Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

An increase in the volume of airway smooth muscle (ASM) as a result of hyperplasia (1, 2) and hypertrophy (3) of smooth muscle cells contributes significantly to the airway wall remodeling occurring in asthma (4). Mathematical modeling studies have shown that the resultant airway wall thickening could itself account for much of the airway hyperresponsiveness (AHR) in asthma (4, 7, 8).

A large number of structurally and functionally diverse inflammatory mediators and growth factors generated or released during episodes of asthma stimulate proliferation of human ASM cells in culture, and are probable contributors to airway wall remodeling in vivo (9, 10). Most, if not all, mitogens for ASM activate extracellular-regulated kinases (ERK), the key mitogenic signaling proteins, which are also known as mitogen-activated protein kinase (MAPK) (11). Persistent activation of the enzymatic activity of these proteins is required for cells to enter into the S phase of the cell cycle (14, 15). The signals downstream of ERK that have been implicated in mitogenesis include phosphorylation of transcription factors including c-jun, Ets, and Elk-1, and synthesis of cell cycle regulatory proteins such as the cyclin D family (16). In particular, cyclin D1 synthesis may be a key regulatory step for progression of cells to the S phase (20). Cyclin D1 binds to cyclin-dependent kinase 4 (cdk4) to form a complex which, upon activation by cyclin-activating kinase (CAK), phosphorylates the retinoblastoma protein (pRb), a key step in traversal from the G₁ to the S phase of the cell cycle (24).

Inhaled glucocorticoids are now considered to be first-line agents in the treatment of chronic asthma. Prophylactic use of inhaled glucocorticoids reduces AHR, but the maximum impact on AHR requires prolonged treatment, whereas airway inflammation resolves within 2 wk (27- 30). This contrast is consistent with the suggestion that the glucocorticoid-induced resolution of long-term, slowly reversible aspects of asthma pathogenesis underlies the slow onset of the maximum effects of these agents on AHR.

Glucocorticoids inhibit mitogen-stimulated proliferation of smooth muscle cells cultured from rabbit (31), bovine (32), and human airways (33). However, little is known of the particular site(s) at which glucocorticoids inhibit the mitogenic signaling pathway. In Swiss 3T3 fibroblasts, treatment for 48 h with glucocorticoids reduces ERK activation stimulated by insulin-like growth factor (IGF)-1 (34), and reduces antigen-stimulated ERK activity in mast cells (35). However, studies with 3T3-F442A fibroblasts suggest that glucocorticoids inhibit ERK activation by some mitogens via inhibitory actions on upstream pathways, and not by inhibiting ERK activity per se (36). In the fibroblast L929 cell line and in lung alveolar epithelial cells, dexamethasone inhibits proliferation, but has no effect on levels of cyclin D1, D2, or D3 in cells in the mid-log phase of cell growth (37, 38). In continuously cycling rat mammary tumor cells arrested with glucocorticoids, subsequent withdrawal of glucocorticoids increases levels of cyclin D1 and c-myc mRNA (39). c-Myc, a transcription factor that is increased in the G₁ phase of the cell cycle, may regulate cyclin D1 expression (20, 40). In lymphoid cells, HeLa cells, osteosarcoma cell lines, and fibrosarcoma cells, glucocorticoids inhibit mitogen-stimulated increases in cyclin D3 (41). However, the effects of glucocorticoids on mitogen-stimulated increases in cyclin D1 levels in quiescent, nontransformed cells have yet to be investigated.

In this study we examined the effects of the glucocorticoids dexamethasone and fluticasone propionate (44, 45) on the mitogen-stimulated intracellular pathways involved in the signaling of proliferation in human cultured ASM. We showed in these nontransformed cells that glucocorticoids reduce thrombin-stimulated increases in cyclin D1 protein and mRNA levels, and reduce pRb phosphorylation, by acting on a mitogen signaling pathway that is downstream of or parallel to the ERK cascade.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture

Human ASM cell cultures were generated from bronchi (0.5-2 cm diameter) obtained from lung-resection patients or heart-lung transplant recipients. Smooth muscle was microdissected with the aid of a binocular operating microscope (×10 magnification) and was enzymatically digested with collagenase and elastase to generate cell suspensions, which were then cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS) as previously described (33). Cells were passaged weekly at a 1:3 split ratio through a 10-min exposure to 0.5% trypsin (in PBS containing 1 mM ethylenediamine tetraacetic acid [EDTA]), generating a cell density of approximately 1 × 10⁴ cells/cm². Cells used for experiments were at passage numbers 4-14, a period in which there was no relationship between cell passage number and responsiveness to growth factors or inhibitors, and during which the expression of smooth muscle -actin is maintained (46).

Immunohistochemistry

The cellular composition of the cultures was determined from expression of smooth muscle-specific -actin and myosin. Cells were subcultured into 8-well glass chamber slides and allowed to grow to confluence, after which the medium was replaced with serum-free DMEM (containing 0.25% [wt/vol] bovine serum albumin [BSA] and 1% [vol/ vol] Monomed A; CSL, Melbourne, Australia) for 7 d to render the cells quiescent and to allow reexpression of contractile proteins (47). Cells were then fixed in ice-cold acetone for 20 s and stored for up to 4 wk at 4°C before staining. Immunohistochemical staining was performed as described previously (46), using monoclonal antibodies to smooth-muscle -actin (mouse monoclonal antibody; Dako Corporation, Carpinteria, CA) and myosin (rabbit polyclonal antibody, provided by Prof. M. Sparrow, Perth, Australia), and was followed by incubation with a horseradish peroxidase-conjugated secondary antibody (Silenus Laboratories, Melbourne, Australia). Cells were also stained for platelet endothelial adhesion molecule (PECAM)-1 (mouse monoclonal anti-CD31 antibody; Dako), a marker present on endothelial cells. Staining of the fixed cells was observed by light microscopy. More than 95% of the cells in each of the cell cultures stained positively for -actin and myosin (44).

