Identification, Purification, and Characterization of a Soluble Interleukin (IL)-13-binding Protein
EVIDENCE THAT IT IS DISTINCT FROM THE CLONED IL-13 RECEPTOR AND IL-4 RECEPTOR alpha -CHAINS*

(Received for publication, November 4, 1996, and in revised form, December 19, 1996)

Jian-Guo Zhang , Douglas J. Hilton Dagger §, Tracy A. Willson , Clare McFarlane , Bronwyn A. Roberts , Robert L. Moritz , Richard J. Simpson , Warren S. Alexander , Donald Metcalf and Nicos A. Nicola

From the Walter and Eliza Hall Institute of Medical Research and the Cooperative Research Centre for Cellular Growth Factors and the  Joint Protein Structure Laboratory of the Walter and Eliza Hall Institute and the Ludwig Institute for Cancer Research, P.O. Royal Melbourne Hospital, Victoria 3050, Australia

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Interleukin-4 (IL-4) and interleukin-13 (IL-13) are structurally and functionally related cytokines which play an important role in the regulation of the immune response to infection. The functional similarity of IL-4 and IL-13 can be explained, at least in part, by the common components that form their cell surface receptors, namely the IL-4 receptor alpha -chain (IL-4Ralpha ) and the IL-13 receptor alpha -chain (IL-13Ralpha ). Soluble forms of the IL-4Ralpha have also been described and implicated in modulating the effect of IL-4. In this paper we describe the presence of a 45,000-50,000 Mr IL-13-binding protein (IL-13BP) in the serum and urine of mice. This protein binds IL-13 with a 100-300-fold higher affinity (KD = 20-90 pM) than does the cloned IL-13Ralpha (KD = 3-10 nM). In addition to this functional difference, the IL-13BP appears to be structurally and antigenically distinct from the IL-13Ralpha . Finally, unlike the cloned receptor, the IL-13BP acts as a potent inhibitor of IL-13 binding to its cell surface receptor, raising the possibility that it may be used to modulate the effects of IL-13 in vivo.


INTRODUCTION

Interleukin-4 (IL-4)1 and interleukin-13 (IL-13) are structurally and functionally related cytokines that share common receptor components (1-3). IL-4 and IL-13 exhibit approximately 30% amino acid sequence similarity and are encoded by genes with a similar intron/exon structure that are closely linked on human chromosome 5 and mouse chromosome 11. Both cytokines are produced by activated T-cells and act to regulate the immune response by, for example, inducing immunoglobulin class switching in B-lymphocytes to IgG1 and IgE isotypes and inhibiting the release of inflammatory mediators by macrophages (2). IL-4 but not IL-13 also regulates a variety of aspects of T-cell differentiation and function (4).

Three components of functional IL-4 and IL-13 receptors have been cloned. These are the IL-4 receptor alpha -chain (IL-4Ralpha ) (5), IL-13 receptor alpha -chain (IL-13Ralpha ) (6), and the IL-2 receptor gamma -chain (IL-2Rgamma ) (7). The relationships between the IL-4 and IL-13 receptors have been gleaned from many studies (5, 6, 8-16) and have recently been synthesized into a simple model by Leonard and colleagues (12). This model, which is consistent with the available data, suggests that there are two classes of IL-4 receptors but a single class of IL-13 receptors. In the case of IL-4, binding occurs initially to the IL-4Ralpha , and this complex then interacts with either IL-2Rgamma or IL-13Ralpha to yield a high affinity receptor capable of signal transduction. Reciprocally, IL-13 is thought to first bind to the IL-13Ralpha and then to recruit IL-4Ralpha to form a functional receptor. IL-2Rgamma does not appear to play a central role in IL-13 receptor function.

In addition to cell surface IL-4Ralpha , many studies have described the presence of a secreted form of the IL-4Ralpha (5, 17-21). mRNA for the cell surface and secreted forms of the IL-4Ralpha are transcribed from a single gene and arise by alternative splicing (5, 21). Soluble forms of many members of the hemopoietin receptor family have been described (22). These include the IL-2 receptor alpha -chain, IL-6 receptor alpha -chain (IL-6Ralpha ), IL-7 receptor alpha -chain, IL-9 receptor alpha -chain, GM-CSF receptor alpha -chain, and LIF receptor alpha -chain (LIFRalpha ). Soluble receptors have been implicated in both enhancing and reducing the biological effect of their cognate cytokine. For example, the complex of IL-6Ralpha and IL-6 is capable of interacting with cell surface gp130 to trigger a variety of biological responses (23), the complex of IL-4Ralpha and IL-4 inhibits the biological response of IL-4 in some settings and augments the response in other situations (24), while the LIFRalpha appears to act only as an inhibitor of LIF function (25, 26). In this study, we describe the presence of a high affinity binding protein for IL-13 in mouse serum and urine. This protein is shown to be functionally, structurally, and antigenically distinct from a recently cloned IL-13Ralpha and to act as a potent inhibitor of IL-13 binding to its cell surface receptor.


