Class II MHC Ags are critical for the initiation of immune responses by presenting Ag to T lymphocytes, leading to their activation and differentiation. The transcriptional activation of class II MHC genes requires the induction of the class II transactivator (CIITA) protein, a master regulator that is essential for both constitutive and IFN-γ-inducible class II MHC expression. The cytokine IL-1β has been shown to inhibit IFN-γ-induced class II MHC expression in various cell types. We investigated the underlying mechanism of this inhibitory effect of IL-1β using human astroglioma cell lines. Our findings demonstrate that IL-1β prevents the expression of class II MHC mRNA and protein upon treatment with IFN-γ. Furthermore, we demonstrate that IFN-γ induction of CIITA mRNA expression is inhibited by treatment of cells with IL-1β. IL-1β suppressed IFN-γ activation of the type IV CIITA promoter in astroglioma cells, indicating that the inhibitory influence of IL-1β is mediated by inhibition of CIITA transcription. IL-1β did not interfere with IFN-γ receptor signal transduction, since tyrosine phosphorylation, nuclear translocation, and DNA binding of STAT-1α to an IFN-γ activation sequence of the type IV CIITA promoter were not affected by IL-1β. As well, IL-1β treatment did not affect the ability of IFN-γ-induced interferon-regulatory factor-1 (IRF-1) to bind the IRF-1 element within the type IV CIITA promoter. This study suggests that IL-1β may play a role in regulating immunoreactivity by inhibiting transcription of the CIITA gene, thereby reducing subsequent class II MHC expression.
Class II MHC molecules play a central role in the immune system by presenting Ag to CD4+ Th cells, leading to T cell activation and differentiation (1). Constitutive expression of class II MHC Ags is restricted to professional APCs, such as dendritic cells, B cells, macrophages, and thymic epithelium. However, class II MHC expression can be induced both in vivo and in vitro upon exposure to the cytokine IFN-γ on a wide variety of cell types, including astrocytes, microglia, pancreatic β-cells, keratinocytes, and endothelial cells (for review, see 2 . Aberrant expression of class II MHC molecules has been described in autoimmune disorders such as multiple sclerosis and has been linked to the progression of neurologic disease development (for review, see Refs. 3 and 4). In the central nervous system (CNS),3 class II MHC expression is absent under normal conditions, while in inflammatory conditions, prominent class II MHC Ag immunoreactivity is detected on microglia as well as astrocytes, brain endothelium, and infiltrating leukocytes (5, 6, 7, 8, 9, 10).
Regulation of class II MHC gene expression occurs primarily at the transcriptional level. Coordinated expression occurs through conserved cis-acting regions, termed W (Z, S, or H), X (X1 and X2), and Y elements, within the proximal promoter of most class II MHC genes (for review, see Refs. 2 and 11). Optimal expression requires the cooperative binding of several constitutively expressed trans-acting factors to the W, X, and Y boxes of the class II promoter (12, 13, 14). Although the presence of DNA binding proteins is necessary, it is not sufficient for class II transcription. Transcription of class II MHC genes (both constitutive and inducible) occurs only in the presence of the recently described class II transactivator (CIITA) (15, 16, 17, 18). CIITA acts through the conserved elements within the proximal promoter of class II MHC genes, but does not bind directly to DNA (19, 20). It has been postulated that the mechanism by which CIITA activates transcription is by connecting the constitutively present trans-acting factors to the basal transcription machinery. In this context, it was recently shown that CIITA can interact with both TFIIB and TAF proteins as well as with the X box binding protein RFX5 through distinct domains within the CIITA molecule (21, 22). Constitutive expression of CIITA is only found in cells that also exhibit constitutive class II MHC expression (15). CIITA is also involved in IFN-γ-induced class II MHC expression; CIITA is not constitutively expressed in class II MHC-negative cells, but can be induced upon stimulation with IFN-γ in a time frame that precedes expression of class II MHC (15, 17, 18, 23, 24). Constitutive, ectopic expression of CIITA cDNA from an expression construct transfected into IFN-γ-inducible cells bypasses the requirement of IFN-γ stimulation for class II MHC expression, indicating that CIITA is the mediating factor of IFN-γ induction of class II MHC (16, 18). In CIITA-deficient mice, both constitutive and IFN-γ-inducible class II MHC expression is lacking, except for low expression in a subset of thymic epithelial cells (25). Thus, CIITA appears to be the master switch for class II MHC expression and may potentially serve as a target for controlling aberrant class II MHC expression.
Numerous cytokines, including TNF-α, TGF-β, IFN-αβ, IL-3, IL-4, IL-1, IL-10, and granulocyte-macrophage CSF, have been shown to modulate class II MHC expression in both a positive and a negative manner depending on the cell type studied (for review, see 2 . Immunosuppressive cytokines such as IFN-α/β, TGF-β, and IL-10 generally have an inhibitory effect on IFN-γ-induced class II MHC expression (24, 26, 27, 28, 29, 30, 31, 32, 33, 34). The CIITA gene is an attractive target for cytokine-mediated inhibition of class II MHC expression. In this regard, we and others have recently shown that TGF-β suppresses IFN-γ-induced class II MHC expression by inhibiting the expression of CIITA mRNA (24, 35). The inhibitory effect of TGF-β on CIITA mRNA expression was mediated at the transcriptional level (24, 36), suggesting that the CIITA promoter may be targeted by TGF-β.
