Int. J. Cancer: 67,357-364 (1996) 0 1996 Wiley-L,iss, Inc. Publication of the International Union Against Cancer Publication de I‘Union Internationale Contre le Cancer STEROID-HORMONE RECEPTORS IN CELL LINES AND TUMOR BIOPSIES OF HUMAN LUNG CANCER Ulrich KAISER',^, Jurgen HOFMANN~, Margret SCHILLI~, Barbel WEGMANN~, Uwe KLoTzI, Steffen WEDEL’, Ellen WOLLMER~, Detlev BRANSCHEID? Adi F. GAZDAR~ and Klaus HAVEMAN” Arvind K. VIRMANI~, lPhilipps Universitat Marburg, Department of Internal Medicine, Marburg, Germany; 2Simmons Cancer Center, Universit)iof Texas Southwestern at Dallas, Dallas, Texas, USA; and 3GrosshansdorfHospital, Department of Thoracic Surgery, Grosshansdo~Germany. Female gender is a significant independent favorable prognostic factor in lung cancer. To study the possible role of sex hormones in lung cancer, the expression of sex-steroid receptors and the glucocorticoid receptor was investigated in 29 lung-cancer cell lines stemming from small-cell lung cancer (SCLC) and non-small-cell lung cancer (NSCLC) by means of immunocytochemistry, ligand-binding assays and RNA expression via polymerase chain reaction. In at least 2 methods of investigation, NSCLC cell lines showed a low expression of estrogen receptor in 6, progesterone receptor in 13 and androgen receptor in I2 out of 17 cases examined; sex-steroidreceptor expression was virtually absent in SCLC cell lines. The glucocorticoid receptor was expressed in all 29 cell lines studied. Additionally, 52 tumor samples from primary lung cancer were investigated for their receptor expression by means of immunohistochernistry. Among patients with primary lung-cancer sex-steroid-receptor expression in tumor biopsies was detected most frequently in female patients (in 69% of I6 cases, vs. 42% of 36 tumors from men) and in patients with adenocarcinoma. Further research will focus on these subgroups. lmmunohistology is a feasible method of studying steroid-receptorexpression in lung cancer. o 1996 Wiley-Lhs, Inc. Analyses of large clinical trials have shown that female gender is a significant independent favorable prognostic factor in small-cell lung cancer (Albain et al., 1990; Johnson et al., 1988). Further multivariate analysis of a German multicenter study comprising 766 patients revealed significance only for women under the age of 60 years (Wolf et al., 1991). Two studies of patients with unresectable non-small-cell lung cancer who were treated with chemotherapy also showed a longer survival rate for women (O’Connell et al., 1986; Finkelstein et al., 1986). To determine the influence and possible therapeutical value of sex hormones in lung cancer, we decided to study the expression of their receptors in lung-cancer cell lines and tumor biopsies, since these are a prerequisite for steroid action. There are basically 3 methods for the detection of steroid-hormone receptors: gene expression, binding assays and immunohistochemistry. The gene expression of receptors has not been investigated in lung cancer so far. Binding assays which use radioactive ligands in whole-cell or cytosolic assays are the standard method for studying receptor expression. Immunohistochemistry uses monoclonal antibodies and has been well established for the detection of estrogen receptors and progesterone receptors in breast cancer. Numerous studies have shown a close correlation between immunocytochemical methods and binding assays in mammary carcinoma (Charpin et UL, 1986). The diagnosis of lung cancer is made predominantly on small bronchoscopic biopsies or cytological specimens. Therefore binding assays for which large amounts of fresh tissue have to be provided cannot readily be performed. Immunohistochemistry may overcome this problem, since this method can be applied to small tissue-biopsy slides. Consequently, another aim of our study was to establish immunocytochemistry on lung-cancer cell lines as a new and additional method of studying receptor expression. Both the RNA expression of steroid receptors and the ligand-binding studies were compared with the results of immunocytochemistry. In addition, 52 samples of primary lung cancer were investigated for receptor expression by immunohistochemistry. MATERIAL AND METHODS Cell lines The cell lines used in this study have been established and characterized in the National Cancer Institute (Carney et al., 1985; Giard et al.. 1973), Dartmouth Medical School (Pettengill et al., 1980), the University of