| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute for Anatomy and Cell Biology II,4 Justus Liebig University, 35385 Giessen, Germany
Department of Anatomy and Cell Biology,5 Phillips University, 35037 Marburg, Germany
Department of Veterinary Anatomy, Histology and Embryology,6 Justus Liebig University, 35392 Giessen, Germany
ABSTRACT
The vital importance of peroxisomal metabolism for regular function of the testis is stressed by the severe spermatogenesis defects induced by peroxisomal dysfunction. However, only sparse information is available on the role and enzyme composition of this organelle in distinct cell types of the testis. In the present study, we characterized the peroxisomal compartment in human and mouse testis in primary cultures of murine somatic cells (Sertoli, peritubular myoid, and Leydig cells) and in GFP-PTS1 transgenic mice with a variety of morphological and biochemical techniques. Formerly, peroxisomes were thought to be absent in late stages of spermatogenesis. However, our results obtained by detection of different peroxisomal marker proteins show the presence of these organelles in most cell types in the testis, except for mature spermatozoa. Furthermore, we demonstrate a strong heterogeneity of peroxisomal protein content in various cell types of the human and mouse testis and show marked differences in structure, abundance, and localization of these organelles in spermatids, depending on their maturation. Highest and selective enrichment of the peroxisomal lipid transporters (ABCD1 and ABCD3) as well as ACOX2, the key regulatory enzyme of the beta-oxidation pathway 2 for side chain oxidation of cholesterol, were found in Sertoli cells, whereas Leydig cells were enriched in catalase and ABCD2. Our results suggest a cell type-specific metabolic function of peroxisomes in the testis and point to an important role for peroxisomes in spermiogenesis and in the lipid metabolism of Sertoli cells.
Leydig cells, Sertoli cells, spermatid, spermatogenesis, testis
Metabolic pathways of peroxisomes are of vital importance for normal spermatogenesis and regular functions of the human testis. This is accentuated by the impaired spermatogenesis and infertility in adult patients with peroxisomal single-enzyme deficiencies, such as X-linked adrenoleukodystrophy (X-ALD) or adrenomyeloneuropathy (AMN, a milder phenotype of X-ALD) [1]. Patients with X-ALD or AMN suffer from a single enzyme defect of ABCD1, an ABC transporter on the peroxisomal membrane (previously known as ALDP) involved in the transport of very long-chain fatty acids (VLCFAs) into the peroxisomal matrix [2]. Men with X-ALD exhibit an adreno-testiculo-leukomyelo-neuropathic complex of symptoms [3], with impairment of the testicular functions occurring in 80% of these patients. Histological analysis of testicular tissue from X-ALD or AMN patients exhibited hypocellularity in the seminiferous tubules, a maturation arrest of spermatogenesis in distinct stages or a "Sertoli cells and spermatogonia"-only phenotype [1]. Similarly, patients with mild forms of peroxisomal biogenesis disorders surviving to adolescence show hypocellular seminiferous tubules with spermatogenesis arrest, vacuolated Sertoli cells, and a complete degeneration of Leydig cells [4]. Moreover, severe forms of peroxisomal biogenesis disorders lead to cryptorchism [4], suggestive of an important function of peroxisomal metabolism also in the development of the testis. Despite these deleterious effects of peroxisomal diseases on development, integrity, and function of the adult testis (fertility), very little is known about peroxisomal metabolism in this organ.
Peroxisomes are ubiquitous cell organelles, generally involved in the metabolism of reactive oxygen species (ROS) and of many lipid derivatives [5]. They are highly versatile organelles, the enzyme composition of which varies in distinct species, organs, or individual cell types, as well as in response to environmental or developmental conditions [6, 7]. Numerous proteins of the peroxin family are required for proliferation and regular biogenesis of mammalian peroxisomes, including the import of peroxisomal matrix proteins (e.g., PEX13 and PEX14 or Pex13p and Pex14p according to the unified peroxin nomenclature [8]; Fig. 1). Peroxisomes house two pathways for β-oxidation (Fig. 1), with a much broader substrate specificity than mitochondria. The two peroxisomal β-oxidation systems consist of a classical peroxisome proliferator-activated receptor-inducible pathway for catabolism of straight-chain acyl-CoAs and eicosanoids by fatty acyl-CoA oxidase (ACOX1), multifunctional protein 1 (EHHADH, also known in the peroxisome field as MFP1) and thiolase A, and a second noninducible pathway catalyzing the oxidation of the cholesterol side chain and 2-methyl-branched fatty acyl-CoA by a branched-chain acyl-CoA oxidase in humans (ACOX2 and ACOX3 in mice), multifunctional protein 2 (HSD17B4, also known as MFP2), and sterol carrier protein X (SCPX). These pathways are capable of degrading a range of bioactive lipids (e.g., leukotrienes and prostaglandins) mediating inflammation, as well as arachidonic acid and oxysterols involved in signaling, and they are also involved in the synthesis of poly-unsaturated fatty acids, which are implicated in signaling processes and apoptosis. In addition, the peroxisomal β-oxidation pathway 2 is involved in the conversion of active gonadal steroids into inactive forms [9]. Peroxisomes are also involved in the synthesis of isoprenoids, such as retinoic acid derivatives, and important membrane lipids, such as cholesterol or plasmalogens. The initial steps of plasmalogen biosynthesis (Fig. 1) are catalyzed by dihydroxyacetonephosphate acyltransferase (GNPAT, previously known as DHAPAT) and alkyl-dihydroxyacetonephosphate synthase (AGPS, previously known as DHAPS), both of which are matrix proteins that are closely associated with the membranes of the peroxisome (for a review on peroxisomal metabolism, see Wanders and Waterham [5]). ABC transporters (ABCDs 1, 2, 3, and 4; Fig. 1) are involved in the transport of metabolites across the peroxisomal membrane.
|
In the testis, peroxisomes have been investigated by routine electron microscopy [10] or by visualization of the marker protein catalase [11–15]. Based on these studies, peroxisomes of the testis were thought to be restricted to Leydig and Sertoli cells. Leydig cell peroxisomes were shown to proliferate upon LH treatment, whereas LH deprivation resulted in a significant decrease in the number of these organelles. In addition, an increase in free cholesterol was noted in peroxisomes and mitochondria after LH treatment. Due to these facts, Mendis-Hadagama and colleagues have speculated that testicular steroid synthesis could occur at least in part in peroxisomes [16].
