THE ANATOMICAL RECORD 244:155-164 (1996) Differential Expression of Placental (P)-Cadherin in Sertoli Cells and Peritubular Myoid Cells During Postnatal Development of the Mouse Testis LI-HSIEN LIN AND ROBERT M. DEPHILIP Department of Cell Biology, Neurobiology, and Anatomy, Ohio State University College of Medicine, Columbus, Ohio ABSTRACT Background: In previous work, RNA transcripts for placental (Pbcadherin, a calcium-dependent cell adhesion molecule, were identified in the rat testis during the first 4 weeks of postnatal development. However, the cells in the testis responsible for P-cadherin expression have not yet been identified. Methods: We used conventional epifluorescence microscopy to examine P-cadherin immunoreactivity in cryostat sections of mouse testis and scanning laser confocal microscopy to localize P-cadherin and p-catenin in wholemount preparations of mouse seminiferous tubules. We used fluorescent phalloidin to identify actin filaments. Results: Sertoli cells expressed P-cadherin on postnatal days 1,3, and 8, but not on any day thereafter. In contrast, peritubular cells did not express P-cadherin during the first week of development, but began to express P-cadherin on postnatal day 8 and continued expression in adulthood. p-Catenin was localized near contact areas between peritubular cells on postnatal days 12 and 15. A mature pattern of actin filament organization in peritubular cells appeared on day 15 and coincided with the uniform appearance of P-cadherin and p-catenin near areas of contact between adjacent peritubular cells. Conclusions: During postnatal development of the testis, the earlier expression of P-cadherin by Sertoli cells is replaced by the subsequent expression of P-cadherin by peritubular cells. The expression of P-cadherin in peritubular cells is correlated temporally with the expression of P-catenin and the development of a mature network of actin filaments and is consistent with a role in intercellular adhesion and junction formation. 0 1996 Wiley-Liss, Inc. Key words: Actin cytoskeleton, Catenin, Cell Adhesion, Seminiferous tubules P-cadherin belongs to a family of cell surface adhesion molecules that are characterized by their dependency on calcium for adhesive function (Takeichi, 1988). Like other “classical” cadherin molecules such as epithelial (El- and neural (N)-cadherin, P-cadherin was identified in functional cell adhesion assays and consists of a large extracellular domain that specifies adhesion, a transmembrane portion that fixes the molecule in the plasma membrane, and a cytoplasmic domain that interacts with actin microfilaments via linking molecules called catenins (Kemler, 1992). It has been suggested that the formation of a cadherin-catenin complex is necessary for full adhesion of cadherins (Ozawa et al., 1990). Cadherins play a “morpho-regulator’, function during development and contribute to the formation of body structure (Edelman, 1988). In this capacity, cadherins allow cells to recognize each other, aggregate, 0 1996 WILEY-LISS. INC. and form definitive structures. One such example of this function is seen during kidney development when mesenchymal cells of the lateral plate mesoderm condense, initiate E-cadherin expression, and participate with E-cadherin-expressing cells of the ureteric bud in nephron formation (Vestweber et al., 1985). Termination of cadherin expression also can have an effect as seen in the development of the neural tube, when certain ectodermal cells cease expression of E-cadherin and define the neural plate (Hatta and Takeichi, 1986). These cells subsequently express N-cadherin and form Received June 19, 1995; accepted September 27, 1995. Address reprint requests to Robert M. DePhilip, Department of Cell Biology, Neurobiology, and Anatomy, 4072 Graves Hall, Ohio State University, Columbus, OH 43210. 156 LIN AND DEPHILIP the neural tube. Cadherins also play a “mechanostatic” role in maintenance of adult structure (Ozawa et al., 1989). This function is seen most easily in epithelial cells, where cadherins are responsible for an initial adhesive event that results in the formation and maintenance of specific intercellular junctions (Gumbiner et al., 1988). We are studying the expression of cadherins in the testis to determine their role in organogenesis and in the establishment of the various cell-cell interactions and junctions that are responsible for normal testicular function. A picture of cadherin expression in the testis began to emerge when cDNA probes and specific antibodies were used to determine the presence of different cadherins during testicular development. Transcripts for N-cadherin have been identified in RNA isolated from rat (Cyr et al., 1992; Wu et al., 1993; Byers et al., 1994) and mouse testis (MacCalman et al., 1993). An antiserum against the putative cell adhesion recognition sequence of N-cadherin was shown to be reactive toward the surfaces of isolated spermatogenic cells and Sertoli cells and was able to reduce germ cell binding to Sertoli cells in vitro by 3 0 4 0 % (Newton et al., 19931, suggesting a role for N-cadherin in germ cell adhesion to Sertoli cells. Transcripts for E-cadherin have also been demonstrated in rat testis (Wu et al., 1993)) and E-cadherin protein has been localized immunohistochemically to germ cells in 8-day-old mouse