- Email: [email protected]
IDENTIFICATION OF REGIONS WITHIN THE THIRD FnIII-LIKE DOMAIN OF THE IL-5Rα INVOLVED IN IL-5 INTERACTION
doi:10.1006/cyto.1999.0663, available online at http://www.idealibrary.com on
IDENTIFICATION OF REGIONS WITHIN THE THIRD FnIII-LIKE DOMAIN OF THE IL-5R INVOLVED IN IL-5 INTERACTION Peter E. Czabotar,1 John Holland,2 Colin J. Sanderson1 Previously, two binding sites for interleukin 5 (IL-5) were identified on the IL-5 receptor alpha chain (IL-5R). They are located within the CD loop of the first fibronectin type III (FnIII)-like domain and the EF loop of the second FnIII-like domain. The first binding site was identified by exploiting the different abilities of human IL-5R (hIL-5R) and mouse IL-5R (mIL-5R) to bind hIL-5. Here we show that ovine IL-5 (oIL-5) has the ability to activate the hIL-5R but not the mIL-5R. By using chimeras of the mIL-5R and hIL-5R we demonstrate that residues within the first and third FnIII-like domains of mIL-5R are responsible for this lack of activity. Furthermore, mutation of residues on hIL-5R to mIL-5R within the predicted DE and FG loop regions of the third FnIII domain reduces oIL-5 activity. These results show that regions of the third FnIII domain of IL-5R are involved in binding, in addition to the regions in domains one and two of the IL-5R that were identified in an earlier study. 2000 Academic Press
Interleukin 5 (IL-5) is a cytokine produced by activated T lymphocytes. This cytokine has a pivotal role in the differentiation and maturation of eosinophils,1 the cells that play a key part in the pathogenesis of asthma and other allergic diseases. For these reasons IL-5 is considered to be an attractive drug target2 and understanding the interaction between IL-5 and its receptor is an important step towards the discovery and design of IL-5 antagonists. IL-5 is one of a group of structurally related cytokines and growth factors. These proteins are composed of four helices arranged in a configuration known as the cytokine helical bundle. IL-5 is unique within this group in that it exists as a homodimer stabilized by disulphide bridges between residues 42 and 84,3,4 and anti-parallel -sheet formation between residues 23–35 and 89–92.5 The IL-5 receptor is composed of a ligand-specific -chain (IL-5R) and a -chain (c) shared with the IL-3 and GMCSF receptor complexes.6,7 IL-5 displays From the 1Molecular Immunology Group, School of Biomedical Sciences, Curtin University of Technology, Perth, WA 6000, Australia; 2Urological Research Centre, QEII Medical Centre, Nedlands, WA 6009, Australia Correspondence to: Peter Czabotar, Division of Virology, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK; E-mail: [email protected] Received 17 November 1999; accepted for publication 20 December 1999 2000 Academic Press 1043–4666/00/070867+06 $35.00/0 KEY WORDS: cytokine receptor/interleukin 5/IL-5/IL5-R CYTOKINE, Vol. 12, No. 7 (July), 2000: pp 867–873
a binding affinity of approximately 500 pM for the IL-5R subunit. c does not provide any detectable affinity for IL-5 in the absence of the IL-5R, but does increase the affinity of the IL-5–IL-5R complex by 2–3-fold. IL-5R is a member of the class 1 cytokine receptor family.8 The extracellular region of these molecules consists of repeats of the fibronectin type III (FnIII)-like domain. IL-5R has three such repeats conventionally labelled from the membrane distal domain. Although the crystal structure of the IL-5R is yet to be determined its structure can be modelled on other members of the class 1 cytokine receptor family.9,10 The residues involved in receptor interactions have been well mapped in IL-5.11–14 However, the corresponding binding sites on IL-5R are not fully understood. This is mainly because the large size of the molecule places limitations on alanine scanning projects. Thus, making use of species differences is potentially more practical, as large numbers of residues are conserved and thus less mutations need to be generated. This technique has previously been used to identify residues within the CD loop of the first FnIIIlike domain that are important in ligand interactions.11 In addition, sequence homology to a growth hormone receptor contact point has been used to identify an IL-5 contact point in the EF loop of the second FnIII-like domain of the IL-5R. In this paper differences between ovine IL-5 (oIL-5) and mouse IL-5 867
868 / Czabotar et al.
Localization of the region in mIL-5R responsible for the lack of oIL-5 activity Chimeras of mIL-5R and hIL-5R (Fig. 2A) were transfected into BAF-B03 cells and selected with mIL-5. Murine IL-5 has the same affinity for all of these chimeras11 and so this selection regime will produce cell populations that express the IL-5R chimeras. Bulk cultures were used in all experiments to minimize differences due to gene copy or positional effects. None of the cells expressing chimeric receptors responded significantly to oIL-5 although they all responded to mIL-5 (Fig. 2B). Thus, several areas of the receptor are involved in the species difference, the previously described FnIII domain 1, and interestingly, the third FnIII domain. As the third FnIII domain has not previously been shown to be involved in ligand interactions it was examined in more detail.
Analysis of loop regions of FnIII domain 3 of the IL-5R Further analysis was concentrated on residues within the loop regions of the IL-5R predicted to be orientated towards IL-5. As multiple sites were responsible for the lack of oIL-5 activity on the mIL-5R it was possible that mutation of a single site within FnIII domain 3 may not result in any detectable oIL-5 activity. For this reason a number of mutations were performed on the hIL-5R to make the third FnIII
H-thymidine inc. (CPS)
Cell proliferation assays using the murine IL-5 responsive cells B13 and BAF (hIL-5R) were performed with Cos-7-expressed oIL-5 and mIL-5. Both of these cell types have the murine cytokine receptor chain, c, the B13 cells express mIL-5R and BAF-B03 cells were transfected with hIL-5R to make the BAF (hIL-5R) line. Thus, any differences in the activities of oIL-5 and mIL-5 are due to differences in the ability of these cytokines to interact with the particular IL-5R expressed by the cell lines. It was found that oIL-5 recognized hIL-5R and triggered proliferation of BAF (hIL-5R) cells to an extent equal to, or exceeding, that observed with mIL-5. In contrast, oIL-5 did not recognize mIL-5R and could not support the growth of B13 cells (Fig. 1). Thus there are regions of the mIL-5R that differ from those in the hIL-5R and these differences prohibit interaction with oIL-5.
A
40 000
20 000
3
Activity of oIL-5 in the mIL-5R and hIL-5R bioassay
0
60 000 H-thymidine inc. (CPS)
RESULTS
60 000
100
1000 Dilution
10 000
100
1000 Dilution
10 000
B
40 000
20 000
3
(mIL-5) activity have been used to identify loop regions within the third FnIII domain of IL-5R that interact with IL-5.
