Ligating FcγR on macrophages results in suppression of IL-12 production. We show that FcγR ligation selectively down-regulates IL-12 p40 and p35 gene expression at the level of transcription. The region responsive to this inhibition maps to the Ets site of the p40 promoter. PU.1, IFN consensus sequence binding protein, and c-Rel form a complex on this element upon macrophage activation. Receptor ligation abolishes the binding of this PU.1-containing activation complex, and abrogates p40 transcription. A dominant-negative construct of PU.1 diminishes IL-12 p40 promoter activity and endogenous IL-12 p40 protein secretion. Thus, the specificity of IL-12 down-regulation following receptor ligation lies in the inhibition of binding of a PU.1-containing complex to the Ets site of the IL-12 promoter. These findings provide evidence demonstrating for the first time the importance of PU.1 in the transcriptional regulation of IL-12 gene expression.
Interleukin 12 is an inducible heterodimeric cytokine made up of two covalently linked subunits, designated p35 and p40. IL-12 drives type I immune responses, primarily by virtue of its ability to induce the production of IFN-γ from T and NK cells. A number of studies have shown that animals deficient in IL-12 are more susceptible to infections by intracellular organisms (1, 2), and conversely that the addition of exogenous IL-12 can promote protective immune responses to these organisms (3, 4). Despite its conspicuous role in driving cellular immune responses, the overproduction of IL-12 can be detrimental to the host. This is especially evident in cases of acute endotoxemia (5) or during autoimmune diseases (6, 7), where uncontrolled IL-12 production can exacerbate disease. Thus, the tight regulation of IL-12 is necessary to allow type I immune responses to occur without the consequent autoimmune sequelae that can accompany the overproduction of IL-12.
We (8) and others (9) have demonstrated that ligating phagocytic receptors on macrophages prevents these cells from making IL-12 in response to a variety of proinflammatory stimuli. The down-regulation of IL-12 production was a property that was shared by all of the phagocytic receptors tested, including the FcγRs, complement receptors, and scavenger receptors. Similarly, the ligation of CD46 (10), CD47 (11), the histamine H2 receptors (12), or the β2 adrenergic receptors (13) also resulted in IL-12 down-regulation. Thus, a down-regulation of IL-12 production may accompany the ligation of a variety of receptors on macrophages.
The regulation of IL-12 transcription is complex and incompletely understood. The genes for IL-12 p40 and p35 are independently regulated. The IL-12 p40 gene is transcriptionally regulated and expressed only in cells making biologically active IL-12 (14). The transcription of IL-12 p40 has been investigated by several groups (15, 16, 17, 18, 19, 20). Three potential transcriptional regulatory elements have been identified in the proximal regions of the IL-12 p40 promoter. An NF-κB half-site has been identified at −132/−122 bp of the murine IL-12 p40 promoter (15). NF-κB complexes composed of either p50/p65 or p50/c-Rel bind to this site following cellular activation to produce IL-12. An Ets site is located at −212/−207 in the human IL-12 p40 promoter, and this site has been identified as a major response region to IFN-γ and LPS stimulation (18, 19). A third potential site for CCAAT/enhancer binding protein β is located at −80/−72 of the murine p40 promoter (17). In a variety of experimental systems, both the NF-κB site and the Ets site are required for optimal IL-12 p40 transcription. Deleting or mutating either of these sites causes a dramatic decrease in IL-12 transcription, suggesting that both of these domains are needed to drive the regulated transcription of p40.
Our understanding of the regulation of the IL-12 p35 gene has lagged behind that of p40. In leukocytes, p35 is made in very low levels unless the cells are primed with IFN-γ, which results in an increase in p35 transcription (21). The human p35 promoter contains several putative transcription factor-binding elements including PU.1, CCAAT/enhancer binding protein, and GAS (IFN-γ activation sequence). The p35 gene appears to initiate transcription from multiple sites (21, 22), and there is evidence to suggest that the p35 gene may also be regulated at the level of post-translational processing (23, 24).