We have been interested in the pathways by which cytokines modulate class II MHC gene expression. IL-1, a cytokine with predominantly proinflammatory properties, has been shown to inhibit IFN-γ-induced class II MHC expression in astrocytes, cerebral endothelial cells, and synovial fibroblasts (37, 38, 39, 40). We have investigated the molecular mechanisms underlying IL-1β-mediated inhibition of class II MHC expression in human astroglioma cells and have found that IL-1β exerts its inhibitory effect by suppressing IFN-γ-induced CIITA mRNA expression. IL-1β inhibition of CIITA mRNA expression results from the ability of this cytokine to inhibit IFN-γ activation of the type IV CIITA promoter. Thus, IL-1 may contribute to the regulation of immunological events within the CNS by reducing class II MHC gene expression.
Materials and Methods
The human CH235-MG and U373-MG astroglioma cell lines were maintained in a 50/50 mixture of DMEM and Ham’s F-12 medium (Life Technologies, Gaithersburg, MD) supplemented with 10% FBS, 10 mM HEPES (pH 7.2), 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (complete medium) as previously described (23, 41).
Human rIFN-γ was a gift from Biogen (Cambridge, MA), human rIL-1β was purchased from Genzyme (Cambridge, MA), and human rIL-1R antagonist (IL-1RA) was purchased from R&D Systems (Minneapolis, MN). Mouse IgG and affinity-purified goat anti-mouse IgG conjugated to FITC were purchased from Southern Biotechnology Associates (Birmingham, AL). mAb to HLA-DR (clone FMN 14) was purchased from Accurate Corp. (Westbury, NY). Mouse anti-human ICAM-1 mAb (clone HA58) conjugated to phycoerythrin was purchased from PharMingen (San Diego, CA). Polyclonal antiserum to STAT-1α was a gift from Berlex Biosciences (Richmond, CA), and monoclonal anti-phosphotyrosine Ab 4G10 and polyclonal STAT-3 antisera were purchased from Upstate Biotechnology (Lake Placid, NY). Polyclonal antisera against IRF-1 and USF-1 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
Immunofluorescence flow cytometry
Cells were plated into 60-mm dishes (Costar, Cambridge, MA) at 1 × 106 cells/dish in complete medium and allowed to adhere overnight. Before stimulation the culture medium was replaced with fresh DMEM/Ham’s F-12 medium supplemented with 1% FBS. Cells were either untreated or treated with IFN-γ for 48 h in the absence or the presence of IL-1β. IL-1β was added 24 h before stimulation with IFN-γ or concurrently with IFN-γ. Cells were trypsinized, washed in cold PBS, and stained for cell surface expression of HLA-DR or ICAM-1 molecules as described previously (24, 42). Negative controls were incubated with irrelevant isotype-matched mouse IgG. We have previously determined that trypsinization does not affect the expression of either HLA-DR or ICAM-1 surface Ags (24, 42). Ten thousand cells were analyzed for each sample using a FACScan flow cytometer. HLA-DR expression is expressed as the percentage of positive cells after subtraction of background staining with irrelevant IgG. Total ICAM-1 expression is expressed in arbitrary units calculated from the percentage of positive cells × mean fluorescence intensity. The ICAM-1 data are expressed in this fashion as cytokine treatment affects both the percentage of positive cells as well as the mean fluorescence intensity (42).
RNA isolation and analysis by RNase protection assay (RPA)
Cells were plated into 100-mm dishes (Costar) at 3 × 106 cells/dish in complete medium and allowed to adhere overnight, then incubated for the indicated time periods with IFN-γ in the absence or the presence of IL-1β. Total cellular RNA was isolated as previously described (24). Preparation and in vitro transcription of HLA-DRα, CIITA, and GAPDH riboprobes have been described previously in detail (23, 24).
RPA was performed using a commercially available kit (Ambion, Austin, TX) according to the manufacturer’s instructions as previously described (23, 24). Ten to twenty micrograms of total RNA were hybridized overnight with riboprobes at 42°C in 20 μl of 40 mM PIPES (pH 6.4), 80% deionized formamide, 400 mM sodium acetate, and 1 mM EDTA. After hybridization, the mixture was treated with RNase A/T1 (1/200 dilution) at 37°C for 30 min. Protected fragments were analyzed by 5% denaturing PAGE (8 M urea), and the gels were exposed to x-ray film for visualization. Quantitation of protected fragments was performed using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA). Values for HLA-DRα and CIITA mRNA were normalized to GAPDH mRNA values, since GAPDH mRNA levels are not affected by cytokine treatment. The sizes of the protected fragments of the HLA-DRα, CIITA, and GAPDH riboprobes in the RPA are 413, 452, and 230 nucleotides in length, respectively.
Nuclear extracts and electrophoretic mobility shift assay (EMSA)
Nuclear extracts from astroglioma cells were prepared as described by Pine et al. (43). Cells were grown in 100-mm dishes, allowed to adhere overnight, and then treated with various combinations of IL-1β and/or IFN-γ. After treatment, cells were washed with cold PBS, harvested by scraping, and pelleted. Cells were resuspended in 5 packed cell volumes (pcv) of buffer A (10 mM KCl, 20 mM HEPES, 1 mM MgCl2, 1 mM DTT, 0.4 mM PMSF, 1 mM NaF, and 1 mM Na3VO4), incubated on ice for 10 min, and pelleted at 1000 × g for 10 min. Pellets were resuspended in 3 pcv of buffer A plus 0.1% Nonidet P-40, incubated on ice for 10 min, and centrifuged at 3,000 × g for 10 min. The nuclear pellet was resuspended in 2 pcv of buffer B (10 mM HEPES, 400 mM NaCl, 0.1 mM EDTA, 1 mM MgCl2, 1 mM DTT, 0.4 mM PMSF, 15% glycerol, 1 mM NaF, and 1 mM Na3VO4) and incubated for 30 min at 4°C with constant gentle mixing. Nuclei were then pelleted at 14,000 × g for 15 min, and extracts were dialyzed for 2 h at 4°C against 1 liter of buffer C (20 mM HEPES, 200 mM KCl, 1 mM MgCl2, 0.1 mM EDTA, 1 mM DTT, 0.4 mM PMSF, 15% glycerol, 1 mM NaF, and 1 mM Na3VO4). Extracts were cleared by centrifugation at 14,000 × g for 15 min at 4°C. Protein concentrations were determined using a Bio-Rad (Richmond, CA) protein assay.
