Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Inhaled glucocorticoids are an established and potent therapy to control asthma; airway epithelium comes in direct contact with inhaled glucocorticoids and is now recognized to be an important target of their action (1). Several studies have demonstrated that glucocorticoids regulate the production of cytokines by bronchial epithelium. Previous studies in our laboratories have shown that glucocorticoids inhibit expression of granulocyte macrophage colony-stimulating factor as well as chemokines including regulated on activation, normal T cells expressed and secreted, and monocyte chemotactic protein-4 in the BEAS-2B bronchial epithelial cell line (4). Glucocorticoids have also been shown to suppress the expression of interleukin (IL)-6 and IL-8 by epithelial cells (8, 9).

The accumulation of leukocytes at sites of inflammation in the airways is a characteristic of asthma, and is believed to be regulated to a large degree by adhesion molecules, including selectins, integrins, and their counterreceptors expressed on the surface of leukocytes, and target cells such as endothelial cells, connective tissue fibroblasts, and bronchial epithelial cells. Recently, we have demonstrated that vascular cell adhesion molecule-1 (VCAM-1) and intercellular adhesion molecule-1 (ICAM-1) expression is induced by cytokines in human bronchial epithelial cells (10). Furthermore, we have demonstrated that VCAM-1 on the BEAS-2B human bronchial epithelial cell line can serve as a substrate for adhesion of eosinophils. In the present report, we show that glucocorticoids are potent inhibitors of VCAM-1 expression, but not ICAM-1, on BEAS-2B cells. The following glucocorticoids were tested: fluticasone propionate (FP), mometasone furoate (MF), beclomethasone dipropionate (BDP), beclomethasone-17-monopropionate (17-BMP), triamcinolone acetonide (TAA), and hydrocortisone (HC). These studies demonstrate that VCAM-1, unlike ICAM-1, is a glucocorticoid-sensitive gene, and suggest that regulation of epithelial VCAM-1 expression may be a previously unrecognized mechanism of glucocorticoid action in the airways.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

The following materials were purchased: Dulbecco's modified Eagle's medium (DMEM), Ham's F12 medium, Ca²⁺- and Mg²⁺-free Hanks' balanced salt solution (HBSS), Versene (Ca²⁺- and Mg²⁺-free HBSS containing 0.02% ethylenediaminetetraacetic acid [EDTA]), trace elements, phosphoethanolamine/ethanolamine, and retinoic acid (Biofluids, Inc., Rockville, MD); insulin, HC, epidermal growth factor, and endothelial cell growth supplement (Collaborative Research, Bedford, MA); actinomycin D, cholera toxin, and triiodothyronine (Sigma Chemical Co., St. Louis, MO); fetal calf serum (FCS), penicillin/streptomycin solution, agarose, and formamide (Life Technologies, Inc., Gaithersburg, MD); fungizone (GIBCO BRL, Gaithersburg, MD); RNAzol B (Tel-Test Inc., Friendswood, TX); chloroform, isopropanol and formaldehyde (Fisher Scientific, Fernwood, NJ); Genescreen membranes and [-³²P] deoxyadenosine triphosphate (Dupont NEN, Boston, MA); and recombinant human tumor necrosis factor- (TNF-) (R&D Systems, Minneapolis, MN). The following glucocorticoids were obtained from the indicated sources: 17-BMP, BDP, and FP (Glaxo-Wellcome, Research Triangle Park, NC); budesonide (BUD) (Astra Draco, Westborough, MA); HC (Sigma Chemical Co.); MF (Schering); and TAA (Rhone-Poulenc Rorer, Collegeville, PA). Stock solutions of glucocorticoids were dissolved in dimethyl sulfoxide (DMSO) at 0.1 M concentration and stored at 20°C.

