The Prostate 28: 182- I 94 ( 1996) Control of LNCaP Proliferation and Differentiation: Actions and Interactions of Androgens, I a,25-Dihydroxycholecalciferol, All-Truns Retinoic Acid, 9-Cis Retinoic Acid, and Phenylacetate Murielle Esquenet, Johannes V. Swinnen, Walter Heyns, and Guido Verhoeven laboratory for Experimental Medicine and Endocrinology, Department of Developmental Biology, Onderwijs en Navorsing, Gasthuisberg, Catholic University of Leuven, leuven, Belgium ABSTRACT: There is increasing evidence that growth and differentiation of prostatic carcinoma cells may be modulated not only by androgens and growth factors but also by vitamin D, retinoids, and phenylacetate (PA). The latter agonists may have a role in the prevention and therapy of prostate cancer but their exact therapeutic potential remains unclear. Since both retinoids and vitamin D act via nuclear receptors, the same way androgens do, we studied the interactions of these compounds with androgen-induced proliferation and differentiation using LNCaP cells as a model of androgen-responsive tumor cells. PA was included because of its suspected different mode of action. [,H]-thyrnidine incorporation was used as a measure of proliferative activity, secretion of prostate-specific antigen (PSA) as a measure of differentiated function. The present data show that la,25-dihydroxycholecalciferol(VD,), all-trans retinoic acid (atRA), 9-cis retinoic acid (~cRA),and PA stimulated LNCaP cell-differentiated function in the presence or absence of androgens. The effects on cell growth were more complicated. In the absence of androgens growth stimulatory effects were observed for the retinoids and under some conditions for VD,. These effects were limited, however, and tended to be more pronounced at low cell densities. In the presence of androgens nearly exclusively growth inhibitory effects were observed. On a molar basis VD, was the most effective antiproliferative agonist (ED,, = M). It completely neutralized the stimulatory effects of androgens. Growth inhibition was not due to a decrease in the concentration of androgen receptor: whereas atRA, 9cRA, and PA did not alter androgen receptor levels, VD, provoked a twofold increase. Neither in the presence nor in the absence of androgens did we observe any cooperativity in the growth stimulatory or inhibitory effects of VD,, atRA, or 9cRA. To test whether treatment with any of the studied agonists resulted in a phenotypic reversion and sustained growth arrest, LNCaP cells were pretreated with VD,, atRA, 9cRA, or PA for 6-12 days and reseeded at equal densities as untreated cells. In all cases tested [,H]-thyrnidine incorporation was restored within 6 days suggesting that none of these compounds caused irreversible growth inhibition. 0 1996 Wiley-Liss, Inc. KEY WORDS: prostate cancer, vitamin D, retinoids, differentiation therapy, prostatespecific antigen INTRODUCTION Prostate cancer is becoming an increasingly important problem in our aging society [l].The mechanisms the Origin and the Of the disease remain poorly understood. It has been well 0 1996 Wiley-Liss, Inc. Received for publication August 9, 1994; accepted December 6, 1994. Address reprint requests to G. Verhoeven, Laboratory for Experimental Medicine and Endocrinology, Department of Developmental Biology, Onderwijs en Navorsing, Gasthuisberg, Catholic University of Leuven, B-3000 Leuven, Belgium. Growth and Differentiation of LNCaP Cells established, however, that during the initial stages of tumor development androgens play at least a permissive role. For many years androgen ablation or blockade has become the therapy of choice for the treatment of metastatic carcinoma . Unfortunately, tumors tend to become androgen insensitive and at this point endocrine therapy fails. In search for new therapies it has become increasingly clear that apart from androgens several other steroidal and nonsteroidal factors such as vitamin D, retinoids, and phenylacetate (PA) influence proliferation and differentiation of prostatic carcinoma cells studied in vitro [3-lo]. Epidemiologic investigations have presented evidence that vitamin D as well as vitamin A may protect against clinical prostate cancer [ll-131 and it has been suggested that both compounds might offer new avenues for the treatment of androgen-dependent as well as androgen-independent prostate cancer. A number of tumor cell lines have proven useful in the study of the response of prostatic carcinoma cells to growth factors [14-191. LNCaP cells, however, are the only ones that remain androgen responsive in vitro [20-221. This cell line was established by Horoszewicz et al.  from a lymph node metastasis of a patient who developed refractory disease following androgen-deprivation therapy. LNCaP cells respond to androgens with increased proliferation and some of their differentiated functions such as production of acid phosphatase (PAP) and prostate-specific antigen (PSA) are also androgen regulated [24-271. LNCaP cells have high concentrations of an androgen receptor, but a point mutation in the ligand binding domain results in an aberrant response to progestins, estrogens, and antiandrogens [28-311. Recently it has been demonstrated that LNCaP cells have receptors for other members of the nuclear receptor family such as vitamins D and A. Proliferative as well as antiproliferative effects of the corresponding ligands have been demonstrated and it has been shown that both vitamins D and A are able to stimulate differentiated functions of LNCaP cells [3,4,71. In the present paper we used the LNCaP model to study whether there are any interactions between androgens and vitamin D or vitamin A, ligands for two other members of the nuclear receptor family. Moreover, we explored whether 9-cis retinoic acid (~cRA), which modulates the effects of vitamins D and A in a number of other systems , influences the effects of these agonists in LNCaP cells. Finally, since it has been claimed that differentiation of prostate cancer cells by agonists such as PA might irreversibly change the malignant behavior of invasive prostate tumor cells [lo], we investigated the reversibility of the ef- 183 fects of vitamin A, vitamin D, and PA