Modulation of Decapentaplegic Gradient During Haltere Specification in Drosophila ⁎ Kalpana Makhijani , Chilukuri Kalyani, Tamarisa Srividya, L.S

Developmental Biology 302 (2007) 243–255 www.elsevier.com/locate/ydbio

Modulation of Decapentaplegic gradient during haltere specification in Drosophila ⁎ Kalpana Makhijani , Chilukuri Kalyani, Tamarisa Srividya, L.S. Shashidhara

Center for Cellular and Molecular Biology, Uppal Road, Hyderabad 500 007, India Received for publication 7 September 2006; revised 12 September 2006; accepted 14 September 2006 Available online 20 September 2006

Abstract

Suppression of wing fate and specification of haltere fate in Drosophila by Ultrabithorax is a classical example of Hox regulation of serial homology. However, the mechanism of Ultrabithorax function in specifying haltere size and shape is not well understood. Here we show that Decapentaplegic signaling, which controls wing growth and shape, is a target of Ultrabithorax function during haltere specification. The Decapentaplegic signaling is down-regulated in haltere discs due to a combination of reduced levels of the Dpp, its trapping at the A/P boundary by increased levels of its receptor Thick-vein and its inability to diffuse in the absence of Dally. Results presented here suggest a complex mechanism adopted by Ultrabithorax to modulate Decapentaplegic signaling. We discuss how this complexity may regulate the final form of the adult haltere in the fly, without compromising features such as cell survival, which is also dependent on Decapentaplegic signaling. © 2006 Elsevier Inc. All rights reserved.

Keywords: Wing; Organ size; Thick-vein; Dally; Phosphorylated SMAD

Introduction Wing size is primarily regulated by the Notch (N) and Wingless (Wg) pathways along the D/V axis and the In Drosophila, wings and halteres are the dorsal appendages Decapentaplegic (Dpp) pathway along the A/P axis. Activa- of the second and third thoracic segments, respectively. In the tion of N in the D/V boundary occurs via interactions between third thoracic segment, homeotic selector gene Ultrabithorax dorsal and ventral cells of the wing pouch to set up the (Ubx) suppresses wing development to mediate haltere devel- organizer (Diaz-Benjumea and Cohen, 1993, 1995; Williams opment (Lewis, 1978). The genetics of Ubx has served as a et al., 1994; Irvine and Weischaus, 1994; Kim et al., 1995). paradigm for understanding Hox function in generating Previous reports have shown that several downstream targets morphological diversity within as well as between organisms. of the N pathway, such as Wingless (Wg), Cut (Ct) and Wing and haltere imaginal primordia are specified in the late Vestigial (Vg) are down-regulated in haltere discs (Weatherbee embryo. At this stage, they show a difference of 2- to 3-fold in et al., 1998; Shashidhara et al., 1999; Mohit et al., 2003). cell number (Cohen, 1993; Held, 2002). This difference is Activation of Dpp in the wing disc is dependent on events further increased to 5-fold during the post-embryonic develop- along the A/P axis controlled by the expression of engrailed mental stages and is subsequently reflected in the sizes of adult (en)(Posakony et al., 1990; Nellen et al., 1996; reviewed in wings and halteres. Interestingly, removal of Ubx even in the Ingham, 1998). en, which is exclusively expressed in the post-embryonic stages not only induces haltere-to-wing trans- posterior compartment, activates hedgehog (Hh)inthat formations at the cellular level, transformed tissue also shows a compartment. Hh is a short-range morphogen that diffuses dramatic increase in size (Morata and Garcia-Bellido, 1976), to activate Smoothened (Smo) in the anterior compartment by suggesting that regulation of haltere size is a continuous process. binding to Patched (Ptc), which normally inhibits the latent signaling activity of Smo. In the presumptive A/P boundary ⁎ Corresponding author. cells, Smo activates Cubitus-interruptus (Ci) by stabilizing its E-mail address: [email protected] (K. Makhijani). full-length isoform, which in turn activates dpp expression.

Dpp functions as a long-range signal and activates genes in a Calleja, personal communication to FlyBase) and ptc-GAL4 (Brand and concentration dependent manner. Recently, we have reported Perrimon, 1993). The UAS lines used are UAS-CD2 (N. Brown, personal Communication to FlyBase), UAS-Dally (provided by H. Nakato), UAS-Dpp that Dpp signaling is down-regulated in the haltere discs, (Frasch, 1995), UAS-Dpp::GFP (Teleman and Cohen, 2000), UAS-FLP (Xu and particularly at the level of the transcription of Dpp itself and Harrison, 1994), UAS-GFP::DPP (Entchev et al., 2000), UAS-nuclear lacZ over-expression of Dpp in the entire haltere pouch induces (Brand and Perrimon, 1993), UAS-Tkv (Lecuit and Cohen, 1998), UAS-Ubx moderate haltere-to-wing transformations at the cellular level (Castelli-Gair et al., 1994) and UAS-dsRNA>Ubx, Df(3R)Ubx-109 (UAS- RNAi (Mohit et al., 2006). Reduced Dpp signaling in haltere may Ubx ; Monier et al., 2005). To prevent early lethality of ptc-GAL4/+; UAS- GFP::Dpp/+ larvae, tub-GAL80ts (McGuire et al., 2003) was used to temporally have limited the size of the haltere by directly affecting its control GAL4 activity. The lacZ reporter constructs used are dpp BS3.0-lacZ growth. However, loss-of-function clones of Ubx generated (Blackman et al., 1991), dally P2-lacZ (Fujise et al., 2001), mtvk00702-lacZ in haltere discs at different time points do not show any (Funakoshi et al., 2001), sal-lacZ (Galant et al., 2002), tkv P906-lacZ (Tanimoto growth advantage. The ratio of Ubx− clones to their wild et al., 2000) and tkv k16713-lacZ (George and Terracol, 1997). type twin spots is comparable to that in wing discs (wherein Ubx is not functional; Supplementary Fig. 1). This suggests Histology that either loss of Dpp is not the reason for reduced haltere size or the effect of down-regulation of Dpp is non-cell- RNA in situ hybridization for dpp and tkv was done as described by Tautz autonomous. and Pfeifle (1989) using its cDNA clones RE20611 (for dpp) and GH25238 (for tkv). The Tyramide Signal Amplification kit (NEN Life Science Products) was Recently, it has been shown that the juxtaposition of cells used for fluorescent RNA in situ hybridization. Immunohistochemical staining receiving different levels of Dpp signaling is essential for cell was performed essentially as described by Patel et al. (1989). The staining proliferation in the wing disc (Rogulja and Irvine, 2005). This protocol for extracellular GFP::DPP was as described in Belenkaya et al. (2004) observation is relevant to understand the mechanism by which with minor modifications: the dissected discs were pre-blocked in PBS–BSA haltere size is reduced and also in the context of differences (0.5% BSA) on ice for 30 min and then were incubated with Rabbit polyclonal anti-GFP (Molecular Probes, 1:100) for 1 h on ice. between wing and haltere discs with respect to the relative In experiments involving Leptomycin B (LMB; Sigma, St. Louis, USA) sizes of anterior and posterior compartments (3:1 in haltere treatment, third instar larvae were dissected and submerged in LMB (50 ng/ml) discs as against 1:1 in wing discs). In this context, here we containing cl-8 medium (Wang and Holmgren, 2000) for 3 h followed by examine the status of the Dpp gradient in haltere discs. We immunostaining. The primary antibodies used are monoclonal anti-β-galacto- report that Ubx indeed inhibits spread of Dpp and thereby sidase (1:100; Developmental Studies Hybridoma bank, DSHB), anti-CD2 (1:200; Caltag Laboratories, CA), anti-Ci (1:5; Motzny and Holmgren, 1995), affects the properties of the gradient. The profile of Dpp anti-engrailed 4D9 (1;200; Patel et al., 1989), anti-Fas III (1:50; DSHB), anti- signaling differs between wing and haltere discs with haltere patched (1:50; Capdevila et al., 1994) and anti-Ubx (1:30; White and Wilcox, discs showing relatively higher levels of Dpp signaling in 1984); polyclonal anti-β-galactosidase (1:500; raised by A. Khar, CCMB, anterior cells immediately adjacent to the A/P boundary India), anti-GFP (1:500, Molecular Probes), anti-Hh (1:1000; Tabata and (hereafter referred to as AB cells) and lower levels of signaling Kornberg, 1994) and anti-phospho Smad 1 (pMAD; 1:500; Persson et al., 1998). Secondary antibodies conjugated to Flourophore (Alexa dyes) were obtained in cells away from the A/P boundary. This difference is from Molecular Probes, USA and used in 1:1000 dilutions. To detect Ci, prominent at the level of pMAD staining in the posterior biotinylated anti-rat IgG (1:500, Jackson ImmunoResearch Laboratories) and compartment. This is due to trapping of Dpp in the AB cells Cy3 conjugated avidin (1:1000, Jackson ImmunoResearch Laboratories) were caused by relatively higher levels of its receptor Tkv in those used. Fluorescence images were obtained on either Zeiss Axioplan2 florescence cells and down-regulation of Dally in both transducing and microscope or Zeiss LSM 510 META confocal microscope, and images were processed using either Axiovision or Adobe Photoshop. The fluorescence receiving cells. Our results suggest a novel mechanism by intensity of dpp RNA in situ hybridization was measured with NIH Image J which the Hox gene Ubx generates differences in sizes of software plot profile tool. serially homologous organs. Comparisons between wing and haltere discs for trafficking of GFP-tagged Dpp were made on projected images of 11 sections each (0.5 μ thick sections). In all experiments, for comparison of fluorescent intensities, wing and haltere (or Materials and methods control and experimental) discs were digitized always at identical microscope and camera settings. Fly strains Chromatin immunoprecipitation (ChIP) The Canton-S strain of Drosophila melanogaster was used as the wild type strain. Recombinant chromosomes and combinations of GAL4 drivers, UAS The ChIP protocol used is modified from Weinmann and Farnham (2002) lines, different mutations and/or markers were by standard genetic techniques. and Birch-Machin et al. (2005). We used CbxHm/+ wing discs for ChIP assays as Unless stated otherwise, all genetic crosses were set up at 25°C. For it provides more chromatin than wild type haltere discs. Late third instar larvae experiments with tub-GAL80ts, embryos were allowed to grow at 20°C for of genotype were dissected in chilled PBS, fixed in 1% formaldehyde for 20 min 24 h (6 days for experiments related top trafficking of GFP-tagged Dpp) and and transformed wing discs were dissected. Chromatin of size range 500 bp to then shifted to 28°C until the wandering third instar larval stage. The mutant 1.2 kb was prepared from about 75 Cbx wing discs. One third of this chromatin fly strains used are abx, pbx1, Ubx1, CbxHm and tkv6 (all described in was immunoprecipitated with Ubx antibody (1:30), and an equal amount of FlyBase). chromatin without antibody was used as control or mock immunoprecipitation. FLP-FRT method (Xu and Rubin, 1993) was used for generating mitotic The immunoprecipitated DNA was subjected to PCR amplification with primers clones of Ubx and tkv. P[FRT]82B Ubx1 is reported in Shashidhara et al. (1999) designed to amplify those regions of dally and tkv genes, which harbor putative and P[FRT]40 tkv6 in Singer et al. (1997). GAL4-UAS system (Brand and Ubx-binding motifs. Primers against sal328 element of salm (Galant et al., Perrimon, 1993) was used for targeted mis-expression of gene products. The 2002) and 3rd exon of Rpl32 served as positive and negative control GAL4 drivers used are dppblk-GAL4 (Morimura et al., 1996), dppBS3.0-GAL4 respectively. Sequences of primers used are given in the Supplementary data (Hazelett et al., 1998), en-GAL4 (Brand and Perrimon, 1993), omb-GAL4 (M. section. K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 245

