doi:10.1006/cyto.2002.1004, available online at http://www.idealibrary.com on

STRUCTURE FUNCTION ANALYSIS OFINTERLEUKIN 7: REQUIREMENT FOR AN

AROMATIC RING AT POSITION 143 OF HELIX D

Johanna C vanderSpek,1,2,3 John A Sutherland,1 Brian M Gill,1 Gullu Gorgun,4

Francine M Foss,4 John R Murphy1,2

The residues located at the carboxyl terminus of helix D in interleukin-7 (IL-7) have previouslybeen targeted as important for recruitment and binding to the � chain component of the IL-7receptor (IL-7R). In this study, Trp 143 of helix D was mutated to His, Phe, Tyr and Pro andthese mutants, along with a W143A mutant previously described, were studied to determine theeffects on activation of DNA synthesis and binding affinity to IL-7R positive 2E8 cells. TheW143F and W143Y mutants were similar to wild type IL-7 in their binding properties andretained 85% and 74% of their activating properties, respectively. In contrast, the W143Hmutant possessed a lower binding affinity and a corresponding decrease in activation, the W143Amutant possessed an over 100-fold decreased binding affinity and some residual activationactivity and the W143P mutant possessed a greatly decreased binding affinity and did notactivate. These results strongly suggest an aromatic residue is required at position 143 for IL-7Rbinding and subsequent signal transduction.

� 2002 Elsevier Science Ltd. All rights reserved.

From the 1Evans Department of Clinical Research and theDepartment of Medicine, Boston Medical Center, Boston,Massachusetts 02118, USA, 2New England Medical Center,Boston, Massachusetts 02118, USA

3Correspondence to: Department of Medicine, Boston MedicalCenter, 650 Albany St., Boston, MA 02118. E-mail: johanna.vanderspek@bmc.org

1043–4666/02/$-see front matter � 2002 Elsevier Science Ltd.All rights reserved.

KEY WORDS: IL-7/mutational analysis3This research was supported in part by Public Health Service Grant

CA-48626 from the NCI, National Institutes of Health.

Interleukin-7 is a cytokine that was originallyidentified as a growth factor for murine B cell pro-genitors in long-term bone marrow cultures.1 It is nowknown to play a critical role in the development of Tand B cells in mice and in T cell development inhumans.2 The IL-7R complex consists of at least twochains, an � chain and the � chain common to thereceptors for IL-2, IL-4, IL-9 and IL-15.3,4 Both chainsare members of the cytokine receptor superfamily. IL-7receptors exist in high, low and ultra low affinity formswith Kds ranging from 38 pM to 10 �M.5–8 Signaltransduction by IL-7 is not well established butinvolves the recruitment and activation of intracellularkinases including the Src and Jak families and PI3–kinase.9–12

IL-7 stimulates the growth of pre-B and T acutelymphoblastic leukemia cells in vitro.13 It also inducesproliferation of chronic lymphocytic leukemia cells and

CYTOKINE, Vol. 17, No. 5 (7 March), 2002: pp 227–233

acute myelogenous leukemia cells, as well as cells frompatients with Sezary syndrome14,15 The responses toIL-7 by malignant cells suggest that IL-7 and theIL-7R might serve as targets for the development ofnovel therapeutics. A model of IL-7 has recently beendescribed in which comparative model building wasperformed using the x-ray crystal structure of IL-4 as atemplate.16 This model suggested that IL-7 exists as afour-helix bundle with an up-up-down-down topology.The authors postulated that helix D was involved withrecruitment of the common � chain for receptor bind-ing and activation. This would also predict that theindoyl ring of the Trp residue at position 143 of helix Dwould be solvent exposed and suggests that the hydro-phobic moment of helix D is oriented toward bulksolvent. In order to further probe this anomalousfeature we have replaced the Trp residue at position143 with either a His, Phe, Tyr or Pro residue. We havestudied the mutants, as well as the pre-existingIL-7(W143A) mutant for stimulation of DNA syn-thesis and binding affinity to 2E8 cells. We demon-strate that the substitution of aromatic residues Phe orTyr for Trp at position 143 of IL-7, results in similarbinding affinities and subsequent stimulation of DNAsynthesis. The presence of a His residue at that positionresults in an intermediate binding affinity anddecreased stimulation. Presence of an Ala or Proresidue at position 143 resulted in mutants with greatly

227

228 / vanderSpek et al. CYTOKINE, Vol. 17, No. 5 (7 March, 2002: 227–233)

decreased binding affinity and stimulation activity.These results indicate that an aromatic residue atposition 143 of helix D is required for IL-7R bindingand is consistent with the hypothesis that helix D isinvolved with binding and activation.

