Abstract
Lipid architecture of the plasma membrane plays an important role in the capacitation process of the sperm cell. During this process, an increase in membrane fluidity takes place, which coincides with a redistribution of cholesterol to the apical region of the head plasma membrane and subsequently an efflux of cholesterol. Cholesterol is also a major player in the formation of lipid rafts or microdomains in the membrane. Lipid rafts favour specific protein–protein interactions by concentrating certain proteins in these microdomains while excluding others. In this study, we investigated the organization of lipid rafts during in vitro capacitation of boar sperm cells. We report on the presence of the lipid raft-specific proteins caveolin-1 and flotillin-1 in sperm cells. Capacitation induced a change in membrane distribution of these proteins. Lipid analysis on detergent-resistant membranes (DRMs) of sperm cells indicated that capacitation induces a lipid raft concentration rather than a disintegration of lipid rafts, because the total amount of lipid in the DRM fraction remained unaltered. Using a proteomic approach, we identified several major DRM proteins, including proteins involved in capacitation-dependent processes and zona pellucida binding. Our data indicate that sperm raft reorganization may facilitate capacitation-specific signalling events and binding to the zona pellucida.
Introduction
Directly after ejaculation, the sperm cell is not able to fertilize the oocyte. Only after activation in the female genital tract is the sperm cell able to reach, bind to and fertilize the oocyte. The sperm cell becomes hypermotile and obtains the ability to penetrate the cumulus oophorus and to bind to the zona pellucida of the oocyte which in turn triggers the acrosome reaction (Yanagimachi, 1994). Although the end results of this activation process (called capacitation) are known, the process is still not well understood at the molecular level. Capacitation in vivo takes place in the female genital tract, but capacitation can also be induced in vitro by incubation in media containing high levels of bicarbonate. Bicarbonate is also present in high concentrations in the female genital tract, although only low levels of bicarbonate are present in epididymal and seminal fluids, suggesting that exposure to bicarbonate is also key to capacitation in vivo (Harrison, 1996). Among the molecular processes that take place during capacitation are the activation of signalling pathways, such as cAMP-dependent protein kinase A (PKA) and protein tyrosine phosphorylation (Visconti et al., 1995a,b).
In addition to effects on proteins, the lipid architecture of the sperm plasma membrane plays an important role in the capacitation process. During capacitation, phospholipid scrambling takes place that induces surface exposure of endogenous aminophospholipids at the surface of sperm cells. This exposure of aminophospholipids in intact cells is restricted to the anterior acrosomal region of the head plasma membrane (Gadella and Harrison, 2002). A concomitant increase in membrane fluidity results in a redistribution of cholesterol to the apical region of the head plasma membrane which in turn allows an efflux of cholesterol (Flesch et al., 2001). In line with this, it is possible that these lipid rearrangements also enable a reorganization of membrane proteins, for instance, into a complex that enables binding to the zona pellucida. Indeed, such lipid rearrangements have been shown to coincide with enhanced sperm–zona-binding capacity (Harkema et al., 1998; Visconti et al., 1999a,b). Furthermore, the increased membrane fluidity may play a role in the induction of the acrosome reaction (Zarintash and Cross, 1996).
Because cholesterol plays such an important role in capacitation, we were interested in a possible role of lipid rafts in this process. Lipid rafts are commonly defined as sphingolipid and cholesterol-rich ordered domains in the membrane, and cholesterol is essential for the formation of these domains (Simons and Ikonen, 1997, 2000; Brown and London, 2000). Lipid rafts concentrate signalling molecules while excluding others and thereby favour specific protein–protein interactions, resulting in the activation of signalling cascades (for a thorough review, see Simons and Toomre, 2000 and references cited therein). Despite the isolation of sperm lipid rafts (Travis et al., 2001; Trevino et al., 2001) and the detected capacitation-induced changes in raft organization (Cross, 2004; Shadan et al., 2004), the possible role of lipid rafts in capacitation and the capacitation-induced signalling events and their importance of cholesterol redistribution and efflux during sperm capacitation remain unknown. For this reason, we investigated the presence and dynamics of cholesterol-rich ordered domains in boar sperm. We report on the presence of lipid raft-specific proteins in sperm cells and their membrane distribution before and after capacitation. Lipid analysis of detergent-resistant membranes (DRMs) from sperm cells was performed to determine whether capacitation induces dispersion (Cross, 2004) or clustering (Shadan et al., 2004) of lipid rafts. Furthermore, we analysed the protein composition of the DRM fraction of sperm cells by 2D electrophoresis combined with nano liquid chromatography-tandem mass spectrometry (LC-MS/MS) and discuss the possible role of these proteins in capacitation.
Materials and methods
Sperm preparation and plasma membrane isolation
Sperm-rich fractions were collected from ejaculates of healthy boars that are commonly used for artificial insemination in the Netherlands. All boars used showed excellent sperm output and were highly fertile. They were held at the Cooperative Centre for Artificial Insemination in Pigs ‘Utrecht en de Hollanden’ (Bunnik, The Netherlands). Semen was filtered through gauze to remove gelatinous material. Sperm cells were washed on a discontinuous Percoll (Pharmacia, Uppsala, Sweden) gradient, as described in Flesch et al. (1998). All solutions used were iso-osmotic (285–315 mOsm/kg) and at room temperature unless stated otherwise.
