INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Glucocorticoids (GC) represent the cornerstone anti-inflammatory treatment of chronic asthma. A small proportion of asthmatics develop a severe form of the disease and require a chronic long-term treatment with oral GC (1). These patients, ascribed as GC-dependent asthmatics, present an ongoing inflammation of the airways usually characterized by an increased number of neutrophils (2, 3), activated T lymphocytes (4), and eosinophils (5, 6). In addition, an increased immunoreactivity for granulocyte/macrophage colony-stimulating factor (GM-CSF) and interleukin (IL)-8 has been found respectively in bronchi and induced sputum of GC-dependent asthma patients (6, 7). GM-CSF stimulates the generation of leukocyte precursors (monocytes, neutrophils, and eosinophils) and activates antigen presenting cells (8). IL-8 activates both neutrophils and monocytes (9), being chemotactic only for the former cells. GC-dependent asthma should be differentiated from GC-resistant asthma. GC-resistant asthmatics are defined as patients whose baseline prebronchodilation FEV₁ of less than 70 to 80% predicted improves by less than 15% after 1 to 2 wk of 40 mg prednisolone daily (10).

The effects of GC are mediated by the GC receptor (GR)  which represses expression of various genes encoding inflammatory mediators (11). In addition to GR, an isoform deficient in hormone binding has been isolated in humans and termed GR (12). Both  and  variants are generated by alternative splicing and diverge at their carboxy-termini. It was reported that GR functions as a dominant negative inhibitor of GR in transfected cells (13, 14). However, this observation was not reproduced by others (15).

It has been proposed that resistance of asthmatic patients to the anti-inflammatory effects of GC could result from an elevated level of GR (19). Glucocorticoid resistance in asthma has also been associated with a qualitative or quantitative deficiency in GR (20). However, expression of GR and GR in GC-dependent asthma has not yet been investigated. In a previous study, in which the  and  variants were not distinguished, GR level in bronchial biopsies was found to be similar between different groups of asthmatics including GC-dependent asthmatics (7).

The aim of the present study was to better understand the inflammatory process in GC-dependent asthma and to assess whether this severe form of the disease was associated with a dysregulation of GR or GR. As mentioned previously, GM-CSF and IL-8 have been implicated in airway inflammation in severe asthma and their synthesis is normally repressed by GC (21). We therefore measured the release of these cyto-kines by peripheral blood mononuclear cells (PBMC) from normal subjects, GC-untreated controlled asthmatics, GC-untreated uncontrolled asthmatics, and GC-dependent asthmatics. We found that PBMC of GC-dependent asthma patients produced increased concentrations of GM-CSF and IL-8 despite the long-term GC treatment. To determine whether the persistent release of these cytokines was associated with a dysregulation of the GR, we evaluated the expression of GR and GR at the messenger RNA (mRNA) and protein level in PBMC of the four groups.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Thirty-three asthmatic subjects were selected according to the criteria of the American Thoracic Society as described previously (22). None of the subjects participating in this study was a current smoker. Subjects who had any bronchial or respiratory tract infection during the month preceding the test were excluded from the study. Patients were excluded from the study if they had had a severe exacerbation of asthma resulting in hospitalization during the month preceding the study. The first group consisted of 11 subjects (age in median, 25th to 75th percentiles: 22, 20 to 25 yr) with mild intermittent asthma who took inhaled short-acting ₂-agonists as needed but no GC and were defined as untreated controlled asthmatics. The second group consisted of eight patients (age in median, 25th to 75th percentiles: 29, 25 to 32 yr) with recent nocturnal and diurnal symptoms, who required more than 4 puffs of ₂-agonists a day, but did not take any GC and were defined as untreated uncontrolled. These patients were subsequently treated with a short course of oral GC (1 mg/kg prednisolone for 10 d). The third group of 14 patients (age in median, 25th to 75th percentiles: 44, 31 to 56 yr) with GC-dependent asthma was defined as previously described (7). In the latter group, all had severe persistent asthma which required a daily dose of inhaled GC (2,000 µg fluticasone propionate), oral prednisone, long-acting ₂-agonists (100 µg salmeterol), and short-acting ₂-agonists as required. These patients were all considered as GC-dependent because, in the past 2 yr, the attempt to wean them from the systemic treatment had always failed. Compliance was checked by using diary cards and by questionnaire during the follow-up visits. For all GC-dependent asthmatics, bone mineral density in the lumbar spine and proximal femur was decreased as assessed by absorptiometry (T score in median, 25th to 75th percentiles was respectively: 1.46, 1.67/1.20; 1.60, 2.10/1.30). These observations suggest that these patients complied with their oral GC treatment.

