- Email: [email protected]
Cholesterol and the biosynthesis of glycosphingolipids are required for sperm activation in Caenorhabditis elegans
Biochimica et Biophysica Acta 1821 (2012) 934–942
Contents lists available at SciVerse ScienceDirect
Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbalip
Cholesterol and the biosynthesis of glycosphingolipids are required for sperm activation in Caenorhabditis elegans Jiangli Dou, Lianwan Chen, Youqiao Hu, Long Miao ⁎ Laboratory of Noncoding RNA, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District, Beijing 100101, China
a r t i c l e
i n f o
Article history: Received 24 September 2011 Received in revised form 28 February 2012 Accepted 16 March 2012 Available online 28 March 2012 Keywords: Caenorhabditis elegans Major sperm protein (MSP) Sperm activation Sperm motility Lipid microdomains
a b s t r a c t Ejaculated mammalian sperm must acquire fertilization capacity after residing into the female reproductive tract, a process collectively known as capacitation. Cholesterol efflux was required for sperm maturation. Different from flagellated sperm, C. elegans sperm are crawling cells. C. elegans sperm are highly enriched with cholesterol though this animal species lacks biosynthetic pathway for cholesterol and its survival requires an exogenous cholesterol supply. The low abundance of cholesterol in C. elegans lipid extract is thought insufficient to form lipid microdomains ubiquitously in this organism. We present evidence that cholesterol is enriched in the plasma membrane of C. elegans spermatids and that cholesterol- and glycosphingolipids (GSLs)-enriched membrane microdomains (lipid microdomains) mediate sperm activation. Disruption of sperm lipid microdomains by acute manipulation of cholesterol in vitro blocks the sperm activation. Restriction of cholesterol uptake also results in the abnormal sperm activation in both males and hermaphrodites. Manipulation of the integrity of lipid microdomains by targeting the biosynthesis of GSLs inhibits sperm activation and the inhibition can be rescued by the addition of exogenous GSLs. The cleavage of glycosylphosphatidylinositol (GPI)-anchored proteins, which are exclusively found in lipid microdomains, also affects sperm activation. We conclude that localized signaling mediated by lipid microdomains is critical for worm sperm activation. Lipid microdomains composed of cholesterol and GSLs have been observed in flagellated sperm of several animal species, thus cholesterol, before its efflux from the plasma membrane, might be needed to assemble into a platform for some more important upstream signal sorting during spermatogenesis than was previously thought. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The correct establishment and maintenance of cell polarity is essential for many biological processes, including asymmetric cell division, cellular morphogenesis and directed cell migration. Typically, actin cytoskeletons are employed to establish cell polarity and microtubules are involved in maintaining the stability of the polarized organization [1]. Nematode spermiogenesis (or sperm activation), in which round sessile spermatids differentiate into asymmetric crawling spermatozoa, is a symmetry-breaking process [2]. However, nematode spermatozoa possess neither actin nor tubulin; instead, the establishment of their polarity and cell motility depend on controlled assembly/disassembly of the major sperm protein (MSP) cytoskeleton [3,4]. After the meiosis, spermatids are transcriptionally and translationally silent and thus sperm activation, motility acquisition and fertilization are performed without new gene expression [5]. Sperm activation is initiated by vas deferens-derived trypsin-like serine protease, TRY-5 [6,7], that activates round sessile spermatid to
⁎ Corresponding author. Tel.: + 86 10 64888425; fax: + 86 10 64888478. E-mail address: [email protected] (L. Miao). 1388-1981/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2012.03.005
assemble the MSP cytoskeleton and to extend a single pseudopod, which is necessary for sperm motility and fertilization competence [8,9]. In addition, PI (3,4,5)P3 signaling on the plasma membrane (PM) is involved in C. elegans sperm activation [10]. The observations that an extracellular activator functions at or within the sperm PM, and that phospholipid in the PM is required during sperm activation, indicate that the lipid environment in the PM is crucial for signal transduction in this process. Whether a specific lipid environment in the PM is essential for sperm activation remains unknown. Increasing evidence indicates that the PM typically contains dynamic microdomains that are resistant to cold detergent extraction and are of lower buoyant density than the total PM. Such membrane microdomains on the cell surface play important roles in signal transduction during the establishment of cell polarity and in directed cell migration [11–13]. Collectively known as lipid microdomains (or membrane microdomains or lipid rafts) in the exoplasmic leaflet of the bilayer, these microdomains are enriched with cholesterol, glycosphingolipids (GSLs) and specific membrane proteins such as GPIanchored proteins [12]. The survival of C. elegans requires an exogenous cholesterol supply because it lacks the de novo biosynthetic pathway for cholesterol [14]. Animals deprived of cholesterol produce a reduced brood size in the second generation, and re-supplying
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
cholesterol restores their fertility [15]. The requirement for trace quantities of cholesterol for the survival of C. elegans and its relatively low abundance in lipid extracts suggests that cholesterol might serve as a signaling molecule rather than being a structural component in most cell membranes of this species [14]. Conversely, a live celllabeling assay with the fluorescent cholesterol analog dehydroergoesterol showed that C. elegans sperm are enriched with cholesterol [16], suggesting that cholesterol might be crucial for sperm function and reproductive success. However, it remains unknown whether cholesterol in the sperm plays a signaling role (e.g., as a precursor to a steroid hormone or other biologically active compound) or whether—in conjunction with GSLs—it might form lipid microdomains providing a platform mediating signal transduction during sperm activation and motility acquisition. Here we demonstrate that cholesterol is enriched in the plasma membrane of C. elegans spermatids and that the integrity of lipid microdomains composed of cholesterol and glycosphingolipids (GSLs) is critical during sperm activation. Disruption of sperm lipid microdomains by acute manipulation of cholesterol in vitro blocks the sperm activation. Restriction of cholesterol update also results in the abnormal sperm activation in both males and hermaphrodites. Manipulation of the integrity of lipid microdomains by targeting the biosynthesis of GSLs inhibits both the membranous organelles' fusion with PM and pseudopod extension, two events necessary for sperm activation. The inhibitory effects can be rescued by the addition of exogenous GSLs. The cleavage of glycosylphosphatidylinositol (GPI)-anchored proteins also affects sperm activation. Collectively, these findings indicate a functional role for lipid microdomains in mediating localized signal transduction during sperm activation and sperm motility acquisition in C. elegans. 2. Materials and methods 2.1. Nematode strains C. elegans strains, provided by the Caenorhabditis Genetics Center, were cultured on Nematode Growth Media (NGM) or cholesterol free NGM plates at 20 °C according to Brenner [17]. The strain him5(e1490)V served as the wild-type animal to facilitate the supply of males for detecting sperm function. The effect of cholesterol depletion on sperm function was performed by mating the him-5(e1490)V males cultured on cholesterol free NGM with normal N2 hermaphrodites or by growing the him-5 hermaphrodites on cholesterol free NGM.
935
Sialoglycosphingolipids, ganglioside GM3 or GM1 (Sigma-Aldrich) was applied in the presence of D-PDMP or NB-DNJ to test the inhibitor specificity during the sperm activation. 2.4. Detection of mating behavior and sperm transfer The cholesterol-depleted males were cultured on single plates and soaked in MitoTracker Red overnight before mating with N2 hermaphrodites for 6 h. The hermaphrodites were observed under fluorescent microscopy (Carl Zeiss). 2.5. Membrane organelle fusion assay FM1-43 (Invitrogen, Carlsbad, CA) was used to detect fusion of membrane organelles (MOs) with the plasma membrane of C. elegans sperm [18] after different treatments. Sperm were incubated with FM1-43 for 1 min, rinsed twice with SM buffer. DIC and rhodaminefiltered microscopy images were obtained as described above. 2.6. Immunofluorescence Sperm were fixed with 4% paraformaldehyde (EMS, Fort Washington, PA) and 0.1% glutaraldehyde (EMS) for 15 min, rinsed three times with phosphate buffered saline and then permeabilized with 0.3% Triton X-100 for 5 min. The fixed cells were blocked with 2% bovine serum albumin (Sigma-Aldrich) for 4 h to prevent non-specific binding. An antibody against UDP-Glucosyl Ceramide Glycotransferase (UGCG; 1:200; Santa Cruz Biotechnology, Santa Cruz, CA) or caveolin-1(1:100; Abgent, CA) was added and incubated overnight at 4 °C followed by staining with a secondary antibody conjugated with Alexa Fluor 488 (1:400; Invitrogen) for 2 h at room temperature. Images were acquired using a confocal microscope (Leica Microsystems, Heidelberg, Germany). 2.7. Large scale culture of C. elegans and preparation of sperm The large scale culture of him-5 worms was carried out according to a standard protocol [19]. C. elegans sperm cells were dissected from isolated male worms and stored at −80 °C. Sperm lysates were analyzed by SDS/PAGE and Western blotting with an antibody against UDP-Glucosyl Ceramide Glycotransferase (Santa Cruz Biotechnology, Santa Cruz, CA).
2.2. C. elegans sperm activation in vitro 2.8. Immuno electron microscopy Virgin males at the L4 stage were transferred to a new NGM plate to prevent mating for 2 days before the C. elegans sperm were dissected into sperm medium (SM) buffer containing monensin (200 nM; Sigma) for 10 min. Sperm activation and migration were recorded in real time at 20 °C using an Axio Imager M2 microscope (Carl Zeiss), equipped with a 63× and 100× differential interference contrast (DIC) objective lenses with appropriate filters. Images were captured with a charge-coupled device (CCD; Andor Technology, Belfast, Northern Ireland) and processed with MetaMorph software (Universal Imaging, Downington, PA) 2.3. Disruption of the integrity of lipid microdomains in C. elegans sperm To extract cholesterol from the sperm, different concentrations of methyl-β-cyclodextrin (MCD, Sigma-Aldrich, St. Louis, MO) were added into the SM buffer for 10 min along with monensin (200 nM). Filipin (Sigma-Aldrich) was used to bind the cholesterol and thus decrease its dynamic distribution. D-threo-1-Phenyl-2decanoylamino-3-morpholino-1-propanol (D-PDMP; Sigma-Aldrich) and N-butyldeoxynojirimycin (NB-DNJ; Sigma-Aldrich) were applied to inhibit the synthesis of glycosphingolipids in C. elegans sperm.
The him-5 male animals at L4 stage were picked to new plates to grow until young adults. The sperm were dissected into the SM buffer or the SM buffer containing monensin on Thermanox Coverslips (Electron Microscopy Sciences). The cells were allowed to adhere to the coverslip for 15 min before being fixed with 4% paraformaldehyde and 0.2% glutaraldehyde overnight at 4 °C. The fixed sperm were rinsed with SM several times before serial ethanol dehydration at 4 °C. The samples were infiltrated with serial LR-White resin (Electron Microscopy Sciences), and LR-White polymerization was carried out under UV irradiation for 48 h at −20 °C. Areas of the plastic wafer containing well preserved sperm were excised and remounted with superglue onto plastic stubs. Ultrathin (~80 nm) sections were cut on Leica ultramicrotome and transferred to 200-mesh grids. Grids containing these sections were blocked with 20% goat serum and 0.5% Tween in PBS for 1 h, followed by incubation with caveolin-1 antibody (Abgent, 1:25) for 2 h. After rinsing, the grids were transferred to droplets of 10 nm immunogold conjugated goat anti-rabbit secondary antibody (Sigma; 1:20). The grids were stained with 2% uranyl acetate for 10 min before observation on an FEI Tecnai Spirit transmission electron microscope, operating at 120 kV.
