Báo cáo Y học: Functional epitope of common c chain for interleukin-4 binding ppt

Functional epitope of common c chain for interleukin-4 bindingJin-Li Zhang, Manfred Buehner and Walter Sebald Theodor-Boveri-Institut fu¨r Biowissenschaften Biozentrum, Physiologische Ch

Functional epitope of common c chain for interleukin-4 binding

Jin-Li Zhang, Manfred Buehner and Walter Sebald

Theodor-Boveri-Institut fu¨r Biowissenschaften (Biozentrum), Physiologische Chemie II, Universita¨t Wu¨rzburg, Germany

Interleukin 4 (IL-4) can act on target cells through an IL-4

receptor complex consisting of the IL-4 receptor a chain and

the common c chain (cc) An IL-4 epitope for ccbinding has

previously been identified In this study, the cc residues

involved in IL-4 binding were defined by alanine-scanning

mutational analysis The epitope comprises ccresidues I100,

L102, and Y103 on loop EF1 together with L208 on loop

FG2 as the major binding determinants These

predomin-antly hydrophobic determinants interact with the

hydro-phobic IL-4 epitope composed of residues I11, N15, and

Y124 Double-mutant cycle analysis revealed co-operative

interaction between cc and IL-4 side chains Several cc residues involved in IL-4 binding have been previously shown to be mutated in X-linked severe combined immunodeficiency The importance of these binding residues for ccfunction is discussed These results provide a basis for elucidating the molecular recognition mechanism in the IL-4 receptor system and a paradigm for other cc-dependent cytokine receptor systems

Keywords: common c chain; interleukin 4; mutagenesis; protein–protein interaction; structure/function

Interleukin-4 (IL-4) is a multifunctional cytokine that plays

a critical role in the regulation of immune responses [1,2] It

induces the generation of Th2-dominated early immune

response [3] and determines the immunoglobulin class

switching to IgE [4] Dysregulation of IL-4 function is

strongly correlated with type I hypersensitivity reactions,

such as allergies and asthma [5] The IL-4 receptor complex

is therefore a potential target for the development of

antiallergic drugs The central role of IL-4 in the

develop-ment of Th2 cells suggests that it may be of benefit in the

treatment of autoimmune disease characterized by an

imbalance of Th cells [6] Its ability to induce growth arrest

and apoptosis in leukemic lymphoblasts in vitro [7] suggests

that IL-4 is also a promising cytokine for the treatment of

high-risk acute lymphoblastic leukemia Understanding the

molecular recognition mechanism in the IL-4 receptor

system is a prerequisite for the rational design of IL-4-like

drugs

IL-4 is one of the short-chain four-helix bundle cytokines

Its effects depend on binding to and signaling through a

receptor complex consisting of a primary high-affinity

binding subunit, the IL-4Ra, and a low-affinity receptor,

depending on the cell type, the common c chain (cc; type I

IL-4 receptor [8]) or IL-13Ra1 chain (type II IL-4 receptor

[9]) All three receptors are members of the type I cytokine receptor superfamily, which is characterized by the presence

of at least one cytokine-binding homology region (CHR) composed of two fibronectin type III domains The membrane distal domain contains a set of four conserved cysteines, and the membrane proximal domain contains a WSXWS motif [10] The fibronectin type III domain is comprised of seven b strands, the sequences of which are conserved between members of the family, while loop sequences connecting the b strands vary between family members and putatively contain residues that mediate distinct intermolecular contacts These loop regions were therefore selected for this mutational analysis

A comprehensive mutational analysis of IL-4 in which single residues were replaced by alanine or charged residues yielded high-resolution data on the binding epitopes for the receptor chains The IL-4 site 1 binding epitope for IL-4Ra consists of a mixed charge pair (E9, R88) as major determinants and five minor determinants located on helices

A, B, and C [11] The importance of site 1-binding determinants and their partner residues on IL-4Ra (D72, Y183 as key binding determinants) was subsequently confirmed and further defined by determining the crystal structure of the 1 : 1 IL-4/IL-4Ra ectodomain (IL-4-binding protein, IL-4BP) complex [12] and by mutational analysis of the IL-4BP binding epitope [13] The results have already been used for the rational design of IL-4 minipro-teins [14] The IL-4 site 2 epitope for cccomprises residues I11 and N15 on helix A together with Y124 on helix D as major binding determinants and three minor determinants K12, R121, and S125 on helices A and D [15] A double mutant of IL-4 that completely inhibits responses induced

by IL-4 and IL-13 by disrupting the binding of the IL-4 site

2 epitope to ccor IL-13Ra1 proved to be a very promising anti-asthma drug [16–18] Two further IL-4 mutants that selectively inhibit IL-4-induced activity on endothelial cells appeared to be good candidate drugs for the treatment of certain autoimmune diseases [6] and high-risk acute lymphoblastic leukemia [7] However, the residues on cc that contribute to IL-4 site 2 binding remain uncertain

Correspondence to W Sebald, Theodor-Boveri-Institut fu¨r

Biowis-senschaften (Biozentrum), Physiologische Chemie II, Universita¨t

Wu¨rzburg, Am Hubland, D-97074 Wu¨rzburg, Germany.

Fax: + 49 931 888 4113, Tel.: + 49 931 888 4111,

E-mail: sebald@biozentrum.uni-wuerzburg.de

Abbreviations: IL-4, interleukin-4; IL-4Ra, interleukin-4 receptor a

chain; IL-4BP, IL-4 binding protein; c c , common c chain; IL-13Ra1,

IL-13 receptor a1 chain; CHR, cytokine-binding homology region;

Jak, Janus kinase; XSCID, X-linked severe combined

immunodefi-ciency; hGHR, human growth hormone receptor; hEPOR, human

erythropoietin receptor; b c , common b chain.

