Identification of IFN Regulatory Factor-1 Binding Site in IL-12 p40 Gene Promoter¹ Free

Interleukin-12 is a key cytokine that functions to protect the host from viral and microbial infections (1, 2). IL-12 is a heterodimer composed of p40 and p35. The former is induced by microbial infection, whereas the latter is constitutively expressed (2, 3). Macrophages and NK cells function to link the innate immune system with the acquired immune system during viral or bacterial infection (4, 5, 6, 7). This linkage is mediated mainly by IFN-γ and IL-12. Bacterial stimuli activate macrophages and subsequently NK cells in the innate immune response to produce IL-12 and IFN-γ, respectively (8). IL-12 subsequently induces NK cells to produce IFN-γ (9, 10, 11), which in turn activates macrophages to present Ags to Ag-specific T cells (12). The involvement of IL-12 in type 1 T cell differentiation is well established in many systems (8, 13, 14, 15, 16, 17). This type of innate immune response and its accompanying Ag-specific T cell response work to eradicate microbial pathogens (1, 4, 5).

The induction of the IL-12 p40 gene in macrophages is regulated by NF-κB in the presence of IFN-γ (3, 18). Signals from IFN-γ receptors are initially transduced by IFN-stimulated gene factor (ISGF)⁶-3 and IFN-γ-activated factor (GAF). Transcription factors such as IFN regulatory factor-1 (IRF-1) are used for further proper regulation of the broad range of genes induced by IFN, e.g., inducible NO and IL-1β converting enzyme (19, 20, 21, 22). In a previous study, we demonstrated that IRF-1 gene-disrupted mice are defective in inducing IL-12 p40 mRNA, suggesting that gene expression is regulated by transcription factor IRF-1 (23, 24). However, the binding site of IRF-1 to the promoter region of the IL-12 p40 gene has not yet been formally determined. We further analyzed within this study the regulation of IL-12 p40 gene expression and showed that IRF-1 directly binds to the IL-12 p40 gene promoter and up-regulates gene expression. Although IL-12 p35 is thought to be constitutively expressed, we show the IL-12 p35 gene is also inducible by LPS stimulation, even in the absence of IFN-γ, and is regulated differently from the IL-12 p40 gene.

Recombinant murine IFN-γ was purchased from Genzyme (Cambridge, MA). Bacterial LPS was purchased from Sigma-Aldrich (Tokyo, Japan). Ab specific for IRF-1 protein was purchased from Santa Cruz Biotechnology (Santa Cruz, CA).

C57BL/6 mice were purchased from Charles River Japan (Yokohama, Japan). IRF-1-deficient mice have been described earlier (25) and were maintained by backcrossing to C57BL/6 mice. ISGF3γ-deficient (p48) mice have been described previously (26). The littermates of each mutant strain were used as control mice. These mutant mice and their littermates were reared under specific pathogen-free conditions in the animal facility of either the University of Tokyo or Ehime University School of Medicine. All mice were used in accordance with our institutional guidelines for animal experimentation.

Cells were cultured at 37°C in a 5% CO₂ humidified air atmosphere. The medium used was RPMI 1640 supplemented with 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, 1-time nonessential amino acid, 50 μM 2-ME, and 10% heat-inactivated FCS.

PECs taken from mice treated with thioglycolate 3 days earlier were allowed to adhere to tissue culture plates for 1 h, and nonadherent cells were removed. Adherent cells were cultured in the presence of medium alone, LPS (10 μg/ml), or LPS plus recombinant mouse IFN-γ (100 U/ml). Total RNA was extracted 16 h later and subjected to Northern blotting with probes for IL-12 p40 (a gift from Dr. H. Yamamoto, Osaka University) and TNF-α (a gift from Dr. T. Yokota, University of Tokyo).

Total cellular RNA was isolated by the guanidinium-thiocyanate method. The procedure for Northern blot analysis is described by Harada et al. (27). To prepare DNA probes, fragments of TNF-α, IL-12 p35, IL-12 p40, and β-actin cDNA were labeled by the random primer method. The specific activity of the IL-12 p35 probe and that of the IL-12 p40 probe were comparable. The intensity of the band was measured by BAS2000 (Fuji Film, Tokyo, Japan).

