Toxicological Sciences 58, 324-338 (2000)
Copyright © 2000 by the Society of Toxicology
Reproductive and Developmental Toxicology |
2,3,7,8-Tetrachlorodibenzo-p-dioxin Inhibits Luminal Cell Differentiation and Androgen Responsiveness of the Ventral Prostate without Inhibiting Prostatic 5
-Dihydrotestosterone Formation or Testicular Androgen Production in Rat Offspring
,1,3
* School of Pharmacy and
Environmental Toxicology Center, University of Wisconsin, Madison, Wisconsin 53706
Received May 15, 2000; accepted August 2, 2000
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In utero and lactational exposure to a single maternal dose of 1 µg 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)/kg causes some overt toxicity and impairs prostate growth in male offspring. As similar effects on the ventral prostate can be caused by decreased testosterone production during perinatal development, we determined whether intratesticular testosterone content, testicular responsiveness to gonadotropin stimulation, or plasma testosterone concentrations were reduced in fetal and newborn rats. Because these endpoints were not affected, the ability of TCDD exposure to inhibit synthesis of the proximal androgen in prostate development, 5
-dihydrotestosterone (DHT), from the circulating precursor testosterone and 5-androstane-3,17ß-diol (3-Diol), was studied on postnatal days (PNDs) 14, 21, and 32. The ability of the ventral prostate to form DHT from 3-Diol was slightly impaired on PND 14, but this transient effect was not statistically significant, and recovery was evident by PND 21. Subsequent experiments used organ culture to study the effects of in vivo TCDD exposure on androgen metabolism, androgen responsiveness, androgen receptor expression, and luminal epithelial cell differentiation after in vitro exposure to graded androgen concentrations. In utero and lactational TCDD exposure had no effect on DHT formation in organ culture, but transiently reduced the androgen -induced expression of prostatic-binding protein subunit C3, decreased ventral prostate epithelial cell androgen receptor expression, and inhibited the formation of androgen responsive luminal epithelial cells. These results suggest that TCDD exposure impairs prostate growth and androgen responsiveness by inhibiting prostatic epithelial cell differentiation.
Key Words: 2,3,7,8-tetrachlorodibenzo-p-dioxin; rat, prostate; development; impaired epithelial differentiation; 5-dihydrotestosterone-forming enzymes; testicular testosterone production.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Exposure of male Holtzman rat offspring in utero and via lactation to a single low dose of TCDD (0.064–1.0 µg/kg maternal body weight) administered on GD 15 resulted in ventral prostate, epididymis, and seminal vesicle growth reductions (Mably et al., 1992a
). In addition, in utero and lactational TCDD exposure decreased neonatal anogenital distance, an external indicator of decreased androgenic status (Mably et al., 1992b), and reduced the responsiveness of the ventral prostate to androgenic growth stimulation in adulthood (Bjerke et al., 1994b). One possible explanation for these results is that in utero and lactational TCDD exposure reduces fetal plasma androgen concentrations. Indeed, our initial study in which male Holtzman rat offspring were exposed to 1 µg TCDD/kg maternal body weight on GD 15 reported reduced plasma testosterone concentrations between GD's 17 and 21, and at 2 h after birth (Mably et al., 1992a). However, an independent study found no effect from GD 15 TCDD exposure on plasma testosterone concentrations in Long-Evans rats on PND 0 (Gray et al., 1995).
Another possible explanation for some of the seemingly antiandrogenic effects of TCDD, particularly with respect to prostate development, is that in utero and lactational exposure inhibits the formation of 5-dihydrotestosterone (DHT). The rat ventral prostate begins to arise before birth from the prostate anlagen in the urogenital sinus region. Prostatic mesenchyme, under the influence of androgenic stimulation, induces the formation of epithelial buds that will eventually develop into secretory glands (Hayward et al., 1996). In order for this process to occur, however, circulating testosterone from the fetal testes must be enzymatically converted to DHT within the nascent prostate (George and Peterson, 1988). The development of the seminal vesicle and epididymis, like that of the prostate, requires androgenic stimulation. However, unlike the prostate, the seminal vesicle and epididymis do not require DHT for normal development (Clark et al., 1993; George and Peterson, 1988; Hayward et al., 1996). If DHT synthesis is globally inhibited during the period from gestational day (GD) 14 to GD 16, normal prostate differentiation and growth will be prevented (Clark et al., 1993), the normal course of the urethral opening to the penis tip will be deflected, and a hypospadia will result (Anderson and Clark, 1990). However, penis development appears normal and hypospadias are not seen in response to TCDD exposure (Mably et al., 1992b). It remains possible, though, that a local decrease in DHT synthesis activity may well play a role in decreasing ventral prostate growth, as in utero and lactational TCDD exposure decreased ventral prostate DHT concentration (Roman et al., 1995).
Prior to and just after birth, the predominant circulating androgen in fetal male rat blood is testosterone (Corpechot et al., 1981). This hormone is converted to DHT by two NADPH requiring isozymes of 5-reductase (5-R), which are the products of different genes, are distinguishable in enzyme assays by their different Michaelis constants and pH optima, and appear to play different roles in ventral prostate androgen metabolism (Berman et al., 1995; Moore and Wilson, 1976; Normington and Russell, 1992; Span et al., 1995). The type-1 5-R isozyme has a Michaelis constant in the micromolar range, a pH optimum in vitro of 6.9, and is the only form found in the ventral prostate epithelium. In contrast, the type-2 5-R isozyme has a Michaelis constant in the nanomolar range, pH optimum in vitro of 5.5, and is found in the developing ventral prostate exclusively in the developing mesenchyme and stroma (Berman et al., 1995; Span et al., 1995). Since androgen signaling in the developing fetal prostate initially occurs via the interaction of androgens with receptors in the mesenchyme, which in turn relay a signal to the epithelium, it appears that the type-2 5-R isozyme is the enzyme form required for this early androgenic activity (Berman et al., 1995; Chung and Cunha, 1983). The type-1 5-R isozyme, on the other hand, with its higher Km, appears to play a catabolic role, which primarily facilitates the elimination of testosterone from the ventral prostate, rather than activating androgen receptors.
