help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Prins, G. S.
Right arrow Articles by van Breemen, R. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Prins, G. S.
Right arrow Articles by van Breemen, R. B.
Endocrinology Vol. 143, No. 9 3628-3640
Copyright © 2002 by The Endocrine Society

ARTICLE

Retinoic Acid Receptors and Retinoids Are Up-Regulated in the Developing and Adult Rat Prostate by Neonatal Estrogen Exposure

Gail S. Prins, William Y. Chang, Yan Wang and Richard B. van Breemen

Department of Urology, College of Medicine (G.S.P., W.Y.C.), and Department of Medicinal Chemistry and Pharmacognosy, College of Pharmacy (Y.W., R.B.v.B.), University of Illinois, Chicago, Illinois 60612

Address all correspondence and requests for reprints to: Gail S. Prins, Ph.D., Department of Urology, University of Illinois, M/C 955, Chicago, Illinois 60612. E-mail: gprins{at}uic.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Exposure to estrogens during the neonatal period interrupts rat prostatic development by reducing branching morphogenesis and by blocking epithelial cells from entering a normal differentiation pathway. Upon aging, ventral prostates exhibit extensive hyperplasia, dysplasia, and massive lymphocytic infiltrate, suggesting that neonatal estrogens may predispose the prostate gland to precancerous lesions. Vitamin A (retinol) and their derivatives (retinoic acids) are known key developmental regulators that bind and activate retinoic acid receptors (RARs). To evaluate whether neonatal estrogenization alters the sensitivity of the developing rat prostate to retinoids, RAR{alpha}, -ß, and -{gamma} cellular localization and protein levels were analyzed over the course of development and into adulthood by immunocytochemistry and Western analysis, whereas mRNA levels were measured using RT-PCR. In addition, intraprostatic retinol and retinoic acid levels were quantitated on d 10 and 90 using HPLC-mass spectroscopy. Male rats were given 25 µg estradiol benzoate or oil on d 1, 3, and 5 of life, and prostatic complexes were removed on d 6, 10, 15, 30, and 90. The RARs localized to distinct cell populations: RARß was expressed within basal epithelial cells, RAR{alpha} was localized to differentiated luminal epithelial cells and smooth muscle cells, and RAR{gamma} was expressed within periductal stromal cells. Over the normal course of development, total protein and mRNA levels for the RARs declined, so that the adult prostate possessed the lowest amounts of RAR. Exposure to estrogens during the neonatal period resulted in an immediate and sustained increase in RAR{alpha} levels and in the number of cells that expressed RARß, whereas RAR{gamma} levels were unaffected. Western analysis confirmed that total prostatic RAR protein levels were significantly increased, whereas RT-PCR demonstrated that RAR{alpha} and RARß mRNA levels were markedly elevated in response to estrogenic exposure. The total prostatic retinol content was tripled by estrogenic exposure on d 10 and 90, indicating that the ability to retain retinoids within the prostate was permanently increased. Intraprostatic levels of 9-cis- and all-trans-retinoic acid levels were reduced on d 10, whereas 13-cis-retinoic acid levels were increased in response to estrogens. In the adult prostates of rats exposed neonatally to estrogen, total retinoic acid levels were doubled due to significant increases in both 9-cis- and 13-cis-retinoic acids compared with those in control prostates. In summary, levels of specific RARs and their activating ligands are increased in the prostate gland after neonatal estrogenic exposure, and this effect is permanent throughout the life of the animal. Thus, we hypothesize that alterations in morphogenesis as well as dysplasia in the adult prostate may be mediated in part through augmentation of transcriptional signals in the retinoid pathway.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
ESTROGENS are believed to play a physiological role in the prostate gland and have been implicated in the induction of prostate cancer (1, 2, 3). Estrogen action is mediated through two specific intracellular receptors, estrogen receptor-{alpha} (ER{alpha}) and ERß, both of which are expressed in the human and rat prostate gland (4, 5, 6, 7, 8). During fetal development, maternal estrogens have a direct effect on the human prostate, leading to squamous metaplasia at the time of birth (9, 10, 11). In the rodent the prostate gland is particularly sensitive to estrogens during early morphogenesis, which occurs during the perinatal period. Work from our laboratory and others has shown that brief exposure of rats to high levels of estrogen during the initial 1–5 d of neonatal life leads to permanent alterations in the prostate gland, a phenomenon referred to as estrogen imprinting or developmental estrogenization (12, 13, 14, 15, 16). The effects are lobe-specific, with the ventral prostate being the most prominently affected (14). The estrogen-induced changes include differentiation defects of the epithelium, which give rise to hyperplasia and dysplasia with aging, reduced expression of secretory genes, reduced responsiveness to androgens in adulthood, stromal hyperplasia, and infiltration of inflammatory cells (12, 14, 17, 18, 19, 20). Thus, neonatal estrogenization in the rat is used as a model to evaluate the role of exogenous and endogenous estrogens as a potential predisposing factor for prostate diseases later in life, including prostatic carcinoma (21, 22). As such, the role of estrogens would involve the initiation of permanent cellular changes early in life, and subsequent events during adulthood would be required for promotion of these changes.

Most recently we have determined that estrogen imprinting of the developing rodent prostate gland is initially mediated through stromal ER{alpha}, which are transiently up-regulated in response to estrogenic exposure (23). This, in turn, leads to transient as well as permanent alterations in the levels of several members of the steroid receptor superfamily that are involved in mediating hormonal regulation of prostate development and growth throughout the life of the animal. Thus, after neonatal estrogen exposure, androgen receptor levels are permanently suppressed, stromal ER{alpha} and progesterone receptors are transiently up-regulated, and epithelial ERß expression exhibits delayed suppression upon glandular maturation (4, 5, 14, 24, 25). However, the potential roles of retinoids and their cognate receptors in prostate development and the estrogenized phenotype have not been previously investigated.

Retinoids are derivatives of vitamin A (retinol) and are major regulators of cell differentiation and tissue morphogenesis (26, 27). Their roles in differentiation and proliferation have logically led to many studies of their ability to prevent cancer. Retinoids have been shown to retard the processes that lead to a malignant phenotype in vitro and those that lead to carcinogenesis in vivo (28, 29). The retinoids believed to be responsible for the majority of effects on cell differentiation and proliferation are isomers of retinoic acid (RA). RAs bind two families of receptors, RA receptor (RAR) and retinoid X receptor (RXR), both of which are members of the steroid-thyroid receptor gene superfamily (30, 31, 32). RAR and RXR each possess three subtypes ({alpha}, ß, and {gamma}), which are differentially expressed during the development of various tissues (27). Although RARs are activated by several RAs, including the all-trans and 9-cis isomers, RXRs are more selective and bind 9-cis isomers exclusively (32, 33, 34). Once activated by ligand binding, RARs and RXRs homo- or heterodimerize before binding to RA response elements of regulated genes. The various dimer species potentially vary in their binding affinities and their recognition of specific response elements, and this is thought to be the basis of the highly pleiotropic effect of RAs.

