- Email: [email protected]
Different glycoforms of the human GPI-anchored antigen CD52 associate differently with lipid microdomains in leukocytes and sperm membranes
BBRC Biochemical and Biophysical Research Communications 338 (2005) 1275–1283 www.elsevier.com/locate/ybbrc
Different glycoforms of the human GPI-anchored antigen CD52 associate differently with lipid microdomains in leukocytes and sperm membranes L. Ermini a, F. Secciani a, G.B. La Sala b, L. Sabatini c, D. Fineschi c, G. Hale d, F. Rosati a,* a Department of Evolutionary Biology, University of Siena, Italy Unit of Human Reproduction, Ospedale S.Maria Nuova, Reggio Emilia, Italy Laboratory of Hematology and Coagulation, Azienda Ospedaliera Universitaria Senese, Via Tufi 3, Siena, Italy d Sir William Dunn School of Pathology, University of Oxford, Oxford, UK b
c
Received 12 October 2005 Available online 24 October 2005
Abstract CD52 is a human GPI-anchored antigen, expressed exclusively in the immune system and part of the reproductive system (epididymal cells). Sperm cells acquire the antigen from the epididymal secretions when transiting in the epididymal corpus and cauda. The peptide backbone of CD52, consisting of only 12 aminoacids, is generally considered no more than a scaffold for post-translational modifications, such as GPI-anchor and especially N-glycosylation which occur at the third asparagine. The latter modification is highly heterogeneous, especially in the reproductive system, giving rise to many different glycoforms, some of which are tissue specific. A peculiar O-glycan-containing glycoform is also found in reproductive and immune systems. We determined to locate CD52 in microdomains of leukocytes and sperm membranes using two antibodies: (1) CAMPATH-1G, the epitope of which includes the last three aminoacids and part of the GPI-anchor of glycoforms present in leukocytes and sperm cells; (2) antigp20, the epitope of which belongs to the unique O-glycan-bearing glycoform also present in both cell types. Using a Brij 98 solubilization protocol and sucrose gradient partition we demonstrated that the CD52 glycoforms recognized by both antibodies are markers of typical raft microdomains in leukocytes, whereas in capacitated sperm the O-glycoform is included in GM3-rich microdomains different from the cholesterol and GM1-rich lipid rafts with which CAMPATH antigen is stably associated. The importance of the association between GM3 and O-glycans for formation of specialized microdomains was confirmed by heterologous CD52 insertion experiments. When prostasomes from human seminal fluid were incubated with rat sperm from different epididymal regions, the CD52 glycoform recognized by anti-gp20 decorated rat epididymal corpus and cauda sperm, associated with the same low-cholesterol GM3-rich sperm membrane fractions as in human sperm. The glycoforms recognized by CAMPATH-1G were not found in rat sperm. The relationship between this differential insertion and differences in glycosylation of rat and human CD52 is discussed. 2005 Elsevier Inc. All rights reserved. Keywords: CD52 antigen; Leukocyte; Sperm; GPI-anchored proteins; Membrane lipid rafts
The concept that specific lipid and lipid-derivatives are involved in determining membrane-ordered microdomains moving within the fluid glycerophospholipid matrix is gaining evidence. These microdomains, characterized by their *
Corresponding author. Fax: +39 577 234476. E-mail address: [email protected] (F. Rosati).
0006-291X/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.10.082
order and buoyant density, are resistant to solubilization by non-ionic detergents and are generally isolated by sucrose-gradient centrifuging. Different major functions have been proposed for them, including membrane trafficking and signalling [1,2]. In the immune system, these microdomains have been termed immunosynapses [3]. Attention has recently also been paid to the adhesion role of glycosyl
1276
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
epitopes carried by proteins or lipids of glycosphingolipidrich microdomains and the term glycosynapses has been coined for them [4–6]. GPI(glycosylphosphatidylinositol)anchored proteins are generally considered preferential inhabitants of membrane microdomains, although the GPI-linkage does not seem sufficient to determine this domain restriction [7]. CD52 is a human GPI-anchored antigen, product of a single gene on chromosome 1 [8]. Its functional role is still largely unknown and its association with microdomains has not been determined. CD52 is expressed exclusively in the immune system (lymphocytes, monocytes, and macrophages) and part of the reproductive system (epididymal cells). Sperm cells also bear the antigen but do not express it. This is because epididymal cells of the corpus and cauda districts secrete it into the surrounding fluid from which it is taken up by sperm in transit [9]. The peptide backbone of CD52 is short, consisting of only 12 aminoacids [10]. Indeed, it is generally considered no more than a scaffold for post-translational modifications such as GPI-anchorage and especially N-glycosylation which occur at the asparagine in third position [10]. The latter modification is highly heterogeneous, especially in the reproductive system [11], giving rise to many different glycoforms, some of which are tissue specific [12,13]. In the past and also recently, O-glycans have been suggested to be part of certain CD52 glycoforms [14–16]. The first and most commonly used anti-CD52 monoclonal antibody, termed CAMPATH-1, was generated in the immune system [17] and recognizes CD52 of the immune and reproductive systems [18]. Purification of CD52 from spleen and seminal plasma for structural characterization of N-glycans and GPI-anchor, and for peptide sequencing has always been done using this antibody [19,10]. The CAMPATH epitope includes the last three aminoacids and the first part of GPI [20]. Other anti-CD52 antibodies such as S-19 [21] and Mab H6-3C4 [13], generated in the reproductive system and directed against the N-linked chains, recognize CD52 glycoform subpopulations specific to the reproductive system [22]. In this panorama, antigp20, a polyclonal antibody that we raised against a purified sialyl-glycoprotein of the human sperm surface, homologous to CD52 [23], is peculiar. Indeed, it interacts strongly with a specific glycoform subpopulation of the CAMPATH antigen present in the immune and reproductive systems [24], a characteristic of which is to bear O-glycans not found in other glycoforms [16]. The antigp20 epitope includes these glycans [16]. Intriguingly, the O-glycoform is prevalently localized in the equatorial region of the head of capacitated sperm, whereas forms recognized by other anti-CD52 antibodies are equally distributed over the whole cell membrane [21,23]. A role in fertility has been suggested for this glycoform [24,25]. The CAMPATH antigen is highly expressed in leukocytes. The humanized form of CAMPATH antibodies has many therapeutic applications where depletion of lymphocytes is required, for example in leukemia, lymphoma,
bone marrow transplant, rheumatoid arthritis, and multiple sclerosis [26–30]. To determine whether high glycosylation heterogeneity and differential insertion in membranes influence the membrane microdomain association of CD52, we investigated the location of CD52 glycoforms recognized by CAMPATH and anti-gp20 with respect to detergent-insoluble glycosphingolipid/cholesterol-enriched microdomains in leukocytes and sperm. The possibility that microdomains are involved in cell-to-cell transfer of CD52 was also investigated by analyzing epididymal rat sperm after fusion with human prostasomes. Co-localization of the antigen glycoforms recognized by CAMPATH and gp20 antigen was also performed in leukocytes and sperm. Materials and methods Antibodies. The following antibodies were used: mouse anti-Lck (Sigma–Aldrich), anti-rabbit IgG conjugated with alkaline phosphatase (BioRad Microscience, Cambridge, MA), anti-mouse IgG conjugated with alkaline phosphatase (Sigma–Aldrich), anti-rat IgG conjugated with alkaline phosphatase (Sigma–Aldrich), streptavidin HRP-conjugated (Amersham Biosciences), biotin-conjugated cholera toxin B subunit (Sigma–Aldrich), anti-rat IgG conjugated with rhodamine (Sigma–Aldrich), fluorescein-conjugated anti-rabbit IgG (Calbiochem, San Diego, CA), CAMPATH-1G (rat IgG2b, clone YTH34.5G2b, Delta Biological S.r.l Rome, Italy, or a gift from G. Hale). Polyclonal anti-gp20 was obtained according to Focarelli et al. [23]. Anti-gp20 IgG were purified by affinity chromatography according to Flori et al. [16] using an AKTA liquid chromatography system. Cell and vesicle preparation procedures. All preparations were performed at 4 C unless otherwise stated. Human white blood cells were obtained from blood samples of healthy donors. Five millilitres aliquots of whole blood were mixed with 1.5 ml of 6% dextran solution in saline (NaCl 0.9%). After sedimentation for 2 h at 37 C, the leukocyte enriched supernatant was collected, washed, and centrifuged at 500g for 10 min Contaminating erythrocytes in the sediment were lysed with 1 ml chilled 0.8% NH4Cl solution [31]. Human ejaculate was obtained from a healthy donor and allowed to liquefy at room temperature for 30 min. Spermatozoa were separated from seminal plasma by centrifugation at 500g for 15 min and then washed twice in phosphate-buffered saline (PBS: 150 mM NaCl, 50 mM KH2PO4, pH 7.4). Spermatozoa were selected by a swim-up procedure. Briefly, 0.5 ml of BWW (Biggers–Whitten–Whittingham) medium [32] containing 35 mg/ml human serum albumin (HSA) was layered over aliquots of spermatozoa and incubated for 1 h at 37 C in 5% CO2 and 95% air. The upper layer, rich in mobile spermatozoa was collected and capacitation followed for 5 h at the same conditions. Seminal plasma, obtained as described previously, was centrifuged at 10,000g for 20 min to eliminate cell debris and residual sperm. The new supernatant was centrifuged at 105,000g for 120 min. The supernatant was discarded, and the pellet containing prostasomes and amorphous material was suspended in PBS containing 0.5 M KCl and centrifuged again at 105,000g for 120 min to remove the amorphous material. Cells and prostasomes were resuspended in 1 ml PBS for immunofluorescence analysis or in buffer A (2 mM Tris–HCl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 1 lg/ml leupeptin, 1 lg/ml pepstatin, 2 lg/ml chymostatin, and 5 lg/ml a2 macroglobulin) for isolation of detergent resistant membranes. Fluorescence microscopy. Leukocytes were washed three times with PBS and fixed for 10 min with 4% paraformaldehyde in PBS at 4 C. Cells (3 · 106) were then smeared onto a glass slide and after blocking the unspecific sites with PBS containing 2% BSA (PBS–BSA) and incubated with CAMPATH-1G and with anti-gp20 IgG diluted 1:25 in the same buffer for up to 30 min at room temperature. After three washings, leu-
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283 kocytes were incubated with respective fluorescein or rhodamine-conjugated secondary antibodies. Aliquots of sperm (3 · 106) were smeared onto ethanol-cleaned glass slides, and allowed to attach at room temperature, taking care to keep them in liquid phase to avoid damaging the sperm membranes. After several washings in PBS, the smears were blocked for 30 min with 2% BSA in PBS and then incubated for 1 h in the same buffer containing IgG anti-gp20 and CAMPATH-1G diluted 1:25. After several washings, the smears were incubated for 1 h with goat antirabbit IgG and goat anti-rat IgG conjugated, respectively, with fluorescein and rhodamine. Slides were then washed, mounted in 20 mM Tris–HCl, pH 8.0, 80% glycerol, and 4% (w/v) N-propyl gallate as anti-fade, and observed with a laser scanning confocal apparatus (TCS 4D, Leica, Heidelberg, Germany). The images were processed in Photoshop 5.5 (Adobe Systems, Mountain View, CA). Detergent extraction and sucrose gradient centrifugation. Cells (1 · 108 leukocyte and 1.5 · 108 spermatozoa) and prostasome suspension were homogenized using a motor-driven Potter–Elvehjem homogenizer (Kontes, Vineland, NJ) with a tight-fitting pestle. The homogenate was then centrifuged at 800g for 10 min, the post-nuclear supernatant (PNS) was pre-incubated for 4 min at 4 C. Brij 98 (Sigma–Aldrich) was then added to a final concentration of 1%. After 5 min of solubilization at 37 C, the PNS (1 ml) was diluted with 2 ml 37 C pre-warmed buffer A containing 2 M sucrose (final sucrose concentration 1.33 M; final Brij 98 concentration 0.33%) and chilled down on ice (55 min) before being placed at the bottom of a step sucrose gradient (0.9–0.8–0.7–0.6–0.5–0.4–0.3–0.2 M sucrose, 450 ll each) in buffer A. Gradients were centrifuged at 50,000 rpm for 16 h in a SW65 rotor (Beckman Instruments Inc.) at 4 C [33]. Five hundred microliter fractions were harvested from the top. The first one was discarded and the others, unless specified, pooled as follows: lowdensity fractions (2–4), intermediate-density fractions (5–7), and heavydensity fractions (8–10) were pooled. Low- and intermediate-density fractions were further centrifuged at 50,000 rpm in a SW65 rotor for 16 h and only the pellet was recovered and analyzed. For cholesterol extraction, leukocytes and sperm were pre-treated with 250 ll of 0.2% saponin in PBS at 4 C for 10 min with occasional mixing. The sample