| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
MRC Human Reproductive Sciences Unit,4 Centre for Reproductive Biology, University of Edinburgh, Edinburgh EH16 4SB, Scotland, United Kingdom
Laboratory for Experimental Medicine and Endocrinology,5 Department of Developmental Biology, Catholic University of Leuven, B-3000 Leuven, Belgium
Institute of Experimental Morphology and Anthropology,6 Bulgarian Academy of Science, 1113 Sofia, Bulgaria
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sertoli cells, spermatogenesis, testosterone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The molecular and biochemical pathways that androgens activate in Sertoli cells remain virtually unknown [10], though secretion of seminiferous tubule fluid [11–13] and maintenance of junctional contacts with meiotic and postmeiotic germ cells [14] are likely endpoints of such effects. To date, only one androgen-regulated gene, the Pem homeobox gene, has been identified in Sertoli cells of rodents [15, 16]. Pem is expressed maximally in Sertoli cells at stages VII–VIII in the mouse and rat testis, but its downstream effects remain unknown, and transgenic overexpression of Pem in mice results in some impairment but no gross effects or obliteration of spermatogenesis or fertility [17].
In searching for candidates for androgen regulation of spermatogenesis, we were prompted to re-examine the expression of D-type cyclins by data demonstrating that cyclin D1 could act as a repressor of ligand-dependent AR activation in the prostate by binding to the N-terminal AF-1 domain of the AR [18]. In addition, stage-specific expression of retinoblastoma 1 (RB1) in Sertoli cells, a downstream target for D-cyclins, was reported to be androgen dependent in adult rats [19]. D-type cyclins are associated with the coordination of G1-phase progression into S-phase of the cell cycle, but the view that D-type cyclins have an exclusively mitogenic role has been questioned by new findings. These include up-regulation of cyclin D3 in quiescent myotubes [20], up-regulation of cyclin D1 and D3 in specific lineages differentiated from CD34+ progenitor cells [21], and the increased abundance of cyclin D2 associated with growth arrest in human and murine fibroblasts [22]. In each of these situations, the D-cyclins are thought to respond to extracellular signals by regulating signal transduction rather than cell division.
Cyclin D1, the most studied member of the D-type cyclin family, can interact with nuclear receptor transcription factors such as the AR [23], estrogen receptor-alpha (ER
) [24], myb-like transcription factor, DMP1 [25], and cofactors such as NcoA/SRC1a [26] and TAF250 [27]. While cyclin D1 activates the transcriptional potential of ER in a ligand-independent manner [24], it acts as a repressor of ligand-dependent AR activation in the prostate [18]. Cyclin D1 thus appears to function differentially depending on the steroid receptors it interacts with. In comparison to cyclin D1, expression of cyclin D2 is uncharacterized with regard to androgen dependence in adult male reproductive organs, including the testes, though it is of interest that male cyclin D2-deficient mice have hypoplastic testes [28]. The expression pattern of cyclin D2 in the normal adult testis is also ambiguous. In the mouse, cyclin D2 immunoexpression was shown in spermatogonia, spermatocytes, and spermatids [29], whereas in situ hybridization revealed slight enrichment of cyclin D2 mRNA levels in germ cell-deficient mouse testes, consistent with Sertoli cell localization [30]. In contrast, no immunoexpression of cyclin D2 was observed in normal adult human testes [31].
With promising indications of AR interactions with cyclin D1 in the prostate, testicular atrophy in cyclin D2-deficient mice, and stage-specific androgen-dependent expression of RB1 in Sertoli cells in the rat, we decided to investigate the cell-specific pattern of expression of D-cyclins in the rat and mouse testis and whether this expression was androgen dependent. As our preliminary studies indicated that only cyclin D2 was immunoexpressed in Sertoli cells in the androgen-deprived rat, our detailed studies focused on this.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Most studies used adult male Wistar rats, aged 80–100 days, but for age time-course studies or neonatal treatments, all-male litters of 8–12 pups (generated by cross-fostering pups on the day of birth) were used. All rats were bred in our own animal house and were maintained in a controlled environment with free access to food (soy-free diet) and water. Sertoli cell-selective AR knockout mice (SCARKO mice) were generated as described later.