RESULTS

Wild type IL-7 contains a tryptophan residue atposition 143 and the W143 mutants, W143A, W143F,W143H, W143P and W143Y were constructed,expressed, purified and refolded as described under‘‘Materials and Method’’. The purified proteins wereseparated by 17% SDS gel electrophoresis and stainedwith Coomassie Blue. The proteins migrated betweenthe 18 kDa and 14 kDa molecular weight markerswhich is in agreement with their calculated molecularweight of 17 kDa (Fig. 1A). In contrast to commer-cially available recombinant human IL-7 (Sigma),native gel electrophoresis indicated the wild type IL-7prepared in these studies existed mainly in monomericform. Similar mobilities were noted for wild type andmutant fractions tested (data not shown). Western blotanalysis indicated the proteins were all immunoreactivewith anti-IL-7 antibodies (Fig. 1B).

Wild type IL-7 and the mutants were tested forstimulation of DNA synthesis in a [3H]thymidineincorporation assay using murine, IL-7 dependent, 2E8cells. A representative assay is shown in Figure 2.Maximum stimulation of DNA synthesis by wild typeIL-7 occurred at 10�8 and 10�9 M, indicating thataccurate refolding had occurred. The active mutantsIL-7(W143Y), IL-7(W143F) and IL-7(W143H) alsoexhibited maximum stimulation at 10�8 and 10�9 M.

In contrast, IL-7(W143P) was not active andIL-7(W143A) showed slight activity at 10�8 M but notat 10�9 M. The results for the mutants are reportedcompared to activation by wild type at the sameconcentration (Fig. 3).

The results for the binding assays are shown inFigure 4. The concentration of wild type IL-7 requiredto displace 50% of the [125I]-labelled IL-7 (IC50) was2.5�10�8 M. The IC50 for IL-7(W143F) andIL-7(W143Y) were comparable at 1.5�10�8 M and4.0�10�8 M, respectively. IL-7(W143H) bound withless affinity and had an IC50 of approximately2.1�10�7 M. The W143A and W143P mutants didnot bind well and required concentrations of greaterthan 1�10�6 M to compete off 50% of the [125I]-IL-7.

Figure 1. SDS-PAGE and immunoblot analysis of purified proteins.

A, 17% SDS-polyacrylamide gel stained with Coomassie Blue. Lane 1, IL-7(W143P); lane 2, IL-7 (W143A); lane 3, IL-7(W143Y); lane 4, IL-7(W143H); lane 5, IL-7(W143W) and lane 6, IL-7(W143F). Molecular weight standards are indicated (103). B, Immunoblot analysis using ananti-IL-7 antibody. Lane 1, IL-7(W143P); lane 2, IL-7(W143A); lane 3, IL-7(W143Y); lane 4, IL-7(W143H); lane 5, IL-7(W143W) and lane 6,IL-7(W143F).

DISCUSSION

The hematopoietins are proteins with a commonfour-helix bundle, up-up-down-down motif. Typically,helix D of the four-helix bundle is involved withbinding to the common � chain of the receptors.Kroemer et.al., 199617 suggested a model for IL-7 inwhich the hydrophobic face of helix D is presented to acompact hydrophobic core of the protein. The result-ing structure contained a topology similar to the othermembers of the hematopoietin family. A more recentmodel of IL-7, in which the x-ray crystal structure ofIL-4 was used as a template, suggests that helix D ofIL-7 presents a hydrophobic surface to bulk solvent.16

This model placed a hydrophobic Trp residue at posi-tion 143 of helix D in a position where it would beexposed to bulk solvent. This unusual topology sug-gested that the Trp residue might be required for IL-7

Mutational analysis of IL-7 helix D / 229

F

old

acti

vati

on o

ver

con

trol

5

10

20

15

25

30

0�10

Log molarity�11�12 �9 �8

Figure 2. Representative dose response analysis of stimulation of (3H)thymidine incorporation by recombinant human IL-7 and related mutants on2E8 cells.