Only recently, work from our groups has revealed that peroxisomes in the testis are present also in spermatogonia in addition to Leydig and Sertoli cells [17–19]. In the present study, we have characterized this cell compartment with more sensitive morphological methods for various peroxisomal markers in situ in mouse and human testis, in GFP-PTS1 transgenic mice, and in primary cell cultures of Sertoli, peritubular myoid, and Leydig cells. Our results show for the first time the presence of the peroxisomal compartment also in later stages of spermatogenesis in the adluminal region of the seminiferous epithelium. We were able to demonstrate maturation-dependent alterations of the peroxisomal compartment during spermatogenesis. In addition, we present evidence for a specific role of Sertoli cell peroxisomes for lipid metabolism in seminiferous tubules.
Animals and Human Tissue Material
Human. Testis biopsies were obtained after written informed consent, immersion fixed with Bouin fixative, and embedded in paraffin. All biopsies used in our study were from the biopsy and tissue repository of the Hessian Center for Reproductive Medicine in Giessen, Germany. General approval for the repository biopsy collection has been granted by the ethics committee of the Medical Faculty of the Justus Liebig University in Giessen. The three biopsies analyzed were from 35- and 39-year-old men and were diagnosed as "normal spermatogenesis" based on histopathological analysis.
Mice. Male C57BL/6J (Charles River Laboratories, Sulzfeld, Germany) mice at the ages of 4–6 mo and 14-day-old animals (for isolation of cells) were used for all experiments. The animals were delivered 2 days prior to experiments and housed under standard conditions with free access to standard laboratory food and water and a 12L:12D cycle. Experiments with laboratory mice were approved by the Government Commission of Animal Care Germany.
Generation of GFP-PTS1 Transgenic Mice
A fusion protein of the green fluorescent protein (GFP) and the peroxisomal targeting signal 1 (PTS1) is frequently used for visualization of peroxisomes in living cells [20]. The transgenic mouse line used in our study has been generated by injecting a GFP-PTS1 cDNA fragment under the control of the murine Rosa26 promoter [17] into the pronucleus of CD1 mouse zygotes. Pronucleus injection and generation of the transgenic mouse line has been performed in the laboratory of A. Zimmer (Department of Neurobiology, University of Bonn, Germany), and further details on this transgenic mouse line will be published elsewhere (Lüers et al., unpublished results).
Processing of Testes for Paraffin Embedding and Sectioning
Five C57BL/6J mice were anesthetized by intraperitoneal injection and perfused through the heart with 4% depolymerized paraformaldehyde containing 2% sucrose in PIPES or PBS, pH 7.4. After fixation, testes were removed and immersion fixed in the same fixative overnight. Complete testes were embedded into paraffin (Paraplast; Sigma, St. Louis, MO) using a Leica TP 1020 automated vacuum infiltration tissue processor (90 min each: 70%, 80%, 90%, 3x 100% ethanol; 2 h each: 2x xylene, 2x paraffin). Paraffin blocks of testes were cut on a Leica RM2135 rotation microtome into sections of 1–3 µm thickness.
An indirect immunofluorescence (IF) protocol was established for immunolabeling of paraffin sections. Sections were deparaffinized and rehydrated. For improved retrieval of peroxisomal antigens and accessibility of epitopes, deparaffinized and rehydrated testis sections were subjected to digestion with trypsin for 15 min at 37°C, followed by microwave treatment for 15 min at 900 W in 10 mM citrate buffer at pH 6.0 (modified according to Grabenbauer et al. [21]). Nonspecific binding sites were blocked with 4% BSA in Tris-buffered saline containing 0.05% Tween 20 (TBST) for 2 h at room temperature, and sections were incubated with primary antibodies in 1% BSA in TBST overnight at 4°C. On the following day, the sections were incubated with fluorochrome-conjugated secondary antibodies. For a complete summary of all antibodies, suppliers, and functions of antigens, see Table A in Supplemental Data available at www.biolreprod.org. Since individual, specific preimmune sera were not available for most antibodies, negative controls were processed in parallel 1) by addition of TBST buffer instead of the first antibodies or 2) by antigen preabsorption of the first antibody (see below and Fig. A in Supplemental Data available online at www.biolreprod.org). Nuclei were visualized with 1 µM TOTO-3 iodide for 30 min at room temperature (Molecular Probes/Invitrogen, Carlsbad, CA). Thereafter, samples were inspected with a Leica fluorescence microscope, and the best preparations were used for confocal laser scanning microscopy (CLSM) with a Leica TCS SP2. Table B in Supplemental Data summarizes the results of immunofluorescence preparations of paraffin sections for all used peroxisomal marker proteins.
Analysis of the Specificity of Catalase Antiserum by Antigen Competition
The polyclonal antiserum against catalase (dilution range 1:100–1:1000; Polysciences Inc.) was preincubated with bovine liver catalase at a final concentration of 6.45 mg/ml (Sigma) for 1 h and centrifuged at 13 000 x g for 15 min at 4°C (Eppendorf centrifuge), and the depleted supernatant was used for immunostaining experiments. Paraffin sections were incubated overnight in parallel 1) with supernatant from the catalase-preabsorption procedure or 2) with the regular antiserum against catalase. Subsequently, sections were incubated with AlexaFluor488-conjugated secondary anti-rabbit antibodies (dilution 1:200) for 1 h, followed by washing in TBST. For results on negative controls, see Figure A in Supplemental Data.
Processing of Transgenic Testes for Immunofluorescence on Frozen Sections
Testes of GFP-PTS1 transgenic mice were fixed by perfusion via the heart with 4% (w/v) paraformaldehyde in 0.15 M HEPES, pH 7.4. Testes were excised and immersed in the same fixative overnight. Thereafter, they were incubated in 25% sucrose for about 2 days and subsequently frozen and stored at –80°C. Cryosections obtained on a Leica microtome CM 3050 were either directly analyzed by CLSM to monitor GFP fluorescence or were subjected to immunofluorescence using antibodies against Pex14p as described above without antigen retrieval and lower detergent concentrations.
