| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Cellular and Physiological Sciences,4 Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z3
Department of Biological Sciences,5 Faculty of Science, Simon Fraser University, Burnaby, British Columbia, Canada V5A 1S6
Centre for Molecular Medicine and Therapeutics,6 University of British Columbia, Vancouver, British Columbia, Canada V5Z 4H4
ABSTRACT
During spermatogenesis, spermatids undergo a "down and up" translocation event in the seminiferous epithelium. This event has been proposed to result from the movement of ectoplasmic specializations, which are formed in Sertoli cells at sites of adhesion to spermatids, along adjacent microtubule tracts. To test the hypothesis that a kinesin is associated with ectoplasmic specializations, we generated antibodies to conserved kinesin sequences and detected kinesins on fixed frozen testis sections and fixed seminiferous epithelial fragments. The antibodies reacted with ectoplasmic specializations related to spermatids, in addition to reacting with other structures in the epithelium known to contain kinesins. At the electron microscopy level, the antibodies reacted with the cytoplasmic face of the endoplasmic reticulum component of ectoplasmic specializations. Based on mRNA transcript screens using mouse GeneChip arrays of testis and Sertoli cells, we identified KIF20 as a candidate kinesin at ectoplasmic specializations. Antibodies generated against a peptide sequence unique to this kinesin reacted at ectoplasmic specializations in testis sections and epithelial fragments, as well as with the endoplasmic reticulum component of ectoplasmic specializations when analyzed by electron microscopy. The antibody reacted on Western blots with full-length KIF20. On Western blots of testis lysates, the antibody reacted with a protein that is not present in other tissues and which migrates at a higher molecular weight than that predicted for KIF20. Our results demonstrate that a kinesin is associated with apical ectoplasmic specializations in Sertoli cells and that the motor may be an isoform of KIF20.
adherens junctions, ectoplasmic specializations, kinesin, Sertoli cells, spermatid translocation, testis
During spermatogenesis, developing spermatids are moved from apical to basal regions of the seminiferous epithelium and then are returned to apical regions, where they eventually are released as spermatozoa into the duct system [1–4]. This "down and up" translocation event results from changes in the depth of apical Sertoli cell invaginations, or crypts, that contain the developing spermatids (Fig. 1). When crypts initially form, they are shallow and, together with the attached elongate-shaped spermatids, occur in apical regions of the epithelium. As spermatogenesis proceeds, the crypts deepen (most dramatically at stages IV–V in the rat), and the attached spermatids, specifically their heads, become located at the epithelial base. Later, the crypts again become shallow, resulting in spermatids being repositioned at the apex of the epithelium.
In regions of the crypts that are attached to spermatid heads, Sertoli cells form complex intercellular adhesion structures termed ectoplasmic specializations that are predicted to be part of the spermatid translocation machinery. Ectoplasmic specializations are tripartite structures consisting of the Sertoli cell plasma membrane, a layer of actin filaments and an attached cistern of endoplasmic reticulum. These three components function as a structural unit that is firmly anchored to the adjacent spermatid head [5]. Although the biologic significance of spermatid translocation is unknown, the mechanism responsible for the process, at least in mammals, is thought to be the transport of ectoplasmic specializations, and therefore the attached spermatids, along adjacent Sertoli cell microtubule tracts (Fig. 1) [6, 7].
|
Because microtubules in Sertoli cells are oriented with their plus ends at the base of the cell and their minus ends at the apex [8], and because spermatids are moved first basally and then apically in the epithelium, we have predicted that if this motility event involves the microtubule-based translocation of developing spermatids, then there should be two types of microtubule-associated motors present at ectoplasmic specializations: one motor, likely a kinesin, to transport the junction/spermatid complexes basally in the epithelium, and another, likely a cytoplasmic dynein, to return spermatids to the apex of the epithelium prior to sperm release. Ectoplasmic specializations that are attached to spermatid heads mechanically detached from the epithelium support microtubule transport in both directions [9]—a result consistent with the prediction that both plus end- and minus end-directed, microtubule-based motor proteins are associated with ectoplasmic specializations. Cytoplasmic dynein is concentrated around apical crypts in the seminiferous epithelium [10] and is associated with the cytoplasmic face of the endoplasmic reticulum component of ectoplasmic specializations [9].