Rat studies To study androgen and/or estrogen regulation of Sertoli cell function, adult rats were administered a single intraperitoneal injection of 75 mg/kg ethane dimethane sulfonate (EDS) in dimethylsulfoxide/water (1:3; v/v). This treatment has been established to destroy all Leydig cells within 3 days and to reduce serum and testicular testosterone to undetectable levels [32, 33]. Quantitatively normal spermatogenesis can be maintained in EDS-treated rats by the subcutaneous injection of 25 mg testosterone esters (TE; Sustanon, Organon Laboratories, Cambridge, UK) in 0.1 ml arachis oil every 3 days, beginning on Day 0 [33, 34]. Control rats were treated with the vehicle alone. Testes from control, EDS-, and EDS + TE-treated rats were studied at Day 6 after EDS/vehicle injection. Our earlier studies have shown that at this time, maximal withdrawal of androgen action on Sertoli cells at stages VI-VIII is demonstrable but with minimal germ cell degeneration [4]. At least six animals in each treatment group were examined.
In other EDS-treated animals, the effect of acute steroid hormone manipulation was assessed. Thus, at 6 days after EDS-induced testosterone withdrawal, groups of three to six rats received a single subcutaneous injection of TE (25 mg) and were killed 4 h later. As the administered TE could be metabolized to estradiol, other studies assessed this possibility in several ways. First, rats were treated with TE ± the ER antagonist (Imperial Chemical Industries [ICI]) 182,780 (3 mg/kg in 1 ml/kg corn oil; kindly provided by Dr. Alan Wakeling, AstraZeneca, Alderley Park, Cheshire, UK) and killed +4 h; animals treated with either vehicle or ICI alone were used as controls. Second, other EDS + 6 day-treated rats were injected subcutaneously with 250 µg/kg estradiol-17ß (E2; Sigma-Aldrich, Dorset, UK) in 0.1 ml corn oil or with 125 µg/kg diethylstilbestrol (DES; Sigma-Aldrich) in 0.1 ml corn oil and killed 4 h after injection.
A group of four adult rats were injected subcutaneously with 100 mg/ kg flutamide (Sigma-Aldrich) in 1 ml/kg corn oil and killed 18 h later. Flutamide is an AR antagonist that binds the AR and prevents its downstream signaling [35]. To establish when cyclin D2 is first expressed, testes from rats aged 10, 15, or 35 days were examined, as a progressive increase in testosterone levels occurs during this period [36].
All rats were killed rapidly by inhalation of carbon dioxide followed by cervical dislocation. Some of the animals were then immediately perfusion-fixed with Bouins as described previously [37, 38]. These studies were performed according to the Animal Scientific Procedures (UK) Act (1986) under Project License approval by the UK Home Office.
Mouse studies: Generation of SCARKO mice The AR knockouts were generated using Cre/loxP technology. ARflox animals with exon 2 of AR floxed were crossed with AMH-Cre animals expressing Cre recombinase (under the AMH gene promoter) selectively in Sertoli cells to generate the SCARKO line. Full details are provided elsewhere [8]. Testes from adult (50 day) wild-type and SCARKO mice were dissected out and fixed in Bouins as described.
