| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
Department of Biochemistry and Molecular Biology3
Department of Cell Biology,4 Georgetown University School of Medicine, Washington, District of Columbia 20057
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, we subsequently confirmed that Hsp90 was primarily expressed in gonocytes, and that it was increased following estrogen exposure. Hsp90 immunolocalization in fetal and prepubertal testes showed an increased expression in fetal gonocytes upon estrogen exposure, but no difference in the subsets of Hsp90-positive germ cells in prepubertal testes. These results demonstrate that prenatal estrogen exposure specifically affects Hsp90 expression in gonocytes. Considering the interaction of Hsp90 with several signaling molecules, changes in its expression levels may lead to subsequent changes in gonocyte development.
developmental biology, environment, estradiol receptor, gamete biology, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The fact that male ERß knockout mice offspring are fertile [10] suggests that ERß does not participate in testis development and functions. However, recent studies have revealed the existence of variant forms of ERß [11], including forms expressed specifically in normal fetal human germ cells [12] and human testicular germ cell tumors [13], and forms that demonstrated their ability to bind several estrogenic compounds, to heterodimerize with the classical ER and ERß, and to induce transcription [14]. The existence of new ERß forms in testis suggests that they may indeed support functions originally attributed to the classical ERs, and that the situation in testis regarding ERs functions might be more complex than originally believed.
The ER signaling pathway comprises cytosolic and nuclear regulatory complexes. The multimeric cytosolic complex is composed of two groups of proteins that associate sequentially: first an immature complex comprising the chaperone proteins heat shock protein 90 (Hsp90) associated with Hsp70, Hip, and Hop; and a mature complex in which Hsp90 is associated with the small cochaperone p23 and one of the immunophilins, peptidyl propyl isomerases such as FKBP51, FKBP52, or cyclophilin 40 (Cyp40) [15]. The mature complex of Hsp90, p23, and an immunophilin has been shown to stabilize and protect the ER from degradation and to maintain it in a high affinity conformation for estrogen binding [15, 16]. Hsp90 will dissociate from the complex upon estrogen binding to the receptor, whereas p23 will translocate with the estrogen-coupled dimerized ERs to the nucleus [15, 17]. There, ER will interact with a different set of regulatory molecules and exert its transcriptional activities [5, 15].
Heat shock proteins are involved in the response of organisms to environmental stress, but have also been implicated in developmental events [18]. Besides its role as a major component of the ER complex, Hsp90 also interacts with other nuclear receptors, and with molecules not generally associated with nuclear receptor pathways [19], such as G proteins [20], tyrosine kinases [21], and Raf [22]. Two Hsp90-related proteins have been identified in mouse testis: Hsp84, which is expressed mainly in somatic cells; and Hsp86, which is found predominantly in germ cells [23, 24]. The expression of Hsp90, considered as the inducible form of Hsp90 and equivalent to Hsp86, was recently reported in rat germ cells [25], whereas an Hsp90-related protein, Hsp105, was found to interact with p53 in rat testis [26]. Heat shock proteins have been shown to increase in testis following heat stress [27]. Moreover, the possibility that Hsp90 actively participates in testis development was clearly illustrated by a study carried out in Drosophila, in which a reduction in Hsp90 function following point mutations was shown to alter spermatogenesis [28]. However, the precise function of Hsp90 and its regulation in testis remains to be determined.
We recently reported that prenatal exposure to estrogenic compounds induces the overexpression of platelet-derived growth factor (PDGF) receptors in neonatal testis; more specifically, in gonocytes [29]. Indeed, ER has been implicated in cross talks with growth factor signaling pathways in both in vitro [30–32] and in vivo [33] experimental models. Estrogens were also shown to affect Hsp90 and Hsp70 expression in female reproductive organs [34, 35]. Considering that Hsp90 plays a critical role in stabilizing and allowing the activity of several molecules involved in proliferation [22] and survival [36] pathways, these observations suggested that Hsp90 expression might also be altered in response to estrogen exposure in testicular germ cells.
