| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gamete Biology |
Gamete Signalling Laboratory, Babraham Institute, Babraham, Cambridge CB2 4AT, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gamete biology, in vitro fertilization, sperm capacitation; sperm maturation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Further support for the cholesterol efflux hypothesis has arisen from the use of ß-cyclodextrins. These compounds have high affinity and specificity for sterols and are capable of removing >80% of membrane cholesterol from cells [15, 16]. The addition of methyl-ß-cyclodextrin (MBCD) to BSA-free media has been shown to induce capacitation-associated protein tyrosine phosphorylation in mouse [6], bull [6], stallion [17], and human [18] spermatozoa and to culminate in the attainment of full fertilizing capacity [19]. The reduction in membrane cholesterol necessary to achieve capacitation may be relatively small (
20%) and does not appear to compromise membrane integrity or sperm viability [16]. Cholesterol is the major sterol in mammalian spermatozoa (exceptions are hamster and mouse spermatozoa, which have 40% desmosterol [6, 20]) and is distributed throughout the plasma membrane with higher concentrations overlying the head region [21, 22].
Traditionally, cholesterol is thought to intercalate between phospholipids and glycolipids in the hydrophobic interior of cell membranes and to increase or decrease order, depending on the degree of saturation or unsaturation of the fatty acyl chains. Recent studies on model membranes, however, have shown that cholesterol also influences the miscibility of different phospholipids. Korlach [23] detected two coexisting phases in equimolar mixtures of dilauroylphosphatidylcholine and dipalmitylphosphatidylcholine that enlarged in the presence of 5 mol % cholesterol. With >10 mol % cholesterol, however, phase separation disappeared. Using atomic force microscopy (AFM) to visualize supported bilayers of sphingomyelin (SM) and dioleoylphosphatidylcholine, Lawrence et al. [24] observed formation of SM-microdomains that enlarged or dissolved depending on the proportion of cholesterol. These, and many other studies on model systems [25], have enforced the view that membranes are a mosaic of liquid ordered and liquid disordered phases and that the former equate with the detergent-resistant complexes or raft fraction now described in various cell types. Rafts are rich in glycosphingolipids, sphingomyelin, glycosylphosphatidylinositol (GPI)-anchored proteins, cholesterol, and various signal transduction proteins, e.g., Src family kinases [26, 27]. A reduction in membrane cholesterol has been reported to destabilize raft structures, leading to changes in their size, distribution, and content of glycoproteins [28].
Direct visualization of membrane rafts in live cells has so far proved difficult and much effort has been directed toward describing changes in the behavior of raft-associated markers following cholesterol efflux. In mast cells, it was found that regions of ordered lipids developed following crosslinking of IgE receptors [29], and in CHO cells micrometer-sized domains containing aggregations of sphingomyelin and DiIC16 appeared after cholesterol removal [30]. Other examples are dispersal of previously clustered foliate receptors [31], inhibition of SNARE aggregation prior to endocytosis [32], and delocalization of PI(4,5)P2 from the actin cytoskeleton [33]. It is fair to point out that other microscopic techniques, particularly those employing green fluorescent protein (GFP)-labeled proteins and fluorescence resonance energy transfer (FRET) techniques, have failed to visualize rafts in cell membranes, suggesting that they may be very small and transient [34]. Hao et al. [30] have cautioned that cell membranes may actually consist of small regions of liquid disordered (fluid) lipids dispersed within large areas of liquid ordered domains, a reversal of the normal concept of a small raft in a large sea of fluid lipid.
In this investigation, we have examined the hypothesis that cholesterol efflux from spermatozoa destabilizes lipid raft structures in the plasma membrane thereby initiating protein phosphorylation and acquisition of the capacitated state. The results support this concept by demonstrating that, once a critical reduction in cholesterol is reached, lipid rafts migrate to and cluster over the anterior sperm head, where they are correctly positioned to activate downstream signaling pathways leading to exocytosis of the acrosomal vesicle.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methyl-ß-cyclodextrin (MBCD), chlortetracycline (CTC), propidium iodide (PI) and FITC-labeled cholera toxin ß subunit (FITC-CTXB) were purchased from Sigma (Poole, UK). Antiphosphotyrosine monoclonal antibody (McAb; clone 4G10) was from UBI (Santa Cruz, CA). Filipin was obtained from Polysciences Inc. (Warrington, PA), 5-(N-octadecanoyl)aminofluorescein (ODAF) from Molecular Probes (Portland, OR), and Gelcode Blue from Pierce (Tattenhall, UK). All other chemicals were of the highest purity available commercially and were purchased from Merck-Eurolab (Lutterworth, UK) or Oxford Glycosystems (Oxford, UK).
Media
Boar spermatozoa were prepared in a TALP medium [35] containing 100 mM NaCl, 3.1 mM KCl, 25 mM NaHCO3, 3 mM NaH2PO4, 21.6 mM Na lactate, 2 mM CaCl2, 0.4 mM MgCl2, 10 mM Hepes, 1 mM Na pyruvate, and 5 mM glucose. Mouse spermatozoa were suspended in Hepes medium bicarbonate (HMB) medium [6] consisting of 119.4 mM NaCl, 4.78 mM KCl, 10 mM NaHCO3, 1.2 mM KH2PO4, 1.7 mM CaCl2, 1.2 mM MgSO4, 25 mM Hepes, 1 mM Na pyruvate, and 5.6 mM glucose. Immediately before each experiment, media were supplemented with MBCD as required, pH adjusted to 7.3 by gassing with 95% air:5% CO2, and sterilized using Stericup filtration (0.22 µm). The osmotic pressure of complete TALP and HMB media was 295–305 mOs/L.