PU.1 belongs to the Ets family of DNA binding proteins (25, 26). It is expressed predominantly in macrophages, B cells, and erythroid cells (27, 28). PU.1 plays important but varied roles in the development of hematopoietic cells. A deletion of the PU.1 gene leads to a failure to produce mature B lymphocytes and macrophages (29). However, the over-expression of PU.1 prevents terminal differentiation of hematopoietic cells and leads to the overproduction of erythroblasts (30). It was shown recently, using retroviral transduction of PU.1 cDNA into hematopoietic progenitors derived from PU.1-deficient mice, that a graded expression of PU.1 can lead to distinct cell fates (i.e., a low concentration of PU.1 protein induces the B cell fate, whereas a high concentration promotes macrophage differentiation and blocks B cell development) (31). In many instances the transcriptional activation by PU.1 depends on the cooperation of PU.1 with other transcription factors. These factors can either become physically associated with PU.1 or they can bind to DNA adjacent to the Ets site upon which PU.1 resides (32, 33, 34, 35, 36).
In the present work, we examine the molecular mechanism of IL-12 down-regulation following FcγR ligation. We show that the decrease in IL-12 production is due to a decrease in the transcription of both the IL-12 p40 and p35 genes. This decrease in transcription maps to the Ets site found in the p40 promoter. In the present work, we characterize a transcription complex, consisting of PU.1, c-Rel, and IFN consensus sequence binding protein (ICSBP),4 that is formed on this site following cellular activation. We demonstrate that the ligation of phagocytic receptors on macrophages results in an inability to form this complex, and a dramatic decrease in IL-12 p40 transcription.
Materials and Methods
Cells and reagents
Bone marrow-derived macrophages (BMMφ) were established as previously described (8). One day before use, fully differentiated BMMφ were removed from the original plates, washed, and allowed to readhere to plastic wells in DMEM containing 10% heat-inactivated FCS, 2 mM l-glutamine (Mediatech, Herndon, VA) and 100 U/ml penicillin and 100 μg/ml streptomycin (Mediatech). The murine macrophage-like cell line RAW264.7 was obtained from American Type Culture Collection (Manassas, VA), and maintained in RPMI 1640 (Mediatech) containing 10% FCS, 2 mM l-glutamine and penicillin/streptomycin. All of the Abs used in the Western blots, immunoprecipitations, and supershift experiments were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Recombinant murine IFN-γ was purchased from Genzyme (Cambridge, MA). LPS from Escherichia coli 0127:B8 was purchased from Sigma (St. Louis, MO). IgG-opsonized erythrocytes (EIgG) were generated by incubating sheep erythrocytes (Lampire, Pipersville, PA) with rabbit anti-sheep erythrocyte IgG (Cappel, Durham, NC) at nonagglutinating titers for 40 min at room temperature, as previously described (37). EIgG were washed with HBSS (Life Technologies, Grand Island, NY) and resuspended at 2 × 108 cells/ml and used the same day.
Macrophage stimulation and receptor ligation
Monolayers of BMMφ were washed and primed with 100 U/ml of IFN-γ for 12 h before stimulation with LPS at a final concentration of 100 ng/ml, in the presence or absence of opsonized erythrocytes. EIgG were added immediately before LPS stimulation at a ratio of 20 erythrocytes per macrophage, which generally resulted in a binding of 1–3 erythrocytes per macrophage after 1 h. RAW264.7 cells were primed with IFN-γ at a final concentration of 200 U/ml and stimulated with LPS at a final concentration of 1 μg/ml. Erythrocytes were added at the same ratio as for BMMφ.
A 3300-bp and a 220-bp fragment of the human IL-12 p40 promoter were cloned into the luciferase reporter construct pXP2 (16). A genomic fragment of the TNF-α gene derived from nucleotides 2849 through 4142 of clone M16441 was cloned into the SmaI site of the pXP2 luciferase reporter plasmid by blunt ended ligation. A 1143-bp genomic fragment of the human IL-12 p35 promoter derived from the clone AF050083 corresponding to nucleotide positions 532 through 1675 was cloned into the pXP2 luciferase vector as a BamHI-PstI fragment. The NF-κB luciferase plasmid was purchased from Stratagene (La Jolla, CA; catalog number 219078). The PU.1 dominant-negative mutant plasmid was kindly provided by Michael L. Atchison (University of Pennsylvania, School of Veterinary Medicine, Philadelphia, PA) (38). All plasmids were isolated using the Qiagen Endotoxic free kit (Qiagen, Valencia, CA).
