In mammalian spermatogenesis, a small subpopulation of As (singly isolated) and Apr (pair of two interconnected) spermatogonia within the basal compartment of the seminiferous epithelia proliferate to maintain a stem cell pool. The remaining spermatogonia give rise to a longer spermatogonial chain (Aal [chains of 4, 8, 16, and 32 cells, etc.]). These Aal spermatogonia differentiate into a larger number of advanced (A1–A4- and B-type) spermatogonia, which subsequently undergo meiosis, generating haploid spermatozoa following their transition from the basal to the adluminal compartment of the seminiferous epithelia (Russell et al., 1990; De Rooij and Russell, 2000; Wong and Cheng, 2005).
Spermatogonial stem cells (SSCs) are continuously maintained by a process of self-renewal in the basal compartment, located between the tight junctions of adjacent Sertoli cells and the continuous basal lamina of seminiferous tubules (De Rooij and Grootegoed, 1998; Brinster 2002; Oatley and Brinster, 2008; Yoshida, 2010). The SSCs are believed to comprise mostly As and Apr cells, and have an active cell-spreading shape. They are a small subset of spermatogonia that express GFRα1, a GPI-anchored receptor for GDNF (glial cell line-derived neurotrophic factor) in the healthy testis (Sada et al., 2009; Suzuki et al., 2009; Nakagawa et al., 2010). It has been shown that GDNF is essential in maintaining the balance between self-renewal and differentiation within the SSC pool (Meng et al., 2000; Naughton et al., 2006; Hofmann, 2008; Oatley and Brinster, 2008). GDNF is mainly produced by Sertoli cells and acts through a GFRα1-RET receptor complex in undifferentiated spermatogonia, which include SSCs (see the review by Sariola and Saarma, 2003). It has been demonstrated that heterozygote Gdnf mutant mice undergo SSC depletion while transgenic male mice overexpressing Gdnf accumulate undifferentiated spermatogonia in the seminiferous tubules (Meng et al., 2000). Moreover, it is likely that the role of GDNF in SSC maintenance is evolutionally conserved, as the SSCs in various mammals are maintained both in recipient mouse testes in vivo and in the presence of GDNF in vitro (Ogawa et al., 1999; Aponte et al., 2005, 2008; Oatley et al., 2005; Ryu et al., 2005; Kanatsu-Shinohara et al., 2008; Sadri-Ardekani et al., 2009). Together, these data suggest that in mammalian spermatogenesis, the dose-dependent interaction between GDNF signals and GFRα1-positive spermatogonia may regulate the size of the SSC pool in the basal compartment of the seminiferous tubules. The dynamics and proliferative patterns of GFRα1-positive spermatogonia, however, remain unclear.
The term spermatogenic “stem” cells describes undifferentiated spermatogonia that are capable of completely reestablishing spermatogenesis in the recipient testes (Brinster and Avarbock, 1994, Brinster and Zimmermann, 1994, Nagano et al., 1999, Oatley and Brinster 2006). The SSC transplantation system involves injection of a donor testis cell suspension into the seminiferous tubules of germ-cell-depleted (e.g., busulfan-treated or W/Wv mutant) recipient males. Since each individual SSC in the injected donor testicular cell suspension has the capacity to clonally form one reestablishing spermatogenic patch in the recipient seminiferous tubules (Dobrinski et al., 1999, Nagano et al., 1999; Zhang et al., 2003; Kanatsu-Shinohara et al., 2006a), this transplantation system can be used as an unequivocal detection assay of SSCs, in both a qualitative and quantitative manner, based on their regenerative capacity (Brinster, 2002, 2007; Oatley and Brinster, 2008). Previous studies have shown that the colonization process of the spermatogonial transplantation assay could be divided into the following three continuous phases: (1) During the first week following transplantation, transplanted cells are randomly distributed throughout the tubules, with some cells reaching the basal compartment; (2) From 1 week to 1 month post-transplant, donor cells in the basement membrane divide and form a monolayer network; and (3) At 1 month post-transplantation, donor cells differentiate extensively and establish a spermatogenic colony in the center of the network, but no significant changes in the number of colonized sites are detected in recipient testes at 1 to 4 months post-transplant (Nagano et al., 1999; Nagano, 2003). These findings suggest that SSC proliferation, followed by expansion and survival selection of each SSC-derived cell patch, occurs between 1 week and 1 month post-transplant in the spermatogonial transplantation assay.
