- Email: [email protected]
Modulation of Sensitivity to Transforming Growth Factor-β1 (TGF-β1) and the Level of Type II TGF-β Receptor in LNCaP Cells by Dihydrotestosterone
EXPERIMENTAL CELL RESEARCH
222, 103–110 (1996)
Article No. 0013
Modulation of Sensitivity to Transforming Growth Factor-b1 (TGF-b1) and the Level of Type II TGF-b Receptor in LNCaP Cells by Dihydrotestosterone1 ISAAC YI KIM, DAVID J. ZELNER, JULIA A. SENSIBAR, HAN-JONG AHN, LINDA PARK, JIN-HO KIM, AND CHUNG LEE2 Department of Urology, Northwestern University Medical School, Chicago, Illinois 60611
Transforming growth factor-b1 (TGF-b1) and androgen are potential physiological regulators of prostate cancer cells. In the present study, we have used LNCaP cells as a model of androgen-responsive prostate cancer to investigate the effects of dihydrotestosterone (DHT) on the sensitivity to TGF-b1. The ability of LNCaP cells to respond to TGF-b has been controversial. In some studies, LNCaP cells were insensitive to TGF-b1 while, in others, they were sensitive to the growth inhibitory effect of TGF-b1. The present study was carried out to establish androgenic conditions that rendered LNCaP cells sensitive to TGF-b1. Cells were cultured in phenol-red-free RPMI 1640 medium supplemented with 10% charcoal-stripped fetal bovine serum. DHT was added at the following concentrations: 0, 10012, 10010, and 1007 M. These concentrations were selected because they represent the zero DHT control, the low-proliferative dose, the high-proliferative dose, and the growth-arrest dose, respectively. The effects of TGF-b1 observed on LNCaP cells included inhibition of cell proliferation, decrease in cell viability, alteration in cell morphology, and enhancement of gene transcriptional activity through activation of a TGF-b responsive promoter. Of the various DHT concentrations investigated in this study, these effects of TGF-b1 on LNCaP cells were consistently demonstrated only at 10010 M. At other concentrations, the effects of TGF-b1 were either minimal or undetectable. Accompanying these effects of TGF-b1, a low but statistically significant level of TGF-b1-specific binding and an increased protein level of TGF-b receptor type II were detected by a competitive binding assay and Western blot analysis, respectively. These results indicate that LNCaP cells can be induced by DHT to respond to TGF-b1 and that DHT modulates the sensi-
1
This work was supported in part by NIH Grants DK 39250 and CA 60553. 2 To whom correspondence and reprint requests should be addressed at Tarry 11-715, Department of Urology, Northwestern University Medical School, 303 E. Chicago Avenue, Chicago, IL 60611.
tivity to TGF-b1 and the level of TGF-b receptor type II in these cells. q 1996 Academic Press, Inc.
INTRODUCTION
Transforming growth factor-b (TGF-b) is a 25-kDa pleiotropic growth factor that has been shown to inhibit the proliferation of most cells, especially those of the epithelial lineage [1]. Among prostate cancer cell lines, TGF-b has been demonstrated to inhibit the proliferation of PC3 and DU145 cells [2, 3]. These findings suggest that TGF-b is a potential regulator of prostate tumor biology. In LNCaP cells, however, the effects of TGF-b is uncertain because some investigators observed an inhibitory effect while others detected no effect [2, 4, 5]. LNCaP cells are androgen-sensitive and are used extensively as an in vitro model for the study of androgen action in the growth of prostate cancer cells [6, 7]. Although LNCaP cells have a mutated androgen receptor at the androgen binding domain and respond not only to androgens but also to anti-androgens, estrogen, and progestins [8, 9], they remain a useful model system for prostate cancer since androgen receptor mutations are a frequent event in many prostate cancer cases [10]. Recent observations have demonstrated that LNCaP cells exhibit a biphasic growth response with androgen stimulation [11]. At low concentrations of dihydrotestosterone (DHT) (10012 to 10010 M), the cells proliferate in a dose-dependent manner. However, at concentrations greater than 1009 M DHT, the rate of proliferation declines in a dose-dependent manner. These observations suggest that DHT may be able to elicit a different biological response in LNCaP cells depending on the concentration used in the culture medium. One of the biological responses that DHT may modulate in LNCaP cells may be the sensitivity to the growth inhibitor TGF-b1 because recent reports have indicated that androgen can profoundly influence TGFb action [12–14].
103
/ m4678$3013
12-08-95 13:08:40
eca
0014-4827/96 $12.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP: Exp Cell
104
KIM ET AL.
Since both androgens and TGF-b are potential physiological regulators of prostate tumor progression, an investigation of the interaction between these two factors will lead to further understanding of the biology of prostate cancer. In the present study, we have used LNCaP cells as an in vitro model of androgen-responsive prostate cancer to investigate the possibility that DHT may modulate the sensitivity to TGF-b1. To establish TGF-b1 sensitivity in these cells, we have deliberately used different but complementary approaches because these cells have been previously reported to be insensitive to the action of TGF-b1 [2, 4]. MATERIALS AND METHODS Cell culture. LNCaP and PC 3 cells were purchased from the American Type Culture Collection (Rockville, MD). In this study, all cells were derived from the 48th through the 51st passages. Cells were routinely maintained in RPMI 1640 containing 10% fetal bovine serum (FBS), penicillin (100 U/ml), and streptomycin (100 mg/ml). Phenol-red-free RPMI 1640 supplemented with dextran-treated charcoal-stripped serum (cFBS) was used in each experiment. Crystalline DHT (Sigma Chemical Co., St. Louis, MO) was dissolved in absolute ethanol at a concentration of 1003 M and was reconstituted in the culture medium at concentrations ranging from 1007 to 10012 M. Human TGF-b1 (Collaborative Research, Bedford, MA) was diluted to 2 mg/ml and added to culture medium at preselected concentrations. Mitogenic assay. Cellular proliferation was assayed by cell counts and [3H]thymidine incorporation [15]. For cell counts, LNCaP cells at 5 1 103/well or 1 1 104/well were plated in 24-well culture plates in phenol-red-free RPMI 1640 supplemented with 10% FBS and allowed to adhere for 24 h. Then, the cultures were washed two times with Hank’s balanced salt solution (HBSS) and cells from previously selected wells were harvested and counted to determine the plating efficiency. Cells in remaining wells were cultured for 2 and 4 days in phenol-red-free RPMI 1640 supplemented with cFBS containing different concentrations of DHT in the presence or absence of TGFb1 at 0.2, 2, 20, and 200 ng/ml. For cultures lasting longer than 2 days, medium was changed every other day. To count cells, the medium was removed and the cells were detached with 0.5 ml of 0.25% trypsin. Once detached, the cells were added to 9.5 ml of isotonic solution (Isoton II, Coulter Corp., Hialeah, FL) and counted with a coulter counter (ZF, Coulter Corp., Hialeah, FL). In subsequent studies, unless otherwise specified, a concentration of TGF-b1 at 2 ng/ ml was selected. For [3H]thymidine incorporation assay, LNCaP cells were seeded in 24-well plates as previously described at 2 1 104 cells/well. Once the plating efficiencies were determined, the cells were treated with 2 ng/ml of TGF-b1 at varying concentrations of DHT for 24 and 48 h. After the preselected time of incubation, 1 mCi/ml of [3H]thymidine (6.7 Ci/mM, Amersham Corp., Arlington Heights, IL) was added to each well and incubated for 4 h. The cells were subsequently treated with 1.0 ml of cold 10% trichloroacetic acid. The precipitates were dissolved in 0.5 ml of 0.4 N NaOH, and the amount of radioactivity was counted in a liquid scintillation counter. Cell viability was determined by a trypan blue dye exclusion test. After treating LNCaP cells with 2 ng/ml of TGF-b1 for 2 days, the percentage of viable cells was calculated by the following equation: % viable Å
trypan-blue negative cells 1 100 . total no. of cells
Photomicrographs of cells were taken directly from cultured plates
/ m4678$3013
12-08-95 13:08:40
eca
under an inverted microscope equipped with camera (Olympus OM 4T, Olympus Camera Co., Woodbury, NY) using Ektachrome 64T films. Prints developed from these films were used as permanent hard copy records. Transient transfection and luciferase assay. Cells were seeded as above in six-well plates at 1 1 105 cells/well. After treatment for 2 days in varying concentrations of DHT, cells were transiently transfected with plasmid p3TP-Lux using lipofectamine according to the manufacturer’s directions (Gibco-BRL, Grand Island, NY). p3TPLux contains three TPA responsive elements from the human collagenase gene and one TGF-b responsive element from the human plasminogen activator inhibitor-1 (PAI-1) promoter linked to the luciferase reporter gene [16]. Throughout the transfection protocol, the cells were constantly maintained in phenol-red-free RPMI 1640 with 10% cFBS and preselected concentrations of DHT. Briefly, 1 mg of p3TP-Lux and 12 ml of lipofectamine were added with 1 ml of transfection medium (Opti-mem, Gibco-BRL) to each well and incubated for 24 h. Subsequently, fresh medium was added and the cells were incubated for an additional 24 h. Finally, 5 ng/ml of TGF-b1 was added and the cultures were maintained for an additional 16 h. This concentration of TGF-b1 and culture conditions were selected according to the original protocol [16]. The extent of the promoter activity of p3TP-Lux was assayed by measuring luciferase activity using a commercial luciferase assay kit (enhanced luciferase assay kit 1800k, Analytical Luminescence Laboratory, San Diego, CA) and a luminometer (Monolight 2010B, Analytical Luminescence Laboratory, San Diego, CA). TGF-b1 binding assay (17). Cells were plated as above in 24-well plates at 2 1 104 cells/well. After culturing for 2 days in phenol-redfree RPMI 1640 supplemented with cFBS and varying concentrations of DHT, they were washed two times with phosphate-buffered saline (PBS) and harvested by scraping. Subsequently, the cells were incubated for 3 h with 200 pM of 125I-TGF-b1 (800–2200 Ci/mM, Amersham Corp, Arlington Heights, IL) in the presence (nonspecific binding) or absence (total binding) of 100-fold molar excess of nonradiolabeled TGF-b1 in binding buffer (PBS containing 0.9 mM CaCl2 , 0.49 mM MgCl2 , and 1 mg/ml bovine serum albumin) at 47C. After incubation, cells were applied to circular 3 MM Watman filter papers and washed three times with the same binding buffer without bovine serum albumin. The remaining radioactivity was counted by gamma counter (Cobra II Autogamma, Packard Instrument Company, Downers Grove, IL). Specific binding of 125I-TGF-b1 was determined by calculating the difference in radioactivity between nonspecific and total binding. The difference in radioactivity was converted to fmol of TGF-b1-specific binding per milligrams of protein based on the specific activity of 125I-TGF-b1. Western blot analysis for TGF-b receptors. Cells were homogenized with PBS at 47C and the protein concentration was determined according to the method of Lowry. Samples were placed in sample buffer (0.0625 M Trizma base, 2% SDS, 5% 2-mercaptoethanol) and boiled for 5 min. Electrophoresis was carried out in 10% SDS–polyacrylamide gel using 100 mg of total protein loaded onto each lane. Following electrophoresis, proteins were transferred to a 0.2-mm nitrocellulose membrane (Bio-Rad). The membrane was incubated overnight in blocking buffer (5% nonfat dry milk, PBS, and 0.1% Tween). Subsequently the membrane was incubated with anti-type II receptor and anti-type I receptor antibodies [17] at a dilution of 1:800 overnight at 47C. After washing with 0.1% PBS–Tween, the membrane was incubated in the presence of goat anti-rabbit horseradish peroxidase-labeled secondary antibody (Bio-Rad) at a dilution of 1:8000 for 4 h at room temperature. After washing several times, immunoreactive proteins were visualized by enhanced chemiluminescence (Amersham) and exposed to Hybond-ECL film (Amersham). Statistical analysis. All numerical data are expressed as mean { SEM with triplicate observations. Differences of means among different treatments were compared by unpaired Student’s t test [18]. A value of P õ 0.05 was considered statistically significant.
