| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Université Paris Descartes,3 Faculté de Médecine, 75014 Paris, France
Service de Biochimie et Biologie Moléculaire,4 Hôpital Lariboisière, 75010 Paris, France
Université Paris Descartes,5 Faculté des Sciences Pharmaceutiques et Biologiques, 75006 Paris, France
INSERM U566,6 CEA, 92260 Fontenay-aux-Roses, France
ABSTRACT
The first prion-like protein doppel, officially designed as prion protein dublet, does not seem to be needed for prion disease progression, whereas its physiological function seems to be related to male fertility. Its expression is primarily detected in the male genital tract, and Prnd-inactivated male mice are sterile. We investigated the location of Doppel in the testis of various species of mammal to determine its physiological function. Doppel is expressed early during ontogenesis, and is found in both germ cells and Sertoli cells in mice, rats, boars, and humans. Doppel is permanently expressed in the Sertoli cells but at different levels according to species. Its expression in testicular germ cells was primarily detected in spermatids, with a transient presence in the acrosome. These data suggest that Doppel may play a physiological role in acrosome biogenesis and may be of use in studies of patients suffering from idiopathic infertility.
acrosome reaction, fertilization, Sertoli cells, spermatid, testis
INTRODUCTION
The first prion-like protein doppel (from downstream prion-like protein), officially designated as prion protein dublet (PRND), was discovered after the study of the prion protein gene knockout (Prnp–/–) mouse lines and a large-scale sequencing program [1, 2]. The ataxia observed in these mouse lines was linked to an overexpression of the Prnd gene, which encodes PRND [1]. The introduction of a wild-type Prnp transgene corrected the ataxic phenotype, suggesting an interaction between PRND and the prion protein. In humans, PRND is located 27 Kb downstream from PRNP, the gene encoding the prion protein, and is probably a consequence of ancestral duplication of this gene.
Although PRND and the prion protein have a 25% sequence homology, mainly in the C-terminal domain, resulting in similar patterns of folding [3, 4], the two proteins exhibit different features. The mature PRND protein is about 125 amino acid residues long and has no region similar to the copper-binding octarepeat region of the prion protein. Unlike the prion protein, PRND does not seem to be needed for prion disease progression or for the generation of the pathogenic form of the prion protein, PrPSc [5, 6]. The prion protein is a ubiquitous glycoprotein that is produced in large amounts in neurons [7] whereas PRND seems to be tissue-specific [8, 9]. In mice, Prnd transcripts are principally found in the central nervous system during embryogenesis and for a short time after birth, and in the testes during adulthood [1, 2, 10]. In the testes, PRND has been found in spermatids [11]. In humans, PRND is found in adult testes—in the Sertoli cells—and also in seminal fluid and spermatozoa [12].
The function of PRND is still unknown. Therefore, several studies have analyzed mouse lines in which Prnd has been invalidated. The invalidation of Prnd (Prnd–/–) in two different mouse lines led to infertile males with different sperm phenotypes. The Prnd–/– mouse line with a 129/Ola background produced low numbers of spermatozoa with poor motility and abnormal nuclei and acrosomes [11], whereas the Prnd–/– mouse line with a mixed C57BL6/CBA background produced normal numbers of motile spermatozoa with an altered chromatin structure and DNA damage [13]. Both Prnd–/– mice phenotypes lost integrity of the sperm head. These data suggest the involvement of PRND during spermiogenesis, particularly during acrosome formation, nucleus modelling, and compaction of the chromatin.
Therefore, we have carefully studied the developmental and spatial expression patterns of PRND to determine its function. Because the cellular distribution of PRND within the testes of humans and mice is different [11, 12], we aimed to determine its testicular location in different species of mammal at different stages of testis development. Also, the sperm head alterations found in the Prnd–/– mice, which are related to the acrosome and nuclear structures, led us to carry out an ultrastructural analysis of germ cells, specifically the spermatid differentiation step.