Shortly after birth, the androgenic prohormone 5-androstane-3, 17ß-diol (3-Diol) transiently becomes the predominant circulating androgen in postnatal rat plasma until just before puberty (Corpechot et al., 1981). This precursor is converted to DHT in the developing ventral prostate by the action of 3-hydroxysteroid dehydrogenase (3-HSD) (Penning et al., 1996). Unlike 5-R, 3-HSD catalyzes a thermodynamically reversible reaction and prefers NAD rather than NADP as the cofactor. Because the enzyme reaction is thermodynamically reversible, 3-HSD can act as a molecular switch that provides a buffering action to resist changes in the intracellular DHT concentration (Penning et al., 1996).
Androgen receptors first appear in rat ventral prostate epithelial cells on GD 19 (Hayward et al., 1996). Early ventral prostate epithelial development (before PND 19) occurs under the inductive influence of the mesenchyme and involves mesenchymal androgen receptors, which when activated, elicit production of paracrine growth factors such as keratinocyte growth factor, which binds to and activates receptors on epithelial cells (Thomson et al., 1997). While epithelial androgen receptors are not required for prostatic development, they are absolutely necessary for secretory function (Donjacour and Cunha, 1993). In addition to androgen-receptor expression, ventral prostate secretory activity requires the formation of luminal epithelial cells from cellular precursors, which presumably are basal cells that are induced along a differentiation pathway by the presence of circulating androgens (Bonkhoff and Remberger, 1996; Robinson et al., 1998).
Previous results indicated that in utero and lactational TCDD exposure inhibits epithelial-bud formation, decreases the number of luminal epithelial cells, decreases luminal epithelial-cell androgen receptor-protein expression, and decreases androgen-responsive mRNA expression in the rat ventral prostate (Roman and Peterson, 1998; Roman et al., 1998). Therefore, the present study was undertaken in order to determine whether these effects might be contributed to by decreased testosterone production by the fetal and early neonatal testis, or decreased postnatal DHT synthesis, or altered androgen metabolism within the ventral prostate. Furthermore, prostatic binding protein subunit C3 (C3) mRNA and protein expression were assayed as indicators of the in vitro response of ventral prostates from TCDD and vehicle-exposed male offspring to testosterone, 3-Diol, and DHT. Luminal epithelial cell differentiation, as judged by the immunohistochemical localization of androgen receptors, cytokeratin 18, and C3 were also assayed in these cultures. The results of the current study indicate that the TCDD-induced decrease in ventral prostate growth cannot be explained by decreased perinatal androgen production in the fetal testis, or by a decreased conversion of the precursor androgens testosterone and 3-Diol to the active androgen DHT within the PND 14 to PND 32 ventral prostate. However, TCDD-induced decreases in androgen responsiveness, androgen receptor expression, and luminal cell development suggest an interference with prostate differentiation that may be caused by TCDD acting directly on the developing prostate to alter gene expression and/or growth factor activity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Animals and treatments.
Pregnant Holtzman rats (Harlan Sprague-Dawley, Inc., Madison, WI) arrived at our facility on GD 12 or 13 (GD 0 = sperm positive) and were housed individually in clear plastic cages (41 x 20 x 20 cm) containing aspen chips for bedding. Rats were maintained in a controlled environment with temperature at 22 ± 1°C, humidity at 35 ± 5%, and 12-h photoperiods (lights on at 0600 h, off at 1800 h). Animals were provided feed (Rat Chow 5012, Purina Mills, St. Louis, MO) and tap water ad libitum. Dams were administered a single oral dose of TCDD (1.0 µg/kg; 98% purity, Cambridge Isotope Laboratories, Woburn, MA) or an equal volume of vehicle (2 ml/kg, 95% corn oil/5% acetone) during the morning of GD 15; the fetal rat testis is just beginning to synthesize androgens at this time (Feldman and Bloch, 1978
; Warren et al., 1975, 1984). Testicular androgen content, human chorionic gonadotropin (hCG)-stimulated testosterone synthesis, and plasma testosterone concentration were evaluated at GDs 18 and 20 and 2 h after birth. For prenatal androgen measurements, dams were sacrificed on GD 18 or 20 by light CO2 overdose followed by decapitation. Uteri were removed and placed on ice, and fetuses were excised sequentially and held on ice until sacrifice. Fetuses and 2-h-old neonates were initially sexed by visual inspection of anogenital distance (AGD; distance between the anus and the genital tubercle, which is longer in males than in females), and sex determination was confirmed by gonadal inspection using a dissecting microscope. Implantation sites were counted after staining of the uterus with 10% (v/v) ammonium sulfide, and percent survival was calculated as number of live offspring at time of sacrifice/number of implantation sites x 100. This study was conducted in 3 consecutive blocks involving 15 to 24 dams each. In each block, dams were dosed with either TCDD or vehicle, and all litters were sacrificed on either GD 18, or 20, or at 2 h after birth. Overall n values were 7–13 litters per treatment per time point.
For analysis of postnatal ventral prostatic 5-R and 3-HSD activities, dams were allowed to deliver and all litters were culled to 10 pups on PND 1 in order to normalize lactational TCDD exposure across litters. Male offspring were sacrificed on PNDs 14, 21, and 32, while the remaining offspring were weaned at PND 21. After weaning, all males from the same litter were housed in 1 or 2 cages (maximum 4 rats per cage). This study consisted of 4 blocks, each involving 2–3 litters per treatment. Overall n values were 3–5 litters per treatment per time point.
For analysis of postnatal ventral-prostate androgen metabolism and androgen responsiveness in organ culture, dams were allowed to deliver and litters were culled to 10 pups in order to normalize lactational TCDD exposure across litters. Male offspring were sacrificed on PNDs 11 and 18. This study consisted of 1 block involving 20 litters. Overall n values were 3 litters per treatment per time point.
