The Interleukin-1 Type 2 Receptor Gene Displays Immediate Early Gene Responsiveness in Glucocorticoid-stimulated Human Epidermal Keratinocytes*

Human epidermal keratinocytes (HEKs) in primary culture (P2–P4) were used to study glucocorticoid (GC)-mediated transcription of the genes encoding the constitutively expressed interleukin-1 type 1 receptor (IL-1R1) and the inducible interleukin-1 type 2 receptor (IL-1R2). Utilizing Northern dot blot analysis and a quantitative reverse transcription-polymerase chain reaction protocol for IL-1R1 and IL-1R2, dexamethasone and, in particular, the budesonide epimer R were shown to effectively and rapidly induce transcription from the IL-IR2 gene when compared with IL-1R1 or β-actin RNA message levels in the same sample. Southern blot analysis of newly generated IL-1R2 reverse transcription-polymerase chain reaction products using end-labeled IL-1R2 intron probes suggested that GC enhancement of IL-1R2 expression was regulated primarily at the level of de novotranscription. GC-induced IL-1R2 gene transcription displayed features characteristic of a classical immediate early gene response, including a signal transduction function, a relatively low basal abundance, a rapid, transient induction, cycloheximide superinduction, actinomycin D suppression, and a rapid decay of IL-1R2 RNA message. Parallel time course kinetic analysis of IL-1R2 RNA message levels with Western immunoblotting revealed tight coupling of de novo IL-IR2 gene transcription with translation of the IL-1R2 RNA message; a newly synthesized (∼46-kDa) IL-1R2 protein was detected in the HEK growth medium as early as 1 h after budesonide epimer R treatment. These data indicate that different GC compounds can variably up-regulate the IL-1R2 response in HEKs through transcription-mediated mechanisms and, for the first time, suggest that a gene encoding a soluble cytokine receptor can respond like an immediate early gene.

The immunosuppressive and anti-inflammatory activities of glucocorticoids (GCs) 1 are not well understood, due largely to limited knowledge of their complex activities at the level of gene expression. Among the cytokines that mediate the cellular immune and inflammatory response, the interleukin-1 (IL-1) signaling system plays a central role in diverse cell types (1)(2)(3). The ϳ17-kDa IL-1 affects target cells initially through two distinct types of transmembrane receptor complexes: (a) an integral type 1 (ϳ80-kDa) IL-1 receptor (IL-1R1) protein, and (b) an external type 2 (ϳ68-kDa) IL-1 receptor (IL-1R2; Refs. 4 -6) protein. IL-1R1 and IL-1R2 proteins, which are both members of the immunoglobulin gene superfamily, each share virtual identity in their extracellular IL-1 ligand binding domain and transmembrane anchor; however, the IL-1R1 has a 213-amino acid cytoplasmic domain essential for signal transduction, whereas the IL-1R2 has a 29-amino acid cytoplasmic terminus incapable of signal transfer (2)(3)(4)(5)(6). Signal transduction through the IL-1R1 can be relayed via cytosolic IL-1R-associated kinase cascades, leading to IB phosphorylation and degradation, followed by NF-B action on target NF-B-regulated gene promoters (7)(8)(9). These target genes can include those coding for: (a) specific IL-1-converting enzymes (10), (b) proteases that cleave the soluble extracellular domain of IL-1R2 (11); (c) the de novo expression of IL-1R1, IL-1R2, and related growth factor genes (2,11,12); and (d) transcriptional activator proteins such as AP1, NF-IL6, and NF-B, which, in turn, induce transcription from IL-1-sensitive, pro-inflammatory IEGs, such as cyclooxygenase-2 (13,14). At least two distinct ligand-mediated mechanisms can modulate autocrine, juxtacrine, or paracrine IL-1 stimulation: (a) the cellular production of the ϳ20-kDa IL-1 receptor antagonist, which competes with the binding of secreted IL-1 to IL-1R1 at the cell surface (15), and (b) the de novo expression of the non-signal-transducing IL-1R2 "decoy" receptor which, through its prominent cysteine-rich extracellular domain, acts as a local membranebound scavenger of IL-1 (but not IL-1 receptor antagonist) or serves as a soluble ϳ46-kDa molecular sink for IL-1 (4, 11) by virtue of an extracellular protease cleavage site.