EMSA was performed using the following oligonucleotides. The 19-bp oligonucleotide ICAM-1-GAS has the sequence GAGGTTTCCGGGAAAGCAG and is derived from the human ICAM-1 promoter sequence −66 to −84 (44). The 30-bp oligonucleotide CIITA-GAS has the sequence TGCCACTTCTGATAAAGCACGTGGTGGCCA and corresponds to the type IV CIITA promoter sequence −119 to −148. The 30-bp CIITA-IRF-1 oligonucleotide has the sequence TGCAGAAAGAAAGTGAAAGGGAAAAAGAAC and corresponds to the type IV CIITA promoter sequence −45 to −74 (45). Two-tenths nanogram of 32P-labeled oligonucleotide (20,000 dpm) was incubated for 30 min at room temperature with 7.5 μg of nuclear extract in a volume of 25 μl containing 50 mM KCl, 2.5 mM MgCl2, 1 mM EDTA, 1 mM DTT, 10 mM Tris-Cl (pH 7.5), 8% glycerol, 1 μg of salmon sperm DNA, and 1 μg of poly(dI-dC). For supershift analysis, 1 μl of Ab was incubated with the nuclear extracts for 30 min in binding buffer, followed by an additional incubation for 30 min at room temperature with labeled oligonucleotide. Bound and free DNAs were resolved by electrophoresis through a 6% polyacrylamide gel at 250 V in 1× TGE (50 mM Tris-Cl, 380 mM glycine, and 2 mM EDTA). Dried gels were exposed to Kodak XAR-5 film (Eastman Kodak, Rochester, NY) at −70°C with intensifying screens.
Human CIITA promoter constructs
The sequence for the primers used to PCR amplify a 1703-bp DNA fragment of the type IV promoter of the human CIITA gene was derived from that reported previously (45). The sense primer is located at the 3′ end of the type III promoter and has the sequence GCCTGGCTCCACGCCCTGCTG, and the antisense primer is located at the 3′ end of the type IV promoter and has the sequence CGCTGTTCCCCGGGCTCCCG. PCR was performed with the Taq PCR Core Kit (Qiagen, Santa Clarita, CA) according to the manufacturer’s instructions with 2.5 U of Taq DNA polymerase, 0.1 μM of each primer, and 600 ng of genomic DNA. The PCR amplification protocol consisted of an initial 1-min melting step at 94°C, followed by 30 cycles with 40-s melting at 92°C, 40-s annealing at 60°C, and 1-min 30-s extension at 75°C, except for the last cycle which contained a 5-min extension step. The resulting 1703-bp fragment was gel purified and ligated into linearized pCRII vector (Invitrogen, Carlsbad, CA). The complete sequence of the insert was obtained by automatic sequencing, which was performed by the University of Alabama at Birmingham CFAR Molecular Biology Core Facility. The 1703-bp insert was released from pCRII by digestion with the restriction enzyme EcoRI and gel purified, and the restriction ends were blunted with the Klenow fragment of DNA polymerase I according to the manufacturer (Promega, Madison, WI). The blunt-ended fragment was ligated into the SmaI site of the pGL2-Basic vector, which contains the gene for luciferase as reporter. The designated name for this construct is hCIITAp1.7. Plasmid constructs containing deletions in the promoter (D1, D2, and D4) were prepared as follows. Aliquots of the hCIITAp1.7 construct were subjected to digestion with the restriction enzyme XhoI, giving rise to a 1020-bp XhoI fragment, which was gel purified and subcloned into pGL2-Basic, generating the deletion construct hCIITAp-D1. The hCIITAp1.7 was also digested with SmaI, generating a 301-bp SmaI fragment, or BstXI/KpnI, generating a 229-bp fragment, that was subcloned into pGL2-Basic, creating the deletion constructs hCIITAp-D2 and hCIITAp-D4, respectively. The deletion construct hCIITAp-D3 was generated by religating the hCIITAp1.7 construct after deleting a 535-bp KpnI fragment, then inserting a 99-bp fragment (−24 to +75) into the Mlu1 site (see Fig. 5). A 1281-bp human ICAM-1 promoter construct containing the gene for luciferase as the reporter was also used as a control (46).