Necropsy and in vitro testis incubations.
Prior to sacrifice by decapitation, male offspring were weighed and, at the 2-h time point, AGD and crown-rump length were measured using a hand-held micrometer. Trunk blood was collected in heparinized capillary tubes, pooled within each litter, centrifuged for 10 min at 12,000 x g at 4°C, and plasma was stored in polypropylene tubes at –80°C. Testes were excised, weighed, and stored in a 24-well culture plate (Corning, Corning, NY) in 0.5 ml ice-cold 0.9% saline, until all offspring from a single litter had been processed. Subsequently, approximately one testis per litter was bisected and frozen immediately in DMEM/F-12 (medium, Sigma, St. Louis, MO) for determination of intratesticular testosterone content. All remaining testes were transferred to a fresh 24-well culture plate containing 25 µl medium per well and bisected with fine dissecting scissors. Twenty-five µl of the appropriate dilution of hCG (Sigma, St. Louis, MO) was then added to each well, followed by 450 µl medium. Final hCG concentrations were 0, 1, 2, 4, 8, 16, 32, 64, and 128 mIU/ml. Testes were incubated on an orbital shaker (150 rpm) for 3 h at 37°C in 5% CO2/95% O2. At the end of the incubation, testes and medium were collected and frozen at –20°C until assay.
Testosterone radioimmunoassay.
Testicular parenchyma and incubation medium were homogenized for 30 s (Tissumizer; Tekmar, Cincinnati, OH). After homogenization, the Tissumizer probe was rinsed with 1 ml fresh medium, and the combined sample and rinse were disrupted for 10 s, using an immersion sonicator (Laboratory Supplies Co. Inc, Hicksville, NY). Testis homogenates (100 µl) and plasma samples (300 µl) were extracted twice with 10 volumes of anhydrous ethyl ether and organic phases were combined and dried in a Speed Vac concentrator (Savant Instruments, Farmingdale, NY). Samples were reconstituted in ethanol. Recovery of [3H]-testosterone from representative spiked samples ranged from 80–90% for testis homogenates and plasma. Standards and samples were assayed in triplicate for testosterone by radioimmunoassay (Robinson et al., 1975). Unlabelled testosterone was purchased from Steraloids (Wilton, NH); [3H]-testosterone was purchased from Amersham (Arlington Heights, IL), and testosterone antiserum was purchased from Endocrine Sciences (Calabasa, CA). The range of quantitation for the RIA was 10–1000 pg/tube.
5-reductase and 3-hydroxysteroid dehydrogenase activities.
Ventral prostates removed from vehicle- and TCDD-exposed offspring on postnatal days (PNDs) 14, 21, and 32 were immediately frozen in liquid nitrogen, and were subsequently stored at –80°C until use. PNDs 21 and 32 ventral prostates were thawed on ice and bisected so that one lobe was used to measure the type-1 5-R isozyme and 3-HSD activities. The remaining ventral prostate lobe was used to measure the activity of the type-2 5-R isozyme. Ventral prostates were homogenized in a buffer system that contained 116 mM NaCl, 4.5 mM KCl, 2.5 mM CaCl2, 1.3 mM MgCl2, 10 mM sodium phosphate salts, 10 mM sodium citrate salts, 2 mM EDTA, 4 mM dithiothreitol (DTT), and 5% (v:v) glycerol. The prostate lobe intended for the analysis of type-1, 5-R and 3-HSD activities was homogenized in 2 ml of the buffer at pH 6.9, whereas the prostate lobe intended for the analysis of type 2 5-R activity was homogenized in 1 ml of the buffer at pH 5.5. On PND 14 type 1 5-R activity was not detectable and separate ventral prostates had to be used for the analysis of type-2 5-R and 3-HSD activities.
To assay the type-1 and type-2 5-R activities, run buffers were prepared that contained 116 mM NaCl, 4.5 mM KCl, 2.5 mM CaCl2, 1.3 mM MgCl2, 10 mM sodium phosphate salts, and 10 mM sodium citrate salts. Type-1 5-R activity was assayed in tubes that contained 50 µl of run buffer and 50 µl of homogenate in the presence of 0.5 µM [3H]-testosterone and 2 mM NADPH at pH 6.9. Type-2 5-R activity was assayed in similar tubes that contained 0.05 µM [3H]-testosterone at pH 5.5. The testosterone concentration was lower than that used for the type-1 5-R assay so that the differing Michaelis constants between the 2 isozymes could be used to increase assay selectivity (Span et al., 1995). 3-HSD activity was determined in 40 mM trizma buffer (pH 8.6) in the presence of 0.5 µM 3H-3-Diol, 2 mM NAD and 50 µl of homogenate in 100 µl of total volume. In each assay (5-R and 3-HSD) the reactions, in duplicate tubes, were stopped at 10, 20, 30, 40, 50, and 60 min by the addition of 10 µl of 3 N NaOH. Zero-time tubes were formed by adding the NaOH to prevent reaction before adding the homogenate. Very little metabolism of the radioactive substrates occurred in the zero-time tubes. All samples were stored at –20°C until they were extracted.
All samples were diluted with 0.9% saline to 1 ml and were extracted once with 2 ml of chloroform:methanol (2:1) and once with chloroform. Organic phases from each sample were combined and evaporated at 45°C under a gentle steam of nitrogen. The aqueous phase of each sample was prepared for liquid scintillation counting to determine the radioactivity. The dried organic phases were redissolved in 40 µl of methanol that contained non-radiolabeled testosterone, dihydrotestosterone, 3-, 3ß-Diols, and either androstenedione (5-R assays) or androstanedione (3-HSD assay). The reconstituted samples (20 µl) were applied to the preabsorbant region of a 19-channel silica gel TLC plate (J. T. Baker, Phillipsburg, NJ). TLC plates were developed with toluene:95% ethanol (94:6) and the bands for the carrier steroids were visualized after staining with anisaldehyde spray reagent (Sigma, St. Louis, MO). The zones for each steroid, and for the determination of background radioactivity were scraped into scintillation vials, such that every lane was counted from the preabsorbant area to a point just beyond the farthest migrating steroid. The results from each vial were then considered in the analysis either as a known steroid, or as background.