Uppsala (Bergh et al., 1985) and the University of Marburg (Beppler et al., 1987). They are listed in Table I. Assays were performed on 12 cell lines of small-cell lung cancer, 16 cell lines of non-small-cell lung cancer, and one cell line established from a malignant mesothelioma. As control, we used the breast-carcinoma cell line MCF-7 (Brooks et al., 1973) and the B-cell line B2077 (obtained from the National Cancer Institute) as cell lines with known receptor expression. Primary tissue Primary tumor tissue specimen were obtained from the Department of Thoracic Surgery at the Grosshansdorf Hospital, Germany (Dr. Branscheid). All tumor samples were derived from tumor resections and were stored in liquid nitrogen at -70°C. Diagnosis was made on representative slides; for confirmation, cryostat sections used in this study were stained with hematoxylin and eosin. As control tissues for immunoreaction, cryostat sections from receptor-positive breast cancer (estrogen receptor, progesterone receptor), normal prostate (androgen receptor) and lymph node (glucocorticoid receptor) were prepared. Out of 61 tumors, 52 were suitable for evaluation, 36 tumors from men and 16 tumors from women. Among the 52 cases, there were 18 squamous-cell carcinomas, 17 adenocarcinomas, 11 large-cell carcinomas, 3 SCLC, 2 carcinoids and 1 bronchoalveolar-cell carcinoma. For distribution of pathological stage and histological grade, see Table V. Culture media and buffer systems Cells were maintained in RPMI-1640 medium (GIBCO, Grand Island, NY) plus 10% FCS (Boehringer, Mannheim, Germany). Prior to receptor determination with immunocytochemistry and ligand assays the cells were grown for 3 days in a 4To whom correspondence and reprint requests should be addressed, at Philipps Universitat Marburg, Department of Internal Medicine. Division of HematologyiOncology, Baldinger Strasse. 35033 Marburg, Germany. Fax: (49) 6421-282700. Abbreviations: SCLC, small-cell lung cancer; NSCLC, non-small-cell lung cancer; ER, estrogen receptor; PR, progesterone receptor; AR, androgen receptor; GR, glucocorticoid receptor; DCM, medium supplemented with charcoal-treated FCS. Received: November 27,1995 and in revised form April 18,1996. 358 KAISER E T A L TABLE I - LUNG-CANCER CELL LINES USED IN THE STUDY SCLC SCLC SCLC SCLC SCLC SCLC SCLC SCLC SCLC SCLC SCLC SCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC NSCLC Cell line Source 24H H60 H69 H82 HI46 H187 N417 H510 H526 H841 H1618 DMS 79 H23 H125 H157 H322 H460 H596 H661 H1155 H1299 A549 EPLC32M1 EPLC272H U1752 Ul810 LCLC-97TMl LCLC103H MSTO-211H Marburg NCI NCI NCI NCI NCI NCI NCI NCI NCI NCI Dartmouth NCI NCI NCI NCI NCI NCI NCI NCI NCI NCI Marburg Marburg UppsaK Uppsala Marburg Marburi Marburg Gender m f m m m m f m m m f n.k. rn m m m m m rn m m m m m m m m m m Histologyiclass classic classic classic variant classic classic variant variant variant variant n.d. n.d. adenocarcinoma adenocarcinoma squamous broncho-alveolar large-cell adenocarcinoma large-cell large-cell I arge-ceI I adenocarcinoma squamous-cell squatnous-cell squamous-cell large-cell large-cell large-cell mesothelioma NCI, National Cancer Institute. m, male; f, female; n.k., not known; n.d., not done. TABLE I1 - ANTIBODIES USED IN THE STUDY Antieen Name Soecies Class Source estrogen receptor ERICA progesterone receptor PGR-1A6 rat monoclonal IgG Abbott (Chicago, IL) mouse monoclonal IgG Novocastra (Newcastle, UIJ) androgen receptor AN 1-15 rat monoclonal IgG Affinity Bioreagents (Freeport, UK) androeen receutor Ab 70 rabbit Dolvclonal Blauer et aL, 1991 mouse-mdnoclonal IgG Affinity Bioreagents gluco&rticoidLreceptor BuGR2 glucocorticoid receptor anti-GR 721-727 mouse monoclonal IgG Flach et al., 1992 steroid-depleted medium (DCM) based on phenol-red-free RPMI-1640 medium with 10% FCS which was freed of steroid hormones by treatment with dextran-coated charcoal. For the passaging of the NSCLC cell lines, a solution of 0.02% EDTA in PBS was used to detach the cells from the culture bottles. Antibodies Mono- and polyclonal antibodies against the androgen receptor, the estrogen receptor, the progesterone receptor and the glucocorticoid receptor used in this study have all been established and characterized, and are listed in Table 11. Immunocytochemistty For immunologic detection, cytospin slides and cryostat sections were prepared and fixed in 10% paraformaldehyde for 10 min. Subsequent to incubation with