In the present study, we determined the identity of the ER-associated proteins expressed in neonatal gonocytes and further examined whether their expression was affected by prenatal exposure to estrogenic compounds. Our results demonstrated that neonatal gonocytes express Hsp90, p23, Cyp40, and the constitutive Hsc70, together with ERß. The inducible Hsp70 could not be detected in gonocytes but was found in whole testis and in Sertoli/myoid cell fractions. Prenatal estrogen exposure induced an increase in Hsp90 that was mostly localized in gonocytes. We found that Hsp90 increased in gonocytes in response to estrogen exposure in a manner similar to the total Hsp90, suggesting that Hsp90 is the isoform involved in the gonocyte response to estrogens. These results suggest that prenatal estrogen exposure specifically alters the expression of Hsp90 in testicular gonocytes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bisphenol A (4,4'-isopropylidenediphenol; BPA) was purchased from ICN Biomedicals, Inc. (Aurora, OH). Diethylstilbestrol (DES), genistein, 17ß-estradiol, and hematoxylin were purchased from Sigma (St. Louis, MO). Coumestrol was purchased from Fluka Chemika (St. Louis, MO). Two antibodies against Hsp90 were purchased, one used for immunoblot analysis, from Stressgen Biotechnology (Victoria, BC, Canada), and one for immunohistochemistry, from BD Biosciences (BD Pharmingen, San Diego, CA). The antibodies against Hsp90, Hsp70 and Hsc70, were purchased from StressGen Biotechnology. The antibody against p23 was from BD Biosciences. Anti Cyp40, ERß, and FKBP51 antibodies were purchased from Affinity BioReagents (Golden, CO). The antibody against Hsp56 (FKBP52) was from Santa Cruz Biotechnology (Santa Cruz, CA) and the anti-actin antibody was from NeoMARKERS (Fremont, CA). The antibody against ER was from Upstate Biotechnology (Lake Placid, NY). Rabbit and mouse immunoglobulin Gs (IgGs) and reagents for immunohistochemistry (Histostain-Plus Kit) were from Zymed Laboratories, Inc. (South San Francisco, CA). The enhanced chemiluminescence (ECL) Western blotting detection kit was from Amersham Pharmacia Biotech (Piscataway, NJ). Crystal-mount was from Biomeda Corp. (Foster City, CA). Restriction enzymes were from Stratagene (La Jolla, CA) and New England Biolabs (Beverly, MA). Primers were synthesized by Biosynthesis (Lewisville, TX). RNA polymerase chain reaction (PCR) kits were from Perkin Elmer (Branchburg, NJ). G3PDH primers (0.45 kilobase [kb] 5'and 3' G3PDH amplimers) were from Clontech (Palo Alto, CA). SYBR Green PCR Master Mix kit was from AB Applied Biosystems (Foster City, CA). Cell culture supplies were purchased from Invitrogen (Grand Island, NY). Electrophoresis reagents and materials were from BioRad (Hercules, CA). All other chemicals used were of analytical grade and were obtained from various commercial sources.
Animals and Treatments
Sprague-Dawley female rats or newborn male pups were purchased from Charles Rivers Laboratories (Wilmington, MA). Pregnant rats were treated daily with corn oil alone or corn oil containing 0.01 to 2 µg kg day DES, 0.1 to 200 mg kg day BPA, 0.1 to 10 mg kg day genistein, or 1 to 100 mg kg day coumestrol, administered by gavage or by s.c. injection (DES only), from Gestational Day 14 (Day Postcoitum ["c] 14) to the day of parturition, and doses were adjusted according to changes in animal weight. Male offspring at Fetal Day 21 (dpc21) and Postnatal Days 3 and 21 (Days Postpartum ["p] 3 and 21) were killed by CO2 inhalation. Testes were collected and either fixed in 3.5% buffered formaldehyde, snap-frozen in liquid nitrogen, or used for cell isolation. The animals were handled according to a protocol that was reviewed and approved by the Georgetown University Animal Committee.
Cell Isolation
Pure gonocyte preparations (100% purity) were isolated from dpp3 rat testis as previously described [29]. Briefly, testes from 40 rat pups of 3-day-old rats were decapsulated, digested first with 0.7 mg/ml collagenase and 0.2 mg/ml hyaluronidase for 30 min in RPMI containing antibiotics (100 U/ml penicillin; 0.1 mg/ml streptomycin) at 37°C;, and then with 0.25% trypsin-1 mM EDTA, and 0.15 mg/ml DNase I three times for 10 min each. Trypsin digestion was stopped by adding fetal bovine serum (FBS), the suspension was filtered through 47-µm nylon mesh, and centrifuged. The cell pellet was resuspended in RPMI containing 5% FBS, and kept overnight in a 5% CO2 incubator at 37°C in 150-mm culture plates to allow adhesion of Sertoli and myoid cells. Nonadherent cells were collected and further purified by sedimentation velocity at unit gravity on a 2% to 4% BSA gradient. Fractions containing more than 80% gonocytes (as judged by their morphology and large size) were pooled and further purified by adding a step of individual cell sorting using a micromanipulator system (TransferMan NK micromanipulator from Eppendorf Scientific; Brinkmann Instruments, Inc., Westbury, NY) equipped with a cell harvesting system coupled to a 15-µm glass transfer micropipette, to selectively collect gonocytes. The final pools of pure gonocytes (3000 to 6000 cells per sorting) were centrifuged, and cell pellets were frozen at –70°C for RNA extraction or protein analysis. Enriched Sertoli cell preparations obtained during the process of cell purification, and containing around 20% myoid cells were also collected to serve as a comparison.
