Cesll surface β1,4-galactosyltransferase-I activates G protein-dependent exocytotic signaling

Successful fertilization requires a complex sequence of cell recognition, adhesion, signaling, and exocytotic steps (Miller and Shur, 1994; Snell and White, 1996; Wassarman, 1999; Yanagimachi, 1994). Mammalian oocytes are surrounded by a glycoprotein coat called the zona pellucida, a structure made up of three glycoproteins termed ZP1, ZP2, and ZP3 (Wassarman, 1999). In the mouse, ZP3 was identified as the egg coat component responsible for binding free-swimming sperm (Wassarman, 1999). Following binding, ZP3 is believed to crosslink its receptor and induce exocytosis of the acrosome by activation of heterotrimeric G proteins (Florman et al., 1989; Leyton and Saling, 1989b; Ward and Kopf, 1993). Release of the acrosome is required for sperm penetration of the zona pellucida. Once inside the zona pellucida, the fertilizing sperm fuses with the egg membrane and induces exocytosis of cortical granules leading to the block to polyspermy and activation of development (Yanagimachi, 1994).

The identity of the receptor for ZP3 on sperm is a matter of considerable debate (Miller and Shur, 1994; Snell and White, 1996). A number of mouse sperm proteins have been suggested to function as ZP3 receptors (Bookbinder et al., 1995; Bork, 1996; Burks et al., 1995; Foster et al., 1997; Leyton and Saling, 1989a; Tsai and Silver, 1996). One well-studied ZP3 binding protein on sperm is b1,4-galactosyltransferase-I (GalTase; Miller and Shur, 1994; Shur, 1998). GalTase is one of at least six gene products that are capable of transferring galactose from uridine 5′-diphospho-galactose (UDP-Gal) in a b1,4 linkage to specific saccharide substrates (Amado et al., 1999). The b1,4-galactosyltransferase-I gene is unusual in that it encodes two protein isoforms that differ in their cytoplasmic domains. The long GalTase isoform consists of 399 amino acids, whereas the short isoform, a product of a downstream AUG translation initiation site, lacks 13 amino acids at the N terminus (Russo et al., 1990; Shaper et al., 1986). The two isoforms have overlapping but distinct biological properties. Both isoforms are able to function during glycoprotein processing in the Golgi complex but only the long isoform, with its additional 13 amino acid cytoplasmic sequence, is able to function as a signaling receptor in response to binding extracellular glycoside ligands (Lopez et al., 1991; Miller et al., 1992; Nguyen et al., 1994; Youakim et al., 1994a).

Sperm from all mammals examined thus far express GalTase on their surface and do so in a spatially restricted domain of the plasma membrane (Larson and Miller, 1997). A variety of reagents that inhibit GalTase or the GalTase binding site on ZP3 inhibit sperm-zona binding, supporting GalTase’s function as a ZP3 receptor (Lopez et al., 1985; Miller et al., 1992; Shur and Hall, 1982a). In addition to binding ZP3, GalTase may also trigger exocytosis of the sperm acrosome. Neoglycoproteins and polyacrylamide polymers terminating in N-acetylglucosamine, the saccharide substrate for GalTase, mimic ZP3 and induce the sperm acrosome reaction, as do GalTase antibodies (Loeser and Tulsiani, 1999; Macek et al., 1991; Nixon et al., 2000). Targeted disruption of the long isoform of GalTase results in sperm with impaired ability to bind ZP3 and undergo acrosomal exocytosis (Lu and Shur, 1997).

A variety of observations suggest that GalTase induces acrosomal exocytosis by activation of heterotrimeric G proteins including: (1) GalTase antibodies that act as agonists activate sperm G proteins, (2) pertussis toxin inhibits both ZP3- and GalTase antibody-induced acrosomal exocytosis, (3) sperm from mice that overexpress GalTase bind more ZP3 and show increased G protein activation, and (4) the GalTase cytoplasmic domain precipitates a G_i-containing complex from sperm lysates (Gong et al., 1995).

Although all of these studies are consistent with GalTase functioning as a signal transducing receptor for ZP3, there has been no report of expressing GalTase in a heterologous cell type, independent of other putative ZP3 receptors, to assess ZP3 binding and G-protein activation. In this study, we used Xenopus oocytes to express GalTase and assess its receptor function. When activated, Xenopus eggs release cortical granules in a manner related to sperm acrosomal exocytosis; exocytosis in both cells is regulated by G protein activation (Blitzer et al., 1993; Kline et al., 1991; Kline et al., 1988; Moore et al., 1993; Quick et al., 1994; Shilling et al., 1990; Shilling et al., 1994; Williams et al., 1992). Using this expression system, we found that eggs expressing GalTase on their surface bound ZP3 but not other zona pellucida glycoproteins and that GalTase agonists triggered G protein activation and pertussis toxin-sensitive exocytosis. Finally, a point mutation in the long GalTase cytoplasmic domain resulted in a loss of G protein activation and cortical granule exocytosis.

