| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Testis |
Biology of Aging Laboratory,3
Department of Geriatrics, Department of Gynecology and Obstetrics,4 University Hospitals, CH-12225 Geneva, Switzerland
INSERM-INRA U 418,5 Hôpital Debrousse, Lyon, 69322 Cedex 05, France
Institut de Biologie,6 INSERM, U419 Nantes, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
apoptosis, BARD1, BRCA1, FSH, infertility, meiosis, spermatogenesis, splicing, steroid hormones, testosterone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the adult rat, the rate of spontaneous apoptosis extends to 75% and occurs mostly in spermatogonia [11–13]. Apoptotic signaling pathways lead to caspase 3 activation and involve Bax and Apaf1 or the FasL/Fas and FADD [14]. Expression of p53 was also reported in spermatogonia and early meiotic prophase cells associated with DNase I expression and commitment for apoptosis [15]. The upregulation of p53 might either lead to the repair of DNA by inducing cell cycle arrest or apoptosis by activating the proapoptotic gene Bax, a pathway well described in spermatogenesis [16].
BARD1, a protein that interacts with p53 in vitro [17], originally identified as binding partner for BRCA1 [18], is also implicated in the apoptotic pathway involving p53 [17]. The expression of BARD1 is highly elevated in testis [19, 20], suggesting a role of BARD1 in spermatogenesis.
The mRNAs of BRCA1 and BARD1 are quite abundant in preparations from whole testis [19, 20], and an additional 2 kilobase (kb) smaller BARD1 transcript of 2 kb is found in testis, but not in other tissues, by Northern blotting. Measurements of BRCA1 mRNA levels in purified somatic cells of the testis and in staged germ cells showed that high-level BRCA1 mRNA expression is limited to the germ cell. Within the germ cell lineage, the high expression was detected in meiotic cells, specifically pachytene spermatocytes and in postmeiotic round spermatids. In contrast, little or no BRCA1 mRNA was found in premeiotic germ cells [21]. In situ hybridization localized BRCA1 mRNA to early meiotic prophase spermatocytes [22]. A partial understanding of BRCA1 function in meiosis comes from studies on meiotic chromosomes where BRCA1 localizes to recombination nodules [23] and the finding that Brca1 –/– mice lack crossover during the pachytene stage and fail to enter diplotene [24].
Although exhaustive studies on the localization and function of BARD1 and BRCA1 proteins during spermatogenesis are not available, portions of the BRCA1 and BARD1 proteins localize to the XIST RNA on the inactive X-chromosome during the pachytene stage [25] and suggest a function in the maintenance of its repressed state. In this present study, we intend to elucidate the function of BARD1 in spermatogenesis, including its potential role in apoptosis suggested by earlier studies on BARD1 and BRCA1.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Two adult male Wistar rats were used in each part of the study. Animals were anesthetized, the tunica albuginea of the right testis was incised from the proximal to the distal pole, and the parenchyma was immersion fixed with 10% formaldehyde in 0.1 M phosphate buffer, pH 7.4. The thoracic aorta was then cannulated and the vasculature flushed with buffered sodium chloride before perfusion fixation of the left and the right testes with 2% glutaraldehyde and 2% paraformaldehyde in 0.05 M phosphate buffer, pH 7.4. Procedures relating to the care and use of animals were approved by the Swiss cantonal veterinary office.
Immunohistochemistry
The immersion-fixed right and left testes were processed for paraffin embedding. Consecutive orthogonal sections (5 µm) across the seminiferous tubules were mounted on polylysine-coated slides. The first section of each set of sections was stained with hematoxylin eosin and used to determine tubule stages. The remaining sections were further processed for immunocytochemical staining. The endogenous peroxidase was quenched by 15 min of incubation in 2% hydrogen peroxide in PBS. Unmasking of the epitope was carried out by boiling deparaffinized rehydrated sections (twice for 5 min) in 10 mM citrate buffer, pH 6.0, using a microwave oven at 5000 W power output. Primary antibodies were as follows: anti-BARD1 N19 polyclonal antibody (1:20) directed to N-terminus of BARD1 (Santa Cruz); anti-BARD1 C20 polyclonal antibody (1: 20), directed to carboxyl-terminus of BARD1 (Santa Cruz); anti-BARD1 JH3 polyclonal antibody (1:20), directed to the amino acids 527–540; anti-BRCA1 D16 (1:50), epitope corresponding to the amino-terminus; anti-BRCA1 M20 (1:20), directed to the carboxyl terminus; anti-p53 (1:20) (Santa Cruz), epitope corresponding to amino acids 1–393. Secondary antibodies were peroxidase coupled and negative controls were processed in an identical manner except that the primary antibody was replaced by PBS.