936
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
3. Results 3.1. Cholesterol is required for C. elegans sperm activation in vitro and in vivo The low abundance of cholesterol in nematode lipid extracts suggests that there is insufficient cholesterol to play a major structural role in nematode membranes [14]. However, lipid microdomains possessing similar physical characteristics to those observed in other organisms were detected in the nervous system of C. elegans [20]. These lipid microdomains are composed of cholesterol, sphingolipids and the proteins UNC-1, UNC-8 and UNC-24, suggesting that lipid microdomains might be assembled in a small subset of cells in C. elegans. A live cell-labeling assay with the fluorescent cholesterol analog dehydroergoesterol showed that C. elegans sperm are enriched with cholesterol [16]. By staining with filipin, which is highly fluorescent and binds specifically to cholesterol, we showed cholesterol was enriched in the PM of non-activated sperm (spermatids) (Fig. 1, top paired images). The enrichment of cholesterol on the PM of spermatids implies that lipid microdomains might exist in spermatids and function during sperm activation. To test whether cholesterol in C. elegans spermatids might be assembled into lipid microdomains necessary for sperm function, we manipulated the integrity of the microdomains using methyl-β-cyclodextrin (MCD), a water-soluble cyclic oligomer that extracts cholesterol from the PM [21]. When spermatids dissected from males were incubated in medium containing the cationic ionophore monensin, an artificial activator [22], the motile spermatozoa with pseudopods were generated after 10 min (Fig. 1B). In contrast, when a range of concentrations of MCD was included with monensin, the sperm activation rate in the system decreased in a dose-dependent manner (Fig. 1B and C). At 25 mM MCD, only 4.6% of sperm were activated, compared with 60.4% activation in untreated controls (P b 0.05). At 27.5 mM MCD, monensininduced sperm activation was totally inhibited and the cells retained their round symmetrical shape. However, when the sperm were treated with a mixture of 27.5 mM MCD and 2 mM cholesterol, the cells could be activated successfully (Fig. 1B). Fluorescent staining with filipin indicates that MCD incubation largely depleted membrane cholesterol from the sperm (Fig. 1D). To confirm that the inhibition of monensin-induced sperm activation was not an artifact of membrane cholesterol depletion by MCD, we also disrupted lipid microdomains using filipin, another lipid microdomain-disrupting agent, which is structurally different from MCD and sequestrates but not depletes cholesterol from the membrane [21,23]. Monensin-induced sperm activation was consistently inhibited by treatment with 400 μg/mL filipin for 10 min (Fig. 1E). Cholesterol efflux from the plasma membrane is required for mammalian sperm capacitation in the female reproductive tract. Interestingly, though the nematode sperm are crawling cells whose motility is totally different from that in flagellated sperm found in mammals, once they are fully activated, the cholesterol level on the PM of spermatozoa decreased dramatically, esp. in the leading edge (Fig.1, A, bottom paired images), as
reported during mammalian sperm capacitation [24,25]. These data suggest that cholesterol efflux might have broad phylogenetic conservation for sperm activation and male reproductive success. However, our data further indicate that the cholesterol on the PM of spermatids as a component of lipid microdomain is required for the initial signal transduction events during sperm activation before the cholesterol efflux occurs, though we cannot exclude the possibility that cholesterol might also function as a signal precursor inside the sperm cell during the sperm activation. Mature C. elegans spermatids contain specialized membranous organelles (MOs), which is a type of intracellular vesicle with similarity to lysosomes [26]. The fusion of MOs with the PM during sperm activation is necessary for spermatozoan motility and male fertility in C. elegans [18,27] and this process is known to be a regulated exocytosis. Upon spermatid activation, the heads of the MOs fuse with the PM and their contents are released into the extracellular environment, leaving a permanent invaginated pore on the PM around the cell body periphery. Using the lipophilic dye FM143 that stains the outer leaflet of the PM [18,28], we were able to observe fusion of MOs with the PM (bright fluorescent puncta around the periphery of the cell body) in mock-treated sperm (Fig. 1F, left paired images). The MO fusion was obviously inhibited when the sperm were treated with MCD (Fig. 1F, middle paired images) or filipin (Fig. 1F, right paired images). These data indicate that MO fusion with PM during the sperm activation is dependent on the signal transduction mediated by the lipid microdomains on the sperm membrane. To further confirm our in vitro acute cholesterol depletion data, we cultured him-5 worm on the cholesterol free NGM for several generations. Animals deprived of cholesterol produced a reduced brood size in the second generation (F2) as reported [15], although the dissected spermatids from F2 males still could be activated partially in vitro by monensin (Fig. 1G). Most of the spermatids from F3 males cultured on cholesterol free NGM could not be activated in vitro by monensin and retained their round symmetrical shape (Fig. 1G). We further tested whether the sperm from F3 males grown on cholesterol free NGM could be activated by the endogenous activators in males and/or hermaphrodite by mating F3 males grown on cholesterol free NGM with normal N2 hermaphrodites. Unlike mock-depletion control where mitotracker-labeled male sperm were activated upon ejaculation and migrated into the spermatheca after mating (Fig. 1H, upper panels), the sperm from F3 males grown on cholesterol free NGM were found in the whole uterus of N2 hermaphrodite, some near the vulva (Fig. 1H, lower panels). These data indicate that the spermatids from F3 males grown on cholesterol free NGM are insensitive to either male-derived activators or hermaphrodite-derived activators, thus fail to respond to the chemotaxis signal derived from the oocyte [29]. In C. elegans, sperm are made in both males and self-fertile hermaphrodites. Hermaphrodites produce sperm during the fourth larval stage (L4) and produce oocytes as adults. Ovulation of the first oocyte pushes the stored spermatids from the gonad into the spermatheca. Once in the spermatheca, they are rapidly activated
Fig. 1. Depletion of cholesterol from the plasma membrane inhibits C. elegans sperm activation. (A) DIC (left) and fluorescence (right) images of cells labeled with filipin for visualizing membrane cholesterol in spermatids (top) or spermatozoon (bottom). Bar = 5 μm. (B) MCD treatment inhibited C. elegans sperm activation. Cells were treated with monensin and a range of concentrations of MCD for 10 min. The cells with pseudopods (indicated by arrows) are spermatozoa and the round ones are spermatids. Bar = 5 μm. (C) Quantification of sperm activation as shown in (B). Results represent the mean ± standard deviation (SD; n = 3 experiments). *P b 0.05; **P b 0.001. (D) Treatment with MCD decreased the amounts of cholesterol indicated with filipin. C. elegans sperm were dissected in sperm medium (SM) alone or SM containing 27.5 mM MCD for 30 min before being fixed with 4% paraformaldehyde for 1 h. The fixed cells were stained with filipin for 2 h. Bars = 5 μm. (E) Pretreatment with filipin for 10 min abolished monensin-induced sperm activation. Bar = 5 μm. (F) MO fusion was inhibited obviously when the sperm were treated with MCD and filipin. Bars = 5 μm. (G) The activation of sperm from males in F3 progeny cultured on cholesterol free NGM plates was inhibited. Spermatids were dissected from males and treated with monensin. Control, sperm from him-5 males cultured on normal NGM plates. Bar = 5 μm. (H) The sperm transfer assay showed that sperm from F3 males cultured on cholesterol free NGM plates had motility defects when mating to hermaphrodites. Upper: N2 hermaphrodites mated with normal him-5 males that were labeled with MitoTracker Red; Lower: N2 hermaphrodites mated with F3 him-5 males cultured on cholesterol free NGM plates and labeled with MitoTracker Red. Bar = 50 μm. (I) The Sperm from hermaphrodites cultured on cholesterol free NGM plates showed motility defects. Upper: hermaphrodites cultured on normal NGM plates (left) and their sperm dissected (right); Lower: hermaphrodites cultured on cholesterol free NGM plates (left) and their sperm dissected (right). The arrows indicated the sperm accumulated in spermatheca. Bar in left = 25 μm, bar in right = 10 μm.
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
into spermatozoa. We found that most of the dissected sperm from F3 hermpharodites grown on cholesterol free NGM were non-activated and retained their round shape (Fig. 1I, lower right), while sperm dissected from control hermaphrodites were motile with clear pseudopod extensions (Fig. 1I, upper right). Compared to control hermaphrodites (Fig. 1I, upper left), F3 hermaphrodites grown on cholesterol free
937
NGM contained obviously less sperm (Fig. 1I, lower left), probably because the ovulation pushed the non-activated spermatids out of spermatheca and due to lack of motility they could not craw back to the spermatheca. These data provide direct evidence that cholesterol depletion also affects self sperm activation triggered by endogenous activator (s) in hermaphrodite.
938
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
Fig. 2. Cleavage of GPI-anchored proteins inhibits C. elegans sperm activation. (A) Monensin-induced sperm activation was inhibited by pretreatment of spermatids with 25 U/mL PI-PLC for 60 min (right). Control pretreatment (SM) had no inhibitory effects on sperm activation (left). (B) Quantification of sperm activation as shown in A. Results represent the mean ± SD (n = 3 experiments). **P b 0.001. (C) Pretreatment of spermatids with PI-PLC inhibited fusion of MOs with the PM. After treatments, sperm were stained with FM 1–43 for 1 min to visualize PM and MOs upon fusion. MO fusion is indicated by the bright puncta (arrows) in the cell body of controls (left paired images). In spermatids pretreated with 25 U/mL PI-PLC, only the PM is stained. Bars = 5 μm.
Due to their size being below the classical diffraction limit of a light microscope, lipid microdomains ranging from 10–200 nm in size have proved difficult to visualize directly. Caveolae, small surface invaginations seen in many cell types, are specialized lipid microdomains that perform a number of signalling functions. The permanent invaginated pore was formed (Fig. S1E) when MO fused with the PM during sperm activation, suggesting that caveolae-like structure might be involved for signal sorting and functions. To test this hypothesis, we employed indirect immunofluorescence microscopy and immuno-EM to examine the localization of caveolin-1 (CAV-1), a widely used marker to indicate the lipid microdomains. As shown in Fig. S1A and B, CAV-1 could be detected in both spermatid and spermatozoon by immunofluorescence microscopy. CAV-1 was localized in the subcellular compartments in the spermatid (Fig. S1A), while in the spermatozoon, the CAV1 was clustered and enriched in the periphery of cell body, where MO fusion took place. Subcellular compartment of CAV-1 was further demonstrated using immuno-EM. Immuno-gold was exclusively localized in the non-fused or fused MOs in both spermatid (Fig. S1C) and spermatozoon (Fig. S1D). Intriguingly, the distribution pattern of the CAV-1 (Fig. S1B) is similar to that of the cholesterol in spermatozoon (Fig. 1, A, bottom paired images), indicating that lipid microdomains composed of cholesterol and CAV-1 might function during sperm activation. Taken together, these results suggest that cholesterol in the spermatids is required for sperm activation in C. elegans, and that it participates in this process mainly as a structural component of lipid microdomains. Currently it is believed that cholesterol functions primarily in cell signaling in C. elegans, possibly as a precursor to steroid hormone(s) or to other physiologically active compounds, and instead of a structural component of the PM. This is based on the following observations: first, cholesterol constitutes only about 1% of the total lipid content of embryo extracts; second, tiny amounts of cholesterol are able to restore the development of cholesterol-depleted arrested larvae into fertile adults. Our data suggest that cholesterol functions as a structural component of lipid microdomains in C. elegans spermatids and that their integrity is important for sperm activation. Because cholesterol and the integrity of lipid microdomains seem to be essential for C. elegans sperm activation (Fig. 1), we next examined whether other lipid microdomain components,
glycosylphosphatidylinositol (GPI) -anchored proteins and glycosphingolipids (GSLs) are involved in sperm activation. 3.2. GPI-anchored proteins are required for C. elegans sperm activation In eukaryotic cells, some cell surface proteins are anchored to the exoplasmic face of the PM via covalent attachment of a glycosylphosphatidylinositol (GPI) anchor. GPI-anchored proteins, with their unique biochemical and biophysical characteristics during the signal transduction, associate with glycosphingolipids and cholesterol to form detergent-resistant lipid microdomains in the outer leaflet of the PM. GPI-anchored proteins are often cross-linked and such crosslinking may stabilize small microdomains and, by coalescence with each other, could generate larger lipid microdomains involved in signaling [30]. The GPI anchor can be cleaved by phosphatidylinositol-specific phospholipase C (PI-PLC) [31], thus releasing the protein from the membrane surface in a soluble form. If GPI-anchored proteins are essential to signal transduction during sperm activation, cleaving of these proteins by PI-PLC would result in the failure of sperm activation. This is indeed what we observed (Fig. 2A and B). Only 6.9 % of sperm were activated by monensin if the spermatids were first incubated with 25 U/mL PI-PLC for 60 min, significantly lower than that in the mock treatment (56.0%, P b 0.001, Fig. 2B). The activated sperm subjected to the PI-PLC treatment also showed an aberrant activation phenotype: i.e., the spermatids initiated sperm activation but failed to extend normal-looking pseudopods (Fig. 2A). We also tested whether cleavage of GPI-anchored protein inhibits MO fusion during sperm activation. We were able to observe fusion of MOs with the PM (bright fluorescent puncta around the periphery of the cell body) in mock-treated sperm (Fig. 2C, left paired images). When sperm were incubated with PI-PLC, followed by monensin treatment, the fusion of MOs with the PM was dramatically inhibited and fluorescence was only detected in an even distribution around the PM of the sperm (Fig. 2C, right paired images). This indicates that GPI-anchored proteins on the exoplasmic leaflet of sperm are required for the localized signal transduction to initiate membrane fusion during sperm activation. Taken together, our results indicated that GPI-anchored proteins in the lipid microdomains are required
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
for C. elegans sperm activation; these proteins participate in both MO fusion and assembly of the MSP cytoskeleton during this process.
