MICROSCOPY RESEARCH AND TECHNIQUE 20:219-231 (1992)
Characterization of Sertoli Cell Perinuclear Filaments CARLOS A. SUAREZ-QUIAN AND MARTIN DYM Georgetown University Medical Center, Department of Anatomy and Cell Biology, Washington, D.C. 20007
KEY WORDS
Sertoli cell, Nucleus, Cytoskeleton
ABSTRACT Sertoli cell nuclei are characterized by deep invaginations and, in addition, the orientation of the nuclei with respect to the wall of the seminiferous tubules varies during the cycle of the seminiferous epithelium. These events may be the result of cytoplasmic filaments acting a t the level of the nuclear capsule and may represent significant changes in Sertoli cell activity. Thus, a study was performed to characterize the nature of the perinuclear filaments of Sertoli cells in vivo and in vitro. In Sertoli cells in vivo, microtubules and microfilaments were often detected in the perinuclear cytoplasm, and these cytoskeletal components were observed to course either parallel to, or abut at, the nuclear capsule. In Sertoli cells in vitro, the nuclear infoldings are retained and the perinuclear cytoskeleton was shown to contain microtubules, f-actin, and intermediate filaments. A fixationpermeabilization protocol employing tannic acid-saponin was used and it significantly enhanced the preservation of cytoskeletal components. The presence of f-actin was demonstrated by using the S1 fragment of muscle myosin to decorate the microfilaments. Treatment of the cultured cells with either microtubule or f-actin depolymerizing agents had no effect on nuclear shape. Thus, at present, the function of the prominent perinuclear cytoskeletal components remains unknown.
INTRODUCTION
serum testosterone levels by passive immunization with LH antisera also decreases Sertoli cell nuclear The orientation of Sertoli cell nuclei in situ varies infoldings (Dym and Raj, 1977). These data suggest during the cycle of the seminiferous epithelium (Re- that changes in nuclear shape and orientation are not gaud, 1901; Elftman, 1950); to date, this change is per- haphazard, but rather a required function of normal haps the most prominent morphological feature of Ser- Sertoli cell activity. toli cells known to vary during the cycle (Parvinen and Ultrastructural observations of Sertoli cells in vivo Ruokonen, 1982). A quantitative analysis of this phe- have revealed a perinuclear zone of filaments (Fawcett, nomena revealed that the nuclei generally adopted one 1975; Dym, 1977; Dym and Cavicchia, 1978). Thereof two configurations (Leblond and Clermont, 1952). fore, the composition and possible involvement of these During the early part of the cycle, nuclei appeared to cytoplasmic filaments in generating and maintaining exhibit a greater length in their parallel axis, with nuclear shape was examined. In addition, the relerespect to the basement membrane of seminiferous tu- vance of Sertoli cell nuclear infoldings in spermatogenbules, than in their perpendicular axis. As the cycle of esis is discussed. seminiferous epithelium progressed, more nuclei gained a perpendicular orientation, and a t stage VIII, MATERIALS AND METHODS where sperm release occurs, the number of perpendicTissue Preparation ularly oriented nuclei became maximal. Pressure on Six adult Sprague-Dawley rats were anesthetized the lateral borders of Sertoli cells, generated by the with Nembutal and the testes fixed by perfusion via rapid increase of developing germ cells undergoing mitosis during the early stages of the cycle, was believed the abdominal aorta with 5% glutaraldehyde in 0.2 M to passively produce the variation in the nuclear shape s-Collidine buffer. After 15 minutes, the testes were removed, cut into 1mm cubes, and fixed for an addi(Leblond and Clermont, 1952). tional two hours in the same solution. The testes pieces Numerous and deep invaginations, often extending throughout its length, are another easily recognized were rinsed three times in buffer and post-fixed for one feature of Sertoli cell nuclei in sections of testes. Such hour in potassium ferrocyanide-reduced OsO, a t room nuclear infoldings develop during maturation of the temperature (Karnovsky, 1971). Next, the testes were Sertoli cell (Ramos and Dym, 1979; Hatier and Grig- rinsed in buffer and dehydrated through a graded senon, 1980), and appear to depend upon normal levels of ries of alcohols, followed by propylene oxide. The testes circulating plasma gonadotropins (Cameron and Markwald, 1975; Chemes et al., 1979). Consistent with these observations is the fact that in diseased testes, where serum testosterone levels decrease, or testosterone inReceived January 17,1990;accepted in revised form May 30, 1990. sensitivity is recognized as the biochemical lesion, Serreprint requests to Dr. Carlos A. Suarez-Quian,Georgetown Univertoli cell nuclear infoldings are diminished (Chemes et sityAddress Medical Center, Department of Anatomy and Cell Biology, 3900 Reservoir al., 1977; Nistal et al., 1982). In addition, depletion of Road, N.W., Washington, D.C.20007.
0 1992 WILEY-LISS, INC.
220
C.A. SUAREZ-QUIAN AND M. DYM
were infiltrated with an 1:lmixture of propylene oxide and Epon, and embedded in Epon. For light microscopic examination, one micron sections were cut, mounted on slides, and stained with toluidine blue. Sections were examined with a Zeiss microscope using a 6 3 Planapo, ~ 1.4 N.A. objective. Photographs were taken with Kodak 4162 film and processed per manufacturer’s instructions. For electron microscopy, 60-90 nm sections were prepared using an LKB ultramicrotome. Sections were counterstained with uranyl acetate and lead citrate and examined with a Jeol 1200 EX electron microscope.
Preparation of Sertoli Cell Cultures Sertoli cell cultures were prepared as described previously (Suarez-Quian et al., 1983) as modified (Hadley et al., 1985), except that cells were plated onto Aclar coverslips (Allied Chemicals) coated with rat tail collagen (Collaborative Research, Walthan, MA). Briefly, testes of 20-day-old rats were excised, the tunica albuginea removed, and the seminiferous tubules subjected to three sequential enzymatic digestions consisting of collagenase and DNAse (Worthington). After each digestion the cells were allowed to pellet at unit gravity and washed in Dulbecco’s minimal essential medium (DMEM). The Sertoli cell enriched cell pellet resulting after the last digestion was diluted in a serum free defined medium (SFDM),consisting of DMEM plus 100 ng/ml FSH (NIADDK-oFSH-16),2 p,g/ml insulin, 5 p,g/ ml human transferrin, 50 ng/ml vitamin A, 200 ng/ml vitamin E, lo-’ M hydrocortisone, lo-’ M testosterone, lo-’ M estradiol, 2 mM glutamine, 5 ng sodium selenate, 1mM sodium pyruvate, and 22 mM sodium lactate. Preparation of In Vitro Cells for Light and Electron Microscopy For conventional light and electron microscopic observation, cells were fixed and processed exactly as described above for tissue, but with the following modifications: After dehydration in 100% alcohol the cells were immersed in a 5050 mixture of ethanol-Epon plus 2% DMP 30 for two hours, followed by fresh Epon plus 2% DMP 30 and the plastic was allowed to hardened in a 60°C oven. After embedding, desired regions of the Epon containing cells were cut and glued to blocks for sectioning. Embedded cells were mounted for sectioning in either of two orientations: knife sectioned cells in a plane parallel to their plane of growth, or knife sectioned cells perpendicular to their plane of growth. Tannic Acid Fixation for Sertoli Cells To enhance the preservation of the cytoskeleton yet maintain membranes of Sertoli cells, an empirical fixation protocol was developed. This recipe is similar to that of Maupin and Pollard (1983), except that concentrations of saponin, tannic acid, and glutaraldehyde were modified. Sertoli cells in vitro were fixed for 30 minutes in the following: 2 mM MgC1, 50 mM KC1, 0.01% saponin, 0.1% tannic acid, 3.0%glutaraldehyde, and 0.1 M sodium phosphate buffer (pH 7.0). All steps
were performed as described above for tissue, except cells were post-fixed with OsO, in 0.1 M sodium phosphate buffer (pH 6.0) for only 15 minutes on ice without en bloc staining in uranyl acetate. Instead, thin sections were stained for four minutes using lead citrate only. Labeling of Sertoli Cells In Vitro With S1 Fragments The S1 fragment of the myosin molecule was prepared as described by Begg et al. (1978). Decoration of Sertoli cell f-actin with S1 was performed as follows: Cells were rinsed twice, two minutes each, in stabilization buffer (SB)consisting of 0.1 M PIPES (pH 6.9),2 mM EGTA, and 2 mM MgC1,. Next, the cells were extracted and decorated with S1 fragments for ten minutes on ice in the following solution: SB plus 0.1% Triton X-100, 0.5 mg/ml TAME (Sigma), and 1-2 mg/ml S1. The cells were then washed in SB and fixed in 1% glutaraldehyde, 0.2%tannic acid in 0.1 M sodium phosphate buffer (pH 7.0) for one hour. Next, the cells were post-fixed in 1% OsO, in 0.1 M sodium phosphate buffer (pH 6.0) at 4°C for 15 minutes. Afterwards, the cells were processed for electron microscopic examination as described above.
