|Home | About | Journals | Submit | Contact Us | Français|
Epididymosomes are small membranous vesicles secreted by epithelial cells within the luminal compartment of the epididymis. In bovine, many proteins are associated with epididymosomes, and some of them, such as the glycosylphosphatidylinositol (GPI)-anchored protein P25b, macrophage migration inhibitory factor (MIF), and aldose reductase (AKR1B1), are transferred to spermatozoa during the epididymal maturation process. P25b is associated with detergent-resistant membrane (DRM) domains of epididymal spermatozoa, whereas MIF and AKR1B1 are cytosolic proteins associated with detergent-soluble fractions. In this study, we tested the hypothesis that DRM domains are also present in the epididymosomes and that P25b DRM-associated proteins in these vesicles are transferred to the DRMs of spermatozoa. The presence of DRMs in epididymosomes was confirmed by their insolubility in cold Triton X-100 and their low buoyant density in sucrose gradient. Furthermore, DRMs isolated from epididymosomes are characterized by the exclusive presence of ganglioside GM1 and by high levels of cholesterol and sphingomyelin. Biochemical analysis indicated that P25b is linked to DRM in epididymosomes, whereas MIF and AKR1B1 are completely excluded from these membrane domains. Proteolytic treatment of epididymosomes and immunoblotting studies showed that P25b is affected by trypsin or pronase proteolysis. In contrast, MIF and AKR1B1 are not degraded by proteases, suggesting that they are localized within epididymosomes. Interaction studies between epididymosomes and epididymal spermatozoa demonstrated that P25b is transferred from the DRM of epididymosomes to the DRM of the caput epididymal spermatozoa as a GPI-anchored protein. Together, these data suggest that specific localization and compartmentalization of proteins in the epididymosomes coordinate the association of epididymal proteins with the different functional structures of spermatozoa.
In order to become fully competent to fertilize, mammalian spermatozoa leaving the testis need to transit along the epididymis . The epididymal sperm maturation process is mainly dependent on the apical secretory activity of principal epithelial cells within the intraluminal compartment of the tubule [2, 3]. Epididymal epithelial cells exhibit region-specific differences in their structure, their function, and their patterns of protein secretion [4–6]. The proteins secreted by caput and corpus epididymides are devoted to the acquisition of self-assured motility and the ability of sperm cells to bind the zona pellucida and to fuse with the egg plasma membrane . Proteins secreted by cauda epididymidis are required to maintain viable and potentially fertile stored sperm cells . During their transit, spermatozoa acquire some of these proteins, a process essential to generate functional male gametes .
Mature spermatozoa possess distinct functional structures, in which the head is involved in zona pellucida recognition and penetration, the midpiece in energy production for motion, and the tail in flagellar motility . It is thus expected that the proteins acquired by spermatozoa during epididymal maturation take a strategic, functional localization in the distinct sperm structures. Subsequently, upon transfer to spermatozoa, most of these epididymal proteins behave as membrane proteins [8, 10] or are incorporated into the intracellular structures of spermatozoa [11–13]. Many of these proteins are anchored to the sperm plasma membrane via a glycosylphosphatidylinositol (GPI) moiety, such as P25b in bull , P26h in hamster , HE5 (CD52) in human , and hyaluronidase and SPAM1 in mouse [17–19]. The exact mechanisms by which secreted epididymal proteins are added to spermatozoa remain elusive.
