BIOgel-elicited peritoneal Mθ were stimulated with classical Thl¹ (IFN-γ) or Th2 (IL-4) cytokines or LPS (Figure 1). RNA and culture supernatant samples were simultaneously prepared from single wells at defined time points over a 72-hour period and were analysed for murine TARC (mTARC) mRNA and protein expression. Cell viability over this time period was independently confirmed by MTT assay (data not shown). mTARC mRNA was almost completely absent from untreated cells at all time points (Figure 1A), and was most potently induced by IL-4, consistent with its reported expression pattern in human Mθ 1525. Onset of mTARC induction by IL-4 was seen at 4 hours and peaked at 24 hours, when mTARC mRNA induction over untreated samples was more than 1,000-fold. IL-13 was also a potent inducer of mTARC mRNA at this time point, as measured by semi-quantitative PCR (data not shown). After 24 hours, there was a dramatic decline in mTARC mRNA levels detected in the IL-4-treated samples, although expression had still not returned to basal level within 72 hours. Neither LPS nor IFN-γ stimulated up-regulation of mTARC mRNA, even though these treatments resulted in significant induction of IP-10 (IFN-γ-inducible protein 10, CXCL10) 42 mRNA [see Additional file 1]. IFN-γ was also found to inhibit mTARC mRNA induction by IL-4 when added either prior to or concurrently with IL-4 treatment [see Additional file 1]. mTARC protein expression closely followed the mRNA profile in time of onset and duration (Figure 1B). As with the mRNA, mTARC protein expression was only detected in IL-4-treated samples. mTARC protein first appeared in the culture supernatants of IL-4-treated cells after 24 hours. Protein levels reached a peak at 48 hours, and had not declined greatly by 72 hours, as previously documented for human PBMCs 25 and T lymphocytes 43.

The kinetics of the IL-4 induction of TARC are such that several different mechanisms of promoter activation might be possible 44, including activation of PI3 kinase 4546 or the IL-4-activated transcription factor, PPARγ 4748. However, mTARC induction by IL-4 was not inhibited by the PI3 kinase inhibitor, wortmannin, or the PKC inhibitor, chelerythrine chloride (data not shown) and mTARC mRNA was not up regulated by the PPARγ agonists, rosiglitazone, troglitazone or 15-deoxy-Δ^12,14-prostaglandin J₂ (data not shown). As STAT6-mediated gene transcription does not require de novo protein synthesis 49, we measured mTARC induction by IL-4 in the presence of the translational inhibitor, cycloheximide. Cells were pre-treated for 2 hours with doses of cycloheximide found to abrogate ³⁵S-methionine incorporation. IL-4 was added and cells cultured for a further 4 or 24 hours before RNA harvest. mTARC mRNA was still induced by IL-4 in the presence of both cycloheximide doses at 4 hours and at 24 hours [see Additional file 2]. Furthermore, cycloheximide treatment resulted in mTARC superinduction, as has been noted for early immediate genes such as COX-2 50. These results indicate that de novo protein synthesis is not a pre-requisite of the IL-4 induction of mTARC and may actually act to reduce mTARC transcription.

To determine the involvement of STAT6 in mTARC regulation, thioglycollate-elicited peritoneal Mθ were isolated from both wild-type and STAT6^-/- 51 mice and treated with IL-4 for 24 hours (Figure 2A). Whereas the wild-type BALB/c Mθ exhibited the characteristic strong IL-4 induction of mTARC mRNA, the STAT6^-/- Mθ did not, demonstrating an absolute requirement of STAT6 for the IL-4-mediated up regulation of mTARC in Mθ. Binding of STAT6 to the endogenous mTARC promoter was also confirmed in ChIP assays using IL-4-treated bone marrow-derived Mθ (Figure 2B) in the presence or absence of a STAT6 blocking peptide. This experiment shows that STAT6 is rapidly recruited to the mTARC promoter in IL-4 treated Mθ.