In this study, we transplanted GFP-positive spermatogenic cells into germ-cell-depleted testes of recipient W/Wv male mice, collected the tubule fragments containing GFP+ cell patches at days 7–21 post-transplant, and then quantitatively examined the number of GFRα1-positive cells in each GFP+ patch by whole-mount immunohistochemistry. Here we show the appearance of several Aal-like GFRα1-positive cell aggregates within a single spermatogenic patch during early colonization. The appearance of GFRα1-positive Aal-like aggregates was positively correlated with regional, high-level expression of GDNF. These data imply that regional changes in GDNF signal levels may regulate the dynamics of GFRα1-positive Aal-like aggregates within each spermatogenic patch, resulting in further selection of survival colonies at later stages.
Temporal Changes in the Number of GFP+ Patches Containing GFRα1-Positive Cells During Early Colonization
In order to examine the proliferative patterns and kinetics of GFRα1-positive cells during early colonization, we transplanted GFP+ spermatogenic cells into recipient W/Wv testes, then retrieved the tubule fragments containing GFP+ cells at days 7, 10, 14, and 21 post-transplant. The phenotypes of GFRα1-positive cells in each GFP+ patch were then examined by whole-mount immunohistochemistry.
First, the number of GFP+ cell patches in the W/Wv testes (top row in Table 1) was quantified. At day 7 post-transplant, more than 40 GFP+ cell clusters (including As–Aalcells) per testis were observed under a fluorescence dissecting stereo-microscope. In addition to these healthy GFP+ cell clusters, however, an abundance of small GFP+ aberrant-shaped cell debris (containing presumptive degenerative cells) was also detectable at this stage. At day 10 post-transplant, the GFP+ aberrant-shaped cell debris was still, albeit rarely, observed in the recipient seminiferous tubules. At this stage, healthy GFP+ cell patches were easily distinguishable from GFP+ debris, with 22.5 ± 4.2 patches per testis. These GFP+ cell patches appeared to expand their GFP+ area along the longitudinal axis, while their number per testis was reduced to 12.5 ± 2.4 at day 14 post-transplant and 9.4 ± 2.3 at day 21 post-transplant, showing a similar level to that observed 1–2 months post-transplant under the present experimental conditions. These observations of the pattern and kinetics of donor cells at early stages of SSC colonization are similar to those in the previous report using ROSA26(LacZ+)-derived spermatogenic cell suspension (Nagano et al., 1999).
Table 1. Temporal Changes in the Number of Donor-derived (GFP+) Spermatogenic Patches Containing GFRα1-positive Cells during Early Colonization in the Spermatogonial Transplantation Assay
Day of post-transplantation
GFP+ spermatogenic cell suspension (1.0 ×108 cells/ml prepared from 1-week-old testes) was transplanted into 3-week-old recipient W/Wv males. At each stage post-transplant, the number of GFP+ patches was counted under a fluorescent stereomicroscope.
The isolated tubule fragments containing GFP+ patches were subjected to whole-mount immunostaining with anti-GFRα1 antibody. The GFP+ (green)/GFRα1-positive (red) signals were photographed (× 200) in multiple focal (Z-stack) planes. The relative ratio and total number of GFRα1-positive patches per testis at each stage were estimated. Data are represented as mean ± S.E.M.
Dunnett tests showed significant differences between day 7 and all later stages after transplantation (p<0.01).