was then subjected to the detergent extraction procedure. Fusion of rat spermatozoa with human prostasomes. Rat sperm, collected in the three epididymal districts (caput, corpus, and cauda), were resuspended in Hepes 2 mM, pH 8.0, containing 0.32 M sucrose and BSA (1%, w/v) to have about 6 · 107 cells/ml 1. When required, prostasomes were added to reach a prostasome to sperm protein ratio of 2:1. These mixtures were incubated at 37 C for 60 min. Sucrose Hepes buffer (0.32 M sucrose), pH 8.0 (1 ml), was then added and sperm were collected by centrifugation for 10 min at 600g. The washing procedure was repeated twice using Hepes 2 mM, pH 8.0, containing 0.6 M KCl [34]. The pellet of fused sperm was resuspended in buffer A for sucrose density gradient analysis. Lipid analysis. Analysis of ganglioside GM1 was performed by directly testing blotted fractions with CTB. Aliquots corresponding to 2 lg protein of each fraction were spotted onto nitrocellulose filters; the membranes were subsequently blocked with TBS (20 mM Tris–HCl, 500 mM NaCl, pH 7.5) containing 3% non-fat powdered milk and then incubated with biotin-conjugated CTB diluted 1:1000 for 1 h at room temperature. After washing, nitrocellulose filters were incubated with HRP-conjugated streptavidin under the same conditions. After extensive rinsing in TBS– Tween 0.2%, the labelled components were developed using Immun-Star HRP Chemiluminescent Detection Kit (BioRad Microscience, Cambridge, MA) following the manufacturerÕs instructions. Analysis of cholesterol and ganglioside GM3 was performed by thin layer chromatography (TLC) after lipid extraction from fraction membranes. Aliquots (50 lg) of protein fractions, obtained by sucrose gradient centrifugation, were lipid extracted with chloroform:methanol (1:2, by volume) as described by Bligh and Dyer [35]. Lipids contained in the chloroform phase were dried under N2 stream and dissolved again in chloroform. Thin layer chromatography (TLC) was carried on high-performance TLC (HPTLC) plates (Silica Gel 60 F-254, Merck) with chloroform:methanol (98:2, v/v). For ganglioside analysis a 2:1:1 chloroform:methanol:water solution was added to the chloroform phase and
1277
centrifuged at 800g for 10 min. The resulting biphasic solution consisted of an upper phase containing polar lipids and a lower phase containing nonpolar lipids. The upper phase was removed, dried under N2 stream, and resuspended in a chloroform:methanol:water solution (15:30:4, v/v/v). The samples were then separated by TLC using a mixture of chloroform:methanol:water (60:40:9, v/v/v). Lipids were visualized using 20% (v/ v) H2SO4 at 150 C or resorcinol reagent [36]. Cholesterol and ganglioside GM3 were quantified by densitometry using commercial cholesterol and GM3 (Sigma–Aldrich) as standard. Electrophoresis and Western blotting. Proteins were concentrated as described by Wessel and Flugge [37]. Samples were separated by Tricine– sodium dodecyl sulfate–polyacrylamide gel electrophoresis [38]. The separation gel consisted of 10% acrylamide with a 3% stacking gel. In other experiments, proteins were separated on 10% polyacrylamide slab gels according to Laemmli [39]. The gels were transferred to nitrocellulose as described [40]. Nitrocellulose sheets were then blocked with 3% non-fat powdered milk in TBS and incubated overnight at 4 C with anti-gp20 IgG (1:100), anti-Lck diluted 1:1000 or CAMPATH-1G diluted 1:100. After several washings with TBS–Tween, the blots were incubated for 1 h at room temperature with the respective secondary antibodies conjugated with alkaline phosphatase. After extensive rinsing in TBS–Tween, labelled proteins were developed using Immun-star Chemiluminescent Detection Kit according to the manufacturerÕs instructions. Pre-immune serum was used as control. Protein concentration was determined using BCA protein assay reagent kit (Pierce Biotechnology, Inc.).
Results The CD52 O-glycoform co-localized with the other CD52 forms in leukocytes but not in capacitated sperm In a first series of experiments we investigated co-localization of the different CD52 glycoforms on leukocytes and capacitated spermatozoa using CAMPATH-1G and antigp20 simultaneously. Equivalent amounts (3 · 106) of cells were smeared on glass slides and incubated simultaneously with CAMPATH-1G and anti-gp20, and then with the corresponding anti-rat or anti-rabbit IgG conjugated with fluorescein or rhodamine. The results demonstrated co-localization of all CD52 glycoforms in leukocyte but confirmed their different localization in capacitated sperm. In fact, as shown in Fig. 1, the signals of the two antibodies overlapped completely in leukocytes whereas in capacitated sperm, overlapping only occurred in the equatorial region of the head and sometimes in the neck region; only the CAMPATH-1G signal was detected on the rest of the sperm surface. In leukocytes all CD52 glycoforms are associated to lipid raft microdomains Rafts have been generally defined by their insolubility in cold Triton X-100 [41] and this characteristic is widely used as a basis for isolating them [1,42]. The detergents Brij 98, 96 or 56 have recently replaced Triton since they are reported to preserve membrane rafts even above 37 C, thus excluding the objection that Triton X-100 used at 4 C induces lipid transitions, artificially creating detergent-resistant protein–lipid complexes [33]. We used Brij 98 for fractioning the leukocyte membranes. Cells were lysed in Brij 98 at 37 C, the post-nuclear supernatant
1278
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
Fig. 1. In peripheral blood leukocytes, the CD52 glycoforms recognized by CAMPATH-1G and anti-gp20 co-localized, whereas in swim-up selected and 5 h capacitated sperm cells the O-glycoform recognized by anti-gp20 localized in the equatorial region of the head and in the first part of the tail. Leukocytes were fixed for 10 min with 4% paraformaldehyde at 4 C and then smeared on glass slides. Sperm were smeared directly on glass slides and allowed to attach at room temperature, taking care to keep them in liquid phase to avoid damaging the sperm membranes. Both cell types were then incubated for 1 h simultaneously with anti-gp20 and CAMPATH-1G followed by goat anti-rabbit IgG and goat anti-rat IgG conjugated, respectively, with fluorescein and rhodamine.
(PNS) was ultracentrifuged in a sucrose density gradient, and 10 fractions were recovered. The first seven fractions were always ultracentrifuged and only the pelletted material was used for the following preparations. This was done for avoiding the presence of soluble proteins in these fractions. Each fraction was analyzed for the presence of cholesterol, GM1, and GM3. GM1 is, in fact, considered in general a marker of lipid rafts and GM3 has been described to reside in these domains in leukocytes [43]. All of them were found to be significantly prevalent in the lowest density fractions 2, 3, and 4, slightly present in the intermediate fractions 5, 6, and 7, and absent in the last three heavy-density ones. For all this, we pooled the three groups of fractions as indicated in Fig. 2 and considered only the fractions 2–4 as typically containing the detergent insoluble lipid rafts. Fractions 2–4 corresponded to 10–20% sucrose density, fractions 5–7 corresponded to 20–30% sucrose density and fractions 8–10 corresponded to 30–40% sucrose density. Protein quantification of the three pooled fractions revealed that the first fractions contained 5% of the total PNS protein, the intermediates 20%, and the last ones the rest. In the following experiments, the pooled fractions were always analyzed under normalized conditions, i.e., using the same protein concentration. Equivalent amounts of protein of the three fractions groups were then run in SDS–PAGE, blotted, and analyzed with CAMPATH-1G, anti-gp20, and an anti-Lck, an antigen known to reside in the lipid rafts [44]. As shown in Fig. 2, the CD52 forms recognized by both the anti-CD52 antibodies were included in the lipid raft domains along with Lck since they prevalently floated in the fractions 2–4. However, unlike CAMPATH, anti-gp20 also revealed the corresponding antigen form in fractions 5–7 (Fig. 2). To confirm the association of CD52 to raft domains we also evaluated the effect of saponin on CD52 localization.
Fig. 2. In leukocytes, all the CD52 glycoforms recognized by CAMPATH-1G and anti-gp20 prevalently partitioned in the low-density sucrose fractions with cholesterol, GM1, GM3, and Lck. After removal of cholesterol by saponin, they shifted to high-density fractions. Postnuclear supernatant of homogenized cells was treated with Brij 98 at 37 C in the presence or absence of 0.2% saponin and then sucrose-gradient fractionated. The pooled fractions, normalized for protein content, were then tested for Lck as well as anti-gp20 and CAMPATH-1G CD52 glycoforms by immunoblot. Fractions were also tested for ganglioside GM1 by direct dot blot analysis with CTB, and for cholesterol and ganglioside GM3 by thin layer chromatography (TLC) after lipid extraction from membrane fractions.
Saponin is known for its ability to disorganize plasma membrane and to promote a shift in protein localization [33]. PNS was dissolved with Brij 98 at 37 C with or without 0.2% saponin and then fractionated on a sucrose gradient. The fractions were then tested for the presence of Lck and CD52. As shown in Fig. 2, after this treatment all the antigen forms were prevalently recovered in the detergent-soluble fractions of cell membranes, as expected. However, part of the antigen recognized by antigp20 was still detectable in the intermediate fractions (Fig. 2). In capacitated human sperm, the O-glycoform is inserted in GM3-rich domains whereas the other CD52 forms are associated to cholesterol/GM1-rich raft microdomains Sperm plasma membranes were then investigated for CD52 localization using the same protocol as for leukocytes. In a first series of experiments, we analyzed swim up selected and 5 h capacitated sperm since they are more homogeneous and bear a specific localization of the glycoform recognized by anti-gp20 [23]. 1.5 · 108 sperm were homogenized, subjected to the detergent extraction procedure and ultracentrifuged in a sucrose density gradient. The 10 fractions recovered were harvested from the top and analyzed for cholesterol content. The fractions were then pooled as described for leuko-
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
1279
cytes after confirmation that cholesterol was maximal in the low-density fractions 2, 3, and 4 (Fig. 3). Analysis of the three blotted fractions groups for the presence of GM1 and GM3 confirmed the association of GM1 with the same cholesterol-rich fractions but also revealed that GM3 was instead prevalently associated with the intermediate 5–7 fractions. Equivalent amounts of the three fraction groups were then separated on Tricine–sodium dodecyl sulfate one-dimensional 10% polyacrylamide gel, blotted, and analyzed using CAMPATH-1G and anti-gp20. The positivity pattern was very different from that found in leukocytes. In fact, only the first antibody produced a broad positive signal around 20 kDa in the first fraction group, whereas anti-gp20 gave a weak signal in the same fractions and strong reactivity in the intermediate GM3-rich fractions (Fig. 3). To be sure of these results, we analyzed all the sucrose gradient fractions for the presence of CAMPATH and gp20 antigen forms and for GM1 and GM3. The results showed that the gp20 glycoform was found prevalent in fraction 5 and to a lesser extent in fraction 6 with GM3, whereas the CAMPATH antigen forms were prevalent in fractions 3–4 with GM1 (data not shown). This different behavior of the CD52 glycoforms was confirmed in more than 20 independent experiments by analyzing the same fractions with the two antibodies sequentially, stripping the first antibody. Furthermore, in order to analyze whether this behavior was specific for Brij solubiliza-
tion we treated the same samples with Triton X-100 at 4 C and analyzed the sucrose gradient fractions in the same way. Again CAMPATH antigen was mainly recovered in the first three fractions and the gp20 form in intermediate fractions. Saponin treatment of sucrose-separated sperm plasma membranes also produced a different pattern with respect to leukocytes. Neither the anti-gp20 glycoform nor the others recognized by CAMPATH shifted in the heavy fractions as they did in leukocytes but all were recovered in the same fractions as before treatment (Fig. 3). Since changes in lipid composition and distribution in the plasma membrane are major capacitation events [45], we investigated whether the different localization of CD52 glycoforms was a consequence of the capacitation process and depended on loss of cholesterol. Equal amounts of washed ejaculated sperm were treated in the same way as capacitated sperm, and lipids and proteins in the sucrose density gradient fractions were investigated in the presence or absence of saponin. As shown in Fig. 4, before capacitation and without saponin, cholesterol, and GM1 were associated with the low- and intermediate-density fractions in almost equivalent amounts and GM3 was prevalent in the intermediate fractions with all the CD52 glycoforms. After removal of cholesterol by saponin, the association pattern of lipids (data not shown) and CD52 (Fig. 4) turned out to be identical to that of the capacitated sperm membranes.