EXPERIMENTAL PROCEDURES

Reagents

Escherichia coli-expressed mouse IL-13 (R & D Systems) and mouse GM-CSF (a kind gift from Schering-Plough) were radioiodinated using the iodine monochloride method (27, 28), and E. coli-expressed mouse IL-4 (R & D Systems) was radiolabeled using the Bolton-Hunter reagent (29). Recombinant soluble mouse IL-4Ralpha was from Genzyme. FLAG peptide (DYKDDDDK) and anti-FLAG M2 affinity gel were from Scientific Imaging Systems. Soluble mouse IL-13Ralpha with an N-terminal FLAG epitope tag was expressed in Chinese hamster ovary cells using the mammalian expression vector pEF-BOS (30) and purified on an anti-FLAG M2 affinity column by elution with FLAG peptide. Anti-IL-13Ralpha polyclonal antiserum was prepared by injecting purified soluble IL-13Ralpha into rabbits, and the rabbits were bled after 3 months.

Gel Filtration Chromatography

Aliquots of samples were incubated with 125I-IL-13 in the presence or absence of unlabeled IL-13 for at least 30 min at 4 °C in a final volume of 200 µl. The mixture was applied to a Superdex 200 10/30 column (Pharmacia Biotech Inc.), equilibrated in 20 mM Tris-buffered saline (pH 7.0) containing 0.02% (v/v) Tween 20 and 0.02% (w/v) sodium azide (Tris-buffered saline). Samples were eluted with Tris-buffered saline at a flow rate of 0.5 ml/min, and 0.5-ml fractions were collected and counted in a Packard gamma -counter.

Cross-linking Assay

Aliquots of samples were incubated with 125I-IL-13 or 125I-IL-4 in the presence or absence of unlabeled competitor in a final volume of 20 µl for at least 30 min at 4 °C. Then 5 µl of a 12 mM bifunctional cross-linker BS3 (Pierce) solution was added and the mixtures were incubated for 30 min at 4 °C. Samples were mixed with 8 µl of 4 × concentrated SDS sample buffer and analyzed by 13% SDS-PAGE under nonreducing conditions. The gels were dried and visualized by either autoradiography or PhosphorImager.

125I-IL-13 Saturation Binding Assay

Binding of 125I-IL-13 to COS-7 cells expressing IL-13Ralpha was performed as described previously (6). Binding of 125I-IL-13 to soluble proteins in mouse serum and urine was determined by a gel filtration-based assay using Sephadex G-50 minicolumns (NICK columns, Pharmacia) to separate free from bound ligand. Scatchard analyses of saturation binding isotherms were performed using the nonlinear curve-fitting program LIGAND (31).

Deglycosylation and V8 Digestion

After SDS-PAGE, the gel, which contained 125I-IL-13 cross-linked to either IL-13BP from mouse urine or purified soluble IL-13Ralpha , was cut out according to the migration positions of prestained molecular weight markers. The gel pieces were added to 10 µl of Milli-Q purified water and 15 µl of 0.1 M sodium phosphate buffer (pH 7.0) containing 10 mM EDTA, 0.1% SDS, 0.01% (v/v) beta -mercaptoethanol and minced into small pieces with a pair of tweezers. Then 15 µl of 0.1 M sodium phosphate buffer (pH 7.0) containing 10 mM EDTA was added to the minced gel mixtures. After 1 h of incubation at 37 °C, the mixtures were heated for 1 min at 95 °C, cooled, and then 4 µl of 10% beta -octylglucoside was added. Treatment with N-glycosidase F (0.5 unit in 2.5 µl, Boehringer Mannheim) overnight at 37 °C completed the deglycosylation. The deglycosylated samples were then centrifuged and the gel pieces removed. The recovered liquid portions of the deglycosylated samples were further digested with protease V8 (Boehringer Mannheim) at concentrations ranging from 5 to 60 µg/ml for 3 h at 37 °C. The samples were analyzed by 15% SDS-PAGE.

Immunoprecipitation

Aliquots of 12-fold concentrated mouse urine or 3 µg/ml purified soluble IL-13Ralpha were incubated with 125I-IL-13 in the presence or absence of 0.5 µg/ml unlabeled IL-13 followed by cross-linking as described above. The cross-linking reactions were terminated by adding 1 M Tris-HCl buffer (pH 7.5) to a final concentration of 40 mM. The cross-linked samples were then mixed with 1:50 diluted control rabbit serum or anti-IL-13Ralpha antiserum which had been preincubated with or without FLAG peptide (100 µg/ml) to eliminate immunoreaction with potential anti-FLAG antibodies in the anti-FLAG-IL-13Ralpha polyclonal antiserum. After 30 min of incubation at 4 °C, the mixtures were added to 40 µl of 50% (v/v) protein G-Sepharose gel slurry (Pharmacia) and incubated for 30 min at 4 °C. The samples were centrifuged and the protein G-Sepharose beads washed with 3 × 0.5 ml of phosphate-buffered saline. For elution, the beads were mixed with 40 µl of 2 × concentrated SDS sample buffer and heated for 2 min at 95 °C. The supernatants were then analyzed by 13% SDS-PAGE under nonreducing conditions.

Production and Preparation of IL-13 Affinity Support

Mouse IL-13 was produced as a N-terminally FLAG-tagged fusion protein in Pichia pastoris and purified by reversed-phase high performance liquid chromatography. To prepare the IL-13 affinity support, the FLAG-tagged IL-13 was first bound to anti-FLAG antibody M2 affinity beads and then covalently linked to the M2 beads by the chemical cross-linker, BS3.

