STAT6 Is Required for the Anti-inflammatory Activity of Interleukin-4 in Mouse Peritoneal Macrophages*

Interleukin-4 (IL-4) is an anti-inflammatory cytokine which inhibits many inducible macrophage functions. The present study demonstrates that the ability of IL-4 to inhibit interferon γ (IFNγ)-dependent gene transcription is dependent upon STAT6. IL-4 suppressed IFNγ-induced expression of the MIG (monokine induced by IFNγ) gene, a C-X-C chemokine, in mouse macrophages. IFNγ-induced expression of MIG mRNA was abolished in peritoneal macrophages fromStat1−/− mice, and the suppression of MIG mRNA by IL-4 was abolished in macrophages from Stat6−/− mice. Transient transfection assays using a reporter gene containing the MIG gene promoter or the IFNγ-responsive element (γRE) from the MIG gene revealed that the IFNγ-dependent transcription was suppressed by IL-4, although IL-4 alone had no transactivating function. IFNγ and IL-4 activated STAT1 and STAT6, respectively, and both proteins were able to bind the γRE motif. Furthermore, STAT6 was associated with the co-activator CREB-binding protein in RAW264.7 cells. These observations indicate that STAT6 is necessary for the IL-4-mediated suppression of IFNγ-induced, STAT1-dependent transcription and suggest that STAT6 may directly suppress the STAT1-dependent transcription by competing with STAT1 for occupancy of the γRE motif and/or by competing with limiting quantities of the transcriptional coactivator.

IL-4 and IFN␥ exhibit antagonistic effects on macrophage gene expression (14 -18). We have previously observed that IFN␥-mediated induction and IL-4-mediated suppression utilize the same regulatory sequences (the ISRE in the IP-10 gene and the GAS motif in the IRF-1 gene) (16,40). Furthermore, IL-4 does not inhibit the activation of STAT1 by IFN␥ and IL-4-induced STAT6 is able to bind the IFN␥-responsive site without transactivation (40). Recent studies have shown that STAT1-dependent transcription requires transcriptional coactivators p300 and CREB-binding protein (CBP), which may be present in limiting amounts (41,42). On the basis of these lines of evidence, we postulated that the inhibitory effect of IL-4 may result, at least in part, from the direct action of IL-4-induced STAT6 on IFN␥-induced STAT1-dependent transcription. The present study was undertaken to determine if IL-4-induced STAT6 is required for the suppressive action of IL-4 on IFN␥induced gene transcription and to extend our analysis to include the IFN␥-responsive element (␥RE), which controls expression of the MIG chemokine gene (43,44), a third form of IFN␥-sensitive nucleotide regulatory motif. The results demonstrate that IL-4-dependent suppression of IFN␥-induced MIG gene expression is abolished in peritoneal macrophages from mice in which the STAT6 gene has been deleted (45). Both the MIG gene promoter and the ␥RE from the MIG promoter are sensitive to the stimulatory action of IFN␥ and the inhibitory effects of IL-4. Furthermore, we find that STAT6 as well as STAT1 can interact with the coactivator CBP. These results suggest that these two factors may compete for coactivator and/or occupancy of the ␥RE site with opposite functional consequences.  (45). Specific pathogen-free C57Bl/6 mice 9 -12 weeks of age were purchased from the Trudeau Institute (Saranac Lake, NY).
Preparation of RNA and Northern Hybridization Analysis-Total cellular RNA was extracted by the guanidine isothiocyanate-cesium chloride method (49). Northern hybridization analysis and cDNA probes for mouse MIG and rat GAPDH were described previously (47,50). Northern blots were also quantified using phosphorescence detection. The relative magnitude of expression was determined for each gene and normalized to values for GAPDH expression in the same experiment.
Nuclear Run-on Assay-Cultures of 6 ϫ 10 7 RAW264.7 cells in 150-mm Petri dishes were treated as indicated in the text for 90 min, and nuclei were isolated as described previously (51). Transcription initiated in intact cells was allowed to complete in the presence of [␣-32 P]UTP (NEN Life Science Products), and the RNA was isolated and hybridized to slot-blotted plasmid containing specific cDNA insert (7 g of DNA/slot) as described elsewhere (52).
