Abstract

A novel α2,9-linked polysialic acid (polySia)-containing glycoprotein of sea urchin sperm flagella was identified and named “flagellasialin.“ Flagellasialin from Hemicentrotus pulcherrimus shows a diverse relative molecular mass on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) of 40–80 kDa. Flagellasialin is a 96-amino acid, threonine-rich, heavily O-glycosylated (80–90% by weight) glycoprotein with a single transmembrane segment at its C-terminus and no apparent cytosolic domain. Of 12 extracellular Thr residues, eight are O-glycosylated and three are nonglycosylated. Flagellasialin is highly expressed in the testis but cannot be detected in the ovary. The amino acid sequences of flagellasialin from three sea urchin species (H. pulcherrimus, Strongylocentrotus purpuratus, and Strongylocentrotus franciscanus) are identical, but some species differences exist in the three core glycan structures to which the sulfated α2,9-linked polyNeu5Ac chain is linked. Finally, the treatment of sperm with a specific antibody against the α2,9-linked polyNeu5Ac structure results in the elevation of intracellular Ca2+ and inhibition of sperm motility and fertilization, implicating flagellasialin as a regulator of these critical processes.

Introduction

The sialic acids (Sia) are a family of 9-carbon carboxylated sugars containing ∼50 members that are derivatives of N‐acetylneuraminic acid (Neu5Ac), N-glycolylneuraminic acid (Neu5Gc), and deaminoneuraminic acid (2-keto-3-deoxy-d-glycero-d-galacto-nononic acid [KDN]) (Schauer et al., 1996; Sato and Kitajima, 1999; Angata and Varki, 2002). Occasionally, Sia are linked to each other to form a polymerized structure, polysialic acid (polySia). PolySia with a degree of polymerization (DP) ranging from 8 to 200 sialyl residues were first demonstrated as capsular components of neuroinvasive bacteria, such as Escherichia coli K1 (Barry and Goebel, 1957; Rohr and Troy, 1980), Neisseria meningitidis group B (NMGB), and N. meningitidis group C (NMGC) (Liu et al., 1971; Bhattacharjee et al., 1975). The α2,8-linked polySia is present in E. coli K1 and NMGB, whereas the α2,9 linkages are present in NMGC. In animals, the α2,8-linked polySia structure has been demonstrated in several glycoproteins of fish eggs (Inoue and Iwasaki, 1978; Kanamori et al., 1990; Sato et al., 1993), eel electroplax (James and Agnew, 1987), mammalian brains, and milk (Finne, 1982; Zuber et al., 1992; Troy, 1996; Yabe et al., 2003). The presence of α2,9-linked polySia has not been known until its recent demonstration in mouse neuroblastoma cells (Inoue et al., 2002) as well as in the sea urchin sperm 40–80 kDa-GP (Miyata et al., 2004). The 40–80 kDa-GP contains a new type of polySia chain—an α2,9-linked polyNeu5Ac chain capped with an HSO3→8Neu5Ac (Neu5Ac8S) residue. In vertebrates, the biological functions of α2,8-linked polySia have been well studied in the neural cell adhesion molecules (NCAMs) (Rutishauser, 1996), and polySia is recognized as a modulator of cell–cell and cell–extracellular matrix interactions by negatively regulating NCAM-mediated interactions due to its bulky, polyanionic nature (Yang et al., 1994; Johnson et al., 2005). However, nothing is known about the biological functions of α2,9-linked polySia in these cells.

In sea urchin egg jelly, there are two major polyanionic molecules, the fucose-sulfate polymer (FSP) and the polySia-glycoprotein, both of which are involved in triggering the sperm acrosome reaction (SeGall and Lennarz, 1979; Vacquier and Moy, 1997; Hirohashi and Vacquier, 2002). The polySia-glycoprotein contains a unique α2,5-Oglycolyl-linked polyNeu5Gc structure (Kitazume et al., 1994). Recently, Sia-rich glycans (SGs) prepared by the alkaline borohydride treatment were shown to increase the intracellular pH of sperm, which potentiates the FSP-induced acrosome reaction (Hirohashi and Vacquier, 2002). In sea urchin sperm, however, no polySia structure has been demonstrated until our finding of the sulfated α2,9-linked polyNeu5Ac structure in the sperm 40–80 kDa-GP (Miyata et al., 2004), although the presence of unique gangliosides containing α2,8-linked di-, tri-, and tetraNeu5Ac structures has been documented in these cells (Ijuin et al., 1996). The 40–80 kDa-GP is exclusively localized in the sperm flagellum (Miyata et al., 2004), leading to the hypothesis that this glycoprotein may be important in sperm motility needed for fertilization.

Results

Purification and biochemical characterization of the 40–80 kDa-GP

Previously, we reported that the major sea urchin sperm sialoglycoprotein had a heterogeneous apparent molecular mass ranging from 40 to 80 kDa (40–80 kDa-GP) and contained a unique polySia, sulfated α2,9-linked polyNeu5Ac (Miyata et al., 2004). To purify the 40–80 kDa-GP, we used several chromatographic procedures as described under Materials and methods. The Triton X-100 extract of sperm was applied to a DEAE-Toyopearl 650 M column and eluted stepwise with 0.2 and 0.4 M NaCl. The 40–80 kDa-GP was enriched in the 0.4 M NaCl fraction (Figure 1A, lane 2, Neu5Ac8S). The 0.4 M NaCl fraction was then subjected to the same column and eluted with a linear gradient of NaCl. The 40–80 kDa-GP eluted between 0.25 and 0.3 M NaCl (Figure 1B). As shown in Figure 1B (Sia), the 0.25–0.3 M NaCl fraction contained 80% of the total Sia content. This result suggests that the 40–80 kDa-GP is the major Sia-containing glycoprotein in sperm. The pooled fractions (overline, Figure 1B, Sia) were then applied to a wheat germ agglutinin (WGA) column. The 40–80 kDa-GP was recovered in the flow-through fraction (Figure 1C, lane 1) and could be separated from other mAb.3G9-positive glycoproteins (Figure 1C, lane 3). The 40–80 kDa-GP cannot be detected by silver staining (Figure 1C, lane 1, Silver) or Coomassie brilliant blue staining (data not shown). However, the 40–80 kDa-GP is readily detected on western blots using mAb.3G9 (Figure 1C, lane 1, Neu5Ac8S).

Fig. 1.

Purification of the 40–80 kDa-GP from Hemicentrotus pulcherrimus sperm. (A) Western blotting of the 0.2 M (lane 1) and 0.4 M NaCl (lane 2) fractions from DEAE-Toyopearl chromatography of the Triton X-100 extract of sperm. Each fraction was subjected to SDS–PAGE followed by the silver staining (Silver) and western blotting with mAb.3G9 (Neu5Ac8S). (B) DEAE-Toyopearl chromatography of the 0.4 M NaCl fraction obtained in (A). The 0.4 M NaCl fraction was applied to the same DEAE-Toyopearl column as in (A) and eluted with a linear gradient of 0.2–0.5 M NaCl in 1% Triton X-100, and 10 mM Tris–HCl (pH 8.0). The elution profiles were monitored by SDS–PAGE followed by silver staining (Silver), western blotting with mAb.3G9 (Neu5Ac8S), and the resorcinol method for the amount of Sia (A580; Sia). The 40–80 kDa-GP-containing fractions as indicated by the bar were pooled. The NaCl concentration is shown by the straight line. (C) WGA-affinity chromatography of the 40–80 kDa-GP-containing fraction in (B). The pooled fraction was applied to the WGA column (2 mL) and washed with 0.5 M NaCl, 0.1% Triton X‐100, and 10 mM Tris–HCl (pH 8.0) and eluted with washing buffer containing 100 mM GlcNAc. The flow-through fraction (lane 1), the wash fraction (lane 2), and the GlcNAc-eluted fraction (lane 3) were subjected to SDS–PAGE followed by silver staining (Silver) and western blotting with mAb.3G9 (Neu5Ac8S).

Fig. 1.

Purification of the 40–80 kDa-GP from Hemicentrotus pulcherrimus sperm. (A) Western blotting of the 0.2 M (lane 1) and 0.4 M NaCl (lane 2) fractions from DEAE-Toyopearl chromatography of the Triton X-100 extract of sperm. Each fraction was subjected to SDS–PAGE followed by the silver staining (Silver) and western blotting with mAb.3G9 (Neu5Ac8S). (B) DEAE-Toyopearl chromatography of the 0.4 M NaCl fraction obtained in (A). The 0.4 M NaCl fraction was applied to the same DEAE-Toyopearl column as in (A) and eluted with a linear gradient of 0.2–0.5 M NaCl in 1% Triton X-100, and 10 mM Tris–HCl (pH 8.0). The elution profiles were monitored by SDS–PAGE followed by silver staining (Silver), western blotting with mAb.3G9 (Neu5Ac8S), and the resorcinol method for the amount of Sia (A580; Sia). The 40–80 kDa-GP-containing fractions as indicated by the bar were pooled. The NaCl concentration is shown by the straight line. (C) WGA-affinity chromatography of the 40–80 kDa-GP-containing fraction in (B). The pooled fraction was applied to the WGA column (2 mL) and washed with 0.5 M NaCl, 0.1% Triton X‐100, and 10 mM Tris–HCl (pH 8.0) and eluted with washing buffer containing 100 mM GlcNAc. The flow-through fraction (lane 1), the wash fraction (lane 2), and the GlcNAc-eluted fraction (lane 3) were subjected to SDS–PAGE followed by silver staining (Silver) and western blotting with mAb.3G9 (Neu5Ac8S).

