Human Reproduction Update Advance Access originally published online on June 10, 2004
Human Reproduction Update 2004 10(4):349-369; doi:10.1093/humupd/dmh026
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ectoplasmic specialization, a testis-specific cell–cell actin-based adherens junction type: is this a potential target for male contraceptive development?
Population Council, 1230 York Avenue, New York, NY 10021, USA
1 To whom correspondence should be addressed. Email: y-cheng{at}popcbr.rockefeller.edu
 |
Abstract |
|---|
 TOP Abstract IntroductionEctoplasmic specialization in...Components of ESReferences |
|---|
The seminiferous tubule of the mammalian testis is largely composed of Sertoli and germ cells, which coordinate with Leydig cells in the interstitium and perform two major physiological functions, namely spermatogenesis and steroidogenesis respectively. Each tubule is morphologically divided into (i) the seminiferous epithelium composing Sertoli and germ cells, and (ii) the basement membrane (a modified form of extracellular matrix); underneath this lies the collagen fibril network, the myoid cell layer, and the lymphatic vessel, which collectively constitute the tunica propia. In the seminiferous epithelium, of rodent testes each type A1 spermatogonium (diploid, 2n) differentiates into 256 elongated spermatids (haploid, 1n) during spermatogenesis. Additionally, developing germ cells must migrate progressively from the basal to the luminal edge of the adluminal compartment so that fully developed spermatids can be released into the lumen at spermiation. Without this timely event of cell movement, spermatogenesis cannot reach completion and infertility will result. Yet developing round elongating/elongated spermatids must remain attached to the epithelium via a specialized Sertoli–germ cell actin-based adherens junction (AJ) type known as ectoplasmic specialization (ES), which is crucial not only for cell attachment but also for spermatid movement and orientation in the epithelium. However, the biochemical composition and molecular architecture of the protein complexes that constitute the ES have only recently been studied. Furthermore, the signalling pathways that regulate ES dynamics are virtually unknown. This review highlights recent advances in these two areas of research. It is expected that, if adequately expanded, these studies should yield new insights into the development of novel contraceptives targeted to perturb ES function in the testis. The potential to specifically target the ES may also mean that contraceptive action could be achieved without perturbing the hypothalamic–pituitary–testicular axis.
Key words: 1-(2,4-dichlorobenzyl)-indazole-3-carbohydrazide / ectoplasmic specialization / male contraception / Sertoli–germ cell AJ dynamics / testis
|
Introduction |
|---|
|
TOPAbstractIntroductionEctoplasmic specialization in...Components of ESReferences |
|---|
The testis is an organ in which millions of germ cells are produced daily. For instance, it is known that
150 x 106 sperm are produced per day in a typical man (Johnson et al., 1970
). The functional unit of the mammalian testis that produces germ cells is the seminiferous tubule where spermatogenesis occurs. Throughout spermatogenesis, a single type A1 spermatogonium (2n, diploid) divides and differentiates into 256 elongated spermatids (1n, haploid) in the seminiferous epithelium (Russell et al., 1990). Spermatogenesis, however, is regulated largely by FSH released from the pituitary gland and testosterone produced by Leydig cells in the interstitium via steroidogenesis (for a review, see de Kretser and Kerr, 1988). Figure 1 (left panel) is a schematic drawing of the seminiferous epithelium in the rat testis that illustrates the relative relationship of germ cells at different stages of their development with Sertoli cells and the different junctional types between Sertoli and germ cells in the epithelium. The right panel of Figure 1 is a schematic drawing that illustrates the three known protein complexes and their associated peripheral proteins found at the apical ectoplasmic specialization (apical ES) in the seminiferous epithelium of the rat testis, which is a testis-specific cell–cell actin-based adherens junction (AJ) type (see below for a detailed description and definition of apical and basal ES). The electron micrographs shown in Figures 2A–C are the magnified view of the seminiferous epithelium using cross-sections of an adult rat testis, which also illustrate the intimate relationships between Sertoli and germ cells (Figure 2B) and the ultrastructural feature of the apical ES (Figure 2C) (see also Figure 2D and Figure 1). The schematic drawing and micrographs shown in Figures 1 and 2 are typical of the testis in adult rats, mimicking many of the same features found in other mammals including mice and men.
|
|
The seminiferous epithelium can be subdivided into (i) the basal compartment and (ii) the adluminal compartment based on the relative location of the blood-testis barrier (BTB) (see Figure 1, left panel), which is also known as the seminiferous epithelium barrier (Figures 1 and 2A, B) (de Kretser and Kerr, 1988
; Russell et al., 1990; Byers et al., 1993). The BTB is constituted largely by tight junctions between Sertoli cells (also known as the Sertoli cell barrier), however, both basal ES, desmosome-like junctions, and gap junctions are also found at the BTB site in the rat and mouse testis (de Kretser and Kerr, 1988; Russell et al., 1990; Byers et al., 1993). Immediately underneath the basal compartment lies the tunica propria (Figure 1, left panel, Figure 2A, B), which is composed of both an acellular zone (i.e. basement membrane and collagen fibril network) and a cellular zone (i.e. myoid cells, lymphatic endothelium). Within the seminiferous epithelium, specialized testis-specific cell–cell actin-based AJ, such as the ectoplasmic specializations (ES) (Figures 1, 2C and D) and tubulobulbar complexes, and cell–cell intermediate filament-based desmosome-like junctions (Figure 1) (note: all three are anchoring junction types, see Table I), are responsible for anchoring developing germ cells onto the epithelium at different stages of the epithelial cycle. Cell junctions in the testis are classified primarily based on their relative locations and the attachment sites used for constituting the cytoskeleton (for a review, see Cheng and Mruk, 2002) (see Table I).