DNA Synthesis

Cells were plated in 24-well plates at approximately 1.5-2 × 10⁴ cells/cm² and allowed to grow for 72-96 h to monolayer confluence in DMEM containing 10% FCS. Thereafter, the cells were growth-arrested in media (containing 0.25% BSA) without FCS for a further 24 h to synchronize the cells in the G₀ phase of the cell cycle before addition of a maximally effective concentration of thrombin (0.3 U/ ml) (48). Glucocorticoids were added 60 min before addition of mitogen, and the ERK kinase (MEK1) inhibitor PD98059 (49, 50) (New England Biolabs, Beverley, MA) was added 30 min before mitogen. A growth supplement containing insulin, transferrin, and selenium (Monomed A, 1% [vol/vol]; CSL Australia) was added to all wells at the time of mitogen stimulation to provide progression factors essential for the mitogenic activity of growth factors such as thrombin, epidermal growth factor (EGF), and basic fibroblast growth factor (bFGF) (33, 48). After stimulation with mitogens for 24-26 h, the cells were incubated for a further 4 h in 1 µCi/ml [³H]thymidine to allow incorporation of radioactivity into DNA, and were then harvested as described previously with a binding harvester (Filtermate; Packard, Melbourne, Australia) (46). Radioactivity was measured by liquid scintillation counting (Topcount; Packard).

Cell Proliferation Assay

In separate experiments, cells were seeded into the wells of a 96-well microtiter plate at a density of 0.5 × 10⁴ cells/ cm², were grown under the conditions described earlier, and were stimulated for 72 h to ensure sufficient time for cell division to take place. Cells were fixed by addition of 10% formalin in PBS for at least 30 min at room temperature. Fixative was aspirated and replaced by 50 µl of prefiltered methylene blue (1% [wt/vol]) in 0.01 M borate buffer (pH 8.5) for 30 min at room temperature (13, 51). Cells were then washed twice with 0.01 M borate buffer (200 µl), after which the stain was solubilized in 100 µl 0.1 M HCl/ethanol (1:1 [vol/vol]). Absorbance was measured at 650 nm with an automated plate reader (Multiskan RC; Labsystems, Helsinki, Finland), and cell number was calculated against the measured optical densities of a standard curve of cell number constructed by seeding cells of the same culture and passage number at densities of 0-5 × 10⁴ cells/cm² at least 6 h before fixation (to ensure adequate adhesion).

Flow Cytometry

Flow-cytometric analyses were done to ascertain whether dexamethasone caused apoptosis of ASM. Cells were grown in 6-well plates and then stimulated for 24 h exactly as described for the studies involving DNA synthesis; the cells were then washed twice in PBS and harvested by incubation in 0.5% trypsin (1 mM EDTA) for 30 min at 37°C. The resulting cell suspension was washed twice in PBS before resuspension in 1 ml of 70% ethanol for storage at 20°C before staining. Approximately 24 h before flow cytometry, the cells were triturated through an 18-gauge needle to ensure a single-cell population. DNA profiles were determined by collecting 10,000 events in a Becton Dickinson FACScan flow cytometer (Becton Dickinson, Franklin Lakes, NJ), using MODFIT software (Verity Software House, Topsham, ME).

Analysis of ERK Activation

Human ASM cells were grown in 6-well plates and stimulated under conditions identical to those used for estimation of DNA synthesis. After incubations with mitogen for 5 min, 2 h, or 12 h, cells were washed in PBS and harvested in ice-cold buffer (containing 10 mM Tris, 150 mM NaCl, 2 mM ethyleneglycol-bis-(-aminoethyl ether)-N,N',-tetraacetic acid [EGTA], 2 mM EDTA, 2 mM dithiothreitol, 1 mM orthovanadate, 1 mM phenylmethylsulfonyl fluoride [PMSF], 10 µg/ml leupeptin, and 10 µg/ml aprotinin, pH 7.4). Lysates were aspirated into Eppendorf tubes, snap frozen in liquid nitrogen, thawed on ice to enhance cell lysis, and then centrifuged at 18,000 × g at 4°C. Supernatants were collected and an aliquot was removed for protein analysis. Lysates were then assayed for ERK activity on the same day, using the Biotrak p42/p44 MAPK enzyme assay system (Amersham, Cardiff, UK). Briefly, 15 µl of lysate was added to 10 µl of a substrate buffer (containing KRELVEPT⁶⁶⁹PAGEAPNALLR, a synthetic peptide that is selective for phosphorylation by MAPK). The reaction was started by addition of magnesium [³²P]adenosine triphosphate ([³²P]ATP), incubated for 30 min at 30°C, and terminated by adding orthophosphoric acid, after which the reaction mixture was pipetted onto filter paper discs. Nonincorporated radioactivity was removed by washing the discs twice with 0.1% (vol/vol) acetic acid, followed by two washes with distilled water (dH₂O). Radioactivity was measured through liquid scintillation counting (Topcount; Packard). ERK activity is expressed as pmol/ min/mg protein.