Purification and N-terminal Amino Acid Sequencing of IL-13BP

270 ml of 8-fold concentrated mouse urine was incubated with 1.0 ml of IL-13 affinity beads for 5 h at room temperature. After removing unbound proteins by centrifugation, the IL-13 affinity beads were washed extensively with 20 mM phosphate-buffered saline (pH 7.0) containing 0.02% (v/v) Tween 20 and 0.02% (w/v) sodium azide (phosphate-buffered saline) followed by additional washes with 5 × 2 ml of 3-fold diluted Actisep elution medium (Sterogenes Bioseparations). The bound protein was then eluted with 6 × 2.5 ml of neat Actisep elution medium. The affinity column eluates were buffer-exchanged into phosphate-buffered saline using PD-10 columns (Pharmacia). Aliquots of fractions were analyzed by SDS-PAGE, after which the gel was silver-stained (32). Fractions were also analyzed for their ability to bind to 125I-IL-13 using the cross-linking protocol described above. The affinity-purified IL-13BP in a total volume of 12 ml (approximately 2-4 µg) was loaded onto the sample cartridge of a Hewlett Packard model G1005A protein sequencer which was operated with the routine 3 sequencer program (33).


RESULTS

Given the expression of transmembrane and secreted forms of many members of the hemopoietin receptor family from alternatively spliced transcripts or via proteolysis (22), we sought evidence for the existence of a soluble form of the IL-13Ralpha . Using a gel filtration-based assay, mouse serum and urine were examined for the presence of an IL-13BP. Gel filtration chromatography of 125I-IL-13 alone resulted in elution of radioactivity in fractions 35-39 (Fig. 1,A and B). Prior addition of mouse serum or urine to the 125I-IL-13 resulted in the presence of an additional peak of 125I-IL-13 eluting earlier in fractions 27-30 (Fig. 1, C and D). The formation of this higher molecular weight peak was competed for by the addition of an excess of unlabeled IL-13 but not by unlabeled IL-4, demonstrating that the interaction with the binding protein was specific (Fig. 1, E and F).


Fig. 1. Analysis of IL-13 binding to IL-13BP by gel filtration chromatography. The gel filtration column was run as described under "Experimental Procedures." A, 125I-IL-13 alone (20,000 cpm); C, E, and G, 125I-IL-13 (20,000 cpm) + 100 µl of mouse serum or in the presence of either 1 µg/ml unlabeled IL-13 or IL-4, respectively; B, 125I-IL-13 alone (50,000 cpm); D, F, and H, 125I-IL-13 (50,000 cpm) + 100 µl mouse urine or in the presence of either 1 µg/ml unlabeled IL-13 or IL-4, respectively.
[View Larger Version of this Image (27K GIF file)]


In contrast to crude mouse urine, purified soluble IL-13Ralpha appeared unable to bind IL-13 as assessed by gel filtration chromatography (data not shown). Given the very low affinity of IL-13 for the IL-13Ralpha and the fast kinetic dissociation rate (6), this result was not surprising but did suggest that the serum and urine binding proteins were either distinct from the IL-13Ralpha or contained other components in addition to IL-13Ralpha . An obvious candidate for a protein capable of interaction with IL-13Ralpha to generate a high affinity IL-13 receptor was the IL-4Ralpha . The presence of soluble IL-4Ralpha in urine and serum has been described in a number of previous studies (17, 18, 20), and this result was confirmed by the cross-linking experiments described below. Despite the presence of an IL-4 binding protein (IL-4BP) in urine and serum, addition of purified IL-4Ralpha to purified IL-13Ralpha did not recapitulate the properties of the IL-13BP found in serum and urine as assessed by gel filtration chromatography (data not shown). Moreover, the apparent molecular weight of the IL-13/IL-13BP complex from serum and urine (60,000-65,000) was not consistent with a complex of IL-13 with IL-13Ralpha and IL-4Ralpha (predicted molecular weight of approx 100,000).

To assess the size of the IL-13BP and its relationship to the IL-4BP, mouse urine was fractionated on a gel filtration column (Fig. 2A). Aliquots from each fraction were then mixed with 125I-IL-13 or 125I-IL-4 in the presence or absence of an excess of unlabeled IL-4 or IL-13 and subjected to cross-linking using the bifunctional reagent BS3. The products of the cross-linking reaction were then resolved by SDS-PAGE and visualized by autoradiography. Cross-linking of 125I-IL-13 to unfractionated mouse urine revealed the presence of a major band that migrated with a Mr of approximately 60,000-65,000 (Fig. 2B). Given that 125I-IL-13 migrates with a Mr of approx 15,000, this would suggest that the IL-13BP has a Mr of 45,000-50,000. Cross-linking to the IL-13BP was specific, since it was abolished by competition with unlabeled IL-13 (Fig. 3B) but not IL-4 (Fig. 4). The 45,000-50,000 Mr IL-13BP eluted from the gel filtration column in fractions 27 to 29, consistent with its molecular weight estimated from the cross-linking experiment. Additional apparently nonspecific lower molecular weight radioactive products were also observed in the cross-linking studies with 125I-IL-13. One of these eluted from the gel filtration column in fractions 29 and 30, after the 45,000-50,000 Mr IL-13BP, but in a similar position to an IL-4BP (Fig. 2C). The complexities of this interaction will be discussed in more detail below. Cross-linking of 125I-IL-4 to the IL-4BP resulted in a species migrating with a Mr of approximately 50,000 (Fig. 2C), suggesting that the binding protein itself had a Mr of approx 35,000.


