| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Cell and Developmental Biology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
fertilization, Sertoli cells, sperm, sperm maturation, spermatid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Zonadhesin is synthesized in the testis as a precursor protein that is subsequently proteolytically processed [4, 5]. The precise site of protein cleavage and the stage or stages of sperm development at which it occurs are not known, but a peptide sequence from native pig zonadhesin isolated from the membranes of ejaculated spermatozoa demonstrated the presence of only three D domains, D1, D2, and D3 [5]. From this observation the loss of the N-terminal MAM and mucin domains is inferred, and the zona pellucida binding site of zonadhesin is therefore located within the D domains of the protein.
Accordingly, we have studied the processing and localization of the zonadhesin protein from spermatogenesis until fertilization. We show that the precursor protein occurs exclusively in the testis and that proteolytic processing results in the formation of three fragments. These are p43 (D1 domain), p97 (D2–D4 domains) and p58 (D4 domain-C-terminal). In spermatozoa, the p43 and p97 fragments exist as disulfide-bonded dimers localized exclusively on the anterior acrosome of acrosome-intact spermatozoa. After the acrosome reaction, zonadhesin is found associated with the acrosomal shroud that is formed from fused, vesiculated plasma and outer acrosomal membranes held together by insoluble acrosomal material [6]. In addition, we show that the recombinant D4 domain can bind zona pellucida, and that there appears to be no domain specificity in zona pellucida binding. We propose that zonadhesin functions during fertilization to anchor the acrosomal shroud to the zona pellucida, allowing the acrosome-reacted spermatozoon to penetrate the zona pellucida.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Complementary DNA Synthesis and Sequencing
Pig zonadhesin cDNAs were kindly provided by Dr. D. Garbers, Southwestern Medical School, University of Texas, Dallas, TX. Total RNA was isolated from adult rabbit testis using the RNeasy mini kit (Qiagen Inc., Valencia, CA) and used in first-strand cDNA synthesis according to the protocols of 5' and 3' rapid amplification of cDNA ends (RACE) system by Gibco BRL (Grand Island, NY). Complementary DNA was subsequently tailed and amplified by polymerase chain reaction (PCR) using nested, gene-specific primers (according to the 3' and 5' RACE method of Gibco BRL). Degenerate primers used in the initial isolation of rabbit zonadhesin from rabbit testis RNA were CTGGAGCTTGTGGTCAAC (sense, base pairs [bp] 5599–5616), and CAGGGTGTCATACAGCTCTG (antisense, bp 6275–6256).
Products were purified using a PCR purification kit (Qiagen) and subsequently cloned using a TOPO TA cloning reaction into pCR2.1-TOPO according to the manufacturer's instructions (Invitrogen, Carlsbad, CA). When the amplified cDNA produced bands on a 1% agarose gel, these were excised and the DNA was purified using the QIAquick gel extraction kit (Qiagen) and cloned into pCR2.1-TOPO according to the manufacturer's instructions.
Sequencing was performed at the University of North Carolina automated sequencing facility on a Model 377 DNA sequencer with an ABI prism rhodamine terminator cycle sequencing ready kit (Applied Biosystems, Foster City, CA). Sequence analysis was carried out with DNAsis software (Hitachi Software Engineering Co. Ltd., South San Francisco, CA).
Zonadhesin Antibodies
Four recombinant proteins, D1, D2/3, D3/4, and the C-terminal were expressed (in bacteria) from PCR-generated cDNA and cloned into the pQE-30 (Qiagen) vector. The proteins were purified on a Ni2+-NTA column according to the method of Lea et al. [7], and each was used to immunize three BALB/c mice. Four peptide sequences were also synthesized and coupled to keyhole limpet hemocyanin (KLH) via a C-terminal cysteine residue. These were each used to immunize three BALB/c mice. Each mouse was immunized with either 3–5 µg recombinant proteins in 250 µl PBS or 10 µg of peptide-KLH conjugate in 50 µl of PBS. These were emulsified with an equal volume of Freund complete adjuvant and immunized s.c. at two sites (Week 0). Similar immunizations were made at Weeks 3 and 5, except that the immunogen was emulsified with Freund incomplete adjuvant. Synthesized recombinant proteins and peptide sequences are summarized in Table 1.
|
Mouse anti-P16C antiserum (diluted 1:1 with PBS) was affinity-purified on a P16C sulfolink column (Pierce, Rockford, IL) and dialyzed against PBS. Preparation of the sulfolink column and purification of the antiserum were performed according to the manufacturer's instructions.