We obtained the sequence for the MDC-Fractalkine-TARC loci on mouse chromosome 8 from the Celera and Sanger Centre 52 databases and identified eleven putative STAT6 binding sites (called STAT6 T1-T11 in order of proximodistal location) within a 1.2 kb region 5' of the mTARC coding sequence (Figure 3). Three of the putative mTARC STAT6 sites (T2, T4 and T6) conform to the published consensus sequence for STAT6 binding, TTC(N)₄GAA 39. The T1 site has TTC(N)₃GAA spacing, to which STATs 1, 3, 4 and 5 535455 can bind, as well as STAT6 3356. The remaining seven sites represent variations of the TTC(N)₄GAA consensus, each possessing 4 central nucleotides flanked by the A and T residues that delineate the binding site. The T1 and T2 sites were found to be highly conserved between human and mouse, suggesting they may be of particular importance in mediating TARC induction by IL-4 in both species.

To analyse the relative contribution of sites T1-T11 to the IL-4 induction of transcription, we initially generated a 5' deletion series of promoter constructs in which these putative STAT6 sites were sequentially lost from a full-length -1173 construct containing all eleven putative sites and the transcriptional start site (Figure 4). The nomenclature of the mTARC promoter plasmids is derived from assigning the start point of transcription of the mTARC gene as + 1 30. These constructs were co-transfected into the RAW 264.7 Mθ cell line with a STAT6 expression vector to ensure transcription factor expression was not rate limiting or variable between samples. When stimulated with IL-4 for 18 hours, these deletion constructs demonstrated differential activation. Activation of full-length -1173 mTARC promoter was reproducibly induced at least 12-fold by IL-4, whereas the minimal mTARC promoter (mTARC -128), which lacks any STAT6 sites, was not at all IL-4-responsive. Deletion of DNA sequences between -649 and -498 (containing T5 and T6 sites) and -467 and -128 (containing T1 and T2 sites) reduces IL-4 induction, indicating that these two regions are important to the overall IL-4 response of the mTARC promoter.

The STAT6 binding capacity of each putative STAT6 site was tested in vitro by EMSA (Figure 5). Nuclear extracts from C57BL/6 BIOgel-elicited Mθ untreated or treated with IL-4 or IFN-γ for 1 hour were prepared and incubated with radiolabelled probes for each individual mTARC STAT6 site. The probes containing the T2 and T4 STAT6 consensus sites were capable of binding a complex of the predicted mobility for STAT6 specifically present in extracts from IL-4-treated samples (Figure 5A). The T1 promiscuous STAT binding site and the T11 imperfect consensus site were also able to bind this complex. Surprisingly, the probe containing the T6 'perfect' STAT6 consensus site did not bind this complex (data not shown). The T3 and T5-T10 probes bound larger complexes that were not specific to IL-4-treated samples (data not shown). The IL-4-induced complex bound by T1, T2, T4 and T11 was supershifted by a STAT6-specific Ab, but not by a STAT1 Ab, further confirming that it contained STAT6 (Figure 5B). Both unlabelled T1 and T4 probes added in 100-fold molar excess were able to efficiently compete the STAT6 complex from the T1, T2, T4 and T11 probes. These experiments show that STAT6 can bind to sites with TTC(N)₃GAA as well as TTC(N)₄GAA spacing in the mTARC promoter and that perfect STAT6 consensus sequence is not the only determinant of STAT6 binding.

The binding capacity of an oligonucleotide in vitro is not always representative of its functional activity within cells 57 and so we compared the relative IL-4-inducibility of individual key mTARC putative STAT6 sites in transient transfection experiments. We generated luciferase reporter plasmids containing trimers of the T1, T2, T4 and T11 sites, which bound STAT6 in EMSA assays, and also of the T6 perfect consensus site, which did not bind STAT6, but has perfect consensus STAT6 binding sequence and might contribute to the activity of sequence region -649 to -498 (Figure 4).