Immunohistochemistry
For rat testis sections, immunohistochemistry was performed on dewaxed 5-µm sections in conjunction with heat-induced antigen retrieval in 0.01 M citrate buffer, pH 6.0 (Sigma-Aldrich), using a pressure cooker; for mouse testis sections, antigen retrieval was not used because retrieval resulted in nonspecific staining. Anti-AR (sc-816; Santa Cruz Biotechnology, Santa Cruz, CA) and anti-cyclin D2 (sc-181; Santa Cruz Biotechnology) were used at 1:200 and 1:100, respectively. Monoclonal anti-cyclin D2 (MS-221-P1; Neomarkers, Fremont, CA), monoclonal anti-p27kip1 (NCL-p27; Novacastra, Newcastle-Upon-Tyne, UK), and anti-RB1 (MAB3186; Chemicon, Temecula, CA) were used at 1:50, while anti-phospho RB1 antibodies (9307, 9301, and 9308; Cell Signaling Technology, Beverly, MA) were diluted 1:100. Appropriate biotinylated secondary antibodies (E0353 and E0464; DAKO, Cambridge, UK) were applied using protocols detailed elsewhere [8].
For double immunostaining with vimentin and cyclin D2, rat testis sections were first stained with anti-cyclin D2 (Santa Cruz Biotechnology) as described previously before incubation with mouse monoclonal anti-vimentin (M0725; DAKO) at 1:100 and detection with a biotinylated rabbit anti-mouse antibody (E0464; DAKO). An ABC-alkaline phosphatase (Vector, Peterborough, UK) system was used with Fast Blue (1 mg/ml Fast Blue BB salt [Sigma-Aldrich] in 0.1 M Tris-(hydroxymethyl)methylamine, pH 8.2, 200 µg/ml Naphthol AS-MX phosphate [Sigma-Aldrich], 2% [v/ v] dimethylformamide buffer) to enable visualization of vimentin protein expression. Sections were counterstained with hematoxylin and aqueous mounted in Hydromount (National Diagnostics, NJ).
To allow comparative evaluation of the immunostaining, sections of tissues from control and treated animals were processed in parallel on at least three occasions to ensure reproducibility of results; on each occasion, tissue sections from three to six animals in each treatment group were run. To ensure comparability, sections from control, EDS-, and EDS + TE-treated rats or from control and SCARKO mice were mounted on the same slide.
Quantitation of RB1 Immunostaining
The optical density of Sertoli cell nuclei was determined using an Olympus BH-2 microscope (Olympus Corp., New Hyde Park, NY) fitted with a constant illuminating light source, a Prior automatic stage (Prior Scientific Instruments, Cambridge, UK), and a Diagnostic Instruments Spot Insight CCD video camera (supplied by Olympus Corp.). Immunostaining intensity was measured using the line profile tool of Image-Pro Plus software v4.5.1 (Media Cybernetics UK, Wokingham UK) with default free-form intensity calibration. Standardization, calibration, calculation, and protocol details were as described elsewhere [6]. Slides from three control and three EDS-treated rat testes were used; values were obtained from 10 Sertoli cell nuclei per tubule, five tubules per each of the three stage groups. Statistical significance was assessed using analysis of variance with replication.
Immunofluorescence/Confocal Microscopy
Sections of rat testis were subjected to antigen retrieval and processed as described previously. Washes between incubation with reagents were in phosphate-buffered saline (PBS) twice before the incubation step with the secondary antibody and then with an initial wash in PBS+0.005% (v/v) Tween 20 followed by PBS for subsequent washes after incubation with the secondary antibody. Cyclin D2 (Neomarkers) and AR (Santa Cruz Biotechnology) antibodies were diluted 1:50 and 1:800, respectively, in PBS/goat serum/BSA and incubated overnight at 4°C. Sections were incubated with both HRP-conjugated goat anti-mouse IgG (P0447; DAKO) diluted 1:100 and biotinylated goat anti-rabbit IgG (E0432; DAKO) diluted 1:500 for 1 h. All further incubations were carried out in the dark. Cyclin D2 signal was enhanced and visualized using a Tyr-Cy5 kit (NEL745; NEN Life Science Products, Boston, MA) according to the manufacturer's instructions. After washing, the fluorochrome streptavidin 488 Alexafluor (Molecular Probes, Inc., Eugene, OR), diluted 1:200, was added to slides for 2 h. Sections were incubated with biotin (from Avidin/ Biotin blocking kit; Vector), counterstained with propidium iodide (Sigma), and mounted in Permafluor (Immunotech-Coulter, High Wycomb, Buckinghamshire, UK). Fluorescent images were captured on a Carl Zeiss laser scanning microscope LSM 510 (Jena, Germany). The alexafluor 488 was detected using the Argon laser (excitation beam, 488 nm) and an emission bandpass filter at 515–530 nm, while the Cy 5 was detected with the helium/neon 2 laser (excitation beam 633 nm) and emission long bandpass filter at 650. Propidium iodide counterstaining was visualized using the HeNe1 laser (excitation beam 543nm) and an emission bandpass filter at 560–615 nm.
