Molecular Identification and Characterization of Xenopus Egg Uroplakin III, an Egg Raft-associated Transmembrane Protein That Is Tyrosine-phosphorylated upon Fertilization*

Here we describe mass spectrometric identification, molecular cloning, and biochemical characterization of a lipid/membrane raft-associated protein that is tyrosine-phosphorylated upon Xenopus egg fertilization. This protein is homologous to mammalian uroplakin III, a member of the uroplakin family proteins (UPs) that constitute asymmetric unit membranes in the mammalian urothelial tissues, thus termed Xenopus uroplakin III (xUPIII). xUPIII contains N-linked sugars and is highly expressed in Xenopus eggs, ovary, urinary tract, and kidney. In unfertilized eggs, xUPIII is predominantly localized to the lipid/membrane rafts and exposed on the cell surface, as judged by surface biotinylation experiments and indirect immunofluorescent studies. After fertilization or hydrogen peroxide-induced egg activation, xUPIII becomes rapidly phosphorylated on tyrosine residue-249, which locates in the carboxyl-terminal cytoplasmic tail of the molecule. Raft localization and tyrosine phosphorylation of xUPIII can be reconstituted in HEK293 cells by coexpression of xUPIII, and Xenopus c-Src, a tyrosine kinase whose fertilization-induced activation in egg rafts is required for initiation of development. In mammals, UPIII is forming a complex with a tetraspanin molecule uroplakin Ib. As another tetraspanin, CD9, is known to be a critical component for sperm-egg fusion in the mouse, we have assumed that xUPIII is involved in sperm-egg interaction. An antibody against the extracellular domain of xUPIII blocks sperm-egg interaction, as judged by the occurrence of egg activation and first cell cleavage. Thus, xUPIII represents an egg raft-associated protein that is likely involved in sperm-egg interaction as well as subsequent Src-dependent intracellular events of egg activation in Xenopus.

A number of animal and plant species have been employed as a model organism for fertilization research. Several molecules from both sperm and eggs of many animals have been identified as those that are involved in sperm-egg interaction, spermegg fusion, and/or subsequent sperm-induced egg activation (1)(2)(3)(4). They include ADAMs (a disintegrin and metalloprotease family of proteins), galactosyltransferase, integrins, and tetraspanins such as CD9 and CD81, all of which may work in the gamete interaction and/or fusion (in mammals) (5). Candidates for the trigger of egg activation include sperm-borne phospholipase C (PLC) 1 and truncated c-Kit (in mammals) as well as egg-associated Src-related tyrosine kinases and PLC␥ (in frogs, sea urchin, and starfish) (6 -11). However, the molecular mechanisms connecting sperm-egg interaction/fusion and egg activation are not well understood in any organisms.
We have previously shown that fertilization in Xenopus laevis involves activation of an egg-associated tyrosine kinase Src, the ubiquitous expression and multiple functions of which in normal cells have been well documented (12)(13)(14). The Src activation occurs within 1 min of insemination, and the active Src interacts with, phosphorylates, and activates PLC␥ (15,16). Pharmacological experiments using specific inhibitors in the intact egg system as well as immunodepletion experiments in cell-free systems demonstrate that Src-dependent PLC␥ activation is required for many features of egg activation: a transient calcium release, inactivation of cytostatic factor, and resumption of meiosis II (15,(17)(18)(19). On the other hand, artificial activation of Src by using hydrogen peroxide in both intact eggs and cell-free systems promotes events of egg activation as noted above (19,20). These indicate that Src activation is sufficient for egg activation in Xenopus. Further analysis has been directed to analyze molecular mechanism related to sperm-induced activation of Src, and we have found that Src is enriched in the egg lipid/membrane rafts and that sperm activate the raft-associated Src in vivo and in vitro (18,21).
Lipid/membrane rafts, also called low density, detergentinsoluble membranes or detergent-resistant membranes, are kinds of microdomains in the plasma membrane and are thought to facilitate signal transduction by cell surface receptors (22)(23)(24). Rafts are known to contain various types of transmembrane and/or signaling molecules. The dynamic arrangement and functional importance of rafts have been described in several cell systems such as lymphocytes (25), neuronal cells (26), and platelets (27). Localization of Src in egg rafts would be important for eliciting sperm-induced egg activation signaling. In fact, we have shown that sperm can activate Src in the isolated rafts prepared from unfertilized eggs (18,21). The results suggest that the raft fractions contain a molecule(s) that can transmit sperm-binding signal to the machinery for Src activation.
In the present study, we demonstrate that a 30-kDa egg raft-associated protein is rapidly phosphorylated on tyrosine residues after fertilization or hydrogen peroxide treatment of Xenopus eggs. Peptide mass fingerprinting and product ion mass fingerprinting show that the phosphorylated protein is a Xenopus homologue of uroplakin III, termed Xenopus uroplakin III (xUPIII). From the deduced amino acid sequence, xUPIII is supposed to be a single transmembrane protein that contains a large amino-terminal extracellular domain and a short carboxyl-terminal cytoplasmic tail. Fertilization-induced phosphorylation occurs in the tyrosine residue 249 of the carboxyl-terminal tail, as revealed by the mass spectrometric analysis. An antibody against the extracellular domain of xUPIII blocks fertilization in a dose-dependent manner. Thus, these data suggest that xUPIII is involved in sperm-egg interaction/fusion via its extracellular domain and in the Src-dependent egg activation signaling via phosphorylation of the tyrosine residue in its cytoplasmic tail.
