An Interleukin (IL)-13 Receptor Lacking the Cytoplasmic Domain Fails to Transduce IL-13-Induced Signals and Inhibits Responses to IL-4*

Interleukin (IL)-13 is a pleiotropic immunoregulatory cytokine that shares many, although not all, of the biological activities of IL-4. The overlapping biological properties of IL-4 and IL-13 appear to be due to the existence of shared components of the receptors, and we and others showed that the IL-4 receptor-α is involved in signal transduction paths activated by both. We show here that expression of the IL-13 receptor-α in two factor-dependent cell lines, the premyeloid FD5 and the T lymphoid CT4.S, conferred the ability to grow continuously in response to IL-13; to respond to IL-13 with tyrosine phosphorylation of JAK1, Tyk2, IL-4Rα, IRS-2, and STAT6; and to respond to IL-4 with tyrosine phosphorylation of Tyk2 in addition to those induced in parental cell lines. Expression of a truncated IL-13 receptor-α that lacked the cytoplasmic domain demonstrated that this domain was essential for IL-13-dependent growth and phosphorylation of the above substrates. Expression of this truncated IL-13 receptor also resulted in an inhibition of biochemical and biological responses to IL-4 that was exacerbated by the presence of IL-13. These dominant inhibitory effects indicate that the extracellular domain of the truncated IL-13 receptor competes with γc for complexes of IL-4 and the IL-4 receptor-α, or, when itself bound to IL-13, competes with IL-4 for the IL-4 receptor-α.

synthesis (3). They induce expression of the low affinity receptor for IgE (Fc⑀RII/CD23) and up-regulate class II major histocompatibility complex expression on both B cells and monocytes (4,5). In monocytes, IL-4 and IL-13 down-regulate Fc␥ receptor surface expression (5) and inhibit synthesis of inflammatory cytokines, including tumor necrosis factor-␣, IL-1␤, IL-6, and IL-8 (5,6). Moreover, they suppress synthesis of IL-12 (5), a critical cytokine for differentiation of uncommitted T cells toward the Th1 phenotype (7). It appears that every cellular response to IL-13 can also be mediated by IL-4. However, the reverse is not true; human T cells and murine T cells and B cells respond to IL-4 but not to IL-13.
The overlapping activities of IL-4 and IL-13 are probably due to the existence of common receptor components, as revealed by studies demonstrating cross-competition between IL-13 and IL-4 for binding to certain cells (8,9). The best characterized form of the IL-4 receptor is composed of a 140-kDa transmembrane glycoprotein (IL-4R␣), which binds IL-4 with a K d of 50 -600 pM, depending on the cell type (10 -13), and the ␥c chain of the IL-2 receptor, which, when associated with IL-4R␣ and IL-4, results in a 2-3-fold increase in affinity for IL-4 (14,15). Neither of these two proteins alone or complexed binds IL-13. A specific binding subunit for IL-13 has been recently cloned in the murine (16) and human (17) systems. The IL-13 receptor-␣ (IL-13R␣) has an apparent molecular mass of 55-65 kDa, and it binds IL-13 with a K d of 2-10 nM when transfected alone into COS cells (16) or 293 fibroblasts (17). However, when human IL-13R␣ was cotransfected with human IL-4R␣ into 293 cells and murine IL-13R␣ was transfected into CTLL-2 cells that express IL-4R␣, the K d for IL-13 appeared to be 400 and 75 pM, respectively (16,17), suggesting that the IL-4R␣ and the IL-13R␣ formed a higher affinity complex with IL-13. Expression of the IL-13R␣ in CTLL-2 cells conferred upon these cells the ability to respond to IL-13 with a brief burst of DNA synthesis; long term growth was not investigated, but it is unlikely to occur, since CTLL-2 cells do not continuously grow in response to IL-4. IL-4 and IL-13 were equally effective in competing for 125 I-labeled IL-13 binding (16,17), indicating that IL-4R␣ was part of the IL-13R-binding complex. Recently, another IL-13-binding protein (IL-13R␤) has been cloned from the Caki-1 human renal carcinoma cell line (18). This protein has a short cytoplasmic domain (only 17 amino acid residues), and bound IL-13 with a K d of 250 pM when transfected alone into COS cells. Cotransfection of IL-13R␤ together with IL-4R␣ did not decrease K d values for binding of IL-13. Neither IL-4R␣, ␥c, nor IL-13R␣ encode kinase sequences in their cytoplasmic domains, but all three have Pro-rich sequences (Box 1) that might be involved in docking JAK family kinases.
