The Murine IL-13 Receptor α2: Molecular Cloning, Characterization, and Comparison with Murine IL-13 Receptor α1¹

Interleukin-13 is a pleiotropic cytokine that is secreted by activated Th2 cells with immunoregulatory activities that partially overlap those of IL-4 (1). Both cytokines are members of the IL-4 cytokine subfamily and are encoded by closely linked genes within the mouse chromosome 11 cytokine gene cluster (2). The redundancy in biologic responses to IL-13 and IL-4 may be explained by shared components in the receptors for IL-4 and IL-13 (3, 4). Evidence for shared receptor components includes experiments in which a mutant IL-4 protein, Y124D, which is a potent antagonist of IL-4, also blocks the biologic activity of IL-13 (3). Abs directed against the IL-4R chain have been shown to inhibit the response of TF-1 cells and monocytes to IL-13 and IL-4, indicating that the receptors for IL-4 and IL-13 require this protein (4, 5). Two additional potential components of the IL-13R have recently been cloned. Hilton et al. (6) have identified NR4 or murine IL-13Rα1³ (mIL-13Rα1),⁴ a murine hematopoietin receptor chain that binds IL-13 with low affinity (kDa = 2–10 nM). Human IL-13Rα1 (hIL-13Rα1) has also been isolated; this component binds IL-13 with low affinity (7, 8, 9). Caput et al. (10) have identified a different human hemopoietin receptor chain, hIL-13Rα2, that binds IL-13 with higher affinity (kDa = 0.25 nM). In the present work, we describe the identification of mIL-13Rα2 cDNA and the genetic mapping of the mIL-13Rα1 and mIL-13Rα2 loci. Furthermore, we have functionally compared the two distinct IL-13-binding polypeptides.

We have found that many characteristics of the hIL-13Rα2 are conserved in the mouse; however, there are some interesting differences. This observation is of particular importance given the known differences between responses in humans and mice to IL-13. A detailed understanding of the mechanisms by which IL-13 regulates immune responses in the mouse may enhance the usefulness of mouse models of human diseases such as asthma and allergy.

mIL-3 and mIL-6 were obtained from titered COS-1 and Chinese hamster ovary (CHO) cell supernatant. All other cytokines were purchased from R&D Systems (Minneapolis, MN). The Ba/F3 cell line (11) was maintained in RPMI 1640, 10% (v/v) FBS, and 0.05% (v/v) CHO mIL-3-conditioned medium. The B9 cell line was cultured in RPMI 1640, 10% (v/v) FBS, 20 nM of β-mercaptoethanol, and 0.01% (v/v) COS-1 mIL-6-conditioned medium (12).

Poly(A)⁺ RNA was prepared from the thymii of 6- to 8-wk-old C3H/HeJ mice, and a cDNA library of 1.5 × 10⁶ primary recombinant phage was prepared according to the manufacturer’s instructions using the Zap Express Kit from Stratagene (La Jolla, CA). A total of 200,000 plaques were screened with a degenerate 17-base oligonucleotide probe of the sequence 5′-dKSRCTCCABK CRCTCCA-3′ (K = G or T; S = C or G; R = A or G; and B = C or G or T) using standard tetramethylammonium chloride hybridization conditions (13).

A partial fragment of the human homologue of this receptor was isolated by PCR using oligonucleotides that had been derived from the mouse sequence. cDNA was prepared from human testis poly(A)⁺ RNA that had been obtained from Clontech (Palo Alto, CA). A DNA fragment of 274 base pairs (bp) was amplified from this cDNA by PCR with the oligonucleotides 5′-dATAGTTAAACCATTGCCACC-3′ and 5′-CTCCATTCGCTCCAAATTCC-3′ using Taq polymerase in 1× Taq buffer containing 1.5 mM MgCl₂ for 30 cycles (94°C for 1 min, 42°C for 1 min, and 72°C for 1 min). The DNA sequence of this fragment was determined, and two oligonucleotides were prepared from an internal portion of this fragment with the following sequence: 5′-dAGTCTATCTTACTTTTACTCG-3′ and 5′-dCATCTGAGCAATAAATATTCAC-3′. These oligonucleotides were used as probes to screen a human testis cDNA library that had been purchased from Clontech (catalog no. HL1161). Filters were hybridized at 52°C using standard 5× SSC hybridization conditions and washed in 2× SSC at 52°C. A total of 22 clones were isolated that hybridized to both oligonucleotides in a screen of 400,000 clones. The DNA sequence was determined from four of the cDNA clones.

