| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
Unité INSERM 581,3 Hôpital Henri Mondor, 94010 Créteil, France
GERM-INSERM U.435,4 Université de Rennes I, Campus de Beaulieu, 35042 Rennes Cedex, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gene regulation, Leydig cells, luteinizing hormone, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
During spermatogenesis, these different cell types, and in particular germ cells, are known to express specific sets of genes. These genes are regulated at either the transcriptional level or the transductional level [1]. Some of them are exclusively detected in haploid cells, whereas others are first expressed before or during meiosis and continue to be expressed in spermatids. Eddy distributed those genes in four categories: 1) male germ cell-specific gene homologues (e.g., glyceraldehyde 3 phosphate dehydrogenase), 2) unique genes expressed exclusively during spermatogenesis (e.g., protamines), 3) germ cell-specific alternate transcripts leading to specific isoforms, and 4) genes developmentally regulated during germ cell differentiation (e.g., c-Abl protooncogene).
A puzzling observation is that, among the genes expressed at postmeiotic stages, there are large sets of oncogenes (for review, see [1]), neuropeptides (for review, see [2]), as well as genes otherwise uniquely expressed during embryogenesis, such as Hox 1.4 [3] or Zfy-1 or -2 genes [4].
Looking for new genes expressed in testis, we characterized a large series of expressed sequence tags (EST) from human testis by partial cDNA sequencing [5]. Among them, we identified a transcript encoding a protein homologous to the drosophila Goliath protein. In this species, the Goliath gene is located in the Gooseberry-Zipper region [6], a locus containing at least five segmentation genes. The drosophila Goliath gene is transcribed in the mesoderm of 8–24-h-old embryos. The authors proposed that the drosophila Goliath protein is probably a transcription factor that regulated gene expression during mesoderm formation [7]. The drosophila embryonic expression led us to investigate the function of this putative developmental transcription factor in rat testis. Recently, the mouse Goliath mRNA, also named G1RP (G1-related protein), was isolated from 32Dcl3 myeloblastic cells and was shown to be induced in apoptosis by IL-3 deprivation [8].
In this study, we have cloned and identified the rat Goliath species, using the human Goliath EST, and analyzed its expression in testis of normal and hypophysectomized rats. Goliath was found to be expressed in Leydig cell mitochondria as well as in germ cells in testis. In Leydig cells, the protein is regulated by LH but not in germ cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Male Sprague-Dawley rats and 7–8-wk-old hypophysectomized male wistar rats were obtained from Iffa Credo (Lyon, France). Six days after surgery, they were treated subcutaneously either with hCG (human choriogonadotropin, 25 U/animal/day) or with vehicle for 3 consecutive days. Complete removal of the pituitary was verified histologically and by measurement of serum testosterone and LH levels using an electrochemiluminescent assay (ECL; Roche Diagnostic, Mannheim, Germany) and a RIA (Biocode SA, Liege, Belgium), respectively. Hypophysectomized rats testis were collected 6 days after surgery or after the additional hCG injection.
For tissue analysis, animals were anesthetized by intraperitoneous pentobarbital injection (6 mg/100 g). One testis was removed, snap frozen in liquid nitrogen, and stored at -80°C until use. The remaining tissues were fixed in vivo by intracardiac perfusion of the rat with 4% paraformaldehyde in PBS. The second testis was removed, postfixed in 4% paraformaldehyde in PBS for 2 h, and embedded in paraffin. Animal manipulations were performed according to the recommendations of the French Ethical Committee and under the supervision of authorized investigators.
Cell Culture
The R2C Leydig cell line was obtained from the American Type Culture Collection (Manassas, VA). Cells were cultured in Ham F10 medium (Invitrogen, Paisley, UK) containing 15% horse serum (Invitrogen), 2.5% fetal bovine serum (Pan Biotech GmbH, Aidenbach, Germany), 100 IU/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine.
Purification of Testis Cell Populations
Sertoli cells (enrichment >98%, based on morphological criteria) and peritubular cells (enrichment >96%, based on morphological criteria) were isolated from 20-day-old Sprague-Dawley rats as described previously [9]. Leydig cells (purity >98%, determined by 3ß-hydroxysteroid dehydrogenase [HSD] assay) and testicular resident macrophages (purity >94%, determined with the specific ED2 antibody) were isolated from 90-day-old Sprague-Dawley rats according to Klinefelter's method [10]. Spermatogonia (purity >90%, based on morphological criteria assessed by electron microscopy) were prepared from 9-day-old Sprague-Dawley rats according to Bellve et al. [11]. Adult rat pachytene spermatocytes, round spermatids (purity >90%, based on morphological criteria assessed by electron microscopy), and late spermatid cytoplasmic fragments (purity >75–85%) were prepared by centrifugal elutriation [12].
