Sexually dimorphic regulation of the Wingless morphogen controls sex-specific segment number in Drosophila

Abstract

Sexual dimorphism is widespread throughout the metazoa and plays important roles in mate recognition and preference, sex-based niche partitioning, and sex-specific coadaptation. One notable example of sex-specific differences in insect body morphology is presented by the higher diptera, such as Drosophila, in which males develop fewer abdominal segments than females. Because diversity in segment number is a distinguishing feature of major arthropod clades, it is of fundamental interest to understand how different numbers of segments can be generated within the same species. Here we show that sex-specific and segment-specific regulation of the Wingless (Wg) morphogen underlies the development of sexually dimorphic adult segment number in Drosophila. Wg expression is repressed in the developing terminal male abdominal segment by the combination of the Hox protein Abdominal-B (Abd-B) and the sex-determination regulator Doublesex (Dsx). The subsequent loss of the terminal male abdominal segment during pupation occurs through a combination of developmental processes including segment compartmental transformation, apoptosis, and suppression of cell proliferation. Furthermore, we show that ectopic expression of Wg is sufficient to rescue this loss. We propose that dimorphic Wg regulation, in concert with monomorphic segment-specific programmed cell death, are the principal mechanisms of sculpting the sexually dimorphic abdomen of Drosophila.

Brachycera, higher diptera that include drosophilidae, exhibit an evolutionary trend toward reduced abdominal size that contributes to swift, maneuverable flight (1). Such reduction is especially pronounced within the infraorder Muscomorpha. Within this group of flies, abdominal reduction is sexually dimorphic such that adult males have fewer segments than females. Lower diptera, which includes mosquitoes and midges, retain ancestral morphology with respect to segment number; both adult males and females generate eight abdominal segments. In Muscomorpha the most posterior adult abdominal segments (all or a subset of segments A5–A8) are modified in females, usually as a telescoping ovipositor, whereas corresponding segments are absent in males (2). In all diptera, segment number is monomorphic during embryogenesis and larval development, reflecting the basal insect body plan of three head, three thoracic, and 11 abdominal segments. For most diptera, only embryonic abdominal segments 1–8 generate adult abdominal tissue (only segments 1–7 in the drosophilidae). The more posterior embryonic segments contribute to the adult genitalia. During pupation, sex-specific developmental programs are deployed that sculpt sexually dimorphic segment morphology and number.

The posterior abdomen of Drosophila melanogaster serves as an excellent model to study the development of these sex-specific morphologies. Posterior abdominal segment identity, morphology, and number in both sexes is regulated by the Hox protein Abdominal-B (Abd-B) (3, 4). Abd-B expression is initiated during embryogenesis in parasegments 10–14 (roughly equivalent to embryonic segments A5–A8) in an increasing gradient, highest in the most posterior parasegments (3, 5, 6). This gradient is maintained through larval and pupal development where Abd-B patterns morphology of the adult posterior abdominal segments and genitalia (7). In classic homeotic fashion, reduced Abd-B expression in hypomorphic mutants or heterozygous null backgrounds transforms segment A7 toward more anterior identity: adult males generate an A7 and that segment in females is morphologically similar to the more anterior segment A6 (8, 9) (Fig. S1C). Conversely, when Abd-B levels are increased or Abd-B is ectopically expressed, anterior segments are transformed toward a posterior identity. For example, in the regulatory mutant Abd-B^Fab-7, Abd-B levels in A6 are equivalent to wild-type A7 levels. As a result, segment A6 of both sexes is transformed to an A7 identity; adult male A6 is absent and female A6 adopts the morphology of wild-type A7 (10) (Fig. S1B).

Because pupal Abd-B expression is sexually monomorphic (Fig. S2), other genetic factors must contribute to patterning sex-specific morphology. The sex-determination transcription factor Doublesex (Dsx), which has distinct male and female isoforms, functions with Abd-B to pattern sex-specific morphology of the adult Drosophila abdomen and genitalia (11–15). For example, Abd-B and Dsx confer sex-specific regulation on bric-a-brac (bab) to generate dimorphic abdominal pigmentation (15, 16). Abd-B and Dsx-female positively regulate bab in segments A5–A7, whereas Abd-B and Dsx-male repress bab in these segments. The role of Dsx in sculpting sexually dimorphic morphologies is evident as homozygous loss-of-function dsx mutants display an intersex phenotype (11). The posterior abdomen of both XX and XY individuals is pigmented as in wild-type males. Furthermore, adult segment A7 develops in both XX and XY intersexes. (Fig. S1D).

