Abstract
The wild-type functions of the Wnt family of genes are still little understood (for review see Nusse and Varmus, Cell 69, 1073-1087, 1992). In Drosophila, the wingless (D-Wnt-1) protein is expressed in segmental stripes: its absence leads to a complete failure of segmentation, loss of engrailed expression and lack of pattern in the cuticle. A predominating hypothesis is that the spatial distribution of wingless is crucial to pattern; it might carry an instructive signal from cells that secrete the protein to cells nearby, or it might form a concentration gradient which acts as a morphogen. We tested these hypotheses by expressing wingless ubiquitously in wingless- embryos. The distribution of wingless protein in these embryos is uniform. Despite this, engrailed expression persists, is confined to the most anterior third of the parasegment, and delineates the parasegment border. The cuticle shows a segmentally reiterated pattern and, dorsally, the denticles are normally distributed and oriented. Because all these position-specific features cannot have been placed by a local source or a differential distribution of wingless protein, we conclude that, in the early embryo, the role of wingless is neither to act as a local instructive signal, nor as a morphogen. We propose an alternative hypothesis that the wild-type function of the wingless protein is to maintain and 'seal' the parasegment borders; in its absence the borders fail to isolate abutting segmental gradients.
