thj(CCAC(Af(i..p(:ItJI!i.J f((1L1 j!!c!Icci ATCIAA1GAG1CCC?CCACCA1CCGACCTCATCCACiTC CAL TICTCCACCI.TCCI:C I

A

I

ACCACCIACAACICTCCACAI?CCPACCCCACACCCCICACCCCI.CIACCIAACCLCCC AGCCCGTICGACC?GICCAICCGCPACGCCCACGAICACCCGCCIGGICLIGLACCCG

C

C

CCGG1GGLCACCGAcACCGLAICCAGGGALACCCACCALCGCA(CGG1CACCCACCCC CACCIGCCCALCLCCCCCAILCCLCCCACGACCAAGACCAGCCGCCCGC(CCCCIL

C

I

cAGLlccc1;GACLcc;GcAcACcAcLAcGLALCcccCcAALAccAAGCLclCcC-?Cc

I

I

[SEQ ID No:i] The polypeptide sequence of wild-type human IgE-Fc is provided herein as SEQ ID No:2, as follows: C5R_L!_!P±fVKItQSSCDLGL1_!P?I_QtCtVSG5!_P(2_JN_IL2DGQVYDVULiJiAS 233 1 C-FII!IHkKS-rIcILvI:!I L-LJKGH vNIv.KPVRII pvGrLDILGLr5QClivI11R1p_tyR5!_!r±isGppj?Lv5ALvI?Lw?GLUK,6IA 163 (Ct_QNLn2LU_3v1.2vtFNEvcpUALH1I!_!Qp_5RrKLscLHvssRLV r_AEc.!1cuL1.52:3 [SEQ ID No:2] Thus, preferably the variant IgE Fc region of the invention is derived from the wild-type human IgE-Fc comprising an amino acid sequence substantially as set out in SEQ ID No:2, or a functiona' variant or fragment thereof. Preferahly, the variant 1gB Fc region is derived from the wild-type human IgE-Fc which is encoded by a nucleic acid sequence substantially as set out in SEQ ID No:i, or a functional variant or fragment thereof.

The Cs3 domain of IgE is defined as amino acid residues 329-441 of SEQ ID No:2.

Thus, the at least one amino acid which is modified may be any one of amino acid residues 329-441 of SEQ ID No:2.

As described in the Examp'es, and as illustrated in Figure iB, the inventors are the to first to demonstrate that the A-B helix of Cc3 does not form part of the CD23 binding site, and so it cannot be directly involved in CD23 binding. Therefore, preferably the at east one amino acid modification in the variant IgE Fc region is not in the A-B helix of the Cs3 domain of IgE. It will be appreciated that the A-B helix of the Cs3 domain of IgE is defined as amino acid residues 344-353 of SEQ ID No:2.

The inventors have observed that a number of amino acid residues from three discontinuous sequences in the Fc region of IgE form a contiguous surface, which together represent the binding site on the Ca3 domain of IgE for the CD23 receptor.

These three spaced apart sequences are located in the E-F helix, the C-D loop and the C-terminal region of the Cc3 domain of IgE. Accordingly, it is preferred that the variant IgE Fc region comprises at least one amino acid modification in the E-F helix of the Cc3 domain of IgE, in the C-D loop of the Cc3 domain of IgE and/or in the C-terminal region of the Ca3 domain of IgE.

It will be appreciated that the E-F helix of the Cc3 domain of IgE is defined as amino acid residues 404-415 of SEQ ID No:2, the C-D loop of the CE3 domain of IgE is defined as amino acid residues 376-380 of SEQ ID No:2, and that the C-terminal region of the CE3 domain of IgE is defined as amino acid residues 433-441 of SEQ ID No:2. It will be appreciated that the numbering of amino acid residues given herein is Jo equivalent to the wild-type sequence of human Ig Fc region as shown in SEQ ID No:2.

Therefore, preferably the variant IgE Fc region comprises at least one amino acid modification at a position corresponding to wild-type IgE Fc region sàlected from the group consisting of: (i) amino acids 404-415 from the E-F helix of IgE; (ii) amino acids 376-380 from the C-D oop of IgE; and (iii) residue 433-441 from the C-terminal region of IgE. Thus, the variant IgE region may comprise at least one amino acid modification at a position corresponding to wild-type IgE Fc region selected from the group consisting of: 376; 377; 378; 379; 380; 404; 405; 406; 407; 408; 409; 410; 411; 412; 413; 414; 415; 433; 434; 435; 436; 437; 438; 439; 440; and 441 of SEQ IDNo:2.

The inventors have analysed the bound structure of the IgE/CD23 complex and used considerable inventive endeavour to determine the solvent-exposed IgE residues that make contact with CD23. Accordingly, the variant IgE region may comprise at least /0 one amino acid modification at a position corresponding to wild-type IgE Fc region selected from the group consisting of: 376; 378; 381; 408; 409; 411 412; 435; 439; and 440.

The inventors observed that mutations of amino acid residues at positions 409, 412, /5 376, 380 and 435 displayed the largest effects on CD23 binding energetics. Therefore, preferably the variant IgE region comprises at least one amino acid modification at a position equivalent to 409,412,376,380 and/or 435 of the wild4ype human IgE Fc region.

The at least one modification may comprise an amino acid deletion or an insertion.

However, preferably the at least one modification comprises an amino acid substitution. Preferably, the at east one modification may comprise a non-conservative amino acid change. For example, the charge (i.e. PH) of amino acid may be changed, or the residue maybe converted from being hydrophilic so that it is hydrophobic, or vice versa. For example, a K residue may be substituted for a D, or an E may be replaced with a R. Alternatively, a smafi amino acid may be replaced with a thrge amino acid (e.g., Y for A).

Ahernatively, a arge amino acid may be replaced with a small or constraining amino acid, for example alanine, gycine, proline or serine (e.g., F may be replaced with 5).

Substitution with alanine maybe preferred, and the skilled person would readily appreciate the codons which encode alanine, and therefore how to modify the wild-type DNA sequence of SEQ ID No: 1 to produce variant Fc regions of the invention.

For example, in one embodiment the variant Fc region may comprise at least one amino acid modification selected from: D409A; E412A; R376A; K38oA and K435A.

The variant human IgE Fe region may comprise more than one modification in the Cs3 domain of the Fe region equivalent to the human IgE Fe region of a wild-type antibody. For examp'e, the variant human IgE Fe region may comprise at least two, three, four or five modifications in the C3 domain, as descnbed herein. Thus, the variant Fc region may comprise a double or treble mutant, and so on. The variant Fe region may comprise a modification at two, three, four, five or more of the amino acids at positions equivalent to 376; 377; 378; 379; 380; 404; 405; 406; 407; 408; 409; 41O 411; 412; 413; 414; 415; 433; 434; 435; 436; 437; 438; 439; 440; and 441 of wild-type human IgE-Fe. The variant Fe region may comprise a modification at each to of the amino acids at positions equivalent to 409,412,376,380 and/or 435 of wild-type human IgE-Fc.

In one embodiment, the p&ypeptide sequence of the vanant IgE Fe region (denoted as D409A) is provided herein as SEQ ID No:3, as follows: /5 (_3k0F'2 IJKII sIcI,k'-'rç)I 0_1IVS0y F' INI5AjI!I)00!V\)VI1Ij_YIiS 233 1T43L0ThtAST.3S2JftS4K11WSUPT YtC.3V1.3.53 43:3 1101 163 0t_QNLi'-'1±LD_SV3Vtt.N1±VQPLiA-511ST_Q P_-Kf±4.5(U_v_.5RL.5vf_-L QKU2SO. .523 C2AVJLAAS 0551V.5RAV.5vN? [SEQ ID No:3] In another embodiment, the polypeptide sequence of the variant IgE Fe region (E412A) is provided herein as SEQ ID No:4, as follows: C5R_1L._P±5VK1LQSSCDG1._.P?T_QL63LVSGYt PGF1N_TL2DGQVYDVL)LU_AS 233 iT!.2LcLtAo ft'V06HW'UR2 YrcQvr Qc.-t_*0.5SLR±.5AUSNPRCV 34:3 SPPOLPIRK3?TI?CLVVDLA?SIKCTVKL?W33SASCKPVKRS?OJKEEKQSOKCTLOVIS?L 433: 163 0t_QNLflOLU_SVQVQLFNEVQ63L1A2.Ho P_-5RTKC3CL_0r3RL_.5f_.5AE4.5Q16UL.5I. 52:3 OOAVHEAA3PSQIVORAVSV2J? [SEQ ID No:4] In another embodiment, the polypeptide sequence of the variant IgE Fe region (R36A) is provided herein as SEQ ID No:, as follows: 0.5R_:L'_T±fvKi.tQ55CD.5'jct._ 233.

SFTQL.5LLA.5TQ3.5..5ft.5.5KHW6322T YtC.5 63CJ_S.5D31R±CAUSNPRJ5V SAYL.5RP 34:3 5'FIHI!IIkV,i'-' 61631234111 j.-'KlvNl v,A634F(FVN 103 PVGTIODWIEGFTYQCRVTHPHIP2ALYR3?TIKI2300JAOEVYAFAIOEWOSSSADKI.5ILA 463 0210516M1'EDISVQWLE4FV4100AAHST305JKTK4S4PFVFSRLEVTOAFWEQKDEFI 523 O2CAVHEAASP6QIVORAVSV2J? [SEQ ID No:5] In another embodiment, the polypeptide sequence of the variant IgE Fe region (K380A) is provided herein as SEQ ID No:6, as follows: 03R3F?P27VKILQS355D(4EFP?TICiLICLVSGYTPG?INIT3LEDGC,1VYDVDL3?A3 263 TT5ECELA3T5!363TL3cKH13DRTYT053JTYQC333'ED31KKCAD3NPRCV3AYL3RP 3:43: 3065LPIRKS?TI?CLVVDLAOSITVKL?WSOASGAPVNHSTRKEEKQRNOTLSVISTL 403 PVGTcDWIEGFTYCCRVTHPH100ALYRS?TKIS200IAOEVYAFAIOEWOO3SIKDK2IIA 463 ci 523 1I0V-1FJ\.rSF;5i0TVc;PTj5VcD [SEQ ID No:6] In another embodiment, the polypeptide sequence of the variant IgE Fc region (K435A) is provided herein as SEQ ID No:7, as follows: 10!CJ vscypcr' TI TiT0' DQ(VV0VDT!3TS 233 T 3/3 061 I!LIkK2)-IICIVIiI J%sKG!I vNI vS2/\SGKPVRIIS 3 123 463 (ST 523 c-coo II!!3%.oi6006Ivc;kr003s_) [SEQ ID No:7] Thus, a preferred variant IgE Fc region of the invention comprises an amino acid sequence substantially as set out in any one of SEQ ID No: 3-7, or a functional variant or fragment thereof.

The IgE Fc variants disdosed herein exhibit reduced binding affinity to the CD23 receptor compared to that of the wild-type or parent Fc region. The term "reduced affinity" compared to the wild-type Fc region can mean that the Fc variant binds to CD23 with a lower equilibrium constant of association (KA or 1(a) or higher equilibrium constant of dissociation (K or 1(d) than the wild-type Fc region when the amounts of variant and wild type region in the binding assay that is used are essentiafly the same. For examp'e, the Fc variant with reduced binding affinity for CD23 may dispthy from about a 5-fold to about io,ooo-fokl reduction in CD23 binding affinity compared to the wild type Fc region, where CD23 binding affinity is determined, for examp'e, by the binding methods disclosed herein.

Reduced affinity can also be defined rdative to an absolute levd of affinity. As shown in Tab'e 1, the preferred variants exhibit surprisingly high Ku values. For example, the equilibrium constant of dissociation (LCD or Kd) of the variant IgE Fc region for CD23 may be at least jiM, lojiM, 25iM or more.

Preferably, the variant IgE Fc region of the first aspect is capable of binding to FcsRI.

Preferably, the binding affinity of the variant Fc region for FccRI is at least the same as that of the wild-type Fc region. For example, the variant IgE Fc region is capable of binding to FesRI with a KA value of at least 1w M', io7 M', 108 M0, 1o M0, or at least 1010 J\j-1 In a second aspect, there is provided a nucleic acid encoding the variant IgE Fc region according to the first aspect.

In one embodiment, the DNA sequence encoding the variant IgE Fc region (D409A) is provided herein as SEQ ID No:8, as follows: GCCGAC1 _!CCCCCCCACCATCCAGC1CC!_C1CCCTCCTCFG_!CCCTACACCCCACCGAC1 CAPCA?CACC? CC SIC SAC CACOC CCACC SCAIC CACCICCACI? CICCACCCCC?CI /0 ACCACCAACAAC?CICCACA?ICCAACCCCACACCCCICACCCCCIACC?AACCCCCCCC ACGCCCI?CCACCIC?ICAICCCCAACICCCCCACCAICACCICIC?CC?CCICCACCTC

C /5 (

CC C St C ?CGC?CA?CCACPAC?ICAICCC?CACCAGASCICCCSCCACICCC?CCACPACCACCIC c

CCTCLGACGIGACCACLG

AAICCCS

[SEQ ID No:8] In another embodiment, the DNA sequence encoding the variant IgE Fc region (E412A) is provided herein as SEQ ID No:g, as follows: JO CCCCACI1CCCSCCCACCAFCCACCICCICFCCCICCFC!_!C1CCCFACACCCCACCCACI AICAACAICACCICCC?CCACCACCCCCACCICAICCACCTCCACIICICCACCCCCICI

A

IGGCIGTCAGASCGCAGCSACACC?GCGAGGICAGCSAIC 5555 ACCIIISACCAC ACCACCAACAACICTCCACAI?CCPACCCCACACCCCICACCCCCIACCIAACCCCCCCC ACCSCCTIGCASC?CI?CAICStCPACICCCCCAGCAICACC?CSSICCICCICCACC?C

S

CACICCAC

CACSIGCCGASCSCCC?CAISCCS?CCACGACCAAGAGCAGCCGCGCGC?SSISGCCCG GAACICTAIGCCI?TSCGAGSCCCS;AsICssscCt;CGACCCGCS;ACAAGCSCACCCICSCC

I

CAGCICCCCGAGCCCCCGCACACCACSACGCASGGCCPACACCGCGCIGCSCC?TC

I

I

[SEQ ID No:9] In another embodiment, the DNA sequence encoding the variant IgE Fc region so (R6A) is provided herein as SEQ ID No:io, as follows: ICICCACCCAGI?CACCI.CGCI.CCCACC CICAACAICSIACAC?CCICCICSCACCCCC GGGCACTIGCSCCGAICCAICCA(ICGIGIGCGICSIC?CIGGSIACACCCGASGSACI AIC}C}CICACJII.JCC!Ci-iCi*iC:*CSt c!1!CAt!Ci*-iCi!j:i-Cr;( j!!IiJ JI! ACACiCACCACC SIC ASS 1. JCCL ?SCACACAAACLSACCISACCS?CACCCACAAC SAC IGGCIGTCAGASCGCAICCSACACC?GCGAGGICAGCSAICAAGGSGACSIIISASSAC

A

ACCCCICTICCAGC ?CIThCAILCSLAAL ICCCCCACCAICACC 2555155155155 ASS SC CCACCCACCCCCACCCFCCCICACCFCC:CC1CStC:t!CCC5J1CCCICFCAAC

CACICCAC c

CACSIGCCSASGSCCC?CAISCGS?CCACGACCAAGASCAGCGGCSCGSS?SSISSCCCG GAACICTAIGCCI?TSSGASSCCCS;ASIGGSCCS;CGACCCGCS;ACAAGCCSSACCCICSCC

T

[SEQ ID No:io] In another embodiment, the DNA sequence encoding the variant IgE Fc region /0 (K3SoA) is provided herein as SEQ ID No:ii, as follows:

I

G.1CGAG1GACT ATCIAAIGAGICC?CCACCAICCGACCICATCCACi.iTC CAL IICICCACCC

A /5

IGCA

AGCCCGTICGACC?GICCAICCGCPACGCCCACLATCACCTLCCIGGICLIGLACCTL (IACJ' CE ACCACLAIJCCL7 Li c CACCIGCCCALCLCCCCCAILCCLCCCACGALCAALACCALCGLCCCCCGILCIL.