Cell Culture

BEAS-2B cells. BEAS-2B cells were a generous gift from Dr. Curtis Harris (National Cancer Institute, Bethesda, MD). This cell line was derived from human bronchial epithelium transformed by an Adenovirus 12-SV40 hybrid virus (11). These cells retain electron microscopic features of epithelial cells and show positive staining with antibodies to cytokeratin but do not form tight junctions (11 and data not shown). BEAS-2B cells were cultured in 25-cm² tissue culture flasks and maintained in F12/10× medium consisting of Ham's F12 nutrient medium containing penicillin (100 U/ml) and streptomycin (100 U/ml) and supplemented with insulin (5 µg/ml), HC (10⁷ M), triiodothyronine (3.1 × 10⁹ M), cholera toxin (10 ng/ml), endothelial cell growth supplement (3.75 µg/ml), epidermal growth factor (12.5 ng/ml), phosphoethanolamine/ethanolamine (5 × 10⁶ M), trace elements (1×), and retinoic acid (0.1 µg/ml). Cells were used between passages 35 and 47 and were plated on six-well cluster plates (Costar, Cambridge, MA) and cultured for at least 48 h in F12/DMEM (which lacks added HC containing 5% heat-inactivated FCS, penicillin (100 U/ml), and streptomycin (100 U/ml) before use in experiments.

Primary normal human bronchial epithelial cells (NHBE). Primary bronchial epithelial cells were obtained using a modification of a previously described technique (12). Specimens of human bronchi were obtained within 24 h from the autopsied lungs of patients without respiratory disorders. Tissues were dissected and rinsed five times in Ca²⁺- and Mg²⁺-free HBSS and incubated overnight at 4°C in 0.1% protease (Calbiochem, San Diego, CA) solution in Ham's F12 medium containing penicillin (100 U/ml), streptomycin (100 U/ml), and fungizone (1 µg/ml). Cells were detached by a gentle jet of 20% FCS in Ham's F12 medium. The suspension was centrifuged at 200 × g for 8 min, and the cell pellet was resuspended in F12/DMEM containing 5% heat-inactivated FCS, penicillin (100 U/ml), and streptomycin (100 U/ml), and cells were plated on collagen (Vitrogen 100; Collagen Biomaterials, Palo Alto, CA)-coated plates. The medium was changed at Day 1 to F12/10 × medium, and changed on alternate days thereafter. The purity of the cultures and identity of the cells were confirmed by light microscopy and immunocytochemical staining using specific monoclonal antibodies directed against cytokeratin. Cells were used between passages 1 and 3.

Flow Cytometric Analysis

In the experiments assessing the inhibitory effects of glucocorticoids on VCAM-1 and ICAM-1 expression on BEAS-2B cells, cells were preincubated with increasing concentrations of glucocorticoids or an equivalent concentration of DMSO diluent for 24 h and then incubated with TNF- (10 ng/ml) for 24 h. Cells were then washed three times with Ca²⁺- and Mg²⁺-free HBSS and treated for 10 min with Versene (Ca²⁺- and Mg²⁺-free HBSS containing 0.02% EDTA) without trypsin, and then removed from plates by repeated pipetting. For each analysis, 1 × 10⁶ cells were incubated in 30 µl of phosphate-buffered saline/ 0.2% bovine serum albumin containing saturating concentrations of each monoclonal antibody and 4 mg/ml of human immunoglobulin (Ig)G (to reduce nonspecific binding) on ice for 30 min, as previously described (10, 13). Monoclonal antibody to ICAM-1 was RR1 (AMAC, Westbrook, ME) and VCAM-1 was BBIG-V1 (R&D Systems). The cells were washed, resuspended in saturating amounts of fluorescein-conjugated goat F(ab')₂ antimouse IgG antibody (Bio Source, Camarillo, CA) for another 30 min, and then washed again. Negative staining with propidium iodide (2 µg/ml) and a combination of scatter characteristics were used to identify a uniform population of viable cells. Fluorescence was measured using an EPICS Profile II flow cytometer (Coulter Electronics, Hialeah, FL) and was expressed as percent of control IgG mean fluorescence intensity by comparison with control staining using an irrelevant isotype-matched mouse monoclonal antibody. For each sample, at least 5,000 events were collected.