A similar technique was also used with immunoprecipitation of ERK and its subsequent reaction with a nonspecific protein substrate, myelin basic protein (MBP), to confirm the data provided by the Amersham kit assay. Briefly, cells were treated as described earlier for the kit assay, but were lysed in 110 µl of ice-cold buffer (50 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid [Hepes], pH 7.4; 0.27 M sucrose; 50 mM NaF; 50 mM -glycerophosphate; 1% [vol/vol] Triton-X 100; 0.5 mM orthovanadate; 2 mM EDTA, 333 kIU/ml aprotinin, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride (Pefa Bloc); 0.5 µg/ml leupeptin), placed on ice for 10 min, and then centrifuged at 12,000 × g for 5 min. The supernatant was collected and stored at 80°C. Immunoprecipitation was achieved by adding of 0.2 µg of goat polyclonal IgG anti-ERK1 antibody (C-16; Santa Cruz Biotechnology, Santa Cruz, CA) to each sample; parallel control samples had no antibody added. After 10 min of equilibration on ice, 20 µl of a 30% (wt/vol) protein G-Sepharose slurry (Pharmacia, Uppsala, Sweden) was added to all samples, and the samples were rotated for 60 min at 4°C. The samples were then centrifuged at 760 × g for 30 s, and washed twice in buffer (containing 50 mM Hepes, pH 7.4; 150 mM NaCl; 1% Triton-X 100). The washed, protein G-complexed ERK was assayed for kinase activity by addition of 20 µg of MBP and 10 µl of kinase assay buffer (180 µM Hepes, pH 7.4; 90 µM ATP; 90 mM MgCl₂; 10 µCi/ ml [-³²P]ATP) to a total assay volume of 30 µl, with the assay run for 30 min at 30°C. The assay was terminated by adding 10 µl of 180 mM H₃PO₄, and the resultant mixture was centrifuged at 760 × g. Twenty-five microliters of aspirate was spotted onto 2-cm² P81 filter paper (Whatman International, Maidstone, UK), washed with orthophosphoric acid, and dried, after which Phosphorylation of MBP was assessed by Phosphorimager analysis and densitometry (PhosphorImager; Molecular Dynamics, Sunnyvale, CA).

Immunoblot Analysis

Cells were grown in 6-well plates as described for the ERK activity assay, and were incubated for 12 h (phosphorylated MAPK) and 20 h (cyclin D1 and pRb) with mitogen with or without inhibitor. The cells were then washed twice in ice-cold PBS and lysed on ice for at least 20 min in buffer (containing 100 mM NaCl; 10 mM Tris-HCl, pH 7.5; 2 mM EDTA; 0.5% [wt/vol] deoxycholate; 1% [vol/vol] Triton X-100; 1 mM PMSF; 1 mM MgCl₂; 100 IU/ml aprotinin, 100 µM orthovanadate), scraped and transferred to Eppendorf tubes, and centrifuged at 760 × g for 5 min, after which the supernatant was aspirated and stored at 20°C. Aliquots were removed for protein assay (Biorad Reagent; Biorad, Sydney, Australia). Identical amounts of protein were separated electrophoretically on 12% sodium dodecylsulfate (SDS)-polyacrylamide gels, and proteins were transferred onto nitrocellulose membranes (Hi-Bond C; Amersham) for Western blotting. The hypo- and hyperphosphorylated forms of pRb were resolved on separate 7.5% gels to enhance detection of the increase in molecular weight produced by phosphorylation. Membranes were blocked in 5% (wt/vol) Carnation milk for 60 min at room temperature, and were then incubated overnight with a phosphospecific anti-p42/p44 MAPK antibody (rabbit polyclonal IgG, 1:1,000; New England Biolabs), anti-cyclin D1 (rabbit polyclonal IgG, 1:1,000; Upstate Biotechnology, Lake Placid, NY), or anti-Rb (rabbit polyclonal IgG; Santa Cruz Biotechnology) at 4°C. Immunoblotted membranes were incubated with horseradish peroxidase (HRP)-conjugated antirabbit IgG (Silenus Laboratories, Hawthorn, Australia), and the HRP activity was visualized with an enhanced chemiluminescence kit (Amersham). Densitometry was performed to quantify relative levels of phosphospecific MAPK and cyclin D1 (Personal Densitometer; Molecular Dynamics).

Northern Blot Analysis

Cells were grown in 175 cm² flasks under the same conditions as used in the proliferation experiments, and were stimulated for 16 h with thrombin at 0.3 U/ml. The supernatant was then aspirated, 3 ml of Trizol reagent (GIBCO BRL, Melbourne, Australia) was added, and total RNA was isolated according to the manufacturer's instructions. Poly A⁺ mRNA was extracted from ~ 70 µg of total RNA, using oligo(dT)₂₅ Dynabeads (Dynal, Oslo, Norway), and was separated on a 1.2% formaldehyde denaturing gel and transferred to Hybond N⁺ nylon membranes (Amersham, UK) by alkali blotting (0.04 M NaOH). Cyclin D1 mRNA was detected by Northern hybridization (Megaprime Labelling Kit; Amersham, Amersham, UK) through use of a 440-bp human complementary DNA (cDNA) probe (52) labeled with [-³²P]deoxycytosine triphosphate ([-³²P]dCTP) (Amersham, UK). The membranes were hybridized overnight at 65°C, and were washed once with 2× standard saline citrate (SSC) + 0.1% SDS at 55°C for 30 min and once with 1 × SSC + 0.1% SDS at 60°C for 30 min, after which they were exposed to autoradiography film (Hyperfilm MP; Amersham) at 20°C for ~ 2 d. After autoradiography, the membranes were stripped and hybridized with a -actin cDNA probe (53), and were exposed as described earlier (n = 6). The autoradiographs were quantitated with a Molecular Dynamics Personal Densitometer. To control for loading of RNA, the cyclin D1 mRNA levels were normalized against the levels of -actin. Although thrombin stimulated -actin expression, the level of -actin expression was unaffected by dexamethasone.

