Glucocorticoid receptor expression in human bronchial ... - CiteSeerX

Research Paper Mediators of Inflammation, 7, 275–281 (1998)

PREVIOUSLY , w e found th at in flam m atory m ediato rs m odulate d th e n um ber and bin din g affin ity o f glucocortico id r ecep tors (GR) in h um an bro nch ia l ep ith e lial cell lin es . In th is s tudy w e in ves tigate d w h e th er s m o kin g an d ch r on ic o bs tructive p ulm onary dis e as e (COPD), bo th ch aracterized by air w ay in flam m ation w ith in cr eas ed lev els o f in flam m ator y m ediators , affe ct GR ch ar acte ris tics in cultur ed h um an bro nch ial ep ith e lial cells (HBEC). A s tatis tically s ignificant diffe r en ce w as foun d be tw ee n th e dis s ociation co ns tan t (K d ) values in HBEC fr om s m o king (K d = 0.98 ± 0.08 n M; n = 6) an d no ns m o kin g co ntr ols (Kd = 0.76 ± 0.10 n M, P = 0.03; n = 5), but n o s ign ifican t diffe r en ce w as foun d betw e en th e m ean n um be r of bin din g s ites . Our r es ults ar e th e firs t in dicatio n th at cultur ed HBEC fr o m s m okers p os s e s s GR w ith a low e r bin din g affin ity. Th is m ay re s ult fr o m th e in flam m ation found in th e airw ays fr om s m o ke rs . Furth er m or e, th es e r es ults p r ovide fur th er e viden ce th at th e bro nch ia l ep ith e lium m ay be an actual tar ge t for in h aled glucocor ticoid th e rap y.

Glucocorticoid receptor expression in human bronchial epithelial cells: effects of smoking and COPD

Key w o r ds : Gluc ocorticoid rec eptors, Human bronchial epithelial ce lls, Chronic obstructive pulmonary disease, Smoking

CA

Introduction The bronchial epithelium has long be en re garde d as a pas sive barrier to protect the underlying tissue .1 Now, bronchial epithelial cells are also know n to play an ac tive role in airw ay inflammation. The y are able to produc e various inflammatory me diators including cytokines and e icosanoids.2 ,3 Bronchial epithe lial c ells re cove re d from asthmatic patients show incre ase d ge ne ex pre ssion for GM-CSF, IL-6 and IL-8. 3 Furthe rmore , incre ase d ex pre ssion of endothelin-1, TNF-a , IL-1ß and MCP-1 has be en demonstrate d.4 ,5 Gluc oc orticoids are w ide ly used in the tre atment of inflammatory pulmonary dise ase s, e.g. bronchial asthma and chronic obstruc tive pulmonary disease (COPD) 6 Rec ent studie s show ed that glucoc ortic oids are able to inhibit the re le as e of some bronchial epithe lial ce ll-derived cytokines.3 ,7 These effe cts of gluc ocorticoids combined w ith the above-mentione d role of bronchial epithelial cells in airw ay inflammation, sugge st that glucocorticoid therapy may suppre ss airw ay inflammation, at least partially, by modulating the func tion of bronchial ep ithe lial cells. This hypothesis is supported by the observation that the greater part of inhale d gluc oc ortic oids pre c ipitate

on the e pithe lium of the larger airw ays.8 A pre re quisite for ste roid re sponsivene ss is the pre senc e of spe cific gluc oc ortic oid re c eptors (GR). In a re ce nt study w e ide ntifie d and charac te rize d spe cific GR in tw o SV-40/ade novirus-transforme d human bronchial e pithe lial c ell line s, BEAS 2B and BEAS S6.9 We also demonstrate d that inflammatory me diators such as IL-1ß and LPS may modulate the number and the binding affinity of GR in BEAS 2B ce lls. Similar re sults w e re found by othe r investigators in human T c ells, othe r c ell lines and murine macrophage s.10 –1 2 The se re sults indicate that inflammatory processes may influe nce the re spons e of the bronchial ep ithe lium to glucoc ortic oid the rap y via loc ally produce d c ytokine s. Here , as an ex te nsion of our findings in c ell line s, w e re p ort on the ide ntific ation and charac te rization of spec ific GR in c ultured human bronchial epithelial ce lls (HBEC). From the above-me ntione d in vitro findings conc erning the effec ts of inflammatory me diators on GR, it may be ex pecte d that inflammation in v ivo may also affec t GR characte ristics. There fore , w e also studied w he the r in c ulture d HBEC from smokers and COPD patie nts, in w hom gene ral inflammation has be e n desc ribe d,1 3 the number or binding affinity of GR is altere d.