939
[34,35]. Strong UGCG signal was detected in C. elegans sperm when fused with GFP [36]. As shown by immunoblot (Fig. S2A) and immunofluorescence microscopy (Fig. S2B), UGCG was detected in the isolated sperm.
3.3. Glycosphingolipid (GSL) biosynthesis is essential for C. elegans sperm activation and sperm migration 3.3.2. Inhibition of GSL synthesis blocks C. elegans sperm activation 3.3.1. UDP-glucose ceramide glucosyltransferase (UGCG) is detected in C. elegans sperm GSLs comprise a group of membrane components containing sugar and lipid moieties. GSLs are thought to be vertically translocated from cytoplasmic leaflet to the exoplasmic leaflet of the PM by unidentified GSL-specific flippase [32]. The localization of GSLs on the exoplasmic leaflet is important in the assembly of signaling molecules and GPIanchored proteins into lipid microdomains during signal transduction [33]. The precursor of most GSLs, glucosylceramide, is synthesized from ceramide and catalyzed by UGCG (UDP-glucose ceramide glucosyltransferase or ceramide glucosyltransferase—CGT—in C. elegans)
The content of GSLs was reduced in cells treated with D-threo-1Phenyl-2-decanoylamino-3-morpholino-1-propanol (D-PDMP), a preferential substrate of UGCG. In the presence of D-PDMP, sphingomyelins rather than GSLs were synthesized by UGCG [37]. Incubation of C. elegans spermatids with D-PDMP blocked both pseudopod formation and membrane fusion during monensin-induced sperm activation (Fig. 3A–C). At 0.3 mM and 0.4 mM D-PDMP, only 8.8% and 6.9% of cells extended pseudopods, respectively; significantly lower than in controls (39.9%, P b 0.05). At 0.4 mM D-PDMP, more than 33% of the cells were arrested at the spiky stage, an intermediate phase of sperm activation in which MSP bundles project, but fail to
Fig. 3. Inhibition of GSL biosynthesis blocked pseudopod formation and membrane fusion of C. elegans sperm while exogenous GSLs rescued these processes. (A) D-PDMP inhibited C. elegans sperm activation. Cells were treated with monensin and a range of concentrations of D-PDMP for 10 min. The white arrow indicates a spike. The rightmost panel shows that D-PDMP exhibited no cytotoxicity to C. elegans sperm. Activation was rescued when the cells were first treated with 1 mM D-PDMP and monensin for 10 min and then perfused with monensin. (B) Quantitative analysis of pseudopod or spike formation in C. elegans sperm as shown in (A). Black bars show cells with pseudopods; grey bars show cells with spikes. The values are the mean ± SD (n = 3 experiments). *P b 0.05; **P b 0.001. (C) MO fusion with the PM was inhibited by D-PDMP in a dose-dependent manner. The cells were stained with FM1-43 for 1 min before observation. (D) Exogenous GSLs rescued the D-PDMP-inhibited sperm activation. Cells were treated with monensin, 1 mM D-PDMP + monensin, 1 mM D-PDMP + 0.15 mM GM1 + monensin, 1 mM D-PDMP + 0.15 mM sialidase-treated GM1 + monensin, or 1 mM D-PDMP + 0.15 mM GM3 + monensin. (E) Exogenous GSLs rescued the D-PDMP-inhibited sperm migration. The time-course of sperm migration after perfusion with D-PDMP or GM1 in the presence of D-PDMP was recorded. The images were extracted from the movie S2. (F) Quantitative analysis of sperm migration in (E). Data represent the mean ± SD (n = 3 experiments). **P b 0.001. Bars = 5 μm.
940
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
coalesce into a pseudopod. Pseudopod extension was totally blocked by 0.5 mM D-PDMP, while 1 mM blocked both pseudopod and spike extension. We infer that the inhibition of sperm activation by DPDMP was caused by decreased GSLs and not by cellular toxicity, because the inhibition at 1 mM D-PDMP could be rescued by monensin perfusion, leading to pseudopod extension (Fig. 3A, rightmost panel). MO fusion during sperm activation was inhibited by D-PDMP in a dose-dependent manner and 1 mM D-PDMP treatment totally blocked fusion of MOs with the PM, as shown by FM 1–43 staining (Fig. 3C). We also found that cells treated with another UGCG inhibitor, N-butyldeoxynojirimycin (NB-DNJ) [38], manifested defects in pseudopod extension (Fig. S3A, B) and in MO fusion with the PM (Fig. S3C) in a manner similar to D-PDMP treatment during sperm activation. These results demonstrate that GSLs are crucial for C. elegans sperm activation, and that they function in both pseudopod formation and membrane fusion during this process. When monensin-activated spermatozoa were perfused with DPDMP, the MSP cytoskeleton in pseudopods disassembled and the leading edge retracted toward the cell body, resulting in rounding of the spermatozoa. Crawling capability was restored when D-PDMP was washed out (Movie S1), further suggesting that the biosynthesis of GSLs play a role in localized MSP cytoskeleton assembly in C. elegans sperm.
3.3.3. Exogenous GSLs rescue blockage including pseudopod formation and membrane fusion When exogenous GSLs (sialoglycosphingolipid, ganglioside GM3 or GM1) were applied, we observed that they were able to rescue the inhibitory effects of D-PDMP on sperm activation (Fig. 3D). Whatever the order of application of D-PDMP, GM1 or monensin, exogenous
GM1 could rescue the inhibitory effects of D-PDMP on membrane fusion and pseudopod extension during monensin-induced sperm activation (Fig. 4A). Neither pseudopods nor MO fusion could be observed in controls where exogenous GM1 was absent (Fig. 4A). Exogenous GSLs also rescued sperm activation inhibited by NB-DNJ (Fig. S3A, B). In addition, exogenous GM1 in the presence of D-PDMP rescued D-PDMPinhibited sperm migration (Fig. 3E, F and Movie S2). To confirm that the exogenous GSLs were incorporated into the sperm PM, sperm were incubated with bodipy-GM1 for 10 min. The bright fluorescent signal of bodipy-GM1 was detected around the periphery (Fig. 4B, left paired images). Like the cholesterol staining patterns in both spermatids and spermatozoa, the bodipy-GM1 was enriched in the spermatid PM with a relatively evenly distribution pattern, while in the spermatozoon the highest fluorescent signal was detected in the periphery of cell body though with detectable signal in the leading edge PM of spermatozoon (Fig. 4B, left paired images). In the presence of 10-fold unlabeled GM1, however, the fluorescent intensity of bodipy-GM1 in sperm decreased dramatically (Fig. 4B, right paired images), indicating that GSLs had been incorporated into the sperm surface. Interestingly, when exogenous GM1 was applied in the absence of D-PDMP, the cells were activated aberrantly by monensin and failed to extend normal-looking pseudopods (Fig. 4C). It is possible that the addition of exogenous GM1 excluded other lipid microdomain-associated proteins thus blocking correct signal transduction during sperm activation. Thus, a balance of GSLs might be important for maintaining the integrity and stability of lipid microdomains required for localized signal modulation. Because the sialic acid synthesis pathway is absent in C. elegans [39], we used sialidase to hydrolyze terminal sialyl residues from GM1. Sialidase-treated GM1 rescued D-PDMP-inhibited sperm activation similarly to GM1 and GM3 (Fig. 3D), indicating that the lipid moiety rather than the sugar moiety of GSLs was responsible for C. elegans
Fig. 4. Exogenous GM1 could be integrated into C. elegans sperm membrane and rescued the MO fusion and sperm activation inhibited by D-PDMP. (A) The activation of C. elegans sperm inhibited by D-PDMP was rescued by supplementation with GM1. MO fusion was indicated by bright puncta of FM1-43 fluorescence. (B) The exogenous GM1 was integrated into the sperm membrane. Left paired images: monensin-activated C. elegans spermatozoa were incubated with 0.03 mM bodipy-GM1 for 5 min and then rinsed with SM buffer. Right paired images: monensin-activated C. elegans spermatozoa were incubated with 0.3 mM GM1 for 5 min then incubated with 0.03 mM Bodipy-GM1 and finally rinsed with SM. (C) The addition of excess GM1 alone inhibited monensin-induced sperm activation. Bars = 5 μm.