Generation of Polyclonal Anti-BARD1 Antibody JH-3
Peptides were generated corresponding to the region including amino acids 527–540 and polyclonal antibodies JH-3 were generated in rabbits. Whole sera were used for Western blotting and immunohistochemistry.
Reverse Transcription-Polymerase Chain Reaction
Total RNA was purified from physically isolated rat germ cells of different stages using Qiagen reagent. RNA was reverse transcribed using oligo-dT and superscript II (Life Technologies). BARD1 primers used were: 56 forward (ccatggaaccagctaccg) and 2307 reverse (tcagctgtcaagaggaagca). GAPDH primers were 176 forward (CTACCCACGGCAAGTTCAAT) and 557 reverse (ACTGTGGTCATGAGCCCTTC). Reverse transcription (RT) products were subjected to PCR analysis using specific primers for rat BARD1 cDNA (covering BARD1 coding regions). PCR cycles were: 94°C 15 min, 94°C 1 min, 56°C 1 min, 72°C 2 min 30 (35 x), then 72°C 10 min. Equal volumes of PCR products were analyzed on 1% agarose gel.
Western Blot
BARD1 expression was monitored by Western blot analysis. Germ cells harvested by elutriation were lysed in SDS loading buffer (4% SDS, 10% glycerol, 4% 2-mercaptoethanol, 0.125 M Tris base, and 0.02% bromophenol blue, pH 6.8) containing protease inhibitors. Protein concentrations were determined using Bio-Rad protein analysis solution (Bio-Rad, Hercules, CA). Protein samples (40 µg) were size separated on 10% discontinuous polyacrylamide gels and transferred overnight to polyvinylidene fluoride membranes. Filters were blocked for 1 h with 5% nonfat milk and then incubated for 2 h with polyclonal BARD1 antibodies (C20: 1:200; JH3:1:1000). N-19, which was used in immunohistochemistry, did not work on Western blots. After washing, membranes were incubated with horseradish-peroxidase-conjugated anti-rabbit antibodies for JH3 conjugated anti-goat antibodies for C20 (1:1000), and the resulting complexes were visualized by enhanced chemiluminescence autoradiography.
Preparation of Sertoli Cell Conditioned Media
Sertoli cells were plated in tissue culture flasks at a density of 2 x 105 cells/cm2 and were cultured for 3 days in HEPES-buffered F12/DMEM supplemented with insulin (10 µg/ml), transferrin (10 µg/ml), vitamin C (10–4 M), vitamin E (10 µg/ml), retinoic acid (3.3 x 10–7 M), retinol (3.3 x 10–7 M), pyruvate (1 mM) (all products from Sigma, La Verpillière, France), 0.2% fetal calf serum (Life Technologies, Cergy-Pontoise, France) in the absence or presence of 10–7 M testosterone (Sigma) and/or 1 ng/ ml bovine NIH FSH-20 obtained through NIDDK (lot no. AFP-7028D). On Day 3, media were replaced by serum-free media supplemented or not with hormones. On Day 5, culture media were recovered, centrifuged to eliminate cell debris, and conserved at –20°C until used. Four Sertoli cell conditioned media (SCCM) were prepared: control (no FSH, no testosterone [F–/T–]), FSH (1 ng/ml) (F+/T–), testosterone 10–7 M (F–/T+), and a combination of both hormones (F+/T+).
Isolation of Rat Sertoli Cells, Pachytene Spermatocytes, Round Spermatids, and Spermatogonia/Preleptotene Spermatocyte Fractions
Sertoli cells from 20-day-old rats and pachytene spermatocytes and round spermatids from 90- to 120-day-old rats were isolated as described previously [26]; the purity of these different fractions was assessed by flow cytometry. In Sertoli cell fractions, 77% ± 1% of the isolated cells were 2C cells, 5.5% ± 1% were 4C cells; in pachytene spermatocyte fractions, 94% ± 3% of cells were 4C cells, 3% ± 2% were 2C cells, and 1% ± 0.5% were 1C cells; in round spermatid fractions, 67.2% ± 5.2% were 1C cells, 10.6% ± 6.6% were 4C cells, and 9.1% ± 1.6% were 2C cells. Procedures relating to the care and use of animals were approved by the French Ministry of Agriculture according to the French regulations for animal experimentation.
