Table of contents

Regulation of phosphorylation of ß-catenin

The 'ß-catenin destruction complex' is central to canonical Wnt/ß-catenin signaling. The scaffolding protein Axin and the tumor suppressor adenomatous polyposis coli protein (APC) are critical components of this complex, required for rapid ß-catenin turnover. The crystal structure of a complex between ß-catenin and the ß-catenin-binding domain of Axin (Axin-CBD) was determined. The Axin-CBD forms a helix that occupies the groove formed by the third and fourth armadillo repeats of ß-catenin and thus precludes the simultaneous binding of other ß-catenin partners in this region. Biochemical studies demonstrate that, when ß-catenin is phosphorylated, the 20-amino acid repeat region of APC competes with Axin for binding to ß-catenin. It is proposed that a key function of APC in the ß-catenin destruction complex is to remove phosphorylated ß-catenin product from the Axin/GSK-3ß active site (Xing, 2003).

In the canonical Wnt/ß-catenin pathway, ß-catenin mediates the transmission of a Wnt signal into the nucleus and the subsequent activation of target genes. In the absence of a Wnt signal, a cytoplasmic protein complex containing glycogen synthase kinase-3ß (GSK-3ß), the adenomatous polyposis coli protein (APC), and the scaffolding protein Axin, among others, catalyzes the phosphorylation of ß-catenin. This complex has been termed the 'ß-catenin destruction complex' because phosphorylation of ß-catenin targets it for degradation by the proteasome. When the pathway is active, binding of Wnt to its receptors leads to the inactivation of the destruction complex and a consequent accumulation of ß-catenin. The ß-catenin translocates to the nucleus, where it binds to DNA-binding proteins of the Tcf/LEF family. Together they turn on the transcription of Wnt-responsive genes. Although ß-catenin levels may also be regulated by other Axin-independent pathways, phosphorylation of ß-catenin by the ß-catenin destruction complex is the central regulatory step of the canonical Wnt/ß-catenin signaling pathway (Xing, 2003 and references therein).

In the ß-catenin destruction complex, GSK-3ß phosphorylates the critical residues in the N terminus of ß-catenin, contingent upon priming phosphorylation by casein kinase I (CKI). By itself, GSK-3ß does not efficiently phosphorylate ß-catenin; thus Axin plays a critical role in bringing GSK-3ß, CKIalpha, and ß-catenin together to efficiently promote the phosphorylation reaction. The importance of Axin in ß-catenin destruction is underscored by the presence of mutations in the human AXIN1 gene in certain human cancers that are associated with increased ß-catenin levels (Xing, 2003 and references therein).

Another essential component of the destruction complex is the tumor suppressor APC. Mutations of APC cause the elevation of cytoplasmic ß-catenin levels and are found in ~85% of colon cancers. The function of APC in the ß-catenin destruction complex is connected with Axin, because the overexpression of Axin in APC-mutant cancer cells is sufficient to down-regulate ß-catenin levels in these cells. APC contains repetitive ß-catenin interaction motifs, including three 15-amino acid repeats (or possibly four) and seven 20-amino acid repeats. It has been shown that APC plays a role in the transportation of ß-catenin from the nucleus to the cytoplasm, where ß-catenin is phosphorylated and degraded. Although it has also been proposed that APC may attenuate ß-catenin levels by recruiting ß-catenin to the ß-catenin destruction complex, it remains unclear how APC plays an essential role in the ß-catenin destruction complex (Xing, 2003 and references therein).

In addition to Axin, APC, GSK-3ß, and CKI, many other proteins, such as protein phosphatase 2A (PP2A), have also been found to play a role in the ß-catenin destruction complex. A central question now is how these proteins interact to form a molecular machine that efficiently phosphorylates and degrades ß-catenin. Specifically, a catalytic machine must be efficient in both substrate recruitment and product release. How does the ß-catenin destruction complex keep ß-catenin in the complex long enough to be phosphorylated, yet release it quickly enough to maintain the efficiency of phosphorylation (Xing, 2003)?

The crystal structure was determined of a complex between the armadillo repeat region of ß-catenin and the ß-catenin-binding domain of Axin, thus revealing the structural basis of the ß-catenin/Axin interaction. This structure suggests that Axin and the 20-amino acid repeat region of APC compete for binding to ß-catenin when they are both involved in the ß-catenin destruction complex. Biochemical studies clearly show that these regions do compete for binding, but only when the 20-amino acid region is phosphorylated. Based on these data, it is suggested that APC is required for both the recruitment of ß-catenin and the removal of the phosphorylated ß-catenin from the Axin/GSK-3ß active site, which explains the critical role of APC in ß-catenin turnover (Xing, 2003).

Wnt signaling increases ß-catenin abundance and transcription of Wnt-responsive genes. The B56 regulatory subunit of protein phosphatase 2A (PP2A) inhibits Wnt signaling. Okadaic acid (a phosphatase inhibitor) increases, while B56 expression reduces, ß-catenin abundance; B56 also reduces transcription of Wnt-responsive genes. Okadaic acid is a tumor promoter, and the structural A subunit of PP2A is mutated in multiple cancers. Taken together, the evidence suggests that PP2A is a tumor suppressor. However, other studies suggest that PP2A activates Wnt signaling. The B56, A and catalytic C subunits of PP2A each have ventralizing activity in Xenopus embryos. B56 is epistatically positioned downstream of GSK3ß and axin but upstream of ß-catenin, and axin co-immunoprecipitates B56 A and C subunits suggesting that PP2A:B56 is in the ß-catenin degradation complex. PP2A appears to be essential for ß-catenin degradation, since ß-catenin degradation is reconstituted in phosphatase-depleted Xenopus egg extracts by PP2A, but not PP1. These results support the hypothesis that PP2A:B56 directly inhibits Wnt signaling and plays a role in development and carcinogenesis (Li, 2001).

The Wnt pathway controls numerous developmental processes via the ß-catenin-TCF/LEF transcription complex. Deregulation of the pathway results in the aberrant accumulation of ß-catenin in the nucleus, often leading to cancer. Normally, cytoplasmic ß-catenin associates with APC and axin and is continuously phosphorylated by GSK-3ß, marking it for proteasomal degradation. Wnt signaling is considered to prevent GSK-3ß from phosphorylating ß-catenin, thus causing its stabilization. However, the Wnt mechanism of action has not been resolved. The regulation of ß-catenin phosphorylation and degradation by the Wnt pathway has been studied. Using mass spectrometry and phosphopeptide-specific antibodies, it has been shown that a complex of axin and casein kinase I (CKI) induces ß-catenin phosphorylation at a single site: serine 45 (S45). Immunopurified axin and recombinant CKI phosphorylate ß-catenin in vitro at S45; CKI inhibition suppresses this phosphorylation in vivo. CKI phosphorylation creates a priming site for GSK-3ß and is both necessary and sufficient to initiate the ß-catenin phosphorylation-degradation cascade. Wnt3A signaling and Dvl overexpression suppress S45 phosphorylation, thereby precluding the initiation of the cascade. Thus, a single, CKI-dependent phosphorylation event serves as a molecular switch for the Wnt pathway (Amit, 2002).

Wnt signals control decisive steps in development and can induce the formation of tumors. Canonical Wnt signals control the formation of the embryonic axis, and are mediated by stabilization and interaction of ß-catenin with Lef/Tcf transcription factors. An alternative branch of the Wnt pathway uses JNK to establish planar cell polarity in Drosophila and gastrulation movements in vertebrates. This study describes the vertebrate protein Diversin that interacts with two components of the canonical Wnt pathway, Casein kinase Iepsilon (CKIepsilon) and Axin/Conductin. Diversin recruits CKIepsilon to the ß-catenin degradation complex that consists of Axin/Conductin and GSK3ß and allows efficient phosphorylation of ß-catenin, thereby inhibiting ß-catenin/Tcf signals. Morpholino-based gene ablation in zebrafish shows that Diversin is crucial for axis formation, which depends on ß-catenin signaling. Diversin is also involved in JNK activation and gastrulation movements in zebrafish. Diversin is distantly related to Diego of Drosophila, which functions only in the pathway that controls planar cell polarity. These data show that Diversin is an essential component of the Wnt-signaling pathway and acts as a molecular switch, which suppresses Wnt signals mediated by the canonical ß-catenin pathway and stimulates signaling via JNK (Schwarz-Romond, 2002).

Compared with Diversin, Diego contains six ankyrin repeats instead of eight, and has 35% amino acid sequence identity within the ankyrin repeats, but little identity (18%) in the residual domains. Diego acts downstream of Frizzled and controls planar polarization of epithelial cells in the eye and wing; such polarization depends upon JNK activity. Coimmunoprecipitation experiments show that Diego interacts with mammalian CKIepsilon but not with Drosophila Axin. Further, Diversin stimulates JNK-dependent transcription in 293 cells. Diversin also promotes JNK activation that is induced by Dishevelled or Wnt11. In zebrafish embryos, injection of Diversin mRNA results in abnormal gastrulation movements, that is, the convergence and extension of injected embryos are defective. Low amounts of Diversin mRNA induce failure of gastrulation movements, but little ventralization, whereas at higher dosages, ventralization is predominant. Similarly, injection of low amounts of Diversin MOs also interfers with gastrulation movements and induces a general undulation of the embryo along its anterior-posterior axis, as revealed by in situ hybridization for myoD at the 5-10 somite stage. Thus, both activation and inhibition of the Wnt/JNK pathway perturb gastrulation movements. Injection of Diego mRNA also induces deficits in convergence and extension movements, but does not affect axis formation at any concentration tested. Diego could not rescue the Diversin MO-induced dorsalization. Taken together, these data indicate that Diversin functions both in the Wnt/ß-catenin and the Wnt/JNK pathway, and that Diego acts only in the Wnt/JNK pathway. Diego and Diversin are therefore structurally and functionally not entirely homologous (Schwarz-Romond, 2002).

Biochemical analysis allows the molecular mechanism by which Diversin functions in the canonical Wnt pathway to be assigned. Efficient ß-catenin degradation requires a two-step mechanism, a priming phosporylation at Ser 45 catalyzed by CKIepsilon or CKIalpha, and subsequent phosphorylation on three equally spaced serine/threonine residues by GSK3ß. Diversin recruits the priming kinase CKIepsilon to the Axin/Conductin-GSK3ß complex. Separate domains of Diversin, the central and C-terminal regions, mediate these two interactions. Both Diversin and GSK3ß bind simultaneously to dimeric Axin/Conductin, and they use identical binding sites. Diversin-mediated recruitment of CKIepsilon allows phosporylation of Ser 45 of ß-catenin, thus creating a classical GSK3ß recognition motif and initiating the subsequent phosphorylation cascade. A minimal fusion molecule that contains the catalytic domain of CKIepsilon and the Axin/Conductin-binding domain of Diversin is fully functional in ß-catenin signaling, showing the role of Diversin as a molecular linker. Diversin is inactive in the presence of Frat-1/GBP, which displaces GSK3ß, showing the importance of GSK3ß in the complex. Taken together, these data demonstrate that Diversin functions in the canonical Wnt pathway by engaging CKIepsilon to the ß-catenin degradation complex, and allows priming phosphorylation and degradation of ß-catenin (Schwarz-Romond, 2002).

Diversin activates the JNK branch of the Wnt-signaling pathway, that controls the establishment of planar cell polarity in Drosophila and gastrulation movements in vertebrates. In zebrafish, inhibition and overexpression of Diversin cause defects in gastrulation movements, that is, a reduction in body length and undulation of the body axis -- these defects are similar to those observed in pipetail (Wnt5a) mutants. Thus, Diversin controls gastrulation movements, as does the Drosophila protein Diego. However, Diego is only in part a functional homolog of Diversin, because it does not interact with Drosophila Axin and has not been implicated in Wnt/ß-catenin signaling. Specific to Diversin in vertebrates is its role at a branchpoint of intracellular Wnt signaling, where it represses the canonical Wnt/ß-catenin pathway and simultaneously activates the JNK pathway (Schwarz-Romond, 2002).

Protein interactions of Armadillo homologs: Interaction with Axin and APC, two components of the ß-catenin destruction complex

The adenomatous polyposis coli (APC) protein binds to the cellular adhesion molecule ß-catenin, a mammalian homolog of Armadillo. Overexpression of APC blocks cell cycle progression. When ß-catenin is present in excess, APC binds to another component of the Wingless pathway, glycogen synthase kinase 3ß, a mammalian homolog of Shaggy/Zeste white 3. APC is a good substrate for GSK3ß in vitro, and the phosphorylation sites map to the central region of APC. Binding of ß-catenin to this region is dependent on phosphorylation of GSK3ß (Rubinfeld, 1996). APC binds to DLG, the human homolog of the Drosophila Discs Large tumor suppressor protein. This interaction requires the C-terminal region of APC and the PDZ domain repeat region of DLG. APC colocalizes with DLG at the lateral cytoplasm in rat colon epithelial cells and at the synapse in cultured hippocampal neurons. These results suggest that the APC-DLG complex may participate in regulation of both cell cycle progression and neuronal function (Matsumine, 1996). Perhaps the PDZ domain of Dishevelled (involved directly in Wingless signaling in Drosophila) interacts with APC and thereby provides a link between DSH and the APC-GSK3-ß-catenin complex (Perrimon, 1996).

The stabilization of beta-catenin is a key regulatory step during cell fate changes and transformations to tumor cells. Several interacting proteins, including Axin (see Drosophila Axin), APC, and the protein kinase GSK-3beta are implicated in regulating beta-catenin phosphorylation and its subsequent degradation. Wnt signaling stabilizes beta-catenin, but it has not been clear whether and how Wnt signaling regulates the beta-catenin complex. Axin has been shown to be dephosphorylated in response to Wnt signaling. The dephosphorylated Axin binds beta-catenin less efficiently than the phosphorylated form. Thus, Wnt signaling lowers Axin's affinity for beta-catenin, thereby disengaging beta-catenin from the degradation machinery (Willert, 1999).