Collection and Preparation of Spermatozoa
Ejaculated semen was collected from boars maintained at the Babraham Institute or obtained from a commercial breeder (JSR Newsham Ltd., Selby, UK). Maintenance of animals and all animal procedures were carried out with approval of local ethical committees in accordance with Home Office (UK) regulations. Semen was filtered through a double layer of surgical gauze, the spermatozoa washed by centrifugation through a step gradient of 35%:70% Percoll [36] and their concentration adjusted to 6 x 107/ml in TALP. Immature spermatozoa were collected from the testes of slaughtered boars by puncturing the external rete testis and draining the fluid into polythene capillary tubing. Caput and cauda epididymidal spermatozoa were obtained by slicing the tissue in TALP medium. Both testicular and epididymal spermatozoa were washed through Percoll as described above. Similarly, mouse (Black 6 strain) cauda epididymidal spermatozoa were collected by mincing the tissue in HMB medium and allowing the contents to disperse for 5 min. Spermatozoa were washed once by centrifugation in 10 ml at 800 x g for 10 min and resuspended in HMB to 2.5 x 107/ml. Sperm concentrations were determined using a hemocytometer. For capacitation studies, spermatozoa were diluted five times in the appropriate medium ± cholesterol acceptors and incubated in 95% air:5% CO2 at 37°C (mouse) or 38°C (boar).
SDS-PAGE, Immunoblotting, and Detection of Phosphorylated Proteins
Total proteins were extracted from spermatozoa by heating samples for 5 min at 100°C in Laemmli sample buffer [37] followed by centrifugation at 9000 x g for 10 min. Reduced proteins were separated in one dimension on 10% SDS-PAGE gels and either transferred to nylon membranes (Immobilon; Millipore, Bedford, MA) by Western blotting techniques or visualized by staining with 0.1% Coomassie Blue R-250 or silver [38]. Phosphorylated proteins were detected on blots using horse radish peroxidase-conjugated 4G10 McAb [39] and visualized by chemiluminescence procedures (Amersham, Little Chalfont, UK).
Two-dimensional SDS-PAGE of proteins was performed using the Millipore Investigator system as described by [40]. Washed boar sperm pellets (1.5–3.0 x 107cells) were dissolved in 25 µl of lysis buffer (9 M urea, 2% NP-40, 20 mM dithiothreitol) for 1 h at 23°C, centrifuged at 9000 x g for 10 min and the supernatant lysate kept on ice until loaded onto isoelectric focusing gels (pH range 3–10). Raft fractions from sucrose gradients (see below) were pooled, dialyzed overnight against water at 4°C and proteins concentrated by ethanol precipitation. Dried pellets were dissolved in 25 µl lysis buffer as described above. Gels were calibrated with Bio-Rad 2D standards (pI range 4.5–8.5, Mr 17.5–76 kDa) and proteins visualized by silver staining.
Assays for Capacitation: CTC Fluorescence and Induction of Acrsomal Exocytosis
The capacitation status of boar spermatozoa was assessed using the CTC fluorescence technique as described [41, 42] and responsiveness of the acrosome to Ca2+ ionophore A23187. For the former, aliquots of CTC-stained spermatozoa were mixed with an equal volume of antifade solution (Citifluor; Agar Ltd., Stansted, UK) and viewed under ultraviolet light (excitation 392 nm, emission 536 nm) with a Zeiss Axiophot photomicroscope equipped with epifluorescence optics. At least 200 spermatozoa were counted for each treatment and were classified into one of three categories: uncapacitated (F pattern) with uniform fluorescence over the entire head region, capacitated (B pattern) with no or weak fluorescence on the postacrosomal region, and acrosome reacted (AR pattern) with weak fluorescence over the entire head except for the equatorial segment.
For assessment of acrosomal responsiveness, spermatozoa were incubated for 60 min in TALP ± 5 mM MBCD, washed once and incubated with 2 µM A23187 for 15 min at room temperature. Exposure of acrosomal membranes and acrosomal contents was assessed by staining spermatozoa with 40 µg/ml FITC-peanut lectin (FITC-PNA) for 5 min and viewing under ultraviolet light.
Fluorescence-Activated Cell Sorting
Cholesterol distribution in the sperm plasma membrane was studied by staining with filipin. Aliquots (100 µl) of boar spermatozoa were mixed with 100 µl of filipin (200 µg/ml in TALP containing 4% [v/v] ethanol), incubated 15 min at room temperature in the dark and washed twice by centrifugation in TALP. Spermatozoa were viewed either by fluorescence microscopy (excitation 365 nm, emission > 420 nm) or analyzed on a fluorescence-activated cell sorting (FACS) Vantage flow cytometer (Becton Dickinson, San Jose, CA) using a FL4 530/30 nm band-pass filter. The system collects fluorescence data in logarithmic mode and light-scatter data in linear mode. Ten thousand cells were counted in each sample at a rate of 50–500 events per sec. Data were analyzed using the Cell Quest package.