Eight healthy subjects (age in median, 25th to 75th percentiles: 29, 25 to 40 yr) were used as a control group. Their pulmonary function was within normal range. Subjects who had any bronchial or respiratory tract infection during the month preceding the test were excluded from the study.

The study was approved by the Ethics Committee of the Montpellier-Nîmes hospital and written informed consent was obtained from all patients.

Isolation of PBMC

PBMC were isolated by Ficoll-Hypaque gradient centrifugation at 1,800 g for 20 min at 20° C. They were removed from the plasma/Ficoll interface and washed twice in RPMI. Purity of the cells was assessed by May-Grünwald-Giemsa staining and was always > 97%. Viability was always > 95%.

IL-8 and GM-CSF Release by PBMC

Freshly isolated PBMC were diluted at a concentration of one million cells/ml in RPMI 1640 medium containing 10% heat-inactivated fetal calf serum (FCS), 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine. Cells were cultivated for 3, 12, and 24 h in the absence or presence of the GC dexamethasone (DEX) at 0.1 µM. Supernatants were then harvested to measure their content in IL-8 and GM-CSF by quantitative sandwich enzyme immunoassays, following the manufacturer's recommendations (R&D Systems, Oxon, UK). The limit of detection was 10 pg/ml for IL-8 and 0.36 pg/ml for GM-CSF.

Isolation and Analysis of Total RNA

Total cellular RNA was obtained by lysing PBMC following a standard protocol. Briefly, the extracted RNA in the aqueous phase was obtained after homogenization of the cells in the reaction mixture containing RNAzol (Bioprobe System, Montreuil, France) and chloroform (Prolabo, Paris, France), followed by centrifugation at 12,000 g for 15 min at 4° C. The RNA extract was precipitated with 1 vol isopropanol at 4° C for 15 min and centrifuged at 12,000 g for 15 min at 4° C. The RNA pellet was washed with 70% ethanol, vacuum dried briefly, solubilized in water, and stored at 80° C until subsequent analysis. The quantity of RNA was calculated by spectrophotometry at 260 nm. The integrity of purified RNA was determined by visualization of the 28 S and 18 S ribosomal RNA bands after electrophoresis of 1 to 2 µg of each RNA sample through a 1% agarose gel.

Reverse Transcription and Polymerase Chain Reaction

An amount of 5 µg of total RNA was subjected to reverse transcription (RT) for 1 h at 37° C. Reaction mixture contained 0.5 mM deoxyadenosine triphosphate (dATP), deoxyguanosine triphosphate (dGTP), deoxycytidine triphosphate (dCTP), deoxythymidine triphosphate (dTTP; Promega, Madison, WI), 10 mM dithiothreitol (DTT; Promega), 20 U RNasin ribonuclease inhibitor (Promega), 0.25 µg oligo dT (Gibco BRL, Gaithersburg, MD), and 200 U of Moloney murine leukemia virus reverse transcriptase (Gibco BRL) in 25 µl supplied buffer. The reaction mixture was heated to 98° C for 5 min to stop RT.