The coding sequence for the long or short version of GalTase was removed from pMGT-239/2615 or pMGT 33/2615 plasmid by digesting with EcoRI and HincII and blunt-end ligating into the BglII site of the pSP64T vector (Krieg and Melton, 1984; Lopez et al., 1991). The DLGT mutation was created using the Transformer Site-Directed Mutagenesis kit (Clontech) according to the manufacturer’s instructions. The sequence of the mutagenic primer was 5′-GTCGGCGGGACCATGGGGTTTGGTGAGCAGTTC-3′ which created a unique NcoI site. The sequence of the selection primer was 5′-GTGCCACCTGATATCTAAGAAACC-3′ which converted an AatII site to an EcoRV site. DH5a cells were transformed with plasmids containing the sequences for short, long and mutant long GalTase, and plasmid DNA was purified.

The coding sequences for the AU1 epitope tag (DTYRYI) and AU5 tag (TDFYLK) were added to the pSP64T plasmids with each GalTase insert using a PCR-based method with Pfu DNA polymerase. The forward primer for both the AU1 and AU5 tags was 5′-GTTACC-ACTAAACCAGCCTCAAGAACACCC-3′. The reverse primer for the AU1 tag was 5′-CTAGATGTAGCGGTACGTGTCTCTCGG-TGTCCCGATGTCCACTGTGAT-3′ and the reverse primer for the AU5 tag was 5′-CTACTTCAAGTAGAAGTCCGTTCTCGGTGTC-CCGATGTCCACTGTGAT-3′. Each PCR product was isolated from an agarose gel, phosphorylated using T4 DNA kinase, and self-ligated. Correct insertion of the tag sequence was confirmed by DNA sequencing.

For in vitro transcription, plasmid DNA was linearized with EcoRI. In vitro transcription was performed by using the Ambion SP6 mMESSAGE mMACHINE^™kit. Linearized template DNA was incubated with SP6 polymerase, an RNase inhibitor, dinucleotide triphosphates and cap analog (m⁷G[5′]ppp[5′]G). After transcription, DNA was removed by digestion with RNase-free DNase. The sample was phenol-chloroform extracted, ethanol precipitated and the pellet was resuspended in diethyl pyrocarbonate (DEPC)-treated water. The pSP64T vector contains a 3′ poly(A)32 tract terminated by a poly (C)30 tail. Transcribed inserts have a terminus that is refractory to deadenylation in mature eggs and GalTase translation can be sustained throughout oocyte maturation (Krieg and Melton, 1984). RNA aliquots of 5 mg were stored at –80°C until use.

To collect oocytes, Xenopus laevis were anesthetized in 8.2 mM 3-aminobenzoic acid ethyl ester and oocytes were recovered from the ovary. Oocytes were manually defolliculated using Dumont forceps. The vegetal hemisphere of oocytes was injected with either 20 nl of DEPC-treated water or 20 nl of 1 mg/ml GalTase RNA in DEPC-treated water using a microinjector (Narishige IM 300 Microinjector). Oocytes were incubated in OR3 medium for 48 hours at 19°C as described (Kline et al., 1991; Kline et al., 1988). Oocytes were matured by incubation in 1 mg/ml insulin and 1 mg/ml progesterone for 8-10 hours at 19°C. Maturing oocytes (eggs) form a white spot on their animal pole indicative of germinal vesicle breakdown. After formation of the white spot, eggs were cultured for an additional 4 hours to complete maturation.