RAW264.7 cells stably transfected with a plasmid containing the luciferase gene driven by a 3.3-kbp fragment of the IL-12 p40 promoter were generated as previously described (16). Transient transfections were performed by electroporation as previously described (16). Briefly, for each condition 0.4 ml of a cell suspension containing 1 × 107 cells was mixed with 50 μg total DNA and electroporated in 0.45-cm electroporation cuvettes (Gene Pulser II; Bio-Rad, Hercules, CA) at 960 microfarads and 340 V in RPMI 1640 without serum. Transfected cells from the different cuvettes were combined and resuspended in RPMI 1640 containing 10% FCS and 10 μM chloroquine. Cells were added to 24-well plates and incubated for 48 h before harvesting. To measure luciferase activity, cells were pelleted by centrifugation and resuspended in 100 μl lysis buffer (20 mM tricine, 1.07 mM MgCO3, 2.67 mM MgSO4, 0.1 mM EDTA, 33.3 mM DTT, 270 μM coenzyme A, 530 μM ATP, 470 μM D-luciferine potassium salt). For cotransfection experiments with the PU.1 dominant-negative mutant, transfected cells from each cuvette were plated separately and luciferase activity was normalized by cotransfection of 3 μg pCMV-β-galactosidase plasmid. Lysates were used for both luciferase and β-galactosidase assays (16).
Cytokine secretion was measured by ELISA, using appropriately diluted culture supernatants. IL-12 p40 was measured using mAbs C15.6 (anti-murine IL-12 p40) and C17.8 (biotinylated anti-murine IL-12 p40) as ELISA capture and detection Abs, respectively, according to protocols provided by PharMingen (San Diego, CA). Recombinant murine IL-12 (Genzyme) was used as a standard.
RNase protection was performed using the mCK-2 RiboQuant Multiprobe RNase Protection Assay system from PharMingen according to the manufacturer’s instructions. A total of 10 μg RNA was used for each determination. The intensity of all bands was determined by densitometry after normalizing to the L32 gene. Gels from three distinct experiments were analyzed and the mean fold decrease following FcγR ligation was determined.
Nuclear extractions for Western analysis and for EMSA assays were performed according the method of Schreiber et al. (39). Briefly, 5–10 × 106 cells were washed and resuspended in 600 μl of buffer containing 10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 1 mM DTT, and 0.5 mM PMSF for 15 min on ice. Cells were lysed in 0.6% Nonidet P-40 with vortexing for 10 s. The homogenate was centrifuged for 30 s in a microfuge and the nuclear pellet was resuspended in ice-cold buffer containing 20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF at 4°C for 15 min with rocking. Following centrifugation in a microfuge for 5 min, the supernatant was either used immediately or frozen at −70°C.
Western blotting and immunoprecipitation
SDS-PAGE was performed according to Laemmli (40) with 20–30 μg of nuclear extract. Western analysis was performed as described (18). Gels were electroblotted to PVDF membranes (Millipore, Bedfore, MA) and blocked in 5% milk in Tris buffer, pH 8.0. Primary Abs were added at the concentration of 1 μg/ml in Tris buffer containing 1% milk powder for 1 h at room temperature. After extensive washing, secondary Ab conjugated to HRP was added at a 1:5000 dilution in 1% milk. After extensive washing, blots were subjected to ECL detection (Amersham Pharmacia Biotech, Piscataway, NJ). For immunoprecipitation analysis nuclear extracts (100 μg) prepared from unstimulated or IFN-γ/LPS-stimulated cells were precleared for 1 h at 4°C using 15 μl of protein A-agarose (Santa Cruz Biotechnology). Extracts were immunoprecipitated with 2 μg/ml of anti-PU.1 mAb (Santa Cruz Biotechnology) plus 15 μl protein A-agarose for 3 h at 4°C. Precipitated protein was washed five times in buffer (10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.5 mM PMSF, 10 mM NaF, 5 mM Na3VO4, 10 μg/ml leupeptin, 10 μg/ml pepstatin A). The final pellet was resuspended in 60 μl of SDS sample buffer containing 5% 2-ME, boiled for 5 min, and electrophoresed on 12% SDS-PAGE.
EMSA and supershifts were performed as described previously (18). The sequence spanning the −222/−196 region of the human p40 promoter 5′-CAAAAGTCATTTCCTCTTAGTTCAT-3′, was used as a probe for PU.1 binding.