Number of fragments examined (Number of testes used)
Estimated number of GFRα1-positive patches per testisb
Next, as many tubule fragments containing GFP+ patches were isolated as possible, pasted onto slide glasses, then stained with an anti-GFRα1 antibody. The number of GFRα1-positive cells and total number of GFP+ cells in each GFP+ patch were examined by focusing through the whole tubule depth (second row in Table 1). At day 7 post-transplant, the ratio of the number of patches containing GFRα1-positive cells relative to the total number of GFP+ patches was only 39.8% (n=40 patches), suggesting that three-fifths of GFP+ patches lacked a potential SSC at this stage. At days 10, 14, and 21 post-transplant, the ratios of the number of patches containing GFRα1-positive cells relative to the total number of GFP+ patches were 83.5% (n=62 patches), 87.0% (n=58), and 100% (n=66), respectively (second row in Table 1). The estimated number of GFRα1-positive patches per testis at days 10, 14, and 21 post-transplant was 19.2, 10.9, and 9.4 patches per testis, respectively (bottom row in Table 1). Therefore, it is estimated that approximately half of GFRα1-positive patches at day 10 post-transplant will survive to day 21 post-transplant.
Correlation between GFRα1-positive Ratio and Total Cell Number in Each GFP+ Patch
The changes in the number of GFRα1-positive cells relative to the total number of GFP+ cells in accordance with the size (total cell number) of each GFP+ patch on days 7, 10, and 14 post-transplant were examined (Fig. 1). As shown in Figure 1, the percentage (%) of GFRα1-positive cells relative to the total number of GFP+ cells in each patch is plotted on the Y-axis, while the X-axis shows the total cell number of each GFP+ patch.
Most GFRα1-negative patches comprised less than 50 cells in total (green triangles in Fig. 1; all of these samples were isolated on days 7 and 10 post-transplant; see Supp. Fig. S1A, which is available online), with the exception of two large GFRα1-negative patches with 54 and 105 GFP+ cells (both samples were isolated on day 14 post-transplant; see Supp. Fig. S1B).
In GFP+ cell patches with less than 10 GFRα1-positive cells, all GFP+ cells were GFRα1-positive (blue broken line in Fig. 1). Unlike the cell-spreading shape of the As and Apr GFRα1-positive cells observed in healthy seminiferous tubules of wildtype mice (Fig. 2A), these Apr–Aal8-like GFRα1-positive cells were spherical in shape and tightly connected with each other (Fig. 2B).
At days 7 to 14 post-transplant, GFRα1-positive patches with a total of 10 to 250 GFP+ cells could be classified into two groups based on the ratio of GFRα1-positive cells to the total number of GFP+ cells. The “high-ratio” group (with at least 60% of GFRα1-positive cells) consisted of the 7 clusters (1 and 6 clusters at days 7 and 10 post-transplant, respectively) that are surrounded by the solid elliptical line in Figure 1 (also see Fig. 3A). The “low-ratio” group (with less than 40% of GFRα1-positive cells) consisted of 13 clusters (2, 7, and 4 clusters at days 7, 10, and 14 post-transplant, respectively) that are surrounded by the broken elliptical line in Figure 1 (also see Fig. 3B). Since it has previously been shown using a retransplantation assay that SSC activity is predominantly enriched in the GFRα1-positive cell fraction in these early stages of colonization (Morimoto et al., 2009), the present data suggest that most “high-ratio” GFRα1-positive patches are likely to be involved in the formation of stable spermatogenic colonies at later stages, with approximately half of the GFRα1-positive patches at day 10 post-transplant appearing to form a stable colony at a later stage (bottom row, Table 1). In contrast, it is likely that the “low-ratio” group includes many unsettled patches that are destined to disappear at later stages. In addition, at 3 months post-transplant, the number of GFRα1-positive cells relative to the total number of GFP-positive cells located in the basal compartment was approximately 1.8 ± 0.5% in the central region of the stable colonies (n= 5).