Evidence that RAs and/or RARs may be involved in the process of prostatic neonatal estrogenization includes studies demonstrating that vitamin A deficiency can mimic epithelial changes associated with estrogen treatment. Vitamin A has a pronounced role in epithelial growth and differentiation, as demonstrated by the hallmark findings of metaplasia and keratinization of tracheal epithelium in vitamin A-deficient rodents (35, 36). Vitamin A deficiency also results in keratinization of genital epithelia. In mice, vitamin A deficiency-induced keratinization of the vaginal epithelium is similar to that seen with estrus and with neonatal estrogenization of the female reproductive tract, suggesting that retinoids may be mediating the action of estrogen (37, 38). In support of this theory, neonatal estrogen induction of proliferation and cornification of the vaginal epithelium can be suppressed by coadministration of retinyl acetate (38, 39). Mariotti et al. (40) demonstrated that squamous metaplasia of the mouse anterior prostate induced by adult estrogen exposure was blocked by the concomitant administration of RA. Conversely, high dose administration of all-trans-RA to neonatal mice led to inhibition of prostate ductal morphogenesis (41), whereas culture of neonatal rat anterior prostates in millimolar levels of 13-cis- or all-trans-RA inhibited ductal growth and branching (42). Thus, dependant on the timing of exposure (adult vs. development), both retinol deficiency as well as high dose exposure to RAs can lead to prostatic abnormalities. Additionally, null mutant mice for RAR{gamma} exhibit prostatic squamous metaplasia, indicating that alterations in receptor expression can interfere with normal prostatic growth (43).

The presence of RA-binding sites in the prostate was demonstrated in studies of the expression patterns of RARs during mouse organogenesis conducted by Dollé et al. (44). Although RAR{alpha} mRNA was ubiquitously expressed, and RARß and -{gamma} were expressed in the genital tubercle, only RARß mRNA was detectable in the prostate mesenchyme and genital tract epithelium. This suggests that RARß might mediate the effects of retinoids on the differentiation status of the prostate. An RT-PCR screen of the rat anterior prostate (coagulating gland) revealed the presence of RARß and -{gamma} on d 0, whereas RAR{alpha} and the RXR isomers were undetectable (42). More recently, Huang et al. (45) detected transcripts for RAR{alpha} and -{gamma} in the mature rat prostate using Northern analysis, indicating that either the rat prostate possesses a different RAR profile than the mouse or that RAR{alpha} and -{gamma} are more prominent in the adult prostate.

In the present study we sought to determine the expression profiles and cellular localization of RAR{alpha}, -ß, and -{gamma} in the rat prostate gland over the course of development and into adulthood to better understand the potential roles of these transcription factors in the hormonal regulation of prostate morphogenesis and growth. We next examined whether neonatal exposure to estrogens could alter the expression patterns and cellular localization of these transcription factors at the protein and mRNA levels. In addition, the intraprostatic levels of retinols and RAs were measured within the developing and adult prostate gland to determine whether estrogenic exposures could affect retinoid uptake, transport, and/or metabolism. Our findings reveal that neonatal estrogen exposure increases both prostatic RAR levels as well as retinol and RA retention, thereby providing a framework for amplified retinoid signaling during the critical developmental period. We postulate that elevation of these transcription factors and their cognate ligands contributes to the homeotic shift and differentiation defects observed within the developing prostate gland and that permanent shifts in their expression may be involved in dysplastic growth upon aging.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Animals
All rats were handled in accordance with the principles and procedures of the Guiding Principles for the Care and Use of Animal Research, and experiments were approved by the institutional animal care committee. Timed pregnant female Sprague Dawley rats were purchased from Zivic-Miller (Pittsburgh, PA), housed individually in a temperature (21 C)- and light (14 h of light, 10 h of darkness)-controlled room, and fed standard Purina rat chow (Ralston Purina Co., St. Louis, MO) ad libitum. They were monitored daily for delivery of pups, and the day of birth was designated d 0. Pups were sexed according to ano-genital distance, and female pups were removed. All males from a single mother were assigned to one of two treatment groups given sc injections of either 25 µg estradiol benzoate (Sigma, St. Louis, MO) in 25 µl peanut oil or oil alone on neonatal d 1, 3, and 5. Pups from both treatment groups were killed by decapitation on d 1–6, 10, 15, 30, or 90. Accessory sex gland complexes were quickly removed and placed in ice-cold PBS. Prostatic complexes or individual lobes were microdissected at 4 C under a dissecting microscope. Tissues designated for RNA or protein isolation or for quantitation of retinoids were snap-frozen in liquid nitrogen. For immunocytochemistry with frozen sections, tissues were arranged on a nylon square, covered with OCT (Miles, Elkhart, IN) compound, frozen in liquefied propane, and subsequently stored in liquid nitrogen.

Immunocytochemistry
Immunocytochemistry was performed according to previously described methods (46). Briefly, frozen prostatic complexes or individual lobes were mounted on precooled chucks (-20 C) in a Reichert-Jung cryostat, and 6-µm sections were thaw-mounted on gelatin-coated glass slides. Whenever possible, individual lobes were sectioned longitudinally to reveal the proximal-distal orientation. The sections were fixed in acetone for 2 d at -70 C, followed by fixation in 2% paraformaldehyde at 4 C. The tissue was incubated with appropriate 2% blocking goat serum and subsequently incubated overnight with primary antibody at 4 C. The specific antibodies, sources, and concentrations used are presented in Table 1Go.


View this table:
[in this window]
[in a new window]
  		Table 1. Antibodies used for immunocytochemistry

 
For single antibody labeling, the primary antibody was reacted with biotinylated secondary antibody (Vector Laboratories, Inc., Burlingame, CA), and the biotin was detected with an avidin-biotin peroxidase kit (ABC-Elite, Vector Laboratories, Inc.) using diaminobenzidine tetrachloride as a chromagen. The sections were stained with Gill’s #3 hematoxylin (1:4) as a blue nuclear counterstain. As a final step, the sections were dehydrated gradually with alcohol, cleared with xylene, and coverslipped with Permount (Fisher Scientific, Itasca, IL).

For dual antibody labeling, a fluorescein isothiocyanate (FITC)-labeled antirabbit secondary antibody was used against antibodies specific for RARs. After blocking the slides with 2% horse serum, antibodies specific against cell markers (smooth muscle actin, macrophage, T cell receptor, CD8+ and CD4+ lymphocytes, granulocytes, luminal, and basal epithelial cells) were incubated on the slides for 1 h at 37 C. Fluorescein (Cy3)-labeled antimouse secondary antibody was used against antibodies specific for the cell markers. The slides were mounted with Vectashield containing 4',6-diamido-2-phenylindole hydrochloride (Vector Laboratories, Inc.) to visualize nuclei. Photographs of slides were taken using a Axioshop (Carl Zeiss, Inc., Thornwood, NJ) equipped with either a Axiocam digital color camera or, for fluorescent images, a Microview (Princeton Instruments, Trenton, NJ) digital black and white camera, followed by color reconstruction in Adobe Photoshop.

As negative controls, rabbit IgG (Vector Laboratories, Inc.) or normal mouse ascites fluid (Sigma) was substituted for primary antibody. To determine specificity, the primary antibody was preincubated with a 10-fold molar excess of the corresponding antigenic peptide for 1 h at room temperature. This preabsorbed antibody was used in place of primary antibody while an adjacent section was incubated with the primary antibody alone. Comparison of staining between the two sections was used to confirm staining specificity. For comparative studies, tissues from different days of age as well as from control and estrogenized rats were always run in parallel to reduce discrepancies related to interassay variability in staining intensity. In most instances, photographs comparing treatments and days were taken from tissues processed on the same glass slide. Tissues from a minimum of three animals at each age and treatment were evaluated to ensure the reproducibility of results.

