Portal de Periódicos da CAPES

Definition of cis-Regulatory Elements of the Mouse Interleukin-5 Gene Promoter

1995; Elsevier BV; Volume: 270; Issue: 29 Linguagem: Inglês

10.1074/jbc.270.29.17541

1083-351X

Hyun Jun Lee, Esteban S. Masuda, Naoko Arai, Ken‐ichi Arai, Takashi Yokota,

Immune Response and Inflammation

We have previously reported that the promoter region of the mouse interleukin-5 (IL-5) gene, extending from a nucleotide position about −1,200 to +33 relative to the transcription initiation site, can mediate transcriptional stimulation by phorbol 12-myristate 13-acetate and dibutyryl cAMP (Bt2cAMP) in mouse thymoma EL-4 cells. Here, we describe identification of four cis-regulatory elements necessary for full activity of the IL-5 promoter, using deletion and mutation analyses. We designated these elements as IL-5A (−948 ∼−933), IL-5P (−117 ∼−92), IL-5C (−74 ∼−56), and IL-5CLE0 (−55 ∼−38). We found that IL-5P bears homology to the binding site for the nuclear factor of activated T cells (NF-AT) and interacted with protein factors in nuclear extracts prepared from EL-4 cells stimulated with phorbol 12-myristate 13-acetate and Bt2cAMP (designated NFIL-5P). NFIL-5P complex was inhibited in the presence of an excess NF-AT and AP1 oligonucleotides and supershifted by antisera raised against NF-ATp, c-Fos, and c-Jun. It thus seems likely that an NF-AT-related factor is involved in the regulation of IL-5 gene transcription. We have previously reported that the promoter region of the mouse interleukin-5 (IL-5) gene, extending from a nucleotide position about −1,200 to +33 relative to the transcription initiation site, can mediate transcriptional stimulation by phorbol 12-myristate 13-acetate and dibutyryl cAMP (Bt2cAMP) in mouse thymoma EL-4 cells. Here, we describe identification of four cis-regulatory elements necessary for full activity of the IL-5 promoter, using deletion and mutation analyses. We designated these elements as IL-5A (−948 ∼−933), IL-5P (−117 ∼−92), IL-5C (−74 ∼−56), and IL-5CLE0 (−55 ∼−38). We found that IL-5P bears homology to the binding site for the nuclear factor of activated T cells (NF-AT) and interacted with protein factors in nuclear extracts prepared from EL-4 cells stimulated with phorbol 12-myristate 13-acetate and Bt2cAMP (designated NFIL-5P). NFIL-5P complex was inhibited in the presence of an excess NF-AT and AP1 oligonucleotides and supershifted by antisera raised against NF-ATp, c-Fos, and c-Jun. It thus seems likely that an NF-AT-related factor is involved in the regulation of IL-5 gene transcription. Interleukin-5 (IL1 1The abbreviations used are: ILinterleukinGM-CSFgranulocyte-macrophage colony stimulating factorPMAphorbol 12-myristate 13-acetateBt2cAMPdibutyryl cAMPNF-ATnuclear factor of activated T cellsEMSAelectrophoretic mobility shift assayCsAcyclosporin ACHXcycloheximideThhelper Tbpbase pair(s)kbkilobaseCLEconserved lymphokine elementPCRpolymerase chain reactionCREcAMP response element. -5) is primarily a T cell-derived lymphokine with multiple regulatory functions, including stimulation of growth and differentiation, on eosinophils (1Yokota T. Coffman R.L. Hagiwara H. Rennick D.M. Takebe Y. Yokota K. Gemmell L. Shrader B. Yang G. Meyerson P. Luh J. Hoy P. Pene J. Briere F. Spits H. Banchereau J. de Vries J. Lee F.D. Arai N. Arai K. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7388-7392Crossref PubMed Scopus (189) Google Scholar, 2Sanderson C.J. Blood. 1992; 79: 3101-3109Crossref PubMed Google Scholar) and, in the mouse, on B cells(3Takatsu K. Tominaga A. Harada N. Mita S. Matsumoto M. Takahashi T. Kikuchi Y. Yamaguchi N. Immunol. Rev. 1988; 102: 107-135Crossref PubMed Scopus (222) Google Scholar). The IL-5 gene is located on mouse chromosome 11 and human chromosome 5, within a cluster of the IL-3, granulocyte-macrophage colony stimulating factor (GM-CSF), IL-4, and IL-13 genes (4Lee J.S. Campbell H.D. Kozak C.A. Young I.G. Somat. Cell. Mol. Genet. 