|
In the rat testis, spermatogenesis per se involves extensive junction restructuring when immature germ cells, such as preleptotene and leptotene spermatocytes, initially residing at the basal compartment, need to migrate to the adluminal compartment traversing the BTB for further development at late stage VIII and early stage IX of the epithelial cycle (Russell, 1977c
) (see Figures 1 and 2). Furthermore, spermatocytes must continue their journey migrating to the adluminal edge of the lumen while differentiating into round and elongated spermatids. As such, the testis per se is a novel in vivo model to study the biology and regulation of junction dynamics. Obviously, such studies, if adequately expanded, will also shed new light on male contraceptive development. For instance, if the key regulatory molecules that are important for the junction restructuring event are identified and characterized, they can become the prime targets for male contraceptive development. In this review, the biology and regulation of ES are discussed. We will not discuss desmosome-like junctions since less is known regarding this junction type, for example its molecular composition, despite earlier morphological studies (Russell, 1977a). A review on the tubulobulbar complex (TBC, a testis-specific actin-based adherens junction type) is also not provided here since it has recently been discussed and eminently reviewed (Russell, 1993). But in this context, it is of interest to note that TBC is another testis-specific AJ type in the seminiferous epithelium found at the site of BTB between Sertoli cells (basal TBC) as well as surrounding the heads of elongating/elongated spermatids (apical TBC) (Russell, 1993) (see also Figure 1, left panel). Apical TBC appears as a pair of tubular structures surrounded by actin with a balloon-like terminal bulbar structure restricted to the concave side of the elongating/elongated spermatid head, which is not visible until several days prior to spermiation at stage VII and is crucial to prevent premature release of elongating/elongated spermatids from the epithelium (for reviews, see Russell, 1993; Vogl et al., 1993). However, the precise structure of basal TBC at the BTB site is not well understood, and the precise biochemical composition of both basal and apical TBC is not known except for a recent report illustrating that cofilin, an actin-severing protein, is associated with the apical TBC in the rat seminiferous epithelium (Guttman et al., 2004). As such, it is clear that much research is needed to dissect this testis-specific AJ type biochemically. Yet ES is the most studied testis-specific AJ type in the seminiferous epithelium. Interestingly, an in-depth review on the protein complexes that constitute ES and their regulation is currently missing. As such, it is the goal of this review to fill in this gap. |
Ectoplasmic specialization in the seminiferous epithelium: basic structure and functions |
|---|
|
TOPAbstractIntroductionEctoplasmic specialization in...Components of ESReferences |
|---|
ES is an actin-based testis-specific AJ type between Sertoli cells at the basal compartment at the site of BTB as well as between Sertoli and germ cells at the adluminal compartment of the seminiferous epithelium in the rat testis (Figure 1, left panel) (for reviews, see Russell, 1980
; Vogl et al., 2000; Toyama et al., 2003). Since most of the earlier morphological studies of the ES were performed in rats (Russell, 1977b; Russell et al., 1990) including several recent biochemical studies (Lee et al., 2003, 2004; Lui et al., 2003b; Siu et al., 2003; Siu and Cheng, 2004), the structures of ES discussed herein are primarily based on the rat model. In the rat testis, the basal ES is situated near the BTB site between adjacent Sertoli cells throughout the epithelial cycle (for reviews, see Russell, 1980; Vogl et al., 2000; Toyama et al., 2003). Likewise, the apical ES is also present throughout the seminiferous epithelial cycle; it first appears between Sertoli cells and round spermatids in late stage VII and early stage VIII in the rat testis (Figure 1) (de Franca et al., 1993; for reviews, see Russell, 1980; Vogl et al., 2000; Toyama et al., 2003). It persists in the seminiferous epithelium and associates with spermatids up to step 19 in the mouse before the appearance of the tubulobulbar complex (for reviews, see Russell, 1980; Vogl et al., 2000; Toyama et al., 2003).
Structurally, apical ES consists of a narrow layer of hexagonally packed, parallel actin bundles sandwiched between Sertoli cell plasma membrane and a cistern of endoplasmic reticulum (ER) on the Sertoli cell side of the plasma membrane between an apposing Sertoli cell and an elongating/elongated spermatid (Figure 2C, D, see also Figure 1) (for reviews, see Russell, 1980; Vogl et al., 2000; Toyama et al., 2003). A typical apical ES found between a Sertoli cell and the head of an elongating spermatid is also shown in an electron micrograph in Figure 2C. Figure 2D is a schematic drawing that illustrates the ultrastructural feature of the apical ES in the seminiferous epithelium of an adult rat testis. It is noted that the basic ultrastructures of the basal and apical ES are virtually identical; however, ribosomes are found in the cytoplasmic side of ER in the basal, but not apical, ES, suggesting their constituent proteins could be different (for a review, see Vogl et al., 2000). Also, the intercellular space between two apposing Sertoli cells having the basal ES is sealed by tight junctions at the BTB site (Figure 1, left panel) (for reviews, see Russell, 1993; Vogl et al., 2000; Toyama et al., 2003). Furthermore, at the apical ES, actin bundles and ER are only found on the Sertoli cell side of the plasma membrane; tight junctions are not present, and little is known about the integral membrane junctional molecules in the apposing elongating/elongated spermatids (for reviews, see Russell, 1980, 1993; Vogl et al., 2000; Toyama et al., 2003). However, recent studies have shown that nectin-2, nectin-3, N-cadherin, laminin 
3, RhoB GTPase, Rab8B, zyxin and others are found at the apical ES site and are associated with elongating/elongated spermatids (Lee et al., 2003, 2004; Lui et al., 2003c; Siu et al., 2003; Lau and Mruk, 2003; Siu and Cheng, 2004). The primary function of ES is to facilitate germ cell movement, with an additional anchoring function to retain germ cells, in particular spermatids, in the epithelium until spermiation (for reviews, see Russell, 1980; Vogl et al., 2000; Toyama et al., 2003). In addition, it likely plays a role in the BTB regulation (for a review, see Vogl et al., 2000). For instance, the intimate relationship between basal ES and tight junction (TJ) was demonstrated by freeze–fracture technique, suggesting that they are present side-by-side in the BTB (Parreira et al., 2002) (Figures 1, 2). Although ES was first identified in the testis almost three decades ago (for reviews, see Russell, 1980; Vogl et al., 2000; Toyama et al., 2003), the precise mechanism(s) that regulates ES dynamics has yet to be elucidated. Recent studies have shed new light on the molecular architecture of the constituent protein complexes at the apical ES site. Also, the underlying mechanisms that regulate this AJ type in the testis have recently been identified and partially characterized.
