The RNA-Binding Protein NANOS2 Is Required to Maintain Murine Spermatogonial Stem Cells

See allHide authors and affiliations

Science  11 Sep 2009:
Vol. 325, Issue 5946, pp. 1394-1398
DOI: 10.1126/science.1172645


Stem cells give rise to differentiated cell types but also preserve their undifferentiated state through cell self-renewal. With the use of transgenic mice, we found that the RNA-binding protein NANOS2 is essential for maintaining spermatogonial stem cells. Lineage-tracing analyses revealed that undifferentiated spermatogonia expressing Nanos2 self-renew and generate the entire spermatogenic cell lineage. Conditional disruption of postnatal Nanos2 depleted spermatogonial stem cell reserves, whereas mouse testes in which Nanos2 had been overexpressed accumulated spermatogonia with undifferentiated, stem cell–like properties. Thus, NANOS2 is a key stem cell regulator that is expressed in self-renewing spermatogonial stem cells and maintains the stem cell state during murine spermatogenesis.

Stem cells are essential for tissue homeostasis and regenerative responses to injury and disease. In the spermatogenic stem cell system, germ cell–intrinsic factors have an essential role in the maintenance of stem cells for the continuation of spermatogenesis throughout life (15). However, the previous loss-of-function studies have some limitations in terms of understanding the mechanism by which stem cells are lost upon the gene deletion, as it could be caused by cell death, defective self-renewal, premature differentiation, or other mechanisms.

For decades, the mammalian spermatogenic stem cell has been characterized by the morphological features of the spermatogonia. The spermatogonial types Asingle (As; isolated single cells), Apaired (Apr; chains of 2 cells), and Aaligned (Aal; chains of 4, 8, 16 or 32 cells) are the most primitive germ cells observed in mature testes and are collectively described as undifferentiated spermatogonia. They give rise to differentiating spermatogonia, which undergo additional divisions and enter a differentiation pathway. It has been proposed that only As spermatogonia represent the stem cells (68); however, there is no As-specific molecular marker, and the presence of stem cells is assayed by long-term colony formation after the transplantation of candidate cells into recipient testes (9). For this reason, undifferentiated spermatogonia containing As to Aal are the smallest population proven to have the properties of stem cells.

Recently, two functionally distinct spermatogonial stem cell populations were identified in mice (10). One is the population that acts as the self-renewing stem cells (actual stem cells), and the other population possesses the potential to self-renew but acts as transit-amplifying cells (potential stem cells) during steady-state spermatogenesis. It was also indicated that Neurogenin3 (Ngn3) might be expressed higher in potential stem cells (10, 11). At present, however, the molecular signatures or morphological features that distinguish actual stem cells from potential stem cells remain elusive.

NANOS, a zinc-finger RNA-binding protein, has been proposed as a conserved factor for germline stem cell function (1215). In adult mouse testes, NANOS2 is predominantly expressed in As and Apr spermatogonia, and it appears to represent a different subset of undifferentiated spermatogonia from those expressing Ngn3 (fig. S1). The heterogeneity observed in undifferentiated spermatogonia may represent functional differences, and the restricted expression of NANOS2 suggests its possible involvement in stem cell function. However, the majority of Nanos2-null germ cells die by apoptosis before birth (16), hindering functional studies of NANOS2 during spermatogenesis.

To test whether NANOS2-expressing spermatogonia have stem cell properties in vivo, we followed the fate of Nanos2-expressing cells using a tamoxifen (TM) inducible Cre/loxP cell-lineage tracing system, and we visualized cells using 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) staining to detect β-galactosidase activity (fig. S2, A and B). Six-week-old transgenic mice were subjected to a TM pulse, resulting in the initial labeling of ~20 to 25% of the NANOS2-expressing undifferentiated spermatogonia (fig. S2, C to F, and table S1). At 10 weeks (1 month after labeling), a large number of X-Gal–positive cells were observed all along the seminiferous tubules (Fig. 1A), which were classified as either differentiated spermatogenic cells (Fig. 1B, asterisks) or spermatogonia (Fig. 1B, arrowheads). Because the completion of spermatogenesis takes ~40 days from the initial step of differentiating spermatogonia (8), these stained cells would reflect stem cell-derived clones and/or transient clones that originated from transit-amplifying cells. At 18 weeks (3 months after labeling), a sufficiently long period for a repeated completion of spermatogenesis, we observed X-Gal–positive patches that contained all stages of spermatogenic cells (Fig. 1, C and D). These patches persisted for prolonged periods of time (Fig. 1E). This indicates that the labeled spermatogonia continuously give rise to differentiating cells while maintaining their own population over long periods. Therefore, the spermatogonia can be referred to as actual stem cells.