Spermatogonia/preleptotene spermatocytes were prepared by centrifugal elutriation of germ cells isolated from 21- to 23-day-old rats, which do not yet produce round spermatids in their testes. These spermatogonia/ preleptotene spermatocytes eluted in the fraction of elutriation in which round spermatids from mature rats are recovered. In this fraction, 64.8% ± 7.3% of cells were 2C cells, 29.1% ± 5.4% were 4C cells, and 5.3% ± 2.8% were 1C cells.
Incubation of Germ Cells in Sertoli Cells Conditioned Media
Pachytene spermatocytes, round spermatids, spermatogonia/preleptotene spermatocytes were incubated for 24 h in the different conditioned media. At the end of the incubation period, the cells were centrifuged and recovered for RNA extraction.
Generation of BARD1ß Expression Clone
BARD1ß expression clone was produced by cloning the sequence of BARD1ß PCR product cDNA into pcDNA3-EGFP in frame with EGFP.
Transfections were performed reproducibly with Effecten (Qiagen). One to 4 µg of DNA was transfected per well of mouse embryonic stem cells, CGR8 [17]. Transfection efficiency was between 20% and 30% as assayed by parallel transfection of GFP cloned into pcDNA3 and analyzed by flow cytometry and fluorescent microscopy. Thirty-six hours after transfection, cells were harvested and apoptosis was monitored by 7-AAD viability probe (BD Pharmingen) and TUNEL assay (Molecular Probes) by flow cytometry (DAKO galaxy pro).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
BARD1 Expression Is Not Correlated with BRCA1 Expression
To determine the cell type and stage-specific expression of BARD1, we performed immunohistochemistry on sections of testis from adult rats. Interestingly, in the testis of young adult rats (23 days) BARD1 protein is expressed in many, but not all, spermatogonia and preleptotene spermatocytes, and it is less abundant during later stages of spermatogenesis (Fig. 1A). Staining was also observed in some cells at later stages of meiosis and in round spermatids (Fig. 1A). During early stages, BARD1 staining is localized to the nucleus, but it is also found in the cytoplasmic prolongations of Sertoli cells. The same staining could be observed when different antibodies against the N-terminal region of BARD1 were used. No staining was observed when the primary BARD1-specific antibody was omitted (Fig. 1C). However, when JH-3, an antibody directed against the middle region of BARD1 was used, spermatogonia showed strong staining but also later stage germ cells were stained.
|
Because it could be expected that BARD1 and BRCA1 act in concert during spermatogenesis, we determined the specific stages at which the BRCA1 protein is expressed by immunohistochemistry with anti-BRCA1 antibodies. Consistent with previous observations of BRCA1 mRNA expression, the BRCA1 protein was predominantly detected at the pachytene stage and rarely in preleptotene spermatocytes or spermatogonia (Fig. 1B). These data demonstrate that BRCA1 expression is mostly limited to the pachytene stage, while BARD1 expression is found in spermatogonia and preleptotene spermatocytes. The elevated expression of BARD1 but not BRCA1 in spermatogonia and primary spermatocytes is indicative of a BRCA1-independent function of BARD1 in spermatogenesis.
Stage-Specific Isoforms of BARD1
To characterize the expression pattern of BARD1 further, isolated cells from different stages of spermatogenesis were analyzed on Western blots. Two forms of BARD1, corresponding to a molecular mass of 97 and 74 kDa, were found on Western blots probed with different antibodies (Fig. 2A). Spermatogonia were prepared from 9-day-old rats, at which age spermatogenesis is not induced [27], and spermatogonia but not primary spermatocytes are present. While in spermatogonia from very young rats (9 days), the 97-kDa BARD1 was expressed, it was less abundant in preleptotene spermatocytes from young rats (23 days). The 74-kDa protein was found in pachytene spermatocytes of young (23 days) and adult rats (3 mo), as well as in round spermatids. Depending on which antibody was used for detection, the 97-kDa and 74-kDa proteins could be detected at the preleptotene stage. This protein expression pattern suggests that a shorter isoform of BARD1 is expressed at all stages with the exception of spermatogonia. The 97-kDa BARD1 was also found in Sertoli cell preparations but might be derived from BARD1 staining of residual bodies phagocytosed by the Sertoli cells (data not shown). Similarly, p53 and Bax immunohistochemical staining on phagocytosed residual apoptotic bodies has been reported previously [15, 28].
