Abstract
There is evidence that estrogens can modulate the activity of prostate epithelial cells. To determine whether estradiol can have a direct influence on rat prostate, this study examined the effects of estradiol-17β (E2) administered alone or in combination with dihydrotestosterone (DHT) to castrated rats for 3 weeks on prostate binding protein (PBP) C1 mRNA expression and androgen receptor (AR) localization. PBP C1 mRNA levels were measured by semi-quantitative in situ hybridization using a 35S-labeled cDNA probe. In intact animals, strong hybridization signal could be observed in prostate sections after 12 hr of exposure to Kodak X-Omat films. In castrated rats, no PBP C1 mRNA could be detected even with longer exposure times, an effect that was prevented by administration of DHT. E2 administered alone induced a detectable hybridization signal, and the concomitant administration of E2 and DHT induced an increase in PBP C1 mRNA that significantly exceeded that obtained in animals that received only DHT. In prostate epithelial cells of intact animals, AR immunostaining was restricted to the nucleus. In castrated animals the alveoli were decreased in size and the epithelial cells were atrophied. AR staining was weak and was detected in both cytoplasm and nucleus. DHT administration completely obviated the effect of castration on epithelial cell histology and on AR immunostaining distribution and intensity. Interestingly, E2 administration alone induced moderate hypertrophy of epithelial cells compared to the histological appearance of cells in untreated castrated rats. Moreover, in E2-treated animals the nuclear staining was much stronger than that detected in untreated castrated rats, whereas the cytoplasmic staining was not modified by the treatment. In animals that received both DHT and E2, the staining was similar to that seen in DHT-treated rats. These results suggest that E2 can influence the activity of rat prostate epithelial cells by mechanisms that remain to be fully clarified.
Keywords
T
During the past few years accumulating evidence suggests that estrogens can play a physiological role in male reproduction (Sharpe 1998; Simpson et al. 2000). It is well documented that prostate tissue from several species contains estrogen receptors (ERs) (Tilley et al. 1980,1985; Swaneck et al. 1982; West et al. 1988). Recently a second ER, called ERβ, has been cloned from a rat prostate library (Kuiper et al. 1996), and the original one is now designated as ERα. ERβ expressed at high levels in rodent and primate prostate (Kuiper et al. 1996,1997; Couse et al. 1997; Pelletier 2000; Pelletier and El-Alfy 2000; Pelletier et al. 2000). Low levels of ERα mRNA have also been reported in rat prostate (Kuiper et al. 1997; Lou et al. 1998). In rat prostate, ERβ was localized by in situ hybridization and immunocytochemistry in epithelial cells in alveoli (Prins and Birch 1997; Pelletier 2000; Pelletier et al. 2000), whereas no ERα expression could be detected.
There is some evidence that estrogen itself can exert an influence on prostate epithelial cell division and differentiation. Estrogen administration to castrated or hypophysectomized dogs induced glandular hyperplasia (Leav et al. 1978; Tunn et al. 1979). In the Noble rat, estrogen synergizes with androgens to induce glandular hyperplasia and dysplasia (Leav et al. 1989; Lou et al. 1998). On the other hand, ERβ knockout mice display local prostate hyperplasia with aging (Krege et al. 1998), suggesting that ERβ might exert a negative regulation on prostate growth.
To study the involvement of estrogens in prostate epithelial cell functions, we evaluated the effects of E2 administered alone or in combination with DHT to castrated adult male rats on the expression of PBP C1 mRNA, as evaluated by in situ hybridization and the immunohistochemical localization of AR.
Materials and Methods
Animals
Thirty adult male Sprague-Dawley rats (Charles River; Wilmington, MA) weighing 225–250 g at the beginning of the experiments were housed under constant temperature (21 ± 1C) and lighting (light on from 0600 to 2000) regimens. They had free access to standard rat chow and tapwater. All the protocols were approved by the Laval University's Animal Welfare Committee.
Treatments and Tissue Preparation
Four groups of animals (six per group) were castrated via the scrotal route. One group of sham-operated rats was used as intact control. The castrated animals were treated twice daily with the vehicle, E2 (0.4 μg/kg bw), DHT (400 μg/kg bw), or DHT in combination with E2 for 3 weeks. The steroids were administered SC in 0.5 ml 1% (w/v) gelatin. The intact animals received only the vehicle (1% gelatin). The steroids were purchased from Steraloids (Wilton, NH). On the morning after the last day of the treatment, animals were perfused transcardially with 200 ml 4% (w/v) paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Ventral prostates were excised and postfixed in the same fixative for 48 hr at 4C. For immunocytochemistry, the tissues were embedded in paraffin. For in situ hybridization, the tissues were placed in 15% sucrose in 0.1 M phosphate buffer before being quickly frozen in isopentane chilled in liquid nitrogen. This experiment was duplicated and essentially the same results were obtained.