To induce capacitation, we incubated Percoll-washed sperm cells in HEPES-buffered Tyrode’s (HBT: 90 mM NaCl, 21.7 mM lactate, 20 mM HEPES, 5 mM glucose, 3.1 mM KCl, 2.0 mM CaCl2, 1.0 mM pyruvate, 0.4 mM MgSO4, 0.3 mM NaH2PO4 and 100 µg/ml of kanamycin; 300 mOsm/kg, pH 7.4) with 15 mM NaHCO3 in equilibrium with 5% CO2 in air in a humidified atmosphere. HBT was supplemented either with 0.1% (w/v) polyvinyl alcohol (PVA) and 0.1% (w/v) polyvinylpyrrolidone (PVP) or with 0.3% (w/v) bovine serum albumin (BSA, delipidated fraction V, Boehringer Mannheim, Almere, The Netherlands). For control conditions, the sperm cells were incubated in Tyrode’s medium without bicarbonate (again supplemented with either PVA/PVP or BSA). The sperm cell suspensions were incubated at 38.5°C in a cell incubator for 3–4 h. To extract cholesterol, we incubated sperm cells with 10 mM methyl-β-cyclodextrin (MBCD; Sigma Aldrich Fluka, Zwijndrecht, The Netherlands) for 30 min at 37°C.
Apical plasma membranes of sperm cells were isolated according to Flesch et al. (1998). Determination of the purity of the isolated plasma membranes was performed using the marker enzymes alkaline phosphatase and acrosin, as previously described (Soucek and Vary, 1984).
Western blotting
Twenty micrograms of protein (plasma membrane fractions and whole cell lysates) were solubilized with Laemmli buffer and separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE). Subsequently, proteins were transferred to a polyvinylidene difluoride (PVDF) membrane. After blocking with 1% (w/v) BSA and 0.05% (v/v) Tween 20 in Tris-buffered saline (TBS; 20 mM Tris, 500 mM NaCl; pH 7.4), the blots were incubated with caveolin-1 antibody [BD Biosciences, San Jose, CA, USA; diluted 1:2000 in incubation buffer (TBS containing 0.1% (w/v) BSA and 0.05% (v/v) Tween 20] or flotillin-1 antibody (BD Biosciences, San Jose, CA, USA; diluted 1:250 in incubation buffer). The original concentration of both primary antibodies was 0.25 mg/ml. The membranes were washed with TBS containing 0.05% (v/v) Tween 20 and subsequently incubated with goat anti-rabbit alkaline phosphatase and goat anti-mouse alkaline phosphatase for the rabbit anti-caveolin-1 and mouse anti-flotillin-1 antibodies, respectively (ECF detection kit, Amersham Biosciences, Uppsala, Sweden; diluted 1:4000 in incubation buffer). Specific antibody binding was detected using ECF™ chemiluminesence substrate (Amersham Biosciences, Uppsala, Sweden) on a STORM analyzer (Molecular Dynamics, Sunnyvale, CA, USA).
Immunofluorescence
Incubated sperm suspensions were labelled with merocyanine (M540) and Yo-Pro-1 and sorted on a FACS Vantage SE (Becton Dickinson, San Jose, CA, USA), as described before (Flesch et al., 2001). Sorted viable M540-negative (low fluorescent) and M540-positive (high fluorescent) cells were sorted in 4°C fixative of 4% paraformaldehyde in phosphate-buffered saline (PBS). Fixed cells were air-dried on microscope slides. Sperm cells were washed and permeabilized in methanol (previously cooled to −20°C) for 2 min. The cells were blocked with 3% (w/v) BSA in PBS (1.7 mM NaH2PO4, 5.0 mM Na2HPO4, 150 mM NaCl; pH 7.4) for 1 h at room temperature and subsequently incubated with caveolin-1 antibody or flotillin-1 antibody [diluted 1:100 to a final concentration of 2.5 µg/ml in PBS containing 0.1% (w/v) BSA and 0.05% (v/v) Tween 20]. Antibody incubations were performed in a humidity chamber at 37°C for 1 h. After washing three to five times with PBS containing 0.05% (v/v) Tween 20, the cover slips were incubated with goat anti-rabbit Alexa Fluor 488 conjugated [Molecular Probes, Leiden, The Netherlands; diluted 1:250 to a final concentration of 1 µg/ml in PBS containing 0.1% (w/v) BSA and 0.05% (v/v) Tween 20] and goat anti-mouse Alexa Fluor 488 conjugated [diluted 1:250 in PBS containing 0.1% (w/v) BSA and 0.05% (v/v) Tween 20]. The washing procedure was repeated, and the cover slips were sealed with Fluorsave (Calbiochem, San Diego, CA, USA). Control experiments were routinely done: (i) the incubation with primary antibodies was omitted and (ii) the incubation with primary antibodies that were preincubated with 20 µg/ml of the antigenic peptides. Both control experiments delivered unlabelled sperm. Slides were examined using an inverted spectral Confocal Laser Scanning Microscope (CLSM; Leica TCS SP, Leica GmbH, Wetzlar, Germany). The samples were illuminated with a 488 nm Ar Laser; emitted photons (510–530 nm range) were captured on a photomultiplier tube. Pinhole was set at 1 µm, and settings were made for maximal linear range of background and maximal emission of the brightest-labelled structures (set for capacitated sperm cells). The same settings were used for the detection of labelling of sperm cells after different incubations. In an area of 1 µm around the confocal plane, 10 confocal sections were made of eight scans per section, and average fluorescence was used for imaging in the extended focus mode (TCS software, Leica GmbH, Wetzlar, Germany).