MATERIALS AND METHODS
Samples
Five human testicular biopsies with normal spermatogenesis, and human testicular germ cells from the mechanical trituration of eight fresh biopsies in ferticult medium (FertiPro NV), were obtained from infertile patients suffering of obstructive azoospermia. All patients were followed for assisted reproduction in the Assisted Medical Procreation Centre of Cochin Hospital (Paris, France). A testicular biopsy from a 22-wk-old human fetus was a gift from Pr. J.P. Barbet of St. Vincent Hospital (Paris, France). All human biopsies were obtained after informed consent from the patients and with the approval from the Medical Ethics Committee of the Hospital Center.
Testes from one sexually immature boar and two adult boars were a gift from C. Cotinot and M. Bonneau (INRA). C57BL/6 mice and Wistar rats were purchased from Charles River Laboratories and housed at the animal house of the faculty of medicine at Paris–Descartes University. Rodent testes of sexually immature animals (one mouse of 10 days of age and one rat of 14 days of age) and mature animals (one of each species) were promptly removed and fixed or stored at –80°C. Experimental procedures were approved by the ethical committee for animal experimentation of the University of Paris–Descartes and were conducted in accordance with the Guidelines for Biomedical Research Involving Animals.
Specific Antibodies
DDC39 was a rabbit polyclonal antiserum raised against a synthetic peptide encompassing amino acids 68–88 of human PRND (Neosystems). Dop151 was a monoclonal antibody obtained by immunizing mice with recombinant human PRND (amino acids 28–152) produced in E. coli. Dop 151 specificity was verified by Western blotting (WB) on the recombinant human PRND and by protein sequencing [12]. G20, a goat polyclonal antibody raised against a peptide mapping within an internal region of human PRND, and its blocking peptide were obtained from Santa Cruz Biotechnology (Sc-16863). These three anti-PRND antibodies were tested in all the species studied by WB and immunochemistry. DDC39 recognizes PRND in all species except rat, Dop151 recognizes only the human protein, and G20 recognizes rodent and human proteins. Mouse IgG was purchased from Sigma. The pAb against P36/TPI, which stains the acrosomal membrane, was a gift from J. Auer [14]. Rabbit polyclonal antibodies against RAB6 (a Golgi vesicle marker) and against RAB5 (an endocytic vesicle marker) were obtained from Santa Cruz Biotechnology. Secondary antibodies were anti-rabbit horseradish peroxidase (HRP)-conjugated antibody (Southern Biotechnologies), goat anti-mouse 10-nm or 20-nm gold particle-conjugated IgG (B.B. International, TEBU-bio), goat anti-rabbit CY3-conjugated IgG (Jackson Immuno Research), anti-mouse HRP- and fluorescein isothiocyanate (FITC)-conjugated IgG (DAKO), and donkey anti-goat HRP-conjugated IgG (Santa Cruz Biotechnology).
Western Blotting
Tissue homogenates obtained as previously described [12] were loaded onto 12% SDS/glycine/polyacrylamide gel. Proteins were electroblotted onto Immobilon P membranes (Millipore) and PRND was detected with the indicated antibodies and corresponding HRP-conjugated secondary antibody. The blots were developed using an enhanced chemiluminescence protocol (Pierce).