Analysis included a subtraction of the background counts and where necessary, the inclusion of polar metabolites recovered in the aqueous phase. In the 5-R assays the extraction efficiency was 
90% and constant for all time points. However, in the 3-HSD assay, there were increasingly greater amounts of radioactivity recovered in the aqueous phase as assay time increased. These were considered to be polar metabolite(s) (mostly of the substrate, 3-Diol) and they were quantified and included in the analysis. Initial reaction velocities for both 5-R isozymes and for 3-HSD were determined by computer-assisted numerical and graphical analysis using the slope of the tangent line that passed through the origin in the applicable time-course curve for DHT. These initial velocities were then used to determine the specific activity of each 5-R isozyme and 3-HSD relative to the total protein concentration in each homogenate. Protein concentrations were determined by a modification of the Lowry protein assay (Bensadoun and Weinstein, 1976).
Ventral prostate organ cultures.
Ventral prostates were obtained from vehicle- and TCDD-exposed offspring on PNDs 11 and 18. The ventral prostate from each rat was immediately placed into DMEM (GIBCO, Grand Island, NY) with 5% of charcoal-dextran-stripped fetal bovine serum (Summit, Ft. Collins, UT) and cultured at 37°C in an atmosphere of 95% air and 5% carbon dioxide. On PND 11 whole prostates were cultured, whereas one-half of the prostate from each rat was sufficient on PND 18. After 24 h the medium was removed and new DMEM that contained 5% charcoal-dextran-stripped fetal bovine serum supplemented with graded concentrations of androgenic steroids was added. Each culture was then exposed to 10–9 M, 10–8 M, or 10–7 M concentrations of [3H]-testosterone, [3H]-DHT, or [3H]-3-Diol for 48 h. The harvesting dates for the cultures initiated on PND 11 and PND 18 then coincided with PND 14 and PND 21, respectively. The tissue culture medium was stored at –20°C prior to solvent extraction and TLC analysis of steroid content (as above). Total RNA was isolated from each cultured ventral prostate and separated by electrophoresis on a 1.4% agarose gel that contained formaldehyde (Krumlauf, 1996). The RNA was transferred from the gel to a nylon membrane, which was probed for prostatic-binding protein C3 subunit and cyclophillin by using [32P]-labeled DNA probes. Triplicate determinations were made for each concentration of each steroid, and band intensities were evaluated by using commercially available phosphoimage analysis equipment and software (Molecular Dynamics, Sunnyvale, CA). The C3 subunit band intensities from each sample were normalized to the appropriate cyclophillin band intensity. To evaluate the effects of in vivo TCDD exposure on the expression of the C3 subunit protein, a separate set of ventral prostates was cultured for 24 h in the absence of steroids, and then for 48 h in the presence of 10–8 M 3-Diol. This steroid was chosen because it is the predominant circulating androgen in rat blood on PNDs 11–21 (Corpechot et al., 1981).
Immunohistochemistry.
Cultured ventral prostates were fixed in Bouin's solution overnight and embedded in paraffin. Tissue sections (5 µm) were evaluated for androgen receptor, prostatic binding protein subunit C3 (C3), or cytokeratin 18 (CK 18) expression by an immunohistochemical avidin-biotin-coupling (ABC) technique (Loeffler and Peterson, 1999). Briefly, sections were deparaffinized and rehydrated through a graded series of alcohol concentrations to PBS. Androgen receptor and C3 antigenicity was recovered by boiling sections in 5% urea (ICN, Aurora, OH) for 30 min, while CK 18 antigenicity was restored by digesting sections with 0.25% trypsin (Difco Labs, Detroit, MI) at 37°C for 30 min. Nonspecific binding was blocked by incubation for 20 min with 2% heat-inactivated goat serum (Vector Labs, Burlingame, CA) in PBSG (0.05 M phosphate buffered saline and 0.1% gelatin, pH 7.3). The primary antibody for the androgen receptor was a rabbit polyclonal antibody directed against a human androgen receptor fusion protein (NH27, kindly provided by Dr. A. Mizokami, University of Occupational and Environmental Health, Osaka, Japan). A rabbit polyclonal antibody directed against the C3 subunit was kindly provided by Dr. T. C. Shao, VA Medical Center, Houston, TX). A commercially available mouse monoclonal antibody directed against CK 18 (Sigma, St. Louis, MO) was used to assess the functional differentiation of ventral prostate luminal epithelial cells. Primary antibodies were diluted in the PBSG-goat serum-blocking solution as follows: anti-androgen receptor 1:1,200, anti-C3 1:20,000, and anti-CK 18 1:200. These were then incubated over the sections in a humid chamber at 4°C overnight. Immunochemical-positive cells were visualized by using the Elite Vectastain ABC kit according to the manufacturer's protocol (Vector Labs, Burlingame, CA). Sections were counterstained for 1 min in Mayer's hematoxylin, rinsed in running tap water until clear, dehydrated through a graded series of alcohol concentrations, cleared with xylene, and mounted with Permount. Negative controls, in which primary antibody was replaced with normal rabbit/mouse IgG, showed no nonspecific staining.
Statistical analysis.