|
Stage-Specific BARD1 mRNA Expression
To determine whether stage-specific isoforms of BARD1 were generated at the protein or the mRNA level, we investigated BARD1 mRNA expression at different stages of spermatogenesis. For this purpose, rat testes were dissected and differentially-staged germ cells were fractionated [26]. Total RNA was prepared from each of the isolated cell types and RT-PCR was performed to monitor BARD1 mRNA expression profiles. Primers were designed for amplification of the entire BARD1 coding region. While control RT-PCR of the GAPDH shows similar expression levels in all cell types tested, BARD1 mRNA was found in spermatogonia and less abundantly in primary spermatocytes and round spermatids. Interestingly, RT-PCR produced two products of slightly different size (Fig. 2B). The primers corresponding to regions of the translation initiation and stop codon were expected to amplify a 2251-base pair (bp) fragment, corresponding to the coding region of wild-type BARD1. This was only found in spermatogonia and was less pronounced in preleptotene stages. A fragment of approximately 2000 bp was amplified in preleptotene and pachytene spermatocytes as well as in round spermatids (Fig. 2B). As observed on the protein level, on the transcript level, two different mRNAs are transcribed in a stage-specific fashion, full-length BARD1 in spermatogonia and a shorter form in all stages except spermatogonia. In preleptotene stages, both BARD1 mRNAs and both proteins of 97 and 74 kDa are expressed. This might reflect the purification procedure, which does not permit the separation of spermatogonia and preleptotene spermatocytes.
A Testis-Specific Isoform of BARD1 Is Generated by Differential Splicing
We cloned and sequenced the RT-PCR fragments corresponding to the 2.2-kb and the approximately 2000-bp fragments. The 2.2-kb fragment was identified as 2251 bp of full-length rat BARD1 cDNA [29]. The 2000-bp fragment encoded a 1964-bp sequence identical to the known rat BARD1 sequence but missing the regions including nucleotides 143–348 and nucleotides 1287–1367. The deleted sequences are homologous to exons 2 and 3, and 5, respectively, of the corresponding human BARD1 sequence (Fig. 3A). To determine whether the equivalent testis-specific isoform, called BARD1ß hereafter, is unique to the rat, RT-PCR was confirmed on RNA isolated from mouse testis. Cloning and sequencing RT-PCR products showed that a similar splice variant exists in the mouse. The abundance of mouse BARD1ß when cloned from total RNA from mouse testis was fourfold lower that full-length BARD1 cDNA. Our data suggest that intron-exon boundaries are conserved between mouse and human sequences and that the same intron-exon boundaries are used in human, mouse, and rat.
|
The missing BARD1 exons 2 and 3 encode part of the RING-finger structure, and exon 5 encodes one of the ankyrin repeats. However, deletion of nucleotides 143–348 leads to a frame shift that would result in translation of 46 amino acid peptide before ending with a stop codon (Fig. 3A). Alternatively, ATG at position 389, 42 nucleotides downstream of the deletion, could serve as translation initiation codon. Translation initiation at nucleotide 389 would result in a protein of approximately 70 kDa, which is consistent with the observed 74-kDa protein detected in preleptotene and pachytene spermatocytes and round spermatids. We used a deletion construct D-RING-EGFP (nucleotides 1–340), designed to be translated from the ATG at position 389 and producing a truncated BARD1-GFP fusion protein [30] and compared its expression with BARD1ß fused to GFP. Both constructs expressed a truncated BARD1-EGFP fusion protein of approximately 97 kDa, which can be detected with antibodies directed against GFP (Fig. 3B), indicating that ATG at position 389 is a functional translation initiation codon in vitro. This supports the assumption that wild-type BARD1 is expressed in spermatogonia and the protein product of the splice variant BARD1ß, schematically presented in Figure 3C, corresponds to the 74-kDa protein expressed in pachytene spermatocytes and round spermatids.
