2022 年 97 巻 1 号 p. 15-25
Continuity of spermatogenesis in mammals is underpinned by spermatogenic (also called spermatogonial) stem cells (SSCs) that self-renew and differentiate into sperm that pass on genetic information to the next generation. Despite the fundamental role of SSCs, the mechanisms underlying SSC homeostasis are only partly understood. During homeostasis, the stem cell pool remains constant while differentiating cells are continually produced to replenish the lost differentiated cells. One of the outstanding questions here is how self-renewal and differentiation of SSCs are balanced to achieve a constant self-renewing pool. In this review, we shed light on the regulatory mechanism of SSC homeostasis, with focus on the recently proposed mitogen competition model in a facultative (or open) niche microenvironment.
Germ cells are the only cell type whose genetic information is passed to the next generation. In addition to genome transmission through the germline, evolution has selected strategies that maximize the number of gametes that successfully produce the next generations of gametes. In many species, males utilize the spermatogenic stem cell (SSC) system to sustain sperm production during their lifetime. SSCs continue to proliferate throughout the reproductive phase to generate cells that undergo differentiation, while simultaneously renewing themselves. Mammalian species produce millions of sperms per day (Tegelenbosch and de Rooij, 1993; Fayomi and Orwig, 2018), while sustaining a sufficiently large self-renewing pool for the considerable extent of spermatogenesis that is required. However, the homeostatic mechanisms that maintain the huge number of stem cells (approximately 35,000 per adult mouse testis) by balancing between self-renewal and differentiation remain poorly understood (Russell et al., 1990; Tegelenbosch and de Rooij, 1993; Fayomi and Orwig, 2018; La and Hobbs, 2019; Yoshida, 2020). In this review, we summarize recent progress in understanding how murine SSCs sustain successive sperm production while maintaining a constant self-renewing pool.
Organization of the testis and spermatogenesisThe murine adult testis comprises two major compartments: the seminiferous tubule, where spermatogenesis occurs, and the interstitial compartment (interstitium) that contains blood vessels and lymphatic space (Russell et al., 1990; Fig. 1A). The most frequent cell type in the interstitium is the Leydig cell, which is the major source of testosterone. Macrophages have also been found in the interstitium (Fig. 1A). The seminiferous tubule is bounded by basement membrane (rich in collagen and laminin), peritubular myoid cells (PMCs) and lymphatic endothelial cells (LECs) (Fig. 1A). Sertoli cells that extend cytoplasmic projections are the only somatic cell type residing among germ cells within the seminiferous tubule (Fig. 1A). Sertoli cells provide much of the structural framework for the organization of the seminiferous epithelium, and support spermatogenesis. Tight junctions between Sertoli cells separate the basal and adluminal compartments within the tubule (França et al., 2016; Fig. 1A). In adult mice, spermatogenesis is initiated upon the production of differentiated cells from SSCs that subsequently undergo mitotic expansion prior to the initiation of meiosis. There are two types of spermatogonia: undifferentiated spermatogonia (Aundiff, which include SSCs) and differentiating spermatogonia (Adiff, In or B), which are considered committed transient-amplifying cells, proceeding to undergo meiotic division of spermatocytes via preleptotene spermatocytes, and maturation of haploid spermatozoa (Russell et al., 1990; Yoshida, 2019).
Spatial heterogeneity of the SSC niche microenvironment of the mouse testis. (A) Cross-sectional view of seminiferous tubules and interstitium. Aundiff preferentially reside adjacent to the vasculature and interstitium. While LECs and PMCs uniformly cover the seminiferous tubules, a subset of LECs (FGF5+) is biased adjacent to the interstitium and vasculature. (B, C) Outside views of seminiferous tubules. (B) The interstitium is shown by dark gray. Aundiff are inside the tubule while Leydig cells, macrophages, vasculature and LECs are outside. FGF5+ LECs are biased toward the vasculature and interstitium. Aundiff preferentially locate in the tubules adjacent to FGF5+ LECs. (C) PMCs and peritubular macrophages are apparently ubiquitous on the seminiferous tubules. (D) In the rectangle defined by the black lines in (B), GFRα1+ cells effectively compete for a limited supply of FGF. The fate of cells consuming larger amounts of FGF will tilt toward self-renewal (proliferation) without differentiation, while cells consuming smaller amounts of FGF will tilt toward differentiation.