Indirect Immunofluorescence
Human testicular cells were washed in phosphate buffered saline (PBS) and fixed with 1% paraformaldehyde (PFA) in PBS by incubation for 1 h at room temperature (RT). The cell suspension was washed three times with PBS (100 mM glycine was included in the second wash) and loaded into the wells of immunofluorescence slides and air-dried. After rehydratation in PBS (or, for RAB5 staining, in PBS with 0.1% Triton x100 for 10 min), the cells were preincubated for 30 min at RT with PBS containing 10% goat serum to prevent nonspecific staining. For double staining, the slides were initially incubated for 1 h with Dop151 antibody diluted 1:50 in PBS supplemented with 0.1% Tween and 1% BSA (PBSTB), washed in PBSTB and then incubated for either 1 h further at 37°C with anti-P36/TPI (diluted 1:100 in PBSTB) or anti-RAB6 (diluted 1:100 in PBSTB) or overnight at 37°C with anti-RAB5 (diluted 1:50 in PBSTB). After washing three times, the cells were incubated with FITC goat anti-mouse IgG (1:50) for 30 min, and then with CY3 goat anti-rabbit IgG (1:200) for 30 min at RT. After three further washes in PBS, the slides were covered with antifading solution (Vectashield; Vector Laboratories) and covered with a coverslip. Controls were prepared either by omitting the primary antibody or by incubating the same amount of mouse IgG instead of Dop151. The slides were viewed with an epifluorescence microscope (Nikon Eclipse E600) using adequate filters for FITC, tetramethylrhodamine isothiocyanate, or FITC/Texas red. Images were collected at 20x and 100x magnification with a digital camera (DXLM1200, Nikon) using ACT-1 (Nikon) software.
Immunohistochemistry
Three testicular biopsies were fixed in 4% PFA and embedded in paraffin. Sections (4 µm) were cut, cleared by incubation in alcohol, and rehydrated. Sections were treated either by acetone for 5 min (G20 staining) or by 3 x 5 minutes by microwave in citrate buffer (DDC39 staining) just before blockade of endogenous peroxidase activity by incubation with 3% H2O2 for 30 min. Sections were incubated for 1 h at RT with a blocking medium (1.5% goat serum in PBSTB or 4% donkey serum in 0.1% Tween in PBS) and then for 1 h further with primary antibodies (DDC39, Dop151, or G20). The sections were washed for 10 min three times in the blocking medium, incubated for 1 h with their respective HRP-conjugated secondary antibodies, and then washed three times. The sections were finally treated with diaminobenzidine by incubation in the dark for 10 min. The sections were washed and rapidly counterstained with Hemalun de Mayer (RAL) and mounted in Eukitt medium (Labonord). Negative controls were prepared by incubating sections either with the secondary antibody alone, with the primary antibody preincubated for 1 h at RT with the corresponding immunogenic peptide (10-fold-excess w/w, for G20 control), with the same amount of mouse IgG as the primary antibody (for Dop151 control), or with the preimmune serum (for DDC39 control). The sections were viewed at 20x and 100x magnification using the Nikon microscope and the images collected as described for indirect immunofluorescence (IIF).
Transmission Electron Microscopy
Two human testicular biopsies were fixed by incubation for 3 h at 4°C in 2% PFA and 0.75% glutaraldehyde in Sorensën buffer pH 7.4, treated and embedded in Unicryl resine (British Biocell International). Ultrathin sections were collected on formvar-coated nickel grids and treated with glycine-PBS (15 mg/ml). They were then incubated for 1 h with a blocking solution (5% BSA in PBS) and then overnight at 4°C with the primary antibody Dop151 or mouse IgG (1:100 dilution in blocking solution). After washing, the sections were exposed to goat anti-mouse IgG antibodies conjugated with colloidal gold particles (10 or 20 nm in size) at a 1:50 dilution by incubation for 1 h at RT. Finally, after washing, the sections were counterstained by incubation with saturated aqueous uranyl acetate for 60 min and examined under a JEOL-1010 electron microscope at 15000x or 20000x magnification.
RESULTS
Anti-PRND Antibodies Screening
All anti-PRND antibodies were tested on all species by WB and immunohistochemistry (IHC). In WB, DCC39 gave the stronger signal when compared with the other anti-PRND antibodies in all species except rat. For rat, G20 was the only antibody usable. In IHC, DDC39 gave a specific staining of PRND in boar testis when compared to the preimmune serum, and only a faint signal in humans and a background on rodent testes. G20, already used by Behrens et al. in mice [11], immunostained PRND in human and rodent testes. As for Dop151, this monoclonal antibody stained only PRND of human testis.