Transient transfection and luciferase assay
Twenty micrograms of the hCIITAp1.7 promoter construct, CIITA deletion constructs, or human ICAM-1 promoter construct were cotransfected with 4 μg of the pCMV-β-galactosidase construct into 3 × 106 cells by electroporation with a Bio-Rad Gene Pulser set at 250 V and 960 μF as previously described (42). After transfection, cells were allowed to recover for 12 h before treatment with IL-1β and IFN-γ for various time periods. Cells were washed with PBS and lysed with 200 μl of lysis buffer containing 25 mM trisphosphate (pH 7.8), 2 mM DTT, 2 mM diaminocyclohexane tetraacetic acid, 10% glycerol, and 1% Triton X-100. Extracts were assayed in triplicate for luciferase activity in a volume of 140 μl containing 40 μl of cell extract, 20 mM Tricine, 0.1 mM EDTA, 1 mM magnesium carbonate, 2.67 mM MgSO4, 33.3 mM DTT, 0.27 mM coenzyme A, 0.47 mM luciferin, and 0.53 mM ATP, and light intensity was measured using a luminometer (Promega, Madison, WI). Luciferase activity was integrated over a 10-s period. Extracts were also assayed in triplicate for β-galactosidase enzyme activity as previously described (42). The luciferase activity of each sample was normalized to β-galactosidase activity before calculating the fold induction value.
Levels of significance for comparisons between samples were determined using Student’s t test distribution.
IL-1β inhibits IFN-γ induction of class II MHC expression in human astroglioma cells
The effect of IL-1β on the inducibility of class II MHC Ags by IFN-γ was examined in two human astroglioma cell lines, CH235-MG and U373-MG. CH235-MG cells do not constitutively express class II MHC gene products, but can be induced to do so by stimulation with IFN-γ (23). Levels of class II MHC expression at the cell surface reach a maximum after 48–72 h of stimulation with IFN-γ (data not shown). IL-1β alone does not influence class II MHC expression in CH235-MG cells; however, pretreatment of cells with IL-1β for 24 h followed by IFN-γ stimulation for 48 h resulted in significant suppression (∼85%) of class II MHC expression (Table I). Simultaneous treatment of CH235-MG cells with IL-1β plus IFN-γ also resulted in inhibition of class II MHC expression (∼46%), although the extent of inhibition was greatest with IL-1β pretreatment. These findings indicate that IL-1β can inhibit IFN-γ-induced class II MHC expression, and that the inhibitory effect of IL-1β occurs in a time-dependent manner. Although IFN-γ induction of class II MHC was less potent in U373-MG cells than in CH235-MG cells, a 24-h pretreatment with IL-1β resulted in significant suppression (∼54%) of class II expression (Table I). Interestingly, in U373-MG cells, simultaneous addition of IL-1β and IFN-γ did not inhibit class II MHC expression. Dose-response experiments were conducted with IL-1β (10 pg/ml to 2 ng/ml) to determine the optimal concentration for inhibition; the results indicated that maximal inhibition was observed using 0.5–1 ng/ml of IL-1β (data not shown); thus, 1 ng/ml of IL-1β was used for the remainder of the study.
|Cell Treatment .||% Positive Cells .||% Inhibition .|
|IFN-γb||38.2 ± 5.9f|
|IL-1β/IFN-γd||5.7 ± 5.8g||85%j|
|IL-1β plus IFN-γe||20.6 ± 7.1g||46%j|
|Controla||1.5 ± 1.3h|
|IFN-γb||18.6 ± 4.4|
|IL-1βc||1.6 ± 0.5|
|IL-1β/IFN-γd||8.5 ± 2.7i||54%j|
|IL-1β plus IFN-γe||18.3 ± 3.6||—|
|Cell Treatment .||% Positive Cells .||% Inhibition .|
|IFN-γb||38.2 ± 5.9f|
|IL-1β/IFN-γd||5.7 ± 5.8g||85%j|
|IL-1β plus IFN-γe||20.6 ± 7.1g||46%j|
|Controla||1.5 ± 1.3h|
|IFN-γb||18.6 ± 4.4|
|IL-1βc||1.6 ± 0.5|
|IL-1β/IFN-γd||8.5 ± 2.7i||54%j|
|IL-1β plus IFN-γe||18.3 ± 3.6||—|
Control medium for 72 h.
Medium for 24 h, then IFN-γ (100 U/ml) for 48 h.
IL-1β (1 ng/ml) for 72 h.
IL-1β for 24 h, then IFN-γ for 48 h.
Medium for 24 h, then IL-1β plus IFN-γ for 48 h.
Mean ± SD of five experiments.
Significantly different from IFN-γ (p < 0.001).
Mean ± SD of three experiments.
Significantly different from IFN-γ (p < 0.05).
Compared with IFN-γ alone.
We next wished to determine the specificity of IL-1β suppression of class II MHC expression. To do this, we used IL-1RA. As shown in Fig. 1, IL-1RA, in a dose-dependent manner, reversed the inhibitory effect of IL-1β on IFN-γ-induced class II MHC expression. We examined the expression of other gene products that are modulated by IFN-γ and/or IL-1β to determine whether the inhibitory effect of IL-1β was restricted to class II MHC expression. IFN-γ is a modest inducer of ICAM-1 on CH235-MG cells, while IL-1β induces expression to a greater extent (Fig. 2). Preincubation of cells with IL-1β for 24 h and then exposure to IFN-γ for 48 h result in a synergistic effect on ICAM-1 expression (Fig. 2), which is abrogated by the inclusion of IL-1RA. Furthermore, we have previously determined that IL-1β and IFN-γ individually are weak inducers of RANTES production by CH235-MG cells, yet synergize for strong induction of RANTES gene expression (47). These results indicate that IL-1β does not globally inhibit IFN-γ responses in CH235-MG cells.
