| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
INSERM U 567 CNRS-UMR 8104,4 Institut Cochin, Département d'Hématologie, Maternité de Port-Royal, Université René Descartes, 75014 Paris, France
Station de Physiologie de la Reproduction et des Comportements,5 UMR INRA-CNRS 6073, 37380 Nouzilly, France
GERM-INSERM U. 435 Université de Rennes I,6 Campus de Beaulieu, 35042 Rennes cedex, Bretagne, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
endoplasmic reticulum, epididymis, gamete biology
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This anatomical and histological regionalization of the epididymis is a reflection of the gene activities controlled by an interplay between hormones of blood origin and products coming from the seminiferous tubules, i.e., factors and spermatozoa themselves. The nature and regulation of the spermatozoon maturation process require identification and characterizing the factors influencing them.
We have recently started to characterize of a group of genes known as Phtf (putative homeodomain transcription factor) [4]. The first member of the Phtf family was discovered in the Insecta class and two others have been identified in mammals (Phtf1 and Phtf2). The function of the Phtf genes is not known yet, but the expression of both Drosophila Phtf and mammalian Phtf1 are closely linked to the male germ cell activity [4, 5]. Phtf1 transcripts have an open reading frame that could encode a polypeptide of 761 amino acids. PHTF1 displayed the properties of an 84-kDa integral membrane protein present in the endoplasmic reticulum (ER) of the meiotic and postmeiotic male germ cells of rodents [6]. In these cells, PHTF1 appears initially during meiosis in an ER domain juxtaposed to the Golgi apparatus. After the Golgi phase of spermiogenesis, it extends throughout the ER system and, by the end of spermiogenesis, PHTF1 has been released from the mature spermatic cell within the residual body. No trace of the protein seems to remain in the immobile spermatozoa released into the lumen of the seminiferous tubule.
Our search for any remaining PHTF1 activity accompanying the organelles still present in the rat maturing germ cell did not detect any reactivity in the maturing spermatozoa. However, we did discover that PHTF1 was present in the epididymal epithelium. We searched for its expression throughout the organ and found that the epithelial principal cells of the caput epididymis are the sites of PHTF1 production. We report the detection of a 56-kDa PHTF1 that is a shorter form than the 84-kDa form found in testis and we characterize its intracellular location. We offer some insights into how its expression is regulated.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
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 (guideline 19/04/1988). Male Sprague-Dawley rats were killed in a CO2 chamber. Testis and epididymides were dissected out as indicated previously [6, 7]. Epididymides were sectioned into three segments, caput, corpus, and cauda. For RNA extraction, fresh tissues were homogenized in 20 volumes of Trizol reagent (Invitrogen, Groningen, Netherlands) at 4°C. Protein extracts were prepared by homogenizing freshly isolated tissues in 5 volumes of denaturing lysis buffer (0.2% SDS, 1% NP-40, 0.5% sodium deoxycholate, 150 mM NaCl, 5 mM EDTA, 0.7 µg/ml pepstatin, 0.5 µg/ml leupeptin, 40 µg/ml bestatin, and 100 µg/ ml PMSF) at 4°C. Extracts were cleared by centrifugation at 3000 x g for 15 min and were stored at –80°C. Protein quantitations were performed with micro BCA assay (Biorad, Marne-la-Coquette, France).
Surgical Procedures
Male rats were anesthetized by intraperitoneal injection of a ketamine-xylazine mixture at a dose of 90 mg/kg (ketamine) and 10 mg/kg (xylazine). The abdominal surface was prepared for aseptic surgery and, after incision, testis and epididymis were exposed. Ligations and sutures were done using silicone-treated nonabsorbable braided no. 4-0 silk. Castrations were performed after exposure of scrotal contents and the testicular vascular supply was ligated without compromising epididymal blood supply. Testes were dissected out from the epididymides and excised. Epididymis and fat pad were returned into the scrotum and the incision was sutured. Experimental rats were killed on Days 1, 2, 4, and 7 postcastration. The epididymis was removed and trimmed from associated fat and connective tissue. After division of the different regions, tissues were homogenized for RNA or protein extractions. In the ligated animals, unilateral efferent ducts were ligated at their junction with extratesticular rete testes without compromising testicular or epididymal blood supply. Each testis was returned into its tunica vaginalis and the incisions were sutured. The rats were killed on Days 1 and 2 postligation.