Western blots
Total proteins (cytoplasmic and nuclear) were isolated by homogenizing ventral prostates in a high salt buffer containing protease inhibitors (10 mM Tris-HCl and 400 mM NaCl, pH 7.4, containing 0.1 mM leupeptin, 1 IU/ml aprotinin, and 2 mM phenylmethylsulfonylfluoride). Cellular debris was removed by centrifugation at 10,000 x g for 10 min at 4 C, and proteins were denatured by boiling in SDS sample buffer [50 mM Tris (pH 6.8), 2% sodium dodecyl sulfate, 5% 2-mercaptoethanol, 10% glycerol, and 0.005% bromophenol blue]. Protein concentration was quantified by the Bradford assay (Bio-Rad Laboratories, Inc., Hercules, CA). Proteins (50 µg each) were separated by SDS-PAGE and electrotransferred onto a nitrocellulose membrane as previously described (46). The membrane was immunoblotted with a pan antibody that recognized all RARs (5 µg/ml; RAR82, Dr. Tuohimaa, University of Tampere, Finland). Biotinylated antigoat secondary antibody (Vector Laboratories, Inc.) was detected using ECL Plus chemiluminescent detection system (Amersham Pharmacia Biotech, Piscataway, NJ). The procedure was repeated three times using separate tissues for each run.

RT-PCR
RT-PCR was performed on d 6, 10, 30, and 90 on ventral prostate tissue from oil control and estrogen-treated rats in duplicate, and each dataset was repeated with three separate sets of tissues. Comparisons between oil and estrogen treatments were performed in parallel. RARs and RPL19, a ubiquitous ribosomal RNA, were reverse transcribed and coamplified to obtain semiquantitative results. For RAR{alpha} and RARß, the receptor and RPL19 cDNAs were amplified in the same tube, whereas for RAR{gamma}, the receptor and RLP19 cDNAs were amplified in parallel tubes, because coamplification interfered with RAR{gamma} efficiency. Total RNA was isolated from prostate tissue using guanidinium thiocyanate-chloroform extraction (RNA STAT-60, Tel-Test, Friendswood, TX). Two micrograms of total RNA were reverse transcribed at 37 C for 60 min in 25 µl first strand buffer (Promega Corp., Madison, WI) with 0.4 mM deoxy-NTPs through use of random hexamers and 200 U murine leukemia virus reverse transcriptase (Promega Corp.). The reaction was terminated by boiling for 5 min, and 1.25 µl RT product were amplified by PCR in a mixture containing 0.15 mM MgCl2, 1x PCR buffer, 0.16 mM deoxy-NTPs, 7.5 pmol of each RAR primer (see Table 2Go), 0.5 U Taq DNA polymerase (Perkin-Elmer, Norwalk, CT), and 10 µCi [32P]deoxy-CTP (Amersham Biotech). Initial experiments determined that gene amplification was in the logarithmic phase at 30 PCR cycles for each set of RAR primers. Amplification of RARs was carried out for 30 cycles in a Perkin-Elmer 9600 thermal cycler at 94 C for 30 sec, 55 C for 30 sec, and 72 C for 45 sec, with a final extension at 72 C for 4 min. As RPL19 reached its logarithmic phase of amplification between 16–24 PCR cycles, 7.5 pmol of each RPL19 primer were added to the PCR reaction after the 10th cycle to allow for coamplification of both the RAR and the internal standard within their respective logarithmic phases. The radiolabeled cDNAs were separated on 3% NuSieve/agarose (3:1) gel (FMC, Rockland, ME), and specific radioactive bands were quantitated on a PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA) after transfer of emitted radioactivity on a phosphor plate. The intensity of the RAR signals was normalized to that of the ribosomal protein RPL19 internal control.


View this table:
[in this window]
[in a new window]
  		Table 2. Primers used for RT-PCR

 
Liquid chromatography-mass spectrometry (LC-MS)
Prostatic cis- and all-trans-retinol and all-trans, 9-cis-, and 13-cis RA were separated and simultaneously quantified using an LC-MS method specifically designed for this study (47). For samples from d 10 rats, 8–10 ventral prostates were pooled to obtain a sample size of approximately 8 mg for extraction and analysis. For d 90 rats, 1 ventral lobe was homogenized and then fractionated to give a sample size equivalent to approximately 8 mg tissue to prevent variance in data due to tissue heterogeneity. Extraction and analysis were repeated on 4 separate sets of tissues from each treatment group. Tissue samples were homogenized and extracted after simple protein precipitation without derivitization. Briefly, ventral prostates from d 10 and 90 oil- and estrogen-treated rats were homogenized on ice in deionized water and extracted twice with ice-cold ethanol, and the combined supernatants were evaporated to dryness under vacuum. LC-MS was carried out using a G1946A LCMSD quadrupole mass spectrometer (Agilent, Palo Alto, CA) equipped with a series 1100 HPLC system. Baseline separation of 9-cis-, 13-cis-, and all-trans-RA was obtained by using a nonporous silica C18 column (Micra, Northbrook, IL). During positive ion atmospheric pressure chemical ionization, selected ion monitoring of the ions m/z 301 and m/z 269 with a dwell time of 889 ms/ion was used to record the most abundant ions. The ion of m/z 301 corresponded to the protonated molecules of RA, whereas the ion of m/z 269 corresponded to loss of water or acetic acid from the protonated molecules of retinol or the internal standard retinyl acetate, respectively. Standard solutions were prepared by dissolving all-trans-RA, all-trans-retinol, and all-trans-retinyl acetate in methanol and diluting with HPLC solvent A. Concentrations were determined by the weight of the standards dissolved. The method had a linear response over a concentration range of at least 3 orders of magnitude. The limit of quantitation was 0.702 pmol all-trans-RA injected on the column. The intraassay coefficients of variance (CV) were 3.6% and 2.2% on 2 different d, and the intraassay CVs of retinol were 3.0% and 1.9%. The interassays CV for RA and retinol were 15% and 4.7%, respectively.

Statistical analysis
For RT-PCR data, the mean ± SD of the relative values of RAR mRNAs (normalized to RPL19) were obtained for each time point and treatment using three separate replicates of tissues. Data were analyzed by two-way ANOVA, followed by the Schiff test to determine statistical significance.

For retinoid analysis, the mean ± SD were obtained from four separate replicates of control and estrogen-treated prostates on d 10 and 90. Data were analyzed by three-way ANOVA for treatment, time, and assay, and statistical significance was determined by Bonferroni’s test.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Total RAR protein
Western blot analysis using a pan antibody that recognizes all three isoforms of RAR revealed two closely migrating bands at 56 and 54 kDa in d 10 rat prostatic homogenates (Fig. 1Go), which correspond to the known molecular weights of RARß (56 kDa) and RAR{alpha} and RAR{gamma} (54 kDa). The doublet was most noticeable on d 10, because the 56-kDa band representing RARß disappeared upon maturation of the gland in the oil control prostates. Similarly, the 54-kDa band representing prostatic RAR{alpha} and RAR{gamma} declined with development in the control prostates, and on d 90 the adult ventral prostate had the lowest levels of total RARs. In contrast, neonatal exposure to estrogen resulted in increased levels of RARs at all time points examined with maximal RAR content seen in the adult prostate (Fig. 1Go). The most noticeable effect of estrogen was a significant increase in the 54-kDa band, with maximal levels observed on d 90. In addition, estrogen treatment increased the intensity of the 56-kDa band on d 10 and resulted in its continued presence on d 30.



View larger version (58K):
[in this window]
[in a new window]
  		Figure 1. Western blot analysis of RARs on total proteins from control and neonatally estrogenized rat ventral prostates using a pan antibody that recognizes all three RAR subtypes. Estrogen treatment enhanced total RAR levels at each time point compared with oil-treated control prostates. On d 10 there was a doublet at 56 and 54 kDa, which corresponds to the known molecular masses of RARß (56 kDa) and RAR{alpha} and -{gamma} (54 kDa). By d 30 the 56-kDa band was absent in the oil control prostates, but was present in the estradiol benzoate (EB)-treated tissue, albeit at lower levels than on d 10. By d 90 the 56-kDa band was absent in both treatment groups. In contrast, although there was a decline in signal intensity of the 54-kDa band in the oil control tissues from d 10 to 90, there was a marked increase in band intensity in the EB-treated prostates, with maximum signal observed on d 90. Due to a fanning effect during the electrophoresis run, the lanes on the left appear to run slower than the lanes on the right, as indicated by molecular mass markers on each side of the blot.

