Conserved Role of nanos Proteins in Germ Cell Development

See allHide authors and affiliations

Science  29 Aug 2003:
Vol. 301, Issue 5637, pp. 1239-1241
DOI: 10.1126/science.1085222


In Drosophila, maternally supplied Nanos functions in the migration of primordial germ cells (PGCs) into the gonad; in mice, zygotic genes are involved instead. We report the cloning and the functional analyses of nanos2 and nanos3 in mice. These genes are differentially expressed in mouse PGCs. nanos2 is predominantly expressed in male germ cells, and the elimination of this gene results in a complete loss of spermatogonia. However, nanos3 is found in migrating PGCs, and the elimination of this factor results in the complete loss of germ cells in both sexes. Hence, although mice and flies differ in their mechanisms for germ cell specification, there seems to be conserved function for nanos proteins among invertebrates and vertebrates.

A Nanos gene, encoding an RNA binding protein, was first identified as a maternal effect gene in Drosophila (1). In the absence of maternal Nanos, PGCs fail to migrate into the gonad and do not become functional germ cells (2, 3). To date, nanos-related genes have been cloned in several invertebrates and vertebrates (48). In these organisms, maternally derived nanos protein plays a critical role in germ cell development. In mice, however, germ cell fate is induced in proximal epiblast cells by the extra-embryonic ectoderm at E6.5 (embryonic day 6.5) (9, 10). The mechanism responsible for the specificity and maintenance of PGCs is not yet known, although two genes, fragilis and stella, have been suggested to play important roles in germ cell specification (11). We have previously cloned the mouse nanos gene, nanos1, which is maternally derived (12). The disruption of nanos1, however, did not affect germ cell development, which led us to search for other nanos-related genes.

An indication of the existence of other nanos-class genes in mammals was provided by a BLAST analysis of the human genome using the conserved zinc-finger motif. Two putative human homologs (NT-011212.2/Hs1911369 and NT-011151.2/Hs11308) were subsequently identified. We used this information to clone two mouse homologs, designated nanos2 and nanos3. The deduced amino acid alignment with other nanos proteins revealed that the zinc-finger motif was the only conserved region (fig. S1, A and B).

The expression of nanos2 and nanos3 was observed in the developing gonads (Fig. 1A). Reverse transcription-polymerase chain reaction (RT-PCR) analyses confirmed the expression patterns of nanos2 and nanos3 in the male gonad; however, nanos3 expression was observed in the bipotential gonad at E11.5, and in the female gonad at E12.5 (13). By examining isolated PGC fractions at various developmental stages (14), we found that nanos3 was expressed in the PGCs as early as E9.5 and expressed in male gonads at a later stage of development, whereas the expression of nanos2 was restricted to the developing male PGCs (Fig. 1B).

Fig. 1.

Expression of nanos genes and phenotype of the nanos2-null mouse. (A) Expression of nanos2 and nanos3 in the developing gonads by whole-mount in situ hybridization. (B) Expression in germ cells at different stages of development, revealed by RT-PCR. (C to E) Morphology and histology of 4-week-old testes. (F to K) Comparison of germ cells in nanos2+/- and nanos2-/- testes, revealed by immunohistochemical staining for the TRA 104 antibody at E14.5 [(F) and (G)], E16.5 [(H) and (I)], E18.5 [(J) and (K)], and 4 weeks [(D) and (E)]. Arrowheads indicate ectopically localized germ cells. (L to Q) TUNEL-positive cells (arrow-heads) are absent in nanos2+/- testes [(L), (N), and (P)] but present in the nanos2-/- testes [(O) and (Q)]. Scale bars: (C), 1 mm; [(D) and (E)], 50 μm; [(F) to (Q)], 100 μm.

To examine the function of nanos2 and nanos3, we generated knockout mice (figs. S2 and S3). Both heterozygous and homozygous knockout mice for either nanos2 or nanos3 were viable and showed no apparent abnormalities. However, the gonads of the homozygous mice showed defects, consistent with the expression patterns of nanos2 and nanos3.

