| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
Population Council,3 New York, New York, 10021
Department of Zoology,4 University of Hong Kong, Hong Kong, Special Administrative Region of China
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-catenin, p120ctn, c-Src (a putative PTK and the product of the transforming, sarcoma-inducing gene of Rous sarcoma virus), Rab 8 (a GTPase), actin, vimentin, but not E-cadherin, afadin, nectin-3, and integrin ß1, suggesting Fer kinase associates not only with the actin-based cell-cell AJ structures, such as the N-cadherin/catenin complex (but not the 
6ß1 integrin/laminin and the afadin/nectin complex), but also with intermediate filament-based cell-cell desmosomes. An induction in Fer kinase expression was detected during Sertoli-germ cell AJ assembly in vitro but not during AF-2364-induced AJ disruption in vivo. Yet this AF-2364-induced Fer kinase plummeting associated with an induction in N-cadherin, ß-catenin, and p120ctn, particularly at the base of the seminiferous epithelium. In summary, Fer kinase structurally associates with the N-cadherin/catenin protein complex in the testis and can possibly be used to mediate signaling function via the cadherin/catenin protein complex.
kinases, Sertoli cells, signal transducers, signal transduction, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
6ß1 integrin/laminin [8, 9] complexes are largely restricted to the AJ at the site of apical ES, a testis-specific AJ, whereas the cadherin/catenin complex is found mostly at the site of basal ES between Sertoli cells as well as between Sertoli and germ cells [8, 10–12]. Yet the cadherin/catenin complex is also detectable at the site of apical ES by fluorescent microscopy but only at stages I–VII of the epithelial cycle [11, 12].
Recently studies by immunoprecipitation using Sertoli-germ cell cocultures and seminiferous tubules with negligible Leydig and myoid cell contaminations have shown that N-cadherin and its associated ß-catenin and p120ctn that constitute the N-cadherin/catenin complex are linking to actin cytoskeleton, possibly via -catenin and/or -actinin [for reviews, see 1, 4, 13–15]. Furthermore, the dynamics of AJs in the testis are regulated by changes in the phosphorylation status of the AJ structural proteins induced by protein kinases and phosphatases possibly via protein kinases A and C [16]. Taken collectively, these data clearly illustrate the significance of protein kinases, in particular those found at the site of AJs, in the regulation of junction dynamics.
The Fujinami poultry sarcoma/feline sarcoma (fps/fes) proto-oncogene encodes a structurally unique nonreceptor protein tyrosine kinase (PTK) called Fps/Fes kinase [17–19]. Fer kinase (94 kDa) is the only other known member of this distinct subfamily in mammalian cells [20, 21]. In contrast to the tissue-specific expression pattern of Fps/Fes kinase, Fer kinase is an ubiquitously expressed nonreceptor PTK also found in the testis [20]. Fps/Fes and Fer kinases have been implicated in the regulation of cell-cell AJs and cell-matrix focal adhesion interactions in multiple epithelia, and they serve similar and even redundant biological function based on recently completed genetic analysis in mice [for review, see 22]. However, there is also a testis-specific and nucleus-associated FerT transcript (the mature protein of FerT is 
51 kDa vs. the 94-kDa Fer kinase) [23–25], which becomes intensively accumulated in primary spermatocytes at the mid- and late-pachytene stage of meiotic prophase in the mouse [23]. Yet FerT is not associated with either spermatogonia, round spermatids, elongating/elongate spermatids, or Sertoli cells [23] as demonstrated by in situ hybridization.
Subsequent studies have shown that FerT is a meiosis-specific nuclear tyrosine kinase [24]. Structurally, Fer kinase and FerT are virtually identical, sharing >90% homology [25]. For instance, both kinases are having the same SH2 and kinase domains except that the FCH (Fps/Fes/Fer/ClP4 homology) and the three coiled-coil (CC) domains from the N-terminus of Fer kinase are missing from FerT (Fig. 1) [for review, see 22]. Also, FerT has a unique N-terminal sequence of 44 amino acid residues vs. Fer kinase (Fig. 1) [for review, see 22; 23, 25]. Unlike FerT, which is largely restricted to the nucleus of pachytene spermatocytes [23] and a putative nuclear PTK [24], Fer kinase is a cytoplasmic PTK found at the site of AJs and was shown to associate with the phosphorylated form of p120ctn, which in turn interacts with the cadherin/catenin complex [26]. As such, Fer kinase may be an important PTK that mediates signals at the site of AJs to the nucleus in much the same way as p120ctn and ß-catenin. Ironically, both catenins are AJ-peripheral signaling molecules (both p120ctn - and ß-catenin are found in ES [10, 12, 27]), yet they are also detected in the nucleus as trafficking molecules to induce transcriptional regulation of genes [for review, see 28, 29]. For instance, p120ctn and ß-catenin interact specifically with Kaiso and Lef1/TCF transcription factor family, respectively, to induce gene activation [for review, see 29]. As such, it is not unprecedented for an AJ-peripheral signaling molecule to be detected in the nucleus.
|
Fer kinase is an important regulator of cell adhesion function. For instance, overexpression of Fer kinase in embryonic fibroblasts can reduce the level of -catenin and ß-catenin associating with E-cadherin, causing a loss of cell adhesion function [30]. Other studies have shown that Fer kinase can induce phosphorylation of p120ctn and ß-catenin, dissociating them from the cadherin/catenin complex [30], causing a loss of cell adhesion function. Furthermore, Fer kinase mediates cross-talk between cadherin and ß1-integrin. For instance, treatment of epithelial cells with a peptide mimicking the juxtamembrane (JMP) region of the N-cadherin's cytoplasmic domain can induce an inhibition of the N-cadherin and ß1-integrin function [31]. Because the blockade of the cadherin's JMP region by the synthetic peptide causes the release of Fer kinase from the cadherin/catenin complex, this free Fer kinase pool becomes accumulated to the ß1-integrin complex, inducing a loss of the integrin-mediated signaling function [31].