Fig. 3. In swim-up-selected and 5 h capacitated human sperm, the CD52 CAMPATH glycoforms prevalently partitioned in the low-density sucrose fractions with GM1, whereas the anti-gp20 form was associated with GM3 in the intermediate-density sucrose fractions. The different forms did not shift position after saponin treatment. Post-nuclear supernatant of the cells was treated with Brij 98 at 37 C in the presence or absence of 0.2% saponin and then sucrose-gradient fractionated. The fractions, normalized for protein content, were then tested for anti-gp20 and CAMPATH-1G CD52 glycoforms by immunoblot. The same fractions were also tested for the presence of the ganglioside GM1 by direct dot blot analysis with CTB, and for cholesterol and ganglioside GM3 by thin layer chromatography (TLC) after lipid extraction from membrane fractions.
Fig. 4. In freshly ejaculated sperm, the membrane was not yet organized like capacitated sperm membranes, except for GM3 and the glycoform recognized by anti-gp20 that were already in the intermediate fractions. By contrast, cholesterol and GM1 were equally distributed in more than one fraction and CAMPATH antigen forms were prevalent in the intermediate fractions. Incubation of sperm with saponin restored the partitioning properties of capacitated sperm. Solubilization and fractionation of these cells were done as in capacitated sperm.
1280
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
Only the CD52 O-glycoform of prostasomes was efficiently inserted in heterologous epididymal sperm, associated with low-cholesterol/GM3 rich microdomains The mechanism by which CD52 is stably inserted in the plasma membrane of sperm transiting in the epididymis is still largely unknown; however, transfer from large aggregates of the antigen or from vesicles containing it (termed prostasomes) in epididymal fluid has often been suggested [46,47]. Transfer capacity of the different forms of CD52 and the role played by microdomains in this transfer were investigated under heterologous conditions, i.e., using rat epididymal sperm and human prostasomes. To do this equal amounts of rat sperm (6 · 107) obtained from the caput, corpus, and cauda of epididymis were challenged with a constant amount of human prostasomes and the rat sperm was then subjected to the same solubilization protocol used for human sperm. In a first series of experiments, we investigated the presence of the antigen in an equivalent Brij-insoluble membrane fraction of sperm recovered from the three epididymal districts. The results showed that the antigen was strongly present in sperm from the cauda, faintly detectable in the sperm from the corpus, and absent in sperm from the caput (Fig. 5A). The Brij solubilized rat cauda sperm membranes were then sucrose density gradient, centrifuged, and the fractions were analyzed for lipid and protein content. The results revealed a distribution pattern of cholesterol, GM1, GM3, and CD52 O-glycoform identical to that of capacitated human sperm. Cholesterol and GM1 were recovered in the low-density fractions, whereas GM3 and the main portion of the O-glycan CD52 form were detected in fractions 5–7. A lower amount of this form was also found in the detergent insoluble fractions 2–4 and the detergent soluble fractions 8–10 (Fig. 5C). Immunofluorescence analysis revealed that the heterologous CD52 was prevalently inserted in the membrane overlaying the acrosome and in the first part of the tail (Fig. 5B). Surprisingly, CAMPATH antigen was not detectable in any fraction after density gradient sucrose fractionation nor in the sperm membrane after immunofluorescence analysis. To be sure that incubation of prostasomes with sperm did not alter the CAMPATH epitope, prostasomes were recovered after incubation and proteins were tested in immunoblot with CAMPATH-1G and anti-gp20. All CD52 glycoforms occurred in prostasomes but the signal intensity of the same amount of prostasomes before and after incubation with the two antibodies was unchanged for CAMPATH-1G and much lower for anti-gp20, thus confirming its insertion in rat sperm. Discussion We looked for CD52 in microdomains of leukocytes and sperm membranes using two antibodies: (1) CAMPATH-1G, the epitope of which includes the last three
Fig. 5. Human CD52 O-glycan-containing glycoform, but not CAMPATH glycoforms, may be transferred to membranes of rat sperm obtained from the epididymis. (A) Transfer occurred in sperm from the epididymal corpus and cauda but not in those from the caput. (B) Transfer of human CD52 O-glycan-containing glycoform occurred in the membrane of the acrosomal region and in part of the tail of rat sperm (C). The human CD52 glycoform co-isolated with GM3 in intermediate-density sucrose fractions. Rat sperm, obtained from the three districts of the epididymis (caput, corpus, and cauda), were incubated with prostasomes for 60 min. They were then collected by centrifugation for 10 min at 600g and the sperm membranes were solubilized and fractionated like those of human sperm. For immunofluorescence analysis, rat cauda sperm were incubated with prostasomes and after centrifugation smeared on glass slides and incubated with anti-gp20 and CAMPATH-1G followed by the corresponding secondary antibodies.
aminoacids and part of the GPI-anchor of glycoforms present in leukocytes and sperm cells [20]; (2) anti-gp20, the epitope of which belongs to a unique O-glycan-bearing glycoform also present in both cell types [16]. Using a Brij 98 solubilization protocol and sucrose gradient partition we demonstrated that the CD52 glycoforms recognized by both antibodies in leukocytes are markers of typical raft microdomains, whereas in sperm the same glycoforms have differential insertion. In fact, most of the antigen recognized by CAMPATH-1G and anti-gp20 in leukocytes was recovered in the low-density, high-cholesterol, GM1rich fraction, along with Lck, known to reside constitutionally in raft microdomains of these cells [44,48]. By our method, this fraction included fractions 2, 3, and 4 at 10–20% sucrose density. GM3, also known to reside in raft microdomains of peripheral blood cells [43], was found to co-isolate with the CAMPATH and anti-gp20 glycoforms. Furthermore, like other typical raft-associated
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
proteins [49], most CD52 glycoforms shifted to the highdensity, detergent-soluble fraction when cholesterol was depleted by saponin. To comparatively analyze CD52 glycoforms in male germ cells, we first used swim-up-selected, 5-h capacitated sperm, due to the specific localization of the CD52 O-glycoform in these cells [23] and the need to use cells as homogeneous as possible. Interestingly, the O-glycoform was found to be differently associated with respect to most of the CAMPATH antigen. In fact, when sperm membranes were fractionated in the same way as those of leukocytes and the fractions pooled, most of the CAMPATH antigen was recovered in the same fractions (fractions 2–4) as in leukocytes but the O-glycan-containing glycoform was mainly recovered in fractions 5–7 at 20–30% sucrose density. The markers of typical raft microdomains, cholesterol, and GM1, were found associated with low-density fractions, where CAMPATH antigen was prevalent. Unlike in leukocytes, however, GM3 was associated with intermediate sucrose density fractions together with the O-glycoform. Many independent experiments with the same blotted samples and the two antibodies were done to be sure of this different partitioning property but the results were always the same. The finding was also confirmed using the cold Triton X-100 solubilization protocol. The distinct partitioning of CD52 glycoforms, cholesterol, GM1, and GM3 in the 20% and 30% sucrose density fractions thus indicates that the different density fractions contain separate microdomains. The high-cholesterol and GM1 content of fractions 2–4 indicated that in capacitated sperm, as in leukocytes, these fractions contain typical raft microdomains. The CAMPATH antigen is therefore also a marker of these domains in capacitated sperm. The question is now to understand the kind of microdomain in which CD52 O-glycoform is included. These microdomains contain low-cholesterol and GM1, and are characterized by a specific prevalence of GM3. Interestingly, an asymmetric distribution of GM1- and GM3-enriched domains has been described as being associated with the leading-edge and uropod of migrating lymphocytes [50] and GM3 has often been indicated as playing a key role in formation of membrane platforms termed glycosynapses involved in cell–cell interaction [6]. Glycosynapse characteristics also include independence from cholesterol and the cis-interacting role of GM3 with oligosaccharide chains of platform constituents [6]. This means that the CD52 O-glycoform and GM3 may together take part in regionalized specific microdomains involved in adhesion mechanisms. This would explain the restricted localization of the CD52 O-glycoform in the equatorial region of the head of capacitated sperm [23] and its suggested role of interacting with the oocyte membrane [23,25]. Since capacitation is associated with an efflux of cholesterol that leads to significant rearrangement of sperm membrane components [51], we also investigated whether the different partition properties of the CD52 glycoforms were a consequence of capacitation-dependent cholesterol
1281
loss. Indeed, during capacitation at least 60–70% of cholesterol is lost and a consequent alteration of lipid raft constituents in sperm membranes has been reported in boar [52] and in human sperm membranes [53,54]. The first difference we noted was that cholesterol and GM1 of freshly ejaculated sperm membranes were not restricted to the same low-density fractions as in capacitated sperm but were equally distributed in fractions 2–4 and 5–7. Instead, GM3 partitioned in the intermediate sucrose density fractions as in capacitated sperm membranes. Unlike in capacitated sperm, in freshly ejaculated sperm the glycoforms recognized by CAMPATH-1G were found prevalently associated with fractions 5–7 along with that recognized by anti-gp20. However, external removal of cholesterol by saponin gave rise to a lipid and CD52 glycoform partition pattern identical to that found in capacitated sperm. On the other hand, treatment of capacitated sperm with saponin did not alter the partition properties of the different CD52 glycoforms. Taken together, these results suggest that the membrane of freshly ejaculated sperm is particularly rich in cholesterol and is highly unstable, and that the progressive capacitation-dependent loss of cholesterol defines its final rigid organization ready for fertilization, when: (1) canonical lipid rafts with which CAMPATH antigen remains stably associated are formed; (2) GM3-rich microdomains containing CD52 O-glycoform segregate in specific regions. The importance of GM3 in forming microdomains in association with the O-glycan-containing glycoform was further confirmed by experiments of heterologous CD52 insertion experiments. Indeed, when prostasomes from human seminal fluid were incubated with rat sperm from different epididymal regions, the CD52 glycoform recognized by anti-gp20 decorated the rat sperm, associated with the same low-cholesterol, GM3-rich sperm membrane fractions as in human sperm. Only sperm from the corpus and cauda accepted the antigen. This is in line with previous reports suggesting that in rats [55,56], as well as in humans and monkeys [47], only corpus and cauda sperm have sufficient maturation status for antigen uptake. Surprisingly, the CAMPATH antigen was not transferred to the rat sperm membrane, despite its abundance in prostasomes. We cannot yet explain this different behavior, but it seems significant that unlike human CD52, rat CD52 does not appear to have N-linked oligosaccharide chains but rather a number of O-linked oligosaccharide chains [56,57]. On the other hand, the CD52 forms recognized by CAMPATH have already been reported to transfer better from large aggregates of the protein than from prostasomes [46]. We therefore deduce that the oligosaccharides bound to peptides could be determinant for their transfer from an external source to membrane microdomains. On the other hand, since neither the GPI-anchor of the human CD52 O-glycoform nor that of rat CD52 is known, these anchors may also play a role. Taken together, the present results indicate a difference in location of CD52 in
1282
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
membrane domains when inserted internally through the secretory pathway or when transferred from an external source, and are in line with our previous suggestions that the CD52 O-glycoform plays a role in fertilization [24,25]. Concerning CAMPATH antigen, indications that it is stably associated with raft microdomains are in line with the possibility that the antigen is involved in signaling processes in leukocytes and sperm. Finally the fact that the O-glycoform segregates differently from the other CD52 glycoforms in sperm rises the question of whether the same may occur in leukocytes under certain conditions. Research into this aspect is now underway study in our laboratory.