C

I 1

I

[SEQ ID No:ii] In another embodiment, the DNA sequence encoding the variant IgE Fc region (K435A) is provided herein as SEQ ID No:12, as follows:

I

GGGCACTICCCCCGACCAICCA(ICCIGIGCCICLIC?CICGLIACCCCALCLACI

A

A

I

A

LciCc(4TTCcC7\ cc'cq'c J'ccGc7\ACJ'CCCCc7\CG7\J'c7\Ci'GCIGGTGi-J'GL-/\ccTh;

C

CALl CCACCACLAAACCLA Ci ACAA (LCAH1GCAAICJ1'TAACLICACGqCCJCCCG

CACCIGCCCACCCCCCCCAICCC

qTiLj(<AiLJ' CCALGAGCCGCGACAAGC CCACCCJ'C CCC 1 nCG!ICTiCACCCCCC_!CCACCrCACCACCGCCGfJv_GCCAAiJGA_A(_rCA_!C

I

[SEQ ID No:12] Thus, the nucleic acid according to the second aspect may comprise a nucleotide sequence substantially as set out in any one of SEQ ID No: 8-12, or a functional is variant or fragment thereof. For each of the nucleic acid sequences represented by SEQ ID No: 8-12, the codon which is shown in bold and underlined represents a modification which would encode alanine. However, it will be appreciated that alanine may be encoded by GCT, GCC, GCA and GCG and so any of these codons would be suitable as a modification to create the same amino acid mutation.

-10 -Similary, the codons for gycine (GGT, GGC, GGA and GGG), profine (CCT, CCC, CCA, CCG) and serine (TCT, TCC, TCA, TCG, AGT and AGC) will be known, and could be used to modi' wild type sequence.

The nucleic acid may be an isolated or purified nucleic acid sequence. The nucleic acid sequence may be a DNA sequence. Also described herein are methods for producing the variant IgE Fc region of the first aspect using the nucleic acids of the second aspect. Thus, the nucleic acid sequence may be incorporated into a genetic construct for cloning purposes. /0

In a third aspect, there is provided a genetic construct comprising the nucleic acid of the second aspect.

Genetic constructs of the invention may be in the form of an expression cassette, /5 which maybe suitable for expression of the encoded polypeptide (i.e. the variant IgE Fc region) in a host cefi. The genetic construct may be introduced in to a host cell without it being incorporated in a vector. For instance, the genetic construct, which may be a nucleic acid molecule, may be incorporated within a liposome or a virus particle. Alternatively, a purified nucleic acid molecule (e.g. histone-free DNA, or naked DNA) may be inserted directly into a host cell by suitable means, e.g. direct endocytotic uptake. The genetic construct may be introduced directly in to cells of a host subject (e.g. a bacterial or eukaryotic cefi) by transfection, infection, àlectroporation, microinjection, cell fusion, protoplast fusion or ballistic bombardment. Alternatively, genetic constructs of the invention may be introduced directly into a host cell using a partide gun. Alternatively, the genetic construct may be harboured within a recombinant vector, for expression in a suitable host cell.

Therefore, in a fourth aspect, there is provided a recombinant vector comprising the genetic construct according to the third aspect.

The recombinant vector may be a plasmid, cosmid or phage. Such recombinant vectors are useful for transfoiming host cells with the genetic construct of the third aspect, and for replicating the expression cassette therein. The skilled technician will appreciate that genetic constructs of the invention may be combined with many types of backbone vector for expression purposes. Recombinant vectors may include a variety of other functional elements including a suitable promoter to initiate gene expression. For instance, the recombinant vector may be designed such that it -11 -autonomously replicates in the cytosol of the host cell. In this case, elements which induce or regulate DNA replication may be required in the recombinant vector.

Alternatively, the recombinant vector maybe designed such that it integrates into the genome of a host cell. In this case, DNA sequences which favour targeted integration (e.g. by homologous recombination) are envisaged.

The recombinant vector may also comprise DNA coding for a gene that may be used as a selectable marker in the cloning process, i.e. to enable selection of cells that have been transfected or transformed, and to enable the selection of cells harbouring /0 vectors incorporating heterologous DNA. Alternatively, the selectable marker gene may be in a different vector to be used simultaneously with vector containing the gene of interest. The vector may also comprise DNA involved with regulating expression of the coding sequence, or for targeting the expressed polypeptide to a certain part of the host cell. /5

In a fifth aspect, there is provided a host cell comprising the genetic construct according to the third aspect, or the recombinant vector according to the fourth aspect.

The host cell may be a bacterial cell. The host cell may be an animal cell. The host cell may be a mammalian cell, for examp'e a mouse or rat cell. It is preferred that the host cell is not a human cell. The host cell may be transformed with genetic constructs or vectors according to the invention, using known techniques. Suitable means for introducing the genetic construct into the host cell will depend on the type of cell.

Thus, according to a sixth aspect, there is provided a method of preparing a variant IgE Fc region, the method comprising-(i) culturing at east one cell according to the fifth aspect under conditions suitable for the expression of a variant IgE Pc region; and (ii) isolating the variant IgE Fc region.

The inventors believe that knowledge of the sequences of IgE Fc region which form the binding site for the CD23 receptor can be harnessed in a method for reducing the binding affinity of any existing IgE antibody or in the production of novel IgE antibodies, in order to reduce or avoid the risks of patients suffering from any CD23-mediated side effects that may be associated with an IgE antibody therapy.

-12 -Hence, in a seventh aspect, there is provided a method of producing a variant IgE Fc region which exhibits a reduced binding affinity for CD23, the method comprising modifying at least one amino acid in the CE3 domain of a wild-type IgE Fc region, wherein the at least one amino acid in the wild-type IgE Fc region which is modified is part of the CD23 binding site, such that the variant IgE Fc region exhibits a reduced binding affinity to CD23 compared to the wild-type IgE Fc region.

Preferably, the variant 1gB Fc region produced by the method is incapable of binding /0 to CD23 (FcsRII), but is capable of binding to FcsRL Preferably, the method does not comprise modifying an amino acid in the A-B helix of the C3 domain of 1gB. Preferably, the method comprises modifying at east one amino acid in the E-F helix, in the C-D loop and/or in the C-terminal region of the /5 Ca domain of IgE. Preferably, the method comprises modifying at least one amino acid at a position corresponding to wild-type 1gB Fc region s&ected from the group consisting of: (i) amino acids 404-415 from the E-F helix of IgE; (ii) amino acids 376- 380 from the C-D loop of IgE; and (iii) residue 433-441 from the C-terminal region of IgE. The method may comprise modifying at least one amino acid at a position corresponding to wild-type 1gB Fc region selected from the group consisting of: 376; 377; 378; 379; 380; 404; 405; 406; 407; 408; 409; 410; 411 412; 413; 414; 415; 433; 434; 435; 436; 43; 438; 439; 440; and 441 of SEQ ID No:2. Preferably, the method comprises modifying at least one amino acid at a position equivalent to 409, 412, 376, 380 and/or 435 of the wild-type human IgE Fc region. The method may comprise modifying at least two, three, four, five or more amino acids in the Cc3 domain relative to a wild-type IgE Pc region.

The wild-type IgE Pc region which is modified in the method may be part of a therapeutically active IgE antibody. Examples of such an 1gB antibody which may be Jo modified include 1gB antibodies directed against tumour antigens, as described by Karagiannis et cii. (Karagiannis, 2011, Cancer Jminuncil Immunotherapy; PMID 22139135). Thus, the method may be used to modify the 1gB antibody such that it cannot bind to CD23, thereby avoiding the risk of CD23-mediated side effects.

is Alternatively, the method may comprise initially creating the variant 1gB Pc region and then fusing it to a functional fragment of an 1gB antibody to produce an Fc -13 -fusion, for example a chimera. The functional fragment may comprise variable regions exhibiting immunospecificity for a target epitope. The functional fragment may be selected from a group consisting of VU, VL, Fd, Fv, Fab, Fab', scFv and F(ab') 2.

Based on the foregoing, it will be appreciated that the variant IgE Fe region of the first aspect or the method of the seventh aspect can be used to form antibody or a functional fragment thereof.

/0 Thus, in an eighth aspect of the invention, there is provided an antibody or a functional fragment thereof comprising the variant IgE Fe region of the first aspect.

Preferably, the antibody or functiona' fragment thereof is incapable of binding to CD23 (FccRII), or exhibits a reduced binding affinity thereto, but is capable of /5 bindingtoFcsRl.

The invention extends both to whole antibodies (i.e. immunoglobulins) with immunospecificity for a certain target epitope, as well as to functional fragments thereof. Such fragments retain at least one antigen binding region of a corresponding full-length antibody as well as the variant IgE Fe region of the invention. The antibody or functional fragment thereof may comprise a monoclonal or polyclonal antibody or functional fragment thereof.

The antibody or functional fragment may be monovalent, divalent or polyvalent.

Monovalent antibodies are dimers (HL) comprising a heavy (H) chain associated by a disulfide bridge with a light chain (L). Divalent antibodies are tetramer (H2L2) comprising two dimers associated by at least one disulfide bridge. Polyvalent antibodies may also be produced, for example by linking multiple dimers. The basic structure of an antibody molecule consists of two identical light chains and two identical heavy chains which associate non-covalently and can be Unked by disulphide bonds. Each heavy and light chain contains an amino-terminal variable region of about 110 amino acids, and constant sequences in the remainder of the chain. The variable region indudes several hypervariable regions, or compementanty determining regions (CDRs), that form the antigen-binding site of the antibody m&ecule and determine its specificity for the antigen in question. On either side of the CDRs of the heavy and light chains is a framework region, a rethtively conserved sequence of amino acids that anchors and orients the CDRs. -14-

The constant region consists of one of five heavy chain sequences (t, y, i, a, or c) and one of two light chain sequences (K or X). The heavy chain constant region sequences determine the isotype of the antibody and the effector functions of the molecule.

Preferably, the antibody or functional fragment thereof is an IgE.

The antibody or fragment thereof may be a human antibody. As used herein, the term "human antibody" can mean an antibody, such as a monoclonal antibody, which comprises substantially the same heavy and light chain CDR amino acid sequences as /0 found in a parUcular human antibody exhibiting immunospecificity for certain epitope. An amino acid sequence, which is substantially the same as a heavy or light chain CDR, exhibits a considerable amount of sequence identity when compared to a reference sequence. Such identity is definitively known or recognizable as representing the amino acid sequence of the particular human antibody.

/ 5 Substantially the same heavy and light chain CDR amino acid sequence can have, for example, minor modifications or conservative substitutions of amino acids. Such a human antibody maintains its function of selectivdy binding to an epitope.

The term "human monodonal antibody" can include a monoclonal antibody with substantially human CDR amino acid sequences produced, for examp'e by recombinant methods such as production by a phage library, by ymphocytes or by hybridoma cells.

The term "humanised antibody" can mean an antibody from a non-human species (e.g. mouse) whose protein sequences have been modified to increase their similarity to antibodies produced naturally in humans.

The antibody may be a recombinant antibody. The term "recombinant human antibody" can include a human antibody produced using recombinant DNA Jo technology.

The term "antigen binding region" can mean a region of the antibody having specific binding affinity for its target antigen. The binding region may be a hypervariable CDR or a functional portion thereof. The term "functional portion" of a CDR can is mean a sequence within the CDR which shows specific affinity for the target antigen.

The functional portion of a CDR may comprise a ligand which specifically binds to its epitope.

-15 -The term "CDR" can mean a hypervariable region in the heavy and light variable chains. There may be one, two, three or more CDRs in each of the heavy and light chains of the antibody. Normally, there are at least three CDRs on each chain which, when configured together, form the antigen-binding site, i.e. the three-dimensional combining site with which the antigen binds or specifically reacts. It has however been postulated that there may be four CDRs in the heavy chains of some antibodies.

The definition of CDR also includes overlapping or subsets of amino acid residues when compared against each other. The exact residue numbers which encompass a /0 particular CDR or a functional portion thereof, will vary depending on the sequence and size of the CDR. Those skilled in the art can routinely determine which residues comprise a particular CDR given the variable region amino acid sequence of the antibody.

/5 The term "functional fragment" of an antibody can mean a portion of the antibody which retains a functional activity. A functional activity can be, for example antigen binding activity or specificity. A functional activity can also be, for exampk, an effector function provided by an antibody constant region. The term "functional fragment" is also intended to include, for example, fragments produced by protease digestion or reduction of a human monoclonal antibody and by recombinant DNA methods known to those skilled in the art. Human monoclonal antibody functional fragments include, for example individual heavy or light chains and fragments thereof, such as VL, VU and Fd; monovalent fragments, such as Fv, Fab, and Fab' bivalent fragments such as F(ab')2; single chain Fv (scFv); and Fc fragments.

