Localization of N-linked carbohydrate chains in glycoprotein ZPA of the bovine egg zona pellucida Keiichi Ikeda 1 , Naoto Yonezawa 1,2 , Keita Naoi 1 , Toshiyuki Katsumata 3 , Seizo Hamano 4 and Minoru Nakano 1,2 1 Graduate School of Science and Technology and 2 Department of Chemistry, Chiba University, Japan; 3 College of Liberal Arts and Science, Tokyo Medical and Dental University, Chiba, Japan; 4 Animal Bio-Technology Center, Livestock Improvement Association, Tokyo, Japan The zona pellucida, a transparent envelope surrounding the mammalian oocyte, consists of three glycoproteins, ZPA, ZPB and ZPC, and plays a role in sperm–egg interactions. In bovines, these glycoproteins cannot be separated unless the acidic N-acetyllactosamine regions of the carbohydrate chains are removed by endo-b-Galactosidase digestion. Endo-b-Galactosidase-digested ZPB retains stronger sperm- binding activity than ZPC. It is still unclear whether ZPA possesses significant activity. Recently, we reported that bovine sperm binds to Man 5 GlcNAc 2 , the neutral N-linked chain in the cow zona proteins. In this study, we investigated the localization of the sperm-ligand active high-mannose- type chain and the acidic complex-type chains in bovine ZPA. Three N-glycopeptides of ZPA, containing an N-gly- cosylation site at Asn83, Asn191 and Asn527, respectively, were obtained from endo-b-Galactosidase-digested ZPA. Of these glycosylation sites, only Asn527 is present in the ZP domain common to all the zona proteins. The carbohydrate structures of the N-linked chains obtained from each N-glycopeptide were characterized by two-dimensional sugar mapping analysis, while considering the structures of the N-linked chains of the zona protein mixture reported previously. Acidic complex-type chains were found at all three N-glycosylation sites, while Man 5 GlcNAc 2 was found at Asn83 and Asn191, but there was very little of this sperm- ligand active chain at Asn527 in the ZP domain of ZPA. Keywords: glycoprotein; N-linked carbohydrate chain; sperm ligand; zona pellucida; ZP domain. The mammalian oocyte is coated with a transparent matrix called the zona pellucida. This matrix plays various roles in the early phase of fertilization: species-specific sperm binding, blocking polyspermy, and protecting the embryo until implantation [1,2]. The zona is composed of three glycoproteins, called ZPA, ZPB and ZPC in the order of the size of their cDNAs [3], and their carbohydrate chains are responsible for species-specific sperm-zona binding [1,4]. In the pig, the carbohydrate structures of O-linked chains of zona protein mixture [5,6] and the acidic and neutral N-linked chains of ZPB/ZPC mixture [7,8] are well charac- terized. Sperm-binding activity has been ascribed to the O-linked chains of the zona protein mixture [9] or to the neutral N-linked chains obtained from the ZPB/ZPC mixture [8]. We have shown that the triantennary/tetraan- tennary chains of the ZPB/ZPC mixture possess greater activity than the diantennary chains [10]. The triantennary/ tetraantennary chains of ZPB are localized at Asn220 in the N-terminal region [10]. Moreover, the isolated N-terminal peptide of ZPB including Asn220 has sperm-binding activity [11]. Yurewicz et al. [12] showed that ZPB and ZPC form heterocomplexes that have sperm-binding activity, but that monomeric ZPB or ZPC does not exhibit this activity. These results suggest that ZPC contributes to the expression of the sperm-binding activity of the neutral N-linked chain of ZPB. In ZPC, the triantennary/tetraantennary chains are mainly localized at Asn271 in the C-terminal region [13]. The sperm-binding activity of porcine ZPA, a minor component, has not been assessed because its purification is difficult. The N-linked chains of the bovine zona protein mixture include neutral (23%) and acidic (77%) chains [14]. We have determined the structures of the neutral chain and the core regions of the acidic chains [14]. The acidic chains are diantennary, triantennary and tetraantennary, fucosylated complex-type chains with a tandem N-acetyllactosamine repeat in the nonreducing regions to which the main sialic acids are attached. The neutral chain is a high-mannose- type chain with five mannose residues (Man 5 GlcNAc 2 ). This high-mannose-type chain exhibits sperm-binding