Introduction

Top Abstract Introduction Methods Results Discussion References

Rheumatoid arthritis is a complex autoimmune disease characterized by chronic inflammation, bone erosion and proliferation of the synovial lining. Inflammatory cytokines such as IL-1 are elevated in the joint fluid of patients with rheumatoid arthritis and, as such, are thought to play a critical role in the progression of the disease (Goddard et al., 1992; Sipe et al. 1994). Exposure of RSF to IL-1 induces the expression of several inflammatory genes and results in the production of a wide variety of pro-inflammatory mediators including IL-8 and PGE₂ (Dayer et al., 1986; Gilman et al., 1988; Roshak et al., 1996a, b). In the case of PGE₂, IL-1 causes the coordinate induction of RSF 85-kDa PLA₂ and mitogen-inducible COX-II mRNA and subsequent increases in protein levels. This results in nanogram quantities of PGE₂ produced by these cells (Angel et al., 1994; Hulkower et al., 1994; Roshak et al., 1996a).

IL-1 is known to act through the activation of the pro-inflammatory transcription factor, NFB (Siebenlist et al., 1994; Thanos and Maniatis, 1995). The NFB family of transcription factors comprises several distinct gene products including the mammalian forms, p65, p50, c-rel and Rel-B (Baldwin, 1996; Siebenlist et al., 1994). These proteins form a variety of homo- and heterodimer pairs, display different affinities for distinct DNA binding motifs and are expressed in varying levels in different tissues. Typically, NFB dimers are confined to the cytoplasm of nonstimulated cells through sequestration of the nuclear localization sequence by its endogenous inhibitor, IB (Miyamoto and Verma, 1995). Upon cellular activation through a variety of stimuli (i.e., cytokines, viral or bacterial products, free radicals or physical stress), IB is phosphorylated and degraded via the ubiquitin proteasome pathway (Li et al., 1995; Palombella et al., 1994; Traenckner et al., 1994). Liberated NFB dimers are then free to translocate to the nucleus, bind to specific B motifs in target gene enhancers and induce the transcription of several pro-inflammatory genes exemplified by the cytokines IL-6 (Liberman and Baltimore, 1990) and IL-8 (Mukaida et al., 1989; Kunsch and Rosen, 1993) and adhesion molecules such as intracellular and vascular cell adhesion molecules (Ledebur and Parks, 1995; Muller et al., 1995; Shu et al., 1993). We recently demonstrated a critical role of the NFB protein, p65, in the IL-1-regulated expression of 85-kDa PLA₂ and COX-II through the use of oligonucleotide decoys and specific antisense (Roshak et al., 1996b).

NFB activation has been described in a variety of inflammatory disease models including airway inflammation (Adcock et al., 1994; Blackwell et al., 1994) and atherosclerosis (Liao et al., 1994) and is thought to significantly contribute to the progression of the disease through the enhanced expression of target inflammatory genes. In rheumatoid synovium, immunohistochemistry identified NFB proteins, p65 and p50, constitutively present in the nuclei of synovial lining cells (Handel et al., 1995). Further, exposure of cultured synovial cells to TNF caused increased nuclear translocation of NFB proteins which led to expression of NFB-dependent genes, IL-6 and intracellular adhesion molecule and proliferation (Fujisawa et al., 1996). This was inhibitable by treatment with the antioxidant, N-acetyl-L-cysteine. Further, glucocorticosteroids and salicylates have recently been shown to suppress NFB activity through transcriptional up-regulation of IB (Auphan et al., 1995; Scheinman et al., 1995) and prevention of its degradation (Kopp and Ghosh, 1994), respectively. Taken together, targeting NFB activation therefore provides an attractive approach for developing novel anti-inflammatory agents.

