Long-range Dpp signaling is regulated to restrict BMP signaling to a crossvein competent zone

Developmental Biology 280 (2005) 187 – 200 www.elsevier.com/locate/ydbio

Long-range Dpp signaling is regulated to restrict BMP signaling to a crossvein competent zone Amy Ralston, Seth S. Blair* Department of Zoology, 250 N. Mills Street, University of Wisconsin, Madison, WI 53706, USA Received for publication 11 June 2004, revised 8 January 2005, accepted 11 January 2005

Abstract The sensitivity of the crossveins of the Drosophila wing to reductions in BMP signaling provides a valuable system for characterizing members of this signaling pathway. We demonstrate here two reasons for that sensitivity. First, the initial stage of posterior crossvein development depends on BMP signaling but is independent of EGF signaling. This is the opposite of the longitudinal veins, which rely of EGF signaling for their initial specification. Second, BMP signaling in the posterior crossvein depends on Decapentaplegic (Dpp) at a stage when it is being produced in the longitudinal veins. Thus, the posterior crossvein will be especially vulnerable to reductions in the levels or range of Dpp signaling. We investigated the roles of the BMP receptor Thickveins (Tkv) and the BMP inhibitor Short gastrulation (Sog) in allowing this long-range signaling. Expression of both is downregulated in the developing posterior crossvein. The Tkv downregulation depends on BMP signaling and may provide a positive feedback by allowing the spread of Dpp. The Sog downregulation is independent of BMP signaling; Sog misexpression experiments indicate that this prepattern is essential for posterior crossvein development. However, this requirement can be overridden by co-misexpression of the BMP agonist Cv-2, indicating the presence of as yet unknown cues; we discuss possible candidates. D 2005 Elsevier Inc. All rights reserved. Keywords: Chordin; BMP receptor; Smad; Cv-2; BMPER

Introduction A key morphological feature of the adult Drosophila wing is the stereotyped array of veins, thickenings of the ectodermal cuticle that serve both as structural supports and as conduits for nutrients, nerves and trachea for the wing. These arise, not via an invasive process, but by localized differentiation within the developing wing epithelium. Because of their relatively invariant positions and the ease with which they can be examined, the wing veins have served as an important model for the study of the genetics of pattern formation (reviewed in Bier, 2000; de Celis, 2003; de Celis and Diaz-Benjumea, 2003). Moreover, the integrity of the veins requires localized signaling via two important and highly conserved signaling pathways. The Drosophila

* Corresponding author. E-mail address: [email protected] (S.S. Blair). 0012-1606/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2005.01.018

Epidermal growth factor receptor (Egfr) and the Bone Morphogenetic Protein (BMP) signaling pathways are critical for vein development, and thus the molecular analysis of vein mutations has identified many novel members of these signaling pathways. There are two main classes of veins, the longitudinal veins, oriented along the proximal distal axis, and the crossveins–the anterior crossvein (ACV) and the posterior crossvein (PCV)–oriented perpendicular to the longitudinal veins (Fig. 1A). These two classes of veins share several features; for instance, both require Egfr and BMP signaling for their appearance in adults (Burke and Basler, 1996; DiazBenjumea and Garcia-Bellido, 1990). However, they also differ in several respects. While the longitudinal vein primoridia first appear during late larval stages, the crossveins do not appear until later pupal stages (Conley et al., 2000; Waddington, 1940). Moreover, the development of the crossveins and especially the PCV can be selectively disrupted by reductions in BMP signaling that leave the

188

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

longitudinal veins largely or entirely intact (de Celis, 1997; Haerry et al., 1998; Khalsa et al., 1998; Nguyen et al., 1998; Ray and Wharton, 2001; Wharton et al., 1999; Yu et al., 1996). Intriguingly, the PCV is also lost after mutation of any of the several crossveinless loci, suggesting that the crossveins follow a qualitatively different genetic program (Conley et al., 2000; reviewed in de Celis and Diaz-Benjumea, 2003). The two crossveinless loci that have been molecularly

characterized encode secreted modulators of BMP signaling. Crossveinless 2 (Cv-2) (Conley et al., 2000), like its vertebrate orthologs (Binnerts et al., 2004; Coffinier et al., 2002; Moser et al., 2003), contains five cysteine-rich (CR) domains that are similar to the putative BMP ligandbinding domains of Short gastrulation (Sog) and its vertebrate homolog Chordin (Larrain et al., 2000). Mammalian Cv-2s can bind BMPs in vitro (Binnerts et al., 2004; Moser et al., 2003). Crossveinless (Cv) is a second Drosophila member of the Twisted gastrulation (Tsg) family (Ross et al., 2001; Vilmos et al., 2001; Shimmi et al., unpublished). Tsg proteins form complexes with BMP ligands and Chordin-like molecules and are therefore important modulators of BMP signaling (Chang et al., 2001; Larrain et al., 2001; Oelgeschla¨ger et al., 2000, 2003; Ross et al., 2001; Scott et al., 2001; Shimmi and O’Connor, 2003). The sensitivity of the PCV to BMP signaling therefore provides a powerful model for identifying new BMP signaling pathway members and suggests that BMP signaling plays a unique role during crossvein development. However, the nature of that role has not been experimentally analyzed. To better understand this sensitivity, we have here performed an analysis of the relative roles of Egfr and BMP signaling during crossvein development, comparing our findings to what is previously known from studies of the longitudinal veins. In the longitudinal veins, short-range BMP signaling maintains vein cell fates established much earlier in development by Egfr signaling. Egfr signaling is activated within the longitudinal vein primordia in mid to late third larval instar and is both necessary and sufficient to induce the expression of vein-specific markers (Diaz-Benjumea and Garcia-Bellido, 1990; Diaz-Benjumea and Hafen, 1994; Gabay et al., 1997; Guichard et al., 1999; MartinBlanco et al., 1999; Roch et al., 1998; Sturtevant and Bier, 1995). Continued Egfr signaling within the longitudinal veins is required during pupal stages to maintain the vein fate (Sturtevant and Bier, 1995). Late in this process, Egfr signaling activates BMP signaling within the longitudinal Fig. 1. Egfr f2 homozygous clones do not affect early PCV development. (A) The wild-type adult wing contains longitudinal veins 1–5 (L1–L5, indicated) and anterior (ACV) and posterior (PCV) crossveins. (B) Egfr f2 homozygous clone (marked by absence of anti-Myc staining) straddling the early PCV in a 25 h AP wing (equivalent to ~21 h AP in wild type). (BV) Heightened pMad is still detectable within Egfr f2 homozygous clone (red outline) shown in panel B. (C) Egfr f2 homozygous clone (marked by absence of anti-Myc staining) in a 25 h AP wing straddling the PCV. (CV) Anti-DSRF staining is still repressed in the PCV within Egfr f2 homozygous clone (red outline) shown in panel B. (D) Egfr f2 homozygous clones (marked by absence of anti-Myc staining) in mid third instar wing imaginal disc. (DV) Anti-DSRF staining, which is just beginning to be downregulated in the primordia of L3 and L4, is upregulated in the Egfr f2 clone (red outline) that straddles the L4 primordium, as well as in other clones in the disc. (E) Egfr f2 homozygous clone (red outline) in a 30 h AP wing disrupts anti-pMad staining in the PCV. (F) Egfr f2 homozygous clone (red outline) in a 30 h AP wing disrupts anti-Dl staining in the PCV.

