IL-4 and IL-13 Induce SOCS-1 Gene Expression in A549 Cells by Three Functional STAT6-Binding Motifs Located Upstream of the Transcription Initiation Site ¹ Free

Allergies have become a major health problem in the past decades, and the incidence of allergic asthma alone has nearly doubled since 1980 in industrialized countries. This substantial increase can be described as an epidemic of dysregulated immunity, and the regulation of immune functions in allergy is consequently an urgent matter (1). The cytokines IL-4 and IL-13 are crucial regulators during allergic sensitization. IL-4 induces differentiation of allergen-specific Th2 cells, and both IL-4 and IL-13 direct class switching toward IgE production (2, 3). Several aspects of ongoing allergic and asthmatic responses to allergens are promoted by IL-4 and IL-13 as well, including activation of eosinophils, basophils, and mast cells; eotaxin synthesis by fibroblasts; and differentiation and hyperreactivity of mucus-producing goblet cells (1, 4). Central to the response of cells to both IL-4 and IL-13 is the activation of cytoplasmic Janus kinases and the subsequent phosphorylation of a number of signaling molecules, including STAT6 (2). STAT6 binds via a single Src homology 2 (SH2) ³ domain first to tyrosine-phosphorylated motifs in the IL-4Rα chain and then to another STAT6 molecule, which results in the formation of active dimers. The dimerized STAT6 complexes translocate to the nucleus where they bind preferentially to IFN-γ-activated sequence (GAS) motifs with the consensus sequence TTC(N₄)GAA (5). The spacing of the symmetrical GAA half-sites which are 4 bp apart (also called an N4 motif) distinguishes STAT6 sites from binding sites of other STAT proteins, where the GAA half-sites are usally 3 bp apart (6). STAT6 is critical in the activation or enhanced expression of many IL-4/IL-13-responsive genes, including those for class II MHC molecules, CD23, IL-4Rα, germline Igε transcripts and, in the mouse, germline Igγ1 transcripts, as well as eotaxin-1, eotaxin-3, found in inflammatory zone-1, and found in inflammatory zone-2 (7, 8, 9, 10, 11, 12, 13, 14, 15).

Although much is known about IL-4-mediated activation of the Janus kinase-STAT signaling pathway, inhibition of this pathway is less well understood. Recent studies suggest that cytokine signaling can be negatively regulated by one or more of three different families of proteins, 1) protein tyrosine phosphatases such as Src homology protein-1 and CD45, 2) the protein inhibitors of activated STATs, and 3) the suppressors of cytokine signaling (SOCS) proteins. A common feature of the SOCS family is that all members contain both an SH2 domain and a 40-residue C-terminal motif referred to as the SOCS box (16, 17, 18). This element has by now been found in >40 proteins from 9 families, where its general function appears to be binding of E3 ubiquitin protein ligases, which ultimately leads to proteasomal destruction of associated proteins (19). Within the SOCS family, SOCS-4 to -7 have been described based on sequence homologies (20), and their functions are less well understood. Cytokine-inducible SH2 protein (CIS), SOCS-1, SOCS-2, and SOCS-3 have been characterized more thoroughly. The mRNAs coding for these proteins are found at low abundancies in resting cells, but transcription is strongly enhanced after stimulation with a wide range of different cytokines both in vitro and in vivo (21, 22, 23, 24). The patterns of SOCS mRNA induction seem to vary depending on which cell type or tissue is being treated with the cytokines.

It has been further observed that the STAT family of transcription factors has a prominent role in the up-regulation of the CIS, SOCS-1, and SOCS-3 genes. The existence of STAT-binding motifs in the promoters was noticed for all of these genes (18, 25, 26, 27, 28). STAT5 is required for erythropoietin- and IL-3-dependent expression of CIS (25), and STAT3 mediates LIF- and IL-6-dependent expression of SOCS-1 (18). Interestingly, such a regulation can be indirect; IFN-γ-activated STAT1 does not regulate SOCS-1 via GAS motifs in the SOCS-1 promoter but rather via STAT1 induced up-regulation of IFN regulatory factor-1, which in turn stimulates transcription from the SOCS-1 locus (26). The SOCS-3 promoter binds STAT5B in response to growth hormone or insulin (29, 30), and an intact STAT1/STAT3 motif is necessary for LIF-induced expression of SOCS-3 (31).