Electron Microscopy—Cytochemical Localization of Catalase Activity with the Alkaline DAB Method
Three mice were anesthetized and perfused via the left ventricle with a mixture of 4% depolymerized paraformaldehyde, 0.05% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4), and 2% sucrose. After fixation, testes were carefully removed, cut in slices with razor blades, postfixed in 1% glutaraldehyde in cacodylate buffer (pH 7.4) for 15 min, and washed three times for 5 min with 0.1 M cacodylate buffer. For cytochemical localization of catalase, specimens were incubated for 3 h at 45°C in the alkaline 3,3-diaminobenzidine (DAB) medium [22], followed by postfixation in 1% aqueous osmium tetroxide overnight. Samples were dehydrated in a series of graded ethanol (70% to absolute) and embedded in Epoxy resin (Agar 100 Resin Kit, Agar Scientific Ltd., Essex, United Kingdom). Ultrathin sections were inspected after contrasting with a LEO 906 electron microscope (Zeiss, Oberkochen, Germany).
Three control (C57BL/6J) and three GFP-PTS1 transgenic mice were anesthetized and perfused, and testes were fixed as described above. Fixed testes were cut into slices with a razor blade and embedded into LR White resin (medium grade) [23]. LR White-filled gelatin capsules were polymerized at 50°C for 3 days. After preparation of semithin sections, blocks were trimmed for areas with defined stages of seminiferous tubules. Ultrathin sections of 80 nm were cut, collected on 100 mesh nickel grids, and thereafter coated on the back side with a 1% formvar film. The grids were dried at 37°C overnight prior to immunostaining. The sections were incubated with blocking solution (1% BSA in TBST) for 30 min at room temperature. Incubation with the primary antibodies was performed on droplets with antibodies (anti-GFP, 1:100 to 1:2000; anti-catalase, 1:500 to 1:10 000; and anti-PEX13, 1:500 to 1:5000) in 0.5% BSA in TBST overnight at room temperature. Thereafter, the sections were incubated with a protein A-gold solution (15-nm colloidal gold particles) for 1 h at room temperature [24]. Negative controls were processed in parallel 1) by addition of TBST buffer instead of the first antibodies or 2) by antigen preabsorption of the first antibody (catalase preabsorption of the anti-catalase antibody; see above). The grids were rinsed on droplets of TBST and subsequently contrasted with uranyl acetate for 2 min and lead citrate for 45 sec. The sections were examined using a LEO 906 electron microscope.
Isolation and Culture of Leydig Cells
Adult (10 mice) and 14-day-old (15 mice) C57BL/6J mice were killed by cervical dislocation, and their testes were removed aseptically. The tunica albuginea was carefully removed, and seminiferous tubules and interstitial cells were dispersed by treating the decapsulated testes with collagenase A (1 mg/ml), hyaluronidase (1 mg/ml), and DNase (20 µg/ml) in Dulbecco modified Eagle medium/Ham F-12 (DMEM/F12; 1:1 [v/v]; Invitrogen, Carlsbad, CA) with 10 mM HEPES (pH 7.4) at 34°C for 20 min in a shaking water bath. All following procedures were carried out under sterile conditions. Isolation of Leydig cells was done according to the method of Schumacher and colleagues with only minor modifications [25]. Detailed protocols for isolation of primary cell lines are given in the online Supplemental Data. Leydig cells were taken for experiments after 3 days of culture. The purity of the resulting Leydig cell preparation was determined by indirect immunofluorescence with an antibody against mitochondrial cytochrome P450 side chain cleavage enzyme, a marker specific for Leydig cells. A total of 85%–90% of the cells were cytochrome P450 positive, indicative of a high enrichment of Leydig cells within the preparation (Fig. C in Supplemental Data). Leydig cells have been collected from three distinct sets of experiments.
Isolation and Culture of Sertoli Cells and Peritubular Myoid Cells
Sertoli cells and peritubular myoid cells from 15 mice (14-day-old C57BL/6J) were isolated essentially according to Monsees and colleagues [26] (for Methods details, see online Supplemental Data). The identity of these cells as peritubular myoid cells was based on phase-contrast morphology and indirect immunofluorescence staining using anti-
-smooth muscle actin as a specific cell marker. The purity of peritubular myoid cells was higher than 95%. To remove germ cells and increase the purity of the Sertoli cell preparation, after 3 days of culture the Sertoli cell monolayer was subjected to hypotonic shock by incubation with 20 mM Tris-HCl (pH 7.5) for 5 min at room temperature [27]. The hypotonic solution was replaced with medium (lacking cytosine arabinoside). The medium was exchanged every day, and the Sertoli cells were used for experiments after another 3 days in culture. The identity of these cells as Sertoli cells was based on immunostaining for vimentin as a specific cell marker. The purity of the Sertoli cell cultures was higher than 95% (see Fig. C in Supplemental Data). Sertoli and peritubular myoid cells were collected from three distinct sets of experiments.
Isolation of Enriched Peroxisomal Fractions from Primary Leydig, Peritubular Myoid, and Sertoli Cells
Distinct cell preparations (18 x 106 Sertoli cells, 12 x 106 Leydig cells, 1 x 107 peritubular myoid cells) were homogenized in homogenization medium (HM: 150 µl of 5 mM MOPS, pH 7.4, 250 mM sucrose, 1 mM EDTA, 0.1% [v/v] ethanol, 0.2 mM dithiothreitol, 1 mM 6-aminocapronic acid) supplemented with protease inhibitors (10% protease inhibitor mix M; Serva, Heidelberg, Germany) with a single stroke (2 min, 1000 rpm) using a Potter-Elvehjem homogenizer (Potter-S; B. Braun, Melsungen, Germany). The homogenate was centrifuged at 1900 x g for 10 min. The resulting supernatant (S1a) was kept on ice, and the pellet was resuspended in 100 µl HM and recentrifuged at 1900 x g, resulting in the supernatant (S1b) and a pellet (P1) with large mitochondria and nuclei. The combined supernatants S1 (S1a and S1b) were further subjected to centrifugation at 50 000 x g for 20 min to yield the enriched peroxisomal fraction (pellet) and the supernatant S2a. The enriched peroxisomal pellet was resuspended in 100 µl HM and recentrifuged again at 50 000 x g for 20 min, yielding the enriched peroxisomal fraction (P2) and the supernatant S2b. The supernatants S2a and S2b were combined (S2). Fractions S1, P1, S2, and P2 were analyzed by Western blotting. The enriched peroxisomal fraction is a mixed-organelle fraction (light mitochondrial fraction [LM], also known as D-fraction) containing a high amount of peroxisomes as well as mitochondria, lysosomes, and a lower amount of microsomal vesicles [28].
