The BMP-like signaling mediated by the ligands Dpp and Gbb is required to reinforce the development of most veins in the Drosophila wing. However, the formation of the cross veins is especially sensitive to reductions in BMP-like signaling. We show here that the formation of the definitive cross veins occurs after the initial specification of the longitudinal veins in a process that requires localized BMP-like activity. Since Dpp and Gbb levels are not detectably higher in the early phases of cross vein development, other factors apparently account for this localized activity. Our evidence suggests that the product of the crossveinless 2 gene is a novel member of the BMP-like signaling pathway required to potentiate Gbb of Dpp signaling in the cross veins. crossveinless 2 is expressed at higher levels in the developing cross veins and is necessary for local BMP-like activity. The Crossveinless 2 protein contains a putative signal or transmembrane sequence, and a partial Von Willebrand Factor D domain similar to those known to regulate the formation of intramolecular and intermolecular bonds. It also contains five cysteine-rich domains, similar to the cysteine-rich domains found in Chordin, Short Gastrulation and Procollagen that are known to bind BMP-like ligands. These features strongly suggest that Crossveinless 2 acts extracelluarly or in the secretory pathway to directly potentiate Dpp or Gbb signaling.

During development, the dorsal and ventral epithelia of the Drosophila wing form a precisely patterned array of veins (Fig. 1A). The specification and maintenance of vein cells requires the cooperation and interaction between several different signaling pathways. The roles played by two such pathways have been relatively well characterized: the MAPK signaling mediated by the Drosophila EGF Receptor (Egfr) stimulates vein formation, while Notch signaling inhibits and refines vein formation. Both these pathways are active from the earliest stages of vein formation, at mid-late third instar, and are required to maintain and refine vein fates until at least 30 hours after pupariation (a.p.) (Shellenbarger and Mohler, 1975; Diaz-Benjumea and Garcia-Bellido, 1990; Sturtevant and Bier, 1995; Martin-Blanco et al., 1999). The mechanisms that localize Egfr signaling to the veins are not completely understood, but involve regulating ligand expression, ligand activation and the sensitivity of cells to active ligand (Sturtevant et al., 1993; Simcox et al., 1996; Guichard et al., 1999; Martin-Blanco et al., 1999; Wessells et al., 1999; Vervoot et al., 1999; Mohler et al., 2000). The Notch ligand Delta is expressed along the veins, apparently in response to high Egfr signaling, induces Notch target gene expression and inhibits vein formation in neighboring cells (Sturtevant and Bier, 1995; de Celis et al., 1997; Huppert et al., 1997).

Fig. 1.

Adult wings. (A) Wild-type wing, showing the location of the ACV, PCV and the LVs (L0-L6); (B) cv-21; (C) cv-22253; (D) cv-21/Df(2R)Pu-D17. This is an extreme example; note loss of the entire ACV and the ends of some LVs. (E) cv-23511/Df(2R)Pu-D17. Note presence of the entire ACV. (F) dppd6cv-23511/Df(2R)Pu-D17. Note additional losses in the ACV (arrow) and ends of some LVs.

Fig. 1.

Adult wings. (A) Wild-type wing, showing the location of the ACV, PCV and the LVs (L0-L6); (B) cv-21; (C) cv-22253; (D) cv-21/Df(2R)Pu-D17. This is an extreme example; note loss of the entire ACV and the ends of some LVs. (E) cv-23511/Df(2R)Pu-D17. Note presence of the entire ACV. (F) dppd6cv-23511/Df(2R)Pu-D17. Note additional losses in the ACV (arrow) and ends of some LVs.

A third pathway, the BMP-like signaling mediated by the ligands Decapentaplegic (Dpp) and Glass Bottom Boat (Gbb, also known as 60A), also plays a role in vein specification. In wing discs, dpp is expressed just anterior to the anteroposterior compartment boundary, between L3 and L4, and plays a role setting up the axes of the wing. It has not yet been determined whether the reception of BMP-like signals initiates vein specification, although high BMP activity has been detected near the fourth longitudinal vein (LV) at late third instar (Tanimoto et al., 2000). However, at later pupal stages, dpp expression is lost from the compartment boundary and rises along the veins; this expression is required for maintaining the fate of most veins and ectopic signaling can induce ectopic veins (Posakony et al., 1990; Yu et al., 1996; de Celis, 1997).

Although Gbb expression is not higher in pupal veins (see below), it also helps maintain vein formation, either by raising general levels of signaling or by interacting with localized modulators of signaling (Doctor et al., 1992; Khalsa et al., 1998). Both Dpp and Gbb vein signals may be mediated largely by the type I receptor Thickveins (Tkv), rather than the alternate type I receptor Saxophone (Sax). Cells lacking Tkv do not form veins (Burke and Basler, 1996), but removal of Sax does not reliably remove veins (Singer et al., 1997).

However, not all veins are equally sensitive to reductions in Dpp and Gbb signaling. The hypomorphic gbb4 mutation shows complete loss of the cross veins (CVs), but only slight loss of the ends of the LVs (Khalsa et al., 1998). Sog encodes a Chordin-like molecule that inhibits BMP-like signaling; both Sog and Chordin are thought to bind to and sequester ligands, preventing the activation of receptors (Francois et al., 1994; Holley et al., 1995; Sasai et al., 1995; Schmidt et al., 1995; Piccolo et al., 1996; Larrain et al., 2000). Overexpressing Sog in the wing specifically blocks formation of the CVs and the ends of the LVs (Yu et al., 1996, 2000). The secreted Tolloid proteases, similar to vertebrate BMP1s, can increase BMP signaling by cleaving and inactivating Chordin or Sog (Shimell et al., 1991; Piccolo et al., 1997; Marques et al., 1997; Scott et al., 1999). Loss of tolkin (also known as tolloid-related) blocks formation of the CVs and the tips of the LVs (Nguyen et al., 1994; Finelli et al., 1995). Overexpressing a dominant negative form of Sax again induces a similar phenotype (Haerry et al., 1998).

Such phenotypes are very reminiscent of the crossveinless class of mutations in Drosophila (reviewed in Garcia-Bellido and de Celis, 1992). As we will show here, strong reductions in crossveinless 2 (cv-2) function remove the posterior CV (PCV), the anterior CV (ACV), and the ends of the LVs (Fig. 1B-E). However, despite the possibility that the crossveinless genes encode novel players in BMP-like signaling, none have been characterized and the sensitivity of CVs to BMP-like signaling has not been explained.

We will present evidence that cv-2 encodes a novel member of the BMP-like signaling pathway, expressed in and required for high levels of BMP-like signaling in the developing cross veins. The Cv-2 protein contains five cysteine-rich domains similar to those known to bind BMP-like ligands, strongly suggesting that Cv-2 directly modulates Dpp or Gbb activity.

Drosophila stocks

The original cv-21cn bw/CyO stock (kindly provided by A. Garcia-Bellido) was homozygous lethal, but outcrossing yielded a homozygous viable stock. All the analyses here used the viable stock, except that shown in Fig. 1C. cv-22253 was obtained from P[lacW] stocks kindly provided by A. Laughon and S. Carroll, while cv-23511 is the P[lacW] insertion l(2)k03511. gbb4 was kindly provided by K. Wharton. gbb4cv-23511 and dppd6cv-23511 chromosomes were created by recombination. The lacZ enhancer traps rhoAA69 (Nambu et al., 1990; kindly provided by S. Spencer), argosrJ472 (kindly provided by M. Freeman), wgen11 (Kassis et al., 1992), dpp10638 and tkvlacZ (Tanimoto et al., 2000; kindly provided by T. Tabata) were used to monitor gene expression. UAS-sog was kindly provided by E. Bier (Yu et al., 1996).

Immunohistology

Dissection and staining with rat anti-DSRF, Promega mouse anti-β-gal and Cappell rabbit anti-β-gal, and guinea pig anti-Dl were performed as previously described (Blair and Ralston, 1997; Micchelli et al., 1997; Blair 2000). Rabbit anti-p-Mad (kindly provided by T. Tabata and P. ten Dijke) was used at 1/2,000 overnight, and rabbit anti-Asense (kindly provided by A. Jarman; Brand et al., 1993) at 1/4,000. For anti-MAPK* staining, tissues were fixed overnight in PBS with 4% formaldehyde, dissected, stored in cold methanol for 2-4 hours and washed with PBS containing 0.1% Tween-20; wings were incubated overnight with 1/200 Sigma mouse anti-MAPK* in PBS-Tween.

In situ hybridization

Tissue was prepared as described by Cadigan et al. (1998), except that the initial fix of partly dissected pupae in PBS-formaldehyde was performed overnight, the wings were fully dissected prior to the next fixation and formaldehyde was not used in the final methanol-PBS-formaldehyde step. Probes were digoxigenin-labeled single-stranded DNA from either the 5′ and 3′ ends of the cv-2 open reading frame (ORF), PCR amplified from cDNA template with a tenfold excess of one primer.