A model is presented for the role of Axin in Wnt signal transduction. In an unstimulated cell, GSK-3beta is active and phosphorylates Axin, which in turn, recruits beta-catenin into the Axin/GSK-3beta complex. By virtue of its proximity to GSK-3beta, beta-catenin is then phosphorylated. Phosphorylated beta-catenin is then targeted for degradation. Upon transduction of the Wnt signal through the Frizzled (Fz) receptors to Dishevelled, GSK-3beta kinase activity is inhibited so that PP2A dephosphorylates Axin. Unphosphorylated Axin, in turn, no longer recruits beta-catenin to the complex. Failure of beta-catenin to associate with the Axin/GSK-3beta complex prevents its phosphorylation by GSK-3beta so that it can accumulate to high levels in the cytoplasm and nucleus and activate transcription in concert with the Tcf/Lef-1 family of transcription factors. GSK-3beta also phosphorylates APC, which may facilitate beta-catenin recruitment into the complex; however, this event has not been shown to be regulated by Wnt signaling (Willert, 1999).

An implication of these results is that the primary target for GSK-3beta phosphorylation is Axin. Previous models have argued that GSK-3beta directly phosphorylates beta-catenin and thereby targets it for degradation. However, GSK-3beta does not bind directly to beta-catenin, and efficient in vitro phosphorylation of beta-catenin by GSK-3beta requires the presence of Axin, which binds both proteins. In contrast, efficient Axin phosphorylation by GSK-3beta does not require additional proteins. Thus, in a Wnt-stimulated cell, beta-catenin fails to be phosphorylated by GSK-3beta because it is not recruited into the Axin/GSK-3beta complex. It should also be noted that phosphorylation of APC by GSK-3beta increases beta-catenin binding to APC; however, in contrast to Axin, phosphorylation of APC has not been shown to be regulated by Wnt signaling (Willert, 1999 and references).

The N-terminal region of Dvl-1 (a mammalian Dishevelled homolog) shares 37% identity with the C-terminal region of Axin, and this related region is named the DIX domain. The functions of the DIX domains of Dvl-1 and Axin were investigated. By yeast two-hybrid screening, the DIX domain of Dvl-1 was found to interact with Dvl-3, a second mammalian Dishevelled relative. The DIX domains of Dvl-1 and Dvl-3 directly bind one another. Furthermore, Dvl-1 forms a homo-oligomer. Axin also forms a homo-oligomer, and its DIX domain is necessary. The N-terminal region of Dvl-1, including its DIX domain, bind to Axin directly. Dvl-1 inhibits Axin-promoted glycogen synthase kinase 3beta-dependent phosphorylation of beta-catenin, and the DIX domain of Dvl-1 is required for this inhibitory activity. Expression of Dvl-1 in L cells induces the nuclear accumulation of beta-catenin, and deletion of the DIX domain abolishes this activity. Although expression of Axin in SW480 cells causes the degradation of beta-catenin and reduces the cell growth rate, expression of an Axin mutant that lacks the DIX domain does not affect the level of beta-catenin or the growth rate. These results indicate that the DIX domains of Dvl-1 and Axin are important for protein-protein interactions and that they are necessary for the ability of Dvl-1 and Axin to regulate the stability of beta-catenin (Kishida, 1999).

Axin was identified as a regulator of embryonic axis induction in vertebrates that inhibits the Wnt signal transduction pathway. Epistasis experiments in frog embryos have indicated that Axin functions downstream of glycogen synthase kinase 3beta (GSK3beta) and upstream of beta-catenin, and subsequent studies have shown that Axin is part of a complex including these two proteins and adenomatous polyposis coli (APC). The roles of different Axin domains in the effects on axis formation and beta-catenin levels have been examined. The regulators of G-protein signaling domain (major APC-binding site) and GSK3beta-binding site are required, whereas the COOH-terminal sequences, including a protein phosphatase 2A binding site and the DIX domain, are not essential. Some forms of Axin lacking the beta-catenin binding site can still interact indirectly with beta-catenin and regulate beta-catenin levels and axis formation. Thus in normal embryonic cells, interaction with APC and GSK3beta is critical for the ability of Axin to regulate signaling via beta-catenin. Myc-tagged Axin is localized in a characteristic pattern of intracellular spots as well as at the plasma membrane. NH2-terminal sequences are required for targeting to either of these sites, whereas COOH-terminal sequences increase localization at the spots. Coexpression of hemagglutinin-tagged Dishevelled (Dsh) reveals strong colocalization with Axin, suggesting that Dsh can interact with the Axin/APC/GSK3/beta-catenin complex, and may thus modulate its activity (Fagotto, 1999).

Glycogen synthase kinase-3 (GSK-3) mediates epidermal growth factor, insulin and Wnt signals to various downstream events such as glycogen metabolism, gene expression, proliferation and differentiation. A GSK-3 beta-interacting protein has been isolated from a rat brain cDNA library using a yeast two-hybrid method. This protein consists of 832 amino acids and possesses Regulators of G protein Signaling (RGS) and Dishevelled (Dsh) homologous domains in its N- and C-terminal regions, respectively. The predicted amino acid sequence of this GSK-3beta-interacting protein shows 94% identity with mouse Axin, which recently has been identified as a negative regulator of the Wnt signaling pathway; therefore, this protein has been called rAxin (rat Axin). rAxin interacts directly with, and is phosphorylated by, GSK-3beta. rAxin also interacts directly with the armadillo repeats of beta-catenin. The binding site of rAxin for GSK-3beta is distinct from the beta-catenin-binding site, and these three proteins formed a ternary complex. Furthermore, rAxin promotes GSK-3beta-dependent phosphorylation of beta-catenin. These results suggest that rAxin negatively regulates the Wnt signaling pathway by interacting with GSK-3beta and beta-catenin and mediating the signal from GSK-3beta to beta-catenin (Ikeda, 1998).

Axin antagonizes the developmental effects of Wnt in vertebrates. Axin simultaneously binds two components of the Wnt pathway: beta-catenin and its negative regulator, glycogen synthase kinase-3beta. In mammalian cells, Axin inhibits Wnt-1 stimulation of beta-catenin/lymphoid enhancer factor 1-dependent transcription. Axin also blocks beta-catenin-mediated transcription in colon cancer cells that have a mutation in the adenomatous polyposis coli gene. These findings suggest that Axin, by forming a complex with beta-catenin and glycogen synthase kinase-3beta, can block signaling stimulated by Wnt or by adenomatous polyposis coli mutations (Sakanaka, 1998).

Regulation of beta-catenin degradation by intracellular components of the wnt pathway was reconstituted in cytoplasmic extracts of Xenopus eggs and embryos. The ubiquitin-dependent beta-catenin degradation in extracts displays a biochemical requirement for axin, GSK3, and APC. Axin dramatically accelerates while dishevelled inhibits beta-catenin turnover. Through another domain, dishevelled recruits GBP/Frat1 to the APC-axin-GSK3 complex. These results confirm and extend models in which inhibition of GSK3 has two synergistic effects: (1) reduction of APC phosphorylation and loss of affinity for beta-catenin and (2) reduction of beta-catenin phosphorylation and consequent loss of its affinity for the SCF ubiquitin ligase complex. Dishevelled thus stabilizes beta-catenin, which can dissociate from the APC/axin complex and participate in transcriptional activation (Salic, 2000).

The central mechanistic features of beta-catenin phosphorylation and ubiquitination have been well described in several systems by genetic approaches and by overexpression in cell lines and embryos. These studies have identified dsh, axin, GSK3, APC, and GBP/FRAT1 as important regulators of beta-catenin levels. However, there is disagreement as to whether APC acts in a positive or negative fashion; whether GBP plays an important role, or whether it just buffers GSK3; how the components interact; what the precise function of axin and APC is in regulating beta-catenin phosphorylation, and finally, the role of dsh in this process. To address these issues, cytoplasmic extracts from Xenopus eggs were used to reconstitute the regulated degradation of beta-catenin. By measuring beta-catenin stability directly rather than steady-state levels of the protein, the complicating effects of transcription and translation present in other systems were avoided. By adding known amounts of purified proteins to extracts and measuring degradation kinetics, a quantitative analysis of wnt signaling can be performed. This system also allows an examination of mutant proteins that might not show phenotypes if expressed in vivo, due to rapid turnover. In extracts, beta-catenin was degraded with a half-life of 1 hr, dropping to less than 15 min in the presence of added axin, which suggestes that axin is rate limiting for degradation. Addition of dsh, a positive regulator epistatically upstream of GSK3, stabilizes beta-catenin, suggesting that the wnt pathway from dsh to beta-catenin had been recapitulated in extracts (Salic, 2000).

This in vitro system allowed for a test of the biochemical requirement for several different components of the wnt pathway, ruling out the necessary involvement of any machinery upstream of protein translation. beta-catenin degradation depends on axin, GSK3, and APC; extracts depleted of any of these proteins cannot degrade beta-catenin. Of the four defined domains in axin, only the C-terminal DIX domain is dispensable for activity. The DIX domain of axin binds a related domain in dsh, and this interaction is required for dsh function; hence, binding of axin and dsh through the DIX domains confers signal-dependent regulation of beta-catenin. Another domain of dsh required for activity is PDZ, which binds GBP. Dsh and GBP synergize to inhibit beta-catenin degradation in extracts and beta-catenin phosphorylation in vitro, which are both axin-dependent processes. The inhibitory effect of GBP and dsh on beta-catenin degradation in extracts is reversed by high concentrations of GSK3 due to titration of dsh/GBP on axin. These observations further support a role for dsh, as an adaptor protein recruiting GBP to the axin/GSK3/APC/beta-catenin complex and inhibiting locally the enzymatic activity of GSK3. These recent results suggest that the axin-GSK3 interaction is highly dynamic and that increasing the local GBP concentration at the level of the complex efficiently promotes the dissociation of GSK3 from its site on axin (Salic, 2000).

The simplest model would have beta-catenin and GSK3 binding to their sites on axin and GBP binding to its site on dsh. When dsh and axin bind via their DIX domains, GSK3 would be inhibited. How does APC fit into this model? In addition to the mechanistic confusions about the role of APC, there are suggestions that APC can be an inhibitor of beta-catenin degradation in some circumstances or an activator in others. Specifically in Xenopus, injection of APC can cause secondary axis formation, the opposite of the ventralizing effects of axin injections. In an in vitro system, there is an absolute requirement for APC in beta-catenin degradation, consistent with genetic experiments in Drosophila and experiments in cultured cells. The confusion is likely to be simply due to the tendency of overexpressed APC to act as a dominant inhibitor. In Xenopus extracts, overexpressed APC is a potent inhibitor of beta-catenin degradation, although these experiments yield a mixture of truncated and full-length APC proteins. The role of APC in beta-catenin degradation is clear from the in vitro experiments. An interaction of APC with the RGS domain of axin is essential for beta-catenin degradation, demonstrating that endogenous APC is in fact a negative regulator of beta-catenin levels in Xenopus. Although beta-catenin binds to axin beads in a purified system, the two proteins show little interaction in the nanomolar range. Extracts contain an activity that promotes the binding of beta-catenin to axin at low concentrations. The activity is inhibited by the RGS domain of axin, suggesting that it is either APC or APC complexed to other proteins. Also, axinDeltaRGS (which cannot bind APC) does not bind beta-catenin in extracts. With purified components in vitro, APC accelerates the binding of beta-catenin to axin, recapitulating the stimulating effect of extracts on binding between the two proteins. Taken together, these results identify APC as the activity stimulating the axin-beta-catenin interaction in extracts (Salic, 2000).

beta-catenin interacts with numerous cellular proteins, some not directly involved in wnt signaling. The requirement for an APC-like molecule makes sense if one realizes that not only the stability of beta-catenin is important but also its availability for other interactions, both transcriptional and cytoskeletal. Aside from participating in beta-catenin degradation, APC also prevents the interaction of beta-catenin with Tcf3, thus inhibiting the transcriptional activation function of beta-catenin. Additionally, APC maintains a pool of beta-catenin in a state that can either lead to its degradation or release in response to a wnt signal. The phosphorylation state of APC is maintained by rapid futile cycles of opposing phosphorylation and dephosphorylation reactions. Changes in the phosphorylation of APC by GSK3 acts as a switch to trigger either beta-catenin degradation or its discharge in the cytoplasm. Although it has been suggested that phosphorylation of axin modulates binding to beta-catenin, no evidence could be found for such a regulation in extracts because (1) APC is the principal mediator of the axin-beta-catenin interaction and (2) using purified axin and beta-catenin, an increase in the affinity of beta-catenin for axin due to phosphorylation of the latter by GSK3 could not be detected. These data suggests that the APC/axin/GSK3/beta-catenin complex exists in two states. In one state, beta-catenin is bound tightly to axin via APC; beta-catenin and APC are both phosphorylated by GSK3. In this situation, phosphorylated beta-catenin can only be removed through SCF-dependent degradation while phosphorylated APC continues to bind beta-catenin molecules avidly. Upon binding of active dsh to axin, GBP interacts with GSK3, removing it from axin competitively. This inhibits phosphorylation on both beta-catenin and APC. The former stabilizes beta-catenin to ubiquitination, and the latter releases intact beta-catenin so that it can interact with other partners. The beta-catenin-binding site on axin is also important because its deletion impairs the function of axin. Although this scenario is written as a set of irreversible steps, it is likely that all binding events are dynamic. The fact that dominant-negative GSK3 blocks degradation of beta-catenin in extracts suggests that the interaction between endogenous axin and GSK3 must be dynamic. The dissociation rate for GSK3 bound to axin must be fast enough to allow its displacement by the dominant-negative mutant on a time scale of perhaps less than 1 hr. The same must be true for the axin-APC interaction, since the RGS domain of axin (which interacts with APC) blocks beta-catenin degradation (Salic, 2000).

The stability of beta-catenin is subject to a system of weak protein interactions and posttranslational modifications. The ability to dissect this pathway in vitro without altering appreciably its kinetics and without simplifying essential steps should aid an understanding of these weak interactions and their physiological consequences. Several remaining questions should be clarified by studies in partially purified systems and ultimately by reconstitution from purified components. These include the mechanism of signal transmission from frizzled to dsh and the mechanism of dsh activation. In particular, does the wnt signal regulate the GBP-dsh or the dsh-axin interaction? It will also be of interest to know whether APC suffices by itself to mediate beta-catenin degradation or whether associated proteins are also required. Finally, it will be of interest to know whether signaling pathways cross-regulate the wnt pathway and how that might work. In vitro reconstitution should be useful in deciphering the molecular events involved in transducing the wnt signal (Salic, 2000).