Isolation of the DRM fraction
DRMs were isolated according to Martens et al. (2000). Briefly, washed sperm cells were resuspended (concentration ∼109 cells/ml) in 2-(N-morpholino) ethanesulfonic acid (Mes) buffer (25 mM Mes, 150 mM NaCl, 1 mM EGTA, protease inhibitors; Complete, Roche Diagnostics GmbH, Mannheim, Germany; pH 6.5) with 1% (v/v) Triton X-100 and kept on ice for 10 min. The suspension was mixed with the same volume of an 80% buffered sucrose solution, and a sucrose gradient (containing 6 ml of 30% sucrose in Mes buffer and 4 ml of 5% sucrose in Mes buffer) was layered on top of it. After 18 h centrifugation (200 000 g, 4°C), the DRM fraction appeared as an opalescent band in the low-density fraction of the gradient and isolated for further analysis.
Lipid analysis
Lipids were extracted according to Bligh and Dyer (1959) from the DRM fractions of control and capacitated sperm cells, and total lipid extracts were used for further analysis. The sterols and molecular species of phosphatidylcholine (PC) and sphingomyelin (SM) were separated according to Brouwers et al. (1998a) with a slightly modified mobile phase of acetonitrile (ACN) : methanol : triethylamine (25:24:1) on two LiChrospher 100 RP18-e columns (5 µm; Merck, Darmstadt, Germany) in series. Lipids were detected with a Varex MKIII light scattering detector (ELSD, Alltech, Deerfield, IL, USA). Identification of molecular species was performed by online (tandem) mass spectrometry on an API-365 triple stage quadrupole mass spectrometer (Sciex, Ontario, Canada), as described previously (Brouwers et al., 1998b).
2D electrophoretic separation of DRM proteins
The proteins in the DRM fraction were precipitated with trichloroacetic acid [10% (v/v)]. The pellet was washed twice with ice-cold acetone and air-dried. The proteins of the DRM fraction and the pellet fraction were solubilized in rehydration buffer [7 M urea, 2 M thio-urea, 4% (w/v) 3-[(3-cholamidopropyl)-dimethylammonio]-propane-sulfonate (CHAPS), 18 mM dithiothreitol (DTT), 2% (v/v) IPG buffer (pH 3–10) (Amersham Biosciences, Uppsala, Sweden) and a trace of bromophenol blue]. In the first dimension, 250 µg protein was subjected to isoelectric focussing on linear Immobuline™ dry strip and IPG™ buffer pH 3–10, 18 cm (Amersham Biosciences, Uppsala, Sweden) with a Multiphor II system [Amersham Biosciences, Uppsala, Sweden; 150 V (30 min), 300 V (1 h), 1500 V (1 h), 3000 V (8 h)].
The focussed strip was rotated in 65 mM DTT in equilibration buffer [50 mM Tris/HCl (from stock 1.5 M Tris/HCl, pH 8.8), 6 M urea, 30% (v/v) glycerol, 2% (w/v) SDS, 0.005% (w/v) bromophenol blue; 10 ml for 15 min] and subsequently in 135 mM iodoacetamide in equilibration buffer (10 ml for 15 min). For the second dimension, SDS–PAGE was performed using 11% acrylamide gels. The 2D gels were silver stained according to Blum et al. (1987) with small modifications. The gels were fixed in 40% (v/v) ethanol with 10% (v/v) acetic acid and washed twice with 30% (v/v) ethanol and once with distilled water. The gels were sensitized for 1 min in 0.02% (w/v) sodium thiosulphate and subsequently washed three times with MilliQ water for 20 s. The gels were incubated in a cold (4°C) 0.1% (w/v) silver nitrate solution and subsequently washed three times with MilliQ water for 20 s. The gels were developed in 1% (w/v) sodium carbonate, 0.017% (v/v) formaldehyde solution, and the staining was terminated with 5% (v/v) acetic acid.
Protein identification
Predominant spots of the DRM fraction that were consistently enriched (at least in four of five gels) and that were absent or only in reduced amounts detectable in the pellet fraction were considered to be of interest and were excised and digested with trypsin. For the in-gel digestion, excised spots were washed with water and destained with a solution containing 30 mM potassium ferricyanide and 100 mM thiosulphate. The spots were washed with water and dehydrated with ACN. Subsequently they were reduced with DTT (6.5 mM DTT in 50 mM ammonium bicarbonate, pH 8.5), dehydrated with ACN and alkylated (54 mM iodoacetamide in 50 mM ammonium bicarbonate, pH 8.5). After alkylation, the gel plugs were washed with 50 mM ammonium bicarbonate and dehydrated with CAN, and this was repeated three times. After the last ACN step, the gel plugs were dried in air and swollen with 12 µl 10 ng/µl trypsin (modified sequencing grade, Roche, Mannheim, Germany) in 50 mM ammonium bicarbonate on ice. After 30 min on ice, the gel pieces were digested overnight at 37°C. Digestion was stopped with 1% (v/v) trifluoroacetic acid (TFA), and the peptides were separated from the gel plug and stored in Eppendorf tubes at −20°C. The peptides (in 50 mM ammonium bicarbonate with 0.1 % (v/v) TFA; injection volume 5 µl) were separated using capillary LC (Waters, Elstree, Hertford, UK) and analysed with a nano-electrospray ionization (ESI) quadruple time of flight (Q-Tof) MS/MS (Q-Tof Ultima GLOBAL, Micromass UK Ltd, Manchester, UK; funded and maintained by BBSRC grant 6/JIF13209). The peptide mixture was trapped on the precolumn [PepMap C18 column 300 µm internal diameter × 5 mm, Dionex (UK) Ltd] with 93% solvent A [0.1% (v/v) formic acid] and 7% solvent B [ACN with 0,1% (v/v) formic acid]. The peptides were gradient eluted (7–90% B) on a PepMap C18 column [75 µm internal diameter × 15 cm; Cat. No. 160396, Dionex (UK) Ltd] over 45 min into the mass spectrometer. The raw data were processed using MassLynx 3.5 software (Micromass). MS/MS spectra were matched to sequences in various protein databases (MSDB, SwissProt and the NCBI nonidentical protein sequence database) using Mascot software (Matrix Science, London, UK). In these searches, static modification of cysteine (carbamidomethylation) and differential modification of methionine (oxidation) were selected. Matching of a peptide to a protein was done by the evaluation of the quality of the raw MS/MS data. We required the presence of good quality y-ion data for at least two peptides for each protein. Finally each of the peptides was used for a BLAST search to confirm that the protein identified by MASCOT was the only relevant match in the nonredundant protein database for a particular peptide sequence (Kinter and Sherman, 2000).