Here we report that antibodies generated against conserved peptide sequences common to most kinesins react with ectoplasmic specializations as well as with other sites in the seminiferous epithelium known to contain kinesins. Mass spectrometry of testis fractions enriched for ectoplasmic specializations and screens of mouse testis and Sertoli cell microarrays indicate that KIF1 kinesins and KIF20 are possible candidate motors. Moreover, antibodies raised against a peptide sequence specific to KIF20 react with the endoplasmic reticulum component of ectoplasmic specializations and with a band on immunoblots of testis and seminiferous epithelium lysates that migrates to a position that is somewhat larger than that predicted for conventional KIF20. Our results are consistent with the prediction that a kinesin is present on ectoplasmic specializations and with the possibility that the motor may be an isoform of KIF20.
Pan-Kinesin Antibody Production
Our strategy to determine whether kinesins were associated with apical ectoplasmic specializations was to generate antibodies to peptide sequences generally conserved among known kinesins and determine whether the antibodies reacted with Sertoli/spermatid junction sites. This approach has been successfully used to study kinesins by immunofluorescence and immunoblotting in other systems [11–13]. Three of eight sequence motifs that are generally conserved in the motor domain of most known kinesins include "HIPYR," "LAGSE," and "FAYGQ" [14].
Peptides (CHIPYRESKLT, LNLVDLAGSE, and GYNTIFAYGQTG) were synthesized commercially by ResGen (Huntsville, AL), and antibodies against each peptide were produced in White Leghorn laying hens by QED Biosciences (San Diego, CA). Eggs were collected for 2 wk prior to injection for isolation of control IgY for each antibody. Peptides were coupled to keyhole limpet hemocyanin (KLH) mixed with Freund Complete Adjuvant, and each peptide was injected into three hens (500 µg peptide per animal). Hens were given four "boost" injections—one injection per week—consisting of 500 µg peptide mixed with Freund Incomplete Adjuvant. Antibodies were affinity purified against their respective peptides and then used on immunoblots and screened on tissue for reactivity at apical ectoplasmic specializations.
Identification of Candidate Kinesins
To identify potential candidate kinesins associated with apical ectoplasmic specializations, we used two strategies.
The first was to use mass spectrometry (2DLCMS) of supernatants from gelsolin-treated rat spermatid/ectoplasmic specializations to identify any kinesins present. In this approach, testicular fractions enriched for spermatids with attached ectoplasmic specializations were treated with gelsolin to fragment the actin layer of the junction plaques and release the endoplasmic reticulum with attached motors into solution [10]. The spermatids then were removed by centrifugation and the supernatants analyzed by mass spectrometry (University of Victoria mass spectrometry facility). Using this approach, we identified two peptide fragments (TVAATNMNETSSR and ANSTGATGARLK) conserved in KIF1 kinesins. These include motors that are large (around 200 kDa), monomeric, single-headed kinesins involved in vesicular trafficking in neurons as well as in mitochondrial transport and vesicle trafficking between Golgi and endoplasmic reticulum generally in cells [15]. Although we attempted to generate antibodies to the sequences identified by mass spectrometry, we were unsuccessful and did not pursue the KIF1 motors further.
The second approach was to screen mouse testis developmental and mouse Sertoli cell GeneChip array databases for kinesin transcripts. The databases were made available to us by Michael Griswold at Washington State University [16]. The testis developmental databases were screened for kinesins whose transcripts increased at Developmental Days 30–35 compared with Day 20. Days 30–35 are the approximate times when spermatid entrenchment occurs during the first wave of spermatogenesis. The Sertoli cell database was screened to see whether any of these kinesins also were high in Sertoli cells. We reasoned that any kinesins identified as high in both screens could be candidate motors for spermatid translocation.
Anti-KIF20 Antibody Production
Antibodies were raised to hydrophilic regions specific to this protein. Specificity was determined by performing blast searches on peptide sequences using the National Center for Biotechnology Information protein database protein-protein BLAST search. Three sequences unique to KIF20 were chosen to generate antibodies: DLRSVVRKDLLSDCS (amino acids 30–44); KEDKADSDLEDSPEDE (amino acids 535–551); and GQASAKKRLGANQENQQ (amino acids 820–836). The peptides were synthesized and conjugated to KLH by Invitrogen (Burlington, ON, Canada). Polyclonal antibodies were generated by Invitrogen using standard protocols in New Zealand white rabbits 3–9 mo of age. On completion of the protocol, ELISA titer readings were obtained and rabbit serum was affinity purified. Of the six antibodies (two for each peptide sequence) produced, only three reacted with the peptide sequences to which they were generated, and only one of the antibodies (C1455), generated against the sequence GQASAKKRLGANQENQQ, reacted monospecifically on immunoblots of testis. This antibody was used for localization studies at the light and electron microscopic levels.