Using a yeast two-hybrid method, a novel protein has been identified that interacts with glycogen synthase kinase 3beta (GSK-3beta) and has 44% amino acid identity with Axin, a negative regulator of the Wnt signaling pathway. This protein has been termed Axil, for Axin like. Like Axin, Axil ventralizes Xenopus embryos and inhibits Xwnt8-induced Xenopus axis duplication. Axil is phosphorylated by GSK-3beta. Axil binds not only to GSK-3beta but also to beta-catenin; the GSK-3beta-binding site of Axil is distinct from the beta-catenin-binding site. Axil enhances GSK-3beta-dependent phosphorylation of beta-catenin. These results indicate that Axil negatively regulates the Wnt signaling pathway by mediating the GSK-3beta-dependent phosphorylation of beta-catenin, thereby inhibiting axis formation (Yamamoto, 1998).

Control of stability of beta-catenin is central in the Wnt signaling pathway. The protein Conductin is found to form a complex with both beta-catenin and the tumor suppressor gene product adenomatous polyposis coli (APC). Conductin induces beta-catenin degradation, although conductin mutants are deficient in complex formation stabilize beta-catenin. Fragments of APC that contain a conductin-binding domain also block beta-catenin degradation. Thus, conductin is a component of the multiprotein complex that directs beta-catenin to degradation and is located downstream of APC. In Xenopus embryos, conductin interfers with wnt-induced axis formation (Behrens, 1998).

Dysregulation of Wnt-beta-catenin signaling disrupts axis formation in vertebrate embryos and underlies multiple human malignancies. The adenomatous polyposis coli (APC) protein, axin, and glycogen synthase kinase 3beta form a Wnt-regulated signaling complex that mediates the phosphorylation-dependent degradation of beta-catenin. A protein phosphatase 2A (PP2A) regulatory subunit, B56, interacts with APC in the yeast two-hybrid system. Expression of B56 reduces the abundance of beta-catenin and inhibits transcription of beta-catenin target genes in mammalian cells and Xenopus embryo explants. The B56-dependent decrease in beta-catenin is blocked by oncogenic mutations in beta-catenin or APC, and by proteasome inhibitors. B56 may direct PP2A to dephosphorylate specific components of the APC-dependent signaling complex and thereby inhibit Wnt signaling. Loss of PP2A function may provide an additional route to activation of Wnt signaling and oncogenesis. Consistent with this, mutations in the gene encoding the beta isoform of the PP2A subunit have been identified in colon and lung cancers (Seeling, 1999).

The apical ectodermal ridge (AER) is an essential structure for vertebrate limb development. Wnt3a is expressed during the induction of the chick AER, and misexpression of Wnt3a induces ectopic expression of AER-specific genes in the limb ectoderm. The genes beta-catenin and Lef1 can mimic the effect of Wnt3a, and blocking the intrinsic Lef1 activity disrupts AER formation. Hence, Wnt3a functions in AER formation through the beta-catenin/LEF1 pathway. In contrast, neither beta-catenin nor Lef1 affects the Wnt7a-regulated dorsoventral polarity of the limb. Thus, two related Wnt genes elicit distinct responses in the same tissues by using different intracellular pathways (Kengaku, 1998).

Various structural components of intercellular junctions have recently been found to represent tumor-suppressor genes (or be related to their products). The tumor-suppressor gene product adenomatous polyposis coli (APC) binds to beta 2-Catenin which is cytoplasmically associated with the cell adhesion molecule E-cadherin (Hulsken, 1994).

Human Tcf-4, a Tcf family member that is expressed in colonic epithelium, transactivates transcription of an artifical promoter only when associated with ß-catenin. Nuclei of cells mutant for Adenomatous polyposis coli tumor suppressor protein are found to contain a stable ß-catenin hTcf-4 complex that is constitutively active. Reintroduction of APC removes ß-catenin from hTcf-4 and abrogates the transcriptional transactivation. Constitutive transcription of Tcf target genes, caused by loss of APC function, may be a crucial event in the early transformation of colonic epithelium (Korinek, 1997).

A novel member of the human frizzled (Fz) gene family was cloned and found to be specifically expressed in 3 of 13 well differentiated (23%), 13 of 20 moderately differentiated (62%), and 12 of 14 poorly differentiated (86%) squamous cell esophageal carcinomas compared with the adjacent uninvolved normal mucosa. The FzE3 cDNA encodes a protein of 574 amino acids and shares high sequence homology with the human FzD2 gene, particularly in the putative ligand binding region of the cysteine-rich extracellular domain. Functional analysis reveals that transfection and expression of the FzE3 cDNA in esophageal carcinoma cells stimulates complex formation between adenomatous polyposis coli (APC) and beta-catenin followed by nuclear translocation of beta-catenin. Furthermore, cotransfection of a mutant construct encoding a FzE3 protein with a C-terminal truncation completely inhibits the interaction of APC with beta-catenin in cells. Finally, coexpression of FzE3 with Lef-1 transcription factor enhances beta-catenin translocation to the nucleus. These observations suggest that FzE3 gene expression may down-regulate APC function and enhance beta-catenin mediated signals in poorly differentiated human esophageal carcinomas (Tanaka, 1998).

The APC tumor suppressor protein plays a critical role in regulating cellular levels of the oncogene product ß-catenin. APC binds to ß-catenin through a series of homologous 15 and 20 amino acid repeats. The crystal structure of a 15 amino acid ß-catenin binding repeat has been determined from APC bound to the armadillo repeat region of ß-catenin. Although it lacks significant sequence homology, the N-terminal half of the repeat binds in a manner similar to portions of E-cadherin and XTcf3, but the remaining interactions are unique to APC. Evidence from structural, biochemical and sequence data is presented, which suggests that the 20 amino acid repeats can adopt two modes of binding to ß-catenin (Spink, 2001).

The homology between the N-terminal regions of the APC 15 and 20mer repeats suggests that the APC 20mers may bind to ß-catenin by a mechanism similar to that of APC-rA (the 15mer repeat A from APC) binding. Alignment of the core homology region of the 20mers with that of the 15mers results in a reasonable alignment with E-cadherin and XTcf3, even outside of the core homology region. Serines from the 20mer C-terminus (e.g. hAPC-1 Ser1276 and 1278) align with the first two phosphoserines seen in the E-cadherin structure (pSer684 and 686) and two glutamic acid residues from XTcf3 (Glu26 and 28). This alignment is consistent with data on the binding of an APC 20mer construct to a series of ß-catenin point mutants. The two mutations that eliminate APC 20mer binding (Lys345Ala and Trp383Ala) are in residues that interact with Glu1034 of APC-rA, at the C-terminus of the 15mer peptide (Spink, 2001).

In order to assess whether the 15 and 20mer repeats of APC bind at the same site on ß-catenin, whether the two classes of repeats could compete for binding to ß-catenin was tested. A construct containing 15mer repeat A (APC-fA) can compete with binding of a construct containing two 20mer repeats (APC-2,3). The efficiency of the observed competition (with significant reduction of APC-2,3 binding even at a 1:1 ratio with APC-fA), suggests an extensive overlap between the binding sites of the two proteins. However, it remains possible that the 15 and 20mer binding sites are distinct but overlapping, or that presentation of the binding repeats within the larger constructs may result in steric clashes, even if their binding sites do not overlap (Spink, 2001).

The theory that the core homology regions of the 15 and 20mer repeats bind at the same site differs from a proposed mechanism for 20mer binding derived from the ß-catenin- E-cadherin complex structure. In that structure, a Ser-Leu-Ser-Ser-Leu (SLSSL) sequence from E-cadherin binds to ß-catenin in a phosphorylation-dependent manner, with phosphoserines at consensus GSK3ß sites within and just N-terminal to the SLSSL sequence interacting with ß-catenin. An SLSSL sequence is conserved in the C-terminal region of the 20 amino acid ß-catenin binding repeats of APC, suggesting that this region of the 20 amino acid repeats can interact with the SLSSL binding region. This hypothesis is consistent with mutagenesis experiments, which show that the binding of ß-catenin to an APC construct containing both 15 and 20mer repeats is affected by mutations in the SLSSL binding site of ß-catenin. Since several hydrophobic residues from the SLSSL region of E-cadherin make contacts with ß-catenin, binding at the SLSSL site need not be entirely dependent upon phosphorylation. However, phosphorylation of the GSK3ß consensus sites in E-cadherin increases its affinity for ß-catenin 1000-fold. Likewise, GSK3ß phosphorylation of the 20 amino acid repeat region of APC has been shown to increase its affinity for ß-catenin (Spink, 2001).

Alignment of the SLSSL sequence of an APC 20mer (e.g. hAPC-1, residues 1278-82) with that from E-cadherin (residues 690-694) results in the alignment of a conserved 20mer serine (hAPC-1 Ser1272) with an upstream GSK3ß site in E-cadherin (pSer686). This serine corresponds to the last residue of the 20mer core homology region. However, E-cadherin pSer686 binds to ß-catenin >18 Å from the end of the core homology region-binding site (APC-rA Ser1028). Thus, it would be impossible for a single 20mer repeat to bind by both mechanisms simultaneously. Since the structural, sequence and mutagenesis data provide evidence for each binding mechanism it is suggested that the 20mers can adopt two distinct binding modes, and that phosphorylation may act as a switch between these modes (Spink, 2001).

What might be the role of two binding modes in ß-catenin degradation? Although the role of APC in the degradation complex is not fully understood, it is likely to include sequestration of ß-catenin from Lef/Tcf before degradation and presentation of ß-catenin for phosphorylation and degradation. The two binding modes could simply increase the number of potential binding sites, and hence the effective affinity between APC and ß-catenin. Alternatively, the two binding modes could have distinct roles; for example, the unphosphorylated 20mers, in conjunction with the 15mers, may mediate the initial binding and sequestering of ß-catenin, whereas subsequent phosphorylation of the 20mers and a switch in their mode of binding to ß-catenin could be important for efficient presentation of ß-catenin to GSK3ß. The existence and roles, if any, of these alternatives await further experimentation (Spink, 2001).

Although Apc is well characterized as a tumor-suppressor gene in the intestine, the precise mechanism of this suppression remains to be defined. Using a novel inducible Ahcre transgenic line in conjunction with a loxP-flanked Apc allele, loss of Apc is shown to acutely activate Wnt signaling through the nuclear accumulation of ß-catenin. Coincidentally, it perturbs differentiation, migration, proliferation, and apoptosis, such that Apc-deficient cells maintain a 'crypt progenitor-like' phenotype. Critically, a series of Wnt target molecules has been confirmed in an in vivo setting, and a series of new candidate targets has been identified within the same setting (Sansom, 2004).

ß-catenin levels were examined within the Cre+Apcfl/fl tissue at day 5. There was no increase in total ß-catenin in the Cre+Apcfl/fl samples. However, levels of dephosphorylated ß-catenin were moderately elevated and, crucially, ß-catenin relocalized to the nuclei in the Cre+Apcfl/fl tissue. To more precisely define the time scale of nuclear relocalization, immunohistochemical analyses were performed at days 1, 2, 3, and 4 following induction of the cre recombinase. This analysis showed that relocalization occurred at day 3, and this was coincident with the observed onset of changes in morphology, proliferation, and apoptosis (Sansom, 2004).

To test whether nuclear ß-catenin was activating transcription of its known target genes, microarray analysis was performed using the affymetrix U74A chip. RNA samples were derived from sibling Cre+Apcfl/fl and Cre+Apc+/+ mice given four daily injections of ß-napthoflavone and killed at days 4 and 5. Of the 100 most significantly up-regulated genes, 10 have been associated with Wnt signaling (either directly or through arrays that had examined targets of the ß-catenin/TCF4 complex). Of the comparable genes up-regulated at day 5, 45 of 47 showed increases, of which 36 were in excess of twofold. These 36 included c-Myc, CD44, Tiam 1, Sema3c, and EphB3, all of which were confirmed changes at day 5. These data are, therefore, consistent with the notion that these are important early changes following nuclear relocalization of ß-catenin. Use of the larger chip set also revealed up-regulation of other Wnt target genes at day 4, including Sox17 and Axin2 (Sansom, 2004).

To validate the results obtained from the microarray analysis, the expression pattern of a subset of dysregulated genes was examined. Up-regulation of CD44, C-Myc, laminin gamma2, EphB2, and EphB3 was confirmed immunohistochemically in the Apc-deficient tissue. Expression of the EphrinB2 ligand, which is normally restricted to the top of the crypts and villi, was reduced in concordance with the reduction in villus differentiation in the Cre+Apcfl/fl tissue. These data therefore confirm, in an in vivo setting, many of the targets of Wnt signaling that have been implicated from in vitro studies. These include up-regulation of CD44, c-Myc, MMP-7 (matrilysin), gamma-2 laminin, Sema3c (confirmed by RT-PCR) Ets-2, EphB2, EphB3, and GPR49. The array analysis also indicates up-regulation of a series of genes that either interact with CD44 or are targets of CD44. These include MMP-7, TIAM1, FGF4 and its receptor, and TASR-2. The up-regulation of TIAM1 is particularly interesting, since TIAM1 has been shown to mediate Ras signaling. Indeed, mice deficient in TIAM1 are resistant to Ras-induced skin tumors (Sansom, 2004).

Deficiency of EphB3 has been shown to lead to abnormal Paneth cell positioning in the crypt. Wnt-mediated up-regulation of EphB3 yields a similar Paneth cell phenotype, confirming a pivotal role for the EphB/ephrinB mutual repulsion system in defining crypt-villus architecture. These results are also consistent with the notion that Apc mutant cells express the same genetic program as cells at positions 1-2 of the crypt, with notable increases in EphB3, MMP7, and Pla2g2a being characteristic of both Paneth cells and the Apc-deficient cells described in this study (Sansom, 2004).