For analysis of plasma membrane integrity, prepared spermatozoa were mixed with an equal volume of PI (25 µg/ml) for 5 min at room temperature and FACS analysis carried out immediately using a FL3 675/20 nm band-pass filter as described above.
Fluorescence Recovery after Photobleaching Analysis of Lipid Diffusion in Sperm Plasma Membranes
Lipid diffusion in the plasma membrane of mouse and boar spermatozoa was measured by fluorescence recovery after photobleaching analysis of lipid diffusion in sperm plasma membranes (FRAP) analysis using 5-(N-octadecanoyl)aminofluorescein (ODAF) as reporter probe [43, 44]. Briefly, 100 µl of spermatozoa were mixed with 100 µl of 0.2 µM ODAF in 2% (v/v) ethanol for 10 min in the dark and washed twice by centrifugation at 600 x g for 4 min in 1 ml of the appropriate buffer. Resuspended spermatozoa were drawn into capillary microslides (CamLab, Over, UK) and ODAF diffusion measured at 23°C using a custom-built photobleaching system with a spot diameter of 1.2 µm. When necessary, spermatozoa were immobilized with 0.01% sodium azide. The diffusion coefficient (D) and percentage recovery (%R) for ODAF were recorded for acrosomal, postacrosomal, midpiece, and principle piece domains of the plasma membrane and computed by the formulas described earlier [43]. Statistical analyses were performed by Excel XP SR-1 t-test for two-tailed samples of unequal variance.
Isolation of Lipid Rafts from Boar Spermatozoa
A low-density Triton X-100 insoluble membrane fraction was prepared from boar spermatozoa as described in [45] for lymphocytes. Briefly, boar spermatozoa (capacitated and noncapacitated) were concentrated to 109 cells/ml in TALP, mixed with an equal volume of ice-cold MBS (50 mM MES [pH 6.5], 150 mM NaCl, 1% Triton X-100, 2 mM Na3VO4, 4 mM AEBSF) and incubated for 20 min at 4°C. The supernatant (1.75 ml) from a low-speed centrifugation step (900 xg for 10 min at 4°C) containing Triton X-100-insoluble material was then mixed with an equal volume of 85% sucrose (w/v) in mMBS (25 mM MES [pH 6.5], 150 mM NaCl, 1 mM Na3VO4 and 2 mM AEBSF). The mixture was overlaid with 6 ml of 35% sucrose followed by 3 ml of 5% sucrose in mMBS and centrifuged at 200 000 x g for 18 h at 4°C in a SW41 Beckman rotor. One-ml fractions were collected from the top of the gradients and used for (i) measurement of light scattering properties by absorbance at 620 nm, (ii) protein determination using a BCA assay kit (Perbio Science, Tattenhall, UK), and (iii) SDS-PAGE and immunoblotting analysis following protein precipitation with 90% acetone at –20°.
Visualization of GM1 Gangliosides on Boar and Mouse Spermatozoa with FITC-CTXB
Suspensions of capacitated and uncapacitated spermatozoa were mixed with an equal volume of FITC-CTXB (10 µg/ml) and incubated for 30 min at 16°C [46]. Spermatozoa were washed once in 5 volumes of TALP or HMB medium, the pellets resuspended in medium containing Citifluor and observed directly by epifluorescence microscopy. Photographs were taken with a SPOT digital camera (Diagnostic Instruments, MI). For each treatment, at least 100 cells were counted and categorized into four different fluorescent patterns (see Results). The data were summarized using Excel XP Analysis ToolPak software.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MBCD-Mediated Cholesterol Efflux Induces Tyrosine Phosphorylation of a Subset of Specific Proteins in Boar Spermatozoa
Because BSA as a capacitating agent may have secondary unknown effects on sperm plasma membranes, we have focused on the greater specificity of water-soluble ß-cyclodextrins for mediating cholesterol efflux from boar spermatozoa. As shown in Figure 1A, increasing concentrations of MBCD (0.1–20 mM) in TALP medium induced phosphorylation of 2 proteins at 26 kDa (p26) and 15 kDa (p15) with a maximum response obtained at 5 and 10 mM. Several constitutively phosphorylated proteins were also present between 34–40 kDa. Modification of p26 was consistently more sensitive to lower levels of MBCD than p15, and high concentrations of MBCD (15–20 mM) suppressed phosphorylation of both proteins. A time course of changes induced by 5 mM MBCD revealed that phosphorylated p26 was detectable after 30 min and reached maximal levels by 60–90 min, whereas p15 required >180 min for optimal modification (Fig. 1B). The presence of 200 µM cholesterol sulfate in the capacitation medium strongly inhibited phosphorylation of p26 and p15 (Fig. 2) as well as having a suppressive effect on one of the constitutively phosphorylated proteins at 34 kDa. Figure 2 also illustrates that some phosphorylation of p26 can take place during incubation of spermatozoa in TALP medium alone and that the addition of 0.5 mM MBCD to suspensions containing 200 µM cholesterol could overcome the inhibition of phosphorylation of p15 but not p26. This suggests differential sensitivity of p15 and p26 to phosphorylation and that it is the rate of cholesterol efflux that determines the onset of capacitation changes in boar spermatozoa. It is accelerated in the presence of cholesterol acceptors such as MBCD and retarded by the addition of exogenous cholesterol sulfate.