|
Components of ES |
|---|
|
TOPAbstractIntroductionEctoplasmic specialization in...Components of ESReferences |
|---|
Multi-protein complexes
To date, three multi-protein complexes, namely the cadherin/catenin, the nectin/afadin/ponsin, and the integrin/laminin complexes are found in the ES (for reviews, see Cheng and Mruk, 2002; Takai and Nakanishi, 2003). For the cadherin/catenin and the nectin/afadin complexes, both are detected at the apical and basal ES sites, whereas the integrin/laminin complex is restricted mostly to the apical ES site. Table II summarizes some of the biochemical features and functions of different classes of proteins at the ES site and their interacting partners, some of which are discussed in this section. It must be noted that many of the proteins found at the ES site are common to AJ in other epithelia, and a deletion of these proteins can be fatal, as shown in results of gene knockout experiments in mice summarized in Table III.
|
|
The cadherin/catenin complex
The cadherin/catenin complex is one of the best studied AJ structural complexes in the testis (for a review, see Cheng and Mruk, 2002
) (Table II).
Cadherin. Neural (N) and epithelial (E)-cadherins are classic cadherins, belonging to the cadherin superfamily (for a review, see Yagi and Takeichi, 2000). Classic cadherins, including N- and E-cadherins, interact with both p120ctn and ß-/-catenin (see Figure 1) (for a review, see Perez-Moreno et al., 2003). Both N- and E-cadherins are found in the seminiferous epithelium of the testis in rats and/or mouse and are products of Sertoli and germ cells; and germ cells express more E-cadherin than Sertoli cells (Wu et al., 1993; Byers et al., 1994; Chapin et al., 2001; Johnson and Boekelheide, 2002b; Lee et al., 2003). In short, these data suggest that cadherins residing on Sertoli and germ cells can indeed confer cell adhesion function. This concept was rather provocative when it first emerged (Byers et al., 1994; Lee et al., 2003), since ultrastructural observations on the ES had suggested that Sertoli cells contributed most, if not all, of the adhesive devices for the ES (Figure 2C). Another recent report indeed supports this concept. For instance, nectin-3 was shown to reside exclusively in elongating/elongated spermatids that can form heterotypic interactions with nectin-2 residing in Sertoli cells at the site of apical ES (Ozaki-Kuroda et al., 2002). While N-cadherin was found at the site of apical and basal ES using fluorescent microscopy (Wine and Chapin, 1999; Johnson and Boekelheide, 2002b; Lee et al., 2003), these classic cadherins apparently were utilizing intermediate filament as the underlying cytoskeleton for attachment (Mulholland et al., 2001; Johnson and Boekelheide, 2002b). Yet we have recently completed a detailed biochemical analysis using co-immunoprecipitation techniques and lysates of testes and seminiferous tubules (note: these tubules were freed of Leydig and myoid cell contamination) to investigate this further. It was shown that virtually all N-cadherin in the Leydig and myoid cell-free seminiferous tubule lysates were using actin as the attachment site (Lee et al., 2003) similar to those found in other epithelia (for reviews, see Gumbiner, 2000; Yagi and Takeichi, 2000; Perez-Moreno et al., 2003). Yet 5% of E-cadherin in seminiferous tubule lysates was indeed associated with vimentin, an intermediate filament structural protein (Lee et al., 2004). Furthermore, when lysates of testes (with Leydig and myoid cells in addition to Sertoli and germ cells) were used for co-immunoprecipitation, it was shown that as much as 50% of the E- and N-cadherins were indeed associated with vimentin in the testis (Lee et al., 2004). Collectively, these analyses reveal that classic cadherins at the ES are using actin predominantly as attachment sites in the seminiferous epithelium with only 5% of E-cadherin using vimentin as its attachment site, whereas cadherins found in the interstitial cells and at the basal portion of the tubules are using intermediate filament as attachment sites (Lee et al., 2003).
Catenins. 
