The gelsolin-phosphoinositide pathway may be part of the normal mechanism by which Sertoli cells regulate sperm release and turnover of the blood-testis barrier. The intercellular adhesion complexes (ectoplasmic specializations)involved with these two processes are tripartite structures consisting of the plasma membrane, a layer of actin filaments and a cistern of endoplasmic reticulum. Gelsolin is concentrated in these adhesion complexes. In addition,phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2) and phosphoinositide-specific phospholipase C are found in the structures. Treatment of isolated spermatid/junction complexes with exogenous phosphoinositide-specific phospholipase C, or with a synthetic peptide consisting of the PtdIns(4,5)P2 binding region of gelsolin, results in the release of gelsolin and loss of actin from the adhesion complexes. We present a model for the disassembly of the actin layer of the adhesion complex that involves the hydrolysis of PtdIns(4,5)P2 resulting in the release of gelsolin within the plaque. Further, we speculate that the hydrolysis of PtdIns(4,5)P2 may result in a local Ca2+ surge via the action of inositol triphosphate on junctional endoplasmic reticulum. This Ca2+ surge would facilitate the actin severing function of gelsolin within the adhesion complex.

Ectoplasmic specializations are actin filament-containing adhesion complexes found at sites of intercellular attachment in the seminiferous epithelium of the testis. They are present only in Sertoli cells and occur at sites of attachment to spermatids in apical regions of the epithelium, and as part of the junction network between neighboring Sertoli cells in basal regions of the epithelium (Fig. 1A). Tight junctions within this basal junction network form the blood-testis barrier. Ectoplasmic specializations are morphologically characterized by the Sertoli cell plasma membrane, a sub-membrane plaque of actin filaments and an attached cistern of endoplasmic reticulum(Fig. 1B,C). The actin filaments are hexagonally packed into bundles oriented parallel to the plasma membrane.

Fig. 1.

Position of ectoplasmic specializations in the seminiferous epithelium of the testis and immunolocalization of gelsolin to the actin component of the structures. (A) Ectoplasmic specializations are present only in Sertoli cells and occur apically at sites of adhesion to spermatids and basally at sites of adhesion to neighboring Sertoli cells. (B,C) Typical appearance of ectoplasmic specializations in transmission electron micrographs. The structures consist of the plasma membrane of the Sertoli cell, a layer of actin filaments and a cistern of endoplasmic reticulum. The junctions shown here are from sites of Sertoli cell attachment to spermatids in the ground squirrel testis. Bars, 100 nm. (D) Immunofluorescence localization of gelsolin to ectoplasmic specializations in frozen sections of perfusion fixed rat testis. Sections were treated with a primary monoclonal antibody to gelsolin and with a secondary antibody to conjugated to Texas Red. Actin filaments were labeled with fluorescent phalloidin. Within the seminiferous epithelium, gelsolin and actin are co-localized at ectoplasmic specializations. Sites of apical and basal ectoplasmic specializations are indicated by the `a' and `b',respectively, in the panel labeled for actin. Specific staining for gelsolin was not observed in any of the controls (not shown). Bar, 10 μm. (E)Immunoelectron microscopic localization of gelsolin to the actin zone of ectoplasmic specializations. Spermatids with attached ectoplasmic specializations were mechanically dissociated from perfusion-fixed testes and treated with a primary antibody to gelsolin and a secondary antibody conjugated to nanogold. The material was embedded and sectioned, and then the sections were silver enhanced and stained. Shown here is an ectoplasmic specialization attached to a spermatid head. Notice that silver grains (small arrows) are associated with the actin zone of the junction plaque. Bar, 500 nm.

Fig. 1.

Position of ectoplasmic specializations in the seminiferous epithelium of the testis and immunolocalization of gelsolin to the actin component of the structures. (A) Ectoplasmic specializations are present only in Sertoli cells and occur apically at sites of adhesion to spermatids and basally at sites of adhesion to neighboring Sertoli cells. (B,C) Typical appearance of ectoplasmic specializations in transmission electron micrographs. The structures consist of the plasma membrane of the Sertoli cell, a layer of actin filaments and a cistern of endoplasmic reticulum. The junctions shown here are from sites of Sertoli cell attachment to spermatids in the ground squirrel testis. Bars, 100 nm. (D) Immunofluorescence localization of gelsolin to ectoplasmic specializations in frozen sections of perfusion fixed rat testis. Sections were treated with a primary monoclonal antibody to gelsolin and with a secondary antibody to conjugated to Texas Red. Actin filaments were labeled with fluorescent phalloidin. Within the seminiferous epithelium, gelsolin and actin are co-localized at ectoplasmic specializations. Sites of apical and basal ectoplasmic specializations are indicated by the `a' and `b',respectively, in the panel labeled for actin. Specific staining for gelsolin was not observed in any of the controls (not shown). Bar, 10 μm. (E)Immunoelectron microscopic localization of gelsolin to the actin zone of ectoplasmic specializations. Spermatids with attached ectoplasmic specializations were mechanically dissociated from perfusion-fixed testes and treated with a primary antibody to gelsolin and a secondary antibody conjugated to nanogold. The material was embedded and sectioned, and then the sections were silver enhanced and stained. Shown here is an ectoplasmic specialization attached to a spermatid head. Notice that silver grains (small arrows) are associated with the actin zone of the junction plaque. Bar, 500 nm.