 
RAR{alpha}
Although the majority of experiments were performed on ventral lobe tissue, no differences were found in the immunolocalization of the RARs in the dorsal or lateral lobes. Immunocytochemistry using a specific RAR{alpha} antibody revealed that signal for RAR{alpha} protein was strongest in the perinuclear region of prostatic cells, with relatively weaker intranuclear staining. Competition experiments with excess antigenic peptide confirmed the specificity of this signal for RAR{alpha} (Fig. 2Go, A and H, insets). On d 6, RAR{alpha} was primarily expressed by periductal stromal cells, whereas interductal stromal cells and epithelial cells were largely negative for this receptor (Fig. 2AGo). As the periductal stromal cell layer thinned with glandular development between d 5–20, the number of RAR{alpha}-positive periductal cells surrounding each epithelial duct likewise declined (Fig. 2BGo). By d 30, in addition to periductal stromal cells RAR{alpha} signal was observed in the fully differentiated luminal epithelial cells of the control prostates (Fig. 2CGo). This dual compartment localization continued into adulthood, when RAR{alpha} immunostaining was equally intense in both the luminal epithelial and periductal stromal cells (Fig. 2DGo).



View larger version (137K):
[in this window]
[in a new window]
  		Figure 2. Immunocytochemistry of RARs in the developing ventral prostate glands of control and neonatally estrogenized rats. A–H, RAR{alpha} antibody (brown reaction product) counterstained with hematoxylin. In the d 6 prostate of oil control rats (A), RAR{alpha} immunolocalized to the perinuclear region of stromal cells closely associated with the epithelial ducts. Interductal stromal cells and epithelial cells were negative (inset shows competition of antibody with antigenic peptide on an adjacent section). On d 15 (B) the stromal compartment has thinned to a tight periductal layer, which is RAR{alpha} positive. By d 30 (C) specific perinuclear RAR{alpha} stain was present in the differentiated epithelial cells in addition to the positive stromal signal of control animals. On d 90 (D) all luminal epithelial cells and periductal stromal cells were strongly RAR{alpha} positive. For comparative purposes, neonatally estrogenized (neoEB) prostates (E–H) were mounted on the same glass slide as control tissue at each time point. On d 6 (E), 15 (F), and 30 (G) the layer of RAR{alpha}-positive stromal cells surrounding each epithelial duct was markedly thicker compared with oil-treated counterparts. By d 90 (H) the estrogenized prostate was filled with cells within the lumen, epithelial, and stromal compartments that were strongly RAR{alpha} positive (e, epithelium; s, stroma). The inset shows an adjacent section stained with antibody plus antigenic peptide for competition of specific signal. I–K (oil) and M–O (neoEB), RARß stain of d 6 (I and M), d 10 (J and N), and d 30 (K and O) prostates revealed that RARß localized to the nuclei of basally located cells in the developing prostates. As development proceeded, the number of RARß-positive cells declined and was observed only intermittently on d 30 (K and O, arrows). At each time point, there were significantly more RARß-positive cells in the estrogenized prostates compared with the oil-treated controls (processed on the same glass slide). The inset in M shows competition of antibody with antigenic RARß peptide on an adjacent section. L and P, Day 10 oil control (L) and estrogenized (P) prostates stained with RAR{gamma} antibody (brown) and hematoxylin showed that RAR{gamma} localized to a thin layer of periductal stromal cells directly adjacent to the basement membrane, with no observable differences between treatment groups. Occasional interductal stromal cells were also positive for RAR{gamma} in the estrogenized prostates (P, arrow). Magnification bars in A–N and L–P are the same as that shown in P and are equal to 20 µm. Magnification bars in K and O are equal to 20 µm.

 
Although neonatal estrogen treatment did not alter the localization pattern of RAR{alpha} across development, the overall number and density of RAR{alpha}-positive prostatic cells were markedly increased in response to estrogenic exposure. Between d 6–30, RAR{alpha} stain intensity in periductal stromal cells was greater than that in control prostates on side by side comparisons (Fig. 2Go, E–G). This was largely due to an increased thickness of the periductal stromal layer in estrogenized prostates at all time points examined, which reflects a proliferation of this compartment as a direct consequence of neonatal estrogen treatment (20). By d 90 there were dramatic differences in RAR{alpha} immunostaining in estrogenized ventral prostates compared with controls. Similar to controls, both epithelial and periductal stromal cells stained positively for nuclear RAR{alpha} with equivalent intensities. However, due to extensive hyperplasia and piling of epithelial cells in the estrogenized prostate, the RAR{alpha}-positive epithelial cell number was greatly increased (Fig. 2HGo). In addition, a large population of stromal cells outside the periductal region as well as cells within the lumen of the prostatic ducts stained positively for RAR{alpha}.

To further characterize the cell types expressing RAR{alpha} in the estrogenized adult prostate, fluorescent double labeling with RAR{alpha} (green) and specific cell markers (red) is shown in Fig. 3Go. Labeling with smooth muscle actin revealed that in the control (Fig. 3AGo) and estrogenized ventral prostate (Fig. 3Go, B and C), periductal smooth muscle cells were positive for RAR{alpha} (yellow arrows). Confirming previous findings, the RAR{alpha}-positive smooth muscle layer was thicker in the estrogen-treated prostates (Fig. 3BGo) compared with oil control prostates. The tissues in Fig. 3Go, A–C, were processed on the same glass slide, and the photographs were taken at the same light and time exposure to allow for comparison of stain intensity between the treatment groups. Under these conditions, it is apparent that the intensity of epithelial signal for RAR{alpha} is much greater in the estrogenized prostate (B and C) than in controls (A). Colocalization of RAR{alpha} (green) with luminal cell cytokeratins 8/18 (red) showed weak perinuclear stain and punctate nuclear signal for RAR{alpha} in luminal cells of the control d 90 prostate (Fig. 3DGo). In comparison, an estrogen-treated d 90 prostate examined on the same glass slide as in D and at the same light exposure revealed a stronger perinuclear RAR{alpha} signal in the piled luminal cells (Fig. 3EGo).



