Identification of Sperm-Binding Sites in the N-Terminal Domain of Bovine Egg Coat Glycoprotein ZP4

The species-selective interaction between sperm and egg at the beginning of mammalian fertilisation is partly mediated by a transparent envelope called the zona pellucida (ZP). The ZP is composed of three or four glycoproteins (ZP1–ZP4). The functions of the three proteins present in mice (ZP1–ZP3) have been extensively studied. However, the biological role of ZP4, which was found in all other mammals studied so far, has remained largely unknown. Previously, by developing a solid support assay system, we showed that ZP4 exhibits sperm-binding activity in bovines and the N-terminal domain of bovine ZP4 (bZP4 ZP-N1 domain) is a sperm-binding region. Here, we show that bovine sperm bind to the bZP4 ZP-N1 domain in a species-selective manner and that N-glycosylation is not required for sperm-binding activity. Moreover, we identified three sites involved in sperm binding (site I: from Gln-41 to Pro-46, site II: from Leu-65 to Ser-68 and site III: from Thr-108 to Ile-123) in the bZP4 ZP-N1 domain using chimeric bovine/porcine and bovine/human ZP4 recombinant proteins. These results provide in vitro experimental evidence for the role of the bZP4 ZP-N1 domain in mediating sperm binding to the ZP.


Introduction
The zona pellucida (ZP), a transparent coat surrounding mammalian oocytes, plays important roles in oogenesis, species-selective sperm recognition, blocking polyspermy and early embryonic development [1,2]. Mammalian fertilisation requires several steps. In most species, capacitated sperm bearing an intact acrosome binds to the ZP (primary sperm-ZP binding). After binding, the sperm undergoes the acrosome reaction, which reinforces the binding to the ZP (secondary sperm-ZP binding). The acrosome-reacted sperm then reaches the perivitelline space and binds to the oolemma. Afterwards, gamete membrane fusion occurs [3,4].
All ZP proteins have a ZP module that is important for the formation of ZP filaments. The ZP module consists of two structurally related immunoglobulin (Ig)-like domains (ZP-N and ZP-C) connected to each other by a short, flexible hinge region (see Figure 1A) [12,13]. ZP proteins are highly heterogeneous because asparagine (N-linked glycosylated) and serine/threonine (O-linked glycosylated) residues are glycosylated to different extents.
In many species, the glycans on ZP proteins are involved in the sperm-ZP binding in a species-selective manner [14][15][16].   polypeptide, and a schematic representation of the truncated bZP4  consisting of the ZP-N1 domain (red) and trefoil domain (blue) and its corresponding porcine (p)ZP4(22-187) (ZP-N1 domain, green; trefoil domain, blue) counterpart examined in this study. Inverted tripods mark potential N-glycosylation sites, and cysteine residues are highlighted in yellow bars in bZP4  and pZP4 . (B) Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the bZP4 , bZP4  and N-glycosylation site mutant, bZP4(25-184) N71Q. These proteins were expressed in Sf9 cells and purified by TALON resin specific to His-tag. The SDS-PAGE gels were silver-stained. Molecular mass standards (kDa) are indicated on the left side of each gel. (C) Western blot analyses of bZP4  and bZP4  N71Q. Both proteins were detected by immunoblotting with anti-His antibody (anti-His), while only bZP4  was detected by Galanthus nivalis agglutinin (GNA). The objective bands are indicated by arrows. Molecular mass standards (kDa) are indicated on the left side of each panel. (D) Adsorption of bZP4  and bZP4  N71Q to plastic wells. The amount of protein necessary for adsorption saturation was examined by detecting the adsorbed protein with an antibody specific to the N-terminal His-tag. The amount of bZP4(25-184) (dark gray) or bZP4(25-184) N71Q (light gray) added to one plastic well is indicated under each bar. The experiment was performed three times and the average ± standard deviation (SD) of absorbance at 405 nm is shown. Absorbances of these proteins reached the same level as that of bZP4  at 0.8 µg (blue). (E) Comparison of sperm-binding activity between bZP4(25-184) and bZP4(25-184) N71Q. Plastic wells were coated with the ZP4 proteins (0.8 µg for each protein) indicated in the graph. The number of sperm bound to wells coated with bZP4(25-464) varied from 35 to 86 but was designated as 100%. Assays were repeated five times. Data are presented as the mean ± SD. There was no significant difference between the activities of bZP4(25-184) and bZP4(25-184) N71Q (* p > 0.05).
By developing a solid support assay, we recently demonstrated the sperm-binding activity of two regions of bZP4: the region extending from Lys-25 to Asp-136, which almost corresponds to the N-terminal ZP-N1 domain, and the one extending from Ser-290 to Lys-340, which consists of the flexible hinge region and the N-terminal part of the ZP-C domain (see Figure 1A) [11]. The sperm-binding activity of bZP4(25-136) is a little higher than that of bZP4(290-340). It was hypothesised that the bZP4(290-340) interaction with bZP3 enhances its sperm-binding activity [11]. Although these results document the dependency of sperm recognition on bZP4 , it is still unknown which region of the bZP4  domain is involved in the sperm-binding activity. In the mouse ZP2, the region extending from residues 52 to 83 in the N-terminal domain includes the sperm-binding site, and the N-glycosylation at the single putative site is not necessary for the sperm-binding activity [9,20]. There are four potential N-glycosylation sites in bZP4: at Asn-71, -202, -218, and -314. The Asn-71 site is located in bZP4 . The role of Asn-71 N-glycosylation in the sperm-binding activity of bZP4  remains to be determined.
In this study, to precisely identify the sperm-binding sites of bZP4(25-136), we analysed the involvement of N-glycosylation at the Asn-71 site in the sperm-binding activity and analysed the sperm-binding sites in bZP4(25-136) using a solid support assay.