Animals, Gametes, Embryos, and Tissues-X. laevis were obtained from the Hamamatsu Seibutsu Kyozai (Hamamatsu, Japan) and maintained in de-chloride tap water at ambient temperature (18 -22°C). Ovulated or gently squeezed eggs were obtained by the method as described previously (21) and immediately washed three times with 1ϫ DeBoer's buffer (DB) containing 110 mM NaCl, 1.3 mM KCl, and 0.44 mM CaCl 2 (pH 7.2, by addition of NaHCO 3 ), kept at ambient temperature, and used within 3 h of spawning. Before experiments, the jelly layer surrounding the eggs was removed by incubation with 1ϫ DB supplemented with 2% cysteine and 0.06 N NaOH for 3-8 min. The resulting jelly coat-free eggs were washed five times with 1ϫ DB and subjected to egg activation treatment (see below). Sperm suspension in 1ϫ DB was prepared from testis of male frogs as described previously (18). Immediately before insemination, the sperm suspension was sedimented by brief centrifugation and incubated with egg jelly water (21). After the incubation, the suspension was washed once with 1ϫ DB, and the resulting jelly water-treated sperm were used for insemination. The concentration of sperm in the suspension was determined by counting sperm number by hematocytometer. Activation of jelly layer-free eggs was done by fertilization with jelly water-treated sperm (about 10 7 sperm/ml) or H 2 O 2 (10 mM) for the times as specified under "Results." After the activation treatments, egg/embryo samples were washed extensively with ice-cold 1ϫ DB, immediately frozen in liquid nitrogen, and kept at Ϫ80°C. To analyze proteins in adult tissues of Xenopus, several organs were surgically removed from sacrificed animals, imme-diately washed extensively with phosphate-buffered saline, and kept at Ϫ80°C until required.
Extraction and Subcellular Fraction of Eggs-To obtain raft and non-raft fractions, egg samples prepared as above were mixed with a 5-fold volume of ice-cold extraction buffer containing 20 mM Tris-HCl (pH 7.5), 1 mM EDTA, 1 mM EGTA, 10 mM ␤-mercaptoethanol, 1 mM Na 3 VO 4 , 10 g/ml leupeptin, 20 M (p-amidinophenyl)methanesulfonyl fluoride hydrochloride (APMSF), 150 mM NaCl, and 250 mM sucrose and homogenized with a 7-ml Dounce tissue grinder (Wheaton). The homogenates were centrifuged at 500 ϫ g for 10 min to remove debris and yolk platelets, and the supernatants were collected and centrifuged at 150,000 ϫ g for 20 min. Concentrated Triton X-100 (25%) was then added to the fluffy layer of the pellet (crude membranes) to yield a final concentration of Triton X-100 at 1%. The mixtures were homogenized again, incubated on ice for 10 min, and mixed with equal volumes of ice-cold raft buffer containing 150 mM NaCl and 85% sucrose (sucrose buffer). The resulting mixtures (5 ml) were layered first with 19 ml of 30% sucrose and second with 12 ml of 5% sucrose in the same buffer. The samples were centrifuged at 144,000 ϫ g for 20 -24 h in an SW28 rotor (Beckman Instruments). After the centrifugation, 3-ml aliquots of 12 fractions were collected from the top to the bottom of the tubes. Usually, fractions 4 -6 were pooled as raft fractions, whereas fraction 12 was pooled as detergent-solubilized non-raft fractions. Alternatively, egg samples were directly homogenized with the Triton X-100-containing buffer and subjected to raft fractionation as described previously (21). In some experiments, raft fractions were diluted with more than 4-fold volume of water and centrifuged at 150,000 ϫ g for 30 min. The pellet fractions were regarded as concentrated rafts and resuspended with an appropriate volume of raft buffer containing 150 mM NaCl and then used for experiments.
To obtain cytosolic and membrane fractions, samples were homogenized in 5-fold volume of ice-cold extraction buffer. The homogenates were centrifuged at 500 ϫ g for 10 min, and the supernatants were centrifuged at 150,000 ϫ g for 20 min. The resulting supernatants were collected as the cytosolic fractions. The pellets were further homogenized with the extraction buffer supplemented with 1% Triton X-100 and 0.1% SDS. The homogenates were kept on ice for 10 min and then centrifuged at 150,000 ϫ g for 20 min. The resulting supernatants were collected as the detergent-solubilized membrane fractions. Protein concentrations were determined by means of the dye binding assay (Bio-Rad). Calibration was made with standard bovine serum albumin (Calbiochem).
Immunoprecipitation, SDS-PAGE, and Immunoblotting-Proteins from egg samples or cultured cells (see below) (50 -500 g, 1 g/l) were immunoprecipitated with an appropriate amount of antibodies as specified under "Results" for 3-6 h at 4°C. After centrifugation at 10,000 rpm for 10 min at 4°C, the immune complexes were adsorbed onto 10 l of protein A-Sepharose beads by gentle agitation for 30 min at 4°C. The beads were washed three times with 500 l of buffer containing 0.1% SDS, 1% Triton X-100, 1% deoxycholate sodium salt, 0.15 M NaCl, 50 mM Tris-HCl (pH 7.5), 1 mM Na 3 VO 4 , 10 g/ml leupeptin, and 20 M APMSF. The washed beads were then treated with Laemmli's SDS sample buffer (29) at 98°C for 5 min. The SDS-denatured proteins were separated by SDS-PAGE and analyzed by immunoblotting or silver stain as described previously (21). Note that silver stain was done with the use of a Bio-Rad kit (Silver Stain Plus). Also note that in immunoblotting analysis, detection of the immune complex between proteins and the primary antibodies of interest was made by enzyme-linked color development with alkaline phosphatase or horseradish peroxidase that were conjugated to the secondary antibodies (Cappel).