Fig. 2. Fractionation of crude mouse urine by gel filtration chromatography. A, 20-fold concentrated mouse urine (0.5 ml) was applied to a Superdex 200 10/30 column previously equilibrated in normal saline containing 0.02% (v/v) Tween 20. Samples were eluted at a flow rate of 0.5 ml/min, and 0.5-ml fractions were collected. B and C, SDS-PAGE analyses of cross-linking of 125I-IL-13 (130,000 cpm) and 125I-IL-4 (200,000 cpm), respectively, to fractions 26-31 from A (10 µl of each fraction was used), respectively.
[View Larger Version of this Image (53K GIF file)]



Fig. 3. SDS-PAGE analyses of 125I-IL-13 cross-linking to IL-13BP, soluble IL-13Ralpha , or IL-4Ralpha . Lanes 1-3, 125I-IL-13 (100,000 cpm) + 2 µg/ml soluble IL-13Ralpha or in the presence of 2 µg/ml unlabeled IL-13 and IL-4, respectively; lanes 4-6, 125I-IL-13 (100,000 cpm) + 2 µg/ml soluble IL-4Ralpha or in the presence of 2 µg/ml unlabeled IL-4 and IL-13, respectively; lanes 7-9, 125I-IL-13 (100,000 cpm) + 2 µg/ml soluble IL-13Ralpha and 1 µg/ml soluble IL-4Ralpha or in the presence of 2 µg/ml unlabeled IL-13 and IL-4, respectively; lanes 10-12, 125I-IL-13 (100,000 cpm) + 20-fold concentrated mouse urine (10 µl) or in the presence of 2 µg/ml unlabeled IL-13 and IL-4, respectively; lanes 13-15, 125I-IL-13 (100,000 cpm) + fraction 27 (5 µl) from Fig. 2A and 1 µg/ml soluble IL-4Ralpha or in the presence of 2 µg/ml unlabeled IL-13 and IL-4, respectively.
[View Larger Version of this Image (37K GIF file)]



Fig. 4. SDS-PAGE analyses of competitive cross-linking of 125I-IL-13 binding to soluble IL-13Ralpha or IL-13BP. A and B, 125I-IL-13 (130,000 cpm) cross-linked to soluble IL-13Ralpha (5 µl of 0.75 µg/ml) or IL-13BP (5 µl of 12-fold concentrated mouse urine), respectively, in the presence of various concentrations of unlabeled IL-13 as indicated. C, plot of densitometric analysis data from A (black-square) and B (square ) expressed as percentage of binding in the absence of unlabeled IL-13.
[View Larger Version of this Image (38K GIF file)]


The ligand-binding specificity of the 45,000-50,000 Mr IL-13BP and purified receptor components was examined further using cross-linking. Although binding of 125I-IL-13 to purified IL-13Ralpha was not detected by gel filtration chromatography, an interaction was observed using cross-linking (Fig. 3, lane 1). Consistent with previous studies (6), this interaction was competed for by unlabeled IL-13 but not unlabeled IL-4 (Fig. 3, lanes 2 and 3). When cross-linking studies were performed using 125I-IL-13 and purified IL-4Ralpha , a product with a Mr of approx 50,000 was observed (Fig. 3, lane 4). Surprisingly, cross-linking of 125I-IL-13 to IL-4Ralpha was in competition with unlabeled IL-4 (Fig. 4, lane 5) but not unlabeled IL-13 (Fig. 3, lane 6). Combining purified IL-13Ralpha and IL-4Ralpha with 125I-IL-13 resulted in the generation of species similar in molecular weight and specificity to reactions containing each receptor alone (Fig. 3, lanes 7-9). No higher molecular weight complexes were observed, suggesting that formation of ternary IL-13/IL-13Ralpha /IL-4Ralpha complexes or their capture with the cross-linker occurred inefficiently in solution, even at high concentrations of each component. Consistent with our initial experiments (Fig. 2B), crude mouse urine contained a major 45,000-50,000 Mr IL-13BP (Fig. 3, lane 10). Cross-linking of 125I-IL-13 to this species was in competition with unlabeled IL-13 (Fig. 3, lane 11) but not unlabeled IL-4 (Fig. 3, lane 12). The lower molecular weight species observed in our initial experiments (Fig. 2B, fractions 29 and 30), was again observed. As before, cross-linking of 125I-IL-13 to this protein was not in competition with unlabeled IL-13 (Fig. 3, lanes 10 and 11); however, like cross-linking of 125I-IL-13 to purified IL-4Ralpha , cross-linking to this protein was competed for by unlabeled IL-4 (Fig. 3, lane 12). Although the addition of purified IL-4Ralpha to purified IL-13Ralpha did not alter the pattern of cross-linking observed to either component alone, we sought to determine whether purified IL-4Ralpha could interact with partially purified 45,000-50,000 Mr IL-13BP from mouse urine (Fig. 2A, fraction 27). As with the purified receptor components, no additional effect of adding IL-4Ralpha was observed (Fig. 3, lanes 13-15).