PBP C1 Probe Preparation
The plasmid containing cDNA, corresponding to the C1 peptide of PBP cloned in the Pst-1 site of pS64, was provided by Dr. M.G. Parker (London, UK). The Pst-1 restriction fragment was radiolabeled with [35S]-CTP (NEN Life Science Products; Boston, MA), as previously described (Pelletier et al. 1988). As control, sections were treated with pancreatic RNase A (20 μg/ml; Boehringer Mannheim; Mannheim, Germany) for 1 hr at 37C before hybridization.
In Situ Hybridization
Frozen sections (10 μm) were serially cut at −20C and mounted on gelatin- and poly-
Immunocytochemistry
AR immunostaining was performed on paraffin sections (two sections/glass slide) as previously described (Pelletier 2000; Pelletier et al. 2000). An affinity-purified rabbit polyclonal antibody reacting with rat AR (N-20; Santa Cruz Biotechnology, Santa Cruz, CA) was used at a concentration of 1 μg/ml. This antibody has been successfully used to localize AR in several tissues, including the prostate (Pelletier and El-Alfy 2000; Pelletier et al. 2000). Control reaction was obtained by substituting preabsorbed antibody with an excess of the peptide used as an antigen (20 μg/ml). After immunostaining the sections were counterstained with hematoxylin. To avoid any variations related to the staining procedure, two sections from each of the 30 prostates (six prostates/group) were stained in the same run. In each run, immunoabsorption controls (one for each experimental group) were also included. We thus proceeded to four runs. The results were evaluated by two independent investigators who were unaware of the treatments.
Results
PBP C1 mRNA Expression
As shown in Figure 1A, in vehicle-treated, sham-operated animals, hybridization with the 35S-labeled PBP C1 cDNA probe induced a strong signal in prostate sections after 12 hr of exposure. Pretreatment of the sections with RNase before hybridization completely prevented any labeling (not shown). In castrated animals no detectable signal could be obtained even after longer exposure times (up to 7 days; not shown). As shown in Figures 1B and 2, administration of DHT to castrated rats completely obviated the effects of castration, the mRNA levels being 18% above the levels measured in vehicle-treated, sham-operated animals (p<0.001). Administration of E2 induced a hybridization signal that could be detected after 12 hr of exposure, the measured PBP C1 mRNA levels corresponding to approximately 3% of the values obtained in intact animals (Figures 1C and 2). Figures 1D and 2 show the effect of the administration of both DHT and E2 on the mRNA levels, which exceeded by 16% (p<0.001) those observed in DHT-treated animals.
X-ray autoradiographs illustrating the expression of PBP C1 mRNA in rat prostate sections (exposure time 12 hr). (
AR Immunostaining
In prostate sections from vehicle-treated, sham-operated animals immunostained for AR localization, strong labeling was detected in nuclei of all secretory epithelial cells. The cytoplasm of the epithelial cells did not exhibit any labeling (Figure 3A). Other immunostained nuclei in the stroma surrounding the alveoli were also consistently observed. In castrated animals, as shown in Figure 3B, the alveoli were markedly reduced in size and appeared dispersed throughout the stroma, which was not modified. The epithelial cells had a cuboidal appearance, with markedly reduced cytoplasm and an increased nuclear-to-cytoplasmic ratio. In contrast to the observations in sham-operated rats, immunostaining was present in both cytoplasm and nuclei, with a marked reduction in nuclear labeling. The staining of stromal cells did not appear to have been modified by castration. As shown in Figure 3C, treatment with DHT completely reversed the effect of castration on the epithelial cells. The histology and AR localization were very similar to what has been observed in vehicle treated, sham-operated animals, with nuclei being strongly immunoreactive. In castrated animals treated with E2, the size of the alveoli was not modified, but the epithelial cells appeared hypertrophied compared to those observed in vehicle-treated castrated rats (compare Figures 3D and Figure 3B). The nucleus:cytoplasmic ratio appeared to be decreased by E2 treatment. Moreover, in the E2-treated animals the intensity of the reaction was very different from that observed in vehicle-treated castrated rats. The nuclei were strongly labeled and light cytoplasmic staining was present (Figure 3D). In the stroma, AR labeling was also stronger. In animals that received both DHT and E2, the results were very similar to those observed in animals treated only with DHT (Figure 3E).
Effects of castration and DHT and E2 administration to castrated rats on PBP C1 mRNA levels as evaluated by semi-quantitative in situ hybridization. INT, intact animals;
Discussion
The present results clearly demonstrate that 3-week administration of E2 can stimulate the mRNA expression of a prostate androgen-dependent protein, PBP C1, in adult castrated rats. The effect of E2 was weak but significant because no hybridization signal could be observed in untreated castrated rats. Moreover, E2 was also effective in stimulating PBP C1 mRNA levels when the activity of epithelial cells was maintained by DHT administration. These data strongly suggest that E2 can directly stimulate the activity of prostate epithelial cells in the absence or presence of circulating androgens.