(Received 14 November 2001, revised 16 January 2002, accepted

21 January 2002)

cc is shared by several important cytokine receptor

complexes, including those for IL-2, IL-4, IL-7, IL-9, IL-15

[8] and also for the recently described new member of the

cytokine family, IL-21 [19] ccalone binds ligands with very

low affinity (Kd 150 lMfor IL-4) [15] Recruitment of cc

into receptor complexes for the above cytokines increases

receptor affinity for binding [20–22] cc participates in

cytokine signaling in several receptor complexes via JAK3

[23] Mutations of either cc or JAK3 result in X-linked

severe combined immunodeficiency (XSCID) which is

characterized by a failure in T and NK cell development

[24] cc-knockout mice have been generated and their

immune system successfully reconstituted by gene therapy

[25,26] Initial attempts at gene therapy for patients with

XSCID had been successful for more than 10 months

[27,28] Thus, defining the IL-4-binding determinants on cc

is important not only for elucidating molecular recognition

and activation mechanisms in the IL-4 receptor system and

possibly providing a paradigm for other cc-dependent

cytokine receptor systems, but also for delineating the

molecular pathology of XSCID

So far, the binding epitopes of human and murine cc

for some cc-dependent cytokines have been studied

A molecular mapping study using the antagonistic

monoclonal antibody PC.B8, which reacts with a

discon-tinuous site on human cc, localized ccbinding residues to

four loops, but did not identify single specific residues for

ligand binding [29] Mutational analysis of murine cc

employing heterodimeric IL-2R and IL-7R on whole cells

suggests that cc epitopes for IL-2 and IL-7 binding

overlap and comprise at least three distinct putative loop

segments of the ccprotein [30] Here we report the effect

of single amino-acid substitutions in the human cc

ectodomain on IL-4 binding Biosensor techniques

employing soluble recombinant IL-4, IL-4-BP and the

wild type or mutant forms of human cc ectodomain

revealed the contributions of ccresidues to IL-4 binding

The possible co-operativity between some residues on the

cc epitope and the IL-4 site 2 epitope was analyzed by

double-mutant cycle analysis

E X P E R I M E N T A L P R O C E D U R E S

Protein expression and purification

The ectodomain of human cc comprising amino-acid

residues 1–232 [20] was expressed with a C-terminal

thrombin cleavage site (LVPRGS) plus a His6tag in SF9

insect cells according to the manufacturer’s instructions

(PharMingen) The protein was isolated from the culture

medium of infected SF9 cells by standard procedures

involving Ni2+/nitriloacetate/agarose (Qiagen), digested

with thrombin (Sigma), and purified by gel-filtration

chromatography through a Superdex 200 HR 10/30

col-umn (Pharmacia) After exhaustive dialysis against water,

the purified protein was freeze-dried and stored at)80 °C

The cDNA for the murine cc ectodomain comprising

residues 1–233 [31] was cloned into the

temperature-regulated expression vector pRpr9 fd [32], expressed in

Escherichia colistrain KS 474, and refolded as described

[33] The refolded protein was purified to homogeneity by

gel-filtration chromatography through a Superdex 200 HR

10/30 column, and stored at)80 °C

The A182, C207 IL-4BP variant was produced in SF9 cells, purified, and biotinylated at C207 as described [32] IL-4 and IL-4 variants were expressed in E coli, refolded, and purified to homogeneity as described [11,34] Protein concentrations were determined by measuring A280, using

an absorption coefficient (e280) ¼ 8860M )1Æcm)1 for IL-4, e280 ¼ 7370M )1Æcm)1 for A124 IL-4, e280=

66 930M )1Æcm)1for IL-4BP, e280 ¼ 61 450M )1Æcm)1for human cc, e280 ¼ 60 170M )1Æcm)1for A103 human cc, and e280 ¼ 45 660M )1Æcm)1for murine cc

Mutagenesis of the ccectodomain cDNA for human ccectodomain was submitted to in vitro cassette mutagenesis employing synthetic double-stranded oligonucleotides The cc variants were expressed and purified as the wild-type human ccectodomain

Biosensor interaction analysis The binding of ccvariants to IL-4/IL-4BP was recorded on a BIAcore 2000 system (Pharmacia Biosensor) as described [15] Briefly, a CM5 biosensor chip was first loaded with streptavidin in flow cells 1 and 2 Subsequently biotinylated A182,C207 IL-4BP was immobilized at the streptavidin matrix of flow cell 2 at a density of 200 resonance units The following reaction cycle was applied using the com-mand COINJECT: (a) IL-4 at 0.1 lMin HBS buffer (10 mM Hepes, pH 7.4, 150 mM NaCl, 3.4 mM EDTA, 0.005% surfactant P20) was perfused over flow cells 1 and 2 at a flow rate of 10 lLÆmin)1at 25°C for 2 min; (b) 0.1 lMIL-4 plus

ccectodomain or ccvariants at 1–10 lMin the same buffer were perfused in the same way for 2 min; (c) HBS buffer alone was perfused for 5 min; (d) free receptors were regenerated by perfusion with 0.1Macetic acid/1MNaCl for 30 s Sensograms were recorded at a data-sampling rate

of 2.5 Hz and evaluated as described [15] Equilibrium binding of ccvariants at 1, 2, 3, 5, 10 lMwas measured for at least three times in duplicate The mean standard deviation (mean r) was 13.8% ± 6.5% for the Kdvalues calculated from the five variant concentrations For the double mutant cycle analysis [35], the same procedure as above was used except that IL-4 variants [15,36] at 0.1 lMand ccvariants at

2, 4, 6, 10, 20 lM were perfused (the mean r was 16.4% ± 7.4% for the Kdvalues) The loss of binding free energy on mutation for IL-4 and ccwas calculated as ddG (kJÆmol)1) ¼ 5.69 log Kd (mutant)/Kd (wild-type) The interaction energy between two residues was calculated by the double-mutant cycle method as in Eqn (1):

ddGint ¼ ddGX-A þ ddGY-B ÿ ddGX-A;Y-B ð1Þ where ddGX-A and ddGY-B are the changes in binding energy on mutation of X to A and Y to B (mutation of IL-4 and ccin this experiment), respectively, and ddGX-A, Y-Bthe change on the simultaneous mutation of X to A and Y to B ddGintis a measure of the co-operativity of the interaction

of the two components mutated ddGint ¼ 0 indicates that the pair of residues analyzed do not interact A positive value of ddGintmeans that two residues interact favorably, and a negative value means that the two residues repel each other [37] The individual errors (2 r, a ¼ 0.95) calculated from the mean for ddG are shown in Table 3