Cos7 cells were resuspended in 10% FCS-containing RPMI 1640 medium at 1 × 10⁶ cells/ml. Further, 10 μg of plasmid DNA was mixed with Cos7 cells in a 1-ml electrochamber (PKG/36; Life Technologies, Gaithersburg, MD). DNA was electroporated to Cos7 cells at 750 V/cm and 880 μF using the Electroporation System I (Life Technologies). The cells for luciferase assay were harvested 60 h after transfection. The same amount of cell extract was used in each experiment.

Electrophoretic mobility shift analysis was done as described by Harada et al. (27). A total of 2 μl of in vitro-translated IRF-1 were incubated with a ³²P-labeled DNA probe in a buffer containing 10 mM Tris-HCl (pH 7.6), 50 mM NaCl, 1 mM DTT, 1 mM EDTA, and 5% glycerol at 25°C for 60 min. Protein-DNA complexes were analyzed by 4% polyaclylamide gel electrophoresis.

A PCR fragment of −133 to +51 in the IL-12 p40 promoter region was labeled with ³²P by incubating with T4 kinase and followed by digestion with SacI and HindIII. The sense strand was mixed with rIRF-1 protein and incubated at 30°C for 1 h. The mixture was digested with DNase I (0.2 U/μl) at 25°C for 1 min, treated with phenol, and precipitated with ethanol. The pellet was dissolved in a loading buffer and analyzed on an 8% polyacrylamide-8 M urea gel (28).

PECs were stimulated with LPS (10 μg/ml) plus recombinant mouse IFN-γ (100 U/ml). At 0 and 2 h later, formaldehyde solution was added directly to the culture at a final concentration of 1%. Cross-linking of proteins on chromatin was allowed to occur at 37°C for 10 min, and the cells were lysed by Nonidet P-40 lysis buffer with protease inhibitors. Chromatin in the lysate was sonicated to an average length of 200–500 bp as determined by agarose gel electrophoresis. IRF-1 protein binding to chromatin was immunoprecipitated, washed, and eluted. Cross-links were reversed by 0.5 M NaCl. After proteinase K digestion, DNA in samples was phenol extracted, ethanol precipitated, and used for PCR amplication to detect IL-12 p40 promoter segment.

We first examined the requirement of IFN-γ for IL-12 p35 and IL-12 p40 gene expression. Although the IL-12 p40 gene is thought to be inducible and IL-12 p35 gene is constitutively expressed, both genes are induced by stimulation with LPS plus IFN-γ (Fig. 1, A and B). However, the requirement for IFN-γ is distinctly different in the two genes. The induction of IL-12 p40 is minimal in the absence of IFN-γ and is enhanced by the presence of IFN-γ, confirming previous findings (2). In contrast, the effect of IFN-γ on IL-12 p35 gene induction is minimal. Therefore, IL-12 p70 production is profoundly influenced by the presence of IFN-γ.

Because the signal from the IFN-γ receptor is transduced by IRF-1 and ISGF3γ (p48), IL-12 p40 and p35 gene induction with LPS plus IFN-γ was evaluated by Northern blot analysis using thioglycorate-induced peritoneal adherent cells from IRF-1^−/− mice, p48^−/− mice, and their littermates as wild-type mice (Fig. 1 C). Costimulation with LPS and IFN-γ induced the IL-12 p40 gene in wild-type mice, whereas the same stimulation failed to induce the gene in IRF-1^−/− mice. Gene induction in p48^−/− mice was achieved by costimulation with LPS plus IFN-γ, and the induction level was comparable to their wild-type littermates. This result suggests that IRF-1 but not p48 is responsible for the regulation of the IL-12 p40 gene. In contrast with IL-12 p40 gene induction, neither IRF-1 nor p48 is required for IL-12 p35 gene induction. The result proves that induction of the IL-12 p40 and p35 genes is differentially regulated by IFN-γ via IRF-1 gene induction. The former is IRF-1 dependent and the latter is IRF-1 independent.