-Catenin is a 100–104 kDa actin-binding and -actinin-binding protein (Knudsen et al., 1995) (for a review, see Rudiger, 1998). The 92 kDa ß-catenin and 82 kDa -catenin, also called plakoglobin, can interact with cadherins at the highly conversed intracellular cytoplasmic domain called catenin binding domain (CDB) (Pierceall et al., 1995; Takayama et al., 1996). But their interactions with actin are mediated via -catenin (for a review, see Zhurinsky et al., 2000). ß- and -catenins share similar structural features, yet ß-catenin is predominantly localized at the site of AJ whereas -catenin is found in both AJ and desmosome-like junctions (for a review, see Zhurinsky et al., 2000) (Figure 1). ß- or -catenin in turn interacts with -catenin, linking cadherins to the underlying actin filament via -catenin and/or -actinin (for a review, see Zhurinsky et al., 2000). All three catenins are found in the testis (Cowin et al., 1986; Byers et al., 1994; Janssens et al., 2001; Lee et al., 2003). ß-Catenin co-localizes with N-cadherin at the site of ES in the basal compartment of the epithelium (Lee et al., 2003); it also co-localizes with classic cadherin at non-ES sites, possibly at the desmosome-like junctions (Mulholland et al., 2001). -Catenin was localized to the basal ES and myoid cells (Byers et al., 1994; Janssens et al., 2001). -Catenin was found at the site of basal ES (Cowin et al., 1986; Byers et al., 1994). The three catenins have been shown to form a complex with E- or N-cadherin in the testis (Lau and Mruk, 2003; Lee et al., 2003, 2004). A transmembrane phosphatase called PTPµ was shown to associate with the catenin/cadherin complexes and might take part in the signalling event (Brady-Kalnay et al., 1995), yet this possibility remains to be investigated in the testis.
p120ctn. More than four isoforms of p120ctn with Mr 65–120 kDa are found in different tissues, including the testis, due to alternative splicing (Mo and Reynolds, 1996; Johnson and Boekelheide, 2002a). p120ctn binds to cadherin at a site different from the ß-/-catenin binding site (Figure 1) (for a review, see Anastasiadis and Reynolds, 2000). It is still not known whether p120ctn facilitates or inhibits cell adhesion function (for a review, see Anastasiadis and Reynolds, 2000). p120ctn is localized to apical and basal ES, to the same site of N-cadherin and ß-catenin, and is also found in desmosome-like junctions, co-localizing with plectin (Wine and Chapin, 1999; Golenhofen and Drenckhahn, 2000; Johnson and Boekelheide, 2002a).
The nectin/afadin/ponsin complex
Nectin/afadin/ponsin is a recently identified putative ES structural unit (for reviews, see Cheng and Mruk, 2002; Takai and Nakanishi, 2003) (Table II). Nectin is an AJ-integral membrane protein that interacts with afadin, which in turn binds to the underlying actin filament (Figure 1). Ponsin is an afadin-binding protein also found in the testis (Mandai et al., 1999; Lebre et al., 2001) (Figure 1).
Nectin. Four nectins, namely nectin-1, -2, -3 and -4, are known to date (for a review, see Takai and Nakanishi, 2003). It is structurally associated with afadin at its cytoplasmic domain (for a review, see Takai and Nakanishi, 2003). Recent studies have shown that nectin-2 and nectin-3 found in spermatids can mediate adhesive interactions with nectin-2 found in Sertoli cells (Bouchard et al., 2000; Ozaki-Kuroda et al., 2002; Mueller et al., 2003). Nectin-2, -3 and -4, but not nectin-1, are found in the testis (for a review, see Takai and Nakanishi, 2003). Nectin-2 and -3 were shown to co-localize with actin and espin to ES (Ozaki-Kuroda et al., 2002; Mueller et al., 2003). Nectin-2–/– mice are infertile with defective Sertoli–spermatid junctions, resulting in sperm malformation (Bouchard et al., 2000; Ozaki-Kuroda et al., 2002; Mueller et al., 2003). Interestingly, nectin-3 immunostaining was still detected in apical ES in nectin-2–/– mice but with disturbed and diminished distribution (Ozaki-Kuroda et al., 2002; Mueller et al., 2003).
Afadin. Afadin has two splice variants, namely l-afadin (afadin) and s-afadin (a rat homologue of human AF-6 protein, which lacks the actin filament-binding domain at its C-terminus), with an apparent Mr of 205 and 190 kDa respectively (Mandai et al., 1997). Afadin is expressed by both Sertoli and germ cells and is localized to the site of apical and basal ES (Ozaki-Kuroda et al., 2002). Interestingly, afadin level was found to be diminished at the site of Sertoli-spermatid junctions, but not the Sertoli–Sertoli cell junctions in nectin-2–/– mice (Ozaki-Kuroda et al., 2002), suggesting that nectin-2 is required for the proper localization of afadin at the apical, but not basal, ES. This also implicates the presence of another yet-to-be identified nectin(s) or nectin-associated protein(s) at the site of basal ES.
Ponsin. Ponsin/SH3P12/CAP belongs to an adaptor family consisting of at least two other members: vinexin and Arg-binding protein 2 (ArgBP2) (for a review, see Kioka et al., 2002). It has at least 13 splice variants, and the Mr 93 kDa variant is the major afadin binding partner at the AJ site (Mandai et al., 1999). Ponsin is an afadin-binding protein (Mandai et al., 1999). It remains to be determined if ponsin is present at the site of ES in the seminiferous epithelium but studies by northern blots have positively identified ponsin in the testis (Mandai et al., 1999).
The integrin/laminin complex
The integrin/laminin complex was initially identified as a junction complex at the site of cell–substratum focal adhesion (for a review, see Hynes, 2002). Yet the integrin/laminin complex is one of the first AJ structural protein complexes found at the site of ES in the testis (Koch et al., 1999; Palombi et al., 1992; Salanova et al., 1995; Siu and Cheng, 2004) (Table II).
