| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Morphology, ICB, Federal University of Minas Gerais, Belo Horizonte, MG, 31270-901, Brazil
b Laboratory of Cellular Biology, Department of Biology, Federal University of Juiz de Fora, ICB, Juiz de Fora, MG, Brazil
c Laboratory of Structural Biology, Department of Physiology, Southern Illinois University School of Medicine, Carbondale, Illinois 62901
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
12 nm apart). Tight junctional strands were sometimes present without an apparent association with organized actin bundles or they were tangential to actin bundles. En face views showed that gap junctions were commonly observed intercalated with tight junction strands. The results taken together suggest a relationship of organized actin with tight junction complexes. However, the occasional examples of tight junction complexes being not perfectly aligned with actin filament bundles suggest that a precise and rigidly organized actin-tight junction relationship described above is not absolutely mandatory for the presence or maintenance of tight junctions. Species variations in tight junction organization are also presented.
male reproductive tract, Sertoli cells, sperm motility and transport, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
30–50 nm) placed saccule of endoplasmic reticulum covers the actin filament bundles. Various proteins such as 
-actinin [5], zonula occludens 1-ZO-1 [6], occludin [7], claudin-11 [8, 9], espin [10], gelsolin, and others [11–14] have been recently described as components of the barrier or associated with the ectoplasmic specialization. Conventional electron microscopic views of the Sertoli-Sertoli barrier are generally displayed perpendicular or slightly oblique to the membranes of the cells that form the barrier. They reveal that the Sertoli cell barrier has bundles of organized actin filaments flanking the junctional region as well as more peripherally placed saccules of endoplasmic reticulum that run parallel to the junctional region. These components of Sertoli junctions were described some time ago [15, 16], and because of their uniqueness, they have drawn considerable attention.
Over the course of 20 years, one of us (L.D.R.) has collected micrographs of both conventional and en face views of the Sertoli-Sertoli junctions. En face views are those that lie in the plane of the membranes of adjoining cells or cross the membranes of two cells at an acute angle relative to the plane in which they lie. Such sections are difficult to find, and thus the collection of micrographs of junctional regions were not obtained in a single systematic study of normal animals. Micrographs in several species with various treatment or genetic conditions were taken and set aside until they were recently assembled as a group and analyzed. Whereas the collection of micrographs is not systematic, sufficient micrographs and sufficient similarities in the various conditions and treatments exist to allow analysis and yield useful information.
En face sections of conventionally fixed Sertoli-Sertoli junctional regions are not by themselves informative. Our laboratory has for the most part postfixed tissue with osmium ferrocyanide [17], a technique that enhances the membranes of the junctions and allows visualization in negative relief of tight and gap junctional particles. This technique does not well preserve the actin filament bundles pattern, but nevertheless, the presence of these bundles is in evidence. En face thin sections of osmium ferrocyanide-fixed tissues at Sertoli-Sertoli junctional regions often show the junctional particles and the actin filaments sufficiently well to provide new evidence of their structure and relationships. We have used both perpendicular (conventional) and en face sections, and occasional freeze fracture images, to describe previously unreported features of Sertoli-Sertoli junctions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The micrographs used in the present study were obtained through a suboptimal regimen, compared with the normal and desirable process used in a well-controlled study. Sectioned Sertoli-Sertoli cell junctions, both perpendicular and en face, were gathered during 20 years of studying normal, treated, and genetically modified animals. Animals were maintained in full compliance with the National Institutes of Health Guide for Care and Use of Laboratory Animals. Thus, the micrographs used in the present study were collected from previous studies, which had a variety of other objectives. Table 1 presents information regarding the animals used in the present study. In some instances, tissues were obtained from animals that had not completely attained puberty, although a Sertoli cell barrier had formed. Our rationale for conducting this study under less than optimal conditions follows.
|
Freeze Fracture
Testes were prepared for freeze fracture according to the method of Weber [18]. Briefly, tissues were perfused-fixed with 2.5% glutaraldehyde, diced, and immersed in the same fixative for an additional 15 min before being cryoprotected by infiltration with 30% glycerol. Tissue was next frozen in freon 22, cooled with liquid nitrogen, and then fractured in a Balzer 301 freeze-fracture apparatus (Balzer's High Vacuum Corp., Nashua, NH). Replicas were shadowed with platinum, coated with carbon, and then isolated by sodium hypochlorite digestion. Fractured membrane faces were described according to the nomenclature used by Branton [19].