Northern Blot Hybridization Analysis

In the experiments assessing the inhibitory effects of FP on VCAM-1 and ICAM-1 messenger RNA (mRNA) expression, cells were preincubated with DMSO diluent or FP (10⁷, 10⁹, and 10¹¹ M) for 24 h and then incubated with TNF- (10 ng/ml) for 2 h. To assess the effect of FP on VCAM-1 mRNA stability, cells were preincubated with DMSO diluent or FP (10¹¹ M) for 24 h, and then incubated with TNF- (10 ng/ml) for 2 h. Actinomycin D (5 µg/ml) was then added to cells at baseline and for 2 and 4 h. Total RNA was extracted from BEAS-2B cells using the RNAzol B extraction technique (14). Samples of RNA (10 µg) were denatured with formaldehyde/formamide and subjected to electrophoresis on 1% agarose formaldehyde gels. Gels were run for 1.5 h at 85 V, and the RNA was transferred onto a GeneScreen nylon membrane (New England Nuclear, Boston, MA). Membranes were cross-linked by exposure to ultraviolet irradiation, and then were subjected to prehybridization treatment for 20 h at 37°C in 2× [1,4-piperazinebis (ethane sulfonic acid)], 50% formamide buffer containing 0.5% sodium dodecyl sulfate (SDS), and 100 µg/ml sonicated salmon-sperm DNA. Blotted membranes were hybridized with ³²P-labeled human VCAM-1 and ICAM-1 complementary DNA (cDNA) probes generated by the random hexamer priming method (0.5 × 10⁶ counts per min/ml). A 2.2-kB VCAM-1 cDNA probe excised from pcDNA3 vector was prepared by purifying plasmid DNA and digesting with restriction endonuclease NotI. A 1.6-kB ICAM-1 cDNA probe excised from Bluescript II SK vector was prepared by purifying plasmid DNA and digesting with restriction endonucleases XbaI and XhoI. Human VCAM-1 plasmid was a gift from Dr. Venkot Gopal (Clonexpress, Inc., Gaithersburg, MD). Human ICAM-1 plasmid was a gift from Dr. Richard Hampton (Johns Hopkins University, Baltimore, MD). Membranes were washed twice at room temperature in 2× saline sodium citrate (SSC), twice at 60°C in 2× SSC/0.1% SDS, again at room temperature in 0.1× SSC, and then subjected to autoradiography at 80°C using an intensifying screen. Quantification of RNA was by densitometry. Autoradiographs were scanned and analyzed with Image 1.53 software (National Institutes of Health Public Software, Bethesda, MD). Loading of lanes and integrity of total RNA were confirmed by ethidium bromide staining.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Analysis of VCAM-1 mRNA in Primary NHBE

Total RNA extracted from NHBE was reverse-transcribed using oligo(dT) as a primer and Moloney murine leukemia virus reverse transcriptase (Life Technologies, Paisley, Scotland). The VCAM-1 primers used were: forward primer 5'-ATTGGGAAAAACAGAAAAGAG-3', and reverse primer 5'-GGCAACATTGACATAAAGTC-3'; these primers produced a 642-bp product. Thermocycler settings were an initial step at 94°C for 5 min to denature and linearize cDNA, followed by 30 cycles of 94°C for 30 s to denature, 56°C for 30 s for annealing, and 72°C for 1 min for polymerization. Amplified products were electrophoresed on a 1.5% agarose gel that was then stained with ethidium bromide. Quantification of RNA was done by densitometry. Results are expressed as percent of maximal signal obtained after standardization compared with -actin.