RESULTS In Vivo Observations A profile of a seminiferous tubule cross section is shown in Figure 1.Staging of the seminiferous tubules is based on the different cell associations and characteristic shapes of acrosome development and spermatids. The perpendicular axis of Sertoli cell nuclei, when compared to the parallel axis, with respect to the basement membrane, is altered during the seminiferous tubule cycle, although all nuclei within the same stage do not exhibit an identical orientation. That is, as shown in Figure 1, a single stage may contain Sertoli cell nuclei exhibiting both orientations. Thus, it is a statistical percentage of the total number of nuclei in one stage versus another that alter their orientation (Leblond and Clermont, 1952). Ultrastructural studies were performed in Sertoli cells in vivo to characterize the composition of the perinuclear filaments (Fig. 2). The Sertoli cell nuclei are readily identified by their vast amount of euchromatin, a prominent nucleolus, and nuclear infoldings. The perinuclear cytoplasm of Sertoli cells, a distance of approximately 0.25 microns, is commonly devoid of cytoplasmic organelles (arrowheads in Fig. 2). Although cytoplasmic organelles may enter nuclear infoldings, they are also excluded from the 0.25 micron perinuclear zone. At higher magnification (Fig. 2B), the perinuclear zone devoid of cytoplasmic organelles may be observed to contain components of the cytoskeleton. These components include microtubules and microfilaments. Generally, however, the preservation of cytoskeletal components in the perinuclear exclusion zone is poor. The reason for this is not known, although both OsO, and glutaraldehyde destruction of microfilaments are possible agents of f-actin depolymerization (Maupin-Szamier and Pollard, 1978; Lehrer, 1981). Another profile of a Sertoli cell nuclei is illustrated
SERTOLI CELL PERINUCLEAR FILAMENTS
22 1
Fig. 1. Light microscopic observations of seminiferous tubules. For orientation purposes the lumen of tubules, Leydig cells (L)residing outside the tubule walls, and a few examples of spermatocytes (SC)and round spermatids (RS)are labeled. The nuclei of Sertoli cells (N)are readily identified in cross-sectional profiles of the seminiferous epithelium. They are characterized by their euchromatin pattern
and one prominent nucleoli. The apical cytoplasm of the Sertoli cells extends to the tubule lumen. It is this area of the Sertoli cells that forms crypts for elongated spermatids. Note that the nuclei of Sertoli cells may adopt one of two configurations, either parallel or perpendicular to the wall of the tubules. Bar = 10 microns; x 1,400.
in Figure 3. During germ cell differentiation a close association exists between elongated spermatids and Sertoli cell nuclei. Elongated spermatids plunge to-
wards Sertoli cell nuclei and often come to rest within nuclear infoldings as shown in Figure 3. It is during the spermatid plunge that Sertoli cell nuclei undergo
Fig. 2. Ultrastructural observations of Sertoli cells in vivo I. Characteristic morphological features of Sertoli cells are readily observed at the ultrastructural level (A). Nuclear infoldings, lack of heterochromatin, and a single reticular nucleoli with its two satellite karyosomes are prominent features of the Sertoli cell nucleus. Cytoplasmic organelles include the Golgi apparatus (G), mitochondria (m), and junctional complexes (JC) between adjacent Sertoli cells. Elongated spermatids (ES) residing within the apical cytoplasm of the Sertoli cell are demonstrated. A perinuclear exclusion zone (arrowheads) of
cytoplasmic organelle is also observed. The square shown in A is illustrated a t higher magnification in B. In B, the nuclear capsule has been section tangentially and nuclear pores are indicated by arrows. Microtubules (arrowheads) are present within 0.25 microns of the nuclear capsule. The diameter of thin filaments in the perinuclear area suggests that they are f-actin and they appear to abut directly against the nuclear capsule. A: Bar = 1micron; x 13,500. B: Bar = 0.25 micron; x 54,000.
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 3. Ultrastructural observations of Sertoli cells in vivo 11. A profile of a Sertoli cell nuclei oriented perpendicular to the base of the seminiferous epithelium is illustrated. Note the head portion of an elongated spermatid (ES)residing within a crypt of the Sertoli cell cytoplasm. The perinuclear cytoplasmic organelle exclusion zone (arrowheads) is quite prominent. Observe also that the exclusion zone is present on the side of the Sertoli cell nucleus adjacent to the elongated spermatid. Bar = 1 micron; x 15,600.
223
224
C.A. SUAREZ-QUIAN AND M. DYM
their change in orientation; the length of the parallel axis diminishes as the perpendicular axes of the nuclei elongate (compare with Fig. 1). At its closest, the spermatid plasma membrane is only 300-400 nm from the outer membrane of the Sertoli cell nuclear envelope (arrowheads in Fig. 3). This distance corresponds to the area inhabited by cytoplasmic perinuclear filaments and is also devoid of cytoplasmic organelles.