Western Blot Analysis
Western blot analysis was performed as previously described [29]. Briefly, isolated gonocytes and Sertoli/myoid cell fractions from three independent cell preparations were pooled and solubilized in Laemmli buffer. Thus, each sample was indicative of the average protein expression levels found in 120 untreated rat pups, corresponding to 10 000–30 000 isolated gonocytes. In whole testes, samples from three different litters were examined, in which each sample was obtained by extracting three pooled testes from two to three pups of the same dam, using Laemmli buffer containing protease inhibitors (PMSF, 1 mM; leupeptin, 0.05 mg/ ml; aprotinin, 0.028 thousand International Units (TIU)/ml; and benzamidine, 1 mM) and sonication to facilitate protein extraction. Protein contents were determined using Coomassie staining quantification, and the proteins were separated by SDS-PAGE on 4%–20% polyacrylamide gels. Immunodetection of ER, ERß, and ER-associated proteins (Hsp90, Hsp90, p23, Cyp40, Hsp70, Hsc70, FKBP52, and FKBP51) as well as actin, which was used to normalize for loading variations, was carried out with specific primary antibodies diluted at 1:400 to 1:1000, and an ECL Western blotting detection kit. Changes in protein expression were measured by comparing the protein:actin ratios in the different conditions using the OptiQuant Acquisition & Analysis program from Packard Biosciences (Meridien, CT). For the experiments on nontreated rats only, representative results are shown. In the study addressing prenatal estrogen exposure, the results show the mean ± SEM of three experiments using rats from three different litters.
RNA Isolation and RT-PCR Analysis
Total RNA was isolated as previously described [29], except that the RT-PCR reactions were carried out using the TaqMan Reverse Transcription Reagents (AB Applied Biosystems). Briefly, total RNA obtained from purified gonocytes (2 µg) and somatic cells or whole testis (10 µg) were reverse transcribed, and the PCR reactions were carried out using an iCycler thermal cycler from BioRad on 2-µl aliquots of cDNAs using the primers described in Table 1. Simultaneous runs of the samples were performed using G3PDH primers (0.45 kb 5'and 3' G3PDH amplimers) as a housekeeping gene to normalize the data. The cycling parameters of the PCR reactions are shown in Table 2. PCR products were then separated on a 1% agarose gel, visualized by ethidium bromide staining, and the image analysis of the bands obtained was performed using the OptiQuant Acquisition & Analysis program.
|
|
In Situ Hybridization Experiments
Probe synthesis and labeling was performed as described before [29]. Briefly, probes of approximately 1000 base pairs (bp) in size were synthesized using a rat testis cDNA Library (Marathon-Ready cDNA kit; Advantage from Clontech) using the following set of primers for Hsp90: 885–906, 5'-CATCACTCAAGAGGAATATGGT-3'; 1873–1854, 5'-TCAGGGTTGATCTCTAGGTG-3'; and PCR kits (Clontech). The cycling conditions were as follows: denaturation at 94°C for 2 min; then annealing with the first five cycles at 94°C for 15 sec, 62°C for 20 sec, and 72°C for 2 min; followed by 30 cycles at 94°C for 15 sec, 60°C for 20 sec, 72°C for 2 min, and an extension phase at 72°C for 7 min. PCR fragments were purified, cloned, and sequenced. A large quantity of plasmid was produced; both sense and antisense products were linearized by restriction enzyme digestion and then labeled using 35S-CTP as described before [29]. In situ hybridization reactions were performed on paraffin sections of dpp3 rat testis samples. Nonsense and sense probes were used as controls and gave much lower signals than those obtained with the antisense probe. The in situ hybridization illustrations show a representative experiment for different treatment conditions.
Real-Time Quantitative PCR
Testes (two testes from two dpp3 siblings per sample) were homogenized in RNAzol B and processed through a conventional RNA extraction procedure as described before [29]. Reverse transcription for the 18S and target gene amplicons was then performed using an RNA PCR kit on the isolated total RNA extracts following the kit specifications. Real-time quantitative PCR (Q-PCR) was performed on an ABI PRISM 7700 Sequence Detector using a SYBR Green PCR Master Mix kit and primers specific for the genes of interest (Table 3). The cycling conditions were an initial step at 50°C for 2 min and 10 min at 95°C, followed by 40 cycles at 95°C for 15 sec and 60°C for 1 min. Direct detection of PCR products was monitored by measuring the increase in fluorescence caused by the binding of SYBR Green dye to double-stranded (ds) DNA, and the Comparative CT method was used to analyze the data. The amounts of the various genes were normalized to the endogenous reference (18S rRNA) by calculating the value of 2
Ct control–Ct treatment, with Ct being the difference between Ct of the gene of interest and that of 18S rRNA. The final data were expressed in a relative unit, with the control sample having a value of 1, representing the levels of the mRNA of interest present in the samples. For each treatment, the mRNA levels were determined in samples from three different litters and the mean ± SEM are presented.
|
Immunohistochemistry
Immunohistological analysis of dpc21, dpp3, and dpp21 rat testes was performed as previously described on 5-µm paraffin sections [29]. Briefly, an antigen retrieval method was used, and the sections were incubated with the primary antibodies diluted (1:80 to 1:300) in PBS containing 10% calf serum overnight at 4°C, then with reagents provided in the Histostain-Plus Kit from Zymed Laboratories. Nonspecific IgGs were used for the treatment of negative control slides. Sections were counterstained with hematoxylin, coated with Crystal-mount, and examined under brightfield microscopy with a BX40 Olympus microscope (Olympus America Inc., Melville, NY) coupled to a Spot RT color CCD camera (Diagnostic Instruments Inc., Sterling Heights, MI). The results obtained for animals born from four to six different mothers were compared, and representative results are shown.