Several agents have been shown to possess NFB modulatory activity. Antioxidants, such as n-acetyl-cysteine and PDCT, have been reported to repress activation of NFB through the inhibition of IB phosphorylation (Kawai et al., 1995). Although these agents have provided researchers with tools to assess NFB activity, they require use at high concentrations, which in some cases are toxic, and they exhibit other activities, which often makes interpretation of results difficult. Better tools would clearly be beneficial in the study of NFB regulation. Recently, the marine natural product, hymenialdisine (from the sponges Axinella verrucosa and Acanthella aurantiaca; Cimino et al., 1982) has been characterized as an inhibitor of NFB activation. Hymenialdisine concentration-dependently inhibited both receptor-mediated (TNF and lipopolysaccharide) and soluble stimuli-mediated (PMA) luciferase expression in an NFB-driven luciferase reporter assay constructed in U937 cells (Breton and Chabot-Fletcher, 1997). Gel-shift analysis of cellular nuclear extracts verified specific reduction in NFB nuclear binding but not in the binding of the transcription factors C/EBP, AP-1 or SP1 by hymenialdisine. Further, hymenialdisine inhibited IL-8 mRNA and protein formation in the TNF-treated U937 cell. These data support a specific inactivation of NFB by hymenialdisine and provide a novel reagent to study NFB participation in inflammatory models.

Herein, we characterize hymenialdisine, for the first time, in a relevant in vitro model of disease, i.e., IL-1-stimulated RSF. We show that hymenialdisine directly inhibits IL-1-mediated NFB activation and nuclear translocation. This leads to reductions in COX-II and 85-kDa PLA₂ mRNA and protein levels and ultimate attenuation of PGE₂ production. Further, we also demonstrate inhibition of IL-1-stimulated IL-8 production, an NFB-mediated process, by hymenialdisine. Alternatively, IL-1-mediated VEGF up-regulation is not affected, which supports the lack of effect on general transcription machinery. Together, these data support the hypothesis that modulation of NFB activation provides a novel approach to therapeutically modifying inflammatory mediators.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials and chemicals. PDCT was purchased from Sigma Chemical Co.(St. Louis, MO). Hymenialdisine (SK&F 108752) and aldisine (SB 203063) were obtained from Suntory Ltd., Japan. RO 32-0432 was synthesized by the Department of Medicinal Chemistry, SmithKline Beecham Pharmaceuticals according to the reported synthesis (Bit et al., 1993)

Human synovial fibroblast culture. Primary cultures of human RSF were obtained by enzymatic digestion of synovium obtained from 10 adult patients with rheumatoid arthritis as described previously. (Roshak et al., 1996a). Cells were cultured in Earle's Minimal Essential Medium which contained 10% fetal bovine serum, 100 U/ml penicillin and 100 µg/ml streptomycin (GIBCO, Grand Island, NY), at 37°C and 5% CO₂. Cultures were used at passages 4 through 9 to obtain a more uniform type I fibroblast population. For some studies, fibroblasts were plated at 5 × 10⁴ cells/ml in 16-mm (diameter) 24-well plates (Costar, Cambridge, MA). Cells were exposed to an optimal dose of IL-1 (1 ng/ml; Roshak et al. 1996a) (Genzyme, Cambridge, MA) for the designated time. Drugs in DMSO vehicle (1%) were added to the cell cultures 15 min before the addition of IL-1. Each study represents one of two to five individual experiments with RSF from different donors unless otherwise noted.

ELISA measurement of PGE₂, IL-8 and VEGF. PGE₂ or IL-8 levels in cell-free medium collected at the termination of the culture period were measured directly by enzyme immunoassay kits purchased from Cayman Chemical Co. (Ann Arbor, MI) and Biosource International (Camarillo, CA) respectively, as described previously (Roshak et al., 1996b). Vascular endothelial cell growth factor levels in cell-free medium were measured with a VEGF ELISA Kit (R&D Systems, Minneapolis, MN) according to the manufacturer's protocol as described previously (Jackson et al., 1997). Sample or standard dilutions were made with experimental medium and results were expressed as nanograms per milliliter of medium as mean ± S.D. of triplicate determinations unless otherwise stated and are subjected to one-way analysis of variance and Duncan's multiple range test (P < .05) for statistical evaluation where indicated.

RSF subcellular fractionation. Human RSF were removed by trypsin/EDTA, resuspended to 1.0 × 10⁸ cells/ml in cold homogenization buffer (0.34 M sucrose, 10 mM HEPES, pH 7.4, 1 mM ethyleneglycol-bis(-aminoethyl ether)-N,N,N,N-tetraacetic acid, 1 mM phenylmethylsulfonyl fluoride, 200 µM leupeptin, 20 µg/ml soybean trypsin inhibitor and 20 µg/ml aprotinin at 4°C) and disrupted on ice by sonication with a Bransonic probe tip as described previously (Roshak et al., 1996a). The homogenate was centrifuged at 100,000 × g for 60 min at 4°C to obtain a supernatant (cytosol) and particulate fraction. The particulate fraction was resuspended in 5 volumes of homogenization buffer. Protein concentration was measured by Bradford analysis (Bio-Rad, Richmond, CA). Both fractions were flash frozen with liquid N₂ and stored at 80°C for analysis.