To ensure that GalTase RNA was being translated, GalTase activity was measured by a GalTase enzyme assay (Shur and Hall, 1982b). Briefly, 10 ml of 80 mM uridine 5′diphosphate-[³H]galactose (UDP-[³H]galactose; 12.54 Ci/mmol; DuPont NEN-Research Products, Boston, MA) was combined with 20 ml of 1 mM unlabelled UDP-galactose and dried in a SpeedVac. Matured eggs were lysed in Hepes-buffered saline with 1% NP-40 and 10 mM MnCl₂ by pipetting 10 times through a small-bore pipette tip. A 40 ml volume of the egg homogenate was added to the tubes containing UDP-[³H]galactose. Assays were performed in Hepes-buffered saline with 10 mM MnCl₂, 30 mM N-acetylglucosamine as substrate, and 1 mM AMP as a competitive inhibitor of nucleotidases. Assays used the homogenates of 5 eggs and were performed in triplicate. The volume of each tube was adjusted to 100 ml using Hepes-buffered saline with 10 mM MnCl₂. The reactions were incubated in a 37°C water bath for 30 minutes. The tubes were mixed and 40 ml aliquots were removed at 0 and 30 minutes and added to 10 ml of stop solution (0.2 M EDTA, 50 mM Tris-HCl, phenol red, pH 7.4), to stop the reaction. Samples (45 ml) were spotted on 3mm Whatman chromatography paper. The substrate and breakdown products were separated from [³H]lactosamine product using paper electrophoresis in the presence of sodium tetraborate. The papers were dried and the origins, containing [³H]lactosamine, were cut out and placed into scintillation vials. ScintiVerse E scintillation cocktail (20 ml; Fisher Scientific) was added and the samples were counted by liquid scintillation spectroscopy. The counts at time zero were subtracted from each 30 minutes time point.

Confocal immunofluorescence microscopy was conducted on oocytes injected with water (control) or GalTase RNA containing either an AU1 or AU5 epitope tag sequence. Oocytes were transferred into approximately 1 ml thin-walled plastic vials containing OCT (Sakura Finetek). These vials were placed into a small container of isopentane placed into liquid nitrogen. After 10 minutes, the samples were stored at –80°C until sectioning. 10 mm sections were dried onto slides, equilibrated with PBS, and fixed in 3.9% paraformaldehyde for 30 minutes. Sections were blocked using 5% nonfat dry milk (SuperBlock, Pierce). A 1/1000 dilution of AU1 or AU5 monoclonal antibody (Covance) was added and incubated 1 hour at room temperature. As a control, the irrelevant antibody was used. Samples were rinsed three times, incubated with rotation for 1 hour in 1/500 dilution of Texas Red-anti-mouse secondary antibody (Jackson Laboratories), and rinsed three times. Stained samples were covered and viewed using a confocal microscope (Zeiss LSM-210) using the He/Ne laser for excitation and rhodamine filters to collect the signal.

Zonae pellucidae were purified from mouse ovarian homogenates and zona proteins were separated and electroeluted from SDS-PAGE gels as described (Miller et al., 1992). Zona pellucida proteins (50 mg) were radiolabeled with 0.25 mCi of Na¹²⁵I using Iodogen (Rockford, IL). Free ¹²⁵I was removed with a desalting column. Between 1-30 nM of each zona protein was incubated in 200 ml of MBS (88 mM NaCl, 1.0 mM KCl, 2.4 mM NaHCO₃, 15.0 mM Hepes, 0.30 Mm CaNO₃, 0.41 mM CaCl₂, 0.82 mM MgSO₄, 10 ml/ml penicillin, 10 mg/ml streptomycin) containing 4% BSA and 125 mg/ml of sodium iodate with three matured eggs in triplicate. Eggs were prepared by injecting oocytes with 20 ng GalTase RNA, incubating for 24-48 hours, and maturing with progesterone and insulin. Incubations were performed in a 96-well plate with cellulose acetate well bottoms (MultiScreenâ-HA, Millipore) for 2 hours at 22°C. Unbound zona proteins were removed by filtration and five washes with cold MBS. Bound zona protein was quantitated with a gamma counter.

Zonae pellucidae were purified from mouse ovarian homogenates and zona proteins were electroeluted from preparative SDS-PAGE gels (Miller et al., 1992). GalTase antibodies were prepared from rabbits immunized with affinity-purified, bacterially expressed murine GalTase (Nguyen et al., 1994) and preimmune and immune sera were dialyzed in F1 medium (31 mM NaCl, 1.8 mM KCl, 0.5 mM NaH₂PO₄, 1.9 mM NaOH, 1.0 mM CaCl₂, 0.06 mM MgCl₂, and 10 mM tricine, pH 7.8; Kline et al., 1988). Alternatively, IgG was prepared by Protein A affinity chromatography. A 1/30 dilution of GalTase antisera, anti-GalTase IgG (25 ng/ml), or ZP proteins (30 ng/ml) were incubated with eggs in F1 medium. Eggs were observed under a dissecting microscope for 20 minutes for both cortical contraction and vitelline envelope elevation. An egg that has undergone cortical contraction has distinct changes in appearance. Before activation, the egg is divided almost equally between the dark pigmented animal hemisphere and the vegetal hemisphere. After cortical contraction, the pigmented cytoplasm appears contracted (Stewart-Savage and Grey, 1982).