Electrophoresis and Blotting
SDS-PAGE (using either 7.5% or 10% gels) was performed as described in [8]. Western blots were performed as described in [9] using Immobilon P as the transfer membrane. Antibody staining of Western blots was carried out as described in [10], except that mouse anti-zonadhesin antibodies were detected with alkaline phosphatase-conjugated goat anti-mouse immunoglobulin (Ig) G at a 1:2000 dilution.
Rabbit Testis Lysate Preparation
Rabbit testis lysates were prepared by homogenizing frozen and thawed rabbit testis (Pel-Freez Biologicals, Rogers, AR) in five pellet volumes (1–2 ml) of TBS (50 mM Tris·Cl pH 7.4, 150 mM NaCl) containing protease inhibitors (4 mM Pefabloc SC, 100 µM leupeptin, and 10 µg/ml aprotinin). The homogenate was made 50 mM for 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid (CHAPS) pH 7.4, rocked for 1 h at 4°C, and the tissue debris was removed by centrifugation (1000 x g for 10 min). The supernatant was then recentrifuged at 4000 x g for 10 min and the postnuclear supernatant was made 1% for sodium deoxycholate. This was incubated for 20 min on ice and then dialyzed against Tris-buffered saline (TBS) at 4°C with extensive changes of buffer.
For electrophoresis, 700 µg of rabbit testis lysate was mixed with an equal volume of sample buffer (final concentration, 62.5 mM Tris·Cl pH 6.8, 2% [w/v] SDS, 5% [v/v] glycerol, 5% [v/v] ß-mercaptoethanol) and heated to 100°C for 5 min. Before loading onto the acrylamide gel, samples were centrifuged (11 000 x g) for 5 min.
Rabbit Spermatozoa Preparation
Epididymal spermatozoa were flushed from rabbit cauda epididymides using McCoy 5A medium (Sigma Aldrich Co., St. Louis, MO). The spermatozoa were collected and washed twice in the same medium and pelleted by centrifugation at 2000 x g for 4 min.
Ejaculated spermatozoa were collected with an artificial vagina and ejaculates with good motility (>90% ± 5%) were processed further. Pooled ejaculates were washed twice by centrifugation at 700 x g for 5 min in 15 ml PBS (pH 7.0) containing protease inhibitor cocktail set I (Calbiochem Novabiochem Corp., San Diego, CA), followed by washing in 25 mM Tris·Cl pH 7.4/150 mM NaCl (Tris/NaCl).
Spermatozoa pellets (epididymal or ejaculated) were resuspended in one pellet volume (100–200 µl) of Tris/NaCl containing 1% Triton X-100 (with protease inhibitor cocktail set I at the manufacturer's suggested dilution) and rocked for 1–1.5 h at 4°C. Afterward, the spermatozoa were centrifuged at 3000 x g for 5 min before the supernatant was removed and diluted 1:1 with Tris/NaCl.
Ejaculated rabbit sperm were capacitated by centrifugation through a Percoll gradient according to the method of O'Rand and Fisher [11].
Zona Pellucida
Heat-solubilized rabbit zonae pellucida (HSRZP) were prepared as described previously [12, 13]. HSRZP (19 µg) were biotinylated using sulfo-NHS-biotin (Pierce) according to the manufacturer's instructions. Unreacted biotin was subsequently removed with Chromaspin-30 columns (Clontech Laboratories Inc., Palo Alto, CA).
Immunofluorescence
Epididymal, ejaculated, and capacitated rabbit spermatozoa, which exhibited greater than 95% ± 5% motility, were washed with McCoy 5A medium, smeared onto slides, and fixed in ice-cold 100% methanol (15 min, 4°C). Slides were then washed in PBS (5 min) and blocked in PBS containing 3% normal goat serum (NGS) and 1 mM EDTA pH 8.0 (30 min at room temperature). Experiments were performed four times using different populations of spermatozoa.
