| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
REVIEW |
Monash Institute of Medical Research, Monash University, 27-31 Wright Street, Clayton, Victoria 3168, Australia
Correspondence should be addressed to K L Loveland; Email: kate.loveland{at}med.monash.edu.au
 |
Abstract |
|---|
 Top Abstract The canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
|
The canonical TGF-ß superfamily signalling pathway |
|---|
|
TopAbstractThe canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
subunit linked with an inhibin ß subunit) and activins (formed as dimers of activin ß subunits; these subunits are also termed inhibin ß subunits), the bone morphogenetic proteins (BMPs), Müllerian inhibiting substance (MIS; also known as anti-Müllerian hormone (AMH)), the growth and differentiation factors (GDFs) and the distantly related glial cell line-derived neurotropic factor (GDNF; this signals through an atypical receptor complex). A review of this family and its mechanisms of function are presented by Lin et al.(2006).
The prototypical signalling receptors for this family are heterotetramers comprised of two type I and type II receptors, each with Ser/Thr kinase activity. The type II receptors are constitutively active, while the type I receptors are most commonly recruited and activated by the ligand-bound form of the type II receptor. Each ligand uses a distinct receptor subset, but many ligands share receptor subunits, so that competition amongst the superfamily for signalling molecules occurs at the level of receptor binding (summarised in Table 1
). Three type I receptors (ALK2, ALK3 and ALK6) and three type II receptors (BMPRII, ActRIIA and ActRIIB) are known to bind BMPs and signal, while activin uses ALK2, ALK4, ActRIIA and ActRIIB (reviewed in Zhao & Garbers 2002, Shi & Massague 2003). As a further example of the complexity of these interactions, activin B (a dimer of activin ß B subunits), activin AB (a dimer of activin ß A and B subunits) and nodal preferentially utilise the ALK7 type I receptor, with nodal binding being particularly enhanced in the presence of the nodal co-receptor, cripto (Reismann et al. 2001). The TGF-ß ligands use a distinct set of type I (ALK1 and ALK5) and type II (TßRII) receptors in addition to a third receptor, termed type III or betaglycan, which appears required for transduction of TGF-ß2 signals, but not for those of TGF-ß1 or TGF-ß3. Thus, the synthesis or lack of betaglycan provides another level for regulating the response of a cell to a set of complex incoming signals, as when several TGF-ß ligands are present simultaneously.
|
The Smad pathway is a well-characterised intracellular response to TGF-ß ligand binding. The R-Smads, Smad1, Smad5 and Smad8 are signalling molecules for BMPs, while Smad2 and Smad3 propagate TGF-ß and activin signals (reviewed in Shi & Massagué 2003). Smad4 is recruited by all activated R-Smads, and it is this heteromeric complex, which translocates into the nucleus to effect changes in gene expression (Fig. 1). Smad6 and Smad7 are each inhibitors of this pathway, with several modes of action, including prevention of signalling complex formation and transcriptional regulation of elements of the signalling pathway.
|
|
TGF-ß superfamily signalling in the embryo and developing testis |
|---|
|
TopAbstractThe canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
). They arise within the proximal epiblast of the mouse embryo in response to extra-embryonic signals that occur approximately 2 days earlier (5.25–5.5 dpc), and as discussed below, have a clear dependence on BMP signalling (Lawson et al. 1999, de Sousa Lopes 2004, Drummond 2005).
|
Analysis of multiple lines of Bmp heterozygous animals has illuminated their functional interrelationships. Mice heterozygous for both Bmp2 and 4 (Bmp2+/–, Bmp4+/–) have reduced PGC numbers compared with either single heterozygote, implying that these two proteins have additive effects (Ying & Zhao 2001). However, BMP4 and Bmp8b double heterozygotes (Bmp4+/–, Bmp8b+/–) have the same PGC number as Bmp4 single heterozygotes, while the number of PGCs in Bmp2 and Bmp8b double heterozygote animals (Bmp2+/–, Bmp8b+/–) is similar to their respective single heterozygotes. This illustration that Bmp8b does not have additive effects with either Bmp2 or Bmp4 indicates that, while DPP class members signal through receptor complexes distinct from those used by 60A class members during PGC formation (Ying et al. 2000, Ying & Zhao 2001), they influence the same fundamental processes (Fig. 2
)
During PGC induction in epiblast cells, it is unclear, which specific receptor complexes transduce BMP2, BMP4 and BMP8b signals. A comprehensive study by de Sousa Lopes et al.(2004) aimed to elucidate the key surface receptors associated with this process. They showed that BMP4 produced in the extra-embryonic ectoderm signals through ALK2 in the visceral endoderm to induce PGC formation from the epiblast. Interestingly, ALK2 deficient embryos are devoid of PGCs, and their numbers are also reduced in heterozygous animals. This phenotype mimics that seen in the aforementioned Bmp4 knockout animal, which consequently, can be rescued by the constitutive expression of active ALK2 in the visceral endoderm (de Sousa Lopes 2004). Targeted disruption of the ALK4 gene in mouse embryos leads to a disorganised epiblast and extra-embryonic ectoderm, followed by aberrant development of the egg cylinder prior to gastrulation (Gu et al. 1998).
The Smads are also vital for embryonic development. Smad1 is localised to the epiblast and visceral endoderm during gastrulation but is not expressed immediately after, at 7.5 dpc in the wild-type mouse, suggestive of a role in PGC specification rather than proliferation and maintenance (Hayashi et al. 2002). Smad1 knockout mice die between 10.5 and 11.5 dpc (Tremblay et al. 2001). These mice produce only a few PGCs, if any supporting the concept that Smad1 is crucial for PGC formation (Tremblay et al. 2001, Hayashi et al. 2002).
Smad5 knockout mice also die in utero. Their embryonic lethal phenotype is seen between 9.5 and 11.5 dpc (Chang et al. 1999) and similar to Smad1 knockouts, they exhibit either greatly reduced numbers of PGCs or none at all. Smad5 may be important for determining, which cells of the epiblast become PGCs, as it is detected in the epiblast of normal mice at 6.5 dpc (Chang et al. 1999, Chang & Matzuk 2001) and is expressed by all three germ layers 1 day later in embryonic development. Smad1 and Smad5 are expressed in the epiblast at the onset of gastrulation, but Smad8 is detected only in the allantois (Chang et al. 1999, Yang et al. 1999, Chang & Matzuk 2001, Tremblay et al. 2001).
