| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Anatomy, University of British Columbia, Vancouver, British Columbia, Canada V6H 3Z6
b Department of Biochemistry, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z3
ABSTRACT
The seminiferous epithelium contains unique actin related cell-cell junctions, termed ectoplasmic specializations (ESs). Turnover of these junctions is fundamental to sperm release and to movement of spermatocytes from basal to adluminal compartments of the epithelium during spermatogenesis. In this study we report several novel observations related to the spatial and temporal distribution of integrin-related signaling molecules at ESs. We confirm the presence of ß1-integrin at these sites and further demonstrate co-localization of integrin linked kinase (ILK). ß1-Integrin and ILK were shown by immunoprecipitation to associate in whole cell lysates of seminiferous epithelium. This observation provides the first evidence for a direct ß1-integrin/ILK interaction in noncultured epithelium. Pan-cadherin and ß-catenin antibodies did not react at ESs. Rather, antibodies reacted with desmosome-like junctions that are present both at basal junctional complexes between Sertoli cells and at sites of attachment to spermatogenic cells. Focal adhesion kinase (FAK), a known integrin-associated molecule, did not codistribute with ß1-integrins and did not associate with these adhesion molecules in immunoprecipitation studies. Although FAK was expressed in the epithelium, it appeared to be limited to the cytoplasm of early spermatogenic cells. Significantly, polyclonal antibodies against phosphotyrosine-containing residues reacted strongly at ESs, with highest levels detected during sperm release and turnover of basal junction complexes. Our observations indicate that ESs share cell signaling features both of cell-cell junctions and of cell-extracellular matrix junctions.
developmental biology, gametogenesis, Sertoli cells, signal transduction, spermatid, spermatogenesis
INTRODUCTION
The seminiferous epithelium contains an architecturally unique type of cell-cell junction, termed an ectoplasmic specialization (ES). As developing spermatids mature and elongate, ESs are formed in Sertoli cells adjacent to spermatid heads. This site of adhesion between the germ cell and Sertoli cell is disassembled as part of the process by which spermatids are released into the tubule lumen [1, 2]. These junctions are also believed to be fundamental in the maintenance of adhesion between neighboring Sertoli cells near the base of the epithelium [2, 3]. Like apical ESs, the structural components of basal ESs are dynamic and undergo cyclic disassembly and reassembly as primary spermatocytes move from basal to adluminal compartments of the epithelium [1].
Molecules reported to be present at ESs include actin [4], vinculin [5, 6], 
-actinin [4], fimbrin [7], and espin [8]. The only transmembrane adhesion molecule known to be present, however, is 6ß1 integrin [9–12]. Although cadherins are expressed in the testes [13, 14], none have yet been convincingly localized to ESs. Furthermore, even though the basic ultrastructural features and some of the molecular components of ESs are known, surprisingly little is known about how these specialized cell-cell junctions are regulated. The presence of ß1-integrin and the apparent absence of known cadherins from these sites indicates to us that the junctions may have signaling properties similar to some known cell/matrix adhesion junctions.
The 6ß1 integrin complex has been previously believed to be mainly a laminin receptor [15]. However, reports have suggested that it may function on the egg surface as a sperm receptor and, therefore, potentially be involved in cell-cell interactions [16]. Support for a functional role for integrins in the seminiferous epithelium is based on the disappearance of integrin expression concurrent with the turnover of ESs [11]. As spermatocytes translocate from the basal to the adluminal compartment, integrin expression disappears and then subsequently reappears [12]. Also, the disappearance of integrin expression has been shown to coincide with disassembly of the apical ESs and spermatid release into the tubule lumen [10]. With reports of ß1 integrin expression coinciding with key spermatogenic events it is possible that integrin-related molecules could also show changing patterns of expression in the seminiferous epithelium and show protein interactions that are analogous to in vitro focal contacts.
Two kinases that have been shown to associate with integrins in vitro include focal adhesion kinase (FAK) [17–19] and integrin linked kinase (ILK) [20, 21]. These submembranous molecules are believed to be closely associated with integrin cytoplasmic tails and to mediate signals to the cytoskeleton. Focal adhesion kinase, a tyrosine kinase, has been shown to interact in vitro with ß1-integrin. Specifically, it becomes phosphorylated at tyrosine residues during integrin activation and has been shown to induce changes in other signaling proteins [17, 18]. Focal adhesion kinase has never been shown to associate with ß1-integrin cytoplasmic tails in vivo or in noncultured cells. This is unlike the situation with ILK, a serine-threonine kinase, which has been shown to have a relatively strong in vivo association with ß1, ß2, and ß3 integrin cytoplasmic domains [20]. Adhesion molecules downstream of ILK and FAK include vinculin and paxillin. These proteins have been shown in vitro to have roles in the regulation of both adhesion and downstream signaling events and have, therefore, been termed linking or docking proteins to which other components of the actin cytoskeleton can associate.