IL-4 and IL-13 activate similar signal transduction pathways. Both induce activation of JAK1, but only IL-4 activates * This work was supported by the Arthritis Society of Canada. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 1 The abbreviations used are: IL, interleukin; IL-13R␣, IL-13 receptor ␣ chain; IL-4R␣, IL-4 receptor ␣ chain; ␥c, ␥ common chain of IL-2, IL-4, IL-7, IL-9, and IL-15 receptors; IRS-2, insulin receptor substrate-2; STAT6, signal transducer and activator of transcription-6; LCCM, L cell conditioned media; IL-4-CM, IL-4-containing conditioned medium; CSF, colony-stimulating factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; JAK, Janus kinase; Ab, antibody; mAb, monoclonal antibody; kb, kilobase pair(s); FCS, fetal calf serum; FITC, fluorescein isothiocyanate. JAK3; Tyk2 is tyrosine-phosphorylated in response to IL-13 but also in response to IL-4 in cells that respond to IL-13, including human myeloid cells and a mouse plasmacytoma cell (19 -21). IL-13 induces tyrosine phosphorylation of the IL-4R␣ chain (20,22) and activates two signaling pathways characteristic of IL-4 responses, namely one involving insulin-receptor substrate 1 and 2 (IRS-1/2) (20) and the other involving STAT6 (23). Activated STAT6 binds to the promoters of several genes known to be regulated by IL-4/IL-13 (24 -28) that contain the consensus STAT6-binding site TTCNNNNGAA (29) and is involved in Th2 differentiation, expression of cell surface markers, IgE switching (30 -32), and IL-4-induced proliferation (30,32). IRS-2 and IRS-1 are phosphorylated on tyrosine residues in response to IL-4 and IL-13 and associate with the p85 subunit of phosphatidylinositol 3Ј-kinase (20,33). IRS-2 is not necessary for proliferative responses to IL-4 (34), although its presence enhanced proliferation in response to IL-4 in certain cell lines (33,35,36). IL-4 and IL-13 also share the characteristic that in lymphohematopoietic cells, they fail to induce tyrosine phosphorylation of Shc or its association with Grb2; modification of Sos1 (37); activation of Erk-1 and Erk-2 mitogen-activated protein kinases (38); activation of p38 mitogenactivated protein kinase (39); or activation of SAP-kinases (40). 2 These data are consistent with a model in which there are two types of IL-4Rs but a single type of IL-13R. IL-4 would first bind to the IL-4R␣, and this complex would interact with either the IL-2R␥ or the IL-13R␣ to yield active receptors. IL-13, in contrast, would first bind to the IL-13R␣, and this complex would then interact with the IL-4R␣ to form an active receptor (20,25).
We report here the crucial role of IL-13R␣, not only in IL-13but also in IL-4-induced responses. Transfection of IL-13R␣ cDNA in FD5, a premyeloid cell line, or CT4.S, a T cell line, resulted in continuous IL-13-dependent growth and IL-13-dependent phosphorylation of JAK1, Tyk2, IL-4R␣, IRS-2, and STAT6. In contrast, a truncated IL-13R␣, encoding the extracellular and transmembrane domains but lacking the cytoplasmic domain, failed to mediate detectable biological or biochemical responses to IL-13. Moreover, cells overexpressing this truncated IL-13R␣ showed a greatly decreased response to IL-4 in terms of JAK1, JAK3, Tyk2, IL-4R␣, IRS-2, and STAT6 phosphorylation and proliferation. This effect was exacerbated by the presence of IL-13 and reflected sequestration of the IL-4R␣ into sterile complexes of the truncated IL-13R␣ and IL-4 or IL-13. These results indicate that the cytoplasmic domain of IL-13R␣ is necessary for signaling by IL-13 and in some cases IL-4 and that the extracellular domain of the IL-13R␣ interacts with the extracellular domain of the IL-4R␣ in the presence of either IL-4 or IL-13.
Recombinant Plasmid-The full-length cDNA sequence encoding the mouse IL-13R␣ was cloned into the mammalian expression vector pEF-BOS (44) (pEB-13R) under the transcriptional control of the elongation factor-1␣ promoter and poly(A) adenylation signal from human granulocyte colony-stimulating factor. The coding sequence was preceded by the IL-3 signal sequence and an N-terminal FLAG epitope tag sequence (16). The cDNA encodes the extracellular domain of IL-13R␣ (amino acids 27-340), the transmembrane region (amino acids 341-364), and the intracellular domain (amino acids 365-424) (Fig. 1A). The receptor sequence encoding the cytoplasmic domain was removed from pEB-13R by digestion with AflII. Three fragments were generated, one of 4.2 kb encoding most of the vector, another of 1.9 kb encoding a small region of the vector together with the extracellular and transmembrane receptor coding sequences, and a third fragment of 0.5 kb encoding the cytoplasmic domain sequence and part of the poly(A). The fragments of 4.2 and 1.9 kb were purified from 0.8% agarose gels; blunt-ended with Klenow polymerase, creating a new in-frame stop codon downstream of the transmembrane sequence; and ligated to each other. The new plasmid (pEB-13R⌬CD) encodes the entire extracellular and transmembrane receptor sequences and the two membrane-proximal amino acid residues of the cytoplasmic domain (Lys, Arg) (Fig. 1B). The identity of the construct was confirmed by restriction mapping and sequencing.