Nuclear run-on assays
The isolation of nuclei and in vitro transcription in the presence of [32P]UTP (3000 Ci/mmol) were performed as previously described (18). [32P] pulse-labeled nuclear RNA was hybridized for 2 days at 60°C to prehybridized nylon filters to which 500 ng of denatured cDNA corresponding to the coding regions of IL-12 p35, IL-12 p40, and TNF-α had been immobilized using a slot blot apparatus (Hoeffer Scientific, San Francisco, CA). After hybridization, filters were washed twice at room temperature with 2× SSC containing 0.1% SDS for 15 min at 50°C, and once with 0.1× SSC containing 0.1% SDS for 30 min at 50°C. Hybridization was quantified using ImageQuant software on a PhosphorImager 44581 (Molecular Dynamics, Sunnyvale, CA).
IL-12 p40 and p35 gene transcription are selectively inhibited following FcγR ligation
We have previously shown that the ligation of phagocytic receptors on macrophages prevented their production of IL-12 in response to proinflammatory mediators, such as LPS (8). In the present work, to examine the mechanism of this effect, we used macrophages that were stimulated with LPS in either the presence or absence of EIgG to ligate the FcγR. Macrophages grown in medium alone without stimuli produced minimal cytokines, as expected (Fig. 1). LPS treatment of IFN-γ-primed macrophages induced the production of a variety of cytokines including IL-10, IL-1α, IL-1β, IL-6, migration inhibitory factor (MIF), and IL-12 p35 and p40. The simultaneous ligation of FcγR on parallel monolayers of stimulated macrophages resulted in marked and specific reduction in mRNA encoding the two subunits of IL-12, p35, and p40 (Fig. 1). The mRNA levels for many other cytokine genes was either unchanged or minimally decreased following FcγR ligation (Fig. 1). Densitometric quantitation of three separate RNase protection assays indicated that whereas IL-12 mRNA levels decreased by >90%, the mRNA levels for MIF, IL-6, and IL-1RA were not significantly diminished following FcγR ligation. A single other gene, IL-10, was increased following FcγR ligation, as previously reported (37). The addition of control unopsonized erythrocytes, which do not bind to specific macrophage receptors, did not affect cytokine mRNA production. A similar alteration in cytokine production was observed when soluble immune complexes (OVA/αOVA) were used to ligate FcγR (data not shown).
To determine whether the reduction of IL-12 mRNA occurred at the level of transcription, nuclear run-on assays were performed on macrophages following IFN-γ/LPS treatment (Fig. 2). Macrophages stimulated in the presence of EIgG exhibited a decrease in the transcription of both the IL-12 p40 and p35 genes (Fig. 2). Transcription of IL-12 in the presence of unopsonized erythrocytes was the same as that observed in macrophages treated with IFN-γ/LPS alone. As a control for these studies, the transcription of TNF-α was also analyzed and found to be minimally reduced following FcγR ligation (Fig. 2). Thus the ligation of FcγR on stimulated macrophages specifically prevented the accumulation of mRNAs for IL-12 p40 and p35, and this regulation occurred at the level of gene transcription.
The promoter activities of the IL-12 p40 and p35 genes are selectively inhibited following FcγR ligation
To further examine the regulation of IL-12 transcription following FcγR ligation, an extended (3.3 kbp) segment of the human IL-12 p40 promoter was cloned immediately upstream of a luciferase reporter gene and the construct was stably transfected into the RAW264.7 macrophage-like cell line, as previously reported (16). To show that RAW264 cells were appropriate recipient cells for these assays, endogenous IL-12 from these cells was measured by ELISA following their stimulation with IFN-γ/LPS. Stimulated cells made relatively high levels of IL-12 p40 and these levels were dramatically reduced when cells were stimulated in the presence of FcγR ligation (Fig. 3,A). Thus, like primary macrophages, these cells respond to LPS stimulation by secreting IL-12, and they also react to receptor ligation by decreasing IL-12 synthesis, indicating that they are appropriate cells for use in these assays. Luciferase activity of the stably transfected cells was measured following stimulation with LPS in the presence or absence of FcγR ligation. Stimulated cells made relatively high levels of luciferase and this activity was markedly diminished following receptor ligation (Fig. 3 B). Control unopsonized erythrocytes failed to abrogate luciferase induction. Thus the RAW264.7 cells used in these assays largely recapitulate the responses of primary cells to LPS stimulation, and luciferase activity from these stably transfected cells is an accurate reflection of IL-12 transcription.