Despite all the study investigating Fer kinase in other epithelia, very little is known regarding its functional role in the testis. Worse, the function of FerT in spermatogenesis is entirely unknown. In light of the recent findings that the cadherin/catenin complex is present side by side with the integrin 6ß1/laminin and afadin/nectin complexes at the site of apical ES [7, 8, 10–12], it is of interest to assess the structural relationship of Fer kinase with these AJ structural proteins in the testis. Yet Fer/FerT-/- homozygous mice lacking both Fer and FerT kinases by targeting the fer locus with a kinase-inactivating missense mutation (ferD743R) develop normally and are still fertile [32], even though the level of phosphorylated cortactin was significantly reduced [32] (cortactin is an AJ-associated actin-binding protein also found in the testis [12]). These results thus suggest that although these two nonreceptor PTKs are important for cell adhesion and other physiological function, their function(s) can be superseded by other PTKs. Yet these data do not negate the significance of Fer kinase in AJ function; instead they illustrate the pivotal role of PTKs in spermatogenesis, that nature has installed multiple PTKs in cells to ensure normal cellular functioning. As such, we thought it pertinent to conduct a careful study to assess the function of Fer kinase in the testis, in particular the structural association of Fer kinase with the known ES structural protein complexes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Male Sprague-Dawley rats were obtained from Charles River (Kingston, NY). The use of animals for the studies reported herein was approved by the Rockefeller University Animal Care and Use Committee, with Protocol no. 00111, 97117, and 95129-R.
Fer Kinase Peptide Synthesis and Polyclonal Antibody Production
A 21-amino acid peptide (NH2-SAPQNCPEEIFTIMMKCWDYK-COOH) corresponding to residues 779–799 and 411–431 from the N-terminus of rat Fer kinase and mouse FerT, respectively, at the catalytic kinase site of the protein (Fig. 1) was synthesized by Genemed Synthesis Inc. (South San Francisco, CA). The sequence of this peptide is identical between Fer kinase and FerT except for one amino acid, indicated in bold, which is Ile788 in rat Fer kinase versus Val420 in mouse FerT [25]. This peptide was subsequently purified by HPLC as described [33–35] and conjugated to keyhole limpet hemocyanin, which was used as an adjuvant, and the protein-peptide complex was injected into rabbits to produce a specific polyclonal antibody against Fer kinase as verified by ELISA. As such, this antibody should cross-react with both Fer kinase and FerT. Yet studies by immunoblotting using lysates from testes in our preliminary experiments had consistently shown that this antibody recognized only a band of 94 kDa, but not 51 kDa, suggesting that this antibody reacted largely with Fer kinase. But because the primer pair used for semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) to estimate the steady-state mRNA level of Fer kinase in cell cultures and testes cannot distinguish between Fer kinase and FerT (see below), both kinases were examined in our RT-PCR experiments.
Antibodies
Primary antibodies against N-cadherin (H-63; Cat: SC-7939, Lot: C081), E-cadherin (H-108; Cat: SC-7870, Lot: K080), -catenin (Cat: SC-7900, Lot: J139), p120ctn (S-19; Cat: SC-1101, Lot: A079), actin (H-196; Cat: SC-7210, Lot: C222), nectin-3 (Cat: SC-14806, Lot: K261), integrin ß1 (Cat: SC-8978, Lot: E221), cSrc (Cat: SC-8056, Lot: C051), and vimentin (V9; Cat: SC-6260, Lot: B252); and bovine antirabbit or goat antimouse IgG-horseradish peroxidase were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against Rab 8 (Cat: R66320-152, Lot: 2) and afadin (Cat: 610732, Lot: 1) were obtained from BD Transduction Laboratories (San Diego, CA).
Primary Sertoli Cell Cultures
Sertoli cells were isolated from 20- [10, 36], 45-, and 90-day-old [37, 38] rats as described. Cells were plated at 0.5 x 106 cells/cm2 on Matrigel (Collaborative Research, Inc., Bedford, MA)-coated 12-well dishes. Cells were incubated in a humidified atmosphere of 95% air/5% CO2 (v/v) at 35°C. At the time of cell plating, cultures were designated time zero. To obtain Sertoli cell cultures with purity greater than 95%, cells were hypotonically treated with 20 mM Tris, pH 7.4, for 2.5 min, 36 h after plating as described [39] to remove contaminating germ cells.
Germ Cell Cultures and Sertoli-Germ Cell Cocultures
Germ cells from 20-, 45-, 60-, and 90-day-old rat testes were isolated by a mechanical procedure as earlier described [40, 41]. These cells were used within 3 h after their isolation. For coculture experiments, Sertoli cells isolated from 20-day-old rat testes were cultured alone for 5 days, allowing them to form a cell epithelium with intact tight junctions (TJs) and AJs before the addition of germ cells isolated from 90-day-old rat testes on Day 6 to initiate AJ assembly [41].
Microdissection of Staged Tubules by Transillumination Microscopy
Testes were isolated from adult rats. Tunica albuginea were removed and the seminiferous tubules were suspended in ice-cold Ham's F-12 Nutrient Mixture/Dulbecco's Modified Eagle's Medium (1:1, v/v). Separation of staged tubules was done under a stereomicroscope (Fisher Scientific, Pittsburgh, PA) at x10 magnification. The wave of the seminiferous tubules was determined as earlier described [42], and tubules were dissected into four zones as follows: pale (stages IX–XII), weak spot (stages XIII–I), dark spot (stages II–VI), and dark (stages VII–VIII) [42], suspended in F12/DMEM and used for RNA extraction and protein lysate preparation.
Treatment of Rats with AF-2364
Adult rats weighing between 280 and 320 g were used. The 1-(2,4-dichlorobenzyl)-indazole-3-carbohydrazide (AF-2364) was suspended in 0.25% methylcellulose (w/v in sterile water) to a final concentration of 20 mg/ml [43, 44]. Rats were fed with a single dose of AF-2364 at 50 mg/kg body weight (b.w.), and rats in groups of three to five were terminated at specified time points [43, 44]. Testes were removed and frozen in liquid nitrogen immediately and stored at -80°C until use.
