Defective glucocorticoid receptor nuclear translocation and altered histone acetylation patterns in glucocorticoid-resistant patients

Transcription

1 Defective glucocorticoid receptor nuclear translocation and altered histone acetylation patterns in glucocorticoid-resistant patients John G. Matthews, MRCP, a Kazuhiro Ito, PhD, DVM, a Peter J. Barnes, DM, DSc, FRCP, and Ian M. Adcock, PhD London, United Kingdom Background: Most chronic inflammatory diseases are well controlled by glucocorticoids. However, a minority of patients fails to respond adequately to this treatment. Objective: We wished to determine whether glucocorticoid insensitivity in a group of steroid-resistant (SR) and steroiddependent (SD) asthmatic subjects resulted from an inability of the glucocorticoid receptor (GR) to translocate into the nucleus. Methods: Glucocorticoid receptor nuclear translocation was determined in PBMCs by immunocytochemistry and GR function measured by suppression of TNF-aeinduced GM- CSF release and effects of dexamethasone on histone acetylation. Results: Glucocorticoid repression of TNF-aeinduced GM- CSF release was reduced in PBMCs from SD and SR patients. This inhibition correlated with a failure of GR to translocate into the nucleus and induce histone acetylation in response to dexamethasone. In addition, dexamethasone failed to inhibit TNF-aeinduced histone acetyltransferase activity, which predominantly targeted histone residues lysine (K)8 and K12. However, in a subset of patients, even high levels of GR nuclear translocation failed to produce histone acetylation in response to dexamethasone. Histone H4 K5 acetylation, a marker of dexamethasone transactivation, was specifically reduced in this group. However, cells from this subset of steroid-insensitive subjects were still capable of inhibiting TNF-aeinduced histone acetylation. Conclusion: We have identified a novel mechanism of glucocorticoid insensitivity in a group of SR and SD subjects. These data suggest that most patients respond to glucocorticoids according to the degree of GR nuclear translocation occurring, but some subjects with steroid resistance have a reduced response because of a failure of steroids to transactivate, rather than transrepress. (J Allergy Clin Immunol 2004;113: ) Key words: Asthma, monocytes, chromatin, immunohistochemistry, ELISA From Thoracic Medicine, National Heart and Lung Institute, Imperial College of Science, Technology and Medicine. a These authors contributed equally to this work. Supported by the Wellcome Trust, the National Asthma Campaign, the Clinical Research Committee (Royal Brompton Hospital), and GlaxoSmithKline (United Kingdom). Received for publication January 7, 2004; revised March 2, 2004; accepted for publication March 2, Reprint requests: Ian M. Adcock, PhD, Thoracic Medicine, Imperial College School of Medicine, National Heart and Lung Institute, Dovehouse Street, London SW3 6LY, United Kingdom. ian.adcock@imperial.ac.uk /$30.00 Ó 2004 American Academy of Allergy, Asthma and Immunology doi: /j.jaci Abbreviations used GR: Glucocorticoid receptor HAT: Histone acetyltransferase HDAC: Histone deacetylase K5: N-terminal lysine number 5 on histone 4 K8: N-terminal lysine number 8 on histone 4 K12: N-terminal lysine number 12 on histone 4 K16: N-terminal lysine number 16 on histone 4 MAPK: Mitogen-activated protein kinase MKP-1: Mitogen-activated protein kinase phosphatase 1 NF-jB: Nuclear factor jb SD: Glucocorticoid-dependent SR: Glucocorticoid-resistant SS: Glucocorticoid-sensitive TSA: Trichostatin A UK: United Kingdom Glucocorticoids are effective anti-inflammatory drugs in several diseases, such as asthma, rheumatoid arthritis, and inflammatory bowel disease, in which multiple inflammatory genes are expressed. In asthma, they suppress the airway inflammatory response and reduce airway hyperresponsiveness. 1 However, their exact molecular mechanisms and cellular targets in the airways remain uncertain. Asthma is characterized by chronic airway inflammation, involving eosinophils, macrophages, mast cells, and CD4 + T-lymphocytes, which is characteristically suppressed by glucocorticoids. Glucocorticoids act by binding to a cytoplasmic receptor (GR), which rapidly translocates into the nucleus. Within the nucleus, the complex binds as a dimer to specific glucocorticoid response elements on DNA within the promoter region of glucocorticoid-responsive genes to enhance transcription (transactivation) 2 after induction of histone acetylation. 3,4 Glucocorticoids induce the acetylation of specific lysine residues (K5 and K16) in histone H4. 3 Glucocorticoids also prevent other transcription factors, such as activator protein 1 and nuclear factor jb (NF-jB), from activating their target genes by inhibition of acetylation of specific lysine residues in histone H4. 3 Although glucocorticoids are effective in controlling asthma in the majority of patients, there is a minority of patients that responds less well, needing high doses of oral glucocorticoids, with a small proportion demonstrating glucocorticoid resistance. 5 Steroid resistance is also found in other inflammatory diseases, such as rheumatoid arthritis and inflammatory bowel disease.