Integrin. Integrin is a transmembrane protein receptor, composed of and ß subunits. To date, eight ß and 18 subunits have been found in mammalian epithelia, and most integrin receptors are restricted to the focal adhesion site between cell and extracellular matrix (for a review, see Hynes, 2002). Interestingly, 6ß1- and 4ß1-integrins are two putative integral membrane proteins at the site of apical ES, co-localzing with actin, on the Sertoli cell side (Palombi et al., 1992; Salanova et al., 1995, 1998; Chapin et al., 2001; Mulholland et al., 2001). The interacting partners of ß1-, 4- and 6-integrins in the testis can be found in Table II. The findings that Sertoli–germ cells are using components of the focal adhesion complex (a cell–matrix actin-based anchoring junction type in other epithelia, for a review, see Cheng and Mruk, 2002) to constitute apical ES (a testis-specific cell–cell actin-based AJ type) are significant since integrins at the focal adhesion site at the level of cell–matrix are crucial to permit rapid cell-matrix restructuring, such as between fibroblasts/macrophages and extracellular matrix, to facilitate cell movement. Since the event of spermiation also requires extensive junction restructuring, it is not surprising that apical ES is utilizing one of the most efficient protein complexes at focal adhesion sites in other epithelia, namely integrin/laminin, to regulate ES dynamics.
Laminin. Laminin is a heterotrimer composed of three different chains, one each of the , ß and chain, to constitute a functional laminin protein, which is the putative binding partner for an integrin receptor in many epithelia (for a review, see Colognato and Yurchenco, 2000). As such, laminin is anticipated to be the binding partner for 6ß1- and 4ß1-integrin at the apical ES site. However, laminin is restricted to the basement membrane in most epithelia. Recently, a non-basement membrane form, laminin 3 (Mr 170 kDa), was found in the testis, which was found at the site consistent with its localization at the apical ES (Iivanainen et al., 1999; Koch et al., 1999). Subsequent studies using fluorescent microscopy with dual fluorescent probes against ß1-integrin and laminin 3 have localized both proteins to the same site at the apical ES in the seminiferous epithelium of the adult rat testis (Siu and Cheng, 2004). More important, these data were also verified by co-immunoprecipitation technique using seminiferous tubules without Leydig and myoid cell contamination, since an anti-ß1-integrin can pull out laminin 3 from the tubule lysates, whereas an anti-laminin 3 can also pull out ß1-integrin but not N-cadherin and nectin-3 (Siu and Cheng, 2004). Collectively, these data clearly illustrate that ß1-integrin can form a bona fide complex with laminin 3 (Siu and Cheng, 2004). Yet the other two chains, namely and ß, that are required to assemble a functional laminin receptor for 6ß1 integrin in the seminiferous epithelium, remain to be identified. In this context, it is of interest to note that besides the 6ß1-/4ß1-integrin/laminin (?)ß(?)3, which is found at the apical ES site, recent studies have shown that other signalling molecules that are usually restricted to the focal adhesion site in other epithelia are also present in the apical ES. This will be discussed in ‘Signalling proteins’ below.
Adaptors
Adaptors are proteins that recruit other signalling and structural proteins to the site of junctions, such as ES. They are also crucial to tether integral membrane proteins, such as cadherins, integrins and nectins, to the underlying cytoskeleton network, such as the actin-based microfilament. Yet few reports are found in the literature that have investigated the role of adaptors in ES dynamics. The known adaptors at the site of ES include -, ß- and -catenins, afadin, -actinin, cortactin, fimbrin, paxillin, vinculin, zonula occludens-1 (ZO-1), and several others (Table II) (for reviews, see Russell and Goh, 1988; Vogl et al., 2000; Cheng and Mruk, 2002).
Signalling proteins
Signalling proteins in the ES are mostly protein kinases, phosphorylating the downstream target proteins at either Tyr, Ser or Thr residues. The putative interacting partners of several signalling molecules found in the seminiferous epithelium of the testis are listed in Table II.
Cellular homologue of a viral protein in Rous Sarcoma virus (RSV) (c-Src) and Fyn
c-Src and Fyn belong to the Src family and have apparent Mr ranging between 52 and 62 kDa (for a review, see Thomas and Brugge, 1997). c-Src (Nishio et al., 1995; Wine and Chapin, 1999; Wang et al., 2000) and Fyn (Maekawa et al., 2002) have been localized to the ES in the seminiferous epithelium.
C-terminal Src kinase (Csk)
Csk is a 50 kDa protein that phosphorylates c-Src at the C-terminal Tyr residue (for a review, see Bjorge et al., 2000). Csk is localized to ES (Wine and Chapin, 1999), almost at the same site of Src, demonstrating a kinase-substrate relationship of these two kinases in regulating ES dynamics.
Focal adhesion kinase (FAK)
FAK is a 120 kDa focal adhesion-associated non-receptor PTK and the only other member of this family known to date is cell adhesion kinase ß (CAKß) (for a review, see Parsons, 2003). FAK has been localized to ES in studies using immunohistochemistry (Wine and Chapin, 1999; Siu et al., 2003). Interestingly, pFAK-Tyr397 is intensely localized to apical ES in stage VII through early stage VIII but not during spermiation in late stage VIII of the epithelial cycle, whereas most of the FAK is restricted to the basal compartment (Siu et al., 2003). It is noted that pFAK-Tyr397 co-localizes with vinculin to the same site in ES (Siu et al., 2003). pFAK is also an important PTK that mediates the ß1-integrin-induced signalling function, regulating ES dynamics in the rat testis (Siu et al., 2003).
