| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
3 Protein Complex in the Regulation of Ectoplasmic Specialization Dynamics in the Rat Testis1
Population Council, Center for Biomedical Research, New York, New York 10021
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
6ß1 integrin in the ES, immunofluorescent microscopy coupled with immunoprecipitation techniques were used to demonstrate that laminin 
3, largely a germ cell product, was present at the apical ES and could form a bona fide complex with ß1-integrin. Furthermore, the structural interactions of MMP-2 and MT1-MMP with laminin 3 and ß1-integrin, but not with N-cadherin or nectin-3, have implicated the crucial role of MMP-2/MT1-MMP in the regulation of integrin/laminin-based ES dynamics. Using an in vivo model to study AJ dynamics where adult rats were treated with 1-(2,4-dichlorobenzyl)-indazole-3-carbohydrazide (AF-2364) to disrupt Sertoli-germ cell adhesive function, an induction of active MMP-2, active MT1-MMP and TIMP-2 but not active MMP-9 was detected between 0.5 and 8 h after AF-2364 treatment. This time frame coincided with the depletion of elongating/elongate spermatids from the epithelium, illustrating the synergistic relationships between MMP-2, MT1-MMP, and TIMP-2 in AJ disassembly. Perhaps the most important of all, the use of a specific MMP-2 and MMP-9 inhibitor, (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid, could effectively delay the AF-2364-induced elongating/elongate spermatid loss from the epithelium, demonstrating the pivotal role of MMP-2 activation in ES disassembly. Collectively, these studies illustrate that the ß1-integrin/laminin 3 complex is a putative ES-structural protein complex, which is regulated, at least in part, by the activation of MMP-2 involving MT1-MMP and TIMP-2 at the apical ES. The net result of this interaction likely regulates germ cell movement in the seminiferous epithelium.
Sertoli cells, signal transduction, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-macroglobulin, tissue inhibitor of metalloproteases-1 (TIMP-1), and cystatin C [4, 5, 7]; in junction-restructuring events during spermatogenesis. More recent studies have also revealed the participation of matrix metalloprotease-9 (MMP-9) and TIMP-1 in the regulation of Sertoli cell tight junction (TJ) dynamics, possibly via their intricate interactions with collagen 3(IV) in the basement membrane regulated by tumor necrosis factor- [8]. As such, it is our belief that proteases and protease inhibitors are involved in the events pertinent to germ cell movement in the seminiferous epithelium during spermatogenesis (for reviews, see [9–11]), yet the precise mechanism by which these events are regulated remains largely unexplored. MMPs are a family of secretory enzymes with more than 25 members (some of which are transmembrane proteins) that degrade virtually all extracellular matrix (ECM) proteins (for reviews, see [12, 13]). Other proteases, protease inhibitors, latent growth factors, growth factor-binding proteins, cell adhesion molecules, and cell-surface receptors are also putative substrates of MMPs (for reviews, see [12, 14, 15]). Moreover, each MMP has distinct but often overlapping substrates. As such, MMPs are pivotal to numerous biological processes. These include embryonic development, morphogenesis, tissue remodeling, and tumor cell invasion (for reviews, see [12, 13–15]). Most MMPs are initially synthesized and secreted as latent/inactive zymogens. Extracellular activation is required to unleash their intrinsic proteolytic activity (for reviews, see [12, 13]). For instance, both pro and active forms of MMP-2 have been detected in Sertoli cell cultures, and its activation is regulated by FSH [16–19]. An earlier study has suggested that the activation of MMP-2 in the testis requires membrane-type 1 (MT1)-MMP and TIMP-2 [17]. MT1-MMP is characterized by the presence of a transmembrane domain for membrane localization [20]. The binding of TIMP-2 to MT1-MMP can form a complex that acts as a receptor for pro MMP-2, which is needed for MMP-2's activation (for reviews, see [12, 13–15]). Although MT1-MMP has been detected at the apical compartment of the seminiferous epithelium [17], it remains to be determined whether MMP-2 and TIMP-2 are indeed localized to the same site as MT1-MMP. Also, the precise mechanism for the activation of MMP-2 in the testis remains to be elucidated.
At the cell-cell contact sites, the extracellular space is filled with an array of macromolecules, largely composed of glycoproteins and polysaccharides, which in turn constitute the ECM (for a review, see [21]). In the testis, the basement membrane is a modified form of ECM (for reviews, see [21, 22]). For instance, the thin sheetlike basement membrane in rodent testes is largely composed of laminin, type IV collagen, heparan sulfate proteoglycan, and entactin; all are ECM proteins, which are in direct contact with the base of Sertoli cells and spermatogonia, surrounding the entire seminiferous tubule (for a review, see [22]). It is of interest to note that ECM, a known structural barrier, undergoes progressive disruption and remodeling in multiple epithelia (for a review, see [12]). This is reminiscent of the cell-cell junction-restructuring events that take place in the seminiferous epithelium pertinent to germ cell movement during the epithelial cycle. This poses an interesting question: Does AJ restructuring, such as at the site of the ES, utilize a similar mechanism(s) to one that is found at the cell-matrix interface, i.e., focal contacts, for its regulation? While most ECM proteins identified to date in the testis are confined to the basement membrane [22, 23], recent studies have shown that the testis uses cell-matrix focal contact proteins for ES regulation [8, 24, 25]. For instance, laminin 3 is a non-basement-membrane-associated laminin chain at the adluminal compartment in the seminiferous epithelium [23]. Although the laminin 3 chain was proposed to be the ligand for 6ß1 integrin at the site of apical ES [1, 8], it remains to be shown if germ, but not Sertoli, cells indeed express and produce laminin 3. On the other hand, additional studies have shown that most of the ES-associated proteins in the seminiferous epithelium, such as integrin-linked kinase (ILK), focal adhesion kinase (FAK), vinculin, and others, are components of the focal adhesion complex found in other epithelia [8, 25]. Thus, it is possible that the proteolytic events that regulate ECM homeostasis also operate at the site of apical ES and regulate ES dynamics in the seminiferous epithelium. In this article, we report findings that illustrate apical ES dynamics are regulated by the intricate interactions between MMP-2, MT1-MMP, TIMP-2. and laminin 3.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sprague-Dawley rats were obtained from Charles River Laboratories (Kingston, MA). The use of rats for studies reported herein was approved by The Rockefeller University Animal Care and Use Committee (protocol numbers 00111 and 03017).
Sertoli Cell Cultures
Sertoli cells were isolated from testes of 20-day-old rats as described [26, 27]. Cells were plated at high density (1 x 106 cells/cm2) on Matrigel (Collaborative Biochemical Products, Bedford, MA)-coated 12-well dishes in 3 ml Ham F12 Nutrient Mixture and Dulbecco modified Eagle medium (F12/DMEM, 1:1, v/v) per well supplemented with different factors as described [27–29]. At this cell density, all three types of junctions, including TJs, AJs, and gap junctions, were formed. Cultures were incubated in a humidified atmosphere of 95% air and 5% CO2 (v/v) at 35°C. Time 0 represents Sertoli cell cultures that were terminated within 
3 h after plating. After 48 h of incubation, cultures were hypotonically treated with 20 mM Tris (pH 7.4 at 22°C) for 2.5 min to lyse residual germ cells [30], to be followed by two successive washes with F12/DMEM to remove cell debris. Media were replaced every 24 h. The purity of these Sertoli cell cultures was routinely analyzed by electron [31] and light microscopy [4, 7] and by reverse transcription-polymerase chain reaction using primer pairs to specific cell markers of Sertoli, germ, and myoid cells as described [32]. For the analysis of cell purity by light microscopy, Sertoli cells were first fixed in OsO4, embedded in Polybed, and 1-µm sections were obtained with a microtome and stained with 1% toluidine blue [4]. To obtain cell lysates, media were removed and cells in each dish were treated at specific time points using 1 ml SDS lysis buffer (0.125 M Tris, pH 6.8, at 22°C containing 1% SDS [w/v], 2 mM EDTA, 2 mM N-ethylmaleimide, 2 mM phenylmethylsulfonyl fluoride [PMSF], 1.6% 2-mercaptoethanol [v/v], 1 mM sodium orthovanadate, and 0.1 µM sodium okadate). The cell suspension was transferred to a microfuge tube, vortexed, incubated for 5–10 min to allow solubilization. Thereafter, samples were centrifuged at 15 000 x g to remove cellular debris, and the clear supernatant was used as cell lysate.
Germ Cell Isolation
Germ cells were isolated from 20- and 60-day-old rats by a mechanical procedure [33] except that elongate spermatids were not removed by omitting the glass wool filtration step. Germ cell preparations had a purity >95% with negligible somatic cell contamination when examined microscopically and assessed by other criteria [32]. Germ cells were lysed in SDS lysis buffer and used for immunoblotting experiments.
Seminiferous Tubule Cultures
Seminiferous tubules were isolated from testes of adult rats (300 g body weight [BW]) with negligible Leydig cell contamination [8, 32, 34, 35]. Thereafter, lysates were obtained by treating tubules with an immunoprecipitation (IP) buffer (0.125 M Tris, pH 6.8, at 22°C containing 1% NP-40 [v/v], 2 mM EDTA, 2 mM N-ethylmaleimide, 2 mM PMSF, 1 mM sodium orthovanadate, and 0.1 µM sodium okadate) using a buffer:tissue ratio of 3:1, vortexed, sonicated, and centrifuged at 15 000 x g for 20 min. The clear supernatant was used as tubule lysate.
Treatment of Rats with AF-2364 to Induce Germ Cell Loss from the Seminiferous Epithelium
AF-2364 was synthesized as described with a purity of greater than 99.8% when assessed by elemental analysis, nuclear magnetic resonance, high-performance liquid chromatography, and mass spectrometry as described [36]. This compound was shown to perturb Sertoli-germ cell adhesion function in the testis, inducing premature loss of germ cells from the seminiferous epithelium, causing reversible infertility in rats [36, 37]. One of the apparent targets of AF-2364 in the seminiferous epithelium is apical ES (for a review, see [1]). Adult rats (250–300 g BW) were fed one dose of AF-2364 at 50 mg/kg BW. The time at which rats were fed with AF-2364 was designated time 0 (control). Thereafter, testes were removed at specified time points from a group of three rats per time point. For immunoblotting, testes were homogenized using a Tissumizer (Tekmar, Cincinnati, OH) for lysate preparation either with SDS lysis buffer (for immunoblotting) or IP buffer (for IP). For immunohistochemistry, testes were immediately frozen in liquid nitrogen and stored at -80°C (for frozen sections) or fixed in 4% paraformaldehyde, v/v, in PBS (10 mM sodium phosphate, 0.5 M NaCl, pH 7.4 at 22°C) (for paraffin sections).
Immunoblotting
Protein concentration was estimated by Coomassie blue dye-binding assay using BSA as a standard [38]. Equal amounts of protein (100 µg per lane) were resolved onto 7.5% or 12.5% T SDS-polyacrylamide gels by SDS-PAGE under reducing conditions [39] and electroblotted onto nitrocellulose papers. For immunoblotting, the following primary antibodies were used: rabbit anti-MMP-2 (cat. #AB809, lot #21082339, which recognized both the 68-kDa pro form and the 64-kDa active form), rabbit anti-TIMP-2 (cat. #AB801-50, lot #22031406) and rabbit anti-MMP-9 (cat. #AB805, lot #21100763, which identified both the 92-kDa pro form and the 84-kDa active form) were from Chemicon (Temecula, CA). Rabbit anti-MT1-MMP (cat. #M3927, lot #072K1219, which recognized the 63-kDa pro form, the 60-kDa active form, and the 45-kDa cleaved form) was from Sigma (St. Louis, MO). Goat anti-laminin 3 (cat. #sc-16601, lot #B282), goat anti-actin (cat. #sc-7210, lot #C222), and goat anti-nectin-3 (cat. #sc-14806, lot #K261) were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Mouse anti-ß1-integrin (cat. #610468, lot #7) was from Transduction Laboratories (Lexington, KY). Mouse anti-N-cadherin (cat. #33-3900, lot #21073914) was from Zymed (Burlingame, CA). All these antibodies cross-reacted with the corresponding rat target proteins as indicated by the manufacturers. After the primary antibody incubation, membranes were incubated with one of the following secondary antibodies: goat anti-rabbit IgG-horseradish peroxidase (HRP) (cat. #sc-2004), goat anti-mouse IgG-HRP (cat. #sc-2005), or donkey anti-goat IgG-HRP (cat. #sc-2004) (Santa Cruz). Target proteins in the blots were visualized using an enhanced chemiluminescence kit (Amersham Pharmacia Biotech, Piscataway, NJ).
Immunoprecipitation and Electron Microscopy
IP was performed essentially as previously described [8, 31, 32]. Lysates of testes and/or seminiferous tubules without incubation with any antibodies or incubated with normal serum served as controls. Electron microscopy was performed at The Rockefeller University Bio-Imaging Resource Center to examine ES in the seminiferous epithelium using procedures as described earlier [31].
Immunohistochemistry
Immunohistochemistry was performed to localize MMP-2, MT1-MMP, and TIMP-2 in the seminiferous epithelium in normal and AF-2364-treated rat testes as described [8, 40] using Histostain SP kits (cat. #95-6143) from Zymed. To prepare frozen sections for MMP-2 immunostaining, testes were removed from rats, immediately frozen in liquid nitrogen, and embedded in O.C.T. embedding compound (Miles Scientific, Elkhart, IN). All tissue blocks were stored at -80°C until use. Frozen sections (6 µm thick) were cut at -20°C in a cryostat and mounted on poly-L-lysine (Sigma, >150 kDa)-coated slides. Sections were air dried at room temperature and then fixed in modified Bouin fixative (4% formaldehyde in picric acid) for 5 min and washed thoroughly with PBS (10 mM sodium phosphate, 0.15 M NaCl, pH 7.4, at 22°C). For the staining of MT1-MMP and TIMP-2, paraffin sections were used because preliminary experiments using frozen sections yielded unsatisfactory results. In brief, testes were fixed in 4% paraformaldehyde in PBS for 24 h, dehydrated, embedded in paraffin, and sectioned to 8 µm. After deparaffinization and rehydration, antigen unmasking was performed by heating sections in 10 mM sodium citrate buffer (pH 6.0) for 10 min at a temperature just below boiling. Sections were then cooled for 20 min. Streptavidin-biotin-peroxidase immunostaining was performed as follows. Briefly, fixed sections were treated with 3% hydrogen peroxide in methanol (v/v) for 5 min to block endogenous peroxidase activity. To minimize nonspecific antibody binding, sections were incubated with serum blocking solution (Zymed) (i.e., 10% nonimmune goat serum). Sections were then incubated with primary antibodies in a moist chamber at 35°C overnight. Primary antibodies were used with the following dilutions: rabbit anti-MMP-2 (1:100) (cat. #AB809, lot #21082339; Chemicon), rabbit anti-MT1-MMP (1:250) (cat. #M3927, lot #072K1219; Sigma) and rabbit anti-TIMP-2 (1:100) (cat. #AB801-50, lot #22031406; Chemicon). Sections were washed thoroughly with PBS and incubated with biotinylated goat anti-rabbit IgG for 30 min and then with streptavidin-peroxidase conjugate for 10 min. Thereafter, sections were incubated with the aminoethylcarbazole mixture (substrate-chromogen mixture) for 5–10 min, counterstained with hematoxylin, and mounted. Sections were examined and photographed in an Olympus BX-40 (Olympus Industrial America, Inc., New York, NY) using UPlanF1 10, 20, and 40x objectives and an 82A blue filter. All micrographs were digitally acquired using an Olympus digital imaging camera (QColor 3 cooled) interfaced to a Macintosh G4 computer running under Mac OS 9.22 using the QCapture (V1.2.0) Software Package (Quantitative Imaging Corp., Burnaby, BC, Canada), and analyzed by Adobe Photoshop (Version 7.0). At least 50–100 sections were examined from each testis and at least three rats were used per time point in each experimental set. Controls consisted of i) sections incubated with PBS instead of the specific primary antibody; ii) sections incubated with normal rabbit serum or purified rabbit IgG at the same dilution as of the specific primary antibody but omitting the primary antibody incubation; iii) the secondary antibody being replaced with normal rabbit serum; and iv) sections incubated with the primary antibody that had been preabsorbed with lysates of seminiferous tubules, crude MMP-2 (Chemicon; cat. #CC071-5 for MMP-2 staining) or crude TIMP-2 (Chemicon; cat. #CC3327-5 for TIMP-2 staining). Control and experimental slides were processed simultaneously in the same session to eliminate interexperimental variations. For preabsorbed controls, 500 µg protein of tubule lysates or 2.5–5 µg crude protein was added to 50 µl of diluted primary antibodies, and incubated overnight at 4°C with agitation so that antibodies were bound to excessive antigens before its use. All control slides yielded nondetectable staining illustrating specificity of the staining results.
Immunofluorescent Microscopy for Colocalization Studies
Immunofluorescent microscopy using dual fluorescent probes was performed essentially as previously described [32]. Two different fluorescent probes, namely fluorescein isothiocyanate (FITC) and Cy3, were used to colocalize laminin 3 with either MMP-2 or ß1-integrin in the same tissue sections. In brief, frozen sections were prepared as described above. Sections were then treated with 10% normal donkey serum (cat. #S30; Chemicon) to minimize nonspecific antibody binding. Sections were subsequently incubated with a goat anti-laminin 3 antibody (1:50) (cat. #sc-16601, lot #B282; Santa Cruz), and followed by a donkey-anti-goat IgG-FITC (1:100) (cat. #AP180F; Chemicon). Thereafter, sections were washed in PBS and incubated with a rabbit anti-MMP-2 antibody (1:100) (cat. #AB809, lot #21082339; Chemicon) or a rabbit anti-ß1-integrin antibody (cat. #AB1952, lot #22111029; Chemicon), followed by a donkey anti-rabbit IgG-Cy3 (1:100) (cat. #AP182C; Chemicon). Sections were then mounted in Vectashield (Vector Laboratories, Burlingame, CA) and fluorescent microscopy was performed using an Olympus BX40 microscope equipped with Olympus UPlanF1 fluorescent optics. All images were digitally acquired using QCapture Imaging Software Package (Version 1.2.0.) and analyzed by Adobe Photoshop (Version 7.0). Controls included i) sections incubated with normal donkey serum instead of the primary antibody, ii) secondary antibody alone without the use of primary antibody, and iii) primary antibody preabsorbed with seminiferous tubule lysates or control peptides provided by the corresponding manufacturers as described above. For all controls, the test failed to yield detectable fluorescent staining.
Treatment of Adult Rats with (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic Acid (MMP-2/MMP-9 Inhibitor I), an Inhibitor of MMP-2 and MMP-9, or cis-9-Octadecenoyl-N-hydroxylamide (OA-Hy, MMP-2 Inhibitor I), an Inhibitor of MMP-2, to Examine Its Effects on the Kinetics of AF-2364-Induced Germ Cell Loss from the Seminiferous Epithelium and Changes in Tubule Diameter
To assess the precise physiological roles of MMPs in AJ dynamics in the seminiferous epithelium, an in vivo model of AJ disruption was used. In brief, inhibitors of MMPs were being used to monitor their effects on AF-2364-induced AJ disruption in vivo. Approximately 3 µM MMP-2/MMP-9 Inhibitor I (Mr 381.5; Calbiochem) or 10 µM MMP-2 Inhibitor I suspended in 200 µl saline was administered intratesticularly to the right testis of adult rats (300 gm BW, n = 3 per time point) at three sites using a 22-gauge needle with 70 µl per site as described [27, 41]. The stock solution of (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid was prepared in DMSO at 5 mg/ml. Assuming the testicular volume is 1.6 ml/testis, 1.8312 µg (3 µM, or 4.8 nmol/testis) (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid/testis was administered to each testis. For OA-Hy, the stock solution was also prepared in DMSO at 10 mg/ml, and 4.76 µg (10 µM or 16 nmol/testis) OA-Hy/testis was administered. These concentrations were selected based on earlier reports [42, 43]. The same volume of DMSO without inhibitors resuspend in 200 µl saline was injected into the left testis, which served as a negative control. About 0.5–1 h post-inhibitor treatment, each rat received a single dose of AF-2364 at 50 mg/kg BW by gavage or vehicle only (0.5% methylcellulose, w/v in PBS). Rats (n = 3) were killed on Days 1, 2, 4, and 6 and testes were removed, frozen in liquid nitrogen, and stored at -80°C. Sections were obtained in a cryostat and stained with Mayer hematoxylin. Cross sections of testes were randomly selected and seminiferous tubules consisting of elongating or elongate spermatids were scored. At least 100 tubules selected randomly from each testis (three testes from three rats were examined, with a total of 300 tubules per time point) were photographed, digitally stored, and printed for scoring. A tubule from rats treated with AF-2364, inhibitor+AF-2364, or inhibitor alone, which contained <50% of the number of elongating or elongate spermatids versus control tubules (for control testis, the number of elongating/elongate spermatids in a typical cross section of a seminiferous tubule at stages I–VI, VII–VIII, XI–XIV were 136 ± 15, 145 ± 18, 106 ± 15, and elongating/elongate spermatids were not found in stages IX–X tubules), was scored as a tubule with significant loss in elongating or elongate spermatids from the epithelium. The percentage of seminiferous tubules (ST) with normal elongating or elongate spermatids after AF-2364, AF-2364+inhibitor, or inhibitor alone treatment was calculated as follows:
To assess the effects of the inhibitor on the changes in tubular diameter induced by AF-2364, the following formula was used:
where STd/ctrl is the diameter of tubules in control rats; STd/AF-2364 with or without inhibitor is the diameter of tubules in AF-2364-treated rats with or without pretreatment with inhibitor. A total of 80 sections of tubules per testis were scored by measuring their diameters and the means from three testes of three different rats were used.
Statistical Analysis
Multiple comparisons were performed using one-way analysis of variance followed by Tukey honestly significant difference test to compare selected pairs of experimental groups so that changes in the expression of a target gene or protein level at a selected time point within an experimental group could be compared between samples. In selected experiments, a Student t-test was also performed by comparing treatment groups with the corresponding controls. Statistical analysis was performed using the GB-STAT Statistical Analysis software package (Version 7; Dynamic Microsystem, Inc., Silver Spring, MD).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It was postulated that pro MMP-2 at the cell surface is activated through a unique mechanism involving MT1-MMP and TIMP-2 in the testis [17, 19]. If this is true, one would expect MMP-2 (Fig. 1), MT1-MMP (Fig. 2), and TIMP-2 (Fig. 3) to be localized to similar sites in the seminiferous epithelium, such as the ES. Indeed, MMP-2, MT1-MMP, and TIMP-2 localized to the same site in the seminiferous epithelium at the apical ES, showing stage-specificity. The strongest staining for each of these target proteins was at stages VII–VIII, largely surrounding the heads of elongate spermatids adjacent to the seminiferous tubule lumen. This was consistent with their localization at the site of apical ES (Figs. 1, A and E–G, 2A, and 3, A and B). In virtually all stages of the epithelial cycle, a very weak staining of MMP-2 was also found associating with spermatocytes and to a lesser extent with round spermatids (Fig. 1). To verify that MMP-2 protein is present in germ cells, immunoblotting was performed using lysates prepared from germ cells isolated from adult rat testis. Both pro and active MMP-2 were indeed detected in germ cell lysates (data not shown). Results of the immunohistochemical localization of MT1-MMP reported herein (Fig. 2) was consistent with a previous report [17], which illustrated that the highest MT1-MMP level in the seminiferous tubule was found at stage VIII preceding spermiation. In stages VIII–XII, some weak staining of MT1-MMP and TIMP-2 was found largely associating with pachytene spermatocytes and elongating spermatids (Figs. 2, A and C, and 3, A–D). Weak MT1-MMP and TIMP-2 staining was also detected in spermatocytes and elongating/elongated spermatids at stages XIV–I (Figs. 2, D and E, and 3E) and at stages V–VI (Figs. 2F and 3F). While TIMP-2 staining was also detected in the interstitium, possibly associating with Leydig cells and blood vessel endothelia, MMP-2 and MT1-MMP staining in this region was not visible (Figs. 1–3). Figures 1B, 2B, and 3, G–I, are control sections in which the corresponding primary antibody was substituted by normal rabbit serum (other controls also yielded negative results, data not shown) indicating the reddish-brown precipitates of MMP-2, MT1-MMP, and TIMP-2 shown in Figures 1–3 were specific to these proteins.
|
|
|
Association of MMP-2 and MT1-MMP with Other AJ-Associated Proteins
Because MMP-2, MT1-MMP, and TIMP-2 colocalized at the site of apical ES in stages VII–VIII (Figs. 1, A and E–G, 2A, and 3, A and B), we next sought to investigate whether MMP-2 and MT1-MMP interacted with components of the ES-associated multiprotein complexes (for a review, see [1]). Using seminiferous tubule lysates, immunoprecipitation was performed using antibodies against MMP-2, MT1-MMP, laminin 3, ß1-integrin, N-cadherin, and nectin-3. It was shown that MMP-2 and MT1-MMP indeed associated with laminin 3 and ß1-integrin (Fig. 4, A and B) but not with N-cadherin (Fig. 4C) and nectin-3 (Fig. 4D).
|
Laminin 3 Forms Complex and Colocalizes with ß1-Integrin at the Site of Apical ES in the Seminiferous Epithelium
Laminin 3 was shown to associate tightly with ß1-integrin in the seminiferous tubule as demonstrated by immunoprecipitation (Fig. 4, A and B). By immunofluorescent microscopy, the localization of laminin 3 in the rat testis was stage specific, with the highest staining found at stages VII–VIII, largely around the heads of elongated spermatids adjacent to the seminiferous tubule lumen at the site of apical ES (Fig. 5A versus Fig. 5, D and G). Weak immunofluorescent staining of laminin 3 was also detected at the site of apical ES in stages XIII–XIV (Fig. 5D) and I–III (Fig. 5G). Furthermore, weak laminin 3 staining was also detected at basal ES at stages I–III (Fig. 5G). The immunofluorescent staining of ß1-integrin in the seminiferous epithelium is shown in Figure 5, B, E, and H. It was noted that ß1-integrin associated mostly with apical ES with weaker staining at the basal ES, consistent with earlier reports [25, 44]. In essence, both laminin 3 and ß1-integrin were found to localize to the same site in the apical ES at stages VII–VIII (Fig. 5C, merged images of Fig. 5, A and B), but this localization was not immediately visible at stages XIII–XIV (Fig. 5F, merged images of Fig. 5, D and E) and I–III (Fig. 5I, merged images of Fig. 5, G and H) because of a weak laminin 3 staining.
|
Colocalization of MMP-2 with Its Putative Substrate, Laminin 3, to Apical ES
To verify that MMP-2 is associated with its natural substrate, laminin 3, immunofluorescent microscopy was used to colocalize these two target proteins. MMP-2 (Fig. 6A) and laminin 3 (Fig. 6B) indeed colocalized to the same site in apical ES, with the strongest staining at stages VII (data not shown) and VIII (Fig. 6C, merged image versus Fig. 6, A and B). The immunofluorescent staining patterns of MMP-2 (MMP-2 was also found in the basal compartment; see arrowheads in Fig. 6A) and laminin 3 (Fig. 6B) were also consistent with results shown in Figures 1 and 5.
|
Relative Distribution of Laminin 3 in Isolated Sertoli and Germ Cells
To confirm and extend the immunofluorescent data of laminin 3, immunoblotting was performed using lysates of Sertoli (20-day-old) and germ (20- and 60-day-old) cells. It was noted that germ but not Sertoli cells indeed expressed laminin 3 (Fig. 7, A–D). To further validate that Sertoli cells did not contribute to the pool of laminin 3 in the testis, Sertoli cells were cultured at high density (1 x 106 cells/cm2) and terminated at specified time points; lysates were obtained for use in subsequent immunoblotting experiments. Laminin 3 was not detected in these Sertoli cell cultures (Fig. 7, C and D) when functional AJs and TJs were being formed in these cultures as assessed by an in vitro germ cell adhesion assay [7] and the measurement of transepithelial electrical resistance across the cell epithelium [27, 34] in parallel experiments (data not shown). Furthermore, ß1-integrin was found to be induced in these cultures as described [8] (data not shown), illustrating functional ESs were formed between Sertoli cells.
|
Changes in MT1-MMP and Laminin 3 During Testis and Germ Cell Development
During testis maturation, the immature (inactive pro form, 63 kDa) form of MT1-MMP remained relatively steady from 20 to 40 days of age except that a mild yet significant decline in the mature (active, 60 kDa) form was detected at 40 versus 20 days of age (Fig. 7, E and F). Both forms plummeted by more than 50% at 90 days of age (Fig. 7, E and F). Like MT1-MMP, laminin 3 also remained relatively steady from 20 to 40 days of age, but then plunged by 80% at 90 days of age (Fig. 7, G and H). Both pro and active MT1-MMP were detected in 20-day-old germ cells (Fig. 7, I and J). During germ cell maturation, the pro form decreased drastically from 20 to 60 days of age (Fig. 7, I and J). In contrast, the active form increased during germ cell aging (Fig. 7, I and J). For laminin 3, its expression in germ cells plummeted by at least 50% from 20 to 60 days of age (Fig 7, A and B).