testis (Wu et al., 1993) and to interstitial cells in fetal and newborn rat testis (Byers et al., 1994). Finally, transcripts for P-cadherin have been identified in rat testis RNA with highest expression occurring during the first 2 weeks of postnatal development and lower levels observed a t weeks 3 and 4 (Cyr et al., 1992; Wu et al., 1993).Although P-cadherin protein has been identified in homogenates of rat testis on immunoblots, P-cadherin expression has not yet been demonstrated in the testis using immunohistochemistry. The purpose of this work was to determine which cells in the developing mouse testis express P-cadherin. For these studies, we examined P-cadherin expression in the mouse rather than in the rat as we did earlier (Wu et al., 1993), in order to take advantage of a wellcharacterized, commercially available monoclonal antibody, PCD-1, that is specific for mouse P-cadherin (Nose and Takeichi, 1986). Conventional epifluorescence microscopy and scanning laser confocal microscopy were used to examine P-cadherin expression in cryostat sections of mouse testis and in wholemount preparations of mouse seminiferous tubules. We show that Sertoli cells and peritubular cells both expressed P-cadherin, but in nonoverlapping patterns. Sertoli cells expressed P-cadherin during the first week of postnatal development, but not thereafter. Peritubular cells did not express P-cadherin during postnatal week 1,but did so during week 2, and later. The appearance of P-cadherin in peritubular cells paralleled the appearance of p-catenin, one of the molecules that links cadherins to the actin cytoskeleton, and occurred near the time that actin microfilaments organized into the pattern characteristic of mature peritubular cells. This work raises interesting questions regarding the function of P-cadherin in Sertoli cells and peritubular cells and about the differential regulation of P-cadherin in these two cell types. MATERIALS AND METHODS Animals ND4 Swiss Webster mice were purchased from Harlan Sprague-Dawley (Indianapolis, IN) and were maintained on a cycle of 12 h of light and 12 h of darkness. Pups were housed with their mothers, who had free access to food and water. Pups were killed by cervical dislocation. Maintenance and experimental use of all animals complied with the NIH Guide for the Care and Use of Laboratory Animals. lmmunoblot Analysis Freshly dissected tissues were homogenized in buffer “0” (O’Farrell, 1975) containing protease inhibitors as described (Wu et al., 1993). Samples were sonicated for 30 sec on ice, clarified by centrifugation at 15,850 xg for 15 min, and boiled for 5 min before electrophoresis on 7.5% polyacrylamide reducing slab gels containing sodium dodecyl sulfate (Laemmli, 1970). Rainbow’” prestained protein markers (Amersham Co., Arlington Heights, IL) and p-galactosidase (molecular mass = 116 kDa) were included in each gel. Proteins were then electrophoretically transferred to BA83 0.2 pm2 pore size nitrocellulose (Schleicher & Schuell, Keene, NH) using a two-step procedure (Wu et al., 1993). Nitrocellulose blots were incubated for 30 min a t room temperature in blocking buffer consisting of Dulbecco’s phosphate-buffered saline (pH 7.2) containing 1.5 mM CaC1, and 1.0 mM MgC1, (DPBS), 5% nonfat dry milk, and 0.1% Tween-20. The rat monoclonal antibody against mouse P-cadherin (PCD-1) (Nose and Takeichi, 1986) was purchased from Zymed Laboratories (South San Francisco, CA). Blots were incubated overnight at 4°C with 5 pglml PCD-1 antibody in blocking buffer. After three washes with DPBS-0.1% Tween-20, blots were incubated for 1.5 h at room temperature in a 1:2500 dilution of alkaline phosphatase-conjugated, goat antirat IgG antiserum (Promega, Madison, WI) in 50 mM TrisHC1-150 mM NaC1, pH 8.2 (TBS) containing 5% nonfat dry milk and 0.1% Tween-20. Blots were then washed three times with TBS containing 0.1% Tween-20 and bound antibodies were detected using a substrate mixture of nitro blue tetrazolium and 5-bromo-4-chloro-3-indoyl phosphate (Sigma Chemical Co., St. Louis, MO). P-Galactosidase in the marker lane was stained with Fount India drawing ink (Pelikan, F.R.A.). A control for nonspecific binding of the secondary antibody was carried out on duplicate blots that were not exposed to primary antibody, but were treated with the secondary antibody and the chromogen. lmmunofluorescenceMicroscopy Tissues were embedded in O.C.T. Compound (Miles, Elkhart, IN) and frozen in liquid nitrogen. Cryostat sections (7-10 pm) were cut, thawmounted on 0.6% gelatin-coated slides, and stored a t -20°C. Sections were rehydrated in DPBS and fixed for 10 min with 3% paraformaldehyde in DPBS. After three washes with DPBS, sections were extracted with 0.5% (v/v) Triton X-100 in DPBS for 5 min at room temperature using gentle agitation. Extraction with Triton X-100 was found empirically to reduce nonspecific immuno- 157 P-CADHERIN IN MOUSE TESTIS reactivity. DPBS containing 5% normal goat serum (Jackson ImmunoResearch Lab., West Grove, PA) was then applied t o sections for 1 h at room temperature. Sections were incubated with 10 pg/ml PCD-1 in DPBS-5% normal goat serum for 1 h at 37°C. After three washes with DPBS, sections were incubated in a 1:30 dilution of rhodamine-conjugated, goat antirat IgG antiserum (HyClone Laboratories, Logan, UT) in