Results (Masucci et al., 1990; St. Johnston et al., 1990; Blackman et al., 1991). Of particular interest here are blk1 (Morimura et al., Differential levels of Decapentaplegic expression in wing 1996) and BS3.0 elements (Blackman et al., 1991). blk1 and haltere discs element comprises of 4 kb of the disc region, while BS3.0 comprises of 10 kb of this regulatory region, with a 1 kb overlap Previously we have reported that Dpp signaling is down- to the blk1 element (FlyBase). We examined the expression regulated in the haltere discs, particularly at the level of the patterns of blk1 element using a dppblk1-GAL4 (Morimura et transcription of dpp itself (Mohit et al., 2006; Fig. 1A). With the al., 1996) driver and of BS3.0 element using a dppBS3.0-lacZ help of fluorescent RNA in situ hybridization and its reporter transgene (Blackman et al., 1991)aswellasadppBS3.0- quantification using Image J of NIH, we have observed that a GAL4 (Hazelett et al., 1998) driver. Temporal and spatial given AB cell in the haltere disc expresses ∼1.5-fold lower expression patterns of these regulatory elements in different levels of dpp than a corresponding cell in the wing disc. genetic backgrounds suggested both cell-autonomous and cell However, we did not observe any difference between wing and non-autonomous down-regulation of dpp in haltere discs haltere imaginal discs in the expression patterns of En, Hh, Ptc (Supplementary Figs. 2–4). We also carried out lineage analysis and Ci, the critical upstream components required to activate studies of dppblk1-anddppBS3.0-GAL4 drivers. We have Dpp expression in AB cells (Figs. 1B–E). As the repression of observed that dppblk1-GAL4 is not activated (except in AB dpp in the haltere disc is partial, it is possible that modulation in cells of the hinge) at any time during haltere development, its upstream components could be too subtle for simple whereas dppBS3.0-GAL4 is down-regulated in the haltere in immunostaining reactions to reveal. It has been observed that, third instar larval stages (Supplementary Figs. 4A, B). Thus, our due to the robust nuclear export mechanism, much of the full- observations suggest that Ubx may allow normal levels of Dpp length isoform of Ci in AB cells too is in the cytoplasm (Wang in the beginning of disc patterning along the A/P axis, while it and Holmgren, 2000). Thus, inhibition of the nuclear export of may down-regulate Dpp levels at the beginning of the third Ci would allow better visualization of its nuclear localization, instar stages. reflecting the activity of Hh signaling (Wang and Holmgren, 2000). We cultured wing and haltere discs in the presence of a Unequal Decapentaplegic signaling between the anterior and nuclear-export blocker (Leptomycin B) for 3 h and then posterior compartments of the haltere disc examined the relative levels of nuclear Ci. We did not observe any difference between the two discs (Fig. 1F), suggesting that Higher levels of Dpp are required to activate Spalt-major Ubx may act directly on Dpp to down-regulate its levels in AB (Sal), while Optomotor-blind (Omb) is activated by relatively cells. lower levels of Dpp (Lecuit et al., 1996; Nellen et al., 1996). Post-embryonic expression of dpp is regulated by a 3′ cis- Sal is completely repressed in haltere discs and is one of the regulatory element (30 kb long), referred to as the disc region direct targets of Ubx (Weatherbee et al., 1998; Galant et al.,

Fig. 1. Ubx down-regulates A/P signaling by down-regulating Dpp. In all panels, wing discs are shown on the left and the haltere discs to the right. (A) dpp RNA in situ pattern. Note levels of dpp transcripts are lower in the haltere disc. (B–E) Wing and haltere discs stained for En (B), Hh (C), Ptc (D) and Ci (E). No difference in the pattern or intensity of staining was observed between wing and haltere discs. (F) Wing and haltere discs cultured for 3 h in the presence of Leptomycin B and then stained for Ci. Levels of nuclear Ci also appear to be similar between wing and haltere discs. 246 K. Makhijani et al. / Developmental Biology 302 (2007) 243–255

2002; Fig. 2A). Consistently, it is not activated in response to boundary (Fig. 2D). As levels of dpp expression in pbx/Ubx1 the over-expression of Dpp in the haltere disc (Fig. 2B). haltere discs are comparable to the levels in abx/Ubx1 Removal of Ubx from the anterior compartment causes the background (Supplementary Figs. 2–3), the absence of activation of Sal expression, which is comparable to the corresponding activation of sal expression suggests inability anterior compartment of wing disc in its intensity and area of of Dpp to efficiently signal to the posterior compartment. the expressing domain (Fig. 2C). However, removal of Ubx We examined this phenomenon in more detail using from the posterior compartment causes partial de-repression of antibodies against the phosphorylated form of MAD (pMAD; Sal in a very small region immediately adjacent to the A/P Persson et al., 1998). Presence of pMAD reflects normal Dpp

Fig. 2. Non-cell-autonomous down-regulation of Dpp signaling in haltere discs. Wing and/or haltere discs stained as indicated on the figure itself. (A) Expression pattern of sal-lacZ in wild type wing (A) and haltere (A′) discs. Sal is not expressed in the haltere pouch. (B) sal expression pattern in omb-GAL4/+; sal-lacZ/+; UAS- Dpp wing and haltere discs. Note ectopic Sal expression in the wing disc, but not in the haltere disc (arrow). (C–D) Expression pattern of sal in sal-lacZ/+; abx/Ubx1 (C) and sal-lacZ/+; pbx/Ubx1 (D) haltere discs. Note de-repression of Sal in both the discs, although the de-repression is only marginal in pbx/Ubx1 background. (E) Wild type wing disc stained for pMAD. Note strong pMAD levels in both anterior and posterior compartments in cells close the A/P boundary. A single row of posterior cells, which show peak staining (arrow), acts as a marker for signaling from AB cells to the posterior compartment. (F) dppBS3.0-GAL4/UAS-lacZ wing disc showing partial overlap between Dpp and pMAD staining in AB cells. (G) Wild type haltere disc. pMAD staining in the anterior compartment is comparable to that of wing disc. However, the posterior compartment does not show any activation of pMAD. (H) dppBS3.0-GAL4/UAS-lacZ haltere disc. PMAD, which is activated only in AB cells, completely overlaps with that of Dpp. In images F, G and H, large peripodial cells are also in focus, which express both Dpp and pMAD, which is not related to the phenomenon discussed in this report. (I–L) abx/Ubx1 (I, J) and pbx/Ubx1 (K, L) haltere discs showing increased levels of pMAD staining. Activation of pMAD in the posterior compartment is observed only when the anterior compartment is transformed (in abx/Ubx1 background), although in this background removal of Ubx is incomplete. K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 247 signaling, at least, up to its internalization and phosphorylation activate pMAD levels in the posterior compartment of haltere of MAD. In wing imaginal discs, pMAD induction is seen as a discs (Fig. 3E). gradient running on both sides of AB cells (Funakoshi et al., 2001; Fig. 2E). Thus, its levels are relatively lower in AB cells. Limited movement of Dpp in haltere discs A striking feature of pMAD staining in wing discs is the appearance of a sharp/steeper peak of pMAD induction in the We then directly examined (with the help of GFP::DPP; posterior compartment (Fig. 2E), the functional significance of Entchev et al., 2000) if Dpp trafficking to cells outside AB cells which is not known. The domain of pMAD activation partially is affected in haltere discs. The quantitative estimation of Dpp overlaps with that of Dpp expression (Fig. 2F), reflecting long- diffusion is conditional on expressing equivalent levels of GFP:: range signaling capabilities of Dpp. PMAD activation in haltere DPP in AB cells of wing and haltere discs. dpp-GAL4 drivers discs differs from wing discs in two ways. (i) pMAD activation are not useful for this purpose as they are differentially is mostly restricted to AB cells and the sharp peak of pMAD expressed between wing and haltere imaginal discs. Our expression seen in the posterior compartment of wing discs is attempts to generate sufficiently large Actin>GFP::DPP clones conspicuously absent (Fig. 2G) and (ii) it completely overlaps in AB cells using FLP-out method were not conclusive as all with Dpp expression (Fig. 2H). discs invariably had more than one clone (data not shown). Interestingly, removal of Ubx from the anterior compartment Unless a disc had a single clone in the normal Dpp expression (in abx/Ubx1 background; Figs. 2I, J) and not from the posterior domain, we would not be able to interpret on the diffusion of compartment (in pbx/Ubx1 background; Figs. 2K, L) was GFP::DPP from AB cells. Levels of expression of ptc-GAL4 required to restore the peak pMAD staining in the posterior are equivalent between wing and haltere discs (Figs. 3F, G). As compartment. Moreover, down-regulation of Ubx levels only in ptc-GAL4/UAS-GFP::DPP combination is early larval lethal, AB cells (using ptc-GAL4/UAS-UbxRNAi; Fig. 3B) was we standardized ptc-GAL4+tubulin-GAL80ts combination to sufficient to induce peak pMAD staining in the posterior express GFP::DPP and monitor its diffusion. We examined the compartment (Fig. 3C). Conversely, over-expression of Ubx in levels of extracellular GFP, which was detected by staining the AB cells of the wing disc caused lowering of pMAD levels in discs with anti-GFP antibodies without using any detergents. the posterior compartment (Fig. 3D). As the level of dpp We observed extracellular GFP in cells away from AB cells in expression is higher in pbx/Ubx1 haltere discs than in wild type wing discs (in both anterior and posterior compartments; Fig. discs (Supplementary Figs. 2–3), these observations suggest 3F), whereas it was mostly restricted to AB cells (wherein GFP:: inability of Dpp to efficiently signal to the posterior compart- DPP is expressed) in haltere discs (Fig. 3G). These observations ment. This is further supported by the observation that over- provide strong evidence that Dpp trafficking is compromised in expression of Dpp in the A/P boundary was not sufficient to haltere discs. Limited Dpp diffusion in haltere discs would also

Fig. 3. Removal of Ubx from AB cells alone is sufficient to activate pMAD in the posterior compartment of the haltere disc. (A–C) Wild type (A) and ptc-GAL4/+; UAS-UbxRNAi Df(109)/+ (B, C) haltere discs. Note knockdown of Ubx in AB cells (B) and activation of pMAD in the posterior compartment (C). Also note significant increase in the size of the haltere, although Ubx is removed only from AB cells. (D) ptc-GAL4/UAS-Ubx wing disc. Over-expression of Ubx in AB cells non-cell- autonomously down-regulates pMAD levels in both anterior and posterior compartments. Note considerable reduction in the size of the wing disc, although Ubx is mis-expressed only in AB cells. (E) tub-GAL80ts/ptc-GAL4; UAS-GFP::Dpp/+ haltere disc. Over-expression of Dpp in AB cells is not sufficient to activate pMAD in the posterior compartment. (F–G) Wing (F) and haltere (G) discs of tub-GAL80ts/ptc-GAL4; UAS-GFP::Dpp/+ larvae stained with anti-GFP antibodies in the absence of any detergent to detect only extracellular GFP::DPP (red). Discs are also imaged for GFP, which reflects both intracellular and extracellular GFP::DPP (green). The images shown here are representative of 5 haltere and 4 wing imaginal discs. The intensities of green fluorescence are comparable between the two discs, an indication of similar levels of GFP::DPP transgene expression from the ptc-GAL4 driver. However, extracellular GFP extends much beyond green staining in the wing disc suggesting long-range diffusion of DPP to both anterior and posterior compartments. In haltere discs, the red staining is mostly detected in the region of GFP::DPP expression indicating limited diffusion. 248 K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 explain why loss-of-function clones of Ubx do not show any growth advantage.

Dally is cell-autonomously down-regulated by Ultrabithorax in haltere discs

Dpp trafficking requires the function of members of the glypican family of heparin sulfate proteoglycans such as Dally and Dally-like (Fujise et al., 2003; Belenkaya et al., 2004). Although Dally and Dally-like have somewhat redundant roles, pMAD levels are greatly affected in dally mutant wing discs than in dally-like mutants (Belenkaya et al., 2004). In wing discs, Dally expression is highest in AB cells and shows gradual decrease in its expression towards the distal cells in both the anterior and the posterior compartments (Fujise et al., 2003; Fig. 4A). In haltere discs, Dally expression is completely repressed in the posterior compartment. In the anterior compartment too, its expression is conspicuously absent from AB cells (Fig. 4B). However, its expression in D/V boundary cells of the anterior compartment is comparable to that of wing discs (Fig. 4B). dally expression was activated in the anterior compartment of abx/Ubx1 haltere discs and in the posterior compartment of pbx/Ubx1 haltere discs (Figs. 4C, D). Conversely, its expression in Cbx wing discs is comparable to that of wild type haltere discs (data not shown). Loss-of-function mitotic clones of Ubx resulted in the up-regulation of dally expression in a cell- autonomous manner in haltere discs (Fig. 4E). Although activation of dally in AB cells of abx/Ubx1 haltere discs explains activation of pMAD in the posterior compartment of those discs (Figs. 2I, J), over-expression of Dally in AB cells using ptc-GAL4 driver did not have the same effect on pMAD levels (data not shown). This, however, could be due to the inherent problem in the experiment itself as over-expression of Dally in wing discs causes a dominant negative effect, probably by sequestering Dpp (Fujise et al., 2003). Nevertheless, over- expression of Dally using en-GAL4 driver resulted in enhanced pMAD levels in the posterior compartment of the haltere disc (Fig. 4F), suggesting that Dally could be a critical limiting factor for proper trafficking of Dpp to the posterior compartment of haltere discs. It is, however, possible that increased pMAD activation in this experiment is due to enhanced sensitivity of haltere cells caused by the trapping of Dpp on the cell surfaces. In ptc-GAL4/UAS-UbxRNAi haltere discs (Fig. 3C), while pMAD is activated in the posterior compartment, the response to loss of Ubx from AB cells is more pronounced in Fig. 4. Cell-autonomous down-regulation of dally in haltere discs. Discs in the anterior compartment than in the posterior compartment. panels A–E are stained for dally-lacZ (green), A–D and F for En (red), E for This could be due to continued repression of dally in the Ubx (red) and F for pMAD (green). (A–B) Wild type wing (A) and (B) haltere posterior compartment. Thus, it is more likely that trafficking of discs. In the wing pouch, dally expression pattern has two components. It is Dpp is a limiting factor in haltere discs rather than the reduced expressed on either side of the D/V boundary and also in AB cells. Both these components of dally expression are differentially manifested in haltere discs. In sensitivity of haltere cells to Dpp. the haltere pouch, its expression along the D/V axis is selectively repressed only in the posterior compartment, while its expression along the entire A/P axis is Increased levels of Thick-vein in the A/P boundary of haltere missing. (C–D) abx/dally-lacZ, Ubx1 (C) and pbx/dally-lacZ, Ubx1 (D) haltere discs discs showing de-repression of dally in AB cells (in panel C) and in the posterior compartment (in panel D). (E) Haltere disc with FLP-induced mitotic clones of Ubx1. Note cell-autonomous up-regulation of dally in Ubx− cells. (F) Over- In the wing disc, Hh signaling, which activates Dpp in AB expression of Dally in the posterior compartment activates pMAD in the cells, also down-regulates its receptor Tkv in those cells posterior compartment, suggesting that absence of Dally is one of the limiting (Funakoshi et al., 2001). This may allow enhanced diffusion of factors for Dpp to signal to the posterior compartment. K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 249

Dpp from the source to the neighboring cells, which express of the Sal domain (Lecuit and Cohen, 1998) and down-regulates relatively higher levels of tkv than cells of the A/P boundary levels of pMAD in the posterior compartment (Fig. 5B). (Brummel et al., 1994; de Celis, 1997). Consistently, over- We, therefore, monitored tkv expression by RNA in situ and expression of Tkv in AB cells of wing discs causes a reduction by using two different tkv-lacZ enhancer trap insertions. In wing