Oligonucleotides consisting of three copies of the EMSA sequence for each of the sites were cloned onto a heterologous promoter, the minimal Emr1 -40 promoter 58, in pGL3 Basic. The Emr1 -40 minimal promoter was selected for its low basal activity in Mθ, lack of inducibility by IL-4 (Figure 8A) and absence of TATA box (similar to mTARC). Constructs containing sites in forward and reverse orientations were identified. When co-transfected with STAT6 into RAW 264.7 cells (Figure 6), trimers of T1 or T6 in either the forward (3XT1 F and 3XT6 F) or reverse orientations (data not shown) were not induced by IL-4 at all. In contrast, trimers of T2, T4 and T11 were induced by IL-4 regardless of insert orientation (data shown for forwards orientation only, 3XT2 F, 3XT4 F and 3XT11 F). The T4 trimer was only minimally induced by IL-4 (~3.5-fold), whereas T2 and T11 trimers were more highly induced than the mTARC -1173 full-length promoter construct. Thus, sites equally capable of binding STAT6 in vitro can have quite disparate capacity for activation by IL-4 in Mθ when assayed by transfection.

To resolve the contribution of individual putative STAT6 sites to overall mTARC promoter IL-4 inducibility, we performed site-directed mutagenesis of key sites within the context of the full-length -1173 promoter. The T1, T2, T4, T5, T6 and T11 sites were mutated, as these sites were implicated as strong candidate functional STAT6 by previous experiments (see above). As the 5' TTC nucleotide triplet is highly conserved amongst both published STAT6 sites 405960 and all mTARC putative sites capable of binding STAT6 in EMSA experiments, this sequence was mutated to 5' AAC in order to abrogate STAT6 binding 61. Mutant and wild-type sequence information was compared using the Genomatix MatInspector 62 to confirm that mutagenesis did not create new transcription factor binding sites that could alter the IL-4 responsiveness of the constructs (data not shown). Each mutant construct was co-transfected with STAT6 into RAW 264.7 cells and stimulated with IL-4 for 18 hours (Figure 7).

This assay confirmed that T1 and T2 are of critical importance to the IL-4 induction of mTARC -1173. Mutation of T2 resulted in complete abrogation of promoter activation by IL-4, whereas T1 mutation reduced promoter fold induction by IL-4 from 12-fold (mTARC -1173 wild-type) to 2-fold. This is consistent with results shown in Figures 4 and 5, but in contrast with the functional activity of the isolated T1 trimer in Figure 6. Mutation of T4 reduced promoter IL-4 induction by 40%, indicating, as suggested by Figures 4 and 6, that T4 contributes less significantly than T1 and T2 to the overall IL-4 induction. Similarly, although T11 bound STAT6 in EMSA and was potently activated by IL-4 as an isolated trimer, it does not appear to have significant activity within the context of mTARC -1173 (Figures 4 and 6). Interestingly, mutation of either T5 or T6 significantly reduced mTARC -1173 promoter IL-4 induction from 12-fold to 4-fold, demonstrating that these sites do contribute to overall IL-4 induction (consistent with Figure 4), even though they cannot bind or be activated by STAT6 when isolated from the promoter. Thus, the functional activity of STAT6 sites in isolation may be distinct from their functional activity within the context of their endogenous promoter. A summary of this data is compiled in Table 1.