All animals used in these studies were reproductively active male Sprague Dawley rats and CD1 male mice. These animals were obtained from the University of British Columbia animal care colony and were maintained according to the guidelines established by the Canadian Council on Animal Care. All experiments were repeated at least twice on separate animals.
All chemicals and reagents were obtained from Sigma-Aldrich Canada unless otherwise indicated. Paraformaldehyde and NaCl were obtained from Fisher Scientific (Vancouver, BC, Canada), and all immunoglobulins and secondary antibodies conjugated to horseradish peroxidase or Texas Red were purchased from Jackson ImmunoResearch (West Grove, PA).
All secondary antibodies conjugated to Alexa fluorochromes for immunofluorescence were obtained from Invitrogen (Carlsbad, CA), whereas EMbed-812 resin was from Electron Microscope Sciences (Hatfield, PA).
Recombinant kinesin protein was obtained from Cytoskeleton Inc. (Denver, CO) and consisted of residues 1–379 of human kinesin heavy chain motor domain with a GST-Tag at the amino terminal end.
Testicular tissue for immunofluorescence was collected from reproductively active male Sprague Dawley rats between 250 and 500 g. Animals were anesthetized using Halothane (Halocarbon Laboratories, River Edge, NJ) inhalation. When the animals were under deep anesthesia, the testes were removed and perfused with warm (33°C) PBS (150 mM NaCl, 5 mM KCl, 0.8 mM KH2PO4, 3.2 mM Na2HPO, pH 7.3) for 2 min, followed by fixation with warm 3% paraformaldehyde in PBS for 30 min. Following fixation, the testes were perfused for an additional 30 min with PBS, frozen in liquid nitrogen, and then transferred to a cryostat. Sections of 5–10 µm thickness were collected on poly-L-lysine-coated slides and immediately immersed in cold acetone (–20°C) for 5 min before being air dried.
Testes were perfusion fixed as described above. The testes were removed from the canula and decapsulated in PBS. The mass of seminiferous tubules was minced with scalpels and then mechanically fragmented by passing them through first an 18-gauge needle and then a 21-gauge needle. Tissue fragments were placed into a 15-ml Falcon tube and allowed to sediment by gravity for 5 min at room temperature. The upper layer, enriched with spermatids with attached ectoplasmic specializations, then was transferred to another tube and pelleted by centrifugation (7300 x g for 4 min). The pellet was resuspended in a small amount of PBS and then added to poly-L-lysine-coated slides and placed in humidity chambers. After 10 min, all excess PBS was removed, and without being allowed to dry, the slides were quickly immersed in cold acetone (–20°C) for 5 min and then air dried.
Sectioned and fragmented tissue was rehydrated and blocked with 5% normal goat serum (NGS) in TPBS-BSA (PBS containing 0.05% Tween-20 and 0.1% BSA) and placed in a humidity chamber for 20 min. This was followed by the addition of 50 µl of primary antibodies made up in TPBS-BSA with 1% NGS (Sigma-Aldrich) overnight at 4°C in a humidity chamber. All primary antibodies were added at a final concentration of 0.01 mg/ml. Slides were washed three times (10 min each wash) with TPBS/BSA buffer, followed by the addition of 50 µl secondary antibody (goat anti-chicken or goat anti-rabbit conjugated to fluorochromes) for 1 h at 37°C. This was followed by an additional three washes (10 min each) with TPBS/BSA buffer. Slides were mounted with Vectashield (Vector Laboratories, Burlingame, CA) containing DAPI. Tissue then was viewed using a Zeiss Axiophot microscope fitted with fluorescent filters. Controls included replacing the primary antibodies with the same concentration of normal IgY or IgG, replacing the primary antibodies with buffer alone, and replacing both the primary and secondary antibodies with buffer alone.
Filamentous actin was labeled using Alexa phalloidin (Invitrogen, Carlsbad, CA). The stain was made up in TPBS-BSA and incubated at room temperature for 8 min before being washed in TPBS-BSA.