To date, very little is known about the presence and distribution of molecules that are capable of signaling in the seminiferous epithelium. We use Western blotting, immunofluorescence, immunoprecipitation, and immunoelectron microscopy to explore the possibility that key members of ß1-integrin signaling pathways are present at cell-cell junctions (ESs) in the seminiferous epithelium as a prerequisite to determining their role in events such as sperm release and turnover of the basal junctional complex. We present the first evidence for a direct in vivo protein interaction between ß1-integrin and ILK. We also provide evidence suggesting that a cadherin/catenin complex is not found at ESs, but rather localizes to desmosome-like junctions. We further show that FAK is not an immediate component of ES signaling but that another tyrosine kinase is likely present. Our results may have implications for signaling events involved in sperm release and turnover of adhesion junctions associated with the "blood-testes" barrier.
MATERIALS AND METHODS
Animals
Tissue for all experiments was obtained from adult Sprague-Dawley rats that ranged in weight from 250 g to 550 g. The animals were obtained from a colony in the Animal Care Facility at the University of British Columbia and were housed according to standards established by the Canadian Council on Animal Care. Animals were anesthetized using halothane administered via the respiratory tract. Testes were removed and rats were killed while under deep anesthesia.
Reagents
Most reagents were obtained from Sigma-Aldrich (Oakville, ON, Canada). Fluorescent phallotoxins were obtained from Molecular Probes (Eugene, OR) and paraformaldehyde was obtained from Fisher Scientific (Vancouver, BC, Canada).
Preparation of Tissue for Immunostaining
Testes were perfusion-fixed via the spermatic artery with 3% paraformaldehyde in PBS (150 mM NaCl, 5 mM KCl, 3.2 mM Na2HPO4, 0.8 mM KH2PO4 pH 7.3) at 33°C for 30 min. Following fixation, testes were perfusion-washed for an additional 30 min with PBS. Both fixative and wash solutions were filtered using 0.22-µm Millipore filters and warmed to 33°C prior to perfusion.
After fixing, the testes were frozen in OCT compound (Fisher Scientific) and then equilibrated to chamber temperature in a cryostat (Brights 5010). For immunofluorescence studies, sections were cut to a thickness of 10 µm, collected on polylysine-coated slides, and then immediately treated with -20°C acetone for 10 min. Sections were then air-dried for 10 min. When immunostaining with probes for ß1-integrin and ILK, testes were immediately frozen after excision from the animal; that is, the tissue was frozen fresh without prior fixation.
Immunofluorescence
Prior to immunostaining, tissue was treated for 30 min with 5% normal goat serum in TPBS (0.1% BSA, 0.1% Tween-20, PBS) to minimize nonspecific binding of secondary antibodies. Incubation times with primary antibodies were either 1–2 h at 37°C or overnight at 4°C. Primary antibodies were used at the following dilutions: polyclonal anti-ß-catenin (C2206, Sigma-Aldrich) at 1:1000, monoclonal anti-pan cadherin (C1821, Sigma-Aldrich) at 1:400, polyclonal anti-ß1 integrin (AB 1952, Chemicon International, Temecula, CA) at 1:200, polyclonal anti-ILK at 1:100 (a gift of S. Dedhar), monoclonal anti-FAK (F15020, Transduction Laboratories, Lexington, KY) at 1:50, monoclonal anti-paxillin (F15020, Transduction Laboratories) at 1:400, monoclonal anti-vinculin (V9131, Sigma) at 1:400, and polyclonal antiphosphotyrosine (P11230, Transduction Laboratories) at 1:100.
After primary antibody staining, tissue sections were washed three times for 10 min with TPBS and then incubated with secondary antibodies (Jackson Immunoresearch Laboratories Inc., West Grove, PA) for 1 h at 37°C at a 1:100 dilution. After secondary antibody staining, sections were washed three times for 10 min with TPBS and then mounted with Vectashield (Vector Laboratories, Burlingame, CA) and a coverslip. In some cases the second wash included rhodamine or fluorescein phalloidin to specifically label f-actin structures. Slides were observed and photographed using a Zeiss Axiophot microscope (Carl Zeiss, Inc., Thornwood, NJ).
Tissue Preparation for Gel Electrophoresis and Immunoblotting
Testes were decapsulated and the seminiferous tubules were placed in a glass Petri dish containing cold PEM (80.0 mM PIPES, 1.0 mM EGTA, 0.1 mM MgCl2, 250 mM sucrose pH 6.8 with KOH). Tissue was minced with two scapel blades in cold PEM containing a cocktail of protease inhibitors (0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin, 0.5 µg/ml aprotinin, 0.1 mM phenylmethylsulfonyl fluoride [PMSF]), for a maximum of 60 min at 4°C. Using fine dissecting tools and a Zeiss Dissecting Microscope fitted with dark-field condenser (Don Mills, ON) epithelia were separated from tubule walls by securing one end of the tubule and using another probe to squeeze the tubule contents out toward the other end. Ejected epithelia were collected with a micropipette and placed in fresh buffer on ice.
Epithelia were lysed both under denaturing (Western blots of whole cell lysates) and native (all coimmunoprecipitations) conditions. Denaturing conditions included 30 min of incubation with NP-40 lysis buffer (30 mM Tris pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% deoxylacholate, 0.1% SDS, 2 mM EDTA) followed by boiling and 2 sec of sonication at 60 MHz. Native lysis conditions used NP-40 lysis buffer (30 mM Tris pH 8, 150 mM NaCl, 1% NP-40, 10% glycerol, 2 mM EDTA) without sonication or boiling. Lysis buffers included a cocktail of protease inhibitors (0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin, 0.5 µg/ml aprotinin, 0.1 mM PMSF and 0.1 mM Na vanadate).