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
sperm activation. These in vitro data collectively suggest that GSLs, as components of lipid microdomains, play essential roles in C. elegans sperm activation and MSP-based cell motility. 4. Discussion Selective disruption of the integrity of lipid microdomains by acute manipulation of cholesterol, using MCD or filipin, resulted in the failure of sperm activation, indicating that the assembly of lipid microdomains incorporating cholesterol plays an essential role in localized signal transduction during sperm activation. The sperm from either males or hermaphrodites cultured on cholesterol free medium could not be activated by in vitro activator monensin or endogenous activator, further indicating that cholesterol play an important role during sperm activation. The importance of lipid microdomains in mediating sperm activation was further supported by the evidence that other lipid microdomain components, GSLs and GPI-anchored proteins, were required during sperm activation. This suggests that the lipid microenvironment—rather than cholesterol itself—is required for localized signal transduction during sperm activation. Exogenous application of GSLs in the presence of D-PDMP rescued the inhibition of sperm activation, while the exogenous supply of GSLs to the sperm in the absence of D-PDMP caused abnormal activation. This implies that the stability and integrity of lipid microdomains are required for sperm activation, presumably because the incorporation of excess GSLs into the microdomains excludes other microdomain-associated proteins. We further demonstrated that GPI-anchored proteins are required for sperm function. Extracellular signaling induces multiple changes, such as membrane rearrangements, membrane fusion, exocytosis and MSP cytoskeleton dynamics, which initiate the protrusion of the leading edge and the generation of the pseudopod during nematode sperm activation. Our results showed that the integrity of lipid microdomains was required for localized signal transduction during MO fusion with the PM, formation of spikes and pseudopod extension. Intriguingly, the fer-1 mutation in C. elegans exhibits failure of MO fusion with the PM, sperm motility defects and consequent male sterility [18]. In the human homolog of FER-1, the protein dysferlin, containing repeated C2 domains characteristic of synaptotagmin, is required for vesicle fusion during Ca 2 +-induced muscle membrane repair. Dysferlin is believed to promote vesicle fusion through interactions with SNARE proteins, a complex localized in lipid microdomains [40]. Furthermore, dysferlin was shown to interact with caveolin-3, another marker for lipid microdomains [41]. Thus, lipid microdomains and its associated components in the membrane bilayer might play critical roles on membrane fusion during sperm activation. Our data from the parasitic nematode Ascaris suum showed that a trypsin-like serine protease secreted by the vas deferens (As_TRY-5) is an endogenous sperm activator [6]. This protease shares a high degree of homology, including a conserved catalytic triad with a channel-activating serine protease (CAP or prostasin). Thus, it is plausible that As_TRY-5 initiates sperm activation through proteolytic cleavage and activation of its receptor, an epithelial sodium channel (ENaC)-like molecule on the PM, resulting in intracellular pH increase. This might act as a second messenger that relays and amplifies the signal to convert the spermatids to spermatozoa. Interestingly, ENaC was shown to be present in and regulated by lipid microdomains [42,43]. Ejaculated mammalian sperm must acquire fertilization capacity after residing into the female reproductive tract, a process collectively known as capacitation. The cholesterol present in sperm membranes was shown to rigidify and stabilize membranes, unfavorable for sperm capacitation [44]. Cholesterol efflux from the sperm membrane, an obligatory step in the sperm capacitation process, might change the biophysical properties of sperm membrane leading to an increase in the membrane fluidity towards a more fusogenic and
941
flexible state. Notably, the change of cholesterol/phospholipid (C/ PL) ratio in the PM has been observed during the sperm capacitation. A relatively high C/PL ratio is present in freshly ejaculated sperm. During capacitation, cholesterol efflux from the sperm membrane is accelerated by cholesterol acceptors in the female reproductive tract, leading to a decrease of C/PL ratio. Simultaneously, phospholipid moves into the sperm membrane, further decreases the C/PL ratio. The similar C/PL ratio change might also occur during the nematode sperm activation and motility acquisition since the leading edge of spermatozoa showed less cholesterol staining by filipin (Fig. 1A, bottom paired images). However, before this occurs, the cholesterol, GSLs and GPI-anchored proteins assembling into lipid microdomain is required in mediating initial localized signal transduction during sperm activation and sperm motility acquisition. The C/PL ratio decreases should be the downstream events during the signal transduction in sperm activation. Lipid rafts were isolated and their reorganization was observed in mammalian flagellated sperm [45,46]. An interesting avenue for future investigation may be to determine whether the initial signaling mediated by the lipid microdomains is required for sperm maturation in other animal species besides C. elegans. Supplementary data to this article can be found online at doi:10. 1016/j.bbalip.2012.03.005. Acknowledgements We gratefully acknowledge Drs. Kexi Yi and Pingsheng Liu for critical reading of the manuscript; Miss Wei Zhuang for excellent technical assistance; other members of our lab for thoughtful comments and criticisms. This project was supported by grants 2010CB912303, 2012CB94502 and 30971648 (to LM) and 30900719 (to JD) from the Chinese government. LM is supported by the Chinese Academy of Sciences 100-talents program. References [1] R. Li, G.G. Gundersen, Beyond polymer polarity: how the cytoskeleton builds a polarized cell, Nat. Rev. Mol. Cell Biol. 9 (2008) 860–873. [2] S.W. L'Hernault, Spermatogenesis, WormBook (2006) 1–14. [3] T.M. Roberts, M. Stewart, Acting like actin. The dynamics of the nematode major sperm protein (msp) cytoskeleton indicate a push-pull mechanism for amoeboid cell motility, J. Cell Biol. 149 (2000) 7–12. [4] T.M. Roberts, Major sperm protein, Curr. Biol. 15 (2005) R153. [5] S. Grunewald, U. Paasch, H.J. Glander, U. Anderegg, Mature human spermatozoa do not transcribe novel RNA, Andrologia 37 (2005) 69–71. [6] Y. Zhao, W. Sun, P. Zhang, H. Chi, M.J. Zhang, C.Q. Song, X. Ma, Y. Shang, B. Wang, Y. Hu, Z. Hao, A.F. Huhmer, F. Meng, S.W. L'Hernault, S.M. He, M.Q. Dong, L. Miao, Nematode sperm maturation triggered by protease involves sperm-secreted serine protease inhibitor (Serpin), Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 1542–1547. [7] J.R. Smith, G.M. Stanfield, TRY-5 is a sperm-activating protease in Caenorhabditis elegans seminal fluid, PLoS Genet. 7 (2011) e1002375. [8] J.J. Fraire-Zamora, R.A. Cardullo, The physiological acquisition of amoeboid motility in nematode sperm: is the tail the only thing the sperm lost? Mol. Reprod. Dev. 77 (2010) 739–750. [9] S.W. L'Hernault, The genetics and cell biology of spermatogenesis in the nematode C. elegans, Mol. Cell. Endocrinol. 306 (2009) 59–65. [10] Y.K. Bae, E. Kim, S.W. L'Hernault, M.M. Barr, The CIL-1 PI 5-phosphatase localizes TRP polycystins to cilia and activates sperm in C. elegans, Curr. Biol. 19 (2009) 1599–1607. [11] S. Manes, A. Viola, Lipid rafts in lymphocyte activation and migration, Mol. Membr. Biol. 23 (2006) 59–69. [12] D. Lingwood, K. Simons, Lipid rafts as a membrane-organizing principle, Science 327 (2010) 46–50. [13] K. Simons, M.J. Gerl, Revitalizing membrane rafts: new tools and insights, Nat. Rev. Mol. Cell Biol. 11 (2010) 688–699. [14] T.V. Kurzchalia, S. Ward, Why do worms need cholesterol? Nat. Cell Biol. 5 (2003) 684–688. [15] M. Merris, W.G. Wadsworth, U. Khamrai, R. Bittman, D.J. Chitwood, J. Lenard, Sterol effects and sites of sterol accumulation in Caenorhabditis elegans: developmental requirement for 4alpha-methyl sterols, J. Lipid Res. 44 (2003) 172–181. [16] V. Matyash, C. Geier, A. Henske, S. Mukherjee, D. Hirsh, C. Thiele, B. Grant, F.R. Maxfield, T.V. Kurzchalia, Distribution and transport of cholesterol in Caenorhabditis elegans, Mol. Biol. Cell 12 (2001) 1725–1736. [17] S. Brenner, The genetics of Caenorhabditis elegans, Genetics 77 (1974) 71–94.
942
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
[18] N.L. Washington, S. Ward, FER-1 regulates Ca2+-mediated membrane fusion during C. elegans spermatogenesis, J. Cell Sci. 119 (2006) 2552–2562. [19] S.W. L'Hernault, T.M. Roberts, Cell biology of nematode sperm, Methods Cell Biol. 48 (1995) 273–301. [20] M.M. Sedensky, J.M. Siefker, J.Y. Koh, D.M. Miller III, P.G. Morgan, A stomatin and a degenerin interact in lipid rafts of the nervous system of Caenorhabditis elegans, Am. J. Physiol. Cell Physiol. 287 (2004) C468–C474. [21] K. Simons, D. Toomre, Lipid rafts and signal transduction, Nat. Rev. Mol. Cell Biol. 1 (2000) 31–39. [22] G.A. Nelson, S. Ward, Vesicle fusion, pseudopod extension and amoeboid motility are induced in nematode spermatids by the ionophore monensin, Cell 19 (1980) 457–464. [23] J.M. Robinson, M.J. Karnovsky, Evaluation of the polyene antibiotic filipin as a cytochemical probe for membrane cholesterol, J. Histochem. Cytochem. 28 (1980) 161–168. [24] F.M. Flesch, J.F. Brouwers, P.F. Nievelstein, A.J. Verkleij, L.M. van Golde, B. Colenbrander, B.M. Gadella, Bicarbonate stimulated phospholipid scrambling induces cholesterol redistribution and enables cholesterol depletion in the sperm plasma membrane, J. Cell Sci. 114 (2001) 3543–3555. [25] S. Shadan, P.S. James, E.A. Howes, R. Jones, Cholesterol efflux alters lipid raft stability and distribution during capacitation of boar spermatozoa, Biol. Reprod. 71 (2004) 253–265. [26] G.D. Zhu, G. Salazar, S.A. Zlatic, B. Fiza, M.M. Doucette, C.J. Heilman, A.I. Levey, V. Faundez, S.W. L'Hernault, SPE-39 family proteins interact with the HOPS complex and function in lysosomal delivery, Mol. Biol. Cell 20 (2009) 1223–1240. [27] S. Ward, Y. Argon, G.A. Nelson, Sperm morphogenesis in wild-type and fertilizationdefective mutants of Caenorhabditis elegans, J. Cell Biol. 91 (1981) 26–44. [28] W.E. Achanzar, S. Ward, A nematode gene required for sperm vesicle fusion, J. Cell Sci. 110 (Pt 9) (1997) 1073–1081. [29] J.W. Edmonds, J.K. Prasain, D. Dorand, Y. Yang, H.D. Hoang, J. Vibbert, H.M. Kubagawa, M.A. Miller, Insulin/FOXO signaling regulates ovarian prostaglandins critical for reproduction, Dev. Cell 19 (2010) 858–871. [30] T. Friedrichson, T.V. Kurzchalia, Microdomains of GPI-anchored proteins in living cells revealed by crosslinking, Nature 394 (1998) 802–805. [31] R.A. Mortara, L.M. Minelli, F. Vandekerckhove, V. Nussenzweig, F.J. RamalhoPinto, Phosphatidylinositol-specific phospholipase C (PI-PLC) cleavage of GPIanchored surface molecules of Trypanosoma cruzi triggers in vitro morphological reorganization of trypomastigotes, J. Eukaryot. Microbiol. 48 (2001) 27–37. [32] G. van Meer, D.R. Voelker, G.W. Feigenson, Membrane lipids: where they are and how they behave, Nat. Rev. Mol. Cell Biol. 9 (2008) 112–124.
[33] S. Hakomori, Bifunctional role of glycosphingolipids. Modulators for transmembrane signaling and mediators for cellular interactions, J. Biol. Chem. 265 (1990) 18713–18716. [34] S. Ichikawa, K. Ozawa, Y. Hirabayashi, Assignment of a UDP-glucose:ceramide glucosyltransferase gene (UGCG) to human chromosome band 9q31 by in situ hybridization, Cytogenet. Cell Genet. 79 (1997) 233–234. [35] H. Zhang, N. Abraham, L.A. Khan, D.H. Hall, J.T. Fleming, V. Gobel, Apicobasal domain identities of expanding tubular membranes depend on glycosphingolipid biosynthesis, Nat. Cell Biol. 13 (2012) 1189–1201. [36] K.H. Nomura, D. Murata, Y. Hayashi, K. Dejima, S. Mizuguchi, E. Kage-Nakadai, K. Gengyo-Ando, S. Mitani, Y. Hirabayashi, M. Ito, K. Nomura, Ceramide glucosyltransferase of the nematode Caenorhabditis elegans is involved in oocyte formation and in early embryonic cell division, Glycobiology 21 (2011) 834–848. [37] N.S. Radin, J.A. Shayman, J. Inokuchi, Metabolic effects of inhibiting glucosylceramide synthesis with PDMP and other substances, Adv. Lipid Res. 26 (1993) 183–213. [38] M. Guerrera, S. Ladisch, N-butyldeoxynojirimycin inhibits murine melanoma cell ganglioside metabolism and delays tumor onset, Cancer Lett. 201 (2003) 31–40. [39] A. Bacic, I. Kahane, B.M. Zuckerman, Panagrellus redivivus and Caenorhabditis elegans: evidence for the absence of sialic acids, Exp. Parasitol. 71 (1990) 483–488. [40] T. Lang, SNARE proteins and ‘membrane rafts’, J. Physiol. 585 (2007) 693–698. [41] C. Matsuda, Y.K. Hayashi, M. Ogawa, M. Aoki, K. Murayama, I. Nishino, I. Nonaka, K. Arahata, R.H. Brown Jr., The sarcolemmal proteins dysferlin and caveolin-3 interact in skeletal muscle, Hum. Mol. Genet. 10 (2001) 1761–1766. [42] W.G. Hill, B. An, J.P. Johnson, Endogenously expressed epithelial sodium channel is present in lipid rafts in A6 cells, J. Biol. Chem. 277 (2002) 33541–33544. [43] W.G. Hill, M.B. Butterworth, H. Wang, R.S. Edinger, J. Lebowitz, K.W. Peters, R.A. Frizzell, J.P. Johnson, The epithelial sodium channel (ENaC) traffics to apical membrane in lipid rafts in mouse cortical collecting duct cells, J. Biol. Chem. 282 (2007) 37402–37411. [44] E. deLamirande, P. Leclerc, C. Gagnon, Capacitation as a regulatory event that primes spermatozoa for the acrosome reaction and fertilization, Mol. Hum. Reprod. 3 (1997) 175–194. [45] 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. [46] N.L. Cross, Reorganization of lipid rafts during capacitation of human sperm, Biol. Reprod. 71 (2004) 1367–1373.