1989; 15: 143-152Crossref PubMed Scopus (45) Google Scholar, 5van Leeuwen B.H. Martinson M.E. Webb G.C. Young I.G. Blood. 1989; 73: 1142-1148Crossref PubMed Google Scholar, 6Morgan J.G. Dolganov G.M. Robbins S.E. Hinton L.M. Lovett M. Nucleic Acids Res. 1992; 20: 5173-5179Crossref PubMed Scopus (178) Google Scholar) which are commonly expressed in T helper 2 (Th2) cells. interleukin granulocyte-macrophage colony stimulating factor phorbol 12-myristate 13-acetate dibutyryl cAMP nuclear factor of activated T cells electrophoretic mobility shift assay cyclosporin A cycloheximide helper T base pair(s) kilobase conserved lymphokine element polymerase chain reaction cAMP response element. Mouse Th cells have been classified into two subsets on the basis of their distinct lymphokine production patterns(7Mosmann T.R. Cherwinski H. Bond M.W. Giedlin M.A. Coffman R.L. J. Immunol. 1986; 136: 2348-2357Crossref PubMed Google Scholar). Th1 cells produce IL-2, interferon γ, and lymphotoxin and promote cell-mediated immunity, whereas Th2 cells produce IL-4, IL-5, IL-6, and IL-10 and promote humoral immunity(8Mosmann T.R. Immunol. Res. 1991; 10: 183-188Crossref PubMed Scopus (105) Google Scholar). Polarized Th1 and Th2 responses have been shown to be a feature of a number of disease states in both mice and humans, and a proper balance between the two subsets is crucial for the effective responses(9Scott P. Kaufmann S.H. Immunol. Today. 1991; 12: 346-348Abstract Full Text PDF PubMed Scopus (336) Google Scholar, 10Sher A. Coffman R.L. Annu. Rev. Immunol. 1992; 10: 385-409Crossref PubMed Scopus (818) Google Scholar, 11Romagnani S. Annu. Rev. Immunol. 1994; 12: 227-257Crossref PubMed Google Scholar). Despite their physiological significance, molecular mechanisms which account for the difference in lymphokine gene expression in Th1 and Th2 cells are not well understood. One mechanism may entail differences in transcription factors interacting with cis-acting elements in lymphokine genes. Another mechanism may involve differences in signal transduction pathways between Th1 and Th2 cells(12Arai N. Naito Y. Watanabe M. Masuda E.S. Yamaguchi-Iwai Y. Tsuboi A. Heike T. Matsuda I. Yokota K. Koyano-Nakagawa N. Lee H.J. Muramatsu M. Yokota T. Arai K. Pharmacol. Ther. 1992; 55: 303-318Crossref PubMed Scopus (27) Google Scholar). There are reports suggesting that Th1 and Th2 cells use different signal transduction pathways; several groups have demonstrated that compounds which elevate intracellular cAMP levels have differential effects on lymphokine production in Th1 and Th2 cells(13Li T.K. Fox B.S. J. Immunol. 1993; 150: 1680-1690PubMed Google Scholar, 14Munoz E. Zubiaga A.M. Merrow M. Sauter N.P. Huber B.T. J. Exp. Med. 1990; 172: 95-103Crossref PubMed Scopus (249) Google Scholar, 15Betz M. Fox B.S. J. Immunol. 1991; 146: 108-113PubMed Google Scholar). We have previously reported that cAMP has differential effects on the production of Th1- and Th2-type lymphokines in EL-4 cells, a cell line which produces both types of lymphokines when stimulated with PMA(16Lee H.J. Koyano-Nakagawa N. Naito Y. Nishida J. Arai N. Arai K. Yokota T. J. Immunol. 