Immunoblot analysis. Cell fractions (25-50 µg protein) and/or recombinant protein standard were analyzed by SDS-polyacrylamide gel electrophoresis (10% gels; Bio-Rad) as described previously (Roshak et al., 1996a, b) and visualized by use of the ECL Western blotting system (Amersham, Arlington Heights, IL). Rabbit polyclonal antiserum against the rh 85-kDa PLA₂ was prepared as described previously (Roshak et al., 1996a, b). Rabbit anti-human COX-II was kindly donated by D. Dewitt (Michigan State University, East Lansing, MI) and used as described previously (Roshak et al., 1996a, b). Positive control standards included a 24-hr lipopolysaccharide-stimulated monocyte particulate fraction (25 µg) containing COX-II protein and rh 85-kDa PLA₂ (Roshak et al., 1996a, b). Rabbit polyclonal antibodies to p65 were obtained from Santa Cruz Biotechnology Inc. (Santa Cruz, CA) and used according to the manufacturer's instructions. Gels were scanned for density with UVP Imagestore 5000 (San Gabriel, CA). Measurement of pixels in the bands was expressed as an area value.

Preparation of nuclear extracts and electrophoretic mobility shift assay. Confluent RSF in T75 flasks were stimulated for 15 min in the presence of IL-1 (1 ng/ml; 37°C), washed twice with phosphate-buffered saline, then removed by trypsinization. Nuclear extracts were prepared according to published methods (Dignam et al., 1983; Osborne et al., 1989) with some modifications. In short, cells were pelleted by centrifugation and resuspended in buffer A, 20 µl/10⁷ cells (10 mM HEPES, pH 7.9, 10 mM KCl, 1.5 mM MgCl₂, 0.5 mM DTT, 0.1% (w/v) Nonidet P-40). The cell suspension was incubated on ice for 10 min, and the nuclei were pelleted by microcentrifugation at 3500 rpm for 10 min at 4°C. The pellet was suspended in 15 µl of buffer C (20 mM HEPES, pH 7.9, 0.42 M NaCl, 1.5 mM MgCl₂, 25% (v/v) glycerol, 0.2 mM EDTA, 0.5 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride) and gently mixed for 20 min at 4°C. The sample was microcentrifuged at 14,000 rpm for 10 min at 4°C, and the resultant supernatant (nuclear extract) was diluted to 75 µl with buffer D (20 mM HEPES, pH 7.9, 50 mM KCl, 20% glycerol, 0.2 mM EDTA, 0.5 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride). Samples were stored at 80°C until analysis.

DNA binding reactions and electrophoretic mobility shift assay (EMSA). A double-stranded oligonucleotide containing the sequence corresponding to the classical NFB consensus site (5 agt tga ggg gac ttt ccc agc c 3) (Santa Cruz Biotechnology Inc.) was end-labeled with -³²P-ATP with T4 kinase (Life Technologies, Gaithersburg, MD). Unincorporated nucleotides were removed by column chromatography over two Sephadex G-50 columns (Pharmacia, Piscataway, NJ). Binding reactions were carried out in a final volume of 25 µl consisting of 10 mM HEPES, pH 7.9, 4 mM Tris-HCl, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 10% glycerol, 1.5 mg/ml bovine serum albumin and 2 µg of poly(dI-dC). Each reaction, containing 10 µg of nuclear extract, and 0.5 ng of ³²P-labeled oligonucleotide probe (~50,000 cpm) was incubated for 20 min at room temperature. Binding reactions were subjected to nondenaturing polyacrylamide electrophoresis through 4% gels in a 1× Tris-borate-EDTA buffer system. Gels were dried and subjected to autoradiography. We previously demonstrated specific NFB binding to IL-1-treated RSF nuclear extracts which was competed by unlabeled NFB motif but not by an unrelated oligonucleotide motif (OCT-1) (Roshak et al., 1996b)

