Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The biologic action of glucocorticoids is mediated through the activation of intracellular glucocorticoid receptors (GR) (1, 2). Two human isoforms of GR have been identified, termed hGR and hGR, which originate from the same gene by alternative splicing of the hGR primary transcript (Figure 1) (3). hGR is the predominant isoform of the receptor and the one that shows steroid-binding activity (4). This isoform is expressed in most human tissues and cell lines at varying levels. Using hGR-specific antibodies, Oakley and colleagues (6) recently detected the expression of the hGR protein in different cell types. In the absence of ligand, hGR resides primarily in the cytoplasm of cells and is held inactive by its binding to heat-shock proteins. Upon hormone binding, hGR is phosphorylated, dissociated from heat-shock proteins, and subsequently translocated to the cell nucleus, where it binds as a homodimer to specific sequences within the promoter region of target genes, leading to enhancement or repression of their transcription (1, 2). In addition to this direct interaction of hGR on the DNA, most of the anti-inflammatory responses of glucocorticoids are known to be mediated through a protein-protein interaction between the hGR and transcription factors, such as activating protein-1 and nuclear factor-B (2).

View larger version (17K): [in this window] [in a new window]   Figure 1.   Predicted structure of the hGR gene and transcription products. The human GR gene has been mapped within chromosome 5. Alternative processing of exon 9 of the hGR primary transcript generates multiple hGR messages (5). The default splicing pathway generates the 7.0- and 5.5-kb hGR messages, whereas the alternative splicing pathway generates the 4.3-kb hGR message. Exons are represented by boxes and introns by lines. Arrowheads represent the primers used in this study.

Contrary to hGR, much less is known about the hGR isoform. Expression of hGR messenger RNA (mRNA) has been detected in various human tissues and cell lines (5, 7). With the recent development of hGR-specific antibodies, positive immunoreactivity for hGR has been detected in different tissues and cell types (8, 11). However, reports analyzing the abundance of hGR relative to hGR are conflicting, underlying the necessity of developing accurate strategies to measure the expression levels of both receptor isoforms. hGR does not show either ligand-binding activity (4, 5) or transactivation of target genes (5, 8). Using transfection studies, however, it has been reported that hGR can inhibit hGR-mediated stimulation of gene expression, acting as a dominant negative inhibitor of hGR activity (5, 7, 12). Although the physiologic significance of hGR is still unknown, recent studies have reported increased hGR immunoreactivity in peripheral blood mononuclear cells and bronchoalveolar lavage cells from patients with glucocorticoid-insensitive asthma (13).

Target tissue sensitivity to glucocorticoids has been demonstrated to parallel changes in receptor levels (14). Glucocorticoids have been shown to downregulate the expression of their own receptor in a variety of tissues and cell lines. Downregulation has been shown to occur at the transcriptional, post-transcriptional, and post-translational levels, depending on the cell type (15). Little information is available on the effect of glucocorticoids on hGR expression in the airways epithelium (10, 18). In addition, to date, most of the reported data concerning the autoregulation of hGR expression by its ligand did not distinguish between hGR and hGR isoforms. Korn and associates (10) reported hormone-induced downregulation of hGR and hGR mRNA expression in a human bronchial epithelial cell line. However, given the contradictory results shown in the literature when analyzing the message and/or protein expression of the hGR isoform, the regulation of hGR and hGR expression levels by ligand, as well as the molecular mechanisms involved, require further studies. Because downregulation of hGR levels after treatment with glucocorticoids may be one of the possible explanations to the secondary glucocorticoid resistance phenomenon (2, 16), the regulation of hGR by hormone is an interesting subject of study. In the present study we examined the expression and regulation by dexamethasone (DEX) of hGR and hGR mRNA isoforms in a human bronchial epithelial cell line (BEAS-2B), as well as the expression of hGR and hGR proteins in BEAS-2B cells, A549 cells, and human nasal primary epithelial cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

RPMI 1640 and Dulbecco's modified Eagle's medium (DMEM), penicillin-streptomycin, L-glutamine, N-2-hydroxyethylpiperazine- N'-ethane sulfonic acid (Hepes) buffer, insuline, transferrine, diethylpyrocarbonate-treated water, random hexanucleotide primers, MgCl₂, deoxynucleotide triphosphates (dNTPs), dithiothreitol (DTT), Moloney murine leukemia virus (MMLV) reverse transcriptase, ribonuclease (RNase) inhibitor, Taq DNA polymerase, Taq Platinum DNA polymerase, and the reverse transcriptase/ polymerase chain reaction (RT-PCR) buffers were obtained from Life Technologies SA (Barcelona, Spain). The PCR primers and the protease inhibitor cocktail tablet were purchased from Boehringer Mannheim (Barcelona, Spain). Epidermal growth factor and type 1 rat-tail collagen were purchased from Upstate Biotechnology (Lake Placid, NY), and fetal calf serum (FCS) and NU-Serum from Biological Industries (Beit Haemek, Israel). Tissue culture flasks and plates were obtained from TPP (Trasadingen, Switzerland), DEX from MSD (Barcelona, Spain), and TRI-Reagent from MRC (Cincinnati, OH). The peroxidase-labeled goat antirabbit secondary antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), the chemiluminescent reagent from Pierce (Rockford, IL), and all other reagents from Sigma Chemical Co. (Madrid, Spain).