Quantification of Spermatogonia in dpp21 Rat Testis
The quantification of spermatogonia in prepubertal testes was performed by two methods. Spermatogonia were counted in 5-µm sections of testes from two to three rats for each treatment, and they were identified according to their morphology, position adjacent to the basement membrane at the periphery of the tubules, and positive staining for Hsp90. In one method, the number of spermatogonia per tubule was counted for 50 tubules (transverse sections), chosen randomly over the surface of one or two sections per sample, using a 20x objective. In the other method, the volume of spermatogonia per testis was evaluated using a point counting method that took into account the number of Sertoli cells, tubule and testis volumes [37], adapted to digital pictures using Adobe Photoshop. Photographs of 5-µm sections were captured using an Olympus DP70 digital camera fitted on an Olympus BX40 microscope, with a final magnification of 65x. A circular visor containing 200 randomly chosen points, which was rotated before counting each new picture at a random angle, was numerically applied on top of the picture. Points falling over spermatogonia, tubules, interstitial spaces, and Sertoli cells were counted, expressed as a percentage of 200, and further used in the equations below:
 |
 | (4) |
Statistical Analysis
Statistical analysis was performed by unpaired t-test using the Instat (v.3.0) package from GraphPad, Inc. (San Diego, CA). In all experiments in which the differences between control rats and pups exposed in utero to estrogenic compounds were examined, both at the mRNA and protein levels, the experimental unit was the mother. In both cases, we compared the responses obtained for pups from at least three independent litters for each condition tested. In experiments in which a single sample was composed of the pooled testes of two to three pups from the same litter, in order to assure sufficient amounts of material, three samples from three different mothers were used and the mean ± SEM were calculated using n = 3, indicative of three mothers (and not the total number of pups).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Gonocytes were isolated from dpp3 rat testes and the mRNA and protein expression of ERß and several ER-associated proteins were examined. RT-PCR analysis of isolated gonocytes showed the presence of ERß, Hsp90, p23, and the immunophilin Cyp40 (Fig. 1A). However, the mRNA for the inducible Hsp70 was not detected in gonocytes, whereas it was clearly present in whole testis and to a lesser extent in Sertoli/myoid cell fractions (Fig. 1B). We then determined the protein levels of ER, ERß, and a panel of known ER-associated proteins by Western blot analysis of purified gonocytes and enriched Sertoli/myoid cells preparations (Fig. 2A). Protein levels were normalized against actin, used as a loading reference, and the expression levels between cell types were compared (Fig. 2A, right panel). Gonocytes contained high protein levels of ERß, Hsp90, p23, and Cyp40, but not Hsp70; only small amounts of FKBP52 (Hsp56), and only trace amounts of FKBP51. However, gonocytes expressed high levels of the constitutive Hsc70, the heat shock cognate protein of 70 kDa. In both cell types, Cyp40 was the predominant form of immunophilin expressed. It is interesting that ER was also detected at very low levels in isolated neonatal gonocytes, whereas the anti-ERß antibody used in this study detected only minute amounts of ERß in enriched Sertoli/ myoid cell preparations from dpp3 rats. Immunohistological analysis of dpp3 rat testis sections, using the same antibodies as those used in the Western blot experiments, showed that the highest levels of Hsp90, p23, and Cyp40 immunoreactivity were found in the cytosol of gonocytes, in agreement with the Western blot results (Fig. 2B). Unfortunately, the antibodies against the other ER-associated proteins did not function in immunohistochemical studies.
|
|
Effect of Prenatal Estrogen Exposure on the mRNA Levels of ER-Associated Proteins
The possibility that estrogenic exposure might affect components of the ER cytosolic complex in gonocytes was examined by measuring mRNA levels of Hsp90 (Fig. 3), p23, and Cyp40 (Fig. 4) in RNA extracts from whole testes of control and estrogen-exposed dpp3 rat pups, using Q-PCR. As shown in Figure 3, BPA and coumestrol exposure induced dose-dependent increases in the Hsp90 mRNA expression, reaching 2.2-fold and 2.6-fold increases (P 
0.05) for their strongest responses, respectively. Genistein induced similar increases in amplitude to those found with BPA and coumestrol, but with an optimal response of a 2.1-fold increase (P 0.05) at the lowest dose used. DES had only a small effect (1.56-fold increase; P 0.05) at the highest dose used. By contrast to Hsp90, the mRNA levels of two other major ER-associated proteins in gonocytes, p23 and Cyp40, did not change in a consistent manner following estrogen exposure (Fig. 4), except for a 40%– 60% decrease observed in Cyp40 mRNA levels following DES exposure.