To induce the acrosome reaction we used the zona-coated coverslip method described in [14] with some modification. HSRZP (400 ng) was air-dried onto microscope slides, which were inverted over a 100-µl drop of capacitated rabbit spermatozoa containing 1 x 106 sperm. After a 45-min incubation at 37°C in 5% CO2, the slides were transferred to cold methanol and fixed for 30 min at 4°C. Slides were further processed according to the method of Richardson et al. [14] using primary antibody (affinity-purified mouse anti-P16C; Table 1) at a 1:2000 dilution and secondary antibody (biotin-conjugated goat anti-mouse IgG) at a 1:200 dilution. Antibody binding was detected using avidin-Texas red at a 1:200 dilution. Control slides were incubated with affinity-purified mouse anti-P16C antiserum preincubated with P16C sulfolink column matrix for 1 h. Before the addition of the antibody to the spermatozoa, the column matrix was removed by centrifugation at 1000 x g for 2 min.
In Situ Hybridization and Immunohistochemistry
In situ hybridization and immunohistochemical staining were performed on paraffin-embedded, Bouin-fixed rabbit testes. Immunohistochemistry sections (8 µm) were probed with affinity-purified mouse anti-P16C antiserum (Table 1) at a 1:1000 dilution. Control sections were probed with absorbed affinity-purified P16C antibody. Antibody binding was detected as described in [15]. For in situ hybridization, sections (8 µm) were pretreated with 0.01 M citrate buffer pH 6.0 and heated for 15 min. After cooling, they were incubated with a 315-bp riboprobe (bp 4186–4501) derived from the zonadhesin D3 domain at a 1:100 dilution for 2 h at 50°C. Sense and antisense probes were synthesized by in vitro transcription according to the Maxiscript protocol (Ambion Inc., Austin, TX) using digoxigenin-11-UTP (Roche Molecular Biochemicals, Indianapolis, IN) as the labeled nucleotide. Probe binding was detected using horseradish peroxidase-conjugated anti-digoxigenin (Roche Molecular Biochemicals) at a 1:100 dilution overnight (4°C). The sections were stained with diaminobenzidine, counterstained with toludine blue, and mounted.
HeLa Cell Transfection and Immunofluorescence
A zonadhesin construct containing the D4 domain was synthesized by PCR and cloned into pDisplay vector (Invitrogen). This vector contains both the Ig kappa-chain secretion signal, a C-terminal myc epitope tag, and the transmembrane anchoring domain of platelet-derived growth factor receptor, and targets proteins to the cell surface. HeLa cells were transiently transfected by electroporation using a Gene Pulser Electroprotocol procedure described at the Bio-Rad (Hercules, CA) web site (http://www.discover.bio-rad.com). Briefly, exponentially growing HeLa cells cultured in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal calf serum (FCS; Hyclone Labs, Inc., Logan, UT) were trypsinized, washed, and resuspended at 1.3 x 107 cells/ml in RPMI medium containing 10 mM dextrose and 0.1 mM dithiothreitol. Aliquots of cells (300 µl) were incubated with 8 µg DNA in 0.4-cm gap-width cuvettes (5 min at room temperature). Electroporation was performed at 0.3 kV with a capacitance of 500 microfarads, after which cells were gently resuspended in 500 µl DMEM containing 10% FCS, and were seeded onto Lab-Tek II chamber slides (Fisher Scientific, Pittsburgh, PA). Transfected cells were incubated for 48 h at 37°C in 5% CO2, after which they were washed in serum-free DMEM, and live cells were incubated with biotinylated HSRZP (2 µg) in serum-free medium for 1 h. Cells were then washed twice in serum-free DMEM, twice in PBS, and fixed in cold methanol (-20°C) for 15 min. After further washing in PBS, cells were labeled with mouse anti-myc (Invitrogen) at a 1:200 dilution in PBS containing 0.1% BSA. Zonadhesin-expressing cells were detected with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG Fc, and zona pellucida binding was detected with goat anti-biotin antibody (Pierce) and rhodamine-conjugated rabbit anti-goat IgG (H + L, Pierce) both at a 1:250 dilution. Slides were mounted in Vectashield (Vector Labs, Inc., Burlingame, CA) and viewed under a microscope.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To obtain rabbit zonadhesin sequence, degenerate primers were made from conserved zonadhesin D4 domain sequences (conserved between pig and human zonadhesin) and were used to amplify rabbit testis cDNA by PCR. The resulting 676 bp product was cloned, sequenced, and shown by homology to pig to encode rabbit zonadhesin (data not shown). To extend this sequence, 5' and 3' RACE methodologies were utilized with nested, gene-specific primers to overlapping fragments to generate a cDNA sequence 7007 bp in length with a predicted open reading frame (ORF) of 2282 amino acids. Alternative splicing produced three 5' sequences, one of which contained a predicted signal sequence of 10 amino acids [16] and a noncanonical Kozak consensus sequence to predict the translational start site [17]. The 3' untranslated region (UTR) sequence was 149 bp in length with a canonical polyadenylation signal sequence (AATAAA) between nucleotides 6970 and 6975. These sequence data are available from GenBank, accession number AF244982.