Gel Electrophoresis and Western Blotting
Tissue samples were mixed with 2x sample buffer (0.125 M Tris-HCl pH 6.8, 4% SDS, 20% glycerol, 5% mercaptoethanol), boiled at 100°C for 5 min, placed on ice, and then loaded into gels. Tissue samples were electrophoresed on 7.5%, 10% or 12% SDS-PAGE minigels and transferred to polyvinylidene difluoride membranes for 4 h or overnight. After transfer, membranes were washed twice for 5 min in TBST (100 mM-Tris-HCl pH 7.5, 0.9% NaCl, 0.1% Tween-20) and then blocked with 10% dried skim milk in TBST for either 1 h at room temperature or overnight at 4°C.
Primary antibodies for Western blotting were used at the following dilutions and were obtained by companies previously specified: monoclonal anti-ß-catenin (1:1000), polyclonal anti-ß-catenin (1:5000), monoclonal anti-pan-cadherin (1:1000), polyclonal anti-pan-cadherin (1:5000), polyclonal anti-ß1-integrin (1:1000), monoclonal anti-ILK (1:500), polyclonal anti-ILK (1:1000), monoclonal anti-FAK (1:500), monoclonal anti-paxillin (1:10 000), monoclonal anti-vinculin (1:2000), and polyclonal anti-phosphotyrosine (1:1000).
Following primary incubations membranes were washed three times for 10 min with TBST followed by three times for 10 min with TBS. Membranes were incubated either with anti-mouse-horse radish peroxidase (HRP) (sc-2031, Santa Cruz Biotechnology, Santa Cruz, CA) or anti-rabbit-HRP (sc-2030, Santa Cruz Biotechnology) at a dilution of 1:5000. Membrane bound antibodies were detected using chemiluminescence with ECL Western blotting detection agents (RPN 2106, Amersham Life Science, Piscataway, NJ) for incubation periods of 1 min and then exposed immediately to radiographic film.
Immunoprecipitation
Whole cells lysates of seminiferous epithelia were precleared for at least 60 min by adding 30–50 µl of a 50% slurry of protein A-sepharose beads (P9424, Sigma) to 1 ml of total cell lysate. After preclearing, the lysates were centrifuged for 5–10 sec at high speed to pellet beads. Supernatants were then transferred to new, prechilled tubes. Primary antibodies were added to cell lysates and incubated overnight without agitation. During the last 60 min of incubation 30–50 µl of 50% protein A or protein G-sepharose was added to the lysates and allowed to incubate while on a platform roller. Tubes were centrifuged at high speed for 15–20 sec to sediment immune complexes; supernatants were saved for Western blotting to assess the extent of specific antigen precipitation. Immune complexes were then washed four to five times by sedimenting the sepharose beads and then resuspending them in lysis buffer at a volume of 10 times that of the beads. In the final wash, beads were resuspended in 20 mM Tris pH 7.5 to remove detergents contained in the wash buffer. Immunoprecipitate complexes were then resuspended in 2x SDS-PAGE sample buffer and boiled for 5 min to elute bound proteins. Beads were pelleted and the supernatants were removed and either used immediately for gel electrophoresis or frozen at -20°C.
Electron Microscopy
Testes were perfused with warm (33°C) fixative (80 mM PIPES, 1.0 mM EGTA, 1.0 mM MgCl2, 0.5% gluteraldehyde, 3.0% paraformaldehyde adjusted to pH 6.8 with KOH) for 30 min. After perfusion, testicular tissue was fragmented with scapel blades and washed in a test tube for 1 h. Cells were sedimented by centrifugation at 11 000 rpm in an Eppendorf centrifuge and the supernatant was replaced with 0.1 M sodium cacodylate (pH 6.9). Samples were washed twice with fresh buffer and postfixed for 1 h on ice in 0.1 M sodium cacodylate containing 1% osmium tetroxide. Samples were further processed using standard techniques for electron microscopy and photographed on a Philips 300 electron microscope.
Immunogold Labeling
Testes were fixed as described elsewhere [22] and embedded in lowicryl. Tissue sections were collected on carbon/formvar-coated nickel grids and treated for 5 min with a pre-block solution (0.01 M glycine, 0.1% BSA, 1:20 normal goat serum, and 0.05% Tween-20). Sections were incubated with pan-cadherin antibodies in antibody solutions (0.01 M glycine and 1% NGS) at a dilution of 1:50 for 2 h at 37°C. After staining, grids were washed three times for 10 min with wash buffer (0.01 M glycine and 0.1% BSA) and then incubated for 60 min at 37°C with gold-conjugate secondary antibodies (G7652, Sigma) diluted 1:20 in secondary antibody solution (0.1% BSA, 0.05% Tween-20, and 5% fetal bovine serum). Sections were washed with PBS and fixed for 10 min with 2% gluteraldehyde in PBS followed by two washes with dH2O. Grids were stained with 1% uranyl acetate, washed, and dried.