We next tested whether the natural configuration of sites within the mTARC promoter could confer IL-4 inducibility on heterologous promoters and hence act as a transferable IL-4 response element. The Emr1 -40 (non TATA box) and the EIF4A1 -40 (TATA box) 63 minimal promoters were selected as neither was induced by IL-4 nor had strong basal activity when transfected into RAW 264.7 cells (Figure 8A). To assess the significance of the distal (non-binding) putative STAT6 sites and potential cooperation between the T5 and T6 sites in enhancing mTARC induction by IL-4, promoter constructs covering two different regions of the mTARC proximal promoter were designed. One region from position -132 (refer to Figure 3) to -526 contained the five most proximal putative STAT6 sites and bisected the sequence containing T5 and T6. A second region from position -132 to -868 contained nine putative STAT6 sites and intact T5-T6 sequence. The T11 site was not included in these constructs because it had not appeared significant in the experiments shown in Figure 7. Constructs were synthesised containing either the five (5XST6) or the nine (9XST6) putative STAT6 sites on each heterologous minimal promoter in both forward and reverse orientations.

When co-transfected with STAT6 into RAW 264.7 and stimulated with IL-4 (Figure 8A), these mTARC promoter elements were able to confer potent IL-4 inducibility on both the Emr1 and the EIF-4A promoters. Interestingly, only sites inserted in the forward orientation were induced by IL-4 (reverse constructs not shown). The 9XST6 constructs gave 95-135-fold induction with IL-4, whereas the 5XST6 constructs gave only 8-25-fold induction with IL-4. This suggests that distal mTARC promoter sites do contribute to the overall IL-4 response, possibly via cooperative interactions 4164656667.

To examine the potency of the most IL-4 inducible construct, Emr1 9XST6 F, we performed dose-response experiments. The full-length mTARC -1173 and the Emr1 9XST6 F construct were transfected into RAW 264.7 cells and treated with IL-4 concentrations ranging from 0.01–20 ng/ml. Both constructs gave characteristic sigmoidal response curves that began to plateau above 5 ng/ml IL-4 (Figure 8B). The Emr1 9XST6 F construct showed greater induction by IL-4 than mTARC -1173 at all concentrations used above 0.25 ng/ml, which appeared to be the lower limit of the response. Importantly, the Emr1 9XST6 F construct still demonstrated a 40-fold induction with concentrations of IL-4 as low as 1 ng/ml, demonstrating that mTARC promoter sequences constitute a very sensitive IL-4 response element.

No de novo protein synthesis was required for the IL-4 induction of mTARC, indicating direct interaction of STAT6 with the mTARC promoter. This was confirmed by ChIP experiments revealing STAT6 bound to the endogenous mTARC promoter in IL-4-treated Mθ. Fulkerson et al. 70 identified mTARC (CCL17) as a CC chemokine induced early in the lung in an allergen challenge model of asthma and using STAT6 deficient mice showed that mTARC expression in this disease model was STAT6 dependent. During the preparation of this manuscript, Wirnsberger et al. 43 reported the involvement of two STAT6 sites (one corresponding to the T2 site in our study) in the IL-4-mediated regulation of TARC in human T lymphocytes, but did not investigate the additional highly conserved proximal T1 site that we identified. The authors identified these STAT6 sites by 'visual inspection' rather than systematic analysis using transcription factor search engines and so the data do not exclude the possibility of more distal functional STAT6 sites that may diverge from strict consensus sequence 81. We identified eleven putative STAT6 sites within the mTARC proximal promoter and determined that five of these had functional activity within the context of the mTARC -1173 construct. Hence the regulation of mTARC by IL-4 is more complex than previous work suggests, involving multiple STAT6 sites. To our knowledge, such interaction between STAT6 sites has only previously been reported for the promoter of the SOCS1 gene 80.

Of the five functional STAT6 sites identified, the two sites conserved between human and mouse, T1 and T2, were found to be important mediators of the mTARC induction by IL-4, even though T1 is not considered a classical STAT6 binding site 43. By site-directed mutagenesis, T2 was shown to be absolutely essential for IL-4 induction of the mTARC promoter and mutation of T1 in the context of the -1173 promoter resulted in an 85% reduction in promoter induction by IL-4. The significance of the T2 site was predicted from its perfect TTC(N)₄GAA consensus STAT6-binding sequence and ability to bind and be activated by STAT6 as an isolated site. In contrast, the T1 site has TTC(N)₃GAA spacing, previously associated with weak or inhibitory activity when occupied by STAT6 3382, and also could not be activated by IL-4 as an isolated site, even though it bound STAT6.