Kinetics of Germ Cell Loss from the Seminiferous Epithelium Induced by AF-2364
The half-times of germ cell disappearance from the seminiferous epithelium for elongate spermatids (including both elongated and elongating spermatids), round spermatids, and spermatocytes following AF-2364 treatment were estimated as follows. In control rats, elongate (both elongating and elongated) spermatids were restricted to the seminiferous epithelium and were not found in the tubule lumen except at late stage VIII; round spermatids and spermatocytes were also restricted to the seminiferous epithelium and were not found in tubule lumen at all stages of the epithelial cycle (see Fig. 3). Yet after AF-2364 treatment, elongate spermatids rapidly detached from the epithelium at 4–8 h post treatment, to be followed by round spermatids (at 1–4 days post treatment) and spermatocytes (at 4–7 days post treatment), and more of these germ cell types were found in the tubule lumen when time progressed.
|
To assess the kinetics of germ cell loss, 600 tubules were examined at high magnification (x40 or x60 objective), and two parameters were used by scoring rat (n = 3 rats per time point) testes at 4 and 8 h and 1, 2, 4, 7, and 15 days after treatment with AF-2364 at 50 mg/kg b.w. by gavage. First, tubules having progressive loss of specific germ cell types, namely elongate (both elongating and elongated were scored as one group) spermatids, round spermatids, and spermatocytes from the epithelium (see those shown in Figs. 9 and 11) versus control testes were scored. This was done by counting tubules, which had more than 10 elongated/elongating spermatids, round spermatids, or spermatocytes in each tubule lumen versus control tubules (in control sections of testes, virtually no germ cells could be detected in the lumen per tubule unless in late stage VIII where fully developed elongated spermatids were found in tubule lumen during spermiation; see Figs. 3, 9, 11) using the following formula:
|
|
where ST, seminiferous tubule; Ctrl, control rats (600 tubules were scored in both control and each treatment group).
The above parameter was used to score tubules from 4 h and up to Day 1 post-AF-2364 treatment when most of the tubules appeared to be normal except that elongate spermatids were detected (at least 10 spermatids/tubule lumen) in 80% of the tubules on Day 1, and some round spermatids were also detected in the lumen on Day 2 (see Fig. 12 vs. Figs. 9 and 11). Second, between Day 2 and up to Day 15 (see Figs. 9, 11), germ cell layers in the seminiferous epithelium were scored in the treatment group vs. control. This parameter was used to score tubules at these time points because virtually no elongate/elongating spermatids were found in the epithelium on Day 4 and thereafter, as such, it became impossible to track elongated spermatid cell numbers found in the tubule lumen when most round spermatids and spermatocytes were still present yet beginning to deplete from the epithelium. This was done by measuring the number of cell layer, such as round spermatids, in the epithelium versus control. For any tubule that had 30–50% cell layers of round spermatids and spermatocytes (in control testes, between four and eight cell layers of round spermatids and about two to three layers of spermatocytes were found; see Figs. 3, 9, 11) less than that of the control, cell loss was considered significant and was scored (see Figs. 3, 9, 11). These data, percent of tubules with at least 10 germ cells of either elongate (both elongated and elongating spermatids were counted altogether) spermatids, round spermatids, or spermatocytes per tubule lumen (from time 0 to Day 1 after AF-2364 treatment), or cell layers of round spermatids and spermatocytes at 30%–50% less than controls (from Day 1 to Day 15 when elongate spermatids were virtually not found in any tubules in this time frame) were plotted against time after treatment of rats with AF-2364 at 50 mg/kg b.w. by gavage to obtain the half-times of cell loss. A total of 600 tubules (about 200 tubules/testis with a total of three testes from three different rats) were scored using the following formula:
|
Immunohistochemistry
Testis sections (6 µm) were cut using a microtome in a cryostat at -20°C and mounted on poly-L-lysine (Mr > 150 kDa)-coated slides. About six cross-sections of testes at different time points within an experimental group were mounted onto one slide. As such, two to three slides could hold all the testes sections from an entire experimental group. To eliminate interexperimental variations, all slides within an experimental group were processed simultaneously in one experimental session including the primary and secondary antibody incubation and the subsequent color development.
For morphological analysis and immunohistochemistry study, at least 300 cross-sections for each time point from at least three different rats (about 100 cross-sections/testis per rat) were examined. Each experiment was repeated at least three times and representatives of these findings are reported herein. For immunohistochemistry, tissue sections on slides were air dried and fixed in 4% paraformaldehyde (w/v in PBS) (10 mM sodium phosphate, 0.15 M NaCl, pH 7.4 at 22°C) for 20 min. The endogenous peroxidase activity was blocked by treatment with 1% H2O2 (v/v in methanol) for 15 min. Nonspecific antibody-binding sites were blocked by preincubation of sections with 10% normal goat serum (v/v in PBS) (Zymed Laboratories, Inc., San Francisco, CA) at room temperature for 30 min. Sections were then incubated with anti-Fer kinase antibody (1:400) overnight at 4°C in a humidified chamber and subsequently with biotinylated goat antirabbit IgG (1:1600) (Vector Laboratories, Burlingame, CA) (Cat: BA1000, Lot: M0823) for 30 min. After incubation with streptavidin-peroxidase complex (Zymed) for 10 min, sections were incubated with 3, 3'-diaminobenzidine tetrahydrochloride and counterstained with hematoxylin (Zymed). The slides were then dehydrated, mounted, and examined under a microscope.
All antibodies were diluted in 10% normal goat serum (v/v in PBS). Controls were performed in adjacent sections as follows: 1) antibody preabsorbed with Fer kinase peptide, 2) substituting the primary antibody with preimmune serum, 3) substituting the primary antibody with 10% normal goat serum (v/v in PBS), and 4) substituting the secondary antibody with 10% normal goat serum (v/v in PBS). The criteria used for defining stages of the seminiferous tubules were based on the position and morphology of the heads of the elongate spermatids within the seminiferous epithelium [45]. Representative micrographs selected from 100–150 cross-sections were obtained by using a BX40 microscope (Olympus, Olympus America Inc., Melville, NY), with images captured using a built-in QImaging MicroPublisher Cooled digital camera (Olympus) interfaced to a G4 computer (Macintosh, Apple Computer Inc., Cupertino, CA). All experiments were repeated at least three times using testis sections from three different rats for each time point.