The presence of α2,9-linked polyNeu5Ac on O-linked glycans in 40–80 kDa-GP

To examine whether the purified 40–80 kDa-GP contains sulfated α2,9-linked polyNeu5Ac chains, we subjected the purified glycoprotein to mild acid hydrolysis/fluorometric anion-exchange high-performance liquid chromatography (HPLC) analysis. As shown in Figure 2A, sulfated and unsulfated oligo/polyNeu5Ac were detected under peaks S1–S5 and 2–15. This HPLC profile was identical to that of the sulfated α2,9-linked polyNeu5Ac-containing glycopeptide (SGP) prepared by extensive Actinase digestion of the delipidated lysate of sperm (Miyata et al., 2004; see also Figure 6A). As previously reported, α2,9- and α2,8-linked oligo/polyNeu5Ac could be separated by HPLC. The oligo/polyNeu5Ac peaks 2–15 in Figure 2A were co-eluted with authentic α2,9-linked, but not with α2,8-linked oligo/polyNeu5Ac (data not shown), indicating that the intrachain linkage of the polySia is α2,9. The confirmative results were also obtained by demonstrating that SGP (see Determination of O-glycosylation sites), which was previously shown to contain exclusively α2,9-linked polyNeu5Ac using the mild acid hydrolysis/fluoromeric HPLC as well as the fluorometric C7/C9 analyses (Miyata et al., 2004), was derived from the 40–80 kDa-GP.

Fig. 2.

Chemical and immunochemical detection of the sulfated α2,9‐linked polyNeu5Ac structure in the 40–80 kDa-GP by mild acid hydrolysis/fluorometric anion-exchange HPLC and western blot analysis. (A) Fluorometric anion-exchange HPLC of DMB derivatives of the mild acid hydrolysate of purified 40–80 kDa-GP. The mild acid hydrolysate of the purified 40–80 kDa-GP was derivatized with DMB and applied to a Mono Q anion-exchange HPLC as described under Materials and methods. The peaks designated as 2, 3, 4, and so on are corresponding to the α2,9-linked diNeu5Ac, triNeu5Ac, tetraNeu5Ac, and so on. The peaks designated as S2, S3, S4, and S5 are corresponding to Neu5Ac8S linked to monoNeu5Ac, diNeu5Ac, triNeu5Ac, and tetraNeu5Ac. The peak S1 corresponds to Neu5Ac8S. The NaCl concentration is shown by the straight line. (B) Western blotting of the Triton-extracted sperm lysate and the purified 40–80 kDa-GP that were untreated and treated with PNGase F or 0.1 M NaOH. The Triton X-100 extract of whole sperm (lanes 1, 3, and 5) and the purified 40–80 kDa-GP (lanes 2, 4, and 6) were analyzed. For PNGase F treatment, the samples were treated with PNGase F at 37°C for 24 h and subjected to SDS–PAGE/Western blotting (lanes 5 and 6). For alkaline treatment, the samples were subjected to SDS–PAGE and electroblotted onto polyvinylidene difluoride (PVDF) membranes. The membranes were untreated (lanes 1 and 2) and treated with 0.1 N NaOH at 37°C for 18 h (lanes 3 and 4). Under these conditions, the β-elimination of O-glycans occurred. Immunodetection was performed with mAb.3G9 (Neu5Ac8S) or anti-α2,9 Neu5Ac (α2,9 Neu5Ac).

Fig. 2.

Chemical and immunochemical detection of the sulfated α2,9‐linked polyNeu5Ac structure in the 40–80 kDa-GP by mild acid hydrolysis/fluorometric anion-exchange HPLC and western blot analysis. (A) Fluorometric anion-exchange HPLC of DMB derivatives of the mild acid hydrolysate of purified 40–80 kDa-GP. The mild acid hydrolysate of the purified 40–80 kDa-GP was derivatized with DMB and applied to a Mono Q anion-exchange HPLC as described under Materials and methods. The peaks designated as 2, 3, 4, and so on are corresponding to the α2,9-linked diNeu5Ac, triNeu5Ac, tetraNeu5Ac, and so on. The peaks designated as S2, S3, S4, and S5 are corresponding to Neu5Ac8S linked to monoNeu5Ac, diNeu5Ac, triNeu5Ac, and tetraNeu5Ac. The peak S1 corresponds to Neu5Ac8S. The NaCl concentration is shown by the straight line. (B) Western blotting of the Triton-extracted sperm lysate and the purified 40–80 kDa-GP that were untreated and treated with PNGase F or 0.1 M NaOH. The Triton X-100 extract of whole sperm (lanes 1, 3, and 5) and the purified 40–80 kDa-GP (lanes 2, 4, and 6) were analyzed. For PNGase F treatment, the samples were treated with PNGase F at 37°C for 24 h and subjected to SDS–PAGE/Western blotting (lanes 5 and 6). For alkaline treatment, the samples were subjected to SDS–PAGE and electroblotted onto polyvinylidene difluoride (PVDF) membranes. The membranes were untreated (lanes 1 and 2) and treated with 0.1 N NaOH at 37°C for 18 h (lanes 3 and 4). Under these conditions, the β-elimination of O-glycans occurred. Immunodetection was performed with mAb.3G9 (Neu5Ac8S) or anti-α2,9 Neu5Ac (α2,9 Neu5Ac).

Fig. 6.

Comparative structural analyses of polySia chains obtained from sea urchin sperm of the three species. A major Sia-containing glycopeptide fraction (SGP) was prepared from sperm of Hemicentrotus pulcherrimus (H.p), Strongylocentrotus purpuratus (S.p), and Strongylocentrotus franciscanus (S.f) as described under Materials and methods. (A) Mono Q-HPLC of DMB derivatives of mild acid hydrolysate of SGP. SGP was mild acid hydrolyzed to release a series of oligo/polyNeu5Ac. These oligo/polyNeu5Ac were DMB labeled and analyzed on an anion-exchange HPLC. Elution was performed with a linear gradient of NaCl (straight line) and monitored with fluorescence (excitation 373 nm, emission 448 nm). See also the legend of Figure 2A. (B) Mono Q-HPLC of sialoglycan alditols obtained from SGP by alkaline borohydride treatment. The sialooligosaccharide alditols obtained by the treatment with 0.1 N NaOH and 1 M NaBH4 were subjected to anion-exchange HPLC as described above. An elution profile was monitored by measuring the absorbance at 210 nm (A210). The arrowheads with numbers show the elution positions of α2,8-linked oligo/polyNeu5Ac with a series of DP. The NaCl concentration is shown by the straight line.

Fig. 6.

Comparative structural analyses of polySia chains obtained from sea urchin sperm of the three species. A major Sia-containing glycopeptide fraction (SGP) was prepared from sperm of Hemicentrotus pulcherrimus (H.p), Strongylocentrotus purpuratus (S.p), and Strongylocentrotus franciscanus (S.f) as described under Materials and methods. (A) Mono Q-HPLC of DMB derivatives of mild acid hydrolysate of SGP. SGP was mild acid hydrolyzed to release a series of oligo/polyNeu5Ac. These oligo/polyNeu5Ac were DMB labeled and analyzed on an anion-exchange HPLC. Elution was performed with a linear gradient of NaCl (straight line) and monitored with fluorescence (excitation 373 nm, emission 448 nm). See also the legend of Figure 2A. (B) Mono Q-HPLC of sialoglycan alditols obtained from SGP by alkaline borohydride treatment. The sialooligosaccharide alditols obtained by the treatment with 0.1 N NaOH and 1 M NaBH4 were subjected to anion-exchange HPLC as described above. An elution profile was monitored by measuring the absorbance at 210 nm (A210). The arrowheads with numbers show the elution positions of α2,8-linked oligo/polyNeu5Ac with a series of DP. The NaCl concentration is shown by the straight line.

Polyclonal and monoclonal antibodies were raised against NMGC capsular polysaccharide consisting of the α2,9-linked polyNeu5Ac structure and designated anti-α2,9 Neu5Ac and 4F7, respectively. Both anti-α2,9 Neu5Ac and mAb.4F7 were reactive with the α2,9-linked polyNeu5Ac chains from NMGC capsular polysaccharide but not with the α2,8-linked polyNeu5Ac chains from E. coli K1 capsular polysaccharide (colominic acid) (data not shown). To immunochemically demonstrate the presence of α2,9-linked oligo/polyNeu5Ac chains in the sperm 40–80 kDa-GP, we performed western blotting of whole sperm lysate and the purified 40–80 kDa-GP using anti-α2,9 Neu5Ac. As shown in Figure 2B, the anti-α2,9 Neu5Ac reacted exclusively with the 40–80 kDa-GP (lane 1, α2,9 Neu5Ac). The same results were obtained with mAb.4F7 (data not shown). These results indicate that the 40–80 kDa-GP contains α2,9-linked polyNeu5Ac. The mAb.3G9 immunostained not only the 40–80 kDa-GP but also several other glycoproteins (Figure 2B, lane 1, Neu5Ac8S), indicating that Neu5Ac8S is a common epitope among several glycoproteins in sea urchin sperm, whereas the α2,9-linked polyNeu5Ac structure exists specifically on the 40–80 kDa-GP.

