Molecular Scaffolds Regulate Bidirectional Crosstalk Between Wnt and Classical Seven-Transmembrane Domain Receptor Signaling Pathways

See allHide authors and affiliations

Science's STKE  31 Jul 2007:
Vol. 2007, Issue 397, pp. pe41
DOI: 10.1126/stke.3972007pe41


Signaling downstream of classical seven-transmembrane domain receptors (7TMRs) had generally been thought to recruit factors that are in large part separate from those recruited by atypical 7TMRs, such as Frizzleds (Fzs), receptors for the Wnt family of glycoproteins. Classical 7TMRs are also known as G protein–coupled receptors (GPCRs) and are mediated by signaling factors such as heterotrimeric guanine nucleotide–binding proteins (G proteins), GPCR kinases (GRKs), and β-arrestins. Over the past few years, it has become increasingly apparent that classical and atypical 7TMRs share these factors, which are often associated with mediating classical 7TMR signaling, as well as the scaffolding proteins that were initially thought to be involved in transmitting atypical 7TMR signals. This sharing of signaling components by agonists that bind classical 7TMRs and those binding to atypical 7TMRs establishes the possibility of extensive crosstalk between these receptor classes. We discuss the evidence for, and against, crosstalk, and examine mechanisms by which this can occur.

Signaling downstream of classical seven-transmembrane domain receptors [7TMRs, also known as G protein–coupled receptors (GPCRs)] was thought to involve factors that are in large part distinct from those that mediate signaling by atypical 7TMRs, such as Frizzleds (Fz), which are receptors for the Wnt family of secreted glycoproteins. It has become increasingly apparent, however, that many factors are shared by classical and atypical 7TMRs, including heterotrimeric guanine nucleotide–binding proteins (G proteins), GPCR kinases (GRKs), β-arrestins, and the scaffolding proteins that mediate Wnt signals, such as Axins and members of the Disheveled (Dvl) family. This sharing of signaling components by classical and atypical 7TMRs raises the possibility of extensive crosstalk between these receptor classes. Evidence supporting crosstalk has been found in certain cell types under specific circumstances. Herein, we review the evidence for and against crosstalk, examine mechanisms by which crosstalk occurs, and speculate on possible additional examples.

Connecting Wnt Signaling to Factors Downstream of Classical 7TMRs

Wnts are secreted glycoproteins that bind to Fzs and the co-receptors LRP5 and LRP6 [low density lipoprotein (LDL) receptor-related protein 5 and 6] , which leads to the activation of two classes of signaling pathways: canonical and noncanonical (1) (Fig. 1). Canonical Wnt signaling culminates in the stabilization of β-catenin and its translocation to the nucleus where, together with its transcriptional coactivators, members of the T cell factor/lymphoid enhancer factor (Tcf/Lef) family, it induces the expression of genes involved in a host of processes, including development and carcinogenesis (2). The noncanonical pathway [(Fig. 1); also known as the planar cell polarity or convergent extension pathway] (35) is a network of pathways (to be discussed below) and is also involved in a wide variety of cellular processes. Notably, Dvls play key roles in the activation of both classes of Wnt pathways (4, 6, 7).

Fig. 1.

