Table of contents

Genetic structure of the wingless locus

In Drosophila, there are cases in which gene products contributing to the same developmental event may derive from closely adjacent transcription units and may even share cis-regulatory sequences. Correct recognition of such genomic organization is central to an understanding of developmental mechanisms. The adult phenotypes of combinations between the mutations spade, Sternopleural, and wingless suggest that they are lesions in functionally related genes within the same chromosomal region. wingless mutations fail to complement the recessive mutation spade. The spade mutation behaves as a lesion in a regulatory site of wingless, sited 5' to the transcription unit, and is concerned with particular postembryonic functions of wingless. The spd mutation affects wing patterning by reducing wg transcription in the hinge and at the wing margin and subsequently by diminishing cell proliferation at the wing hinge (Buratovich, 1997).

While showing wingless-like phenotypes in combination with Sternopleural, even lethal alleles of wingless complement the recessive lethality of Sternopleural alleles. Mutations in Sternopleural increase the severity of wingless phenotypes in many wingless-dependent processes during postembryonic development; this interaction can occur when the only functional copies of Sp or wg are located in either opposing chromosomes or on the same chromosome. This is inconsistent with previous attempts to define Sp as a regulatory allele of wg and attempts to explain the phenotypes that result from combinations of Sp and wg by means of transvection. A new EMS-induced allele of Sternopleural is more severe than the original allele, which also argues for Sp being a separate, mutable genetic locus rather than a regulatory allele of wg. A revertant of Sternopleural (Sp[Rv1]) behaves as a genetic null allele of wg, but causes ventral-to-dorsal transformations in combination with wg(P), which is not observed in combinations of wg null alleles with wg(P). Because wg(P) is the result of an inversion and because inversions inhibit transvection, the increased severity observed in Sp(Rv1)/wg(P) in comparison to wg(null)/Sp(Rv1) animals cannot be explained by an absence of transvection. Therefore, the two Sternopleural mutations most reasonably define an independent gene located 3' to the wingless gene and having strong functional synergism with it (Buratovich, 1997).

Functional analysis of Wingless reveals a link between intercellular ligand transport and dorsal-cell-specific signaling

How is a molecule that associates tightly with membranes and extracellular matrix able to promote direct response in cells at a distance from the source of gene product? Various observations suggest that Wg protein is actively moved between and across cells in a process that requires endocytic components. This form of transcytotic movement is directed laterally in the plane of the polarized epidermal epithelium and therefore differs from apical-basal transcytosis across polarized epithelia. Wg protein is produced in one row of cells near the posterior of every segment and is secreted and distributed throughout the segment to generate wild-type pattern elements. Ventrally, epidermal cells secrete a diverse array of anterior denticle types and a posterior expanse of naked cuticle; dorsally, a stereotyped pattern of fine hairs is secreted. Three new wg alleles are described that promote naked cuticle cell fate but show reduced denticle diversity and dorsal patterning. These mutations cause single amino acid substitutions in a cluster of residues that are highly conserved throughout the Wnt family. Each of the three lesions represents a single amino acid substitution within the coding sequence. All three residues affected are highly conserved throughout the Wnt family; alanine 136 and cysteine 242 are absolutely invariant in Wnts from C. elegans to humans, while glycine 258 is conserved among most members of the family. By manipulating expression of transgenic proteins, it is demonstrated that all three mutant molecules retain the intrinsic capacity to signal ventrally but fail to be distributed across the segment. Thus, movement of Wg protein through the epidermal epithelium is essential for proper ventral denticle specification and this planar movement is distinct from the apical-basal transcytosis previously described in polarized epithelia. Furthermore, ectopic overexpression of the mutant proteins fails to rescue dorsal pattern elements. Thus a region of Wingless has been identified that is required for both the transcytotic process and signal transduction in dorsal cell populations, revealing an unexpected link between these two aspects of Wg function (Dierick, 1998).