Cloning and Sequencing
The rat Goliath cDNA was isolated from a rat liver cDNA library constructed in the 
TriplEx vector (Clontech, BD Biosciences, Palo Alto, CA), using the human Goliath cDNA as a probe (accession no. AY083998). MilleGen Biotechnologies (Labege, France) established the complete sequence of the rat Goliath cDNA clone (1443 base pairs; accession no. AY190520).
Northern Blot Analysis
Rat tissues total RNA were obtained and analyzed by Northern blot as previously described [13]. The blots were hybridized with either the rat Goliath cDNA or the mouse ß-actin cDNA (accession no. AK075973) labeled with 
[32P] dCTP according to the Random Prime protocol (Invitrogen). Northern blots were washed in 0.5x SSC (1x SSC: 150 mM sodium chloride, 15 mM sodium citrate, pH 7), 0.1% SDS, at 68°C. The specific signals were detected using a PhosphorImager STORM 840 (Molecular Dynamics, Amersham Biosciences, Little Chalfont, UK).
Antibodies
The obtainment of an affinity-purified rabbit polyclonal antibody, raised against amino acids 223–369 of human Goliath, has been described elsewhere (A. Guais et al., unpublished results). Other antibodies were kindly provided by the cited groups and used at the dilutions indicated: anti-MIF (macrophage inhibitory factor) monoclonal mouse antibodies (dilution 1:100 for immunohistochemistry) or polyclonal rabbit (dilution 1:1000 for Western blot) by Dr. A. Meinhardt [14] (Philipps University of Marburg, Marburg, Germany), the anti-porcine P450 17- (17-hydroxylase/17-20-lyase) rabbit antiserum (dilution 1:500) [15] by Dr. A. Payne (Stanford University School of Medicine, Stanford, CA), and the anti-bovine 3 ß-HSD (2,3 ß-hydroxysteroid-dehydrogenase 
5/4-isomerase) rabbit antiserum (dilution 1:400) [16] by Dr. G. Defaye (CEA/DMBS/BRCE, Grenoble, France). Rabbit control Ig G (1.6 µg/µl), anti-rat ED1 and ED2 monoclonal antibodies (dilutions 1:50 and 1:800, respectively), anti-GAPDH (glyceraldehyde 3-phosphate dehydrogenase) monoclonal antibody (dilution 1:1000), anti-rat P450-ssc (P450 side-chain cleavage) rabbit antiserum (dilution 1:2000), anti-actin (dilution 1:1000), and anti-cytochrome C oxidase subunit I monoclonal antibody (dilution 1:50) were obtained from Vector Laboratories (Burlingame, CA) Serotec (Oxford, UK), Chemicon International Inc. (Temecula, CA), Sigma (St. Louis, MO), and from Molecular Probes (Eugene, OR), respectively.
Western Blot Analysis
Rat tissues and cells were homogenized by using an Ultraturrax homogenizer (Janke and Kunkel IKA Labortechnik, Staufen, Germany) and kept on ice for 30 min in 50 mM Tris-HCl buffer pH 7.5 containing 150 mM NaCl, 1% Nonidet P-40, 0.1% SDS, 1 mM sodium orthovanadate, 1 mM sodium fluoride, 3 mM phenylmethylsulfonylfluoride (PMSF), 1 µg/ml leupeptin, 10 µg/ml aprotinin, and 1 µg/ml pepstatin. After centrifugation at 13 000 x g for 2 min, supernatants were frozen and subsequently assayed for protein content (Bradford method, Bio-Rad, Hercules, CA). Following denaturation for 5 min at 90°C, protein from the supernatants (50 µg/lane) were separated on 12% SDS-PAGE, blotted onto an Immobilon-P polyvinylidene difluoride (PVDF) transfer membrane (Millipore, Billerica, MA), and probed with the anti-Goliath purified antibody (1:1000), anti-P450-ssc antibody (1:500), anti-actin (1:1000), anti-MIF (1:1000), or anti-GAPDH (1:1000). Membrane treatment was done according to the enhanced chemifluorescence protocol (ECF; Amersham). Goat anti-rabbit or mouse Ig G, coupled to alkaline phosphatase (1:10 000), were used as secondary antibodies. The membrane was incubated for 10 min in ECF solution, and fluorescent signals were visualized using a PhosphorImager STORM 840 (Molecular Dynamics).