0962-9351/98/040275-07 $9.00 © 1998 Carfax Publishing Ltd

275

M. M. Verheggen,1,2 P. W. C. Adriaansen-Soeting,1 C. A. Berrevoets,3 P. Th. W. van Hal,1,2 A. O. Brinkmann,3 H. C. Hoogsteden2 and M. A. Versnel1, CA Departments of 1Immunology; 2Pulmonary Medicine; and 3Endocrinology and Reproduction, Erasmus University and University Hospital Rotterdam-Dijkzigt, PO Box 1738, 3000 DR Rotterdam, The Netherlands

Corresponding Author Te l: (+31) 10 408808 6 Fax : (+31) 10 4367601 Email: versnel @immu.fgg.eur.nl

M. M. Ve rhe g g e n et al. Table 1. Clinical characteristics of subject groups Subjects

Nonsmoking controls Smoking controls Nonsmoking COPD patients Smoking COPD patients

Number

Age (years)

5 6 3 5

71 ± 2 59 ± 3 67 ± 1 60 ± 3

Smoking (pack-years)

³ ³

0 35 0 35

FEV (% of predicted) 87 ± 6 90 ± 4 68 ± 6 52 ± 3

Data are shown as mean ± SEM.

Materials and Methods Subjects A total of 19 subje cts w ere entere d into the study (Table 1). Of the subjec ts eight fulfilled ATS c riteria for COPD.1 4 COPD patie nts w ere nonallerg ic w ith complaints of chronic c ough and sputum production. Chronic cough means that c oughing occ urs on most days during at least three conse cutive months in tw o cons ecutive years. One -se cond forced ex piratory volume (FEV1 ) w as me as ure d in all subje cts be fore entry into the study. COPD patie nts suffere d from major airw ay obstruction, that is, the FEV1 w as < 75% of pre dicte d. No re versibility of airw ay obstruc tion in these e ight patients w as obtain ed after inhalation of 0.5 mg te rbutaline . Controls (n = 11) w e re subjec ts w ho de nied any symptoms of asthma or COPD, and had normal lung function parame te rs. The COPD and control group c ontain ed five and six smoking individuals, re spec tive ly. From the nonsmoking c ontrols one individual stopped smoking 12 years ago. Smoking individuals w ere current smoke rs w ith a smoking history of at least 35 pack-ye ars.

Isolation and culture conditions of HBEC Bronchial tissue w as obtaine d from patie nts undergoing surgery for lung cance r and use d immediately for c ulture of HBEC by a c ell c ulture method desc ribed pre viously.1 5 Briefly, bronchial tissue distant from the tumour w as c ut into pie c es and inc ubate d e ither overnight at 4°C or 1 h at 37°C in 0.1% prote ase XIV (Sigma, St Louis, MO). Subsequently, e pithe lial cells w ere ge ntly sc rape d from the tissue samples, w as he d tw ice in c ulture medium and plated onto 35-mm dishes at a dens ity of 2.5 3 10 5 ce lls/dish. HBEC w e re c ulture d in a 1:1 mix ture of Dulbec co’s modified Eagle’s me dium and Ham’s F12 (DMEM/F12)(GIBCO, Paisle y, UK), w ith supple ments as de scribed pre viously.15 Ce lls w ere charac te rized as epithe lial c ells by immunofluore sce nce staining using a mouse monoclonal antib ody dire cte d against a number of human cytokeratins (CK-1; DAKOpatts, Glostrup, Denmark). At least 99% of the isolate d c ells staine d positive for cytoke ratin. 276

Mediators of Inflammation · Vol 7 · 1998

When dishes w ere c onfluent, HBEC w ere passage d to 75 c m 2 flasks and use d for ex perime nts after confluenc e. Tw enty-four hours be fore performing GR binding ex pe rime nts the me dium w as re place d by a basal me dium of DMEM/F12 (1:1) w ith pe nicillin G sodium and streptomycin sulphate , but w ithout hydrocortisone or othe r supplements to pre ve nt influenc e of endogenous ste roids on the number and affinity of GR.

Cell lines BEAS 2B is a SV-40/ade novirus transformed human bronchial epithelial ce ll line , w hich w as kindly provide d by Dr J. Le chne r (Inhalation Tox icology Research Institute , Albuquerque, NM).16 Ce lls w ere maintained in a complete keratinocyte grow th me dium (KGM) containin g bovine p ituitary ex trac t, EGF, pe nicillin G sodium and streptomycin sulphate (GIBCO). The COS-1 cell line w as used as a ne gative c ontrol for GR We ste rn blot ex perime nts. The COS-1 c ell line, derive d from the kidne y of a male adult Afric an gre en monke y (American Typ e Culture Colle ction, Rockville, MD) w as c ultured in DMEM w ith 5% FCS.