Epididymosomes are small membranous vesicles released via apocrine secretion by principal epithelial cells within the intraluminal compartment of the epididymis [20–22]. These vesicles show structural similarity with exosomes and have a diameter of 50 to 500 nm. Epididymosomes have been described in an increasing number of mammalian species, including human , mouse , sheep [24, 25], rat [12, 26], hamster [15, 27], and bull [14, 28, 29]. The content of proteins in epididymosomes is complex in number and electrophoretic behavior [28–30]. The transfer of a specific set of functional proteins to spermatozoa from epididymosomes suggests that these vesicles are involved in the epididymal sperm maturation process. Among the epididymosome-associated proteins we have identified in bovine, P25b, macrophage migration inhibitory factor (MIF), and aldose reductase (AKR1B1) are transferred to specific structures of the spermatozoa [14, 29, 30]. In this context, we have demonstrated that P25b is associated with detergent-resistant membrane (DRM) domains, also known as lipid rafts, in bovine epididymal spermatozoa. MIF and AKR1B1 are excluded from these membrane domains and are found exclusively in the detergent-soluble fractions . The present study confirms that DRMs also are present in bovine epididymosomes and that epididymosomal-associated P25b protein is predominantly detected in DRM fractions and transferred to the DRM of spermatozoa. Specific localization and compartmentalization of proteins in the epididymosomes, particularly in DRM domains, may thus coordinate the association of epididymal proteins with the different functional structures of the spermatozoa.
Epididymides from sexually mature bulls were obtained from a commercial slaughterhouse. The epididymides were dissected and kept on ice for the duration of sperm and epididymosome preparation as described previously . Briefly, fluid from the caput epididymidis was recovered from neatly cut tubules and by applying a gentle pressure to the proximal portion of the caput epididymidis. Spermatozoa were isolated from the caput epididymal fluid by centrifugation at 500 × g for 5 min and were washed extensively in saline solution. Fluid from the cauda epididymidis was obtained by retrograde flushing by applying air pressure with a syringe in the proximal scrotal segment of the vas deferens. Cauda epididymal fluids were centrifuged twice at 3000 × g to remove spermatozoa and remaining debris and then centrifuged at 120000 × g for 2 h at 4°C. The pellets were resuspended in 0.15 M NaCl and centrifuged again at 120000 × g. The pellets containing isolated cauda epididymosomes were resuspended in 0.15 M NaCl. The Coomassie brilliant blue method  was used to determine the protein concentration of the epididymosomes using bovine serum albumin as the standard.
Subcellular fractions of bovine epididymal spermatozoa were prepared according to Noland et al. . The complete procedure was performed at 4°C. Caput epididymal spermatozoa were washed in HBS (150 mM NaCl, 10 mM Hepes, pH 7.2) supplemented with protease inhibitors (500 mM PMSF and 10 mg/ml each aprotinin, leupeptin, and pepstatin A) and subjected to nitrogen cavitation at 750 psi for 10 min at 4°C. After cavitation, the sperm suspensions were centrifuged at 10000 × g for 25 min to pellet the demembranated spermatozoa. The supernatant obtained from the 10000 × g centrifugation was further centrifuged at 100000 × g for 60 min to recover the plasma membrane fraction (pellet) and the cytosolic fraction (supernatant). Demembranated spermatozoa in the 10000 × g pellet were suspended in HBS, kept on ice for ultrasonic treatment, and then placed on the top of a 75% Percoll solution and centrifuged at 700 × g for 15 min. Along the density gradient, sperm components were distributed as three layers: the upper layer was mainly enriched in membranes (acrosomal, nuclear, and remnants of plasma membranes); the intermediate layer contained flagellae and the pellet enriched in sperm heads. The different samples containing the flagellae and the head separately were washed and recovered in HBS by centrifugation at 10000 × g for 30 min. Sample containing membranes was centrifuged at 100000 × g and pooled with the first plasma membrane fraction.
