Wingless, Decapentaplegic and EGF Receptor signaling pathways interact to specify dorso-ventral pattern in the adult abdomen of Drosophila

While much effort has been devoted to understanding the mechanisms of pattern formation in the embryo and imaginal discs of Drosophila, the adult abdomen has attracted little attention. However, its unique mode of development raises many intriguing questions about its patterning and offers a number of practical advantages. The abdominal epidermis has a simple geometry (each segment can be regarded as a two-dimensional rectangular field), a wide range of identifiable cell fates, and completely uniform polarity. The absence of cell competition in the abdomen greatly simplifies clonal analysis. Most importantly, in contrast to the imaginal discs (Serrano and O’Farrell, 1997), patterning and growth in the abdomen appear to be uncoupled. Even drastic pattern changes have little effect on cell proliferation, and do not affect the final shape and size of abdominal segments.

The adult abdominal epidermis develops during the pupal stage from groups of cells called histoblast nests, which differ from imaginal discs in two important respects. First, abdominal histoblasts do not invaginate during embryogenesis, but remain part of the larval epidermis and secrete larval cuticle. Second, they do not proliferate during the larval stages. After pupariation, histoblasts multiply rapidly and migrate to replace the polyploid larval epidermal cells (LEC). As the individual nests grow and merge, LEC are destroyed only upon contact with histoblasts, so that the continuity of the pupal epidermis is maintained at all times. The replacement of LEC by histoblasts is completed by 40-42 hours after puparium formation (APF) (Madhavan and Madhavan, 1980). The epidermis of each abdominal segment is produced by three bilateral pairs of histoblast nests: the anterior dorsal nests produce the tergite, the posterior dorsal nests form the flexible intertergal cuticle, and the ventral nests produce the sternite and pleura. In addition, a spiracular nest produces a small patch of epidermis around each spiracle.

The antero-posterior (AP) patterning of abdominal segments is controlled partly by the Hedgehog (Hh) morphogen (Struhl et al., 1997a,b; Kopp et al., 1997) and partly by an hh-independent mechanism of unknown identity (Kopp et al., 1997; A. Kopp and I. Duncan, unpublished). The patterning function of Hh is mediated primarily by optomotor-blind (omb) (Kopp and Duncan, 1997), which encodes a putative T-box transcription factor (Pflugfelder et al., 1992). Surprisingly, the secreted morphogens Wingless (Wg) and Decapentaplegic (Dpp), which act downstream of Hh in the long-range AP patterning of imaginal discs, are not involved in AP patterning in the abdomen (Shirras and Couso, 1996; Struhl et al., 1997a,b; Kopp et al., 1997). Rather, Hh functions directly to specify cell fates along the AP axis (Struhl et al., 1997b).

The functions of Wg and Dpp in the adult abdomen are limited to dorso-ventral (DV) patterning. Wg has been shown by Shirras and Couso (1996) to promote tergite and sternite, as opposed to pleural, cell fates. Here, we report that signaling by the Drosophila EGF receptor (DER) acts in concert with Wg to promote tergite and sternite identities, and that Wg and DER activities are opposed by Dpp signaling, which promotes pleural identity. We show that Wg and Dpp limit each other’s expression domains by mutual transcriptional repression, and that cell fates are determined by the balance between competing Wg and Dpp activities. Finally, we demonstrate that Dpp signaling controls tergite pigmentation near the dorsal midline.

In all figures, anterior is up and dorsal is to the right. Preparation of adult abdominal cuticles and pupal body walls, antibody and X-Gal staining and in situ hybridization were performed as described (Kopp et al., 1997). For X-Gal/antibody double staining, unfixed pupal abdomens were stained in standard X-Gal staining solution for 10-20 minutes at room temperature, fixed in 3% formaldehyde in PBS for 20 minutes, and processed according to standard antibody staining procedures.

wg expression was detected using the enhancer trap lines P{ry^+t7.2=PZ}wg^r0727 (G. Rubin, BDGP) and P{ry^+t7.2=PZ}wg⁰²⁶⁵⁷ (A. Spradling, BDGP). Both lines (referred to as wg-lacZ) accurately reproduce the pattern of wg transcript expression in the pupal abdomen (not shown), and were used interchangeably. dpp expression was detected using the P{ry^+t7.2=PZ}dpp¹⁰⁶³⁸ enhancer trap (dpp-lacZ; Twombly et al., 1996) and by RNA in situ hybridization. The omb-GAL4² enhancer trap was isolated by Y. H. Sun (unpublished). The expression of GAL4 lines was monitored using the UAS-lacZ reporter of Brand and Perrimon (1993). We find that most enhancer traps, including dpp-lacZ and wg-lacZ, become expressed ubiquitously in LEC after 10-12 hours APF, and thus cease to be useful for monitoring gene expression in LEC.

The UAS lines used were UAS-DN-DER (O’Keefe et al., 1997); UAS-wg (Lawrence et al., 1995); UAS-dpp (dpp^UAS.cSa) (Staehling-Hampton et al., 1994); UAS-DN-dTCF (dTCF^ΔN) (van de Wetering et al., 1997); UAS-λtop (constitutively activated DER; Queenan et al., 1997); and UAS-arm^S10 (constitutively activated Armadillo; Pai et al., 1997). GAL4 drivers used were T155-GAL4 (Harrison et al., 1995) and dpp-GAL4 (GAL4^dpp.blk1) (Staehling-Hampton et al., 1994). UAS/GAL4 combinations were kept at 17°C until the late larval or early pupal stage, when they were shifted to the desired temperature (25-31°C). Flp-out clones were generated using Tub>f⁺>dpp and hs-FLP122 (Zecca et al., 1995), Act>CD2>GAL4 (Pignoni and Zipursky, 1997) and hs-FLP1 (Dang and Perrimon, 1992). Clones were induced by heat shocking hs-FLP122; Tub>f⁺>dpp or hs-FLP1/Act>CD2>GAL4; UAS-wg/UAS-lacZ or hs-FLP1/ Act>CD2>GAL4; UAS-dpp/UAS-lacZ first or second instar larvae at 37°C for 10-30 minutes in a circulating water bath.

top^CA/top^2W74 heterozygotes show a severe loss of DER function at 25°C, but not at 18°C (Clifford and Schupbach, 1992). Mad¹² is a presumed null allele of Mothers against dpp, the transcription factor that activates dpp target genes (Sekelsky et al., 1995). stubby chaete (stc), a mutation that reduces the size of bristles and trichomes (Struhl et al., 1997b), was used as a cell-autonomous marker for Mad¹² clones.

The cuticle of the adult abdomen is differentiated into three distinct regions along the DV axis (Fig. 1A). The ventralmost structure of a typical abdominal segment is the sclerotized, unpigmented sternite. The ventro-lateral position is occupied by the flexible, unsegmented pleural cuticle, which is also present mid-ventrally between the sternites of consecutive segments. Dorsally, each segment is composed of a sclerotized, pigmented tergite and flexible, unpigmented intertergal cuticle that is normally folded underneath the tergite. All cells in the sternites, pleura and tergites secrete 3-4 trichomes per cell. The wide-based, curved trichomes secreted by pleural cells are distinct from the thinner, straighter sternal and tergal trichomes (Fig. 1B). In addition, sternites and tergites, but not the pleura, contain arrays of bristles. Dorsoventral patterning is also present within tergites, as the dark pigment band at the posterior edge of each tergite is wider medially than laterally. Some segments deviate from the typical pattern. For example, the first abdominal segment (A1) lacks a sternite. Also, in the male, A7 lacks both a sternite and a tergite, A6 lacks bristles on its sternite, and A5 and A6 have uniformly darkly pigmented tergites (Fig. 1A).

The adult abdominal epidermis is formed during the first 40-42 hours of pupal development. At pupariation, the abdominal epidermis is composed predominantly of the polyploid LEC, which are easily distinguishable from the much smaller, diploid histoblasts. At this stage, the anterior dorsal histoblast nest (aDHN) contains 13-19 cells, the posterior dorsal nest (pDHN) contains 5-8 cells and the ventral nest (VHN) contains 9-13 cells (Madhavan and Madhavan, 1980). Histoblasts begin to proliferate and migrate to supplant the LEC soon after pupariation. At 18-20 hours APF, the aDHN and pDHN merge to form a single dorsal histoblast nest (DHN) (Fig. 2). The DHN merges with the VHN and the spiracular anlage between 20 and 28 hours APF (Fig. 3). The spiracle, located at the lateral midline, marks the boundary between ventral and dorsal histoblasts, and eventually the boundary between the pleura and the tergite. The fusion of histoblast nests of consecutive segments begins at 28 hours APF and proceeds until 40-42 hours APF, when the formation of a continuous adult epidermis is completed by the fusion of contralateral nests at the dorsal and ventral midlines. Morphological differentiation of the epidermis into sternite, tergite and pleural territories becomes evident shortly thereafter. These regions can be distinguished at 45 hours APF by differences in the shape and arrangement of cells and by the pattern of developing adult muscles (Fig. 4).

The origin of the adult wg and dpp patterns can be traced to the early pupal stage. At the time of fusion of the aDHN and pDHN (18 hours APF), wg and dpp expression domains encompass both adult and larval cells, and are limited to the posterior region of the anterior compartment (Fig. 2). Within this zone, the patterns of wg and dpp are largely complementary along the DV axis. wg is expressed in a dorsal-posterior sector of the aDHN and the adjacent dorso-lateral LEC, as well as in a ventral sector of the VHN and the adjacent ventro-lateral LEC (Shirras and Couso, 1996; Fig. 2A,D,E). wg is not expressed in the dorsal part of the VHN, in the ventral part of the aDHN or in the lateral LEC between the two nests. wg expression is also weak or absent in the LEC near the dorsal and ventral midlines. dpp is expressed in a dorsal sector of the VHN and in a few cells at the dorsal margin of the aDHN. dpp expression is also seen in the lateral LEC between the VHN and aDHN, and in the dorsal LEC between contralateral dorsal nests (Fig. 2A-C).

The pupal expression of wg evolves from the pattern present in the larva, where wg is expressed in a circumferential stripe along the AP compartment boundary (not shown). The early pupal pattern develops by gradual elimination of expression at the ventral, dorsal and lateral midlines (Shirras and Couso, 1996, and our observations). dpp is not expressed in the epidermis of third instar larvae, as judged by in situ hybridization and lack of expression of the dpp¹⁰⁶³⁸ enhancer trap. However, the expression of dpp in the pupa is reminiscent of the embryo, where it is expressed in mid-dorsal and ventro-lateral stripes (St Johnston and Gelbart, 1987). Thus, the pupal expression of dpp may reflect some memory of this embryonic pattern.

The patterns of wg and dpp expression established by 18 hours APF are maintained during the subsequent growth of the histoblast nests. At the time of fusion of the VHN and DHN (24-28 hours APF), wg expression is seen in sectors in the ventral third of the VHN and in the dorsal half of the DHN (Fig. 3A,D,E). dpp is expressed in a stripe in the dorsal two-thirds of the VHN and in a group of 30-40 cells at the dorsal DHN margin (Fig. 3A-C). dpp expression also extends transiently into the ventral DHN margin; this expression lasts for only a few hours, and encompasses about 15 cells at its peak (Fig. 3B).

The complementary expression patterns of wg and dpp are retained in the newly formed adult epidermis at 40-42 hours APF. dpp is expressed in a transverse stripe in the presumptive pleura (Fig. 4A-C) and in a wedge-shaped stripe along the dorsal midline of the tergite (Fig. 4A,D). The limits of pleural expression of dpp coincide precisely with the sternite-pleura and tergite-pleura boundaries. wg is expressed in the sternite and in the medial tergite, but is excluded from the dorsal midline (Fig. 4A,E,F). Neither gene is expressed in a large lateral region of the tergite (Fig. 4A,B,E). The expression of wg and dpp remains restricted to the posterior region of the anterior compartment, with sharply defined posterior and graded anterior boundaries.

Double labelling with Engrailed (En) shows that the posterior limit of dpp expression coincides with the compartment boundary (Fig. 5H). Based on morphological landmarks and on the pattern of lacZ expression in wg-lacZ/dpp-lacZ pupae (not shown), the same appears to be true for wg.

The expression of wg and dpp in the pupal abdomen is regulated along the AP axis by hh, which activates both genes, and en, which represses them. The gain-of-function allele hh^Mir causes ectopic expression of hh in a stripe at the anterior edge of the DHN, and low level ectopic expression in the anterior compartment of the VHN (Kopp et al., 1997; Fig. 5A). In hh^Mir heterozygotes, wg expression in the presumptive tergite is duplicated, and dpp expression in the pleura is expanded to the anterior (Fig. 5B,D). In addition, dpp expression at the dorsal midline becomes wider, while wg expression recedes laterally (Fig. 5C,D). In the gain-of-function mutant en^Apa, en is expressed ectopically in anterior compartment histoblasts in all abdominal segments except A1, where its expression is nearly normal (Kopp et al., 1997; Fig. 5E). In en^Apa heterozygotes, dpp and wg are almost completely repressed in A2 and more posterior segments, but are expressed normally in A1 (Fig. 5F,G).

Shirras and Couso (1996) presented evidence that wg functions in the pupal abdomen to promote tergite and sternite identities. They found that ectopic wg expression in hs-wg animals causes the expansion of tergite and sternite territories at the expense of the pleura, whereas loss of wg function in wg^ts animals causes the reciprocal phenotype. We have obtained similar results. Almost the entire pleura is transformed to tergite and sternite identities following ectopic expression of wg in the pleural territory in dpp-GAL4/UAS-wg pupae (Fig. 6C). Expression of a constitutively active form of the Wg transcriptional effector Armadillo (Arm^S10; Pai et al., 1997) in dpp-GAL4/UAS-arm^S10 pupae produces an identical phenotype (not shown). Interference with Wg signaling has the opposite effect. dTCF, the product of the pangolin gene, is required to activate transcription of wg target genes (Bienz, 1997; Cavallo et al., 1997; Nusse, 1997; Riese et al., 1997; van de Wetering et al., 1997). We used the GAL4 enhancer trap T155 (Harrison et al., 1995), which is expressed ubiquitously in the pupal abdomen (not shown), to drive the expression of a dominant-negative form of dTCF (DN-dTCF; van de Wetering et al., 1997). When T155/UAS-DN-dTCF pupae are raised at 29°C, the entire sternite and most of the tergite are transformed to pleura (Fig. 6D). Strikingly, a small patch of lateral tergite retains tergal identity, while more medial regions are completely transformed.

We find that dpp functions reciprocally and antagonistically to wg in the pupal abdomen. Ectopic expression of dpp expands the pleura at the expense of sternites and tergites. In T155/UAS-dpp pupae raised at 29°C, a large lateral region of the tergite is transformed to pleural identity, while the sternite is reduced to about one-quarter of its normal size (Fig. 6B).

Interestingly, the number of bristles remains unchanged, resulting in a higher than normal bristle density in the tergite and in the development of bristles in the pleural tissue near the reduced sternite (Fig. 6B). A similar, but weaker phenotype is seen following overexpression of dpp in dpp-GAL4/UAS-dpp pupae (not shown).

To examine the effects of loss of dpp function, we generated mitotic clones homozygous for Mad¹², a null allele of Mothers against dpp. The Mad gene product is required for the transcription of dpp target genes (Sekelsky et al., 1995; Heldin et al., 1997). Surprisingly, only six out of several hundred clones located in the pleura were transformed to sternite or tergite identity, as gauged by trichome morphology and the presence of bristles (Fig. 6F). Other clones retained pleural identity, or displayed a weak phenotype such as mild sclerotization (not shown). Furthermore, the Dpp receptor components encoded by punt (put) and thickveins (tkv) (Ruberte et al., 1995) also appear dispensable for pleural specification. We tested the effects of loss of Put function in clones homozygous for the strong hypomorphic allele put¹³⁵ (Ruberte et al., 1995) and in temperature-sensitive put¹³⁵/put¹⁰⁴⁶⁰ heterozygotes (Theisen et al., 1996). Loss of Tkv function was examined in clones homozygous for tkv⁷, a null (or possibly antimorphic) allele (de Celis, 1997; Twombly et al., 1996). None of these mutants had any effect in the pleura (not shown). We conclude that, although Dpp is able to promote pleural identity, its function in pleural specification is not essential.

To study the effects of localized dpp expression, we generated ‘flp-out’ dpp-expressing clones in the abdomen of f^36ahs-FLP122; Tubα1>f⁺>dpp flies (Zecca et al., 1995). When such clones, which are marked by f⁻ bristles, are located in the sternite or lateral tergite, they cause transformation of nearby sternal or tergal tissue to pleural identity (Fig. 7A,B). This effect is opposite to that of wg-expressing clones, which cause transformation of the pleura to tergite or sternite (Shirras and Couso, 1996, and Fig. 7D). To test the autonomy of dpp and wg, we induced flp-out clones in hs-FLP1/Tubα1>CD2>GAL4; UAS-dpp/UAS-lacZ and hs-FLP1/Tubα1>CD2>GAL4; UAS-wg/UAS-lacZ flies. In these genotypes, cells that express dpp or wg are marked by coincident expression of lacZ. We observe that cells well away from the source of Dpp or Wg are transformed, indicating that both genes function at a distance (Fig. 7D-F).

We find that wg and dpp repress one another at the transcriptional level in the pupal abdomen, as they do in the leg and antennal imaginal discs (Brook and Cohen, 1996; Jiang and Struhl, 1996; Johnston and Schubiger, 1996; Penton and Hoffmann, 1996; Theisen et al., 1996). wg-lacZ expression is reduced dramatically in the abdominal histoblasts of T155/UAS-dpp pupae raised at 29°C from the late third larval instar (Fig. 8A). Conversely, the expression of dpp-lacZ is virtually eliminated in the histoblasts of dpp-GAL4/UAS-wg pupae raised under the same regime (Fig. 8B,C). The expression of wg and dpp in the LEC was not visibly affected in these experiments. We see no change in the expression of dpp-lacZ in T155/UAS-dpp pupae, or in the expression of wg-lacZ in dpp-GAL4/UAS-wg pupae (not shown). Thus, we find no evidence that wg and dpp are subject to autoactivation, as they are in the leg disc (Brook and Cohen, 1996).

Wg and Dpp also appear to compete directly with one another in specifying opposite cell fates. As described above, ectopic Wg expression expands tergite and sternite territories at the expense of the pleura (Fig. 6C), whereas ectopic Dpp expands the pleura at the expense of tergites and sternites (Fig. 6B). However, when wg and dpp are coexpressed in T155/UAS-wg UAS-dpp pupae, abdominal pattern is restored almost to wild type (Fig. 6E). Since the UAS-wg and UAS-dpp transgenes are not subject to mutual repression, we conclude that pleural and tergite/sternite cell fates are specified by the balance between competing Wg and Dpp activities.

Direct competition between Wg and Dpp presumably accounts for the incomplete transformations observed in flp-out clones. Although dpp-expressing clones in the lateral tergite are associated with tergite to pleura transformations, parts of these clones can secrete tergite cuticle (Fig. 7A). Moreover, even very large dpp-expressing clones located in the medial tergite have no effect on cell fates (Fig. 7C). In both cases, it is likely that the effects of ectopic Dpp are overwhelmed by endogenous Wg. Conversely, small wg-expressing clones surrounded by pleural tissue often secrete pleural cuticle (Fig. 7G), presumably because ectopic Wg is out-competed by endogenous Dpp.

The behavior of the lateral tergite was initially puzzling to us. This region lies between wg- and dpp-expressing territories (the medial tergite and pleura, respectively), but expresses neither gene (Fig. 4). Because of this, and because the ventral DHN margin is closer to the source of Dpp than it is to the source of Wg, we expected the lateral tergite to be very sensitive to changes in wg expression. However, this region turns out to be the least sensitive to loss of Wg function. Reduction in Wg activity in either wg^ts (Shirras and Couso, 1996) or T155/UAS-DN-dTCF pupae (Fig. 6D) transforms the entire medial tergite to pleura, yet fails to transform part of the lateral tergite. This observation cannot be explained by the effects of Wg and Dpp alone. Indeed, our results indicate that signaling by the Drosophila EGF Receptor (DER) pathway is also involved in specifying tergite and sternite identities.

Ubiquitous expression of a dominant negative form of DER (DN-DER; O’Keefe et al., 1997) in T155; UAS-DN-DER pupae raised at 25-29°C causes expansion of the pleura at the expense of tergites and sternites (Fig. 9A). A similar phenotype is caused by a temperature-sensitive combination of DER alleles, top^CA/top^2W74, at 25°C (Fig. 9B). When DN-DER is coexpressed with DN-dTCF, the entire tergite is transformed to pleura (Fig. 9D). Conversely, ubiquitous expression of a constitutively active form of DER (λtop; Queenan et al., 1997) in dpp-GAL4; UAS-λtop or T155; UAS-λtop pupae transforms the pleura to tergite and sternite identities (Fig. 9C). These observations suggest that DER signaling functions in concert with Wg to promote tergite and sternite fates, and is particularly important in the lateral tergite region.

We find that expression of wg-lacZ and dpp-lacZ is not visibly affected in T155; UAS-DN-DER pupae at 29°C (not shown). This suggests that DER signaling does not regulate wg or dpp expression, but rather interacts synergistically with Wg and/or antagonizes Dpp directly at the level of cell fate determination.

The involvement of DER in tergite specification may explain the surprising resistance of the tergite to the effects of ectopic dpp. While ectopic expression of wg can cause transformation of the entire pleura to tergite and sternite (Shirras and Couso, 1996; Fig. 6C), ectopic dpp is able to transform only the lateral edge of the tergite to pleura (Fig. 6B). No further transformation is achieved by increasing the dosage of UAS-dpp (not shown). However, a strong synergistic effect is observed when dpp and DN-DER are coexpressed in T155/UAS-dpp; UAS-DN-DER pupae. In the resulting adults, the tergite is reduced to a very small patch near the dorsal midline; importantly, the lateral tergite is completely transformed to pleura (Fig. 9E).

Madhavan and Madhavan (1995) report that histoblast proliferation is inhibited in top^CA/top^2W74 heterozygotes raised at the restrictive temperature. However, direct examination of the pupal abdominal epidermis shows that the patterning defects we observe in T155; UAS-DN-DER pupae are not due to reduced histoblast proliferation (not shown). The proliferation defects observed by Madhavan and Madhavan likely result from a stronger reduction in DER function than was achieved in our experiments.

In addition to promoting pleural development, Dpp signaling contributes to DV patterning within the tergite. The pigment band that covers the posterior edge of the tergite widens toward the anterior at the dorsal midline (Fig. 10A). This widening coincides with the mid-dorsal stripe of dpp expression (Fig. 4D). Two lines of evidence indicate that dpp is responsible for pigmentation near the dorsal midline. First, ectopic expression of dpp in T155/UAS-dpp pupae leads to expansion of the pigment band (Fig. 10B). Second, Mad¹² clones at or near the dorsal midline show loss of pigmentation (Fig. 10C,D). Surprisingly, a partial loss of pigmentation is often observed in Mad¹² clones at positions lateral to the anterior inflection of the pigment band, suggesting that Dpp signaling contributes to pigmentation over a wide medial region of the tergite.

Away from the dorsal midline, the width of the pigment band is controlled by the expression of optomotor-blind (omb) (Kopp and Duncan, 1997), which encodes a putative T-box transcription factor (Pflugfelder et al., 1992). omb expression is activated along the posterior edge of the tergite by Hh protein secreted by posterior compartment cells (Kopp and Duncan, 1997). omb may also be responsible for pigmentation along the dorsal midline. We find that the omb-GAL4² enhancer trap (Y. H. Sun, unpublished), which appears to reflect accurately the pattern of omb transcript expression in the pupal abdomen, is expressed more extensively near the dorsal midline than elsewhere in the tergite (Fig. 10E). Also, omb⁻ hemizygotes lack midline as well as posterior pigmentation (Kopp and Duncan, 1997). However, ectopic dpp expression in T155/UAS-dpp pupae does not consistently upregulate omb (not shown).

To understand better the regulation of dpp in the abdomen, we tested genomic fragments from the 3’ region of dpp (Blackman et al., 1991; R. K. B., unpublished) for the ability to drive lacZ expression in the pupal epidermis. We find that dpp expression in the histoblasts and in the LEC is controlled by separate enhancer elements (Table 1; Fig. 11).

Histoblast expression is regulated by two distinct regions (Table 1; Fig. 11B-F). Fragments from between 109.5 kb and 113.5 kb on the standard dpp genomic map (Blackman et al., 1991; Merli et al., 1996) drive lacZ expression in the developing pleura, but not in the sternite or most of the tergite (Fig. 11B,F). Accordingly, we refer to this region as the pleural enhancer. The dpp-GAL4 reporter of Staehling-Hampton et al. (1994) is driven by a fragment that extends from 106 to approximately 110.8 kb, and is expressed in a similar pattern (Fig. 11E). Unlike the endogenous dpp pattern, some of the fragments from the 109.5-113.5 kb region drive persistent, rather than transient, expression in the lateral tergite (Fig. 11B,E). The tergite expression is controlled in part by a distinct element, located between 112.3 kb and 113.5 kb (Fig. 11C).

A second enhancer region active in histoblasts (the ‘circumferential enhancer’) is located between 117.2 kb and 118.9 kb. This fragment drives expression in a stripe that extends around almost the entire segment, interrupted only at the ventral midline and near the spiracle (Table 1; Fig. 11D,F). Presumably the activity of this enhancer is normally repressed in the tergite and sternite territories by other regulatory regions. We have not identified sequences responsible for dpp expression along the dorsal midline.

Both the pleural and circumferential histoblast enhancers are responsive to hh. Expression of the BS 3.21 reporter construct, which is representative of the pleural enhancer (Table 1), is strongly expanded to the anterior in the hh^Mir gain-of-function mutant, whereas expression of the BS 4 construct, which contains the circumferential enhancer, is duplicated (not shown). Both enhancers are repressed by wg, although to differing extents. BS 4 expression in the tergite (but not in the pleura) is completely eliminated in hs-wg pupae grown at high temperature overnight, whereas BS 3.21 expression is only weakly affected (not shown). dpp expression in the LEC is controlled by an entirely separate region (Table 1; Fig. 11A,F). Fragments located between 98.5 kb and 106.9 kb drive expression in a correct dpp pattern in the LEC, but not in the histoblasts (Fig. 11A). Interestingly, this region is devoid of imaginal disc enhancers (Blackman et al., 1991). The fragments BS 1.1 (98.5-100.3 kb), BS 2 (100.2-104.5 kb) and BS 2.1 (104.7-106.9 kb) produce very similar expression patterns, suggesting that dpp expression in the LEC is controlled by several redundant enhancers. Unlike the endogenous dpp gene, the BS 2 and BS 2.1 reporters are also expressed in the third instar larval epidermis (not shown; Blackman et al., 1991).