Results
Localization of raft proteins
We investigated the presence and dynamics of lipid rafts in boar sperm cells.
We determined the presence of caveolin-1 and flotillin-1 in both total sperm cell extracts and apical plasma membrane extracts by western blot analysis.
The purity of the apical plasma membrane fractions was checked with enzyme assays using well-established techniques in our group (see Flesch et al., 1998 for full validation of this method). Alkaline phosphatase was used as a marker for the plasma membrane and proved to be 18 ± 2 (mean value ± SEM, n = 5) times enriched in plasma membrane preparations. For the determination of the acrosomal contamination, the marker enzyme acrosin was used. The apical plasma membranes were highly purified because only 2 ± 1% (mean value ± SEM, n = 5) acrosin activity was detected compared with whole sperm samples.
Both caveolin-1 and flotillin-1 could readily be detected in boar sperm cells (Figure 1). We examined three conditions to determine the effects of capacitation on the presence of raft proteins in the apical plasma membrane fraction versus the whole cell: (i) sperm cells that were not incubated (indicated as t = 0), (ii) sperm cells incubated in Tyrode’s medium without bicarbonate (indicated as 0 mM bicarbonate) and (iii) sperm cells incubated in Tyrode’s medium supplemented with bicarbonate to induce capacitation (indicated as 15 mM bicarbonate). Flotillin-1 was enriched in the apical plasma membrane fraction of the sperm cell in all three conditions, whereas this enrichment was not detected for the monomeric form of caveolin-1 (Figure 1). This different partitioning behaviour may be a result of the presence of caveolin-1 but the absence of flotillin-1 in the tail (conform Figure 2). On the other hand, the degree of caveolin-1 oligomerization in the apical plasma membranes was higher than in the whole cell homogenate (Figure 1).
Western blot for caveolin-1 (A) and flotillin-1 (B). Both apical plasma membrane fractions (indicated as PM) and total cell extract (indicated as SP) were analysed. Sperm cells that were not incubated were used (–), sperm cells incubated in absence of bicarbonate (0) and sperm cells incubated with 15 mM bicarbonate (15), to induce capacitation. On each lane, 20 µg protein was loaded. A representative immunoblot is depicted (n = 4). Note that a higher degree of caveolin-1 oligomerization was noted for the apical plasma membranes when compared with whole cells (oligomerized forms of caveolin-1 are indicated with >). The recovery of caveolin-1 and flotillin-1 in the isolated apical plasma membrane fractions are indicated as % of immunostaining detected with a densitometer in the whole cell homogenates (mean values, n = 4). Always the SD values were <10% of the mean values indicated. A separate lane of the blot contains a caveolin-1- or flotillin-1-containing protein sample delivered along with their primary antibodies (+C).
Immunofluorescence detection of caveolin-1 (A–C) and flotillin-1 (D–I) in sorted M540 low-fluorescent (A, D and G) and M540 high-fluorescent sperm cells (B, E and H) and in methyl-β-cyclodextrin-treated (MBCD-treated) sperm cells (C, F and I) with confocal laser scanning microscope. Flow cytometric analysis and sorting of low (A, D and G) and high M540 fluorescent (B, E and H) viable cell subpopulations. The relative proportion of the three M540/Yo-Pro-1 subpopulations are provided from five independent experiments (mean % ± SD). After sorting, the cells were fixed, permeabilized and the localization of the raft proteins caveolin-1 (A and B) and flotillin-1 (D, E, G and H) were determined. The immunolabelling patterns of sorted sperm subpopulations indicated are representative for >95% of the cells. Note that <2% of the negative-stained M540 cells showed weak labelling of flotillin-1 in the tail (in contrast to D and G). To induce raft disruption, we incubated sperm cells with 10 mM MBCD for 30 min. The MBCD-treated cells were fixed, permeabilized and immunolabelled for caveolin-1 (C) and flotillin-1 (F and I). Arrows indicate the orientation of the apical tip of the sperm heads. Panels G–I are immunofluorescent pictures for flotillin-1 labelling that are merged with phase contrast images. All sperm images were made in the extended focus mode of the confocal microscope using a pinhole of one micron. No fluorescence could be detected after incubations with just secondary antibody or with primary antibody that was preincubated with antigenic peptides (data not shown).