SOCS-1 expression is induced by many stimuli (21, 22, 23, 24), but it is supposed to be particularly important for limiting IFN-γ-dependent T cell activation, and indeed SOCS-1^−/− mice die of a complex inflammation that depends on IFN-γ signals (32). However, SOCS-1^−/−/IFN-γ^−/− double-knockout mice also show signs of dysregulated immunity, which underlines that regulation of other cytokines by SOCS-1 is critical as well (33). In the present study, we investigated the regulation of SOCS-1 expression by IL-4 and IL-13, mediated by the shared signaling protein STAT6. Consensus elements for STAT6 binding have been noticed in the murine SOCS-1 promoter (18, 27), but thus far no proof for binding of STAT6 to any of the SOCS gene promoters and no STAT6-dependent regulation of SOCS gene expression have been shown. The present work presents evidence that SOCS-1 expression is up-regulated in A549 human lung adenocarcinoma cells by IL-4 and IL-13. The stimulatory effect was mediated by binding of STAT6 to three adjacent GAS motifs ∼600 bp upstream of the transcription initiation site. It is further shown that SOCS-1 can autoregulate its expression via a negative feedback loop.

The epithelial human lung carcinoma line A549 and the human T cell leukemia line Jurkat were cultured in RPMI 1640 plus 10% FCS, 100 U/ml penicillin and 100 μg/ml streptomycin at 37°C in a humidified atmosphere containing 5% CO₂.

Recombinant human IL-4 and IL-13 were obtained from R&D Systems (Minneapolis, MN).

Cells were pelleted and resuspended in ice-cold lysis buffer (25 mM Tris-HCl (pH 7.5), 140 mM NaCl, 1 mM EDTA, 10 mM Na₂P₂O₇, 0.5% Nonidet P-40). After end-over-end rotation for 45 min, the cell lysates were centrifuged at 10,000 × g for 20 min. Cleared supernatants were incubated for 2 h or overnight at 4°C with primary Ab, and the immune complexes were collected on protein A/G-agarose (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h. Precipitates were washed twice with lysis buffer and twice with salt buffer (0.5 M LiCl in 100 mM Tris-HCl, pH 8.0) and boiled for 5 min in SDS sample buffer (50 mM Tris-HCl (pH 6.8), 2% SDS, 5% 2-ME, 10% glycerol). The samples were subjected to SDS-PAGE on 12% or gradient gels (4–20%), followed by Western blot analysis with the secondary Ab and by detection using enhanced chemiluminescence (Pierce, Rockford, IL).

For immunoprecipitation anti-SOCS-1 Ab (Santa Cruz Biotechnology) was used. For Western blot detection anti-SOCS-1 (Santa Cruz Biotechnology), Ig was used as primary Ab. HRP-coupled anti-rabbit-IgG (Bio-Rad, Hercules, CA) was used as secondary Ab.

Total RNA from A549 cells was isolated using the TRIzol reagent (Life Technologies, Gaithersburg, MD). Using SuperScript II (RNase H⁻) reverse transcriptase (Life Technologies), 4 μg of total RNA was reverse transcribed according to the manufacturer’s instructions. Real time PCR analysis was done on a Rotorgene 2000 (Corbett Research, Sydney, Australia) using iQ SYBR Green Supermix (Bio-Rad). Primer sequences used for real time analysis of SOCS-1 expression were 5′-TTGGAGGGAGCGGATGGGTGTAG-3′ (forward primer) and 5′-AGAGGTAGGAGGTGCGAGTTCAGGTC-3′ (reverse primer) for SOCS-1 and 5′-GGCACCATTGAAATCCTGAGTGATGTG-3′ (forward primer) and 5′-TTGCGGACACCCTCCAGGAAG C-3′ for large ribosomal protein P0, which was used as reference gene. The specificity of the PCR was checked by recording a melting curve and by sequencing the amplicons using an ABI-prism automated sequencing machine (Applied Biosystems, Foster City, CA). Induction ratios (x) were calculated using the formula x = 2^−ΔΔCt, where Ct represents the mean threshold cycle of all replicate analyses of a given gene and ΔCt represents the difference between the Ct values of the gene in question (SOCS-1) and the Ct value of the reference gene (large ribosomal protein P0). ΔΔCt is the difference between the ΔCt values of the samples induced with IL-4 or IL-13 and the ΔCt of the uninduced sample.