|
|
Cholesterol Efflux Stimulates Protein Phosphorylation in Mature but Not Immature Spermatozoa
Spermatozoa leaving the testis are infertile largely because of their poor motility and inability to respond appropriately to capacitating media. Both these properties develop during passage through the caput epididymidis until maximum levels of fertilizing capacity are reached in the cauda epididymidis. As shown in Figure 3, maturation of signaling pathways is also reflected in phosphorylation of p15 and p26 proteins. Proteins from freshly collected testicular boar spermatozoa were unreactive to McAb 4G10, including the constitutively phosphorylated group at 34–40 kDa. During incubation in TALP medium alone, however, these proteins became phosphorylated but no further changes took place even in the presence of 5 mM MBCD. Spermatozoa from the caput and cauda epididymidis, on the other hand, responded to 5 mM MBCD by showing significant phosphorylation of both p15 and p26 proteins (Fig. 3), the effect being greater with the fully mature cauda cells. Once again, p26 was phosphorylated more rapidly than p15. These results indicate that the intracellular signaling pathways that mediate the stimulus provided by cholesterol efflux are assembled during posttesticular development of spermatozoa in the epididymis.
|
Correlation Between Cholesterol Efflux from the Plasma Membrane and Capacitation as Assessed by CTC Staining and Acrosomal Responsiveness to Ca2+ Ionophore
To quantify the relative amounts of cholesterol removed from the sperm's plasma membrane after incubation in MBCD and any subsequent effects on membrane integrity, samples of spermatozoa were stained with filipin or PI and analyzed by fluorescence microscopy and FACS. On uncapacitated spermatozoa, filipin staining was observed throughout the surface membrane, with localized concentrations of fluorescence, 0.5µ–1.5 µ diameter, on the acrosomal region (Fig. 4, A and B; described in detail in [22]). Incubation in 0.5 mM MBCD for 60 min slightly reduced the intensity of filipin staining (Fig. 4B), but in the presence of 5 mM MBCD, fluorescence was very weak (Fig. 4, C and D). FACS analysis of the latter spermatozoa indicated that 70% of cholesterol had been removed (Fig. 4E). Despite this, the plasma membrane retained its integrity as revealed by only a small increase in the proportion of PI-positive cells from 5% to 10% after treatment with 5 mM MBCD (Fig. 4F).
|
The capacitation status of spermatozoa was assessed using the CTC assay and their ability to undergo an acrosome reaction in response to Ca2+ ionophore. The CTC assay has been validated in several species and a strong correlation has been found between fertilizing capacity and the incidence of spermatozoa showing the so-called B staining pattern (see Materials and Methods); uncapacitated spermatozoa remain in the F pattern. To establish a link between cholesterol efflux and capacitation, boar spermatozoa were incubated in TALP ± MBCD under the same conditions described above for induction of phosphorylation and stained with CTC. As shown in Figure 5A, >75% of spermatozoa incubated in TALP medium alone for up to 180 min retained the F pattern with only 22% in the B pattern. The inclusion of 0.5 mM MBCD in the medium, however, increased the incidence of B-pattern spermatozoa to 55% after 180 min (Fig. 5B). In the presence of 5 mM MBCD, the rate of appearance of B-pattern spermatozoa increased substantially to the extent that 90% were present after only 60 min (Fig. 5C). Thus, the time frame for transition of spermatozoa from uncapacitated F pattern to capacitated B pattern correlates closely with that found for tyrosine phosphorylation of proteins p15 and p26.
|
Further evidence that 5 mM MBCD had induced capacitation was shown by enhanced responsiveness of spermatozoa to Ca2+ ionophore A23187. Only 10.6% ± 3.3% (n = 7) of control spermatozoa incubated in TALP stained positively with FITC-PNA lectin after exposure to 2 µM A23187, whereas after MBCD treatment, 31.9% ± 6.4% (n = 10) of spermatozoa bound the lectin over the acrosomal region. These figures are comparable with those reported by Tardif et al. [47], who used BSA to induce capacitation in boar spermatozoa.
Effects of Cholesterol Efflux on Lipid Diffusion in Sperm Plasma Membranes
Having established the conditions for capacitating boar spermatozoa with MBCD, the remaining experiments examined the effects of cholesterol efflux on lipid diffusion and raft organization in the plasma membrane. Several glycoproteins are known to change their surface distribution during sperm capacitation [48], but it is not known how the large concentration gradients generated within the membrane are maintained. One possibility is that cholesterol efflux, coupled with a breakdown in transverse lipid asymmetry, leads to major changes in lateral diffusion rates thereby enabling directed transport mechanisms. To test this hypothesis, boar spermatozoa were loaded with the fluorescent reporter probe ODAF, which intercalates into the outer leaflet of the lipid bilayer, and its diffusion measured by FRAP analysis after MBCD treatment. As reported previously [44], lipid diffusion is 5 times faster in the plasma membrane overlying the sperm head than the tail and this distinction is maintained during incubation in TALP medium without MBCD (Fig. 6A). Unexpectedly, the only significant difference in D values between incubated controls (0 mM, 60 min) and MBCD-treated spermatozoa (5 mM, 60 min) was on the postacrosomal region (Fig. 6A). The %R values, however, were significantly reduced to 40%– 50% on the acrosome, postacrosome, and midpiece regions (Fig. 6B). Very high concentrations of MBCD (20 mM) on the other hand, reduced D values significantly below controls on both head and tail regions indicating the formation of large areas of immobile phase lipids. These spermatozoa all showed the dead pattern of labeling with ODAF, indicating a permeabilized membrane [43].