Immunofluorescence studies were performed to determine the localization of these raft proteins. Sperm cells were incubated in the presence and absence of bicarbonate and stained with M540 and Yo-Pro to differentiate capacitating from noncapacitated and deteriorated cells, respectively (Flesch et al., 2001). M540-negative, viable cells and M540-positive, viable cells were sorted in fixative (Figure 2). Immunofluorescence experiments showed that both caveolin-1 and flotillin-1 were present in the apical area of the sperm head (Figure 2A–I), whereas only caveolin-1 was detected in the tail of the sperm cell as well (Figure 2A and B). Both proteins were present in a punctuate pattern in M540-negative, viable cells (Figure 2A, D and G). Similar patterns were almost exclusively predominant (>95% of the sperm cells) in sperm samples that were not subjected to a capacitation treatment (data not shown). In M540-positive, viable cells (indicating sperm capacitation; Flesch et al., 2001), a redistribution of the proteins in the apical head that resulted in a concentration to the apical ridge area was detected (Figure 2B, E and H). This change in fluorescence distribution pattern of raft proteins was compared with the patterns of sperm cells incubated with MBCD. MBCD disrupts rafts by its ability to extract cholesterol from the membrane. MBCD-treated cells showed a disappearance of the punctuate pattern as well as a diffusion of raft proteins from the apical ridge area (Figure 2C, F and I). Consequently to this, the fluorescent signal per surface area in the MBCD-treated cells was weaker as compared with control and capacitated sperm cells.
Lipid analysis
The effect of capacitation on quantity and constitution of lipid rafts was also determined by lipid analysis. If capacitation induced a disruption of lipid rafts, the amount of lipid in the DRM fraction of capacitated sperm cells would be expected to decrease when compared with uncapacitated cells. On the other hand, no changes are expected to be observed when lipid rafts aggregate. To discriminate between these two options, we determined the amount of lipid including cholesterol in the DRM fractions of both control (incubation without bicarbonate) and capacitated sperm cells. To this end, we extracted the lipids of the light buoyant fractions, and the lipid composition was analysed with high-performance liquid chromatography-MS (HPLC-MS). With this method, we were able to analyse the PC species (the major phospholipid class present in sperm cells), the SM species and cholesterol simultaneously. The DRM fraction of sperm cells was enriched in cholesterol, as described before in other cell types (Simons and Ikonen, 2000): the sterol/PC ratio increased significantly from 0.50 ± 0.06 (mean value ± SD, n = 3) in total sperm cells to 0.79 ± 0.05 (mean value ± SD, n = 3) in the DRM fraction (P < 0.05). Interestingly, the lipid composition of the DRM fraction did not change after capacitation, neither qualitatively (molecular species of lipids) (Figure 3) nor quantitatively: the amount of lipids (PC, SM and cholesterol) in the DRM fraction of capacitated sperm cells, as percentage of the control incubation (0 mM bicarbonate), was 106 ± 10% (mean value ± SD, n = 3, P = 0.4). In contrast, the DRM fraction derived from MBCD-treated sperm cells showed a significant decrease in all lipids analysed (cholesterol, PC and SM): in the DRM fraction 66 ± 13% (SD, n = 3) of lipid was present, as percentage of the control incubation (P = 0.002), demonstrating disruption of DRMs. These data support and extend the work from Shadan et al. (2004).
Separation of sterols and molecular species of phosphatidylcholine (PC) and sphingomyelin (SM) by reverse phase high-performance liquid chromatography (HPLC). Lipids were detected by evaporative light scattering detection and identified by online electrospray mass spectrometry. Top panel: detergent-resistant membrane (DRM) fraction of noncapacitated, control sperm cells and capacitated sperm cells. Bottom panel: DRM fraction of control sperm cells and methyl-β-cyclodextrin-treated (MBCD-treated) sperm cells. No differences were observed in either lipid composition or amount of lipid in the DRM fraction after capacitation: the amount of lipids (PC, SM and cholesterol) in the DRM fraction of capacitated sperm cells, as percentage of the control incubation (0 mM bicarbonate), was 106 ± 10% (mean value ± SD, n = 3, P = 0.4). In contrast, the amount of lipid in MBCD-treated cells was severely decreased: in the DRM fraction, 66 ± 13% (SD, n = 3) of lipid was present, as percentage of the control incubation (P = 0.002). Representative chromatograms of one experiment shown.
Remarkably the capacitation-dependent and albumin-mediated cholesterol depletion was noted in the pellet fraction of the gradient. The lipids that reside here represent the lipid composition of the non-DRM fraction. In capacitated sperm, this non-DRM lipids show a decrease of 20 ± 4% of cholesterol and no decrease in phospholipids when compared with noncapacitated sperm (n = 3, data not shown).
Identification of DRM proteins
Because lipid rafts are able to concentrate proteins involved in signal transduction events (Simons and Toomre, 2000), we investigated which proteins are enriched in lipid rafts of sperm cells. The proteins in the DRM fraction were separated by 2D-gel electrophoresis (Figure 4). For comparison, the nonraft fraction (pellet of the gradient) was analysed as well. Fifteen major spots that were reproducibly present (al least in four of five gels) in the DRM fraction and absent or only marginally present in the pellet fraction were considered to be of interest. For protein spots 11, 12, 13 and 15 (Figure 4A), these criteria were somewhat arbitrary because a large protein smear appeared at pI 9.5 in the second dimension of the pellet fraction (Figure 4B). Nevertheless, spots 11, 12, 13 and 15 were selected because they were the major spots of the DRM fraction. All numbered spots were cut out and analysed by trypsin digestion followed by nano LC-MS/MS to enable protein identification from MS/MS data.
2D gel of proteins in the detergent-resistant membrane (DRM) fraction (A) and pellet (nonraft) fraction (B) of sperm cells. The proteins present in the light-buoyant (gel A) fraction and the proteins of the pellet fraction (gel B) of the sucrose gradient were separated with 2D electrophoresis and silver stained. The spots that were considered to be of interest (major, DRM-specific proteins) are labelled with numbers in gel A and were subjected to analysis by nano liquid chromatography mass spectrometry (LC-MS). For the identification of spots, see Table I. The localization of the investigated DRM spots is indicated with circles on the nonraft gel (B).