Statistical Analysis

Data are expressed as means ± SEM. Statistical significance was assessed with the paired t test or analysis of variance test followed by Bonferroni-Dunn analysis; P < 0.05 was considered significant. Statistical analysis of data by flow cytometry was performed using percent of control mean fluorescence intensity values.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Glucocorticoids on TNF--Induced VCAM-1 Expression

Previous studies have shown that stimulation of BEAS-2B cells with TNF- and other cytokines induces expression of the adhesion molecule VCAM-1. Because we have found that glucocorticoids inhibit some epithelial cell responses, we tested the effects of several glucocorticoids on TNF--induced VCAM-1 expression on BEAS-2B cells. BEAS-2B cells were preincubated with various concentrations of glucocorticoids for 24 h before stimulation with 10 ng/ml of TNF- for another 24 h. Data in Figure 1 show that all glucocorticoids tested inhibited VCAM-1 expression induced with 10 ng/ml of TNF- in a concentration-dependent manner. -estradiol, used as a control steroid, did not change TNF--induced VCAM-1 expression. The concentrations at which each glucocorticoid produced 50% inhibition (IC₅₀) are listed in Table 1, showing a rank order of potency MF  FP > BUD  TAA > BDP  17-BMP  HC. The most potent of the glucocorticoids tested were MF and FP, both of which had IC₅₀s under 10 pM. BUD, TAA, and BDP had intermediate potency, and HC and the BDP metabolite 17-BMP were the least potent of the steroids tested.

View larger version (22K): [in this window] [in a new window]   Figure 1.   Inhibition of TNF--induced VCAM-1 expression by glucocorticoids in BEAS-2B cells. (A) Representative fluorescence histograms showing the inhibition of VCAM-1 expression (black histograms) on BEAS-2B cells. Cells were incubated with or without FP for 24 h before stimulation with 10 ng/ml of TNF- for another 24 h. IgG1 controls are gray histograms. Control expression of VCAM-1 was 107 ± 3 with buffer and 327 ± 32 after stimulation with TNF- (mean fluorescence intensity: percentage of control IgG1 ± SEM). Values for IgG1 controls were 1.07 ± 0.03. (B) Inhibition of TNF--induced VCAM-1 expression by various glucocorticoids in BEAS-2B cells. Cells were preincubated with indicated glucocorticoids for 24 h before stimulation with 10 ng/ml of TNF- for another 24 h. Values are presented as means ± SEM percent inhibition (n = 3).

FP was chosen for the subsequent studies because it is the most potent and widely used inhaled glucocorticoid in asthma. The kinetics of inhibition of VCAM-1 expression by FP are shown in Figure 2. FP was added to BEAS-2B cells at either 48, 24, 18, 6, or 3 h before, or 0, 3, or 6 h after, stimulation with TNF- (10 ng/ml). FP inhibited VCAM-1 expression completely even when FP was added at the same time as the TNF- stimulus. Moreover, FP inhibited VCAM-1 expression by 50% even when added as late as 6 h after stimulation with TNF-.

View larger version (22K): [in this window] [in a new window]   Figure 2.   Kinetics of the inhibition of VCAM-1 expression on BEAS-2B cells by FP. BEAS-2B cells were treated with 107 M FP before or after stimulation with 10 ng/ml of TNF- for 24 h. Values are presented as means ± SEM percent inhibition (n = 3).

Effect of Glucocorticoids on TNF--Induced ICAM-1 Expression

Data in Figure 3 show that preincubation with FP (10⁷ M) for 24 h did not inhibit the increase of ICAM-1 and ICAM-1 mRNA expression induced by TNF- in BEAS-2B cells. In data not shown, BUD (10⁷ M, n = 3) also did not inhibit increased expression of ICAM-1 and ICAM-1 mRNA.

View larger version (17K): [in this window] [in a new window]   View larger version (34K): [in this window] [in a new window]   Figure 3.   Effect of FP on TNF--induced expression of ICAM-1 (A) and ICAM-1 mRNA (B) in BEAS-2B cells. (A) Flow cytometry analysis of ICAM-1 expression. Cells were preincubated with 107 M FP for 24 h before stimulation with 10 ng/ml of TNF- for another 24 h. Values presented are means ± SEM fluorescence intensity as percentages of control IgG1 (n = 3). (B) Northern blot analysis of ICAM-1 mRNA expression. Cells were pretreated with 107 M FP for 24 h before stimulation with 10 ng/ ml of TNF- for 2 h. (I) Representative autoradiography of ICAM-1 mRNA expression. (II) Densitometric analysis of the autoradiography. Values are presented as means ± SEM (n = 3). (III) Ethidium bromide staining of representative gel showing equal loading of lanes and integrity of mRNA.