Washed spermatozoa (500 × 106) or cauda epididymosomes (the equivalent of 1 mg of proteins) were treated with ice-cold TNE (10 mM Tris, pH 7.5; 150 mM NaCl; and 5 mM ethylenediaminetetraacetic acid) containing 1% (v/v) Triton X-100 for 30 min on ice. In order to remove whole sperm and sperm particulates, the Triton X-100-treated sperm suspension was centrifuged at 2000 × g for 5 min. Supernatants from sperm or Triton X-100-treated epididymosomes were homogenized by ultrasonic treatment prior to centrifugation at 200000 × g in a three-step discontinuous sucrose gradient (42.5%, 30%, and 5%) as described previously [31, 34, 35]. Briefly, an aliquot of 0.4 ml of Triton X-100-treated sperm or epididymosome extracts were mixed with an equal volume of 85% (w/v) sucrose and overlaid on 2.4 ml of 35% (w/v) sucrose and 1 ml of 5% (w/v) sucrose. Centrifugation was performed at 200000 × g in a SW60Ti rotor at 4°C for 16 h. Eleven fractions (numbered 1–11) were collected starting from the top of the tube: the low-density fractions 3–5 contained the DRMs, the intermediate-density fractions were 6–8, and the high-density fractions 9–11 contained the Triton-soluble (T-S) proteins. Phospholipid composition, cholesterol concentration, and light scattering (wavelength of 400 nm) of DRMs and T-S proteins were determined. The Coomassie brilliant blue method was used to determine the protein concentration of each fraction.
Lipids were extracted from epididymosomes by the Folch method . Briefly, epididymosomes (equivalent of 1 mg of proteins) were mixed with 8 ml chloroform:methanol (2:1, v/v) and then diluted with 1.5 ml of distilled water. The organic and aqueous phases were separated by centrifugation at 500 × g for 10 min following 1 h of incubation at room temperature. The organic phase was collected and evaporated under nitrogen flow; the dried material was then dissolved in chloroform:methanol (2:1, v/v). The upper aqueous phase was mixed with 8 ml of chloroform:methanol (2:1, v/v) and centrifuged at 500 × g for 10 min at 4°C, after which the organic phases were collected and pooled. The levels of cholesterol were quantified in DRMs and T-S fractions isolated from epididymosomes .
The different classes of membrane phospholipids were separated as described previously [23, 38] by high-performance thin-layer chromatography (HPTLC; Merck). The HPTLC plates were incubated with a mix of methyl acetate, propan-2-ol, chloroform, methanol, and 0.25% (w/v) aqueous KCl (25:25:25:10:9, v/v). The spots of phospholipids were visualized by treatment of HPTLC plates with a solution containing 10% (w/v) CuSO4 and 8% (v/v) H3PO4 and by heat treatment at 180°C. Densitometric analysis of the phospholipid classes was carried out using an image analyzer system (Alpha Innotech Corp., San Leandro, CA). The relative intensities of different phospholipids were quantified by reference to different concentrations of phospholipid standards (phosphatidyl serine, phosphatidyl inositol, phosphatidyl choline, Lα-phosphatidyl ethanolamine, and sphingomyelin; Sigma-Aldrich, Oakville, ON, Canada). Three samples of epididymosomes were independently analyzed, and each analysis was performed in duplicate. Standard dosage values gave curves with linear-regression coefficients (R2) ≥0.90. The concentrations of phospholipids and cholesterol (μg/mg protein) were expressed as mean of triplicates ± SEM.
The ganglioside GM1 is known to be specifically targeted by the cholera toxin B subunit (CTB) and exclusively associated with lipid rafts, or DRMs, of somatic and germinal cells [31, 39, 40]. Fractions collected after ultracentrifugation of epididymosomes on a discontinuous sucrose gradient were dotted onto nitrocellulose membranes as described previously . Membranes were blocked in 5% (v/v) skim milk in PBS containing 0.05% (v/v) Tween-20 (Sigma-Aldrich) and then incubated with horseradish peroxidase (HRP)-conjugated CTB (Sigma-Aldrich) in PBS containing 3% (w/v) bovine serum albumin. Cholera toxin B subunit-probed GM1 was visualized using the ECL detection kit (Roche Diagnostics, Québec City, QC, Canada).