CYTOKINE, Vol. 12, No. 7 (July, 2000: 867–873)
0
Figure 1. Activity of mIL-5 (shaded bars) and oIL-5 (open bars) in the hIL-5R (A) and mIL-5R (B) bioassays. Values represent counts/s of incorporated 3H-thymidine as described in the methods section. These results represent the activity of varying dilutions of Cos-7 supernatant containing the respective cytokine. Error bars represent the standard deviation of duplicates. These results are representative of at least three separate experiments.
domain more like that of the mouse. The residues mutated are shown in Figure 3. The mutated hIL-5Rchains were transfected into BAF-B03 cells, the cells selected for expression using mIL-5. The ability of oIL-5 to trigger proliferation of cells expressing the mutant form of hIL-5R was compared to that obtained with cells expressing the wild-type hIL-5R. A reduction in the proliferative response would indicate that the mutated residues were involved in binding IL-5. As expected, all cells expressing the mutated hIL5R-chains displayed a similar response to mIL-5 and this resembled that obtained with cells expressing the wild-type hIL-5R (Fig. 4). In contrast, marked differences were seen when oIL-5 was used. Cells expressing mutant 1 and mutant 2 displayed a similar response to oIL-5 as cells expressing wild-type hIL-5R (Fig. 5), but cells expressing mutant 3 and mutant 4 exhibited a reduction in oIL-5 responsiveness. The combination of mutations 3 and 4 further reduced the ability of cells to respond to oIL-5. Thus, the residues on hIL-5R mutated in mutant 3 and mutant 4 are involved in the interaction of oIL-5 with hIL-5R.
Binding sites on the IL-5R / 869
A FnIII dom 1 FnIII dom 2
FnIII dom 3 TM
Chimera 1 ClaI
TM
Chimera 2 ClaI
TM
Chimera 3 ClaI
TM
Chimera 4 MluI B
30 000 20 000 10 000
hIL5Rα
Chimera 4
Chimera 3
Chimera 2
0
Chimera 1
3
H-thymidine inc. (CPS)
40 000
Figure 2. (A) Schematic representation of the extracellular domains of hIL-5R/mIL-5R chimeras. Grey areas represent regions of hIL-5R sequence and solid areas represent regions of mIL-5R sequence. The location of the three FnIII-like domains are indicated. The transmembrane domain (TM) and the introduced restriction sites used to construct these chimeras11 are also shown. (B) Activity of mIL-5 (shaded bars) and oIL-5 (open bars) on cells expressing chimeras of hIL-5R and mIL-5R. Values represent counts/s of incorporated 3H-thymidine. These values represent the activity of these cells in response to a 1/333 or a 1/100 dilution of Cos-7 supernatants containing either mIL-5 or oIL-5 respectively. These results are representative of three separate experiments.
DISCUSSION Murine IL-5 and hIL-5 share 70% sequence similarity at the amino acid level. Both bioassays and direct binding assays indicate that mIL-5 is able to interact with the hIL-5R with the same efficacy as hIL-5.15,16 However, hIL-5 interacts with the mIL-5R with only 1% of the binding and biological activity displayed by mIL-5.11,16 This difference in activity has previously been used to identify regions of the IL-5R involved in interacting with IL-5.11 We have shown here that like hIL-5, oIL-5 is less able to interact with the mIL-5R than the hIL-5R, indicating there are different amino acids in the hIL-5R and the mIL-5R at the IL-5:IL5R contact points, which restrict the interaction oIL-5 of with mIL-5R. To localize the points of contact between IL-5 and the IL-5R-chain we developed an oIL-5 bioassay using cells expressing chimeras of the hIL-5R and mIL-5R. This assay revealed that contact points,
between IL-5 and the IL-5R-chain, were located in the first and third FnIII-like domains. Further analysis was concentrated on the third domain. In particular, we analysed the loop regions of this domain which, on other members of the class 1 cytokine receptor family, are located on the ligand binding face of the molecule.17–19 Cells expressing hIL-5R mutated in either the DE (Mut 3) or FG (Mut 4) loops of FnIII domain 3 were less able to respond to oIL-5 than cells expressing wild-type hIL-5R. This activity was further reduced when hIL-5R was mutated at both of these loop regions. Clearly, the mutations made to hIL-5R in the DE and FG loops of FnIII domain 3 disrupts the binding sites for oIL-5. Furthermore, the additive effect of mutating both loop regions indicates that the amino acids altered in Mut3 and Mut4 disrupt two separate points at which IL-5 makes contact with the IL-5R-chain. The FG loop region of the membrane proximal FnIII-like domain, on other members of the class 1 cytokine receptor family, has been shown to be involved in ligand interactions. This has been demonstrated by crystallographic analysis of receptor:ligand complexes for the growth hormone receptor (GHR),17 prolactin receptor (PRLR)18 and erythropoietin receptor (Epo R).19 In addition, mutagenesis has been used to confirm the importance of this loop region on the IL-6R-chain,20,21 granulocyte colony stimulating factor receptor (GCSFR)22 and the granulocyte macrophage colony stimulating factor receptor -chain (GMR-).10 The importance of this loop in GMR- is of particular interest as this molecule is closely related to IL-5R. In addition, the residue identified as being involved in ligand contact in GMR- (Arg-280) has been conserved in IL-5R (Arg-297). In light of these findings it is reasonable to expect that residues within the FG loop of IL-5R are acting as contact points for IL-5. Unlike the FG loop, the DE loop of the membrane proximal FnIII-like domain has not been found to play a direct role in ligand interactions in other members of the class 1 cytokine receptor family. X-ray crystallography of the growth hormone GH:GHR complex and the GH:PRLR complex has demonstrated that this loop region is not directly involved in ligand interaction in these instances.17,18 Mutagenesis analysis of other class 1 receptor family members has concentrated only on those loop regions which have been shown to be important when GH contacts GHR. As a result the DE loop of the membrane proximal FnIII domain, and other loop regions not used by GHR, have been largely ignored in the structure/function analyses of these proteins. Although our observations implicate the DE loop region of the membrane proximal FnIII-like domain as
870 / Czabotar et al.
Figure 3.
CYTOKINE, Vol. 12, No. 7 (July, 2000: 867–873)
Amino acid sequence of the third FnIII-like domain of the hIL-5R.
Those residues differing on the mIL-5R are displayed above the sequence. The location of the predicted -pleated sheets are labelled A–G.8 Those residues mutated from hIL-5R to mIL-5R residues on hIL-5R mutants are highlighted and labelled. The location of the ClaI site used to make chimera 1 and the beginning of the transmembrane domain (TM) are indicated in italics. The conserved WSXWS motif is boxed.
16 000 H-thymidine inc. (CPS)
20 000 15 000 10 000 5000 0 10
12 000
8000
4000
3
3
H-thymidine inc. (CPS)
25 000
0
100
1000 10 000 Reciprocal dilution
100 000
2
10
100 1000 Reciprocal dilution
10 000
Figure 4. Titrations of mIL-5 on cells expressing hIL-5R (+), Mut 1 ( ), Mut 2 ( ), Mut 3 ( ), Mut 4 ( ) or Mut 3+4 ( ).
Figure 5. Titrations of oIL-5 on cells expressing hIL-5R (+), Mut 1 ( ), Mut 2 ( ), Mut 3 ( ), Mut 4 ( ) or Mut 3+4 ( ).
Response values represent counts/s of incorporated 3H-thymidine. Cytokine concentration is expressed as a dilution of Cos-7 cell supernatant containing mIL-5. Error bars represent the standard deviation of triplicates. The ED50 values for these curves are reciprocal dilutions of 1061 (hIL-5R), 1146 (Mut 1), 732.6 (Mut 2), 1108 (Mut 3), 1040 (Mut 4) and 1050 (Mut 3+4). These results are representative of at least two separate experiments.
