Drosophila Sex lethal Gene Initiates Female Development in Germline Progenitors

See allHide authors and affiliations

Science  12 Aug 2011:
Vol. 333, Issue 6044, pp. 885-888
DOI: 10.1126/science.1208146


Sex determination in the Drosophila germ line is regulated by both the sex of the surrounding soma and cell-autonomous cues. How primordial germ cells (PGCs) initiate sexual development via cell-autonomous mechanisms is unclear. Here, we demonstrate that, in Drosophila, the Sex lethal (Sxl) gene acts autonomously in PGCs to induce female development. Sxl is transiently expressed in PGCs during their migration to the gonads; this expression, which was detected only in XX PGCs, is necessary for PGCs to assume a female fate. Ectopic expression of Sxl in XY PGCs was sufficient to induce them to enter oogenesis and produce functional eggs when transplanted into an XX host. Our data provide powerful evidence that Sxl initiates female germline fate during sexual development.

Primordial germ cells (PGCs) are able to differentiate into eggs or sperm. It is thought that PGCs do not assume a sexual fate until they reach the gonads, where sexual dimorphism is imposed by both the sex of the surrounding soma and cell-autonomous cues (13). In Drosophila, pole cells or PGCs differentiate to a male fate in response to JAK/STAT signaling from the gonadal soma (46). The method by which female sexual development is initiated in pole cells, however, has not been elucidated. To clarify the mechanism that initiates a female fate in pole cells, we first identified a female-specific marker for this cell type. Although several sex-specific markers, including mgm-1, disc proliferation abnormal, and minichromosome maintenance 5, have been reported, they are all expressed only in male pole cells after gonad formation (stage 15), based on signals from the male gonadal soma (4, 5, 7). We previously showed that lesswright (lwr), a gene that regulates posttranslational modification of proteins by small ubiquitin-related modifiers, is expressed in pole cells during embryogenesis (8, 9). lwr is not characterized by sex-specific expression. When a dominant-negative form of lwr (lwrDN) (10) was expressed in the pole cells of either sex, however, apoptosis was induced only in female (XX) pole cells during migration to the gonads. This effect caused a significant reduction in the number of XX pole cells in the gonads (Fig. 1, fig. S1, and table S1) (11). Introduction of female-specific germline apoptosis induced by a dominant-negative form of lwr (f-gal) provides a previously uncharacterized marker of female sexual identity in migrating pole cells.

Fig. 1

Female-specific germline loss induced by a dominant-negative form of lwr. (A) Average numbers of pole cells in the gonads of female (magenta) and male (blue) control, lwrDN, Sxl o/e, Sxl o/e lwrDN, Sxl, Sxl lwrDN, tra-2, and tra-2 lwrDN embryos (stage 15) are shown (29). More than 40 gonads were examined in each case. Error bars represent SD. Significance was calculated using the Student’s t test (*P < 0.01; **P > 0.05). (B to G) Gonads in female (B to D) and male (E to G) control (B and E), lwrDN (C and F), and Sxl o/e lwrDN (D and G) embryos (stage 15) were stained for the germline marker Vasa (magenta). (H to K) Female (H and I) and male (J and K) control (H and J) and lwrDN (I and K) embryos (stage 13) were stained for Vasa (magenta) and cleaved caspase-3 (green). Arrowheads indicate pole cells that were positive for cleaved caspase-3. Scale bars, 10 μm.

Sex determination is controlled by the Sex lethal (Sxl) gene, which is first expressed at the blastodermal stage in the embryonic soma (1214). Sxl encodes an RNA binding protein involved in alternative splicing and translation. In the soma of XX embryos, it functions through transformer (tra) and transformer-2 (tra-2), which in turn regulate alternative splicing of the doublesex (dsx) gene to produce a female-specific form of Dsx (15, 16). In male (XY) embryos, this pathway is turned off, and a male-specific form of Dsx is produced by default (16). These Dsx proteins determine the sexual identity of somatic tissues (1517). Previous reports, however, suggested that Sxl does not induce female sexual development in the germ line, as it does in the soma (18, 19). Although Sxl is autonomously required for female sexual development (16, 17, 20), constitutive mutations in Sxl (SxlM) that cause XY animals to undergo sexual transformation from male to female do not necessarily interfere with male germline development (18, 19). Moreover, tra, tra-2, and dsx are not required for female germline development (2022). Finally, female-specific Sxl expression has been detected later in gametogenesis, but not in early germline development (12, 13, 20).