Primers were designed to amplify a region corresponding to the 3′ untranslated region of either IL-13Rα2 (A25) or IL-13Rα1 (NR4) to test for single-strand conformation polymorphisms (SSCPs) between mouse strains. These SSCPs were analyzed as described previously (14). Briefly, oligonucleotides were radiolabeled with [³²P]ATP using polynucleotide kinase, and genomic DNA from a series of mouse strains was amplified using standard protocols (anneal at 55°C for 1 min, extend at 72°C for 2 min, and denature at 94°C for 1 min for 40 cycles, with a final extension at 72°C). A volume of 2 μl of the amplified reaction was added to 8.5 μl of stop solution (United States Biochemical, Cleveland, OH), denatured at 94°C for 5 min, and immediately placed onto ice. An aliquot of 2 μl of each reaction was loaded on a nondenaturing 6% acrylamide-sequencing gel and electrophoresed in 0.5× Tris-borate/EDTA buffer for 2 to 3 h at 40 watts in a 4°C cold room. A primer pair with the sequence 5′-dCCACCAATTTCTTGACATAGAGC-3′ (forward) and 5′-dACTTTAGTCAAAAACCCAACACT-3′ (reverse) identified a polymorphism between C57BL/6J and Mus spretus for IL-13Rα2, and a primer pair with the sequence 5′-dGCATTCTCCATTTGTTATCTG-3′ (forward) and 5′-dTCCCCTTGTCTTTGTCCTCTA-3′ (reverse) identified a polymorphism between C57BL/6J and M. spretus for IL-13Rα1. These primers were used to analyze DNA prepared from the BSS backcross (15). The strain distribution patterns were analyzed using the Map Manager Program (16).

Northern blots of poly(A)⁺ RNA from various tissues (Clontech) were performed as recommended by the manufacturer. An RT-PCR analysis of mIL-13Rα2 RNA expression in murine tissues was performed as follows: Poly(A)⁺ RNAs were purchased from Clontech, and 2 μg of each RNA was treated with RQ1 RNase-free DNase (Promega, Madison, WI) according to the manufacturer’s instructions. Following DNase inactivation, each sample was divided into two tubes; random hexamer-primed first strand cDNA synthesis (Superscript II Preamplification System; Life Technologies, Gaithersburg, MD) was conducted according to the manufacturer’s instructions, except that the reverse transcriptase was omitted from one tube in each sample pair. Reaction products were purified using the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) as described by the manufacturer and recovered in 50 μl of double-distilled H₂O. We conducted 50-μl PCRs using 5 μl of the recovered material and Advantage cDNA Polymerase Mix (Clontech) according to the manufacturer’s instructions. A total of 25 reaction cycles were performed at 94°C for 1 min, at 64°C for 1 min, and at 72°C for 1 min for actin and at 96°C for 1 min, at 56°C for 1 min and at 72°C for 1 min for IL-13Rα2 in a Perkin-Elmer Cetus Thermal Cycler (Branchburg, NJ). The oligonucleotide primers that were used for the amplification of mIL-13Rα2 were 5′-dATGGCTTTTGTGCATATCAGATGCT-3′ (forward) and 5′-dCAGGTGTGCTCCATTTCATTCTAAT-3′ (reverse), which generate an 800-bp fragment; the primers used for γ-actin were 5′-dCAGAGCAAGAGGGGTATCCTGA (forward) and 5′-dTGATCCACATCTGCTGGAAGGT-3′ (reverse), which generate a 900-bp fragment. An aliquot of each reaction was electrophoresed on a 0.8% agarose (w/v) gel and transferred to nitrocellulose. The membranes were hybridized at 55°C for 18 h to end-labeled oligonucleotides that were specific for mIL-13Rα2 5′dTGCCACCAGAATTCCTTCATATTAG-3′ (internal) and γ-actin 5′dTGAAGATCCTGACTGAACGGGGCTACAGCTTT-3′ (internal) followed by autoradiography.