The primers used for amplification of GR message were as follows: 5'-CCTAAGGACGGTCTGAAGAGC-3' (upstream) and 5'-GCCAAGTCTTGGCCCTCTAT-3' (downstream), corresponding to nucleotides 2158-2178 and 2616-2635 of GR complementary DNA (cDNA) (14).

The primers used for amplification of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were as follows: GAPDH sense 5'-TCGCCAGCCGAGCCACAT-3', GAPDH antisense 5'-GGAACATGT-AAACCATGTAGTTG-3'. Polymerase chain reaction (PCR) was performed with 4 µl and 8 µl of RT reaction mixture to analyze GAPDH and GR mRNA levels, respectively. Control PCR were carried out with no RT reaction mixture. The reactions contained 0.5 U Taq DNA polymerase, 0.2 µM of each oligonucleotide primer, 0.2 mM deoxyribonucleoside triphosphate (dNTP), 50 mM KCl, 10 mM Tris-HCl pH 8.3, 0.1% Triton X-100, 2 mM MgCl₂ in a final volume of 50 µl. PCR conditions were 30 cycles of 30 s at 95° C, 30 s at 62° C, and 30 s at 72° C. These were followed by a final extension step at 72° C for 10 min. Amplified DNA fragments were electrophoretically fractionated on 1.7% agarose gels containing 0.5 µg/ml ethidium bromide and visualized under ultraviolet light. GR and GAPDH PCR products were semi-quantified by densitometric scanning using a monochrome charge-coupled device (CCD) camera RS-170 (COHU, San Diego, CA) coupled to NIH Image analysis software (NIH, Bethesda, MD). Amount of GR mRNA was normalized to that of GAPDH mRNA.

Nested PCR

Nested PCR was used to amplify GR. The first round of PCR was performed using 8 µl of the cDNA from the reverse transcription. The primers used for the first round were as follows: 5'-CCTAAGGACGGTCTGAAGAGC-3' (upstream) and 5'-CCACGTATCCTAA-AAGGGCAC-3' (downstream), corresponding to nucleotides 2158 to 2178 and 2503 to 2523 of GR cDNA (14). The PCR mix contained 0.2 µM of each outer primer or 0.2 µM of the GAPDH primers, together with the same reagents as previously described. PCR conditions were 30 cycles of 30 s at 95° C, 30 s at 61° C, and 30 s at 72° C. These were followed by a final extension step at 72° C for 10 min.

Nested PCR was initiated with 4 µl of the first-round PCR products. The primers used were as follows: 5'-AGCACATCTCACAC-ATTAAT-3' (upstream) and 5'-TATAGTTGTCGATGAGCATC-3' (downstream), corresponding to nucleotides 2338 to 2357 and 2455 to 2471 of GR cDNA. The PCR mix was as described previously. Samples were subjected to 30 cycles of 30 s at 95° C, 30 s at 54° C, and 30 s at 72° C. These were followed by a final extension step at 72° C for 10 min.

Amplified DNA fragments were revealed as described previously. Quantities of GR product from the nested PCR and of GAPDH product from one round of PCR were estimated by densitometric scanning. Amount of GR mRNA was normalized to that of GAPDH mRNA.

Culture and Transfection of A549 Cells

A549 human lung carcinoma cells were cultivated in Ham's F12 medium containing 10% heat-inactivated FCS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine.

For transfection, 4 µg of DNA was diluted in 20 mM Hepes pH 7.4, 150 mM NaCl, complexed with a replication-deficient adenovirus, transferrin-polylysine, and poly-L-lysine as described previously (23), and added to one million cells. A549 cells were transfected with the expression vectors for human GR (pRShGR) or human GR (pRShGR) (kindly provided by R. Evans, Salk Institute, San Diego, CA). After transfection, the cells were cultivated for 24 h.