Materials

Fluticasone propionate, RU 486, and dexamethasone were dissolved in 100% dimethylsulfoxide (DMSO) to a stock concentration of 10 mM, and were then diluted in PBS or DMEM. The maximum concentration of DMSO vehicle in the cell supernatant was 0.01%. DMSO concentrations of 0.3% or more were required to produce modulations of basal and/or thrombin-stimulated [³H]thymidine incorporation. PD98059 was dissolved to 50 mM in 100% DMSO and was then diluted in DMEM, yielding a maximum final supernatant concentration of DMSO of 0.1%. Thrombin was prepared in PBS containing 0.25% BSA.

Statistical Analysis

Each experiment was conducted with at least three different cell cultures obtained from different individuals. Each treatment for the DNA synthesis assays was applied in triplicate, whereas the cell proliferation assays were performed in sextuplicate. Results are expressed as grouped data from multiple cultures, and are expressed as the mean ± SE of n cultures. Results for DNA synthesis, cyclin D1 levels, and ERK activity assays are expressed as a percentage of the thrombin response for that culture. In the cell proliferation assays, fold increments were calculated by dividing the response of treated wells by that of the unstimulated wells on the same plate. Values of the negative log of the concentration required to produce a 50% inhibition of the thrombin response (pIC₅₀) were calculated for the inhibitory effect of glucocorticoids on DNA synthesis, cell proliferation, and cyclin D1 protein levels, using nonlinear curve fitting available in the Graphpad Prism 2.01 software system (Graphpad Software Inc., San Diego, CA). Grouped data were analyzed, after log transformation, by analysis of variance (ANOVA), with Dunnet's post hoc paired comparisons used to identify individual differences. Paired t tests were conducted after normalization by log transformation for those experiments in which a single treatment was analyzed. Differences were considered statistically significant at a two-tailed P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Fluticasone Propionate Inhibits Thrombin-Stimulated Increases in DNA Synthesis and Cell Number in a Concentration-Dependent Manner

We have previously shown that thrombin-mediated ASM cell proliferation is inhibited by synthetic glucocorticoids such as dexamethasone, methylprednisolone, and the physiologic glucocorticoid hydrocortisone (33). We analyzed and compared the effects of the inhaled glucocorticoid fluticasone propionate (0.01-100 nM) on human ASM cell proliferation with those of dexamethasone (0.1-100 nM). Glucocorticoids were added 60 min before exposure of cultured ASM cells to thrombin at 0.3 U/ml (Figure 1A). Thrombin-stimulated [³H]thymidine incorporation was significantly inhibited by fluticasone propionate, with a threshold inhibitory concentration of 0.01-0.1 nM, and a maximal inhibitory effect of approximately 85% was observed at concentrations of 1 nM or more. Dexamethasone was inhibitory at concentrations above 1 nM, and had a maximum inhibitory effect of approximately 75%. The pIC₅₀ value for inhibition of DNA synthesis by fluticasone propionate was significantly greater than that for dexamethasone (Table 1).

View larger version (14K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 1.   (A) Effect of glucocorticoids on DNA synthesis. Cells were grown to monolayer confluence in 24-well plates in 10% FCS, then stimulated for 28 h with thrombin at 0.3 U/ml, after a 24 h serum-deprivation step to synchronize the cells in the G0 phase of the cell cycle. Fluticasone propionate (0.01-100 nM, n = 4, filled squares) or dexamethasone (0.1-1,000 nM, n = 4, filled circles) was added 60 min before thrombin addition. Incorporation of [3H]thymidine (1 µCi/ml) was used as an index of DNA synthesis. Results are expressed as a mean percentage ± SE of the control thrombin response. (B) Effect of glucocorticoids on cell proliferation. Cells were grown as above in 96-well plates and stimulated with thrombin at 0.3 U/ml for 72 h in the absence or presence of fluticasone propionate (0.1-10 nM, n = 7, closed squares) or dexamethasone (1-100 nM, n = 6, closed circles). Cells were then fixed in 10% formalin and stained with methylene blue as described in MATERIALS AND METHODS. Cell number was obtained by reference to a cell number calibration curve generated by seeding cells at a density of 0-5 × 104 cells/cm2. Results are expressed as mean fold increments ± SE over the basal cell number. Stars indicate P < 0.05 versus control thrombin response.

The effect of the glucocorticoids on cell proliferation induced by thrombin was examined by measuring methylene blue uptake after a 72-h incubation of the cells (Figure 1B). Thrombin stimulated a 75% increase in cell number, from an average initial cell density of 1.1 × 10⁴ cells/cm². Fluticasone propionate significantly inhibited thrombin-stimulated increases in cell number, with a threshold concentration of 0.1-1 nM, and was significantly more potent than dexamethasone in this respect (Table 1). Both dexamethasone (100 nM) and fluticasone propionate (10 nM) completely inhibited the thrombin-stimulated increase in cell number. Neither agent had any effect on the basal cell number over the 72 h duration of the experiment (data not shown), and there were no cytotoxic effects as detected by trypan blue uptake (data not shown), confirming our previous observations of a lack of cytotoxicity of glucocorticoids in ASM (33). Furthermore, flow-cytometric analyses indicated that dexamethasone either alone or in combination with thrombin did not elicit an apoptotic response, since there was no detectable change in the population of cells with a sub-G₀/G₁ DNA content. This was established through the absence of cells having a fluorescence level below the first peak (at 50 channels [i.e., G₀/G₁ DNA content]) and greater than that of the debris (at 10 channels) (Figure 2).

