Interactive Fly, Drosophila

Wingless and the patterning of the dorsal mesothorax

On each half of the dorsal mesothorax (heminotum), 11 large bristles (macrochaetae) occupy precisely constant positions. The location of each macrochaeta is specified during the third instar larval and early pupal stages by the emergence of its precursor cell (sensory mother cell: SMC) at a precise position in the imaginal wing discs, the precursors of the epidermis of most of the mesothorax and wings. The accurate positioning of SMCs is thought to be the culmination of a multistep process in which positional information is gradually refined. The GATA family transcription factor Pannier and the Wnt secreted protein Wingless are known to be important for the patterning of the notum. Thus, both proteins are necessary for the development of the dorsocentral mechanosensory bristles. Pannier has been shown to directly activate the proneural genes achaete and scute by binding to the enhancer responsible for the expression of these genes in the dorsocentral proneural cluster. Moreover, the boundary of the expression domain of Pannier appears to delimit the proneural cluster laterally, while antagonism of Pannier function by U-shaped, a Zn-finger protein, sets its limit dorsally. Therefore, Pannier and U-shaped provide positional information for the patterning of the dorsocentral cluster. In contrast and contrary to previous suggestions, Wingless does not play a similar role, since the levels and vectorial orientation of its concentration gradient in the dorsocentral area can be greatly modified without affecting the position of the dorsocentral cluster. Thus, Wingless has only a permissive role on dorsocentral achaete-scute expression. Evidence is provided indicating that Pannier and U-shaped are main effectors of the regulation of wingless expression in the presumptive notum (Garcia-Garcia, 1999).

An enhancer that directs expression specifically at the DC proneural cluster is present within a 5.7 kb fragment of AS-C DNA. Different subfragments were assayed for enhancer activity in vivo. A 1.4 kb subfragment (AS1.4DC) directs lacZ transcription from a minimal hsp70 promoter in the DC proneural cluster: beta-galactosidase and Scute endogenous accumulations precisely colocalized at this cluster. This fragment and the corresponding region of the AS-C from D. virilis were sequenced. Stretches of conserved DNA were present throughout the fragment, although they appeared to cluster within three regions. Subfragments containing each one of these regions were assayed for DC enhancer activity. Only the most 3' subfragment (PB0.5DC) shows such an activity, but to a much lesser extent than AS1.4DC. Interestingly, the activity is usually limited to only one cell, which is the posterior DC SMC. However, when assayed with the sc promoter, the PB0.5DC fragment directs lacZ activity in most cells of the DC cluster. Consequently, the sequences essential for specifying transcription in the DC cluster are contained within the PB0.5DC subfragment, although additional sequences that reinforce this expression are present in the larger AS1.4DC fragment. The AS1.4DC fragment was used to study DC enhancer activity (Garcia-Garcia, 1999).

The Pnr protein, which is a GATA-1 transcription factor, is known to regulate ac-sc expression at the DC cluster by acting directly or indirectly through the DC enhancer. The sequence of AS1.4DC was examined: within it, seven putative GATA-1 factor binding sites were found. Three of them fit the vertebrate consensus sequence (WGATAR: sites 1, 2 and 4); three comply with the consensus obtained in a random oligonucleotide selection experiment performed with Pnr protein (GATAAG: sites 3, 5 and 6), and one fits both consensus sequences (site 7). Interestingly, sites 5, 6 and 7 are within the PBO.5DC subfragment and two of them are conserved in D. virilis. Site-directed mutagenesis of site 7 strongly decreases enhancer activity of the AS1.4DC-lacZ construct (abbreviated DC-lacZ). Additional mutagenesis of other sites displaying the vertebrate consensus does not further reduce the residual activity. However, mutagenesis of all seven sites completely abolishes activity. These data suggest that Pnr interacts with some of these sites and that this interaction is essential for DC-lacZ activity. The capacity of Pnr to activate transcription of an AdhCAT reporter gene linked to either the complete AS1.4DC enhancer fragment or to each of the three subfragments was tested in transfection assays performed in chicken embryonic fibroblast (CEF) cells. Pnr stimulates transcription to similar levels from the complete enhancer and from subfragment PB0.5DC. In contrast, no stimulation was detected with the other subfragments. Notably, PB0.5DC displays DC enhancer activity in flies and contains three putative Pnr-binding GATA sites. Mutagenesis of only one of these (site 7) does not affect AdhCAT activity. But simultaneous removal of two sites (either sites 5 and 7, or 6 and 7) strongly impairs activity and mutagenesis of all three sites essentially abolished it. This suggests that a minimum of two GATA sites are necessary for transcriptional activation. Further evidence for a direct interaction of Pnr with the GATA sites of the enhancer was obtained in electrophoretic mobility-shift assays (EMSA) conducted with two different GST-Pnr fusion proteins that included the DNA binding domain of Pnr. Additonally, it has been shown that although relatively high levels of Wg protein are necessary for full DC-lacZ activity, the precise levels of this protein and the orientation of its gradient do not convey information for the position and the shape of the DC cluster (Garcia-Garcia, 1999).