Northern analysis. Total RNA was isolated from RSF with Trizol reagent (Gibco/BRL, Bethesda MD) according to the manufacturer's protocol and quantitated by spectrophotometry. RNA (20 µg) were subjected to electrophoresis in 1% agarose gel containing formaldehyde. RNA molecular weight markers (Gibco/BRL) were also included flanking the samples. After electrophoresis, gels were rinsed twice by shaking for 15 min in 300 ml of distilled water followed by a 7-min incubation in 200 ml 50 mM NaOH. Gels were next incubated in 300 ml 10× SSC (1× = 0.15 M sodium chloride, 0.015 M sodium citrate, pH 7.0) for 20 min. RNA was transferred to Hybond N+ (Amersham, Amersham, UK) by vacuum blotting (Bio-Rad, Hercules, CA) in 10× SSC according to the manufacturers' protocol. After transfer, RNA was fixed to the membrane by UV cross-linking (0.12 J/cm²). RNA samples and markers were visualized on the membrane by staining with 0.02% methylene blue in 0.3 M sodium acetate, pH 5.5, for 5 min followed by destaining in distilled water for 15 min. Hybridizations were carried out in bottles in a Hybaid oven (Hybaid Ltd, Middlesex, UK). Filters were prehybridized (20 ml/blot) in 6× SSC, 5× Denhardt's [50× = 10 mg/ml Ficoll (400), 10 mg/ml polyvinylpyrrolidone, 10 mg/ml bovine serum albumin], 0.5% SDS, 0.1 mg/ml denatured salmon sperm DNA for 3 hr or more at 68°C. Hybridizations were done in prehybridization solution (10 ml/blot) containing 30 ng of denatured specific DNA probe labeled to 1 to 2 × 10⁹ dpm/µg with ³²P (see below) at 68°C for 18 hr. After hybridization, blots were washed twice with 100 ml 2× SSC, 0.1% SDS for 15 min at 68°C in bottles, and once with 100 ml 1× SSC, 0.1% SDS for 30 min at 68°C in bottles. Blots were then removed from bottles and washed once with 200 ml 0.2× SSC, 0.1% SDS for 15 min at 68°C in a tray. Filters were analyzed on a phosphorimager (Molecular Dynamics, Sunnyvale, CA).

DNA probes. 85-kDa PLA₂ probe was a 2.5-kb HindIII to SstI cDNA fragment, COX-II probe was a 1.2-kb EcoRI fragment of the murine PGHS-2 cDNA clone kindly provided by Dr. David DeWitt, Michigan State University. Labeled probes were prepared from 20 ng of cDNA by random priming with a Rediprime kit and 50 µCi ³²P-dCTP (Amersham, Arlington Heights, IL). Unincorporated nucleotides were removed by gel filtration with Quick Spin columns (Boehringer Mannheim, Indianapolis, IN).

	Results

Top Abstract Introduction Methods Results Discussion References

The effect of inhibitors on IL-1-induced RSF PGE₂ formation. Antioxidants such as PDCT have been effective but nonselective inhibitors of NFB activation, presumably functioning through the inhibition of IB degradation. Because we have previously demonstrated NFB participation in IL-1-induced RSF PGE₂ production, PDCT was evaluated for its ability to inhibit PGE₂ formation in the RSF system. Confluent RSF in Earle's Minimal Essential Medium with 10% fetal bovine serum were incubated with DMSO vehicle or PDCT (1-300 µM) for 15 min at 25°C before addition of IL-1 (24 hr, 37°C). Cells were monitored and toxicity was observed at concentrations greater than 30 µM, as assessed by morphology and trypan blue exclusion. Figure 1 shows that pretreatment with PDCT resulted in a concentration-dependent decrease in IL-1-stimulated PGE₂ production (IC50 ~ 15 µM). Hymenialdisine (0.03-10.0 µM), an analog, aldisine (10 µM), or DMSO vehicle alone were evaluated for their effect on IL-1-stimulated PGE₂ production. Cells were incubated with the respective reagents 15 min before exposure to IL-1 for 24 hr. Pretreatment of RSF with hymenialdisine, but not aldisine, resulted in a concentration-dependent inhibition of IL-1-stimulated PGE₂ release (fig. 2; IC₅₀ = 0.6 µM ± 0.2; confidence limits of 0.002-1.116). No toxicity was noted. Breton and Chabot-Fletcher (1997) mentioned that an analog of hymenialdisine, debromohymenialdisine, is an inhibitor of PKC (DiMartino et al., 1995). However, in their system this was not shown to be the primary mechanism of action for hymenialdisine because a selective, nonisotype specific, PKC inhibitor, RO 32-0432 (IC₅₀ vs. human neutrophil PKC,14 nM) did not have an effect in the TNF-stimulated NFB-luciferase reporter assays and did not affect TNF-stimulated U937 IL-8 production by the same cells. As a control, RO 32-0432 was also examined in the IL-stimulated RSF system. Figure 2 shows that RO 32-0432 (100 nM) had no effect on IL-1-stimulated prostanoid synthesis, which demonstrates the lack of participation of PKC.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   The antioxidant, PDCT, concentration-dependently inhibits IL-1-stimulated PGE2 formation. RSF were preincubated with increasing concentrations of PDCT (1-300 µM) for 15 min at 25°C before the addition of IL-1 for 24 hr (1 ng/ml, 37°C). Cell-free medium was removed and analyzed for PGE2 levels by ELISA as described under "Methods." Data represent the mean ± S.D., n = 3 of one experiment.