View larger version (24K): [in this window] [in a new window]   Figure 2.   Representative flow-cytometric DNA content profiles of cells incubated under identical conditions to those used for DNA synthesis experiments (Figure 1). DNA staining by propidium iodide (PI) is plotted on the x-axis versus cell number on the y-axis. An increase in PI staining represents cells with increased DNA content. The two peaks, from left to right of the figure, represent cells in the G0/G1 and G2/M phases of the cell cycle, respectively. Cells undergoing S phase are indicated by the population of cells midway between the two peaks. Neither dexamethasone nor thrombin caused the appearance of cells with a sub-G0/G1 DNA content (n = 3). The proportion of cells in the S phase is increased by thrombin and reduced by dexamethasone.

Dexamethasone Inhibits DNA Synthesis via a Glucocorticoid Receptor-Dependent Action

In order to ascertain the role of the glucocorticoid receptor in the inhibitory effects of the glucocorticoids, increasing concentrations (1-1,000 nM) of the glucocorticoid receptor antagonist RU 486 (54) were added to cells 30 min before dexamethasone addition (and therefore 90 min before mitogen stimulation). RU 486 (10-1,000 nM) restored the thrombin response in a concentration-dependent manner (Figure 3). A full restoration of the response was not achieved, since the direct inhibitory effects of RU 486 itself on both basal and thrombin-stimulated cells precluded the use of concentrations of 3 µM or more (data not shown). RU 486 is known to have partial agonist activity at the glucocorticoid receptor (55), perhaps by enhancing the binding of receptor coactivators such as L7/SPA and decreasing the binding of corepressor molecules such as N-CoR or SMRT (56).

View larger version (12K): [in this window] [in a new window]   Figure 3.   Inhibition of the glucocorticoid effect on ASM cell DNA synthesis by the glucocorticoid receptor antagonist RU 486. The line graph presents data from cells that were pretreated with dexamethasone (100 nM) and stimulated with thrombin (0.3 U/ml), and which had also been incubated with increasing concentrations of RU 486 (1-1,000 nM, added 30 min before dexamethasone). Results are expressed as the mean percentage ± SE of the control thrombin-stimulated response (filled histogram). (Stars indicate P < 0.05 versus thrombin response; crosses indicate P < 0.05 versus thrombin response in the presence of dexamethasone [100 nM; crosshatched histogram].)

Effect of Glucocorticoids on Phosphorylation of the Restriction Protein pRb by Thrombin

The retention of the inhibitory effect of glucocorticoids on ASM cell DNA synthesis when added as late as 18 h after addition of thrombin (33) is compatible with an action on the restriction point, which occurs in the mid- to late-G₁ phase, the duration of which is 22 h in ASM cells stimulated with thrombin (48). Hyperphosphorylation of pRb is widely accepted as the biochemical event that defines the restriction point, after which the cell is committed to DNA synthesis (57, 58).

Hyperphosphorylation of pRb results in an increase in its molecular weight that can be detected electrophoretically. Thrombin (0.3 U/ml) stimulated a large increase in the levels of both hypo- and hyperphosphorylated pRb (Figure 4). Dexamethasone (100 nM) pretreatment had no marked effect on the level of pRb in unstimulated cells, but reduced the levels of both hyper- and hypophosphorylated pRb in thrombin-treated cells.

View larger version (23K): [in this window] [in a new window]   Figure 4.   Dexamethasone (Dex 100 nM) inhibits phosphorylation of the restriction protein pRb and reduces the increase in the total level of pRb in response to thrombin (0.3 U/ml) stimulation. pRb was resolved on a 7.5% SDS-polyacrylamide gel electrophoresis (PAGE) gel, and the resulting Western blot is representative of experiments with five cultures.

Fluticasone Propionate or Dexamethasone Inhibits Thrombin-Stimulated Increases in the Levels of the Cell Cycle Protein Cyclin D1

We investigated the effects of the glucocorticoids used in the study on the key cell cycle regulatory protein cyclin D1, since this cyclin partners with cdk4, which phosphorylates pRb (24). We measured cyclin D1 levels after 20 h of thrombin stimulation, a time point late in the G₁ phase of the cell cycle (33). After a 20 h incubation, thrombin at 0.3 U/ml stimulated a 4-fold increase in the level of cyclin D1 over the levels in nonstimulated cells (Figures 5A and 5B). Fluticasone propionate or dexamethasone each significantly and concentration-dependently inhibited the thrombin response, and completely inhibited increases in cyclin D1 levels at concentrations of 10 nM. Fluticasone propionate was significantly more potent than dexamethasone in reducing cyclin D1 protein levels (Table 1).

View larger version (13K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 5.   (A) Thrombin-stimulated increases in the levels of cyclin D1 (20-h incubation) are inhibited concentration-dependently by dexamethasone (1-100 nM). Representative blot of identical experiments performed in three to five cultures. Cyclin D1 protein appears as a single band at ~ 36 kD. (B) Grouped data for cyclin D1 levels calculated from densitometry measurements and expressed as mean percentages ± SE of the thrombin-stimulated response. The cells were stimulated with thrombin (0.3 U/ml) in the absence and presence of fluticasone propionate (0.1-10 nM, n = 5, filled squares), or dexamethasone (1-100 nM, n = 3, filled circles). Stars indicate P < 0.05 versus the control thrombin response.

The serum-free-medium growth supplement Monomed A stimulates a low level of DNA synthesis, which is inhibited by the glucocorticoids we examined (data not shown). This low level of DNA synthesis may be linked to the basal levels of cyclin D1 that we detected, which were also reduced by fluticasone propionate and dexamethasone (Figure 5A; grouped data not shown).