on the proliferation of LNCaP cells. MATERIALS AND METHODS Cell Culture The LNCaP human prostatic cancer cell line was obtained from the American Type Culture Collection (ATCC, Rockville, MD). The cells were routinely maintained in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS), 3 mM L-glutamine, 100 pg/ml streptomycin, and 100 U/ml penicillin. All media and supplements were obtained from GIBCO BRL (Paisley, Scotland). The cultures were trypsinized twice weekly and were maintained in a humidified atmosphere of 5% C 0 2 in air. Where indicated FCS was pretreated with dextran-coated charcoal (DCCFCS; 0.25% [w/v] dextran; 2.5% [wh] charcoal) . [3H]-Mibolerone, unlabeled mibolerone, and R1881 (methyltrienolone) were purchased from Dupont-New England Nuclear (Boston, MA). la,25dihydroxy-cholecalciferol (VD,) was obtained from Solvay Duphar (Weesp, Holland). g-cis-retinoic acid (9cRA) was kindly provided by Dr. Uskokovic (Hoffmann-La Roche, Nutley, NJ). All-trans retinoic acid (atRA) and phenylacetate (PA) were obtained from Sigma Chemical Co. (St. Louis, MO). Cell Proliferation Assays Cell proliferation was evaluated either by measurements of [,H]-thyrnidine incorporation, by cell counting, or by DNA measurements. For the [3H]-thymidineincorporation,experiments, cells were seeded in 96-well multidishes (Nunc, Roskilde, Denmark) at a density of 5 x lo3 cells per well in RPMI-1640 with 10% FCS. The next day the medium was changed to RPMI-1640 supplemented with 10% DCC-FCS. Fourty-eight hours later the medium was changed to RPMI-1640 with 5% DCC-FCS supplemented with the compounds of interest. Steroids, vitamin D, and vitamin A derivatives were added in ethanol. The final ethanol concentration did not exceed 0.1%. PA was dissolved in distilled water, brought to pH 7.0 by the addition of NaOH, and stored in aliquots at -20°C. Depending on the experiment, incubation was continued for 3 up to 6 days. For incubation periods of 6 days, the medium was changed after 3 days. After this incubation, [3H]-thymidine (specific activity: 2 Ci/mmol; ICN, Irvine, CA) was added (1 pWwe11). Six hours later, the media were removed; the cultures were washed twice with 100 pl ice-cold phosphate buffered saline (PBS), trypsinized, and harvested on Whatman filter paper (grade 934AH, Whatman, Maidstone, England) using 184 Esauenet et al. a multiple-cell harvester (Nunc). The filters were collected and counted in 3 ml counting scintillant (InstaGel, Packard Instrument Company, Meriden, CT). In some experiments, proliferation was evaluated by cell counting. To this end 6 x lo4 cells were seeded in 12-well multidishes. The culture conditions were identical to those described for the [3H]-thymidine incorporation experiments. After the incubation periods, cells were detached by trypsinization and counted using a Coulter Counter (model ZB.0067). DNA content of the cultures was measured using the method of Labarca and Paigen . Study of the Reversibility of the Effects of VD,, atRA, 9cRA, or PA on LNCaP Cell Proliferation To study the reversibility of the growth-suppressing effects of VD,, RA, and PA, LNCaP cells were seeded in 80 cm2 flasks at a density of 1.25 X lo6 cells in RPMI-1640 with 10% FCS. After 24 hr the medium was changed to RPMI-1640 with 5% FCS supplemented or not with the test compounds. The medium was changed every 3 days. This pretreatment was continued for 6 or 12 days. Thereafter, cells were trypsinized, counted in a Biirker Chamber, and reseeded in 96-well plates at a density of 5x103 cells per well. These cells were cultured for an additional period of 3-6 days in RPMI-1640 supplemented with 5% FCS, 5% DCC-FCS, or 5% DCC-FCS + lo-'' M R1881. After this incubation time, [,H]-thyrnidine incorporation was studied as described above. Study of the Effects of VD,, RAs, and PA on Androgen Receptor Binding Androgen receptor binding was studied in intact LNCaP cells pretreated with various test compounds using a single point displaceable binding assay as described previously . Briefly, cells were plated in 6-well plates in medium with 10% FCS and allowed to adhere overnight. The medium was changed to medium containing 10% DCC-FCS and 48 hr later pretreatment with the different test compounds was started in RPMI-1640 supplemented with 5% DCCFCS. After an incubation time of 6 days, cultures were carefully washed with serum-free medium and incubation was continued for an additional 24 hr in serumfree medium in the presence of the respective test compounds. Finally, after three more washes with serum-free medium, androgen receptor binding was measured. To this end [3H]-mibolerone was added at a concentration of 5 x lop9 M, in the absence (tracer) or presence (overload) of unlabeled mibolerone (lop6 M) and the cells were incubated for 24 hr at 37°C. At the end of the incubation the medium was discarded and the cells were washed three times with ice-cold PBS containing 0.2% bovine serum albumin. Cells were recovered in 1 ml of 1%(v/v) Triton X-100 in 0.1 N NaOH. After transfer to scintillation vials, 3 ml of counting scintillant was added and the radioactivity was counted. Displaceable binding was calculated and expressed per milligram cell protein. Protein concentrations were measured using the Coomassie protein assay reagent (Pierce, Rockford, IL). Statistics Determination of PSA Secretion by LNCaP Cells The secretion of PSA was used as a parameter of differentiated cell function in LNCaP cultures. To this end LNCaP cells were seeded at a density of 6 x lo4 cells per well in 12-well multidishes in RPMI-1640 medium supplemented with 10% FCS. The following day the medium was changed to RPMI-1640 containing 10% DCC-FCS. Fourty-eight hours later medium with 5% DCC-FCS and supplemented with the indicated concentrations of the studied test compound was added to each well. After 3 days, the media were replaced by media containing the same test compounds but supplemented with lo-'' M R1881 as indicated. The cells were cultured for another 3 days, the conditioned media were collected, centrifuged to remove residual cells (800 rpm; 10 min), and stored at -20°C. PSA levels were determined using the Tandem-R immunoradiometric assay (Hybritech, San