| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Male Reproductive Tract |
Gamete Signalling Laboratory, The Babraham Institute, Cambridge CB2 4AT, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
epididymis, gamete biology, male reproductive tract, sperm, sperm maturation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To address these questions we have investigated the properties of secreted epididymal antigens recognized by the monoclonal antibody (McAb) 2D6. In previous work it was shown that during the early stages of fertilization a 24-kDa antigen on cauda spermatozoa recognized by 2D6 McAb is transferred to, and mixes with, components of the oolemma [14, 15]. We have hypothesized that this glycoprotein may be involved in the initial stages of egg activation by forming complexes with egg components to initiate, for example, calcium oscillations in the cytoplasm [15]. Thus, glycoproteins of epididymal origin would have a direct effect on egg activation. The present results, however, show that the 24-kDa antigen on cauda spermatozoa is a member of the ß-defensin superfamily of small pore-forming peptides of which several are secreted specifically by the epididymis and have putative antibacterial functions [16–19].
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Unless otherwise stated, all routine chemicals were supplied by Sigma-Aldrich (Poole, UK) or Merck (Poole, UK). "Immobilon-P" PVDF membrane was supplied by Millipore (Watford, UK) and synthetic oligonucleotides by Sigma Genosys (Pampisford, UK). Culture supernatants containing McAbs to rat sperm membrane antigens 2D6 and 2B1, together with rabbit polyclonal antibodies (PAbs) to PBP and 2D6 antigens, were prepared and used as described previously [20, 21]. The 2D6 PAb was raised against the 24-kDa form of the protein by electroeluting the antigen from SDS gels. A guinea pig PAb to rat CD52 glycoprotein was a generous gift of Dr. Christianne Kirchhoff (Institute for Hormone and Fertility Research, Hamburg, Germany). FITC-conjugated rabbit anti-mouse IgG (FITC-RAM), peroxidase-conjugated rabbit anti-mouse IgG (Px-RAM), FITC-conjugated sheep anti-rabbit IgG (FITC-SAR), and peroxidase-conjugated sheep anti-rabbit IgG (Px-SAR) were purchased from DAKO Ltd (Ely, UK). Rats (Wistar strain) and mice (Balb/c strain) were supplied by the small-animal facility at the Babraham Institute and hamsters by Harlan U.K. Ltd. (East Sussex, UK). Ejaculated ram, bull, and boar semen were collected from animals maintained at the Institute using an artificial vagina. A cDNA library from whole rat epididymis was custom-made in pSPORT 1 vector by the Deutsches Ressourcenzentrum für Genomforschung GmbH (RZPD; Berlin, Germany). 4,2-Aminoethylbenzenesulfonyl-fluoride (AEBSF or "Pefabloc"; Melford Ltd, Ipswich, UK) was used as a protease inhibitor.
Immunocytochemical Analysis of Tissue Sections
Small pieces of epididymal tissue were snap frozen in iso-pentane at -80°C and embedded in Cryo-M-Bed (Bright) medium. Sections 8–12 µm thick were cut on a cryostat, dried onto clean glass slides, and stored at -20°C until analysis. Dried sections were immersed in PBS containing 3% (w/v) BSA at room temperature for at least 1 h to block nonspecific binding sites, washed 2 x 10 min in PBS, and probed with 2D6 McAb supernatant for 1–16 h at 4°C. Slides were washed 3 x 10 min in PBS followed by incubation with FITC-RAM diluted 1:100 in PBS/1% (w/v) BSA for 3 h in the dark at room temperature. After washing 3 x 10 min in PBS/0.1% (w/v), BSA sections were mounted in Citifluor antifade (Citifluor, London, UK) and viewed with UV light using a Zeiss Axiophot photomicroscope (Carl Zeiss, Oberkochen, Germany) fitted with epifluorescence illumination and a SPOT-RT camera (Diagnostic Instruments, Sterling Heights, MI).
Immuno-Gold Electron Microscopy
Pieces of rat epididymis from the proximal caput, distal caput, and cauda regions were immersed in 3% (w/v) paraformaldehyde in 0.1 M phosphate buffer containing 5% (w/v) sucrose for 2 h at room temperature. Tissues were then transferred to 0.2 M sodium cacodylate buffer (pH 7.2) and cryoprotected by immersion in 2 M sucrose containing 5% (w/v) polyvinylpyrrolidene overnight at 4°C. Frozen sections were cut at -20°C using a Reichert FC4 cryo-ultramicrotome and incubated in undiluted 2D6 McAb supernatant for 3–5 h at 4°C. Control sections were incubated in culture medium only. After thorough washing in PBS/0.1% (w/v) BSA, sections were incubated in rabbit anti-mouse IgG diluted 1:1000 for 30 min followed by colloidal gold (5 or 15 nm)-conjugated goat anti-rabbit IgG for 1 h. After further washing, sections were counterstained with 2% (w/v) uranyl acetate, embedded in 0.3% (w/v) methyl cellulose, and air dried. Sections were viewed using a JOEL 100C transmission electron microscope as described [22].