Contents lists available at SciVerse ScienceDirect
Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbalip
Cholesterol and the biosynthesis of glycosphingolipids are required for sperm activation in Caenorhabditis elegans Jiangli Dou, Lianwan Chen, Youqiao Hu, Long Miao ⁎ Laboratory of Noncoding RNA, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District, Beijing 100101, China
a r t i c l e
i n f o
Article history: Received 24 September 2011 Received in revised form 28 February 2012 Accepted 16 March 2012 Available online 28 March 2012 Keywords: Caenorhabditis elegans Major sperm protein (MSP) Sperm activation Sperm motility Lipid microdomains
a b s t r a c t Ejaculated mammalian sperm must acquire fertilization capacity after residing into the female reproductive tract, a process collectively known as capacitation. Cholesterol efflux was required for sperm maturation. Different from flagellated sperm, C. elegans sperm are crawling cells. C. elegans sperm are highly enriched with cholesterol though this animal species lacks biosynthetic pathway for cholesterol and its survival requires an exogenous cholesterol supply. The low abundance of cholesterol in C. elegans lipid extract is thought insufficient to form lipid microdomains ubiquitously in this organism. We present evidence that cholesterol is enriched in the plasma membrane of C. elegans spermatids and that cholesterol- and glycosphingolipids (GSLs)-enriched membrane microdomains (lipid microdomains) mediate sperm activation. Disruption of sperm lipid microdomains by acute manipulation of cholesterol in vitro blocks the sperm activation. Restriction of cholesterol uptake also results in the abnormal sperm activation in both males and hermaphrodites. Manipulation of the integrity of lipid microdomains by targeting the biosynthesis of GSLs inhibits sperm activation and the inhibition can be rescued by the addition of exogenous GSLs. The cleavage of glycosylphosphatidylinositol (GPI)-anchored proteins, which are exclusively found in lipid microdomains, also affects sperm activation. We conclude that localized signaling mediated by lipid microdomains is critical for worm sperm activation. Lipid microdomains composed of cholesterol and GSLs have been observed in flagellated sperm of several animal species, thus cholesterol, before its efflux from the plasma membrane, might be needed to assemble into a platform for some more important upstream signal sorting during spermatogenesis than was previously thought. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The correct establishment and maintenance of cell polarity is essential for many biological processes, including asymmetric cell division, cellular morphogenesis and directed cell migration. Typically, actin cytoskeletons are employed to establish cell polarity and microtubules are involved in maintaining the stability of the polarized organization [1]. Nematode spermiogenesis (or sperm activation), in which round sessile spermatids differentiate into asymmetric crawling spermatozoa, is a symmetry-breaking process [2]. However, nematode spermatozoa possess neither actin nor tubulin; instead, the establishment of their polarity and cell motility depend on controlled assembly/disassembly of the major sperm protein (MSP) cytoskeleton [3,4]. After the meiosis, spermatids are transcriptionally and translationally silent and thus sperm activation, motility acquisition and fertilization are performed without new gene expression [5]. Sperm activation is initiated by vas deferens-derived trypsin-like serine protease, TRY-5 [6,7], that activates round sessile spermatid to
⁎ Corresponding author. Tel.: + 86 10 64888425; fax: + 86 10 64888478. E-mail address: [email protected] (L. Miao). 1388-1981/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2012.03.005
assemble the MSP cytoskeleton and to extend a single pseudopod, which is necessary for sperm motility and fertilization competence [8,9]. In addition, PI (3,4,5)P3 signaling on the plasma membrane (PM) is involved in C. elegans sperm activation [10]. The observations that an extracellular activator functions at or within the sperm PM, and that phospholipid in the PM is required during sperm activation, indicate that the lipid environment in the PM is crucial for signal transduction in this process. Whether a specific lipid environment in the PM is essential for sperm activation remains unknown. Increasing evidence indicates that the PM typically contains dynamic microdomains that are resistant to cold detergent extraction and are of lower buoyant density than the total PM. Such membrane microdomains on the cell surface play important roles in signal transduction during the establishment of cell polarity and in directed cell migration [11–13]. Collectively known as lipid microdomains (or membrane microdomains or lipid rafts) in the exoplasmic leaflet of the bilayer, these microdomains are enriched with cholesterol, glycosphingolipids (GSLs) and specific membrane proteins such as GPIanchored proteins [12]. The survival of C. elegans requires an exogenous cholesterol supply because it lacks the de novo biosynthetic pathway for cholesterol [14]. Animals deprived of cholesterol produce a reduced brood size in the second generation, and re-supplying
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
cholesterol restores their fertility [15]. The requirement for trace quantities of cholesterol for the survival of C. elegans and its relatively low abundance in lipid extracts suggests that cholesterol might serve as a signaling molecule rather than being a structural component in most cell membranes of this species [14]. Conversely, a live celllabeling assay with the fluorescent cholesterol analog dehydroergoesterol showed that C. elegans sperm are enriched with cholesterol [16], suggesting that cholesterol might be crucial for sperm function and reproductive success. However, it remains unknown whether cholesterol in the sperm plays a signaling role (e.g., as a precursor to a steroid hormone or other biologically active compound) or whether—in conjunction with GSLs—it might form lipid microdomains providing a platform mediating signal transduction during sperm activation and motility acquisition. Here we demonstrate that cholesterol is enriched in the plasma membrane of C. elegans spermatids and that the integrity of lipid microdomains composed of cholesterol and glycosphingolipids (GSLs) is critical during sperm activation. Disruption of sperm lipid microdomains by acute manipulation of cholesterol in vitro blocks the sperm activation. Restriction of cholesterol update also results in the abnormal sperm activation in both males and hermaphrodites. Manipulation of the integrity of lipid microdomains by targeting the biosynthesis of GSLs inhibits both the membranous organelles' fusion with PM and pseudopod extension, two events necessary for sperm activation. The inhibitory effects can be rescued by the addition of exogenous GSLs. The cleavage of glycosylphosphatidylinositol (GPI)-anchored proteins also affects sperm activation. Collectively, these findings indicate a functional role for lipid microdomains in mediating localized signal transduction during sperm activation and sperm motility acquisition in C. elegans. 2. Materials and methods 2.1. Nematode strains C. elegans strains, provided by the Caenorhabditis Genetics Center, were cultured on Nematode Growth Media (NGM) or cholesterol free NGM plates at 20 °C according to Brenner [17]. The strain him5(e1490)V served as the wild-type animal to facilitate the supply of males for detecting sperm function. The effect of cholesterol depletion on sperm function was performed by mating the him-5(e1490)V males cultured on cholesterol free NGM with normal N2 hermaphrodites or by growing the him-5 hermaphrodites on cholesterol free NGM.
935
Sialoglycosphingolipids, ganglioside GM3 or GM1 (Sigma-Aldrich) was applied in the presence of D-PDMP or NB-DNJ to test the inhibitor specificity during the sperm activation. 2.4. Detection of mating behavior and sperm transfer The cholesterol-depleted males were cultured on single plates and soaked in MitoTracker Red overnight before mating with N2 hermaphrodites for 6 h. The hermaphrodites were observed under fluorescent microscopy (Carl Zeiss). 2.5. Membrane organelle fusion assay FM1-43 (Invitrogen, Carlsbad, CA) was used to detect fusion of membrane organelles (MOs) with the plasma membrane of C. elegans sperm [18] after different treatments. Sperm were incubated with FM1-43 for 1 min, rinsed twice with SM buffer. DIC and rhodaminefiltered microscopy images were obtained as described above. 2.6. Immunofluorescence Sperm were fixed with 4% paraformaldehyde (EMS, Fort Washington, PA) and 0.1% glutaraldehyde (EMS) for 15 min, rinsed three times with phosphate buffered saline and then permeabilized with 0.3% Triton X-100 for 5 min. The fixed cells were blocked with 2% bovine serum albumin (Sigma-Aldrich) for 4 h to prevent non-specific binding. An antibody against UDP-Glucosyl Ceramide Glycotransferase (UGCG; 1:200; Santa Cruz Biotechnology, Santa Cruz, CA) or caveolin-1(1:100; Abgent, CA) was added and incubated overnight at 4 °C followed by staining with a secondary antibody conjugated with Alexa Fluor 488 (1:400; Invitrogen) for 2 h at room temperature. Images were acquired using a confocal microscope (Leica Microsystems, Heidelberg, Germany). 2.7. Large scale culture of C. elegans and preparation of sperm The large scale culture of him-5 worms was carried out according to a standard protocol [19]. C. elegans sperm cells were dissected from isolated male worms and stored at −80 °C. Sperm lysates were analyzed by SDS/PAGE and Western blotting with an antibody against UDP-Glucosyl Ceramide Glycotransferase (Santa Cruz Biotechnology, Santa Cruz, CA).
2.2. C. elegans sperm activation in vitro 2.8. Immuno electron microscopy Virgin males at the L4 stage were transferred to a new NGM plate to prevent mating for 2 days before the C. elegans sperm were dissected into sperm medium (SM) buffer containing monensin (200 nM; Sigma) for 10 min. Sperm activation and migration were recorded in real time at 20 °C using an Axio Imager M2 microscope (Carl Zeiss), equipped with a 63× and 100× differential interference contrast (DIC) objective lenses with appropriate filters. Images were captured with a charge-coupled device (CCD; Andor Technology, Belfast, Northern Ireland) and processed with MetaMorph software (Universal Imaging, Downington, PA) 2.3. Disruption of the integrity of lipid microdomains in C. elegans sperm To extract cholesterol from the sperm, different concentrations of methyl-β-cyclodextrin (MCD, Sigma-Aldrich, St. Louis, MO) were added into the SM buffer for 10 min along with monensin (200 nM). Filipin (Sigma-Aldrich) was used to bind the cholesterol and thus decrease its dynamic distribution. D-threo-1-Phenyl-2decanoylamino-3-morpholino-1-propanol (D-PDMP; Sigma-Aldrich) and N-butyldeoxynojirimycin (NB-DNJ; Sigma-Aldrich) were applied to inhibit the synthesis of glycosphingolipids in C. elegans sperm.
The him-5 male animals at L4 stage were picked to new plates to grow until young adults. The sperm were dissected into the SM buffer or the SM buffer containing monensin on Thermanox Coverslips (Electron Microscopy Sciences). The cells were allowed to adhere to the coverslip for 15 min before being fixed with 4% paraformaldehyde and 0.2% glutaraldehyde overnight at 4 °C. The fixed sperm were rinsed with SM several times before serial ethanol dehydration at 4 °C. The samples were infiltrated with serial LR-White resin (Electron Microscopy Sciences), and LR-White polymerization was carried out under UV irradiation for 48 h at −20 °C. Areas of the plastic wafer containing well preserved sperm were excised and remounted with superglue onto plastic stubs. Ultrathin (~80 nm) sections were cut on Leica ultramicrotome and transferred to 200-mesh grids. Grids containing these sections were blocked with 20% goat serum and 0.5% Tween in PBS for 1 h, followed by incubation with caveolin-1 antibody (Abgent, 1:25) for 2 h. After rinsing, the grids were transferred to droplets of 10 nm immunogold conjugated goat anti-rabbit secondary antibody (Sigma; 1:20). The grids were stained with 2% uranyl acetate for 10 min before observation on an FEI Tecnai Spirit transmission electron microscope, operating at 120 kV.