In Vitro Observations The characterization of Sertoli cell perinuclear filaments was performed in primary cell cultures. The shape of nuclei of cultured Sertoli cells plated onto rat tail collagen is shown in Figure 4A,B. Sertoli cells in vitro maintain their characteristic highly infolded nuclei if cultured in the presence of fetal calf serum or in serum free media but containing androgens and FSH. Cells fixed with 5% glutaraldehyde buffered in 0.2 M s-Collidine and prepared for electron microscopy reveal a perinuclear zone of filaments (data not shown). Generally, however, the preservation of the perinuclear zone of filaments is not consistent using this fixation protocol. The results of using tannic acid in combination with saponin, a mild detergent, in the primary fixation to enhance visualization of perinuclear filaments are shown in Figures 5 and 6. Using this fixation protocol various artifacts are noted: distended vesicles appear in the cytoplasm, cytoplasmic ground substance is generally absent, and mitochondria exhibit a less dense consistency. Presumably, the detergent saponin partially disrupted the cell membranes and rendered sertoli cells susceptible to osmotic damage. Nevertheless, preservation of cytoskeletal components in the presence of tannic acid is significantly enhanced. For example, microtubules were readily detected within the perinuclear cytoplasmic organelle exclusion zone and were prominent features of the nuclear infoldings (small arrowheads in Fig. 5B). In favorable profiles, microtubules appeared to end at the nuclear capsule, often in the vicinity of nuclear pores (white arrows in Fig. 6A,B). The presence of f-actin as a component of Sertoli cell perinuclear filamentous material was also determined. Sertoli cells were cultured, the f-actin decorated with the fragment of myosin S1, and the cells prepared for electron microscopy as described in Material and Methods. Images of S1 decorated f-actin at or near the Sertoli cell nuclei are shown in Figure 7A,B. The barbed appearance of S1 decorated f-actin may be detected more readily in isolated filaments (arrow in Fig. 7B). F-actin bundles are recognized as exhibiting a ropelike appearance following decoration (asterisk in Fig. 7B). Both deep and shallow nuclear invaginations contain S1-decorated f-actin and the filaments appear to terminate at the nuclear envelope or nuclear matrix. Shallow infoldings may represent an intermediate step in the formation, or dissolution of a nuclear invagination. Bundles of decorated f-actin, on the other hand, course alongside the perimeter of the nuclear capsule. Non-decorated intermediate filaments are also shown in Figure 7C. Wisp-like configurations are noted far-
ther away from the nucleus. Individual intermediate filaments are also seen to end at the nuclear capsule. To test whether Sertoli cell nuclear shape is affected by microtubule or microfilament disrupting agents, Sertoli cells were cultured in the presence of either colchicine or vinblastine and cytochalasin D, respectively. These drugs had no effect on nuclear shape (data not shown). The only detectable effect of microtubule depolymerizing agents noted was the formation of an intermediate filament perinuclear cap (Fig. 8). This perinuclear cap of intermediate filaments is not generally present in the absence of microtubule depolymerizing agents.
DISCUSSION Because the orientation of the Sertoli cell nuclei in vivo varies with different stages of the cycle of the seminiferous epithelium, and the shape itself of the nucleus is characterized by deep invaginations, a study was initiated to identify the motive force responsible for these phenomena. Specifically, the Sertoli cell perinuclear filaments of in vivo and in vitro cells were characterized by morphological means. The results revealed that the nuclei of Sertoli cells in vivo and in vitro are surrounded by microtubules, microfilaments, and intermediate filaments. However, treatment of Sertoli cells in vitro with either microtubule or microfilament depolymerizing drugs had no effect on the morphology of nuclear shape, although the cytoskeletal elements susceptible to these drugs, respectively, depolymerized and disappeared. Thus, these results suggest that prominent perinuclear cytoskeletal elements do not play a role in modulating Sertoli cell nuclear invaginations, and perhaps are also not involved in modifying nuclear orientation. The demonstration that microfilaments are present in the perinuclear zone of the Sertoli cell in vivo and in vitro revealed in the present study is new (Fig. 2). Previous investigators have reported microtubules (Vogl, 1988) and intermediate filaments (Amlani and Vogl, 1988) in this zone, but did not detect f-actin. Although we emphasize that the presence of microfilaments in the perinuclear zone is a rare event when observed in conventionally fixed tissue, microfilaments are certainly present. One likely explanation of the fact that previous investigators did not detect f-actin filaments in the perinuclear zone is that both OsO, and glutaraldehyde are known t o destroy f-actin filaments devoid of actin-binding-proteins (Maupin-Szamier and Pollard, 1978; Lehrer, 1981). Treatment with tannic acid would in theory significantly enhance the preservation of microfilaments (Maupin and Pollard, 1983). However, penetration of tannic acid requires prior permeabilization of the plasma membrane with detergents. The permeabilization of the plasma membrane may lead to changes in the cytoplasmic milieu which may significantly alter the preservation of f-actin in the perinuclear zone. Thus, a question remains whether failure to observe cytoskeletal elements in vivo represents the state of activity of Sertoli cells at different stages of the cycle, or merely expresses inadequate fixation. Also, the specific functional significance of these microfilaments remains to be elucidated. However, that
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 4. Sertoli cell morphology in vitro. The shape of the nuclei of Sertoli cells in vitro is readily observed at the light (A) and electron (B) microscopic levels in cells sectioned parallel to their plane of growth. Note that nuclear infoldings are prominent and that the in vivo euchromatin pattern is maintained. Also, a single nucleoli is demonstrated in favorable profiles. A: Bar = 10 microns; x 1,310.B: Bar = 2 microns; X 5,200.
225
226
C.A. SUAREZ-QUIAN AND M. DYM
Fig. 5. Sertoli cell perinuclear filaments, tannic acid fixation I. A profile of a Sertoli cell nucleus from a cultured cell fix-permeabilized with tannic acid-saponin and sectioned parallel to the plane of growth is illustrated. A higher magnification view of the enclosed rectangle in A is shown in B. The enclosed area represents the cytoplasm contained within a nuclear infolding. Note the prominent microtubules,
some of which appear to terminate at the nuclear capsule. In B, the large arrowheads indicate the nuclear pores prsent in the nuclear capsule and the small arrowheads display a microtubule that appears to extend from one part of the nuclear capsule to another. A: Bar = 1 micron; x 13,500. B: Bar = 0.5 micron; x 50,000.
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 6. Sertoli cell perinuclear filaments, tannic acid fixation 11. Profiles of Sertoli cell nuclei of cells fixed-permeabilized with tannic acid-saponin and sectioned parallel to their plane of growth are illustrated in A and B. In A, the nuclear envelope was cut tangentially, whereas in B the nuclear envelope was cut in cross section. The white
227
arrows in A indicate the nuclear pores. In A and B arrowheads indicate the course of microtubules. Note that microtubules appear to circle the nucleus, emanate from (or end at) the nuclear capsule, or extend from one nuclear area to another within a nuclear infolding. Bar = 0.5 microns; x 23,000.
228
C.A. SUAREZ-QUIAN AND M. DYM
Fig. 7. Sertoli cell perinuclear filaments, S1 decoration. In A, a deep nuclear invagination containing a bundle of decorated f-actin filaments (asterisk) is illustrated. Individual filaments illustrating a polarity away from the nuclear capsule may be observed (arrow).In B, a shallow nuclear infolding containing decorated f-actin filaments is shown. A bundle of decorated f-actin filaments running along the
perimeter of the nuclear capsule is indicated by the asterisk. The black-on-white arrow indicates that the barbed end of one microfilament is away from the nuclear envelope. In C , intermediate filaments (arrows) that do not decorate with S1 are indicated. A: Bar = 0.25 microns; x 90,000. B,C: Bar = 0.5 microns; x 50,000.
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 8. Intermediate filaments. The effect of a 24-hour incubation of Sertoli cells with M colchicine is illustrated. Colchicine had no effect on nuclear morphology. The only significant effect of colchicine on Sertoli cells detected was that intermediate filaments coalesced into a perinuclear cap. Bar = 0.25 microns; x 130,000.