SSC activity resides within a minor fraction of Aundiff, which show heterogeneous morphology. Some fractions of Aundiff are singly isolated (called Asingle or As), and many others are interconnected via intercellular bridges, forming syncytia of variable lengths including Apaired (Apr) and Aaligned (Aal) (Huckins, 1971; Oakberg, 1971; de Rooij, 1973) (reviewed in de Rooij, 2017; Yoshida, 2018, 2019, 2020). The behavior of SSCs remains to be fully elucidated. However, lineage-tracing assays have revealed that long-term-retained SSCs express Gfrα1, Nanos2, Id4, Bmi1, Pax7, Pdx1, Eomes, T and Shisa6. (Sada et al., 2009; Nakagawa et al., 2010; Aloisio et al., 2014; Hara et al., 2014; Komai et al., 2014; Tokue et al., 2017; La et al., 2018; Yoshida, 2018, 2019, 2020; Sharma et al., 2019). Among these, Gfrα1 and Nanos2 are critical for the long-term maintenance of SSCs (Sada et al., 2009, 2012). In contrast, recent single-cell RNA-seq experiments have been conducted to characterize multiple and sequential transcriptional states within Aundiff (Hammoud et al., 2015; Chen et al., 2018; Green et al., 2018; Hermann et al., 2018; La et al., 2018; Suzuki et al., 2021). Although a consensus has not been established with respect to the markers defining the identity of SSCs, expression of GFRα1 defines a small fraction of Aundiff demonstrating SSC activity, comprising mainly As and Apr along with few Aal (Hofmann et al., 2005; Nakagawa et al., 2010; Chan et al., 2014; Hara et al., 2014; Helsel et al., 2017; Lord and Oatley, 2017; Garbuzov et al., 2018; Yoshida, 2018, 2019, 2020). GFRα1+ Aundiff are primitive populations that are able to self-renew and give rise to GFRα1– Aundiff (NGN3+ and RARγ+), whereas GFRα1– Aundiff are primed for differentiation and rarely contribute to the self-renewing pool during homeostasis (Yoshida et al., 2004; Nakagawa et al., 2007, 2010; Gely-Pernotet al., 2012; Ikami et al., 2015; Carrieri et al., 2017; Yoshida, 2018, 2019, 2020). GFRα1– Aundiff, however, remain capable of reversion to GFRα1+ spermatogonia and the self-renewing pool, a process that gains prominence during regeneration after damage or transplantation (Nakagawa et al., 2007, 2010; Nakamura et al., 2021). GFRα1– Aundiff further differentiate to KIT+ differentiating spermatogonia that are committed to differentiation (Schrans-Stassen et al., 1999).
Spatial correlation between GFRα1+ cells and the interstitium and vasculatureIntravital live-imaging studies have demonstrated that GFRα1+ spermatogonia continually interconvert between the states of As, Apr and Aal through incomplete division and intercellular bridge breakdown. They actively migrate between Sertoli cells and, probably, the differentiating spermatogonia in the basal compartment of seminiferous tubules. The migration is seemingly random, with a majority of GFRα1+ spermatogonia showing continuous crawling over the basement membrane near the interstitium (Yoshida et al., 2007; Hara et al., 2014). Therefore, SSC regulation takes place in a facultative (or open) niche microenvironment that is distinct from a specialized region where stem cells are gathered (known as a definitive or closed niche), such as the Drosophila testis and ovary (Morrison and Spradling, 2008; Stine and Matunis, 2013; Inaba et al., 2015; Yoshida, 2018, 2019, 2020). In the basal compartment of seminiferous tubules, although GFRα1+ spermatogonia show higher densities in areas near the interstitium and vasculature, they do not cluster but rather intersperse between their differentiating progeny, such as NGN3+ and KIT+ spermatogonia (Chiarini-Garcia et al., 2001; Yoshida et al., 2007; Hara et al., 2014). In the present review, we address two key questions: why do GFRα1+ spermatogonia prefer the area facing the interstitium, and which somatic cell components regulate SSC homeostasis in the open niche?