Reporter Plasmids-A sequence encoding the 5Ј-flanking promoter/ enhancer region of the MIG gene was cloned by PCR using Ultima DNA polymerase (Perkin-Elmer) from a mouse genomic DNA using a set of primers corresponding to the mouse MIG genomic sequence (53) (upstream primer; 5Ј-TCGCCATATAGTGTCATGTCCC-3Ј, downstream primer; 5Ј-CTGTGTTGGAGTGAAGTCCG-3Ј). The amplified PCR fragment was subcloned into luciferase reporter construct (pGL2-B, Promega) and the nucleotide sequence confirmed. The sequences corresponded to nucleotide residues Ϫ328 through ϩ34 (53). The 5Ј-half-site of the ␥RE motif at position Ϫ198 to Ϫ189 of the MIG promoter was mutated by using two-step PCR amplification (54). The mutant sequence utilized was TCCCggAggATccACT. Lowercase letters represent the mutant sequence, and the underlined sequence is the 5Ј-half-site of the ␥RE. One copy of the ␥RE motif (sense strand: 5Ј-gATCCCTTAC-TATAAACTCCCCGTTTATGTGAAATGGA-3Ј) was placed in front of a heterologous thymidine kinase promoter (55) linked to the luciferase gene (pTK-81Luc). The luciferase control plasmid under control of the cytomegalovirus promoter was provided by Dr. G. Sen (Dept. of Molecular Biology, Cleveland Clinic Foundation).
Transient Transfection-Luciferase reporter gene plasmids were transiently transfected into RAW264.7 cells by the DEAE-dextran method as described previously (20,56). Briefly, 15 g of reporter plasmid DNA and 3 g of reference plasmid (pRL-TK, Promega) were transfected in the presence of DEAE-dextran (300 g/ml). To standardize transfection efficiencies, the transfected cells were harvested, pooled, and seeded in four 60-mm culture dishes. After 24 h, the cells were stimulated with IFN␥ and/or IL-4 for 8 h prior to analysis of luciferase activity. Firefly and Renilla luciferase activities were assayed using reagents provided by Promega according to the manufacturer's instructions. Twenty g of extract protein were utilized in each assay.
Electrophoretic Mobility Shift Assay (EMSA)-Nuclear extracts were prepared as described previously using a modified method of Dignam et al. (20,40,57). The following oligonucleotides were used in EMSA: ␥RE: For binding reactions, nuclear extracts (5 g of protein) were incubated in 12.5 l of total volume containing 20 mM HEPES (pH 7.9), 50 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol, 5% glycerol, 200 g/ml bovine serum albumin, and 1.25 g of poly(dI-dC) for 15 min at room temperature. 32 P-Labeled oligonucleotide (0.5 ng, 5 ϫ 10 5 cpm) was then added to the reaction mixture and incubated for 15 min at room temperature. The reaction products were analyzed by electrophoresis in a 5% polyacrylamide gel with 0.25ϫ TBE buffer (22.3 mM Tris, 22.2 mM borate, 0.5 mM EDTA). In some experiments, rabbit antibody to NFB1 (p50), RelA (p65), STAT1, and STAT6 were added prior to electrophoresis. The dried gels were analyzed by autoradiography and by phosphorescence detection.

IL-4 Suppresses IFN␥-induced Expression of MIG
mRNA-IL-4 is known to suppress IFN␥-induced gene expression in mononuclear phagocytes (14 -18). The MIG gene encodes a C-X-C Chemokine known to be transcriptionally regulated by IFN␥ (43,44,53). To determine whether IL-4 can suppress MIG expression, thioglycollate-elicited peritoneal macrophages were pretreated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with different doses of IFN␥ for 4 h prior to preparation of total RNA and Northern hybridization analysis. IFN␥ induced and IL-4 suppressed levels of MIG mRNA (Fig. 1A). IL-4 did not alter the kinetics of IFN␥-induced MIG mRNA expression but reduced the steady state mRNA levels over the entire time course (Fig. 1B). Quantitatively and qualitatively similar effects of IFN␥ and IL-4 on MIG mRNA were seen in the macrophage-like cell lines RAW264.7 (Fig. 1C) and ANA-1 (Fig. 1D). The suppression by IL-4 was mediated predominantly by inhibiting transcription as assessed by nuclear run-on assay in RAW264.7 cells (Fig. 2).