229
230
C.A. SUAREZ-QUIAN AND M. DYM
significant numbers of microfilaments are present and appear t o end at the nuclear capsule suggest a consequential role in some aspect of nuclear activity. An alternative means used to study Sertoli cell filaments at the ultrastructural level was to examine Sertoli cells in vitro. Using tannic acid to enhance the preservation of cytoplasmic fibers, the close association between microtubules and the nuclear capsule was noted. In particular, microtubules were detected terminating at the nuclear capsule in close proximity to the nuclear pores. When Sertoli cell cultures were prepared and the f-actin filaments decorated with S1, microfilaments were observed to form a prominent component of the cytoskeleton in the perinuclear zone. These results are consistent with the observation that microfilaments are located in the perinuclear zone of Sertoli cells in vivo. The enhanced visualization of the decorated filaments is probably the result of two factors: 1)administration of S1 to in vitro cells is significantly much easier to perform; 2) there is much less likelihood that variations exist between in vitro cells regarding the penetration of the S1 fragment to decorate f-actin. Furthermore, intermediate filaments were readily distinguished from the decorated f-actin filaments. Since thermodynamic considerations favor a rounded nucleus, it is intriguing to speculate first on the function of Sertoli cell nuclear invaginations and second on their control. First, nuclear infoldings may occur to maximize the exchange area between the nuclear contents and the cytoplasm. Consistent with this observation are the tracks of microtubules observed within nuclear infoldings and in the perinuclear cytoplasm. The microtubules may represent the structural basis on which cytoplasmic and nuclear components travel to their final destination. Evidence is present in the literature that microtubules form a vital link in the transit of components throughout the cytoplasm (Vale et al., 1985a,b). That exchange of components exists between the cytoplasm and the nucleus is amptly documented and is known to occur at nuclear pores. The observations that microtubules appear to terminate at, or extremely close to, nuclear pores is consistent with this hypothesis. Second, the regulation of nuclear infoldings is known to be dependent on normal levels of circulating androgens (Dym and Raj, 1977; Dym, 1977; Chemes et al., 1977). In addition, nuclear infoldings appear during maturation, correlative evidence that invaginations are hormonally dependent (Cameron and Markwald, 1975; Ramos and Dym, 1979; Hatier and Grignon, 1980). A similar requirement was noted for nuclear invaginations of Sertoli cells in vitro in this invagination. These data imply that nuclear invaginations are regulated by hormones and consequently are required in normal spermatogenesis. In this context it is important to note that earlier studies proposed that folliclestimulating hormone exercises control over the Sertoli cell cytoskeleton (Dedman et al., 1979; Welsh et al., 1980). Therefore, our results may provide the structural basis by which extracellular signals recruit the cytoskeleton and directly alter nuclear activity. Whether the microtubule tracks and/or individual mi-
crofilaments observed in the perinuclear cytoplasm work in concert or independently, however, remains to be determined.
ACKNOWLEDGMENTS This work was funded by NIH grants HD23484 to C.A.S.-Q. and HD16260 to M.D.
REFERENCES Amlani, S., and Vogl, A.W. (1988) Changes in the distribution of microtubules and intermediate filaments in mammalian Sertoli cells during spermatogenesis. Anat. Rec., 220:143-160. Begg, D.A., Rodewald, R., and Rebhun, L.I. (1978) Visualization of actin filament polarity in thin-section. J. Cell Biol., 79:846-852. Cameron, D.F., and Markwald, R.R. (1975) Histochemical and ultrastructural observations on normal and follicle stimulating hormone-injected prepubertal rat Sertoli cells. In: Hormonal Regulation of Spermatogenesis. F.S. French, V. Hansson, E.M. Ritzen, and S.N. Nayfeh, eds. Plenum Publishing Corp., New York, pp. 479493. Chemes, H.E., Dym, M., Fawcett, D.W., Javadpour, N., and Sherins, R.J. (1977) Pathophysiological observations of Sertoli cells in patients with germinal aplasia or severe germ cell depletion. Ultrastructural findings and hormone levels. Biol. Reprod., 17:108-123. Chemes, H.E., Dym, M., and Fhj, H.G.M. (1979)Hormonal regulation of Sertoli cell differentiation. Biol. Reprod., 21:251-262. Dedman, J.R., Brinkley, B.R., and Means, A.R. (1979) Regulation of microfilaments and microtubules by calcium and cyclic AMP. In: Advances in Cyclic Nucleotide Research, Vol. 11.P. Greengard and G.A. Robinson, eds. Raven Press, New York, pp. 131-174. Dym, M. (1977) The role of the Sertoli cell in spermatogenesis. In: Male Reproductive System. R. Yates and M. Gordon, eds. Mason Press, New York, pp. 155-169. Dym, M., and Raj, H.G.M. (1977) Response of adult rat Sertoli cells and Leydig cells to depletion of luteinizing hormone and testosterone. Biol. Reprod., 17:676-696. Dym, M., and Cavicchia, J.C. (1978) Functional morphology of the testis. Biol. Reprod., 181-15. Elftman, H. (1950) The Sertoli cell cycle in the mouse. Anat. Rec., 106381-393. Fawcett, D.W. (1975) Ultrastructure and function of the Sertoli cell. In: Handbook of Physiology, Vol. 5. D.W.Hamilton and R.O. Greep, eds. American Physiological Society, Washington, D.C., pp. 21-55. Suara-Quian, C.A., Kleinman, H.K., and Hadley, M.A., Byers, S.W., Dym, M. (1985) Extracellular matrix regulates Sertoli cell differentiation, testicular cord formation, and germ cell development in vitro. J. Cell Biol., 101:1511-1522. Hatier, R., and Grignon, G. (1980) Ultrastructural study of Sertoli cells in rat seminiferous tubules during intrauterine life and the postnatal period. Anat. Embryol., 16011-28. Karnovsky, M.J. (1971) Use of ferrocyanide-reduced osmium tetroxide in electron microscopy. J. Cell Bio., Abstract 284. Leblond, C.P., and Clermont, Y. (1952) Definition of the stages of the cycle of the seminiferous epithelium. Ann. N.Y. Acad. Sci., 55548572. Lehrer, S.S. (1981)Damage to actin filaments by glutaraldehyde: Protection by tropomyosin. J . Cell Bio., 90459-466. Maupin-Szamier, P., and Pollard, T.D. (1978) Actin filament destruction by osmium tetroxide. J. Cell Bio., 77:837-852. Maupin, P., and Pollard, T.D. (1983) Improved preservation and staining of HeLa cell actin filaments, clathrin-coated membranes, and other cytoplasmic structures by tannic acid-glutaraldehyde-saponin fixation. J . Cell Biol., 9651-62. Nistal, M., Paniasua, R., Abaurrea, M.A., and Santamaria, L. (1982) Hyperplasia and the immature appearance of Sertoli cells in primary testicular disorders. Hum. Path. 13:3-12. Parvinen, M., and Ruokonen, A. (1982) Regulation of the seminiferous epithelium. Endocr. Rev. 3:404-417. Ramos, AS., and Dym, M. (1979) Ultrastructural differentiation of rat Sertoli cells. Biol. Reprod., 21:909-922. Regaud, C. (1901) Etudes sur la structure des tubes seminiferes et sur la spermatogenese chez les mamiferes. Arch. Anat. Microsc. Morphol. Exp., 4101-155. Suarez-Quian, C.A., Dym, M., Makris, A., Brumbaugh, J., Ryan, K.J., and Canick, J.A. (1983) Estrogen synthesis by immature rat Sertoli cells in vitro. J . Androl., 4:203-209.
SERTOLI CELL PERINUCLEAR FILAMENTS Vale, R.D., Schnapp, B.J., Reese, T.S., and Sheetz, M.P. (1985a) Organelle, bead, and microtubule translocations promoted by soluble factors from the squid giant axon. Cell, 40559-569. Vale, R.D., Reese, T.S., and Sheetz, M.P. (1985133 Identification of a novel force-generating protein, kinensin, involved in microtubulebased motility. Cell, 42:39-50. Vogl, A.W. (1988) Changes in the distribution of microtubules in rat Sertoli cells during spermatogenesis. Anat. Rec., 222:34-41.
231
Welsh, M.J., Van Sicle, M., and Means, A.R. (1980) Possible involvement of cyclic AMP, calcium and cytoskeleton in control of protein secretion by Sertoli cells. In: Testicular Development, Structure, and Function. A. Steinberger and E. Steinberger, eds. Raven Press, New York,pp. 89-98.
