|Home | About | Journals | Submit | Contact Us | Français|
Throughout spermatogenesis, a select pool of germ cells, the leptotene spermatocytes, must traverse the blood–testis barrier (BTB) to enter the adluminal compartment of the seminiferous epithelium. This event requires extensive restructuring of cell junctions, and it must also coincide with germ cell cycle progression in preparation for primary spermatocyte meiosis. Recent findings show that cell-cycle-associated kinases and phosphatases, including mitogen-activated protein kinases (MAPKs), participate in the pathways that also direct germ cell adhesion and movement. Our new biochemical model explains, in part, how two distinct cellular events, BTB restructuring and spermiation, are coordinated to maintain spermatogenesis and fertility. In this way, MAPKs would synchronize cell cycle progression in primary spermatocytes with junction remodeling and cell migration across the BTB.
In the rodent testis, spermatogenesis begins shortly after birth and takes place within the seminiferous tubule, the functional unit of the testis in which spermatozoa will eventually be produced from more primitive germ cells known as spermatogonia (Figure 1). Germ cell development is a complex, highly regulated and step-wise process, and it is divided into four distinct steps (i.e. mitosis, meiosis, spermiogenesis and spermiation) which collectively take ~54 days to complete in the rat. These events are also described as a cycle of cellular changes, referred to as stages of the seminiferous epithelial cycle, and they occur within a defined region of the epithelium. In the rat, there are fourteen stages, denoted by roman numerals I–XIV, and each stage is characterized by a unique arrangement of differentiating germ cells, namely spermatogonia, spermatocytes and spermatids. Spermatogonia, defined as undifferentiated germ cells, are composed of three types: A-single (also known as germ-line stem cells), A-paired and A-aligned, and they are found adjacent to the basement membrane. These germ cells enter mitosis, and some will eventually differentiate into type A, intermediate and type B spermatogonia [1,2]. However, only type B spermatogonia further differentiate into leptotene spermatocytes, the germ cells that cross the blood–testis barrier (BTB) at stages VIII–XI of the epithelial cycle to gain entry into the adluminal compartment of the seminiferous epithelium [3,4] (Figure 1). Once in the adluminal compartment, these germ cells develop into pachytene spermatocytes and enter meiosis I (MI), which is followed by meiosis II (MII) at stage XIV and the formation of spermatids (Figure 1). Spermatids, which generally lie closer to the tubule lumen, undergo extensive morphological changes (defined as spermiogenesis), including chromosomal condensation and formation of the acrosome, tail and residual body. At the completion of spermiogenesis, elongated spermatids line the tubule lumen, awaiting release from the seminiferous epithelium at spermiation.
Spermatogenesis is further complicated by the requirement for germ cells, as they differentiate, to traverse the seminiferous epithelium in an orderly manner. This process requires extensive interactions between Sertoli (somatic ‘nurse-like’ cells) and germ cells, as well as tight coordination of germ cell movement and differentiation. For example, germ cell entry into the adluminal compartment of the seminiferous epithelium is restricted to leptotene spermatocytes preceding transition into G2–MI, a cellular event that occurs during stages VIII–XI of the epithelial cycle. Thus even the slightest disturbance or misregulation of cell cycle progression can arrest spermatogenesis, leading to infertility.
Here, we discuss recent developments in the field of cell cycle dynamics with special emphasis on the role of cell cycle kinases and phosphatases in junction remodeling and cell migration as they relate to spermatogenesis. We begin by reviewing the current status of research on kinase and phosphatase families known to be important in cell cycle control, followed by a short perspective on how these proteins might also participate in cell migration and junction dynamics. Finally, we provide a biochemical model that describes how these events are coordinated and regulated in the seminiferous epithelium.
The movement of germ cells across the seminiferous epithelium is a tightly regulated process that requires precise timing of cell division events that rely on the involvement of distinct kinases and phosphatases. For instance, the mitogen-activated protein kinase (MAPK) signaling pathway regulates cell cycle events, as well as the transit of primary spermatocytes across the BTB, pointing to a role for MAPKs in synchronizing changes in BTB remodeling during germ cell division. In this section, we briefly review several kinase and phosphatase families that direct cell cycle control and germ cell movement.
