Patterning of the embryo — the process by which asymmetrically organized cell types and organs are generated from a single cell to make up an individual — is controlled by a surprisingly limited number of signalling pathways (for example, the Hedgehog, Wingless, FGF, TGF-β and EGFR pathways). What’s amazing is that patterning of structures as different as the limb or appendage and hair or bristles (for vertebrates or invertebrates, respectively), in species as different as the fruitfly Drosophila melanogaster and mammals, is controlled by the same conserved pathways. Specificity is achieved by the fine cross-regulation of these pathways by one another. Negative feedback loops have been described whereby activation of these signalling pathways leads to expression of negative regulators. These inhibitors may directly antagonize the signal, possibly by interfering with ligand binding to the receptor at the start of the signal cascade in the case of soluble inhibitors or membrane inhibitors, or by interfering with the downstream signalling cascade in the case of intracellular inhibitors. Alternatively, inhibitors may generate an antagonizing signal. But, until now, a negative feedback loop had not been described for the Wingless pathway.

Because the phenotype of Drosophila with a loss-of-function mutation in the naked cuticle (nkd) gene resembles that of wingless (wg) gain-of function mutants and wg transgenic animals, Nkd has been proposed to be an antagonist of Wg signalling. Genetic evidence also indicated that nkd might be a direct transcriptional repressor of Engrailed (a target of Wg), or might somehow influence Wg transport. Matthew Scott and colleagues (Nature 403, 789–795; 2000) now show that Nkd indeed opposes the effect of a Wg signal. The nkd mutation, which is lethal in the embryo and causes multiple segmental defects, has been described previously and named after the absence of denticles in the mutant Drosophila embryos. Scott and colleagues have now cloned the nkd gene, and show that its expression pattern parallels that of wg in the Drosophila embryo and larva, indicating that nkd may be a target for Wg. They confirm this by monitoring the expression of nkd in gain-of-function and loss-of-function wg mutants. Overexpression of nkd (right-hand panel in figure below) and decreased wg activity (middle panel) in the Drosophila larva produce similar adult phenotypes (such as the absence of wings in the middle and right-hand panels). Scott et al.’s results indicate that Nkd interferes with Wg signalling, which they also show to be the case for Wnt signalling in the frog Xenopus laevis. So, as is the case for Hedgehog signalling with Patched, EGFR signalling with Argos, Kekkon and Sprouty, and TGF-β signalling with Dad, Wg signalling now has its own inhibitor and its own negative feedback loop.

The question now will be how Nkd inhibits Wg signalling. The evidence so far is that it does not, at least initially, interfere with the expression or transport of Wg, and so it most likely interferes with the signalling cascade downstream of the Wg receptors, Frizzled and Dishevelled. But where in the cascade, and how, does it act? Nkd bears a region of homology to calcium-binding EF-hand domains, which might provide a hint to its mode of action. However, Nkd seems to lack other structural features of EF-hand-containing proteins such as EF-hand repeats or myristoylation sequences, so this clue might be misleading.