DPBS-5% normal goat serum for 30 rnin at room temperature. After three washes with DPBS, some sections were incubated with 5 pg/ml Hoechst 33258 for 5 min to visualize nuclei (Haneji and Koide, 1988). All sections were mounted using nine parts glycerol and one part DPBS before examining with conventional epif luorescence microscopy. Control sections in which the primary antibody was omitted were examined in each experiment. Wholemount Preparation of Seminiferous Tubules for lmmunohistochemistry The tunica albuginea of the testis was removed and seminiferous tubules were gently rinsed in a plastic culture dish containing 2 ml of ice cold DPBS. Seminiferous tubules were fixed in two steps, first by gently adding an equal volume of ice cold DPBS containing 3%paraformaldehyde into the dish (final concentration of paraformaldehyde = 1.5%)and incubating on ice for 30 min, followed by replacement of this solution with 3%paraformaldehyde and incubation for an additional 30 min. Continuous, gentle agitation was used in all steps described below. After washing three times with DPBS, each testis was cut into smaller pieces (-1-2 mm3) and extracted with 0.05% (dv) Triton-X 100 in DPBS for 5 min at room temperature. Tubule fragments were rinsed and 2-3 tubules, each 1-2 mm in length, were transferred to one well of a 96-well tissue culture plate. Tubule fragments were incubated in DPBS-5% normal goat serum for 1 h at room temperature, followed by incubation with primary antibodies in DPBS a t 37°C for 1 h. The PCD-1 antibody was used a t a concentration of 10 pg/ml. Mouse monoclonal antibody against chicken p-catenin (Johnson et al., 1993) was applied as an undiluted supernatant of a hybridoma culture. After incubation, tubule fragments were rinsed and incubated with rhodamine-conjugated, goat antirat IgG (1:lOO) or goat antimouse IgG (1:lOOO)antiserum, as appropriate, at room temperature for 30 min. Tubule fragments were washed three times with DPBS prior to mounting in glycerol-DPBS medium. The preparations were examined using conventional epifluorescence microscopy as well as the MRC-600 Confocal Laser Scanning Imaging System (Bio-Rad Laboratories, Life Science Group, Melville, NY). Confocal images were captured and photographed using a digital film recorder (GCC Technologies, Bedford, MA). Control wholemount tubules were prepared without incubation with primary antibodies and were examined in each experiment, using the same gain and black level settings as used for samples incubated with primary antibodies. Actin filaments were visualized after incubating fixed and detergent-extracted tubules for 5 min with 0.1 pg/ml fluoresceinlabeled phalloidin (#P-5282, Sigma Chemical Co.). RESULTS lmmunodetection of P-cadherin in Mouse Testis and Epididymis The specific reactivity of PCD-1 toward P-cadherin in mouse skin has been demonstrated (Nose and Takeichi, 1986).To establish the reactivity of PCD-1 toward mouse testis, we examined homogenates of mouse testis on immunoblots. We included homogenates of mouse epididymis in our analysis because high expression of P-cadherin RNA transcripts in the rat epididymis has been reported (Cyr and Robaire, 1991). PCD-1 identified an immunoreactive band with a molecular mass of 118 kDa in homogenates of skin, testis, and epididymis prepared from 7-day-old mice (Fig. 1A). PCD-1 did not react with homogenates of mouse liver. The expression of P-cadherin in skin and its absence in liver is consistent with previous immunoblot results using this antibody (Nose and Takeichi, 1986). PCD-1 reactivity in cryostat sections of mouse epididymis was concentrated at contact sites between adjacent epithelial cells (Fig. 1B) and suggested that epithelial cells are responsible for the expression of P-cadherin RNA transcripts that have been detected in RNA isolated from the intact organ (Cyr and Robaire, 1991). The P-cadherin reactivity in mouse skin was associated with cells in the outer root sheath and hair matrix of follicles (Fig. 1C) and was consistent with previous localization of P-cadherin in human skin (Hirai et al., 1989). No specific reactivity of antibody PCD-1 was found in cryostat sections of mouse liver (Fig. 1D). Next, we explored the spatial distribution of P-cadherin immunoreactivity in mouse testis during the first 15 days of postnatal development in cryostat sections. Germ cells, Sertoli cells, and peritubular cells were identified after counterstaining nuclei with Hoechst 33258, a fluorescent dye for DNA (Haneji and Koide, 1988). Germ cell nuclei were large and contained evenly fluorescent chromatin, whereas Sertoli cell nuclei were small and contained clusters of brightly fluorescent heterochromatin. Peritubular cell nuclei were flattened and separated from the seminiferous epithelium by a space occupied by the basement membrane between Sertoli cells andperitubular cells. On postnatal day 1, P-cadherin immunoreactivity appeared in both the basal and central regions of the developing seminiferous tubules (Fig. 2A). P-cadherin in the basal region of day 1 tubules was assigned to Sertoli cells because a direct correlation could be made between P-cadherin reactivity and Hoechst-stained Sertoli cell nuclei (Fig. 2A,B). P-cadherin in the central region of day 1 tubules was also attributed to Sertoli cells, specifically to Sertoli cell processes