Western Blot Analysis and Relative Quantification
Protein concentration was determined according to Bradford [29] using BSA as standard. Protein samples (10 µg) were separated on 12% SDS-polyacrylamide gels and transferred onto polyvinylidene fluoride membranes. Nonspecific protein-binding sites were blocked with Tris-buffered saline (TBS) containing 10% nonfat milk powder and 0.05% Tween-20 (blocking buffer). The blots were incubated for 2 h at room temperature with primary antibodies and thereafter for 1 h at room temperature with alkaline phosphatase-conjugated secondary antibodies. Alkaline phosphatase activity was detected using the Immun-StarTM AP (no. 170-5018) substrate from Bio-Rad (München, Germany) and exposure of the blots to Kodak Biomax MR films. Bands on films were quantified with the Gel Doc 2000 system from Bio-Rad. All Western blot analyses were performed three times and represent data from three individual experiments.
RNA Analysis by Semiquantitative RT-PCR
Total RNA was prepared from Leydig, peritubular myoid, and Sertoli cells using the RNeasy kit (Qiagen, Hilden, Germany). First-strand cDNA was synthesized from 3.5 µg total RNA with oligo(dT)12–18 primers using Superscript II reverse transcriptase (no. 18064–022; Invitrogen, Karlsruhe, Germany). The PCR was set up in a final volume of 50 µl using 2 µl cDNA. All primers were tested and PCR conditions optimized with gradient PCR on a Bio-Rad iCycler prior to parallel analysis of cDNA samples from distinct testicular cells. Primer sequences are summarized in Table 1. Bands on gels were quantified with the Bio-Rad Gel Doc 2000 system. All RT-PCR experiments were performed three times and represent data from three individual RNA isolation experiments.
|
Knowledge of the distribution and enzyme composition of peroxisomes is necessary to understand the physiological role of this subcellular compartment in distinct testicular cell types. It is also a prerequisite for investigations on the molecular pathogenesis of testicular pathologies and of impaired fertility due to peroxisomal dysfunction. Since only sparse information is available on peroxisomes in the testis, we have characterized this cell compartment in mouse and human testis using a variety of peroxisomal markers with morphological, biochemical, and molecular biological techniques.
Peroxisomal Proteins Are Heterogeneously Distributed in Distinct Cell Types of the Mouse Testis
Catalase in general is an abundant peroxisomal marker protein and has been frequently used for the detection of peroxisomes by immunohistochemistry (IHC) on paraffin sections or by cytochemical activity staining for this enzyme on the ultrastructural level in a variety of tissues [30]. Using peroxidase-based IHC (ABC method), we could confirm that catalase was present in Leydig cells (Fig. B in Supplemental Data). By using immunofluorescence for the localization of several peroxisomal marker proteins, however, we could detect peroxisomes also in somatic cells (Sertoli cells and peritubular cells) and germ cells (spermatogonia, spermatocytes, and round and elongated spermatids) in the seminiferous tubules of adult mouse testis (Fig. 2). In agreement with the peroxidase-based IHC results, catalase immunoreactivity was most intense in Leydig cells in immunofluorescence preparations. In addition, a punctuate peroxisomal staining pattern could also be observed in the basal compartment of the germinal epithelium and in peritubular myoid cells (Fig. 2A). Only with very high concentrations of the catalase antibody (1:100) and prolonged exposure times, leading to overexposure of Leydig cells in the images, was a weak punctuate staining for catalase also seen in spermatocytes and spermatids (Fig. A in Supplemental Data). A comparable distribution of immunoreactivity was observed for peroxisomal thiolase A, with strong signal in Leydig cells and a fine punctuated staining of lower intensity in the germinal epithelium (Fig. 2B). In comparison to catalase, clear thiolase A immunoreactivity was present in a punctuate pattern in suprabasal layers of the germinal epithelium. In contrast, the peroxisomal ABC transporter ABCD3, which is one of the most abundant integral membrane proteins of peroxisomes in hepatocytes, was expressed in the periphery of seminiferous tubules, with highest abundance in Sertoli cells (Fig. 2C). In Leydig cells, ABCD3 was barely detectable. ABCD1, a second ABC transporter in the peroxisomal membrane, was selectively enriched in Sertoli cells, as shown by an ABCD1/vimentin double-immunofluorescence staining (Fig. 2D).
|
In comparison to the above-mentioned metabolic enzymes and transporters, the peroxisomal biogenesis proteins PEX13 and PEX14 were detected in all cell types of murine testis, except mature spermatozoa (Fig. 2, E and F). However, expression patterns of both proteins with respect to signal intensities were different in distinct cell types. PEX13 was most abundant in germ cells, with weaker staining of Sertoli, peritublar myoid, and Leydig cells (Fig. 2E), whereas PEX14 was most abundant in the basal compartment of the germinal epithelium, with significant labeling also of Leydig cells (Fig. 2F).
In addition to individual small peroxisomes, large and strongly immunoreactive structures were observed with all antibodies against peroxisomal proteins at the luminal surface of the germinal epithelium in the region of late spermatids (Figs. 2E and 4).
|
Our results with antibodies against peroxisomal proteins were further substantiated by fluorescence analysis of cryosections of GFP-PTS1 transgenic mice, in which the GFP is targeted to the peroxisomal matrix via the C-terminal peroxisomal targeting signal 1. This transgenic mouse strain exhibits high expression levels of the GFP transgene in all germ cells, allowing straightforward detection of peroxisomes in frozen sections without further embedding and antibody labeling procedures. As depicted in Figure 2G, import competent peroxisomes are present throughout the germinal epithelium.