1993; 151: 6135-6142PubMed Google Scholar). Effects of cAMP on PMA-induced expression of IL-2 and IL-5 genes were opposite; i.e. cAMP activated the IL-5 gene synergistically with PMA, while it suppressed PMA-induced expression of the IL-2 gene, by modulating respective promoter activities. Mechanisms for differential regulation of the IL-2 and IL-5 genes by cAMP in EL-4 cells may be related to those in Th1/Th2 cells and should provide some insight at the molecular level. Transcriptional control of the IL-2 gene has been extensively studied, and the major regulatory elements have been shown to reside within a region of approximately 300 bp upstream of the transcription initiation site(17Serfling E. Barthelmas R. Pfeuffer I. Schenk B. Zarius S. Swoboda R. Mercurio F. Karin M. EMBO J. 1989; 8: 465-473Crossref PubMed Scopus (196) Google Scholar, 18Ullman K.S. Northrop J.P. Verweij C.L. Crabtree G.R. Annu. Rev. Immunol. 1990; 8: 421-452Crossref PubMed Scopus (491) Google Scholar). Of several transcription factors involved in regulation of the IL-2 gene, NF-AT is essential for transcription of the IL-2 gene upon T cell activation and plays a major role in various aspects of regulation of the IL-2 gene, including T cell-specific expression(19Shaw J.P. Utz P.J. Durand D.B. Toole J.J. Emmel E.A. Crabtree G.R. Science. 1988; 241: 202-205Crossref PubMed Scopus (10) Google Scholar). NF-AT has also been implicated in the transcriptional regulation of other lymphokine genes, such as GM-CSF, IL-3, IL-4, and tumor necrosis factor-α, hence, a role in the coordinated expression of these genes was suggested(20Masuda E.S. Naito Y. Arai N. Arai N. Immunologist. 1993; 1: 198-203Google Scholar, 21Rao A. Immunol. Today. 1994; 15: 274-281Abstract Full Text PDF PubMed Scopus (490) Google Scholar). We also have data that NF-AT is a major target of the inhibitory action of cAMP on the IL-2 promoter(22Tsuruta L. Lee H.J. Masuda E.M. Koyano-Nakagawa N. Arai K. Arai N. Yokota T. J. Immunol. 1995; 154: 5255-5264PubMed Google Scholar). On the other hand, details on the transcriptional control of the IL-5 gene have not been explored at the molecular level. Within the 5'-flanking region, the IL-5 gene is associated with the conserved lymphokine elements (CLE)0, CLE1, and CLE2 which are also found in the promoters of IL-3, GM-CSF, and IL-4 genes, and may have a role in the regulation of coordinated lymphokine expression (23Arai K. Lee F. Miyajima A. Miyatake S. Arai N. Yokota T. Annu. Rev. Biochem. 1990; 59: 783-836Crossref PubMed Scopus (1176) Google Scholar, 24Miyatake S. Shlomai J. Arai K. Arai N. Mol. Cell. Biol. 1991; 11: 5894-5901Crossref PubMed Google Scholar) (Fig. 1). In this study, we identified four cis-regulatory elements which are necessary for full activity of the IL-5 promoter in response to PMA and Bt2cAMP stimulation in the mouse thymoma line EL-4. We showed that one element, the IL-5P element, shares homology with the binding site of NF-AT of the IL-2 gene and interacts with the inducible nuclear factor NFIL-5P which is closely related to the transcription factor NF-AT. The mouse thymoma cell line EL-4 TB6 was maintained in RPMI 1640 medium supplemented with 2 mM glutamine, 50 units/ml penicillin, 50 εg/ml streptomycin, 50 εM 2-mercaptoethanol, and 5% fetal bovine serum under 5% CO2. Jurkat, Ba/F3, and COS-7 cells were cultured in the same medium containing 10% fetal bovine serum. Recombinant mIL-3 (1 ng/ml) was additionally supplemented for maintenance of Ba/F3 cells. NIH3T3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and the antibiotics described above. Anti-mouse NF-ATp (αNF-ATp), a rabbit polyclonal antiserum raised against recombinant NF-ATp, was purchased from Upstate Biotechnology Incorporated. The Fos antibody (αFos), a rabbit polyclonal IgG reactive with c-Fos, Fos B, Fra-1, and Fra-2 and Jun antibody (αJun), a rabbit polyclonal IgG reactive with c-Jun, Jun B, and Jun D and their cognate peptides purchased from Santa Cruz Biotechnology. The DNA sequence of both strands of the region extending from position −1174 to −545 was determined using an Applied Biosystems 373A DNA sequencer. Plasmids pUC00Luc and pmIL5Luc(1.2) were described previously(16Lee H.J. Koyano-Nakagawa N. Naito Y. Nishida J. Arai N. Arai K. Yokota T. J. Immunol. 