To determine the IRF-1 regulatory region of the IL-12 p40 gene, the promoter region in the published DNA sequence at −677 to +51 of the IL-12 p40 gene (18, 29) was obtained by PCR using an IL-12 p40 genomic clone of C57BL/6 origin as a template. The DNA region obtained was inserted into the 5′-upstream region of a luciferase gene lacking the promoter region (basic vector, pLuc) and the reporter plasmid (p40–677-pLuc) was constructed. RAW264.7 cells were transfected with this plasmid. The transfectant was stimulated with LPS plus IFN-γ and the luciferase activity of the cells was measured. As shown in Fig. 2,A, the treatment activated the transfectant and resulted in the induction of luciferase activity. This result proved that the chosen promoter region of the IL-12 p40 gene is appropriate for further study. Because it has been shown that the LPS signal required for IL-12 p40 gene activation was mediated by NF-κB (18), and that the p65 subunit of the NF-κB complex is a potent transcriptional activator in the apparent absence of the p50 subunit (30), we used a NF-κB p65 expression plasmid (CMIN-p65, a gift of Dr. Ruben, Roche Institute of Molecular Biology, Nutley, NJ) as a substitute for treatment with LPS. The reporter plasmid containing the IL-12 p40 (p40–677-pLuc) promoter region was electrically transfected to Cos7 cells and luciferase activity was measured. Transfection of the plasmid to Cos7 cells gave baseline enzyme activity. Cotransfection of CMIN-p65 with p40–677-pLuc to Cos7 cells did not augment luciferase activity. On the other hand, cotransfection of IRF-1 expression plasmid (pAct-1, containing actin-promoter and full length of IRF-1 cDNA) with p40–677-pLuc to Cos7 cells augmented the luciferase activity. This augmentation was further enhanced by the copresence of CMIN-p65 (Fig. 2 B). These results show that the binding region for IRF-1 is present within the DNA sequence at −677 to +51 of the IL-12 p40 gene.

The putative IRF-1 binding site is contained in the published DNA sequence at −75 to −56 in the upstream region of the IL-12 p40 gene (AGTTTCTACTTTGGGTTTCC) (18, 29, 31). The −133 to +51 DNA region of the IL-12 p40 gene (which contains the NF-κB half site) was obtained by PCR and inserted into the 5′-upstream region of pLuc. The constructed plasmid (p40–133-pLuc) was electrically transfected to Cos7 cells and luciferase activity was measured (Fig. 2 C). Cotransfection of pAct-1 with p40–133-pLuc to Cos7 cells augmented the luciferase activity. These results show that the IRF-1 binding site in the IL-12 p40 gene is present in −133 to +51 of the 5′-upstream region of the gene.

Whether IRF-1 protein binds to the putative IRF-1 binding site of the IL-12 p40 promoter region was directly tested by EMSA. An in vitro-translated IRF-1 protein, which contains a His tag, was incubated with the −133 to −31 region of the IL-12 p40 promoter (which lacks a TATA box), and the resulting solution was applied to the gel. As shown in Fig. 3, the recombinant IRF-1 protein formed a band, and the addition of Ab specific for the His tag of the recombinant IRF-1 altered the electrophoretic mobility of the band. The result shows that IRF-1 binds to the used promoter region of the IL-12 p40 gene.

To detect the critical promoter region for IRF-1 function, deleted segments of the promoter region were obtained by PCR and inserted into the 5′-upstream region of the luciferase gene of pLuc after confirming the sequence of the PCR products. Thus constructed plasmids were transfected into Cos7 cells with pAct-1 plasmid and, the luciferase activity of the transfectants was measured (Fig. 4 A). The promoter activity was obtained in the −83 to +51 region, partially reduced in the −73 to +51 region, and completely absent in the −63 to +51 region. This result shows that the important DNA segment for IRF-1 promoter activity is present between −83 to −63 of the 5′-upstream region of the IL-12 p40 gene, and consistent with our prediction.

We accidentally found that the DNA sequence at −63 to −61 of the IL-12 p40 gene is the critical binding site of IRF-1 by introducing point mutation. The wild-type DNA sequence of the promoter region of the p40–133-pLuc plasmid was replaced by three kinds of mutant sequences. The resultant plasmids were subjected to the luciferase assay by cotransfection with an IRF-1-expression plasmid (Fig. 4, B and C). The augmentation of luciferase activity by IRF-1 was observed with plasmid containing a wild-type promoter, while the augmentation was reduced with the promoters containing a mutation. This point was further tested by EMSA using the promoters containing a mutation (Fig. 4 D). The result was consistent with that obtained by luciferase assay described previously. These results show that the sequence of the −63 to −61 of the 5′-upstream region is critical for the binding of IRF-1 transcription factor to the promoter of the IL-12 p40 gene.