A 1433-bp DNA fragment containing the human SOCS-1 promoter was amplified from genomic DNA (Roche, Basel, Switzerland) using the primers 5′-CCCAAGCTTGCGCAGGGAGGGCAGTCGA-3′ and 5′-AGTCTCTAGAGGGGCCAGCCGGAGGGGTG-3′ which contained restriction sites for HindIII and BglII. The fragment was gel excised using the Concert kit (Life Technologies). The purified PCR product was digested with HindIII and BglII and was cloned into the HindIII- and BamHI (generates same overhang as BglII)-digested pLUC vector generating the construct WT. Mutations were introduced into the three STAT6-binding motifs contained within the promoter sequence of SOCS-1. The motifs were mutated from TTC(N)₄GAA to TAT(N)₄GAA (13). Three single mutants, MA, MB, MC, three double mutants, MAB, MAC, MBC, and one triple mutant, MABC, were thus constructed. Mutagenesis was conducted using the Quikchange kit (Stratagene, La Jolla, CA). Mutated inserts were cloned back into the original vector to exclude disruption of the luciferase gene by mispriming of the primers used for mutagenesis. The deletion mutant ΔGAS was generated by cutting WT with HindIII, followed by filling up the 5′-overhang with Klenow fragment (Promega, Milwaukee, WI). Then the linearized vector was digested with XhoI, and the resulting fragments were gel excised. The smaller fragment, representing the SOCS-1 promoter sequence, was cut at the natural SmaI site 685 bp upstream of the translation initiation site. This cut off the upstream part of the promoter, which contains the three GAS motifs. Thus the longer fragment was gel excised and ligated back into the linearized WT construct.

All plasmids were analyzed by digestion with restriction enzymes and DNA sequencing. Plasmids for transfection were purified with Maxi-prep kits (Qiagen, Chatsworth, CA).

A DNA fragment of the coding sequence of SOCS-1 was amplified by PCR using cDNA from IL-4-treated Jurkat cells. The fragment was gel excised using the Concert kit (Life Technologies). The purified PCR product was digested with NheI and XhoI and cloned into the pcDNA3.1 vector (Invitrogen, Carlsbad, CA). The plasmid was analyzed by digestion with restriction enzymes and DNA sequencing. Plasmids for transfection were purified with Maxi-prep kits.

The day before transfection, 5 × 10⁵ cells were seeded into 24-well culture plates in fresh medium. Transient transfection of A549 cells was conducted using Effectene Transfection Reagent (Qiagen) according to the manufacturer’s protocol. After 24 h, cells were washed and cultured overnight in the presence or absence of 50 ng/ml IL-4 or 12 ng/ml IL-13 before luciferase assays were conducted in triplicates according to the manufacturer’s instructions using the Promega Luciferase Assay System. Luciferase activity was normalized to the levels of the corresponding uninduced samples.

Nuclear extracts from unstimulated A549 cells or from cells that had been stimulated for 30 min with 50 ng/ml IL-4 or 12 ng/ml IL-13 were prepared according to the method described by Andrews and Faller (34). One double-stranded oligonucleotide probe containing the STAT6-binding site in the eotaxin-3 promoter between positions −86 and −45 (14), one probe containing the first GAS motif in the SOCS-1 promoter between positions −704 and −660 (5′-GGC CCC GCC CAG TTT CCG AGG AAC TGG GCC GGG GTG GAG GCG C-3′ was used as the forward oligo, and 5′-GAG GCG CCT CCA CCC CGG CCC AGT TCC TCG GAA ACT GGG CGG G-3′ was used as the reverse oligo), one probe containing the second GAS motif in the SOCS-1 promoter between positions −646 and −616 (5′-GCG CCG CCA GGG GTT CCT CTG AAG CCT GTG GTC AGG CCG CCG C-3′ was used as the forward oligo, and 5′-GAG GCG GCG GCC TGA CCA CAG GCT TCA GAG GAA CCC CTG GCG G-3′ was used as the reverse oligo), and one probe containing the third GAS motif in the SOCS-1 promoter between positions −616 and −589 (5′-GTG GTC AGG CCG CCG CTT CCC GGG AAG CCC GAG CCA AGA CCA G-3′ was used as the forward oligo and 5′-GAG CTG GTC TTG GCT CGG GCT TCC CGG GAA GCG GCG GCC TGA C-3′ was used as the reverse oligo), were end-labeled using [³²P]dCTP (Amersham, Arlington Heights, IL) and Klenow polymerase (Roche Molecular Biochemicals). The nucleoprotein binding reaction was performed as described using 5 μg of nuclear extracts. For oligonucleotide competition assays, a 50-fold molar excess of cold oligonucleotide was added to the binding reaction 30 min before the radiolabeled probe. For supershift experiments, extracts were preincubated with 2 μg of Ab for 20 min before the radiolabeled probe was added. All Abs used in supershift experiments were from Santa Cruz Biotechnology.