Preparation of cytosol Flasks containin g approx imately 10 3 10 6 c ells w ere w ashe d tw ice w ith c old phosphate buffere d saline (PBS) and HBEC w ere scrap ed in 0.5 ml of buffer A (40 mM Tris-HCI pH 7.4, 1 mM EDTA, 10% (v/v) glyce rol, 10 mM dithiothre itol, 10 mM Na 2 MoO4 ) supplemente d w ith 0.6 mM phe nylmethylsulphonyl fluoride (PMSF), 0.5 mM bacitracine and 0.5 mM leupep tin. The c ells w ere lyse d by fre ezing/thaw ing thre e times and the homogenate w as c entrifuged for g .1 7 The sup ernatant w as 10 min at 400 000 3 collecte d and store d at –80°C. Protein conc entrations w ere dete rmined ac cording to the method of Bradford 18 and range d from 1 to 2 m g/ m l. Cytosol pre parations w ere used for both We ste rn blotting and binding assays.

Glu co co rtico id re ce pto rs in b ro n ch ia l e pith e lia l ce lls

Western blotting Samples (20 m l cytosol containin g 1 m g/ m l prote in) w ere mix ed w ith 5 m l of 5 3 sample buffer (1 3 sample buffer = 50 mM Tris-HCI p H 6.8, 10% (v/v) glyce rol, 2% (w /v) SDS, 10 mM dithiothre itol, 0.001% (w /v) bromophe nol blue), boile d for 2.5 min and se parated on a 7% SDS-PAGE ge l. Afte r e le ctrophore sis the ge ls w ere blotted to nitroc ellulose membrane in blotting buffer (16.5 mM Tris-HCl, pH 8.3, 150 mM glycine and 20% v/v me thanol). The membranes w ere blocked for 1 h at RT w ith 1% block solution (Boe hringe r, Mannheim, Ge rmany) and incubated for 1 h at room te mpe rature w ith a c ommercially available GR p olyclonal rabbit antibody, no. 57 (PA1–511; Affinity Biore agents, Ne shanic Station, NJ), dilute d 1:250 in 0.5% block solution. Membranes w ere w ashe d tw ice w ith PBS/T w ee n 0.1% and tw ice w ith 0.5% block solution, then incubated w ith horse radish perox idase-c onjugated goat anti-rabbit-IgG (Sigma) for 1 h at RT. After 4 3 15 min w as hes (PBS/T w een 0.1%) prote ins w e re de te c te d w ith Boe hringer-Mannhe im’s Che milumine sc enc e Weste rn Blotting kit and membranes w ere ex pose d to X-ray film.

10% glycerol) and pellets w ere solubilized in Solue ne 350 (Packard, Meride n, CT). There after, sample s w e re transferre d to scintillation vials and radioactivity w as dete rmined by liquid sc intillation c ounting. Specific binding w as c alc ulated as the diffe re nc e be tw e en the totally bound radioactivity and the nonspe cific ally bound radioactivity at e ach gluc oc ortic oid c oncentration. Nons pecific binding w as calculate d from the aliquots containin g the 100-fold molar ex ce ss nonradioac tive dex ame thasone , assuming that nonspecific binding w as nons aturable and linearly re lated to the conc entration of fre e gluc oc orticoid.

Analysis Binding curves w e re construc te d from incre asing conc entrations of 3 H-dex amethasone . GR number and K d values w ere dete rmined by Scatchard analysis of these data. Data w ere ex p re sse d as me an fmol/mg prote in spe cific binding ± the standard error of the me an (SEM). The Wilc ox on Matche d-Pair Signe dRanks te st w as used to assess the e quality of GR number and Kd distributions in HBEC from patie nts w ith COPD and controls. A P-value of le ss than 0.05 w as considere d signific ant.

Steroids 3

H-labe lle d dex ame thasone (1,2,4,6,7 [ 3 H] dex amethas one ; spe cific activity 81 Ci/mmol) w as obtaine d from Ame rsham (Buckinghamshire , UK). Nonradioactive dex ame thasone w as kept in a stock solution of 2 3 10 – 3 M in e thanol (Duchefa bv. Haarlem, the Netherlands).

Results Expression of glucocorticoid receptor protein in normal HBEC Cytosolic fractions w e re isolated from HBEC, BEAS 2B and COS-1 ce lls, and GR leve ls w ere ex amined by Weste rn blotting. The GR polyclonal rabbit antibody

Binding assays GR numbers and K d values w ere dete rmined ac cording to established methods.1 9 Six serial doubling dilutions (50 m l) w ere pre pared in duplicate in PBS to final 3 H-labelled dex amethasone conce ntration s of 32, 16, 8, 4, 2, and 1 nmol/l, re spe ctively. For measureme nts of nonspec ific bindin g, paralle l dilutions of 3 H-labe lle d dex amethasone plus a 100-fold molar ex c ess of nonradioa ctive dex ame thasone w e re pre pared. To e ach of the 24 tube s 150 m l of the c ytosol pre paration w as added and the mix ture w as inc ubated overnight at 4°C. Binding e quilibrium w as re ached at all conc entrations after incubation overnight at 4°C. Subsequently, 160 m l aliquots of each inc ubation mix ture w ere transferred to albumin (0.1%) c oate d tubes, 750 m l of protamine dihydrochloride solution (0.5 mg/ml) w as added and tubes w ere ce ntrifuge d for 15 min at 4000 rpm. From the re maining inc ubation mix ture 20 m l w as used to establish the ex act conce ntration of 3 H-labelled dex amethas one. Tube s c ontainin g prote in pe lle ts w ere w ashe d thre e time s w ith 1 ml of inc ubation buffer (50 mM Tris-HCI pH 7.4, 1.5 mM EDTA, 1.5 mM DTT,