Diego, CA) according to the manufacturer's instructions. To account for differences in cell number, results were expressed per microgram DNA. Statistical analysis was performed using analysis of variance followed by Tukey's t test. A value of Pc0.05 was considered statistically significant. RESULTS Effects of Androgens on the Proliferation of LNCaP Cells In a first series of experiments optimal conditions for the study of the effects of androgens on the proliferation of LNCaP cells were delineated. Cells were cultured in DCC-FCS to prevent interference of serum-derived androgens. Increasing concentrations of the synthetic androgen R1881 were added since natural androgens are rapidly metabolized in LNCaP cells . After various periods of time [3H]-thymidine incorporation as well as cell number were measured as a parameter of proliferative activity. In agreement with previous reports [20-221 [,HIthymidine incorporation showed a biphasic response to increasing concentrations of R1881 (Fig. 1A). Max- Growth and Differentiation of LNCaP Cells 185 h m 50 ? IL-1 2: 0 C -12-11 . -10 - 9 -8 -7 -6 B T A T L C -12-11 -10 - 9 -8 -7 -6 R 1 8 8 1 ( l o g M) Fig. 1. Effect of androgen concentration on the proliferation of LNCaP cells. LNCaP cells were seeded in 96-well plates at a density of 5 x I O3 cells per well for [3H]-thymidineincorporation (A) or at a comparable density of 6 X lo4 cells per well in a 12-well plate for cell counts (B). [3H]-thymidine incorporation (A) was measured after 24 hr (O),48 hr (V), 72 hr (W), and 96 hr (A)of incubation in the presence of increasing concentrations of the synthetic androgen R I 88 I. Cell counts (B) were determined after 72 hr (0) and 6 days (0)of incubation with R1881. Results shown represent the mean k SD of incubations performed in triplicate. imal stimulatory effects were observed at a concentration of lo-'' M. At higher concentrations [,H]-thymidine incorporation dropped sharply. No changes in optimal R1881 concentrations were observed as a function of time of incubation, but the effect was more pronounced after longer incubation periods (Fig. 1A). It may be noted that a dose-dependent increase in [,H]-thymidine incorporation was already apparent after an incubation period of 24 hr. Maximal incorporation (four to fivefold stimulation as compared to the control) was observed after 72 hr of incubation. A similar dose-response relationship was observed when cell numbers were used as a parameter of the proliferative activity of LNCaP cells (Fig. 1B). However, more prolonged exposure to androgens (6 days) was required to document this response. For these reasons [,H]-thymidine incorporation after 3 days of incubation was selected to evaluate androgen effects in all subsequent experiments. and the factors determining the nature of these effects are only partially understood. To analyze the actions of the mentioned compounds in detail LNCaP cells were seeded at three different densities (lo3, 5 x lo3, and lo4 cells per well) and exposed for six days to increasing concentrations of VD,, atRA, 9cRA, and PA. All experiments were performed in DCC-FCS. Moreover, eventual interactions of the test compounds with androgens were evaluated by adding a maximal stimulatory concentration of R1881 (lo-'" M) to half of the culture wells during the last 3 days of incubation. [,H]-thymidine incorporation was used as a measure of proliferative activity. The results obtained with cells seeded at a concentration of 5 x lo3 cells per well are summarized in Figure 2. Under these conditions VD, inhibited [,H]-thymidine incorporation. This effect became evident from a concentration of lop9 M on. In the absence of androgens [,H]-thymidine incorporation was reduced to approximately 50% of the control value. In the presence of androgens a reduction to less than 30% of the control value was observed. At concentrations of VD, equal or higher than lop8 M the stimulatory effect of androgens was completely neutralized. Roughly the same effects were observed when LNCaP cells were seeded at higher densities (lo4 cells/well). At lower cell densities (lo3cells/well)stimulatory effects of VD, were noted at a concentration Effects of VD,, RAs, and PA on C3H+Thymidine Incorporation in LNCaP Cells Grown in the Presence or Absence of Androgens As summarized in the introduction, evidence has been presented that vitamin D, vitamin A, and PA may modulate growth and differentiation of prostate tumor cells and cell lines [3-lo]. The effects of the first two compounds may be stimulatory or inhibitory 186 Esquenet et al. 40 //' 30- "D3 I ". 1 i- atRA i' k \ 20 - lo I -12 L C -10 -8 -6 - 30 - 20 - 10 n- 4 C \. - 40 -12 -10 -8 -6 ++ J 75 60 45 \* 5 C -12 -10 -8 -6 C -5 -4 -3 -2 CONCENTRATION ( l o g M ) Fig. 2. Dose-dependent effects of VD,, atRA, 9cRA. and PA on [ 'HI-thymidine incorporation by LNCaP cells in the presence o r absence of androgens. Cells were seeded in 96-well plates at a density of 5 x lo3 cells per well. The cultures were treated as described in Materials and Methods. [ 'HI-thymidine incorporation was measured after 6 days of incubation with increasing concen- trations of the indicated compounds in the absence of androgens (0)or in the presence of 10- l o M R I88 I (0)during the last 3 days of culture. Results shown represent the mean SD of incubations performed in triplicate. Values which are statistically different (W0.05)from the respective control values are indicated by an asterisk. of lo-" M. These effects were limited (25% in the presence of androgens and 100% in the absence of androgens) and were replaced by inhibitory effects at higher concentrations of VD, (not shown). In the absence of androgens atRA stimulated [,HIthymidine incorporation at concentrations ranging from lo-'' up to M. In the presence of androgens limited stimulatory effects were observed at M and lo-'' M whereas inhibitory effects became obvious at concentrations exceeding lop8 M. Again stimulatory effects tended to be more pronounced when cells were seeded at lower densities. In the presence of androgens, however, inhibitory effects were consistently observed at concentrations equal or higher than lop8 M (not shown). 