activity and inhibits in vitro fertilization [15]. The native bovine zona proteins cannot be separated into their three components because their molecular masses are almost identical (74 kDa) and their carbohydrate chains are very heterogeneous [16]. When they are digested with endo- b-Galactosidase to remove the acidic N-acetyllactosamine repeat, the three components become separable [16–18]. Sperm-binding activity is mainly ascribed to ZPB, while Correspondence to M. Nakano, Department of Chemistry, Faculty of Science, Chiba University, 1-33, Yayoi-cho, Inage-ku, Chiba-shi, Chiba, Japan 263-8522. Fax: + 81 43 2902874, E-mail: mnakano@scichem.s.chiba-u.ac.jp Abbreviations: BrCN, cyanogen bromide; LCA, Lens culinaris agglutinin; ConA, Canavalia ensiformis agglutinin. Enzymes: b-Galactosidase (EC 3.2.1.23); keratan-sulfate endo-1,4-b-Galactosidase (EC 3.2.1.103); sialidase (EC 3.2.1.18); trypsin (EC 3.4.21.4); N-glycanase (EC 3.5.1.52) (Received 22 April 2002, revised 21 June 2002, accepted 12 July 2002) Eur. J. Biochem. 269, 4257–4266 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03111.x ZPC has weak activity. However, it is unclear whether ZPA possesses sperm-binding activity using in vitro competition and sperm-bead binding assays [18]. We deduced the amino acid sequence of the bovine ZPA by cDNA cloning and sequencing, and revealed that bovine ZPA has four potential N-glycosylation sites [18], whereas porcine ZPA has six potential sites [3]. Here, we describe the localization of the N-linked chains in bovine ZPA. MATERIALS AND METHODS Preparation of zona protein mixture Zonae pellucidae of bovine eggs were isolated from frozen ovaries as described previously [16] and were solubilized in H 2 Oat70°C for 30 min and lyophilized. The yield of bovine zona proteins from one ovary is approximately one- tenth of that of porcine zona proteins. Total 1.2 mg of protein mixture (60 000 zonae) from 2400 ovaries were used for the present experiment. The heat solubilized zonae were digested with Escherichia freundii endo-b-Galactosidase (Seikagaku co., Tokyo, Japan) in 0.5 M ammonium acetate (pH 5.6) at 37 °C for 48 h [8]. Tryptic digestion and cyanogen bromide cleavage of zona protein mixture After the endo-b-Galactosidase-digested zona protein mix- ture was reduced and carboxymethylated as described previously [13], the protein mixture (400 lg) was digested with trypsin (enzyme/substrate 1 : 100, w/w) in NaCl/P i (pH8.0)at37°C for 18 h. The protein mixture (400 lg) was also cleaved by cyanogen bromide (BrCN) in 70% formic acid at room temperature for 18 h in the dark and the products were desalted by using a Nucleosil 300–7C 18 column (4 · 150 mm; Machery-Nagel, Duren, Germany). Elution was performed by a gradient of acetonitrile from 0 to 50% in 0.1% trifluoroacetic acid at a flow rate of 1mLÆmin )1 at 37 °C. Lectin affinity chromatography of glycopeptides Tryptic peptides and BrCN-peptides dissolved in 500 lL of buffer A (1 m M calcium chloride, 1 m M magnesium chloride, 150 m M sodium chloride, 0.02% sodium azide, 10 m M Tris/HCl, pH 7.2) were separately applied to a Lens culinaris agglutinin (LCA) agarose column (13 · 23 mm; Seikagaku co.) equilibrated with buffer A. After elution with 21 mL of buffer A, the materials retained in the column were eluted with 0.2 M methyl-a- D -manno- pyranoside/buffer A. The flow-through fraction was applied to a Canavalia ensiformis agglutinin (ConA) agarose column (13 · 23 mm; Seikagaku co.). Elution conditions were the same as those in the case of the LCA column. Fractionation of N-glycopeptides The LCA-binding fraction and the ConA-binding fraction were separately applied to a Chemcosorb 3C 8 HPLC column (4.6 · 150 mm; Chemco, Osaka, Japan) equili- brated with 0.1% trifluoroacetic acid and eluted by a linear gradient from 0 to 50% acetonitrile in 0.1% trifluoroacetic acid over 50 min at a flow rate of 1 mLÆmin )1 at 37 °C. Effluent was monitored at 230 nm. The N-terminal amino acid sequence of each peak was determined by automated Edman degradation using a PPSQ-21 protein sequencer (Shimadzu, Kyoto, Japan). Purification of ZPA The endo-b-Galactosidase-digested zona protein mixture was fractionated into three components (ZPA–C) on the Nucleosil 300–7C 18 column [18] at 40 °C. N-glycanase digestion of ZPA We showed