Transillumination-Assisted Microdissection of Seminiferous Tubules
Seminiferous tubules (ST) were isolated on a cooled microscope stage as detailed previously [4], using the transillumination method developed by Parvinen [39]. Aliquots of 5 cm ST at stages I–V, VI–VIII, and IX– XIV from one control, one EDS + TE, and two EDS-treated rat testes were snap-frozen and stored at –70°C or immediately processed for RNA isolation as described in the following.
RNA Extraction, Reverse Transcription, and Real-Time Semiquantitative PCR
Total RNA was extracted from frozen or freshly isolated ST using the RNeasy kit (QIAGEN Ltd, West Sussex, UK) according to manufacturer's instructions, incorporating a DNase treatment step to minimize genomic DNA contamination. RNA quality and concentration were determined by assays from the RNA 6000 Nano Labchip kit (Agilent Technologies UK Limited, Cheshire, UK) used with an RNA ladder (Ambion Ltd, Cambridgeshire, UK) and analyzed on the Agilent 2100 Bioanalyzer. Total RNA from each sample was reverse transcribed by Multiscribe RT with random hexamer primers. 18S rRNA was the endogenous control using TaqMan 18S rRNA control primer/probe (VIC-labeled) set, while Ccnd2 (cyclin D2) transcripts were detected using specific primers and FAM-labeled probe from the TaqMan gene expression assays (Assay ID: Rn01492397_m1); both sets of primers/probe and cDNA were used in conjunction with Taqman PCR Universal master mix in a multiplex PCR reaction. PCR was performed using an ABI Prism 7900HT sequence detector, and data analysis used the Sequence Detector software (version 2.1) to determine threshold cycle (CT, number of cycles for signal to cross threshold). The mRNA expression levels ofCcnd2, normalized to endogenous 18S rRNA reference and relative calibrator (i.e., control seminiferous tubules stages I-V; CI-V), was calculated using 2-
CT, where CTsample = CTsample-averageCTCI-V and CTsample = CTcyclinD2-CT18SrRNA. The value of 2-CT should represent linearly the relative mRNA levels of samples to the calibrator. All reagents, equipment, and accessories for real-time PCR were purchased from Applied Biosystems (Cheshire, UK). Experiments on 96-well plates were run in triplicate with total RNA (no RT) and water only as negative controls.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In control rats, stage-specific immunoexpression of AR was evident in all animals, with most intense expression at stages VI–VIII and lowest expression at stages IX–XIV (Fig. 1, Table 1), as reported previously [6]. In contrast, immunoexpression of cyclin D2 was completely absent from most control testes, though occasional lightly immunopositive Sertoli cell nuclei were evident at stages IX– XIV (Fig. 1, Table 1). At Day 6 after EDS treatment, Sertoli cell immunoexpression of AR was reduced markedly such that immunoexpression was completely absent at stages IX-XIV, only occasionally evident at stages I-V, and just detectable at stages VI–VIII (Fig. 1, Table 1). In contrast, intense immunostaining for cyclin D2 was now evident in Sertoli cells at stages IX–XIV in EDS-treated animals and barely visible in stages I–VIII (Fig. 1, Table 1). The disappearance/reduction in AR (Table 1) and the appearance of cyclin D2 immunoexpression in EDS-treated animals was prevented completely when TE was administered from the time of EDS treatment (Fig. 1, Table 1). The two different anti-cyclin D2 antibodies (Santa Cruz Biotechnologies and Neomarkers) that were used gave identical staining patterns in these treatment groups except for nonspecific staining of spermatid cytoplasm with the Santa Cruz antibody; nonspecificity was indicated by the lack of EDS-treatment effect on spermatid staining and was confirmed by demonstrating that it could not be blocked by preincubation with a 5-fold excess of the C-17 peptide immunogen, whereas the latter treatment completely eliminated the immunostaining in Sertoli cell nuclei at stages IX–XIV in EDS-treated rats.