AP: Exp Cell
MODULATION OF TGF-b1 SENSITIVITY IN LNCaP BY DHT
FIG. 1. Effect of increasing doses of TGF-b1 on cell numbers. Cells were seeded at 5 1 103/well in 24-well plates and cultured in medium containing TGF-b1 at 0, 0.2, 2, 20, and 200 ng/ml and the following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M. Cells were counted at the end of 2 days by Coulter counter. The results are presented as percentage of control. The value in each group represents the mean of triplicate observations. The actual cell numbers were as following. 0 ng/ml TGF-b1 (control): cFBS, 10160 { 528; 10012 M DHT, 10800 { 680; 10010 M DHT, 15680 { 512; 1007 M DHT, 9600 { 584. 0.2 ng/ml TGF-b1: cFBS, 10240 { 203; 10012 M DHT, 11280 { 304; 10010 M DHT, 13040 { 696; 1007 M DHT, 9200 { 408. 2 ng/ml TGF-b1: cFBS, 10160 { 576; 10012 M DHT, 11120 { 304; 10010 M DHT, 9040 { 472; 1007 M DHT, 9440 { 600. 20 ng/ml TGF-b1: cFBS, 10320 { 440; 10012 M DHT, 10640 { 648; 10010 M DHT, 6720 { 256; 1007 M DHT, 9360 { 408. 200 ng/ ml TGF-b1: cFBS, 10320 { 440; 10012 M DHT, 10640 { 648; 10010 M DHT, 6720 { 256; 1007 M DHT, 9360 { 408. The vertical bar denotes SEM. *P õ 0.05 compared to the corresponding value in cultures grown in the absence of TGF-b1.
105
creased the cell count only at 10010 M DHT after 2 and 4 days of treatment. After 2 days of TGF-b1 treatment at 2 ng/ml, the cell count at 10010 M DHT was approximately 58% of control. Such a magnitude of decrease in cell count is comparable to that of other prostate cancer cell lines when they were treated with a similar dose of TGF-b1 [2, 3]. Following the initial decrease at Day 2, a similar magnitude of decrease in cell count was observed after 4 days of TGF-b1 treatment at 10010 M DHT. Effect of TGF-b1 and DHT on DNA synthesis and cell death. To analyze the effect of TGF-b1 on the active process of DNA synthesis, changes in [3H]thymidine incorporation were investigated. Based on the above study, in which a significant decrease in cell count was detectable after 2 days of TGF-b1 treatment, the present study was done in an attempt to investigate whether an effect can be detected at 24 and 48 h after the addition of 2 ng/ml TGF-b1 under varying concentrations of DHT. Figure 3 shows that TGF-b1 inhibited
RESULTS
Effect of TGF-b1 and DHT on cell number. Figure 1 shows the effect of increasing doses of TGF-b1 on the LNCaP cell count after 2 days of treatment in the presence of varying concentrations of DHT. Changes in the cell count were observed only when the cells were treated with TGF-b1 in the presence of 10010 M DHT. A slight decrease in the cell count was observed following TGF-b1 treatment at 0.2 ng/ml. However, a drastic decrease in cell count was observed when the cells were treated with TGF-b1 at concentrations of 2 ng/ml or greater. In the presence of 20 and 200 ng/ml TGF-b1, the cell count decreased to approximately 40% of control. In subsequent experiments, unless otherwise specified, a single dose of 2 ng/ml TGF-b1 was used. Figure 2 shows the change in cell number after LNCaP cultures were treated with 2 ng/ml TGF-b1 under varying concentrations of DHT for 2 and 4 days. Again, the results show that TGF-b1 significantly de-
/ m4678$3013
12-08-95 13:08:40
eca
FIG. 2. Effect of TGF-b1 and DHT on cell numbers. Cells were seeded at 1 1 104/well in 24-well plates and cultured in RPMI 1640 supplemented with 10% cFBS and following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M. To preselected wells, 2 ng/ml TGF-b1 was added. Media were changed every 2 days. Cells were counted at the end of the 2- and 4-day period by Coulter counter. The results are presented as percentage of control. The value in each group represents the mean of triplicate observations. The actual cell numbers were as following. Day 2 with no TGF-b1 (control): cFBS, 36267 { 2010; 10012 M DHT, 37060 { 1180; 10010 M DHT, 57330 { 2270; 1007 M DHT, 41330 { 2460. Day 2 with 2 ng/ ml TGF-b1: cFBS, 34000 { 4380; 10012 M DHT, 31460 { 3750; 10010 M DHT, 33460 { 4160; 1007 M DHT, 42400 { 2830. Day 4 with no TGF-b1 (control): cFBS, 42000 { 3600; 10012 M DHT, 45200 { 2880; 10010 M DHT, 67460 { 3420; 1007 M DHT, 41060 { 8080. Day 4 with 2 ng/ml TGF-b1: cFBS, 45600 { 2010; 10012 M DHT, 41200 { 4210; 10010 M DHT, 41600 { 5060; 1007 M DHT, 37330 { 4270. The vertical bar denotes SEM. *P õ 0.05 compared to the corresponding value in cultures grown in the absence of TGF-b1.
AP: Exp Cell
106
KIM ET AL.
FIG. 3. Effect of TGF-b1 and DHT on [3H]thymidine incorporation. Cells were seeded at 2 1 104/well in 24-well plates and cultured in medium containing TGF-b1 (2 ng/ml) and following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M. At 24 and 48 h in culture, [3H]thymidine (1 mCi/ml) was added and incubated for an additional 4 h. Radioactivity in the acid insoluble fraction was counted using a liquid scintillation counter. The results are presented as percentage of control. Each point represents the mean of triplicate observations. The actual counts per minute (cpm) were as follow. 0 h: cFBS, 40220 { 3239; 10012 M DHT, 43794 { 4028; 10010 M DHT, 65367 { 1066; 1007 M DHT, 28440 { 1313. 24 h with no TGF-b1 (control): cFBS, 40220 { 3239; 10012 M DHT, 43794 { 4028; 10010 M DHT, 65367 { 1066; 1007 M DHT, 28440 { 1313. 24 h with 2 ng/ml TGF-b1: cFBS, 39717 { 603; 10012 M DHT, 40729 { 2803; 10010 M DHT, 34000 { 1111; 1007 M DHT, 23918 { 398. 48 h with no TGF-b1 (control): cFBS, 39955 { 7913; 10012 M DHT, 39346 { 6838; 10010 M DHT, 67781 { 6664; 1007 M DHT, 13639 { 2801. 48 h with 2 ng/ml TGF-b1: cFBS, 41670 { 3120; 10012 M DHT, 41155 { 2821; 10010 M DHT, 47253 { 3171; 1007 M DHT, 15528 { 405. The vertical bar denotes SEM. *P õ 0.05 compared with the corresponding value in cultures grown in the absence of TGF-b1.