Collection of Luminal Contents from the Caput and Cauda Epididymidis
Luminal contents were collected from the distal caput and cauda epididymidis either by micropuncture or by slicing the whole tissue. In the latter procedure, the epididymis was cleaned of excess blood and superficial fluid by gently blotting onto Kleenex paper and immersed in 2 ml PBS containing 1 mM Pefabloc. Three to four incisions were made with fine scissors, and luminal fluid was allowed to exude from the cut ends of the duct. The suspension was gently agitated for 5 min, and the released spermatozoa + fluid was decanted away from tissue fragments and centrifuged at 1000 x g for 5 min. The resulting supernatant was clarified by recentrifugation at 10 000 x g for 5 min and the sperm-free caput or cauda plasma stored frozen at -20°C. The first sperm pellet was then washed once more in eight volumes of PBS/Pefabloc, and proteins were extracted into five volumes of 1% (v/v) NP40 in PBS/Pefabloc at 4°C for 45 min. Detergent extracts of spermatozoa were clarified by centrifugation at 14 000 x g for 10 min and stored frozen at -20°C.
Alternatively, luminal fluid was collected by micropuncture techniques. Although laborious, this method is superior to slicing the tissue, as there is less contamination from blood and lymph. All manipulations were carried out at room temperature using a 10x dissecting microscope. A small area of the surface of the epididymal duct was exposed by careful dissection with fine forceps, washed with PBS, and gently blotted with tissue paper. Glass micropuncture pipettes (25–50-µm tip diameter) filled with PBS/1 mM Pefabloc were inserted into the duct lumen at several superficial sites and the contents drawn out by mild vacuum aspiration. Spermatozoa and their surrounding plasma were then separated and extracted with detergent as described previously.
Quantitation of 2D6 McAb Binding to Spermatozoa by Fluorescence-Activated Cell Sorting (FACS)
The amount of 2D6 McAb bound to spermatozoa was quantified by FACS analysis using a Beckton Dickinson 2 laser FACScalibur, and the resulting data were analyzed on Cell Quest software (Beckton Dickinson, Franklin Lakes, NJ). Spermatozoa from the specified epididymal region were washed twice in PBS/1 mM Pefabloc as described previously and resuspended in PBS/0.1% (w/v) BSA. The final washed sperm pellet was then treated as follows: (A) Resuspension in 150 µl of 2D6 McAb supernatant plus 50 µl PBS/0.1% (w/v) BSA for 1 h with gentle rotation, washed twice as previously described, and incubated in 100 µl FITC-rabbit anti-mouse IgG (diluted 1:20 in PBS/0.1% [w/v] BSA). Sperm were washed twice more in PBS/0.1% (w/v) BSA and analyzed by FACS. (B) Resuspension in 500 µl dilute caudal plasma (CEP; protein concentration 1 mg/ml) for 60 min at room temperature, washed, and treated as described in (A). In some experiments, CEP was dialyzed against PBS/1 mM Pefabloc at 4°C for 48 h. Prior to FACS analysis, samples were diluted to 
5 x 104 sperm/ml in PBS/BSA, and the instrument was operated at a flow rate of 100–500 cells/sec. "Gating" was used to remove cell clumps and debris from the data analyzed. Gates were set using washed caudal sperm (no first layer antibody) as a control. This analysis ignored signals with very low forward and sideways scatter (predominantly debris) or high forward scatter (frequently due to clumped cells or aggregated debris).
Lymphocytes were prepared from two rat spleens minced in 10 ml 
MEM (Gibco, Paisley, UK) containing 10% (v/v) fetal calf serum (FCS; Gibco, UK). Tissue pieces were disaggregated in a Potter homogenizer (four to five strokes). The cell suspension was centrifuged at 1000 x g for 3 min and the supernatant removed and centrifuged at 1500 x g for 5 min. The pellet was resuspended in 8 ml MEM/FCS and divided into two equal aliquots each of which was layered onto 3 ml Ficoll-Paque (Pharmacia, Milton Keynes, UK). These were centrifuged at 1000 x g for 40 min, and the lymphocyte band (present at the interphase between the clear lower and cloudy upper layers) was collected and centrifuged at 1500 x g for 5 min. The resulting pellet, consisting of washed purified lymphocytes, was treated as described previously for spermatozoa.
Red blood cells were prepared by mixing 4 ml of whole blood with 55 µl of acid dextrose citrate solution (403.8 mM NaCl, 16 mM NaH2PO4, 161 mM glucose, and 15.6 mM citrate). The suspension of red cells was then washed and treated as described previously for spermatozoa.
Electrophoresis and Western Blotting
Proteins were separated by SDS-PAGE under nonreducing or reducing (by heating at 100°C for 5 min with 2 mM dithiothreitol [DTT]) conditions. Gels were stained with Coomassie blue R250 using standard procedures. Western blots were prepared from parallel gels, blocked with 4% (w/v) nonfat dry milk ("Marvel") in PBS for at least 3 h at room temperature, and probed with McAb supernatants or polyclonal antisera (diluted 1:1000) for 2 h. After washing three times in PBS, blots were probed with the appropriate Px-RAM or Px-SAM (diluted 1:5000) for 2 h. Bound antibody was detected on x-ray film by chemiluminescence procedures (NEN Life Sciences, Boston, MA).