Measurements and Determinations
The following measurements as illustrated in Figure 1 were made directly from micrographs (means ± SEM):
|
We determined from en face sections the percentage of junctional strands related to a single actin bundle. Only regions of en face sections in which both en face tight junctions and actin bundles could be clearly visualized were used for this determination. In particular, we determined the percentage of junctional strands that were related to the actin bundles zat the lateral extent of the bundle or central region of the bundle. Finally, we determined the percentage of clearly visible junctional strands extending more than 1 µm that did not run parallel with underlying actin filament bundles.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A drawing showing a sectioned Sertoli cell at its base and its association with other adjoined cells is shown for orientation purposes (Fig. 2).
|
Conventional and en face sections of ectoplasmic specialization-tight junction complexes were seen with electron microscopy near the base of the seminiferous tubules at the interface of adjacent Sertoli cells (Fig. 3, a and b, respectively). The expanse of the captured en face region was variable. The most information could be obtained from expansive en face sections, although some information could be obtained from slightly oblique sections of Sertoli-Sertoli junctions. Also, sections generally perpendicular to the plane of the junctional complex were useful for obtaining information, given they could be compared with en face sections.
|
Although postfixation with osmium ferrocyanide often allowed visualization of both tight junctions and ectoplasmic specializations, for reasons unknown to us, only some tissues fixed in this manner displayed sufficient quality to visualize both; most were inadequate in this respect.
Orientation and Measurements
Perpendicular sections generally revealed that bundles of actin of one Sertoli cell were directly opposite those of an adjacent Sertoli cell, although they sometimes exhibited a staggered appearance. The individual actin filaments were often difficult to discern because osmium ferrocyanide was used to postfix tissues. Actin filament bundles were generally parallel to one another (Fig. 4a), but occasionally, they diverged from one or another, or they curved slightly away from one another, or they were simply in nonparallel alignment (Fig. 4b).
|
Actin filament bundles in en face sections appeared as electron-dense bands exhibiting a width that ranged from 25.6 to 73.3 nm (mean ± SEM of 46.8 ± 2.5 nm), separated from one another by an electron translucent region (i.e., a space between actin bundles). In this space between actin bundles, the measurements taken in en face views were close to those obtained in conventional views (Fig. 4c). For example, in mice, measurements of this region yielded a mean of 52.9 ± 2.3 nm in en face sections, and a mean of 50.3 ± 7.5 nm in measurements of perpendicular sections. The means of the measurements of the space between actin bundles, taking into account both en face and conventional views are shown in Table 2. Apparently, there were no great differences between species.
|
General Features of Ectoplasmic Specialization in the Various Mammals Studied
Tight junction strands as seen in en face preparations were generally parallel with the course of the actin filament bundles. About 24.3% of tight junction strands were slightly misdirected from actin filament bundles, especially within a short distance of anastomosing with neighboring tight junction strands (Fig. 4a).
Actin filament bundles were generally but not always strictly parallel with one another. Actin filament bundles sometimes divided, and a tight junction branch sometimes occurred at the division and followed the actin filament branch (Fig. 4b).
Bundles of actin filaments were present, and entire sheets of actin could be viewed. Three or more tight junction strands were related to a sheet of actin (Fig. 4a). Two patterns of actin distribution appear to be related to junction strands. The first appears as a single bundle associated with two tight junction strands, and each strand is located at the edge of the actin bundle (Fig. 4a). In the second, a single bundle is associated with one centrally positioned junction strand in en face (Fig. 4b) and conventional (Fig. 4c) views. Table 2 shows the percentage of frequencies of their occurrence.
