PerspectiveWnt signaling

Context-Dependent Activation or Inhibition of Wnt-β-Catenin Signaling by Kremen

See allHide authors and affiliations

Science Signaling  26 Feb 2008:
Vol. 1, Issue 8, pp. pe10
DOI: 10.1126/stke.18pe10


Wnt-β-catenin signaling controls critical events in metazoan development, and its dysregulation leads to cancers and developmental disorders. Binding of a Wnt ligand to its transmembrane co-receptors Frizzled (Fz) and low-density lipoprotein (LDL) receptor–related protein (LRP) 5 or LRP6 inhibits the degradation of the transcriptional coactivator β-catenin, which translocates to the nucleus to regulate gene expression. The secreted protein Dickkopf1 (Dkk1) inhibits Wnt signaling by binding to LRP5 and LRP6 and blocking their interaction with Wnt and Fz. Kremen 1 and 2 (Krm1 and 2, collectively termed Krms) are single-pass transmembrane Dkk1 receptors that synergize with Dkk1 to inhibit Wnt signaling by promoting the endocytosis of LRP5 and LRP6. A study now suggests that Krms, in the absence of Dkk1, potentiate Wnt signaling by maintaining LRP5 and LRP6 at the plasma membrane. It is proposed that the absence or presence of Dkk1 determines whether Krms will activate or inhibit Wnt-β-catenin signaling, respectively. Here, we speculate that the proposed context-dependent positive and negative roles for Krms could promote biphasic Wnt signaling in response to a shallow gradient of Dkk1, resulting in the generation of precise and robust borders between cells during development. Identification of a context-dependent role for Krms in Wnt-β-catenin signaling offers insight into the mechanism of Wnt signaling and has important developmental implications.

Wnt signaling specifies many cell and tissue fates in metazoan development, and dysregulation of this pathway leads to cancers and developmental disorders [reviewed in (1)]. The best-characterized form of Wnt signaling is the Wnt-β-catenin, or canonical Wnt, pathway. During Wnt-β-catenin signaling, a Wnt ligand binds to the transmembrane co-receptors Frizzled (Fz) and low-density lipoprotein receptor-related protein (LRP) 5 or LRP6 and initiates a process that leads to the stabilization and nuclear translocation of β-catenin. In the nucleus, β-catenin binds to transcription factors of the T cell factor/lymphoid enhancer factor (TCF/LEF) family and activates a Wnt-β-catenin transcriptional program. Here, we discuss the significance of a study by Hassler et al. in which the authors propose that Kremen 2 (Krm2) is a context-dependent regulator of LRP5 and LRP6 that either promotes or inhibits Wnt-β-catenin signaling (2). Specifically, we discuss the molecular mechanism and biological consequences of the context-dependent regulation by Krm2 of Wnt-β-catenin signaling.

Krm1 and 2 (collectively termed Krms) were initially characterized as negative regulators of Wnt signaling (3). The secreted protein Dickkopf1 (Dkk1) [reviewed in (4)] inhibits Wnt signaling by directly binding to LRP5 and LRP6, thereby blocking formation of an active receptor signaling complex that consists of Wnt, Fz, and LRP5 or LRP6 (5, 6). Krms are single-pass transmembrane, high-affinity receptors for Dkk1 that augment Dkk1-mediated inhibition of Wnt signaling by promoting the endocytosis of LRP5 and LRP6 (3). The initial in vivo characterization of Krms indicated that they inhibit Wnt-β-catenin signaling through a Dkk1-dependent mechanism (3, 7). In Xenopus embryos, overexpression of Krm2 antagonizes Wnt8-mediated axis duplication in a Dkk1-dependent manner (7). In the same system, Krm2 overexpression rescues a partial loss-of-function phenotype that is mediated by an inhibitory antibody against Dkk1, whereas Dkk1 overexpression partially rescues the phenotype caused by morpholino-mediated knockdown of Krm1 and Krm2 (7). However, the authors noticed the presence of krm2 mRNA expression in some tissues that lacked expression of dkk1, which suggested the possibility of Dkk1-independent roles for Krms.