The major conclusion of this report is that much of the dorso-ventral (DV) patterning of the adult abdomen is determined by antagonistic interactions between Dpp, which specifies pleural cell fate, and Wg and DER signaling, which together specify tergite and sternite fates. Expression of wg and dpp is activated at the posterior edge of the anterior compartment by Hh signaling. Within this zone, wg and dpp are expressed in complementary patterns along the DV axis: wg is expressed in the presumptive sternite and in the medio-lateral region of the tergite (Shirras and Couso, 1996), whereas dpp is expressed in the presumptive pleura and at the dorsal midline of the tergite. Although the pattern of DER activation in the abdomen has not been determined, DER signaling is most important in the lateral tergite, a region where neither wg nor dpp are expressed.

Our results confirm the conclusion of Shirras and Couso (1996) that Wg promotes tergite and sternite fates at the expense of the pleura. In addition, we show that DER signaling acts co-operatively with Wg in specifying tergite and sternite identities. Our observations suggest that the requirements for Wg and DER within the tergite are complementary: Wg function appears to be most important in the medial region, whereas DER signaling is most important laterally. Thus, reduction of Wg signaling transforms the medial tergite to pleura, but leaves part of the lateral tergite untransformed (Shirras and Couso, 1996; this report). Conversely, reduction of DER signaling affects only the lateral tergite margin. Simultaneous inhibition of Wg and DER activities results in the transformation of the entire tergite to pleura.

Consistent with its spatial requirement, wg is expressed in the medial, but not the lateral, regions of the tergite. Although we have not been able to determine the pattern of DER activation, we suspect that it is highest in the lateral tergite. DER is activated in somatic tissues by two known ligands, encoded by vein and spitz (Perrimon and Perkins, 1997; Schweitzer and Shilo, 1997). Surprisingly, ectopic expression of either Vein (in T155; UAS-vn pupae) or an activator of Spitz (Rhomboid; in hs-rho pupae) had no effect on abdominal pattern, although both caused extensive abnormalities in the wing (not shown). Whatever ligand is utilized in the abdomen, an appealing possibility is that its expression is induced jointly by Wg and Dpp, as has been suggested for vein in the midgut (Szüts et al., 1998). Such joint induction would serve to promote high DER activity specifically in the lateral tergite.

The effects of Wg and DER in the abdomen are opposed by Dpp, which is expressed in the presumptive pleura and promotes pleural fate at the expense of sternites and tergites. Interestingly, the transformation of dorsal cuticle to pleura caused by ectopic Dpp is more extensive in the posterior than in the anterior compartment. This higher sensitivity of posterior cells may account for the development of pleural cuticle between the sternites of successive segments in wild type, as much of this pleura is located in the posterior compartment.

We find that disruption of Dpp signaling by Mad, put or tkv mutations has little or no effect on the development of the pleura. Although it is possible that Dpp acts through a novel signal transduction pathway in the abdomen, it seems more likely that dpp is not essential for pleural specification. One possibility is that pleura is the ground state of the abdominal epidermis, and that Dpp functions only to block tergite and sternite specification by Wg and DER. Alternatively, additional factors may act in parallel with Dpp to promote pleural identity, similar to the way Wg and DER cooperate in specifying sternite and tergite fates.

We find that Wg and Dpp antagonize one another at two levels. First, their expression in complementary domains is maintained by mutual transcriptional repression. Second, Wg and Dpp appear to compete directly with one another in influencing the fate of target cells. Since both proteins function nonautonomously, we suggest that cell fates along the DV axis are determined largely by the relative levels of Wg and Dpp at each position. The relationship of DER to Wg and Dpp signaling is unclear. We show that DER is not involved in regulating wg or dpp expression. However, it is not known whether DER has an independent ability to specify tergite and sternite identities, or functions by promoting Wg activity and/or antagonizing Dpp activity.

The pattern established by Dpp and Wg/DER signaling appears to be further refined by differential cell adhesion. Cells transformed to pleural identity by ectopic Dpp are invariably found adjoining the normal pleura. Conversely, cells transformed to sternite or tergite identity by ectopic Wg are usually, but not always, associated with a normal sternite or tergite (Shirras and Couso, 1996; this report). These observations suggest that cells either migrate toward cells with similar identities, or are eliminated when surrounded by cells of different identity. Either mechanism would produce sharp sternite/pleura and tergite/pleura boundaries, even if the initial readout of Dpp and Wg/DER signaling were somewhat imprecise.