As an internal control for organelle distribution, we also detected mitochondria with an antibody against complex III of the respiratory chain (OxPhosIII; Fig. 2H). In comparison to peroxisomal enzymes, mitochondrial complex III was enriched in spermatocytes I and also was abundant in the mitochondrial sheath of step 16 spermatids (Fig. 3B).
|
Cell Type-Specific Differences in Abundance of Peroxisomal Proteins Are Conserved Between Mouse and Man
Indirect immunofluorescence preparations of paraffin sections of human testis with antibodies against different peroxisomal marker proteins showed a staining pattern similar to that observed in adult mouse testis (Fig. 3). Catalase was mainly detected in Leydig cells (inset) and the basal region of the seminiferous tubules (labeling in Sertoli cells; Fig. 3, A and B). In contrast to catalase and similarly to mouse samples, mitochondrial complex III was clearly detectable in spermatocytes I (Fig. 3B). The distribution pattern of Acyl-CoA oxidase I, the rate-limiting enzyme of the β-oxidation pathway I, was almost identical to that of catalase, with strongest abundance in Sertoli cells (Fig. 3C). Similarly to mouse preparations, peroxisomal thiolase A, the terminal enzyme of the β-oxidation pathway I, could be detected in addition to Sertoli cells also in suprabasal layers of the germinal epithelium in human testis (Fig. 3D). Furthermore, the peroxisomal biogenesis proteins PEX13 and PEX14 showed protein abundance patterns similar to those in mouse testis, with labeling of all cell types in different intensities (Fig. 3, E, F, and inset in G). Specific staining of lysosomes and of autophagic vacuoles using anti-LAMP2 was strongest in Sertoli cells in addition to the labeling of the acrosomes in spermatids (Fig. 3F), and it did not colocalize with peroxisomal marker proteins (PEX13, PEX14). Similarly to murine testis, both ABC transporters, ABCD1 and ABCD3 (Fig. 3, G and H), were selectively enriched in Sertoli cells. Double immunofluorescence with ABCD1/vimentin revealed an almost exclusive localization of ABCD1 in Sertoli cells (Fig. 3H). Large aggregates, similar to those seen in mouse testis, were also present in human samples in late spermatids (Fig. 3F: PEX14 and Fig. 3G: ABCD3 and inset PEX13).
Peroxisomes Aggregate in Clusters During Spermatid Maturation
For analysis of alterations of the peroxisomal compartment during spermatogenesis or different steps in spermiogenesis, distinct stages of the seminiferous tubules must be compared. In mice, the process of spermatogenesis progresses along the longitudinal axis of the seminiferous tubules, and the synchronization of the spermatogenic cycle allows for the classification of different tubule segments in 12 distinct stages ([31]; for a review, see Russel et al. [32]). Since PEX14 labeling was most sensitive for the identification of peroxisomes in all cell types of the seminiferous tubules, we have used immunofluorescence preparations of paraffin sections with this marker for analysis of peroxisomal alterations during the spermatogenic cycle (Fig. 4, A–C) or a combination of fluorescence analysis of cryosections of GFP-PTS1 transgenic mice with PEX14 immunolabeling (Fig. 4, D, E, G, and H). During the course of spermiogenesis, peroxisomes could be clearly identified as single organelles in round and early elongating spermatids (steps 1–13). In contrast, less numerous, large and intensely labeled peroxisomal structures appeared in late elongated spermatids (steps 15 and 16). Colocalization of PEX14 and GFP-PTS1 in the same particles verified the peroxisomal nature of these structures (Fig. 4, D and E). Similar structures were also labeled with catalase and ABCD3 antibodies (Fig. 5, A and B). Higher-magnification images revealed aggregates and network-like structures positive for catalase and ABCD3. Similar peroxisomal aggregates were also found in PEX13 and PEX14 preparations (Figs. 2E and 3, F and G inset). Upon careful analysis of peroxisomal aggregates in spermatids of distinct stages of the seminiferous epithelium, a significant difference in the number and spatial localization of peroxisomal structures with respect to the nuclei was noted (Fig. 4, D–F). During the progress of spermatid maturation, peroxisomes disappeared as individual organelles (stages II–III; Fig. 4, A and D), decreased in number, and aggregated to larger clusters (stages VI–VII; Figs. 4, B and E, and 5, A and B). Furthermore, they were transported from central regions in the spermatid cytoplasm to a basal location beneath the nuclei of the mature spermatids of step 16 (stage VIII; Fig. 4F). In addition, large peroxisomal aggregates were also found in residual bodies (Fig. 4G, arrows).
|
All results obtained by fluorescence microscopy were corroborated by ultrastructural analysis. The specificities of all antibodies were high on the ultrastructural level in immunocytochemical preparations as shown by selective staining of peroxisomes for catalase or PEX13 in various testicular somatic cell types (Fig. 5, I, M, and N). In addition, we could show immunoreactivity of catalase in small, elongated peroxisomes in spermatogonia (Fig. 5L). Furthermore, we found specific labeling for GFP and PEX13 on membrane-bound structures resembling peroxisomes in all stages of spermatogenesis and spermiogenesis (except for mature spermatozoa). Only rarely were single nonspecific gold particles found in appropriate negative controls for all antibodies (data not shown).