[3H]thymidine incorporation in cultures grown only at 10010 M DHT. The level of DNA synthesis in cultures grown at this dose of DHT decreased to Ç40% of control by 24 h after initiation of TGF-b1 treatment. After 48 h of TGF-b1 treatment, [3H]thymidine incorporation increased slightly to Ç50% of control. However, the increase was not statistically significant. The cytotoxic effect of TGF-b1 in LNCaP cells was investigated by performing trypan blue dye exclusion assay at the end of a 2-day treatment with 2 ng/ml TGF-b1 (Table 1). In the absence of TGF-b1 treatment, LNCaP viability varied according to the level of DHT used in the culture and followed a bell-shaped dose– response curve as observed earlier [19]. In the presence of 2 ng/ml TGF-b1, a statistically significant decrease in cell viability occurred only at 10010 M DHT (Ç60% of control). Effect of TGF-b1 and DHT on LNCaP morphology. Figure 4 is a collection of representative photomicro-
/ m4678$3013
12-08-95 13:08:40
eca
graphs showing the effect of TGF-b1 on LNCaP cell morphology. The normal morphology of LNCaP cells, a typical spindle shape with occasional pseudopodiumlike extensions, was observed in the presence of either 10010 M DHT or 2 ng/ml TGF-b1 alone (Figs. 4a–4c, and 4e). The normal LNCaP cell morphology was also observed at all other concentrations of DHT (0, 10012, and 1007 M) either in the presence or absence of 2 ng/ ml TGF-b1 (data not shown). However, alteration in the morphology of these cells appeared when they were treated with the combination of 10010 M DHT and 2 ng/ ml TGF-b1 (Figs. 4d and 4f). These changes included shrinkage of cytoplasm and increased frequency of cytosolic vacuoles. Effect of TGF-b1 and DHT on activation of a TGFb-responsive promoter. In addition to decrease in cell proliferation, decrease in cell viability, and alteration in cell morphology, activation of a TGF-b-responsive promoter that is present in the plasmid, p3TP-Lux, is also a distinct effect of TGF-b1 [16]. Thus, to further support the above results regarding TGF-b1 sensitivity in LNCaP cells, they were transiently transfected with the plasmid p3TP-Lux. Because this plasmid contains the luciferase reporter gene that is under the control of a TGF-b-responsive promoter composed of three TPAresponsive elements from the human collagenase gene and one TGF-b-responsive element from the human PAI-1 gene [16], luciferase activity was assayed as an extent of LNCaP response to TGF-b1. Following the transfection, cells were treated with varying doses of DHT in the presence or absence of 5 ng/ml TGF-b1. The levels of luciferase activity in cultures that were not treated with TGF-b1 were used as respective controls. Figure 5 shows that, in the presence of TGF-b1, treatment of LNCaP cells with 10010 M DHT yielded the highest level of induction of luciferase activity (Ç300% of control). These results indicate that, as far TABLE 1 Effect of TGF-b1 on Viability of LNCaP Cells % viable cellsa Concentrations of DHT (M) 0 (cFBS) 10012 10010 1007
no TGF-b1 (control) 92.5 90.4 96.0 83.7
{ { { {
0.87 1.00 0.29 1.09c
/TGF-b1 (2 ng/ml) 90.8 91.0 61.2 84.3
{ { { {
1.30 1.04 2.73b 2.92
% of control 98.2 100.7 63.8 100.7
{ { { {
1.39 1.13 2.85 3.50
Note. Each value represents mean { SEM of three separate experiments. a Cell viability was assessed by the trypan blue dye exclusion test. b The value is significantly different from the corresponding value in control group (P õ 0.05). c The value is significantly different from the corresponding value in cFBS (P õ 0.05).
AP: Exp Cell
MODULATION OF TGF-b1 SENSITIVITY IN LNCaP BY DHT
107
FIG. 4. Effect of DHT and TGF-b1 on LNCaP cell morphology. The cells were treated with TGF-b1 (2 ng/ml) and zero DHT control (cFBS) or 10010 M DHT for 48 h. (a) Zero DHT control without TGF-b1 (331); (b) 10010 M DHT without TGF-b1 (331); (c) zero DHT control with TGF-b1 (331); (d) 10010 M DHT with TGF-b1 (331); (e) zero DHT control with TGF-b1 (661); (f) 10010 M DHT with TGF-b1 (661).
as activation of a TGF-b-responsive promoter is concerned, sensitivity to TGF-b1 in LNCaP is the highest when they are treated with 10010 M DHT. Effect of DHT on binding of 125I-TGF-b1 in LNCaP cells. Results of the above experiments have indicated that, by four different but complementary criteria, the sensitivity of LNCaP cells to TGF-b1 can be best demonstrated when cells are treated with 10010 M DHT. Because this sensitivity may be reflected by LNCaP’s ability to bind TGF-b1, we investigated the level of TGF-b1-specific binding under varying con-
/ m4678$3013
12-08-95 13:08:40
eca
centrations of DHT. Table 2 shows the result of a competition binding assay between 125I-labeled and unlabeled TGF-b1. LNCaP cells had a detectable level of TGF-b1-specific binding only at 10010 M DHT (Ç8.45 fmol/mg proteins). However, the level of TGF-b1 binding was not abundant enough to warrant Scatchard analysis. Effect of DHT on the expression of TGF-b receptors. To further characterize the TGF-b1-specific binding site(s) detected in LNCaP cells at 10010 M DHT, we performed Western blot analysis for TGF-b receptor
AP: Exp Cell
108
KIM ET AL.
FIG. 5. Effect of TGF-b1 and DHT on gene transcriptional activation of p3TP-Lux. Cells were seeded at 1 1 105/well in six-well plates and cultured for 2 days with the following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M DHT. Subsequently, the cells were transfected transiently with p3TP-Lux by lipofection. At 48 h after the transfection, cells were incubated with 5 ng/ml of TGF-b1 for 16 h. Luciferase activity was measured with a Luminometer and expressed as percentage control. Each column represents the mean of triplicate observations. The vertical bar denotes SEM. *P õ 0.05 when compared with the corresponding value in cultures grown in the absence of TGF-b1.
type II based on the understanding that type II receptor is mainly responsible for ligand binding [20, 21]. Figure 6a shows a dose-dependent increase in type II receptor protein at DHT concentrations between 0 and 10010 M. At DHT concentrations greater than 10010 M, the level of type II receptor protein decreased. Regarding the mechanism of action of TGF-b receptors, it has been postulated that both type I and type II TGF-b receptors are required for signaling [20]. Thus, we also conducted Western blot analysis for type I receptor. Figure 6b shows that the expected 55-kDa immunoreactive band was observed in PC 3 cells but not in LNCaP cells. Since we have demonstrated that TGFb1 signaling occurs in LNCaP cells, it is probable that type I receptor is present in these cells but at a level not detectable by Western blot analysis.
of TGF-b1 are realized only in the presence of 10010 M DHT in the culture medium. In addition, at this concentration of DHT, the presence of specific binding for TGF-b1 and an increase in TGF-b receptor type II in these cells have further supported the existence of TGF-b sensitivity in LNCaP cells. These observations, taken together, provide a valuable insight into the growth control mechanisms of LNCaP cells, in vitro. LNCaP cells express a functional, but mutated, androgen receptor that is sensitive not only to androgen but also to anti-androgens, estrogen, and progestins [8, 9]. Nevertheless, because the androgen receptor has a normal DNA binding domain, the cell line has been used widely as a model of androgen-responsive prostate cancer to study the cellular mechanism of androgen action [22–24]. The present observations have provided evidence to indicate that DHT can modulate the sensitivity of LNCaP cells to TGF-b1. Four concentrations of DHT were used in the present study: 0, 10012, 10010, and 1007 M. These concentrations were selected because they represent the zero DHT control, the lowproliferative dose, the high-proliferative dose, and the growth-arrest dose, respectively [19]. The results of the present study demonstrated that TGF-b1 treatment decreased cell count in a dose-dependent manner only at 10010 M DHT; the lowest level of cell count (Ç40% of control) was achieved at 20 and 200 ng/ml of TGFb1. In addition, in cultures with 2 ng/ml of TGF-b1 and 10010 M DHT, [3H]thymidine incorporation was Ç40% of control, cell viability was Ç60% of control, and cell morphology was altered. Finally, treating these cells with TGF-b1 and 10010 M DHT resulted in Ç3-fold increase in the activity of a TGF-b-responsive promoter present in p3TP-Lux. These series of independent but complementary observations demonstrated that LNCaP cells responded to TGF-b1 only in the presence of the high-proliferative dose of DHT, 10010 M. At other concentrations of DHT, the effects of TGF-b1 were minimal and statistically insignificant. Therefore, these re-
DISCUSSION
TABLE 2
Results of the present study have demonstrated that DHT modulates the sensitivity of LNCaP cells to TGFb1 in culture. The effects of TGF-b1 on these cells have been shown by four different but complementary criteria: (i) inhibition of cell growth, (ii) reduction in cell viability, (iii) alteration in cell morphology, and (iv) activation of a TGF-b-responsive promoter. These different approaches were employed to establish the sensitivity of LNCaP cells to TGF-b1 for two reasons. First, previous reports have indicated that these cells do not respond to TGF-b1 treatment [2, 4]. Second, some of the effects of TGF-b1 observed in this study were modest. The present results showed that the various effects
TGF-b1-Specific Binding in LNCaP Cells
/ m4678$3013
12-08-95 13:08:40
eca
Concentrations of DHT (M)
TGF-b1-specific binding (fmol/mg protein)a
0 (cFBS) 10010 1007
Undetectable 8.45 { 1.27 Undetectable
Note. Each value represents mean { SEM of three separate experiments. a Specific binding of 125I-TGF-b1 was determined by the difference in radioactivity between samples in the presence (nonspecific binding) and absence (total binding) of 100-fold molar excess of unlabeled TGF-b1. The specific binding in radioactive counts was then converted to fmol/mg of proteins.