In the prospective notum, the stripe of diffusible Wg protein straddles the lateral border of the domain of expression of pnr. This is compatible with the location of the Wg source being on the border of, but still within, the pnr domain. In accordance with this location, pnr appears to activate wg, since it has been found that a wg-lacZ construct, which reproduces the notal band of Wg accumulation, is not expressed in pnr mutant discs and is ectopically expressed in the dorsalmost area of the disc in a pnr dominant gain-of-function combination. In contrast, other data suggest that Pnr represses wg. Thus, the notal wg stripe is expanded dorsally in strong hypomorphic pnr combinations. Moreover, in flies in which pnr is overexpressed there was no expansion of the domain of WG mRNA, which in fact accumulates in a stripe that is even narrower than that seen in the wild type. The repressing effects appeared to be restricted to the domain of accumulation of Ush, which suggests the participation of Pnr/Ush heterodimers in the repression. Consistent with this assumption, the Pnr^D1 mutant protein, which is incapable of interacting with Ush, promotes wg expression within the entire dorsalmost area of the disc in pnr mutants animals. Interestingly, Pnr D1 can not induce the expansion of the wg expression domain in the presence of wild-type Pnr, suggesting that Pnr+/Ush heterodimers interfere with the Ush-resistant function of Pnr^D1. Such interference may also account for the repression of the Pnr^D1-mediated dorsal expansion of DC-lacZ expression by Pnr+. Taken together, these results suggest that during development of the wing disc, Pnr is necessary both for activation of wg and (together with Ush) for its repression in the dorsalmost region of the presumptive notum. This dorsal repression probably takes place from the start of wg expression, since the earliest detectable accumulation of WG mRNA is already restricted to the presumptive mid notal region. A wg-lacZ enhancer trap line, which shows expression throughout the dorsalmost part of the early third instar wing discs and posterior refinement to the notal stripe, might have a reduced sensitivity to the repression by Pnr/Ush (Garcia-Garcia, 1999).

A model is provided for the dorsal-lateral patterning of the DC area by Pnr and Ush. In the third instar wing disc and in the dorsalmost part of the prospective notum, Ush is present at high concentrations and the Pnr/Ush heterodimers are relatively abundant. These heterodimers would act as repressors and prevent activation of downstream genes. In the DC area, defined along the dorso-lateral axis by lower concentrations of Ush and the presence of Pnr, there is sufficient free Pnr to activate genes like ac-sc, DC-lacZ and wg. ac-sc is transcribed in the more dorsal part of the area because its activation requires relatively high concentrations of Pnr. wg is only transcribed at the edge of the Pnr domain because its expression is very sensitive to both Pnr and Pnr/Ush, and consequently low concentrations of the former are sufficient for activation and low concentrations of the latter, even in the presence of high concentrations of free Pnr, impose repression. The inability of extra doses of the activator Pnr to revert the repression by Pnr/Ush in the dorsalmost region of the notum suggests that activator and repressor do not compete for overlapping sites at the DC as-sc and notal wg enhancers. The presence of Pnr/Ush at their site(s) would block the activating effect of bound Pnr. Additional inputs, notably decapentaplegic, are known to act on the DC enhancer (Garcia-Garcia, 1999).