The Effects of IFN␥ and IL-4 on MIG Expression Require STAT1 and STAT6, Respectively-STAT1 is believed to be important for most if not all IFN␥-induced gene expression (46,58), while STAT6 is a primary signaling molecule linked with response to IL-4 (45,59,60). To examine the role of IFN␥induced STAT1 and IL-4-activated STAT6 in control of MIG expression, macrophages from Stat1Ϫ/Ϫ and Stat6Ϫ/Ϫ mice were used to measure each response (45,46) (Fig. 3). As expected, IFN␥-induced MIG mRNA expression was completely abolished in peritoneal macrophages from Stat1Ϫ/Ϫ mice ( Fig.  3A) but remained intact in macrophages from Stat6Ϫ/Ϫ (Fig.  3B). Whereas IFN␥-induced MIG mRNA was inhibited by IL-4 in wild type macrophages, the suppressive effect of IL-4 was entirely lost in macrophages from the Stat6Ϫ/Ϫ mice (Fig. 3, B and C).
The ␥RE Motif in the MIG Promoter Is Sensitive to Both IFN␥ and IL-4 -IFN␥-induced transcription of the MIG gene has been shown to depend upon the ␥RE motif located at positions Ϫ198 to Ϫ171 in the MIG promoter (44). To determine whether the suppression of IFN␥-induced MIG gene expression by IL-4 is dependent upon the ␥RE motif, 328 nt of the 5Ј-flanking sequence of the MIG gene were linked to the luciferase reporter gene and transiently transfected into RAW264.7 cells. Twentyfour hours after transfection, the cultures were stimulated with IFN␥ alone or in the presence of IL-4 for 8 h prior to analysis of luciferase activity. Although IL-4 alone had no effect on luciferase activity, reporter gene expression was markedly elevated in cells stimulated with IFN␥ (Fig. 4A). IL-4 added along with IFN␥ suppressed the induced luciferase activity by approximately 50%. IL-4 did not modulate the luciferase activity from the cells transfected with cytomegalovirus promoter-driven luciferase construct (data not shown).
To determine if the ␥RE motif of the MIG gene was the target site for IL-4, one copy of this motif was placed in front of the TK promoter (TK-81) linked to the luciferase reporter gene and transiently transfected in RAW264.7 cells (Fig. 4B). As was observed for the natural MIG promoter, IFN␥ stimulated lucif-erase expression. IL-4 alone had no effect but suppressed the response to IFN␥ by approximately 50%. Luciferase activity from TK-81 alone was not affected by IFN␥ and/or IL-4 (data not shown).
IL-4-induced STAT6 Interacts with Coactivator CBP-The promoter analysis of the MIG gene indicates that the ␥RE motif is essential for both the IFN␥-induced transcriptional activation of the MIG promoter and the suppressive effect of IL-4. Furthermore, STAT6 is required for the suppressive activity of IL-4. Since STAT1-dependent transcription has been demonstrated to require the presence of the coactivator p300 or CBP (41,42), it is conceivable that suppression might involve competition between STAT1 and STAT6 for limiting quantities of the coactivator. To explore this possibility, we assessed whether STAT6 can interact with CBP in vivo using a coimmunoprecipitation analysis (Fig. 5). Consistent with previous results (41,42), STAT1 was present in immunoprecipitates using antibody to CBP in whole cell lysates from RAW264.7 cells, and the association appeared to be independent of stimulation (Fig. 5A). The same CBP immunoprecipitates also contained STAT6 (Fig. 5B). These results indicate that STAT6 as well as STAT1 binds CBP and are consistent with the possibility that STAT6 may compete for coactivator CBP.