that extend between gonocytes to reach the center of the cords. It is unlikely that P-cadherin reactivity in the center of the cords is associated with the gonocyte surface because the surface contour of gonocytes is smooth, whereas the P-cadherin reactivity in the center of the tubules is irregular. We never observed P-cadherin reactivity associated with the smooth contours of germ cells, despite the numerous opportunities to view such profiles in sectioned material. The absence of P-cadherin reactivity within the cytoplasm of gonocytes resulted in large areas within the cords that exhibited no fluorescent signal. On postnatal day 3 P-cadherin reactivity was 158 LIN AND DEPHILIP Fig. 1 , Immunoblot detection and immunofluorescence localization of P-cadherin in mouse tissues. Immunoblot in A contains protein (30 pg) isolated from testis (t), epididymis (el, skin ( s ) ,liver (1) of a postnatal day 7 mouse. Arrow on the right indicates P-cadherin immunoreactivity at 118 kDa. Positions of molecular mass markers are indicated on the left in kDa. Bands around 50 kDa seen in all lanes at the asterisk represent nonspecific reactivity of the secondary antibody. Immunofluorescence of P-cadherin was localized on cryostat sections of day 5 epididymis (B) and day 7 skin (C),but not day 7 liver (D). Arrowheads in B indicate three of many examples of P-cadherin reactivity at areas of lateral contact between epididymal cells. In C, a cross section through the basal region of five hair follicles is shown and P-cadherin reactivity associated with cells in the outer root sheath (0s)and the hair matrix (hm) is indicated. B, x 250; C, x 400; D, x 325. concentrated in the basal region of the cords and was associated again with Sertoli cells (Fig. 2C). Central regions of the tubules on day 3 often contained clusters of gonocytes that excluded P-cadherin reactivity, demonstrating clearly that P-cadherin was not present a t sites of contact between gonocytes. By postnatal day 8, the majority of gonocytes had migrated from the central regions of the tubules to the basement membrane. P-cadherin reactivity re-appeared in the central region of the tubules from which gonocytes had departed and again occupied both the basal and central regions, suggesting that Sertoli cell processes had regained access to the center of the tubule (compare A and E, Fig. 2). An additional feature of P-cadherin reactivity on day 8 was a thin, discontinuous line of fluorescent signal between Sertoli cell nuclei and peritubular myoid cell nuclei (Fig. 2E) that was not seen on day 1, nor on day 3. During subsequent development of the seminiferous tubules, this linear reactivity of P-cadherin a t the margins of the tubules intensified, whereas the intratubular reactivity diminished, so that by day 12, P-cadherin in the testis was seen exclusively at the tubular margins and was absent within the tubules (Fig. 3A). The association of P-cadherin with peritubular cells was suggested when tangential sections of day 15 tubules were examined (Fig. 3B), and P-cadherin reactivity could be seen outlining a network of large cells a t the tubule margin. The identification of these cells as peritubular cells was confirmed in wholemount preparations of tubules, processed for immunohistochemistry and examined with conventional fluorescence microscopy (Fig. 3C). The large size of these cells and their location on the surface of seminiferous tubules identified them as peritubular cells and distinguished them from Sertoli cells that are smaller and are located below the tubule surface. Furthermore, sinusoidal endothelial cells were removed during isolation of seminiferous tubules for wholemount examination and did not contribute to P-cadherin reactivity on the tubule surface. P-cadherin continued to be associated with peritubular cells on postnatal day 60, the latest age of development studied here (not shown). No P-cadherin P-CADHERIN IN MOUSE TESTIS 159 Fig. 2. Immunofluorescence localization of P-cadherin during early postnatal development of the mouse testis. P-cadherin reactivity in cryostat sections from 1-,3-, and 8-day-old mouse testes is presented in A, C, and E,respectively. DNA was visualized in the same sections by counterstaining with Hoechst dye 33258 and the corresponding micrographs are shown in B, D, and F. Arrows in A and B indicate the same five Sertoli cells and demonstrate that P-cadherin is associated with Sertoli cells. Arrows in E indicate the discontinuous line of P-cadherin reactivity between Sertoli cell nuclei and pertibular cell nuclei and should be compared to the corresponding area indicated by arrows in F. In E and F, note the presence of P-cadherin reactivity and the absence of cell nuclei in the center of the tubules. Sc, P-cadherin reactivity associated with Sertoli cells in the basal region of a seminiferous tubule sectioned tangentially. g, germ cell nucleus. A, B, E, F, x 500; C and D, x 450. immunoreactivity was ever detected in the interstitial region of the testis at anv " aae- studied. postnatal development, wholemount preparations of day 7 and dav 15 seminiferous tubules were compared using confocd microscopy. A focal plane at the sirface of day tubules revealed that only a few of the large, flattened peritubular cells were outlined by P-cadherin reactivity-(Fig. 4A). The P-cadherin