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

veins (Conley et al., 2000) by inducing the vein-specific expression of the BMP-like ligand Decepentaplegic (Dpp) (de Celis, 1997; Yu et al., 1996). This Dpp expression is required for the maintenance of Egfr signaling and the continued development of most portions of the longitudinal veins (de Celis, 1997; Posakony et al., 1991; Ray and Wharton, 2001). The generally expressed BMP ligand Glass bottom boat (Gbb; formerly known as 60A) also contributes to signaling and is required for the maintenance of the distal portions of the longitudinal veins (Khalsa et al., 1998; Ray and Wharton, 2001). The Thickveins (Tkv) type I BMP receptor is apparently more critical for this signaling than the alternative type I receptor Saxaphone (Burke and Basler, 1996; Ray and Wharton, 2001; Singer et al., 1997). We will show that the development of the PCV differs from that of the longitudinal veins in two important respects. First, heightened BMP signaling precedes Egfr signaling within the PCV, and Egfr signaling is not required for the early stages of PCV development. This is the opposite of what occurs in the longitudinal veins, whose initial development is dependent on Egfr. Second, the initial stages of BMP signaling in the PCV depend in part on long-range signaling by Dpp produced outside the PCV region, likely from the longitudinal veins. Thus, the PCV will be especially sensitive to manipulations that reduce the levels or range of BMP signaling. This long-range Dpp signaling from the longitudinal veins into the PCV provides a potential mechanism to spatially coordinate development of these different veins. We have investigated the mechanisms that allow Dpp to signal over long range to the presumptive PCV, examining the roles of the BMP receptor Tkv and of Sog. Tkv is required for Dpp signaling and vein formation in the wing (Burke and Basler, 1996; Tanimoto et al., 2000), but high levels of Tkv limit the range of Dpp signaling, presumably by binding Dpp and preventing its movement to more distant cells (de Celis, 1997; Haerry et al., 1998; Lecuit and Cohen, 1998). Sog and Chordin attenuate BMP signaling by binding and sequestering BMP ligands (Larrain et al., 2000, 2001; Oelgeschla¨ger et al., 2000, 2003; Piccolo et al., 1996; Ross et al., 2001; Scott et al., 2001; Shimmi and O’Connor, 2003). We show that the expression of both Tkv and Sog is high within intervein regions, but lowered within the presumptive crossveins, where BMP signaling is highest. We show that the Tkv expression pattern is a consequence of heightened BMP signaling within the PCV. Therefore, although the Tkv expression pattern is a contributing factor to long-range BMP signaling, it is unlikely to be its cause. sog expression is, however, independent of BMP signaling in the PCV, and our tests suggest that restricted sog expression is important for PCV patterning. However, we will show experimental conditions where unpatterned sog expression does not disrupt BMP signaling in the crossveins, indicating the existence of unknown regulators of BMP signaling.

189

Materials and methods Drosophila stocks Gal4/UAS-mediated transgene overexpression (Brand and Perrimon, 1993) was achieved using engrailed-gal4 (en-gal4) or heat shock-gal4 (hs-gal4). UAS stocks were UAS-GFP.nls (II), UAS-dppIR (Inaki et al., 2002) and UASsog (on chromosome III) (Yu et al., 1996). EP(2)1103 was used to overexpress cv-2 (Conley et al., 2000). The mutant and FRT stocks used were as follows: y w hsflp; Dp(2;2)B16 pM FRT 40 /CyO and dpp H46 FRT 40 /CyO23-dpp + (Ray and Wharton, 2001). y w. cv-2 1 . gbb 1 /SM5a-TM6,Tb. gbb 4 / SM5a-TM6,Tb. rho ve vn 1 . y w hsflp; FRT 42D M(2)53 pM/ CyO and FRT 42D Egfr f2 /CyO. gbb 1 is a null allele and gbb 4 is a hypomorphic allele (Wharton et al., 1999). Egfr f2 (also known as IK, IK35 and 1K35) contains a stop codon before the kinase domain (Clifford and Schupbach, 1989). Developmental staging The developmental stages of pupae were measured in hours from white prepupal stage. White prepupae were defined as individuals that had ceased movement, everted anterior spiracles, but had not yet begun tanning of cuticle. Individual white prepupae were picked from crosses and reared at 258C. Mitotic recombinant clones Bottles of progeny from the cross y w hsflp; FRT 42D M(2)53 pM/CyO  FRT 42D Egfr f2 /CyO were heat shocked to induced mitotic recombination after a 5 day egg-lay. White prepupae were picked after two or three additional days at 258C. Expression of the kM (Myc) clone marker was induced by heat shocking pupae at 388C for 3 h before dissection. Bottles of progeny from the cross y w hsflp; Dp(2;2)B16 kM FRT 40 /CyO  dpp H46 FRT 40 /CyO23-dpp + were heat-shocked at 388C for 2–3 h after a 3 day egg-lay. Wandering third instar to white prepupal stages were picked after three additional days at 258C. Expression of the kM clone marker was induced by heat shocking pupae at 388C from 26 to 29 h after being picked. Pupae were dissected and fixed as described below. Immunofluorescent localization Pupal wing dissection, fixation and staining were performed as previously described (Blair, 2000; Blair and Ralston, 1997) with the following modifications: Igepal was used instead of Nonidet P-40 in the fixation butter, and the fixation time was reduced to 4–6 h at 48C. Rabbit anti-pMad (Persson et al., 1998) was used at 1:2,000, rat anti-DSRF (Affolter et al., 1994) at 1:1,000, mouse anti-DSRF (Geneka) at 1:1,000, mouse anti-Myc (Santa Cruz) at 1:1,000 and rat anti-Tkv (Teleman and Cohen, 2000) at 1:100. Secondary

190

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

and tertiary antibodies were used as described in Blair and Ralston (1997). In situ hybridization In situ hybridizations in pupal wings were performed as described for discs (http://www.stanford.edu/~rnusse/ ownpage/InsituHybrid.html) with the following modifications. Pupal cases were removed from pupae in PBS and pupae were then fixed in 4% formaldehyde/PBS for 1 h at 48C. Pupal wings were then dissected from cuticle in PBT (PBS + 0.1% Tween 20) containing 4% formaldehyde and 0.1% deoxycholate, then processed as described up through the 2 min methanol wash. Following this step, pupal wings were digested with 4 Ag/mL Proteinase K in PBT for 10 min at room temperature, followed by five quick washes with PBT, a 20 min postfix in PBT containing 5% non-EM grade formaldeyde at room temperature, and then processed further as described. The hybridization buffer contained 5 DEPC-treated SSC, 50% formamide, 1% sonicated herring sperm DNA, 0.02% tRNA and 0.1% Tween 20. Dig-labeled riboprobes were prepared from cDNAs (kindly provided by M.B. O’Connor) according to manufacturer’s instructions (Roche). Following ethanol precipitation, probes were resuspended in 100 AL hybridization buffer. tkv and sog probes were used at 1:25 and dpp probe at 1:50.

Results The initial stages of PCV development are independent of Egfr signaling At 19 h after pupariation (AP), the longitudinal veins can be detected using a variety of markers, including reduced expression of the of intervein-specific transcription factor Drosophila Serum Response Factor (DSRF; also known as Blistered) (Montagne et al., 1996). The presumptive crossveins do not express vein markers at this stage and, like intervein cells, continue to express high levels of DSRF (Conley et al., 2000). The presumptive crossveins can be detected, however, using antiserum generated against the active, phosphorylated form of the BMP signaling effector Mothers against Dpp (anti-pMad) (ibid.). At 19 h AP, higher levels of nuclear pMad are visible in the crossveins as well as the tips of the longitudinal veins and a subset of cells along the anterior wing margin; lower levels are also visible in portions of the longitudinal veins (see Fig. 4A). During subsequent stages, pMad levels increase in all the veins. DSRF is gradually lost from the crossveins, beginning around 19–21 h AP. This period, from 19–21 h AP, is therefore the earliest stage during which vein-specific markers can be detected within the developing crossveins. Egfr signaling is critical for the initial stages of longitudinal vein development. However, our previous