2 J ALLERGY CLIN IMMUNOL VOLUME 113, NUMBER 6 Matthews et al 1101 A spectrum of glucocorticoid responsiveness in airway inflammatory diseases may exist, reflecting several mechanisms, secondary either to disease activity itself or to the effects of therapy, with the glucocorticoid-resistant asthmatic at 1 extreme of this spectrum. At a molecular level, resistance to the anti-inflammatory effects of glucocorticoids can be induced by several mechanisms. 6 The reduction in glucocorticoid responsiveness observed in cells from these subjects has been ascribed to a reduced number of GRs, 7 altered affinity of glucocorticoid binding to GR, 8 reduced ability of GR to bind to DNA, 9 increased expression of inflammatory transcription factors, 10 or increased expression of the b isoform of GR that binds to DNA but does not bind glucocorticoids. 11 We have investigated the mechanism for reduced glucocorticoid-responsiveness in asthmatic patients by studying whether GR nuclear translocation and subsequent histone acetylation is impaired in patients with steroid-dependent and steroid-resistant asthma. MATERIALS AND METHODS Patients Six glucocorticoid-sensitive subjects (SS; controlled on inhaled steroids), 11 glucocorticoid-dependent asthmatic subjects (SD; controlled only on oral steroids), 9 steroid-resistant asthmatic subjects who showed no response even to high doses of oral steroids (SR), and 10 normal nonasthmatic control subjects were studied (Table I). All subjects were matched for age, and no current smokers were included in the study. The SD (2.7 ± 2 pack years) and SR (0.1 ± 0.2 pack years) groups included some exsmokers. Asthmatics were defined as patients who had shown within the previous year a 15% increase in FEV 1 to 200 lg inhaled albuterol via a metered dose inhaler. All patients were receiving inhaled albuterol on an asrequired basis. The mean dose per day for albuterol was 900 ± 428 lg (SS), 1810 ± 844 lg (SD), and 618 ± 204 lg (SR) for each group. SD patients were on maximal inhaled corticosteroid therapy, with most patients using a dry powder inhaler (2500 ± 327 lg fluticasone; n = 8) or a nebulizer (2.7 ± 0.7 mg budesonide; n = 3). SD patients also required a maintenance dose of oral prednisolone (24 ± 5 mg daily) for the control of their asthma. SR patients were on maximal inhaled corticosteroid therapy (1380 ± 308 lg fluticasone via a dry powder inhaler) and failed to have a >15% improvement in FEV 1 after a 2-week course of 40 mg prednisolone (mean response ÿ5.6 ± 3 L). The study was approved by the Ethics Committee of the Royal Brompton Hospital, and written informed consent was obtained from all subjects. PBMC isolation PBMCs were isolated by Ficoll-Hypaque (Pharmacia-Biotek, St Albans, United Kingdom [UK]) density centrifugation. Viability was assessed by trypan blue exclusion and was consistently >97%. Cells at a concentration of /ml in RPMI with HEPES (20 mmol/l) supplemented with 10% FCS, penicillin, streptomycin, and amphotericin B were cultured for 18 hours at 378C in medium alone, TNF-a (10 ng/ml, R&D Systems, Abingdon, UK), or TNF-a and dexamethasone (10 ÿ6 mol/l). Thereafter, plates were centrifuged, supernatants were collected, and GM-CSF was measured by sandwich ELISA. 12 Western blotting PBMCs were isolated and cells harvested and washed repeatedly before lysis as previously described. 13 Total cellular proteins (30 lg) were size-fractionated by 7% SDS-PAGE, and Western blot analysis was performed as previously described by using 1:1000 dilution of anti-gr antibody (Santa Cruz, Wembley, UK), which detects both GRa and GRb isoforms. 14 Immunocytochemistry PBMCs ( ) were cultured in 8-well slide chambers in the presence or absence of dexamethasone (10 ÿ6 mol/l) or trichostatin A (TSA:100 ng/ml). Cells were washed with HBSS and air-dried for 30 minutes at room temperature. Cells were then fixed in ice-cold acetone-methanol (50/50, wt/wt) for 10 minutes. Slides were air-dried and incubated with blocking buffer (20% normal swine serum in PBS, 0.1% saponin; Dako, Ely, UK) for 20 minutes, followed by 1 hour incubation with primary antibody solution (PBS, 0.1% saponin, 1% BSA). Antibodies against panacetylated H4, H4-K5, H4-K8, H4- K12, and H4-K16 (Serotec, Cambridge, UK) were used at 1:100 to 1:300 dilution, or GRa (Santa Cruz) at 1:50. Slides were washed twice and incubated with biotinylated swine antirabbit IgG (1:200; Dako) for 45 minutes. Slides were washed again before incubation with fluorescein isothiocyanate-conjugated streptavidin (1:100) for 45 minutes. The slides were washed twice more before counterstaining with 20% hematoxylin and mounting. Stained cells were observed by confocal microscopy. Confocal scanning laser microscopy images were collected with a Leica confocal microscope equipped with a 488/514-nm dual-band argon ion laser (Leica Microsystems, Milton Keynes, UK). An oil-immersion objective was used, and images were collected by using TCSNT (Leica) software. Positively stained nuclei and total cells (500) were counted on each slide. The observer was blinded to the clinical status of the subject. Histone acetylation activity Cells were plated at a density of