Peptide blocking of antibodies was performed using peptides to which the KIF20 antibodies were raised. The protocol was identical to procedures for immunofluorescence. However, primary antibodies were incubated with either 5 µl of peptide or ddH20 (final concentration of peptide twice that of the antibody) or 40 µl of peptide or ddH20 (final concentration of peptide 10 times that of antibody) at identical concentrations overnight on a LabQuake shaker (Labindustries, Inc., Berkeley, CA) at 4°C. After incubation, these dilutions were added to the tissue per normal protocols.
Rat tissue was homogenized in RIPA lysis buffer (150 mM NaCl, 50 mM Tris, pH 7.4; 5 mM EDTA, 1% Nonidet P-40, 1% deoxycholic acid [sodium salt], 10% SDS, protease inhibitor capsule [Roche]) using an electric homogenizer with 10 gentle passes at setting 3. Western blots were performed using our standard protocols [17]. Membranes were blocked with 4% milk for 8 h at 4°C and incubated with primary antibodies overnight at 4°C in 1% BSA. All antibodies were used at a final concentration of 0.001 mg/ml. The following day, membranes were extensively washed in TBST, and then secondary antibodies conjugated to horseradish peroxidase were added in 1% BSA in TBST for 1 h at room temperature. Membranes were extensively washed in TBST and TBS, then were reacted with ECL reagents (Amersham) and exposed to Kodak Biomax film for visualization. Controls consisted of replacing the primary antibodies with IgG at identical primary antibody concentrations.
Nitrocellulose membranes were incubated with 10 ml peptide solution in TBST at 10 µg/ml for 1 h. Three 5-min washes with TBST then were performed prior to blocking with 4% nonfat milk overnight. The following day, the membrane was placed into a Bio-Rad multi-protean II multiscreen apparatus that then was placed into a humidity chamber. Each lane of the Bio-Rad apparatus was incubated with a primary antibody for 4 h at room temperature, washed, and then incubated with a secondary antibody for 1 h. Blots then were washed and treated identically to Western blots.
Transfection of Cells with Plasmids Encoding KIF20
As an additional control for specificity of our #1455 antibody, we wanted to confirm that the antibody does react with KIF20. We transfected COS-7 cells with plasmids encoding the motor and analyzed cell lysates for reaction with our antibody. Full-length Kif20a cDNA cloned into pCMV-HA (Image ID 5720408; Clonetech Laboratories Inc.) was obtained from Open Biosystems. Polymerase chain reaction-based cloning was used to generate the correct reading frame, the plasmid was sequenced to ensure accuracy, and then transfected into COS-7 cells using Fugene 6 reagent (Roche). Twenty-four hours after transfection, cells were lysed in RIPA buffer, and Western blot analysis was performed.
Gelsolin-Treated Ectoplasmic Specializations
To obtain a testicular fraction enriched for ectoplasmic specialization components, spermatids with attached junctions were isolated from seminiferous epithelia and then treated with the actin-severing protein gelsolin to release plaque components into solution, as described elsewhere [10, 18].
Immunoprecipitations were performed as follows. A 500-µl aliquot of homogenized isolated seminiferous epithelium or other tissue was diluted to 1 mg/ml in TBS with protease inhibitors. Normal rabbit IgG (NRIgG; 5 µg) and 25 µl protein A beads (Sigma) were added to the sample, which was then incubated at 4°C for 1 h. This was followed by centrifugation at setting 10 in an Eppendorf desktop centrifuge at 4°C for 3 min. Following centrifugation, the supernatant was transferred to a new tube, and 5 µg primary antibody was added. After incubating at 4°C for 4 h, 25 µl protein A beads was added and the tube incubated overnight 4°C. The following day, the beads were washed five times in PBS buffer (pH 7.3) and the supernatants discarded. Sample buffer (45 µl) was added to the beads and boiled for 10 min. The remainder of the protocol was identical to Western blot analysis.