Immunohistochemistry
Rat paraffin sections (5 µm thick) were mounted on Superfrost plus-coated slides (O. Kindler GmbH, Freiburg, Germany). Slides were deparaffinized twice in xylene and rehydrated in graded ethanol (100%, 90%, 70%) and water baths. For antigen retrieval, slides were treated for 2 x 10 min with trypsin (Sigma, 1 mg/ml) and then microwaved (12 min 750 W, 2 x 10 min 350 W) in 10 mM citrate buffer, pH 2.5. To reduce eventual nonspecific binding, the slides were incubated for 40 min at room temperature with blocking buffer (PBS, 1% BSA, 10% normal serum of the secondary antibody species). Slides were then incubated for 2 h with primary antibodies and, after four washings in PBS, incubated for 30 min in the dark with either Cy3-conjugated sheep anti-rabbit Ig (1:500; Sigma), FITC (fluorescein isothiocyanate)-conjugated goat anti-rabbit Ig (1:150; Vector Laboratories) or FITC-conjugated donkey anti-mouse Ig (1:200; Jackson ImmunoResearch, West Grove, PA) antibodies. Slides were mounted in a glycerol-based antifade mounting medium (Vectashield, Vector Laboratories). DNA was stained using 4,6-diamidino-2-phenylindole (DAPI). Sections were observed using a Zeiss microscope (Carl Zeiss, Inc., Oberkochen, Germany). For the double labeling, the antibodies reactions were performed independently and successively to avoid secondary antibodies cross-reactions.
Subcellular Fractionation
R2C Leydig cells (approximately 80 million cells) were rinsed twice in ice-cold PBS, resuspended in 1 ml lysis buffer (250 mM sucrose, 10 mM Tris pH 7.5, 1 mM EDTA, 0.1 mM PMSF, 2 µg/ml leupeptin, 4 µg/ml aprotinin, 1 µg/ml pepstatin) and homogenized in a Teflon-glass homogenizer with 30 strokes by hand at 4°C. Subcellular fractionation was performed by sequential centrifugations. The homogenate was centrifuged four times at 1000 x g for 5 min at 4°C to remove unbroken cells and nuclei. Postnuclear supernatants were pooled and centrifuged for 15 min at 10 000 x g at 4°C. The pellet was resuspended and centrifuged twice under the same conditions. The final pellet P1 containing the mitochondria was resuspended in the same buffer. The supernatants were pooled centrifuged at 100 000 x g for 1 h at 4°C to obtain the cytosol S2 and a pellet P2 containing the plasma membranes. The pellet was resuspended in lysis buffer and centrifuged twice as before.
Immunoelectron Microscopy
Immunogold labeling was performed with two different methods: cryosectioning according to Tokuyasu [17] and LR White embedding. For cryosectioning, (Fig. 4C, a), Leydig R2C cells were fixed with 4% formaldehyde and 0.2% glutaraldehyde in PBS, pH 7.4. Cells were prepared as previously described [18–20], embedded in 10% gelatin for 30 min on ice, and cut into small gelatin blocks that were infiltrated with 2.3 M sucrose and frozen in liquid nitrogen. Ultrathin cryosections were performed using a Leica Ultracut E cryoultramicrotome (Leica Microsystems S.A., Rueil-Malmaison, France) with a cryochamber attachment. For immunolabeling, the sections were preincubated in phosphate buffer containing 0.05 M ammonium chloride and 1% BSA and subsequently incubated with a rabbit anti-Goliath antibody diluted 1:100 in 0.1% BSA-PBS for 1 h at room temperature. Sections were then incubated with protein A-10 nm colloidal gold (purchased from J.W. Slot, University of Utrecht, Utrecht, The Netherlands) diluted 1:65 in 0.1% BSA-PBS for 45 min. Sections were washed in phosphate buffer, fixed with 1% glutaraldehyde for 5 min, rinsed with water, and then prestained with 2% neutral uranyl acetate. Sections were stained in a mixture of 2% methyl cellulose containing 3% aqueous uranyl and analyzed using a Philips Tecnai 12 (Philips Electron Optics, Eindhoven, The Netherlands). In negative control experiments, the primary antibody was omitted.