Statistical analysis of data was performed using either the SAS statistical package (SAS Institute, Cary, NC) or Statistica/w 5.0 (Statsoft, Tulsa OK). Data presented are means ± SE; when multiple animals from a single litter were analyzed, data presented are litter means ± SE, so that each n represents an independent litter. Dam body weights were analyzed by repeated-measures analysis of variance (ANOVA), and survival was analyzed by Chi-square analysis. Fetal offspring body and testis weights, plasma testosterone concentrations, and intratesticular testosterone content were analyzed by Student's (homogeneous variance) or Cochran's (heterogeneous variance) t-test, whereas hCG-stimulated androgen production data were analyzed by 2-way ANOVA. Normalized prostatic binding protein-C3 subunit expression data obtained from ventral prostate organ cultures that were stimulated by 3 different androgens were analyzed by ANOVA. Steroid concentration was considered nested for each treatment within the stimulating steroid category. Levene's test was used to assess homogeneity of variance; when necessary, square root transformations were performed. Pup-body-weight data, and prostatic-binding protein-C3 data for PND 14 did not pass Levene's test, and were therefore analyzed by the Kruskal-Wallis procedure. All other statistical analyses were done using Student's t-test. Significance was set at p < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Overt Toxicity
Treatment of dams with 1.0 µg TCDD/kg on GD 15 did not affect maternal body-weight gain or produce any visible signs of maternal toxicity during gestation. When assessed on GD 18, offspring survival was not significantly decreased by TCDD exposure. However, when assessed on GD 20 and at 2 h after birth, offspring survival was significantly decreased by 10% and 15%, respectively (p < 0.05, results not shown). Intestinal hemorrhaging was noted in 9% and 14% of live offspring on GD 20 and in 2-h-old offspring, respectively, and some stillborn offspring of TCDD-treated dams exhibited moderate to severe subcutaneous edema. TCDD exposure had no effect on body weights, absolute testis weights, or relative testis weights on GD 18 or at 2 h after birth (Table 1
). A significant 14% decrease in absolute testis weight was noted on GD 20, although this decrease was no longer significant when testis weight was corrected for body weight (Table 1). In the postnatal study, there was no effect of TCDD exposure on prenatal mortality, as the ratio of births to implantation sites was 0.87 and 0.86 in the vehicle- and TCDD-exposed groups, respectively. On PND 1, body weights of male offspring exposed to TCDD were significantly decreased by 9.4% when compared to vehicle-exposed male offspring. This difference was maintained during the postnatal period as male offspring in TCDD-exposed litters weighed about 12% less than their vehicle-exposed counterparts on PNDs 7, 14, and 21 (range 10.6–12.9%). Overall, these results indicate that exposure to a single maternal dose of 1 µg TCDD/kg administered on GD 15 caused overt prenatal and postnatal toxicity to the offspring.
|
Anogenital Distance
Neonatal AGD, or the distance from the anterior edge of the anus to the base of the genital tubercle, is androgen-dependent (Neumann et al., 1970
) and is substantially longer in males than females. GD 15 TCDD exposure did not decrease AGD in male rats at 2 h after birth. In addition, there was no effect of TCDD exposure on crown-rump length (CRL) or relative AGD (AGD/CRL) at this time point.
Testosterone Measurements
Plasma testosterone concentrations in control animals decreased by approximately 50% between GDs 18 and 20, and increased 5-fold between GD 20 and at 2 h after birth (Table 2). This later time point corresponds to the time of the expected postnatal testosterone surge. The same pattern of early postnatal changes in plasma testosterone concentration was observed in TCDD-exposed animals. In contrast to results reported previously (Mably et al., 1992a), GD 15 exposure to 1.0 µg TCDD/kg did not decrease plasma testosterone concentrations at any of the time points examined; in fact, testosterone concentrations in TCDD-exposed animals ranged from 98 to 118% of control (Table 2). Similarly, no significant decreases in intratesticular testosterone content were detected, as values from TCDD-exposed animals ranged from 87–124% of control on a per testis basis (Table 2). Even though testes from TCDD-exposed offspring were smaller than those from control offspring on GD 20 (Table 1
), there was no effect of in utero and lactational TCDD exposure at any time when intratesticular testosterone content was expressed on a testis-weight basis (Table 2).
|
In vitro testosterone production by bisected testes from control and TCDD-exposed male offspring was stimulated by hCG on GD 18, GD 20, and at 2 h after birth in a dose-dependent manner (Fig. 1
). However, in contrast to effects of high-dose TCDD exposure in the adult rat on gonadotropin-stimulated testicular testosterone production (Kleeman et al., 1990), the ability of the perinatal testis to produce testosterone in response to graded doses of hCG was not impaired by low-dose, in utero TCDD exposure.
|
Ventral Prostate Weights and 5
-Reductase and 3-Hydroxysteroid Sehydrogenase ActivitiesIn utero and lactational TCDD exposure caused substantial inhibition of ventral prostate growth (Fig. 2A
). In vehicle-exposed offspring, ventral prostate wet weight increased from 7 mg on PND 14 to 89 mg on PND 32. However, ventral prostate wet weight in TCDD-exposed male offspring was only 51%, 40%, and 38% of that in vehicle-exposed offspring on PNDs 14, 21, and 32, respectively. As DHT is required for prostate growth, the ability of the ventral prostate to biosynthesize DHT from the precursor substrates, testosterone and 3-Diol, was investigated. These metabolic transformations require the activity of 5-R and 3-HSD (Fig. 3).
|
|
Perinatal exposure to TCDD increased 5
-R type-2-specific activity (relative to tissue protein) 2.5-fold on PND 14, 3.6-fold on PND 21, and 5.1-fold on PND 32 (increase not statistically significant on PND 32, Fig. 2B). Therefore, the reductions in prostate size and protein content were not accompanied by a decrease in this specific activity during the postnatal period. The catabolic 5-R type-1 isozyme was not readily detectable on PNDs 14 and 21. This result may be expected because 3-Diol is the most abundant circulating androgen at these times (Corpechot et al., 1981), and there is little circulating testosterone for the prostate to eliminate. On PND 32, however, type-1 5-R activity was detectable in ventral prostate homogenates from 2 of 3 vehicle-exposed rats and 2 of 2 TCDD-exposed rats. Similar to the effects of in utero and lactational TCDD exposure on the activity of type-2 5-R, the activity of the type-1 isozyme was increased rather than decreased by TCDD exposure. On PND 32 the 5-R type-1 activity in each of 2 TCDD-exposed rats was more than 3 times greater than that in each of the 2 vehicle-exposed rats that had any detectable activity (Vehicle: 0.26 ± 0.07 nmoles/min/g protein, n = 2; TCDD: 0.97 ± 0.06 nmoles/min/g protein, n = 2). Therefore, the TCDD-induced decrease in ventral-prostate weight does not result from a decrease in the activity of either type-1 or type-2 5-R during the postnatal period.