During fetal mouse development, the Smad2 gene appears to be ubiquitously expressed (Weinstein et al. 1998, Heyer et al. 1999). Mice with a targeted disruption of this gene die between 7.5 and 8.5 dpc, yet most Smad2 knockouts develop PGCs (Tremblay et al. 2001), suggesting that upstream signals that activate Smad2, such as activins, TGF-ßs or nodal, may not be vital for PGC development.
As expected, for a protein essential to such a multifaceted signalling pathway, expression of Smad4 is also ubiquitous. Smad4 knockouts display defects in development of the visceral endoderm and are embryonic lethal between 6.5 and 8.5 dpc (Sirard et al. 1998, Yang et al. 1998). The phenotype of Smad4 knockouts is similar to most BMP knockouts (Sirard et al. 1998), highlighting the crucial role Smad4 plays in propagating signals from the numerous interacting TGF-ß pathways.
TGF-ßs and activins in proliferation and apoptosis
Mouse PGCs move passively to the developing genital ridge between 8.5 and 10.5 dpc, and then migrate actively during 10.5–11.5 dpc. PGCs are proliferating during this period and continue to do so when they arrive at the sexually indifferent gonad. TGF-ßs and activins have been implicated in the migration, survival and proliferation of cells during this period of development. Among the TGF-ß isoforms identified, three are expressed in mammals (Cupp et al. 1999). In a very general sense, the primary actions of the TGF-ßs are to enhance formation of the extracellular matrix and inhibit proliferation of most cells (Lawrence 1996), inhibiting progression into late G1 of the cell cycle. In the fetal testis, TGF-ßs have been shown to exhibit multiple developmentally regulated effects.
TGF-ß1 and TGF-ß2 protein expression appeared limited to Sertoli cells of the 14.5 dpc rat testis, but was undetectable in gonocytes, whilst TGF-ß3 was found in cells surrounding the seminiferous cords at the interface between the testis and mesonephros (Cupp et al. 1999). In the rat, the actions of TGF-ß during embryonic testis development at 14 dpc are considered to be mainly on gonocytes, and by 16.5 dpc on both Leydig cells and gonocytes (Olaso et al. 1998). Both TGF-ß1 and TGF-ß2 have been shown to cause apoptosis of rat gonocytes at 14 dpc (Olaso et al. 1998). TGF-ß1 also reduced the size of rat testis organ cultures at both 13 and 14 dpc, probably by regulating first germ cell and then somatic cell numbers (Cupp et al. 1999).
Additional studies illustrate the changing responsiveness to TGF-ß that is a feature of fetal testis development. Experiments using single cell suspensions produced from PGC-containing fragments of 8.5 dpc mouse embryos showed that TGF-ß1 decreases the number of PGCs (Godin & Wylie 1991), but has no effect on PGC number at 10.5 dpc (de Felici et al. 1992). Another study revealed that TGF-ß1 and activin inhibit proliferation of both isolated pre-migratory and post-migratory mouse PGCs in vitro, and these cells contain mRNAs encoding appropriate receptor subunits (Richards et al. 1999). Chemotropic substances released by the genital ridges at around 10.5 dpc can attract PGCs; in particular TGF-ß1 mimics the chemotropic effect of genital ridges in culture (Godin & Wylie 1991). In this experiment, TGF-ß1 inhibited proliferation of PGCs from 8.5 dpc mouse embryos cultured on embryonic fibroblast feeder layers. While these experimental systems have unveiled roles in proliferation, apoptosis and migration for this signalling pathway, the in vivo consequences of multiple inputs from this signalling pathway will arise from a more complex set of signals that have been reproduced in these experiments.
Between 11.5 and 12.5 dpc, the PGCs become encased in Sertoli cells and surrounded by peritubular myoid cells and are thus sequestered inside testicular cords. Signals from the Sertoli cells direct PGCs into the male lineage, and they are now termed gonocytes, the precursors of the spermatogonial stem cells. At present, it is understood that TGF-ß signalling directly influences the survival of these cells. When 13.5 dpc fetal rat gonads were cultured as explants with TGF-ß1, a concentration-dependant decrease in the number of gonocytes was observed after 2 days in culture (Olaso et al. 1998). This phenomenon was also seen using 3 day post partum (dpp) testis explants (when the gonocytes are re-entering the cell cycle) but not with 17.5 dpc explants (when gonocytes are quiescent), indicating the influence of TGF-ß1 is reduced during the period of germ cell quiescence. Bromodeoxyuridine labelling indices of both gonocytes and Sertoli cells at 13.5 dpc did not vary with either TGF-ß1 or TGF-ß2 addition. However, a marked increase in the number of TUNEL-positive gonocytes was recorded when these factors were added, indicating an increase in DNA fragmentation and apoptosis in these cells rather than a lack of proliferation (Olaso et al. 1998). Rat gonocytes begin to arrest in G1 of the mitotic cell cycle around 13.5 dpc and are fully arrested by 16.5 dpc. They remain quiescent until shortly after birth, as in the mouse, when they begin the process of spermatogenesis, which will ultimately give rise to mature spermatozoa in the adult (Yao et al. 2003).
Müllerian inhibiting substance
Testis fate is determined by the expression of Sry, a gene present on the Y chromosome, which initiates the differentiation of Sertoli cells and their structural organisation into testicular cords (Koopman et al. 1990, Yao et al. 2003). This is the point in the mouse between 10.5 and 11.5 dpc where the male and female pathways of germ cell differentiation diverge; in the absence of the Sry gene, ovarian fate proceeds and the germ cells spontaneously enter meiosis by 13.5 dpc in females (Yao et al. 2003). MIS (also known as AMH) is secreted by pre-Sertoli cells of the embryonic XY gonad and, as its name suggests, induces the regression of the Mü llerian duct of the female reproductive tract (Donahoe et al. 1987, Ross et al. 2003). Male mice lacking the MIS gene are sterile due to blockage of the efferent ducts; retention of female ducts in mature males (pseudo-hermaphroditism) blocks sperm efflux during ejaculation (Behringer et al. 1994). MIS exerts its actions on somatic cells of the fetal testis via ALK2 and ALK6 receptor complexes, which had been previously shown to function as MIS type I receptors (Clarke et al. 2001). Mice lacking the genes encoding either ALK2 or ALK6 have been generated. ALK2 knockouts die before 9.5 dpc, well before Müllerian duct regression takes place, whilst ALK6 knockouts survive to adulthood and do not exhibit any defects in Müllerian duct regression, suggesting this receptor may not be essential for this event (Clarke et al. 2001). AMHRII (also termed MISR-II), the gene encoding the MIS type II receptor is expressed in fetal Sertoli cells and mesenchymal cells of the Müllerian ducts, fetal ovaries and testes and in granulosa cells of the post-natal ovary (di Clemente et al. 1994, Lee et al. 1999). Intercrossing female mice with a floxed ALK3 gene and heterozygote males carrying the Cre recombinase under regulation of the AMHRII promoter resulted in the conditional knockout of ALK3 in the developing Müllerian ducts. The resulting males developed as pseudohermaphrodites, containing uteri and oviducts (Jamin et al. 2002) a similar phenotype to what has been observed with the MIS knockout males. This finding suggests that ALK3 inactivation was enough to block Müllerian duct regression and that this receptor subunit can physically interact with the MISR-II to transduce MIS signals (Jamin et al. 2002). Thus, MIS signalling can be effected through the same type I receptors used by the BMPs.