the primary antibody for 10 min at room temperature, immunodetection was performed by the avidin-biotin-complex method using a biotinylated secondary antibody (Vector, Burlingame, CA). Each preparation was performed twice. As a negative control, immunoglobulins of the same Ig class, species and in the same concentration were employed. Predominantly nuclear expression was detected when using the anti-receptor antibodies. Cell lines were graded as positive (+) if more than 50% of all cells exhibited clear nuclear expression. Reactivity in 10 to 50% of cells was regarded as weakly positive [(+)I. Receptor determination by binding assay For receptor measurement by ligand binding, whole-cell assays were used. Receptor values were calculated according to the method of Scatchard. NSCLC cell lines and the MCF-7 cell line were seeded in multiwell dishes (Nunc, Roskilde, Denmark) at a density of 50 x lo3cells per well in 0.5 ml of growth medium. After culturing for one day, cells were grown for 3 more days in 0.5 ml of DCM. On the day of the assay, cell numbers were determined from quadruplicate cultures. The wells were subsequently washed with fresh DCM and supplied each with 200 pl DCM. A volume of 25 pI of tritiated ligands in PBS-BSA was added to achieve final concentrations ranging from 0.1 to 25 nM in quadruplicate cultures; (6.7-3H(N))-dexamethasone(DuPont de Nemours, Dreieich, Germany), (6.7-3H(N))-estradio1, ( 17a-methyl-3H)-methyl-trienoloneand ( 1 7 ~ m e t h y l - ~ H ) promegestone were used as ligands. In case of tritiated methyltrienolone, the assay medium was supplemented with 1 pM triamcinolone acetonide (Sigma, St. Louis, MO) to prevent binding to the glucocorticoid receptor. Thereafter the probes were adjusted to 250 p1 by adding 25 pl of PBS-BSA with or without unlabeled ligands. Parallel runs in the pres- STEROID RECEPTORS IN LUNG CANCER 3.59 ence of an excess (1 pM) of unlabeled ligands (Sigma) allowed Estrogen receptor the determination of unspecific binding. The wells were Immunoreaction with the monoclonal antibody (MAb) diincubated at 37°C in 5% C02/95% air. After 2 hr the cellular rected to the estrogen receptor (ER-ICA) resulted in strong monolayers were washed twice with cold DCM. The cells were nuclear positivity for approximately 80% of MCF-7 cells. B lysed with 200 p,l 1 M NaOH per well at room temperature lymphoblasts were completely negative. Among the lungovernight. Then 200 p1 1 M HCI and 100 pl 1 M HEPES buffer cancer cell lines, 7 of the 29 cell lines (24%), 2 of the 12 SCLC were added for neutralization. The radioactivity of the soh- and 5 of the 17 NSCLC cell lines expressed estrogen receptor. tions was determined in a beta-scintillation counter. Estrogen-receptor-specific DNA fragments with the exSCLC cell lines which grow in suspension cultures were pected size of 0.4 kb could be detected in MCF-7 (positive cultivated for 3 days in DC medium and washed. Portions of control) and in 12 of 23 (52%) lung-cancer cell lines. All 10 2 x lo6 cells in 200 p,I of D C medium were pipetted into SCLC cell lines were negative, while 12 of the 13 NSCLC cell multiwell dishes (Nunc). Incubations with the ligands were lines demonstrated RNA expression (Fig. 2). performed as described for the NSCLC cell lines. Thereafter Receptor content of the MCF-7 cell line by ligand binding 5O-pl volumes of a 0.05% Trypan-blue solution was added to was determined as 46000 sites per cell (KD0.1 nM). In 8 of the facilitate the following manoeuvre. The cell suspensions were 28 cell lines (29%) estrogen receptor was detected. In 2 of the laid on 2-ml cushions of cold PBS-BSA/20% glycerol in 5-rnl receptor-positive cell lines quantification was not possible; in tubes. The tubes were then centrifuged at 1000 g for 10 min one case, receptor content was less than 1000 binding sites per and the liquid contents were discarded. The radioactivity of cell. One of the 12 SCLC cell lines and 7 of the 16 NSCLC cell the resultant cell pellets was determined as described above. lines were positive. Cell lines were graded as positive (+) if specificbinding sites Comparing the methods employed the data suggest that could be detected. Quantification was expressed in binding SCLC cell lines do not express the estrogen receptor, whereas sites per cell. In some cases specific binding sites could be estrogen-receptor-RNA expression is common in NSCLC cell detected though not quantified, which means that low