Abstract
Spermatogonial stem cells (SSCs) and progenitor spermatogonia encompass the undifferentiated spermatogonial pool in mammalian testes. In rodents, this population is comprised of Asingle, Apaired and chains of 4–16 Aaligned spermatogonia. Although traditional models propose that the entire Asingle pool represents SSCs, and formation of an Apaired syncytium symbolizes irreversible entry to a progenitor state destined for differentiation; recent models have emerged that suggest that the Asingle pool is heterogeneous, and Apaired/Aaligned can fragment to produce new SSCs. In this review, we explore evidence from the literature for these differing models representing SSC dynamics, including the traditional ‘Asingle’ and more recently formed ‘fragmentation’ models. Further, based on findings using a fluorescent reporter transgene (eGfp) that reflects expression of the SSC-specific transcription factor ‘inhibitor of DNA binding 4’ (Id4), we propose a revised version of the traditional model in which SSCs are a subset of the Asingle population; the ID4-eGFP bright cells (SSCultimate). From the SSCultimate pool, other Asingle and Apaired cohorts arise that are ID4-eGFP dim. Although the SSCultimate possess a transcriptome profile that reflects a self-renewing state, the transcriptome of the ID4-eGFP dim population resembles that of cells in transition (SSCtransitory) to a progenitor state. Accordingly, at the next mitotic division, these SSCtransitory are likely to join the progenitor pool and have lost stem cell capacity. This model supports the concept of a linear relationship between spermatogonial chain length and propensity for differentiation, while leaving open the possibility that the SSCtransitory (some Asingle and potentially some Apaired spermatogonia), may contribute to the self-renewing pool rather than transition to a progenitor state in response to perturbations of steady-state conditions.
Background
The actions of the undifferentiated spermatogonial population provide the foundation for continuity and robustness of spermatogenesis. Spermatogonial stem cells (SSCs) are a rare subset of the population intermingled with an abundant pool of progenitors that transiently amplify in number before transitioning to a differentiating state at periodic intervals in response to a retinoic acid (RA) pulse. Self-renewal activity by SSCs sustains a reservoir from which the next cohort of progenitors will arise. Whether self-renewal is a symmetric or asymmetric mode remains undefined and likely varies based on circumstances (reviewed by Oatley & Brinster (2008)). Signals that communicate the changing requirements for SSC self-renewal or progenitor production are thought to originate from support cells that constitute a stem cell niche unit (reviewed by Oatley & Brinster (2012)). The stem cell niche microenvironment exists within the seminiferous tubule, in the basal compartment of the seminiferous epithelium between the basement membrane and the Sertoli cell tight junctions. Sertoli cells not only provide architectural support, but are also thought to be the major source of growth factors that are known to be integral for maintenance of the undifferentiated population, including glial cell line-derived neurotrophic factor (GDNF) and fibroblast growth factor (FGF2) (Meng et al. 2000, Chen & Liu 2015). Growth factor support is also thought to be provided by somatic cells in the interstitial space such as the Leydig and peritubular myoid cells (Piquet-Pellorce et al. 2000, Oatley et al. 2009, Chen et al. 2016). Indeed, the supplementation of media with these growth factors (GDNF and FGF2) is required for the maintenance of an SSC pool in primary cultures of rodent undifferentiated spermatogonia (Kanatsu-Shinohara et al. 2003, 2005, Kubota et al. 2004, 2011, Hamra et al. 2005, Ryu et al. 2005, Wu et al. 2009, Kaucher et al. 2012).
In many mammalian species, cytokinesis is incomplete following mitotic division of progenitor spermatogonia, thus leading to the formation of doublets (Apaired) and chains (Aaligned) of up to 16 interconnected cells (Huckins 1971, Oakberg 1971). Upon reception of an RA signal, the progenitor spermatogonia but not SSCs respond by transitioning to a differentiating state. Although the vast majority of progenitors encompassed in chains of 4, 8 and 16 cells experience RA-induced differentiation, as represented by hallmark expression of the RA-responsive genes Stra8 and Kit (Zhou et al. 2008, Busada et al. 2015), some paired and single spermatogonia are resistant to this signal (Tagelenbosch & de Rooij 1993, Ikami et al. 2015). The differentiating spermatogonia undergo several further rounds of mitosis, from A1 to A4, then transition to intermediate and type B spermatogonia; again accompanied by mitotic division, resulting in an exponential increase of cell number in relation to the ‘parent’ SSC. It is from type B spermatogonia that spermatocytes form, accompanied by two rounds of meiosis to generate haploid cells. From spermatocytes arise round and then elongate spermatids, and then finally spermatozoa (reviewed by Oatley & Brinster (2008)). It should be noted that the SSC pool and the mode of amplification for the progenitor spermatogonial populations in humans and non-human primates differ slightly compared to those in rodents. The SSCs in primate species are distinguished as Apale or Adark; that are thought to be active and reserve SSCs respectively (Clermont 1969, Clermont & Antar 1973). However, the presumptive subsets may actually constitute a single population, and, in the human at least, progenitors produced from the SSC mitotic division transition directly into type B differentiating spermatogonia (Clermont 1966). Although the discrete details of spermatogonial activities are slightly different between primates and rodents, the overarching principles and kinetics are similar. In this review, we will focus on studies using the mouse because genetically tractable models exist to study the spermatogonial populations in detail.
Although we do have some understanding of the journey of the SSC through to formation of the spermatozoon, numerous challenges exist in attempting to study the SSC population specifically, and thus in monitoring SSC dynamics. First, the rarity of SSCs, which are thought to make up only 0.03%, possibly less, of cells in the entire testis (Tagelenbosch & de Rooij 1993), makes them notoriously difficult to monitor. In addition to this, historically, there has been a lack of markers available with which to distinguish SSCs from the undifferentiated progenitors that also reside in the same compartment of the seminiferous epithelium. This difficulty is primarily a consequence of the closely related molecular and morphological profiles of the two cell types (de Rooij & Russell 2000, Grisanti et al. 2009, Chan et al. 2014, Hermann et al. 2015). In addition, although putative SSC markers have been identified, the specificity of a number of these for the SSC population has been contested; in part as a result of the different methodologies utilized to assess SSC purity. Notwithstanding the aforementioned challenges, a number of models depicting the dynamics of the SSC population have arisen in the field, including a revised version of the traditional Asingle model, and a ‘fragmentation’ model that proposes fluid interchangeability between Asingle, Apaired and Aaligned cell types. In this review, we will first discuss the tools that are currently available to assess dynamics of the SSC population; including recently identified putative SSC markers, and the spermatogonial transplantation and lineage tracing techniques that have been used to both indirectly and directly assess SSC dynamics. In explaining these tools, we will explore the experimental evidence for the more recently formulated models of SSC dynamics, as well as recap traditional models that were generated in the 1970s, and, in light of these models, speculate as to what characteristics encompass a male germline stem cell.