Response values represent counts/s of incorporated 3H-thymidine. Cytokine concentration is expressed as a dilution of Cos-7 cell supernatant containing oIL-5. Error bars represent the standard deviation of triplicates. The ED50 values for these curves are reciprocal dilutions of 426.3 (hIL-5R), 647.8 (Mut 1), 492.2 (Mut 2), 172.4 (Mut 3), 60.26 (Mut 4) and 22.17 (Mut 3+4). These results are representative of at least two separate experiments.
playing a part in ligand contact it is possible that its role is structural. Mutation of this region may be disrupting the orientation of a ligand contact point in another region of IL-5R. Numerous observations support this possibility. Firstly, modelling IL-5R to the structure of EpoR reveals that the residues mutated in the DE loop, in the work reported here, may influence the orientation of the BC loop region (Fig. 6). In particular, intramolecular interactions may occur between Thr-271, which is predicted by the modelling to lie within the E -strand, and residues located within the B -strand. Secondly, the BC loop of the membrane proximal FnIII domain is utilized by numerous class 1 cytokine receptors for ligand interaction including GHR,17 PRLR,18 EpoR,19,26 IL-6R20,21 and GCSFR.22 Thirdly, the BC loop region of IL-5R is highly conserved between the human and mouse IL-5R, indicating that gross alterations to this region
have not been tolerated during evolution. This would be expected if it was an important region of the molecule. This high degree of conservation resulted in only one residue of the BC loop being mutated in this work (Mutation 2, Ile-247-Asp). Further analysis is required to determine whether a ligand contact point in the BC loop, DE loop, or some other region of the IL-5R is being disrupted by mutation 3. Such work could include the alanine scanning of these loops. In conclusion, this is the first demonstration that the third FnIII-like domain of the IL-5R is involved in interactions with IL-5. We have provided evidence that the DE and FG loop regions of this domain are involved in these interactions although further work is required to elucidate the actual role they play. This understanding is important for the discovery and design of IL-5 antagonists which are potentially useful for the treatment of asthma and other allergic diseases.
Binding sites on the IL-5R / 871
generously provided by Dr Jan Tavernier (Ghent, Belgium). Chimeras 1–4 used here are the equivalent of H1, H2, H4 and H5 used previously.11 The plasmid pEE6 containing cDNA that encodes hIL-5R was mutated using the Altered Sites Mutagenesis Kit (Clonetech). Mutations were confirmed by sequencing. The selection primer used in mutagenesis was designed to remove a unique EcoR1 site. Mutagenesis primers were designed to mutate the predicted loop regions of the third FnIII domain of the hIL-5R to a mIL-5R sequence. Selection primer: 5 -GCGAGCTCGAC TTCATTGATCCG-3 Mut 1: 5 -GCCATTGATCAAGTAAATCCTCC ACGGAATGTCACAGC-3 Mut 2: 5 -GTGTCTGCTTTTCCAGACCA TTGCTTTGATTATG-3 Mut 3: 5 -GCAGATAGAAAAATTGATCGCC AATAAATTCATCTCAATAATTGATG-3 Mut 4: 5 -GCAGTGAGCTCCCCGTGCAGAAT GCCAGGGCTCTGGAGTG-3 Figure 6. Predicted molecular structure of the third FnIII domain of the IL-5R. Structural co-ordinates were obtained by modelling the sequence from hIL-5R to the EpoR19 using the Swiss-Model database.23,24 This was then viewed with Rasmol.25 Viewing orientation is from the predicted point of view of the ligand. The molecule is displayed in structural format with -pleated sheets being detailed by yellow arrows. The loop regions discussed in the text are coloured red (B –C loop), green (D –E loop) and purple (F –G loop). Residues mutated in this work from hIL-5R to mIL-5R are labelled and displayed in spacefill format.
MATERIALS AND METHODS Mammalian expression vectors encoding IL-5 IL-5 DNA encoding oIL-5 (generously supplied by Dr Seow from CSIRO, Australia) and mIL-5 was cloned into the mammalian expression vector pEE6.27 These clones were sequenced and found to be identical to the published genebank sequence except for nucleotides 247 (A–G substitution), 386 (C–G substitution) and 387 (G–C substitution) of oIL-5. These differences result in a substitution from a lysine to glutamate at amino acid 66 and from threonine to serine at amino acid 112.
Production of IL-5 in Cos-7 cell cultures Fifteen micrograms of plasmid encoding IL-5 was electroporated into 2106 Cos-7 cells (260 V, 960 F). These cells were transferred to tissue culture flasks containing 5 ml of RPMI+5% FCS and incubated for 72 h at 37C and 5% CO2. The supernatants were then harvested and centrifuged at 2700g for 10 min to remove cell debris.
Plasmids expressing chimeric and mutant forms of the IL-5R The plasmid pSV-SPORT containing cDNA encoding chimeric forms of the human and murine IL-5R were
Assessment for hIL-5R and mIL-5R activity of oIL-5 and mIL-5 Cos-7 supernatants containing oIL-5 and mIL-5 were assessed for activity using mIL-5 and hIL-5 bioassays. The mIL-5R bioassay was performed using B13 cells, a mIL-5dependent cell line which contains the mIL-5R and mc.28 For the hIL-5R bioassay we used the BAF-B03 cell line transfected with the hIL-5R. BAF-B03 cells are a subclone of the mIL-3-dependent pro-B cell line BA/F329 and therefore already posses mc. When forced to express the hIL-5R they are able to proliferate in response to IL-5.30 These cells were washed three times in RPMI+10% FCS to remove residual IL-5 from the culture medium. BAF (5103) (hIL5R) or 1104 B13 cells were then added to the wells of 96-well plates containing 50 l of test substrate at varying dilutions and incubated at 37C and 5% CO2. After 48 h, 10 l of RPMI+10% FCS containing 0.33 Ci of 3 H-thymidine was added to each well and the cells incubated a further 4 h. The cells were then harvested onto glass fibre filters and level of 3H-thymidine incorporation determined on a Packard Matrix 9600 direct -counter.
Production of BAF-B03 cells expressing IL-5R BAF-B03 cells (106) were electroporated with 15 g of mammalian expression vector encoding either WT, chimeric or mutant hIL-5R (960 F, 260 V). Cells were allowed to recover from transfection in recovery medium (RPMI+10% FCS+2 l/ml baculovirus supernatant containing mIL-3). This recovery typically took 24–48 h. Cells were then washed twice and resuspended in selection medium (RPMI+10% FCS+10 l/ml Cos-7 supernatant containing mIL-5). Healthy populations of cells were usually obtained after 7–9 days of selection. These were then tested for IL-5 responsiveness using the same procedure used to test for B13 activity.
872 / Czabotar et al.
Data analysis Dose response curves were fitted to data from bioassays using Slide Write Plus for Windows, Version 4.0 (32-Bit Edition) (Advanced Graphics Software, Inc., Carlsbad, CA, USA). The equation y=a0+a1/(1+(x/a2)^a3) was used in this fitting. The ED50 value for the curve was taken as the value obtained for a2 from the fitting.
Protein modelling Protein modelling was performed using the automated comparative protein modelling server at the SWISS-MODEL web site (http://expasy.hcuge.ch/swissmod/SWISS-MODEL. html).23,24 PDB files derived from this modelling were viewed on RasWin Molecular Graphics, Windows Version 2.6.25
Acknowledgements We thank Ms Susanne Peroni for assistance with tissue culture. Thanks also goes to Dr Ward Lutz and Dr Richard Lipscome for useful discussion and support and to Dr Deirdre Coombe for critical review of this manuscript. C. J. Sanderson is supported by a fellowship from the National Health and Medical Research Council of Australia. P. E. Czabotar was in receipt of an Australian Post Graduate Award during the course of this work.