N-Glycosylation Is Not Required for bZP4(25-184) Sperm-Binding Activity
Recently, we showed that bZP4 has a predominant binding activity for bovine sperm in a solid support assay [11]. The bZP4 polypeptide expressed in Sf9 cells included residues Lys-25 to Arg-464 and consisted of five regions: N-terminal ZP-N-like domain (Lys-25 to Pro-135, ZP-N1), trefoil domain (Asp-136 to Tyr-181), ZP-N domain (Gly-182 to Ala-289, ZP-N), hinge region (Ser-290 to Gln-311) and ZP-C domain (Pro-312 to Arg-464, ZP-C) ( Figure 1A). The sperm-binding activity of bZP4  was shown to be similar to that of bZP4 , indicating that the trefoil domain is dispensable for the spermbinding activity [11]. The purification yield of bZP4(25-184) was a little higher than that of bZP4 . Therefore, bZP4(25-184) was chosen as a starting material in this study.
Previous studies suggested that α-mannose (Man) residues at the nonreducing termini of high-Man-type N-linked chains of bovine ZP are essential for sperm-binding [21][22][23]. Recombinant bZP4 expressed in Sf9 cells has pauci-Man-type N-glycans with α-Man residues at the nonreducing termini [17]. Additionally, there is a potential N-glycosylation site at Asn-71 of bZP4(25-184) ( Figure 1A). Since recombinant bZP4 was not detected by Amaranthus caudatus agglutinin, the population of recombinant bZP4 possessing Olinked chains was low [17]. To investigate the role of N-glycosylation on bZP4  in sperm binding, Asn-71 was mutated to Gln. The bZP4(25-184) N71Q mutant was expressed in Sf9 cells and purified to near homogeneity ( Figure 1B). The mutation of the N-glycosylated Asn-71 residue was confirmed, as Galanthus nivalis agglutinin (GNA) failed to recognise bZP4(25-184) N71Q. In contrast, bZP4  was recognised by GNA, indicating that the Asn-71 residue was N-glycosylated ( Figure 1C). The sperm-binding activity of the recombinant proteins was examined by adsorbing them to plastic wells. We first assessed the amount of bZP4(25-184) and bZP4(25-184) N71Q required to saturate the adsorption of the proteins to the plastic wells and found that 0.8 µg of protein was enough for saturation ( Figure 1D). The number of bovine sperm bound to the wells coated with bZP4(25-184) was 40% lower than that measured in bZP4(25-464)-coated wells, as reported previously ( Figure 1E) [11]. The number of bovine sperm bound to wells coated with bZP4(25-184) N71Q was not significantly different from that obtained with bZP4(25-184) coating ( Figure 1E), indicating that N-glycosylation is not necessary for bZP4  sperm-binding activity.