Every GFRα1+ spermatogonium resides on the basement membrane in direct contact with Sertoli cells (Fig. 1A). Sertoli cells demonstrate a dual function involving the promotion of both self-renewal and differentiation by producing many extrinsic factors, such as GDNF, WNT and retinoic acid (RA) (Meng et al., 2000; Ikami et al., 2015; Tokue et al., 2017; Sharma and Braun, 2018). Although temporal heterogeneity of the seminiferous epithelium is evident, as discussed later, it remains unknown whether Sertoli cells also show spatial heterogeneity in association with the interstitum and vasculature (Fig. 1A).
Peritubular myoid cellsPMCs uniformly surround the seminiferous tubules and exhibit flattened and thin structures, similar to a cobblestone pattern (Fig. 1A, 1C), thereby providing structural support and aiding the peristaltic contraction of seminiferous tubules to enable movement of spermatozoa and luminal fluid to the rete testis (Maekawa et al., 1996). PMCs are known to secrete paracrine factors, including GDNF, that are important for SSCs (Chen et al., 2016). However, their correlation with the local heterogeneity of SSCs remains unknown (Fig. 1C).
Leydig cellsAndrogens, such as testosterone, are produced in Leydig cells, which are most commonly found in the interstitium (Fig. 1A). Androgens are essential for spermatogenesis because transgenic animals with a disrupted androgen receptor exhibit various abnormalities related to fertility, such as germ cell arrest at meiotic stages (Yeh et al., 2002; De Gendt et al., 2004; O’Hara and Smith, 2015). Although androgens are known to affect spermatogenesis in Sertoli cells and PMCs, the downstream effectors of this paracrine signaling network have not been fully elucidated (O’Hara and Smith, 2015; Heinrich and DeFalco, 2019). Apart from testosterone, Leydig cells also produce factors that directly target SSCs, such as CSF1 and IGF1 (Huang et al., 2009; Oatley et al., 2009; Wang et al., 2015). When IGF1 is added to in vitro-cultured spermatogonia, it increases their proliferation (Kubota et al., 2004; Wang et al., 2015). However, its role in the regulation of spermatogenesis and SSCs under in vivo conditions remains unknown.
MacrophagesTesticular macrophages have been implicated in the maintenance of the immunosuppressive environment, testosterone production and spermatogenesis (Yee and Hutson, 1985; Gaytan et al., 1994; Cohen et al., 1996; Pollard et al., 1997; Nes et al., 2000; Fijak and Meinhardt, 2006). Testicular macrophages comprise interstitial macrophages that intermingle with Leydig cells and are associated with vasculature, and peritubular macrophages that intermingle with PMCs (Fig. 1A–1C; DeFalco et al., 2015). DeFalco et al. (2015) observed that macrophages were enriched in regions overlying Aundiff and that transient depletion of macrophages resulted in fewer spermatogonia. Thus, macrophages may influence the regulation of spermatogonia, likely via the regulation of CSF1 expression and RA synthesis. Recently, Lokka et al. (2020) revealed the heterogeneity and developmental origin of testicular macrophages using single-cell analyses. However, what type of macrophage is required for spermatogenesis and SSC regulation, and whether macrophages regulate the balance between self-renewal and differentiation of SSCs, are still under discussion. This topic warrants future studies.