|
To ensure that these results were not specific to boar spermatozoa, the experiments were repeated on mouse spermatozoa prelabeled with ODAF and incubated with MBCD (0.5 mM) under conditions known to initiate phosphorylation and capacitation [6]. Preliminary experiments showed that phosphorylated proteins had similar molecular masses to those described by Visconti et al. [4]. After incubation in 0.5 mM MBCD, D values were not substantially different from incubated controls (Fig. 6C) and were significantly lower only on the acrosome. However, when MBCD was increased to 6 mM significant decreases in D values were observed on midpiece and principal piece regions (Fig. 6, C and D).
Thus, although cholesterol efflux under conditions compatible with capacitation has little measurable effect on bulk lipid diffusion rates in the outer leaflet of the plasma membrane, the increase in the proportion of immobile phase lipids suggests subtle changes in bilayer organization. Complete removal of cholesterol with high concentrations of MBCD exacerbates this situation so that lipid diffusion is significantly reduced.
Characterization and Stability of Membrane Lipid Rafts in Spermatozoa Following Cholesterol Efflux
FRAP analysis does not have sufficient resolution to reveal nanometer-sized heterogeneities within membranes such as those thought to exist between liquid-ordered and liquid-disordered microdomains [49]. To investigate the hypothesis that cholesterol efflux during capacitation perturbs the equilibrium between raft and nonraft microdomains, leading to redistribution of their components, boar spermatozoa were incubated with 5 mM MBCD, extracted with cold Triton-X100, and the low density insoluble fraction recovered from sucrose gradients. In these experiments, 170 ml of sperm suspension (1.2 x 107/ml) were required to obtain sufficient material for analysis. In control samples, an opalescent band was obtained immediately below the 5%:35% sucrose interface, which is the characteristic buoyant density of membrane lipid rafts [34, 50]. The putative raft fractions 4 and 5 showed the highest light-scattering properties at 620 nm, consistent with a high content of lipids, whereas nonraft fractions 6–12 showed little or no absorbance (Fig. 7, A and B). Conversely, most of the recovered protein was present in the nonraft fractions with only 15% in the low-density raft fraction.
|
Following treatment of spermatozoa with 5 mM MBCD, the amounts of lipid and protein in raft fractions 4 and 5 were similar to, or slightly higher than, controls (Fig. 7, A and B). Thus, incubation of boar spermatozoa under conditions that induce capacitation appears to have little effect on the lipid and protein content of membrane rafts. Increasing the concentration of MBCD to 50 mM, however, had a severe effect on the stability of lipid rafts, causing a substantial decrease in the amount of low-density opalescent material at the 5%:35% sucrose interface and complete loss of raft-associated proteins (Fig. 7, A and B). All spermatozoa treated in this way were PI positive, indicating that the plasma membrane had been permeabilized (results not shown).
The number of the proteins associated with raft fractions 4 + 5 from control spermatozoa was analyzed by two-dimensional SDS-PAGE. Results revealed that it was enriched in a subset of 15–20 proteins ranging from 10kDa to 80 kDa that were minor components in the total profile of proteins (estimated to be >200) extractable from control spermatozoa (Fig. 8). This is consistent with current views that a restricted range of membrane proteins (predominantly those with GPI anchors) are found in lipid rafts [27]. On one-dimensional SDS-PAGE gels, a protein at 70 kDa stained strongest with Coomassie Blue (results not shown) and was considered to be a potential marker for localization studies on membrane rafts. Consequently, it was subjected to internal peptide sequencing by MALDI-TOF MS/MS analysis for further identification. The sequences obtained (peptide 1, YPSL/IDPH; peptide 2, NTGDL/IA), however, were not recognizable in any mammalian protein databases (SWISSPROT, EMBL) and attempts to localize the 70-kDa protein on spermatozoa with a polyclonal antibody were unsuccessful due to high levels of background staining with the preimmune serum (results not shown).
|
To determine if any of the inducible or constitutively phosphorylated proteins described earlier (see Figs. 1 and 2) were preferentially associated with raft or nonraft fractions, immunoblots of proteins in fractions 1–12 isolated from control spermatozoa were probed with McAb 4G10. Both the constitutive group at Mr 34–40 kDa and the induced p15 and p26 proteins were found predominantly in fractions 8–12 (Fig. 9A). Only a weak reaction was present in fractions 4–7 and this did not change significantly following cholesterol efflux under capacitating conditions. Therefore, both the constitutive and inducible phosphorylated proteins are found primarily in the nonraft fractions.