specific lines. However, protein data as measured by ligand-binding binding could be demonstrated at all ligand concentrations; assays and immunocytochemistry suggest that less than half of however, Scatchard plots could not be evaluated. These cases the NSCLC cell lines express detectable amounts of the and cell lines which had receptor expression of less than 1000 estrogen receptor. binding sites per cell were interpreted as weakly positive [(+)I. Progesterone receptor Preparation of RNA The MCF-7 cell line showed heterogenous nuclear reaction For polymerase-chain-reaction (PCR) analysis, cells grow- in the majority of cells with the anti-progesterone-receptor ing in RPMI-1640 medium plus 10% FCS were collected in the MAb in immunocytochemistry. No immunoreaction was demlogarithmic growth phase. Cells were washed twice in PBS and onstrated in the B-lymphoblast cell line. Nineteen of the 28 were dissolved in 6 M guanidinium thiocyanate/25 nM sodium cell lines tested, 19 (68%) expressed the progesterone recepcitrate/0.5% sodium Iauryl sarcosine. This solution was over- tor, among them 6 of the 11 SCLC cell lines and 13 of the 17 laid on a CsTFA cushion (Pharmacia, Uppsala, Sweden) and NSCLC cell lines. total RNA was pelleted in a Beckman centrifuge (24 hr, 15”C, Using PCR, MCF-7 cells were positive (0.6 kb) as were 16 of 125,OOOg). For reverse transcription, PolyA-RNA was isolated the 24 lung-cancer cell lines (66%). In addition, 5 of the 10 from 75 pg total RNA using paramagnetic oligo (dT)25 beads SCLC cell lines examined and 11 of the 14 NSCLC cell lines (Dynal, Oslo, Norway). expressed progesterone-receptor RNA. Ligand-binding studies revealed a receptor content of 6000 PCR sites/cell (K, 0.3 nM) in MCF-7 cells. Of the 27 lung-cancer PolyA-RNA isolated from 75 pg total RNA was reverse cell lines, 15 (56%) exhibited receptor expression; in 10 cases transcribed with Superscript RT (GIBCO-BRL) in 30 p1 among the positive cell lines, quantification could not be reaction volume. cDNAs were tested by amplification of performed; 2 of the 11SCLC cell lines and 13 of the 16 NSCLC GAPDH (5’ primer: CGT CTT CAC CAT GGA GA; 3‘ cell lines showed receptor expression. primer: CGG CCA TCA CGC CAC AGT TT) in the presence In conclusion, 5 SCLC cell lines showed RNA expression, 6 of [a3’P]-dCTP and were equilibrated to give similar GAPDH cell lines were positive for immunocytochemistry. Overall data signals. For the amplification of specific cDNAs the following suggest low expression of progesterone receptor in SCLC cell primer pairs were used: lines, whereas in 13 of 17 NSCLC cell lines at least 2 methods demonstrated receptor expression. hGR: S’primer: CCA TCC AGC TCC TCA ACA GCA A 3’primer: TCG ACT TTC TTT AAG GCA ACC Androgen receptor hAR: S’primer: TGG ACA CGA CAA CAA CCA GC Immunoreaction with the anti-androgen-receptor MAb was 3 ’primer: CTG GTA GAA GCG TCT TGA GC strictly nuclear in the MCF-7 line. B-lymphoblast cells were hPR 5’primer: AAC TCA AGA CCT CAT AAT CC negative for androgen-receptor expression. Of the twenty-nine 3’primer: CAT TCA GTA TTC 7TG GAT GAG C lung cancer cell lines, 16 (55%) revealed positivity; 4 of the 12 hER S’primer: CAA GAC ATA CCA AGA GAA GC SCLC cell lines and 12 of the 17 NSCLC cell lines revealed 3’primer: GTT GCT GTC ACT ACT CAG GC. nuclear positivity. Cell lines were graded as positive if a distinct fragment could RNA expression was demonstrated in the MCF-7 cell line at be detected. A faint band was evaluated as weakly positive. 0.5 kb. Of the 23, 15 (65%) cell lines were positive, while 4 of the 10 SCLC cell lines and 11 of the 13 NSCLC cell lines expressed androgen-receptor RNA (Figs. 3,4). RESULTS In the ligand-binding studies, MCF-7 cells had a receptor Receptor expression in lung-cancer cell lines content of 14000 sitesicell (K, 0.3 nM). Out of 27 cell lines, 15 The results are summarized in Table 111. Data on receptor (56%) showed receptor expression; in 2 cases quantification quantification by ligand-binding assays are listed in Table IV. was not possible, in 1 case receptor content was less than 1000 Using all 3 detection methods, receptor expression was deter- binding sites per cell; 3 of the 11 SCLC cell lines and 12 of the mined as follows (Figs. 1-5). 16 NSCLC cell lines were positive. 