Transient transfections were also performed on this cell line using both the p40 and the p35 promoters to drive luciferase activity. Transiently transfected cells were primed with IFN-γ and then stimulated with LPS. Stimulation of primed cells with LPS resulted in substantial luciferase activity driven by the extended (3.3 kbp) IL-12 p40 promoter. This activity was significantly diminished by ligating the FcγR with EIgG (Fig. 4,A), although the effect was not as dramatic as that observed with stably transfected cells. Control unopsonized erythrocytes had no effect on LPS-driven luciferase activity. LPS treatment also stimulated the production of luciferase driven by the p35 promoter, and this activity was also diminished by FcγR ligation (Fig. 4,B), although not as much as the p40 promoter. To identify the promoter regions involved in IL-12 down-regulation, the IL-12 p40 promoter was truncated from 3.3 kbp to 222 bp 5′ of the transcription initiation site. This truncated promoter, which includes three of the putative transcription factor binding sites of the p40 promoter is the minimal promoter capable of driving IL-12 transcription in response to IFN-γ and LPS (16). This minimal promoter was sufficient to drive luciferase production, albeit not as well as the extended promoter (Fig. 4,C). However, the ligation of phagocytic receptors down-regulated luciferase activity to an extent that was comparable to that observed with the extended promoter. These results suggest that the down-regulation of IL-12 transcription following receptor ligation is mediated by the proximal 222 bases of the IL-12 p40 promoter. As controls for these studies, luciferase driven by the TNF-α promoter and luciferase driven by a consensus NF-κB sequence were examined in parallel. Neither TNF-α nor NF-κB-driven luciferase activity was significantly diminished by FcγR ligation (Fig. 4, D and E, respectively). These results demonstrate that the ligation of phagocytic receptors causes a selective decrease in the transcription of the IL-12 p40 and p35 genes.
The nuclear translocation of PU.1, ICSBP, and IFN regulatory factor 1 (IRF-1) is inhibited following FcγR ligation
To further examine the decrease in IL-12 transcription following FcγR ligation, Western blotting analysis was performed to measure the nuclear translocation of a variety of transcription factors that have been reported to interact with the IL-12 p40 promoter. Nuclear extracts were isolated 4 h following cellular activation in the presence or absence of FcγR ligation. The nuclei of unstimulated cells contained abundant ets-2 but only modest amounts of PU.1 and ICSBP. Little or no detectable levels of IRF-1 or NF-κB transcription factors were detected in the nuclei of unstimulated cells (Fig. 5, lane 1). When cells were stimulated with IFN-γ/LPS all seven of the factors examined were detected in high levels in nuclear extracts (Fig. 5, lane 2). The ligation of FcγR on stimulated cells resulted in a decrease in the levels of PU.1, ICSBP, and IRF-1 (Fig. 5, lane 3). This decrease was specific to these factors, as the levels of the three NF-κB members (c-Rel, p50, p65) were not diminished by FcγR ligation, and were detected in levels that were comparable to those observed in cells stimulated with IFN-γ/LPS alone (Fig. 5, lane 2 vs lane 3). Consistent with the nuclear accumulation of NF-κB in activated cells, the cytoplasmic extracts of unstimulated cells contained ample IκB, which was rapidly degraded following IFN-γ/LPS stimulation, as previously reported (41). The degradation of cytoplasmic IκB was not reversed by FcγR ligation. Thus FcγR ligation does not influence the extent nor the kinetics of NF-κB nuclear translocation. However, FcγR ligation did exert a dramatic effect on the nuclear levels of PU.1, ICSBP, and IRF-1, which were severely reduced following FcγR ligation.