Molecular analyses

Plasmid rescue and inverse PCR techniques used have been described elsewhere (Wilson et al., 1989; Rehm, http://www.fruitfly.org/EST/). cv-2 sequence and cDNAs were isolated from a λgt10 imaginal disc library (kindly provided by G. Rubin). Initial sequence was obtained after PCR amplification of the library using nested primers. The cDNAs were isolated using two probes. Template was PCR amplified from genomic DNA from a 200 bp region immediately upstream of the predicted ORF and from a 250 bp region lying 30 bp further downstream, within the ORF; these were gel purified, sequenced and used to generate digoxigenin probes. Each probe was hybridized against duplicate filters according to standard protocols (Sambrook et al., 1989).

The timing of cross vein formation

Morphologically, the LVs first become visible from 4-8 hours after pupariation (a.p.) (Waddingon, 1940; Mohler and Swedberg, 1964; Murray et al., 1984, 1995; Fristrom et al., 1993). The dorsal and ventral surfaces of the everting wing disc come together to form the wing blade, and LV ‘proveins’ form as broad gaps or lacunae between the dorsal and ventral epithelia. The pattern of LV proveins differs from the mature vein pattern: the veins are broader and somewhat incomplete, and proximally the L3 and L4 proveins fuse into a central, single provein. Molecular markers of veins, such as rhomboid-lacZ (rho-lacZ) (Nambu et al., 1990; Sturtevant et al., 1993), argos-lacZ (Sawamoto et al., 1994; Schweitzer et al., 1995), anti-Delta (anti-Dl) (Huppert et al., 1997), and an antibody against the activated form of MAPK (anti-MAPK*; Gabay et al., 1997), or of interveins, such as anti-Drosophila Serum Response Factor (DSRF, also known as Blistered) (Montagne et al., 1996), define the LV proveins as early as mid-late third instar. However, even with these markers, distinct L3 and L4 proveins were difficult to detect in the proximal region, where L3 and L4 fuse (Fig. 2A). The provein lacunae are lost from 8-16 hours a.p. as the wing inflates, separating the dorsal and ventral surfaces. Only as the dorsal and ventral surfaces of the wing reappose, at approximately 16-26 hours a.p., are the definitive veins visible. MAPK activity remains high in the definitive veins, but is lost from the veins and become high in the intervein regions after 30 hours a.p. (Gabay et al., 1997; Guichard et al., 1999; Martin-Blanco et al., 1999; S. S. B., unpublished observations).

Fig. 2.

Development of the CVs. (A) Anti-DSRF (red) and anti-Dl (green) staining in a rho-lacZ (blue) 5 hour a.p. wing. Note the ACV-like provein between L3 and L4; high rho-lacZ marks the cells of the ACV campaniform sensillum (arrow). Separate L3 and L4 veins are poorly defined proximal to the ACV-like provein. (B,C) Anti-DSRF (red) and anti-Dl (green) staining in 20 hour (B) and 24 hour (C) a.p. wings. (B) Anti-DSRF staining is very slightly reduced in the ACV and PCV, but anti-Dl staining is absent. (C) Note strong reductions of anti-DSRF and gains in anti-Dl staining in the CVs. (D) Anti-DSRF (red) and anti-MAPK* (green) staining in the veins of a 24 hour a.p. wing. (E) Anti-β-gal staining (green) in a 28 hour a.p. rho-lacZ wing; note the initiation of very faint staining in the CVs. (F) Anti-β-gal staining (green) in a 30 hour a.p. argos-lacZ wing; note the lack of staining in the CVs.

Fig. 2.

Development of the CVs. (A) Anti-DSRF (red) and anti-Dl (green) staining in a rho-lacZ (blue) 5 hour a.p. wing. Note the ACV-like provein between L3 and L4; high rho-lacZ marks the cells of the ACV campaniform sensillum (arrow). Separate L3 and L4 veins are poorly defined proximal to the ACV-like provein. (B,C) Anti-DSRF (red) and anti-Dl (green) staining in 20 hour (B) and 24 hour (C) a.p. wings. (B) Anti-DSRF staining is very slightly reduced in the ACV and PCV, but anti-Dl staining is absent. (C) Note strong reductions of anti-DSRF and gains in anti-Dl staining in the CVs. (D) Anti-DSRF (red) and anti-MAPK* (green) staining in the veins of a 24 hour a.p. wing. (E) Anti-β-gal staining (green) in a 28 hour a.p. rho-lacZ wing; note the initiation of very faint staining in the CVs. (F) Anti-β-gal staining (green) in a 30 hour a.p. argos-lacZ wing; note the lack of staining in the CVs.

The development of the ACV and PCV differ; a temporary, ACV-like provein is formed during the formation of the LV proveins, but no PCV provein is visible (Waddington, 1940). The L3 and L4 proveins join at the future site of the ACV, as shown by heightened rho-lacZ expression in the cells of the ACV campaniform sensillum (Fig. 2A). Anti-Dl and anti-DSRF detected a distinct, CV-like structure extending between L3 and L4 in this region, but no PCV-like gene expression could be detected at 5-6 hours a.p. (Fig. 2A). However, the ACV-like provein was no longer visible with anti-Dl at 16-20 hours a.p., even though the LVs can be identified at these stages (Fig. 2B).

The definitive ACV and PCV first became visible at 19-22 hours a.p. as broad regions with slightly reduced anti-DSRF staining (Fig. 2B,C); this reduction was only rarely observed at 19-20 hours a.p. The DSRF suppression begins during the morphological formation of the definitive CVs (Mohler and Swedberg, 1964). By 23-26 hours a.p. narrow, well-defined CVs were visible with anti-Dl, anti-DSRF (Fig. 2C) and anti-MAPK* (Fig. 2D; also see Martin-Blanco et al., 1999). Although, in our hands, the reduction of anti-DSRF in the CVs preceded detectable levels of anti-MAPK* staining, DSRF is suppressed in a cell-autonomous manner by Egfr signaling (Roch et al., 1998) so its reduction may correspond to increased Egfr activity. rho-lacZ and argos-lacZ did not appear in the CVs until 28-32 hours a.p. (Fig. 2E,F). wingless-lacZ was expressed in the developing CVs (Phillips and Whittle, 1993; Blair, 1994), beginning at 25-26 hours a.p. (not shown). wingless plays only a weak role in the formation of the CVs (C. A. C. and S. S. B., unpublished data).

Mad is activated in the developing cross veins

To follow the signaling mediated by Dpp and Gbb, we used an antiserum specific to the phosphorylated, activated form of Mothers Against Dpp (Mad), the receptor-activated Smad in Drosophila (anti-p-Mad; Persson et al., 1998; Tanimoto et al., 2000). We did not detect any patterned anti-p-Mad staining during early pupal stages (5-6 hours a.p.), when the temporary ACV-like provein is formed. However, at 19 and 21 hours a.p., we observed nuclear staining in broad regions at the future sites of the CVs and in the tips of the LVs; some weaker staining was also observed in the portion of the LVs near the CV attachment sites (including the portion of L4 between the ACV and PCV) and along L2 (Fig. 3A,B). Interestingly, the CV anti-p-Mad staining typically preceded the reduction of anti-DSRF staining in the CVs, suggesting that Mad activation precedes the suppression of DSRF expression mediated by Egfr. At 24 and 26 hours a.p. anti-p-Mad-stained nuclei along all the LVs, but staining was still stronger near the CVs (Fig. 3C). At approximately 36 hours a.p., staining was observed throughout the wing, but again was stronger in the CVs (Fig. 3D). Strong staining was also observed in both the nuclei and axons of the PNS of the wing at 18-36 hours a.p.

Fig. 3.

Anti-p-Mad staining in normal (A-D) and mutant (E-I) pupal wings. At all stages staining is higher in the axons and nuclei of the neurons of the anterior margin. (A) 19 hours a.p. Note the stronger nuclear staining in regions in and around the CVs and the end of L5, and lower staining in L4 between the ACV and the PCV. (B) 21 hours a.p. Composite of two images of the same wing, taken in slightly different focal planes. Note the strong staining in the CVs and in the ends of all the LVs (arrows). (C) 26 hours a.p. Strong staining is now present along all the LVs, but is still higher in the CVs. (D) 36 hours a.p. Staining is present throughout the wing, but is still higher in the CVs. (E-G) cv-21 homozygotes at 19 (E), 26 (F)and 36 (G) hours a.p. Staining is lost from normal site of the PCV (?) but, as expected from the adult phenotype, is still present in the ACV and the end of L5 (arrow). (H,I) cv-21/Df(2R)Pu-D17 wings at 25 (H) and 36 (I) hours a.p. Staining is lost from the normal site of the PCV at 25 hours a.p., but from the sites of both CVs at 36 hours a.p. (?).

Fig. 3.

Anti-p-Mad staining in normal (A-D) and mutant (E-I) pupal wings. At all stages staining is higher in the axons and nuclei of the neurons of the anterior margin. (A) 19 hours a.p. Note the stronger nuclear staining in regions in and around the CVs and the end of L5, and lower staining in L4 between the ACV and the PCV. (B) 21 hours a.p. Composite of two images of the same wing, taken in slightly different focal planes. Note the strong staining in the CVs and in the ends of all the LVs (arrows). (C) 26 hours a.p. Strong staining is now present along all the LVs, but is still higher in the CVs. (D) 36 hours a.p. Staining is present throughout the wing, but is still higher in the CVs. (E-G) cv-21 homozygotes at 19 (E), 26 (F)and 36 (G) hours a.p. Staining is lost from normal site of the PCV (?) but, as expected from the adult phenotype, is still present in the ACV and the end of L5 (arrow). (H,I) cv-21/Df(2R)Pu-D17 wings at 25 (H) and 36 (I) hours a.p. Staining is lost from the normal site of the PCV at 25 hours a.p., but from the sites of both CVs at 36 hours a.p. (?).