In-gel Protein Digestion and Mass Spectrometry-Proteins in SDSpolyacrylamide gels were visualized by using the Bio-Rad Silver Stain Plus kit. The bands corresponding to proteins were excised and then were destained by treatment with 15 mM potassium hexacynoferrate (III), K 3 [Fe(CN) 6 ], 50 mM sodium thiosulfate, Na 2 S 2 O 3 , for 10 min at room temperature. Destained proteins in gels were reduced by incubating with 10 mM EDTA, 10 mM dithiothreitol, 100 mM ammonium bicarbonate for 1 h at 50°C and alkylated by treatment with 10 mM EDTA, 40 mM iodoacetamide, 100 mM ammonium bicarbonate for 30 min at room temperature. They were digested in-gel with lysyl endopeptidase from A. lyticus (Wako Pure Chemical Industries) in 100 mM Tris-HCl (pH 8.9) for 15 h at 37°C. Resulting peptide fragments were extracted from gels and then concentrated in vacuo. After desalting with Zip-TipC18 (Millipore), peptide fragments were subjected to mass spectrometric analysis. Mass spectra were acquired by direct infusion analysis on Micromass Q-Tof2 hybrid quadrupole time-of-flight mass spectrometer equipped with a nano-electrospray ionization source, in positive mode. Tandem mass spectrometry (MS/MS) was performed by collision-induced dissociation using argon as the collision gas.
Identification of EST Consensus Sequence-Identification of EST consensus sequence using mass spectrometric data was performed by peptide mass fingerprinting and product ion mass fingerprinting run by the MASCOT program (Matrix Science) in our in-house server. An EST consensus sequence data base, Tentative Consensus sequence data base of X. laevis (Version 3.1), was downloaded from the file transfer protocol server of the Institute for Genomic Research to our in-house MASCOT server.
cDNA Cloning and Sequencing of Xenopus Uroplakin III Gene-Total RNA was isolated from Xenopus liver or ovary according to the described method. mRNA was enriched using poly(A) columns (Amersham Biosciences) according to the manufacturer's instruction and reverse-transcribed with a reverse transcriptase provided by a kit (SuperScript TM , Invitrogen) and oligo(dT) [12][13][14][15][16][17][18] primers. The resulting first strand DNA was used as template for PCR reactions of Xenopus uroplakin III gene. Oligonucleotides used for PCR were: the sense primer, 5Ј-GGG CTG CTG ATG TGA GAG TGT ACC TGA CA-3Ј; the antisense primer, 5Ј-CAC AGG GAA GGT ATT CCT CCT CCT CCG CAT-3Ј; each corresponds to a part of the 5Ј-or the 3Ј-untranslated regions of the Xenopus uroplakin III sequence deposited in the Tentative Consensus data base (The Institute of Genomic Research Gene Index, www.tigr.org/tigr-scripts/tgi/T_reports.cgi?speciesϭXenopus), respectively. Pfu polymerase was used to amplify the intervening sequence. The PCR sample mixtures were amplified on a thermal cycler using programs with annealing temperature ranges of 60 -95°C. A fragment of the expected size was excised from a 2% agarose gel, cloned into the pCR2.1-TOPO vector (Invitrogen), and sequenced.
cDNA Constructs-cDNA encoding the extracellular domain (ED: amino acid residues 1-191) of xUPIII was obtained by PCR, using the pCR2.1-TOPO vector containing the full-length cDNA for xUPIII prepared as above. PCR primers used were: the sense primer, 5Ј-GGA TCC CCG GGA ATT CCC ATG GGT CCT TGG AGG TAA-3Ј and the antisense primer, 5Ј-GAC AGA ATT CCC ACC ACT CCT TCT GCC AGG-3Ј. Both primers contained a restriction site for EcoRI as indicated by underlining. After PCR amplification, the products were digested with EcoRI and ligated into the cognate site of the bacterial expression plasmid pGEX-2T (Amersham Biosciences) so that the products were expressed as glutathione S-transferase (GST) fusion protein (GST⅐xUPIII-ED). Alternatively, cDNA encoding the full-length UPIII was amplified by PCR using the pCR2.1-TOPO vector carrying xUPIII sequence. PCR primers used were: the sense primer, 5Ј-CTG CGA ATT CAT GGG TCC TTG GAG GTA TCT-3Ј and the antisense primer, 5Ј-AAA AGC TTG GCC TGC TGG GTA GCC GCA TAG-3Ј. The sense primer contained a restriction site for EcoRI, whereas the antisense primer contained that for HindIII, each as indicated by an underline. The products were digested with EcoRI and HindIII and subcloned into the pCMV vector (Stratagene, La Jolla, CA) that had been digested with the same restriction enzymes. Both constructs were confirmed in their orientation, length, and sequence by DNA sequencing. Preparation of cDNA encoding Xenopus Src2 gene (xSrc2) and construction of mammalian expression plasmids that will express FLAG-tagged wild type or kinase-active xSrc2 were done as described previously. 2 Expression of GST Fusion Xenopus Uroplakin III Extracellular Domain-Cultures of Escherichia coli strain DH5␣ were transfected with the pGEX-2T plasmid harboring the cDNA for the extracellular domain of xUPIII (xUPIII-ED) and grown at 37°C in L-broth (10 mg/ml Bactotryptone, 5 mg/ml Bacto-yeast extract, 5 mg/ml NaCl) containing 50 g/ml ampicillin. When the growth of bacteria reached at log phase, isopropyl-1-thio-␤-D-galactopyranoside (1.5 mM) was added, and the cultures were further continued for 2 h. Bacteria were collected by centrifugation at 2000 ϫ g for 5 min, and the pellet was sonicated in the presence of extraction buffer containing 1% Triton X-100, 20 mM Tris-HCl (pH 7.5), 1 mM EDTA, 1 mM EGTA, 10 mM ␤-mercaptoethanol, 10 g/ml leupeptin, and 20 M APMSF. The resulting insoluble material or "inclusion body" was collected, separated by SDS-PAGE on 8% gels, and used as antigen for immunization of rabbits.