BARD1 Expression Is Correlated with Apoptosis
The stage-specific pattern of BARD1 expression during spermatogenesis was reminiscent of the reported occurrence of apoptosis [14]. To confirm this correlation, we performed TUNEL experiments and BARD1 staining on serial sections from testes of a young adult rat. TUNEL assays primarily produced positive staining in spermatogonia and preleptotene spermatocytes coinciding specifically with anti-BARD1 staining when using antibody N-19 (Fig. 4A). TUNEL staining was also observed at later stages of spermatogenesis, in round spermatids, coinciding with BARD1 staining. Only a few TUNEL positive cells could be detected at the pachytene stage. Therefore, the spatial and developmental distribution of BARD1 staining coincides with apoptosis at all stages of spermatogenesis.
|
To further demonstrate the correlation between BARD1 expression and apoptosis induction, we also tested the coexpression of BARD1 with other proteins involved in apoptosis pathways, such as increased p53, Bax, and activated caspase 3. Staining with BARD1 antibodies against the amino-terminus colocalized with cells positive for activated caspase 3, tested on adjacent sections, mostly in spermatogonia (Fig. 4B), while expression of activated caspase 3 was also observed at later stages. When using an antibody against the middle region of BARD1, JH-3, staining was observed in spermatogonia and later stages of spermatogenesis (Fig. 4B). Similar staining was observed for the proapoptotic protein Bax. At these stages, the expression of BARD1 was primarily cytoplasmic, which is consistent with an apoptotic function, as reported recently [30, 31]. The increase of a C-terminal but not N-terminal epitope at the pachytene stage and postmeiotic stages is consistent with the expression of a new splice form BARD1ß at those stages and suggests that BARD1 and BARD1ß expression is associated with apoptotic events at all stages of spermatogenesis.
BARD1ß, a Potent Apoptosis Inducer
To test whether BARD1ß is capable of apoptosis induction, the BARD1ß sequence was cloned into pcDNA and was transiently transfected into mouse embryonic stem cells, which were previously reported to readily induce apoptosis in response to exogenous expression of BARD1 [17]. The testis-specific splice variant BARD1b induced apoptosis in vitro as well as the full-length BARD1 sequence, pcBARD1. Apoptosis was quantified by vital staining of living cells with 7'AAD and analyzed by flow cytometry (Fig. 5A). Interestingly, the RING-finger deficient splice form BARD1ß is as efficient in apoptosis induction as BARD1. This apoptotic capacity could be due to the lack of the RING finger, thus liberating BARD1 from apoptosis-competing binding to BRCA1, which was found previously [17]. Indeed, cotransfection of BARD1ß with BRCA1 did not lead to a reduction of apoptosis induction, while cotransfection of BRCA1 with BARD1 decreased apoptotic capacity of BARD1 (Fig. 5, A and B). Apoptosis induction by BARD1ß could be due to increased protein stability of BARD1ß, as compared with full-length BARD1, because an N-terminal deletion renders BARD1 more stable [30].
|
Apoptosis induction by BARD1 is dependent on a functional p53 [17]; we therefore tested whether BARD1ß expression affected p53 stability. Indeed, p53 is upregulated in cells that are transfected with BARD1 and even more so in cells that are transfected with BARD1ß (Fig. 5C). From these data, we concluded that BARD1ß, like BARD1, induces apoptosis when overexpressed in vitro, through stabilization of p53. It is likely that BARD1ß might have a similar function in vivo.
