Wnt ligands are lipid-modified, secreted glycoproteins that control multiple steps during embryogenesis and adult-tissue homeostasis. Little is known about the mechanisms underlying Wnt secretion. Recently, Wntless (Wls/Evi/Srt) was identified as a conserved multi-pass transmembrane protein whose function seems to be dedicated to promoting the release of Wnts. Here, we describe Wls accumulation in the Golgi apparatus of Wnt/Wingless (Wg)-producing cells in Drosophila, and show that this localization is essential for Wg secretion. Moreover, Wls localization and levels critically depend on retromer, a conserved protein complex that mediates endosome-to-Golgi protein trafficking in yeast. In the absence of the retromer components Dvps35 or Dvps26, but in presence of Wg, Wls is degraded and Wg secretion impaired. Our results indicate that Wg, clathrin-mediated endocytosis and retromer sustain a Wls traffic loop from the Golgi to the plasma membrane and back to the Golgi, thereby enabling Wls to direct Wnt secretion.

Targeting of Wnt signaling represents a promising anti-cancer therapy. However, the consequences of systemically attenuating the Wnt pathway in an adult organism are unknown. Here, we globally prevent Wnt secretion by genetically ablating Wntless. We find that preventing Wnt signaling in the entire body causes mortality due to impaired intestinal homeostasis. This is caused by the loss of intestinal stem cells. Reconstitution of Wnt/β-catenin signaling via delivery of external Wnt ligands prolongs the survival of intestinal stem cells and reveals the essential role of extra-epithelial Wnt ligands for the renewal of the intestinal epithelium. Wnt2b is a key extra-epithelial Wnt ligand capable of promoting Wnt/β-catenin signaling and intestinal homeostasis. Wnt2b is secreted by subepithelial mesenchymal cells that co-express either Gli1 or Acta2. Subepithelial mesenchymal cells expressing high levels of Wnt2b are predominantly Gli1 positive.

Wnt proteins are members of a conserved family of secreted signaling ligands and play crucial roles during development and in tissue homeostasis. There is increasing evidence that aberrant Wnt production is an underlying cause of dysregulated Wnt signaling, however little is known about this process. One protein known to play a role in secretion is the transmembrane protein Wntless (Wls). However, the mechanism by which Wls promotes Wnt secretion is a riddle. It is not known which Wnt family members require Wls and what the structural requirements are that make some of them reliant on Wls for secretion. Here we present a systematic analysis of all known Drosophila Wnt family members with respect to their dependence on Wls function for secretion. We first show that the glycosylation status of Wg at conserved sites does not determine its dependence on Wls. Moreover, in apparent contrast to murine wls, Drosophila wls is not a target gene of canonical Wnt signaling. We then show that all Wnts, with the exception of WntD, require Wls for secretion. All Wnts, with the exception of WntD, also contain a conserved Serine residue (in Wg S239), which we show to be essential for their functional and physical interaction with Wls. Finally, all Wnts, with the exception of WntD, require the acyltransferase Porcupine for activity and for functionally interacting with Wls. Together, these findings indicate that Por-mediated lipidation of the S239-equivalent residue is essential for the interaction with, and secretion by, Wls.

Thiopurines are a standard treatment for childhood leukemia, but like all chemotherapeutics, their use is limited by inherent or acquired resistance in patients. Recently, the nucleoside diphosphate hydrolase NUDT15 has received attention on the basis of its ability to hydrolyze the thiopurine effector metabolites 6-thio-deoxyGTP (6-thio-dGTP) and 6-thio-GTP, thereby limiting the efficacy of thiopurines. In particular, increasing evidence suggests an association between the NUDT15 missense variant, R139C, and thiopurine sensitivity. In this study, we elucidated the role of NUDT15 and NUDT15 R139C in thiopurine metabolism. In vitro and cellular results argued that 6-thio-dGTP and 6-thio-GTP are favored substrates for NUDT15, a finding supported by a crystallographic determination of NUDT15 in complex with 6-thio-GMP. We found that NUDT15 R139C mutation did not affect enzymatic activity but instead negatively influenced protein stability, likely due to a loss of supportive intramolecular bonds that caused rapid proteasomal degradation in cells. Mechanistic investigations in cells indicated that NUDT15 ablation potentiated induction of the DNA damage checkpoint and cancer cell death by 6-thioguanine. Taken together, our results defined how NUDT15 limits thiopurine efficacy and how genetic ablation via the R139C missense mutation confers sensitivity to thiopurine treatment in patients. Cancer Res; 76(18); 5501-11. Ó2016 AACR.