Rat testes were perfusion fixed (3% paraformaldehyde in PBS), and then the seminiferous tubules were mechanically fragmented by aspirating the material through 18-gauge and 21-gauge syringe needles. After letting the material settle for 10 min in a conical tube, the upper layer of sedimented material enriched with spermatids with attached ectoplasmic specializations was collected, and equal volumes were placed in four 1-ml microtubes. The material was concentrated by centrifugation (7300 x g for 4 min) and then resuspended in buffer (PBS, 0.1% BSA) containing primary antibodies. The anti-LAGSE antibody was used at 0.1 mg/ml, and the anti-1455 antibody was used at 0.1–0.5 mg/ml. The samples were incubated overnight at 4°C, washed three times with buffer, and then incubated at room temperature with second antibody conjugated to colloidal gold. After 2 h, samples were washed with PBS, fixed for 1 h with 1% paraformaldehyde and 0.1% gluteraldehyde, and then washed with 0.1 M sodium cacodylate (pH 7.3). Samples were postfixed for 1 h in 1% OsO4/0.1 M sodium cacodylate on ice and then processed using standard protocols for electron microscopy.
Controls consisted of replacing primary antibody with normal chicken IgY (NCIgY) or NRIgG, replacing the primary antibody with buffer alone, or replacing the primary and secondary antibodies with buffer alone.
Pan-kinesin Antibodies React with Their Respective Peptides, Conventional Kinesin, and Ectoplasmic Specializations
The pan-specific antibodies generated against LAGSE, HIPYR, and FAYGQ react with the peptide sequences to which each was raised (Fig. 2a) and with recombinant conventional kinesin on immunoblots (Fig. 2b). Significantly, in seminiferous tubules of fixed frozen sections of rat testis, the antibodies concentrate at apical Sertoli/spermatid ectoplasmic specializations (Fig. 3). Staining of these regions is particularly prominent in association with spermatids and attached ectoplasmic specializations that together have been mechanically separated from the epithelium (Fig. 4). The antibodies also label the manchette (Fig. 4b), a microtubule-rich structure present in spermatids and reported to contain kinesins [19, 20]. Staining is absent in tissue sections and at Sertoli/spermatid ectoplasmic specializations in epithelial fragments when NCIgY was substituted for the primary antibodies, when the primary antibody was replaced by buffer alone, and when both primary and secondary antibodies were replaced by buffer alone (Figs. 3, d–h and 4, d–g).
|
|
|
Of the three probes, the LAGSE antibody consistently resulted in the strongest signal at ectoplasmic specializations. In sections, staining appears to outline apical crypts containing elongate spermatids (Fig. 5a). The antibody recognizes multiple bands on immunoblots of seminiferous epithelium and testis that are not present on control blots when the primary antibody is replaced with NCIgY (Fig. 5b). This result is expected, because this pan-kinesin antibody would be predicted to recognize multiple kinesin types.
|
Because fluorescence staining at apical ectoplasmic specializations was more intense with the LAGSE antibody than with the HIPYR or FAYGQ antibodies, localization with the LAGSE antibody was pursued at the ultrastructural level. In electron micrographs of isolated spermatids immunolabeled prior to embedding, gold particles were present predominantly on the cytoplasmic face of the endoplasmic reticulum of ectoplasmic specializations (Fig. 6a). Controls were negative (Fig. 6, b–d).
|
The C1455 Antibody Detects Full Length KIF20
To more clearly define the type of kinesin present at the adherens junctions between Sertoli cells and spermatids, we screened data from mouse testis and Sertoli cell microarrays to identify potential candidates for the actual type of kinesin involved with spermatid translocation. Kinesin transcripts that increase at Days 30–35 (the time of spermatid entrenchment) in the testis database compared with Day 20 (prior to the generation of elongate spermatids) include Kif9, Kif17, Kif7, and Kif20. Relative levels of expression (signals) at Day 20 and at Days 30–35 are 444 and 764 for Kif9, 52 and 1441 for Kif17, 280 and 366 for Kif7, and 46 and 212 for Kif20 (data provided to us by Jim Shima and Dr. Mike Griswold, Washington State University [16]). When the Sertoli cell database was scanned for kinesin transcripts, only the Kif20 transcript had an appreciable level of expression at a signal of 498. We concluded that KIF20 was a candidate for spermatid translocation, because transcripts for this motor increase in the testis during the period when translocation first occurs in the testis, and the transcript also is present in Sertoli cells.
Antibody production was performed using specific peptides chosen from the KIF20a mouse sequence. In total, six antibodies were produced against three peptide sequences. Of these, only the antibody termed C1455 resulted in positive results when screened on testis tissue and immunoblots; therefore, only data for this antibody are presented here.