9cRA slightly stimulated [,H]-thymidine incorporation in the absence of androgens at concentrations between lo-'' and M. Against this effect be- came more pronounced at lower cell densities (lo3 cells/well; not shown). At higher concentrations of 9cRA inhibitory effects prevailed particularly in the presence of androgens. PA provoked a dose-dependent inhibition of [,HIthymidine incorporation at all cell densities studied both in the presence and absence of androgens. In the presence of androgens the effect became apparent at a concentration of lop5 M. In the absence of androgens a concentration of M was required. To explore whether VD,, atRA, and 9cRA might display any cooperativity in their effects on [,H]-thymidine incorporation, LNCaP cells were exposed to lop9 M VD,, M atRA, and M 9cRA or to combinations of these agonists. The effects were studied both in the presence and absence of androgens. As illustrated in Table I no cooperativity was observed neither for the stimulatory effects observed * 187 Growth and Differentiation of LNCaP Cells TABLE 1. Effect of Combined Compounds on Basal and Androgen-Induced Proliferation of LNCaP Cellst Control lop9 M VD, lo-' M atRA 10-7 M ~ C R A l o p 9 M VD, + lop8 M atRA 10-9 M VD, + 10-7 M ~ C R A lo-' M atRA + M 9cRA No androgen lop'' M R1881 100 f 1 87 ? 6 168 ? 10' 171 ? 9* 118 ? 5* 153 t 4* 159 ? 10* 100 f 8 44 ? 1* 37 t 4* 30 ? 1' 30 ? 2* 25 ? 5* 28 -C 3* ~~ ~ 'LNCaP cells were seeded in 96-well plates at a density of 5 x lo3 cells per well and cultured as described in Materials and Methods. [3H]-thymidine incorporation was measured after 6 days of incubation with the mentioned compounds. Where indicated lo-''' M R1881 was added during the last 3 days. Results shown represent the mean 2 SD of incubations performed in triplicate and are expressed in % compared to the respective controls. Values which are statistically different ( R 0 . 0 5 ) from the respective control values. in the absence of androgens nor for the inhibitory effects observed in the presence of R1881. h m I 0 4 Influence of VD,, atRA, 9cRA, and PA on the Responsiveness of LNCaP Cells to Androgens In order to explore whether VD,, RAs, or PA might affect the characteristic biphasic dose-response curve of [,H]-thymidine incorporation to androgens, LNCaP cells were exposed to increasing concentrations of R1881 in the presence or absence of lop8 M VD,, M atRA, lop6M 9cRA, and 5.10p3M PA. As shown in Figure 3 basal and androgen-stimulated [,H]-thyrnidine incorporation was suppressed but maximal androgen responses were still observed at lo-" M R1881. The decrease in androgen responsiveness was apparently not related to a change in androgen receptor concentration. In fact, as summarized in Table 11, VD, provoked a 2.5-fold increase rather than a decrease in androgen receptor levels whereas the other compounds tested displayed only marginal effects. Effects of VD,, atRA, 9cRA, and PA on PSA Secretion PSA was monitored as a parameter of the differentiated secretory functions of LNCaP cells. As expected androgens markedly stimulated PSA secretion (Table 111). In the absence of androgens all concentrations of VD,, atRA, 9cRA, and PA tested displayed stimulatory effects. A twofold increase was observed with VD, and PA whereas a fourfold increase was noted with the RAs. Even in the presence of R1881 a 50% increase in PSA secretion was observed with all compounds tested. 25 K x E za 20 15 - 10 - 5 - I z I C -12 -11 I -10 % - -9 -8 -7 R1881 ( l o g M ) ., Fig. 3. Dose-response curves of R I 881 in the presence of inhibitory concentrations of VD,, atRA, ScRA, and PA. 5 x IO3 LNCaP cells were seeded in 96-well plates and incubated as described in Materials and Methods. After a 3-day incubation period with each of the test compounds (0, control; V, lo-* M VD,; lo-' M atRA; 0, M 9cRA; A, 5.10-3 M PA), the medium was changed t o medium containing both the test compound and increasing concentrations of R I88 I. [ ,H]-thymidine incorporation was measured 3 days later. Values represent the mean ? SD of incubations performed in triplicate. Reversibility of the Growth Inhibitory Effects of VD,, atRA, 9cRA, and PA In order to explore whether the investigated compounds might induce terminal differentiation and ir- 188 Esquenet et al. TABLE II. Effect of VD, atRA, 9cRA, and PA on Androgen Receptor Level in LNCaP Cellst Receptor binding (% of control) Control VD, lo-' M lo-* M M 100 ? 7 244 243 264 * 38 ? 2 6 43 atRA lo-* M 137 5 9 120 1 * 143 * 22 M 9cRA lop7 M lo-' M 106 ? 7 PA * 10-3 M 5.1OP3 M 110 6 84 ? 12 +LNCaPcells were seeded in 6-well plates and incubated as described in Materials and Methods. After an incubation period of 6 days in the presence of the mentioned compounds androgen receptor binding was measured by a single point monolayer binding assay. Results shown are expressed as percent compared to control and are expressed as mean SD. Receptor binding in the control was 617 2 53 fmol/mg protein. * reversible growth arrest of prostatic tumor cells, LNCaP cells were exposed for 6 days to VD,, atRA, 9cRA, and PA. All pretreatments were performed in 5% FCS. Thereafter control cultures and pretreated cells were trypsinized and identical numbers of viable cells were seeded in multidishes. The growth potential of these cells was examined by exposing them to DCC-FCS in the absence or presence of R1881 or to FCS. [3H]-thymidine incorporation or cell DNA was measured at the end of a 3-day or 6-day incubation period. Pretreatment with VD, reduced [,H]-thymidine incorporation at the end of the 3-day incubation period (Fig. 4). A limited stimulatory effect of R1881 and FCS could still be observed but this effect was not reflected in an increase in the DNA content of the cell cultures (Table IV). At the end of the 6-day incubation period [,H]-thymidine incorporation was further decreased in cultures maintained in DCC-FCS. A marked response to R1881 and FCS was observed, however, and in pretreated cells grown in FCS [3H]thymidine incorporation equalled that observed in control cultures. The DNA contents of the cultures also reflected a marked response to R1881 and FCS although the levels of DNA remained clearly below those observed in nonpretreated cells. Similar results were observed for cells pretreated with PA. In this case [3H]-thymidine incorporation after 6 days was TABLE 111. Effect of VD,, atRA, 9cRA. and PA on PSA Secretion by LNCaP Cellst No androgen Control VD3 M lo-* M M 0.216 * 0.044 lo-'' M R1881 2.103 ? 