|
|
Acute (+4 h) Effects of Androgen or Estrogen Treatment on Immunoexpression of AR and Cyclin D2 on Day 6 After EDS Treatment
The stage-specific increase in immunoexpression of cyclin D2 that was evident at Day 6 after EDS treatment was reversed within 4 h by injection of TE (Fig. 1); this suppression was complete in three rats but was incomplete in two rats (Table 1). This acute TE-induced reduction in cyclin D2 expression was accompanied by induction of the normal stage-specific pattern of immunoexpression of AR (Table 1). Coadministration of the ICI ER antagonist together with TE failed to block the TE-induced changes (Fig. 1, Table 1), suggesting that the TE effects were not due to conversion of the TE to estradiol. This conclusion was confirmed directly by showing that acute treatment with either estradiol (Fig. 1, Table 1) or DES (Table 1) was unable to suppress cyclin D2 immunoexpression that occurred at Day 6 after EDS treatment.
Neonatal Expression of Cyclin D2 and AR in the Rat
In accord with the literature [6], intense AR immunoexpression was evident in Leydig cells and peritubular myoid cells of rats on Day 5, whereas AR immunoexpression was only weakly detectable in Sertoli cells at Day 5 (not shown). However, the intensity of AR immunoexpression in Sertoli cells increased progressively thereafter, and stage specificity of expression was evident from Day 15 through to adulthood. Correspondingly, Sertoli cell nuclear immunoexpression of cyclin D2 emerged only after Day 10, was evident in Day 15 animals, and peaked around puberty with highest expression at Day 35 in the animals tested before decreasing dramatically to no expression in the adult (Fig. 2). Consecutive serial sections of testis sections from Day 25, Day 35, and adult testes from rats probed with either cyclin D2 or AR antibodies demonstrated that Sertoli cells at specific stages of the spermatogenic cycle exhibited either high intensity of AR immunoexpression and no or weak immunoexpression of cyclin D2 or, conversely, weak AR immunoexpression and intense immunoexpression of cyclin D2 (Fig. 2).
|
SCARKO Mice
In testes of adult wild-type and SCARKO mice, occasional spermatogonia stained positive for cyclin D2 (Fig. 3, top row, left inset), but in wild-type mice, no immunoexpression of cyclin D2 was evident in Sertoli cell nuclei (Fig. 3, top row, left). In contrast, in testes of adult (Fig. 3, top row, right) and 12-day (not shown) SCARKO mice, stage-specific (i.e., variation in intensity between tubules) Sertoli cell nuclear immunoexpression of cyclin D2 was evident in all animals (n = 3 per group). The positioning of Sertoli cell nuclei around the lumen of Day 50 SCARKO testes is typical of the SCARKO adults [8], and the immunostaining of p27kip1, a marker of differentiated Sertoli cells [36, 40], was used to illustrate this (Fig. 3, bottom row).
|
Coimmunolocalization of Cyclin D2 and Vimentin in the Rat
Although it was clear that cyclin D2 immunoexpression occurred in Sertoli cell nuclei at stages IX–XIV in EDS-treated rats, our studies did not exclude the possibility of immunoexpression also in some spermatogonia, as found in mice [29]. To evaluate this possibility, double immunostaining with cyclin D2 and vimentin, a Sertoli cell cytoplasmic cytoskeletal marker, was performed. This confirmed that cyclin D2 was immunolocalized exclusively to nuclei of vimentin-positive (Sertoli) cells at stages IX-XIV in EDS-treated rats (Fig. 4).
