Sex Hormone-Binding Globulin Mediates Prostate Androgen Receptor Action via a Novel Signaling Pathway
Victor D. H. Ding, David E. Moller, William P. Feeney, Varsha Didolkar, Atif M. Nakhla, Linda Rhodes, William Rosner and Roy G. Smith
Estradiol (E2) and 5-androstan-3,17ß-diol (3-diol) have been implicated in prostate hyperplasia in man and dogs, but neither of these steroids bind to androgen receptors (ARs). Recently, we reported that E2 and 3-diol stimulated generation of intracellular cAMP via binding to a complex of sex hormone-binding globulin (SHBG) and its receptor (RSHBG) on prostate cells. We speculated that this pathway, involving steroids normally found in the prostate, was involved in the indirect activation of ARs. Using the dog as a model to test this hypothesis in normal prostate, we investigated whether E2, 3-diol, and SHBG stimulated the production of the androgen-responsive protein, arginine esterase (AE), the canine equivalent of human prostate-specific antigen. In cultured dog prostate tissue preincubated with SHBG, E2 and 3-diol stimulated AE activity. These effects were blocked by hydroxyflutamide, an AR antagonist, and by 2-methoxyestradiol, a competitive inhibitor of E2 and 3-diol binding to SHBG. In the absence of exogenous steroids and SHBG, AE also was significantly increased by treatment with forskolin or 8-Bromoadenosine-cAMP. These observations support the hypothesis that in normal prostate, E2 and 3-diol can amplify or substitute for androgens, with regard to activation of the AR via the RSHBG by a signal transduction pathway involving cAMP. Because both E2 and 3-diol are involved in the pathogenesis of benign prostatic hyperplasia in dogs and implicated in benign prostatic hyperplasia in man, antagonism of the prostatic SHBG pathway may offer a novel and attractive therapeutic target.
SEX hormone-binding globulin (SHBG) is a plasma glycoprotein that binds sex steroids with high affinity, thereby regulating their plasma concentrations (1). Also, by binding to a specific receptor (RSHBG) on prostate cell membranes, SHBG participates directly in cellular signaling pathways (2, 3, 4). Two steroids, 17ß-estradiol (E2) and 5-androstan-3,17ß-diol (3-diol), bind to the SHBG-RSHBG complex and stimulate cAMP production (3, 4, 5). Because cAMP transduces downstream signals implicated in modulation of cell growth (6, 7, 8, 9) and regulation of specific gene transcription and expression (10, 11, 12, 13, 14, 15, 16), activation of this pathway by E2 and 3-diol may prove to play a critical role in prostate function.
The androgen receptor (AR) is a tissue-specific transcription factor that is directly activated by binding to testosterone and 5-dihydrotestosterone (DHT). However, its indirect activation by FSH, by polypeptide growth factors, and through a cAMP pathway has been reported (17, 18, 19, 20, 21). Androgens cause an increase in the production of specific proteins in the prostate. For example, in humans, synthesis of prostate-specific antigen (PSA) is stimulated by DHT (22). In dogs, DHT increases prostate arginine esterase (AE) synthesis (23, 24, 25, 26, 27, 2.
The steroids, DHT, E2 and 3-diol, implicated in prostate growth, all bind to SHBG with high affinity; but only E2 and 3-diol activate SHBG-RSHBG in prostate tissue, causing rapid accumulation of intracellular cAMP (3, 4, 29). Traditionally, 3-diol has been thought to be a biologically inert metabolite of DHT; however, this metabolite was recently shown to exhibit hormonal activity (3, 30). Further, for more than a decade, it has been known that administration of 3-diol induces benign prostatic hyperplasia (BPH) in dogs (31, 32, 33, 34, 35). E2 also has been implicated in contributing to the pathogenesis of BPH. In castrated dogs, E2 prevents prostate regression; moreover, BPH can be induced experimentally by estrogens in intact dogs and monkeys (32, 33, 34, 35, 36). In humans, accumulation of E2 in the prostate is associated with advancing age (37). E2 also has been implicated in contributing to the growth and progression of human prostate cancer (3. In this regard, E2 binding to SHBG has been shown to stimulate cAMP accumulation in a human prostate cancer (LNCaP) cell line (5).
Prostate growth is known to be mediated by DHT activation of AR and consequent increase in the production of PSA (22). To address the mechanism by which E2 or 3-diol stimulation, via the SHBG-RSHBG pathway, might contribute to prostate growth in the dog, we asked whether the increase in cAMP was linked functionally to an androgen-responsive pathway in primary cultures of dog prostate. We selected the canine equivalent of PSA, AE as a marker of activation of the AR. As anticipated, AE was stimulated by DHT, but most importantly, in the presence (but not in the absence) of SHBG, both E2 and 3-diol mimicked the effect of DHT. While this work on normal prostate was in progress, similar observations were made using human BPH tissue, suggesting that this pathway has physiological relevance in the growth of both normal and hyperplastic tissue (39).