REFERENCES 1. Karlen S, De Boer ML, Lipscombe RJ, Lutz W, Mordvinov VA, Sanderson CJ (1998) Biological and molecular characteristics of interleukin-5 and its receptor. [Review] [159 refs]. Int Rev Immunol 16:227. 2. Van Oosterhout AJ, Ladenius AR, Savelkoul HF, Van Ark I, Delsman KC, Nijkamp FP (1993) Effect of anti-IL-5 and IL-5 on airway hyperreactivity and eosinophils in guinea pigs. Am Rev Resp Dis 147:548. 3. Minamitake Y, Kodama S, Katayama T, Adachi H, Tanaka S, Tsujimoto M (1990) Structure of recombinant human interleukin 5 produced by Chinese hamster ovary cells. J Biochem 107:292. 4. McKenzie AN, Ely B, Sanderson CJ (1991) Mutated interleukin-5 monomers are biologically inactive. Mol Immunol 28:155. 5. Milburn MV, Hassell AM, Lambert MH, Jordan SR, Proudfoot AE, Graber P, Wells TN (1993) A novel dimer configuration revealed by the crystal structure at 2.4 A resolution of human interleukin-5. Nature 363:172. 6. Takaki S, Tominaga A, Hitoshi Y, Mita S, Sonoda E, Yamaguchi N, Takatsu K (1990) Molecular cloning and expression of the murine interleukin-5 receptor. EMBO J 9:4367. 7. Tavernier J, Devos R, Cornelis S, Tuypens T, Van der Heyden J, Fiers W, Plaetinck G (1991) A human high affinity interleukin-5 receptor (IL5R) is composed of an IL5-specific alpha chain and a beta chain shared with the receptor for GM-CSF. Cell 66:1175. 8. Bazan JF (1990) Structural design and molecular evolution of a cytokine receptor superfamily. Proc Natl Acad Sci USA 87:6934. 9. Lyne PD, Bamborough P, Duncan D, Richards WG (1995) Molecular modeling of the GM-CSF and IL-3 receptor complexes. Prot Sci 4:2223.
CYTOKINE, Vol. 12, No. 7 (July, 2000: 867–873) 10. Rajotte D, Cadieux C, Haman A, Wilkes BC, Clark SC, Hercus T, Woodcock JA, Lopez A, Hoang T (1997) Crucial role of the residue R280 at the F -G loop of the human granulocyte/ macrophage colony-stimulating factor receptor alpha chain for ligand recognition. J Exp Med 185:1939. 11. Cornelis S, Plaetinck G, Devos R, Van der Heyden J, Tavernier J, Sanderson CJ, Guisez Y, Fiers W (1995) Detailed analysis of the IL-5-IL-5R alpha interaction: characterization of crucial residues on the ligand and the receptor. EMBO J 14:3395. 12. Tavernier J, Tuypens T, Verhee A, Plaetinck G, Devos R, Van der Heyden J, Guisez Y, Oefner C (1995) Identification of receptor-binding domains on human interleukin 5 and design of an interleukin 5-derived receptor antagonist. Proc Natl Acad Sci USA 92:5194. 13. Graber P, Proudfoot AE, Talabot F, Bernard A, McKinnon M, Banks M, Fattah D, Solari R, Peitsch MC, Wells TN (1995) Identification of key charged residues of human interleukin-5 in receptor binding and cellular activation. J Biol Chem 270:15762. 14. Morton T, Li J, Cook R, Chaiken I (1995) Mutagenesis in the C-terminal region of human interleukin 5 reveals a central patch for receptor alpha chain recognition. Proc Natl Acad Sci USA 92:10879. 15. Clutterbuck E, Shields JG, Gordon J, Smith SH, Boyd A, Callard RE, Campbell HD, Young IG, Sanderson CJ (1987) Recombinant human interleukin 5 is an eosinophil differentiation factor but has no activity in standard human B cell growth factor assays. Eur J Immunol 17:1743. 16. Plaetinck G, Van der Heyden J, Tavernier J, Fache I, Tuypens T, Fischkoff S, Fiers W, Devos R (1990) Characterization of interleukin 5 receptors on eosinophilic sublines from human promyelocytic leukemia (HL-60) cells. J Exp Med 172:683. 17. de Vos AM, Ultsch M, Kossiakoff AA (1992) Human growth hormone and extracellular domain of its receptor: crystal structure of the complex. Science 255:306. 18. Somers W, Ultsch M, De Vos AM, Kossiakoff AA (1994) The X-ray structure of a growth hormone-prolactin receptor complex [see comments]. Nature 372:478. 19. Livnah O, Stura EA, Johnson DL, Middleton SA, Mulcahy LS, Wrighton NC, Dower WJ, Jolliffe LK, Wilson IA (1996) Functional mimicry of a protein hormone by a peptide agonist: the EPO receptor complex at 2.8 A [see comments]. Science 273:464. 20. Kalai M, Montero-Julian FA, Grotzinger J, Wollmer A, Morelle D, Brochier J, Rose-John S, Heinrich PC, Brailly H, Content J (1996) Participation of two Ser-Ser-Phe-Tyr repeats in interleukin-6 (IL-6)-binding sites of the human IL-6 receptor. Eur J Biochem 238:714. 21. Yawata H, Yasukawa K, Natsuka S, Murakami M, Yamasaki K, Hibi M, Taga T, Kishimoto T (1993) Structurefunction analysis of human IL-6 receptor: dissociation of amino acid residues required for IL-6-binding and for IL-6 signal transduction through gp130. EMBO J 12:1705. 22. Layton JE, Iaria J, Smith DK, Treutlein HR (1997) Identification of a ligand-binding site on the granulocyte colonystimulating factor receptor by molecular modeling and mutagenesis. J Biol Chem 272:29735. 23. Peitsch MC (1996) ProMod and Swiss-Model: Internetbased tools for automated comparative protein modelling. Biochem Soc Trans 24:274. 24. Guex N, Peitsch MC (1997) SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714. 25. Sayle RA, Milner-White EJ (1995) RASMOL: biomolecular graphics for all. Trends Biochem Sci 20:374. 26. Middleton SA, Johnson DL, Jin RZ, McMahon FJ, Collins A, Tullai J, Gruninger RH, Jolliffe LK, Mulcahy LS (1996) Identification of a critical ligand binding determinant of the human erythropoietin receptor—evidence for common ligand binding motifs in the cytokine receptor family. J Biol Chem 271:14045. 27. Stephenst PE, Cockett MI (1989) The construction of a highly efficient and versatile set of mammalian expression vectors. Nucl Acids Res 17:7110.
Binding sites on the IL-5R / 873 28. Rolink AG, Melchers F, Palacios R (1989) Monoclonal antibodies reactive with the mouse interleukin 5 receptor. J Exp Med 169:1693. 29. Palacios R, Steinmetz M (1985) IL-3-dependent mouse clones that express B-220 surface antigen, contain Ig genes in germ-line configuration, and generate B lymphocytes in vivo. Cell 41:727.