Expression vectors for mIL-13Rα1 (6) and IL-13Rα2 were constructed in pLNCX (17). Ba/F3 cells were transfected with 20 μg of either mIL-13Rα2 or mIL-13Rα1 linearized expression plasmids by electroporation using the Gene Pulsar (Bio-Rad, Richmond, CA) at 270 V and 960 μF. Pools of stable transfectants (Ba/F3.IL-13Rα2 or Ba/F3.IL-13Rα1) were isolated by selection in 1.0 mg/ml of G418 (Life Technologies).

Expression vectors were also constructed that encode either full-length mIL-13Rα2 or mIL-13Rα1 fused in frame at the C terminus to sequences encoding either T7 or hemagglutinin (HA) epitope tags and inserted into the COS-1 expression vector, pED. COS monolayers that had been transiently transfected with mIL-13Rα2.T7 or mIL-13Rα1.HA expression plasmids or mock transfected were radiolabeled with l-[³⁵S]methionine and l-[³⁵S]cysteine (Amersham, Arlington Heights, IL). Proteins containing the T7 epitope tag were precipitated from detergent-solubilized cell extracts by incubation with a murine T7.Tag mAb that had been conjugated to agarose (Novagen, Madison, WI). Proteins containing the HA epitope tag were precipitated by incubation with a murine mAb against the HA peptide epitope (Boehringer Mannheim, Indianapolis, IN) and with goat anti-mIgG that had been conjugated to Sepharose 4B (Zymed, San Francisco, CA). Immunoprecipitated material was examined by reducing SDS-PAGE that was followed by treatment with Amplify (Amersham) and fluorography.

Expression vectors encoding either the extracellular domain of IL-13Rα2 (amino acids (aa) 1–332) or IL-13Rα1 fused to a Gly-Ser-Gly spacer and the sequence encoding the hinge-CH2-CH3 regions of hIg-γ1 were also constructed in pED. The mIL-13Rα2.Fc or mIL-13Rα1.Fc protein was produced from stably transfected CHO cells and purified via protein A Sepharose chromatography (Pharmacia, Uppsala, Sweden) (18).

A BIACORE 2000 instrument (Biacore AB, Uppsala, Sweden) was used to show IL-13 binding to IL-13Rα2.Fc (19). Purified IL-13Rα2.Fc, IL-11Rα.Fc, or hIgG1 (The Binding Site, Birmingham, U.K.) in 10 mM sodium acetate (pH 4) were each coupled to the carboxymethyl dextran layer of a CM5 research-grade chip using standard amine-coupling procedures. Typically, ∼10,000 to 17,000 resonance units (RUs) were immobilized by this procedure. The running buffer used for screening cytokine binding was 10 mM sodium acetate (pH 7.4), 300 mM sodium chloride, 3.4 mM EDTA, and 0.005% (v/v) Tween 20. Purified IL-13, IL-4, IL-7, and IL-15 (R&D Systems) were separately injected at 10 μg/ml in series over the immobilized IL-13Rα2.Fc, IL-11Rα.Fc, and IgG1 for 10 min at 2 μl/min. Binding was quantified as the increase in RUs at 60 s after the end of injection compared with a baseline established at 20 s before injection.