reactivity associated with Sertoli cells within day 7 tubules was dem- u Demonstration of P-Cadherin in Seminiferous Tubules using Confocal Microscopy To visualize P-cadherin both on the surface and through the thickness of seminiferous tubules during 160 LIN AND DEPHILIP 8 (Fig. 2E). By day 15, all peritubular cells were outlined by P-cadherin, creating an impressive network of reactivity a t contact sites between adjacent cells (Fig. 4C). P-cadherin reactivity associated with peritubular cells also could be demonstrated when the focal plane passed through the diameter of day 15 tubules (Fig. 4D). Here, a single layer of peritubular cells was immunoreactive at the lateral margins of the tubule, whereas the multiple layers of cells within the tubules were not reactive. Thus P-cadherin reactivity on peritubular cells and the lack of reactivity on Sertoli cells and germ cells within day 15 tubules were demonstrated both in cryostat sections (Fig. 3B) and wholemount preparations (Fig. 4D). Actin Filament Organization and p-Catenin Expression in Peritubular Cells during Postnatal Development Fig. 3. Immunofluorescence localization of P-cadherin during later postnatal development of mouse testis. P-cadherin reactivity in cryostat sections from 12- and 15-day-old mouse testis is presented in A and B, respectively. Arrows in A and B indicate the discontinuous line of P-cadherin reactivity surrounding seminiferous tubules (t) cut in cross section. In B, P-cadherin reactivity is observed outlining peritubular cells (p) in tubules cut tangentially. There is no P-cadherin reactivity within the seminiferous tubules at these ages. In C, the continuous network of P-cadherin reactivity outlining peritubular cells in tubules from 15-day-old mouse testis is demonstrated in wholemount preparations examined with conventional microscopy. The plane of focus is at the surface of the tubules. A, x 325; B and C , x 500. onstrated again when the focal plane passed through the tubule diameter (Fig. 4B) and should be compared to the P-cadherin reactivity on cryostat sections on day The adhesive function of cadherins is dependent on an association with actin filaments that is mediated by catenins (Kemler, 1993). Having demonstrated that P-cadherin is expressed at contact sites between all peritubular cells on day 15, it was appropriate to examine the organization of actin filaments in peritubular cells and, in addition, to ask whether any catenin molecule could be localized at the peritubular cell surface. Actin filament bundles in mature peritubular cells are arranged in a characteristic orthogonal pattern and have been studied in adults of several species using fluorescent phallotoxins (Vogl and Soucy, 1985; Vogl et al., 1985; Maekawa et al., 1994). The postnatal development of peritubular cell actin bundles has been studied in rats, where the adult pattern appears on, or slightly before postnatal day 22 in the Sprague-Dawley strain (Russell et al., 1989), or on day 30 in the Wistar strain (Maekawa et al., 1995). To correlate the development of the actin cytoskeleton in peritubular cells with P-cadherin expression, we incubated wholemount preparations of seminiferous tubules with fluorescent phalloidin and examined the preparations using confocal microscopy. Actin bundles perpendicular to the long axis of seminiferous tubules appeared first and were detected in the majority of peritubular cells on postnatal day 7 (Fig. 5A). Perpendicular actin bundles were still the predominant type in peritubular cells on day 12 (Fig. 5B). On day 15, actin bundles parallel to the long axis of tubules were found in the majority of peritubular cells and joined the perpendicular bundles to create the mature orthogonal network (Fig. 5C). At all ages, strong fluorescent signal corresponding to cortical actin defined the perimeter of peritubular cells. p-Catenin immunoreactivity outlined several peritubular cells on day 12 (not shown) and the majority of peritubular cells on day 15 (Fig. 5D), the same day that P-cadherin appeared uniformly at peritubular cell contact sites (Fig. 4C) and that the mature actin network was established (Fig. 5C). DISCUSSION P-cadherin has been localized to two cell types in the developing mouse testis-Sertoli cells within the seminiferous tubules and peritubular cells in the tubule boundary tissue. However, the temporal expression of P-cadherin in these two cells was quite different. P-cadherin expression in Sertoli cells was observed on post- P-CADHERIN IN MOUSE TESTIS 161 Fig. 4. Immunofluorescence localization of P-cadherin in wholemount preparations of seminiferous tubules using confocal microscopy. P-cadherin reactivity is demonstrated in tubules from 7- (A and B) and 15- (Cand D) day-old mouse testis. The plane of focus is either at (A and C) or below (B and D) the surface of the tubules. In A, an example of P-cadherin reactivity outlining a large peritubular cell is indicated (arrow). In B, P-cadherin reactivity associated with cells within the seminiferous tubules is shown. In C, P-cadherin reactivity outlines all peritubular cells on the surface of day 15 tubules. In D, the plane of focus is through the diameter of a day 15 tubule and demonstrates P-cadherin reactivity associated with peritubular cells (example at arrow) at the lateral margins of the tubule. There