Bovine Sperm
Recognise bZP4(25-184) but not pZP4  in the Solid Support Assay Next, we determined whether bovine sperm recognise pZP4  as it is similar to bZP4(25-184) ( Figure 1A and Figure S1). Both fragments were expressed in Sf9 cells and purified to near homogeneity, as revealed by SDS-PAGE ( Figures 1B and 2B). The amount of pZP4  required to saturate the adsorption of the protein to the plastic wells was 0.8 µg, which was the same as that of bZP4(25-184) ( Figure 2C) [11]. The number of bovine sperm bound to wells coated with pZP4(22-187) was significantly decreased compared with the number bound to bZP4(25-184)-coated wells ( Figure 2D). This result indicated that bovine sperm recognised bZP4(25-184) but not pZP4(22-187). The amino acid sequences of bZP4 and pZP4 share 75% identity [17]. The bovine and porcine N-terminal ZP-N1 domains share 58% identity, whereas 88% identity was found between the bovine and porcine trefoil domains ( Figure S1). Since the N-glycosylation was not essential for sperm binding and since we previously showed that O-glycosylation was not detected [17], the species-selectivity was most likely mediated by residues that were not identical in bZP4(25-135) and pZP4 .
The translational Met was numbered as 1 for the bZP4, pZP4 and hZP4 fragments. There were gaps in the bovine, porcine and human amino acid sequence alignment ( Figure S1). Therefore, the numbering of corresponding residues did not match amongst the three fragments as shown below.
Since we could not make deletional mutants of bZP4(25-135) because of the disulphide bonds, our strategy followed that used to identify sperm-binding sites in the hZP2 N-terminal domain [9]. Two chimeric fragments were prepared ( Figure 2A): bZP4(25-73)/pZP4(73-187) in which the bovine sequence extending from Asp-74 to Thr-184 of bZP4  was replaced by the porcine sequence extending from Gly-73 to Thr-187, and pZP4(22-67)/bZP4(69-184) in which the porcine sequence extending from Leu-68 to Thr-187 of pZP4  was replaced by the bovine sequence extending from Leu-69 to Thr-184 ( Figure 2A). These chimeric fragments were expressed in Sf9 cells and purified to near homogeneity, as shown by SDS-PAGE ( Figure 2B). The amount of protein required for saturation of the adsorption on plastic wells was 0.8 µg for both proteins ( Figure 2C). The number of bovine sperm bound to wells coated with either bZP4(25-73)/pZP4(73-187) or pZP4(22-67)/bZP4(69-184) was reduced by about 50% compared with that bound to bZP4(25-184)-coated wells ( Figure 2D). Interestingly, there was no significant difference in the sperm-binding activities of bZP4(25-73)/pZP4(73-187) and pZP4(22-67)/bZP4(69-184) ( Figure 2D), suggesting that both regions (residues 25 to 73 and residues 74 to 184) were involved in the sperm-binding activity. The amount of each recombinant ZP4 protein added to one plastic well is indicated under each group of bars. The amounts of proteins necessary for adsorption saturation were examined by detecting the adsorbed proteins with an antibody specific to the N-terminal His-tag. The experiment was performed three times and the average ± standard deviation (SD) of absorbance at 405 nm is shown. (D) Spermbinding activity of the four recombinant ZP4 proteins shown in (C). Plastic wells were coated with each of the proteins (0.8 µg for each protein). The number of sperm bound to the wells coated with bZP4(25-184) varied from 33 to 65 but was designated as 100% for each experiment. Assays were repeated five times. Data are presented as the mean ± SD, with statistical significance between bZP4(25-184) and pZP4(22-187) indicated as p < 0.01 (**) on the line connecting the two bars. There was no significant difference between bZP4(25-73)/pZP4(73-187) and pZP4(22-67)/bZP4(69-184).