Here we have studied the effects of the GCs dexamethasone (DEX), the novel nonhalogenated budesonide epimer R (BUDeR), and an equimolar racemic mixture of budesonide S and R (BUDr; Fig. 1) on IL-1R1 and IL-1R2 RNA message and receptor protein induction using human epidermal keratinocytes (HEK) cells at passages P2-P4 in a primary culture test system. HEK cells respond to a wide variety of cytokines and lipid mediators (12,14,16) and provide a useful model to study IL-1 biology because they both synthesize and secrete IL-1 and respond to it via intrinsic IL-1R1-and IL-1R2mediated pathways (17,18). 2 Northern dot blot analysis, quantitative reverse transcription-polymerase chain reaction (RT-PCR), and multiplex RT-PCR using human-specific IL-1R1, IL-1R2, and ␤-actin primer sets radiolabeled to high specific activity (Ͼ10 9 dpm/g) indicated that after only 1 h of HEK cell stimulation, DEX and, in particular, BUDeR elicited a strong induction of the non-signal-transducing IL-IR2 RNA message but not of the IL-1R1 transcript. Southern blotting and RT-PCR unprocessed transcript assay (19,20) indicated that IL-1R2 RNA increases were primarily the result of de novo IL-1R2 gene transcription. Temporal analysis of IL-1R2 RNA message abundance with newly synthesized IL-1R2 protein indicated that the induction of the IL-1R2 gene expression pathway was a relatively rapid event, fulfilling each of the criteria for the classical cellular IEG response (21). A ϳ46-kDa IL-1R2 protein was detected by Western immunoblot analysis in HEK whole cell extracts (WCXTs) and in 200-fold concentrated HEK extracellular growth medium as early as 1 h, and increasing to 12ϩ h after BUDeR treatment. These results suggest that rapidly induced IL1-R2 gene transcription, tight coupling to IL-1R2 message translation, and secretion of the IL1-R2 protein can contribute to the anti-inflammatory potential of GC compounds in HEKs by scavenging and antagonizing extracellular cytokine IL-1 activity. Enhanced local up-regulation of the IL-1R2 signaling pathway by nonhalogenated 16␣,17␣substituted GCs such as BUDeR may make these compounds particularly useful anti-inflammatory agents in epidermal keratinocytes that play key structural and physiological roles in primary immune defense and the inflammatory response in vivo.
HEK Cells in Culture-HEKs respond productively to a wide variety of cytokines, lipid mediators, and GCs. Cryopreserved normal HEK cells (obtained from pooled donors (Clonetics CC-2504; HEK-Neo pooled) and received as frozen primary cultures) were grown to 80% confluence (ϳ2-3 ϫ 10 6 HEK cells/T-75 flask) in 20 ml of KGM supplemented with a serum containing 7.5 mg/ml bovine pituitary extract, 0.1 g/ml human epidermal growth factor, 5 mg/ml insulin, 50 g/ml gentamycin/amphotericin, and 0.5 mg/ml hydrocortisone in T-75 flasks at 37°C in 5% CO 2 according to the manufacturer's specifications (Clonetics). HEK cells responded well to ligands between P2 and P4, after which there was a graded decline in both the growth rate and responsiveness to ligand induction. The HEK cell cultures used here, which were obtained from pooled neonatal or adult donors, revealed little difference in their individual responses to either DEX, BUDr, or BUDeR stimulation, as measured by the induction of the IL-1R2 RNA message (Figs. [3][4][5][6]. For maximal ligand induction, cells were deprived for 24 h to 0.5% of normal serum levels before the addition of GC test compounds. 100 nM (final concentration) DEX, BUDeR, or BUDr were added to HEKs, and cells were harvested at 0, 1, 3, 6, 12, and 24 h. For each treatment condition, HEKs were cultured in duplicate T-75 flasks for the par-2 N. G. Bazan and W. J. Lukiw, unpublished observations. allel isolation of either total RNA for Northern blot and RT-PCR analysis or WCXTs for Western immunoblot detection.