Cyclins (cyclins A–I) are important regulators of the cell cycle, and each cell cycle transition is regulated by a different set of cyclins. The cyclin B–cyclin-dependent kinase 1 (CDK1) complex, for example, regulates entry into and proper progression through mitosis and meiosis [5,6]. Cyclins are regulatory subunits that have the ability to activate the catalytic subunit of CDKs. CDK1, a cell cycle regulator, must bind to a specific cyclin, assembling a cyclin–CDK complex that is maintained in an inactive state by phosphorylation of Thr14 and/or Tyr15 within the CDK1 ATP-binding loop. However, cyclin–CDK activation is required to drive cell cycle progression, and this event is triggered by cell division cycle 25 (CDC25)-mediated dephosphorylation of these two residues. CDC25 family members (CDC25A–C) are dual-specificity phosphatases that localize to both the nucleus and cytoplasm; they require specific activation by cyclin B–CDK1, thereby creating a positive-feedback loop . In addition, they are regulated by other mechanisms, including changes in cellular localization (in part by binding to 14-3-3 proteins, important regulators of cell survival, proliferation and apoptosis ) and protein–protein interactions. Among the cyclins expressed in mammalian germ cells, only cyclins A1 (encoded by Ccna1) and B1-3 (encoded by Ccnb1-3) are expressed during meiosis [9,10]. As expected, Ccna1−/− mice are sterile owing to a failure of cyclin B–CDK1 activation and subsequent meiotic arrest . Likewise, prolonging the expression of cyclin B3, which localizes specifically to leptotene and zygotene spermatocytes in the normal mouse testis , until the end of meiosis reduces sperm counts, increases germ cell apoptosis and disrupts spermatogenesis . Abnormal, round spermatids are also observed within the seminiferous tubules of these animals. Of note, fertility is unaltered in Cdc25C−/− mice , suggesting that other CDC25 proteins might compensate for the loss of CDC25C. Collectively, these findings illustrate that cyclins are crucial for normal spermatogenesis.
Of equal importance in cell cycle progression and cell division is the Polo-like kinase (PLK) family of Ser/Thr kinases, which comprises four proteins in humans (PLK1–4) . PLK1, the best studied member, phosphorylates cyclin B1 and CDC25C, resulting in their nuclear translocation [15,16]. The essential nature of PLK1 in mitosis is illustrated by the failure of Plk1 null embryos to develop beyond the eight-cell stage . In general, the highest PLK1 levels are detected in tissues containing rapidly proliferating cells (e.g. thymus and testis) , and PLK1 overexpression also correlates with tumorigenesis and carcinoma invasion . Interestingly, PLK1 interacts with and phosphorylates β-catenin , a cytoplasmic protein that functions in cell adhesion and Wnt-mediated transcriptional activation. PLK1 can also interact with and regulate 20S proteasome activity . Thus PLK1 is a dual-regulator of the cell cycle and proteolysis. By contrast, PLK2 and PLK3 can bind calcium and integrin binding 1 (CIB1). CIB1 activates p21-activated kinase 1 (PAK1, a focal adhesion protein implicated in cytoskeleton reorganization), resulting in decreased cell migration . CIB1 is also essential for fertility , and it might coordinate cell cycle progression, junction restructuring and cell migration during spermatogenesis.
In addition to PLKs, Aurora proteins (Aurora-A, -B and -C) comprise another small family of Ser/Thr kinases with important roles in cell division ranging from spindle assembly to cytokinesis, and all members, except for Aurora-C, are expressed ubiquitously . Aurora-A is essential for spindle assembly , and recently it was identified as the enzyme responsible for PLK1 phosphorylation [26,27]. By contrast, Aurora B is not required for spindle assembly; instead, Aurora B functions in spindle checkpoint activation . Considerably less is known about Aurora-C, a protein that is expressed predominantly in the testis , largely by meiotically dividing germ cells . What is known, however, is that Aurora-C functions during late spermatogenesis (i.e. spermiogenesis) and is crucial for male fertility . Further investigation is required to gain a better understanding of PLKs and Aurora kinases in cell cycle control during spermatogenesis.