Indirect Immunofluorescent Study-Localization of xUPIII on the surface of Xenopus eggs was determined by using a rabbit antibody to the extracellular domain of xUPIII (anti-xUPIII-ED antibody). All the following manipulations were carried out at room temperature. Dejellied unfertilized eggs were incubated with 1ϫ DB containing 10 mg/ml bovine serum albumin for 10 min and then treated with the anti-xUPIII-ED antibody (500-fold diluted antiserum) for 20 min. After the antibody treatment, the eggs were washed several times with 1ϫ DB containing 10 mg/ml bovine serum albumin and then treated with a goat anti-rabbit IgG conjugated with Alexa 488 (Molecular Probes, SDS-gels containing pp30 prepared as in B were cut out and subjected to in-gel digestion with lysyl endopeptidase followed by mass spectrometric analysis as described under "Experimental Procedures." Shown is the deconvoluted mass spectrum of the digest. The arrowheads indicate ion signals assigned to be a part of Xenopus uroplakin III ( Fig. 2A). Two ions with a mass value difference of 80 (‚ ϭ 80), which reflects the absence or the presence of one phosphate in the peptide fragment with the same sequence (see panel D), are also indicated. D, deconvoluted product ion mass spectrum of a phosphopeptide from pp30. A peptide fragment with an 80-Da increase to the cognate peptide as shown in panel C was further analyzed by MS/MS. Shown is a part of the spectrum that has been annotated to show amino acid residues. A phosphotyrosine residue could be detected following the Ile-Thr sequence.
Eugene, OR) at 5 g/ml in the same buffer solution for 15 min in the dark. After the treatment, the eggs were washed several times with 1ϫ DB and then directly examined under a confocal laser-scanning microscope (Model LSM510, Carl Zeiss, Oberkochen, Germany). To determine the localization of xUPIII on the surface of egg plasma membranes, we manually removed vitelline envelopes from the antibodytreated egg samples and examined the vitelline envelope-free eggs under the microscope.
Surface Bbiotinylation of Eggs-Dejellied unfertilized eggs were treated with 1ϫ DB containing 2 mg/ml sulfo-NHS-biotin for 5-10 min at room temperature. After the treatment, egg samples were washed several times with 1ϫ DB and subjected to extraction to prepare either the cytosolic and the membrane fractions or the raft and the non-raft fractions (see above). To detect biotinylated proteins in egg samples, we performed immunoprecipitation and immunoblotting with a mouse monoclonal antibody against biotin (Clone BN-34, Sigma).
Fertilization Assay in the Presence of anti-xUPIII ED Antiserum-When the effect of the anti-xUPIII ED antibody on fertilization was examined, a group of 10 dejellied, unfertilized eggs was placed in a polystyrene well (diameter: 8 mm) filled with 100 l of 0.1ϫ modified Ringer's solution and an equal volume of the preimmune antiserum or anti-xUPIII-ED antiserum diluted with 0.1ϫ modified Ringer's solution. The final concentrations of antiserum were dilutions of 1:100 -1: 10. After the incubation for 30 min, 100 l of serum/0.1ϫ modified Ringer's solution was replaced with an equal volume of jelly watertreated sperm. The final concentrations of sperm were 0.5-2.0 ϫ 10 7 /ml. We determined the rate of successful fertilization by scoring those that underwent cortical contraction within 30 min and first cleavage within 100 min after insemination.
Transfection of HEK293 Cells-HEK293 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C in a humidified 5% CO 2 atmosphere. Cells of 20 -30% confluence in 100-mm dishes were transfected with 1-2 g of plasmid DNA/dish as specified under "Results" using Effectene TM reagent (Qiagen, Hilden, Germany) according to the manufacturer's standard protocol. The transfection treatment proceeded for 24 h at 37°C, and the resultant transfected cells were serum-starved in Dulbecco's modified Eagle's medium for 18 h prior to cell extraction. After the transfection treatment, cells were washed twice with ice-cold phosphate-buffered saline and lysed in extraction buffer containing 20 mM Tris-HCl, pH 7.5, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 10 mM ␤-mercaptoethanol, 1 mM Na 3 VO 4 , 10 g/ml leupeptin, and 20 M APMSF. Cell lysates were vortex-mixed (10 s, 2 times), sonicated (30 s, 2 times), clarified by ultracentrifugation (300,000 ϫ g, 10 min), diluted to a protein concentration of 1 mg/ml with the extraction buffer, and used as Triton X-100-solubilized whole cell extracts.

RESULTS
To look for egg raft-associated proteins that are phosphorylated on tyrosine residues after fertilization, we performed immunoblotting of the egg raft fractions, which had been prepared from unfertilized eggs, fertilized eggs, or hydrogen peroxide (H 2 O 2 )-treated eggs, with anti-phosphotyrosine antibody. As shown in Fig. 1A, there were several proteins in the raft fractions to be tyrosine-phosphorylated after the egg activation treatments. They include the 57-kDa xSrc and the 145-kDa PLC␥. In addition, there was a markedly tyrosine-phosphorylated protein that migrated with M r 30,000 in both fertilized and H 2 O 2 -treated eggs, so we termed this protein pp30 and conducted further experiments. The pp30 band was not detected in the non-raft fractions of any egg samples (not FIG. 2. Sequence of deduced amino acids of xUPIII and alignment with UPIIIs from mammalain species. A, amino acid sequence of pp30/xUPIII (265 amino acids) deduced from a Tentative Consensus sequence (TC71622) as well as from cDNA obtained from a Xenopus liver library is shown. The amino acid sequences matching those of the known partial sequences by MS or MS/MS analysis are underlined. Note that two ions containing oxidized or non-oxidized methionine 112 have been identified. An arrowhead between positions 10 and 11 marks the end of putative signal peptide. The circled asparagine residue at position 81 denotes a potential N-glycosylation site. An arrowhead between positions 188 and 189 marks the potential digestion site with cathepsin B, which prefers the glycine-arginine-arginine sequence (see "Discussion"). The dashed bar denotes the potential transmembrane domain. The asterisk indicates the position of tyrosine 249, which is phosphorylated in pp30. B, alignment of xUPIII and mammalian UPIIIs (human, NP008884; bovine, I45986; mouse, AAF34681; and rat, XP235546). The dashes indicate gaps inserted to optimize the alignments. The asterisks indicate identity. The dots indicate amino acids that are identical in more than three UPIII proteins. The ovals indicate the positions of potential N-glycosylation sites. The highly conserved nature of the conserved domains as well as the transmembrane domains is seen between all UPIIIs (underlined); of particular, note the glycine-arginine-arginine sequence.