View larger version (77K):
[in this window]
[in a new window]
  		Figure 3. Double-labeled immunofluorescence of RARs with several cell markers was used to identify the cell types expressing RAR proteins. A, Day 90 control ventral prostate immunostained with RAR{alpha}-FITC (green) and RAR{alpha} and -{gamma} smooth muscle actin-Cy3 (red) shows that specific nuclear RAR{alpha} expression is low and localizes to the smooth muscle cells (arrow) and to punctate sites within the epithelium (e). B and C, Day 90 neoEB ventral prostate immunostained with RAR{alpha}-FITC (green) and RAR{alpha} and -{gamma} smooth muscle actin-Cy3 (red) shows a thicker layer of periductal smooth muscle cells all strongly positive for RAR{alpha} (B, arrow) and strong RAR{alpha} staining within epithelial and stromal (s) regions of the tissue (C). D and E, To confirm localization of RAR{alpha} within luminal epithelial cell, RAR{alpha}-FITC (green) and cytokeratins 8- and 18-Cy3 (red) were colabeled in d 90 ventral prostates. In control rats, RAR{alpha} appeared in luminal cell nuclei as a punctate signal (D). In neoEB prostates, RAR{alpha} signal was increased and localized to the perinuclear region of all luminal cells as well as punctate signal within the nuclei (E). F–H, To determine whether infiltrating immune cells expressed RAR{alpha}, d 90 neoEB ventral prostates were colabeled with RAR{alpha}-FITC (green) and antibodies to CD8 lymphocytes-Cy3 (red) (F), T cell receptor {alpha}/ß-Cy3 (red) (G), or granuloctes-Cy3 (red; H). Except for occasional granulocytes that stained for RAR{alpha} (H, arrow), the immune cells (red) did not colocalize for RAR{alpha} (green), as seen by a lack of yellow signal. I, Day 10 neoEB ventral prostate immunostained with RARß-FITC (green) and cytokeratin 5-Cy3 (red) shows that RARß is localized to the nuclei of basal cells. J, Day 90 neoEB ventral prostate immunostained with RAR{gamma}-FITC (green) and RAR{alpha} and-ß actin-Cy3 (red) shows that RAR{gamma} colocalizes to the periductal smooth muscle cells. The yellow color indicates that RAR{gamma} is present within the cytoplasm of the smooth muscle cells where the actin molecules are located, as opposed to the nuclear location of the other RARs. K, Day 90 neoEB ventral prostate immunostained with RAR{gamma}-FITC (green) and macrophage CD1-Cy3 (red) demonstrates that the occasional cells within the stromal region that are positive for RAR{gamma} are not fixed macrophages.

 
It is important to note that in adulthood, the prostatic ducts of the estrogenized rats are often filled with sloughed epithelial and inflammatory cells, and infiltrating lymphocytes and macrophages frequently flood the interstitium and penetrate into the epithelial ducts (14, 19, 23). To determine whether the large population of RAR{alpha}-positive cells in the estrogenized adult prostate was due to RAR{alpha} expression in infiltrating immune cells as has been shown with other tumors (48), colocalization with cell markers was performed (Fig. 3Go, F–H). Lymphocytes were stained with CD8+ antibodies (Fig. 3FGo), CD4+ antibodies (not shown), and T cell receptors {alpha} and ß (Fig. 3GGo). In all instances, there was no evidence of RAR{alpha} expression within lymphocytes, and RAR{alpha} signal (green) was confined to epithelial cells, as revealed by 4',6-diamido-2-phenylindole hydrochloride counterstain. Occasional granulocytes were immunopositive for RAR{alpha} (Fig. 3HGo, yellow arrow); however, the majority of the granulocyte population was RAR{alpha} negative. Thus, it was concluded that the large number of RAR{alpha}-positive cells within the d 90 estrogenized prostate are not immune cells and must be of prostatic origin.

To determine whether increased RAR{alpha} levels in the estrogenized prostate were a result of elevated RAR{alpha} mRNA expression, semiquantitative RT-PCR was performed. In the normal ventral prostate, RAR{alpha} mRNA expression levels remained constant throughout morphogenesis (d 6–30) and declined to the lowest expression levels in the d 90 adult prostates (Fig. 4Go). Neonatal exposure to estrogens resulted in increased RAR{alpha} mRNA expression at each time point, which reached statistical significance on d 90 (P < 0.05). Examination of RAR{alpha} mRNA levels across time in the estrogen-treated animals revealed that the maturation-associated decline in prostatic RAR{alpha} mRNA did not occur in the estrogenized prostates, and expression levels in adulthood remained at the high levels found in developing prostates (Fig. 3Go).



View larger version (60K):
[in this window]
[in a new window]
  		Figure 4. RAR{alpha} mRNA levels in oil control prostates ({blacksquare}) and neonatally estrogenized (EB) rat ventral prostates (). Top, Representative 32P-labeled gel of RT-PCR products for RAR{alpha} and RPL-19 mRNA fragments after transfer to a phosphorimager plate. Bottom, Graphic representation of RAR{alpha} mRNA levels normalized to RPL19 from three separate experiments. In control prostates, RAR{alpha} mRNA levels peaked on d 10 and declined thereafter. In the estrogenized prostates, although RAR{alpha} mRNA expression was higher than control levels at all points examined, the difference was most dramatic and significant on d 90. Bars represent the mean ± SD. a, P < 0.05 vs. d 90 oil-treated prostate.

 
RARß
The antibody used to detect RARß recognizes the ß1 and ß2 isoforms of this receptor, but does not cross-react with ß4, the smaller splice variant that acts as a dominant negative receptor (49). During normal postnatal development of the prostate, RARß immunolocalized to the nuclei of basally located epithelial cells (Fig. 2Go, I–K). The frequency of these immunopositive cells was greatest on d 6 (Fig. 2IGo) and declined with development. By adulthood, basally located cells immunostaining for RARß were rare in the ventral prostate gland (Fig. 2KGo). In the estrogenized ventral prostates the number of RARß-immunopositive cells was consistently elevated compared with that in control prostates. On d 6 the prostatic ducts were primarily solid epithelial cords, and the majority of the epithelial cells were strongly immunopositive for RARß (Fig. 2MGo). The specificity of this signal was confirmed by competition with excess peptide (Fig. 2MGo, inset). As development proceeded in the estrogenized prostates, the number of RARß-positive cells declined; however, in direct comparisons with control tissues, there were always more basally located cells that were RARß positive in the prostates exposed to neonatal estrogens (Fig. 2Go, N and O). As we have previously reported an accumulation of basal epithelial cells in estrogenized prostates (33), we suspected that these RARß-positive cells were undifferentiated basal epithelial cells. To test this, double-immunofluorescent labeling was performed using antibodies specific for RARß (green) and basal cell cytokeratin 5 (red) with results from d 10 shown in Fig. 3IGo. The FITC-labeled RARß antibody localized to the nuclei of basal epithelial cells, whereas the Cy3-labeled cytokeratin antibody immunolabeled the cytoplasm of these same cells, confirming that the basal cells express RARß.

To determine whether RARß mRNA expression was concomitantly altered by neonatal exposure to estrogens, semiquantitative RT-PCR was performed. In the control ventral prostate, RARß mRNA levels were highest during postnatal development on d 6 and 10 and markedly declined by d 30 to low expression levels (Fig. 5Go). In the normal adult prostate, RARß mRNA levels were extremely low and barely detectable by PCR. Neonatal estrogen exposure elevated RARß mRNA levels within the prostate at all time points examined, with the most significant effect noted on d 10 when expression levels peaked (Fig. 5Go). Thereafter, RARß message levels declined, albeit at a slower rate compared with controls. Thus, in the adult prostate of estrogenized rats, RARß was readily detectable by RT-PCR.



View larger version (48K):
[in this window]
[in a new window]
  		Figure 5. RARß mRNA levels in control prostates ({blacksquare}) and neonatally estrogenized rat ventral prostates (). Top, Representative 32P-labeled gel of RT-PCR products for RAR{alpha} and RPL-19 mRNA fragments after transfer to a phosphorimager plate. Bottom, Graphic representation of RAR{alpha} mRNA levels normalized to RPL19 from three separate experiments. In control prostates, RAR{alpha} mRNA levels were highest on d 6 and 10 and significantly declined with maturation of the gland. In estrogenized prostates, RAR{alpha} expression was higher at all time points, with significant differences observed on d 10 and 30. Bars represent the mean ± SD. a, P < 0.05 vs. d 10 oil-treated prostate; b, P = 0.08 vs. d 10 oil-treated prostate; c, P < 0.05 vs. d 30 oil-treated prostate; d, P < 0.05 vs. d 10 EB-treated prostate.