The rabbit zonadhesin ORF predicts a large, multiple-domain protein (Fig. 1A) containing four different types of domains: MAM, mucin, D, and EGF. At the N-terminal of the protein are two MAM domains, one partial and one complete spanning amino acids 36–314. These are followed by a mucin-like domain (amino acids 315–498) that spans 183 amino acids and largely consists of 20 imperfect seven-amino acid repeats with the consensus sequence (P/T T V P P/T E P/E). Carboxyl to this are five D domains, one partial (D0, amino acids 499–610) and four complete (D1–D4, amino acids 611–2189). Like other D domain-containing proteins, rabbit zonadhesin shows conservation of nearly all cysteine residues, and the TFDG and GLCG motifs (present in the D1, D2, and D4 domains). In addition, the CXXC motif occurring in the D1, D2, and D3 domains is conserved. On the carboxyl side of the D domains is an EGF-like domain (amino acids 2209–2220), following which is a hydrophobic sequence that is predicted to encode a transmembrane spanning domain (amino acids 2237–2256) and a short intracellular sequence of only 26 amino acids.
|
Comparison of rabbit zonadhesin with zonadhesin cloned from other species shows that rabbit zonadhesin is 62% identical to pig zonadhesin [5] (Fig. 1B) with an identical domain structure, and 30% identical to mouse zonadhesin (Fig. 1C). The low level of identity seen with mouse zonadhesin is caused by mouse zonadhesin having three MAM domains and the addition of 20 partial D3 repeats [4].
Processing of Rabbit Zonadhesin
It has previously been reported that zonadhesin is processed on mature spermatozoa such that the N-terminal third of the protein is removed and two fragments are generated, both of which encode D domain sequence [5]. In order to address the likelihood of full-length zonadhesin performing a role either in spermatogenesis or during epididymal maturation, we have examined the occurrence of unprocessed zonadhesin during these events. CHAPS extracts of rabbit testis lysate were separated by 7.5% reducing SDS-PAGE and Western blotted. Using antibodies raised to the zonadhesin C-terminal recombinant protein (Table 1), full-length zonadhesin was detected as a protein of approximately 243 kDa (Fig. 2, lane 1). When probed with antibodies to either the D1 or D3/4 domain recombinant proteins (Table 1), in addition to full-length zonadhesin, smaller proteins in the range of 130–40 kDa were detectable (data not shown). Preimmune antiserum detected no immunoreactive proteins (Fig. 2, lane 2).
|
When Triton X-100 extracts of cauda epididymal spermatozoa were separated on 7.5% SDS-PAGE, Western blotted, and probed with domain-specific antisera, full-length zonadhesin was no longer detectable. Instead, zonadhesin occurred predominantly as three distinct fragment sizes. Using antibodies to the D3/4 domains (Table 1), two immunoreactive proteins were detectable at 97 kDa and 85 kDa (Fig. 2, lane 3). In addition to these, some larger proteins were weakly immunoreactive and probably resulted from the incomplete reduction of the sperm lysate. Antibodies to the D1 domain (Table 1) recognized immunoreactive proteins at 43 kDa (Fig. 2, lane 4) and antibodies to the C-terminal weakly recognized a single band of 58 kDa. (Fig. 2, lane 5, indicated with *). No immunoreactivity was detectable with preimmune serum (lane 6).
To determine whether any further processing of zonadhesin occurred after ejaculation, membrane proteins were extracted from ejaculated rabbit sperm using 1% Triton X-100. On Western blotting and probing with the D1, D3/4, and CT antisera (Table 1) as before, the same size fragments were evident (Fig. 2, lanes 7–10). Probing these blots with antiserum raised to D2/3 recombinant (Table 1) revealed proteins of 97 and 85 kDa that were identical to those detected with the D3/4 antibodies (data not shown).
In order to confirm the identities of the different sized fragments we synthesized four peptide sequences (Table 1) specific for different D domains, immunized mice, and produced mouse antisera. When used on Western blots of ejaculated rabbit sperm lysates the D1 peptide antibody recognized both the 43 and 62 kDa proteins, and the D2 and D3 peptide antisera recognized the 97- and 85-kDa bands (data not shown).
