Looking beyond the Wnt pathway for the deep nature of β-catenin

Abstract

After two decades of stardom, one would think that β-catenin has revealed all of its most intimate details. Yet the essence of its duality has remained mysterious--how can a single protein both be the core link between cadherins and the cytoskeleton, and the nuclear messenger for Wnt signalling? On the basis of the available evidence and on molecular and evolutionary considerations, I propose that β-catenin was a born nuclear transport receptor, which by interacting with adhesion molecules acquired the property to coordinate nuclear functions with cell-cell adhesion. While Wnt signalling diverted this activity, the original pathway might still function in modern eukaryotes.

Figure 1

Structural and functional similarities between importin and exportin and β-catenin. (A) Diagram of the basic structure of an importin and exportin. The HEAT repeats form two surfaces—a convex outer surface rich in hydrophobic residues, which interacts with FG-repeats of the Nups, and a groove that binds to transport cargos (red). (A′) Three-dimensional structure of importin-β. (B) Summary diagram of β-catenin showing the components involved in nuclear transport, gene regulation and cell–cell adhesion. Direct binding of Arm repeats to Nups is responsible for rapid nuclear import and export. The groove formed by the Arm repeats (blue) can bind to TCF/LEF1 and several other nuclear factors, as well as the scaffold proteins Axin and APC, which constitute the β-catenin degradation complex. These interactions are mutually exclusive with cadherin binding. The carboxy-terminal activation domain (CTA) binds to several components involved in transcription and chromatin remodelling and is essential for classical Wnt signalling. (B′) 3D structures of β-catenin. (C) General diagram of classical Ran-dependent import and export, and of a hypothetical nuclear transport that is based on simple diffusion, in which directionality would be dictated by the compartmentalization of a sink for cargos. Sinks might bind to chromatin, membrane or cytoplasmic proteins, or they might be due to protein degradation. (D) Potential mechanisms of transport and transcriptional regulation by β-catenin. The original pathway might have involved passive shuttling of the β-catenin-cargo complex and delivery to chromatin. Modern β-catenin can de-repress target genes simply by competing out co-repressors (Groucho in the case of TCF/LEF1). In addition, β-catenin can directly recruit transcription factors (as well as histone deacetylases and components of chromatin remodelling complexes, not shown) and positively regulate transcription. Last, β-catenin could mediate export of nuclear factors. TCF–DNA interactions can be released by TCF phosphorylation, and similar reactions could regulate other DNA-binding partners of β-catenin. Degradation in the cytoplasm could serve as a sink for the diffusion reaction. APC, adenomatous polyposis coli protein; CTA, carboxy-terminal transactivation domain; HEAT, Huntingtin, elongation factor 3, protein phosphatase 2A, TOR1; NFκB, nuclear factor kappa B; Nup, nucleoporin; Sox, Sry-related HMG box proteins; TCF/LEF, T-cell factor/ lymphoid enhancer-binding factor 1.

Figure 2

Hypothetical evolutionary and functional relationship between β-catenin and nuclear transport receptors. Importins and exportins on one side and β-catenin and other Arm proteins on the other side might have evolved from a common transporter prototype, which could shuttle DNA-binding proteins, delivering them to chromatin and/or retrieving them back to the cytoplasm. Such an ancient importin probably had a groove enriched with acidic residues (red) selective for DNA-binding proteins. An important evolution of HEAT-repeat-containing importins and exportins might have been the acquisition of Ran binding, which regulates cargo assembly and release on either side of the nuclear pore and thus mediates efficient unidirectional transport. Along the branch of the Arm proteins, importin-α might have lost its ability to translocate on its own and become dependent on importin-β1 and CAS for nuclear import and export, respectively. The importin-α/-β1 pair became the main pathway for import of cargos carrying a classical NLS sequence composed of one or two small clusters of basic residues. β-catenin (and other related Arm proteins) might have been specialized carriers for a subset of cargos, and in addition, might have acquired the ability to interact with adhesion molecules, thus establishing a primitive signalling pathway between the cell surface and the nucleus. Note that a physiological role of β-catenin in nuclear transport has not yet been demonstrated and putative cargos have not yet been identified (question mark). This putative basal mechanism, based on competition between cargos and membrane anchors, might have been refined by eliminating excess free β-catenin through the Axin-based degradation complex. β-catenin might also have acquired transcriptional co-activation properties by bridging its DNA-binding partners (such as TCF) with transcription factors (TFs). The cadherin–β-catenin signalling pathway might have been high-jacked by Wnts and their receptors LRP5/6 and Frizzled (Fz)—by inhibiting the degradation complex, this pathway builds a large pool of cytosolic β-catenin and thus might bypass the basal control by cell adhesion. Note that several proteins harbouring HEAT or Arm repeats might have lost the property to cross the nuclear pore and have evolved different functions. These side branches of the two families are omitted in the diagram. CAS, cellular apoptosis susceptibility protein homologue; HEAT, Huntingtin, elongation factor 3, protein phosphatase 2A, TOR1; LRP5/6, low-density lipoprotein receptor-related protein 5/6; NLS, nuclear localization signal; TCF, T-cell factor; Wnt, combination of Wingless and mammalian proto-oncogene int1

Figure 3

Cell-adhesion-dependent β-catenin signalling. We propose that the state of adhesion controls specific phosphorylation reactions. The simplest mechanism could involve increased/decreased accessibility to kinases and phosphatases depending on the degree of cadherin clustering. Under conditions of strong adhesion, β-catenin is stable—it is protected within the cadherin complex, both by the tight clustering of the cadherins, and by the recruitment of actin-binding proteins (ABPs) and actin filaments. When adhesion is loose or absent, cadherins are de-clustered and connections with the cytoskeleton are dissolved. The cadherin–β-catenin bond is then exposed to various regulators. Both cadherin and β-catenin undergo multiple modifications, in particular serine/threonine and tyrosine phosphorylation (pS and pY). Cadherin serine/threonine phosphorylation tends to stabilize its interaction with β-catenin, whilst phosphorylation of specific tyrosine residues destabilizes it [2]. Phosphorylation of β-catenin decreases its interaction with α-catenin, and perhaps also with cadherins [2]. Note that multiple regulators, including kinases and phosphatases, can directly associate with cadherin and β-catenin [91]. Endocytosis of the cadherin complex (not represented) is probably also implicated in β-catenin signalling [95]. The pool of β-catenin released from the cadherin complex could be biochemically marked by intramolecular interaction [97], phosphorylation and interaction with other components. This activated β-catenin would signal more efficiently, either due to protection from degradation and rebinding to cadherins, increased nuclear transport, stronger binding to TCF or higher transactivation capabilities.

François Fagotto