Molecular modeling of the IL-4–IL-4BP–ccternary

complex

The present model is based on the crystal structure of the

complex of IL-4 and IL-4BP (PDB entry 1IAR [12]),

augmented by the model of ccderived from human growth

hormone receptor (hGHR), as obtained from an older

model (T Mueller, & W Kammer, personal

communica-tion, Universita¨t Wu¨rzburg, Germany)

complex of IL-4–IL-4BP–cc This old model was based on

the structure of free IL-4 (PDB entry 1HIK [38]) and of

models of the extracellular domains of IL-4Ra and cc

obtained by analogy modeling following the structure of the

hGHR complex (PDB entry 3HHR [39]) The 3HHR data

were obtained from the protein databank (PDB [40]) The

old model was built in such a way that all cysteine residues

formed proper disulfide bonds, and all evidence from

mutation experiments available at the time was used to

adjust the binding epitopes of the receptor chains The

resulting alignment required some nontrivial rebuilding with

insertions and deletions, and, consequently, the resulting

model of the IL-4 receptor complex had to be extensively

energy refined The program O [41] was used for model

building, and the programX-PLOR[42] for energy refinement

The differences between the experimentally determined

binary complex and the corresponding components of the

old model were significant in detail, but the gross changes

were small enough that the binding topology of cccould be

transferred to the new model without major problems

The local program DISDM2 was used (H J Hecht, &

M Buehner, unpublished results) to build and adjust the

present model using the data of mutational analysis of IL-4

and cc The program runs under Open-VMS and uses

Datagraph VTC 8002 and VTC 8003 terminals for display

All model building was performed manually For online

refinement of conformational energy, the program EREF

was used [43], which is called from withinDISDM2

R E S U L T S

Site-specific mutagenesis of amino acids

in the ccectodomain

Alanine substitutions were targeted to residues in four

putative interconnecting loops and the interdomain segment

of the human ccectodomain based on the published models

[44–46], and sequence alignment performed between ccand

several cytokine receptors, the major ligand-binding

deter-minants of which were identified These include the hGHR

[39,47], the human erythropoietin receptor (hEPOR

[48,49]), IL-4BP [12], and the human gp130 (hgp130

[50,51]) Eighteen ccvariants were generated with

amino-acid substitutions in the AB1, EF1, BC2, FG2 loops and the

interdomain segment (Fig 1) A deletion mutant lacking

residues 1–33 of the N-terminus of cc, named ccCHR, was

also generated to find out whether this N-terminal region of

ccis required for ligand binding

All human cc wild-type or variant proteins could be

purified to apparent homogeneity by Ni2+/nitrilotriacetate/

agarose and gel filtration The wild-type human cc

ectodo-main expressed in SF9 cells was recovered as monomeric

and dimeric species [52,53] The murine cc ectodomain

expressed in E coli occurred as a monomer (Fig 2) Initial

biosensor studies showed that the different forms of human and murine proteins exhibited similar binding affinity for the IL-4–IL-4BP complex The mixture of monomeric and dimeric human c interacts with the complex with a K of

Fig 1 Amino-acid substitutions in the ectodomain of the human com-mon c chain (c c ) The amino-acid sequence of c c is shown with boxed portions indicating predicted b-strands which are designated by the letter below the box Residues substituted in this study are indicated by asterisks.

Fig 2 Gel-filtration analysis of the human c c ectodomain expressed in SF9 cells and the murine c c ectodomain expressed in E coli The samples were applied to a Superdex 200 HR 10/30 column and eluted with the same buffer The two peaks of human c c represent dimer (A) and monomer (B).

4 lM, and murine cc with a Kd of 1.6 lM (Fig 3 and

Table 1) In addition, different preparations of wild-type

human ccectodomain consistently showed a Kdof 4 lM

irrespective of the monomer to dimer ratio (data not

shown) Therefore, the mixtures of dimeric and monomeric

human ccprotein were used for all biosensor measurements

The ccepitope for IL-4 binding

The method of measuring the binding of ccto IL-4–IL-4BP

by biosensor was established previously [15] The

dissocia-tion constant Kd evaluated from the concentration

dependence of equilibrium binding proved to be very

reliable for measuring the interaction of the ccectodomain

with IL-4–IL-4BP The measured Kdfor interaction of cc

ectodomain variants with IL-4–IL-4BP are compiled in

Table 1 Eight cc variants including cc CHR exhibited

unchanged binding characteristics Changes in binding

affinity were observed in 11 cc variants The Kd of six

variants was too high to be reliably determined A rough

estimate yields Kdvalues of about 200–300 lMfor I100A,

L102A, Y103A and L208A, and Kdvalues of about 500–

1000 lM for C160A and C209A The Kd values of five

variants, N128A, H159A, L161A, E162A, and G210A,

were found to be increased threefold to fourfold compared

with the Kdof wild-type cc, suggesting that these residues are

part of the ccbinding interface, but do not play a key role in

binding The loss of binding affinity of the four variants

I100A, L102A, Y103A and L208A is not likely to be caused

by extended structural alterations, as I100A, L102A, and

L208A were reported to bind to IL-2 and IL-7 with the

same affinity as wild-type cc, and the Y103A mutation

resulted in only twofold to threefold reduced IL-2 and IL-7

binding [30] Thus, the four residues I100, L102, Y103 and

L208 are hot spots on cc, contributing > 9 kJÆmol)1each

The five minor residues investigated contribute only 2.9–

3.5 kJÆmol)1 The two cysteine variants C160A and C209A

exhibited a largely reduced binding affinity (Kd> 500 lM) This may be caused by structural perturbation of the protein A direct role in binding, however, cannot be excluded for these residues

Double-mutant cycle analysis of the IL-4–ccinterface The co-operativity of the interaction of some residues on the IL-4 site 2 epitope and the ccectodomain was determined in this experiment Double-mutant cycles were constructed only for the mutants with minimal effects on binding (Tables 2 and 3), because the Kdvalues for the interaction between variants of the main binding residues I100, L102, Y103, and L208 of cc and IL-4 variants as well as the interaction between variant I11A of IL-4 and ccvariants were too high to be reliably determined Interaction between residues on IL-4 and cccan be grouped according to their coupling energies (Table 3) The main binding determinant

of IL-4 Y124 failed to exhibit positive coupling energies with any of the cc residues analyzed IL-4 Y124 probably interacts with the main functional side chains of the receptor located on loop EF1 (I100, L102, Y103), the binding of the alanine variants of which was too weak to be analyzed by this approach Remarkably, IL-4 S125 neighboring Y124 does show coupling to receptor N128 in addition to that to

Fig 3 Sensograms recording the binding of human and murine c c

ectodomains to the IL-4–IL-4BP complex IL-4BP was immobilized on

the biosensor matrix At time zero, perfusion with 100 n M IL-4 was

initiated The saturation binding of IL-4 was arbitrarily set as zero.