These data demonstrate that proteolytic processing of zonadhesin gives rise to three fragments: p97 (97 and 85 kDa, D2–D4 sequence), p43 (43 kDa, D1 sequence), and p58 (58 kDa, D4 and C-terminal sequence).
Zonadhesin in Spermatozoa
In addition to the D domains of rabbit zonadhesin being rich in cysteine (8.5%), the D1, D2, and D3 domains each contain the CXXC motif. To demonstrate whether native zonadhesin occurred in spermatozoa as disulfide-bonded complexes we separated Triton X-100 extracts of ejaculated sperm lysates by 7% nonreducing SDS-PAGE. When blots were probed with antibodies raised either to D1 or D3 peptide sequences (see Table 1), the same size proteins were detected, 138 kDa and 120 kDa (Fig. 3, lanes 1 and 2). When probed with antibodies raised to an intracellular sequence (K11C; Table 1) a single protein of 44 kDa was detected (Fig. 3, lane 3). The disparity in apparent molecular weight between reduced and nonreduced C-terminal fragments (58 kDa compared with 44 kDa) is likely the result of intramolecular disulfide bond formation. Similarly, the sperm surface protein fertilin ß demonstrated an increase in molecular weight between nonreduced and reduced protein conformations [18]. No proteins were detectable with preimmune antiserum (Fig. 3, lane 4).
|
Localization of Zonadhesin in Testis and Spermatozoa
Testis To examine the localization of zonadhesin during spermatogenesis we performed immunohistochemical staining of serial sections of rabbit testis using affinity-purified antibody to the D3 peptide sequence, P16C (see Table 1). Zonadhesin was present in all stages of spermatogenesis, appearing first at low levels in pachytene spermatocytes, with increasing amounts detectable in round and elongating spermatids. Spermatogonia and early primary spermatocytes did not express zonadhesin (Fig. 4A). A study of the pattern of staining demonstrated the presence of zonadhesin throughout the cytoplasm of germ cells (Fig. 4, A and C). In round spermatids zonadhesin was detected around the periphery of the nucleus (Fig. 4C, arrows) in a pattern characteristic of the developing acrosome. In addition to germ cell expression of zonadhesin, stage-specific expression of the protein was also detected in Sertoli cells. Somatic cell synthesis of zonadhesin appeared to be restricted exclusively to stages 2 and 3 of the cycle in the rabbit seminiferous epithelium [19] when round spermatids initiate and complete elongation (Fig. 4B). When the seminiferous epithelium is sectioned longitudinally, staining is localized to the region of the lateral processes from adjacent Sertoli cells (Fig. 4D, arrows). Probing of rabbit testis sections with P16C antiserum absorbed with P16C peptide showed negligible staining (Fig. 4A, inset).
|
When the cell specificity of zonadhesin gene expression was determined by in situ hybridization, binding of a D3 antisense riboprobe (Fig. 4E) to the testis section adjacent to that probed with the D3 antibody (Fig. 4B) also revealed zonadhesin expression in all tubules. In contrast to protein expression, however, zonadhesin mRNA was most abundant in primary spermatocytes, specifically preleptotene/leptotene and pachytene spermatocytes (in all stages), with round spermatids showing lower levels of expression (Fig. 4E). Low levels of staining were detectable in elongating spermatids and Leydig cells. Positive staining at the base of the germinal epithelium indicated that transcription occurred in both or either spermatogonia and Sertoli cells. Despite comparison of in situ hybridized sections with adjacent sections that had been stained with hematoxylin and eosin (Fig. 4, E and F), it was not possible to distinguish between these cell types after RNA detection. Testis sections were probed with a D3 sense riboprobe as control and showed very low levels of background staining (data not shown).