We made further chimeric proteins considering that the residues 47 to 49 and 64 are conserved between bovines and pigs ( Figure S1). Using pZP4(22-187) as a backbone, the porcine sequence was replaced by the bovine sequence to express bZP4(41-46), bZP4(65-68) and both bZP4(41-46) and bZP4(65-68) instead of the corresponding porcine sequences ( Figure 3A). The chimeric fragments were purified to near homogeneity ( Figure 3B) and were adsorbed to plastic wells ( Figure 3C). The sperm-binding activities of chimeric proteins containing bZP4(41-46), bZP4(65-68) or both bZP4(41-46) and bZP4(65-68) were comparable and significantly higher than that of pZP4(22-187) ( Figure 3E), whereas they were similar to that of bZP4(25-73)/pZP4(73-187) ( Figures 2D and 3E), suggesting that the residues 41 to 46 (site I) and 65 to 68 (site II) were involved in bZP4(25-184) spermbinding activity. The amount of each chimeric ZP4 protein added to one plastic well is indicated under each group of bars. As a standard, 0.8 µg of bZP4(25-184) was added to a well (red bar). The amounts of proteins necessary for adsorption saturation were examined by detecting the adsorbed proteins with an antibody specific to the N-terminal His-tag. The experiment was performed three times and the average ± standard deviation (SD) of absorbance at 405 nm is shown. (D,E) Sperm-binding activity of each chimeric ZP4 protein shown in (A). Plastic wells were coated with each chimeric ZP4 protein (0.8 µg for each protein) indicated in the graph. The number of sperm bound to the wells coated with bZP4(25-184) varied from 33 to 65 but was designated as 100% for each experiment. Assays were repeated at least four times. Data are presented as the mean ± SD, with statistical significance between two bars indicated as p < 0.05 (*) and p < 0.01 (**) on each line connecting two bars.
The sperm-binding activities of pZP4(22-78)/bZP4(80-81)/pZP4(81-85)/bZP4(87-89)/pZP4(89-187) and pZP4(22-72)/bZP4(74)/pZP4(74-78)/bZP4(80-81)/pZP4(81-85)/ bZP4(87-89)/pZP4(89-187) were much lower than that of pZP4(22-106)/bZP4(108-123)/ pZP4(123-187) and were not significantly different from that of pZP4(22-187) ( Figure 4D). Because six residues at positions 74, 80, 81 and 87 to 89 in the 74 to 93 region of the bovine sequence are not identical to the corresponding residues in the porcine sequence, this result suggested that the residues 74 to 93 were not involved in the sperm-binding activity of the ZP-N1 domain. An unknown technical problem prevented the generation of a plasmid with the Ser-94 to Arg mutation, thus, the involvement of Ser-94 in the sperm-binding activity was not clarified. Taken together, the results strongly suggested that the region extending from residues 108 to 123 (site III) is a sperm-binding site of the ZP-N1 domain of bZP4. The chimeric proteins were expressed in Sf9 cells and purified by TALON resin specific to His-tag. Gels were silver-stained. Molecular mass standards (kDa) are indicated on the left side of each panel. (C) Adsorption of each chimeric ZP4 protein shown in (A) to plastic wells. The amount of each chimeric ZP4 protein added to one plastic well is indicated under each group of bars. As a control, 0.8 µg of bZP4(25-184) was added to a well (red bar). The amounts of proteins necessary for adsorption saturation were examined by detecting the adsorbed proteins with an antibody specific to the N-terminal His-tag. The experiment was performed three times and the average ± standard deviation (SD) of absorbance at 405 nm is shown. (D) Sperm-binding activity of each chimeric ZP4 protein shown in (A). Plastic wells were coated with each chimeric ZP4 protein (0.8 µg for each protein). The number of sperm bound to wells coated with bZP4(25-184) varied from 33 to 65 but was designated as 100% for each experiment. Assays were repeated at least three times. Data are presented as the mean ± SD, with statistical significance between two bars indicated as p < 0.05 (*) and p < 0.01 (**) on each line connecting two bars.