Cell Culture and Transfections

BEAS-2B cells, a human bronchial epithelial cell line, were cultured as previously reported (19), with slight modifications. Briefly, cells were cultured in collagen-coated culture flasks in RPMI 1640 medium containing 100 IU/ml penicillin, 100 µg/ml streptomycin, 2 µg/ml amphotericin B, 1% FCS, 1 mM L-glutamine, 50 nM Na₂-SeO₃, 10 mM Hepes buffer, 1 µg/ml hydrocortisone, 10 µg/ml insuline, 10 µg/ml transferrine, and 10 ng/ml epidermal growth factor.

A549 cells, an epithelial cell line from human lung carcinoma, were cultured as previously reported (20).

Human nasal primary epithelial cells were isolated after protease digestion of nasal mucosa specimens obtained from four patients undergoing nasal corrective surgery. Patients did not receive corticosteroid treatment before surgery. Nasal epithelial cells were seeded onto six-well plates coated with type 1 rat-tail collagen and cultured as previously reported (21).

COS-7 cells were grown in DMEM supplemented with 10% NU-Serum, 100 IU/ml penicillin, and 100 µg/ml streptomycin. Cells were transfected with the hGR expression vector pCMVhGR, which contains the hGR-specific coding sequences and the hGR 3'-untranslated region (UTR), using the calcium phosphate coprecipitation technique as previously described (11).

All cultured cells were maintained in a 5% CO₂ humidified atmosphere at 37°C. BEAS-2B, A549, and COS-7 cells were passaged every 3 to 4 d. Experiments were initiated when cell cultures reached 80 to 90% confluence.

Isolation of Total RNA

Total RNA from BEAS-2B cells was isolated using a rapid extraction method (TRI-Reagent) derived from the method of Chomczynski and Sacchi (22). The integrity of the purified RNA was determined by visualization of the 28S and 18S ribosomal RNA bands after electrophoresis on a 1% agarose denaturing gel stained with ethidium bromide. Total RNA was quantified by densitometric analysis with respect to five known amounts of total RNA loaded in parallel.

RT-PCR

Total RNA (4 µg) from BEAS-2B cells was reverse transcribed to complementary DNA (cDNA) in a buffer containing 50 mM Tris-HCl (pH 8.3); 75 mM KCl; 5 mM MgCl₂; 10 mM DTT; 1.5 µg random hexanucleotide primers; 2 mM each of deoxyadenosine triphosphate, deoxythymidine triphosphate, deoxyguanidine triphosphate, and deoxycytidine triphosphate; 40 units RNase inhibitor; and 300 units MMLV RT, in a final volume of 30 µl. This mixture was incubated for 1 h at 37°C and finally heated for 10 min at 95°C to inactivate the RT. Random hexanucleotide primers were chosen for first-strand synthesis of cDNA because they prime all species of RNA present and allow the amplification of any desired target sequence from the reverse transcription mix. Thus, the resulting cDNAs were used for the amplification by PCR of specific targets: hGR, hGR, and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH). This "simple PCR" amplified specific targets without using any quantitative approach. All PCR reactions were performed before reaching the plateau phase. Primers were designed to span introns, so that any genomic DNA product would be distinguished from the target cDNA by size difference (Table 1). The specific hGR and hGR mRNAs were amplified using specific antisense primers (RGR and RGR, respectively) that shared the same sense primer (FGRc) (Figure 1). For all amplified mRNAs, the PCR reaction buffer contained 3 µl of a 1:16 (GAPDH, hGR) or 1:2 (hGR) dilution of the RT reaction, 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 0.5 mM (GAPDH) or 1.5 mM (hGR, hGR) MgCl₂, 0.2 mM dNTPs, 0.6 µM (GAPDH) or 1.2 µM (hGR, hGR) forward and reverse primer (Table 1), and 1 U Taq DNA polymerase (GAPDH, hGR) or 2 U Taq Platinum DNA polymerase (hGR). The mixture was overlaid with mineral oil and amplified in a thermocycler (PTC-100; MJ Research, Waterton, MA). After an initial incubation for 4 min at 94°C, samples were subjected to 25 (GAPDH), 28 (hGR), or 38 (hGR) cycles of 1 min at 94°C, 1 min at 60°C, and 1.5 min at 72°C (GAPDH, hGR) or 68°C (hGR). A final extension step at 72°C (GAPDH, hGR) or 68°C (hGR) for 7 min was performed. All PCR components used in the reaction were tested for possible contamination of DNA in a simultaneous PCR reaction containing all the reagents except the target cDNA. Amplified cDNA fragments (PCR products) were electrophoretically fractioned on a 1% agarose gel stained with ethidium bromide. DNA bands were visualized after exposure to ultraviolet (UV) light. Sequencing of the hGR and hGR PCR products confirmed their homology with respect to their predicted cDNA sequences (4).