The term "VI. fragment" can mean a fragment of the Bght chain of a human monodonal antibody which includes all or part of the Ught chain variable region, including the CDRs. AVL fragment can further include light chain constant region sequences.

The term "VU fragment" can means a fragment of the heavy chain of a human monoclonal antibody which includes all or part of the heavy chain variable region, including the CDRs.

The term "Fd fragment" can mean the Ught chain vailable and constant regions coupled to the heavy chain variable and constant regions, i.e. VL CL and VII CH-i. -16-

The term "Fv fragment" can mean a monovalent antigen-binding fragment of a human monoclonal antibody, including all or part of the variable regions of the heavy and light chains, and absent of the constant regions of the heavy and light chains. The variable regions of the heavy and light chains include, for example, the CDRs. For example, an Fv fragment includes all or part of the amino terminal variable region of about 110 amino acids of both the heavy and light chains.

The term "Fab fragment" means a monovalent antigen-binding fragment of a human monoclonal antibody that is larger than an Fv fragment. For example, a Fab fragment /0 includes the variable regions, and all or part of the first constant domain of the heavy and light chains. Thus, a Fab fragment additionally includes, for example, amino acid residues from about 110 to about 220 of the heavy and light chains.

The term "Fab' fragment" can means a monovalent antigen-binding fragment of a is human monoclonai antibody that is larger than a Fab fragment. For example, a Fab' fragment inchides afi of the Bght chain, afi of the variable region of the heavy chain, and all or part of the first and second constant domains of the heavy chain. For example, a Fab' fragment can additionally include some or all of amino acid residues 220 to 330 of the heavy chain.

The term "F(ab')2 fragment" can mean a bivalent antigen-binding fragment of a human monodonal antibody. An F(ab')2 fragment includes, for example, all or part of the variable regions of two heavy chains-and two light chains, and can further include all or part of the first constant domains of two heavy chains and two Bght chains.

The term "single chain Fv (scFv)" can mean a fusion of the variable regions of the heavy (VH) and light chains (VL) connected with a short Unker peptide.

One skilled in the art knows that the exact boundaries of a fragment of a human monodonal antibody are not important, so thng as the fragment maintains a functional activity. Using well-known recombinant methods, one skilled in the art can engineer a polynucleotide sequence to express a functional fragment with any endpoints desired for a particular application. A functional fragment of the antibody may comprise fragments with substantially the same heavy and light chain variable regions as the human antibody.

The functional fragment may include fragments wherein at least one of the binding -17 -region sequences has substantiafly the same amino acid sequence as the binding region sequences of the antibody. The functional fragment may comprise any of the fragments selected from a group consisting of VII, VL, Fd, Fv, Fab, Fall, scFv, F (all)2 and Fc fragment. Thus, in some embodiment, the invention comprises a functional fragment of an antibody (e.g. Fv, Fab, and Fall, bivalent fragments such as single chain Fv (scFv), fused to the Fc region of the first aspect.

The functional fragment may comprise any one of the antigen binding region sequences of the VL, any one of the antigen binding region sequences of the VH, or a /0 combination of VL and VII antigen binding regions of a human antibody. The appropriate number and combination of VH and VL antigen binding region sequences may be determined by those skilled in the art depending on the desired affinity and specificity and the intended use of the functional fragment. Functional fragments of antibodies may be readily produced and isolated using methods well is known to those skilled in the art. Such methods include, for example, proteolytic methods, recombinant methods and chemical synthesis. Proteob4ic methods for the isolation of functional fragments comprise using human antibodies as a starting material. Enzymes suitable for proteolysis of human immunoglobulins may include, for example, papain, and pepsin. The appropriate enzyme may be readily chosen by one skilled in the art, depending on, for example, whether monovalent or bivalent fragments are required. For example, papain cleavage results in two monovalent Fall fragments that bind antigen and an Fc fragment. Pepsin deavage, for example, results in a bivalent F (ab) fragment. An F (ab')2 fragment of the invention maybe further reduced using, for examp'e, D'Vl' or 2-mercaptoethan& to produce two monova'ent Fab' fragments.

Functional fragments produced by proteolysis may be purified by affinity and column chromatographic procedures. For example, undigested antibodies and Fc fragments may be removed by binding to protein A. Additionafly, functional fragments may be purified by virtue of their charge and size, using, for example, ion exchange and gel filtration chromatography. Such methods are well known to those skilled in the art.

The human antibody or functional fragment thereof may be produced by recombinant methodology. Preferably, one initially isolates a polynucleotide encoding desired regions of the antibody heavy and fight chains. Such regions may include, for example, all or part of the variable region of the heavy and light chains.

Preferably, such regions can particularly include the antigen binding regions of the -18 -heavy and Ught chains, preferaNy the antigen binding sites, most preferably, the CDRs.

The polynucleotide encoding the human antibody or functional fragment of the invention may be produced using methods known to those skilled in the art. The polynucleotide encoding the antibody or a functional fragment thereof may be directly synthesized by methods of oligonucleotide synthesis known in the art.

Alternatively, smaller fragments may be synthesized and joined to form a larger functional fragment using recombinant methods known in the art. /0

As used herein, the term "immunospecificity" can mean the binding region is capable of immunoreacting with a certain target epitope, by specifically binding therewith.

The term "immunoreact" can mean the binding region is capable of eliciting an /5 immune response upon binding with an epitope.

The term "epitope" can mean any region of an antigen with ability to elicit, and combine with, a binding region of the antibody or fragment thereof.

Since IgE antibodies play a central role in the initiation and regulation of allergic disorders, clinical applications of the variant IgE Fc region of the first aspect, and of the antibodies (i.e. modified antibodies) of the eighth aspect, include their use in treating a wide range of medical conditions (including cancer, autoimmune disease or infectious disease), while simultaneously avoiding the d&eterious aergic side effects that are often observed in patients who are treated with IgE antibody therapies. In addition, they may also have utility in diagnosis.

Accordingly, in a ninth aspect of the invention, there is provided a variant IgE Fc region according to the first aspect, or an antibody or a functional fragment thereof according to the eighth aspect, for use in therapy or diagnosis.

In a tenth aspect, there is provided provided a variant IgE Fc region according to the first aspect, or an antibody or a functional fragment thereof according to the eighth aspect, for use in treating, preventing or ameliorating a CD23-mediated side effect.

In an eleventh aspect, there is provided a method for treating, preventing or ameliorating a CD23-mediated side effect in a subject, the method comprising -19 -administering, to a subject in need of such a treatment, a therapeutically effective amount of a variant IgE Fe region according to the first aspect, or an antibody or a functional fragment thereof according to the eighth aspect.

Preferably, the variant IgE Fc region, or the antibody or a functional fragment thereof is used for preventing a CD23-mediated side effect. The CD23-mediated side effect, which is prevented or avoided, may be caused by IgE antibody treatment or therapy, of which there are many types. The CD23-mediated side effect which is prevented is one which may otherwise be caused during the treatment of cancer, autoimmune /0 disease or infectious disease with a therapeutic IgE antibody. Thus, the antibody or fragment thereof may be a therapeutic antibody.

A CD23-mediated side effect which may be prevented may include CD23-dependent epitope spreading, which will be known to the skilled person. Epitope spreading is is the process by which an antibody response to one epitope on an antigen leads to the production of antibodies specific for other epitopes on the same antigen, or for epitopes on entirely unrelated antigens. This results from the internalisation of who'e antigen and subsequent display of a range of peptides derived from that antigen, leading to the generation of T cells with different epitope specificities. Simultaneous processing of two unrelated antigens by an antigen-presenting cell can lead to the production of antibodies directed against both antigens (Gould and Sutton, 2008, Nature Reviews [mrnunology, 8, 205-217).

It will be appreciated that antibodies, fragments, peptides and nucleic acids according to the invention (collectively referred to herein as "agents") may be used in a monotherapy (e.g. the use of an antibody or fragment thereof atone), for treating, ameliorating or preventing a CD23-mediated side effect. Alternativdy, agents according to the invention may be used as an adjunct to, or in combination with, known therapies for treating, ameborating, or preventing CD23-mediated side effects.

The agents according to the invention may be combined in compositions having a number of different forms depending, in particular, on the manner in which the composition is to be used. Thus, for example, the composition may be in the form of a powder, tablet, capsule, liquid, ointment, cream, g&, hydrogel, aerosol, spray, micellar solution, transdermal patch, liposome suspension or any other suitable form that may be administered to a person or animal in need of treatment. It will be -20 -appreciated that the vehicle of medicaments according to the invention slioukl be one which is well-tolerated by the subject to whom it is given, and preferably enables delivery of the agents across the blood-brain barrier.

Medicaments comprising agents of the invention maybe used in a number of ways.

For instance, oral administration may be required, in which case the agents may be contained within a composition that may, for example, be ingested orally in the form of a tablet, capsule or liquid. Compositions comprising agents and medicaments of the invention may be administered by inhalation (e.g. intranasally). Compositions /0 may also be formulated for topical use. For instance, creams or ointments may be applied to the skin.

Agents and medicaments according to the invention may also be incorporated within a slow-or delayed-release device. Such devices may, for example, be inserted on or /5 under the skin, and the medicament may be released over weeks or even months.

The device may be located at least adjacent the treatment site. Such devices may be particularly advantageous when long-term treatment with agents used according to the invention is required and which would normally require frequent administration (e.g. at least daily injection).

In a preferred embodiment, agents and medicaments according to the invention may be administered to a subject by injection into the blood stream or directly into a site requiring treatment. Injections may be intravenous (bolus or infusion) or subcutaneous (bolus or infusion), or intradermal (bolus or infusion).

It will be appreciated that the amount of the antibodies, fragments, peptides and nucleic acids (i.e. agent) that is required is determined by its biological activity and bioavailability, which in turn depends on the mode of administration, the physiochemical properties of the agent, and whether it is being used as a monotherapy or in a combined therapy. The frequency of administration will also be influenced by the half-life of the agent within the subject being treated. Optimal dosages to be administered may be determined by those skilled in the art, and will vary with the particular agent in use, the strength of the pharmaceutical composition, the mode of administration, and the advancement of the disease. Additional factors depending on the particular subject being treated will result in a need to adjust dosages, including subject age, weight, gender, diet, and time of administration.

-21 -Gener&ly, a daily dose of between 0.001 jig/kg of body weight and 10mg/kg of body weight of agent according to the invention may be used for treating, ameliorating, or preventing a CD23-mediated side effect. More typically, the daily dose of agent is between 0.01 jig/kg of body weight and 1mg/kg of body weight, more preferably between o.ipg/kg and ioo jig/kg body weight, and most preferably between approximately o.ijig/kg and iojig/kg body weight.

The agent may be administered before, during or after onset of the side effect. Daily doses may be given as a single administration (e.g. a single daily injection).

Alternatively, the agent may require administration twice or more times during a day.

As an example, agents may be administered as two (or more depending upon the severity of the condition being treated) daily doses of between 0.07 jig and oomg (i.e. assuming a body weight of 70 kg). A patient receiving treatment may take a first dose upon waking and then a second dose in the evening (if on a two dose regime) or at 3-or 4-hourly interv&s thereafter. Alternativ&y, a stow release device may be used to provide optimal doses of agents according to the invention to a patient without the need to administer repeated doses. Known procedures, such as those conventionally employed by the pharmaceutical industry (e.g. in vivo experimentation, clinical trials, etc.), may be used to form specific formulations of the agents according to the invention and precise therapeutic regimes (such as daily doses of the agents and the frequency of administration).

In a twelfth aspect of the invention, there is provided a pharmaceutical composition comprising a variant IgE Fc region according to the first aspect, or an antibody or fragment thereof according to the eighth aspect; and optionally a pharmaceutically

acceptable vehide.

The composition may be CD23-mediated side effect treatment composition. The term "CD23-mediated side effect treatment composition" can mean a pharmaceutical Jo formulation used in the therapeutic amelioration, prevention or treatment of a CD23-mediated side effect.

The invention also provides in a thirteenth aspect, a process for making the composition according to the tenth aspect, the process comprising combining a therapeutically effective amount of a variant IgE Fc region according to the first -22 -aspect, or an antibody or a functional fragment thereof as defined in the eighth aspect, with a pharmaceutically acceptable vehicle.

A "subject" may be a vertebrate, mammal, or domestic animal. Hence, medicaments according to the invention maybe used to treat any mammal, for example livestock (e.g. a horse), pets, or may be used in other veterinary applications. Most preferably, the subject is a human being.

A "therapeutically effective amount" of the antibody or fragment thereof is any /0 amount which, when administered to a subject, is the amount of agent that is needed to treat CD23-mediated side effect, or to produce the desired effect.

For example, the therapeutically effective amount of antibody or fragment thereof used may be from about 0.001 ng to about 1 mg, and typically from about 0.01 ng to /5 about too ng. It is preferred that the amount of antibody or fragment is an amount from about 0.1 ng to about 10 ng, and most preferably from about 0.5 ng to about 5 ng.

A "pharmaceutically acceptable vehicle" as referred to herein, is any known compound or combination of known compounds that are known to those skilled in the art to be useful in formulating pharmaceutical compositions.

In one embodiment, the pharmaceutically acceptable vehide may be a solid, and the composition may be in the form of a powder or tablet. A sobd pharmaceuticafly acceptable vehide may include one or more substances which may also act as flavouring agents, lubricants, sohibiBsers, suspending agents, dyes, fillers, glidants, compression aids, inert binders, sweeteners, preservatives, dyes, coatings, or tablet-disintegrating agents. The vehicle may also bean encapsulating materiaL In powders, the vehicle is a findy divided solid that is in admixture with the finely divided active agents according to the invention. In tablets, the active agent may be mixed with a vehicle having the necessary compression properties in suitable proportions and compacted in the shape and size desired. The powders and tablets preferably contain up to 99% of the active agents. Suitable solid vehicles include, for example calcium phosphate, magnesium stearate, talc, sugars, lactose, dextrin, starch, gdatin, ceBulose, polyvinylpyrroBdine, low melting waxes and ion exchange resins. In another embodiment, the pharmaceutical vehicle may be a gel and the composition may be in the form of a cream or the like.