Semiquantitative RT-PCR
Semiquantitative RT-PCR was performed using RNA extracted from cells or tissues essentially as previously described [46]. In brief, 2 µg of total RNA was reverse transcribed into cDNAs using 1 µg oligo(dT)15 in a final reaction volume of 25 µl. Three microliters of this reverse transcribed product was used for RT-PCR. The primers used for the amplification of target genes are shown in Table 1. The cycling parameters for PCR were as follows: denaturation at 94°C for 1 min, annealing at 49–63°C (Table 1) for 2 min, and extension at 72°C for 3 min. A total of 21–28 cycles was performed, and the reaction was ended with a 15-min extension at 72°C. To enhance the detection sensitivity, the sense primers were 5' end labeled with [-32P] ATP. Autoradiography was performed using X-OMAT AR films (Kodak, Rochester, NY). Autoradiograms were densitometrically scanned at 600 nm, and data were normalized against S16 or ß-actin for statistical analysis.
|
Immunoblotting
Lysates were obtained from testes using a tissue:lysis buffer ratio of 1:5 with lysis buffer A (0.125 M Tris, 1% SDS [w/v], 1.2% mercaptoethanol [v/v], 1 mM EDTA, 2 mM PMSF, 1 mM N-ethylmaleimide, pH 6.8 at 22°C). Protein concentrations were determined using BSA as a standard [47]. About 70 µg protein was denatured in SDS sample buffer (0.125 M Tris, 20% glycerol [v/v], 1% SDS [w/v], and 1.6% 2-mercaptoethanol [v/v], pH 6.8 at 22°C) and resolved in 7.5% T SDS-polyacrylamide gels under reducing conditions as described [48, 49]. To eliminate interexperimental variation, all samples within an experimental group were processed simultaneously for lysate preparation, SDS-PAGE, and immunoblotting. Following electrophoresis, proteins in gels were electroblotted onto nitrocellulose membrane, and proteins were detected by an enhanced chemiluminescence kit from Amersham Pharmacia Biotech (Piscataway, NJ) as described [10].
Immunoprecipitation
Immunoprecipitation was performed essentially as described [50, 51] with minor modifications. Briefly, 400 µg protein (100–300 µl sample volume) was first pretreated with normal rabbit serum at 1:150 dilution for 4–6 h at room temperature with agitation in a rotating device, and the nonspecific interacting proteins were removed by the addition 5 µl of Protein A/G-PLUS agarose (Santa Cruz Biotechnology). Following a brief centrifugation, the supernatant was then incubated with an anti-Fer kinase antibody at a dilution of 1:150 overnight with agitation. The specificity of this antibody was confirmed by immunoblots and immunoelectrophoresis as earlier described [52, 53]. Thereafter, 20 µl of Protein A/G-PLUS agarose was added to the sample tube and incubated for another 4 h to precipitate the immunocomplexes, which were washed three times by PBS to remove nonspecifically bound proteins. The immunocomplexes bound to the Protein A/G-PLUS agarose were then removed by denaturing the sample in SDS sample buffer, and samples were resolved by SDS-PAGE in 7.5% T SDS polyacrylamide gels. Proteins were electroblotted onto nitrocellulose membrane and immunoblotted with the corresponding antibodies as described [10, 54].
Statistical Analysis
Multiple comparisons were performed using one-way ANOVA followed by Tukey's honest significant different (HSD) test to compare selected pairs of experimental groups using the GB-STAT statistical analysis software package (version 7.0) (Dynamics Microsystems, Inc., Silver Spring, MD). All experiments reported herein were repeated at least three times with different batches of cells. Each time point had triplicate cultures in each experiment. For in vivo studies, each time point had four to six rats per time point.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because the primer pair used for RT-PCR overlapped between Fer kinase and FerT, it is important to note that any changes detected in our studies reported herein for Fer kinase represent Fer kinase/FerT. Yet the antibody apparently cross-reacted with only Fer kinase, FerT was not detected in studies by immunohistochemistry (see Materials and Methods). Fer kinase was found to be predominantly expressed by germ cells versus Sertoli cells in immature rats at 20 days of age (Fig. 2, A and E). During maturation, its steady-state mRNA level in germ cells increased significantly with aging (Fig. 2, B and E). Such an increase in Fer kinase expression was more pronounced during Sertoli cell maturation (Fig. 2, C and E). The steady-state Fer kinase level also increased steadily during maturation of the testis, which peaked at 45–60 days of age but plummented at 90 days of age to a level similar to immature rats (Fig. 2, D and E).
|
Localization of Fer Kinase in Normal Adult Rat Testis
The localization of immunoreactive Fer kinase in the seminiferous epithelium of adult rat testes was shown in Figure 3, A–E, in which Fer kinase appeared as brownish precipitates. Figure 3, A and B, shows the corresponding negative controls stained with either preimmune serum (Fig. 3A) or the anti-Fer kinase antibody preabsorbed with the synthetic Fer kinase peptide (Fig. 3B) by incubating the antibody (1:150) in 0.5 ml sample volume with 10 µg of the synthetic peptide overnight at 4°C before its use for immunostaining. Fer kinase was found to express in a cell type-specific manner in different stages. For instance, in stages I–VIII, immunoreactive Fer kinase was largely confined to round spermatids (Fig. 3, C and D), and virtually no staining could be detected in elongate spermatids (Fig. 3, C and D). Yet Fer kinase became associated largely with spermatocytes and elongating spermatids during stages IX–XIV (Fig. 3, C and E). No signal was detected in spermatogonia or cells in the interstitium (Fig. 3, C–E). This pattern of localization is also consistent with the localization of Fer kinase but not FerT in the epithelium using this antibody because FerT was shown to be restricted to pachytene spermatocytes in mid- and late prophase stage of meiosis in the epithelium [23–25]. Furthermore, this pattern of staining also suggests that Fer kinase associates closely with the nucleus except in developing spermatids (Fig. 3E) when Fer kinase appears to associate with the ES. Perhaps this study needs to be expanded to the ultrastructural level by electron microscopy (EM) to pin-point the precise localization of Fer kinase.