Western Blotting

Freshly isolated PBMC and untransfected or transfected A549 cells were washed with cold phosphate-buffered saline (PBS) and lysed in 10 mM Tris-HCl, pH 7.4, 50 mM NaCl, 5 mM ethylenediaminetetraacetic acid (EDTA), 1% Nonidet P-40, and 10 µg/ml phenylmethylsulfonyl fluoride (PMSF). Cell extracts were transferred in microcentrifuge tubes, mixed, and left on ice for 10 min. After one cycle of freeze/thaw, they were centrifuged at 12,000 g for 5 min at 4° C. A sample of the supernatant was taken for protein estimation and the remainder adjusted to 1× Laemmli dissociation buffer. An amount of 50 µg of total protein was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 4 to 12% gradient gels (Novex, San Diego, CA) and blotted onto nitrocellulose membranes. These were blocked with PBS containing 3% bovine serum albumin (BSA), 0.1% Tween 20, and then probed with either polyclonal antibodies directed against a common epitope (amino acids 245 to 259) of human GR and GR (Affinity Bioreagents, Golden, CO) or with polyclonal antibodies recognizing specifically GR (kind gift of J. A. Cidlowski, NIH, Research Triangle Park, NC). Anti-GR/ and anti-GR antibodies were used respectively at a 1:100 dilution and at a 1:500 dilution. After serial washes with PBS containing 0.1% Tween 20, membranes were incubated with peroxidase-conjugated secondary antibodies at a dilution of 1:15,000 (Sigma, St. Louis, MO). In some experiments, to analyze the quantity of a control protein unaffected by GC treatment, blots were also probed with an anti--actin monoclonal antibody diluted 1:5,000 (Sigma) and peroxidase-conjugated anti-mouse antibodies diluted 1:5,000 (Dako, Glostrup, Denmark). Revelation was performed with an enhanced chemiluminescence system (NEN, Boston, MA) followed by autoradiography. Autoradiographic films were analyzed by densitometric scanning using a monochrome CCD camera RS-170 (COHU) coupled to the NIH Image analysis program. The amount of GR in PBMC was normalized relatively to that in A549 cells. Therefore, 10 µg of total protein from A549 cells were loaded on each gel. Results were expressed as arbitrary units (AU)/µg total protein. As positive control for the detection of GR, 5 µg of total protein from A549 cells transfected with a GR expression vector were loaded on the gels.

Study Design

Using Western blot analyses, we assessed the amount of GR and GR protein in PBMC isolated from eight normal subjects, 11 untreated controlled asthmatics, eight untreated uncontrolled asthmatics, and 14 GC-dependent asthmatics. The other assays were performed on a subset of each group owing to limitation in the amount of cells recovered from certain subjects. We evaluated by RT-PCR the level of GR and GR mRNA in PBMC isolated from 6 of 8 normal subjects, 6 of 11 untreated controlled asthmatics, 6 of 8 untreated uncontrolled asthmatics, and 6 of 14 GC-dependent asthmatics. We also determined expression of GR protein in PBMC isolated from 6 of 8 previously untreated uncontrolled asthmatics before and after 10 d of treatment with an oral dose of 1 mg/kg prednisolone. Using quantitative sandwich enzyme immunoassays, we measured IL-8 and GM-CSF release by PBMC isolated from 8 of 8 normal subjects, 6 of 11 untreated controlled asthmatics, 6 of 8 untreated uncontrolled asthmatics, and 9 of 14 GC-dependent asthmatics. We studied IL-8 and GM-CSF release by PBMC recovered from 6 of 8 patients with previously untreated uncontrolled asthma before and after 10 d of treatment with an oral dose of 1 mg/kg prednisolone. We then investigated the in vitro effect of DEX on IL-8 and GM-CSF production by PBMC of 8 of 8 normal subjects, 6 of 11 untreated controlled asthmatics, 6 of 8 untreated uncontrolled asthmatics, and 9 of 14 GC-dependent asthmatics.