After 120 s, perfusion was continued with 100 n M IL-4 plus c c

ecto-domain In different cycles, 5 l M human (a) or murine (b) c c

ecto-domains were applied Perfusion with buffer alone started at time

240 s The ruler indicates resonance units (RU) corresponding to

1–10 l M murine c c ectodomain The resonance unit for 5 l M human c c

corresponded to that for 1.6–2 l M murine c c

Table 1 Equilibrium binding between c c ectodomain mutants and IL-4–IL-4BP The dissociation constants K d were evaluated from equilibrium binding between wild-type (wt) or mutants (mut) of the c c

ectodomain and immobilized IL-4BP saturated with IL-4 The loss of free energy of binding on mutation was calculated as ddG (kJÆmol)1) ¼ 5.69 log K d (mut)/K d (wt).

Alanine variant

Equilibrium binding

ddG (kJÆmol)1)

K d (l M ) K d (mut)/K d (wt) Murine c c (wt) 1.6

Loop 1 (AB1)

Loop 3 (EF1)

Loop 4 (ID)

Loop 5 (BC2)

Loop 6 (FG2)

G210 The IL-4 side chain of N15 functionally interacts with

the central receptor side chain N128, and also with H159

located at the periphery of the functional ccepitope The

relative positions of the coupling side chains as proposed by

our theoretical model of the ternary complex (see below) are

presented in the open-book view in Fig 4A,B The two

receptor side chains N1128 and H159 are 12 A˚ apart in the

cc model This could indicate that our cc model is inaccurate, because this model does not completely fit the results of the double-mutant cycle analysis Alternatively, the interaction of IL-4 side chain N15 with H159 (coupling energy only 0.8 kJÆmol)1) may be indirect Of particular interest is the IL-4 side chain of R121, which, after being substituted with D or E, leads to a selective IL-4 agonist specifically impaired in IL-13Ra1 binding [6,7,16,17,34] The IL-4 R121 was distinct in showing positive coupling during interaction with the ccside chain L161

Model of the structure of the IL-4–IL-4BP–ccternary complex

In a series of steps, ccwas adapted to achieve a good fit to the core structure (the binary complex; Fig 5) The procedure started with moving the whole chain (Ôrigid bodyÕ) Then domains and subdomains were moved indi-vidually The binary core complex was changed as little as possible, being an experimentally determined structure and thus the most reliable part of the model, but some minor changes in side chain orientation could not be avoided for proper adaptation An important point was to keep the C-terminal domains of the receptor chains close together, as this was expected to be essential for dimer formation and thereby signaling through the membrane The structures of

ccand the ternary complex were modeled so that residues that exhibit positive coupling energies during double-mutant cycle analysis were placed close to each other Occasionally, however, there was a Ôconflict of interestÕ between the requirements of interaction and those of dimerization

D I S C U S S I O N

This mutational analysis defines human ccresidues involved

in IL-4 binding The residues are located in the EF1, BC2, and FG2 loops and the interdomain segment of cc The functional binding epitope of cc includes residues I100, L102, Y103, and L208 as major binding determinants and five residues, N128, H159, L161, E162, and G210, as minor determinants Our results also show that the truncated cc CHR has the same binding affinity as the complete cc ectodomain, indicating that the short N-terminal region of

ccis not required for ligand binding This is true for most type I cytokine receptors, except for hgp130 [51] and granulocyte colony-stimulating factor receptor [54] There-fore, ccCHR, the short form of the ccectodomain, may be

Table 2 Double mutant cycle analysis of interaction between c c and

IL-4 The dissociation constants K d were evaluated from equilibrium

binding between wild-type (wt) or mutants (mut) of the c c ectodomain

and immobilized IL-4BP saturated with wild-type or mutants of IL-4.

The loss of free energy of binding on mutation was calculated as

ddG ¼ 5.69 log K d (mut)/K d (wt) ddG sum is the sum of the losses of free

energy of binding upon mutation for IL-4 and c c separately ND,

Sensogram could not be evaluated because of weak binding.

IL-4

variants

c c chain

variants

K d

(l M )

ddG (kJÆmol)1)

ddG sum

(kJÆmol)1)

Table 3 Co-operativity between residue pairs in the interaction interface of c c and IL-4 The coupling energy between a pair of residues was calculated

as ddG int ¼ ddG sum ) ddG (data from Table 2) according to eqn (1) The underlined values indicate favorable interaction The numbers in parentheses are the calculated errors (2 r, a ¼ 0.95) ND, Sensogram could not be evaluated because of weak binding.

ccchain

variants

ddG of IL-4 variants (kJÆmol)1)

Fig 5 Model of IL-4–IL-4BP–c c ternary complex The structures of IL-4, IL-4BP, and CHR of c c are depicted as ribbons and colored blue, red, and green, respectively The major binding residues on c c and IL-4 site II epitopes are represented by sticks The figure was generated using

Fig 4 Open-book view of complementary functional IL-4 (site 2) (A) and c c (B) binding epitopes, and missense mutations in the putative loops of c c

implicated in patients with XSCID (62) (C) The structures of IL-4 and c c from our model are depicted as ribbons The mutated residues are represented by space-filling models The colors of residues in IL-4 and c c binding sites indicate the loss of binding free energy [ddG (kcalÆmol)1) ¼ 1.36 log (K d variant/K d wild-type)] due to alanine substitution (see Tables 1 and 2; 1 kcalÆmol)1 ¼ 4.18 kJÆmol)1) The data for I11, K12, and Y124 were taken from Letzeler

2 et al [15] The letters in parentheses in (C) indicate the other mutations found in the same position The figure was produced with MOLSCRIPT and RASTER 3 D

better suited to form crystals of IL-4–IL-4BP–ccthan the

complete ccectodomain for solving the structure of the

low-affinity complex by X-ray diffraction

It appears that binding of cc to IL-4 is sustained

predominantly by hydrophobic interactions Of the nine

residues involved in IL-4 binding, five, in particular all four

major determinants, are hydrophobic We propose that

residues I100, L102, and Y103 of loop EF1, and L208 of

FG2 form a hydrophobic cluster to interact with the

hydrophobic epitope composed of residues I11, N15, and

Y124 on helices A and D of IL-4 ([12,15]; Figs 4 and 5)