Discussion
The structures and functions of the three proteins (ZP1, ZP2 and ZP3) present in mouse ZP have been studied extensively. ZP2 and ZP3 are the main ZP glycoproteins [7] and are involved in the interaction with sperm [24]. The less abundant ZP1 is not directly involved in sperm binding but covalently crosslinks ZP filaments to maintain the structural integrity of the ZP matrix [25,26]. Although human ZP1, ZP2 and ZP3 biological activities are similar to those of the corresponding mouse ZP glycoproteins [27], the lack of ZP4 in the mouse has hindered the study of the protein and its biological function remains unknown.
To evaluate the role of individual bovine ZP proteins in sperm recognition, we previously developed a solid support assay by coating plastic plates with individual ZP proteins and found that ZP2 and ZP3 are not involved in sperm binding, whereas ZP4 multimerisation is important for the sperm-binding activity [11]. We also identified the N-terminal ZP-N1 domain of ZP4 as responsible for the sperm-binding activity of the protein [11]. The present study aimed to decipher the sperm-ZP binding mechanism in detail. We took advantage of the solid support assay system to analyse sperm-binding sites in the ZP-N1 domain.
We found that bZP4(25-184) sperm-binding activity was independent of Asn-71 Nglycosylation, as mutation of the site did not change sperm-binding activity. The spermbinding activity is species-selective, as bovine sperm failed to recognise both pZP4  and hZP4 . We further utilised this species-selective sperm recognition and by analysing bovine/porcine and bovine/human chimeric proteins, identified three sites, site I: bZP4(41-46), site II: bZP4(65-68) and site III: bZP4(108-123), in the N-terminal ZP-N1 domain of bZP4 , that are important for sperm-binding activity in bovines. No crystal structure of the N-terminal ZP-N1 domain of bZP4 ) is yet available. However, we generated a 3D model of bZP4(25-135) ( Figure 6A).
A single domain in the ZP2 N-terminus is needed for human and mouse sperm recognition [8,9]. In humans, the region extending from residues 56 to 87, namely between Cys-55 and Cys-88, of hZP2 was identified as a sperm-binding site by using chimeric human/mouse recombinant proteins of hZP2(39-154) [9]. In mice, the corresponding residues from 52 to 83 in the N-terminal domain of mZP2 were suggested to be a spermbinding site [9]. Recently, the crystal structure of the mZP2 N-terminal domain, mZP2(35-138) was obtained [27] ( Figure 6B). The 3D model of bZP4  and the crystal structure of mZP2  are similar, and their β-strand cores make two antiparallel β-sheets ( Figure 6A,B). Importantly, the residues 41 to 68, including bZP4(25-135) sites I and II, almost overlap with the 52 to 83 region of mZP2. This suggests that the sperm-binding regions of the ZP-N1 domain of bZP4 and of the mZP2 N-terminal domain overlap. A notable difference is that site III is involved in bZP4(25-135) sperm binding in addition to sites I and II. The site III corresponds to the loop region comprising residues 116 to 130 in the mZP2 sequence which connects the two antiparallel β-strands F and G in mZP2(35-138) ( Figure 6B) [27]. In the crystal structure of mZP2 , residues 52 to 83 and the loop region (residues 116 to 130) are close to each other and seem to form an interaction surface ( Figure 6B) [27]. Since the involvement of the 116 to 130 region of mZP2 or of the corresponding hZP2 region in sperm-binding has not been reported, the role of bZP4(25-135) site III in sperm-binding is, so far, unique to the ZP-N-like domain of bZP.  [28]. Three sperm-binding sites proposed in the present study are highlighted in red (Sites I and II) and in yellow (Site III). (B) Crystal structure of the N-terminal ZP-N1 domain of mZP2 (PDB 5II6) [27]. Mouse sperm-binding region is highlighted in gray. (C) A predicted structural model of the ZP-N1 domain of pZP4. The model was made using AlphaFold2 [28]. (D) Superimposition of the predicted 3D models of the ZP-N1 domains of bZP4 and pZP4.
The difference in the sperm-binding activities between pZP4(22-39)/bZP4(41-61)/ pZP4(61-187) and pZP4(22-39)/bZP4(41-63)/pZP4(63-187) was not significant ( Figure 3D) and the difference in the sperm-binding activities between pZP4(22-48)/bZP4(50-68)/ pZP4(68-187) and pZP4(22-55)/bZP4(57-68)/pZP4(68-187) was not significant ( Figure 3D). These results suggested that the residues 50 to 56, 62 and 63, which are located between site I and site II, were not important for the sperm-binding activity of bZP4 ). The 3D model of bZP4  proposes a hypothesis that the loop region between site I and site II is involved in sperm binding, but the results in this study did not support this hypothesis. We made a 3D model of the N-terminal ZP-N1 domain of pZP4, pZP4(22-136) ( Figure 6C). The 3D models of the ZP-N1 domains of bZP4 and pZP4 are very similar and could be superimposed ( Figure 6D). The probabilities of the 3D model prediction in the loop areas on the left side of the β-strand cores were low for both bZP4 and pZP4. So, it is not yet possible to discuss the molecular mechanisms of the species-selective recognition of the bovine and porcine ZP-N1 domains by bovine sperm based on the 3D models.
Both site I and site II contain Pro at positions 46 and 66, respectively, which are not conserved in the porcine sequence. This raises the possibility that the Pro residues are directly or indirectly involved in bZP4(25-135) sperm-binding activity. However, this remains to be clarified. The identity between site III of bZP4 and the corresponding region of pZP4 is 25%, whereas the identity between the other region of bZP4  and the corresponding region of pZP4 is 63%. This suggests that the unique structural elements of bZP4  site III are important for species-selective sperm recognition.
ZP1 and ZP4 share sequence similarities and present a similar domain organisation [29]. It is believed that they have arisen from a common gene after a gene duplication event that occurred after the fish lineage [30]. Considering that human sperm do not recognise mouse ZP, ZP4 was initially proposed as a sperm ligand in humans. However, this possibility was not supported by the study of transgenic animals expressing mZP1-3 and hZP4, as hZP4 failed to bind human sperm [31]. In addition, sperm binding is normal in ZP4-knock out rabbits [32]. ZP4 is also dispensable for rat fertilisation [33]. However, recent work has suggested that ZP4 functions diverge amongst vertebrates [34]. The ZP4 biological role in species possessing ZP1 might be different from that in bovines which lack ZP1. Although the findings of the present in vitro biochemical study suggest that in bovines ZP4 plays a role in sperm recognition, sperm factors recognising the sperm-binding sites in bZP4  remain to be identified and further in vivo validation of these observations in cow fertilisation is necessary.