Statistical Analysis

Nonparametric tests were used to analyze the data. The Kruskal-Wallis test and the Dunn's post hoc test were used to compare cytokine release by PBMC from the different groups. The Mann-Whitney test was used to compare cytokine release when PBMC of GC-dependent asthmatics were treated or not by DEX in vitro. The Wilcoxon test was used for paired comparisons of cytokine release and GR expression, when previously untreated uncontrolled asthmatics were given a short course of oral GC. Statistical significance was set at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Characteristics of the Patients

Demographic characteristics of normal subjects, untreated controlled, untreated uncontrolled, and GC-dependent asthmatics are shown in Table 1. GC-dependent asthmatics were older than the other subjects, but a significant difference in age was found only between GC-dependent asthmatics and untreated controlled asthmatics (p < 0.001). GC-dependent patients had asthma of a longer duration than less severe asthmatics, but difference was not statistically significant. Their oral GC requirement was highly variable. Airflow impairment remained high in GC-dependent asthmatics despite their chronic GC treatment and was similar to that in untreated uncontrolled patients.

Eight previously untreated uncontrolled asthmatic patients were treated for 10 d with an oral GC (1 mg/kg of prednisolone). After treatment, all of them experienced an improvement of FEV₁ (median, 25th to 75th percentiles before treatment: 52, 47 to 56; median, 25th to 75th percentiles after treatment: 79, 77 to 89; p = 0.0078).

IL-8 and GM-CSF Production

We measured the amount of IL-8 and GM-CSF released by PBMC isolated from the different study groups and cultured for 3, 12, and 24 h. At any time point, the amount of IL-8 released by PBMC of GC-dependent asthmatics and untreated uncontrolled asthmatics was higher than that of normal subjects (p < 0.001) but not statistically different from that of the other study groups (Figure 1A). At any time point, PBMC of GC-dependent asthmatics also released higher concentrations of GM-CSF than PBMC isolated from untreated controlled asthmatics (p < 0.01) and normal subjects (p < 0.001) (Figure 1B). After 24 h in culture, the amount of GM-CSF released by PBMC of untreated uncontrolled asthmatics was higher than that of normal subjects (p < 0.05) (Figure 1B). Higher concentrations of IL-8 and GM-CSF were also released by cells from untreated uncontrolled asthmatics as compared with untreated controlled asthmatics (Figure 1), and the amount of IL-8 produced by PBMC of untreated controlled asthmatics was greater than that of normal subjects (Figure 1A). These latter differences may be of importance even though these do not appear statistically significant with the test used.

View larger version (20K): [in this window] [in a new window]   Figure 1.   IL-8 and GM-CSF release by PBMC. Concentrations of IL-8 (A) and GM-CSF (B) released by PBMC isolated from eight normal subjects, six untreated controlled asthmatics, six untreated uncontrolled asthmatics, and nine GC-dependent asthmatics. Individual data are shown. Horizontal lines represent the median.

The amounts of IL-8 and GM-CSF produced by PBMC of previously untreated uncontrolled asthmatics were reduced after a short course of oral GC (p = 0.0078) (Figure 2).

View larger version (17K): [in this window] [in a new window]   Figure 2.   Effect of a short-course oral GC treatment on IL-8 and GM-CSF production. Concentrations of IL-8 (A) and GM-CSF (B) released by PBMC of eight previously untreated uncontrolled asthmatics before and after a short-course oral GC treatment (prednisolone at 1 mg/kg for 10 d). Individual data are shown; p = 0.0078.

An in vitro DEX treatment of 24 h inhibited release of IL-8 (p = 0.0012) and GM-CSF (p < 0.0001) by PBMC of GC-dependent asthmatics (Figure 3). Similar results were obtained at 3 and 12 h, and with PBMC isolated from the other asthmatic groups (data not shown).