Molecular components of ectoplasmic specializations include α6β1 integrin (Palombi et al.,1992; Salanova et al.,1995), vinculin (Grove and Vogl, 1989), fimbrin (Grove and Vogl, 1989), α-actinin(Franke et al., 1978), espin(Bartles et al., 1996), and myosin VIIa (Hasson et al.,1997). Integrin-linked kinase (ILK) also is present at the sites,whereas focal adhesion kinase (FAK) is not(Mulholland et al., 2001), nor is myosin II (Vogl and Soucy,1985).

Turnover of ectoplasmic specializations is related to two changes in intercellular adhesion that are fundamental to the process of spermatogenesis. At basal sites, turnover is correlated with the loss of attachment between adjacent Sertoli cell plasma membranes and the movement of spermatocytes from basal to adluminal compartments of the epithelium(Russell, 1977). At apical sites, disassembly is associated with sperm release(Russell, 1984). Little is known about how the structures are regulated or how the three elements (plasma membrane, actin filaments, endoplasmic reticulum) of the structures are functionally interrelated.

Here we report that gelsolin is a component of ectoplasmic specializations. In addition, we report that phosphatidylinositol 4,5-bisphosphate(PtdIns(4,5)P2) and phosphoinositide-specific phospholipase C (PLCγ) are present in the structures. Treatment of isolated ectoplasmic specializations with exogenous PLCγ or with a synthetic peptide of the PtdIns(4,5)P2 binding region of gelsolin results in the release of gelsolin and loss of filamentous actin from the adhesion junctions. Our results support a model for the disassembly of junction-related actin filaments during sperm release and turnover of the blood-testis barrier that involves the gelsolin-phosphoinositide pathway. Moreover, we include in our model the possibility that the endoplasmic reticulum component of ectoplasmic specializations may participate in gelsolin-mediated filament disassembly by regulating Ca2+ levels within the filament layer.

Chemicals and reagents

Unless otherwise indicated, most reagents used in the study were from Sigma Chemical Co. (St Louis, MO).

Animals

Animals used in this study were reproductively active Sprague-Dawley rats and New Zealand White rabbits. They were acquired and maintained in accordance with guidelines established by the Canadian Council on Animal Care.

Immunofluorescence

For immunolocalization of gelsolin, testes were perfusion (rat) or immersion (rabbit) fixed with 3% paraformaldehyde in PBS (150 mM NaCl, 5 mM KCl, 0.8 mM KH2PO4, 3.2 mM Na2HPO4, pH 7.3) and then cryosectioned. Sections were single or double labeled with Alexa 488 phalloidin (Molecular Probes, Eugene,OR) for filamentous actin and with mouse monoclonal antibodies generated against gelsolin (0.0625 μg/ml Sigma antibody; 0.0049 μg/ml Transduction Laboratories antibody) (Sigma; BD Transduction Laboratories, Mississauga, ON). Secondary antibodies consisted of goat anti-mouse IgG conjugated to Texas Red. Controls included replacing primary antibodies with equivalent concentrations of normal mouse IgG, replacing primary antibody with buffer alone, and replacing both primary and secondary antibodies with buffer alone.

For immunolocalization of PLCγ, fixed frozen sections were single or double labeled with 50 μg/ml mouse anti-Phospholipase Cγ IgG(Transduction Laboratories) and with Alexa 568 phalloidin (Molecular Probes,Eugene, OR). Secondary antibodies consisted of goat anti-mouse IgG conjugated to Alexa 488 (Molecular Probes). Controls were similar to those described for gelsolin immunostaining.

To immunolocalize PtdIns(4,5)P2, rat testes were perfusion fixed with 3% paraformaldehyde in PBS containing 2 mM EGTA. The tissue was cut into small pieces and then spermatids with attached junction plaques were mechanically dissociated from the epithelium by asperating the pieces through a graded series of syringe needles. Large fragments were allowed to settle and spermatids with attached ectoplasmic specializations that were still in suspension were removed and attached to polylysine-coated slides. These spermatid/junction complexes were then single- or double-labeled with 10 μg/ml of purified mouse monoclonal anti-PtdIns(4,5)P2 IgM (Echelon Research Laboratories,Salt Lake City, UT) and with Alexa 568 phalloidin (Molecular Probes). Secondary antibodies consisted of goat anti-mouse IgM conjugated to Alexa 488(Molecular Probes). Controls (not shown) included replacing the specific antibody with the equivalent concentration of normal mouse IgM, replacing the primary antibody with buffer alone, or replacing both primary and secondary antibodies with buffer alone.

Immunoelectron microscopic localization of Gelsolin

Testes were perfusion (rat) or immersion (rabbit) fixed with 3%paraformaldehyde in PBS. The tissue was cut into small pieces and then spermatids with attached junction plaques were mechanically dissociated from the epithelium by asperating the pieces through a graded series of syringe needles. Large fragments were allowed to settle and then were removed. Cells remaining in solution were concentrated by centrifugation, treated with 50 mM glycine, and then labeled first with a primary antibody to gelsolin (167μg/ml Sigma antibody; 23 μg/ml Transduction Laboratories antibody) and then with a secondary goat anti-mouse antibody (1:40 dilution) conjugated to nanogold (1.4 nm) (Nanoprobes, Yaphank, NY). Controls were similar to those for immunofluorescence. The material was embedded in Unicryl (British BioCell International, Cardif, UK) and all sections treated with a silver enhancement system (HQ SILVER, Nanoprobes).