hGR-specific primers amplified a unique hGR band of 824 base pairs (bp) (Figure 2A) which, according to the model of alternative splicing of the hGR primary transcript predicted by Oakley and coworkers (5), is derived from the amplification of the 7.0- and 5.5-kb hGR messages (Figure 1). hGR-specific primers amplified two PCR products: a 1,002-bp band derived from the 4.3-kb hGR message and a 3.64-kb fragment derived from the 7.0-kb hGR message (Figures 1 and 2A). However, the intensity of the latter band was not as strong as would be expected because our PCR amplification conditions did not favor the production of this large fragment. The analysis of hGR and hGR mRNA expression by simple PCR, although not quantitative, already revealed that hGR mRNA expression was much higher than that of hGR. Thus, about 10 more cycles were needed for the PCR of hGR to obtain UV fluorescence intensities similar to the hGR PCR products. GAPDH-specific primers amplified a band of 594 bp (Figure 2A). To ensure that the RNA was effectively reverse transcribed to cDNA for each condition and that the drug treatment by itself did not have any effect on the housekeeping gene GAPDH expression, the GAPDH PCR was routinely performed in each experiment.

View larger version (33K): [in this window] [in a new window]   Figure 2.   Simple PCR and RT-competitive PCR analysis of hGR and hGR mRNAs. Total RNA isolated from BEAS-2B cells was reverse transcribed to cDNA. (A) Simple PCR. cDNA was amplified using either hGR-, hGR-, or GAPDH-specific primers (Table 1). The PCR products were electrophoresed on 1% agarose gels stained with ethidium bromide. No cDNA was added in lane 1. Lanes 2-5 correspond to PCR reactions of four different samples. The sizes (in bp) of the DNA marker and the PCR products are indicated in the left and right margins, respectively. (B) RT-competitive PCR. Four serial dilutions of the competitors (internal standard, GR-S, or GR-S) were added to a constant amount of target cDNA. After amplification, the PCR products were resolved by agarose gel electrophoresis and ethidium bromide staining. The relative amounts of target and GR-S products, which differed in size, were analyzed and compared in each sample. After correction for size differences, the initial amounts of target and competitor (GR-S or GR-S) products were assumed to be equal in the reactions where their molar ratios were judged to be equal (log ratio = 0). Because the amount of GR-S added to each PCR reaction was known, the absolute initial amount of target could be determined. Insets show representative gels of the hGR and hGR PCRs, with the molecular weight standard (lane 1) and four increasing dilutions (lanes 2-5) of GR-S (1,091 bp) or GR-S (1,118 bp), which compete with a constant amount of the target cDNA (hGR, 824 bp; hGR, 1,002 bp).

RT-Competitive PCR

To measure hGR and hGR mRNA expression in BEAS-2B cells we developed a competitive PCR technique in which, after the RT step, known amounts of an exogenous DNA were added to the amplification mixture (RT-competitive PCR) (23). The exogenous molecule, called competitor or internal standard, was coamplified in competition with the target in the same test tube. Because the initial amount of internal standard was known, this technique allowed the determination of the initial amount of target cDNA (Figure 2B).

Construction and cloning of the internal standard. A heterologous DNA fragment, i.e., a 1,037-bp fragment (corresponding to nucleotides 191-1227) of the intercellular adhesion molecule-3 cDNA, was amplified by PCR using two composite primers (Table 2). One composite primer contained the forward common primer to hGR and hGR cDNAs, linked to a 24-mer that annealed to one strand of the heterologous fragment. The other composite primer contained the reverse specific primers for hGR and hGR cDNAs, linked to a 24-mer that annealed to the opposite strand of the heterologous fragment. After 35 cycles of PCR amplification (1 min at 94°C, 1 min at 60°C, and 1.5 min at 72°C) the hGR-specific primer sequences were incorporated into the heterologous fragment. PCR products were then electrophoresed on a gel and the 1,118-bp band, corresponding to the hGR internal standard (GR-S), was excised from the gel, purified by electroelution using the Biotrap Starterkit (Schleicher & Schuell, Dassel, Germany), and precipitated.