FIG. 1. Western blot of 20 m g protein illustrating immunoreactive GR levels in cultured human bronchial epithelial cells (BE) and in the cell line BEAS 2B. One representative experiment out of three is shown. The COS-1 cell line (20 m g) was used as a negative control. Mediators of Inflammation · Vol 7 · 1998

277

M. M. Ve rhe g g e n et al.

FIG. 2. (A) Binding curve of 3H-labelled dexamethasone for cultured HBEC. The specific binding is represented by the difference between the total and nonspecific binding. The data points represent mean values of duplicate determinations. (B) Scatchard plot of the specific binding of 3H-labelled dexamethasone to HBEC. On the ordinate is given the ratio of the number of specifically bound 3H-labelled dexamethasone (Bs) to the number of free 3H-labelled dexamethasone (F). One representative experiment out of 11 is shown. The Kd and the number of glucocorticoid binding sites per mg protein (R) were calculated using the negative inverse of the slope and the 3 intercept, respectively.

re c ognized a prominent band at ~ 97 kDa in the HBEC and BEAS 2B pre p aration s (Fig. 1). This e stimate d molecular mass is consis te nt w ith that re porte d pre viously for the human GR.20 In the GR ne gative ce ll line COS-1, no immunore active protein w as dete ctable. After de monstrating the pre se nce of GR protein, the numbe r of GR and K d values w e re studie d in cultured HBEC from controls (n = 11). Using e stablishe d me thods to ide ntify GR, w e could demonstrate spe cific binding of 3 H-labe lle d dex ame thasone by these ce lls. A typic al binding c urve for culture d HBEC is show n in Fig. 2A. After Scatchard analysis of the data, the line ar re gre ssion line obtaine d indicated a single class of GR (Fig. 2B).

GR binding in HBEC COPD patients, smokers and controls HBEC w e re is olated from bronchus tissue from e ight COPD patients and 11 c ontrols. The COPD and control group c ontain ed five and six smoking individuals, re spe ctive ly. Ce lls w ere cultured for one pas sage. There after, cytosols w ere pre p ared and the numbe r of gluc ocorticoid bindin g sites and Kd w e re dete rmine d. Results are pre sented in Table 2 and Fig. 3. A significant diffe re nc e w as found be tw een the Kd values in HBEC from smoking (Kd = 0.98 ± 0.08 nM) and nonsmoking controls (Kd = 0.76 ± 0.10 nM, P = 0.03), but no significant differe nc e w as found be tw e en the mean number of binding sites (70.5 ±

Table 2. Glucocorticoid binding in HBEC from smoking and nonsmoking COPD patients and controls

R (fmol/mg protein) Kd (nM)

Controls Nonsmoking (n = 5)

COPD patients Smoking (n = 6)

Nonsmoking (n = 3)

Smoking (n = 5)

87.2 ± 16.7 0.76 ± 0.10

70.5 ± 14.0 0.98 ± 0.08*

79.7 ± 22.6 0.74 ± 0.19

68.4 ± 12.8 0.71 ± 0.15

Data are shown as mean ± SEM. R: number of glucocorticoid binding sites; Kd : dissociation constant. *: significantly higher (P < 0.05) than in nonsmoking controls.

278



FIG. 3. Total number of GR (A) and Kd values (B) in cultures of HBEC from nonsmoking controls (n = 5), smoking controls (n = 6), nonsmoking COPD patients ( n = 3) and smoking COPD patients (n = 5). Data are shown as mean ± SEM. * P < 0.05 for smoking controls versus nonsmoking controls. sm: smoking.

14.0 and 87.2 ± 16.7 fmol/mg protein, re spec tively, P = 0.3). Furthermore , value s me as ure d in nonsmoking and smoking COPD patients (P = 0.6 and P = 0.6, re spective ly) did not differ signific antly. Nor w as any significant diffe re nce observe d in gluc oc orticoid binding site s or Kd be tw een nonsmoking COPD patients and nonsmoking controls (P = 0.6 and P = 0.6, re spec tively) or be tw e en smoking COPD patients and smoking c ontrols (P = 0.9 and P = 0.1, re spective ly).

Discussion In this study w e identified and characte rize d spec ific GR in primary c ulture s of HBEC from COPD patients and c ontrols. The pre senc e of GR prote in w as demonstrate d using Weste rn blot analysis and 3 H-dex ame thasone binding studies. A significant decre as e w as found in the binding affinity of GR in HBEC from smoking c ontrols compare d w ith nonsmoking controls, but no signific ant differe nc e w as found be tw e en the mean numbe r of spec ific glucocortic oid binding site s. Furthermore , values me as ure d in nons moking and smoking COPD patie nts did not differ signific antly. Nor w as any diffe re nce observe d be tw e en the mean numbe r of binding sites and the Kd values in HBEC from nonsmoking COPD patients and the value s me asured in nonsmoking controls or be tw e en the values meas ure d in smoking COPD patients and smoking c ontrols.