Chemicals and reagents
Tissue culture medium RPMI-1640, FBS, and sodium pyruvate were purchased from GIBCO-BRL (Gaithersburg, MD). All steroid compounds were obtained from Steraloids Inc. (Wilton, NH). Highly purified canine SHBG was prepared as previously described (3). All reagents and apparatus for electrophoresis and Western blot analysis were purchased from Novex (San Diego, CA). The enhanced chemiluminescence system for Western blot detection was obtained from Amersham Life Science (Arlington Heights, IL). 8-Br-cAMP, benzoyl arginine ethyl ester, forskolin, and isobutyl-methylxanthine were obtained from Sigma Chemical Company (St. Louis, MO). Hydroxyflutamide was obtained from the Merck Chemical Data collection. The anti-AE antiserum was a gift of Dr. J. Y. Dubé, Laboratory of Hormonal Bioregulation, Laval University Hospital Research Center, Sainte-Foy, Québec, Canada. AE protein standard was a gift of Dr. Alan Partin, Johns Hopkins University, Baltimore, MD. All other reagents were of analytical grade.
Prostatic tissue explants
Unless otherwise specified, prostate tissue was obtained from pure-bred male beagle dogs, 2–3 yr of age (3). Dogs were euthanized 7 days after surgical castration. All procedures for the humane handling, care, and treatment of research animals were done according to humane animal use procedures approved by the Merck Institutional Animal Care and Use Committee. The prostatic tissue was removed and brought to the laboratory under sterile conditions. It was divided into approximately 5-mm3 cubes and placed in 100-mm culture dishes (Corning Glass Works, Corning, NY) in RPMI-1640 medium with 5% FBS containing 1 mM sodium pyruvate, 100 U/ml penicillin, 100 mg/ml streptomycin sulfate, and 0.25 mg/ml amphotericin B for 2 days at 37 C in an atmosphere of 95% air and 5% CO2. Tissue was then minced into 1–2 mm3 portions and transferred to 24-well plates in serum-free medium (0.5 ml/well) for approximately 18 h before beginning each experiment.
Tissue processing and AE activity determination
Prostatic minces were incubated in the presence or absence of steroid hormones for 2 days. To avoid having any SHBG in the culture medium, serum-free medium was used, as previously reported (3, 4). Fresh medium and hormones were replaced at 24 h. To saturate the SHBG receptor, prostatic minces were incubated with 50 nM purified dog SHBG (3) for 3 h. Subsequently, appropriate concentrations of steroids were added after washing to remove excess SHBG. Other treatments with hydroxyflutamide (100 nM), 2-methoxyestradiol (2MeOE2), were added 15 min before addition of steroids. All the steroids and the antagonists used in this study were initially dissolved in 100% ethanol. Tissue minces were harvested and homogenized using a polytron (Tekmar Co., Cincinnati, OH) for 30 sec in 0.5 ml cold PBS (10 mM, pH 7.2) containing 0.2% Triton X-100. The homogenates were centrifuged at 25,000 x g for 45 min at 4 C to remove particulate matter, and the supernatants were stored at -80 C. The protein concentration of the supernatant was determined using the Bio-Rad Protein microassay procedure. AE activity was determined as described previously (40, 41, 42, 43), with modifications. The substrate concentration was 1 mM, and the reaction was carried out at room temperature (24 C) in 1 ml 0.01 M Tris/HCl buffer at pH 8.0. Total protein amounts (5–20 µg) were used to determine the enzyme activity by following the increase in optical density at 253 nm upon the hydrolysis of benzoyl arginine ethyl ester (42). AE activity present in tissue homogenates was found to be stable after sample storage at -80 C and a single freeze/thaw cycle did not affect the enzyme activity.
Western blotting analysis
Samples were resolved by electrophoresis through 4–20% SDS-polyacrylamide gradient gels, followed by electroblotting onto 100% methanol-treated PVDF membrane (0.45 µm, Immobilon P from Millipore, Bedford, MA). The membranes were blocked with 5% nonfat milk in PBS for 1 h. Immunostaining was performed by incubating with rabbit anti-AE polyclonal antibodies at 1:2,000 dilution in washing buffer (0.25% gelatin, 5 mM EDTA, 0.15 M NaCl, 0.05% Tween 20, 50 mM Tris/HCl, pH 7.4) for 1 h. The membranes were washed for 30 min, then incubated with horseradish peroxidase-conjugated donkey antirabbit IgG (Amersham) at 1:3,000 dilution for 40 min. The Western blots were developed using enhanced chemiluminescence procedures similar to those described by Amersham. All steps were carried out at room temperature.