Approaches for studying SSC dynamics
Assessment of SSC content in spermatogonial populations using transplantation analyses
Spermatogonial transplantation provides a means not only to establish the number of regenerative cells within a selected spermatogonial population, but also to assess specificity of putative SSC markers prior to their utilization for alternative non-quantitative methodologies that can be used to assess SSC dynamics; such as lineage tracing (Helsel & Oatley 2017). The spermatogonial transplantation technique uses a donor population of testis cells that constitutively express a marker such as LacZ. After isolation using a variety of techniques including magnetic or fluorescence activated cell sorting (MACS or FACS), an isolated testis cell population is microinjected into the rete testis of a recipient mouse that lacks endogenous spermatogenesis (Brinster & Avarbock 1994, Brinster & Zimmermann 1994). From 2 months following transplantation, colonization of the recipient testes by donor SSCs can be visualized if they possess a marker transgene (e.g. X-gal staining for LacZ expressing cells). The colonies represent regeneration of continual spermatogenesis, a hallmark of stem cell capacity for the male germline. As one colony in the recipient testes is representative of regeneration by one SSC (Dobrinski et al. 1999, Zhang et al. 2003, Kanatsu-Shinohara et al. 2006), direct quantitative comparisons can be made between different populations of interest. As such, this technique can retrospectively provide assessments of different subpopulations within the undifferentiated pool.
As already mentioned, spermatogonial transplantation is also valuable for assessing the purity of populations selected with putative SSC markers, prior to utilization of these markers for alternative methodologies such as lineage tracing. Specifically, populations of cells selected with the marker of interest (using MACS or FACS) can be transplanted alongside an unselected cell population, or alternatively alongside the opposing population within the undifferentiated pool that does not express the marker of interest (i.e. positive and negative populations). Although the capture of SSCs within a marker-expressing population depicts enrichment of SSCs, it is not necessarily representative of SSC purity. More recently, spermatogonial transplantation has been used in a limiting dilution (LD) manner to provide a means to directly calculate SSC purity in the donor population (Helsel et al. 2017). In this paradigm, spermatogonia that are selected for transplantation are diluted into populations of 1000, 100 and 10 cells; with the rationale that only highly pure populations of SSCs will generate colonies following transplantation of such restricted numbers of spermatogonia. The colonization efficiency of SSCs following transplantation is estimated to be between 5 and 12% (Nagano et al. 1999, Nagano 2003, Ogawa et al. 2003); with the capacity for colonization likely being limited by niche availability and saturation (Oatley et al. 2011b) as well as successful homing of SSCs through the blood-testis barrier to the available niches (Nagano 2003). Using an estimated colonization efficiency of 5%, a pure population of SSCs would produce only one colony in every second LD10 transplanted testis. Using colony counts from the LD10 testes, the ‘purity’ of the donor SSC population can be assessed using the calculation: (# cells transplanted ÷ (# colonies of donor-derived spermatogenesis/5% colonization efficiency) = X), where 1 in every X cells is an SSC.
Monitoring SSC dynamics using reporter transgenes and lineage tracing
In contrast to the quantitative readout of SSC content that can be made by transplantation, current approaches for lineage tracing provide only a qualitative assessment of whether self-renewing cells are contained within the labeled population. However, lineage tracing has the advantage of facilitating observation of in vivo dynamics during steady-state conditions. Lineage tracing involves the utilization of animal models containing a transgene reporter system that is drug inducible. For example, a tamoxifen-induced Cre transgene in which expression is driven by the regulatory elements of a putative SSC-specific marker will lead to irreversible activation of a reporter transgene in cells expressing the marker, and all daughter cells forming from the induced cell also inherit expression of the label. As such, this technique has been used in conjunction with live imaging to monitor the formation of chains of cells from a single spermatogonia, and further, the presumptive fragmentation of single cells from spermatogonial chains (Nakagawa et al. 2007, 2010, Hara et al. 2014). This technique has also been used to infer specificity of a marker of interest to the SSC population by tracking the presence of fluorescent cells in the testes at long-term intervals (months to a year) following tamoxifen labeling (Hara et al. 2014, Sun et al. 2015). However, only a portion of a marker expressing cell population appears to be amenable to labeling after systemic treatment with a drug inducer (Nakagawa et al. 2007, Sun et al. 2015), thus quantitative assessment of the portion of the population in question that possess SSC capacity is challenging. Also, a potentially unrealized nuance of tamoxifen-induced labeling is the effect that the drug may have as a stressor to the system (Verma & Krishna 2017), which could actually lead to a non-steady-state condition after administration.
Markers of the SSC pool
In order to study the dynamics of the SSC population; either via direct monitoring using lineage tracing or retrospective assessment via transplantation analyses, it is clearly integral to have markers that can accurately identify and distinguish these stem cells from their immediate undifferentiated progenitor counterparts. Unfortunately, a consensus has not been reached in the field as to the specificity of purported SSC markers that have been used to study the population dynamics. Certainly, the development of contesting models across the field may have arisen from the assessment of different undifferentiated subpopulations that were considered at the time to be pure SSCs but later found to be heterogeneous populations. The primary markers used to formulate the ‘fragmentation’ and ‘revised Asingle’ models of SSC dynamics are GDNF family receptor alpha 1 (Gfra1) and neurogenin3 (Ngn3) and inhibitor of DNA binding 4 (Id4) respectively. The evidence for specificity of these markers specifically is discussed below.
