| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Cell and Developmental Biology, Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Científicas (CSIC), Madrid, Spain
ABSTRACT
In the seminiferous epithelium, numerous cell interactions between Sertoli cells and Sertoli-germ cells are established by specialized proteins so as to maintain the functionality of the testis. Exogenous estrogen exposure can result in alterations to these interactions and cause pathologies, including impaired spermatogenesis and tumorigenesis. In the present study, with the aim of finding markers of the action of estrogenic compounds in the mammalian testis, we focused on investigating molecules that are linked to cellular junctions. We found that the testicular vinexin (sorbin and SH3 domain-containing protein 3, encoded by the Sorbs3 gene) pattern underwent significant changes after developmental exposure to 17beta-estradiol (E2). Vinexin is an adaptor protein that is implicated in cell adhesion and actin-cytoskeletal reorganization. We characterized, at the protein and mRNA levels, the expression patterns of vinexin isoforms during testis development and in defined cell types from the seminiferous tubule. The protein expression patterns of vinexin-interacting proteins flotillin 1 and vinculin were also analyzed. Thus, we have identified a novel association between a vinexin isoform and germ cells, which contrasts with the predominant localization of the gamma isoform in Sertoli cells. The effects of E2 on the testes of developmentally exposed mice were evident, with total depletion of the germ-cell-associated vinexin isoform and a noticeable decrease in Sertoli-cell-related vinexin gamma.
estradiol, flotillin 1, gene regulation, Sertoli cells, spermatogenesis, vinculin, vinexin
Estrogen plays different roles in the development and differentiation of spermatogenesis. However, exposure to high levels of estrogen or to compounds acting as estrogens (xenoestrogens) during embryonic and pubertal periods can alter testis development [1]. Alterations in Sertoli-Sertoli and Sertoli-germ cell interactions within the mouse and rat seminiferous epithelium, as a consequence of exposure to exogenous estrogenic compounds, have been reported [2–5], mainly regarding the effects on testis-specific actin-based junctions, ectoplasmic specializations (ES), after estrogenic exposure. Thus, neonatal treatment with 17ß-estradiol (E2) and ß-estradiol 3-benzoate results in altered or deleted apical ES between Sertoli cells and late spermatids [3, 5]. Total or partial deletion of ES provokes the exfoliation of some spermatids and abnormal spermatids and spermatozoa are observed. It has also been reported that the administration to newborn mice of the endocrine disruptor diethylstilbestrol (DES) results in a failure to form basal ES (between Sertoli cells), and a consequent delay in the formation of the blood-testis barrier [2, 4]. Interactions between cells and the extracellular matrix (ECM) in the seminiferous epithelium are necessary for the functioning of the testis. As a result, numerous studies have focused on the highly specialized proteins implicated in adhesion, attachment, and communication between adjacent cells in the seminiferous tubule [6–8].
The purposes of our study were to identify and to characterize, at the molecular level, potential markers of the effects triggered by estrogenic compounds on the exposed mouse testis. Within this prospecting work, we explored proteins that are linked to cell-cell interactions. Thus, the present study includes the analysis of the aqueous-insoluble protein fractions from the testes, in which membrane-associated and cytoskeletal proteins cofractionate in a preferential manner. Using an experimental model in which mice are developmentally exposed to E2, we found a dramatic alteration of the testicular protein pattern of vinexin (encoded by the Sorbs3 gene). Vinexin has been implicated as playing important roles in cell adhesion and actin-cytoskeletal organization [9]. Vinexin belongs to a family of adaptor proteins [10]. The molecular architecture of vinexin-family proteins is highly conserved. All members share a Sorbin Homology (SoHo) domain in the N-terminal region, which has been shown to bind the lipid raft-associated protein flotillin 1, and three Src Homology 3 (SH3) domains in the C-terminal region, which allow the binding of specific signaling and cytoskeletal molecules (including vinculin). To date, three different isoforms of vinexin (
, ß, and 
) have been characterized [9, 11]. Kioka et al. [9] first isolated the and ß isoforms. Both isoforms contain a common C-terminal sequence with three SH3 domains, while the ß isoform does not include the N-terminal SoHo domain. Recently, a new isoform of vinexin, the isoform, has been characterized in mouse fetal gonads [11]. Vinexin differs from the isoform in that only a few amino acids are missing from the N-terminal region.