, IL-7; , IL-7(W143F); +, IL-7(W143Y); , IL-7(W143H); , IL-7(W143P) and X, IL-7(W143A).

function. To test this hypothesis, mutants were con-structed in which the Trp residue at position 143 wasmutated and the mutants were tested for binding toand stimulation of [3H] thymidine incorporation inIL-7R positive 2E8 cells in-vitro. The IL-7 mutants,IL-7(W143F) and IL-7(W143Y) possessed the aro-matic residues Phe or Tyr, instead of Trp at position143. These proteins possessed similar binding affinitiesand slightly diminished stimulation of DNA synthesiscompared to wild type IL-7. The IL-7(W143A) andIL-7(W143P) mutants possessed much lower bindingaffinities and corresponding decreased stimulation(Figs. 2, 3, 4 and Table 1). The IL-7(W143H) mutationpossessed approximately a 10-fold decreased bindingaffinity and an intermediate activation activity. Thesefindings support the hypothesis that an aromatic resi-due at position 143 may be required for � chain bindingand activation.

The aromatic side chains are known to packneatly against nearby � strand backbones.18 The �chain component of the IL-7 receptor is composed ofseven � strands.19 If Trp143 is indeed involved withbinding to the � chain then substitution with a non-aromatic residue would be expected to diminish thisinteraction, resulting in decreased binding and acti-vation. Our results are consistent with this inter-

pretation. Pro is a known helix breaker and itsintroduction into helix D would be anticipatedto disrupt structure and affect receptor binding.Importantly, substitution of W147 with an Ala residuealso affected receptor binding. Since Ala readily adaptsto �-helical and �-sheet structure, it is unlikely that anAla substitution at W143 would alter the tertiarystructure of helix D. Were the side of Ala buried in thehydrophobic core of IL-7 in a fashion similar to W147as modelled by Kroemer et al.17 one would not antici-pate changes in either IL-7 (W143A) binding or stimu-lation of DNA synthesis. Previous studies on NALM-6cells suggested that an Ala residue at position 143resulted in an IL-7 mutant with a binding affinity of10-fold less than wild type.16 This study indicates thatan Ala at position 143 results in an IL-7 molecule with�100-fold less binding affinity in 2E8 cells. The bind-ing was weak and did not displace radiolabelled IL-7,nor was the activation by wild type IL-7 inhibitedby competitive binding of IL-7(W143A) (data notshown). The IL-7(W143A) mutant might not providethe necessary packing interaction required for high-affinity binding. The intermediate binding and acti-vation of the IL-7(W143H) mutant indicates that a Hisresidue, which is also an aromatic ring structure, maynot pack as well against the IL-7 receptor � chain.

230 / vanderSpek et al. CYTOKINE, Vol. 17, No. 5 (7 March, 2002: 227–233)

%

Con

trol

act

ivat

ion

10

20

30

110

100

90

80

70

60

50

40

120

0A P H

IL-7 mutant (W143)F Y

Figure 3. Wild type activation versus mutant.

, activation at 10�9 M and , activation at 10�8 M.

% M

axim

um

bin

din

g

10

20

30

70

40

50

60

80

90

0�6.5 �6.0 �5.5 �5.0 �4.5�7.0�7.5�8.0�8.5

Log molarity inhibitor

Figure 4. Competitive displacement of 125I-labelled IL-7.

, IL-7; , IL-7(W143F); +, IL-7(W143Y); , IL-7(W143H); , IL-7(W143P) and X, IL-7(W143A).

Unlike Trp, Tyr and Phe at neutral pH, His is slightlypositive in charge and this could affect receptor bindingas hydrophobic interactions with solvent might not

exist as a force to cause packing. In summary, theresults of this study suggest that the aromatic Trpresidue at position 143 of helix D in IL-7 is involved

Mutational analysis of IL-7 helix D / 231

with binding and biologic activity. The binding isfacilitated by packing of the hydrophobic aromaticring to the � subunit of the IL-7 receptor.

MATERIALS AND METHODS

Bacterial strains, plasmids and eukaryotic celllines

The bacterial strains, plasmids and eukaryotic cell linesare listed in Table 2. The plasmids are all pET11d derivatives(Novagen, Madison WI, USA) and induction of proteinexpression was under control of the T7 promoter. The parentplasmid on which mutagenesis was performed was pES5which encodes wild type hIL-7.20 Eschericia coli TOP 10 was

the host for plasmid propagation and HMS174 was the hostwhen protein inductions were performed.