-23 -However, the pharmaceutical vehicle may be a liquid, and the pharmaceutical composition is in the form of a solution. Liquid vehicles are used in preparing solutions, suspensions, emulsions, syrups, elixirs and pressurized compositions. The active agent according to the invention may be dissolved or suspended in a pharmaceutically acceptable liquid vehicle such as water, an organic solvent, a mixture of both or pharmaceutically acceptable oils or fats. The liquid vehicle can contain other suitable pharmaceutical additives such as solubilisers, emulsifiers, buffers, preservatives, sweeteners, flavouring agents, suspending agents, thickening /0 agents, colours, viscosity regulators, stabilizers or osmo-regulators. Suitable examples of liquid vehicles for oral and parenteral administration include water (partially containing additives as above, e.g. cellulose derivatives, preferably sodium carboxymethyl cellulose solution), alcohols (including monohydric alcohols and polyhydric alcohols, e.g. glycols) and their derivatives, and oils (e.g. fractionated /5 coconut oil and arachis oil). For parenteral administration, the vehicle can also be an oily ester such as ethy' oleate and isopropyl myristate. Sterile liquid vehicles are useful in sterile Uquid form compositions for parenteral administration. The Bquid vehicle for pressurized compositions can be a halogenated hydrocarbon or other pharmaceutically acceptable propellant.

Liqtud pharmaceutical compositions, which are sterile solutions or suspensions, can be utilized by, for example, intramuscular, intrathecal, epid ural, intraperitoneal, intravenous and particularly subcutaneous injection. The agent may be prepared as a sterile solid composition that may be dissolved or suspended at the time of administration using sterile water, saline, or other appropriate sterile injectaNe mcdiii m.

The agents and compositions of the invention may be administered orafly in the form of a sterile solution or suspension containing other s&utes or su spending agents (for examp'e, enough saline or glucose to make the solution isotonic), bile salts, acacia, gelatin, sorbitan monoleate, polysorbate 8o (oleate esters of sorbitol and its anhydrides copolymerized with ethylene oxide) and the like. The agents used according to the invention can also be administered orally either in liquid or solid composition form. Compositions suitable for oral administration include solid forms, such as pills, capsules, granules, tablets, and powders, and Uquid forms, such as solutions, syrups, elixirs, and suspensions. Forms useful for parenteral administration include sterile solutions, emulsions, and suspensions.

-24 -It will be appreciated that the invention extends to any nucleic acid or peptide or variant, derivative or analogue thereof, which comprises substantially the amino acid or nucleic acid sequences of any of the sequences referred to herein, including functional variants or functional fragments thereof. The terms "substantially the amino acid/nucleotide/peptide sequence", "functional variant" and "functional fragment", can be a sequence that has at least 40% sequence identity with the amino acid/nucleotide/peptide sequences of any one of the sequences referred to herein, for example 40% identity with the sequence identified as SEQ ID No:2 (i.e. wild type /0 human IgE Fc region) or the nucleotide identified as SEQ ID No:i (i.e. encoding wild type human 1gB Fc region), or 40% identity with the polypeptides identified as SEQ ID No:3-7 (i.e. 1gB Fc variant regions) or the nucleotides identified as SEQ ID No:8- 12 (i.e. encoding 1gB Fc variant regions), and so on.

/5 Amino acid/polynucleotide/polypeptide sequences with a sequence identity which is greater than 50%, more preferably greater than 65%, 70%, 75%, and still more preferably greater than 8o% sequence identity to any of the sequences referred to are also envisaged. Preferably, the amino acid/polynucleotide/polypeptide sequence has at least 85% identity with any of the sequences referred to, more preferably at least 90%, 92%, 95%, 97%, 98%, and most preferably at least 99% identity with any of the sequences referred to herein.

The skilled technician wifi appreciate how to calculate the percentage identity between two amino acid/polynudeotide/polypeptide sequences. In order to caktulate the percentage identity between two amino acid/polynucleotide/polypeptide sequences, an alignment of the two sequences must first be prepared, followed by cakifiation of the sequence identity value. The percentage identity for two sequences may take different values depending on:-(i) the method used to align the sequences, for exampk, ClustalW, BLAST, FASTA, Smith-Waterman (impkmented in different programs), or structural alignment from 3D comparison; and (ii) the parameters used by the alignment method, for example, local vs global alignment, the pair-score matrix used (e.g. BLOSUM62, PAJ\'125o, Gonnet etc.), and gap-penalty, e.g. functional form and constants.

Having made the alignment, there are many different ways of cakulating percentage identity between the two sequences. For example, one may divide the number of identities by: (i) the length of shoitest sequence; (ii) the length of alignment; (iii) the -25 -mean length of sequence; (iv) the number of non-gap positions; or (iv) the number of equivalenced positions excluding overhangs. Furthermore, it will be appreciated that percentage identity is also strongly length dependent. Therefore, the shorter a pair of sequences is, the higher the sequence identity one may expect to occur by chance.

Hence, it will be appreciated that the accurate alignment of protein or DNA sequences is a complex process. The popular multiple alignment program ClustaIW fhompson et a!., 1994, Nucleic Acids Research, 22, 4673-4680; Thompson et a!., 1997, Nucleic Acids Research, 24, 4876-4882) is a preferred way for generating /0 multiple alignments of proteins or DNA in accordance with the invention. Suitable parameters for ClustaIW maybe as follows: For DNA alignments: Gap Open Penalty = 15.0, Gap Extension Penalty = 6.66, and Matrix = Identity. For protein alignments: Gap Open Penalty = 10.0, Gap Extension Penalty = 0.2, and Matrix = Gonnet. For DNA and Protein alignments: ENDGAP = -1, and GAPDIST = 4. Those skilled in the /5 art will be aware that it may be necessary to vary these and other parameters for optimal sequence alignment.

Preferably, calculation of percentage identities between two amino acid/polynucleotide/polypeptide sequences may then be calculated from such an alignment as (N/T)ioo, where N is the number of positions at which the sequences share an identical residue, and T is the total number of positions compared including gaps but excluding overhangs. Hence, a most preferred method for cakidating percentage identity between two sequences comprises (i) preparing a sequence alignment using the CthstalW program using a suitabk set of parameters, for exampk, as set out above; and (ii) inserting the values of N and T into the foflowing formida:-Sequence Identity = (N/T)ioo.

Alternative methods for identifying similar sequences will be known to those sHIed in the art. For exampk, a substantially similar nucleotide sequence will be encoded by a sequence which hybridizes to the sequences shown in SEQ ID No's: 1, 8-12 or their complements under stringent conditions. By stringent conditions, we mean the nucleotide hybridises to filter-bound DNA or RNA in 3x sodium chloride/sodium citrate (SSC) at approximately 45°C followed by at least one wash in o.2x SSC/o.i% SDS at approximately 20-65°C. Alternatively, a substantially similar polypeptide may differ by at least 1, but less than 5, 10, 20,50 or 100 amino acids from the sequences shown in SEQ ID No: 2,3-7.

-26 -Due to the degeneracy of the genetic code, it is dear that any nucleic acid sequence described herein could be varied or changed without substantially affecting the sequence of the protein encoded thereby, to provide a functional variant thereof.

Suitable nucleotide variants are those having a sequence altered by the substitution of different codons that encode the same amino acid within the sequence, thus producing a silent change. Other suitable variants are those having homologous nucleotide sequences but comprising all, or portions of, sequence, which are altered by the substitution of different codons that encode an amino acid with a side chain of similar biophysical properties to the amino acid it substitutes, to produce a /0 conservative change. For example small non-polar, hydrophobic amino acids include glycine, alanine, leucine, isoleucine, valine, proline, and methionine. Large non-polar, hydrophobic amino acids include phenylalanine, tiyptophan and tyrosine. The polar neutral amino acids include serine, threonine, cysteine, asparagine and glutamine. The positively charged (basic) amino acids include lysine, arginine and is histidine. The negatively charged (acidic) amino acids include aspartic acid and glutamic acid. It will therefore be appreciated which amino acids may be replaced with an amino acid having similar biophysical properties, and the skilled technician will know the nucleotide sequences encoding these amino acids All of the features described herein (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined with any of the above aspects in any combination, except combinations where at least some of such features and/or steps are mutually exclusive.

For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying diagrammatic drawings, in which:-Figure ishows NMR mapping of the CD23 and 1gB interaction surfaces. (A) A small number of residues from Cc3 show chemical shift perturbation upon addition of derCD23. Increasing amounts of unlabelled derCD23 were added to a 2ooLM sample of l5N4abelled Cc3; five spectra of the titration are overlaid (red, zero derCD23; magenta, 50 tiM; blue, 100 kiM; cyan, 150 pM; green, 200 pM). Insets show magnified views of the indicated regions. (B) The NMR-derived derCD23 interaction site on Cs3 was mapped onto the structure of IgE-Fc (iF6A; Garman et al., 2000, Nature 406:259-266) and shown as surface representation. For comparison, the -27 -residues of IgE that interact with FcERI are indicated in green. (C) The IgE interaction surface on CD23 was defined previous'y (Hibbert eta]., 2005,JErp Med 202:751-760) and shown here as a surface representation. The interacting surfaces of 1gB and CD23 are co]oured according to ekctrostatic potentia' and coded such that regions with a potentia' <-4 k11T are red, whereas those >4 k11T are blue (ku, Boltzmann constant; T, absolute temperature). The IgE-CD23 interaction site is formed by comp'ementary charged surfaces; Figure 2 shows competition binding experiments between derCD23 and sFccRla for /0 IgE-Fc. (A-D) Surface plasmon resonance and (B-F) TR-FRET-based competitive binding experiments were used to investigate the mechanisms of receptor interactions. The binding of derCD23 was tested against (A) IgE-Fc immobilized on a sensor surface and (B) IgE-Fc captured on an FcR1a-immobilized surface; the start of the derCD23 injection is indicated with an arrow. DerCD23 binding to / s immobi]ized IgE-Fc with a K of 2.3 pM; no measureab]e binding was observed for derCD23 to IgE-Fc complexed to FcaRIa. (C) The binding of IgE-Fc to immobilized derCD23 was compared with (D) the binding of a complex of IgE-Fc/sFcsRIft to the same surface; the start of the injection of the complex is indicated with an arrow.

IgE-Fc binds to derCD23 with a Ku of 2.4 pM, but the IgE-Fc/sFcsRlct complex does not bind to derCD23. AU SPR binding experiments were performed using identica' two-fold serial dilutions of ligands, from 4° pM to 78nM. (B) Binding between terbium-labelled derCD23 and Alexa 647-labelled IgE-Fc was measured in a solution TR-FRET assay in the presence of increasing concentrations of the ay-fusion protein (the FcERIU extracellular region fused to an IgG4 Fc) (Shi et al., 1997, Biochemistry, 36, 2112-2122); zero (bthck), o.nM (red), 2.5nM (b]ue) and nM (green). (F) These data are a'so shown as a double reciproca] p]ot to illustrate that addition of the inhibitor results in a decrease in the apparent binding maximum (the va]ue at X=o), rather than a change in the apparent binding affinity (the value at Y=o); this is characteristic of aflosteric inhibitors; Figure 3 shows that so]ub]e CD23 does not cross-]ink 1gB bound to FcPJ on mast cells. The ability of soluble CD23 to engage 1gB on B cells and mast cells was tested.

(A) After preincubation of IgE, the addition of anti-IgE antibody resifits in activation of the FcaRIa LAD-2 mast cell Une, as measured by rdease of-hexosaminidase.

Neither monomeric derCD23 nor trimeric triCD23 is able to cross-link 1gB and activate mast cells in this assay. (B) In contrast, triCD23 effectively cross-links mIgE -28 -on the surface of 1gW human tonsiflar B cells, resulting in activation of these cefis and increased secretion of IgE. B cell cultures were activated with IL-4 and anti-CD40, and soluble CD23 was added at 1 jtM; supernatants were harvested 12 days after activation, tested for 1gB levels and compared with lev&s for ces treated with IL-4/anti-CD40 alone (*=p<o.os; **=pco.ol); Figure 4 shows the temperature dependence and ionic strength dependence of the interaction between IgE-Fc and derCD23. (A) The temperature dependence of the IgE-Fc and derCD23 interaction is demonstrated by p'otting the natural log of the Jo equihbrium constant as a function of the inverse of absohite temperature (the vant Hoff p'ot). A linear fit of the data is shown (red line), and 95% confidence evels indicated (dashed lines). The slope of this plot estimates the enthalpic changes associated with binding. At a temperature of 298K, the thermodynamic parameters of binding are as follows: AG= -32.2 (± 0.5) kJ moF1, AH= -5.0 (± 0.3) moP, -TAS= - / 5 27.2 kJ moP; these values are consistent with a strong electrostatic component to binding energetics. (B) The effects of ionic strength on the binding of IgE-Fc to derCD23. Binding affinities are measured over a range of ionic strengths and p'otted as log (Ks) versus log (ionic strength). The steep s'ope (-0.94) of the curve confirms a strong dectrostatic contribution to binding; Figures shows the structure of the derCD23-Fc3-4 complex. The two molecules of derCD23 (light and dark blue Ca traces with surfaces) bind one to each heavy chain between the Cs (dark red and green) and C4 domains (light red and green). The carbohydrate is shown in all-atom representation (red and yellow, without surfaces) and can be seen behind the (red) Ce3 domain. The adjacent N-and C-termini of each derCD23 m&ecule, the former being the connection to the "stalk", the latter to the "taB" region, can be seen at the extreme left and nght of the figure. (The complex shown here comprises chains A (red), B (green), G (light blue) and H (dark blue)); Figure 6 shows salt bridges and hydrogen bonds at the derCD23-Fc23-4 interface.