The expression of the vinexin isoforms is differentially regulated. During embryonic development, vinexin mRNA has been detected by in situ hybridization, mainly in the mouse testis (12.5 days postcoitum). In contrast, vinexin ß is expressed more ubiquitously [12]. Nevertheless, recent analyses by Western blot and RT-PCR have demonstrated the presence of vinexin in the mouse fetal testis (E12.5), whereas vinexin protein was not detected [11], which contradicts the results obtained by Kawauchi et al. [12]. The vinexin and proteins have also been detected in the mouse adult testis [11]. It is interesting to note that Northern blot analysis of human tissues has shown that vinexin expression is highly prevalent in the testis [13]. Taken together, these data suggest important roles for the vinexin isoforms in testis development.
Since vinculin (encoded by the Vcl gene) and flotillin 1 (encoded by the Flot1 gene) have been described as vinexin-interacting proteins, they were investigated as candidate markers of estrogen effects in the mouse testis. Vinculin is an actin-binding cytoskeletal protein that localizes to cell-ECM and cell-cell adhesion sites, with crucial roles in the maintenance and regulation of cell adhesion and migration. Vinexin is able to bind vinculin through its first and second SH3 protein-interacting domains [9]. Vinculin has been detected in the testis [14]. In addition, it has been demonstrated in vitro that flotillin 1 binds vinexin via the SoHo domain in its N-terminal region [15]. Flotillin 1 is a protein that localizes to lipid rafts, that is, specialized microdomains of the plasma membrane [16–18], and has been related to cell interactions [19]. Recently, Evans et al. [20] have described the presence of flotillin 1 protein in Sertoli and germ cells from the rat testis.
To shed light on the relevance of the presence of vinexin, we analyzed the expression patterns of the vinexin isoforms during postnatal testis development and in isolated cell types from the seminiferous epithelium. In addition, the expression patterns of flotillin 1 and vinculin proteins were studied.
All mice were bred and maintained in the Animal Care Facility at the Centro de Investigaciones Biológicas (CSIC) on a 12L:12D cycle. CD1 mice were exposed in vivo to E2 (Sigma-Aldrich), which was administered in the drinking water at a final concentration of 0.16 mg/L (for a calculated daily dosage of 40 µg/kg of body weight). The initial stock solution of E2 was prepared in ethanol.
Mice were exposed to E2 following a defined protocol. Briefly, mothers were exposed 2 wk before mating and exposure was maintained during pregnancy and for 4 wk after birth. All male offspring were killed at 30 days postnatal (dpn). Control animals were treated in the same way but with 0.2% ethanol in place of the E2 solution. Animal care and killing were performed according to the regulations of the CSIC Bioethics Committee.
Isolation of Testicular Cell Types
Sertoli cells were isolated from CD-1 mice as previously described [21], with minor modifications. Testes from 17-dpn males were decapsulated in PBS, cut into smaller fragments, and digested for 30 min at 32°C in DMEM:Ham F12 (1:1; Gibco-BRL) that contained 2% fetal bovine serum (FBS), 0.2 mg/ml collagenase-dispase (Roche), and 0.1 mg/ml DNase I (Roche), with shaking. The resultant seminiferous tubule fragments were washed with DMEM:F12. Two more digestions were carried out under the above conditions and the tubule fragments were washed with DMEM:F12. This material was passed repeatedly through an 18.5G needle and the disaggregated cells were collected by filtration through a 70-µm cell strainer (BD Falcon). The cells were then incubated with shaking for 30 min at 32°C in DMEM:F12 that contained 2% FBS, 0.4 mg/ml hyaluronidase I-S (Sigma), and 0.1 mg/ml DNase I, and centrifuged (200 x g for 10 min). The centrifuged cells were suspended in DMEM:F12 with 10% FBS and the Sertoli cells were allowed to settle for 20 min at 32°C. Settled cells were recovered and cultured at 32°C in a 5% CO2 atmosphere in DMEM:F12 supplemented with 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 1x ITSS (insulin transferrin sodium selenite; Sigma). Three days later, the germ cells residually attached to Sertoli cells were removed at 20°C with a 3-min hypotonic treatment in 20 mM Tris-HCl (pH 7.4).