RESULTS
Distribution of Actin as a Marker for ESs
The distribution of actin filaments, as determined by fluorescent phallotoxin staining, was used as a marker for basal and apical junction sites (Fig. 1). Actin was concentrated both at Sertoli-germ cell ESs (Fig. 1a, S/Gj), Sertoli-Sertoli cell ESs (Fig. 1a, S/Sj) and developing junctional sites (Fig. 1d, dj). Actin was also found at intercellular bridges (Fig. 1b, ij) and tubulobulbar processes (Fig. 1c, tb).
|
Cadherin and ß-Catenin Co-Localize at Non-ES Sites
ß-Catenin and cadherins were highly expressed in the seminiferous epithelium. Staining for ß-catenin (Fig. 2i, a' and b', arrow) and pan-cadherin (Fig. 2i, a'' and b'', arrow) was restricted to the area of Sertoli cell-Sertoli cell junctional complexes. Overlay staining showed pan-cadherin to be codistributed with ß-catenin throughout the epithelium with nearly identical spatial and temporal distribution. In both double-label experiments and single-label experiments specific staining for neither molecule was detectable at Sertoli cell/germ cell ESs during any stage of spermatogenesis (Fig. 2i, a'–e', a''–e'', arrowhead). Even with high concentrations of primary antibodies (1:10) and extended incubation periods (2–3 h at 37°C) specific staining was not observed at these sites (data not shown). At all stages, the staining pattern for the cadherin and catenin molecules appeared punctate rather than linear and continuous. Single-labeled tissue sections counterstained with phalloidin (data not shown) revealed that neither molecule codistributed with actin found at Sertoli/Sertoli junctional sites. In addition to the intense staining observed at basal sites, discrete punctate staining for the cadherin and catenin was observed in more apical positions (Fig. 2i, e'–e''', arrow). This staining pattern appeared to localize to areas of contact between adjacent cells and was observed in association with most early germ cells including spermatocytes and round spermatids but was not apparent in association with elongate spermatids. Specificity of the catenin probe was verified by a convenient internal marker, a blood vessel in the interstitium, which showed intense staining coincident with the borders of cells likely to be smooth muscle cells (data not shown).
|
Coimmunoprecipitations suggested a cadherin/catenin interaction at basal parts of the epithelium. ß-Catenin immunoprecipitations probed with pan-cadherin antibodies showed the presence of specific cadherin at 135 kDa (Fig. 2iia). Similarly, pan-cadherin immunoprecipitates probed with ß-catenin antibodies indicated a specific band at 95 kDa (Fig. 2iib). Normal mouse immunoglobulin G (MsIgG) and normal rabbit IgG (RbtIgG) were used for negative immunoprecipitation controls.
Immunoelectron Microscopy Suggests That Cadherins Are Not Localized at ESs
Ultrastructural immunolabeling supported light-level observations that pan-cadherin did not localize to ESs (Fig. 3, a and b). Rather, staining occurred at desmosome-like junctions as identified by the presence of intermediate filaments (Fig. 3, c and d; arrow).
|
ß1-Integrin and ILK Interact at ESs
The temporal and spatial distribution of the ß1 integrin (Fig. 4i) and ILK (Fig. 4ii) in unfixed cryosections was consistent with that of ES actin both at Sertoli cell/germ cell junctions (S/Gj) and Sertoli cell/Sertoli cell junctions (S/Sj). Although preservation of morphology was poorer in these tissue sections (integrin and ILK antibodies showed specific staining only on unfixed tissue), integrin expression was clearly coincident with the presence of junctional actin at ESs. Both integrin and ILK staining was most abundant on dorsal sides of elongate spermatids and was most evident at stage IX (Fig. 4i, c, and Fig. 4ii, c; S-Gj). Both molecules were detected during the commencement of ES formation (Fig. 4i, d, and Fig. 4ii, d; dj); that is, during stage IX when round spermatids become polarized and elongate in shape. Although the most intense staining of ILK was coincident with staining for ß1 integrin, ILK was also expressed at nonjunctional sites or areas where ß1-integrin or actin was not detectable. Specifically, faint ILK staining could be observed at the margins of spermatocytes and spermatogonia (Fig. 4ii, e). In general, ILK was more difficult to localize at basal sites because it did not appear to form a linear contour that follows the junctional complex as defined by actin staining. Integrin expression was not detected at non-ES sites such as intracellular bridges and tubulobulbar complexes. Intense staining of the tubule wall, which was dissected away for biochemical studies, appeared nonspecific and was accounted for by normal IgG and secondary antibody alone negative controls.
|
Coimmunoprecipitations supported an in vivo interaction for ß1-integrin and ILK. When polyclonal ILK immunoprecipitates were probed for ß1-integrin a specific band was observed at 140 kDa (Fig. 4iiia). ß1-Integrin immunoprecipitates probed with monoclonal ILK antibodies (Fig. 4iiib) showed several lower molecular weight species but a major band at 60 kDa, suggesting the detection of specific ILK. Normal IgG immunoprecipitates did not contain any detectable ß1-integrin or ILK.