|
|
In situ hybridization experiments using 35S probes were performed on dpp3 rat testis sections in order to determine the cellular localization of Hsp90 mRNA in neonatal testes from control and estrogen-exposed rats. These experiments showed that most Hsp90 mRNA was localized in the center of the seminiferous cords, corresponding to the gonocytes (Fig. 5B). While exposure to the phytoestrogens genistein (Fig. 5E) and coumestrol (Fig. 5F) led to Hsp90 mRNA increase primarily in gonocytes, exposure to BPA induced a more generalized increase throughout the testis, with a pattern of expression indicative of gonocytes as well as interstitial Leydig cells (Fig. 5C). The effects of DES were variable, with some samples showing a decrease (Fig. 5D) and others showing an increase (not shown).
|
Quantification of ER-Associated Protein Levels in Neonatal Testis Following Prenatal Estrogen Exposure
Considering that mRNAs and proteins do not always follow the same expression patterns, we used semiquantitative Western blot analysis to determine the protein levels of ERß and Hsp90, p23, Cyp40, and Hsc70 in whole testis by calculating the ratios of the protein of interest versus actin levels in three samples from independent litters for each treatment, and by comparing the ratios obtained in control (given a value of 1) and treated rats. As shown in Figure 6, Hsp90 presented significant changes in response to estrogen exposure, with increased expression from 1.5-fold to 2-fold in whole testis extracts (1.8-fold, 2.0-fold, 1.5-fold; P 0.05 for genistein, BPA, and coumestrol, respectively), in agreement with the changes found in mRNA expression. Although animals from three different litters showed similar trends in their responses averaging to the significant variations mentioned above, the amplitudes of their responses varied noticeably, as illustrated by the sizes of the standard errors. It should also be noted that each individual sample represented the protein extract from three pooled testes, collected from two to three pups of the same litter, in order to reduce the possibility of a single outlier pup in a litter showing an atypical response. Thus, the results presented herein reflected changes in six to nine pups issued from three dams for each treatment. By contrast, Cyp40 protein levels were not altered by the treatments, even with DES (Fig. 6), demonstrating that changes in mRNA levels are not always associated with changes in protein levels. Similarly, there was no change in either p23 or ERß protein levels (Fig. 6). Hsc70 protein expression was significantly increased by BPA treatment, and showed a small but not significant increase with the other compounds.
|
Changes in Hsp90 Protein Expression in Histological Sections of dpp3 Rat Testis Following Estrogen Exposure
Because Hsp90 was the protein the most affected by estrogen exposure, we further examined its cellular localization and levels of expression by immunohistochemistry. In preliminary studies, we determined that a dilution of 1: 200 of the anti-Hsp90 antibody would allow signal detection without reaching saturation, by performing reactions on control samples with dilutions ranging from 1:50 to 1: 500. Hsp90 expression was examined in testis sections from pups of six different litters for one dose for each compound and vehicle. In control samples, Hsp90 was predominantly expressed in gonocytes, and more specifically in their cytosol (Fig. 7A). As shown in Figure 7, C–F, exposure to either BPA (200 mg kg day), DES (1 µg kg day), genistein (10 mg kg day), or coumestrol (50 mg kg day) resulted in increased Hsp90 protein expression as compared with control samples (corn oil; Fig. 7A), primarily in gonocyte cytosol. Moreover, a slight increase in Hsp90 staining intensity could also be seen in Sertoli cell cytosol and in the interstitium. These results are in agreement with the mRNA changes observed by in situ hybridization following phytoestrogen exposure. However, the changes observed following BPA exposure in Hsp90 protein and mRNA expression differed partially, because the mRNA increase found in the interstitium appeared much stronger than the corresponding protein increase in this testicular compartment. For DES, all samples examined for protein expression showed an increase in Hsp90 protein levels similar to that of the representative sample shown in Figure 7D. This is in contrast with the results obtained on mRNA expression, which were variable between samples, ranging from a decrease to a small increase in expression. In general, all estrogen-exposed samples examined for a given treatment gave comparable results in protein expression.
|
Effects of Estrogen Exposure on the Expression of Hsp90 in Neonatal Gonocytes
Neither of the anti-Hsp90 antibodies used in the experiments described above was specified to preferentially recognize Hsp90 or Hsp90ß. Thus, we performed additional experiments with an antibody that specifically recognizes Hsp90 and does not cross-react with Hsp90ß to determine whether the changes we observed applied to this isoform. As shown in Figure 8A, semiquantitative Western blot analysis of Hsp90 expression gave results that closely mirrored those obtained with the regular antibody, suggesting that a large part of the increases observed correspond to changes in Hsp90 expression. Further analysis of Hsp90 protein expression in isolated gonocytes and Sertoli/myoid cell preparations (Fig. 8B) indicated that the majority of Hsp90 was expressed in gonocytes. Indeed, RT-PCR analysis using a set of probes targeting a sequence unique for rat Hsp90 showed expression of Hsp90 mRNA in isolated dpp3 gonocytes (Fig. 8C). Immunohistochemical analysis confirmed that Hsp90 is primarily expressed in gonocytes (Fig. 8D), following a pattern identical to that obtained with the general Hsp90 antibody. Moreover, when we examined sections from control and estrogen-exposed rats using the Hsp90-specific antibody, we obtained results that were identical to those presented in Figure 7 (data not shown), showing that estrogen exposure induces an increase in the expression of Hsp90.