To examine whether sulfated α2,9-linked polyNeu5Ac chains are present on the N- or O-glycans, we performed PNGase F digestion and alkaline treatment of the whole sperm lysate and the purified 40–80 kDa-GP. The immunostaining of the 40–80 kDa-GP with both mAb.3G9 and anti-α2,9 Neu5Ac was lost on alkaline treatment (Figure 2B, lanes 3 and 4) but not on PNGase F digestion (Figure 2B, lanes 5 and 6). These results indicate that sulfated α2,9‐linked polyNeu5Ac chains are present on the O-glycans of the 40–80 kDa-GP, consistent with our previous findings (Miyata et al., 2004). From these results, we confirm that the 40–80 kDa-GP contains sulfated α2,9-linked polyNeu5Ac structures on its O-glycans.

cDNA and deduced amino acid sequences for the 40–80 kDa-GP

The 40–80 kDa-GP was resistant to several proteases that are used for conventional peptide mapping, such as trypsin, V8 protease, and endopeptidase Lys-C. Therefore, the 40–80 kDa-GP was hydrolyzed with diluted formic acid for the peptide mapping. Under the hydrolysis conditions, peptide bonds are known to be preferentially cleaved at aspartic acid residues (Inglis, 1983). The acid hydrolysate of the purified 40–80 kDa-GP was applied to an ODS column, and the obtained peptides were sequenced (data not shown). The sequences of the resulting peptides 1–4 are summarized in Table I. By reverse transcription polymerase chain reaction (RT–PCR) using degenerate primers based on the sequence of peptide 4, and by the subsequent 3′- and 5′-rapid amplification of cDNA ends (RACE) reactions, the full-length cDNA was cloned (accession number DQ381467). The initial codon at nucleotide 1 is assigned at the start of translation, because there are two upstream in-frame stop codons. A polyadenylation signal (AATAAA) is located at nucleotides 799–804, 22 nucleotides upstream from poly A tail (data not shown). Thus, the cDNA contains an open reading frame of 288 bp, encoding 96 amino acids. The deduced amino acid sequence includes the four peptide sequences obtained from the purified 40–80 kDa-GP (Figure 3 and Table I). The deduced amino acid sequence has a putative 26-amino acid signal peptide at the NH2-terminus, a 47-amino acid extracellular region, and a 23-amino acid transmembrane segment at the COOH-terminus with no cytosolic C-terminus. The cleavage of the putative signal sequence at Ala26 would produce a 70-amino acid mature polypeptide chain. Although its mobility on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) suggests a heterogeneous molecular mass of 40–80 kDa, the calculated molecular mass of the putative mature protein is only 7.2 kDa. This suggests that about 80–90% of the glycoprotein consist of glycan chains. This glycoprotein is also characterized by the high content of Thr residues (∼20%). Edman degradation analysis of the peptides obtained from the purified 40–80 kDa-GP suggests that at least 7 of 12 Thr residues in the extracellular region are modified, most likely by O-linked glycans (Table I). Because it has no Ser residues in the extracellular region, all O-linked glycans may be exclusively linked to Thr residues. There are 15 negatively charged residues and 0 positively charged residue in the protein part of this molecule. There are no consensus sites for protein kinase-A phosphorylation. There is one potential N-glycosylation site in the extracellular region; however, the following observation suggests that the protein has no N-linked glycan chain. First, the amino acid residue at the potential N-glycosylation site was determined by Edman degradation to be Asn64 (Table I, peptide 3). Second, the mobility of purified 40–80 kDa-GP on SDS–PAGE was unaltered after PNGase F treatment (Figure 2). No homologous protein was found in various protein and gene databases; thus, we named this novel glycoprotein “flagellasialin.“

Table I.

Amino acid sequences of peptide fragments obtained from the 40–80 kDa-GP by the formic acid treatment and their location (amino acid number) in the deduced amino acid sequence of flagellasialin

Peptide Sequence Locationa 
MXXNVAAEEEGXD 42–54 
TPXVXVTVD 31–39 
VPXVVNETEVXEEP 59–72 
TPXVXVTVDEDMXXNVAAEEEGXD 31–54 
Peptide Sequence Locationa 
MXXNVAAEEEGXD 42–54 
TPXVXVTVD 31–39 
VPXVVNETEVXEEP 59–72 
TPXVXVTVDEDMXXNVAAEEEGXD 31–54 

X, not detected.

a

See Figure 3.

Fig. 3.

The deduced amino acid sequence of the 40–80 kDa-GP or flagellasialin. The deduced amino acid sequence is numbered from the initiation Met. The putative signal sequence is underlined. The putative transmembrane domain is marked by a dashed underline. One potential N-glycosylation site is italicized at N64. Two amino acid sequences that were determined by the Edman degradation of peptide fragments obtained from purified 40–80 kDa-GP are double underlined (see Table I).

Fig. 3.

The deduced amino acid sequence of the 40–80 kDa-GP or flagellasialin. The deduced amino acid sequence is numbered from the initiation Met. The putative signal sequence is underlined. The putative transmembrane domain is marked by a dashed underline. One potential N-glycosylation site is italicized at N64. Two amino acid sequences that were determined by the Edman degradation of peptide fragments obtained from purified 40–80 kDa-GP are double underlined (see Table I).

To learn about the expression of flagellasialin, we isolated total RNA from the testis and ovary of Hemicentrotus pulcherrimus. Northern blots using digoxigenin (DIG)-labeled full-length flagellasialin cDNA demonstrated that the expression of a 1.0-kb flagellasialin mRNA was detected in the testis but not in the ovary (Figure 4).

Fig. 4.

Northern blot analysis of the flagellasialin mRNA. Total RNA fractions isolated from the testis (lane 1) and ovary (lane 2) of Hemicentrotus pulcherrimus were analyzed by northern blotting using full-length flagellasialin cDNA as a probe (left) or visualized with ethidium bromide (right) to show equal loading. The positions of 18S and 28S rRNA are indicated on the right.

Fig. 4.

Northern blot analysis of the flagellasialin mRNA. Total RNA fractions isolated from the testis (lane 1) and ovary (lane 2) of Hemicentrotus pulcherrimus were analyzed by northern blotting using full-length flagellasialin cDNA as a probe (left) or visualized with ethidium bromide (right) to show equal loading. The positions of 18S and 28S rRNA are indicated on the right.

Determination of O-glycosylation sites

At least 7 of 12 Thr residues in the extracellular region of flagellasialin possess O-linked glycans (Table I). To determine the substitution on the glycoprotein, we performed matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) and Edman degradation of sialoglycopeptides (SGPs). SGP was prepared from H. pulcherrimus sperm homogenates by Actinase digestion and subsequent column chromatography and was the only glycopeptide fraction that contained sulfated α2,9-linked polyNeu5Ac structures (Miyata et al., 2004). Only two monosaccharide components GalNAc and Sia were detected in the SGP, and asialo-SGP, which were prepared by a reverse-phase HPLC of mild acid hydrolysate of the SGP. Three major amino acid sequences from asialo-SGP are summarized in Table II and are assignable to amino acid residues 31–39, 40–50, and 51–63 (Figure 3). MALDI-TOF-MS analysis showed that each fragment contained two or three GalNAc residues. The sites of GalNAc substitution could be determined as summarized in Table II, because substituted Thr residues were easily distinguished from unsubstituted ones by Edman degradation. Thus, seven Thr residues (Thr33, 35, 43, 44, 53, 57, 61) are O-glycosylated, and two Thr residues (Thr31, 37) are not substituted. In addition, the amino acid sequence of peptide 3 (Table I) indicates that Thr66 and Thr69 are unglycosylated and O-glycosylated, respectively, although we could not detect peptide 3 in the asialo-SGP fraction. No information was obtained about Thr28. Taken together, of 12 threonine residues in the predicted extracellular region of flagellasialin, eight residues (Thr33, 35, 43, 44, 53, 57, 61, 69) were O-glycosylated and three residues (Thr31, 37, 66) were not glycosylated. All GalNAc residues in SGP are sialylated (Miyata et al., 2004), and therefore, all seven O-glycosylated Thr residues in SGP appear to contain sulfated α 2,9-linked poly Neu5Ac chains.

Table II.

Identification of O-glycosylation sites of flagellasialin by MALDI-TOF-MS and Edman degradation of asialo-glycopeptides derived from flagellasialin

Observed mass (m/z) [M + Na]+ Number of GalNAc Assignmenta 
1362.6 (31) TPTVTVTVD (39) 
1639.6 (40) EDMTTNVAAEE (50) 
1991.8 (51) EGTDEPTDVPTVV (63) 
Observed mass (m/z) [M + Na]+ Number of GalNAc Assignmenta 
1362.6 (31) TPTVTVTVD (39) 
1639.6 (40) EDMTTNVAAEE (50) 
1991.8 (51) EGTDEPTDVPTVV (63) 

Numbers in parentheses are amino acid numbers in the deduced flagellasialin amino acid sequence.

a

O-GalNAc-glycosylated Thr residues are indicated in boldface.

Comparison of flagellasialin from three species of sea urchin sperm

To investigate the evolutionary conservation of flagellasialin, we cloned its cDNA orthologs from two other species, Strongylocentrotus purpuratus (accession number DQ381468) and Strongylocentrotus franciscanus (accession number DQ381469) using primers specific to H. pulcherrimus. The cDNA and amino acid sequences for three species showed very high similarity to each other. The nucleotide sequence identities of the flagellasialin cDNA for S. purpuratus and S. franciscanus to that of H. pulcherrimus were 99.7 and 100%, respectively. The deduced amino acid sequences of flagellasialin from these three sea urchin species were 100% identical, indicating that flagellasialin is highly conserved in sea urchin sperm.