cells/ml and exposed to 0.05 mci/ml of 3 H acetate (Amersham-Pharmacia, Amersham, UK). After incubation for 10 minutes at 378C, cells were stimulated for 6 hours with 10 ÿ6 mol/l dexamethasone. Histones were isolated and separated by electrophoresis on SDS-16% polyacrylamide gel. Gels were stained with Coomassie brilliant blue, and the core histones (H2A, H2B, H3, and H4) were excised. The radioactivity in extracted core histones was determined by liquid scintillation counting and normalized to protein concentration. Data analysis Data are expressed as means ± SEMs of n independent observations. The results obtained before or after drug treatment were compared by ANOVA and Bonferroni multiple comparison test. Correlations were calculated by using linear iterative regression with the PRISM curve-fitting program (GraphPad Instat software program; GraphPad Software Inc, San Diego, Calif). Repeatability of measurements was analyzed by using the Bland and Altman method. 15 Blood samples were taken on a second follow-up visit and analyzed in the same manner. The mean difference and the standard deviation of the differences between individual measures at the first and second occasion were calculated, and 95% of differences were within 2 standard deviations (repeatability coefficient). RESULTS Effect of dexamethasone on TNF-aeinduced GM-CSF production in PBMCs The basal levels of GM-CSF production from PBMCs (10 6 cells) from normal subjects and SS, SD, and SR

3 1102 Matthews et al J ALLERGY CLIN IMMUNOL JUNE 2004 patients were similar in all groups. GM-CSF release was enhanced by TNF-a (10 ng/ml) in all groups to a similar extent (normal, 124 ± 10; SS, 137 ± 25; SD, 164 ± 17; SR, 103 ± 9 pg/ml). Dexamethasone (10 ÿ6 mol/l) inhibited TNF-aeinduced GM-CSF production in all subject groups according to the in vivo efficacy of glucocorticoids in each subject group. Dexamethasone produced a similar level of reduction in GM-CSF release in both the normal (66% ± 4%) and the SS groups (64% ± 6%). This reduction was significantly less in the SD group (40% ± 5%; P <.01 vs normal and P <.05 vs SS) and in the SR group (29% ± 5%; P <.0001 vs normal and P <.001 vs SS; Fig 1, A). GR expression in PBMCs To determine whether the failure to respond to glucocorticoids was caused by a reduced expression of GR, we measured the levels of GR in PBMCs isolated from normal subjects and SS, SD, and SR patients by Western blotting. A wide variation in GR expression was observed between subjects and was similar to that previously measured by ligand binding assays. 8 There was no significant difference in the level of expression of GR in any of the patient groups examined when compared with b-actin expression (Fig 1, B). The antibody used detects both GRa and GRb isoforms; however, it was unclear whether the lower of 2 bands was GRb or the B-form of GRa. 16 Importantly, the expression of neither band was enhanced in the SD and SR groups overall compared with normal and SS subjects (Fig 1, B). GR nuclear localization We did not see any obvious differences between the responses of lymphocytes and monocytes when the methodology was being optimized. Therefore, we decided to use unfractionated human PBMC separated from whole blood. This finding agrees with previous studies that show that steroid resistance is seen in both monocytes and T cells in these patients. 5 PBMCs were treated with or without dexamethasone (10 ÿ6 mol/l) for 6 hours, and GR nuclear translocation was measured by immunofluorescence and confocal microscopy. At baseline, cells from all groups showed very low levels of nuclear staining (normal, 5.8 ± 1; SS, 7 ± 1.3; SD, 5 ± 0.6; SR, 4.2 ± 0.7). Cells isolated from normal (70 ± 2; n = 10) and SS (63 ± 3%; n = 6) subjects showed marked nuclear translocation of GR at 6 hours. In contrast, the SD (50% ± 6%; n = 11; P <.05) and SR (36% ± 7%; n = 9; P <.01) groups showed significantly less nuclear translocation compared with normal subjects (Fig 1, C, D). At shorter time points (60 minutes), there was complete (;100%) GR nuclear translocation in the normal and SS groups and a concomitant reduction in the SD and SR groups. The reduction seen over time was caused by either GR recycling or degradation. 16 Six hours was chosen as the time point for further examination because the effects on histone acetylation were maximal. Histone acetylation induced by dexamethasone in PBMCs As a readout of GR function, we examined the ability of dexamethasone to induce histone acetylation. PBMCs isolated from normal (68% ± 3%; n = 10) and SS (55% ± 6%; n = 6) subjects showed a marked increase in the number of nuclei staining positive for acetylated histone H4 as measured by immunocytochemistry after dexamethasone treatment (10 ÿ6 mol/l; Fig 2, A, B). In contrast, reduced histone acetylation was seen in both the SD (35% ± 6%; n = 11; P <.001 vs normal) and SR (33% ± 4%; n = 9; P <.001 vs normal) groups (Fig 2, A, B). PBMCs treated in the same manner but not exposed to dexamethasone all showed low levels of nuclear staining (normal, 7.4% ± 1%; SS, 9% ± 1%; SD, 8% ± 1%; SR, 9% ± 2%). PBMCs isolated from normal (3.2 ± 0.4-fold increase; n = 10) and SS (2.7 ± 0.6-fold increase; n = 6) subjects showed a marked increase in the ratio of stimulated versus nonstimulated histone acetylation. In contrast, there was significantly less 3 H-acetate incorporation seen in both the SD (2.0 ± 0.3-fold increase; n = 11) and SR (2.0 ± 0.3-fold increase; n = 9; P <.05) groups (Fig 2, C). Specific targeting of histone H4 lysine residues by TNF-a and dexamethasone We determined the pattern of lysine acetylation after dexamethasone (10 ÿ6 mol/l) and TNF-a (10 ng/ml) stimulation and compared this with results from unstimulated cells. Dexamethasone predominantly targeted acetylation on histone H4 K5 and K16 in all subjects with little acetylation of K8 and K12 residues compared with unstimulated cells (Fig 2, D). However, TNF-a predominantly induced K8 and K12 acetylation in all groups (Fig 2, E). The SD and SR subjects had markedly reduced staining for K5 acetylation after dexamethasone treatment. This defect in K5 acetylation in SD and SR groups was specific for dexamethasone stimulation and did not result from a generalized defect in total acetylation, because TNF-a stimulation of K8 and K12 acetylation was normal. Relationship between GR nuclear translocation and histone acetylation There was a significant correlation between GR nuclear translocation and the increase in number of nuclei staining positive for acetylated histone H4 in PBMCs isolated from normal and SS subjects after dexamethasone treatment (r 2 = 0.38; P =.01; Fig 3, A). The SD and SR groups, however, demonstrated 2 populations: 1 that had a similar relationship between nuclear GR and acetylated H4 (group 1, r 2 = 0.75; P =.0001) and a second (group 2, r 2 = 0.47; NS) that showed little histone acetylation even after high levels of GR nuclear translocation (mean GR percent positive nuclei, group 2, was 62% ± 7%, vs 70% ± 2% for normal and 63 ± 3 for SS). The 95% CIs for the 2 SD and SR groups did not overlap.

4 J ALLERGY CLIN IMMUNOL VOLUME 113, NUMBER 6 Matthews et al 1103 FIG 1. Glucocorticoid receptor function and expression in PBMCs from normal subjects (N) and SS, SD, and SR asthmatic patients. a, Inhibition of TNF-aeinduced GM-CSF release from PBMCs by dexamethasone (10 ÿ6 mol/l). Data are expressed as means ± SEMs. n $ 6 in each group. *P <.01 and P <.001 vs N and àp <.05 and P <.01 vs SS. b, Representative Western blot analysis of GRa expression in PBMCs from N, SS, SD, and SR subjects. b-actin expression was used to control for protein loading. Lower panel shows quantitation of GRa/b-actin expression. Bars represent means ± SEMs. n $ 6 in each group. c, Representative immunocytochemical analysis of nuclear translocation of GR in PBMCs from N, SS, SD, and SR subjects after dexamethasone treatment (10 ÿ6 mol/l, 6 hours). d, Graphical representation of the results shown in panel c. The number of nuclei staining positive for GR was compared in each group. The number of nuclei staining positive at baseline (b) is included. Data are expressed as means ± SEMs. kp <.05 vs N. n $ 6 in each group. s, Stimulated. Decreased acetylation of lysine 5 in histone H4 in PBMCs from group 2 insensitive subjects To confirm the specificity of this inability to induce histone acetylation and to indicate whether changes in GR-associated histone acetyltransferase (HAT) or histone deactylases (HDACs) were responsible for altered K5 acetylation, PBMCs were treated with the histone deacetylase inhibitor TSA. TSA (100 ng/ml, 6 hours) induced similar levels of K8 and K12 acetylation in all subjects. However, the SD and SR groups had reduced K5 acetylation compared with the normal and SS subjects. When the SD and SR groups were subdivided into group 1 and group 2 patterns, we found that the reduced acetylation of K5 was specific for group 2 subjects (P <.001;Fig 3, B). In contrast, group 1 patients still maintained the ability to acetylate histone H4 K5. There was also a significant reduction in K16 acetylation after TSA treatment in SD and SR subjects compared with normal (P <.05) and SS subjects (P <.01). As with the defect in K5 acetylation, this was specific to group 2 subjects (P <.001 compared with group 1; data not shown). Inhibition of TNF-aeinduced histone acetylation by dexamethasone We have shown a reduced ability of dexamethasone to induce K5 histone acetylation in the SD and SR groups. However, many of the effects of glucocorticoids are caused by transrepression and suppression of inflammatory gene expression rather than transactivation. Therefore, we examined the effect of dexamethasone on TNFaeinduced histone acetylation in these patients. TNF-a induced similar levels of histone acetylation activity in each patient group (normal, 2.7 ± 0.6-fold increase; SS, 2.9 ± 0.8; SD, 2.0 ± 0.3; SR, 1.8 ± 0.4). However, the inhibition by dexamethasone (10 ÿ8 mol/l) was significantly reduced in the SR group (11% ± 8%; P <.05) compared with the normal (33% ± 6%), SS (49% ± 12%), and SD (24% ± 11%) groups (Fig 3, C). When SD and SR subjects were classified according to group 1 or group 2, group 1 subjects had a significantly