The H-bonds associated with the four salt bridges are shown in red, additional H-bonds present in all six independent interactions are shown in green, and a further H-bond present in /6 mcilecules is shosrn in yellow; Figure 7 shows composite images of the derCD23 and sFccRla complexes with FcE3-4 to show the mutual incompatibility of their binding modes. a) derCD23-FcE3- -29 - 4 and sFcRIa-Fc3-4 [PDB iF6A] complexes superposed on their (C4)2 domain pairs. The receptors are shown as surfaces (derCD23 light & dark blue; sFcaRla red) and the FcE3-4 structures are shown as Ca traces (in corresponding colours). The closed (derCD23-binding) and open (sFcrRla-binding) conformations of the CE3 domains maybe seen. b) Steric clashes between the sFcRIa structure (red Ca trace) and both chains of the derCD23-Fca3-4 complex (blue) are indicated (orange surfaces). c) Steric clashes of both derCD23 molecules (blue Ca traces) with the sFcERIa-FcE3-4 complex (red) are indicated (green surfaces); Figure 8 shows that sFceRla displaces derCD23 and vice versa in a FRET inhibition assay. a) Competition of derCD23 binding to IgE-Fc (blue) and FcE3-4 (red) by unlabeled sFcaRla. b) Competition of sFcaRla binding to IgE-Fc (blue) and Fce3-4 (red) by unlabeled derCD23. (All data: n=3; error bars ± SEM); Figure 9 shows electron density in the interface region. Electron density 2F0 -map contoured at la for key side-chains involved in salt bridges at the derCD23/Fcc3-4 interface. The side-chains shown in blue are from derCD23, those in purple from Fc3-4; Figure ma) The six independent FcE3-4 heavy chains superposed on their CE4 domains, showing the variation in the relative positions of the Ca3 domains. b) The six independent derCD23 molecules superposed, showing virtually no overall structural variation; Figure ii shows a comparison of FcE3-4 heavy chains taken from various FcE3-4 and IgE-Fc structures superposed by their CE4 domains upon the set of six independent heavy chains of the derCD23-Fcc3-4 complex (shown lightly coloured), a) Free IgE-Fc (light and dark red) [PDB 2WQR; 28], b) sFccRlcz-IgE-Fc complex (light and dark cyan) [PDB 2Y7Q; 21], c) sFccRIa-FcE3-4 complex (light and dark purple) [PDB iF6A; 20], d) Single chain of free Fcc3-4 (green), the most closed structure observed to date [PDB 3HAO; 27]; and Figure 12 shows the superposition of apo-and calcium-bound derCD23 structures.

The apo-CD23 (dark blue) and calcium-bound CD23 (dark red) structures [PDB 2H2R, 2H2T; 40] superposed upon the six derCD23 molecules from the FcE3-4 -30 -complex (lightly coloured), showing differences in loops 1 and 4. The Ca2E ion is shown (red sphere); note the disordered loop 4 between residues Ser252 and Asp258 in the calcium-bound structure.

Examples

Example 1

Materials and Methods Protein expression and purification: Recombinant human IgE-Fc (comprising /0 domains Cs2-Cs4)(Wan et al., 2002, Nat Immunol 3:681-686), the cry-fusion protein (the FcERIU extracelhdar region fused to an IgG4 Fc) (Shi et at, 1997, Biochemistry 36, 2112-2122), soluble FcaRlct (Price et al., 2005, J Biol Chern 280, 2324-2330), derCD23 (Hibbert et at, 2005, JExp Med 202, 751-760) and the Cc3 domain (Price et al., 2005, J Biol Chem 280, 2324-2330) were each produced and purified as / 5 described previous'y. Primers for site-directed mutagenesis were obtained from Sigma-Genosys (Sigma Lifescience), and mutagenesis was performed using the QuikChange II Kit (Stratagene) according to the manufacturer's instructions. Mutant constructs were expressed and purified using the same methods as the wildtype proteins. MAb 7.12 was produced from a B cell hybridoma (Kanowith-Klein et at, 1988, C/in immunol Immunopathol 48, 214-224), and was a kind gift of Dr. Rebecca Beavil (King's College London, UK).

NMR spectroscopy: NMR spectroscopy was performed on protein samples in a buffer containing 25mM Tris, 125 mM NaCl, 4mM Cad2, pH 6.8, at protein concentrations between 120 and 900 tM. Data were collected at 25°C on Bruker spectrometers equipped with CryoProbes operating at oo and 700 MHz. For chemic& shift perturbation experiments, unlabelled derCD23 ligands were concentrated to 2mM and then added in small aliquots to samples of 200 jiM 15N..labelled CE3 until saturation was seen. The NMR chemical shifts of the urea denatured and native state Cs3 domain are available from the BioMagResBank database under accession numbers 18482 and 18483.

Surface plasmon resonance: All experiments were performed on a Biacore Tioo instrument (GE Healthcare), essentially as described previously (Hibbert et al., 2005, is JExp Med 202, 751-760; Holdomet al., 2011,NatStructMo/BiollS, 571-576).

Specific binding surfaces were prepared using standard amine coupling methods for -31 -derCD23 and the ay-fiision protein, whereas IgE-Fc was biotinylated (one biotin per molecule) and captured on a streptavidin surface. CoupBng densities were kept low (coo RU) to minimize potential andity effects. Ligands in HBS (10mM HEPES, pH 7.4, 150mM Nacl, 4mM Cad2, 0.005% surfactant p20) were injected at 25 gl/min with a 1-mm association phase followed by a 15-mm dissociation phase.

Thermodynamic measurements were performed at 288K, 293K, 298K and 303K.

For the sandwich binding experiments, approximately 9ORU of IgE-Fc was captured on an ay-fusion protein surface during a 1-mm injection of a ionM IgE-Fc sample; after a 3-mm stabilization period, o-iooi1\'l derCD23 was injected for 1-mm followed Jo by a 15-mm dissociation phase. All measurements were done independenfly at least twice, using standard double reference subtraction methods for data ana'ysis (Myszka, 1999, JMoI Recogrzit 12, 279-284). Curve fitting and other analyses were performed using MicroCal Origin 8.o software (OriginLab Corporation).

/ 5 FRET assay: Inhibition assays were performed by competing 1LM terbium-chethte labelled derCD23 and o-2opjM Alexa 647-labelled IgE-Fc with a range of concentrations of unlabelled cry-fusion protein. Protein mixtures were prepared in Lanthascreen buffer (Invitrogen) in triplicate, in 384-well plates (Greiner Bio-One), and equilibrated overnight at room temperature. FRET measurements were made on an Artemis plate reader (Berthold Technologies). TR-FRET ratios were calculated for each well as the emission of acceptor at 665nm divided by the emission of donor at 62onm and then multiphed by 10,000. Apparent}(D and Brnax values were derived from non-linear curve fitting of inhibition titrations, and data are presented as double reciprocal plots to highlight the differences in apparent Bniax values upon addition of the inhibitor.

Mast cell degranulation assay: The human mast cell line LAD-2 (NIH) was cultured in Stem Pro-34 SFM medium (Gibco) supplemented with L-glutamine (2mM), penicillin (to U/ml), and streptomycin (i pg/ml; all from Life Technologies). LAD-2 Jo cells were suspended in culture medium at i.6x1o6 cells/ml and 50111 aliquots were placed in v-bottomed 96-we pthtes (Greiner Bio-One). Cefis were primed by addition of 2.5nM IgE (NIBSC) or a buffer-only control for one hour, before addition of cross-linking reagents. Polyclonal rabbit anti-human IgE (Dako) was added at 2onM and soluble CD23 constructs at 0.1, 1 and lotM, and incubated for 1 hour at 37°C. Supernatants were harvested and tested for -hexosaminidase rdease, measured fluorometrically as described previously (Hammond, 2006, In Cell Biology, -32 -J.E. Cebs, editor. Elsevier. Amsterdam). Controls inchided cells treated with wash buffer plus i% Triton-X for total release, with buffer-only to measure background release, typically about io% of total release, with 2.5nM IgE-only, and with 1o1iM CD23-only. The level of degranulation measured for Triton-X treated cefis was defined as ioo% release and all samp'es were defined relative to that.

B cell activation assays: Human tonsillar B cells were cultured in 24-well plates (Nunc) at 5 x ios cells/ml in RPMI with penicillin (ioolU/ml), streptomycin (loog/ml), glutamine (2mM) (Invitrogen), io% FCS (Hyclone, Perbio Biosciences io Ltd), insulin (51g/ml) and transferrin (35pg/ml) (Sigma-Aldrich). Cells were activated with IL-4 (200IU/ml) (R&D Systems), anti-CD40 antibody (ig/mfl (G28.5; ATCC), and either ltM derCD23 or 1pM triCD23. Supernatants were harvested on day 12 for IgE measurements. IgE ELISA assays were performed as described previously (Cooper et al., 2012, J Immunol 188, 3199-3207), using is Maxisorp pbtes (Nunc) coated with polyclonal mouse anti-human IgE (1:7000) (Dako), in pH 9.8 carbonate buffer (o.2M Na2CO3, o.2M NaHCO3). Unbound sites were blocked with 2% mHk powder in PBS + 0.05% Tween2o (Sigma-Aldrich).

Binding was detected by mouse anti-human IgE-HRP (1: iooo) (Dako) in i% milk powder in PBS-T for 2 hours at 37°C. Standard curves were derived using human serum IgE (NIBSC), with a minimal detectable concentration of about 2ng/ml.

Results and discussion In an earlier study, the inventors identified the IgE binding site on CD23 using NMR chemical shift perturbation studies (Hibbert et al., 2005, J Exp Med 202, 751-760).

In this work, the inventors have now performed the reciprocal NMR binding experiment, mapping the interaction site of CD23 onto the Cs3 domain from IgE.

Using an approach described by Schiilman eta!. (Schulman et aL, 1997, Nat Stnwt Biol 4, 630-634), they then assigned the backbone resonances of the molten globule Cs3 domain by first performing resonance assignments of Cs3 denatured in 6M urea and then, through gradual titration of buffer conditions, tracking those resonances to the native state Cs3 domain. Next, the inventors titrated unlabelled monomeric CD23 protein (derCD23) against an SN-labelled CE3 sample and used the assigned NMR spectra to identi residues that were affected by addition of Ugand. A small number of C3 residues showed peak shifting and line broadening during the derCD23 titration, consistent with an interaction characterized by fast/intermediate exchange kinetics (see Fig. iA).

--

When mapped onto the surface of the protein, the identified residues from three discontinuous sequences (amino acids 405-407, 409-411 and 413 from the E-F helix of the CE3 domain of IgE, amino acids 377-380 from the C-D loop, and residue 436 from the C-terminal region) form a contiguous surface representing the binding site on Cs3 for CD23 (Figure 1B). This region is at the end of the CE3 domain, near to the interface with C84, in contrast to the interaction site for FcsRJ, which is at the other end of CE3 near the interface with CE2 (Garman et al., 2000, Nature 406, 259-266; Holdom et a]., 2011, J'Jat Struct Mo! Bio! i8, 571-576) (Fig. iB). /0

The identification of this CD23 interaction site on IgE provides a structural explanation for the experimentally observed 2:1 (CD23:IgE) stoichiometry, as the dimeric IgE-Fc can bind to two separate CD23 lectin head domains. The two CD23 interactions were shown to have slightly different binding affinities and /5 thermodynamic characteristics, as was also observed for the FcaRI-IgA interaction.

The two binding affinities imply an asymmetry of the two CD23 binding sites, which may possibly be aBostericay induced. CD23's capacity for inducing a conforrnationa change in IgE is discussed further below.

In support of the NMR-based interaction site mapping, the inventors then ana'ysed the binding characteristics of the IgE-CD23 interaction and performed extensive site-directed mutagenesis of residues at the binding interface. Surface plasmon resonance (SPR) based measurements of binding theniiodynamics indicate a1 interaction with a strongly favorable entropic contribution (at 298K, AG = -32.2 (±0.5) kJ mol-1, AH =-.o (±0.3) kJ moF', -TAS= -27.2 Id mol) (Figure 4A). Binding experiments performed over a range of different ionic strengths gave a measure of the contributions of electrostatic forces to the overall binding energies; a plot of log(KA) versus log(ionic strength) shows an unusually steep slope of -0.94 (Fig. 4B) (in comparison, the same analysis of the IgE-Fc-sFccRla interaction gives a slope of -0.13). Both of these measurements are consistent with the IgE-CD23 binding event being strongly electrostatic in character.

Following the NMR mapping of interaction epitopes for both proteins, the inventors used site-directed mutagenesis to validate the interaction site in the context of the is full IgE-Fc construct and to define the energetic contributions of individual residues.

Ten mutants from derCD23 and eleven mutants from IgE-Fc (domains Cc2-4) were -34 -produced, purified and characterized; their binding affinities were measured using an SPR assay (Hibbert et al., 2005). Table 1 summarizes the results of the site-directed mutagenesis studies.

Table 1 -Effects of mutations on the IgE-CD2 interaction IgE-Fc Ku AAG derCD23 Ku AG mutation (tiM) (kJ mo11) mutation (LM) (U mo11) wildtype 2.3 -wildtype 2.3 -D409A 26.3 +6.o D227A 30.9 +6.4 E412A 24.2 +.8 E257A 26.7 +6.i R376A 19.7 + R224A 26.2 +6.o K38oA 13.3 +4.3 Ri88A 25.0 +5.9 K435A 5.0 +1.9 Yi8gA 15.6 +4.7 K352A 3.8 +1.2 K276A 10.9 +3.9 R351A 2.6 +0.3 L226A 6.6 +2.6 D347A 2.5 +0.2 D236A.8 +2.3 P439A 2.5 +0.2 D192A 4.3 +i.6 Q535A 2.5 +0.2 E265A 2.5 +0.2 Q538A 2.4 +0.1 Mutations on both proteins that affect binding are entirely consistent with the NMR-defined interaction sites. Charged residues have the largest energetic effect on binding. CD23 mutations D227A, E257A, R224A and Ri88A all show a change in binding free energy (AAG) of about +6 kJ moF' (Table i). Uncharged residues also contribute to the binding energy; a prominently exposed tyrosine residue Y189) in the center of CD23's IgE binding site makes a substantial contribution to binding energy. The CD23 binding surface on IgE is also predominantly dectrostatic, with residues D4o9, E412, R376 and K38o showing the largest effects on CD23 binding energetics.

An earlier study had implicated IgE residue K352 from the A-B helix as part of the binding interface. However, the inventors' NMR perturbation experiments indicate that the A-B helix is near the site of interaction, but in the assay reported herein that directly measures binding affinity, a KjzA mutation had only a very small effect on CD23 binding (see Table 1). The high-resolution crystal structure of the IgE-CD23 complex indicates that 1(352 is not part of the binding site. The inventors postulate -35 -that it is likely that the effects of the non-conservative K352E mutation previous'y observed resulted from local conformational changes rather than a direct effect from 1(352 on CD23 binding. Because the NMR data indicated a site on Cs3 very near to the C4 interface (Fig. iB) and because binding sites from several other immunoglobulin-receptor interactions involve sites analogous to the C83-Cs4 interface, the inventors also made a pair of mutations in the F-G loop of the Cs4 domain, close to the CD23 binding site in Cs3. However, neither Q535A nor Q538A appear to affect CD23 binding, leading the inventors to believe that CD23's binding surface on IgE is largely restricted to residues from Cc3. /0

Earlier studies indicated that soluble CD23 can compete with FcsRI binding, and this was attributed to steric competition for an overlapping binding site within the Cs3 domain. However, the data described herein show that the CD23 and FccRI binding sites are spatially distinct and suggest that the mechanism of mutual inhibition must /5 be allosteric in nature. The inventors performed a set of competitive binding assays to confirm this experimentally. Firstly, using an SPR assay, they showed that derCD23 can bind to IgE-Fc immobilized to an SPR chip but cannot bind to IgE-Fc captured by immobilized FccRIa (Fig. 2A,B); a positive control, a Fab fragment of the anti-IgE antibody 7.12, directed against the Cs2 domain, did bind to FcsRTcz-captured IgE-Fc (data not shown). Secondly, they showed that IgE-Fc can bind to immobilized derCD23, but an IgE-Fc-sFccRla complex cannot bind to derCD23 (Fig. 2C,D). These data indicate that CD23 and FccRI interactions with IgE are mutually inhibitory.

