| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Population Council (J.Z., C.W., W.X., D.D.M., N.P.Y.L., C.Y.C.), Center for Biomedical Research, New York, New York 10021; and Department of Zoology (W.M.L.), The University of Hong Kong, Hong Kong, China
Address all correspondence and requests for reprints to: C. Yan Cheng, Ph.D., Population Council, 1230 York Avenue, New York, New York 10021. E-mail: Y-cheng{at}popcbr.rockefeller.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-actinin, vinculin, afadin, and p130 Crk-associated protein; and AJ-integral membrane proteins at the ectoplasmic specialization (ES, a testis-specific cell-cell actin-based AJ type) site: N-cadherin, ß-catenin, integrin ß1, and nectin 3. Second, MTMR2, instead of structurally interacting with phosphatidylinositol 3-kinase, a protein and lipid kinase, was shown to associate only with c-Src, a nonreceptor protein tyrosine kinase, as demonstrated by both coimmunoprecipitation and fluorescent microscopy at the site of apical ES, but none of the kinases, adaptors, and AJ-integral proteins that were examined. Collectively, these results suggest that the MTMR2/c-Src is an important phosphatase/kinase protein pair in AJ dynamics in the testis. Because c-Src is known to associate with the cadherin/catenin protein complex at the ES in the testis, we next sought to investigate any changes in the protein-protein interactions of this protein complex during androgen suppression-induced germ cell loss. Indeed, there was a loss of N-cadherin and ß-catenin association, accompanied by a surge in Tyr phosphorylation of ß-catenin, during germ cell loss from the epithelium. Third, and perhaps the most important of all, during natural recovery of the epithelium after removal of testosterone-estradiol implants when spermatids were reattaching to Sertoli cells, an increase in N-cadherin and ß-catenin association was detected with a concomitant loss in the increased Tyr phosphorylation in ß-catenin. In summary, these results illustrate that the cadherin/catenin is a crucial cell adhesion complex that regulates AJ dynamics in the testis, and its functionality is likely modulated by the MTMR2/c-Src protein complex. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Testosterone (T) is a male sex hormone that determines male phenotype and fertility. Its testicular level is approximately 100-fold of that in serum (14). A reduction of testicular T level by increasing serum T and estradiol-17ß (E) levels in male rats using TE implants can suppress the development of step 8 spermatids and beyond, leading to their detachment from the epithelium possibly by compromising ES function (for reviews, see Refs. 12 , 15). After TE implants are removed, and if rats are permitted to recover in the presence of excessive T or allowed to undergo spontaneous natural recovery (NR), a restoration of T level in the testes can be achieved, which is also accompanied by a reversal of spermatogenesis. Collectively, these observations illustrate that this is a novel model to study ES dynamics (for a review, see Ref. 12).
Myotubularin-related protein 2 (MTMR2) is a lipid phosphatase at 3D site within the myotubularin (MTM) family having a catalytic site found in protein tyrosine phosphatases (PTP) and dual specific phosphatases (16, 17, 18). Its mutation causes the demyelinating neuropathy Charcot-Marie-Tooth disease type 4B1 (CMT 4B1) (16, 19). MTMR2 dephosphorylates phosphatidylinositol 3-phosphate and phosphatidylinositol 3,5-bisphosphate, both in vivo and in vitro, and the product of the latter can act as a feedback regulator of MTMR2 (20, 21). Interestingly, MTMR2 is expressed by Sertoli and germ cells in the testis, and its level is induced during testicular maturation or when Sertoli-germ cell AJs are assembled in vitro (22, 23), implicating its role in AJ dynamics. A recent study has shown that MTMR2–/– mice displayed not only neuropathy but also azoospermia, accompanied by a significant reduction in testis size and weight, implicating the pivotal role of MTMR2 in AJ function and spermatogenesis (24). Besides forming homodimers between MTMR2 (25), MTMR2 can interact with several proteins in other epithelia such as MTMR5 (26). Moreover, neurofilament light chain protein and disc large 1 (Dlg1) are also structurally associated with MTMR2 in the brain (24, 27). However, a binding partner of MTMR2 has yet to be found in the testis. We therefore sought to identify the putative interacting partner(s) of MTMR2 in the seminiferous epithelium and to assess whether MTMR2 interacts with any protein kinases at the ES site to regulate cell adhesion function.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Antibodies
Antibody against rat (r)MTMR2 was raised in a rabbit using a 22-amino-acid peptide (residues 156–177 of rMTMR2) of NH2-TKVNERYELCDTYPALLAVPAN-COOH as earlier described (22), which shared 90.9% identity with human MTMR2 at residues 228–249 of NH2-TKINERYELCDRYPALLVVPAN-COOH. The italicized amino acids represent those that are different between the two species. Other antibodies used for immunoblotting studies are listed in Table 1
.
|
Primary Sertoli cell cultures
Sertoli cells were isolated from 20-d-old rat testes as described (29). Cell number was determined by using a hemocytometer. Sertoli cells adjusted to the desired cell density were suspended in serum-free Ham’s F12 nutrient mixture (F12) and DMEM (F12/DMEM, 1:1, vol/vol) supplemented with 15 mM HEPES, 1.2 g/liter sodium bicarbonate, 10 µg/ml bovine insulin, 5 µg/ml human transferrin, 2.5 ng/ml epidermal growth factor, 20 mg/liter gentamicin, and 10 µg/ml bacitracin. Twelve-well dishes were precoated with Matrigel (Collaborative Biochemical Products, Bedford, MA), diluted 1:6 (vol/vol) in F12/DMEM, and dried overnight at 35 C. Cells were plated on dishes at a density of 0.5 x 106 cell/cm2 as described (30) with 3 ml F12/DMEM per well and incubated in a humidified atmosphere of 5% CO2 and 95% air (vol/vol) at 35 C. Approximately 48 h after cell plating, Sertoli cells were hypotonically treated with 20 mM Tris (pH 7.4) to lyse residual germ cells (31), followed by two washes with F12/DMEM. The purity of Sertoli cells used for our studies was at least 95% as characterized by light and electron microscopy (32, 33).
Sertoli-germ cell cocultures
Germ cells were isolated from 90-d-old rat testes by a mechanical procedure using sequential filtrations as described (34). For germ cell lysate preparation, cells were terminated within 3 h after their isolation. For cocultures, freshly isolated germ cells were suspended in F12/DMEM supplemented with 6 mM sodium lactate, 2 mM sodium pyruvate, 20 mg/liter gentamicin, and 10 µg/ml bacitracin. Cell number was determined by a hemocytometer. Germ cells at desired cell number were reconstituted in F12/DMEM in the absence or presence of T (2 x 10–8 or 2 x 10–7 M) (Sigma Chemical Co., St. Louis, MO) and cyproterone acetate (1 x 10–6 M) (Sigma). Germ cells were cocultured with Sertoli cells at a 1:1 ratio in which Sertoli cells had been cultured alone for 5 d at 0.5 x 106 cells/cm2 forming an intact epithelium (33, 35). Cocultures were terminated at 24 and 48 h thereafter when functional anchoring junctions, such as desmosome-like junctions and apical ES, were formed (3, 36, 37).
Immunohistochemistry
Testes isolated from normal rats and from animals of different treatment groups were frozen immediately in liquid nitrogen. Frozen testes were embedded in OCT compounds (Miles Scientific, Elkhart, IN), and 8-µm sections were cut using a microtome in a cryostat at –20 C and mounted on poly-L-lysine-coated slides. Histostain-SP kits obtained from Zymed Laboratories (South San Francisco, CA) were used for immunohistochemistry as described (38, 39). Sections were air dried at room temperature, fixed in Bouin’s fixative (4% formaldehyde in picric acid) for 5 min, and washed with PBS [10 mM sodium phosphate, 0.15 M sodium chloride (pH 7.4) at 22 C]. Sections were then treated with 3% hydrogen peroxide solution (vol/vol, in methanol) for 20 min to eliminate endogenous peroxidase activity and blocked with a serum blocking solution (10% nonimmune goat serum, Zymed). Rabbit anti-rMTMR2 serum (1:300, in 1% nonimmune goat serum) or preimmune rabbit serum (control, 1:300, in 1% nonimmune goat serum) was added onto sections, and slides were incubated in a humidified chamber at 35 C overnight. Thereafter, sections were rinsed in PBS and incubated with biotinylated goat antirabbit IgG for 30 min, to be followed by streptavidin peroxidase conjugate for 10 min. Color development was performed using aminoethylcarbazole mixture. Sections were counterstained using hematoxylin and mounted with Histomount (Zymed). All sections from different rats within an experimental group were processed simultaneously in two to three microscopic slides (with two to four sections per slide) to eliminate interexperimental variations. Micrographs were obtained using an Olympus DP70 12.5 MPa Digital Camera interfaced to an HP Vectra VL800 Workstation and the QCapture Suite V2.60 software package from Quantitative Imaging Corp. (Burnaby, Canada). Digital images were subsequently processed using Adobe Photoshop (version 7.0). All immunohistochemistry experiments were repeated at least eight to 12 times over a period of 10 months using either frozen or paraffin sections from different rats, and the results depicted in Figs. 1–3 are the representatives of these analyses.
|
|
|
Treatment of rats with AF-2364 to induce germ cell loss from the seminiferous epithelium
AF-2364 was synthesized as previously described (6) with a purity greater than 99.8% when examined by elemental analysis. Male rats [250–300 g body weight (BW)] were fed with a single dose of AF-2364 at 50 mg/kg BW as described (6, 7), which was shown to induce germ cell loss from the epithelium, in particular round, elongating, and elongated spermatids (44), by perturbing cell adhesion function in the epithelium (for reviews, see Refs. 2, 3). Rats at time 0 served as controls. Testes were removed at specific time points, frozen in liquid nitrogen, and stored at –80 C until used.
Induction of germ cell loss from the seminiferous epithelium using an in vivo androgen suppression model in adult rats with steroid implants
Ethyl vinyl acetate (DuPont ELVAX 770) implants (3 or 4 cm in length) filled with T (Sigma) and 0.4-cm implants filled with E (Sigma) were prepared as described (45). A single 3-cm T and a 0.4-cm E implant were placed under the skin along the dorsal surface of each male adult rat (n = 3 rats for each specified time point) to suppress the endogenous T level in the testis, which was shown to induce germ cell loss from the seminiferous epithelium (for reviews, see Refs. 12 , 15). Rats were under anesthesia using ketamine HCl (
70 mg/kg BW) during the surgery. The surgical area (2 cm2) was shaved to remove hair, cleansed with 70% ethanol and Betadine (two times each). A small incision (1.5 cm) was made using a sterile scalpel. Implants were carefully inserted under the skin, and the surgical site was stitched using Ethilon PC-S sterile, nonabsorbable surgical suture with a 19 PC prime needle (Ethicon Inc., Somerville, NJ). After 28 d, TE implants were removed from all rats under anesthesia as described above. In one group, rats (n = 3 rats for each time point) received four 4-cm T implants (without E) to permit rapid T recovery and resumption of spermatogenesis, whereas in the other group, rats were allowed to have spontaneous (i.e. natural) recovery without any steroid implants. Rats had free access to standard chow and water with a 12-h light, 12-h dark cycle. Rats at time 0 were designated controls. Testes were removed at specific time points, frozen in liquid nitrogen, and stored at –80 C until used.
Co-IP
Approximately 500 µg of protein from testis lysates, prepared as described above from normal rats, rats that underwent spontaneous recovery (NR) on d 42, or rats that received four 4-cm T implants to undergo rapid recovery on d 30, were used for IP. Samples were pretreated with 2 µg rabbit IgG (or mouse or goat IgG, depending on the source of the primary antibody) for 3 h, followed by an incubation with 10 µl protein A/G plus agarose (Santa Cruz Biotechnology, Santa Cruz, CA) to eliminate nonspecific interactions between lysates and IgG. After centrifugation at 1000 x g for 5 min to precipitate the IgG-nonspecifically bound protein complex, supernatants were collected for subsequent co-IP. In brief, 2 µg of an anti-rMTMR2 antibody, IgG against a target protein (e.g. c-Src), or normal rabbit IgG (negative control), was added and incubated with the supernatant at 4 C overnight. The immunocomplexes were then precipitated using 20 µl protein A/G plus agarose on a rotator (40 rpm) for 6 h. Immunoprecipitates were washed with 300 µl IP buffer (four times, by gentle resuspension of the pellet in the IP buffer followed by centrifugation at 1000 x g for 5 min). Proteins in the immunocomplexes were extracted in SDS-sample buffer [0.125 M Tris (pH 6.25) at 22 C containing 1% SDS (wt/vol), 1.6% 2-mercaptoethanol (vol/vol), and 20% glycerol (vol/vol)] at 100 C for 10 min, and samples were resolved by SDS-PAGE onto 7.5% T SDS-polyacrylamide gels. After proteins were electroblotted onto nitrocellulose papers, immunoblotting was performed as earlier described using different antibodies (38). Lysates from normal testes and seminiferous tubules served as positive controls to assess the specificity of the primary antibodies and the electrophoretic mobility of the target proteins.
Immunofluorescence microscopy
Immunofluorescence microscopy was performed as earlier described (32, 46) to examine whether rMTMR2 and c-Src colocalized to the same site in the seminiferous epithelium, verifying results of IP that rMTMR2 and c-Src are indeed part of a structural complex at the ES. Frozen sections were fixed in Bouin’s fixative (4% formaldehyde in picric acid) for 5 min and washed with PBS [10 mM sodium phosphate, 0.15 M sodium chloride (pH 7.4) at 22 C]. To eliminate interexperimental variation, all sections obtained from testes in different treatment groups within an experimental set were processed simultaneously in two slides because each slide holds at least three to four cross-sections of testes. This is necessary so that all sections within an experimental set can be stained, colorimetrically developed, and photographed simultaneously. Sections were blocked with 10% normal goat serum for 30 min. Sections were then incubated with a rMTMR2 antibody at a dilution of 1:200 in PBS containing 1% goat serum and then with a p-Src-Tyr416 antibody (Upstate Biotechnology, Lake Placid, NY; catalog no. 05-677, lot 26419) at a concentration of 20 µg/ml in PBS containing 1% goat serum, to be followed by 1:50 fluorescein isothiocyanate (FITC)-goat-antirabbit IgG conjugate (Zymed, catalog no. 62-6111, lot 30677924) and 1:50 goat antimouse IgG-CY3 conjugate (Zymed, catalog no. 81-6515, lot 31079880R). Sections were then mounted by using Vectashield mounting medium with 4',6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Inc., Burlingame, CA; catalog no. H-1500, lot Q0220). Images were digitally captured by using an Olympus BX40 microscope with UPlanF1 fluorescent optics (Olympus Corp.) and an Olympus DP70 12.5 MPa digital camera. All data were analyzed using Adobe Photoshop (version 7.0). Controls included the use of either IgG or 1% goat serum instead of the primary antibodies, which failed to yield detectable fluorescence, illustrating that the staining was specific.
Statistical analysis
Statistical analysis was performed using the GB-STAT Statistical Analysis Software Package (version 7.0, Dynamic Microsystems, Inc., Silver Spring, MD).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, A–H, which appeared as reddish-brown precipitates. Figure 1, B and C, are negative controls showing the staining with the preimmune serum. Figure 1C corresponds to the boxed area of Fig. 1B, illustrating that the staining shown in Fig. 1A was specific to rMTMR2. rMTMR2 was shown to associate with both Sertoli cells and germ cells in virtually all stages of the epithelial cycle (Fig. 1, A and D–H). However, rMTMR2 was detected predominantly at the site consistent with its localization at the apical ES at stages XI–IV of the epithelial cycle in the rat testis (Fig. 1, A and D–H). For instance, it localized to the heads of elongating spermatids at stage XII–IV at the apical ES (Fig. 1, D–H), as well as at the basal ES (Fig. 1, A and D–H). However, rMTMR2 was not detected at the apical ES at the Sertoli-elongating/elongate spermatid interface at stage V–VII, yet it became intensely localized to this site at stage VIII just before spermiation (Fig. 1F, compare the stage VIII tubule on the left and the stage V tubule on the right, vs. Fig. 1, A, D, E, G, and H). Furthermore, some staining of rMTMR2 was found in the interstitium, surrounding blood vessels. This latter finding suggests that rMTMR2 may also play a role in interstitial Leydig cells and blood vessel endothelium. These results are consistent with an earlier report, demonstrating rMTMR2 (formerly reported as rMTM) in round spermatids of stages VII and IX and Sertoli cells (22). Indeed, studies by immunoblotting using the lysates of testes, seminiferous tubules, germ cells, and brain (Fig. 1I) conclusively identified rMTMR2 in these tissues/cells that exists as three isoforms ranging between 66 and 68 kDa. Interestingly, germ cells were shown to produce rMTMR2 (Fig. 1I) consistent with the immunohistochemical data shown in Fig. 1, A–H. It is noted that these germ cells had negligible somatic cell contaminations when assessed by RT-PCR using primer pairs specific to protein markers of Leydig, peritubular myoid, and Sertoli cells (data not shown) as earlier reported (32, 43). The specificity of this antibody was also shown by immunoblot analysis (Fig. 1J).
Changes in the localization and protein level of rMTMR2 during AF-2364-induced AJ disruption in the seminiferous epithelium
AF-2364 (single dose, 50 mg/kg BW) was fed to adult rats at time 0 to induce germ cell depletion from the seminiferous epithelium by perturbing AJ function. Even before germ cell depletion was clearly visible in the epithelium, there was an increase in immunoreactive rMTMR2 in the epithelium, such as at the apical ES at the Sertoli-spermatid interface by d 1 after treatment compared with rats at time 0 (Fig. 2, A–C). When the event of germ cell depletion became more pronounced, more rMTMR2 was detected at the Sertoli-round spermatid interface (Fig. 2, D–G). A negative control using a preimmune serum is shown in Fig. 2H, illustrating the specificity of the rMTMR2 staining. Furthermore, these immunohistochemistry results were consistent with the immunoblot analysis shown in Fig. 2, I and J, when the rMTMR2 protein level was quantified in testis lysates. For instance, a surge of rMTMR2 was detected on d 1 after treatment (Fig. 2I). When more spermatids were being released to the lumen, the level of rMTMR2 also surged, increasing by almost 10-fold by d 15 after treatment when tubules were virtually devoid of germ cells except for some spermatogonia (Fig. 2, I, J, and G). By 60 d, when germ cells began to repopulate the epithelium (for reviews, see Refs. 2 , 6), the rMTMR2 protein level also subsided (Fig. 2I), illustrating that the resumption of spermatogenesis and possibly AJ recovery are associated with a tumbling of rMTMR2. Figure 2I, bottom, shows an anti-ß-actin staining using the same blot as the top, which served as the protein loading control.
Changes in the localization and protein level of rMTMR2 in the rat testis during androgen suppression-induced AJ disruption and its recovery
When a 3-cm T and a 0.4-cm E implant were inserted into adult rats at time 0 to suppress the endogenous T level in the testis, this procedure would induce AJ disruption and germ cell loss from the seminiferous epithelium (12, 15, 47). On d 28, both implants were removed and four 4-cm T implants were inserted to the same site in half of the rats (n = 3 for each time point) to induce rapid recovery of androgen level, whereas the other half of the animals (n = 3 per time point) were allowed to undergo spontaneous recovery (NR) without any steroid implants (for a review, see Ref. 12). Similar to the AF-2364 model, there was a progressive surge in rMTMR2 staining in the epithelium during androgen suppression-induced germ cell loss (Fig. 3, C and D vs. A and B). For instance, when most of the tubules were still devoid of elongating/elongate spermatids by d 42 (note that TE implants were removed on d 28) and permitted to undergo spontaneous recovery, the staining of rMTMR2 was very intense in the epithelium (Fig. 3E), yet rMTMR2 staining was greatly reduced when spermatids began to reappear in the epithelium in rats that either received four T implants or underwent NR (Fig. 3, F–H vs. E). Based on the observation of such a rapid recovery of spermatids in the epithelium, within approximately 20 d, it is logical to conclude that most spermatocytes were not depleted from the epithelium by this treatment. Figure 3I shows a normal testis stained with preimmune rabbit serum that served as a negative control. Such changes in immunohistochemical localization of rMTMR2 during androgen suppression-induced spermatid loss and the subsequent recovery were verified by immunoblotting shown in Fig. 3, J and K. For instance, rMTMR2 was induced by d 8 after receiving TE implants and reached its peak by 42 d in the NR group but peaked by 30 d in the four-T-implants group and declined gradually during the recovery phase. The lower panels of Fig. 3J are the same blots as the upper panels but stained with an anti-ß-actin antibody to illustrate equal protein loading.
Effects of androgen suppression-induced AJ disruption and its recovery on the protein levels of selected kinases and adaptors in the seminiferous epithelium in vivo
To investigate possible MTMR2 interacting protein partners, including kinases and adaptors, immunoblot analysis was first performed to assess whether any signaling proteins/adaptors were induced. As shown in Fig. 4, A and B, an increase in protein levels were found for phosphatidylinositol 3-kinase (PI 3-K) p85, c-Src, C-terminal Src kinase (Csk), -actinin, vinculin, and p130Cas, whereas no significant changes were detected for casein kinase (CK) II, axin, and zyxin during androgen-induced AJ disruption. PI 3-K p85, Csk, and vinculin displayed a similar trend; their protein levels were induced before the epithelium was fully recovered naturally on d 42 and on d 30 in the high-dose T recovery group, but tumbled thereafter (Fig. 4, A and B), which was somewhat later than MTMR2 (Fig. 3J). c-Src, -actinin, and p130Cas were also induced, yet their induction was detected earlier, such as by d 20–28 after TE implants were placed to deplete androgen, and their recoveries were also earlier (Fig. 4, A and B), similar to MTMR2 (Fig. 3J). These findings suggest that the expression of these two groups of proteins may be regulated differently at the time of androgen depletion-induced germ cell loss and its recoveries. The bottom panel in Fig. 4A is the corresponding blot stained for ß-actin to illustrate equal protein loading.
|
), but neither PI 3-K p85 nor vinculin (Fig. 5, B and C), was associated with MTMR2. All other kinases and adaptors that were examined also failed to associate with MTMR2 (Fig. 5D).
|
illustrates the colocalization pattern of rMTMR2 and p-Src-Tyr416, the active form of c-Src, in a stage VIII tubule. rMTMR2 appeared as green fluorescence with an FITC conjugate (Fig. 6A). p-Src-Tyr416 appeared as red fluorescence with a Cy3 conjugate (Fig. 6B). It is apparent that p-Src-Tyr416 has a more restricted distribution in the epithelium and was confined to the apical ES. This is somewhat different from c-Src because c-Src was localized both at the apical ES as well as at the interface of Sertoli cell and round spermatids/spermatocytes at stages VI–VIII as recently reported (13). The merged image of Fig. 6, A and B, is shown in Fig. 6C, illustrating rMTMR2 and p-Src-Tyr416 were indeed colocalized to the same site at the apical ES. Figure 6D is a DAPI staining.
|
3, and nectin 3/afadin in the rat testis3, and nectin 3/afadin, using testis lysates from the androgen depletion model (Fig. 7A). N-Cadherin, ß-catenin, ß1-integrin, and afadin were all induced during androgen suppression-induced germ cell loss, whereas nectin 3 was reduced and laminin-3 displayed no significant change. Except for laminin-3, all other structural proteins began to recover in 2 d after TE implants removal (i.e. d 30 after rats had received TE implants that were removed on d 28). This pattern of changes in AJ structural protein complexes illustrates that many AJ proteins were induced during germ cell loss, possibly trying to rescue the epithelium, and they returned to their normal levels during recovery. A representative ß-actin blot was also shown in the lower panel in Fig. 7A to verify equal protein loading.
|
-catenin in Sertoli-germ cell cocultures with and without treatments with T or cyproterone acetate (CPA) plus T). It was shown that the levels of both MTMR2 and c-Src were induced when cocultures were treated with either 0.02 or 0.2 µM T (the latter represents the normal intratesticular T level) (14). More importantly, this androgen-induced MTMR2 and c-Src protein productions were blocked by CPA (Fig. 8, A and B), illustrating the specificity of the androgen treatment. Because -catenin was shown to be T inducible and c-Src was known to bind to the cadherin/catenin complex via catenin (13), the corresponding -catenin blot served as an internal positive control. Indeed, -catenin was induced by T treatment in cocultures and such induction could be blocked by CPA (Fig. 8, A and B). The bottom panel was the same blot as the upper panels but reprobed with an anti-ß-actin antibody that served as a protein loading control.
|
, A and B, there was a significant loss in the interactions between N-cadherin and ß-catenin (Fig. 9, A–C) regardless of whether co-IP was performed using an anti-cadherin or an anti-catenin antibody during androgen suppression-induced germ cell loss from the epithelium. This effect appears to be specific because when spermatids reattached to the epithelium during the recovery phase, the interactions between cadherins and catenins were reestablished, making them indistinguishable from normal testes (Fig. 9, A–C). More important, it was shown that when ß-catenin lost its ability to associate with N-cadherin during androgen suppression-induced germ cell loss, the Tyr-phosphorylation status of ß-catenin was significantly increased. However, when the cadherin-catenin interaction was reestablished during the recovery phase, the induced Tyr phosphorylation of ß-catenin subsided (Fig. 9, B and C). Interestingly, when co-IP was performed using a c-Src antibody and the resultant blot was probed with an anti-ß-catenin antibody, we also detected a loss of c-Src association with ß-catenin during germ cell loss (Fig. 9, B and C). And more c-Src was found to reassociate with ß-catenin during the recovery (Fig. 9B). Based on the findings reported herein, we now propose a molecular model that illustrates the significance of T on the integrity of the cadherin-catenin protein complex at the ES site. Although an increase or a decrease in endogenous T can induce the production of N-cadherin, ß-catenins, cSrc, and MTMR2 (Fig. 10) as reported herein, the integrity of the cadherin/catenin complex is compromised when the T level is suppressed because of a loss of protein-protein interactions between N-cadherin and ß-catenin, as well as between ß-catenin and cSrc, as illustrated in Fig. 10. An increase in T not only induces the production of N-cadherin, ß-catenin, MTMR2, and c-Src proteins, but it also stabilizes protein-protein interactions between these proteins at the ES site (Fig. 10).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
First, it has been known for decades that T produced by Leydig cells under the regulation of LH is essential to spermatogenesis (for reviews, see Refs. 12 , 48). TE implants that induce a supra-T level in the systemic circulation can suppress the intratesticular T level. This, in turn, causes detachment of step 8 spermatids and beyond from Sertoli cells in the seminiferous epithelium, so that spermatids are released prematurely to the tubule lumen (for reviews, see Refs. 12 , 15). When the testicular T level is restored either by T implants that replace the TE implants or simply by TE implant removal to permit naturally spontaneous recovery, cell adhesion function and spermatogenesis can be restored. Thus, this model was used in this study. Interestingly, it was reported that actin bundles found between the Sertoli cell membrane and smooth endoplasmic reticulum at the site of apical ES remained qualitatively normal during TE treatment that depleted endogenous T level in the testis (47). This observation is significant because it suggests that cell adhesion molecules at the apical ES per se are the target of this treatment rather than the underlying cytoskeleton. It is also possible that changes in the cell adhesion function of these ES-structural proteins can contribute to the detachment of germ cells from the epithelium rather than an extensive restructuring of the ES structural protein/actin cytoskeleton complex. Indeed, it was shown in our studies that the event of germ cell loss during intratesticular T suppression induced by TE implants correlated to a loss of protein-protein interactions between cadherins and catenins as well as between c-Src and ß-catenin. Perhaps most importantly, it is apparent that such a loss of protein-protein interaction is the result of an increase in Tyr phosphorylation of ß-catenin. This is consistent with the findings based on studies in other epithelia that an increase in ß-catenin phosphorylation can affect cell adhesion function of the cadherin/catenin protein complex (for reviews, see Refs. 2 , 3 , 49 , 50). This is likely mediated by c-Src, which is a nonreceptor protein tyrosine kinase. It is plausible that once c-Src induces phosphorylation of ß-catenin, it is dissociated from the cadherin/catenin protein complex, which also explains the surge in c-Src (to induce ß-catenin phosphorylation) and its loss of association with ß-catenin (when it finishes its job) at the time of germ cell loss. Second, it was observed that many phosphatases, kinases, and adaptors, as well as integral membrane proteins that were examined in this androgen-suppression model, were induced at the time of germ cell loss, consistent with the trend found in the AF-2364 model (13, 32, 38, 43, 46). Collectively, these results suggest that the mechanisms that induce germ cell loss from the epithelium in both models are similar and that the signaling pathways that were shown to regulate AJ dynamics in the seminiferous epithelium are not nonspecific toxic effects of AF-2364. The fact that the effects of AF-2364 in inducing germ cell loss are much more rapid than the TE-induced affects, 1–7 d vs. 15–30 d, seemingly suggests that AF-2364 is a more potent chemical entity to induce germ cell loss vs. the androgen suppression-induced changes in the epithelium. Needless to say, much work remains to be done to delineate the downstream action of the signaling pathways that regulate Sertoli-germ cell attachment during spermatogenesis. From the findings reported herein, one of these mechanisms is mediated via changes in protein-protein interactions of the cell adhesion molecules at the Sertoli cell-spermatid interface.