After microscopic observation, eggs were fixed, dissected and stained with Periodic Acid Shiff’s reagent (PAS) to detect cortical granules. The procedure of fixation and PAS staining was essentially as described previously (Humason, 1979). After activation, eggs were fixed immediately in Smith’s fixative for 24 hours. Eggs were washed overnight in running water and transferred to De Boer’s Ringer Solution (110 mM NaCl, 1.3 mM KCl, 0.44 mM CaCl₂, pH 7.3). Prior to embedding, eggs were dehydrated through a series of increasing concentrations of ethanol in a vacuum. Eggs were embedded with JB4, sectioned and stained with PAS.

To determine if pertussis toxin could block GalTase-induced egg activation, pertussis toxin (4 mg/ml) was added to oocytes during their incubation in OR3 and remained present during maturation (a total of 72-hours incubation in pertussis toxin), as described (Kline et al., 1991). Oocytes were matured in progesterone and insulin.

After injection and maturation, eggs were homogenized using a Dounce homogenizer in a buffer containing Tris-EDTA membrane buffer (50 mM Tris-HCl, 1 mM EDTA, 0.1 mM phenylmethylsulfonylfluoride, 1 mM pepstatin A, 1 mM leupeptin, pH 7.4). The homogenate was centrifuged three times at 2,000 g for 10 minutes at 5°C. The supernatants were combined into an ultracentrifuge tube and centrifuged at 200,000 g for 1 hour at 5°C. The supernatant was discarded and the pellet was resuspended in 200 ml of Tris-EDTA membrane buffer. Membrane proteins were solubilized in 1% SDS and the protein concentration was measured using the BCA protein assay with BSA as a standard, according to the manufacturer (Pierce). Samples were snap-frozen and stored at –80°C.

The modified GTPg[³⁵S] binding assay is similar to that described previously (Gong et al., 1995). Briefly, oocyte membranes (50 ml from a 200 mg/ml protein concentration solution) were incubated in a total volume of 100 ml with 0.1 mM GTPgS (Sigma) in 50 mM Tris-HCl, 1 mM EDTA, 20 mM MgCl_2, 26 mM GDP, and 8 mM b-mercaptoethanol for 15 minutes at 30°C. The preparation was diluted with 200 ml of the above buffer lacking GTPgS. From the diluted membranes, 50 ml was removed and combined with 15 ml of ligand in 96-well culture plates with cellulose acetate bottoms (MultiScreenâ-HA, Millipore). Either anti-GalTase serum, preimmune serum, or soluble zona glycoproteins were added so that the final concentration after adding GTPg[³⁵S] was 1:40 antiserum dilution or 30 ng/ml of zona proteins. After incubation at 30°C for another 15 minutes, buffer (35 ml) containing GTPg[³⁵S] was added (final concentration 20 nM) and the mixture was incubated for 15 minutes at 30°C. To remove unbound GTPg[³⁵S], the membranes were rapidly filtered and washed with 4 volumes of 200 ml Tris-EDTA buffer containing 10 mM Mg²⁺. The filters containing the washed membranes were punched from the plates and bound GTPg[³⁵S] was quantitated by a scintillation counter.

Oocyte membranes were subjected to pertussis toxin-catalyzed ADP-ribosylation as described (Kline et al., 1991). Briefly, oocyte membranes were resuspended to a concentration of 200-400 mg/ml in ice-cold buffer containing 50 mM Tris-HCl, 1 mM EDTA, 0.1 mM phenylmethylsulfonylfluoride, 1 mM pepstatin A, 1 mM leupeptin, at pH 7.5. Activated pertussis toxin and NAD⁺ were added to the membranes to a final concentration of 5 mg/ml and 5.3 mM, respectively. Pertussis toxin-catalyzed ADP-ribosylation was allowed to proceed by incubating the reaction mixture at 30°C for 60 minutes. The treated membranes were then assayed for soluble zona pellucida protein and GalTase antibody-induced GTPg[³⁵S] binding as described above. Untreated controls were prepared identically except that pertussis toxin was replaced with buffer in the reaction mixture.