In late-stage Drosophila embryos, epidermal cells secrete a segmentally repeating pattern of cuticular structures: six rows of uniquely shaped, hook-like denticles interspersed with naked cuticle on the ventral surface and fine hairs on the dorsal surface. Wg signaling during earlier stages is required for correct specification of these cuticular pattern elements. wg null mutants secrete no ventral naked cuticle and instead produce denticles consisting primarily of a single morphology; this denticle type resembles the large refractile denticles found in the fifth row of the wild-type denticle belt. wg mutant denticles are arranged in a segmentally repeating pattern of reversed polarity. In contrast, on the dorsal surface, no segmental pattern is observed and all cells secrete a single type of hair. Ectopic Wg restores belts of diverse denticle types when provided at low level in a wg null mutant, and produces uniform naked cuticle when provided at high level. Thus the ventral denticle-secreting cells, which lie at the greatest distance from the wg-expressing row of cells, may serve as indicators of long-range Wg protein transport, while the naked cuticle region, which is roughly centered over the wg-expressing row of cells, serves as a marker for short-range, high-level Wg activity. Alternatively, it is possible that denticle diversity is generated indirectly by a short-range relay mechanism triggered by Wg. However, experiments with a temperature-sensitive wg allele have shown that denticle diversity is specified by Wg activity at early stages of development, making a multiple step process less likely. Three novel alleles of wg secrete some ventral naked cuticle and alter denticle diversity to different degrees, but dorsally they show little or no Wg-mediated patterning. Two alleles, wgPE6 and wgNE1, are temperature sensitive. At 25 degrees C, mutant embryos display more segmentation than a wg loss-of-function mutant and produce denticle morphologies other than row-5 type, but only at a low frequency. The wgNE2 allele produces the same pattern at both 18 degrees C and 25 degrees C: mutant embryos secrete patches of naked cuticle and belts containing predominantly row-5-type denticles, with a low frequency of other denticle morphologies (Dierick, 1998).

The restricted cuticular patterning of the mutant embryos is mirrored by molecular markers of Wg activity. wgNE2 mutants, and in wgPE6 and wgNE1 mutants at 18 degrees C, produce protein that accumulates at high level in the wg-expressing cells and shows little accumulation in neighboring cells. Arm protein, which normally accumulates in broad stripes centered over the wg-expressing row of cells shows no accumulation when examined in mutant flies. Thus, as in shabire mutants, which are mutant for a protein involved in exocytosis, the range of cells over which Wg can signal is reduced in flies bearing the three novel alleles. Expression of the Wg-responsive gene engrailed also provides an assay for Wg signal transduction in neighboring, non-wg- expressing cells. Wg activity is required for the maintenance of en expression in the adjacent posterior two rows of cells; in the absence of wg gene function, epidermal en expression is initiated normally but decays during stages 9 and 10. In mutant flies, en is expressed normally at early stages but, by stage 11, no epidermal en expression is reproducibly detected. In wgNE2 mutants, and in wgPE6 and wgNE1 at 18 degrees C en expression decays in a defined pattern: stabilized en expression can be seen only in the most ventral cells of the en stripe, with little or no stabilization in dorsal and dorsolateral cells. Stabilization of en expression in neighboring cells indicates that the mutant molecules can move through the secretory pathway to the cell surface. The autocrine function of Wg is not disrupted significantly by these mutations. Wg activity is required for wg autoregulation; in the absence of wg gene function, wg expression decays by stage 10. All three partial function alleles, at either temperature, express WG mRNA beyond stage 11. In contrast, the paracrine Wg response, as measured by stabilization of later en expression, is eliminated in dorsal cell populations and is limited in ventral populations to those cells that directly contact the wg-expressing cells. Thus these mutational changes appear to perturb those activities of Wg that require transit between or through cells (Dierick, 1998).

