A large number of developmental processes in organisms ranging from worms to humans depend on signaling by the Wnt family of secreted proteins. More recent studies have also implicated Wnt signaling pathways in the control of stem cell maintenance in the adult, and dysregulated Wnt activity results in a variety of human cancers.1,2 The Drosophila gene wingless (Wg) is a founding member of the Wnt family, and its activities in patterning the embryonic fly epidermis and developing wing have been studied extensively.3
Wnts belong to a class of developmentally important molecules known as morphogens: secreted signaling proteins that display a gradient of activity. Responding cells adopt different fates based on the concentration of morphogen (generally a function of their distance from the source of the signal). As such, factors that influence the secretion and mobility of morphogens such as Wnts are clearly of major importance. Yet to date, most studies of Wnt signaling have focused on the proteins mediating cellular responses to Wnt ligands (such as Frizzled and LRP receptors, intracellular proteins Axin, APC, Dishevelled, GSK-3 beta and beta-catenin, and the TCF/LEF family of transcription factors).
Several recent advances have begun to shed light on the mechanisms that determine the amount and activity of secreted Wnts available to target cells. For example, Wnts were recently shown to undergo a lipid modification that is essential to their activity.4 Palmitate is added at a conserved cysteine residue, probably by the product of the porcupine gene.5 Heparan sulfate proteoglycans (HSPGs) have been implicated in the movement of Wnts through tissues by stabilizing the proteins or otherwise influencing their transport from cell to cell.1,6
|Figure 1. Factors influencing Wnt signaling. In addition to the signal transduction cascade that occurs in Wnt responding cells (shown in cell at right), recent work has identified factors involved in regulating Wnt secretion and transport. In particular, Wls/Evi activity is required for release of Wnt family proteins from cells that produce them (shown in cell at far left). In the absence of Wls/Evi function, Wnt ligands are retained in Wnt producing cells and Wnt signaling fails to occur.|
Two recent articles in the journal Cell address another essential variable in Wnt signaling: the release of Wnt proteins from cells that produce them.7,8 Using genetic screens in Drosophila designed to identify novel Wnt pathway elements, both reports identified the same molecule, named Wntless (Wls)7 or Evi (evenness interrupted) (Figure 1).8 Wls/Evi is a novel, seven pass transmembrane protein with orthologs in both worm and human. Loss of Wls/Evi function blocks Wnt dependent processes, both in fly, where it produces patterning defects in the embryo, and in human cultured embryonic kidney (HEK) cells. Banziger et al. used siRNA to deplete endogenous Wls/Evi in HEK cells genetically modified to produce Wnt3a or to respond to Wnt signals via a TCF/LEF controlled reporter gene. Bartscherer et al. performed analogous experiments using Wnt1. Mixing of the Wnt producing and responding (reporter) cell populations demonstrated that Wls/Evi activity is required in cells that produce Wnt, but not in cells responding to Wnt. Both groups also found that absence of Wls/Evi led to increased levels of Wnt ligand in the mutant cells, implying either increased transcription or impaired release of Wg/Wnt. Because the distribution of ligand was shifted from outside the cells to inside, the authors concluded that Wg/Wnt ligands are aberrantly retained in cells in the absence of functional Wls/Evi. Similarly, Wnt ligand was found in the culture medium of Wnt3a transfected HEK cells, but was not seen after siRNA knockdown of Wls/Evi.7
To exclude the possibility that Wls/Evi might be required generally for secretion, both studies confirmed that secretion of unrelated ligands, such as hedgehog, were not affected in Wls/Evi mutants. Banziger et al. further examined specificity among Wnt family ligands. All Wnts tested, including Wnt3a, Wnt1, and Wnt5a, required Wls/Evi function.
The exact role of Wls/Evi in the release of Wnt ligands from cells remains uncertain. Banziger et al. detected tagged Wnt protein in the Golgi apparatus and in vesicles between the Golgi and the plasma membrane in cells lacking Wls/Evi function, whereas Bartscherer et al. noted excess Wnt associated with the cell surface. Importantly, the secretion defect seen by both groups precluded analysis of whether Wls/Evi might also affect activity of Wnt proteins, an interesting question that will need to be addressed in future studies.