that ZPA is specifically cleaved between Ala167 and Asp168 by an unknown protease on fertilization, giving a short N-terminal fragment (36–167) and a long C-terminal fragment (168–637), which are crosslinked via a disulfide bond [16,17]. This site is also cleaved in at least half of the ZPA molecules during preparation of the zona from ovarian eggs. The molecular masses of the N- and C-terminal fragments of the endo-b-Galactosidase-digested ZPA are 21 and 63 kDa, respectively [12], and the time course of the N-glycanase (glycopeptidase F; Takara Shuzo, Kyoto, Japan) digestion of these fragments could be monitored by SDS/PAGE on 15 and 8% acrylamide gels, respectively, under reducing conditions [17]. The digestion was performed in 0.1% SDS, 10 m M o-phenanthroline, 100 m M sodium phosphate (pH 8.6) at 37 °C and aliquots were removed at various times between 0 min and 22 h and subjected to SDS/PAGE. To detect the shift of the 21 kDa band, 2.5 mU of N-glycanase was added to 4 lg (56 pmol) of ZPA. To detect the shift of the 63 kDa band, 0.25 mU of N-glycanase was added to 1 lg (14 pmol) of ZPA. Isolation of core regions of N-linked oligosaccharides from glycopeptides After reduction and carboxymethylation [13], the purified ZPA (21 lg) was digested with trypsin under the conditions described above. Tryptic digests were applied to a Chem- cosorb 3C 8 column. Because elution conditions were the same as in the case of the separation of N-glycopeptides from the zona protein mixture (Fig. 1), the elution times of the tryptic N-glycopeptides from ZPA were the same as those shown in Fig. 1A and B. Each fraction containing N-glycopeptide was digested with N-glycanase (0.5 mU) in 0.1 M Tris/HCl (pH 8.6) at 37 °C for 18 h. The digests were applied to tandem-linked columns of AG50W-X8 (10 · 15 mm; Bio-Rad, California, USA) and AG3 (10 · 15 mm, Bio-Rad), and the columns were eluted with water. The flow-through fraction was collected. After lyophilization, the N-linked chains were then digested with Jack bean b-Galactosidase (Seikagaku co.) in 50 lLof 0.1 M sodium citrate (pH 4.1) at 37 °C for 24 h. After the pH of the digestion mixture was adjusted to 8.0 by 1 M Tris/ HCl (pH 8.6), the solutions were applied to the AG3 column. The flow-through fraction with water was collected and lyophilized. Pyridylamination of N-linked oligosaccharides The N-linked oligosaccharides thus obtained were modified with 2-aminopyridine and sodium cyanoborohydride [20]. 4258 K. Ikeda et al.(Eur. J. Biochem. 269) Ó FEBS 2002 The excess reagents were removed by phenol/chloroform (1 : 1, v/v) extraction performed three times. The aqueous phase was applied to a Cellulofine GCL25-sf HPLC column (7.5 · 600mm;Seikagakuco.),andelutedwith10% acetonitrile/10 m M ammonium acetate at a flow rate of 0.3 mLÆmin )1 . The fluorescence intensity of the effluent was monitored with excitation at 320 nm and emission at 400 nm. Separation of pyridylaminated neutral and acidic chains The pyridylaminated N-linked oligosaccharides were sepa- rated into neutral and acidic fractions by using a TSK-gel DEAE-5PW HPLC column (7.5 · 75 mm; Tosoh, Tokyo, Japan) equilibrated with NH 3 /H 2 O (pH 9.0). Elution was performed by a linear gradient from 0 to 100% of 0.5 M CH 3 COONH 4 (pH 8.0) over 60 min at a flow rate of 0.5 mLÆmin )1 . The fluorescence intensity was monitored with excitation at 310 nm and emission at 375 nm. Sugar mapping analysis of the pyridylaminated N-linked chains The pyridylaminated neutral fraction of N-linked chains was chromatographed on a Shim-pack CLC-ODS column (6 · 150 mm; Shimadzu) by a linear gradient from 0.1 to 0.25% 1-butanol in 10 m M sodium phosphate (pH 3.8) for 60 min at a flow rate of 1 mLÆmin )1 at 55 °C[14].The fluorescence intensity was monitored with excitation at 320 nm and emission at 400 nm. Major peaks were further chromatographed on a size-fractionation HPLC column of TSK gel Amide-80 (4.6 · 250 mm; Tosoh). Elution was performed by a linear gradient decrease in acetonitrile from 65 to 50% in 0.5 M acetic acid/triethylamine (pH 7.3) over 60 min at a flow rate of 1 mLÆmin )1 at 40 °C. The fluorescence intensity was monitored with excitation at 310 nm and emission at 375 nm. The pyridylaminated acidic fractions were also analyzed using both columns after digestion with