A schematic view of the integration of common signaling elements from the pathways activated by ligands binding to classical and atypical 7TMRs. (Right) The canonical pathway leading to stabilization of β-catenin and activation of β-catenin/Tcf–dependent gene expression. Also shown are noncanonical pathways and the signaling complexes that lead to activation of RhoA/ROCK and Rac/JNK, the latter mediated, at least in B cells, by GCKR. GCKR can also, by unclear mechanisms, lead to stabilization of β-catenin, thereby activating β-catenin/Tcf–dependent gene expression. (Left) The binding of ligands to classical 7TMRs such as the α-adrenergic or PGE2 receptors leads to the activation of the G proteins Gq and Gs, respectively. This results in Akt-mediated inhibition of GSK-3 and stabilization of β-catenin. Also shown is the β-arrestin (β-arr)–mediated internalization of classical and atypical 7TMRs (broken red arrows), which, for the classical receptors, can lead to the activation of downstream signaling pathways. The functional consequences of β-arrestin–mediated internalization of Fzs are not clear. Solid red arrows represent new molecular interactions (for example, Akt and αs entering the Axin complex) and the dissociation of GSK-3 from the Axin complex. Black arrows and blunt end lines depict activation and inhibition, respectively. Broken black arrows indicate both translocation and activation. See text for further details. ASK1, apoptosis-stimulated kinase 1; MAP2K, mitogen-activated protein kinase kinase (MAP3K, similarly); MEK, mitogen-activated and extracellular signal-regulated kinase kinase; MKK4, mitogen-activated protein kinase kinase kinase 4; PDK1, phosphoinositide-dependent protein kinase; PI3K, phosphatidylinositol 3-kinase.

Heterotrimeric G proteins of the Gαi (which includes Gαo) and Gαq families couple to Fzs and are necessary for signal transmission in both the canonical and noncanonical pathways (8, 9). Thus, activation of these G proteins by Wnt might be expected to recruit factors that are activated downstream of classical 7TMRs that couple to these same G proteins [for example, mitogen-activated protein (MAP) kinases]. However, this is generally not true for the canonical Wnt pathway because the scaffolding proteins that mediate canonical Wnt signaling, rather than G proteins, largely determine which downstream mediators will be recruited (6, 10). Thus, the spatial constraints provided by these scaffolds lead to fidelity of signaling in the canonical pathway.

β-Arrestins are scaffold proteins that regulate signaling by both classical and atypical 7TMRs (1013). β-Arrestin-1 and β-arrestin-2 (also known as arrestin-2 and arrestin-3, respectively) are ubiquitously expressed (although not in the eye) and act in a receptor-specific manner. β-Arrestins bind to GRK-phosphorylated classical 7TMRs and cause receptor desensitization, first, by steric blockade of further G protein–coupling, and second, by triggering receptor internalization (10). Internalization of ligand-bound classical 7TMRs is not only responsible for receptor desensitization, but is also involved in other processes, such as receptor resensitization, receptor degradation, and in mediating signaling by the internalized receptor. Indeed, β-arrestins are key scaffolding proteins for various kinases and adaptor proteins that not only facilitate activation of downstream signaling cascades, but also determine the specificity of signaling by classical 7TMRs (10, 13).

Chen et al. (14) first proposed a central role for β-arrestins in Wnt signaling when they demonstrated that β-arrestin-1 binds to phosphorylated Dvl, and that coexpression of β-arrestin-1 with either Dvl1 or Dvl2, even in the absence of Wnt stimulation, synergistically activated a β-catenin/Tcf–dependent reporter. This is consistent with a role for β-arrestin in activation of the canonical Wnt pathway, a contention recently supported by Bryja et al. (15). Indeed, these authors demonstrated that β-arrestins are necessary for maximal Wnt3a-induced phosphorylation of Dvl [a key step in activation typically mediated by the δ and ε isoforms of casein kinase I (CKI), or Par-1, or both (7)] and subsequent signaling to β-catenin.

Perhaps it is not surprising that β-arrestins meditate signals downstream of Wnts, given that the receptors for Wnts are 7TMRs. What is surprising, however, is the atypical role that arrestins play in serving as a critical link holding together a complex of scaffold proteins, including Axin and Dvl, which are necessary for canonical Wnt signaling (15). β-Arrestin appears to be necessary for the recruitment of Dvl to the Axin complex, which leads to the inhibition of glycogen synthase kinase-3 (GSK-3) and stabilization of β-catenin, although the mechanisms involved are unknown. These studies demonstrate that β-arrestin is necessary, but not sufficient, for canonical Wnt signaling.