Lymphatic endothelial cellsHalf a century ago, observation of the interstitial tissue of the testis via electron microscopy demonstrated that the interstitium contained lymphatic space and that endothelial cells were located at the interface of the interstitial and tubular compartments of the testis (Dym and Fawcett, 1970, 1971; Fawcett et al., 1973; Clark, 1976). However, markedly few studies have been conducted to address the physiological functions of lymphatics and endothelial cells in spermatogenesis and SSCs. The organization of the lymphatics within the testis is not based on a typical vessel structure but comprises irregular channels or sinusoids that are incompletely bounded by endothelial cells. The lymphatic fluids seem to percolate through the cells of the interstitium; however, the biological function of the lymphatics remains unknown (Russell et al., 1990). The peritubular LECs (CD34+) are located on the outside of the PMCs and cover the tubule uniformly, albeit incompletely, in mice and humans (Fig. 1A; Kuroda et al., 2004; Kitadate et al., 2019). The LECs are observed as thin cells in cross-sections and are recognized as large and flattened cells in whole-mount preparations of seminiferous tubules (Fig. 1A, 1B; Kitadate et al., 2019). Interestingly, primary cultured LECs (CD34+) are capable of promoting the proliferation of cultured spermatogonia in vitro (Seandel et al., 2007; Kim et al., 2008; Kitadate et al., 2019). A subset of LECs express FGF5 and cover 60% of the surface of the tubules, with a bias toward areas facing the interstitium (Fig. 1A, 1B; Kitadate et al., 2019). Across the basement membrane and PMCs, FGF5+ LECs are spatially correlated with GFRα1+ spermatogonia that are furnished with reception machinery consisting of FGF receptors and their downstream signal-transducing machinery. Indeed, GFRα1+ spermatogonia are thought to internalize and consume FGF5 as a mitogen. While heparan sulfates (HSs) showing affinity to FGFs are located on the basement membrane, interestingly, GFRα1+ spermatogonia are rich in HS chains with higher levels of sulfation (and higher affinities for FGFs) on their surface, suggesting that GFRα1+ spermatogonia internalize FGFs from the basement membrane as they randomly migrate (Kitadate et al., 2019). Therefore, FGF5+ LECs can provide spatial heterogeneity, which may promote a higher density of GFRα1+ spermatogonia near the vasculature and the interstitium.
Stem cell homeostasis is a critical state: it is a boundary between the expansion and contraction of a stem cell pool. To achieve long-term homeostasis, the stem cell population must be balanced by cell duplication (through division) and loss (through differentiation). In principle, a balance can be achieved at the level of individual cell divisions, such that one daughter cell is retained as a stem cell while the other differentiates (a mechanism called division asymmetry). Alternatively, asymmetry could be enforced at the population level (called population asymmetry), without the occurrence of any asymmetric division (Klein and Simons, 2011; Simons and Clevers, 2011).
Clonal fate analysis of individual GFRα1+ spermatogonia based on indelible genetic labeling has revealed that population asymmetry maintains the homeostasis without any evidence of asymmetric divisions (Klein et al., 2010; Hara et al., 2014). GFRα1+ spermatogonia follow highly variable fates: some GFRα1+ spermatogonia lead to the formation of more than one GFRα1+ spermatogonium, whereas others produce only differentiating cells. However, their collective behavior establishes a balance between self-renewal and differentiation at the population level (Klein et al., 2010; Hara et al., 2014).
Neutral competition between GFRα1+ spermatogoniaSeminiferous tubules harbor abundant populations of GFRα1+ spermatogonia with equal potential to generate a long-lived lineage of cellular offspring. However, considerable fractions of GFRα1+ spermatogonia are lost through differentiation in a stochastic manner, leading to a neutral competition between GFRα1+ clones. Such population dynamics may be underpinned by competition between GFRα1+ spermatogonia for the availability of limited amounts of niche factors or other local niche environments (Klein et al., 2010; Klein and Simons, 2011; Hara et al., 2014). Only a minority fraction of such an equipotent stem cell population will persist in the long term, resulting in the production of additional stem cells among their offspring. All other stem cells, despite their equipotency with the long-term survivors, are forced to exit from the niche by their proliferative neighbors (through failure to receive the niche factors), following which they undergo differentiation and are cleared from the system. In this scheme, the short- or long-lived stem cells are considered a simple outcome of the stochastic competition. In this case, the outcomes are not based on differences in their intrinsic characteristics.