Pachytene spermatocytes, round spermatids, and elongating spermatids from adult CD-1 mouse testes were separated by gravity sedimentation with the STA-PUT system using a continuous BSA gradient (2–4% in PBS-glucose) [22]. The purity of the cell fractions was estimated by examination under a microscope. To confirm that isolated germ cells were devoid of somatic cell contamination, enriched populations of cultured Sertoli cells, pachytene spermatocytes, and round and elongating spermatids were analyzed by RT-PCR. Primers for mouse transferring (Trf) (MGI:98821), 17ß-hydroxysteroid dehydrogenase (Hsd17b1) (MGI: 105077), and fibronectin (Fn1) (MGI: 95566) were used to amplify the specific cDNAs as markers for Sertoli, Leydig, and myoid cells, respectively (Supplemental Figure 1, available online at http://www.biolreprod.org).
Spermatozoa were obtained from the corpus epididymis. Fragments of epididymis were maintained in ice-cold PBS for 30 min for maximum recovery of released spermatozoa.
CD1 male mice at different developmental stages (8, 12, 18, and 30 dpn) and male mice that were exposed to E2, as described above, were orchidectomized. In order to analyze the aqueous-soluble and insoluble protein fractions, a sequential extraction protocol was used. Whole testes were homogenized in an ice-cold solution of 40 mM Tris (pH 7.0), 2 mM EDTA (tissue weight: buffer volume ratio of 1:3) that was supplemented with a mixture of protease inhibitors (Complete Mini; Roche). After extraction on ice for 1 h, the homogenates were centrifuged at 15 000 x g at 4°C for 60 min. The supernatants were saved (hydrophilic fraction) and the sediment was washed several times with Tris-EDTA solution. The sediment was directly re-extracted by incubation for 30 min at 8°C in SDS-Laemmli sample buffer that contained 6 M urea. This final homogenate was centrifuged at 15 000 x g at 8°C for 60 min, and the supernatant was saved (hydrophobic fraction). Both supernatants were processed as described below.
Total protein extracts were prepared by homogenization of either whole testis or isolated testicular cells in SDS-Laemmli buffer that contained 6 M urea and a protease inhibitor cocktail.
Protein concentrations were measured according to the Bradford method, using BSA as the standard.
Total RNA was extracted from testes at different developmental stages (8, 12, 18, 30, and 120 dpn), from isolated testicular cells, and from somatic tissues using TRIzol (Invitrogen) according to the manufacturer's instructions.
For immunoblot analyses, the following primary antibodies were used: 1) affinity-purified rabbit polyclonal antibodies raised against vinexin (kindly provided by Dr. Kioka, [9]); 2) goat polyclonal antibodies raised against protamine 2 (sc-23104; Santa Cruz Biotechnology); and 3) mouse monoclonal antibodies raised against flotillin 1 (BD Transduction Laboratories) and human vinculin (V9131; Sigma). Peroxidase-conjugated anti-rabbit (Amersham Biosciences), anti-goat (Santa Cruz Biotechnology) or anti-mouse immunoglobulins (Amersham Biosciences) were used as the secondary antibodies.
All protein samples were finally dissolved in Laemmli buffer that contained mercaptoethanol as a reducing agent. Urea-free samples were boiled for 5 min, and urea-containing samples were incubated for 1 h at room temperature. Protein samples were separated by SDS-PAGE [23] on 8%, 10% or 12% polyacrylamide gels. The Mini-Protean II electrophoresis apparatus (Bio-Rad) was used. Gels were electrotransferred to a nitrocellulose membrane (Bio-Rad) by routine methods [24], using the Mini Trans-Blot Cell (Bio-Rad). As a loading control, gel replicates for each analysis were stained with Coomassie blue. Transfer efficiency and protein loading were verified by membrane staining with Ponceau S.
Transferred membranes were blocked in TBS-T (Tris-buffered saline, 0.1% Tween-20) that contained 5% blocking reagent (Amersham Biosciences), and further assayed with the different antibodies. ECL (Amersham Biosciences) was used to develop the blots, according to the manufacturer's instructions. The molecular mass (Mr) of each protein was determined by comparison with prestained Mr calibration kits (Bio-Rad). Signal intensities were quantified by means of a densitometer (model GS-800; Bio-Rad) and the Quantity-One software (Bio-Rad).
All Western blot experiments were repeated at least once. Moreover, at least three replicates were performed within the developmental exposure to E2 and cell-type analyses.