The enzyme prostaglandin D2 (PGD2) is more highly expressed in the testis relative to the ovary at 12.5 dpc (Adams and McLaren 2002). This signalling molecule can partially masculinise female embryonic gonads when added in organ cultures, as indicated by induction of MIS expression in female somatic cells gonads at the age of sex determination (Adams & McLaren 2002). Therefore, the local synthesis of PGD2 by both somatic and germ cells of the testis may provide a key feedback loop ensuring that male gonadal differentiation follows.
Gene expression profiles in the embryonic gonad
The migratory phase of mouse germ cell development lasts for approximately 4 days (8.5–12.5 dpc; Buehr 1997). At the end of this period, PGCs within the gonad undergo sex determination upon exposure to gender-specific signals, continue to proliferate, and then become quiescent at around 14.5 dpc. During this dynamic period of cellular progression, PGC numbers increase from less than 100 at 8.5 dpc to approximately 25 000 at 13.5 dpc, the result of numerous division cycles (Tam & Snow 1981, Zhao & Garbers 2002). Concomitant with this growing PGC population in the testis are changes in the expression of many genes, including those related to TGF-ß superfamily signalling. The cues that govern these steps in PGC development are likely to result from a complex interaction between several TGF-ß superfamily ligands and inputs from other signalling pathways.
Changes in the level and sites of expression of TGF-ß signalling-related genes during testis development have been documented using in situ hybridisation and microarray analyses. A large body of data relating to the mouse testis and to specific testis cell types has been collected using Affymetrix arrays (see Small et al. 2005, gene expression omnibus (GEO) data repository, NCBI, available at: www.ncbi.gov/geo/). This information can be interrogated to gain insights concerning the key regulatory changes that occur in the TGF-ß signalling pathway as the testis develops. For example, the ActRIIB mRNA level does not change remarkably between the time of sex determination (11.5 dpc) and germ cell quiescence (16.5 dpc) in either the male or female gonad (Fig. 3). In contrast, the activin ßA subunit mRNA shows a continuous male-specific elevation up to the day of birth. Similarly, the inhibin subunit shows a profile of male-specific increase up to birth. Note that, while the expression profiles for these ligands follow a similar pattern, the inhibin numerical data indicates a tenfold higher signal than that recorded for activin ßA subunit mRNA. These data, while requiring verification using independent means (see Loveland et al. 2006), provide a clear indication that the level of ligand, not receptor subunit, is likely to be regulated during these developmental processes. The profile for Smad signalling does not show distinct sex differences during fetal life. However, in both embryonic testis and ovary Smad4, the common Smad, appears to be more highly expressed in comparison to the other Smads, based on the data of Small et al.(2005).
|
|
TGF-ß superfamily signalling in post-natal spermatogenesis |
|---|
|
TopAbstractThe canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
The Sertoli cells are specialised ‘nurse cells’ that provide the nutritional and architectural support required for adult germ cell development. They proliferate continuously until around 14 dpp, when they undergo a significant morphological and functional transformation to their final, differentiated phenotype. The inter Sertoli tight junctions, which initially form between 7 and 14 dpp, separate the meiotic and post-meiotic germ cells from factors present in the lymph and blood circulation, and secretions from the apical surface of this seminiferous epithelium create a specialised environment for their development and a lumen for passage out of the testis by spermatozoa.
Meiotic metaphase begins with each spermatocyte producing two diploid secondary spermatocytes, which rapidly undergo a second meiotic division to yield four haploid spermatids, clonal siblings, which remain linked by their persistent cytoplasmic bridges. The round spermatids first appear at 20 dpp; they undergo further differentiation and morphological changes to first form mature sperm by 35 dpp. In the adult mouse testes, the process is continuously repeated, so that each type A spermatogonium, located at the basement membrane, undergoes six mitotic divisions, two meiotic divisions and more than 2 weeks of differentiation to become spermatozoa that are released into the lumen at the centre of the seminiferous tubule (reviewed in McCarrey 1993, de Rooij & Russell 2000, Kerr et al. 2005).
Initiation of spermatogenesis and the first wave
In vitro studies using both organ culture and dispersed cell preparations have furthered our understanding of TGF-ß superfamily signalling in spermatogenesis, particularly at the time that undifferentiated spermatogonia first commit to differentiation at the onset of the first wave of spermatogenesis.
Activins
Activins and inhibins have long been identified as hormonal regulators of testicular development through their impact on pituitary secretion of follicle-stimulating hormone (FSH; reviewed in Cook et al. 2004, de Kretser et al. 2004). Their potential to act in a paracrine manner to influence spermatogonial proliferation was first identified over 15 years ago. In studies with cells obtained from adult animals, inhibin was reported to decrease the level of DNA synthesis in spermatogonia of the Chinese hamster, (Mather et al. 1990), while activin elevated proliferation rates of rat spermatogonia that were placed in co-culture with Sertoli cells purified from a 20 dpp animal (van Dissel-Emiliani et al. 1989). While these studies were effected without the benefit of the current suite of molecular probes that permit a refined assessment of somatic and germ cell maturation states, they firmly demonstrated the potential for inhibin and activin levels, and TGF-ß superfamily signalling pathways, to influence male germ cell development.