FAK, Vinculin, and Paxillin Interact but Do Not Co-Localize to ESs
Expression of FAK was not detected at ESs at any spermatogenic stage. Expression was highest in the cytoplasm of germ cells juxtaposing the extracellular matrix, the spermatogonia (Fig 5i, b and d; arrowhead), where staining intensity was comparable to that of endothelial cells of blood vessels in the interstitial (data not shown). Vinculin (Fig. 5ii) staining had a pattern similar to that of actin, with greatest concentrations detected at ES sites. Vinculin staining appeared less abundant along the dorsal aspect of the elongate spermatids just prior to spermiation but did not codistribute with staining for FAK during any stage of spermatogenesis.
|
Paxillin was expressed at high levels throughout both Sertoli cells and germ cells (Fig. 5iii). Paxillin staining did not concentrate at junctional sites (Fig. 5iii, b; arrow), but appeared diffuse and cytoplasmic, and was highly expressed in the tubulobulbar processes (Fig. 5iii, c; arrow) and at the apical aspect of Sertoli cells after spermiation (Fig. 5iii, d). Primary spermatocytes showed considerably less paxillin staining than any other cell type (Fig. 5iii, a; arrow). Apical staining of Sertoli cells was considerably less intense by stage X–XIV (Fig. 5iii, e).
Although FAK, vinculin, and paxillin did not co-localize to ES sites, biochemical data suggested direct vinculin/paxillin (Fig. 5, iva and ivb) and FAK/paxillin (Fig. 5, ivc and ivd) interactions. Precipitates with an anti-vinculin antibody showed a strong signal at 68 kDa when probed with paxillin (Fig. 5iva). However, paxillin immunoprecipitates (Fig. 5ivb) appeared to contain relatively less vinculin. Paxillin associated with a polyclonal FAK immunoprecipitates as recognized by a band at 68 kDa (Fig. 5ivd). Paxillin antibody precipitates revealed a band at 120 kDa when probed for FAK, suggesting a stable interaction. Based on light-level data, protein interactions between vinculin and paxillin likely occurs at tubulobulbar processes because this is the only location in the epithelium that showed co-localization. It is possible that paxillin exists at ESs and interacted with vinculin, however, immunocytochemistry studies (data not shown) showed this to be unlikely. In addition, the interaction between paxillin and FAK is likely restricted to spermatogonia because this is the only site in the epithelium that showed light-level coexpression. To further qualify distribution of FAK, single cell staining (Fig. 6) showed that FAK (red) does not co-localize with junctional actin (green). Biochemical data did not show a direct ß1/FAK interaction (Fig. 6, ii and iii).
|
High Levels of Phosphotyrosine at ESs
Immunostaining with a rabbit polyclonal antibody revealed phosphotyrosine-containing proteins to be widely expressed throughout the seminiferous epithelium (Fig. 7i), including nonjunctional components of the epithelium. Expression was concentrated at nonjunctional cell borders of most cells including germ cells (Fig. 7i, c; arrowhead) and nonjunctional cell borders of Sertoli cells (Fig. 7i, nj). Expression, however, was most intense at S-S junctions (Fig. 7i, a) and S-G junctions (Fig. 7i, b), sites where ESs are known to occur. Unlike the other molecules studied, phosphotyrosine showed staining around both the dorsal and ventral components of the spermatid head. Stage VII/VIII phosphotyrosine staining of the S-Sj appeared distributed along only a small portion of the length of the junction (Fig. 7i, c') and differed from the more linear staining pattern (Fig. 7i, a; S-Sj) that was coincidental with the majority of basal actin. Phosphotyrosine also appeared concentrated at tubulobulbar complexes (Fig. 7i, c; tb) but disappeared after sperm release (Fig. 7i, d). Phosphotyrosine staining was not obvious in developing junctions (Fig. 7i, c and d).
|
Polyclonal antibodies against phosphotyrosine residues showed 5 bands including 55, 70, 120, 130, and 140 kDa (Fig. 7ii). Bands at 70 and 120 kDa were suspected to be the phosphotyrosine proteins, paxillin and FAK, respectively. To investigate this possibility, paxillin and FAK immunoprecipitates, under denaturing conditions, were probed for phosphotyrosine-containing residues. Interestingly, immunoprecipitates for paxillin (Fig. 7iii) showed a high level of reactivity, whereas polyclonal FAK precipitates appeared to have no detectable tyrosine phosphorylation (Fig. 7iv).
DISCUSSION
Previous experiments have shown that in the adult rat seminiferous epithelium there is a curious absence of several key adhesion molecules that have been implicated with signaling events associated with cellular adhesion. Some members of the cadherin family, such as E-cadherin, have been detected at cell-cell junctions in many types of epithelia [23] and have been shown to promote cellular adhesion, but appear not to be expressed in the rat seminiferous epithelium [24]. Some investigators have suggested that N-cadherin may be present at adhesion sites between Sertoli cells and early spermatids when grown in vitro [24, 25] and at sites of spermatid/Sertoli cell adhesion in whole seminiferous epithelium [26]. The - and ß-catenins, two cadherin-associated molecules, have been reported to occur at basal S/S junctional complexes but are absent from apical S/G junctions [24]. The only adhesion molecule found to be present both at apical and basal ESs is ß1-integrin [9–12]. Results presented here indicate that epitopes reactive with pan-cadherin and ß-catenin antibodies were concentrated at desmosomes and not at ESs. The only staining detected at apical sites could be accounted for with negative staining controls.