Transfections and Screening of Protein Expression-FD5 and CT4.S cells, grown in RPMI containing 10% FCS, 3% IL-4-CM, or Dulbecco's modified Eagle's medium containing 10% FCS, 5% IL-4-CM, respectively, were washed and resuspended in transfection buffer (25 mM Hepes, 0.75 mM Na 2 HPO 4 , 140 mM KCl, 5 mM NaCl, 2 mM MgCl 2 , 0.5% Ficoll 400) at a concentration of 1.3 ϫ 10 7 cells/ml. For each transfection, 1 ϫ 10 7 cells were mixed with 1 g of pPGK/Neo, a plasmid conferring neomycin resistance, alone or together with 10 g of either pEB-13R or pEB-13R⌬CD cDNA and were subjected to electroporation using a Bio-Rad gene pulser at 960 microfarads and 280 V or 300 V for FD5 or CT4.S, respectively. In parallel, 10 7 cells were electroporated without DNA to subsequently monitor neomycin-induced death. After transfection, these groups of cells were cultured in the appropriate media for 48 h and then transferred to selection media in 96-well plates at 10 4 cells/well. FD5 cells were selected in RPMI containing 10% FCS, 100 g/ml G418, and 2% LCCM, to avoid selection for responsiveness to IL-4. CT4.S cells were selected in Dulbecco's modified Eagle's medium containing 10% FCS, 1 mg/ml G418, and 2% IL-2-CM to avoid inhibitory effects of IL-4. Individual colonies of neomycin-resistant cells were cloned and propagated for assays. Cells of individual clones were tested for expression of full-length (FD5-13R or CT4.S-13R) or truncated IL-13R␣ (FD5-13R⌬ or CT4.S-13R⌬) by FACScan analysis using M2, a mAb against FLAG, and goat-anti-mouse IgG-FITC and confirmed by immunoprecipitation and immunoblotting with anti-FLAG mAb. A representative clone from each cell population was recloned by limit dilution and used for further experiments.
Cell Proliferation Assay-Cytokine-induced proliferation of cells was assessed by [ 3 H]thymidine incorporation into de novo synthesized DNA (45), cell counting, and microscopic visualization. For assessment of DNA synthesis, cells were washed three times with Hanks' buffered salts supplemented with 2% (v/v) FCS. CT4.S or their transfectants were plated at 500 cells/Terisaki well, and varying concentrations of factor were added from a dilution series. Each point was repeated in triplicate. Cells were incubated at 37°C for 5 days, pulsed for 6 h with [ 3 H]thymidine at a final concentration of 15 Ci/ml, harvested, and counted on a scintillation counter. Results were expressed as a percentage of the maximal [ 3 H]thymidine incorporation observed in cultures stimulated with 5% IL-2-CM. To determine long term growth and morphological changes, parental and transfected cells were plated at low density with different factors and counted, and in some cases photographed, at regular intervals.