To confirm that Mad activation was induced by BMP-like signaling, we overexpressed Sog, which can inhibit BMP-like signaling in pupal wings (see Yu et al., 1996). en-Gal4 was used to drive UAS-sog in the posterior of the wing. Adult wings lacked the ACV and the PCV, but these levels of Sog were not sufficient to block formation of the ends of the LVs; as expected, heightened anti-p-Mad staining was not detected near the PCV in 19-20 hours a.p. pupal wings, while staining in the ACV was reduced and in the LVs was largely normal (not shown).

Localized ligand expression cannot apparently account for the early stages of Mad activation in the CVs. dpp is first expressed along the LVs at 18-20 hours a.p., but is not detected in the CVs until some time after 24 hours a.p., as shown by in situ hybridization (Yu et al., 1996; de Celis, 1997; Martin-Blanco et al., 1999) and dpplacZ enhancer traps (not shown). As in imaginal discs, anti-Gbb staining (Doctor et al., 1992; Khalsa et al., 1998) appeared largely uniform at 21 and 24 hours a.p., and was not higher in the CVs (Fig. 4A). The localized activation of Mad is apparently not due to localized receptor expression. Expression of the BMP receptor tkv is ubiquitous during pupal stages, but is lower in the veins and higher in the cells immediately flanking the veins; the lowered expression is thought to be mediated by Dpp signaling (de Celis, 1997). tkv-lacZ expression was slightly reduced in the CVs at 22.5 hours a.p. (Fig. 4B).

Fig. 4.

(A) Anti-Gbb staining in a 21 hour a.p. wing. Staining is slightly higher in the interveins, but not higher near the CVs. (B) Anti-βgal staining in a tkv-lacZ wing at 22.5 hours a.p. Staining is slightly reduced in the LVs and the PCV.

Fig. 4.

(A) Anti-Gbb staining in a 21 hour a.p. wing. Staining is slightly higher in the interveins, but not higher near the CVs. (B) Anti-βgal staining in a tkv-lacZ wing at 22.5 hours a.p. Staining is slightly reduced in the LVs and the PCV.

cv-2 alleles

cv-21 was isolated by Nicoletti (Lindsley and Zimm, 1992); it failed to complement Df(2)AA21, Df(2)Pu-D17, Df(2R)PI12, Df(2R)PK1, Df(2)PL3 and l(2)57DEFb-1 57, localizing it to chromosome bands 57D11-E1. Two P(lacW) enhancer trap element insertions were also isolated with recessive CV-less phenotypes. Both were mapped to the 57D region by in situ hybridization on polytene chromosomes and failed to complement cv-21. A majority of the excisions generated by remobilizing the P elements reverted the phenotypes in both lines, indicating that the P-element insertions were responsible for the mutations.

The cv-21 allele generated the strongest wing phenotype; wings typically lacked most or all of the PCV and, in some cases, the posterior of the ACV (Fig. 1B). cv-22253 flies typically lacked part or all of the PCV (Fig. 1C). cv-23511 overlapped wild type (not shown). Phenotypes of all three alleles were strengthened when placed over a deficiency; cv-21/Df(2)Pu-D17 always lacked the PCV, variably lacked some or all of the ACV and, on occasions, lacked the ends of the LVs (Fig. 1D). A small percentage of lethal P-element excisions were also generated, but none had appreciably stronger wing phenotypes than their parent P-element lines. No other defects were detected in adults or embryos in any of these lines.

cv-2 affects Dpp and Gbb signaling

As discussed above, dpp and gbb mutations both disrupt CV formation. We found that weak cv-2 alleles were strengthened by dpp and gbb loss-of-function mutations. cv-22253/cv-23511 flies never lacked the entire PCV, but 50% of gbb4cv-22253/cv-23511 flies lacked the entire PCV. Similarly, cv-23511/Df(2R)Pu-D17 only rarely disrupted the ACV, but dppd6cv-23511/Df(2R)Pu-D17 commonly did (Fig. 1E,F). However, cv-2 cannot dominantly enhance earlier dpp-dependent patterning in the wings: dppd5Df(2R)Pu-D17/dpphr4 wings looked no worse than dppd5/dpphr4 wings (not shown).

To provide a more direct link between cv-2 and Dpp and Gbb signaling, we examined Mad activation in mutant pupal wings. In cv-21 adults, the PCV was more reliably disrupted than the ACV; the anti-p-Mad staining normally found near the PCV in 19, 22, 26 and 36 hours a.p. wings was lost or disrupted in cv-21 homozygotes (Fig. 3E-G), as was the reduction of anti-DSRF in the PCV. In adults of the stronger allelic combination cv-21/Df(2R)Pu-D17, the ACV was also often lost along with the ends of some of the LVs. Interestingly, we did not detect disruption of the ACV or LV anti-p-Mad staining cv-21/Df(2R)Pu-D17 pupal wings at 21 or 25 hours a.p. (Fig. 3H); only at 36 hours a.p. was staining lost from the ACV (Fig. 3I). This indicates that cv-2 is required not only to initiate Mad activity in the PCV, but also to maintain that activity in the ACV.

Genomic region

We used plasmid rescue and inverse PCR to isolate genomic DNA flanking the two P-element insertions; these mapped within a fully sequenced 190 kb contig generated by the Berkeley Drosophila Genome Project from P1 DS01261 and BACR48K03. A map of the cv-2 region of this contig is shown in Fig. 5A. The two P elements lie 6.8 kb from one another, 12.2 kb (3511) and 5.4 kb (225-2) upstream of a novel transcript. This transcript is likely to encode cv-2; as shown below it is expressed in CVs and other tissues in a pattern similar to that seen in the enhancer traps, and its expression is lost or reduced in cv-2 mutants.

Fig. 5.

(A) Top: Genomic organization and P-element insertions near cv-2. Arrows show direction of transcription. Bottom: Exon-intron structure of cv-2. B. Sequence of cv-2, starting at the predicted TATA box, ending at the predicted poly(A) site. The putative hydrophobic signal sequence (yellow), cysteine-rich domains (CR, green), partial VWFd domain (red) and potential Furin cleavage site (box) are marked. Exon splice sites are underlined. Sequence shown is based on identical genomic sequences from the Berkeley Drosophila Genome Project (AC007175) and Adams et al. (2000) (AE003453) (see Genbank accession number AF284429). Sequence derived from the cDNA contained three amino acid differences, changing A432 to S, N482 to D and C572 to R (see Genbank accession number AF288223). (C,D) Alignments of Cv-2; cysteines are shaded green, while amino acids shaded black and grey are conserved or similar, respectively, in half or more of the sequences. (C) Comparison of Cv-2 CRs with Chordin and Procollagen CRs known to bind BMPs (aligned using MAP; Huang et al., 1994), and of Cv-2 CR5 with the canonical Pfam VWFc domain (aligned using Pfam, Bateman et al., 2000). Less well conserved amino acids in the canonical Pfam VWRc sequence are shown in lower case. (D) Comparison of the Cv-2 VWFd domain with the VWFd2 domain from human VWF (aligned using ClustalW; Thompson et al., 1994). Cv-2 matches only a portion of the VWFd2 domain (underlined red) and a similar region in other VWFd domains (not shown). (E) Schematic comparison of Cv-2 with Kielin, Sog, Chordin and CRIM1. Kielin has a region similar to the amino terminal domain of Thrombospondin (TSPN), while human and mouse (but not C. elegans) CRIM1s have a region similar to a insulin-like growth factor binding-protein domain (IGFBP).

Fig. 5.

(A) Top: Genomic organization and P-element insertions near cv-2. Arrows show direction of transcription. Bottom: Exon-intron structure of cv-2. B. Sequence of cv-2, starting at the predicted TATA box, ending at the predicted poly(A) site. The putative hydrophobic signal sequence (yellow), cysteine-rich domains (CR, green), partial VWFd domain (red) and potential Furin cleavage site (box) are marked. Exon splice sites are underlined. Sequence shown is based on identical genomic sequences from the Berkeley Drosophila Genome Project (AC007175) and Adams et al. (2000) (AE003453) (see Genbank accession number AF284429). Sequence derived from the cDNA contained three amino acid differences, changing A432 to S, N482 to D and C572 to R (see Genbank accession number AF288223). (C,D) Alignments of Cv-2; cysteines are shaded green, while amino acids shaded black and grey are conserved or similar, respectively, in half or more of the sequences. (C) Comparison of Cv-2 CRs with Chordin and Procollagen CRs known to bind BMPs (aligned using MAP; Huang et al., 1994), and of Cv-2 CR5 with the canonical Pfam VWFc domain (aligned using Pfam, Bateman et al., 2000). Less well conserved amino acids in the canonical Pfam VWRc sequence are shown in lower case. (D) Comparison of the Cv-2 VWFd domain with the VWFd2 domain from human VWF (aligned using ClustalW; Thompson et al., 1994). Cv-2 matches only a portion of the VWFd2 domain (underlined red) and a similar region in other VWFd domains (not shown). (E) Schematic comparison of Cv-2 with Kielin, Sog, Chordin and CRIM1. Kielin has a region similar to the amino terminal domain of Thrombospondin (TSPN), while human and mouse (but not C. elegans) CRIM1s have a region similar to a insulin-like growth factor binding-protein domain (IGFBP).