MTMR2 structurally interacts with c-Src but not PI 3-K p85 in the rat testis
MTMR2 is a member of the MTM family of phosphatase/pseudophosphatase with a PTP/dual specific phosphatase catalytic site, CX5R (for a review, see Ref. 18). It can dephosphorylate phosphatidylinositol 3-phosphate and phosphatidylinositol 3,5-bisphosphate at D3-phosphate (20). MTMR2 is also highly expressed in the testis (22), and its level in the testis is significantly induced during maturation or during Sertoli-germ cell AJ assembly using cells cocultured in vitro (22, 23). Furthermore, a case study in which a CMT 4B patient was azoospermic (18), along with MTMR2–/– mice being found to have defects in retaining spermatids in the epithelium, have illustrated its significance in spermatogenesis. Interestingly, MTMR2–/– mice at 4–6 months of age were azoospermic, in which more than 90% of the tubules were devoid of elongating/elongate and round spermatids in the seminiferous epithelium, leading to a significant reduction in testicular weight and size by approximately 30–40% vs. normal mice (24). Herein MTMR2 was shown to be induced significantly during AJ disruption after AF-2364 or TE implant treatment, and MTMR2 was localized to the apical ES in the epithelium, further supporting the notion that it may play a crucial role in ES dynamics during spermatogenesis. In the brain, or kidney 293 cells, at least three putative binding partners of MTMR2 have been identified thus far. First, MTMR5 (26), a catalytically inactive MTM protein, may function as an adaptor and/or enhancer that guides MTMR2 to the specific subcellular localization as well as enhances its phosphatase activity by acting as a binding partner (26). Other studies have shown that MTMR5–/– mice were azoospermic, similar to MTMR2–/– mice, where germ cells lost their ability to attach to the seminiferous epithelium (51), suggesting this MTMR2/MTMR5 protein complex may be an important regulator of AJ dynamics in the testis. Second, neurofilament light chain protein, a neurofilament protein that associates with CMT 2E, a neuropathological condition typified by axonal losses, is structurally associated with MTMR2, and they appear to be crucial for signal transmission along the nerve fibers (27). Third, Dlg1 (also known as synapse-associated protein-97) (24), a scaffolding molecule with a putative PDZ (PSD95/DLG/ZO-1) domain, is expressed in brain and testes, as well as in several junction types (52). Interestingly, in studies by co-IP to identify the putative binding partner(s) of MTMR2 in the testis, it was shown that lipid phosphatase MTMR2 interacted with nonreceptor protein tyrosine kinase c-Src instead of the adaptor subunit p85 of PI 3-K. This is somewhat unexpected because PI 3-K is a lipid and protein kinase and is apparently a natural interacting partner of MTMR2. Yet a recent report has shown that a substrate-trapped mutant of MTM1 (MTM1 belongs to the MTM family having the highest homology with MTMR2), but not the wild-type MTM1, can interact with VPS 34, a class III PI 3-K (53), supporting the notion that PI 3-K is not the likely natural interacting partner of MTMR2. Interestingly, proteins that are Tyr phosphorylated by exogenously expressed c-Src cannot be reversely dephosphorylated when coexpressing with MTM1, yet they can be dephosphorylated by other protein phosphatases, such as PTP-PEST (PTP rich in Pro, Glu, Ser and Thr) and TC-PTP (T-cell PTP), in yeasts (53). This seemingly suggests that MTMs, such as MTM1, may not use c-Src as a binding partner to form a regulatory unit similar to the MTMR2-c-Src complex identified in the rat testis. However, the interaction of MTM and other kinases may be regulated either via a lipid second messenger, such as phosphatidylinositol 5-phosphate, a product of MTMR2 that may increase the enzymatic activity of MTMs (20) or a yet-to-be identified adaptor. Another surprising finding is that because MTMR2 has a PDZ-binding domain near its C-terminus, and it can interact with the PDZ domain of Dlg1 in COS-7 cells (24), it is expected to interact with other proteins having the PDZ domain at the ES. Yet MTMR2 failed to interact with zonula occludens 1 (ZO-1), another PDZ-domain-containing protein, as reported herein. Additionally, MTMR2 failed to interact with -actinin, another lipid-associated adaptor and an actin-binding and bundling protein. These findings also illustrate that among many adaptors and kinases that were identified in the apical ES, only c-Src physically interacts with MTMR2 in the seminiferous epithelium. This result obtained by co-IP was also confirmed by a colocalization study using fluorescent microscopy. Although MTMR2 failed to interact directly with the known adaptors at the apical ES, including ZO-1, -actinin, p130Cas, Wiskott-Aldrich syndrome protein (WASP), and others, but interacted only with c-Src, it must be noted that c-Src was recently shown to associate with some of these adaptors such as Wiskott-Aldrich syndrome protein and the cadherin-catenin protein complex (13). As such, MTMR2 can still indirectly interact with other adaptors to affect cell adhesion function of the cadherin-catenin complex via c-Src, and this possibility must be vigorously investigated in future studies.
