Search

Signalling proteins in the BMP-decapentaplegic (dpp), WNT-wingless (wg) and Shh-hedgehog (hh) families have been implicated in limb and appendage development in both invertebrates and vertebrates. In Drosophila, dpp protein (Dpp) induces distal outgrowth and patterning of legs and wings, but the molecular responses to Dpp are not well characterized. Analysis of clones mutant for the Dpp receptors encoded by punt or thickveins (tkv) reveals that repression of wg expression is one critical function of Dpp signalling in leg and wing discs. Distal clones that lie on the anterior edge of the anterior-posterior compartment boundary ectopically express wg and cause pattern abnormalities, suggesting that Dpp represses Hh activation of wg in the distal primordia of the leg and wing. By repressing wg expression in the leg, Dpp signalling limits the region that responds to high levels of Wg and Dpp to the site of distal outgrowth. Such negative regulatory feedback loops between signalling molecules are likely to be critical for limb patterning in other species.

We have used the GAL4-UAS expression system to increase the level of expression of the Drosophila gene decapentaplegic (dpp) in a pattern approximating its normal pattern in leg and wing imaginal discs. Intermediate increases of dpp expression have little effect in wing discs but high levels of dpp overexpression lead to reduction of the scutellum and duplication of posterior wing structures. In leg discs intermediate increases cause supernumerary outgrowths of ventral leg structures in the anterior-ventral region. Greater increases of dpp expression cause the loss of ventral leg structures with the concomitant fusion of left and right dorsal forelegs. The defects observed in both legs and wings appear to arise through dose-dependent effects of dpp on wingless (wg) expression. A high level of dpp overexpression in the wing disc causes reduction of wg expression in the presumptive scutellar region, consistent with the subsequent reduction of the scutellum. An intermediate increase of dpp expression in leg discs induces the expansion of wg expression into the ventral outgrowths. At higher dpp expression levels, ventral wg expression in leg discs is eliminated, consistent with the loss of ventral leg cuticle. In the leg disc end knob and in the wing margin primordium, where wg and dpp cooperate in producing distal outgrowth, dpp overexpression has no detectable effect either on patterning or on wg expression. We propose that a critical role for dpp in other regions of the leg and wing discs is to reduce or block the expression of wg. This role of dpp is supported by the observation that ectopic wg expression is detected in imaginal discs where dpp signaling is compromised by lowering the activity of one of its receptors, tkv. This antagonism between dpp and wg expression may be critical to assigning only one disc region as the distal organizer.

Little is known about the signal transduction pathways by which cells respond to mammalian TGF-beta s or to decapentaplegic (dpp), a Drosophila TGF-beta-related factor. Here we describe the genetic and molecular characterization of Drosophila schnurri (shn), a putative transcription factor implicated in dpp signaling. The shn protein has eight zinc fingers and is related to a human transcription factor, PRDII/MBPI/HIV-EP1, that binds to nuclear factor-kappa B-binding sites and activates transcription from the HIV long terminal repeat (LTR). shn mRNA is expressed in a dynamic pattern in the embryo that includes most of the known target tissues of dpp, including the dorsal blastoderm, the mesodermal germlayer and parasegments 4 and 7 of the midgut. Mutations in shn affect several developmental processes regulated by dpp including induction of visceral mesoderm cell fate, dorsal/ventral patterning of the lateral ectoderm and wing vein formation. Absence of shn function blocks the expanded expression of the homeodomain protein bagpipe in the embryonic mesoderm caused by ectopic dpp expression, illustrating a requirement for shn function downstream of dpp action. We conclude that shn function is critical for cells to respond properly to dpp and propose that shn protein is the first identified downstream component of the signal transduction pathway used by dpp and its receptors.

Dorsal-ventral patterning within the ectoderm of the Drosophila embryo requires seven zygotic genes, including short gastrulation (sog). Here we demonstrate that sog, which is expressed in the ventrolateral region of the embryo that gives rise to the nerve cord, is functionally homologous to the chordin gene of Xenopus, which is expressed in the dorsal blastopore lip of the embryo and in dorsal mesoderm, in particular the notochord. We show by injections of messenger RNA that both sog and chordin can promote ventral development in Drosophila, and that sog, like chordin, can promote dorsal development in Xenopus. In Drosophila, sog antagonizes the dorsalizing effects of decapentaplegic (dpp), a member of the transforming growth factor-beta family. One of the dpp homologues in vertebrates, bmp-4, is expressed ventrally in Xenopus and promotes ventral development. We show that dpp can promote ventral fates in Xenopus, and that injection of sog mRNA counteracts the ventralizing effects of dpp. These results suggest the molecular conservation of dorsoventral patterning mechanisms during evolution.