To learn about the similarity of glycan structures of flagellasialin among these three species, we performed western blotting of whole sperm lysates with mAb.3G9 and anti-α2,9 Neu5Ac (Figure 5). For the α2,9-linked polyNeu5Ac epitope, smears at molecular masses of 40–80, 35–70, and 30–60 kDa were detected for H. pulcherrimus, S. purpuratus, and S. franciscanus, respectively (Figure 5, α2,9 Neu5Ac). For the Neu5Ac8S epitope, several components were detected other than smears that correspond to those observed with anti-α2,9 Neu5Ac (Figure 5, Neu5Ac8S). These results suggest that the occurrence of the sulfated α2,9-linked polyNeu5Ac structure is a common feature of flagellasialin. It should be noted that the molecular mass of flagellasialin was different among these three sea urchin species. Because the core proteins are identical, the glycan structures of flagellasialin must be different among these three species. To examine how different these glycan structures are, we analyzed SGP prepared from the sperm of these three spices by three methods. First, mild acid hydrolysates of SGP were subjected to fluorometric anion-exchange HPLC. Similar HPLC profiles were obtained for the three species, and sulfated α2,9-oligoNeu5Ac (peaks S1–S4) and unsulfated ones (peaks 2–12 or 13) were detected in all species (Figure 6A). Second, SGPs from the three species were treated with alkaline borohydride, and the released sialoglycan alditols were separated on an anion-exchange HPLC. Because sulfated α2,9-polyNeu5Ac chains are not hydrolyzed by the alkaline borohydride, the precise DP of polySia chains can be determined. The anion-exchange HPLC profiles of the sialoglycan alditols from three species are shown in Figure 6B. All three SGPs gave a series of peaks with the DP ranging from 1 to 20. However, the distributions of chain lengths of polySia were different among the three species. Nearly identical HPLC profiles were obtained for H. pulcherrimus and S. purpuratus. In contrast, the population of short sialyl chains (DP<5, oligoSia) in the SGP from S. franciscanus was larger than those of the other two species. The higher population of oligoNeu5Ac with short DPs for S. franciscanus may explain the lower molecular masses (30–60 kDa) of flagellasialin of this species. Third, carbohydrate compositions of SGP from three species were analyzed by gas–liquid chromatography (GLC) (Table III). We previously showed (Miyata et al., 2004) that the reducing terminal structure of H. pulcherrimus flagellasialin was →6GalNAcα1→Thr and that the sulfated α2,9-linked polyNeu5Ac chains were attached to the 6-position of a GalNAc residue. From Table III, the molar proportion of Neu5Ac+Neu5Ac8S to N-acetylgalactosaminitol (GalNAc-ol) in S. franciscanus (16.8:1.0) was smaller than those in H. pulcherrimus and S. purpuratus (18.7:1.0 and 21.1:1.0), consistent with the lower molecular masses of S. franciscanus flagellasialin. It is also noted that Gal and GlcNAc other than GalNAc, Neu5Ac, and Neu5Ac8S were detected in S. purpuratus and S. franciscanus SGPs, whereas no Gal or GlcNAc was detected in H. pulcherrimus SGP (Table III). Taken together, differences in the O-glycan core structures as well as in the DP distribution of the polyNeu5Ac chains might be involved in the species-specific difference in the molecular masses of flagellasialin among these three species.

Fig. 5.

Immunodetection of flagellasialin from sperm from three sea urchin species. The Triton X-100 extracts of Hemicentrotus pulcherrimus (lane 1), Strongylocentrotus purpuratus (lane 2), and Strongylocentrotus franciscanus (lane 3) sperm were subjected to SDS–PAGE (1 µg protein/lane) and immunoblotted with mAb.3G9 (Neu5Ac8S) or anti-α2,9Neu5Ac (α2,9 Neu5Ac).

Fig. 5.

Immunodetection of flagellasialin from sperm from three sea urchin species. The Triton X-100 extracts of Hemicentrotus pulcherrimus (lane 1), Strongylocentrotus purpuratus (lane 2), and Strongylocentrotus franciscanus (lane 3) sperm were subjected to SDS–PAGE (1 µg protein/lane) and immunoblotted with mAb.3G9 (Neu5Ac8S) or anti-α2,9Neu5Ac (α2,9 Neu5Ac).

Table III.

Carbohydrate composition of SGP from three species of sea urchin sperm

 Hemicentrotus pulcherrimus Strongylocentrotus purpuratus Strongylocentrotus franciscanus 
GalNAc 1.0a 1.0a 1.0a 
Gal 0.0 0.1 0.2 
GlcNAc ND 0.2 0.5 
Neu5Acb 16 18 14 
Neu5Ac8Sc 2.7 3.1 2.8 
 Hemicentrotus pulcherrimus Strongylocentrotus purpuratus Strongylocentrotus franciscanus 
GalNAc 1.0a 1.0a 1.0a 
Gal 0.0 0.1 0.2 
GlcNAc ND 0.2 0.5 
Neu5Acb 16 18 14 
Neu5Ac8Sc 2.7 3.1 2.8 

ND, not detected.

a

Values are molar ratios relative to GalNAc taken as 1.0.

b

Values of unsulfated plus 8-O-sulfated Neu5Ac.

c

8-O-sulfated Neu5Ac.

Effects of anti-α2,9 Neu5Ac on fertilization, sperm motility, and intracellular Ca2+

To learn of the potential roles of flagellasialin in H. pulcherrimus sperm physiology, we investigated the effects of pre-incubation with anti-α2,9-linked polyNeu5Ac antibodies (α2,9 Neu5Ac) (Figure 7A). Anti-α2,9 Neu5Ac inhibited the fertilization of eggs in a dose-dependent manner, whereas pre-immune sera had no effect on fertilization. Because we previously showed that flagellasialin was localized in sperm flagellum (Miyata et al., 2004), we suspected its involvement in sperm motility. As shown in Figure 7B, sperm swam in circular paths in HASW without antibodies. When sperm was treated with anti-α2,9 Neu5Ac, swimming in circular paths was disrupted, whereas very little or no effect of the treatment with pre-immune sera was observed (Figure 7B). The lengths of swimming trajectories per 0.2 s were measured with and without antibody treatments. As shown in Figure 7C, trajectory lengths were significantly decreased by anti-α2,9 Neu5Ac treatment, whereas they were not altered by the pre-immune sera.

Fig. 7.

Significance of flagellasialin containing the α2,9-linked polyNeu5Ac in sperm fertility and motility. (A) Effects of anti-α2,9-linked polySia on sperm fertility. Hemicentrotus pulcherrimus sperm was pre-incubated with pre-immune sera (pre) or anti-α2,9-linked polyNeu5Ac (α2,9), followed by the fertilization of fresh eggs. Four independent experiments were performed on different days, and the figure shows representative data. (B) Sperm motility. Sperm was incubated with HASW (HASW), pre-immune sera (pre), or anti-α2,9-linked polyNeu5Ac antibodies (α2,9) for 1 min. Each panel shows the time-lapse photographs, showing the trajectory of swimming sperm heads during 0.2 s. (C) Sperm motility. The swimming length of sperm during 0.2 s was measured and a value expressed as relative length to that in HASW set to 100%. Bars represent ± SD. The number of observed sperm is indicated in parentheses. Statistics: ns, p > 0.05; *, P<0.001. (D) Sperm motility. Sperm were photographed (1 s exposures) using dark field conditions 1 min and 3 min after mixing with 0.1–1.0 µg/mL of mAb.4F7 (IgG), mAb.12E3 (IgM), mAb.3G9 (IgM), and mAb.kdn3G (IgG).

Fig. 7.

Significance of flagellasialin containing the α2,9-linked polyNeu5Ac in sperm fertility and motility. (A) Effects of anti-α2,9-linked polySia on sperm fertility. Hemicentrotus pulcherrimus sperm was pre-incubated with pre-immune sera (pre) or anti-α2,9-linked polyNeu5Ac (α2,9), followed by the fertilization of fresh eggs. Four independent experiments were performed on different days, and the figure shows representative data. (B) Sperm motility. Sperm was incubated with HASW (HASW), pre-immune sera (pre), or anti-α2,9-linked polyNeu5Ac antibodies (α2,9) for 1 min. Each panel shows the time-lapse photographs, showing the trajectory of swimming sperm heads during 0.2 s. (C) Sperm motility. The swimming length of sperm during 0.2 s was measured and a value expressed as relative length to that in HASW set to 100%. Bars represent ± SD. The number of observed sperm is indicated in parentheses. Statistics: ns, p > 0.05; *, P<0.001. (D) Sperm motility. Sperm were photographed (1 s exposures) using dark field conditions 1 min and 3 min after mixing with 0.1–1.0 µg/mL of mAb.4F7 (IgG), mAb.12E3 (IgM), mAb.3G9 (IgM), and mAb.kdn3G (IgG).

The effects of the addition of monoclonal antibody against α2,9-linked polyNeu5Ac (4F7) on sperm motility were investigated (Figure 7D). Sperm motility was also inhibited by mAb.4F7 (IgG) in a dose-dependent and time-dependent manner. At 1.0 µg/mL mAb.4F7, sperm stopped swimming within 1 min following the antibody addition. No effects were observed when mAb.12E3 (IgM) recognizing α2,8-linked polyNeu5Ac with DP of ≥5, mAb.3G9 (IgM) recognizing Neu5Ac8S, and mAb.kdn3G (IgG) recognizing KDNα2,3Gal were used.