Similar hydrophobic determinants have been found in

several type I cytokine receptors, including hGHR [39],

hEPOR [48], hgp130 [50], and the human common b chain

(hbc[55–57]) Two of the three loops EF1, BC2 and FG2 of

these receptors appear to establish two major functional

interfaces with the ligands, and the binding is dominated by

one or two hydrophobic aromatic residues For example,

W104 and W169 in loops EF1 and BC2 of hGHR, F93 and

F205 in loops EF1 and FG2 of hEPOR, F169 in loop EF1

of hgp130, and Y365 and Y421 in loops BC2 and FG2 of bc

are all key residues in binding interactions (Fig 6) In terms

of cc, Y103 is homologous to W104 of hGHR, to F93 of

hEPOR and to F169 of hgp130, and the FG2 loop

containing L208 may have a similar function to the loop

containing W169 in hGHR In this regard, Y103 and L208

may have the most important role in the hydrophobic

cluster for binding to IL-4

The two ccvariants, C160A and C209A, exhibited very

high Kd values (> 490 lM and > 900 lM, respectively)

The two cysteines may form a disulfide bond between loops

BC2 and FG2 This prediction is consistent with our model

and one of the published models [46] of cc The contribution

of the two residues to binding could not be directly

determined The disulfide bond may be only important for

maintaining the structural integrity of cc However, it

cannot be ruled out that the disulfide group participates directly in binding These questions may be answered when the structures of both free cc and the IL-4–IL-4BP–cc ternary complex are solved

Double-mutant cycle analysis could identify co-operativ-ity between two side chains [35], and predict a more detailed map of interacting residues without knowledge of the structures of the two proteins analyzed Unfortunately, the coupling energies between the major determinants on ccand IL-4 site 2 cannot be measured because of the low binding affinity of the alanine variants Nevertheless, our experiment revealed favorable interactions between several pairs of cc and IL-4 side chains The results support our prediction of hydrophobic interaction between the functional ccepitope and IL-4 site 2 reasonably well Accordingly, the binding epitope of cccan be divided into two functional interfaces (Figs 4A,B and 5): (a) I100, L102, and Y103 on the EF1 loop interact mainly with IL-4 Y124 and S125; (b) L208 and other residues on the BC2 and FG2 loops interact mainly with IL-4 N15, and probably I11 (coupling with ccresidues could not be determined) The most important is the interface on the EF1 loop of cc, because the partner residue IL-4 Y124 is a key determinant for binding (contributing 10.9 kJÆmol)1) [15], and the Y124D mutant exhibits a complete antagonist activity [36] The IL-4 R121 which is more important for IL-13Ra1 binding [6,7,16,17] was found

to interact with L161 on the BC2 and FG2 interface of cc Its interaction with the binding residues on the EF1 loop of

cccould not be excluded Therefore, it would be interesting

to determine the IL-13Ra1 epitope for IL-4 binding and compare it with the ccepitope defined in this experiment

It is unfortunate for our modeling process that the most effective mutations did not yield interaction data Therefore,

we had to rely on the residues of the weaker (but measurable) interaction which, although they are expected

to work over larger distances and thus provide less stringent constraints than desirable, nevertheless lead to a quite reasonable model as far as the gross features are concerned For all the details on a truly atomic scale, however, we have

to await the crystal structure of the ternary complex This study focuses on the molecular description of the mechanism of recognition between human IL-4 and cc Nevertheless, it will be important to understand how the cc mutations and the associated changes in IL-4 binding affect the biological activity of cc during IL-4 signaling or the signaling of the cytokines that depend on cc Previous experiments with IL-4 mutant proteins [15] revealed that substitutions in the cc binding epitope lead to partial agonists and IL-4 antagonists The binding affinity of such mutants to the receptor on whole cells was at most threefold reduced compared with wild-type IL-4 (see, e.g [6]), indicating that cc binding contributes only marginally to IL-4 binding affinity with the whole receptor complex (see also [21]) Remarkably, only a lowering of the ccbinding affinity of more than 100-fold, measured by Biosensor in certain IL-4 mutants, produced partial agonist activities of less than 20% It could be predicted that the ccmutant proteins with reduced IL-4 binding affinity will exhibit the same alterations in biological activity as the complementary IL-4 proteins Furthermore, some point mutations have been reported to abrogate or diminish the high affinity of IL-4 (and also IL-2 and IL-7) for Epstein-Barr virus-transformed B cell lines or cells transfected with

Fig 6 Alignment of loops of different cytokine receptors involved in

ligand binding The structure-based sequence alignment of hIL-4Ra,

hGHR, hEPOR, and hgp130 was taken from Hage et al [12] The

sequences of hc c and hb c were aligned manually The central part of the

EF1, BC2, and FG2 loops of receptors are selectively shown Key

residues for binding are underlined The deletions are marked Ô–Õ The

interaction of the receptor with the respective ligand is classified as

follows: a AD or AC helix interface involved in receptor binding;

b polarity of the interface; c affinity of ligand–receptor interaction.

mutant sequences derived from patients with XSCID

[58–61] Two ccmutants (A134V and R202C) were found

to produce twofold and fourfold reduced IL-4 and IL-2

binding, and to be less effective in modulating Jak3

activation stimulated by IL-4 and IL-2, respectively

[60,61] A134 is located at the periphery of the ccepitope

identified in this study and has not been included in the

present experiment

Some of the residues in the ccepitope for IL-4 binding as

identified in this study (Y103, L161, L208 and G210) have

been found to be mutated in patients with XSCID

(Fig 4B,C [62]) The XSCID phenotype seems to be caused

predominantly by the disruption of IL-7 and/or IL-15

signaling [28,63] Thus, the ccepitopes for binding of IL-4

and of IL-7 and/or IL-15 most likely share Y103, L161,

L208 and G210 as binding determinants Remarkably,

L161 and G210 of ccare only minor determinants for IL-4

binding The severe deficiency produced in XSCID may

result from the particular substitutions (G210R and L161S;