Peroxisomes in germ cells were often elongated and dump-bell shaped and were similar to or even smaller in diameter (50–100 nm) than segments of the endoplasmic reticulum. Similar to light microscopic results, their distribution and shape changed depending on the maturation of spermatids. In spermatid development up to step 13 (Fig. 5H, step 9) peroxisomes appeared as small, individual structures. Individual peroxisomes in early stages of spermatid development were difficult to identify in postembedding labeling experiments, since they are very small and were only rarely exposed on the surface of ultrathin sections in these cell types. To obtain a rough estimation of the probability of the presence of peroxisomes on the surface of these ultrathin preparations, we counted the peroxisome number in 100 round spermatids in a paraffin section stained for PEX14 using regular fluorescence microscopy (number of peroxisomes in 5 x 20 round spermatids of five distinct seminiferous tubules). In 100 spermatid profiles, 1874 fluorescent particles were present (range of 15–23 peroxisomes per spermatid profile). Thereafter, a thickness of 1.3 µm for this section was determined by an xzy-scan (vertical scan) with a CLSM (pinhole: airy 1, objective: 63x, zoom: 8). By mathematical extrapolation, this would implicate for a DAB-stained ultrathin section of 80-nm thickness a value of 0.92–1.42 peroxisomes per round spermatid profile and a minimum probability of 0.0115–0.0178 (value for DAB sections divided by 80-nm section thickness) on the surface of postembedding labeling preparations (a single peroxisome on the section surface per 56 to 87 spermatids). These derived, nonempirical values help to explain the scarcity of peroxisomal profiles on the electron microscopic images in comparison to the enumerated abundance in the paraffin-sectioned material.
In contrast to early spermatids, in later stages of spermiogenesis (step 15–16 spermatids), aggregation of peroxisomal profiles was noted. These clusters were positively labeled with gold particles in immunocytochemical preparations stained for detection of catalase, PEX13, or GFP (testis sections of transgenic animals; Fig. 5, E and F). Labeling was present on round or tubular profiles and also on double-membraned loop structures (Fig. 5G). Cytochemical detection of catalase activity on the ultrastructural level also revealed large clusters of catalase-positive profiles in step 16 spermatids, including catalase-positive, double-membraned loops (Fig. 5, C and D).
Heterogeneity of Peroxisomal Enzymes Is Preserved in Primary Cell Cultures and Cytospin Preparations of Isolated Leydig, Peritubular Myoid, and Sertoli Cells
After isolation of primary Leydig, peritubular myoid, and Sertoli cells from 14-day-old (P14) mice and Leydig cells from adult mice, the purities of the cultures were determined by immunofluorescence stainings using antibodies against cell type-specific markers (Fig. C in Supplemental Data). More than 95% of cells in Sertoli and peritubular myoid cell cultures and more than 85% of juvenile and adult Leydig cell cultures were positive for cell type-specific markers.
To confirm the morphological results obtained in situ and in isolated cell cultures by immunofluorescence, Western blot analysis was performed using distinct subcellular fractions obtained by differential centrifugation from homogenized cell preparations. The peroxins PEX13 and PEX14 were detected in adult Leydig cells and P14 Sertoli and peritubular myoid cells, whereas the protein levels of both peroxins were low in P14 Leydig cells (Fig. 6A). In accordance with the results obtained by immunofluorescence, ABCD1 was mainly present in Sertoli cells (Fig. 6B). In contrast, high levels of catalase were present in adult Leydig and peritubular myoid cells, whereas the abundance of this enzyme was low in Sertoli cells (Fig. 6A). Catalase was barely detectable in P14 Leydig cells, but a specific band of expected size could be observed after prolonged exposure times of films (Fig. 6B).
|
Using semiquantitative RT-PCR, the steady-state levels for the mRNAs encoding peroxisomal proteins were determined in isolated cell cultures. For calculations of differences in mRNA expression levels, the RT-PCR band intensities of peroxisome-related genes were normalized for the band intensity of 28rRNA of the same cDNA preparation (Fig. 7A). Messenger RNAs for Abcd1 and Abcd3 (Fig. 7B) were present in high amounts in Sertoli and peritubular myoid cells (1.4-fold compared with adult Leydig cells). In contrast, the expression of Abcd2 mRNA was strongest in Leydig cells, whereas expression levels of Abcd4 and the genes encoding the peroxins Pex13 and Pex14 were similar in all cell types (Fig. 7, B and C). Catalase (Cat) mRNA levels were comparable in adult Leydig, P14 Sertoli, and peritubular myoid cells. However, the expression level of catalase in P14 Leydig cells was only about 20% of that of adult Leydig cells (Fig. 7D). Most mRNAs for peroxisomal β-oxidation enzymes (Acox1, Ehhadh, and Thiolase A for the β-oxidation pathway I, and Hsd17b4 and ScpX for the β-oxidation pathway II) were expressed at comparable levels in distinct cell types (Fig. 7, D and E). However, the mRNA levels for acyl-CoA oxidase 2 (Acox2), the rate-limiting enzyme for cholesterol side chain cleavage, was elevated about 4- and 6-fold in P14 Leydig- and P14 Sertoli cells, respectively, compared with adult Leydig cells, whereas Acox2 expression in P14 peritubular myoid cells was decreased to 20% of that of adult Leydig cells (Fig. 7E). The mRNA for acyl-CoA oxidase 3 (Acox3), which is the rate-limiting enzyme for the β-oxidation of branched-chain fatty acids, was not altered.
|
The expression of mRNAs of two enzymes involved in the biosynthesis of ether lipids, dihydroxyacetonephosphate acyltransferase (Gnpat), and alkyl-dihydroxyacetonephosphate synthase (Agps) was about 1.3-fold higher in P14 compared with adult Leydig cells (Fig. 7F).
The significance of peroxisomal metabolism for male fertility is accentuated by the impairment of spermatogenesis and severe testicular pathologies present in patients with peroxisomal dysfunction [1]. However, only sparse information is available on this organelle in the testis, and its physiological function in the testis is not understood to date. Because catalase is the most abundant peroxisomal protein in many tissues, it is commonly used as a marker to identify peroxisomes. Using this marker, peroxisomes of the testis had been identified in Leydig cells [11–15]. Only recently, peroxisomes were discovered also in Sertoli cells and spermatogonia by use of an alternative marker protein (ABCD3) in combination with more sensitive methods [17–19]. Due to the fact that with this marker, peroxisomes could not be visualized in germ cells in later stages of spermatogenesis, the significance of these organelles for normal spermatogenesis has been questioned. The aim of this study was to verify the presence of peroxisomes in germ cells and to obtain information on the possible functions of peroxisomes in distinct cell types of the testis.