360 KAISER E T A L TABLE 111- STEROID-RECEPTOR CONTENT OF LUNG-CANCER CELL LINES' Estrogen receptor Cell line IC 24H H60 H69 H82 H146 H187 N417 H510 H526 H84 1 H1618 DMS7V H23 H125 H157 H322 H460 H5Y6 H661 H1155 H1299 A549 EPLC 32M1 EPLC272H U1752 U1810 LCLC-97TM1 LCLC103H MSTO-211H Progesterone receptor Ligand PCR Cell line (+) 24H H60 H69 H82 H146 H187 N417 H510 H526 H841 H1618 DMS79 H23 H125 H157 H322 H460 H596 H661 H1155 H1299 A549 EPLC 32M1 EPLC272H U1752 U1810 LCLC-V7TM1 LCLC103H MSTO-211H - ti] - + - - - - - + Androgen receptor Cell line 24H H60 H69 H82 H146 HI87 N417 H510 H526 H841 H1618 DMS79 H23 H125 H157 H322 H460 H596 H661 H1155 H1299 A549 EPLC 32M1 EPLC272H U1752 U1810 LCLC-97TM1 LCLC103H MSTO-211H IC Liaand PCR - - - + - - ++ + ++ t + - (+I - 111 >+' Ligand + - PCR + +- (i) n.d. +- + n.d. - + + + + + + n.d. n.d. n.d. +- - n.d. + (+I + + (+I +- + Glucocort~co~d receptor (+) n.d. - (+I IC (+I + (+) + + +++ n.d. - Cell line - - + n.d. - + + n.d. +- + + (+) n.d. - + n.d. n.d. + + + (+I + n.d. (+) i+) 24H H60 H69 H82 H146 H187 N417 H5 10 H526 H841 H1618 DMS79 H23 H125 H157 H322 H460 H5Y6 H661 H1155 H1299 A549 EPLC 32M1 EPLC272H U1752 U1810 LCLC-Y7TM1 LCLC103H MSO-211H IC + + n.d. n.d. + + + + + + + + + n.d. + + + + + + + + + n.d. n.d. - n.d. n.d. + Ligand PCR + + + + + + + + + + + n.d. + + + + n.d. + + + + + - n.d. (+) ++ + + + + + + + + + + + + + + + + + + n.d. + + n.d. n.d. + + + + + + + + 'As measured by immunocytochemistry (IC), ligand-receptor binding (Ligand) and RNA expression (PCR) detected via RT-PCR amplification. Comparing the results, 2 SCLC cell lines were positive for the androgen receptor with all 3 methods employed. In 2 further cases expression data were incomplete, These results suggest that androgen-receptor expression in SCLC is uncommon. A few cases, however, may express the androgen receptor. In 12 of the 17 NSCLC cell lines, at least 2 methods demonstrated receptor expression, although in 4 of these cases R N A expression was low. Nevertheless, androgen receptor in NSCLC cell lines is a common event. Glucocorticoid receptor In immunocytochemistry, glucocorticoid-receptor expression was high in MCF-7 cells with both MAbs described above. B-lymphoblast cells showed an intense nuclear reaction pat- STEROID RECEPTORS IN LUNG CANCER 361 TABLE IV - STEROID-RECEPTORCONTENT OF LUNG-CANCER CELL LINES’ Cell line 24H H60 H69 H82 H146 H187 N417 H5 10 H526 H841 H1618 DMS 79 H23 HI25 H157 H322 H460 H596 H661 H1155 H1299 ‘4549 EPLC 32M1 EPLC272H U1752 U1810 LCLC-97TM1 LCLC103H MSTO-211H ER Kd PR Kd AR Kd (+I n.d. - 0.5 1 1.9 0.3 1500 0.3 3500 0.4 12000 10000 18000 111000 1.3 0.7 1.8 41000 200 4000 69000 27000 0.3 2 0.5 0.5 0.3 19000 18000 2.5 0.5 - 3.4 1.5 0.6 4.7 1 0.8 - (+I 1.6 - 0.4 - n.d. - GR Kd 3400 1900 2000 1100 5400 5000 n.d. 5.1 1.3 3.2 3.1 5.4 4.7 (A0 2.8 3700 24000 3 1000 36000 211000 42000 85000 186000 86000 20400 42000 314000 141000 147000 157000 47000 139000 66000 63000 18.1 14 3.5 10 7.7 10 7 8.2 2.6 9.8 1.6 9.3 8.2 8.7 7.6 12 7.7 7.3 5.4 - ‘As measured by ligand-receptor binding. Receptor content in binding sites per cell (parts of the results of the glucocorticoid receptor have been published in Hofmann et al. (1995)). FIGURE1 - Immunocytochemistry of lung-cancer cell line NSCLC-EPL32M1 with anti-GR. Scale bar, 10 km. tern. Among the lung-cancer cell lines, 22 of the 22 lines (95%) revealed glucocorticoid-receptor expression, while 10 out of 10 SCLC and 11 of the 12 NSCLC cell lines were positive. In PCR the MCF-7 cell line was positive at 1.15 kb, as were all 24 (100%) lung-cancer cell lines (Fig. 5 ) . In the ligand-binding studies, the MCF-7 cell line was measured as having a receptor content of 19000 sites/cell (K, 1.2 nM). Of 28 cell lines measured, 27 (96%) were receptorpositive; among them quantification was not possible in 1 case; 10 of 11 SCLC cell lines and all 17 NSCLC cell lines were positive. All 3 methods demonstrated GR expression in lung-cancer cell lines, in SCLC and in NSCLC cell lines. FIGURE2 - cDNA from NSCLC cell lines as indicated were amplified with specific primers for the human estrogen receptor (for primer sequences, see “Material and Methods”). The length of the amplification product is 0.4 kb, as expected. The size marker (M) consisted of HaeIII fragments of PhiX 174 DNA. The MCF-7 cell line was included as positive control; for negative control, c~~~ was substituted with water, Immunohistochemistryin primary lung cancer Using immunohistochemistry, glucocorticoid receptor was detected in 49 cases (94%), progesterone receptor in 23 cases (44%), androgen receptor in 9 cases (17%), and