|
Distribution of GM1 Gangliosides in Spermatozoa During Capacitation
GM1 gangliosides are major components of lipid rafts [51]. They are well-known receptors for cholera toxin ß subunit (CTXB) and can be visualized by FITC-conjugated derivatives of the latter. When noncapacitated boar spermatozoa were probed with FITC-CTXB, fluorescence was observed predominantly over the tail region, with only a very weak signal on the head (pattern a; Fig. 10A). Following incubation in TALP medium + MBCD (0.5 mM and 5 mM), however, strong binding of FITC-CTXB was observed in different regions of the sperm head as well as the tail. The head labeling was classified into three additional staining patterns. Pattern b were labeled more strongly over the postacrosome than on the tail, pattern c showed stronger fluorescence on the midpiece and postacrosome with a characteristic semicircular ring of fluorescence outlining the anterior boundary of the equatorial segment. In these cells, the central area of the equatorial segment was unstained. Pattern d spermatozoa showed weak staining over the tail and midpiece with strong fluorescence on the acrosomal and postacrosomal regions; again, the central area of the equatorial segment was weakly stained. Analysis of the relative proportions of the above staining patterns in control spermatozoa showed a decline in the proportion of pattern a spermatozoa from 79% at 0 min to 52% after 60 min incubation (Fig. 10B). Concomitantly, patterns b, c, and d increased. In the presence of MBCD, there was a sharp decrease in the proportion of a pattern staining spermatozoa to <30% while the incidence of B + c + d patterns amounted to 64% after 60 min. The above results suggest that GM1 gangliosides on uncapacitated boar spermatozoa are distributed principally over the tail, but following cholesterol efflux, they diffuse progressively onto the sperm head with the exception of the equatorial segment.
|
In view of the significance of these observations, they were repeated on mouse spermatozoa incubated in HMB medium ± 0.5 mM MBCD and stained with FITC-CTXB as described above. Essentially, similar results were obtained (Fig. 11). With incubated control samples, >70% of spermatozoa labeled weakly over the entire surface or with slightly stronger fluorescence on the midpiece and postacrosomal areas (patterns i + ii, Fig. 11). Less than 8% labeled on the acrosome (patterns iii + iv). In the presence of 0.5 mM MBCD, the proportion of spermatozoa with patterns i + ii decreased to 29% while those showing patterns iii + iv increased to 70%. This trend was continued with 6 mM MBCD such that >90% of spermatozoa were in the iii + iv category. Thus, cholesterol efflux from boar and mouse spermatozoa elicits significant changes in the distribution of GM1 gangliosides concomitant with triggering tyrosine phosphorylation of specific proteins, a chain of events that culminates in the attainment of the capacitated state.
|
As a corollary to these experiments, we also investigated if capacitation-associated cholesterol efflux had any observable effects on the surface topography of the plasma membrane as viewed by AFM. AFM has recently been used to image the partitioning of alkaline phosphatase (a characteristic raft protein) into SM-rich domains in artificial bilayers prepared from phospholipid mixtures [52]. Unfortunately, consistent effects of MBCD treatment on either boar or mouse spermatozoa could not be obtained (results not shown).
Sphingomyelinase (SMase) Treatment of Spermatozoa Induces Protein Tyrosine-Phosphorylation in the Absence of Exogenous Sterol Acceptor
Together with cholesterol and glycosphingolipids, sphingomyelin (SM) is also a major component of membrane rafts [51]. Its hydrolysis by SMase has been shown to accelerate sterol efflux from the plasma membrane in cultured fibroblasts [53, 54] and to induce acrosomal exocytosis in human spermatozoa [55]. Furthermore, its degradation products (ceramide, sphingosine, and sphingosine-1-phosphate) have been claimed to have a direct role in initiating signal transduction processes involved in apoptosis, differentiation, and senescence [56, 57]. To determine whether SMase treatment would have an effect on cholesterol efflux and tyrosine-phosphorylation of p15 and p26 proteins, boar spermatozoa were incubated with the enzyme and subsequently stained directly with filipin for FACS analysis or proteins extracted with SDS and Western blots probed with McAb 4G10. Neither BSA nor MBCD were present in the external medium. As shown in Figure 12A, increasing amounts of SMase induced progressive tyrosine phosphorylation of both the aforementioned proteins without causing a detectable change in cholesterol content as assessed by binding of filipin (Fig. 12B). As observed in previous experiments, p26 was more sensitive to phosphorylation than p15. Subsequent FRAP analysis of SMase-treated spermatozoa did not reveal any significant effect on ODAF diffusion in any surface domain (results not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MBCD-Induced Protein Phosphorylation During Capacitation of Boar Spermatozoa