5 * 0.016' 0.074' 0.019' 3.759 ? 0.509' 3.556 5 0.011' 2.860 -+ 0.149' 0.876 0.882 0.874 2 0.040' 3.578 0.051' 3.115 t 0.297' 3.052 5 0.413' * 0.028' * 0.073' 0.927 5 0.123' 1.076 ? 0.14T 0.881 ? 0.080' PA 10-3 M 5.10-3 M 0.291 0.478 0.403 0.324 atRA lop9 M lo-* M lop7 M 9cRA lop8 M 10-7 M M 2 0.440 0.778 * 0.070' ? 0.049' * 3.314 3.281 3.448 3.099 3.456 5 ? * 0.267' 0.443' 0.725' 0.432' * 0.150' +LNCaP cells were incubated in the presence of varying concentrations of the indicated compounds for 6 days as outlined in Materials and Methods. Conditioned media of the last 3 days were collected and PSA levels were determined. DNA content of the cultures was determined simultaneously. Values are expressed in ng/pg DNA. Results shown are expressed as the mean k SD of incubations performed in triplicate. "Values which are statistically different (P<0.05) from the respective control. not significantly different in control cultures and pretreated cultures and the amount of DNA in the latter cultures was close to that observed in the control cultures. It may be noted that pretreatment with atRA M) in medium supplemented with FCS resulted in an inhibition of [,H]-thymidine incorporation (Fig. 5, 3-day incubation period) rather than in a stimulation as observed in medium supplemented with DCC-FCS (Fig. 2). After 6 days of incubation the inhibition disappeared. At lop7 M the inhibitory effects of atRA became even more obvious and at that concentration a 50% reduction in cell DNA content was observed at the end of the 6-day incubation period (Table IV). A comparable degree of growth inhibition was observed with l o p 6 M 9cRA. To investigate whether a more prolonged pretreatment might produce irreversible growth inhibition, LNCaP cells were exposed for 12 days to M VD,. This resulted in a more marked inhibition of [3H]thymidine incorporation at the end of the 3-day incubation period (Fig. 6). Moreover, at this time point cells did not respond to androgens or FCS. After 6 days of incubation, however, only cells maintained in DCC-FCS still displayed diminished [3H]-thymidine incorporation, whereas in LNCaP cells treated with Growth and Differentiation of LNCaP Cells 189 Pretreatment with VD, ( 6 days) 60 t T a P r e t r e a t m e n t with PA ( 6 days) G 0 I inc. 6 inc. 3 d 2 E La inc. 6 d inc. 3 d d T 45 T D 15 L DCC-FCS DCC-FCS + Rl88l FCS DCC-FCS DCC-FCS + FCS Rl88l Fig. 4. Effect of retrea ment with VD, or PA on LNCaP cell proliferation in the presence and absence of androgens. LNCaP cells were cultured for a period of 6 days in the absence (open bars) or in the presence of of lop8 M VD, or 5.10-' M PA (hatched bars) as described in Materials and Methods. Thereafter the cells were reseeded in 96-well plates at a density of 5 x lo3 cells per well. icubation (inc.) was continued for 3 or 6 days in RPMI-I 640 supplemented with 5% DCC-FCS, 5% DCC-FCS 10- M R I88 I, or 5% complete FCS. At the end of this period (3H]-thymidine incorporation was measured. Results shown represent the mean ? SD (n = 3) for a representative experiment. R1881 or FCS [,H]-thyrnidine incorporation was completely restored. [3H]-thymidine incorporation at a concentration of lo-'' M. A 3-day incubation period with lo-'' M R1881 was selected for further experiments. The remarkable shape of this dose-response curve remains poorly understood. The present data clearly show that the shape of this curve was not affected by the duration of androgen exposure (24-96 hr). Moreover, it is worth noting that a classical hyperbolic response curve was observed for androgen-stimulated parameters of differentiated function of LNCaP cells such as the induction of PSA and diazepam-binding inhibitor (DBI) mRNA . In the absence of androgens the studied agonists (VD,, atRA, 9cRA, PA) displayed variable effects on cell proliferation. PA consistently inhibited [,H]-thymidine incorporation with an ED,, value of 5 mM comparable to that reported by Samid et al. [lo]. The effects of vitamin D and retinoids were more variable and depended not only on the concentration of these agonists but also on cell density which may explain some of the controversy in the literature [3,4,7]. VD, DISCUSSION One of the main aims of this study was to explore the interactions between androgens, the main growth-promoting hormones for prostate tumor cells, and VD,, atRA, 9cRA, and PA, a series of agonists with potential antiproliferative activity. LNCaP cells were selected for these experiments since they represent the only widely employed system which allows the study of androgen-regulated growth of prostate cancer cells in vitro [20-231. The proliferative response of LNCaP cells to androgens markedly depends on culture conditions [36-381 and accordingly, optimal conditions to quantitate this response were first delineated. In agreement with previous reports [20-223 cells maintained in DCC-FCS displayed a steep biphasic dose-response curve after 3 days of incubation with the synthetic androgen R1881 with a maximal stimulation of + ' 190 Esquenet et al. TABLE IV. Influence of 6-Day Pretreatment With Various Agonists on Basal and Androgen-Induced Cell Growth* 3-Day incubation Pretreatment Control VD3 lo-' M 10-7 M atRA lo-' M M 9cRA M lo-' M PA M 5.10-" M DCC-FCS DCC-FCS 6-Day incubation + R1881 FCS * 0.22 3.44 t 0.37 4.50 t 0.30 1.62 ? 0.08 2.11 f 0.17 1.76 0.32 1.88 f 0.09 1.88 f 0.22 1.56 ? 0.04 1.91 f 0.21 1.93 0.13 * 2.19 t 0.19 1.99 0.09 1.64 f 0.21 n.d. 2.66 f 0.17 1.88 0.09 3.75 2.08 * 5.51 DCC-FCS 5.41 1.02 DCC-FCS + R1881 * 2.53 2.95 * 0.66 FCS 8.68 14.7 f 2.79 1.44 f 0.33 2.04 f 0.47 4.46 t 0.37 5.11 t 0.44 6.69 f 0.17 2.77 0.65 1.96 f 0.15 1.44 f 0.35 1.89 t 0.14 5.48 4.23 2.37 f 0.25 n.d. 2.66 t 0.27 n.d. 2.76 1.52 * 0.74 * 0.29 0.13 3.92 t 0.18 1.99 f 0.95 2.76 f 0.27 3.00 t 0.46 * f * 0.25 * f * 0.97 f 0.48 8.24 4.44 f f 0.22 0.50 8.66 6.76 f 0.59 * 0.31 10.88 ? 0.57 7.62 t 0.32 * 1.59 * 0.62 10.07 * 0.49 11.28 5.78 10.58 ? 