Effect of FP on TNF--Induced VCAM-1 mRNA Expression

To determine whether the inhibition of VCAM-1 expression by glucocorticoids was due to reduced levels of mRNA, Northern blot experiments were performed. BEAS-2B cells were preincubated with 10¹¹, 10⁹, or 10⁷ M concentrations of FP for 24 h before stimulation with TNF- (10 ng/ml) for 2 h. As shown in Figure 4, FP significantly inhibited TNF--induced VCAM-1 mRNA expression in a concentration-dependent manner. An FP concentration of 10¹¹ M inhibited VCAM-1 mRNA expression by 63%, suggesting that the IC₅₀ was below 10 pM as was found in the flow cytometry experiments (see Table 1).

View larger version (35K): [in this window] [in a new window]   Figure 4.   Inhibition of TNF--induced VCAM-1 mRNA expression by FP in BEAS-2B cells. Cells were pretreated with either DMSO diluent control or indicated concentrations (1011 to 107 M) of FP for 24 h before stimulation with 10 ng/ml of TNF- for 2 h. (I) Representative autoradiography of VCAM-1 mRNA expression. (II) Densitometric analysis of the autoradiography. Values are presented as means ± SEM (n = 3). Statistically significant differences from control assessed with Bonferroni-Dunn analysis are indicated by an asterisk (*P < 0.001). (III) Ethidium bromide staining of representative gel showing equal loading of lanes and integrity of mRNA.

Effect of Actinomycin D on Inhibition of TNF--Induced VCAM-1 mRNA Expression by FP

To determine whether glucocorticoid inhibition of VCAM-1 expression results from reduced stability of VCAM-1 mRNA, cells were preincubated with 10¹⁰ M FP or an equivalent dilution of DMSO for 24 h before they were stimulated with TNF- (10 ng/ml) for 2 h, and then treated with 5 µg/ ml of actinomycin D for 0, 2, or 4 h. As shown in Figure 5, after treatment with actinomycin D for 4 h, the amount of TNF--induced VCAM-1 mRNA decreased 17%, whereas FP treatment enhanced the decrease in the amount of VCAM-1 mRNA up to 53% (half-life < 4 h), suggesting that FP reduced the stability of VCAM-1 mRNA.

View larger version (43K): [in this window] [in a new window]   Figure 5.   Effect of actinomycin D on TNF--induced VCAM-1 mRNA expression in BEAS-2B cells. Cells were pretreated for 24 h with 1010 M FP or an equivalent amount of DMSO diluent control before stimulation with 10 ng/ml of TNF- for 2 h. Cells were then incubated with 5 µg/ml of actinomycin D for 0 to 4 h. (I) Representative autoradiography of VCAM-1 mRNA expression. Lanes are (left to right): Control 0 h, control 2 h, control 4 h, FP 0 h, FP 2 h, and FP 4 h. Actinomycin D was present in all cases. (II) Densitometric analysis of the autoradiography. Values presented are the mean percentages of initial levels ± SEM (n = 4). Statistically significant differences between two groups assessed with a paired t test are indicated by an asterisk (*P < 0.05).

Effect of FP on VCAM-1 mRNA Expression in Primary Human Bronchial Epithelial Cells

The effect of FP on the expression of VCAM-1 mRNA by primary bronchial epithelial cells was examined by RT- PCR analysis. Cells were pretreated with 10⁷ M of FP or an equivalent amount of DMSO for 24 h, and then stimulated with 100 ng/ml of TNF- for 24 h. As is shown in Figure 6, FP reduced TNF--induced VCAM-1 mRNA to undetectable levels.