The important observations by Visconti and coworkers [3, 4] that tyrosine phosphorylation of a specific group of proteins in mouse spermatozoa correlated closely with acquisition of acrosomal responsiveness and fertilizing capacity established objective criteria for assessing the capacitation process. Phosphorylation was dependent on optimal levels of Ca2+ and HCO3– in the external medium and was regulated via cAMP activation of PKA. The upstream event that triggered these pathways was efflux of cholesterol from the plasma membrane irrespective of whether it was mediated by BSA or cyclodextrins. In mouse spermatozoa, an 20% decrease in membrane cholesterol was sufficient to initiate phosphorylation when BSA (3 mg/ml) was used as the sterol acceptor, whereas with MBCD (1 mM), an 60% reduction was necessary [6, 7]. A similar magnitude of cholesterol efflux was observed in the present investigation with MBCD treatment of boar spermatozoa. This suggests that BSA has additional effects on sperm membranes that augment the capacitation process. Cyclodextrins are known to be more efficient at removing cholesterol from membranes than protein acceptors because of their lower activation energies and the proximity with which they can approach the hydrophobic bilayer [16]. They neither bind to nor insert into the plasma membrane and, when saturated with cholesterol, can be used to deliver the sterol to membranes by back exchange. In somatic cells, at least two kinetic pools of cholesterol efflux have been described; a fast pool in the plasma membrane that is readily accessible to cyclodextrins and a slow pool that replenishes the fast pool from an intracellular source. In spermatozoa, which do not have endoplasmic reticulum or Golgi membranes, replenishment of the fast cholesterol pool could take place from in-plane rafts or even the cytoplasmic droplet. Non-apoptotic-related scrambling of the lipid bilayer has been reported in spermatozoa during capacitation [8] that supports the notion of cholesterol exchange between inner and outer leaflets, but more work needs to be done on the kinetics of cholesterol efflux from sperm membranes with MBCD and how it affects the balance of different sterol pools. As found in the mouse [3, 4], concentrations of MBCD greater than optimum suppressed phosphorylation probably through uncontrolled leakage of intracellular ions, ATP, etc., following a general loss of membrane integrity. In boar spermatozoa, the constitutively modified proteins at 34–40 kDa were dephosphorylated as well as p26 and p15, and virtually all spermatozoa became PI positive. Dephosphorylation may be due, in part, to release or activation of protein phosphatases, of which PP1 is present in spermatozoa [58]. Thus, there is a safe window for efflux of cholesterol from spermatozoa that is compatible with induction of capacitation and preservation of viability.
The unresponsiveness of immature testicular and caput epididymidal spermatozoa to MBCD treatment suggests that either the downstream signaling pathways have not been assembled correctly or that the transducing systems within the plasma membrane itself are nonfunctional. Currently, the balance of evidence favors the latter interpretation, as it is known that substantial remodeling of the sperm's plasma membrane takes place during epididymal maturation, especially in the content of phospholipids, sterols, and extrinsic glycoproteins [59]. Furthermore, epididymal fluid contains sterol-binding proteins [60, 61] that are potential reservoirs for back exchange of cholesterol with the sperm plasma membrane. Because it is important that spermatozoa do not capacitate or acrosome react prematurely in the epididymis, maintaining high levels of plasma membrane cholesterol would effectively decapacitate them until they are exposed to sterol acceptors in the female reproductive tract. The higher activity of PP1 in caput than cauda spermatozoa would also suppress protein phosphorylation [58].
Unlike mouse spermatozoa, in which proteins were phosphorylated at similar rates during capacitation [3], the kinetics of modification of p15 was noticeably slower than for p26. This may reflect their different content of tyrosine residues or their separate subcellular distribution. Immunofluorescence studies on mouse, hamster, and human spermatozoa have shown that many tyrosine phosphorylated proteins are localized within the flagellum, a distribution that is consistent with their identification as AKAPs and Ca2+-binding proteins attached to the fibrous sheath [62]. Sequential protein phosphorylation has been described in capacitating mouse spermatozoa, beginning in the principal piece and then extending into the midpiece, the onset of the latter coinciding with the ability to bind to the zona pellucida [63]. After fusion with the oolemma, sperm tail proteins were rapidly dephosphorylated. In capacitated boar spermatozoa, however, Tardif et al. [64] identified phosphorylated proteins (principally p32) within the acrosome rather than the tail. It is likely that p32 and p26 are the same protein, but until sequence data become available, this will remain conjectural. Recently, a second protein (TK-32) of similar size to p32 has been described [65]. Its relationship to p26 is unclear, however, as it shows tyrosine kinase activity. It is also unclear why phosphorylation of p15 has not been described hitherto. One possible explanation is that it requires higher levels of cholesterol efflux than p26 and is easily overlooked. The localization of p26 and p15 is not known, but we speculate that they are found in different subcellular compartments. Because the plasma membranes overlying the sperm head and tail have different compositions [59], differential rates of cholesterol efflux could lead to variable phosphorylation of p26 and p15. Last, the fact that all the phosphorylated proteins partitioned into the nonraft fractions on sucrose gradients is consistent with their association with cytoskeletal components.
Effects of Cholesterol Efflux on Plasma Membrane Fluidity and Lipid Raft Stability
Despite a substantial literature on the subject, the concept of lipid rafts in cell membranes is somewhat contentious [66], as their presence depends on insolubility in cold Triton-X 100 and low density on a sucrose gradient. Major unresolved problems are visualization of rafts in live cells and their presence at 37°C. Notwithstanding these difficulties, it is generally agreed that lipid rafts are enriched in glycosphingolipids, cholesterol, and GPI-containing glycoproteins; that they are very dynamic entities ranging in size from a few hundred to tens of thousands of molecules; and are linked to the cytoskeleton. The finding that certain classes of Src kinases are sequestered preferentially into lipid raft has enhanced their importance as organizing centers for signal transduction.