Immunoprecipitation, Electrophoresis, and Immunoblotting Analysis-To investigate the biochemical effects of stimulation with different cytokines, selected clones of FD5, FD5-13R, and FD5-13R⌬ were placed in RPMI, 10% FCS, 0.2% LCCM for 16 h, washed twice with serum/factor-free media containing 20 mM Hepes, and incubated at 1-3 ϫ 10 7 cells/ml of the same media at 37°C for 1 h. Cells were then 2 I. Foltz, personal communication. stimulated at 37°C with either synthetic IL-4 (20 g/ml), synthetic IL-13 (20 g/ml), synthetic GM-CSF (10 g/ml) for 10 min, recombinant insulin (5 g/ml) for 2 min, or left untreated as control. Cells were lysed in lysis buffer (20 mM Tris, pH 7.6, 150 mM NaCl, 1% Nonidet P-40, 2 mM EDTA, 1 mM sodium orthovanadate, 1 mM sodium molybdate, 10 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 0.7 g/ml pepstatin, 10 g/ml aprotinin). Lysates were cleared by centrifugation and incubated with antisera against STAT6, IRS-2, JAK1, JAK2, JAK3, or Tyk2 or with mAbs against IL-4R␣ or FLAG for 2 h at 4°C followed by an additional hour with Protein A-or Protein G-Sepharose. The beads were washed three times with cold lysis buffer containing decreasing concentrations of detergent and heated in Laemmli sample buffer (46) containing dithiothreitol at a final concentration of 40 mM. The samples were subjected to SDSpolyacrylamide gel electrophoresis using 7.5 or 8.5% polyacrylamide gels. The gels were then equilibrated in transfer buffer (20 mM Tris, 150 mM glycine, 20% methanol) and transferred to polyvinylidene difluoride membranes by electrophoresis toward the anode at 200 mA for 2 h in a Bio-Rad transblot apparatus (47). The polyvinylidene difluoride membranes were blocked in 4% bovine serum albumin in TBST (10 mM Tris, pH 7.5, 150 mM NaCl, 0.1% Tween 20) for 4 -16 h at room temperature. The blots were then washed twice with TBST and incubated for 2 h with anti-Tyr(P) mAb (0.25 g/ml) or anti-FLAG mAb (1 g/ml) in TBST, 1% bovine serum albumin. The blots were thoroughly washed with TBST and incubated for 1 h with goat anti-mouse IgG antibody-conjugated to horseradish peroxidase (DAKO A/S, Denmark) diluted 1:10,000 in TBST. The blots were thoroughly washed and subsequently developed with the enhanced chemiluminescence assay as described by the manufacturer (ECL kit, Amersham Corp.). Afterward, blots were stripped with erasing buffer (62.5 mM Tris-Cl, pH 6.8, 2% SDS, 100 mM 2-mercaptoethanol) for 1 h at 55°C, washed twice with TBST, blocked with 4% bovine serum albumin in TBST, and incubated for 2 h with antisera against STAT6 (1:2000) The blots were washed several times in TBST and incubated for 1 h in goat anti-rabbit IgG antibody-conjugated to horseradish peroxidase (DAKO A/S, Denmark), 1:10,000 in TBST. Proteins were detected as described previously.

Generation of Transfectants Expressing the Wild-type or
Truncated IL-13R␣-The expression plasmid pEB-13R containing the coding sequence for the full-length IL-13R␣ cDNA (16) (Fig. 1A) was digested with AflII to create the expression plasmid pEB-13⌬CD containing the extracellular, transmembrane and two membrane-proximal amino acid sequences of IL-13R␣ cDNA under the transcriptional control of the elongation factor-1␣ promoter (Fig. 1B). The truncated and full-length receptor expression plasmids were used to transfect FD5 and CT4.S cells, both of which are unresponsive to IL-13 but grow continuously in IL-4. Transfectants were grown in LCCM or IL-2-CM, respectively, rather than IL-4, to avoid selection for IL-4 responsiveness, which could complicate interpretation of results. Twelve neomycin-resistant colonies from each of the transfections with cDNA encoding the full-length or the truncated IL-13R␣ were individually analyzed for IL-13R␣ expression by fluorometry using anti-FLAG mAb and FITC conjugated to goat-anti-mouse antibody. Of the colonies of FD5 transfectants, 7 of 12 were positive for expression of the fulllength IL-13R␣, and 5 of 12 were positive for expression of the truncated IL-13R␣. Both forms of the IL-13R␣ were expressed as surface proteins, although in all five clones expressing IL-13R⌬CD, the transfected receptor was expressed at levels that were 10 -20-fold higher than those seen in all seven clones expressing the full-length receptor in FD5 (Fig. 1C). Likewise, in the case of CT4.S cells, analysis of four neomycin-resistant colonies from transfections of cDNA encoding the full-length or the truncated IL-13R␣ indicated that three of four and two of four, respectively, expressed the receptor. The three clones positive for the full-length IL-13R␣ expressed approximately 10-fold less receptor than the 2 clones expressing the truncated receptor (Fig. 1D). The same differences in expression level was observed when wild-type and truncated receptor proteins were immunoprecipitated from the transfected cells and immunoblotted with anti-FLAG mAb (results not