Do any of the more traditional functions of β-arrestins, such as mediating receptor internalization, play a role in Wnt signaling? Two studies suggest that this is the case. First, internalization of Fz4 following stimulation with Wnt5a (a noncanonical Wnt) is dependent on β-arrestin-2 and protein kinase C (PKC) (16). Second, blockade of internalization by hyperosmolar sucrose prevented Wnt-stimulated β-catenin signaling in the murine fibroblast-like L cell (17). Although this observation suggests that receptor internalization is a key part of canonical Wnt signaling, it is not known whether β-arrestins directly mediate this process. Finally, the consequences of internalization are not clear, and could involve the usual roles of β-arrestins, such as mediating receptor down-regulation or enhancing signaling, or as-yet-uncharacterized roles. Thus the role of β-arrestins in internalization of each of the 10 Fzs, the branch (or branches) of Wnt signaling that are affected by Fz internalization, and the consequences of receptor internalization remain to be determined.

Of potentially greater interest, Bryja et al. suggest that the recruitment of β-arrestins by Wnt signaling opens up the possibility of all manner of crosstalk mediated by β-arrestins between Wnt-dependent pathways and factors thought to be primarily downstream of classical 7TMRs (15). That said, the classical mediators downstream of β-arrestin–mediated internalization of classical 7TMRs are the MAPKs, and in particular, the extracellular signal–regulated kinases (ERKs) (10), yet there is relatively little evidence, despite the intense investigation of the canonical Wnt pathway, that ERKs or other kinases, including Src family members, directly interact with canonical Wnt signaling (1821). Thus, β-arrestins appear to perform unique functions and target different downstream mediators in Wnt signaling as opposed to classical 7TMR signaling. Therefore, it seems unlikely that β-arrestins mediate direct crosstalk between the pathways. Rather, if crosstalk occurs, it seems more likely that it would be driven by Dvl.

There is indeed a precedent for this because Dvl mediates the activation of Akt (22) and the c-Jun N-terminal kinase (JNK) family of MAPKs by canonical and noncanonical Wnt signaling, respectively (7, 19). Thus, both canonical and noncanonical pathways use Dvls, (6, 7, 19), but the complexes assembled on Dvl differ, and in this way, downstream factors are differentially activated (6). The mechanisms by which Dvls discriminate between specific downstream factors are not understood, but involve differential phosphorylation of Dvls by upstream kinases and the subsequent assembly of specific complexes of scaffold proteins and their various binding partners [(Fig. 1), and see below] (6, 7).

What are the signaling intermediates from Dvl to JNK activation in the noncanonical pathway? Dvl is necessary to activate two arms of the pathway, the first, which involves the Rho family guanosine triphosphatase (GTPase) member RhoA, and the second, which is mediated by the Rho GTPase Rac (4, 23, 24) (Fig. 1). Dvl forms two complexes, one including the formin-homology (FH) protein Daam1 (disheveled-associated activator of morphogenesis) and RhoA that leads to the subsequent activation of Rho-associated kinase (ROCK) and cytoskeletal reorganization (25), and the other including Rac that leads to JNK activation (26), although how this signal was transmitted to JNK activation was not known. Kehrl and co-workers reported that Wnt stimulation in B cells leads to the Dvl-dependent assembly of a signaling complex that includes adenomatous polyposis coli (APC), Asef [a guanine nucleotide exchange factor (GEF) for Rac], Rac, and the kinase GCKR [germinal center kinase (GCK)-related], and that this complex activates JNK (27). GCKR is a MAP4K (MAP kinase kinase kinase kinase) that is a member of the large (28 member) GCK family of Sterile 20-like (Ste20-like) kinases. Strikingly, unlike CKIε and Par-1, which phosphorylate Dvls, resulting in activation of the canonical pathway and inhibition of the noncanonical pathway, GCKR, acting downstream of Dvl, activates both pathways (27).

Dvl also interacts with other members of the Ste20-like kinase families. For example, Dvl mediates activation of p21-activated kinase 1 (PAK1) (28) in response to activation of the muscle-specific receptor tyrosine kinase (MuSK), which is located at the neuromuscular junction in skeletal muscle. MuSK, Dvl, and PAK1 form a ternary complex, raising the possibility that interactions between Dvl and other MAP4Ks could be a more general paradigm by which Dvls signal to MAPKs.