View larger version (12K): [in this window] [in a new window]   Figure 3.   Effect of DEX treatment in vitro on IL-8 and GM-CSF release. Concentrations of IL-8 (A) and GM-CSF (B) produced by PBMC isolated from GC dependent asthmatics (n = 9) and cultivated in the absence or presence of 0.1 µM DEX for 24 h. Individual data are shown. Horizontal lines represent the median. *p = 0.0012 versus DEX; #p < 0.0001 versus DEX.

GR mRNA Levels

To evaluate GR mRNA expression in PBMC, RT-PCR was performed with primers specific for GR and GAPDH cDNAs. Expression of the housekeeping gene GAPDH was determined to provide an internal control for RT and PCR efficiencies. A representative electrophoretic analysis of the PCR products obtained is shown in Figure 4A. Densitometric scanning of the data indicated that the amount of GR mRNA was similar in normal subjects and in the three groups of asthmatics (Figure 4B).

View larger version (21K): [in this window] [in a new window]   Figure 4.   GR mRNA expression in PBMC of normal subjects, untreated controlled, untreated uncontrolled, and GC-dependent asthmatics. (A) Representative results of RT-PCR for GR and GAPDH mRNAs in PBMC of one control subject (lane 1), one untreated controlled asthmatic (lane 2), one untreated uncontrolled asthmatic (lane 3), and two GC-dependent asthmatics (lanes 4 and 5). M, pBR322 DNA-MspI digest. (B) PCR products were semiquantified by densitometric scanning and normalized relatively to the amount of GAPDH. Results are expressed in AU.

GR mRNA Levels

To evaluate GR mRNA expression in PBMC, RT-PCR was performed with primers specific for GR and GAPDH cDNAs. No GR PCR product was detected after a first round of amplification in the four groups of subjects studied (data not shown). Detection of GR transcripts required the use of nested PCR as shown on the representative electrophoretic analysis (Figure 5A). GAPDH mRNA expression was determined after the first round of PCR. Densitometric analysis of the data indicated that the amount of GR mRNA was similar in normal subjects and in the three groups of asthmatics (Figure 5B).

View larger version (21K): [in this window] [in a new window]   Figure 5.   GR mRNA expression in PBMC of normal subjects, untreated controlled, untreated uncontrolled, and GC-dependent asthmatics. (A) Representative results of nested PCR for GR mRNA and RT-PCR for GAPDH mRNA. Starting material was total RNA purified from PBMC of one control subject (lane 1), one untreated controlled asthmatic (lane 2), one untreated uncontrolled asthmatic (lane 3), and one GC-dependent asthmatic (lane 4). M, pBR322 DNA-MspI digest. (B) PCR products were semiquantified by densitometric scanning and normalized relatively to the amount of GAPDH. Results are expressed in AU.

GR and GR Protein Levels

To investigate the relative protein level of GR and GR, Western blots were performed initially with antibodies recognizing both forms of the receptor which can be distinguished by their difference in size (94 kD for GR versus 90 kD for GR). In each gel, a constant amount (10 µg) of protein from A549 human lung epithelial cells was loaded to provide a qualitative and quantitative reference for GR expression (Figure 6A, lane 1). Indeed, A549 cells express a single receptor species of the same size as overexpressed GR and of higher size than overexpressed GR (data not shown). PBMC isolated from the four groups of subjects expressed similar amount of an immunoreactive protein of the same size as GR (Figures 6A and 6B). A short course of oral GC downregulated GR (p = 0.0312), but not -actin, in PBMC of previously untreated uncontrolled asthmatics (Figure 7).

View larger version (25K): [in this window] [in a new window]   Figure 6.   GR protein expression in PBMC of normal subjects, untreated controlled, untreated uncontrolled, and GC-dependent asthmatics. (A) Representative Western blot analysis of GR in PBMC from one control subject (lane 2), two untreated controlled asthmatics (lanes 3 and 4), two untreated uncontrolled asthmatics (lanes 5 and 6), and three GC-dependent asthmatics (lanes 7, 8, and 9). Fifty micrograms of proteins from whole cell extracts were separated on 4 to 12% SDS-polyacrylamide gels and transferred to nitrocellulose membranes which were then probed with anti-GR/ antibodies. In lane 1, 10 µg of total protein from A549 cells were loaded on the gel as reference for GR expression. (B) Signals corresponding to GR on the various Western blots were semiquantified by densitometric scanning and normalized relatively to the amount of GR in A549 cells. Data are expressed as AU/µg of total protein.