Also note the complete conservation of the tyrosine residue equivalent to tyrosine-249 in xUPIII, as indicated by a rectangle.
shown), suggesting that pp30 and its unphosphorylated form are localized predominantly in the raft fractions before and after egg activation treatments.
To identify pp30, we purified it by immunoprecipitation by using the anti-phosphotyrosine antibody PY99. As shown in Fig. 1B, successful concentration and purification of pp30 from the SDS-solubilized raft fractions of H 2 O 2 -treated eggs was confirmed by means of silver staining of proteins as well as anti-phosphotyrosine immunoblotting of the immunoprecipitates. Gels containing pp30 were excised and directly treated with lysyl endopeptidase. The in-gel digest was subjected to mass spectrometry. The peptide mass list and product ion mass list obtained by MS/MS were searched against the non-redundant data base of NCBI by using the MASCOT program. The peptide mass fingerprinting assigned pp30 as a channel protein. However, in product ion mass fingerprinting (MS/MS Ions Search) by the MASCOT, none of the product ion mass lists obtained from MS/MS of five peptide fragments matched to the channel protein. Therefore, product ion mass lists of five peptide fragments were then searched against an EST consensus sequence data base, the Tentative Consensus sequence data base of X. laevis. All MS/MS data from five peptide fragments matched to an EST consensus sequence, TC71622 ( Fig. 2A). Peptide mass fingerprinting demonstrated that the preparation of pp30 includes at least two proteins; one is the channel protein, and the other is a hypothetical protein coded by TC71622.
The EST consensus sequence TC71622 contains an open reading frame encoding a polypeptide consists of 265 amino acid residues (Fig. 2A). The calculated molecular size of the full-length polypeptide, which includes the signal sequence, is 29,239 Da, which is near the relative molecular size (30 kDa) estimated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). According to the nucleotide sequence, we performed PCR using Xenopus liver and ovary cDNA as template. Although there are four nucleotide differences between the open reading frame nucleotide sequences of the PCR product and the TC71622, all nucleotide substitutions do not affect the translated amino acid sequence. The deduced amino acid sequence is homologous to a mammalian protein, UPIII. Thus, we concluded that pp30 is a Xenopus homologue of uroplakin III, xUPIII. MS/MS analysis demonstrated that a tyrosine residue (Tyr 249 ) in the carboxyl-terminal region of xUPIII is phosphorylated (Fig. 1D).
Shown in Fig. 2B is an alignment of amino acid sequences of xUPIII and UPIIIs from mammalian species. As reported for mammalian UPIIIs, xUPIII is predicted as a single transmembrane protein that has a long extracellular domain in its amino-terminal sequence (residues 1-191) followed by a transmembrane domain (residues 192-219). The phosphorylated Tyr 249 is included in a short cytoplasmic tail in the carboxyl terminus (residues 220 -265), and this tyrosine residue is conserved in all UPIIIs. The extracellular domain of xUPIII contains only one potential N-glycosylation site at position 81 (Asn-Tyr-Thr), whereas mammalian UPIIIs contain three glycosylation sites (Fig. 2B). Although the entire sequence identity between xUPIII and other UPIIIs is not so high (e.g. xUPIII versus human UPIII is 37%), a region containing dibasic amino acids (Arg 187 -Arg 188 ) near the transmembrane region (residues 181-205), called the "conserved domain" (30), is highly identical to other UPIIIs (xUPIII versus all other UPIIIs is 84%). The functional importance of the conserved domain in UPIII has not yet been demonstrated (see "Discussion").
To analyze the structure and function of xUPIII in Xenopus eggs, we prepared two different rabbit polyclonal antibodies. One antibody was raised against the bacterially expressed GST fusion protein containing the extracellular domain (residues 1-191) of xUPIII and termed anti-xUPIII-ED antibody. Another antibody was raised against an xUPIII carboxyl-terminal peptide that corresponds to residues 244 -265 and termed anti-xUPIII-CT antibody. Immunoblotting analysis of various Xenopus tissues with these antibodies demonstrated that a 30-kDa immunoreactive protein (i.e. xUPIII) is present in urinary tract, ovary, kidney (Fig. 3A), and egg (Figs. 3B and 4). We detected no or little xUPIII band in testis, lung, liver, heart, and skeletal muscle (Fig. 3A). It should be noted that although we obtained cDNA encoding xUPIII from liver and ovary cDNA libraries, we failed to detect xUPIII protein in liver (Fig. 3A).
It has been shown that mammalian UPIIIs are highly Nglycosylated and that removal of the N-linked sugars results in a dramatic change of mobility of UPIII in SDS-PAGE (31). To determine whether xUPIII is also N-glycosylated or not, raft fractions were prepared from Xenopus unfertilized eggs and subjected to deglycosylation treatment with either endoglycosidase H or N-glycosidase F. The resulting protein samples were analyzed by immunoblotting with anti-xUPIII antibodies. As shown in Fig. 3B, N-glycosidase F treatment caused a mobility shift of xUPIII from 30 to 25 kDa. Endoglycosidase H treatment did not show such an effect. The result clearly demonstrates that UPIII contains an N-linked sugar(s), as suggested by its deduced amino acid sequence. Importantly, not only anti-xUPIII CT antibody but also anti-xUPIII ED antibody recognizes both intact and deglycosylated forms of xUPIII in immunoblotting.