Harvesting of HEK and Preparation of WCXTs-All extraction procedures were performed at 4°C on wet ice. After each incubation period with GC test compounds, the KGM was decanted and replaced with 20 ml of Dulbecco's phosphate-buffered saline (Life Technologies, Inc.) containing 1 mM phenylmethylsulfonyl fluoride (Sigma P-7626), 0.05 g/l aprotinin (Sigma A-6279), and 0.025 g/l leupeptin (Sigma L-2884). HEKs were scraped into a suspension of phosphate-buffered saline containing these enzyme inhibitors and pelleted at 4°C by centrifugation at 1400 ϫ G av for 10 min. HEK cellular pellets were gently resuspended in 200 l of a hypotonic buffer consisting of 20 mM HEPES (pK a ϭ 7.55 at 20°C), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 1% (v/v) aprotinin (which was made 0.5% (v/v) with SDS (Sigma L-4509) just before use). HEK WCXTs were thoroughly resuspended using a 200 l Pipetman with a cut-off tip followed by repeated pipetting on wet ice for 5 min. HEK WCXT protein concentrations were determined using a dotMETRIC protein microassay (Chemicon, Temecula, CA; dot sensitivity, 0.3 ng of protein) using bovine serum albumin (bovine serum albumin radioimmunoassay grade; Sigma A-7888) as a standard.
Isolation of Total Keratinocyte RNA-After each GC induction time point, the KGM was completely replaced with 7.5 ml of TRIzol reagent (Refs. 22 and 23; Life Technologies, Inc. catalog number 15596-026). Samples were then incubated for 5 min, with gentle shaking at 25°C to dissociate nucleoprotein complexes. After the addition of 1.5 ml of reagent grade chloroform, suspensions were transferred to sterile, diethylpyrocarbonate (Sigma D5758)-treated 15-ml conical tubes, vigorously shaken for 5 min, incubated at 25°C for 15 min, and centrifuged at 5000 ϫ G av for 15 min at 4°C. The upper aqueous phase, containing total extracted RNA, was transferred into sterile RNase-free 15-ml conical tubes, and 4 ml of 100% (v/v) ACS reagent grade iyl alcohol were added. RNA was precipitated overnight at Ϫ80°C and then centrifuged at 9000 ϫ G av for 10 min at 0°C. Supernatants were vacuum aspirated, and visible RNA pellets were washed twice with 10 ml of 80% ACS reagent grade ethanol (v/v) and pelleted by centrifugation at 9000 ϫ G av for 10 min at 4°C. Supernatants were aspirated again, and the total RNA pellet was dried under a vacuum for ϳ10 min over a bed of anhydrous calcium sulfate (Drierite; W. A. Hammond). The partially dessicated RNA pellet was then resuspended in RNase-free water to ϳ1 g/l by incubating for 15 min at 60°C in an Eppendorf 5436 thermomixer. Total RNA concentrations were determined spectrophotometrically at A 260 , and samples had A 260 :A 280 ratios of Ն2.1.