results indicated that known targets of Egfr signaling are not expressed in the crossveins until well after the appearance of pMad (Conley et al., 2000). To test whether earlier Egfr signaling is required for the initial stages of PCV development, we generated clones of Egfr f2 , a null allele. Egfr mutant clones have reduced viability (DiazBenjumea and Garcia-Bellido, 1990), so to ensure the survival of clones, we used the Minute technique (reviewed in Blair, 2003). Since this Minute mutation delays development, we examined pupal wings at 25 h AP, which is developmentally equivalent to wild-type 21 h AP, by morphology and by the expression of vein and intervein markers. Egfr f2 clones that encompassed only the PCV did not disrupt Mad phosphorylation (5/5 clones, Figs. 1B,BV). Interestingly, Egfr f2 clones also did not heighten the expression of intervein-specific DSRF within the early PCV (5/5 clones, Figs. 1C,CV). This contrasts strongly with the situation in the longitudinal vein primordia. At mid to late third instar, DSRF expression is downregulated in the longitudinal veins by Ras-mediated Egfr signaling (Roch et al., 1998). To confirm that the longitudinal vein phenotype is a defect in the initiation, rather than the maintenance, of longitudinal vein development, we examined Egfr f2 clones during mid third larval instar, when the suppression of DSRF expression is barely visible. Such clones raised DSRF expression within the longitudinal veins; in fact, DSRF expression was even heightened in clones lying in intervein regions (Figs. 1D,DV). Thus, DSRF regulation is quite different during the initial stages of longitudinal vein and PCV specification, and its suppression in the PCV is initially independent of Egfr signaling. Nonetheless, beginning at 24 h AP, markers of Egfr signaling, such as activation of Map Kinase, appear in the crossveins (Conley et al., 2000). Consistent with this observation, we find that Egfr signaling is required during later stages of PCV development. When examined at 30 h AP, Egfr f2 clones that encompassed the PCV reduced pMad (Fig. 1E) and the vein-specific marker Delta (Huppert et al., 1997) (Fig. 1F). Thus, while Egfr signaling does not initiate PCV development, it is required during a later maintenance stage. Early expression of BMP ligands in the pupal wing These results suggest that localized BMP signaling plays a critical, early role in crossvein patterning, prior to Egfr signaling. Either Gbb or Dpp ligands might be responsible for BMP signaling during the early stages of crossvein development. Both are expressed in the pupal wing at this stage, gbb ubiquitously (Conley et al., 2000) and dpp in vein cells (de Celis, 1997; Yu et al., 1996). However, previous reports suggest that dpp is not expressed within the crossveins until some time between 20 and 30 h AP (de Celis et al., 1997; Yu et al., 1996), apparently after the initial stages of crossvein formation.

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

To confirm this and to determine the precise timing of the onset of dpp expression in the PCV, we performed an in situ time course with antisense dpp probe in pupal wings aged 5.5, 19, 24, 26 and 28 h AP (Figs. 2A–E). At 5.5 h AP, the compartment boundary expression of dpp was already modified, forming two distinct domains of expression in the third and fourth longitudinal proveins (Fig. 2A). At 19 h AP and later, dpp was expressed in all of the longitudinal veins. However, it was not detectable in cells of the presumptive PCV until 26–28 h AP (Figs. 2D,E). To visualize low levels of dpp transcript, we allowed the alkaline phosphatase reaction to proceed longer than was standard, but were still unable to detect dpp in the presumptive PCV at 19 h AP (Fig. 2B). Thus, if dpp is required for formation BMP signaling during the initial stages of PCV formation, it is diffusing into this region from the adjacent longitudinal veins. The situation at the ACV is more complex. At 5.5 h AP, there was slightly increased dpp expression at the intersection of the third and fourth longitudinal proveins (Fig.

191

2A). This is the position of the ACV campaniform sensillum, and thus lies near the cells that will form the ACV (Conley et al., 2000). At 19 and 24 h AP, dpp was expressed in a stripe of cells that ran parallel to the third and fourth longitudinal veins, intersecting the presumptive ACV (Figs. 2B,C). However, this stripe did not encompass the entire ACV and was not congruent with the cells of the ACV or the pattern of pMad. This stripe was no longer detectable at 26 or 28 h AP (Figs. 2D,E). If dpp expression ever becomes congruent with the ACV, as it does for the other veins, it must do so at a later stage of development. Nonetheless, this stripe of dppexpressing cells could serve as a source of Dpp during the initial stages of ACV development. Dpp and Gbb are both required for initial BMP signaling in the crossveins To test the requirement for Dpp during the initial stages of crossvein development, we generated dpp null clones, examining wings between 22–26 h AP. Consistent with

Fig. 2. Expression of dpp during crossvein development. (A) At 5.5 h AP, dpp transcript is expressed in longitudinal veins 3 and 4 (indicated), and at high levels in an ACV-like region (arrowhead). (B) At 19 h AP, dpp is expressed in all longitudinal veins and in a horizontal streak of cells in the region of the ACV (arrowhead). (C) At 24 h AP, dpp expression resembles expression of dpp at 19 h AP. (D) At 26 h AP, dpp is faintly detectable in cells of the PCV (arrowhead). dpp expression is no longer detectable near the ACV. (E) At 28 h AP, dpp is expressed more robustly in the PCV (arrowhead).

192

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

what had been previously observed in adult wings (de Celis, 1997; Posakony et al., 1991; Ray and Wharton, 2001), clones limited to one surface of the wing had little or no effect on venation. However, the dpp produced on one surface can signal to the other during pupal stages, as we have observed using FLPout clones expressing ectopic dpp (not shown). We therefore examined the regions of overlap between clones on the dorsal and ventral surfaces. Ray and Wharton (2001) previously showed that in adult wings the PCV is lost in a nearly cell-autonomous manner from overlapping dorsal and ventral dpp clones; crossvein tissue is only found 2–3 cells inside the edge of the region of overlap. However, when we examined the initial stages of PCV development, we found no consistent cell-autonomous disruption of the PCV within regions of overlap (Figs. 3A,AV, arrows, anti-DSRF). This is in agreement with our in situ data and suggests that the initial stage of BMP signaling in the PCV does not rely on the expression of dpp within the PCV itself. The nearly cell-autonomous loss observed in adults thus must occur after the initial stages of PCV development, likely after 26 h AP when dpp expression begins to appear along the PCV. We did, however, observe two cases with disruption of a portion of the PCV. In one, there was slight disruption of

Fig. 3. Dpp is not required within the PCV for early PCV development. The figures show the outlines of dpp null clones on one (green) or the other (red) surface of 21–26 h AP pupal wings; regions of overlap are shaded red. (A,AV) Vein development identified by the absence of anti-DSRF staining. Despite the presence of overlapping clones over substantial portions of the PCV, PCV development within the clone is unperturbed (arrows). However, some defects are observed at the posterior end of the PCV (arrowheads), largely outside a small clone. Note also that clones overlap over a substantial portion of a nearby section of the fourth longitudinal vein. (B,BV). Vein development identified by anti-Dl staining. Overlapping clones are limited to the portion of the fourth longitudinal vein normally intersected by the PCV. The robust anti-Dl staining observed in the posterior of the PCV (see arrows) is strongly reduced in the anterior of the PCV (arrowheads).

the posterior end of the PCV. While some of this disruption lay within a small region of overlap, it also extended well outside the clones (Figs. 3A,AV, arrowheads, anti-DSRF). In this case, overlapping clones also covered a nearby region of the fourth longitudinal vein. In the other, the anterior of the PCV was disrupted despite the absence of overlapping clones in that region (Figs. 3B,BV, arrowheads; anti-Dl). In this case, overlapping clones covered the region of the fourth longitudinal vein normally intersected by the PCV. Both these results suggest that the Dpp produced in the longitudinal veins contributes to signaling in the early PCV. Unfortunately, in several hundred wings examined, these were the only cases we found with large regions of overlap on the longitudinal veins adjacent to the PCV. We therefore used a different strategy to remove dpp, expressing doublestranded dpp RNA using UAS-dppIR (Inaki et al., 2002). To avoid the possibility of interfering with the early phase of dpp expression on the anterior side of the A/P compartment boundary, we used the posteriorly expressed engrailed (en)gal4 driver. Because en expression extends a short distance into the anterior compartment at pupal stages (Blair, 1992), en-gal4 is also expressed in most of the cells of the ACV (assessed using UAS-GFP; Figs. 4A,AV). When examined at 19–21 h AP, pMad was lost from the PCV and the posterior half of the ACV and severely reduced in the distal tips of the fourth and fifth longitudinal veins (16/16 wings, Fig. 4B). The vein-specific suppression of DSRF expression was also lost from the affected regions of the crossveins and the distal ends of the fourth and fifth longitudinal veins (compare Figs 4AW,4BV). Thus, BMP signaling in the early PCV depends on Dpp at a stage when dpp is expressed at high levels only in the longitudinal veins. As shown above, Dpp does eventually become expressed in the PCV. Intriguingly, loss of BMP signaling during the early stages of PCV formation blocks this later expression. Mutations in the secreted BMP modulator cv-2 remove detectable Mad phosphorylation during the initial stages of PCV formation while leaving signaling in the adjacent longitudinal veins intact (Conley et al., 2000). Unlike wildtype wings, cv-2 wings lack dpp expression in the PCV at 29 h AP (data not shown). Thus, BMP signaling within the PCV reinforces itself via a positive feedback loop by inducing localized expression of dpp. The generally expressed Gbb also contributes to BMP signaling during the initial stages of crossvein development. Rare escapers of the gbb 1 /gbb 4 mutant genotype appear developmentally delayed, as assessed by morphology and marker expression. We therefore examined wings at early and later stages of development. pMad was absent from the PCV and reduced in the ACV at 21 h AP; pMad was also lost from the distal tips of the fourth and fifth longitudinal veins (3/3 wings; Fig. 4C). A similar phenotype was observed at 28 h AP (2/2 wings; not shown), indicating that the 21 h AP phenotype was not simply an artifact of developmental delay. gbb is therefore