Peptide competition experiments

Testicular fractions enriched for spermatids with attached ectoplasmic specializations were isolated generally as described elsewhere(Miller et al., 1999). The method involved manually collecting epithelia from seminiferous tubules in PEM buffer (80 mM Pipes, 1.0 mM EGTA, 1.0 mM MgCl2, pH 6.8) containing 250 mM sucrose, 10 μg/ml soybean trypsin inhibitor, 0.5 μg/ml leupeptin,0.5 μg/ml pepstatin, 0.1 mM PMSF, and then mechanically fragmenting the material by asperation through syringe needles. The fragments were loaded onto three-step sucrose gradients (30-60% sucrose in PEM buffer) and then the gradients were centrifuged. The fractions were resuspended in cold buffer (3 mM EGTA, 25 mM Hepes, 80 mM KCl, 0.5 mM DTT, pH 7.0) and then the suspension was divided into four equal volumes that were incubated on ice for 10 minutes. Following this, the cells were pelleted by centrifugation and then resupended in carrier buffer alone, or carrier buffer containing 40 μM of control peptides (QRLFGKDEL or FRVKLKQGQR) or the PtdIns(4,5)P2binding region of gelsolin (QRLFQVKGRR) directly conjugated to rhodamine B(Cunningham et al., 1996). The control peptide QRLFGKDEL consisted of the first four residues of the specific peptide followed by residues thought to localize the peptide to the ER. The other control peptide had the same sequence as the specific peptide, but in random order. The specific peptide consisted of residues 160-169 of human plasma gelsolin with rhodamine B conjugated to the N-terminus. CaCl2 was adjusted in each tube to result in a calculated 10 μM free Ca2+. The suspensions were incubated at 37°C for 15 minutes with gentle agitation every 30 seconds. Cells were pelleted by centrifugation and equivalent volumes of supernatant collected from each tube and relative actin concentrations compared qualitatively by immunoblots using a monoclonal anti-actin antibody (Sigma) used at 0.01 mg/ml. To access qualitatively the amount of gelsolin in the supernatants, we stripped the blots probed for actin and re-probed them using a polyclonal rabbit anti-mouse gelsolin antibody (gift from Dr Toshi Azuma, Brigham and Women's Hospital,Boston, Massachusetts) used at 1:2000. Unlike the antibodies used for the morphological work, this probe reacted with rat gelsolin on Westerns.

To determine whether the three synthetic peptides could bind to ectoplasmic specializations, spermatids with attached ectoplasmic specializations were treated for 30 minutes with buffer alone (3 mM EGTA, 25 mM Hepes, 80 mM KCl,0.5 mM DTT, pH 6.5) or with buffer containing 20 μM synthetic peptides. The cells were washed and then examined with a fluorescence microscope.

PLCγ experiments

Testicular fractions enriched for spermatids with attached ectoplasmic specializations were obtained as described for the peptide competition experiments, pooled, and then diluted in 1 ml of buffer (3 mM EGTA, 25 mM Hepes, 80 mM KCl, 0.5 mM DTT) not containing Ca2+. Equal volumes of suspensions were added to the required number of treatment tubes and then the cells pelleted by centrifugation. The supernatants were discarded and the cells in each of four tubes were resuspended in 500 μl buffer containing a calculated 11 μM free Ca2+ (3 mM EGTA, 25 mM Hepes, 80 mM KCl,0.5 mM DTT, 2.92 mM CaCl2). Cells to be used in the no Ca2+ control were resuspended in 500 μl buffer containing no Ca2+ (3 mM EGTA, 25 mM Hepes, 80 mM KCl, 0.5 mM DTT). The cells were allowed to sit on ice for 10 minutes. Following this, the tubes again were centrifuged and the pellets resuspended in 100 μl of treatment buffers containing the appropriate calculated amounts of Ca2+ (11 μM or 1.5 mM) and PLCγ or buffer alone. The reaction mixtures were incubated for 15 minutes at 37°C with gentle agitation every 30 seconds. Following incubation, cells were pelleted by centrifugation and equivalent volumes of supernatant were removed from each tube and assayed, by immunoblot, for actin and gelsolin as described above for the peptide competition experiments.

While doing a screen for a number of candidate actin binding proteins that could be involved with actin dynamics in ectoplasmic specializations in the rat, we found that two different monoclonal antibodies to gelsolin reacted with apical and basal sites known to contain the unique adhesion complexes(Fig. 1D). Staining was dramatically co-distributed with actin filaments, as indicated by fluorescent phallotoxin staining, in double-labeled material. Results were confirmed at the ultrastructural level where staining was restricted to the region of the adhesion complex containing the actin layer(Fig. 1E). Staining in this region was present both on the plasma membrane and endoplasmic reticulum side of the actin layer, as well as within the layer itself. Although both antibodies reacted in a site-specific fashion on fixed frozen sections,neither reacted with a band corresponding to purified bovine gelsolin, nor with any other bands, on immunoblots of rat testis (data not shown). Using native gels did not improve immunoreactivity, nor did mild protease treatment or fixation of the blots. We repeated all the experiments using rabbit testis. Immunolocalization results at the light(Fig. 2A) and the electron microscopic (Fig. 2B) levels were identical to those obtained using rat tissue. Significantly, the two antibodies reacted with a single band that migrated slightly ahead of purified bovine plasma gelsolin on immunoblots of rabbit testis(Fig. 2C). Rabbit cytoplasmic gelsolin is known to be 25 amino acids smaller than human plasma gelsolin(Kwiatkowski et al., 1986). This result, together with the immunolocalization data, indicated to us that the antibodies were monospecific. Our results are generally consistent with a previous report of gelsolin in the apical cytoplasm of human Sertoli cells(Rousseauz-Prevost et al., 1997). We conclude that gelsolin is a major component of Sertoli cell ectoplasmic specializations and that it is localized to the actin-containing region between the plasma membrane and the endoplasmic reticulum.