The addition of activin to fragment cultures of 3 dpp rat testes led to a significant increase in the number of gonocytes and a decrease in the number of Sertoli cells after 3 days in culture, the timeframe, which represents the very onset of spermatogenesis. The addition of follistatin (a soluble inhibitor of activin) in the presence of FSH (which acts on receptors present only on Sertoli cells) resulted in an increase in the number of spermatogonia, with no significant effect on gonocyte numbers compared to unsupplemented media (Meehan et al. 2000). These data suggest that a reduction in the level of activin signalling may be required for gonocytes to differentiate into spermatogonia. This is further indicated by the fact that rodent gonocytes themselves produce activin ßA subunit mRNA in fetal life and store protein until after birth. However, this protein becomes undetectable and both follistatin and Bambi mRNAs are switched on in the cells that have differentiated into spermatogonia (Meehan et al. 2000, Loveland et al. 2003, Loveland et al. 2005). Since BAMBI is a natural transmembrane dominant-negative receptor (Onichtchouk et al. 1999), the implication of its expression on spermatogonia is that it specifically downregulates activin and other TGF-ß ligand signalling inputs into these cells. In contrast, production of the soluble secreted proteins, activin, inhibin or follistatin, could influence any neighbouring somatic or germ cells. A recent analysis of activin A levels in the developing rat testes showed the ßA subunit was at relatively high levels in homogenates of the 3 dpp testis and showed a peak at 6 dpp, the time window enveloping the onset of spermatogenesis (Buzzard et al. 2004). Unfortunately, this study did not examine earlier timepoints, such as 1 dpp, when the gonocytes are present but quiescent and non-motile in the rat testis, so the relative contribution from germ cells to the pool of activin ßA subunit cannot be evaluated from these data.
The capacity for activin to enhance the FSH-mediated stimulation of Sertoli cell proliferation in the rat has been established through in vitro studies from two independent laboratories (Boitani et al. 1995, Buzzard et al. 2003). This effect has been linked to the discrete upregulation in the synthesis of the ActRII mRNA in Sertoli cells between 7 and 11 dpp, in accord with data that have identified the transient timing when activin addition stimulates their proliferation in vitro (Fragale et al. 2001).
The mechanisms that underpin production and secretion of activin and inhibin are complex and are discussed in many reviews (see Loveland & Robertson (2005) and references contained therein). The regulation of activin and inhibin secretion by Sertoli cells is not only modulated by FSH, but also by other factors synthesised in the juvenile and adult testis, such as the pro-inflammatory cytokine, interleukin-1 (IL-1). For example, addition of IL-1 and IL-1ß to immature (20 dpp) and adult rat Sertoli cell cultures stimulates activin A protein release and suppresses inhibin B (Okuma et al. 2005a, b). The addition of FSH significantly modifies these responses by acting directly through the FSH receptors that reside exclusively on Sertoli cells. In contrast, IL-1 and IL-1ß are constitutively produced by the Sertoli cells examined in these experiments, and their production (mRNA or protein) can be further stimulated by addition of developing germ cells and residual bodies to these cultures (Okuma et al. 2005a,b). Hence, there exists a complex autocrine and paracrine regulation on the production of activin and inhibin by Sertoli cells, which in turn act in an autocrine and paracrine manner to regulate testicular development and function.
GDNF
In vitro culture techniques have verified and exploited the role for GDNF as an essential maintenance factor for undifferentiated spermatogonia, first identified using in vivo mouse models (see ‘Adult spermatogenesis and male fertility: insights from genetically modified animals’). The long-term cultures of germline stem cells derived from newborn mouse testes have been successful only when cultured in the presence of GDNF (Kanatsu-Shinohara et al. 2003, 2005, discussed further in ‘Spermatogonial stem cells’). GDNF is produced and secreted by Sertoli cells, and it acts on undifferentiated spermatogonia, which express the c-ret/GFR1 receptor complex (Tadokoro et al. 2002).
BMPs
To date, proliferating Sertoli cells of the juvenile rodent testes have been shown to produce or secrete activin, inhibin and BMP4 subunits, all of which influence germ cell behaviour by signalling through common receptor molecules. Spermatogonia in the early post-natal rodent testis are known to express the activin typeIIB, ALK2, ALK3 and ALK4 receptor subunits (de Winter et al. 1992, Puglisi et al. 2004).
In enriched germ cell cultures from day 4 mouse testes, which contain only undifferentiated spermatogonia, BMP4 addition appeared to induce germ cell differentiation by increasing the level of functional c-kit receptor in preparations of isolated spermatogonia (Pellegrini et al. 2003). In a different experiment, fragment cultures from day 7, mouse testes exhibited increased spermatogonial proliferation in the presence of BMP2 and FSH (Puglisi et al. 2004).
TGF-ß ligands can also influence germ cell survival before and around the time of spermatogenesis onset. TGF-ß2 protein is expressed in the gonocytes of the rodent neonatal testis along with TGF-ß receptors (Olaso et al. 1997), and culture of 3 dpp rat testis fragments in the presence of TGF-ß1 or TGF-ß2 increases the rate of germ cell apoptosis (Olaso et al. 1998). TGF-ß3 protein is also detectable in gonocytes (Olaso et al. 1999) but its functional impact on these cells has not been reported. However, the downregulation of TGF-ß3 mRNA and protein appears to be required for formation of Sertoli cell tight junctions (Liu et al. 2003). The addition of TGF-ß3 to cultures of Sertoli cells isolated from 20 dpc rats perturbed their capacity to form tight junctions and increased the levels of phosphorylated p38 mitogen-activated protein (MAP) kinase, an alternative signalling pathway activated by the TGF-ß superfamily, without affecting the level of Smad proteins. In contrast, the inhibition of p38 MAP kinase phosphorylation in these cultures allowed the tight junction formation to proceed, implicating TGF-ß3 action in the regulating the dynamics of Sertoli cell tight junctions (Liu et al. 2003).
Spermatogonial stem cells
The capacity of the male to continuously produce sperm during adult life depends on the establishment of a spermatogonial stem cell population early during testis development and its subsequent maintenance. Although two models have been proposed for maintenance of the spermatogonial stem cell population (discussed in de Rooij 1998), the mechanism essentially translates at a molecular level as a decision to balance spermatogonial stem cell self-renewal division vs unequal cell division to yield one stem cell and one cell committed to differentiation.
Presently, the absence of a definitive spermatogonial stem cell marker restricts our understanding of stem cell biology to what is inferred from the results of transplantation of isolated spermatogonia into a recipient testis devoid of endogenous spermatogenesis. Analysis of the restoration of spermatogenesis usually facilitated by cells expressing a reporter gene is measured by the number and size of colonies derived following transplantation. This is interpreted as an indication of the number of cells injected that possessed stem cell capacity, or more precisely, the number of spermatogonial stem cells that were able to find a niche, and how efficiently they proliferated before they committed to spermatogenesis.
Transplantation analysis has been routinely used since the mid 1990s (Brinster & Nagano 1998), and has been developed as a particularly informative tool in the analysis of the effect of growth factors on spermatogonial stem cell potential following in vitro culture prior to transplantation.