In rare instances, tight junctions were present without the presence of actin filament bundles (Fig. 5, a and b). The shortest distance between tight junction strands seen in conventional and en face sections was recorded as 12.5 nm (e.g., mouse and dog; Fig. 5, c and d), and close to half the width of a single bundle of actin filaments (25.7 nm).
|
Gap junctions were frequently seen intercalated among tight junction rows. Tight junctions that intercalated gap junctions often showed anastomotic connections (Fig. 5, d and e). En face sections and freeze fracture revealed that in such regions, the tight junction strands approximated one another and maintained a relatively constant distance of 23.9 nm ± 1.8 nm. The width of the space between tight junction rows was occupied by approximately one to three rows of junctional particles (mean of 2.5 ± 0.2 nm).
Cisternae of endoplasmic reticulum on two adjoining Sertoli cells were usually complementary, but in a minority of regions of ectoplasmic specialization only one Sertoli cell possessed a cistern (Fig. 6a). Thus, some junctional regions may not be lined by endoplasmic specializations. In some instances, no actin was apparent on one or both sides of a tight junction.
|
Specific Features of Ectoplasmic Specialization in the Various Mammals Studied
Mouse In most junctional regions the endoplasmic reticulum was similar to that described for a variety of other species. However, a form that appeared vesicular was noted in both normal and abnormal testes. The vesicular form often coexisted with the more conventional saccular form. Regions of the vesicles of the vesicular form were often closer to the plasma membrane than was noted when the saccular elements were found. Vesicular elements were often situated between actin filament bundles (Fig. 6b). Continuity was noted in some vesicles forming the ectoplasmic specialization and the rough endoplasmic reticulum (Fig. 6b).
Dog An atypically large space (Table 2) was present between the actin bundles and endoplasmic reticulum (Fig. 6c).
Opossum No cisternae of endoplasmic reticulum flanked the actin filaments. Actin filaments were not found in bundles (Fig. 6d).
Rat Tight junctions were often observed between small elements that form adhering-type junctions (Fig. 6e).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Numerous molecular features of the ectoplasmic specialization have been recently shown (referenced in the Introduction), but the most poorly studied is the fundamental relationship of the ectoplasmic specialization to the surface tight junctions. Based on the relationship described above, it is tempting to speculate that junctional particles within the membrane are linked, most likely indirectly, to the actin of the filament bundles of the ectoplasmic specialization. Links between actin, plasma membrane, and endoplasmic reticulum have been grossly visualized [20, 21], but such links must certainly possess greater molecular complexity and phylogenetic diversity, than anticipated by electron microscopy observations.
We have shown examples from a variety of species suggesting occasional variations in the relationship of actin filaments to tight junctions. The variability suggests that there is not a simple, unifying model that dictates how actin and tight junctions are related at this site. First, for example, some actin bundles are related to the center of the junctional strand; others to the lateral aspect. Second, the spacing of some junctions is so close that the bundle must bind to several junctions. Third, some bundles of actin are not apparently related to any junction. This situation was observed in the opossum, which shows no apparent organization of actin at Sertoli-Sertoli junctions. The existence of these examples, albeit a minority of the total ectoplasmic specialization-tight junction relationships, suggests either that tight junction strands can be organized or possibly maintained by means other than direct or indirect linkage to bundles of actin. However, overall, the main finding of the present study is a relationship between the parallel course of actin bundles and the parallel course of the tight junctions that is commonly seen in a variety of species.
Numerous components are associated with the actin-tight junction complex between Sertoli cells. Proteins such as ZO-1 and espin are described to be present at the cytoplasmic surface of tight junction regions [6, 10, 22, 23], and their functional roles are still under investigation. In guinea pigs, the distribution of the two isoforms of the tight junction-associated ZO-1 protein was shown by light microscopy to have a correspondence with the F-actin and G-actin distributions along the Sertoli cell membrane [6]. These authors suggest that the spatial organization of the subsurface actin accompanying cell junctions may affect the ZO-1 isoforms-plasma membrane association [6]. The espins are actin-binding and actin-bundling proteins localized parallel with actin bundles [23], and were recently considered a major actin-bundling protein of the Sertoli cell-spermatid ectoplasmic specialization [10].