936
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
3. Results 3.1. Cholesterol is required for C. elegans sperm activation in vitro and in vivo The low abundance of cholesterol in nematode lipid extracts suggests that there is insufficient cholesterol to play a major structural role in nematode membranes [14]. However, lipid microdomains possessing similar physical characteristics to those observed in other organisms were detected in the nervous system of C. elegans [20]. These lipid microdomains are composed of cholesterol, sphingolipids and the proteins UNC-1, UNC-8 and UNC-24, suggesting that lipid microdomains might be assembled in a small subset of cells in C. elegans. A live cell-labeling assay with the fluorescent cholesterol analog dehydroergoesterol showed that C. elegans sperm are enriched with cholesterol [16]. By staining with filipin, which is highly fluorescent and binds specifically to cholesterol, we showed cholesterol was enriched in the PM of non-activated sperm (spermatids) (Fig. 1, top paired images). The enrichment of cholesterol on the PM of spermatids implies that lipid microdomains might exist in spermatids and function during sperm activation. To test whether cholesterol in C. elegans spermatids might be assembled into lipid microdomains necessary for sperm function, we manipulated the integrity of the microdomains using methyl-β-cyclodextrin (MCD), a water-soluble cyclic oligomer that extracts cholesterol from the PM [21]. When spermatids dissected from males were incubated in medium containing the cationic ionophore monensin, an artificial activator [22], the motile spermatozoa with pseudopods were generated after 10 min (Fig. 1B). In contrast, when a range of concentrations of MCD was included with monensin, the sperm activation rate in the system decreased in a dose-dependent manner (Fig. 1B and C). At 25 mM MCD, only 4.6% of sperm were activated, compared with 60.4% activation in untreated controls (P b 0.05). At 27.5 mM MCD, monensininduced sperm activation was totally inhibited and the cells retained their round symmetrical shape. However, when the sperm were treated with a mixture of 27.5 mM MCD and 2 mM cholesterol, the cells could be activated successfully (Fig. 1B). Fluorescent staining with filipin indicates that MCD incubation largely depleted membrane cholesterol from the sperm (Fig. 1D). To confirm that the inhibition of monensin-induced sperm activation was not an artifact of membrane cholesterol depletion by MCD, we also disrupted lipid microdomains using filipin, another lipid microdomain-disrupting agent, which is structurally different from MCD and sequestrates but not depletes cholesterol from the membrane [21,23]. Monensin-induced sperm activation was consistently inhibited by treatment with 400 μg/mL filipin for 10 min (Fig. 1E). Cholesterol efflux from the plasma membrane is required for mammalian sperm capacitation in the female reproductive tract. Interestingly, though the nematode sperm are crawling cells whose motility is totally different from that in flagellated sperm found in mammals, once they are fully activated, the cholesterol level on the PM of spermatozoa decreased dramatically, esp. in the leading edge (Fig.1, A, bottom paired images), as
reported during mammalian sperm capacitation [24,25]. These data suggest that cholesterol efflux might have broad phylogenetic conservation for sperm activation and male reproductive success. However, our data further indicate that the cholesterol on the PM of spermatids as a component of lipid microdomain is required for the initial signal transduction events during sperm activation before the cholesterol efflux occurs, though we cannot exclude the possibility that cholesterol might also function as a signal precursor inside the sperm cell during the sperm activation. Mature C. elegans spermatids contain specialized membranous organelles (MOs), which is a type of intracellular vesicle with similarity to lysosomes [26]. The fusion of MOs with the PM during sperm activation is necessary for spermatozoan motility and male fertility in C. elegans [18,27] and this process is known to be a regulated exocytosis. Upon spermatid activation, the heads of the MOs fuse with the PM and their contents are released into the extracellular environment, leaving a permanent invaginated pore on the PM around the cell body periphery. Using the lipophilic dye FM143 that stains the outer leaflet of the PM [18,28], we were able to observe fusion of MOs with the PM (bright fluorescent puncta around the periphery of the cell body) in mock-treated sperm (Fig. 1F, left paired images). The MO fusion was obviously inhibited when the sperm were treated with MCD (Fig. 1F, middle paired images) or filipin (Fig. 1F, right paired images). These data indicate that MO fusion with PM during the sperm activation is dependent on the signal transduction mediated by the lipid microdomains on the sperm membrane. To further confirm our in vitro acute cholesterol depletion data, we cultured him-5 worm on the cholesterol free NGM for several generations. Animals deprived of cholesterol produced a reduced brood size in the second generation (F2) as reported [15], although the dissected spermatids from F2 males still could be activated partially in vitro by monensin (Fig. 1G). Most of the spermatids from F3 males cultured on cholesterol free NGM could not be activated in vitro by monensin and retained their round symmetrical shape (Fig. 1G). We further tested whether the sperm from F3 males grown on cholesterol free NGM could be activated by the endogenous activators in males and/or hermaphrodite by mating F3 males grown on cholesterol free NGM with normal N2 hermaphrodites. Unlike mock-depletion control where mitotracker-labeled male sperm were activated upon ejaculation and migrated into the spermatheca after mating (Fig. 1H, upper panels), the sperm from F3 males grown on cholesterol free NGM were found in the whole uterus of N2 hermaphrodite, some near the vulva (Fig. 1H, lower panels). These data indicate that the spermatids from F3 males grown on cholesterol free NGM are insensitive to either male-derived activators or hermaphrodite-derived activators, thus fail to respond to the chemotaxis signal derived from the oocyte [29]. In C. elegans, sperm are made in both males and self-fertile hermaphrodites. Hermaphrodites produce sperm during the fourth larval stage (L4) and produce oocytes as adults. Ovulation of the first oocyte pushes the stored spermatids from the gonad into the spermatheca. Once in the spermatheca, they are rapidly activated
Fig. 1. Depletion of cholesterol from the plasma membrane inhibits C. elegans sperm activation. (A) DIC (left) and fluorescence (right) images of cells labeled with filipin for visualizing membrane cholesterol in spermatids (top) or spermatozoon (bottom). Bar = 5 μm. (B) MCD treatment inhibited C. elegans sperm activation. Cells were treated with monensin and a range of concentrations of MCD for 10 min. The cells with pseudopods (indicated by arrows) are spermatozoa and the round ones are spermatids. Bar = 5 μm. (C) Quantification of sperm activation as shown in (B). Results represent the mean ± standard deviation (SD; n = 3 experiments). *P b 0.05; **P b 0.001. (D) Treatment with MCD decreased the amounts of cholesterol indicated with filipin. C. elegans sperm were dissected in sperm medium (SM) alone or SM containing 27.5 mM MCD for 30 min before being fixed with 4% paraformaldehyde for 1 h. The fixed cells were stained with filipin for 2 h. Bars = 5 μm. (E) Pretreatment with filipin for 10 min abolished monensin-induced sperm activation. Bar = 5 μm. (F) MO fusion was inhibited obviously when the sperm were treated with MCD and filipin. Bars = 5 μm. (G) The activation of sperm from males in F3 progeny cultured on cholesterol free NGM plates was inhibited. Spermatids were dissected from males and treated with monensin. Control, sperm from him-5 males cultured on normal NGM plates. Bar = 5 μm. (H) The sperm transfer assay showed that sperm from F3 males cultured on cholesterol free NGM plates had motility defects when mating to hermaphrodites. Upper: N2 hermaphrodites mated with normal him-5 males that were labeled with MitoTracker Red; Lower: N2 hermaphrodites mated with F3 him-5 males cultured on cholesterol free NGM plates and labeled with MitoTracker Red. Bar = 50 μm. (I) The Sperm from hermaphrodites cultured on cholesterol free NGM plates showed motility defects. Upper: hermaphrodites cultured on normal NGM plates (left) and their sperm dissected (right); Lower: hermaphrodites cultured on cholesterol free NGM plates (left) and their sperm dissected (right). The arrows indicated the sperm accumulated in spermatheca. Bar in left = 25 μm, bar in right = 10 μm.
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
into spermatozoa. We found that most of the dissected sperm from F3 hermpharodites grown on cholesterol free NGM were non-activated and retained their round shape (Fig. 1I, lower right), while sperm dissected from control hermaphrodites were motile with clear pseudopod extensions (Fig. 1I, upper right). Compared to control hermaphrodites (Fig. 1I, upper left), F3 hermaphrodites grown on cholesterol free
937
NGM contained obviously less sperm (Fig. 1I, lower left), probably because the ovulation pushed the non-activated spermatids out of spermatheca and due to lack of motility they could not craw back to the spermatheca. These data provide direct evidence that cholesterol depletion also affects self sperm activation triggered by endogenous activator (s) in hermaphrodite.
938
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
Fig. 2. Cleavage of GPI-anchored proteins inhibits C. elegans sperm activation. (A) Monensin-induced sperm activation was inhibited by pretreatment of spermatids with 25 U/mL PI-PLC for 60 min (right). Control pretreatment (SM) had no inhibitory effects on sperm activation (left). (B) Quantification of sperm activation as shown in A. Results represent the mean ± SD (n = 3 experiments). **P b 0.001. (C) Pretreatment of spermatids with PI-PLC inhibited fusion of MOs with the PM. After treatments, sperm were stained with FM 1–43 for 1 min to visualize PM and MOs upon fusion. MO fusion is indicated by the bright puncta (arrows) in the cell body of controls (left paired images). In spermatids pretreated with 25 U/mL PI-PLC, only the PM is stained. Bars = 5 μm.
Due to their size being below the classical diffraction limit of a light microscope, lipid microdomains ranging from 10–200 nm in size have proved difficult to visualize directly. Caveolae, small surface invaginations seen in many cell types, are specialized lipid microdomains that perform a number of signalling functions. The permanent invaginated pore was formed (Fig. S1E) when MO fused with the PM during sperm activation, suggesting that caveolae-like structure might be involved for signal sorting and functions. To test this hypothesis, we employed indirect immunofluorescence microscopy and immuno-EM to examine the localization of caveolin-1 (CAV-1), a widely used marker to indicate the lipid microdomains. As shown in Fig. S1A and B, CAV-1 could be detected in both spermatid and spermatozoon by immunofluorescence microscopy. CAV-1 was localized in the subcellular compartments in the spermatid (Fig. S1A), while in the spermatozoon, the CAV1 was clustered and enriched in the periphery of cell body, where MO fusion took place. Subcellular compartment of CAV-1 was further demonstrated using immuno-EM. Immuno-gold was exclusively localized in the non-fused or fused MOs in both spermatid (Fig. S1C) and spermatozoon (Fig. S1D). Intriguingly, the distribution pattern of the CAV-1 (Fig. S1B) is similar to that of the cholesterol in spermatozoon (Fig. 1, A, bottom paired images), indicating that lipid microdomains composed of cholesterol and CAV-1 might function during sperm activation. Taken together, these results suggest that cholesterol in the spermatids is required for sperm activation in C. elegans, and that it participates in this process mainly as a structural component of lipid microdomains. Currently it is believed that cholesterol functions primarily in cell signaling in C. elegans, possibly as a precursor to steroid hormone(s) or to other physiologically active compounds, and instead of a structural component of the PM. This is based on the following observations: first, cholesterol constitutes only about 1% of the total lipid content of embryo extracts; second, tiny amounts of cholesterol are able to restore the development of cholesterol-depleted arrested larvae into fertile adults. Our data suggest that cholesterol functions as a structural component of lipid microdomains in C. elegans spermatids and that their integrity is important for sperm activation. Because cholesterol and the integrity of lipid microdomains seem to be essential for C. elegans sperm activation (Fig. 1), we next examined whether other lipid microdomain components,
glycosylphosphatidylinositol (GPI) -anchored proteins and glycosphingolipids (GSLs) are involved in sperm activation. 3.2. GPI-anchored proteins are required for C. elegans sperm activation In eukaryotic cells, some cell surface proteins are anchored to the exoplasmic face of the PM via covalent attachment of a glycosylphosphatidylinositol (GPI) anchor. GPI-anchored proteins, with their unique biochemical and biophysical characteristics during the signal transduction, associate with glycosphingolipids and cholesterol to form detergent-resistant lipid microdomains in the outer leaflet of the PM. GPI-anchored proteins are often cross-linked and such crosslinking may stabilize small microdomains and, by coalescence with each other, could generate larger lipid microdomains involved in signaling [30]. The GPI anchor can be cleaved by phosphatidylinositol-specific phospholipase C (PI-PLC) [31], thus releasing the protein from the membrane surface in a soluble form. If GPI-anchored proteins are essential to signal transduction during sperm activation, cleaving of these proteins by PI-PLC would result in the failure of sperm activation. This is indeed what we observed (Fig. 2A and B). Only 6.9 % of sperm were activated by monensin if the spermatids were first incubated with 25 U/mL PI-PLC for 60 min, significantly lower than that in the mock treatment (56.0%, P b 0.001, Fig. 2B). The activated sperm subjected to the PI-PLC treatment also showed an aberrant activation phenotype: i.e., the spermatids initiated sperm activation but failed to extend normal-looking pseudopods (Fig. 2A). We also tested whether cleavage of GPI-anchored protein inhibits MO fusion during sperm activation. We were able to observe fusion of MOs with the PM (bright fluorescent puncta around the periphery of the cell body) in mock-treated sperm (Fig. 2C, left paired images). When sperm were incubated with PI-PLC, followed by monensin treatment, the fusion of MOs with the PM was dramatically inhibited and fluorescence was only detected in an even distribution around the PM of the sperm (Fig. 2C, right paired images). This indicates that GPI-anchored proteins on the exoplasmic leaflet of sperm are required for the localized signal transduction to initiate membrane fusion during sperm activation. Taken together, our results indicated that GPI-anchored proteins in the lipid microdomains are required
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
for C. elegans sperm activation; these proteins participate in both MO fusion and assembly of the MSP cytoskeleton during this process.