Finay, they also tested the ability of the receptors to compete for binding to IgE in a solution TR-FRET experiment (Selvin, 2002, Arinu Rev Biophys Biornol Struct 31, 275-302). This assay can be performed under equilibrium binding conditions, allowing a different set of mechanistic properties to be tested than in the kinetic SPR experiments. Under eqniHbrium conditions, different inhibition patterns are observed for competitive and aflosteric inhibitors. A competitive inhibitor affects the apparent binding affinity, with inhibitor I reducing the apparent affinity by a ratio of Jo (i+[I]/Ki); whereas an allosteric inhibitor affects the apparent Bniax of the interaction without changing the apparent Ku (Fersht, 1999, Structure and Mechanism in Protein Science. New York, W.H. Freeman.). Using Fcr.RIa as the inhibitor of the IgE-Fc-derCD23 interaction resulls in a decrease of apparent Bnax without affecting the apparent I(f) of the interaction (Fig. 2E,F). These experiments confirm mutual inhibition by the two IgE receptors and offer experimental evidence that an aflosteric mechanism is inv&ved.

-36 -Given the location of the CD23 binding site, the most obvious mechanism for allostery is a conformational change around the interface between the Cs3 and Cr4 domains. Crystal structures of IgE-Fc and IgE-Fc-FcRIa comp'exes indicate that the Ca domains can exist in "open" and "closed" states, with only an open state being capable of binding FccRI. A detailed study of the open and closed states concluded that it is the motions around the Cs3 A-B helix, sifting at the Ca3-Ca4 interface, that control the orientation of the two Cc3 domains.

/0 Soluble trimeric CD23 has been shown to bind to and cross-link membrane IgE on B cells, resulting in B cell activation. However, it is essential that trimeric CD23 not cross-link IgE bound to FcsRl on the surface of mast cells. If this were to occur then high evels of CD23 would result in mast cefi activation in the absence of allergens.

The data from binding experiments (Fig. 2B) predict that s&uble CD23 cannot / 5 directly cross-Bnk IgE bound to FcRI on mast ces. The inventors tested this prediction in a mast cell degranulation assay using the FcsRP LAD-2 human mast cell line. In this assay, cells are first primed by adding IgE, followed by addition of potential cross-linking reagents and measurement of release of the mast cell granule- associated enzyme 13-hexosaminidase. An anti-IgE antibody results in FcgRI-mediated activation of the mast cefi and robust 13-hexosaminidase release, but the addition of either the monomeric derCD23 or a trimeric CD23 construct (triCD23) fails to induce mast cell degranulation (Fig. 3A). In contrast, trimeric CD23 effective'y cross-links IgE on B ces, resulting in activation of these cells and increased production of s&uble IgE (Fig. 3B).

Immunoglobufins have ev&ved two separate sites for binding to receptors. One site is near the hinge region in IgG and at the Cc2-Cs3 interface in IgE, while the other is at the interface of the C-terminal domain and the penultimate domain: the Cr3-Cr4 interface in IgE. A mechanism of communication has evolved within the IgE Jo molecule between these two distant sites to prevent simultaneous engagement of CD23 and Fcr.RI. This may be a unique property of IgE. Since IgE and CD23 both exist in membrane bound and soluble forms, and s6luble FcRIa has also recently been shown to exist at functionally relevant concentrations, there is considerable potential for receptor cross-regulation. Mutually exclusive receptor binding assures independent functions for IgE-FccRI and IgE-CD23 interactions.

-37 -IgE is a clinicafly important drug target. An anti-IgE antibody (omalizumab) is an effective therapy, currently used in the treatment of moderate to severe asthma that is not controlled by corticosteroids. Omalizumab binds to the Ca3 domain of IgE and competitively inhibits FcRI binding, although its in vivo activity rehes on more than just inhibition of this interaction. Results presented here demonstrate that IgE is amenable to allosteric inhibition, an approach that may have significant advantages over competitive inhibition, and thy the foundation for the devdopment of allosteric modulators of IgE-receptor interactions.

/0 Summary

IgE, the antibody that mediates allergic responses, acts as part of a self-regulating protein network. Its unique effector functions are controed through interactions of its Fc region with two ceihilar receptors, FcsRI on mast cells and basophils and CD23 on B cells. IgE cross-linked by allergen triggers mast cell activation via FccRI, while / 5 IgE-CD23 interactions contral IgE expression evels. The inventors have determined the CD23 binding site on IgE, using a combination of NMR chemical shift mapping and site-directed mutagenesis. They have shown that the CD23 and FcERI interaction sites are at opposite ends of the Cs3 domain of IgE, bitt that receptor binding is mutually inhibitory, mediated by an allosteric mechanism. This prevents CD23-mediated cross-linking of 1gB bound to FccRI on mast cells and resulting antigen-independent anaphylaxis. The mutually inhibitory nature of receptor binding provides a degree of autonomy for the individual activities mediated by IgE-FcslU and IgE-CD23 interactions.

It will be appreciated that by inhibiting binding between the Fc region on IgE and CD23 (which has been dearly demonstrated herein), one can be confident about inhibiting the downstream functional activity and avoiding CD23-mediated side effects, which are often experienced when using therapeutic IgE antibodies.

Examnle2 Materials and Methods Protein expression and purification: Recombinant human derCD23 (Ser156 -Glu298) was expressed, refolded and dialyzed into 25 mM Tris-HC1 pH 7.5 is (purification buffer') as previously described (Hibbert, 2005, 1. Exp. Med., 202, 751- 760). It was purified on a heparin-sepharose column (GE Healthcare) pre- -38 -equilibrated with purification buffer, ehiting with 25 mM Tris-HC1 pH 7.5, 200 mM NaC1. Fractions were pooled, concentrated to 1 ml and loaded onto a HiLoad 16/60 Superdex G75 column (GE Healthcare), pre-equilibrated and subsequently washed with purification buffer. Folding was assessed by iD-1H NMR at 500 MHz (large dispersion and strong signals of methyl groups between 1.0 and -1.0 ppm). Human IgE-Fc (N265Q, N371Q) was expressed in NS0 cells and purified by affinity chromatography with sFccRla-IgG4-Fc fusion protein as previously described (Shi, 1997, Biochemistry, 36, 2112-2122). The genes for recombinant human FcE3-4 (Cys328-Lys547, with N-terminal ADP) and sFcRIa-Cys-His (Vall-Lys176, with C-Jo terminal cysteine and His6 tag) were synthesized by DNA2.o and cloned as Hindlil/EcoRl fragments into proprietary mammalian expression vectors. The DNAs were transiently transfected into HEK293 cells using 293-fectin (Life Technologies) according to the manufacturers' instructions and the supernatants were harvested 6 days post transfection. Human FcE3-4 was purified by cation /5 exchange chromatography on a SPHP matrix (GE Healthcare) in 50 mM NaOAc buffer pH 6.0, followed by gel filtration on a Superdex S200 matrix (GE Healthcare) in PBS pH 7.4. Human sFcsRla.-Cys-His was purified on a Ni-NTA column (Qiagen) followed by gel filtration on a Superdex 5200 matrix (GE Healthcare) in PBS pH 7.4 and stored under nitrogen to prevent reactivity of the free cysteine residue.

Crystallization and data collection: Fcs3-4 was concentrated to 20 mg/ml, and derCD23 to i8 mg/mi, in 25 mM Tris-HCI pH 7.5, 20mM NaCI, o.o% sodium azide (crystallization buffer'). The complex was formed with 0.4 mM derCD23 (6.2 mg/mI), 0.2 mM Fca3-4 (io mg/ml) and 4 mM CaCI2, dikited with equal volume of 3% PEG 8,000, 0.1 M Tris-HC1 pH 7.5 as the precipitant. Crystals grew to -400 jim within 8 days at 295 K, were flash-cooled to looK (using 14% PEG 8,000, 0.1 M Tris-HC1 pH 7.5, 30% PEG 200) and data collected at beamlines 102 and 104, Diamond Light Source (Harwell, U.K.).

FRET assay: Labeling of the proteins is described in ST Materials and Methods.

Inhibition assays were performed by competing % of 1 jiM terbium labeled derCD23 and 5 % of 5 jiM Alexa Fluor 647 labeled IgE-Fc or Fc83-4 with a dilution series of unlabeled sFccRla-Cys-His. Assays were conducted in 384 well hi-base, white plates (Greiner BioOne) using Lanthascreen buffer (Invitrogen) as a diluent. The plate was left to incubate for 1 hr at room temperature with shaking and read by the Artemis plate reader (Berthold Technologies). TR-FRET ratios were then calculated for each -39 -w&l as the emission of acceptor at 665 nm divided by the emission of donor at 620 nm multiplied by 1o. Data were analyzed using GraphPad Prism 5. Similarly, 2 nM terbium labeled sFcsRla and 10 nM Alexa Fluor 647 labeled IgE-Fc or FcE3-4 were competed with a dilution series of unlabeled derCD23. Assays were conducted as described above with the exception of an overnight incubation. Positive controls for the two experiments were provided by displacement of labeled sFcrRla by unlabeled sFcERIcz, and labeled derCD23 by unlabeled derCD23; these defined the start-and end-point values for each titration, which were treated as zero and ioo% inhibition for calculation of the "% inhibition" values for displacement of one receptor by the /0 other (Figs. 8a &b).

Crystaflization: Crystals with a different morphology were also found after 6 days with 12% PEG 4,000, 0.1 M Tris-HC1 pH 8.75, 0.2 M sodium acetate trihydrate as precipitant after streak seeding. Diffraction data were collected to 3.6A and the / 5 crystal form was found to be the same as that reported in Table 2.

Crystallographic data collection and structure determination: Indexing, integration and merging of data to 3.iA res&ution was carried out with the HKL2000 suite of programs (Otwinowski, 1997, Methods EnzymoL, 276, 307-326). Nine previously determined Fcc3-4 dimers (Wurzburg & Jardetzky, 2009, J. Mo!. Bid., 393, 176-190) were used as search models in molecular replacement. However, due to the high degree of conformational flexibility of the two Fcs3-4 chains (Wurzburg & Jardetzky, 2009, 3. Mo!. Biol., 393, 176-190), fourteen additional dimers were manually generated from crystallographically independent chains and were also used as search models. Solutions were only found using the "closed" conformations of the Fc3-4 dimer. For example, using a dimer generated from chain A of PDB 3H9Z, two copies of the Fcc3-4 dimer were identified in the asymmetric unit using PHASER (MacCoy, 2007, J. App!. Crystallogr., 4°, 658-674). Subsequenfly, four derCD23 (PDB 2H2R) molecules were located. A third derCD23-FcE3-4 complex could only be found by searching with one of the previous'y located comp'exes.

Due to the high degree of NCS, reflections were selected for the Ri±. set in thin resolution shefis (Fabiola, 2006, Acta Grystallogr D, 62, 227-238) using SETOOLS (Winn, 2011, Acta Crystallogr D, 67, 235-242). Iterative cydes of refinement using PHENIX (Adams, 2011, Methods, 55, 94-106), REFMACS (Murshudov, 2011, Acta Crystalloyr D, 67, 355-367) and BUSTER-TNT (Bricogne, BUSTER-TNT, Global Phasing Ltd.) ahemated with manna] mode] building with COOT (Ems]ey, 2010, Aeta Crystallogr D,, 66, 486-501). To minimize bias, the model was built into zF0 -F composite omit, 2F0 -F and F0 -F electron density maps. Initially during refinement, tight NCS restraints were used. These were gradually relaxed, and finally Local Structure Similarity Restraints (LSSR or local NCS') (Murshudov, 2011, Acta Crystdllogr D, 67, 355-367) were applied. TLS groups (Painter, 2008, Acta Crystallogr D, 62, 439-450) were generated using the TLSMD web server (Painter, 2006, JAppi Crystallogr, 39, 109-111). Carbohydrate atoms and a single water molecule were subsequently incorporated into the structure. No electron density was /0 observed for FcE3-4 residues 363-364 of chain B; 367-371 and 421 -428 of chain C; 367-37°, 419 -422, 429 -430,479 -480 and 517-519 of chain E; and derCD23 residues 256 -257 of chains G, H, I, J and K. Therefore, these residues were not built into the model. Refinement statistics are shown in Table 2.

/s Labeling of proteins for FRET assay: IgE-Fc and Fcc3-4 were labeled with acceptor fluorophore by reacting 4 mg/rn] protein in 100 rnM sodium bicarbonate, so mM NaCI, pH 9.3, with a -fo]d molar excess of Alexa F]uor 647 succinimidy] ester (Invitrogen). After 1 hr incubation at room temperature with agitation, excess unreacted fluorophore was removed by dialyzing into HBS (in mM Hepes, 150 mM NaC1, 4mM CaC12, pH 7.4). derCD23 and the sFcsRla-Cys-His mutant were labeled with donor fluorophore. derCD23 at 3.5 mg/rn] in 100mM Hepes, 125 rnM NaCI, pH 8.3 was reacted with a -fo]d molar excess of terbium cholate isothiocyanate (Invitrogen) for a 3 hr incubation at mom temperature with agitation. Excess fluorophore was removed by dia]yzing into HBS. sFcsRIa-s-His at 5 rng/m in PBS (20 mM phosphate buffer saline, pH 7.4) was reacted with a io-fold molar excess of terbium chelate maleimide (Invitrogen) overnight at 4°C via thiol coupling.