1993; 151: 6135-6142PubMed Google Scholar). pmIL5Luc(1.2) was linearized by SalI digestion and a series of nested 5' deletions was generated by BAL-31 exonuclease treatment. Then, 7-bp SalI linkers were attached and the SalI-HindIII fragments were inserted into pUC00Luc. The 5' border of each deletion mutant was determined by DNA sequencing and was named according to its 5'-end point. To generate pmIL5Luc(1.8), the 5'-EcoRI site of the pmIL5EHE (16Lee H.J. Koyano-Nakagawa N. Naito Y. Nishida J. Arai N. Arai K. Yokota T. J. Immunol. 1993; 151: 6135-6142PubMed Google Scholar) was replaced by the SalI site and the 1.8-kb SalI-HindIII fragment was inserted into pUC00Luc. The linker-scanning mutants were generated by replacing nucleotide sequences in various positions in the 1.2-kb IL-5 promoter with a 7-bp SalI or 6-bp XhoI linker sequences, using polymerase chain reaction (PCR). The resulting linker-scanning mutants contained 3-7-bp substitutions out of 6 or 7 bp at various positions in the context of the 1.2-kb IL-5 promoter (Fig. 3). These mutants were named according to the positions at which substitution linkers were introduced. To make the linker-scanning mutants LS971/965, PCR was performed with pmIL5Luc(1.2) as the template, M13 universal primer ((M13(−47)), and 1SA, an antisense strand oligonucleotide primer containing a SalI recognition site at nucleotide position −972 (Table 1). The resulting PCR product was digested with SacI (at the polylinker region) and SalI and was inserted into the SacI-SalI site of p964, a deletion mutant with a 5'-end point of −964 (Fig. 2). LS946/940, LS939/933, LS110/104, LS103/97, LS91/85, LS69/63, and LS57/51 were generated by essentially the same procedure. For the latter five mutants, p161, a deletion mutant whose 5'-end point is −161, was used as the template to minimize PCR-derived sequence and a sequence further upstream was attached using the SacI-NsiI (at nucleotide position −141) fragment of pmIL5Luc(1.2). To obtain LS954/949 and LS932/927, a pair of PCR amplifications was performed for each construct, as described(25Ausubel F.M. Brent R.E. Kingston D.M. Moor J.G. Seidman J.A. Sttruhl K. Current Protocols in Molecular Biology. John Wiley and Sons, New York1991Google Scholar). Fragments upstream of the introduced mutations were generated using pmIL5Luc(1.2) as the template and M13(−47), and primers containing the XhoI recognition sites at the 5'-ends of the antisense strand oligonucleotides (Table 1). To generate downstream fragments, we used sense-strand oligonucleotide primers containing the XhoI linker sequence at the 5'-ends to overlap with the XhoI site of the upstream fragment, and antisense-strand oligonucleotide primer C1A (Table 1). The resulting upstream and downstream fragments were digested with SacI and XhoI, XhoI and BclI (at nucleotide position −800), respectively, and ligated between the SacI and BclI sites of pmIL5Luc(1.2). LS123/118, LS117/112, LS80/74, LS62/57, LS37/32 were generated essentially by the same procedure used for LS955/949 and LS933/927. For these constructions, we used p161 as the template and GLprimer2, a primer corresponding to the antisense strand of the luciferase structural gene (Promega Corp., Madison, WI) instead of C1A to generate downstream fragments. The resulting PCR fragments were cloned into the NsiI and HindIII (at nucleotide position +33) sites of pmIL5Luc(1.2). To generate pM1 to pM4 containing mutations in the IL-5P site, two steps of PCR were performed