Triton X-114 Partitioning, PI-PLC Digestion, and Effects of Dissociating Reagents on Binding of 2D6 Antigen to the Plasma Membrane of Cauda Spermatozoa
To determine the strength of association between the cauda form of 2D6 antigen and the plasma membrane, spermatozoa were exposed to Triton X-114, PI-PLC, and different dissociating reagents followed by an analysis of its partitioning between the insoluble (membrane bound) and soluble fractions. For this purpose, washed cauda sperm pellets were resuspended in 1% (v/v) Triton X-114 in PBS/1 mM Pefabloc, incubated at 4°C for 30 min, and centrifuged at 1400 x g for 10 min [23]. The supernatant was removed and overlaid on a 10% (w/v) sucrose cushion containing 0.06% Triton X-114 at 4°C. Samples were heated to 37°C for 20 min to induce phase separation and centrifuged at 750 x g for 10 min at room temperature. Aqueous and detergent phases were removed for SDS-PAGE analysis and Western blotting. 2B1 glycoprotein, which partitions strongly into the detergent phase [24], was used as an internal control.
To investigate the possibility that the form of the 2D6 antigen on cauda spermatozoa had a GPI anchor, washed sperm suspensions (100 µl) were incubated with 5 U/ml of phosphatidylinositol-specific phospholipase C (PI-PLC) for 2 h at 37°C. Samples were centrifuged and SDS extracts prepared of the supernatant and pellets for SDS-PAGE and Western blotting. CD52 glycoprotein, which contains a GPI anchor [25], was used as an internal control.
The effects of dissociating reagents on partitioning of cauda sperm 2D6 antigen between soluble and membrane-bound fractions was investigated by resuspending washed sperm pellets in (A) 1 M KCl, pH 7.2; (B) 4 M urea adjusted to pH 7.2; (C) PBS adjusted to pH 3.0 with HCl; (D) PBS adjusted to pH 10.0 with NaOH; (E) PBS, pH 7.2, containing 0.4% (v/v) ß-mercaptoethanol; (F) PBS containing 0.1 M EDTA pH 7.2; and (G) PBS, pH 7.2, containing 1% (v/v) NP40. Sperm suspensions were incubated at room temperature for 30 min and centrifuged at 10 000 x g for 20 min. Supernatants were removed and pellets reextracted with 1% SDS in PBS for 5 min at room temperature. Supernatants and pellet extracts were analyzed by SDS-PAGE/Western blotting and probed with 2D6, 2B1, and PBP antibodies.
Analysis of the Epitope Recognized by 2D6 McAb (Protein versus Carbohydrate)
The cauda form of 2D6 antigen is highly glycosylated [21], and since many monoclonal antibodies recognize specific sugar epitopes, a preliminary analysis was carried out to determine whether 2D6 monoclonal antibody was binding to protein or carbohydrate moieties. For this purpose a Western blot containing reduced proteins extracted from washed cauda sperm was incubated at 37°C for 16 h in 5 mM sodium phosphate, pH 7.6, containing 0.1% (v/v) Triton X-100, 1 mM Pefabloc, and 10 U/ml of N-glycanase to cleave N-linked carbohydrate chains [26]. A parallel control was incubated under the same conditions but without the enzyme. Blots were washed in several changes of PBS followed by probing with either 2D6 McAb or 2D6 polyclonal antibody as described previously.
O-linked carbohydrates were released by alkaline hydrolysis as described by Duk et al. [27] but with the omission of the preliminary acid step, as it appeared to cause degradation of the antigen. Western blots containing reduced cauda sperm proteins, previously separated by SDS-PAGE, were immersed in 55 mM NaOH, pH 12.0, at 4°C for various times up to 1 h, washed three times in PBS, and probed with 2D6 McAb and 2D6 polyclonal antibodies as described previously. As a positive control, parallel blots were probed with 2B1 McAb, which recognizes a peptide-based epitope [28].
Purification and Peptide Sequence Analysis of 2D6-Reactive Antigen from Cauda Spermatozoa
Washed caudal sperm were resuspended in PBS/1 mM PefaBloc containing 1 M KCl and incubated at room temperature for 20 min to release non-covalently-bound proteins from the plasma membrane. Spermatozoa were washed twice more in PBS and extracted with 1% NP-40 as described previously. Proteins were separated by SDS-PAGE in the first dimension under nonreducing conditions and in the second dimension under reducing conditions. Gels were either stained directly with Gelcode Blue Reagent (Pierce Chemical Co., Rockford, IL) or Western blotted and stained with Coomassie blue. Protein bands at 24 kDa were outlined with pencil and blots destained and probed with 2D6 McAb using 4-chloronaphthol as dye reagent [29]. This procedure enabled the reduced 24-kDa form of 2D6 antigen to be located on the blot with high precision. Subsequently, its position on Coomassie blue-stained gels and blots could be ascertained with accuracy.
Fifteen pieces of PVDF membrane containing the 24-kDa form of 2D6 antigen were pooled and subjected to N-terminal sequence analysis by automatic Edman degradation on an Applied Biosystems 470A gas-phase sequenator equipped with a 120A on-line phenylthiohydantoin analyzer. For internal peptide sequencing, gel slices containing the 24-kDa form of 2D6 antigen were lightly stained with 15–20% Gelcode and trypsinized in situ, and released peptides were analyzed by tandem mass spectroscopy (MALDI-TOF MS/MS) (M-Scan Ltd, Wokingham, UK). Two peptides were successfully sequenced, and the analysis was repeated three times.