ß1-Integrin and ILK Are Present at ESs
Our studies have shown that expression of ß1 integrin is codistributed with actin filaments within ESs as indicated by phalloidin staining. We have further demonstrated that ß1-integrin is expressed during ES development and shows a close spatial and temporal distribution to that of ILK. Maximum immunofluorescence staining of both molecules is observed adjacent to spermatids deep within the Sertoli cell crypts. Just prior to sperm release, staining occurs along the dorsal side of the spermatid head in the same location as actin staining in ESs. Staining for ß1 and ILK also appears codistributed with newly forming ESs, which develop in close association with early spermatids during stage IX of spermatogenesis. The close spatial and temporal association of ß1-integrin, ILK and ES actin filaments at ES sites indicates that the ß1-integrin/ILK complex may participate in junction maintenance, turnover during spermiation, and movement of spermatocytes through basal junction complexes. Coimmunoprecipitation of these two molecules provides biochemical support for a direct ß1-integrin/ILK interaction in vivo. This interaction is likely to occur at junctional sites because this is the only instance in which these two molecules are coexpressed.
As ß1-integrin and ILK appear to be coexpressed throughout the spermatogenic stages, it is possible that the ß1-integrin/ILK complex is a constitutive one rather than an induced one that is involved with continuous structural changes of apical ESs as they conform to shape changes of the spermatid head. The only location in the epithelium where ß1-integrin and ILK are not codistributed is in the early germ cells. Here, some ILK staining occurs at the periphery of primary spermatocytes. This staining could be attributed to ILK associating with other known substrates such as the ß2- and ß3-integrin cytoplasmic tails [20], both of which have not yet been identified as being present in the adult rat seminiferous epithelium.
Focal Adhesion Kinase Is Not Associated with ESs
FAK is not detectable at junctional sites in the adult rat seminiferous epithelium as indicated by the absence of staining at sites labeled with fluorescent phalloidin. FAK staining is restricted to basal germ cells and appears diffusely, but intensely, expressed throughout the cytoplasm of spermatogonia and primary spermatocytes. Expression of FAK in spermatogonia cytoplasm is more intense than that in spermatocytes and is comparable to FAK staining in endothelial cells, a convenient positive control found in the interstitial blood vessels (data not shown). To further corroborate that FAK does not associate with ß1-integrin, in vivo coimmunoprecipitations failed to show a direct interaction in whole cell lysates of these proteins. It is interesting that a direct in vivo association in tissue or primary epithelia between ß1-integrin and FAK has never been shown despite their strong interaction in cultured cells [18, 27, 28]. By reason of the fact that FAK has been shown to be present at cell adhesion complexes in many other systems, the absence of this molecule at ESs (an actin-related intracellular adhesion junction) is a significant finding and indicates that this molecule does not play a direct role in signaling events involved with spermiation and turnover of the blood testes barrier.
Phosphotyrosine Proteins Are Concentrated at ESs
The absence of FAK at ESs indicates to us that another major tyrosine kinase may be present. The observation that a probe for phosphotyrosine intensely labels both S/S junctions and S/G junctions is consistent with this possibility. Significantly, phosphotyrosine staining dramatically highlights basal junctions coincident with movement of spermatocytes through the blood-testes barrier, an event known to occur during stage VII/VIII of spermatogenesis in the rat. This observation indicates that a major phosphotyrosine containing protein may be involved with the rearrangement of the S/S junctional complex. In addition, because phosphotyrosine was prominent at S/G junctions just prior to spermiation, the same or a similar phosphotyrosine-containing protein could be functionally important for breakdown of the junction. When whole cell lysates were probed with polyclonal phosphotyrosine antibodies, several prominent bands appeared at about 55, 70, 120, 130, and 140 kDa. Although the presence of a band at about 120 kDa seems to suggest the presence of the tyrosine kinase, FAK, this does not appear to be the case. FAK immunoprecipitates appeared not to contain detectable phosphotyrosine, while phosphorylated paxillin is present in relatively high amounts in seminiferous epithelia cell lysates. With the detection of a major phosphotyrosine band at about 70 kDa and the abundance of paxillin detected at the light level, it is possible that paxillin is one of the major phosphotyrosine-containing proteins in the adult rat seminiferous epithelia.
Paxillin Is Highly Expressed in the Seminiferous Epithelium but Is Not Concentrated at ESs
Paxillin, a vinculin-binding protein, has a nearly ubiquitous spatial distribution throughout the seminiferous epithelium cytoplasm. The most outstanding observation regarding paxillin expression throughout spermatogenesis is its increase in abundance during stages VII/VIII in round spermatids and in the columnar portion of Sertoli cells. Curiously, paxillin does not appear to be abundantly present in the spermatocytes. The recruitment of paxillin to the tubulobulbar complexes, structures that are believed to have an anchoring function between Sertoli cells and germ cells, just prior to spermiation suggests that paxillin could be functionally involved with the turnover of the actin network and therefore a possible candidate for involvement with sperm release. The coimmunoprecipitation of paxillin and FAK, but the absence of immunostaining for FAK in apical regions of the epithelia, suggests that a direct FAK/paxillin interaction occurs in the basal portions of the epithelium.