The binding of a PU.1-containing complex to the Ets domain of the IL-12 p40 promoter is inhibited following FcγR ligation
EMSA were performed using a 27-base probe corresponding to nucleotides −222/−196 surrounding the Ets domain (−212/−207) of the IL-12 p40 promoter. In resting cells a single factor bound to this site (Fig. 6, lane 2). This factor has previously been identified to be PU.1 (18). In cells stimulated with IFN-γ/LPS, a second higher-order complex was formed on this site (Fig. 6, lane 3, arrow designated PU.1+). The ligation of FcγR resulted in a dramatic loss of transcription factor binding to this probe (Fig. 6,A, lane 4). Neither the lower (resting) band nor the higher (activation complex) band was detected in these extracts. As a control, extracts from IFN-γ/LPS stimulated cells incubated with unopsonized erythrocytes were analyzed and found to behave similarly to LPS-stimulated cells, showing both the resting band and the activation complex (Fig. 6,A, lane 5). Supershift assays were performed to characterize the complex formed on the Ets-2 site following cellular activation. Ab to PU.1 caused a dramatic decrease in the binding of all factors to this site, indicating that both the lower and the upper activation complex are composed of PU.1, and furthermore that interfering with PU.1 abolishes all DNA binding activity to this site. The addition of Abs to ICSBP or c-Rel increased the intensity of the lower band relative to the intensity of the upper band (Fig. 6,A). Abs to several of the other transcription factors tested did not affect the intensity of either band, indicating that this increase was specific to ICSBP and c-Rel. These results suggested that the activation complex formed on this site contains ICSBP, and c-Rel in addition to PU.1. This conclusion was supported by immunoprecipitation experiments, where Ab to PU.1 was used to immunoprecipitate the PU.1 complex from the nuclei of stimulated cells. This complex was subsequently analyzed by Western blot under denaturing and reducing conditions. As shown in Fig. 6,B, Ab to PU.1 precipitated not only PU.1, but also ICSBP and c-Rel. To determine the step at which PU.1 was down-regulated, we examined the steady-state mRNA expression of PU.1 following FcγR ligation. PU.1 mRNA was constitutively expressed and was not affected at all by FcγR cross-linking, whereas both IL-12 p40 and p35 mRNA expression were characteristically inhibited (Fig. 6 C), suggesting that FcγR ligation targets PU.1 subsequent to mRNA transcription.
To determine the precise DNA sequence requirements for the binding of these two complexes, a sequential mutagenesis approach was taken in which bases of the Ets-2 domain were altered. Gel shift analyses were then performed (Fig. 7). Mutations encompassing the Ets-2 site abrogated the binding of both complexes, whereas mutations outside of this site affected neither complex (Fig. 7). Importantly, all mutations that prevented the binding of PU.1 (lower complex) also prevented the formation of the upper PU.1+ complex, suggesting that the formation of the activation complex depends on PU.1 binding to the DNA. Furthermore, there were no mutations that affected only the binding of the activation complex, indicating that flanking sequences adjacent to the Ets element were not required for the binding of the activation complex.
Inhibition of PU.1 binding by a dominant-negative mutant abolishes IL-12 p40 promoter activity
To further correlate the decrease in IL-12 production with alterations in PU.1 activity, a dominant-negative construct of PU.1 (PU.1-DN) was cotransfected into RAW264.7 cells along with a luciferase construct driven by the IL-12 p40 promoter. Transfected cells were stimulated with IFN-γ/LPS to induce the production of luciferase driven by the IL-12 p40 promoter. In cells cotransfected with the PU.1-DN construct, IL-12 p40 promoter-driven luciferase activity was decreased to near background levels (Fig. 8,A). Control cells transfected with the vector alone (pCB6) were fully capable of expressing IL-12-driven luciferase activity. The inhibition of transcription by PU.1-DN was specific to IL-12 p40, because neither the TNF-α promoter-driven luciferase activity (Fig. 8,B) nor NF-κB (Fig. 8,C) was diminished by this construct. Furthermore, the sequence element(s) responsive to the inhibitor appears to be located within the proximal 222 bp of the IL-12 p40 promoter, because similar to the 3.3-kb parental construct, the −222 construct was also inhibited by ectopic expression of the PU.1-DN (Fig. 8 D). Thus IL-12 p40 transcription depends on PU.1 binding to the Ets site of the IL-12 p40 promoter, and the inhibition of this binding prevents IL-12 p40 transcription.
Interfering with PU.1 functions results in a diminution of endogenous IL-12 p40 gene expression
It is generally believed that in transient transfections, the reporter gene is not usually associated with a particular chromatin structure like the endogenous gene. To determine whether the observed effects of the PU.1-DN on the IL-12 p40 reporter gene in RAW264.7 cells could also be exerted on the endogenous IL-12 p40 gene, we transfected RAW cells with the p40 reporter gene with or without the PU.1-DN. Cell-free supernatants were collected following appropriate stimulation, and the endogenously produced murine IL-12 p40 and TNF-α were measured by ELISA. As shown in Fig. 9, the synthesis of endogenous mIL-12 p40 protein was inhibited by ∼70% by PU.1-DN, whereas murine TNF-α was not affected. Thus, interfering with PU.1 activity leads to a selective inhibition of both endogenous and exogenous IL-12 p40 gene expression.