Works from several laboratories have emphasized that a relationship exists between tight junctions and underlying actin filaments [24–27]. In general, the subplasma membrane cytoskeleton may act as a scaffold to anchor and position associated junctional plaque membrane proteins. In rats, cytochalasin D [18, 20], antilaminin [28], or glycerol [29] treatment disrupts actin filaments and tight junctions to affect the functional integrity of the Sertoli cell barrier. Cytochalasin D works by specifically disrupting F-actin, demonstrating that actin perturbation and a permeability breach are directly related. Further work is necessary to determine the cascade of events from actin to junctional protein disruption.
Are the ectoplasmic specializations related to maintenance of basolateral tight junctions? It is difficult to understand the necessity for the extensive cytoskeletal actin complex in this region. Other regions where tight junctions are present do not have highly organized actin like that seen at Sertoli-Sertoli junctions, yet they do have actin that is apparently important for tight junction integrity [30]. It has been proposed that the ectoplasmic specialization is important for stiffening this region of the Sertoli-Sertoli interface such that the interface could be maintained in relationship to basal compartment germ cells that would be moving at a specific time through such junctions to the adluminal compartment of the testis [1]. This is the most comprehensive explanation of their function because it is consistent with the function of similar-appearing ectoplasmic specializations located at the interface of elongating spermatids [4, 31]. The ectoplasmic specializations facing the acrosomal region of elongating spermatids function as an attachment plaque that holds the heads of elongating spermatids as they invaginate the lateral Sertoli cell to form crypts. Tight junctions are not present at this site. A trypsin-sensitive junction [32] in which the junctional links traverse a narrowed intercellular space has been documented [5]. The ectoplasmic specializations at their interface with elongated and elongating spermatids are removed or dissolved as the spermatids are released from their attachment to the Sertoli cell [1, 33].
Gap junctions were frequently found between tight junction strands. They were detected in thin sections, mainly in en face views, and in freeze-fracture preparations, where they appeared as small clusters of packed particles as previously described for other tissues by freeze-fracture methodology [34]. Gap junctions related to Sertoli cell tight junctions are larger and more frequently seen in animals devoid of germ cells [35] and in developing animals more often than in adults [4]. To our knowledge, this is the first report showing individual gap junction particles without the use of tracers to highlight them.
The endoplasmic reticulum showed variations in arrangement and morphology. In mice, the conventional saccular form of this organelle seen in ectoplasmic specializations coexisted with a less frequent vesicular form. We believe that these different aspects were not artifacts of fixation because they were present when tissues showed excellent fixation. A vesicular form of the endoplasmic reticulum was also observed in areas of developing Sertoli cell tight junctions in young mice and described as sites of membrane apposition that precede the reticular organization [42].
Tight junction strands in epithelia are known to circumscribe the apico-lateral regions of many epithelial cells and to create a permeability constraint to the flow of substances from the apex to the base of a cell. This condition is also similar to the barrier formed by Sertoli cells, the only differences being the location of the junction (basolaterally) and the number of rows of junctions present [4]. We are aware of one other site, in the stria vascularis, where the number of rows of tight junctions is equivalent to that observed at the interface of Sertoli cells [30]. There is no equivalent to the Sertoli-Sertoli ectoplasmic specialization of the testis, so it appears that ectoplasmic specializations are not important in the maintenance of a high number of junctional strands. It appears that the highly organized ectoplasmic specializations are not necessary to maintain the circumferential orientation of junctions, because this is a property of other junctional sites where actin shows no apparent organization similar to that of an ectoplasmic specialization. Thus, it appears that the ectoplasmic specialization is related to cell-cell junctions but it is multifunctional, and it possesses slightly different structural and functional properties at each site at which it is present.
|
ACKNOWLEDGMENTS |
|---|
We acknowledge Bruno Garzon for the fine illustrations used in this paper.
|
FOOTNOTES |
|---|
Received: 13 March 2002.
First decision: 26 March 2002.
Accepted: 29 April 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. P.Y. Lee and C. Y. Cheng Ectoplasmic specialization, a testis-specific cell-cell actin-based adherens junction type: is this a potential target for male contraceptive development? Hum. Reprod. Update, July 1, 2004; 10(4): 349 - 369. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