Glycosylation and Acid Phosphatase Assays
Fifty micrograms of total protein were acetone precipitated and 100 µl of denaturing buffer was added and the reaction was incubated with Endo-H or PNGase F glycosidase for 2 h at 37°C as indicated by the manufacturer (New England Biolabs). For dephosphorylation experiments, calf intestinal phosphatase (New England Biolabs) was used at 0.1 U per µg of protein in 100 µl of 50 mM Tris-HCl, 100 mM NaCl, 10 mM MgCl2, 1 mM DTT, pH 7.9, at 25°C for 2 h. Reactions were precipitated with acetone and electrophoresed by SDS-PAGE gel.
Western Blot and Fluorescence Microscopy Analyses
A total of 20 µg of protein was electrophoresed by 5–15% or 10% discontinuous SDS-PAGE and transferred to Hybond-P membranes (Amersham). Membranes were washed twice with distilled water, stained with Ponceau S, and blocked 1 h with 5% nonfat dry milk in TBS-T. Anti-PHTF1, the polyclonal mp71 antibody [6], or normal sera were used diluted at 1:1000. The commercial monoclonal myc antibody 9E10 (Roche, France) was used diluted at 1:5000. Secondary anti-rabbit or anti-mouse HRP-conjugated antibodies (Roche) were used diluted at 1:5000. Three TBS-T washed were performed between steps. Reactive bands were detected by enhanced chemiluminescence Western blotting detection kit (Amersham).
Bouin-fixed, paraffin-embedded slides were treated as previously described [6]. Dilutions were as follows: mp71 was used at 1:100; antigiantin [8] (gift of Dr. Y. Ikehara, Fukuoka, Japan), anti-p58 (Sigma), and anticalnexin (gift of Dr. Fanen, Créteil, France) were used diluted at 1:250; anti-TGN38 and anti-ß-adaptin were used at 1:150 (both were a gift of Dr. Benichou, Paris, France). The secondary antibodies were commercial anti-rabbit tetramethyl rhodamine isothiocyanate (TRITC)-conjugated, anti-rabbit fluorescein isothiocyanate (FITC)-conjugated as indicated in figures or anti-mouse FITC-conjugated. Images shown in Figures 5 and 6, a and b, were obtained with a Zeiss Axioplan 2 (plan apo, NA 1.3) microscope and processed with Metamorph (Adobe Systems). Confocal images were obtained with a Leica TCS SP2 device with a 63x objective (plan apo, NA 1.4). Slides were edited with Photoshop (Adobe Systems, Mountain View, CA).
|
|
Transient Expression of Phtf1 in IMCE-20 Cells
Immortalized canine epididymal cells IMCE-20 were cultured as indicated by Telgmann et al. [9]. Medium was changed every 48 h and transfections were performed with 2 µg of plasmid per ml of medium by using Fugene-6 reagent as indicated by suppliers (Roche). Plasmids containing the mouse Phtf1 coding sequence with myc tags in N- or C-terminal extremities were elsewhere described [6].