To determine whether the C1455 antibody was reactive against the peptide to which it was raised, the peptide was attached to an Immobilon-P Transfer membrane (Millipore, Billerica, MA) and then reacted with the antibody using standard immunoblotting techniques. The C1455 antibody reacted specifically with its peptide when compared to IgG controls (Fig. 7a). Of the other five antibodies produced, three did not react against their raised peptide, whereas the remaining two were not specific against their individual peptides (data not shown).
|
To verify that the 1455 antibody detected full-length KIF20, full-length HA-tagged Kif20a cDNA was transfected into COS-7 cells. Western blot analysis was performed, and the C1455 antibody reacted with a band at the same molecular weight as blots reacted with the anti-HA antibody (Fig. 7b). When compared to nontransfected cells, no band of comparable molecular weight was identified, indicating the overexpression of the KIF20a protein band was at an appropriate molecular weight of approximately 110 kDa.
Both C1455 and LAGSE Antibodies Detect a 128-kDa Protein in the Testis
On Western blots of rat testis and seminiferous epithelium, the C1455 antibody reacted specifically with a protein band at approximately 128 kDa (Fig. 7c) that was not present in other tissues that were tested (brain, lung, liver, kidney, spleen, and heart; data not shown). A prominent band at a much lower molecular weight also was present but was accounted for in control NRIgG blots.
Because KIF20 contains the LAGSE sequence and a band at approximately 128 kDa appears in blots of testis stained with the LAGSE antibody, we were interested in determining whether the LAGSE antibody recognized the protein identified by the 1455 antibody. To test the hypothesis that the LAGSE and C1455 antibodies were reactive with the same protein, we used the C1455 antibody to immunoprecipitate the antigen from testis and reacted an immunoblot of the precipitated protein with the LAGSE antibody. The LAGSE antibody reacted positively and specifically with a band at 128 kDa (Fig. 7d).
C1455 Reacts with Ectoplasmic Specializations
The C1455 antibody reacts with apical ectoplasmic specializations on fixed frozen sections of testicular tissue (Fig. 8a). Similar sites were labeled on spermatids with attached ectoplasmic specializations that together were mechanically separated from the seminiferous epithelium (Fig. 8e). No labeling was observed in material in which the primary antibody was replaced with the same concentration of normal IgG (Fig. 8, b and f) when the primary antibody was replaced with buffer alone (Fig. 8, c and g) or when both primary and secondary antibodies were replaced with buffer alone (Fig. 8, d and h).
|
Peptide Blocking Inhibits Antibody Staining
As an additional specificity control, the C1455 antibody was blocked using the peptide to which it was raised and then applied to sectioned and fragmented tissue (Fig. 9). In this material, there was only nonspecific staining throughout the tissue and no specific staining at apical ectoplasmic specializations.
|
C1455 Staining Is Stage Specific
If C1455 detects a kinesin that is associated with spermatid entrenchment, then labeling should be most intense at ectoplasmic specializations when spermatids are moved to the base of the epithelium. As predicted, ectoplasmic specializations associated with elongate spermatids from stages of spermatogenesis during which downward transport occurs were highly reactive with the C1455 antibody, whereas ectoplasmic specializations associated with later-stage spermatids prior to release were less or not reactive (Fig. 10).
|
C1455 Labels Testicular Fractions Enriched for Ectoplasmic Specializations
If C1455 reactivity is associated with ectoplasmic specializations, then testicular fractions enriched for components of these junction plaques should be strongly reactive with the antibody. To obtain testicular fractions enriched for ectoplasmic specializations, spermatids with attached ectoplasmic specializations were treated with gelsolin to disassemble the actin layer (Fig. 11a) and to release associated components and the endoplasmic reticulum into solution. Spermatids and any remaining attached junction components were precipitated, and both the supernatants and pellets reacted on immunoblots with the C1455 antibody. As predicted, the C1455 antibody reacted more strongly with the supernatant from gelsolin-treated spermatid/junction complexes than with the pellet (Fig. 11b). Densitometry revealed that the reactive band in the supernatant lane was 23% more intense than the reactive band in the pellet lane.