The regulatory relationships we describe for hh, wg and dpp in the adult abdomen are similar to those that operate in the leg, antennal and genital imaginal discs (Held, 1995; Brook and Cohen, 1996; Jiang and Struhl, 1996; Johnston and Schubiger, 1996; Penton and Hoffmann, 1996; Morimura et al., 1996; Theisen et al., 1996; Freeland and Kuhn, 1996; Chen and Baker, 1997; Hepker et al., 1997; Sanchez et al., 1997). As in these discs, wg and dpp are activated by hh and repressed by en, while their expression along the DV axis is controlled by mutual repression. However, in contrast to the imaginal discs, wg and dpp play no role in AP patterning in the abdomen (Shirras and Couso, 1996; Struhl et al., 1997a,b; Kopp et al., 1997). One might speculate that the mechanisms that pattern the body wall are less derived than those that pattern appendages. Thus, the primitive functions of wg and dpp may have been to control DV patterning, while AP pattern was specified directly by hh. With the origin of appendages, wg and dpp could have been co-opted to control both AP patterning and proximo-distal outgrowth.

Functional interactions among the Wg, Dpp and DER signaling pathways differ in different developmental contexts. Although Wg and Dpp have opposing functions in DV patterning, they co-operate in proximo-distal patterning of legs and antennae (Campbell et al., 1993; Lecuit and Cohen, 1997) and in endoderm induction in the midgut (Bienz, 1997; Riese et al., 1997). Wg and DER co-operate in tergite and sternite specification, yet interact antagonistically in the AP patterning of the embryonic epidermis (O’Keefe et al., 1997; Szüts et al., 1997). Finally, Dpp and DER act co-operatively in wing vein development (de Celis, 1997) and in inductive interactions in the midgut (Szüts et al., 1998), but play antagonistic roles in abdominal patterning and in cell fate specification in tracheal placodes (Wappner et al., 1997; Chen et al., 1998).

The dark pigment band in the posterior tergite is specified by Hh signaling (Struhl et al., 1997a,b; Kopp et al., 1997), which is mediated primarily by omb (Kopp and Duncan, 1997). The pigment band widens in the medial tergite, and usually continues into a narrow stripe along the dorsal midline. We show that this medial pigmentation is specified by Dpp signaling, possibly by activation of the same target, omb. Surprisingly, a partial requirement for Dpp signaling extends further laterally than the midline pigment stripe, suggesting that Hh and Dpp may interact synergistically to promote pigmentation over much of the medial tergite.

Comparison of abdominal pigmentation in different Drosophila species reveals a number of recurrent pattern motifs. These include a mid-dorsal gap in the pigment band, which distinguishes the subgenus Drosophila from the subgenus Sophophora, and a triangular patch of pigment at the lateral tergite margin, which is seen mainly in the repleta species group (Patterson and Stone, 1952). Both of these motifs are present in D. mulleri (Fig. 10F). Our observations suggest that dpp and its target genes may have played an important role in the evolution of these patterns. The lateral pigment patch strikingly recapitulates the expression of one of the dpp genomic enhancers in D. melanogaster (Fig. 11C), while the midline gap in the pigment band coincides with the mid-dorsal expression of dpp. Thus, these pattern elements could be accounted for if pigmentation were promoted by moderate levels of Dpp signaling, but repressed by higher levels.

In the ventral abdomen, Dpp specifies pleural identity, whereas dorsally it normally promotes tergite pigmentation. What determines which effect Dpp will have? It is unlikely that timing or level of expression is important, since dpp is expressed at similar levels in the presumptive pleura and at the dorsal midline, and expression is sustained in both locations throughout pupal development. One possibility is that Wg and DER signaling are too strong dorsally to be completely overcome by Dpp. Consequently, tergite develops here rather than pleura, and Dpp is able to influence only pigmentation. Alternatively, dorsal and ventral histoblasts may be pre-programmed to respond differently to Dpp.

Wg and DER signaling also have different effects in ventral and dorsal histoblasts: ventrally they specify sternite, whereas dorsally they specify tergite. Again, it is not clear what determines these different outcomes. Factors other than Wg, Dpp or DER are almost certainly involved. Transformations of sternite to tergite occur in clones lacking the Hox protein cofactor Extradenticle (Gonzalez-Crespo and Morata, 1995) and in certain Polycomb-group mutants (our unpublished observations), suggesting that differential expression of some currently unknown homeodomain protein may determine the response of histoblasts to Wg and DER signaling.

We are grateful to Konrad Basler, Mariann Bienz, Carrie Brachmann, Peter Lawrence, Anthea Letsou, Chris Merli, Elizabeth Noll, Richard Padgett, Mark Peifer, Gert Pflugfelder, Trudi Schupbach, Gary Struhl, Y. Henry Sun, Lawrence Zipursky and Kathy Matthews and the Bloomington stock center for various fly stocks. We also thank the anonymous reviewers for comments and members of the Duncan laboratory for support and discussions. This work was supported by NSF grants DCB 90-18618 and MCB 93-17701 to R.K.B. and NIH grant GM32318 to I.D.