Measurement of cAMP level
Levels of cAMP were measured using commercial enzyme-linked immunosorbent assay kits (Oxford Biomedical Research, Inc., Oxford, MI), as described previously (3). All samples contained isobutyl-methylxanthine (100 µM).
Statistical analysis
The significance of differences between treatment and control groups was assessed using Student’s t test. Values are reported as the mean ± SEM (SE of the mean).
Stimulation of AE by DHT
As expected, AE activity increased as a function of the dose of DHT (Fig. 1A). The basal level of AE activity in castrated (7 days) dog prostate was approximately 3 µmol/min·mg protein. AE activity was not significantly stimulated by 1 nM DHT; however, at 10 nM and 100 nM DHT, AE activity was significantly increased. Thus, DHT stimulates AE activity in a dose-dependent manner; and although a concentration of 100 nM may not represent a maximal response, this concentration was used as a reference in subsequent experiments.
Figure 1. Effect of steroid hormones on AE activity. (A), DHT; (B), E2 or 3-diol ± SHBG; (), tissue incubated with dog SHBG for 3 h to saturate SHBG receptors before treatment with steroids; (), no SHBG treatment; **, P < 0.05, compared with the control; *, P < 0.05, compared with the same treatment in the absence of SHBG. Each data point represents the mean ± SEM of six determinations derived from two independent experiments.
Stimulation of AE through the SHBG receptor complex
After preincubation with SHBG, to allow formation of the SHBG-RSHBG complex, treatment with E2 or 3-diol led to an increase in AE activity similar to that observed with DHT (Fig. 1B). By contrast, in the absence of SHBG, E2 and 3-diol did not stimulate AE activity. Thus, the stimulation of AE, caused by E2 and 3-diol, required the SHBG-RSHBG complex.
The effects on AE protein were measured by Western blot using a specific anti-AE antibody. AE migrated as a 29-KDa molecular mass species on a 4–20% polyacrylamide gel under nonreducing conditions. The AE protein in prostate tissue treated with DHT, or SHBG plus E2, was increased substantially, relative to the respective controls (Fig. 2). In tissue treated with SHBG or E2 alone, the AE activity increased only slightly, compared with control.
Figure 2. Western blot of prostate tissue homogenates treated with steroid hormones. Samples were resolved under nonreducing conditions on a 4–20% linear gradient polyacrylamide gel, followed by electrophoretic blotting on a PVDF membrane. AE, arginine esterase standard; lane 1, control; lane 2, DHT (100 nM); lane 3, DHT + hydroxyflutamide; lane 4, SHBG; lane 5, SHBG + E2; lane 6, E2 (lanes 1–6, 200 ng protein of tissue homogenate were loaded in each well). Similar results were obtained in two additional experiments.
Effect of AR antagonists on SHBG-mediated stimulation of AE activity.
Hydroxyflutamide (100 nM), a specific AR antagonist, inhibited the DHT-mediated increase in AE (Fig. 3). Similarly, the increased activity after treatment with SHBG+E2 or SHBG+3-diol also was blocked by hydroxyflutamide (Fig. 3). Western blotting revealed parallel results in AE protein.
Figure 3. Effect of hydroxyflutamide on AE activity. Prostate minces treated with steroid only or with 100 nM hydroxyflutamide (HF) plus steroid. The SHBG receptor of prostate tissue was saturated with dog SHBG before treatment with E2 and 3-diol. (), tissue treated in the absence of HF; (), tissue treated in the presence of HF; control, vehicle only; *, P < 0.05, compared with the same treatment in the absence of HF. Each data bar represents the mean ± SEM of six determinations derived from two independent experiments.
Effect of an SHBG-cAMP pathway antagonist
In confirmation of our previous studies (3), intracellular cAMP levels were stimulated markedly by the combination of SHBG, E2, and 3-diol (Fig. 4A). In a dose-dependent manner, 2MeOE2 inhibited the accumulation of cAMP stimulated by E2 or 3-diol (Fig. 4A). Importantly, the stimulation of AE by E2 and 3-diol also was blocked by 2MeOE2 (Fig. 4B). Thus, 2MeOE2 blocks both the accumulation of intracellular cAMP and the stimulation of AE activity that is initiated by E2 or 3-diol and suggests that cAMP is a necessary intermediator between E2- or 3-diol-SHBG-RSHBG and AE stimulation. Indeed, both the cAMP analog, 8-Br-cAMP, and the inducer of cAMP forskolin stimulated AE activity (Fig. 5).