5 1104 Matthews et al J ALLERGY CLIN IMMUNOL JUNE 2004 FIG 2. Histone acetylation induced by 6 hours of treatment with dexamethasone (10 ÿ6 mol/l) or TNF-a (10 ng/ ml) in PBMCs isolated from normal (N), SS, SD, and SR subjects. a, Representative immunocytochemical analysis of panacetylated histone H4 expression in PBMCs isolated from N, SS, SD, and SR subjects. b, Graphical representation of the results in panel a. The numbers of nuclei staining positive for panacetylated histone H4 were compared in each group and are presented as means ± SEMs. n $ 6 subjects in each group. *P <.01 vs normal. P <.05 vs SS. c, Histone acetylation activity was measured in these cells after dexamethasone treatment and presented as the ratio to the amount of 3 H-acetate incorporation in resting nonstimulated cells. Data are expressed as means ± SEMs. n $ 6 for each group. àp <.05 vs N. d, Specificity of histone H4 lysine acetylation after dexamethasone treatment in N, SS, SD, and SR subjects. Acetylation was determined by nuclear staining, and results are expressed as means ± SEMs of at least 6 subjects in each group. e, Specificity of histone H4 lysine acetylation after TNF-a treatment in N, SS, SD, and SR subjects. reduced amount of inhibition (7.3% ± 7%) compared with group 2 subjects (39% ± 11%), who had a degree of inhibition similar to that of normal subjects (Fig 3, D). These data suggest that the defect in dexamethasoneinduced K5 acetylation seen in group 2 subjects is distinct from the inability to repress TNF-aeinduced histone acetylation seen in group 1 subjects. Therefore, we reassessed the ability of dexamethasone (10 ÿ8 mol/l) to suppress TNF-aeinduced GM-CSF release in the steroidinsensitive groups according to their subdivision as group 1 and group 2 patients. Both group 1 and group 2 subjects were less able to suppress GM-CSF release (group 1, 14% ± 4%; group 2, 14% ± 10%; P <.05) compared with normal (35% ± 5%) and steroid-sensitive subjects (26% ± 3%; Fig 3, E). Repeatability of measures Nine subjects (3 normal, 4 SD, 2 SR) were followed up with repeat testing at a mean time between tests of 10 ± 2 months. There was good agreement between repeated measurements in both the functional assays (inhibition of GM-CSF, P <.01; r 2 = 0.75) and immunohistochemistry (GR nuclear localization, r 2 = 0.68, P <.01; acetylated histone H4, r 2 = 0.96, P <.0001; K5, r 2 = 0.92, P <.0001), with all measurements falling within 2 standard deviations. Importantly, subjects found to fall within group 1 or group 2 remained within these groups on repeat testing (Fig 4, A). The defect in TSA-induced histone H4 K5 acetylation was also maintained over this period (Fig 4, B). DISCUSSION PBMCs from SD and SR patients had a reduced ability to suppress TNF-aeinduced GM-CSF release compared with normal and SS subjects. In most SD and SR subjects, this failure to attenuate GM-CSF release was correlated with a reduced ability of GR to translocate to the nucleus and induce histone acetylation in response to dexamethasone. These subjects were also unable to repress TNFaeinduced histone acetylation (group 1). In addition, we have also identified a distinct subgroup within SD and SR asthmatics that was unable to induce histone H4 acetylation despite normal levels of GR nuclear translocation (group 2). These subjects had a specific defect in that K5, and to a lesser extent, K16, showed reduced acetylation with dexamethasone or even with the nonselective HDAC inhibitor TSA. This finding suggests that

6 J ALLERGY CLIN IMMUNOL VOLUME 113, NUMBER 6 Matthews et al 1105 FIG 3. Relationship between GR nuclear translocation and histone acetylation after dexamethasone treatment (10 ÿ6 mol/l, 6 hours). a, The number of nuclei from PBMCs isolated from 10 normal (N), 6 SS, 11 SD, and 9 SR subjects staining positive for GR are plotted against the number of nuclei staining positive for acetylated histone H4. Regression lines with 95% confidence limits are plotted for N and SS subjects and for group 1 (Gp 1) SD and SR subjects and group 2 (Gp 2) SD and SR subjects. In most subjects, the degree of histone acetylation correlates with GR nuclear translocation (group 1). Group 2 SD and SR subjects show little histone acetylation even after high levels of GR nuclear translocation. b, Decreased acetylation of K5 in histone H4 in PBMCs from group 2 insensitive subjects. PBMCs were treated with TSA (100 ng/ml) for 6 hours or left untreated as a control, and cells with nuclei staining positive for acetylation on specific histone H4 lysine residues were measured by fluorescent immunocytochemistry and confocal microscopy. Data are expressed as means ± SEMs. c, Effect of dexamethasone (10 ÿ8 mol/l) on inhibition of TNF-aeinduced HAT activity in N, SS, SD, and SR subjects. Results are expressed as means ± SEMs. n = at least 4 in each group. *P <.05 vs normal and SS. d, Effect of dexamethasone (10 ÿ8 mol/l) on inhibition of TNF-aeinduced HAT activity in N, SS, group 1, and group 2 subjects. Results are expressed as means ± SEMs. n = at least 4 in each group. *P <.05 vs group 2. e, Inhibition of TNF-aeinduced GM-CSF release from PBMCs by dexamethasone (10 ÿ8 mol/l) in N, SS, group 1 steroid-insensitive subjects and group 2 steroid-insensitive subjects. Data are expressed as