Site-directed mutagenesis of hIL-7Mutagenesis was performed using a QuikChange site-

directed mutagenesis kit (Stratagene, techservices@stratagene.com). Oligonucleotides were purchased fromGibco Life Sciences (primers@lifetech.com). The oligo-nucleotide pairs used to introduce the mutations are listed inTable 3. The conditions used for the PCR mutagenesis were:One cycle at 95�C for 30 s followed by 25 cycles of 95�C, 30 s,37�C, 1 min and 68�C for 1 min. The IL-7(W143A) mutantwas created previously by Ala scanning mutagenesis.16

TABLE 1. Comparison of binding IC50s and % wild typeactivation for hIL-7 and mutants

IL-7 or mutant IC50

% activation at 10�9 Mcompared to 10�9

M wild type

IL-7 2.5�10�8 M 100IL-7(W143F) 1.5�10�8 M 86IL-7(W143Y) 4.0�10�8 M 74IL-7(W143H) 2.1�10�7 M 26IL-7(W143A) >10�6 M 4IL-7(W143P) >10�6 M 1

TABLE 2. Plasmids, bacteria and eukaryotic cells

Plasmids, eukaryotic cellsand bacterial strains Protein product or description source or reference

pES5 human IL-7 Cosenza et al., 1997pET IL-7(W143P) IL-7(W143P) This studypET IL-7(W143A) IL-7(W143A) Cosenza et al., 2000pET IL-7(W143Y) IL-7(W143Y) This studypET IL-7(W143H) IL-7(W143H) This studypET IL-7(W143W) IL-7(W143W) This studypET IL-7(W143F) IL-7(W143F) This study2E8 cells murine immature B lymphocytes

IL-7 dependentATCC TIB-239

E. Coli TOP10 F�mcrA � (mrr-hsdRMS-mcrBC)�80 LacZ�M15�lacX74 deoRrecA1 endA1

Invitrogen

E. Coli HMS174 F�, recA, rK12�, mK12

+, Rif Novagen

TABLE 3. Oligonucleotides used for site-directed mutagenesis of hIL-7

Oligonucleotide protein product

GAAATCAAAACTTGC CCG AACAAAATCCTGATG IL-7(W143P)CATCAGGATTTTGTT CGG GCAAGTTTTGATTTCGAAATCAAAACTTGC TAC AACAAAATCCTGATG IL-7(W143Y)CATCAGGATTTTGTT GTA GCAAGTTTTGATTTCGAAATCAAAACTTGC CAC AACAAAATCCTGATG IL-7(W143H)CATCAGGATTTTGTT GTG GCAAGTTTTGATTTCGAAATCAAAACTTGC TTC AACAAAATCCTGATG IL-7(W143F)CATCAGGATTTTGTT GAA GCAAGTTTTGATTTC

Expression and purification of hIL-7 and mutantsIL-7 and mutant forms were expressed from the T7

promoter in host HMS174. Expression was induced by theaddition of � coliphage derivative, CE6. The proteinsexpressed under these conditions formed inclusion bodiesand were subjected to three washes as described previously21.The final inclusion body pellet was resuspended in 2.0 ml ofdenaturing solution (7M Guanidine Hydrochloride, 50 mMTris-Cl, pH 8.8, 20 mM dithiothreitol) and pulse sonicatedfor 10 min , 10% cycle, 1.5 s on, 1.0 s off using a FisherScientific ). The sample was then placed in microcentrifugetubes and centrifuged at high speed for 10 min. The super-natent was loaded onto a high resolution, Sephacryl S-200(Pharmacia) sizing column that had been equilibrated withdenaturing buffer. Elution was in the same denaturing buffer.

232 / vanderSpek et al. CYTOKINE, Vol. 17, No. 5 (7 March, 2002: 227–233)

0.5 ml Fractions were refolded by dialysis against 50 mMTris-Cl, pH 8.0, 50 mM NaCl, 5 mM reduced glutathione,1 mM oxidized glutathione, 100 mM L-Arg. SDS denaturinggel electrophoresis was performed to determine which frac-tions contained the purified IL-7 and mutants. Selectedfractions were also subjected to native gel electrophoresis.Western blot analysis was performed using anti-human IL-7antibody (R & D systems, Minneapolis MN, USA). Proteinconcentration was determined using Pierce Reagent (Pierce,Rockford, IL, USA) and purified fractions were stored at�70 �C. The purified IL-7, IL-7(W143A), IL-7(W143H),IL-7 (W143F), IL-7 (W143P) and IL-7 (W143Y) andimmunoblot are shown in Figure 1.