Arthrobacter ureafaciens sialidase (Nacalai tesque, Kyoto, Japan) in 0.1 M ammonium acetate (pH 5.0) at 37 °C for 24 h. The glucose units of the pyridylaminated N-linked chains were calculated from their retention times using the pyridylaminated glucose oligomers as the standard [14,21]. RESULTS Identification of N-glycosylation sites The acidic chains of the bovine zona protein mixture are diantennary, triantennary and tetraantennary, sialylated complex-type chains with a fucose residue at their reducing ends, while the neutral chain is a high-mannose-type chain [14]. The endo-b-Galactosidase-digested zona protein mix- ture was cleaved using trypsin and BrCN. The tryptic peptides and BrCN-peptides were separately applied to an LCA agarose column, which binds to the Fuc(a1–6)Glc- NAc of the complex-type chains [22]. The glycopeptides bound to the LCA column were eluted with 0.2 M a-methyl- D -mannopyranoside. The flow-through fraction from the LCA column was applied to a ConA agarose column, which binds to the high-mannose-type chain. The glyco- peptides retained in the column were eluted with 0.2 M a-methyl- D -mannopyranoside. The LCA-binding fraction (Fig. 1A,C) and the ConA-binding fraction (Fig. 1B,D) were then fractionated by reverse-phase HPLC separately. N-terminal amino acid sequence analysis of each peak of the tryptic peptides revealed that the peptides in the three peaks in Fig. 1A and B (peak a at 19 min, peak b at 34 min and peakcat38min)arefragmentsofZPA.Thatis, Fig. 1. Reverse-phase HPLC of N-glycopep- tides retained to LCA and ConA agarose. Reduced and carboxymethylated zona protein mixture was cleaved by trypsin and BrCN and appliedtoanLCAagarosecolumn.Theflow- throughfractionfromLCAcolumnwas applied to a ConA agarose column. Fractions eluted with 0.2 M methyl-a- D -mannopyrano- side from both columns were subjected to Chemcosorb 3C 8 HPLC. Elution was per- formed with a linear gradient of acetonitrile (broken line) in 0.1% trifluoroacetic acid. (A and B) Tryptic peptides; (C and D) BrCN- peptides; (A and C) LCA binding fraction; (B and D) ConA binding fraction. Peptides were detected at 230 nm. Arrowheads indicate the elution positions of the N-glycopeptides from ZPA. Ó FEBS 2002 Localization of N-linked chains in ZPA (Eur. J. Biochem. 269) 4259 VLXRTDPNIK (peak a), MLXCTYVLDP (peak b) and AQXLTLQEALTQGYNLLIEN (peak c) correspond to the N-terminal sequences of tryptic peptides 525–534, 81–92 and 189–210, respectively, in which X is a glycosylated Asn residue (Fig. 3). Furthermore, the three peaks in Fig. 1C (peak 1 at 34 min, peak 2 at 38 min and peak 3 at 40 min) were shown to be fragments of ZPA. That is, LXCTYVLDPE (peak 1), GWTVTVGDGE (peak 2) and LINTNVESLP (peak 3) correspond to the N-terminal sequence of BrCN-peptides 82–112, 178–211 and 468–546, respectively (Fig. 3). The absence of peak 3 in Fig. 1D agrees with the fact that there is only a small amount of the high-mannose-type chain at Asn527 (see below). Collec- tively, Asn83, Asn191 and Asn527 of ZPA are glycosylated. Determination of the number of N-glycosylation sites We then investigated the number of N-glycosylation sites and the localization of N-linked chains in ZPA using the purified ZPA. In the 15% gel, the 21 kDa band shifted to 16 kDa within 2 min and no further time-dependent shift was observed within 22 h (Fig. 2A), indicating that one N-glycosylation site is present in the N-terminal fragment. This fragment has a potential N-glycosylation site at Asn83 (Fig. 3). In the 8% gel, the 63 kDa band shifted to 59 and 56 kDa bands within five minutes, but the 59 kDa band did not converge on the 56 kDa within 22 h (Fig. 2B). The carbohydrate chains linked to one of the two sites may be resistant to N-glycanase, although the reason is uncertain. These results confirmed that two N-glycosylation sites are present in the C-terminal fragment and that the N-glycosylation sites of ZPA are Asn83, Asn191 and Asn527. Sugar mapping of the carbohydrate chains from each N-glycosylation site Three tryptic N-glycopeptides from the endo-b-Galactos- idase-digested ZPA were eluted at the same positions as peaks a–c in Fig. 1. N-linked chains were obtained by N-glycanase digestion of these tryptic peptides. We iden- tified the structures of the N-linked chains of ZPA by two- dimensional