Western Analysis and Immunoblotting-30 g of control or GCtreated HEK WCXTs or HEK KGM concentrated 200-fold using Dulbecco's phosphate-buffered saline and Centricon-10 concentrators (Amicon, Beverly, MA) were analyzed under reducing conditions on 10% acrylamide Tris-glycine SDS gels (Bio-Rad Ready-Gels 161-0907); proteins were transferred onto Hybond-P (Amersham RPN2020F) polyvinylidene difluoride transfer membranes using a mini-Trans-Blot electrophoretic transfer cell (Bio-Rad 170-3930). Membranes were blocked and probed with an anti-human IL-1R2 rat monoclonal primary antibody (Cdw121b) that exhibited no cross-reactivity with human IL-1R1 (Genzyme, Cambridge, MA). Bound primary antibodies were detected with an anti-rat IgG peroxidase-linked secondary antibody (Amersham NA932) and developed with an ECL Plus Western blotting analysis system, according to the manufacturer's instructions (Amersham RPN2132).  -TGC-CTG-AGG-TCT-TGG-AAA-AAC-3Ј  IL-1R1 reverse (exon 12)  5Ј-TGT-GGT-CCC-TGT-GTA-AAG-TCC-3Ј  340  IL-1R2 forward (exon 4)  5Ј-TCC-ATG-TGC-AAA-TCC-TCT-CTT-3Ј  IL-1R2 reverse (exon 8)  5Ј-TCC-TGC-CGT-TCA-TCT-CAT-ACC-3Ј  Data Analysis and Quantitation-For cold multiplex-PCR, agarose gels were stained in 0.5 g/ml ethidium bromide, and images were either photographed using Polaroid instant photography or digitized using the Bio-Rad Gel Doc 1000 UV Gel Documentation System. For hot multiplex-PCR, dried polyacrylamide gels were exposed to phosphorimager storage screens, and signals were analyzed via a raster scanning laser on a GS-250 molecular imager (Bio-Rad). To quantitate Western analysis signals Coomassie Blue-stained gels were placed on a UV/ white light conversion screen (Bio-Rad 1707538) and photometrically digitized using the GS-250 molecular imager. Relative intensities of the IL-1R1 and IL-1R2 RNA signals were quantitated against the ␤-actin RNA signal in the same sample using phosphorimager analysis and the data acquisition packages provided with each instrument. All p values were derived from protected t tests or least square means from a two-way factorial analysis of variance (ANOVA).
IL-1R Promoter DNA and IL-1R2 RNA Sequence Analysis-PCR primers were designed using Hitachi Oligo DNA sequence analysis software (Version 5.0). Human IL-1R1 and IL-1R2 gene promoter 5Ј untranslated region and IL-1R2 mRNA sequence data were obtained through GenBank™ accession numbers L09701, U14177, U14178 and X59770, respectively. RNA sequence analysis was performed using both Kodak/IBI/Pustell (a modified Version 2.04) and Omiga 1.1.3 (Oxford Molecular) RNA subsequence analysis packages.

RESULTS
Utilizing human-specific IL-1R1 and IL-1R2 cold or hot primer sets (Table I) and purified IL-1R1 and IL-1R2 cDNA templates, it was determined that at 31 cycles of PCR amplification, the masses of the IL-1R1 and IL-1R2 input cDNAs were linear functions of both the IL-1R1 and the IL-1R2 primary PCR products (Fig. 2). Similarly, quantitative ␤-actin PCR analysis using human-specific ␤-actin primers at a 60°C

FIG. 4. Time course of induction of human IL-1R2 genes with DEX and BUDeR as monitored by Northern dot blot analysis (A) and by quantitative RT-PCR using end radiolabeled primers (B).