Fes-related protein (Fer) kinase
Fer kinase is a 94 kDa cytoplasmic protein tyrosine kinase (PTK) (for a review, see Greer, 2002). The truncated form of Fer kinase is called FerT (51 kDa), which is restricted to pachytene spermatocytes (Fischman et al., 1990; Keshet et al., 1990). Fer kinase and FerT are non-receptor PTK in the testis (Fischman et al., 1990; Keshet et al., 1990; Chen et al., 2003). Fer kinase is a stage-specific protein, being at its highest levels in stages I–XIII (Chen et al., 2003); it largely associates with spermatocytes and elongating, but not elongated, spermatids in stages IX–XIV, and is restricted to round spermatids in stages I–VIII (Chen et al., 2003). Fer kinase is also associated with ES, apparently both the basal and apical ES site, as demonstrated by immunohistochemistry and immunoprecipitation technique (Chen et al., 2003), in contrast to FerT, which is restricted to pachytene spermatocytes (Keshet et al., 1990; Hazan et al., 1993). Interestingly, Fer kinase–/– mice are viable and fertile (Craig et al., 2001), seemingly suggesting that while Fer kinase is a crucial PTK in spermatogenesis, its function can be superseded by other kinases in the testis. As such, these negative findings by gene knockout experiments do not negate the significance of Fer kinase, they indeed reinforce the significance of PTK in AJ dynamics.
Integrin-linked kinase (ILK)
ILK is a 59 kDa serine/threonine kinase that binds to the cytoplasmic tail of ß1-integrin (Hannigan et al., 1996). ILK has been localized to apical ES, virtually to the same site of ß1-integrin and vinculin (Chapin et al., 2001; Mulholland et al., 2001; Beardsley and O'Donnell, 2003). In the testis, ILK and ß1-integrin form an interacting complex (Mulholland et al., 2001). Recent studies have shown that this is an important signalling molecule that mediates the integrin downstream signalling function, affecting ES dynamics in the testis (Mulholland et al., 2001).
Protein kinase A (PKA)
PKA has four regulatory (R) subunits (RIa, RIb, RIIa and RIIb) and three catalytic (C) subunits (, ß, and ), which have similar Mr ranging between 40 and 55 kDa (for reviews, see Chin et al., 2002; Robinson-White and Stratakis, 2002). PKA is found in the testis and different subunits have distinct distribution in the seminiferous epithelium (Pariset et al., 1989; Lonnerberg et al., 1992). For instance, the expression of RIb is highest in stages VIII–XI, whereas RIIa and RIIb are highest in stages VII–VIII (Lonnerberg et al., 1992). It is noteworthy that the relative abundance of different subunits changes during germ cell maturation, suggesting their differential role in spermatogenesis. For instance, RII subunits are predominantly expressed in late-stage germ cells, such as elongating spermatids (Pariset et al., 1989; Landmark et al., 1993). While the function of PKA in ES dynamics remains to be studied, it is a known crucial regulator of Sertoli cell TJ dynamics (Li et al., 2001). Likewise, PKG was recently shown to be a crucial regulator of Sertoli cell TJ dynamics (Lee and Cheng, 2003).
Remarks
While several reports are found in the literature confirming the presence of signalling molecules at the site of ES, mostly by immunohistochemistry as reviewed herein, how these proteins regulate ES integrity and restructuring during spermatid movement is largely unknown. Much of the information in this area of research relies on studies of the same proteins found in other epithelia many of which are confined to the cell-matrix focal adhesion sites, such as FAK and ILK. Nonetheless, these recent findings have opened new opportunities to study ES dynamics. For instance, inhibitors can now be used to assess whether a disruption of these kinases can perturb spermatid movement or spermiation. If it does, these inhibitors can become the basis of developing novel male contraceptives. Obviously, this is an area of research that deserves much attention in future studies in order to delineate the mechanism(s) that regulates ES dynamics in the testis. Table III also summarizes results of recent findings using gene knockout mice, illustrating the crucial function of many proteins found in the ES as well as in AJ sites in other epithelia.
Motor proteins
Three types of motor proteins, namely dynein, kinesin and myosin, are found in locomotive cells, such as fibroblasts and macrophages, facilitating cellular movement. In the testis, with the exception of kinesin (a microtubule-based motor protein), both dynein and myosin are found in the apical ES and these proteins apparently function as transporters to assist spermatid translocation across the seminiferous epithelium during spermatogenesis (Redenbach and Vogl, 1991; Redenbach et al., 1992; Beach and Vogl, 1999). In essence, motor proteins work in concert with other partner proteins, such as Kelch-like ECH Associating Protein 1 (Keap1), that they use microtubule found in Sertoli cells at the apical ES site as a track to facilitate the movement of elongating/elongated spermatids across the epithelium (Figure 3) (Table II) (Redenbach and Vogl, 1991; Beach and Vogl, 1999; Guttman et al., 2000; Velichkova et al., 2002).
|
Cytoplasmic dynein
Cytoplasmic dynein is a microtubule-associated motor protein complex with Mr
1000–2000 k Da (for review, see King, 2000). It is known to generate the minus-end-directed movement along the microtubules (for reviews, see King, 2000; Kamal and Goldstein, 2002). Cytoplasmic dynein associates with the cytoplasmic face of ER at the site of ES, co-localizing with actin (Collins and Vallee, 1989; Hall et al., 1992; Maeda et al., 1998; Miller et al., 1999; Guttman et al., 2000). It was postulated that cytoplasmic dynein was used as a vehicle and the microtubule as a track to assist spermatid movement across the seminiferous epithelium (Redenbach and Vogl, 1991; Redenbach et al., 1992; Beach and Vogl, 1999; Guttman et al., 2000) (Figure 3). However, the positive-end-directed microtubule-based motor protein, likely a kinesin family member, remains to be found at the ES site in the testis.
Myosin VIIa
Myosin VIIa, which hydrolyses ATP and can bind actin, is a 250 kDa motor protein in a large family with at least 15 different classes (for a review, see Sellers, 2000). Two forms of myosin VII, namely VIIa and VIIb, which lacks the coiled-coil domains, are known to date (Chen et al., 2001). Myosin VIIa is a motor protein in the testis (Hasson et al., 1995). It co-localizes with Keap1, a negative regulator of transcription factor NF-E2-Related Factor 2 (Nrf2) (Itoh et al., 1999), at the site of ES (Velichkova et al., 2002).