mapping on HPLC referring to the reported structures of the unfractionated zona protein mixture [14]. To obtain the core carbohydrate chains of the acidic chains, the N-linked chains were further digested with b-Galactosidase [14]. After pyridylamination, the N-linked chains were separated into neutral and acidic fractions by DEAE-5PW HPLC (Fig. 4). The elution position of the acidic fraction was the same as that of the monosialylated chain, and all the acidic chains in this fraction were neutralized by sialidase digestion. From the peak area, the molar ratios of the neutral chains to the acidic chains linked to Asn83, Asn191 and Asn527 were estimated to be 78 : 22, 67 : 33 and 67 : 33, respectively. The neutral chains were subjected to two-dimensional sugar mapping and the acidic chains were also analyzed by mapping after digestion with sialidase. The glucose units in the mapping are summarized in Table 1. The high-mannose-type chain, Fig. 3. N-glycosylation sites of bovine ZPA. The amino acid sequence of bovine ZPA, as deduced from cDNA and N-terminal amino acid sequence analyses [16,18]. Solid and dotted underlines indicate the N-terminal amino acid sequences of BrCN and tryptic N-glycopep- tides, respectively. Filled black boxes indicate the N-glycosylation sites determined by the present experiment and the broken box indicates a potential N-glycosylation site that is not glycosylated. The arrow indicates the N-terminus of mature ZPA and arrowhead indicates the specific cleavage site on fertilization. The putative processing site for furin or furin-like enzymes is double underlined. Closed area of Leu366-Ser632 indicates ZP domain [18,25] and the putative trans- membrane domain is from Thr683 to Leu697. Fig. 2. Time course of N-glycanase digestion of ZPA. At least half of the ZPA molecules are specifically cleaved by an unknown protease during preparation, giving rise to N-terminal 21 kDa fragment and C-terminal 63 kDa fragment, which are crosslinked through a disulfide bond. This ZPA specimen was digested with N-glycanase at 37 °C. At various times, aliquots were removed and subjected to SDS/PAGE on a 15% for the 21 kDa fragment (A) and 8% for the 63 kDa fragment (B) separating gel under reducing conditions. Proteins were silver- stained. The bands moving faster on the SDS gel than the band of undigested ZPA fragments (0 min) are indicated by arrowheads. 4260 K. Ikeda et al.(Eur. J. Biochem. 269) Ó FEBS 2002 which is the intact neutral chain [14], was found at Asn83 and Asn191, but was rare at Asn527. The other chains in the neutral fraction in Table 1 must be neutralized by endo-b-Galactosidase digestion, as their sialic acids are originally linked to the N-acetyllactosamine repeats. While all the chains in the neutral fraction were found previously by analysis of the unfractionated zona protein mixture [14], the monosialylated complex-type chains in the acidic fraction on DEAE-HPLC (sialidase digests of the acidic fraction in Table 1) were found in this analysis of ZPA. These sialic acids must be linked to the nonreducing terminal b-Gal residues of the core regions of the complex- type chains, as these sialylated b-Gal residues were unsusceptible to b-Galactosidase. In a previous experiment [14], sugar mapping analysis was applied to the N-linked chains from the zona protein mixture after digestion with sialidase, endo-b-Galactosidase and b-Galactosidase. Therefore, the nonreducing terminal b-Gal residues in the sialidase digests of the acidic fraction in Table 1 have been eliminated and the monosialylated chains converge on the chains in the neutral fraction. DISCUSSION The molar ratios of ZPA/ZPB/ZPC in the porcine and bovine zona pellucida are 1 : 3 : 3 and 1 : 1 : 2, respectively [17,23]. Recently, the neutral N-linked chains in porcine ZPB and ZPC were localized [10,13], but no information on the structures of these carbohydrate chains or their local- ization in the minor component porcine ZPA is available, because its purification is difficult. In this study, we localized a neutral high-mannose-type chain and acidic complex-type chains in bovine ZPA in order to fill a gap in the structural information on the zona glycoproteins. As with porcine ZPB and ZPC, bovine ZPA possesses three N-glycosylation sites (Fig. 2) determined to be Asn83, Asn191 