is no P-cadherin reactivity associated with cells within seminiferous tubules from the 15-day-oldmouse. The width of the tubule in C and D is exaggerated because the specimen was compressed under the coverslip to better demonstrate the P-cadherin network. A-D, x 500. natal days 1 , 3 , and 8, but not on day 12 or thereafter. P-cadherin expression in peritubular cells was first observed on day 8, became progressively stronger, and was seen as late as postnatal day 60. The detection of P-cadherin protein in the testis seen here immunohistochemically should be compared with previous identification of P-cadherin RNA transcripts (Cyr et al., 1992; Wu et al., 1993). The level of P-cadherin transcripts is highest during the first 2 weeks of postnatal development and lower at weeks 3 and 4. It is likely that P-cadherin mRNA identified during week 1is produced by Sertoli cells, whereas transcripts expressed a t lower levels during weeks 3 and 4 are synthesized by peritubular cells. Functions for P-cadherin in the early postnatal development of Sertoli cells can be suggested by considering three features of the seminiferous tubules during this period. First, Sertoli cells and germ cells undergo an important change in relative position as gonocytes move from the center of the cords to their permanent position adjacent to the basement membrane (McGuinness and Orth, 1992). This migration of gonocytes is thought to be crucial for providing the proper number of germ cells to sustain spermatogenesis. P-cadherin may help re-establish contact between Sertoli cells as gonocytes migrate between them and thus maintain tubule structure during this period of cell migration. Second, Sertoli cells in the newborn mouse demonstrate gap junctions and rudimentary occluding junctions (Nagano and Suzuki, 1976). The gap junctions decrease in number and size between birth and the end of week 1,whereas the occludingjunctions continue to form and eventually divide the seminiferous tubule into apical and basal compartments. P-cadherin expressed during postnatal week 1 may facilitate gap junction formation between Sertoli cells, a role proposed for N-cadherin in mouse sarcoma cells (Matsuzaki et al., 1990) and for E-cadherin in mouse epidermal cells (Jongen et al., 1991). Since P-cadherin expression within seminiferous tubules declines to an undetectable level between day 8 and day 12, before the Sertoli cell occluding junctions mature around day 16 (Nagano and Suzuki, 1976), any involvement of P-cadherin in occluding junction formation must be early, rather than late in this process. Finally, the loss of P-cadherin expression by Sertoli cells between day 8 162 LIN AND DEPHILIP Fig. 5. Distribution of actin filament bundles and p-catenin immunoreactivity in peritubular cells. Wholemount preparations of seminiferous tubules were incubated with fluorescein-conjugated phalloidin and examined using confocal microscopy. Actin filament distribution was observed in peritubular cells from 7- (A), 12- (B),and 15- (C) day-old tubules. Arrows in A and B indicate actin bundles that are perpendicular to the long axis of the seminiferous tubule. In C, actin filaments parallel to the long axis of tubules joined the perpendicular fibers to create the mature lattice network in the majority of peritubular cells. Black arrowheads in A, B, and C indicate cortical actin at the perimeter of peritubular cells. In D, p-catenin immunoreactivity outlined the majority of peritubular cells on the surface of 15-day-old tubules. A-D, x 500. and day 12 coincides with their decreased potential for proliferation. The labelin index of mouse Sertoli cells after incorporation of [BHI-thymidine declines from 11%on postnatal day 6 t o 0.1%on day 12 (Kluin et al., 1984).Since P-cadherin expression has been associated with cells having a high rate of proliferation (Hirai et al., 1989), the loss of P-cadherin by Sertoli cells may signal a more mature, postmitotic stage of differentiation. Whatever its role in Sertoli cells, P-cadherin may function independently of known catenins, since we were unable to detect p-catenin within mouse seminiferous tubules (not shown), and only weak staining with antibodies against a-and p-catenin was reported in rat seminiferous tubules during fetal development and on postnatal day 10 (Byers et al., 1994). The localization of P-cadherin and 6-catenin at areas of intercellular contact suggests an adhesive function for P-cadherin in peritubular cells. Furthermore, the appearance of a mature lattice network of actin bundles at about the same time that P-cadherin and p-catenin completely outline the peritubular cells is consistent with the idea that a P-cadherin-p-catenin-actin fil- ament complex is being formed to achieve intercellular adhesion and junction formation. The identification of discrete junctions between peritubular cells using electron microscopy is challenging because of the limited profiles that are available in sectioned material. Nevertheless, cytoplasmic dense bodies, responsible for cellular adhesion and similar to those seen in smooth muscle cells, have been described between mouse peritubular cells (Ross, 1967). It is reasonable to predict that strong junctions between peritubular cells are necessary to maintain the integrity of the peritubular cell sheath as it undergoes the