The hypertrophy of epithelial cells after administration of E2 suggests that estrogens can directly stimulate the activity of secretory epithelial cells. These results are in agreement with several reports indicating that administration of estrogen to castrated or hypophysectomized animals could exert a stimulatory influence on prostate epithelial cells. In castrated or hypophysectomized dogs, estrogen induced hypertrophy of epithelial cells (Leav et al. 1978; Tunn et al. 1979; Merk et al. 1980,1986; Kwan et al. 1982). Similarly, in castrated rats, estrogens increase epithelial cell height in the ventral prostate (Salander and Tisell 1976; Thompson et al. 1979; Timms and Chandler 1985). Moreover, it has been shown that, in the Noble rat prostate, androgen-supported estrogen could be responsible for epithelial proliferation and dysplasia (Leav et al. 1989; Lou et al. 1998). Our data also clearly show that, in DHT-treated animals, E2 could further increase PBP C1 mRNA expression. It thus appears reasonable to hypothesize that estrogens can potentiate the effects of androgens on prostate epithelial cell activity. Recently, Yeh et al. (1998) have reported that E2 can activate androgen target genes in the prostate via an interaction with the AR complex.
AR immunolocalization showed that, in intact animals, the staining was restricted to nuclei in epithelial cells. After castration there was a marked reduction in nuclear labeling and, contrary to what was observed in sham-operated rats, cytoplasmic labeling was consistently found. These results are in agreement with previous findings indicating a decrease in nuclear androgen-binding sites in ventral prostate in 2-week castrated rats (Prins 1989). The presence of immunoreactive material in the cytoplasm of epithelial cells may therefore reflect retention of AR in the cytoplasmic compartment. It is well known that binding of a steroid (including androgens) to its receptor molecule results in activation of the receptor in the cytoplasmic compartment (Clark et al. 1992). The activated receptor-steroid complex, which has a high affinity for various nuclear binding sites, then migrates to the nucleus. In castrated rats that were treated with DHT, no cytoplasmic staining could be detected, whereas strong nuclear staining similar to that observed in sham-operated animals was present. The histology of the epithelial cells was also very similar to that observed in intact animals. This appears to be a morphological confirmation that the activation of AR by circulating androgens leads to a translocation of the receptor to the nucleus.
Immunocytochemical localization of AR in rat prostate sections. (
Of great interest was the finding that E2 administration induced a marked increase in nuclear AR labeling and a decrease in cytoplasmic AR staining. This suggests that estrogens, even in the absence of circulating androgens, can activate AR, leading to transfer of the receptors from the cytoplasm to the nucleus. Such an activation of AR might result from direct interaction of E2 with AR. Other mechanisms, such as an increase in AR biosynthesis, might be involved in the increase in nuclear staining. We have recently observed that administration of estrogens to castrated rats induced a marked increase in prostate AR mRNA levels (unpublished). Using subcellular fractionation, Blondeau et al. (1982) have reported that, in castrated rat prostate, AR could be found in the cytosolic fraction and that the injection of DHT or E2 4 hr before sacrifice could increase nuclear AR concentrations. On the other hand, it cannot be totally excluded that cytoplasmic staining might result from artifactual redistribution or unbound AR, which might occur during the immunocytochemical procedure.
We have previously reported that administration of E2 during one week to castrated rats did not induce significant changes in PBP C1 mRNA levels (Pelletier et al. 1988). It therefore appears that longer exposure to E2 is required to positively modulate PBP C1 mRNA in castrated animals. Because androgens can rapidly (within 12 hr) stimulate PBP C1 mRNA in castrated rats, it is unlikely that the effect of E2 might be related to an activation of AR although, as mentioned above, interaction of E2 with AR might explain the changes in AR staining distribution. We have recently observed that chronic (3-week) administration of E2 to castrated rats could decrease ERβ expression in the prostate (unpublished data). Because, in ERβ knockout mouse, hyperplasia of prostate epithelium occurs with aging, it has been proposed that ERβ might negatively regulate epithelial cell activity in the prostate (Weihua et al. 2001). Such an effect on ERβ might explain the stimulatory influence of E2 that occurs after 3 weeks of administration.
It clearly appears, on the basis of the present experiments, that E2 administration can stimulate the expression of an androgen-dependent protein and interact with AR in rat prostate epithelial cells. The mechanism(s) of action of E2, which might involve interaction with ERβ and/or AR, remains to be fully clarified. Other studies involving use of anti-estrogens and anti-androgens would help to clarify the exact role of estrogens in prostate regulation.