Fig. 5. Increased levels of tkv expression in AB cells of the haltere disc. (A–B) Wild type (A) and dppblk1-GAL4/UAS-Tkv (B) wing discs. Note pMAD levels are decreased in non-AB cells, suggesting that increased Tkv in the A/P boundary reduces the diffusion of Dpp to the neighboring cells. (C–E) Expression pattern of tkv in wing and haltere discs visualized by using enhancer trap lines tkvk16713-lacZ (C) and tkvP906-lacZ (D) and by tkv RNA in situ hybridization (E). In wing discs, tkv expression is minimum in AB cells and increased levels of tkv expression are observed in cells away from the A/P boundary. In haltere discs, while overall levels of tkv are lower compared to the wing discs, AB cells show as much tkv expression as the rest of the pouch cells. (F) mtv-lacZ pattern in wild type wing and haltere discs. Note higher levels of mtv expression in AB cells of the wing disc, while its levels are higher in non-AB cells of the haltere disc. (G–H) tkv RNA in situ hybridization on ptc-GAL4/+; UAS-UbxRNAi Df(109)/+ haltere disc (G) and CbxHm/+ wing disc (H). Removal of Ubx from AB cells of the haltere disc causes down-regulation of tkv expression, while ectopic Ubx in the wing disc causes up-regulation of tkv specifically in AB cells and down-regulation in other cells. (I–J) tkvP906-lacZ (I) and tkvP906-lacZ/ptc-GAL4; UAS-UbxRNAi Df(109)/+ (J) haltere discs. Knockdown of Ubx expression in AB cells causes cell-autonomous down-regulation of tkv.(K–L) tkv6 P[FRT]40A/arm-lacZ P[FRT]40A, ptc-GAL4; UAS-FLP/+ wing (K) and haltere (L) discs. Mitotic clones of hypomorphic tkv allele were generated specifically in AB cells using ptc-GAL4 and UAS-FLP. Note increase in pMAD levels in non-AB cells adjacent to the clones in both anterior and posterior compartments of the wing disc. However, there was only a marginal increase in pMAD levels in the haltere disc and it was only in AB cells. 250 K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 discs, tkv expression complements that of dpp expression. Its as in the posterior compartment of haltere discs. Indeed, specific levels are lowest in AB cells, which gradually increase in cells knockdown of Ubx from AB cells results in activation of away from the A/P boundary (Figs. 5C–E). Both tkvk16713-lacZ pMAD in posterior cells (Fig. 3C), which could be due to a and tkvP906-lacZ enhancer trap insertions recapitulate RNA in combination of reduced tkv and increased dally levels in those situ pattern in wing discs, although tkvP906-lacZ insertion cells. expresses at lower levels (Figs. 5C, D). In the haltere disc, although overall levels of tkv expression are lower compared to dally and thick-vein are potential direct targets of the wing disc (Mohit et al., 2006; Figs. 5C, E), the expression of Ultrabithorax tkv is uniform in the central region of the haltere (Figs. 5D, E). Among the two enhancer trap insertions, only tkvP906-lacZ Expression data and genetic analysis suggest that Ubx cell- recapitulates this pattern of tkv expression (Fig. 5D). Relatively autonomously regulates the expression patterns of dally and higher levels of pMAD in all dpp-expressing cells of the haltere tkv (Figs. 4 and 5). We examined genomic sequences of dally disc (Fig. 2H) could be due to higher levels of Tkv activity in and tkv for (a) consensus Ubx-binding sequences T-T-A-A-T-T/ AB cells of haltere discs. G-A/G and (b) their reverse complements (a total of eight 7-mer In wing discs, Hh activates mother of thick-vein (mtv)inAB motifs; Ekker et al., 1991). We considered only those clusters of cells (Fig. 5F), which in turn functions as a negative regulator of Ubx-binding sites that were conserved between D. melanoga- tkv (Funakoshi et al., 2001). In haltere discs too, tkv and mtv ster and D. pseudoobscura. We identified 5 clusters of Ubx- expressions are mutually exclusive, although, unlike in wing discs (Figs. 5E, F), tkv is expressed in AB cells of haltere discs (Fig. 5E′), while mtv is not expressed in those cells (Fig. 5F′). However, over-expression of Mtv specifically in AB cells using ptc-GAL4 driver of haltere discs caused only a marginal decrease in tkv expression (data not shown). In contrast, knockdown of Ubx in AB cells of haltere discs alone was sufficient to down-regulate tkv levels in those cells. We observed near-complete down-regulation of tkv in AB cells (at the levels of both tkv RNA in situ and tkvP906-lacZ expression patterns) in ptc-GAL4/UAS-UbxRNAi, Df(3R)Ubx- 109 haltere discs. We also observed up-regulation of tkv in AB cells of Cbx wing discs, wherein Ubx is ectopically expressed (Figs. 5G–H, J). Increased levels of Tkv in AB cells may sequester Dpp in the A/P boundary itself and thereby reducing Dpp signaling to distant cells. To reverse this phenomenon, we generated mitotic clones of tkv specifically in the A/P boundary using ptc-GAL4/ UAS-FLP combination. We used a hypomorphic allele of tkv (tkv6) for this purpose as complete removal of tkv affects the cell survival (Gibson and Perrimon, 2005; Shen and Dahmann, Fig. 6. dally and tkv are potential direct targets of Ubx. (A, C) Gene regions of 2005a). We observed increased levels of pMAD in cells dally (A) and tkv (C) showing potential Ubx-binding motifs (shown in red). 6 immediately adjacent to the mitotic clones of tkv . While it was Exons are shown in green and introns in white. Motifs 1, 2, 4 and 5 of dally and equivalent in both anterior and posterior compartments of the 1, 4 and 5 of tkv were individually validated by chromatin immunoprecipitation. wing disc (Fig. 5K), activation of pMAD was relatively higher (B, D) Results of representative chromatin immunoprecipitation experiments showing enrichment of motif 2 of dally (B) and motif 1 of tkv (D) in the in the anterior compartment of the haltere disc (Fig. 5L). In wing chromatin immunoprecipitated with anti-Ubx antibodies. Primary antibodies discs, as tkv levels are lower in AB cells and pMAD is activated were omitted in the control sets. Results with motifs 1, 4 and 5 of dally were in cells away from AB cells, we expect marginal change in inconclusive, while motifs 4 and 5 of tkv did not show any enrichment in the pMAD levels in response to removal of tkv from AB cells. immunoprecipitated chromatin. Ubx-binding motif located on cis-regulatory Indeed, levels of activation of pMAD in anterior cells adjacent region of sal was used as a positive control (which shows enrichment in both the experiments) and sequence of Rpl32 as negative control (whose levels are equal to AB cells were more pronounced in haltere discs than in wing between the control and immunoprecipitated chromatin). Following are the discs. Thus, these observations are an indication of limited quantitative estimations of enrichments (ratio of band intensities of experimental signaling abilities of Dpp in haltere discs due to its retention in over the control ChIP) from three sets of experiments. AB cells by high levels of tkv. E1/C1 E2/C2 E3/C3 However, it is expected that lower levels of tkv outside the Rpl32 0.922 0.958 0.99 AB domain would reduce sensitivity of haltere cells to Dpp and sal 2.709 2.414 1.649 thereby levels of tkv alone may explain absence of pMAD and dally 2.072 3.545 NE other effects. However, the discrepancy between anterior and tkv 2.72 NE 1.883 posterior compartments in their response to reduced tkv levels is NE =not estimated, E=experimental i.e. immunoprecipitation with Ubx more likely to be due to the absence of dally in AB cells as well antibodies. C=control i.e. mock immunoprecipitation without any antibodies. K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 251 binding motifs (each cluster with a minimum of 4 motifs), Effect of reduced Decapentaplegic signaling on haltere growth sparsely distributed on 4 different introns of dally (Fig. 6A). and differentiation We identified 4 such clusters of Ubx-binding motifs concentrated on the second intron towards the 3′ end of tkv Epithelial cells of the wing disc that do not receive Dpp (Fig. 6C). Chromatin immunoprecipitation experiments sug- signals undergo extrusion from the rest of the epithelial cells gested that, for both dally and tkv, at least one of the and die (Gibson and Perrimon, 2005; Shen and Dahmann, predicted sites is indeed a Ubx-binding site in vivo (Figs. 6B, 2005a). Although Dpp signaling is lower in haltere discs, D). Binding of Ubx to regulatory sequences of dally and tkv epithelial cells of the wild type wing and haltere discs do not was further validated by electromobility shift assays using show any difference in shape or size (Figs. 7A, B). Furthermore, purified Ubx protein (data not shown). The ChIP enriched epithelial cells within Ubx− clones (which allows comparison sites of dally, tkv as well as spalt (positive control) were of wing and haltere cells in the same disc rather than from two band-shifted. Rpl32, which was used as a negative control in different discs) are appropriately positioned within the haltere ChIP experiments, was used as negative control in electro- disc epithelium and do not show any obvious change in cell mobility shift assays. Rpl32 did not show any bandshift. shape or size (Fig. 7D). Thus, although Dpp levels and signaling Although sal is a previously known direct target of Ubx are lower in haltere epithelium, the levels are sufficient for cell (based on EMSA, foot-printing and transgenic reporter survival and to maintain wing disc-type cytoskeletal organiza- assays), this is the first time it has been shown to be an in tion. As Ubx− cells do show complete transformation of haltere vivo target of Ubx by ChIP experiments. Considering that it cells to that of wing-type cells in adult flies and over-expression has been tested by several independent assays, sal would be of Dpp in the haltere disc does cause transformations at the an ideal candidate gene as a positive control in whole-genome cellular level, it is likely that differences in the cell shape and ChIP experiments (ChIP-on-chip). size between wing and haltere are manifested during metamor- While above results from ChIP experiments are suggestive phosis. Indeed, Roch and Akam (2000) have observed that wing of dally and tkv being potential direct targets of Ubx, further and haltere cells diverge in their shape and size during pupal in vivo validation is needed to confirm these results. It development. requires generating transgenic lines for the regulatory regions Functions of En and Hh are required to specify compart- (with and without Ubx-binding motifs) and investigating ments, cells of which show compartment-specific sorting their regulation in wing and haltere discs in different genetic behavior (Morata and Lawrence, 1975; Garcia-Bellido et al., backgrounds. 1976). Cells of the A/P boundary play an important role to