|
Changes in Hsp90 Protein Expression in Histological Sections of dpc21 and dpp21 Rat Testis Following Estrogen Exposure
In order to determine whether the effects observed on Hsp90 expression following prenatal estrogen exposure were sustained or were transient throughout testis development, we also examined Hsp90 expression in control and estrogen-exposed rats in fetal testis (dpc21), corresponding to the end of estrogen exposure, and at dpp21, an age when several types of more-mature germ cells are present. Examination of testes from control rats showed that Hsp90 was primarily expressed in gonocytes prior to birth (Fig. 9B), and that it was strongly expressed in spermatogonia and in pachytene spermatocytes in prepubertal testis (Fig. 9H). However, Hsp90 was not visible in leptotene spermatocytes (Fig. 9H), whereas the preleptotene spermatocytes visible in some of the sections expressed high levels of Hsp90 (Fig. 9I). Moreover, the levels of Hsp90 protein were higher in fetal samples exposed to estrogenic compounds than in controls (Fig. 9, C–F), suggesting that the Hsp90 overexpression observed at dpp3 had already occurred in utero and was maintained after birth. By contrast, the Hsp90 protein levels observed in individual germ cells in prepubertal testis appeared similar in control and treated samples (Fig. 9, H–L), suggesting that the difference in Hsp90 expression observed in gonocytes was transient and probably limited to early developmental stages. However, samples from estrogen-exposed testes were found to contain 2-fold to 3-fold more spermatogonia than those from control rats using both counting methods. The numbers of spermatogonia/tubule (mean ± SEM) were, respectively: control, 5.6 ± 1.8; BPA, 12.2 ± 4.0; genistein, 12.4 ± 2.9; coumestrol, 11.9 ± 2.3; and DES, 14.6 ± 3.9. The calculated volumes of spermatogonia per testis (mean ± SEM) were, respectively: control, 2.83 ± 0.04; BPA, 6.97 ± 0.10; genistein, 8.32 ± 1.04; coumestrol, 5.25 ± 0.71; and DES, 5.22 ± 1.01. Moreover, the number of spermatogonia with mitotic figures observed in sections from treated rats was higher overall than the number observed in control rats (Fig. 9, I–L). These results indicate the persistence of differences between control and estrogen-treated rats in the progeny of the gonocytes until prepuberty.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
knockout ( ERKO) mice, the males of which are infertile primarily due to abnormal efferent ductile development and fluid movements in the epididymis [38], and of aromatase knockout (ArKO) mice that present alterations of testis morphology and germ cell numbers [39, 40] have clearly established that endogenous estrogen plays a physiological role in testis development and function [41]. Indeed, the dichotomy of estrogen action in testis is exemplified in our previous studies in which we found that 17ß-estradiol and PDGF induced neonatal testicular gonocyte proliferation in vitro [8], whereas prenatal exposure to several estrogenic compounds was associated with an increased expression of the PDGF receptors (PDGFRs) in neonatal testis, with the strongest effect being the overexpression of PDGFRß in gonocytes [29]. Taken together, these results suggested that endogenous estrogen may play an important physiological role in gonocyte development, while exogenous estrogens might exert some of their detrimental effects by altering the expression of key signaling molecules in gonocytes. In the present study, we examined the possibility that estrogen exposure might alter components of its own pathway in gonocytes by determining which ER-associated proteins were expressed in neonatal gonocytes, and then by comparing their expression levels in control and estrogen-exposed rats.
We found that, under physiological conditions, neonatal (dpp3) gonocytes expressed high levels of Hsp90, more specifically the Hsp90 isoform, which is in agreement with a recent report that used Hsp90 as a marker for dpp4 rat gonocytes in culture [25]. By contrast, only low levels of Hsp90 were found in Leydig and Sertoli cells, which have been shown to express mostly Hsp90ß [5, 27], suggesting that the antibody used in this study, which was raised against human testicular extracts, preferentially recognized Hsp90, further explaining the similitude of the results obtained using this antibody and the Hsp90-specific antibody. Our study also demonstrated that Hsp90 was expressed in prenatal gonocytes, and in specific subsets of germ cells in prepubertal testis, the spermatogonia and pachytene spermatocytes, suggesting a role for this protein in the initial steps of germ cell development and in specific developmental events in more mature germ cells, in agreement with a study that reported the preferential mRNA expression of Hsp86, a homolog of Hsp90, in pachytene spermatocytes of mouse testis [23]. In the same order of idea, another Hsp90-related protein, Hsp105, was found in adult rat testis and isolated spermatozoa, but only in rats older than 5 wk, suggesting the existence of yet another Hsp90 isoform specific for pubertal and adult germ cells [42].