Carrier-free mrIL-13 was purchased from R&D Systems and radioiodinated with ¹²⁵I-labeled N-succinimidyl p-iodobenzoate (New England Nuclear, Boston, MA) according to the manufacturer’s instructions. ¹²⁵I-labeled mIL-13 (¹²⁵I-mIL-13) was purified by fractionation over a PD-10 column (Pharmacia) followed by ultrafiltration on a Centricon 3 microconcentrator (Amicon, Beverly, MA). ¹²⁵I-mIL-13 was >99% TCA precipitable and displayed a specific activity of 10,000 cpm/ng by bioassay on B9 cells; >90% ¹²⁵I-mIL-13 could be bound by Ba/F3.IL-13Rα2 cells. For binding assays, 2.5 × 10⁵ to 1 × 10⁶ Ba/F3.IL-13Rα2 or Ba/F3 cells were incubated with 0 to 10 nM ¹²⁵I-mIL-13 in 100 μl of media with or without 100× excess unlabeled IL-13 at 4°C for 4 h. Cell-bound ¹²⁵I-mIL-13 was separated from free ¹²⁵I-mIL-13 by centrifugation over serum, and radioactivity was determined for each pellet and supernatant. Curve fitting of the binding data was performed using the Ligand program and plotted in the Scatchard format (20).

The stimulation of proliferation of B9, Ba/F3, Ba/F3.IL-13Rα2, and Ba/F3.IL-13Rα1 cells in response to IL-13 or IL-4 was measured by [³H]thymidine (New England Nuclear) incorporation into DNA. Cells (0.5–1 × 10⁴/well) were seeded into 96-well plates in 200 μl of media with and without growth factors. After incubation for 3 days, 1 μCi/well of [³H]thymidine was added, and the cells were incubated for an additional 4 h. The incorporated radioactivity was determined using an LKB 1205 plate reader (LKB Pharmacia, Gaithersburg, MD).

The hemopoietin receptor superfamily is characterized by a conserved aa motif, WSXWS, in the extracellular domain (21). We screened an adult murine thymus cDNA library using a degenerate 17-base oligonucleotide that encodes a subset of sequences containing this motif. Clone A25 was identified as a positive that did not hybridize to probes derived from known cytokine receptors.

Sequencing of the 1.5-kilobase (kb) murine A25 cDNA revealed a 1149-bp open reading frame that was predicted to encode a 383-aa protein. The predicted sequence contains a putative signal peptide, an extracellular domain with four potential N-linked glycosylation sites, a transmembrane domain, and a short cytoplasmic domain (Fig. 1). The extracellular domain contains an N-terminal fibronectin type III domain followed by a typical cytokine receptor module at residues 139 to 321. This cytokine receptor module includes a WSEWS motif and the four cysteine residues that are the most conserved features of the hemopoietin receptor superfamily. The A25 domain structure is shared with the murine receptors for granulocyte-macrophage CSF, IL-3, IL-5, and mIL-13Rα1 (NR4). The extracellular region of A25 is most closely related to mIL-13Rα1, with 29% identical aa residues (Fig. 1). However, in contrast to mIL-13Rα1, the short cytoplasmic tail of A25 contains no box 1 or box 2 signaling motifs (22). A25 shows 59% overall aa identity with hIL-13Rα2 (10), suggesting that it is the murine orthologue (Fig. 1).

The genes encoding both mIL-13Rα2 and mIL-13Rα1 were mapped by SSCP analysis (14). IL-13Rα2 and IL-13Rα1 were both found to map to chromosome X, with log of the odds likelihood scores of 26.8 and 18.5, respectively. No recombinants were found between IL-13Rα2 and DXMit34 in the 89 progeny that were scored. The position of IL-13Rα2 with respect to flanking microsatellite markers is: DXMit4 − 6.4 ± 2.5 cM − (IL-13Rα2, DXMit34) − 7.9 ± 2.9 cM − DXMit120. No recombinants were found between IL-13Rα1 and angiotensin receptor-2 (Agtr2) in the 80 progeny that were scored. The position of IL-13Rα1 with respect to flanking markers is: DXMit85 − 3.8 ± 2.1 cM − (IL-13Rα1, Agtr2) − 3.8 ± 2.1 cM − DXMit49.