HBEC are c onside re d to play an important role in airw ay inflammation. 1 – 5 Inhaled glucocortic oids are used to suppre ss airw ay inflammation and, concomitantly, to improve clinic al paramete rs. The finding that HBEC possess func tional GR allow s us to hypothesize that the clinical re spons e to inhale d gluc ocorticoids, w hich mainly pre cipitate in the larger airw ays, re sults, at le ast partly, from the modulation of airw ay epithelial cell functions. From the K d value of the GR in HBEC observe d in our studies, w e ex pec t that effe ctive in vivo gluc ocortic oid conc entrations should be around 1 nM. In a re c ent study by Van den Bosch e t a l. 2 1 it w as show n that at least 90 min afte r inhalation of 1.6 mg budesonide, lung tissue conce ntrations ranged from 2.1 to 8.9 nM. There fore , w e can assume that the rapeutic al inhalation of gluc ocorticoids re sults in such conc entrations of gluc oc ortic oids in lung tissue that interaction w ith GR in HBEC w ill occur. In a pre vious study w e found that the SV-40/ade novirus transformed human bronchial e pithelial ce ll line BEAS 2B c ontain ed a higher numbe r of gluc oc orticoid binding site s compared w ith peripheral blood mononuclear ce lls (PBMC).8 The numbe r and quality of GR in targe t ce lls may dete rmine the ex te nt of gluc ocortic oid re sponsivene ss.22 HBEC w ere obse rved to contain le ss GR than the BEAS 2B c ell line (87.4 ± 16.5 and 370 ± 14 fmol/mg protein, re spe ctive ly). This differe nc e may be c ause d by the SV-40/ade novirus transformation. HBEC contain a re lative ly high numMediators of Inflammation · Vol 7 · 1998

279

M. M. Ve rhe g g e n et al.

be r of GR c omp ared w ith PBMC, w hich contain approx imately 30 fmol/mg protein (unpublishe d re sults). It w ould be intere sting to inve stigate w hether the se diffe re nc es in GR number re sult in differe nc es in the functional re sponse to gluc ocortic oids. As both epithelial ce lls and PBMC are able to produce IL-1ß, the inhibition of IL-1ß production by glucocorticoids may be a use ful paramete r to study this issue . Few data are available on the number of GR in other epithelial c ells. A human e pithelial duct ce ll line w as found to contain 83–92 fmol GR/mg prote in. 23 Pre viously, w e demonstrate d that the inflammatory me diators IL-1ß and LPS incre ase d the GR numbe r and K d in BEAS 2B ce lls.8 Similar re sults w ere found in othe r cell types.9 –1 1 These re sults suggest that inflammatory p roc esses may indirec tly influenc e the re sponse of bronchial epithelium to gluc oc orticoid therapy via production of cytokines by infiltratin g ce lls. It has be e n suggeste d that in COPD, bronchial inflammation is re sponsible for the de ve lopme nt of airw ay hyperreactivity, and chronic airflow limitation. 1 2 Cigarette smoking causes an inflammatory re action in the airw ays and c an le ad to the de ve lopme nt of COPD.1 3 We found a significantly de cre ase d binding affinity of the GR in smoking controls w hen compared w ith values found in nons moking c ontrols. These re sults indicate that the inflammation pre se nt in the airw ays of smoke rs may influenc e the binding affinity of GR re c eptors. The finding that in v ivo inflammation found in smokers affec ts GR characte ristic s in HBEC is in ac cordanc e w ith our above me ntione d findings in v itro . How ever, w e did not find any diffe re nc e in the binding affinity of the GR in smoking COPD patie nts w he n compared w ith values in nonsmoking COPD patients. A study by Linden e t a l. 13 show ed that obstructe d smoking COPD patients had signific antly low er c onc entrations of inflammatory c ells in the bronchial lavage fluid compare d w ith nonobstructe d smoking individuals. The re fore , it w as suggeste d that obstructe d and nonobstructe d smokers may diffe r in the type of airw ay inflammation pre sent. The p re senc e of low e r conce ntrations of inflammatory ce lls in obstruc te d smoking COPD patients in the study by Linden e t a l. may ex plain w hy in our study no differe nc es in GR binding affinity w ere found be tw ee n smoking and nonsmoking obstructe d COPD patie nts. Furthermore , w e did not find any diffe re nc e in the binding affinity of the GR in COPD patients w hen c ompared w ith values found in controls. Differe nt inflammatory mediators may be involved in COPD and in smoking and it re mains to be established w hich mediators ac tually alter GR characte ristics. The ac tual me chanisms by w hich inflammatory mediators, pre sent in the airw ays of smoking controls can induce a dec re ase in GR binding affinity in HBEC re main to be e stablishe d. Others have show n that altere d ex pre ssion of glucocorticoid-re gulate d 280