To biotinylate surface proteins, cauda epididymosomes were washed with PBS, pH 7.4, and incubated at room temperature with 1 mg/ml Sulfo-NHS-LC-biotin (Pierce, Brockville, ON, Canada) for 30 min and then at 4°C for 8–10 h. The biotinylated epididymosomes were diluted with 15 volumes of 0.15 M NaCl and centrifuged at 120000 × g for 2 h. Pellets were resuspended in 0.15 M NaCl, separated in aliquots, and kept at −80°C until use. Biotinylated epididymosomes were coincubated with epididymal spermatozoa to determine the electrophoretic pattern of transferred proteins. Caput epididymal spermatozoa were resuspended at a concentration of 250 × 106 spermatozoa/ml in a buffered solution containing 0.15 M NaCl and 10 mM MES-PIPES, pH 6.5, and were incubated with biotinylated cauda epididymosomes for 3 h. The protein transfer from epididymosomes to spermatozoa was shown previously to be optimum at pH 6.5 in vitro . After coincubation, spermatozoa were extensively washed in 0.15 M NaCl to eliminate unbound biotinylated epididymosomes.
Proteins from DRM fractions, which were isolated from biotinylated epididymosomes or caput epididymal spermatozoa previously coincubated with biotinylated epididymosomes, were first solubilized in 1% Triton X-100 for 60 min at 37°C. Following Triton X-100 treatment, DRM fractions were centrifuged at 200000 × g for 60 min in order to eliminate insoluble material. Biotinylated proteins from solubilized DRMs as well as T-S fractions were precipitated with streptavidin-agarose (Pierce). The affinity precipitates were recovered by centrifugation, washed, boiled in SDS sample buffer, and subjected to SDS-PAGE and Western blot analysis to detect P25b proteins.
Epididymosomes, purified MIF proteins, or recombinant AKR1B1 proteins (a gift from Dr. M.A. Fortier, Université Laval, Laval, QC, Canada) were left untreated or were treated with 25 μg/ml trypsin (Sigma) or 15 μg/ml Pronase E (Streptomyces griseus protease XIII; Sigma) for 30 min at room temperature in 0.15 M NaCl.
Caput spermatozoa coincubated with biotinylated epididymosomes were subjected to nitrogen cavitation and ultrasound treatment to isolate membrane fractions as described above. Spermatozoa and membrane fractions were washed once in 150 mM NaCl then incubated for 15 min at 30°C in either 150 mM NaCl containing 5 U/ml phospholipase C or in 150 mM NaCl alone. The phospholipase C used was from Bacillus cereus and is specific for phosphatidyl-inositol (Sigma). After enzymatic treatment, membrane fractions were pelleted at 120000 × g for 60 min, and spermatozoa were pelleted at 4000 × g for 5 min. Both pellets and supernatants were analysed by SDS-PAGE and immunoblotted in order to detect biotinylated proteins.
Sperm proteins collected after sucrose gradient centrifugation or subcellular fractionation were precipitated by the MeOH/CHCl3 method as described previously  and solubilized in SDS-PAGE sample buffer (2% [v/v] SDS, 2.5% [v/v] β-mercaptoethanol, and 50 mM Tris, pH 6.8). The sperm protein extracts were resolved by SDS-PAGE  and transferred onto nitrocellulose membrane. To detect P25b protein, macrophage MIF and AKR1B1 membranes were incubated with a rabbit anti-P26h/P25b antiserum from our laboratory , and specific antibodies directed against MIF (generously provided by Dr. M. Nishibori from the Department of Pharmacology, Okayama University of Japan, Okayama City, Japan) or against AKR1B1 (a gift from Dr. M.A. Fortier ). Membranes then were incubated with HRP-conjugated anti-rabbit IgG antisera (1:10000) for 1 h at room temperature. To detect biotinylated proteins, membranes were incubated with neutravidin-conjugated HRP (Pierce). Targeted proteins were visualized using the ECL detection kit.