estrogen receptor in 4 of the 52 tumor samples (8%). When tumors were grouped according to histology, sex-steroid-receptor expression was most frequent in adenocarcinoma, as indicated in Table V. Out of 16 cases, 11 (69%) exhibited sex-steroidreceptor expression. Receptor expression was found in 47% of the 17 squamous-cell carcinomas and 23% of the 13 large-cell carcinomas. When tumors were grouped according to the gender of the patients, there was strikingly higher expression of sex-steroid receptors in tumors from women than in tumors from men: KAISER ETAL. 362 TABLE V - STEROID-RECEPTOR EXPRESSION IN PRIMARY LUNG CANCER DETERMINED BY IMMUNOHISTOCHEMISTRY FIGURE3-cDNA from NSCLC cell lines as indicated were amplified with specific primers for the human androgen receptor (for primer sequences, see “Material and Methods”). The length of the amplification product is 0.5 kb, as expected. For description of marker and control, see Figure 2. FIGURE4-cDNA from SCLC cell lines as indicated were amplified with specific primers for the human androgen receptor (for primer sequences, see “Material and Methods”). The length of the amplification product is 0.5 kb, as expected. For description of marker and control, see Figure 2. Gender Histology Sl m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m m f squamous-cell squamous-cell squamous-cell squamous-cell squamous-cell squamous-cell squarnous-cell squamous-cell squamous-cell squarnous-cell squarnous-cell squamous-cell squamous-cell squatnous-cell adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma broncho-alveolar large-cell large-cell large-cell large-cell large-cell large-cell large-cell large-cell large-cell large-cell large-cell SCLC squamous-cell squamous-cell squamous-cell adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma adenocarcinoma large-cell large-cell carcinoid carcinoid SCLC SCLC IIIA I I I1 I I IIIB IIIA I1 I1 IIIA IIIA IIIA I 11 IIIB I IIIA IIIB IIIA IIIA IIIA IIIA IIIA I IV I1 IIIA IIIA I1 IIIB I1 I1 IIIA IV I IIIB I1 IIIA I1 IIIA IIIA 1 IIIB IIIA I IIIA IIIA I I I I f f f f f f f f f X 2 2 3 2 3 2 2 2 3 I X X X X X X + - + + - - + + + + + + + - - - - + - + - + + - - - + - + + - - - + - + + + + + + + + + + + + + + + st, staging. Clinical staging of lung cancer was performed according to the International System for Staging Lung Cancer (Mountain, 1986). G, cytological grading of tumour cells, 1-111; X, no grading performed. +, distinct nuclear immuno reaction in tumour cells; -, no immuno reaction detectable. FIGURE5 - cDNA from NSCLC cell lines as indicated were amplified with specific primers for :he human glucocorticoid receptor (for primer sequences, see Material and Methods”). The length of the amplification product is 1.15 kb, as expected. For description of marker and control, see Figure 2. 69% of the 16 tumors from female patients vs. 42% of the 36 tumors from male patients. DISCUSSION Differences in survival among men and women with lung cancer have led to the assumption that sex hormones and their receptors may play a role in the biology of this tumor. Bhatavdekar et al. (1994) have shown that serum levels of various hormones, including FSH, LH and testosterone, in lung cancer patients differ from the levels in controls. A review of the literature reveals that in primary lung cancer only limited data are available on hormone-receptor expression, which is the prerequisite for hormone action. Binding assays have been performed to some extent in lung cancer. The glucocorticoid receptor has been reported to be expressed in a large number of non-small-cell lung cancers. However there is large variation in reported results, depending on the binding assay used. Beattie et al. (1985) reported 75% of NSCLC to be positive for the glucocorticoid receptor, while Chaudhuri et al. (1982) found that 30% of NSCLC expressed STEROID RECEPTORS IN LUNG CANCER 363 the glucocorticoid receptor. No standard threshold for recep- than in the cell lines. Loss of protein by storing procedures tor positivity has been established. Inter-observer differences cannot be completely ruled out. The glucocorticoid receptor is are also striking for detection of the androgen receptor. While frequently expressed in primary tumors. Among the sexBeattie et al. (1985) found androgen receptors in almost 75% steroid receptors, expression is lowest for estrogen receptor, of 26 cases, Chaudhuri et al. (1982) and Cagle et al. (1990) while a higher expression rate is seen for progesterone reported such expression in less than 20% of NSCLC. Data for receptor. Our data confirm the observation that steroidexpression of the estrogen and the progesterone receptors are receptor expression is highest in adenocarcinoma (Beattie et more conclusive. The estrogen receptor was found in 10 to al., 1985). 