Spermatogonial stem cell culture aims to establish conditions that support proliferation of stem cells, but not differentiation. As spermatogonia have proven difficult to maintain in culture for long periods of time, culture studies initially assessed the effect of growth factors on spermatogonial stem cells for short periods in vitro. Activin A was identified as a mitogenic factor for rat spermatogonia (Mather et al. 1990), and a transgenic mouse model (BK mouse) with reduced levels of bioactive activin shows a delay in the onset of spermatogenesis (Brown et al. 2000, discussed in ‘Adult spermatogenesis and male fertility: insights from genetically modified animals’), however, a specific effect of activin on spermatogonial stem cells, rather than on spermatogonia at other stages of differentiation, had not been demonstrated. Nagano et al.(2003) addressed this by culturing an enriched population of spermatogonia from immature and cryptorchid mice on a STO cell feeder line for 7 days in the presence of activin A. They observed a significant reduction in the number of cells with spermatogonial stem cell capacity following transplantation. Culture with BMP4 had a similarly negative impact on the presence of spermatogonial stem cells, while conversely, culture with GDNF significantly increased repopulation following transplantation. These data are in excellent agreement with other in vitro and in vivo analyses, which clearly identify GDNF as the factor that is essential for maintenance of spermatogonial stem cells (discussed in ‘Initiation of spermatogenesis and the first wave’ and ‘Adult spermatogenesis and male fertility: insights from genetically modified animals’).
The use of feeder layers to maintain spermatogonia in culture gives the additional complication that such feeder layers themselves provide growth stimuli. Nagano et al.(2003) aptly addressed this by looking at the effect of 13 different feeder layers. Not surprisingly, the ability to maintain spermatogonial stem cells varied with the feeder cells used. For example, when spermatogonia were cultured the Sertoli cell-derived TM4 feeder layer, transplantation potential was reduced. Interestingly, this TM4 cell line is likely to produce activin A, as the activin ßA subunit transcript is detected by RT-PCR in these cells (our unpublished data).
Addition of GDNF to serum-free culture media has permitted long-term culture of stem cell-competent spermatogonia (Kanatsu-Shinohara et al. 2003), although this result depended on the mouse strain tested. Maintenance in culture of spermatogonial stem cells derived from C57Bl/6 mouse strain requires addition of soluble GFR1- (a GDNF receptor) as well as basic fibroblast growth factor, whereas GDNF alone is sufficient to maintain spermatogonial stem cells from DBA mice. Other studies have reported similar enhancement of stem cell maintenance when culture media is supplemented with both GDNF and GFR1- (Kubota et al. 2004, Ryu et al. 2005), in accord with data from analysis of genetically modified mice (discussed further in ‘Glial cell line-derived neurotropic factor’).
Spermatogonial stem cells isolated from immature pup testes have a better short-term survival rate than do adult spermatogonial stem cells (Nagano et al. 2003). A key difference between these populations is that the c-kit receptor tyrosine kinase, understood to be absent from the surface of mouse adult spermatogonial stem cells, appears present in a proportion of spermatogonia with repopulation potential isolated from immature testes (Ohbo et al. 2003). This finding is in accord with the observation that the first wave of spermatogenesis is shorter than adult waves (de Rooij 1998), and growth factor profiles within the testis differ between the immature and adult testis (Buzzard et al. 2004). In a recent report by Pellegrini et al.(2003), addition of BMP4 to preparations of spermatogonia from 4 dpp mouse testes upregulated c-kit expression, implicating BMP4 in the commitment of undifferentiated spermatogonia to spermatogenesis, characterised by the onset of c-kit expression. However, in the vitamin A-deficient (VAD) mouse testis, high levels of BMP4 correlated with arrested spermatogenesis due to VAD (Baleato et al. 2005). In this model, all spermatogonia are understood to be spermatogonial stem cells, and the provision of retinol to these VAD mice resulted in a concomitant reduction in BMP4 mRNA levels and alleviation of the spermatogenic arrest. The authors of this study suggested that in the VAD testis, that BMP4 maintains spermatogonia in an undifferentiated state.
It is promising that mechanisms of spermatogonial stem cell maintenance and proliferation appear to be conserved across species. Although most studies addressing the role of TGF-ß superfamily regulation of spermatogonial stem cell potential focus on the mouse, a recent report that rat spermatogonial stem cells are maintained and proliferate in culture in conditions that were optimised for mouse spermatogonial stem cells (which required GDNF supplementation (Ryu et al. 2005)) not only suggest that the role of TGF-ß superfamily ligands in regulating spermatogonial stem cells may be conserved across species, but will also potentially expand possibilities for culturing spermatogonial stem cells of other species, with a view to developing agricultural and medical applications (Ryu et al. 2005).
Adult spermatogenesis and male fertility: insights from genetically modified animals
Several lines of genetically modified mice have been generated to examine the impact of TGF-ß superfamily signalling components on normal physiology (Chang et al. 2002). In many cases, a reproductive phenotype has been identified, leading us to understand that specific family members have roles at precise stages of cellular development during spermatogenesis (Chang et al. 2001, 2002, Kulkarni et al. 2002). Selected examples are reviewed briefly here.
BMPs
BMP7, 8a and 8b are expressed in spermatogonia and early spermatocytes in the juvenile testis, with expression shifting to round spermatids of the adult testis. Targeted inactivation of these genes revealed that BMP8b is important for the onset of spermatogenesis in the juvenile testis, while BMP8a did not affect the onset, but is necessary for maintaining spermatogenesis in the adult. Deletion of BMP8a results in increased apoptosis of meiotic germ cells in 20–50% of the seminiferous tubules, compromising spermiogenesis and resulting in loss of fertility. When BMP8a null mice were crossed with BMP7 heterozygous mice, the proportion of tubules exhibiting the degenerative phenotype increased to 40–80%, indicating a supplementary role of BMP7 in maintaining spermatogenesis (Zhao et al. 1996, 1998, 2001). BMP4 knockout mice die at birth and initial analysis of their testes showed there were no germ cells present (Lawson et al. 1999). On the other hand, heterozygous mice survive to adulthood and at 4 weeks of age display normal testicular phenotype. However, the adult heterozygotes have lower testis weights and display a decline in sperm number and motility due to the degeneration of meiotic germ cells, similar to the phenotype observed with BMP8a null mice (Hu et al. 2004). These knockout models also revealed the importance of BMP4, 7, 8a and 8b in maintaining epididymal function; the degeneration of the epididymis in null mice may be partly responsible for the observed defects in sperm motility.