Peroxisomes Are Present in Germ Cells and Undergo Significant Alterations During Spermiogenesis
Using antibodies against PEX13 and PEX14 in mouse and human testis, we could show the presence of peroxisomes in germ cells up to late stages of spermiogenesis (step 16 spermatids). Colabeling of GFP fluorescence in peroxisomes of GFP-PTS1 transgenic mice with PEX14 in these organelles at the light microscopic level indicates that the labeled peroxisomal membrane structures are indeed small, but import competent peroxisomes. Clear evidence for the peroxisomal nature of these particles was obtained by postembedding protein A-gold immunocytochemistry with anti-GFP, anti-PEX13, or anti-catalase antibodies.
The presence of functional peroxisomes in maturing spermatids suggests that these organelles might be involved in the biosynthesis of plasmalogens (ether lipids) for protection of spermatids against reactive oxygen species. Azoospermia due to spermatogenic arrest at the level of spermatocytes in mice with a knockout of the Gnpat (DHAPAT) gene, encoding a peroxisomal enzyme of ether lipid synthesis, support this hypothesis [33]. In addition, peroxisomes in germ cells might cooperate with the endoplasmic reticulum in the synthesis of polyunsaturated fatty acids in larger quantities, which is necessary for the integration into sphingomyelin that is essential for normal sperm function [34, 35]. Close spatial association between ER segments and small, tubular peroxisomal profiles were frequently found in maturing spermatids at the ultrastructural level in our study, supporting this hypothesis.
Futhermore, peroxisomes in the germinal epithelium might also be involved in the control of tubular polyamine abundance or D-aspartate levels in elongated spermatids, since polyamine oxidase and D-aspartate oxidase are peroxisomal enzymes [5, 36]. In this respect, for the seminiferous epithelium it is of interest that distinct polyamine levels (e.g., spermidine) are suggested to play an important role in cell proliferation and apoptosis [37].
The immunoreactivity of catalase is clearly present in spermatogonia and Sertoli cells but could only be detected with high concentrations of antibodies in spermatocytes or early spermatids in light microscopical preparations. Catalase immunoreactivity is clearly present in late spermatids (steps 14–16), where it is localized in network-like clusters. These clusters are consistently labeled with antibodies against peroxisomal proteins (PEX13, PEX14, ABCD3, catalase, and thiolase) and contain the imported GFP in the peroxisomal matrix in GFP-PTS1 transgenic animals. Peroxisomal profiles in these clusters are similar to or often smaller in size than endoplasmic reticulum and frequently exhibit double-membraned loop structures. Clusters of small peroxisomes, peroxisomal tubular profiles with low catalase content and double-membraned loops, are not peculiarly new structures (for a review, see Gorgas [38]). In contrast, they are frequently found in tissues closely associated with lipid metabolism (such as lipid-synthesizing glands) [39, 40], in fetal tissues, or in conditions associated with peroxisome proliferation (for a literature survey, see discussion sections in Baumgart et al. [41, 42]). Serial section analysis revealed that double-membraned loops represent terminal cup-shaped segments of the peroxisomal compartment and are sometimes devoid of catalase but are labeled for peroxisomal membrane proteins [39, 42].
Our results show that only few, but aggregated, peroxisomal clusters are present in residual bodies (Fig. 4G), most probably being phagocytosed and degraded by Sertoli cells. Selective staining of Sertoli cells for the lysosomal protein cathepsin D [18] and the autophagosomal protein LAMP2 in addition to labeling of the acrosome (Fig. 3F) supports this hypothesis. The genes involved in the regulation of the aggregation and clustering process of peroxisomes in late spermatids are unknown. However, it is known from studies with knockout animals that PEX11 proteins exert strong effects on peroxisomal abundance, size, and structure. Indeed, absence of PEX11/β leads to a decrease in peroxisomal number and an increase in the size of the particles and, more interestingly, overexpression of PEX11
leads to strong reduction and clustering of almost all peroxisomes in an individual cell [43, 44]. In addition, Pex11 mRNA is strongly expressed in the testis [43]. Future studies on PEX11 proteins in germ cells will reveal whether PEX11 is involved in the clustering process.
Peroxisomal Enzyme Content Is Heterogeneous, Resulting in Different Metabolic Functions of this Organelle in Distinct Cell Types of the Testis
Peroxisomes are versatile organelles and change their enzyme compositions according to the needs of specific cell types and organs (for a review, see Baumgart [6]). At least in hepatocytes, the abundance and the enzyme composition of this intracellular compartment are differentiation dependent, a process that seems to be influenced by PPAR [45]. Similarly, in the present study we have also noted significant differences in the expression of mRNAs encoding for peroxisomal ABC transporters and enzymes or in protein composition of these organelles in distinct cell types of the testis. These differences were conserved between mouse and humans, suggestive of a complementary or alternative function of peroxisomal metabolism in different testicular cell types. Marked differences have been observed in the protein or mRNA distribution of the peroxisomal ABC transporters (ABCD1, Abcd2, and ABCD3), the acyl-CoA oxidase 2 (Acox2) of the β-oxidation pathway II, and catalase.
Peroxisomal Metabolism in Cells of the Seminiferous Epithelium: Sertoli Cell Peroxisomes as Protectors Against Lipid Toxicity
Acox2 mRNA and the ABC transporters ABCD1 and ABCD3 are highly expressed in Sertoli cells compared with other cell types. These results are in agreement with the observation that in patients with X-linked adrenoleukodystrophy, first pathologic alterations seem to occur in Sertoli cells as "vacuolation" before the spermatogenetic arrest develops and Leydig cells are altered [3]. In addition, in mice with a Sertoli cell-specific knockout of Pex5 encoding the cytoplasmic receptor for the import of peroxisomal matrix proteins containing a PTS1, an accumulation of lipids in Sertoli cells already develops at P10 and precedes the arrest of spermatogenesis [46]. The importance of peroxisomal lipid metabolism for normal function of the seminiferous epithelium is further substantiated by the fact that knockout mice with single-enzyme deficiencies in the β-oxidation enzymes ACOX1 and HSD17B4 (also named MFP2) are infertile (for a review, see Baes and Van Veldhoven [47]). Indeed, Hsd17b4 knockouts gradually developed a complete fatty degeneration of the seminiferous epithelium, whereas Leydig cell function was preserved [6].