and Asn527 in amino acid sequence analyses of tryptic and BrCN peptides. The acidic property of the N-linked chains of bovine zona proteins is due to sialic acids [14], whereas many sulfates are linked to the N-acetyllactosamine repeats of porcine zona proteins [7,24]. The N-linked chains of the bovine zona Fig. 4. Anion-exchange HPLC of N-linked chains from tryptic N-gly- copeptides. After b-Galactosidase digestion, N-linked chains of each tryptic N-glycopeptide from endo-b-Galactosidase digested ZPA were pyridylaminated and applied to a DEAE-5PW HPLC column equili- brated with NH 3 /H 2 O (pH 9). Elution was performed by a linear gradient and from 0 to 100% of 0.5 M CH 3 COONH 4 (pH 8) for 60 min at a flow rate of 0.5 mLÆmin )1 . The fluorescence intensity was monitored with excitation at 310 nm and emission at 375 nm. Arrow heads and arrows indicate the elution positions of neutral chains and monosialylated acidic chains, respectively. (A) peak a (Asn527); (B) peak b (Asn83) and (C) peak c (Asn191). Fig. 5. N-glycosylation sites in ZP domain among ZP components. N-glycosylation sites of bovine ZPA and porcine ZPB [10,11] and ZPC [13] in the ZP domains are shown by filled black boxes. Arrows and C indicate the linkage sites of sperm-ligand carbohydrate chains and positions of cysteine residues, respectively. Ó FEBS 2002 Localization of N-linked chains in ZPA (Eur. J. Biochem. 269) 4261 proteins possess one to several sialic acid residues and most of these are eliminated on fertilization [14]. However, participation of the sialic acids in bovine sperm-zona binding has not been demonstrated. Bovine sperm-binding activity is mainly ascribed to the high-mannose-type chain with five mannose residues [15]. The ZP domain is a module, approximately 260 resi- dues long, common to all three zona protein components [25–27] and this domain contains eight conserved cysteine residues, which form four disulfide linkages [25]. Bovine ZPA has only one N-glycosylation site in the ZP domain, while porcine ZPB and ZPC have three N-glycosylation Table 1. Glucose units of the N-linked chains from each of the N-glycosylation sites of bovine ZPA. The values in parentheses are glucose units of authentic pyridylaminated carbohydrate chains [14]. Molar ratio of the N-linked chains were calculated from the peak area. Proposed structures are from references [14,21]. PA, pyridylamino. Number of glucose unit Site N-linked chain Shim-pack CLC-ODS Amide 80 Structure Molar ratio 4262 K. Ikeda et al.(Eur. J. Biochem. 269) Ó FEBS 2002 sites in the domain (Fig. 5). Furthermore, porcine ZPA has one potential N-glycosylation site in the domain, Asn530. Therefore, the N-glycosylation sites in the ZP domain are not conserved in the three zona protein components. In the pig, triantennary and tetraantennary neutral complex-type chains have sperm-binding activity [10] and these active chains are mainly localized in the N-terminal region of the ZP domain of ZPB and the C-terminal region of the ZP domain of ZPC (Fig. 5). In the cow, ZPB exhibits the strongest sperm-binding activity among the components and ZPC exhibits sperm-binding activity approximately one-sixth that of ZPB [18]. However, it is still unclear whether bovine ZPA plays a role in sperm-zona binding. That is, ZPA exhibits weak sperm-binding activity in the competition assay, but does not exhibit activity in the sperm-bead binding assay [18]. Table 1 shows that bovine ZPA possesses the sperm- ligand active high-mannose-type chain, but there is very Table 1. (Continued). Number of glucose unit Site N-linked chain Shim-pack CLC-ODS Amide 80 Structure Molar ratio Ó FEBS 2002 Localization of N-linked chains in ZPA (Eur. J. Biochem. 269) 4263 Table 1. (Continued). Number of glucose unit Site N-linked chain Shim-pack CLC-ODS Amide 80 Structure Molar ratio In 83A1, 191A1 and 527A1, nonreducing terminal b-Gal residue attaches to GlcNAc residue of the a1-6 branch (*) or of the a1-3 branch (**). In 83A3, 191A3 and 527A3, nonreducing terminal b-Gal residue attaches to GlcNAc residue of the a1-6 branch ( # ) or of the b1-4 branch ( ## ). 4264 K. Ikeda et al.(Eur. J. Biochem. 