|
C1455 Labels the Cytoplasmic Face of Ectoplasmic Specializations
In spermatids with attached junction plaques that have been mechanically isolated from fixed seminiferous epithelium and then processed for immunoelectron microscopy, colloidal gold particles conjugated to the secondary antibodies were associated predominantly with the cytoplasmic face of the endoplasmic reticulum of ectoplasmic specializations (Fig. 12a). Gold particles were absent from similar locations in controls (Fig. 12, b–d).
|
C1455 Staining Is Specific Across Species
The C1455 antibody was tested on cryosections of fixed mouse testis (Fig. 13), and the results were consistent with those observed in rat tissue. Western blot analysis (Fig. 13, insets) of mouse testis compared with isolated rat seminiferous epithelial tissue and whole testis also displays a specific band at approximately 128 kDa with an additional lower-molecular weight band, the same as that observed in rat.
|
Here we present data consistent with the hypothesis that a kinesin is associated with apical ectoplasmic specializations in Sertoli cells and with the possibility that an isoform of KIF20 may be the kinesin responsible for spermatid entrenchment. The presence of microtubule-based motor proteins at apical ectoplasmic specializations, which are structures that occur at sites of adhesion to spermatids, is consistent with the hypothesis that microtubule-based transport machinery is coupled to intercellular junctions in the seminiferous epithelium to position and translocate elongate spermatids in the seminiferous epithelium.
Ectoplasmic specializations are a unique class of actin-related intercellular adherens junctions [21, 22]. They are tripartite structures that consist of regions of the Sertoli cell plasma membrane involved with adhesion to adjacent cells, a subsurface layer of hexagonally packed actin filaments arranged in bundles parallel to the plasma membrane, and a cistern of endoplasmic reticulum anchored to the actin layer. Integral membrane adhesion molecules present at the sites consist of 
6β1 integrins [23, 24] and nectin-2 [25]. N-cadherin also has been reported to be present [26, 27]. At apical sites of Sertoli cell attachment to spermatids, the ligand for the integrin is thought to be 
3-laminin [28]. The ligand for nectin-2 is nectin-3 in the adjacent spermatid plasma membrane [25]. Recently, vascular endothelial cadherin has been reported to be present in the spermatids at ectoplasmic specializations [29]. Intact ectoplasmic specializations remain attached to spermatids mechanically separated from the seminiferous epithelium [5, 30, 31], indicating that the three components of ectoplasmic specializations (Sertoli cell plasma membrane, actin filaments, and endoplasmic reticulum) and spermatid heads function as single mechanical units.
The microtubule-based spermatid translocation hypothesis states that ectoplasmic specializations, together with the attached spermatid heads, are moved along adjacent Sertoli cell microtubule tracts by motor proteins anchored to the cytoplasmic face of the endoplasmic reticulum component of the junctions [6, 8]. Although the hypothesis has not yet been tested in vivo, there are a number of observations consistent with this proposal. Microtubules in Sertoli cells are abundant [32–34], are arranged parallel to the direction of spermatid translocation [34], and appear structurally linked to ectoplasmic specializations in electron micrographs [35]. Microtubules bind to ectoplasmic specializations attached to spermatids mechanically separated from the epithelium, and the binding is ATP dependent [31]. In motility assays, also using isolated spermatids with attached junctions, ectoplasmic specializations are capable of transporting microtubules [36]. Significantly, microtubules move both in the plus end direction and in the minus end direction across ectoplasmic specializations [9], a result consistent with the presence of at least two types of motor proteins on the structures and with the down and up directions of spermatid translocation in vivo. Cytoplasmic dynein is concentrated around apical crypts containing spermatids in sections of seminiferous epithelium [10], and immunoelectron microscopy indicates that the motor is localized to the cytoplasmic face of the endoplasmic reticulum component of the junctions [9]. Here we present similar evidence that a kinesin also is associated with ectoplasmic specializations.
Antibodies generated to peptide sequences conserved in the head region of most kinesins react with apical ectoplasmic specializations in fixed frozen sections of seminiferous epithelium and with ectoplasmic specializations attached to spermatids mechanically separated from fixed seminiferous epithelium. Significantly, the LAGSE antibody reacts specifically with the cytoplasmic face of the endoplasmic reticulum of the junctions when sections are analyzed by immunoelectron microscopy. Because the LAGSE sequence also occurs in a subset of nonmotor peptides, it is impossible for us to rule out the possibility that the probe is reacting nonspecifically; however, the antibody does react with other regions of the seminiferous epithelium in which kinesins are concentrated, the antibody does recognize conventional kinesin on immunoblots, and the other two pan-specific antibodies also react with ectoplasmic specializations, but more weakly. Results with the pan-kinesin antibodies are consistent with the prediction that a kinesin is associated with the junction plaques.