GFRA1 is part of the receptor complex for the growth factor GDNF that is well known to influence SSC maintenance (Meng et al. 2000, Kubota et al. 2004, Chen et al. 2016). However, whether expression is SSC-specific is arguable because the outcomes of multiple studies suggest that expression traverses into the progenitor population. Outcomes of immunohistochemical and gene expression analyses suggest that GFRA1 expression is present in all single and paired, and even some aligned undifferentiated spermatogonia (Ebata et al. 2005, Grasso et al. 2012). Although this does not conclusively suggest that GFRA1 is a pan-undifferentiated marker, transplantation analysis of the sorted GFRA1+ spermatogonial population has revealed little to no enrichment in the number of regenerative cells when compared to an unsorted spermatogonial population (~2.5-fold increase in pups, and a decrease in adults) (Buageaw et al. 2005, Ebata et al. 2005, Grisanti et al. 2009). Alternatively, lineage tracing analyses have demonstrated that derivatives of GFRA1+ labeled cells do persist in the mouse testis for up to a year following tamoxifen injection, suggesting that at least a portion of this population are true SSCs (Hara et al. 2014), but labeling of the entire population for quantitative assessment is not possible using this approach. Although it has been suggested that the inconsistencies between transplantation and lineage tracing data could be a consequence of sensitivity of the GFRA1 surface receptor to the enzyme treatments used prior to transplantation, primary cultures of undifferentiated spermatogonia are regularly exposed to trypsin at the time of sub-culturing but remain responsive to exogenous GDNF (Oatley et al. 2006), thus suggesting that this is not the case. Moreover, GFRA1+ spermatogonia isolated from testes using enzymatic treatments respond to GDNF exposure, thus demonstrating that at least some of the cells still possess an intact receptor complex (Hofmann et al. 2005). Together, these data infer that although a percentage of the GFRA1+ population is indeed SSCs, this population also consists of progenitors that lack regenerative (i.e. stem cell) capacity. In contrast to GFRA1, NGN3 is thought to be expressed primarily by progenitors, where it interacts with the transcription factor STAT3 to regulate differentiation (Kaucher et al. 2012). Indeed, it has been demonstrated that NGN3+ spermatogonia have a greater propensity for RA-induced differentiation than their NGN3- counterparts (Nakagawa et al. 2010). Despite this, it appears that at least a subset of NGN3 expressing cells are SSCs, as previous studies demonstrated self-renewing properties in NGN3-expressing spermatogonia using lineage tracing analyses (Yoshida et al. 2004). Analysis of SSC dynamics by a single research group using these markers has led to the formulation of the so-called ‘fragmentation model’ to explain SSC dynamics (Hara et al. 2014). At present, unequivocal functional evidence that GFRA1 and NGN3 are specific for SSCs and progenitors respectively is lacking. Thus, it is important to factor in this nuance when deriving independent interpretations of whether the findings support a model in which fragmentation of chained spermatogonia produces cells with a functional capacity to behave as SSCs.
The basis for development of the revised Asingle model of SSC dynamics is the monitoring of the ID4-expressing spermatogonial population. ID4 is a transcription factor that our laboratory has identified to be heterogeneously expressed within the undifferentiated spermatogonial population, primarily within Asingle cells (Oatley et al. 2011a, Chan et al. 2014, Sun et al. 2015, Helsel et al. 2017). Using a mouse line containing an Id4-eGfp transgene, along with spermatogonial transplantation, we have demonstrated that the ID4+ population of primary spermatogonial cultures encompasses the veritable entirety (>95%) of the SSC population (Chan et al. 2014). Further, Sun and coworkers (Sun et al. 2015) used lineage tracing techniques to independently establish that at least a portion of the ID4-expressing spermatogonial population possess long-term self-renewal capacity in vivo, in both neonatal and adult testes. Although the ID4+ spermatogonial population as a whole is highly enriched for SSCs, it is important to note that heterogeneity exists even within this pool of cells; as depicted by a spectrum of ID4-eGFP fluorescence (bright to dim) (Chan et al. 2014, Helsel et al. 2017), and the differential expression of factors such as TSPAN8 (Mutoji et al. 2016). The significance of heterogeneity in terms of levels of ID4 expression will be discussed briefly in the context of the Asingle model. However, it is important to note that in using a LD spermatogonial transplantation assay on post-natal day 6–8 spermatogonia, we have established that 1 in 0.94 cells within the subset of the population that contains the highest levels of ID4 (i.e. ID4-eGFPBright) is an SSC. Thus, the population is highly enriched for, if not pure, stem cells (Helsel et al. 2017). Although lineage tracing data are suggestive that the ID4-expressing spermatogonial population also encompasses SSCs in the adult testes (Sun et al. 2015), similar assessments of population purity using transplantation analyses have not yet been conducted.
Models of SSC dynamics
Traditional Asingle model
For several decades, the ‘Asingle model’ that was devised from experimentation in both mice and rats has served as the predominant explanation of SSC dynamics in mammalian testes (Fig. 1) (de Rooij 1969, Huckins 1971, Oakberg 1971, de Rooij 1973). These experiments used morphological assessment of whole mount seminiferous tubules and 3H-thymidine tracing to propose that stem cell activity is restricted to the Asingle spermatogonia in the niche (Oakberg 1971). As mentioned previously, these Asingle spermatogonia can undergo mitotic division to produce new single cells, or an Apaired syncytium in which the cells remain connected by an intercellular bridge. The Asingle model predicts that the incomplete abscission of an intercellular bridge following mitotic division is an indication of irreversible transition to a progenitor state. Accordingly, these studies stated that there was no evidence for reversion of chains of spermatogonia back to the Asingle cell type. The Asingle model of SSC dynamics has provided the basis for several decades of research, however the capacity for experimentation on the SSC population has now far exceeded dependence on morphological assessments. As such, more modern models of SSC dynamics build upon, and in some cases contradict the original hypothesis. As will be discussed below, both of the recently developed models of SSC dynamics have used fluorescent reporter transgenes under the control of putative SSC-specific markers to challenge the notion that the entirety of the Asingle population is synonymous with the SSC population, and further, that reversion into a self-renewing cell type does not occur once steps toward differentiation have been initiated.