Other neighboring genes are unlikely to contribute to the cv-2 wing phenotype. 9-10 kb proximal to the putative cv-2 sequence is a gene highly similar to Drosophila antigen 5-related, encoding a protein thought to act as an allergen or antimicrobial agent (Kovalick et al., 1998). 3.2-6.7 kb distal to the 3511 insertion is a gene encoding a protein highly similar to GTP-binding mitochondrial and bacterial Eftu elongation factors, and a partially sequenced EST (LD27358) with limited similarity to GTP-binding proteins that is predicted by Fgenesh (Salamov and Solovyev, unpublished; see http://genomic.sanger.ac.uk/) to correspond to the 3′ end of the elongation factor. 0.2-6.1 kb distal to the elongation factor lie two orthologs of the peroxisomal acyl-CoA oxidases (Dmel\acox57Dp and Dmel\acox57Dd). Gene prediction programs identified no other likely coding regions in this region. Four

Drosophila lines contain EP insertions (Rorth, 1996) just distal to the cv-23511 insertion (EP(2)1103, EP(2)2395; EP(2)2634; EP(2)0796), and another contains the partially male-sterile ms(2)05235 (trail mix) P-element insertion (Castrillon et al., 1993). The EP lines are homozygous viable and have wild-type wings; EP(2)1103 and ms(2)05235 complement cv-2.

cv-2 message and protein

Fgene (Solayev et al., 1995; see http://genomic.sanger.ac.uk/) predicted that the putative cv-2 sequence contains six exons (Fig. 5A); a strong TATA box is found 550 bp upstream and a predicted poly(A) site 1.2 kb downstream of the coding sequence. The predicted splice sites and the translation start and stop sites were first confirmed by PCR sequencing directly from an amplified imaginal disc cDNA library. This library was also screened using probe from the 5′ end of the predicted sequence. A cDNA was isolated which, when sequenced, confirmed the predicted open reading frame.

The amino half of the putative Cv-2 protein (Fig. 5B) contains five adjacent cysteine-rich domains (CRs), similar to those found in a number of secreted and transmembrane proteins, including vertebrate Chordin, Sog and Procollagens (Sasai et al., 1994; Francois et al., 1994), the mammalian and C. elegans CRIM1 proteins (Kolle et al., 2000), and Xenopus Kielin (Matsui et al., 2000). The CRs in Chordin and Procollagen IIA bind BMPs and can mediate Chordin-like biological activity during axis formation in Xenopus embryos (Zhu et al., 1999; Larrain et al., 2000). The spacing of cysteines is especially well conserved in the CxxCxC and CCxxC motifs (Fig. 5C). The second Cv-2 CR is the most divergent in terms of cysteines, lacking two. The Chordin and Sog CRs also contain a conserved tryptophan (Francois and Bier, 1995) that is shared by the second and fifth Cv-2 CRs. The CRs are similar to Von Willebrand Factor Type C domains (VWFc), but most regions in these CRs are shorter than the equivalent regions in the canonical VWFc (e.g. Fig. 5D), and the first and second Cv-2 CRs only weakly match VWFc domains.

The portion of Cv-2 carboxy terminal to the CRs contains a region highly similar to the amino two thirds of the Von Willebrand Factor Type D domain (VWFd; Fig. 5B,E). VWFd domains are found in a number of secreted proteins, including Von Willebrand Factors and Mucins, and are involved in the regulation of disulfide bonds and the formation of protein multimers (see Discussion). The similarity ends at Cys646; the remaining portion of Cv-2 shares no significant similarity with any known protein, but does contain a potential Furin cleavage site (Fig. 5B; Nakayama, 1997).

The amino terminus of Cv-2 contains a hydrophobic region predicted to act as either a signal peptide, with a likely cleavage site between amino acids 53 and 54, or as an uncleaved transmembrane anchor (Nielsen et al., 1997). Cv-2 does not contain any obvious secretory retention signals, such as KDEL, suggesting that this protein acts extracellularly, being either secreted or membrane bound. This, and the presence of domains typical of secreted proteins, makes it likely that Cv-2 acts either extracellularly or in the secretory pathway.

Despite sharing similar CR domains, Cv-2 is not an obvious ortholog of Chordin, Sog or CRIM1. The number and arrangement of CRs differs, and the portion of Cv-2 outside the CRs is not similar to Chordin, Sog or CRIM1, as these lack a VWFd domain (Fig. 5E). Cv-2 is most similar to the Xenopus Kielin protein, in that Kielin contains a region of repeated CR domains, followed by a VWFd domain (Matsui et al., 2000; Fig. 5E). Kielin is much longer than Cv-2, however, having 28 CR domains, a full rather than a partial VWFd domain, and an amino-terminal region, similar to the amino-terminal domain of Thrombospondin, that is lacking in Cv-2 (Fig. 5E). Nonetheless, BLAST analyses indicate that Cv-2’s VWFd domain is more similar to the VWFd domain of Kielin than to other VWFd domains in the database, suggesting that the two might share a common ancestor (see Discussion).

cv-2 RNA and enhancer trap expression in normal and mutant flies

To provide further evidence that this transcript is cv-2, we generated probes to the 5′ and 3′ ends of the putative cv-2 open reading frame, as well as to the neighboring EST (LD27358) and elongation factor coding regions, and performed in situ hybridization. LD27358 and the elongation factor were expressed uniformly in the wing disc (not shown). However, the putative cv-2 transcript was expressed in a dynamic pattern, including expression in and near the developing CVs, that is identical to a combination of the lacZ expression patterns seen in both of the cv-2 enhancer trap lines. Moreover, this expression was lost or reduced in cv-2 mutants. In late third instar wing discs, both enhancer trap lines showed weak general expression and slightly heightened expression along the anteroposterior (AP) boundary in the region of dpp expression (Figs 6A, 7A). The cv-23511 line also showed strong expression in three distinct regions just distal to the prospective notum, as identified by the position of notal bristle precursors (Fig. 6A). In pupal wings, cv-23511 was expressed generally at low levels, but was also found at the tips of the LVs and, beginning at 25 hours a.p., in a broad region surrounding the CVs (Fig. 6B-D). cv-22253 retained a slight emphasis along the AP boundary in pupal wings, but lacked CV expression (not shown).

Fig. 6.

(A-D) Anti-β-gal staining (green) in the cv-23511 enhancer trap. (A) A late third instar wing disc, counterstained with anti-Asense (red) to identify bristle precursors. Note the anti-β-gal staining in three regions of notum. The anterior-proximal notal region of cv-23511 staining contains and lies distal to the posterior notopleural bristle (pNP), and lies posterior to the anterior notopleural bristle (aNP). The posterior-proximal region of cv-23511 staining lies distal to the trichoid sensillum (tri1). Faint staining is also observed along the AP boundary in the prospective wing blade. B-D. Pupal wings counterstained with anti-DSRF (red). Staining in and around the CVs is shown at 25 (B) and 28 (C) hours a.p. Staining in the ends of the LVs is shown at 32 hours a.p. (D), but is also present at earlier stages. E-H. In situ hybridization with cv-2 probe. (E) Late third instar wing disc. Expression is similar to that in the cv-23511 enhancer trap. (F-H) Staining in pupal wings. Faint diffuse staining is visible around the CVs (arrows) and the ends of the LVs at 22 hours a.p. (F), and becomes stronger and more tightly defined at 27 (G) and 28 (H) hours a.p.

Fig. 6.

(A-D) Anti-β-gal staining (green) in the cv-23511 enhancer trap. (A) A late third instar wing disc, counterstained with anti-Asense (red) to identify bristle precursors. Note the anti-β-gal staining in three regions of notum. The anterior-proximal notal region of cv-23511 staining contains and lies distal to the posterior notopleural bristle (pNP), and lies posterior to the anterior notopleural bristle (aNP). The posterior-proximal region of cv-23511 staining lies distal to the trichoid sensillum (tri1). Faint staining is also observed along the AP boundary in the prospective wing blade. B-D. Pupal wings counterstained with anti-DSRF (red). Staining in and around the CVs is shown at 25 (B) and 28 (C) hours a.p. Staining in the ends of the LVs is shown at 32 hours a.p. (D), but is also present at earlier stages. E-H. In situ hybridization with cv-2 probe. (E) Late third instar wing disc. Expression is similar to that in the cv-23511 enhancer trap. (F-H) Staining in pupal wings. Faint diffuse staining is visible around the CVs (arrows) and the ends of the LVs at 22 hours a.p. (F), and becomes stronger and more tightly defined at 27 (G) and 28 (H) hours a.p.

The putative cv-2 transcript was expressed in a pattern resembling a combination of the enhancer traps. There was weak general expression at all stages. Expression in late third instar wing discs was also strong in the three regions near the prospective notum, and weakly emphasized the AP boundary (Fig. 6E). At 19-21 hours a.p. expression was stronger in broad regions in and around the developing CVs; this expression resolved into narrower regions by 28 hours a.p. (Fig. 6F-H). Heightened expression was also observed at the tips of the LVs in pupal wings. This pattern is thus very similar to that of anti-p-Mad staining at these stages.