Fig. 2.

Immunolocalization of gelsolin to ectoplasmic specializations in rabbit testis. (A) Fluorescence of fixed-frozen sections of rabbit testis labeled for actin and gelsolin. Apical and basal ectoplasmic specializations are indicated by the `a' and `b', respectively, in the actin panel. Bar, 10 μm. (B)Immunoelectron microscopic localization of gelsolin to the actin filament-containing region of an apical ectoplasmic specialization. The material was processed and labeled exactly as described in the legend toFig. 1. As in the rat, the silver grains (small arrows) are associated with the actin zone of the junction plaque. Bar, 250 nm. (C) Immunoblot of rabbit testis and purified bovine gelsolin (Sigma). The antibodies from Sigma and from Transduction Laboratories reacted with a single band on immunoblots of rabbit testis, and this band migrated slightly ahead of purified bovine plasma gelsolin. A similar band was not present on control blots in which the primary antibody was replaced with a similar concentration of normal mouse IgG (not shown). The lines indicate the top and bottom of the gel. Loading densities were 0.05μg of purified gelsolin and 100 μg of decapsulated testis homogenate.

Fig. 2.

Immunolocalization of gelsolin to ectoplasmic specializations in rabbit testis. (A) Fluorescence of fixed-frozen sections of rabbit testis labeled for actin and gelsolin. Apical and basal ectoplasmic specializations are indicated by the `a' and `b', respectively, in the actin panel. Bar, 10 μm. (B)Immunoelectron microscopic localization of gelsolin to the actin filament-containing region of an apical ectoplasmic specialization. The material was processed and labeled exactly as described in the legend toFig. 1. As in the rat, the silver grains (small arrows) are associated with the actin zone of the junction plaque. Bar, 250 nm. (C) Immunoblot of rabbit testis and purified bovine gelsolin (Sigma). The antibodies from Sigma and from Transduction Laboratories reacted with a single band on immunoblots of rabbit testis, and this band migrated slightly ahead of purified bovine plasma gelsolin. A similar band was not present on control blots in which the primary antibody was replaced with a similar concentration of normal mouse IgG (not shown). The lines indicate the top and bottom of the gel. Loading densities were 0.05μg of purified gelsolin and 100 μg of decapsulated testis homogenate.

The presence of gelsolin in ectoplasmic specializations has significant implications for the assembly and disassembly of the actin plaques. Gelsolin is a potent Ca2+-dependent actin severing and capping protein(Sun et al., 1999). The presence of gelsolin within the actin containing region of ectoplasmic specializations, structures that are stable during most of the long process of spermatogenesis, indicates to us that much of the protein may be `inhibited'until it is required for actin disassembly or reorganization. In addition to the lack of Ca2+, the only known inhibitors of the severing and capping functions of gelsolin are certain phospholipids(Janmey and Stossel, 1987;Meerschaert et al., 1998;Sun et al., 1999), the most notable of which is PtdIns(4,5)P2.

To determine whether PtdIns(4,5)P2is present in ectoplasmic specializations, we treated isolated spermatids, to which the adhesion complexes remained attached, with an antibody to the phospholipid. The antibody positively reacted with regions that also labeled with probes for junction related actin filaments (Fig. 3A). Similar staining was not observed when normal mouse IgG was substituted for primary antibody. We also labeled fixed sections of rat testis with antibodies to phosphoinositide-specific phospholipase C (PLCγ). This probe specifically labeled regions of the seminiferous epithelium known to contain both basal and apical ectoplasmic specializations, and was stage specific. At stages when the junctions are stable, staining at the junctions was weak (Fig. 3B). Significantly, staining was most intense during the period of spermatogenesis(stage VII in rat) when the adhesion complexes are disassembling apically and turning over basally (Fig. 3C). The antibody reacted specifically with one band on blots of rat testis and rat seminiferous epithelium (Fig. 3D).

Fig. 3.