as described(25Ausubel F.M. Brent R.E. Kingston D.M. Moor J.G. Seidman J.A. Sttruhl K. Current Protocols in Molecular Biology. John Wiley and Sons, New York1991Google Scholar). First, fragments upstream and downstream of point mutations were generated, using the same procedure as for LS124/118, except for use of sense and antisense oligonucleotides (Table 2, PM1 to PM4) used in the electrophoretic mobility shift assays (EMSAs) instead of primers containing the XhoI linker sequence. Second step PCR amplifications were done using one-tenth of the fragments generated by first step PCR and M13(−47) and GLprimer2. The resulting fragments were inserted into the NsiI and HindIII sites of pmIL5Luc(1.2). To increase fidelity, PCR amplifications were performed for 15 cycles (four cycles with annealing at 45°C and 11 cycles at 65°C) with a large amount (100 ng) of template. The PCR-derived sequences of each construct were confirmed by DNA sequencing. The expression plasmids of NF-ATc and NF-ATx were constructed by inserting corresponding cDNAs downstream of the SRα promoter in pME18S(26Masuda E.S. Naito Y. Tokumitsu H. Campbell D. Saito F. Hannum C. Arai K. Arai N. Mol. Cell. Biol. 1995; 15: 2697-2706Crossref PubMed Scopus (198) Google Scholar).Tabled 1 Open table in a new tab Figure 2:Deletion analysis of the 5'-upstream region of the mouse IL-5 promoter. EL-4 cells were transfected with the serial deletion mutants shown schematically in the left panel. The cells were stimulated 24 h after transfection with either PMA (solid bars) or PMA and Bt2cAMP (cross-hatched bars), cell lysates were prepared 16 h later, and tested for luciferase activity, as described under "Materials and Methods." Results represent the mean value of duplicate transfections. Two other independent experiments gave similar results. The regions responding to PMA and Bt2cAMP are indicated by shadowed boxes.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Tabled 1 Open table in a new tab Transient transfections were done using a DEAE-dextran procedure, as described previously(16Lee H.J. Koyano-Nakagawa N. Naito Y. Nishida J. Arai N. Arai K. Yokota T. J. Immunol. 1993; 151: 6135-6142PubMed Google Scholar). At 24 h after transfection, cells were stimulated with 1 mM Bt2cAMP (Sigma) and 10 ng/ml PMA (Calbiochem). After another 16 h, the cells were harvested for luciferase assays. Luciferase activity was determined using the luciferase assay substrate (Promega) with a luminometer (Berthold, Postfach, Germany). Protein concentration in an aliquot of each sample was measured using BCA reagents (Pierce), and the results were calculated as (relative light unit-background)/εg protein. Cells were stimulated for 3 h with various drugs, as described in figure legends, and nuclear extracts were prepared either by the method of Dignam et al.(27Dignam J.D. Levovitz R.M. Roeder R.G. Nucleic Acids Res. 1983; 11: 1475-1489Crossref PubMed Scopus (9164) Google Scholar) or as described(28Masuda E.S. Tokumitsu H. Tsuboi A. Shlomai J. Hung P. Arai K. Arai N. Mol. Cell. Biol. 1993; 13: 7399-7407Crossref PubMed Scopus (126) Google Scholar). EMSAs were done using the double-stranded oligonucleotides given in Table 2. Each single-stranded oligonucleotide was purified on denaturing polyacrylamide gels before annealing. The annealed oligonucleotides were 32P-labeled with Klenow fragment and purified on a 12% polyacrylamide gels. The DNA-binding reactions were performed at room temperature for 30 min with 3-5 εg of nuclear extracts, 0.5 εg of poly(dI-dC), 10 mM HEPES, pH 7.9, 10% glycerol, 1 mM EDTA, 1 mM dithiothreitol, 100 mM KCl, and 50,000 counts/min of probe (∼0.5 ng) in a total volume of 10 εl. The samples were resolved on a 