Contrary to previous observations, we found that Sxl was expressed in XX but not XY pole cells during their migration to the gonads (Fig. 2). In the soma, Sxl transcripts are first expressed from the establishment promoter (Sxl-Pe) in a female-specific manner (12, 14). Using a probe specific to the early transcript derived from Sxl-Pe, we detected in situ hybridization signals in migrating XX pole cells at around stage 9/10 (Fig. 2, B to E). We used transgenic embryos, which expressed enhanced green fluorescent protein (EGFP) under the control of the Sxl-Pe promoter (Fig. 2, F to I) (23), to further confirm this female-specific Sxl-Pe activation. We used reverse transcription polymerase chain reaction and sequencing analyses in pole cells to detect early Sxl transcripts that had the same sequence as the transcripts expressed in the soma (11).

Fig. 2

Female-specific expression of Sxl in migrating pole cells. (A) An embryo (stage 4) stained for early Sxl transcripts. No specific signals were observed in pole cells (arrowhead). (Right) Zoomed-in view. (B and C) Female (B) and male (C) embryos (stage 9) stained for early Sxl transcripts. Early transcripts were detected in pole cells only in females. Arrowheads indicate pole cells. (D and E) Confocal images of a female embryo (stage 10) double stained for Vasa (D) and early Sxl transcripts (E). (F to I) Female (F and G) and male (H and I) embryos (stage 9) carrying Sxl-Pe-EGFP were stained for Vasa (magenta) (F and H) and GFP (green) (G and I). GFP was detected in pole cells of female embryos (arrowheads). Scale bars, 10 μm.

Next, we determined whether Sxl feminized early pole cells using f-gal as a marker for female identity. We found that the loss-of-function mutation SxlfP7B0 (24) repressed f-gal in XX pole cells (Fig. 1A). This repression is unlikely to result from sexual transformation of the soma, because an amorphic tra-2 mutation, which alters somatic sex (19, 22, 25), did not affect f-gal (Fig. 1A). Conversely, when we forced the expression of Sxl together with lwrDN in pole cells from stage 9 onward by using nanos-Gal4 (26) and UAS-Sxl (27), f-gal was ectopically observed in XY pole cells (Fig. 1, A, D, and G). We found that Sxl alone did not induce apoptosis or developmental defects in pole cells (Fig. 1A) (11). These observations suggest that female sexual identity of migrating pole cells is regulated cell-autonomously by Sxl.

We then determined whether Sxl induced female development in XY pole cells. Because XY soma produces signals that direct XX germline cells to a male fate (4, 5, 7), we transplanted XY pole cells expressing Sxl into XX females and examined their developmental fate. Even in the presence of a gain-of-function Sxl mutation (SxlM1) that causes XY soma to transform from male to female, XY (or XO) pole cells enter the spermatogenic pathway when transplanted into XX females (18, 19). These results suggest that Sxl is not sufficient to activate female germline development. SxlM1 mutations, however, do not affect transcription from the Sxl-Pe promoter, but instead structurally alter the late transcript from the Sxl maintenance promoter (Sxl-Pm), which allows Sxl protein production in both males and females (28). Consistent with this observation, we detected Sxl transcripts derived from Sxl-Pe in the pole cells of only female SxlM1 embryos (fig. S7, D and E). Thus, the SxlM1 mutation does not result in Sxl expression in XY pole cells as early as in XX pole cells.

Instead, we used nanos-Gal4 and UAS-Sxl to induce Sxl expression in XY pole cells. We transplanted three types of XY pole cells, each characterized by a different duration of Sxl expression: (i) XY pole cells in which Sxl was expressed from stage 9 until stage 16/17 using maternal nanos-Gal4 (XY-mSxl), (ii) XY pole cells in which Sxl was expressed from stage 15/16 onward using zygotic nanos-Gal4 (XY-zSxl), and (iii) XY pole cells in which Sxl was expressed from stage 9 onward using both maternal and zygotic nanos-Gal4 (XY-mzSxl) (29). We found that XY-mzSxl and XY-mSxl pole cells entered the oogenic pathway and produced mature oocytes in XX females (Fig. 3, C and D, and Table 1). These oocytes contributed to progeny production (Fig. 3, E to H, and Table 1). Thus, the XY pole cells produced functional eggs, even though oogenesis and egg production were reduced compared with XX pole cells (Table 1). In contrast, XY-zSxl pole cells did not enter the oogenic pathway in almost all (92.3%) of the XX female hosts (Table 1) and instead were characterized by a tumorous phenotype, an indication of XY germline cells that have maintained male characteristics (Fig. 3, A and B) (18). Control XY pole cells from the embryos expressing Sxl only in the soma (XY-nullo-Sxl) showed a similar phenotype to that of XY-zSxl pole cells (Table 1). These observations demonstrate that Sxl expression in XY pole cells during embryogenesis induces functional egg differentiation in the female soma.