Association of Fer Kinase to AJ-Associated Proteins
Using an anti-Fer kinase antibody for immunoprecipitation using lysates of testes, Sertoli cells, and seminiferous tubules, it was shown that Fer kinase associated with N-cadherin (but not E-cadherin), -catenin, p-120ctn, c-Src, and actin but not nectin-3, afadin, and integrin ß1 (Fig. 4), suggesting its involvement in the N-cadherin/catenin-based AJ dynamics but not the nectin/afadin, integrin/laminin, or E-cadherin-based AJ complexes at the site of ES [1, 6, 8]. Interestingly, Fer kinase was shown to associate with vimentin (Fig. 4), seemingly suggesting that it may also play a role in the dynamics of desmosome-like junctions between Sertoli and germ cells. It also associated with Rab 8, a GTPase (Fig. 4), which was shown to be an important trafficking molecule in regulating Sertoli-germ cell AJ dynamics [55].
|
Changes in the Steady-State mRNA Levels of Fer Kinase/FerT, N-Cadherin, ß-Catenin, and E-Cadherin in Staged Seminiferous Tubules
Tubules were isolated by transillumination stereomicroscopy and divided into four groups. These included the dark zone (stages VII–VIII) (Fig. 5A), pale zone (stages IX–XII) (Fig. 5B), weak spot zone (stages XIII–I) (Fig. 5C), and dark spot zone (stages II–VI) (Fig. 5D). The steady-state Fer kinase mRNA level was detected in all stages, yet its level was the lowest in stages VII–VIII (Fig. 6, A and B). This pattern of Fer kinase expression was consistent with results of the immunohistochemistry shown in Figure 3 that an increase in Fer kinase immunostaining was detected at stages XIII–I (Fig. 6, A and B), a reflection in the increase of Fer kinase associated with elongating spermatids (Fig. 3, C and E). In contrast to Fer kinase/FerT, the steady-state mRNA levels for N-cadherin, ß-catenin, and E-cadherin peaked at stages II–VIII and II–VIII versus VII–VIII, respectively (Fig. 6, A and B). This result is also consistent with recent findings by immunofluorescent microscopy that cadherin is a stage-specific protein being highest at stages I–VII [11]. Taken collectively, these results also illustrate a reciprocal relationship between Fer kinase/FerT and the cadherin/catenin complex, showing that an increase in Fer kinase is associated with a decline in the cadherin/catenin complex.
|
|
Changes in the Steady-State Fer Kinase/FerT mRNA Level During Sertoli-Germ Cell AJ Assembly In Vitro
Germ cells were added onto the Sertoli cell epithelium (Sertoli cells had been cultured alone for 5 days forming an epithelium with intact TJs and AJ) to initiate AJ assembly; it was shown that this event associated with a transient but significant induction of Fer kinase/FerT mRNA at 1–3 h at the time germ cells attached onto the Sertoli cell epithelium (Fig. 7, A and B) [41]. This transient induction was not detected in Sertoli cell cultures without germ cell addition (data not shown), illustrating this event is specific to Sertoli-germ cell AJ assembly. This time frame of Fer kinase induction coincides with the attachment of germ cells onto the Sertoli cell epithelium initiating AJ assembly [41, 56, 57].
|
Changes in the Steady-State Fer Kinase/FerT mRNA and Fer Kinase Protein Levels in Testes During AF-2364-Induced AJ Disruption
Rats were fed with 50 mg AF-2364/kg b.w. to induce germ cell loss from the seminiferous epithelium by perturbing the cell adhesion function between Sertoli and germ cells [43, 44]. It was shown that this event of germ cell loss was accompanied by a reduction of Fer kinase/FerT steady-state mRNA and protein levels, which became visible by Day 1 post treatment. By Day 15, the level of Fer kinase in the treated testes was <10% of the level found in control testes (Fig. 8, A and B). This observation was also consistent with results of immunohistochemistry (Fig. 9, A–F). For instance, elongated spermatids became depleted from the epithelium prematurely in some tubules (Fig. 9B versus Fig. 9A, control) by 8 h post AF-2364 treatment. The tubule marked with an asterisk in Fig. 9B appeared to be a stage V tubule, yet spermatids were found in the tubule lumen. On Day 1, the stage of the cycle was not distinguishable in some tubules and the amount of Fer kinase/FerT declined gradually (Fig. 9C). And by Day 4, most of the elongate spermatids were absent from the epithelium with the appearance of multinucleated cells (Fig. 9D). From Day 7 onward, immunoreactive Fer kinase was virtually undetectable in the seminiferous epithelium, and the tubules were devoid of advanced germ cells with only spermatogonia at the basal compartment of the epithelium (Fig. 9, E and F). Obviously, the plunge of Fer kinase/Fer T in the testis during AF-2364-induced germ cell loss from the epithelium can be attributed exclusively to the loss of round and elongating spermatids from the epithelium because these germ cell types contribute significantly to the pool of Fer kinase/FerT in the testis (see Figs. 2 and 3). Yet it is of interest to note that in some tubules undergoing germ cell depletion, clusters of spermatids displayed intensive staining of Fer kinase (Fig. 9D), which was not seen in normal tubules.
|
Effects of AF-2364 on the Level of N-Cadherin andthe Distribution of N-Cadherin, ß-Catenin, and p120ctn in the Testes during AF-2364-Induced AJ Disassembly
A transient induction on the steady-state mRNA level of N-cadherin was detected in the testis between 4 h and 4 days post AF-2364 treatment (Fig. 10A, upper panel; Fig. 10B). Yet there was a slight delay in the induction of N-cadherin protein (Fig. 10A, lower panel; Fig. 10B), which was also induced but was not detected until Day 4 post- treatment and persisted until Day 15 (Fig. 10A lower versus upper panel; Fig. 10B). Similar results were obtained by immunohistochemistry (Fig. 11, A-H). In control testes, immunoreactive N-cadherin was confined largely to the seminiferous epithelium close to the basal compartment between Sertoli cells and spermatogonia/spermatocytes at the base of the seminiferous tubules (Fig. 11B), which is consistent with our earlier studies by immunofluorescent microscopy [10] with very weak staining at the adluminal compartment. Figure 11A is the negative control in which the primary antibody was substituted with normal rabbit serum.