Sperm motility is altered by changes in intracellular Ca2+ (Darszon et al., 2005). We thus investigated the effects of mAb.4F7 on intracellular Ca2+. As shown in Figure 8, mAb.4F7 caused a dose-dependent increase in intracellular Ca2+. Fab fragment of mAb.4F7 did not increase intracellular Ca2+ (data not shown). This effect was dependent on the external Ca2+, as no Ca2+ increase occurred in Ca2+-free seawater (data not shown). The pre-incubation of mAb.4F7 with SGP blocked the increase in intracellular Ca2+, indicating that the effect is specific to α2,9-linked polyNeu5Ac structure on flagellasialin.

Fig. 8.

Effects of mAb.4F7 on intracellular Ca2+. Fura-2-loaded sperm were used for intracellular Ca2+ measurements. mAb.4F7 (0–10 µg) was pre-incubated with or without SGP (13 µg as Sia). After 10 min, various concentrations of mAb.4F7 were added (at the arrow). The Ca2+ signal is presented as the F340/F380 ratio.

Fig. 8.

Effects of mAb.4F7 on intracellular Ca2+. Fura-2-loaded sperm were used for intracellular Ca2+ measurements. mAb.4F7 (0–10 µg) was pre-incubated with or without SGP (13 µg as Sia). After 10 min, various concentrations of mAb.4F7 were added (at the arrow). The Ca2+ signal is presented as the F340/F380 ratio.

Discussion

In a previous report, we demonstrated the presence of novel sulfated α2,9-linked polyNeu5Ac chains in sea urchin sperm flagella (Miyata et al., 2004). In this study, we purified this unique polySia-containing protein and cloned the cDNA coding for its protein backbone. We named it flagellasialin, because it is a heavily sialylated glycoprotein that is localized to the flagellum of sea urchin sperm. Flagellasialin is a novel glycoprotein consisting of 96 amino acids. It has a 26-residue signal peptide at the NH2-terminus and a single putative transmembrane segment without a cytosolic domain at its COOH-terminus. The molecular mass calculated from the amino acid sequence of the mature protein is 7.2 kDa, although it is estimated to be 40–80 kDa based on its relative mobility on SDS–PAGE. This suggests that about 80–90% of this glycoprotein may consist of glycan chains. It should be noted that sulfated α2,9-linked polyNeu5Ac-containing O-glycans show polydispersity in chain length ranging from 2 to >25 Neu5Ac residues. This polydispersity in the polySia chains may explain the polydisperse nature of flagellasialin on SDS–PAGE. Flagellasialin does not stain with silver or Coomassie brilliant blue dye but is detected only by immunostaining with antibodies against glycan structures, that is, anti-Neu5Ac8S (mAb.3G9) and anti-α2,9 Neu5Ac. Few glycoproteins have been reported exhibiting similar polydisperse nature with no detectability by standard gel staining methods, except for fish egg polysialoglycoproteins (Kitajima et al., 1986). Flagellasialin is also characterized by its high content of Thr residues (∼20%). Of 12 Thr residues in the extracellular region, eight are O-glycosylated and three were not glycosylated. The overall structure of flagellasialin is shown in Figure 9. Flagellasialin is devoid of an intracellular domain, and the putative extracellular domain is covered with sulfated α2,9-linked polyNeu5Ac chains. Therefore, flagellasialin looks as if it is a carrier protein of the sulfated α2,9-linked polyNeu5Ac structures exposed on the outer surface of the sperm flagellum.

Fig. 9.

Overall structure of flagellasialin. Flagellasialin consists of a 47-amino acid extracellular domain and a 23-amino acid transmembrane segment with a short or no cytosolic domain. The signal peptide is cleaved off in this figure. In the extracellular domain, of 12 Thr residues, 8 residues are O-glycosylated, and most contain HSO3→8Neu5Acα2,9(Neu5Ac2,9)n–2Neu5Acα2,6GalNAcα1→Thr (n = at least 2) structure.

Fig. 9.

Overall structure of flagellasialin. Flagellasialin consists of a 47-amino acid extracellular domain and a 23-amino acid transmembrane segment with a short or no cytosolic domain. The signal peptide is cleaved off in this figure. In the extracellular domain, of 12 Thr residues, 8 residues are O-glycosylated, and most contain HSO3→8Neu5Acα2,9(Neu5Ac2,9)n–2Neu5Acα2,6GalNAcα1→Thr (n = at least 2) structure.

Flagellasialin is 100% conserved among the sea urchins H. pulcherrimus, S. purpuratus, and S. franciscanus. Interestingly, besides the conserved polypeptide sequences, the three flagellasialins share the highly conserved glycan structure, HSO3→8Neu5Acα2,9(Neu5Ac2,9)n-2Neu5Ac (n = ≥2). There are few glycoproteins that share a highly conserved glycan structure among different animal species. For example, hyosophorins (carbohydrate-rich egg cortical vesicular glycoproteins) from two species of medaka fish, Oryzias latipes and Oryzias melastigma, share the completely identical peptide sequence but bear very different glycan structures (Taguchi et al., 1993, 1994). This suggests that not only the polypeptide part, but also the sulfated α2,9-linked polyNeu5Ac chain in flagellasialin, is of functional importance.

There are, however, subtle differences in core glycan structures of flagellasialin among the three sea urchin species examined herein. The carbohydrate compositions of O‐glycans of flagellasialin are different: flagellasialin from S. purpuratus and S. franciscanus contains Gal and GlcNAc residues besides GalNAc and Neu5Ac, whereas H. pulcherrimus flagellasialin contains only GalNAc and Neu5Ac. The proportions of (Gal + GlcNAc)/GalNAc for flagellasialin from H. pulcherrimus, S. purpuratus, and S. franciscanus are 0, 0.3, and 0.7, respectively. Notably that the higher the proportions of Gal and GlcNAc residues are, the lower the molecular masses of flagellasialin appear to be. The molecular masses for flagellasialin molecules from H. pulcherrimus, S. purpuratus, and S. franciscanus become lower in this order. With regard to the chain length of polySia chain, short chains of α2,9-linked polyNeu5Ac, say DP <5, are relatively more abundant in S. franciscanus flagellasialin as compared with those of two other species. Whether such subtle species-specific differences are of functional importance remains unknown.

In animal sperm, Ca2+ concentrations control flagellar motility, the acrosome reaction, and the fusion of sperm and egg plasma membranes (Darszon et al., 2005). Keeping intracellular Ca2+ low is necessary for normal sperm motility, because high internal Ca2+ impairs sperm motility (Gibbons and Gibbons, 1980). In sea urchin sperm, both the K+-dependent Na+/Ca2+ exchanger (suNCKX) and the plasma membrane Ca2+ ATPase (suPMCA) may keep Ca2+ low (Su and Vacquier, 2002; Gunaratne et al., 2006). suNCKX localizes to the plasma membrane of the sperm flagellum, whereas suPMCA is concentrated in the sperm head plasma membrane. The treatment of sea urchin sperm with the inhibitors of suNCKX or suPMCA results in the elevation of intracellular Ca2+ and loss of motility (Su and Vacquier, 2002; Gunaratne et al., 2006). In this study, we showed that mAb.4F7 inhibited sperm motility and increased intracellular Ca2+. The antibody may block sperm motility due to the elevation of intracellular Ca2+. The fact that the Fab fragment of mAb.4F7 does not increase intracellular Ca2+ suggests the importance of aggregations of flagellasialin caused by IgG for the elevation of intracellular Ca2+. Notably, sperm motility was not inhibited by mAb.3G9, although this epitope co-exists with the mAb.4F7 epitope on the same sulfated α2,9-linked polyNeu5Ac chains. This result indicates that the inhibition of the sperm motility is specific to the α2,9-linked polyNeu5Ac structure.

Although we do not know how flagellasialin influences the intracellular Ca2+ and sperm motility, it is interesting to speculate on the possibility that the α2,9-linked polySia regulates sperm cationic channels, like voltage-sensitive sodium and calcium channels. Vaithianathan and others (2004) have recently shown that the α2,8-linked polySia modulates the activity of specific ionotropic glutamate receptor, called the AMPA receptor. The binding of the α2,8-linked polySia to the AMPA receptor prolongs the channel open time. It has been also shown that the α-subunit of the voltage-sensitive sodium channel is modified by α2, 8-linked polySia (Zuber et al., 1992), and the possible involvement of functional regulation of the sodium channel is suggested. Because there are many channel proteins in the sperm flagellum (Darszon et al., 2005), it is thus interesting to see whether the α2,9-linked polySia on flagellasialin influences electrophysiological properties of sperm.