Activin and inhibin
While deletion of the coding sequence of the mature ßA subunit (encoded by the Inhba gene) produces a mouse that does not survive beyond birth (Matzuk et al. 1992), deletion of the mature coding domain of the activin ßB subunit (encoded by the Inhbb gene) yields fertile mice that do survive to adulthood and have an apparently normal testis phenotype (Vassali et al. 1994). In order to understand the impact of activin in the post-natal animal, transgenic mice were produced, termed BK mice that have reduced levels of bioactive activin through insertion of the mature activin ßB coding sequence into the Inhba locus. In this way, the activin ßB protein is produced with the ßA precursor and so is processed similar to the wild-type activin ßA protein. However, activin B protein has a lower biopotency than activin A due to its lower affinity for activin receptors and different physiological properties (Brown et al. 2000). The insertion of Inhbb coding region into the Inhba locus rescued the Inhba null mice from craniofacial malformations and neonatal lethality, but BK mice exhibit lower testes weights and volume than their wild-type siblings. In addition, the delayed onset of fertility experienced by the BK mice revealed the importance of activin A in establishing spermatogenesis in the mouse. A related transgenic mouse line, the activin typeIIA receptor (ActRIIA) knockout, with compromised signalling from activin and some BMPs, also has reduced testis size and delayed fertility onset. However, the BK male mice show a 50% increase in FSH levels, while the ActRIIA null mice show a reduction in FSH levels, suggesting that activin A alone, independent of FSH levels, can influence germ cell and Sertoli cell development and ultimately, fertility (for a detailed review see Chang et al. (2001)). However, the prospect for altered signalling by other ligands of this superfamily, in the absence of activin A, remains unexplored.
Further, evidence of the importance of regulating TGF-ß superfamily signalling during germ cell development is derived from analysis of transgenic mice overexpressing the activin ßA subunit by placement of the hInhba subunit under the metallothionein promoter (Tanimoto et al. 1999). Preferential extopic expression of activin ßA subunit in the pachytene spermatocytes and round spermatids was observed from 3 weeks of age. These animals were initially fertile and then became sterile by 10 weeks of age due to an inability to produce sperm. The testes were smaller in size and histology revealed the tubules were deficient in secondary spermatocytes and round spermatids and exhibited vacuolation.
Mice that do not produce the inhibin subunit (resulting from deletion of the Inha gene) develop testicular somatic cell tumours by 4 weeks of age and die by 12 weeks from cachexia, a wasting syndrome induced by elevated activin levels (Matzuk et al. 1994). These mice cannot make inhibin, illustrating that inhibin acts as a tumour suppressor, but its precise role in normal adult testis function cannot be determined from this model, as the tumours disrupt the architecture of the testis. However, the histology of these testes prior to 4 weeks of age, when activin and presumably inhibin are governing developmental events, has not been reported. To elucidate, if inhibin plays a role in testicular development, inhibin A protein was overexpressed in the livers of mice using a GeneSwitch system (Pierson et al. 2000), which drives production of linked coding sequences of activin ßA and inhibin protein subunits. The inhibin A protein produced was secreted and acted on the pituitary to reduce levels of circulating FSH. When inhibin A protein was induced in 3-week-old Inha knockout mice, the mice did not develop tumours and were fertile, displaying a normal testis phenotype and undetectable levels of activin A. When inhibin A protein was induced in 3-week-old wild-type mice for 8 weeks, it exhibited reduced testis weights, reduced seminiferous tubule volume and diameter and were fertile (Pierson et al. 2000). Transgenic mice, over-expressing the inhibin subunit using the metallothionein-I (MT-) promoter were also reported to be fertile and displayed reduced FSH levels. The transgene was expressed in the pituitary, liver, testis and kidney, however a testis phenotype was not recorded (Yomogida et al. 2003). The phenotypes observed in the inhibin overexpressing mice are similar to that of the ActRIIA null mice mentioned earlier. Hence, whether inhibin has a direct role in testis development or whether its role is in modulating activin activity cannot be deciphered in these transgenic models. It appears that the relationship between the gonadal hormones, inhibin, activin and FSH, is very sensitive, so that disruption of one affects the expression or action of other, hence the resulting phenotype cannot be attributed to the effect of a single gene product.
Glial cell line-derived neurotropic factor
Genetically modified mice have also been informative in identifying a role for GDNF in spermatogonial stem cell biology. GDNF is secreted by Sertoli cells and its expression is regulated by FSH (Tadokoro et al. 2002). Mice with one allele of the GDNF gene exhibit depletion in stem cell reserves, revealing that a threshold level of GDNF is required for spermatogonial stem cell maintenance in vivo (Meng et al. 2000). The first wave of spermatogenesis is initiated, but testicular atrophy is observed in the testis of these mice at 5 weeks of age, with many tubules containing spermatocytes and spermatids but lacking spermatogonia. Subsequently, ongoing adult spermatogenesis failed and male mice were rendered infertile as the spermatogonial stem cell population became exhausted, and by 8 weeks, depletion of germ cells resulted in Sertoli cell only tubules.
Conversely, transgenic mice overexpressing GDNF in the testis are also infertile, with a testicular phenotype showing overpopulation of undifferentiated spermatogonia and a block in the progression into differentiating spermatogonia followed by germ cell apoptosis (Meng et al. 2000). In this model, over-expression of hGDNF in mouse testes was driven using the testis-specific elongation factor-1 (EF-1) promoter. The spermatogonial stem cells proliferated but did not appear to respond to normal differentiation cues, including retinoic acid, leading to disruption of the seminiferous epithelium and spermatogenic failure. At 2–3 weeks of age, the tubules displayed a chimeric phenotype with large cell clusters of type A spermatogonia. These mice exhibited lower testis weights at 4 and 8 weeks of age and did not become fertile as expected by 8 weeks. The germ cell clusters degenerated such that at 10 weeks of age, the tubules were occupied by a ring of spermatogonial cells with no spermatozoa observed. These transgenic mice ultimately fall subject to non-metastatic testicular tumours (Meng et al. 2000, Meng et al. 2001). Interestingly, overexpression of hGDNF in Sertoli cells of 12-day old mice by in vivo transfection using electroporation also resulted in spermatogonial cell clusters in the seminiferous tubules 4 weeks after transfection. These cells were confirmed to be spermatogonial stem cells as they were able to colonise the tubules of busulphan-treated mouse testes and progress through spermatogenesis (Yomogida et al. 2003).
