Off-track takes Frizzled off the canonical path

Abstract

EMBO J 30 18, 3729–3740 (2011); published online July192011 [PMC free article] [PubMed] Like many other signalling pathways, the Wnt pathway is seen as part of a network that integrates extracellular information mediated by a variety of cell surface proteins to produce cell context-specific outputs. The identification of Frizzled as Wnt receptors (Bhanot et al, 1996) provided the essential link between extracellular Wnts and the intracellular components of Wnt signal transduction. Recent observations, however, indicate a rather complex interplay of this ‘core' receptor element with an increasing number of co-receptors. In this issue of The EMBO Journal, Tolwinski and Borchers (Peradziryi et al, 2011) describe PTK7/Off-track as a novel co-receptor element that might govern specificity in cellular Wnt responses. The β-catenin/TCF pathway is often described as canonical because it mediates the core activities of Wnts in cell specification and tumourigenesis. This distinction is necessary because Frizzled receptors had previously been shown to participate in planar cell polarity (PCP), a process whereby cells acquire polarity within the plane of epithelia (Strutt, 2008). PCP is the best characterised of various alternative Wnt signalling pathways that have collectively been termed non-canonical. Over the years, vertebrate Wnts have been classified into canonical and non-canonical Wnts. For example, vertebrate Wnt3a is an archetypical canonical Wnt while vertebrate Wnt5a has been implicated in PCP signalling. This distinction can be useful but is almost certainly an oversimplification, as the effect of various Wnts is strongly dependent on the nature of the Frizzled receptors present (van Amerongen and Nusse, 2009). It is worth mentioning that, in Drosophila, no Wnt has so far been shown to be required for PCP signalling. One feature that has often been noted is that activation of the two pathways is somewhat exclusive. How PCP and canonical signalling affect each other and what the relative contribution of various Frizzled and Wnts to either pathway is are questions currently under active investigation. Despite their central importance, Frizzled proteins are far from being the only receptors that modulate or control Wnt signalling. Indeed, members of the LRP (LDL-related protein) family are essential co-receptors for canonical signalling. For example, Arrow, the sole Drosophila LRP, is absolutely required for Wingless signal transduction (Wehrli et al, 2000). A current view is that LRP family members form a tripartite complex with Wnt and Frizzled. As arrow mutations do not cause a PCP phenotype in flies, and vertebrate LRPs do not modulate PCP signalling, this class of co-receptors is seen as strictly canonical. Another class of single-pass transmembrane proteins, including ROR1 and ROR2, also act as Wnt co-receptors although their effect on canonical signalling is inhibitory (Green et al, 2008). Initially, ROR receptors were thought to act by merely sequestering Wnt receptors. However, a recent paper suggests that, in addition to suppressing canonical signalling, these receptors could actively promote PCP signalling in a Frizzled-dependent manner (Grumolato et al, 2010). One simple model is that the ROR2/Fz/Wnt complex could trigger PCP signalling while the LRP/Fz/Wnt complex would activate canonical signalling. Individual Wnts would favour the formation of one or the other type of complex thus explaining why some Wnts tend to activate PCP while others activate canonical signalling. Tolwinski, Borchers and colleagues describe another transmembrane receptor that inhibits canonical signalling and could act in the PCP pathway (Peradziryi et al, 2011). This protein, which is known as protein tyrosine kinase-7 (PTK7) in vertebrates and Off-track (Otk) in Drosophila, has been shown previously to be deregulated in cancers and to be implicated in various morphogenetic movements (see references in Peradziryi et al, 2011). The authors' interest in Otk was piqued by the mutant phenotype in Drosophila: reduction of zygotic otk leads to a mild excess in canonical Wingless signalling. Conversely, otk overexpression inhibits Wingless signalling, although only in regions where Wnt4 is expressed. Such inhibition is almost certainly mediated by Wnt4, because Wnt4 overexpression inhibits canonical signalling in an Otk-dependent manner. This, and the results of immunoprecipitation experiments, led to the suggestion that Otk could preferentially bind to Wnt4 in Drosophila. More immunoprecipitation experiments with Xenopus proteins show that PTK7 binds to Wnt3a and Wnt8 (which activate canonical signalling in Xenopus embryos), and that this interaction requires a Frizzled. Thus, PTK7 could inhibit canonical signalling by displacing LRP from the activating tripartite complex. This is similar to, yet distinct from, the proposed action of ROR2. ROR2 disrupts the canonical Wnt/Frizzled/LRP signalling complex but, at the same time, promotes the formation of a PCP-specific Wnt/Frizzled/ROR2 signalling complex. PTK7 is unlikely to act in the same manner, since it does not bind to Wnt5a and Wnt11, two ‘PCP-specific' Wnts. Yet the authors suggest that, through an unknown mechanism, PTK does participate in PCP signalling. In support of this assertion, overexpression of Wnt4 or Otk disrupts PCP signalling in Drosophila wings. However, this observation is weakened by the fact that otk or wnt4 loss-of-function has no effect on PCP. Thus, further work is needed to determine whether PTK7/Otk is really a component of a PCP signalling pathway. In particular, it will be necessary to determine its mode of action and to assess the significance of its similarity to receptor tyrosine kinases (RTKs), especially in light of observations, suggesting that activity of its kinase domain is not required for its role in Wnt signalling. Intriguingly, RORs are also homologous to RTKs, and there may be useful information to be gleaned from comparing directly the function of ROR and PTK7. Here, the power of Drosophila genetics may help as a single ROR is encoded by its genome (http://flybase.org/). Genetic analysis could assess whether ROR and PTK7/Otk overlap in function or perform distinct roles. The present study adds to the growing sentiment that the nature and abundance of receptors and co-receptors are key to determining the cellular effects of Wnts (Figure 1). In addition to the receptors discussed above, other Wnt receptors have been reported including Crypto, which too acts in concert with Frizzled albeit in an as yet uncharacterised fashion (Tao et al, 2005), and Ryk, which, in contrast to ROR and PTK7, activates a Frizzled-independent pathway (van Amerongen and Nusse, 2009). Other extracellular regulators of Wnts include secreted antagonists and glypicans (Kawano and Kypta, 2003; Lin, 2004). It is also worth mentioning the recent demonstration that R-spondins, secreted molecules previously known to potentiate canonical Wnt signalling, act through LGR4 and LGR5, two 7-pass transmembrane receptors that have remained orphaned up to now (Carmon et al, 2011; de Lau et al, 2011). The cell surface is rich with factors that control and modulate Wnt signalling activity. We can look forward to understanding how the abundance and activity of these factors is controlled and integrated to provide the cell with meaningful information about its environment. Figure 1 Many receptors modulate Wnt signalling. Frizzled proteins are considered the main Wnt receptors. The outcome of the Wnt-Frizzled interaction is determined by co-receptors. Thus, Wnt/Frizzled/LRP activates the canonical β-catenin/TCF signalling ...