Materials and methods

Materials

Hemicentrotus pulcherrimus were collected from Chiba coast or purchased from the local fisheries at Fukushima, Japan, and maintained in aquaria in the Center for Education and Research of Field Sciences, Shizuoka University. Strongylocentrotus purpuratus and S. franciscanus were collected at San Diego, CA. Sperm and eggs were collected by the intracoelomic introduction of adults with 0.55 M KCl. An mAb.kdn3G (IgG) recognizing KDNα2,3Gal was prepared as previously described (Sato and Kitajima, 1999). An mAb.12E3, which recognizes α2,8-linked oligo/polyNeu5Ac, was generously gifted by Dr. Tatsunori Seki (Medical school of Juntendo University, Japan) and prepared as previously described (Sato and Kitajima, 1999). An mAb.3G9, which recognizes 8-O-sulfated Neu5Ac (Neu5Ac8S), was prepared as previously described (Ohta et al., 1999). Arthrobacter ureafaciens sialidase was purchased from Nacalai (Kyoto, Japan). Clostridium perfringens sialidase was purchased from Sigma (St. Louis, MO). Actinase E was purchased from Kaken (Tokyo, Japan). Endo-N-acylneuraminidase (Endo-N) was prepared from bacteriophage K1F as previously reported (Hallenbeck et al., 1987). 1,2-Diamino-4,5-methylenedioxybenzene (DMB) was purchased from Dojindo (Kumamoto, Japan). 4-Methylumbelliferyl derivatives of the following sulfated and unsulfated Sia were kindly provided by Dr. K. Furuhata (Kitasato University): Neu5Ac, 8-O-sulfated Neu5Ac (Neu5Ac8S), 9-O-sulfated Neu5Ac (Neu5Ac9S), Neu5Gc, 8-O-sulfated Neu5Gc (Neu5Gc8S), and 9-O-sulfated Neu5Gc (Neu5Gc9S). Colominic acid (α2,8-linked Neu5Ac polymer) was purchased from Wako (Osaka, Japan). 2,9-Linked Neu5Ac polymer was isolated from capsular polysaccharide of NMGC (American Type Culture Collection, Manassas, VA) as previously described (Gotschlich et al., 1969; Bundle et al., 1974). DEAE-Toyopearl was purchased from Tosoh (Tokyo, Japan). Sephacryl S-300, S‐100, Sephadex G-25, G-50 and protein A-Sepharose were obtained from Amersham Tokyo (Japan). WGA-agarose column was obtained from Honen (Tokyo, Japan).

Production of antibodies against α2,9-linked polyNeu5Ac

NMGC was grown in chemically defined media (Kenny et al., 1967). NMGC (1 × 108 cells/mouse) fixed with formaldehyde together with Freund’s complete adjuvant was intraperitoneally injected into the NZBWF1 mouse (female, 4 weeks old, Japan SLC, Hamamatsu, Japan). The animal was boosted twice with NMGC (1 × 108 cells/mouse) mixed with Freund’s incomplete adjuvant every 2 weeks. Blood was collected 1 week after the last boost, and the serum was prepared. The IgG fraction was prepared from the antiserum and designated anti-α2,9 Neu5Ac. Using spleen from mice producing antibodies against NMGC, an mAb.4F7, which was an IgG type, was prepared according to described procedures (Ohta et al., 1999). Fab fragment was generated by papain digestion and were purified by protein A-Sepharose.

Purification of the 40–80 kDa-GP from H. pulcherrimus sperm

Hemicentrotus pulcherrimus sperm were collected as undiluted semen (dry sperm) and kept on ice. Dry sperm (12 mL) were suspended in 100 mL of HEPES-buffered artificial seawater [HASW; 444 mM NaCl, 9 mM KCl, 30 mM MgCl2, 22 mM MgSO4, 10 mM CaCl2, 5 mM NaHCO3, 10 mM HEPES (pH 8.0)]. Coelomocytes were removed by centrifugation at 200 × g for 5 min, and sperm cells sedimented at 1000 × g for 10 min. To solubilize membrane glycoproteins, we suspended the sperm pellet in 150 mL of 10 mM Tris–HCl (pH 7.5), 1% Triton X-100, 150 mM NaCl, 5 mM EDTA, protease inhibitors (1 µg/mL leupeptin, 2 µg/mL antipain, 10 µg/mL benzamidine, 1 µg/mL pepstatin, 0.9 µg/mL aprotinin) on ice for 1 h. After centrifugation at 10,000 × g for 30 min, the supernatant containing solubilized sperm glycoproteins was applied to a DEAE-Toyopearl 650 M column (2.2 × 15 cm) equilibrated with buffer A (1% Triton X-100, 10 mM Tris–HCl, pH 8.0). The column was washed with buffer A containing 0.2 M NaCl and then eluted with buffer A containing 0.4 M NaCl. The 0.4 M NaCl fraction was diluted with equal volume of buffer A and applied to a DEAE-Toyopearl 650 M column and eluted with a linear gradient of NaCl (0.2–0.5 M) in buffer A. The eluents were analyzed for Sia by the resorcinol method (Svennerholm, 1957). SDS–PAGE followed by silver staining using Silver Stain II Kit (Wako) and western blotting using mAb.3G9 (see SDS–PAGE and immunostaining) was performed to detect the 40–80 kDa-GP. The 40–80 kDa-GP-containing fractions were pooled and applied to a WGA-agarose column (2 mL, Honen). The column was washed with 50 mL of buffer B (0.5 M NaCl, 0.1% Triton X-100, 10 mM Tris–HCl, pH 8.0) and eluted with 10 mL of buffer B containing 100 mM GlcNAc. The flow-through fractions were used for amino acid sequencing.

Amino acid sequencing of the purified 40–80 kDa-GP

Peptide fragments of the purified 40–80 kDa-GP were obtained by formic acid hydrolysis as previously described (Inglis, 1983). The 40–80 kDa-GP (200 µg as Sia) was hydrolyzed in 1.3 mL of 0.53 M formic acid at 108°C for 2 h and subjected to a reverse-phase HPLC. The column (Wokopak Handy ODS, 4.6 × 150 mm) was equilibrated with solvent A (0.1% aqueous trifluoroacetic acid) and eluted at 0.2 mL/min with solvent A for 0–10 min, and with gradients of solvent A/solvent B (0.09% trifluoroacetic acid in acetonitrile) 100:0–40:60 (v/v) for 10–130 min. The elution profile was monitored by absorbance at 210 nm. The resulting peptides were sequenced by automated Edman degradation on Procise HT (Applied Biosystems, Foster City, CA).

cDNA cloning of the 40–80 kDa-GP

Total RNA was extracted from H. pulcherrimus, S. purpuratus, and S. franciscanus testis using Trizol reagent following the manufacturer’s protocol (Invitrogen, Carlsbad, CA). Single-strand cDNA was synthesized with random primers and used as a template for the PCR. Degenerated primers (sense: 5′-AC(A/C/G/T)GT(A/C/G/T)GA(C/T) GA(A/G)GA(C/T)ATG-3′; antisense: 5′-CC(C/T)TC(C/T) TC(C/T)TC(A/C/G/T)GC(A/C/G/T)GC-3′) were designed based on the sequences of the 40–80 kDa-GP-derived peptides. PCR products were extended by 3′- and 5′-RACE as previously described (Nakata et al., 2000). The signal sequence was predicted using the SignalP website (www.cbs.dtu.dk/services/SignalP/). The transmembrane segment was predicted using the SOSUI website (http://sosui.proteome.bio.tuat.ac.jp/sosuiframe0.html). cDNA clones for the 40–80 kDa-GP, designated flagellasialin, of S. purpuratus and S. franciscanus were amplified by PCR using the following primers: 5′-CCTTCTGGCCAGAGATTTGATC-3′ (nucleotides –71 to –50 for H. pulcherrimus) and 5′-CAACGCCTTAGAGTAAACACCC-3′ (nucleotides 713 to 734 for H. pulcherrimus).

Northern blot analysis

Total H. pulcherrimus testis RNA (20 µg) was separated on a 1% agarose gel containing 5% formaldehyde and transferred to a nylon membrane (Hybond N+, Amersham Biosciences, Piscataway, NJ). DIG-labeled cDNA for flagellasialin (nucleotides –71 to 734) was prepared according to the manufacturer’s protocol (Roche, Mannheim, Germany) and used as a probe. Hybridization and visualization were performed using a peroxidase-conjugated anti-DIG antibody (1:3000 diluted) as previously described (Trayhurn et al., 1995).

Structural analysis of SGP

A major Sia-containing glycopeptide fraction (SGP) was prepared from sperm of the three sea urchin species as previously described (Miyata et al., 2004), except that Triton X-100 extracts of sperm, instead of acetone powder, were used for the Actinase digestion for S. purpuratus and S. franciscanus.

Analysis of O-glycosylation sites using MALDI-TOF-MS and Edman degradation

SGP (260 µg as Sia) derived from species H. pulcherrimus was hydrolyzed with 0.1 N trifluoroacetic acid at 80°C for 2 h to remove Sia residues. The asialo-SGP was subjected to reverse-phase HPLC and sequenced by automated Edman degradation. For MALDI-TOF-MS, the HPLC-purified glycopeptides (20 µL) were dried and dissolved in 2 µL of water/methanol (1:1, v/v). Samples were spotted onto a MALDI sample target plate with a matrix consisting of a saturated solution of α-cyano-4-hydroxycinnamic acid (Sigma) prepared in 50% acetonitrile/0.1% trifluoroacetic acid. MALDI-TOF-MS measurements were performed in the positive-ion reflector mode by using 4700 Proteomics Analyzer (Applied Biosystems).

Comparative analysis of glycan structures of SGP isolated from the three species

The monosaccharide composition of SGP was determined by the GLC as previously described (Nomoto et al., 1982). The species of Sia was determined by DMB derivatization/reverse-phase HPLC method as previously described (Kitazume-Kawaguchi et al., 1997; Sato et al., 1998; Miyata et al., 2004). PolySia structure was analyzed by the mild acid hydrolysis/fluorometric anion-exchange HPLC method and the alkaline borohydride treatment/anion-exchange HPLC method as previously described (Sato et al., 1999; Miyata et al., 2004).