In summary, Apc has been shown to be a critical determinant of cell fate in the murine small intestinal epithelium. Acute activation of Wnt signaling immediately produces many of the phenotypes associated with early colorectal lesions: failed differentiation, increased proliferation, and aberrant migration. Within a short time scale, multiple processes are affected: interactions with the cellular matrix, interactions with the basement membrane, increased proliferation, and failure of positional cues (EphB/ephrinB) (Sansom, 2004).

The adenomatous polyposis coli (APC) protein is inactivated in most colorectal tumours. APC loss is an early event in tumorigenesis, and causes an increase of nuclear ß-catenin and its transcriptional activity. This is thought to be the driving force for tumour progression. APC shuttles in and out of the nucleus, but the functional significance of this has been controversial. APC truncations have been shown to be nuclear in colorectal cancer cells and adenocarcinomas, and this correlates with loss of centrally located nuclear export signals. These signals confer efficient nuclear export as measured directly by fluorescence loss in photobleaching (FLIP), and they are critical for the function of APC in reducing the transcriptional activity of ß-catenin in complementation assays of APC mutant colorectal cancer cells. Importantly, targeting a functional APC construct to the nucleus causes a striking nuclear accumulation of ß-catenin without changing its transcriptional activity. This evidence indicates that the rate of nuclear export of APC, rather than its nuclear import or steady-state levels, determines the transcriptional activity of ß-catenin (Rosin-Arbesfeld, 2003).

Wnt signaling plays an important role in both oncogenesis and development. Activation of the Wnt pathway results in stabilization of the transcriptional coactivator ß-catenin. Recent studies have demonstrated that axin, which coordinates ß-catenin degradation, is itself degraded. Although the key molecules required for transducing a Wnt signal have been identified, a quantitative understanding of this pathway has been lacking. This study developed a mathematical model for the canonical Wnt pathway that describes the interactions among the core components: Wnt, Frizzled, Dishevelled, GSK3ß, APC, axin, ß-catenin, and TCF. Using a system of differential equations, the model incorporates the kinetics of protein-protein interactions, protein synthesis/degradation, and phosphorylation/dephosphorylation. Initially a reference state of kinetic, thermodynamic, and flux data was defined from experiments using Xenopus extracts. Predictions based on the analysis of the reference state were used iteratively to develop a more refined model from which the effects of prolonged and transient Wnt stimulation were analyzed on ß-catenin and axin turnover. Several unusual features of the Wnt pathway were predicted, some of which were tested experimentally. An insight from this model, which was confirmed experimentally, is that the two scaffold proteins axin and APC promote the formation of degradation complexes in very different ways. This study could also explain the importance of axin degradation in amplifying and sharpening the Wnt signal, and it was shown that the dependence of axin degradation on APC is an essential part of an unappreciated regulatory loop that prevents the accumulation of ß-catenin at decreased APC concentrations. By applying control analysis to the mathematical model, this study demonstrates the modular design, sensitivity, and robustness of the Wnt pathway and derives an explicit expression for tumor suppression and oncogenicity (Lee, 2003).

Theory and quantitation are mutually dependent activities. It would seem unlikely that one would go to the trouble to measure detailed kinetic quantities without a specific model to test, and it is equally unlikely that realistic models can be constructed without the constraints of quantitative experimental data. The intent in trying to reproduce a substantial part of the Wnt pathway in Xenopus egg extracts was to acquire the kind of detailed kinetic data required to build a realistic model. There are several unusual advantages to the extract system that contributed to this effort. The Xenopus egg extract is essentially neat cytoplasm; it reproduces the in vivo rate of β-catenin degradation and responds to known regulators as expected from in vivo experiments. Kinetic experiments with high time resolution are possible in this system, since a well-stirred extract is presumably synchronous in ways in which collections of cells may not be. In extracts it is possible to precisely set the level of components by depletion or addition. The direct output of the canonical Wnt pathway is an easily measured cytoplasmic event, the degradation of β-catenin. Thus, in this unusual system it is possible to acquire quantitative information about signaling pathways, not achievable in vivo. At the same time, these extracts have limitations. The receptor events were not considered, and it is likely that reactions at the plasma membrane contribute to dynamic features. Also, the analysis is incomplete, since there are other components of Wnt signaling, such as casein kinase Iδ, casein kinase Iɛ, and PAR1, as well as cross-talk from other pathways, that influence the behavior of the system. The multiple phosphorylation steps were oversimplified. A simple interconversion of the phosphorylated and unphosphorylated complex of axin, APC, and GSK3β was assumed, whereas in reality multiple phosphorylation states exist within the complex; the states may be random or sequential. The information needed to provide a much more specific model of phosphorylation interconversions is not available at this time, although the model could easily be extended. Finally, there is the question of what Wnt process are being studied. Events in the cytoplasm of unfertilized eggs are being examined. Though endowed with all of the core components of the Wnt pathway, the egg is, as far as is known, transcriptionally silent and not involved in Wnt signaling, though this system is active very soon in embryogenesis. Thus, there is no biological in vivo behavior with which to compare the in vitro behavior. Nevertheless, the basic core circuitry is intact and is presumably prepared for the early Wnt events in the embryo. All the properties of the egg extract system are very similar to that circuitry in vertebrate somatic cells (Lee, 2003).

To build a mathematical description of the Wnt signaling system, Lee (2003) started with the basic circuitry discerned from previous studies in Xenopus embryos and mammalian cells, whose similarity to the in vitro system they had already confirmed. A system of differential equations was derived that described the time-dependent variations of the system variables, i.e., the concentrations of the pathway components and their complexes. Parameters of the model are binding constants of proteins, rate constants of phosphorylations and dephosphorylations, rate constants of protein degradation, and rates of protein synthesis. Model reduction was achieved by considering conservation relations and by applying rapid equilibrium approximations for selected binding processes. Despite these simplifications, the model consists of a nonlinear system of differential equations whose solution requires the use of computers. Not all of these parameters were accessible to measurement. To circumvent this problem, not only kinetic parameters characterizing individual steps were used as primary inputs, but quantities that are more easily accessible from experiment, such as the overall flux of β-catenin degradation. This allowed rate constants to be derived as well as protein concentrations in a reference state, where there was no Wnt signal. This state serves as a starting point for predicting the system behavior during Wnt signaling as well as after experimental perturbations (Lee, 2003).

The basic model reproduced quantitatively the behavior of the reference state, including perturbations of this state achieved by varying the concentration of axin, GSK3β, and TCF. It also reproduced extensions of this to the signaling state. A wide variety of different sets of experimental data could be simulated by the same model, employing the same sets of kinetic parameters. This process was approached iteratively. For example, the early model did not include nonaxin-dependent degradation of β-catenin, but inclusion of this process improved the fit to the experimental data. More significantly, addition of this process had interesting biological implications, which are discussed (Lee, 2003).

In many ways, one of the most peculiar findings was the very low concentration of axin in the Xenopus extracts. Axin levels in other organisms may similarly be very low: Drosophila axin can be detected by Western blotting only following its immunoprecipitation. Although theoretical and experimental studies have shown that axin is inhibitory at high concentrations, both indicate that axin is not present at the optimal concentration for the highest rate of β-catenin degradation. Therefore, axin levels are not set for optimality of β-catenin degradation, but are presumably optimized for some other purpose. Theoretically, axin levels must be held below the very sharp threshold of Dsh inhibition. Experimentally, these thresholds, which blunt Wnt signaling, are observed but are not as sharp as expected, and this may indicate some other compensatory effects. These thresholds would limit axin concentration to well below 1 nM if activated Dsh were constrained to concentrations of below 1 μM. Under these circumstances, it can be expected that axin would never be found at concentrations approaching those of other Wnt pathway components (50-100 nM) (Lee, 2003).

The low concentration of axin relative to other components (such as GSK3β, Dsh, and APC) has another design feature potentially very general and important for the modularity of metazoan signaling pathways. Axin is a critical node point for controlling β-catenin levels, but it also interacts with components shared with several other important pathways. The interaction of these components with axin fluctuates due to Wnt signals (reflecting changes in binding as well as changes in axin levels), yet because the concentration of axin is so low, there will be no appreciable change in the overall levels of GSK3β, Dsh, or APC (all these components important in other pathways would otherwise be driven to fluctuate). The very low axin concentration thus isolates the Wnt pathway from perturbing other systems, a simple mechanism to achieve modularity. Other scaffold proteins may serve similar functions in other pathways. These insights follow from a very simple measurement of axin concentration and suggest the utility of measuring the levels of signaling pathway components in different cell types and circumstances. Since quantitative and kinetic features may be important in defining modules, it suggests that qualitative circuit diagrams of signal transduction may overlook very important design features. Modularity within the Wnt pathway can be defined by an extension of a summation theorem which argues that the steady state of an entire pathway would have control coefficients that added to zero. When the Wnt pathway is broken down to several subpathways, it is found that within these subpathways the control coefficients would sum to zero at steady state. While some of this subdivision is obvious (i.e., the kinase phosphatase module involving the phosporylation of APC and axin complexed to GSK3β), in other cases, such as the β-catenin module, it is much less obvious. Here the reactions include the phosphorylation of β-catenin in the APC/axin/GSK3β complex, the release and degradation of β-catenin, and the synthesis and nonaxin-dependent degradation of β-catenin. Balanced perturbation of these subpathways as a whole will not affect the overall flux of β-catenin degradation. It is not clear whether this concept of modularity might be extended usefully in two other directions: modularity in systems not at steady state, i.e., with transients, and estimates of linkage between pathways by some definition of nonzero summations expressing the degree of independence or modularity (Lee, 2003).

This paper marks one of the first extensions of metabolic control theory to signal transduction. Metabolism and signal transduction seem very different, the former involving the transfer of mass and the latter the transfer of information. In addition, metabolic pathways generally involve dedicated components and the specificity of interaction of substrates and enzymes is very high. Signaling pathways share many components; interactions are often weak. Metabolism, which has had a long history of quantitative study, was a natural field for the development of control theory, and this theory has been successful in converting the specific information about the behavior of enzymes in a pathway to the overall behavior of metabolic circuits. Control coefficients are useful measures of the impact of one process or quantity on another. In its application to metabolism, it allowed erroneous concepts, such as the notion of a rate-determining step, to be disposed. In signal transduction, control coefficients might play a similar role. Here they can be used to indicate quantitively the effects of a particular reaction on some other property, such as flux through the pathway or concentration of another component. By this definition, certain rate constants, such as the phosphorylation and dephosphorylation of APC and axin, have a major influence on the levels of β-catenin, while others, such as the degradation rate of phosphorylated β-catenin, have little effect. The sign and magnitude of these control coefficients give some indication what gene products could be oncogenes or tumor suppressors. By this criterion APC, GSK3β, and axin are potent tumor suppressors, whereas β-catenin is an oncogene. Dsh would be expected to exert only moderate oncogenic effects. Clearly the effects of certain gene products are dependent on context, including their rate of synthesis and steady-state concentration. As understanding of pathways improves, the effect of mutation or pharmacologic inhibition could be estimated quantitatively using control coefficients. The differences between cell types and organisms could be exploited to better predict mutagenic and pharmacologic impact on signal propagation (Lee, 2003).

Despite considerable progress in identifying components of the Wnt pathway, many important mechanistic details are still lacking. This analysis has shown that Dsh seems to act to prevent the phosphorylation of the axin/APC complex, not the phosphorylation of β-catenin. Dsh (complexed to the GSK-3 binding protein FRAT1 or GBP, which has no Drosophila homolog) does not seem to be a general GSK3β inhibitor, like Li+, but rather is focused on the two scaffolding proteins. This was apparent from the biphasic nature of both the theoretical and experimental curves, which suggests that Dsh inhibits the rephosphorylation of axin/APC, but still allows many cycles of β-catenin phosphorylation, ubiquitination, and degradation. This mechanism was further proven by a timing-of-addition experiment. It needs to be further confirmed and extended by looking specifically at individual phosphorylation sites on all the components of the complex. Another insight into the mechanisms of complex formation and control of β-catenin degradation concerns the inhibition of β-catenin degradation at concentrations of axin approaching those of other components. This suggests that axin binds APC, GSK3β, and β-catenin in random order. The axin concentration is limited by other factors; owing to the low concentration of axin, random binding is not likely ever to be a problem. The situation for APC seems very different. The concentration of APC is comparable to that of the other components. Overexpression studies show no inhibitory effects. These theoretical and experimental observations suggest that APC as a scaffold must be very different from axin as a scaffold. Most likely, APC binds components in an ordered manner (Lee, 2003).

Although inappropriate activation of the Wnt/ß-catenin pathway has been implicated in the development of hepatocellular carcinoma (HCC), the role of this signaling in liver carcinogenesis remains unclear. To investigate this issue, a mutant mouse strain, Apc(lox/lox), was constructed in which exon 14 of the tumor-suppressor gene adenomatous polyposis coli (Apc) is flanked by loxP sequences. i.v. injection of adenovirus encoding Cre recombinase (AdCre) at high multiplicity inactivated the Apc gene in the liver and resulted in marked hepatomegaly, hepatocyte hyperplasia, and rapid mortality. ß-Catenin signaling activation was demonstrated by nuclear and cytoplasmic accumulation of ß-catenin in the hepatocytes and by the induction of ß-catenin target genes (glutamine synthetase, glutamate transporter 1, ornithine aminotransferase, and leukocyte cell-derived chemotaxin 2) in the liver. To test a long-term oncogenic effect, mice were inoculated with lower doses of AdCre, compatible with both survival and persistence of ß-catenin-activated cells. In these conditions, 67% of mice developed HCC. ß-Catenin signaling was strongly activated in these Apc-inactivated HCCs. The HCCs were well, moderately, or poorly differentiated. Indeed, their histological and molecular features mimicked human HCC. Thus, deletion of Apc in the liver provides a valuable model of human HCC, and, in this model, activation of the Wnt/ß-catenin pathway by invalidation of Apc is required for liver tumorigenesis (Colnot, 2004).