Metabolic Labeling and Immunoprecipitation
Twenty-four-hour transiently transfected cells were washed twice in media lacking serum. Metabolic labeling was performed by incubating cells for 4 h in Met-free media containing 50 µCi/ml of 35[S]-Met. Cells were chased in normal culture medium and, 24 h after, the cells were washed in PBS and incubated in denaturing lysis buffer (0.2% SDS, 1% NP-40, 0.5% sodium deoxycholate, 150 mM NaCl, 5 mM EDTA, 0.7 µg/ml pepstatin, 0.5 µg/ml leupeptin, 40 µg/ml bestatin, and 100 µg/ml PMSF) at 4°C for 30 min. After a 30-min centrifugation at 15 000 x g, supernatants were recovered and proteins were acetone precipitated. Treatments with glycosidases were then performed. Lysates in denaturing buffer were diluted by adding 10 volumes of lysis buffer without SDS and clearing for 30 min with 10 µl of protein G-Sepharose beads (Amersham) at 4°C. Extracts were then immunoprecipitated overnight with the same quantity of Sepharose beads and 3 µg of anti-myc antibody (9E10; Roche). Pellets were washed five times with lysis buffer without SDS, and finally recovered in 30 µl of Laemmly buffer.
RNA Analyses
For Northern blots, 3–30 µg of total RNAs were run in 1% agarose-formaldehyde gels. Staining with ethidium bromide was used for detection of 18 S rRNA. Hybridization probes for Phtf1 that covered all the coding region were elsewhere described [4]. Reverse transcriptions were performed using random hexamers and Super ScriptII reverse transcriptase as indicated by the manufacturer (Invitrogen). PCRs were performed by using primers Phtf1sens, Phtf1antisens for 32 cycles EP2sens, EP2antisens for 32 cycles Gapdssens, Gapdsantisens for 28 cycles (the sequences are listed in Table 1). Each PCR was performed as follows: an initial 96°C denaturation for 3 min; 96°C for 15 sec; 55°C (Gapds) or 58°C (Phtf1 and EP2) for 15 sec; and 72°C for 1 min for 28–32 cycles; a final extension was carried out at 72°C for 7 min. Amplified products (289 base pairs [bp] for Phtf1, 192 bp for EP2, and 1.09 kb for Gapds) were analyzed on agarose gels. Primers used for exon recognition were designed based on the mouse Phtf1 cDNA sequence are indicated in Table 1 and the size of the PCR products are shown in Figure 1. Amplifications were performed for 30 cycles as follows: 96°C for 15 sec, 56°C for 15 sec, and 72°C for 30 sec.
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We dissected the caput, corpus, and cauda regions of epididymis. The total RNA was isolated and prepared for Northern blot analysis. A representative blot of three independent experiments is shown in Figure 1a. Hybridization with a probe targeting the whole coding region of Phtf1 revealed a single band that predominated in the caput epididymis (Fig. 1a). Little or no expression was observed in the corpus or cauda (Fig. 1a). RT-PCR analysis confirmed that the levels of Phtf1 mRNAs were highest in the caput (Fig. 2b).
|
We then compared the structure of testicular and epididymal transcripts. Based on our estimation that the epididymal transcript level was approximately 40-fold less abundant in the epididymis than in the testis (data not shown), we loaded the gel to obtain similar quantities of the Phtf1 mRNAs to assess their respective sizes. Northern blot revealed no perceptible differences in the size of the Phtf1 messengers (approximately 3 kilobases [kb]) (Fig. 1b). A series of RT-PCRs was then used to analyze the presence of all the known coding exons. No differences were found between the Phtf1 transcripts in testis and epididymis (Fig. 1c). The splice form lacks exon 3, that had already been observed in testicular RNA (unpublished results), was also observed in the caput epididymis (Fig. 1c). Taken together, these results suggest that both messages carry the same information.