Different models to explain the dynamics of the SSC pool. The traditional model proposes that all Asingle spermatogonia comprise the SSC pool. These Asingle cells can either self-renew, or transition to a progenitor state; as represented by the formation of an Apaired syncytium. Apaired progenitors undergo further mitotic divisions to produce longer chains of spermatogonia (Aaligned) that will respond to retinoic acid (RA) by transitioning to a differentiating state. In the traditional Asingle model, formation of Apaired signifies irreversible commitment to a differentiating pathway. In contrast to the Asingle model, the fragmentation model proposes that the self-renewing SSC pool is comprised of interchanging Asingle, Apaired and Aaligned spermatogonia, all of which express GFRA1. In this model, the transition to progenitor is signified by loss of GFRA1 expression and gain of NGN3 expression. Finally, the revised Asingle model proposes that a specified subset of the Asingle pool (i.e. cells with high levels of ID4 expression) represent the SSCultimate population that serve as the self-renewing reservoir in steady-state conditions. From the SSCultimate pool, transitory cells (i.e. spermatogonia with lower levels of ID4) arise. The continuum of transition to a progenitor state involves declining expression of key self-renewal genes in SSCtransitory that will allow for the next mitotic division to result in a true Apaired syncytium. In all models, RA drives Apaired and Aaligned progenitors into a differentiating state at periodic intervals, characterized by the attainment of Kit expression. Following this, spermatogonia transition through A1, A2, A3, A4, intermediate and type B stages, before entering into meiosis as spermatocytes, then producing haploid spermatids and finally spermatozoa.
Citation: Reproduction 154, 2; 10.1530/REP-17-0034
Fragmentation model
Based on the outcomes of lineage tracing and live imaging analyses using mice containing the aforementioned GFRA1 and NGN3 reporter transgenes, the ‘fragmentation model’ of SSC dynamics has emerged over the last 10 years (Fig. 1) (Nakagawa et al. 2007, 2010, Hara et al. 2014). This model predicts that spermatogonia are constantly ‘breaking off’ from Apaired and Aaligned chains to replenish the self-renewing SSC pool (Hara et al. 2014); a direct contradiction to traditional schools of thought that depicted these chains of cells as progenitors that were irreversibly committed to a differentiating fate (de Rooij & Kramer 1968, de Rooij 1969, 1973, Huckins 1971, Oakberg 1971, De Rooij 1988, De Rooij et al. 1989, de Rooij & Grootegoed 1998, de Rooij & Russell 2000). Thus, the fragmentation model proposes that the primary stem cell pool, believed to be encompassed by all GFRA1 expressing spermatogonia, is not only comprised solely of Asingle cells, but also of pairs and short spermatogonial chains. In this paradigm, the GFRA1+ spermatogonia can alternatively transition into a GFRA1-/NGN3+ progenitor state, in which the propensity for differentiation is increased. Furthermore, it has been proposed that in times when regeneration is pertinent; the NGN3+ progenitor population can also experience fragmentation, with single progenitor spermatogonia breaking off from chains and reverting to a GFRA1+ state to re-join the self-renewing pool (Nakagawa et al. 2010). Indeed, recent independent research has suggested a capacity for NGN3+ cells to presumably revert to a GFRA1+ state (Zhang et al. 2016), a process that was only observed during circumstances of regeneration, and appeared to be reliant on the presence of Doublesex and Mab-3-related transcription factor 1 (DMRT1). Considering that NGN3 is expressed by subsets of Asingle, Apaired and Aaligned spermatogonia, these findings do not prove unequivocally that fragmentation is the mode by which new GFRA1+ cells are generated. The possibility that NGN3+/GFRA1− Asingle cells revert to a GFRA1+ state has not been ruled out.
Although the propensity for alteration or reversion of spermatogonial expression profiles in response to different requirements within the niche is a concept that is becoming more readily accepted; the notion that these reverting cells are ‘fragments’ broken off from spermatogonial chains is controversial. Currently, evidence for these phenomena are based primarily on observations of fragmentation using live imaging of testes in mice that are maintained in a stress condition of long-term anesthesia, and computer generated biophysical models (Nakagawa et al. 2007, 2010, Hara et al. 2014). However, as yet, no direct functional evidence has been provided for re-established self-renewal capacity in the fragmented cells. Further qualms underlying the fragmentation model rest with the aforementioned uncertainty surrounding the specificity of GFRA1 and NGN3 as markers for the SSC and progenitor populations respectively. Moreover, this model of SSC dynamics is not easily harmonized with previously published data that has demonstrated that the propensity for RA-induced differentiation is correlated with spermatogonial chain length; i.e. 100% of Aaligned chains consisting of 8–16 spermatogonia, 94% of Aaligned chains containing four spermatogonia, 64% of Apaired spermatogonia and 0% of Asingle spermatogonia differentiate in response to RA (Tagelenbosch & de Rooij 1993); findings that suggest a more linear relationship between the stem cell state and spermatogonial chain length. Furthermore, the computational models generated based on live imaging and lineage tracing analysis do not take into account the stages of the seminiferous cycle in which the undifferentiated-to-differentiating transition occur and different cohorts of spermatogonia are known to randomly divide.
Revised Asingle model to explain SSC dynamics
Based on the analysis of the ID4+ spermatogonial population as well as recent findings using other markers for a subset of Asingle with stem cell properties, we propose a revised version of the traditional model to describe SSC dynamics in rodents (Fig. 1). This paradigm posits that a subset of the Asingle population are functionally true SSCs, and further suggests that some plasticity may exist between the self-renewing SSC pool and spermatogonia that are in a transitioning state between SSC and progenitor. Early evidence for heterogeneity within the Asingle population came from transplantation analyses of single cell suspensions from whole testes. Although the number of Asingle cells in the adult mouse testis is estimated to be approximately 35,000 (Tagelenbosch & de Rooij 1993), the number of regenerated spermatogenic colonies in recipient testes following transplantation of an unselected total donor testis cell population is less than 10% of this value, at approximately 3000 cells (Nagano 2003). Further supporting the notion that not all Asingle cells are SSCs, a number of research groups have reported heterogeneity in expression of various genes within the Asingle population; including Id4 (Oatley et al. 2011a, Chan et al. 2014, Hermann et al. 2015), Paired box 7 (Pax7) (Aloisio et al. 2014), Gfra1 (Grisanti et al. 2009), Nanos2 (Suzuki et al. 2009), Ngn3 (Yoshida et al. 2004, Kaucher et al. 2012) and Oct4 (Ohmura et al. 2004).