Thrombin-Stimulated ERK Activation, Increases in Cyclin D1, and Increases in Incorporation of [³H]thymidine Are Inhibited by PD98059, an MEK1 Inhibitor

The importance of the ERK pathway in thrombin-stimulated mitogenesis was investigated with PD98059 (49, 50). ERK activity was assayed in ASM cell lysates prepared after 5 min, 2 h, or 12 h of incubation with thrombin (Table 2), corresponding to time points in the early- to mid-G₁ phase at which ERK activity is reported to be still necessary for mitogens to stimulate proliferation (14, 15). Basal ERK activity was highest in the 5 min incubation lysates (e.g., ~ 150-250 pmol/min/mg protein at 5 min stimulation, versus ~ 30-80 pmol/min/mg protein after 12 h incubation), reflecting stimulation of ERK activity by the growth supplement Monomed A. The approximately 3-fold stimulation by thrombin of basal ERK activity was attenuated after a 5-min incubation, and was prevented after a 12 h incubation with PD98059 10 µM (Table 2). Moreover, the phosphorylation of ERK1 and ERK2 in ASM cell lysates, as measured by Western blotting (12 h incubation), was completely inhibited by 10 µM PD98059 (Figure 6 and Table 2). This concentration of PD98059 was therefore considered suitable for investigating the role of MEK1 after thrombin stimulation. PD98059 had no significant effect on the basal incorporation of [³H]thymidine, but reduced thrombin-stimulated [³H]thymidine incorporation by 40% (Figure 7A). PD98059 also inhibited both the basal and the thrombin (0.3 U/ml)-stimulated 2.5-fold increase in levels of cyclin D1 (Figure 7B).

View larger version (17K): [in this window] [in a new window]   Figure 6.   Effect of dexamethasone (Dex, 100 nM) and PD98059 (10 µM) on thrombin-stimulated phosphorylation of ERKs. Cells were stimulated for 12 h and Western blotting was performed on the cell lysates for the phosphorylated forms of ERK1 and ERK2. The results are representative of those obtained in four experiments with four different cultures.

View larger version (13K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 7.   Effect of the MEK1 inhibitor PD98059 (10 µM, 30-min pretreatment) on thrombin-stimulated increases in DNA synthesis (A) and cyclin D1 levels (B). Cells were prepared as described in Figure 1a, and were stimulated for 24-28 h for [3H]thymidine incorporation analysis (n = 4) and for 20 h for Western blotting of cyclin D1 levels (n = 3). Results are expressed as mean percentages ± SE of the control thrombin response (single star indicates P < 0.05 versus basal; double stars indicate P < 0.01 versus thrombin response).

Effect of Glucocorticoids on Stimulation of ERK Activity by Thrombin

The inhibition of ERK activation by PD98059, and the subsequent reduction in cyclin D1 levels and DNA synthesis, raised the possibility that the glucocorticoids we examined reduced cyclin D1 levels by inhibiting ERK. Incubation of ASM with 100 nM dexamethasone, a concentration that completely inhibits cell proliferation, had no effect on thrombin-stimulated ERK activity (Table 2) or phosphorylation (Figure 6). Furthermore, experiments using the immune-complex kinase assay in which ERK was immunoprecipitated from the crude lysate before assay for MBP kinase activity confirmed that glucocorticoids do not inhibit the thrombin-stimulated increase in ERK activity at 5 min (thrombin + dexamethasone 100 nM, 105% ERK activity versus thrombin control, n = 2), 30 min (thrombin + dexamethasone 100 nM 112 ± 10% ERK activity versus thrombin control, n = 3), or 2 h incubation (thrombin + dexamethasone 100 nM, 72 ± 18% ERK activity versus thrombin control, n = 3). Dexamethasone had no effect on ERK activity in unstimulated cells at any of the incubation time points (110%, 82 ± 32%, and 110 ± 53% ERK activity for the 5 min, 30 min, and 2 h incubations, respectively).

Effect of Glucocorticoids on Cyclin D1 mRNA Levels

Dexamethasone reduced both the basal and thrombin-stimulated increases in cyclin D1 mRNA levels during a 16 h incubation (100 nM) (Figure 8). Although -actin mRNA levels were also regulated by thrombin, the amount of -actin detected in the presence of thrombin was not affected by dexamethasone (161 ± 15% versus 166 ± 18%, n = 6, as compared with unstimulated ASM cells), thus validating the comparisons of cyclin D1 mRNA levels. Similarly, dexamethasone had no effect on the -actin level in untreated ASM cells (122 ± 11% as compared with control 100%, n = 6).

View larger version (23K): [in this window] [in a new window]   Figure 8.   Inhibition by dexamethasone (100 nM) of thrombin-stimulated cyclin D1 mRNA levels. Cells were grown to confluence, serum-deprived for 24 h, and then stimulated for 16 h with thrombin at 0.3 U/ ml. Poly A+ mRNA was subjected to Northern blotting and probed for cyclin D1. The results are representative of those obtained in four experiments with four different cultures.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The glucocorticoids dexamethasone and fluticasone propionate inhibit the proliferation of human ASM cells in culture by causing a noncytotoxic, nonapoptotic arrest of cells in the G₁ phase of the cell cycle. The inhaled glucocorticoid fluticasone propionate has at least a 10-fold greater inhibitory potency than dexamethasone, which is in accord with reports of its greater antiinflammatory potency (44, 45). Thrombin-stimulated phosphorylation of the restriction protein pRb, a key step in passage of cells through the restriction point, was attenuated by dexamethasone. This finding is compatible with the marked inhibition by the glucocorticoids in our study of the thrombin-stimulated increase in cyclin D1 protein and mRNA levels. In contrast, dexamethasone had no effect on ERK activation, despite the clear implication of this enzyme cascade in ASM cell proliferation via regulation of cyclin D1 levels in experiments using the MEK1 inhibitor PD98059.