In order to identify the structures of the spermatozoa in which epididymosome-associated proteins P25b, MIF, and AKR1B1 are preferentially transferred, caput epididymal spermatozoa were subjected to nitrogen cavitation and sonication procedures allowing enrichment of membrane, cytosolic, head, and flagellar fractions. As expected, P25b was detected exclusively in the membrane fractions. In contrast, MIF and AKR1B1 were both strictly associated with the cytosolic/soluble fractions (Fig. 1). Together, these data established that epididymosome-associated proteins transferred to spermatozoa are localized in the different functional structures of the spermatozoa.
Based on the above observations, we assessed the hypothesis that DRMs are also present in epididymosomes and are involved in particular compartmentalization of P25b. In order to verify the presence of DRM domains in the epididymosomes, these vesicles were isolated from cauda epididymal fluid, treated with cold Triton X-100, and then centrifuged on a discontinuous sucrose gradient. Eleven fractions were collected from the sucrose gradient, starting at the top. As shown in Figure 2A, the ganglioside GM1 was mainly detected in fractions 3–5, whereas it was completely excluded from fractions 9 to 11. Given that DRM domains are characterized by an increased ability to disperse the light beam compared with detergent-soluble material, light scattering of each fraction was determined by measuring the absorbance at a wavelength of 400 nm [45, 46]. As expected, the optical density of GM1-containing fractions was significantly higher than the other fractions (Fig. 2). This result reflects a higher content of insoluble particulates present in these fractions.
To further characterize the GM1-containing DRM domains isolated from epididymosomes, lipid and protein compositions were determined in pools of fractions 3–5 (DRM domains) and fractions 9–11 (T-S domains). As shown in Table 1, the cholesterol levels in a constant amount of epididymosomes (equivalent to 1 mg of protein) were evaluated at 102.4 ± 5.7 μg in DRM domains and 55.7 ± 1.0 μg in T-S domains. In the same preparations, the content of phospholipids was evaluated at 51.2 ± 4.4 μg in DRM and 69.5 ± 8.0 μg in T-S. Analysis from these data indicates that cholesterol represented 67% and 45% of total lipids in DRM and T-S, respectively. The cholesterol:phospholipid ratio in DRM was thus 2.5-fold greater (2.0 ± 0.4) than in T-S (0.8 ± 0.1). Among the phospholipid classes present in DRM, sphingomyelin (SM; 25.5 ± 0.7 μg) was the most abundant, whereas phosphatidylcholine (PC; 14.1 ± 2.5 μg), phosphatidylethanolamine (PE; 7.1 ± 0.7 μg), phosphatidylserine (PS; 2.3 ± 0.5 μg), and phosphatidylinositol (PI; <1 μg) were present at relatively low levels. In contrast, in T-S fractions, SM content was relatively low (5.0 ± 0.1 μg) compared with PC, which was the most abundant (32.4 ± 6.7 μg), whereas PE (15.7 ± 1.5 μg), PS (12.8 ± 1.0 μg), and PI (3.6 ± 0.3 μg) were present at relatively low levels. In addition, the protein contents in DRM and T-S fractions were evaluated at 77.9 ±8.0 μg and 861.5 ± 72.5 μg, respectively, both in an equivalent of 1 mg of epididymosomal proteins. The protein:lipid ratio in DRM was 14-fold lower (0.5 ± 0.03) than in T-S (6.9 ± 0.6).
Given that P25b, MIF, and AKR1B1 are associated with epididymosomes and transferred to different compartment of spermatozoa, we examined whether a specific compartmentalization of these proteins is also established in these vesicles. As shown in Figure 3, P25b was predominantly detected in DRM sucrose fractions (fractions 3–5) but was barely or not detected in the T-S fractions (fractions 9–11) isolated from epididymosomes. In contrast to P25b, MIF and AKR1B1 were totally excluded from DRM and were associated with the T-S fractions. In order to determine whether P25b, MIF, and AKR1B1 behave as surface-exposed proteins in the epididymosomes, we tested the possibility that these epididymosome-associated proteins can be digested by exposing intact epididymosomes to different proteolytic treatments. To confirm this premise, caudal epididymosomes were incubated with trypsin (25 μg/ml) or pronase E (15 μg/ml). Immunoblotting analysis of protease-exposed epididymosomes indicated that P25b proteins are cleaved by both proteases (Fig. 4). In contrast to P25b proteins, treatments with trypsin or pronase E were unable to cleave epididymosome-associated MIF and AKR1B1 proteins. This suggests that MIF and AKR1B1 behave as internal proteins. To further confirm that these proteins are insensitive to direct protease hydrolysis due to their internal localization, soluble forms of MIF and AKR1B1 were treated with trypsin or pronase E. Both proteins were affected by protease treatments, confirming that these proteins are sensitive to protease cleavage in a soluble form but not when they are associated to epididymosomes.