Proteins with specific functions
Espin
Espin is an ES-associated protein of Mr 110 kDa (Bartles et al., 1996). It has three actin-binding sites that link actin filaments to each other and to cell membranes at the site of ES (Bartles et al., 1996; Chen et al., 1999). It is co-localized with actin and vinculin to the site of basal and apical ES (Bartles et al., 1996; Chen et al., 1999; O'Donnell et al., 2000). It is noted that actin bundles start to appear at the time of espin accumulation (Chen et al., 1999) illustrating that it is a crucial actin bundling protein.
Gelsolin and scinderin
Gelsolin and scinderin, an apparent Mr of 91 and 80 kDa respectively, are actin-severing proteins belonging to the same family with at least six other members, such as advillin (for a review, see Kwiatkowski, 1999). Scinderin, also called adseverin, is the closest homologue of gelsolin (for review, see Kwiatkowski, 1999). Gelsolin co-localizes with actin to the same site in ES (Rousseaux-Prevost et al., 1997; Guttman et al., 2002). In addition, scinderin (Pelletier et al., 1999), and two other gelsolin regulators, phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2] and phosphoinositide-specific phospholipase C (PLC)-, are also found in the same site (Guttman et al., 2002), suggesting that gelsolin and its regulators are physically and intimately co-existing in ES. Indeed, treatment of ES-enriched fractions either with exogenous PLC or with a synthetic peptide corresponding to the PI(4,5)P2 binding region of gelsolin can lead to actin filament disassembly, releasing gelsolin (Guttman et al., 2002).
Phosphatidylinositol 4,5-bisphosphate
PI(4,5)P2 belongs to a group of phospholipids with at least seven known derivatives (for a review, see Yin and Janmey, 2003). PI(4,5)P2 co-localizes with actin, gelsolin, and PLC to the same site in mechanically isolated spermatids (Guttman et al., 2002). PI (4, 5) P2 is a crucial regulator of junction dynamics via its effects on intracellular calcium level and PKC.
PLC
PLC is member of a large protein family consisting of at least 11 members in mammals, such as ß1 to ß4, 1 and 2, 
1 to 4 and 
1. The molecular weights of PLC range from 85 to 150 kDa and PLC has an apparent Mr of 148 kDa (for review, see Rebecchi and Pentyala, 2000). It hydrolyses PI(4,5)P2 to two second messengers, namely inositol 1,4,5-trisphosphate (IP3), which releases calcium from the intracellular store, and diacylglycerol (DAG), a PKC activator (for review, see Rebecchi and Pentyala, 2000). Other studies have shown that PKC is a crucial regulator of setor; cell tight junction dynamics (Li et al 2001).
Remarks
As reviewed in sections ‘Motor proteins’ and ‘Proteins with specific functions’, it is obvious that ES is equipped with the necessary motor and regulatory proteins to induce spermatid movement in the epithelium. Yet it remains unclear how these proteins can interact together to facilitate spermatid movement (see ‘The role of ES in germ cell movement across the seminiferous epithelium’ below).
Cytoskeletal proteins
Cytoskeletal proteins, such as actin, tubulin and vimentin, are constituents of cytoskeletons (Table II). They are crucial not only to confer cell shape and cell rigidity, but also to facilitate cell movement by changes in polymerization/depolymerization and via the intricate interactions with different proteins, such as adaptors and signalling molecules, at the ES site. Recent studies using adenovirus to induce overexpression of -tubulin in Sertoli cells in vivo have shown that in addition to its structural role as a cytoskeleton, it is crucial for the attachment of the heads of elongating/elongated spermatids onto the epithelium at the site of apical ES (Fleming et al., 2003a,b). Additionally, overexpression of Sertoli cells with -tubulin apparently induces germ cell apoptosis, suggesting that -tubulin may be crucial to cell viability in the epithelium (Fleming et al., 2003b). It is not certain based on these data whether -tubulin overexpression recruits signalling molecules harmful to cell viability to the ES sites where spermatids attach onto the epithelium, or whether -tubulin per se can induce apoptosis via a yet-to-be defined mechanim(s), such as by disrupting the intracellular trafficking system. Nonetheless, this latest finding has illustrated that cytoskeletal proteins may have potential functional roles in the seminiferous epithelium other than their structural function. Figure 1 shows the three cytoskeletal elements in the testis and their interacting partners (see also Table II).
The role of ES in germ cell movement across the seminiferous epithelium
The events of junction restructuring that take place in the seminiferous epithelium during spermatogenesis have made the testis as one of the most dynamic organs in the mammalian body. As such, the testis is equipped with some of the most flexible cell–cell junctions to facilitate germ cell movement in the seminiferous epithelium (for reviews, see Russell, 1980; Cheng and Mruk, 2002) (Table I). Furthermore, desmosome-like junctions that are found between Sertoli cells and spermatogonia/spermatocytes are not detected between Sertoli cells and spermatids beyond step 9 (Russell, 1977a). Instead, they are equipped with ES (Russell, 1977b). In short, this seemingly suggests that the mechanism(s) that regulates the movement of preleptotene and leptotene spermatocytes across the blood-testis barrier that occurs at late stage VIII through early stage IX of the epithelial cycle in the rat (Russell, 1977c) is regulated differently from those involved for spermatid movement at the ES site. Indeed, recent studies have shown that ES is crucial for the spermatid movement across the epithelium. Such up-and-down movement of spermatids across the epithelium during the epithelial cycle is likely regulated by the intricate interactions of several motor proteins that work in concert with the underlying cytoskeletons at the ES (Miller et al., 1999; Guttman et al., 2000; Velichkova et al., 2002). A brief review is presented herein.