|
For LR White embedding (Fig. 4C, b), Leydig R2C cells were fixed with 4% paraformaldehyde and 0.2% glutaraldehyde in PBS pH 7.4 for 1 h, washed in phosphate buffer containing 0.05 M ammonium chloride for 30 min, dehydrated by increasing ethanol series, and embedded in LR White (Polysciences, Niles, IL) at 37°C for 5 days. Indirect immunolabeling was performed on ultrathin sections obtained using a Reichert Ultracut (Leica Microsystems S.A.) and collected on Formvar-coated nickel grids. The sections were blocked with phosphate buffer containing 5% BSA and 1% fish gelatin and incubated with the polyclonal serum against Goliath diluted 1:100 and then with 10 nm protein A-gold diluted 1:65. Ultrathin sections were contrasted with uranyl acetate for 15 min and examined in a Philips Tecnai 12.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We screened 106 clones of a rat liver cDNA library (Clontech), constructed in TriplEx vector, using the human Goliath cDNA as a probe and isolated 10 clones. The sequence of the longest insert (1443 bp; accession no. AY190520) revealed an open reading frame of 419 amino acids encoding a putative 46-kDa protein. Alignment of this proteic sequence with the drosophila sequence (accession no. AY069169) exhibited 42% of identical and 57% of similar residues over the first 300 amino acids. Rat Goliath protein shares 98% of identity with mouse protein (accession no. NP067515) [8]. Thus, the rat sequence was considered as a possible orthologue of the drosophila Goliath protein.
Analysis of this sequence on the Pfam (protein family) server reveals a N-terminal PA (protease associated) domain and a C-terminal ring-H2 domain. The PA domain is described as a putative protein-interacting domain composed of a ß-sheet and peripheral -helices [21, 22]. The ring-H2 domain is a subfamily of ring-finger domain with a specific C3H2C3 motif binding a zinc atom and has been involved mainly in the formation of macromolecular assemblages [23]. The hydropathic profile reveals three hydrophobic domains: a small leader sequence (amino acids 1–24), a potential central transmembrane domain (amino acids 192–220), and a third short hydrophobic domain in the C-terminal part of the protein (amino acids 381–411), suggesting a transmembranous location of the protein. Recently, Mahon and Bateman [21] described a new family of proteins, the RZF-related proteins, which could compose a new family of receptors or signal transducers. This family is characterized by the following domains: PA, transmembrane, and ring-H2 domains. Thus, Goliath could be a new member of this family.
This protein is likely to be encoded by only one gene since, in rat, a single EST species has been obtained from various tissues. Furthermore, a single localization has been described for the Goliath gene on chromosome 13 (NW_042846.1).
Goliath Expression in Rat Tissues
Northern blot analysis of Goliath mRNA expression revealed a 1.85-kb transcript in all analyzed tissues (Fig. 1A), although the expression level was very low in intestine and kidney as compared to the ß-actin signal. Using anti-Goliath polyclonal antibody, which cross-reacts with the rat and mouse proteins, we confirmed its ubiquitous expression. Actin and GAPDH were used as loading control. Three main isoforms (28, 46, and 50 kDa) were detected at various levels of expression in the 11 tissues tested (Fig. 1B): The 28-kDa band was expressed in all tissues, and the 46- and 50-kDa bands were weakly and unevenly expressed in most of the tissues tested; 36- and 34-kDa tissue-specific isoforms were detected only in intestine. The discrepancy between the expression levels of Goliath mRNA and protein indicate a posttranscriptional regulation.
|
Goliath Immunolocalization in the Rat Testis
We further investigated Goliath expression in adult rat testis by using immunohistochemistry on paraffin sections. As shown in Figure 2A, Goliath was expressed in the interstitium and the seminiferous tubules of testis. The three main cell types present in the interstitial tissue are Leydig cells, macrophages, and endothelial vessel cells. Goliath was not detected in endothelial vessel cells. In order to discriminate the cells expressing Goliath, we performed a series of double immunohistochemistry (Fig. 2B) using Goliath antibody in combination with antibodies against the Leydig-specific 3ß-HSD (Fig. 2B, a) or against the macrophage ED1 and ED2 proteins [24] (Fig. 2B-b). The distribution of Goliath (Fig. 2B, c and d) systematically overlapped with 3ß-HSD's expression (Fig. 2B, e) but not with macrophage markers (Fig. 2B, f). Thus, Goliath is specifically expressed in Leydig cells and is present mainly in the cytoplasm.