View larger version (46K): [in this window] [in a new window]   Fig. 2.   Hymenialdisine inhibits IL-1-induced RSF PGE2 production in a concentration-dependent manner. RSF were incubated with various concentrations of hymenialdisine (0.03-10.0 µM), aldisine (10 µM) or RO 32-0432 (100 nM) for 15 min at 25°C before stimulation with IL-1 for an additional 24 hr (1 ng/ml, 37°C). Cell-free medium was removed and analyzed for PGE2 levels by ELISA as described under "Methods." Data are expressed as % stimulated control (nonstimulated PGE2 control, 0.7 ng/ml ± 0.06; stimulated PGE2 level, 17.4 ng/ml ± 2.7) and represent the mean ± S.D., n = 3 of one of four experiments. * indicates significant difference form IL-1 control at P < .05.

Effect of hymenialdisine on IL-1-mediated activation of NFB in RSF. We have previously demonstrated that 15 min is the peak time for acute activation of NFB dimers by IL-1 in this system (Roshak et al., 1996b). RSF from three donors were pretreated with DMSO vehicle, hymenialdisine at its approximate IC₅₀, 1 µM, or aldisine (1 µM), for 15 min at 25°C before exposure to IL-1 (1 ng/ml, 15 min at 37°C). Nuclear extracts were prepared as described under "Methods." Figure 3A shows the EMSA of one representative experiment and shows that NFB is constitutively present in unstimulated RSF nuclear extracts as previously reported (Roshak et al., 1996b). Exposure to IL-1 results in a 2- to 3-fold increase in NFB binding activity in the nuclear fraction as confirmed by scanning gel densitometry (area pixel values unstimulated control,1980; IL-1-stimulated control, 4804; fig 3A). We have previously demonstrated that binding to the NFB classical motif by stimulated RSF nuclear proteins is specific (Roshak et al., 1996b). Similarly, in this study, binding to the NFB classical motif was inhibited by incubation with excess unlabeled NFB consensus oligonucleotide (40×). Pretreatment of RSF with hymenialdisine or aldisine resulted in an ~80% (area pixel value 2500) or 35% (area pixel value 3820) reduction in the IL-1-stimulated NFB binding, respectively. Percent reduction is obtained by comparing the IL-1-stimulated value, corrected for constituitive levels (subtraction of 1980), to the corrected hymenialdisine value. Addition of hymenialdisine (1 µM) to binding reactions containing untreated RSF nuclear extracts did not directly interfere with NFB binding to the classical motif (data not shown).

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Hymenialdisine directly effects NFB activation. (A) EMSA analysis of the effect of hymenialdisine on binding of RSF nuclear proteins to the classical NFB motif. RSF were preincubated with DMSO vehicle, aldisine (1 µM) or hymenialdisine (1 µM), for 15 min at 25°C before stimulation with IL-1 (1 ng/ml) for 15 min at 37°C and preparation of nuclear extracts as described under "Methods." Nuclear extracts (10 µg) were incubated with radiolabeled classic NFB oligonucleotide probe (0.5 ng) in the presence or absence of cold NFB consensus as competitor (40×) and subjected to EMSA as described under "Methods." One representative of three individual studies. (B) Western analysis of the effect of hymenialdisines on cytosolic p65 levels in IL-1-stimulated RSF. Western analysis was performed as described under "Methods" with cytosols (30 µg/lane) from RSF treated with DMSO vehicle (lane 1), DMSO and IL-1 (1 ng/ml, 15 min at 37°C) (lane 2) or 1 µM hymenialdisine for 15 min at 25°C (lane 3) before stimulation with IL-1.