We also examined the expression of murine and human IL-13Rα2 RNA by Northern blotting and RT-PCR (Fig. 2). Two transcripts of 1.5 and 2.1 kb were detected in the murine spleen, brain, and 7-day-old embryo. The presence of mIL-13Rα2 RNA was further confirmed in the spleen and brain by RT-PCR. A single transcript of 2.1 kb was seen in the human liver, lung, and thymus, and a smaller 1.5-kb transcript was observed in the placenta, brain, and heart. An intensely labeled transcript in human testis was detected and was shown to be a single band of 2.1 kb upon shorter exposure. Testis RNA from a second pooled source showed the same intensity of signal. Apparent differences in transcript size in the RNA preparations of different human tissues did not correlate with changes in actin transcript mobility. The sequence comparisons of four independent cDNA clones from human testis were identical in the coding region and ranged in size from 1.1 to 1.3 kb.

Results from cross-linking studies of hIL-13 and hIL-13R have shown specific binding proteins of similar molecular masses from a variety of tissue sources. Therefore, we wished to directly compare the observed molecular masses of noncross-linked mIL-13Rα1 and mIL-13Rα2. The mIL-13Rα2 and mIL-13Rα1 cDNA were modified to encode epitope tags at the carboxyl terminus of each protein and inserted into a COS expression vector. COS cells that had been transiently transfected with these plasmids were biosynthetically labeled with l-[³⁵S]methionine and l-[³⁵S]cysteine and examined by immunoprecipitation with mAb specific for the T7 (for IL-13Rα2) or HA (for IL-13Rα1) epitopes. SDS-PAGE and fluorography resulted in the detection of both polypeptides; each polypeptide had a similar apparent molecular mass of ∼56 kDa (Fig. 3).

The specificity of the interaction between IL-13 and mIL-13Rα2 was investigated by surface plasmon resonance using a BIACORE 2000 instrument. IL-4, IL-5, IL-7, IL-9, IL-15, and IL-13 were screened for binding to immobilized mIL-13Rα2.Fc, and only IL-13 binding was detected. None of these cytokines were found to interact with IL-11Rα.Fc or human IgG1 (Fig. 4). Similar results were obtained for hIL-13Rα2.Fc protein; these results were consistent with results reported previously (10).

To determine the affinity of IL-13 for mIL-13Rα2, we used Ba/F3 cells that had been stably transfected with the mIL-13Rα2 expression plasmid. IL-13 was iodinated and bound to the transfected cells. A Scatchard analysis of the binding data from three separate experiments revealed that these cells expressed ∼5000 IL-13-binding sites of a single K_d of 0.5 to 1.2 nM (Fig. 5). No binding was detected on untransfected Ba/F3 cells.

Hilton and colleagues have reported that the expression of mIL-13Rα1 (NR4) by their Ba/F3 subline was not sufficient to allow a proliferative or survival response to IL-13 (6). In contrast, the expression of mIL-13Rα1 in CTLL cells did confer a proliferative response to IL-13 (6). The potential signaling functions of hIL-13Rα2 have not yet been investigated. Therefore, we examined the proliferative responses of IL-3-dependent Ba/F3 cells that had been transfected with mIL-13Rα2 or mIL-13Rα1 to assess the relative contributions of each protein to receptor function. Ba/F3 cells expressing mIL-13Rα2 were unable to proliferate in response to IL-13, even though they expressed high numbers of receptors per cell (Figs. 5 and 6). Surprisingly, Ba/F3 cells that had been transfected with mIL-13Rα1 responded to IL-13 and proliferated with a half-maximal response of 200 pg/ml. The half-maximal proliferation to IL-4 occurred at ∼1.2 ng/ml and was unchanged from untransfected Ba/F3 cells even in the presence of high levels of IL-13.

The effect of a dimeric form of mIL-13Rα2, the IL-13Rα2.Fc fusion protein, on the proliferative response of Ba/F3.IL-13Rα1 cells to IL-13 was examined. mIL-13Rα2.Fc was added at various concentrations to cultures of Ba/F3.IL-13Rα1 cells that contained a saturating concentration (3 ng/ml) of mIL-13. Figure 7 shows that the concentration of fusion protein giving a 50% level of inhibition of IL-13-stimulated proliferation is 8 ng/ml. Similar results were obtained using B9 cells (data not shown). In contrast, mIL-13Rα1.Fc was at least 100-fold less effective in neutralizing the proliferative effect of IL-13 than mIL-13Rα2.Fc (Fig. 7). This difference in potency is consistent with the lower affinity (K_d = 2–10 nM) of IL-13 for mIL-13Rα1 (6) when compared with the higher affinity of IL-13 for mIL-13Rα2 (K_d = 0.5 nM).