4% nondenaturing polyacrylamide gel at 120 V in 1 × Tris-glycine-EDTA buffer, and the results were visualized by autoradiography. Extracts from transfected COS-7 cells were prepared as described(26Masuda E.S. Naito Y. Tokumitsu H. Campbell D. Saito F. Hannum C. Arai K. Arai N. Mol. Cell. Biol. 1995; 15: 2697-2706Crossref PubMed Scopus (198) Google Scholar). We have previously reported that the 5'-flanking region of the IL-5 promoter extending from the SphI site at a nucleotide position about −1,200 to +33 relative to the transcription initiation site mediated induction by PMA and Bt2cAMP in EL-4 cells(16Lee H.J. Koyano-Nakagawa N. Naito Y. Nishida J. Arai N. Arai K. Yokota T. J. Immunol. 1993; 151: 6135-6142PubMed Google Scholar). Since only a partial nucleotide sequence has been reported, we first determined the complete nucleotide sequence of this region and found a number of potential regulatory elements (Fig. 1). Next, to identify the region responsible for inducible expression of the IL-5 promoter in response to PMA and Bt2cAMP, the luciferase reporter plasmids carrying sequential 5'-deletions of the mouse IL-5 promoter were transiently transfected into EL-4 cells. The plasmid pUC00Luc, which has no upstream sequence derived from the IL-5 gene(16Lee H.J. Koyano-Nakagawa N. Naito Y. Nishida J. Arai N. Arai K. Yokota T. J. Immunol. 1993; 151: 6135-6142PubMed Google Scholar), showed an almost negligible response to PMA or a combination of PMA and Bt2cAMP (Fig. 2). Removal of the 5' sequence up to position 964 did not affect the promoter activity responding to PMA and Bt2cAMP, but deletions beyond this position through −932 did reduce the response by 50%. Further deletions up to position −140 led to no further reduction. Remarkably, the subsequent removal of nucleotides −140 through −80 diminished the promoter activity by 90%. Thus, the response to a combination of PMA and Bt2cAMP appears to decline in two stages, at the −964 to −933 (region I) and at the −140 to −80 (region II) (Fig. 2). On the other hand, the IL-5 promoter responded weakly to PMA alone, and removal of the 5' sequences up to position −140 did not affect promoter activity responding to PMA, while deletions beyond this position almost completely abolished the response. Thus, the sequences between −140 to −80 (region II) of the IL-5 promoter are required for a response, albeit a weak one, to the PMA signal. To further delineate regulatory elements in the IL-5 promoter, we dissected regions shown to be important for promoter activity. Using the linker-scanning method, we replaced the sequence in various positions between −972 and −932, and between −124 and −31 in the 1.2-kb IL-5 promoter with a 7-bp SalI or 6-bp XhoI linker sequence (see "Materials and Methods"). We did not introduce mutations around −140, because we found that in further deletion analysis, the proximal regulatory region is located between −127 and −103 (data not shown). Consistent with results of the 5'-deletion analysis, mutations introduced between −948 and −933 (designated IL-5A) reduced promoter activity by 60% in response to PMA and Bt2cAMP but not to PMA alone (Fig. 3). Mutations introduced in the region extending from −117 to −92 (designated as IL-5P) reduced promoter activity by about 80%, in response to a combination of PMA and Bt2cAMP as well as to PMA alone. Mutation analyses allowed for definition of third and fourth elements, extending from −74 to −56 (designated IL-5C) and from −55 to −38 (designated IL-5CLE0), respectively. The IL-5C and IL-5CLE0 elements are located in close proximity, and mutation analysis has heretofore shown no clear boundary between them. However, we defined IL-5CLE0 in analogy with CLE0 of the GM-CSF gene, and