The purified GR-S was cloned using the Original TA Cloning kit (Invitrogen, San Diego, CA), ligated into the pCR 2.1 plasmid (Invitrogen), and transformed into competent Escherichia coli INV F' cells, according to the manufacturer's instructions. Minipreparations of plasmid DNA were performed for the selected colonies. To confirm the presence of insert, the isolated plasmids were digested with EcoRI and HindIII and the digested products were resolved on a 1% agarose gel. One clone was selected and the band corresponding to the predicted insert was excised from the gel, purified by electroelution, and precipitated. The quantity of GR-S obtained was determined by running an aliquot of the GR-S on a gel and comparing the intensity of the band to a dilution series of a DNA marker (phage- DNA digested with HindIII) containing known amounts of DNA.

Competitive PCR of hGR and hGR. The primers used in the competitive PCR reaction were FGRc and RGR for the amplification of hGR cDNA, and FGRc and RGR to amplify the hGR cDNA (Table 1). Three-fold serial dilutions of the GR-S, starting with 1.2 × 10⁶ (hGR) or 4 × 10³ (hGR) copies, were loaded into each PCR reaction tube with a constant amount of the target cDNA (Figure 2B). All other components of the mixture, as well as the PCR reaction conditions, were similar to those previously described for the "simple PCR" of hGR and hGR. Because GR-S contains the complementary sequences to hGR and hGR primers, GR-S and the target cDNA competed for the same pair of primers and therefore for amplification. The resulting products of the competitive PCR were electrophoretically fractioned on a 1.3% agarose gel stained with ethidium bromide. DNA bands were visualized after exposure to UV light. Images from electrophoresed gels were captured in a computer-assisted imaging system, and the relative quantities of hGR or hGR to competitors were quantified and compared by densitometric analysis using the software Bio-Profil (Vilber Lourmat, Marne La Vallée, France). To correct the differences in nucleotide number, the density ratio of the target band to the GR-S or GR-S band was multiplied by a correction factor that was 1.32 (1,091/ 824) for hGR and 1.12 (1,118/1,002) for hGR. The logarithm of the corrected ratio was then plotted versus the logarithm of the initial amount of GR-S added in the PCR reaction. At the competition equivalence point (log ratio = 0), the initial concentration of the target corresponded to the initial amount of the added GR-S. Finally, the value obtained at the equivalence point was divided by a factor of 2 because the competitor is double-stranded whereas the target RNA is single-stranded.

Western Blotting

BEAS-2B, A549, COS-7, and human nasal epithelial cells were harvested from subconfluent culture flasks, centrifuged, and washed with cold phosphate-buffered saline (pH 7.4). The final cell pellets were resuspended in 200 to 300 µl of lysate buffer containing one protease inhibitor cocktail table (Complete), 50 mM Hepes buffer, 0.05% Triton X-100, and 0.62 mM phenylmethylsulfonyl fluoride. Cells were then sonicated by two 15-s bursts on ice using a Branson sonifier (Branson, Danbury, CT) and the homogenate was immediately centrifuged at 20,000 × g for 10 min at 4°C. Supernatants were collected and stored at 70°C. The totals of 50 µg of protein from COS-7 cells and 150 µg of protein from other cells were resolved by electrophoresis through 7% Tris-acetate gels and electrophoretically transferred to nitrocellulose following the instruction manual. To check for equal loading and transfer efficiency, membranes were stained with Ponceau S (0.5% in 1% acetic acid) and then blocked overnight at 4°C in a blocking buffer containing 5% nonfat dry milk in Tris-buffered saline (T-TBS) (10 mM Tris-HCl [pH 7.4], 150 mM NaCl, and 0.05% Tween-20). Membranes were then incubated for 2 h at room temperature with 1:2,000 dilution in blocking buffer of the primary antibody. For detection of hGR protein we used the well-characterized antipeptide hGR antibody 57, raised against epitopes common to both receptor isoforms (11). For specific detection of hGR protein, we used an antibody, BShGR, raised against a peptide corresponding to the 15 nonhomologous amino acids of the carboxy terminus of the hGR protein (11). The characterization and specificity of both antibody 57 and antibody BShGR have been extensively studied (11). Blots were then washed in T-TBS and subsequently incubated for 2 h at room temperature with a horseradish peroxidase-labeled goat antirabbit secondary antibody diluted 1:40,000 in blocking buffer. After washing with T-TBS, blots were incubated with an enhanced chemiluminiscent substrate and then exposed to films. Autoradiograms were analyzed by densitometry.