IL-4-induced STAT6 Binds to the ␥RE Motif-An alternative possibility is that IL-4-induced STAT6 might compete for binding of STAT1 to the ␥RE motif. Since STAT6 can bind to a number of sequence motifs without transactivating function (20,38,40), the occupancy of the ␥RE site by STAT6 might result in suppression and not transactivation. Indeed, although IFN␥ could stimulate reporter gene expression from constructs containing the ␥RE linked to the TK promoter, treatment with IL-4 was entirely inactive. To further test this hypothesis, we determined the ability of STAT1 or STAT6 to bind with the ␥RE motif. Nuclear extracts were prepared from untreated, IFN␥-treated, and/or IL-4-treated RAW264.7 cells and analyzed by EMSA using a radiolabeled oligonucleotide corresponding to the ␥RE motif (Fig. 6A). Several constitutive ␥RE binding activities were detected in nuclear extracts from untreated cells (designated as I and II). IFN␥ induced two additional ␥RE binding activities (lane 2, a low and a high mobility complex designated as ␥RF-1 and ␥RF-2, respectively). Interestingly, IL-4 induced a single binding complex on the ␥RE motif (lane 3). When the cells were co-stimulated with IFN␥ and IL-4, the pattern of complex formation was essentially the same as seen in nuclear extracts from the cells treated with individual stimuli. As reported previously, IL-4 did not effect the tyrosine phosphorylation or nuclear localization of STAT1 in response to IFN␥ or vice versa (40).
To identify the proteins participating in formation of the ␥RE binding complexes, antibody supershift assays were performed (Fig. 6B). The constitutive binding activities (I and II) in nu-FIG. 2. IL-4 suppresses rate of the IFN␥-induced MIG transcription. RAW264.7 cells were untreated (UT) or pretreated with IL-4 (10 ng/ml) for 15 min, followed by followed by stimulation with or without IFN␥ (100 units/ml) for 90 min prior to isolate nuclei. Transcription was assessed by nuclear run-on assay as described under "Experimental Procedures." Radiolabeled RNA transcripts were hybridized with slot blotted cDNA encoding MIG or GAPDH as indicated. Similar results were obtained in two separate experiments.

FIG. 1. IL-4 suppresses IFN␥-induced MIG mRNA expression in macrophages.
A, thioglycollate broth-elicited peritoneal macrophages were pretreated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with different concentrations of IFN␥ for 4 h prior to preparation of total RNA and analysis of specific mRNA levels by Northern hybridization as described under "Experimental Procedures." Five g of total RNA were analyzed in each lane. Similar results were obtained in two separate experiments. B, time dependence of MIG mRNA expression in IFN␥-and/or IL-4-stimulated macrophages. Thioglycollate broth-elicited peritoneal macrophages were pretreated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with IFN␥ (100 units/ml) for the indicated times prior to analysis of specific mRNA levels as described above. Northern blots were quantified by phosphorimage analysis, and relative mRNA levels are presented as percentage of maximum expression as described under "Experimental Procedures." C and D, RAW264.7 (C) or ANA-1 (D) macrophages were unstimulated or pretreated with IL-4 (10 ng/ml) for 15 min prior to stimulation with increasing concentrations of IFN␥ as indicated for 4 h. Levels of MIG mRNA were determined by Northern hybridization as described above. Similar results were obtained in two separate experiments. clear extracts from untreated cells were reactive with antibodies to NFB1 (p50) and RelA (p65) (lanes 4 and 5). This result is consistent with previous observations of low level of NFB binding activity in nuclear extracts from unstimulated macrophages (61) and that NFB1 (p50) and RelA (p65) recognize and bind to some SBE motifs with low affinity (50,62). The most slowly migrating constitutive complex represents nonspecific binding since specific competitor could not eliminate the complex (data not shown). When nuclear extracts from IFN␥treated cells were analyzed, antibody to STAT1 supershifted the IFN␥-induced ␥RE binding complexes without affecting NFB1 and RelA (lane 7). A combination of antibodies to STAT1 and NFB1 supershifted the inducible and constitutive binding complexes (lane 11). When nuclear extracts from IL-4treated cells were analyzed, antibody to STAT6 supershifted the IL-4-induced complex (lane 15). The residual binding activity in nuclear extracts treated with anti-STAT6 and anti-NFB1 (lane 19) appears to be the supershifted-NFB1 as seen in lanes 4 and 11. These results indicate that both IL-4-induced STAT6 and IFN␥-induced STAT1 recognize the ␥RE motif. Two different complexes containing STAT1 were identified; the more rapidly migrating complex (␥RF-2) is likely to be STAT1

FIG. 3. STAT6 is indispensable for the IL-4-mediated suppression of the IFN␥-induced, STAT1-dependent expression of MIG mRNA.