Phenotypes of GFRα1-Positive Spermatogonia in the “High-Ratio” and “Low-Ratio” Groups of GFRα1-Positive Patches
Interestingly, in all 7 patches in the high-ratio GFRα1-positive group, the formation of Aal-like GFRα1-positive aggregates was consistently observed on day 10 post-transplant (Fig. 3A). This observation was similar to the formation of the Aal8-like GFRα1-positive cell cluster at the previous stage (Fig. 2B; blue broken line in Fig. 1). As the number of cells within each GFP+ patch increased, we found that the GFRα1-positive morula-like clusters consisted of chains of not only Aal4, Aal8, Aal16, and Aal32-like cells, but also Aal12-like cells (possibly Aal4 + Aal8), although their intercellular connections are unclear. These Aal-like aggregates were scattered throughout a single GFP+ patch (Fig. 3A). Even in GFP+ patches with more than 100 cells, many GFRα1-positive cells were spherical in shape and tightly connected with each other, suggesting the synchronous maintenance of a GFRα1-positive state in each aggregate. Moreover, in patches with more than 1,000 cells, such Aal-like GFRα1-positive cell aggregates were frequently detectable (bottom right panel in Fig. 3A, also see Fig. 4B). With regards to the formation of spherical Aal4-like or Aal8-like GFRα1-positive cell aggregate at early stages (i.e., day 7 post-transplant; Fig. 2B), these data raise the possibility that a single, donor-derived GFRα1-positive spermatogonium may proliferate in an Aal-like aggregated manner during the early stages of the colonization.
In contrast, the 13 patches in the low-ratio GFRα1-positive group displayed considerable variation in both cell number and location of GFRα1-positive cells within each GFP+ patch (Fig. 3B). In some cases, a single GFRα1-positive cell aggregate was located away from the major GFP+ cell clusters without any GFRα1-positive cells. Anti-c-KIT staining confirmed that almost all GFRα1-negative cells were c-KIT-positive (Fig. 3C), suggesting that GFRα1-negative spermatogonia may possibly arise from GFRα1-positive cells in each GFP+ patch.
Phenotypes of GFRα1-Positive Spermatogonia in Stable Colonies at Later Stages
The phenotypes of GFRα1-positive spermatogonia in stable colonies 2 to 3 months after transplantation were examined. In GFP+ patches with active spermatogenesis, a large number of GFRα1-positive cells showed As and Apr spermatogonia with a cell-spreading shape (Fig. 4A), similar to those observed in the normal wildtype testes (Fig. 2A). Some GFRα1-positive spermatogonia, however, still formed an Aal-like cell aggregate with a spherical shape (“Aal16” in Fig. 4A), especially in the region near the edge of the GFP+ colony (arrows in Fig. 4B). These GFRα1-positive cells displayed Aal-like aggregates with a chained or morula-like structure (white arrows in Fig. 4B), similar to those observed in the early phases of the spermatogonial transplantation (see Fig. 3A). These data suggest that Aal-like aggregation of GFRα1-positive spermatogonia frequently occurs in the border region of the stable colony at later stages.
In order to quantitatively examine the changes in the number of chained cells in each GFRα1-positive cell cluster during SSC transplantation, we calculated the number of chained cells in each GFRα1-positive patch (n= 8 patches [within a range of 44–258 GFRα1-positive cells]) on day 10 post-transplant (Fig. 5A). These data were also compared with those at 3 months post-transplant (n=7 colonies) and in wild-type normal testes (n=6 testes). At day 10 post-transplant, the relative ratio of GFRα1-positive As cells was significantly lower at 14.7 ± 4.7% (n=8) than the 37.1 ± 5.1% observed at 3 months post-transplant (P < 0.01; Tukey tests; Fig. 5B). In contrast, the relative ratio of GFRα1-positive Aal-like cells with more than 9 cells on day 10 was significantly higher at 21.7 ± 3.7% than the 6.6 ± 2.6% observed at 3 months post-transplant (P < 0.01; Tukey tests; Fig. 5C). In addition, these two values on day 10 post-transplant were significantly different from those in normal testes (P < 0.01; Tukey tests).
Reduced Patterns of Immunoreactive GDNF Expression in Seminiferous Epithelia Associated With Colony Expansion
Previous studies have demonstrated that GDNF is more highly expressed in the W/Wv testes than in wild-type testes (Tadokoro et al., 2002; Sato et al., 2011). In order to understand the spatiotemporal changes in GDNF expression levels, we examined immunoreactive GDNF-positive signals in Sertoli cells during the early phases of colonization of the W/Wv testes using immunohistochemical techniques. Anti-GDNF staining confirmed a high-level of GDNF-positive signals in Sertoli cells of the W/Wv seminiferous tubules (Fig. 6B), compared to the low levels observed in the normal seminiferous tubules of the wildtype testes (Fig. 6A).