shown). Differential expression of full-length and truncated IL-13R␣ might be due to higher copy number integration of pEB-13R⌬CD cDNA, higher stability of its mRNA, slower turnover of the translated protein, or other reasons. Investigation of this phenomenon is under way. Assays of responses to IL-4 and IL-13 indicated that clones within each set responded equivalently in terms of proliferation and survival. Therefore, only one clone of each transfected type was used for biochemical experiments. The responses of FD5 and CT4.S cells that were transfected with pPGK/Neo alone were indistinguishable from untransfected cells in terms of proliferation and survival in response to IL-4-CM, indicating that Neo expression plasmid did not interfere with IL-4 signaling. Importantly, endogenous murine IL-4R␣ was expressed at equivalent levels on the surface of parental or transfected FD5 (Fig. 1E) and CT4.S (Fig. 1F)  Fluorescence intensity (F.I.) was detected using a FACScan instrument (C). CT4.S (black thin line), CT4.S-13R (gray thick line), and CT4.S-13R⌬ cells (black thick line) were stained, and the receptor was detected as described in C (D). FD5, FD5-13R, and FD5-13R⌬ cells were incubated sequentially with anti-IL-4R␣ antibody and goat anti-rat Ig-FITC (white) or with the second antibody alone (shaded gray), and fluorescence intensity was detected as described in C (E). CT4.S.1, CT4.S-13R, and CT4.S-13R⌬ cells were stained as described in E, and fluorescence intensity was detected as described in C (F). and CT4.S cells grew continuously in IL-4 but failed to survive in IL-13. As demonstrated with representative clones in Fig.  2A, FD5-13R cells not only grew continuously in IL-13 as well as they did in IL-4, but they also exhibited the same flattened morphology observed when FD5 cells were cultured in IL-4. In contrast, cells of a representative clone expressing the truncated IL-13R␣ (FD5-13R⌬) failed to grow in IL-13, indicating that the cytoplasmic domain of IL-13R␣ was necessary for transduction of signals leading to cell growth. Moreover, cell counting experiments (Fig. 3A) demonstrated that clones expressing the full-length IL-13R␣ (FD5-13R) grew continuously in IL-13, at a similar rate to that observed in IL-4, whereas clones expressing the truncated IL-13R␣ (FD5-13R⌬) failed to respond to IL-13. Similar results were obtained using the T cell line CT4.S and clones derived by transfecting full-length or truncated IL-13R␣. As shown in Fig. 2B, clones of CT4.S-13R cells grew continuously in IL-13 as well as in IL-4, whereas cells of clones expressing the truncated IL-13R␣ (CT4.S-13R⌬) failed to grow or survive in response to IL-13. As expected, cell growth correlated with DNA synthesis. Fig. 3B shows that clones of CT4.S-13R cells incorporated [ 3 H]thymidine in response to IL-13 as well as to IL-4, but clones of CT4.S-13R⌬ showed no detectable response to IL-13.

Effect of Expression of Truncated IL-13R⌬ on Responses to IL-4 -
The demonstration that expression of the truncated IL-13R␣ did not permit biological responses to IL-13 prompted us to test whether cells expressing the truncated IL-13R␣ lacking cytoplasmic sequences responded normally to IL-4. We observed that clones of both FD5-13R⌬ and CT4.S-13R⌬ cells grew much more slowly in response to IL-4 than did the parental lines or clones of the respective transfectants expressing the full-length IL-13R␣. Clones of FD5-13R⌬ cultured in IL-4 not only failed to grow but also failed to exhibit the flattened morphology that is characteristic of the parental cells grown in IL-4 or of clones of FD5-13R grown in IL-4 or IL-13 ( Fig. 2A). Seven-day cultures of a representative clone of FD5-13R⌬ stimulated with IL-4 contained 10-fold fewer cells than parallel cultures of IL-4-stimulated cells expressing the full-length IL-13R␣ (Fig. 3A). Similarly, the number of cells of a representative clone of CT4.S-13R⌬ cells generated in the presence of IL-4 were much lower than those in parallel cultures of parental CT4.S cells (Fig. 2B). Furthermore, the level of [ 3 H]thymidine incorporation induced by IL-4 in CT4.S-13R⌬ cells was 50 and 60% lower than the incorporation induced by the same stimulation in CT4.S and CT4.S-13R cells, respectively (Fig. 3B). This negative effect on growth was specific for IL-4, since FD5, FD5-13R, and FD5-13R⌬ grew equally well in GM-CSF or CSF-1, and CT4.S, CT4.S-13R, and CT4.S-13R⌬ grew equally well in IL-2 (Figs. 2 and 3).
We next determined whether the presence of IL-13 influenced the suppression of IL-4 responses seen in cells expressing the truncated IL-13R⌬. As shown in Fig. 3A, the presence of IL-13 suppressed the IL-4-dependent growth by a further 10 -100-fold to background levels. This effect of IL-13 on IL-4induced growth was specific for cells expressing truncated IL-13R⌬, since IL-13 had no inhibitory effect on the response to IL-4 on FD5-13R or parental FD5 cells.