Fig. 7. Wing and haltere discs do not differ at the cellular level, while they show differential growth due to down-regulation of Dpp. (A–B) Cross-section along the Z-axis of wild type wing (A) and haltere (B) discs stained for Fas-III showing cellular organization in the columnar epithelium. The two discs are identical at the levels of cellular organization in the epithelium, cell shape and cell size. (C) Haltere disc with hsFLP-induced mitotic clones of Ubx1. Green marks Ubx expression and is absent in large patches of cells. (D) Cross-section along the Z-axis of a region in the haltere disc shown in panel C. Note both Ubx+ cells and Ubx− cells show similar cellular organization in the epithelium, cell shape and cell size. (E–F) Wing and haltere discs showing hsFLP-induced mitotic clones of only arm-lacZ. Note the clones respect the compartment boundary in both the discs, although the clones in the wing discs show a smooth margin along the A/P boundary and such clones in the haltere discs have a wiggly margin. (G–H) Haltere discs with hsFLP-induced mitotic clones of Ubx1 (along with the clonal marker arm-lacZ) generated at 48–72h AEL. Both Ubx− clones and their twin spots respect the compartment boundary, although Ubx− cells are sorted out and hence clones are mostly round in shape. (I) omb-GAL4/+; UAS-Dpp::GFP/+ wing and haltere discs. Note over-growth in both wing and haltere discs. 252 K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 maintain compartmental identities (Dahmann and Basler, 2000). 2002). Thus, during haltere evolution, certain wing-patterning It has been observed that Dpp signaling within the anterior genes may have come under the regulation of Ubx. compartment is more critical for this purpose than signaling Ubx may have acquired its present role in haltere develop- from the anterior to the posterior compartment (Shen and ment through many intermediate steps, and several wing Dahmann, 2005b). As mentioned above, although Dpp patterning genes may have come under its regulation during signaling to the posterior compartment is compromised, AB these changes. Some of these steps/genes are likely to affect the cells do show high levels of pMAD due to higher levels of tkv. wing size, others its pattern, and still others both wing size and This may compensate for reduced Dpp levels and thereby pattern. Dpp signaling is critical to growth and pattering along maintain a compartment boundary. Consistently, clonal ana- the A/P axis of the Drosophila wing disc. We, therefore, lyses with a lineage marker indicated that haltere cells do examined in detail how Ubx modulates Dpp signaling to specify respect the compartment boundary (Fig. 7F). Furthermore, haltere development. Similar and complementary results have mitotic clones of Ubx1 allele and their wild type twin spot also been reported by Crickmore and Mann (2006) and de Navas et respect the compartment boundary, although Ubx− clones show al. (in press). sorting behavior of their own (Figs. 7G, H). These observations suggest that, even in the absence of comparable levels of Dpp Ultrabithorax acts at multiple levels in the Decapentaplegic signaling outside AB cells, wing and haltere discs are patterned signaling pathway in a similar way along the A/P axis. In addition to the overall reduction in the size of the haltere We have observed that differences between wing and haltere disc, the posterior compartment of haltere discs is relatively discs at the level of dpp transcription are less pronounced smaller. The ratio of the size of the anterior to the size of the compared to the degree of down-regulation of Dpp signaling, posterior compartment is approximately 3:1 in wild type haltere suggesting that Ubx affects Dpp signaling at multiple levels. discs (n=12), while it is approximately 1:1 in wing discs Indeed, Ubx modulates the expression patterns of dally and tkv (n=10). This is consistent with the observation that the effect of and preliminary experiments suggest that they both are potential Ubx on pMAD levels is more pronounced in the posterior direct targets of Ubx (Fig. 6). compartment than the anterior compartment. Over-expression It has been observed that over-expression of activated Tkv of Dpp in AB cells using ptc-GAL4 driver caused increase in represses dpp at the transcription level, whereas over-expression the size of the anterior compartment, while there was no change of the dominant negative form of Tkv causes the opposite effect in the size of the posterior compartment (Fig. 3E). Over- (Haerry et al., 1998). Thus, increased levels of tkv in AB cells expression of Dpp using omb-GAL4, however, caused reduc- may have led to increased Dpp signaling in those cells and tion in the differences between anterior and posterior compart- thereby causing reduced levels of dpp. Thus, as dpp does not ments in the haltere discs (Fig. 7I). The ratio of the size of the appear to be a direct target of Ubx (see Supplementary data), anterior to the size of the posterior compartment was reduced to increased levels of tkv in AB cells of the haltere may explain <1.5:1 (n=11). Interestingly, omb-GAL4 expresses only in a lower levels of dpp transcripts in those cells. small part of the posterior compartment of haltere discs, Here we have shown that expression patterns of dpp, tkv, immediately adjacent to the A/P boundary. This suggests that mtv and dally are also modified in haltere discs. Sal, a target one of the main limitations for Dpp to signal to the posterior of Dpp signaling, is a known direct target of Ubx during haltere compartment is crossing the A/P boundary. development (Galant et al., 2002). Thus, Ubx appears to act at Taken together, our results suggest modulation of Dpp multiple levels to regulate Dpp signaling. Although Ubx morphogen gradient in haltere discs by Ubx by acting at modulates Dpp signaling at several levels, modulation of tkv multiple levels. This complexity may allow change in the organ and dally expression appears to be critical for changing the Dpp size and probably also the organ shape, without compromising morphogen gradient. The down-regulated Tkv and Dally in features such as cell survival, which is also dependent on Dpp non-AB cells cause less Dpp signaling in those cells. Increased signaling. Tkv activity in AB cells may cause increased Dpp signaling and thereby decreased dpp levels, which in turn further reduces Dpp Discussion signaling to non-AB cells. Similar to Dpp signaling, Wg signaling (Mohit et al., 2003) Evolution of size and forms of appendages in different insect and Egfr/Ras pathway (Pallavi et al., 2006) too are regulated at groups is one of the main topics in evolutionary developmental multiple levels. Thus, it appears that Ubx acts at multiple levels biology (Evo–Devo). It is generally believed that the Hox in the hierarchy of signaling pathways to specify haltere protein Ubx specifies the differences between forewing and development. It is, however, not clear why such an elaborate hindwing in all insect groups. One of the less-studied aspects of mechanism may have evolved to down-regulate a given homeotic genes is their function to regulate the size and shape of pathway in haltere discs. Unlike in abdominal segments, organs. Interestingly, Ubx protein itself has not evolved among where homeotic genes of Bx-C completely repress appendage the diverse insect groups, although there are significant development, Ubx modifies the fate of the dorsal appendage in differences in Ubx sequences between Drosophila and T3. This necessitates that all signaling events still need to be crustacean Arthropods (Weatherbee et al., 1999; Grenier and functioning. Indeed, total loss of N, Dpp or Wg pathways leads Carroll, 2000; Galant and Carroll, 2002; Ronshaugen et al., to total loss of both wing (Williams et al., 1993) and haltere K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 253 appendages (data not shown). Ubx may modulate signaling that make the organ. This is reflected in the fact that there is no events in a complex way so that at different times during single target gene known so far that mimics Ubx function in development varied amounts of signals are still available to specifying haltere development. All classical haltere-to-wing or specify haltere development. Such complex modulation of wing-to-haltere transformations are always associated with loss wing-patterning events may help canalization of haltere fate and or gain of Ubx, respectively. make any one mutation unable to reverse the fate. All the major signaling pathways operating during organ development such as Wg, Dpp and Egfr/Ras pathways are Change in the morphogen gradient and its effect on organ size pleiotropic in nature and their modulation would affect more and shape than one morphological feature. For example, Dpp signaling pathway regulates growth and proliferation of cells depending Signaling from organizers such as A/P and D/V signaling on the amount of Dpp signal received by the cells, in addition to centers of Drosophila wing disc regulates growth and patterning cell survival and maintenance of the compartment boundary. of the entire wing blade. Here we have shown that Ubx down- Thus, down-regulation of Dpp pathway may affect all functions regulates A/P signaling by modifying the Dpp morphogen of Dpp, unless it is regulated in such a way (temporally and/or gradient. Reduced Dpp signaling leads to reduced growth and quantitatively) that only one of the functions of Dpp is affected. hence smaller size of the haltere. Thus, Ubx may ensure a global Here, we have shown how Ubx modulates Dpp pathway to effect to bring in changes at the organ level by regulating both effect changes in growth pattern, without affecting the cell D/V (Shashidhara et al., 1999; Mohit et al., 2003) and A/P survival and compartment boundary in the haltere epithelium. signaling pathways. The activation of pMAD in the posterior compartment of Acknowledgments abx/Ubx1 haltere discs suggests that inability of Dpp to signal to the posterior compartment may form the basis of differences We thank RAH White for help in ChIP experiments and E in the compartment sizes. Furthermore, we have observed that Sanchez-Herrero for sharing unpublished results. We thank N Ubx− clones and their twin spots are invariably of similar sizes Rangarajan for excellent help on confocal microscope; S Cohen, (Supplementary Fig. 1), which rule out the possibility that M Gonza'lez-Gaita'n, Y Graba, I Guerrero, C-H Heldin, A Khar, haltere cells are intrinsically less responsive to Dpp signaling. H Nakato, N Patel, E Sanchez-Herrero, M Sémériva, T Tabata, Furthermore, as repression of Dally is more pronounced in non- J Treisman, RAH White, Bloomington Stock Center and Deve- AB cells of the posterior compartment, the Dpp gradient is lopmental Studies Hybridoma Bank for fly stocks and skewed towards the anterior compartment and may help in re- antibodies; members of the laboratory for technical help, shaping (the posterior compartment of haltere disc is relatively advice and discussions. KM is supported by a fellowship from smaller compared to its counterpart in the wing disc) the haltere. CSIR, India and was a recipient of Development Traveling Thus, Ubx-mediated modulation of tkv and dally expression is Fellowship to visit Cambridge. critical for determining haltere size and shape. Interestingly, at the levels of D/V signaling too, Ubx completely represses Wg Appendix A. Supplementary data signaling by repressing Wg itself in the posterior compartment, while it partially represses Wg signaling in the anterior Supplementary data associated with this article can be found, compartment (Mohit et al., 2003). Increase in the size of the in the online version, at doi:10.1016/j.ydbio.2006.09.029. posterior compartment has been observed in different genetic RNAi 1 backgrounds. In ptc-GAL4/UAS-Ubx or abx/Ubx discs, References this could be due to increased Dpp signaling (note pMAD activation is non-cell-autonomously restored), whereas in pbx/ Belenkaya, T.Y., Han, C., Yan, D., Opoka, R.J., Khodoun, M., Liu, H., Lin, X., 1 Ubx discs, it could be due to restoration of Wg signaling. Thus, 2004. Drosophila Dpp morphogen movement is independent of dynamin- these results suggest compartment-specific function of Ubx in mediated endocytosis but regulated by the glypican members of heparan – shaping the haltere. sulfate proteoglycans. Cell 119, 231 244. Berger, C., Pallavi, S.K., Prasad, M., Shashidhara, L.S., Technau, G.M., 2005. A critical role for Cyclin E in cell fate determination in the central nervous Organ identity vs. cellular identities system of Drosophila melanogaster. Nat. Cell Biol. 7, 56–62. Birch-Machin, I., Gao, S., Huen, D., McGirr, R., White, R.A.H., Russell, S., Hox proteins AbdA and AbdB specify the lineage of the 2005. Genomic analysis of heat shock factor targets in Drosophila. Genome embryonic NB6-4 neuroblast in abdominal segments by down- Biol. 6, R63. Blackman, R.K., Sanicola, M., Raftery, L.A., Gillevet, T., Gelbart, W.M., 1991. regulating CycE (Berger et al., 2005). Differential expression of An extensive 3′ cis-regulatory region directs the imaginal disk expression of CycE is both required and sufficient to generate segmental decapentaplegic, a member of the TGF-beta family in Drosophila. differences in NB6-4 lineage. Similarly, specification of the Development 111, 657–666. larval oenocyte is dependent on the regulation of just one Brand, A.H., Perrimon, N., 1993. Targeted gene expression as a means of principal target, Rho, by the homeotic gene abdominal A altering cell fates and generating dominant phenotypes. Development 118, 401–415. (Brodu et al., 2002). In contrast, modulation of an organ identity Brodu, V., Elstob, P.R., Gould, A.P., 2002. abdominal A specifies one cell type would be complicated as they are intricately linked to the in Drosophila by regulating one principal target gene. Development 129, collective proliferation behavior and identities of all the cells 2957–2963. 254 K. Makhijani et al. / Developmental Biology 302 (2007) 243–255