Expression and Purification of Recombinant ZP Proteins
Translational Met was numbered as 1 for bZP4, pZP4 and hZP4 fragments. There were gaps of amino acids in the alignment of bovine, porcine and human amino acid sequences. Therefore, the numbering of corresponding residues did not match amongst the three fragments ( Figure S1).

Expression of Recombinant ZP Proteins
The preparation of recombinant viruses was performed according to the previous report [11]. Sf9 cells were routinely propagated in Sf-900II serum-free medium (Invitrogen, Carlsbad, CA, USA). Each of the constructed baculovirus transfer plasmids was transfected along with flashBAC DNA (Oxford Expression Technologies, Oxford, UK) into Sf9 insect cells according to the manufacturer's protocol. Sf9 cells were infected with each recombinant virus, and the expression and secretion of each recombinant protein into the culture supernatant were verified as previously reported [11]. All recombinant ZP proteins were expressed as secretory proteins using a signal peptide derived from pBACgus6.

Purification of Recombinant ZP Proteins from Culture Supernatants
Purification of recombinant ZP proteins was also performed according to the previous report [11]. Briefly, for large-scale protein expression, 200 mL of Sf9 cells (1.0 × 10 6 cells/mL) were infected with each recombinant virus and cultured for 48 h at 27 • C in suspension. Each recombinant protein was purified with TALON metal affinity resin ® (Takara) using the N-terminal His-tag of the proteins.

SDS-PAGE
SDS-PAGE was performed on 12.5% (w/v) separating gels under reducing conditions according to the Laemmli method [36]. The gels were silver-stained. Standard proteins with a broad molecular mass range (Takara) were used to estimate apparent protein molecular masses. A prestained molecular mass marker (SMOBIO Tech, Hsinchu City, Taiwan) was used for Western blots.