These results suggest that specific localization and compartmentalization of proteins in the epididymosomes could be essential to direct the transfer of epididymal proteins in the different functional structures of maturing spermatozoa.
In order to confirm that compartmentalization of proteins in epididymosomes is involved in the association of epididymal proteins with different subcellular compartments of bovine spermatozoa, the transfer of P25b proteins from the DRM of epididymosomes to the DRM of spermatozoa was examined. Surface-exposed proteins of epidymosomes were biotinylated and coincubated with caput epididymal spermatozoa. The content of biotinylated proteins in the DRM and T-S fractions of epididymosomes were analyzed after ice-cold Triton X-100 treatment and subsequent fractionation on sucrose gradient. The detection of biotinylated proteins by immunoblotting revealed a single 30-kDa band localized in DRM fractions (Fig. 5A). Most of the biotinylated proteins in epididymosomes were detected in the T-S fractions. In order to determine whether the 30-kDa biotinylated protein in DRM fractions corresponded to P25b, biotinylated proteins of DRM or T-S fractions were affinity precipitated using streptavidin coupled to agarose beads. Immunoblotting analysis was performed to detect biotinylated proteins using HRP-conjugated neutravidin, or to detect P25b proteins using a rabbit anti-P26h/P25b antiserum. As shown in Figure 5B, biotinylated proteins of both DRM and T-S fractions were affinity precipitated. Western blot analysis of P25b in the precipitated fractions established that biotinylated P25b corresponds to the 30-kDa biotinylated protein band associated with DRM in epididymosomes. In parallel, biotinylated epididymosomes were coincubated with caput epididymal spermatozoa to study functional interactions. Detergent-resistant membrane and T-S material were isolated from coincubated sperm cells after treatment with Triton X-100 and fractionation on sucrose gradient. As shown in Figure 5C, detection of biotin molecules established that only a few selected proteins were transferred from epididymosomes to spermatozoa. Moreover, the biotinylated P25b protein is associated with the DRM fractions of spermatozoa. Three major proteins with molecular weights between 30 and 45 kDa were transferred from epididymosomes and detected in the T-S fractions (Fig. 5C). To further determine that the compartmentalization of P25b proteins in epididymosomes coordinates the association of P25b in the DRM of spermatozoa, biotinylated proteins transferred to DRM or T-S fractions were affinity precipitated with streptavidin beads in sperm. Western blots probed with the anti-P25b antiserum showed that P25b was biotinylated and exclusively detected in sperm DRM fractions (Fig. 5D).
To further establish that P25b is specifically transferred as a GPI-anchored protein to the sperm membrane, biotinylated epididymosomes were coincubated with caput epididymal spermatozoa. Membrane-enriched fraction was then isolated from these spermatozoa by nitrogen cavitation and ultrasonic treatment. To verify whether biotinylated P25b protein is GPI anchored, total spermatozoa as well as membrane fraction were treated in the absence or presence of phospholipase C and centrifuged to separate pellet from supernatant. As shown in Figure 6, the 30-kDa proteins corresponding to the biotinylated P25b were associated with the pellets of untreated spermatozoa or membrane-enriched fraction. Western blot analysis indicates that phospholipase C treatment induced a significant loss of the biotinylated P25b in the sperm and membrane pellets. Following phospholipase C treatment, biotinylated P25b was found in the supernatant fraction as a soluble protein. This result corroborates the fact that epididymosome-associated P25b is specifically transferred and anchored to the sperm membrane.