There are several MAPK subfamilies in mammalian cells: the extracellular signal-regulated kinases (ERK1/2 and ERK5), c-Jun N-terminal kinases (JNKs) and p38s. Each subfamily forms a branch of MAPK signaling consisting of MAPK kinase kinases (MKKKs), MAPK kinases (MKKs) and MAPKs. Of the MAPK family, ERK1/2 are Ser/Thr kinases with broad cellular localization patterns. ERK1/2 are expressed by both spermatogonia and spermatocytes, and they have important roles in mitosis and meiosis in the testis. For example, treatment of pachytene spermatocytes with okadaic acid, a Ser/Thr phosphatase inhibitor known to induce premature G2–MI transition and chromatin condensation, resulted in ERK1/2 activation , illustrating that ERK1/2 has a role in G2–MI. MAPK/ERK kinases (MEK)1/2 lie upstream of ERK1/2; MEK1 is expressed in spermatogonia and leptotene/zygotene spermatocytes, but MEK2 is expressed by all germ cells throughout spermatogenesis . Surprisingly, Mek2 null mice are fertile , suggesting that MEK1 might be able to compensate for the loss of MEK2 or that MEK2 might have additional functions outside of germ cell differentiation. Likewise, p90RSK2, an ERK1/2 effector, is abundantly expressed in primary spermatocytes, and its activation is required for chromatin condensation . Given the requirement for chromatin condensation in meiosis, these studies illustrate the importance of ERK1/2 and p90RSK2 in meiosis. In mouse spermatocytes, ERK1/2–p90RSK2 phosphorylation and activation triggers the activation of never in mitosis gene a (NIMA)-related kinase 2 (Nek2) , a Ser/Thr kinase that can bind and phosphorylate β-catenin . Although an extracellular signal is believed to be responsible for meiotic MAPK activation in spermatocytes , this signal has not yet been defined. Interestingly, this role might be fulfilled by paracrine factors: both epidermal growth factor (EGF) and fibroblast growth factor (FGF) have been implicated in p90RSK2 phosphorylation in several cell lines (e.g. 293T, COS7 and Ba/F3) [36,37]. Similar factors are likely to activate MAPK signaling during the transit of leptotene spermatocytes across the BTB because these germ cells are known to secrete several cytokines (e.g. transforming growth factor-β3 [TGF-β3] and tumor necrosis factor α [TNFα]) into the BTB microenvironment to regulate germ cell movement [38,39]. As such, cytokines might activate ERK1/2–p90RSK2 to facilitate the entry of primary spermatocytes into metaphase I, and this might also promote migration of leptotene spermatocytes across the BTB.
As discussed, kinases and phosphatases maintain accuracy of germ cell division throughout spermatogenesis. At the same time, germ cells must also traverse the seminiferous epithelium, a process that involves cycles of junction disassembly and assembly. Therefore, tight coordination of these cellular events is necessary for fertility. At this point, there is no better example of the significance of crosstalk among cell cycle progression, junction restructuring and cell migration processes than tumorigenesis, which results from misregulation of these cellular events. Over the past several years, many reports have linked cell-cycle-associated kinases and phosphatases (e.g. CDKs and ERK1/2) with cell migration and junction dynamics during tumorigenesis.
Recently, much attention has focused on Aurora-A activation by human enhancer of filamentation-1 (HEF1) , a well-studied scaffolding protein that belongs to the Crk-associated substrate (CAS) family, which participates in actin cytoskeleton remodeling, cell migration, cell cycle progression, apoptosis and tumorigenesis . HEF1 (also known as CAS-L) also associates with CAS/HEF1-associated signal transducer (Chat, a protein implicated in the regulation of cell adhesion)  and has a defined role in integrin-mediated focal adhesion and cell migration . These findings are interesting because they clearly link events of cell division with cell adhesion and movement. Similarly, PAK1 is thought to regulate centrosome maturation via Aurora-A phosphorylation . Other focal adhesion proteins, such as zyxin  and focal adhesion kinase (FAK) [46–48], which both physically interact with HEF1 [49,50], and paxillin , have been implicated in mitotic control. Moreover, integrins, which function immediately upstream of these proteins, can regulate cyclin B–CDC2 activity to facilitate cell movement . Collectively, these results point to the existence of a unique mechanism by which cell division, adhesion and migration are synchronized. The possibility that Aurora proteins, in particular Aurora C, which is highly expressed in the testis, might also function during the timely movement of germ cells across the seminiferous epithelium would provide new insights into spermatogenesis.