Fig. 3

Ectopic Sxl expression induced a female fate in XY pole cells. (A to D) Ovaries from EGFP-vasa adults transplanted with XY (A), XY-zSxl (B), XY-mzSxl (C), or XY-mSxl (D) pole cells were stained for Vasa (magenta) and GFP (green) (29). Asterisks denote GFP-negative clones derived from the transplanted pole cells: single asterisk, XY and XY-zSxl pole cells produced a tumorous phenotype in the germarium; double asterisk, XY-mzSxl and XY-mSxl cells underwent oogenesis and formed egg chambers. (E to H) XY-mSxl pole cells contributed to progeny production. Ovary (F) and testis (H) from F1 progenies derived from the transplanted XY-mSxl pole cells were GFP-negative, whereas F1 progenies from host germline cells expressed GFP (green) in ovary (E) and testis (G) (29). (I and J) Distal regions of testes stained for Vasa (green). Adult males were derived from nos-Gal4/nos-Gal4 females mated with nos-Gal4/nos-Gal4 (control) (I) or UAS-Sxl/UAS-Sxl (nos-Gal4>UAS-Sxl) (J) males. In nos-Gal4>UAS-Sxl testes, spermatogenesis proceeded normally, and sperm was produced. Scale bars, 40 μm.

Table 1

Sxl induced female fate in XY pole cells. Pole cell transplantation experiments are described in the supporting online material (29).

View this table:

We then used Sxl-specific double-stranded RNA (UAS-SxlRNAi) under the control of maternal nanos-Gal4 to reduce Sxl activity in XX pole cells during embryogenesis. Introducing UAS-SxlRNAi resulted in tumorous and agametic phenotypes in female adults, indicating that the XX germ line lost female characteristics (fig. S8) (18, 19). Taken together, our results show that Sxl acts as a master gene necessary and sufficient to induce female development in pole cells.

We found that XY-mzSxl pole cells adopted a male fate and executed spermatogenesis when they developed in an XY male soma (Fig. 3, I and J, and fig. S9). This observation suggests that the male soma plays a dominant role in determining the male germline fate, overriding the feminizing effect of Sxl. Another possibility is that the XX female soma plays a critical role in maintaining the Sxl-initiated female germline fate. Indeed, an XX germ line in the male soma shows a male gene-expression profile, whereas an XY germ line in the female soma exhibits a female expression profile, although these germ lines does not execute gametogenesis (4, 30). Thus, female germline development requires interactions between the germline and somatic cells, in addition to germline-autonomous mechanisms involving Sxl.

In mice, germline sexual identity is also regulated by both germline-autonomous and somatic signals (6, 7). In the coelenterate Hydra, the germline sex is not influenced by the surrounding soma, and the germ line determines the phenotypic sex of the polyp (31, 32). Thus, germline-autonomous regulation of sex has probably been present throughout the evolution of animals, and somatic control may have evolved with the emergence of mesodermal tissues, including gonadal soma. Sxl does not appear to play a key role in sex determination in non-drosophilid animals (15, 17). Nevertheless, future studies should determine whether Sxl homologs are expressed in the germ line of non-drosophilids. Moreover, it would be of particular interest to identify downstream targets of Sxl in the Drosophila germ line and to test whether these genes have a widespread role in germline sex determination.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S9

Tables S1 to S5


References and Notes

  1. Supporting online text is available on Science Online.
  2. Materials and methods are available as supporting material on Science Online.
  3. Acknowledgments: We thank the researchers who provided us with flies and antibodies and members of our laboratory for their valuable comments. We also thank the Drosophila Genetic Resource Center (Kyoto), the Bloomington and Vienna Drosophila RNAi Stock Centers, and the Developmental Studies Hybridoma Bank for fly stocks and antibodies. This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (Japan) to S.K. and a Research Fellowship for Young Scientists from the Japan Society for the Promotion of Science to K.H.
View Abstract

Navigate This Article