The landmark Drosophila melanogaster genetic screen carried out by Christiane Nüsslein-Volhard and Eric Wieschaus in 1980 (Nature 287: 795–801) identified many mutants with defects in the embryonic cuticle (a secretion product of the epidermis). Twenty-five years later we are still trying to clone some of these genes and determine their functions. New work in Nature by Elisabeth Knust and colleagues (Nature, 414: 638–642 2001), and Yuh-Nung Jan and colleagues (Nature, 414: 634–637 2001), solves the mystery of one of these genes, stardust.

stardust mutants have severe embryonic defects similar to both crumbs and shotgun mutants (Crumbs encodes a transmembrane protein essential for establishing epithelial polarity, and Shotgun encodes DE-Cadherin, which is a component of the apical adherens junction). The mutant embryos in all three cases are literally blown to pieces (see picture). Until recently, however, the molecular nature of stardust was unknown. In the December 6th issue of Nature two groups show that the stardust gene encodes a MAGUK (membrane associated guanylate kinase) protein. MAGUK proteins contain one or several PDZ domains, one SH3 (Src homology) domain and one guanylate kinase domain. Jan and colleagues identified two MAGUK sdt isoforms, whereas Knust and colleagues identified one of these and an additional sdt isoform that only contains the guanylate kinase domain.

As crumbs (crb) mutants look similar to stardust (sdt) mutants, both groups went on to see whether the two proteins interact. Sdt does indeed bind to the carboxyl terminus of Crb. In addition, loss of either sdt or crb prevents formation of the SAC (subapical complex), a process essential for the formation of apical adherens junctions (a process that also includes the recruitment of E-Cadherin) and leads to a loss of epithelial polarity.

Interestingly, neuroblast polarity is unaffected. Neuroblasts form by delaminating from the epithelial layer; once they have separated from this layer, however, neither sdt or crb are needed to ensure their polarity.

One by one the pieces that will allow us to understand how epithelial polarity is generated are falling into place. It seems that Crb recruits Sdt into a complex (or vice versa), which is essential for epithelial cells to establish their initial polarity. As the molecular function of Sdt and its partners becomes clearer, we will be able to determine how this complex affects the recruitment of E-Cadherin and hence cell adhesion.