We next determined whether xUPIII localizes to the raft fractions of Xenopus eggs. To this end, sucrose density gradient centrifuge fractions of unfertilized eggs were analyzed by immunoblotting with anti-xUPIII-ED antibody. As shown in Fig.  4A, xUPIII was clearly detected as highly concentrated bands in the raft fractions (fractions 3-6), whereas no clear band was detected in the non-raft fractions (fractions 10 -12). A similar pattern of the anti-xUPIII-ED immunoblots was obtained with the sucrose fractions of fertilized and H 2 O 2 -treated eggs (not shown). Thus, we conclude that xUPIII is localized to the raft fractions of eggs before and after egg activation. By using anti-xUPIII antibodies, we could also confirm that pp30, which has been initially identified by an anti-phosphotyrosine antibody (Fig. 1, A and B), is xUPIII. As shown in Fig. 4B, anti-xUPIII-ED antibody recognized pp30 that had been prepared by immunoprecipitation of the raft fractions of H 2 O 2 -treated eggs with antiphosphotyrosine antibody. Reciprocally, the same anti-phosphotyrosine antibody efficiently recognized xUPIII that had been prepared by immunoprecipitation of the raft fractions of H 2 O 2treated eggs with anti-xUPIII-ED antibody. Similar results were obtained when anti-xUPIII-CT antibody was used instead of anti-xUPIII-ED antibody (data not shown).
We next performed a whole mount immunocytochemical analysis of live, dejellied, and unfertilized Xenopus eggs with anti-xUPIII-ED antibody. The antibody-specific indirect fluorescent signal was evident over the entire egg surface, as visualized by confocal laser-scanning microscopy (Fig. 4D). No visible signal was detected when the preimmune antibody was used at the same antibody concentration (Fig. 4E). To further evaluate the localization of xUPIII, we set a focal plane of the antibody-treated egg to the cell surface (Fig. 4F) and analyzed magnified images. As shown in Fig. 4G, the antibody-specific signal was not uniformly localized on the egg surface but rather showed a scattered pattern, which avoids multiple areas of more than 1-m diameter. This may reflect a localization of UPIII on the tip and/or the other areas of microvilli in the plasma membranes but not areas just above the cortical granules in the egg cytoplasm. Again, such localization pattern was not obtained with the preimmune antibody (Fig. 4H). It is possible that xUPIII is present on the vitelline envelope, which is present just above the egg plasma membranes. Thus, we . The immunoprecipitates were separated by SDS-PAGE on 12.5% gels and analyzed by immunoblotting (IB) with either anti-phosphotyrosine or anti-xUPIII-ED antibody. The positions of xUPIII, the mobility of which on SDS-gels matches exactly with that detected by PY99, are indicated. Also indicated are the positions of heavy chains (H.C.) and light chains (L.C.) of IgG used for immunoprecipitation. D-I, indirect immunofluorescent study of xUPIII. Xenopus unfertilized eggs were treated successively with the primary antibody (anti-xUPIII-ED antibody or preimmune antibody) and the second antibody (Alexa 588-conjugated anti-rabbit IgG) and were directly analyzed under the laser-scanning confocal microscopy as described under "Experimental Procedures." The cortical area of the egg is stained by the anti-xUPIII-ED antibody (D) but not by the preimmune antibody (E). The same sample as in D was analyzed in the same view but in a different focal plane to show staining image of the egg surface (F). G and H, magnified views of a part of the egg surface stained with the anti-xUPIII-ED antibody (G) or the preimmune antibody (H). I, another magnified view of the egg surface stained with the anti-xUPIII-ED antibody after manual removal of vitelline membranes. Scale bars, 100 m in D-F, 250 m in G-I. removed the vitelline envelope manually from the same egg as analyzed in Fig. 4G and re-examined it. As shown in Fig. 4I, we observed a similar image of the antibody-specific signal. Thus, we concluded that xUPIII is exposed on the surface of the egg plasma membranes.
To further evaluate the cell surface localization of xUPIII on egg plasma membranes, we performed surface biotinylation of Xenopus eggs. Dejellied unfertilized eggs were treated with sulfo-NHS-biotin, a membrane non-permeable biotinylation reagent, and fractionated into rafts and non-rafts. Immunoblotting showed that a protein band of 30 kDa was biotinylated in both the raft fractions (Fig. 5A, lanes 3-6) and the non-raft fractions (lanes 10 -12). This protein was identified as xUPIII because anti-xUPIII-ED antibody could immunoprecipitate effectively the biotinylated 30-kDa protein, and reciprocally, anti-biotin antibody could immunoprecipitate effectively xUPIII (Fig. 5B). Fig. 5B also shows that in the raft fractions, xUPIII is a predominantly biotinylated protein.
Cell surface localization of xUPIII as determined above suggests that this protein is involved in sperm-egg interaction and/or fusion. To assess this possibility, we examined the effect of anti-xUPIII-ED antibody on fertilization of Xenopus eggs. Unfertilized eggs were preincubated with the different concentrations of antiserum and then inseminated with the different concentrations of sperm. Fertilization was scored by occurrence of cortical contraction (within 20 min of insemination) and first cell cleavage (within 90 min of insemination). As shown in Fig. 6, anti-xUPIII-ED antibody, which would bind to the extracellular domain of xUPIII, inhibited fertilization in a concentration-dependent manner (solid lines). The inhibitory effect of the antibody was more evident when the concentration of fertilizing sperm was reduced (Fig. 6). Anti-xUPIII-CT antibody, which would bind to the cytoplasmic tail of xUPIII (data not shown). Preimmune rabbit serum at the highest concentration (1:10 dilution) also did not show inhibitory effect (Fig. 6, dotted lines). In these conditions, preimmune IgG could bind to the egg surface to some extent (data not shown), indicating that the effect of anti-xUPIII ED antibody is not simply due to the binding of IgG to the egg surface but rather due to the specific binding to xUPIII on the egg surface. The effect of anti-xUPIII-ED antibody was not seen when sperm, but not eggs, were pretreated with the antibody (data not shown).