PRND Expression in the Testis of Various Species
We detected by WB using a pAb raised against the N-terminus part of human PRND (DDC39) a large heterodisperse band in murine, rat, boar, and human adult testis extracts. The apparent molecular weight of testicular PRND (between 23 and 42 kDa) was less heterogeneous in mice and rats compared to the other species (Fig. 1). Treatment of testicular extracts with PNGaseF generated different isoforms of PRND according to the species. In human and boar testes, we detected a major band at 16–22 kDa and a minor band at 14–15 kDa, whereas in rat and mouse we observed only the 14–15 kDa band (Fig. 1).
|
Location of PRND in Immature Testis of Different Species
We investigated the developmental expression of PRND by studying fetal and juvenile testis from humans, boars, mice, and rats (Fig. 2). In a 22-wk-old human fetus in which the seminiferous cords were already fully organized in the differentiating testis, we observed a strong PRND immunoreactivity in the cytoplasm of Sertoli cells but not in gonocytes (Fig. 2A). We found similar results in immature boar differentiating testes (Fig. 2B). However, in mice, we observed no labeling of PRND in 18-day-old fetal testes in which the seminiferous cords were beginning to be structured (data not shown). In 10-day-old mouse testis, we observed PRND immunoreactivity within the Sertoli cells of the seminiferous cords or nascent tubules but at a much lower level than in immature boar or human testes (Fig. 2C). When seminiferous tubules began to be structured in 14-day-old rats, PRND labeling was very faint (Fig. 2D).
|
Location of PRND in Mature Testis of Different Species
We confirmed the expression of PRND in the seminiferous tubules of human testes by IHC using two different anti-PRND antibodies (Dop151 and G20) (Fig. 3, A1 and A2). We observed a strong immunoreactivity in the Sertoli cells (Fig. 3, A3 and A4) but found no staining in the interstitial tissue.
|
In boar testes, the seminiferous tubules were stained by the DDC39 antibody, whereas we found no obvious staining in the interstitial tissue (Fig. 3B1). Different patterns of PRND labeling were observed depending on the tubule sections, which correspond to different stages of spermatogenesis. We observed a very strong labeling near the tubule lumen, the area in which the late elongated spermatids showed condensed nuclei (Fig. 3B2). In the sections where round or early-elongated spermatids with no condensed chromatin were observed, we found PRND staining principally in the cytoplasm of the Sertoli cells, around the germ cells (Fig. 3, B3 and B4).
In the rat and mouse testis sections, we observed PRND expression in the seminiferous tubules, with the staining pattern also being dependent on the different stages of spermatogenesis (Fig. 3, C1 and D1). Intense PRND staining was visible around the tubule lumen when late elongated spermatids were present (Fig. 3, C2 and D2) and in the residual bodies left after the spermiation process (Fig. 3, C3 and D3). We also detected PRND expression at the base of the germinative epithelium in midpachytene spermatocytes, which exhibited a granular cytoplasmic staining (Fig. 3D3). As for human and porcine testes, the Sertoli cell cytoplasm also appeared stained between the germ cells (Fig. 3, C4 and D4)
Subcellular Location of PRND in Human Testicular Germ Cells
We used IIF to investigate the presence of PRND on isolated germ cells recovered from triturated human testicular biopsies (Fig. 4). We detected either a granule-shaped (Fig. 4, A1 and A3) or a crescent-shaped (Fig. 4A2) PRND staining in the acrosomal region of round spermatids. Sometimes the staining of PRND and P36 (an acrosomal membrane antigen) was partially colocalized (Fig. 4A2). By contrast, we detected no PRND labeling in fully differentiated testicular spermatozoa (Fig. 4A4). Double labeling with PRND and RAB5, an endocytic vesicle marker [15], did not evidence any colocation of these proteins in round spermatids (Fig. 4B). In early round spermatids, we observed a separate cytoplasmic granular staining of PRND and RAB5 (Fig. 4B1). In late round spermatids, we detected PRND expression in the acrosome granule, whereas RAB5 appeared to be located at the acrosome membrane (Fig. 4, B2, B3, and B4). Some diffuse labeling of PRND and RAB5 could be also observed in the cytoplasm of late round spermatids, but without actual colocalization for these two proteins. Also, double labeling with PRND and RAB6 (a Golgi vesicle marker) on testicular biopsies did not show any colocalization in round spermatids (Fig. 4C). Unlike RAB5, RAB6 was not observed in the acrosome of spermatids.