To achieve a steady state, the production rate of stem cells (input) and the differentiation rate (output) must be completely balanced or must exist in equilibrium. The stem cells maintain their numbers while producing differentiating cells, referred to as “numerical homeostasis” (Lander, 2009). Although the number (i.e., density) of GFRα1+ cells is heterogeneous locally, probably attributable to the spatial heterogeneity in relation to the interstitium and vasculature, the average density of GFRα1+ cells over extended tubule lengths remains highly constant macroscopically (Kitadate et al., 2019).
Mitogen competition underpins numerical homeostasisRecently, we proposed a mechanism that could explain GFRα1+ cell density homeostasis and that was elucidated using the “mitogen competition model” (Kitadate et al., 2019). In this model, GFRα1+ spermatogonia, which are deemed an equipotent population that can stochastically select between alternative fates (namely self-renewal or differentiation), consume FGFs secreted by LECs. GFRα1+ spermatogonia that consume substantial amounts of FGFs show a higher probability of self-renewal, while those that consume a lower amount of FGFs are more inclined to undergo differentiation, a process which is mediated by the expression of cell cycle-promoting and anti-differentiation genes regulated by FGF signaling (Fig. 1B, 1D; Kitadate et al., 2019). The supply level of FGFs impacts on the fate balance as well as the density homeostasis of GFRα1+ spermatogonia, while individual GFRα1+ spermatogonia consume a finite amount of FGFs. As result, GFRα1+ cell density homeostasis is self-organized by the competition for FGFs.
This simple model was proposed based on the following findings (Kitadate et al., 2019). First, the FGF5 protein was detected in the lysosomes of GFRα1+ spermatogonia. Second, there was a linear correlation observed between the supply level of FGFs (FGF4/5/8 gene dosages) and the average density of GFRα1+ spermatogonia in the steady state. Importantly, the minimal theory of mitogen competition quantitatively encompasses a wide range of experimental data, in addition to the dependence of the homeostatic GFRα1+ cell density on FGF supply under steady-state conditions, and the non-trivial oscillatory recovery of GFRα1+ cell density toward a steady state after perturbation. This oscillatory behavior arises because of the dynamic feedback system between GFRα1+ cell density and FGF concentration. The reduction of GFRα1+ cell density leads to decreased FGF consumption, resulting in FGF accumulation, leading to an overshoot of GFRα1+ cell density. Indeed, such a non-trivial overshoot phenotype was also observed in an independent quantitative study of Aundiff after perturbation (van Keulen and de Rooij, 1974).
GFRα1+ cell density homeostasis by mitogen competition is achieved in response to changing circumstances by balancing between self-renewal and differentiation (Kitadate et al., 2019; Jörg et al., 2021). This feedback system may be advantageous for enhancing robustness by avoiding both stem cell exhaustion, which would lead to reduced fertility, and excess accumulation, which would lead to wastage of energy. The system does not always demonstrate a stable hard-wired and steady-state phenotype; it may also exhibit oscillatory behavior after perturbation.
In a closed (or definitive) niche, self-renewal-promoting factors are localized in distinct areas where stem cells gather or anchor (Morrison and Spradling, 2008; Spradling et al., 2011; Stine and Matunis, 2013; Yoshida, 2018, 2019, 2020). In the Drosophila testis, ectopic activation of niche signaling is often sufficient to induce germline stem cell (GSC) tumors (Kiger et al., 2001; Tulina and Matunis, 2001). Thus, it is likely that stem cell fate is predominantly determined by niche signaling and by spatial correlation between the stem cells and the niche signaling. In the Drosophila ovary, multiple niche compartments orchestrate stepwise GSC progeny differentiation (Tu et al., 2021). Spatially distinct sub-compartments exert control on self-renewal and multiple differentiation in a stepwise manner, and self-renewal and differentiation are supported by somatic microenvironments. In contrast, in open niche-supported mouse spermatogenesis, SSCs (GFRα1+ cells) actively migrate and are intermingled between differentiating progeny (Hara et al., 2014). However, the following question remains to be answered: how do different types of cells intermingle with each other to achieve numerical homeostasis balance between both self-renewal and differentiation without confusion?