Immunoblot Analysis
SDS-PAGE resolved proteins were processed for immunoblot analysis. After transfer of resolved proteins to Immobilon-P membranes (Millipore, Bedford, MA, USA), the membranes were blocked with 3% bovine serum albumin (BSA) in Tris-buffered saline (TBS) for 1 h at room temperature, rinsed once with TBS and incubated with anti-Histag antibody (Wako, Kyoto, Japan) 3000-fold diluted with TBS containing 1% BSA for 2 h. The membranes were then washed three times for 15 min each with TBS containing 0.05% Tween 20 (T-TBS) and then incubated for 1.5 h with horseradish peroxidase (HRP)conjugated rabbit anti-mouse IgG (Wako) 1000-fold diluted with TBS containing 1% BSA. After washing three times for 15 min each with T-TBS, the blots were developed using 3,3 ,5,5 -tetramethyl benzidine (TMB) (SeraCare, Gaithersburg, MD, USA).

Lectin Blot Analysis
Alternatively, SDS-PAGE resolved proteins were transferred to Immobilon-P membranes. The membranes were blocked with TBS containing 3% BSA for 1 h and then incubated for 2 h with 1 µg/mL of biotin-conjugated GNA (EY Laboratories, San Mateo, CA, USA) in T-TBS. Membranes were washed three times for 15 min each with T-TBS and incubated for 1 h with 0.5 µg/mL of HRP-conjugated streptavidin (Sigma-Aldrich) in T-TBS. Membranes were then washed three times for 15 min each with T-TBS, followed by color development using TMB.

Measurement of the Sperm-Binding Activity of Recombinant ZP Proteins by the Solid Support Assay
This procedure was performed according to the previous report [11] as described briefly below.

Adsorption of Recombinant ZP Proteins to Plastic Wells
The amount of each recombinant ZP protein that was enough for saturated adsorption to a well was investigated by adding 50 µL of protein solution at different concentrations to a 96-well plate (Nalge Nunc, Rochester, NY, USA) and incubating the plate overnight at 4 • C. As a control, 0.8 µg of bZP4 was added to a well. The wells were rinsed with phosphate-buffered saline (PBS: 40 mM KH 2 PO 4 , 150 mM NaCl, pH 7.4). The wells were blocked with 3% BSA in TBS for 1 h at room temperature. After washing the wells three times with TBS, an antibody against His-tag (Wako) 3000-fold diluted with TBS containing 1% BSA was added to each well and incubated for 1 h at room temperature. The wells were washed three times with T-TBS and incubated with HRP-conjugated rabbit anti-mouse antibody (Wako) 1000-fold diluted with TBS containing 1% BSA. After washing three times with T-TBS, 2,2 -Azinobis (3-ethylbenzothiazolin-6-sulfonic acid) (ABTS; Roche) was added to each well as a substrate of HRP. After incubating for 1 h at room temperature, absorbance at 405 nm was measured with a plate reader (TECAN, Mannedorf, Swiss).

Sperm Binding to Recombinant ZP Proteins Adsorbed to Plastic Wells
The designated amount of each protein was added to a 96-well plate (Nalge Nunc) and incubated overnight at 4 • C. As a negative control, 50 µL of elution buffer alone, used for TALON column chromatography, was adsorbed. The solution in the wells were discarded, and the wells were washed once with PBS and then blocked with 3% BSA in TBS at 38.5 • C for 2 h. The frozen Holstein bull sperm straws supplied for artificial insemination were purchased from Animal Genetics Japan Co., Ltd. (Matsuzaka, Japan). Frozen bovine sperm was thawed and washed twice in pre-warmed (38.5 • C) Brackett and Oliphant (BO) solution without BSA [11,37]. The bovine sperm were then capacitated by incubation in BO solution containing BSA for 30 min. Capacitation and subsequent incubations were carried out at 38.5 • C under 2% CO 2 . Aliquots (50 µL) containing 4 × 10 5 capacitated sperm were transferred into the wells, and the plates were incubated for 2 h. The wells were washed three times with BO solution and then 50 µL of 70% glycerol in PBS was added to each well, and the sperm bound to the wells were recovered by 20 strokes of vigorous pipetting. The number of sperm in 0.1 µL of suspension was determined using a haemocytometer. The number of sperm bound to the wells not coated with recombinant ZP proteins (0 to 5) was subtracted from the number of sperm bound to the wells coated with recombinant ZP proteins. The average number of sperm in 0.1 µL of suspension in the 100% sperm-binding control of experiments is shown in the legends for figures.

Statistical Analysis
Welch's t-test was used to determine whether there was a significant difference in the sperm count between two groups. Differences were considered to be significant at p < 0.05.