the remaining sequence as IL-5C. These two elements are the most crucial for promoter activity, because mutations in these regions almost completely abolished promoter activity in response to PMA as well as to PMA and Bt2cAMP. Deletion and mutation analyses revealed that downstream sequences of −124 are critical for IL-5 promoter activity. Comparison of the mouse and human IL-5 promoters showed a high degree of homology within this region (Fig. 4A). Interestingly, closer inspection revealed that IL-5P is homologous to the binding site for NF-AT (Fig. 4B). NF-AT has also been implicated in transcriptional regulation of other lymphokine genes(20Masuda E.S. Naito Y. Arai N. Arai N. Immunologist. 1993; 1: 198-203Google Scholar, 21Rao A. Immunol. Today. 1994; 15: 274-281Abstract Full Text PDF PubMed Scopus (490) Google Scholar). We recently noted that the NF-AT site is a major target of the inhibitory action of cAMP on the IL-2 promoter(22Tsuruta L. Lee H.J. Masuda E.M. Koyano-Nakagawa N. Arai K. Arai N. Yokota T. J. Immunol. 1995; 154: 5255-5264PubMed Google Scholar). Thus, we speculate that IL-5P may be involved in coordinated and differential regulation of the IL-5 gene and focused on further characterization of IL-5P. To identify nuclear factors that can mediate inducibility through IL-5P, EMSAs were done using a probe corresponding to the region extending from −119 to; 89 of the IL-5 promoter (Table 2). Nuclear extracts prepared from EL-4 cells stimulated with a combination of PMA and Bt2cAMP formed several complexes with the IL-5P probe (Fig. 5A). The major complex (designated NFIL-5P) appeared only when we used extracts prepared from stimulated cells (compare lanes 1 and 2), and this complex formation was specifically inhibited in a dose-dependent manner by excess unlabeled IL-5P oligonucleotide (lanes 2-5) but not by the Sp1 oligonucleotide (lanes 6-8). Other minor complexes with different mobilities could be detected, but appearance was variable, depending on different nuclear extract preparations, and they did not seem to be specific for IL-5P because binding was also inhibited by the Sp1 oligonucleotide (lanes 6-8). When EMSAs were done with nuclear extracts from EL-4 cells stimulated with different combinations of PMA and Bt2cAMP, the NFIL-5P complex was induced by either PMA or Bt2cAMP. The combination of both agents intensified the band (Fig. 5B, lanes 2-4). Interestingly, the NFIL-5P complex migrated with a different mobility, depending on the stimulation given. The control Sp1 binding was not altered by any of these stimuli (lanes 5-8). To identify specific bases critical for NFIL-5P complex formation, segments of IL-5P sequence were mutated and tested for effects on complex formation (Fig. 6A). The PM1 mutant oligonucleotide with a mutation in the GGA sequence did not form the NFIL-5P complex, but did retain the ability to form other minor complexes (Fig. 6B, lane 4). NFIL-5P binding was gradually restored as mutations were introduced into the segments closer to the 3' portion of the IL-5P sequence (compare lanes 6, 9, 12, and 15). Competition analyses gave consistent results (data not shown). To search for a possible correlation between the in vitro formation of the NFIL-5P complex and in vivo activity of the IL-5 promoter, we introduced site-specific mutations corresponding to mutations introduced into the IL-5P probe for the EMSAs described above, in the context of the 1.2-kb IL-5 promoter. Constructs were transiently transfected into EL-4 cells and analyzed for luciferase activity (Fig. 6C). The results are shown along with the intensity of the NFIL-5P band elicited by each probe (Fig. 6B), as determined by PhosphorImage analyzer (Molecular