The revised Asingle model proposes that in steady-state conditions, a subset of the Asingle population that expresses high levels of self-renewal factors (e.g. ID4) are the foundation for continuity of the spermatogenic lineage, essentially serving as ultimate SSCs (SSCultimate). In addition to self-renewal, these SSCultimate can also give rise to SSCtransitory (Asingle and Apaired) that possess lower levels of expression for self-renewal factors and are set on a continuum of transition to a progenitor state as true Apaired and Aaligned. Thus, the differentiation from SSC to progenitor is not a ‘light-switch’ phenomenon, but rather a gradual transition between two functionally distinct extremes. Evidence supporting this model was generated first using whole mount imaging of live seminiferous tubules to show that spermatogonia with high levels of ID4 expression (i.e. ID4-eGFPBright cells) are primarily Asingle spermatogonia within the testes, and further using transplantation analysis to demonstrate that these ID4-eGFP bright cells encompass over 85% of the SSC population. In contrast, ID4-eGFP ‘dim’ cells were identified primarily as Apaired and some Asingle cells, and encompassed less than 15% of the self-renewing population when assessed via transplantation analyses (Helsel et al. 2017). Although a small percentage of Apaired syncytium were identified with bright ID4-eGFP epifluorescence, in our opinion these are likely ‘false pairs’ that arise temporarily in the process of self-renewal when abscission of the intercellular bridge has not been resolved or abscission has completed but migration apart has not occurred (de Rooij & Russell 2000). Importantly, Aaligned cohorts appear to possess low to undetectable levels of ID4-eGFP (Helsel et al. 2017), and we have demonstrated previously that the ID4-eGFP negative population obtained from primary spermatogonial culture has little capacity for regeneration. These findings suggest that Apaired and Aaligned, with low to no expression of specific self-renewal factors such as ID4, represent the pool of undifferentiated progenitors. These experiments not only demonstrated a direct correlation between ID4 expression, spermatogonial ‘chain identity’ and self-renewal capacity; but have used transcriptome profiling to demonstrate that distinct gene expression signatures exist between the SSCultimate, SSCtransitory and progenitor populations (Chan et al. 2014, Helsel et al. 2017). Notably, independent evidence for the revised Asingle model has also recently come to light following characterization of expression of the Wnt/β-catenin signaling inhibitor ‘SHISA6’ in the spermatogonial population. As demonstrated by the research group of Dr Yoshida (Tokue et al. 2017), SHISA6 acts to maintain the self-renewing pool by impeding the differentiating transition, however its expression is restricted to all but a small subset of the GFRA1+ population. Indeed, SHISA6-expressing cells were primarily identified to be Asingle. We propose that these SHISA6 expressing cells are likely to be the SSCultimate that also express high levels of ID4, while the GFRA1+/SHISA6− cells represent the SSCtransitory and progenitor populations. While further investigation into the association between ID4 and SHISA6 is warranted, our recently published transcriptome profiling studies suggest that Shisa6 is uniquely expressed by the ID4-eGFPBright population that is highly enriched, if not pure, SSCs with the transcript levels being 8-fold higher compared to the ID4-eGFPDim population that is in transition to become progenitors that lack stem cell capacity (Helsel et al. 2017).
The revised Asingle model agrees with the traditional model in that reversion to a stem cell state from the progenitor pool does not occur in steady-state conditions or during situations of regeneration caused by external insults. However, this model does support the notion that some plasticity may exist for cells in the beginning stages of transition from the ultimate SSC state to the progenitor state (i.e. SSCtransitory spermatogonia). Indeed, a recent independent study demonstrated that NGN3 expressing spermatogonia that are rarely found to express high levels of ID4 in steady-state conditions may be able to re-establish the stem cell expression profile and contribute to the self-renewing pool when regeneration of the germline is required, such as following cytotoxic insult (Zhang et al. 2016). Based on the lack of regenerative capacity in the ID4-eGFP negative population (as assessed by transplantation analysis using populations derived from primary spermatogonial culture), and their transcriptome profile (Chan et al. 2014), along with the relationship between spermatogonial chain identity and RA-induced differentiation (Tagelenbosch & de Rooij 1993); we propose that the source of these reverted regenerative cells are likely Asingle but potentially also the Apaired that are in the SSCtransitory state.
What makes an SSC an SSC? – speculations based on the revised Asingle model
Based on transcriptome comparisons between subpopulations of SSCs and progenitors, one can begin to form a picture as to how the self-renewing reservoir is maintained in the testes during steady-state conditions, and the changes that occur upon transition to, and formation of, a progenitor state. First, the SSCultimate possess high expression levels of key self-renewal factors including GFRA1, and, likely as a direct consequence of this, are also characterized by high levels of expression of a number of GDNF responsive genes that are known to be involved in self-renewal; including Bcl6b, Lhx1, Etv5 and Id4 itself (Chan et al. 2014, Helsel et al. 2017). Although genes involved in self-renewal are upregulated in these SSCultimate cells, expression of genes that are associated with a progenitor state and elevated propensity for differentiation such as Sohlh1, Ngn3 and Kit, are repressed (Chan et al. 2014, Helsel et al. 2017). In addition to promoting self-renewal and preventing differentiation, the gene expression profile of the SSCultimate also appears to be integral for maintaining a ‘slow-cycling’ state; suggested by the rare incorporation of EdU (Zhang et al. 2016) that is typical of stem cells in many lineages (Fuchs 2009).
As the transition to progenitor begins at the SSCtransitory state, levels of self-renewal genes (e.g. Id4 and Gfra1) are reduced, concomitant with increased expression of genes controlling the progenitor state (e.g. Sohlh1 and Ngn3) (Helsel et al. 2017). Such findings suggest that GFRA1 expression, like ID4, may exist as a spectrum in which SSCultimate are GFRA1-high and SSCtransitory are GFRA1-low. The heterogeneity in GFRA1 expression between these two populations would again lend to the notion that results produced from studies using GFRA1-based reporter transgenes to follow SSC dynamics may be confounded by combining two functionally distinct subpopulations (Hara et al. 2014). In conjunction with the decreased levels of ID4 expression, EdU incorporation begins to become more frequent upon initiation of the progenitor transition, indicating more rapid cycling of these cells (Zhang et al. 2016). Although we have demonstrated that progenitor spermatogonia, which have lost ID4-eGFP expression, exist as chains of spermatogonia (Aaligned) with expression profiles distinct from that of the SSC population (Chan et al. 2014, Helsel et al. 2017); factors that act specifically to regulate the maintenance of intercellular bridges, or abscission following a self-renewing division in SSCs, remain uncharacterized. Testis expressed 14 (TEX14) has been identified as indispensable for the formation and maintenance of intercellular bridges in mice (Greenbaum et al. 2006, 2011), however, any differential expression between SSC and progenitor populations is yet to be reported.