Tyrosine Phosphorylation of IL-4R␣, IRS-2, and STAT6 -To probe the biochemical basis of these effects on growth and morphology, we examined tyrosine phosphorylation of IL-4R␣, IRS-2, and STAT6 in response to IL-13 and IL-4 in cells where the full-length or the truncated IL-13R␣ was exogenuously introduced. Fig. 4A shows that both IL-4 and IL-13 induced tyrosine phosphorylation of IL-4R␣ in FD5-13R, while in the parental cell line only IL-4 was effective. In FD5-13R⌬ cells, IL-13 failed to induce tyrosine phosphorylation of IL-4R␣. Moreover, there was no detectable phosphorylation of IL-4R␣ in response to IL-4. Very similar results were obtained in analysis of a second signal transduction component phosphorylated in response to IL-4 and IL-13, IRS-2. Fig. 4B shows that IL-13 induced significant phosphorylation of IRS-2 only in FD5-13R cells. As expected, IL-4 induced phosphorylation of IRS-2 in both FD5 and FD5-13R cells. However, once again, the response of FD5-13R⌬ cells to IL-4 was dramatically reduced, with little if any tyrosine phosphorylation of IRS-2. This effect was specific, as shown by the equivalent response to insulin, which is a potent activator of IRS-2 phosphorylation and was included in the experiment as an internal reference for each cell population. Equivalent results were obtained with a third component of the IL-4 and IL-13 signaling pathway, STAT6. Fig. 4C shows that IL-13 induced phosphorylation of STAT6 only in FD5 cells expressing the full-length IL-13R␣ (FD5-13R). Likewise, whereas IL-4 induced phosphorylation of STAT6 in FD5 and FD5-13R cells, the level of phosphorylation of STAT6 in FD5-13R⌬ was almost undetectable.
Phosphorylation of JAK Family Kinases-We observed previously that in cells of hematopoietic origin IL-4 induce tyrosine phosphorylation of only JAK1 and JAK3 kinases and in some cells also Tyk2, whereas IL-13 induces tyrosine phosphorylation of JAK1 and Tyk2 (20). Both IL-4 and IL-13 induce tyrosine phosphorylation of Jak2 in solid carcinoma cells and vascular endothelial cells (48 -50). Therefore, we investigated the ability of IL-4 and IL-13 to induce tyrosine phosphorylation of JAK family kinases in myeloid cells expressing the fulllength or truncated IL-13R␣. As expected, in FD5-13R cells, IL-13 induced tyrosine phosphorylation of JAK1 (Fig. 5A) and Tyk2 (Fig. 5D) but failed to induce tyrosine phosphorylation of JAK3 (Fig. 5C), although these cells expressed both IL-13R␣ and ␥c chain. In FD5-13R cells, IL-4 induced tyrosine phosphorylation of JAK1 to a level that was higher than that observed in parental FD5 cells (Fig. 5A), maintained the inducible level of phosphorylation of JAK3 (Fig. 5C), and gained the ability to induce phosphorylation of Tyk2 (Fig. 5D). In agreement with the lack of induced tyrosine phosphorylation of IL-4R␣, IRS-2, and STAT6 observed in FD5-13R⌬ cells (Fig. 4), both IL-4 and IL-13 failed to induce detectable phosphorylation of JAK1 (Fig. 5A), JAK3 (Fig. 5C), or Tyk2 (Fig. 5D) in these cells.