APC plays a critical role in the Wnt signaling pathway by tightly regulating ß-catenin turnover and localization. The central region of APC is responsible for APC-ß-catenin interactions through its seven 20 amino acid (20aa) repeats and three 15 amino acid (15aa) repeats. Using isothermal titration calorimetry, the binding affinities were determined of ß-catenin with an APC 15aa repeat fragment and each of the seven 20aa repeats in both phosphorylated and unphosphorylated states. Despite sequence homology, different ß-catenin binding repeats of APC have dramatically different binding affinities with ß-catenin and thus may play different biological roles. The third 20aa repeat is by far the tightest binding site for ß-catenin among all the repeats. The fact that most APC mutations associated with colon cancers have lost the third 20aa repeat underlines the importance of APC-ß-catenin interaction in Wnt signaling and human diseases. For every 20aa repeat, phosphorylation dramatically increases its binding affinity for ß-catenin, suggesting phosphorylation has a critical regulatory role in APC function. In addition, CD and NMR studies demonstrate that the central region of APC is unstructured in the absence of ß-catenin and Axin, and suggest that ß-catenin may interact with each of the APC 15aa and 20aa repeats independently (Liu, 2006).

The APC tumor suppressor controls the stability and nuclear export of β-catenin (β-cat), a transcriptional coactivator of LEF-1/TCF HMG proteins in the Wnt/Wg signaling pathway. β-cat and APC have opposing actions at Wnt target genes in vivo. The β-cat C-terminal activation domain associates with TRRAP/TIP60 and mixed-lineage-leukemia (MLL1/MLL2) SET1-type chromatin-modifying complexes in vitro, and β-cat promotes H3K4 trimethylation at the c-Myc gene in vivo. H3K4 trimethylation in vivo requires prior ubiquitination of H2B, and ubiquitin is found necessary for transcription initiation on chromatin but not nonchromatin templates in vitro. Chromatin immunoprecipitation experiments reveal that β-cat recruits Pygopus, Bcl-9/Legless, and MLL/SET1-type complexes to the c-Myc enhancer together with the negative Wnt regulators, APC, and βTrCP. Interestingly, APC-mediated repression of c-Myc transcription in HT29-APC colorectal cancer cells is initiated by the transient binding of APC, βTrCP, and the CtBP corepressor to the c-Myc enhancer, followed by stable binding of the TLE-1 and HDAC1 corepressors. Moreover, nuclear CtBP physically associates with full-length APC, but not with mutant SW480 or HT29 APC proteins. It is concluded that, in addition to regulating the stability of β-cat, APC facilitates CtBP-mediated repression of Wnt target genes in normal, but not in colorectal cancer cells (Sierra, 2006).

The data presented here support a model in which the APC tumor suppressor functions directly to counteract β-cat-mediated transcription at Wnt target genes in vivo. This possibility was first suggested by the finding that full-length APC cycles on and off the c-Myc enhancer in conjunction with β-cat and associated coactivators in LiCl-treated C2C12 cells. In contrast, the enhancer complex appears to be stable and does not cycle in HT29 CRC cells, which contain a Class II APC mutant protein that is unable to degrade β-cat. Most strikingly, the binding of the full-length APC protein to the c-Myc gene in HT29-APC cells correlates with the rapid disassembly of the Wnt enhancer complex in vivo and the subsequent decline in steady-state c-Myc mRNA levels, both of which significantly precede the drop in β-cat protein levels that occurs as a result of proteolytic degradation in the cytoplasm. Thus, the effect of APC on c-Myc transcription appears to be immediate and direct, and may serve to coordinate the switch between the β-cat coactivator and TLE1 corepressor complexes (Sierra, 2006).

The β-cat enhancer complex includes the Wnt coactivators Pygopus and Bcl-9/Lgs, which control the retention of β-cat in the nucleus and may also function directly in transcription. The observation that APC can also regulate nuclear transport of β-cat raises the possibility that these factors may reside within a larger regulatory complex that chaperones β-cat in and out of the nucleus and mediates its release from the DNA. Indeed, sequential ChIP (re-ChIP) data indicate that the mutant APC in HT29 colorectal cancer cells exists in a stable complex with β-cat and LEF-1 at the active c-Myc gene. This finding is unexpected because β-cat cannot bind simultaneously to APC and LEF-1, and thus, if the full-length APC is part of a larger β-cat:LEF enhancer complex, it may interact with other subunits. Alternatively, the full-length APC and β-cat may exist in different complexes that rapidly exchange at the enhancer. The current data indicate that targeting is mediated by the N-terminal half of the APC protein, and that CtBP and βTrCP appear only in conjunction with the full-length APC protein. How APC is recruited to Wnt enhancers remains an open and important question (Sierra, 2006).

The ChIP experiments also suggest that APC-mediated inhibition of c-Myc transcription in HT29 cells occurs in two steps, initiated by transient binding of APC, βTrCP, CtBP, and YY1 to the enhancer, and followed by stable binding of the TLE-1 and HDAC1 corepressors. The transient recruitment of APC and CtBP, at the time when β-cat, Bcl-9, Pygo, and other Wnt enhancer factors leave the DNA, strongly suggests a role for these factors in the exchange of Wnt coactivator and corepressor complexes. In this respect it is interesting that CtBP was shown recently to associate with APC, both in vivo and in vitro. The results confirm a high-affinity interaction between CtBP and the full-length APC protein induced in HT29-APC cells, as well as with the native (full-length) APC protein in 293 cells. Consequently, APC may function to recruit CtBP to Wnt enhancers. Although both CtBP and TLE-1 are well-established corepressors of Wnt target genes, the different functions of the two types of corepressors remain unclear, and the ChIP data suggest that they act at distinct steps. Together, these data suggest that APC counteracts β-cat function in the nucleus, as well as in the cytoplasm, and may facilitate turnover of the enhancer complex at responsive genes by recruiting βTrCP and CtBP (Sierra, 2006).

The Wnt pathway is a conserved signal transduction pathway that contributes to normal development and adult homeostasis, but is also misregulated in human diseases such as cancer. The tumor suppressor Adenomatous Polyposis Coli (APC) is an essential negative regulator of Wnt signaling inactivated in over 80% of colorectal cancers. APC participates in a multi-protein 'destruction complex' that targets the proto-oncogene β-catenin for ubiquitin-mediated proteolysis; however, the mechanistic role of APC in the destruction complex remains unknown. Several models of APC function have recently been proposed, many of which have emphasized the importance of phosphorylation of high affinity β-catenin binding-sites (20 amino acid repeats; 20Rs) on APC. This study tested these models by generating a Drosophila APC2 mutant lacking all β-catenin binding 20Rs and performing functional studies in human colon cancer cell lines and Drosophila embryos. The results are inconsistent with current models, as it was found that β-catenin binding to the 20Rs of APC is not required for destruction complex activity. In addition, an APC2 mutant was generated lacking all β-catenin binding-sites (including the 15Rs), and a direct β-catenin/APC interaction was found to be also not essential for β-catenin destruction, although it increases destruction complex efficiency in certain developmental contexts. Overall, these findings support a model whereby β-catenin binding sites on APC do not provide a critical mechanistic function per se, but rather dock β-catenin in the destruction complex to increase the efficiency of β-catenin destruction. Furthermore, in Drosophila embryos expressing some APC2 mutant transgenes a separation of β-catenin destruction and Wg/Wnt signaling outputs was observed, and it is suggested that cytoplasmic retention of β-catenin likely accounts for this difference (Yamulla, 2004).

Protein interactions of Armadillo homologs: Nuclear transport and interaction with TCF

Control of the nuclear localization of specific proteins is an important mechanism for regulating many signal transduction pathways. When the Wnt signaling pathway is activated, beta-catenin localizes to the nucleus and interacts with TCF/LEF-1 (T-cell factor/lymphocyte enhancer factor-1) transcription factors, triggering activation of downstream genes. The role of regulated nuclear localization in beta-catenin signaling is still unclear. beta-catenin has no nuclear localization sequence (NLS). Although it has been reported that beta-catenin can piggyback into the nucleus by binding to TCF/LEF-1, there is evidence that in vivo its import is independent of TCF/LEF-1. Therefore, the mechanism for beta-catenin nuclear localization remains to be established. Beta-catenin nuclear import has been analyzed in an in vitro assay using permeabilized cells. Beta-catenin docks specifically onto the nuclear envelope in the absence of other cytosolic factors. Docking is not inhibited by an NLS peptide and does not require importins/karyopherins, the receptors for classical NLS substrates. Rather, docking is specifically competed by importin-beta/beta-karyopherin, indicating that beta-catenin and importin-beta/beta-karyopherin both interact with common nuclear pore components. Nuclear translocation of beta-catenin is energy dependent and is inhibited by nonhydrolyzable GTP analogs and by a dominant-negative mutant form of the Ran GTPase. Cytosol preparations contain inhibitory activities for beta-catenin import that are distinct from the competition by importin-beta/beta-karyopherin and may be involved in the physiological regulation of the pathway. Beta-catenin is imported into the nucleus by binding directly to the nuclear pore machinery, similar to importin-beta/beta-karyopherin or other importin-beta-like import factors, such as transportin. These findings provide an explanation for how beta-catenin localizes to the nucleus without an NLS, independent of its interaction with TCF/LEF-1. This is a new and unusual mechanism for the nuclear import of a signal transduction protein. The lack of beta-catenin import activity in the presence of normal cytosol suggests that its import may be regulated by upstream events in the Wnt signaling pathway (Fagotto, 1998).

Wnt signaling is thought to be mediated via interactions between beta-catenin and members of the LEF-1/TCF family of transcription factors. The mechanism of transcriptional regulation by LEF-1 in response to a Wnt-1 signal has been studied under conditions of endogenous beta-catenin in NIH 3T3 cells, and an examination made as to whether association with beta-catenin is obligatory for the function of LEF-1. Wnt-1 signaling confers transcriptional activation potential on LEF-1 by association with beta-catenin in the nucleus. By mutagenesis, specific residues in LEF-1 have been identified that are important for interaction with beta-catenin. The amino-terminal 56 residues of LEF-1 are necessary and sufficient for mediating the interaction with beta-catenin. Two transcriptional activation domains in beta-catenin have been delineated whose function is augmented in specific association with LEF-1. To examine whether the carboxyl terminus of beta-catenin is the sole determinant for transcriptional activation in association with LEF-1, the deletion mutant delta C was generated. Expression of this mutant beta-catenin with LEF-1 in transfected Neuro2A cells stimulates the activity of the LEF-CAT reporter gene to a level similar to that obtained with the wild-type beta-catenin. Since this deletion removes the epitope recognized by the anti-beta-catenin antibody, the accumulation of the mutant protein cannot be compared to that of the wt protein. Nevertheless, the data suggest that beta-catenin may contain additional sequences that contribute to transcriptional activation (Hsu, 1998).

Alternatively, overexpression of exogenous forms of beta-catenin may result in an increase in the pool of free cytosolic endogenous beta-catenin, which would obscure the mapping of transcriptional activation domains. To identify transcriptional activation domains in beta-catenin, fusion proteins were generated in which various portions of beta-catenin were linked to the GAL4 DNA-binding domain. Expression of GAL4 fusion proteins containing either the amino- or carboxyl-terminal region of beta-catenin stimulate transcription of a GAL-CAT reporter gene in transfected Neuro2A cells by a factor of either 8 or 14, respectively. In comparison, a 50-fold activation of the reporter gene is observed with a GAL4-VP16 expression plasmid in which the transcriptional activation domain of the viral protein VP16 is linked to GAL4 as a positive control. A similar accumulation of both GAL4-beta-catenin fusion proteins was confirmed by electrophoretic mobility shift assays with nuclear extracts from COS cells transfected with the corresponding expression plasmids. Thus, beta-catenin contains two distinct transcription activation domains that can function in a heterologous context. However, the levels of transcriptional activation by the GAL4-beta-catenin fusion proteins are significantly lower than those observed in cotransfections of Neuro2A cells with LEF-1 and beta-catenin, raising the possibility that the transcriptional activation domains of beta-catenin collaborate with one another or with LEF-1. A Wnt-1 signal and beta-catenin association are not required for the architectural function of LEF-1 in the regulation of the T-cell receptor alpha enhancer, which involves association of LEF-1 with a different cofactor, ALY. Thus, LEF-1 can assume diverse regulatory functions by association with different proteins (Hsu, 1998).

A crystal structure of beta-catenin bound to the beta-catenin binding domain of Tcf3 (Tcf3-CBD) has been determined. The Tcf3-CBD forms an elongated structure with three binding modules that runs antiparallel to beta-catenin along the positively charged groove formed by the armadillo repeats. Structure-based mutagenesis defines three sites in beta-catenin that are critical for binding the Tcf3-CBD and are differentially involved in binding APC, cadherin, and Axin. The structural and mutagenesis data reveal a potential target for molecular drug design studies (Graham, 2000).

As seen in the beta-catenin/Tcf complex crystal structure, beta-catenin provides a rigid platform that presents main chain and side chain groups for Tcf recognition. The armadillo repeat region forms an ideal structural platform that restrains the positions of Cbeta atoms as well as all main chain atoms. In this sense, beta-catenin is analogous to double-stranded DNA. It forms a right-handed superhelix with a groove that spirals along the helix. The beta-catenin binding domain of XTcf3 (XTcf3-CBD) extends along much of this groove, binding to beta-catenin with three different modules. Recognition of a peptide spread out in a rigid groove allows the peptide to bind with high specificity by exposing all its functional groups for recognition. In addition, binding in the groove allows a higher binding affinity due to a larger protein-protein interface -- the Kd between beta-catenin and hTcf4-CBD is in a low nM range. Since all of the critical residues in the beta-catenin/Tcf3 interface are conserved in the Tcf/LEF-1 family, it is expected that the binding mode observed in this structure serves as a framework for all beta-catenin/Tcf and beta-catenin/LEF-1 interactions. Furthermore, the beta-catenin/Tcf-CBD structure, together with mutagenesis data, suggest that binding within the groove of the armadillo repeat region of beta-catenin may be a common theme for some other CBDs, such as those of cadherins and APC. It is noteworthy that another armadillo repeat protein, Karyopherin, also uses its groove to recognize NLS (nuclear localization signal) peptides in extended conformations (Graham, 2000).