30. Coombe DR, Nakhoul AM, Stevenson SM, Peroni SE, Sanderson CJ (1998) Expressed luciferase viability assay (ELVA) for the measurement of cell growth and viability. J Immunol Meth 215:145
IDENTIFICATION OF REGIONS WITHIN THE THIRD FnIII-LIKE DOMAIN OF THE IL-5R INVOLVED IN IL-5 INTERACTION Peter E. Czabotar,1 John Holland,2 Colin J. Sanderson1 Previously, two binding sites for interleukin 5 (IL-5) were identified on the IL-5 receptor alpha chain (IL-5R). They are located within the CD loop of the first fibronectin type III (FnIII)-like domain and the EF loop of the second FnIII-like domain. The first binding site was identified by exploiting the different abilities of human IL-5R (hIL-5R) and mouse IL-5R (mIL-5R) to bind hIL-5. Here we show that ovine IL-5 (oIL-5) has the ability to activate the hIL-5R but not the mIL-5R. By using chimeras of the mIL-5R and hIL-5R we demonstrate that residues within the first and third FnIII-like domains of mIL-5R are responsible for this lack of activity. Furthermore, mutation of residues on hIL-5R to mIL-5R within the predicted DE and FG loop regions of the third FnIII domain reduces oIL-5 activity. These results show that regions of the third FnIII domain of IL-5R are involved in binding, in addition to the regions in domains one and two of the IL-5R that were identified in an earlier study. 2000 Academic Press
Interleukin 5 (IL-5) is a cytokine produced by activated T lymphocytes. This cytokine has a pivotal role in the differentiation and maturation of eosinophils,1 the cells that play a key part in the pathogenesis of asthma and other allergic diseases. For these reasons IL-5 is considered to be an attractive drug target2 and understanding the interaction between IL-5 and its receptor is an important step towards the discovery and design of IL-5 antagonists. IL-5 is one of a group of structurally related cytokines and growth factors. These proteins are composed of four helices arranged in a configuration known as the cytokine helical bundle. IL-5 is unique within this group in that it exists as a homodimer stabilized by disulphide bridges between residues 42 and 84,3,4 and anti-parallel -sheet formation between residues 23–35 and 89–92.5 The IL-5 receptor is composed of a ligand-specific -chain (IL-5R) and a -chain (c) shared with the IL-3 and GMCSF receptor complexes.6,7 IL-5 displays From the 1Molecular Immunology Group, School of Biomedical Sciences, Curtin University of Technology, Perth, WA 6000, Australia; 2Urological Research Centre, QEII Medical Centre, Nedlands, WA 6009, Australia Correspondence to: Peter Czabotar, Division of Virology, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK; E-mail: [email protected] Received 17 November 1999; accepted for publication 20 December 1999 2000 Academic Press 1043–4666/00/070867+06 $35.00/0 KEY WORDS: cytokine receptor/interleukin 5/IL-5/IL5-R CYTOKINE, Vol. 12, No. 7 (July), 2000: pp 867–873
a binding affinity of approximately 500 pM for the IL-5R subunit. c does not provide any detectable affinity for IL-5 in the absence of the IL-5R, but does increase the affinity of the IL-5–IL-5R complex by 2–3-fold. IL-5R is a member of the class 1 cytokine receptor family.8 The extracellular region of these molecules consists of repeats of the fibronectin type III (FnIII)-like domain. IL-5R has three such repeats conventionally labelled from the membrane distal domain. Although the crystal structure of the IL-5R is yet to be determined its structure can be modelled on other members of the class 1 cytokine receptor family.9,10 The residues involved in receptor interactions have been well mapped in IL-5.11–14 However, the corresponding binding sites on IL-5R are not fully understood. This is mainly because the large size of the molecule places limitations on alanine scanning projects. Thus, making use of species differences is potentially more practical, as large numbers of residues are conserved and thus less mutations need to be generated. This technique has previously been used to identify residues within the CD loop of the first FnIIIlike domain that are important in ligand interactions.11 In addition, sequence homology to a growth hormone receptor contact point has been used to identify an IL-5 contact point in the EF loop of the second FnIII-like domain of the IL-5R. In this paper differences between ovine IL-5 (oIL-5) and mouse IL-5 867
868 / Czabotar et al.
Localization of the region in mIL-5R responsible for the lack of oIL-5 activity Chimeras of mIL-5R and hIL-5R (Fig. 2A) were transfected into BAF-B03 cells and selected with mIL-5. Murine IL-5 has the same affinity for all of these chimeras11 and so this selection regime will produce cell populations that express the IL-5R chimeras. Bulk cultures were used in all experiments to minimize differences due to gene copy or positional effects. None of the cells expressing chimeric receptors responded significantly to oIL-5 although they all responded to mIL-5 (Fig. 2B). Thus, several areas of the receptor are involved in the species difference, the previously described FnIII domain 1, and interestingly, the third FnIII domain. As the third FnIII domain has not previously been shown to be involved in ligand interactions it was examined in more detail.
Analysis of loop regions of FnIII domain 3 of the IL-5R Further analysis was concentrated on residues within the loop regions of the IL-5R predicted to be orientated towards IL-5. As multiple sites were responsible for the lack of oIL-5 activity on the mIL-5R it was possible that mutation of a single site within FnIII domain 3 may not result in any detectable oIL-5 activity. For this reason a number of mutations were performed on the hIL-5R to make the third FnIII
H-thymidine inc. (CPS)
Cell proliferation assays using the murine IL-5 responsive cells B13 and BAF (hIL-5R) were performed with Cos-7-expressed oIL-5 and mIL-5. Both of these cell types have the murine cytokine receptor chain, c, the B13 cells express mIL-5R and BAF-B03 cells were transfected with hIL-5R to make the BAF (hIL-5R) line. Thus, any differences in the activities of oIL-5 and mIL-5 are due to differences in the ability of these cytokines to interact with the particular IL-5R expressed by the cell lines. It was found that oIL-5 recognized hIL-5R and triggered proliferation of BAF (hIL-5R) cells to an extent equal to, or exceeding, that observed with mIL-5. In contrast, oIL-5 did not recognize mIL-5R and could not support the growth of B13 cells (Fig. 1). Thus there are regions of the mIL-5R that differ from those in the hIL-5R and these differences prohibit interaction with oIL-5.
A
40 000
20 000
3
Activity of oIL-5 in the mIL-5R and hIL-5R bioassay
0
60 000 H-thymidine inc. (CPS)
RESULTS
60 000
100
1000 Dilution
10 000
100
1000 Dilution
10 000
B
40 000
20 000
3
(mIL-5) activity have been used to identify loop regions within the third FnIII domain of IL-5R that interact with IL-5.