While interaction between IL-13 and both purified soluble IL-13Ralpha and the 45,000-50,000 Mr binding protein was demonstrable by cross-linking, only the interaction with the binding protein was detectable using gel filtration chromatography. These results suggested that the affinity of IL-13 for the binding protein was higher than for soluble IL-13Ralpha . To test this formally, saturation binding experiments were performed. Scatchard transformations revealed that, consistent with previous studies (6), the equilibrium dissociation constant (KD) of IL-13 for the IL-13Ralpha expressed by Chinese hamster ovary cells was approximately 10 nM (data not shown). However, the affinity of IL-13 for the serum and urinary binding protein was 100-300-fold higher, with KD values ranging from 20 to 90 pM (data not shown). The difference in affinity was confirmed in cross-linking experiments in which 125I-IL-13 was mixed with increasing concentrations of unlabeled IL-13 prior to cross-linking to soluble IL-13Ralpha (Fig. 4A) or urinary binding protein (Fig. 4B). Densitometric analysis of these data demonstrated that half-maximal inhibition of binding to soluble IL-13Ralpha occurred with approximately 100 ng/ml IL-13, a 40-fold higher concentration than required to inhibit 50% of the cross-linking to the 45,000-50,000 Mr urinary binding protein.

The structural relationship between soluble IL-13Ralpha and the 45,000-50,000 Mr urinary IL-13BP was examined by cross-linking 125I-IL-13 to both proteins and isolating the resultant complexes from SDS-polyacrylamide gels. Each complex was then subjected to exhaustive deglycosylation using N-glycanase F and digestion with various concentrations of Staphylococcus V8 protease. The products of these treatments were then resolved by further SDS-PAGE. The untreated and deglycosylated complexes of IL-13 and the soluble IL-13Ralpha appeared slightly larger than the corresponding complexes with the urinary binding protein (compare Fig. 5, lane 1 with lane 10 and lane 2 with lane 9). In addition, the products of V8 proteolysis of the two complexes were clearly different, emphasizing the structural difference between the IL-13Ralpha and the 45,000-50,000 Mr urinary IL-13BP (compare Fig. 5, lanes 3-5 with lanes 6-8).


Fig. 5. Deglycosylation and V8 digestion of 125I-IL-13 cross-linked to IL-13BP or soluble IL-13Ralpha . Lanes 1 and 2 or lanes 10 and 9, mouse urinary IL-13BP or soluble IL-13Ralpha cross-linked to 125I-IL-13 before and after N-glycosidase F treatment, respectively; lanes 3-5 or lanes 8-6, deglycosylated IL-13BP/125I-IL-13 complex (from lane 2) or deglycosylated IL-13Ralpha /125I-IL-13 complex (from lane 9) after further digestion with protease V8 at concentrations of 5, 25, and 60 µg/ml, respectively.
[View Larger Version of this Image (87K GIF file)]


A rabbit polyclonal antiserum was raised against purified soluble recombinant IL-13Ralpha . This antiserum was capable of immunoprecipitating the cross-linked product of 125I-IL-13 with IL-13Ralpha (Fig. 6, lane 10), while no immunoprecipitation was observed with a preimmune rabbit serum (Fig. 6, lane 12). Immunoprecipitation of the radioactive complex was not inhibited by the FLAG peptide which is present in our soluble recombinant IL-13Ralpha (Fig. 6, lane 11) but was inhibited by an excess of IL-13 (Fig. 6, lanes 13 and 14). In contrast to the IL-13Ralpha complex, the complex between the 45,000-50,000 Mr urinary IL-13BP and 125I-IL-13 was not recognized by the rabbit antiserum to IL-13Ralpha (Fig. 6, lanes 3-9), suggesting that these proteins are antigenically as well as structurally and functionally distinct.


Fig. 6. Immunoprecipitation by anti-IL-13Ralpha antiserum of 125I-IL-13 cross-linked to soluble IL-13Ralpha but not of 125I-IL-13 cross-linked to IL-13BP. Lane 1, IL-13BP (5 µl of 12-fold concentrated mouse urine) cross-linked to 125I-IL-13 (125,000 cpm); lane 2, soluble IL-13Ralpha (5 µl of 3 µg/ml) cross-linked to 125I-IL-13 (125,000 cpm); lanes 3-5, IL-13BP (30 µl of 12-fold concentrated mouse urine) cross-linked to 125I-IL-13 (750,000 cpm) and immunoprecipitated with a control rabbit serum or with an anti-IL-13Ralpha polyclonal antiserum in the presence or absence of 100 µg/ml FLAG peptide, respectively; lanes 6-8, IL-13BP (30 µl of 12-fold concentrated mouse urine) cross-linked to 125I-IL-13 (750,000 cpm) in the presence of 0.5 µg/ml unlabeled IL-13 and immunoprecipitated with an anti-IL-13Ralpha polyclonal antiserum in the presence or absence of 100 µg/ml FLAG peptide, respectively; lanes 9-11, soluble IL-13Ralpha (30 µl of 3 µg/ml) cross-linked to 125I-IL-13 (750,000 cpm) and immunoprecipitated with a control rabbit serum or with an anti-IL-13Ralpha polyclonal antiserum in the presence or absence of 100 µg/ml FLAG peptide, respectively; lanes 12-14, soluble IL-13Ralpha (30 µl of 3 µg/ml) cross-linked to 125I-IL-13 (750,000 cpm) in the presence of 0.5 µg/ml unlabeled IL-13 and immunoprecipitated with an anti-IL-13Ralpha polyclonal antiserum in the presence or absence of 100 µg/ml FLAG peptide, respectively.
[View Larger Version of this Image (54K GIF file)]


Soluble receptors for a variety of cytokines have been described (22). In some cases these act to augment a biological response (23, 24), while in other situations they may inhibit the biological effect (24-26). To gain information on whether the purified soluble IL-13Ralpha or the urinary binding protein could influence IL-13 action we examined their effects on binding of 125I-IL-13 to peritoneal macrophages. Fig. 7 demonstrates that even though both the urinary IL-13BP and the soluble IL-13Ralpha could inhibit the binding of 125I-IL-13 to cell surface receptors expressed by macrophages the latter was far less efficient, raising the possibility that the soluble IL-13BP may be used to modulate the effects of IL-13 in vivo.