|
We then carried out transmission electron microscopy (TEM) analysis of human testicular biopsies to determine the precise location of PRND in the acrosome of spermatids (Fig. 5). Using Clermont's classification of the different stages of human spermatid differentiation [16], we found PRND as early as stage 1 in round spermatids, within the granule located inside the acrosomal vesicle (Fig. 5, A and B). During stages 2 and 3 of round spermatid differentiation, in which the acrosome grows and spreads onto the nuclear surface, we observed PRND staining in the acrosomal matrix (Fig. 5, C, D, and E) but not in the subacrosomal perinuclear space (asterisk in Fig. 5, C and D). We also found intense PRND labeling in a cytoplasmic vesicle-like structure not far from the acrosome (arrow in Fig. 5D). At stage 4, when the nucleus begins its elongation, and later in the differentiated spermatozoa, we no longer detected PRND in the acrosome (Fig. 5G).
|
DISCUSSION
In a previous study in humans [12], we found that PRND, the first prion-like protein, was primarily located in the male genital apparatus, in both somatic (Sertoli cells) and germinal (mature ejaculated spermatozoa) cells. In other species, such as mice, PRND expression appeared in the spermatids, precisely at the late stages of spermiogenesis [11]. Such an apparent interspecies variation led us to study the location of PRND in mature and immature testes of humans, boars, and rodents (mice and rats).
By WB, we found that PRND has an apparent molecular weight ranging from 23 to 42 kDa, reflecting the high glycosylation of this protein. Murine PRND appears to be the least heterogeneous of the PRND proteins among the four species. After PNGaseF treatment, we found the same migration profile for both boar and human PRND, demonstrating that PRND is N-glycosylated in boars and suggesting, by comparison with human PRND, already reported to be N- and O-glycosylated [12], that boar PRND could also be an O-glycosylated protein. By contrast, as already reported in mice [1], PRND is apparently not O-glycosylated in rats, because N-deglycosylation led to a single molecular form of the protein.
We found that PRND is expressed in both fetal and adult testes of humans and boars. This is consistent with the expression of PRND transcripts in human fetal tissues, including the testes, and in adult tissues, exclusively the testes [17]. PRND appears to be expressed by the same cellular type, such as Sertoli cells, at these two different periods of life. The expression of PRND in the Sertoli cells of human fetal testes (present study) and of testis of patients suffering from Sertoli-cell-only syndrome [12] suggests that PRND expression in Sertoli cells is independent of the spermatogenesis.
Analysis of the location of PRND in the testes of other species, such as rats, mice, and boars (this study), and cattle [18], showed that no differences exist between the species. However, the PRND staining intensity of the Sertoli cells, both during testicular differentiation and in adulthood, varies among species. Although Behrens et al. [11] did not mention the presence of PRND in the Sertoli cells of adult mouse testes, Westaway et al. [19] reported low PRND staining of Sertoli cells of testes chemically depleted of germ cells, consistent with our observations. Whether this variation in PRND staining reflects true differences in the levels of PRND expression or relies on the immunodetection remains to be determined.
Although the principal location of PRND in humans appear to be the Sertoli cells, in the present study we also observed labeling of PRND in the acrosome and cytoplasm of spermatids isolated from different human testicular biopsies. The acrosomal labeling had both a granular form and a crescent form, the latter being similar to the IIF staining of a condensed cytosolic structure located between the inner acrosomal membrane and the nuclear envelope of spermatids and implicated in acrosome-nuclear docking [20]. Our TEM observations allowed us to rule out a similar subacrosomal location for PRND. Indeed, during the acrosome formation in round spermatids PRND was found within the acrosomal granule/matrix, with PRND becoming undetectable once the acrosome has finished spreading over the nucleus in early elongating spermatids. This transient locating of PRND in the acrosomal matrix is unusual because PRND has been described as a GPI-linked membrane protein [1], as seen in neuroblastoma cell lines transfected with Prnd [8] and in human ejaculated spermatozoa [12].