Results Overa]] tono]ogy of the comn]ex The structure of the cornpex was determined at 3.iA resolution and revea's one derCD23 molecule bound to each heavy chain at the interface between C3 and CE4, making contact with residues from both domains (Fig. and Table 2; Fig. 9 shows the electron density map at the interface).

-41 -Table 2 -CIystMIogI'aDhic data collection and refinement statistics Data processing statistics Beamline 104, Diamond Light Source Wavelength (A) 0.9763 Space group P212121 Unit cell parameters (A) a = 62.89, b = 110.75, C = 376.56 Resolution range A) 50.0 -3.10 (3.21 -3.10) Observations 162,066 Unique reflections 48,290 Average redundancy 3.6 (4.0) Completeness (%) 92.8 (z') Wilson B factor (A2) 117 1/o(l) 14.5 (1.68) Rmeret o.o86 (o.88) Refinement statistics Resolution range (A) 29.9 -3.1 Total No. of reflections 48,257 No. of working reflections 45,820 No. of test reflections 2,437 R factort 0.2649 Rworking 0.2640 0.2831 No. of atoms Protein 16,107 Others 389 R.m.s. bond-length deviation (A) 0.041 R.m.s. bond-angle deviation (°) 1.29 Mean B factor A2) 111 Main chain 107 Side Chain ii Others 131 R.m.s. backbone B factor deviation 2.6 Ramachandran statistics (%)e Favored 92.9 Allowed 100 Outliers 0 -42 -Values in parentheses are for the outer resolution shell.

tRrnei.ge = EI1i, <1>1/ E<J> R factor = II F0(hlcL) I -I Fc'h/cL) II / E,,k(I F0(ikl) I Carbohydrate and one water molecule.

R.m.s. deviation between B factors for bonded main-chain atoms.

As defined by MolProbity (Chen, 2010, Acta Ciwstaiiogr D 66:12-21)..

The crystal form contained three independent copies of the comp'ex in the Jo asymmetric unit and the six heavy chains (labàled Ato F) bound to six derCD23 heads (labeled G to L) showed essentiafly identical modes of interaction, exemplified by the structure depicted (Fig. 5; chains A, B, G & H). The angle between the CE3 and Ca4 domains varies only slightly between the six heavy chains (Fig. ioa) and there are virtually no differences between the six derCD23 head domains (Fig. tob). All three /5 complexes thus dispthy approximate two-fold symmetry, although one (consisting of chains C, D, I & J) includes the two "extremes" in terms of the angle between the C53 and CE4 domains, but these differ by only 7 degrees. On all six heavy chains, electron density was present for five N-linked sugar units at Asn394 ([N-acetyghicosamine]2[Mannose]3; Fig. 5), and an additional two mannose residues were visible on chain C. Flexibility in the C23-CE4 inter-domain angle has been well-documented through studies of unliganded Fce3-4 structures in different crystal forms, as well as the unliganded IgE-Fc structure and the sFcERIcz complexes with both Fc23-4 and IgE-Fc. Amongst all of these structures the inter-domain angle varies over a range of approximately 25 degrees between the most "open" (in the sFcrRla complexes) and the most "closed" (chain Din the derCD23 complex reported here) (Fig. iia-d).

Previously, the most closed conformation had been seen in one of the unliganded Fcc3-4 structures (chain D of PDB 3HAO), and the conformations seen in the JO derCD complex all range within 3 or 4 degrees of this closed conformation (Fig. nd).

Nature of the IRE/CD22 interface The 2:1 stoichiometry agrees with earlier measurements in solution by analytical ultracentrifugation, recorded for both Fcc3-4 and IgE-Fc, but direct contact with CE4 was not anticipated by previous mutagenesis or peptide inhibition studies. Three salt -43 - bridges between FcE3-4 and derCD23 (Asp4o9-Arg188; G1u412-ArglS8; G1u412-Arg224) and a potential fourth (G1u414-His186 if protonated), together with four additional hydrogen bonds (Arg376-Tyr189; Asp4o9-Tyr189; Arg44o-Ser254, both side-chain and main-chain) dominate the interaction (Fig. 6 and Fig. 9). In the four salt bridges, all negative charges reside on IgE, all positive charges on CD23. Arg44o, while H-bonding to Ser254 of derCD23, retains its intra-chain salt bridge interaction with G1u529 of Cr4 that is seen in other, unliganded Fcr3-4 structures. These salt bridging and hydrogen bonding residues of IgE are not conserved in the other human antibody classes (Fig. 12), consistent with the specificity of CD23 for IgE. /0

Remarkably few additional residues make well-defined van der Waals contacts in all six interfaces (fle4n, G1y413, Pro439, G1u529, G1n535 in Fcr3-4; Trp184, Vah8, Asp227 in derCD23) and the surface complementarity is poor (calculated Sc values for the six interfaces range from 0.63 to 0.71); indeed, there is a substantial cavity at /5 the interface between G1u412 (in Fcr3-4), Arg224 and the Cys259-Cys273 disulphide bridge (in derCD23). The buried surface area for each interaction ranges from 86o to 89oA2 for all except one (chains F & L) at 920A2; in this latter case the derCD23 loop at residues Ser256 and G1u257 is not as disordered as it is in the other five independent views of the interaction. The Cr3 domain dominates the interface with 63% of the contact area, a further 25% involves the linker region (residues 437 to 440) and Cr4 contributes 12%. Al! of the salt-bridge and hydrogen bond interactions involve the Cr3 domain with the exception of Arg44o in the linker region, while Cr4 contributes on'y van der Waals interactions. Earlier studies had impflcated the AB loop of Cr3, and residue Lys352 in particular, in CD23 binding, but this loop is not directly involved at all and Lys352 makes no contact with CD23; any effects of mutagenesis in this oop must therefore be indirect. It is the EF oop/helix of the Cr3 domain that is central!y placed at the interface (Fig. 6). Surprisingly, no Ca2-ions were found bound to derCD23 in the complex; the role of Ca2+in the CD23-IgE interaction is discussed below.

Conformational change upon CD2 binding to IgE-Fc In the free state, IgE-Fc adopts an asymmetrically bent structure with the (Cr2)2 domain pair folded back against one of the Cr3 domains, but although the two Cr3-Cr4 pairs have different inter-domain angles, both are more open than that seen in is the derCD23-Fce3-4 complex (Fig. ha). To assess whether the presence of the Cr2 domains might be expected to affect derCD23 binding, the (Cr2)2 domain pair was modeled onto the derCD23-Fcr3-4 complex. This was achieved by superimposing the Cr3 domain of free IgE-Fc that contacts the (Cr2)2 domain pair, onto the Cr3 domain (chain A) of the derCD23-Fcr3-4 complex. One of the derCD23 molecules (chain G) lay very close to the Cr2 domains, with a minor rearrangement of the N-terminal residues of Cr2 required to prevent a steric clash. Contact between side chains of this derCD23 molecule and Cr2 in a complex with IgE-Fc (or indeed IgE) certainly cannot be ruled out; the other bound derCD23 molecule (chain I-I) however, lay far from the Cr2 domains. Modeling of the Cr2 domains onto the Fcr3-4 complex also revealed /0 that the AB loop of Cr2 (chain A) wouki have to adopt a sflghtly different conformation to avoid dashing with Cr3 (chain B) in the very closed conformation that this Cr3 domain adopts in the comp'ex.

Mutual exclusion of FcrRI and CD2.g binding to IgE-Fc /5 The closed conformation for the Cr3 domains in the derCD23 complex is dearly incompatible with FcrRI binding. This may be seen by superimposing the derCD23-Fcr3-4 and the sFcrRla-Fcr3-4 (Garman, 2000, Nature, 406:259-266) comp'exes on their (Cr4)2 domain pair, and noting the very different orientations of the Cr3 domains (Fig. 7a). The steric clashing of sFcrRla with the Cr3 domains in their CD23-bound conformation, and of both derCD23 molecules with the Cr3 and Cr4 domains in their sFcrRla-bound conformations, is clear from the superposition of the complexes upon one or other of the Cr3 domains (Figs. 7th & c). Thus IgE cannot bind both FcrRI and CD23 simultaneously: binding of sFcrRlct causes conformational changes that are incompatible with binding of either derCD23 molecule, and the binding of either derCD23 molecule ensures that the two sub-sites required for high-affinity binding of FcrRI are not both accessible.

This surprising conclusion was verified experimentally in solution using acceptor fluorophore-labeled IgE-Fc or Fcr3-4, and competitively displacing either donor fluorophore-labded derCD23 by imlabded sFcrRla, or labeled sFcrRla by unthbeled derCD23 (Fig. 8). It can be seen that sFcrRla readfly and completely displaces derCD23 from the complex wfth either IgE-Fc or Fcr3-4. Similarly, derCD23 can displace sFcrRla almost comp'etely from either IgE-Fc or Fcr3-4, but at very much higher concentrations as expected from the considerably lower binding affinity of -45 -derCD23. Positive contrcils for these FRET measurements were provided by displacement of labeled sFcaRla by udabeled sFcaRla, and labeled derCD23 by unlabeled deICD23.

Discussion The interaction between IgE and CD23 is critically involved in the allergic response at several stages, including allergen presentation, the regulation of IgE synthesis and transport of IgE and immune complexes across epithelial barriers in the gut and airways. At the cell surface, mCD23 is trimeric, and 5CD23 fragments shed from the /0 membrane that contain sufficient stalk region are also trimeric, although the structure of the trimer has only been modeled based either upon the structures of other C-type lectins or guided by NMR chemical shift data. In the crystal structure of the complex reported here, two derCD23 "heads" bind to IgE, one to each heavy-chain at a location between the Cc3 and Ce4 domains, and remote from the FccRI /5 binding site. The interaction is predominantly hydrophilic and dominated by salt bridges between positively charged CD23 residues and negatively charged IgE residues, despite the overall net positive charge (+9) of the CE3 domain. The site on CD23, in agreement with that identified by NMR chemical shift mapping by titration of N-labekd derCD23 with monomeric CE3 (titration with IgE-Fc ed to the formation of high molecular weight oligomers), is diametrically opposed to the connection to the a-helical coiled-coil stalk region (Fig. 5). This topology is such that an IgE molecule could not engage two heads from the same (modeled) CD23 trimer (as depicted in earlier cartoons), but could readily cross-link two mCD23 molecules.

(The distance of i36A between the Ca atoms of the N-terminal Phe158 residues of the two derCD23 molecules, which are immediately adjacent to the top of the predicted a-hehcal coiled-coil stalk region, is such that it would require unraveling of about 30 residues, or almost one third of the stalk, to allow two heads from the same CD23 trimer to bind to the same IgE-Fc). Furthermore, the exposed location of the IgE-binding site on derCD23 strongly suggests that trimeric 5CD23 could cross-link IgE Jo molecules either in solution or as mIgE on the cell surface, but a definitive statement must await determination of the structure of the CD23 trimer.

The binding of derCD23 in the deft between the CE3 and CE4 domains fixes the angle between these two domains. Compansons between FcE3-4 and IgE-Fc structures (the latter including CE2 domains), both free and bound to sFcaRla, show that while the (CE4)2 domain pair display no conformational variation and provide a fixed point -46 -of reference, the Cc3 domains can adopt a range of "open" and "closed" conformations. The most open of these have been found in the complexes of Fcc3-4 and IgE-Fc with sFccRla (Fig iia-d) and occur in both c-chains since this receptor engages with both Cc3 domains (Fig. 7a). In contrast, CD23 binding causes the Cc3 domains to adopt the most closed conformation that has yet been observed in any Fca3-4 or IgE-Fc structure. This opening and closing of the CE3 domains in fact results from a flexing within the C23 domain, an effect first noted in earlier structural comparisons. The AB loop/helix of C23 interacts closely with C24 (Fig. 6), and a change in the relative orientation of the EF loop/hefix, which is contacted by /0 derCD (Fig. 6), can clearly be seen when the FccRI and CD23 complexes are compared (Fig. 7a). This intra-domain flexibility may be a result of the unique packing/stability profile of Cc3 compared with all other CH domains, and in fact the isolated Cc3 domain displays molten globule-like characteristics.

/5 The consequence of these extreme open (FccRI-bound) and closed (CD23-bound) conformations for the Cr3 domains of IgE is that the binding of the two receptors are incompatible with each other (Figs. 7a-c). Clearly this is not, as previously thought, due to overlapping binding sites, but to a surprising allosteric linkage between the two distant sites. The idea that the open and closed conformations might interact differently with Feclil and CD23 has been previously considered following a surprising discovery that CD23 bound more strongly to IgE at 4°C than 37°C, while the opposite was true for FccRI. It was hypothesised that a change in the relative proportion of these open and closed conformational states might account for this temperature effect, and it will certainly be revealing to explore further the temperature dependence of the kinetics and thermodynamics of these receptor interactions as well as the conformational dynamics of the IgE molecule, particularly with respect to the Cr3 domains.

IgE-Fc and IgE additionally contain the Cr2 domains, and it is important to consider their effect upon the crystallographic results presented here for Fcr3-4. In IgE-Fc the (Cr2)2 domain pair packs asymmetrically against one of the Cr3 domains in the bent IgE-Fc structure, and moves together with that Cr3 domain when it opens up to accommodate sFcrRla binding. Modeling the (Cr2)2 domains onto the derCD23-Fca3-4 complex shows that although there is no steric conflict that would prevent is binding of either derCD23 molecule, the Cr2 domains lie immediately adjacent to one -47 -of the derCD23 m&ecules and suggest that an interaction (stabilizing or destabilizing) might occur. This is consistent with the observed 2:1 stoichiometry for derCD23 binding to both FcE3-4 and IgE-Fc. Akhough there is no evidence of any significant difference between the binding affinity of derCD23 for FcE3-4, IgE-Fc or IgE, or between the binding of the two derCD23 molecules to IgE-Fc, a slightly faster on-rate for derCD23 binding to Fce3-4 compared with both IgE-Fc and IgE has been reported. The biphasicity or dual affinity of CD23 binding to IgE that has been reported is a function of its oligomeric state and ability to engage 1gB through more than one head, although as pointed out above, two heads of one CD23 almost /0 certainly cannot engage the same 1gB molecule and the avidity effect must result from trimeric CD23 binding to more than one 1gB molecule.