means ± SEMs. n $ 4 in each group. P <.05 vs normal and SS. these patients have a defective K5-selective HAT. In contrast with group 1 subjects, these subjects were able to repress TNF-aeinduced histone acetylation. Although it was not possible to distinguish group 1 and group 2 subjects clinically in baseline characteristics, we hypothesize that these data suggest that the inability to respond to glucocorticoid therapy is caused by at least 2 distinct molecular mechanisms, 1 involving a failure to transrepress and the other an inability to transactivate. Subjects who are unable to transactivate might be expected to have fewer glucocorticoid side effects if all side effects were caused by GR transactivation; however, not all side effects are driven by transactivation. 17 We did not formally assess side effect profiles in these patients. Larger trials are needed to characterize these groups of patients prospectively for severity and side effects. TNF-aeinduced GM-CSF could be released from either lymphocytes or monocytes, and we observed similar results for GR translocation in both lymphocytes and monocytes. These data suggest that steroid resistance is seen in both monocytes and T cells in these patients. 5 In the future, it would be interesting to examine each cell population individually to determine whether subtle differences between the cell types exist. Overall, inhibition of TNF-aeinduced GM-CSF release was not a good method of distinguishing the 2 groups, which suggests that both transactivation and transrepression are involved in the regulation of this cytokine in PBMCs and may explain why functionally these groups behave similarly. GM-CSF release has been reported to be under both transcriptional and posttranscriptional control, 18,19 and glucocorticoids may regulate GM-CSF production at both levels. Thus, GR may control GM-CSF expression by preventing NF-jBemediated HAT activity directly or by recruiting HDACs, 3,20 or by induction of the dual mitogen-activated protein kinase (MAPK) phosphatase 1 (MKP-1), 21,22 thereby affecting mrna stability. Increased gene transcription is associated with an increase in histone acetylation, whereas hypoacetylation

7 1106 Matthews et al J ALLERGY CLIN IMMUNOL JUNE 2004 FIG 4. Repeatability of measurements. Peripheral blood mononuclear cells were isolated from 9 subjects on average 10 months after initial testing and the relationship between panhistone H4 acetylation and GR nuclear translocation measured. a, Immunocytochemistry was measured in 3 normal (N) subjects, 3 group 1 (Gp 1) subjects, and 3 group 2 (Gp 2) subjects. Subjects remained within defined groups 1 and 2. b, After TSA treatment (100 ng/ml, 6 hours), K5 acetylation was measured in group 1 and group 2 subjects. *P <.05 vs group 1. is correlated with reduced transcription or gene silencing. 23 We have previously shown a distinct pattern of histone H4 acetylation in an epithelial cell line (A549) by IL-1b, which acetylated K8 and K12, whereas dexamethasone acetylated K5 and K16. 3 Dexamethasone at low concentrations inhibits histone acetylation induced by IL- 1b. 3 The data presented here confirm the specific targeting of H4 lysines by dexamethasone. We also show that TNFa in PBMCs, like IL-1b in A549 cells, induced acetylation of H4 K8 and K12 predominantly, whereas dexamethasone markedly acetylated K5 and K16 with no effect on K8 and K12. The transrepressive action of dexamethasone is thought to account for most of the anti-inflammatory actions of steroids. 24 Our data suggest that for most SD and SR subjects, a failure of GR to translocate to the nucleus results in reduced transrepression. Reduced GR nuclear translocation has been reported to be caused by excessive expression of cytokines (eg, IL-1a or TNF-a) 25 ; increased GR nitrosylation, preventing GR-hsp90 dissociation 26 ; or altered GR phosphorylation status. 27 We have previously demonstrated that phosphorylation of GR, mediated by p38 MAPK, is associated with reduced responsiveness to dexamethasone in PBMCs. 28 It is also possible that GRb expression may be elevated in these patients and that this may modify GR nuclear translocation. 5,11 Recently, Goleva et al 29 suggested that interactions between GR and signal transducers and activators of transcription 5 are important in the failure of GR to translocate into the nucleus and suppress proliferation in response to IL-2 stimulation in murine helper T (HT)e2 cells. Interestingly, this effect was blocked by p38 MAPK inhibition. Further investigation is required to determine which of these possible mechanisms are occurring. The clinical implication for these patients is that enhancing GR nuclear translocation with long-acting b-agonists, for example may improve steroid responsiveness in these group 1 patients but not in group 2 patients. However, in a subset of these patients (group 2), repression of TNF-aeinduced histone acetylation is not reduced at relatively low concentrations of dexamethasone TABLE I. Clinical characteristics of patients Normal SS SD SR N Age (y) 31 ± 2 37 ± 6 33 ± 4 39 ± 6 Sex (M/F) 6/4 2/4 3/8 1/8 Atopy Baseline FEV 1 98 ± 2 70 ± 8 60 ± 7 51 ± 5 (% predicted) Prednisolone