CYTOKINE, Vol. 12, No. 7 (July, 2000: 867–873)
0
Figure 1. Activity of mIL-5 (shaded bars) and oIL-5 (open bars) in the hIL-5R (A) and mIL-5R (B) bioassays. Values represent counts/s of incorporated 3H-thymidine as described in the methods section. These results represent the activity of varying dilutions of Cos-7 supernatant containing the respective cytokine. Error bars represent the standard deviation of duplicates. These results are representative of at least three separate experiments.
domain more like that of the mouse. The residues mutated are shown in Figure 3. The mutated hIL-5Rchains were transfected into BAF-B03 cells, the cells selected for expression using mIL-5. The ability of oIL-5 to trigger proliferation of cells expressing the mutant form of hIL-5R was compared to that obtained with cells expressing the wild-type hIL-5R. A reduction in the proliferative response would indicate that the mutated residues were involved in binding IL-5. As expected, all cells expressing the mutated hIL5R-chains displayed a similar response to mIL-5 and this resembled that obtained with cells expressing the wild-type hIL-5R (Fig. 4). In contrast, marked differences were seen when oIL-5 was used. Cells expressing mutant 1 and mutant 2 displayed a similar response to oIL-5 as cells expressing wild-type hIL-5R (Fig. 5), but cells expressing mutant 3 and mutant 4 exhibited a reduction in oIL-5 responsiveness. The combination of mutations 3 and 4 further reduced the ability of cells to respond to oIL-5. Thus, the residues on hIL-5R mutated in mutant 3 and mutant 4 are involved in the interaction of oIL-5 with hIL-5R.
Binding sites on the IL-5R / 869
A FnIII dom 1 FnIII dom 2
FnIII dom 3 TM
Chimera 1 ClaI
TM
Chimera 2 ClaI
TM
Chimera 3 ClaI
TM
Chimera 4 MluI B
30 000 20 000 10 000
hIL5Rα
Chimera 4
Chimera 3
Chimera 2
0
Chimera 1
3
H-thymidine inc. (CPS)
40 000
Figure 2. (A) Schematic representation of the extracellular domains of hIL-5R/mIL-5R chimeras. Grey areas represent regions of hIL-5R sequence and solid areas represent regions of mIL-5R sequence. The location of the three FnIII-like domains are indicated. The transmembrane domain (TM) and the introduced restriction sites used to construct these chimeras11 are also shown. (B) Activity of mIL-5 (shaded bars) and oIL-5 (open bars) on cells expressing chimeras of hIL-5R and mIL-5R. Values represent counts/s of incorporated 3H-thymidine. These values represent the activity of these cells in response to a 1/333 or a 1/100 dilution of Cos-7 supernatants containing either mIL-5 or oIL-5 respectively. These results are representative of three separate experiments.
DISCUSSION Murine IL-5 and hIL-5 share 70% sequence similarity at the amino acid level. Both bioassays and direct binding assays indicate that mIL-5 is able to interact with the hIL-5R with the same efficacy as hIL-5.15,16 However, hIL-5 interacts with the mIL-5R with only 1% of the binding and biological activity displayed by mIL-5.11,16 This difference in activity has previously been used to identify regions of the IL-5R involved in interacting with IL-5.11 We have shown here that like hIL-5, oIL-5 is less able to interact with the mIL-5R than the hIL-5R, indicating there are different amino acids in the hIL-5R and the mIL-5R at the IL-5:IL5R contact points, which restrict the interaction oIL-5 of with mIL-5R. To localize the points of contact between IL-5 and the IL-5R-chain we developed an oIL-5 bioassay using cells expressing chimeras of the hIL-5R and mIL-5R. This assay revealed that contact points,
between IL-5 and the IL-5R-chain, were located in the first and third FnIII-like domains. Further analysis was concentrated on the third domain. In particular, we analysed the loop regions of this domain which, on other members of the class 1 cytokine receptor family, are located on the ligand binding face of the molecule.17–19 Cells expressing hIL-5R mutated in either the DE (Mut 3) or FG (Mut 4) loops of FnIII domain 3 were less able to respond to oIL-5 than cells expressing wild-type hIL-5R. This activity was further reduced when hIL-5R was mutated at both of these loop regions. Clearly, the mutations made to hIL-5R in the DE and FG loops of FnIII domain 3 disrupts the binding sites for oIL-5. Furthermore, the additive effect of mutating both loop regions indicates that the amino acids altered in Mut3 and Mut4 disrupt two separate points at which IL-5 makes contact with the IL-5R-chain. The FG loop region of the membrane proximal FnIII-like domain, on other members of the class 1 cytokine receptor family, has been shown to be involved in ligand interactions. This has been demonstrated by crystallographic analysis of receptor:ligand complexes for the growth hormone receptor (GHR),17 prolactin receptor (PRLR)18 and erythropoietin receptor (Epo R).19 In addition, mutagenesis has been used to confirm the importance of this loop region on the IL-6R-chain,20,21 granulocyte colony stimulating factor receptor (GCSFR)22 and the granulocyte macrophage colony stimulating factor receptor -chain (GMR-).10 The importance of this loop in GMR- is of particular interest as this molecule is closely related to IL-5R. In addition, the residue identified as being involved in ligand contact in GMR- (Arg-280) has been conserved in IL-5R (Arg-297). In light of these findings it is reasonable to expect that residues within the FG loop of IL-5R are acting as contact points for IL-5. Unlike the FG loop, the DE loop of the membrane proximal FnIII-like domain has not been found to play a direct role in ligand interactions in other members of the class 1 cytokine receptor family. X-ray crystallography of the growth hormone GH:GHR complex and the GH:PRLR complex has demonstrated that this loop region is not directly involved in ligand interaction in these instances.17,18 Mutagenesis analysis of other class 1 receptor family members has concentrated only on those loop regions which have been shown to be important when GH contacts GHR. As a result the DE loop of the membrane proximal FnIII domain, and other loop regions not used by GHR, have been largely ignored in the structure/function analyses of these proteins. Although our observations implicate the DE loop region of the membrane proximal FnIII-like domain as
870 / Czabotar et al.
Figure 3.
CYTOKINE, Vol. 12, No. 7 (July, 2000: 867–873)
Amino acid sequence of the third FnIII-like domain of the hIL-5R.
Those residues differing on the mIL-5R are displayed above the sequence. The location of the predicted -pleated sheets are labelled A–G.8 Those residues mutated from hIL-5R to mIL-5R residues on hIL-5R mutants are highlighted and labelled. The location of the ClaI site used to make chimera 1 and the beginning of the transmembrane domain (TM) are indicated in italics. The conserved WSXWS motif is boxed.
16 000 H-thymidine inc. (CPS)
20 000 15 000 10 000 5000 0 10
12 000
8000
4000
3
3
H-thymidine inc. (CPS)
25 000
0
100
1000 10 000 Reciprocal dilution
100 000
2
10
100 1000 Reciprocal dilution
10 000
Figure 4. Titrations of mIL-5 on cells expressing hIL-5R (+), Mut 1 ( ), Mut 2 ( ), Mut 3 ( ), Mut 4 ( ) or Mut 3+4 ( ).
Figure 5. Titrations of oIL-5 on cells expressing hIL-5R (+), Mut 1 ( ), Mut 2 ( ), Mut 3 ( ), Mut 4 ( ) or Mut 3+4 ( ).
Response values represent counts/s of incorporated 3H-thymidine. Cytokine concentration is expressed as a dilution of Cos-7 cell supernatant containing mIL-5. Error bars represent the standard deviation of triplicates. The ED50 values for these curves are reciprocal dilutions of 1061 (hIL-5R), 1146 (Mut 1), 732.6 (Mut 2), 1108 (Mut 3), 1040 (Mut 4) and 1050 (Mut 3+4). These results are representative of at least two separate experiments.