Fig 4C [62]); this could be more disruptive than an alanine

substitution Alternatively, L161 and G210 may be major

determinants for IL-7 and/or IL-15 binding More detailed

molecular information is needed on how far the ccepitopes

for binding of IL-4, IL-2, IL-7, IL-9, IL-15, and IL-21 differ

or coincide Then the severity of the clinical manifestation in

patients with XSCID can possibly be correlated with cc

mutations in major or minor binding determinants

The common nature of cc raises the possibility that

common residues for binding different ligands may exist in

this receptor Indeed, some common residues contributing

to binding of different ligands have been found in hbc[55,56]

and hgp130 [51] Our result and the mutagenesis analysis of

the binding of the murine ccchain to IL-2 and IL-7 [30]

show that Y103 of ccis a key ligand-interacting residue for

IL-2, IL-4, and IL-7 Y103 is probably a common critical

residue for all cc-dependent receptor systems In addition, in

that study [30], the counterpart of three dominated residues

I100, L102 and L208 of human ccfor IL-4 binding were

reported not to be important for IL-2 and IL-7 binding

These residues are probably unique to IL-4 binding, as

suggested by the fact that cc binding sites for different

cytokines overlap but are not identical [29,64] However, it

cannot be ruled out that some of the binding residues of cc

defined in our study also participate in IL-2 and IL-7

binding, as, in the aforementioned study, only one residue

(Y103) was shown to be directly involved in IL-2 and IL-7

binding Y103A or Y103R mutations resulted in only

slightly (twofold to threefold) reduced IL-2 and IL-7

binding [30] The difference between these results and our

own may partly originate from the different methods

applied Therefore, further studies will be required to

determine whether Y103 and other residues identified in

the present study are also involved in binding of other

cc-dependent cytokines

The co-ordinate of the model of the IL-4–IL-4BP–cc

ternary complex is available from the authors

A C K N O W L E D G E M E N T S

The authors are grateful to Dr J Nickel for helpful discussion,

Dr Siddiqi for drawing the figures, and W Ha¨delt and C So¨der for

excellent technical assistance This work was supported by the SFB 487,

TP B2, and by the Fonds der Chemischen Industrie.

R E F E R E N C E S

1 Paul, W.E (1991) Interleukin-4: a prototypic immunoregulatory lymphokine Blood 77, 1859–1870.

2 Nelms, K., Keegan, A.D., Zamorano, J., Ryan, J.J & Paul, W.E (1999) The IL-4 receptor: signaling mechanisms and biologic functions Annu Rev Immunol 17, 701–738.

3 Paul, W.E & Seder, R.A (1994) Lymphocyte responses and cytokines Cell 76, 241–251.

4 Coffman, R.L., Lebman, D.A & Rothman, P (1993) Mechanism and regulation of immunoglobulin isotype switching Adv Immunol 54, 229–270.

5 Romagnani, S (1994) Mechanism and regulation of immuno-globulin isotype switching Curr Opin Immunol 6, 838–846.

6 Shanafelt, A.B., Forte, C.P., Kasper, J.J., Sanchez-Pescador, L., Wetzel, M., Gundel, R & Greve, J.M (1998) An immune cell-selective interleukin 4 agonist Proc Natl Acad Sci USA 95, 9454–9458.

7 Srivannaboon, K., Shanafelt, A.B., Todisco, E., Forte, C.P., Behm, F.G., Raimondi, S.C., Pui, C.H & Campana, D (2001) Interleukin-4 variant (BAY 36–1677) selectively induces apoptosis

in acute lymphoblastic leukemia cells Blood 97, 752–758.

8 Sugamura, K., Asao, H., Kondo, M., Tanaka, N., Ishii, N., Nakamura, M & Takeshita, T (1995) The common gamma-chain for multiple cytokine receptors Adv Immunol 59, 225–277.

9 Gauchat, J.F., Schlagenhauf, E., Feng, N.P., Moser, R., Yamage, M., Jeannin, P., Alouani, S., Elson, G., Notarangelo, L.D., Wells, T., Eugster, H.P & Bonnefoy, J.Y (1997) A novel 4-kb inter-leukin-13 receptor alpha mRNA expressed in human B, T, and endothelial cells encoding an alternate type-II interleukin-4/inter-leukin-13 receptor Eur J Immunol 27, 971–978.

10 Bazan, J.F (1990) Structural design & molecular evolution of a cytokine receptor superfamily Proc Natl Acad Sci USA 87, 6934–6938.

11 Wang, Y., Shen, B.J & Sebald, W (1997) A mixed-charge pair

in human interleukin 4 dominates high–affinity interaction with the receptor alpha chain Proc Natl Acad Sci USA 94, 1657–1662.

12 Hage, T., Sebald, W & Reinemer, P (1999) Crystal structure of the interleukin-4/receptor alpha chain complex reveals a mosaic binding interface Cell 97, 271–281.

13 Zhang, J.L., Simeonowa, I., Wang, Y & Sebald, W (2002) The high-affinity interaction of human IL-4 and the receptor a chain is constituted by two independent binding clusters J Mol Biol 315, 399–407.

14 Domingues, H., Cregut, D., Sebald, W., Oschkinat, H & Serrano,

L (1999) Rational design of a GCN4-derived mimetic of inter-leukin-4 Nat Struct Biol 6, 652–656.

15 Letzelter, F., Wang, Y & Sebald, W (1998) The interleukin-4

site-2 epitope determining binding of the common receptor gamma chain Eur J Biochem 257, 11–20.

16 Tony, H.P., Shen, B.J., Reusch, P & Sebald, W (1994) Design of human interleukin-4 antagonists inhibiting interleukin-4-depend-ent and interleukin-13-dependinterleukin-4-depend-ent responses in T-cells and B-cells with high efficiency Eur J Biochem 225, 659–665.

17 Schnarr, B., Ezernieks, J., Sebald, W & Duschl, A (1997) IL-4 receptor complexes containing or lacking the gamma C chain are inhibited by an overlapping set of antagonistic IL-4 mutant pro-teins Int Immunol 9, 861–868.

18 Grunewald, S.M., Werthmann, A., Schnarr, B., Klein, C.E., Brocker, E.B., Mohrs, M., Brombacher, F., Sebald, W & Duschl, A (1998) An antagonistic IL-4 mutant prevents type I allergy in the mouse: inhibition of the IL-4/IL-13 receptor system completely abrogates humoral immune response to allergen and development of allergic symptoms in vivo J Immunol 160, 4004–4009.