0.06 *LNCaP cells were pretreated with the mentioned test compound for a period of 6 days as outlined in Materials and Methods. Cells were reseeded in 12-well multidishes at a density of 6 x lo4 cells per well and were incubated in RPMI-1640 medium supplemented with 5% DCC-FCS, 5% DCC-FCS + lo-'' M R1881, or full serum. Three or 6 days later DNA content of the cultures was determined. Values are expressed in kg DNA/well. Results shown represent the mean ? SD of incubations performed in triplicate. n.d., not determined. displayed mainly growth inhibitory effects. At low cell densities, however, limited stimulatory effects (up to twofold) were observed. These effects were maximal at lo-" M. Stimulatory effects were observed with both atRA and 9cRA over a broad range of concentrations (lo-'' up to lop6 M and lo-" M up to M, respectively). Again these effects tended to be more pronounced at lower cell densities and to decrease at higher cell densities or at higher concentrations of the retinoids. In general it should be stressed that in comparison with androgens the observed proliferative effects of vitamin D and retinoids were variable and limited. In the presence of androgens growth-stimulating effects became even less prominent (atRA) or disappeared (9cRA). Interestingly, concentrations of atRA with obvious stimulatory effects in the absence of androgens (e.g., lop7 M) caused 50% growth inhibition in the presence of these hormones, whereas limited stimulatory effects became evident at very low concentrations of atRA which were inactive in the absence of androgens (Fig. 2). A comparable complex response pattern to RA was observed by Fong et al.  using dihydrotestosterone (DHT) as an androgen. The antiproliferative effects of vitamin D and retinoids were clearly dose dependent. On a molar basis VD, was by far the most efficient growth inhibitor investigated (ED,, approximately lop9 M). This compound completely neutralized the proliferative effect of androgens. Some more variability was observed for atRA and 9cRA (ED,, between lo-' M and lop7 M and between lop7 M and M, respectively). The ED,, for PA was approximately lop3 M. It may be concluded that the antiproliferative activities of the studied compounds are maintained and become even more pronounced in the presence of androgens. The mechanism of this growth inhibition remains elusive. The present data do not support any direct interaction with the androgen-response apparatus of the cells. Androgen receptor levels were maintained in the presence of atRA, 9cRA, and PA and were even markedly increased in the presence of VD,. Moreover, in the presence of suboptimal concentrations of the inhibitors the remarkable shape of the dose-response curve to androgens remained unchanged: maximal [3H]-thymidine incorporation was still observed at lo-'' M R1881. Finally, the studied compounds opposed the growth-promoting effects of androgens on LNCaP cells but enhanced their effects on differentiated secretory functions. In fact, in contrast with the complexity of their effects on the proliferation of LNCaP cells, VD,, atRA, 9cRA, and PA stimulated PSA secretion under all the conditions investigated. Stimulatory effects on PSA secretion have previously been reported for VD, and atRA [4,7]. Our data show that these effects were modest when compared with those of androgens. They remained significant, however, even in the presence of lo-'' M R1881. Moreover, for atRA and 9cRA, they could be observed both under conditions Growth and Differentiation of LNCaP Cells 191 P r e t r e a t m e n t with atRA ( 6 d a y s ) I inc. 3 d 60 h 0 4 451 k ': E a U a 30 e: P r e t r e a t m e n t with 9 c R A ( 6 d a y s ) U Pnc. 0 5 w 5 6o E E 45 inc. 6 d t DCC-FCS DCC-FCS + FCS DCC-FCS R1881 DCC-FCS + FCS Rl88l Fig. 5. Effect of pretreatment with atRA or 9cRA on LNCaP cell proliferation in the presence or absence of androgens. LNCaP cells were cultured for a period of 6 days in the absence (open bars) or in the presence of lo-* M atRA or M 9cRA (hatched bars). Thereafter the cells were reseeded in 96-well plates at a density of 5 X lo3 cells per well. Incubation (inc.) was continued for 3 or 6 days in RPMI- I640 supplemented with 5% DCC-FCS, 5% DCCFCS lo-'' M R1881, or 5% complete FCS. A t the end of this period [ 'HI-thymidine incorporation was measured. Results shown represent the mean ? SD (n = 3) for a representative experiment. which stimulate proliferation and under conditions which inhibit proliferation indicating that proliferation and differentiation are not mutually exclusive. Little is known on the mechanisms by which androgens, vitamin D, and retinoids influence the proliferation of LNCaP cells. All these factors are ligands for members of a superfamily of nuclear receptors that modulate gene expression by binding to specific DNA sequences located in the regulatory regions of target genes [40,41]. Receptors for androgens, atRA, 9cRA, and VD, have been shown to be present in prostate tumor cells [3,4,7,9]. It remains unclear, however, whether the mentioned ligands act directly on genes related to cell cycle control or whether they act indirectly by modulating the secretion of autocrine acting growth-promoting or growth-inhibiting factors. The slowness of the effects rather suggests an indirect mechanism of action. The receptors for VD, (VDR), atRA (RAR), and 9cRA (RXR) share a family of similar hormone response elements configured as direct or everted repeats of the half-site consensus PuGGTCA. Both the VDR and the RAR form heterodimers with the RXR which can also function as an independent receptor for 9cRA [42-471. In a number of systems synergistic effects have been observed between 9cRA and the ligand occupying the other half of the heterodimeric receptor complex [48,49]. Such cooperative effects could not be demonstrated in the LNCaP system. In fact the proliferative andor antiproliferative effects of VD,, atRA, and 9cRA were more or less additive. Since the effects of 9cRA strongly resembled those of atRA and since 9cRA binds with a higher affinity to the RAR than to the RXR , it is quite possible that the observed effects were mediated by the RAR. In a number of systems retinoids, vitamin D, and other