Fig. 7. Comparison of the ability of IL-13BP and soluble IL-13Ralpha to inhibit the binding of IL-13 to its cell surface receptor. 105 cpm of 125I-IL-13 or 125I-GM-CSF were incubated for 40 min at 4 °C with the indicated dilution of partially purified urinary IL-13BP (A) or with the indicated concentration of soluble IL-13Ralpha (B). The labeled ligands and the soluble IL-13Ralpha or IL-13BP were then added to 50 µl of medium containing 1.5 × 106 peritoneal cells from a GM-CSF transgenic mouse. Incubation was continued for a further 2 h at 4 °C before cell-associated and free 125I-IL-13 or 125I-GM-CSF were separated by centrifugation of cells through 200 µl of fetal calf serum. The resulting cell pellets and supernatants were then counted in a gamma -counter, and specific binding was calculated as the difference between the 125I-ligand bound in the absence and presence of the unlabeled competitors. In turn this was expressed as a percentage of that observed in the absence of either soluble IL-13Ralpha or IL-13BP (bullet ). As a control, both soluble IL-13Ralpha and IL-13BP showed no inhibition of 125I-GM-CSF binding (open circle ) to peritoneal cells.
[View Larger Version of this Image (12K GIF file)]


The soluble IL-13BP was purified from 2.2 liters of mouse urine by affinity chromatography on immobilized IL-13 as described under "Experimental Procedures." Approximately 2-4 µg of the purified protein was obtained and electrophoresed as a single silver staining protein of Mr 45,000-50,000 on SDS-polyacrylamide gel (Fig. 8A). Chemical cross-linking of this protein incubated with 125I-IL-13 revealed a complex of Mr 60,000-65,000 (Fig. 8B) similar to that seen with crude urine preparations (Fig. 2). Moreover, chemical cross-linking of this purified preparation incubated with unlabeled IL-13 and visualization of the products by silver staining of the SDS-polyacrylamide gel showed that a significant proportion of the major protein band of Mr 45,000-50,000 could interact with IL-13 and shift to a band of Mr 60,000-65,000 (Fig. 8C, lanes 3 and 4).


Fig. 8. Properties of purified IL-13BP. A, SDS-PAGE of purified IL-13BP visualized by silver staining. B, purified IL-13BP cross-linked to 125I-IL-13 with BS3 and visualized by autoradiography. C, lane 1, IL-13 incubated with BS3; lane 2, partially purified side-fraction of IL-13BP; lane 3, purified IL-13BP incubated with BS3; and lane 4, purified IL-13BP cross-linked to IL-13 with BS3 (lanes 1-4 visualized by silver staining). Open and solid arrows indicate the positions of IL-13BP and IL-13/IL-13BP complex, respectively.
[View Larger Version of this Image (37K GIF file)]


The purified protein was subjected to N-terminal sequencing and generated a single major 27-amino-acid sequence (with an initial yield of approx 50 pmol), namely EIKVNPPQDFEILDPGLLGYLYLQWKP~. This sequence is clearly different from the N-terminal amino acid sequence of the cloned mouse IL-13Ralpha (6). An initial search of the GenBankTM data base identified a human expressed sequence tag (accession number R52795[GenBank]/clone 41648) derived from an infant brain cDNA library with an identical sequence stretch except for conservative amino acid changes at positions 13 (Lright-arrowV), 17 (Lright-arrowY), and 26 (Kright-arrowQ). Moreover, this sequence was preceded by a typical leader sequence of 27 amino acids and signal sequence cleavage site in the expressed sequence tag suggesting that its predicted mature amino acid sequence would start at the same position as that identified for the IL-13BP. The 3' end of clone 41648 (the sequence of which is represented by expressed sequence tag R52796[GenBank]) was also deposited in the GenBankTM data base and revealed sequences suggesting that this protein was a member of the type I cytokine receptor family including the diagnostic sequence WSEWS.


DISCUSSION

At least one form of the cell surface receptor complex for IL-13 is composed of the IL-4Ralpha (5) and the recently cloned IL-13Ralpha (6). Despite its low affinity, the specificity of the interaction between IL-13 and the soluble IL-13Ralpha was demonstrable by cross-linking, since IL-13 but not IL-4 was capable of competing for binding. In the course of the cross-linking studies, an unexpected, puzzling, and to date unexplained observation was made. 125I-IL-13 was found to cross-link to purified recombinant IL-4Ralpha . This interaction was capable of being competed for by IL-4 but not by unlabeled IL-13. The possibility that iodination of IL-13 on tyrosine residues using the iodine monochloride technique altered its receptor specificity was thought to be unlikely, since a similar result was observed with IL-13 labeled on lysine residues using the Bolton-Hunter reagent.