269) Ó FEBS 2002 little of it at the N-glycosylation site, Asn527, in the ZP domain. This may explain why the sperm-binding activity of ZPA is weak, if any. The localization of the sperm- ligand active high-mannose-type chain in bovine ZPB and ZPC should be clarified. Besides the zona proteins, the ZP domain has been found in several proteins: TGF-b type III receptor (betaglycan) from fetal and adult tissues [28,29], uromodulin (urinary Tamm-Horsfall glycoprotein of pregnant woman) [30], the major zymogen granule membrane protein (GP-2) [31], and the proteins in the uterus and oviduct [32,33]. Furthermore, tectorins in the tectorial membranes responsible for hearing also have a ZP domain [34,35]. In each protein, the ZP domain is generally present next to a putative transmem- brane region [25]. The structural characterization of the carbohydrate chains in the ZP domain of these proteins is of interest. ACKNOWLEDGMENT This study was supported in part by a Grant-in Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. REFERENCES 1. Wassarman, P.M. (1988) Zona pellucida glycoproteins. Annu. Rev. Biochem. 57, 415–442. 2. To ¨ pfer-Petersen, E. (1999) Carbohydrate–based interactions on the route of a spermatozoon to fertilization. Hum. Reprod. Update 5, 314–329. 3. Harris, J.D., Hibler, D.W., Fontenot, G.K., Hsu, K.T., Yurewicz, E.C. & Sacco, A.G. (1994) Cloning and characterization of the zona pellucida genes and cDNAs from a variety of mammalian species: The ZPA, ZPB and ZPC gene families. DNA Seq. 4, 361– 393. 4. Litscher, E.S. & Wassarman, P.M. (1993) Carbohydrate-mediated adhesion of eggs and sperm during mammalian fertilization. Trends Glycosci. Glycotechn. 5, 369–388. 5. Hokke, C.H., Damm, J.B.L., Kamerling, J.P. & Vliegenthart, J.F.G. (1993) Structure of three acidic O-linked carbohydrate chains of porcine zona pellucida glycoproteins. FEBS Lett. 329, 29–34. 6. Hokke, C.H., Damm, J.B.L., Penninkhof, B., Aitken, R.J., Kamerling, J.P. & Vliegenthart, J.F.G. (1994) Structure of O-linked carbohydrate chains of porcine zona pellucida glyco- proteins. Eur. J. Biochem. 221, 491–512. 7. Noguchi, S. & Nakano, M. (1992) Structure of the acidic N-linked carbohydrate chains of the 55 kDa glycoprotein family (PZP3) from porcine zona pellucida. Eur. J. Biochem. 209, 883–894. 8. Noguchi, S., Hatanaka, Y., Tobita, T. & Nakano, M. (1992) Structural analysis of the N-linked carbohydrate chains of the 55 kDa glycoprotein family (PZP3) from porcine zona pellucida. Eur. J. Biochem. 204, 1089–1100. 9. Yurewicz, E.C., Pack, B.A. & Sacco, A.G. (1991) Isolation, composition, and biological activity of sugar chains of porcine oocyte zona pellucida 55K glycoproteins. Mol. Reprod. Dev. 30, 126–134. 10. Kudo, K., Yonezawa, N., Katsumata, T., Aoki, H. & Nakano, M. (1998) Localization of carbohydrate chains of pig sperm ligand in the glycoprotein ZPB of egg zona pellucida. Eur. J. Biochem. 252, 492–499. 11. Yonezawa, N., Mitsui, S., Kudo, K. & Nakano, M. (1997) Involvement of N-linked carbohydrate chains of pig zona pellucida in sperm-egg binding. Eur. J. Biochem. 248,86–92. 12. Yurewicz, E.C., Sacco, A.G., Gupta, S.K., Xu, N. & Gage, D.A. (1998) Hetero-oligomerization-dependent binding of pig oocyte zona pellucida glycoproteins ZPB and ZPC to boar sperm mem- brane vesicles. J. Biol. Chem. 273, 7488–7494. 13. Yonezawa, N., Fukui, N., Kudo, K. & Nakano, M. (1999) Localization of neutral N-linked carbohydrate chains in pig zona pellucida glycoprotein ZPC. Eur. J. Biochem. 260, 57–63. 14. Katsumata, T., Noguchi, S., Yonezawa, N., Tanokura, M. & Nakano, M. (1996) Structural characterization of the N-linked carbohydrate chains of the zona pellucida glycoproteins from bovine ovarian and fertilized eggs. Eur. J. Biochem. 240, 448– 453. 15. Amari,S.,Yonezawa,N.,Mitsui,S.,Katsumata,T.,Hamano,S., Kuwayama, M., Hashimoto, Y., Suzuki, A., Takeda, Y. & Nakano, M. (2001) Essential role of the nonreducing terminal a-mannosyl residues of the N-linked carbohydrate chain of bovine zona pellucida glycoproteins in sperm-egg binding. Mol. Reprod. Dev. 59, 221–226. 16. Noguchi, S., Yonezawa, N., Katsumata, T., Hashizume, K., Kuwayama, M., Hamano, S., Watanabe, S. & Nakano, M. (1994) Characterization of the zona pellucida glycoproteins from bovine ovarian and fertilized eggs. Biochim. Biophys. Acta 1201, 7–14. 17. Iwamoto,K.,Ikeda,K.,Yonezawa,N.,Noguchi,S.,Kudo,K., Hamano, S., Kuwayama, M. & Nakano, M. (1999) Disulfide formation in the bovine zona pellucida glycoproteins during fer- tilization: an evidence for the involvement of cystine cross-linkages in zona hardening. J. Reprod. Fertil. 