Database searches led to the identification of 12 of these spots, as shown in Table i. The raw MS/MS data of all 12 spots showed the presence of good quality y-ion data for at least two peptides for each protein. Spot numbers 5, 7, 8, 13 and 15 were considered as definite identifications because the pI value and molecular mass corresponded with literature. Spot 1, 3, 4, 9, 11, 12 and 14 are probable identifications, whereas their pI value or molecular mass did not correspond exactly with the literature. Most of the identified proteins are known to be present in sperm cells, and some, like spermadhesin AQN-3, are even sperm specific. However, here we report for the first time the presence of peroxiredoxin 5 in sperm cells.
Tandem mass spectrometry (MS/MS) identification of detergent-resistant membrane (DRM) proteins
| Spot | Protein assigned from database search of sequenced peptides | Accession number | Number of peptides matched | Comment |
|---|---|---|---|---|
| 1 | Fertilin beta | 28564477 | 6 | B |
| 3 | Alpha-s1-casein precursor | 1070620 | 3 | B |
| 4 | sp32 precursor | 1082952 | 15 | B |
| 5 | Aldose (aldehyde) reductase | 584742 | 10 | A |
| 7 | Superoxide dismutase | 66364 | 6 | A |
| 8 | Peroxiredoxin 5 | 10305336 | 12 | A |
| 9 | sp32 precursor | 1082952 | 5 | B |
| 11 | Spermadhesin AQN-3 | 114083 | 3 | A |
| 12 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 13 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 14 | sp32 precursor | 1082952 | 9 | B |
| 15 | Preproacrosin | 1480413 | 2 | A |
| Spot | Protein assigned from database search of sequenced peptides | Accession number | Number of peptides matched | Comment |
|---|---|---|---|---|
| 1 | Fertilin beta | 28564477 | 6 | B |
| 3 | Alpha-s1-casein precursor | 1070620 | 3 | B |
| 4 | sp32 precursor | 1082952 | 15 | B |
| 5 | Aldose (aldehyde) reductase | 584742 | 10 | A |
| 7 | Superoxide dismutase | 66364 | 6 | A |
| 8 | Peroxiredoxin 5 | 10305336 | 12 | A |
| 9 | sp32 precursor | 1082952 | 5 | B |
| 11 | Spermadhesin AQN-3 | 114083 | 3 | A |
| 12 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 13 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 14 | sp32 precursor | 1082952 | 9 | B |
| 15 | Preproacrosin | 1480413 | 2 | A |
Proteins were cut out from silver-stained 2D gels, digested with trypsin and microsequenced by MS/MS. To assign an excised spot to a particular protein, at least two peptides with validated MS/MS data from a single spot matched the database sequence for the assigned protein. A, definitive protein identification. Protein mass and pI on gel were the same, as previously reported. B, probable identification by MS/MS data. Different molecular weight and/or pI value on gel than expected. Differences can be caused by alternative splicing or posttranslational modifications. Note that spots 1, 4, 7 and 11–15 have been identified previously from sperm (see text for references).
Tandem mass spectrometry (MS/MS) identification of detergent-resistant membrane (DRM) proteins
| Spot | Protein assigned from database search of sequenced peptides | Accession number | Number of peptides matched | Comment |
|---|---|---|---|---|
| 1 | Fertilin beta | 28564477 | 6 | B |
| 3 | Alpha-s1-casein precursor | 1070620 | 3 | B |
| 4 | sp32 precursor | 1082952 | 15 | B |
| 5 | Aldose (aldehyde) reductase | 584742 | 10 | A |
| 7 | Superoxide dismutase | 66364 | 6 | A |
| 8 | Peroxiredoxin 5 | 10305336 | 12 | A |
| 9 | sp32 precursor | 1082952 | 5 | B |
| 11 | Spermadhesin AQN-3 | 114083 | 3 | A |
| 12 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 13 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 14 | sp32 precursor | 1082952 | 9 | B |
| 15 | Preproacrosin | 1480413 | 2 | A |
| Spot | Protein assigned from database search of sequenced peptides | Accession number | Number of peptides matched | Comment |
|---|---|---|---|---|
| 1 | Fertilin beta | 28564477 | 6 | B |
| 3 | Alpha-s1-casein precursor | 1070620 | 3 | B |
| 4 | sp32 precursor | 1082952 | 15 | B |
| 5 | Aldose (aldehyde) reductase | 584742 | 10 | A |
| 7 | Superoxide dismutase | 66364 | 6 | A |
| 8 | Peroxiredoxin 5 | 10305336 | 12 | A |
| 9 | sp32 precursor | 1082952 | 5 | B |
| 11 | Spermadhesin AQN-3 | 114083 | 3 | A |
| 12 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 13 | Spermadhesin AQN-3 | 114083 | 2 | A |
| 14 | sp32 precursor | 1082952 | 9 | B |
| 15 | Preproacrosin | 1480413 | 2 | A |
Proteins were cut out from silver-stained 2D gels, digested with trypsin and microsequenced by MS/MS. To assign an excised spot to a particular protein, at least two peptides with validated MS/MS data from a single spot matched the database sequence for the assigned protein. A, definitive protein identification. Protein mass and pI on gel were the same, as previously reported. B, probable identification by MS/MS data. Different molecular weight and/or pI value on gel than expected. Differences can be caused by alternative splicing or posttranslational modifications. Note that spots 1, 4, 7 and 11–15 have been identified previously from sperm (see text for references).