Activation assays2E8 cells (immature murine B lymphocytes, IL-7

dependent) were maintained in RPMI 1640 (HycloneLaboratories, Logan UT) supplemented with 5% fetal bovineserum, 2 mM glutamine, 50 �g/ml streptomycin, 50 IU/mlpenicillin, 0.05 mM 2–mercaptoethanol and 1 nM recom-binant hIL-7. For the assays the 2E8 cells were washed twicein the above medium without IL-7 and then seeded, in thesame medium, without IL-7 at a concentration of 1�105

cells per well of a 96 well, V-bottomed, microtitre plate(Linbro). Wild type IL-7 or the mutant IL-7s were added tofinal concentrations ranging from 10�7 down to 10�12 M.The cells were incubated at 37�C, 5% CO2 for 48 h and then1 �Ci [3H] thymidine (6.7 Ci/mmole, DuPont, NEN) wasadded to each well. The cells were pulsed for 6 h at 37�C, 5%CO2 and then centrifuged at 170�g for 5 min. The mediumwas aspirated from above the cell pellets and the cells werelysed in 60 �l 0.4 M KOH. The samples were incubated atroom temperature for 10 min and then 140 �l 10% trichloro-acetic acid was added. The samples were incubated anadditional 10 min and then harvested on glass fibrefilters using a PhD cell harvester (Cambridge Technology,Cambridge MA, USA). Radioactivity was determined usingstandard methods. Each molecule was tested in at least twoassays and assays were performed in quadruplicate.

Binding assaysDisplacement of [125I]-labelled recombinant IL-7 from

2E8 cell receptors was basically performed following theprocedure of Wang and Smith22. The IL-7 was iodinated andpurified using IODO-GEN tubes (#28601ZZ) and D-Saltcolumns (#43243ZZ) from Pierce (Rockford, IL, USA).I-125 Radionuclide (NEZ033A) was purchased from NewEngland Nuclear Life Science Products (Boston, MA, USA).One nmole of IL-7 was iodinated following Pierce instruc-tions. The concentration of radiolabelled IL-7 was deter-mined by activation assays. The 2E8 cells were washed incomplete RPMI 1640, without IL-7 and then 3�107 2E8cells were incubated with 300 pM [125I]-labelled recombinantIL-7 in the presence of varying concentrations of IL-7 or therelated IL-7 (W143) mutants. The cells were overlaid on amixture of 80% 550 fluid (Accumetric Inc., Elizabethtown,KY, USA) 20% paraffin oil (density=1.03 g/ml) in 0.4 mlpolyethylene tubes (Fisher Scientific, Cat. #02–681–229). Thecells were incubated on ice for 1 h and then the tubescentrifuged for 2 min. The cell pellets contained the bound

ligand and the aqueous phase which remained on top of theoil contained the unbound ligand. Samples were counted in aBeckman Gamma 5500 counter.

REFERENCES

1. Namen AE, Lupton S, Hjerrild K, Wignall J, MochizukiDY, Schmierer A, Mosley B, March CJ, Urdal D, Gillis S (1988)Stimulation of B-cell progenitors by cloned murine interleukin-7.Nature 333:571–573.

2. Hofmeister R, Khaled AR, Benbernou N, Rajnavolgyi E,Muegge K, Durum SK (1999) Interleukin-7: physiological roles andmechanisms of action. Cytokine Growth Factor Rev 10:41–60.

3. Noguchi M, Nakamura Y, Russell SM, Ziegler SF, TsangM, Cao X, Leonard WJ (1993) Interleukin-2 receptor gamma chain:a functional component of the interleukin-7 receptor. Science 262:1877–1880.

4. Kondo M, Takeshita T, Higuchi M, Nakamura M, Sudo T,Nishikawa S, Sugamura K (1994) Functional participation of theIL-2 receptor gamma chain in IL-7 receptor complexes. Science263:1453–1454.