SDS–PAGE and immunostaining

Sperm extracts with 1% Triton X-100 and purified flagellasialin (0.2 µg as Sia) were treated with or without PNGase F (1 mU, Takara, Kyoto, Japan) at 37°C for 24 h and incubated with Laemmli buffer (Laemmli, 1970) containing 5% mercaptoethanol at 65°C for 15 min. The samples were then electrophoresed on 7.5% polyacrylamide gels and electroblotted onto polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA). After treatment with or without 0.1 N NaOH at 37°C for 18 h, we blocked the membranes with 10 mM sodium phosphate (pH 7.2)/0.15 M NaCl (PBS), containing 0.05% Tween 20, and 1% bovine serum albumin (Sigma) at 25°C for 1 h and then incubated with mAb.3G9 and anti-α2,9 Neu5Ac. As the secondary antibody, peroxidase-conjugated anti-mouse IgM (1:5000 dilution, Zymed Laboratories, San Francisco, CA), or anti-mouse IgG + IgM (1:5000 dilution, American Qulex, San Clemente, CA) was used at 37°C for 1 h, and the color development was carried out as previously described (Sato et al., 2000).

Fertilization experiments

Antibody was dialyzed against HASW before use. Dry sperm were kept on ice until use and diluted 5000–10,000-fold into HASW immediately before use. Nine microliters of sperm suspension was mixed with 1 µL of the antibody and pre-incubated at 15°C for 10 min. Then, 10-µL aliquots of the sperm suspension were added to 90 µL of 10% (v/v) egg suspension and incubated at 15°C for 10 min. More than 200 eggs were microscopically observed, and the number of eggs with elevated fertilization envelopes was counted (Maehashi et al., 2003).

Sperm motility

Sperm was diluted into 0.5% (w/v) polyvinylpyrrolidone in HASW. Sperm motility was observed under dark field conditions. Nine microliters of sperm suspension was mixed with 1 µL of the antibody and photographed with a 0.2- and 1.0-s exposure 1 and 3 min after mixing with antibody. The lengths of swimming trajectories during 0.2 s of 50 individual sperm cells were measured.

Measurement of intracellular Ca2+

Sperm were loaded with fura-2 (Molecular Probes, Eugene, OR) and washed as previously described (Hirohashi and Vacquier, 2002). For the measurements, 20 µL of fura-2-loaded sperm was diluted into 1 mL of HASW. Ca2+ measurements were at 16°C under constant stirring in a FluoroMax-3 fluorometer (Jobin Yvon-Spex, Park Avenue, NJ) with excitation at 340 and 380 nm, and emission at 500 nm. The ratio of intensities at 340 and 380 nm (F340/F380) reports relative intracellular Ca2+. In this article, intracellular Ca2+ is expressed as the F340/F380 ratio.

Acknowledgments

This research was supported in part by Grants-in-Aid for Japan-Canada Joint Health Research Program (to K.K.), Japan Human Science Foundation (to K.K.), CREST of Japan Science and Technology Corporation (to K.K.), the 21st Century COE Program (to K.K.), Young Scientists (B) (16770073) (to C.S.), Scientific Research for Priority Areas (17046006) (to C.S.) from the Ministry of Education, Science, Sports and Culture, and NIH HD12986 (V.D.V.).

Conflict of interest statement

None declared.

  • DP

    degree of polymerization

  • GalNAc

    N-acetylgalactosamine

  • HASW

    HEPES-buffered artificial seawater

  • HPLC

    high-performance liquid chromatography

  • mAb

    monoclonal antibody

  • Neu5Ac

    N-acetylneuraminic acid

  • Neu5Ac8S

    8-O-sulfated Neu5Ac

  • Neu5Gc

    N-glycolylneuraminic acid

  • NMGB

    Neisseria meningitidis group B

  • NMGC

    Neisseria meningitidis group C

  • polySia

    polysialic acid

  • SDS–PAGE

    sodium dodecyl sulfate–polyacrylamide gel electrophoresis

  • SG

    sialic acid-rich glycoprotein

  • SGP

    sialoglycopeptide from the Actinase E digest

  • Sia

    sialic acid

  • WGA

    wheat germ agglutinin

References

Angata
T.
Varki
A.
Chemical diversity in the sialic acids and related α-keto acids: an evolutionary perspective
Chem. Rev.
 , 
2002
, vol. 
102
 (pg. 
439
-
469
)
Barry
G.T.
Goebel
W.F.
Colominic acid, a substance of bacterial origin related to sialic acid
Nature
 , 
1957
, vol. 
179
 