In addition to Aurora kinases, which have dual roles in cell division and adhesion, CDK1 and CDK2 also phosphorylate the membrane-associated guanylate kinase homologs (MAGUK) scaffolding protein discs, large homolog 1 (DLG1) , which has been long known to regulate cell polarity, proliferation and junction dynamics . In addition to the junctional complex, DLG1 also concentrates at the midbody in mammalian cells , suggesting that DLG1 probably transits between two distinct intracellular compartments to mediate cell cycle and junction events. However, the mechanism underlying this shuffling between compartments remains undefined. Similarly, human enhancer of invasion clone 10 (HEI10), a negative regulator of cell motility, can interact functionally with cyclin B1 and is an in vitro substrate of cyclin B1–CDC2 . Recent studies have provided another example of a protein with dual roles in cell division and adhesion: CD2-associated protein (CD2AP) has a role in a late phase of cytokinesis because its knockdown by small interfering RNA (siRNA) transfection resulted in incomplete separation of daughter cells ; as a regulator of the actin cytoskeleton, CD2AP also forms a functional complex with several cell junction proteins, namely E-cadherin, p120 catenin, zonula occludens-1 (ZO-1) and nephrin in Madin-Darby canine kidney (MDCK) cells and glomerular lysates .
In addition, MAPKs also have functions outside of cell cycle control. For instance, MEK1/2 can bind IQ-motif-containing GTPase-activating protein 1 (IQGAP1) to regulate actin polymerization . Restructuring of Sertoli cell junctions, which is crucial for the movement of primary spermatocytes across the BTB, is also regulated by ERK1/2 and p38 MAPK. TGF-β3 activates p38 MAPK , thereby affecting the Sertoli cell tight junction barrier in vitro and in vivo. The involvement of p38 MAPK in BTB dynamics was subsequently confirmed through the use of a rat model that utilized the environmental toxicant cadmium: inhibition of intrinsic p38 MAPK activity by a specific p38 inhibitor can block cadmium-induced BTB disruption and delay germ cell loss from the epithelium . This finding reinforced p38 MAPK’s role in cell junction dynamics. TGF-β3 can also activate ERK without affecting endogenous p38 MAPK levels, but surprisingly this activity fails to compromise BTB function in vivo; instead, Sertoli–germ cell junction remodeling and germ cell loss from the seminiferous epithelium occurs . Furthermore, Sertoli–Sertoli (i.e. BTB) and Sertoli–germ cell junction restructuring requires ERK and p38 MAPK activation that results from TGF-β3–TβRI (a Ser/Thr kinase receptor) interacting with TGF-β-activated kinase 1 (TAK1)-binding protein 1 (TAB1, an adaptor protein that activates TAK1) and CD2AP. By contrast, the interaction of TGF-β3–TβRI with CD2AP, but not TAB1, activated only ERK , illustrating that different mechanisms are in place to disassemble Sertoli–Sertoli and Sertoli–germ cell junctions.
Several recent findings support a role for TGF-β beyond junction remodeling. For instance, both cyclin B and CDK1 are known targets of TGF-β signaling [64,65], and TGF-β can promote cell cycle arrest by effectively inhibiting CDK2 . TGF-β1 can also upregulate the expression of p21waf1/cip1 , which binds and inhibits CDK2–cyclin A/E complexes . A role for TGF-β in the G2–M transition is supported in part by a recent study that showed that Smad3-mediated loss of TGF-β signaling can abrogate G2–M progression in bone marrow stromal cells . In another in vitro study, TGF-β increased the number of germ cells present at metaphase II . Taken collectively, these results illustrate that TGF-β elicits different regulatory cues depending on the cellular context. Most important, these findings demonstrate the need for strict control of TGF-β activity because misregulation can affect cell division, perturb cell adhesion and enhance cell movement, events that occur throughout oncogenic transformation and invasion.
At this point of our discussion, we ask how entry into meiosis and movement of primary spermatocytes across the BTB are coordinated. Because leptotene spermatocytes must cross the BTB before the G2–MI transition, we speculate that there must be a mechanism of exogenous control at the BTB, in addition to surveillance at the level of the nucleus, to allow coordination of these two cellular events. Cytokines such as TGF-β will certainly have an important role, but additional experiments are needed to assess, for instance, whether the loss of specific cytokines can prevent BTB restructuring and prohibit spermatocyte entry into the adluminal compartment via effects on the cell cycle. This hypothesis is based on the finding that TGF-β3 perturbs Sertoli–Sertoli cell junctions at the BTB . In vitro experiments should provide additional insight because the co-culture of Sertoli and germ cells can facilitate meiotic progression [71,72]. It is worth noting that the inhibition of MAPK activity in Sertoli, but not germ, cells barred the onset of metaphase I in primary spermatocytes , demonstrating the essential role for Sertoli–germ cell interactions in meiosis. It is also possible that androgens participate in cell cycle progression, junction remodeling and cell migration. Two recent studies linked testosterone to BTB dynamics [74,75]; likewise, cyclins and CDKs are implicated in the regulation of androgen receptor (AR) transcriptional activity . Specifically, AR transcriptional activity was repressed by cyclin D3–CDK11p58, which forms a functional complex with AR , thereby suggesting that cyclins and CDKs regulate BTB integrity indirectly. Nonetheless, synchronization of the cell cycle, junction restructuring and cell movement is likely to be accomplished by crosstalk between cytokines and/or androgens, cell cycle and cell junction proteins.