|
There was a progressive increase in N-cadherin immunostaining from 4 h to 15 days posttreatment (Fig. 11, C–H versus Fig. 11B). Two days after AF-2364 treatment, some immunoreactive N-cadherin was detected surrounding the round and elongating spermatids in the lumen of the seminiferous tubules (Fig. 11E). Also, the basal staining intensified accordingly (Fig. 11, E and F). The accumulation of N-cadherin became very intense near the basal lamina by Day 7 and Day 15, forming an almost continuous immunoreactive arc at or near the basal compartment of the seminiferous epithelium when most of the elongated and round spermatids were depleted from the epithelium (Fig. 11, G and H), apparently in an effort used by the seminiferous epithelium to retain the spermatocytes and spermatogonia in the epithelium. When cross-sections of rat testes following AF-2364 treatment were stained for ß-catenin and p120ctn, a progressive accumulation of immunoreactive ß-catenin and p120ctn in the basal seminiferous epithelium, similar to N-cadherin as shown in Figure 11, was detected when spermatids were detaching from the epithelium (data not shown). On Day 2, some immunoreactive ß-catenin and p120ctn, similar to N-cadherin (Fig. 11), was also found surrounding the round and elongating spermatids in the lumen of the tubules (data not shown). By Days 7 and 15 when virtually no spermatids and a few spermatocytes were found in tubules, both ß-catenin and p120ctn were confined largely to the basal compartment of the epithelium (data not shown).
Kinetics of Germ Cell Loss from the Seminiferous Epithelium of Adult Rat Testes after AF-2364 Treatment
Adult rats were fed with a single dose of 50 mg AF-2364/kg b.w. by gavage; this treatment induced germ cell loss from the seminiferous epithelium because by Days 7–15, >95% of the tubules were devoid of spermatids and spermatocytes [43, 44]. Yet the kinetics of germ cell loss, in particular elongate and round spermatids, are not known. To obtain this information, about 600 tubules from 100 different cross-sections from three different rats (i.e., 200 tubules/testis per rat) were stained by hematoxylin and photographed at high magnification; all stages of the cycle were selected randomly. It must be noted that within 24 h after treatment, it became somewhat difficult to distinguish stages of the tubules as shown in Figures 9 and 11 because elongate and some round spermatids detached from the epithelium, and ended up in the tubule lumen. The number of elongated/elongating spermatids, round spermatids, and spermatocytes in the tubule lumen of each tubule in the treatment group as well as the cell layers for round spermatids and spermatocytes in the epithelium were scored and compared with control testes using two different formulae shown in Materials and Methods. It was noted that elongated spermatids began to deplete from the seminiferous epithelium as early as 4 h post-treatment (Fig. 12). But significant loss of round spermatids was not detected until Day 1 posttreatment (Fig. 12). And it took at least 4–7 days before the loss of spermatocytes became visible (Fig. 12). On Day 4, many multinucleated cells appeared in the tubules, and virtually no normal tubule was found (Fig. 9D). On Day 15, most tubules were almost empty except for spermatogonia and Sertoli cells (Fig. 9F). The half-time for depletion of elongate spermatids from the epithelium was estimated to be 6.5 h versus 3 days and 6.5 days for round spermatids and spermatocytes, respectively (Fig. 12). These results seemingly suggest that the ES, between Sertoli cells and elongate spermatids, is the primary target of AF-2364 and most susceptible to this treatment, yet the AJ and/or desmosome-like structures between spermatocytes and Sertoli cells can also be affected, but they apparently require a longer period of exposure to AF-2364.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
PTKs are involved in cell cycle regulation, signal transduction, cell-cell interactions, and cell differentiation [for reviews, see 22, 58–60]. More recent studies using specific inhibitors have shown that the interplay of PTKs and protein tyrosine phosphatases (PTPs) that determine the intracellular phosphoprotein content are also crucial to the regulation of Sertoli cell TJ dynamics [16, 33]. For instance, it was shown that Sertoli cell TJ dynamics, at least in vitro, are regulated by protein kinases A and C [16]. Furthermore, both Sertoli and germ cells express myotubularin and produce the protein encoding by this mRNA, which is also a putative PTP [16, 33, 38]. To expand these earlier studies, we selected two closely related nonreceptor PTKs for our investigation, namely Fer kinase and FerT. It was shown that during the assembly of Sertoli-germ cell AJs in vitro, this event indeed was associated with an induction of Fer kinase/FerT steady-state mRNA level, seemingly suggesting that these kinases are involved in AJ assembly. Other recently completed studies have shown that the assembly of Sertoli-germ cell AJs in vitro also associates with a transient induction of E-cadherin, N-cadherin, ß-catenin, and p120ctn [10]. Because of such physical intimacy between Fer kinase and catenins as demonstrated by immunoprecipitation study as shown herein, it is not entirely unexpected that an induction in one member of this multiprotein complex induces the other constituent proteins as well. It is likely that an increase in Fer kinase in this multiprotein complex can activate p120ctn, which when tyrosine phosphorylated becomes a positive regulator of the cadherin/catenin-based AJ structures [61, 62].
Fer kinase may also regulate tight junction (TJ) dynamics by interacting with cortactin [32, 63, 64], an actin-binding protein, to induce changes in the cytoskeleton network. Perhaps these studies could be expanded and validated when a specific Fer kinase inhibitor is available to assess its downstream signaling pathway. Certainly one may argue that the coculture system used in our study is indeed a suitable model to study Sertoli-germ cell AJ assembly. To address this issue, we offer the following explanations. First, earlier studies have shown that when Sertoli cells were cocultured with germ cells in vitro, desmosome-like junctions were detected between these cells by electron microscopy within 24–48 h [57]. Subsequent studies confirmed this earlier observation that ES-like AJ structures were also detected in these cocultures in vitro within 2 days [56]. We have since developed an in vitro cell adhesion assay to study the biochemical and molecular events of Sertoli-germ cell AJ assembly and have characterized this system in our laboratory [41, 55, 65]. More recent studies by fluorescent microscopy have shown that both N-cadherin and ß-catenin colocalized to the same sites when these cells were cultured in vitro [10]. Second, cocultures of Sertoli and germ cells in vitro can also induce a transient surge in vinculin [54], a putative ES-protein [for reviews, see 1, 66], suggesting these cells are indeed assembling ES structures.