The origin of acrosomal PRND still needs to be determined. It is unknown whether PRND travels from the Golgi, as already described for other acrosomal proteins [21], or from the plasma membrane via an endocytic pathway [22], as hypothesized for a PRND form delivered by the Sertoli cells. Unfortunately, double labeling of round spermatids with PRND and RAB6, a Golgi-acrosome vesicular traffic marker, or RAB5, an early endocytic pathway marker, failed to detect any colocalization of PRND with these RAB proteins and did not allow any conclusion to be drawn on that point. However, PRND labeling observed within spermatocytes in rodents suggests a PRND synthesis by the germ cells themselves rather than endocytosis of PRND delivered from the Sertoli cells. Interestingly, Pou4f1 (synonym: Brn3a) and Pou4f2 (synonym: Brn3b) transcription factors involved in the transcriptional regulation of PRND [23] are also expressed in the spermatogonia and spermatids of rat testis [24]. Whether these factors also participate in the regulation of Prnd transcription in the testis remains to be investigated.
In the present study, we showed that PRND was transiently present in spermatids and was not detected in fully differentiated testicular spermatozoa. Interestingly, we previously found PRND on epididymal and ejaculated spermatozoa [12]. One explanation could be that PRND was acquired during the passage of the maturing spermatozoa through the epididymis, as is known to happen for other GPI proteins anchored on spermatozoa. In this respect, it is interesting to note that we have observed a staining of PRND in epithelial cells of the boar epididymis (unpublished data), suggesting a possible epididymal origin of PRND.
Currently, PRND function remains unknown. The double location in both germinal and somatic cells from the male genital tract suggests that PRND plays a role in male fertility. Two Prnd–/–mice lines have been constructed [11, 13]. In both lines, the mice survive until adulthood without displaying any objective abnormalities or developmental problems. The males exhibited normal testis development but were finally infertile. Thus, at least in mice, PRND seems to be essential for male fertility but not for reproductive apparatus development. These two infertile transgenic mouse lines mainly show abnormalities in spermiogenesis, with large consequences for fertilization [11] or early embryo development [13]. The transient presence of PRND that we observed within the developing acrosome in human spermatids, together with the lack of a well-formed acrosome in spermatids from Prnd–/– mice in the study of Behrens et al. [11], strongly suggests a role of PRND in acrosome biogenesis. The other Prnd–/– mouse line resulted in spermatozoa with morphologically normal acrosome but with abnormally high levels of DNA strand breaks (DSBs) [13]. In germ cells, transient DSBs have normally been found both in early elongating spermatids during the chromatin condensation process and in primary spermatocytes during meiotic recombinations [25]. Because we detected PRND in these two cell types, the possible role of PRND in these processes should be investigated.
In conclusion, we studied the location of PRND in different species, and confirmed its dual expression in both Sertoli cells and germ cells. In the Sertoli cells, PRND is permanently expressed, but at different levels according to the species, whereas its expression in testicular germ cells was detected after puberty, principally in spermatids with a transient presence in the acrosome. These data show that PRND may be implicated in normal acrosome genesis and may be of use to elucidate the cause of certain idiopathic infertility.
FOOTNOTES
1 Supported by funds from Université Paris 5 (Bonus Qualité Recherche) and grant EA1752 from the Direction de la Recherche et des Etudes Doctorales. 
2 Correspondence: Catherine Serres, Laboratoire de Biologie de la Reproduction, Faculté de Médecine, site de Cochin, 24 rue du Faubourg Saint-Jacques, 75014 Paris, France. FAX: 33 1 44 41 23 02; catherine.serres{at}univ-paris5.fr
Received: 28 September 2005.
First decision: 20 October 2005.
Accepted: 17 January 2006.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