Signaling by TGF beta-related factors requires ligand-induced association between type I and type II transmembrane serine/threonine kinases. In Drosophila, the saxophone (sax) and thick veins (tkv) genes encode type I receptors that mediate signaling by decapentaplegic (dpp), a member of the bone morphogenetic protein (BMP) subgroup of TGF beta-type factors. In this report, we demonstrate that the Drosophila punt gene encodes atr-II, a previously described type II receptor that on its own is able to bind activin but not BMP2, a vertebrate ortholog of dpp. Mutations in punt produce phenotypes similar to those exhibited by tkv, sax, and dpp mutants. Furthermore, punt will bind BMP2 in concert with tkv or sax, forming complexes with these receptors. We suggest that punt functions as a type II receptors for dpp and propose that BMP signaling in vertebrates may also involve sharing of type II receptors by diverse ligands.

Inductive interactions between germ layers are an essential feature of the development of many organisms. In several species these interactions are mediated by members of the transforming growth factor-beta (TGF beta) family. In amphibians, different concentrations of activin can induce different types of mesoderm in the animal cap assay. In Drosophila, a member of the TGF beta family, decapentaplegic (dpp), acts as an inductive signal. Midway through embryogenesis, dpp is expressed in the visceral mesoderm, and enhances the expression of the homeotic gene labial in the underlying midgut endoderm. Earlier in development, however, dpp expression is limited to the dorsal ectoderm. At this stage in development, thickveins, a dpp receptor, is expressed in the mesoderm, and this suggests that ectodermal dpp might not only be required for development of dorsal ectoderm, but could also act inductively to mediate pattern formation in the underlying mesoderm. Here we show, by expressing dpp ectopically in the ectoderm and mesoderm and by examining dpp null mutant embryos, that dpp regulates expression of mesodermal genes.

The patterns of homeotic gene expression in the Drosophila midgut visceral mesoderm are instrumental in several morphogenetic events, including the formation of the gastric caeca and the positioning of the three midgut constrictions. We demonstrate that a potent regulator of homeotic gene expression in the visceral mesoderm is the secreted growth factor-like molecule encoded by the decapentaplegic (dpp) gene. Ectopic dpp in the visceral mesoderm caused changes in the gene expression of Sex combs reduced, Antennapedia, Ultrabithorax (Ubx), and abdominal-A (abd-A) and disrupted the formation of the gastric caeca and the first and third midgut constrictions. Ectopic dpp also induced expression of teashirt, wingless (wg) and the endogenous dpp gene in the visceral mesoderm and enhanced labial expression in the adjacent endoderm. The patterns of gene expression and the formation of the second midgut constriction in the presence of ectopic dpp are most consistent with a dpp-induced transformation of virtually the entire midgut to cell fates normally seen only in the parasegment (ps)7 and ps8 regions of the midgut. We conclude that dpp is a primary signal in maintaining Ubx expression in the visceral mesoderm in a pattern different from Ubx expression in the embryonic ectoderm and in providing a cell-cell communication mechanism by which Ubx expression influences gene expression across germlayers and across the ps7 to ps8 parasegment boundary in the visceral mesoderm.

Drosophila sequences at chromosomal positions 25D (Brk25D) and 43E (Brk43E) are similar to the TGF beta type I receptor serine/threonine kinases and are expressed broadly during embryogenesis. Brk25D binds dpp protein and bone morphogenetic protein 2 with high affinity. Mutations affecting Brk25D map to the gene thick veins and block the expression of two decapentaplegic-responsive (dpp-responsive) genes, dpp and labial, in the embryonic midgut. Defects in Brk25D receptor function combined with reduced expression of dpp ligand produce mutant phenotypes in the embryo and adult. Brk43E is the product of the gene saxophone, which also interacts with dpp. We conclude that dpp signaling in vivo is mediated by at least two receptors, Brk25D and Brk43E.

Reported assays of the bone morphogenetic proteins (BMPs) have not in general revealed specific functions for the different proteins, belying the specificity implied by the evolutionary conservation and distinct expression patterns of the genes encoding BMPs. We have used assays of developmental function to show that the two Drosophila homologues of the BMPs, decapentaplegic (dpp) and 60A, that both induce ectopic bone formation in mammalian assay systems, have distinct effects in Drosophila development. A binary expression system using the yeast transcriptional activator GAL4 directed identical patterns of tissue and temporally specific dpp and 60A expression. When dpp enhancer elements drove GAL4 expression, GAL4-responsive dpp transgenes rescued dpp mutant phenotypes, but GAL4-responsive 60A transgenes did not. Ectopic ectodermal expression of dpp during gastrulation respecified the dorsal/ventral pattern of the embryo. In contrast, ectopic 60A expression had no detectable effects on embryonic development but led to defects in adult structures or lethality during metamorphosis. Expression of 60A in cells expressing dpp did not interfere with dpp functions, indicating that dysfunctional heterodimers did not form at sufficient levels to inhibit dpp. These specific developmental responses in Drosophila indicate that in vivo functions of BMP-like factors can be more specific than indicated by the ectopic bone formation assays and that the Drosophila embryo provides an assay system sensitive to the structural differences that contribute to BMP specificity in vivo.