prote ins appe ars to be mediated via inte raction of the modulatory domain of the GR w ith transcriptional factors, such as AP-1. 2 4 – 2 6 Ove rex pre ssion of AP-1 inte rfere s w ith the function of the modulatory domain of the GR. Bec ause cytokines can induc e ele vated leve ls of AP-1, it has be en suggeste d that this may provide a plausible ex planation for the de cre ase d ligand binding affinity of nuclear GR for glucocortic oids found in T c ells afte r incubation w ith IL-2 and IL-4. 24 – 2 6 Pe rhaps, a similar ex planation could be given for the de cre ase d GR binding affinity in HBEC induc ed by inflammatory mediators in the airw ays of smoking c ontrols. Another ex planation for de cre ase d GR binding affinity c ould be a modulating effect of inflammatory mediators in the airw ays of smoking controls on the ex pre ssion and phosp horylation of he at shock proteins, w hich are associate d w ith the unliganded GR.27 ,28 Bacte rial p roduc ts and cytokines can re gulate the ex pre ssion and p hosphorylation of he at shock prote ins and this may modulate gluc ocortic oid binding to the GR.27 ,28 Further studies are ne cessary to clarify the ex act mechanism of the effects of these inflammatory mediators found in the airw ays of smoking controls on GR. We did not find any diffe re nc e in the number of GR be tw e en COPD patie nts and controls, nor be tw e en smoking and nonsmoking controls. How e ver, it cannot be ex clude d that differe nc es, pre sent be tw een the groups in vivo , disappeared during isolation and culture of these cells. We tried to clarify this is sue and te ste d cultures of HBEC at subsequent p assages. Similar numbers of GR and K d values w e re found in subse que nt pas sages. Othe rs have found that inc re as ed GR numbers and de cre ase d GR binding affinity in T cells from ste roid re sistant asthmatic patients re verted to normal after 48 h in culture.2 9 It w ould be inte re sting to study GR numbe r and affinity dire ctly on the bronchial tissue or direc tly after c ell isolation. How e ve r, performance of a 3 H-dex ame thasone binding assay, the most se nsitive me thod to dete rmine GR numbe r and affinity, direc tly after cell is olation is not possible, be cause of the low ce ll numbe r obtaine d. We are c urre ntly w orking on a me thod to quantify GR number in small ce ll sample s by flow cytome try using se veral GR spe cific antib odie s. To our know le dge, no other studie s have be en published on GR numbe r and K d values in HBEC from smoking and nonsmoking COPD patients and controls. Until now, most investigations have conc entrated on the analysis of GR in PBMC, c omparing asthmatics w ith he althy individuals.30 ,31 No diffe renc es in GR numbe r or GR binding affinity w ere found be tw ee n PBMC from as thmatic s and controls.23 ,2 9 Rece ntly, Kam e t a l. 10 show ed that PBMC from ste roid re sistant asthmatics had a significantly re duced GR binding affinity and an inc re as ed GR number w hen c ompare d w ith normal subjec ts. Others have claime d that the clinic al re sponse to gluc o-


cortic oids can not be ex plained by abnormal GR number or affinity.3 2 The y suggeste d that the ability of GR to bind to the ir DNA binding sites (GRE) is impaired. Further studies are ne ce ssary to clarify w hether the clinic al re sponse to inhale d gluc ocortic oids is re lated to the number and binding affinity of GR and w hich c ells are involved in this re sponse. As only a subgroup of COPD patients re sponds to inhaled gluc oc ortic oids, it is of intere st to look in re sistant patie nts for c orrelations be tw e en the clinical re sp onse to glucoc ortic oids and the number of GR and their K d in bronchial epithelial c ells.33 In c onclusion, w e de monstrated that c ulture d HBEC posse ss a single class of spe cific GR and that the binding affinity of GR in HBEC from smoking c ontrols w as significantly dec re as ed w hen c ompared w ith values found in nons moking c ontrols. These re sults provide furthe r evide nce that the bronchial epithelium may be an ac tual targe t for gluc ocorticoid therapy. Furthe rmore , our findings are the first indication that alte re d GR charac te ristics are pre se nt in c ulture d HBEC from smoke rs. We hyp othesize that this may re sult from the inflammation found in the airw ays from smoke rs.