The present study also showed the presence of high levels of the cochaperone p23, and the immunophilin Cyp40 in the cytosol of neonatal gonocytes. Although the association of these proteins with Hsp90 has been well described in other cell models, this is the first time, to our knowledge, that their presence is demonstrated in germ cells. The colocalization of these proteins with Hsp90, and their low levels in somatic cells, suggest that ER complexes are more abundant in germ cells than in somatic cells in neonatal testis. Western blot analysis of purified gonocytes showed that ERß was strongly expressed in these cells, in agreement with earlier reports [6]. Moreover, the recent discovery of new forms of ERß in fetal and tumoral human germ cells [12, 13] suggests that such variant forms might also be expressed in rat gonocytes. Sertoli cells were found to express only minimal levels of ERß in our experiments, in contrast to previous reports [7], probably due to antibody differences or to a down-regulation of the receptor in these cells during the cell isolation procedures. It is interesting that low levels of ER were also found in gonocytes, which was not reported in previous studies [7], probably due to differences in methodologies, antibodies used, and detection limits of the assays, especially in view of the low levels present. Thus, while ERß is probably responsible for most of the ER-dependent effects of estrogens in gonocytes, one cannot entirely exclude a potential role of ER in these cells. While revising this manuscript, a new report appeared that examined the morphology and number of various cell types in testis of neonate mice in which either ER or ERß genes had been disrupted [43]. This study showed that ERß–/– mice had a greater number of gonocytes, but normal numbers of Leydig and Sertoli cells, as well as normal testicular testosterone levels, compared with the wild type. No differences were found in any of these parameters between ER–/– mice and wild-type mice. This study, which concluded that endogenous levels of estrogen can inhibit male germ cell growth via ERß, further emphasizes that more studies are needed to clarify the role of ERß and estrogen in germ cells during testis development.
The absence of Hsp70 in gonocytes, together with studies reporting its expression in specific subsets of adult germ cells [44], emphasizes its tight developmental regulation. This indicates that the expression of Hsp70 isoforms is differentially regulated in testicular cells, with Hsc70 expressed both in germ cells from neonatal to adult stages as well as in neonatal somatic cells, while the expression of Hsp70 is restricted to somatic cells in neonatal testis and to specific germ cell types in adult testis. In neonatal testis, ER appears to be exclusively associated with the Hsp90-Hsc70 complex in germ cells, while somatic cells contain all Hsp90 and Hsp70 forms, further suggesting that these proteins participate in different types of functions during testis development.
Most of the work reported on Hsps addressed cellular responses to heat stress, especially in studies related to testis, an organ that is highly sensitive to temperature changes, in which Hsps have been shown to rescue unfolded proteins from denaturation [15, 45, 46]. These proteins are also required for the proper maturation of numerous proteins under normal physiological conditions [15, 46]. Moreover, Hsp90 induction has been shown to occur in response to effectors such as hormones, growth factors, and in specific pathologies [15].
In the present study, we found that prenatal exposure to four different estrogenic compounds induced the overexpression of Hsp90 in gonocytes, both at the mRNA and protein levels. The comparison of the changes in mRNA and protein levels of Hsp90 observed in response to estrogen exposure indicate that BPA and genistein induced the larger and most consistent increases, in a fashion similar to our previous observations on PDGFR expression [29]. With coumestrol, the increase in mRNA levels was larger than that of the protein, probably reflecting an increase in RNA stability not accompanied entirely by increased translation. With DES, the results obtained were not as reproducible from animal to animal as those obtained for the other estrogenic compounds, possibly because DES is an extremely potent estrogenic compound that acts on both ERs [47], and that some pregnant rats were more sensitive to it than others. In this regard, it is noteworthy that a recent study using organ cultures of testes from 14.5-day-old rat fetuses reported that incubation with DES induced a decrease in the total number of fetal gonocytes in an ER-dependent manner, largely due to increased apoptosis [48], whereas the same concentration of estradiol had only a minor effect, indicating that fetal gonocytes were responsive to estrogens, but that DES was more potent on these cells than the natural hormone estradiol.