GTPase assays were performed using methods similar to those described elsewhere (Ward et al., 1992). In a 1.5 ml microfuge tube, 10 μl of GalTase antisera, preimmune sera, or solubilized zona proteins were added to yield a final concentration of 1:30 or 30 ng/ml in the GTPase assay, respectively. Membrane protein (40 ml of a 200-ng/ml stock in 2′ reaction buffer) was added. The 1′ reaction buffer was 50 mM Tris-HCl, 0.2 mM EGTA, 10 mM MgCl₂, 0.2% BSA, 1.5 mM App(NH)p, 5 mM dithiothreitol, 0.2 mM ATP, 5 mM phosphocreatine, 100 units/ml phosphocreatine kinase, and 2 mM GDP. [g-³²P]GTP (0.5 mM final concentration) was added as a substrate in 50 ml. The samples were incubated at 30°C for 15 minutes. The assay was stopped by adding 1 ml of ice-cold stop buffer (5% activated charcoal, 0.5% BSA, 0.1% dextran, 20 mM K₂HPO₄ buffer, pH 7.5). After 15 minutes incubation on ice, the samples were centrifuged for 2 minutes at 15,000 g at 24°C and 350 ml of the supernatant was aspirated and placed into a scintillation vial. The samples were counted without adding scintillation cocktail using Cerenkov’s method. Background counts were obtained by deleting membranes from the assay.

The data for GalTase assay, egg activation assay, GTPgS binding assay and GTPase assay were evaluated by analysis of variance using SigmaStat software (SPSS, Inc.). Individual comparisons were made using Tukey’s Studentized range (Neter and Wasserman, 1974). Differences at a probability of P<0.05 were considered statistically significant.

Previous studies have suggested that sperm GalTase functions as a ZP3 binding protein and that crosslinking GalTase with ZP3 induces G protein-dependent signaling. It has also been suggested that amino acids found in the long isoform of GalTase, but not the short isoform, are required for this signal transduction capacity. To examine these possibilities more directly, Xenopus oocytes were used as a heterologous system to express murine GalTase isoforms and assess the ability to bind zona glycoproteins, activate G-proteins, and trigger exocytosis.

The coding sequence for the long version of GalTase was inserted into the pSP64T vector (Fig. 1). Capped RNA (10-20 ng) was injected into Stage VI Xenopus oocytes and the oocytes were matured. To confirm that injected RNA was expressed, GalTase enzyme activity was measured in egg homogenates. Injection of murine GalTase RNA resulted in a 200-fold increase in GalTase activity (Fig. 2).

Because the two forms of GalTase appear to be preferentially localized to different regions, immunofluorescence was used to determine the location of murine long GalTase expressed in eggs. For convenience, an AU1 epitope tag was added to the GalTase sequence (Fig. 1). GalTase-injected oocytes were matured and fixed. Sections were analyzed by confocal microscopy. Most of the fluorescence signal detected on eggs expressing long GalTase was found on the plasma membrane (Fig. 3A). Controls using an irrelevant antibody (AU5) or water-injected oocytes had no signal (Fig. 3D). The AU1 antibody was also used in immunoblots of egg membrane protein and GalTase glycosylation variants from 48-54 kDa, as well as a poorly glycosylated 45 kDa GalTase band, were detected (data not shown). Similar results were observed when long GalTase was tagged with the AU5 epitope and the AU5 antibody was used for immunofluorescence (data not shown). Therefore, long murine GalTase was localized to the plasma membrane of eggs where it could act as a receptor.

To determine if plasma membrane GalTase bound ZP3, purified zona pellucida proteins were labeled with ¹²⁵I and incubated (1-30 nM) with mature eggs expressing GalTase. Unbound zona proteins were removed by filtration. Eggs expressing GalTase bound ZP3, but not ZP2 (Fig. 4). A 100-fold excess of unlabelled ZP3 was able to reduce binding by 77%, demonstrating that ZP3 binding to eggs was saturable (data not shown). Although ZP3 binding was observed consistently, the affinity of ZP3 for GalTase-expressing eggs varied among different ZP3 preparations, precluding a calculation of a reliable binding affinity. These results showed that murine GalTase protein expressed in heterologous cells bound ZP3 but not other zona pellucida glycoproteins.

Signaling events leading to cortical granule exocytosis in eggs are similar (although not identical) to signaling events during sperm acrosomal exocytosis. Intracellular Ca²⁺ and pH are elevated in both cells leading to exocytosis of cortical granules or the acrosome (Arnoult et al., 1999; Busa and Nuccitelli, 1985; Florman et al., 1998; Kline, 1988; Miyazaki et al., 1993; Miyazaki et al., 1992; Nuccitelli et al., 1993). Furthermore, G proteins are required for exocytosis of the sperm acrosome, and activation of G proteins in eggs triggers exocytosis of cortical granules (Florman et al., 1989; Kline et al., 1991; Kline et al., 1988; Moore et al., 1993; Shilling et al., 1990; Shilling et al., 1994; Ward and Kopf, 1993; Williams et al., 1992). Therefore, we reasoned that Xenopus eggs may be useful cells in which to study exocytotic signaling, and hypothesized that GalTase would initiate G protein-dependent signaling events in Xenopus eggs leading to cortical granule exocytosis.