The cv-2 enhancer traps expressed lacZ in other imaginal tissues (Fig. 7B,C). The pattern in leg and antennal discs was similar to that of dpp; expression in both was higher along the dorsal or lateral sides, respectively, of the AP boundary. Fainter expression was also observed in eye discs posterior to the morphogenetic furrow.

Fig. 7.

(A) Anti-βgal staining in a late third instar cv-22253 enhancer trap. Note the faintly emphasized staining along the AP boundary. (B,C) X-gal staining of late third instar leg (B) and eye-antennal (C) discs in the cv-23511 enhancer trap. Note staining along the AP boundary in leg and antenna; dorsal expression is stronger in both. Also note the fainter staining posterior to the morphogenetic furrow in the eye.

Fig. 7.

(A) Anti-βgal staining in a late third instar cv-22253 enhancer trap. Note the faintly emphasized staining along the AP boundary. (B,C) X-gal staining of late third instar leg (B) and eye-antennal (C) discs in the cv-23511 enhancer trap. Note staining along the AP boundary in leg and antenna; dorsal expression is stronger in both. Also note the fainter staining posterior to the morphogenetic furrow in the eye.

Expression of the cv-2 message was altered in cv-21 and cv-22253 lines (Fig. 8). At late third instar, the heightened expression observed in the three regions near the notum was largely lost in cv-22253 and consistently reduced in cv-21 (Fig. 8B,C). Mutant adults have no obvious defects in the notum or hinge, so this loss is unlikely to be a secondary effect of a structural change. Expression near the CVs was also reduced in both these lines (Fig. 8D). However, general expression was not totally lost from either cv-22253 or cv-21, consistent with their hypomorphic phenotypes.

Fig. 8.

In situ hybridization with a cv-2 probe in normal mutant wing discs and pupal wings. (A-C) Late third instar wings discs. (A) Wild-type wing disc; control discs were stained in the same tubes as the mutant discs and marked with a cut (X). (B) cv-22253; notal staining is reduced or absent. (C) cv-21; notal staining is consistently reduced. D. 28 hours a.p. cv-22253 pupal wing; CV staining is reduced or absent.

Fig. 8.

In situ hybridization with a cv-2 probe in normal mutant wing discs and pupal wings. (A-C) Late third instar wings discs. (A) Wild-type wing disc; control discs were stained in the same tubes as the mutant discs and marked with a cut (X). (B) cv-22253; notal staining is reduced or absent. (C) cv-21; notal staining is consistently reduced. D. 28 hours a.p. cv-22253 pupal wing; CV staining is reduced or absent.

Misexpression of cv-2

EP elements contain multiple UAS sequences coupled to a heat-shock promoter; once inserted into the genome, these can be used in conjunction with Gal4 enhancer traps to drive the expression of neighboring genes (Rorth, 1996). As noted above, several EP lines contain EP insertions upstream of cv-2, close to the insertion site of the cv-23511 P element. Crossing the EP(2)1103 line to either patched-Gal4 or engrailed-Gal4 lines resulted in strong misexpression of the cv-2 transcript at levels equal to or greater than endogenous levels (Fig. 9A,B). Adults from these crosses were fully viable and no wing phenotypes were apparent (not shown). This suggests that wild-type levels of Cv-2, while required for CV formation, are not normally sufficient to induce ectopic venation. However, the stronger ap-Gal4 driver did generate a range of wing phenotypes; the most severely affected wings had ectopic venation near L2, vein deltas at the tips of the LVs, thickening of the ACV and, rarely, the partial loss of the PCV (Fig. 9C,D).

Fig. 9.

Misexpression of cv-2 using EP(2)1103. (A,B) In situ hybridization with a cv-2 probe in wing discs in which the EP(2)1103-driven misexpression of cv-2 was induced using patched-Gal4 (A) or engrailed-Gal4 (B). Levels of message in regions of ectopic expression (induced) were equal to or greater than levels in the regions of normal notal expression (normal). Adult wings from such flies were normal (not shown). (C,D) Adult wings in which cv-2 misexpression was induced using the stronger ap-Gal4 driver. (C) Note ectopic venation at the tips of the LVs and the thickening of the ACV (arrows). (D) Note partial loss of the PCV (arrow); this phenotype was rare.

Fig. 9.

Misexpression of cv-2 using EP(2)1103. (A,B) In situ hybridization with a cv-2 probe in wing discs in which the EP(2)1103-driven misexpression of cv-2 was induced using patched-Gal4 (A) or engrailed-Gal4 (B). Levels of message in regions of ectopic expression (induced) were equal to or greater than levels in the regions of normal notal expression (normal). Adult wings from such flies were normal (not shown). (C,D) Adult wings in which cv-2 misexpression was induced using the stronger ap-Gal4 driver. (C) Note ectopic venation at the tips of the LVs and the thickening of the ACV (arrows). (D) Note partial loss of the PCV (arrow); this phenotype was rare.

As shown both morphologically (Waddington, 1940; Mohler and Swedburg, 1964) and using molecular markers (see above), the definitive CVs are not formed until long after the initial specification of the LVs. The CVs therefore must form within territory that has already been specified as intervein. The CVs must also interconnect with existing LVs at a time when the Delta expressed by the LVs is thought to inhibit vein formation in adjacent cells (de Celis et al., 1997; Huppert et al., 1997). Mechanisms must exist that override both intervein specification and the lateral inhibition of veins, allowing the formation of continuous, interconnected vein tissue.

Our results show that BMP-like signaling plays a special role in the formation of the CVs from within intervein territory. BMP-like signals also help maintain the connections between the LVs and the margin of the wing. cv-2 is a critical factor in these processes, as it is expressed more highly in the CVs and the ends of the LVs and is required for the high levels of BMP-like signaling observed in these regions. The structure of the Cv-2 protein strongly suggests that these effects are direct, and that Cv-2 is a novel player in the BMP-like signaling pathway.

Mad signaling and the formation of cross veins

While MAPK signaling appears to play the primary role in the initial specification of most or all of the LV proveins in third instar discs, BMP-like signaling helps maintain the LVs during later stages. After the first day of pupal development, dpp is lost from the L3-L4 intervein and is expressed along the LVs and, later, along the CVs; this expression is required to reinforce vein formation (Posakony et al., 1990; Yu et al., 1996; de Celis, 1997). We observed that Mad is indeed activated in all of the veins during pupal stages. However, Mad activity was higher and appeared earlier in the CVs and the ends of the LVs, the regions that are especially sensitive to reductions in BMP-like signaling. Mad activation may play the initial role during CV formation, preceding the activation of MAPK. Higher Mad activity was detected in the CVs before the loss of DSRF expression which is thought to be mediated by MAPK activity (Roch et al., 1998). Ectopic Dpp signaling is capable of inducing vein formation in regions lacking high MAPK activity (Yu et al., 1996; de Celis, 1997).

CV formation is sensitive to reductions in Gbb and Dpp expression (Posakony et al., 1991; Khalsa et al., 1998), and the genetic interactions of both dpp and gbb with cv-2 mutations shown above indicate that both Gbb and Dpp play a role in the local activation of Mad in the CVs. However, localized expression of gbb or dpp cannot apparently account for the initial activation of Mad near the CVs. At the stage when Mad activity was first detected in the CVs, dpp is expressed along the LVs but is not detectable in the CVs (Yu et al., 1996; de Celis, 1997; Martin-Blanco et al., 1999), and anti-Gbb staining did not obviously emphasize the CVs (Fig. 4A). While it is possible that other BMP-like ligands locally activate Mad, Screw, the only other known BMP-like ligand known in Drosophila, has not been detected in pupae (Arora et al., 1994). It is therefore likely that other localized factors potentiate Mad activity in the CVs. Our results suggest that Cv-2 is such a factor.

A similar situation occurs during axis formation in the embryo, as extracellular modulators are thought to help localize BMP-like signaling (reviewed in Podos and Ferguson, 1999). Mad activity is highest on the extreme dorsal side of the embryo, but Dpp is expressed in a much broader dorsolateral domain and the other BMP-like ligand Screw is expressed ubiquitously. Ventrally expressed Sog inhibits Screw activity and thus provides a further dorsal bias in signaling (Neul and Ferguson, 1998; Nguyen et al., 1998). Sog may also act as a long-range activator of signaling in a process that requires cleavage by the secreted, dorsally expressed Tolloid protease (Zusman et al., 1988; Ashe and Levine, 1999). The dorsally secreted Twisted gastrulation protein (Tsg) also modulates signaling (Mason et al., 1994, 1997), perhaps by altering the cleavage of Sog by Tolloid (Yu et al., 2000), or more directly by binding ligands and potentiating signaling downstream of Sog (Oelgeschlager et al., 2000).