treatment (% improvement) ND 43 ± 11 ND ÿ5.6 ± 3 Albuterol (lg/d) ± ± ± 204 Bronchodilator response (% increase in FEV 1 ) ND 20 ± 3 25 ± 5 51 ± 9 Inhaled corticosteroids (lg/d) ± ± ± 308 Oral prednisolone (mg/d) ± ± 1 ND, Not determined. (10 ÿ8 mol/l) and is similar to that seen in normal subjects and SS patients. We postulate, therefore, that gene induction by glucocorticoids may play an important role in the control of inflammation and symptoms in SD and SR subjects with severe asthma. Indeed, Hawrylowicz 30 reported that dexamethasone-induced IL-10 release is decreased in cells from patients with SR asthma and proposed that failure to induce this anti-inflammatory cytokine was responsible for the failure to respond to steroids. In this report, we have focussed on the ability of dexamethasone to stimulate histone acetylation. We have measured histone acetylation as a marker of GR-induced transactivation. We can infer from these data that some subjects with steroid resistance have a reduced response to steroids, possibly because of a failure of steroids to induce specific acetylation of histone residues and subsequently switch on anti-inflammatory gene expression. Although it

8 J ALLERGY CLIN IMMUNOL VOLUME 113, NUMBER 6 Matthews et al 1107 is not known what GR-inducible genes are critical for the anti-inflammatory effect of glucocorticoids to work, several genes are increased by glucocorticoids, including inhibitor of jba, MKP-1, secretory leukocyte inhibitor protein, and IL-10. 5,6 Our data would suggest that attenuated induction of histone acetylation by dexamethasone would result in reduced expression of anti-inflammatory genes and, subsequently, enhanced inflammation. This is the first study that we are aware of to use immunocytochemistry to demonstrate the degree of nuclear localization of GR in SR and SD asthma. Previous studies, largely based on ligand binding experiments, have not demonstrated any difference in nuclear translocation in steroid-insensitive asthma. However, a reduction in the numbers of nuclear translocated GR available for DNA interaction has been measured in these patients, and this is in agreement with the current study. 9 The exact mechanism of GR-mediated repression is unknown. However, the failure to respond to glucocorticoids in these subjects has demonstrated that the ability to acetylate histone H4 K5, and to a lesser extent K16, is likely to be an important mechanism of glucocorticoid action in a subgroup of glucocorticoid-insensitive subjects. The identity of the HATs that regulate K5 acetylation in these cells has not yet been elucidated but may include SRC-1, Tip60, or CBP. Novel pharmacologic targeting of specific histone acetylation sites has potential therapeutic value for inflammatory diseases and treatment of glucocorticoid insensitivity. 31 These data suggest that some subjects with steroid resistance have a reduced response to steroids because of a failure of steroids to switch on anti-inflammatory gene expression, rather than a defect in switching off inflammatory genes. Our data can explain an inability to induce transactivation in CR patients but not transrepression in all patients. Group 1 patients are clearly unable to repress TNF-aeinduced histone acetylation and GM-CSF release because of a failure to translocate GR into the nucleus, where it is required to affect gene transcription. However, group 2 patients are clearly able to repress GM-CSF release. There may be a possible role of transactivation in GM-CSF repression after induction of MKP-1, for example, and alterations in mrna stability induced by changes in p38 MAPK activity, 21,22 but this is unlikely to account for all of the repressive effects of dexamethasone in this study. Dexamethasone was unable to suppress TNF-aeinduced histone acetylation in group 2 patients. It is possible that, examining whole cell activation, we may miss important events occurring at the GM-CSF promoter that show repression. Alternatively, we can hypothesize that the event that dexamethasone targets in group 2 patients is downstream of NF-jB activation and activation of histone acetylation. Nissen and Yamamoto 32 have previously reported that dexamethasone can reduce RNA polymerase II phosphorylation and thereby attenuate gene expression. In summary, the ability of GR to translocate to the nucleus is limiting in most SD and SR asthmatics, and further studies to examine the mechanism of this defect are needed. Drugs that can enhance GR nuclear translocation, such as long-acting b-agonists, may be of benefit in these subjects. In contrast, patients with reduced histone acetylation activity induced by dexamethasone may require targeting of specific GR-associated HAT activities to restore glucocorticoid function. We would like to thank Donald Payne and Andrew Bush, Royal Brompton Hospital, for help with clinical characterization of some patients. REFERENCES 1. Barnes PJ. Inhaled glucocorticoids for asthma. N Engl J Med 1995;332: Karin M. New twists in gene regulation by glucocorticoid receptor: is DNA binding dispensable? Cell 1998;93: Ito K, Barnes PJ, Adcock IM. Glucocorticoid receptor recruitment of histone