Wingless function during dorsal closure

A genetic system has been developed based upon the hobo transposable element in Drosophila melanogaster. hobo, like the better-known P element, is capable of local transposition. A hobo enhancer trap vector has been mobilized and two unique alleles of decapentaplegic (dpp) have been generated . A detailed study of one of those alleles (dpp^F11) is reported. This is the first application of the hobo genetic system to understanding developmental processes. LacZ expression from the dpp^F11 enhancer trap accurately reflects dpp mRNA accumulation in leading edge cells of the dorsal ectoderm. Combinatorial signaling by the Wingless (Wg) pathway, the Dpp pathway, and the transcriptional coactivator Nejire (CBP/p300) regulates dpp^F11 expression in these cells. This analysis of dpp^F11 suggests a model for the integration of Wg and Dpp signals that may be applicable to other developmental systems. This analysis also illustrates several new features of the hobo genetic system and highlights the value of hobo, as an alternative to P, in addressing developmental questions (Newfeld, 2002).

During early stages of embryogenesis, wg and dpp are expressed in undifferentiated dorsal ectoderm. wg mRNA expression, in 15 stripes along the entire dorsal-ventral axis of the embryo (including the dorsal ectoderm), begins at stage 8. wg expression persists in this striped pattern through stage 17. dpp mRNA is expressed on the dorsal side of the embryo along the entire anterior-posterior axis, beginning at stage 4. dpp mRNA expression persists in a large portion of the dorsal ectoderm through stage 8 and resolves into leading edge cell-specific expression in stage 12 embryos. The embryonic expression pattern of nej has not been reported. However, some information can be inferred from nej mutant phenotypes. nej zygotic mutant embryos show visible defects in the tracheal system at stage 12. The tracheal system is derived from the dorsal ectoderm, suggesting that nej is expressed in this tissue prior to stage 12 (Newfeld, 2002).

Analysis of dpp^F11 suggests that dpp expression in leading edge cells is initiated by prior episodes of wg and dpp expression in the undifferentiated dorsal ectoderm. The maintenance of dpp expression in leading edge cells appears to require continuous input from wg and from a dpp feedback loop. The initiation and maintenance of dpp expression in leading edge cells also require continuous nej activity. Overall, these data are consistent with the following combinatorial signaling model: Med (signaling for the Dpp pathway) interacts with Arm (signaling for the Wg pathway) via the transcriptional coactivator Nej. This multimeric complex initiates and, with continuous signaling, maintains dpp expression in leading edge cells (Newfeld, 2002).

These data extend previous studies of dpp expression in leading edge cells and Dpp signaling in several ways. A role for Wg signaling in the regulation of dpp expression in the leading edge has been suggested. dpp leading edge expression is not maintained in arm² zygotic mutants and does not initiate in arm² germline clones. nej and Med are involved in the regulation of dpp expression in leading edge cells. While nej³ enhances dpp wing phenotypes, Med¹ enhances nej³ embryonic phenotypes. This study suggests a role for nej in mediating combinatorial signaling by the Wg and Dpp pathways (Newfeld, 2002).

Several questions remain about the combinatorial regulation of dpp expression by Wg, Dpp, and Nej. One question is, how is Nej recruited to bridge the two pathways? Numerous studies have shown that p300/CBP transcriptional coactivation functions are stimulated by its phosphorylation but the site of phosphorylation has never been mapped. Perhaps Zeste white3 (a serine-threonine kinase in the Wg pathway) or Thickveins (a serine-threonine kinase in the Dpp pathway) are involved in recruiting Nej to participate in combinatorial signaling (Newfeld, 2002).

A second question concerns the nature of the enhancer element that directs dpp expression in leading edge cells. Using reporter genes, a 54-nucleotide candidate enhancer has been identified near the dpp^F11 transgene insertion that drives lacZ expression in a subset of leading edge cells. The region contains two sets of conserved, overlapping consensus-binding sites for dTCF (a transcriptional partner for Arm in the Wg pathway) and Mad (a transcriptional partner for Med in the Dpp pathway). No JNK-pathway-binding sites are in the region, suggesting that JNK regulation of dpp expression in leading edge cells is independent of Wg and Dpp signaling (Newfeld, 2002).

Interestingly, there is also a consensus Brinker (Brk) binding site in the candidate enhancer. Brk is a transcriptional repressor of Dpp target genes and one mechanism by which Dpp signaling activates its target genes is to relieve Brk repression. The genetic data cannot discriminate between the possibility that combinatorial signaling by the Wg and Dpp pathways regulates dpp expression in leading edge cells by direct activation or by relief of Brk repression (Newfeld, 2002).

Table of contents

The Interactive Fly resides on the
Society for Developmental Biology's Web server.