GalTase-expressing oocytes were matured and transferred to F1 fertilization medium (Kline et al., 1988) to which several multivalent GalTase agonists were added. Total zona proteins, purified ZP3, and antibodies raised against recombinant GalTase were used as ligands. Controls included ZP1, ZP2 and preimmune antibodies as well as water-injected eggs. Egg activation was scored using a dissecting microscope and confirmed by cortical granule histochemical staining. When Xenopus eggs are activated, the cortex of the animal pole undergoes a transient contraction, resulting in a retraction of the pigmented zone and elevation of the vitelline envelope (Stewart-Savage and Grey, 1982). Only eggs in which all such events occurred were scored as activated.

Purified ZP3 and GalTase agonistic antibodies stimulated both cortical contraction and envelope elevation in eggs expressing GalTase (Figs 5, 6). Appropriate ligands activated 50-60% of eggs expressing GalTase (Fig. 6A). All eggs that appeared activated by cortical contraction and vitelline envelope elevation had also released their cortical granules (Fig. 5B), demonstrating that GalTase activated signaling events leading to exocytosis. Control eggs treated with preimmune antibodies or injected with water had low activation rates (Fig. 6B). Over 90% of the control eggs could be activated by subsequent treatment with the calcium ionophore A23187, demonstrating the viability of control eggs (data not shown).

Previous studies suggested that ZP3 crosslinks its sperm receptor to trigger acrosomal exocytosis (Leyton and Saling, 1989b). Similarly, GalTase requires crosslinking to induce exocytotic signaling in sperm (Macek et al., 1991). Consistent with a requirement for crosslinking, monovalent F_ab fragments of GalTase antibodies were unable to activate GalTase-expressing eggs (Fig. 6A).

Pertussis toxin blocks induction of the acrosome reaction by zona proteins or GalTase antibody (Gong et al., 1995; Ward and Kopf, 1993). In this study, pertussis toxin pretreatment of GalTase-expressing eggs prevented ZP3- or GalTase antibody-induced egg activation (Fig. 6A). GTPλ [³⁵S] binding and GTPase activity were used to measure G protein activation in GalTase-expressing eggs (Neer, 1995). Heterotrimeric G proteins are activated when GDP bound to the Ga subunit is replaced by GTP and we used the poorly hydrolyzable radiolabeled analog GTPg[³⁵S] to measure activation. Egg membrane proteins were incubated with GTPλ [³⁵S] and either ZP3 or GalTase antibodies. Both ligands stimulated GTPλ [³⁵S] binding, but neither F_ab fragments of GalTase antibodies nor ZP2 increased GTPλ [³⁵S] binding above controls (Fig. 7). The magnitude of increase was similar to that observed in sperm membranes in response to soluble zona pellucida proteins. The GalTase-dependent increase in GTPλ [³⁵S] binding was sensitive to pertussis toxin treatment of eggs or membrane preparations. As an independent confirmation of G protein activation, GTPase activity in the egg membrane preparations was assayed, using [λ -³²P]GTP as a substrate. An activated Ga subunit hydrolyzes bound GTP before reassociating with the β λ subunits to reform the heterotrimer. GTPase activity was increased by ZP3 and GalTase antibody, compared to controls using other zona glycoproteins or preimmune antibody (Fig. 8). The increase in both GTPg[³⁵S] binding and [γ -³²P]GTP hydrolysis was comparable to that previously reported for sperm membranes in response to soluble zona glycoproteins (Gong et al., 1995; Ward et al., 1992).

Results described above suggest that GalTase crosslinking, either directly or indirectly, activates G proteins and support previous reports in which the cytoplasmic domain of long GalTase was shown to precipitate the G_i heterotrimer from sperm (Gong et al., 1995). The ability of GalTase to activate G proteins is unexpected because GalTase has a single transmembrane domain whereas traditional G protein coupled receptors have seven transmembrane domains. However, there are reports that groups of basic amino acids within single transmembrane domain receptors are able to associate with and activate G proteins (Okamoto et al., 1990; Okamoto and Nishimoto, 1991; Okamoto et al., 1991; Okamoto et al., 1993). More recently, clusters of basic amino acid residues necessary for G protein activation have also been identified in seven transmembrane receptors (Frandberg et al., 1998; Lee et al., 1996; Okamoto and Nishimoto, 1992; Wade et al., 1999; Wang, 1999; Xie et al., 1997). More detailed structural features necessary for G protein activation by single transmembrane receptors were identified. Within a 20 amino acid peptide, two sequence motifs have been implicated in G protein activation. First, at the N terminus of this peptide, a region with two basic amino acids is necessary. Second, at the C terminus, a BBXB or BBXXB motif, where B is a basic amino acid and X is any amino acid is necessary (Lee et al., 1996; Okamoto et al., 1990; Okamoto and Nishimoto, 1992; Okamoto et al., 1991; Wade et al., 1999; Wang, 1999). Both the long and short forms of GalTase have a QRACR sequence in the cytoplasmic domain near the plasma membrane that could serve as a BBXXB motif (although Q is not a basic amino acid), but only the long form has a sequence (MRFR) compatible with the first motif.