Discussion
In this article, the presence and dynamics of lipid rafts were assessed during sperm capacitation in vitro. It has been known for years that lipids, and especially cholesterol, play an important role in the capacitation process (Visconti et al., 1999a,c). Bicarbonate-induced sperm capacitation causes a lateral redistribution of cholesterol, which is followed by a lipoprotein-mediated cholesterol efflux (Flesch et al., 2001). Gadella and Harrison (2002) showed that bicarbonate-mediated cholesterol redistribution and uptake by albumin is preceded by and dependent on phospholipid scrambling.
Lipid rafts are defined as cholesterol-enriched domains in the membrane, and cholesterol plays a structural role in lipid rafts (Simons and Ikonen, 2000). Therefore, we were interested in the organization of lipid rafts during sperm capacitation. Because lipid rafts are known to play a role in many signalling events by assembling functional protein complexes (Simons and Toomre, 2000), we isolated and identified the main proteins in the DRM fraction of sperm cells.
Distribution of caveolin-1 and flotillin-1
A first indication for the presence of lipid rafts is the detection of the raft marker proteins caveolin-1 (Okamoto et al., 1998; Smart et al., 1999) and flotillin-1 (Bickel et al., 1997; Volonte et al., 1999). It has already been shown that caveolin-1 is present in mouse, guinea pig and rat sperm cells (Travis et al., 2001; Trevino et al., 2001; Evans et al., 2003), that flotillin-1 is present in rat sperm (Evans et al., 2003) and that flotillin-2 is present in human sperm (Cross, 2004). Western blot analysis revealed that flotillin-1 and caveolin-1 proteins are also present in boar sperm cells (Figure 1), indicating the presence of lipid rafts. Both proteins were detected in the apical plasma membrane fraction, and flotillin-1 was enriched in the apical plasma membrane fraction compared with the total cell extract.
Flesch et al. (2001) showed that bicarbonate induces a redistribution of cholesterol to the apical plasma membrane of the sperm cell. Here, we did not observe a difference in enrichment of caveolin-1 and flotillin-1 in the plasma membrane fractions of capacitated sperm cells compared with the control cells with western blot experiments, indicating that there is no redistribution of raft proteins from other subcellular membrane regions to the apical plasma membrane.
The raft proteins were also detected by immunolocalization using confocal microscopy. The enrichment of flotillin-1 to the plasma membrane overlying the apical head was confirmed with immunofluorescence microscopical data (Figure 2D and G), clearly showing the presence of flotillin-1 at the apical site of the sperm head. A similar enrichment in the sperm head was observed for caveolin-1. However, it was also shown that, besides being enriched in the apical head area, caveolin-1 was also present in the tail. The different partition behaviour of the two raft proteins (Figure 1) may be a result of the presence of caveolin-1 but the absence of flotillin-1 in the tail (Figure 2). The tail membranes are not isolated in the apical plasma membrane preparations (Flesch et al., 1998). On the other hand, caveolin-1 appears to be oligomerized to a higher degree in the apical sperm membrane isolate when compared with the whole cell fraction (Figure 1).
These findings correspond with the localization of caveolin-1 in mouse and Guinea-pig sperm described by Travis et al. (2001) and Trevino et al. (2001), although the latter describes the presence of caveolin-1 in the equatorial head surface of mouse sperm. In absence of bicarbonate, caveolin-1 and flotillin-1 both appeared in a punctate pattern (Figure 2); this may indicate that lipid rafts are dispersed over these plasma membrane areas. Capacitation induced a redistribution of both proteins that concentrated at the apical ridge area of the sperm head when compared with sperm cells incubated in bicarbonate-free media.
Reordering of lipid rafts during capacitation
Cross (2004) showed that the distribution of raft markers can shift towards the nonraft fraction during capacitation, which may represent a disruption of lipid rafts. However, we demonstrated that the amount of total lipids isolated in the DRM fraction did not change upon capacitation, indicating that lipid rafts are not disrupted during capacitation. Thus, the redistribution of caveolin and flotillin represents a capacitation-dependent reorganization of lipid rafts. Their concentration in the apical ridge surface area may form a support for the proposal that small microdomains cluster into a lager membrane domain, as proposed by Flesch et al. (2001). This was already shown by Shadan et al. (2004). We also observed the effect of MBCD on caveolin and flotillin distribution as well as on lipid composition of DRM fraction. MBCD disrupts lipid rafts by its ability to extract cholesterol from the membrane (Scheiffele et al., 1997; Ilangumaran and Hoessli, 1998). MBCD-treated cells did not show the same redistribution of the raft proteins (Figure 2): MBCD-treated sperm cells lost the punctate pattern, and in comparison with capacitated sperm cells, they had a weaker and more diffuse labelling. This may indicate that the cholesterol depletion of sperm cells using MBCD causes the dissociation of rafts (see also Flesch et al., 2001). Further to this, we isolated the DRM fraction from an equal amount of sperm cells incubated in the presence and the absence of bicarbonate. Lipid analysis of PC species (the major phospholipid class in mammalian sperm membranes), SM species and cholesterol revealed that there was no detectable difference, neither qualitatively nor quantitatively, between the DRM fractions of control and capacitated sperm cells, demonstrating that there is no disruption of lipid rafts (Figure 4). This is in contrast to the MBCD-treated cells in which all lipid classes, but especially cholesterol, were significantly decreased in the DRM fraction. Taken together, our immunofluorescence data and lipid analysis, along with work from others (Shadan et al., 2004), indicate that lipid rafts are present in sperm cells and that capacitation induces a redistribution of these microdomains. This results probably in a concentration of lipid rafts at the apical ridge area of the sperm head.
MBCD induces a disruption of lipid rafts and can be considered to nonphysiologcally induce cholesterol extraction, whereas capacitation induces a redistribution of cholesterol followed by a physiological efflux of cholesterol in the presence of a lipid acceptor like BSA.