30% of NSCLC (Beattie et al., 1995; Chaudhuri et al., 1982; The clinical impact of these findings remains to be deterCagle et al., 1990; Kobayashi et al., 1982). The progesterone mined. In one sub-group of positive cell lines, receptor receptor was found in 20 to 25% of NSCLC by means of expression as determined by ligand-binding studies was low. ligand-binding assays (Beattie et al. 1985; Chaudhuri et al., However, in another sub-group, the glucocorticoid receptor 1982; Cagle et al., 1990). and the androgen receptor, particularly, were expressed to a Data on small-cell lung cancer are sparse. At least 0.5 to 1 g considerable level. Hence, these cell lines may represent of tissue have to be available for binding assays. Since the tumors that are sensitive to hormonal treatment. Only 3 cell diagnosis of small-cell lung cancer is most frequently made on lines investigated were from female patients; however, data small bronchoscopic biopsies, binding assays are often not from primary tumors are more conclusive. The better prognosis for women with lung cancer may be related to increased feasible in this tumor type. We chose a set of lung-cancer cell lines to study receptor expression of steroid receptors, as seen in tumors from women. expression using 3 different methods, namely immunocytochem- However, other aspects have to be considered, such as differistry, ligand-binding assays and RNA expression. A survey of ing metabolisation to cytostatic drugs in men and women. The epidemiologic data and our own findings suggest that receptor expression on a panel of SCLC and NSCLC cell lines shows expression of sex-steroid receptors in a sub-group of cell sex hormones may influence lung cancer. Proliferation studies are currently undertaken to evaluate the influence of estrogen lines. Among a number of conclusions, our results indicate strik- and progesterone on lung-cancer cell lines. Depending on ing differences in sex-hormone-receptor expression between these results, a sub-set of patients may be defined whose SCLC and NSCLC cell lines. Expression of estrogen, androgen tumors may be susceptible to hormonal influence. Earlier in and progesterone receptors is rare in SCLC cell lines. In at vitro studies have indicated a growth-inhibitory effect of least 2 methods of investigation, NSCLC cell lines showed low dexamethasone (Hofmann et al., 1995) and a growth-promotexpression of estrogen receptor in 6 of progesterone receptor ing effect of testosterone in lung-cancer cell lines (Maasberg et in 13 and of androgen receptor in 12 out of 17 cases examined. nl., 1989). In a pilot study performed at our institute, a Glucocorticoid-receptor expression is substantial in NSCLC as temporary growth-inhibitory effect of the anti-androgene fluwell as SCLC cell lines. For detection of the glucocorticoid tamide combined with an LHRH-agonist could be detected in receptor and the sex steroid receptors, immunochemistry patients with extensive small-cell lung cancer. appears as a feasible method. In a few cases, discrepancies Adenocarcinomas of the lung and other tumors from female between results of immunocytochemistry and of PCR were patients preferentially express sex-steroid receptors. Further detected. Since these were cases in which “borderline” expres- research on the functional activity of sex steroids will focus on sion was observed, the different threshholds of positivity may these sub-groups. account for these findings. In a few cases with positive RNA expression but lack of protein detection receptor expression ACKNOWLEDGEMENTS may have been too low to be detected. Receptor-expression studies in primary tumors using immuWe thank Mrs. B. Safran for excellent laboratory assistance. nohistochemistry widely confirm the data obtained from cell This work was supported by a grant from the Deutsche lines. However, expression rates for all receptors are lower Krebshilfe, Bonn, Germany (project W30/93/ka2). REFERENCES ALBAIN,K.S., CROWLEY, J.J., LE BLANC,M. and LIVINGSTON, R.B., Determinants of improved outcome in small-cell lung cancer: an analysis of the 2,580-patient