Residual genomic DNA from total RNA samples was digested with DNase (RQ1 RNase-free DNase; Promega). For vinexin analysis, the RNA was reverse-transcribed using a common primer (RT, 5'-GGTGGCTTCGGGCTCATAG-3') for the , , and 
vinexin isoforms, and SuperScript II Reverse Transcriptase (Life Technologies), according to the manufacturer's instructions. Primer pairs for the detection by PCR of specific vinexin isoforms were designed using the Primer Express software (Applied Biosystems). The oligonucleotide sequences were: nested Sorbs3: 5'-ATGCTGGGTCTGGATTCAAG-3'; Sorbs3 , 5'-GCCGCTCTCTTCCTTGTTG-3'; Sorbs3 /, 5'-GATGTAGCCCTTCTTAGTCC-3'; and Sorbs3 , 5'-GGCGCAGGGTCATGAAACCT-3'.
To confirm the consistency of the Sertoli and germ cell isolations, oligo(dT) was used to primer reverse transcription of the RNAs. The cDNAs were amplified using specific primers designed for: ribosomal protein S16 (encoded by Rps16 gene; positive control) transferrin, 17ß-hydroxysteroid dehydrogenase, and fibronectin: Rps16-F, 5'-TTCTGGGCAAGGAGCGATT-3'; Rps16-R, 5'-GATGGACTGTCGGATGGCA-3'; Trf-F, 5'-CCGGGTTAAGGCTGTACTGA-3'; Trf-R, 5'-GTGTCATCCCTGAACAGAAGG-3'; Hsd17b1-F, 5'-GTTATGAGCAAGCCCTGAGC-3'; Hsd17b1-R, 5'-CGCATTGCAGTCAAGAAGAG-3'; Fn1-F, 5'-TTTGCTCCTGCACGTGTT-3'; and Fn1-R, 5'-CTGTGTATACTGGTTGTAGGTGTGG-3'.
PCR was carried out in a Gene Amp PCR System 2700 (Applied Biosystems) under the following general conditions: 94°C for 2 min, followed by 30 cycles of 94°C for 15 sec, 60°C for 30 sec, and 72°C for 1 min, with a final step of 72°C for 10 min. Twenty cycles were performed to avoid saturated amplification of vinexin . For each sample, a replicate was run, omitting the reverse-transcription step, and a template-negative control was run for each primer combination. RT-PCR products were loaded on 2% agarose gels and stained with ethidium bromide. The same starting quantity of total RNA as used in the RT-PCR experiments was electrophoretically separated in 1% agarose gels under denaturing conditions (8% formaldehyde) as a loading control.
Testicular Protein Expression Pattern of Vinexin Is Altered by Developmental Exposure to Exogenous E2
We studied the expression of vinexin protein in the testes of mice exposed to E2. The exposure period in the experimental mice started 2 wk before mating (the mothers being treated) and ceased at postnatal day 30, when the offspring were killed. Thus, male mice were exposed during embryonic and postnatal development. The usual detrimental exposure of estrogenic compounds is by intake. This implies that the potential effects, detected in any particular tissue, cell or developmental stage could be due to E2 or to an indirect effect of estrogen intake.
According to our research strategy, aqueous-insoluble protein fractions were prepared from testes of mice exposed to E2 and of controls, and then assayed by Western blot. The results obtained showed a dramatic change in the vinexin protein pattern (Fig. 1). Antibodies raised against vinexin recognized two protein bands in nonexposed testes. These bands corresponded to different vinexin isoforms, designated as vinexin (76-kDa band) and vinexin (the highest molecular mass band), following the description of Matsuyama et al. [11]. E2 exposure at 0.16 mg/L induced the total depletion of vinexin in extracts from exposed testes, as well as a conspicuous decrease in vinexin . Exposure to lower dosages of E2 (0.02 mg/L and 0.04 mg/L in the drinking water) did not cause any change in the protein patterns of the vinexin isoforms compared to nonexposed mice (data not shown).
|
Under the same experimental conditions, the vinexin-interacting protein flotillin 1 did not undergo significant changes after E2 exposure (Fig. 1). However, we detected a slight reduction in the vinculin protein level.