Lipid rafts have been described in sea urchin [67], mouse, and guinea pig spermatozoa [68], but their behavior in relation to capacitation has not been defined. In boar spermatozoa, the opalescent band at the 5%:35% sucrose interface was rich in lipids and contained a restricted subset of proteins, both characteristic features of lipid rafts. The level of cholesterol efflux that induced capacitation of boar spermatozoa (as assessed by tyrosine phosphorylation and CTC staining) had few detectable effects on the stability of these rafts; in fact, if anything, they enlarged. Further depletion of cholesterol, however, had the reverse effect, leading to their dispersal. In many respects, this is consistent with the FRAP data, which showed that only after treatment with high levels of MBCD was there a significant decrease in diffusion coefficients and formation of large areas of immobile lipids in the plasma membrane. As mentioned earlier (see Introduction), this paradoxical effect of cholesterol is well known from work on artificial bilayers. Extrapolating this data to biological membranes is difficult because of the enormous increase in complexity, but it has been shown that saponin treatment of BeWo carcinoma cells causes alkaline phosphatase (a raft marker) to become completely soluble in Triton X-100, implying redistribution into the nonraft regions of the plasma membrane [69]. Similarly, cholesterol depletion of Jurkat cells with 10 mM MBCD disrupts lipid rafts coincident with protein tyrosine phosphorylation and translocation of PKC from the cytoplasm to the surface [27]. Ilangumaran and Hoessli [70] have proposed a model in which MBCD preferentially extracts cholesterol from the glycerophospholipid-rich (nonraft) domains, leaving behind a core raft structure surrounded by a cholesterol collar. We speculate that a similar situation takes place in boar spermatozoa following treatment with 5 mM MBCD. Only with higher concentrations of MBCD would sufficient cholesterol be removed from raft cores to destabilize them and disperse their constituents. Thus, a decrease in nonraft cholesterol may actually be the crucial factor in initiating protein phosphorylation during sperm capacitation.
The finding that SMase was as effective as MBCD in initiating phosphorylation of p26 and p15 proteins without a detectable decrease in total membrane cholesterol implies that either SM breakdown products (mainly sphingosine-1-phosphate and ceramide) have signaling roles via alternative pathways, or else SM breakdown influences the flux of cholesterol between raft and nonraft domains. Cross [55] found that added ceramide or SMase treatment of human spermatozoa enhanced acrosomal responsiveness to progesterone with only small changes in their cholesterol content. The effects of SMase may also be indirect due to its known ability to deform lipid bilayers into hexagonal II phase structures and create disorder not unlike that following cholesterol efflux [71]. It remains to be determined if SMase perturbs partitioning of cholesterol between raft and nonraft domains.
Redistribution of GM1 Gangliosides During Capacitation of Boar and Mouse Spermatozoa
A particularly striking effect of MBCD-mediated capacitation was the increase in the proportion of spermatozoa showing FITC-CTXB binding to the acrosomal region. Although the possibility cannot be excluded that this reflects unmasking of binding sites, the balance of evidence favors the interpretation that they arise from polarized migration and clustering of GM1 gangliosides within the plane of the plasma membrane. First, there was loss of staining on the sperm tail (patterns a and b) concomitant with appearance of GM1 gangliosides on the acrosome (patterns c and d). Second, spermatozoa were not fixed and many remained motile after staining. This argues against exposure of internal antigens caused by permeabilization of the plasma membrane. Third, precedents for polarized migration of plasma membrane glycoproteins and glycolipids during sperm maturation and capacitation are well known, e.g., PH20 (a GPI-anchored protein [48, 72]), CE9 [73], and seminolipid [74]. More specifically, during capacitation of rat spermatozoa, Roberts et al. [75] found that binding of FITC-CTXB changed from the postacrosomal region to extend over the whole head and tail. Similarly, in somatic cells, cholesterol efflux leads to changes in surface distribution of many GPI-anchored proteins [49, 76]. Lipid rafts are now thought to be heterogeneous in composition [77] and clustering would provide a mechanism for signaling complexes to come together to initiate, for example, membrane fusion and exocytosis of the underlying acrosomal vesicle. In this regard, it is interesting that the equatorial segment, a region of the sperm surface that does not vesiculate during the acrosome reaction, remained largely devoid of GM1 gangliosides. The FITC-CTXB pattern d observed here is reminiscent of a subclass of capacitated boar spermatozoa described by Flesch et al. [8], in which there was bicarbonate-dependent clustering of filipin-sterol complexes over the anterior head. The authors speculated that this represented a large raft structure. This hypothesis remains to be demonstrated, but it is noteworthy that, in cultured CHO cells, removal of cholesterol promoted coalescence of microdomains into large micrometer-scale domains [30]. This supports the view that lipid rafts diffuse as separate entities and that super-rafts may form in membranes following cholesterol efflux [77].
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Roy Jones, Laboratory of Molecular Signalling, The Babraham Institute, Cambridge CB2 4AT, UK. FAX: 44 0 1223 496022. roy.jones{at}bbsrc.ac.uk
Received: 10 December 2003.
First decision: 4 January 2004.
Accepted: 9 March 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