Cv-2 and vein formation

cv-2 is expressed at higher levels in the developing CVs and the ends of the LVs, and is required for Mad activation in these regions. The structure of the Cv-2 protein is consistent with a direct role in Dpp or Gbb signaling. Cv-2 contains a putative signal or transmembrane region, and CR and VWFd domains typical of secreted proteins, suggesting that it acts either extracellularly or in the secretory pathway. Moreover, the strong similarity between the five closely apposed CR domains in Cv-2 and the CR domains in Chordin, Sog and Procollagen (Fig. 5C) suggest that Cv-2 may directly bind Dpp or Gbb. Chordin and Procollagen CRs bind BMPs and can mediate Chordin-like inhibition of BMP signaling (Zhu et al., 1999; Larrain et al., 2000). Xenopus Tsg contains a partial CR that can also bind BMPs (Oelgeschlager et al., 2000).

As Sog overexpression induces cv-2-like phenotypes (Yu et al., 1996), Sog and Cv-2 have opposing effects on vein formation. Sog is expressed in the interveins during CV formation and is likely to diffuse to non-expressing cells (Yu et al., 1996). To raise BMP-like activity, localized factors may have to overcome the Sog-mediated inhibition of signaling.

Thus, one possible role for Cv-2 is to protect or release Gbb or Dpp from Sog. This is similar to the role proposed for Xenopus Tsg (Oelgeschlager et al., 2000). While clones lacking sog do not induce ectopic CVs (Yu et al., 1996), such clones might be rescued by the diffusion of Sog from outside the clone. Sog can act over long distances; overexpression of Sog in the anterior compartment of the wing can block PCV formation in the posterior (not shown).

Another possible role for Cv-2 is in the activation or stability of the ligands themselves. BMPs are initially expressed as inactive precursors that must be cleaved and stabilized during secretion (reviewed in Constam and Robertson, 1999). Interestingly, Thrombospondin, which also contains a Chordin-like CR region, is required for the activation of a cleaved but latent form of TGFβ? in vivo (Crawford et al., 1998).

Cv-2 also contains a partial VWFd domain. VWFd domains are found in a number of secreted proteins, including Von Willebrand Factors (VWFs) and Mucins, and both directly bridge protein multimers and regulate the formation of intermolecular and intramolecular disulfide bonds (reviewed in Sadler, 1998; Perez-Vilar and Hill, 1999). Cv-2 lacks the portion of the VWFd domain that has been proven to directly bridge VWFs or Mucins. Cv-2 does, however, contain a CGLCG motif, which in VWFs and Mucins is required for the formation of protein multimers; this region is similar to a motif found in disulfide isomerases, and thus may regulate the formation of disulfide bonds (Mayadas and Wagner, 1992; Azuma et al., 1993; Dong et al., 1994; Perez-Vilar and Hill, 1998).

The presence of multiple CR domains and a VWFd domain makes Cv-2 different from Chordin, Sog, Tsg or CRIM1, but similar to the recently identified Xenopus protein Kielin which, like Cv-2, can regulate BMP-dependent patterning (Fig. 5E; Matsui et al., 2000). Kielin and Cv-2 are not identical, however (Fig. 5E). The much longer 2,327 amino acid Kielin has 28 CR domains instead of the five found in Cv-2, followed by a full rather than a partial VWFd domain, and has an amino-terminal region, similar to the amino-terminal domain of Thrombospondin, which is lacking in Cv-2. Thus, while it is possible that these proteins are homologs, one or both would have to have been severely modified from its original form.

Cv-2 and Kielin also differ functionally. While Cv-2 is required for high levels of BMP-like signaling in the CVs, ectopic Kielin inhibits some (but not all) forms of BMP-like signaling during axis formation in Xenopus (ibid). This functional difference may be caused by the structural differences between these proteins. Alternatively, other factors may regulate the ability of these two proteins to either inhibit or potentiate signaling. This would not be surprising in a situation where multiple ligand-binding proteins compete for available ligand. The cleavage of Chordin or Sog by Tolloid apparently lowers their affinity for ligand. In the Drosophila embryo, this is thought to convert Sog from an inhibitor to an activator of signaling, as Sog is required for the highest levels of BMP-like activity in a process that requires Tolloid (Zusman et al., 1988; Ashe and Levine, 1999). Xenopus Tsg potentiates BMP signaling, possibly by protecting ligand that would otherwise be sequestered by the cleaved form of Chordin; however, Tsg may also sequester ligand by forming a complex with full-length Chordin (Oelgeschlager et al., 2000).

The activity of Cv-2 thus may be highly context dependent and, indeed, our evidence suggests that other locally expressed factors are required for Cv-2’s activity. While cv-2 is necessary for Mad activity surrounding the CVs, misexpression of cv-2 message at levels at or above normal levels had little effect on vein formation or other types of Mad-dependent development. Flies that had misexpressed cv-2 throughout embryogenesis, larval and pupal life were viable and appeared normal. Only the very high levels of wing expression driven by ap-Gal4 induced ectopic vein phenotypes, and these were limited to the regions near where Mad activity is normally high. It is therefore unlikely that Cv-2 alone is responsible for localizing Mad activation. Other extracellular regulators of Mad activity are similar in this regard: for instance, ectopic Tsg expression has little effect on the Drosophila embryo (Mason et al., 1997), and overexpression in the wing causes a mild phenotype similar to that observed after misexpression of Cv-2 (Yu et al., 2000). The remaining crossveinless loci provide obvious candidates for the missing factors.

We thank the Keck Neural Imaging Center and Dr James Pawley for use of their confocal microscopes. This work was supported by grants to S. S. B. from NIH (R01-NS28202) and NSF (IBN-9305209, IBN-9723564), and to W. G. from NIH (R37-GM28669-20).