5. Page TH, Lali FV, Groome N, Foxwell BM (1997)Association of the common gamma-chain with the human IL-7receptor is modulated by T cell activation. J Immunol 158:5727–5735.

6. Foxwell BM, Taylor-Fishwick DA, Simon JL, Page TH,Londei M (1992) Activation induced changes in expression andstructure of the IL-7 receptor on human T cells. Int Immunol 4:277–282.

7. Park LS, Friend DJ, Schmierer AE, Dower SK, Namen AE(1990) Murine interleukin 7 (IL-7) receptor. Characterization on anIL-7-dependent cell line. J Exp Med 171:1073–1089.

8. Armitage RJ, Ziegler SF, Friend DJ, Park LS, Fanslow WC(1992) Identification of a novel low-affinity receptor for humaninterleukin-7. Blood 79:1738–1745.

9. Seckinger P, Fougereau M (1994) Activation of src familykinases in human pre-B cells by IL-7. J Immunol 153:97–109.

10. Venkitaraman AR, Cowling RJ (1992) Interleukin 7 recep-tor functions by recruiting the tyrosine kinase p59fyn through asegment of its cytoplasmic tail. Proc Natl Acad Sci USA 89:12083–12087.

11. Zeng YX, Takahashi H, Shibata M, Hirokawa K (1994)JAK3 Janus kinase is involved in interleukin 7 signal pathway. FEBSLett 353:289–293.

12. Dadi H, Ke S, Roifman CM (1994) Activation ofphosphatidylinositol-3 kinase by ligation of the interleukin-7 recep-tor is dependent on protein tyrosine kinase activity. Blood 84:1579–1586.

13. Tuow I, Pouwels K, van Agthoven T, van Gur R, Budel L,Hoogerbrugge H, Delwel R, Goodwin R, Namen A, Lowenberg B(1990) Interleukin-7 is a growth factor of precursor B and T acutelymphoblastic leukemia. Blood 75:2097–2101.

14. Digel W, Schmid M, Heil G, Conrad P, Gillis S, Porzsolt F(1991) Human interleukin-7 induces proliferation of neoplastic cellsfrom chronic lymphocytic leukemia and acute leukemias. Blood 78:753–759.

15. Merle-Beral H, Schmitt C, Mossalayi D, Bismuth G,Dalloul A, Mentz F (1993) Interleukin-7 and malignant T cells.Nouv Rev Fr Hematol 35:231–232.

16. Cosenza L, Rosenbach A, White JV, Murphy JR, Smith T(2000) Comparative model building of interleukin-7 usinginterleukin-4 as a template: a structural hypothesis that displaysatypical surface chemistry in helix D important for receptoractivation. Protein Sci 9:916–926.

17. Kroemer RT, Doughty SW, Robinson AJ, Richards WG(1996) Prediction of the three-dimensional structure of humaninterleukin-7 by homology modeling. Protein Eng 9:493–498.

Mutational analysis of IL-7 helix D / 233

18. Richardson JS, Richardson DC (1989) Principals andPatterns of Protein Conformation. In Fasman GD (ed) Prediction ofProtein Structure and the Principals of Protein Conformation.Plenum Press, New York. Pp. 1–99.

19. Raskin N, Jakubowski A, Sizing ID, Olson DL, Kalled SL,Hession CA, Benjamin CD, Baker DP, Burkly LC (1998) Molecularmapping with functional antibodies localizes critical sites on thehuman IL receptor common gamma (gamma c) chain. J Immunol161:3474–3483.

20. Cosenza L, Sweeney E, Murphy JR (1997) Disulfide bondassignment in human interleukin-7 by matrix-assisted laser

desorption/ionization mass spectroscopy and site-directed cysteine toserine mutational analysis. J Biol Chem 272:32995–33000.

21. vanderSpek JC, Mindell JA, Finkelstein A, Murphy JR(1993) Structure/function analysis of the transmembrane domain ofDAB389-interleukin-2, an interleukin-2 receptor-targeted fusiontoxin. The amphipathic helical region of the transmembrane domainis essential for the efficient delivery of the catalytic domain to thecytosol of target cells. J Biol Chem 268:12077–12082.

22. Wang HM, Smith KA (1987) The interleukin 2 receptor.Functional consequences of its bimolecular structure. J Exp Med166:1055–1069.