Fig. 1

Pulse-chase studies of Nanos2-expressing spermatogonia. (A to D) Detection of the Nanos2-expressing cell lineage by X-Gal staining. Insets in (B) and (D) are cross sections of the representative tubules containing X-Gal–positive cells. The stained differentiated spermatogenic cells within the inner layer (asterisks) and spermatogonia on the periphery (arrowheads) are indicated in (B). TM was injected at 6 weeks of age. Scale bars, 5 mm in (C); 1 mm in the main panel of (D); 100 μm in the inset of (D). (E) Average number of X-Gal–positive patches per testis is shown with the SE. In Ngn3 lineage, the number of these patches has been reported to be 6.1 ± 0.7 per testis at 3 months after pulse (10). N number of times an experiment was independently performed.

In addition, we noticed distinct differences in the cell fates between the Nanos2 lineage and the previously reported Ngn3 lineage (10). In the Ngn3 lineage, the majority of labeled spermatogonia acted as transit-amplifying cells that immediately proceed to differentiation without self-renewal (10) (summarized in table S1). Eventually, a small fraction of labeled spermatogonia in the Ngn3 lineage formed persistent patches; however, the number was more than 10-fold smaller than that of the Nanos2 lineage [Fig. 1E compared with (10)]. The difference is not due to the efficiency of pulse-labeling method, because the number of initially labeled spermatogonia was similar in both cases [table S1 compared with (10)]. These results indicate that the NANOS2-expressing subpopulation of undifferentiated spermatogonia contains a higher proportion of actual stem cells than transit-amplifying cells that are represented by Ngn3 expression.

To elucidate the function of NANOS2 during spermatogenesis, we employed a conditional knockout (cKO) strategy (Fig. 2A and fig. S3). Histological analysis of the Nanos2-cKO testes revealed a progressive defect in spermatogenesis with age (Fig. 2, B to G). The histology and immunostaining with germ cell markers showed various degrees of germ cell depletion in the 8-week-old Nanos2-cKO mice (Fig. 2, D, E, H, and I). By 12 weeks after birth, most of these tubules were devoid of any germ cells (compare Fig. 2F with Fig. 2G). Hence, postnatal deficiency of Nanos2 results in the gradual loss of the germ cell population within a few cycles of spermatogenesis.

Fig. 2

Progressive loss of spermatogenesis in Nanos2-cKO mice. (A) Strategy to generate mice lacking postnatal Nanos2 expression. By introducing the transgene (floxed 3xFlag-tagged Nanos2) into the Nanos2-null background, the Nanos2 expression is rescued. For conditional deletion of postnatal Nanos2, TM was injected into 4-week-old Nanos2−/−; Tg+; Ert2-Cre mice for 5 consecutive days, and the testes were harvested at the indicated time points. The blue line segments indicate the periods at which germ cells express Nanos2. (B to G) Histology of control and Nanos2-cKO testes. Tubules lacking spermatogonia (arrowheads), those with only elongated spermatids (arrow), and those without any germ cells (asterisks) are indicated. (H and I) Immunostaining of 8-week-old testes with TRA98 (a germ cell marker for spermatogonia, spermatocytes, and round spermatids, shown in green), anti-PLZF (a marker for undifferentiated spermatogonia, shown in magenta) antibodies, and 4′,6′-diamidino-2-phenylindole (DAPI) (shown in blue). The dotted lines show outlines of seminiferous tubules. (J) The number of PLZF-positive undifferentiated spermatogonia per seminiferous tubule in sections. More than 100 tubules from 20 independent microscopic fields were scored (n = 3). The results were normalized using the control of each stage. Asterisks indicate **P < 0.01. (K to N) Five-week-old testes of control and Nanos2-cKO were examined by whole-mount immunostaining with anti-GFRA1 (green) and anti-NANOS3 (red) antibodies. Scale bars, 100 μm.

To determine the cause of germ cell loss observed in Nanos2-cKO mice, we counted the number of promyelocytic leukemia zinc-finger (PLZF)–positive undifferentiated spermatogonia during the 4 weeks after TM administration and found that their number quickly declined within 2 weeks (Fig. 2J). Furthermore, the most primitive set of spermatogonia indicated by GFRΑ1 (glial cell line–derived neurotrophic factor family receptor alpha 1) expression were lost immediately after Nanos2 deletion, although more differentiated types of undifferentiated spermatogonia that express NANOS3 still remained (Fig. 2, K to N). Hence, the germ cell–loss phenotype in Nanos2-cKO mice was caused by the depletion of spermatogonial stem cells that produce differentiating spermatogenic cells.

The stem cell depletion in Nanos2-cKO mice could occur by apoptosis and/or premature differentiation of these cells (17) (figs. S4 and S5). To address the role of NANOS2 in stem cell maintenance, we designed a gain-of-function study. To achieve the continuous expression of Nanos2 in the germ cell lineage, we crossed CAG-floxed CAT-3xFlag-Nanos2 transgenic mice with Nanos3-Cre mice (fig. S6A). The histology (Fig. 3, A and B) and the expression of germ cell or meiotic markers (fig. S6, B to E) revealed that most tubules in the Nanos2-overexpressing testes retained spermatogonia-like cells on the periphery, whereas differentiated germ cells were absent or markedly reduced in number. To characterize these remaining cells, we examined spermatogonial markers. In Nanos2-overexpressing testes, the number of PLZF-expressing undifferentiated spermatogonia was significantly increased (Fig. 3, E to G), whereas KIT (kit oncogene)–expressing differentiating spermatogonia were rarely observed (compare Fig. 3C and Fig. 3D). Furthermore, these PLZF-positive cells in Nanos2-overexpressing mice had lower proliferation rates (Fig. 3H) and wild-type levels of apoptosis (Fig. 3I). Taken together, these data suggest that Nanos2-overexpression results in the accumulation of undifferentiated spermatogonia due to blocked differentiation rather than hyperproliferation or a reduction in apoptosis. Similar results were obtained when Nanos2 was induced after birth in Ngn3-Cre instead of Nanos3-Cre mice (17) (fig. S7).