deacetylase 2 inhibits interleukin-1beta-induced histone H4 acetylation on lysines 8 and 12. Mol Cell Biol 2000;20: Lambert JR, Nordeen SK. Steroid-selective initiation of chromatin remodeling and transcriptional activation of the mouse mammary tumor virus promoter is controlled by the site of promoter integration. J Biol Chem 1998;273: Leung DY, Bloom JW. Update on glucocorticoid action and resistance. J Allergy Clin Immunol 2003;111: Adcock IM, Lane SJ. Corticosteroid-insensitive asthma: molecular mechanisms. J Endocrinol 2003;178: Sher ER, Leung DY, Surs W, Kam JC, Zieg G, Kamada AK, et al. Steroid-resistant asthma: cellular mechanisms contributing to inadequate response to glucocorticoid therapy. J Clin Invest 1994;93: Kam JC, Szefler SJ, Surs W, Sher ER, Leung DY. Combination IL-2 and IL-4 reduces glucocorticoid receptor-binding affinity and T cell response to glucocorticoids. J Immunol 1993;151: Adcock IM, Lane SJ, Brown CR, Peters MJ, Lee TH, Barnes PJ. Differences in binding of glucocorticoid receptor to DNA in steroidresistant asthma. J Immunol 1995;154: Adcock IM, Lane SJ, Brown CR, Lee TH, Barnes PJ. Abnormal glucocorticoid receptor-activator protein 1 interaction in steroid-resistant asthma. J Exp Med 1995;182: Leung DY, Hamid Q, Vottero A, Szefler SJ, Surs W, Minshall E, et al. Association of glucocorticoid insensitivity with increased expression of glucocorticoid receptor beta. J Exp Med 1997;186: Seldon PM, Barnes PJ, Meja K, Giembycz MA. Suppression of lipopolysaccharide-induced tumor necrosis factor-alpha generation from human peripheral blood monocytes by inhibitors of phosphodiesterase 4: interaction with stimulants of adenylyl cyclase. Mol Pharmacol 1995;48: Lane SJ, Adcock IM, Richards D, Hawrylowicz C, Barnes PJ, Lee TH. Corticosteroid-resistant bronchial asthma is associated with increased c-fos expression in monocytes and T lymphocytes. J Clin Invest 1998; 102: Adcock IM, Gilbey T, Gelder CM, Chung KF, Barnes PJ. Glucocorticoid receptor localization in normal and asthmatic lung. Am J Respir Crit Care Med 1996;154: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1: Yudt MR, Cidlowski JA. The glucocorticoid receptor: coding a diversity of proteins and responses through a single gene. Mol Endocrinol 2002; 16: Schacke H, Docke WD, Asadullah K. Mechanisms involved in the side effects of glucocorticoids. Pharmacol Ther 2002;96: Bergmann MW, Staples KJ, Smith SJ, Barnes PJ, Newton R. Glucocorticoid inhibition of GM-CSF from T cells is independent of control by NF-jB and CLE0. Am J Respir Cell Mol Biol 2004;30: Smith PJ, Cousins DJ, Jee YK, Staynov DZ, Lee TH, Lavender P. Suppression of granulocyte-macrophage colony-stimulating factor

9 1108 Matthews et al J ALLERGY CLIN IMMUNOL JUNE 2004 expression by glucocorticoids involves inhibition of enhancer function by the glucocorticoid receptor binding to composite NF-AT/activator protein-1 elements. J Immunol 2001;167: Ito K, Jazrawi E, Cosio B, Barnes PJ, Adcock IM. p65-activated histone acetyltransferase activity is repressed by glucocorticoids: mifepristone fails to recruit HDAC2 to the p65-hat complex. J Biol Chem 2001;276: Lasa M, Abraham SM, Boucheron C, Saklatvala J, Clark AR. Dexamethasone causes sustained expression of mitogen-activated protein kinase (MAPK) phosphatase 1 and phosphatase-mediated inhibition of MAPK p38. Mol Cell Biol 2002;22: Kassel O, Sancono A, Kratzschmar J, Kreft B, Stassen M, Cato AC. Glucocorticoids inhibit MAP kinase via increased expression and decreased degradation of MKP-1. EMBO J 2001;20: Imhof A, Wolffe AP. Transcription: gene control by targeted histone acetylation. Curr Biol 1998;8:R Barnes PJ, Adcock IM. How do corticosteroids work in asthma? Ann Intern Med 2003;139: Pariante CM, Pearce BD, Pisell TL, Sanchez CI, Po C, Su C, et al. The proinflammatory cytokine, interleukin-1alpha, reduces glucocorticoid receptor translocation and function. Endocrinology 1999;140: Galigniana MD, Piwien-Pilipuk G, Assreuy J. Inhibition of glucocorticoid receptor binding by nitric oxide. Mol Pharmacol 1999;55: Munck A, Mendel DB, Smith LI, Orti E. Glucocorticoid receptors and actions. Am Rev Respir Dis 1990;141:S Irusen E, Matthews JG, Takahashi A, Barnes PJ, Chung KF, Adcock IM. p38 Mitogen-activated protein kinase-induced glucocorticoid receptor phosphorylation reduces its activity: role in steroid-insensitive asthma. J Allergy Clin Immunol 2002;109: Goleva E, Kisich KO, Leung DY. A role for STAT5 in the pathogenesis of IL-2-induced glucocorticoid resistance. J Immunol 2002;169: Hawrylowicz C, Richards D, Loke TK, Corrigan C, Lee T. A defect in corticosteroid-induced IL-10 production in T lymphocytes from corticosteroid-resistant asthmatic patients. J Allergy Clin Immunol 2002;109: Turlais F, Hardcastle A, Rowlands M, Newbatt Y, Bannister A, Kouzarides T, et al. High-throughput screening for identification of small molecule inhibitors of histone acetyltransferases using scintillating microplates (FlashPlate). Anal Biochem 2001;298: Nissen RM, Yamamoto KR. The glucocorticoid receptor inhibits NFkappaB by interfering with serine-2 phosphorylation of the RNA polymerase II carboxy-terminal domain. Genes Dev 2000;14: