Notch signaling is used for cell-fate determination in many different developmental contexts. This study shows that the master control gene for sex determination in Drosophila, Sex-lethal, negatively regulates the N-signaling pathway in females. In genetic assays, reducing Sxl activity suppresses the phenotypic effects of N mutations, while increasing Sxl activity enhances the effects. Sxl appears to negatively regulate the pathway by reducing N protein accumulation, and higher levels of N are found in Sxl⁻ clones than in adjacent wild-type cells. The inhibition of N expression does not depend on the known downstream targets of Sxl; however, it was found that Sxl protein can bind to N mRNAs. Finally, these results indicate that downregulation of the N pathway by Sxl contributes to sex-specific differences in morphology and suggest that it may also play an important role in follicle cell specification during oogenesis (Penn, 2007).

While it has long been known that Sxl must control some aspects of sexual dimorphism by mechanisms that are independent of the Sxl→tra→dsx-fru regulatory cascade, understanding of what these morphological features might be and of how this might be accomplished has remained rudimentary. In these studies reported here, a regulatory link has been uncovered between Sxl and the N-signaling pathway. Sxl impacts the functioning of this pathway in a sex-specific fashion by negatively regulating N itself (Penn, 2007).

Several lines of evidence support the conclusion that the N-signaling pathway is a target for Sxl regulation. N and Sxl show genetic interactions in a variety of different developmental contexts. In the ovary, egg-chamber packaging defects are induced when homozygous N^ts1 females are placed at the nonpermissive temperature. Eliminating one copy of the Sxl gene dominantly suppresses these egg-chamber packaging defects. In female wing discs, N is haploinsufficient for the formation of the tip of the wing blade. This haploinsufficiency is sensitive to the Sxl gene dose. The N wing phenotype is suppressed when females have only one functional Sxl gene, while it is exacerbated when females have three functional Sxl genes. Like wing development, N is 'haploinsufficient' in females for bristle formation in the A5 sternite, and bristle number is increased in heterozygous flies. This bristle phenotype is suppressed when the N⁻/+ females have only one Sxl gene, while it is enhanced when the females have three Sxl genes. Finally, the female lethal effects of a combination of loss of function N alleles can be suppressed by reducing the Sxl dose. Taken together, these genetic interactions argue that Sxl must negatively regulate the N pathway. Moreover, in each of these contexts, the regulatory interactions between Sxl and N must be independent of both the Sxl→tra→dsx-fru regulatory cascade and of the msl dosage compensation system. The reason for this is that Sxl is not haploinsufficient for either tra splicing or for turning off the msl-2 dosage compensation system, and in females heterozygous for Sxl, both of these regulatory pathways are fully in the female mode. Likewise, adding an extra dose of Sxl would not hyperfeminize tra nor would it further repress msl-2 translation. In this context, it should also be pointed out that Sxl negatively regulates its own expression through binding sites in the UTRs of Sxl mRNAs. Because of this negative autoregulatory feedback loop, the levels of Sxl protein in both Sxl⁻/+ and Sxl^dup/+ females are maintained close to that in wild-type females. Thus, the effects of Sxl on N activity are likely to be underestimated in genetic interaction experiments (Penn, 2007).

These is a substantial upregulation of N protein in Sxl⁻ follicle clones. This upregulation is independent of the Sxl→tra→dsx-fru regulatory cascade; however, in this case, it is suspected that two factors likely contribute to the observed increase in N protein. The first is the loss of Sxl regulation, while the second is the activation of the msl-2 dosage compensation system in the complete absence of Sxl activity. As the latter is expected to generate only a 2-fold increase in N expression, it would not fully account for the effects of losing Sxl activity in the clones (c.f., the N levels in adjacent stage 10 Sxl⁺ and Sxl⁻ follicle cells) (Penn, 2007).