Response values represent counts/s of incorporated 3H-thymidine. Cytokine concentration is expressed as a dilution of Cos-7 cell supernatant containing oIL-5. Error bars represent the standard deviation of triplicates. The ED50 values for these curves are reciprocal dilutions of 426.3 (hIL-5R), 647.8 (Mut 1), 492.2 (Mut 2), 172.4 (Mut 3), 60.26 (Mut 4) and 22.17 (Mut 3+4). These results are representative of at least two separate experiments.
playing a part in ligand contact it is possible that its role is structural. Mutation of this region may be disrupting the orientation of a ligand contact point in another region of IL-5R. Numerous observations support this possibility. Firstly, modelling IL-5R to the structure of EpoR reveals that the residues mutated in the DE loop, in the work reported here, may influence the orientation of the BC loop region (Fig. 6). In particular, intramolecular interactions may occur between Thr-271, which is predicted by the modelling to lie within the E -strand, and residues located within the B -strand. Secondly, the BC loop of the membrane proximal FnIII domain is utilized by numerous class 1 cytokine receptors for ligand interaction including GHR,17 PRLR,18 EpoR,19,26 IL-6R20,21 and GCSFR.22 Thirdly, the BC loop region of IL-5R is highly conserved between the human and mouse IL-5R, indicating that gross alterations to this region
have not been tolerated during evolution. This would be expected if it was an important region of the molecule. This high degree of conservation resulted in only one residue of the BC loop being mutated in this work (Mutation 2, Ile-247-Asp). Further analysis is required to determine whether a ligand contact point in the BC loop, DE loop, or some other region of the IL-5R is being disrupted by mutation 3. Such work could include the alanine scanning of these loops. In conclusion, this is the first demonstration that the third FnIII-like domain of the IL-5R is involved in interactions with IL-5. We have provided evidence that the DE and FG loop regions of this domain are involved in these interactions although further work is required to elucidate the actual role they play. This understanding is important for the discovery and design of IL-5 antagonists which are potentially useful for the treatment of asthma and other allergic diseases.
Binding sites on the IL-5R / 871
generously provided by Dr Jan Tavernier (Ghent, Belgium). Chimeras 1–4 used here are the equivalent of H1, H2, H4 and H5 used previously.11 The plasmid pEE6 containing cDNA that encodes hIL-5R was mutated using the Altered Sites Mutagenesis Kit (Clonetech). Mutations were confirmed by sequencing. The selection primer used in mutagenesis was designed to remove a unique EcoR1 site. Mutagenesis primers were designed to mutate the predicted loop regions of the third FnIII domain of the hIL-5R to a mIL-5R sequence. Selection primer: 5 -GCGAGCTCGAC TTCATTGATCCG-3 Mut 1: 5 -GCCATTGATCAAGTAAATCCTCC ACGGAATGTCACAGC-3 Mut 2: 5 -GTGTCTGCTTTTCCAGACCA TTGCTTTGATTATG-3 Mut 3: 5 -GCAGATAGAAAAATTGATCGCC AATAAATTCATCTCAATAATTGATG-3 Mut 4: 5 -GCAGTGAGCTCCCCGTGCAGAAT GCCAGGGCTCTGGAGTG-3 Figure 6. Predicted molecular structure of the third FnIII domain of the IL-5R. Structural co-ordinates were obtained by modelling the sequence from hIL-5R to the EpoR19 using the Swiss-Model database.23,24 This was then viewed with Rasmol.25 Viewing orientation is from the predicted point of view of the ligand. The molecule is displayed in structural format with -pleated sheets being detailed by yellow arrows. The loop regions discussed in the text are coloured red (B –C loop), green (D –E loop) and purple (F –G loop). Residues mutated in this work from hIL-5R to mIL-5R are labelled and displayed in spacefill format.
MATERIALS AND METHODS Mammalian expression vectors encoding IL-5 IL-5 DNA encoding oIL-5 (generously supplied by Dr Seow from CSIRO, Australia) and mIL-5 was cloned into the mammalian expression vector pEE6.27 These clones were sequenced and found to be identical to the published genebank sequence except for nucleotides 247 (A–G substitution), 386 (C–G substitution) and 387 (G–C substitution) of oIL-5. These differences result in a substitution from a lysine to glutamate at amino acid 66 and from threonine to serine at amino acid 112.
Production of IL-5 in Cos-7 cell cultures Fifteen micrograms of plasmid encoding IL-5 was electroporated into 2106 Cos-7 cells (260 V, 960 F). These cells were transferred to tissue culture flasks containing 5 ml of RPMI+5% FCS and incubated for 72 h at 37C and 5% CO2. The supernatants were then harvested and centrifuged at 2700g for 10 min to remove cell debris.
Plasmids expressing chimeric and mutant forms of the IL-5R The plasmid pSV-SPORT containing cDNA encoding chimeric forms of the human and murine IL-5R were
Assessment for hIL-5R and mIL-5R activity of oIL-5 and mIL-5 Cos-7 supernatants containing oIL-5 and mIL-5 were assessed for activity using mIL-5 and hIL-5 bioassays. The mIL-5R bioassay was performed using B13 cells, a mIL-5dependent cell line which contains the mIL-5R and mc.28 For the hIL-5R bioassay we used the BAF-B03 cell line transfected with the hIL-5R. BAF-B03 cells are a subclone of the mIL-3-dependent pro-B cell line BA/F329 and therefore already posses mc. When forced to express the hIL-5R they are able to proliferate in response to IL-5.30 These cells were washed three times in RPMI+10% FCS to remove residual IL-5 from the culture medium. BAF (5103) (hIL5R) or 1104 B13 cells were then added to the wells of 96-well plates containing 50 l of test substrate at varying dilutions and incubated at 37C and 5% CO2. After 48 h, 10 l of RPMI+10% FCS containing 0.33 Ci of 3 H-thymidine was added to each well and the cells incubated a further 4 h. The cells were then harvested onto glass fibre filters and level of 3H-thymidine incorporation determined on a Packard Matrix 9600 direct -counter.
Production of BAF-B03 cells expressing IL-5R BAF-B03 cells (106) were electroporated with 15 g of mammalian expression vector encoding either WT, chimeric or mutant hIL-5R (960 F, 260 V). Cells were allowed to recover from transfection in recovery medium (RPMI+10% FCS+2 l/ml baculovirus supernatant containing mIL-3). This recovery typically took 24–48 h. Cells were then washed twice and resuspended in selection medium (RPMI+10% FCS+10 l/ml Cos-7 supernatant containing mIL-5). Healthy populations of cells were usually obtained after 7–9 days of selection. These were then tested for IL-5 responsiveness using the same procedure used to test for B13 activity.
872 / Czabotar et al.
Data analysis Dose response curves were fitted to data from bioassays using Slide Write Plus for Windows, Version 4.0 (32-Bit Edition) (Advanced Graphics Software, Inc., Carlsbad, CA, USA). The equation y=a0+a1/(1+(x/a2)^a3) was used in this fitting. The ED50 value for the curve was taken as the value obtained for a2 from the fitting.
Protein modelling Protein modelling was performed using the automated comparative protein modelling server at the SWISS-MODEL web site (http://expasy.hcuge.ch/swissmod/SWISS-MODEL. html).23,24 PDB files derived from this modelling were viewed on RasWin Molecular Graphics, Windows Version 2.6.25
Acknowledgements We thank Ms Susanne Peroni for assistance with tissue culture. Thanks also goes to Dr Ward Lutz and Dr Richard Lipscome for useful discussion and support and to Dr Deirdre Coombe for critical review of this manuscript. C. J. Sanderson is supported by a fellowship from the National Health and Medical Research Council of Australia. P. E. Czabotar was in receipt of an Australian Post Graduate Award during the course of this work.