IL-1β inhibits IFN-γ-induced class II MHC mRNA expression
To study the mechanism by which IL-1β inhibits class II MHC surface expression, we sought to determine whether IL-1β exerts its effect at the level of class II MHC mRNA expression. CH235-MG cells were incubated with IFN-γ alone for 12–48 h or were pretreated with IL-1β for 24 h, then incubated with IFN-γ for 12–48 h. RNA was isolated and analyzed by RPA for class II MHC mRNA expression. Fig. 3,A shows the time course of class II MHC mRNA induction after IFN-γ treatment; optimal levels are detected 48 h after stimulation with IFN-γ (lane 10). IL-1β pretreatment inhibits IFN-γ induction of class II MHC mRNA at all time points tested, although the inhibition is more pronounced at the later time points (∼70% at 36 h and ∼80% at 48 h; Fig. 3 B). Thus, the inhibitory influence of IL-1β is evident on class II MHC mRNA expression. The RPA shown in this figure is overexposed for GAPDH because the signal for class II MHC is weaker. However, quantitation of the original gels was performed on a Phosphor Imager to arrive at accurate values. Comparable results were obtained using U373-MG cells (data not shown).
IFN-γ-induced CIITA mRNA expression is inhibited by IL-1β
The regulation of class II MHC gene expression occurs predominantly at the transcriptional level, and the CIITA protein is a potent activator of class II MHC transcription (for review, see 2 . To determine whether IL-1β inhibits IFN-γ-induced class II MHC mRNA levels by preventing CIITA mRNA induction, we examined the effect of IL-1β pretreatment on IFN-γ-induced CIITA mRNA expression. Consistent with class II MHC expression, CH235-MG cells are constitutively negative for CIITA (Fig. 4,A, lane 1), but can be induced to express CIITA mRNA by IFN-γ (lane 2). IL-1β pretreatment, in a dose-dependent manner, inhibits IFN-γ-induced CIITA mRNA levels (lanes 3–7). Strong inhibition is observed using 300 and 1000 pg/ml IL-1β (Fig. 4 B), which are comparable to the IL-1β concentrations needed for inhibition of IFN-γ-induced class II MHC mRNA and protein expression. The RPAs for CIITA mRNA expression are overexposed for GAPDH because the signal for CIITA is much weaker. Comparable findings were observed in U373-MG cells (data not shown).
IL-1β suppresses IFN-γ inducibility of the type IV CIITA promoter
The CIITA gene is controlled by four independent CIITA promoters, leading to CIITA transcripts with four distinct first exons (45). Two promoters direct constitutive expression in dendritic cells and B cells, respectively, while another mediates IFN-γ-induced expression. The type IV CIITA promoter, which is the IFN-γ-responsive promoter, contains a number of potential cis-acting elements. These include an NF-GMa site, a GAS element, an E box, and an IRF-1 site in both the mouse and human promoters. In addition, the type IV human promoter contains an NF-κB site, and the mouse promoter has two AP-1 sites (45). We wished to determine whether IL-1β could act directly on the type IV CIITA promoter and interfere with IFN-γ induction of CIITA gene transcription. To accomplish this, the full-length IFN-γ-inducible type IV promoter of the human CIITA gene was cloned (see details in Materials and Methods). The schematic diagram shown in Fig. 5 depicts cytokine response elements that have been identified through the Mat Inspector program (48). Interestingly, three clusters of GAS and IRF elements are found in the type IV CIITA promoter. The proximal cluster is located −55 to −142 bp upstream of the transcription start site (pIRF, pGAS), the medial cluster is located −591 to −683 bp (mIRF, mGAS), and the distal cluster is located −769 to −854 bp upstream of the transcription start site (dIRF, dGAS). To analyze CIITA promoter activity in response to IFN-γ, the 1703-bp fragment was subcloned into the pGL2-Basic vector to drive expression of the luciferase reporter gene. The resulting plasmid construct, hCIITAp1.7, was transiently transfected into U373-MG astroglioma cells, and IFN-γ induction of luciferase activity was determined (Fig. 6). A very low basal transcriptional activity of the hCIITAp1.7 construct was detected, and a 5.7-fold induction of CIITA promoter activity was observed upon stimulation with IFN-γ. To determine whether all three GAS/IRF elements were required to confer IFN-γ inducibility on the CIITA promoter, deletion constructs were generated from the 5′ and 3′ ends of the full-length hCIITAp1.7 construct, and their inducibility by IFN-γ was examined (see Figs. 5 and 6). The hCIITAp-D1 was tested to determine whether deletion of 683 bp upstream of the distal GAS element had any influence on IFN-γ inducibility. Similar levels of IFN-γ-induced luciferase activity were observed with hCIITAp-D1 (4.9-fold induction) compared with hCIITAp1.7. The hCIITAp-D2 is deleted of the distal and medial GAS/IRF elements and contains the NF-κB, NF-GMa, proximal GAS, E box, and proximal IRF-1 elements. Using this construct, IFN-γ inducibility was maintained (6.2-fold induction), indicating that 237 bp of the type IV CIITA promoter is sufficient to confer IFN-γ inducibility. Further deletion of the NF-κB element in hCIITAp-D4 did not significantly influence IFN-γ inducibility of the type IV CIITA promoter (5.1-fold induction). The hCIITAp-D3 construct contains the distal and medial GAS/IRF elements and lacks the downstream 437-bp region of the type IV CIITA promoter. The hCIITAp-D3 was not inducible by IFN-γ treatment, indicating that the distal and medial GAS/IRF elements cannot mediate CIITA transcription in response to IFN-γ. These results collectively indicate that the IFN-γ inducibility of the CIITA gene in astrocytes is dependent on elements contained within a 154-bp fragment of the type IV CIITA promoter.