|
In seminiferous tubules, Goliath was strongly expressed in germ cells and dependent on the spermatogenesis stages (Fig. 2C). Four expression profiles could be distinguished: 1) a faint cytoplasmic dotlike labeling in primary spermatocytes (Fig. 2C, a); 2) a strong cytoplasmic dotlike labeling around nuclei at stages VI, VII, and VIII in round spermatids at steps 6–8 (Fig. 2C, a); 3) a dotlike labeling in cytoplasm of elongating spermatids at steps 9–14 (Fig. 2C, b) from stages X to XIV; 4) a very intense staining of elongated spermatids (steps 15–19) in the cytoplasm and flagella (Fig. 2C, c) present from stages III to VIII. At higher magnification, we observed a condensation of Goliath in a ring structure in the middle piece of the flagella in elongated spermatids (data not shown). Thus, Goliath is expressed mainly in haploid germ cells during spermatogenesis.
Goliath Isoforms in the Different Testis Cell Types
We analyzed the expression of Goliath isoforms by Western blot in the different testis cell types (Fig. 3). In whole testis, three isoforms (50, 46, and 28 kDa) were expressed. The 50- and 28-kDa isoforms were expressed mainly in Leydig and Sertoli cells. The 46-kDa isoform was present in all cell types, including round spermatids, pachytenes spermatocytes, residual bodies, Sertoli cells, and peritubular cells but not Leydig cells. A 36-kDa band appeared in Sertoli cells but was not observed in whole testis probably because of the weak proportion of Sertoli cells (5% of the total cells) in this tissue [25]. An additional band was observed in Leydig cells at 26 kDa. No Goliath isoform was detected in macrophages in accordance with the immunohistochemistry experiments.
|
Intracellular Goliath Localization
In all cell types, Goliath was located mainly in the cytoplasm (Fig. 2, B and C), especially in elongated spermatids, where the dotlike signal is characteristic of a cytoplasmic organelle. As a first approach, we wanted to determine whether Goliath colocalized with cytoskeleton- and organelle-specific markers: tubulin for cytoskeleton, calnexin for endoplasmic reticulum, LAMP-1 (lysosome-associated membrane protein type 1) for lysosomes, and cytochrome C oxidase located in the inner mitochondrial membrane. On testis sections, it can be observed that Goliath signal colocalized only with the cytochrome C oxidase in Leydig cells (Fig. 4A). In a second series of experiments, using R2C Leydig cells, a tumor cell line, we performed differential centrifugations to isolate the different cells compartments (Fig. 4B). We noticed that the 50- and 28-kDa isoforms are specifically expressed in mitochondria, whereas the 46-kDa isoform is present in cytosolic and microsomal fractions. We observed that Goliath is enriched 450 times in the same fraction as the cytochrome C oxidase (Fig. 4C). Finally, immunoelectron microscopy on rat R2C cell ultrathin sections revealed a specific labeling of mitochondria (Fig. 4D). All together, these data clearly indicate that Goliath is associated with mitochondria.
Goliath Regulation by LH
Since most of the testicular functions are under the control of the hypophysis and because of the strong expression of Goliath in Leydig and germ cells, we wanted to determine whether pituitary hormones control Goliath expression. We chose to perform short-term hypophysectomy on adult rats, an experiment that induced few modifications of the testis [26]. Six days after hypophysectomy, serum testosterone was undetectable in hypophysectomized rats, and we observed a decrease in the cytoplasmic space of Leydig cells, whereas no modification of the tubular cells was detectable. In parallel, we analyzed the expression of Leydig cell markers (Fig. 5A). MIF is a paracrine factor expressed at a constant level in Leydig cells in testis throughout hypophysectomy [14]. As depicted in our experiment, the 3ß-HSD enzyme protein level decreases but is not completely LH dependent [27] because a basal protein level is observed after 6 days of hypophysectomy [28]. P450-ssc and P450-17 enzymes are known to be under the tight regulation of LH/hCG. These proteins are undetectable 2 days after hypophysectomy [27]. In our experiment, no signal was observed 6 days after hypophysectomy. In the same time, Goliath expression in Leydig cells was reduced to a nearly complete extinction, whereas it remained strongly expressed in germ cells at all stages (Fig. 5B). We could not detect any regulation of Goliath in Sertoli cells since the initial signal was too weak. But according to the literature, Sertoli cells are not affected by 6-day-long hypophysectomy [29]. Similar results were obtained at 12 days after hypophysectomy, a time point at which the germ cell number was drastically reduced (data not shown).