Vinexin Protein Pattern During Mouse Testis Development
The results presented above led us to characterize further the postnatal developmental protein patterns of the vinexin isoforms. Total and sequential protein extracts were prepared from different developmental stages of mouse testes (8, 12, 18, and 30 dpn) and assayed by Western blotting (Fig. 2). The anti-vinexin antibody recognizes different vinexin forms [9]. Similarly, Matsuyama et al. [11] have used a different anti-vinexin antibody to detect diverse isoforms in various tissues. In both studies, vinexin ß (in addition to other proteins with lower Mr than the or forms) was detected on the basis of molecular mass. The vinexin ß isoform is ubiquitously expressed in different tissues, while other isoforms are expressed in defined tissues and developmental stages. Since the vinexin ß isoform has been characterized as lacking the SoHo domain, which allows binding to lipid raft proteins, such as flotillin, we focused the present study on the isoforms that contain both the SoHo and SH3 domains. We used hydrophilic and hydrophobic proteins that were extracted sequentially, since the distinction between the two forms could provide preliminary information about the subcellular locations of the different vinexin isoforms. Vinexin isoform (76 kDa) was detected predominantly in the hydrophilic fraction, although a proportion of the protein was also detected in the hydrophobic fraction (Fig. 2B). The level of the protein associated with the hydrophilic fraction seemed to decrease in the testes of prepubertal (8 dpn) to 30 dpn mice. In contrast, it was not very clear whether the protein level in the total extracts decreased in parallel with development (Fig. 2A). It was interesting to note that the vinexin isoform, assigned as by Matsuyama et al. [11], was first detected at 30 dpn in the hydrophobic fraction (asterisk in Fig. 2B). However, isoform was not detected in total protein extracts at 30 dpn. The probable reason for the exclusive presence of vinexin in the hydrophobic fraction is that the sequential protein extraction protocol for obtaining the hydrophobic fraction favors enrichment of this isoform. The expression pattern of the vinexin isoform seemed to be related to the presence of elongating spermatids in the seminiferous epithelium at 30 dpn [22].
|
The vinexin-interacting proteins flotillin 1 and vinculin were also studied. Flotillin 1 was detected only in the hydrophobic fraction of the mouse testis homogenate, and the protein level decreased in the testes of prepubertal (8 dpn) to 30 dpn mice (Fig. 2, A and B). Vinculin was present in both the aqueous-soluble and insoluble-associated fractions. The vinculin level in the hydrophilic fraction and total extracts seemed to decrease during testis development (Fig. 2, A and B). In the hydrophobic fraction, vinculin showed low but uniform levels during development.
Differential Levels of Vinexin Isoforms in Testicular Cell Types
Sertoli cells, pachytene spermatocytes, round spermatids, elongating spermatids, and epididymal sperm were isolated, verified to be free of contaminating cell types (Supplemental Fig. 1, available online at http://www.biolreprod.org) and analyzed. Total protein extracts from all isolated cell types were prepared and assayed by Western blotting.
Vinexin was detected in Sertoli cells (Fig. 3A). This fact was consistent with the level decrease observed during testis postnatal development in the hydrophilic fraction, as a consequence of a relatively lower proportion of Sertoli cells in late than in early stages of testicular development. Elongating spermatids showed a strong immunopositive signal (Fig. 3A), which may correspond to the isoform designated as vinexin by Matsuyama et al. [11]. Protein bands detected in both elongating spermatids and the hydrophobic fractions of adult testes showed the same electrophoretic mobility (Fig. 3B).
|
We conclude that the vinexin isoform detected in elongating spermatids corresponds to vinexin (Fig. 3B). Accordingly, vinexin would be testis-specific for germ cells, and would have the highest accumulation in elongating spermatids. However, the vinexin isoform was scarcely present in spermatozoa from the corpus epididymis (Fig. 3B).
The vinexin-associated proteins, flotillin 1 and vinculin, were also studied. Both proteins were detected at high levels in Sertoli cells (Fig. 3A), which was also consistent with the observed decrease in the levels of these proteins in the testis during postnatal development. In fact, the germ cell populations presented no signal for flotillin 1 and contained only low levels of vinculin.
These results show that hydrophilic fraction-associated vinexin , flotillin 1, and vinculin display similar patterns in the seminiferous epithelium, their levels decreasing from 8 dpn to 30 dpn. All three proteins are strongly expressed in Sertoli cells.