Figure 4. Effect of 2MeOE2 treatment on cAMP levels present in prostate tissue. (A), Concentration dependent on 2MeOE2 (nM) inhibition of cAMP accumulation induced by E2 and 3-diol in the presence of SHBG. (B), Effect of 2MeOE2 on AE activity. Prostate tissue minces saturated with SHBG, were treated either with steroid hormone alone () or in the presence of 1 µM 2MeOE2 (). Control, tissue AE activity without treatment; SHBG only, tissue saturated with SHBG; SHBG+E2, SHBG-saturated tissue treated with 100 nM E2 alone () or in the presence of 2MeOE2 (); SHBG+3-diol, SHBG-saturated tissue treated with 3-diol alone () or in the presence of 2MeOE2 (); *, P < 0.05, compared with the same treatment in the absence of 2MeOE2. Each data bar represents the mean ± SEM of six determinations derived from two independent experiments.
Figure 5. Effect of stimulating the cAMP pathway on AE activity. AE activity was measured in the homogenates of prostate tissue treated with 100 nM DHT and compared with the effects of 50 µM forskolin (Fsk), a stimulator of cAMP production, or 50 µM 8-Br-cAMP. **, P < 0.05, compared with the control. Each data bar represents the mean ± SEM of six determinations derived from two independent experiments.
Androgens are important mediators of prostate growth, but although androgens require ARs to mediate their effects, the converse may not be true. Both in transformed cell lines and in appropriately transfected cells, evidence has been presented showing that the AR can be activated by substances other than steroids (19, 20, 21). However, the relevance of these observations to actual physiology has been problematical. The old observations that E2 and 3-diol induced prostate hyperplasia in dogs (31, 32, 33, 34, 35), together with our demonstration that both these steroids caused increases in cAMP in dog prostate, led us to hypothesize that these events might be connected (3). We speculated that perhaps ARs in the prostate were being activated by a cAMP-mediated pathway.
To test our hypothesis, we investigated the effects of E2 and 3-diol on an androgen-responsive protein in the dog prostate. AE is a dog prostate-specific protein under androgenic control that can be employed as a prostate marker, analogous to the use of PSA in humans (23, 24, 25, 26, 27, 2. We first confirmed that AE in dog prostate tissue was stimulated by DHT and that this effect could be antagonized by the specific AR antagonist, hydroxyflutamide. We then demonstrated that E2 and 3-diol caused AE increase in prostate in the presence (but not in absence) of SHBG. The effects were not mediated by the estrogen receptor because, in the absence of SHBG, E2 alone had no effect, and 3-diol does not bind to the estrogen receptor. The possible conversion of 3-diol to DHT does not explain the effects of 3-diol, because increases in AE required preincubation of prostate tissue with SHBG; 3-diol alone was ineffective. 2MeOE2 a competitive inhibitor of E2 and 3-diol binding to SHBG, blocked stimulation of cAMP by E2 and 3-diol and prevented increases in AE caused by these steroids, suggesting that cAMP was involved in increasing AE. That 8-Br-cAMP and an inducer of cAMP (forskolin) also increased AE is consistent with this explanation. Because the AR selective antagonist, hydroxyflutamide, blocks E2, 3-diol, and cAMP induction of AE, it seems that these effects are mediated through ARs. Furthermore, these results support the notion that ARs can be activated, even in the absence of androgens, by natural steroids that activate the cAMP pathway. Blockage of ligand-independent AR activation by antiandrogen is in agreement with several other recent reports.
Based on our findings, we propose a novel mechanism whereby E2 and 3-diol can increase the production of androgen-responsive proteins in a physiologically relevant system. The AR, like other steroid hormone receptors, is a phosphoprotein (44, 45), and its activation state can potentially be modulated by phosphorylation-dephosphorylation, resulting in augmented or ligand-independent activation (17, 19, 21, 46, 47, 48, 49). The phosphorylation state of the AR has been shown to be increased upon hormone binding; however, changes in cAMP levels and activation of protein kinase A also have recently been implicated in causing ligand-independent activation of transfected ARs when expressed in either CV-1 or human prostate (PC-3) cells (17, 20, 21). Consistent with our results in normal prostate, it is therefore plausible that activation of the SHBG-RSHBG pathway can lead to androgen-independent AR activation via stimulation of protein kinase A. It is also possible that up-regulation of AR expression might occur via a cAMP response element present within the regulatory region of the AR gene (10, 13). This could potentially augment AE expression, causing an increase in AR concentrations and a corresponding increase in basal transcription.
A number of lines of evidence have implied a role for E2 in the pathogenesis of BPH or in androgen-independent progression of prostate cancer (31, 32, 33, 34, 35, 36, 37, 3. These studies indicate that E2 synergizes with 3-diol, but not DHT, in induction of canine BPH. Because E2 and 3-diol are the only two known steroids that activate the SHBG-RSHBG pathway in prostate tissue (3), and we have shown that both are capable of activating pathways normally considered androgen responsive, antagonism of the pathway by which SHBG leads to the induction of androgen-responsive genes may be a valuable therapeutic target for the treatment or prevention of BPH or prostate cancer. To our knowledge, this is the first demonstration in a physiologically relevant system that androgenic events can be observed in the complete absence of exogenous androgens.