As described above and shown in Figure 2C, D, apical ES is composed of thin layers of hexagonally packed actin filaments and endoplasmic reticulum (ER) on the Sertoli cell side lying against the heads of elongating/elongated spermatids in the epithelium (Figure 3) (for reviews, see Russell, 1980; Vogl et al., 2000). At the cytoplasmic face of the ER, microtubules are also arranged parallel to the long axis of the Sertoli cell with the plus end pointing to the basal epithelium and the minus end directed to the apical side (Redenbach and Vogl, 1991) (Figure 3). It was postulated that the intrinsic polarity of microtubules at the cytoplasmic side of the ER provides a guide for the directed movement of spermatids across the epithelium from the basal to the adluminal compartment and vice versa during the epithelial cycle (Vogl, 1988; Redenbach and Vogl, 1991; Fleming et al., 2003a,b; Kann et al., 2003) (Figure 3). Yet microtubules alone are not sufficient for spermatid translocation unless motor proteins that act as vehicles provide the needed driving force, moving spermatids along the microtubules that act as a track. Indeed, cytoplasmic dynein, a minus end-directed motor protein (for a review, see Kamal and Goldstein, 2002), was recently identified in ES (Hall et al., 1992; Miller et al., 1999; Guttman et al., 2000). In addition, cytoplasmic dynein links to ER, which is part of the structural component of ES (Guttman et al., 2000). As such, forces generated by cytoplasmic dynein can assist spermatid translocation. At this point, the precise mechanism by which motor proteins generate the needed driving force is not known. Yet based on studies in migrating fibroblasts or macrophages, changes in the rigidity and fluidity of a migrating cell can be the result of polymerization/depolymerization of the actin network, coupling with the intricate interactions between focal adhesion complexes, GTPases, and ATPases. Thus it is tempting to postulate that the mechanical forces known to be generated by fibroblasts for cell movement are likely utilized by Sertoli cells to uplift spermatids, yet this possibility must be vigorously tested experimentally. No plus end-directed motor protein has been identified to date in ES, even though it is likely to be a kinesin family protein (for a review, see Kamal and Goldstein, 2002). Myosin VIIa, another actin filament-dependent motor protein, is also found in ES (Velichkova et al., 2002) (Figure 3). As yet, the functional role of myosin VIIa in spermatid movement is not known. While the event of spermatid movement across the seminiferous epithelium remains obscure, it is obvious that different motor proteins, which interact with the underlying cytoskeletons, play a crucial role. The significance of microtubule-based cell vertical mobility in the seminiferous epithelium utilizing the ES has recently been demonstrated in two recent studies in which rats were transfected with adenovirus carrying -tubulin inducing overexpression of -tubulin in Sertoli cells in vivo (Fleming et al., 2003a,b). For instance, overexpression of -tubulin in Sertoli cells in vivo led to retention of spermatids and residual bodies in the epithelium by disrupting the events of spermatid release and residual body elimination (Fleming et al., 2003b), as well as inducing germ cell apoptosis (Fleming et al., 2003b), possibly as a result of intracellular trafficking disruption. Needless to say, many open questions remain unanswered. For instance, unlike fibroblasts or macrophages, Sertoli and germ cells are not actively migrating cells per se; since the movement of developing germ cells must co-ordinate with their developmental status, what signal(s) is involved that dictates the timely movement of germ cells across the epithelium in a given stage of the epithelial cycle? Also, what is the mechanism that regulates the opening and closing of junctions to facilitate cell movement? Do germ cells elicit any signals to Sertoli cells requesting the needed uplifting driving force? To answer these questions, the molecular architecture of the motor proteins regulating the structural complex and the pertinent signalling and structural proteins at the site of ES must be delineated. Figure 3 illustrates a current concept of germ cell movement at the site of ES.
Regulation of ES dynamics
The precise mechanism(s) that regulates ES dynamics is not known. Yet recent studies have shown that ES can be regulated by several parameters via different molecules and pathways. (Figure 4).
|
Phosphorylation and dephosphorylation
The importance of phosphorylation in the regulation of ES was implicated by localization studies using antibodies specific to phosphor-Tyr, -Ser or -Thr residues, which yielded a pattern of localization, resembling the location of ES in the seminiferous epithelium (Wine and Chapin, 1999
; Chapin et al., 2001; Mulholland et al., 2001; Maekawa et al., 2002). These studies have shown that many of the ES constituent proteins are phosphorylated. In addition, several PTK, such as c-Src, Csk and FAK, and serine/threonine kinase, such as ILK, are also localized to the ES (Wine and Chapin, 1999; Mulholland et al., 2001; Siu et al., 2003). Equally important, the putative substrates of protein kinases, such as paxillin and p120ctn, are also found at the same site (Wine and Chapin, 1999). Collectively, these findings implicate the role of phosphorylation and dephosphorylation in the regulation of ES dynamics. For instance, the non-phosphorylated form of FAK was found to be restricted to the basal ES, whereas most of the pFAK-Tyr397 was localized at the apical ES (Siu et al., 2003), suggesting that there is a rapid intracellular trafficking system in the seminiferous epithelium to facilitate such intracellular protein compartmentalization. Future research must include the identification of downstream pathway(s) following phosphorylation of specific target proteins by kinases.
Calcium ions
The importance of calcium ions in maintaining cadherin-mediated adhesion function is well documented (for a review, see Steinberg and McNutt, 1999). Furthermore, calcium ions can activate gelsolin activity, changing it from an inactive state to an active state, inducing its inherent actin filament-severing ability (for a review, see Kwiatkowski, 1999). Also, hydrolysis of PI(4,5)P