A, thioglycollate broth-elicited peritoneal macrophages from wild type mice or from Stat1Ϫ/Ϫ mice were untreated (UT) or stimulated with IFN␥ (100 units/ml) for 4 h. Levels of murine MIG mRNA were determined by Northern hybridization. Five g of total RNA were analyzed in each lane. B, thioglycollate broth-elicited peritoneal macrophages from wild type mice or from Stat6Ϫ/Ϫ mice were untreated (UT) or pretreated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with IFN␥ (100 units/ml) for 4 h prior to analysis of specific mRNA levels by Northern hybridization. Five g of total RNA were analyzed in each lane. C, Northern blots were quantified by phosphorimage analysis and relative mRNA levels are presented as percentage of the IFN␥-induced expression. Each column and bar represents the mean Ϯ S.E. from three independent experiments.

FIG. 4. Effects of IFN␥ and IL-4 on transcription from the MIG promoter.
A, the luciferase reporter construct (shown schematically) was transiently transfected into RAW264.7 cells as described under "Experimental Procedures." Twenty-four h after transfection, the cells were pretreated with IL-4 (10 ng/ml) for 15 min prior to stimulation with IFN␥ (100 units/ml) for 8 h, followed by analysis of cell extracts for luciferase activity. The relative luciferase activity is presented as a percentage of activity obtained in cells stimulated with IFN␥ (100 units/ml) alone. Each column and bar represents the mean Ϯ S.E. from three independent experiments. B, RAW264.7 cells were transiently transfected with a luciferase reporter plasmid containing one copy of the MIG ␥RE motif linked to the TK promoter (TK-81) as shown schematically. After 24 h, the cells were either unstimulated (UT) or treated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with IFN␥ (100 units/ml) for 8 h prior to measurement of luciferase activity. The relative luciferase activity is presented as percentage of maximum activity obtained in cells transfected with the reporter plasmid stimulated with IFN␥ (100 units/ml). Each column and bar represents the mean Ϯ S.E. from three independent experiments. The -fold induction of stimulated versus unstimulated samples is also indicated over the column.
STAT6 Binds to the 3Ј-Half-site of ␥RE-1 Motif-The ␥RE motif consists of two GAS-like motifs (Fig. 7D), and both sites have been shown to be required for IFN␥-induced transcription of the MIG gene (44). To determine whether the binding of IL-4-induced STAT6 to the ␥RE motif also requires both 3Ј-and 5Ј-half-sites, the 5Ј-half-site of the ␥RE motif was mutated and tested for binding activity by EMSA (Fig. 7A). Although the IFN␥-induced ␥RF-1 binding complex was markedly reduced on the mutant oligonucleotide (lane 6), IL-4-induced STAT6 binding remained intact (lane 7). Interestingly, the IFN␥-induced high mobility complex (␥RF-2) substantially increased when the mutant oligonucleotide was used as the probe (lane 6). The IFN␥-induced transcriptional activity of the ␥RE was almost completely abolished when a luciferase reporter construct containing a mutant 5Ј-half-site was analyzed by transient transfection (Fig. 7B), indicating that the formation of the ␥RF-1 complex is essential for IFN␥-induced MIG promoter activity. The residual IFN␥-induced promoter activity in the mutant luciferase construct was still sensitive to IL-4. To determine if the 3Ј-half-site of the ␥RE motif alone is sufficient for STAT6 binding, oligonucleotides corresponding to the 3Ј or the 5Ј-half-site of the ␥RE motif were independently synthesized and tested for IFN␥-and IL-4-induced binding activity by EMSA. Consistent with the results shown in Fig. 7A, IL-4-induced STAT6 was able to form a complex on the 3Ј-half-site (lane 15), while IFN␥ stimulated primarily the formation of the high mobility complex ␥RF-2 (lane 14). Interestingly, when the 5Ј-half-site was used as probe (lanes [17][18][19][20], neither IFN␥ nor IL-4 induced any detectable complex formation. These results indicate that while the 3Ј-half-site of the ␥RE motif is sufficient for STAT6 binding and for ␥RF-2, formation of ␥RF-1 requires both the 3Ј-and 5Ј-half-sites. Taken together with the previous observation that the ␥RE is not responsive to IL-4 alone (see Fig. 4), the data from the association of STAT6 with CBP and the ␥RE binding suggest two potential possibilities for the antagonistic functions of STAT6 on IFN␥-induced STAT1-dependent transcription of the MIG gene. 1) IL-4-induced STAT6 could compete for the limiting amount of coactivator CBP, which is necessary for the STAT1-dependent transcription; and/or 2) STAT6 could reduce the frequency of IFN␥-induced transcription by competing for occupancy of the ␥RE by STAT1.