We next tested whether IL-1β could inhibit IFN-γ-induced CIITA transcription. The hCIITAp1.7, hCIITAp-D1, hCIITAp-D2, and hCIITAp-D4 constructs were transfected into U373-MG cells, and IFN-γ-inducible activity was tested after pretreating the cells with IL-1β. IL-1β treatment inhibited IFN-γ-induced hCIITAp1.7 promoter activity by approximately 70% (Fig. 6). This inhibitory effect was still evident when the hCIITAp-D1 construct was used (∼53% inhibition; Fig. 6). As well, IFN-γ induction of the hCIITAp-D2 construct was inhibited by about 60% in the presence of IL-1β (Fig. 6). We tested the effect of IL-1β on the hCIITAp-D4 construct to determine whether the inhibitory effect was mediated through the NF-κB element. IL-1β inhibited IFN-γ-induced activation of the hCIITAp-D4 construct by approximately 70%, indicating that the NF-κB element is not involved in IL-1β inhibition of CIITA promoter activity (Fig. 6). IL-1β treatment alone had no effect on the transcriptional activation of any of the CIITA promoter constructs (data not shown). Thus, IL-1β inhibits IFN-γ-induced transcription of the CIITA gene, and the element(s) mediating the inhibitory response resides within the 154-bp region of the type IV CIITA promoter. The CIITA promoter constructs were also tested in CH235-MG cells; however, the cells did not survive the transfection procedure well, and reproducible results could not be obtained. To confirm that treatment with IL-1β lead to selective inhibition of IFN-γ-induced CIITA promoter activity, we tested the effect of IL-1β on IFN-γ-induced ICAM-1 promoter activity in U373-MG cells. We have previously shown that a 1281-bp human ICAM-1 promoter is responsive to both IFN-γ and IL-1β stimulation (46). U373-MG cells were transfected with the human ICAM-1 promoter construct, then incubated with medium for 34 h, with IL-1β (1 ng/ml) for 34 h, with medium for 24 h, then with IFN-γ (100 U/ml) for 10 h, or with IL-1β for 24 h, then with IFN-γ for 10 h, which are the exact conditions used for IL-1β suppression of CIITA promoter activity (shown in Fig. 6). ICAM-1 promoter activity in U373-MG cells was enhanced in the presence of IFN-γ (2.5-fold induction) and IL-1β (1.8-fold induction), and a 5.2-fold increase was observed when cells were pretreated with IL-1β for 24 h, then exposed to IFN-γ for 10 h (data not shown). These results indicate that IL-1β pretreatment has a selective inhibitory effect on type IV CIITA promoter activity.
Influence of IL-1β on IFN-γ activation of STAT-1α
Macrophages from STAT-1α-deficient mice cannot be induced to express CIITA or class II MHC in response to IFN-γ stimulation (49), and in cells depleted of STAT-1α by antisense oligonucleotides, IFN-γ induction of CIITA and class II MHC is impaired (23). Furthermore, the proximal GAS element in the type IV CIITA promoter has been shown to bind STAT-1α and is critical for IFN-γ-inducible activation of the CIITA promoter (50). To investigate whether IL-1β treatment interferes with IFN-γ induction of STAT-1α activation, we first tested whether IL-1β influenced the ability of IFN-γ to induce tyrosine phosphorylation of STAT-1α. CH235-MG or U373-MG cells were incubated with IFN-γ in the absence or the presence of IL-1β, then tyrosine phosphorylation of STAT-1α was assessed. A 30-min incubation with IFN-γ induces tyrosine phosphorylation of STAT-1α, and neither a 24-h pretreatment with IL-1β nor simultaneous addition of IL-1β with IFN-γ affected IFN-γ-induced tyrosine phosphorylation of STAT-1α (data not shown). Thus, IL-1β does not interfere with the ability of IFN-γ to induce tyrosine phosphorylation of STAT-1α in the astroglioma cells.
It is possible that IL-1β may affect the ability of tyrosine-phosphorylated STAT-1α to translocate into the nucleus and bind to the GAS element. To test this, nuclear extracts were prepared from CH235-MG cells stimulated with various combinations of IFN-γ and IL-1β, and binding to an oligonucleotide containing the proximal GAS element from the type IV CIITA promoter was assessed. It should be noted that in addition to the GAS element, this oligonucleotide contains an E box sequence that binds the constitutively expressed transcription factor USF-1, which is also critical for IFN-γ induction of the CIITA promoter (50). As shown in Fig. 7,A, nuclear extracts from unstimulated cells formed a DNA-protein complex (lane 1, complex B), and IFN-γ induced the formation of a slower migrating complex (lane 4, complex A). The specificity of complex formation was examined by competition experiments using a 100-fold molar excess of CIITA-GAS oligonucleotide as well as an oligonucleotide containing the GAS element from the human ICAM-1 promoter. Complex B was competed away by the CIITA-GAS oligonucleotide (Fig. 7,A, lane 2), but not by the ICAM-1-GAS oligonucleotide (lane 3). Competition of IFN-γ-stimulated extracts revealed that the CIITA-GAS oligonucleotide competed away both complex A and complex B (lane 5), while the ICAM-1-GAS oligonucleotide competed for only complex A (lane 6). As determined by supershift analysis, complex B is composed of USF-1 (data not shown), while complex A is STAT-1α (see Fig. 7 B, lanes 5 and 6). We interpret the competition results as follows. Complex B binds to the E box contained within the CIITA-GAS oligonucleotide and thus is competed away by excess CIITA-GAS oligonucleotide. The ICAM-1-GAS oligonucleotide, which does not contain an E box element, does not compete for complex B. In the IFN-γ-stimulated extracts, both complexes A and B are effectively competed away by CIITA-GAS oligonucleotide, while the ICAM-1-GAS oligonucleotide competes away only complex A (STAT-1α).