 
RAR{gamma}
Immunocytochemistry using specific antibodies showed that RAR{gamma} localized to the perinuclear region of stromal cells immediately adjacent to the basement membrane in the developing prostate gland (Fig. 2LGo). The specificity of this signal was confirmed with competition studies using excess peptide where no signal was observed (data not shown). The weak signal observed on d 10 remained through development into adulthood, where it was confined to the thin periductal smooth muscle layer. In contrast to RAR{alpha} and RARß, there was no difference in staining intensity for RAR{gamma} protein in prostates of rats exposed to neonatal estrogens (Fig. 2PGo). Occasionally, select interstitial cells stained intensely for RAR{gamma} in both control and estrogenized prostates (Fig. 2PGo, arrow). To determine whether these cells were macrophages that normally reside in the prostate interstitium, double-fluorescence immunolabeling was performed using an antibody specific to macrophages (ED2). However, the FITC-labeled RAR{gamma} antibody did not colocalize to the same cells as the Cy3-labeled ED2 antibody, indicating that these occasional cells were not resident macrophages.

Similar to the immunocytochemistry results, relative RAR{gamma} mRNA expression in the rat ventral prostate was constant throughout the postnatal developmental period to adulthood, and neonatal exposure to estrogens had no effect on RAR{gamma} mRNA levels at any time point examined (Fig. 6Go).



View larger version (40K):
[in this window]
[in a new window]
  		Figure 6. RAR{gamma} mRNA levels in control ({blacksquare}) and neonatally estrogenized rat ventral prostates (). Top, Representative 32P-labeled gels of RT-PCR products for RAR{alpha} and, in a parallel reaction, for RPL-19 mRNA fragments after transfer to a phosphorimager plate. Bottom, Graphic representation of RAR{gamma} mRNA levels normalized to RPL19 from three separate experiments. RAR{gamma} expression levels remained constant in control prostates throughout development and levels were not altered by neonatal estrogen treatment. Bars represent the mean ± SD.

 
Intraprostatic retinol and RA levels
Circulating retinol (vitamin A) is taken up by cells, retained by cellular retinol-binding proteins (CRBP), and serves as a precursor to RA isomers. After metabolism to retinal by retinol dehydrogenase, specific retinaldehyde dehydrogenase enzymes present within the prostate irreversibly metabolize the retinal isomers to RA isomers (50, 51). The mechanism of interconversion between the various RA isomers is unclear, but may involve tissue-specific isomerases. RARs are transcriptionally activated by several naturally occurring RAs, including all-trans-, 9-cis-, and, to a lesser degree, 13-cis-RA (31, 42, 50). To determine the levels of these specific ligands for the RARs within the rat prostate and whether they are altered in response to neonatal estrogen exposure, we developed an LC-MS method that was capable of simultaneously isolating and quantitating retinol and the biologically relevant isoforms of RAs within small samples of prostate tissue (47). We successfully identified cis- and trans-retinol and 9-cis-, 13-cis-, and all-trans-RA within both developing (d 10) as well as adult tissues. Retinol was present at similar levels in both the developing and adult ventral prostate lobes in the cis- and all-trans isomeric forms (Table 3Go), which indicates that stable levels of this precursor are maintained within the normal prostate gland. Although 9-cis-, 13-cis-, and all-trans-RA were present in both the developing and adult prostates, total RA levels were higher in the younger prostate than in the adult gland (Table 3Go). Interestingly, the levels of 9-cis- and 13-cis-RA were 2- to 5-fold higher than those of the all-trans isomer in the rat prostate (Table 3Go).


View this table:
[in this window]
[in a new window]
  		Table 3. Effects of neonatal estrogen on intraprostatic levels of retinol and retinoic acids

 
After neonatal exposure to estrogens, there was a significant increase in intraprostatic total retinol levels (Table 3Go). This effect was both immediate (d 10) and sustained into adulthood, indicating a permanent imprint in prostatic retinol retention. Although total RA levels were not immediately increased in the estrogenized prostates, the levels were doubled on d 90 compared with control prostates. When the levels of the individual RA isomers were analyzed, there were significant alterations observed on both d 10 and 90 after neonatal exposure to estrogens (Table 3Go). On d 10 there was a significant reduction in 9-cis- and all-trans-RA and a significant increase in 13-cis-RA in response to estrogen. On d 90, both 9-cis- and 13-cis-RA isomers were significantly elevated in response to estrogen treatment, whereas the levels of all-trans-RA were half the control prostate levels.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The present study documents and characterizes various ligands (retinol and RAs) and receptors (RAR{alpha}, -ß, and -{gamma}) of the RA signaling pathway in developing and adult rat prostates. We observed that the separate RAR{alpha}, -ß, and -{gamma} subtypes are differentially localized to distinct cell populations within the rat prostate gland. In the developing rat prostate, RARß is the predominant epithelial RAR and is localized to basal cell nuclei. This finding is similar to that of Dolle et al. (44), who observed that RARß mRNA was the predominant RAR in the embryonic mouse prostate, and supports the hypothesis that RARß may mediate the effects of retinoids on the differentiation of the epithelium. Upon maturation of the gland, the number of RARß-positive cells markedly declines. As the epithelium functionally differentiates between d 15 and d 30, RAR{alpha} expression commences within the luminal epithelial cells, and by adulthood RAR{alpha} is the predominant epithelial RAR within the prostate gland. Both RAR{alpha} and RAR{gamma} are found within prostatic stromal cells during development. RAR{alpha} localized to the periductal mesenchymal cells and, upon differentiation, smooth muscle cells, which thinned to a single cell layer as the gland matured. RAR{gamma} also localized to smooth muscle cells immediately adjacent to the basement membrane; however, its expression did not change with development of the gland. The fixed macrophage population and other infiltrating immune cells were mostly negative for these transcription factors within the prostate. As the prostatic cell population changes upon glandular development, the RAR levels shift accordingly, with an overall decrease in total RAR levels in the adult prostate. Thus, we predict that the normal adult prostate gland is less sensitive to retinoids than the immature developing prostate, which coincides with the known function of retinoids as regulators of tissue morphogenesis and differentiation. Furthermore, analysis of intraprostatic levels of retinoids showed similar levels of retinol in d 10 and 90 ventral prostates, but lower levels of total RAs in the adult tissues compared with developing glands, which also supports a greater role for retinoids during normal prostatic development and a reduced function of this signaling pathway in the adult prostate gland.

After neonatal exposure to estradiol, we observed a significant increase in prostatic RAR levels attributable to increases in RAR{alpha} and RARß in both developing and adult prostates. Thus, the effect of estrogens on RAR is similar to the effect on AR, in that it is both immediate and sustained (14, 24), but contrasts with the effects on ER{alpha} (4) and progesterone receptor, which are transient (25), and the effect on ERß effect, which is delayed (5). Previous studies have shown that estrogens are capable of inducing several RAR and RXR genes in cervical epithelium, breast cancer cells, and in vitro systems and have suggested that their induction may be one of the mediators of estrogenic effects (52, 53, 54). In the present study RAR{alpha} expression was increased within prostatic stromal cells immediately after neonatal estrogen exposure, and with maturation of the gland, epithelial cell RAR{alpha} expression was markedly enhanced. Similarly, the normal maturational decline in prostatic RAR{alpha} mRNA levels did not occur in the estrogenized prostates, resulting in continued high expression into adulthood, i.e. a permanent imprint. RARß levels were also elevated in all estrogenized prostates compared with controls, mostly as a result of an increased number of basal cells within the estrogenized tissue (24). Thus, in total, the prostates of estrogen-exposed rats have more cells expressing receptors for retinoid ligands than the normal prostate and can be considered more sensitive to retinoid signals both during development and into adulthood. As a delicate balance of retinoid levels and RARs is required for normal tissue homeostasis, we propose that the amplification of retinoid signaling after neonatal estrogen exposure may in part mediate the developmental and differentiation defects as well as the aging-associated dysplasia observed within this gland.