On PNDs 14, 21, and 32, the specific activity of 3-HSD in all ventral prostates examined, regardless of TCDD-exposure, was at least 20-fold greater than that of the 5-R type-2 isozyme (compare ordinate on Figs. 2B and 2C). In addition, the ventral prostates removed from TCDD-exposed rats on PND 14 exhibited a 60% reduction in the amount of 3-HSD specific activity, when compared to that in vehicle-exposed rats (Fig. 2C). The magnitude of the TCDD exposure-induced decrease in 3-HSD-specific enzyme activity, which was not statistically significant on PND 14, was reduced when the prostates were obtained on PND 21, and there was no difference between vehicle and TCDD-exposed rats in the specific activity of 3-HSD on PND 32.
Androgen Metabolism
Ventral prostates obtained from vehicle- and TCDD-exposed rats on PNDs 11 and 18 were not exposed to any steroid for the 24 h in organ culture, and subsequently were exposed to [3H]-testosterone, [3H]-3-Diol, or [3H]-DHT for 48 h. At the end of the androgen exposure period, the radiolabeled steroid products were extracted from the culture medium on PNDs 14 and 21, respectively, and separated by TLC. While multiple radiolabeled products were recovered in the medium, the results indicate that the activities of 5-R and 3-HSD in the organ-cultured ventral prostates were apparently not inhibited by TCDD exposure. After 48 h of exposure to [3H]-testosterone or [3H]-3-Diol, similar amounts of DHT were recovered from the culture medium of ventral prostates from vehicle- and TCDD-exposed offspring on PNDs 14 and 21 (Figs. 4 and 5). In addition, similar amounts of DHT were recovered when vehicle- and TCDD-exposed ventral prostates were exposed to [3H]-DHT in organ culture. This suggests that TCDD exposure did not substantially enhance the ventral prostate's ability to metabolize DHT in organ culture. Indeed after 48 h, most of the [3H]-DHT was recovered from the culture medium in both treatment groups as the unmetabolized substrate on PND 14 (Fig. 4). Similar amounts of DHT remained, and equal amounts of polar metabolites were formed by ventral prostates of both treatment groups after 48 h of exposure to [3H]-DHT that ended on PND 21 (Fig. 5). Thus, there was never as much as a 2-fold difference between the vehicle- and TCDD-exposed ventral prostate cultures in the amount of DHT recovered, with any substrate at either time point. Therefore, these results do not suggest that there would be any substantial decrease in the amount of DHT available to vehicle- and TCDD-exposed prostates in the presence of similar concentrations of the DHT precursors. Rather, these results complement those of the direct enzyme assays for 5-R and 3-HSD activity. Taken together, all results support the conclusion that the effect of in utero and lactational TCDD exposure on ventral prostate development in the Holtzman rat is not associated with any lasting reduction in its ability to form DHT during the postnatal period, although there was a small decrease in 3-HSD activity on PND 14 that was not statistically significant.
|
|
Responsiveness of the Ventral Prostate to Androgenic Stimulation
Ventral prostates from rats exposed to TCDD in utero and via lactation display histological alterations similar to those which can be caused by androgen deprivation (Hayward et al., 1996
; Roman et al., 1998). Yet, maternal TCDD treatment does not impair testicular androgen production perinatally or at later times (Roman et al., 1995), nor does it significantly affect androgen metabolism within the ventral prostate postnatally. This suggests that TCDD impairs the prostate's ability to proliferate and differentiate in response to normal androgen stimulation. To test the hypothesis that in utero and lactational TCDD exposure impairs the prostate's ability to respond to normal androgen stimulation, pregnant rats were given 1 µg TCDD/kg or vehicle on GD 15, and ventral prostates were removed on PNDs 11 and 18. These organs were cultured for 1 day in 5% stripped fetal bovine serum, transferred to fresh medium containing vehicle or graded concentrations of testosterone, DHT, or 3-Diol for 2 days, and harvested at the equivalent of PND 14 and 21. On both PNDs evaluated, exposure to each androgen increased the abundance of C3 mRNA in explant cultures from control rats (Fig. 6). In contrast, there was almost no androgen-dependent stimulation of C3 mRNA synthesis in ventral prostates removed from TCDD-exposed male offspring on PND 11 and cultured until PND 14 (Fig. 6A). When the ventral prostates were removed from TCDD-exposed offspring on PND 18 and cultured until PND 21, the amount of androgen-dependent C3 mRNA synthesis was significantly less than that produced by similar organ cultures from vehicle-exposed offspring (Fig. 6B). These results indicate that exposure to TCDD in utero and via lactation caused a decrease in ventral prostate androgen responsiveness. Significantly, the decrease in androgen responsiveness was evident on both postnatal days even in cultures exposed to DHT, and a similar degree of recovery from the TCDD-induced inhibition was evident for all androgens evaluated on PND 21. These findings reinforce the hypothesis that TCDD-induced decreases in ventral prostate weight are not the result of decreased DHT synthesis between PNDs 14 and 21.