The structural comparison of the derCD23 and sFcaRla complexes (Figs. 7b & c) also shows that sFccRla binding prevents binding of both derCD23 molecules, and /5 similarly that binding of either derCD23 molecule will prevent sFcaRla binding. The experimental data presented here (Fig. 8) confirm this observation. sFcrRla. readily displaces derCD23 and, as expected, at approximately equimolar (5 jiM) concentrations of sFcaRla and IgE-Fc (or Fca3-4) the displacement of derCD23 is virtually complete since this concentration of sFccRla is very much greater than the Ku 1 nM for sFceRla binding. (The difference between the two curves in Fig.8a may reflect the slightly lower affinity reported for sFceRlct binding to FcE3-4 compared with IgE-Fc). derCD23 can similarly cause almost complete displacement of sFceRla, although much higher concentrations are required, greater than ioojiM, in order to overcome its lower affinity (Ku 1O -106 M).

CD23 belongs to the C-type (calcium-dependent) lectin superfamily and the presence of Ca2t is known to enhance the affinity for IgE approximately 7-fold, although it is not essential for binding. A crystal structure of the head domain with a single bound Ca2+ ion has been solved, together with the Ca2+_free form, and an NMR structure Jo also reports Ca2+ binding, but at an alternative site. However, no Ca2+ ions were observed in any of the six derCD23 molecules in the FcE3-4 complex, despite the presence of 2 mM Ca2+ in the crystallization medium. In this regard, the existence of a disordered loop in derCD23 at the edge of the interface (residues 256-257 of loop 4, following earlier terminology) is intriguing. The crystal structure of the lectin head i5 domain showed Ca2+ bound at a site involving residues G1u249 and Thr251 of loop 4.

-48 -In that structure, residues 253-257 of loop 4 were also disordered, paradoxically becoming ordered in the absence of Ca2+ due to a rearrangement of the side-chain of Arg253, which occupied the Ca24 site. The conformation of loop 4 seen in the complex differs from either the apo-or Ca2._bound structures; together with the adjacent ioop 1, they represent the only main-chain conformational changes in derCD23 upon Fca3-4 binding. The r&e of Ca2 in IgE binding to CD23 thus remains unres&ved at present, but it may involve reorganization of residues of the flexible loop 4 in the context of the complex, providing additional protein-protein interactions and thus enhancing affinity. There are precedents for the stabilizing effects of Ca24 upon this /0 and other loops in C-type lectin domains, such as mannose-binding protein in which carbohydrate binding is affected, and others where the binding of a protein partner is affected.

What is the functional significance of IgE's inability to engage both FccRI and CD23 /5 simultaneously? Although the affinity of derCD23 for IgE is considerably lower than that of FcaRJ, avidity effects for trimenc CD23 can considerably enhance its ability to compete with FcERI binding. Indeed, a recombinant 5CD23 species trimerized via a leucine-zipper motif, termed IzCD23, more effectively inhibited IgE binding to mast cells expressing FcaRI. The mutual exclusion of FcERI and CD23 binding to 1gB is essential to prevent mast cell and basophil activation by trimeric 5CD23, which could otherwise cross-Unk FcERI-bound 1gB on these cells. Similary, it ensures that FceRI cannot be cross-Unked either by sohible IgE-CD23 complexes, or 1gB bound to mCD23 (on B cells or APCs) in the absence of allergen. The mutual exclusion of FcERI and CD23 binding is thus an important aspect of 1gB biology, allowing it to function independently through its two receptors.

Furthermore, the location of the CD23 binding sites at points where the two heavy chains are most widely separated (in contrast to the FcaRI binding site where the chains approach each other most dosely) and the connections to the stalk region (Fig. so), maximize the propensity for cross-linking of mIgE on B cells committed to 1gB synthesis bytrimeric 5CD23, and also mCD23 on B cells by 1gB or 1gB-allergen complexes. The inventors have hypothesized that through such interactions, the former leading to up-regulation of 1gB synthesis and the latter to down-regulation, CD23 contributes to the mechanism of IgE homeostasis, and this notion has received is experimental support from studies with monomeric and oligomeric 5CD23 species.

The structure of the complex is also consistent with the co-crosslinking of mCD21 and -49 -mIgE by trimeric 5CD23 (proposed to enhance IgE up-regulation), since CD21 binds to the "tail" sequence that is, although only partially present in derCD23, located adjacent to the connection to the stalk [Fig. 5;]. However, the precise spatial arrangement of the IgE and CD21 binding sites on each derCD23 head and the relative positions of the heads in the CD23 trimer (as yet only modeled), together with the fact that two heads from one trimer almost certainly cannot engage a single IgE molecule, undoubtedly places topological constraints upon the formation of signaling complexes of these three molecules at the B cell surface.

/0 The crystal structure of the derCD23-FcE3-4 complex reported here provides the first structural basis for designing inhibitors with the potential to interfere with IgE regulation and other CD23-mediated processes. Of wider significance to IgE and allergic disease, the structure also reveals the mechanism for an allosteric connection between the distantly located binding sites for IgE's two principal receptors, and Is demonstrates that allosteric inhibition is a viable strategy to target either of these receptor interactions for therapeutic purposes.

Summary

The role of immunoglobulin E (IgE) in allergic disease mechanisms is performed principally through its interactions with two receptors, FceRI on mast cells and basophils, and CD23 (FceRII) on B cells. The former mediates allergic hypersensitivity, the latter regulates IgE levels, and both receptors, also expressed on antigen presenting cells, contribute to allergen uptake and presentation to the immune system. The inventors have solved the crystal structure of the soluble lectin-like "head" domain of CD23 (derCD23) bound to a sub-fragment of IgE-Fc consisting of the dimer of Cc3 and Cz4 domains (FcE3-4). One CD23 head binds to each heavy chain at the interface between the two domains, explaining the known 2:1 stoichiometry and suggesting mechanisms for cross-linking membrane-bound trimeric CD23 by IgE, or membrane IgE by soluble trimeric forms of CD23, both of which may contribute to the regulation of IgE synthesis by B cells. The two symmetrically located io binding sites are distant from the single FceRI binding site, which lies at the opposite ends of the CE3 domains. Structural comparisons with both free IgE-Fc and its FcERI comp'ex reveal not only that the conformational changes in IgE-Fc required for CD23 binding are incompatible with FcERI binding, but also that the converse is true. The two binding sites are allosterically linked. The inventors demonstrate experimentally the reciprocal inhibition is of CD23 and FcRI binding in solution, and suggest that the mutual exclusion of receptor binding allows IgE to function independently through its two receptors.

Claims (14)

1. -50 -Claims 1. A variant IgE Fc region comprising at least one amino acid modification in the Cs3 domain relative to a wild-type IgE Fc region, wherein the at least one amino acid in the wild-type 1gB Fc region which is modified is pan of a CD23 binding site, and wherein the variant 1gB Fc region exhibits a reduced binding affinity to CD23 compared to the wild-type IgE Fe region.
2. 2. A variant 1gB Fc region according to claim 1, wherein the variant IgE Fe region exhibits ablated binding to CD23 (FcRII), but is capable of binding to FcRI. /0
3. 3. A variant IgE Fe region according to either claim 1 or 2, wherein the variant 1gB Fe region is derived from the wild-type human 1gB Fc comprising an amino acid sequence substantially as set out in SEQ ID No:2, or a functiona' variant or fragment thereof. /5
4. 4. A variant 1gB Fe region according to any preceding claim, wherein the variant 1gB Fe region is derived from the wild-type human IgE Fc region which is encoded by a nucleic acid sequence substantially as set out in SEQ ID No:i, or a functiona' variant or fragment thereof.
5. A variant 1gB Fe region according to any preceding claim, wherein the at east one amino acid which is modified is any one of amino acid residues 329-441 of SEQ ID No:2.
6. 6. A variant 1gB Fe region according to any preceding claim, wherein the at east one amino acid modification in the variant IgE Fe region is not in the A-B helix of the Cs3 domain of 1gB.
7. 7. A variant 1gB Fe region according to any preceding claim, wherein the variant 1gB Fe region compnses at least one amino acid modification in the B-F hehx of the Cs3 domain of 1gB, in the C-D loop of the CE3 domain of IgE and/or in the C4erminal region of the Cc3 domain of 1gB.
8. 8. A variant IgE Fe region according to any preceding claim, wherein the variant 1gB Fe region comprises at least one amino acid modification at a position corresponding to wild-type 1gB Fe region selected from the group consisting of: (i) -51 - amino acids 404-415 from the E-Fhelix of IgE; (ii) amino acids 376-380 from the C-D loop of IgE; and (iii) residue 433-441 from the C-terminal region of IgE.
9. 9. A variant IgE Fc region according to any preceding claim, wherein the variant IgE region comprises at least one amino acid modification at a position corresponding to wild-type IgE Fc region selected from the group consisting of: 376; 377; 378; 379; 380; 404; 405; 406; 407; 408; 409; 410; 411 412; 413; 414; 415; 433; 434; 435; 436; r; 438; jç; 440; and 441 of SEQ ID No:2, and preferaNy at position 376; 378; 381; 408; 409; 411 412; 435; 439; and/or 440 of SEQ ID Nb:2. /0
10. 10. A variant IgE Fc region according to any preceding claim, wherein the variant IgE region comprises at least one amino acid modification at a position equivalent to 409,412,376, 380 and/or 435 of the wild-type human IgE Fc region.

    /5
11. ii. A variant IgE Fc region according to any preceding claim, wherein the at least one modification comprises an amino acid substitution.
12. 12. A variant IgE Fc region according to any preceding claim, wherein the at least one modification comprises a non-conservative substitution.
13. 13. A variant IgE Fc region according to any preceding claim, wherein the variant Fc region comprises at least one amino acid modification selected from: D409A; E412A; R376A; K380A and K435A.
14. 14. A variant IgE Fc region according to any preceding claim, wherein the variant human IgE Fc region comprises more than one modification in the Ca3 domain of the Fc region equivalent to the human IgE Fc region of a wild-type antibody.is. A variant IgE Fc region according to any preceding claim, wherein the variant human IgE Fc region comprises at least two, three, four or five modifications at any of the amino acids at positions eqtuvalent to 376; 377; 378; 379; 380; 404; 405; 406; 407; 408; 409; 410 411; 412; 413; 414; 415; 433; 434; 435; 436; 437; 438; 439; 440; and 441 of wild-type human IgE-Fc.16. A variant IgE Fc region according to any preceding claim, wherein the variant Fc region comprises a modification at each of the amino acids at positions equivalent to 409, 412,376,380 and/or 435 of wild4ype human IgE-Fc.-52 - 17. A variant IgE Fc region according to any preceding claim, wherein the variant IgE Fc region comprises an amino acid sequence substantially as set out in any one of SEQ ID No: 3-7, or a functional variant or fragment thereof.i8. A variant IgE Fc region according to any preceding claim, wherein the equiBbrium constant of dissociation (K or lcd) of the variant IgE Fc region for CD23 is at east 5gM, 10gM, 25gM or greater./0 19. A variant IgE Fc region according to any preceding claim, wherein the variant IgE Fe region is capable of binding to FccRI with a KA value of at least icr M-1.20. A nucleic acid encoding the variant IgE Fe region according to any one of claims 1-19. /521. A nucleic acid according to claim 20, wherein the nucleic acid comprises a nucleotide sequence substantially as set out in any one of SEQ ID No: 8-12, or a functional variant or fragment thereof.22. A genetic construct comprising the nucleic acid according to either claim 20 or claim 21.23. A recombinant vector comprising the genetic construct according to claim 22.24. A host cell comprising the genetic construct according to either claim 20 or claim 21, or the recombinant vector according to claim 23.25. A method of preparing a variant IgE Fe region, the method comprising-(i) culturing at least one cell according to claim 24 under conditions suitable for the expression of a variant IgE Fe region; and (ii) is&ating the variant IgE Fe region.26. A method of producing a variant IgE Fc region which exhibits a reduced binding affinity for CD23, the method comprising modifying at east one amino acid in the C3 domain of a wild-type IgE Fc region, wherein the at east one amino acid in the wild-type IgE Fe region which is modified is part of the CD23 binding site, and wherein the variant IgE Fc region exhibits a reduced binding affinity to CD23 compared to the wild-type IgE Fe region.-53 - 27. A method according to claim 26, wherein the variant IgE Fe region produced by the method is as defined in any one of claims 1-19.28. A method according to either claim 25 or claim 26, wherein the wild-type IgE Fe region which is modified is part of a therapeutically active IgE antibody.29. A method according to either claim 25 or claim 26, wherein the method comprises initially creating the variant IgE Fe region and then fusing it to a /0 functional fragment of an IgE antibody to produce an Fe fusion, for example a chimera.30. An antibody or a functional fragment thereof comprising the variant IgE Fc region of any one of claims 1-19. /531. An antibody or a functional fragment thereof according to claim 30, wherein the antibody or functional fragment thereof is incapable of binding to CD23 (FcaRII), or exhibits a reduced binding affinity thereto, but is capable of binding to FccRI.32. An antibody or a functional fragment thereof according to either claim 30 or claim 31, wherein the antibody or functional fragment thereof is monoclonal or polyclonal.33. An antibody or a functional fragment thereof according to any one of claims 30-32, wherein the functional fragment comprises any of the fragments selected from a group consisting of VH, VL, Fd, Fv, Fab, Fab', scFv, F (ab')2 and Fc fragment.34. A variant IgE Fe region according to any one of claims 1-19, or an antibody or a functional fragment thereof according to any one of claims 30-33, for use in therapy or diagnosis.35. A variant IgE Fc region according to any one of claims 1-19, or an antibody or a functional fragment thereof according to any one of claims 30-33, for use in treating, preventing or ameliorating a CD23-mediated side effect.-54 - 36. A variant IgE Fc region, or an antibody or a function& fragment thereof according to claim 35, wherein the CD23-mediated side effect is prevented in IgE antibody treatment or therapy.37. A variant IgE Fc region, or an antibody or a functional fragment thereof according to either claim 35 or 36, wherein the CD23-mediated side effect is prevented during the treatment of cancer, autoimmune disease or infectious disease.38. A variant IgE Fc region, or an antibody or a functional fragment thereof /0 according to any one of claims 35-37, wherein the CD23-mediated side effect is CD23-dependent epitope spreading.39. A pharmaceutical composition comprising a variant IgE Fc region according to any one of claims 1-19, or an antibody or a functional fragment thereof according /5 to any one of claims 30-33; and optionally a pharmaceutically acceptable vehicle.40. A process for making the composition according to daim 39, the process comprising combining a therapeutically effective amount of a variant 1gB Fc region according to any one of claims 1-19, or an antibody or a functional fragment thereof according to any one of claims 30-33, with a pharmaceutically acceptable vehicle.