Characterization of Sertoli Cell Perinuclear Filaments CARLOS A. SUAREZ-QUIAN AND MARTIN DYM Georgetown University Medical Center, Department of Anatomy and Cell Biology, Washington, D.C. 20007
KEY WORDS
Sertoli cell, Nucleus, Cytoskeleton
ABSTRACT Sertoli cell nuclei are characterized by deep invaginations and, in addition, the orientation of the nuclei with respect to the wall of the seminiferous tubules varies during the cycle of the seminiferous epithelium. These events may be the result of cytoplasmic filaments acting a t the level of the nuclear capsule and may represent significant changes in Sertoli cell activity. Thus, a study was performed to characterize the nature of the perinuclear filaments of Sertoli cells in vivo and in vitro. In Sertoli cells in vivo, microtubules and microfilaments were often detected in the perinuclear cytoplasm, and these cytoskeletal components were observed to course either parallel to, or abut at, the nuclear capsule. In Sertoli cells in vitro, the nuclear infoldings are retained and the perinuclear cytoskeleton was shown to contain microtubules, f-actin, and intermediate filaments. A fixationpermeabilization protocol employing tannic acid-saponin was used and it significantly enhanced the preservation of cytoskeletal components. The presence of f-actin was demonstrated by using the S1 fragment of muscle myosin to decorate the microfilaments. Treatment of the cultured cells with either microtubule or f-actin depolymerizing agents had no effect on nuclear shape. Thus, at present, the function of the prominent perinuclear cytoskeletal components remains unknown.
INTRODUCTION
serum testosterone levels by passive immunization with LH antisera also decreases Sertoli cell nuclear The orientation of Sertoli cell nuclei in situ varies infoldings (Dym and Raj, 1977). These data suggest during the cycle of the seminiferous epithelium (Re- that changes in nuclear shape and orientation are not gaud, 1901; Elftman, 1950); to date, this change is per- haphazard, but rather a required function of normal haps the most prominent morphological feature of Ser- Sertoli cell activity. toli cells known to vary during the cycle (Parvinen and Ultrastructural observations of Sertoli cells in vivo Ruokonen, 1982). A quantitative analysis of this phe- have revealed a perinuclear zone of filaments (Fawcett, nomena revealed that the nuclei generally adopted one 1975; Dym, 1977; Dym and Cavicchia, 1978). Thereof two configurations (Leblond and Clermont, 1952). fore, the composition and possible involvement of these During the early part of the cycle, nuclei appeared to cytoplasmic filaments in generating and maintaining exhibit a greater length in their parallel axis, with nuclear shape was examined. In addition, the relerespect to the basement membrane of seminiferous tu- vance of Sertoli cell nuclear infoldings in spermatogenbules, than in their perpendicular axis. As the cycle of esis is discussed. seminiferous epithelium progressed, more nuclei gained a perpendicular orientation, and a t stage VIII, MATERIALS AND METHODS where sperm release occurs, the number of perpendicTissue Preparation ularly oriented nuclei became maximal. Pressure on Six adult Sprague-Dawley rats were anesthetized the lateral borders of Sertoli cells, generated by the with Nembutal and the testes fixed by perfusion via rapid increase of developing germ cells undergoing mitosis during the early stages of the cycle, was believed the abdominal aorta with 5% glutaraldehyde in 0.2 M to passively produce the variation in the nuclear shape s-Collidine buffer. After 15 minutes, the testes were removed, cut into 1mm cubes, and fixed for an addi(Leblond and Clermont, 1952). tional two hours in the same solution. The testes pieces Numerous and deep invaginations, often extending throughout its length, are another easily recognized were rinsed three times in buffer and post-fixed for one feature of Sertoli cell nuclei in sections of testes. Such hour in potassium ferrocyanide-reduced OsO, a t room nuclear infoldings develop during maturation of the temperature (Karnovsky, 1971). Next, the testes were Sertoli cell (Ramos and Dym, 1979; Hatier and Grig- rinsed in buffer and dehydrated through a graded senon, 1980), and appear to depend upon normal levels of ries of alcohols, followed by propylene oxide. The testes circulating plasma gonadotropins (Cameron and Markwald, 1975; Chemes et al., 1979). Consistent with these observations is the fact that in diseased testes, where serum testosterone levels decrease, or testosterone inReceived January 17,1990;accepted in revised form May 30, 1990. sensitivity is recognized as the biochemical lesion, Serreprint requests to Dr. Carlos A. Suarez-Quian,Georgetown Univertoli cell nuclear infoldings are diminished (Chemes et sityAddress Medical Center, Department of Anatomy and Cell Biology, 3900 Reservoir al., 1977; Nistal et al., 1982). In addition, depletion of Road, N.W., Washington, D.C.20007.
0 1992 WILEY-LISS, INC.
220
C.A. SUAREZ-QUIAN AND M. DYM
were infiltrated with an 1:lmixture of propylene oxide and Epon, and embedded in Epon. For light microscopic examination, one micron sections were cut, mounted on slides, and stained with toluidine blue. Sections were examined with a Zeiss microscope using a 6 3 Planapo, ~ 1.4 N.A. objective. Photographs were taken with Kodak 4162 film and processed per manufacturer’s instructions. For electron microscopy, 60-90 nm sections were prepared using an LKB ultramicrotome. Sections were counterstained with uranyl acetate and lead citrate and examined with a Jeol 1200 EX electron microscope.
Preparation of Sertoli Cell Cultures Sertoli cell cultures were prepared as described previously (Suarez-Quian et al., 1983) as modified (Hadley et al., 1985), except that cells were plated onto Aclar coverslips (Allied Chemicals) coated with rat tail collagen (Collaborative Research, Walthan, MA). Briefly, testes of 20-day-old rats were excised, the tunica albuginea removed, and the seminiferous tubules subjected to three sequential enzymatic digestions consisting of collagenase and DNAse (Worthington). After each digestion the cells were allowed to pellet at unit gravity and washed in Dulbecco’s minimal essential medium (DMEM). The Sertoli cell enriched cell pellet resulting after the last digestion was diluted in a serum free defined medium (SFDM),consisting of DMEM plus 100 ng/ml FSH (NIADDK-oFSH-16),2 p,g/ml insulin, 5 p,g/ ml human transferrin, 50 ng/ml vitamin A, 200 ng/ml vitamin E, lo-’ M hydrocortisone, lo-’ M testosterone, lo-’ M estradiol, 2 mM glutamine, 5 ng sodium selenate, 1mM sodium pyruvate, and 22 mM sodium lactate. Preparation of In Vitro Cells for Light and Electron Microscopy For conventional light and electron microscopic observation, cells were fixed and processed exactly as described above for tissue, but with the following modifications: After dehydration in 100% alcohol the cells were immersed in a 5050 mixture of ethanol-Epon plus 2% DMP 30 for two hours, followed by fresh Epon plus 2% DMP 30 and the plastic was allowed to hardened in a 60°C oven. After embedding, desired regions of the Epon containing cells were cut and glued to blocks for sectioning. Embedded cells were mounted for sectioning in either of two orientations: knife sectioned cells in a plane parallel to their plane of growth, or knife sectioned cells perpendicular to their plane of growth. Tannic Acid Fixation for Sertoli Cells To enhance the preservation of the cytoskeleton yet maintain membranes of Sertoli cells, an empirical fixation protocol was developed. This recipe is similar to that of Maupin and Pollard (1983), except that concentrations of saponin, tannic acid, and glutaraldehyde were modified. Sertoli cells in vitro were fixed for 30 minutes in the following: 2 mM MgC1, 50 mM KC1, 0.01% saponin, 0.1% tannic acid, 3.0%glutaraldehyde, and 0.1 M sodium phosphate buffer (pH 7.0). All steps
were performed as described above for tissue, except cells were post-fixed with OsO, in 0.1 M sodium phosphate buffer (pH 6.0) for only 15 minutes on ice without en bloc staining in uranyl acetate. Instead, thin sections were stained for four minutes using lead citrate only. Labeling of Sertoli Cells In Vitro With S1 Fragments The S1 fragment of the myosin molecule was prepared as described by Begg et al. (1978). Decoration of Sertoli cell f-actin with S1 was performed as follows: Cells were rinsed twice, two minutes each, in stabilization buffer (SB)consisting of 0.1 M PIPES (pH 6.9),2 mM EGTA, and 2 mM MgC1,. Next, the cells were extracted and decorated with S1 fragments for ten minutes on ice in the following solution: SB plus 0.1% Triton X-100, 0.5 mg/ml TAME (Sigma), and 1-2 mg/ml S1. The cells were then washed in SB and fixed in 1% glutaraldehyde, 0.2%tannic acid in 0.1 M sodium phosphate buffer (pH 7.0) for one hour. Next, the cells were post-fixed in 1% OsO, in 0.1 M sodium phosphate buffer (pH 6.0) at 4°C for 15 minutes. Afterwards, the cells were processed for electron microscopic examination as described above.