Statistical Data Analysis

Basal expression of hGR or hGR mRNA is expressed as the arithmetic mean ± standard error of the mean (SEM) of 10⁶ copies of hGR cDNA or 10³ copies of hGR cDNA per microgram of total RNA. Results from DEX experiments are expressed as arithmetic means ± SEM of percentage of control (media-treated cells). Statistical comparisons were performed using analysis of variance (ANOVA) with the Dunnett t test comparisons in time-course experiments, and a nonparametric test (Wilcoxon signed rank) in all other experiments. P < 0.05 was regarded as statistically significant (24).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

BEAS-2B Cells

At 24 h before initiating cell treatment and during cell treatment, media were replaced by supplemented media without hydrocortisone. Time-course studies (1, 6, and 24 h) with or without DEX (10⁶ M) were performed to analyze hGR and hGR mRNA and protein expression. Because the maximal effects on mRNA and protein expression were found at 6 and 24 h, respectively, these incubation times were selected for the DEX dose-response studies. For the analysis of hGR and hGR mRNA expression, BEAS-2B cells were treated with DEX (from 10⁹ to 10⁵ M) for 6 h; whereas for the analysis of hGR and hGR proteins, cells were incubated with DEX (10¹⁰, 10⁸, and 10⁶ M) for 24 h.

Expression and regulation of hGR and hGR mRNAs. Basal expression (0 h) of hGR and hGR mRNAs were 2.8 ± 0.5 × 10⁶ copies and 1 ± 0.1 × 10³ copies, respectively (n = 9). Expression of either hGR or hGR mRNA at 1, 6, and 24 h was not significantly different from basal expression levels.

DEX at 10⁶ M caused downregulation of hGR mRNA levels (n = 9) at 6 h (54.8 ± 8.3% of control; P < 0.01) and 24 h (57.6 ± 4.9% of control; P < 0.01) (Figure 3A, left graph). At 6 h, DEX caused a dose-related downregulation of hGR mRNA whose maximal effect was reached at 10⁶ M (54.8 ± 2.1% of control; n = 6; P < 0.05) (Figure 3B, left graph).

View larger version (42K): [in this window] [in a new window]   Figure 3.   Effect of DEX on hGR and hGR mRNA expression in BEAS-2B cells. Cultured cells were incubated in either the absence (dashed lines) or presence ( filled symbols) of 106 M DEX for 1, 6, and 24 h (n = 9) (A), or in the absence or presence of 109 M to 105 M DEX for 6 h (n = 6) (B). Total cellular RNA was isolated and reverse transcribed to cDNA. hGR and hGR mRNAs were amplified by competitive PCR using hGR- or hGR-specific primers. The PCR products were electrophoresed on agarose gels and the relative amounts of hGR or hGR to competitors were compared by densitometric analysis. ANOVA with the Dunnett t test was used for the time-course studies and the Wilcoxon test for the dose-response studies (*P < 0.05 compared with media-treated cells).

DEX 10⁶ M also caused downregulation of hGR mRNA levels at 6 h (54.8 ± 7.1% of control; n = 9; P < 0.01). After 24 h, hGR mRNA levels in DEX-treated cells were similar to those of media-treated cells (Figure 3A, right graph). Similarly to hGR and at 6 h of incubation, DEX provoked a dose-related downregulation of hGR mRNA expression whose maximal effect was reached at 10⁶ M (33.7 ± 8.6% of control; n = 5; P < 0.05) (Figure 3B, right graph).

Expression and regulation of hGR and hGR proteins. Western blotting analysis of hGR using antibody 57 detected, as previously reported (11), the 94-kD hGR protein in all samples examined. DEX at 10⁶ M (n = 3) downregulated hGR protein levels at 6 h (69.4 ± 11.5% of control; P < 0.05) and 24 h (16 ± 4.5% of control; P < 0.01) (Figure 4A). At 24 h, DEX caused a dose-related downregulation of hGR protein whose maximal effect was reached at 10⁶ M (19.5 ± 8.7% of control; n = 5; P < 0.01) (Figure 4B).

View larger version (38K): [in this window] [in a new window]   Figure 4.   Effect of DEX on hGR protein expression in BEAS-2B cells. Cultured cells were incubated either in the absence (, dashed line) or presence (+, open symbols) of 106 M DEX for 1, 6, and 24 h (n = 3) (A), or in the absence or presence of 1010 M to 106 M DEX for 24 h (n = 5) (B). Total cellular proteins were extracted and analyzed by Western blotting with antibody 57. Insets show representative time-course (left) and dose-response (right) experiments. ANOVA with the Dunnett t test was used for the time-course studies and the Wilcoxon test for the dose- response studies (*P < 0.05, **P < 0.01, compared with media-treated cells).

The use of BShGR antibody efficiently detected the 90-kD hGR protein in COS-7 cells transfected with the pCMVhGR expression vector. This 90-kD protein was confirmed to be hGR because it disappeared with BShGR preabsorbed with the peptide antigen (11). However, Western blotting analysis of hGR using BShGR antibody failed to detect hGR protein in BEAS-2B cells (data not shown).