AP: Exp Cell
MODULATION OF TGF-b1 SENSITIVITY IN LNCaP BY DHT
109
FIG. 6. TGF-b receptors type II and type I expression in LNCaP cells. Cells were cultured in varying concentrations of DHT for 2 days. Subsequently, 100 mg of each protein sample was separated by electrophoresis in 10% SDS–polyacrylamide gel, and immunoreactive bands were visualized by enhanced chemiluminescence. (a) Expression of type II receptor. Lane 1, cFBS; lane 2, 10012 M DHT; lane 3, 10011 M DHT; lane 4, 10010 M DHT; lane 5, 1009 M DHT; lane 6, 1008 M DHT; lane 7, 1007 M DHT. (b) Expression of type I receptor. Lane 1, PC 3 prostate cancer cell line; lane 2, cFBS; lane 3, 10012 M DHT; lane 4, 10011 M DHT; lane 5, 10010 M DHT; lane 6, 1009 M DHT; lane 7, 1008 M DHT; lane 8, 1007 M DHT.
sults indicate that DHT modulates TGF-b1 sensitivity in LNCaP cells. The influence of DHT on the TGF-b1 responsiveness in LNCaP cells can provide a partial explanation for the inconsistencies in the literature regarding TGF-b1 sensitivity in these cells. Wilding et al. and Carruba et al. reported that TGF-b1 did not affect the proliferation of LNCaP cells and that no TGF-b1-specific binding was detectable in these cells [2, 5]. On the other hand, Schuurmans et al. showed that TGF-b1 blunted the growth stimulating effects of epidermal growth factor (EGF) and transforming growth factor-a (TGF-a) [4]. These observations are consistent with the concept that the sensitivity of LNCaP cells to TGF-b1 can be manifested only under certain culture conditions. An inspection of these conditions suggests that TGF-b1 sensitivity is associated with a proliferative state of these cells. It is known that EGF and TGF-a promote LNCaP proliferation [26, 27]. In the present study, LNCaP cells became sensitive to TGF-b1 only in the presence of the high-proliferative dose of DHT. Therefore, the concept that LNCaP cells are sensitive to the action of TGF-b1 under a proliferative state warrants further investigation. Although the mechanism through which LNCaP cells acquire sensitivity to TGF-b1 at 10010 M DHT remains unclear, it is likely that the mechanism involves some components of the signaling system that are necessary for LNCaP cells to respond to TGF-b1. It has been demonstrated that TGF-b signaling is mediated through the interaction with membrane receptors. There are three different types of ubiquitously expressed TGF-b receptors: type I, II, and III [reviewed in 28]. Of these, type I and type II receptors are directly involved in TGF-b signal transduction [reviewed in 21, 28]. Current understanding states that type II receptor
/ m4678$3013
12-08-95 13:08:40
eca
is necessary for ligand binding and for subsequent formation of a heteromeric complex with type I receptor before TGF-b signaling can occur [15, 19]. Results of the present study demonstrated that DHT, at 10010 M, induced specific binding of TGF-b1, although at a low level, and increased the cellular contents of type II receptor. Since type II receptor has been shown to be responsible for the ligand binding [19, 20], the increase in type II receptor is consistent with the appearance of TGF-b1-specific binding at 10010 M DHT. Furthermore, since type II receptor is necessary for TGF-b signaling, an increase in the level of the receptor at 10010 M DHT will likely lead to an increased capacity to respond to TGF-b. An investigation to determine the biological significance of induction of type II receptor protein by 10010 M DHT in LNCaP cells will be pursued in our future studies. Efforts to detect type I receptor in LNCaP cells were not successful in the present study. Nevertheless, in this study, the effects of TGF-b1 signaling were demonstrated by four different but complementary criteria. These results suggest that the machinary for TGF-b signaling within LNCaP cells is intact when DHT is present at the dose of 10010 M. In summary, results of the present study have provided evidence to demonstrate that LNCaP cells are sensitive to TGF-b1 and express increased level of TGF-b receptor type II protein when the high-proliferative dose of DHT is added to the culture medium. These results indicate that DHT can modulate both TGF-b1 sensitivity and type II receptor protein level in LNCaP cells in culture. The present system, therefore, can be used as a model for the study of interaction between androgen and TGF-b1 in prostate cancer cells. We thank Dr. Jon Lomasney of Northwestern University Medical
AP: Exp Cell
110
KIM ET AL.
School (Chicago, IL) for his help in 125I-TGF-b1 binding assay and in reviewing the manuscript, and Dr. Joan Massague of Memorial Sloan-Kettering Cancer Center (New York, NY) for kindly providing p3TP-Lux. Finally, we thank Dr. Kohei Miyazono of Ludwig Institute for Cancer Research (Uppsala, Sweden) for providing TbRI and TbRII antibodies and experimental protocols.
REFERENCES 1. Wilding, G. (1991) Cancer Surv. 11, 147–163. 2. Wilding, G., Zugmeier, G., Knabbe, C., Flanders, K., and Gelmann, E. (1989) Mol. Cell. Endocrinol. 62, 79–87. 3. Franzen, P., Ichijo, H., and Miyazono, K. (1993) Exp. Cell Res. 207, 1–7. 4. Schuurmans, A. L. G., Bolt, J., Veldscholte, J., and Mulder, E. (1991) J. Steroid Biochem. Mol. Biol. 40, 193–197. 5. Carruba, G., Leake, R. E., Rinaldi, F., Chalmers, D., Comito, L., Sorci, C., Pavone-Macaluso, M., and Castagnetta, L. A. M. (1994) Steroids 59, 412–420. 6. Horoszewicz, J. S., Leong, S. S., Kawinski, E., Karr, J. P., Rosenthal, H., Chu, M. T., Mirand, E. A., and Murphy, G. P. (1983) Cancer Res. 43, 1809–1817. 7. Hiipakka, R. A., and Liao, S. (1995) in Androgen Receptors and Action. (Degroot, L. J., Ed.), pp. 2336–2351, Saunders, New York. 8. Wilding, G., Chen, M., and Gelmann, E. P. (1989) The Prostate 14, 103–115. 9. Kokontis, J., Ito, K., Hipakka, R. A., and Liao, S. (1991) Receptor 1, 271–279. 10. Taplin, M. E., Bubley, G. J., Shuster, T. D., Frantz, M. E., Spooner, A. E., Ogata, G. K., Keer, H. N., and Balk, S. P. (1995) New Engl. J. Med. 332, 1393–1398. 11. Sonnenschein, C., Olea, N., Pasanen, M. E., and Soto, A. M. (1989) Cancer Res. 49, 3474–3481.
12. Kyprianou, N., and Isaacs, J. T. (1989) Mol. Endocrinol. 3, 1515–1522. 13. Kyprianou, N., and Isaacs, J. T. (1988) Endocrinology 123, 2124–2131. 14. Knabbe, C., Klein, H., Zugmaier, G., and Voigt, K. D. (1993) J. Steroid. Biochem. Mol. Biol. 47, 137–142. 15. Fong, C. J., Sherwood, E. R., Braun, E. J., Berg, L. A., Lee, C., and Kozlowski, J. M. (1992) The Prostate 21, 121–131. 16. Wrana, J. L., Attisano, L., Carcamo, J., Zentella, A., Doody, J., Laiho, M., Wang, X., and Massague, J. (1992) Cell 71, 1003– 1014. 17. Franzen, P., ten Dijke, P., Ichijo, H., Yamashita, H., Schultz, P., Heldin, C.-H., and Miyazono, K. (1993) Cell 75, 681–692. 18. Steele, R. E. D., and Torrie, J. H. (1960) Principles and Procedures of Statistics, pp. 99–131. McGraw–Hill, New York. 19. Lee, C., Sutkowski, D. M., Sensibar, J. A., Zelner, D., Kim, I., Amsel, I., Shaw, N., Prins, G. S., and Kozlowski, J. M. (1995) Endocrinology 136, 796–803. 20. Wrana, J. L., Attisano, L., Wieser, R., Ventura, F., and Massague, J. (1994) Nature 370, 341–347. 21. Derynck, R. (1994) TIBS 19, 548–553. 22. Veldscholte, J., Berrevoets, C. A., and Mulder, E. (1994) J. Steroid Biochem. Mol. Biol. 49, 341–346. 23. Henttu, P., and Vihko, P. (1993) Cancer Res. 53, 1051–1058. 24. Krongrad, A., Wilson, C. M., Wilson, J. D., Allman, D. R., and McPhaul, M. J. (1991) Mol. Cell. Endocrinol. 76, 79–88. 25. Gleave, M. E., Hsieh, J-T., Wu, H-C., Eschenbach, A. C., and Chung, L. W. K. (1992) Cancer Res. 52, 1598–1605. 26. Macdonald, A., and Habib, F. K. (1992) Br. J. Cancer 65, 177– 182. 27. Limonta, P., Moretti, R. M., Dondi, D., Marelli, M. M., and Motta, M. (1994) J. Steroid Biochem. Mol. Biol. 49, 347–350. 28. Kingsley, D. M. (1994) Genes Dev. 8, 133–146.