Effects of AF-2364-Induced AJ Disruption in the Testis on the Protein Levels of MMP-2, MT1-MMP, TIMP-2, and MMP-9 In Vivo
To further strengthen the hypothesis that MMP-2, MT1-MMP, and TIMP-2 are involved in ES dynamics, an in vivo model of AJ disruption (for review, see [1]) using AF-2364 was utilized to assess changes in these proteins during AJ disruption in the seminiferous epithelium. When adult rats were treated with a single dose of AF-2364 at 50 mg/kg BW by gavage, both active MMP-2 (64 kDa) and active MT1-MMP (60 kDa) were induced within 1 h and 30 min, respectively, when compared with control rats at time 0, and remained at these levels for up to 8 h after treatment (Fig. 8, A and B). It should be noted that data obtained from normal rat testes were not different from rats at time 0 (data not shown). Induced MMP-2 and MT1-MMP plummeted rapidly thereafter (Fig. 8, A and B). When active MMP-2 and active MT1-MMP proteins were induced after treatment, the protein level of 68-kDa pro MMP-2 remained steady up to 8 h after treatment and decreased slightly thereafter (Fig. 8, A and B), whereas the 63-kDa pro MT1-MMP protein level in the adult rat testis was very low and barely detectable. In addition, its protein level remained unaltered after treatment (Fig. 8, A and B). The protein level of the 45-kDa cleaved MT1-MMP, which was devoid of its catalytic domain [45] and known to be detected in extracts of germ cells (but not Sertoli and myoid cells) [17], remained steady for up to 2 h after treatment and decreased drastically thereafter (Fig. 8, A and B). For TIMP-2, an induction in its protein level was detected within 30 min to 2 h, which persisted until Day 15 with its highest stimulation being 2.5-fold within 24 h after treatment (Fig. 8, A and B). In contrast with MMP-2, the protein levels of pro and active MMP-9 remained relatively steady after AF-2364 treatment except that they became undetectable at 15 days posttreatment (Fig. 8, A and C). The bottom panel of Figure 8A is an immunoblot for ß-actin, illustrating equal protein loading. It should be noted that all three experiments yielded identical results and the data shown in Figure 8, B–D, are representative results of these studies.
|
Changes in the Protein Level of Laminin 3 During AF-2364-Induced AJ Disruption in the Testis In Vivo
A gradual yet significant reduction in the protein level of laminin 3, the binding partner of 6ß1 integrin at the apical ES, was detected after AF-2364 treatment (Fig. 8, A and D). Its protein level became undetectable by 15 days posttreatment (Fig. 8, A and D), when virtually no spermatids were found in the epithelium. This is consistent with results shown in Figure 7, A–D, illustrating that laminin 3 chain is a product of germ cells and a possible binding partner of ß1-integrin. This was in contrast with ß1-integrin, which is a Sertoli cell product [25, 44].
Immunohistochemical Localization of MMP-2, MT1-MMP, and TIMP-2 in the Seminiferous Epithelium of Adult Rat Testis During AF-2364-Induced AJ Disruption
We proceeded to examine the localization of MMP-2 (Fig. 9), MT1-MMP (Fig. 10), and TIMP-2 (Fig. 11) in the seminiferous epithelium after AF-2364-induced AJ disruption. It was found that AF-2364 induced increases in the protein levels of MMP-2, MT1-MMP, and TIMP-2 in the seminiferous epithelium and that these changes were largely confined to the heads of elongating/elongated spermatids at cell-cell contact sites between spermatids and Sertoli cells, consistent with its localization at the apical ES, from 1–8 h posttreatment (Figs. 9, A–E; 10, A–E; and 11, A–C). It is of interest to note that immunoreactive MMP-2, MT1-MMP, and TIMP-2, which were all detected in the apical ES near the tubule lumen at stages VII–VIII in the normal testis, were found to be stimulated within the epithelium from 2–8 h posttreatment (Fig. 9, A–E, versus Fig. 1, E–G; Fig. 10, A–E, versus Fig. 2A; and Fig. 11, A–C, versus Fig. 3B). From 4 to 15 days posttreatment, tubules were gradually devoid of elongating/elongate and round spermatids and the number of spermatocytes was also significantly reduced. During this time, very low levels of immunoreactive MMP-2 and MT1-MMP were detected in the seminiferous tubule. The immunostaining patterns shown in Figures 9–11 for MMP2, MT1-MMP, and TIMP-2, respectively, were consistent with the immunoblotting data shown in Figure 8, A and B, illustrating that a transient induction of these three proteins occurred prior to visible germ cell loss from the epithelium as detected by histological analysis. These results thus suggest that there must be an increase in proteolysis prior to a loss of cell adhesive function. It should also be noted that, while immunoreactive MMP-2 was still detectable in testes by immunoblotting on Days 2, 4, and 15 after AF-2364 treatment (Fig. 8A), a very weak MMP-2 staining was found in the epithelium at these time points (Fig. 9, F–H).