DISCUSSION
The initiation, magnitude, and resolution of immune-mediated inflammation are regulated at least in part through the action of pro-and anti-inflammatory cytokines (1,2). For example, the Th1-derived cytokine IFN␥ promotes the pro-inflammatory function of macrophages through stimulation of new gene expression, and this is often antagonized by the Th2-derived cytokine IL-4 (2-5). IFN␥ and IL-4 are both known to activate distinct members of the STAT family of transcription factors, each of which may exhibit differential transactivating potential depending upon the sequence motif to which they bind (20,34,(37)(38)(39)(40). In consideration of this, we postulated that the antagonistic action of IL-4 on IFN␥-stimulated gene transcription is mediated, at least part, through a direct antagonistic action of IL-4-activated STAT6. The results from the present study demonstrate that STAT6 is indispensable for thenegativeregulatoryfunctionofIL-4onIFN␥-induced,STAT1dependent transcription of the MIG gene in mouse macrophages, and both factors appear to mediate their action through the ␥RE motif found in the MIG gene promoter. These conclusions are based on the following observations. 1) IFN␥induced MIG mRNA expression was suppressed by IL-4 in primary macrophages and in ANA-1 or RAW264.7 macrophage-like cell lines. 2) IFN␥-induced MIG gene expression was abolished in peritoneal macrophages from Stat1Ϫ/Ϫ mice, while the suppressive effect of IL-4 was abolished in macrophages from Stat6Ϫ/Ϫ mice. 3) IL-4 suppressed the IFN␥induced transcription of a luciferase reporter gene linked to either a 328-base pair fragment flanking the transcriptional start site of the MIG gene or to a single copy of the ␥RE motif from the MIG gene in the context of a heterologous promoter; IL-4 alone showed no independent transactivating function on either construct.
Our findings provide support for two possible mechanisms through which STAT6 may function to directly suppress IFN␥induced gene expression. One mechanism of STAT6-mediated suppression may involve the participation of transcriptional co-activators such as CBP or p300. It has been reported recently that STAT1-dependent transcription requires either CBP and/or p300 (41,42). Thus the finding that STAT6 can interact with CBP raises the possibility that STAT6 might sequester limiting quantities of CBP and thereby antagonize the STAT1-dependent transcription (41). The association of both STAT1 and STAT6 with CBP was not altered by stimulation with either IFN␥ or IL-4. Thus, these data do not provide any direct evidence of competition between the two STATs for CBP. Nevertheless, further studies of the functional conse- Whole cell extracts were prepared from RAW264.7 cells either untreated (UT) or pretreated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with or without IFN␥ (100 units/ml) for 30 min before lysis the cells. Whole cell extracts (ϳ500 g of protein) were immunoprecipitated (IP) with anti-CBP antibody or normal rabbit IgG as indicated. CBP-associated proteins were then eluted, and half of the eluted protein was individually separated by 7.5% SDS-PAGE. One g of whole cell extract was also loaded onto the gel (Input). Proteins were transferred to membrane and each blots were developed with anti-STAT1␣ (A) and anti-STAT6 (B) antibodies, respectively. The blot in A was subsequently stripped and immunoblotted with anti-CBP antibody (C). Similar results were obtained in two separate experiments.
quences of STAT6-CBP interaction will be required to determine the physiological significance.
The present results are also consistent with a scenario in which STAT6 and STAT1 compete for occupancy of the ␥RE regulatory site, as has been suggested previously in studies of the effects of IL-4 on IFN␥-stimulated IRF-1 expression (40). Because STAT6 has poor transactivating function on the MIG promoter, occupancy of the STAT1 binding site by STAT6 would result in reduced transcription. The nature of the regulatory sequence motif that determines both induction and suppression of the MIG gene (␥RE) provides some unique features to the potential interactions between SBEs and STAT1 or STAT6. First, the formation of the ␥RF-1 complex requires both the 3Ј-and 5Ј-half-sites of the ␥RE motif while STAT6 only recognizes the 3Ј-half-site. Second, the mutant 5Ј-half-site version of the ␥RE motif will form ␥RF-2 but not ␥RF-1 and is functionally inactive in supporting IFN␥-driven reporter gene transcription. Thus, the ␥RF-1 complex but not the ␥RF-2 complex is essential for the IFN␥-induced transcriptional activity of the MIG gene. Finally, IL-4-induced STAT6 is able to bind to the ␥RE (3Ј-half-site) without transactivating function. These findings suggest that binding of STAT6 on the 3Ј-halfsite may prevent the formation of the ␥RF-1 complex, resulting in decreased transcriptional frequency.