|
We sought to reverse the hypophysectomy effect on Goliath expression. Six days following hypophysectomy, rats were injected with hCG (25 U), once a day, from Day 6 to Day 8. Each of the hCG-injected rats exhibited an increasing testosterone level along the experiment from 15 to 72 ng/ml 6 h after the first injection and the last injection, respectively. We analyzed Goliath and P450-ssc marker expression. After a nearly complete extinction of the marker expression in the hypophysectomized animals, the three hCG injections restored the P450-ssc expression (Fig. 6), as already described [27]. Similarly, Goliath expression, which was barely detectable in Leydig cells of hypophysectomized animals, was recovered after 3 hCG injections (Fig. 6), whereas no variation of Goliath expression was detected in germ cells. Therefore, Goliath expression appears to be under the control of LH in Leydig cells with no effect in germ cells.
|
We analyzed the profile of Goliath isoform expression in whole testis of hypophysectomized rats. Out of the three Goliath bands, only the 50-kDa band was down regulated in the hypophysectomized rat and restored to the initial level after the hCG injections (Fig. 7). MIF expression remained stable whatever the conditions. This 50-kDa isoform, specific to Leydig and Sertoli cells (Fig. 3), was likely to correspond to the Leydig cell signal detected by immunohistochemistry, and the weak 50-kDa remaining signal after hypophysectomy might correspond to Sertoli cell signal. The 46-kDa isoform, present only in germ cells and Sertoli cells, remained constant. The Leydig cell 28-kDa isoform was not regulated by hCG/LH and might be related to the weak Goliath signal remaining in Leydig cells after hypophysectomy. As expected, the P450-ssc enzyme was clearly down-regulated after hypophysectomy and recovered within 3 days of hCG injection. Thus, one can conclude that the Goliath 50-kDa protein is under the control of LH/hCG in testis in contrast to the other Goliath isoforms.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because of its structural organization, Goliath belongs to the RZF-related protein family, which composes a new family of receptors, or signal transducers, that is very conserved across species (from yeast, plants, worms, chicken, mouse, human, and rat) indicating a major role in the cell [21]. Members of this protein family would mediate their effect through the extracellular or luminal PA domain and the cytosolic ring-H2 domain. Both domains, separated by a transmembrane sequence, could interact with proteins on opposite sides of the membrane. Given that Goliath protein is mitochondrial, one can assume that the PA domain is inside the mitochondria or inside the intermembrane space and that the ring-H2 is cytoplasmic or matrix faced. Remarkably, no mitochondrial addressing signal is detected using Mitoprot II 1.beta [30], iPSORT [31], or TargetP [32] programs. Thus, it is possible that the Goliath mitochondrial addressing signal is distributed throughout the entire sequence of the protein as previously described for the heme lyase or the carnitine palmitoyltransferase 1 [33, 34].
The rat Goliath protein is widely distributed as three main isoforms (28, 46, and 50 kDa). The open reading frame of Goliath mRNA encodes a 46-kDa protein. Using in vitro transcription/translation/maturation, we have previously shown that the human and mouse mRNA are translated in a 46-kDa protein that is glycosylated in a 50-kDa form and further cleaved into a mature 28-kDa protein (A. Guais et al., unpublished results). Germ cells exclusively expressed the 46-kDa precursor protein, whereas Leydig cells expressed only the 50-kDa premature and, to a lesser extent, the 28-kDa processed isoform. Thus, in rat Leydig cells, Goliath follows the same process as in humans, whereas the germ cells do not mature the protein.
As Goliath, several genes are described to be expressed both in germ cells and in somatic cells in adult rat or mouse. These include the Meig 1 gene [35], the proenkephalin gene [36], the farnesyl Ppi synthase gene FPP [37], the Cu/Zn superoxide dismutase gene SOD1 [38], aromatase gene [39], and Estrogen receptor gene [40]; all these proteins are encoded by somatic and germ cell-specific transcripts, issued from two distinct promoters on the same gene. Goliath probably does not fit in this scheme since it appears to be transcribed into a single mRNA.