In conclusion, although progress in characterizing SSC dynamics has certainly been achieved since derivation of the Asingle model in the 1970s, the field has yet to come to agreement on a single paradigm for this process. These inconsistencies are at least in part a result of utilizing different markers for the SSC population, as well as the different strategies used to assess SSC specificity. Despite opposing theories on which cells within the undifferentiated population actually encompass the self-renewing pool, the revised Asingle and fragmentation models do agree that simply being a ‘single’ undifferentiated spermatogonia is not equivalent to being an SSC; in other words, the Asingle population is heterogeneous in terms of marker gene expression and functional capacity. We propose that in steady-state conditions, a subset of the Asingle population; distinguished by high levels of ID4 expression, act as the self-renewing reservoir that maintains continual sperm production (SSCultimate). In the continuum of adopting a true progenitor state, Asingle pass through a transition state (SSCtransitory) where expression of self-renewal driving genes declines, and expression of genes driving differentiation increases. Thus, at the next cell division, we envision that the SSCtransitory adopt a true progenitor state at which they are responsive to RA signaling. However, in circumstances where germline regeneration is pertinent, it is possible that these transitioning cells can respond to uncharacterized extrinsic signals to revert to the self-renewing pool rather than transition to a progenitor state at the next mitotic division, thereby providing a means for rapid restoration of the spermatogenic lineage.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this review.
Funding
This work was supported by grant HD061665 awarded to J M O form the Eunice Kennedy Shriver National Institutes of Child Health and Human Development.
References
Aloisio GM, Nakada Y, Saatcioglu HD, Pena CG, Baker MD, Tarnawa ED, Mukherjee J, Manjunath H, Bugde A & Sengupta AL et al. 2014 PAX7 expression defines germline stem cells in the adult testis. Journal of Clinical Investigation 124 3929–3944. (doi:10.1172/JCI75943)
Brinster RL & Avarbock MR 1994 Germline transmission of donor haplotype following spermatogonial transplantation. PNAS 91 11303–11307. (doi:10.1073/pnas.91.24.11303)
Brinster RL & Zimmermann JW 1994 Spermatogenesis following male germ-cell transplantation. PNAS 91 11298–11302. (doi:10.1073/pnas.91.24.11298)
Buageaw A, Sukhwani M, Ben-Yehudah A, Ehmcke J, Rawe VY, Pholpramool C, Orwig KE & Schlatt S 2005 GDNF family receptor alpha1 phenotype of spermatogonial stem cells in immature mouse testes. Biology of Reproduction 73 1011–1016. (doi:10.1095/biolreprod.105.043810)
Busada JT, Chappell VA, Niedenberger BA, Kaye EP, Keiper BD, Hogarth CA & Geyer CB 2015 Retinoic acid regulates Kit translation during spermatogonial differentiation in the mouse. Developmental Biology 397 140–149. (doi:10.1016/j.ydbio.2014.10.020)
Chan F, Oatley MJ, Kaucher AV, Yang Q-E, Bieberich CJ, Shashikant CS & Oatley JM 2014 Functional and molecular features of the Id4+ germline stem cell population in mouse testes. Genes and Development 28 1351–1362. (doi:10.1101/gad.240465.114)
Chen S-R & Liu Y-X 2015 Regulation of spermatogonial stem cell self-renewal and spermatocyte meiosis by Sertoli cell signaling. Reproduction 149 R159–R167. (doi:10.1530/REP-14-0481)
Chen L-Y, Willis WD & Eddy EM 2016 Targeting the Gdnf Gene in peritubular myoid cells disrupts undifferentiated spermatogonial cell development. PNAS 113 1829–1834. (doi:10.1073/pnas.1517994113)
Clermont Y 1966 Renewal of spermatogonia in man. American Journal of Anatomy 118 509–524. (doi:10.1002/aja.1001180211)
Clermont Y 1969 Two classes of spermatogonial stem cells in the monkey (Cercopithecus aethiops). American Journal of Anatomy 126 57–71. (doi:10.1002/aja.1001260106)
Clermont Y & Antar M 1973 Duration of the cycle of the seminiferous epithelium and the spermatogonial renewal in the monkey Macaca arctoides. American Journal of Anatomy 136 153–165. (doi:10.1002/aja.1001360204)
de Rooij DG 1969 Further evidence for the proposed way of spermatogonial stem cell renewal in the rat and the mouse. Zeitschrift fur Zellforschung und Mikroskopische Anatomie 99 134–138. (doi:10.1007/BF00338803)
de Rooij DG 1973 Spermatogonial stem cell renewal in the mouse. I. Normal situation. Cell and Tissue Kinetics 6 281–287. (doi:10.1111/j.1365-2184.1973.tb01617.x)
de Rooij DG & Kramer MF 1968 Spermatogonial stemcell renewal in rats and mice. Zeitschrift fur Zellforschung und Mikroskopische Anatomie 85 206–209. (doi:10.1007/BF00325035)
de Rooij DG & Grootegoed JA 1998 Spermatogonial stem cells. Current Opinion in Cell Biology 10 694–701. (doi:10.1016/S0955-0674(98)80109-9)
de Rooij DG & Russell LD 2000 All you wanted to know about spermatogonia but were afraid to ask. Journal of Andrology 21 776–798.