939
[34,35]. Strong UGCG signal was detected in C. elegans sperm when fused with GFP [36]. As shown by immunoblot (Fig. S2A) and immunofluorescence microscopy (Fig. S2B), UGCG was detected in the isolated sperm.
3.3. Glycosphingolipid (GSL) biosynthesis is essential for C. elegans sperm activation and sperm migration 3.3.2. Inhibition of GSL synthesis blocks C. elegans sperm activation 3.3.1. UDP-glucose ceramide glucosyltransferase (UGCG) is detected in C. elegans sperm GSLs comprise a group of membrane components containing sugar and lipid moieties. GSLs are thought to be vertically translocated from cytoplasmic leaflet to the exoplasmic leaflet of the PM by unidentified GSL-specific flippase [32]. The localization of GSLs on the exoplasmic leaflet is important in the assembly of signaling molecules and GPIanchored proteins into lipid microdomains during signal transduction [33]. The precursor of most GSLs, glucosylceramide, is synthesized from ceramide and catalyzed by UGCG (UDP-glucose ceramide glucosyltransferase or ceramide glucosyltransferase—CGT—in C. elegans)
The content of GSLs was reduced in cells treated with D-threo-1Phenyl-2-decanoylamino-3-morpholino-1-propanol (D-PDMP), a preferential substrate of UGCG. In the presence of D-PDMP, sphingomyelins rather than GSLs were synthesized by UGCG [37]. Incubation of C. elegans spermatids with D-PDMP blocked both pseudopod formation and membrane fusion during monensin-induced sperm activation (Fig. 3A–C). At 0.3 mM and 0.4 mM D-PDMP, only 8.8% and 6.9% of cells extended pseudopods, respectively; significantly lower than in controls (39.9%, P b 0.05). At 0.4 mM D-PDMP, more than 33% of the cells were arrested at the spiky stage, an intermediate phase of sperm activation in which MSP bundles project, but fail to
Fig. 3. Inhibition of GSL biosynthesis blocked pseudopod formation and membrane fusion of C. elegans sperm while exogenous GSLs rescued these processes. (A) D-PDMP inhibited C. elegans sperm activation. Cells were treated with monensin and a range of concentrations of D-PDMP for 10 min. The white arrow indicates a spike. The rightmost panel shows that D-PDMP exhibited no cytotoxicity to C. elegans sperm. Activation was rescued when the cells were first treated with 1 mM D-PDMP and monensin for 10 min and then perfused with monensin. (B) Quantitative analysis of pseudopod or spike formation in C. elegans sperm as shown in (A). Black bars show cells with pseudopods; grey bars show cells with spikes. The values are the mean ± SD (n = 3 experiments). *P b 0.05; **P b 0.001. (C) MO fusion with the PM was inhibited by D-PDMP in a dose-dependent manner. The cells were stained with FM1-43 for 1 min before observation. (D) Exogenous GSLs rescued the D-PDMP-inhibited sperm activation. Cells were treated with monensin, 1 mM D-PDMP + monensin, 1 mM D-PDMP + 0.15 mM GM1 + monensin, 1 mM D-PDMP + 0.15 mM sialidase-treated GM1 + monensin, or 1 mM D-PDMP + 0.15 mM GM3 + monensin. (E) Exogenous GSLs rescued the D-PDMP-inhibited sperm migration. The time-course of sperm migration after perfusion with D-PDMP or GM1 in the presence of D-PDMP was recorded. The images were extracted from the movie S2. (F) Quantitative analysis of sperm migration in (E). Data represent the mean ± SD (n = 3 experiments). **P b 0.001. Bars = 5 μm.
940
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
coalesce into a pseudopod. Pseudopod extension was totally blocked by 0.5 mM D-PDMP, while 1 mM blocked both pseudopod and spike extension. We infer that the inhibition of sperm activation by DPDMP was caused by decreased GSLs and not by cellular toxicity, because the inhibition at 1 mM D-PDMP could be rescued by monensin perfusion, leading to pseudopod extension (Fig. 3A, rightmost panel). MO fusion during sperm activation was inhibited by D-PDMP in a dose-dependent manner and 1 mM D-PDMP treatment totally blocked fusion of MOs with the PM, as shown by FM 1–43 staining (Fig. 3C). We also found that cells treated with another UGCG inhibitor, N-butyldeoxynojirimycin (NB-DNJ) [38], manifested defects in pseudopod extension (Fig. S3A, B) and in MO fusion with the PM (Fig. S3C) in a manner similar to D-PDMP treatment during sperm activation. These results demonstrate that GSLs are crucial for C. elegans sperm activation, and that they function in both pseudopod formation and membrane fusion during this process. When monensin-activated spermatozoa were perfused with DPDMP, the MSP cytoskeleton in pseudopods disassembled and the leading edge retracted toward the cell body, resulting in rounding of the spermatozoa. Crawling capability was restored when D-PDMP was washed out (Movie S1), further suggesting that the biosynthesis of GSLs play a role in localized MSP cytoskeleton assembly in C. elegans sperm.
3.3.3. Exogenous GSLs rescue blockage including pseudopod formation and membrane fusion When exogenous GSLs (sialoglycosphingolipid, ganglioside GM3 or GM1) were applied, we observed that they were able to rescue the inhibitory effects of D-PDMP on sperm activation (Fig. 3D). Whatever the order of application of D-PDMP, GM1 or monensin, exogenous
GM1 could rescue the inhibitory effects of D-PDMP on membrane fusion and pseudopod extension during monensin-induced sperm activation (Fig. 4A). Neither pseudopods nor MO fusion could be observed in controls where exogenous GM1 was absent (Fig. 4A). Exogenous GSLs also rescued sperm activation inhibited by NB-DNJ (Fig. S3A, B). In addition, exogenous GM1 in the presence of D-PDMP rescued D-PDMPinhibited sperm migration (Fig. 3E, F and Movie S2). To confirm that the exogenous GSLs were incorporated into the sperm PM, sperm were incubated with bodipy-GM1 for 10 min. The bright fluorescent signal of bodipy-GM1 was detected around the periphery (Fig. 4B, left paired images). Like the cholesterol staining patterns in both spermatids and spermatozoa, the bodipy-GM1 was enriched in the spermatid PM with a relatively evenly distribution pattern, while in the spermatozoon the highest fluorescent signal was detected in the periphery of cell body though with detectable signal in the leading edge PM of spermatozoon (Fig. 4B, left paired images). In the presence of 10-fold unlabeled GM1, however, the fluorescent intensity of bodipy-GM1 in sperm decreased dramatically (Fig. 4B, right paired images), indicating that GSLs had been incorporated into the sperm surface. Interestingly, when exogenous GM1 was applied in the absence of D-PDMP, the cells were activated aberrantly by monensin and failed to extend normal-looking pseudopods (Fig. 4C). It is possible that the addition of exogenous GM1 excluded other lipid microdomain-associated proteins thus blocking correct signal transduction during sperm activation. Thus, a balance of GSLs might be important for maintaining the integrity and stability of lipid microdomains required for localized signal modulation. Because the sialic acid synthesis pathway is absent in C. elegans [39], we used sialidase to hydrolyze terminal sialyl residues from GM1. Sialidase-treated GM1 rescued D-PDMP-inhibited sperm activation similarly to GM1 and GM3 (Fig. 3D), indicating that the lipid moiety rather than the sugar moiety of GSLs was responsible for C. elegans
Fig. 4. Exogenous GM1 could be integrated into C. elegans sperm membrane and rescued the MO fusion and sperm activation inhibited by D-PDMP. (A) The activation of C. elegans sperm inhibited by D-PDMP was rescued by supplementation with GM1. MO fusion was indicated by bright puncta of FM1-43 fluorescence. (B) The exogenous GM1 was integrated into the sperm membrane. Left paired images: monensin-activated C. elegans spermatozoa were incubated with 0.03 mM bodipy-GM1 for 5 min and then rinsed with SM buffer. Right paired images: monensin-activated C. elegans spermatozoa were incubated with 0.3 mM GM1 for 5 min then incubated with 0.03 mM Bodipy-GM1 and finally rinsed with SM. (C) The addition of excess GM1 alone inhibited monensin-induced sperm activation. Bars = 5 μm.