The position of the actual binding site was further determined by DNase I footprinting assay (Fig. 5,A). The −72 to −58 region was protected from DNase I digestion, which demonstrates that this region is the binding site of IRF-1 protein to the IL-12 p40 promoter. Actual binding of IRF-1 to IL-12 p40 promoter of PECs in vivo was demonstrated by ChIP assay. The anti-IRF-1 Ab precipitated the IL-12 p40 promoter region from LPS plus IFN-γ-stimulated PECs (Fig. 5 B).

The gene expression of IL-12 p40 was induced by stimulation with LPS in macrophages that had been pretreated with IFN-γ (2, 18, 32). IL-12 p40 was shown to be one of the IFN-γ-inducible genes that are activated by a phosphorylated complex, GAF, which consists of dimerized STAT1α and possibly another DNA binding protein (19, 20). IRF-1 is one of the GAF-activated proteins and has been shown to modulate the cellular response to IFN-γ (19). IL-12 is a heterodimer composed of p40 and p35. The former is induced by microbial infection, whereas the latter is known to be constitutively expressed (2, 3). Although IL-12 p40 gene induction requires costimulation with IFN-γ, the IL-12 p35 gene is rapidly and strongly induced in PECs during bacterial infection even in the absence of IFN-γ (Y. Asano, unpublished observation). Because signals from IFN-γ receptor are transduced by transcription factors, IRF-1 and p48, we assessed the requirement of these transcription factors for IL-12 p40 and p35 genes expression in the present study. We demonstrated that induction of IL-12 p40 and p35 genes is differentially regulated by IFN-γ via IRF-1 but not by p48. IL-12 p40 is IRF-1-dependent and IL-12 p35 is IRF-1-independent.

We analyzed the promoter region required for the regulation by the transcription factor IRF-1. Stimulation of RAW cells transfected with the −677 to +51 region of the IL-12 p40 gene resulted in the induction of luciferase activity. This finding suggested that the transcription factors induced by LPS and IFN-γ bind to this DNA region, and it was further confirmed by cotransfection experiments using NF-κB and IRF-1 expression plasmids. The cotransfection of CMIN-p65 and pACt-1 induced the luciferase activity in a synergistic manner confirming that a LPS and IFN-γ responsive region is present in this promoter segment. Although IL-12 p40 is barely induced in PECs by stimulation with IFN-γ alone, IRF-1 (pAct-1) augmented the −677 promoter activity of IL-12 p40 in the absence of CMIN-p65 (Fig. 2,B). This finding suggests the possibility that the negative regulatory element might be present in the upstream area of the promoter region. Because we found that IRF-1 binds to the promoter region of the IL-12 p40 gene, we further determined the binding site of IRF-1. We showed by an EMSA using the DNA segment at −133 to −31 of the IL-12 p40 promoter region that the important DNA segment for IRF-1 binding is present between −72 to −58 of the 5′-upstream region of the IL-12 p40 gene (Figs. 3 and 5). In addition, we found through the mutation experiment that the DNA sequence at −63 to −61 of IL-12 p40 gene is the critical site for binding of IRF-1 (Fig. 4). This is the first direct demonstration of the binding site of IRF-1 in the promoter region of the IL-12 p40 gene. The copresence of NF-κB and IRF-1 maximally up-regulates the gene transcription, consistent with the observation of in vitro induction of the gene by LPS plus IFN-γ.

Although we demonstrated the direct binding of IRF-1 protein to the promoter of the mouse IL-12 p40 gene in the present study, the copresence of NF-κB (CMIN-p65) with IRF-1 (pAct-1) resulted in the maximum augmentation of the promoter activity. It is demonstrated that IRF-1 forms a complex with other nuclear factors during stimulation with IFN-γ or LPS and the complex up-regulates the promoter activity of the human IL-12 p40 gene (33). However, the formation of an IRF-1 complex does not always up-regulate promoter activity. We determined in a previous study that an IRF-1-containing complex was found in the nuclear extract of Plasmodium berghei-infected PECs and the complex bound to the IRF-1-binding motif. The IL-12 p40 transcription and IL-12 p70 production was inhibited in these PECs (34). This finding also suggests the possibility of the presence of a negative regulatory component in the IL-12 p40 promoter region.

We thank Dr. Hiroshi Yamamoto, Dr. Takashi Yokota, Taeko Fukuda, and the Genetic Institute for an IL-12 p40 probe, TNF-α probe, animal care, and recombinant mouse IL-12, respectively. We also thank Dr. Tadatsugu Taniguchi for his warm support.