Immunolocalization of PtdIns(4,5)P2 and PLCγ in rat ectoplasmic specializations. (A) Phase and fluorescence images of isolated spermatids with attached ectoplasmic specializations that have been labeled with an antibody raised against PtdIns(4,5)P2. The spermatids have also been treated with fluorescent phallotoxin to label actin. Notice that the probe for PtdIns(4,5)P2 stains the region surrounding the head containing an ectoplasmic specialization that labels with the probe for actin. Specific staining for PtdIns(4,5)P2was not observed in any of the controls (not shown). Bar, 5 μm. (B,C)Immunofluorescence localization of PLCγ in fixed frozen sections of rat seminiferous epithelium at stage V (B) and stage VII (C) of spermatogenesis. The locations of apical and basal ectoplasmic specializations are indicated by the `a' and `b', respectively, in the actin panels. At stage V, the probe for PLCγ reacts weakly at ectoplasmic specializations. The situation is much different at stage VII when apical ectoplasmic specializations are disassembling as part of the sperm release process and basal ectoplasmic specializations are turning over to allow the next generation of spermatocytes through junction complexes between Sertoli cells. At this stage, apical and basal regions containing ectoplasmic specializations, indicated by the actin staining, also react with the probe for PLCγ. Specific staining was not observed in any of the controls (data not shown). The intense staining of the tubule wall (asterisk) is caused by nonspecific staining by the secondary antibody. Bar, 10 μm. (D) Immunoblot of rat testis and rat seminiferous epithelium. The PLCγ antibody reacts specifically with a single band in each lane (∼148 kDa). The minor bands indicated by the asterisk are nonspecific and are present in blots treated with normal mouse IgG instead of primary antibody. The lines indicate the top and bottom of the gel. Loading densities were 80 μg of testis homogenate and 80 μg of seminiferous epithelium.

Fig. 3.

Immunolocalization of PtdIns(4,5)P2 and PLCγ in rat ectoplasmic specializations. (A) Phase and fluorescence images of isolated spermatids with attached ectoplasmic specializations that have been labeled with an antibody raised against PtdIns(4,5)P2. The spermatids have also been treated with fluorescent phallotoxin to label actin. Notice that the probe for PtdIns(4,5)P2 stains the region surrounding the head containing an ectoplasmic specialization that labels with the probe for actin. Specific staining for PtdIns(4,5)P2was not observed in any of the controls (not shown). Bar, 5 μm. (B,C)Immunofluorescence localization of PLCγ in fixed frozen sections of rat seminiferous epithelium at stage V (B) and stage VII (C) of spermatogenesis. The locations of apical and basal ectoplasmic specializations are indicated by the `a' and `b', respectively, in the actin panels. At stage V, the probe for PLCγ reacts weakly at ectoplasmic specializations. The situation is much different at stage VII when apical ectoplasmic specializations are disassembling as part of the sperm release process and basal ectoplasmic specializations are turning over to allow the next generation of spermatocytes through junction complexes between Sertoli cells. At this stage, apical and basal regions containing ectoplasmic specializations, indicated by the actin staining, also react with the probe for PLCγ. Specific staining was not observed in any of the controls (data not shown). The intense staining of the tubule wall (asterisk) is caused by nonspecific staining by the secondary antibody. Bar, 10 μm. (D) Immunoblot of rat testis and rat seminiferous epithelium. The PLCγ antibody reacts specifically with a single band in each lane (∼148 kDa). The minor bands indicated by the asterisk are nonspecific and are present in blots treated with normal mouse IgG instead of primary antibody. The lines indicate the top and bottom of the gel. Loading densities were 80 μg of testis homogenate and 80 μg of seminiferous epithelium.

As an alternative approach to verifying the presence of PtdIns(4,5)P2 at the junction plaque, we labeled unfixed spermatid/junction complexes, in the absence of Ca2+, with a synthetic peptide of the PtdIns(4,5)P2 binding domain of gelsolin directly conjugated to rhodamine B(Cunningham et al., 1996). Staining of regions known to contain ectoplasmic specializations in controls was weak or absent (Fig. 4A-C,A′-C′), whereas staining with the specific peptide was relatively intense (Fig. 4D,D′).

Fig. 4.

Peptide binding and competition experiments. Paired phase (A-D) and fluorescence (A′-D′) micrographs of spermatids with attached ectoplasmic specializations that were mechanically isolated from rat seminiferous epithelium and treated for 30 minutes with buffer (A,A′) or with buffer containing 20 μM synthetic peptides directly conjugated to rhodamine B. The peptides consisted of two control sequences (QRLFGKDEL(B,B′) and FRVKLKQGQR (C,C′) and the PtdIns(4,5)P2 binding region of gelsolin (QRLFQVKGRR(D,D′)). Note that the PtdIns(4,5)P2 binding region of gelsolin labels regions surrounding the spermatid head more strongly than the control peptides or buffer alone. Bar, 5 μm. (E) Shown here are the results of a peptide competition experiment. From left to right, the lanes are of supernatants collected from spermatids with attached ectoplasmic specializations treated with buffer alone, sequence QRLFGKDEL (Control Peptide 1), sequence FRVKLKQGQR (Control Peptide 2), and sequence QRLFQVKGRR (Gelsolin Peptide). The blots were probed with antibodies to actin and gelsolin. Notice that the amount of actin and gelsolin present in the supernatant from material treated with the PtdIns(4,5)P2 binding region of gelsolin(QRLFQVKGRR) is greater than in supernatants treated with control peptides or buffer alone.

Fig. 4.