The MTMR2/c-Src complex regulates ES dynamics likely via its interaction with the cadherin/catenin protein complex
The cadherin/catenin, the integrin/laminin, and the nectin/afadin/ponsin protein complexes are the three currently known structural protein complexes that confer cell adhesion function in the seminiferous epithelium at the apical ES site (for reviews, see Refs. 2 , 10). Other studies have shown that the phosphorylation status of these protein complexes, in particular the cadherin/catenin complex, is a determining factor to confer cell adhesion function (49). For instance, an increase in Tyr phosphorylation of the cadherin/catenin complex is associated with a loss of interaction between cadherin, catenin, and the actin cytoskeleton, leading to a loss of cell adhesion function (49, 50, 54). In earlier studies, an induction of N-cadherin and ß-catenin was detected in the AF-2364 model (32, 43, 44) at the time of germ cell loss, similar to the results reported herein during TE treatment-induced germ cell loss. Yet it is not known from these earlier studies whether any changes in the phosphorylation status of these proteins indeed occurred. Furthermore, the cadherin/catenin complex was localized to both the basal and apical ES sites (32, 43, 55, 56) and was recently shown to interact with the c-Src/Csk/CK II kinase complex (13). Because both MTMR2 and c-Src were induced during TE treatment at the time of germ cell loss and both MTMR2 and the phosphorylated form of Src (p-Src-Tyr416) were colocalized to the apical ES and to a lesser extent to the basal ES at stage VIII, together with the fact that c-Src is a binding partner of the cadherin/catenin protein complex, it seems logical to speculate that the cadherin/catenin protein complex may be the target of the MTMR2/c-Src complex. We have now shown that the N-cadherin/ß-catenin protein complex was disassociated in vivo during androgen depletion-induced AJ disruption, accompanied by an increase in Tyr phosphorylation of ß-catenin. This result is consistent with earlier reports in other in vitro systems (e.g. Madin Darby canine kidney cells) that a loss of cell adhesion function of the cadherin-based complex is mediated by an increase in ß-catenin phosphorylation (49, 54). Interestingly, a decrease in c-Src association with ß-catenin was also detected, suggesting that c-Src was moving away from ß-catenin once it phosphorylated its substrate. It is obvious that the 6ß1-integrin/laminin-3 complex can also be another target of the MTMR2/c-Src complex. For instance, it was recently shown that phospho-focal adhesion kinase-Tyr397 could bind to ß1-integrin, vinculin, and c-Src (38, 46), suggesting c-Src may be the linker that associates MTMR2 with ß1-integrin indirectly via phospho-focal adhesion kinase. There are also reports illustrating that Src and another lipid/protein phosphatase, PTEN (phosphatase and tensin homolog deleted on chromosome ten), can regulate integrin signaling (57, 58), supporting the notion that MTMR2/c-Src may regulate the integrin/laminin complex at the apical ES. In short, results of this study have clearly demonstrated that c-Src is a putative binding partner of MTMR2 in the testis, in particular at the apical ES, and the cadherin/catenin protein complex is one of its targets.
MTMR2 and c-Src can be induced by T in Sertoli-germ cell cocultures in vitro
Results of the in vivo study presented herein suggest that androgen plays a crucial role in regulating both MTMR2 and c-Src in the testis. To further confirm this observation, the levels of these proteins were quantified in Sertoli-germ cells cocultured in vitro. It was noted that both MTMR2 and c-Src were indeed induced by T in the cocultures, and this dose-dependent induction could be blocked by CPA, an antiandrogen (for a review, see Ref. 59). These data thus demonstrate unequivocally that both MTMR2 and c-Src, as well as cadherin as shown in an earlier study (32), were androgen-regulated proteins. However, much research is needed to understand how androgens regulate the interaction between these proteins, adaptors, and the cadherin/catenin complex.
Issues on changes in levels of target proteins (e.g. kinases, phosphatases, and adaptors) in the seminiferous epithelium as a result of germ cell loss
An interesting issue arises regarding results obtained from this androgen suppression model and the AF-2364 model that correlates changes in target proteins, such as MTMR2 and c-Src, with the event of germ cell loss from the seminiferous epithelium pertinent to AJ disruption. For instance, during the progressive loss of germ cells from the epithelium, the distribution of different cell types in the testis is altered at different time points in both models. This may confound our interpretation, at least to some extent, regarding changes in specific target proteins and/or genes by quantifying their steady-state protein and/or mRNA levels (for a review, see Ref. 60). For instance, if a protein, such as rMTMR2, is also produced by germ cells in the seminiferous epithelium (22), an induction of this protein during germ cell loss as quantified by immunoblotting was likely an underestimate because the cytosol that was being analyzed had been contributed largely by Sertoli cells and other somatic cells (e.g. peritubular myoid and Leydig cells) instead of germ cells as in normal testes, even though the same amount of protein was being analyzed (and the blot was reprobed with an anti-actin antibody staining to assess equal protein loading). However, this is an inherent problem with these models, which is somewhat difficult to resolve. First, although we had done some preliminary kinetic studies to assess the loss of different germ cell types (e.g. elongate, elongating, and round spermatids and spermatocytes), after AF-2364 treatment, which showed that elongate/elongating spermatids were depleted from the epithelium much more rapidly than round spermatids and spermatocytes (6.5 h vs. 3–6 d) (44), we do not have detailed morphometric analysis results on the time-dependent loss of specific germ cell types in both models to correlate these changes to each treatment regimen in either model. Even if these studies were completed, each animal would still respond differently, and there would have been variations between animals responding to the same treatment. Second, although germ cells were depleted from the epithelium, they probably would still be included in the immunoblot analysis because many germ cells would still be in the tubule lumen, in particular in the early phase of germ cell loss. As such, it is important to use other approaches to confirm changes of a specific target protein (e.g. MTMR2) in the epithelium during germ cell loss, such as with the use of immunohistochemistry and fluorescent microscopy as reported herein.
Another concern related to the data interpretation using these models is that an induction of a target protein, either a kinase, a phosphatase, or an adaptor, is in fact the result of AJ disruption rather than a reflection of germ cell loss. The best argument against this possibility is the result of studies using inhibitors. For instance, it was shown that Sertoli-germ cell AJ dynamics are regulated by the ß1-integrin/RhoB/Rho-associated protein kinase/LIMK (Lin-11 Isl-1 Mec3 kinase)/cofilin signaling pathway (61) and the ß1-integrin/laminin-3/membrane-type 1 matrix metalloproteinase/matrix metalloproteinase 2 protein complex (46) because these proteins were significantly induced during AF-2364-mediated germ cell loss from the epithelium. Indeed, the use of either Y27632, a specific inhibitor for Rho-associated protein kinase or 2R-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid, an inhibitor against matrix metalloproteinase 2, could significantly delay the AF-2364-induced germ cell loss from the epithelium (46, 61), validating these interpretations. Likewise, an induction of the MTMR2/c-Src protein complex as reported herein can indeed induce an increase in ß-catenin phosphorylation, which plausibly leads to a loss of protein-protein interactions between N-cadherin and ß-catenin protein as confirmed by experiments using co-IP, thereby inducing germ cell loss from the epithelium.