Expression of Phtf1 During Postnatal Development and Incidence of Testicular Factors
To find out whether Phtf1 is constitutively expressed in the epididymis or whether its concentration increases during postnatal development, RNA from caput and cauda epididymides of rats of different ages was assayed to determine the Phtf1 RNA level by semiquantitative reverse transcription (RT)-PCR. EP2, an androgen-dependent and caput epididymidis-specifically expressed gene [10], was used as a positive control. In the caput epididymidis, the specific Phtf1 band showed a marked rise around 4–6 wk of life and has reached adult levels at 90 days. EP2 mRNA time course showed a similar profile of expression (Fig. 2a). In the cauda epididymidis, little Phtf1 expression was detected at any stages (Fig. 2a). As circulating hormones and/or testicular factors could account for the increase in Phtf1 expression in the caput epididymidis, we measured the Phtf1 RNA level in unilaterally ligated or castrated adult male rats. Figure 2b shows typical results from four different experiments using unilaterally ligated adult male rats. The concentration of Phtf1 RNA remained relatively unchanged, indicating that efferent duct ligation had no impact on Phtf1 expression at 1 or 2 days postoperation (only Day 1 is shown; Fig. 2b). In contrast, castration caused a drastic decrease in the Phtf1 mRNA content as early as after only 2 days of treatment (data not shown), being undetectable by 4 days (Fig. 2b). EP2 control signal was no longer detectable at 4 days postcastration [10]. Amplification bands appearing in corpus could correspond to approximate dissection, as the EP2 band was also present (Fig. 2b, castrated).
Western Blot Analyses of PHTF1
Total protein extracts of the testes; the efferent ducts; the caput, corpus, and cauda regions of the epididymis; and the deferent ducts were prepared and resolved by SDS PAGE. Blots were incubated with the mp71 antibody (directed against amino acids 351–452 of PHTF1 [6]). This antibody revealed a specific reactive band of approximately 56 kDa in the caput epididymis, which was clearly shorter than the 84-kDa PHTF1 band in the testis (Fig. 3a). A 45-kDa band present throughout the epididymis was judged to be nonspecific (Fig. 3a), as it was also identified by the preimmune antiserum (data not shown). The 56-kDa epididymal protein matched the pattern of RNA expression, as it was only seen in the caput epididymidis and paralleled the rise of Phtf1 mRNAs throughout postnatal development. It first appeared in caput protein extracts of Day-30 animals and had reached its maximum by Day 90 (Fig. 3b). Castration severely affected the synthesis of the caput reactive protein, which was undetectable by Day-4 postoperation (Fig. 3b).
|
Because Phtf1 was highly conserved in mammals [4], we searched for Phtf1 expression in the canine epididymal IMCE-20 cell line isolated by Telgmann et al. [9]. RT-PCR indicated that Phtf1 mRNA was present (data not shown) and mp71 identified two specific majority bands, neither revealed by the preimmune serum (data not shown) nor in protein extracts from HeLa cells. One of these bands migrated as the 56-kDa rat epididymal band and the other migrated slightly higher than the rat testicular band (Fig. 3a). The presence of the 56-kDa epididymal protein could make using this cell line a valuable tool for studying PHTF1.
When protein extracts were treated with Endo-H or PNGase F glycosidases, both the testicular and epididymal proteins acquired a faster migrating pattern (Fig. 4a). The 84-kDa protein from the testis now migrated at about 76 kDa, indicating the presence of glycosylation cores linked to lateral asparagine chains (Fig. 4a). The epididymal protein must also have been N-glycosylated, as its apparent molecular weight shifted from 56 to 52 kDa (Fig. 4a). Neither the testicular nor the caput epididymal proteins showed any change in size after treatment using calf alkaline phosphatase (data not shown).
|
As the mRNA coding sequences for testicular and epididymal Phtf1 were similar and the size of the epididymal protein could not be assigned to the splice variants, we hypothesized that proteolysis could generate the 56-kDa isoform. Therefore, we transfected the canine epididymal cell line IMCE-20 with plasmid constructions encoding for N- or C-terminal myc-tagged Phtf1. Transfection with the N-terminal-tagged Phtf1 resulted in the appearance of a form of about 45 kDa and transfection with the C-terminal fusion revealed a 66-kDa form.