To identify candidate motors, we screened data from mouse testis and Sertoli cell microarrays to identify potential candidates for the actual type of kinesin involved with spermatid translocation. KIF20 appeared the most likely candidate, because transcripts for this motor increase in the testis during the period when translocation first occurs in the testis, and the transcript is abundant in Sertoli cells.
KIF20 (previously known as Rab6KIFL) is a kinesin that binds to, and is regulated by, the small guanosine triphosphatase (GTPase) Rab6 [37]. Based on sequence, the motor protein has a predicted weight of around 100 kDa [37]. Although initially localized to the Golgi, where it is thought to be involved with vesicular transport through the organelle, the motor has been since localized to the midzone of the mitotic spindle and to the cleavage furrow in dividing cells [38]. Significantly, transcripts for the motor are highly expressed in the testis [37, 39], and the motor also has been localized to intercellular adhesion sites in cultured pancreatic ductal adenocarcinoma cells [39].
Our results raise the interesting possibility that an isoform of KIF20 may be associated with apical ectoplasmic specializations in the testis. An antibody we generated to a peptide sequence specific to KIF20 reacts with sites known to contain apical ectoplasmic specializations in fixed testis sections and in fixed epithelial fragments. Standard controls for specificity are negative, and antibody binding is dramatically reduced when blocked with peptide. At the ultrastructural level, reactivity at the junctions predominantly is with the cytoplasmic face of the endoplasmic reticulum. On blots, the antibody reacts with the peptide to which it was generated and with full-length KIF20. However, on immunoblots of testis and seminiferous epithelium, the antibody reacts with a peptide band at approximately 128 kDa. Although 128 kDa is a higher molecular weight than that predicted for KIF20, this result is not entirely unexpected. Other proteins, such as espin [40] and some kinesins [19], also are known to have different molecular weights in testis than in other tissues. The observation that the protein is present in the pellet in gelsolin-treated spermatid/junction complexes suggests that there may be some degree of contamination from the supernatant, that disassembly of ectoplasmic specializations is incomplete, or that KIF20 also is present in germ cells. Our results are consistent with the conclusion that at least one of the kinesins associated with ectoplasmic specializations may be an isoform of KIF20.
The presence of a kinesin at ectoplasmic specializations in Sertoli cells at sites of attachment to elongate spermatids has long been predicted based on the microtubule-based spermatid translocation hypothesis. Results from our previous motility assays [9] and from our current approaches reported here are consistent with the prediction that a kinesin is present at these unique junction sites. The possibility that the kinesin is an isoform of KIF20 is exciting, because it identifies KIF20 as a candidate for the motor responsible for the translocation of spermatids deep into the epithelium—a novel function for this motor protein. Also, it suggests to us that one of the elements involved in regulating spermatid translocation may be Rab6. Eventually, genetic or molecular approaches will be required to test the microtubule translocation hypothesis in vivo and to determine the biologic significance of the translocation process itself. The translocation machinery in Sertoli cells or events that occur in spermatogenic cells as a result of translocation may provide molecular targets for contraception.
FOOTNOTES
1Supported by a Canadian Institutes of Health Research grant (MOP 62728) to A.W.V. 
Correspondence: 2A. Wayne Vogl, Department of Cellular and Physiological Sciences, Division of Anatomy and Cell Biology, Faculty of Medicine, Life Sciences Centre, 2350 Health Science Mall, Vancouver, BC, Canada V6T 1Z3. FAX 604 822 2395; e-mail: vogl@interchange.ubc.ca
3These authors contributed equally to this work.
Received: 27 June 2007.
First decision: 12 July 2007.
Accepted: 31 August 2007.
REFERENCES
griswold/microarray/). (June 27, 2007).This article has been cited by other articles:
|
|
 |
|
  D. D. Mruk, B. Silvestrini, and C. Y. Cheng Anchoring Junctions As Drug Targets: Role in Contraceptive Development Pharmacol. Rev., June 1, 2008; 60(2): 146 - 180. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