References 1. Mattoli S, Masie ro M, Calabrò F, Me zze tti M, Ple bani M, Alle gra L. Eic osanoid re le ase from human b ronc hial e p ithe lial ce lls upon ex p osure to tolue ne diisocyanate in v itro . J Ce ll Phy s io l 1990; 142: 379 –385. 2. Salari H, Chan-Ye ung M. Re le ase of 15-hydrox yeicosate trae noic acid (15-HETE) and p ros taglandin E2 (PGE2 ) by culture d human bronchial e p ithe lial ce lls. Am J Re s pir Ce ll Mo l Bio l 1989; 1: 245 –280. 3. Marini M, Vittori E, Holle mborg J, Mattoli S. Ex p re ssion of the p ote nt inflammatory cytokine s , granulocyte -macrop hage -c olony -stimulating factor and inte rle ukin-6 and inte rle ukin-8, in bronc hial e p ithe lial c e lls of p atie nts w ith asthma. J Alle rg y Clin Im m u n o l 1992; 89 : 1001–1009. 4. Vittori E, Marini M, Fasoli A. Inc rease d ex p ress ion of e ndothe lin in bronchial e pithe lial ce lls of asth matic p atie nts and e ffe c ts of c orticos te roids. Am Re v R e s p ir Dis 1992; 146: 1320 –1325. 5. Be cker S, Quay J, Kore n HS, Haskill JS. Cons titutive and stimulate d MCP1, GROa , b , and g ex p re ss ion in human airw ay e pithe lium and bronchoalve olar mac rop hage s. Am J Phy s io l 1994; 266 (Lung Ce ll Mol Physiol 10): L278 – L286. 6. Corrigan C. Me chanism of glucoc or tic oid ac tion in asthma: too little , to o late . Clin Ex p Alle rg y 1992; 22: 315 – 317. 7. Sousa AR, Poston RN, Lane SJ, Nakhos te e n JA, Lee TH. De te ction of GMCSF in asthmatic bronchial e p ithe lium and dec rease b y inhale d corticos te roids. Am R e v Re s p ir Dis 1993; 147: 1557–1561. 8. Barne s PJ, Pederse n S. Efficac y and s afe ty of inhale d cortic oste roids in asthma. Am Re v Re s pir Dis 1993; 148: S1 –S26. 9. Ve rhe gge n MM, van Hal PThW, Adriaans e n-So e ting PWC, Goe nse BJA, Hoogste den HC, Brinkmann AO, Ve rsne l MA. Ex p res sion of glucocorticoid rec e p tors in human bronc hial e p ithe lial c e ll line s; modulation by IL-1b , TNF-a and LPS. Eu r Re s p ir J 1996; 9: 2036 –2043. 10. Kam JC, Sze f le r SJ, Surs W, Sh e r ER, Le ung DY M. Combination of IL-2 and IL-4 reduce s glucoc orticoid re c e p tor-binding affinity and T ce ll re sponse to glucoc orticoids. J Im m u n o l 1993; 151 : 3460 –3466. 11. Rakas z E, Gal A, Biró J, Balas G, Falus A. Modulation of glucoc ortic os te roid binding in human lymp hoid, mono c ytoid and he p atoma ce ll line s b y inflammatory c ytokine s inte rle ukin (IL)-1b , IL-6 and tumour ne cros is fac tor (TNF)-a . Sc a n d J Im m u n o l 1993; 37: 684 –689. 12. Salkow s ki CA, Voge l SN. INF-g me diate s inc rease d gluc oc ortic oid re ce p tor ex p ress ion in murine macrop hage s. J Im m u n o l 1992; 148: 2770 –2777. 13. Linde n M, Rasmuss e n JB, Piitulaine n E, e t a l. Airw ay inf lammation in smoke rs w ith nonob s truc tive and obstructive chronic bronchitis. Am