Three additional sites (T4, T5 and T6) were found to support the IL-4 induction of mTARC to a lesser extent than T1 and T2, even though both T4 and T6 are perfect consensus STAT6-binding sites. Interestingly, the T5 imperfect STAT6 site was found to have equivalent function to T6 within the context of the mTARC -1173 promoter and the significance of both sites did not correlate with their capacities for STAT6 binding in EMSA. Thus, the findings of this paper sound a cautionary note for prediction of STAT6 site function from sequence data alone and emphasise the importance of adopting multiple approaches for the analysis of STAT6-mediated gene transcription. Notably, STAT6 has also been shown to bind to an imperfect palindrome site in the delta-opioid receptor 81, suggesting that functional STAT6 sites in other genes are also not restricted to a simple TTC(N)₄GAA consensus sequence and cannot be identified as such.

The involvement of five closely-apposed STAT6 sites in the IL-4 induction of mTARC suggests that overall promoter induction may result from interactions between sites. Cooperation between proximal STAT6 sites in the SOCS1 promoter has been described 80 and other STAT proteins have similarly been shown to interact to activate gene transcription 4164656667. This arrangement of functional sites may stabilise STAT6 binding to low-affinity or non-canonical sites, such as T5, enhancing the overall transcriptional response to IL-4. Obligate stabilising interactions between STAT6 dimers bound at the juxtaposed T5 and T6 sites may explain the lack of STAT6 binding to these sites in isolation, as well as the comparable reduction in promoter induction by IL-4 when either site was mutated.

From our analyses of the functional activity of mTARC STAT6 sites, we were able to generate transferable promoter elements that conferred potent IL-4 responsiveness on two heterologous promoters. Constructs containing five mTARC putative STAT6 sites and intervening promoter sequences in their endogenous arrangement (5XST6) were induced up to 95-fold by IL-4, whereas constructs containing nine sites (9XST6) were induced up to 135-fold. The significantly greater activation of the 9XST6 constructs containing both T5 and T6 sites than the 5XST6 constructs lacking T6 may have resulted from interactions between the T5 and T6 sites. Although the 5XST6 constructs contained the T1, T2, T4 and T5 functionally significant STAT6 sites, stimulation with IL-4 resulted in promoter activity that was only 10–20% that of the 9XST6 constructs. Thus, distal promoter regions play an important role in mTARC induction by IL-4. Importantly, the potent IL-4 inducibility of these heterologous promoter constructs suggests that mTARC-derived IL-4 responsive elements could be used to direct IL-4-inducible gene expression in vivo with potential scientific and therapeutic benefit.

All experiments were carried out according to Home Office guidelines and with appropriate local ethical approval. Male C57BL/6, BALB/c or STAT6^-/- 51 mice (8–12 weeks) were injected i.p. with 1 ml BIOgel beads (Bio-Rad) in PBS 2% w/v or thioglycollate broth (Difco) and elicited peritoneal macrophages (Mθ) were harvested by lavage with PBS/5 mM EDTA on day 4. Cells were usually plated at 2 × 10⁶ Mθ per 35 mm dish in OptiMEM (Invitrogen) (determined by MTT¹ assay 8485 to promote optimal cell viability over 72 hours) supplemented with L-glutamine and penicillin/streptomycin. Cells were washed 2 hours after plating and used in assays after 24 hours at 37°C. Cells were treated with recombinant cytokines (R&D Systems) or LPS (S. typhimurium, SigmaAldrich) at the concentrations and time periods stated. RAW 264.7 cells were cultured in RPMI (Invitrogen) supplemented with L-glutamine, penicillin/streptomycin, and 10% foetal calf serum. Cells were plated at 2 × 10⁶ cells per 35 mm dish and left at 37°C for 30 minutes to adhere before cytokine treatment.