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

193

Fig. 4. gbb and dpp are required during early crossvein development. (A) en-gal4/UAS-GFP.nls (phenotypically wild type) wing at 19 h AP. pMad is visible at high levels in the crossveins and in the distal tips of the longitudinal veins. Lower levels of pMad are also visible in the longitudinal veins. (AV) Overlay of pMad (red) and nuclear GFP expression (green) reveals that GFP is expressed in all cells of the posterior compartment, as well as in the entire ACV. (AVV) AntiDSRF staining; note the slight decrease in the PCV. (B) en-gal4 UAS-dppIR wing at 19 h AP. pMad is not visible in the PCV or distal tips of L4 or L5 and is partially reduced in the ACV in 19 h AP following expression of dpp-interfering RNA. (BV) DSRF expression in same wing shown in panel B reveals truncation of distal longitudinal veins 4 and 5 and loss of PCV. (C) In gbb 1 /gbb 4 mutants, pMad is not detectable in the PCV or in the distal tips of longitudinal veins 4 and 5 at 21 h AP and is severely reduced in the ACV.

required for the initial stages of BMP signaling within the early presumptive PCV and contributes to signaling in the ACV. This is consistent with previously observed adult phenotypes, which show that gbb is required for the crossveins and the distal portions of the fourth and fifth longitudinal veins (Haerry et al., 1998; Khalsa et al., 1998; Ray and Wharton, 2001; Wharton et al., 1999; Yu et al., 2000). Longitudinal veins are required for the initial stages of PCV development Our results above suggest that the Dpp produced in the longitudinal veins is required for the formation of the PCV. Since dpp expression in the longitudinal veins depends on Egfr signaling (de Celis, 1997), we predicted that loss of Egfr signaling within adjacent longitudinal veins would disrupt early BMP signaling in the PCV. Homozygous Egfr f2 clones could disrupt the development of the longitudinal veins at pupal stages, as assessed by expression of the intervein marker DSRF (Figs. 5A,AV). Interestingly, one 25 h AP wing (the equivalent of 21 h AP in non-Minute flies) contained a large Egfr clone that encompassed a portion of the PCV and a large portion of the adjacent fifth longitudinal vein. This disrupted not only the longitudinal vein but also the posterior half of the PCV (Figs. 5A,AV). Some residual PCV development was observed just inside the clone, nearest to the remaining longitudinal veins. This contrasts strongly with our

previous results at this developmental stage, examining clones that encompassed only the PCV (Figs. 1C,CV). Thus, although Egfr signaling is not required cell-autonomously during early PCV development, Egfr signaling is required in the adjacent longitudinal vein for proper PCV patterning. Another way to investigate the role of the longitudinal veins during early crossvein patterning is to use mutants that lack longitudinal veins. vein (vn) encodes a neuregulin-like Egfr ligand (Schnepp et al., 1996), and rhomboid (rho) an intramembrane protein required for the proteolytic activation of other EGF-like ligands (Bang and Kintner, 2000; Lee et al., 2001; Urban et al., 2001, 2002; Wasserman et al., 2000). The rho ve mutation and the rho ve vn 1 mutant combination disrupt the initial development of most of the veins (Guichard et al., 1999; MartinBlanco et al., 1999; Sturtevant et al., 1993). Consistent with the adult phenotype (Fig. 5B), rho ve vn 1 wings examined at 19 (not shown), 21 or 24 h AP lacked most of the longitudinal veins as assessed using DSRF (Figs. 5CV– DV). Only the most proximal vein tissue was intact 19 and 21 h AP (Fig. 5CV). In wild-type wings, DSRF is downregulated in the crossveins by 21 h AP. In rho ve vn 1 wings, downregulation could not be detected at 19, 21, 24 or 26 h AP (Fig. 5CV– DV). Similarly, pMad was absent from the presumptive PCV at 19, 21 and 24 h AP (5/5 wings, 3/3 wings and 7/7 wings, respectively) (Figs. 5C–D). pMad was also severely reduced in the presumptive ACV (5/5 wings, 2/3 wings

194

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

Tkv is downregulated in the PCV in response to BMP signaling Our results indicate that long-range Dpp signaling from the longitudinal veins is critical for PCV development. In the wing imaginal disc, Dpp acts over long range, but its range is thought to be limited by binding its receptor Tkv (Lecuit and Cohen, 1998). Tkv likely plays a similar role in the pupal wing, reducing the range of Dpp signaling from the longitudinal veins. At 24 h AP, tkv is expressed at higher levels in the cells immediately flanking the veins and reducing Tkv levels results in abnormally thick veins (de Celis, 1997). However, we found that at 19 h AP the putative Tkv bbarrierQ between the longitudinal veins and intervein regions is reduced in the region of the PCV. At this stage, Tkv protein was high in intervein regions but lowered in a broad zone defining the PCV (Figs. 6A–AW). However, it is unlikely that the downregulation of Tkv is responsible for the initiation of BMP signaling in the PCV. At 17 h AP, the expression of Tkv protein is unpatterned (Fig. 7B), so the Tkv expression pattern does not prefigure the PCV. Moreover, the reduction of Tkv expression in the PCV requires heightened BMP signaling. The cv-2 mutation blocks anti-pMad staining in the developing PCV (Conley et al., 2000). Whereas tkv message was repressed in the PCV of wild-type wings (Fig. 5C), mutation of cv-2 blocked the downregulation of tkv message in the PCV at 19 and 21 h AP (Fig. 6D and not shown). Thus, the downregulation of tkv is a result of previously localized BMP signaling. Nonetheless, this downregulation may provide a positive feedback loop that helps reinforce BMP signaling by allowing Dpp diffusion from the longitudinal veins and BMP signaling within the PCV region.

Fig. 5. Early PCV development requires Egfr signaling in the adjacent longitudinal veins. (A) Egfr f2 homozygous clone (marked by absence of Myc) straddling a portion of L5 and the adjacent PCV. (AV) Anti-DSRF staining in the clone (red outline) is derepressed in the longitudinal vein and portions of the PCV. DSRF is weakly repressed in part of the PCV nearest to wild-type longitudinal vein tissue, likely due to diffusion of Dpp from longitudinal vein 4. (B) Adult rho ve vn 1 wings lack most veins, although faint venation in the proximal wing is sometimes detectable. (C) pMad in 21 h AP rho ve vn 1 pupal wings is not detectable in the regions of the presumptive crossveins (arrowheads) but is detectable in cells along the anterior margin and at low levels in proximal vein structures. (CV) DSRF expression in the wing shown in panel C reveals that the longitudinal veins are more severely truncated than at the earlier stage. DSRF is not downregulated in the crossveins. (D) In 24 h AP wings, pMad is not detectable in the presumptive crossveins but is visible in proximal veins and part of L2, consistent with the timing of Mad phosphorylation in wild-type longitudinal veins. (DV) DSRF is complementary to pMad in same wing as shown in panel D but is not downregulated in the crossveins.

and 5/7 wings, respectively). Since Egfr signaling is not cell-autonomously required in the early PCV, the rho ve vn 1 mutations must indirectly disrupt PCV development through their effects on longitudinal vein development.