SOCS-1 is expressed in a number of cell types on IL-4 stimulation, and we recently identified it as having the highest score in a screen of primary human T cells for IL-4-induced genes using DNA array technology (D. Hebenstreit et al., unpublished observations). To test whether SOCS-1 expression could be induced in the human lung cell line A549, cells were stimulated with IL-4 and IL-13 for different times. SOCS-1 mRNA and protein levels were quantified by real time-PCR and Western blot analysis, respectively (Fig. 1). Both cytokines stimulated SOCS-1 expression with similar kinetics. SOCS-1 mRNA levels peaked after 1 h with induction at 5- to 7-fold increased levels in comparison with uninduced levels. After a longer induction time, SOCS-1 mRNA and protein expression stayed at reduced but still elevated levels.

The existence of GAS motifs with the consensus TTC(N)₄GAA within the murine SOCS-1 promoter was noticed previously, but no functional assessments have been conducted (18, 26, 27). The putative promoter sequence of human SOCS-1 (27, 35) contains three potential GAS motifs of the sequence TTC(N)₄GAA, located at positions −691 to −681 (motif A), −646 to −636 (motif B), and −616 to −606 (motif C) relative to the transcriptional start site. To test whether these motifs could interact with STAT6, EMSA experiments were conducted. Nuclear extracts prepared from cytokine-stimulated cells were incubated with radiolabeled probes containing the putative STAT6-binding sites (probes A, B, and C). IL-4 and IL-13 induced the formation of a nucleoprotein complex with all three probes (Fig. 2,A), where IL-13 seemed to be slightly more effective than IL-4. To test the specificity of the binding reactions, cold competitors were added in 50 molar excess. As Fig. 2 A shows, addition of the unlabeled original probes resulted in a loss of factor binding to the radiolabeled oligos. Additionally, we added mutated cold oligonucleotide probes MA, MB, and MC as competitors. The sequence of these probes differs from the wild-type probes by two nucleotide substitutions in the inverted repeat region of the STAT6-binding site, which has been shown to abrogate STAT6 binding (13). These oligos were unable to block binding to the radiolabeled probes. Finally, oligos corresponding to the known STAT6-binding site of the human eotaxin-3 promoter (14) were used as a positive control. These could block binding of the radiolabeled probes. Generally, motifs A and C displayed strong affinity toward the nucleoprotein, whereas motif B was bound very weakly.

Another EMSA was conducted to directly compare binding of wild-type and mutant oligos. This time either radiolabeled wild-type probes or radiolabeled mutant probes were added to the nuclear extracts (Fig. 2,B). The mutant probes were unable to interact with the IL-4-IL-13-induced factor. Thus, interaction of STAT6 with the GAS motifs A to C was dependent on intact canonical STAT6-binding motifs. Preincubation of extracts with Abs directed against STAT6 resulted in disappearance of the nucleoprotein complex (Fig. 2 C) while faint supershifted bands became visible. No bandshifts were observed with Abs against the NF-κB subunit p52. These data collectively demonstrate that IL-4 and IL-13 can stimulate interaction of STAT6 with each of the three GAS motifs in the SOCS-1 promoter.

The relevance of the three STAT6 binding sites for IL-4- and IL-13-induced SOCS-1 gene expression was assessed in transient transfection experiments using SOCS-1 reporter gene constructs. A 1.4-kb DNA fragment from position −743 to +690 relative to the transcriptional start site of SOCS-1 was cloned 5′ of the firefly luciferase reporter gene. This fragment contained all three GAS motifs with an additional 52 bp upstream of motif A, as well as the beginning of the second exon of the SOCS-1 gene (Fig. 3). A549 cells were transiently transfected with this construct and stimulated with IL-4 or IL-13 overnight. Both cytokines led to an ∼5- to 7-fold stronger luciferase activity than that of unstimulated cells (Fig. 4,A). To confirm that the IL-4-IL-13-inducibility was mediated by the STAT6 sites, the same nucleotide substitutions as used for the mutant EMSA probes were introduced by site-directed mutagenesis. We generated three single mutants, each of which had one of the three GAS motifs mutagenized (MA, MB, and MC), three double mutants, which featured two mutagenized motifs at one time (MAB, MAC, and MBC), and one triple mutant with all three motifs changed (Fig. 3). Apart from the wild-type construct, seven additional plasmids were thus obtained. When cells were transfected with the single mutants, inducibility by IL-4 and IL-13 was diminished in all cases compared with the WT construct (Fig. 4,A). This demonstrates that all three GAS motifs are functional and add to the total inducibility of the WT construct. Consistently, luciferase activity of MA was the lowest, with induction rates ranging only slightly above the uninduced ones. Transfections of the double mutants showed that the overall induction rates were further decreased compared with the single mutants. Again, double mutants with a mutated motif A were the ones with the lowest luciferase activity. Finally, the triple mutant was not inducible, either by IL-4 or by IL-13. These data demonstrate that all three STAT6-binding motifs are functional and are necessary for the full wild-type expression of SOCS-1. Motif A appeared to be most important in mediating IL-4-IL-13-induced SOCS-1 transcription, whereas motif B seemed to be the one with the least pronounced effect. This agrees well with the data obtained in EMSA studies (Fig. 2), in which motif A formed the strongest band of a nucleoprotein complex, followed by motifs C and B.