19 Asao, H., Okuyama, C., Kumaki, S., Ishii, N., Tsuchiya, S., Foster, D & Sugamura, K (2001) The common gamma-chain is

an indispensable subunit of the IL-21 receptor complex.

J Immunol 167, 1–5.

20 Takeshita, T., Asao, H., Ohtani, K., Ishii, N., Kumaki, S.,

Tanaka, N., Munakata, H., Nakamura, M & Sugamura, K.

(1992) Cloning of the gamma chain of the human IL-2 receptor.

Science 257, 379–382.

21 Russell, S.M., Keegan, A.D., Harada, N., Nakamura, Y.,

Noguchi, M., Leland, P., Friedmann, M.C., Miyajima, A., Puri,

R.K., Paul, W.E & Leonard, W.J (1993) Interleukin-2 receptor

gamma chain: a functional component of the interleukin-4

receptor Science 262, 1880–1883.

22 Kondo, M., Takeshita, T., Higuchi, M., Nakamura, M., Sudo, T.,

Nishikawa, S & Sugamura, K (1994) Functional participation

of the IL-2 receptor gamma chain in IL-7 receptor complexes.

Science 263, 1453–1454.

23 Leonard, W.J & O’Shea, J.J (1998) Jaks and STATs: biological

implications Annu Rev Immunol 16, 293–322.

24 Leonard, W.J (1996) The molecular basis of X-linked severe

combined immunodeficiency: defective cytokine receptor

signal-ing Annu Rev Med 47, 229–239.

25 Lo, M., Bloom, M.L., Imada, K., Berg, M., Bollenbacher, J.M.,

Bloom, E.T., Kelsall, B.L & Leonard, W.J (1999) Restoration of

lymphoid populations in a murine model of X-linked severe

combined immunodeficiency by a gene-therapy approach Blood

94, 3027–3036.

26 Soudais, C., Shiho, T., Sharara, L.I., Guy-Grand, D., Taniguchi,

T., Fischer, A & Di Santo, J.P (2000) Stable and functional

lymphoid reconstitution of common cytokine receptor gamma

chain deficient mice by retroviral-mediated gene transfer Blood 95,

3071–3077.

27 Cavazzana-Calvo, M., Hacein-Bey, S., de Saint Basile, G., Gross,

F., Yvon, E., Nusbaum, P., Selz, F., Hue, C., Certain, S.,

Casa-nova, J.L., Bousso, P., Deist, F.L & Fischer, A (2000) Gene

therapy of human severe combined immunodeficiency (SCID)-X1

disease Science 288, 669–672.

28 Leonard, W.J (2000) X-linked severe combined

immunodefi-ciency: from molecular cause to gene therapy within seven years.

Mol Med Today 6, 403–407.

29 Raskin, N., Jakubowski, A., Sizing, I.D., Olson, D.L., Kalled,

S.L., Hession, C.A., Benjamin, C.D., Baker, D.P & Burkly, L.C.

(1998) Molecular mapping with functional antibodies localizes

critical sites on the human IL receptor common gamma (gamma c)

chain J Immunol 161, 3474–3483.

30 Olosz, F & Malek, T.R (2000) Three loops of the common

gamma chain ectodomain required for the binding of interleukin-2

and interleukin-7 J Biol Chem 275, 30100–30105.

31 Kumaki, S., Kondo, M., Takeshita, T., Asao, H., Nakamura, M.

& Sugamura, K (1993) Cloning of the mouse interleukin 2

receptor gamma chain: demonstration of functional differences

between the mouse and human receptors Biochem Biophys Res.

Commun 193, 356–363.

32 Shen, B.J., Hage, T & Sebald, W (1996) Global and local

determinants for the kinetics of interleukin- 4/interleukin-4

receptor alpha chain interaction A biosensor study employing

recombinant interleukin-4-binding protein Eur J Biochem 240,

252–261.

33 Sondermann, P & Jacob, U (1999) Human Fcgamma receptor

IIb expressed in Escherichia coli reveals IgG binding capability.

Biol Chem 380, 717–721.

34 Kruse, N., Shen, B.J., Arnold, S., Tony, H.P., Muller, T &

Sebald, W (1993) Two distinct functional sites of human

inter-leukin 4 are identified by variants impaired in either receptor

binding or receptor activation EMBO J 12, 5121–5129.

35 Schreiber, G & Fersht, A.R (1995) Energetics of protein–protein

interactions: analysis of the barnase–barstar interface by

single mutations and double mutant cycles J Mol Biol 248,

478–486.

36 Kruse, N., Tony, H.P & Sebald, W (1992) Conversion of human interleukin-4 into a high affinity antagonist by a single amino acid replacement EMBO J 11, 3237–3244.

37 Schreiber, G., Frisch, C & Fersht, A.R (1997) The role of Glu73

of barnase in catalysis and the binding of barstar J Mol Biol 270, 111–122.

38 Mueller, T., Oehlenschlaeger, F & Buehner, M (1995) Human interleukin-4 and variant R88Q: phasing X-ray diffraction data

by molecular replacement using X-ray and nuclear magnetic resonance models J Mol Biol 247, 360–372.

39 de Vos, A.M., Ultsch, M & Kossiakoff, A.A (1992) Human growth hormone and extracellular domain of its receptor: crystal structure of the complex Science 255, 306–312.

40 Bernstein, F.C., Koetzle, T.F., Williams, G.J.B., Meyer, E.F Jr,, Brice, M.D., Rodgers, J.R., Kennard, O., Shimanouchi, T & Tasumi, M (1977) The Protein Data Bank: a computer-based archival file for macromolecular structures J Mol Biol 112, 535– 542.

41 Jones, T.A., Zou, J.Y., Cowan, S.W & Kjelgaard, M (1991) Improved methods for binding protein models in electron density maps and the location of errors in these models Acta Crystallogr A47, 110–119.

42 Bruenger, A.T (1992) X-Plor, Version 3.1 A System for X-Ray Crystallography and NMR Yale University Press, New Haven,

CT, USA.

43 Jack, A & Levitt, M (1978) Refinement of large structures by simultaneous minimization of energy and R factor Acta Crys-tallogr A34, 931–935.