Finally, like the two known targets for translational regulation by Sxl, msl-2, and Sxl, N mRNA has multiple Sxl binding sites in its UTRs. Moreover, as would be expected if Sxl directly downregulated N protein accumulation by controlling the translation of N message, Sxl binds to N mRNAs in ovaries. It is interesting to note that the configuration of Sxl binding sites in N mRNAs is quite similar to msl-2. Both mRNAs have two Sxl binding sites in the 5′UTR and four in the 3′UTR. In spite of the similarity in the number and distribution of Sxl binding sites, Sxl repression of N mRNA translation must differ from its repression of msl-2 mRNA translation because unlike Msl-2, N protein is expressed in females. One factor that might account for this difference is that repression of msl-2 mRNA translation by Sxl depends upon corepressors that interact with sites in the 3′UTR located adjacent to the Sxl binding sites; however, these putative corepressor recognition sequences are not present next to the Sxl binding sites in the N 3′UTR (Penn, 2007).

The N signaling pathway plays a central role in fly development because of its ability to specify alternative cell fates. Since most of the tissues and cell types in which the N pathway functions are present in both males and females, an obvious question is how Sxl can deploy this pathway to generate sex-specific differences in morphology. The results indicate that in common tissues, Sxl is able to generate sex-specific differences by changing the level of N activity. Thus, in the A5 sternite, the number of bristles in females is greater than in males, and this difference is due to the downregulation of N by Sxl in female flies. As in other parts of the adult cuticle, bristle formation in A5 depends upon the level of N activity. The number of bristles is inversely proportional to N activity, and N heterozygous females have a greater number of bristles than wild-type females. This difference can be suppressed by reducing Sxl activity and magnified by increasing Sxl activity. Excess Sxl activity can also cause an increase in the number of A5 bristles in females that are wild-type for N. It is reasonable to suppose that this general downregulation of N by Sxl will contribute to other morphological differences between males and females that are independent of the Sxl→tra→dsx-fru regulatory cascade such as bristle number in other parts of the adult cuticle, size of tissues and organs, and perhaps some as yet unknown aspects of nervous system development (Penn, 2007).

Since the ovary is only present in females the developmental context for Sxl-N regulatory interactions is different from most other tissues in the fly. Like the wing and sternites, Sxl negatively regulates N in the ovarian follicular epithelium. When Sxl activity is lost in follicle cells, a variety of defects were observed in the development of this epithelium, including egg-chamber packaging defects, ectopic polar cells, and extra-long interfollicular stalks. This spectrum of phenotypes closely resembles those seen when there is excess N activity and argues that N must be inappropriately upregulated in the follicular epithelium when Sxl is lost. Consistent with this suggestion, elevated levels of N protein are found in Sxl clones. With the possible caveat that the MSL dosage compensation system is likely activated in the absence of Sxl and thus probably contributes to the upregulation of N protein, these observations suggest that Sxl plays an important role in mediating N specification of cell fate as the follicular epithelium develops. This view is supported by the reciprocal patterns of N and Sxl protein accumulation in the germarium of wild-type females. Follicle cells expressing high levels of N in the germarium have only little cytoplasmic Sxl, while lower levels of N are found in follicle cells that have high amounts of cytoplasmic Sxl. If, as is suspected, Sxl regulates N at the level of translation, the turnover of cytoplasmic Sxl and/or its relocalization to the nucleus would be expected to lead to the upregulation of N protein expression. Conversely, in cells that retain abundant cytoplasmic Sxl, N expression should remain repressed. Since the cells in the germarium that are induced to express high levels of N are thought to be the progenitors of the stalk and polar cells, releasing N mRNA from translation inhibition by Sxl would be expected to facilitate the specification of these cell types by the N-signaling pathway (Penn, 2007).

This raises the question of why cytoplasmic Sxl turns over and/or is targeted to the nucleus in these particular cells. In the germline and in the wing disc, turnover and nuclear targeting of Sxl protein are known to be mediated by the hh signaling pathway. It seems possible that hh signaling might also promote the turnover/nuclear targeting of Sxl in these particular somatic follicle cells. Consistent with this idea, overexpression of hh in follicle cells leads to at least one of the phenotypes that is seen when Sxl activity is lost (or N is ectopically activated), the expansion of interfollicular stalks. If hh is responsible for the turnover/nuclear targeting of Sxl, the Sxl gene would provide a mechanism for linking the hh- and N-signaling pathways in the specification of stalk and polar cell fates. Further studies will be required to test this model (Penn, 2007).