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
sperm activation. These in vitro data collectively suggest that GSLs, as components of lipid microdomains, play essential roles in C. elegans sperm activation and MSP-based cell motility. 4. Discussion Selective disruption of the integrity of lipid microdomains by acute manipulation of cholesterol, using MCD or filipin, resulted in the failure of sperm activation, indicating that the assembly of lipid microdomains incorporating cholesterol plays an essential role in localized signal transduction during sperm activation. The sperm from either males or hermaphrodites cultured on cholesterol free medium could not be activated by in vitro activator monensin or endogenous activator, further indicating that cholesterol play an important role during sperm activation. The importance of lipid microdomains in mediating sperm activation was further supported by the evidence that other lipid microdomain components, GSLs and GPI-anchored proteins, were required during sperm activation. This suggests that the lipid microenvironment—rather than cholesterol itself—is required for localized signal transduction during sperm activation. Exogenous application of GSLs in the presence of D-PDMP rescued the inhibition of sperm activation, while the exogenous supply of GSLs to the sperm in the absence of D-PDMP caused abnormal activation. This implies that the stability and integrity of lipid microdomains are required for sperm activation, presumably because the incorporation of excess GSLs into the microdomains excludes other microdomain-associated proteins. We further demonstrated that GPI-anchored proteins are required for sperm function. Extracellular signaling induces multiple changes, such as membrane rearrangements, membrane fusion, exocytosis and MSP cytoskeleton dynamics, which initiate the protrusion of the leading edge and the generation of the pseudopod during nematode sperm activation. Our results showed that the integrity of lipid microdomains was required for localized signal transduction during MO fusion with the PM, formation of spikes and pseudopod extension. Intriguingly, the fer-1 mutation in C. elegans exhibits failure of MO fusion with the PM, sperm motility defects and consequent male sterility [18]. In the human homolog of FER-1, the protein dysferlin, containing repeated C2 domains characteristic of synaptotagmin, is required for vesicle fusion during Ca 2 +-induced muscle membrane repair. Dysferlin is believed to promote vesicle fusion through interactions with SNARE proteins, a complex localized in lipid microdomains [40]. Furthermore, dysferlin was shown to interact with caveolin-3, another marker for lipid microdomains [41]. Thus, lipid microdomains and its associated components in the membrane bilayer might play critical roles on membrane fusion during sperm activation. Our data from the parasitic nematode Ascaris suum showed that a trypsin-like serine protease secreted by the vas deferens (As_TRY-5) is an endogenous sperm activator [6]. This protease shares a high degree of homology, including a conserved catalytic triad with a channel-activating serine protease (CAP or prostasin). Thus, it is plausible that As_TRY-5 initiates sperm activation through proteolytic cleavage and activation of its receptor, an epithelial sodium channel (ENaC)-like molecule on the PM, resulting in intracellular pH increase. This might act as a second messenger that relays and amplifies the signal to convert the spermatids to spermatozoa. Interestingly, ENaC was shown to be present in and regulated by lipid microdomains [42,43]. Ejaculated mammalian sperm must acquire fertilization capacity after residing into the female reproductive tract, a process collectively known as capacitation. The cholesterol present in sperm membranes was shown to rigidify and stabilize membranes, unfavorable for sperm capacitation [44]. Cholesterol efflux from the sperm membrane, an obligatory step in the sperm capacitation process, might change the biophysical properties of sperm membrane leading to an increase in the membrane fluidity towards a more fusogenic and
941
flexible state. Notably, the change of cholesterol/phospholipid (C/ PL) ratio in the PM has been observed during the sperm capacitation. A relatively high C/PL ratio is present in freshly ejaculated sperm. During capacitation, cholesterol efflux from the sperm membrane is accelerated by cholesterol acceptors in the female reproductive tract, leading to a decrease of C/PL ratio. Simultaneously, phospholipid moves into the sperm membrane, further decreases the C/PL ratio. The similar C/PL ratio change might also occur during the nematode sperm activation and motility acquisition since the leading edge of spermatozoa showed less cholesterol staining by filipin (Fig. 1A, bottom paired images). However, before this occurs, the cholesterol, GSLs and GPI-anchored proteins assembling into lipid microdomain is required in mediating initial localized signal transduction during sperm activation and sperm motility acquisition. The C/PL ratio decreases should be the downstream events during the signal transduction in sperm activation. Lipid rafts were isolated and their reorganization was observed in mammalian flagellated sperm [45,46]. An interesting avenue for future investigation may be to determine whether the initial signaling mediated by the lipid microdomains is required for sperm maturation in other animal species besides C. elegans. Supplementary data to this article can be found online at doi:10. 1016/j.bbalip.2012.03.005. Acknowledgements We gratefully acknowledge Drs. Kexi Yi and Pingsheng Liu for critical reading of the manuscript; Miss Wei Zhuang for excellent technical assistance; other members of our lab for thoughtful comments and criticisms. This project was supported by grants 2010CB912303, 2012CB94502 and 30971648 (to LM) and 30900719 (to JD) from the Chinese government. LM is supported by the Chinese Academy of Sciences 100-talents program. References [1] R. Li, G.G. Gundersen, Beyond polymer polarity: how the cytoskeleton builds a polarized cell, Nat. Rev. Mol. Cell Biol. 9 (2008) 860–873. [2] S.W. L'Hernault, Spermatogenesis, WormBook (2006) 1–14. [3] T.M. Roberts, M. Stewart, Acting like actin. The dynamics of the nematode major sperm protein (msp) cytoskeleton indicate a push-pull mechanism for amoeboid cell motility, J. Cell Biol. 149 (2000) 7–12. [4] T.M. Roberts, Major sperm protein, Curr. Biol. 15 (2005) R153. [5] S. Grunewald, U. Paasch, H.J. Glander, U. Anderegg, Mature human spermatozoa do not transcribe novel RNA, Andrologia 37 (2005) 69–71. [6] Y. Zhao, W. Sun, P. Zhang, H. Chi, M.J. Zhang, C.Q. Song, X. Ma, Y. Shang, B. Wang, Y. Hu, Z. Hao, A.F. Huhmer, F. Meng, S.W. L'Hernault, S.M. He, M.Q. Dong, L. Miao, Nematode sperm maturation triggered by protease involves sperm-secreted serine protease inhibitor (Serpin), Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 1542–1547. [7] J.R. Smith, G.M. Stanfield, TRY-5 is a sperm-activating protease in Caenorhabditis elegans seminal fluid, PLoS Genet. 7 (2011) e1002375. [8] J.J. Fraire-Zamora, R.A. Cardullo, The physiological acquisition of amoeboid motility in nematode sperm: is the tail the only thing the sperm lost? Mol. Reprod. Dev. 77 (2010) 739–750. [9] S.W. L'Hernault, The genetics and cell biology of spermatogenesis in the nematode C. elegans, Mol. Cell. Endocrinol. 306 (2009) 59–65. [10] Y.K. Bae, E. Kim, S.W. L'Hernault, M.M. Barr, The CIL-1 PI 5-phosphatase localizes TRP polycystins to cilia and activates sperm in C. elegans, Curr. Biol. 19 (2009) 1599–1607. [11] S. Manes, A. Viola, Lipid rafts in lymphocyte activation and migration, Mol. Membr. Biol. 23 (2006) 59–69. [12] D. Lingwood, K. Simons, Lipid rafts as a membrane-organizing principle, Science 327 (2010) 46–50. [13] K. Simons, M.J. Gerl, Revitalizing membrane rafts: new tools and insights, Nat. Rev. Mol. Cell Biol. 11 (2010) 688–699. [14] T.V. Kurzchalia, S. Ward, Why do worms need cholesterol? Nat. Cell Biol. 5 (2003) 684–688. [15] M. Merris, W.G. Wadsworth, U. Khamrai, R. Bittman, D.J. Chitwood, J. Lenard, Sterol effects and sites of sterol accumulation in Caenorhabditis elegans: developmental requirement for 4alpha-methyl sterols, J. Lipid Res. 44 (2003) 172–181. [16] V. Matyash, C. Geier, A. Henske, S. Mukherjee, D. Hirsh, C. Thiele, B. Grant, F.R. Maxfield, T.V. Kurzchalia, Distribution and transport of cholesterol in Caenorhabditis elegans, Mol. Biol. Cell 12 (2001) 1725–1736. [17] S. Brenner, The genetics of Caenorhabditis elegans, Genetics 77 (1974) 71–94.
942
J. Dou et al. / Biochimica et Biophysica Acta 1821 (2012) 934–942
[18] N.L. Washington, S. Ward, FER-1 regulates Ca2+-mediated membrane fusion during C. elegans spermatogenesis, J. Cell Sci. 119 (2006) 2552–2562. [19] S.W. L'Hernault, T.M. Roberts, Cell biology of nematode sperm, Methods Cell Biol. 48 (1995) 273–301. [20] M.M. Sedensky, J.M. Siefker, J.Y. Koh, D.M. Miller III, P.G. Morgan, A stomatin and a degenerin interact in lipid rafts of the nervous system of Caenorhabditis elegans, Am. J. Physiol. Cell Physiol. 287 (2004) C468–C474. [21] K. Simons, D. Toomre, Lipid rafts and signal transduction, Nat. Rev. Mol. Cell Biol. 1 (2000) 31–39. [22] G.A. Nelson, S. Ward, Vesicle fusion, pseudopod extension and amoeboid motility are induced in nematode spermatids by the ionophore monensin, Cell 19 (1980) 457–464. [23] J.M. Robinson, M.J. Karnovsky, Evaluation of the polyene antibiotic filipin as a cytochemical probe for membrane cholesterol, J. Histochem. Cytochem. 28 (1980) 161–168. [24] F.M. Flesch, J.F. Brouwers, P.F. Nievelstein, A.J. Verkleij, L.M. van Golde, B. Colenbrander, B.M. Gadella, Bicarbonate stimulated phospholipid scrambling induces cholesterol redistribution and enables cholesterol depletion in the sperm plasma membrane, J. Cell Sci. 114 (2001) 3543–3555. [25] S. Shadan, P.S. James, E.A. Howes, R. Jones, Cholesterol efflux alters lipid raft stability and distribution during capacitation of boar spermatozoa, Biol. Reprod. 71 (2004) 253–265. [26] G.D. Zhu, G. Salazar, S.A. Zlatic, B. Fiza, M.M. Doucette, C.J. Heilman, A.I. Levey, V. Faundez, S.W. L'Hernault, SPE-39 family proteins interact with the HOPS complex and function in lysosomal delivery, Mol. Biol. Cell 20 (2009) 1223–1240. [27] S. Ward, Y. Argon, G.A. Nelson, Sperm morphogenesis in wild-type and fertilizationdefective mutants of Caenorhabditis elegans, J. Cell Biol. 91 (1981) 26–44. [28] W.E. Achanzar, S. Ward, A nematode gene required for sperm vesicle fusion, J. Cell Sci. 110 (Pt 9) (1997) 1073–1081. [29] J.W. Edmonds, J.K. Prasain, D. Dorand, Y. Yang, H.D. Hoang, J. Vibbert, H.M. Kubagawa, M.A. Miller, Insulin/FOXO signaling regulates ovarian prostaglandins critical for reproduction, Dev. Cell 19 (2010) 858–871. [30] T. Friedrichson, T.V. Kurzchalia, Microdomains of GPI-anchored proteins in living cells revealed by crosslinking, Nature 394 (1998) 802–805. [31] R.A. Mortara, L.M. Minelli, F. Vandekerckhove, V. Nussenzweig, F.J. RamalhoPinto, Phosphatidylinositol-specific phospholipase C (PI-PLC) cleavage of GPIanchored surface molecules of Trypanosoma cruzi triggers in vitro morphological reorganization of trypomastigotes, J. Eukaryot. Microbiol. 48 (2001) 27–37. [32] G. van Meer, D.R. Voelker, G.W. Feigenson, Membrane lipids: where they are and how they behave, Nat. Rev. Mol. Cell Biol. 9 (2008) 112–124.
[33] S. Hakomori, Bifunctional role of glycosphingolipids. Modulators for transmembrane signaling and mediators for cellular interactions, J. Biol. Chem. 265 (1990) 18713–18716. [34] S. Ichikawa, K. Ozawa, Y. Hirabayashi, Assignment of a UDP-glucose:ceramide glucosyltransferase gene (UGCG) to human chromosome band 9q31 by in situ hybridization, Cytogenet. Cell Genet. 79 (1997) 233–234. [35] H. Zhang, N. Abraham, L.A. Khan, D.H. Hall, J.T. Fleming, V. Gobel, Apicobasal domain identities of expanding tubular membranes depend on glycosphingolipid biosynthesis, Nat. Cell Biol. 13 (2012) 1189–1201. [36] K.H. Nomura, D. Murata, Y. Hayashi, K. Dejima, S. Mizuguchi, E. Kage-Nakadai, K. Gengyo-Ando, S. Mitani, Y. Hirabayashi, M. Ito, K. Nomura, Ceramide glucosyltransferase of the nematode Caenorhabditis elegans is involved in oocyte formation and in early embryonic cell division, Glycobiology 21 (2011) 834–848. [37] N.S. Radin, J.A. Shayman, J. Inokuchi, Metabolic effects of inhibiting glucosylceramide synthesis with PDMP and other substances, Adv. Lipid Res. 26 (1993) 183–213. [38] M. Guerrera, S. Ladisch, N-butyldeoxynojirimycin inhibits murine melanoma cell ganglioside metabolism and delays tumor onset, Cancer Lett. 201 (2003) 31–40. [39] A. Bacic, I. Kahane, B.M. Zuckerman, Panagrellus redivivus and Caenorhabditis elegans: evidence for the absence of sialic acids, Exp. Parasitol. 71 (1990) 483–488. [40] T. Lang, SNARE proteins and ‘membrane rafts’, J. Physiol. 585 (2007) 693–698. [41] C. Matsuda, Y.K. Hayashi, M. Ogawa, M. Aoki, K. Murayama, I. Nishino, I. Nonaka, K. Arahata, R.H. Brown Jr., The sarcolemmal proteins dysferlin and caveolin-3 interact in skeletal muscle, Hum. Mol. Genet. 10 (2001) 1761–1766. [42] W.G. Hill, B. An, J.P. Johnson, Endogenously expressed epithelial sodium channel is present in lipid rafts in A6 cells, J. Biol. Chem. 277 (2002) 33541–33544. [43] W.G. Hill, M.B. Butterworth, H. Wang, R.S. Edinger, J. Lebowitz, K.W. Peters, R.A. Frizzell, J.P. Johnson, The epithelial sodium channel (ENaC) traffics to apical membrane in lipid rafts in mouse cortical collecting duct cells, J. Biol. Chem. 282 (2007) 37402–37411. [44] E. deLamirande, P. Leclerc, C. Gagnon, Capacitation as a regulatory event that primes spermatozoa for the acrosome reaction and fertilization, Mol. Hum. Reprod. 3 (1997) 175–194. [45] 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. [46] N.L. Cross, Reorganization of lipid rafts during capacitation of human sperm, Biol. Reprod. 71 (2004) 1367–1373.