Connecting Classical 7TMRs to Wnt Signaling

Bryja et al. also suggest that the interactions of β-arrestins with core components of Wnt pathways (for example, Dvls and Axins) may allow classical 7TMRs to recruit components of the canonical Wnt pathway. Some data support this direction of crosstalk, which leads to the activation of β-catenin/Tcf signaling (2935). However, not only has a role for β-arrestins not been identified, there is no clear role for Dvl. Rather, Axin and its scaffolding functions seem to be critical (29, 30) (Fig. 1). Axin forms a complex with APC, GSK-3, CKI, and β-catenin, and, when cells are not stimulated, GSK-3 phosphorylates β-catenin, targeting it for ubiquitination and degradation. Stimulation of α-adrenergic receptors on cardiomyocytes (acting through Gq) or prostaglandin E2 (PGE2) receptors on colon cancer cells that lack APC (acting through Gs) results in the stabilization of β-catenin and the activation of β-catenin/Tcf–dependent transcription. However, the mechanism of β-catenin stabilization in response to classical 7TMR stimulation is different from that employed by canonical Wnt signaling. Akt is recruited to the Axin complex, where it phosphorylates and inactivates GSK-3 (29, 30), a process that appears to play little or no role in Wnt-dependent stabilization of β-catenin (36). PGE2-dependent β-catenin stabilization also involves the association of the αs subunit with Axin, leading to the dissociation of GSK-3 from the Axin complex (29). This is especially interesting because Gαi and Gαq disrupt the interaction between GSK-3 and Axin in canonical Wnt signaling, but the mechanism involved is not known (37). Stimulated classical 7TMRs may also recruit cAMP-dependent protein kinase (PKA) and one or more PKCs to inhibit GSK-3, but it is not clear whether this can lead to stabilization of β-catenin (38).

The phenotypes produced by activation of the canonical pathway by classical 7TMRs are quite distinct, depending on the cell type. In cancer cells, not surprisingly, cell proliferation is observed (29, 32, 35). In cardiomyocytes, which are thought of as terminally differentiated cells, stabilization of β-catenin results in hypertrophic growth (30). That β-catenin was necessary for this hypertrophic response was later confirmed in vivo using either cardiac-specific conditional deletion of the gene encoding β-catenin (39) or knock-out of the gene encoding Dvl-1 (40). Not surprisingly, however, whether β-catenin plays a role in mediating hypertrophy depends on which receptors are activated. For example, β-catenin was not necessary for angiotensin II–induced hypertrophy (41).

In summary, it is clear that, under certain conditions, Wnt signaling recruits factors that are classically viewed as signaling in "non-Wnt" pathways. These include β-arrestins, which appear to be important in signal transmission to β-catenin/Tcf. For the most part, however, β-arrestins do not serve as important links from Wnt signaling to the pathways downstream of classical 7TMRs that β-arrestins typically activate. However, the Wnt pathway components Dvl and APC mediate GCKR-dependent activation of both the canonical and noncanonical pathways in B cells, and given the large number of factors with which Dvl can interact, there likely will be more examples. Thus far, the mechanisms of recruitment by Wnt pathways of targets for classical 7TMRs appear to be largely indirect. This relative fidelity of signaling in the Wnt pathways is in contrast to the several examples of classical 7TMRs that recruit components of the canonical Wnt pathway. Given that some of these instances have been in cell types with mutations in key factors that could promote the stabilization of β-catenin, their general applicability to noncancerous or nontransformed cells is unclear. Overall, however, fidelity within the Wnt pathways and the relative lack thereof in classical 7TMR pathways may speak to the fine-tuning that has occurred to control Wnt signaling, given the consequences of perturbations in Wnt pathways on the health of the organism.


  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.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.

Stay Connected to Science Signaling

Navigate This Article