In addition to increases in RAR levels, the present results document that neonatal estrogenic exposure increases intraprostatic retinol stores and significantly alters the intracellular ratios of specific RA isomers, which implies that estrogens interfere with the retention and metabolism of retinoids within the prostate. Similar to the effect on RAR expression, the effect on retinoids is both immediate and sustained into the adult prostate gland. Unlike humans, in whom retinol is primarily derived from hepatic carotinoid metabolism, rodents derive their circulating retinol directly from the diet. After passive uptake by cells, retinol is normally retained within target cells through high affinity binding to CRBPs (50), and one possible explanation for the observed effects is that estrogens increase cellular levels of CRBPs and allow the cells to retain more retinol, which serves as an intracellular pool for retinoid signaling. The ratios of RA isomers, which directly bind to receptors, are also altered by estrogenic exposure both during development and in the adult prostate gland. Thus, on d 10, the 9-cis- and all-trans-RA levels decrease, whereas the 13-cis-RA levels increase in response to estrogens. In the adult prostate, total RA levels are doubled in the estrogenized prostate compared with control tissue due to a marked increase in 9-cis- and 13-cis-RA levels. It is known from in vitro studies that RARs bind to several retinoids, including all-trans, 9-cis-, and, to a lesser degree, 13-cis-RA, whereas RXRs are activated exclusively by 9-cis-RA isomers (32, 34); thus, differential receptor activation may occur in response to the altered RA metabolism. As both 9-cis and all-trans isomer levels were reduced on d 10, whereas 13-cis-RA increased, it is possible that RAR activation was limited at that time. On d 90, 9-cis and 13-cis levels were doubled, which may allow more access to RAR activation at that time. However, it is important to note that absolute levels of specific RAs have not been shown to be limiting for RAR activation in vivo (55). Future studies are aimed at analyzing CRBPs, RXRs, and the activities of critical retinoid-synthesizing and -metabolizing enzymes within these tissues with the aim of dissecting out the various levels of estrogenic regulation of this signaling system.

Previous in vitro studies with neoplastic and nonneoplastic prostate cells lines revealed divergent responses to all-trans-RA, with growth promotion of human PC-3 and DU-145 cells, but growth inhibition in canine prostate cell lines (56). Interestingly, a direct association was not observed with growth response and RAR expression levels in these separate cell lines. Studies that have evaluated retinoid and RAR content in normal and cancerous prostate samples are limited (57). Pasquali et al. (51) reported that the RA content of human benign prostatic hyperplasia samples was 2-fold higher than normal tissues, whereas in prostatic adenocarcinoma samples, levels were 5–8 times lower than those in normal prostates. A recent study on RAR and RXR expression in human prostate cancer compared with adjacent benign tissue found that although all six receptors were present in human prostate tissue, RARß and RXRß exhibited a loss of expression in the cancerous foci (58). These findings suggest a causative relationship between the loss of retinoid signaling and aberrant growth mechanisms within the adult human prostate gland.

As we observed an amplification of the retinoid signaling system in estrogenized rat prostates that develop dysplasia with aging, our results seemed paradoxical to the postulated function of retinoids as antitumorigenic (59, 60, 61). This might be explained by both hormone dosage and timing of exposure in the different experimental systems. With regard to dose, a biphasic response has been reported for the effects of RA on human prostatic epithelial cells in vitro (62), where low doses (0.03 nM) stimulated proliferation and higher doses (>3 nM) inhibited cellular proliferation and enhanced differentiation. In the present study the increase in intraprostatic retinoid levels was relatively modest, with a doubling in levels in the picomoles per milligram tissue range, and perhaps this increase is within the stimulatory range for the abnormal epithelial proliferation seen in the adult estrogenized prostate, whereas retinoid treatments at pharmacological levels may be inhibitory for prostatic tumor growth (61).

With regard to timing, prostates were undergoing morphogenesis at the time of neonatal estrogenic exposure, which led to immediate increases in prostatic RARs and retinoid levels. Although vitamin A deficiency leads to squamous metaplasia in the adult prostate (21), elevated retinol during the developmental period may have divergent effects. In support of this, two studies that directly examined the effects of retinoids on the developing prostate gland found that RAs inhibited ductal morphogenesis in vivo (41) and in vitro (42). Additionally, in the seminal vesicles, 13-cis was more potent than 9-cis in suppressing branching (63), which is significant in that intraprostatic 13-cis-RA levels were elevated on d 10 in the present study. Thus, we postulate that amplification of retinoid signaling during prostatic morphogenesis may be directly involved in branching, patterning, and differentiation defects that occur after early estrogen exposure.

There is ample evidence that RAs regulate homeobox genes involved in the morphogenesis of developing structures (64, 65, 66). Several hox genes are regulated by RAs, and specific RA response elements have been identified in the promoter regions of a number of hox clusters (67, 68). We have recently shown that expression of prostatic hox-13 genes are altered by neonatal estrogen exposure (16), and it is conceivable that this is mediated through the elevated RAR and/or retinoids present within the estrogenized prostate. Exposure to abnormal levels of RA has been shown to interrupt normal morphogenesis of several structures and lead to homeotic shifts, i.e. acquisition of adjacent structure morphology, secondary to disruption of hox gene expression. This is highly significant, because the estrogenized prostate is characterized as possessing a proximalized phenotype wherein the morphology of the proximal ducts extends toward the distal tips of the gland (16, 20, 24). RAs have been shown to shift the proximodistal developmental competence of the developing chick limb bud (69), whereas manipulation of RA levels in vitro led to an immature lung phenotype characterized by a failure to form distal buds (70). Likewise, a proximalized phenotype was reported for rat lungs exposed fetally to RAs (71), a phenomenon found to be mediated through changes in hoxa-2, hoxb-6, and sonic hedgehog expression (72). RARß has been shown to mediate the inhibition of distal bud formation in the developing lung (73), which is notable in light of the increased RARß expression found in estrogenized prostate. Thus, it is reasonable to propose that amplification of retinoid signaling in the estrogenized prostate leads to alterations in hox-13 expression and formation of a proximalized prostatic phenotype.

As shown for other developing structures (74), continuous branching morphogenesis of the prostate gland is dictated by time- and region-specific expression of master regulatory genes. In addition to tight feedback loops among the molecules involved, a growing body of evidence supports a role for steroids in regulating key developmental genes (75, 76, 77, 78, 79, 80, 81). We propose that in the prostate gland, androgens, estrogens, and retinoids act as upstream regulators of several of the critical developmental genes and that timely expression of their cognate receptors is critical for normal morphogenesis. Transient and permanent disturbances in the expression of these steroid transcription factors, as occurs after neonatal estrogen exposure, may consequently lead to abnormal or untimely expression of the prostatic master regulatory genes during the developmental critical windows. As development is an Einbahnstrasse, this will, in turn, lead to permanent disturbances in the structure and differentiation of this organ. These molecular alterations provide a mechanistic basis to explain how pre/neonatal insults create chronic functional defects in the prostate gland that, in turn, can predispose the gland to aging-associated dysplasia.


   Acknowledgments
 
The authors thank Lynn Birch, M.S., for superb technical assistance, and Manu Patel, Ph.D., for statistical consultation.


   Footnotes
 
This work was supported by NIH NIDDK Grants 40780 (to G.S.P.), 09873 (to W.Y.C.), and R24-CA-83124 (to R.B.v.B.).

Abbreviations: CRBP, Cellular retinol-binding protein; CV, coefficient of variance; ER, estrogen receptor; FITC, fluorescein isothiocyanate; LC-MS, liquid chromatography-mass spectrometry; RA, retinoic acid; RAR, retinoic acid receptor; RXR, retinoid X receptor.