Received July 25, 1995 Revised version received September 11, 1995
/ m4678$3013
12-08-95 13:08:40
eca
AP: Exp Cell
222, 103–110 (1996)
Article No. 0013
Modulation of Sensitivity to Transforming Growth Factor-b1 (TGF-b1) and the Level of Type II TGF-b Receptor in LNCaP Cells by Dihydrotestosterone1 ISAAC YI KIM, DAVID J. ZELNER, JULIA A. SENSIBAR, HAN-JONG AHN, LINDA PARK, JIN-HO KIM, AND CHUNG LEE2 Department of Urology, Northwestern University Medical School, Chicago, Illinois 60611
Transforming growth factor-b1 (TGF-b1) and androgen are potential physiological regulators of prostate cancer cells. In the present study, we have used LNCaP cells as a model of androgen-responsive prostate cancer to investigate the effects of dihydrotestosterone (DHT) on the sensitivity to TGF-b1. The ability of LNCaP cells to respond to TGF-b has been controversial. In some studies, LNCaP cells were insensitive to TGF-b1 while, in others, they were sensitive to the growth inhibitory effect of TGF-b1. The present study was carried out to establish androgenic conditions that rendered LNCaP cells sensitive to TGF-b1. Cells were cultured in phenol-red-free RPMI 1640 medium supplemented with 10% charcoal-stripped fetal bovine serum. DHT was added at the following concentrations: 0, 10012, 10010, and 1007 M. These concentrations were selected because they represent the zero DHT control, the low-proliferative dose, the high-proliferative dose, and the growth-arrest dose, respectively. The effects of TGF-b1 observed on LNCaP cells included inhibition of cell proliferation, decrease in cell viability, alteration in cell morphology, and enhancement of gene transcriptional activity through activation of a TGF-b responsive promoter. Of the various DHT concentrations investigated in this study, these effects of TGF-b1 on LNCaP cells were consistently demonstrated only at 10010 M. At other concentrations, the effects of TGF-b1 were either minimal or undetectable. Accompanying these effects of TGF-b1, a low but statistically significant level of TGF-b1-specific binding and an increased protein level of TGF-b receptor type II were detected by a competitive binding assay and Western blot analysis, respectively. These results indicate that LNCaP cells can be induced by DHT to respond to TGF-b1 and that DHT modulates the sensi-
1
This work was supported in part by NIH Grants DK 39250 and CA 60553. 2 To whom correspondence and reprint requests should be addressed at Tarry 11-715, Department of Urology, Northwestern University Medical School, 303 E. Chicago Avenue, Chicago, IL 60611.
tivity to TGF-b1 and the level of TGF-b receptor type II in these cells. q 1996 Academic Press, Inc.
INTRODUCTION
Transforming growth factor-b (TGF-b) is a 25-kDa pleiotropic growth factor that has been shown to inhibit the proliferation of most cells, especially those of the epithelial lineage [1]. Among prostate cancer cell lines, TGF-b has been demonstrated to inhibit the proliferation of PC3 and DU145 cells [2, 3]. These findings suggest that TGF-b is a potential regulator of prostate tumor biology. In LNCaP cells, however, the effects of TGF-b is uncertain because some investigators observed an inhibitory effect while others detected no effect [2, 4, 5]. LNCaP cells are androgen-sensitive and are used extensively as an in vitro model for the study of androgen action in the growth of prostate cancer cells [6, 7]. Although LNCaP cells have a mutated androgen receptor at the androgen binding domain and respond not only to androgens but also to anti-androgens, estrogen, and progestins [8, 9], they remain a useful model system for prostate cancer since androgen receptor mutations are a frequent event in many prostate cancer cases [10]. Recent observations have demonstrated that LNCaP cells exhibit a biphasic growth response with androgen stimulation [11]. At low concentrations of dihydrotestosterone (DHT) (10012 to 10010 M), the cells proliferate in a dose-dependent manner. However, at concentrations greater than 1009 M DHT, the rate of proliferation declines in a dose-dependent manner. These observations suggest that DHT may be able to elicit a different biological response in LNCaP cells depending on the concentration used in the culture medium. One of the biological responses that DHT may modulate in LNCaP cells may be the sensitivity to the growth inhibitor TGF-b1 because recent reports have indicated that androgen can profoundly influence TGFb action [12–14].
103
/ m4678$3013
12-08-95 13:08:40
eca
0014-4827/96 $12.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP: Exp Cell
104
KIM ET AL.
Since both androgens and TGF-b are potential physiological regulators of prostate tumor progression, an investigation of the interaction between these two factors will lead to further understanding of the biology of prostate cancer. In the present study, we have used LNCaP cells as an in vitro model of androgen-responsive prostate cancer to investigate the possibility that DHT may modulate the sensitivity to TGF-b1. To establish TGF-b1 sensitivity in these cells, we have deliberately used different but complementary approaches because these cells have been previously reported to be insensitive to the action of TGF-b1 [2, 4]. MATERIALS AND METHODS Cell culture. LNCaP and PC 3 cells were purchased from the American Type Culture Collection (Rockville, MD). In this study, all cells were derived from the 48th through the 51st passages. Cells were routinely maintained in RPMI 1640 containing 10% fetal bovine serum (FBS), penicillin (100 U/ml), and streptomycin (100 mg/ml). Phenol-red-free RPMI 1640 supplemented with dextran-treated charcoal-stripped serum (cFBS) was used in each experiment. Crystalline DHT (Sigma Chemical Co., St. Louis, MO) was dissolved in absolute ethanol at a concentration of 1003 M and was reconstituted in the culture medium at concentrations ranging from 1007 to 10012 M. Human TGF-b1 (Collaborative Research, Bedford, MA) was diluted to 2 mg/ml and added to culture medium at preselected concentrations. Mitogenic assay. Cellular proliferation was assayed by cell counts and [3H]thymidine incorporation [15]. For cell counts, LNCaP cells at 5 1 103/well or 1 1 104/well were plated in 24-well culture plates in phenol-red-free RPMI 1640 supplemented with 10% FBS and allowed to adhere for 24 h. Then, the cultures were washed two times with Hank’s balanced salt solution (HBSS) and cells from previously selected wells were harvested and counted to determine the plating efficiency. Cells in remaining wells were cultured for 2 and 4 days in phenol-red-free RPMI 1640 supplemented with cFBS containing different concentrations of DHT in the presence or absence of TGFb1 at 0.2, 2, 20, and 200 ng/ml. For cultures lasting longer than 2 days, medium was changed every other day. To count cells, the medium was removed and the cells were detached with 0.5 ml of 0.25% trypsin. Once detached, the cells were added to 9.5 ml of isotonic solution (Isoton II, Coulter Corp., Hialeah, FL) and counted with a coulter counter (ZF, Coulter Corp., Hialeah, FL). In subsequent studies, unless otherwise specified, a concentration of TGF-b1 at 2 ng/ ml was selected. For [3H]thymidine incorporation assay, LNCaP cells were seeded in 24-well plates as previously described at 2 1 104 cells/well. Once the plating efficiencies were determined, the cells were treated with 2 ng/ml of TGF-b1 at varying concentrations of DHT for 24 and 48 h. After the preselected time of incubation, 1 mCi/ml of [3H]thymidine (6.7 Ci/mM, Amersham Corp., Arlington Heights, IL) was added to each well and incubated for 4 h. The cells were subsequently treated with 1.0 ml of cold 10% trichloroacetic acid. The precipitates were dissolved in 0.5 ml of 0.4 N NaOH, and the amount of radioactivity was counted in a liquid scintillation counter. Cell viability was determined by a trypan blue dye exclusion test. After treating LNCaP cells with 2 ng/ml of TGF-b1 for 2 days, the percentage of viable cells was calculated by the following equation: % viable Å
trypan-blue negative cells 1 100 . total no. of cells
Photomicrographs of cells were taken directly from cultured plates
/ m4678$3013
12-08-95 13:08:40
eca
under an inverted microscope equipped with camera (Olympus OM 4T, Olympus Camera Co., Woodbury, NY) using Ektachrome 64T films. Prints developed from these films were used as permanent hard copy records. Transient transfection and luciferase assay. Cells were seeded as above in six-well plates at 1 1 105 cells/well. After treatment for 2 days in varying concentrations of DHT, cells were transiently transfected with plasmid p3TP-Lux using lipofectamine according to the manufacturer’s directions (Gibco-BRL, Grand Island, NY). p3TPLux contains three TPA responsive elements from the human collagenase gene and one TGF-b responsive element from the human plasminogen activator inhibitor-1 (PAI-1) promoter linked to the luciferase reporter gene [16]. Throughout the transfection protocol, the cells were constantly maintained in phenol-red-free RPMI 1640 with 10% cFBS and preselected concentrations of DHT. Briefly, 1 mg of p3TP-Lux and 12 ml of lipofectamine were added with 1 ml of transfection medium (Opti-mem, Gibco-BRL) to each well and incubated for 24 h. Subsequently, fresh medium was added and the cells were incubated for an additional 24 h. Finally, 5 ng/ml of TGF-b1 was added and the cultures were maintained for an additional 16 h. This concentration of TGF-b1 and culture conditions were selected according to the original protocol [16]. The extent of the promoter activity of p3TP-Lux was assayed by measuring luciferase activity using a commercial luciferase assay kit (enhanced luciferase assay kit 1800k, Analytical Luminescence Laboratory, San Diego, CA) and a luminometer (Monolight 2010B, Analytical Luminescence Laboratory, San Diego, CA). TGF-b1 binding assay (17). Cells were plated as above in 24-well plates at 2 1 104 cells/well. After culturing for 2 days in phenol-redfree RPMI 1640 supplemented with cFBS and varying concentrations of DHT, they were washed two times with phosphate-buffered saline (PBS) and harvested by scraping. Subsequently, the cells were incubated for 3 h with 200 pM of 125I-TGF-b1 (800–2200 Ci/mM, Amersham Corp, Arlington Heights, IL) in the presence (nonspecific binding) or absence (total binding) of 100-fold molar excess of nonradiolabeled TGF-b1 in binding buffer (PBS containing 0.9 mM CaCl2 , 0.49 mM MgCl2 , and 1 mg/ml bovine serum albumin) at 47C. After incubation, cells were applied to circular 3 MM Watman filter papers and washed three times with the same binding buffer without bovine serum albumin. The remaining radioactivity was counted by gamma counter (Cobra II Autogamma, Packard Instrument Company, Downers Grove, IL). Specific binding of 125I-TGF-b1 was determined by calculating the difference in radioactivity between nonspecific and total binding. The difference in radioactivity was converted to fmol of TGF-b1-specific binding per milligrams of protein based on the specific activity of 125I-TGF-b1. Western blot analysis for TGF-b receptors. Cells were homogenized with PBS at 47C and the protein concentration was determined according to the method of Lowry. Samples were placed in sample buffer (0.0625 M Trizma base, 2% SDS, 5% 2-mercaptoethanol) and boiled for 5 min. Electrophoresis was carried out in 10% SDS–polyacrylamide gel using 100 mg of total protein loaded onto each lane. Following electrophoresis, proteins were transferred to a 0.2-mm nitrocellulose membrane (Bio-Rad). The membrane was incubated overnight in blocking buffer (5% nonfat dry milk, PBS, and 0.1% Tween). Subsequently the membrane was incubated with anti-type II receptor and anti-type I receptor antibodies [17] at a dilution of 1:800 overnight at 47C. After washing with 0.1% PBS–Tween, the membrane was incubated in the presence of goat anti-rabbit horseradish peroxidase-labeled secondary antibody (Bio-Rad) at a dilution of 1:8000 for 4 h at room temperature. After washing several times, immunoreactive proteins were visualized by enhanced chemiluminescence (Amersham) and exposed to Hybond-ECL film (Amersham). Statistical analysis. All numerical data are expressed as mean { SEM with triplicate observations. Differences of means among different treatments were compared by unpaired Student’s t test [18]. A value of P õ 0.05 was considered statistically significant.