RESULTS In Vivo Observations A profile of a seminiferous tubule cross section is shown in Figure 1.Staging of the seminiferous tubules is based on the different cell associations and characteristic shapes of acrosome development and spermatids. The perpendicular axis of Sertoli cell nuclei, when compared to the parallel axis, with respect to the basement membrane, is altered during the seminiferous tubule cycle, although all nuclei within the same stage do not exhibit an identical orientation. That is, as shown in Figure 1, a single stage may contain Sertoli cell nuclei exhibiting both orientations. Thus, it is a statistical percentage of the total number of nuclei in one stage versus another that alter their orientation (Leblond and Clermont, 1952). Ultrastructural studies were performed in Sertoli cells in vivo to characterize the composition of the perinuclear filaments (Fig. 2). The Sertoli cell nuclei are readily identified by their vast amount of euchromatin, a prominent nucleolus, and nuclear infoldings. The perinuclear cytoplasm of Sertoli cells, a distance of approximately 0.25 microns, is commonly devoid of cytoplasmic organelles (arrowheads in Fig. 2). Although cytoplasmic organelles may enter nuclear infoldings, they are also excluded from the 0.25 micron perinuclear zone. At higher magnification (Fig. 2B), the perinuclear zone devoid of cytoplasmic organelles may be observed to contain components of the cytoskeleton. These components include microtubules and microfilaments. Generally, however, the preservation of cytoskeletal components in the perinuclear exclusion zone is poor. The reason for this is not known, although both OsO, and glutaraldehyde destruction of microfilaments are possible agents of f-actin depolymerization (Maupin-Szamier and Pollard, 1978; Lehrer, 1981). Another profile of a Sertoli cell nuclei is illustrated
SERTOLI CELL PERINUCLEAR FILAMENTS
22 1
Fig. 1. Light microscopic observations of seminiferous tubules. For orientation purposes the lumen of tubules, Leydig cells (L)residing outside the tubule walls, and a few examples of spermatocytes (SC)and round spermatids (RS)are labeled. The nuclei of Sertoli cells (N)are readily identified in cross-sectional profiles of the seminiferous epithelium. They are characterized by their euchromatin pattern
and one prominent nucleoli. The apical cytoplasm of the Sertoli cells extends to the tubule lumen. It is this area of the Sertoli cells that forms crypts for elongated spermatids. Note that the nuclei of Sertoli cells may adopt one of two configurations, either parallel or perpendicular to the wall of the tubules. Bar = 10 microns; x 1,400.
in Figure 3. During germ cell differentiation a close association exists between elongated spermatids and Sertoli cell nuclei. Elongated spermatids plunge to-
wards Sertoli cell nuclei and often come to rest within nuclear infoldings as shown in Figure 3. It is during the spermatid plunge that Sertoli cell nuclei undergo
Fig. 2. Ultrastructural observations of Sertoli cells in vivo I. Characteristic morphological features of Sertoli cells are readily observed at the ultrastructural level (A). Nuclear infoldings, lack of heterochromatin, and a single reticular nucleoli with its two satellite karyosomes are prominent features of the Sertoli cell nucleus. Cytoplasmic organelles include the Golgi apparatus (G), mitochondria (m), and junctional complexes (JC) between adjacent Sertoli cells. Elongated spermatids (ES) residing within the apical cytoplasm of the Sertoli cell are demonstrated. A perinuclear exclusion zone (arrowheads) of
cytoplasmic organelle is also observed. The square shown in A is illustrated a t higher magnification in B. In B, the nuclear capsule has been section tangentially and nuclear pores are indicated by arrows. Microtubules (arrowheads) are present within 0.25 microns of the nuclear capsule. The diameter of thin filaments in the perinuclear area suggests that they are f-actin and they appear to abut directly against the nuclear capsule. A: Bar = 1micron; x 13,500. B: Bar = 0.25 micron; x 54,000.
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 3. Ultrastructural observations of Sertoli cells in vivo 11. A profile of a Sertoli cell nuclei oriented perpendicular to the base of the seminiferous epithelium is illustrated. Note the head portion of an elongated spermatid (ES)residing within a crypt of the Sertoli cell cytoplasm. The perinuclear cytoplasmic organelle exclusion zone (arrowheads) is quite prominent. Observe also that the exclusion zone is present on the side of the Sertoli cell nucleus adjacent to the elongated spermatid. Bar = 1 micron; x 15,600.
223
224
C.A. SUAREZ-QUIAN AND M. DYM
their change in orientation; the length of the parallel axis diminishes as the perpendicular axes of the nuclei elongate (compare with Fig. 1). At its closest, the spermatid plasma membrane is only 300-400 nm from the outer membrane of the Sertoli cell nuclear envelope (arrowheads in Fig. 3). This distance corresponds to the area inhabited by cytoplasmic perinuclear filaments and is also devoid of cytoplasmic organelles.