The goal of this study was to examine the down-regulation of IL-12 production that occurs as a consequence of phagocytic receptor ligation. We have previously used defined particles to show that the ligation of several phagocytic receptors on macrophages prevents these cells from producing IL-12 in response to proinflammatory stimuli. In these previous studies IL-12 p70 levels in activated cells were decreased to virtually undetectable levels following receptor ligation (8). Because IL-12 is an important cytokine in directing appropriate immune responses to intracellular pathogens, an understanding of IL-12 regulation is essential to begin to dissect the initiation of type I immune responses. In the present study we use FcγR ligation as a model of phagocytic receptor interaction. We show that the decrease in cytokine production following receptor ligation is specific to IL-12 p40 and p35, occurs at the level of gene transcription, and involves a lack of binding of a large transcription complex to the Ets element in the IL-12 p40 promoter. These observations begin to explain at the molecular level the down-regulation of IL-12 production following receptor-mediated phagocytosis.
We have taken several experimental approaches to show that the receptor-mediated down-regulation of IL-12 occurs at the level of transcription. These approaches included RNase protection, nuclear run-on, and luciferase reporter assays. In all three of these experimental systems, the ligation of FcγRs had a profound inhibitory effect on IL-12 p40 gene transcription, often reducing transcription to background levels. Because the IL-12 p40 gene is expressed only in cells making biologically active IL-12 and because its transcription has been most closely correlated with IL-12 p70 production, the regulation of p40 was the initial focus of this work. We show that IL-12 p40 production was inhibited following FcγR ligation. However, we also analyzed the transcription of IL-12 p35 following receptor ligation. Although this gene is expressed in much lower levels than p40, its expression was also significantly decreased by FcγR ligation. By the luciferase assay, p35 promoter activity was decreased by ∼50%, and by both RNase protection and nuclear run-on assays the levels were also substantially reduced. Thus the ligation of FcγR on macrophages decreases the transcription of both the IL-12 p40 and IL-12 p35 genes. The decrease in transcription was specific to IL-12. By RNase protection assays several other genes, including IL-1α, IL-1β, IL-6, and MIF, continued to be transcribed following FcγR ligation (Fig. 1). Similarly the transcription of TNF-α, measured by either nuclear run-on or luciferase expression was minimally affected by FcγR ligation. Thus receptor ligation does not render macrophages globally refractory to proinflammatory stimuli; rather it specifically inhibits IL-12 production.
Nuclear extracts were analyzed 4 h after receptor ligation by Western blotting to determine whether the accumulation of transcription factors was affected by FcγR ligation. The levels of two IRFs, ICSBP and IRF-1, were markedly diminished in the nuclei of cells following FcγR ligation. This observation is consistent with previous studies showing that knockout mice lacking either of these factors were defective in their ability to produce IL-12 p40 (2). Similarly, our recent studies (42) demonstrate that the overexpression of ICSBP could increase IL-12 p40 transcription. Thus the IRFs are a vital component of IL-12 p40 transcription, and FcγR ligation can functionally mimic the genetic deficiencies in the IRFs. A third transcription factor was also dramatically affected by FcγR ligation. This was PU.1, a member of the Ets family of transcription factors. An Ets site is located at −212 of the human IL-12 p40 promoter. This site with flanking sequences (−292 to −196) has previously been shown to be an important element in controlling the transcription of IL-12 in the RAW264.7 macrophage cell line (17). The extended Ets element binds to a high m.w., multicomponent complex named F1 (18). In RAW cells, the F1 complex appears to contain a number of transcription factors including Ets-2, IRF-1, c-Rel, and ICSBP. The over expression of some of these factors is sufficient to activate the IL-12 p40 promoter through the Ets site independently of the signals provided by LPS and IFN-γ (19, 42). However, in primary macrophages an F1 complex has not been identified. In the present study we describe a smaller complex induced by IFN-γ/LPS that binds to the Ets element but requires less physical space (−222 to −196). In addition to PU.1 this complex also contains ICSBP and c-Rel. We were unable to detect IRF-1 in this complex by either EMSA or Western analysis. The nature of the differences between the transformed macrophage cell line and primary macrophages used in the present work is presently unclear.