De Rooij DG 1988 Regulation of the proliferation of spermatogonial stem cells. Journal of Cell Science: Supplement 10 181–194. (doi:10.1242/jcs.1988.supplement_10.14)
De Rooij DG, Van Dissel-Emiliani FM & Van Pelt AM 1989 Regulation of spermatogonial proliferation. Annals of the New York Academy of Sciences 564 140–153. (doi:10.1111/j.1749-6632.1989.tb25894.x)
Dobrinski I, Ogawa T, Avarbock MR & Brinster RL 1999 Computer assisted image analysis to assess colonization of recipient seminiferous tubules by spermatogonial stem cells from transgenic donor mice. Molecular Reproduction and Development 53 142–148. (doi:10.1002/(SICI)1098-2795(199906)53:2<142::AID-MRD3>3.0.CO;2-O)
Ebata KT, Zhang X & Nagano MC 2005 Expression patterns of cell-surface molecules on male germ line stem cells during postnatal mouse development. Molecular Reproduction and Development 72 171–181. (doi:10.1002/mrd.20324)
Fuchs E 2009 The tortoise and the hair: slow-cycling cells in the stem cell race. Cell 137 811–819. (doi:10.1016/j.cell.2009.05.002)
Grasso M, Fuso A, Dovere L, de Rooij DG, Stefanini M, Boitani C & Vicini E 2012 Distribution of GFRA1-expressing spermatogonia in adult mouse testis. Reproduction 143 325–332. (doi:10.1530/REP-11-0385)
Greenbaum MP, Yan W, Wu MH, Lin YN, Agno JE, Sharma M, Braun RE, Rajkovic A & Matzuk MM 2006 TEX14 is essential for intercellular bridges and fertility in male mice. PNAS 103 4982–4987. (doi:10.1073/pnas.0505123103)
Greenbaum MP, Iwamori T, Buchold GM & Matzuk MM 2011 Germ cell intercellular bridges. Cold Spring Harbor Perspectives in Biology 3 a005850. (doi:10.1101/cshperspect.a005850)
Grisanti L, Falciatori I, Grasso M, Dovere L, Fera S, Muciaccia B, Fuso A, Berno V, Boitani C & Stefanini M et al. 2009 Identification of spermatogonial stem cell subsets by morphological analysis and prospective isolation. Stem Cells 27 3043–3052. (doi:10.1002/stem.206)
Hamra FK, Chapman KM, Nguyen DM, Williams-Stephens AA, Hammer RE & Garbers DL 2005 Self renewal, expansion, and transfection of rat spermatogonial stem cells in culture. PNAS 102 17430–17435. (doi:10.1073/pnas.0508780102)
Hara K, Nakagawa T, Enomoto H, Suzuki M, Yamamoto M, Simons Benjamin D & Yoshida S 2014 Mouse spermatogenic stem cells continually interconvert between equipotent singly isolated and syncytial states. Cell Stem Cell 14 658–672. (doi:10.1016/j.stem.2014.01.019)
Helsel AR & Oatley JM 2017. Transplantation as a quantitative assay to study mammalian male germline stem cells. Methods in Molecular Biology 1463 155–172. (doi:10.1007/978-1-4939-4017-2_12)
Helsel AR, Yang Q, Oatley MJ, Lord T, Sablitzky F & Oatley JM 2017 ID4 levels dictate the stem cell state in mouse spermatogonia. Development 144 624–634. (doi:10.1242/dev.146928)
Hermann BP, Mutoji KN, Velte EK, Ko D, Oatley JM, Geyer CB & McCarrey JR 2015 Transcriptional and translational heterogeneity among neonatal mouse spermatogonia. Biology of Reproduction 92 54. (doi:10.1095/biolreprod.114.125757)
Hofmann MC, Braydich-Stolle L & Dym M 2005 Isolation of male germ-line stem cells; influence of GDNF. Developmental Biology 279 114–124. (doi:10.1016/j.ydbio.2004.12.006)
Huckins C 1971 The spermatogonial stem cell population in adult rats. Cell Proliferation 4 335–349. (doi:10.1111/j.1365-2184.1971.tb01544.x)
Ikami K, Tokue M, Sugimoto R, Noda C, Kobayashi S, Hara K & Yoshida S 2015 Hierarchical differentiation competence in response to retinoic acid ensures stem cell maintenance during mouse spermatogenesis. Development 142 1582–1592. (doi:10.1242/dev.118695)
Kanatsu-Shinohara M, Ogonuki 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. (doi:10.1095/biolreprod.103.017012)
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. (doi:10.1095/biolreprod.104.036400)
Kanatsu-Shinohara M, Inoue K, Miki H, Ogonuki N, Takehashi M, Morimoto T, Ogura A & Shinohara T 2006 Clonal origin of germ cell colonies after spermatogonial transplantation in mice. Biology of Reproduction 75 68–74. (doi:10.1095/biolreprod.106.051193)
Kaucher AV, Oatley MJ & Oatley JM 2012 NEUROG3 is a critical downstream effector for STAT3-regulated differentiation of mammalian stem and progenitor spermatogonia. Biology of Reproduction 86 164. (doi:10.1095/biolreprod.111.097386)
Kubota H, Avarbock MR & Brinster RL 2004 Growth factors essential for self-renewal and expansion of mouse spermatogonial stem cells. PNAS 101 16489–16494. (doi:10.1073/pnas.0407063101)
Kubota H, Wu X, Goodyear SM, Avarbock MR & Brinster RL 2011 Glial cell line-derived neurotrophic factor and endothelial cells promote self-renewal of rabbit germ cells with spermatogonial stem cell properties. FASEB Journal 25 2604–2614. (doi:10.1096/fj.10-175802)
Meng X, Lindahl M, Hyvönen ME, Parvinen M, de Rooij DG, Hess MW, Raatikainen-Ahokas A, Sainio K, Rauvala H & Lakso M et al. 2000 Regulation of cell fate decision of undifferentiated spermatogonia by GDNF. Science 287 1489–1493. (doi:10.1126/science.287.5457.1489)
Mutoji K, Singh A, Nguyen T, Gildersleeve H, Kaucher AV, Oatley MJ, Oatley JM, Velte EK, Geyer CB & Cheng K et al. 2016 TSPAN8 expression distinguishes spermatogonial stem cells in the prepubertal mouse testis. Biology of Reproduction 95 117. (doi:10.1095/biolreprod.116.144220)