contractions responsible for the movement of testicular fluid and spermatozoa. Our demonstration of 6-catenin in peritubular cells of the mouse is consistent with previous work reporting p- and a-,but not y-catenin in developing and mature peritubular cells of the rat (Byers et al., 1994). The question of whether germ cells also express P-cadherin remains unanswered. In sections containing germ cell aggregates that excluded Sertoli cells, it was clear that P-cadherin was not localized at contact sites between adjacent germ cells. Therefore, it is un- P-CADHERIN IN MOUSE TESTIS likely that the observed variations in immunostaining during the first 15 postnatal days can be attributed to changes in the germ cell population that occur during this period (BellvB et al., 1977). However, we cannot rule out the possibility that P-cadherin is expressed at contact sites between germ cells and Sertoli cells. We have shown that germ cells in the 8-day-old mouse testis express E-cadherin (Wu et al., 1993) and have observed E-cadherin on germ cells throughout the first week of postnatal development (not shown). Because adhesion between cells expressing different types of cadherin has been observed (Volk et al., 1987), an interaction between E-cadherin-expressing germ cells and P-cadherin-expressing Sertoli cells during the first week of postnatal development must be considered. In addition, cadherins have been shown t o bind to integrins (Cepek et al., 1994), although this binding is not a general feature of cadherin activity. Integrin subunits have been identified in the developing testis (Palombi et al., 1992; Frojdman and Pelliniemi, 19941, and it is possible that P- or E-cadherin interact with integrins during gonadogenesis. Finally, aside from discussion of P-cadherin function in Sertoli and peritubular cells, these results raise interesting questions regarding the differential regulation of P-cadherin in two cell types that are separated from each other only by a basement membrane. Although their microenvironments may differ, Sertoli cells and peritubular cells are subject to the same endocrine environment, yet Sertoli cells express P-cadherin when peritubular cells do not and vice versa. The loss of P-cadherin expression by Sertoli cells coincides with a decrease in the amount of bioactive Mullerian inhibiting substance to an undetectable level (Taketo et al., 1993). Mullerian inhibiting substance is produced by immature Sertoli cells and is responsible for regression of the female reproductive ducts in male embryos, but has an inductive effect on Sertoli cells. It will be important to determine whether Sertoli cells autoregulate their expression of P-cadherin via Mullerian inhibiting substance. When considering the regulation of P-cadherin in peritubular cells, it may be significant that the expression of P-cadherin parallels the maturation of peritubular cells, a process that is androgendependent (Hovatta, 1972).The androgen regulation of P-cadherin in peritubular cells can be tested in vivo and in vitro. The developing testis may prove to be a valuable model system to study the differential regulation of the P-cadherin gene. A unique feature of this system is that Sertoli cells and peritubular cells can act as positive and negative controls of P-cadherin expression for each other at specific stages of testis development. ACKNOWLEDGMENTS We thank Margaret J. Wheelock and Keith R. Johnson (University of Toledo) for their gift of antibody against p-catenin. We thank Kenneth H. Jones (Ohio State University) for his review of the manuscript before submission. This work was supported by NIH Grant HD-19735. LITERATURE CITED Bellve, A.R., J.C. Cavicchia, C.F. Millette, D.A. OBrien, Y.M. Bhatnagar, and M. Dym, 1977 Spermatogenic cells of the prepuberal 163 mouse: Isolation and morphological characterization. J . Cell Biol., 74:68-85. Byers, S.W., S. Sujarit, B. Jegou, S. Butz, H. Hoschutzky, K. Herrenknecht, C. MacCalman, and O.W. Blaschuk 1994 Cadherins and cadherin-associated molecules in the developing and maturing rat testis. Endocrinology, 134:630-639. Cepek, K.L., S.K. Shaw, C.M. Parker, G.J. Russell, J.S. Morrow, D.L. Rimm, and M.B. Brenner 1994 Adhesion between epithelial cells and T lymphocytes mediated by E-cadherin and the aEP7integrin. Nature, 372:190-193. Cyr, D.G., O.W. Blaschuk, and B. Robaire 1992 Identification and developmental regulation of cadherin messenger ribonucleic acids in the rat testis. Endocrinology, 131:139-145. Cyr, D.G., and Robaire, B. 1991 Developmental regulation of epithelial- and placental-cadherin mRNAs in the rat epididymis. In: The Male Germ Cell: Spermatogonium to Fertilization. New Academy of Sciences, New York, pp. 399-408. Edelman, G.M. 1988 Morphoregulatory molecules. Biochemistry, 27: 3533-3543. Frojdman, K., and L.J. Pelliniemi 1994 Differential distribution of the as subunit of integrins in the development and sexual differentiation of the mouse testis. Differentiation, 57:21-29. Gumbiner, B., B. Stevenson, and A. Grimaldi 1988 The role of the cell adhesion molecule uvomorulin in the formation and maintenance of the epithelial junctional complex. J . Cell Biol., 107:1575-1587. Haneji, T., and S.S. Koide 1988 Identification of cell types in Sertoli cell-enriched cultures by immunocytochemistry and DNA-specific fluorochrome Hoechst 33342. Histochemistry, 8 9 5 - 6 1 . Hatta, K., and M. Takeichi 1986 Expression of N-cadherin adhesion molecules associated with early morphogenetic events in chick development. Nature, 320:447-449. Hirai, Y., A. Nose, S. Kobayashi, and M. Takeichi 1989 Expression and role of E- and P-cadherin adhesion molecules in embryonic histogenesis 11. Skin morphogenesis. Development, 105:271-277. Hovatta, 0. 