AP: Exp Cell
MODULATION OF TGF-b1 SENSITIVITY IN LNCaP BY DHT
FIG. 1. Effect of increasing doses of TGF-b1 on cell numbers. Cells were seeded at 5 1 103/well in 24-well plates and cultured in medium containing TGF-b1 at 0, 0.2, 2, 20, and 200 ng/ml and the following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M. Cells were counted at the end of 2 days by Coulter counter. The results are presented as percentage of control. The value in each group represents the mean of triplicate observations. The actual cell numbers were as following. 0 ng/ml TGF-b1 (control): cFBS, 10160 { 528; 10012 M DHT, 10800 { 680; 10010 M DHT, 15680 { 512; 1007 M DHT, 9600 { 584. 0.2 ng/ml TGF-b1: cFBS, 10240 { 203; 10012 M DHT, 11280 { 304; 10010 M DHT, 13040 { 696; 1007 M DHT, 9200 { 408. 2 ng/ml TGF-b1: cFBS, 10160 { 576; 10012 M DHT, 11120 { 304; 10010 M DHT, 9040 { 472; 1007 M DHT, 9440 { 600. 20 ng/ml TGF-b1: cFBS, 10320 { 440; 10012 M DHT, 10640 { 648; 10010 M DHT, 6720 { 256; 1007 M DHT, 9360 { 408. 200 ng/ ml TGF-b1: cFBS, 10320 { 440; 10012 M DHT, 10640 { 648; 10010 M DHT, 6720 { 256; 1007 M DHT, 9360 { 408. The vertical bar denotes SEM. *P õ 0.05 compared to the corresponding value in cultures grown in the absence of TGF-b1.
105
creased the cell count only at 10010 M DHT after 2 and 4 days of treatment. After 2 days of TGF-b1 treatment at 2 ng/ml, the cell count at 10010 M DHT was approximately 58% of control. Such a magnitude of decrease in cell count is comparable to that of other prostate cancer cell lines when they were treated with a similar dose of TGF-b1 [2, 3]. Following the initial decrease at Day 2, a similar magnitude of decrease in cell count was observed after 4 days of TGF-b1 treatment at 10010 M DHT. Effect of TGF-b1 and DHT on DNA synthesis and cell death. To analyze the effect of TGF-b1 on the active process of DNA synthesis, changes in [3H]thymidine incorporation were investigated. Based on the above study, in which a significant decrease in cell count was detectable after 2 days of TGF-b1 treatment, the present study was done in an attempt to investigate whether an effect can be detected at 24 and 48 h after the addition of 2 ng/ml TGF-b1 under varying concentrations of DHT. Figure 3 shows that TGF-b1 inhibited
RESULTS
Effect of TGF-b1 and DHT on cell number. Figure 1 shows the effect of increasing doses of TGF-b1 on the LNCaP cell count after 2 days of treatment in the presence of varying concentrations of DHT. Changes in the cell count were observed only when the cells were treated with TGF-b1 in the presence of 10010 M DHT. A slight decrease in the cell count was observed following TGF-b1 treatment at 0.2 ng/ml. However, a drastic decrease in cell count was observed when the cells were treated with TGF-b1 at concentrations of 2 ng/ml or greater. In the presence of 20 and 200 ng/ml TGF-b1, the cell count decreased to approximately 40% of control. In subsequent experiments, unless otherwise specified, a single dose of 2 ng/ml TGF-b1 was used. Figure 2 shows the change in cell number after LNCaP cultures were treated with 2 ng/ml TGF-b1 under varying concentrations of DHT for 2 and 4 days. Again, the results show that TGF-b1 significantly de-
/ m4678$3013
12-08-95 13:08:40
eca
FIG. 2. Effect of TGF-b1 and DHT on cell numbers. Cells were seeded at 1 1 104/well in 24-well plates and cultured in RPMI 1640 supplemented with 10% cFBS and following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M. To preselected wells, 2 ng/ml TGF-b1 was added. Media were changed every 2 days. Cells were counted at the end of the 2- and 4-day period by Coulter counter. The results are presented as percentage of control. The value in each group represents the mean of triplicate observations. The actual cell numbers were as following. Day 2 with no TGF-b1 (control): cFBS, 36267 { 2010; 10012 M DHT, 37060 { 1180; 10010 M DHT, 57330 { 2270; 1007 M DHT, 41330 { 2460. Day 2 with 2 ng/ ml TGF-b1: cFBS, 34000 { 4380; 10012 M DHT, 31460 { 3750; 10010 M DHT, 33460 { 4160; 1007 M DHT, 42400 { 2830. Day 4 with no TGF-b1 (control): cFBS, 42000 { 3600; 10012 M DHT, 45200 { 2880; 10010 M DHT, 67460 { 3420; 1007 M DHT, 41060 { 8080. Day 4 with 2 ng/ml TGF-b1: cFBS, 45600 { 2010; 10012 M DHT, 41200 { 4210; 10010 M DHT, 41600 { 5060; 1007 M DHT, 37330 { 4270. The vertical bar denotes SEM. *P õ 0.05 compared to the corresponding value in cultures grown in the absence of TGF-b1.
AP: Exp Cell
106
KIM ET AL.
FIG. 3. Effect of TGF-b1 and DHT on [3H]thymidine incorporation. Cells were seeded at 2 1 104/well in 24-well plates and cultured in medium containing TGF-b1 (2 ng/ml) and following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M. At 24 and 48 h in culture, [3H]thymidine (1 mCi/ml) was added and incubated for an additional 4 h. Radioactivity in the acid insoluble fraction was counted using a liquid scintillation counter. The results are presented as percentage of control. Each point represents the mean of triplicate observations. The actual counts per minute (cpm) were as follow. 0 h: cFBS, 40220 { 3239; 10012 M DHT, 43794 { 4028; 10010 M DHT, 65367 { 1066; 1007 M DHT, 28440 { 1313. 24 h with no TGF-b1 (control): cFBS, 40220 { 3239; 10012 M DHT, 43794 { 4028; 10010 M DHT, 65367 { 1066; 1007 M DHT, 28440 { 1313. 24 h with 2 ng/ml TGF-b1: cFBS, 39717 { 603; 10012 M DHT, 40729 { 2803; 10010 M DHT, 34000 { 1111; 1007 M DHT, 23918 { 398. 48 h with no TGF-b1 (control): cFBS, 39955 { 7913; 10012 M DHT, 39346 { 6838; 10010 M DHT, 67781 { 6664; 1007 M DHT, 13639 { 2801. 48 h with 2 ng/ml TGF-b1: cFBS, 41670 { 3120; 10012 M DHT, 41155 { 2821; 10010 M DHT, 47253 { 3171; 1007 M DHT, 15528 { 405. The vertical bar denotes SEM. *P õ 0.05 compared with the corresponding value in cultures grown in the absence of TGF-b1.