proliferation inhibitors may induce terminal differentiation. In HL-60 cells, e.g., RA induces a granulocyte differentiation accompanied by growth arrest . Some data suggest that in prostate cancer cells too differentiation therapy might result in phenotypic reversion and sustained growth arrest. PC-3 cells treated with PA for 1 week in culture, e.g., apparently loose the ability to form rapidly growing tu- + 192 Esquenet et at. 0 - e inc. 3 d ” 60 - 45 - i I - I U 0 4 m p 30 - U 0 5 E: 15- - P 72 77 2. r I 0 0 DCC-FCS DCC-FCS + FCS R1881 DCC-FCS DCC-FCS + FCS R1881 Fig. 6. Effect of prolonged pretreatment with VD, on LNCaP cell proliferation in the presence and absence of androgens. LNCaP cells were cultured for a period of I 2 days in the absence (open bars) or in the presence of IO-* M VD, (hatched bars). Pretreated cells were reseeded in 96-well plates a t a density of 5 x lo3 cells per well and incubation (inc.) was continued for 3 or 6 days in medium supplemented with 5% DCC-FCS. 5% DCC-FCS + 10- l o M R I88 I, or 5% complete FCS. A t the end of this period [ ’HI-thymidine incorporation was measured. Results shown repreSD (n= 3) for a representative experiment. sent the mean mors in recipient athymic mice [lo]. Similarly, in a clonal growth assay the effects of VD, and retinoids on prostatic epithelial cells seem to be irreversible [5,6]. The present data indicate that, at least in LNCaP cells, growth suppression by VD,, retinoids, or PA is reversible. When cells pretreated for 6 or even 12 days with the mentioned growth inhibitors are reseeded at the same density (5 x lo3 cells/well in 96well multidishes) as untreated LNCaP cells, [,H]-thymidine incorporation and cell proliferation resume within approximately 6 days. Further experiments will be required to exclude that this is a specific property of LNCaP cells. It seems more likely, however, that the reversible or irreversible character of the effects of the studied agonists depends on the selected test conditions. In summary, we demonstrated that each of the substances tested could severely alter the growth responses of the LNCaP cell line both in the presence and in the absence of androgens, and induce differentiated functions as measured by an increase in PSA secretion. However, withdrawal of these substances restored the initial growth response within 6 days. Further studies using an in vivo model may be required to clarify the potential role of these compounds as therapeutic factors in limiting prostate cancer development. Belgian National fund for Scientific Research (N.F.W.O.). The excellent technical assistence of J. Rosseels is kindly acknowledged. ACKNOWLEDGMENTS This research was supported by grant “Geconcerteerde Onderzoeksactie van de Vlaamse Gemeenschap”, grant 3.0015.88 from the Belgian National Fund for Scientific Research (N.F.W.O.), and by a grant from the Vereniging voor Kankerbestrijding. Dr. J. Swinnen is a senior research assistant of the * REFERENCES 1. Carter HB, Coffey DS: The prostate: An increasing medical problem. Prostate 16:39-48, 1990. 2. Garnick MB: Prostate cancer: Screening, diagnosis, and management. Ann Intern Med 118:804-818, 1993. 3. Miller GJ, Stapleton GE, Ferrara JA, Lucia MS, Pfister S, Hedlund TE, Upadhya P: The human prostatic carcinoma cell line LNCaP expresses biologically active, specific receptors for la,25-dihydroxyvitamin D,. Cancer Res 52:515-520, 1992. 4. Skowronski RJ, Peehl DM, Feldman D: Vitamin D and prostate cancer: 1,25 dihydroxyvitamin D, receptors and actions in human prostate cancer cell lines. Endocrinology 132:1952-1960, 1993. 5. Peehl DM, Skowronski RJ, Leung GK, Wong ST, Stamey TA, Feldmand D: Antiproliferative effects of 1,25-dihydroxyvitamin D, on primary cultures of human prostatic cells. Cancer Res 54:805-810, 1994. 6. Peehl DM, Wong ST, Stamey TA: Vitamin A regulates proliferation and differentiation of human prostatic epithelial cells. Prostate 23:69-78, 1993. 7. Fong C-J, Sutkowski DM, Braun EJ, Bauer KD, Sherwood ER, Lee C, Kozlowski JM: Effect of retinoic acid on the proliferation and secretory activity of androgenresponsive prostatic carcinoma cells. J Urol 149:11901194, 1993. 8. Stearns ME, Wang M, Fudge K: Liarozole and 13-cisretinoic acid anti-prostatic tumor activity. Cancer Res 5313073-3077, 1993. 9. Dahiya R, Boyle B, Park H-D, Kurhanewicz J, Macdonald JM, Narayan P: 13-cis-retinoic acid-mediated growth inhibition of DU-145 human prostate cancer cells. Biochem Mol Biol Int 32:l-12, 1994. 10. Samid D, Shack S, Myers CE: Selective growth arrest and phenotypic reversion of prostate cancer cells in Growth and Differentiation of LNCaP Cells vitro by nontoxic pharmacological concentrations of phenylacetate. J Clin Invest 91:2288-2295, 1993. 11. Hanchette CL, Schwartz GG: Geographic patterns of prostate cancer mortality: Evidence for a protective effect of ultraviolet radiation. Cancer 70:2861-2869, 1992. 12. Reichman ME, Hayes RB, Ziegler RG, Schatzkin A, Taylor PR, Kahle LL, Fraumeni JF: Serum vitamin A and subsequent development of prostate cancer in the first national health and nutrition examination survey epidemiologic follow-up study. Cancer Res 50:23112315, 1990. 13. Willett WC, Hunter DJ: Vitamin A and cancers of the breast, large bowel, and prostate: Epidemiologic evidence. Nutr Rev 52:S53-S59, 1994. 14. Wilding G, Valverius E, Knabbe C, Gelmann EP: Role of transforming growth factor-a in human prostate cancer cell growth. Prostate 15:l-12, 1989. 15. Schuurmans ALG, Bolt J, Veldscholte J, Mulder E: Regulation of growth of LNCaP human prostate tumor cells by growth factors and steroid hormones. J Steroid Biochem Mol Biol 40:193-197, 1991. 16. Knabbe C, Kellner U, Schmahl M, Voight K-D: Growth factors in human prostate cancer cells: Implications for an improved treatment of prostate cancer. J Steroid Biochem Mol Biol 40:185-192, 1991. 17. Hofer DR, Shenvood ER, Bromberg WD, Mendelsohn J, Lee C: Autonomous growth of androgen-independent human prostatic carcinoma cells: Role of transforming growth factor a. Cancer Res 51:2780-2785, 1991. 18. Nakamoto T, Chang C, Li A, Chodak GW: Basic fibroblast growth factor in human prostate cancer cells. Cancer Res 52:571-577, 1992. 