Evidence that germ cells exposed to high levels of GDNF sustain their ability to differentiate was obtained through transplantation analysis. Spermatogonia transferred from a mouse in which Sertoli cells over-express GDNF results in normal spermatogenesis in a normal recipient testis (Yomogida et al. 2003); once spermatogonia are removed from the hyper-stimulated environment, they are able to differentiate. As a method of in vivo spermatogonial stem cell amplification, this can potentially provide a highly efficient means of generating a large number numbers of stem cells for experimental application compared to the numbers obtained from methods currently available.
|
Conclusions: what is left to learn and why we need to know it |
|---|
|
TopAbstractThe canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
As touched on earlier, additional difficulties in understanding the cues that govern testis development and germ cell maturation arise from increasing reports that mouse strain differences are noted when genetically modified mice are crossed onto different genetic backgrounds. During culture experiments, medium composition also has demonstrable effects on spermatogonial stem cell properties, leading to difficulties in comparison of data between laboratories. Spermatogonial stem cells isolated from the mouse strains C57Bl/6J and 129/Sv (Kubota et al. 2004) and ICR, C57Bl/6J X DBA/2 F1 (Kanatsu-Shinohara et al. 2003) mice fail to proliferate in culture, while those on a DBA/2Jx ROSA strain, when cultured with GDNF, will do so (Kubota et al. 2004). While this adds complexity, it also offers a valuable opportunity for using direct comparison of mouse strains to identify factors required for maintenance of stem cell properties and proliferation. As more functional data are gathered, and techniques and experimental systems are refined to detect outcomes of the set of integrated signals delivered within the TGF-ß superfamily network, we can apply this knowledge to key outcomes in reproductive health, spanning from testicular cancer to fertility control.
|
Acknowledgements |
|---|
|
TopAbstractThe canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
|
Footnotes |
|---|
All authors contributed equally to this work
|
References |
|---|
|
TopAbstractThe canonical TGF-ß...TGF-ß superfamily...TGF-ß superfamily...Conclusions: what is left...AcknowledgementsReferences |
|---|
Adams IR & McLaren A 2002 Sexually dimorphic development of mouse primordial germ cells: switching from oogenesis to spermatogenesis. Development 129 1155–1164.
Baleato RM, Aitken RJ & Roman SD 2005 Vitamin A regulation of BMP4 expression in the male germ line. Developmental Biology 286 78–90.[CrossRef][ISI][Medline]
Behringer RR, Finegold MJ & Cate RL 1994 Mullerian-inhibiting substance function during mammalian sexual development. Cell 79 415–425.[CrossRef][ISI][Medline]
Bernard DJ, Chapman SC & Woodruff TK 2002 Inhibin binding protein (InhBP/p120), betaglycan and the continuing search for the inhibin receptor. Molecular Endocrinology 16 207–212.
Boitani C, Stefanini M, Fragale A & Morena AR 1995 Activin stimulates Sertoli cell proliferation in a defined period of rat testis development. Endocrinology 136 5438–5444.[Abstract]
Brinster RL & Nagano M 1998 Spermatogonial stem cell transplantation, cryopreservation and culture. Cell and Developmental Biology 9 401–409.[CrossRef]
Brown CW, Houston-Hawkins DE, Woodruff TK & Matzuk MM 2000 Insertion of Inhbb into the Inhba locus rescues the Inhba-null phenoytpe and reveals new activin functions. Nature Genetics 25 453–457.[CrossRef][ISI][Medline]
Buehr M 1997 The primordial germ cells of mammals: some current perspectives. Experimental Cell Research 232 194–207.[CrossRef][ISI][Medline]
Buzzard JJ, Farnworth PG, de Kretser DM, O’Connor AE, Wreford NG & Morrison JR 2003 Proliferative phase Sertoli cells display a developmentally regulated response to activin in vitro. Endocrinology 144 474–483.
Buzzard JJ, Loveland KL, O’Bryan MK, O’Connor AE, Bakker M, Hayashi T, Wreford NG, Morrison JR & de Kretser DM 2004 Changes in circulating and testicular levels of inhibin A and B and activin A during postnatal development in the rat. Endocrinology 145 3532–3541.
Chang H & Matzuk MM 2001 Smad5 is required for mouse primordial germ cell development. Mechanisms of Development 104 61–67.[CrossRef][ISI][Medline]
Chang H, Huylebroeck D, Verschueren K, Guo Q, Matzuk MM & Zwijsen A 1999 Smad5 knockout mice die at mid-gestation due to multiple embryonic and extraembryonic defects. Development 126 1631–1642.[Abstract]
Chang H, Lau AL & Matzuk MM 2001 Studying TGF-beta superfamily signaling by knockouts and knockins. Molecular and Cellular Endocrinology 180 39–46.[CrossRef][ISI][Medline]
Chang H, Brown CW & Matzuk MM 2002 Genetic analysis of the mammalian transforming growth factor-beta superfamily. Endocrine Reviews 23 787–823.
Clarke TR, Hoshiya Y, Yi SE, Liu X, Lyons KM & Donahoe PK 2001 Mullerian inhibiting substance signaling uses a bone morphogenetic protein (BMP)-like pathway mediated by ALK2 and induces SMAD6 expression. Molecular Endocrinology 15 946–959.
Cook RW, Thompson TB, Jardetzky TS & Woodruff TK 2004 Molecular biology of inhibin action. Seminars in Reproductive Medicine 22 269–276.[CrossRef][ISI][Medline]
Coucouvanis E & Martin GR 1999 BMP signaling plays a role in visceral endoderm differentiation and cavitation in the early mouse embryo. Development 126 535–546.[Abstract]
Cupp AS, Kim G & Skinner MK 1999 Expression and action of transforming growth factor beta (TGFbeta1, TGFbeta2 and TGFbeta3) during embryonic rat testis development. Biology of Reproduction 60 1304–1313.
de Felici M, Dolci S & Pesce M 1992 Cellular and molecular aspects of mouse primordial germ cell migration and proliferation in culture. International Journal of Developmental Biology 36 205–213.[ISI][Medline]
de Kretser DM, Buzzard JJ, Okuma Y, O’Connor AE, Hayashi T, Lin SY, Morrison JR, Loveland KL & Hedger MP 2004 The role of activin, follistatin and inhibin in testicular physiology. Molecular and Cellular Endocrinology 225 57–64.[CrossRef][ISI][Medline]
de Rooij DG 1998 Stem cells in the testis. International Journal of Experimental Pathology 79 67–80.[CrossRef][ISI][Medline]
de Rooij DG & Russell LD 2000 All you wanted to know about spermatogonia but were afraid to ask. Journal of Andrology 2 776–798.