References [1] K. Simons, E. Ikonen, Functional rafts in cell membranes, Nature 387 (1997) 569–572. [2] S.K. Pierce, Lipid rafts and B-cell activation, Nat. Rev. Immunol. 2 (2002) 96–105. [3] S.K. Bromley, W.R. Burack, K.G. Johnson, K. Somersalo, T.N. Sims, C. Sumen, M.M. Davis, A.S. Shaw, P.M. Allen, M. Dustin, The immunological synapse, Annu. Rev. Immunol. 19 (2001) 375–396. [4] K. Iwabuchi, K. Handa, S. Hakomori, Separation of ‘‘glycosphingolipid signaling domain’’ from caveolin-containing membrane fraction in mouse melanoma B16 cells and its role in cell adhesion coupled with signaling, J. Biol. Chem. 273 (1998) 33766–33773. [5] S. Hakomori, Traveling for the glycosphingolipid path, Glycoconj. J. 17 (2000) 627–647. [6] S. Hakomori, Glycosynapses: microdomains controlling carbohydrate-dependent cell adhesion and signaling, An. Acad. Bras. Cienc. 76 (2004) 553–572. [7] D.F. Legler, M.A. Doucey, P. Schneider, L. Chapatte, F.C. Bender, C. Bron, Differential insertion of GPI-anchored GFPs into lipid rafts of live cells, FASEB J. 19 (2005) 73–75. [8] M. Tone, K.F. Nolan, L.A. Walsh, Structure and chromosomal location of mouse and human CD52 genes, Biochim. Biophys. Acta 1446 (1999) 334–340. [9] C. Kirchhoff, G. Hale, Cell-to-cell transfer of glycosylphosphatidylinositol-anchored membrane proteins during sperm maturation, Mol. Hum. Reprod. 2 (1996) 177–184. [10] A. Treumann, M.R. Lifely, P. Schneider, M.A.J. Ferguson, Primary structure of CD52, J. Biol. Chem. 270 (1995) 6088–6099. [11] S. Schroter, P. Derr, H.S. Conrad, M. Nimtz, G. Hale, C. Kirchhoff, Male specific modification of human CD52, J. Biol. Chem. 274 (1999) 29862–29873. [12] A.B. Diekman, E.J. Norton, V.A. Westbrook, K.L. Klotz, S. NaabyHansen, J.C. Herr, Anti-sperm antibodies from infertile patients and their cognate sperm antigens: a review. Identity between SAGA-1, the H6-3C4 antigen, and CD52, Am. J. Reprod. Immunol. 4 (2000) 134–143. [13] A. Hasegawa, Y. Fu, H. Tsubamoto, Y. Tsuji, H. Sawai, S. Komori, K. Koyama, Epitope analysis for human sperm-immobilizing monoclonal antibodies, MAb H6-3C4, 1G12 and campath-1, Mol. Hum. Reprod. 9 (2003) 337–343. [14] H. Valentin, C. Gelin, L. Coulombel, D. Zoccola, J. Morizet, A. Bernard, The distribution of the Cdw52 molecule on blood cells and characterization of its involvement in T cell activation, Transplantation 54 (1992) 97–104. [15] A. Hasegawa, K. Koyama, Antigenic epitope for sperm-immobilizing antibody detected in infertile women, J. Reprod. Immunol. (2005) (Epub ahead of print). [16] F. Flori, C. Della Giovampaola, R. Focarelli, F. Secciani, G.B. La Sala, A. Nicoli, G. Hale, F. Rosati, Epitope analysis of immunoglobulins against gp20, a GPI-anchored protein of the human sperm
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30] [31]
[32]
[33]
surface homologous to leukocyte antigen CD52, Tissue Antigens 66 (2005) 209–216. G. Hale, M.Q. Xia, H.P. Tighe, M.J.S. Dyer, H. Waldmann, The CAMPATH-1 antigen (CDw52), Tissue Antigens 35 (1990) 118–127. G. Hale, P.D. Rye, A. Warfor, I. Lauder, A. Brito-Babapulle, The glycosylphosphatidylinositol-anchored CDw52 is associated with the epididymal maturation of human spermatozoa, J. Reprod. Immunol. 23 (1993) 189–205. M.Q. Xia, G. Hale, M.R. Lifely, M.J. Ferguson, D. Campbell, L. Packman, H. Waldmann, Structure of the CAMPATH-1 antigen, a GPI-anchored glycoprotein which is an exceptionally good target for complement lysis, Biochem. J. 29 (1993) 633–640. G. Hale, Synthetic peptide mimotope of the CAMPATH-1 (CD52) antigen, a small glycosylphosphatidylinositol-anchored glycoprotein, Immunotechnology 1 (1995) 175–187. A.B. Diekman, E.J. Norton, K.L. Klotz, V.A. Westbrook, H. Shibahara, S. Naaby-Hansen, C.J. Flickinger, J.C. Herr, N-linked glycan of a sperm CD52 glycoform associated with human infertility, FASEB J. 13 (1999) 1303–1313. E.J. Norton, A.B. Diekman, V.A. Westbrook, D.W. Mullins, K.L. Klotz, L.L. Gilmer, T.S. Thomas, D.C. Wright, J. Brisker, V.H. Engelhard, C.J. Flickinger, J.C. Herr, A male genital tract-specific carbohydrate epitope on human CD52: implications for immunocontraception, Tissue Antigens 60 (2002) 354–364. R. Focarelli, A. Giuffrida, S. Capparelli, M. Scibona, F. Menchini Fabris, F. Francavilla, S. Francavilla, C. Della Giovampaola, F. Rosati, Specific localization in the equatorial region of gp20, a 20 kDa sialylglycoprotein of the capacitated human spermatozoon acquired during epididymal transit which is necessary to penetrate zona-free hamster eggs, Mol. Hum. Reprod. 4 (1998) 119–125. R. Focarelli, S. Francavilla, F. Francavilla, C. Della Giovampaola, A. Santucci, F. Rosati, A sialoglycoprotein, gp20, of the human capacitated sperm surface is a homologue of the leukocyte CD52 antigen: analysis of the effect of anti-CD52 monoclonal antibody (CAMPATH-1) on capacitated spermatozoa, Mol. Hum. Reprod. 5 (1999) 46–51. V. Giuliani, C. Pandolfi, R. Santucci, F. Pelliccione, B. Macerola, R. Focarelli, F. Rosati, C. Della Giovampaola, F. Francavilla, S. Francavilla, Expression of gp20, a human sperm antigen of epididymal origin, is reduced in spermatozoa from subfertile men, Mol. Reprod. Dev. 69 (2004) 235–240. C.M. Lockwood, S. Thiru, J.D. Isaacs, G. Hale, H. Waldmann, Long-term remission of intractable systemic vasculitis with monoclonal antibody therapy, Lancet 341 (1993) 1620–1622. T. Moreau, J. Thorpe, D. Miller, I. Moseley, G. Hale, H. Waldmann, D. Clayton, M. Wing, N. Scolding, A. Compston, Preliminary evidence from magnetic resonance imaging for reduction in disease activity after lymphocyte depletion in multiple sclerosis, Lancet 344 (1994) 298–301. A. Osterborg, A.S. Fassas, A. Anagnostopoulos, M.J. Dyer, D. Catovsky, H. Mellstedt, Humanized CD52 monoclonal antibody Campath-1H as first-line treatment in chronic lymphocytic leukaemia, Br. J. Haematol. 93 (1996) 151–153. J.D. Isaacs, V.K. Manna, N. Rapson, K.J. Bulpitt, B.L. Hazleman, E.L. Matteson, E.W. St Clair, T.J. Schnitzer, J.M. Johnston, CAMPATH-1H in rheumatoid arthritis: an intravenous dose-ranging study, Br. J. Rheumatol. 35 (1996) 231–240. P. Rebello, G. Hale, Pharmacokinetics of CAMPATH-1H: assay development and validation, J. Immunol. Methods 260 (2002) 285–302. A. Garcia, J. Niubo, M.A. Benitez, M. Viqueira, J.L. Perez, Comparison of two leukocyte extraction methods for cytomegalovirus antigenemia assay, J. Clin. Microbiol. 34 (1996) 182–184. J.D. Biggers, W.K. Whitten, D.G. Whittingham, The culture of mouse embryos in vitro, in: J.C. Daniel (Ed.), Methods in Mammalian Embriology, Freeman, S.Francisco, 1971, pp. 86–116. P. Drevot, C. Langlet, X.J. Guo, A.M. Bernard, O. Colard, J.P. Chauvin, R. Lasserre, H.T. He, TCR signal initiation machinery is
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
[34]
[35] [36] [37]
[38]
[39] [40]
[41]
[42] [43]
[44]
[45]
[46]
pre-assembled and activated in a subset of membrane rafts, EMBO J. 21 (2002) 1899–1908. C.A. Palmerini, C. Saccardi, E. Carlini, R. Fabiani, G. Arienti, Fusion of prostasomes to human spermatozoa stimulates the acrosome reaction, Fertil. Steril. 80 (2003) 1181–1184. E.G Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Med. Sci. 37 (1959) 911–917. T. Miettinen, I.T. Takki-Luukkainen, Use of butyl acetate in determination of sialic acid, Acta Chem. Scand. 13 (1959) 856–858. D. Wessel, U.I. Flugge, A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids, Anal. Biochem. 138 (1984) 141–143. H. Schagger, G. von Jagow, Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa, Anal. Biochem. 166 (1987) 368–379. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of the bacteriophage T4, Nature 227 (1970) 680–685. H. Towbin, T. Staehelin, J. Gordon, Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc. Natl. Acad. Sci. USA 76 (1979) 4350–4354. D.A. Brown, J.K. Rose, Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface, Cell 68 (1992) 533–544. D.A. Brown, E. London, Functions of lipid rafts in biological membranes, Annu. Rev. Cell Dev. Biol. 14 (1998) 111–136. M. Sorice, A. Longo, T. Garofalo, V. Mattei, R. Misasi, A. Pavan, Role of GM3-enriched microdomains in signal transduction regulation in T lymphocytes, Glycoconjugate J. 20 (2004) 63–70. B. Kovacs, M.V. Maus, J.L. Riley, G.S. Derimanov, G.A. Koretzky, C.H. June, T.H. Finkel, Human CD8+ T cells do not require the polarization of lipid rafts for activation and proliferation, Proc. Natl. Acad. Sci. USA 99 (2002) 15006–15011. F.M. Flesch, J.F.H.M. Brouwers, P.E.F.M. Nivelstein, A.J. Verkleij, L.M.G. van Golde, B. Colenbrander, B.M. Gadella, Bicarbonate stimulated phospholipid scrambling induces cholesterol redistribution and anables cholesterol depletion in the sperm plasma membrane, J. Cell Sci. 114 (2001) 3543–3555. I.A. Rooney, J.E. Heuser, J.P. Atkinson, GPI-anchored complement regulatory proteins in seminal plasma, J. Clin. Invest. 97 (1996) 1675– 1686.
1283
[47] C.H. Yeung, A. Schroter, C. Wagenfeld, C. Kirchhoff, S. Kliesch, D. Poser, G.F. Wainaner, E. Nieschlag, T.G. Cooper, Interaction of the epididymal protein CD52 (HE5) with epididymal spermatozoa from men and cynomologous monkeys, Mol. Reprod. Dev. 48 (1997) 267– 275. [48] S. Ilangumaran, A. Briol, D.C. Hoessli, CD44 selectively associates with active Src family protein tyrosine kinases Lck and Fyn in glycosphingolipid-rich plasma membrane domains of human peripheral blood lymphocytes, Blood 91 (1998) 3901–3908. [49] K. Simons, D. Toomore, Lipid rafts and signal transduction, Nat. Rev. Mol. Cell Biol. 1 (2000) 31–39. [50] C. Gomez-Mouton, J.L. Abad, E. Mira, R.A. Lacalle, E. Gallardo, S. Jimenez-Baranda, I. Illa, A. Bernad, S. Manes, A.C. Martinez, Segregation of leading-edge and uropod components into specific lipid rafts during T cell polarization, Proc. Natl. Acad. Sci. USA 98 (2001) 9642–9647. [51] B.K. Davis, R. Byrne, K. Bedigian, Studies on the mechanism of capacitation: albumin-mediated changes in plasma membrane lipids during in vitro incubation of rat sperm cells, Proc. Natl. Acad. Sci. USA 77 (1980) 1546–1550. [52] S. Shadan, S.P. James, A.E. Howes, R. Jones, Cholesterol efflux alters lipid raft stability and distribution during capacitation of boar spermatozoa, Biol. Reprod. 71 (2004) 253–265. [53] N.L. Cross, Reorganization of lipid rafts during capacitation of human sperm, Biol. Reprod. 71 (2004) 1367–1373. [54] S.B. Sleight, P.V. Miranda, N.W. Plaskett, B. Maier, J. Lysiak, H. Scrable, J.C. Herr, P.E. Visconti, Isolation and proteomic analysis of mouse sperm detergent-resistant membrane fractions. Evidence for dissociation of lipid rafts during capacitation, Biol. Reprod. 73 (2005) 721–729. [55] J.M. Rifkin, G.E. Olson, Characterization of maturation-dependent extrinsic proteins of the rat sperm surface, J. Cell Biol. 100 (1985) 1582–1591. [56] P. Derr, C.H. Yeung, T.G. Cooper, C. Kirchhoff, Synthesis and glycosylation of CD52, the major ‘‘maturation associated’’ antigen of rat spermatozoa, in the cauda epididymis, Reproduction 121 (2001) 435–446. [57] E.D. Eccleston, T.W. White, J.B. Howard, D.W. Hamilton, Characterization of a cell surface glycoprotein associated with maturation of rat spermatozoa, Mol. Reprod. Dev. 37 (1994) 110–119.