Re v R e s p ir Dis 1993; 148: 1226 –1232. 14. Ame ric an Thorac ic Soc iety. Standards for th e diagnosis and c are of p atie nts w ith c hronic obs truc tive p ulmonary dis e ase (COPD) and asthma. Am Re v Re s pir Dis 1987; 136: 225 –244. 15. Ve rhe gge n MM, de Bont HI, Adriaanse n-Soe ting PWC, e t a l. Ex p re ssion of lip oc ortins in human b ronc hial e p ithe lial c e lls: e ffe c ts of II-1ß, TNF-a , LPS and dex ame thasone . Me d In fla m 1996; 5: 210 –217. 16. Re dde l RR, Ke Y, Ge rw in BI, e t a l. Transformation of human bronc hial e p ithe lial ce lls by infe c tion w ith SV40 or ade nov irus -12 SV40 hybrid virus or trans fec tion via strontium p hos p hate cop re c ipitation w ith a p las mid c ontaining SV40 e arly region gene s . Ca n ce r R e s 1988; 48: 1904 –1909. 17. Be rre voe ts CA, Ve lds c hlote J, Mulde r E. Effe c ts of antiandroge ns on transformation and transc rip tion ac tivation of w ild-typ e and mutate d (LNCaP) androge n rec e p tors . J Ste ro id Bio ch e m Mo le c Bio l 1993; 6: 731–736. 18. Bradford MM. A rapid and s e nsitive me thod for th e quantific ation of mic rogram quantitie s of prote in utilizing th e p rinc iple of p rote in-dye binding. An n a l Bio ch e m 1976; 71 : 248 –254. 19. Ve ldscholte J, Voorhors t-Ogink MM, Bolt-de Vrie s J, Van Rooij HCJ, Trap man J, Mulde r E. Unusual s p e c ific ity of th e androge n rec e ptor in the human p ros tate tumor c e ll line LNCaP: high affinity for p rostage nic and e s troge nic s te roids. Bio c him e t Bio p hy s Ac ta 1990; 1052 : 187–194. 20. Cidlow ski JA, Be llingham DL, Pow e ll-Olive r FE, Lubahn DB, Sar M. Nove l antip e p tide antibodie s to the human gluc oc ortic oid re c e p tor: re cognition of multip le rece ptor forms in v itro and dis tinct localization of c ytop lasmic and nucle ar rec e p tors. Mo l En do cr in o l 1990; 90: 1427–1436. 21. Van de n Bosc h JMM, We s te rmann CJJ, Aumann J, Eds bä cker S, Tö nne ss on M, Se lroos O. Re lationship be tw e e n lung tis sue and blood c once ntrations of inhale d bude sonide . Bio ph a rm Dru g Dis po s 1993; 14: 455 –459. 22. Homo-De larc he F. Gluc oc ortic oid rec e p tors and ste roid s e ns itivity in normal and ne op las ic human lymphoid tiss ues: a rev iew. Ca n c e r Re s 1984; 44: 431–437. 23. Kurokaw a R, Hatake yama S, Kyakumoto S, Ota M. Cytosol gluc oc ortic oid re c e p tor in the ne op las tic e p ithe lial duc t c e ll line from human salivary gland (HSG). Bio ch e m In t 1986; 4: 671–679. 24. Yang-Ye n H-F, Chambard J-C, Sun Y-I, Sme al T, Sc hmidt TJ, Drouin J, Karin M. Transc rip tional inte rfe re nc e be tw e e n c -Jun and the gluc oc ortic oid re c e p tor: mutual inhibtion of DNA binding due to dire ct p rote in-p rote in inte raction. Ce ll 1990; 62: 1205 –1215. 25. Jonat C, Rahmsdorf HJ, Park K-K, e t a l. Antitumor p romotion and antiinf lammation: dow n-modulation of AP-1 (Fos /Jun) activity by gluc oc orticoid hormone . Ce ll 1990; 62: 1189 –1204. 26. Sc h u¨ le R, Rangarajan P, Klie w e r S, e t a l. Func tional antagonism be tw e e n onc op rote in c-Jun and the glucoc ortic oid re c e p tor. Ce ll 1990; 62: 1217–1226. 27. Finc ato G, Pole ntarutti N, Sic a A, Mantovani A, Colotta F. Ex p res sion of a he at-inducible gene of the HSP70 family in human mye lomono c ytic ce lls: re gulation by bac te rial p roduc ts and c ytokine s. Blo o d 1991; 77: 579 –586. 28. Satoh J, Kim SU. Cytokine s and grow th fac tors induc e HSP27 p hos phorylation in human astroc yte s. J Ne u ro p a tho l Ex p Ne u ro l 1995; 54: 504 –512. 29. Sh e r ER, Leung DY M, Surs W, e t a l. Ste roid-re sistant asthma. Cellular me c hanis ms c ontributing to inade quate res pons e to gluc oc ortic oid th e rap y. J Clin In ve s t 1994; 93: 33 –39. 30. Tsai BS, Watt G, Koe snadi K, Tow nle y RG. Lymp hocyte gluc oc ortic oid re c e p tors in asthmatic and c ontrol subje c ts . Clin Alle rg y 1984; 14: 363 –371. 31. Lane SJ, Lee TH. Gluc oc orticoid re c e p tor c harac te ris tics in mono c yte s of p atie nts w ith c orticoste roid-re s istant bronc hial asthma. Am Re v Re s pir Dis 1991; 143: 1020 –1024. 32. Adcoc k LM, Lane SJ, Brow n CR, Pe te rs MJ, Lee TH, Barne s PJ. Diffe re nc e s in binding of glucoc ortic oid rec e p tor to DNA in ste roid-re s istant asthma. J Im m u n o l 1995; 154: 3500 –3505. 33. Callahan C, Dittus RS, Katz BP. Oral cortic os te roid the rap y for patie nts w ith stable c hronic obs truc tive p ulmonary dise ase : a me ta-analysis. An n In te rn Me d 1991; 114: 216 – 223. ACKNOWLEDGEMENTS. We grate fully acknow le dge th e continuou s s upport of Profe s sor Dr R. Be nne r. Dr J. Le c hne r kindly p rovide d the BEAS 2B ce ll line . We th ank Mr T. M. van Os for e x c e lle nt photograp hic ass istanc e and Ms P. C. Asse ms for se c retarial assis tanc e . This study w as sup p or te d by the Ne th e rlands Asthma Foundation (90.44).

Received 20 April 1998; accepted 29 April 1998


281