|
Investigation of the Coimmunoexpression of AR and Cyclin D2 Using Confocal Microscopy
Though our findings from studies with EDS indicated an inverse relationship between immunoexpression of AR and cyclin D2, it was impossible to confirm this impression directly in adult rats because induction of detectable cyclin D2 immunoexpression by EDS treatment also resulted in loss of AR immunoexpression and the converse when TE was administered or adult control tissue was used (Table 1). We therefore applied confocal microscopy to two situations in which AR and cyclin D2 were both immunoexpressed in Sertoli cells at reasonable intensity in the same sections, namely, 1) control rats at 35 days of age and 2) adult rats treated 18 h previously with the AR antagonist flutamide (Table 1). In both situations, it was shown that immunoexpression of AR and cyclin D2 never coincided (Fig. 5) and that, only rarely in Day 35 animals, a few tubules had Sertoli cells that were immunopositive for cyclin D2 and for AR (Fig. 5). It is emphasized that immunoexpression of cyclin D2 in Sertoli cells of flutamide-treated rats was comparable in intensity and stage-specific pattern to that seen in rats at 6 days after EDS treatment, indicating that blockade of AR signaling, without loss of AR immunoexpression, is sufficient to induce stage-specific cyclin D2 immunoexpression.
|
RT-PCR of Ccnd2 Transcripts
To confirm the stage-dependent immunohistochemical data, ST from control, EDS- and EDS + TE-treated rats were microdissected and grouped into stages I–V, VI–VIII, and IX–XIV. Semiquantitative PCR was performed on cDNA from staged ST and demonstrated a 3-fold increase in Ccnd2 mRNA transcript levels in ST from EDS-treated animals at stages IX–XIV but not at other stages (Fig. 6).
|
Immunoexpression of RB1
In addition to the corroborative evidence obtained for altered cyclin D2 RNA expression that corresponded to the stage-dependent immunohistochemistry findings, expression of an immediate downstream target of cyclin D2, namely, RB1, was examined. We used a monoclonal antibody raised against RB1 epitope 300-380 (RB1300-380), which is reported to recognize underphosphorylated and highly phosphorylated forms of RB1 [41]. Intense RB1 expression was found in controls in spermatogonia, and this expression was unaffected by androgen deprivation (not shown). Additionally, stage-specific Sertoli cell nuclear immunostaining for RB1300-380 was found, peaking at stages VII–VIII, declining rapidly in stages IX–XIV, and then rising slowly at stages I–V (Fig. 7, A and B), a pattern similar to that for AR immunoexpression. The stage-dependent pattern of RB1300-380 immunostaining in Sertoli cell nuclei was attenuated after treatment with EDS (Fig. 7, A and B) or flutamide (not shown) due to increased immunoexpression specifically at stages IX–XIV; this change was prevented by androgen replacement (6 day) in EDS-treated rats (not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The stage-dependent pattern of cyclin D2 immunoexpression, which in adult rats could only be visualized after androgen withdrawal, precisely mirrored the stage-dependent pattern of AR immunoexpression in Sertoli cells seen in the normal adult rat testis. Thus, immunoexpression of cyclin D2 was highest at stages IX–XIV, when AR immunoexpression is normally lowest, and was nondetectable at stages VI–VIII, when AR immunoexpression is normally at its highest [6]. By using confocal microscopy on these same samples, it was shown that, at the level of individual Sertoli cell nuclei, the peaks of immunoexpression of AR or cyclin D2 did not coincide, with coimmunoexpression occurring rarely, if at all. Moreover, the present findings show that absence of detectable cyclin D2 immunoexpression in stages IX–XIV of the normal adult rat testis is completely dependent on functional AR signaling, as flutamide treatment, which disrupts AR signaling but did not disrupt AR protein immunoexpression, led to appearance of cyclin D2 immunoexpression at these stages.