Cloning and DNA Sequence Analysis of the 24-kDa Antigen
To obtain the full cDNA sequence of the 24-kDa antigen recognized by 2D6 McAb, the rat epididymal cDNA library was screened with a 32P-labeled degenerate oligonucleotide probe 5' CARGGIGARTAYCARACIGAYCCIGCIACIGGIAA 3' (Y = C or T, R = G or A, I = inosine) encoding a peptide sequence derived from the purified 24-kDa antigen. Bacteria containing inserts that hybridized to the probe were grown in LB ampicillin and plasmids isolated using a Qiagen miniprep kit. The plasmids were then restricted using BamH1 and EcoR1 as directed by the manufacturers (Promega, Southampton, UK) and those with the longest inserts selected for further analysis. DNA sequencing was performed using an ABI automatic sequencer.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The epithelium lining the efferent ductules, initial segment, and proximal caput epididymidis did not stain with 2D6 McAb/FITC-RAM (Fig. 1A). However, a strong signal was detected in the microvilli and apical cytoplasm of principal cells in the distal caput epididymidis and thereafter in the epithelium throughout the corpus and cauda epididymidis (Fig. 1, B and C). Only background fluorescence was present in basal cells and peritubular connective tissue. A noteworthy feature was the abrupt transition between the negatively staining epithelium in the proximal caput region and the positively staining epithelium in the adjoining distal caput. Despite the strong reaction in the epithelium, only a weak reaction was observed in the duct lumen in the distal caput (Fig. 1B). The strength of the fluorescent signal associated with spermatozoa increased rapidly in a distal direction so that a very intense reaction was present in the lumen of the distal corpus and cauda epididymidis (Fig. 1C).
|
At the ultrastructural level, gold-conjugated RAM IgG-2D6 McAb complexes were associated with stacked arrays of membranes in the Golgi region of principal cells throughout the distal caput epididymidis (Fig. 1D). Gold particles were also associated with small membrane-bound vesicles in the apical cytoplasm that were either empty or weakly electron dense (Fig. 1, D and E). Large empty vesicles adjacent to the Golgi were generally unlabeled. The surfaces of microvilli protruding from the apical surface of principal cells were encrusted with gold particles that were also present in the duct lumen, where they had a random distribution (Fig. 1F). Membrane-bound vesicles could not be positively identified in the lumen and gold particles in patterns consistent with the presence of such vesicles were not apparent. In the corpus and cauda epididymidis, the plasma membranes of spermatozoa were heavily labeled with gold particles, especially on cross sections of the flagellum (Fig. 1H).
Binding of 2D6 McAb to Spermatozoa from Different Levels of the Epididymis
There was no detectable binding of 2D6 McAb to immature spermatozoa from the proximal caput epididymidis. However, the proportion of spermatozoa that were 2D6 positive increased progressively in a distal direction from the distal caput to the cauda epididymidis (Fig. 2). Although a single peak of labeled cauda spermatozoa was present in the example shown in Figure 2, a bimodal distribution was frequently obtained, indicating heterogeneity in the strength of signal. Heterogeneity was more apparent in sperm samples collected from the distal caput and proximal corpus epididymidis, presumably because a certain amount of time is necessary to expose all spermatozoa to newly secreted glycoproteins.
|
Identification of Secreted and Membrane-Bound Forms of 2D6 Antigens in Luminal Plasma and Spermatozoa from the Caput and Cauda Epididymidis
Proteins in epididymal plasma and detergent extracts of spermatozoa collected by micropuncture were analyzed under nonreducing and reducing conditions (Fig. 3). Under nonreducing conditions, 2D6 McAb recognized a high-molecular-mass protein of 250 kDa in caput plasma. In cauda plasma a strong reaction was obtained at 200 kDa, and weaker reactions were obtained with two lower-molecular-mass components at 160 and 48 kDa. In detergent extracts of caput spermatozoa a very weak signal was present only at 250 kDa. This protein, and two additional ones at 48 and 24 kDa, was present in extracts of cauda spermatozoa.
|
Under reducing conditions a significantly different protein pattern was obtained to that described above. In caput plasma 2D6 McAb recognized proteins at 200 kDa together with two additional components at approximately 60 and 70 kDa. In extracts of caput sperm a signal was obtained only over a protein with a molecular mass of 65 kDa. In contrast, 2D6 McAb bound strongly to several proteins in cauda plasma and cauda sperm extracts. In cauda plasma, proteins of 32, 28, 24, and 20 kDa gave the strongest reaction, whereas in the sperm extracts only the 24-kDa component bound 2D6 McAb.
The pronounced effects of disulfide bond cleavage on the size of the proteins recognized by 2D6 McAb suggests that they may be secreted as a multimeric complex and are progressively "processed" by reduction and endoproteolysis as they pass through the distal regions of the duct. Further evidence to support this interpretation is shown by Western blots taken from two-dimensional nonreducing/reducing SDS gels of cauda sperm proteins (Fig. 4, A and B). Before reduction the major protein component identified by 2D6 McAb had a molecular mass of 48 kDa, whereas after reduction it was displaced from the diagonal and migrated at 24 kDa. In addition, when the reduced form of 2D6 antigen (24 kDa) from cauda spermatozoa was electroeluted from SDS gels, incubated at 4°C for 2 days in 40 mM Tris-glycine buffer (pH 8.2), and reanalyzed ±2 mM DTT, it migrated at 250 kDa (Fig. 4C). Intermediate-size forms were not detected. This suggests that the 24-kDa antigen has strong tendencies to self-associate.
|
In the course of this work it was found that the method of collection of CEP had a significant influence of its composition. When luminal fluid was collected from the cauda epididymidis by making three to four small incisions into the tissue with iridectomy scissors, nonreducing SDS-PAGE analysis of proteins in the CEP revealed two strongly staining proteins at 160 and 48 kDa that were only weakly present in CEP collected by micropuncture (results not shown). In addition, a diffusely staining protein at 22 kDa was detected in CEP from minced tissue that was absent from micropuncture samples. This suggests that caution should be exercised in collecting CEP by uncontrolled mincing of the tissue.