Therefore, as with pigmentation, Dsx acts in conjunction with Abd-B to regulate the morphology of the posterior abdomen and the reduction of adult male segment A7. The target genes regulated by Abd-B and Dsx to generate sexually dimorphic segment number have not been identified previously. Here, we report that the morphogen Wingless (Wg) is regulated in a sex-specific and segment-specific manner. We provide evidence that sexually dimorphic Wg regulation acts in concert with increased segment-specific (but monomorphic) programmed cell death to sculpt the posterior abdomen of Drosophila, leading to the loss of male segment A7 during pupation. We provide additional evidence that regulation of signaling through the Drosophila epidermal growth factor receptor (DER) may also contribute to sex-specific abdominal morphology.

Results

Reduction of Terminal Male Segment During Early Pupation.

Adult abdominal segments are derived from imaginal cells (histoblasts) born during embryogenesis (17, 18). As they proliferate, these cells replace the larval epidermal cells (LECs), which are removed from the epithelium by apoptosis and extrusion (19, 20). As in embryogenesis, cells of each segment are compartmentalized into anterior and posterior populations (21). The sclerotized cuticle plates of the adult abdomen (dorsal tergites and ventral sternites) are generated by a subpopulation of anterior cells. Other anterior cells produce flexible pleural cuticle that joins the plates laterally as well as a lateral pair of spiracles; openings to the adult tracheal system. Cells of the posterior compartment generate pleural cuticle that joins tergites and sternites along the anterior–posterior axis of the abdomen. Anteroposterior and dorsoventral patterning occurs in all segments, generating regional and segment-specific morphologies mediated by segmentation, Hox, and sex-determination factors (22–26). For example, the first abdominal segment (A1) does not form a sternite in either sex and the sixth abdominal sternite (A6) of males lacks bristles and adopts a characteristic horseshoe shape (Fig. 1A). The male A6 tergite is noticeably enlarged compared with more anterior segments and males lack both A7 tergite and sternite. Female A7 histoblasts fail to fuse at the dorsal midline, producing lateral plates (hemitergites) with characteristic triangular morphology (Fig. 1A).

Fig. 1.

Sexually dimorphic segment number in D. melanogaster. (A) Male and female adult D. melanogaster cuticles. The terminal seventh abdominal segment (A7) is modified in females and absent in males. (B) DAPI- stained male and female pupae (26 h APF). Development of male A7 is initially equivalent to female A7. (C and D) Engrailed-gal4; uas-GFP expression (anti-GFP, green) time-course analysis reveals caudal movement of the male A6 posterior compartment (C, arrowhead) and rapid loss of anterior A7. DAPI staining in blue reveals pupal morphology and loss of aA7 cells. Note male pA6 buckles around the developing A7 spiracle (asterisks) before completely passing this structure. No caudal movement of pA6 is observed in female pupae, however the female A7 does retract into the abdomen, giving the appearance of reduction (D). Figs. S1 and S2 show complete time course.

We found that the absence of adult male A7 is neither the result of a failure to generate histoblasts nor is it due to the nonproliferation of histoblasts during early pupation (Fig. 1B). However, by midpupation (40 h after puparium formation, APF) male A7 is no longer present (Fig. 1C). To trace how and when the terminal male segment is lost, we followed the fate of cells within the posterior compartment of each segment during the 24- to 40-h APF developmental interval. The Engrailed (En) transcription factor is restricted to the posterior compartment of each segment, where it promotes posterior segmental identity/fate (7, 22, 27, 28). We monitored En expression in En-gal-4/uas-GFP pupae. En expression is sexually monomorphic at the beginning of this period and is expressed in both proliferating histoblasts and LECs of each abdominal segment as previously reported (7) (Fig. 1 C and D and Figs. S3 and S4 for full time series). However, once the border cells (LECs that separate developing adult segments) undergo apoptosis, male posterior A6 (pA6) histoblasts begin moving caudally. Concomitantly, the number of anterior A7 (aA7) histoblast cells rapidly decreases (Fig. 1C). By 40 h APF, pA6 cells have passed the A7 spiracle and are juxtaposed to or associated with pA7. As a result, no A7 tergite is produced in adult males and the A7 spiracle is associated with the A6 tergite. Posterior movement of pA6 cells is not observed in female pupae (Fig. 1D). These observations indicate that the absence of an adult male A7 segment is the result of dramatic tissue reorganization during early pupation. Such reorganization may be the result of differential cellular proliferation, apoptosis, or both.

Sex-Specific Regulation of the Morphogen Wingless Is Correlated with Male Reduction.

Engrailed functions in a hierarchical regulatory cascade establishing segment compartment boundaries and identities during embryogenesis. Additionally, En cooperates with other segmentation genes to maintain compartment boundaries during pupation (22, 24, 27, 28). To test whether the observed dynamic changes in male En expression are due to dimorphic regulation of other segmentation genes, we surveyed expression of key members of this cascade with known roles in adult abdominal patterning (22, 23, 25, 26). Anterior and posterior compartment identities are maintained via reciprocal interactions of neighboring cells. En-expressing cells produce the diffusible ligand Hedgehog (Hh), which binds to the receptor Patched (Ptc) on anterior cells. This interaction leads to activation and elevated expression of the transcription factor Cubitus interruptus (Ci), which in turn promotes anterior fate by regulating a number of target genes, including the signaling molecules Wingless (Wg) and Decapentaplegic (Dpp). We found that expression of Hh, Ptc, Ci, and Dpp are each sexually monomorphic, as is expression of another regulator of segment polarity, Optimotor-blind (Omb) (Fig. S2).

However, expression of Wg is selectively absent from pupal male A7 histoblasts (Fig. 2E). Wingless, best known for its role in establishing and maintaining segment polarity during embryogenesis, is a highly pleiotropic signaling protein with many independent roles in development (29–32). Most relevant, Wg is required for pattern formation in the adult Drosophila abdominal epidermis (33). When Wg function is abrogated during pupal development in a temperature-sensitive Wg mutant, the anterior compartment of all abdominal segments fails to generate characteristic cuticle and bristles (Fig. S1F). A requirement for Wg function in anterior adult segment development is therefore consistent with an association between the absence of Wg expression in segment A7 of male pupae and adult segment loss.

Fig. 2.

Wingless expression is sexually dimorphic and controlled by Abd-B and Dsx levels. (A and E) Wg (Anti-Wg, red) is expressed in dorsal and ventral histoblast nests (and the developing spiracle) in all female segments but is absent from the dorsal and ventral nests of male. In Abd-B^Fab-7 pupae Abd-B levels in A6 are increased compared with wild type. (B and F) Wg is additionally repressed in male, but not female, A6 (C and G). In a heterozygous Abd-B loss-of-function mutant (Abd-B^M2), decreased Abd-B levels lead to derepression of Wg in male A7. Wg expression remains unaltered in female A7. (D and H) Similarly, in loss-of-function Dsx pupae, Wg expression is restored in males and unaltered in females. All pupae are 26 h APF.

To further test the association between Wg expression and loss of the terminal male segment as well as to investigate the potential genetic control of sex-specific Wg expression, we examined whether male segment-specific Wg regulation responded to changes in Abd-B and Dsx expression. Indeed, in gain-of-function Abd-B^Fab-7 males Wg was also repressed in the A6 segment (Fig. 2F). Conversely, in heterozygous Abd-B or homozygous Dsx loss of function males, Wg expression was derepressed in male A7 (Fig. 2 G and H). Wingless expression was unaltered in females of each genotype (Fig. 2 B–D). These results suggest that these two regulators govern the sex- and segment-specific repression of Wg expression.

Segment and Sex-Specific Transformation of Segment Boundaries.

To understand the developmental consequences of male-specific Wg regulation as they relate to segment loss, we examined its role in maintaining segment polarity. Because Wg- and En-expressing cells are involved in a reciprocal regulatory loop, the absence of Wg could affect male pA6 En expression (27). We investigated the possibility that all, or a part, of the caudal movement of male pA6 results from compartmental transformation, that is, from the de novo expression of En in cells formerly fated to anterior A7. There is experimental precedent for similar transformations during Drosophila embryonic development (34, 35).

To examine this possibility, we simultaneously monitored En expression indirectly via the Engal4; uas-GFP reporter and directly using an anti-En antibody. GFP expression in this assay requires two rounds of transcription and translation (first gal4, then GFP); therefore, endogenous En expression should precede GFP production if compartmental transformation occurs (35). We observed de novo endogenous En expression in a subset of male aA7 cells between 28 and 36 h APF (Fig. 3A and Fig. S5). A similar transformation also occurs in male aA6 in the Abd-B^Fab-7 regulatory mutant (Fig. 3C and Fig. S6). However, no de novo En expression occurs in either wild-type or Abd-B^Fab-7 females (Fig. 3 B and C and Figs. S7 and S8). Therefore, during pupal development, a subset of male anterior A7 cells is transformed to posterior fate and becomes part of the more anterior segment A6.

Fig. 3.

Compartmental transformation contributes to male A7 reduction. En expression monitored indirectly via En-gal-4; uas-GFP (anti-GFP, green) and directly with anti-En antibody (red) reveals de novo expression of En in cells formerly fated as male anterior A7. (A) In males, as pA6 caudal movement occurs, a subset of aA7 cells begins expressing En protein before GFP, indicating they recently acquired posterior fate. (B) No de novo En expression is observed in females. (C) Sex-specific compartmental transformation also occurs in A6 of Abd-B^Fab-7 males (Left). No de novo En expression is observed in female Abd-B^Fab-7 pupae (Right). Arrowheads point to pA6. Figs. S5–S8 show complete time-course analyses.

To address whether male-specific repression of Wg contributes to the observed compartmental transformation of aA7, we monitored En expression in Wg-ts pupae shifted to the nonpermissive temperature at the onset of pupariation. Compared with heterozygote siblings, posterior compartments of all segments in both sexes were several cells wider in homozygote Wg-ts pupae, suggesting that all segments underwent a transformation similar to that observed in wild-type male A7 (Fig. S9B). Furthermore, these results argue that sex and segment-specific repression of Wg permit the compartmental transformation observed in male aA7.

If compartmental transformation were the sole mechanism responsible for reduction of male A7, the width of pA6 (the number of En-expressing cells) should continue to increase over time as more aA7 cells are incorporated into pA6. However, the width of male pA6 does not increase beyond 12–15 nuclei; compared with 6–8 nuclei in more anterior segments and female A6 (Fig. 3 and Figs. S5–S8). Therefore, additional mechanisms must contribute to the reduction of male segment A7.

We considered the possibility that reduction of male A7 may result from a wave-like migration of En expression (pA6) across the field of aA7 cells. As aA7 cells are transformed to pA6 fate, some pA6 cells may likewise be transformed to aA6 identity. To investigate this, we performed lineage-tracing analyses in w; En-gal4/uas-GFP; uas-flp/Act5Cfrt-stop-frt-LacZ flies. In these flies, GFP is expressed in cells with posterior identity (En positive) and gal4 activated flippase mediates excision of a stop cassette leading to constitutive and autonomous LacZ expression (36). Any cell, regardless of current identity, derived from an En-expressing cell, will express LacZ. Therefore, posterior cells transformed to anterior identity should express LacZ but no longer express GFP. No such transformation was observed in male pA6 or other segments (Fig. S9C). We conclude that, in addition to partial segmental transformation, other mechanisms must also contribute to the loss of male A7.

Roles of Mitosis and Apoptosis in Male Segment Reduction.

We next assayed the possible contributions of cell proliferation and programmed cell death to male segment loss. We compared the number of mitotic and apoptotic events for segments A5–A7 in both sexes during the period of male A7 reduction. Males had fewer mitotic and higher apoptotic events in the dorsal nest of all segments compared with female counterparts (Fig. S10A). Importantly, male A7 mitosis was significantly and strongly reduced compared with more anterior segments and female A7. These data are consistent with a mitogenic role for Wg in certain developmental contexts and suggest that repression of Wg in male A7 contributes to decreased cell proliferation as a component of segmental reduction (37).

Previous studies of pupal epithelial morphogenesis found ∼3% of histoblast cells undergo programmed cell death during histoblast nest proliferation (20). Our results revealed that apoptosis is higher in all male segments compared with female counterparts; however, there was no significant quantitative increase in cell death in male A7 dorsal nests compared with more anterior segments or female A7 (Fig. S10B). However, beginning ∼26 h APF, we observed a stereotypical pattern of cell death in both sexes in cells surrounding the developing spiracle of all segments (Fig. S10 D and E). Apoptosis in spiracle-associated cells was significantly higher for both sexes in A7 compared with A6 (Fig. S10C). The increased cell death was greatest at 30–34 h APF, the interval during which male A7 is reduced. Therefore, we propose that sex-specific Wg repression acts in concert with a sexually monomorphic, segment-specific increase in apoptosis to drive the reduction of the terminal male segment.

Roles of Wingless and Apoptosis in Male Segment Reduction.

We next investigated functional contributions of Wg regulation and apoptosis to male segment reduction through misexpression of either Wg or the antiapoptotic bacculovirus gene p35. Two Dsx-gal4 drivers have been reported with dynamic expression throughout larval, pupal, and adult tissues (38, 39). Importantly for our assays, these drivers are expressed throughout the developing imaginal abdomen epithelium.

As previously reported, we observed that blocking abdominal and genital apoptosis led to failed male genital rotation and occasional supernumerary bristles on the A6 sternite (40). Additionally, the A7 spiracle, which in wild-type males is associated with the A6 tergite, remained posterior to A6 in Dsx-gal4/uas-p35 adults (compare Fig. 4A to 4D). No A7 tergite developed; however, there was excess pleural tissue posterior to the A6 tergite. Furthermore, the A6 tergite was significantly narrower than A6 of control males (Fig. 4 E and F). These results indicate that apoptosis aids in reducing the terminal male abdominal segment by eliminating A7 cells and suggest a possible mechanism for the characteristically enlarged wild-type male A6 tergite. The abdominal epithelium is continuous between segments and tightly associated with the overlying pupal membrane. As A7 histoblasts are removed by apoptosis, our data suggest that the abdominal epithelium expands caudally (perhaps passively via adhesive forces). Indeed, following loss of male A7, the density of nuclei in male A6 is noticeably lower than in more anterior segments (Fig. 1 and Fig. S2).

Fig. 4.

Contributions of cell death and Wg regulation to male terminal segment reduction. (A) Ventral view of wild-type male A6. Note association of A7 spiracle (arrowhead) with A6 tergite and characteristic horseshoe morphology of the A6 sternite (arrow). (B) Loss-of-function Abd-B^M2/Abd-B^iabC6 transheterozygote male. A7 segment is restored (arrowhead) and A6 sternite transformed toward anterior identity. (C) Loss-of-function Dsx¹/Dsx¹⁸ transheterozygote. A7 is partially restored in this hypomorphic combination (arrowhead) and the A6 sternite is transformed toward anterior identity. (D) The A7 spiracle (arrowhead) remains in a posterior position to the A6 tergite in Dsx-gal4/uas-p35 males. (E and F) Blocking apoptosis leads to significant decrease in A6 tergite width compared with control. Each panel shows two segments (Upper) A5 and (Lower) A6 for (E) uas-p35/+; average A5 width = 220.5 μm (± 5.7 μm), average A6 width = 283.7 μm (± 5.9), and (F) Dsx-gal4/uas-p35; average A5 width = 215.0 μm (± 3.5 μm), average A6 width = 229.5 μm (± 3.3 μm) (n = 10 for each measurement). (G) uas-Wg/+; Dsx-gal4/+ males generate A7 tergite (arrowhead) and sternite (arrow). Additionally, the A6 sternite is transformed toward anterior identity. (H) Ectopic Wg expression prevents transformation of aA7. At 34 h APF, A7 cells are still present in uas-Wg/+; Dsx-gal4/+ males and expansion of pA6 (En-positive cells, arrowhead) has not occurred. (I) uas-vn/+; Dsx-gal4/+ males likewise generate A7 tergite (arrowhead) and sternite (arrow) and the A6 sternite is transformed. (J) Combined ectopic Wg and vn leads to less robust A7 male transformation. Note absence of A7 tergite.