The defects in nanos2-null mice were observed only during male spermatogenesis. The size and weight of the testes in 4-week-old null mice were reduced; the weight was about 30% of that of the wild type (Fig. 1C), and no germ cells were detected (Fig. 1, D and E). In contrast, female gonads were morphologically and functionally normal and the homozygous female mouse was fertile. To define the onset and cause of germ cell deficiency, we examined embryonic male gonads. At E14.5, testicular cords were well organized even in the nanos2-null testis, and TRA 104-positive germ cells were normally localized in testicular cords (Fig. 1, F and G). However, from E15.5, some nanos2-null germ cells were localized outside of seminiferous tubules (Fig. 1, H and I) and the number of germ cells was reduced. In E18.5 testes, the number was greatly reduced (Fig. 1, J and K) and no germ cells were seen after 4 weeks (Fig. 1, D and E). Apoptotic cells were observed after E15.5 in nanos2-null testes (Fig. 1, L to Q). Although the number of terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL)-positive cells had not increased, apoptosis was detectable until the germ cells had completely disappeared.

In the nanos3-null mice, the size of the ovaries and testes were both greatly reduced (the weight of the testes was reduced to about 30% of that of the wild type, which was similar to that observed in nanos2-null mice) (Fig. 2, A and D). No germ cells were observed in sections (Fig. 2, B, C, E, and F). Germ cells were absent even in the E15.5 ovaries and testes (Fig. 2, G and J). Additionally, only a few germ cells were detected in the E12.5 genital ridge (Fig. 2, H, I, K, and L). When we examined E9.5 embryos by alkaline phosphatase (ALP) staining, we observed more migrating PGCs than we observed in E12.5 embryos (Fig. 2, M and N), indicating that the specification and derivation of PGCs were normal but that PGCs were not subsequently maintained during migration. An equivalent number of PGCs were detected in the wild type and in the nanos3-null embryo at E7.5, the earliest stage for PGC detection by ALP (Fig. 2O). Although the initial expression of nanos3 has not been defined by either RT-PCR or whole-mount in situ hybridization, it appears that PGCs might divide at least two to three times after birth, but their number declines to almost zero before E11.5 (fig. S4). This indicates that nanos3 is involved in the maintenance of PGCs by supporting their proliferation and/or by suppression of cell death. To further address this question, we subjected sections of nanos3-null embryos between E9.5 and E10.5 to the TUNEL reaction together with PGC identification by immunostaining with the 4C9 antibody. We did not detect double-positive cells (Fig. 2, P and Q), suggesting that apoptosis is not the cause of PGC depletion in nanos3-null embryos.

Fig. 2.

Phenotype of the nanos3-null mouse. The morphology and the histology of 4-week-old testis (A to C) and 6-week-old ovaries (D to F) were compared. (G to L) TRA 104-positive germ cells were compared in testes at E15.5 [(G) and (J)]. [(H), (I), (K), and (L)] SSEA1-positive germ cells were compared in testes [(H) and (K)] or ovaries [(I) and (L)] at E12.5. (M to O) ALP-stained PGCs were compared in E9.5 embryos [(M) and (N)] and E7.5 embryos (O). (P and Q) Dorsal mesentery was double stained by the 4C9 antibody (red) to detect PGC and TUNEL reagent (green) for apoptotic cells at E9.5. Genotypes are indicated. Unless otherwise indicated, scale bars represent 50 μm.

We have demonstrated here an evolutionarily conserved function of nanos proteins in germ cell development. Notably, a mechanism to maintain germ cells is highly conserved, but the mechanism or mechanisms determining germ cell fate are distinct, especially in mammals. Recent studies have begun to elucidate the molecular pathways by which nanos proteins maintain germ cell development through their association with Pumilio in Drosophila (2, 14) and through a translational repression mechanism (15). Human NANOS1 has also recently been shown to interact with human PUMILIO-2 and these are coexpressed in spematogonia (16), indicating a similar function in the human testis. Advances in proteomics are likely to identify additional components of these pathways. Finally, nanos2- and nanos3-deficient mice may provide excellent model systems to evaluate the cause of sterility and infertility in humans.

Supporting Online Material

Materials and Methods

Figs. S1 to S4


References and Notes

Stay Connected to Science

Navigate This Article