During the transit through the epididymis, testicular spermatozoa undergo a series of biochemical modifications mediated by the interactions of spermatozoa with various macromolecules that compose the epididymal luminal fluid [8, 47]. Acquisition of new proteins is one of the essential modifications occurring during this maturation process, thereby generating functional male gametes . Previous studies have demonstrated that some of these proteins, such as P25b, HE5 (CD52), SPAM1, and MIF, are associated with small membranous vesicles, named epididymosomes, secreted by epididymal epithelial cells in the epididymal lumen [12, 14, 19, 23]. Our previous data have revealed that epididymosomes interact with spermatozoa and transfer some of the epidymosome-associated proteins to maturing spermatozoa [14, 29]. These proteins are crucial to various sperm functions, as exemplified by P26h/P25b proteins, which are a determinant in sperm-zona pellucida interactions [43, 49]; HE5 (CD52) proteins, implicated in protection of spermatozoa from being phagocytosed by leukocytes in the female genital tract [16, 50, 51]; SPAM1 proteins, involved in cumulus penetration via hyaluronidase activity, zona pellucida binding, and the signaling involved in acrosomal exocytosis ; or MIF proteins, required for sperm motility . Thus, understanding the transfer mechanisms of these proteins is of physiological relevance to better understanding the complex processes of gamete interactions.
Because spermatozoa are functionally regionalized cells , this means that the sperm proteins newly acquired during the epididymal transit must be localized in specific structures of the male gamete according to their respective functions. In this context, we reported previously that P25b, MIF, and AKR1B1 are differently compartmentalized in spermatozoa . P25b is a GPI-anchored protein that was found to be associated with DRM domains in the plasma membrane of epididymal spermatozoa. MIF and AKR1B1, however, were excluded from these domains but associated with the membrane/cytosolic detergent-soluble material . The experimental data presented here support the hypothesis that DRM is involved in epididymal sperm maturation by allowing protein transfer from the epididymosomes to specific compartments of spermatozoa.
Our data demonstrated that DRM domains are present in the epididymosomes. Detergent-resistant membranes isolated from these vesicles share similar properties with DRMs isolated from somatic and germinal cell models. Structurally, DRM domains are enriched in cholesterol and sphingomyelin but are relatively poor in other structural lipids, such as phospholipids. They are also characterized by the segregation of some proteins, often presenting a GPI anchor . As a consequence, the lipid rafts or DRMs are involved in various processes important in cell physiology, such as signal transduction, protein and lipid trafficking, endocytosis, and even regulation of cellular entry of pathogens . Although rafts are mainly characterized in cell membranes, increasing evidence shows that these domains are also present in membranous vesicles. Notably, it was recently demonstrated that rafts are present in exosomal membranes secreted by reticulocytes, lymphoid B-cell line Daudi, and human erythroleukemia cell line . As reported in the present study, ganglioside GM1, a well-characterized raft marker, is exclusively associated with the low-density sucrose fraction collected after centrifugation of vesicular extracts from cold Triton X-100-treated epididymosomes. The GM1-containing fractions isolated from these vesicular extracts show a higher absorbance at 400 nm, which is in agreement with high lipid content. These DRM fractions are also particularly enriched in cholesterol and sphingomyelin but contain fewer phospholipids than the detergent-soluble fraction. Moreover, P25b, a GPI-anchored protein associated with the sperm membrane, is mostly associated with DRM domains isolated from epididymosomes. It is well known that GPI-anchored proteins are preferentially associated with the DRM and may be used as DRM markers . Together, these data demonstrate the presence of DRM domains in the membrane of bovine epididymosomes.