First, we determined if the short version of GalTase could activate eggs. ZP3 or GalTase antibodies did not activate eggs expressing short GalTase (Fig. 6C). This is in agreement with the requirement for both basic amino acid motifs within the 20 amino acid peptide (Ikezu et al., 1995; Nishimoto, 1993; Nishimoto et al., 1993; Okamoto et al., 1990). However, the majority of short GalTase was located in the cytoplasm (Fig. 3B). Consequently, we focused on the basic sequence unique to the long form of GalTase, specifically the MRFR sequence at the cytoplasmic N terminus. The two arginine codons were mutated so that they encoded alanine residues (DLGT) and mutant RNA was injected into oocytes. Mutant GalTase retained enzyme activity and was expressed on the plasma membrane (Figs 2, 3C). However, mutant GalTase did not activate eggs in response to ZP3 or GalTase antibodies (Fig. 6D), nor ¹⁰ did it increase GTPg[³⁵S] binding in response to GalTase ligands (Fig. 7D). Therefore, the ability of ₅ GalTase to activate G proteins and exocytotic signaling requires the basic MRFR amino acid sequence that is unique to the long isoform of GalTase.

Sperm must undergo acrosomal exocytosis in order to penetrate the zona pellucida and fertilize eggs. Triggered by binding to the zona pellucida glycoprotein ZP3, the acrosome reaction is a G protein-mediated process. Herein, we show that the long isoform of GalTase fulfills criteria expected of a ZP3 binding protein when expressed on Xenopus oocytes. Furthermore, ZP3 binding to GalTase leads to G protein activation, as assessed by increased GTP binding and GTP hydrolysis, egg activation, and cortical granule exocytosis. Pertussis toxin blocked GalTase-dependent egg activation and mutagenesis of the GalTase cytoplasmic domain inhibited the ability of ZP3 to activate eggs or to activate G proteins.

Previous studies have suggested that ZP3 induces the acrosome reaction by crosslinking its receptor. Similarly, GalTase antibodies, but not their F_ab fragments, are able to induce acrosomal exocytosis in sperm and cortical granule exocytosis in GalTase-expressing Xenopus oocytes (present study). Recent results demonstrate that GalTase readily forms dimers (Yamaguchi and Fukuda, 1995), a process that, on sperm, may be facilitated by binding multiple oligosaccharides on a ZP3 molecule (Miller et al., 1992). ZP3-induced dimerization may be analogous to that induced when one growth hormone molecule binds to two copies of its receptor (Cunningham et al., 1991) or to the ability of heptahelical G protein-coupled receptors to form dimers (Mohler and Fritschy, 1999).

The ability of GalTase crosslinking to activate G proteins is at odds with the traditional view of heptahelical G protein-coupled receptors (Neer, 1995). However, some studies indicated that small peptides from single pass transmembrane proteins interact with G proteins. The insulin–like growth factor-II/mannose 6-phosphate receptor (IGF-II/M-6-RP) passes the plasma membrane once and has been reported to be G protein coupled (Ikezu et al., 1995; Nishimoto, 1993; Nishimoto et al., 1993; Okamoto et al., 1990; Okamoto et al., 1993), although others have not confirmed these results (Korner et al., 1995). IGF–II/M-6-PR has a region of similarity in its cytoplasmic domain to mastoparan, a peptide that activates G_i and G_o proteins (Korner et al., 1995; Nishimoto et al., 1993; Okamoto et al., 1990). Mutagenesis of IGF–II/M-6-PR identified a 20 amino acid sequence that contains two basic motifs necessary for binding to G proteins. Similar basic amino acid sequences necessary for G protein activation have been identified in heptahelical receptors (Frandberg et al., 1998; Lee et al., 1996; Okamoto and Nishimoto, 1992; Wade et al., 1999; Wang, 1997; Wang, 1999). Finally, several studies have shown that the cytoplasmic domain of the EGF receptor is coupled to G proteins, perhaps through a related sequence (Liang and Garrison, 1991; Sun et al., 1997; Yang et al., 1993).