Binding of 2D6 Antigen to the Sperm Plasma Membrane
Secreted epididymal glycoproteins are known to bind to spermatozoa with varying tenacities. To elucidate the strength and nature of the binding of the 24-kDa form of 2D6 antigen to the sperm plasma membrane, cauda epididymidal spermatozoa were washed in PBS followed by resuspension in different dissociating agents, at extremes of pH or in detergents. Plasma membrane proteins were then solubilized with 1% NP-40, and the amount of 2D6 antigen remaining was assessed by SDS-PAGE/Western blotting. As an internal control, parallel blots were probed with a polyclonal antibody to PBP antigen (also known as PEPB or raf kinase inhibitory protein [30, 31]). PBP is easily released from spermatozoa under mild dissociating conditions and can be quantitatively recovered in the supernatant [31]. As shown in Figure 5, incubation of cauda spermatozoa in 1 M KCl or 4 M urea or 0.4% ß-mercaptoethanol or PBS (pH 3.0) or PBS (pH 10.0) or 0.1 M EDTA did not solubilize detectable amounts of 2D6 antigen, whereas >95% of PBP was removed with 1 M KCl, 4 M urea, PBS (pH 10.0), and 0.1 M EDTA. These observations suggest that 2D6 antigen is bound strongly to the sperm plasma membrane, possibly by hydrophobic interactions.
|
Two methods widely used for assessing protein-lipid interactions are partitioning of the antigen into Triton X-114 and its solubilization by PI-specific PLC (PI-PLC). The latter procedure indicates the presence of a GPI anchor, although not all GPI-anchored proteins on cell surfaces are susceptible to PI-PLC cleavage owing to problems of steric hindrance. Following solubilization and partitioning of cauda sperm plasma membrane proteins with Triton X-114, the distribution of 2D6 antigen between the aqueous and detergent phases was analyzed by SDS-PAGE/Western blotting. 2B1 glycoprotein was used as an internal control, as it has been shown to contain a GPI anchor and to partition strongly into the detergent phase [24]. As shown in Figure 6, the 24-kDa subunit form of 2D6 antigen is present in approximately equal proportions in the aqueous and detergent layers, whereas >90% of 2B1 glycoprotein is found in the Triton X114 layer.
|
Digestion of washed cauda spermatozoa with PI-PLC failed to release detectable quantities of 2D6 into the medium (Fig. 7). In these experiments we used CD52 antigen as the internal control (although 2B1 glycoprotein has a GPI anchor, it is resistant to cleavage by PI-PLC, whereas CD52 antigen is released in significant amounts by the enzyme [13]). When parallel blots were probed with a polyclonal antibody to rat CD52, a positive reaction was obtained in the supernatant from PI-PLC treated sperm. Preextraction of cauda spermatozoa with 1 M KCl to remove loosely bound glycoproteins from the plasma membrane did not make 2D6 antigen susceptible to cleavage by PI-PLC (results not shown).
|
Preliminary Identification of the Epitope Recognized by 2D6 McAb: Protein versus Carbohydrate
Previous work demonstrated the presence of significant amounts of carbohydrate associated with the 24-kDa form of 2D6 antigen [21]. Since sugar structures are well known to be highly antigenic, it was necessary to obtain information regarding the nature of the epitope recognized by the McAb. For this purpose, Western blots containing cauda sperm proteins were incubated with N-glycanase or subjected to alkaline hydrolysis followed by probing with 2D6 McAb. N-glycanase digestion, which removes N-linked carbohydrates from glycoproteins, had no significant effect on the binding of 2D6 McAb (Fig. 8). Alkaline hydrolysis, however, eliminated it completely (Fig. 9). The continued presence of the 24-kDa antigen on blots after alkaline hydrolysis was demonstrated using a polyclonal antibody prepared to the 24-kDa antigen excised from the gel (Fig. 9). As an additional control, parallel blots were probed with 2B1 McAb, which has been shown to recognize a peptide epitope [28]. Alkali treatment had no detectable effect on the ability of 2B1 McAb to recognize its antigen (Fig. 9). These results suggest that the 24-kDa antigen contains significant amounts of O-linked carbohydrates and that it is these sugar chains that constitute all, or part of, the epitope site recognized by 2D6 McAb.
|
|
Species Specificity of the 2D6 Carbohydrate Epitope
Western blots containing reduced NP-40-solubilized proteins from mouse, hamster, bull, boar, and ram spermatozoa were probed with 2D6 McAb to determine the species specificity of the O-linked carbohydrate epitope. As shown in Figure 10, a positive reaction was obtained only with rat sperm proteins.