In this study, we have shown for the first time that glucocorticoids reduce mitogen-stimulated cyclin D1 mRNA levels. In a murine lymphoid cell line, dexamethasone is reported to destabilize cyclin D3 mRNA (59), possibly by upregulating ribonucleases (60). In contrast to this observation in lymphoid cells, the expression of cyclins D1, D2, or D3 in the L929 murine fibroblast cell line was unaffected by dexamethasone (37, 38, 61). Similarly, in a rat mammary tumor cell line grown in dexamethasone, withdrawal of this glucocorticoid resulted in the reexpression of c-myc, and, after a delay, of cyclin D1 (39). However, the latter study did not evaluate the effects of glucocorticoids on cyclin D1 expression in quiescent cells subjected to mitogen stimulation.

The IC₅₀ values for inhibition of thrombin-stimulated levels of cyclin D1 protein by fluticasone propionate and dexamethasone were 0.6 and 2.3 nM, respectively. The value for dexamethasone agrees with the IC₅₀ values of 1.40 nM for inhibition of thrombin-stimulated DNA synthesis, whereas fluticasone propionate was more potent as an inhibitor of DNA synthesis (IC₅₀ = 0.04 nM). The potencies of both glucocorticoids show agreement with those for inhibition of thrombin-induced increases in cell number (dexamethasone = 7.9 nM; fluticasone propionate = 0.2 nM). Although the reason for the greater potency of fluticasone propionate in inhibiting DNA synthesis remains unclear, the glucocorticoid-glucocorticoid receptor complex formed by fluticasone propionate (t_1/2 = 10 h) has a half-life 10 times longer than that of dexamethasone (t_1/2 = 1 h) (62).

Comparison of IC₅₀ values in the current study with glucocorticoid inhibition of interleukin (IL)-5-mediated eosinophil viability (fluticasone propionate = 1.3 nM, dexamethasone = 94 nM) (63), or with glucocorticoid inhibition of E-selectin promoter induction by IL-1 (fluticasone propionate = 0.13 nM, dexamethasone = 2.7 nM) (64) gives results consistent with our observation that fluticasone propionate has greater potency than dexamethasone. Furthermore, fluticasone propionate has an absolute binding affinity (K_D) of approximately 0.5 nM for the human glucocorticoid receptor, which is 14- to 18-fold higher than that of dexamethasone (62). These findings indicate that both the affinity and receptor-ligand complex stability of fluticasone propionate differ from those of dexamethasone, and that both of these factors may influence the relative potencies of these glucocorticoids in the different assays cited previously.

The inhibitory effect of the glucocorticoids in our study on cyclin D1 protein levels in human ASM cells at concentration ranges identical to those that inhibited cell proliferation is consistent with the suggested key role for increases in cyclin D1 levels in mitogenic signaling in bovine tracheal smooth muscle cells (23). The MEK1 activation inhibitor PD 98059 is widely used to investigate the role of the ERK pathway in mitogen signaling in smooth muscle cells (12, 65), and several studies have failed to detect any nonspecific effects of this inhibitor on related protein kinases (49, 50). The foregoing studies used concentrations of PD98059 in the range of 3-50 µM. We chose to use PD98059 at 10 µM, on the basis of its being the lowest concentration that provided almost complete inhibition of thrombin-stimulated ERK1/2 phosphorylation and activity, and of PD98059 being reported to retain specificity for MEK1 at this concentration (49, 50). Inhibition by PD98059 of thrombin-induced increases in cyclin D1 protein levels suggests that MEK1 and ERK are upstream regulators of cyclin D1 protein levels in human ASM cells, in accord with conclusions drawn from experiments using a hamster fibroblast cell line, CCL39 (17). The importance of ERK activation for the signaling of cell progression to the S phase was further confirmed by the inhibitory effect of PD98059 on thrombin-stimulated DNA synthesis.

The foregoing observations raised the possibility that glucocorticoids could act by inhibiting ERK, either directly or through upstream kinases. Nevertheless, the lack of effect of dexamethasone on either activity levels or phosphorylation of ERK, as assessed at several time points and with three different methods, indicates that glucocorticoids do not act by inhibiting ERK activity. Furthermore, a separate study of the time-course of the inhibitory effect of the MEK1 inhibitor PD98059 indicated that mitogen-stimulated entry of ASM cells into the S phase became independent of the MEK1/ERK pathway at 4 to 8 h after initial mitogen exposure (Fernandes and Stewart, unpublished observations). In contrast, glucocorticoid addition as late as 18 h after thrombin addition prevents entry into the S phase (33), indicating that glucocorticoids are still active at times in G₁ at which progression appears to be independent of the MEK1/ERK cascade. Collectively, these observations indicate that glucocorticoids act independently of ERK regulation to inhibit thrombin-stimulated entry of human ASM cells into the S phase of the cell cycle.

The foregoing findings are particularly significant in view of the paucity of data on the effect of glucocorticoids on the ERK pathway. Exposure of Swiss 3T3 fibroblasts to dexamethasone (100 nM) for 48 h inhibited the stimulation of ERK by IGF (34). However, during this prolonged incubation, dexamethasone decreased the levels of insulin receptor substrate, which is upstream of ERK, indicating that the inhibitory effect on ERK may have been indirect. In a rat basophilic leukemia cell line (RBL2H3), concentrations of dexamethasone as low as 10 nM added 6 h beforehand suppressed the stimulation by antigen or serum of phosphorylation of Raf1, MEK1, and ERK2, as well as supressing increases in ERK activity, without affecting IgE receptor levels or early tyrosine kinase activation (35). These findings suggest that an early step in the signaling pathway, preceding Raf1, was the target for inhibition by glucocorticoids in the RBL2H3 cells. However, glucocorticoids do not regulate ERK activity or ERK levels in macrophages stimulated by lipopolysaccharide (66), or in EGF- or phorbol myristate acetate-stimulated 3T3-F442A fibroblasts (36). The absence of an effect on ERK activity in human ASM cells indicates that the actions of glucocorticoids are cell-type specific, and may depend on the involvement of upstream components of the signaling pathway.