[3H]thymidine incorporation in cultures grown only at 10010 M DHT. The level of DNA synthesis in cultures grown at this dose of DHT decreased to Ç40% of control by 24 h after initiation of TGF-b1 treatment. After 48 h of TGF-b1 treatment, [3H]thymidine incorporation increased slightly to Ç50% of control. However, the increase was not statistically significant. The cytotoxic effect of TGF-b1 in LNCaP cells was investigated by performing trypan blue dye exclusion assay at the end of a 2-day treatment with 2 ng/ml TGF-b1 (Table 1). In the absence of TGF-b1 treatment, LNCaP viability varied according to the level of DHT used in the culture and followed a bell-shaped dose– response curve as observed earlier [19]. In the presence of 2 ng/ml TGF-b1, a statistically significant decrease in cell viability occurred only at 10010 M DHT (Ç60% of control). Effect of TGF-b1 and DHT on LNCaP morphology. Figure 4 is a collection of representative photomicro-
/ m4678$3013
12-08-95 13:08:40
eca
graphs showing the effect of TGF-b1 on LNCaP cell morphology. The normal morphology of LNCaP cells, a typical spindle shape with occasional pseudopodiumlike extensions, was observed in the presence of either 10010 M DHT or 2 ng/ml TGF-b1 alone (Figs. 4a–4c, and 4e). The normal LNCaP cell morphology was also observed at all other concentrations of DHT (0, 10012, and 1007 M) either in the presence or absence of 2 ng/ ml TGF-b1 (data not shown). However, alteration in the morphology of these cells appeared when they were treated with the combination of 10010 M DHT and 2 ng/ ml TGF-b1 (Figs. 4d and 4f). These changes included shrinkage of cytoplasm and increased frequency of cytosolic vacuoles. Effect of TGF-b1 and DHT on activation of a TGFb-responsive promoter. In addition to decrease in cell proliferation, decrease in cell viability, and alteration in cell morphology, activation of a TGF-b-responsive promoter that is present in the plasmid, p3TP-Lux, is also a distinct effect of TGF-b1 [16]. Thus, to further support the above results regarding TGF-b1 sensitivity in LNCaP cells, they were transiently transfected with the plasmid p3TP-Lux. Because this plasmid contains the luciferase reporter gene that is under the control of a TGF-b-responsive promoter composed of three TPAresponsive elements from the human collagenase gene and one TGF-b-responsive element from the human PAI-1 gene [16], luciferase activity was assayed as an extent of LNCaP response to TGF-b1. Following the transfection, cells were treated with varying doses of DHT in the presence or absence of 5 ng/ml TGF-b1. The levels of luciferase activity in cultures that were not treated with TGF-b1 were used as respective controls. Figure 5 shows that, in the presence of TGF-b1, treatment of LNCaP cells with 10010 M DHT yielded the highest level of induction of luciferase activity (Ç300% of control). These results indicate that, as far TABLE 1 Effect of TGF-b1 on Viability of LNCaP Cells % viable cellsa Concentrations of DHT (M) 0 (cFBS) 10012 10010 1007
no TGF-b1 (control) 92.5 90.4 96.0 83.7
{ { { {
0.87 1.00 0.29 1.09c
/TGF-b1 (2 ng/ml) 90.8 91.0 61.2 84.3
{ { { {
1.30 1.04 2.73b 2.92
% of control 98.2 100.7 63.8 100.7
{ { { {
1.39 1.13 2.85 3.50
Note. Each value represents mean { SEM of three separate experiments. a Cell viability was assessed by the trypan blue dye exclusion test. b The value is significantly different from the corresponding value in control group (P õ 0.05). c The value is significantly different from the corresponding value in cFBS (P õ 0.05).
AP: Exp Cell
MODULATION OF TGF-b1 SENSITIVITY IN LNCaP BY DHT
107
FIG. 4. Effect of DHT and TGF-b1 on LNCaP cell morphology. The cells were treated with TGF-b1 (2 ng/ml) and zero DHT control (cFBS) or 10010 M DHT for 48 h. (a) Zero DHT control without TGF-b1 (331); (b) 10010 M DHT without TGF-b1 (331); (c) zero DHT control with TGF-b1 (331); (d) 10010 M DHT with TGF-b1 (331); (e) zero DHT control with TGF-b1 (661); (f) 10010 M DHT with TGF-b1 (661).
as activation of a TGF-b-responsive promoter is concerned, sensitivity to TGF-b1 in LNCaP is the highest when they are treated with 10010 M DHT. Effect of DHT on binding of 125I-TGF-b1 in LNCaP cells. Results of the above experiments have indicated that, by four different but complementary criteria, the sensitivity of LNCaP cells to TGF-b1 can be best demonstrated when cells are treated with 10010 M DHT. Because this sensitivity may be reflected by LNCaP’s ability to bind TGF-b1, we investigated the level of TGF-b1-specific binding under varying con-
/ m4678$3013
12-08-95 13:08:40
eca
centrations of DHT. Table 2 shows the result of a competition binding assay between 125I-labeled and unlabeled TGF-b1. LNCaP cells had a detectable level of TGF-b1-specific binding only at 10010 M DHT (Ç8.45 fmol/mg proteins). However, the level of TGF-b1 binding was not abundant enough to warrant Scatchard analysis. Effect of DHT on the expression of TGF-b receptors. To further characterize the TGF-b1-specific binding site(s) detected in LNCaP cells at 10010 M DHT, we performed Western blot analysis for TGF-b receptor
AP: Exp Cell
108
KIM ET AL.
FIG. 5. Effect of TGF-b1 and DHT on gene transcriptional activation of p3TP-Lux. Cells were seeded at 1 1 105/well in six-well plates and cultured for 2 days with the following concentrations of DHT: zero DHT control (cFBS), 10012, 10010, and 1007 M DHT. Subsequently, the cells were transfected transiently with p3TP-Lux by lipofection. At 48 h after the transfection, cells were incubated with 5 ng/ml of TGF-b1 for 16 h. Luciferase activity was measured with a Luminometer and expressed as percentage control. Each column represents the mean of triplicate observations. The vertical bar denotes SEM. *P õ 0.05 when compared with the corresponding value in cultures grown in the absence of TGF-b1.
type II based on the understanding that type II receptor is mainly responsible for ligand binding [20, 21]. Figure 6a shows a dose-dependent increase in type II receptor protein at DHT concentrations between 0 and 10010 M. At DHT concentrations greater than 10010 M, the level of type II receptor protein decreased. Regarding the mechanism of action of TGF-b receptors, it has been postulated that both type I and type II TGF-b receptors are required for signaling [20]. Thus, we also conducted Western blot analysis for type I receptor. Figure 6b shows that the expected 55-kDa immunoreactive band was observed in PC 3 cells but not in LNCaP cells. Since we have demonstrated that TGFb1 signaling occurs in LNCaP cells, it is probable that type I receptor is present in these cells but at a level not detectable by Western blot analysis.
of TGF-b1 are realized only in the presence of 10010 M DHT in the culture medium. In addition, at this concentration of DHT, the presence of specific binding for TGF-b1 and an increase in TGF-b receptor type II in these cells have further supported the existence of TGF-b sensitivity in LNCaP cells. These observations, taken together, provide a valuable insight into the growth control mechanisms of LNCaP cells, in vitro. LNCaP cells express a functional, but mutated, androgen receptor that is sensitive not only to androgen but also to anti-androgens, estrogen, and progestins [8, 9]. Nevertheless, because the androgen receptor has a normal DNA binding domain, the cell line has been used widely as a model of androgen-responsive prostate cancer to study the cellular mechanism of androgen action [22–24]. The present observations have provided evidence to indicate that DHT can modulate the sensitivity of LNCaP cells to TGF-b1. Four concentrations of DHT were used in the present study: 0, 10012, 10010, and 1007 M. These concentrations were selected because they represent the zero DHT control, the lowproliferative dose, the high-proliferative dose, and the growth-arrest dose, respectively [19]. The results of the present study demonstrated that TGF-b1 treatment decreased cell count in a dose-dependent manner only at 10010 M DHT; the lowest level of cell count (Ç40% of control) was achieved at 20 and 200 ng/ml of TGFb1. In addition, in cultures with 2 ng/ml of TGF-b1 and 10010 M DHT, [3H]thymidine incorporation was Ç40% of control, cell viability was Ç60% of control, and cell morphology was altered. Finally, treating these cells with TGF-b1 and 10010 M DHT resulted in Ç3-fold increase in the activity of a TGF-b-responsive promoter present in p3TP-Lux. These series of independent but complementary observations demonstrated that LNCaP cells responded to TGF-b1 only in the presence of the high-proliferative dose of DHT, 10010 M. At other concentrations of DHT, the effects of TGF-b1 were minimal and statistically insignificant. Therefore, these re-
DISCUSSION
TABLE 2
Results of the present study have demonstrated that DHT modulates the sensitivity of LNCaP cells to TGFb1 in culture. The effects of TGF-b1 on these cells have been shown by four different but complementary criteria: (i) inhibition of cell growth, (ii) reduction in cell viability, (iii) alteration in cell morphology, and (iv) activation of a TGF-b-responsive promoter. These different approaches were employed to establish the sensitivity of LNCaP cells to TGF-b1 for two reasons. First, previous reports have indicated that these cells do not respond to TGF-b1 treatment [2, 4]. Second, some of the effects of TGF-b1 observed in this study were modest. The present results showed that the various effects
TGF-b1-Specific Binding in LNCaP Cells
/ m4678$3013
12-08-95 13:08:40
eca
Concentrations of DHT (M)
TGF-b1-specific binding (fmol/mg protein)a
0 (cFBS) 10010 1007
Undetectable 8.45 { 1.27 Undetectable
Note. Each value represents mean { SEM of three separate experiments. a Specific binding of 125I-TGF-b1 was determined by the difference in radioactivity between samples in the presence (nonspecific binding) and absence (total binding) of 100-fold molar excess of unlabeled TGF-b1. The specific binding in radioactive counts was then converted to fmol/mg of proteins.