The XTcf3-CBD can be roughly divided into three binding modules: an N-terminal beta hairpin (residues 7-15), an extended region that contains two charged 'buttons' (residues 16-29), and an alpha helix (residues 40–52). The crystal structure and the previous Tcf4 mutagenesis data allowed for the location of three hot spots at the beta-catenin/Tcf3 interface that are critical for the binding of XTcf3 to beta-catenin. Two of them are charged buttons in which the lysine residues at postions 312 and 435 bind acidic residues in the extended region of the XTcf3-CBD. The other hot spot is a hydrophobic pocket that binds Leu-48 of XTcf3. By inserting two negatively charged aspartate residues in this hydrophobic pocket, XTcf3 could be prevented from binding beta-catenin. In addition, to mimic the effects of a small compound bound in the groove of beta-catenin, in which the alpha helix docks, a double mutation was designed to build a 'dam' in the beta-catenin groove. This A295W/I296W mutation also prevents XTcf3 binding to beta-catenin (Graham, 2000).

These results are completely consistent with previously reported Tcf4 mutagenesis data. A series of peptides of the human Tcf4 CBD were produced containing deletions and mutations at conserved residues, and the ability of these mutant peptides to compete with the wild-type Tcf4 peptide for beta-catenin binding was measured. Because of the high similarity between human Tcf4 and Xenopus Tcf3, these results can be mapped onto the current structure. This comparison demonstrates that the extended region of Tcf is the minimal unit for binding to beta-catenin, whereas the beta hairpin region is largely dispensable. The alpha helix makes important contributions to the binding affinity of Tcf4 to beta-catenin, since a mutation in this region dramatically reduces the binding affinity of the Tcf4 CBD for beta-catenin. Based on these results, it is suggested that the docking of the extended region to the positively charged groove of beta-catenin may be the first step in the beta-catenin/Tcf recognition, and that the subsequent binding of the alpha helix to beta-catenin through hydrophobic interactions provides a critical component of the high affinity binding (Graham, 2000).

In general, the formation of secondary structure may restrain the conformation of a peptide chain and thus allow the protein to present its surface features in a more specific manner. It is interesting to note that the beta-catenin binding domain of XTcf3 contains a beta hairpin module and an alpha helix at the N- and C-terminal ends, respectively, flanking the essential central extended region. One of the features of the beta-catenin/Tcf interaction is that, while the central extended region provides the minimal recognition domain, the two flanking structural modules appear to help restrain the central extended region to binding in its docking site. In the beta-catenin/XTcf3 structure, the residues C-terminal to Glu-29 of XTcf3 are quite flexible and can potentially be stretched. This helps to explain a paradox in the mutagenesis data. While mutation of the second charged button of beta-catenin (K312D) abolishes the binding to XTcf3, the corresponding mutation in Tcf4 (E24A) does not have dramatic effects on beta-catenin/Tcf4 binding. It is interesting to note that there are several negatively charged residues (Glu-26, Glu-28, and Glu-29) C-terminal to Glu-24. It is possible that residues in this region can adopt alternative conformations so that another negatively charged residue can match with the Lys-312 button when Glu-24 is mutated (Graham, 2000).

beta-catenin plays multiple roles in cell regulation by forming complexes with more than two dozen protein partners. It was of interest to know whether any of the beta-catenin residues required for interacting with XTcf3 are also used to bind other well defined CBDs of different partners, even though there is no apparent reported sequence homology. Of the proteins tested, APC behaves most like XTcf3. Mutations in the region that binds the XTcf3 helix, as well as in the button region, significantly impair APC binding. In contrast, C-cadherin binding to beta-catenin requires the buttons but is not affected by mutations that inhibit binding of the XTcf3 helix. This is in agreement with deletion studies demonstrating that armadillo repeats 1-3, which include residues that bind the XTcf3 helix, are not required for E-cadherin binding. Finally, Axin does not bind to the beta-catenin buttons defined here, but does require contact with the XTcf3 helix binding region. Intriguingly, a single amino acid change in the hydrophobic pocket that interacts with the helix (F253D) is not sufficient to abrogate cadherin or APC binding, but does prevent the binding of Axin to beta-catenin. This result indicates that Axin/beta-catenin binding is more dependent on an interaction in the hydrophobic region than cadherin/beta-catenin or APC/beta-catenin binding, which may be due to the absence of interactions between Axin and the beta-catenin buttons. The observations that these proteins share overlapping binding sites are consistent with earlier reports that binding of Tcf, APC, and cadherin is mutually exclusive. Since Axin and APC can simultaneously bind to beta-catenin, it is plausible that APC can utilize the charged buttons for binding while at the same time Axin might bind to the helical binding region of beta-catenin (Graham, 2000).

Previous truncation studies have suggested that Tcf and cadherin bind to corresponding regions of beta-catenin. Although no apparent homology was found between Tcf-CBD and the cadherin-CBD, the similar requirement of XTcf3 and C-cadherin for binding to the charged buttons, Lys-312 and Lys-435, suggests that they may share a conserved beta-catenin binding motif. It has been shown that an ~30 amino acid region in the E-cadherin cytoplasmic domain is necessary and sufficient for beta-catenin binding. The N-terminal half of the cadherin-CBD is remarkably similar in sequence to the critical extended region of Tcf4-CBD that binds the charged buttons. In particular the two button binding residues in XTcf3-CBD (Asp-16 and Glu-24), as well as a critical residue identified in LEF-1 (corresponding to Phe-21 in XTcf3), are also conserved in the cadherin-CBD. Thus, it is proposed that the N-terminal half of the cadherin-CBD binds to beta-catenin in a conformation similar to that of the extended region of the XTcf3-CBD (Graham, 2000).

A structural model is suggested for the cadherin/beta-Catenin/alpha-Catenin complex in cell adhesion. alpha-catenin is an essential link between adhesion junctions and the cytoskeleton. While the N-terminal region of alpha-catenin is responsible for interacting with beta-catenin, its C-terminal region binds to actin, directly or via alpha-actinin. alpha-catenin binds to the junction of the N-terminal domain and the armadillo repeat region of beta-catenin. Biochemical studies and a crystal structure of a chimeric protein of alpha-catenin (residue 57-164) and beta-catenin (residues 118-151) have resulted in a model for the interaction of the beta-catenin junction segment with alpha-catenin. In the crystal structure of the alpha-/beta-catenin chimera, residues 121-141 of beta-catenin form an alpha helix, and this helix interacts with three alpha helices in alpha-catenin by forming a four-helix bundle. Since the residues N-terminal to residue 150 were not visible in the earlier reported crystal structures of the armadillo repeat region of murine beta-catenin, it was not possible to predict the relative orientation between alpha- and beta-catenin. In the crystal structure, residues 134-149 could be visualized, and amino acids 134-161 have been shown to form a long kinked helix (Graham, 2000).

Although the structure beyond residue 134, which is the first amino acid in this construct, could not be viewed, the structure of this region in the chimeric protein indicates that this helix continues further toward the N terminus of beta-catenin. Since alpha-catenin binds this helix in a four-helix bundle, this orients alpha-catenin approximately perpendicular to the beta-catenin superhelix. Consolidating the E-cadherin/beta-catenin complex and beta-catenin/alpha-catenin complex models, E-cadherin and alpha-catenin sit on roughly opposite surfaces of beta-catenin. In addition, because it is suggested that the E-cadherin peptide chain runs antiparallel to the groove of the armadillo repeat superhelix, the N terminus of beta-catenin, and thus alpha-catenin, are pointing away from the plasma membrane (Graham, 2000).

Structural and mutagenesis studies suggest that the region of beta-catenin that binds the XTcf3 helix might be a promising place for the design of small molecule inhibitors. There is a hydrophobic pocket lined by Phe-253 and Phe-293 that is required for XTcf3 to bind beta-catenin. Mutations in this pocket, or in the shallow groove nearby, have no effect on C-cadherin binding, indicating that a drug binding in this region might not be deleterious for cell adhesion. As with all chemotherapy agents, the key issue will be to develop molecules and conditions that will adversely impact the cancer cells without harming normal cells (Graham, 2000).

beta-Catenin and gamma-catenin (plakoglobin), vertebrate homologs of Drosophila armadillo, function in cell adhesion and the Wnt signaling pathway. In colon and other cancers, mutations in the APC tumor suppressor protein or beta-catenin's amino terminus stabilize beta-catenin, enhancing its ability to activate transcription of Tcf/Lef target genes. Though beta- and gamma-catenin have analogous structures and functions and like binding to APC, evidence that gamma-catenin has an important role in cancer has been lacking. APC is shown in this study to regulate both beta- and gamma-catenin and gamma-catenin functions as an oncogene. In contrast to beta-catenin, for which only amino-terminal mutated forms transform RK3E epithelial cells, wild-type and several amino-terminal mutated forms of gamma-catenin have similar transforming activity. gamma-Catenin's transforming activity, like beta-catenin's, is dependent on Tcf/Lef function. However, in contrast to beta-catenin, gamma-catenin strongly activates c-Myc expression and c-Myc function is crucial for gamma-catenin transformation. These findings suggest APC mutations alter regulation of both beta- and gamma-catenin, perhaps explaining why the frequency of APC mutations in colon cancer far exceeds that of beta-catenin mutations. Elevated c-Myc expression in cancers with APC defects may be due to altered regulation of both beta- and gamma-catenin. Furthermore, the data imply beta- and gamma-catenin may have distinct roles in Wnt signaling and cancer via differential effects on downstream target genes (Kolligs, 2000).

The data presented here are the first to suggest that beta- and gamma-catenin may have differential effects on Tcf/Lef target genes. Specifically, it was found wild-type beta-catenin has a roughly twofold greater effect and S33Y mutated beta-catenin a roughly 15-fold greater effect than wild-type gamma-catenin in activating gene expression from a model promoter construct containing three Tcf-binding sites upstream of a minimal c-Fos promoter. In contrast, the ability of wild-type gamma-catenin to activate c-MYC reporter gene constructs is similar to that of S33Y beta-catenin. S33Y is a cancer-derived missense substitution in beta-catenin's presumptive GSK3beta phosphorylation sequences. gamma-Catenin activates endogenous c-Myc gene expression in RK3E cells more strongly than does the S33Y mutant beta-catenin protein. The underlying mechanisms for their differential effects on the reporter gene constructs and on endogenous c-Myc are not yet clear, though differences in the interactions of the distantly related amino- and carboxy-terminal domains of gamma- and beta-catenin with specific transcription factors, coactivators, and/or other chromatin-associated proteins are among the possible explanations. For instance, gamma-catenin may enhance or facilitate the binding of certain transcription factors to promoters, whereas beta-catenin may cooperate with other factors. The presence or absence of specific DNA-binding sites for certain transcription factors in regulatory elements of a particular Tcf/Lef-regulated target gene might account for its differential activation by beta- or gamma-catenin. Alternatively, beta- and gamma-catenin may differ in their ability to interact with certain chromatin remodeling proteins, some of which likely have differential effects on specific genes in vivo. Regardless of the particular mechanisms underlying their differential effects on c-Myc and potentially other target genes, the data presented here support the view that beta- and gamma-catenin are likely to have distinct but complementary roles in Wnt signaling and cancer development (Kolligs, 2000).

Transcriptional activation of Wnt/Wg-responsive genes requires the stabilization and nuclear accumulation of ß-catenin, a dedicated coactivator of LEF/TCF enhancer-binding proteins. Recombinant ß-catenin strongly enhances binding and transactivation by LEF-1 on chromatin templates in vitro. Interestingly, different LEF-1 isoforms vary in their ability to bind nucleosomal templates in the absence of ß-catenin, owing to N-terminal residues that repress binding to chromatin, but not nonchromatin, templates. Transcriptional activation in vitro requires both the armadillo (ARM) repeats and the C terminus of ß-catenin, whereas the phosphorylated N terminus is inhibitory to transcription. A fragment spanning the C terminus (CT) and ARM repeats 11 and 12 (CT-ARM), but not the CT alone, functions as a dominant negative inhibitor of LEF-1-ß-cat activity in vitro and can block ATP-dependent binding of the complex to chromatin. LEF-1-ß-cat transactivation in vitro is repressed by inhibitor of ß-catenin and Tcf-4 (ICAT), a physiological inhibitor of Wnt/Wg signaling that interacts with ARM repeats 11 and 12, and by the nonsteroidal anti-inflammatory compound, sulindac. None of these transcription inhibitors (CT-ARM, ICAT, or sulindac) can disrupt the LEF-1-ß-cat complex after it is stably bound to chromatin. It is concluded that the CT-ARM region of ß-catenin functions as a chromatin-specific activation domain, and that several inhibitors of the Wnt/Wg pathway directly modulate LEF-1-ß-cat activity on chromatin (Tutter, 2001).

Although some DNA-binding proteins recognize their binding sites in chromatin efficiently in vitro, others must be incubated with specific chromatin remodeling complexes or chromatin-modifying enzymes to activate transcription from fully-assembled chromatin templates. Therefore, it was important to assess whether the recombinant LEF-1-ß-cat complex can activate transcription from a preassembled nucleosomal template. LEF-1-ß-cat transactivation is very inefficient when the complex is incubated with the pBRE template after nucleosome assembly, but transcription is enhanced significantly in the presence of purified recombinant p300. Activation by p300 is specific because it does not enhance transcription without enhancer factors, or when incubated with LEF-1-ß-catDeltaC. It was also asked whether LEF-1-ß-cat activity could be enhanced by ATP-dependent chromatin remodeling complexes. Although the complex could not be activated with a purified SWI/SNF fraction, LEF-1-ß-cat activity is stimulated by a partially-purified chromatin remodeling fraction (RMF), which contains the hSWI/SNF and hACF/ISWI remodeling complexes and is devoid of p300. The effect of RMF is similar to that observed with recombinant p300, and in combination the two fractions function synergistically. LEF-1-ß-cat activation under these conditions can still be repressed selectively by CT-ARM, and not by the CT fragment of ß-catenin. Enhanced binding in the presence of the RMF fraction is more pronounced with LEF-1-ß-cat than with LEF-1 alone. Thus LEF-1-ß-cat can strongly activate transcription from fully assembled chromatin templates when incubated with p300 and chromatin remodeling enzymes (Tutter, 2001).