Effect of cycloheximide and actinomycin-D on hGR and hGR mRNA expression. To investigate whether DEX-induced downregulation of hGR and hGR mRNA expression was dependent on de novo protein synthesis, we examined the effect of DEX in the presence of the protein synthesis inhibitor cycloheximide (CHX). BEAS-2B cells were preincubated with CHX (10 µg/ml) for 1 h before the addition of DEX (10⁶ M) or culture media for 6 h (Figure 5A). CHX alone (n = 6) resulted in a superinduction of both hGR (6-fold) and hGR (2.5-fold) mRNA expression, compared with non-CHX-treated cells. Under the same conditions, CHX had no effect on the expression of GAPDH mRNA (data not shown). Interestingly, even in the presence of CHX, DEX downregulated hGR mRNA levels (73 ± 8.5% of CHX-treated cells; P < 0.05) (Figure 5A, left graph) and hGR mRNA levels (76 ± 8.8% of CHX-treated cells; P < 0.05) (Figure 5A, right graph). This observation suggests that new protein synthesis is not needed for DEX-induced downregulation of hGR gene expression.

View larger version (31K): [in this window] [in a new window]   Figure 5.   Effects of CHX and Act-D on hGR and hGR mRNA expression in BEAS-2B cells. (A) Cultured cells were preincubated with CHX (10 µg/ml) for 1 h, and then incubated with culture media or 106 M DEX for 6 h. RT-competitive PCR for hGR (left graph) and hGR (right graph) was performed and the PCR products were analyzed by densitometry. Wilcoxon test; *P < 0.05, compared with media-treated cells, C; #P < 0.05, compared with CHX treated cells. (B) BEAS-2B cells were cultured with Act-D (5 µg/ml) for 30 min. Thereafter, 106 M DEX ( filled symbols) or media alone (open symbols) was added. RT- competitive PCRs for hGR (left graph) and hGR (right graph) were performed and their products analyzed by densitometry. ANOVA with the Dunnett t test; *P < 0.05 compared with media-treated cells.

To investigate whether glucocorticoids downregulate hGR and/or hGR mRNA steady-state levels by decreasing mRNA half-life, studies using the transcription inhibitor actinomycin-D (Act-D) were carried out. BEAS-2B cells were pretreated with Act-D (5 µg/ml) for 30 min. Culture media or 10⁶ M DEX were then added to the cells and the decay of both hGR (Figure 5B, left graph) and hGR mRNA (Figure 5B, right graph) levels was analyzed after 3, 6, and 12 h of incubation with Act-D (n = 6). Interestingly, DEX treatment resulted in the stabilization of hGR mRNA which was statistically significant at 12 h. The hGR mRNA half-life was 3.5 h in the absence of DEX and increased to 6 h in its presence. Although DEX induced a mild but significant stabilization of hGR mRNA after 12 h of incubation, the hGR mRNA half-life was almost not modified by DEX (3.5 h) compared with controls (3 h). Under the same conditions, DEX had no effect on the stability of GAPDH mRNA (data not shown).

A549 Epithelial Cells

Time-course studies (1, 6, and 24 h) with or without DEX (10⁶ M) were carried out to analyze hGR and hGR protein expression. Western blotting analysis of hGR using antibody 57 detected the 94-kD hGR protein in all examined samples. DEX (10⁶ M, n = 3) decreased hGR protein levels at 6 h (61.1 ± 7.4% of control; P < 0.05) and 24 h (14.5 ± 4.1% of control; P < 0.01) (Figure 6). Using the BShGR antibody no expression of hGR protein was detected in A549 cells (data not shown).

View larger version (30K): [in this window] [in a new window]   Figure 6.   Effect of DEX on hGR protein expression in A549 cells. Cultured cells were incubated in the absence (, dashed line) or presence (+, open symbols) of 106 M DEX for 1, 6, and 24 h (n = 3). Total cellular proteins were extracted and analyzed by Western blotting using antibody 57. Inset shows a representative experiment. ANOVA with the Dunnett t test; (*P < 0.05, **P < 0.01, compared with media-treated cells).

Human Nasal Primary Epithelial Cells

A limited number of epithelial cells (5 to 10 × 10⁶ cells) were obtained from the nasal mucosa specimens. Because maximal downregulation of hGR protein in BEAS-2B and A549 cells was reached after 24 h of incubation with DEX, and Western blotting analysis of hGR and hGR proteins required high amounts of cell protein extracts, cultured nasal epithelial cells were incubated with or without 10⁶ M DEX for only 24 h. Antibody 57 detected the 94-kD hGR protein in all examined samples. DEX (10⁶ M) decreased hGR protein expression at 24 h (28 ± 7.4% of control; n = 4; P < 0.01) (Figure 7). However, no hGR protein expression was detected in nasal epithelial cells (data not shown).