Hassler et al. now provide evidence for a Dkk1-independent role for Krm2 in Xenopus neural crest (NC) induction (2). The authors observed krm2 expression in the lateral neural plate where Wnt3a and Wnt8, but not dkk, transcripts were localized. NC cells are derived from the ectodermal layer of the neural tube and migrate throughout the vertebrate embryo during development [reviewed in (8)]. Cells derived from the NC contribute to the formation of a diverse range of cell and tissue types in the adult including melanocytes, the skeleton and connective tissue of the head and neck, and the peripheral nervous system. Overexpression of Krm2 in Xenopus embryos induced ectopic NC formation, whereas knockdown of Krm2 inhibited NC formation (2). Interestingly, knockdown of LRP6 also inhibited NC formation, which suggested that both Krm2 and LRP6 positively regulate formation of the NC (2). This result was unexpected given previous work that showed that Krms antagonize LRP6 function. To explore the mechanism underlying this developmental phenomenon, the authors studied the effects of Krm1 overexpression in cultured mammalian cells and found that Krm1 synergized with LRP6 overexpression to promote β-catenin–mediated transcription (2). In contrast, Krm1 overexpression sensitized cells to Dkk1-mediated inhibition of Wnt signaling. Thus, in the absence of Dkk1, Krms stimulated Wnt-β-catenin signaling, whereas in the presence of Dkk1, Krms antagonized Wnt-β-catenin signaling. These results may explain how both Krm2 and LRP6 promote NC induction in the absence of Dkk1.

Insight into how Krm2 promotes Wnt-β-catenin signaling may be gleaned from the authors’ findings that Krm2 directly interacts with LRP6 in the absence of Dkk1, and that Krms promote the cell-surface localization of LRP6 (2). The authors demonstrated that the overexpression of Krm2 promoted the cell-surface localization of exogenous LRP6 in human embryonic kidney 293 (HEK293) cells, and that knockdown of Krm2 decreased the abundance of LRP6 in Xenopus NC explants. Hassler et al. suggest that Krm may act as a switch that determines whether LRP5 and LRP6 localize to the plasma membrane or are endocytosed (Fig. 1). In the absence of Dkk1, Krms may sequester and maintain mature LRP5 and LRP6 at the plasma membrane. Binding of Dkk1 to Krm may induce an alternative conformation of Krm that promotes the endocytosis of LRP5 and LRP6. To validate this model, it will be important to test whether knockdown of Krm, in the context of endogenous LRP5 and LRP6, inhibits a cell’s response to canonical Wnts. Nonetheless, this work suggests a novel role for Krms in positively regulating the intracellular trafficking of LRP5 and LRP6 that merits further investigation.

Fig. 1.

A molecular model for the context-dependent activation or inhibition of Wnt-β-catenin signaling by Krms. (A) In the absence of Dkk1, Krm potentiates Wnt-β-catenin signaling by maintaining LRP5 or LRP6 at the plasma membrane. (B) In the presence of Dkk1, Krm augments Dkk1-mediated Wnt inhibition by promoting the endocytosis of LRP5 or LRP6.

One developmental implication of Krms’ context-dependent positive and negative regulation of Wnt-β-catenin signaling is that Krms may act to generate a biphasic response to a shallow Dkk1 gradient. In the context of a tissue responding to a uniform concentration of Wnt ligand in the absence of Krms, cells that secrete Dkk1 will generate a shallow gradient of Wnt inhibition that is defined by Dkk1 diffusion (Fig. 2A). Krms may play a role in establishing distinct borders between cells in which Wnt signaling is transduced and cells in which Wnt signaling is inhibited by Dkk1 (Fig. 2B). In cells exposed to a quantity of Dkk1 that is above a critical concentration, Krms act as Dkk1 receptors and synergize with Dkk1 (by promoting LRP5 and LRP6 endocytosis) to inhibit Wnt signaling (3). In contrast, in cells exposed to subthreshold quantities of Dkk1, Krms enhance the sensitivity of these cells to Wnts and potentiate Wnt signal transduction (2). In this model, the critical concentration of Dkk1 reflects the affinity of Krm for Dkk1. Such a mechanism to promote bistable responses to a shallow Dkk1 gradient could play a role in developing precise and robust borders between one cell type and another.