Notably, the amount of cholesterol in the DRM fraction did not change during bicarbonate/BSA incubations compared with control incubations. This implies that albumin-mediated extraction of cholesterol from the sperm cell exclusively takes place in the nonraft surface area of sperm cells, in contrast to MBCD that also extracts cholesterol from the raft area (Figure 4). Our findings support the study of Shadan et al. (2004), who showed that, despite cholesterol extraction during capacitation, lipid rafts do not disintegrate during capacitation and speculated that a preferential loss of cholesterol from the nonraft pool may be the stimulus that promotes raft clustering over the anterior sperm head.
Protein composition of the sperm DRM fraction
Because caveolin-1 is a potent cholesterol-binding protein (Murata et al., 1995), the previously reported redistribution of cholesterol during capacitation may be linked to the observed redistribution of caveolin-1 (Flesch et al., 2001). Note that the apical ridge area of the sperm head is the specific site of sperm–zona binding as well as the initiation site of the acrosome reaction after this binding (for review, see Flesch and Gadella, 2000).
Thus, the abovementioned redistributions of both cholesterol and marker proteins may represent a reordering of lipid rafts that is required for gamete interaction. One function of lipid rafts is to concentrate signalling molecules, because functional cell signalling protein complexes are often assembled in lipid rafts (Foster et al., 2003). Protein complexes have also been suggested to operate during gamete interaction (for instance, during zona pellucida binding) (Thaler and Cardullo, 1996). Therefore, we identified the main proteins that are present in the DRM fraction of sperm to shed more light on the function of lipid rafts in sperm cells. The major proteins found in the DRM fraction (Table i) can be roughly divided into two groups: (i) proteins proposed to be involved in the (primary) binding to the zona pellucida (fertilin beta; sp32 precursor, spermadhesin AQN-3 and preproacrosin; Baba et al., 1994; Calvete et al., 1996; Cho et al., 1998; Howes and Jones, 2002) and (ii) proteins involved in redox balance (aldose reductase, superoxide dismutase and peroxiredoxin 5; de Lamirande and Gagnon, 1995; Rhee et al., 1999; Mura et al., 2003). The proteins of spots 2, 6 and 10 could not be identified, most likely because of the incomplete porcine genomic database. Many spots, like sp32 precursor and AQN3, have different molecular weight and/or differential pI values than the theoretical values, suggesting posttranslational modifications or alternative splicing. Of the three spots (numbers 11, 12 and 13) that were identified as AQN-3, only the protein with the lowest molecular weight (number 24) had the correctly predicted molecular weight and pI. The other two spots are probably differentially glycosylated forms of AQN-3, because this secretory protein has a conserved glycosylation sequence on an asparagine residue (Calvete et al., 1993).
The primary zona pellucida binding proteins are located at the apical plasma membrane of the sperm head (Yanagimachi, 1994), matching the area where redistribution of both caveolin-1 and flotillin-1 takes place after capacitation. This makes it attractive to speculate on raft distribution as the molecular mechanism responsible for the capacitation induced zona pellucida affinity of sperm cells (Yanagimachi, 1994; Harkema et al., 1998). Because lipid rafts are functioning as platforms for signalling receptors, clustering of these platforms could enhance protein interactions, thus triggering signalling. For sperm cells, clustering of lipid rafts that contain the zona pellucida-binding proteins could result in a concentration of those receptors, thereby enhancing the ability of the sperm cell to bind to the oocyte. Furthermore, the noted redistribution of raft proteins indicates that this concentration takes place at the apical tip of the sperm head, the place where binding to the zona pellucida is supposed to take place (Yanagimachi, 1994).
Superoxide dismutase, aldose reductase and also peroxiredoxin 5 are involved in maintaining the redox balance in cells (de Lamirande and Gagnon, 1995; Rhee et al., 2001; Mura et al., 2003). It has been shown that both capacitation and the acrosome reaction are redox-regulated processes (de Lamirande and Gagnon, 2003; Baker and Aitken, 2004). Again, concentrating these proteins by the clustering of lipid rafts would facilitate activating the signalling pathways involved in these processes.
In conclusion, there are some indications that suggest a formation of large raft structure(s) during capacitation because of the clustering of small microdomains. First, Flesch et al. (2001) showed a bicarbonate-dependent redistribution of cholesterol to the anterior head. Second, we show that both raft proteins caveolin-1 and flotillin-1 redistributed to the same area. Furthermore, the amount of lipid in the DRM fraction does not change during capacitation, indicating that no disintegration of lipid rafts takes place during this process. Clustering of lipid rafts could enhance certain protein interactions. This is of considerable note as some proteins involved in redox signalling and proteins proposed to be involved in zona pellucida binding have been found to be present in the DRM fraction. Further investigation should reveal the physiological consequence of bicarbonate-induced redistribution of lipid rafts and the concomitant relocalization of raft proteins to the apical plasma membrane, where the first interaction with the zona pellucida of the oocyte takes place.
Acknowledgements
This research was supported by a PhD grant for RAvG from the Dutch Medical Research Council (NWO-MW grant number 903-44-156).
References
Author notes
1Department of Biochemistry and Cell Biology, Institute of Biomembranes, Utrecht University, Utrecht, The Netherlands, 2The Reproductive Biology and Genetics Group, Division of Medical Sciences, The University of Birmingham, Birmingham, 3Department of Medical Biochemistry and Immunology and Biostatistics and Bioinformatics Unit, Cardiff University, Cardiff, 4School of Chemical Sciences, The University of Birmingham, Birmingham, UK and 5Department of Farm Animal Health, Graduate School of Animal Health, Utrecht University, Utrecht, the Netherlands