Two distinct members of the hemopoietin receptor family have been shown to specifically bind IL-13 with differing affinities. The lower affinity subunit, IL13-Rα1, has been shown to participate in the formation of an IL-13R complex in combination with the IL-4R (6, 7, 8, 9). A higher affinity subunit, which had only been described previously in humans, is designated hIL13-Rα2 (9). However, the contribution of this IL-13-binding chain to receptor function has not been elucidated.

We have used motif-based cloning to identify a novel member of the hemopoietin receptor family, designated mIL-13Rα2, which encodes the murine orthologue of hIL-13Rα2. Comparisons of the human and murine receptors show that they are structurally conserved, share 59% aa identity, and specifically bind IL-13 with high affinity. Within the hemopoietin receptor family, we show that the extracellular region of mIL-13Rα2 is most closely related to mIL-13Rα1. In contrast, the cytoplasmic domains of mIL-13Rα2 and mIL-13Rα1 are dissimilar. Importantly, the cytoplasmic domains of both the murine and human IL-13Rα2 are short and devoid of box 1 or box 2 signaling motifs. However, hIL-13Rα2 does contain a putative consensus phosphorylation site, Y₃₆₉PKM (23), which is potentially suitable as a docking site for an SH2-containing signaling molecule.

We also show that two IL-13Rα2 mRNA transcripts of 1.5 and 2.1 kb are expressed in the murine brain, spleen, and day 7 embryo, although lower levels were detected in the liver, lung, and testis with longer exposure (data not shown). Isolation of IL-13Rα2 from a thymus cDNA library also indicated that this gene is expressed in the murine thymus at a low level (1/200,000 cDNA clones). Interestingly, the transcript is also expressed in RNA isolated from mouse embryos, with strong expression at day 7. This observation suggests that IL-13 may have an important role in embryogenesis.

An hIL-13Rα2 mRNA transcript of either 1.5 or 2.1 kb was detected and was found to have a broad range of expression in primary tissues. A particularly high level of expression was observed in human testis. The cDNA clones that were isolated by us from the testis were identical in sequence to the cDNA that was isolated by Caput et al. (10) from human renal carcinoma cells as well as to several partial cDNA clones that were isolated from human brain RNA (data not shown). The difference in tissue-specific transcript expression between mice and humans may be related to differences in IL-13 function between the two species. For example, IL-13 has been shown to induce IgE class switching in human but not in murine B cells. Another observation is that this receptor is also expressed in tissues that are not typically associated with immune function, such as the testis and brain. It will be interesting to determine which cells in these tissues are expressing IL-13Rα2, and whether they are cells that might be associated with immune surveillance.

The full-length murine and human cDNAs isolated by us correspond to the smaller 1.5-kb transcripts. Zhang et al. have determined the partial sequence of a soluble form of IL-13-binding protein in mouse serum and urine that matches that of the predicted N terminus of mIL-13Rα2 (2A). This soluble form may be generated by proteolytic processing. Alternatively, the larger 2.1-kb mIL-13Rα2 transcript may encode this form.

Using SSCP analysis, we localized the genes for mIL-13Rα1 and IL-13Rα2 to the murine X chromosome. This finding is consistent with the reported chromosomal locations of hIL-13Rα1 and hIL-13Rα2 on the X chromosome (7, 25). Although the hIL-13Rα2 gene has been precisely mapped to Xq24 (25), the precise location of the hIL-13Rα1 on the X chromosome has not been determined. We have found that mIL-13Rα1 is nonrecombinant with mAgtr2. Since linkage relationships in this subchromosomal region are highly conserved between mice and humans, our data predict that the hIL-13Rα1 would also be tightly linked to hAgtr2, which maps to human chromosome Xq22–23 (26). The presence of these two genes on the X chromosome suggests that they should be considered as candidate genes in X-linked immune disease.