Hymenialdisine inhibits IL-1-induced up-regulation of the prostanoid metabolizing enzymes. In the following studies, hymenialdisine was used at a concentration greater than its PGE₂ inhibitory IC₅₀ value (3 µM) to maximally effect the expression levels of the COX-II and 85-kDa PLA₂ genes. The effect of hymenialdisine on the mRNA levels for COX-II and 85-kDa PLA₂ in IL-1-stimulated RSF was analyzed by Northern blotting. Figure 4A shows one representative of two studies where COX-II mRNA was undetectable in unstimulated control cells and highly induced after 8 hr stimulation with IL-1 as reported previously (Roshak et al., 1996a, b). Hymenialdisine reduced this induction to levels evident in unstimulated control RSF. Pretreatment with aldisine had no effect (data not shown). A minor amount of 85-kDa PLA₂ mRNA was detectable in the absence of IL-1 stimulation and a marked induction followed the 8-hr treatment with IL-1 (fig. 4B). This induction was completely blocked by hymenialdisine. PGE₂ levels measured in cell-free media reflect the Northern blot data (unstimulated control PGE₂, 0.6 ng/ml; IL-1-stimulated control PGE₂, 10.7 ng/ml; 3 µM hymenialdisine + IL-1 PGE₂, 0.6 ng/ml).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Treatment of RSF with hymenialdsine inhibits IL-1-induction of COX-II and 85-kDa PLA2 transcripts. RSF were exposed to DMSO vehicle (lanes 1 and 2) or 3 µM hymenialdisine for 15 min at 25°C before an 8-hr exposure to IL-1 (1 ng/ml) for all but the untreated control cultures (lane 1). Total RNA was isolated and subjected to Northern analysis [COX-II (A); 85-kDa PLA2 ( B)] as described under "Methods." One representative of two individual studies.

View larger version (44K): [in this window] [in a new window]   Fig. 5.   Treatment of RSF with hymenialdisine inhibits IL-1-stimulated up-regulation of COX-II and 85 kDa-PLA2 protein levels. Panel A shows the COX-II Western analysis of particulate fractions (40 µg/lane) from RSF exposed to DMSO vehicle or 3 µM hymenialdisine for 15 min at 25°C followed by treatment with IL-1 (1 ng/ml; 8 hr) in all but the unstimulated control cultures. The COX-II immunoreactive protein migrated at ~80 kDa. Panel B shows a representative 85-kDa PLA2 Western analysis of cytosolic fractions (30 µg/lane) from RSF treated in the same manner but incubated with IL-1 for 24 hr. The baculovirus expressed recombinant human 85-kDa PLA2 standard migrated at ~110 kDa. One representative of two experiments.

Hymenialdisine inhibits IL-1-induced RSF IL-8 but not VEGF production. To verify that hymenialdisine was acting specifically on NFB-regulated genes and not as a general inhibitor of transcription in this system, the effect of the compound on other IL-1-inducible genes was evaluated. RSF produce IL-8 and VEGF in response to IL-1 exposure (Roshak et al. 1996b; Jackson et al.,1997). Transcription of IL-8 has been shown to be highly regulated by NFB (Kunsch and Rosen, 1993) whereas VEGF expression is not reported to be NFB dependent (Tischer et al., 1991). Incubation of RSF from one representative donor with IL-1 caused a marked production of IL-8 which was ~60% inhibited by pretreatment with hymenialdisine (10 µM) but not the inactive analog, aldisine (10 µM; fig. 6A). In contrast, although stimulation with IL-1 resulted in a 1.7-fold increase in VEGF levels in the cell-free medium, this was not affected by pretreatment with 10 µM hymenialdisine, a concentration which resulted in total inhibition of PGE₂ production (fig. 6B).

View larger version (48K): [in this window] [in a new window]   Fig. 6.   Hymenaldisine inhibits RSF IL-1-induced IL-8 but not VEGF production. RSF were incubated with various concentrations of hymenialdisine (0.1-10.0 µM) or aldisine (10 µM) for 15 min at 25°C before stimulation with IL-1 for an additional 24 hr (1 ng/ml, 37°C). Cell-free medium was removed and analyzed for IL-8 (A) or VEGF (B) levels by ELISA as described under "Methods." Data are expressed as % stimulated control [(A) nonstimulated IL-8 control, 0.7 ± 0.04 ng/ml; IL-1-stimulated IL-8 level, 100.6 ± 10.1 ng/ml; (B) unstimulated VEGF control, 0.57 ± 0.06 ng/ml; stimulated VEGF level, 1.0 ± 0.05 ng/ml] and represent the mean ± S.D., n = 3 of one experiment. *indicates significant difference from IL-1 control at P < .05.