|
Immunohistochemistry of the Cultured Ventral Prostates for Androgen Receptor
Ventral prostates obtained on PNDs 11 and 18 were deprived of androgens for 24 h before being exposed to 10–8 M 3
-Diol until the calendar days that corresponded to PNDs 14 and 21, respectively. It was apparent that development of these cultured prostates was not as advanced as are ventral prostates obtained from 14- and 21-day-old rats. However, the cultured prostates displayed morphologically normal epithelium that was organized into ducts surrounded by stroma (Figs. 7–9). On "PND 14" androgen receptor protein was detected in the stroma and epithelium of ventral prostates from vehicle-exposed rats that were organ-cultured in the presence of 10–8 M 3-Diol (Fig. 7A). While the staining was mostly stromal, some ductal lumens were lined by androgen receptor-stained epithelial cells. In similarly organ-cultured ventral prostates from TCDD-exposed rats evaluated on "PND 14", androgen-receptor staining was present in the stroma, but very few of the epithelial cells contained androgen receptors (Fig. 7B). On "PND 21" nearly all epithelial cells stained intensely for the androgen receptor in organ-cultured ventral prostates from vehicle-exposed rats (Fig. 7C). In contrast, androgen-receptor staining in the ventral prostates that had been exposed to TCDD in vivo occurred predominantly in the stroma. Only weak androgen-receptor staining was present in periluminal epithelial cells, and many of these cells appeared to be completely unstained (Fig. 7D).
|
|
|
Luminal Epithelial Cell Differentiation in Cultured Ventral Prostates
In the cultured ventral prostates described above, immunohistochemical staining for cytokeratin-18 expression was used as a marker for luminal epithelial cells, a highly differentiated cell type responsible for androgen-dependent protein secretion (Hayward et al., 1996
; Hsieh et al., 1992). On "PND 14", very little cytokeratin-18 expression was detected in ventral prostates from either vehicle- or TCDD-exposed offspring that were cultured in the presence of 10–8 M 3-Diol (Figs. 8A and 8B). This result may, in part, reflect a lack of development due to 24 h of androgen deprivation in organ culture before beginning androgen exposure. On "PND 21", specific staining for cytokeratin 18 in cultured prostates from vehicle-exposed rats was exclusively confined to the innermost layer of epithelial cells, the luminal epithelial cells (Fig. 8C). In specimens from TCDD-exposed prostates, on the other hand, almost no specific staining for cytokeratin 18 was observed on "PND 21" (Fig. 8D). These results suggest a blockage of luminal epithelial cell development. Such a blockage has previously been reported in ventral prostates of TCDD-exposed neonates prepared for immunohistochemical examination immediately after removal from the animal, without organ culture (Roman et al., 1998). Therefore, despite any delay in development due to the brief period of organ culture in the absence of androgens, these results do reflect those that occur following exposure of the ventral prostate to the normal circulating levels of androgens in vivo.
Immunohistochemistry of the Cultured Vventral Prostates for Prostatic Binding Protein-C3 Subunit
Sections from vehicle-exposed prostates that had been cultured in the presence of 3-Diol were evaluated for immunohistochemical C3 expression. Less than one-half of the ducts from vehicle-exposed offspring examined on "PND 14" displayed the C3 subunit (Fig. 9A). Staining was largely confined to epithelial cells, but in some cases the ductal lumens were also stained. In contrast to the vehicle-exposed ventral prostates, however, exposure of TCDD-treated ventral prostates to 10–8 M 3-Diol stimulated little or no C3 staining in the epithelium or ductal lumens, even though numerous unstained epithelial ducts were present (Fig. 9B). The periductal smooth muscle sheaths appeared to be thicker in the TCDD-exposed ventral prostates than in the vehicle-exposed prostates already described. This last observation is consistent with previous in vivo results (Roman et al., 1998).
Vehicle-exposed ventral prostates obtained on PND 18 and cultured with 3-Diol to "PND 21" exhibited increased ductal development and more extensive staining for the C3 subunit compared to the corresponding results on "PND 14." Exposure to 10–8 M 3-Diol caused a majority of ducts to exhibit epithelial C3 (Fig. 9C). Unlike the results from "PND 14", however, most ductal lumens were also stained, indicating that most ducts were secreting prostatic binding protein. Results from sections of similarly cultured ventral prostates from TCDD-exposed rats corresponded to the previous analysis of cytokeratin-18 expression in that C3 staining was found in relatively few ducts, and such staining was more diffuse (Fig. 9D). In most ducts, a majority of the periluminal epithelial cells did not stain for either cytokeratin-18 or -C3, and prostatic binding protein-C3 secretion into the lumen was not nearly as prevalent as it was in the vehicle-exposed organs. In addition, the smooth muscle sheaths of ducts in sections from TCDD-exposed ventral prostates appeared to be thicker than those sheaths from vehicle-exposed prostates. These results compliment the Northern blot analyses, and indicate that TCDD-exposed prostatic epithelium synthesized less C3 protein in response to 3-Diol stimulation in organ culture than did prostatic epithelium from vehicle-exposed control rats.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In Utero TCDD Exposure Does Not Interfere with Perinatal Androgen Production
Interference with testicular androgen production during the period from GD 18 to 2 h after birth is not the underlying cause of the TCDD-induced decrease in ventral prostate weight. Exposure of male Holtzman rats to TCDD in utero and via lactation did not cause decreases in plasma testosterone concentration, intratesticular testosterone content or in vitro hCG-stimulated testicular testosterone production over this period of evaluation. These findings are in contrast to those presented in an earlier study with this rat strain (Mably et al., 1992a
), in which significant decreases in plasma testosterone concentrations were observed on GDs 18 and 20 (77–79% of control) and at 2 h after birth (40% of control). However, the literature contains support for the current data. Gray et al. (1995) reported that male reproductive system development in Long Evans rats was adversely affected in the absence of a decreased plasma testosterone concentration or decreased in vitro, gonadotropin-stimulated testicular testosterone production on PND 0. In addition, the present results, as well as several other studies (Bjerke and Peterson, 1994; Bjerke et al., 1994b; Gray et al., 1995; Roman et al., 1995), in contrast to the initial report (Mably et al., 1992a), indicate no significant TCDD-induced decrease in relative anogenital distance, a sensitive bioassay of perinatal androgen status (Neumann et al., 1970). Finally, volume of the medial preoptic nucleus in the brain, another sensitive index of perinatal androgen status (Ito et al., 1986), was not altered by in utero and lactational TCDD exposure (Bjerke et al., 1994a). Taken together with the literature cited, the most recent results indicate that effects of in utero and lactational TCDD exposure on the male rat reproductive system are not accompanied by an impaired testicular androgen synthesis capacity during the perinatal period. It is important to note that most previously described effects of TCDD exposure on the developing prostate occurred in response to in utero TCDD exposure alone (Bjerke and Peterson, 1994). Unfortunately, technical problems prevent the evaluation of androgen status between GDs 15 and 17, except perhaps by measurements of gene expression. This is significant because the critical window for androgen exposure may occur during this time period (Clark et al., 1993).