Our results show that within the seminiferous epithelium, several enzymes of the β-oxidation pathways and of lipid transporters are most abundant in Sertoli cells. This is in accordance with the localization of PPAR, a nuclear receptor and the transcriptional regulator of genes for peroxisomal β-oxidation pathway I enzymes in the seminiferous tubules, which also shows strongest expression in Sertoli cells [48]. PPAR-mediated proliferation of peroxisomes could result in a rapid degradation of potential PPAR lipid ligands, leading to the maintenance of a signaling-lipid homeostasis in the seminiferous epithelium by a feedback mechanism. As mentioned above, a precise control of the homeostasis of lipid derivatives in the seminiferous tubules seems to be essential for the survival of the seminiferous epithelium. Prostaglandins especially are potential regulators of spermatogenesis [49]. This is of relevance also to our results, since the β-oxidation enzymes of pathway I are able to degrade these eicosanoids (for a current review, see Wanders and Waterham [5]). Indeed, we found in our study that thiolase A is also present in germ cells in the seminiferous epithelium until late stages of spermiogenesis. With the presence of these enzymes, germ cells also might be able to regulate their intracellular levels of the lipid signaling molecules.
Calatase in Leydig Cells as an Antioxidative Enzyme for the Protection of Steroid Synthesis?
In Leydig cells, peroxisomes are often elongated, tubular, and are sometimes hardly larger in diameter than dilated segments of the smooth endoplasmic reticulum [11, 12]. As indicated also by our results, peroxisomes in Leydig cells contain the highest amount of catalase protein and are often interconnected with each other (Figs. 1–3). A network-like distribution of these organelles was already proposed by Mendis-Handagama et al. [12]. Due to significant proliferation or numerical reduction of this cell organelle by LH treatment or LH deprivation and the presence of SCP2 in peroxisomes, these authors also have speculated that testicular steroid synthesis could occur at least in part in peroxisomes [16].
In contrast to the enrichment of the ABC transporters ABCD1 and ABCD3 in Sertoli cells, peroxisomes in Leydig cells (already at P14) show the highest expression levels of Abcd2. Only sparse information is available on the substrates transported by this ABC transporter family member. However, it is tempting to speculate that these might be intermediates of sterol synthesis on their way from the endoplasmic reticulum or mitochondria to peroxisomes. Indeed, narrow spatial contacts are observed between the endoplasmic reticulum, mitochondria, and peroxisomes on the ultrastructural level (data not shown) and also in immunofluorescence preparations of isolated Leydig cells (Fig. C in Supplemental Data), a prerequisite for the shuttle of intermediates between these cell compartments.
For steroid synthesis, the high abundance of catalase also may be necessary and beneficial for Leydig cells, since catalase is an antioxidant enzyme with a very high capacity to metabolize H2O2. Increased levels of intracellular H2O2 in Leydig cells have been shown to inhibit steroidogenesis via blockage of the mitochondrial cytochrome P450 side chain cleavage activity and StAR protein expression [50]. Since peroxisomes house a variety of other antioxidative enzyme systems [51], further studies are necessary to clarify their importance in the regulation of Leydig cell functions.
The Heterogeneity in Peroxisomal Enzyme Content Is Conserved in Mouse and Humans
Our study demonstrates that peroxisomes can be visualized by immunofluorescence in Bouin-fixed, paraffin-embedded human tissue. The enzymatic heterogeneity of peroxisomes is conserved in mouse and humans. The distribution pattern of peroxisomes in the human seminiferous epithelium is best recognized with PEX13 or PEX14 as marker enzymes. Comparable to the mouse, the peroxisomal staining pattern is dependent on stages of the seminiferous epithelium, with the peroxisomal compartment undergoing similar significant alterations during maturation of spermatids. Clusters of peroxisomal profiles also appear in late spermatids prior to segregation into residual bodies in human preparations. The cell type-specific conservation of peroxisomal metabolic pathways suggests that mouse models can indeed be used to investigate the molecular pathogenesis of peroxisome-related infertility. All described mouse models with knockout of genes involved in peroxisomal biogenesis exhibit almost identical phenotypes to those observed in corresponding patients (for a recent review, see Baes and Van Veldhoven [47]. Similar to human peroxisomal disorders, male knockout mice with single-enzyme defects in peroxisomal lipid metabolism are infertile [33, 46, 52].
Future studies are needed to further clarify the physiological role of peroxisomal metabolism in the testis, the involvement of peroxisomes in the regulation of spermatogenesis, and the molecular pathogenesis of peroxisome-associated spermatogenesis defects. Based on our observations, the use of mouse models seems to be appropriate for comparative investigations of peroxisomal functions in the testis with relevance to humans.
ACKNOWLEDGMENTS
We thank Dr. Robert L. Snipes for careful reading of the manuscript and suggestions in English. The help of Dr. Klaus-Peter Valerius with animal perfusions and the excellent technical assistance of Gabriele Thiele, Magdalena Gottwald, and Elke Richter are gratefully acknowledged. Furthermore, we are indebted to Profs. Denis I. Crane, Alfred Völkl, and Richard A. Rachubinski for providing some antibodies (Table A in Supplemental Data available at www.biolreprod.org ).
FOOTNOTES
1Supported by LOM (Leistungs-orientierte Mittel) funds and the PhD program of the Faculty of Medicine of the Justus Liebig University Giessen. 
Correspondence: 2Eveline Baumgart-Vogt, Institut für Anatomie und Zellbiologie II, Fachbereich Medizin, Justus Liebig Universität, 35385 Giessen, Germany. FAX: 49 641 99 47109; e-mail: Eveline.Baumgart-Vogt{at}anatomie.med.uni-giessen.de
3Current address: Institute of Cell Biology, Swiss Federal Institute of Technology Zürich (ETHZ), 8093 Zurich, Switzerland.
Received: 8 March 2007.
First decision: 29 April 2007.
Accepted: 4 September 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