Our findings in rats are reinforced by those obtained in SCARKO mice, in which we have shown that selective withdrawal of androgen action on Sertoli cells via transgenesis [8] also leads to appearance of stage-specific Sertoli cell immunoexpression of cyclin D2, whereas no Sertoli cell immunoexpression was detectable in control (wild-type) adult mice. The latter finding concurs with an earlier finding in mice [29], though this study also reported intense immunoexpression of cyclin D2 in spermatogonia and in some spermatocytes (using the same antibody as one of the two that we used presently); nevertheless, we observed immunoexpression of cyclin D2 in the occasional spermatogonia in both wild-type and SCARKO adult testes in the present studies, and this is in agreement with that reported by Beumer and colleagues [29]. The difference in spermatocyte staining is possibly explained by the fact that the earlier study used antigen retrieval, whereas we did not, as we had found that this caused nonspecific background staining in our mouse testis sections, although it did not appear to alter the pattern of cyclin D2 immunoexpression in Sertoli cells without antigen retrieval (unpublished data). Without antigen retrieval, cyclin D2 immunoexpression in mouse germ cells may be masked because of the binding of other proteins, such as cyclin-dependent kinases (cdks), to the antibody epitope of cyclin D2, although this is clearly not the case for adult Sertoli cells in SCARKO mice, in which its function is unrelated to cell division. However, we did use antigen retrieval for rat testis sections and never observed cyclin D2 immunoexpression in either spermatogonia or spermatocytes, as has also been reported for the human testis [31]. This may indicate a species difference between mice compared with rats and humans. While we are unable to explain the differences in germ cell immunoexpression of cyclin D2 between the present and the earlier [29] mouse study (see also the following discussion), there is unanimity in showing that Sertoli cell immunoexpression of cyclin D2 is virtually nondetectable in the normal adult mouse, rat, and human testis. However, once androgen action on Sertoli cells is blocked, regardless of whether this is accomplished via transgenesis, EDS treatment, or flutamide administration, Sertoli cell immunoexpression of cyclin D2 becomes evident in a stage-specific pattern. This implies first that immunoexpression of cyclin D2 is dependent on androgen signaling and, second, that cyclin D2 may itself play a role in regulating stage-dependent expression of the AR and/or its signaling ability.
The treatment- and stage-dependent pattern of cyclin D2 immunoexpression in Sertoli cells of EDS-treated rats was robust, with identical results for immunoexpression of cyclin D2 in the present studies using two different antibodies raised in different species, and we could demonstrate that Sertoli cell immunoexpression could be blocked by preabsorption with the immunizing peptide. Furthermore, stage-specific immunoexpression of cyclin D2 was induced selectively in Sertoli cell nuclei in every instance in which androgen action on adult Sertoli cells was experimentally reduced, regardless of the method or species used (see the previous discussion), and these findings were corroborated by semiquantitative PCR of staged seminiferous tubules from control and treated adult rats. Although the general consensus of the role of cyclin D2 is in G1 to S phase progression in the cell cycle, this unusual discovery of its up-regulation of expression in situations of testosterone deprivation in adult male rats must surely denote a role in androgen regulation that is independent of cell proliferation.
We sought to obtain supporting evidence for a possible androgen-regulated pathway by which cyclin D2 regulates spermatogenesis by examining the expression of an immediate downstream target of cyclin D2, namely, RB1. In cell cycle control, RB1 is a transcriptional repressor of genes required for progression in S-phase and is inactivated via phosphorylation by cyclin D-cdk4/6 complexes. Like Yan et al. [19], who used a phospho-specific RB1 antibody specific for RB1332-344 (their immunohistochemistry results were described but not shown), we have shown that in control animals, there is strong phospho-RB1300-380 expression in spermatogonia at stages III–IV and X–XI (not affected by androgen deprivation) and also stage-specific Sertoli cell nuclear immunostaining that peaks at stages VII–VIII, declines rapidly in stages IX–XIV, and then rises slowly at stages I–V; this pattern of immunoexpression is very similar to that for the AR [6]. The stage-specific pattern of Sertoli cell nuclear RB1 staining was attenuated after androgen deprivation by either EDS treatment or flutamide administration and could be restored by testosterone replacement in EDS-treated animals.