206
Bhattacharjee
A.K.
Jennings
H.J.
Kenny
C.P.
Martin
A.
Smith
I.C.
Structural determination of the sialic acid polysaccharide antigens of Neisseria meningitidis serogroups B and C with carbon 13 nuclear magnetic resonance
J. Biol. Chem.
 , 
1975
, vol. 
250
 (pg. 
1926
-
1932
)
Bundle
D.R.
Jennings
H.J.
Kenny
C.P.
Studies on the group-specific polysaccharide of Neisseria meningitidis serogroup X and an improved procedure for its isolation
J. Biol. Chem.
 , 
1974
, vol. 
249
 (pg. 
4797
-
4801
)
Darszon
A.
Nishigaki
T.
Wood
C.
Trevino
C.L.
Felix
R.
Beltran
C.
Calcium channels and Ca2+ fluctuations in sperm physiology
Int. Rev. Cytol.
 , 
2005
, vol. 
243
 (pg. 
79
-
172
)
Finne
J.
Occurrence of unique polysialosyl carbohydrate units in glycoproteins of developing brain
J. Biol. Chem.
 , 
1982
, vol. 
257
 (pg. 
11966
-
11970
)
Gibbons
B.H.
Gibbons
I.R.
Calcium-induced quiescence in reactivated sea urchin sperm
J. Cell Biol.
 , 
1980
, vol. 
84
 (pg. 
13
-
27
)
Gotschlich
E.C.
Liu
T.Y.
Artenstein
M.S.
Human immunity to the meningococcus. 3. Preparation and immunochemical properties of the group A, group B, and group C meningococcal polysaccharides
J. Exp. Med.
 , 
1969
, vol. 
129
 (pg. 
1349
-
1365
)
Gunaratne
H.J.
Neill
A.T.
Vacquier
V.D.
Plasma membrane calcium ATPase is concentrated in the head of sea urchin spermatozoa
J. Cell Physiol.
 , 
2006
, vol. 
207
 (pg. 
413
-
419
)
Hallenbeck
P.C.
Vimr
E.R.
Yu
F.
Bassler
B.
Troy
F.A.
2nd
Purification and properties of a bacteriophage-induced endo-N-acetylneuraminidase specific for poly-α-2,8-sialosyl carbohydrate units
J. Biol. Chem.
 , 
1987
, vol. 
262
 (pg. 
3553
-
3561
)
Hirohashi
N.
Vacquier
V.D.
Egg sialoglycans increase intracellular pH and potentiate the acrosome reaction of sea urchin sperm
J. Biol. Chem.
 , 
2002
, vol. 
277
 (pg. 
8041
-
8047
)
Ijuin
T.
Kitajima
K.
Song
Y.
Kitazume
S.
Inoue
S.
Haslam
S.M.
Morris
H.R.
Dell
A.
Inoue
Y.
Isolation and identification of novel sulfated and nonsulfated oligosialyl glycosphingolipids from sea urchin sperm
Glycoconj. J.
 , 
1996
, vol. 
13
 (pg. 
401
-
413
)
Inglis
A.S.
Cleavage at aspartic acid
Methods Enzymol.
 , 
1983
, vol. 
91
 (pg. 
324
-
332
)
Inoue
S.
Iwasaki
M.
Isolation of a novel glycoprotein from the eggs of rainbow trout: occurrence of disialosyl groups on all carbohydrate chains
Biochem. Biophys. Res. Commun.
 , 
1978
, vol. 
83
 (pg. 
1018
-
1023
)
Inoue
S.
Poongodi
G.L.
Suresh
N.
Jennings
H.J.
Inoue
Y.
Discovery of an α2,9-polyNeu5Ac glycoprotein in C-1300 murine neuroblastoma (Clone NB41A3)
J. Biol. Chem.
 , 
2002
, vol. 
278
 (pg. 
8541
-
8546
)
James
W.M.
Agnew
W.S.
Multiple oligosaccharide chains in the voltage-sensitive Na channel from electrophorus electricus: evidence for α2,8-linked polysialic acid
Biochem. Biophys. Res. Commun.
 , 
1987
, vol. 
148
 (pg. 
817
-
826
)
Johnson
C.P.
Fujimoto
I.
Rutishauser
U.
Leckband
D.E.
Direct evidence that neural cell adhesion molecule (NCAM) polysialylation increases intermembrane repulsion and abrogates adhesion
J. Biol. Chem.
 , 
2005
, vol. 
280
 (pg. 
137
-
145
)
Kanamori
A.
Inoue
S.
Iwasaki
M.
Kitajima
K.
Kawai
G.
Yokoyama
S.
Inoue
Y.
Deaminated neuraminic acid-rich glycoprotein of rainbow trout egg vitelline envelope. Occurrence of a novel α-2,8-linked oligo (deaminated neuraminic acid) structure in O‐linked glycan chains
J. Biol. Chem.
 , 
1990
, vol. 
265
 (pg. 
21811
-
21819
)
Kenny
C.P.
Ashton
F.E.
Diena
B.B.
Greenberg
L.
A chemically defined protein-free liquid medium for the cultivation of some species of Neisseria
Bull. World Health Org.
 , 
1967
, vol. 
37
 (pg. 
569
-
573
)
Kitajima
K.
Inoue
Y.
Inoue
S.
Polysialoglycoproteins of Salmonidae fish eggs. Complete structure of 200-kDa polysialoglycoprotein from the unfertilized eggs of rainbow trout (Salmo gairdneri)
J. Biol. Chem.
 , 
1986
, vol. 
261
 (pg. 
5262
-
5269
)
Kitazume
S.
Kitajima
K.
Inoue
S.
Troy
F.A.
2nd
Cho
J.W.
Lennarz
W.J.
Inoue
Y.
Identification of polysialic acid-containing glycoprotein in the jelly coat of sea urchin eggs. Occurrence of a novel type of polysialic acid structure
J. Biol. Chem.
 , 
1994
, vol. 
269
 (pg. 
22712
-
22718
)
Kitazume-Kawaguchi
S.
Inoue
S.
Inoue
Y.
Lennarz
W.J.
Identification of sulfated oligosialic acid units in the O-linked glycan of the sea urchin egg receptor for sperm
Proc. Natl. Acad. Sci. U. S. A.
 , 
1997
, vol. 
94
 (pg. 
3650
-
3655
)
Laemmli
U.K.
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
Nature
 , 
1970
, vol. 
227
 (pg. 
680
-
685
)
Liu
T.Y.
Gotschlich
E.C.
Dunne
F.T.
Jonssen
E.K.
Studies on the meningococcal polysaccharides. II. Composition and chemical properties of the group B and group C polysaccharide
J. Biol. Chem.
 , 
1971
, vol. 
246
 (pg. 
4703
-
4712
)
Maehashi
E.
Sato
C.
Ohta
K.
Harada
Y.
Matsuda
T.
Hirohashi
N.
Lennarz
W.J.
Kitajima
K.
Identification of the sea urchin, 350-kDa sperm-binding protein as a new sialic acid-binding lectin that belongs to the heat shock protein 110 family: implication of its binding to gangliosides in sperm lipid rafts in fertilization
J. Biol. Chem.
 , 
2003
, vol. 
278
 (pg. 
42050
-
42057
)
Miyata
S.
Sato
C.
Kitamura
S.
Toriyama
M.
Kitajima
K.
A major flagellum sialoglycoprotein in sea urchin sperm contains a novel polysialic acid, an α2,9-linked poly-N-acetylneuraminic acid chain, capped by an, 8-O-sulfated sialic acid residue
Glycobiology
 , 
2004
, vol. 
14
 (pg. 
827
-
840
)
Nakata
D.
Close
B.E.
Colley
K.J.
Matsuda
T.
Kitajima
K.
Molecular cloning and expression of the mouse N-acetylneuraminic acid, 9-phosphate synthase which does not have deaminoneuraminic acid (KDN) 9-phosphate synthase activity
Biochem. Biophys. Res. Commun.
 , 
2000
, vol. 
273
 (pg. 
642
-
648
)
Nomoto
H.
Iwasaki
M.
Endo
T.
Inoue
S.
Inoue
Y.
Matsumura
G.
Structures of carbohydrate units isolated from trout egg polysialoglycoproteins: short-cored units with oligosialosyl groups
Arch. Biochem. Biophys.
 , 
1982
, vol. 
218
 (pg. 
335
-
341
)
Ohta
K.
Sato
C.
Matsuda
T.
Toriyama
M.
Lennarz
W.J.
Kitajima
K.
Isolation and characterization of low density detergent-insoluble membrane (LD-DIM) fraction from sea urchin sperm
Biochem. Biophys. Res. Commun.
 , 
1999
, vol. 
258
 (pg. 
616
-
623
)
Rohr
T.E.
Troy
F.A.
Structure and biosynthesis of surface polymers containing polysialic acid in Escherichia coli
J. Biol. Chem.
 , 
1980
, vol. 
255
 (pg. 
2332
-
2342
)
Rutishauser
U.
Polysialic acid and the regulation of cell interactions
Curr. Opin. Cell Biol.
 , 
1996
, vol. 
8
 (pg. 
679
-
684
)
Sato
C.
Fukuoka
H.
Ohta
K.
Matsuda
T.
Koshino
R.
Kobayashi
K.
Troy
F.A.
2nd
Kitajima
K.
Frequent occurrence of pre-existing α2→8-linked disialic and oligosialic acids with chain lengths up to, 7 Sia residues in mammalian brain glycoproteins. Prevalence revealed by highly sensitive chemical methods and anti-di-, oligo-, and poly-Sia antibodies specific for defined chain lengths
J. Biol. Chem.
 , 
2000
, vol. 
275
 (pg. 
15422
-
15431
)
Sato
C.
Inoue
S.
Matsuda
T.
Kitajima
K.
Development of a highly sensitive chemical method for detecting alpha2→8-linked oligo/polysialic acid residues in glycoproteins blotted on the membrane
Anal. Biochem.
 , 
1998
, vol. 
261
 (pg. 
191
-
197
)
Sato
C.
Inoue
S.
Matsuda
T.
Kitajima
K.
Fluorescent-assisted detection of oligosialyl units in glycoconjugates
Anal. Biochem.
 , 
1999
, vol. 
266
 (pg. 
102
-
109
)
Sato
C.
Kitajima
K.
Glycobiology of di- and oligosialyl glycotopes
Trends Glycosci. Glycotechnol.
 , 
1999
, vol. 
11
 (pg. 
371
-
390
)
Sato
C.
Kitajima
K.
Tazawa
I.
Inoue
Y.
Inoue
S.
Troy
F.A.
Structural diversity in the α2→8-linked polysialic acid chains in salmonid fish egg glycoproteins
J. Biol. Chem.
 , 
1993
, vol. 
268
 (pg. 
23675
-
23684
)
Schauer
R.
Kelm
S.
Reutter
G.
Roggentin
P.
Shaw
L.
Rosenberg
A.
Biochemistry and role of sialic acids
Biology of the Sialic Acids
 , 
1996
New York and London
Plenum Press
(pg. 
7
-
67
)
SeGall
G.K.
Lennarz
W.J.
Chemical characterization of the component of the jelly coat from sea urchin eggs responsible for induction of the acrosome reaction
Dev. Biol.
 , 
1979
, vol. 
71
 (pg. 
33
-
48
)
Su
Y.H.
Vacquier
V.D.
A flagellar K+-dependent Na+/Ca2+ exchanger keeps Ca2+ low in sea urchin spermatozoa
Proc. Natl. Acad. Sci. U. S. A.
 , 
2002
, vol. 
99
 (pg. 
6743
-
6748
)
Svennerholm
L.
Quantitative estimation of sialic acids. II. A colorimetric resorcinol-hydrochloric acid method
Biochim. Biophys. Acta
 , 
1957
, vol. 
24
 (pg. 
604
-
611
)
Taguchi
T.
Seko
A.
Kitajima
K.
Inoue
S.
Iwamatsu
T.
Khoo
K.H.
Morris
H.R.
Dell
A.
Inoue
Y.
Structural studies of a novel type of tetraantennary sialoglycan unit in a carbohydrate-rich glycopeptide isolated from the fertilized eggs of Indian Medaka fish, Oryzias melastigma
J. Biol. Chem.
 , 
1993
, vol. 
268
 (pg. 
2353
-
2362
)
Taguchi
T.
Seko
A.
Kitajima
K.
Muto
Y.
Inoue
S.
Khoo
K.H.
Morris
H.R.
Dell
A.
Inoue
Y.
Structural studies of a novel type of pentaantennary large glycan unit in the fertilization-associated carbohydrate-rich glycopeptide isolated from the fertilized eggs of - Oryzias latipes
J. Biol. Chem.
 , 
1994
, vol. 
269
 (pg. 
8762
-
8771
)
Trayhurn
P.
Thomas
M.E.
Duncan
J.S.
Black
D.
Beattie
J.H.
Rayner
D.V.
Ultra-rapid detection of mRNAs on northern blots with digoxigenin-labelled oligonucleotides and CDP-Star, a new chemiluminescence substrate
Biochem. Soc. Trans.
 , 
1995
, vol. 
23
 pg. 
494S
 
Troy
F.A.
Rosenberg
A.
Sialobiology and the polysialic acid glycotope: occurrence, structure, function, synthesis, and glycopathology
Biology of the Sialic Acids
 , 
1996
New York and London
Plenum Press
(pg. 
95
-
144
)
Vacquier
V.D.
Moy
G.W.
The fucose sulfate polymer of egg jelly binds to sperm REJ and is the inducer of the sea urchin sperm acrosome reaction
Dev. Biol.
 , 
1997
, vol. 
192
 (pg. 
125
-
135
)
Vaithianathan
T.
Matthias
K.
Bahr
B.
Schachner
M.
Suppiramaniam
V.
Dityatev
A.
Steinhaer
C.
Neural cell adhesion molecule-associated polysialic acid potentiates alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor currents
J. Biol. Chem.
 , 
2004
, vol. 
279
 (pg. 
47975
-
47984
)
Yabe
U.
Sato
C.
Matsuda
T.
Kitajima
K.
Polysialic acid in human milk
J. Biol. Chem.
 , 
2003
, vol. 
278
 (pg. 
13875
-
13880
)
Yang
P.
Major
D.
Rutishauser
U.
Role of charge and hydration in effects of polysialic acid on molecular interactions on and between cell membranes
J. Biol. Chem.
 , 
1994
, vol. 
269
 (pg. 
23039
-
23044
)
Zuber
C.
Lackie
P.M.
Catterall
W.A.
Roth
J.
Polysialic acid is associated with sodium channels and the neural cell adhesion molecule N-CAM in adult rat brain
J. Biol. Chem.
 , 
1992
, vol. 
267
 (pg. 
9965
-
9971
)