Fig. 3

Accumulation of PLZF-positive spermatogonia after Nanos2 overexpression. (A and B) Histological cross sections at 8 weeks. Rarely observed differentiated germ cells in Nanos2-overexpressing (OE) mice are due to incomplete recombination by Cre (asterisk). H-E, hematoxylin and eosin staining. (C and D) Eight-week-old testes were stained with antibodies against KIT (magenta) and DAPI (blue). KIT is expressed in differentiating spermatogonia (arrowhead) and Leydig cells (asterisk). (E and F) Testes were stained with antibody for PLZF (magenta) and DAPI (blue). Scale bars, 100 μm. (G) Calculation of PLZF-positive cells per seminiferous tubule in 8-week-old mice. (H and I) Quantification of proliferative (H) and apoptotic (I) spermatogonia in 4-week-old mice. The number of BrdU- and cleaved PARP-positive cells were scored per number of PLZF-positive cells, respectively. More than 15 independent microscopic fields were counted (n = 3). The mean values are shown with the SE. **P < 0.01.

To further analyze the outcome of Nanos2 overexpression, we performed whole-mount immunostaining of seminiferous tubules with several molecular markers representing distinct populations of undifferentiated spermatogonia (Fig. 4A). We first studied the morphological features of Nanos2-overexpressing cells. In the control tubules, As, Apr, Aal-4, and Aal-8 spermatogonia were observed at comparatively similar proportions, whereas the proportion of As and Apr spermatogonia was higher in Nanos2-overexpressing tubules (Fig. 4B). Subsequently, we characterized the protein expression patterns of these cells. In control testes, PLZF expression was observed in most undifferentiated spermatogonia (fig. S8, A to C), whereas GFRA1 was expressed preferentially in As and Apr spermatogonia (Fig. 4, C to E, and fig. S1, A to C). Ngn3-EGFP (Fig. 4, I to K, and fig. S1, D to F) and NANOS3 (fig. S8, G to I) expression were generally found in the longer-chained cells. In the Nanos2-overexpressing testes, most of FLAG-NANOS2–positive cells expressed PLZF (fig. S8, D to F) and GFRA1 (Fig. 4, F to H), but exhibited no or lower levels of Ngn3-EGFP (Fig. 4, L to N) and NANOS3 (fig. S8, J to L) expression. Taken together, it appears that the Nanos2-overexpressing cells have properties similar to the most primitive set of undifferentiated spermatogonia, which indicates that NANOS2 is a cell-intrinsic factor to maintain the unique, undifferentiated state of spermatogonial stem cells.

Fig. 4

Nanos2-overexpressing cells display characteristics of the most primitive undifferentiated spermatogonia. (A) Schematic view of the relation between the typical morphology and gene-expression patterns observed in each group of undifferentiated spermatogonia during normal spermatogenesis. The gray and yellow color gradients in each panel represent the morphological changes and NANOS2 protein levels with respect to spermatogonial differentiation, respectively. (B) Frequency of the undifferentiated spermatogonia classified by the number of chained cells. A total of 420 (control) and 574 (Nanos2-overexpression) clusters were counted. (C to H) Whole-mount immunostaining with anti-NANOS2 (green) and anti-GFRA1 (red) antibodies. The dotted lines show outlines of seminiferous tubules. (I to N) Double-staining with the use of antibodies for NANOS2 (red) and GFP (green) to detect Ngn3-EGFP. Scale bar, 100 μm.

In this report, our genetic studies demonstrate a role of NANOS2 in the murine spermatogenic stem cells (17) (fig. S9). As indicated in Drosophila, Nanos may function by either blocking translation or facilitating RNA degradation of target mRNAs (18, 19). Our current study provides valuable information regarding the in vivo nature of adult-tissue stem cells and the importance of posttranscriptional regulation in stem cell functions.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S9

Table S1


References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. We thank Y. Nishimune for the TRA98 antibody; S. Chuma and N. Nakatsuji for the SCP3 antibody; S. Yoshida for the Ngn3-EGFP and Ngn3-Cre transgenic mice; S. Evans for the MER-Cre-MER construct; and T. Shinohara, R. Behringer, and Y. Hiromi for critical reading of the manuscript. This study was supported in part by the Genome Network Project of the Ministry of Education, Culture, Sports, Science and Technology of Japan and by Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists to A.S.
View Abstract

Navigate This Article