The influence of IL-1β on STAT-1α and USF-1 binding was next examined. IL-1β treatment alone for 24 h did not affect USF-1 (complex B), nor were any other complexes formed over the CIITA-GAS oligonucleotide (Fig. 7,B, compare lanes 1 and 3). Translocation of tyrosine-phosphorylated STAT-1α from the cytoplasm to nucleus and subsequent binding to the CIITA-GAS element were not influenced by IL-1β (Fig. 7 B, compare lanes 5 and 8). Anti-STAT-3 Ab was included as a control for the anti-STAT-1α Ab. Taken together, these data indicate that IL-1β does not alter the constitutive expression of USF-1 (complex B), nor does IL-1β treatment affect IFN-γ-mediated signal transduction pathways that culminate in the activation of STAT-1α (complex A).
Influence of IL-1β on IRF-1 binding to the CIITA promoter
The transcription factor IRF-1 has been shown to be critical for IFN-γ inducibility of CIITA and class II MHC gene expression (50, 51). Since the IL-1β inhibitory response localized to the 154-bp region of the type IV CIITA promoter that contains the proximal IRF-1 binding site, we assessed whether IL-1β influenced IRF-1 expression. IFN-γ stimulation induced the formation of two DNA-protein complexes (Fig. 8, compare lanes 1 and 2), and complex 1 was supershifted upon the addition of Ab against IRF-1 (lane 3). Pretreatment of cells with IL-1β for 24 h, then exposure to IFN-γ for 2 h, did not affect complex 1 formation (lane 4). IL-1β treatment alone did not induce complex formation over the CIITA proximal IRF-1 element (data not shown). These results indicate that IL-1β does not interfere with the ability of IFN-γ-induced IRF-1 to bind to the proximal IRF-1 element in the type IV CIITA promoter.
We have investigated the potential of IL-1β in modulating the induction of class II MHC Ags by IFN-γ. In contrast to the widely appreciated proinflammatory effects of this cytokine (for review, see 52 , we found that IL-1β inhibited IFN-γ induction of class II MHC expression on human astroglioma cell lines. Transcription of class II MHC genes by IFN-γ depends on the induction of CIITA (16). IL-1β treatment was found to diminish CIITA mRNA induction after IFN-γ treatment. Furthermore, using a variety of human type IV CIITA promoter constructs, we demonstrated that IL-1β inhibited the IFN-γ inducibility of this promoter. These results collectively indicate that IL-1β inhibits IFN-γ induction of class II MHC expression by inhibiting CIITA gene transcription.
Muhlethaler-Mottet et al. (50) have recently demonstrated that three cis-acting elements within the type IV CIITA promoter are essential for activation by IFN-γ. They are the proximal GAS site, which binds STAT-1α; the E box (adjacent to the proximal GAS site), which binds USF-1; and the proximal IRF element, which binds IRF-1. STAT-1α and IRF-1 binding occur following IFN-γ stimulation, while USF-1 is constitutively expressed. Furthermore, STAT-1α binds to the proximal GAS element only in the presence of USF-1, and both factors bind cooperatively to the GAS/E box motif in the type IV CIITA promoter. Our results in the human astroglioma cell lines indicate that a 154-bp fragment of the type IV CIITA promoter contains all of the elements necessary for responsiveness to IFN-γ (proximal GAS element, E box, and proximal IRF element). Interestingly, the distal and medial GAS/IRF elements do not contribute to IFN-γ activation of the type IV CIITA promoter, since deletion of these elements had no effect on CIITA gene expression by IFN-γ (Fig. 6). As well, the distal and medial GAS/IRF elements are not sufficient to mediate IFN-γ activation of the CIITA promoter, since construct hCIITAp-D3 is not activated in response to IFN-γ (Fig. 6). It should be noted that neither distal nor medial GAS elements have an adjacent E box motif, which may explain the lack of function of the hCIITAp-D3 construct. As well, using probes containing either distal or medial GAS sequences, only very weak binding of STAT-1α could be detected by EMSA (data not shown). Again, the absence of the E box sequence probably explains the lack of STAT-1α binding to those sites and reinforces the contention that it is the cooperative interaction between STAT-1α and USF-1 that controls the specific activation of the type IV CIITA promoter by IFN-γ (50).
IL-1β inhibition of IFN-γ-induced CIITA gene transcription was observed using the hCIITAp-D2 construct, which contains 237 bp of the CIITA promoter including the NF-κB, NF-GMa, proximal GAS, E box, and proximal IRF-1 elements (see Fig. 5). To determine whether the NF-κB element was involved in mediating the inhibitory effect of IL-1β, we tested the hCIITA-D4 construct, which lacks the NF-κB element. As shown in Fig. 6, this construct was still inhibited in the presence of IL-1β (70% inhibition). These results suggest that the inhibitory influence of IL-1β is mediated independently of the NF-κB element in the type IV CIITA promoter. As the hCIITAp-D4 construct contains the proximal GAS element, E box, and proximal IRF-1 element, we wished to determine whether IL-1β treatment impaired expression/binding of STAT-1α, USF-1, and/or IRF-1. As shown in this study, IL-1β did not interfere with IFN-γ-induced signal transduction events (i.e., tyrosine phosphorylation of STAT-1α), nor were translocation of STAT-1α to the nucleus and binding to the proximal CIITA GAS element impaired (Fig. 7). As well, IL-1β did not affect the ability of USF-1 to bind to the E box or interact with STAT-1α (Fig. 7). In addition, we found no effect of IL-1β on the induction of the IFN-γ-activated transcription factor, IRF-1, as assessed by EMSA using an oligonucleotide containing the proximal IRF sequence (Fig. 8). To date, our results demonstrate that IL-1β does not affect the expression of the three transcription factors known to be essential for IFN-γ inducibility of the type IV CIITA promoter (STAT-1α, USF-1, and IRF-1), nor does IL-1β inhibition require the NF-κB element of the CIITA promoter. Other possibilities for the mechanism of IL-1β inhibition include utilization of the NF-GMa element (which is intact in the hCIITA-D4 construct) or influencing other as yet unknown transcription factors involved in CIITA gene expression. Future studies will be directed toward the elucidation of the cis- and trans-acting factors involved in IL-1β-mediated inhibition of CIITA transcription.