Considering that the affinities of estradiol and DES for ERs are of the same order of magnitude [47], the study mentioned above [48] and our results suggest that the actions and targets of DES might be more complex than those of the endogenous hormone or less potent estrogenic compounds such as BPA, genistein, and coumestrol. Nevertheless, our results showed that a common response of neonatal testis to prenatal estrogen exposure was an increase in Hsp90 (Hsp90) expression, mostly localized in gonocytes. Such an Hsp90 increase has previously been reported to occur in ewe uterus in response to estradiol, as part of the normal events occurring during parturition [34], and in mouse uterus in response to estradiol and BPA, where estradiol and BPA were shown to increase Hsp90 mRNA expression using both similar and distinct pathways [35, 49]. Thus, in testis as well as in uterus, estrogens appear to regulate their own signaling pathway by altering the expression of Hsp90, a critical component of the ER macromolecular complex, potentially leading to an increased responsiveness of the cells to estrogens. In this context, the phenomenon we observed in response to estrogen exposure could be interpreted as the exacerbated version of a normal physiological response, as proposed for the ewe uterus [34]. Indeed, upon examination of testes from estrogen-exposed fetuses, we found that Hsp90 protein expression was already elevated in late gestation fetal testes. Thus, it is possible that the elevated maternal levels of estradiol, known to occur during the last one-third of gestation [50], would normally enhance the ability of the developing testis to further respond to the estradiol produced in the neonatal testis by the Sertoli cells [51], and that in utero exposure to high levels of exogenous estrogens resulted in the overactivation of this mechanism.
Our study revealed that Hsp90 was the only ER-associated protein to be altered in a significant and consistent manner in gonocytes. Cyp40 protein expression was, interestingly, not affected by any of the treatments used, although DES was found to significantly decrease its mRNA expression, probably due to an effect on mRNA stability, in contrast with breast cancer tumor cells, in which Cyp40 mRNA expression was found to increase in response to estradiol [52]. Thus, although the ER cytosolic complex present in gonocytes appears similar to that found in other cell systems, its mode of regulation appears to be different.
That Hsc70 levels did not change following exposure to three of the compounds tested fits with its usual ubiquitous and constitutive expression and its role in protein folding [53]. However, a recent study has reported that Hsc70 was actively released from tumor cells in response to cytokine stimulation, postulating that Hsc70 might have yet unknown regulatory functions [54]. In this regard, our experiments showed that BPA induced a significant increase in Hsc70 protein levels, suggesting that Hsc70 might also be involved in the response to specific stressors in testis. These results also suggest that BPA has two modes of action in testis, one common with other estrogenic compounds, and one unique to BPA. It is interesting that BPA has been found to play such a dual effect in mouse uterus [49]. Considering that p23, Cyp40, Hsc70, and ERß were two to four times more abundant in gonocytes than in Sertoli/myoid cell extracts and the interstitium, the lack of effect observed for these proteins probably characterizes the gonocyte response to estrogens. Taken together, these results pinpoint at Hsp90 as the sole component of the ER complex affected by estrogen exposure in neonatal gonocytes.
In view of the multiple roles of Hsp90, including the responses to toxicants [55], and the regulation of cellular events associated with development [18], differentiated functions of tissues, aging, and cancer [14, 46], it is easy to envision that an increase in Hsp90 expression might affect not only ER, but also other Hsp90 client proteins expressed in gonocytes. In this regard, several proteins belonging to the PDGF cascade, such as Raf1 and ERK1/2, have been shown to be sensitive to Hsp90 inhibitors and to depend on Hsp90 for their activity [22, 56]. As a consequence, an increase in Hsp90 expression could not only change the response of neonatal germ cells to endogenous estrogen, but also their response to PDGF.
In the present study, we found that the testes of prepubertal rats exposed in utero to estrogens contain 2-fold more spermatogonia than the testes of control rats, suggesting that gestational estrogen exposure triggered long-term effects in the germ line stem cell population. Indeed, long-term morphological and functional alterations in male reproductive tissues have been reported following gestational exposure to estrogens, including an increase in reproductive tract tumors in male mice in response to DES [57], and the demasculinization of the reproductive system in rats following exposure to genistein [60]. Further studies are in progress to determine whether the difference in the number of spermatogonia is still visible in adult rats.
From our study and other studies, it is clear that estrogens can exert both beneficial and detrimental effects on testis development, with prenatal exposure to exogenous estrogens leading to the unregulated stimulation of a physiological process or desensitization of the system, thus preventing a subsequent physiological function of estradiol to occur. Another possibility is that the estrogenic compounds used in this study might interact with molecules and cellular pathways not directly related to ER or estradiol, but still involved in the ability of the cells to respond to physiological signals. While most existing studies have focused on recording changes in morphology and reproductive function of the testis, the goal of the present work was to identify specific molecular targets of estrogens in testis. This study, together with our previous work [29], showed that prenatal exposure to estrogenic compounds was able to trigger changes constrained to few signaling molecules, such as Hsp90 and PDGFRs. Indeed, Hsp90 might represent a molecular bridge between the estradiol and PDGF pathways in gonocytes. More studies will be needed to identify the proteins directly interacting with Hsp90 in gonocytes, and to determine the functional and developmental consequences of the changes induced by prenatal estrogen exposure.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Martine Culty, Department of Biochemistry and Molecular Biology, Georgetown University School of Medicine, 3900 Reservoir Road NW, Washington, DC 20057. FAX: 202 687 7855; cultym{at}georgetown.edu
Received: 26 March 2004.
First decision: 24 April 2004.
Accepted: 24 June 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ß in neonatal rat testis: identification of gonocytes as targets of estrogen exposure. Biol Reprod 2003 68:867-880