View larger version (31K): [in this window] [in a new window]   Figure 7.   Effect of DEX on hGR protein expression in human nasal primary epithelial cells. Cultured cells were incubated in the absence () or presence (+) of 106 M DEX for 24 h. Total cellular proteins were extracted and analyzed by Western blotting using antibody 57. Inset is a representative experiment showing the 94-kD hGR protein in 150 µg of total protein of either BEAS-2B cells or primary epithelial cells from four nasal mucosa specimens (NM1-NM4). ANOVA with the Dunnett t test; **P < 0.01 compared with media-treated cells.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Few studies have analyzed the expression of hGR- and hGR-specific isoforms (5, 6, 8). In this report, we studied the expression and regulation of hGR and hGR isoforms in human respiratory epithelial cells. By RT-PCR analysis we confirmed that hGR and hGR transcripts were coexpressed in the human bronchial epithelial cell line BEAS-2B. Detection of hGR- and hGR-specific products with the predicted size confirmed the proposed alternative splicing model of the hGR gene (5). We found that hGR mRNA was about 2,800-fold more expressed than hGR mRNA. Oakley and colleagues (5) reported that hGR mRNA was 200- to 500-fold more expressed than hGR mRNA in two adult tissues and two human cell lines. Although both studies reflect a large difference between hGR and hGR expression levels, we should take into account that hGR amplification products are derived from two hGR messages, whereas hGR amplification products come only from one hGR message (Figure 1). Cell-specific differences or differences in the technique used may account for the apparent discrepancy in the hGR/hGR ratio between our results and those of Oakley and associates (5). Thus, whereas Oakley and coworkers (5) used external standards to quantify the hGR transcripts, we used an exogenous internal standard that was the same for the amplification of hGR and hGR. Another advantage in using internal standards is that the problem of tube-to-tube variation in amplification efficiency is avoided. Dahia and colleagues (9) quantified the expression of both isoforms of hGR mRNA in corticotropin-secreting tumors by means of RT-PCR followed by Southern blotting of hGR and hGR PCR products, demonstrating that the expression of hGR was at a much lower level than hGR. Northern blot analysis using hGR- and hGR-specific probes has also revealed that both hGR messages (7.0 and 5.5 kb) are much more abundant than the hGR transcript (4.3 kb) (5). Whatever the technique used when examining hGR and hGR mRNA expression, special attention should be taken in the design of probes and primers. Probes or primers that do not discriminate between hGR and hGR may not selectively differentiate hGR and hGR messages, thus leading to discrepant results. The low levels of hGR mRNA that we and others have found are consistent with the results obtained in the Western blotting experiments. Although hGR protein was detected in BEAS-2B and A549 cell lines as well as in nasal primary epithelial cells, no expression of hGR was found in these cells. In line with our results, recent reports indicate that in most cells hGR protein is much less abundant than hGR protein (8, 11, 13).

Downregulation of GR expression by different glucocorticoids has been described both in vitro (10, 25) and in vivo (17, 18, 25, 32, 33). However, most of the reports analyzing the regulation of hGR expression by ligand do not distinguish between hGR and hGR isoforms. In the present study we report a dose-dependent downregulation of both hGR and hGR mRNA levels by DEX in BEAS-2B cells. This pattern of downregulation is similar to that reported by Korn and associates (10) in another human bronchial epithelial cell line (BET-1A) using the glucocorticoid budesonide. It is worth noting that hGR does not show ligand-binding activity (4, 5); therefore, downregulation of hGR mRNA is likely to be mediated through binding of the glucocorticoid to the hGR isoform (12).

To further characterize the mechanism of homologous downregulation of hGR and hGR mRNAs in BEAS-2B cells, we studied their regulation after inhibiting translation with CHX. CHX alone caused 6- and 2.5-fold inductions of hGR and hGR mRNA levels, respectively. Superinduction of mRNA expression by CHX has previously been reported not only for the hGR message (25, 29, 30) but also for many mammalian-cell mRNAs (34). CHX-mediated stabilization of hGR and hGR mRNAs could be explained in two different ways: (1) the degradation of hGR and hGR mRNAs is linked to their ongoing translation to protein, thus when translation is inhibited the message is stabilized; and (2) a labile transacting factor is normally required to degrade hGR and hGR mRNAs and is destroyed or inactivated after treatment with CHX. Importantly, DEX still induced downregulation of hGR and hGR mRNAs in the presence of CHX, indicating agreement with previous reports (25, 29) that downregulation does not require ongoing protein synthesis.