Our immunohistochemical studies using serial sections of the W/Wv testes transplanted with GFP+ cells revealed no appreciable reduction of GDNF-positive signals in the seminiferous epithelia with only several GFP+ cells in its basal compartment on day 10 post-transplant (Fig. 6C, c1), showing a similar level to those without any donor germ cells (Fig. 6C, c2). On day 14 post-transplant, a small patch of GFP+ spermatogenic cells was observed in some transversal sections of the seminiferous tubules and, on day 21 post-transplant, the spermatogenic cells, which included several advanced germ cells, were observed to be located in the inner layer separate from the basal lamina. Anti-GDNF immunostaining clearly showed a gradual reduction in immunoreactive GDNF signals in Sertoli cells with these donor-derived spermatogenesis (Fig. 6D, E). Such a reduced pattern of immunoreactive GDNF expression associated with the proliferation and differentiation of donor-derived germ cells was also seen in the presumptive edges of the stable colonies at 2 to 3 months after transplantation (Fig. 6F). Briefly, the level of GDNF immunoreactivity was higher in the presumptive edges of the spermatogenic colonies with only a few spermatogonia (Fig. 6F, f1) compared to the level in the central region of the same colony that exhibited complete spermatogenesis (Fig. 6F, f2).
It has previously been shown that during the early colonization of the spermatogonial transplantation, donor spermatogenic cells reach the basement membrane during the initial week after transplantation, and from 1 week to 1 month post-transplant the donor cells rapidly proliferate and form a monolayer network leading to the formation of a stable spermatogenesis colony from 1 month post-transplant (Nagano et al., 1999). Moreover, a previous sequential transplantation assay has revealed that SSCs rapidly proliferate from day 7 post-transplant, with proliferation activity significantly decreasing by 1 month post-transplant. This decline in activity leads to a stable density of SSCs, albeit active colony expansion, from 1 month post-transplant (Nagano, 2003). These data clearly suggest that SSC proliferation, expansion, and survival selection of each patch occur between 1 week and 1 month post-transplant. In order to expand these observations into potential SSC dynamics, the patterns and kinetics of GFRα1-positive cells at these early stages of the colonization were examined. Previous serial transplantation experiments have shown that SSC activity is predominantly enriched in the GFRα1-positive cell fraction in early-stage colonies (i.e., 11.5±2.0 [GFRα1-positive] versus 0.3±0.1 [GFRα1-negative] colonies/104 injected cells sorted from primary recipients at 2–4 weeks post-transplant; Morimoto et al., 2009). In addition, they also confirmed that SSCs are consistently positive for GFRα1 when the sorted cells are collected from primary recipients at 3 to 4 months post-transplant (Morimoto et al., 2009).
In the present study, we estimated the putative numbers of GFRα1-positive patches per testis on days 10, 14, and 21 post-transplant to be 19.2, 10.9, and 9.4, respectively. These findings suggest that approximately half of GFRα1-positive patches on day 10 post-transplant contribute to the survival colonies from day 21 post-transplant. From days 10 to 14 post-transplant, GFP+ patches in the range of approximately 10 to 250 cells could be classified into “high-ratio” or “low-ratio” groups, based on the number of GFRα1-positive cells relative to the total number of GFP+ cells. The high ratio group tended to be a more stable colony, which was more likely to survive at later stages (Fig. 1). Moreover, all GFP+ patches with more than 250 cells displayed a relatively low ratio of GFRα1-positive cells, which was similar to that of presumptive stable colonies at later stages. These data, therefore, imply that the survival selection of the donor-derived spermatogenic patch may occur before, and on day 14 post-transplant (for patch sizes up to approximately 250 cells), after which the cell patches are stably colonized in the recipient testes.