In some experiments, the background luciferase activity in cells transfected with mutant SOCS-1 promoter constructs was higher than in cells transfected with the wild-type construct (Fig. 4,B). This fits the observation that the radiolabeled probes bound to an unknown nucleoprotein complex (Fig. 2, A (top) and B) in some EMSA experiments, regardless whether wild-type or mutated probes were used. The unknown complex migrated at a lower molecular mass than STAT6, was not bound by anti-STAT6 Ab (data not shown), and was not dependent on IL-4 or IL-13 stimulation. Because the insertion of mutations into the GAS motifs seemed to some extent to facilitate binding of the unknown complex (Fig. 2,B, particularly probes C/MC) this might explain the higher constitutive luciferase activity seen in the reporter gene assays with mutated constructs. This subject was further investigated by constructing a deletion mutant of the SOCS-1 promoter. This construct, designated ΔGAS, features a deletion of 132 bp from the 5′ part of the SOCS-1 wild-type promoter construct, thereby removing all three GAS motifs (Fig. 3,A). Like the MABC construct, the ΔGAS construct was unable to respond to IL-4-IL-13 stimulation (Fig. 5,A). Furthermore, when considering the unnormalized luciferase activity, the higher background activity seen with some of the mutants was gone for ΔGAS (Fig. 5 B). Altogether, these data suggest that the mutations that we inserted into the three GAS motifs on the putative SOCS-1 promoter facilitate binding of an unknown complex to the motifs which seems to lead to increased background activity of the promoter. Nonetheless, this experimental observation seems to have no implications regarding STAT6 dependence of IL-4-IL-13-stimulated expression of SOCS-1.

It is widely assumed that SOCS proteins can act as their own inhibitors via a negative feedback loop (21, 22, 23, 24). The finding that SOCS-1 is a potent inhibitor of IL-4 signal transduction (36, 37) leads to the question of whether overexpression of SOCS-1 can reduce IL-4-IL-13-induced transcription from the SOCS-1 promoter. To address this question, we constructed an expression vector of SOCS-1 by cloning its coding sequence under the control of a CMV promoter. We transiently transfected A549 cells with the wild-type SOCS-1 promoter luciferase construct and cotransfected the SOCS-1 expression vector or an empty vector as control. Cells were then stimulated with IL-4 or IL-13 or were left unstimulated. Cotransfection of the expression vector reduced both IL-4- and IL-13-induced luciferase activity to background levels, when compared with the empty vector control (Fig. 5,A). When the triple-mutant promoter construct was transfected, IL-4-IL-13 inducibility of the luciferase was gone (Fig. 5 B).

To confirm these results, A549 cells were transiently transfected with the SOCS-1 expression vector or with the empty vector as control and were treated with IL-4-IL-13 or were left untreated. Real time PCR was used to quantify SOCS-1 mRNA levels in total RNA preparations. The primers used for the PCR were designed to amplify part of the 3′-untranslated region of the SOCS-1 transcript. Because the 3′-untranslated region was not present in the expression vector, only endogenous SOCS-1 mRNA was detected. As shown in Fig. 5 C, mock-transfected cells expressed ∼3-fold higher quantities of endogenous SOCS-1 mRNA than SOCS-1-transfected cells.