Peptide binding and competition experiments. Paired phase (A-D) and fluorescence (A′-D′) micrographs of spermatids with attached ectoplasmic specializations that were mechanically isolated from rat seminiferous epithelium and treated for 30 minutes with buffer (A,A′) or with buffer containing 20 μM synthetic peptides directly conjugated to rhodamine B. The peptides consisted of two control sequences (QRLFGKDEL(B,B′) and FRVKLKQGQR (C,C′) and the PtdIns(4,5)P2 binding region of gelsolin (QRLFQVKGRR(D,D′)). Note that the PtdIns(4,5)P2 binding region of gelsolin labels regions surrounding the spermatid head more strongly than the control peptides or buffer alone. Bar, 5 μm. (E) Shown here are the results of a peptide competition experiment. From left to right, the lanes are of supernatants collected from spermatids with attached ectoplasmic specializations treated with buffer alone, sequence QRLFGKDEL (Control Peptide 1), sequence FRVKLKQGQR (Control Peptide 2), and sequence QRLFQVKGRR (Gelsolin Peptide). The blots were probed with antibodies to actin and gelsolin. Notice that the amount of actin and gelsolin present in the supernatant from material treated with the PtdIns(4,5)P2 binding region of gelsolin(QRLFQVKGRR) is greater than in supernatants treated with control peptides or buffer alone.

To test the hypothesis that gelsolin in ectoplasmic specializations may be bound to PtdIns(4,5)P2, we mechanically dissociated rat spermatids, to which ectoplasmic specializations of Sertoli cells remained attached, from the seminiferous epithelium and incubated the cells with a synthetic peptide of the PtdIns(4,5)P2 binding region of gelsolin or with PLCγ. In the first set of experiments, we predicted that the specific peptide would compete with endogenous gelsolin for binding to PtdIns(4,5)P2 and, in the presence of Ca2+,would result in increased actin disassembly when compared with controls. Relative to supernatants collected from cells treated with buffer alone or with two control peptides, more actin was present in blots of supernatants collected from cells treated with specific peptide(Fig. 4E). Importantly, when blots were re-probed with an antibody that recognizes rat gelsolin on western blots, more gelsolin was detected in supernatants collected from cells treated with specific peptide than in blots of supernatants treated with control peptides or buffer alone.

In the second set of experiments, we predicted that exogenously added PLCγ would hydrolyze PtdIns(4,5)P2 to inositol(1,4,5)-trisphosphate (Ins(1,4,5)P3) and diacylglycerol,thereby releasing gelsolin. In the presence of Ca2+, actin in the adhesion complexes associated with spermatid heads should disassemble and the amount of actin in solution should increase relative to controls. The buffer systems used contained calculated micromolar and millimolar levels of Ca2+. Millimolar levels were included in the design to swamp any effect of PLCγ treatment on Ca2+ release from junction-related ER that would indirectly activate any gelsolin not bound to PtdIns(4,5)P2. Following incubation, spermatid/junction complexes were pelletted by centrifugation and equivalent volumes of supernatants from experimental and control cells collected and analyzed, by immunoblotting, for actin. In some experiments, an equivalent volume of cells was removed from each tube prior to centrifugation and stained with fluorescent phallotoxin to label filamentous actin. Generally, less actin was visible in the adhesion complexes associated with cells treated with PLCγ in the presence of Ca2+ than in control cells. At mM levels, Ca2+ alone increased actin in supernatants relative to other controls. Significantly, treatment with PLCγ in the presence ofμM and mM Ca2+ increased the amount of actin in supernatants collected from spermatid/junction complexes relative to supernatants collected from all control cells (Fig. 5E). When the same blots of supernatants were re-probed for gelsolin, PLCγ in the absence of Ca2+ increased the level of gelsolin in supernatants without a corresponding increase in the amount of actin. At mM levels, Ca2+ alone increased the amount of gelsolin in supernatants. At μM levels, the effect was reduced and the amount of gelsolin present in supernatants was less than when samples were treated with PLCγ in the presence or absence of Ca2+.

Fig. 5.

PLCγ in the presence of Ca2+ results in the loss of filamentous actin from rat ectoplasmic specializations. Spermatids with attached ectoplasmic specializations were incubated in the presence or absence of either PLCγ in the presence and absence of Ca2+. Shown in panels A,A′ to D,D′ are paired phase and fluorescence micrographs of spermatids with attached adhesion complexes fixed and labeled with fluorescent phalloidin for filamentous actin immediately after isolation(A,A′) or after incubation with or without PLCγ in the presence or absence of a calculated 11 μM free Ca2+ (B,B′-D,D′)The least amount of filamentous actin is associated with cells treated with PLCγ in the presence of Ca2+. (E) Treatment of spermatid/adhesion complexes with PLCγ in the presence of Ca2+ resulted in increased levels of actin and gelsolin in supernatants, relative to controls, as assessed by immunoblot. In these experiments, spermatids with attached junction complexes were incubated in buffers with or without PLCγ in the presence or absence of a calculated 1.5 mM (upper two blots) or 11 μM free Ca2+ (lower two blots). Cells were removed from solution by centrifugation and equivalent volumes of supernatants assessed for actin and gelsolin on immunoblots. The amount of actin and gelsolin are greatest in supernatants from spermatid/junction complexes treated with both PLCγ and Ca2+. Significantly,treatment with PLCγ in the absence of Ca2+ results in increased gelsolin in supernatants, but not in increased actin.

Fig. 5.