|
Uptake of 2D6-Reactive Antigens from CEP: Cell and Species Specificity of Binding
To investigate the specificity of binding of secreted 2D6 antigens to the sperm plasma membrane and the conditions that regulate it, immature spermatozoa from the proximal caput epididymidis were resuspended in CEP(1 mg/ml) diluted in PBS, pH 7.4, and incubated in vitro. As shown in Figure 11A, a significant amount of 2D6-reactive antigens had bound to spermatozoa after 60 min and were localized to the tail (Fig. 11B). Prior dialysis of CEP against PBS had no significant effect on binding (Fig. 11A), nor had the presence of 20 mM DTT or 0.5 M KCl (results not shown). Uptake was also cell-type specific, as neither erythrocytes nor lymphocytes stained with 2D6 McAb after incubation in CEP (Fig. 12, A–D). However, when mouse proximal caput spermatozoa were incubated with rat CEP, they became 2D6 reactive (Fig. 12, E–H). Similar results were obtained with hamster caput spermatozoa (results not shown). In both species the 2D6-reactive antigens localized to the flagellum and had a patchy appearance similar to that observed on rat cauda spermatozoa. Note that in Figure 12F, a small proportion (20–30%) of mouse caput spermatozoa stained on the head region. This antigen appears to be intra-acrosomal and disappears during epididymal maturation. Hence, there is no reaction on the Western blot shown in Figure 10, which represents cauda sperm.
|
|
Sequence Analysis of the 24-kDa Antigen from Cauda Spermatozoa
To obtain clues to the functionality of the 24-kDa form of 2D6 antigen, it was subjected to a primary sequence analysis followed by database searches to identify potential orthologues in other tissues. For this purpose, the reduced 24-kDa subunit was excised from nonreducing/reducing two-dimensional PAGE gels and trypsinized in situ, and released peptides were subjected to mass spectrometry (MS/Q-TOF) analysis. Trypsinization consistently produced one major peptide with a charge:mass ratio of 648 m/z. Since the protein is known to be heavily glycosylated, this suggests that a restricted number of arginine/lysine cleavage sites are available to the protease, probably because of steric hindrance. Repeated (4x) MS/MS analysis of the major peptide yielded the amino acid sequence Q/KGEYQ/KTDPATGK. On one occasion an additional small peptide was obtained that sequenced as I/LI/LDI/LK.
A synthetic degenerate oligonucleotide probe based on the major peptide sequence was used to screen a filter-based cDNA library prepared from whole rat epididymis in pSPORT1 vector. Twenty-seven independent clones were selected that hybridized to the oligonucleotide probe. A single clone containing a plasmid with the largest insert was chosen for further analysis. The insert encoded a 653 nucleotide mRNA, representing a 111-amino-acid open reading frame (Fig. 13A), that included the two polypeptide sequences determined by the mass spectrometry analysis (amino acids 40–51 and 60–64 as highlighted in bold text). A Kyte Doolittle analysis of the amino acid sequence (Fig. 13B) indicated that the protein had a hydrophobic N-terminus suggestive of a signal peptide. Further analysis of this region using the program SignalP [32] supported this conclusion and predicted the signal peptide to be 20 amino acids in length (underlined in Fig. 13A). The presence of a putative signal peptide is consistent with its being a secreted protein [33]. The hydrophobicity plot also indicates the presence of a further hydrophobic region on either side of residue 60. As this region of hydrophobicity is too short to represent a transmembrane domain, it likely to be buried in the middle of a globular region of the protein.
|
Comparison of the protein sequence of the 24-kDa subunit with other proteins in the NCBI databases (Fig. 14) using the computer program BLAST [34] revealed 100% identity with rat epididymal protein E3 recently described by Rao et al. [35]. Significant similarity was also found to epididymal protein ESP13.2, which has been cloned from both human and the macaque monkey [36]. The sequence similarity between the rat 24-kDa protein and macaque ESP13.2 is not sufficiently high for the two proteins to be homologues. All three proteins, however, show weak similarity to other members of the ß-defensin family (Fig. 14), especially in the central core region, which retains a group of six conserved cysteine residues. Particularly noteworthy among these sequences is that of Bin1b, which has been cloned from rat epididymis and reported to have antimicrobial activity on the basis of an anti-sense RNA assay [17]. ß-Defensins are generally small proteins (approximately 60 amino acids, including the signal peptide), whereas 2D6 is 111 amino acids long. On the basis of this evidence, therefore, we tentatively assign the 24-kDa form of 2D6 antigen to a protein superfamily that includes the ß-defensins.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Morphological Evidence for Secretion of Epididymal Glycoproteins
The current light-microscopic and ultrastructural investigations of 2D6-reactive antigens in the epithelium of the epididymis have shed more light on the mechanism of glycoprotein secretion. 2D6 immunoreactivity first appears in the apical cytoplasm and microvilli of principal cells in the distal caput epididymidis, a noticeable feature being the abrupt transition between this region and the preceding proximal caput and initial segment. 2D6 immunoreactivity is also present in the epithelium throughout the distal regions of the duct, although it is somewhat obscured in the cauda epididymidis because of the very strong reaction in the lumen. At the ultrastructural level, gold-labeled 2D6 antigen is associated with the Golgi cisternae and small membrane-bound vesicles found in the apical cytoplasm. These vesicles are frequently confluent with the surface membrane, and their contents are of a similar electron density to the glycocalyx on the microvilli, which label heavily with gold particles. This evidence suggests that 2D6 antigen is secreted in a relatively dilute and soluble form by classic exocytotic pathways and is then concentrated in the luminal fluid by selective transport of water in the reverse direction. Gold particles in the lumen are usually associated with small electron-dense aggregates, and there is no indication that they are contained within or attached to membrane-bound vesicles (epididymosomes), which are thought to be formed by apocrine secretion. The latter process is characterized by the presence of apical blebs several microns in length protruding into the lumen from the surface of principal cells, and even after 30 years of detailed morphological studies on the epididymis, they remain a contentious phenomenon. A long established view is that apical blebs are artifacts of the fixation and embedding procedures commonly used in electron microscopy [37]. Recently, Hermo and Jacks [9] have argued that apical blebs in the epididymis are not fixation artifacts but are manifestations of genuine apocrine secretion and are the probable origins of epididymosomes that have been described in the lumen of the rat and bull epididymis [8, 9]. This long-standing and vexatious problem remains to be resolved. Although the present ultrastructural analysis was limited in scope, apical blebbing was not observed in any electron micrographs of the distal caput epithelium.