	Discussion

Top Abstract Introduction Methods Results Discussion References

We showed previously that IL-1 induces the migration of NFB proteins to the nucleus in RSF (Roshak et al., 1996b). Further, we demonstrated NFB involvement in the IL-1-mediated up-regulation of 85-kDa PLA₂ and COX-II resulting in subsequent prostanoid formation. Specific antisense against the NFB protein, p65, but not p50 or c-rel caused a marked reduction in the expression of these genes, reducing PGE₂ production and demonstrating a key role of this transcription factor (Roshak et al., 1996b).

With the recent description of hymenialdisine as a novel inhibitor of NFB activation, this natural product was evaluated for its effect in a physiologically relevant in vitro model of disease. Hymenialdisine produced a potent concentration-dependent inhibition of IL-1-induced RSF PGE₂ production (IC₅₀, 0.6 µM) whereas the less active analog aldisine was without significant effect. This was considerably more potent than the nonspecific antioxidant, PDCT (IC₅₀, 15 µM), which acts partly through modulation of NFB. Electrophoretic mobility shift assays and Western analysis of p65 confirmed a direct effect of hymenialdisine on IL-1-induced RSF NFB activation and translocation to the nucleus. Pretreatment with hymenialdisine before IL-1 exposure clearly resulted in a decrease in nuclear protein binding to the radiolabeled B motif. Complimentary Western analysis of cytosolic p65 levels demonstrated that hymenialdisine inhibited IL-1-stimulated RSF p65 nuclear translocation.

Hymenialdisine inhibition of NFB activation correlated with RSF reduced ability to transcribe 85-kDa PLA₂ or COX-II in response to IL-1. This resulted in a marked depletion of protein levels for both enzymes, which severely compromised prostanoid formation by these cells. These data are consistent with our previous findings with specific p65 antisense that inhibition of NFB activation results in a reduction in IL-1-driven transcriptional up-regulation of the prostanoid-metabolizing enzymes.

Inhibition was not restricted to PGE₂ production because IL-1-stimulated IL-8 formation was also reduced by exposure to hymenialdisine. This was not unexpected because IL-8 is known to be an NFB-regulated gene (Kunsch and Rosen, 1993) and this suggests a possible common mechanism of action of hymenialdisine on NFB activation. Higher concentrations of hymenialdisine were required for IL-8 inhibition than for PGE₂ reduction. This could be because these studies were optimized to observe PGE₂ production and not IL-8. Further, PGE₂ formation requires the up-regulation of two distinct enzymes before the conversion of AA to prostanoid, whereas IL-8 is directly synthesized. In addition, several other transcription factors are known to participate in the regulation of IL-8 expression, e.g., NF-IL6 (Matsusaka et al., 1993), and one cannot rule out that these may function in a compensatory fashion in the absence of NFB.

Hymenialdisine appears to act specifically at the level of NFB inhibition and not as a general inhibitor of transcription. Breton and Chabot-Fletcher (1997) recently reported that hymenialdisne did not inhibit the transcription of the housekeeping gene, G3PDH or of PAI-1, a TNF-stimulated, NFB-independent gene. Similarly in RSF, exposure to hymenialdisine did not affect the IL-1-induced up-regulation of VEGF whose promoter region contains AP-1 binding sites but not NFB sites (Tischer et al., 1991) and appears to be activated thru src-associated kinase pathways (Mukhopadhyay et al., 1995). Finally, the lack of effect of hymenialdisine on IL-1-stimulated VEGF production also demonstrates that hymenialdisine is functioning downstream of the initial IL-1 receptor signaling.

In conclusion, hymenialdisine effectively attenuated the formation of RSF PGE₂ in response to IL-1 acting predominately through the inhibition of NFB activation and 85-kDa PLA₂ and COX-II gene transcription. These data provide strong support that modulation of NFB activation would be therapeutically beneficial in attenuating pro-inflammatory mediator formation.