This study represents the first characterization of elements necessary for activation of the SOCS-1 promoter by Th2 cytokines. In the murine SOCS-1 promoter, serial promoter deletions have indicated localization and functionality of signal peptide-1, IFN-stimulated regulatory elements, and GAS elements in the 5′-flanking region of the murine SOCS-1 gene (27). Furthermore, the same study provided evidence for a second, 5′-located exon, separated by a 509-bp intron from exon 2. Exon 1 and part of exon 2 contain an open reading frame of 115 nucleotides, ending 1 nucleotide upstream of the major open reading frame. This upstream open reading frame was shown to repress the translation of the downstream major open reading frame (27). The human SOCS-1 locus also produces a bicistronic transcript in which two exons are separated by a 550-bp intron (27, 35). As with murine SOCS-1 (38), the human gene, too, contains an upstream open reading frame that exerts translational repression. It has been shown that IFN-γ induced up-regulation of the murine SOCS-1 gene is mediated by STAT1 (26). The region mediating stimulatory effects of IFN-γ on transcription was localized to the −88/−60 region containing three tandem GAAA units, named variant IFN-γ-responsive element, whereas four GAS sites located further upstream were not related to the IFN-γ response (26).

Here, we provide evidence that STAT6 binds to three GAS motifs in the human SOCS-1 promoter, which are located −691 to −606 bp upstream of the transcriptional start site. Intact GAS motifs are necessary for IL-4- and IL-13-induced up-regulation of SOCS-1 expression in A549 cells. Furthermore, SOCS-1 is able to repress its own expression. These conclusions are based on the following data. Real time PCR studies and Western blot experiments showed that SOCS-1 expression can be induced in A549 cells on treatment with IL-4 and IL-13. The experiments show that shortly after stimulation, mRNA levels of SOCS-1 reach a peak followed by stabilization at an elevated expression level. These expression kinetics are similar to those published for other cell types and other members of the SOCS family for a variety of different stimuli (21). In EMSAs, binding of STAT6 to all three of the GAS motifs was shown. Disruption of these sites led to a loss of IL-4/IL-13 inducibility of SOCS-1 promoter reporter gene constructs. Each individual GAS site was found to contribute to the total induction of SOCS-1 expression. The close proximity of the three motifs to each other (A to B, 36 bp; B to C, 21 bp) suggests a possible cooperative and/or competitive binding effect. In several systems, STAT6 needs cooperation of other factors for efficient transcriptional activation, e.g., C/EBP, in the human IgE class-switching promoter (21). However, oligomerized STAT6 sites confer good inducibility of reporter genes when combined with a minimal promoter in the absence of other regulatory elements (39, 40, 41), which suggests that STAT6 dimers can also interact with each other. The observed binding of an unknown factor to or very close to the GAS motifs may be relevant in this context.

It has been shown previously that overexpression of CIS, SOCS-1, and SOCS-3 results in inhibition of signaling by a wide range of cytokines, hormones, and growth factors (24). Thus, by down-regulating the signaling pathways that stimulated their expression, SOCS proteins act as part of classical negative feedback loops. We have found in cotransfection studies that overexpression of SOCS-1 blocks IL-4-IL-13-stimulated transcription of a reporter gene under control of the SOCS-1 promoter. This negative autoregulation could be responsible for the characteristic expression kinetics of SOCS-1 mRNA, which features an early peak shortly after induction, followed by a lower but nonetheless elevated expression level. The early peak might be a direct consequence of the lag between activation of transcription and attainment of the final state of functional protein, when SOCS-1 becomes able to oppose cytokine stimulation, thereby limiting its own induction. The molecular mode of action by which SOCS-1 represses its own IL-4-IL-13-induced expression remains to be elucidated. A likely mechanism would be the direct interaction of SOCS-1 with one or more of the Janus kinases, as has been reported for SOCS-1 and SOCS-3 (42, 43, 44). An alternative mechanism would be the inhibition of cytokine signaling by competing with STATs for common phosphotyrosine motifs within the cytoplasmic domains of cytokine receptors, as was suggested for CIS (45). A third possibility encompasses promotion of the degradation of specific signaling proteins, as was shown before (19).

In summary, this study reveals another immediate-early gene that is induced by IL-4 and IL-13 via STAT6. By mediating expression of SOCS-1, STAT6 directly takes part in negative regulation of IL-4 and IL-13 signaling. STAT6 is considered an interesting target for novel therapeutic strategies in allergy and asthma (46, 47, 48, 49). It will have to be considered in this context that STAT6 appears to be involved in inducing SOCS-1 expression and that loss of SOCS-1 activity results in severe pathological symptoms in vivo.