Nagano MC 2003 Homing efficiency and proliferation kinetics of male germ line stem cells following transplantation in mice. Biology of Reproduction 69 701–707. (doi:10.1095/biolreprod.103.016352)
Nagano M, Avarbock MR & Brinster RL 1999 Pattern and kinetics of mouse donor spermatogonial stem cell colonization in recipient testes. Biology of Reproduction 60 1429–1436. (doi:10.1095/biolreprod60.6.1429)
Nakagawa T, Nabeshima Y & Yoshida S 2007 Functional identification of the actual and potential stem cell compartments in mouse spermatogenesis. Developmental Cell 12 195–206. (doi:10.1016/j.devcel.2007.01.002)
Nakagawa T, Sharma M, Nabeshima Y, Braun RE & Yoshida S 2010 Functional hierarchy and reversibility within the murine spermatogenic stem cell compartment. Science 328 62–67. (doi:10.1126/science.1182868)
Oakberg EF 1971 Spermatogonial stem-cell renewal in the mouse. Anatomical Record 169 515–531. (doi:10.1002/ar.1091690305)
Oatley JM & Brinster RL 2008 Regulation of spermatogonial stem cell self-renewal in mammals. Annual Review of Cell and Developmental Biology 24 263–286. (doi:10.1146/annurev.cellbio.24.110707.175355)
Oatley JM & Brinster RL 2012 The germline stem cell niche unit in mammalian testes. Physiological Reviews 92 577–595. (doi:10.1152/physrev.00025.2011)
Oatley JM, Avarbock MR, Telaranta AI, Fearon DT & Brinster RL 2006 Identifying genes important for spermatogonial stem cell self-renewal and survival. PNAS 103 9524–9529. (doi:10.1073/pnas.0603332103)
Oatley JM, Oatley MJ, Avarbock MR, Tobias JW & Brinster RL 2009 Colony stimulating factor 1 is an extrinsic stimulator of mouse spermatogonial stem cell self-renewal. Development 136 1191–1199. (doi:10.1242/dev.032243)
Oatley MJ, Kaucher AV, Racicot KE & Oatley JM 2011a Inhibitor of DNA binding 4 is expressed selectively by single spermatogonia in the male germline and regulates the self-renewal of spermatogonial stem cells in mice. Biology of Reproduction 85 347–356. (doi:10.1095/biolreprod.111.091330)
Oatley MJ, Racicot KE & Oatley JM 2011b Sertoli cells dictate spermatogonial stem cell niches in the mouse testis. Biology of Reproduction 84 639–645. (doi:10.1095/biolreprod.110.087320)
Ogawa T, Ohmura M, Yumura Y, Sawada H & Kubota Y 2003 Expansion of murine spermatogonial stem cells through serial transplantation. Biology of Reproduction 68 316–322. (doi:10.1095/biolreprod.102.004549)
Ohmura M, Yoshida S, Ide Y, Nagamatsu G, Suda T & Ohbo K 2004 Spatial analysis of germ stem cell development in Oct-4/EGFP transgenic mice. Archives of Histology and Cytology 67 285–296. (doi:10.1679/aohc.67.285)
Piquet-Pellorce C, Dorval-Coiffec I, Pham M & Jégou B 2000 Leukemia inhibitory factor expression and regulation within the testis. Endocrinology 141 1136–1141. (doi:10.1210/endo.141.3.7399)
Ryu BY, Kubota H, Avarbock MR & Brinster RL 2005 Conservation of spermatogonial stem cell self-renewal signaling between mouse and rat. PNAS 102 14302–14307. (doi:10.1073/pnas.0506970102)
Sun F, Xu Q, Zhao D & Degui Chen C 2015 Id4 marks spermatogonial stem cells in the mouse testis. Scientific Reports 5 17594. (doi:10.1038/srep17594)
Suzuki H, Sada A, Yoshida S & Saga Y 2009 The heterogeneity of spermatogonia is revealed by their topology and expression of marker proteins including the germ cell-specific proteins Nanos2 and Nanos3. Developmental Biology 336 222–231. (doi:10.1016/j.ydbio.2009.10.002)
Tagelenbosch RAJ & de Rooij DG 1993 A quantitative study of spermatogonial multiplication and stem cell renewal in the C3H/101 F1 hybrid mouse. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 290 193–200. (doi:10.1016/0027-5107(93)90159-D)
Tokue M, Ikami K, Mizuno S, Takagi C, Miyagi A, Takada R, Noda C, Kitadate Y, Hara K & Mizuguchi H et al. 2017 SHISA6 confers resistance to differentiation-promoting Wnt/β-catenin signaling in mouse spermatogenic stem cells. Stem Cell Reports 8 561–575. (doi:10.1016/j.stemcr.2017.01.006)
Verma R & Krishna A 2017 Effect of tamoxifen on spermatogenesis and testicular steroidogenesis. Biochemical and Biophysical Research Communications 486 36–42. (doi:10.1016/j.bbrc.2017.02.092)
Wu Z, Falciatori I, Molyneux LA, Richardson TE, Chapman KM & Hamra FK 2009 Spermatogonial culture medium: an effective and efficient nutrient mixture for culturing rat spermatogonial stem cells. Biology of Reproduction 81 77–86. (doi:10.1095/biolreprod.108.072645)
Yoshida S, Takakura A, Ohbo K, Abe K, Wakabayashi J, Yamamoto M, Suda T & Nabeshima Y 2004 Neurogenin3 delineates the earliest stages of spermatogenesis in the mouse testis. Developmental Biology 269 447–458. (doi:10.1016/j.ydbio.2004.01.036)
Zhang X, Ebata KT & Nagano MC 2003 Genetic analysis of the clonal origin of regenerating mouse spermatogenesis following transplantation. Biology of Reproduction 69 1872–1878. (doi:10.1095/biolreprod.103.019273)
Zhang T, Oatley J, Bardwell VJ & Zarkower D 2016 DMRT1 is required for mouse spermatogonial stem cell maintenance and replenishment. PLoS Genetics 12 e1006293. (doi:10.1371/journal.pgen.1006293)
Zhou Q, Li Y, Nie R, Friel P, Mitchell D, Evanoff RM, Pouchnik D, Banasik B, McCarrey JR & Small C et al. 2008 Expression of stimulated by retinoic acid gene 8 (Stra8) and maturation of murine gonocytes and spermatogonia induced by retinoic acid in vitro. Biology of Reproduction 78 537–545. (doi:10.1095/biolreprod.107.064337)