Neither IL-4 nor IL-13 induced significant phosphorylation of JAK2 in any of the cells, although JAK2 activation pathways were intact as GM-CSF induced tyrosine phosphorylation of JAK2 (Fig. 5B). The findings that IL-4-induced phosphorylation of JAK3 was equivalent in FD5 and FD5-13R cells (Fig.  5C) but IL-4-induced phosphorylation of JAK1 was higher in FD5-13R cells than in parental FD5 cells (Fig. 5A) and that both IL-4 and IL-13 induced comparable levels of phosphorylation of JAK1 in FD5-13R cells (Fig. 5A) indicate that the ␥c is the limiting binding protein in the complex and that the IL-13R is expressed at a higher concentration than IL-4R␣. DISCUSSION In this study, we have demonstrated that expression of the IL-13R␣ in premyeloid and T cells confers the ability to respond to IL-13 with continuous growth, characteristic morphological changes, and tyrosine phosphorylation of JAK1, Tyk2, STAT6, IRS-2, and IL-4R␣. The cytoplasmic domain of IL-13R␣ was necessary for all of these biological and biochemical responses to IL-13. Strikingly, cell surface expression of a truncated IL-13R␣ that included the complete extracellular and transmembrane domains but lacked the cytoplasmic region exerted a profound inhibitory effect on biological and biochemical responses to IL-4. The two parental cell lines used in our studies were of different origin, namely premyeloid FD5 and T lymphoid CT4.S; both grew continuously in IL-4 but not in IL-13, indicating that these cells express functional IL-4R␣ and ␥c chain but not an IL-13R␣. Expression of the full-length IL-13R␣ was sufficient to confer responsiveness to IL-13 by all criteria tested. These included the ability to grow continuously in IL-13 (Figs. 2 and 3) and the ability to respond to IL-13 with the same morphological changes induced by IL-4. The induction of a flattened morphology and adherence by IL-4 has been previously reported in the IL-2-dependent line CTLL-2, when transfected with a chimeric IL-4R␣, including an essential small region of its cytoplasmic domain (51). The biological responses of cells transfected with the full-length IL-13R␣ correlated with the ability of IL-13 to induce tyrosine phosphorylation of Tyk2, JAK1, IL-4R␣, IRS-2, and STAT6, all events that have been reported to follow stimulation of other cells with IL-13 (19,20,22,27). Cells transfected with a truncated IL-13R␣ lacking the cytoplasmic domain showed no detectable biological or biochemical responses to IL-13, although they responded normally to unrelated cytokines such as GM-CSF or CSF-1 in the case of FD5-13R⌬ or IL-2 in the case of CT4.S-13R⌬ (Figs. 2 and 3). The importance of the cytoplasmic domain of IL-13R␣ is consistent with that attributed to the cytoplasmic domain of IL-5R␣ (52)(53)(54) or more distantly related IL-3R␣ 3 or GM-CSF-R␣ (55,56). Deletion of the cytoplasmic domains of IL-5R␣, IL-3R␣, GM-CSF-R␣, and also Epo-R (57-60) and IL-2R (61)(62)(63)(64)(65)(66) abolish responsiveness to the respective cytokine. Our results are consistent with the notion that one important function of the cytoplasmic domain of the IL-13R␣ is recruitment of Tyk2 into the signaling complex. Thus, cells expressing the fulllength IL-13R␣ gained the ability to respond not only to IL-13 but also to induce tyrosine phosphorylation of Tyk2 in response to IL-4 (Fig. 5D). This observation fits with the notion that both IL-4 and IL-13 can form a complex with the IL-13R␣ and the IL-4R␣ and that the IL-13R␣ contributes the capacity to induce tyrosine phosphorylation of Tyk2. Our results showed that (i) IL-4-induced phosphorylation of JAK1 was higher in FD5-13R cells than in parental FD5 cells (Fig. 5A); (ii) IL-4-induced phosphorylation of JAK3 was equivalent in FD5 and FD5-13R cells (Fig. 5C); and (iii) both IL-4 and IL-13 induced comparable phosphorylation of JAK1 in FD5-13R cells (Fig. 5A). Together, the results indicate that in FD5 cells the ␥c chain is the limiting protein in determining the amount of functional IL-4 receptor and that IL-13R␣ is expressed at higher levels than IL-4R␣ in FD5-13R cells.
Expression of the truncated IL-13R␣ in clones of either FD5-13R⌬ or CT4.S-13R⌬ resulted in striking ablation of proliferative responses to IL-4. IL-4-induced proliferation of CT4.S-13R⌬ cells, as determined by [ 3 H]thymidine incorporation after 5 days stimulation (Fig. 3B), was decreased by 50%, as compared with the parental cell line. With longer periods in culture, the decreased rate of IL-4-stimulated growth of CT4.S-13R⌬ clones was even more evident (Fig. 2B). Unfortunately, we were unable to correlate IL-4-stimulated growth of CT4.S cells with tyrosine phosphorylation of specific substrates due to low detectability of signaling proteins in these cells. However, in FD5 and its derivative cell clones, both biological and biochemical responses to IL-4 could be investigated. Clones of FD5-13R⌬ cells exhibited profound inhibition (90%) of IL-4stimulated growth over 7 or 12 days in culture ( Figs. 2A and  3A). This negative effect of expression of truncated IL-13R␣ on IL-4-induced growth of FD5-13R⌬ corresponded to the lack of detectable IL-4-induced tyrosine phosphorylation of JAK1, JAK3, IL-4R␣, IRS-2, and STAT6 (Figs. 4 and 5). The observation that FD5-13R⌬ grew, albeit slowly in IL-4, despite the lack of detectable IL-4-induced tyrosine phosphorylation events probably reflects the fact that occupancy of only a low number of receptors is required for growth and that this level of receptor occupancy does not generate detectable levels of tyrosine phosphorylation.