The rapid phosphorylation of xUPIII on tyrosine residues in the egg rafts suggests that a raft-associated tyrosine kinase is responsible for its phosphorylation. As a 57-kDa Src tyrosine kinase, termed xSrc, also localizes to the egg rafts and is activated rapidly upon egg activation, we wanted to determine whether xSrc interacts with and phosphorylates xUPIII. To this end, we performed transient coexpression of xSrc and xUPIII in HEK293 cells. In Fig. 7A, the cytosol and the detergent-solubilized membrane fractions were prepared from HEK293 cells expressing xUPIII or vector alone and analyzed by immunoprecipitation and immunoblotting with anti-xUPIII-ED antibodies. The results clearly demonstrate that xUPIII is expressed efficiently in the membrane fractions. We next analyzed the sucrose density gradient fractions of HEK293 cells expressing both wild type xSrc and xUPIII. As shown in Fig. 7B, both xSrc and xUPIII were detected in the raft fractions (fractions 3-6). Thus, we concluded that functional expression of both xSrc and xUPIII are successfully done in HEK293 cells. We should note, however, that a larger amount of xUPIII is present in non-raft fractions (Fig. 7B, lanes 10 -12, see "Discussion"). To determine the functional interaction between xSrc and xUPIII in HEK293 cells, we performed immunoprecipitation and immunoblotting of the membrane fractions. As shown in Fig. 7C, physical interaction between kinase-active xSrc and xUPIII was detected only when these proteins were coexpressed in HEK293 cells. Tyrosine phosphorylation of xUPIII was also observed under the same conditions, whereas an overall increase of tyrosine phosphorylation in HEK293 cells was solely dependent on the expression of xSrc. These results indicate that xSrc and xUPIII can interact with each other to facilitate tyrosine phosphorylation of xUPIII. A group of 10 Xenopus unfertilized eggs were dejellied and pretreated in the presence of the indicated concentrations of rabbit antiserum against the xUPIII-ED (anti-xUPIII-ED, solid lines) or the preimmune rabbit serum (dotted lines) as described under "Experimental Procedures." After the preincubation, the eggs were subjected to insemination with the indicated concentrations of jelly water-treated sperm. Concentrations of sperm used are 2.0 ϫ 10 7 /ml (closed circles), 1.0 ϫ 10 7 /ml (closed triangles), and 0.5 ϫ 10 7 /ml (closed squares). Fertilization was determined by the appearance of cortical contraction and first cleavage. Under different sperm concentrations, fertilization of eggs in the absence of anti-xUPIII ED antiserum was more than 90% and taken as 100%. In all experimental conditions, more than 20 eggs were analyzed.

DISCUSSION
Here we show that pp30, a raft-associated 30-kDa protein that is tyrosine-phosphorylated in fertilized or H 2 O 2 -treated Xenopus eggs, is xUPIII. Our finding is based on the followings: 1) peptide mass fingerprinting analysis and annotation to the Tentative Consensus sequence encoding a hypothetical protein (Figs. 1 and 2A); 2) the sequence homology with mammalian UPIIIs (Fig. 2B); 3) the immunoreactivity with two different anti-xUPIII antibodies (Fig. 3); 4) the presence of the anti-xUPIII immunoreactive p30 in the urinary tract (Fig. 3A); and 5) the presence of N-linked sugars as mammalian UPIIIs (Fig.  3B). UPIII is a member of the uroplakin family proteins involving two tetraspanin molecules UPIa and UPIb (32), and two single transmembrane molecules UPII and UPIII (33,34).
The finding that pp30 is xUPIII was surprising to us because mammalian UP family proteins are known to be expressed exclusively in the urothelial tissues (31,35). In the urothelial tissues, all UP family proteins act as protein subunits of urothelial plaques that cover more than 90% of the apical urothelial surface (36). By constructing these UP proteinmediated rigid structures, the bladder epithelium gains the highly impermeable barrier (37,38). In the present study, by using specific antibodies to xUPIII, we confirmed that xUPIII is expressed in not only the urinary tract but also the egg, ovary, and kidney in adult Xenopus (Fig. 3A). Therefore, to our knowledge, this is the first report showing that one of UP family proteins, UPIII, is expressed in the reproductive tissues.
Since the initial cDNA cloning of UPIII in bovine tissues (31), it has been discussed that UPIII has a cytoplasmic domain containing possible sites of Ser/Thr/Tyr phosphorylations and thus may have a specific function connecting extracellular signals to intracellular functions. In this connection, our present study provides the first evidence that UPIII function could be modulated through tyrosine phosphorylation of the cytoplasmic domain. MS/MS analysis demonstrated that Tyr 249 in the cytoplasmic tail of xUPIII is phosphorylated. MS/MS analysis also gave a signal for the very carboxyl-terminal fragment of xUPIII containing another tyrosine residue (Tyr 259 ); however, its corresponding phosphopeptide was not detected. Thus, we think that Tyr 259 is not phosphorylated in xUPIII. The fact that only Tyr 249 , but not Tyr 259 , is completely conserved in all UPIIIs so far identified (Fig. 2B) suggests that phosphorylation of Tyr 249 serves a function of universal importance. The amino acid sequence surrounding Tyr 249 ( 244 PGDITYSSTLA 254 ) does not match with those that will bind to a certain kind of the Src homology 2 domains or the phosphotyrosine-binding domains. Further study will be directed to identify the role of the phosphorylated Tyr 249 in xUPIII function.