Sog is downregulated in the PCV in a manner independent of BMP signaling Sog, a secreted BMP signaling antagonist, is another factor that could impede long-range signaling by Dpp during crossvein development. sog is expressed at higher levels in intervein cells from 18–20 h AP but is expressed at much lower levels in a broad region defining the presumptive PCV (Fig. 6E) (Yu et al., 1996), suggesting that Sog does not interfere with BMP signaling from the longitudinal veins during normal PCV development. Global misexpression of sog results in crossveinless adult wings (Yu et al., 1996, 2000). However, it is unclear exactly what stage of crossvein development is affected in these wings. We therefore exposed flies bearing hs-gal4 and UAS-sog transgenes to a 1 h heat shock pulse at different times during pupal development and examined the resultant adult wings for crossvein defects. Although the ACV was never disrupted with this regimen, the PCV was disrupted in flies that had received a 1-h heat shock between 16 and 21 h AP (Table 1). No defects were observed in flies that were heat shocked before or after this developmental window. This

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

195

Fig. 6. tkv and sog are downregulated in the nascent crossvein. (A) pMad in the crossveins at 19 h AP in wild-type wing. (AV) In the same wing shown in panel A, Tkv is expressed in intervein cells and is downregulated in a broad region surrounding the crossveins. (AW) Merge of panels A and AV. (B) pMad in the ACV (arrowhead) but not the PCV at 17 h AP in wild-type wing. (BV) In the same wing shown in panel B, Tkv is expressed at low levels throughout the wing and does not delineate the PCV. Tkv may be upregulated near the presumptive ACV. (C) tkv1 transcript is expressed at lower levels in the PCV (arrowheads) of wild-type pupal wings at 21 h AP. (D) tkv1 transcript is not downregulated in the PCV (arrowheads) in cv-2 1 wings at 21 h AP. (E) sog transcript is downregulated in the PCV of wild-type wings at 19 h AP. (F) sog transcript is downregulated in the presumptive PCV of cv-2 1 pupal wings at 19 h AP. (G) Overexpression of sog using en-gal4 blocks pMad in the PCV at 20 h AP. (H) Overexpression of cv-2 using en-gal4 and EP(2)1103 does not expand pMad at 21 h AP. (I) Co-overexpression of sog and cv-2 using en-gal4 rescues pMad in the PCV at 19 h AP.

phenotype was most penetrant in flies heat shocked at 17– 18 h AP. Any delay between the heat shock and the production of effective levels of Sog is likely to be short. In hs-gal4 UAS-GFP wing discs, GFP was visible 1 h after an identical heat shock (W.E. Jones and S.S. Blair, unpublished). Similarly, overexpression of sog using en-gal4 eliminated pMad within the crossveins at 20 h AP (Fig. 6G).

These results underscore the importance of a restricted sog expression pattern during early crossvein development. Interestingly, we found that downregulation of sog within the PCV, while critical for the development of the PCV, does not itself require BMP signaling, since that downregulation was still apparent in cv-2 mutant pupal wings (Fig. 6F). This indicates that the region that will form

196

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200 Table 1 Percentage of flies with a PCV defect after transient overexpression of sog Age during heat shock

Genotype hs-gal4  UAS-sog

hs-gal4  y w

16–17 17–18 18–19 19–20 20–21 21–22 22–26

13%, 81%, 65%, 54%, 29%, 0%, 0%,

0%, 0%, 0%, 0%, 0%, 0%, 0%,

h h h h h h h

AP AP AP AP AP AP AP

n n n n n n n

= = = = = = =

15 16 20 28 21 11 26

n n n n n n n

= = = = = = =

6 27 22 23 29 13 42

White prepupae from hs-gal4  UAS-sog or hs-gal4  y w crosses were picked and heat shocked 1 h at 378C at indicated times AP. Pupae were then reared at 258C and adult wings scored. Loss of at least 1/3 of the PCV was scored as a disrupted PCV.

the PCV region is specified by positional information that is independent of BMP signaling. Nonetheless, we also found conditions where crossveins can form in the presence of overexpressed Sog. While posterior expression of sog normally causes loss of crossveins in adults and loss of crossvein pMad in pupal wings, both these phenotypes can be rescued by simultaneously expressing the BMP signaling agonist encoded by cv-2 (Fig. 6I). cv-2 is normally expressed at higher levels in the developing crossveins (Conley et al., 2000), and posterior misexpression of cv-2 alone has little effect on crossvein formation (Conley et al., 2000) or pMad (Fig 6H). It is not clear by what mechanism co-overexpression of sog and cv-2 leads to rescue of crossvein development. However, these results indicate that the expression patterns of neither cv-2 nor sog are required for the normal positioning of the nascent crossveins. Therefore, under normal circumstances, a restricted pattern of sog expression is critical for allowing the initiation of BMP signaling in the PCV, but other factors must contribute to positioning the PCV.

Discussion

Fig. 7. Model of crossvein development in wild-type and rho ve wings. (A) In wild-type wings at or before 19 h AP, Dpp is produced by Egfr signaling in the longitudinal veins and signals to a zone that is uniquely responsive, called the crossvein competent zone. The initiation phase of crossvein development is characterized by phosphorylation of Mad within crossvein territories. Expression of Tkv, Sog and Cv-2 during this stage helps localize or stabilize BMP signaling in the crossveins. This is followed by a crossvein maintenance phase beginning at 24–28 h AP, when Egfr signaling becomes active within the crossvein, activating MAPK* and inducing the expression of rho and dpp. (B) Model of how the PCV becomes abnormally bent (arrowhead) in a rho ve mutant adult wing. Truncation of longitudinal vein 5 alters the positioning of the Dpp source during the initiation phase of crossvein development, skewing the region of Mad activation. Subsequent stages of crossvein development reinforce the slanted morphology of the crossvein.

The crossveins are especially sensitive to reductions in BMP signaling. While this makes the crossveins a powerful system for the isolation and analysis of members of the BMP pathway, prior to our studies, the basis of this sensitivity was poorly understood. Our work provides two reasons for this sensitivity. First, the initial development of crossveins requires BMP signaling, but not Egfr signaling. Second, the initial BMP signaling in the PCV depends on long-range signaling by the Dpp produced outside the PCV region. BMP signaling is required for the early stages of crossvein development Our results show that Egfr signaling is not required for the initial stages of PCV development or BMP signaling. Removal of BMPs, on the other hand, blocks not only BMP signaling in the PCV but also the initial stages of

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

crossvein development. This is consistent with our earlier findings that markers of BMP signaling appear prior to markers of Egfr signaling in the crossveins (Conley et al., 2000). In contrast, the longitudinal veins are initiated by Egfr signaling and apparently only require BMP signaling during a late maintenance phase. This difference is likely to produce greater sensitivity of the PCV to reductions in BMP signaling. There is precedent for BMP signaling initiating vein formation, at least in an experimental setting. Ectopically expressed Dpp is sufficient to activate ectopic vein formation and the expression of rho (de Celis, 1997; Yu et al., 1996), which is both a target and a stimulator of Egfr signaling (Sturtevant et al., 1993). Dpp can also apparently induce veins in rho ve vn 1 wings that lack the ability to upregulate rho (de Celis, 1997), and thus Dpp signaling may be sufficient to cause vein formation in the absence of heightened Egfr signaling. After this initial BMP-dependent period, the development of the PCV appears similar to that of the longitudinal veins. BMP signaling in the crossveins leads to the activation of Egfr signaling (Conley et al., 2000) and that signaling is required for continued crossvein development (this study). Moreover, dpp is eventually expressed within the PCV, and this expression requires early BMP signaling. While we have not explicitly tested the links between Egfr signaling and dpp expression in the crossveins, the earliest marker of Egfr signaling (MAPK activation; Conley et al., 2000) appears slightly before detectable dpp message in the PCV. Thus, it is likely that here, as in the longitudinal veins (de Celis, 1997), Egfr activation induces Dpp expression. Both the activation of Egfr signaling and the BMP signaling mediated by the eventual expression of dpp in the crossveins provide a positive feedback loop that helps maintain the PCV fate (Fig. 7A). Long-range Dpp signaling from outside the PCV region Our results show that Dpp is required for the initial activation of BMP signaling in the crossveins. However, we also showed that clonal removal of dpp from the early PCV does not reliably disrupt the initial stages of PCV formation. This latter result is consistent with our in situ data: during the initial stages of crossvein development, dpp expression is only detected in the longitudinal veins and a stripe of cells that intersects the ACV, but it is not in the PCV. Thus, the Dpp must be moving into the PCV region from nearby cells. The most likely source is the longitudinal veins; this is consistent not only with the expression patterns, but also with the dpp clones shown in Fig. 3, and with loss of the PCV observed after removal of the longitudinal veins. If so, the Dpp must travel over a surprisingly long range from the longitudinal veins to the PCV. Dpp can signal over the breadth of the late third instar wing disc (e.g. Tanimoto et al., 2000), but in the pupal wing, the range of Dpp signaling from the longitudinal veins is usually quite limited (Ray and