Examination of the cytoplasmic domain of long GalTase reveals that it contains both basic motifs proposed to be responsible for G protein binding. Furthermore, these regions are conserved in murine, bovine, human and porcine long GalTase (Mengel-Gaw et al., 1991; Russo et al., 1990; Shaper et al., 1986) (E. A. Landers and D. J. M., unpublished results). The short GalTase isoform, which is unable to function as a signal transducing receptor and unable to bind G protein heterotrimers in sperm lysates, contains only one of these two motifs. Replacement of two arginine residues in the N-terminal domain unique to the long isoform abrogated the ability of GalTase to trigger egg activation, as well as the increase in GTPase activity and GTPγ [³⁵S] binding. However, the mutant GalTase was still located on the oocyte surface, suggesting these residues are not required for cell surface localization. The mechanism responsible for surface localization of GalTase remains obscure. Rather than direct activation, it is possible that GalTase activates G proteins indirectly through adaptor proteins, in a mechanism similar to that utilized by some protein tyrosine kinase receptors (Pawson and Scott, 1997). This is a novel concept because no adaptor proteins have been described that are intermediates between heptahelical receptors and heterotrimeric G proteins. Perhaps serving as a link to protein kinase C and the cytoskeleton, a protein called SSeCKS (Src suppressed C kinase substrate) was identified using a yeast two hybrid screen with the GalTase cytoplasmic domain as bait (M. Wassler, C. Foote, I. Gelman, and B. D. S., unpublished observations).

Although GalTase stimulates exocytosis in both sperm and GalTase-expressing Xenopus eggs, it is noteworthy that exocytotic signaling is not identical in both gametes. Pertussis toxin inhibits the ZP3-induced acrosome reaction and opening of ligand-activated Ca²⁺-gated Cl⁻ channels in eggs expressing G protein-coupled receptors, but it does not inhibit the normal egg activation process triggered by the fertilizing sperm (Blitzer et al., 1993; Gong et al., 1995; Kline et al., 1991; Quick et al., 1994; Ward and Kopf, 1993).

GalTase binding depends on the availability of specific N-acetylglucosamine residues on O-linked oligosaccharides of ZP3 (Miller et al., 1992). Recent compositional studies using the entire zona pellucida found terminal N-acetylglucosamine residues on N-linked oligosaccharides, but did not detect any terminal N-acetylglucosamine residues on O-linked oligosaccharides (Easton et al., 2000). However, the O-linked oligosaccharide/s that bind sperm are only a fraction of the total O-linked oligosaccharides of ZP3 (Florman and Wassarman, 1985). As the authors discuss, all zona pellucida glycoproteins were pooled and few O-linked oligosaccharides were recovered. Presumably only the most abundant O-linked oligosaccharides would be detectable. Interestingly, the distribution of oligosaccharides in the zona pellucida is heterogeneous; the outer half of the zona matrix has a different glycoside composition than the internal zona matrix. This implies a temporal regulation of glycosyltransferase and/or glycosidase expression during secretion of the zona pellucida, as well as a functional heterogeneity on the zona matrix (Aviles et al., 2000; Skutelsky et al., 1994). In any event, the heterogeneous distribution of glycoside chains in the zona and the relatively low abundance of sperm-binding oligosaccharides indicate that they are difficult to detect in compositional analyses.

Studies of GalTase function by expression in heterologous cells are consistent with studies of GalTase function in transgenic mice. Overexpressing the long GalTase isoform produces sperm that bind more soluble ZP3 than wild-type sperm, undergo accelerated G protein activation, and undergo the zona-induced acrosome reaction at a higher frequency (Youakim et al., 1994b). In contrast, sperm from mice with a targeted disruption in GalTase are unable to bind normal amounts of ZP3, are severely compromised in their ability to undergo an acrosome reaction, and penetrate the zona pellucida only 7% as efficiently as wild-type sperm (Lu and Shur, 1997). Despite these defects, GalTase-null males are fertile, illustrating that successful fertilization requires redundant interactions. Nevertheless, GalTase fulfills criteria expected of a sperm receptor specific for ZP3 and, when bound by ZP3, can activate heterotrimeric G proteins leading to exocytosis.

This work was supported by National Sciences Foundation (IBN 94-18077) and the Illinois Agricultural Experiment Station as part of Hatch project no. ILLU-35-0335 (to D. J. M.) and by the National Institutes of Health (HD38311 to D. J. M. and HD23479 and HD22590 to B. D. S.). We are grateful to members of our laboratories for critical reading of the manuscript.