PLCγ in the presence of Ca2+ results in the loss of filamentous actin from rat ectoplasmic specializations. Spermatids with attached ectoplasmic specializations were incubated in the presence or absence of either PLCγ in the presence and absence of Ca2+. Shown in panels A,A′ to D,D′ are paired phase and fluorescence micrographs of spermatids with attached adhesion complexes fixed and labeled with fluorescent phalloidin for filamentous actin immediately after isolation(A,A′) or after incubation with or without PLCγ in the presence or absence of a calculated 11 μM free Ca2+ (B,B′-D,D′)The least amount of filamentous actin is associated with cells treated with PLCγ in the presence of Ca2+. (E) Treatment of spermatid/adhesion complexes with PLCγ in the presence of Ca2+ resulted in increased levels of actin and gelsolin in supernatants, relative to controls, as assessed by immunoblot. In these experiments, spermatids with attached junction complexes were incubated in buffers with or without PLCγ in the presence or absence of a calculated 1.5 mM (upper two blots) or 11 μM free Ca2+ (lower two blots). Cells were removed from solution by centrifugation and equivalent volumes of supernatants assessed for actin and gelsolin on immunoblots. The amount of actin and gelsolin are greatest in supernatants from spermatid/junction complexes treated with both PLCγ and Ca2+. Significantly,treatment with PLCγ in the absence of Ca2+ results in increased gelsolin in supernatants, but not in increased actin.

The results presented here provide insight into the molecular mechanism of junction-related actin filament disassembly related to sperm release and turnover of basal junction networks between Sertoli cells. The mechanism may involve, at least in part, gelsolin and the phosphoinositide pathway. Gelsolin and PLCγ are concentrated in ectoplasmic specializations, and PtdIns(4,5)P2 is also present in the structures. In addition, treatment of isolated ectoplasmic specializations either with the PtdIns(4,5)P2 binding domain of gelsolin or with PLCγ in the presence of Ca2+ increases the disassembly of the junction-related actin filaments relative to controls. Significantly, both of these treatments result in an increase in the amount of free gelsolin. Although our results do not rule out the possibility that related proteins,such as scinderin, may be involved with turnover of cortical actin in Sertoli cells, as has been suggested (Pelletier et al., 1999), gelsolin is the first member of this group of actin-severing proteins to be localized specifically to ectoplasmic specializations.

The fact that the severing and capping functions of gelsolin are Ca2+ dependent (Yin and Stossel, 1979), together with the general finding that Ins(1,4,5)P3 stimulates the release of Ca2+from intracellular stores (Berridge,1993), may account for the presence of a cistern of endoplasmic reticulum as an integral part of the adhesion complex. The endoplasmic reticulum of this complex is suspected to regulate Ca2+ levels(Franchi and Camatini, 1985;Pelletier et al., 1999);however, strong experimental evidence that the cistern actually can sequester and release the cation locally is still lacking. It is possible that, in stable plaques, the ER may function to maintain low levels of local Ca2+ within the actin layer, thereby inhibiting the severing function of gelsolin, and perhaps of related proteins. The capping function of gelsolin can occur at lower Ca2+ concentrations than the severing function (Janmey et al., 1985). This observation may account for the presence of gelsolin among actin filaments in stable actin plaques, in addition to being located on either side of the actin layer where presumably the gelsolin could be bound to PtdIns(4,5)P2 in adjacent membranes. Gelsolin caps on actin filaments may account for the effect of increased Ca2+ noted in our experiments. During plaque disassembly, hydrolysis of PtdIns(4,5)P2 by PLCγ would not only release gelsolin bound to PtdIns(4,5)P2, but may generate a local surge in Ca2+ release from the ER, through the action of Ins(1,4,5)P3 (Fig. 6). This Ca2+ surge would stimulate the severing function of gelsolin within the actin plaque.

Fig. 6.

Putative model for gelsolin activation at the time of ectoplasmic specialization disassembly in Sertoli cells. The model involves hydrolysis of PIP2 (phosphatidylinositol 4,5-bisphosphate) resulting both in the release of gelsolin and a surge in local levels of Ca2+. Although PIP2 is shown in association with the plasma membrane, the model is not meant to exclude the possibility that PIP2 is also present in the membrane of the ER. IP3, inositol (1,4,5)-trisphosphate;IP3R, inositol (1,4,5)-trisphosphate receptor; PIP2,phosphatidylinositol 4,5-bisphosphate; PLC, phosphoinositide-specific phospholipase C.

Fig. 6.

Putative model for gelsolin activation at the time of ectoplasmic specialization disassembly in Sertoli cells. The model involves hydrolysis of PIP2 (phosphatidylinositol 4,5-bisphosphate) resulting both in the release of gelsolin and a surge in local levels of Ca2+. Although PIP2 is shown in association with the plasma membrane, the model is not meant to exclude the possibility that PIP2 is also present in the membrane of the ER. IP3, inositol (1,4,5)-trisphosphate;IP3R, inositol (1,4,5)-trisphosphate receptor; PIP2,phosphatidylinositol 4,5-bisphosphate; PLC, phosphoinositide-specific phospholipase C.

How actin filament disassembly, involving gelsolin, is initiated and coupled to signaling pathways associated directly with intercellular adhesion molecules in the plasma membrane remains to be determined. In addition to a role in filament disassembly, it is possible that gelsolin is involved with nucleation and elongation of actin filaments during junction assembly and during the changes in actin filament rearrangement that occur within ectoplasmic specializations during spermatogenesis. The finding that gelsolin is a molecular component of structures related to sites of intercellular adhesion in Sertoli cells may have general implications for the molecular mechanisms underlying assembly and disassembly of actin complexes associated with intercellular adhesion junctions generally in cells.

This work was funded by Operating Grant number MT-13389 awarded to A.W.V. by the Canadian Institutes of Health. We thank Calvin Roskelley for many helpful suggestions with the work and with the manuscript.