Pretreatment with the glucocorticoid receptor antagonist RU 486 (54) prevented glucocorticoid-induced arrest of ASM cell proliferation at the G₁ phase of the cell cycle, indicating a role for glucocorticoid receptor activation in this phase. The point in the mitogenic signaling pathway between ERK activation and cyclin D1 expression at which activated glucocorticoid receptors inhibit passage is not known; however, the inhibitory effects of glucocorticoids on the transcription factors activation protein (AP-1) and nuclear factor-kappa B (NF-B) are well documented (28, 67). AP-1 is formed by the hetero- or homodimerization of the protooncogene proteins c-fos and c-jun, a process that may occur after phosphorylation of c-jun by the MAPK family (68), but is preferentially mediated by stress-activated protein kinase (69). The formation of active AP-1 complexes is a necessary step for mitogenic signaling in many cell types (70, 71); the formation of such complexes may be necessary for cyclin D1 promoter activation (19, 72), and could provide the link between the kinase-driven G₀-to-G₁-phase transition and the cyclin/ pRb-driven G₁-to-S-phase transition of the cell cycle (40). Activated glucocorticoid receptors are known to form a protein-protein complex with AP-1 (73, 74) and thereby prevent its action on genes required for cell progression to the S phase. Thus, the role of AP-1 interactions in the inhibition of pRb phosphorylation warrants further investigation.

There are many potential explanations for the reduced level of pRb phosphorylation in the presence of dexamethasone, since an intricate series of steps lead to the generation of the active cyclin D1/cdk4 complexes that phosphorylate pRb (20, 24, 26, 57, 58). We have not explored all these possible explanations, since the inhibition by glucocorticoids of cyclin D1 protein levels appeared to provide an adequate explanation for the G₁-phase arrest of ASM cells. Nevertheless, upregulation of the levels of the cyclin-dependent kinase inhibitors (cdki) represents an alternative or additional mechanism by which glucocorticoids could modulate pRb phosphorylation. In murine L929 fibroblasts, glucocorticoids inhibited signaling of cell proliferation by stimulating the production of p21^Cip1, a cdki that binds to and prevents the activation of cyclin D/cdk4 complexes (37).

In the P1798 murine lymphoid cell line, in which cyclins D1 and D2 are not major contributors to proliferation (59), dexamethasone decreased the expression of cyclin D3 and the cyclin D-associated kinase cdk4, as well as of c-myc, another transcription factor whose levels are increased during G₁ (41, 75). Because c-myc has been implicated in signaling upstream of cyclin D1 expression (40), it represents a potential site of regulation by glucocorticoids of cyclin D1 expression. However, the overexpression of cyclin D3, cdk4, or c-myc individually in P1798 cells was not sufficient to prevent the growth-arresting effect of dexamethasone, whereas the combined expression vectors for cyclin D3 and c-myc surmounted the growth arrest (75). These findings suggest that in this cell type there is some independence of the effects of c-myc and cyclin D (75). Thus, it is clear that dexamethasone may act on multiple sites in the cell cycle to produce growth arrest. Irrespective of the contributions of these various mechanisms, a reduction in the phosphorylation of pRb is sufficient to explain the glucocorticoid-induced arrest of ASM cells in the G₁ phase of the cell cycle, given the well-established importance of pRb phosphorylation in the passage of normal mammalian cells through the restriction point of the G₁ phase (20, 24, 57, 58). Glucocorticoids appear to have no effect on the levels of another cyclin involved in pRb phosphorylation during the G₁-to-S-phase transition, cyclin E (K.-S. Lee and A. G. Stewart, unpublished observations).

The use of glucocorticoids as antiasthma agents results in relatively acute (days or weeks) antiinflammatory action in the airway (28), yet there is little insight about why the maximum benefits of these agents in asthma occur only after prolonged administration (months or years) (27, 76). Similarly, there is no explanation for the finding that the use of inhaled glucocorticoids early in the course of asthma produces a better medium-to-long term outcome than their delayed use in this disease (77, 78). Recent attention has turned to the morphologic changes occurring in asthma as a potential explanation for these observations (9, 32). Subepithelial collagen deposition is reduced in patients with mild asthma after a period of inhaled fluticasone propionate treatment as short as 3-6 wk, and is associated with diminished AHR (79). Remodeling of ASM may be less readily or rapidly reversible, but bronchial biopsy specimens are too shallow to provide quantifiable information on such changes. Airway wall remodelling has a marked influence on AHR, is a product of chronic inflammation in the airway, and if attenuated by glucocorticoids may well explain the greater benefit of prolonged treatment and the need for early administration of steroids in antiasthma treatment regimens to optimize outcome. Interestingly, patients with glucocorticoid-resistant asthma have an abnormal glucocorticoid receptor-AP-1 interaction (80), which is hypothesized to affect a full range of the antiinflammatory actions of steroids in airway cells. Impairment of the glucocorticoid receptor-AP-1 interaction, and therefore a reduction in the antimitogenic action of glucocorticoids on ASM, may contribute to the gradual worsening of obstruction in these patients.

In conclusion, our results show that glucocorticoids induce a G₁-phase block in cell cycle progression in cultured human ASM cells, preventing phosphorylation of the restriction protein pRb. This inhibition may be explained by a reduction in cyclin D1 protein and mRNA levels, through an action at a site that is either downstream of or parallel to the ERK cascade.