Southwest Oncology Group data base. J. din. OncoL, 8, 1563-1574 (1990). BEATTIE, C.W., HANSEN, N.W. and THOMAS, P.A., Steroid receptors in human lung cancer. Cancer Res., 45,4206-4214 (1985). human cell line (MCF-7) from breast carcinoma. J. biol. Chem., 248, 6521-6523 (1973). CAGLE,P.T., MODY, D.R. and SCHWARTZ, M.R., Estrogen and progesterone receptors in bronchogenic carcinoma. Cancer Res., 50, 6632-6635 (1990). CARNEY, D.N.. GAZDAR, A.F., BEPPLER, G., GUCCION, J.G., MAFUNBEPPLER,G., JAOUES,G., NEUMANN, K., AUMULLER, G., GROPP, GOS,P.J., MOODY,T.W., ZWEIG,M.H. and MINNA,J.D., EstablishK., Establishment, growth properties, and morpho- ment and identification of small-cell-lung-cancer cell lines having C. and HAVEMANN, logical characteristics of permanent human small-cell-lung-cancer cell classic and variant features. Cancer Res., 45,2913-2923 (1985). lines. J. CancerRes. clin. Uncol., 113,3140 (1987). CHARPIN, C., MARTIN,P.M., JACQUIMIER, J., LAVAUT,M.N., POURBERG,J., NILSSON, K., EKMAN, R. and GIOVANELLA, B., Establishment REAU-SCHNEIDER, N. and TOGA,M., Estrogen-receptor immunocytoand characterization of cell lines from human small-cell and large-cell chemical assay (ER-ICA): computerized image analysis system, immucarcinomas of the lung. Acta path. microbiol. immunol. scand. sect. A., noelectron microscopy, and comparisons with estradiol-binding assays 93,133-147 (1985). in 115 breast carcinomas. CancerRes., 46,4271~4277s(1986). BHATAVDEKAR, J.M., PATEL,D.D., CHIKLIAR, P.R., MEHTA,R.H., CHAUDHURI, P.K., THOMAS,P.A., WALKER,M.J., BRIELE,H.A., T.P., DAS GUPTA,Y.K. and BEATTIE, C.W., Steroid receptors in human VORA,H.H., KARELIA, N.H., GHOSH,N., SHAH,N.G., SUTHAR, NEEMA,J.P. and BALAR,D.B., Levels of circulating peptide and lung-cancer cytosols. Cancer Leu., 16,327-332 (1982). steroid hormones in men with lung cancer. Neoplusma, 41. 101-103 FINKELSTEIN, D.M., ETTINGER, D.S. and RUCKDESCHEL, J.C., Long(1994). term survivors in metastatic non-small-cell lung cancer. A n Eastern BLAUER, M., VAALASTI, A., PAULI,S.L., YLIKOMI, T., JOENSUU, T. and Cooperative Oncology Group Study. J. clin. Uncol., 4,702-709 (1986). TUOHIMAA, P., Location of androgen receptors in human skin. J. FLACH, H., KAISER, U. and WESTPHAL, H.M., Monoclonal anti-peptide Invest. Dermatol., 97,264-268 (1991). antibodies to the lucocorticoid receptor. J. Steroid Biochem. molec. BROOKS, S.C., LOCKE,E.R. and SOULE,H.D., Estrogen receptor in a Biol., 42,467-474 8992). 364 KAISER ETAL. GIARD,J.D., AARONSON, S.A., TODARO, G.J., ARNSTEIN, P., KERSEY, dent proliferation, and Sa-reductase activity of small-cell lung cancer J.H., DOSIK,H. and PARKS, W.P., In vitro cultivation of human tumors: cell lines. IntJ Cancer, 43,685-691 (1989). establishment of cell lines derived from a series of solid tumors. J. not. MOUNTAIN, C.F., The new international system for staging lung Cancerhi., 51,1417-1423 (1973). cancer. Chest, 89 (Suppl), 2258-233s (1986). HOFMANN, J., KAISER, U., MAASBERG, M. and HAVEMA”,K., Gluco- O’CONNEL, J.P., KRIS,M.G., GRALLA, R.J., GROSHEN, S., TRUST,A., corticoid receptors and growth-inhibitory effects of dexamethasone in FIORE.J.J., KELSEN,D.P., HEELAN,R.T. and GOLBEY,R.B., Frehuman lung-cancer cell lines Europ. J. Cancer, 31A, 2053-2058 (1995). quency and prognostic importance of pretreatment clinical characterJOHNSON, B.E., STEINBERG, S.M., PHLEBS,R., EDISON,M., VEACH, istics in patients with advanced non-small-cell lung cancer treated with S.R. and IHDE, D.C., Female patients with small-cell lung cancer live combination chemotherapy. J. clin. Oncol., 4,1604-1614 (1986). PETTENGILL. O.S., SORENSON, G.D., WURSTER-HILL, D.H., CURPHEY, longer than male patients.Amer. J. Med., 85,194-196 (1988). T.J., NOLL,W.W., CATE, C.C. and MAURER,L.H., Isolation and KOBAYASHI, S.. MIZUNO, T., TOBIOKA, N.. ICHIMURA, H., SAMOTO, T., growth characteristics of continuous cell lines from small-cell carciA., WAKABAYASHI, S., UMEMURA, S., FUKUOKA, noma of the lung. Cancer, 45,906-918 (1980). TANAKA, H., MASAOKO, H. and NAGAI,H., Sex-steroid receptors in diverse human tumors. WOLF,M., HOLLE,R., HANS,K., DRINGS,P. and HAVEMANN, K., Gann, 73,439-445 (1982). Analysis of prognostic factors in 766 patients with small-cell lung MAASBERG, M., ROTSCH,M., JAQUES, G., ENDERLE-SCHMIDT, U., cancer (SCLC): the role of sex as a predictor for survival. Brit. J. WEHLE,R., HAVEMANN, K., Androgen receptors, androgen-depen- Cancer, 63,986-992 (1991).