Patterns of decapentaplegic (dpp) transcripts derived from the intact gene were compared to the patterns of transcripts generated by partial dpp transgenes in Drosophila embryos. Sequences closest to the dpp coding regions, the dpp hin region, were sufficient to express lacZ-tagged mRNA in patterns indistinguishable from the patterns of endogenous dpp expression in the dorsal and terminal cells at the blastoderm stage, in the dorsal ectoderm during germ band elongation, and in narrow stripes of ectodermal cells along the dorsal edge of the ectoderm and at the boundary between the lateral and ventral neurogenic regions during germ band shortening. The latter pattern of expression responded to the segment polarity genes naked and wingless. However, these dpp sequences were not sufficient to drive lacZ-tagged mRNA expression in other cells normally expressing dpp, including cells in the gnathal segments, the clypeolabrum, the foregut, the midgut visceral mesoderm, and the hindgut. Two separate regulatory regions were found in the dpp hin region. A 479 bp region upstream of the promoter was necessary for the segmented pattern of expression in the lateral ectoderm and for expression in the midgut endoderm. Cis-acting elements in the 2 kbp second intron directed expression in the dorsal and terminal regions of the blastoderm, acted on a heterologous promoter, the P-element promoter, and responded to pattern information derived from the maternal effect dorsal/ventral patterning genes.

The Drosophila decapentaplegic (dpp) gene, a member of the transforming growth factor-beta family, is required for dorsal/ventral pattern formation and midgut and imaginal disk development. We have identified a 3-kb upstream regulatory region necessary for dpp expression in the visceral mesoderm of the gastric caeca primordia and a second 2.5-kb upstream regulatory region necessary for dpp expression in the midgut visceral mesoderm corresponding to a portion of abdominal segments 1 and 2 (parasegment 7). These regulatory regions act over a distance of up to 10-kb on all four of the dpp promoters examined. Absence of dpp expression in the gastric caeca primordia caused defective development of the gastric caeca and a concomitant partial reduction in larval and pupal viability. Absence of dpp expression in the visceral mesoderm of parasegment 7 caused a reduction in the length of the central portion of the larval gut and a change in the morphology of the midgut cells in this region but had little effect on the survival of the animals to the adult stage. However, a larval lethal phenotype was observed when both the central portion of the larval midgut and the gastric caeca were defective.

Both decapentaplegic (dpp) protein and 60A protein have been implicated in pattern formation during Drosophila melanogaster embryogenesis. Within the C-terminal domain, dpp and 60A are similar to human bone morphogenetic protein 2 (75% identity) and human osteogenic protein 1 (70% identity), respectively. Both recombinant human bone morphogenetic protein 2 and recombinant human osteogenic protein 1 have been shown to induce bone formation in vivo and to restore large diaphyseal segmental defects in various animal models. We examined whether the Drosophila proteins, dpp and 60A, have the capacity to induce bone formation in mammals by using the rat subcutaneous bone induction model. Highly purified recombinant dpp and 60A induced the formation of cartilage, bone, and bone marrow in mammals, as determined by histological observations and by measurements of the specific activity of alkaline phosphatase and calcium content of the implants, thereby demonstrating that related proteins from phylogenetically distant species are capable of inducing bone formation in mammals when placed in sites where progenitor cells are available.

The decapentaplegic (dpp) gene of Drosophila melanogaster is required for pattern formation in the embryo and for viability of the epithelial cells in the imaginal disks. The dpp protein product predicted from the DNA sequence is similar to members of a family of growth factors that includes transforming growth factor beta (TGF-beta). We have produced polyclonal antibodies to a recombinant dpp protein made in bacteria and used a metallothionein promoter to express a dpp cDNA in Drosophila S2 cells. Similar to other proteins in the TGF-beta family, the dpp protein produced by the Drosophila cells was proteolytically cleaved, and both portions of the protein were secreted from the cells. The amino-terminal 47-kilodalton (kDa) peptide was found in the medium and in the proteins adhering to the plastic petri dish. The carboxy-terminal peptide, the region with sequence similarity to the active ligand portion of TGF-beta, was found extracellularly as a 30-kDa homodimer. Most of the 30-kDa homodimer was in the S2 cell protein adsorbed onto the surface of the plastic dish. The dpp protein could be released into solution by increased salt concentration and nonionic detergent. Under these conditions, the amino-terminal and carboxy-terminal portions of dpp were not associated in a stable complex.