Temporally dynamic nature of the open niche microenvironmentThe seminiferous epithelium appears highly organized, and comprises Sertoli cells, various generations of spermatogonia in the basal compartment, and spermatocytes and spermatids in the adluminal compartment (Fig. 1A). Different generations of germ cells are associated at fixed combinations in a given area of the seminiferous tubules, termed a ‘cycle of the seminiferous epithelium’ or ‘seminiferous epithelial cycle’ (or simply ‘the cycle’) (Fig. 2A, 2B; Lebrond and Clermont, 1952; Oakberg, 1956a, 1956b; Russell et al., 1990). In mice, a cycle of 8.6 days is divided into stages I to XII (Fig. 2A, 2B; Russell et al., 1990). Typical germ cell combinations are derived from temporally controlled production of differentiating spermatogonia from Aundiff in the basal compartment every 8.6 days, while Aundiff are present throughout the cycle (de Rooij and Russell, 2000; Fig. 2A, 2B). The periodically produced differentiating spermatogonia undergo mitotic expansion, meiosis and spermiogenesis, and finally become spermatozoa that are released into the tubule lumen over a total of four cycles (Fig. 2A; Russell et al., 1990).
Temporal heterogeneity of the SSC niche microenvironment. (A) Murine seminiferous epithelial cycle depicted with colored germ cells (orange shows spermatogonia (including Aundiff, Adiff, In and B); yellow shows spermatocytes, spermatids and spermatozoa). In a series of spermatogenesis, defined combinations among different generations of spermatogenic cells are fixed at each stage. (B) Seminiferous epithelium at each indicated stage shows distinct germ cell associations, while spermatogonia (orange) reside in the basal compartment throughout the cycle. (C) Subpopulations in spermatogonia (including Aundiff, Adiff, In and B) and preleptotene spermatocytes (pL) are characterized by expression of GFRα1, NGN3/RARγ or KIT; GFRα1+ cells are the most primitive population that can self-renew in homeostasis; NGN3+/RARγ+ cells are primed for differentiation while retaining stem cell potential; KIT+ cells (i.e., differentiating spermatogonia) are unidirectionally proceeding toward meiosis. The cells’ different availability for extracellular signals is illustrated. (D) Spermatogonial microenvironment in the basal compartment. RA and GDNF are periodically produced by Sertoli cells in a spatially uniform but periodically fluctuating manner along the cycle. The combination between temporal extracellular signals and different availability of spermatogonia generates different sets of spermatogonial subpopulations. The groupings of cell populations in (D) correspond with particular stages (IV, VI, VIII, X, XII) that are situated directly above them in (B).
The four key transitions (spermatogonial differentiation, meiotic initiation, spermatid elongation and sperm release) all occur in proximity in stages VII and VIII (Oakberg, 1956b). How is this coordination achieved? Endo et al. (2017) revealed that RA concentration rises in stages VII and VIII, thereby coordinating the four key transitions. It has been known since 1925 that vitamin A (VA), a derivative of RA, is required for normal spermatogenesis (reviewed in Livera et al., 2002), and a lack of VA blocks the transition from Aundiff to Adiff. However, when VA or RA then becomes available, a series of spermatogenesis is reinitiated via stimulation of spermatogonial differentiation in a synchronized manner throughout the entire testis (van Pelt and de Rooij, 1990; Gaemers et al., 1996). Therefore, RA is essential for spermatogonial differentiation, and its periodicity underpins the temporal coordination of spermatogenesis. The precise mechanism regulating local RA concentration remains an interesting open question. Given that RA metabolism enzymes are mainly expressed in Sertoli cells and germ cells, especially pachytene spermatocytes and round spermatids, cooperative metabolism by the combination of germ cells and Sertoli cells is likely to establish a periodic fluctuation in local RA concentration (Hogarth and Griswold, 2010; Sugimoto et al., 2012; Endo et al., 2017).