The localization of vinculin, a known paxillin binding protein, to both S/S junctions and S/G junctions occurs throughout the spermatogenic stages. Vinculin also occurs at other actin related structures such as the tubulobulbar complexes. These structures are the only site during spermatogenesis, where the codistribution of vinculin and paxillin is obvious by immunofluorescence, thus suggesting a location in the epithelium that could account for their direct interaction. However, even though paxillin is not detectable at junctional sites, the abundance of this molecule throughout the epithelium indicates that a direct interaction may be occurring elsewhere in the epithelium. Because FAK and paxillin are not present at ESs, it is likely that a FAK/paxillin protein complex is not directly associated with control of junctional actin architecture but may be associated with other developmental roles elsewhere in the epithelium.
Hypothetical Model for Junctional Adhesion and Control Molecules
Based on the distribution of ß1-integrin signaling molecules in the seminiferous epithelium and the hypothesis that ESs are structurally maintained and controlled by an integrin mediated pathway, a hypothetical model of construction can be proposed. With the consideration that the S/S and S/G interfaces are different in design we suggest independent models to represent adhesion molecules and junctional components that are likely to be involved in signaling events. Paired transmission electron micrographs with each model depict the ultrastructure of both S/G junctions (Fig. 8a) and S/S junctions (Fig. 8b). At both sites of adhesion ESs consists of an endoplasmic reticulum (ER), a hexagonal array of actin, and a submembranous space. The junctional complex found between adjacent Sertoli cells also contains desmosome-like junctions and tight junctions in close association with the ES actin.
|
In both the S/G and S/S junctional model we have included the transmembrane 6ß1 dimer with integrin cytoplasmic tail extending into the Sertoli cytoplasm. We have also included the presence of a cadherin/ß-catenin complex at S/S junctional complexes but not at S/G junctional sites or in association with the actin component of the basal ES. We have shown the cytoplasmic tail of the integrin subunit to interact, directly or indirectly, with a putative tyrosine kinase which, based on our data, is not FAK. Because vinculin typically does not directly interact with the ß1-integrin but is known to link directly to the actin cytoskeleton, we have diagrammed it downstream of a possible tyrosine kinase and made the speculation that it signals to a member of the Ras family, possibly Rho A. Molecules such as these could cause direct rearrangement of the actin cytoskeleton.
At both the S/S junctions and S/G junctions, we have removed FAK from the ES to indicate that it does not concentrate at junctional sites but, rather, is found to be highly expressed throughout spermatogonia cytoplasm. Paxillin, but not FAK, appears with phosphotyrosine residues. It is also shown to be present in the cytoplasm of both the Sertoli cell and the germ cell but with question at S/S junctions. At S/G junctions, paxillin and vinculin can be seen in close association with the actin network surrounding the tubulobulbar processes and are diagrammed to have a direct interaction. An unidentified transmembrane molecule is indicated opposite to the integrin dimer as a possible ligand of the ß1-integrin. With the known association of the 6ß4 integrin with hemi-desmosome-like junctions, we have included the integrin dimer as a molecule that may be localized to the base of the Sertoli cells. This molecule along with other unidentified transmembrane molecules may function in maintaining adhesion between the seminiferous epithelium and the limiting membrane.
Conclusion
The presence and spatial codistribution of ß1-integrin, ILK, vinculin, and phosphotyrosine at actin-related cell-cell junctions (ESs) in the rat seminiferous epithelium supports the conclusion that integrin-related signaling cascades may be involved with the regulation of ESs and thereby play a significant role in controlling sperm release and turnover of basal junctions between neighboring Sertoli cells. The absence of epitopes reactive with pan-cadherin and ß-catenin antibodies at ESs indicates that conventional cadherin/catenin signaling cascades may not be directly involved with controlling ESs. It is conceivable, however, that integrins and cadherin-related molecules cooperatively contribute toward signaling pathways that are responsible for regulation of the Sertoli-Sertoli junctional complex and thereby allow cadherins to indirectly contribute to ES regulation. The absence of FAK from junctional sites is intriguing and suggests that this molecule may not be directly associated with the control of junction turnover but could assume a different function during spermatogenesis. Our data have provided the first evidence showing the presence of a ß1-integrin/ILK interaction and the absence of conventional cadherins from an in vivo actin related adhesion junction.
FOOTNOTES
1 Correspondence: David J. Mulholland, 2660 Oak Street, Jack Bell Research Centre, Vancouver, BC, Canada V6H 3Z6. FAX: 604 875 5654;djm{at}interchange.ubc.ca 
Accepted: September 5, 2000.
Received: June 13, 2000.