Murine serum and urine were found to contain a protein capable of binding IL-13. However, such a protein was not detected in human plasma and urine (data not shown). The mouse urinary IL-13BP was characterized in detail and shown to be structurally, antigenically, and functionally distinct from the cloned mouse IL-13Ralpha (6). These differences could not be explained in terms of an association of the IL-13Ralpha and IL-4Ralpha in these biological fluids but rather suggested the existence of an independent protein capable of binding IL-13. Several lines of evidence support the notion that the mouse urinary IL-13BP was different from the cloned mouse IL-13Ralpha (6): (a) the size of the binding protein was smaller than the extracellular domain of the cloned IL-13Ralpha , and the pattern of V8 protease digestion of the two molecules was clearly different; (b) an antiserum raised against the extracellular domain of the cloned IL-13Ralpha failed to recognize the binding protein found in urine; (c) most dramatically, the binding protein interacted with IL-13 with a 100-300-fold higher affinity than the soluble IL-13Ralpha (KD = 20-90 pM versus 3-10 nM) (6); and (d) N-terminal amino acid sequences of the two proteins were different. Moreover, unlike the soluble IL-13Ralpha , the urinary binding protein was an efficient inhibitor of IL-13 binding to its cell surface receptor.

These characteristics suggest that the soluble IL-13BP should be an effective inhibitor of IL-13 biological activities and it may serve to dampen the proinflammatory activities of IL-13 such as macrophage and eosinophil activation and, in humans, IgE production by B-lymphocytes (2, 3, 34).

The recent cloning of an alternate IL-13 cell surface receptor from human renal carcinoma cells by Caput et al. (35) has indicated that the IL-13BP is almost certainly a murine soluble analogue of this human receptor chain. This conclusion is based on the near identity of the first 27 amino acids of the predicted mature form of the human receptor with that of the soluble IL-13BP, the high affinity with which it binds IL-13 (KD = 250 pM), and its absolute specificity for binding IL-13 but not IL-4.

Caput et al. (35) presented evidence that their IL-13 receptor shows little alteration of binding affinity or specificity when it is co-expressed with IL-4Ralpha and, surprisingly, a diminution of IL-13-binding when it is co-expressed with gamma c. This contrasts with our data with IL-13Ralpha (6), where higher affinity IL-13/IL-4-cross-reactive receptors were generated by co-expression with IL-4 receptor.

The evidence of two alternative IL-13-binding receptor chains and the possible involvement of IL-4Ralpha and gamma c in forming IL-13 receptor complexes points to a very complex pattern of IL-13 and IL-4 recognition and signal transduction at different cell surfaces. The existence of a soluble IL-13 receptor adds a further layer of complexity in the control of IL-13 responses. Issues yet to be addressed include whether the IL-13 receptor described by Caput et al. (35) is capable of functional signaling and how the soluble form of this receptor is generated and controlled (e.g. by alternative splicing or proteolytic release from the cell surface). Clearly, very careful reconstitution experiments with all of the receptor components described above will be required to begin to understand this complex regulatory system.


FOOTNOTES

*   This work was supported by the Anti-Cancer Council of Victoria, Australia; AMRAD Operations Pty. Ltd., Melbourne, Australia; The National Health and Medical Research Council, Canberra, Australia; The J. D. and L. Harris Trust; National Institutes of Health Grant CA-22556 and the Australian Federal Government Cooperative Research Centres Program.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    To whom correspondence should be addressed: The Walter and Eliza Hall Institute of Medical Research, P.O. Royal Melbourne Hospital, Victoria 3050, Australia. Tel.: 61-3-9345-2621; Fax: 61-3-9347-0852; E-mail: hilton{at}wehi.edu.au.
§   Supported by a Queen Elizabeth II Postdoctoral Fellowship from the Australian Research Council.
1   The abbreviations used are: IL, interleukin; IL-13BP, interleukin-13-binding protein; IL-4BP, interleukin-4-binding protein; GM-CSF, granulocyte-macrophage colony-stimulating factor;PAGE, polyacrylamide gel electrophoresis.

ACKNOWLEDGEMENTS

We thank Dale Cary and Naomi Sprigg for excellent technical assistance. We are grateful to Wendy Carter and Karen Mackwell for preparing anti-IL-13Ralpha antiserum and to animal technicians for collecting mouse urine.