Collectively, these data illustrate that this is a novel in vitro model to study the biology of Sertoli-germ cell AJ assembly. Based on the results of immunoprecipitation studies reported herein regarding the association of Fer kinase with other known AJ structural proteins at the site of ES, we now propose a revised molecular model depicting the three AJ structural protein complexes in the testis (Fig. 1). In this context it is of interest to note that although a surge in Fer kinase during Sertoli-germ cell AJ assembly was detected similar to N-cadherin, ß-catenin and p120ctn [10], a time-dependent decline in Fer kinase was detected in the epithelium during AJ disassembly, unlike these other AJ proteins, which were also induced. An immediate explanation is not known, but it is possible that reduced phosphorylation of the AJ structural proteins may be needed to perturb cell adhesion function at the site of ES, as such a plunge in Fer kinase was detected during AF-2364-induced AJ disruption.
Fer Kinase Associates with Cell-Cell Actin-Based AJs and Intermediate Filament-Based Desmosome-Like Junction in the Testis
It has been known for almost two decades that cell adhesion function in the testis anchoring germ cells, particularly spermatocytes and round spermatids, onto Sertoli cells is contributed largely by desmosome-like junction, which is an intermediate filament-based cell-cell anchoring (adhering) junction type [5, 6]. Yet the biochemical composition and molecular constituent of this junction type in the testis remains to be explored. In other epithelia, desmosome junctions are constituted by desmocollins, desmogleins, desmoplakin, plakophilin, and plakoglobin [for review, see 1]. With the exception of plakoglobin, none of the above-mentioned desmosome-associated proteins has been identified or studied at the biochemical and molecular level in the testis. Once the cadherin/catenin complex (the cadherin/catenin complex is a major component of AJ, which is an actin-based cell-cell anchoring junction type) was identified in the testis during the 1990s, several studies were conducted to investigate whether the cadherin/catenin complex was being used in the testis to constitute the desmosome-like junctions [for reviews, see 1, 4]. Studies by immunofluorescent microscopy and immunogold EM have shown the association of cadherin/catenin complex with the intermediate filament bundles [8, 11]. Yet studies by fluorescent microscopy have positively identified cadherin at the site of apical ES at the adluminal edge of the tubule between elongate spermatids and Sertoli cells, consistent with its localization with ES [11, 12], which is a putative actin-based AJ type [for reviews, see 1, 6, 66].
Taken collectively, it remains unsettled whether cadherin/catenin is being used for the construction of AJs, such as ES, or desmosome-like junctions in the testis. A recent study by immunoprecipitation with and without the use of a membrane permeable cross-linker demonstrated that the cadherin/catenin complex is indeed an actin-based AJ structure [10], similar to all other epithelia. Still, this recent study has not excluded the possibility that other cadherins (note: at least 24 cadherins have been identified in the testis [67]) may be using the intermediate filament bundle as attachment sites. Also, a small percentage of the N-cadherin/catenin or E-cadherin/catenin complexes can indeed attach to vimentin-based intermediate filament such as via their interactions with Fer kinase. As reported herein, Fer kinase was shown to associate with the actin-based N-cadherin/catenin but not the E-cadherin/catenin complex, yet Fer kinase was also associated with actin and vimentin, further strengthening the possibility that at least some cadherin/catenin complexes indeed could use intermediate filament bundle as their attachment site.
This result also seemingly suggests that Fer kinase regulates only the N-cadherin/catenin complex but not the E-cadherin/catenin complex. This is not entirely unexpected and unprecedented because the three ES structural units are not uniformly distributed among Sertoli and germ cells; as such, they can indeed be regulated differentially. For instance, although the relative N-cadherin ratio in Sertoli:germ cells is 1:1, E-cadherin is more predominant in germ cells with a Sertoli:germ ratio of 1:3 both at the protein and steady-state mRNA level [10]. Also, although nectin-2 (nectin/afadin complex is one of the three known ES structural units in the testis; see Fig. 1) is found in both Sertoli and germ cells, nectin-3 is restricted only to germ cells, permitting homotypic and heterotypic interactions between Sertoli and germ cells as nectin-2:nectin-2 and nectin-2:nectin-3 [7], implicating the nectin-2/afadin and nectin-3/afadin complex can be regulated differentially, similar to N-cadherin/catenin and E-cadherin/catenin complex. Furthermore, the 6ß1 integrin subunit resides exclusively in Sertoli cells [for review, see 66], and the binding partner in germ cells is not even known except that a recent study has implicated that laminin 3 [9] is possibly one of the three chains that constitutes the binding partner for 6ß1 integrin. In this connection, it is of interest to note that Fer kinase also associates with vimentin, an intermediate filament marker protein [for reviews, see 68–70], suggesting it is also involved in desmosome dynamics (Fig. 1). Furthermore, the demonstration of Fer kinase's association with Rab8 GTPase is of interest because a recent study has shown that Rab8B is a crucial regulator of AJ dynamics in the testis [55]. Other studies have shown that GTPases are crucial molecules that also determine the recycling of AJ structural proteins, such as E-cadherin [for review, see 71]. Nonetheless, it remains to be investigated regarding the precise biochemical interactions between PTK and Rab 8B and the physiological consequences thereafter.
Participation of Fer Kinase in Spermatogenesis and Intracellular Trafficking: Is Fer Kinase a Cytosolic or Nuclear PTK or Both?
Studies by immunohistochemistry and in situ hybridization have shown that FerT, a nuclear PTK, is restricted to primary spermatocytes at the mid- and late pachytene stages of meiotic prophase in the seminiferous epithelium of the rat testis [23, 24]. Using an antibody that reacted almost exclusively with Fer kinase, we have shown that Fer kinase displays a drastically different staining pattern in the seminiferous epithelium vs. FerT. For one, although staining of Fer kinase is found in pachytene spermatocytes in stages XIII–XIV of the epithelial cycle, Fer kinase associates largely with round spermatids in the seminiferous epithelium at stages I–VIII of the epithelial cycle and also with elongating spermatids (but not elongate spermatids at stages VII–VIII) in stages X–XIV. For another, Fer T kinase is a stage-specific nonreceptor PTK, being highest at stages IX–I vs. FerT, which is found in mid- and late pachytene spermatocytes of all stages in the epithelium with the weakest staining at stages VII–VIII at the time of or near spermiation [23]. Although Fer kinase associates with N-cadherin and -catenin, its stage specificity is different from both N-cadherin and catenin. For instance, unlike Fer kinase, which is high in stages IX–I, N-cadherin/ß-catenin and E-cadherin peak at stages II-VIII and VII-VIII, respectively. For N-cadherin, it is localized to the basal compartment of the epithelium in all stages and to the heads of elongate spermatids in stages I-VII at the site of ES [11]. ß-Catenin localizes largely to the basal epithelium in all stages [8]. Taken collectively, these results seemingly suggest that FerT, the testis-specific Fer kinase that is restricted to pachytene spermatocytes, is pertinent to the regulation of early spermatocyte development during meiosis, yet Fer kinase is involved in the downstream development from late spermatocytes to early spermatids.