Different glycoforms of the human GPI-anchored antigen CD52 associate differently with lipid microdomains in leukocytes and sperm membranes L. Ermini a, F. Secciani a, G.B. La Sala b, L. Sabatini c, D. Fineschi c, G. Hale d, F. Rosati a,* a Department of Evolutionary Biology, University of Siena, Italy Unit of Human Reproduction, Ospedale S.Maria Nuova, Reggio Emilia, Italy Laboratory of Hematology and Coagulation, Azienda Ospedaliera Universitaria Senese, Via Tufi 3, Siena, Italy d Sir William Dunn School of Pathology, University of Oxford, Oxford, UK b
c
Received 12 October 2005 Available online 24 October 2005
Abstract CD52 is a human GPI-anchored antigen, expressed exclusively in the immune system and part of the reproductive system (epididymal cells). Sperm cells acquire the antigen from the epididymal secretions when transiting in the epididymal corpus and cauda. The peptide backbone of CD52, consisting of only 12 aminoacids, is generally considered no more than a scaffold for post-translational modifications, such as GPI-anchor and especially N-glycosylation which occur at the third asparagine. The latter modification is highly heterogeneous, especially in the reproductive system, giving rise to many different glycoforms, some of which are tissue specific. A peculiar O-glycan-containing glycoform is also found in reproductive and immune systems. We determined to locate CD52 in microdomains of leukocytes and sperm membranes using two antibodies: (1) CAMPATH-1G, the epitope of which includes the last three aminoacids and part of the GPI-anchor of glycoforms present in leukocytes and sperm cells; (2) antigp20, the epitope of which belongs to the unique O-glycan-bearing glycoform also present in both cell types. Using a Brij 98 solubilization protocol and sucrose gradient partition we demonstrated that the CD52 glycoforms recognized by both antibodies are markers of typical raft microdomains in leukocytes, whereas in capacitated sperm the O-glycoform is included in GM3-rich microdomains different from the cholesterol and GM1-rich lipid rafts with which CAMPATH antigen is stably associated. The importance of the association between GM3 and O-glycans for formation of specialized microdomains was confirmed by heterologous CD52 insertion experiments. When prostasomes from human seminal fluid were incubated with rat sperm from different epididymal regions, the CD52 glycoform recognized by anti-gp20 decorated rat epididymal corpus and cauda sperm, associated with the same low-cholesterol GM3-rich sperm membrane fractions as in human sperm. The glycoforms recognized by CAMPATH-1G were not found in rat sperm. The relationship between this differential insertion and differences in glycosylation of rat and human CD52 is discussed. 2005 Elsevier Inc. All rights reserved. Keywords: CD52 antigen; Leukocyte; Sperm; GPI-anchored proteins; Membrane lipid rafts
The concept that specific lipid and lipid-derivatives are involved in determining membrane-ordered microdomains moving within the fluid glycerophospholipid matrix is gaining evidence. These microdomains, characterized by their *
Corresponding author. Fax: +39 577 234476. E-mail address: [email protected] (F. Rosati).
0006-291X/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.10.082
order and buoyant density, are resistant to solubilization by non-ionic detergents and are generally isolated by sucrose-gradient centrifuging. Different major functions have been proposed for them, including membrane trafficking and signalling [1,2]. In the immune system, these microdomains have been termed immunosynapses [3]. Attention has recently also been paid to the adhesion role of glycosyl
1276
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
epitopes carried by proteins or lipids of glycosphingolipidrich microdomains and the term glycosynapses has been coined for them [4–6]. GPI(glycosylphosphatidylinositol)anchored proteins are generally considered preferential inhabitants of membrane microdomains, although the GPI-linkage does not seem sufficient to determine this domain restriction [7]. CD52 is a human GPI-anchored antigen, product of a single gene on chromosome 1 [8]. Its functional role is still largely unknown and its association with microdomains has not been determined. CD52 is expressed exclusively in the immune system (lymphocytes, monocytes, and macrophages) and part of the reproductive system (epididymal cells). Sperm cells also bear the antigen but do not express it. This is because epididymal cells of the corpus and cauda districts secrete it into the surrounding fluid from which it is taken up by sperm in transit [9]. The peptide backbone of CD52 is short, consisting of only 12 aminoacids [10]. Indeed, it is generally considered no more than a scaffold for post-translational modifications such as GPI-anchorage and especially N-glycosylation which occur at the asparagine in third position [10]. The latter modification is highly heterogeneous, especially in the reproductive system [11], giving rise to many different glycoforms, some of which are tissue specific [12,13]. In the past and also recently, O-glycans have been suggested to be part of certain CD52 glycoforms [14–16]. The first and most commonly used anti-CD52 monoclonal antibody, termed CAMPATH-1, was generated in the immune system [17] and recognizes CD52 of the immune and reproductive systems [18]. Purification of CD52 from spleen and seminal plasma for structural characterization of N-glycans and GPI-anchor, and for peptide sequencing has always been done using this antibody [19,10]. The CAMPATH epitope includes the last three aminoacids and the first part of GPI [20]. Other anti-CD52 antibodies such as S-19 [21] and Mab H6-3C4 [13], generated in the reproductive system and directed against the N-linked chains, recognize CD52 glycoform subpopulations specific to the reproductive system [22]. In this panorama, antigp20, a polyclonal antibody that we raised against a purified sialyl-glycoprotein of the human sperm surface, homologous to CD52 [23], is peculiar. Indeed, it interacts strongly with a specific glycoform subpopulation of the CAMPATH antigen present in the immune and reproductive systems [24], a characteristic of which is to bear O-glycans not found in other glycoforms [16]. The antigp20 epitope includes these glycans [16]. Intriguingly, the O-glycoform is prevalently localized in the equatorial region of the head of capacitated sperm, whereas forms recognized by other anti-CD52 antibodies are equally distributed over the whole cell membrane [21,23]. A role in fertility has been suggested for this glycoform [24,25]. The CAMPATH antigen is highly expressed in leukocytes. The humanized form of CAMPATH antibodies has many therapeutic applications where depletion of lymphocytes is required, for example in leukemia, lymphoma,
bone marrow transplant, rheumatoid arthritis, and multiple sclerosis [26–30]. To determine whether high glycosylation heterogeneity and differential insertion in membranes influence the membrane microdomain association of CD52, we investigated the location of CD52 glycoforms recognized by CAMPATH and anti-gp20 with respect to detergent-insoluble glycosphingolipid/cholesterol-enriched microdomains in leukocytes and sperm. The possibility that microdomains are involved in cell-to-cell transfer of CD52 was also investigated by analyzing epididymal rat sperm after fusion with human prostasomes. Co-localization of the antigen glycoforms recognized by CAMPATH and gp20 antigen was also performed in leukocytes and sperm. Materials and methods Antibodies. The following antibodies were used: mouse anti-Lck (Sigma–Aldrich), anti-rabbit IgG conjugated with alkaline phosphatase (BioRad Microscience, Cambridge, MA), anti-mouse IgG conjugated with alkaline phosphatase (Sigma–Aldrich), anti-rat IgG conjugated with alkaline phosphatase (Sigma–Aldrich), streptavidin HRP-conjugated (Amersham Biosciences), biotin-conjugated cholera toxin B subunit (Sigma–Aldrich), anti-rat IgG conjugated with rhodamine (Sigma–Aldrich), fluorescein-conjugated anti-rabbit IgG (Calbiochem, San Diego, CA), CAMPATH-1G (rat IgG2b, clone YTH34.5G2b, Delta Biological S.r.l Rome, Italy, or a gift from G. Hale). Polyclonal anti-gp20 was obtained according to Focarelli et al. [23]. Anti-gp20 IgG were purified by affinity chromatography according to Flori et al. [16] using an AKTA liquid chromatography system. Cell and vesicle preparation procedures. All preparations were performed at 4 C unless otherwise stated. Human white blood cells were obtained from blood samples of healthy donors. Five millilitres aliquots of whole blood were mixed with 1.5 ml of 6% dextran solution in saline (NaCl 0.9%). After sedimentation for 2 h at 37 C, the leukocyte enriched supernatant was collected, washed, and centrifuged at 500g for 10 min Contaminating erythrocytes in the sediment were lysed with 1 ml chilled 0.8% NH4Cl solution [31]. Human ejaculate was obtained from a healthy donor and allowed to liquefy at room temperature for 30 min. Spermatozoa were separated from seminal plasma by centrifugation at 500g for 15 min and then washed twice in phosphate-buffered saline (PBS: 150 mM NaCl, 50 mM KH2PO4, pH 7.4). Spermatozoa were selected by a swim-up procedure. Briefly, 0.5 ml of BWW (Biggers–Whitten–Whittingham) medium [32] containing 35 mg/ml human serum albumin (HSA) was layered over aliquots of spermatozoa and incubated for 1 h at 37 C in 5% CO2 and 95% air. The upper layer, rich in mobile spermatozoa was collected and capacitation followed for 5 h at the same conditions. Seminal plasma, obtained as described previously, was centrifuged at 10,000g for 20 min to eliminate cell debris and residual sperm. The new supernatant was centrifuged at 105,000g for 120 min. The supernatant was discarded, and the pellet containing prostasomes and amorphous material was suspended in PBS containing 0.5 M KCl and centrifuged again at 105,000g for 120 min to remove the amorphous material. Cells and prostasomes were resuspended in 1 ml PBS for immunofluorescence analysis or in buffer A (2 mM Tris–HCl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 1 lg/ml leupeptin, 1 lg/ml pepstatin, 2 lg/ml chymostatin, and 5 lg/ml a2 macroglobulin) for isolation of detergent resistant membranes. Fluorescence microscopy. Leukocytes were washed three times with PBS and fixed for 10 min with 4% paraformaldehyde in PBS at 4 C. Cells (3 · 106) were then smeared onto a glass slide and after blocking the unspecific sites with PBS containing 2% BSA (PBS–BSA) and incubated with CAMPATH-1G and with anti-gp20 IgG diluted 1:25 in the same buffer for up to 30 min at room temperature. After three washings, leu-
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283 kocytes were incubated with respective fluorescein or rhodamine-conjugated secondary antibodies. Aliquots of sperm (3 · 106) were smeared onto ethanol-cleaned glass slides, and allowed to attach at room temperature, taking care to keep them in liquid phase to avoid damaging the sperm membranes. After several washings in PBS, the smears were blocked for 30 min with 2% BSA in PBS and then incubated for 1 h in the same buffer containing IgG anti-gp20 and CAMPATH-1G diluted 1:25. After several washings, the smears were incubated for 1 h with goat antirabbit IgG and goat anti-rat IgG conjugated, respectively, with fluorescein and rhodamine. Slides were then washed, mounted in 20 mM Tris–HCl, pH 8.0, 80% glycerol, and 4% (w/v) N-propyl gallate as anti-fade, and observed with a laser scanning confocal apparatus (TCS 4D, Leica, Heidelberg, Germany). The images were processed in Photoshop 5.5 (Adobe Systems, Mountain View, CA). Detergent extraction and sucrose gradient centrifugation. Cells (1 · 108 leukocyte and 1.5 · 108 spermatozoa) and prostasome suspension were homogenized using a motor-driven Potter–Elvehjem homogenizer (Kontes, Vineland, NJ) with a tight-fitting pestle. The homogenate was then centrifuged at 800g for 10 min, the post-nuclear supernatant (PNS) was pre-incubated for 4 min at 4 C. Brij 98 (Sigma–Aldrich) was then added to a final concentration of 1%. After 5 min of solubilization at 37 C, the PNS (1 ml) was diluted with 2 ml 37 C pre-warmed buffer A containing 2 M sucrose (final sucrose concentration 1.33 M; final Brij 98 concentration 0.33%) and chilled down on ice (55 min) before being placed at the bottom of a step sucrose gradient (0.9–0.8–0.7–0.6–0.5–0.4–0.3–0.2 M sucrose, 450 ll each) in buffer A. Gradients were centrifuged at 50,000 rpm for 16 h in a SW65 rotor (Beckman Instruments Inc.) at 4 C [33]. Five hundred microliter fractions were harvested from the top. The first one was discarded and the others, unless specified, pooled as follows: lowdensity fractions (2–4), intermediate-density fractions (5–7), and heavydensity fractions (8–10) were pooled. Low- and intermediate-density fractions were further centrifuged at 50,000 rpm in a SW65 rotor for 16 h and only the pellet was recovered and analyzed. For cholesterol extraction, leukocytes and sperm were pre-treated with 250 ll of 0.2% saponin in PBS at 4 C for 10 min with occasional mixing. The sample was then subjected to the detergent extraction procedure. Fusion of rat spermatozoa with human prostasomes. Rat sperm, collected in the three