Vinexin mRNA Expression Pattern
The results obtained prompted us to study the vinexin expression pattern at the transcriptional level. The cDNA sequences that correspond to vinexin and (accession numbers AF064806 and AB190911, respectively) have been described previously [9, 11]. In addition, a search of the GenBank database revealed a new cDNA sequence from the adult male testis (accession no. AK029703), which is not yet characterized and which we call vinexin . These sequences were used to design specific primers for the detection by RT-PCR of corresponding vinexin transcripts. Figure 4A shows the primer localization in the depicted exonic structure of each isoform (, , and ). The vinexin transcript was similar to vinexin but differed by the absence of 45 nucleotides in the 5'-end of exon 5 and in the polyadenylation site. The putative ORF of vinexin should encode a smaller polypeptide than vinexin (78 versus 82 kDa), containing the SoHo and three SH3 domains of the vinexin isoforms but lacking several amino acids at the N-terminus and C-terminus. This should result in a truncated third SH3 domain.
|
During mouse testis development, the vinexin transcript level seemed to decrease (Fig. 4B). Nevertheless, the transcripts that correspond to the vinexin and isoforms showed marked increases from 30 dpn.
To clarify this developmental expression pattern, different purified populations of germ cells and Sertoli cells were analyzed (Fig. 4C). Vinexin transcripts were expressed mainly in Sertoli cells. The transcripts that corresponded to the and isoforms were detected in germ cell populations from pachytene spermatocytes to elongating spermatids, whereas in Sertoli cells, the levels of transcripts for the isoform were low and were nonexistent for the isoform. These results agree with the expression patterns observed during postnatal development. Expression of vinexin in the testis correlated with that observed for the isoform at the protein level. However, the data are insufficient to allow us to deduce whether or not the protein detected by Western blotting, designated as the form by Matsuyama et al. [11], is translated from the or transcript.
We also studied the expression patterns of vinexin transcripts in several somatic tissues (Fig. 4D). Vinexin was the most ubiquitously expressed, being detected in the liver, muscle, brain, lung, and ovary. It is interesting that the transcript presented a more restricted pattern, being detected only in the testis and scarcely in the ovary.
Exposure to E2 Does Not Trigger Significant Changes in The Vinexin mRNA Expression Pattern in The Testis
As described above, the level of vinexin isoform decreased in the testes of mice exposed to E2, and the spermatid-associated vinexin isoform was absent (Fig. 1). However, at the mRNA level, vinexin maintained its level of expression independently of E2 exposure, and the and transcripts slightly decreased their expression levels in response to E2 exposure. Transcripts for vinexin and were present in germ cells (Fig. 4C). The decrease observed in E2-exposed mice could be due to a loss of germ cells in the seminiferous epithelium of exposed animals. This loss, which mostly affected the elongating spermatids, in which it is mainly present (Fig. 3A), could explain the absence of the vinexin isoform assigned as the form by Matsuyama et al. [11]. The presence of spermatocytes and some round spermatids in the histological analysis of testes from E2-exposed animals (data not shown) suggests that detachment of spermatids is a cytological consequence of exposure to E2 rather than a failure of germ cell differentiation.
In order to address this question, we looked for the presence of a classical protein marker of elongating spermatids, protamine 2, in testes exposed to E2. Western blot analysis revealed that protamine 2 was not present in testis protein extracts from exposed animals (Fig. 5B). Thus, both proteins, the elongating spermatid-associated vinexin isoform and protamine 2, were found to be absent after E2 exposure. This result indicates the absence or strong depletion of elongating spermatids in the seminiferous epithelium of E2-exposed mice.
|
Vinexin is an adaptor protein that is implicated in cell adhesion [10]. It can interact with other proteins that are known to participate in cell interactions, e.g., flotillin 1, a component of lipid rafts [16], and vinculin, an abundant cytoskeletal protein found in cell-cell and cell-substrate interactions [25]. Moreover, through its three SH3 (src homology 3) domains in the C-terminal region, vinexin can bind other proteins, creating large signaling complexes and directing the subcellular localization of specific molecules. In this sense, vinexin can play different roles depending on its binding partners and its subcellular location.