Fps/Fes nonreceptor tyrosine kinase is the only other member of the Fer kinase subfamily [for review, see 22], and both Fps/Fes and Fer kinases have been shown to be involved in the vesicular trafficking, colocalization with several Rab GTPase proteins [72]. Furthermore, Rab 8, a crucial GTPase that mediates intracellular organelle transport and secretion [73] in other epithelia, has recently been shown to be a putative product of both Sertoli and germ cells, possibly involved in AJ dynamics [55]. Because Fer kinase/FerT share similar structural features with Fps/Fes tyrosine kinase [for review, see 22], we anticipate Fer kinase also associates with Rab8. Indeed, studies by immunoprecipitation clearly illustrate Fer kinase is forming a multiprotein complex associated with N-cadherin, -catenin, and Rab8 in addition to its association with p120ctn, suggesting the multiprotein complex can also regulate intracellular trafficking in addition to its role in AJ dynamics. Taking these reports collectively, it is not unprecedented regarding the unusual distribution pattern of Fer kinase in the seminiferous epithelium, in particular, that this PTK also associates with the nucleus instead of being restricted to the site of AJs. For instance, both ß-catenin and p120ctn are putative AJ-associated signaling molecules found at the site of ES [10, 12, 27, 74], yet these two catenins are also known signal transducers executing transcriptional regulation of genes in the testis and are also found in the nucleus [for reviews, see 28, 29]. Given the fact that Fer kinase indeed physically associates with both p120ctn and -catenin at the site of AJs, we postulate that Fer kinase is a cytosolic nonreceptor PTK, yet it can transduce signals to the nucleus in much the same way as p120ctn and ß-catenin.
Effects of AF-2364 on the AJ Dynamics in the Testis: Is the Cadherin/Catenin/p120ctn/Fer Kinase Multiprotein Complex the Primary Target of AF-2364?
Unlike N-cadherin and its associated structural proteins, such as testin, which became greatly induced during AF-2364-induced germ cell loss [43, 44] from the seminiferous epithelium, Fer kinase/FerT was not induced. This can largely be attributed to the fact that germ cells contribute significantly to the pool of this nonreceptor PTK in the epithelium behind the blood-testis barrier. These data are also consistent with results of immunohistochemistry, illustrating that the AF-2364-induced germ cell loss from the epithelium is associated with a drastic decline of immunoreactive Fer kinase. Yet some intense staining of Fer kinase associated with spermatids was detected in the epithelium at 4 days post AF-2364 treatment when round spermatids began to deplete from the epithelium (see Fig. 9D vs. Fig. 12). Taken collectively, these data seemingly suggest that Fer kinase is not involved in Sertoli-germ cell AJ disassembly. Fer kinase apparently contributes to the normal functioning of the cadherin/catenin complex during AJ assembly by phosphorylating p120ctn and PTP1B, which in turn dephosphorylates ß-catenin [for review, see 22], leading to an increase in cell adhesion. Yet the disruption of AJs leads to a loss of Fer kinase, rather than an induction, possibly because of the loss of germ cells, which are the major contributor of this nonreceptor PTK in the epithelium.
In this context, it is of interest to note that a significant induction in N-cadherin, ß-catenin, and p120ctn in the seminiferous epithelium was observed during germ cell loss when the level of Fer kinase declined, yet it is known that germ cells contribute at least half of these three proteins in the seminiferous epithelium [10]. Such differential responses of ß-catenin and p120ctn versus Fer kinase may indeed be a physiological response of the seminiferous epithelium to the AF-2364 treatment, sensing the loss of spermatids; the increased cadherin/catenin complexes are being used to retain spermatocytes and most spermatogonia in the tubules. This is why we detected an intensive staining of N-cadherin near the basal compartment between Days 7 and 15 when N-cadherin formed an almost uninterrupted immunoreactive ring surrounding the entire tubule.
Taking these results collectively, it is apparent that the role of Fer kinase/FerT in the testis associates with the AJ assembly but not its disassembly. As such, an unavoidable question arises and should be addressed: Is the cadherin/catenin/p120ctn/Fer kinase multiprotein AJ complex a primary target of AF-2364 in the testis? Based on the available data, such a possibility is unlikely. First, if the cadherin/catenin/p120ctn/Fer kinase multiprotein complex is the primary target AJ structure of AF-2364, an administration of AF-2364 would induce extensive damage to other organs of the treated animals because the cadherin/catenin complex is almost universal in all epithelia [for reviews, see 1, 3]. Yet recently completed toxicity studies and serum microchemistry analysis have shown that this compound, at doses effective to induce reversible male infertility, is neither nephrotoxic nor hepatotoxic [1, 43]. For instance, both acute toxicity studies to assess the toxicity of AF-2364 in rats and mice, and mutagenicity study performed by licensed toxicologists according to Food and Drug Administration guidelines have shown that AF-2364 is safe for further development as a male contraceptive (unpublished observations). Furthermore, histological analysis of kidney, liver, prostate, brain, seminal vesicle, and epididymis reveal no alterations of gross morphology in which cadherin/catenin complexes have been positively identified [43]. Also, studies by semiquantitative RT-PCR and protein immunoblotting following AF-2364 treatment revealed no induction (or reduction) of E-cadherin, N-cadherin, and ß-catenin in kidney, liver, and brain (Mruk and Cheng, unpublished observations).
A recent study has shown that the 6ß1 integrin at the site of ES may be the initial target of AF-2364 [54]. The fact that Fer kinase/FerT can induce cross-talks between the cadherin/catenin complex and integrin [31, 75, 76] seemingly suggests that the presence of AF-2364 somehow activates the integrin/laminin complex at the site of ES [