REFERENCES

  1. Howard, M., and Harada, N. (1994) in Guidebook to Cytokines and Their Receptors (Nicola, N. A., ed), pp. 44-46, Oxford University Press
  2. McKenzie, A. N. J., and Zurawski, G. (1994) in Guidebook to Cytokines and Their Receptors (Nicola, N. A., ed), pp. 92-95, Oxford University Press
  3. Zurawski, G., and de Vries, J. E. (1994) Immunol. Today 15, 19-26 [CrossRef][Medline] [Order article via Infotrieve]
  4. de Vries, J. E., and Zurawski, G. (1995) Int. Arch. Allergy Appl. Immunol. 106, 175-179
  5. Mosley, B., Beckmann, M. P., March, C. J., Idzerda, R. L., Gimpel, S. D., VandenBos, T., Friend, D., Alpert, A., Anderson, D., Jackson, J., Wignall, J. M., Smith, C., Gallis, B., Sims, J. E., Urdal, D., Cosman, D., and Park, L. S. (1989) Cell 89, 335-348
  6. Hilton, D. J., Zhang, J-G, Metcalf, D., Alexander, W. S., Nicola, N. A., and Willson, T. A. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 497-501 [Abstract/Free Full Text]
  7. Takeita, T., Asano, H., Ohtani, K., Ishii, N., Kumaki, S., Tanaka, N., Munakata, H., Nakamura, M., and Sugamura, K. (1992) Science 257, 379-382 [Medline] [Order article via Infotrieve]
  8. Zurawski, S. M., Vega, F., Huyghe, B., and Zurawski, G. (1993) EMBO J. 12, 2663-2670 [Abstract]
  9. Zurawski, S. M., Chormarat, P., Djossou, O., Bidaud, C., McKenzie, A. N. J., Miossec, P., Banchereau, J., and Zurawski, G. (1995) J. Biol. Chem. 270, 13869-13878 [Abstract/Free Full Text]
  10. He, Y-W., and Malek, T. R. (1995) J. Immunol. 155, 9-12 [Abstract]
  11. He, Y-W., Adkins, B., Furse, R. K., and Malek, T. R. (1995) J. Immunol. 154, 1596-1605 [Abstract/Free Full Text]
  12. Lin, J-X., Migone, T-S., Tsang, M., Friedmann, M., Weatherbee, J. A., Zhou, L., Yamaauchi, A., Bloom, E. T., Mietz, J., John, S., and Leonard, W. J. (1995) Immunity 2, 331-339 [Medline] [Order article via Infotrieve]
  13. Matthews, D. J., Clark, P. A., Herbert, J., Morgan, G., Armitage, R. J., Kinnon, C., Minty, A., Grabstein, K. H., Caput, D, and Callard, R. (1995) Blood 85, 38-42 [Abstract/Free Full Text]
  14. Obiri, N. I., Debinski, W., Leonard, W. J., and Puri, R. K. (1995) J. Biol. Chem. 270, 8797-8804 [Abstract/Free Full Text]
  15. Smerz-Bertling, C., and Duschl, A. (1995) J. Biol. Chem. 270, 966-970 [Abstract/Free Full Text]
  16. Vita, N., Lefort, S., Laurent, P., Caput, D., and Ferrara, P. (1995) J. Biol. Chem. 270, 3512-3517 [Abstract/Free Full Text]
  17. Fanslow, W. C., Clifford, K., Vanden Bos, T., Teel, A., Armitage, R. J., and Beckmann, M. P. (1990) Cytokine 2, 398-401 [Medline] [Order article via Infotrieve]
  18. Fernandez-Botran, R., and Viteffa, E. S. (1991) J. Exp. Med. 174, 673-681 [Abstract]
  19. Keegan, A. D., Beckmann, M. P., Park, L. S., and Paul, W. E. (1991) J. Immunol. 146, 2272-2279 [Abstract/Free Full Text]
  20. Christie, G., Dacey, I., and Weston, B. J. (1995) Cytokine 7, 305-310 [CrossRef][Medline] [Order article via Infotrieve]
  21. Sempowski, G. D., Beckmann, M. P., Derdak, S., and Phipps, R. P. (1994) J. Immunol. 152, 3606-3614 [Abstract/Free Full Text]
  22. Heaney, M. L., and Golde, D. W. (1996) Blood 87, 847-857 [Free Full Text]
  23. Taga, T., Hibi, M., Hirata, Y., Yamata, H., Natsuka, S., Yasukawa, K., Totsuka, T., Yamasaki, K., Hirano, T., and Kishomoto, T. (1989) Cold Spring Harbor Symp. Quant. Biol. 54, Pt. 2, 713-722
  24. Sato, T. A., Widmer, M. B., Finkelman, F. D., Madani, H., Jacobs, C. A., Grabstein, K. H., and Maliszewski, C. R. (1993) J. Immunol. 150, 2717-2723 [Abstract/Free Full Text]
  25. Layton, M. J ., Cross, B. A., Metcalf, D., Ward, L. D., Simpson, R. J., and Nicola, N. A. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 8616-8620 [Abstract]
  26. Tomida, M., Yamamoto-Yamaguchi, Y., and Hozumi, M. (1993) FEBS Lett. 334, 193-197 [CrossRef][Medline] [Order article via Infotrieve]
  27. Contreras, M. A., Bale, W. F., and Spar, I. L. (1983) Methods Enzymol. 92, 277-292 [Medline] [Order article via Infotrieve]
  28. Hilton, D. J., and Nicola, N. A. (1992) J. Biol. Chem. 267, 10238-10247 [Abstract/Free Full Text]
  29. Bolton, A. E., and Hunter, W. M. (1973) Biochem. J. 133, 529-539 [Medline] [Order article via Infotrieve]
  30. Mizushima, S., and Nagata, S. (1990) Nucleic Acids Res. 18, 5322 [Medline] [Order article via Infotrieve]
  31. Munson, P. J., and Rodbard, D. (1980) Anal. Biochem. 107, 220-239 [Medline] [Order article via Infotrieve]
  32. Heakeshoven, J., and Dernick, R. (1988) Electrophoresis 9, 28-32 [Medline] [Order article via Infotrieve]
  33. Reid, G. E., Hong, J., Eddes, J. G., Moritz, R. L., and Simpson, R. J. (1995) Electrophoresis 16, 1120-1130 [Medline] [Order article via Infotrieve]
  34. Luttmann, W., Knoechel, B., Foerster, Matthys, H., Virchow, J. C., Jr., and Kroegel, C. (1996) J. Immunol. 157, 1678-1683 [Abstract]
  35. Caput, D., Laurent, P., Kaghad, M., Lelias, J.-M., Lefort, S., Vita, N., and Ferrara, P. (1996) J. Biol. Chem. 271, 16921-16926 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.