The deduced amino acid sequence of xUPIII suggests that its extracellular domain contains at least one N-glycosylation site (Fig. 2). Deglycosylation experiments support this possibility (Fig. 3B). This type of glycosylation in xUPIII seems to be universally occurred in Xenopus tissues, where UPIII is expressed, because immunoblotting analysis has shown that all FIG. 7. xUPIII is localized to the raft fractions and associates with Xenopus Src when coexpressed in HEK293 cells. A, expression of xUPIII in HEK293 cells. HEK293 cells, grown in a 100-mm dish, were transfected with pCMV vector alone or pCMV-xUPIII (2 g of DNA/dish) as described under "Experimental Procedures." After 24 h of transfection, the cells were collected and fractionated into the cytosolic and membrane fractions. The cell fractions (each 100 g of protein/lane) were immunoprecipitated with the antibody against xUPIII-ED. The immunoprecipitates were separated by SDS-PAGE on 10% gels and analyzed by immunoblotting with antibodies to xUPIII-ED. The arrowheads indicate the positions of xUPIII. B, xUPIII expressed in HEK293 cells is localized to the raft fractions. HEK293 cells were co-transfected with pCMV-xUPIII and p3xFLAG-CMV-14-xSrc (each 2 g of DNA/100-mm dish) for 24 h and subjected to raft preparation by sucrose density gradient fractionation. Proteins (each equivalent to the protein amount from ϳ0.1 dish/lane) were separated by SDS-PAGE on 12.5% gels and analyzed by silver staining or immunoblotting with antibodies to the FLAG epitope (IB: xSrc) or UPIII-ED (IB: xUPIII). The positions of the raft fractions, FLAG-tagged wild type xSrc and xUPIII are indicated. C, functional interaction between xUPIII and xSrc co-expressed in HEK293 cells. HEK293 cells were transfected with empty vector alone (lane 1) or pCMV-xUPIII and/or p3xFLAG-CMV-14 kinase-active xSrc (lanes 2-4) (each 2 g of DNA/100-mm dish) for 24 h. After transfection, Triton X-100-solubilized whole cell lysates (500 g/lane) were prepared and immunoprecipitated with antibodies to the xUPIII-ED (IP: xUPIII) or FLAG epitope (IP: xSrc). The immunoprecipitates were analyzed by immunoblotting with the anti-phosphotyrosine antibody (IB: PY99), anti-FLAG antibody (IB: xSrc) or anti-xUPIII-ED antibody (IB: xUPIII). The whole cell lysate (10 g/lane) was also analyzed directly by immunoblotting with antibodies to xUPIII-ED, the FLAG epitope (for xSrc), or phosphotyrosine (PY99). The positions of xUPIII and xSrc are indicated. When co-immunoprecipitation of xSrc and xUPIII was analyzed, antibodies used were chemically cross-linked to beads to avoid the appearance of the heavy chains that would mask xSrc. xUPIII bands detected are those migrating at 30 kDa on SDS-PAGE (Fig. 3A). The functional importance of this modification is not known.
Immunoblotting analyses have demonstrated that xUPIII localizes predominantly to the egg rafts (Fig. 4A). Localization of xUPIII to rafts is also seen when it is expressed in HEK293 cells (Fig. 7B). However, as opposed to the case in Xenopus eggs, a larger amount of the expressed xUPIII is present in non-raft fractions in HEK293 cells (Fig. 7B). We think that xUPIII, when expressed alone, poorly localizes to the raft fractions in HEK293 cells. In fact, xUPIII can be localized to the rafts of HEK293 cells more efficiently when coexpressed with Xenopus uroplakin Ib (UPIb). 3 In support with this, it has been reported that in 293T cells, bovine UPIII localizes to the cell surface membranes only when expressed together with bovine UPIb (39,40). Thus, we suggest that in Xenopus eggs, as in the case of the cell expression system, xUPIII is coexpressed with UPIb to be targeted to the raft fractions.
In mammals, UP proteins form two different heterodimers involving the UPIa/UPII and UPIb/UPIII pairs (34,37). The fact that UPIb, a partner of UPIII, is a tetraspanin family protein is of our special interest because CD9, a member of the tetraspanin family, plays a critical role in sperm-egg fusion in the mouse (41)(42)(43). The tetraspanin family proteins are expressed on the cell surface and possess four conserved transmembrane domains (44). Considering the fact that anti-xU-PIII-ED antibody inhibits fertilization of Xenopus eggs (Fig. 6), it is attractive to surmise that xUPIII is involved in sperm-egg interaction and/or fusion via interaction with a tetraspanin molecule(s) such as UPIb or CD9. Expression of UPIb mRNA in Xenopus embryo has already demonstrated (GenBank TM accession number BC043899). Thus, in Xenopus eggs, UPIb could be the most likely binding partner of xUPIII. We are now investigating this possibility by preparing a specific antibody to Xenopus UPIb (xUPIb) and by preparing mammalian cells expressing xUPIII and xUPIb.
Biotinylation and immunofluorescence studies of dejellied Xenopus eggs have shown that xUPIII is a predominantly biotinylated protein in the egg rafts (Fig. 5) and that xUPIII is exposed on the egg surface (Fig. 4, D-I). These results are consistent with the fact that the anti-xUPIII-ED antibody inhibits fertilization (Fig. 6). Therefore, it is suggested that xU-PIII functions as a part of sperm-interacting machinery on the egg rafts. Until now, however, no report has been available as to the direct role of UPIII protein in any kind of ligand-receptor, cell-cell, or cell-substratum interactions. It is noteworthy that some reports have demonstrated the roles of UPIa and UPIb in urinary tract infection by uropathogenic bacteria (45,46). A more recent report has shown that the bacterial infection event involving UPIa occurs through rafts (47). On the other hand, a growing body of knowledge indicates the roles of rafts in cell infection events by not only pathogenic bacteria but also viruses or microorganisms (48). One possible structural implication of xUPIII function is the presence of dibasic amino acids (Arg 187 -Arg 188 ) in its extracellular domain ( Fig. 2A). As has been discussed by Wu and Sun (31), this extracellular dibasic sequence could be an enzymatic cleavage site. In this connection, Mizote et al. (49) have reported that Xenopus egg fertilization requires proteolytic activity derived from sperm and that Xenopus eggs can be activated by the purified, tryptic sperm protease with the similar substrate specificity to cathep-sin B or by the authentic cathepsin B. Therefore, it will be interesting to determine whether they are acting on xUPIII.