Adams
,
M. D.
et al.  (
2000
).
The genome sequence of Drosophila melanogaster
.
Science
287
,
2185
2195
.
Arora
,
K.
,
Levine
,
M. S.
and
O’Connor
,
M. B.
(
1994
).
The screw gene encodes a ubiquitously expressed member of the TGF-beta family required for specification of dorsal cell fates in the Drosophila embryo
.
Genes Dev
.
8
,
2588
2601
.
Ashe
,
H. L.
and
Levine
,
M.
(
1999
).
Local inhibition and long-range enhancement of Dpp signal transduction by Sog
.
Nature
398
,
427
431
.
Azuma
,
H.
,
Hayashi
,
T.
,
Dent
,
J. A.
,
Ruggeri
,
Z. M.
and
Ware
,
J.
(
1993
).
Disulfide bond requirements for assembly of the platelet glycoprotein Ib-binding domain of von Willebrand factor
.
J. Biol. Chem
.
268
,
2821
2827
.
Bateman
,
A.
,
Birney
,
E.
,
Durbin
,
R.
,
Eddy
,
S. R.
,
Howe
,
K. L.
and
Sonnhammer
,
E. L. L.
(
2000
).
The Pfam protein families database
.
Nucleic Acids Res
.
28
,
263
266
.
Blair
,
S. S.
(
1994
).
A role for the segment polarity gene shaggy-zeste white 3 in the specification of regional identity in the developing wing of Drosophila
.
Dev. Biol
.
162
,
229
244
.
Blair
,
S. S.
and
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
.
Blair
,
S. S.
(
2000
).
Imaginal discs
. In
Drosophila Protocols
(ed.
W.
Sullivan
,
M.
Ashburner
and
R. S.
Hawley
), pp.
159
-
173
, Cold Spring Harbor:
Cold Spring Harbor Laboratory Press
.
Brand
,
M.
,
Jarman
,
A. P.
,
Jan
,
L. Y.
and
Jan
,
Y. N.
(
1993
).
asense is a Drosophila neural precursor gene and is capable of initiating sense organ formation
.
Development
119
,
1
17
.
Burke
,
R.
and
Basler
,
K.
(
1996
).
Dpp receptors are autonomously required for cell proliferation in the entire developing Drosophila wing
.
Development
122
,
2261
2269
.
Cadigan
,
K. M.
,
Fish
,
M. P.
,
Rulifson
,
E. J.
and
Nusse
,
R.
(
1998
).
Wingless repression of Drosophila frizzled 2 expression shapes the Wingless morphogen gradient in the wing
.
Cell
93
,
767
777
.
Castrillon
,
D. H.
,
Gonczy
,
P.
,
Alexander
,
S.
,
Rawson
,
R.
,
Eberhart
,
C. G.
,
Viswanathan
,
S.
,
DiNardo
,
S.
and
Wasserman
,
S. A.
(
1993
).
Toward a molecular genetic analysis of spermatogenesis in Drosophila melanogaster: characterization of male-sterile mutants generated by single P element mutagenesis
.
Genetics
.
135
,
489
505
.
Constam
,
D. B.
and
Roberston
,
E. J.
(
1999
).
Regulation of Bone Morphogenetic Protein activity by pro domains and proprotein convertases
.
J. Cell Biol
.
144
,
139
149
.
Crawford
,
S. E.
,
Stellmach
,
V.
,
Murphy-Ullrich
,
J. E.
,
Ribeiro
,
S. M. F.
,
Lawler
,
J.
,
Hynes
,
R. O.
,
Boivin
,
G. P.
and
Bouck
,
N.
(
1998
).
Thrombospondin-1 is a major activator of TGF-β1 in vivo
.
Cell
93
,
1159
1170
.
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.
,
Bray
,
S.
and
Garcia-Bellido
,
A.
(
1997
).
Notch signaling regulates veinlet expression and establishes boundaries between veins and interveins in the Drosophila wing
.
Development
124
,
1919
1928
.
Diaz-Benjumea
,
F. J.
and
Garcia-Bellido
,
A.
(
1990
).
Behaviour of cells mutant for an EGF receptor homologue of Drosophila in genetic mosaics
.
Proc. R. Soc. Lond. B
242
,
35
44
.
Doctor
,
J. S.
,
Jackson
,
P. D.
,
Rashka
,
K. E.
,
Visalli
,
M.
and
Hoffmann
,
F. M.
(
1992
).
Sequence, biochemical characterization, and developmental expression of a new member of the TGF-β superfamily in Drosophila melanogaster
.
Dev. Biol
.
151
,
491
505
.
Dong
,
Z.
,
Thoma
,
R. S.
,
Crimmins
,
D. L.
,
McCourt
,
D. W.
,
Tuley
,
E. A.
and
Sadler
,
J. E.
(
1994
).
Disulfide bonds required to assemble functional von Willebrand factor multimers
.
J. Biol. Chem
.
269
,
6753
6758
.
Finelli
,
A. L.
,
Xie
,
T.
,
Bossie
,
C. A.
,
Blackman
,
R. K.
and
Padgett
,
R. W.
(
1995
).
The tolkin gene is a tolloid BMP-1 homologue that is essential for Drosophila development
.
Genetics
141
,
271
281
.
Francois
,
V.
,
Solloway
,
M.
,
O’Neill
,
J. W.
,
Emery
,
J.
and
Bier
,
E.
(
1994
).
Dorsal-ventral patterning of the Drosophila embryo depends on a putative negative growth factor encoded by the short gastrulation gene
.
Genes Dev
.
8
,
2602
2616
.
Francois
,
V.
and
Bier
,
E.
(
1995
).
Xenopus chordin and Drosophila short gastrulation genes encode homologous proteins functioning in dorsal-ventral axis formation
.
Cell
80
,
19
20
.
Fristrom
,
D. K.
,
Wilcox
,
M. E.
and
Fristrom
,
J.
(
1993
).
The distribution of PS integrins, laminin A and F-actin during key stages in Drosophila wing development
.
Development
117
,
509
523
.
Gabay
,
L.
,
Seger
,
R.
and
Shilo
,
B. Z.
(
1997
).
In situ activation pattern of Drosophila EGF receptor pathway during development
.
Science
277
,
1103
1106
.
Garcia-Bellido
,
A.
and
de Celis
,
J. F.
(
1992
).
Developmental genetics of the venation pattern of Drosophila
.
Ann. Rev. Genet
.
26
,
277
304
.
Guichard
,
A.
,
Biehs
,
B.
,
Sturtevant
,
M. A.
,
Wickline
,
L.
,
Chacko
,
J.
,
Howard
,
K.
and
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.
and
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
.
Holley
,
S. A.
,
Jackson
,
P. D.
,
Sasai
,
Y.
,
Lu
,
B.
,
De Robertis
,
E. M.
,
Hoffmann
,
F. M.
and
Ferguson
,
E. L.
(
1995
).
A conserved system for dorsal-ventral patterning in insects and vertebrates involving sog and chordin
.
Nature
376
,
249
253
.
Huang
,
X.
(
1994
).
On global sequence alignment
.
Computer Appl. Biosci
.
10
,
227
235
.
Huppert
,
S. S.
,
Jacobsen
,
T. L.
and
Muskavitch
,
M. A.
(
1997
).
Feedback regulation is central to Delta-Notch signalling required for Drosophila wing vein morphogenesis
.
Development
124
,
3283
3291
.
Kassis
,
J. A.
,
Noll
,
E.
,
VanSickle
,
E. P.
,
Odenwald
,
W. F.
and
Perrimon
,
N.
(
1992
).
Altering the insertional specificity of a Drosophila transposable element
.
Proc. Natl Acad. Sci. USA
89
,
1919
1923
.
Khalsa
,
O.
,
Yoon
,
J.
,
Torres-Schumann
,
S.
and
Wharton
,
K. A.
(
1998
).
TGF-b/BMP superfamily members, gbb-60A and dpp, cooperate to provide pattern information and establish cell identity in the Drosophila wing
.
Development
125
,
2723
2734
.
Kolle
,
G.
,
Georgas
,
K.
,
Holmes
,
G. P.
,
Little
,
M. H.
and
Yamada
,
T.
(
2000
).
CRIM1, a novel gene encoding a cysteine-rich repeat protein, is developmentally regulated and implicated in vertebrate CNS development and organogenesis. Mech. Dev
.
90
,
181
193
.
Kovalick
,
G. E.
,
Schreiber
,
M. C.
,
Dickason
,
A. K.
and
Cunningham
,
R. A.
(
1998
).
Structure and expression of the antigen 5-related gene of Drosophila melanogaster
.
Insect Biochem. Mol. Biol
.
28
,
491
500
.
Larrain
,
J.
,
Bachiller
,
D.
,
Lu
,
B.
,
Agius
,
E.
,
Piccolo
,
S.
and
De Robertis
,
E. M.
(
2000
).
BMP-binding modules in chordin: a model for signalling regulation in the extracellular space
.
Development
127
,
821
830
.
Lindsley
,
D. L.
and
Zimm
,
G. G.
(
1992
).
The Genome of Drosophila melanogater
.
San Diego
:
Academic Press
.
Marques
,
G.
,
Musacchio
,
M.
,
Shimell
,
M. J.
,
Wunnenberg-Stapleton
,
K.
,
Cho
,
K.W.
and
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.
,
Roch
,
F.
,
Noll
,
E.
,
Baonza
,
A.
,
Duffy
,
J. B.
and
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.
and
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
.
Mason
,
E. D.
,
Williams
,
S.
,
Grotendorst
,
G. R.
and
Marsh
,
J. L.
(
1997
).
Combinatorial signaling by twisted gastrulation and decapentaplegic
.
Mech. Dev
.
64
,
61
75
.
Matsui
,
M.
,
Mizuseki
,
K.
,
Nakatani
,
J.
,
Nakanishi
,
S.
and
Sasai
,
Y.
(
2000
).
Xenopus Kielin: A dorsalizing factor containing muliple chordin-type repeats secreted from the embryonic midline
.
Proc. Natl Acad. Sci. USA
97
,
5291
5296
.
Mayadas
,
T. N.
and
Wagner
,
D. D.
(
1992
).
Vicinal cysteines in the prosequence play a role in von Willebrand factor multimer assembly
.
Proc. Natl Acad. Sci. USA
89
,
3531
3535
.
Micchelli
,
C. A.
,
Rulifson
,
E. J.
and
Blair
,
S. S.
(
1997
).
The function and regulation of cut expression on the wing margin of Drosophila: Notch, Wingless, and a dominant negative role for Delta and Serrate
.
Development
124
,
1485
1495
.
Mohler
,
J. D.
and
Swedberg
,
G. L. S.
(
1964
).
Wing vein development in crossveinless-like strains of Drosophila melanogaster
.
Genetics
50
,
1403
1419
.
Mohler
,
J.
,
Seecoomar
,
M.
,
Agarwal
,
S.
,
Bier
,
E.
and
Hsai
,
J.
(
2000
).
Activation of knot (kn) specifies the 3-4 intervein region in the Drosophila wing
.
Development
127
,
55
63
.
Montagne
,
J.
,
Groppe
,
J.
,
Guillemin
,
K.
,
Krasnow
,
M. A.
,
Gehring
,
W. J.
and
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
.