View larger version (13K): [in this window] [in a new window]   Figure 7.   Downregulation of GR after a short-course oral GC treatment of uncontrolled asthmatics. (A) Representative Western blot analysis of GR and -actin expression in PBMC of one previously untreated uncontrolled asthmatic before and after oral GC treatment (prednisolone at 1 mg/kg for 10 d). (B) Relative amounts of GR in PBMC of six previously untreated uncontrolled asthmatics before and after oral GC treatment. Individual data are shown and expressed in AU; p = 0.0312. (C ) Relative amounts of -actin in PBMC of the same patients before and after the treatment. Individual data are shown and expressed in AU.

No signal was obtained using anti-GR antibodies, confirming that GR expression in PBMC of normal subjects, untreated controlled asthmatics, untreated uncontrolled asthmatics, and GC-dependent asthmatics was undetectable by direct Western blotting (Figure 8, lanes 3-7). The absence of signal was not caused by inappropriate experimental conditions because GR was detected in A549 cells transfected with a GR expression vector (Figure 8, lane 1). In addition, the anti-GR antibodies did not reveal any protein in A549 cells overexpressing GR (Figure 8, lane 2), indicating that they do not cross-react with GR and confirming that A549 cells do not express GR protein at a level detectable by direct Western blotting.

View larger version (25K): [in this window] [in a new window]   Figure 8.   GR protein expression in PBMC of normal subjects, untreated controlled, untreated uncontrolled, and GC-dependent asthmatics. Representative Western blot analysis of GR in PBMC from one control subject (lane 3), one untreated controlled asthmatic (lane 4), one untreated uncontrolled asthmatic (lane 5), and two GC-dependent asthmatics (lanes 6 and 7). In lane 1, 5 µg of total protein from A549 cells transfected with a GR expression vector were loaded as positive control. In lane 2, 5 µg of total protein from A549 cells transfected with a GR expression vector were loaded as negative control.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Despite the long-term use of oral and inhaled GC, a persistent inflammation affects the airways of GC-dependent asthmatics (2, 6, 7). Here, we show that high concentrations of IL-8 and GM-CSF were released by PBMC of GC-dependent asthmatics whereas production of these two cytokines by PBMC of previously untreated uncontrolled asthmatics was inhibited after a short course of oral GC. It is unlikely that the absence of inhibition of IL-8 and GM-CSF release by oral GC in GC-dependent asthmatics is due to lack of compliance. Indeed, bone mineral density of GC-dependent asthmatics was decreased and attempts to wean them from the systemic treatment were unsuccessful, suggesting that these patients complied in taking their oral GC. Higher levels of IL-8 and GM-CSF were also released by cells from untreated uncontrolled asthmatics as compared with untreated controlled asthmatics, and IL-8 was produced in greater quantity by PBMC of untreated controlled asthmatics than by normal subjects. Although the latter differences were not statistically significant, overall our data suggest that production of IL-8 and GM-CSF may be related to the severity of the disease. A similar observation was reported by Shute and coworkers who measured IL-8 in blood and bronchial biopsies of asthmatic subjects (24).

In the present study, IL-8 and GM-CSF were used as markers for cell activation and response to GC treatment. Persistent production of these cytokines by PBMC is in agreement with previous reports of extrathoracic inflammation in asthma (25). We found that neutrophil counts were significantly higher in blood samples from GC-dependent asthmatics as compared with untreated controlled asthmatics (data not shown). Because PBMC of GC-dependent asthmatics released a higher amount of GM-CSF, but a similar level of IL-8 as compared with untreated controlled asthmatics (Figure 1), neutrophilia in GC-dependent asthma may result in part from increased GM-CSF production. However, neutrophilia in GC-dependent asthma may also be the consequence of GC treatment (26). In this regard, GC have been shown to increase neutrophil survival and to enhance the survival effect of GM-CSF (27).