We used the transcription inhibitor Act-D to assess whether DEX caused downregulation of hGR and hGR mRNAs through destabilization of their messages. In the absence of hormone, hGR mRNA half-life (3.5 h) was similar to that of hGR mRNA (3 h). Previous studies analyzing GR mRNA stability have reported half-lives ranging from 2 to 5 h, depending on the cell type (26, 28, 30, 33). In our study, DEX stabilized both hGR mRNA and more importantly hGR mRNA, almost doubling the hGR mRNA half-life up to 6 h, but with little effectif anyon the hGR mRNA half-life (3.5 h). The role of the glucocorticoid on GR mRNA stability is controversial: some reports have shown that GR mRNA turnover is unaffected by glucocorticoid treatment (26, 30, 33), and others have reported a hormone-dependent decrease in the GR message stability (28, 31). These discrepancies could be due to the use of different cell types and/or differences in the experimental approach used. The molecular mechanisms that regulate hGR message stability are unknown. The stability of eukaryotic mRNAs is determined by their poly(A) tail and also by specific cis-acting elements which can be located in the UTR and/or the protein coding region (34). The mRNAs of the steroid receptor gene family are known to contain exceptionally long and well conserved 3'-UTRs with multiple copies of the pentanucleotide AUUUA (28), which have a potent mRNA destabilizing effect (34). Our sequence analysis of the hGR mRNA 3'-UTR revealed the existence of nine AUUUA sequences plus one nonameric UUAUUUAAU element with the AUUUA core. The 3'-UTR of hGR mRNA, which is shorter than that of hGR, contained four AUUUA sequences. The A/T content of the 3'-UTR was the same for both hGR and hGR mRNAs (65%). Despite the differences in the AUUUA content, both messages had a similar pattern of mRNA decay. It is worth noting that the half-life of a transfected hGR mRNA which lacked the 3'-UTR (31) was considerably longer than that of the endogenous mRNA, suggesting a strong destabilizing effect of the hGR 3'-UTR. Regulatory factors, such as estrogens, glucocorticoids, or iron, are known to alter the stabilization of a number of messages by increasing or decreasing the expression or the activity of specific binding proteins, which interact with specific sequences within the 3'-UTR (34). Consistent with the results we obtained in the Act-D and CHX experiments, it is tempting to speculate that DEX could inhibit the expression or the activity of a destablizing protein that, through binding with specific elements within the 3'-UTR, would finally stabilize hGR and hGR mRNAs. However, the existence of such a glucocorticoid-induced binding protein and its binding site within the hGR message have yet to be identified. In addition to this post-transcriptional regulation of hGR expression, the Act-D studies suggest that glucocorticoid- induced downregulation of both isoforms of hGR mRNA is basically a transcriptional event. Indeed, several investigators have reported a decreased rate of transcription of the hGR gene after glucocorticoid treatment (26, 29, 31, 33). Previous investigations suggest that interaction of the hormone-bound GR to intragenic cis elements may result in a decreased rate of transcription of the hGR gene. These elements have been located within the 3' end of the receptor-coding sequence of the hGR cDNA (31), as well as within the 3'-UTR of a rat GR cDNA clone (25). A future challenge to explain the homologous downregulation of hGR and hGR is the identification of such downregulating sequences.

To determine whether DEX-induced downregulation of hGR and hGR mRNAs was extensive to hGR and hGR proteins, we performed Western blotting analysis in BEAS-2B, A549, and primary nasal epithelial cell protein extracts. We report that DEX strongly downregulated the expression of hGR protein in BEAS-2B and A549 cells, showing a similar pattern of downregulation over time. In addition, DEX downregulated hGR protein in primary epithelial cells from four different nasal mucosa specimens. It is worth noting that downregulation of hGR protein in BEAS-2B cells was more pronounced than downregulation of its mRNA. This phenomenon has also been reported in a pituitary tumor cell line (AtT-20) after incubation with triamcinolone acetonide for up to 72 h, where Vig and colleagues found oscillatory changes in GR mRNA levels and consistently low levels of GR protein (29). The fact that DEX-induced downregulation of hGR protein is more pronounced than downregulation of its mRNA suggests that, in addition to a transcriptional regulation, a post-translational regulation of the hGR protein is likely to exist. In fact, glucocorticoids are known to decrease the GR protein half-life by 2-fold (15, 17, 27).

In summary, we report the expression and regulation of hGR and hGR isoforms in human respiratory epithelial cells. In BEAS-2B cells, both hGR and hGR mRNAs were downregulated over time by DEX in a dose-dependent manner, through a mechanism independent of ongoing protein synthesis. Our results indirectly suggest that DEX-induced downregulation of hGR and hGR mRNAs is likely due to a decreased rate of hGR gene transcription, which effect would be partly counteracted by the stabilization of hGR mRNA and also, although to a lesser degree, of hGR mRNA. DEX strongly downregulated the expression of hGR protein both in epithelial cell lines (BEAS-2B and A549) and in primary epithelial cells, which suggests that hGR expression may also be regulated at the post-translational level. Future studies are aimed at better understanding the precise molecular determinants involved in the homologous repression of hGR gene transcription, as well as those regulating the hGR at post-transcriptional and post-translational levels.