Importantly, ectopic Wingless expression is sufficient to restore male A7 cuticle development (Fig. 4G). Male uas-Wg/+; Dsx-gal4/+ flies likewise developed excess tissue posterior to A6, and A7 spiracles remained posterior to the A6 tergite. In contrast to uas-p35 males, this excess tissue was sclerotized and pigmented and had therefore adopted an anterior cuticle fate. Ventral histoblasts were also transformed, generating an A7 sternite. Additionally, ectopic Wg expression resulted in partial transformation of the male A6 sternite toward more anterior identity. Compared with wild-type males, the A6 sternite of uas-Wg/+; Dsx-gal4/+ males was rectangular and developed supernumerary bristles. This transformation was similar, although not as complete, as that observed in loss-of-function Abd-B and Dsx males (Fig. 4 A–C and G). Whereas Wg is expressed in both dorsal and ventral histoblast nests of male A6 (Fig. 1), these data suggest that in addition to adult segment number, regulation of Wg expression may also contribute to sexually dimorphic A6 sternite morphology. Male-specific A6 sternite morphology may result from subtle regional repression or regulation of Wg levels in ventral A6 and warrants further scrutiny. By 34 h APF, A6 En expression in male uas-Wg/+; Dsx-gal4/+ pupae had not expanded caudally and a population of anterior A7 cells was still present (Fig. 4H).

The transformed A7 cuticle of uas-Wg/+; Dsx-gal4/+ males is narrower than more anterior segments and devoid of bristles. Because we hypothesize the combination of Wg repression and segment-specific apoptosis contributes to segment loss, the failure of ectopic Wg to robustly rescue A7 development may reflect persistent segment-specific cell death. Unfortunately, we were unable to assay the combined phenotype of Wg misexpression and apoptosis inhibition as uas-wg/+; uasp35/+ was lethal in combination with Dsx–gal4 and other drivers tested.

Discussion

We have shown that the reduction of the adult Drosophila male terminal abdominal segment during pupation is the result of a combination of segment compartmental transformation, suppression of cell proliferation, and increased programmed cell death. These processes are controlled by segment-specific regulation of at least two developmental pathways. In the first case, expression of the Wingless morphogen is selectively repressed in male pupae in the developing A7 segment. In the second case, programmed cell death is heightened in the terminal segment of both sexes. We suggest that male-specific repression of Wg suppresses cell proliferation below a threshold capable of countering the monomorphic increase in apoptosis. At the same time, the absence of Wg promotes a partial transformation of A7 tissue to more anterior fate.

The mechanism through which Wg is repressed in male segment A7 has not yet been fully elucidated. Our genetic analyses suggest that negative regulation of Wg is mediated both by Abd-B and the Dsx male isoform. Whether regulation of Wg expression by either protein is direct remains to be determined. However, there is precedent for these two transcription factors acting together to control sexually dimorphic pigmentation in the Drosophila abdomen through direct regulation of bab (15, 16). Additionally, Abd-B and Dsx act together to regulate a number of targets in the genital disk (12, 13, 41). Therefore, we anticipate that Abd-B and Dsx act upon a tissue-specific enhancer of the Wg locus to mediate its repression in male segment A7.

Less clear is the mechanism of the genetic regulation of the monomorphic increase in pupal A7 apoptosis. Our data show a clear increase in spiracle-associated cell death in A7 of both sexes. It has been reported that posterior embryonic apoptosis is positively regulated by Abd-B directly through activation of the proapoptotic gene reaper (42). If similar Abd-B–mediated regulation of apoptosis occurs during pupal development, there must be additional regional controls that restrict these events to those cells surrounding the developing spiracles.