Confocal microscopy was used to compare the subcellular localization of the mutant Wg proteins with that of the wild-type Wg protein. Wg protein normally achieves a punctate distribution over several cell diameters on either side of the wg-expressing cells. Vesicular staining can be detected in both apical and basal planes of focus within the epidermal cell layer. A comparable view of a wgPE6 homozygote at 25 degrees C reveals that the mutant protein accumulates at high levels in the wg-expressing row of cells, with no punctate staining detectable in neighboring cells. This protein distribution is consistent with the severely restricted signaling activity observed in mutant embryos at 25 degrees C. A superficially similar distribution is observed in wgIL114 mutant embryos at restrictive temperature and in porcupine mutant embryos, which appear to be defective in export of Wg protein. However, in these situations, no Wg signaling activity can be detected, whereas wgPE6 homozygotes at 25 degrees C clearly retain some Wg function, as measured by cuticle pattern and autocrine effects on wg expression. In wg PE6 homozygotes at 18 degrees C, Wg protein can be detected over a broader domain of cells, encompassing the wg-expressing row and its immediate neighbors. In contrast to the vesicular appearance of wild-type Wg, the wg PE6 mutant protein appears to accumulate preferentially around the cell membranes. This is particularly obvious in the more basal plane of focus, where high levels of Wg staining surround the basolateral cell membranes in both wg-expressing and non-wg-expressing cells. However, punctate staining is also detected within these cells, suggesting that some protein can be internalized properly in vesicles (Dierick, 1998).

Recent work has implicated the putative Wg receptor, Dfrizzled2 (Dfz2), as a potential candidate for a cell surface molecule involved somehow in Wg transport. Overexpression of Dfz2 in the wing imaginal disc produces a broader distribution of the endogenous Wg protein, enhancing its activity in cells at a distance from the wg-expressing domain of cells. This contrasts with other growth factor receptors, which sequester ligand when overexpressed. Although Dfz2 can act as an extracellular chaperone for Wg in the context of the wing disc, it does not appear to play an analogous role in the embryonic epidermis. Embryos overexpressing Dfz2 hatch into viable larvae with cuticle patterns that do not display Wg hyperactivity. Another class of potential extracellular chaperone are proteoglycans: glycosaminoglycans appear to be essential for Wg signal transduction. The extracellular sugar groups are thought to increase the local concentration of Wg ligand and thereby improve signaling efficiency. Overexpressing Wg can bypass the requirement and rescue pattern defects caused by mutations in glycosaminoglycan biosynthesis. In ventral epidermal cells, the three novel mutant molecules described here show substantial signaling activity when ectopically expressed. This supports the idea that their primary defect is impaired transport. In dorsal cell populations, however, none of the three mutant molecules are able to trigger normal molecular and morphological responses to Wg even at high levels of expression. Thus there appears to be a link between Wg protein transport and dorsal Wg signal transduction: each of three amino acid substitutions alters both aspects of Wg function simultaneously. It is formally possible that these defects represent altered ligand interaction with two independent sets of molecules. However, for reasons of parsimony, it is proposed that each mutation decreases binding affinity for a single extracellular chaperone, which is exclusive to the transport machinery in ventral epidermal cells but is shared with signal transduction machinery in dorsal cells. A corollary of this model is that the ventral and dorsal signaling receptors must be distinct, since ventrally the proposed chaperone is not essential for signal transduction. Furthermore, the autocrine function of Wg in maintaining its own expression may not require this chaperone, as dorsal wg autoregulation is not disrupted by the mutant lesions. This model would also explain a previously observed temporal link between Wg specification of dorsal pattern elements and ventral denticle diversity. Temperature-shift experiments with the wg IL114 allele show that denticle diversity and all dorsal pattern are generated by Wg signaling activity during stages 9 and 10, while Wg activity during stages 11 and 12 directs cells to secrete naked cuticle. Embryos that are shifted down to permissive temperature at stage 11 show small expanses of naked cuticle in a lawn of uniform denticles, with no detectable rescue of dorsal pattern elements; this pattern is very similar to that of the partially functional wg mutants described here. These observations could be explained if the proposed extracellular chaperone is present only at early stages of development. Late stage restoration of Wg function then would be unable to rescue either aspect of the pattern: ventrally, the restored Wg protein would not be transported properly to perform long-range diversity generation, and dorsally it would not signal efficiently. During wild-type embryonic patterning, such temporal changes in a chaperone might contribute to the restriction of Wg activity in later stages of development. This would favor accumulation of high-level Wg to promote naked cuticle cell fate close to the wg-expressing row and would protect denticle cell fates from respecification to naked cuticle by inappropriate Wg activity (Dierick, 1998 and references).

Table of contents

wingless continued: Biological Overview | Evolutionary Homologs | Transcriptional regulation |Targets of Activity | Protein Interactions | mRNA Transport | Developmental Biology | References

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

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