|
|
|
Effects of a MMP-2 and MMP-9 inhibitor, (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid (MMP-2/MMP-9 Inhibitor I), on the Kinetics of AF-2364-Induced Germ Cell Loss from the Seminiferous Epithelium and Changes in Tubule Diameter
To further expand the notion on the significance of proteases in AJ dynamics, a specific MMP-2 and MMP-9 inhibitor, (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid, that can block the activation of MMP-2 and MMP-9, was used to pretreat testes to determine if it could delay AF-2364-induced germ cell loss from the epithelium. When rats were treated with AF-2364, the loss of elongating/elongate spermatids from the epithelium was clearly visible by 4 h (Figs. 9, 10, and also see [46]). The apical ES is apparently the primary site affected by AF-2364 [24, 46]. For instance, the percentage of tubules with elongating/elongate spermatids was significantly reduced 1 day after AF-2364 treatment (Fig. 12, A, D, G, J versus C, F, I, L, and Fig. 12M). In contrast, round spermatids and spermatocytes remained relatively unaltered in the seminiferous epithelium at 1–2 day posttreatment (Fig. 12, A and D versus C and F). The tubule diameter was also reduced significantly, 17% and 30% on Days 2 and 6 after AF-2364 treatment (Fig. 12, D versus F and N), respectively. When 3 µM (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid was injected intratesticularly 30 min to 1 h before AF-2364 treatment (50 mg/kg BW by gavage), a significant delay in the loss of elongating/elongated spermatids from the seminiferous epithelium was observed on Days 1 and 2 after treatment because most of the tubules found at these time points still consisted of elongating/elongate spermatids versus rats treated with AF-2364 alone (Fig. 12, B and E versus A and D, and Fig. 12M). Although (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid could effectively delay the AF-2364-induced germ cell loss (largely elongating and elongate spermatids) from the seminiferous epithelium, its efficacy disappeared after 2 days; because germ cell depletion on Days 4 and 6 after inhibitor pretreatment was similar to rats treated with AF-2364 alone (Fig. 12, H and K versus G and J, and Fig. 12M). It is of importance to note that pretreatment of testes with the MMP-2/MMP-9 inhibitor alone had no effect on the germ cell population in the epithelium (Fig. 12, C, F, I, and L). Also, another MMP-2 inhibitor, OA-Hy, when administered to testes at 10 µM to block MMP-2 activation prior to AF-2364 treatment, displayed similar effects as the MMP-2/MMP-9 inhibitor, in the sense that it delayed the loss of elongating/elongate spermatids from the epithelium (data not shown). In short, these results clearly indicate the pivotal role of proteases and protease inhibitors in AJ dynamics, such as at the site of apical ES, in the seminiferous epithelium.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The detection of immunoreactive MMP-2, MT1-MMP, and TIMP-2 at the apical ES by immunohistochemistry has provided a strong argument that the activation of pro MMP-2 in the seminiferous epithelium is mediated via a unique mechanism involving MT1-MMP and TIMP-2 [12, 14, 17]. For instance, most MMPs are activated extracellularly through the cleavage of their pro domains by either other activated MMPs or serine proteases, yet MMP-2 is likely activated at the cell surface with an initial activation of the transmembrane pro MT1-MMP by either intracellular furinlike serine proteases, cell surface plasmin, or nonproteolytic conformational changes. Although activated MT1-MMP is inhibited by the N-terminal domain of TIMP-2, it is also known to bind to it, forming a MT1-MMP-TIMP-2 complex. The C-terminal domain of TIMP-2 of this MT1-MMP/TIMP-2 complex then acts as a cell surface receptor for the hemopexin domain of pro MMP-2. Following the formation of this trimolecular complex, the bound pro MMP-2 can also be activated by another uninhibited MT1-MMP, which cleaves most portions of the MMP-2 pro peptide. The remaining residual pro peptide is then removed by another cell surface-activated MMP-2 to yield a fully activated and mature MMP-2 (for reviews, see [12, 14], see also Fig. 13). As such, the colocalization of the three protein partners, MMP-2, MT1-MMP, and TIMP-2, at the site of the apical ES as shown herein, strongly favors the notion that a similar mechanism is being used for MMP-2 activation, especially at cell-cell contact sites between Sertoli cells and elongate spermatids at the apical ES. This in turn regulates apical ES dynamics. Although a previous study has shown that MMP-2 was produced by somatic cells in the testis [17], weak immunoreactive MMP-2 as well as MMP-2 protein was also detected in germ cells. An explanation for the presence of MMP-2 protein in germ cells is not immediately known. However, it may be the result of endocytosis by germ cells. Such speculation is not unprecedented. For instance, internalization of the 2-macroglobulin/MMPs complex or the thrombospondin 2/MMP-2 complex via endocytosis is part of an important metabolic clearance pathway of these protein complexes in fibroblasts [47, 48].
|
In this context, it is of interest to note that plasmin (a serine protease with trypsinlike substrate specificity), a product of plasminogen (a protein known to be produced by rat seminiferous tubules) following its cleavage by urokinase plasminogen activator (for a review, see [11]), is also a putative activator of both surface-bound pro MMP-2 and pro MT1-MMP [49]. Because urokinase-type plasminogen activator (a serine protease) is also a stage-specific Sertoli cell product, being highest at stages VII–VIII [50], it is likely that it also takes part in the regulation of germ cell movement at spermiation by activating MMPs via plasmin. In short, results presented in this report favor the notion that MMP-2, MT1-MMP, and TIMP-2 are working synergistically to regulate ES dynamics in the seminiferous epithelium.
Laminin 3/ß1-Integrin Complex Exists as a Multiprotein Functional Unit at the Site of Apical ES in the Seminiferous Epithelium
Although an increasing number of nonmatrix proteins, such as cell surface proteins and transmembrane receptors, continue to be identified as MMP substrates (for reviews, see [12, 14, 15]), ECM components are still the major substrates. Because results presented herein have demonstrated the presence of MMP-2, MT1-MMP, and TIMP-2 at the apical ES, it is logical to identify their putative substrates at this cell-cell contact site. The discovery of a non-basement-membrane-associated ECM protein, namely laminin 3 at apical ES, has not only provided a clue that it may be a putative MMP substrate, it is also a potential binding ligand for 6ß1 integrin, an ES structural protein in the testis [25, 44]. Recently, a novel laminin 3 chain was found to localize in the apical portion of the seminiferous epithelium [23], yet its cellular origin or binding partner in the seminiferous epithelium was not known. In this article, we have demonstrated the presence of laminin 3 at the site of apical ES, colocalizing with ß1-integrin. Equally important, we have shown that germ cells exclusively produce laminin 3, which can interact with ß1-integrin to form the laminin 3/ß1-integrin complex. To the best of our knowledge, no previous study can be found in the literature that aims at identifying the in vivo binding partner for ß1-integrin at the site of apical ES. Furthermore, we also have provided compelling evidence that this laminin 3 chain is a MMP-2 and MT1-MMP substrate at the site of apical ES. It is obvious that a biochemical study to assess the proteolytic cleavage of laminin 3 by MMP-2/MT1-MMP at the apical ES site in vivo is needed. In this context, it is of interest to note that laminin 3 is also found at the surface of epithelial cells in various tissues, such as the epididymis, ductus deferen, oviduct, and lung [23]. It is also noteworthy to mention that ES is a cell-cell actin-based AJ structure unique to the testis with 6ß1 integrin, the transmembrane receptor, residing in Sertoli cells [1, 25]. Its association with laminin 3, which is produced by and resides exclusively in germ cells as shown herein, seemingly suggests that laminin 3/ß1-integrin is a putative ES-regulatory unit [1]. This type of regulation of junction dynamics imposed by laminin is not unprecedented. For instance, laminin 5 found at the leading edge of wound outgrowth by keratinocytes was known to regulate integrin-mediated adhesion and signaling [51]. Laminin 10/11 also plays a crucial role on integrin-dependent Rac activation during cell migration [52]. An earlier study has demonstrated that laminin may regulate Sertoli cell function by altering its intracellular [Ca2+] level [53]. However, it is crucial that, in the future, studies identify the two other and ß chains that form the functional laminin receptor protein with 3 (note: a functional laminin receptor is composed of laminin , ß, and chains) for 6ß1 integrin in the seminiferous epithelium. This in turn can facilitate further studies to delineate the precise functional role of laminin 3 in the testis, especially at the site of apical ES.