REFERENCES
6;ß1 receptor is localized at specific sites of cell-to-cell contact in rat seminiferous epithelium. Biol Reprod 1995; 52:79–87.[Abstract]6ß1 integrins, and their formation in controlled by follicle-stimulating hormone. Biol Reprod 1998; 58:371–378.6ß1 interact with laminin and an ADAM. J Cell Biol 1999; 144:549–561.This article has been cited by other articles:
|
|
 |
|
  H. H. N. Yan, D. D. Mruk, E. W. P. Wong, W. M. Lee, and C. Y. Cheng An autocrine axis in the testis that coordinates spermiation and blood-testis barrier restructuring during spermatogenesis PNAS, July 1, 2008; 105(26): 8950 - 8955. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 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]
|
|
|
|
 |
|
 T. J Kaitu'u-Lino, P. Sluka, C. F H Foo, and P. G Stanton Claudin-11 expression and localisation is regulated by androgens in rat Sertoli cells in vitro Reproduction, June 1, 2007; 133(6): 1169 - 1179. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Wakayama, Y. Sai, A. Ito, Y. Kato, M. Kurobo, Y. Murakami, E. Nakashima, A. Tsuji, Y. Kitamura, and S. Iseki Heterophilic Binding of the Adhesion Molecules Poliovirus Receptor and Immunoglobulin Superfamily 4A in the Interaction Between Mouse Spermatogenic and Sertoli Cells Biol Reprod, June 1, 2007; 76(6): 1081 - 1090. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Aivatiadou, E. Mattei, M. Ceriani, L. Tilia, and G. Berruti Impaired Fertility and Spermiogenetic Disorders with Loss of Cell Adhesion in Male Mice Expressing an Interfering Rap1 Mutant Mol. Biol. Cell, April 1, 2007; 18(4): 1530 - 1542. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Hyenne, J. C Harf, M. Latz, B. Maro, U. Wolfrum, and M.-C. Simmler Vezatin, a ubiquitous protein of adherens cell-cell junctions, is exclusively expressed in germ cells in mouse testis Reproduction, March 1, 2007; 133(3): 563 - 574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Beardsley, D. M Robertson, and L. O'Donnell A complex containing {alpha}6{beta}1-integrin and phosphorylated focal adhesion kinase between Sertoli cells and elongated spermatids during spermatid release from the seminiferous epithelium. J. Endocrinol., September 1, 2006; 190(3): 759 - 770. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. H. N. Yan and C. Y. Cheng Laminin {alpha} 3 Forms a Complex with beta3 and {gamma}3 Chains That Serves as the Ligand for {alpha} 6beta1-Integrin at the Apical Ectoplasmic Specialization in Adult Rat Testes J. Biol. Chem., June 23, 2006; 281(25): 17286 - 17303. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P Sluka, L O'Donnell, J R Bartles, and P G Stanton FSH regulates the formation of adherens junctions and ectoplasmic specialisations between rat Sertoli cells in vitro and in vivo. J. Endocrinol., May 1, 2006; 189(2): 381 - 395. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. P.Y. Lee, D. D. Mruk, C.-h. Wong, and C. Y. Cheng Regulation of Sertoli-Germ Cell Adherens Junction Dynamics in the Testis Via the Nitric Oxide Synthase (NOS)/cGMP/Protein Kinase G (PRKG)/{beta}-Catenin (CATNB) Signaling Pathway: An In Vitro and In Vivo Study Biol Reprod, September 1, 2005; 73(3): 458 - 471. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Beall, K. Boekelheide, and K. J. Johnson Hybrid GPCR/Cadherin (Celsr) Proteins in Rat Testis Are Expressed With Cell Type Specificity and Exhibit Differential Sertoli Cell-Germ Cell Adhesion Activity J Androl, July 1, 2005; 26(4): 529 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Wolski, C. Perrault, R. Tran-Son-Tay, and D. F. Cameron Strength Measurement of the Sertoli-Spermatid Junctional Complex J Androl, May 1, 2005; 26(3): 354 - 359. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-H. Wong, W. Xia, N. P. Y. Lee, D. D. Mruk, W. M. Lee, and C. Y. Cheng Regulation of Ectoplasmic Specialization Dynamics in the Seminiferous Epithelium by Focal Adhesion-Associated Proteins in Testosterone-Suppressed Rat Testes Endocrinology, March 1, 2005; 146(3): 1192 - 1204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Mruk and C. Y. Cheng Sertoli-Sertoli and Sertoli-Germ Cell Interactions and Their Significance in Germ Cell Movement in the Seminiferous Epithelium during Spermatogenesis Endocr. Rev., October 1, 2004; 25(5): 747 - 806. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. P.Y. Lee and C. Y. Cheng Adaptors, Junction Dynamics, and Spermatogenesis Biol Reprod, August 1, 2004; 71(2): 392 - 404. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K.Y. Siu and C. Y. Cheng Extracellular Matrix: Recent Advances on Its Role in Junction Dynamics in the Seminiferous Epithelium During Spermatogenesis Biol Reprod, August 1, 2004; 71(2): 375 - 391. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. P.Y. Lee and C. Y. Cheng Ectoplasmic specialization, a testis-specific cell-cell actin-based adherens junction type: is this a potential target for male contraceptive development? Hum. Reprod. Update, July 1, 2004; 10(4): 349 - 369. [Abstract] [Full Text] [PDF]
|
|