Total RNA was prepared by scraping cells into RNAzol (Biogenesis) according to manufacturer's instructions. RNA was DNAse-treated (Promega) and 1 μg was reverse transcribed using primer (dT)₁₅ for cDNA synthesis (Roche) and M-MLV-RT enzyme (Promega). Resultant cDNA was used in 20 μl final volume semi-quantitative PCR reactions using Red Hot DNA Polymerase enzyme and reaction mix (ABGene), supplemented with 1.5 mM MgC1₂, 0.2 mM each dNTP and 0.3 μM forward and reverse primers. Primers for semi-quantitative PCR were as follows: β₂M¹(β₂-microglobulin): 5' TGACCGGCTTGTATGCTATC and 5' CAGT GTGAGCCAGGATATAG (222 bp) and mTARC: 5' ATGAGGTCACTTCA GATGCT and 5' AGGTCACGGCCTTGGGTT TT (284 bp) mTARC: 5' GCTGCCGTCATTTTCTGCCT and 5' GCTAAACGCTTTCATTAAAT. IP-10: 5' GCTGCCGTCATTTTCTGCCT and 5' GCTAAACGCTTTCATTAAAT. β₂M was amplified from 1 μl (~20 ng) cDNA in a 24 cycle PCR program with cycling conditions as follows: 95°C for 30 seconds, 54°C for 30 seconds, 72°C for 30 seconds. Normalised template volumes were subjected to 31–35 PCR cycles using mTARC and IP-10 primers.

Real-time PCR was carried out with 2.5 μl (~50 ng) cDNA in multiplex reactions containing both mTARC and HPRT¹ (hypoxanthine guanine phosphoribosyl transferase) primers (at 0.3 μM each) and fluorogenic probes (at 0.2 μM each) in 25 μl final volumes of Universal MasterMix kit (PE Applied Biosystems). Primer and probe sequences were as follows: HPRT: 5' GACCGG TCCCGTCATG, probe ACCCGCAGTCCCAGCGTCGTG and 5' TCATAACCTGGTTCATC ATCGC and mTARC: 5' GGGATGCCATCGTGTTTCTG, probe CTGTCCAGGGCAAGCT CATCTGTGC and 5' GCCTTCTTCACATGTTTGTCTTTG. PCR and TaqMan analyses were performed using the ABI/PRISM 7700 sequence detector system (PE Applied Biosystems). Reactions were carried out in triplicate on 96-well plates with PCR cycling conditions: 50°C for 2 minutes, 95°C for 10 minutes, then 95°C for 15 seconds, 62°C for 1 minute for 50 cycles. Relative mTARC cDNA copy number was interpolated from a standard curve generated from cycle threshold values of serial dilutions of a known mTARC-containing cDNA sample and normalised to relative HPRT copy number.

C57BL/6 BIOgel-elicited Mθ were cultured in 1 ml supplemented OptiMEM, which was removed at specified time points, centrifuged at 16000 g at 4°C for 1 minute to remove cell debris and then frozen in aliquots at -70°C. ELISA was carried out with 50 μl each supernatant sample using the mouse TARC/CCL17 Quantikine kit (R&D Systems). Samples were analysed in duplicate at 450 nm. Levels of mTARC protein in unknown samples were quantified by interpolation from a standard curve generated in the same experiment.

Balb/c bone marrow Mθ were cultured for 7 days in L cell-conditioned medium. Cells were treated with 10 ng/ml recombinant murine IL-4 for the times indicated prior to formaldehyde fixation and sonication as described previously 86. Immunoprecipitation was performed using the M20 anti-STAT6 antibody (Santa Cruz) followed by reversal of cross-linking and PCR amplification of the proximal TARC promoter using the mTARC primers 5' GAGGTGACCTATTTAACCTCTC and 5' GAAGTAAGTCATTGCCCTTACC An M20 STAT6 blocking peptide (Santa Cruz) was used to confirm the specificity of the immunoprecipitation.

C57BL/6 thioglycollate-elicited Mθ were treated with 0, 0.5, or 1 μg/ml cycloheximide (BDH) for 2 hours prior to addition or not of 20 ng/ml recombinant murine IL-4. Cells were incubated at 37°C for 4 or 24 hours before RNA was harvested and processed for real-time PCR.

Mouse chromosome 8 sequence information was obtained from the Celera database and verified using the Sanger Centre database 52. Murine sequence was aligned with the homologous human chemokine cluster on chromosome 16q13 to search for conserved sequence amongst the human and murine TARC promoters. Sequences were analysed for potential transcription factor binding sites using Genomatix MatInspector 62.

Murine TARC promoter fragments were PCR amplified from 200 ng C57BL/6 genomic DNA sample using a 1:1 ratio of Herculase Enhanced DNA Polymerase (Stratagene) and Taq Polymerase (Gibco BRL) in 25 μl reactions containing 1% DMSO, 0.2 μM each dNTP and 100 ng/ml each primer. The reverse primer for full-length (-24 to -1173) and 5' deletion series mTARC promoter constructs was: 5' AACTAAGCTTTCTTCATGGGTCCTTCTGCCT. Forward primer sequences were as follows: -1173: 5' AACTACGCGTTCTGCTAGCACATG AGTGCAA, -839: 5' AACTACGCGTTGGGTATGCTGAAAATTCACA, -649: 5' AACTAC GCGTTTCCAACCCACTAAAGCACA, -498: 5' AACTACGCGTGCCT CTTCTACTGAAC A, -467: 5' AACTACGCGTTCACCCTCCTGTGGACCTTCT and -128: 5' AACTACGCGTT GGAAGGATTATAGGAGGGGA. Purified PCR products were digested with Mlu I and Hind III and cloned into pGL3 Basic (Promega), which contains the firefly luciferase gene.

Site-directed mutagenesis was carried out by 2-stage PCR using mTARC -1173 forward and reverse cloning primers noted above and forward and reverse primers designed to introduce 2-base pair mutations within selected mTARC STAT6 sites. The mutant primers were as follows (base changes underlined): T1: 5' AGTTATCTCAAA

AA

TT

AA

TT

AA

TT

AA

TT

AA

TT

AA

TT

Primers and oligonucleotides for 9XST6 and 5XST6 derivative constructs were designed to create 5' and 3' Mlu I sites for bi-directional cloning onto Emr1 -40 58 and EIF4A1 -40 63 minimal promoters in the pGL3 Basic vector (Promega). The minimal promoters were modified by addition of an Mlu I linker sequence (5' CACGCGTGGT AC) for ease of cloning. The reverse primer sequence for the 9XST6 and 5XST6 PCR reactions was: 5' AACTACGCGTAGAGAAATTCAAAGAATTTGAGA and the forward primers were: 9XST6: 5' AACTACGCGTCTCTTTTCTGGGCACTGTTGT and 5XST6: 5' AACTACGCG TTGGGTTCCCCAAAATAGGGT.

Trimerised mTARC STAT6 site constructs were generated by random self-ligation of 5'-phosphorylated, PAGE-purified trimers of the T1, T2, T4, T6 and T11 site oligonucleotides used as probes in the EMSA experiments (see below for sequences). These trimers were designed to create 5' and 3' Mlu I sites when annealed to form double-stranded DNA for cloning onto the Emr1 -40 minimal promoter (as described above).