197

Wharton, 2001). Anti-pMad staining in the longitudinal veins is only slightly broader than the domain of Dpp expression (Conley et al., 2000) and in pupal wings ectopic Dpp only triggers pMad 1–3 cell diameters from the misexpressing cells (A. Ralston, unpublished). The center of the PCV is approximately 6–7 cell diameters from the dpp-expressing longitudinal veins. Either the PCV is much more sensitive to low levels of Dpp or the restrictions on Dpp transport or diffusion are reduced in this region. The long range of this Dpp signaling provides another explanation for the high sensitivity of the PCV to changes in Dpp signaling. Any manipulation that reduces the amount of available ligand or its movement into the PCV would likely disrupt signaling. In contrast, signaling in the longitudinal veins should be more robust, since they are themselves a source of Dpp, while the ACV lies nearer to the longitudinal veins and intersects an intervein stripe of dpp expression. Disruption of this long-range Dpp signaling thus provides a mechanism for generating crossveinless phenotypes. This is intriguing, as it may help explain some conflicting data on cv-2. cv-2 encodes a secreted molecule that contains potential BMP-binding domains (Conley et al., 2000). Its vertebrate homologs can bind BMPs and, in various vertebrate assays, misexpression of Cv-2-like molecules inhibits BMP signaling, presumably by sequestering ligand from the receptor (Binnerts et al., 2004; Coffinier et al., 2002; Moser et al., 2003; A. Ralston, unpublished). However, during crossvein development, Cv-2 acts as a signaling agonist: it is expressed in the crossveins, is required for the initial stages of crossvein BMP signaling (Conley et al., 2000), and can counteract the effects of Sog misexpression (this study). One possible explanation is that Cv-2 aids in the long-range diffusion or transport of Dpp from the longitudinal veins into the PCV by forming a complex that protects Dpp from other Dpp-binding proteins. In this view, Dpp is freed from the complex by some other factor in the crossveins. This model is similar to one proposed to account for Sog’s ability to both inhibit and strengthen BMP signaling in the early Drosophila embryo. Prior to gastrulation, dpp is expressed in a broad dorsal region of the embryo, yet BMP signaling is preferentially activated in a smaller group of cells at the extreme dorsal side (Eldar et al., 2002; Ross et al., 2001). sog is expressed in ventrolateral cells; its loss causes not only gains of BMP signaling in dorso-lateral cells, but also loss of the peak signaling in extreme dorsal cells (Decotto and Ferguson, 2001; Eldar et al., 2002; Ferguson and Anderson, 1992; Marquez et al., 1997; Ross et al., 2001). Similarly, misexpressed Sog acts as an antagonist at short range but as an agonist at long range (Ashe and Levine, 1999). Loss of tsg also results in a similar combination of gains and losses in signaling, and Tsg can form a complex with Sog and Dpp (Eldar et al., 2002; Mason et al., 1994; Ross et al., 2001). It has therefore been suggested that the complex containing Sog, Tsg and Dpp can diffuse for longer distances than unbound Dpp. In the dorsal-most cells, Tld is

198

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

thought to cleave Sog and free Dpp from the complex for signaling (Marques et al., 1997; Shimmi et al., 2003). This type of model has the advantage that it does not require any special properties of Sog or Cv-2 beyond their ability to bind BMPs. Moreover, the model may explain why the agonist actions are only evident in certain situations, where longrange signaling is a factor and where other molecules can free BMPs from Sog or Cv-2. Gbb is also required for the initial stages of crossvein development. Gbb is ubiquitously expressed at this stage (Conley et al., 2000) and thus may simply raise the levels of signaling throughout the wing, as it is thought to do at earlier stages of disc development (Khalsa et al., 1998; Ray and Wharton, 2001; Wharton et al., 1999). If so, then those regions of the wing that are especially sensitive to reductions in Gbb, such as the crossveins, may simply be those most critically dependent on high levels of signaling. However, clones lacking gbb only disrupt the adult PCV if they encompass adjacent longitudinal vein material (Ray and Wharton, 2001). This observation raises the intriguing possibility that the Dpp and Gbb produced in the longitudinal veins act in concert, possibly via the formation of heterodimers. Shaping the PCV competent zone An intriguing feature of PCV development is the lack of any congruence between the patterns of ligand expression and the initial sites of BMP signaling. Despite its presence in all the longitudinal veins, Dpp only triggers intervein BMP signaling in a zone that is uniquely competent to respond to BMP signaling (Fig. 7A). This is similar to, although more extreme than, the situation in the early Drosophila embryo. One way of locally increasing signaling into the PCV would be to increase diffusion of Dpp from the longitudinal veins. During wing disc development, Dpp diffusion is thought to be limited by binding to Tkv (Lecuit and Cohen, 1998). We have observed reduced tkv expression in crossvein regions; this could allow increased diffusion of Dpp while still providing enough receptor for signaling. Hypomorphic tkv alleles cause expansion of the veins, especially surrounding the PCV, possibly due to excessive diffusion of Dpp from the longitudinal veins (de Celis, 1997). Overexpression of tkv throughout the posterior compartment can disrupt the formation of L5 and the PCV (Haerry et al., 1998), likely because of reductions in Dpp diffusion during larval and pupal stages. However, we think it is unlikely that the downregulation of tkv in the PCV accounts for the initial pattern of BMP signaling, as tkv downregulation depends on high levels of BMP signaling. Thus, while tkv downregulation might help reinforce the initial pattern of BMP activation, other factors must initiate and localize BMP signaling in the crossvein competent zone. Another way of increasing signaling would be to reduce the expression of BMP inhibitors. sog expression is reduced

in the crossvein regions, and this reduction is independent of BMP signaling. Disruption of this expression pattern by uniform overexpression of sog blocked the initial stages of crossvein development. Reduced Sog expression is not, however, single-handedly responsible for defining the crossvein competent zone, as normally positioned crossveins can form after unpatterned misexpression of both sog and cv-2. There must therefore be other factors that can restrict the region of BMP activation under these conditions. Our finding that the repression of sog in the early PCV is independent of BMP signaling certainly shows that the PCV region is defined by some earlier patterning system. There are a number of candidates for such factors. cv, which encodes a Tsg-like molecule (Ross et al., 2001; Vilmos et al., 2001; Shimmi et al., unpublished) and the Tolloid-related protease (also known as Tolkin) (Finelli et al., 1995; Nguyen et al., 1994) are both required for crossvein development. However, neither is expressed at higher levels in the crossveins (Serpe et al., unpublished; Shimmi et al., unpublished). Reduction of cdc42 induces a small number of ectopic crossveins (Baron et al., 2000; Genova et al., 2000), and it has therefore been suggested that some Cdc42regulated process provides spatial cues for the crossveins, perhaps via regulation of JNK (Marcus, 2001). However, there is no evidence as yet that Cdc42 activity is spatially regulated during vein formation, and puckered-lacZ, a marker of JNK activity (Martin-Blanco et al., 1998), was undetectable in 18–22 h AP wings (Blair, unpublished). Nonetheless, there are additional candidates, including three crossveinless mutants that have not been molecularly characterized (crossveinless c, crossveinless d and detached) and two other crossveinless-like mutations that were isolated by Mohler (1965) but subsequently lost. Coordination of longitudinal and crossvein development We have shown that the longitudinal veins are required for the initial stages of BMP signaling in the PCV. Thus, the longitudinal veins serve as a source of a BMP-activating signal during early PCV development. Most likely, this is Dpp, as dpp expression in the veins depends on Egfr signaling (de Celis, 1997). These results may explain an observation first made by Waddington (1940). If the distal end of either the fourth or fifth longitudinal vein is truncated just past its normal intersection with the PCV, such as occurs in rho ve or abrupt mutants, the PCV takes an abnormal course to join the remaining portion of the longitudinal vein (Fig. 7B). In this situation, the longitudinal vein still provides an instructive cue for the PCV. The range of this cue is limited, however, and the PCV is usually lost if too much of the adjacent longitudinal veins are removed, as we saw with a large Egfr clone on L5 (Figs. 5A,AV). There are a few published adult wings in which the PCV extends to a region lacking longitudinal vein; these exceptional cases are likely caused

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

by the loss of the longitudinal vein after the time of PCV specification. Dpp signaling provides a mechanism by which longitudinal and crossvein development can be spatially coordinated.