de Sousa Lopes SM, Roelen BA, Monteiro RM, Emmens R, Lin HY, Lawson KA & Mummery CL 2004 BMP signaling mediated by ALK2 in the visceral endoderm is necessary for the generation of primordial germ cells in the mouse embryo. Genes and Development 18 1838–1849.
de Winter JP, Themmen AP, Hoogerbrugge JW, Klaij IA, Grootegoed JA & de Jongh FH 1992 Activin receptor mRNA expression in rat testicular cell types. Molecular and Cellular Endocrinology 83 R1–8.[CrossRef][ISI][Medline]
di Clemente N, Wilson C, Faure E, Boussin L, Carmillo P, Tizard R, Picard JY, Vigier B, Josson N & Cate R 1994 Cloning, expression and alternative splicing of the receptor for anti-Mullerian hormone. Molecular Endocrinology 8 1006–1020.[Abstract]
Donahoe PK, Cate RL, MacLaughlin DT, Epstein J, Fuller AF, Takahashi M, Coughlin JP, Ninfa EG & Taylor LA 1987 Mullerian inhibiting substance: gene structure and mechanism of action of a fetal regressor. Recent Progress in Hormone Research 43 431–467.[ISI][Medline]
Drummond AE 2005 TGF beta signalling in the development of ovarian function. Cell and Tissue Research 322 107–115.[CrossRef][ISI][Medline]
Fragale A, Publisi R, Morena AR, Stefanini M & Boitani C 2001 Age-dependent activin receptor expression pinpoints activin A as a physiological regulator of rat Sertoli cell proliferation. Molecular Human Reproduction 7 1107–1114.
Godin I & Wylie CC 1991 TGF beta 1 inhibits proliferation and has a chemotropic effect on mouse primordial germ cells in culture. Development 113 1451–1457.[Abstract]
Gray PC, Harrison CA & Vale W 2003 Cripto forms a complex with activin and type II activin receptors and can block activin signaling. PNAS 100 5193–5198.
Gu Z, Nomura M, Simpson BB, Lei H, Feijen A, van den Eijnden-van Raaij J, Donahoe PK & Li E 1998 The type I activin receptor ActRIB is required for egg cylinder organization and gastrulation in the mouse. Genes and Development 12 844–857.
Gu Z, Reynolds EM, Song J, Lei H, Feijen A, Yu L, He W, MacLaughlin DT, van den Eijnden-van Raaij J, Donahoe PK & Li E 1999 The type I serine/threonine kinase receptor ActRIA (ALK2) is required for gastrulation of the mouse embryo. Development 126 2551–2561.[Abstract]
Hayashi K, Kobayashi T, Umino T, Goitsuka R, Matsui Y & Kitamura D 2002 SMAD1 signaling is critical for initial commitment of germ cell lineage from mouse epiblast. Mechanisms of Development 118 99–109.[CrossRef][ISI][Medline]
Heyer J, Escalante-Alcalde D, Lia M, Boettinger E, Edelmann W, Steward CL & Kucherlapati R 1999 Postgastrulation Smad2-deficient embryos show defects in embryo turning and anterior morphogenesis. PNAS 96 12595–12600.
Hill CS 1996 Signalling to the nucleus by members of the transforming growth factor-beta (TGF-beta) superfamily. Cell Signal 8 533–544.[CrossRef][ISI][Medline]
Hogan BL 1996 Bone morphogenetic proteins in development. Current Opinions in Genetics and Development 6 432–438.[CrossRef]
Hu J, Chen YX, Want D, Qi X, Li TG, Hao J, Mishina Y, Garbers DL & Zhao GQ 2004 Developmental expression and function of Bmp4 in spermatogenesis and in maintaining epididymal integrity. Developmental Biology 276 158–171.[CrossRef][ISI][Medline]
Jamin SP, Arango NA, Mishina Y, Hanks MC & Behringer RR 2002 Requirement of Bmpr1a for Mullerian duct regression during male sexual development. Nature Genetics 32 408–410.[CrossRef][ISI][Medline]
Kanatsu-Shinohara M, Oronuki N, Inoue K, Miki H, Ogura A, Toyokuni S & Shinohara T 2003 Long-term proliferation in culture and germline transmission of mouse male germline stem cells. Biology of Reproduction 69 612–616.
Kanatsu-Shinohara M, Miki H, Inoue K, Ogonuki N, Toyokuni S, Ogura A & Shinohara T 2005 Long-term culture of mouse male germline stem cells under serum- or feeder-free conditions. Biology of Reproduction 72 985–991.
Kerr JB, Loveland KL, O’Bryan MK & de Kretser DM 2006 Cytology of the testis and intrinsic control mechanisms. In Knobil and Neill’s Physiology of Reproduction, 3rd edn edn, pp 827–947. Ed. JD Neill. USA: Elsevier Science.
Koopman P, Munsterberg A, Capel B, Vivian N & Lovell-Badge R 1990 Expression of a candidate sex-determining gene during mouse testis differentiation. Nature 348 450–452.[CrossRef][Medline]
Kubota H, Avarbock MR & Brinster RL 2004 Growth factors essential for self-renewal and expansion of mouse spermatogonial stem cells. PNAS 101 16489–16494.
Kulkarni AB, Thyagarajan T & Letterio JJ 2002 Function of cytokines within the TGF-beta superfamily as determined from transgenic and gene knockout studies in mice. Current Molecular Medicine 2 303–327.[CrossRef][Medline]
Lawrence DA 1996 Transforming growth factor-beta: a general review. European Cytokine Networks 7 363–374.
Lawson KA, Dunn NR, Roelen BA, Zeinstra LM, Davis AM, Wright CV, Korving JP & Hogan BL 1999 Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes and Development 13 424–436.
Lee MM, Seah CC, Masiakos PT, Sottas CM, Preffer FI, Donahoe PK, Maclaughlin DT & Hardy MP 1999 Mullerian-inhibiting substance type II receptor expression and function in purified rat Leydig cells. Endocrinology 140 2819–2827.
Lewis KA, Gray PC, Blount AL, MacConell LA, Wiater E, Bilezikjian LK & Vale W 2000 Betaglycan binds inhibin and can mediate functional antagonism of activin signaling. Nature 404 411–414.[CrossRef][Medline]
Lin SJ, Lerch TF, Cook RW, Jardetzky TS & Woodruff TK 2006 The structural basis of TGF-ß, bone morphogenic protein, and activin ligand binding. Reproduction 132 179–190.
Liu WY, Lee WM & Chent CY