Different availability of differentiation-promoting extracellular signalsA subset of Aundiff become Adiff, while the others remain as Aundiff, upon stimulation with a high concentration of RA during the stages VII and VIII (Vernet et al., 2006; Mark et al., 2008; Sugimoto et al., 2012). The following question remains to be answered: how does Aundiff exhibit such a heterogeneous response to RA? Ikami et al. (2015) have revealed that GFRα1+ spermatogonia do not respond to RA signals, whereas NGN3+ spermatogonia become committed as differentiating spermatogonia upon RA stimulation. RA signals are received by RARγ, which is highly expressed in NGN3+ but not expressed in GFRα1+ cells. When RARγ is ectopically expressed in GFRα1+ cells, they differentiate into KIT+ cells upon stimulation with RA. Therefore, differences in the abilities to exhibit reactions with RA mediated by the expression of RARγ in NGN3+ cells are critical for the asymmetric fate of Aundiff (i.e., to differentiate into differentiating spermatogonia) regardless of the ubiquitous exposure to RA (Fig. 2C, 2D; Ikami et al., 2015).
Asymmetry of the fate of GFRα1+cells: NGN3+ (RARγ+) state or GFRα1+stateA potential strategy for spermatogonia to remain in the stem cell state may involve escape from the differentiation-promoting RA signal. The transition from GFRα1+ (RARγ–) to NGN3+ cells that express RARγ is the first step of differentiation to gain competence in response to RA. However, the following question remains to be answered: why does only a fraction of GFRα1+ cells become NGN3+, while others remain undifferentiated?
Tokue et al. (2017) found a mechanism that might underpin the generation of variation in the differentiation competency among GFRα1+ cells. First, they found that WNT/β-catenin signaling components including receptors, transducers and the downstream target were expressed in both GFRα1+ cells and NGN3+ cells, and promoted the differentiation of GFRα1+ to NGN3+. Second, WNT6 (one of the WNT ligands activating the canonical pathway) showed seminiferous epithelial cycle-related expression in Sertoli cells, which was highest at stages I to VI, suggesting that the differentiation of GFRα1+ into NGN3+ cells is regulated temporally in accordance with the seminiferous epithelial cycle. The authors further identified SHISA6 as a WNT inhibitor expressed in a limited subset of the GFRα1+ cell population. In this case, differential expression of SHISA6 in the GFRα1+ cell population may confer heterogeneous resistance to uniformly distributed WNT. In particular, GFRα1+ cells with higher levels of SHISA6 remained undifferentiated, whereas those with lower SHISA6 levels tended to undergo differentiation. Interestingly, SHISA6 expression is upregulated in in vitro-cultured spermatogonia with FGF5 supplementation (Kitadate et al., 2019). However, further studies are warranted on the in vivo relationship between FGFs and SHISA6.
Multiple temporally distinct signals promote stepwise differentiationWNT and RA signals periodically fluctuate along the seminiferous epithelial cycle. During stages when a certain signal is active, the corresponding ligand (or metabolizing enzymes for RA) shows uniform spatial distribution all around the tubule circumference, suggesting that the basal compartment is uniformly exposed to these signals (Fig. 2B–2D; Vernet et al., 2006; Ikami et al., 2015; Takase and Nusse, 2016; Tokue et al., 2017). While Aundiff are present throughout the cycle, a fraction of Aundiff may receive WNT and RA signals and undergo stepwise differentiation in the basal compartment (Fig. 2D). Taken together, these observations imply that murine spermatogonia dynamically change their reception machinery in response to environmental factors during multistep differentiation, which facilitates the co-existence of both stem cells and differentiated cells in the same spatial region. Previous histological studies and recent comprehensive gene expression analyses have shown that Sertoli cells residing at different stages of the cycle exhibit characteristic heterogeneity (Kerr, 1988a, 1988b; Ye et al., 1993; Johnston et al., 2008; Hasegawa and Saga, 2012; Green et al., 2018), which strengthens the foundation for understanding the temporal heterogeneity of the niche microenvironment involved in the stepwise differentiation along the cycle.