Acknowledgments The authors thank members of the Blair lab for insightful conversations, Dr. James Pawley for use of the confocal microscope, Drs. Ryu Ueda, Stephen Cohen, Osamu Shimmi, and Michael B. O’Connor and other members of the fly community for reagents, and Bill Feeny for help with Fig. 3. This work was supported by grant IBN0077912 from NSF and grant RO1 NS028202 from NIH.

References Affolter, M., Montagne, J., Walldorf, U., Groppe, J., Kloter, U., de la Rosa, M.E., Gehring, W.J., 1994. The Drosophila SRF homolog is expressed in a subset of tracheal cells and maps within a genomic region required for tracheal development. Development 120, 743 – 753. Ashe, H.L., Levine, M., 1999. Local inhibition and long-range enhancement of Dpp signal transduction by Sog. Nature 398, 427 – 431. Bang, A.G., Kintner, C., 2000. Rhomboid and Star facilitate presentation and processing of the Drosophila TGF-a homolog Spitz. Genes Dev. 14, 177 – 186. Baron, M., O’Leary, V., Evans, D.A.P., Hicks, M., Hudson, K., 2000. Multiple roles of the Dcdc42 GTPase during wing development in Drosophila melanogaster. Mol. Gen. Genet. 264, 98 – 104. Bier, E., 2000. Drawing lines in the Drosophila wing: initiation of wing vein development. Curr. Opin. Genet. Devel. 10, 393 – 398. Binnerts, M.E., Wen, X., Cante´-Barrett, K., Bright, J., Chen, H.-T., Asundi, V., Sattari, P., Tang, T., Boyle, B., Funk, W., Rupp, F., 2004. Human Crossveinless-2, a novel inhibitor of Bone Morphogenetic Proteins. Biochem. Biophys. Res. Commun. 315, 272 – 280. Blair, S.S., 1992. Engrailed expression in the anterior lineage compartment of the developing wing blade of Drosophila. Development 115, 21 – 33. Blair, S.S., 2000. Imaginal Discs. In: Sullivan, W., Ashburner, M., Hawley, R.S. (Eds.), Drosophila Protocols. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp. 159 – 173. Blair, S.S., 2003. Mosaic techniques in Drosophila. Development 130, 5065 – 5072. Blair, S.S., Ralston, A., 1997. Smoothened-mediated Hedgehog signalling is required for the maintenance of the anterior–posterior lineage restriction in the developing wing of Drosophila. Development 124, 4053 – 4063. Brand, A.H., Perrimon, N., 1993. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401 – 415. Burke, R., Basler, K., 1996. Dpp receptors are autonomously required for cell proliferation in the entire developing Drosophila wing. Development 122, 2261 – 2269. Chang, C., Holtzman, D.A., Chau, S., Chickering, T., Woolf, E.A., Holmgren, L.M., Bodorova, J., Gearing, D.P., Holmes, W.E., Brivanlou, A.H., 2001. Twisted gastrulation can function as a BMP antagonist. Nature 410, 483 – 487. Clifford, R.J., Schupbach, T., 1989. Coordinately and differentially mutable activities of torpedo, the Drosophila melanogaster homolog of the vertebrate EGF receptor gene. Genetics 123, 771 – 787. Coffinier, C., Ketpura, N., Tran, U., Geissert, D., De Robertis, E.M., 2002. Mouse Crossveinless-2 is the vertebrate homolog of a Drosophila

199

extracellular regulator of BMP signaling. Gene Expression Pattern 2, 189 – 194. Conley, C.A., Silburn, R., Singer, M.A., Ralston, A., Rohwer-Nutter, D., Olson, D.J., Gelbart, W., Blair, SS., 2000. Crossveinless 2 contains cysteine-rich domains and is required for high levels of BMP-like signaling during the formation of the cross veins in Drosophila. Development 127, 3947 – 3959. de Celis, J.F., 1997. Expression and function of decapentaplegic and thick veins during the differentiation of the veins in the Drosophila wing. Development 124, 1007 – 1018. de Celis, J.F., 2003. Pattern formation in the Drosophila wing: the development of the veins. BioEssays 25, 443 – 451. de Celis, J.F., Diaz-Benjumea, F.J., 2003. Developmental basis for vein pattern variations in insect wings. Int. J. Dev. Biol. 47, 653 – 663. Decotto, E., Ferguson, E.L., 2001. A positive role for Short gastrulation in modulating BMP signaling during dorsoventral patterning in the Drosophila embryo. Development 128, 3831 – 3841. Diaz-Benjumea, F., Garcia-Bellido, A., 1990. Behaviour of cells mutant for an EGF receptor homologue of Drosophila in genetic mosaics. Proc. R. Soc. Lond., Ser. B 242, 36 – 44. Diaz-Benjumea, F.J., Hafen, E., 1994. The sevenless signalling casette mediates Drosophila EGF receptor function during epidermal development. Development 120, 569 – 578. Eldar, A., Dorfman, R., Weiss, D., Ashe, H., Shilo, B.Z., Barkai, N., 2002. Robustness of the BMP morphogen gradient in Drosophila embryonic patterning. Nature 419, 304 – 308. Ferguson, E.L., Anderson, K.V., 1992. Localized enhancement and repression of the activity of the TGF-h family member, decapentaplegic, is necessary for dorsal–ventral pattern formation in the Drosophila embryo. Development 114, 583 – 597. Finelli, A.L., Xie, T., Bossie, C.A., Blackman, R.K., Padgett, R.W., 1995. The tolkin gene is a tolloid BMP-1 homologue that is essential for Drosophila development. Genetics 141, 271 – 281. Gabay, L., Seger, R., Shilo, B.Z., 1997. In situ activation pattern of Drosophila EGF receptor pathway during development. Science 277, 1103 – 1106. Genova, J.L., Jong, S., Camp, J.T., Fehon, R.G., 2000. Functional analysis of cdc42 in actin filament assembly, epithelial morphogenesis, and cell signaling during Drosophila development. Dev. Biol. 221, 181 – 194. Guichard, A., Biehs, B., Sturtevant, M.A., Wickline, L., Chacko, J., Howard, K., Bier, E., 1999. Rhomboid and Star interact synergistically to promote EGFR/MAPK signaling during Drosophila wing vein development. Development 126, 2663 – 2676. Haerry, T.E., Khalsa, O., O’Connor, M.B., Wharton, K.A., 1998. Synergistic signaling by two BMP ligands through the SAX and TKV receptors controls wing growth and patterning in Drosophila. Development 125, 3977 – 3987. Huppert, S.S., Jacobsen, T.L., Muskavitch, M.L., 1997. Feedback regulation is central to Delta-Notch signalling required for Drosophila wing vein morphogenesis. Development 124, 3283 – 3291. Inaki, M., Kojima, T., Ueda, R., Saigo, K., 2002. Requirements of high levels of Hedgehog signaling activity for medial-region cell fate determination in Drosophila legs: identification of pxb, a putative Hedgehog signaling attenuator gene repressed along the anterior– posterior compartment boundary. Mech. Dev. 116, 3 – 18. Khalsa, O., Yoon, J.W., Torres-Schumann, S., Wharton, K.A., 1998. TGFh/BMP superfamily members, Gbb-60A and Dpp, cooperate to provide pattern information and establish cell identity in the Drosophila wing. Development 125, 2723 – 2734. Larrain, J., Bachiller, D., Lu, B., Agius, E., Piccolo, S., De Robertis, E.M., 2000. BMP-binding modules in chordin: a model for signalling regulation in the extracellular space. Development 127, 821 – 830. Larrain, J., Oelgeschla¨ger, M., Ketpura, N.I., Reversade, B., Zakin, L., De Robertis, E.M., 2001. Proteolytic cleavage of Chordin as a switch for the dual activities of Twisted gastrulation in BMP signaling. Development 128, 4439 – 4447. Lecuit, T., Cohen, S.M., 1998. Dpp receptor levels contribute to shaping the