View larger version (39K): [in this window] [in a new window]   Figure 6.   VCAM-1 mRNA analysis of primary NHBE by RT- PCR. Cells were pretreated for 24 h with 107 M of FP or an equivalent amount of DMSO diluent control before stimulation with 100 ng/ml of TNF- for 24 h. (I) Representative image of VCAM-1 mRNA expression in primary human bronchial epithelial cells is from one of three similar experiments. (II) -actin expression in same samples showing equal loading of lanes. (III) Densitometric analysis of the VCAM-1 mRNA expression. Values presented are means ± SEM (n = 3).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have previously reported (10) that VCAM-1 is upregulated in BEAS-2B cells following stimulation with TNF-, IL-1, and IL-4, but not with IL-5 or interferon- (IFN-). The present study demonstrates that glucocorticoids markedly inhibited TNF--induced VCAM-1 expression induced in the human bronchial epithelial cell line BEAS-2B. Glucocorticoids attenuated both the expression of VCAM-1 protein and VCAM-1 mRNA, as detected by flow cytometry and Northern blotting, respectively.

We initially investigated the effect of several glucocorticoids on VCAM-1 expression induced in BEAS-2B cells. All glucocorticoids tested inhibited VCAM-1 expression in a concentration-dependent manner. We used glucocorticoids that were developed for topical use for treatment of asthma. Of the glucocorticoids now available for treatment of asthma by inhalation, FP was the most potent. Based on estimated IC₅₀ values, FP was 80-fold more potent than BUD, 120-fold more potent than TAA, and 1,200-fold more potent than BDP. Concentrations of FP as low as 10¹¹ M inhibited expression of VCAM-1 protein by 74% and inhibited expression of VCAM-1 mRNA by 63%, whereas 10⁷ M FP inhibited VCAM-1 protein expression by 100% and VCAM-1 mRNA expression by 89%. The large differences in potency among inhaled glucocorticoids in vitro in suppressing VCAM-1 expression are not reflected in vivo when the potency of these drugs in reversing symptomatic endpoints (e.g., forced expiratory volume at 1 s, eosinophil number, or bronchial reactivity) is compared. It seems likely that inhalation of any of these drugs will lead to saturation or near saturation of glucocorticoid receptors (GRs) in epithelial cells and other airway cells. In such a case, other parameters (such as the residence time of the drug in the airways and the off rate of the steroid from receptors) may determine clinical potency. In our previous studies (10) we have shown that VCAM-1 expression on BEAS-2B can mediate eosinophil adhesion. In those studies we also found that primary airway epithelial cell cultures also expressed VCAM-1, albeit at significantly lower levels than in BEAS-2B cells after stimulation with TNF-. In the present study we found that VCAM-1 mRNA induced in primary airway epithelial cells was inhibited by FP. Although the relevance of epithelial VCAM-1 expression in vivo is unknown, the results of the present study suggest that inhibition of VCAM-1 expression in bronchial epithelial cells by glucocorticoids may exert an anti-inflammatory effect.

The rank order of potency of glucocorticoids for inhibiting VCAM-1 expression on bronchial epithelial cells was similar to our previous results using other assays, including assays of basophil histamine release and eosinophil survival (15). However, VCAM-1 expression was more sensitive to some glucocorticoids (FP, MF, and HC) but not others (BUD, TAA, and BDP) compared with basophil histamine release and eosinophil survival. For example, the IC₅₀s of FP varied over two orders of magnitude. They were: 0.09 nM for histamine release, 0.25 nM for eosinophil survival, and 0.0068 nM for VCAM-1 expression. The IC₅₀s of HC were 170 nM for histamine release, 2,500 nM for eosinophil survival, and 12 nM for VCAM-1 expression. However, the IC₅₀ of BDP for VCAM-1 expression was not lower than in the other assays: 1.0 nM for histamine release, 30 nM for eosinophil survival, and 8.1 nM for VCAM-1 expression. Whether this results from cell type differences or the differences in molecular targets is unknown. Interestingly, 17-BMP was less potent than BDP despite the suggestion from others that 17-BMP is the active metabolite responsible for the efficacy of BDP in the lung (16).