Cyclic fluctuation of the self-renewal-promoting factor GDNFThe periodicity of SSCs (GFRα1+ spermatogonia) along the cycle remains unknown. However, it is likely that the asymmetry of their fate, leading to the generation of either GFRα1+ or NGN3+ (RARγ+) states of Aundiff, is also regulated in vivo by GDNF, a ligand of GFRα1, in a manner similar to FGF. GDNF is an extracellular factor that critically regulates the balance of self-renewal and differentiation of SSCs (Meng et al., 2000). GDNF is expressed in Sertoli cells and PMCs in a spatially uniform manner around the tubule circumference and in a temporally fluctuating manner in accordance with the seminiferous epithelial cycle, presenting with the highest and lowest expression at stages IX–XII–I and V–VIII, respectively (Sato et al., 2011; Caires et al., 2012; Chen et al., 2016; Tokue et al., 2017; Sharma and Braun, 2018; Parekh et al., 2019). Studies using in vitro-cultured SSCs have revealed that GDNF can promote proliferation while repressing differentiation (Kanatsu-Shinohara and Shinohara, 2013; Parekh et al., 2019). A recent study has reported that cyclical fluctuations in GDNF expression are necessary for proper homeostasis of SSCs (Sharma and Braun, 2018). While GFRα1+ spermatogonia are present throughout the cycle, it remains an interesting question whether, as suggested for FGFs, the amount of GDNF consumed by individual SSCs determines their fate selection between remaining as GFRα1+ cells or becoming differentiation-priming GFRα1− (NGN3+/RARγ+) cells. If this is the case, it remains an open and puzzling question whether and how the temporally fluctuating expression of GDNF can regulate the density homeostasis of SSCs (Fig. 2C, 2D). Future studies should examine whether the aforementioned mitogen competition model fits the function of GDNF in the periodic fluctuation of Aundiff along the cycle.
Despite the heterogeneity, in the extremely long (~1.7 m) structure of seminiferous tubules, a substantial number of dispersed equipotent GFRα1+ spermatogonia can migrate long distances, thereby expanding or diminishing their territories via mitogen competition by establishing a balance between self-renewal and differentiation in a stochastic manner. Thus, numerous and equipotent GFRα1+ spermatogonia can undergo proliferation: that is, they possess the ability of self-renewal, and produce long-lasting clones as stem cells. However, one may ask: what is the benefit of this system? Maintenance of sufficient stock is clearly required for the considerable extent of spermatogenesis that is observed. Homeostasis maintaining the GFRα1+ cell density, attributed to mitogen competition, which is a feedback system, may demonstrate the potential advantages of enhancing robustness by preventing both stem cell depletion, leading to infertility, and excess reservation, leading to wastage of energy.
In this review, we shed light on the relationship between SSCs and both the temporal heterogeneity of the seminiferous epithelium and the spatial heterogeneity in the interstitium and vasculature. In the seminiferous tubules, which are a simple repeated structure, mitogen competition within a spatially and temporally heterogeneous environment is likely to create a difference between equivalent SSC populations, which potentially generates both the self-renewing and differentiating progeny. Among the goals of future research will be to further dissect the mechanism controlling the symmetry breaking of GFRα1+ spermatogonia and the differences between GFRα1+ and NGN3+ spermatogonia. Although additional extrinsic factors implicated in self-renewal and differentiation of SSCs have been reported, such as CSF1, Activin A and CXCL12 (Oatley et al., 2009; Kanatsu-Shinohara et al., 2012; Yang et al., 2013; Young et al., 2015), it remains unknown how these factors impact on in vivo SSC homeostasis in the adult. Perhaps synergy between mathematical modeling and experimentation will lead to the discovery of mechanisms involved in stem cell density homeostasis, which would be paradigmatic for stem cell regulation in other open niche-supported tissues.
We thank the members of the Yoshida laboratory for discussions and encouragement. Our studies discussed in this review article were supported in part by Grants-in-Aid for Scientific Research (KAKENHI; grant numbers 20H05548 to K.Y.; JP16H02507, JP18H05551, JP20116004, JP24247041 and JP25114004 to S.Y.) and by AMED-CREST (JP17gm1110005h0001 to S.Y.).