epididymal districts (caput, corpus, and cauda), were resuspended in Hepes 2 mM, pH 8.0, containing 0.32 M sucrose and BSA (1%, w/v) to have about 6 · 107 cells/ml 1. When required, prostasomes were added to reach a prostasome to sperm protein ratio of 2:1. These mixtures were incubated at 37 C for 60 min. Sucrose Hepes buffer (0.32 M sucrose), pH 8.0 (1 ml), was then added and sperm were collected by centrifugation for 10 min at 600g. The washing procedure was repeated twice using Hepes 2 mM, pH 8.0, containing 0.6 M KCl [34]. The pellet of fused sperm was resuspended in buffer A for sucrose density gradient analysis. Lipid analysis. Analysis of ganglioside GM1 was performed by directly testing blotted fractions with CTB. Aliquots corresponding to 2 lg protein of each fraction were spotted onto nitrocellulose filters; the membranes were subsequently blocked with TBS (20 mM Tris–HCl, 500 mM NaCl, pH 7.5) containing 3% non-fat powdered milk and then incubated with biotin-conjugated CTB diluted 1:1000 for 1 h at room temperature. After washing, nitrocellulose filters were incubated with HRP-conjugated streptavidin under the same conditions. After extensive rinsing in TBS– Tween 0.2%, the labelled components were developed using Immun-Star HRP Chemiluminescent Detection Kit (BioRad Microscience, Cambridge, MA) following the manufacturerÕs instructions. Analysis of cholesterol and ganglioside GM3 was performed by thin layer chromatography (TLC) after lipid extraction from fraction membranes. Aliquots (50 lg) of protein fractions, obtained by sucrose gradient centrifugation, were lipid extracted with chloroform:methanol (1:2, by volume) as described by Bligh and Dyer [35]. Lipids contained in the chloroform phase were dried under N2 stream and dissolved again in chloroform. Thin layer chromatography (TLC) was carried on high-performance TLC (HPTLC) plates (Silica Gel 60 F-254, Merck) with chloroform:methanol (98:2, v/v). For ganglioside analysis a 2:1:1 chloroform:methanol:water solution was added to the chloroform phase and
1277
centrifuged at 800g for 10 min. The resulting biphasic solution consisted of an upper phase containing polar lipids and a lower phase containing nonpolar lipids. The upper phase was removed, dried under N2 stream, and resuspended in a chloroform:methanol:water solution (15:30:4, v/v/v). The samples were then separated by TLC using a mixture of chloroform:methanol:water (60:40:9, v/v/v). Lipids were visualized using 20% (v/ v) H2SO4 at 150 C or resorcinol reagent [36]. Cholesterol and ganglioside GM3 were quantified by densitometry using commercial cholesterol and GM3 (Sigma–Aldrich) as standard. Electrophoresis and Western blotting. Proteins were concentrated as described by Wessel and Flugge [37]. Samples were separated by Tricine– sodium dodecyl sulfate–polyacrylamide gel electrophoresis [38]. The separation gel consisted of 10% acrylamide with a 3% stacking gel. In other experiments, proteins were separated on 10% polyacrylamide slab gels according to Laemmli [39]. The gels were transferred to nitrocellulose as described [40]. Nitrocellulose sheets were then blocked with 3% non-fat powdered milk in TBS and incubated overnight at 4 C with anti-gp20 IgG (1:100), anti-Lck diluted 1:1000 or CAMPATH-1G diluted 1:100. After several washings with TBS–Tween, the blots were incubated for 1 h at room temperature with the respective secondary antibodies conjugated with alkaline phosphatase. After extensive rinsing in TBS–Tween, labelled proteins were developed using Immun-star Chemiluminescent Detection Kit according to the manufacturerÕs instructions. Pre-immune serum was used as control. Protein concentration was determined using BCA protein assay reagent kit (Pierce Biotechnology, Inc.).
Results The CD52 O-glycoform co-localized with the other CD52 forms in leukocytes but not in capacitated sperm In a first series of experiments we investigated co-localization of the different CD52 glycoforms on leukocytes and capacitated spermatozoa using CAMPATH-1G and antigp20 simultaneously. Equivalent amounts (3 · 106) of cells were smeared on glass slides and incubated simultaneously with CAMPATH-1G and anti-gp20, and then with the corresponding anti-rat or anti-rabbit IgG conjugated with fluorescein or rhodamine. The results demonstrated co-localization of all CD52 glycoforms in leukocyte but confirmed their different localization in capacitated sperm. In fact, as shown in Fig. 1, the signals of the two antibodies overlapped completely in leukocytes whereas in capacitated sperm, overlapping only occurred in the equatorial region of the head and sometimes in the neck region; only the CAMPATH-1G signal was detected on the rest of the sperm surface. In leukocytes all CD52 glycoforms are associated to lipid raft microdomains Rafts have been generally defined by their insolubility in cold Triton X-100 [41] and this characteristic is widely used as a basis for isolating them [1,42]. The detergents Brij 98, 96 or 56 have recently replaced Triton since they are reported to preserve membrane rafts even above 37 C, thus excluding the objection that Triton X-100 used at 4 C induces lipid transitions, artificially creating detergent-resistant protein–lipid complexes [33]. We used Brij 98 for fractioning the leukocyte membranes. Cells were lysed in Brij 98 at 37 C, the post-nuclear supernatant
1278
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
Fig. 1. In peripheral blood leukocytes, the CD52 glycoforms recognized by CAMPATH-1G and anti-gp20 co-localized, whereas in swim-up selected and 5 h capacitated sperm cells the O-glycoform recognized by anti-gp20 localized in the equatorial region of the head and in the first part of the tail. Leukocytes were fixed for 10 min with 4% paraformaldehyde at 4 C and then smeared on glass slides. Sperm were smeared directly on glass slides and allowed to attach at room temperature, taking care to keep them in liquid phase to avoid damaging the sperm membranes. Both cell types were then incubated for 1 h simultaneously with anti-gp20 and CAMPATH-1G followed by goat anti-rabbit IgG and goat anti-rat IgG conjugated, respectively, with fluorescein and rhodamine.
(PNS) was ultracentrifuged in a sucrose density gradient, and 10 fractions were recovered. The first seven fractions were always ultracentrifuged and only the pelletted material was used for the following preparations. This was done for avoiding the presence of soluble proteins in these fractions. Each fraction was analyzed for the presence of cholesterol, GM1, and GM3. GM1 is, in fact, considered in general a marker of lipid rafts and GM3 has been described to reside in these domains in leukocytes [43]. All of them were found to be significantly prevalent in the lowest density fractions 2, 3, and 4, slightly present in the intermediate fractions 5, 6, and 7, and absent in the last three heavy-density ones. For all this, we pooled the three groups of fractions as indicated in Fig. 2 and considered only the fractions 2–4 as typically containing the detergent insoluble lipid rafts. Fractions 2–4 corresponded to 10–20% sucrose density, fractions 5–7 corresponded to 20–30% sucrose density and fractions 8–10 corresponded to 30–40% sucrose density. Protein quantification of the three pooled fractions revealed that the first fractions contained 5% of the total PNS protein, the intermediates 20%, and the last ones the rest. In the following experiments, the pooled fractions were always analyzed under normalized conditions, i.e., using the same protein concentration. Equivalent amounts of protein of the three fractions groups were then run in SDS–PAGE, blotted, and analyzed with CAMPATH-1G, anti-gp20, and an anti-Lck, an antigen known to reside in the lipid rafts [44]. As shown in Fig. 2, the CD52 forms recognized by both the anti-CD52 antibodies were included in the lipid raft domains along with Lck since they prevalently floated in the fractions 2–4. However, unlike CAMPATH, anti-gp20 also revealed the corresponding antigen form in fractions 5–7 (Fig. 2). To confirm the association of CD52 to raft domains we also evaluated the effect of saponin on CD52 localization.
Fig. 2. In leukocytes, all the CD52 glycoforms recognized by CAMPATH-1G and anti-gp20 prevalently partitioned in the low-density sucrose fractions with cholesterol, GM1, GM3, and Lck. After removal of cholesterol by saponin, they shifted to high-density fractions. Postnuclear supernatant of homogenized cells was treated with Brij 98 at 37 C in the presence or absence of 0.2% saponin and then sucrose-gradient fractionated. The pooled fractions, normalized for protein content, were then tested for Lck as well as anti-gp20 and CAMPATH-1G CD52 glycoforms by immunoblot. Fractions were also tested for ganglioside GM1 by direct dot blot analysis with CTB, and for cholesterol and ganglioside GM3 by thin layer chromatography (TLC) after lipid extraction from membrane fractions.
Saponin is known for its ability to disorganize plasma membrane and to promote a shift in protein localization [33]. PNS was dissolved with Brij 98 at 37 C with or without 0.2% saponin and then fractionated on a sucrose gradient. The fractions were then tested for the presence of Lck and CD52. As shown in Fig. 2, after this treatment all the antigen forms were prevalently recovered in the detergent-soluble fractions of cell membranes, as expected. However, part of the antigen recognized by antigp20 was still detectable in the intermediate fractions (Fig. 2). In capacitated human sperm, the O-glycoform is inserted in GM3-rich domains whereas the other CD52 forms are associated to cholesterol/GM1-rich raft microdomains Sperm plasma membranes were then investigated for CD52 localization using the same protocol as for leukocytes. In a first series of experiments, we analyzed swim up selected and 5 h capacitated sperm since they are more homogeneous and bear a specific localization of the glycoform recognized by anti-gp20 [23]. 1.5 · 108 sperm were homogenized, subjected to the detergent extraction procedure and ultracentrifuged in a sucrose density gradient. The 10 fractions recovered were harvested from the top and analyzed for cholesterol content. The fractions were then pooled as described for leuko-
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
1279
cytes after confirmation that cholesterol was maximal in the low-density fractions 2, 3, and 4 (Fig. 3). Analysis of the three blotted fractions groups for the presence of GM1 and GM3 confirmed the association of GM1 with the same cholesterol-rich fractions but also revealed that GM3 was instead prevalently associated with the intermediate 5–7 fractions. Equivalent amounts of the three fraction groups were then separated on Tricine–sodium dodecyl sulfate one-dimensional 10% polyacrylamide gel, blotted, and analyzed using CAMPATH-1G and anti-gp20. The positivity pattern was very different from that found in leukocytes. In fact, only the first antibody produced a broad positive signal around 20 kDa in the first fraction group, whereas anti-gp20 gave a weak signal in the same fractions and strong reactivity in the intermediate GM3-rich fractions (Fig. 3). To be sure of these results, we analyzed all the sucrose gradient fractions for the presence of CAMPATH and gp20 antigen forms and for GM1 and GM3. The results showed that the gp20 glycoform was found prevalent in fraction 5 and to a lesser extent in fraction 6 with GM3, whereas the CAMPATH antigen forms were prevalent in fractions 3–4 with GM1 (data not shown). This different behavior of the CD52 glycoforms was confirmed in more than 20 independent experiments by analyzing the same fractions with the two antibodies sequentially, stripping the first antibody. Furthermore, in order to analyze whether this behavior was specific for Brij solubiliza-
tion we treated the same samples with Triton X-100 at 4 C and analyzed the sucrose gradient fractions in the same way. Again CAMPATH antigen was mainly recovered in the first three fractions and the gp20 form in intermediate fractions. Saponin treatment of sucrose-separated sperm plasma membranes also produced a different pattern with respect to leukocytes. Neither the anti-gp20 glycoform nor the others recognized by CAMPATH shifted in the heavy fractions as they did in leukocytes but all were recovered in the same fractions as before treatment (Fig. 3). Since changes in lipid composition and distribution in the plasma membrane are major capacitation events [45], we investigated whether the different localization of CD52 glycoforms was a consequence of the capacitation process and depended on loss of cholesterol. Equal amounts of washed ejaculated sperm were treated in the same way as capacitated sperm, and lipids and proteins in the sucrose density gradient fractions were investigated in the presence or absence of saponin. As shown in Fig. 4, before capacitation and without saponin, cholesterol, and GM1 were associated with the low- and intermediate-density fractions in almost equivalent amounts and GM3 was prevalent in the intermediate fractions with all the CD52 glycoforms. After removal of cholesterol by saponin, the association pattern of lipids (data not shown) and CD52 (Fig. 4) turned out to be identical to that of the capacitated sperm membranes.