Dynamics). Introduction of PM1, PM2, PM3, and PM5 mutations decreased IL-5 promoter activity by about 80%, in response to PMA and Bt2cAMP, whereas the PM4 mutation caused only a 50% decrease in promoter activity. Therefore, the in vivo activity of the IL-5 promoter correlated with the binding activity of NFIL-5P to the IL-5P site, i.e. the bases which were required for formation of the NFIL-5P complex in vitro were the same as those required for promoter activity. These results strongly suggest that NFIL-5P is the factor which regulates the IL-5 promoter through the IL-5P sequence. The IL-5 gene is expressed mainly in activated T cells. To determine the cell-type specificity of the NFIL-5P complex, EMSAs were done using nuclear extracts from a variety of cell types either unstimulated or stimulated with PMA, A23187, and Bt2cAMP (Fig. 7). The NFIL-5P complex appeared only in the stimulated EL-4, Jurkat T cell line, and pro-B cell line Ba/F3 cells (left panel, lanes 2, 4, and 6, respectively), while the AP1 complex was detected in all cell types tested (right panel). We next asked whether NFIL-5P and NF-AT were related. NF-AT is composed of a pre-existing cytoplasmic component, referred to as NF-ATp (29McCaffrey P.G. Perrino B.A. Soderling T.R. Rao A. J. Biol. Chem. 1993; 268: 3747-3752Abstract Full Text PDF PubMed Google Scholar) or NF-ATc(30Flanagan W.M. Corthesy B. Bram R.J. Crabtree G.R. Nature. 1991; 352: 803-807Crossref PubMed Scopus (954) Google Scholar), which is translocated into the nucleus in response to a calcium signal, and a newly synthesized nuclear component, which belongs to the AP1 family (30Flanagan W.M. Corthesy B. Bram R.J. Crabtree G.R. Nature. 1991; 352: 803-807Crossref PubMed Scopus (954) Google Scholar, 31Jain J. McCaffrey P.G. Valge A.V. Rao A. Nature. 1992; 356: 801-804Crossref PubMed Scopus (429) Google Scholar). In EMSAs using nuclear extracts from EL-4 cells stimulated with PMA and Bt2cAMP, NFIL-5P and NF-AT complexes migrated with a similar mobility (Fig. 8A, lanes 1 and 7). NFIL-5P complex formation was completely inhibited by excess unlabeled NF-AT as well as by IL-5P oligonucleotides (lanes 2 and 5) but not by those mutated in GGA sequences (lanes 3 and 6). As expected, AP1 oligonucleotide also inhibited NFIL-5P complex formation (lane 4). Conversely, NF-AT binding was inhibited specifically by both excess unlabeled NF-AT and IL-5P oligonucleotides, indicating that NFIL-5P is an NF-AT-related complex (lanes 7-11). NF-AT binding is known to be inhibited by either cyclosporin A (CsA) (30Flanagan W.M. Corthesy B. Bram R.J. Crabtree G.R. Nature. 1991; 352: 803-807Crossref PubMed Scopus (954) Google Scholar) or the protein synthesis inhibitor cycloheximide (CHX)(19Shaw J.P. Utz P.J. Durand D.B. Toole J.J. Emmel E.A. Crabtree G.R. Science. 1988; 241: 202-205Crossref PubMed Scopus (10) Google Scholar, 32McCaffrey P.G. Jain J. Jamieson C. Sen R. Rao A. J. Biol. Chem. 1992; 267: 1864-1871Abstract Full Text PDF PubMed Google Scholar). We then tested sensitivity of the NFIL-5P complex to CsA and CHX; treatment of CsA inhibited the induction of NFIL-5P complex in cells stimulated with PMA and Bt2cAMP, in a dose-dependent manner similar to that seen for NF-AT binding (Fig. 8B, top and central panels, lanes 2 and 3). Induction of both the NFIL-5P and NF-AT complexes was almost completely inhibited by treatment with 10 εg of CHX (lanes 4 and 5). This inhibition was not caused by nonspecific effects of these reagents on nuclear proteins, since there was no change in Sp1 binding, under the same conditions (bottom panel). To determine whether an NF-AT component is involved in the NFIL-5P complex, supershift EMSAs were done (Fig. 8C). Both NFIL-5P and NF-AT complexes (la