It has been proposed that elevated level of GR might cause corticoresistance in asthma patients (19, 28). This hypothesis is based on GR expression studies in tissues obtained from corticoresistant asthmatics and on the inhibitory effect of GR on GR-mediated transcription in vitro (13, 14). However, this latter observation was not reproduced by others (15). Moreover, to inhibit GR-mediated gene regulation, GR has to be more abundant than GR and conflicting data concerning the relative levels of the two isoforms were obtained. In one study conducted with various human tissues and HeLa cells, the amount of GR was found to be equal or higher than that of GR (29). In contrast, in other studies, level of GR in HeLa cell and human lymphocytes was found to be relatively lower than that of GR (15, 16). In agreement with those latter results, quantitative RT-PCR experiments have demonstrated that, in all human tissues and cell lines analyzed so far, the GR mRNA was 200- to 500-fold less represented than the GR mRNA (14). We report herein that GR protein is largely predominant over GR in PBMC of GC-dependent asthmatics. We further confirmed this observation at the mRNA level because GR was revealed by a simple PCR whereas detection of GR required nested PCR. Moreover, GR mRNA expression was not different in GC-dependent asthmatics than in the other groups of asthmatics. These data do not support a role of GR in the pathogenesis of GC-dependent asthma and, therefore, there is a clear distinction with GC-resistant asthma.

After a treatment with inhaled or oral GC, GR mRNA was shown to be downregulated in PBMC and endobronchial biopsies of mild to moderate asthmatics (30, 31). We confirm this observation at the protein level: after a short course of oral GC, downregulation of GR protein occurred in PBMC of previously untreated uncontrolled asthmatics. In contrast, we found that in PBMC from GC-dependent asthmatics, despite chronic exposure to GC, expression of GR mRNA and protein was similar to that in normal subjects, untreated controlled, and untreated uncontrolled asthmatics. This finding indicates that long-term GC treatment does not downregulate GR in GC-dependent asthmatics. Therefore, the incapacity of GC to inhibit IL-8 and GM-CSF release in these patients is not due to a low level of GR. It makes sense that GC-dependent patients retain GR expression because they require a GC therapy. Indeed, withdrawal of GC aggravates their symptoms (3). In addition, although oral GC are not entirely satisfactory to treat GC-dependent asthmatics, they reduce eosinophilic inflammation (3), a key pathophysiological feature of asthma (32). Glucocorticoid-dependent asthmatics also develop side effects further indicating that they respond to GC treatment and therefore express GR. Interestingly, inhibition of IL-8 and GM-CSF release occurred when PBMC of GC-dependent asthmatics were treated in vitro by the GC dexamethasone. Thus, yet undefined factors in vivo are likely responsible for maintaining a high production of IL-8 and GM-CSF and preventing downregulation of GR by GC in PBMC of GC-dependent asthmatics. Previous work has suggested that the transcription factor nuclear factor kappa B (NF-B) plays an important role in the induction of various cytokines, including IL-8 and GM-CSF (33, 34) and that it is target for GC-mediated gene repression (11). Moreover, it was demonstrated that NF-B DNA binding activity is reduced in bronchial mucosa of mild to moderate asthmatics after treatment with an inhaled GC (35). We are currently investigating whether an excess of NF-B activity could explain the persistent production of IL-8 and GM-CSF in GC-dependent asthma and whether other proinflammatory cytokines are upregulated in this severe form of the disease.

In conclusion, in GC-dependent asthma, the ongoing release of IL-8 and GM-CSF despite GC treatment is not the result of low expression of GR or overexpression of GR.