Fig. 3. In swim-up-selected and 5 h capacitated human sperm, the CD52 CAMPATH glycoforms prevalently partitioned in the low-density sucrose fractions with GM1, whereas the anti-gp20 form was associated with GM3 in the intermediate-density sucrose fractions. The different forms did not shift position after saponin treatment. Post-nuclear supernatant of the cells was treated with Brij 98 at 37 C in the presence or absence of 0.2% saponin and then sucrose-gradient fractionated. The fractions, normalized for protein content, were then tested for anti-gp20 and CAMPATH-1G CD52 glycoforms by immunoblot. The same fractions were also tested for the presence of the ganglioside GM1 by direct dot blot analysis with CTB, and for cholesterol and ganglioside GM3 by thin layer chromatography (TLC) after lipid extraction from membrane fractions.
Fig. 4. In freshly ejaculated sperm, the membrane was not yet organized like capacitated sperm membranes, except for GM3 and the glycoform recognized by anti-gp20 that were already in the intermediate fractions. By contrast, cholesterol and GM1 were equally distributed in more than one fraction and CAMPATH antigen forms were prevalent in the intermediate fractions. Incubation of sperm with saponin restored the partitioning properties of capacitated sperm. Solubilization and fractionation of these cells were done as in capacitated sperm.
1280
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
Only the CD52 O-glycoform of prostasomes was efficiently inserted in heterologous epididymal sperm, associated with low-cholesterol/GM3 rich microdomains The mechanism by which CD52 is stably inserted in the plasma membrane of sperm transiting in the epididymis is still largely unknown; however, transfer from large aggregates of the antigen or from vesicles containing it (termed prostasomes) in epididymal fluid has often been suggested [46,47]. Transfer capacity of the different forms of CD52 and the role played by microdomains in this transfer were investigated under heterologous conditions, i.e., using rat epididymal sperm and human prostasomes. To do this equal amounts of rat sperm (6 · 107) obtained from the caput, corpus, and cauda of epididymis were challenged with a constant amount of human prostasomes and the rat sperm was then subjected to the same solubilization protocol used for human sperm. In a first series of experiments, we investigated the presence of the antigen in an equivalent Brij-insoluble membrane fraction of sperm recovered from the three epididymal districts. The results showed that the antigen was strongly present in sperm from the cauda, faintly detectable in the sperm from the corpus, and absent in sperm from the caput (Fig. 5A). The Brij solubilized rat cauda sperm membranes were then sucrose density gradient, centrifuged, and the fractions were analyzed for lipid and protein content. The results revealed a distribution pattern of cholesterol, GM1, GM3, and CD52 O-glycoform identical to that of capacitated human sperm. Cholesterol and GM1 were recovered in the low-density fractions, whereas GM3 and the main portion of the O-glycan CD52 form were detected in fractions 5–7. A lower amount of this form was also found in the detergent insoluble fractions 2–4 and the detergent soluble fractions 8–10 (Fig. 5C). Immunofluorescence analysis revealed that the heterologous CD52 was prevalently inserted in the membrane overlaying the acrosome and in the first part of the tail (Fig. 5B). Surprisingly, CAMPATH antigen was not detectable in any fraction after density gradient sucrose fractionation nor in the sperm membrane after immunofluorescence analysis. To be sure that incubation of prostasomes with sperm did not alter the CAMPATH epitope, prostasomes were recovered after incubation and proteins were tested in immunoblot with CAMPATH-1G and anti-gp20. All CD52 glycoforms occurred in prostasomes but the signal intensity of the same amount of prostasomes before and after incubation with the two antibodies was unchanged for CAMPATH-1G and much lower for anti-gp20, thus confirming its insertion in rat sperm. Discussion We looked for CD52 in microdomains of leukocytes and sperm membranes using two antibodies: (1) CAMPATH-1G, the epitope of which includes the last three
Fig. 5. Human CD52 O-glycan-containing glycoform, but not CAMPATH glycoforms, may be transferred to membranes of rat sperm obtained from the epididymis. (A) Transfer occurred in sperm from the epididymal corpus and cauda but not in those from the caput. (B) Transfer of human CD52 O-glycan-containing glycoform occurred in the membrane of the acrosomal region and in part of the tail of rat sperm (C). The human CD52 glycoform co-isolated with GM3 in intermediate-density sucrose fractions. Rat sperm, obtained from the three districts of the epididymis (caput, corpus, and cauda), were incubated with prostasomes for 60 min. They were then collected by centrifugation for 10 min at 600g and the sperm membranes were solubilized and fractionated like those of human sperm. For immunofluorescence analysis, rat cauda sperm were incubated with prostasomes and after centrifugation smeared on glass slides and incubated with anti-gp20 and CAMPATH-1G followed by the corresponding secondary antibodies.
aminoacids and part of the GPI-anchor of glycoforms present in leukocytes and sperm cells [20]; (2) anti-gp20, the epitope of which belongs to a unique O-glycan-bearing glycoform also present in both cell types [16]. Using a Brij 98 solubilization protocol and sucrose gradient partition we demonstrated that the CD52 glycoforms recognized by both antibodies in leukocytes are markers of typical raft microdomains, whereas in sperm the same glycoforms have differential insertion. In fact, most of the antigen recognized by CAMPATH-1G and anti-gp20 in leukocytes was recovered in the low-density, high-cholesterol, GM1rich fraction, along with Lck, known to reside constitutionally in raft microdomains of these cells [44,48]. By our method, this fraction included fractions 2, 3, and 4 at 10–20% sucrose density. GM3, also known to reside in raft microdomains of peripheral blood cells [43], was found to co-isolate with the CAMPATH and anti-gp20 glycoforms. Furthermore, like other typical raft-associated
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
proteins [49], most CD52 glycoforms shifted to the highdensity, detergent-soluble fraction when cholesterol was depleted by saponin. To comparatively analyze CD52 glycoforms in male germ cells, we first used swim-up-selected, 5-h capacitated sperm, due to the specific localization of the CD52 O-glycoform in these cells [23] and the need to use cells as homogeneous as possible. Interestingly, the O-glycoform was found to be differently associated with respect to most of the CAMPATH antigen. In fact, when sperm membranes were fractionated in the same way as those of leukocytes and the fractions pooled, most of the CAMPATH antigen was recovered in the same fractions (fractions 2–4) as in leukocytes but the O-glycan-containing glycoform was mainly recovered in fractions 5–7 at 20–30% sucrose density. The markers of typical raft microdomains, cholesterol, and GM1, were found associated with low-density fractions, where CAMPATH antigen was prevalent. Unlike in leukocytes, however, GM3 was associated with intermediate sucrose density fractions together with the O-glycoform. Many independent experiments with the same blotted samples and the two antibodies were done to be sure of this different partitioning property but the results were always the same. The finding was also confirmed using the cold Triton X-100 solubilization protocol. The distinct partitioning of CD52 glycoforms, cholesterol, GM1, and GM3 in the 20% and 30% sucrose density fractions thus indicates that the different density fractions contain separate microdomains. The high-cholesterol and GM1 content of fractions 2–4 indicated that in capacitated sperm, as in leukocytes, these fractions contain typical raft microdomains. The CAMPATH antigen is therefore also a marker of these domains in capacitated sperm. The question is now to understand the kind of microdomain in which CD52 O-glycoform is included. These microdomains contain low-cholesterol and GM1, and are characterized by a specific prevalence of GM3. Interestingly, an asymmetric distribution of GM1- and GM3-enriched domains has been described as being associated with the leading-edge and uropod of migrating lymphocytes [50] and GM3 has often been indicated as playing a key role in formation of membrane platforms termed glycosynapses involved in cell–cell interaction [6]. Glycosynapse characteristics also include independence from cholesterol and the cis-interacting role of GM3 with oligosaccharide chains of platform constituents [6]. This means that the CD52 O-glycoform and GM3 may together take part in regionalized specific microdomains involved in adhesion mechanisms. This would explain the restricted localization of the CD52 O-glycoform in the equatorial region of the head of capacitated sperm [23] and its suggested role of interacting with the oocyte membrane [23,25]. Since capacitation is associated with an efflux of cholesterol that leads to significant rearrangement of sperm membrane components [51], we also investigated whether the different partition properties of the CD52 glycoforms were a consequence of capacitation-dependent cholesterol
1281
loss. Indeed, during capacitation at least 60–70% of cholesterol is lost and a consequent alteration of lipid raft constituents in sperm membranes has been reported in boar [52] and in human sperm membranes [53,54]. The first difference we noted was that cholesterol and GM1 of freshly ejaculated sperm membranes were not restricted to the same low-density fractions as in capacitated sperm but were equally distributed in fractions 2–4 and 5–7. Instead, GM3 partitioned in the intermediate sucrose density fractions as in capacitated sperm membranes. Unlike in capacitated sperm, in freshly ejaculated sperm the glycoforms recognized by CAMPATH-1G were found prevalently associated with fractions 5–7 along with that recognized by anti-gp20. However, external removal of cholesterol by saponin gave rise to a lipid and CD52 glycoform partition pattern identical to that found in capacitated sperm. On the other hand, treatment of capacitated sperm with saponin did not alter the partition properties of the different CD52 glycoforms. Taken together, these results suggest that the membrane of freshly ejaculated sperm is particularly rich in cholesterol and is highly unstable, and that the progressive capacitation-dependent loss of cholesterol defines its final rigid organization ready for fertilization, when: (1) canonical lipid rafts with which CAMPATH antigen remains stably associated are formed; (2) GM3-rich microdomains containing CD52 O-glycoform segregate in specific regions. The importance of GM3 in forming microdomains in association with the O-glycan-containing glycoform was further confirmed by experiments of heterologous CD52 insertion experiments. Indeed, when prostasomes from human seminal fluid were incubated with rat sperm from different epididymal regions, the CD52 glycoform recognized by anti-gp20 decorated the rat sperm, associated with the same low-cholesterol, GM3-rich sperm membrane fractions as in human sperm. Only sperm from the corpus and cauda accepted the antigen. This is in line with previous reports suggesting that in rats [55,56], as well as in humans and monkeys [47], only corpus and cauda sperm have sufficient maturation status for antigen uptake. Surprisingly, the CAMPATH antigen was not transferred to the rat sperm membrane, despite its abundance in prostasomes. We cannot yet explain this different behavior, but it seems significant that unlike human CD52, rat CD52 does not appear to have N-linked oligosaccharide chains but rather a number of O-linked oligosaccharide chains [56,57]. On the other hand, the CD52 forms recognized by CAMPATH have already been reported to transfer better from large aggregates of the protein than from prostasomes [46]. We therefore deduce that the oligosaccharides bound to peptides could be determinant for their transfer from an external source to membrane microdomains. On the other hand, since neither the GPI-anchor of the human CD52 O-glycoform nor that of rat CD52 is known, these anchors may also play a role. Taken together, the present results indicate a difference in location of CD52 in
1282
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
membrane domains when inserted internally through the secretory pathway or when transferred from an external source, and are in line with our previous suggestions that the CD52 O-glycoform plays a role in fertilization [24,25]. Concerning CAMPATH antigen, indications that it is stably associated with raft microdomains are in line with the possibility that the antigen is involved in signaling processes in leukocytes and sperm. Finally the fact that the O-glycoform segregates differently from the other CD52 glycoforms in sperm rises the question of whether the same may occur in leukocytes under certain conditions. Research into this aspect is now underway study in our laboratory.