Using either technique, IL-1R2 was found to be maximally induced at 3 h. In these experiments, BUDeR was a consistently stronger inducer of IL-1R2 gene transcription than DEX (C). Actinomycin D (ACTD) strongly suppressed DEX-or BUDeR-mediated induction of the IL-1R2 transcript. Cycloheximide (CHX) consistently superinduced IL-1R2 but not IL-1R1 RNA message. Exposure time for IL-1R1 and IL-1R2 or ␤-actin dot blots was 26 and 8 h, respectively; exposure time for IL-1R1, IL-1R2, and ␤-actin RT-PCR signals was 1 h. Significance over IL-1R1 signal at zero time, dashed line; *, p Ͻ 0.04; **, p Ͻ 0.01; ***, p Ͻ 0.001 (ANOVA). annealing temperature has been described previously (23). To ascertain whether the forward and reverse primer sets specific for human IL-1R1 and IL-1R2 were compatible in the same PCR reaction conditions at a common 60°C annealing temperature, pilot experiments were performed using mixed IL-1R1 and IL-1R2 primer sets (Table I) after 3 h of either DEX or BUDeR treatment of HEK cells. Results of a cold multiplex PCR experiment using only these two primer sets are shown in Fig. 3A. The gel reveals that only two bands in the GC-treated HEKs were generated at 340 and 574 bp, corresponding to the expected size of the IL-1R1 and IL-1R2 primary PCR products from published DNA sequences (GenBank™ accession numbers L09701 and X59770, respectively). Control HEK samples (no DEX, BUDeR, or BUDr) yielded a strong IL-1R1 (304-bp) signal but only a weak 574-bp band, corresponding to basal levels of the IL-1R2 RNA message under control (uninduced) conditions. In agreement with previous observations (24), this finding demonstrates that human-specific IL-1R1 and IL-1R2 RNA messages could be both reliably and simultaneously quantitated in single 0.5-ml PCR reaction tubes. Notably, simultaneous 3-h incubation of combinations of DEX ϩ BUDr or DEX ϩ BUDeR in serum-deprived HEKs showed no additive function in either IL-1R1 or IL-1R2 RNA message induction (Fig. 3A).
Hot multiplex-PCR at 31 cycles may provide a more precise quantitation of the levels of specific RNA messages within complex populations of cDNA (as derived from the reverse transcription of total RNA) because: (a) the fluorescence of ethidium bromide-stained cold gels is nonlinear with respect to varying DNA concentrations, making RT-PCR products difficult to quantify, and (b) only 5Ј-end radiolabeled primers labeled to high specific activity are incorporated into the primary PCR product in direct proportion to the amount of existing template in the original cDNA sample. Omission of the reverse transcriptase mix yielded no PCR signals; similarly, RNasefree DNase treatment of the HEK RNA extracts yielded only the expected primary PCR bands at 340 and 574 bp. The induction of the IL-1R1, IL-1R2, and ␤-actin RNA messages with BUDeR, DEX, and BUDr in HEK cells using this technique is shown in Fig. 3B. Data on the abundance IL-1R1 and IL-1R2 RNA messages were normalized against the RNA signal in the same sample for ␤-actin, a moderate-to-highly abundant DNA transcript (23,25). IL-1R1 and IL-1R2 RNA message levels in control ranged from less than ϳ1% to ϳ4% of signal for ␤-actin RNA (Fig. 4). The strongest GC induction at 3 h of the IL-1R2 RNA over the IL-1R2 RNA control signal was by BUDeR (9.2-fold; p Ͻ 0.02), followed by DEX (6.6-fold, p Ͼ 0.04) and BUDr (5.2-fold, p Ͼ 0.06). For each of the GCs studied, the IL-1R1 RNA signal level was slightly depressed at the 1 and 3 h time points when compared with control levels at zero time ( Figs. 3 and 4). Using either Northern dot blot analysis (Fig.  4A) or the hot quantitative RT-PCR at 31 cycles for both IL-1R1 and IL-1R2 (Fig. 4B), time points of 0, 1, 3, 6, 12, and 24 h established a time course over which DEX and BUDeR induced transcription from the IL-1R2 gene in HEK cells. De novo IL-1R2 gene transcription was almost completely abolished by pretreating HEKs with the DNA transcriptional inhibitor actinomycin D (Fig. 4, ACTD) at 1 g/ml in KGM. Cycloheximide (Fig. 4, CHX; 10 g/ml) superinduced IL-1R2 but not IL-1R1 RNA message abundance (Fig. 4, A-C).