Adhesion between Sertoli cells and elongating/elongated spermatids is maintained by the apical ectoplasmic specialization (apical ES), a hybrid-like, testis-specific type of anchoring junction . Aside from the apical ES, there is no other anchoring device present between Sertoli cells and elongating/elongated spermatids, and adhesion at the apical ES is maintained by three multi-protein complexes: α6β1 integrin–laminin α3β3γ3, nectin–afadin and cad-herin–catenin [78,79]. In addition to maintaining cell adhesion, the apical ES also facilitates spermatid movement, a process that involves the turnover of adhesion proteins, resulting in junction restructuring and germ cell movement . Thus it is not surprising that several proteins normally found at focal contacts and known to function in cell migration are also found at the apical ES [78,79]. Other proteins are known to have crucial roles in spermiation. For example, a surge in matrix metallopro-tease-2 (MMP-2) at the apical ES at stage VIII , possibly activated by TNFα , is thought to cleave laminin into biologically-active fragments to facilitate spermiation  (Figure 2). These laminin fragments can also regulate basal ES dynamics and BTB remodeling, thereby providing a unique mechanism to explain how spermiation at stage VIII is coordinated with the transit of primary spermatocytes across the BTB at stages VIII–XI  (Figure 2). Such a regulatory mechanism is also supported by findings from other systems that illustrated that laminin chains are putative MMP substrates . Biologically active laminin fragments can also regulate distinct cellular processes ranging from apoptosis to cell migration . Indeed, when added to Sertoli cells in vitro with negligible elongating/elongated spermatid contamination (i.e. in the absence of the apical ES), recombinant laminin β3 (domain I) and/or γ3 (domain IV) fragments perturb tight junction barrier function in a dose-dependent manner . This effect is mediated by a loss of integral membrane proteins at the BTB (e.g. occludin and ZO-1) and at the hemidesmosome (e.g. β1 integrin). At this point, it should also be noted that α6β1 integrin and laminin α3β3γ3 do not localize to the BTB. Consistent with these observations, transient overexpression of laminin γ3 (domain IV) in Sertoli cells with a functional tight junction barrier and hemidesmosomes is associated with a loss of occludin and β1 integrin but with an increase in ERK activity , events that disrupt tight junction function. These findings essentially reveal that laminin fragments released from the apical ES at spermiation regulate BTB dynamics directly or indirectly via hemidesmosomes (Figure 2). They also illustrate the existence of a functional loop among the apical ES, BTB and hemidesmosome in the seminiferous epithelium. To confirm these results, β1 integrin was silenced in Sertoli cells. RNA interference (RNAi)-mediated loss of β1 integrin from hemidesmosomes not only altered occludin and N-cadherin localization at the BTB, leading to junction disassembly, but also resulted in increased occludin endocytosis . In summary, the seminiferous epithelium coordinates spermiation with BTB restructuring via a functional loop that connects apical ES, BTB and hemidesmosome function.
It is an exciting time for research aimed at understanding how the BTB is remodeled during the transit of leptotene spermatocytes throughout spermatogenesis. As discussed, cell cycle events, under the control of kinases and phosphatases, are likely to participate directly or indirectly in the movement of leptotene spermatocytes across the BTB. Cytokines and testosterone also have important roles in this process. Nevertheless, in the future it will be interesting to identify additional proteins and signaling cascades that trigger the BTB to ‘open’ transiently, thereby allowing the entry of differentiating leptotene spermatocytes into the adluminal compartment for further development. We have also described a unique biochemical pathway, involving laminin cleavage, to explain in part how sperm release is coordinated with BTB restructuring to maintain spermatogenesis and fertility. Although many questions remain unanswered (Box 1), future studies should provide important insight into these and other related questions in coming years.
Studies from the authors’ laboratory were supported by grants from the National Institutes of Health (NICHD, R01 HD056034; R03 HD051512; U54 HD029990 Project 5).