References [1] K. Simons, E. Ikonen, Functional rafts in cell membranes, Nature 387 (1997) 569–572. [2] S.K. Pierce, Lipid rafts and B-cell activation, Nat. Rev. Immunol. 2 (2002) 96–105. [3] S.K. Bromley, W.R. Burack, K.G. Johnson, K. Somersalo, T.N. Sims, C. Sumen, M.M. Davis, A.S. Shaw, P.M. Allen, M. Dustin, The immunological synapse, Annu. Rev. Immunol. 19 (2001) 375–396. [4] K. Iwabuchi, K. Handa, S. Hakomori, Separation of ‘‘glycosphingolipid signaling domain’’ from caveolin-containing membrane fraction in mouse melanoma B16 cells and its role in cell adhesion coupled with signaling, J. Biol. Chem. 273 (1998) 33766–33773. [5] S. Hakomori, Traveling for the glycosphingolipid path, Glycoconj. J. 17 (2000) 627–647. [6] S. Hakomori, Glycosynapses: microdomains controlling carbohydrate-dependent cell adhesion and signaling, An. Acad. Bras. Cienc. 76 (2004) 553–572. [7] D.F. Legler, M.A. Doucey, P. Schneider, L. Chapatte, F.C. Bender, C. Bron, Differential insertion of GPI-anchored GFPs into lipid rafts of live cells, FASEB J. 19 (2005) 73–75. [8] M. Tone, K.F. Nolan, L.A. Walsh, Structure and chromosomal location of mouse and human CD52 genes, Biochim. Biophys. Acta 1446 (1999) 334–340. [9] C. Kirchhoff, G. Hale, Cell-to-cell transfer of glycosylphosphatidylinositol-anchored membrane proteins during sperm maturation, Mol. Hum. Reprod. 2 (1996) 177–184. [10] A. Treumann, M.R. Lifely, P. Schneider, M.A.J. Ferguson, Primary structure of CD52, J. Biol. Chem. 270 (1995) 6088–6099. [11] S. Schroter, P. Derr, H.S. Conrad, M. Nimtz, G. Hale, C. Kirchhoff, Male specific modification of human CD52, J. Biol. Chem. 274 (1999) 29862–29873. [12] A.B. Diekman, E.J. Norton, V.A. Westbrook, K.L. Klotz, S. NaabyHansen, J.C. Herr, Anti-sperm antibodies from infertile patients and their cognate sperm antigens: a review. Identity between SAGA-1, the H6-3C4 antigen, and CD52, Am. J. Reprod. Immunol. 4 (2000) 134–143. [13] A. Hasegawa, Y. Fu, H. Tsubamoto, Y. Tsuji, H. Sawai, S. Komori, K. Koyama, Epitope analysis for human sperm-immobilizing monoclonal antibodies, MAb H6-3C4, 1G12 and campath-1, Mol. Hum. Reprod. 9 (2003) 337–343. [14] H. Valentin, C. Gelin, L. Coulombel, D. Zoccola, J. Morizet, A. Bernard, The distribution of the Cdw52 molecule on blood cells and characterization of its involvement in T cell activation, Transplantation 54 (1992) 97–104. [15] A. Hasegawa, K. Koyama, Antigenic epitope for sperm-immobilizing antibody detected in infertile women, J. Reprod. Immunol. (2005) (Epub ahead of print). [16] F. Flori, C. Della Giovampaola, R. Focarelli, F. Secciani, G.B. La Sala, A. Nicoli, G. Hale, F. Rosati, Epitope analysis of immunoglobulins against gp20, a GPI-anchored protein of the human sperm
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30] [31]
[32]
[33]
surface homologous to leukocyte antigen CD52, Tissue Antigens 66 (2005) 209–216. G. Hale, M.Q. Xia, H.P. Tighe, M.J.S. Dyer, H. Waldmann, The CAMPATH-1 antigen (CDw52), Tissue Antigens 35 (1990) 118–127. G. Hale, P.D. Rye, A. Warfor, I. Lauder, A. Brito-Babapulle, The glycosylphosphatidylinositol-anchored CDw52 is associated with the epididymal maturation of human spermatozoa, J. Reprod. Immunol. 23 (1993) 189–205. M.Q. Xia, G. Hale, M.R. Lifely, M.J. Ferguson, D. Campbell, L. Packman, H. Waldmann, Structure of the CAMPATH-1 antigen, a GPI-anchored glycoprotein which is an exceptionally good target for complement lysis, Biochem. J. 29 (1993) 633–640. G. Hale, Synthetic peptide mimotope of the CAMPATH-1 (CD52) antigen, a small glycosylphosphatidylinositol-anchored glycoprotein, Immunotechnology 1 (1995) 175–187. A.B. Diekman, E.J. Norton, K.L. Klotz, V.A. Westbrook, H. Shibahara, S. Naaby-Hansen, C.J. Flickinger, J.C. Herr, N-linked glycan of a sperm CD52 glycoform associated with human infertility, FASEB J. 13 (1999) 1303–1313. E.J. Norton, A.B. Diekman, V.A. Westbrook, D.W. Mullins, K.L. Klotz, L.L. Gilmer, T.S. Thomas, D.C. Wright, J. Brisker, V.H. Engelhard, C.J. Flickinger, J.C. Herr, A male genital tract-specific carbohydrate epitope on human CD52: implications for immunocontraception, Tissue Antigens 60 (2002) 354–364. R. Focarelli, A. Giuffrida, S. Capparelli, M. Scibona, F. Menchini Fabris, F. Francavilla, S. Francavilla, C. Della Giovampaola, F. Rosati, Specific localization in the equatorial region of gp20, a 20 kDa sialylglycoprotein of the capacitated human spermatozoon acquired during epididymal transit which is necessary to penetrate zona-free hamster eggs, Mol. Hum. Reprod. 4 (1998) 119–125. R. Focarelli, S. Francavilla, F. Francavilla, C. Della Giovampaola, A. Santucci, F. Rosati, A sialoglycoprotein, gp20, of the human capacitated sperm surface is a homologue of the leukocyte CD52 antigen: analysis of the effect of anti-CD52 monoclonal antibody (CAMPATH-1) on capacitated spermatozoa, Mol. Hum. Reprod. 5 (1999) 46–51. V. Giuliani, C. Pandolfi, R. Santucci, F. Pelliccione, B. Macerola, R. Focarelli, F. Rosati, C. Della Giovampaola, F. Francavilla, S. Francavilla, Expression of gp20, a human sperm antigen of epididymal origin, is reduced in spermatozoa from subfertile men, Mol. Reprod. Dev. 69 (2004) 235–240. C.M. Lockwood, S. Thiru, J.D. Isaacs, G. Hale, H. Waldmann, Long-term remission of intractable systemic vasculitis with monoclonal antibody therapy, Lancet 341 (1993) 1620–1622. T. Moreau, J. Thorpe, D. Miller, I. Moseley, G. Hale, H. Waldmann, D. Clayton, M. Wing, N. Scolding, A. Compston, Preliminary evidence from magnetic resonance imaging for reduction in disease activity after lymphocyte depletion in multiple sclerosis, Lancet 344 (1994) 298–301. A. Osterborg, A.S. Fassas, A. Anagnostopoulos, M.J. Dyer, D. Catovsky, H. Mellstedt, Humanized CD52 monoclonal antibody Campath-1H as first-line treatment in chronic lymphocytic leukaemia, Br. J. Haematol. 93 (1996) 151–153. J.D. Isaacs, V.K. Manna, N. Rapson, K.J. Bulpitt, B.L. Hazleman, E.L. Matteson, E.W. St Clair, T.J. Schnitzer, J.M. Johnston, CAMPATH-1H in rheumatoid arthritis: an intravenous dose-ranging study, Br. J. Rheumatol. 35 (1996) 231–240. P. Rebello, G. Hale, Pharmacokinetics of CAMPATH-1H: assay development and validation, J. Immunol. Methods 260 (2002) 285–302. A. Garcia, J. Niubo, M.A. Benitez, M. Viqueira, J.L. Perez, Comparison of two leukocyte extraction methods for cytomegalovirus antigenemia assay, J. Clin. Microbiol. 34 (1996) 182–184. J.D. Biggers, W.K. Whitten, D.G. Whittingham, The culture of mouse embryos in vitro, in: J.C. Daniel (Ed.), Methods in Mammalian Embriology, Freeman, S.Francisco, 1971, pp. 86–116. P. Drevot, C. Langlet, X.J. Guo, A.M. Bernard, O. Colard, J.P. Chauvin, R. Lasserre, H.T. He, TCR signal initiation machinery is
L. Ermini et al. / Biochemical and Biophysical Research Communications 338 (2005) 1275–1283
[34]
[35] [36] [37]
[38]
[39] [40]
[41]
[42] [43]
[44]
[45]
[46]
pre-assembled and activated in a subset of membrane rafts, EMBO J. 21 (2002) 1899–1908. C.A. Palmerini, C. Saccardi, E. Carlini, R. Fabiani, G. Arienti, Fusion of prostasomes to human spermatozoa stimulates the acrosome reaction, Fertil. Steril. 80 (2003) 1181–1184. E.G Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Med. Sci. 37 (1959) 911–917. T. Miettinen, I.T. Takki-Luukkainen, Use of butyl acetate in determination of sialic acid, Acta Chem. Scand. 13 (1959) 856–858. D. Wessel, U.I. Flugge, A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids, Anal. Biochem. 138 (1984) 141–143. H. Schagger, G. von Jagow, Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa, Anal. Biochem. 166 (1987) 368–379. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of the bacteriophage T4, Nature 227 (1970) 680–685. H. Towbin, T. Staehelin, J. Gordon, Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc. Natl. Acad. Sci. USA 76 (1979) 4350–4354. D.A. Brown, J.K. Rose, Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface, Cell 68 (1992) 533–544. D.A. Brown, E. London, Functions of lipid rafts in biological membranes, Annu. Rev. Cell Dev. Biol. 14 (1998) 111–136. M. Sorice, A. Longo, T. Garofalo, V. Mattei, R. Misasi, A. Pavan, Role of GM3-enriched microdomains in signal transduction regulation in T lymphocytes, Glycoconjugate J. 20 (2004) 63–70. B. Kovacs, M.V. Maus, J.L. Riley, G.S. Derimanov, G.A. Koretzky, C.H. June, T.H. Finkel, Human CD8+ T cells do not require the polarization of lipid rafts for activation and proliferation, Proc. Natl. Acad. Sci. USA 99 (2002) 15006–15011. F.M. Flesch, J.F.H.M. Brouwers, P.E.F.M. Nivelstein, A.J. Verkleij, L.M.G. van Golde, B. Colenbrander, B.M. Gadella, Bicarbonate stimulated phospholipid scrambling induces cholesterol redistribution and anables cholesterol depletion in the sperm plasma membrane, J. Cell Sci. 114 (2001) 3543–3555. I.A. Rooney, J.E. Heuser, J.P. Atkinson, GPI-anchored complement regulatory proteins in seminal plasma, J. Clin. Invest. 97 (1996) 1675– 1686.
1283
[47] C.H. Yeung, A. Schroter, C. Wagenfeld, C. Kirchhoff, S. Kliesch, D. Poser, G.F. Wainaner, E. Nieschlag, T.G. Cooper, Interaction of the epididymal protein CD52 (HE5) with epididymal spermatozoa from men and cynomologous monkeys, Mol. Reprod. Dev. 48 (1997) 267– 275. [48] S. Ilangumaran, A. Briol, D.C. Hoessli, CD44 selectively associates with active Src family protein tyrosine kinases Lck and Fyn in glycosphingolipid-rich plasma membrane domains of human peripheral blood lymphocytes, Blood 91 (1998) 3901–3908. [49] K. Simons, D. Toomore, Lipid rafts and signal transduction, Nat. Rev. Mol. Cell Biol. 1 (2000) 31–39. [50] C. Gomez-Mouton, J.L. Abad, E. Mira, R.A. Lacalle, E. Gallardo, S. Jimenez-Baranda, I. Illa, A. Bernad, S. Manes, A.C. Martinez, Segregation of leading-edge and uropod components into specific lipid rafts during T cell polarization, Proc. Natl. Acad. Sci. USA 98 (2001) 9642–9647. [51] B.K. Davis, R. Byrne, K. Bedigian, Studies on the mechanism of capacitation: albumin-mediated changes in plasma membrane lipids during in vitro incubation of rat sperm cells, Proc. Natl. Acad. Sci. USA 77 (1980) 1546–1550. [52] S. Shadan, S.P. James, A.E. Howes, R. Jones, Cholesterol efflux alters lipid raft stability and distribution during capacitation of boar spermatozoa, Biol. Reprod. 71 (2004) 253–265. [53] N.L. Cross, Reorganization of lipid rafts during capacitation of human sperm, Biol. Reprod. 71 (2004) 1367–1373. [54] S.B. Sleight, P.V. Miranda, N.W. Plaskett, B. Maier, J. Lysiak, H. Scrable, J.C. Herr, P.E. Visconti, Isolation and proteomic analysis of mouse sperm detergent-resistant membrane fractions. Evidence for dissociation of lipid rafts during capacitation, Biol. Reprod. 73 (2005) 721–729. [55] J.M. Rifkin, G.E. Olson, Characterization of maturation-dependent extrinsic proteins of the rat sperm surface, J. Cell Biol. 100 (1985) 1582–1591. [56] P. Derr, C.H. Yeung, T.G. Cooper, C. Kirchhoff, Synthesis and glycosylation of CD52, the major ‘‘maturation associated’’ antigen of rat spermatozoa, in the cauda epididymis, Reproduction 121 (2001) 435–446. [57] E.D. Eccleston, T.W. White, J.B. Howard, D.W. Hamilton, Characterization of a cell surface glycoprotein associated with maturation of rat spermatozoa, Mol. Reprod. Dev. 37 (1994) 110–119.