We analyzed the effects of hypophysectomy and hCG treatment on Goliath expression. Six days after hypophysectomy, Goliath expression is almost absent in Leydig cells and is restored following hCG injections. This expression pattern is similar to the expression of the Leydig cell-specific markers, P450-ssc or P450-17 enzymes, known to be under the tight regulation of LH/hCG [27]. MIF is expressed whatever the LH conditions. A large part of the literature describes a strong atrophy of Leydig cells after hypophysectomy, corresponding mainly to an important reduction of the intracellular organelles volume (for review, see [41]). All these structural alterations induce a cellular dysfunction, for instance, reduction of the testosterone secretion capacity. All these phenomena occurred within the first 6 days of hypophysectomy. However, Tapanainen et al. [42] described that Leydig cells undergo a moderate apoptosis after rat hypophysectomy: The interstitial cells apoptosis is 2.4 times enhanced as compared to normal after 2 days. But our observation of MIF expression corroborates the presence of atrophied Leydig cells. After 3 days of LH injection, we demonstrate that Goliath expression is recovered in Leydig cells. Keeney et al. [43] described a light Leydig cell proliferation after LH removal followed by LH restoration. Thus, the recovery of Goliath expression could be partly due to these new Leydig cells. But considering the intensity and the distribution of the signal, the old Leydig cells also expressed Goliath. Furthermore, after EDS treatment, rat adult Leydig cells died from apoptosis within 3 days. In that case, the new ones began to regenerate only 14 days after the EDS treatment (for review, see [44]). The time point used in our experiment does not allow us to see the appearance of the new Leydig cells. All together, we conclude that the modification of Goliath expression is not due to apoptosis followed by regeneration; rather, Goliath is regulated by LH/hCG in Leydig cells. After 6 days of hypophysectomy, the spermatocytes, spermatids, and Sertoli cells still expressed Goliath in a LH-independent manner at a stage where no major macroscopic morphological modification is observed in these cells [29, 45, 46].
Western blot analysis of whole testis extracts shows that only the 50-kDa Goliath isoform is down-regulated in hypophysectomized rats testis and recovered after hCG injection, whereas the 28- and 46-kDa isoforms remained constant. The precursor isoform expression remains stable in total testis extracts since it is expressed mainly by germ cells and Sertoli cells in which Goliath does not exhibit any regulation. Such a constant expression of a final protein product (28 kDa) might reflect a long half-life responsible for the remaining Goliath signal observed in Leydig cells by immunohistochemistry after hypophysectomy.
As well as Goliath, P450-aromatase is positively regulated by LH and also by steroids [40]. But no differential regulation of P450-aromatase between Leydig and germ cells has been described. To explain the differential regulation of Goliath by LH, a possibility might be that cell-specific transcription factors regulate Goliath gene promoter in a differential manner in Leydig cells and other testis cells.
Recently, Baker and Reddy [8] suggested that the mouse Goliath homologue could be involved in the IL-3 withdrawal-induced apoptosis of myeloid precursor cells. In testis, apoptosis regulates the total cell number [47, 48], but apoptosis of adult Leydig cells has not been described in normal physiological conditions [49]. Since Goliath is strongly expressed in normal adult Leydig cells, it would not be involved in apoptosis of those cells if any. Apoptosis is a strong phenomenon in adult testis since it is responsible for the loss of up to 75% of mature sperm cells [50]. Nevertheless, Goliath is probably not involved in this process for two reasons: 1) Apoptosis is observed mainly in spermatogonia in which Goliath is absent [48, 51], and 2) hypophysectomy, or GnRH-antagonist treatment, increases germ cells apoptosis [45, 50] after 1 wk at a stage where there is no obvious modification of Goliath expression. Thus, in testis, Goliath does not seem to be involved in apoptosis.
The pathway in which Goliath is involved remains elusive. In Leydig cells, Goliath mitochondrial localization and regulation by LH/hCG would be relevant to steroid production that takes place partly in mitochondria. However, the postmeiotic expression and the mitochondrial flagella localization could rather mean that Goliath plays a role in energy production or in flagella movement for spermatozoid mobility. Such a differential role for a mitochondria protein has not been observed as yet and needs further investigation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: F. Bulle, Unité INSERM 581, Hôpital Henri Mondor, 94010 Créteil, France. FAX: 33 1 48 98 09 08; bulle{at}im3.inserm.fr
Received: 9 May 2003.
First decision: 28 May 2003.
Accepted: 17 September 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