DNase I footprint analysis of these transcription reactions revealed that the LEF-1-ß-cat complex also binds very poorly on its own to the pBRE enhancer when added to the template after the chromatin template has been fully assembled, and under these conditions binding of the complex is enhanced considerably by the addition of RMF. Enhanced binding of LEF-1-ß-cat to chromatin in the presence of RMF is completely inhibited by apyrase, indicating that an ATP-dependent chromatin remodeling step is required. Interestingly, RMF-enhanced binding can also be competed by the CT-ARM fragment and not by the CT fragment. In these experiments the CT and CT-ARM inhibitors, or apyrase, were added together with the LEF-1-ß-cat complex and the RMF fraction after the completion of nucleosome assembly. In contrast, purified recombinant p300 does not affect the binding of LEF-1-ß-cat to chromatin. Taken together, these data indicate that LEF-1-ß-cat transactivation requires p300 and chromatin remodeling, and that the CT-ARM fragment can block both ATP-dependent binding of the LEF-1-ß-cat complex to preassembled chromatin, as well as the transcriptional activity of the complex after it has bound stably to chromatin (Tutter, 2001).

ß-Catenin is a mediator of the Wnt-signaling pathway. In many cancers, ß-catenin is stabilized and accumulates in the nucleus where it associates with lymphoid-enhancing factor 1/ T-cell transcription factors to activate genes involved in cell transformation. Adenomatous polyposis coli (APC) protein can regulate ß-catenin localization by nuclear export. In vitro transport assays have been used to test whether cellular ß-catenin can exit the nucleus independent of APC and the CRM1 export receptor. In digitonin-permeabilized SW480 (APCmut/mut) tumor cells, nuclear ß-catenin decreases >60% in export reactions in the absence of exogenous factors. Under similar conditions, nuclear c-ABL is only exported after the addition of cytosolic extract, and the export is blocked by the CRM1-specific inhibitor, leptomycin B. The nuclear export of ß-catenin is not blocked by leptomycin B treatment, revealing a CRM1- and APC-independent pathway. The export of ß-catenin is sensitive to lower temperatures and the removal of ATP, indicating an active process. Ectopically expressed yellow fluorescent protein-ß-catenin also displays CRM1-independent export. Conversely, the overexpression of the CRM1 transporter moderately stimulates export of nuclear ß-catenin, confirming that ß-catenin exits the nucleus by at least two distinct pathways. The shuttling ability of tumor cell ß-catenin has implications for its regulation and its role in transferring signals between the nucleus and plasma membrane (Eleftheriou, 2001).

The APC protein is located in both the cytoplasm and the nucleus. Nuclear export of APC is mediated by two intrinsic, leucine-rich, nuclear export signals (NESs) located near the amino terminus. Each NES was able to induce the nuclear export of a fused carrier protein. Both APC NESs are independently able to interact with the Crm1 nuclear export factor and substitute for the HIV-1 Rev NES to mediate nuclear mRNA export. Both APC NESs function within the context of APC sequence: an amino-terminal APC peptide containing both NESs interacts with Crm1 and shows nuclear export in a heterokaryon nucleocytoplasmic shuttling assay. Also, mutation of both APC NESs results in the nuclear accumulation of the full-length, ~320-kDa APC protein, further establishing that the two intrinsic APC NESs are necessary for APC protein nuclear export. Moreover, endogenous APC accumulates in the nucleus of cells treated with the Crm1-specific nuclear export inhibitor leptomycin B. Together, these data indicate that APC is a nucleocytoplasmic shuttle protein whose predominantly cytoplasmic localization requires NES function and suggests that APC may be important for signaling between the nuclear and cytoplasmic compartments of epithelial cells (Neufeld, 2000).

ß-catenin is a multifunctional protein involved in both cell adhesion and transcriptional activation. Transcription mediated by the ß-catenin/Tcf complex is involved in embryological development and is upregulated in various cancers. The crystal structure at 2.5 Å resolution has been determined of a complex between ß-catenin and ICAT, a protein that prevents the interaction between ß-catenin and Tcf/Lef family transcription factors. ICAT contains a 3-helix bundle that binds armadillo repeats 10-12 and a C-terminal tail that, similar to Tcf and E-cadherin, binds in the groove formed by armadillo repeats 5-9 of ß-catenin. ICAT selectively inhibits ß-catenin/Tcf binding in vivo, without disrupting ß-catenin/cadherin interactions. Thus, it should be possible to design cancer therapeutics that inhibit ß-catenin-mediated transcriptional activation without interfering with cell adhesion (Graham, 2002).

In the canonical Wnt signaling pathway, ß-catenin activates target genes through its interactions with Tcf/Lef-family transcription factors and additional transcriptional coactivators. The crystal structure of ICAT, an inhibitor of ß-catenin-mediated transcription, bound to the armadillo repeat domain of ß-catenin, has been determined. ICAT contains an N-terminal helilical domain that binds to repeats 11 and 12 of ß-catenin, and an extended C-terminal region that binds to repeats 5-10 in a manner similar to that of Tcfs and other ß-catenin ligands. Full-length ICAT dissociates complexes of ß-catenin, Lef-1, and the transcriptional coactivator p300, whereas the helical domain alone selectively blocks binding to p300. The C-terminal armadillo repeats of ß-catenin may be an attractive target for compounds designed to disrupt aberrant ß-catenin-mediated transcription associated with various cancers (Daniels, 2002).

The Bicoid-related transcription factor Pitx2 is rapidly induced by the Wnt/Dvl/ß-catenin pathway and is required for effective cell-type-specific proliferation by directly activating specific growth-regulating genes. Wnt signaling, in acting upstream of Pitx2, directly induces Pitx2 gene expression, based on the recruitment of LEF1 to evolutionary-conserved sites in the Pitx2 gene 5'-regulatory regions, with a regulated exchange of HDAC1 for ß-catenin occurring on these Pitx2 sites. Regulated exchange of HDAC1/ß-catenin converts Pitx2 from repressor to activator, analogous to control of TCF/LEF1. Pitx2 then serves as a competence factor required for the temporally ordered and growth factor-dependent recruitment of a series of specific coactivator complexes that prove necessary for Cyclin D2 gene induction (Kioussi, 2002).

Activation of the Wnt pathway results in rapid recruitment of the Pitx2 gene and binding of Pitx2 to promoters of specific growth control genes. This linkage between the Wnt pathway and Pitx2 gene expression provides an insight into the molecular mechanisms of cell type-specific proliferation, based on the required actions of Pitx2 to activate specific, critical growth-control gene targets acting in G1. Based on in vivo studies, as well as the actions in pituitary and muscle cell models, Pitx2 is required for normal proliferation when expressed in heterologous cells; Pitx2 can actually inhibit proliferation. It is speculated that this may occur by squelching coregulatory factors required by DNA binding transcription factors that exert analogous functions to Pitx2. To subserve its proliferative effects, Pitx2 must bind to its cognate DNA sites and requires an N-terminal activation domain, but not the C terminus. Together, these data suggests that three independent events underlie Pitx2-dependent activation of cell type-specific proliferation: Wnt-dependent activation of Pitx2; Wnt and growth factor-dependent relief of Pitx2 repression function; and serial recruitment of a series of specific coactivator complexes that act in a promoter-specific manner, analogous to effects of β-catenin on LEF1 (Kioussi, 2002).

Renal dysplasia, the major cause of childhood renal failure in humans, arises from perturbed renal morphogenesis and molecular signaling during embryogenesis. Induction of molecular crosstalk between Smad1 and ß-catenin occurs in the TgAlk3QD mouse model of renal medullary cystic dysplasia. The finding that Myc, a Smad and ß-catenin transcriptional target and effector of renal epithelial dedifferentiation, is misexpressed in dedifferentiated epithelial tubules provides a basis for investigating coordinate transcriptional control by Smad1 and ß-catenin in disease. Enhanced interactions occur between a molecular complex consisting of Smad1, ß-catenin and Tcf4 and adjacent Tcf- and Smad-binding regions located within the Myc promoter in TgAlk3QD dysplastic renal tissue, and Bmp-dependent cooperative control of Myc transcription by Smad1, ß-catenin and Tcf4. Analysis of nuclear extracts derived from TgAlk3QD and wild-type renal tissue revealed increased levels of Smad1/ß-catenin molecular complexes, and de novo formation of chromatin-associated Tcf4/Smad1 molecular complexes in TgAlk3QD tissues. Analysis of a 476 nucleotide segment of the 1490 nucleotide Myc genomic region upstream of the transcription start site demonstrated interactions between Tcf4 and the Smad consensus binding region and associations of Smad1, ß-catenin and Tcf4 with oligo-duplexes that encode the adjacent Tcf- and Smad-binding elements only in TgAlk3QD tissues. In collecting duct cells that express luciferase under the control of the 1490 nucleotide Myc genomic region, Bmp2-dependent stimulation of Myc transcription is dependent on contributions by each of Tcf4, ß-catenin and Smad1. These results provide novel insights into mechanisms by which interacting signaling pathways control transcription during the genesis of renal dysplasia (Hu, 2005)

Emerin is a type II inner nuclear membrane (INM) protein of unknown function. Emerin function is likely to be important because, when it is mutated, emerin promotes both skeletal muscle and heart defects. One function of Emerin is to regulate the flux of ß-catenin, an important transcription coactivator, into the nucleus. Emerin interacts with ß-catenin through a conserved adenomatous polyposis coli (APC)-like domain. When GFP-emerin is expressed in HEK293 cells, ß-catenin is restricted to the cytoplasm and ß-catenin activity is inhibited. In contrast, expression of an emerin mutant, lacking its APC-like domain (GFP-emerinDelta), dominantly stimulates ß-catenin activity and increases nuclear accumulation of ß-catenin. Human fibroblasts that are null for emerin have an autostimulatory growth phenotype. This unusual growth phenotype arises through enhanced nuclear accumulation and activity of ß-catenin and can be replicated in wild-type fibroblasts by transfection with constitutively active ß-catenin. These results support recent findings that suggest that INM proteins can influence signalling pathways by restricting access of transcription coactivators to the nucleus (Markiewicz, 2006).

MACF1 (microtubule actin cross-linking factor 1) is a multidomain protein that can associate with microfilaments and microtubules. MACF1 was highly expressed in neuronal tissues and the foregut of embryonic day 8.5 (E8.5) embryos and the head fold and primitive streak of E7.5 embryos. MACF1-/- mice die at the gastrulation stage and displaye developmental retardation at E7.5 with defects in the formation of the primitive streak, node, and mesoderm. This phenotype was similar to Wnt-3-/- and LRP5/6 double-knockout embryos. In the absence of Wnt, MACF1 associated with a complex that contains Axin, ß-catenin, GSK3ß and APC. Upon Wnt stimulation, MACF1 appears to be involved in the translocation and subsequent binding of the Axin complex to LRP6 at the cell membrane. Reduction of MACF1 with small interfering RNA decreased the amount of ß-catenin in the nucleus, and led to an inhibition of Wnt-induced TCF/ß-catenin-dependent transcriptional activation. Similar results were obtained with a dominant-negative MACF1 construct that contained the Axin-binding region. Reduction of MACF1 in Wnt-1-expressing P19 cells resulted in decreased T (Brachyury) gene expression, a DNA-binding transcription factor that is a direct target of Wnt/ß-catenin signaling and required for mesoderm formation. These results suggest a new role of MACF1 in the Wnt signaling pathway (Chen, 2006).

β-catenin is the central signalling molecule of the canonical Wnt pathway, where it activates target genes in a complex with LEF/TCF transcription factors in the nucleus. The regulation of β-catenin activity is thought to occur mainly on the level of protein degradation, but it has been suggested that β-catenin nuclear localization and hence its transcriptional activity may additionally be regulated via nuclear import by TCF4 and BCL9 and via nuclear export by APC and axin. Using live-cell microscopy and fluorescence recovery after photobleaching (FRAP), the impact of these factors on the subcellular localization of β-catenin, its nucleo-cytoplasmic shuttling and its mobility within the nucleus and the cytoplasm were directly analysed. TCF4 and BCL9/Pygopus recruit β-catenin to the nucleus, and APC, axin and axin2 enrich β-catenin in the cytoplasm. Importantly, however, none of these factors accelerates the nucleo-cytoplasmic shuttling of β-catenin, i.e. increases the rate of β-catenin nuclear import or export. Moreover, the cytoplasmic enrichment of β-catenin by APC and axin is not abolished by inhibition of CRM-1-dependent nuclear export. TCF4, APC, axin and axin2 move more slowly than β-catenin in their respective compartment, and concomitantly decrease β-catenin mobility. Together, these data indicate that β-catenin interaction partners mainly regulate β-catenin subcellular localization by retaining it in the compartment in which they are localized, rather than by active transport into or out of the nucleus (Krieghoff, 2006).

Protein interactions of Armadillo homologs: Interaction with p300/CBP

The molecular mechanisms of transcriptional activation by beta-catenin are only poorly understood. The closely related acetyltransferases p300 and CBP potentiate beta-catenin-mediated activation of the siamois promoter, a known Wnt target. beta-catenin and p300 also synergize to stimulate a synthetic reporter gene construct, whereas activation of the cyclin D1 promoter by beta-catenin is refractory to p300 stimulation. Axis formation and activation of the beta-catenin target genes siamois and Xnr-3 in Xenopus embryos are sensitive to the E1A oncoprotein, a known inhibitor of p300/CBP. The C-terminus of beta-catenin interacts directly with a region overlapping the CH-3 domain of p300. p300 could participate in alleviating promoter repression imposed by chromatin structure and in recruiting the basal transcription machinery to promoters of particular Wnt target genes (Hecht, 2000).