Bartles, J. R., Wierda, A. and Zheng, L.(
1996
). Identification and characterization of espin, an actin-binding protein localized to the F-actinrich junctional plaques of Sertoli cell ectoplasmic specializations.
J. Cell Sci.
109
,
1229
-1239.
Berridge, M. J. (
1993
). Inositol triphosphate and calcium signalling.
Nature
361
,
315
-325.
Cunningham, C. C., Janmey, P. A. and Vegners, R.(
1996
). Fluorescent derivatives of inositol lipid binding peptides. In
Peptides
(ed. R. Ramage and R. Epton),pp.
317
-318. Mayflower, London.
Franchi, E. and Camatini, M. (
1985
). Evidence that a Ca2+ chelator and a calmodulin blocker interfere with the sturcture of inter-Sertoli junctions.
Tissue Cell
17
,
13
-25.
Franke, W. W., Grund, C., Fink, A., Weber, K., Jokusch, B. M.,Zentgraf, H. and Osborn, M. (
1978
). Location of actin in the microfilament bundles associated with junctional specializations between Sertoli cells and spermatids.
Biol. Cell
31
,
7
-14.
Grove, B. D. and Vogl, A. W. (
1989
). Sertoli cell ectoplasmic specializations: a type of actin-associated adhesion junction?
J. Cell Sci.
93
,
309
-323.
Hasson, T., Walsh, J., Cable, J., Mooseker, M. S., Brown, S. D. M. and Steel, K. P. (
1997
). Effects of shaker-1 mutations on myosin-VIIa protein and mRNA expression.
Cell Motil. Cytoskel
.
37
,
127
-138.
Janmey, P. A., Chaponnier, C., Lind, S. E., Zaner, K. S.,Stossel, T. P. and Yin, H. L. (
1985
). Interactions of gelsolin and gelsolin-actin complexes with actin. Effects of calcium on actin nucleation, filament severing, and end blocking.
Biochemistry
24
,
3714
-3723.
Janmey, P. A. and Stossel, T. P. (
1987
). Modulation of gelsolin function by phosphatidylinositol 4,5-bisphosphate.
Nature
325
,
362
-364.
Kwiatkowski, D. J., Stossel, T. P., Orkin, S. H., Mole, J. E.,Colten, H. R. and Yin, H. L. (
1986
). Plasma and cytoplasmic gelsolins are encoded by a single gene and contain a duplicated actin-binding domain.
Nature
323
,
455
-458.
Meerschaert, K., De Corte, V., de Ville, Y., Vandekerckhove, J. and Gettermans J. (
1998
). Gelsolin and functionally similar actin-binding proteins are regulated by lysophosphatidic acid.
EMBO
17
,
5923
-5932.
Miller, M. G., Mulholland, D. J. and Vogl, A. W.(
1999
). Rat testis motor proteins associated with spermatid translocation (dynein) and spermatid flagella (kinesin-II).
Biol. Reprod.
60
,
1047
-1056.
Mulholland, D. J., Dedhar, S. and Vogl, A. W.(
2001
). Rat seminiferous epithelium contains a unique junction(Ectoplasmic specialization) with signaling properties both of cell/cell and cell/matrix junctions.
Biol. Reprod.
64
,
396
-407.
Palombi, F., Salanova, M., Tarone, G., Farini, D. and Stefanini,M. (
1992
). Distribution of β1 integrin subunit in rat seminiferous epithelium.
Biol. Reprod.
47
,
1173
-1182.
Pelletier, R.-M., Trifaro, J.-M., Carbajal, M. E., Okawara, Y. and Vitale, M. L. (
1999
). Calcium-dependent actin filanment-severing protein scinderin levels and localization in bovine testis,epididymis, and spermatozoa.
Biol. Reprod.
60
,
1128
-1136.
Rousseaux-Prevost, R., Delobel, B., Hermand, E., Rigot, J.-M.,Danjou, P., Mazeman, E. and Rousseaux, J. (
1997
). Distribution of gelsolin in human testis.
Mol. Reprod. Dev.
48
,
63
-70.
Russell, L. D. (
1977
). Movement of spermatocytes from the basal to the adluminal compartment of the rat testis.
Am. J. Anat.
148
,
313
-328.
Russell, L. D. (
1984
). Spermiation — the sperm release process: ultrastructural observations and unresolved problems. In
Ultrastructure of Reproduction
(ed. J. Van Blerkom and P. M. Modda), pp.
46
-66. Martinus Nijhoff Publishers.
Salanova, M., Stefanini, M., De Curtis, I. and Palombi, F.(
1995
). Integrin receptor α6β1 is localized at specific sites of cell-to-cell contact in rat seminiferous epithelium.
Biol. Reprod.
52
,
79
-87.
Sun, H. Q., Yamamaoto, M., Mejillano, M. and Yin, H. L.(
1999
). Gelsolin, a multifunctional actin regulatory protein.
J. Biol. Chem.
274
,
33179
-33182.
Vogl, A. W. and Soucy, L. J. (
1985
). Arrangement and possible function of actin filament bundles in ectoplasmic specializations of ground squirrel Sertoli cells.
J. Cell Biol.
100
,
814
-825.
Yin, H. L. and Stossel, T. P. (
1979
). Control of cytoplasmic actin gel-sol transformation by gelsolin, a calcium-dependent regulatory protein.
Nature
281
,
583
-586.