Our results demonstrate that vinexin, flotillin 1, and vinculin are all expressed in the postnatal testis. On the one hand, during postnatal development, hydrophilic fraction-associated vinexin , flotillin 1, and vinculin showed similar expression patterns, decreasing their protein levels from 8 dpn to 30 dpn (Fig. 2). On the other hand, they were also associated with Sertoli cells (Fig. 3A). Nevertheless, in germ cells, we found low quantities of these proteins. Evans et al. [20] have analyzed flotillin 1 in the rat seminiferous epithelium and have detected the protein in nonflagellated spermatogenic cell fractions. Our results show little or no detection of flotillin 1 in the mouse testis germ cell stages analyzed. The vinexin , flotillin 1, and vinculin expression patterns allow us to speculate that all three proteins interact with each other in the setting up of cellular junctions in the seminiferous epithelium.
To investigate further the importance of vinexin during mouse testis development, we characterized the expression patterns in the whole testis and in defined cell types of the seminiferous epithelium. Using polyclonal antibodies against vinexin, we detected two vinexin isoforms in the whole testis, which were designated as and according to Matsuyama et al. [11]. Vinexin was detected mainly in the aqueous-soluble (hydrophilic) protein fraction, which includes cytoplasmic proteins, and its level decreased during testis development. However, a weaker signal, the level of which increased during development, was observed in the aqueous-insoluble (hydrophobic) protein fraction, which includes membrane and cytoskeletal proteins. This increase may be related to the recruitment of vinexin to junction complexes during formation, which are associated with the plasma membrane. Taking into account that the proportions of each cell type in the different stages of the first wave of spermatogenesis are changing [22], we argue that the overall reduction observed in the hydrophilic protein level from 8 dpn to 30 dpn (Fig. 2B) is due to a reduction of the relative percentage of Sertoli cells in the seminiferous epithelium during testis development. In fact, the analysis of isolated testicular cells demonstrated that vinexin was strongly present in Sertoli cells, while vinexin was present exclusively in germ cells, mainly in elongating spermatids (Fig. 3A). Therefore, vinexin was only detected in whole testes from 30 dpn (Fig. 2B), when elongating spermatids start to differentiate in the seminiferous epithelium [22]. It is interesting to note that vinexin was detected in the hydrophobic extract, which suggests a relationship with the membrane or hydrophobic fraction-associated cytoskeletal systems of germ cells.
We studied the expression patterns at the transcriptional level of the different isoforms of vinexin, including the new isoform of vinexin . The sequence of this new cDNA differed from the and sequences (Fig. 4A). The expression pattern of vinexin transcripts was consistent with data obtained at the protein level and data derived from male fetal gonads [11], in which the vinexin isoform was associated with somatic cells and was not detected in germ cells. Transcripts that corresponded to the vinexin and isoforms were present from pachytene spermatocytes to elongating spermatids (Fig. 4C). We determined the presence of vinexin transcripts in different somatic and gonad tissues. Vinexin transcripts showed a restricted expression pattern, being detected in the testis and, at a lower level, in the ovary. In contrast, vinexin and showed more ubiquitous expression patterns (Fig. 4D). The results presented in the present study show that two protein isoforms are detectable with polyclonal anti-vinexin antibodies. According to Matsuyama et al. [11], we assigned vinexin to the smallest 76-kDa band and vinexin to the largest band (Figs. 1, 2B, and 3). It is clear that the vinexin isoform arises as a result of transcript translation, since the Mr of 76 kDa calculated for the smallest band, as detected by anti-vinexin antibodies, is in agreement with the expected size for the theoretically translated protein from the vinexin transcript, and the developmental expression pattern of vinexin mRNA correlates with the protein pattern of the isoform. Nevertheless, at the present time we do not have enough data to confirm that the highest Mr protein detected in the Western blots and associated with germ cells corresponds to the translated vinexin transcript or is translated from the new isoform. We detected only two proteins from the three cDNAs that we studied at the transcriptional level. Several explanations can be put forward: the corresponding transcript is not translated, the transcript is translated into a truncated protein, or the protein is degraded. In addition, the observed accumulation of mRNAs for vinexin and in pachytene spermatocytes (Fig. 4C) and the predominance of the highest Mr isoform in elongating spermatids (Fig. 3B) strongly suggest an uncoupling of transcription and translation. In fact, we did not observe significant reductions in the vinexin and mRNA levels after E2 exposure (Fig. 5A) in spite of the loss of the germ-cell-associated vinexin isoform.