The inhibitory effect of glucocorticoids on VCAM-1 expression in BEAS-2B epithelial cells contrasts with our previous studies using endothelial cells, in which glucocorticoids did not inhibit VCAM-1 expression induced by TNF-, IL-1, IL-4, or IL-13 (17, 18). However, FP has been reported to inhibit TNF--induced VCAM-1 expression on endothelial cells (19). Dexamethasone has been reported to inhibit VCAM-1 expression induced by TNF- or IL-1 on synovial fibroblasts, whereas it did not inhibit VCAM-1 expression induced by TNF- or IL-1 on renal tubular epithelial cells (20, 21). Similarly, the effects of glucocorticoids on ICAM-1 expression have been conflicting among studies using different cell types. In the present study, FP and BUD did not inhibit TNF--induced ICAM-1 and ICAM-1 mRNA expression in BEAS-2B cells. Several other investigators have found no effect of glucocorticoids on ICAM-1 expression using a variety of cell types, including the NCI-H292 human bronchial epithelial cell line stimulated with TNF-, a human vascular endothelial cell line stimulated with TNF- and IFN-, human umbilical vein endothelial cells stimulated with lipopolysaccharide, and a renal tubular epithelial cell line stimulated with TNF- or IL-1 (17, 21). On the other hand, dexamethasone was found to inhibit ICAM-1 expression on the NCI-H292 human bronchial epithelial cell line, the SW-1353 human chondrosarcoma cell line, or the A-549 human adenocarcinoma cell line when IL-1, IFN-, or 12-O-tetradecanoyl phorbol-13-acetate were used as stimuli (22, 25, 26). Thus, numerous previous studies have shown that glucocorticoid effects vary depending on cell type, stimulus, and species.

A considerable amount of new information is now available regarding the mechanisms of glucocorticoid regulation of gene expression. Glucocorticoids can act at the transcriptional and post-transcriptional levels. GR complexes may bind to positive glucocorticoid responsive elements (GREs) on target genes to stimulate transcription, as in the case of 2-receptors (27, 28), or to negative GREs to inhibit gene transcription, as in the case of IL-6 (29). Alternatively, the GR can inhibit gene expression indirectly by interacting with and interfering with the function of transcription factors, such as AP-1 and nuclear factor (NF)-B. It has been found that two tandem binding sites for NF-B in the VCAM-1 promotor are necessary for cytokine-mediated transcriptional response, the activation of which may therefore be downregulated by glucocorticoids at the transcriptional level (30). GR can suppress NF-B-mediated responses further by inducing expression of IB, an inhibitor of NF-B (31). In addition to transcriptional regulation, glucocorticoids may regulate gene expression at several post-transcriptional levels, including destabilization of mRNA. Regions of the 3'- untranslated region of mRNA referred to as AU response elements may be required for glucocorticoids to increase mRNA turnover (32). The VCAM-1 3'-untranslated region contains such AU-rich regions. On the basis of studies with actinomycin D, the effect of glucocorticoids on VCAM-1 mRNA expression in bronchial epithelial cells may be regulated partially at the post-transcriptional level.

The in vivo contribution that VCAM-1 expression makes to airway inflammation in bronchial epithelium is unknown. The expression of VCAM-1 that we have observed in primary bronchial epithelial cells was low (10), suggesting that if it does play a role, that role may be more likely via activation of very late antigen (VLA)-4-positive cells than via adhesion responses. The VCAM-1/VLA-4 interaction has been shown to enhance eosinophil superoxide anion generation (35) and prolong eosinophil survival (36). It therefore seems possible that inhaled glucocorticoids may block the VCAM-1/VLA-4 interaction in bronchial epithelium, resulting in a suppression of the function or survival of VLA-4 positive cells, such as eosinophils, in the airways.

In conclusion, this study has provided evidence for downregulation of VCAM-1, but not ICAM-1, expressed in human bronchial epithelial cells by glucocorticoids. These findings suggest a potential role of inhaled glucocorticoids in asthma in suppressing inflammatory reactions in the airways by regulating VCAM-1 expression on bronchial epithelium.