Characterization of Secreted and Membrane-Bound Forms of 2D6 antigen
The demonstration that 2D6 McAb recognizes an O-linked carbohydrate epitope has important implications for identification of its cognate antigens in luminal fluid and on spermatozoa. One interpretation of the multiple bands on Western blots shown in Figure 3 is that 2D6 antigen is secreted in the distal caput epididymidis as a high-molecular-mass complex (250 kDa) that is "processed" in a variety of ways (e.g., by endoproteolytic cleavage or deglycosylation) to produce several lower-molecular-weight components, some of which are found in the luminal fluid and some on the sperm surface. One of these processed proteins (the 24-kDa component, which is a subunit of a cross-linked homodimer of 48 kDa) is the major 2D6-reactive antigen on cauda sperm. The remaining proteins that were present initially on distal caput sperm are either released or processed further so that the carbohydrate epitope is no longer recognized by 2D6 McAb. Support for this interpretation is provided by the finding that the purified 24-kDa component from spermatozoa self-associates to form a high-molecular-mass aggregate (250 kDa) when incubated for 48 h at 4°C. This tendency for aggregation could conceivably produce a profile of related glycoproteins of different molecular weights.
An alternative hypothesis is that the 2D6 carbohydrate epitope is generic to several different glycoproteins that are secreted in different regions of the epididymis and that change their reactivity to the McAb during epididymal maturation. Thus, initially a 250-kDa component is secreted in the distal caput epididymidis that binds to spermatozoa in the lumen of the duct. In the corpus and cauda epididymidis, however, this component is removed or the epitope modified, and instead 2D6 McAb now recognizes several unrelated proteins secreted in the corpus and cauda regions, only one of which (the 48-kDa component) interacts with the sperm plasma membrane. Implicit in this interpretation is that although 2D6 McAb recognizes several different antigens, they all bind to the same area of the sperm plasma membrane (i.e., the tail). Several precedents are known of endoproteolytic cleavage and ectodomain shedding of peptides from surface membrane antigens in the epididymis, most notably fertilin [38] and CE9 [39]. The fate of the peptides released is not known, nor is it known if cleavage is important for activating the portion of the protein remaining attached to the membrane (as found with protease-activated receptors for the G-protein-coupled family [40, 41]).
On the basis of the evidence presented here we cannot distinguish between these two hypotheses. However, during preparation of this work for publication we became aware of a sequence in the Genbank database describing a rat sperm isoantigen (designated E-3) that was identical to that for the 24-kDa antigen. A subsequent publication describing the cloning and expression of this isoantigen by Northern blot analysis revealed that E-3 transcripts are tissue specific and detectable only in the corpus and cauda epithelium [35]. This evidence clearly favors the common epitope hypothesis, although processing of the 2D6 carbohydrate antigen still has to be invoked to explain why the high-molecular-mass components observed in fluid and on spermatozoa in the distal caput lose their reactivity to the McAb in the cauda region. The epididymis is well known to be a rich source of exoglycosidases [4], and it has long been speculated that they function to remove oligosaccharides from sperm surface glycoproteins [42]. If this interpretation is correct, then the current work represents strong evidence for carbohydrate epitope processing during sperm maturation in the epididymis. Equally intriguing is that 2D6 reactivity is species specific, hinting that in the rat epididymal glycoproteins are preferentially O-glycosylated rather than N-glycosylated.
Binding of 2D6-Reactive Antigens to the Sperm Plasma Membrane
In nearly all examples of secreted epididymal glycoproteins binding to the sperm surface, only qualitative information is available, whereas in the present work we have used FACS analysis to quantify the interaction between 2D6-reactive antigens and the sperm's plasma membrane.
The progressive increase in the proportion of spermatozoa that stain positively with 2D6 McAb/FITC-RAM as they pass through the distal caput and corpus epididymidis indicates uptake of the antigen from the fluid onto the sperm plasma membrane, predominantly onto the tail and to a lesser extent the postacrosomal region. The considerable heterogeneity in the strength of the fluorescence signal on spermatozoa from the distal caput and proximal corpus regions reflects different amounts of 2D6 on the sperm surface. As spermatozoa move into the cauda they become more homogeneous in this respect, although there was significant variation between animals, and it was not uncommon for two peaks of labeled cells to be present. Given the high viscosity of rat epididymal luminal fluid, it is not unexpected that efficient mixing of spermatozoa with newly secreted protein would be time dependent and facilitated by peristaltic contractions of the smooth muscle surrounding the duct.