The lack of effect of TCDD on perinatal androgen production was somewhat unanticipated considering that TCDD exposure of adult male rats has been shown to decrease plasma testosterone concentrations by interfering with gonadotropin-stimulated testicular testosterone production (Kleeman et al., 1990; Moore et al., 1985). Differences between responses of the perinatal and adult testes to TCDD exposure may be due to a dosage effect. The ED50 for decreased plasma androgen concentrations in the adult rat is 15 µg TCDD/kg, whereas the dose of TCDD used in the present study was 1.0 µg TCDD/kg maternal body weight. In Long Evans rats, about 0.15% of a maternal dose of 1.0 µg TCDD/kg administered on GD 15 is present in the fetal compartment on GD 16 (Hurst et al., 1996). The approximately 300-g pregnant Holtzman dams in the present study were dosed on average with 0.3 µg TCDD. If Holtzman rats are similar to Long Evans rats, about 0.00045 µg of TCDD would have been distributed between each of about 10 fetuses. Since the body weights of GD 16 Holtzman fetuses are approximately 0.5 g, a rough estimate of the actual dose to each GD 16 fetus in this study is 0.09 µg/kg. Fetal body burdens later in gestation would most likely be lower than at GD 16, considering the rate of growth of the fetus. This estimated GD 16 fetal body burden is considerably lower than the NOAEL for decreased plasma androgen concentrations in adulthood (6.25 µg/kg, Moore et al. 1985).
Effects of in Utero and Lactational TCDD Exposure on Ventral Prostate Androgen Metabolizing Enzymes
The mechanism by which in utero and lactational TCDD exposure impairs development of the ventral prostate does not involve decreased circulating testosterone concentrations during the period just prior to, or immediately after, birth. However, testosterone is not the primary proximal mediator of ventral prostate development. To be effective in the ventral prostate mesenchyme, circulating testosterone must be converted in situ to DHT (George and Peterson, 1988; Wilson and Lasnitzki, 1971). Prior to birth, testosterone is the predominant androgen in the fetal circulation (Corpechot et al., 1981), and it is converted to DHT by 2 isozymes of 5-R within the developing prostate (Berman et al., 1995).
In the fetal rat, the type-2 isozyme, which is expressed exclusively in mesenchymal cells, is probably the most important of the 2 isozymes for androgenic signaling (Berman et al., 1995). In the present study, the anticipated response was a decrease in 5-R activity following in utero and lactational exposure to TCDD. However, no TCDD exposure-induced decrease in the activity of either 5-R isozyme was detected in the ventral prostate postnatally and similar results have been obtained for the dorsolateral prostate (Theobald et al., in press). In the adult rat, but not the fetal rat, this isozyme is subjected to feed-forward control, a mechanism by which type-2 5-R activity can increase in response to DHT (Berman et al., 1995). Therefore, it does not seem likely that the increases in this enzyme activity, which occurred in TCDD-exposed male offspring on PNDs 14, 21, and 32, could reflect decreased ventral prostate DHT levels, even if the amount of the type-2 isozyme expressed were to be considered androgen-responsive at these early postnatal time points.
Shortly after birth, 3-Diol, rather than testosterone, becomes the predominant circulating androgen, and remains so almost until the rat reaches puberty (Corpechot et al., 1981). Thus, in the early postnatal period, most of the prostatic DHT content is derived from 3-Diol by the action of 3-HSD. Since the activity of this enzyme, unlike that of 5-R is readily reversible, it can act as a molecular switch which controls the amount of DHT available for receptor signaling (Penning et al., 1996). In the early postnatal period, up to PND 32, ventral prostate homogenates contained substantially more 3-HSD activity than 5-R activity, which indicates the importance of this enzyme. In addition, the ability of neonatal ventral prostate homogenates to form polar metabolites over a period of 60 min when [3H]-3-Diol is used as a substrate, but not during a similar 1-h incubation when [3H]-testosterone is used as a substrate, indicates that the ventral prostate is geared up for the elimination of 3-Diol, but not testosterone. This observation again suggests the relative importance of the role that 3-Diol plays within the neonatal ventral prostate, compared to that of testosterone.
On PND 14, the earliest day examined for enzyme activities in ventral prostate homogenates, in utero and lactational TCDD exposure transiently reduced the conversion of 3-Diol to DHT to about 40% of its value in vehicle-exposed offspring. Since previous results demonstrated a 37% decrease in ventral prostate DHT concentration (ng/g tissue wet weight) on PND 21 in response to in utero and lactational TCDD exposure (Roman et al., 1995), it is possible that reductions in 3-HSD activity, or a delay in the onset of its expression might have functional significance. Thus, greater effects on 3-HSD activity than those observed on PND 14 could be present just after birth. However, the ventral prostates were quite small on PND 14 and the activity of 3-HSD was low and difficult to measure. Therefore, an accurate assay of this enzyme activity during the first week after birth was not considered to be technically feasible. By PND 32, however, 3-HSD activity (the present study) and ventral prostate DHT concentration (Roman et al., 1995) have returned to control values.
It is not clear whether the observed alterations in 5-R and 3-HSD activi