Treatment of the protein extracts using Endo-H or PNGase F glycosidases revealed a shift of the 66-kDa band to 62 kDa when the myc tags were in the C-terminal position, indicating that the original PHTF1 fragment had been glycosylated. In contrast, the 45-kDa band was not affected by this treatment, indicating that the N-terminal fragment was not modified by glycosylation (Fig 4b). As the 6-myc tags fused to these PHTF1 proteins added about 10 kDa, the N-terminal PHTF1 sequence could be expected to be about 35 kDa in size and the C-terminal deglycosylated segment to migrate at around 52 kDa. This latter segment matches that identified by the mp71 antibody in the caput epididymis (Fig. 4a) and in IMCE cells (Fig. 3a).
Epididymal 56-kDa PHTF1 Immunolocalized in the Epithelial Principal Cells
Further experiments were carried out to determine the histological distribution of the epididymal 56-kDa PHTF1. Bouin-fixed paraffin sections of adult rat epididymides were immunostained with preimmune and mp71 sera. The preimmune serum-stained sections showed uniform background labeling extending over all three regions of the epididymis (Fig. 5a). In accordance with the Western blot analysis, the mp71 serum bound to a protein present in the epithelium of the caput but not in the corpus or cauda regions (Fig. 5a). This finding confirmed that the 56-kDa PHTF1 protein was not present beyond the caput segment (Fig. 5a). Analysis of epididymides from prepubertal rats revealed that the region-specific adult pattern had already been reached as early as postnatal Day 35 (Fig. 5b). Labeling was present in the initial segment and the caput region, and a clear frontline was detected between the caput and corpus segments (Fig. 5b).
Epididymal epithelium includes at least five different cell types [11]. PHTF1 expression was confined to the principal cells of the epithelium (Fig. 6, a and b). Within these cells, the labeling was located in a supranuclear, lamellar, cup-shaped organelle that probably corresponded to the Golgi apparatus. To confirm this, epididymal histological sections were treated to obtain simultaneous staining of PHTF1 and the Golgi membrane component giantin. Confocal microscopic analysis revealed that PHTF1 was situated luminally with respect to giantin (Fig. 6d). Mouse slides stained for p58, a cis Golgi-associated protein, showed similar results (data not shown). We then looked for the correlation between PHTF1 and components of the trans-Golgi network (TGN) by using antibodies directed against TGN 38 or against the AP1 aptamer complexes associated with clathrin-coated vesicles. The mp71 antibody labeled a region situated in a cis compartment with respect to the TGN (Fig. 6e) or to the AP1 clathrin vesicles (data not shown). As PHTF1 had previously been located in a domain of the endoplasmic reticulum [6] and because this organelle was also shown to be located between the Golgi saccules and the TGN [12], we looked for the correspondence between the PHTF1 and calnexin signals. Staining caput slides with anticalnexin antibodies revealed a tubulo-vesicular network that was present throughout the cell but more concentrated in the neighborhood of the nucleus and that left an almost empty space corresponding to the location of the TGN structures (Fig. 6f). The PHTF1-positive region found correspondence with calnexin-positive structures (Fig. 6, e and f, and inner frames), indicating that PHTF1 is located in a domain of the endoplasmic reticulum. All together, these finding suggest that the PHTF1 recognized in the principal cells is in a restricted ER domain juxtaposed to the trans Golgi saccules.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Using RNA analyses, the Phtf1 transcript was detected in the caput epididymis, which matched the immunofluorescence data. Surprisingly, Western blot using protein extracts from caput epididymis revealed that this PHTF1 protein was shorter than expected. The close correlation between RNA- and protein-expression data, such as the appearance of PHTF1 during postnatal development, and the effect of castration are in favor of there being a 56-kDa PHTF1 isoform in the epididymis. As no difference could be detected between testicular and epididymal Phtf1 RNA, we suggest that the 56-kDa epididymal isoform could result from proteolytic cleavage. By being transiently transfected into epididymal IMCE cells, expression vectors containing the mouse Phtf1 cDNA linked to the myc tag at the N or C extremities, we demonstrated that the 56-kDa form could derive from a specific cleavage. The PHTF1 sequence contains