BARD1 Expression Is Implicated in Pathologies of Defective Spermatogenesis
Several pathologies are known that result in infertility due to abortive spermatogenesis, such as obstructive azoospermia, Sertoli cell syndrome, and cryptorchidy. Those diseases are marked by increased levels of apoptosis of germ cells at different stages of spermatogenesis. We have analyzed the expression of BARD1 in sections of human testis samples of those pathologies. An upregulation of BARD1 expression is found in round spermatids and in apoptotic germ cells phagocytosed by the Sertoli cells on sections from patients suffering from azoospermia (Fig. 6A). The expression of BARD1 clearly coincides with apoptosis, as can be determined by the formation of apoptotic bodies. Cryptorchidism in humans is known to prevent all spermatogenesis in the undescended testis and germ cell development is inhibited. We investigated the expression of BARD1 in the cryptorchid testis and the contralateral descended testis of the same individual (Fig. 6B). Within the descended testis, heavy BARD1 staining could be observed, and increased apoptosis and BARD1 expression was clearly associated with apoptotic cells and phagocytosed apoptotic bodies within the Sertoli cells. This is consistent with the reported reduced fertility of individuals with cryptorchidy. In the cryptorchid testis, which contains only Sertoli cells, no BARD1 staining could be detected (Fig. 6, B, a and b), indicating that BARD1 is not expressed in Sertoli cells per se. These data are consistent with our finding that BARD1 mRNA is not expressed in Sertoli cells and supports our interpretation that BARD1 antibody staining in Sertoli cells is derived from the incorporation of apoptotic germ cells.
|
BARD1 expression therefore is clearly associated with germ cell apoptosis in vivo and with apoptosis in pathologies marked by reduced spermatogenesis.
Regulation of BARD1 Transcription
Since BARD1 is associated with developmental and pathological events of apoptosis, it was interesting to investigate the conditions that drive BARD1 expression. We have observed that BARD1 is expressed at all stages of spermatogenesis, although to a different extent. We hypothesized that BARD1 expression is linked to the presence or absence of hormones driving sexual maturation [32]. In particular, FSH and testosterone have been described sufficient for triggering spermatogenesis in vitro [33, 34]. This implies that BARD1 expression is induced in spermatogonia and primary spermatocytes through the paracrine signaling of the Sertoli cells. To test this hypothesis, we cultured isolated germ cells, specifically spermatogonia/preleptotene stage, pachytene, and round spermatids, either in normal culture medium, as defined previously [26], or in conditioned culture medium, obtained from Sertoli cells that were preincubated with hormones, to mimic in vivo paracrine signaling and maturation. FSH and testosterone were added to the culture medium. The response of BARD1 to hormone exposure, FSH or testosterone or both, was monitored on the mRNA level. RT-PCR was performed to detect BARD1 mRNA expression in RNA isolated from the different cultures (Fig. 7).
|
Interestingly, in spermatogonia/preleptotene germ cells, no reduction of expression was observed with the addition of testosterone. FSH, however, led to a complete repression of BARD1, but the combination of FSH and testosterone had no effect, suggesting that testosterone compensated the effect of FSH. In all conditions, two forms of BARD1 mRNA were detected that can be correlated with the wild-type full-length BARD1 and with BARD1b, respectively, when primers amplifying regions including nucleotides 56– 2307 were used (Fig. 7A). At the pachytene stage, the addition of hormones did not result in a variation of the expression level of BARD1. At the round spermatid stage, addition of hormones produced a similar repression pattern, as observed at the spermatogonia/preleptotene stage, but testosterone had no inhibitory effect on FSH-induced repression of BARD1. Both forms of BARD1, full length and BARD1ß, were obtained, although BARD1ß was increased proportionally. These results indicate that FSH acts as a transcriptional repressor of BARD1 in spermatogonia/primary spermatocytes and round spermatids but has no effect on pachytene stage spermatocytes. Interestingly, however, testosterone, by itself, does not have any influence on BARD1 mRNA expression at any of the stages tested, but in combination with FSH, it annihilates the repressive effect of FSH.
In contrast, when cells were grown in preconditioned medium, medium from Sertoli cells that had been exposed to the hormones, the addition of FSH and FSH plus testosterone, but not testosterone alone, resulted in suppression of BARD1 expression in spermatogonia/preleptotene spermatocytes and in round spermatids. No effect of either hormone was observed under these conditions in pachytene spermatocytes (Fig. 7A). Interestingly, testosterone alone, when applied through Sertoli cell-conditioned media, led to BARD1 repression in round spermatids. Furthermore, testosterone had no compensatory effect on BARD1 repression by FSH, but even a repressive effect by itself in round spermatids. These data indicate that FSH, when applied to Sertoli cells, could act as a messenger for BARD1 repression at the spermatogonia/preleptotene stage and in round spermatids.
To test whether other genes, the products of which presumably interact with BARD1, were also hormonally regulated in this in vitro system, we monitored the expression of p53 mRNAs (Fig. 7B). The expression of p53 showed only minor, insignificant changes under conditions of FSH and/or testosterone addition. No changes were observed for the control gene ß-tubulin, which remained unchanged in the various conditions.