In Vitro Observations The characterization of Sertoli cell perinuclear filaments was performed in primary cell cultures. The shape of nuclei of cultured Sertoli cells plated onto rat tail collagen is shown in Figure 4A,B. Sertoli cells in vitro maintain their characteristic highly infolded nuclei if cultured in the presence of fetal calf serum or in serum free media but containing androgens and FSH. Cells fixed with 5% glutaraldehyde buffered in 0.2 M s-Collidine and prepared for electron microscopy reveal a perinuclear zone of filaments (data not shown). Generally, however, the preservation of the perinuclear zone of filaments is not consistent using this fixation protocol. The results of using tannic acid in combination with saponin, a mild detergent, in the primary fixation to enhance visualization of perinuclear filaments are shown in Figures 5 and 6. Using this fixation protocol various artifacts are noted: distended vesicles appear in the cytoplasm, cytoplasmic ground substance is generally absent, and mitochondria exhibit a less dense consistency. Presumably, the detergent saponin partially disrupted the cell membranes and rendered sertoli cells susceptible to osmotic damage. Nevertheless, preservation of cytoskeletal components in the presence of tannic acid is significantly enhanced. For example, microtubules were readily detected within the perinuclear cytoplasmic organelle exclusion zone and were prominent features of the nuclear infoldings (small arrowheads in Fig. 5B). In favorable profiles, microtubules appeared to end at the nuclear capsule, often in the vicinity of nuclear pores (white arrows in Fig. 6A,B). The presence of f-actin as a component of Sertoli cell perinuclear filamentous material was also determined. Sertoli cells were cultured, the f-actin decorated with the fragment of myosin S1, and the cells prepared for electron microscopy as described in Material and Methods. Images of S1 decorated f-actin at or near the Sertoli cell nuclei are shown in Figure 7A,B. The barbed appearance of S1 decorated f-actin may be detected more readily in isolated filaments (arrow in Fig. 7B). F-actin bundles are recognized as exhibiting a ropelike appearance following decoration (asterisk in Fig. 7B). Both deep and shallow nuclear invaginations contain S1-decorated f-actin and the filaments appear to terminate at the nuclear envelope or nuclear matrix. Shallow infoldings may represent an intermediate step in the formation, or dissolution of a nuclear invagination. Bundles of decorated f-actin, on the other hand, course alongside the perimeter of the nuclear capsule. Non-decorated intermediate filaments are also shown in Figure 7C. Wisp-like configurations are noted far-
ther away from the nucleus. Individual intermediate filaments are also seen to end at the nuclear capsule. To test whether Sertoli cell nuclear shape is affected by microtubule or microfilament disrupting agents, Sertoli cells were cultured in the presence of either colchicine or vinblastine and cytochalasin D, respectively. These drugs had no effect on nuclear shape (data not shown). The only detectable effect of microtubule depolymerizing agents noted was the formation of an intermediate filament perinuclear cap (Fig. 8). This perinuclear cap of intermediate filaments is not generally present in the absence of microtubule depolymerizing agents.
DISCUSSION Because the orientation of the Sertoli cell nuclei in vivo varies with different stages of the cycle of the seminiferous epithelium, and the shape itself of the nucleus is characterized by deep invaginations, a study was initiated to identify the motive force responsible for these phenomena. Specifically, the Sertoli cell perinuclear filaments of in vivo and in vitro cells were characterized by morphological means. The results revealed that the nuclei of Sertoli cells in vivo and in vitro are surrounded by microtubules, microfilaments, and intermediate filaments. However, treatment of Sertoli cells in vitro with either microtubule or microfilament depolymerizing drugs had no effect on the morphology of nuclear shape, although the cytoskeletal elements susceptible to these drugs, respectively, depolymerized and disappeared. Thus, these results suggest that prominent perinuclear cytoskeletal elements do not play a role in modulating Sertoli cell nuclear invaginations, and perhaps are also not involved in modifying nuclear orientation. The demonstration that microfilaments are present in the perinuclear zone of the Sertoli cell in vivo and in vitro revealed in the present study is new (Fig. 2). Previous investigators have reported microtubules (Vogl, 1988) and intermediate filaments (Amlani and Vogl, 1988) in this zone, but did not detect f-actin. Although we emphasize that the presence of microfilaments in the perinuclear zone is a rare event when observed in conventionally fixed tissue, microfilaments are certainly present. One likely explanation of the fact that previous investigators did not detect f-actin filaments in the perinuclear zone is that both OsO, and glutaraldehyde are known t o destroy f-actin filaments devoid of actin-binding-proteins (Maupin-Szamier and Pollard, 1978; Lehrer, 1981). Treatment with tannic acid would in theory significantly enhance the preservation of microfilaments (Maupin and Pollard, 1983). However, penetration of tannic acid requires prior permeabilization of the plasma membrane with detergents. The permeabilization of the plasma membrane may lead to changes in the cytoplasmic milieu which may significantly alter the preservation of f-actin in the perinuclear zone. Thus, a question remains whether failure to observe cytoskeletal elements in vivo represents the state of activity of Sertoli cells at different stages of the cycle, or merely expresses inadequate fixation. Also, the specific functional significance of these microfilaments remains to be elucidated. However, that
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 4. Sertoli cell morphology in vitro. The shape of the nuclei of Sertoli cells in vitro is readily observed at the light (A) and electron (B) microscopic levels in cells sectioned parallel to their plane of growth. Note that nuclear infoldings are prominent and that the in vivo euchromatin pattern is maintained. Also, a single nucleoli is demonstrated in favorable profiles. A: Bar = 10 microns; x 1,310.B: Bar = 2 microns; X 5,200.
225
226
C.A. SUAREZ-QUIAN AND M. DYM
Fig. 5. Sertoli cell perinuclear filaments, tannic acid fixation I. A profile of a Sertoli cell nucleus from a cultured cell fix-permeabilized with tannic acid-saponin and sectioned parallel to the plane of growth is illustrated. A higher magnification view of the enclosed rectangle in A is shown in B. The enclosed area represents the cytoplasm contained within a nuclear infolding. Note the prominent microtubules,
some of which appear to terminate at the nuclear capsule. In B, the large arrowheads indicate the nuclear pores prsent in the nuclear capsule and the small arrowheads display a microtubule that appears to extend from one part of the nuclear capsule to another. A: Bar = 1 micron; x 13,500. B: Bar = 0.5 micron; x 50,000.
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 6. Sertoli cell perinuclear filaments, tannic acid fixation 11. Profiles of Sertoli cell nuclei of cells fixed-permeabilized with tannic acid-saponin and sectioned parallel to their plane of growth are illustrated in A and B. In A, the nuclear envelope was cut tangentially, whereas in B the nuclear envelope was cut in cross section. The white
227
arrows in A indicate the nuclear pores. In A and B arrowheads indicate the course of microtubules. Note that microtubules appear to circle the nucleus, emanate from (or end at) the nuclear capsule, or extend from one nuclear area to another within a nuclear infolding. Bar = 0.5 microns; x 23,000.
228
C.A. SUAREZ-QUIAN AND M. DYM
Fig. 7. Sertoli cell perinuclear filaments, S1 decoration. In A, a deep nuclear invagination containing a bundle of decorated f-actin filaments (asterisk) is illustrated. Individual filaments illustrating a polarity away from the nuclear capsule may be observed (arrow).In B, a shallow nuclear infolding containing decorated f-actin filaments is shown. A bundle of decorated f-actin filaments running along the
perimeter of the nuclear capsule is indicated by the asterisk. The black-on-white arrow indicates that the barbed end of one microfilament is away from the nuclear envelope. In C , intermediate filaments (arrows) that do not decorate with S1 are indicated. A: Bar = 0.25 microns; x 90,000. B,C: Bar = 0.5 microns; x 50,000.
SERTOLI CELL PERINUCLEAR FILAMENTS
Fig. 8. Intermediate filaments. The effect of a 24-hour incubation of Sertoli cells with M colchicine is illustrated. Colchicine had no effect on nuclear morphology. The only significant effect of colchicine on Sertoli cells detected was that intermediate filaments coalesced into a perinuclear cap. Bar = 0.25 microns; x 130,000.