Unexpectedly, the present study also suggests that donor-derived GFRα1-positive cells appear to proliferate in a morula-like structure and frequently form several Aal-like spherical cell aggregates, especially in the high ratio group during early phases of colonization (Figs. 2B, 3A). Although we could not confirm the existence of an intercellular connection between two cells due to the tightly compacted aggregates, all these cells appeared to maintain a GFRα1-positive (i.e., undifferentiated) state in a synchronized manner from earlier stages (Fig. 2B). This finding is in clear contrast to GFRα1-positive spermatogonia, which are mostly As and Apr cells with an active cell-spreading shape in stable colonies at later stages (Fig. 4A, left) and in normal wildtype testes (Fig. 2A; Sada et al., 2009; Suzuki et al., 2009; Nakagawa et al., 2010). The existence of these two distinct phenotypes of GFRα1-positive spermatogonia is clearly consistent with the findings of a previous study that showed two distinct types of mouse SSCs in activated (corresponding to Aal-like cell aggregate) and non-activated (corresponding to spreading As and Apr cells in healthy testes) states, which change according to their microenvironment (Morimoto et al., 2009).
The morphology of Aal-like cell aggregates is reminiscent of that of GS (germline stem) cell clusters in vitro (Kanatsu-Shinohara et al., 2003, 2006b, 2011; Kubota et al., 2004; Yeh et al., 2007; Araki et al., 2010). A morula-like aggregate was previously shown to be caused by the proliferation of non-adherent GS cells in an anchorage-independent manner in vitro (Kanatsu-Shinohara et al., 2006b). Interestingly, the growth of these non-adherent GS cells is clearly dependent on a high concentration of GDNF in vitro. Since GFRα1 mediates secreted GDNF-positive signals to regulate their proliferation and survival (Hofmann et al., 2005; He et al., 2007), these data suggest that such an Aal-like clump formation of GFRα1-positive spermatogonia may be closely associated with regional levels of GDNF signals in the seminiferous tubules of the W/Wv testes transplanted with SSCs.
The present comparative immunohistochemical study confirmed that immunoreactive GDNF-positive signals were higher in the empty seminiferous tubules in W/Wv mutant than those in the normal seminiferous tubules with active spermatogenesis (Tadokoro et al., 2002; Sato et al., 2011) (Fig. 6A, B). Moreover, these GDNF-positive signals were gradually down-regulated in the seminiferous epithelia with colonized germ cells from days 10–14 post-transplant (Fig. 6C–E). In the presumptive edges of the stable colony, which contained only a small number of GFP+ spermatogonia in the basal compartment, GDNF-positive signals were still higher than in the central region of the same colony that contained more advanced spermatogenic cells (Fig. 6F). Taken together, these data suggest that high levels of GDNF expression in the almost empty seminiferous epithelia are positively correlated with the regional appearance of Aal-like GFRα1-positive cell aggregates that are frequently seen in the almost empty seminiferous epithelia at both early and later stages of the colonization. Since GDNF signaling has dose-dependent functions in the maintenance of the SSC pool (Meng et al., 2000; Naughton et al., 2006; Hofmann, 2008; Oatley and Brinster, 2008), this idea suggests that a regional reduction in immunoreactive GDNF signals at the site of a spermatogenic patch is possibly associated with the rapid reduction in the ratio of GFRα1-positive cells in each GFP+ patch during and after the critical time window at 10 to 14 days after transplantation (Fig. 1; Table 1). This in turn suggests that the regional reduction in GDNF-expression levels around GFRα1-positive cell clusters may induce their differentiation into GFRα1-negative advanced spermatogenic cells in each patch at this period, subsequently affecting the further selection of the survival colonies at later stages. In vivo and in vitro time-lapse analyses using GFRα1-knock-in mice are required to clarify the exact dynamics of a single GFRα1-positive cell in the empty seminiferous epithelia.
Animal Care and Use
All animal experiments in this study were carried out in strict accordance with the Guidelines for Animal Use and Experimentation as set out by the University of Tokyo. The procedures were approved by the Institutional Animal Care and Use Committee of the Graduate School of Agricultural and Life Sciences in the University of Tokyo (approval ID: P11-500).