STAT6 exhibits the broadest DNA binding specificity of all STAT family members (20,(33)(34)(35)(36)(37)(38)(39)(40). However, binding of STAT6 to IFN␥-responsive motifs does not lead to transcriptional activation despite the fact that STAT6 can interact with CBP. In contrast, STAT6 functions as a transcriptional activator when bound to a subset of SBEs including those found in the IgE or IL-1 receptor antagonist genes (20,34). Thus, the nucleotide sequence of the DNA binding site is a critical determinant of the transcriptional activating function of STAT6. This may be an important distinguishing feature of STAT6, which allows both positive and negative effects on gene transcription.
The ability of STAT6 to compete with STAT1 for binding to the ␥RE was not detectable by EMSA (Figs. 6 and 7). This result is, however, not surprising since the oligonucleotide concentration in the reaction is in great excess relative to the concentration of STAT1 or STAT6, a condition under which competition between the protein factors will not occur. Reduction of the oligonucleotide concentration to a level at which competition might occur does not allow detection of the complexes (data not shown). Thus EMSA analysis of ␥RE binding activities using nuclear extracts from macrophages treated with IFN␥ and IL-4 would not be expected show reduced formation of the ␥RF-1 complex. STAT1 has been shown to exhibit a cooperative binding activity with two or more low affinity SBE motifs, which is mediated by the N-terminal region of the protein (63,64). Thus, it is likely that the IFN␥-induced ␥RF-1 complex is a tetrameric form of STAT1. The higher mobility ␥RF-2 complex appears to be a STAT1 homodimer, which binds to the 3Ј-half-site of the ␥RE motif. Interestingly, STAT6 formed only a single complex despite its ability to bind a relatively wide range of SBE motifs (20,(33)(34)(35)(36)(37)(38)(39)(40). Whether STAT6 lacks cooperativity in the Nterminal region or the ␥RE motif is insufficient for the cooperative binding of STAT6 remains to be determined.
Although IL-4-mediated anti-inflammatory function has been found to include both post-transcriptional and translational events (8,12,66), the majority of reports have identified transcription as a primary molecular target (8,14,16,40,(67)(68)(69)(70)(71). The data in the present study further demonstrate that STAT6 is essential for IL-4-mediated suppression of IFN␥induced MIG gene expression. Indeed, we have also observed that IL-4-mediated suppression of IFN␥-induced IP-10, IRF-1, ICAM-1, and inducible nitric-oxide synthase gene expression are also abolished in peritoneal macrophages from Stat6Ϫ/Ϫ FIG. 6. IFN␥ and IL-4 stimulate different ␥RE binding activities in RAW264.7 cells. A, nuclear extracts were prepared from RAW264.7 cells either untreated (UT) or pretreated with IL-4 (10 ng/ml) for 15 min, followed by stimulation with or without IFN␥ (100 units/ml) for 30 min prior to the preparation of nuclear extracts. Five g of nuclear extract protein were utilized for assay of DNA binding activities by EMSA using radiolabeled oligonucleotides containing the ␥RE element from the MIG gene. Constitutive binding activities (designated as I and II) and IFN␥and IL-4-induced binding activities are indicated as filled and open arrows, respectively. Nonspecific binding activity is designated as ns. B, IL-4-induced ␥RE binding complex contains STAT6. Five g of nuclear extract protein prepared from RAW264.7 cells were utilized in each reaction with or without addition of antibody (1 g) to STAT1, STAT6, NFB1, or RelA and assayed for DNA binding activity as described above. mice (data not shown). IL-13-mediated suppression of inducible nitric-oxide synthase gene expression has also been shown to depend upon STAT6 (72). This collection of studies indicates that IL-4-or IL-13-induced STAT6 can provide negative regulatory function, which is necessary for the anti-inflammatory cytokine action.