The failure to detect tyrosine phosphorylation of JAK3 in FD5-13R⌬ in response to IL-4 ( Fig. 5C) suggests that the presence of the extracellular domain of the truncated IL-13R␣ competitively inhibited the IL-4-induced assembly of complexes of IL-4R␣ together with ␥c and its associated JAK3. This would occur if the extracellular domain of the IL-13R␣ competed with ␥c for the complex of IL-4 and IL-4R␣. There is evidence that the affinity of binding of IL-4 in the IL-4R␣⅐␥c complex is greater than that of IL-4 in the IL-4R␣⅐IL-13R complex (9,12,13), although that might not be the case in all cells (16); however, the truncated IL-13R␣ is expressed on the surface of the transfected cells at relatively high levels and thus is likely to be effectively competing with ␥c and sequestrating complexes of IL-4⅐IL-4R␣ into sterile complexes of IL-4⅐IL-4R␣⅐IL-13R⌬.
The failure of cells expressing truncated IL-13R␣ to respond to either IL-13 or IL-4 with tyrosine phosphorylation of Tyk2 or JAK1 (Fig. 5) is consistent with the notion that the cytoplasmic domain of IL-13R␣ is necessary to provide a docking site for 3 P. Orban, M. Levings, and J. W. Schrader, manuscript in preparation.
FIG. 5. Tyrosine phosphorylation of JAK family kinases. FD5, FD5-13R, and FD5-13R⌬ cells were stimulated with IL-4, IL-13, or GM-CSF (G) or left untreated as control (C). Cells were lysed, immunoprecipitated with specific antibodies, immunoblotted with anti-Tyr(P) mAb, and stripped and reprobed with the specific immunoprecipitating antibody. A, lysates of the equivalent of 3 ϫ 10 7 cells were lysed in lysis buffer and diluted to 0.1% Nonidet P-40. Lysates were immunoprecipitated with anti-JAK1 Ab and immunoblotted with anti-Tyr(P) mAb. The blot was stripped and reprobed with anti-JAK1 Ab. B, lysates of the equivalent of 1 ϫ 10 7 cells were immunoprecipitated with anti-JAK2 Ab and immunoblotted with anti-Tyr(P) mAb. The blot was stripped and reprobed with anti-JAK2 Ab. C, lysates of the equivalent of 1 ϫ 10 7 cells were immunoprecipitated with anti-JAK3 Ab and immunoblotted with anti-Tyr(P) mAb. The blot was stripped and reprobed with anti-JAK3 Ab. D, lysates of the equivalent of 2 ϫ 10 7 cells were immunoprecipitated with anti-Tyk2 Ab and immunoblotted with anti-Tyr(P) mAb. The blot was stripped and reprobed with anti-Tyk2 Ab.
Tyk2 and for subsequent trans-phosphorylation and activation of JAK1. Failure to activate the JAK kinases is likely to account for the absence of IL-13-induced phosphorylation of IL-4R␣, IRS-2, and STAT6 and failure of IL-13 to stimulate proliferation and survival of cells. The inhibitory effect of expression of IL-13R⌬ on responses to IL-4 that we observed also argues against the notion that homodimerization of IL-4R␣ is sufficient for IL-4 signaling (67,68).
The striking dominant inhibitory effect of expression of the truncated IL-13R␣ in response to IL-4 is thus explained by competition with ␥c for complexes of IL-4⅐IL-4R␣, which it sequestrates into sterile complexes containing IL-4, IL-4R␣, and the truncated IL-13R␣. The even greater suppression of IL-4 responses seen in cells expressing the truncated IL-13R␣ in the presence of a combination of IL-4 and IL-13 is explained by the affinity of the complex of IL-13 and truncated IL-13R␣ for the IL-4R␣. The membrane-bound complex of truncated IL-13R␣ and IL-13 effectively outcompetes soluble IL-4 for binding to the IL-4R␣, thus sequestrating all the IL-4R␣ into sterile complexes of IL-13 and truncated IL-13R␣. Taken together, our observations on the responses to either IL-4 or a combination of IL-4 plus IL-13 of cells expressing the truncated IL-13R␣ strongly support a model where the IL-4R␣ and the IL-13R␣, associated respectively with JAK1 and Tyk2 kinases, are components of both the IL-13 receptor and one class of IL-4 receptor. In the case of IL-4 signaling, the IL-4R␣ binds IL-4, and the complex recruits either the IL-13R␣ or the ␥c, which brings, respectively, Tyk2 or JAK3 to join the JAK1 associated with IL-4R␣. In the case of IL-13 signaling, the IL-13R␣ binds IL-13, and the complex recruits the IL-4R␣, which brings JAK1 to join Tyk2 in the active receptor.