Murray
,
M. A.
,
Schubiger
,
M.
and
Palka
,
J.
(
1984
).
Neuron differentiation and axon growth in the developing wing of Drosophila melanogaster
.
Dev. Biol
.
104
,
259
273
.
Murray
,
M. A.
,
Fessler
,
L. I.
and
Palka
,
J.
(
1995
).
Changing distributions of extracellular matrix components during early wing morphogenesis in Drosophila
.
Dev. Biol
.
168
,
150
165
.
Nakayama
,
K.
(
1997
).
Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins
.
Biochem J
.
327
,
625
635
.
Nambu
,
J. R.
,
Franks
,
R. G.
,
Hu
,
S.
and
Crews
,
S. T.
(
1990
).
The single-minded gene of Drosophila is required for the expression of genes important for the development of CNS midline cells
.
Cell
63
,
63
75
.
Neul
,
J. L.
and
Ferguson
,
E. L.
(
1998
).
Spatially restricted activation of the SAX receptor by SCW modulates DPP/TKV signaling in Drosophila dorsal-ventral patterning
.
Cell
95
,
483
494
.
Nguyen
,
T.
,
Jamal
,
J.
,
Shimell
,
M. J.
,
Arora
,
K.
and
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.
and
Arora
,
K.
(
1998
).
Interaction of a BMP activity gradient in Drosophila embryos depends on synergistic signaling by two type I receptors, SAX and TKV
.
Cell
95
,
495
506
.
Nielsen
,
H.
,
Engelbrecht
,
J.
,
Brunak
,
S.
and
von Heijne
,
G.
(
1997
).
Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites
.
Protein Engineering
10
,
1
6
.
Oelgeschlager
,
M.
,
Larrain
,
J.
,
Geissert
,
D.
and
De Robertis
,
E. M.
(
2000
).
The evolutionarily conserved BMP-binding protein Twisted gastrulation promotes BMP signaling
.
Nature
403
,
757
763
.
Perez-Villar
,
J.
and
Hill
,
R. L.
(
1998
).
Identification of the half-cystine residues in porcine submaxillary mucin critical for mulimerization through the D-domains
.
J. Biol. Chem
.
273
,
34527
34534
.
Perez-Villar
,
J.
and
Hill
,
R. L.
(
1999
).
The structure and assembly of secreted mucins
.
J. Biol. Chem
.
274
,
31751
31754
.
Persson
,
U.
,
Izumi
,
H.
,
Souchelnytskyi
,
S.
,
Itoh
,
S.
,
Grimsby
,
S.
,
Engstrom
,
U.
,
Heldin
,
C. −H.
,
Funa
,
K.
and
ten Dijke
,
P.
(
1998
).
The L45 loop in type I receptors for TGF-β family members is a critical determinant in specifying Smad isoform activation
.
FEBS Lett
.
434
,
83
87
.
Phillips
,
R. G.
and
Whittle
,
J. R. S.
(
1993
).
Wingless expression mediates determination of peripheral nervous system elements in late stages of Drosophila wing disc development
.
Development
118
,
427
438
.
Piccolo
,
S.
,
Sasai
,
Y.
,
Lu
,
B.
and
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
.
Piccolo
,
S.
,
Agius
,
E.
,
Lu
,
B.
,
Goodman
,
S.
,
Dale
,
L.
and
De Robertis
,
E. M.
(
1997
).
Cleavage of Chordin by Xolloid metalloprotease suggests a role for proteolytic processing in the regulation of Spemann organizer activity
.
Cell
91
,
407
416
.
Podos
,
S. D.
and
Ferguson
,
E. L.
(
1999
).
Morphogen gradients: new insights from Dpp
.
Trends Genet
.
15
,
396
402
.
Posakony
,
L. G.
,
Raftery
,
L. A.
and
Gelbart
,
W. M.
(
1990
).
Wing formation in Drosophila melanogaster requires decapentaplegic gene function along the anterior-posterior compartment boundary
.
Mech. Dev
.
33
,
69
82
.
Roch
,
F.
,
Baonza
,
A.
,
Martin-Blanco
,
E.
and
Garcia-Bellido
,
A.
(
1998
).
Genetic interactions and cell behaviour in blistered mutants during proliferation and differentiation of the Drosophila wing
.
Development
125
,
1823
1832
.
Rorth
,
P.
(
1996
).
A modular misexpression screen in Drosophila detecting tissue-specific phenotypes
.
Proc. Natl Acad. Sci. USA
93
,
12418
12422
.
Sadler
,
J. E.
(
1998
).
Biochemistry and genetics of Von Willebrand Factor
.
Ann. Rev. Biochem
.
67
,
395
424
.
Sambrook
,
J.
,
Fritsch
,
E. F.
and
Maniatis
,
T.
(
1989
).
Molecular Cloning, 2nd Edition
.
Cold Spring Harbor: Cold Spring Harbor Laboratory Press
.
Sasai
,
Y.
,
Lu
,
B.
,
Steinbeisser
,
H.
,
Geissert
,
D.
,
Gont
,
L. K.
and
De Robertis
,
E. M.
(
1994
).
Xenopus chordin: a novel dorsalizing factor activated by organizer-specific homeobox genes
.
Cell
79
,
779
790
.
Sasai
,
Y.
,
Lu
,
B.
,
Steinbeisser
,
H.
and
De Robertis
,
E. M.
(
1995
).
Regulation of neural induction by the Chd and Bmp-4 antagonistic patterning signals in Xenopus
.
Nature
376
,
333
336
.
Sawamoto
,
K.
,
Okano
,
H.
,
Kobayakawa
,
Y.
,
Hayashi
,
S.
,
Mikoshiba
,
K.
and
Tanimura
,
T.
(
1994
).
The function of argos in regulating cell fate decisions during Drosophila eye and wing vein development
.
Dev. Biol
.
164
,
267
276
.
Schmidt
,
J.
,
Francois
,
V.
,
Bier
,
E.
and
Kimelman
,
D.
(
1995
).
Drosophila short gastrulation induces an ectopic axis in Xenopus: evidence for conserved mechanisms of dorsal-ventral patterning
.
Development
121
,
4319
4328
.
Schweitzer
,
R.
,
Howes
,
R.
,
Smith
,
R.
,
Shilo
,
B. Z.
,
Freeman
,
M.
(
1995
).
Inhibition of Drosophila EGF receptor activation by the secreted protein Argos
.
Nature
376
,
699
702
.
Scott
I. C.
,
Blitz
I. L.
,
Pappano
W. N.
,
Imamura
Y.
,
Clark
T. G.
,
Steiglitz
B. M.
,
Thomas
C. L.
,
Maas
S. A.
,
Takahara
K.
,
Cho
K. W.
and
Greenspan
D. S.
(
1999
).
Mammalian BMP-1/Tolloid-related metalloproteinases, including novel family member mammalian Tolloid-like 2, have differential enzymatic activities and distributions of expression relevant to patterning and skeletogenesis
.
Dev. Biol
.
213
,
283
300
.
Shellenbarger
,
D. L.
and
Mohler
,
J. D.
(
1975
)
Temperature-sensitive mutations of the
Notch locus in Drosophila melanogaster. Genetics
81
,
143
162
.
Shimell
,
M. J.
,
Ferguson
,
E. L.
,
Childs
,
S. R.
and
O’Connor
,
M. B.
(
1991
).
The Drosophila dorsal-ventral patterning gene tolloid is related to human bone morphogenetic protein 1
.
Cell
67
,
469
481
.
Simcox
,
A. A.
,
Grumbling
,
G.
,
Schnepp
,
B.
,
Bennington-Mathias
,
C.
,
Hersperger
,
E.
and
Shearn
,
A.
(
1996
).
Molecular, phenotypic, and expression analysis of vein, a gene required for growth of the Drosophila wing disc
.
Dev. Biol
.
177
,
475
489
.
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
.
Solovyev
,
V. V.
,
Salamov
,
A. A.
and
Lawrence
,
C. B.
(
1995
).
Identification of human gene structure using linear discriminant functions and dynamic programming
. In
Proceedings of the Third International Conference on Intelligent Systems for Molecular Biology
(ed.
C.
Rawling
,
D.
Clark
,
R.
Altman
,
L.
Hunter
,
T.
Lengauer
and
S.
Wodak
), pp.
367
-
375
., Cambridge, UK: AAAI Press.
Sturtevant
,
M. A.
,
Roark
,
M.
and
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
.
Sturtevant
,
M. A.
and
Bier
,
E.
(
1995
).
Analysis of the genetic hierarchy guiding wing vein development in Drosophila
.
Development
121
,
785
801
.
Tanimoto
,
H.
,
Itoh
,
S.
,
ten Dijke
,
P.
and
Tabata
,
T.
(
2000
).
Hedgehog creates a gradient of Dpp activity in Drosophila wing imaginal discs
.
Mol. Cell
5
,
59
71
.
Thompson
,
J. D.
,
Higgins
,
D. G.
and
Gibson
,
T. J.
(
1994
).
CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice
.
Nuc. Acids Res
.
22
,
4673
4680
.
Vervoot
,
M.
,
Crozatier
,
M.
,
Valle
,
D.
and
Vincent
,
A.
(
1999
).
The COE transcription factor Collier is a mediator of short-range Hedgehog-induced patterning of the Drosophila wing
.
Curr. Biol
.
9
,
632
639
.
Waddington
,
C. H.
(
1940
).
The genetic control of wing development in Drosophila
.
J. Genet
.
41
,
75
139
.
Wessells
,
R. J.
,
Grumbling
,
G.
,
Donaldson
,
T.
,
Wang
,
S.-H.
and
Simcox
,
A.
(
1999
).
Tissue-specific regulation of Vein-EGF Receptor signaling in Drosophila
.
Dev. Biol
.
216
,
243
259
.
Wilson
,
C.
,
Pearson
,
R. K.
,
Bellen
,
H. J.
,
O’Kane
,
C. J.
,
Grossniklaus
,
U.
and
Gehring
,
W. J.
(
1989
).
P-element-mediated enhancer detection: an efficient method for isolating and characterizing developmentally regulated genes in Drosophila
.
Genes Dev
.
3
,
1301
1313
.
Yu
,
K.
,
Sturtevant
,
M. A.
,
Biehs
,
B.
,
Francois
,
V.
,
Padgett
,
R. W.
,
Blackman
,
R. K.
and
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.
and
Bier
,
E.
(
2000
).
Processing of the Drosophila Sog protein creates a novel BMP inhibitory activity
.
Development
127
,
2143
2154
.
Zhu
,
Y.
,
Oganesian
,
A.
,
Keene
,
D. R.
and
Sandell
,
L. J.
(
1999
).
Type IIA procollagen containing the cysteine-rich amino propeptide is deposited in the extracellular matrix of prechondrogeneic tissue and binds to TGF-beta1 and BMP-2
.
J. Cell Biol
.
144
,
1069
1080
.
Zusman
,
S.
,
Sweeton
,
D.
and
Wieschaus
,
E.
(
1988
).
Short gastrulation, a mutation causing delays in stage-specific cell shape changes during gastrulation in Drosophila melanogaster. Dev. Biol
.
129
,
417
427
.