We have taken three different experimental approaches to examine the binding of transcription factors to the Ets element of the IL-12 promoter, and to determine their role in controlling IL-12 production following receptor ligation. These approaches included the following: 1) nuclear translocation of putative Ets-binding transcription factors, 2) EMSA using the Ets element of the p40 promoter, and 3) a dominant-negative construct encoding the DNA binding domain of PU.1, which lacks the transactivation domain of the wild-type molecule. All of these approaches pointed to the critical role of PU.1 in activating IL-12 gene transcription. In the p40 promoter, PU.1 occupies the Ets site even in resting cells. Resting cells do not transcribe p40 and, therefore, in this state PU.1 is not a transcriptional activator. However, in IFN-γ/LPS-activated cells a larger-molecular-weight complex is formed on this site (Fig. 6). This activation complex consists of not only PU.1 but also ICSBP and c-Rel. This was determined by EMSA coupled with the use of Abs specific to ICSBP and c-Rel, which revealed a decrease in the intensity of the upper (activation) complex relative to the lower (PU.1) complex, indicating that the Abs may have disrupted the higher-order structure of the activation complex. In addition, an anti-PU.1 Ab could immunoprecipitate a complex that also contained ICSBP and c-Rel, further supporting the EMSA data. We hypothesize that the formation of the PU.1-containing activation complex is necessary for IL-12 p40 transcription, and that blocking the formation of this complex by FcγR ligation prevents IL-12 biosynthesis. We could experimentally mimic this blockade by cotransfecting cells with a dominant-negative construct encoding the DNA binding domain of PU.1. Cells were transfected with this construct along with a reporter plasmid driven by the IL-12 p40 promoter. The inhibition of PU.1 binding to the Ets site of the IL-12 p40 promoter by the dominant-negative construct prevented the synthesis of luciferase, indicating that the Ets site is a critical element in the regulation of IL-12.
It should be pointed out that while the formation of the activation complex on the Ets element was required for IL-12 p40 transcription, it is probably not sufficient. Previous studies have shown a requirement for NF-κB in driving IL-12 p40 transcription (15), and recent studies by Sanjabi et al. (43) have shown that mice lacking c-Rel fail to produce IL-12. We hypothesize that the Ets site cooperates with the NF-κB site to drive IL-12 p40 transcription. This could be achieved through several hypothetical ways. For example, c-Rel has been shown to be present in both the ets-related complex and the downstream NF-κB complex (18, 19). It could be a bridging molecule to link the two complexes. In this case, one would predict a drastic disruption of the interaction between the two sites if the NF-κB site to which c-Rel binds is mutated. Alternatively, ICSBP could play an adaptor’s role by virtue of its known interactions with PU.1, which binds to both the ets site and the PU.1 site immediately adjacent to the NF-κB site (18, 19). In our hands, FcγR ligation did not affect IκB degradation or the translocation of NF-κB (p65, p50, and c-Rel) into the nucleus of activated cells (Fig. 5). Furthermore, luciferase driven by a consensus NF-κB site was not decreased by FcγR ligation (Fig. 4 E). Thus, global alterations in NF-κB do not account for the ablation of IL-12 production following FcγR ligation
Our data for the first time demonstrate a reciprocal link between FcγR, an important mediator in both innate and adaptive immunity, and PU.1, a transcription factor indispensable for myeloid differentiation. The molecular basis of FcγR-mediated selective targeting of PU.1 is not presently understood, but these studies are on-going. Our data indicate that mRNA synthesis of PU.1 is not affected by FcγR ligation (Fig. 6 C), suggesting that the PU.1 deficiency occurs downstream at the protein level. The mechanism of PU.1 inactivation may involve an alteration in its phosphorylation, because the phosphorylation status of PU.1 has been reported to affect its association with other transcription factors and its transcriptional activity (32).
In summary, we have shown that FcγR ligation leads to an inhibition of IL-12 transcription and we conclude that the primary mechanism responsible for this decreased transcription lies in the disruption of the formation of a PU.1-containing complex. These results suggest that an important component of the regulation of IL-12 biosynthesis in myeloid cells lies with the myeloid-specific transcription factor, PU.1.
This work was supported by National Institutes of Health Grants AI-46805 (to D.M.M.); AI-134412, CA-32898, and CA-10815 (to G.T.); and AI-45899 and CA-79772 (to X.J.M.). F.S.S. was supported by the M.D./Ph.D. Program at Temple University School of Medicine.
Abbreviations used in this paper: ICSBP, IFN consensus sequence binding protein; BMMφ, bone marrow-derived macrophages; EIgG, IgG-opsonized erythrocytes; IRF-1, IFN regulatory factor 1; PU.1-DN, dominant-negative construct of PU.1; PU.1-DNM, PU.1-DN mutant; MIF, migration inhibitory factor.