Brummel, T., Twombly, V., Marques, G., Wrana, J.L., Newfeld, S.J., Attisano, to direct the pattern of retinal differentiation in the eye disc. Development L., Massague, J., O'Connor, M.B., Gelbart, W.M., 1994. Characterization 125, 3741–3751. and relationship of Dpp receptors encoded by the Saxophone. Thick veins Held Jr., L.I., 2002. Imaginal Discs. Cambridge University Press, p. 85. genes in Drosophila. Cell 78, 251–261. Ingham, P.W., 1998. Boning up on Hedgehog's movements. Nature 394, Capdevila, J., Pariente, F., Sampedro, J., Alonso, J.L., Guerrero, I., 1994. 16–17. Subcellular localization of the segment polarity protein Patched suggests an Irvine, K., Weischaus, E., 1994. Fringe, a boundary-specific signaling molecule, interaction with the Wingless receptor complex in Drosophila embryos. mediates interactions between dorsal and ventral cells during Drosophila Development 120, 987–998. wing development. Cell 79, 595–606. Castelli-Gair, J., Greig, S., Micklem, G., Akam, M., 1994. Dissecting the Kim, J., Irvine, K.D., Carroll, S.B., 1995. Cell recognition, signal induction, and temporal requirements for homeotic gene function. Development 120, symmetrical gene activation at the dorsal/ventral boundary of the developing 1983–1995. Drosophila wing. Cell 82, 795–802. Cohen, S.M., 1993. Imaginal disc development. In: Bate, M., Martinez-Arias, A. Lecuit, T., Cohen, S.M., 1998. Dpp receptor levels contribute to shaping the Dpp (Eds.), The Development of Drosophila melanogaster. Cold Spring Harbor morphogen gradient in the Drosophila wing imaginal disc. Development Laboratory Press, pp. 747–841. 125, 4901–4907. Crickmore, M.A., Mann, R.S., 2006. Hox control of organ size by regulation of Lecuit, T., Brook, W.J., Ng, M., Calleja, M., Sun, H., Cohen, S.M., 1996. Two morphogen production and mobility. Science 313, 63–68. distinct mechanisms for long-range patterning by Decapentaplegic in the Dahmann, C., Basler, K., 2000. Opposing transcriptional outputs of Hedgehog Drosophila wing. Nature 381, 387–393. signaling and Engrailed control compartmental cell sorting at the Droso- Lewis, E.B., 1978. A gene complex controlling segmentation in Drosophila. phila A/P boundary. Cell 100, 411–422. Nature 276, 565–570. de Celis, J.F., 1997. Expression and function of decapentaplegic and thick veins Masucci, J.D., Miltenberger, R.J., Hoffmann, F.M., 1990. Pattern-specific during the differentiation of the veins in the Drosophila wing. Development expression of the Drosophila decapentaplegic gene in imaginal disks is 124, 1007–1018. regulated by 3′ cis-regulatory elements. Genes Dev. 4, 2011–2023. de Navas, L.F., Garaulet, D.L., Sánchez-Herrero, E., in press. The Ultrabithorax McGuire, S.E., Le, P.T., Osborn, A.J., Matsumoto, K., Davis, R.L., 2003. Hox gene of Drosophila controls haltere size by regulating the dpp pathway. Spatiotemporal rescue of memory dysfunction in Drosophila. Science 302, Development. 1765–1768. Diaz-Benjumea, F.J., Cohen, S.M., 1993. Interaction between dorsal and ventral Mohit, P., Bajpai, R., Shashidhara, L.S., 2003. Regulation of Wingless and cells in the imaginal disc directs wing development in Drosophila. Cell 75, Vestigial expression in wing and haltere discs of Drosophila. Development 741–752. 130, 1537–1547. Diaz-Benjumea, F.J., Cohen, S.M., 1995. Serrate signals through Notch to Mohit, P., Makhijani, K., Madhavi, M.B., Bharathi, V., Lal, A., Sirdesai, G., establish a Wingless-dependent organizer at the dorsal/ventral compartment Ram Reddy, V., Palaparthi, R., Kannan, R., Dhawan, J., Shashidhara, L.S., boundary of the Drosophila wing. Development 121, 4215–4225. 2006. Modulation of AP and DV signalling pathways by the Homeotic gene Ekker, S.C., Young, K.E., Von Kessler, D., Beachy, P.A., 1991. Optimal DNA Ultrabithorax during haltere development in Drosophila. Dev. Biol. 291, sequence recognition by the Ultrabithorax homeodomain of Drosophila. 356–367. EMBO J. 10, 1179–1186. Monier, B., Astier, M., Sémériva, M., Perrin, L., 2005. Steroid-dependent Entchev, E.V., Schwabedissen, A., González-Gaitán, M., 2000. Gradient modification of Hox function drives myocyte reprogramming in the Dro- Formation of the TGF-β Homolog Dpp. Cell 103, 981–991. sophila heart. Development 132, 5283–5293. Frasch, M., 1995. Induction of visceral and cardiac mesoderm by ectodermal Morata, G., Garcia-Bellido, A., 1976. Developmental analysis of some mutants Dpp in the early Drosophila embryo. Nature 374, 464–467. of the bithorax system of Drosophila. Roux's Arch. Dev. Biol. 179, Fujise, M., Izumi, S., Selleck, S.B., Nakato, H., 2001. Regulation of Dally, an 125–143. integral membrane proteoglycan and its function during adult sensory organ Morata, G., Lawrence, P.A., 1975. Control of compartment development by the formation of Drosophila. Dev. Biol. 235, 433–448. engrailed gene in Drosophila. Nature 255, 614–617. Fujise, M., Takeo, S., Kamimura, K., Matsuo, T., Aigaki, T., Izumi, S., Nakato, Morimura, S., Maves, L., Chen, Y., Hoffmann, F.M., 1996. decapentaplegic H., 2003. Dally regulates Dpp morphogen gradient formation in the Dro- overexpression affects Drosophila wing and leg imaginal disc development, sophila wing. Development 130, 1515–1522. wingless expression. Dev. Biol. 177, 136–151. Funakoshi, Y., Minami, M., Tabata, T., 2001. mtv shapes the activity gradient of Motzny, C.K., Holmgren, R., 1995. The Drosophila Cubitus interruptus protein the Dpp morphogen through regulation of thickveins. Development 128, and its role in the Wingless, Hedgehog signal transduction pathways. Mech. 67–74. Dev. 52, 137–150. Galant, R., Carroll, S.B., 2002. Evolution of a transcriptional repression domain Nellen, D., Burke, R., Struhl, G., Basler, K., 1996. Direct and long-range action in an insect Hox protein. Nature 415, 910–913. of a DPP morphogen gradient. Cell 85, 357–368. Galant, R., Walsh, C.M., Carroll, S.B., 2002. Hox repression of a target gene; Pallavi, S.K., Kannan, R., Shashidhara, L.S., 2006. Negative regulation of Egfr/ Extradenticle-independent and additive action through multiple monomer Ras pathway by Ultrabithorax during haltere development in Drosophila. binding sites. Development 129, 3115–3126. Dev. Biol. 296, 340–352. Garcia-Bellido, A., Ripoll, P., Morata, G., 1976. Developmental compartmen- Patel, N.H., Martin-Blanco, E., Coleman, K.G., Poole, S.J., Ellis, M.C., talization in the dorsal mesothoracic disc of Drosophila. Dev. Biol. 48, Kornberg, T.B., Goodman, C.S., 1989. Expression of Engrailed proteins in 132–147. arthropods, annelids and chordates. Cell 58, 955–968. George, H., Terracol, R., 1997. The vrille gene of Drosophila is a maternal Persson, U., Izumi, H., Souchelnytskyi, S., Itoh, S., Grimsby, S., Engstrom, U., enhancer of decapentaplegic and encodes a new member of the bZIP family Heldin, C.-H., Funa, K., ten Dijke, P., 1998. The L45 loop in type I receptors of transcription factors. Genetics 146, 1345–1363. for TGF-β family members is a critical determinant in specifying Smad Gibson, M.C., Perrimon, N., 2005. Extrusion and death of DPP/BMP- isoform activation. FEBS Lett. 434, 83–87. compromised epithelial cells in the developing Drosophila wing. Science Posakony, L.G., Raftery, L.A., Gelbart, W.M., 1990. Wing formation in Dro- 307, 1785–1789. sophila melanogaster requires decapentaplegic gene function along the Grenier, J.K., Carroll, S.B., 2000. Functional evolution of the Ultrabithorax anterior–posterior compartment boundary. Mech. Dev. 33, 69–82. protein. Proc. Natl. Acad. Sci. U. S. A. 97, 704–709. Roch, F., Akam, M., 2000. Ultrabithorax and the control of cell morphology in Haerry, T.E., Khalsa, O., O'Connor, M.B., Wharton, K.A., 1998. Synergistic Drosophila halteres. Development 127, 97–107. signaling by two BMP ligands through the SAX, TKV receptors controls Rogulja, D., Irvine, K.D., 2005. Regulation of cell proliferation by a morphogen wing growth, patterning in Drosophila. Development 125, 3977–3987. gradient. Cell 123, 449–461. Hazelett, D.J., Bourouis, M., Walldorf, U., Treisman, J.E., 1998. Decapenta- Ronshaugen, M., McGinnis, N., McGinnis, W., 2002. Hox protein mutation and plegic and Wingless are regulated by Eyes absent and Eyegone and interact macroevolution of the insect body plan. Nature 415, 914–917. K. Makhijani et al. / Developmental Biology 302 (2007) 243–255 255