Spermatozoa To determine the localization of zonadhesin during maturation of rabbit spermatozoa, immunofluorescence was performed on methanol-fixed, washed cauda epididymal spermatozoa, washed ejaculated spermatozoa, capacitated spermatozoa, and sperm that had been induced to undergo a zona pellucida-induced acrosome reaction. Each sample exhibited greater than 95% motility before fixation and was probed with affinity-purified antiserum to the D3 peptide sequence, P16C (Table 1). Figure 5A shows capacitated rabbit spermatozoa in which zonadhesin was localized exclusively to the anterior acrosome in more than 98% of spermatozoa. No significant staining of the postacrosomal region, midpiece, or principal piece of the tail was observed. Similarly, epididymal and ejaculated spermatozoa showed an identical pattern of staining (data not shown). In spermatozoa that had undergone an acrosome reaction on the zona pellucida (Fig. 4, C–E), zonadhesin staining was no longer evident (Fig. 5E, labeled with *). Instead, D3 peptide sequence was recognized on the many acrosomal shrouds that had been lost from acrosome-reacted spermatozoa and remained attached to the zona pellucida (Fig. 5C, arrows). The loss of zonadhesin with the acrosome could also be observed in sperm that were in the process of undergoing the acrosome reaction. In these sperm (Fig. 5E, labeled with an arrowhead), the acrosomal membranes were seen to be lifting away from the sperm head (shown in Fig. 5D inset by arrowhead). No staining of spermatozoa was observed when D3 antiserum was preincubated with an excess of D3 peptide (Fig. 5F). When live spermatozoa were probed with antibody to the D3 peptide sequence, no immunofluorescence was observed (data not shown).
|
Zonadhesin-Zona Pellucida Binding
Isolation of native pig zonadhesin D1, D2, and D3 domains from a zona pellucida affinity column [5] suggests that each of these domains is capable of binding zona pellucida, yet the function of the D4 domain remains unclear. We therefore assayed the ability of the D4 domain to bind zona pellucida. Because enzymatic deglycosylation of the native protein indicates zonadhesin to be a glycosylated protein (data not shown) and bacterially expressed zonadhesin recombinant proteins are insoluble, we performed an in vitro zona pellucida binding assay on mammalian HeLa cells. Recombinant D4 domain was transiently transfected into HeLa cells and synthesized from the vector, pDisplay, which targets proteins to the cell surface. Mock transfected cells received pDisplay containing no insert. To assay for zona pellucida binding to the surface of D4-transfected HeLa cells we performed a double-label experiment. Live cells were incubated for 1 h with biotinylated heat-solubilized rabbit zona pellucida, washed to remove unbound zona pellucida, and fixed. Zona pellucida binding was then detected with goat anti-biotin antibody and rhodamine-conjugated anti-goat IgG. Zonadhesin-expressing cells were detected after fixation with an antibody to the myc epitope and FITC-conjugated goat anti-mouse IgG. Figure 6, A–D shows the same HeLa cells expressing recombinant D4 domain (A and C) and binding heat-solubilized zona pellucida (B and D). At higher magnification (E and F) it is evident that zona pellucida binding is restricted to sites of zonadhesin expression. There was no evidence of significant antigen patching occurring on addition of zona pellucida. Mock transfected cells and transfected cells that did not express the D4 domain bound no zona pellucida (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In a previous study [4] it was suggested that the N-terminal (MAM and mucin) domains of zonadhesin may function in preventing nonspecific interactions between spermatozoa and the female reproductive tract. Our evidence shows that, at least in rabbit, this is unlikely to be the case. Full-length zonadhesin is present in the testis (as a reduced protein of approximately 243 kDa), but the protein is processed and the N-terminal domains are removed prior to spermatozoa reaching the female reproductive tract. Processing of the protein appears to begin in the testis, as evidenced by the presence of multiple D domain-containing fragments when lysates are probed with D1- or D3/4-specific antibodies. How these fragments relate to either full-length or processed zonadhesins is unclear, but our inability to detect p97, p43, and p58 protein fragments in testis suggests that zonadhesin processing is incomplete when sperm leave the testis. However, by the time spermatozoa reach the cauda epididymis, processing of zonadhesin has been completed and the N-terminal domains (approximately 61 kDa) have been lost. The proteolytically cleaved fragments remaining are p97 (97 and 85 kDa, D2–D4 sequence), p43 (43 kDa, D1 sequence), and p58 (58 kDa, D4 and C-terminal sequence). The exact sites at which protease cleavage occurs is not known; however, there are several dibasic processing site motifs [20] occurring in appropriate regions of the protein. These sites can be cleaved by dibasic processing endoproteases and have been indicated in the processing of proteins such as von Willebrand factor precursor [21]. Two of the processed fragments of rabbit zonadhesin (p97 and p43) are comparable with those found for pig zonadhesin, where the p105 fragment is composed of a D2/D3 sequence and the p45-containing D1 sequence [5].