19. MacDonald A, Habib FK: Divergent responses to epidermal growth factor in hormone sensitive and insensitive human prostate cancer cell lines. Br J Cancer 65: 177-182, 1992. 20. Berns EMJJ, de Boer W, Mulder E: Androgen-dependent growth regulation of and release of specific protein(s) by the androgen receptor containing human prostate tumor cell line LNCaP. Prostate 9:247-259, 1986. 21. Schuurmans ALG, Bolt J, Mulder E: Androgens stimulate both growth rate and epidermal growth factor receptor activity of the human prostate tumor cell LNCap. Prostate 12:55-63, 1988. 22. Kokontis J, Takakura K, Hay N, Liao S: Increased androgen receptor activity and altered c-myc expression in prostate cancer cells after long-term androgen deprivation. Cancer Res 54:1566-1573, 1994. 23. Horoszewicz JS, Leong SS, Kawinski E, Karr JP, Rosenthal H, Chu TM, Mirand EA, Murphy GP: LNCaP model of human prostatic carcinoma. Cancer Res 43:1809-181 8, 1983. 24. Schulz P, Bauer HW, Fittler F: Steroid hormone regulation of prostatic acid phosphatase expression in cultured human prostatic carcinoma cells. Biol Chem Hoppe Seyler 366:1033-1039, 1985. 25. Henttu P, Vihko P: Steroids inversely affect the biosynthesis and secretion of human prostatic acid phosphatase and prostate-specific antigen in the LNCaP cell line. J Steroid Biochem Mol Biol41:349-360, 1992. 26. Montgomery BT, Young CY-F, Bilhartz DL, Andrews PE, Prescott JL, Thompson NF, Tindall DJ: Hormonal regulation of prostate-specific antigen (PSA) glycopro- 193 tein in the human prostatic adenocarcinoma cell line, LNCaP. Prostate 21:63-73, 1992. 27. Lin M-F, Garcia-Arenas R, Chao Y-C, Lai MMC, Pate1 PC, Xia X-Z: Regulation of prostatic acid phosphatase expression and secretion by androgen in LNCaP human prostate carcinoma cells. Arch Biochem Biophys 300:384-390, 1993. 28. Wilding G, Chen M, Gelmann EP: Aberrant response in vitro of hormone-responsive prostate cancer cells to antiandrogens. Prostate 14:103-115, 1989. 29. Olea N, Sakabe K, Soto AM, Sonnenschein C: The proliferative effect of ”anti-androgens” on the androgensensitive human prostate tumor cell line LNCaP. Endocrinology 126:1457-1463, 1990. 30. Veldscholte J, Berrevoets CA, Ris-Stalpers C, Kuiper GGJM, Jenster G, Trapman J, Brinkmann AO, Mulder E: The androgen receptor in LNCaP cells contains a mutation in the ligand binding domain which affects steroid binding characteristics and response to antiandrogens. J Steroid Biochem Mol Biol 41:665-669, 1992. 31. Veldscholte J, Berrevoets CA, Mulder E: Studies on the human prostatic cancer cell line LNCaP. J Steroid Biochem Mol Biol 49:341-346, 1994. 32. Giguere V: Retinoic acid receptors and cellular retinoid binding proteins: Complex interplay in retinoid signaling. Endocr Rev 15:61-79, 1994. 33. Green B, Leake RE: “Steroid Hormones: A Practical Approach.” Oxford: IRL Press, Oxford University Press, 1987. 34. Labarca C, Paigen K: A simple, rapid, and sensitive DNA assay procedure. Anal Biochem 102:344-352, 1980. 35. Verhoeven G, Cailleau J: Testicular peritubular cells secrete a protein under androgen control that inhibits induction of aromatase activity in sertoli cells. Endocrinology 123:2100-2110, 1988. 36. Schulz P, Wolf D, Arbusow V, Bojar H, Klobeck H-G, Fittler F: The synthetic androgen mibolerone induces transcient suppression of the transformed phenotype in an androgen responsive human prostatic carcinoma cell line. Andrologia 22(Suppl 1P):56-65, 1990. 37. Langeler EG, van Uffelen CJC, Blankenstein MA, van Steenbrugge GJ, Mulder E: Effect of culture conditions on androgen sensitivity of the human prostatic cancer cell line LNCaP. Prostate 23:213-223, 1993. 38. Kirschenbaum A, Ren M, Levine AC: Enhanced androgen sensitivity in serum-free medium of a subline of the LNCaP human prostate cancer cell line. Steroids 58: 439-444, 1993. 39. Swinnen JV, Esquenet M, Heyns W, Rombauts W, Verhoeven G: Androgen regulation of the messenger RNA encoding diazepam-binding inhibitorlacyl-CoA-binding protein in the human prostatic adenocarcinoma cell line LNCaP. Mol Cell Endocrinol Mol. Cell. Endocrinol. 104:153-162,1994. 40. Evans Rh4: The steroid and thyroid hormone receptor superfamily. Science 240:889-895, 1988. 41. O’Malley B: The steroid receptor superfamily: More excitement predicted for the future. Mol Endocrinol 4:363-369, 1990. 42. Yu VC, Deisert C, Andersen B, Holloway JM, Devary OV, Naar AM, Kim SY, Boutin J-M, Glass CK, Rosenfelk MG: RXRP: A coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D recep- 194 Esquenet et al. tors to their cognate response elements. Cell 67:12511266, 1991. 43. Zhang X-K, Hoffmann B, Tran PB-V, Graupner G, Pfahl M: Retinoid X receptor is an auxiliary protein for thyroid hormone and retinoic acid receptors. Nature 355:441-446, 1992. 44. Leid M, Kastner P, Lyons R, Nakshatri H, Saunders M, Zacharewski T, Chen J-Y, Staub A, Garnier J-M, Mader S, Chambon P: Purification, cloning, and RXR identity of the HeLa cell factor with which RAR or TR heterodimerizes to bind target sequences efficiently. Cell 68:377-395, 1992. 45. Bugge TH, Pohl J, Lonnoy 0, Stunnenberg HG: RXRa, a promiscuous partner of retinoic acid and thyroid hormone receptors. EMBO J 11:1409-1418, 1992. 46. Marks MS, Hallenbeck PL, Nagata T, Segars JH, Appella E, Nikodem VM, Ozato K: H-2RIIBP (RXRp) heterodimerization provides a mechanism for combinatorial diversity in the regulation of retinoic acid and 47. 48. 49. 50. 51. thyroid hormone responsive genes. EMBO J 11:14191435, 1992. Kliewer SA, Umesono K, Mangelsdorf DJ, Evans Rh4: Retinoid X receptor ineracts with nuclear receptors in retinoic acid, thyroid hormone and vitamin D, signaling. Nature 355:446-449, 1992. Rosen ED, O’Donnell AL, Koenig RJ: Ligand-dependent synergy of thyroid hormone and retinoid X receptors. J Biol Chem 267:22010-22013, 1992. Carlberg C, Bendik I, Wyss A, Meier E, Sturzenbecker LJ, Grippo JF, Hunziker W: Two nuclear signalling pathways for vitamin D. Nature 361:657-660, 1993. Mangelsdorf DJ: Vitamin A receptors. Nutr Rev 52:S32S44, 1994. Eppinger TM, Buck J, Hammerling U: Growth control or terminal differentiation: Endogenous production and differential activities of vitamin A metabolites in HL-60 cells. J Exp Med 178:1995-2005, 1993.