A number of cytokines have been shown to inhibit class II MHC expression, including TGF-β, IFN-β, and IL-10 (24, 27, 29, 33, 34, 36). However, the mechanisms employed by each of these factors in regulating class II MHC expression differ. IL-10 inhibits class II MHC expression on the cell surface of human monocytes by preventing class II MHC molecules from reaching the plasma membrane as well as by causing an accumulation of internalized class II MHC complexes in intracellular vesicles (34). Thus, the inhibitory effect of IL-10 occurs on post-translational events, both the exocytosis and the recycling of class II MHC molecules. IFN-β was shown to inhibit IFN-γ-induced class II MHC transcription without affecting CIITA mRNA induction (33). Since the inhibitory effect of IFN-β on HLA-DRA expression required the presence of interferon-stimulated gene factor 3γ, it was suggested that IFN-β induces a gene product that interferes with CIITA trans-activation of the HLA-DRA promoter, or that the CIITA protein is not fully functional in IFN-β-treated cells. In contrast, the inhibitory effect of TGF-β on IFN-γ-induced class II MHC expression is mediated via transcriptional inhibition of the promoter of class II MHC genes (27, 29), which is caused by TGF-β suppression of IFN-γ-induced CIITA mRNA expression (24). TGF-β does not destabilize CIITA mRNA, suggesting an effect at the level of CIITA mRNA transcription. Indeed, a recent report has demonstrated that TGF-β inhibits IFN-γ-mediated induction of CIITA promoter activity in fibroblasts (36). The results obtained with IL-1β in this study demonstrate similarities with those of TGF-β; i.e., both cytokines inhibit IFN-γ induction of CIITA promoter activity, thereby inhibiting CIITA mRNA expression. IL-1 and TGF-β use distinct receptors and downstream adaptor proteins to mediate their pleiotropic effects (for review, see Refs. 53 and 54). It will be of interest to determine whether IL-1β and TGF-β signaling events converge at the CIITA promoter to inhibit IFN-γ-induced expression of this gene.
During inflammatory responses in the CNS, IL-1β is one of the first cytokines to be produced at sites of inflammation (for review, see 55 . In general, IL-1 is considered to be a proinflammatory cytokine due to its pronounced stimulatory effect on the expression of a variety of proinflammatory molecules, such as the adhesion molecules ICAM-1 and VCAM-1, chemokines, TNF-α, and nitric oxide synthase (for review, see 52 . The ultimate in vivo outcome of the suppressive effect of IL-1β on class II MHC expression in astrocytes is unclear at this time due to the controversy regarding the role of the astrocyte as an APC within the CNS. Class II MHC-positive astrocytes have been shown by some to function as APCs in vitro (56, 57, 58, 59), while other groups report that class II MHC-positive astrocytes are unable to induce T cell proliferation (60, 61). There are also reports that class II MHC-positive astrocytes transmit a suppressive and/or apoptotic signal to CD4+ T cells (60, 62), possibly due to the lack of B7 expression. However, a recent study demonstrated that class II MHC-positive astrocytes are effective APCs for Th2 cell activation (57). If class II MHC-positive astrocytes do, in fact, transmit an apoptotic signal to autoreactive T cells, then inhibition of class II MHC expression by IL-1β could be viewed as perpetuating immune responsiveness within the CNS. As well, if the function of class II MHC-positive astrocytes is to promote activation of Th2 cells and secretion of IL-4 and IL-10 (57), then inhibition of this response by IL-1β may also be detrimental within the CNS. However, if class II MHC-positive astrocytes do promote the activation of naive autoreactive Th1 cells (58), then IL-1β down-regulation of class II MHC expression on these cells could inhibit aberrant immune responses within the CNS. Understanding the mechanism by which IL-1β affects CIITA expression, class II MHC expression, and the subsequent APC function of the astrocyte may aid in the control of inflammatory processes in the CNS.
We thank Sue B. Wade for excellent secretarial assistance, and Dr. Yi-Ju Lee (University of Manchester) for help with the initiation of this project.
This work was supported by National Multiple Sclerosis Society Grant RG-2205-B-9 (to E.N.B.) and National Institutes of Health Grant NS-36765 (to E.N.B.). We acknowledge the support of the University of Alabama at Birmingham Flow Cytometry Core Facility (Grant AM-20614) and the University of Alabama at Birmingham Center for AIDS Research Molecular Biology Core Facility (Grant AI-27767).
Abbreviations used in this paper: CNS, central nervous system; CIITA, class II transactivator; IL-1RA, interleukin-1 receptor antagonist; IRF-1, interferon-regulatory factor-1; USF-1, upstream stimulatory factor-1; RPA, ribonuclease protection assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; EMSA, electrophoretic mobility shift assay; pcv, packed cell volume; GAS, IFN-γ activation sequence.