Although it is clear from studies of pig zonadhesin that proteolytic processing occurs, what is less evident is how these fragments are associated in spermatozoa. We have therefore studied nonreduced lysates of spermatozoa and demonstrated the formation of complexes of zonadhesin fragments. p97 (D2–D4) and p43 (D1) have been demonstrated to occur as disulfide-bonded complexes of 138 and 120 kDa. The existence of two different complex sizes presumably arises from the doublet of p97 proteins seen on reducing gels. Unexpectedly though, the D4-C-terminal fragment (p58) did not appear to associate with this complex despite the presence of numerous cysteine residues in the D domains and C-terminal. Perhaps this is in part due to the presence or absence of the CXXC motif in each domain. Whereas the D1, D2, and D3 domains each have at least one CXXC motif, the D4 domain and C-terminal do not. It is therefore possible that the lack of association between the p97/p43 complex and the D4-C-terminal fragment is due to the absence of the viscinal cysteines in the CXXC motif. Similarly, von Willebrand factor, a protein containing D domains that is analogous to zonadhesin, is known to form high molecular weight multimers by disulfide bond formation using the viscinal cysteine residues of the CXXC motif. This process is believed to be distinct from the formation of disulfide bonds through single cysteine residues [22].
Therefore, we conclude that in spermatozoa, zonadhesin occurs as two noncovalently bound subunits of p138 and p44. The p138 subunit (consisting of the disulfide-bonded p97 and p43) contains domains D0 (partial), D1, D2, D3, and D4 (partial), whereas the p44 subunit contains the remaining part of D4 and the C-terminal domains (EGF, transmembrane, and intracellular sequences). An approximate p61 N-terminal fragment is lost as spermatozoa leave the testis.
Studying the expression pattern of the zonadhesin gene and protein in the testis we have shown that RNA production is at high levels in primary spermatocytes (from preleptotene/leptotene to pachytene). In later stages of spermatogenesis, the levels of zonadhesin mRNA decline, suggesting either that transcription continues but at lower levels in haploid spermatids, or that transcription ceases but the mRNA persists. These findings contrast with those of Hardy and Garbers [5] who showed that pig zonadhesin is exclusively a haploid expressed gene. It is possible that this disparity arises from species-specific differences in the pattern of gene expression. Coincident with mRNA synthesis, zonadhesin protein was also present but at low levels in pachytene spermatocytes. Subsequently, as the level of mRNA declined, more protein appeared present in both round and elongating spermatids, which could imply some degree of translational control of zonadhesin expression. Germ cell expression of zonadhesin occurred in all stages of the cycle and was localized throughout the cell cytoplasm and in the developing acrosome. We cannot rule out the possibility that zonadhesin was also present in the plasma membrane of the cell but electron microscopy would be needed to resolve this possibility. In contrast to germ cell synthesis, Sertoli cells synthesized zonadhesin in a stage-specific manner, in stage 2 and 3 tubules [19]. Protein was localized around the developing germ cells, in particular round spermatids, in the region of the lateral processes of adjacent Sertoli cells. Using in situ hybridization and hematoxylin and eosin staining of serial sections we were unable to show definitively that Sertoli cells synthesized zonadhesin mRNA, probably because of the section thickness. Further studies using separated testicular cells and PCR will be needed to confirm Sertoli cell synthesis of zonadhesin mRNA. Nevertheless, it is interesting to speculate on the function of zonadhesin in the testis. If full-length zonadhesin is expressed on the cell surface of both spermatids and Sertoli cells, it may function in cell adhesion through the MAM and mucin domains. From studies of other MAM domain-containing proteins, it is proposed that these domains can bring about homophilic protein-protein interactions [23]. Testicular zonadhesin expressed on the surface of Sertoli and germ cells may therefore likewise interact to promote cell-cell interaction. In addition, mucin domains have been shown to interact with the selectin family of carbohydrate binding proteins [24]. Because rat Sertoli cells are known to express L-selectin on their surface [25–27], it is possible that the zonadhesin mucin-like domain and L-selectin can contribute to the binding of germ cells to the Sertoli cell during spermatogenesis. Loss of the 61-kDa N-terminal fragment as zonadhesin is processed may contribute to the release of testicular spermatozoa from the Sertoli cell.