229
230
C.A. SUAREZ-QUIAN AND M. DYM
significant numbers of microfilaments are present and appear t o end at the nuclear capsule suggest a consequential role in some aspect of nuclear activity. An alternative means used to study Sertoli cell filaments at the ultrastructural level was to examine Sertoli cells in vitro. Using tannic acid to enhance the preservation of cytoplasmic fibers, the close association between microtubules and the nuclear capsule was noted. In particular, microtubules were detected terminating at the nuclear capsule in close proximity to the nuclear pores. When Sertoli cell cultures were prepared and the f-actin filaments decorated with S1, microfilaments were observed to form a prominent component of the cytoskeleton in the perinuclear zone. These results are consistent with the observation that microfilaments are located in the perinuclear zone of Sertoli cells in vivo. The enhanced visualization of the decorated filaments is probably the result of two factors: 1)administration of S1 to in vitro cells is significantly much easier to perform; 2) there is much less likelihood that variations exist between in vitro cells regarding the penetration of the S1 fragment to decorate f-actin. Furthermore, intermediate filaments were readily distinguished from the decorated f-actin filaments. Since thermodynamic considerations favor a rounded nucleus, it is intriguing to speculate first on the function of Sertoli cell nuclear invaginations and second on their control. First, nuclear infoldings may occur to maximize the exchange area between the nuclear contents and the cytoplasm. Consistent with this observation are the tracks of microtubules observed within nuclear infoldings and in the perinuclear cytoplasm. The microtubules may represent the structural basis on which cytoplasmic and nuclear components travel to their final destination. Evidence is present in the literature that microtubules form a vital link in the transit of components throughout the cytoplasm (Vale et al., 1985a,b). That exchange of components exists between the cytoplasm and the nucleus is amptly documented and is known to occur at nuclear pores. The observations that microtubules appear to terminate at, or extremely close to, nuclear pores is consistent with this hypothesis. Second, the regulation of nuclear infoldings is known to be dependent on normal levels of circulating androgens (Dym and Raj, 1977; Dym, 1977; Chemes et al., 1977). In addition, nuclear infoldings appear during maturation, correlative evidence that invaginations are hormonally dependent (Cameron and Markwald, 1975; Ramos and Dym, 1979; Hatier and Grignon, 1980). A similar requirement was noted for nuclear invaginations of Sertoli cells in vitro in this invagination. These data imply that nuclear invaginations are regulated by hormones and consequently are required in normal spermatogenesis. In this context it is important to note that earlier studies proposed that folliclestimulating hormone exercises control over the Sertoli cell cytoskeleton (Dedman et al., 1979; Welsh et al., 1980). Therefore, our results may provide the structural basis by which extracellular signals recruit the cytoskeleton and directly alter nuclear activity. Whether the microtubule tracks and/or individual mi-
crofilaments observed in the perinuclear cytoplasm work in concert or independently, however, remains to be determined.
ACKNOWLEDGMENTS This work was funded by NIH grants HD23484 to C.A.S.-Q. and HD16260 to M.D.
REFERENCES Amlani, S., and Vogl, A.W. (1988) Changes in the distribution of microtubules and intermediate filaments in mammalian Sertoli cells during spermatogenesis. Anat. Rec., 220:143-160. Begg, D.A., Rodewald, R., and Rebhun, L.I. (1978) Visualization of actin filament polarity in thin-section. J. Cell Biol., 79:846-852. Cameron, D.F., and Markwald, R.R. (1975) Histochemical and ultrastructural observations on normal and follicle stimulating hormone-injected prepubertal rat Sertoli cells. In: Hormonal Regulation of Spermatogenesis. F.S. French, V. Hansson, E.M. Ritzen, and S.N. Nayfeh, eds. Plenum Publishing Corp., New York, pp. 479493. Chemes, H.E., Dym, M., Fawcett, D.W., Javadpour, N., and Sherins, R.J. (1977) Pathophysiological observations of Sertoli cells in patients with germinal aplasia or severe germ cell depletion. Ultrastructural findings and hormone levels. Biol. Reprod., 17:108-123. Chemes, H.E., Dym, M., and Fhj, H.G.M. (1979)Hormonal regulation of Sertoli cell differentiation. Biol. Reprod., 21:251-262. Dedman, J.R., Brinkley, B.R., and Means, A.R. (1979) Regulation of microfilaments and microtubules by calcium and cyclic AMP. In: Advances in Cyclic Nucleotide Research, Vol. 11.P. Greengard and G.A. Robinson, eds. Raven Press, New York, pp. 131-174. Dym, M. (1977) The role of the Sertoli cell in spermatogenesis. In: Male Reproductive System. R. Yates and M. Gordon, eds. Mason Press, New York, pp. 155-169. Dym, M., and Raj, H.G.M. (1977) Response of adult rat Sertoli cells and Leydig cells to depletion of luteinizing hormone and testosterone. Biol. Reprod., 17:676-696. Dym, M., and Cavicchia, J.C. (1978) Functional morphology of the testis. Biol. Reprod., 181-15. Elftman, H. (1950) The Sertoli cell cycle in the mouse. Anat. Rec., 106381-393. Fawcett, D.W. (1975) Ultrastructure and function of the Sertoli cell. In: Handbook of Physiology, Vol. 5. D.W.Hamilton and R.O. Greep, eds. American Physiological Society, Washington, D.C., pp. 21-55. Suara-Quian, C.A., Kleinman, H.K., and Hadley, M.A., Byers, S.W., Dym, M. (1985) Extracellular matrix regulates Sertoli cell differentiation, testicular cord formation, and germ cell development in vitro. J. Cell Biol., 101:1511-1522. Hatier, R., and Grignon, G. (1980) Ultrastructural study of Sertoli cells in rat seminiferous tubules during intrauterine life and the postnatal period. Anat. Embryol., 16011-28. Karnovsky, M.J. (1971) Use of ferrocyanide-reduced osmium tetroxide in electron microscopy. J. Cell Bio., Abstract 284. Leblond, C.P., and Clermont, Y. (1952) Definition of the stages of the cycle of the seminiferous epithelium. Ann. N.Y. Acad. Sci., 55548572. Lehrer, S.S. (1981)Damage to actin filaments by glutaraldehyde: Protection by tropomyosin. J . Cell Bio., 90459-466. Maupin-Szamier, P., and Pollard, T.D. (1978) Actin filament destruction by osmium tetroxide. J. Cell Bio., 77:837-852. Maupin, P., and Pollard, T.D. (1983) Improved preservation and staining of HeLa cell actin filaments, clathrin-coated membranes, and other cytoplasmic structures by tannic acid-glutaraldehyde-saponin fixation. J . Cell Biol., 9651-62. Nistal, M., Paniasua, R., Abaurrea, M.A., and Santamaria, L. (1982) Hyperplasia and the immature appearance of Sertoli cells in primary testicular disorders. Hum. Path. 13:3-12. Parvinen, M., and Ruokonen, A. (1982) Regulation of the seminiferous epithelium. Endocr. Rev. 3:404-417. Ramos, AS., and Dym, M. (1979) Ultrastructural differentiation of rat Sertoli cells. Biol. Reprod., 21:909-922. Regaud, C. (1901) Etudes sur la structure des tubes seminiferes et sur la spermatogenese chez les mamiferes. Arch. Anat. Microsc. Morphol. Exp., 4101-155. Suarez-Quian, C.A., Dym, M., Makris, A., Brumbaugh, J., Ryan, K.J., and Canick, J.A. (1983) Estrogen synthesis by immature rat Sertoli cells in vitro. J . Androl., 4:203-209.
SERTOLI CELL PERINUCLEAR FILAMENTS Vale, R.D., Schnapp, B.J., Reese, T.S., and Sheetz, M.P. (1985a) Organelle, bead, and microtubule translocations promoted by soluble factors from the squid giant axon. Cell, 40559-569. Vale, R.D., Reese, T.S., and Sheetz, M.P. (1985133 Identification of a novel force-generating protein, kinensin, involved in microtubulebased motility. Cell, 42:39-50. Vogl, A.W. (1988) Changes in the distribution of microtubules in rat Sertoli cells during spermatogenesis. Anat. Rec., 222:34-41.
231
Welsh, M.J., Van Sicle, M., and Means, A.R. (1980) Possible involvement of cyclic AMP, calcium and cytoskeleton in control of protein secretion by Sertoli cells. In: Testicular Development, Structure, and Function. A. Steinberger and E. Steinberger, eds. Raven Press, New York,pp. 89-98.