The finding that p300/CBP serves as a coactivator for beta-catenin in vertebrates is unexpected given that in Drosophila dCBP has been shown to negatively regulate Wingless-signaling. The apparent discrepancy between the function of vertebrate p300/CBP and dCBP could be explained most easily if vertebrate CBP and p300, or dCBP and its mammalian orthologs, were functionally different from one another. Species-specific differences are known for certain aspects of Wnt signaling. Also, CBP and p300 are differentially engaged in retinoic acid and cAMP responses in mammalian cells. However, the results presented here show that CBP and p300 can both serve as cofactors for ß-catenin. An interaction between LEF-1 and the CH-3 domain of p300, which corresponds to the dCBP-2 region used in the Drosophila studies, could not be detected. Rather, ß-catenin can interact with p300 in the absence of TCFs (Drosophila homolog, Pangolin). These findings indicate that dCBP differs from p300 and CBP and that in vertebrates p300 or CBP enter the beta-catenin-TCF complex through an interaction with beta-catenin, whereas dCBP may be brought to a promoter by its interaction with dTCF (Hecht, 2000).

Aside from the difference in the TCF interaction, the genetic studies in Drosophila and the experiments in vertebrates may also reveal a more complicated involvement of p300/CBP in Wnt signaling. For example, upon binding to dTCF, Armadillo may form a ternary complex and reprogram the activity of dCBP that is already present. Alternatively, p300 or CBP may function as coactivators of beta-catenin only initially. Over time their activity could change and lead to the downregulation of target genes as reported for the interferon-beta enhanceosome, which eventually is destabilized and disassembled by CBP through acetylation of an architectural component, the HMG I(Y) protein. Similarly, coactivator complexes associated with the promoter-bound estrogen receptor are dissociated after acetylation by p300/CBP, which leads to the attenuation of the hormone response. Thus, one could hypothesize that p300 performs a stimulatory function and also provides a shut-off mechanism in Wnt signaling (Hecht, 2000).

ß-catenin plays a pivotal role in the transcriptional activation of Wnt-responsive genes by binding to TCF/LEF transcription factors. Although it has been suggested that the COOH-terminal region of ß-catenin functions as an activation domain, the mechanisms of activation remain unclear. To screen for potential transcriptional coactivators that bind to the COOH-terminal region of ß-catenin, a novel yeast two-hybrid system, the Ras recruitment system (RRS) that detects protein-protein interactions at the inner surface of the plasma membrane, was used. RRS is based on the ability of mammalian Ras to rescue the growth defect of the yeast temperature-sensitive cdc25-2 strain, in which the endogenous Ras is inactive at the nonpermissive temperature (37°C) due to the lack of a functional Cdc25 guanyl nucleotide exchange factor. For RRS screening, a bait protein of interest is fused at the COOH terminus of mammalian activated Ras. This activated Ras lacks the membrane localization signal [Ras(61)deltaF], whereas library cDNAs are fused to the v-Src myristoylation sequence targeted to the plasma membrane. A protein-protein interaction between the bait and library protein results in the recruitment of Ras to the membrane and complementation of the cdc25-2 mutation. Using this system, the CREB-binding protein (CBP) was isolated. From armadillo (arm) repeat 10 to the COOH terminus of ß-catenin is involved in binding to CBP, whereas ß-catenin interacts directly with the CREB-binding domain of CBP. ß-Catenin synergizes with CBP to stimulate the activity of a synthetic reporter in vivo. Conversely, ß-catenin-dependent transcriptional activation is repressed by E1A, an antagonist of CBP function, but not by an E1A mutant that does not bind to CBP. The activation of Wnt target genes such as siamois and Xnr3 in Xenopus embryos is also sensitive to E1A. These findings suggest that CBP provides a link between ß-catenin and the transcriptional machinery, and possibly mediates the oncogenic function of ß-catenin (Takemaru, 2000).

Nuclear export of ßCatenin

Activation of the Wnt pathway induces ß-catenin to localize inside the nucleus, where it interacts with transcription factors such as TCF/LEF-1. Regulation of the pathway occurs through a ß-catenin-degrading complex based on Axin and the tumor suppressor APC. ß-catenin import occurs independently of nuclear import factors. APC, which can shuttle in and out of the nucleus, has been proposed to be responsible for reexport of ß-catenin in a CRM1-dependent manner. The export factors CRM1 and CAS are known to bind tightly to their cargo only through cooperative association with RanGTP in the nucleus. ß-catenin export has been studied in vivo and in semipermeabilized cells. ß-catenin contains three export sequences. Export is insensitive to leptomycin B, a specific inhibitor of the CRM1-mediated pathway. It does not require nuclear RanGTP, and it can be reconstituted in the absence of additional soluble factors; this is consistent with nondirectional translocation of ß-catenin. Further observations suggest that ß-catenin subcellular distribution in vivo may depend primarily on retention through interaction with other cellular components. Evidence is shown that reexport is required for degradation of nuclear ß-catenin and that nuclei lack Axin, an essential component of the degradation machinery. It is concluded that ß-catenin is exported independently of the CRM1 pathway. A model of free, nondirectional nuclear translocation for ß-catenin is presented, its localization being regulated by retention in the nucleus and degradation in the cytoplasm (Wiechens, 2001).

ß-catenin export was found to rely on two regions, located in the N terminus and C terminus, respectively. Either of these two terminal domains could mediate efficient export of the central arm repeat region, and both could induce export of GST. The C-terminal domain appears to be the most active one and can account by itself for the export of full-length ß-catenin. The two export sequences correspond to regions with other attributed functions; the N-terminal sequence contains the GSK-3 consensus site involved in regulation of ß-catenin stability, and the C-terminal domain has been characterized for its transactivation activity. However, so far these regions have not been implicated in transport or localization of ß-catenin. They have no homology with known export signals or molecules involved in nuclear transport. Export via these terminal domains does not appear to follow any conventional pathway; the export processes they mediate are not negatively affected by excess NES, are insensitive to leptomycin B, and do not require RanGTP. Both pathways, however, are saturable. This indicates that they rely on specific interactions, which remain to be characterized. The terminal domains may be transported by specific exportins with the use of an unconventional RanGTP-independent mechanism. Alternatively, they may be capable of direct translocation through the nuclear pore. Other molecules, such as RCC1 and NTF2, which show no sequence similarity to the repeats of the importin/exportin family, can indeed cross the nuclear pore independently of other transporters (Wiechens, 2001).

The occurrence of independent export domains may further ensure a tight control on nuclear ß-catenin. Note that two of these domains, i.e., the arm repeats and the C terminus, overlap with interaction domains for XTCF-3 and for components of the transcription machinery, respectively, while the N terminus includes the GSK-3 phosphorylation site and could be influenced by Wnt signaling. ß-catenin distribution and activity may thus not only depend on ß-catenin levels but also on regulation of multiple interactions and/or modifications of these domains (Wiechens, 2001).

The Hippo pathway regulates Wnt/beta-catenin signaling

Several developmental pathways contribute to processes that regulate tissue growth and organ size. The Hippo pathway has emerged as one such critical regulator. However, how Hippo signaling is integrated with other pathways to coordinate these processes remains unclear. This study shows that the Hippo pathway restricts Wnt/β-Catenin signaling by promoting an interaction between TAZ (a yorkie homolog, sharing approximately 50% sequence identity with YAp, with a similar topology, containing two central WW domains and a C-terminal transactivation domain) and DVL in the cytoplasm. TAZ inhibits the CK1δ/epsilon-mediated phosphorylation of DVL, thereby inhibiting Wnt/β-Catenin signaling. Abrogation of TAZ levels or Hippo signaling enhances Wnt3A-stimulated DVL phosphorylation, nuclear β-Catenin, and Wnt target gene expression. Mice lacking Taz develop polycystic kidneys with enhanced cytoplasmic and nuclear β-Catenin. Moreover, in Drosophila, Hippo signaling modulates Wg target gene expression. These results uncover a cytoplasmic function of TAZ in regulating Wnt signaling and highlight the role of the Hippo pathway in coordinating morphogenetic signaling with growth control (Varelas, 2010).

The Hpo/Wts kinase, which in vertebrates is represented by MST and LATS, respectively, converges on the transcriptional regulator Yorkie in Drosophila, or the vertebrate homologs, TAZ and YAP. Recent studies have focused on the transcriptional roles for Yorkie, TAZ, and YAP in controlling growth and cell fate choice. However, these proteins are also found in the cytoplasm, where TAZ has been shown to associate with the plasma membrane and actin cytoskeleton. TAZ cytoplasmic localization is mediated by LATS-dependent phosphorylation of serine 89, which drives binding to the cytoplasmic retention factor, 14-3-3. Altogether, these findings raised the possibility that TAZ has cytoplasmic functions. This study demonstrates that one such function is to constrain Wnt signaling by inhibiting CK1δ/epsilon-mediated DVL phosphorylation (see model). Although the ability of CK1δ/epsilon to phosphorylate DVL has been well established, the precise role of this event in Wnt signaling remains to be fully understood. This study has observed that CK1δ/epsilon-independent signaling can contribute to Wnt-induced β-Catenin stabilization. Nevertheless, the data clearly show that DVL phosphorylation and Wnt pathway activation caused by the loss of TAZ expression requires CK1δ/epsilon. Abrogation of LATS expression, which leads to the nuclear accumulation of TAZ, reduced TAZ-DVL interaction, enhanced DVL phosphorylation, and promoted Wnt pathway activation in cells. Furthermore, it was shown that loss of hpo or wts function or enhanced expression of yki in Drosophila wing imaginal discs also modulates Wnt signaling. Thus, by promoting the cytoplasmic localization of TAZ, the Hippo pathway not only prevents nuclear TAZ activity, but also leads to inhibition of Wnt signaling by antagonizing DVL activation. This study has focused on TAZ, but it will be interesting to determine whether the related protein, YAP, has a similar activity in the Wnt pathway. Indeed, functional redundancy is suggested by the Taz/Yap double-knockout mice, which display early embryonic lethality (Varelas, 2010).

Wnts are known to act throughout mammalian kidney development, and hyperactivation of Wnt/β-Catenin signaling causes severe kidney defects. For example, loss of Apc function results in nuclear β-Catenin accumulation that leads to cystic kidneys, and transgenic mice expressing an activated mutant of β-Catenin also develop cysts. Taz mutant mice also develop polycystic kidneys, and increased cytoplasmic and nuclear β-Catenin was found in both cystic and precystic regions of Tazgt/gt kidneys, indicating Wnt pathway activation. Moreover, cysts formed in humans with ADPKD, caused by mutations in PC1 or PC2, display a gene expression profile consistent with Wnt pathway activation. β-Catenin associates with PC1 (Lal, 2008), and TAZ has been proposed to indirectly regulate the activity of PC2 via the SCF ubiquitin ligase β-TrCP. It was confirmed that TAZ and TAZ (1-393, ΔWW), the DVL-binding mutant that cannot rescue the effect of loss of TAZ on Wnt signaling, both interact with β-TrCP. Thus, it is concluded that TAZ modulates Wnt signaling via direct regulation of DVL; however, TAZ may also function indirectly to regulate the PC complex via β-TrCP. Whether TAZ may have other additional roles in Wnt signaling remains to be determined. Altogether, these results suggest that, in the kidney, TAZ restricts canonical Wnt signaling that otherwise would disrupt tubule morphogenesis and the formation of cysts (Varelas, 2010).

Defects in cilia are also known to be involved in the pathogenesis of polycystic kidney disease. Indeed, recent analyses of Taz mutant kidneys have revealed that Taz localizes to cilia, and that cystic regions in Taz −/− kidneys have underdeveloped cilia. Interestingly, primary cilia have recently been implicated in restraining Wnt signaling (Corbit, 2008), and DVL has been reported in cilia (Park, 2008). These findings raise the intriguing possibility that TAZ may link suppression of DVL activation in the kidney to ciliogenesis. Of note, no evidence was found that the MDA-MB-231 breast cancer cells used in this study possess cilia, indicating that cilia, per se, are not required for TAZ inhibition of Wnt signaling. Nevertheless, it will be interesting to examine the relationship between TAZ regulation of ciliogenesis and DVL modulation in kidney epithelia (Varelas, 2010).

The development of multicellular organisms requires cell-cell communication and the coordinated action of distinct signaling pathways for proper patterning of the body and organs. In the Drosophila wing imaginal disc, the Wg morphogen gradient coordinates the expression of an array of developmentally important genes. These studies reveal that the Hippo pathway functions to restrain Wg signaling, and that loss of hpo or wts increases the levels of Arm (β-Catenin) outside the normal expression domain and enhances expression of Wg target genes, particularly at a distance from the Wg source. Thus, Hippo pathway activation may, in effect, serve to restrain the sensitivity of cells to Wg signals, thus sharpening the Wg response gradients to promote appropriate tissue patterning events. How the Hippo kinase cascade is regulated by upstream pathways remains unclear. However, genetic studies place a protocadherin, Fat, and an ERM domain protein, Expanded, upstream of Hpo/Wts kinases. Furthermore, and in accordance with analysis of hpo and wts mutants, fat and expanded mutants display enhanced Wg target gene expression independent of Wg ligand production. The results suggest that this occurs via a direct molecular link. In this regard, it is intriguing to note that mice null for Fat4, a mammalian homolog of Drosophila fat, also develop polycystic kidneys. Direct communication between the Hippo and Wnt pathways may thus be an ancient mechanism that coordinates morphogen signaling with tissue growth control in animals (Varelas, 2010).

This study shows that TAZ inhibits Wnt signaling via DVL, providing for a direct molecular link between the Hippo and Wnt signaling pathways. Other studies in Drosophila have implicated the Hippo pathway as a positive regulator of Notch signaling, and an essential role for TAZ in mediating TGFβ-dependent Smad signaling has been shown. Furthermore, TAZ is induced by BMP signaling to regulate mesenchymal cell differentiation. Altogether, these studies suggest a critical and direct role for TAZ in integrating signals from numerous morphogen signaling pathways (Varelas, 2010).

Dishevelled-KSRP complex regulates Wnt signaling through post-transcriptional stabilization of beta-catenin mRNA

Table of contents

armadillo continued: Biological Overview | Regulation | Protein Interactions | Developmental Biology | Effects of Mutation | References

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.