REFERENCES 1. Karlen S, De Boer ML, Lipscombe RJ, Lutz W, Mordvinov VA, Sanderson CJ (1998) Biological and molecular characteristics of interleukin-5 and its receptor. [Review] [159 refs]. Int Rev Immunol 16:227. 2. Van Oosterhout AJ, Ladenius AR, Savelkoul HF, Van Ark I, Delsman KC, Nijkamp FP (1993) Effect of anti-IL-5 and IL-5 on airway hyperreactivity and eosinophils in guinea pigs. Am Rev Resp Dis 147:548. 3. Minamitake Y, Kodama S, Katayama T, Adachi H, Tanaka S, Tsujimoto M (1990) Structure of recombinant human interleukin 5 produced by Chinese hamster ovary cells. J Biochem 107:292. 4. McKenzie AN, Ely B, Sanderson CJ (1991) Mutated interleukin-5 monomers are biologically inactive. Mol Immunol 28:155. 5. Milburn MV, Hassell AM, Lambert MH, Jordan SR, Proudfoot AE, Graber P, Wells TN (1993) A novel dimer configuration revealed by the crystal structure at 2.4 A resolution of human interleukin-5. Nature 363:172. 6. Takaki S, Tominaga A, Hitoshi Y, Mita S, Sonoda E, Yamaguchi N, Takatsu K (1990) Molecular cloning and expression of the murine interleukin-5 receptor. EMBO J 9:4367. 7. Tavernier J, Devos R, Cornelis S, Tuypens T, Van der Heyden J, Fiers W, Plaetinck G (1991) A human high affinity interleukin-5 receptor (IL5R) is composed of an IL5-specific alpha chain and a beta chain shared with the receptor for GM-CSF. Cell 66:1175. 8. Bazan JF (1990) Structural design and molecular evolution of a cytokine receptor superfamily. Proc Natl Acad Sci USA 87:6934. 9. Lyne PD, Bamborough P, Duncan D, Richards WG (1995) Molecular modeling of the GM-CSF and IL-3 receptor complexes. Prot Sci 4:2223.
CYTOKINE, Vol. 12, No. 7 (July, 2000: 867–873) 10. Rajotte D, Cadieux C, Haman A, Wilkes BC, Clark SC, Hercus T, Woodcock JA, Lopez A, Hoang T (1997) Crucial role of the residue R280 at the F -G loop of the human granulocyte/ macrophage colony-stimulating factor receptor alpha chain for ligand recognition. J Exp Med 185:1939. 11. Cornelis S, Plaetinck G, Devos R, Van der Heyden J, Tavernier J, Sanderson CJ, Guisez Y, Fiers W (1995) Detailed analysis of the IL-5-IL-5R alpha interaction: characterization of crucial residues on the ligand and the receptor. EMBO J 14:3395. 12. Tavernier J, Tuypens T, Verhee A, Plaetinck G, Devos R, Van der Heyden J, Guisez Y, Oefner C (1995) Identification of receptor-binding domains on human interleukin 5 and design of an interleukin 5-derived receptor antagonist. Proc Natl Acad Sci USA 92:5194. 13. Graber P, Proudfoot AE, Talabot F, Bernard A, McKinnon M, Banks M, Fattah D, Solari R, Peitsch MC, Wells TN (1995) Identification of key charged residues of human interleukin-5 in receptor binding and cellular activation. J Biol Chem 270:15762. 14. Morton T, Li J, Cook R, Chaiken I (1995) Mutagenesis in the C-terminal region of human interleukin 5 reveals a central patch for receptor alpha chain recognition. Proc Natl Acad Sci USA 92:10879. 15. Clutterbuck E, Shields JG, Gordon J, Smith SH, Boyd A, Callard RE, Campbell HD, Young IG, Sanderson CJ (1987) Recombinant human interleukin 5 is an eosinophil differentiation factor but has no activity in standard human B cell growth factor assays. Eur J Immunol 17:1743. 16. Plaetinck G, Van der Heyden J, Tavernier J, Fache I, Tuypens T, Fischkoff S, Fiers W, Devos R (1990) Characterization of interleukin 5 receptors on eosinophilic sublines from human promyelocytic leukemia (HL-60) cells. J Exp Med 172:683. 17. de Vos AM, Ultsch M, Kossiakoff AA (1992) Human growth hormone and extracellular domain of its receptor: crystal structure of the complex. Science 255:306. 18. Somers W, Ultsch M, De Vos AM, Kossiakoff AA (1994) The X-ray structure of a growth hormone-prolactin receptor complex [see comments]. Nature 372:478. 19. Livnah O, Stura EA, Johnson DL, Middleton SA, Mulcahy LS, Wrighton NC, Dower WJ, Jolliffe LK, Wilson IA (1996) Functional mimicry of a protein hormone by a peptide agonist: the EPO receptor complex at 2.8 A [see comments]. Science 273:464. 20. Kalai M, Montero-Julian FA, Grotzinger J, Wollmer A, Morelle D, Brochier J, Rose-John S, Heinrich PC, Brailly H, Content J (1996) Participation of two Ser-Ser-Phe-Tyr repeats in interleukin-6 (IL-6)-binding sites of the human IL-6 receptor. Eur J Biochem 238:714. 21. Yawata H, Yasukawa K, Natsuka S, Murakami M, Yamasaki K, Hibi M, Taga T, Kishimoto T (1993) Structurefunction analysis of human IL-6 receptor: dissociation of amino acid residues required for IL-6-binding and for IL-6 signal transduction through gp130. EMBO J 12:1705. 22. Layton JE, Iaria J, Smith DK, Treutlein HR (1997) Identification of a ligand-binding site on the granulocyte colonystimulating factor receptor by molecular modeling and mutagenesis. J Biol Chem 272:29735. 23. Peitsch MC (1996) ProMod and Swiss-Model: Internetbased tools for automated comparative protein modelling. Biochem Soc Trans 24:274. 24. Guex N, Peitsch MC (1997) SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714. 25. Sayle RA, Milner-White EJ (1995) RASMOL: biomolecular graphics for all. Trends Biochem Sci 20:374. 26. Middleton SA, Johnson DL, Jin RZ, McMahon FJ, Collins A, Tullai J, Gruninger RH, Jolliffe LK, Mulcahy LS (1996) Identification of a critical ligand binding determinant of the human erythropoietin receptor—evidence for common ligand binding motifs in the cytokine receptor family. J Biol Chem 271:14045. 27. Stephenst PE, Cockett MI (1989) The construction of a highly efficient and versatile set of mammalian expression vectors. Nucl Acids Res 17:7110.
Binding sites on the IL-5R / 873 28. Rolink AG, Melchers F, Palacios R (1989) Monoclonal antibodies reactive with the mouse interleukin 5 receptor. J Exp Med 169:1693. 29. Palacios R, Steinmetz M (1985) IL-3-dependent mouse clones that express B-220 surface antigen, contain Ig genes in germ-line configuration, and generate B lymphocytes in vivo. Cell 41:727.
30. Coombe DR, Nakhoul AM, Stevenson SM, Peroni SE, Sanderson CJ (1998) Expressed luciferase viability assay (ELVA) for the measurement of cell growth and viability. J Immunol Meth 215:145