200

A. Ralston, S.S. Blair / Developmental Biology 280 (2005) 187–200

Dpp morphogen gradient in the Drosophila wing imaginal disc. Development 125, 4901 – 4907. Lee, J.R., Urban, S., Garvey, C.F., Freeman, M., 2001. Regulated intracellular ligand transport and proteolysis control EGF signal activation in Drosophila. Cell 107, 161 – 171. Marcus, J.M., 2001. The development and evolution of crossveins in insect wings. J. Anat. 199, 126 – 211. Marques, G., Musacchio, M., Shimell, M.J., Wunnenberg-Stapleton, K., Cho, K.W., O’Connor, M.B., 1997. Production of a DPP activity gradient in the early Drosophila embryo through the opposing actions of the SOG and TLD proteins. Cell 91, 417 – 426. Martin-Blanco, E., Gampel, A., Ring, J., Virdee, K., Kirov, N., Tolkovsky, A.M., Martinez Arias, A., 1998. Puckered encodes a phosphatase that mediates a feedback loop regulating JNK activity during dorsal closure in Drosophila. Genes Dev. 12, 557 – 570. Martin-Blanco, E., Roch, F., Noll, E., Baonza, A., Duffy, J.B., Perrimon, N., 1999. A temporal switch in DER signaling controls the specification and differentiation of veins and interveins in the Drosophila wing. Development 126, 5739 – 5747. Mason, E.D., Konrad, K.D., Webb, C.D., Marsh, J.L., 1994. Dorsal midline fate in Drosophila embryos requires twisted gastrulation, a gene encoding a secreted protein related to human connective tissue growth factor. Genes Dev. 8, 1489 – 1501. Mohler, J.D., 1965. Preliminary genetic analysis of crossveinless-like strains of Drosophila melanogaster. Genetics 51, 641 – 651. Montagne, J., Groppe, J., Guillemin, K., Krasnow, M.A., Gehring, W.J., Affolter, M., 1996. The Drosophila Serum Response Factor gene is required for the formation of intervein tissue of the wing and is allelic to blistered. Development 122, 2589 – 2597. Moser, M., Binder, O., Wu, Y., Aitsebaomo, J., Ren, R., Bode, C., Bautch, V.L., Conlon, F.L., Patterson, C., 2003. BMPER, a novel endothelial cell precursor-derived protein, antagonizes Bone Morphogenetic Protein signaling and endothelial cell differentiation. Mol. Cell. Biol. 23, 5664 – 5679. Nguyen, T., Jamal, J., Shimell, M.J., Arora, K., O’Connor, M.B., 1994. Characterization of tolloid-related-1: a BMP-1-like product that is required during larval and pupal stages of Drosophila development. Dev. Biol. 166, 569 – 586. Nguyen, M., Park, S., Marques, G., Arora, K., 1998. Interpretation of a BMP activity gradient in Drosophila embryos depends on synergistic signaling by two type I receptors, SAX and TKV. Cell 95, 495 – 506. Oelgeschl7ger, M., Larrain, J., Geissert, D., De Robertis, E.M., 2000. The evolutionarily conserved BMP-binding protein Twisted gastrulation promotes BMP signalling. Nature 405, 757 – 763. Oelgeschla¨ger, M., Reversade, B., Larrain, J., Little, S., Mullins, M.C., De Robertis, E.M., 2003. The pro-BMP activity of Twisted gastrulation is independent of BMP binding. Development 130, 4047 – 4056. Persson, U., Izumi, H., Souchelnytskyi, S., Itoh, S., Grimsby, S., Engstrom, U., Heldin, C.-H., Funa, K., ten Dijke, P., 1998. The L45 loop in type I receptors for TGF-h family members is a critical determinant in specifying Smad isoform activation. FEBS Lett. 434, 83 – 87. Piccolo, S., Sasai, Y., Lu, B., De Robertis, E.M., 1996. Dorsoventral patterning in Xenopus: inhibition of ventral signals by direct binding of Chordin to BMP-4. Cell 86, 589 – 598. Posakony, L.G., Raftery, L.A., Gelbart, W.M., 1991. Wing formation in Drosophila melanogaster requires decapentaplegic gene function along the anterior–posterior compartment boundary. Mech. Dev. 33, 69 – 82. Ray, R.P., Wharton, K.A., 2001. Context-dependent relationships between the BMPs gbb and dpp during development of the Drosophila wing imaginal disk. Development 128, 3913 – 3925.

Roch, F., Baonza, A., Martin-Blanco, E., Garcia-Bellido, A., 1998. Genetic interactions and cell behaviour in blistered mutants during proliferation and differentiation of the Drosophila wing. Development 125, 1823 – 1832. Ross, J.R., Shimmi, O., Vilmos, P., Petryk, A., Kim, H., Gaudenz, K., Hermanson, S., Ekker, S.C., O’Connor, M.B., Marsh, J.L., 2001. Twisted gastrulation is a conserved extracellular BMP antagonist. Nature 410, 479 – 482. Schnepp, B., Grumbling, G., Donaldson, T., Simcox, A., 1996. Vein is a novel component in the Drosophila Epidermal growth factor receptor pathway with similarity to the neuregulins. Genes Dev. 10, 2302 – 2313. Scott, I.C., Blitz, I.L., Pappano, W.N., Maas, S.A., Cho, K.W., Greenspan, D.S., 2001. Homologues of Twisted gastrulation are extracellular cofactors in antagonism of BMP signalling. Nature 410, 475 – 478. Shimmi, O., O’Connor, M.B., 2003. Physical propertics of Tld, Sog, Tsg, and Dpp protein interactions are predicted to help create a sharp boundary in Bmp signals during dorsoventral patterning of the Drosophila embryo. Development 130, 4673 – 4682. Singer, M.A., Penton, A., Twombly, V., Hoffmann, F.M., Gelbart, W.M., 1997. Signaling through both type I DPP receptors is required for anterior–posterior patterning of the entire Drosophila wing. Development 124, 79 – 89. Sturtevant, M.A., Bier, E., 1995. Analysis of the genetic hierarchy guiding wing vein development in Drosophila. Development 121, 785 – 801. Sturtevant, M.A., Roark, M., Bier, E., 1993. The Drosophila rhomboid gene mediates the localized formation of wing veins and interacts genetically with components of the EGF-R signaling pathway. Genes Dev. 7, 961 – 973. Tanimoto, H., Itoh, S., ten Dijke, P., Tabata, T., 2000. Hedgehog creates a gradient of DPP activity in Drosophila wing imaginal discs. Mol. Cell 5, 59 – 71. Teleman, A.A., Cohen, S.M., 2000. Dpp gradient formation in the Drosophila wing imaginal disc. Cell 103, 971 – 980. Urban, S., Lee, J.R., Freeman, M., 2001. Drosophila rhomboid-1 defines a family of putative intramembrane serine proteases. Cell 107, 173 – 182. Urban, S., Lee, J.R., Freeman, M., 2002. A family of Rhomboid intramembrane proteases activates all Drosophila membrane-tethered EGF ligands. EMBO J. 21, 4277 – 4286. Vilmos, P., Gaudenz, K., Hegedus, Z., Marsh, J.L., 2001. The Twisted gastrulation family of proteins, together with the IGFBP and CCN families comprise the TIC superfamily of cysteine rich secreted factors. Mol. Pathol. 54, 317 – 323. Waddington, C.H., 1940. The genetic control of wing development in Drosophila. Drosophila. J. Genetics 41, 75 – 139. Wasserman, J.D., Urban, S., Freeman, M., 2000. A family of rhomboid-like genes: Drosophila rhomboid-1 and roughoid/rhomboid-3 cooperate to activate EGF receptor signaling. Genes Dev. 14, 1651 – 1663. Wharton, K.A., Cook, J.M., Torres-Schumann, S., de Castro, K., Borod, E., Phillips, D.A., 1999. Genetic analysis of the Bone Morphogenetic Protein-related gene, gbb, identifies multiple requirements during Drosophila development. Genetics 152, 629 – 640. Yu, K., Sturtevant, M.A., Biehs, B., Francois, V., Padgett, R.W., Blackman, R.K., Bier, E., 1996. The Drosophila decapentaplegic and short gastrulation genes function antagonistically during adult wing vein development. Development 122, 4033 – 4044. Yu, K., Srinivasan, S., Shimmi, O., Biehs, B., Rashka, K.E., Kimelman, D., O’Connor, M.B., Bier, E., 2000. Processing of the Drosophila Sog protein creates a novel BMP inhibitory activity. Development 127, 2143 – 2154.