Although laminin 3 was shown to be an exclusive germ cell product (Fig. 7), largely associating with apical ES at the interface between Sertoli cells and elongating/elongate spermatids (Figs. 5 and 6), its level in 20-day-old germ cells when elongating/elongate spermatids were not present was almost twice as high as 60-day-old rats (Fig. 7, A and B). We offer the following explanation for this interesting yet important observation. First, it must be noted that some laminin 3 staining was indeed detected in other areas of the seminiferous epithelium in adult rat testes, such as at the site of basal ES (see Fig. 5G), suggesting that a small quantity of this protein is indeed expressed by primary spermatocytes and/or spermatogonia. The drastic increase in laminin 3 staining in stages VII–VIII at the apical ES as shown by immunofluorescent microscopy could be the result of Sertoli cell-elongate spermatid interactions possibly related to the event of spermiation. This latter possibility is currently being investigated using Sertoli-germ cell cocultures in vitro. Second, the decline in laminin 3 level in testis lysates being analyzed during aging shown in Figure 7, G and H, could be due to a relative increase in other testicular proteins, but possibly not in laminin 3, as animals age. And because the same amount of proteins (100 µg) was used for comparison by immunoblottings, the contribution of laminin 3 became lessened. It is plausible that both possibilities contribute to the results reported in Figure 7 versus those shown in Figures 5 and 6. Furthermore, we had not stained for laminin 3 in the 20- and 40-day-old rat testes but it is likely the level of laminin 3 is differentially regulated during development (see Fig. 7G). For instance, when elongating/elongate spermatids are absent in immature rats, laminin 3 may be targeted to primary spermatocytes and spermatogonia at basal ES.
Interactions of Laminin 3 and ß1-Integrin with MMP-2 and MT1-MMP at the Site of Apical ES: What Is the Functional Significance of Their Interactions?
Both MMP-2 and MT1-MMP were found to structurally associate with laminin 3 and ß1-integrin at the site of apical ES, but not with N-cadherin or nectin-3, two putative ES proteins. Thus, these results suggest their involvement in the regulation of laminin/integrin-based ES dynamics. There is accumulating evidence that the cleavage of laminin-5 by MMP-2 in breast epithelial cells or by MT1-MMP in prostate carcinoma cells results in the release of biologically active laminin fragments, which in turn can induce cell migration [54–56]. Needless to say, the proteolytic cleavage of laminin 3 (and/or the other two yet-to-be identified and ß chains) by MMP-2 and MT1-MMP in the testis remains to be shown. It is possible that laminin 3 may be cleaved by MMP-2 or MT1-MMP at the apical ES to generate the biologically active fragments that are needed to facilitate germ cell movement or spermiation through a yet-to-be defined mechanism, which probably involves 6ß1 integrin as a signal transducer. Obviously, this is a research area that needs to be vigorously investigated in future studies.
It has been suggested that pericellular localization of MMPs can assist in their activation, recruiting MMPs to a specific cellular location and to confine proteolysis to a specific site (for a review, see [12]). Such restricted localization of MMPs to the cell surface can be achieved by restricted expression of membrane-bound MT1-MMPs or by binding to cell surface docking receptors, such as integrin (for review, see [12]). For instance, the localization of activated MMP-2 to the surface of invasive cancer cells is via its specific binding with vß3 integrin [57]. It was postulated that MT1-MMP-bound vß3 integrin participated cooperatively in facilitating MMP-2 activation in breast carcinoma cells [58]. ß1-Integrin has also been shown to associate with MT1-MMP in ovarian carcinoma cells [59]. An additional study has also demonstrated the linking of MT1-MMP with ß1 and vß3 integrins in endothelial cells [60]. Such interactions have been shown to be essential for the localization of MT1-MMP to the cell-cell contact sites or motile cellular structures [60]. Furthermore, the ECM, in cooperation with integrin, is known to affect the expression and activation of MMPs in endothelial and tumor cells [45, 59, 60]. Herein, we provide evidence that there is physical and possibly biochemical association between MMP-2/MT1-MMP and ß1-integrin/laminin 3 at the apical ES. Because the apical ES is found between Sertoli and developing spermatids with 6ß1 integrin being the transmembrane receptor residing in Sertoli cells and laminin 3 in germ cells, it is tempting to speculate that ß1-integrin may recruit MMP-2 and MT1-MMP to the apical ES. These in turn can interact with laminin 3, such as via proteolysis, releasing the biologically active fragments to regulate ES dynamics.
AF-2364 Induced ES Disruption Is Regulated by the Interplay of MMP-2, MT1-MMP, and TIMP-2
AF-2364 is a potential male contraceptive that exerts its effects by inducing germ cell loss from the seminiferous epithelium, apparently by disruptin