44 Bamborough, P., Hedgecock, C.J & Richards, W.G (1994) The interleukin-2 and interleukin-4 receptors studied by molecular modelling Structure 2, 839–851.

45 Gustchina, A., Zdanov, A., Schalk-Hihi, C & Wlodawer, A (1995) A model of the complex between interleukin-4 and its receptors Proteins 21, 140–148.

46 Kroemer, R.T & Richards, W.G (1996) Homology modeling study of the human interleukin-7 receptor complex Protein Eng 9, 1135–1142.

47 Clackson, T & Wells, J.A (1995) A hot spot of binding energy in

a hormone–receptor interface Science 267, 383–386.

48 Livnah, O., Stura, E.A., Johnson, D.L., Middleton, S.A., Mulcahy, L.S., Wrighton, N.C., Dower, W.J., Jolliffe, L.K & Wilson, I.A (1996) Functional mimicry of a protein hormone by a peptide agonist: the EPO receptor complex at 2.8 A˚ Science 273, 464–471.

49 Middleton, S.A., Barbone, F.P., Johnson, D.L., Thurmond, R.L., You, Y., McMahon, F.J., Jin, R., Livnah, O., Tullai, J., Farrell, F.X., Goldsmith, M.A., Wilson, I.A & Jolliffe, L.K (1999) Shared and unique determinants of the erythropoietin (EPO) receptor are important for binding EPO and EPO mimetic pep-tide J Biol Chem 274, 14163–14169.

50 Horsten, U., Muller-Newen, G., Gerhartz, C., Wollmer, A., Wijdenes, J., Heinrich, P.C & Grotzinger, J (1997) Molecular modeling-guided mutagenesis of the extracellular part of gp130 leads to the identification of contact sites in the interleukin-6 (IL-6) J Biol Chem 272, 23748–23757.

51 Kurth, I., Horsten, U., Pflanz, S., Dahmen, H., Kuster, A., Grotzinger, J., Heinrich, P.C & Muller-Newen, G (1999) Acti-vation of the signal transducer glycoprotein 130 by both IL-6 and IL-11 requires two distinct binding epitopes J Immunol 162, 1480–1487.

52 Hoffman, R.C., Castner, B.J., Gerhart, M., Gibson, M.G., Rasmussen, B.D., March, C.J., Weatherbee, J., Tsang, M., Gustchina, A., Schalk-Hihi, C., Reshetnikova, L & Wlodawer, A (1995) Direct evidence of a heterotrimeric complex of human interleukin-4 with its receptors Protein Sci 4, 382–386.

53 Baker, D.P., Whitty, A., Zafari, M.R., Olson, D.L., Hession, C.A., Miatkowski, K., Avedissian, L.S., Foley, S.F., McKay,

M.L., Benjamin, C.D & Burkly, L.C (1998) The murine

anti-human common gamma chain monoclonal antibody CP.B8

blocks the second step in the formation of the intermediate affinity

IL-2 receptor Biochemistry 37, 14337–14349.

54 Hiraoka, O., Anaguchi, H., Asakura, A & Ota, Y (1995)

Requirement for the immunoglobulin-like domain of granulocyte

colony-stimulating factor receptor in formation of a 2: 1 receptor–

ligand complex J Biol Chem 270, 25928–25934.

55 Woodcock, J.M., Zacharakis, B., Plaetinck, G., Bagley, C.J.,

Qiyu, S., Hercus, T.R., Tavernier, J & Lopez, A.F (1994) Three

residues in the common beta chain of the human GM-CSF, IL-3

and IL-5 receptors are essential for GM-CSF and IL-5 but not

IL-3 high affinity binding and interact with Glu21 of GM-CSF.

EMBO J 13, 5176–5185.

56 Woodcock, J.M., Bagley, C.J., Zacharakis, B & Lopez, A.F.

(1996) A single tyrosine residue in the membrane-proximal

domain of the granulocyte-macrophage colony-stimulating factor,

interleukin (IL) -3, and IL-5 receptor common beta-chain is

necessary and sufficient for high affinity binding and signaling by

all three ligands J Biol Chem 271, 25999–26006.

57 Haman, A., Cadieux, C., Wilkes, B., Hercus, T., Lopez, A., Clark,

S & Hoang, T (1999) A single tyrosine residue in the

membrane-proximal domain of the granulocyte-macrophage

colony-stimu-lating factor, interleukin (IL)-3, and IL-5 receptor common

beta-chain is necessary and sufficient for high affinity binding and

signaling by all three ligands J Biol Chem 274, 34155–34163.

58 DiSanto, J.P., Dautry-Varsat, A., Certain, S., Fischer, A & de Saint Basile, G (1994) Interleukin-2 (IL-2) receptor gamma chain mutations in X-linked severe combined immunodeficiency disease result in the loss of high-affinity IL-2 receptor binding Eur.

J Immunol 24, 475–479.

59 Ishii, N., Asao, H., Kimura, Y., Takeshita, T., Nakamura, M., Tsuchiya, S., Konno, T., Maeda, M., Uchiyama, T & Sugamura,

K (1994) Impairment of ligand binding and growth signaling of mutant IL-2 receptor gamma-chains in patients with X-linked severe combined immunodeficiency J Immunol 153, 1310–1317.

60 Kumaki, S., Ishii, N., Minegishi, M., Tsuchiya, S., Cosman, D., Sugamura, K & Konno, T (1999) Functional role of interleukin-4 (IL-4) and IL-7 in the development of X-linked severe combined immunodeficiency Blood 93, 607–612.

61 Sharfe, N., Shahar, M & Roifman, C.M (1997) An interleukin-2 receptor gamma chain mutation with normal thymus morpho-logy J Clin Invest 100, 3036–3043.

62 Puck, J.M (1996) A database of gamma c-chain defects causing human X-SCID Immunol Today 17, 507–511.

63 Malek, T.R., Porter, B.O & He, Y.W (1999) Multiple gamma c-dependent cytokines regulate T-cell development Immunol Today 20, 71–76.

64 He, Y.W., Adkins, B., Furse, R.K & Malek, T.R (1995) Expression and function of the gamma c subunit of the IL-2, IL-4, and IL-7 receptors Distinct interaction of gamma c in the IL-4 receptor J Immunol 154, 1596–1605.