Shashidhara, L.S., Agrawal, N., Bajpai, R., Bharathi, V., Sinha, P., 1999. Wang, Q.T., Holmgren, R.A., 2000. Nuclear import of Cubitus interruptus is Negative regulation of dorsoventral signaling by the homeotic gene regulated by Hedgehog via a mechanism distinct from Ci stabilization, Ci Ultrabithorax during haltere development in Drosophila. Dev. Biol. 212, activation. Development 127, 3131–3139. 491–502. Weatherbee, S.D., Halder, G., Kim, J., Hudson, A., Carroll, S., 1998. Shen, J., Dahmann, C., 2005a. Extrusion of cells with inappropriate Dpp Ultrabithorax regulates genes at several levels of the wing-patterning signaling from Drosophila wing disc epithelia. Science 307, 1789–1790. hierarchy to shape the development of the Drosophila haltere. Genes Dev. Shen, J., Dahmann, C., 2005b. The role of Dpp signaling in maintaining the 12, 1474–1482. Drosophila anteroposterior compartment boundary. Dev. Biol. 279, 31–43. Weatherbee, S.D., Nijhout, H.F., Grunert, L.W., Halder, G., Galant, R., Selegue, Singer, M.A., Penton, A., Twombly, V., Hoffmann, F.M., Gelbart, W.M., 1997. J., Carroll, S., 1999. Ultrabithorax function in butterfly wings and the Signaling through both type I DPP receptors is required for anterior– evolution of insect wing patterns. Curr. Biol. 9, 109–115. posterior patterning of the entire Drosophila wing. Development 124, Weinmann, A.S., Farnham, P.J., 2002. Identification of unknown target genes of 79–89. human transcription factors using chromatin immunoprecipitation. Methods St. Johnston, R.D., Hoffmann, F.M., Blackman, R.K., Segal, D., Grimaila, R., 26, 37–47. Padgett, R.W., Irick, H.A., Gelbart, W.M., 1990. Molecular organization of White, R., Wilcox, M., 1984. Protein products of the Bithorax complex in the decapentaplegic gene in Drosophila melanogaster. Genes Dev. 4, Drosophila. Cell 39, 163–171. 1114–1127. Williams, J.A., Paddock, S.W., Carroll, S.B., 1993. Pattern formation in a Tabata, T., Kornberg, T.B., 1994. Hedgehog is a signaling protein with a key role secondary field: a hierarchy of regulatory genes subdivides the in patterning Drosophila imaginal discs. Cell 76, 89–102. developing Drosophila wing disc into discrete sub-regions. Development Tanimoto, H., Itoh, S., ten Dijke, P., Tabata, T., 2000. Hedgehog creates a 117, 571–584. gradient of DPP activity in Drosophila wing imaginal discs. Mol. Cell 5, Williams, J.A., Paddock, S.W., Vorwerk, K., Carroll, S.B., 1994. Organization 59–71. of wing formation and induction of a wing-patterning gene at the dorsal/ Tautz, D., Pfeifle, C., 1989. A non-radioactive in situ hybridization method for ventral compartment boundary. Nature 368, 299–305. the localization of specific RNAs in Drosophila embryos reveals translational Xu, T., Harrison, S.D., 1994. Mosaic analysis using FLP recombinase. Methods control of the segmentation gene hunchback. Chromosoma 98, 81–85. Cell Biol. 44, 655–681. Teleman, A.A., Cohen, S.M., 2000. Dpp gradient formation in the Drosophila Xu, T., Rubin, G.M., 1993. Analysis of genetic mosaics in developing and adult wing imaginal disc. Cell 103, 971–980. Drosophila tissues. Development 117, 1223–1237.