PCR primers that lie in exons 4 and 8 of the IL-1R2 gene and from the ␤-actin coding region (Table I) were then used for the direct amplification of 0, 1, 3, and 6 h BUDeR-induced HEK RNA using a single-tube integrated RT-PCR. Samples were then subjected to agarose gel electrophoresis, Southern blotting, and probing with an end-labeled ␤-actin oligonucleotide probe to demonstrate equivalent RNA message levels and RT-PCR efficiencies, an end-labeled oligonucleotide from exon 7 to measure processed IL-1R2 RNA levels, and end-labeled oligonucleotides from introns 7 and 8, (separating exons 7 and 8 and exons 8 and 9, respectively) of the IL-1R2 gene to measure the abundance of unspliced IL-1R2 transcripts. To demonstrate the quantitative nature of this assay for IL-1R2-amplified sequences, mixtures of total RNA from control and GC-treated HEKs were subjected to the same RT-PCR and Southern blotting conditions and IL-1R2 exon 7 hybridization protocols and

FIG. 5. BUDeR induces unprocessed IL-1R2 DNA transcripts and processed IL-1R2 RNA message in HEK cells. HEK cells were induced with
BUDeR for 0, 1, 3, and 6 h; total RNA was then isolated and subjected to RT-PCR using primers for both IL-1R2 and ␤-actin (Table I). Primary PCR products were then separated by electrophoresis, Southern blotted, and hybridized to labeled oligonucleotides derived from the ␤-actin coding region, intron 7 of the IL-1R2 gene, or exon 7 of the IL-1R2 gene. Membranes were phosphorimaged, and exposure times were adjusted for comparison of each probe. Lanes A-E contain total RT-PCR products from HEK cells treated for 3 h with BUDeR and contain 100%, 75%, 50%, 25%, and 0% of the total HEK RNA. These samples were subjected to electrophoresis, Southern transferred, and hybridized with a labeled IL-1R2 exon 7 probe. The linearity of this assay using the standard curve generated by samples A-E is shown in the bottom panel. were analyzed along with the other samples. Quantitation of IL-1R2 signal demonstrates the linearity of this assay over the range of IL-1R2 induction observed (Fig. 5, right panel). No unspliced IL-1R2 transcript and little processed IL-1R2 RNA were detectable at 0 h; however both unprocessed (intron 7-probed) and processed (exon 7-probed) IL-1R2 transcripts are detectable in HEK RNA at 1 and 3 h after GC induction. Similar results were obtained using an IL-1R2 intron 8 probe (data not shown). This suggests that GCs are inducing IL-1R2 RNA message, at least in part, by increasing productive transcription from the IL-1R2 gene.
The parallel induction of the ϳ68-kDa IL-1R2 protein by DEX and BUDeR is shown in Fig. 6. De novo IL-1R2 protein synthesis was almost completely blocked with cycloheximide when HEK cells were pretreated at 10 g/ml in KGM. The rapid appearance of both a ϳ68-kDa (cell-associated) and a ϳ46-kDa (soluble) IL-1R2 protein in HEK WCXTs, suggesting rapid IL-1R2 extracellular cleavage, was closely associated with increasing IL-1R2 RNA message output at 3 h. This suggested a tightly regulated transcription-to-translation mechanism operating in the GC-triggered IL-1R2 signaling pathway. Induction coupling of the IL-1R2 RNA message to translation into IL-1R2 protein is shown at the 3 h time point in Fig. 7. Newly synthesized ϳ46-kDa IL-1R2 protein was rapidly shed into the HEK pericellular medium and was abundantly detected in both HEK WCXTs (Fig. 8A) and in 200-fold concentrated HEK KGM as early as 1 h after GC induction (Fig. 8B).