In contrast to P25b, DRM domains isolated from epididymosomes completely exclude MIF and AKR1B1. Moreover, P25b, but not MIF and AKR1B1, is cleaved after proteolytic treatment of intact epididymosomes. These results confirm that P25b is a surface-exposed protein and strongly suggest that MIF and AKR1B1 are both compartmentalized in the lumen of the epididymosomes but not at the membrane surface. The localization of MIF in the epididymosomal lumen is in accordance with previous data showing an internal localization of MIF in epididymosomes using electronic microscopy .
One of the most relevant conclusions we reported here is that compartmentalization of proteins in epididymosomes is critical to coordinating the transfer of proteins in specific sperm compartments or structures. This is supported by the fact that P25b, MIF, and AKR1B1 have a similar compartmentalization in epididymosomes and epididymal spermatozoa. In this context, P25b protein is the major biotinylated protein primarily associated with DRM in epididymosomes and with an SDS-PAGE migration corresponding to a molecular weight of 30 kDa. Most of the biotinylated proteins, however, are associated with the detergent-soluble fractions of epididymosomes. As shown previously , only a specific subset of the biotinylated proteins were transferred to caput epididymal spermatozoa following coincubation with cauda epididymosomes. Among these proteins, the biotinylated P25b, which is associated with the DRM of epididymosomes, is exclusively and directly transferred to the DRM domains of caput epididymal spermatozoa. Our data also confirm that P25b is exclusively associated with and transferred to the membrane structure as a GPI-anchored protein, as detected by Western blot analyses performed on different subcellular fractions of spermatozoa prepared by nitrogen cavitation and sonication. These data are in accordance with previous reports showing that DRM domains are mainly restricted to the plasma membrane overlying the acrosome [56, 57] and with our previous findings that P25b is principally associated with the acrosomal cap . The membrane localization of P25b in the head of the spermatozoa also corroborates the suggested involvement of this protein in zona pellucida recognition and binding [58, 59]. Furthermore, large amounts of biotinylated proteins are associated with the Triton-soluble material; however, only few selected proteins (detected as three bands in SDS-PAGE) were transferred to the spermatozoa. The exact identification of these transferred proteins is actually in progress.
The concept that protein compartmentalization in epididymosomes is essential to properly transfer proteins to specific sperm compartments or structures is further supported by the fact that MIF and AKR1B1 are found as cytosoluble proteins and are completely excluded from the plasma membrane, as determined by sperm fractionation. These data are in accordance with the potential functions of these proteins. It is reported that MIF is associated with the outer dense fibers of midpiece and principal-piece rat sperm flagellum, suggesting that MIF had a potential role in modulating sperm motility of the maturing spermatozoa [11, 12]. Unexpectedly, MIF was not detected in the flagellar fractions. This suggests that MIF is not covalently associated with the outer dense fibers of the sperm flagellum. AKR1B1 is an enzyme of the polyol pathway that reduces glucose in sorbitol, which is used as an energy source by cells [11, 12]. Although cellular localization of AKR1B1 in spermatozoa remains to be determined, it is hypothesized that intracellular localization of AKR1B1 would be involved in cytoplasmic accumulation of sorbitol within the cytoplasm, conferring a protection of epididymal spermatozoa against hypertonic conditions and enhancing sperm survival during epididymal transit and storage [11, 12].
Together, these data strongly suggest that specific localization and compartmentalization of proteins in the epididymosomes coordinate the association of epididymal proteins with the different functional structures of the spermatozoa. These findings provide new insight into the molecular mechanisms that confer sperm fertilization competence during the epididymal transit of testicular sperm cells.
Drs. M. Nishibori (Okayama University of Japan) and M.A. Fortier (Université Laval) are both acknowledged for generous gifts of anti-MIF and anti-AKR1B1 antibodies, respectively.
1Supported by the Canadian Institutes of Health Research (CIHR) and Natural Sciences Engineering Research Council of Canada grants (to R.S.). J.G. is supported by a PhD scholarship from the CIHR.