Green (B6-Tg[CAG-EGFP] mice (1-week-old; SLC, Japan) and WBB6F1-W/Wv mice (3-week-old) were used in this study.
For spermatogonial transplantation, cell suspensions (including spermatogonia) (1.0 ×108 cells/ml) were prepared from the testes of 1-week-old Green C57BL6 mice and transplanted into the testes of 3-week-old recipient W/Wv mice (approximately 3 μl of cell suspension per recipient testis), as described previously (Brinster et al., 1994; Ogawa et al., 1997). On days 7, 10, 14, and 21 and also 1 to 3 months after transplantation, all recipient W/Wv testes were removed from the tunica albuginea and roughly dissected in cold PBS under an Olympus fluorescent stereomicroscope (SZX16 plus U-LH100HG). After the number of GFP+ patches was quantified, the tissues were processed by the immunohistochemical techniques described below.
For whole-mount immunohistochemistry, the number of GFP+ patches was quantified and the testicular tissues containing the GFP+ patches were fixed in 4% PFA for 12 hr and washed with cold PBST several times. After permeabilization steps using a blocking solution containing Tween-20 detergent, the seminiferous tubule fragments containing the GFP+ patches were comprehensively collected under a fluorescent stereomicroscope, and incubated with anti-GFRα1 antibodies (1:100 dilution; R&D Systems, Minneapolis, MN) at 4°C for 12 hr. After washing with PBST, the samples were incubated with Alexa-350/594 conjugated secondary antibodies, including DAPI, at room temperature for 2 hr. After counter-staining with DAPI, the samples were analyzed under Olympus fluorescent microscope (BX51N-34-FL2) and stereomicroscope (SZX16 plus U-LH100HG) systems. Whole-mount samples stained with anti-GFRα1 (red) were photographed (× 200) in each GFP+ (green) patch in multiple focal (Z-stack) planes (see Supp. Fig. S2) and then, in all tubule samples containing a whole GFP+ patch (on days 7–21 post-transplantation), the number of GFRα1-positive cells and the total number of GFP+ cells were calculated for each GFP+ patch. In GFRα1-positive cell clusters, the number of chained cells in each cluster was also estimated for each GFP+ patch. In some samples, after taking photographs, the same specimens were re-stained with anti-mouse c-KIT antibody (1:100 dilution; R&D Systems) as described above.
For section immunohistochemical staining, testes were isolated at various stages, and fixed in 4% PFA for 12 hr at 4°C. The specimens were dehydrated in ethanol, cleared in xylene, and then routinely embedded in paraffin. The continuous deparaffinized sections were incubated with rabbit anti-GDNF antibody (1:200 dilution; Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C for 12 hr. The reaction was visualized with biotin-conjugated secondary antibody in combination with Elite ABC kit (Vector Laboratories, Burlingame, CA). After counterstaining with hematoxylin (blue), the DAB-based HRP staining (brown) and GFP fluorescence (green) images were merged. Immunohistochemical staining in each sample was conducted at least three times to confirm its reproducibility. As a negative control, we confirmed no positive signals in the case of anti-GDNF antibody pre-incubated with GDNF peptide (sc-328P; Santa Cruz Biotechnology) prior to use in section immunostaining.
All quantitative data are represented as mean ± S.E.M. Data analysis was conducted with the graphics and statistics program PRISM v5.0 (GraphPad Software Inc., San Diego, CA). The ANOVA statistic was used to determine whether an overall difference existed among the numbers of cells. Where differences existed, the Tukey or Dunnett test was used to compare the number of cells relative to the total cell number. In all cases, differences were considered to be statistically significant when P < 0.05.
The authors thank Dr. Kenshiro Hara (Yoshida's lab at the National Institute for Basic Biology, Japan) for his comments on and critical reading of the manuscript, Drs. Mayuko Inagaki-Ishii and Yoshiko Kuroda for their kind and technical support, and Ms. Itsuko Yagihashi for her secretarial assistance. Kyoko Harikae and Yoshimi Aiyama are DC1 JSPS Research Fellows.