All plasmids were sequenced and prepared using Qiagen or QBIOgene endotoxin-free kits to ensure levels of endotoxin contamination were ≤ 0.1 EU/μg. Transient transfections were carried out by electroporation of RAW 264.7 as described previously 33. Ten million RAW 264.7 cells were routinely co-transfected with 5 μg specific mTARC luciferase promoter construct DNA and 3 μg of a murine STAT6 expression vector (gift of Yoshihiro Ohmori 87). In all experiments, 1 μg β-galactosidase expression vector was co-transfected as a control for transfection efficiency. Cells were incubated for 16–18 hours post-transfection +/- 20 ng/ml recombinant murine IL-4 before whole cell extracts were harvested in 250 μl reporter lysis buffer (Promega). Transfection results are shown as Relative Luciferase Activity (luciferase/β-galactosidase) or fold induction of this value with IL-4 treatment, as appropriate.

C57BL/6 BIOgel-elicited Mθ were plated at l.5 × 10⁷ cells per 90 mm dish and treated +/-20 ng/ml either recombinant murine IL-4 or recombinant murine IFN-γ for 1 hour. Cells were washed and incubated at 4°C in PBS supplemented with 5 mM EDTA and 0.2 mM PMSF for 15 minutes. Nuclear extracts were prepared by addition to each dish of 1 ml nuclear extract buffer I containing 10 mM HEPES pH 7.9, 1.5 mM MgCl₂, 10 mM KC1, 0.5 mM DTT, 0.2 mM PMSF and additional protease inhibitors. Cells were incubated at 4°C for 15 minutes, then scraped into this buffer and vortexed gently. Cells were centrifuged at 2000 g for 30 seconds to recover the nuclei and the supernatant was discarded. Pellets were resuspended in 25 μl nuclear extract buffer II containing 20 mM HEPES pH 7.9, 25% glycerol, 420 nM NaCl, 0.5 mM DTT, 0.2 mM PMSF and protease inhibitors and left on ice for 20 minutes before being centrifuged at 16000 g for 15 minutes at 4°C. Supernatants from identical cytokine treatments were pooled and frozen at -70°C until used. Protein concentration of extracts was quantified using a BCA kit (Pierce) and 3 μg extract were used per EMSA assay.

Complementary forward and reverse oligonucleotides were designed for each mTARC putative STAT6 site. The forward sequences were as follows: T1 probe: 5' ATCTCAAATTCTTTGAATTTCTCTAG, T2 probe: 5' AGTGACCATTCTCTGGAAAGCCACAA, T3 probe: 5' TACTGAACATCCCTGTAACCTGCTCA, T4 probe: 5' TAGGCCTCTTCTACTGAACATCCCT G, T5 probe: 5' CCTCTGGGTTCCCCAAAATAGGGTGG, T6 probe: 5' GGCTCAGCTTCAGCTGAACCTCTGGG, T7 probe: 5' GTGCTCTATTTCTGGGAGCTGAGCTT, T8 probe: 5' CTATGGGCTTCTAGGGCAGCCTCTGG, T9 probe: 5' AAACTCTTTTCTGGGCACTGTTGTTA, T10 probe: 5' ATAGTGACTTGTCCTGAAACTCTTTT and T11 probe: 5' GGTTCTTTTTCTCAGTAAACCCTGGC. Forward and reverse oligonucleotides were annealed and end-labelled with ³²P (Amersham Pharmacia Biotech). Nuclear extracts were incubated with labelled oligonucleotides in sample buffer with final concentrations of 20 mM HEPES pH 7.9, 6.25% glycerol, 0.5 mM EDTA, 50 μg/ml poly dIdC, 50 μg/ml BSA and 12.5 μg/ml salmon sperm DNA for 15 minutes at room temperature. For supershift assays, 400 ng specific or control antibodies (Santa Cruz) were also added to the EMSA reactions. For cold competition assays, 100-fold molar excess amounts of unlabelled oligonucleotides were added to appropriate reactions. Samples were separated on a 7% polyacrylamide gel run at 250 V (20 mA) for 3 hours at 4°C and visualised after exposure to X-ray film.