Summary and concluding remarks
In summary, we have demonstrated that MTMR2 is a stage-specific lipid phosphatase that is structurally associated with c-Src and that both proteins are colocalized to the apical ES at stage VIII of the epithelial cycle before spermiation. It is likely that the MTMR2/c-Src is an important protein complex that regulates the cell adhesion function of the cadherin/catenin complex via a lipid second messenger and possibly other adaptor(s) and/or kinases(s) by altering protein-protein interactions of the structural protein complexes at the ES site, facilitating cell movement during spermatogenesis.
|
Footnotes |
|---|
First Published Online December 9, 2004
Abbreviations: AJ, Adherens junction; BW, body weight; CK, casein kinase; CMT, Charcot-Marie-Tooth disease; CPA, cyproterone acetate; Csk, C-terminal Src kinase; DAPI, 4',6-diamidino-2-phenylindole; Dlg1, disc large 1; E, estradiol-17ß; ES, ectoplasmic specialization; FITC, fluorescein isothiocyanate; IP, immunoprecipitation; MTM, myotubularin; MTMR2, myotubularin-related protein 2; NR, spontaneous natural recovery; PI 3-K, phosphatidylinositol 3-kinase; PTP, protein tyrosine phosphatase, r, rat; T, testosterone; T, total gel concentration (g/100 ml) = acrylamide + methylene bisacrylamide; TE, testosterone-estradiol; ZO-1, zonula occludens 1.
Received September 7, 2004.
Accepted for publication December 1, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-macroglobulin, and testins: biosynthesis and differential regulation by germ cells. Mol Cell Endocrinol 89:127–140[CrossRef][Medline]
3 protein complex in the regulation of ectoplasmic specialization dynamics in the rat testis. Biol Reprod 70:945–964This article has been cited by other articles:
 |
|
 |
|
  W.-Y. Lui and W. M Lee Molecular mechanisms by which hormones and cytokines regulate cell junction dynamics in the testis J. Mol. Endocrinol., August 1, 2009; 43(2): 43 - 51. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Mruk, B. Silvestrini, and C. Y. Cheng Anchoring Junctions As Drug Targets: Role in Contraceptive Development Pharmacol. Rev., June 1, 2008; 60(2): 146 - 180. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Tash, R. Chakrasali, S. R. Jakkaraj, J. Hughes, S. K. Smith, K. Hornbaker, L. L. Heckert, S. B. Ozturk, M. K. Hadden, T. G. Kinzy, et al. Gamendazole, an Orally Active Indazole Carboxylic Acid Male Contraceptive Agent, Targets HSP90AB1 (HSP90BETA) and EEF1A1 (eEF1A), and Stimulates Il1a Transcription in Rat Sertoli Cells Biol Reprod, June 1, 2008; 78(6): 1139 - 1152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Koduri, S. A. Hild, L. Pessaint, J. R. Reel, and B. J. Attardi Mechanism of Action of l-CDB-4022, a Potential Nonhormonal Male Contraceptive, in the Seminiferous Epithelium of the Rat Testis Endocrinology, April 1, 2008; 149(4): 1850 - 1860. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Cheng, S. C. Watkins, and W. H. Walker Testosterone Activates Mitogen-Activated Protein Kinase via Src Kinase and the Epidermal Growth Factor Receptor in Sertoli Cells Endocrinology, May 1, 2007; 148(5): 2066 - 2074. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Aivatiadou, E. Mattei, M. Ceriani, L. Tilia, and G. Berruti Impaired Fertility and Spermiogenetic Disorders with Loss of Cell Adhesion in Male Mice Expressing an Interfering Rap1 Mutant Mol. Biol. Cell, April 1, 2007; 18(4): 1530 - 1542. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Xia, D. D Mruk, W. M Lee, and C Y. Cheng Unraveling the molecular targets pertinent to junction restructuring events during spermatogenesis using the Adjudin-induced germ cell depletion model J. Endocrinol., March 1, 2007; 192(3): 563 - 583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Goossens, B. Janssens, G. Vanpoucke, R. De Rycke, J. van Hengel, and F. van Roy Truncated isoform of mouse {alpha}T-catenin is testis-restricted in expression and function FASEB J, March 1, 2007; 21(3): 647 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. P Y Lee, D. D Mruk, W. Xia, and C Y. Cheng Cellular localization of sphingomyelin synthase 2 in the seminiferous epithelium of adult rat testes J. Endocrinol., January 1, 2007; 192(1): 17 - 32. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. P Y Lie, W. Xia, C. Q F Wang, D. D Mruk, H. H N Yan, C.-h. Wong, W. M Lee, and C Y. Cheng Dynamin II interacts with the cadherin- and occludin-based protein complexes at the blood-testis barrier in adult rat testes J. Endocrinol., December 1, 2006; 191(3): 571 - 586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. H. N. Yan and C. Y. Cheng Laminin {alpha} 3 Forms a Complex with beta3 and {gamma}3 Chains That Serves as the Ligand for {alpha} 6beta1-Integrin at the Apical Ectoplasmic Specialization in Adult Rat Testes J. Biol. Chem., June 23, 2006; 281(25): 17286 - 17303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. H. N. Yan and C. Y. Cheng Blood-testis barrier dynamics are regulated by an engagement/disengagement mechanism between tight and adherens junctions via peripheral adaptors PNAS, August 16, 2005; 102(33): 11722 - 11727. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.-Y. Lui, Y. H. Cheng, D. D Mruk, C. H. Cheng, M. Y. Mo, W. M Lee, and C Y. Cheng {alpha}2-Macroglobulin expression in the liver in response to inflammation is mediated by the testis J. Endocrinol., June 1, 2005; 185(3): 497 - 505. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