In summary, our data suggest that, at the preleptotene and round spermatid stages, BARD1 is directly or indirectly repressed by FSH through paracrine signaling of the Sertoli cells. The same treatment affects only mild repression of BRCA1 and only in preleptotene cells (data not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
BARD1 knockouts have been generated but are lethal during the early stages of embryogenesis and thus cannot shed light on the function of BARD1 during spermatogenesis. We therefore performed experiments to further dissect the specific local and temporal expression pattern of BARD1 on the mRNA and the protein level. We found that a novel isoform of BARD1, generated by differential splicing, was expressed in the testis in addition to wild-type BARD1. The testis-specific new splice variant, BARD1ß, shows a deletion of exons 2, 3, and 5. The translation of BARD1ß results in a frame shift after excision of exons 2 and 3 and would end with a stop codon (Fig. 3A). However, downstream of the splice acceptor site, an ATG at position 389 could act as an alternative translation initiation site. Western blot analysis is consistent with this site of translation initiation because the proteins detected with anti-BARD1 antibodies correspond to the expected size for the presumed translation product of BARD1ß. Expression of a deletion construct, missing regions 1–380, results in a protein of similar size as translation of BARD1ß. It is therefore likely that the 74-kDa protein, expressed at all stages of spermatogenesis with the exception of spermatogonia and detected with antibodies against the C-terminal portion of BARD1, is a translation product of the differentially spliced form BARD1ß.
The temporal expression patterns of BARD1 mRNAs and BARD1 proteins isoforms show a similar distribution, a 97-kDa protein corresponds to full-length BARD1 and is expressed in spermatogonia, as is the full-length BARD1 transcript. BARD1ß mRNA is expressed at all stages except in spermatogonia, as is a 74-kDa protein, which is recognized by C-terminal antibodies, suggesting that the 74-kDa is derived from differential splicing rather than from posttranslational modification of BARD1. Interestingly, the presumed protein product of BARD1ß lacks the RING-finger domain, required for the interaction of BARD1 with BRCA1 [18, 35], but retains regions sufficient for its apoptotic function [30].
Expression of BARD1ß in vitro resulted in similar levels of apoptosis as BARD1. However, the in vivo apoptosis induced by BARD1ß could be more efficient due to increased stability of BARD1ß compared with full-length BARD1, as determined on Western blot analysis using antibodies against different regions of BARD1 (data not shown) and due to the lack of BRCA1 interaction domain. BRCA1 could act as an inhibitor of BARD1-induced apoptosis [17]; therefore, we demonstrate that BRCA1 reduces BARD1-induced apoptosis but has no affect on the apoptotic capacity of BARD1ß (Fig. 5, A and B). BARD1b not only induces apoptosis but also shows p53 stabilization, as does wild-type BARD1 (Fig. 5C). Therefore, BARD1ß acts in a p53-dependent apoptosis pathway. It could be speculated that BARD1ß acts as apoptosis inducer in stages where BRCA1 and BARD1 are coexpressed.
This hypothesis stipulates that, during premeiotic stages, in the absence of BRCA1, full-length BARD1 proteins are expressed and presumably are involved in apoptotic functions. To ensure apoptosis induction by BARD1 in the presence of the competitor BRCA1 [17], the testis-specific splice variant BARD1ß is expressed.
Besides the importance of apoptosis in normal spermatogenesis, the loss of germ cell homeostasis, or deregulated apoptosis, can result in infertility. In line with these thoughts, BARD1 is associated with pathologies of reduced germ cell maturation due to increased apoptosis. Loss or reduction of apoptosis, on the other hand, can lead to uncontrolled proliferation and it could be speculated that BARD1 plays a role as tumor suppressor in germ cell cancers.