several consensus sites (at least 15) that can be recognized by a furin/subtilisin-type protease. These enzymes are responsible for the cleavage of prohormones, receptors, plasma proteins, viral proteins, bacterial toxins, and growth factors [13, 14]. The high number of consensus cleavage motifs suggests that this modification could be involved in processing not only the PHTF1 epididymal isoform but also other isoforms, particularly the shorter PHTF1 protein observed in the testis [6]. PHTF1 sequence analysis revealed the presence of at least four potential N-glycosylation sites. In accordance with this prediction, both testicular and epididymal proteins were found to be posttranslationally modified by N-linked glycosylases. Adding a core composed of Glc3Man9GlcNAc2 to lateral chains of asparagine is the most frequent modification accounting in ER [15, 16]. Glycosylation of testicular PHTF1 accounted for about 8–10 kDa of the mature protein. In the epididymis, the band identified by the mp71 antibody was 4–5 kDa higher than the deglycosylated one. Investigation of PHTF1 processing in the ICME epididymal cell line indicated that the glycosylation occurred only in the C-terminal part. Taken together, these results suggest that testicular and epididymal PHTF1 isoforms are not only differentially cleaved but also glycosylated in a different way. Recently isolated murine epididymal epithelial cells [17] could to be a useful tool to study these modifications.
In the rat, PHTF1 expression rises after 4–6 wk of life, the time during which the epididymal epithelium reaches the adult morphological differentiation [11]. This differentiation coincides with a considerable increase in the levels of circulating androgens and with the arrival of the first wave of immature spermatids to the epididymis. Because castration abolishes PHTF1 expression and that efferent duct ligation does not show effect, maintenance of PHTF1 expression depends on circulating factors originating from testis. However, the addition of testosterone or dihydrotestosterone did not enhance the transcription of Phtf1 in IMCE cell lines (data not shown). Thus, for the moment, the basis of the influence of circulating factors in PHTF1 expression remains to be elucidated.
PHTF1 was detected in the principal cells of the initial segment and caput epididymis. Absorption and secretion are among the most important tasks of these cells. They possess a well-developed Golgi apparatus comprising stacks of six to eight saccules [12]. The endoplasmic reticulum is closely juxtaposed not only to the cis, but also to the trans, face of the Golgi [18]. A close relationship of this type has also been described in detail in other cells [19, 20]. We found that, in this cell, the distribution of PHTF1 matched that of the ER saccules adjacent to the trans face of the Golgi system. A similar location had previously been reported in differentiating spermatic cells [6]. The slight differences between these cells could be minimized if we bear in mind that the Golgi apparatus of the meiotic germ cells undergoes considerable remodeling. Proteins destined for the Golgi apparatus (either to remain there or to be reconducted) accumulate in ER domains immediately adjacent to the cis face of the organelle [21, 22]. Unlike the cis surface, which is known to be involved in transport, the relevance of the ER domain adjacent to the trans face of the Golgi is still obscure to date. ER is known to have associations not only with the Golgi system but also with the mitochondria, lysosomes, and plasma membrane [11, 19, 23]. Regions in close contact with mitochondria, e.g., could accumulate steroidogenic enzymes such as 3ß-hydroxysteroid dehydrogenase, which would also have rapid access to the sterol products liberated by the inner mitochondrial membrane [24]. We can thus hypothesize that PHTF1-positive structures are involved in a metabolic process that release or have direct access to a soluble product that either enters or emerges from trans Golgi saccules or TGN. The elucidation of the role of epididymal PHTF1 could greatly contribute to the knowledge of the function of this epithelium.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Raich Natacha, INSERM U 567-UMR 8104, 123 boulevard de Port-Royal, 75014 Paris, France. FAX: 33 014 325 1167; raich{at}infobiogen.fr
3 Current address: UCSF Comprehensive Cancer Center, 2340 Sutter Street, Box 0128, Room N-171, San Francisco, CA 94143-0128
Received: 18 March 2004.
First decision: 13 April 2004.
Accepted: 11 August 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