We have shown in the present study that developmental exposure to E2 triggers the depletion of the germ-cell-associated vinexin isoform, in addition to a conspicuous reduction in vinexin (Fig. 1). At the same time, we observed that the protein level of vinculin was slightly reduced. Intriguingly, the classic marker protein of elongating spermatids, protamine 2, was not detected in the testes of mice exposed to E2 (Fig. 5B). There are two possible explanations for these coincidental observations. First, the elongating spermatids that carry both proteins would be absent as a result of a delay in the spermatogenesis process. In this sense, Hosoi et al. [4] have shown the absence of certain germ cell types during spermatogenesis after exposure to the estrogen compound DES, as a consequence of a delay in the differentiation process. The second possible explanation for the depletion of both proteins is that the elongating spermatids were lost by detachment in the exposed testis. It has been described that protamine 2 starts to be translated at spermatid maturation stage 12 in mice [26]. Thus, spermatids in previous maturation stages would start to detach from the seminiferous epithelium. Supporting this last possibility is the fact that testis exposure to E2 can initiate the disruption of apical ectoplasmic junctions in the seminiferous epithelium, with the consequent loss of spermatids [3, 5].
Proteins that are usually restricted to the cell-extracellular matrix focal adhesion site in other epithelia have been demonstrated to be involved in restructuring the apical ES [27]. One possible location of vinexin is at sites of intercellular attachment between Sertoli cells and spermatids, where vinculin is known to be concentrated, at least on the Sertoli cell side. Any vinexin isoforms present in late spermatids at these sites may be internalized at tubulobulbar complexes by Sertoli cells during the sperm release process, thereby accounting for the absence of this protein from epididymal spermatozoa. Tubulobulbar complexes are plasma membranes and cytoplasmic evaginations from spermatids to Sertoli cell invaginations [28]. These structures are considered to be dynamic modifications of Sertoli cell ectoplasmic specializations [29, 30]. Tubulobulbar complexes suffer resorption in Sertoli cells associated with spermiation [31]. The absence of specific isoforms of vinexin in spermatozoa with respect to spermatids may be associated with the resorption of the membrane and associated cytoplasm arising from tubulobulbar-ES complexes. Estrogen exposure triggers the depletion of ES in the seminiferous epithelium [2–5], although the molecular mechanism by which this is achieved is unknown. Several authors have addressed the molecular mechanisms that underlie physiological junction dynamics in the seminiferous epithelium, especially at the ES. Recent results implicate the MAPK (mitogen-activated protein kinase) cascade in the ES disassembly and reassembly processes [32]. Interestingly, it has been strongly suggested that vinexin is involved in the regulation of the MAPK cascade as a scaffold protein, through its interaction with c-Raf and RAF1 [11]. Therefore, we can hypothesize that the imbalance of vinexin levels after E2 exposure may be involved in ES alteration via MAPK cascade destabilization.
In conclusion, we have characterized the expression pattern of a germ-cell-specific vinexin isoform that is absent from the testis at 30 dpn after developmental exposure to E2. This finding suggests the possibility that vinexin is a molecular marker of estrogen-induced damage in the male gonad. However, further analyses must be performed to elucidate whether the expression changes in vinexin isoforms are directly or indirectly caused by E2 exposure, that is, whether the analyzed vinexin isoforms are involved in the actions of estrogenic compounds or alternatively, whether these changes are a consequence of estrogen exposure. Evidence for this last hypothesis is provided in the present study.
ACKNOWLEDGMENTS
The authors thank Dr. Noriyuki Kioka (Kyoto University, Japan) for providing anti-vinexin antibodies, J.F. Escolar for technical assistance, and Dr. Patrick D. Groves and Dr. Massimo De Felici for manuscript review and suggestions.
FOOTNOTES
3These authors contributed equally to this work. 
1Supported by grants from the EC (QLK4-CT-2002-02403) and the MEC (BFU2004-03977/BFI).
Correspondence: 2Jesús del Mazo, Department of Cell and Developmental Biology, Centro de Investigaciones Biológicas, Ramiro de Maeztu 9, 28040 Madrid, Spain. FAX: 34 91 5360432; e-mail: jdelmazo{at}cib.csic.es
Received: 9 January 2007.
First decision: 15 February 2007.
Accepted: 11 June 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