Our BIACORE data demonstrate that mIL-13Rα2 specifically binds IL-13. In particular, we were unable to detect the binding of IL-4 at high concentrations to mIL-13Rα2. Our affinity measurements for the mIL-13Rα2 that is expressed in Ba/F3 cells ranged from 0.5 nM to 1.2 nM and are similar but not identical with the 0.25 and 0.44 nM affinity that was determined for the hIL-13Rα2 in COS and Caki cells, respectively (10). This discrepancy may reflect a true difference between the human and murine proteins or may reflect experimental variation (e.g., differences due to methods of iodination). These affinities are in contrast to the lower affinity that was measured for cells expressing either mIL-13Rα1 (2–10 nM) (6) or hIL-13Rα1, which could not be detected unless both IL-13Rα1 and IL-4R were coexpressed (7).

The structural differences between the cytoplasmic domains of the high and low affinity IL-13-binding chains suggest that these chains are functionally distinct. To test this possibility, we examined the proliferative responses to IL-13 in Ba/F3 cells that had been transfected with either the low affinity mIL-13Rα1 chain or the high affinity mIL-13Rα2 chain. These cells express the IL-4R, and the mitogenic response to IL-4 was unchanged by the heterologous expression of these IL-13-binding proteins. We found that the expression of the IL-13Rα1 chain led to a proliferative response to IL-13, presumably by the formation of an IL-13Rα1/IL-4R complex. In contrast, expression of the IL-13Rα2 was not sufficient to generate a mitogenic response to IL-13.

The extracellular region of mIL-13Rα2 has been fused to the Fc portion of hIgG1 to generate a soluble form of this IL-13-binding protein. We have demonstrated that this molecule is a specific inhibitor of IL-13 in vitro and can block the IL-13-stimulated proliferation of Ba/F3 cells that have been transfected with the IL-13Rα1. We have also made an IL-13Rα1.Fc fusion protein and examined the ability of this fusion protein to inhibit the proliferation of an IL-13-dependent cell line. This inhibition requires 100-fold more IL-13Rα1.Fc fusion protein than IL-13Rα2.Fc, which is consistent with its lower affinity for IL-13. We have found that soluble IL-13Rα2.Fc fusion protein can inhibit the action of IL-13 in murine models in vivo (27).

The effects of the IL-13Rα2.Fc fusion protein and the relative affinities of the two IL-13-binding proteins raise the possibility that the IL-13Rα2 is a dominant negative inhibitor of or a decoy for IL-13. This function would be paralleled by IL-1, the IL-1 type II receptor (28), and the recently described decoy receptor for TRAIL (29, 30). However, our observations are insufficient to conclude that the IL-13Rα2 performs such a role in physiologic circumstances.

It is also possible that an undiscovered molecule acts with the IL-13Rα2 to form a second IL-13-signaling complex. In addition, the IL-13Rα2 may have a complementary role with the mIL-13Rα1 to facilitate the formation of a high affinity signaling complex that has not yet been detected in our work (7).

IL-13 has been shown to have various effects on APCs in vitro, including the modulation of cell surface molecules, the inhibition of inducible nitric oxide synthase and cyclooxygenase-2 expression, the stimulation of intracellular Ca²⁺ flux, and the induction of IgE class switching in human B cells (31, 32, 33, 34, 35). The role of the high affinity IL-13-binding protein, IL-13Rα2, in these processes is unclear. Additional functional studies are required to understand the relative roles of the two IL-13-binding proteins.

The demonstration in the present work of two X chromosome-linked genes for both a high and a low affinity IL-13-binding protein that are structurally conserved between humans and mice suggests that the regulation of IL-13 responses may be more complex than previously appreciated. The conservation of an additional chain that could specifically enhance IL-13 binding, contribute to an additional IL-13R complex, or block IL-13 response raises the possibility that IL-13, rather than IL-4, may be a more important mediator in vivo on cells expressing an IL-13R.

We thank John Morris, Andrew Long, and Paul Burgess for excellent technical assistance in transfections and conditioned media production and Lucy Rowe for help with data analysis.