AP: Exp Cell
MODULATION OF TGF-b1 SENSITIVITY IN LNCaP BY DHT
109
FIG. 6. TGF-b receptors type II and type I expression in LNCaP cells. Cells were cultured in varying concentrations of DHT for 2 days. Subsequently, 100 mg of each protein sample was separated by electrophoresis in 10% SDS–polyacrylamide gel, and immunoreactive bands were visualized by enhanced chemiluminescence. (a) Expression of type II receptor. Lane 1, cFBS; lane 2, 10012 M DHT; lane 3, 10011 M DHT; lane 4, 10010 M DHT; lane 5, 1009 M DHT; lane 6, 1008 M DHT; lane 7, 1007 M DHT. (b) Expression of type I receptor. Lane 1, PC 3 prostate cancer cell line; lane 2, cFBS; lane 3, 10012 M DHT; lane 4, 10011 M DHT; lane 5, 10010 M DHT; lane 6, 1009 M DHT; lane 7, 1008 M DHT; lane 8, 1007 M DHT.
sults indicate that DHT modulates TGF-b1 sensitivity in LNCaP cells. The influence of DHT on the TGF-b1 responsiveness in LNCaP cells can provide a partial explanation for the inconsistencies in the literature regarding TGF-b1 sensitivity in these cells. Wilding et al. and Carruba et al. reported that TGF-b1 did not affect the proliferation of LNCaP cells and that no TGF-b1-specific binding was detectable in these cells [2, 5]. On the other hand, Schuurmans et al. showed that TGF-b1 blunted the growth stimulating effects of epidermal growth factor (EGF) and transforming growth factor-a (TGF-a) [4]. These observations are consistent with the concept that the sensitivity of LNCaP cells to TGF-b1 can be manifested only under certain culture conditions. An inspection of these conditions suggests that TGF-b1 sensitivity is associated with a proliferative state of these cells. It is known that EGF and TGF-a promote LNCaP proliferation [26, 27]. In the present study, LNCaP cells became sensitive to TGF-b1 only in the presence of the high-proliferative dose of DHT. Therefore, the concept that LNCaP cells are sensitive to the action of TGF-b1 under a proliferative state warrants further investigation. Although the mechanism through which LNCaP cells acquire sensitivity to TGF-b1 at 10010 M DHT remains unclear, it is likely that the mechanism involves some components of the signaling system that are necessary for LNCaP cells to respond to TGF-b1. It has been demonstrated that TGF-b signaling is mediated through the interaction with membrane receptors. There are three different types of ubiquitously expressed TGF-b receptors: type I, II, and III [reviewed in 28]. Of these, type I and type II receptors are directly involved in TGF-b signal transduction [reviewed in 21, 28]. Current understanding states that type II receptor
/ m4678$3013
12-08-95 13:08:40
eca
is necessary for ligand binding and for subsequent formation of a heteromeric complex with type I receptor before TGF-b signaling can occur [15, 19]. Results of the present study demonstrated that DHT, at 10010 M, induced specific binding of TGF-b1, although at a low level, and increased the cellular contents of type II receptor. Since type II receptor has been shown to be responsible for the ligand binding [19, 20], the increase in type II receptor is consistent with the appearance of TGF-b1-specific binding at 10010 M DHT. Furthermore, since type II receptor is necessary for TGF-b signaling, an increase in the level of the receptor at 10010 M DHT will likely lead to an increased capacity to respond to TGF-b. An investigation to determine the biological significance of induction of type II receptor protein by 10010 M DHT in LNCaP cells will be pursued in our future studies. Efforts to detect type I receptor in LNCaP cells were not successful in the present study. Nevertheless, in this study, the effects of TGF-b1 signaling were demonstrated by four different but complementary criteria. These results suggest that the machinary for TGF-b signaling within LNCaP cells is intact when DHT is present at the dose of 10010 M. In summary, results of the present study have provided evidence to demonstrate that LNCaP cells are sensitive to TGF-b1 and express increased level of TGF-b receptor type II protein when the high-proliferative dose of DHT is added to the culture medium. These results indicate that DHT can modulate both TGF-b1 sensitivity and type II receptor protein level in LNCaP cells in culture. The present system, therefore, can be used as a model for the study of interaction between androgen and TGF-b1 in prostate cancer cells. We thank Dr. Jon Lomasney of Northwestern University Medical
AP: Exp Cell
110
KIM ET AL.
School (Chicago, IL) for his help in 125I-TGF-b1 binding assay and in reviewing the manuscript, and Dr. Joan Massague of Memorial Sloan-Kettering Cancer Center (New York, NY) for kindly providing p3TP-Lux. Finally, we thank Dr. Kohei Miyazono of Ludwig Institute for Cancer Research (Uppsala, Sweden) for providing TbRI and TbRII antibodies and experimental protocols.
REFERENCES 1. Wilding, G. (1991) Cancer Surv. 11, 147–163. 2. Wilding, G., Zugmeier, G., Knabbe, C., Flanders, K., and Gelmann, E. (1989) Mol. Cell. Endocrinol. 62, 79–87. 3. Franzen, P., Ichijo, H., and Miyazono, K. (1993) Exp. Cell Res. 207, 1–7. 4. Schuurmans, A. L. G., Bolt, J., Veldscholte, J., and Mulder, E. (1991) J. Steroid Biochem. Mol. Biol. 40, 193–197. 5. Carruba, G., Leake, R. E., Rinaldi, F., Chalmers, D., Comito, L., Sorci, C., Pavone-Macaluso, M., and Castagnetta, L. A. M. (1994) Steroids 59, 412–420. 6. Horoszewicz, J. S., Leong, S. S., Kawinski, E., Karr, J. P., Rosenthal, H., Chu, M. T., Mirand, E. A., and Murphy, G. P. (1983) Cancer Res. 43, 1809–1817. 7. Hiipakka, R. A., and Liao, S. (1995) in Androgen Receptors and Action. (Degroot, L. J., Ed.), pp. 2336–2351, Saunders, New York. 8. Wilding, G., Chen, M., and Gelmann, E. P. (1989) The Prostate 14, 103–115. 9. Kokontis, J., Ito, K., Hipakka, R. A., and Liao, S. (1991) Receptor 1, 271–279. 10. Taplin, M. E., Bubley, G. J., Shuster, T. D., Frantz, M. E., Spooner, A. E., Ogata, G. K., Keer, H. N., and Balk, S. P. (1995) New Engl. J. Med. 332, 1393–1398. 11. Sonnenschein, C., Olea, N., Pasanen, M. E., and Soto, A. M. (1989) Cancer Res. 49, 3474–3481.
12. Kyprianou, N., and Isaacs, J. T. (1989) Mol. Endocrinol. 3, 1515–1522. 13. Kyprianou, N., and Isaacs, J. T. (1988) Endocrinology 123, 2124–2131. 14. Knabbe, C., Klein, H., Zugmaier, G., and Voigt, K. D. (1993) J. Steroid. Biochem. Mol. Biol. 47, 137–142. 15. Fong, C. J., Sherwood, E. R., Braun, E. J., Berg, L. A., Lee, C., and Kozlowski, J. M. (1992) The Prostate 21, 121–131. 16. Wrana, J. L., Attisano, L., Carcamo, J., Zentella, A., Doody, J., Laiho, M., Wang, X., and Massague, J. (1992) Cell 71, 1003– 1014. 17. Franzen, P., ten Dijke, P., Ichijo, H., Yamashita, H., Schultz, P., Heldin, C.-H., and Miyazono, K. (1993) Cell 75, 681–692. 18. Steele, R. E. D., and Torrie, J. H. (1960) Principles and Procedures of Statistics, pp. 99–131. McGraw–Hill, New York. 19. Lee, C., Sutkowski, D. M., Sensibar, J. A., Zelner, D., Kim, I., Amsel, I., Shaw, N., Prins, G. S., and Kozlowski, J. M. (1995) Endocrinology 136, 796–803. 20. Wrana, J. L., Attisano, L., Wieser, R., Ventura, F., and Massague, J. (1994) Nature 370, 341–347. 21. Derynck, R. (1994) TIBS 19, 548–553. 22. Veldscholte, J., Berrevoets, C. A., and Mulder, E. (1994) J. Steroid Biochem. Mol. Biol. 49, 341–346. 23. Henttu, P., and Vihko, P. (1993) Cancer Res. 53, 1051–1058. 24. Krongrad, A., Wilson, C. M., Wilson, J. D., Allman, D. R., and McPhaul, M. J. (1991) Mol. Cell. Endocrinol. 76, 79–88. 25. Gleave, M. E., Hsieh, J-T., Wu, H-C., Eschenbach, A. C., and Chung, L. W. K. (1992) Cancer Res. 52, 1598–1605. 26. Macdonald, A., and Habib, F. K. (1992) Br. J. Cancer 65, 177– 182. 27. Limonta, P., Moretti, R. M., Dondi, D., Marelli, M. M., and Motta, M. (1994) J. Steroid Biochem. Mol. Biol. 49, 347–350. 28. Kingsley, D. M. (1994) Genes Dev. 8, 133–146.
Received July 25, 1995 Revised version received September 11, 1995
/ m4678$3013
12-08-95 13:08:40
eca
AP: Exp Cell