1972 Effect of androgens and antiandrogens on the development of the myoid cells of the rat seminiferous tubules (organ culture). Z. Zellforsch., 131:299-308. Johnson, K.R., J.E. Lewis, D. Li, J . Wahl, A.P. Soler, K.A. Knudsen, and M.J. Wheelock 1993 P- and E-cadherin are in separate complexes in cells expressing both cadherins. Exp. Cell Res., 207: 252-260. Jongen, W.M.F., D.J. Fitzgerald, M. Asamoto, C. Piccoli, T.J. Slaga, D. Gros, M. Takeichi, and H. Yamasaki 1991 Regulation of connexin 43-mediated gap junctional intercellular communication by Ca2+ in mouse epidermal cells is controlled by E-cadherin. J. Cell Biol., 114:545-555. Kemler, R. 1992 Classical cadherins. Semin. Cell Biol., 3:149-155. Kemler, R. 1993 From cadherins to catenins: Cytoplasmic protein interactions and regulation of cell adhesion. Trend Genet, 9:317321. Kluin, P.M., M.F. Kramer, and D.G.D. Rooij 1984 Proliferation of spermatogonia and Sertoli cells in maturing mice. Anat. Embryol., 169:73-78. Laemmli, U.K. 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4.Nature, 227:680-685. MacCalman, C.D., D.A. OBrien, S. Byers, and O.W. Blaschuk 1993 N-cadherin expression in the seminiferous epithelium of the mouse testis. Endocr. J., 1:519-525. Maekawa, M., H. Kazama, K. Kamimura, and T. Nagano 1995 Changes in the arrangement of actin filaments in myoid cells and Sertoli cells of rat testes during postnatal development and after experimental cryptorchidism. Anat. Rec., 241 59-69. Maekawa, M., T. Nagano, and T. Murakami 1994 Comparison of actin-filament bundles in myoid cells and Sertoli cells of the rat, golden hamster and mouse. Cell Tissue Res., 275:395-398. Matsuzaki, F., R.-M. MBge, S.H. Jaffe, D.R. Freidlander, W.J. Gallin, J.I. Goldberg, B.A. Cunningham, and G.M. Edelman 1990 cDNAs of cell adhesion molecules of different specificity induce changes in cell shape and border formation in cultured S180 cells. J . Cell Biol., 110:1239-1252. McGuinness, M.P., and J.M. Orth 1992 Reinitiation of gonocyte mitosis and movement of gonocytes to the basement membrane in testes of newborn rats in vivo and in vitro. Anat. Rec., 233527537. Nagano, T., and F. Suzuki 1976 The postnatal development of the junctional complexes of the mouse Sertoli cells as revealed by freeze-fracture. Anat. Rec., 185~403-418. Newton, S.C., O.W. Blaschuk, and C.F. Millette 1993 N-cadherin mediates Sertoli cell-spermatogenic cell adhesion. Dev. Dyn., 197: 1-13. Nose, A. and M. Takeichi 1986 A novel cadherin cell adhesion mole- 164 LIN AND DEPHILIP cule: its expression patterns associated with implantation and organogenesis of mouse embryos. J . Cell Biol., 103r2649-2658. O'Farrell, P.H. 1975 High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem., 250:4007-4021. Ozawa, M., H. Baribault, and R. Kemler 1989 The cytoplasmic domain of the cell adhesion molecule uvomorulin associates with three independent proteins structurally related in different species. EMBO J., 8r1711-1717. Ozawa, M., M. Ringwald, and R. Kemler 1990 Uvomorulin-catenin complex formation is regulated by a specific domain in the cytoplasmic region of the cell adhesion molecule. Proc. Natl. Acad. Sci. USA, 87:4246-4250. Palombi, F., M. Salanova, G. Tarone, D. Farina, and M. Stefanini 1992 Distribution of p l integrin subunit in rat seminiferous epithelium. Biol. Reprod., 47:1173-1182. Ross, M.H. 1967 The fine structure and development of the peritubular contractile cell component in the seminiferous tubules of the mouse. Am. J. Anat., 121:523-558. Russell, L.D., A. Bartke, and J.C. Goh 1989 Postnatal development of the Sertoli cell barrier, tubular lumen, and cytoskeleton of Sertoli and myoid cells in the rat, and their relationship to tubular fluid secretion and flow. Am. J. Anat., 184:179-189. Takeichi, M. 1988 The cadherins: Cell-cell adhesion molecules controlling animal morphogenesis. Development, 102:639-655. Taketo, T., J. Safeed, T. Manganaro, M. Takahashi, and P.K. Donahoe 1993 Miillerian inhibiting substance production associated with loss of oocytes and testicular differentiation in the transplanted mouse XX gonadal primordium. Biol. Reprod., 49:13-23. Vestweber, D., R. Kemler, and P. Ekblom 1985 Cell-adhesion molecule uvomorulin during kidney development. Dev. Biol., 112; 213-221. Vogl, A.W., and L.J. Soucy 1985 Arrangement and possible function of actin filament bundles in ectoplasmic specializations of ground squirrel. J . Cell Biol., 100:814-825. Vogl, A.W., L.J. Soucy, and G.J. Lew 1985 Distribution of actin in isolated seminiferous epithelia and denuded walls of the rat. Anat. Rec., 213:63-71. Volk, T., 0. Cohen, and B. Geiger 1987 Formation of heterotypic adherens-type junctions between L-CAM-containing liver cells and A-CAM-containing lens cells. Cell, 50:987-994. Wu, J.-C., C.W. Gregory, and R.M. DePhilip 1993 Expression of E-cadherin in immature rat and mouse testis and in rat Sertoli cell cultures. Biol. Reprod., 49:1353-1361.