117, 395–402. 18. Yonezawa, N., Fukui, N., Kuno, M., Shinoda, M., Goko, S., Mitsui, S. & Nakano, M. (2001) Molecular cloning of bovine zona pellucida glycoproteins ZPA and ZPB and analysis for sperm-binding component of the zona. Eur. J. Biochem. 268, 3587–3594. 19. Laemmli, U.K. (1971) Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature (London) 227, 680–685. 20. Kosakai, M. & Yoshizawa, Z. (1982) High performance liquid chromatography of pyridylamino derivatives of sulfated oligo- saccharides in the deamination products of porcine and whale heparins. J. Biochem. (Tokyo) 92, 295–303. 21. Tomiya, N., Awaya, J., Kurono, M., Endo, S., Arata, Y. & Takahashi, N. (1988) Analysis of N-linked oligosaccharides using a two-dimensional mapping technique. Anal. Biochem. 171, 73–90. 22. Kornfeld, K., Reitman, M.L. & Kornfeld, R. (1981) The carbo- hydrate-binding specificity of pea and lentil lectins. Fucose is an important determinant. J. Biol. Chem. 256, 6633–6640. 23. Nakano, M., Yonezawa, N., Hatanaka, Y. & Noguchi, S. (1996) Structure and function of the N-linked carbohydrate chains of pig zona pellucida glycoproteins. J. Reprod. Fertil. Supplement 30, 25–34. 24. Nakano, M., Hatanaka, Y., Kobayashi, N., Noguchi, S., Ishikawa, S. & Tobita, T. (1990) Further fractionation of the glycoprotein families of porcine zona pellucida by anion-exchange HPLC and some characterization of the separated fractions. J. Biochem. (Tokyo) 107, 144–150. 25. Bork, P. & Sander, C. (1992) A large domain common to sperm receptors (ZP2 and ZP3) and TGF-b type III receptor. FEBS Lett. 300, 237–240. 26. Nakano, M. & Yonezawa, N. (2001) Localization of sperm ligand carbohydrate chains in pig zona pellucida glycoproteins. Cells Tissues Organs 168, 65–75. 27. Prasad, S.V., Skinner, S.M., Carino, C., Wang, N., Cartwright, J. & Dunber, B.S. (1999) Structure and function of the proteins of the mammalian zona pellucida. Cells Tissues Organs 166, 148–164. Ó FEBS 2002 Localization of N-linked chains in ZPA (Eur. J. Biochem. 269) 4265 28. Pepin, M C., Beauchemin, M., Plamondon, J. & O’Connor- McCourt, M.D. (1994) Mapping of the ligand binding domain of the transforming growth factor b receptor type III by deletion mutagenesis. Proc. Natl Acad. Sci. USA 91, 6997–7001. 29. Derynck, R. & Feng, X H. (1997) TGF-b receptor signaling. Biochim. Biophys. Acta 1333, F105–F150. 30. van Rooijen, J.J.M., Voskamp, A.F., Kamerling, J.P. & Vlie- genthart, J.F.G. (1999) Glycosylation sites and site-specific gly- cosylation in human Tamm-Horsfall glycoprotein. Glycobiology 9, 21–30. 31. Fukuoka, S. & Freedman, S.D., Yu, H., Sukhatme, V.P. & Scheele, G.A. (1992) GP-2/THP gene family encodes self-binding glycosylphosphatidylinositol-anchored proteins in apical secretory compartments of pancreas and kidney. Proc. Natl Acad. Sci. USA 89, 1189–1193. 32. Chen,D.,Xu,X.,Zhu,L.,Angervo,M.,Li,Q.,Bagchi,M.K.& Bagchi, I.C. (1999) Cloning and uterus/oviduct-specific expression of a novel estrogen-regulated gene (ERG1). J. Biol. Chem. 274, 32215–32224. 33. Kasik, J.W. (1998) A cDNA cloned from pregnant mouse uterus exhibits temporo-spatial expression and predicts a novel protein. Biochem. J. 338, 947–950. 34. Legan, P.K., Rau, A., Keen, J.N. & Richardson, G.P. (1997) The mouse tectorins. Modular matrix proteins of the inner ear homologous to components of the sperm-egg adhesion system. J. Biol. Chem. 272, 8791–8801. 35. Verhoeven, K., Laer, L.V., Kirschhofer, K., Legan, P.K., Hughes, D.C., Schatterman, I., Verstreken, M., Hauwe, P.V., Couche, P., Chen,A.,Smith,R.J.H.,Somers,T.,Officiers,F.E.,Heyning, P.V., Richardson, G.P., Wachtler, F., Kimberling, W.J., Willems, P.J., Govaerts, P.J. & Camp, G.V. (1998) Mutations in the human a-tectorin gene cause autosomal dominant non-syndromic hearing impairment. Nat. Genet. 19,60–62. 4266 K. Ikeda et al.(Eur. J. Biochem. 269) Ó FEBS 2002 . describe the localization of the N-linked chains in bovine ZPA. MATERIALS AND METHODS Preparation of zona protein mixture Zonae pellucidae of bovine eggs were. Essential role of the nonreducing terminal a-mannosyl residues of the N-linked carbohydrate chain of bovine zona pellucida glycoproteins in sperm -egg binding. Mol.