Once spermatozoa leave the testis and enter the epididymis they begin the process of maturation that continues through the epididymis and then into the female reproductive tract as they undergo capacitation. Despite the remodeling of the plasma membrane that occurs during this time, zonadhesin remains localized on the anterior sperm head in cauda epididymal, ejaculated, and capacitated spermatozoa. However, once spermatozoa encounter the zona pellucida and the acrosome reaction is induced, the plasma and outer acrosomal membranes vesiculate and lift away from the sperm as a shroud [6, 14, 28]. In rabbits these membranes can be distinguished by phase microscopy as acrosomal shrouds. It is with the shrouds that zonadhesin is associated after the acrosome reaction, allowing the zonadhesin-negative, acrosome-reacted spermatozoa to continue penetrating the zona pellucida and the shroud to remain anchored on the zona pellucida. The inability of D3 antibodies to detect zonadhesin on the surface of live spermatozoa suggests that the antigen is most likely located in the outer acrosomal membrane rather than the plasma membrane.
In addition to studying localization of zonadhesin on spermatozoa, we have also considered whether recombinant rabbit zonadhesin D4 domain has the ability to bind homologous zona pellucida. Although there is some evidence for D domain-specific function in other proteins, in pig, zonadhesin fragments containing D1, D2, and D3 domains bind the zona pellucida [1]. At present, the function of the D4 domain has not been reported. In order to assay whether this domain is capable of binding zona pellucida, we have performed a novel qualitative zona pellucida binding assay in HeLa cells. Native zonadhesin D domains are glycosylated, and because they have a high cysteine content (8.5%) they are insoluble when produced as recombinant proteins in bacteria. In consequence, we used mammalian cell-synthesized recombinant protein, which produces a more physiologically representative model of sperm-zona binding. Our results show that HeLa cell-synthesized D4 domain can bind rabbit zona pellucida, which in turn suggests that zonadhesin D domains do not exhibit specificity for D domain function. Because the level of sequence identity between rabbit zonadhesin D domains is not high (35%–42%), but structurally the domains are similar, it is probable that the binding of zonadhesin to zona pellucida does not occur in a sequence-specific manner.
In conclusion, our results demonstrate the proteolytic processing and localization of zonadhesin from spermatogenesis to sperm-zona pellucida binding. In doing so we have shown that it is unlikely that zonadhesin performs a role in the initial recognition and binding of spermatozoa to the zona pellucida. Rather, zonadhesin binds the zona pellucida after the acrosome reaction has been initiated and may be one of the proteins that anchors the acrosomal shroud to the zona pellucida, thereby allowing the spermatozoa to continue penetration and fertilization to proceed.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Support was provided by a National Institutes of Health (NIH) grant and a postdoctoral fellowship through the Center for Recombinant Gamete Vaccinogens, University of Virginia, and the U.S. Public Health Service (U54HD29099) to I.A.L.; and by CONRAD and the NIH International Training and Research Program in Population and Health to P.S. 
2 Correspondence: Isabel A. Lea, Department of Cell and Developmental Biology, University of North Carolina at Chapel Hill, 210 Taylor Hall, CB 7090, Chapel Hill, NC 27599. FAX: 919 966 1856; ialea{at}email.unc.edu
Accepted: July 17, 2001.
Received: May 22, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/ß dimerization, and lack of essential role of fertilin in sperm-egg fusion. Dev Biol 2000; 222:289-295[CrossRef][Medline]
. J Biol Chem 1995; 270:14247-14250This article has been cited by other articles:
|
|
 |
|
  D. Busso, D. J Cohen, J. A Maldera, A. Dematteis, and P. S Cuasnicu A Novel Function for CRISP1 in Rodent Fertilization: Involvement in Sperm-Zona Pellucida Interaction Biol Reprod, November 1, 2007; 77(5): 848 - 854. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. van Gestel, I. A. Brewis, P. R. Ashton, J. F. Brouwers, and B. M. Gadella Multiple proteins present in purified porcine sperm apical plasma membranes interact with the zona pellucida of the oocyte Mol. Hum. Reprod., July 1, 2007; 13(7): 445 - 454. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. E. Olson, V. P. Winfrey, M. Bi, D. M. Hardy, and S. K. NagDas Zonadhesin Assembly into the Hamster Sperm Acrosomal Matrix Occurs by Distinct Targeting Strategies During Spermiogenesis and Maturation in the Epididymis Biol Reprod, October 1, 2004; 71(4): 1128 - 1134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. G. Robertson, J. Harris, M. J. Naylor, S. R. Oakes, J. Kindblom, K. Dillner, H. Wennbo, J. Tornell, P. A. Kelly, J. Green, et al. Prostate Development and Carcinogenesis in Prolactin Receptor Knockout Mice Endocrinology, July 1, 2003; 144(7): 3196 - 3205. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