In spermatogonia, we detected phosphorylation of RB1 at Ser780, Ser 807/811 (not shown) in addition to amino acid residues between 300 and 380, while other later germ cells appeared to only phosphorylate Ser780 of RB1. Perhaps phosphorylation of RB1 at Ser780 and Ser807/811 by cyclin D-cdk4/6 allows cell cycle progression in spermatogonia, while phosphorylation of RB1 residues between 300 and 380 in adult Sertoli cells does not. However, phosphorylation of RB1300-380 may be sufficient to relieve repression of androgen-regulated genes selectively without affecting repression of cell cycle genes in Sertoli cells. Indeed, it is known that phosphorylation of RB1 on different residues inactivates different molecular functions [43], and full phosphorylation of RB1 may require the action of more than one cyclin-cdk complex, whereas differential phosphorylation by different cdks might translate into selective regulation of downstream effector pathways [44]. We therefore postulate that loss of stage-dependent phosphorylation of RB1300-380 in Sertoli cells in the absence of androgens may result in dysregulation in transcriptional repression of Sertoli cell genes that are essential for supporting spermatogenesis at specific stages, especially at stages IX–XIV.
The present results demonstrate a difference in Sertoli cell nuclear immunoexpression of cyclin D2 in the presence or absence of androgens as well as a change in absolute levels of mRNA transcripts levels. We consider that this robust, stage-specific up-regulation in cyclin D2 expression induced by blockade of androgen action points to a role for cyclin D2 in stage-specific Sertoli cell functions that are regulated by androgens and AR signaling. This role is clearly unrelated to cell division, as cyclin D2 immunoexpression was only detected after terminal differentiation of Sertoli cells, but may be responsible in adulthood for the stage-specific pattern of RB1 phosphorylation at residues 300-380 via cyclin D2-cdk action, which in turn could affect the activity of RB1 as a transcriptional repressor of androgen-regulated genes. If this is the case, identification of the androgen-regulated, non-cell-cycle-related genes that RB1 is able to repress transcription of may provide much needed insight into the mechanisms via which androgens regulate spermatogenesis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence. FAX: 44 131 242 6231; r.sharpe{at}hrsu.mrc.ac.uk
3 Current address: MRL Department of Safety Assessment, Merck & Co Inc, WP45-120 West Point, PA 19486
Received: 5 November 2004.
First decision: 23 November 2004.
Accepted: 10 January 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. H. Tahtamouni, N. H. Mustafa, A. A. Alfaouri, I. M Hassan, M. Y. Abdalla, and S. R. Yasin Prevalence and risk factors for anabolic-androgenic steroid abuse among Jordanian collegiate students and athletes Eur J Public Health, July 4, 2008; (2008) ckn062v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P J O'Shaughnessy, L Hu, and P J Baker Effect of germ cell depletion on levels of specific mRNA transcripts in mouse Sertoli cells and Leydig cells Reproduction, June 1, 2008; 135(6): 839 - 850. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Domanskyi, F.-P. Zhang, M. Nurmio, J. J. Palvimo, J. Toppari, and O. A. Janne Expression and localization of androgen receptor-interacting protein-4 in the testis Am J Physiol Endocrinol Metab, February 1, 2007; 292(2): E513 - E522. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. F. Sneddon, N. Walther, and P. T. K. Saunders Expression of Androgen and Estrogen Receptors in Sertoli Cells: Studies Using the Mouse SK11 Cell Line Endocrinology, December 1, 2005; 146(12): 5304 - 5312. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