Fig. 2.

The context-dependence of Krms’ effects may generate a biphasic response to a shallow Dkk1 gradient. (A) In the absence of Krm, a concentration gradient of Dkk1 generates a linear decrease in Wnt inhibition (and thus an increase in Wnt signaling) as a function of distance from Dkk1-secreting cells. (B) In the presence of Krm, the same Dkk1 concentration gradient leads to biphasic inhibition of Wnt signaling. Krm synergizes with Dkk1 to promote Wnt inhibition in cells exposed to quantities of Dkk1 that are above a threshold concentration, whereas Krm potentiates Wnt signal transduction in cells exposed to quantities of Dkk1 that are below this threshold. In both (A) and (B), cells are exposed to a uniform concentration of Wnt ligand.

The context-dependent role described for Krms and their ability to promote biphasic signaling responses during development may be a general characteristic of other LRP5 and LRP6 modulators such as Dkk2 and Wise, though the mechanistic details may differ. Unlike Dkk1, an increased abundance of Dkk2 promotes Wnt signaling in Xenopus embryos and cultured cells by the direct binding of Dkk2 to LRP6 and the activation of Wnt-β-catenin signaling (9). Interestingly, Krm2 abrogates Dkk2-mediated activation of Wnt signaling in cultured cells (10). In contrast to results from overexpression studies that suggest a positive role for Dkk2, Dkk2-null mice have bone and corneal defects that suggest that Dkk2 inhibits Wnt signaling in vivo (11, 12). Importantly, a positive role for endogenous Dkk2 has not been identified. Wise is a secreted protein that binds to the extracellular domain of LRP6 and acts as either a positive or negative regulator of Wnt-β-catenin signaling depending on the cellular context (13). In Xenopus ectodermal explants in which Wnt ligands are absent, the increased abundance of Wise activates Wnt-β-catenin signaling (13). However, overexpression of Wise antagonizes Wnt-β-catenin signaling in the context of Wnt8 coexpression (13). In cultured cells, Wise synergizes with certain Wnts but antagonizes others (14). Context-dependent signaling is by no means limited to the Wnt pathway; the secreted protein Twisted gastrulation (Tsg) regulates bone morphogenetic protein (BMP) signaling in a context-dependent manner (15). Intriguingly, the dual function of Tsg is suggested to provide a mechanism for generating sharp borders during embryogenesis (15).

Although Krm2, Dkk2, and Wise promote Wnt-β-catenin signaling in certain experimental contexts and inhibit it in others, it is not known whether the endogenous proteins retain these dual roles in vivo. In Xenopus embryos, Krm-mediated inhibition of Wnt signaling has been demonstrated with Krm1 and Krm2 knockdown displaying synergy with Dkk1 inhibition and a phenotype consistent with loss of endogenous canonical Wnt repression in anteroposterior patterning (7). Analysis of thymic epithelial cells derived from Krm1-null mice also suggests a role for Krm1 as a Wnt inhibitor in vivo (16). Conversely, the argument that endogenous Krms act as positive regulators of Wnt-β-catenin requires further study. Although both Krm2 and LRP6 promote NC induction (2), it is not known whether this induction occurs through canonical Wnt signaling or through noncanonical Wnt signaling. Because LRP6 and Dkk1 regulate noncanonical, Wnt–planar cell polarity signaling (17, 18), it will be important to determine whether Krms also play such roles, possibly in NC induction. Future work should formally test whether endogenous Krms promote Wnt-β-catenin signaling in vivo. In summary, the study by Hassler et al. provokes some intriguing questions regarding how and why certain Wnt signaling components are context-dependent agonists and antagonists (2). We predict that the answers to these questions will have profound implications for understanding the mechanism of Wnt signaling and its roles in development.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
View Abstract

Stay Connected to Science Signaling

Navigate This Article