Lastly, the 1286-nucleotide IL-1R2 RNA message contains several structural features resembling IEG transcription products, including enrichment in adenine-uridine-rich elements, RNA instability elements associated with rapidly degraded RNA messages (Refs. 26 and 27; Fig. 9). Primary cultures of HEK cells at P2-P4 may provide one preferred test system for evaluating GC-mediated IL-1, IL-1R1, and IL-1R2 gene signaling pathways because these epidermal keratinocytes were found to be at least 10-fold more responsive to both DEX and BUDeR stimulation than HeLa ATCC CCL-2, a transformed, IL-1-responsive human epithelioid cell line. No GC-mediated IL-1R2 genetic response was noted in WI38 (ATCC CCL-75) cells, a diploid human lung fibroblast cell line, when treated under identical assay conditions (Fig. 10).

DEX, BUDeR, GC Receptor Binding, and IL-1R
Gene Expression-Despite a nonsymmetrical 16␣,17␣-acetyl substitution and a lack of a 9␣-fluoro atom in its steroid nucleus, BUDeR has been shown to have an 11-fold higher binding affinity for the GC receptor when compared with DEX (28). BUDeR also shows greater potency than DEX in modulating NF-kB-DNA binding (14) and in stimulating transcription from synthetic target genes regulated by GC-responsive elements in their promoters (29). Stereospecificity of the BUDeR 16␣,17␣-acetyl substitution at the GC receptor may also be critical in GCreceptor complex activation because BUDeR was found to induce IL-1R2 gene transcription 1.8-fold greater (p Ͻ 0.02) than an equivalent concentration of BUDr, a 1:1 racemic mixture of the budesonide R and S epimers (Fig. 1). Notably, no synergism was noted when combinations of the GCs (DEX ϩ BUDr or DEX ϩ BUDeR) were used in treating HEK cells (Fig. 3), suggesting the recruitment of single DEX-, BUDr-, or BUDeRmediated signaling pathways during IL-1R2 gene induction via productive GC-GC receptor interaction (14, 30 -36). BUDeR was found to be the most efficient in rapidly up-regulating IL-1R2 RNA message and IL-1R2 protein abundance; a small effect was also noted on the depression of IL-1R1 RNA and protein levels (Figs. 4 and 7). The high ratio of topical to systemic activity of BUDeR when compared with DEX (28,29), indicates that BUDeR may be pharmacologically preferred when inhibiting IL-1 biology and IL-1R expression in the treatment of inflammatory conditions associated with epidermal keratinocytes and related cell types.
GCs and the Induction of IL-1R2 Gene Transcription-GC compounds can inhibit the activation of genes coding for cytokine signaling such as IL-1 and cell surface receptors required in the inflammatory response by limiting the availability of transcription factors to access their target promoters (31-33). A dramatic inhibition of cis-acting transcription factor AP1-like-, ␥-interferon activation sequence-, and especially NF-B-DNA binding by the GCs DEX and BUDeR, temporally correlating to a reduction in cyclooxygenase-2 message synthesis, has recently been demonstrated in HEK cell lines (14). Specifically, AP1 and NF-B are required for the activation of many cytokine and cytokine receptor genes (1,8,31). GCs, via the GCreceptor complex, can directly interact with transcription factor AP1 through AP1's N-terminal domain to moderate gene activation (32,33); similarly, GCs like DEX can inhibit NF-B activation by induction of the IB-␣ gene and IB-␣ inhibitory protein, which ultimately sequesters the NF-B regulatory element as an inactive cytoplasmic complex (34,35). However, GCs appear not to be just broad spectrum pro-inflammatorygene repressors, but rather utilize pleiotropic strategies to potentiate anti-inflammatory responses. These processes include the transcription-mediated up-regulation of several cell surface and soluble ligand receptors, such as the ␤2-adrenergic receptor (36) and the induction of the IL-1R2 in human B and T lymphocytes (4), mononuclear phagocytes (6, 37), polymorphonuclear cells, IL-1R2-transfected fibroblasts (38) and in epidermal cell lines (16 -18).