Received February 13, 2002.

Accepted for publication May 14, 2002.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Sinowatz F, Ameselgruber W, Plendl J, Kolle S, Neumuller CH, Boos G 1995 Effects of hormones on the prostate in adult and aging men and animals. Microsc Res Tech 30:282–292[CrossRef][Medline]
  2. Castagnetta LAM, Carruba G 1995 Human prostate cancer: a direct role for oestrogens. Ciba Found Symp 191:269–289[Medline]
  3. Griffiths K 2000 Estrogens and prostatic disease. Prostate 45:87–100[CrossRef][Medline]
  4. Prins GS, Birch L 1997 Neonatal estrogen exposure up-regulates estrogen receptor expression in the developing and adult rat prostate lobes. Endocrinology 138:1801–1809[Abstract/Free Full Text]
  5. Prins GS, Marmer M, Woodham C, Chang WY, Kuiper G, Gustafsson JA, Birch L 1998 Estrogen receptor-ß messenger ribonucleic acid ontogeny in the prostate of normal and neonatally estrogenized rats. Endocrinology 139:874–883[Abstract/Free Full Text]
  6. Lau KM, Leav I, Ho SM 1998 Rat estrogen receptor-{alpha} and -ß, and progesterone receptor mRNA expression in various prostatic lobes and microdissected normal and dysplastic epithelial tissues of the Noble rats. Endocrinology 139:424–427[Abstract/Free Full Text]
  7. Ekman P, Barrack E, Greene G, Jensen E, Walsh P 1983 Estrogen receptors in human prostate: evidence for multiple binding sites. J Clin Endocrinol Metab 57:166–176[Abstract]
  8. Leav I, Lau KM, Adams J, NcNeal J, Taplin M, Wang J, Singh H, Ho SM 2001 Comparative studies of the estrogen receptors ß and {alpha} and the androgen receptor in normal human prostate gland, dysplasia, and in primary and metastatic carcinoma. Am J Pathol 1:79–92
  9. Brody H, Goldman SF 1940 Metaplasia of the epithelium of the prostatic glands, utricle and urethra of the fetus and newborn infant. Arch Pathol Lab Med 29:494
  10. Zondek T, Zondek LH 1975 The fetal and neonatal prostate. In: Goland M, ed. Normal and abnormal growth of the prostate. Springfield: Thomas; 5–28
  11. Driscoll SG, Taylor SH 1980 Effects of prenatal maternal estrogen on the male urogenial system. Obstet Gynecol 56:537–542[Abstract/Free Full Text]
  12. Rajfer J, Coffey DS 1978 Sex steroid imprinting of the immature prostate. Invest Urol 16:186–190[Medline]
  13. Pylkkanen L, Santti R, Newbold R, McLachlan J 1991 Regional differences in the prostate of the neonatally estrogenized mouse. Prostate 18:117–129[Medline]
  14. Prins GS 1992 Neonatal estrogen exposure induces lobe-specific alterations in adult rat prostate androgen receptor expression. Endocrinology 130:3703–3714[Abstract]
  15. Putz O, Schwartz CB, Kim S, LeBlanc GA, Cooper RL, Prins GS 2001 Neonatal low- and high-dose exposure to estradiol benzoate in the male rat. I. Effects on the prostate gland. Biol Reprod 65:1496–1505[Abstract/Free Full Text]
  16. Prins G, Birch L, Habermann H, Chang W, Tebeau C, Putz O, Bieberich C 2001 Influence of neonatal estrogens on rat prostate development. Reprod Fertil Dev 13:241–252[CrossRef][Medline]
  17. Pylkkanen L, Makela S, Valve E, Harkonen P, Toikkanen S, Santti R 1993 Prostatic dysplasia associated with increased expression of C-MYC in neonatally estrogenized mice. J Urol 149:1593–1601[Medline]
  18. Prins GS, Woodham C, Lepinske M, Birch L 1993 Effects of neonatal estrogen exposure on prostatic secretory genes and their correlation with androgen receptor expression in the separate prostate lobes of the adult rat. Endocrinology 132:2387–2398[Abstract]
  19. Prins GS 1997 Developmental estrogenization of the prostate gland. In: Naz RK, ed. Prostate: basic and clinical aspects, chapt 10. Boca Raton, FL: CRC Press; 247–265
  20. Chang WY, Wilson MJ, Birch L, Prins GS 1999 Neonatal estrogen stimulates proliferation of periductal fibroblasts and alters the extracellular matrix composition in the rat prostate. Endocrinology 140:405–415[Abstract/Free Full Text]
  21. Bern HA 1963 Epithelial metaplasia in the prostate and other genital structures of male mammals. Natl Cancer Inst Monogr 12:43–45
  22. Santti R, Newbold RR, Makela S, Pylkkanen L, McLachlan JA 1994 Developmental estrogenization and prostatic neoplasia. Prostate 24:67–78[Medline]
  23. Prins GS, Birch L, Couse JF, Choi I, Katzenellenbogen B, Korach KS 2001 Estrogen imprinting of the developing prostate gland in mediated through stromal estrogen receptor {alpha}: studies with {alpha}ERKO and ßERKO mice. Cancer Res 61:6089–6097[Abstract/Free Full Text]
  24. Prins GS, Birch L 1995 The developmental pattern of androgen receptor expression in rat prostate lobes is altered after neonatal exposure to estrogen. Endocrinology 136:1303–1314[Abstract]
  25. Sabharwal V, Putz O, Prins GS 2000 Neonatal estrogen exposure induces progesterone receptor expression in the developing prostate gland. J Urol 163:97 (Abstract)
  26. Sporn M, Roberts A 1983 Role of retinoids in differentiation and carcinogenesis. Cancer Res 43:3034–3040[Free Full Text]
  27. Ruberte E, Dolle P, Chambon P, Morriss-Kay G 1991 Retinoic acid receptors and cellular retinoid binding proteins. II. Their differential pattern of transcription during early morphogenesis in mouse embryos. Development 111:45–60[Abstract]
  28. Bollag W 1979 Retinoids and cancer. Cancer Chem Pharmacol 3:207–215[Medline]
  29. Lotan R 1980 Effects of vitamin A and its analogs (retinoids) on normal and neoplastic cells. Biochim Biophys Acta 605:33–91[Medline]
  30. Petkovich M, Brand N, Krust A, Chambon P 1987 A human retinoic acid receptor which belongs to the family of nuclear receptors. Nature 330:444–450[CrossRef][Medline]
  31. Giguere V, Ong E, Segui P, Evans R 1987 Identification of a receptor for the morphogen retinoic acid. Nature 330:624–629[CrossRef][Medline]
  32. Mangelsdorf D, Ong E, Kyck J, Evans R 1990 Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345:224–229[CrossRef][Medline]
  33. Levin A, Sturzenbecker L, Kazmer S, Bosakowski T, Huselton C, Allenby G, Speck J, Kratzeisen C, Rosenberger M, Lovey A, Grippo J 1992 9-Cis retinoic acid stereoisomer binds and activates the nuclear receptor RXR{alpha}. Nature 355:359–361[CrossRef][Medline]
  34. Chambon P 1996 A decade of molecular biology of retinoic acid receptors. FASEB J 10:940–954[Abstract]
  35. McDowell E, Keenan K, Huang M 1984a Effects of vitamin A-deprivation on hamster tracheal epithelium. Virchows Arch Cell Pathol 45:197–219
  36. McDowell E, Keenan K, Huang M 1984b Restoration of mucociliary tracheal epithelium following deprivation of vitamin A. Virchows Arch Cell