Direct evidence for uptake of 2D6-reactive antigens from epididymal secretions was obtained by incubating proximal caput spermatozoa (which do not bind 2D6 McAb) in vitro with caudal fluid. Since neither rat erythrocytes nor lymphocytes stain positively with 2D6 McAb after incubation in caudal plasma, it suggests that uptake is cell-type specific and mediated either by a specific receptor or by unusual properties of the plasma membrane of spermatozoa. Both these possibilities could also account for the fact that uptake of 2D6 in vitro is not species specific since mouse (and hamster, results not shown) spermatozoa absorb significant amounts of the antigen. It is of interest that regardless of the species, uptake of 2D6 antigens in vitro mimics that found in vivo as they bind to the tail plasma membrane with a characteristic patchy appearance. Uptake in vitro is unaffected by prior dialysis of the CEP or by the presence of high salt or reducing reagents, suggesting that the interaction with the plasma membrane is primarily hydrophobic. This conclusion is further supported by the finding that only detergents are effective in solubilizing the 24-kDa antigen from mature cauda spermatozoa, and in phase separation experiments, 50% of it partitions into Triton X114. Both these properties are characteristic of hydrophobic interactions. If the antigen had bound covalently to an integral membrane receptor, then it would have migrated with a higher molecular mass after SDS-PAGE. That a significant proportion of 24-kDa antigen remains in the aqueous layer after detergent-phase separation suggests significant structural heterogeneity, conceivably mediated by differences in the level of glycosylation, that could influence the conformational changes necessary to interact with the plasma membrane.
The failure of PI-PLC to release the 24-kDa antigen from cauda spermatozoa suggests absence of a GPI anchor. Although several sperm proteins with known GPI anchors are PI-PLC resistant (e.g., rat 2B1), the characteristic motif for a lipid attachment site is not present in the primary sequence of 2D6 glycoprotein. The mechanism of binding of the 24-kDa antigen to the sperm plasma membrane, therefore, must involve either a covalent attachment to a preexisting membrane receptor or a protein-lipid interaction that is specific for the sperm tail. The credibility of the latter possibility is enhanced by recent photobleaching experiments that have shown that lipid diffusion in the tail plasma membrane is 5 times slower than on the sperm head and has a higher proportion of immobile-phase lipid [43]. Taken together, it would seem that some special properties of the sperm plasma membrane determine binding of epididymal glycoproteins rather than the glycoproteins in question.
Sequence Analysis of 2D6 Antigen and Preliminary Assignment of Functionality
As mentioned earlier, a sequence analysis of the 24-kDa subunit purified from cauda spermatozoa revealed that it is identical to isoantigen E-2 [35] and that it belongs to a family of small epididymal peptides, 68 to 123 residues long, that show significant sequence similarity to somatic ß-defensins. The defensins are a family of cationic antimicrobial peptides with a broad spectrum of activity, which, together with several other families of antimicrobial peptides, such as the cathelicidins, lactoferrin, and lysozyme, make up an important component of the innate immune system [44–46]. Two principal families of defensins have been described: -defensins, which are expressed by neutrophils, and ß-defensins, which are produced by epithelial cells. Both types of defensins possess antimicrobial activity and seem to play a role in the bridging of the responses of the innate and adaptive immune systems by activating receptors on dendritic and T cells [46]. They contain a central conserved cysteine-rich core domain but differ in the pattern of disulfide bonding within that region. Genome analysis has suggested that there are many more defensin-like proteins present in mammals than previously suspected. In human, 31 genes have been identified that encode proteins with defensin-like motifs [47].
In addition to 24 kDa/E-2, several other ß-defensin-like molecules appear to be specific to the epididymis. These include Bin-1b in the rat [17] and ESC42, HE2, and EP2 in primates [16, 18, 48]. Of these, HE2 and Bin-1b have been shown to possess antimicrobial action in vitro. Interestingly, although ESC42 and ESP13.2 possess an N-terminal defensin-like domain, they share with the 24 kDa/E-2 a C-terminal domain that is absent from any true ß-defensin, suggesting that they may form a small subfamily. The role of this C-terminal domain is unclear, although it is noteworthy that ESC42 has been shown to bind to the surface of maturing sperm [16], whereas neither Bin-1B nor ESP13.2 has been reliably detected on the sperm surface. Thus, it is possible that the C-terminal domain has a role in the interaction of the protein with the sperm plasma membrane.
It remains to be determined why those epididymal defensin-like molecules that bind to the sperm surface do not appear to have antimicrobial activity and vice versa. An intriguing speculation is that the latter group protects the vas deferens and epididymis from bacterial infections, whereas the former group may have a role during fertilization (e.g., initiating cortical granule release) as a result of activation through intermixing with components on the oolema.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Roy Jones, Gamete Signalling Laboratory, The Babraham Institute, Cambridge CB2 4AT, UK. FAX: +44 (0)1223 496022; roy.jones{at}bbsrc.ac.uk
3 Current address: Department of Cell Biology, University of Virginia, PO Box 800732, Charlottesville, VA 22908
Received: 28 April 2003.
First decision: 4 June 2003.
Accepted: 28 July 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