Our results suggest that in HEKs at P2-P4, BUDr, DEX, and especially BUDeR trigger an up-regulation of basal IL-1R2 gene expression, because low levels of both processed IL-1R2 RNA and IL-1R2 protein were detected in control HEK total RNA and WCXTs at zero time (Ref. 39;. No significant induction of IL-1R1 RNA message was noted. The fact that the human IL-1R1 and IL-1R2 genes are encoded by multiple, tandemly linked DNA elements located in an IL-1R coding-rich region on chromosome 2q12-13 is noteworthy; however, each IL-1R promoter appears to be under individual transcriptional control (4,40). The IL-1R1 and IL-1R2 gene promoters are encoded by three or two different 5Ј exons, respectively, at this locus (40,41), therefore depending on the selection of variably spliced IL-1R gene 5Ј regulatory regions, AP1-and/or NF-kB-DNA binding may be alternately utilized to promote transcription from a specific IL-1R gene isotype. Notably, each IL-1R2 promoter exon but not every IL-1R1 promoter exon contains a GC-responsive element consensus sequence (40). These latter features may allow greater flexibility in the IL-1R response to different concentrations or combinations of extracellular signaling ligands.
The GC-induced IL-1R2 Gene Behaves Like an IEG-Newly generated IL-1R2 RNA message contains features characteristic of the transcription products of IEGs (21). These include expression in diverse cell types, a signal transduction function, relatively low basal abundance, rapid transient induction after treatment with GCs, cytokines, and mitogens such as phorbol 12-myristate 13-acetate (4,24,38), cycloheximide superinduction, actinomycin D repression, and rapid decay of the IL-1R2 RNA message. Moreover, RNA sequence analysis of the 1286nucleotide IL-1R2 RNA message (GenBank™ accession number X59770) reveals 17 adenine-uridine-rich RNA instability elements typical of cytokine, lymphokine, and proto-oncogene RNA messages that are transiently expressed and rapidly degraded (21, 26, 27; Fig. 9). Northern analysis, RT-PCR assay, unprocessed DNA transcript assay, the rapid disappearance of IL-1R2 RNA signal after a 3-h GC stimulation of HEKs with either DEX or BUDeR, and actinomycin D suppression of this GC-induction suggest that the de novo GC-mediated IL-1R2 expression pathway is regulated primarily at the level of IL-1R2 gene transcription, although post-transcriptional IL-1R2 RNA message stabilization may provide auxiliary controls in other cell types (11,16). Parallel time course kinetic analysis of newly synthesized RNA levels with Western immunoblotting also revealed a tight coupling of de novo IL-IR2 gene transcription with translation of the IL-1R2 RNA message; a newly synthesized (ϳ46-kDa) IL-1R2 protein was detected in the HEK pericellular environment as early as 1 h after BUDeR induction.
In conclusion, the established function of IL-1R2 protein is to act as a molecular trap to capture extracellular IL-1 and thereby compromise the IL-1 signaling system (6,11,42,43) as a strong negative extracellular regulator of IL-1 action. Our evidence of the rapid coupling of IL-1R2 gene transcription to IL-1R2 RNA translation into protein, followed by the shedding of the ϳ46-kDa IL-1R2 into the extracellular space, suggests that GC-triggered HEKs can deal rapidly with local IL-1 levels via a classical IEG response. When compared with DEX, nonfluorinated GCs bearing asymmetric 16␣,17␣-acetyl substitutions such as BUDeR may elicit a stronger induction of the IL-1R2 cytokine scavenger system. Such initiator elements of the IL-1R2 signal transduction pathway may therefore present future targets for pharmacologic design in light of altered IL-1R2 gene expression in pathological conditions of the gastrointestinal tract (44), in focal cerebral ischemia (3,45), in neuroinflammation (46), and in neurodegenerative disorders of the brain (47,48).