Male germ cell apoptosis and development is regulated by different growth-promoting or -repressing signals, one of the regulators being hormones. Testosterone deprivation leads to germ cell apoptosis [5] and estradiol increases germ cell survival [36, 37]. We speculated that BARD1 expression was hormonally regulated because BARD1 mutations are implicated in female gynecological cancers [38, 39] and its expression pattern in ovary and uterine tissue reflected a hormonal modulation [19]. Here we report hormonally regulated transcriptional activation or repression of BARD1. Interestingly, BARD1 transcription is repressed by the addition of FSH or FSH and testosterone to isolated germ cells, but not testosterone alone (Fig. 7A). But either FSH or testosterone are capable of transcriptional repression of BARD1 when added in the presence of Sertoli cells. Sertoli cells can convert testosterone to estradiol in the presence of FSH, and estradiol is required for progression of spermatogenesis [40]. Our data therefore suggest that the repressive effect on BARD1 transcription is due to estradiol rather than testosterone.
In pachytene spermatocytes, no effect is observed by either hormones added, neither directly nor as conditioned medium of Sertoli cells. In round spermatids, however, a similar response is observed as in spermatogonia/preleptotene spermatocytes. It is important to emphasize that the pattern of expression of the two BARD1 splice forms is influenced by additional factors in vivo (Fig. 2), which are not reproduced in vitro (Fig. 7). Interestingly, BRCA1 expression parallels the expression of BARD1. This could be due to similar transcriptional repression of BRCA1 or to BARD1 mRNA or protein affecting BRCA1 transcription, although the latter possibility is unlikely, considering experiments performed with xBARD1 and xBRCA1 [41]. As expected, the expression of p53 mRNA was only slightly, if at all, modulated. Our data are consistent with a model that predicts that BARD1 is repressed by FSH and/or estradiol produced in Sertoli cells but not by testosterone. BARD1 acts as mediator between paracrine signals and apoptosis by transcriptional response to hormone levels, as increased transcript and protein levels of BARD1 lead to increased p53 levels and apoptosis [17]. A pathway of FSH-induced apoptosis to caspase activation and apoptosis induction has been described previously [34]. This would mean that cells in contact with Sertoli cells are more prone to be exposed to the suppressive effect of FSH even in the presence of testosterone, consistent with survival rather than apoptosis and consistent with a limited support function of Sertoli cells.
Although the correlation of BARD1 expression and apoptosis in spermatogenesis is intriguing, it remains to be considered that components of the apoptotic machinery in male germ cells are required for and activated during spermatogenesis without completion of apoptosis, as demonstrated for Drosophila [28]. Whether similar mechanisms exist in mammals remains to be determined.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Irmgard Irminger-Finger, Biology of Aging Laboratory, Department of Geriatrics, University of Geneva, 2 ch. Petit-Bel-Air, 1225 Chene-Bourg/Geneva, Switzerland. FAX: 41 22 305 5455; irmgard.irminger{at}medecine.unige.ch
Received: 15 March 2004.
First decision: 6 April 2004.
Accepted: 25 June 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. A. Thasni, S. Rakesh, G. Rojini, T. Ratheeshkumar, G. Srinivas, and S. Priya Estrogen-dependent cell signaling and apoptosis in BRCA1-blocked BG1 ovarian cancer cells in response to plumbagin and other chemotherapeutic agents Ann. Onc., April 1, 2008; 19(4): 696 - 705. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Li, S. Ryser, E. Dizin, D. Pils, M. Krainer, C. E. Jefford, F. Bertoni, R. Zeillinger, and I. Irminger-Finger Oncogenic BARD1 Isoforms Expressed in Gynecological Cancers Cancer Res., December 15, 2007; 67(24): 11876 - 11885. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Falender, R. N. Freiman, K. G. Geles, K. C. Lo, K. Hwang, D. J. Lamb, P. L. Morris, R. Tjian, and J. S. Richards Maintenance of spermatogenesis requires TAF4b, a gonad-specific subunit of TFIID Genes & Dev., April 1, 2005; 19(7): 794 - 803. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RING-EGFP (97 kDa), and BARD1ß-EGFP (97 kDa) and analysis on Western blots probed with anti-GFP. Protein of identical size is detected after transfection of delta-RING-EGFP, and BARD1ß-EGFP. C) Schematic presentation of BARD1 and presumed BARD1ß protein. BARD1 is presented with RING finger (green), ankyrin repeats (blue), and BRCT repeats (red), as well as approximate location of BARD1 antibodies. BARD1b is presented with presumed location of introns (triangles), translated region (filled heavy line), transcribed region (thin line), and deleted regions (dotted lines). Epitopes recognized by particular antibodies (
) are indicated