The inability of ectopic Wg to completely restore development of male A7 may reflect dimorphic regulation of other signaling pathways required for proper segment development. Active DER is necessary for differentiation of the lateral tergite (25). We were unable to detect expression of DER or two of its known ligands (spitz and vein) in LacZ enhancer trap stocks. However, ectopic expression of the ligand Vein was sufficient to promote male A7 cuticle development (as well as transformation of A6 sternite morphology) to an extent comparable to that of ectopic Wg expression (Fig. 4I). These data suggest that in addition to Wg regulation, sexually dimorphic DER activity may also contribute to male A7 reduction. Combined Wg and vn misexpression did not produce a more robust A7 transformation. Rather, although the A7 spiracle remained in a position posterior to A6, no A7 cuticle formed in uas-vn/Y; uas-Wg/+; Dsx-gal4/+ males (Fig. 4J). Wg and DER function cooperatively but antagonistically through complimentary expression to pattern developing tissues (43, 44). Therefore, simultaneous misexpression of both Wg and vn may abrogate function of both signaling pathways. More detailed expression studies of components of DER signaling will be necessary to confirm the role of its regulation in this process.

The mechanisms described here may also provide valuable insight into the evolution of segment number. Absence of male postabdominal segments and reduction through modification to the corresponding female segments is a trend, if not a synapomorphy, of the higher diptera. This trait is suggested to have evolved as a means to swift and maneuverable flight within this monophyletic group of flies, the Brachycera, which originated ∼200 Mya (45). The findings here, and future observations in Drosophila, offer testable hypotheses to investigate the evolutionary origins of this character.

Materials and Methods

Fly Stocks.

Otherwise wild-type control flies were w¹¹¹⁸. Fly stocks were obtained from the Bloomington Stock Center except for Dsx-gal4 (gifts from B. Baker, HHMI, Chevy Chase, MD, and S. Goodwin University of Oxford, Oxford, United Kingdom), uas-vn (gift from A. Simcox, The Ohio State University, Columbus, OH) and uas-wg (gift from R. Nusse, Stanford University, Stanford, CA). Stocks and pupae were cultured at 20 °C, except Wg-ts mutant temperature-shift experiments. This stock was cultured at 17 °C and 0hr APF pupae were transferred at 25 °C and allowed to develop to appropriate time points for immunohistochemistry (30 h APF equivalent to 36 h APF at 20 °C).

Lineage Tracing Experiment.

A uas-GFP; act5C-FRT-stop-FRT-lacZ stock was generated from Bloomington stocks 4775 and 6355. This stock was crossed with En-gal4; uas-flp to generate w; En-gal4/uas-GFP; uas-flp/Act5Cfrt-stop-frp-LacZ. Upon activation by En-driven Gal4, flippase mediates excision of the stop cassette in Act5Cfrt-stop-frt-LacZ. This leads to lineage autonomous expression of LacZ, allowing identification of cell lines that no longer, but once did, express En.

Immunocytochemistry.

Pupae were collected after larval movement had ceased (0 h APF) and were incubated in a humid chamber at 20 °C to specified hours APF. Pupae were dissected bilaterally and fat body and other internal tissue removed by pipetting, leaving only the abdominal epidermis adhered to the pupal membrane. Samples were fixed (4% paraformaldehyde, 0.2% deoxycholic acid; PBS) at room temperature for 1 h and blocked in BSA (10%) before overnight primary antibody incubation (4 °C). Primary antibodies were chicken anti-GFP (1:500; abCam), rabbit anti-En (1:1,000; Santa Cruz), rabbit anti–β-galactosidase (1:500; MP Biomedicals), mouse anti–Abd-B (1A2E9) (1:500; Developmental Studies Hybridoma Bank), rabbit antiactive–Caspase-3 (1:50; Cell Signaling), rabbit antiphospho–Histone-3 (1:1,000; Cell Signaling), and mouse anti-Wg (4D4) (1:500 Developmental Studies Hybridoma Bank). Secondary antibodies were Alexa-Fluor (Invitrogen) goat antichicken-488, antimouse-555, and antirabbit-555 (all at 1:500). Nucleii were stained with diamidino-2-phenylindole. Samples were removed from pupae cases and mounted in depression-well microscope slides (0.8 mm) in Vectashield under a coverslip. Imaging was performed on a Nikon 90i microscope equipped with an Optigrid structured illumination accessory (Qioptiq). Image stacks were projected to generate final images using the NIS Elements software package.

Cuticle Preparations.

For dorsal and ventral cuticle preparations, adults were incubated in 10% KOH (to dissolve internal tissue) for 1 h at 65 °C and washed in PBS. Abdomens were dissected in PBS and were cleared and mounted in polyvinyl lactophenol under a coverslip.