Journal of Cell Science 113, 911-919 (2000) 911 Printed in Great Britain © The Company of Biologists Limited 2000 JCS0743

COMMENTARY Cross-regulation of the Wnt signalling pathway: a role of MAP

Jürgen Behrens Max-Delbrück-Center for Molecular Medicine, Robert-Rössle-Str. 10, 13122 Berlin, Germany *Author for correspondence (e-mail: [email protected])

Published on WWW 21 February 2000

SUMMARY

The Wnt signal transduction pathway regulates various Wnt-controlled cell fate decision in the early aspects of embryonal development and is involved in cancer Caenorhabditis elegans embryo. Moreover, TAK1 activates formation. Wnts induce the stabilisation of cytosolic β- NLK, which phosphorylates TCFs bound to β-catenin. This catenin, which then associates with TCF transcription blocks nuclear localization and DNA binding of TCFs. factors to regulate expression of Wnt-target genes. At Since TAK1 is activated by TGF-β and various cytokines, various levels the Wnt pathway is subject to cross- it might provide an entry point for regulation of the Wnt regulation by other components. Recent evidence suggests system by other pathways. In addition, alterations in that a specific MAP pathway involving the MAP TAK1-NLK might play a role in cancer. kinase kinase kinase TAK1 and the MAP kinase NLK counteract Wnt signalling. In particular, homologues of TAK1 and NLK, MOM-4 and LIT-1, negatively regulate Key words: Wnt pathway, MAP kinase, TAK1, NLK

INTRODUCTION THE Wnt SIGNAL TRANSDUCTION PATHWAY

A variety of biochemical pathways that relay information Wnts are a family of secreted glycoproteins that act on from the extracellular milieu to the cytosol and cause neighbouring cells in a paracrine fashion and thereby mediate alterations in gene expression have been identified in recent cellular interactions during development (Cadigan and Nusse, years. In most cases, secreted factors bind to and activate 1997). The Wnt pathway is strikingly conserved in different specific receptors on the cell surface, which couple to adaptor species and has been studied in C. elegans, Drosophila proteins in the cytoplasm; these interact with additional melanogaster, Zebrafish, Xenopus laevis, chicken, mouse and downstream components in a sequential fashion. Most of the human cancer cells (Bienz, 1998; Cox and Peifer, 1998; Han, pathways end up in the cell nucleus and activate or repress 1997; Miller and Moon, 1996; Moon and Miller, 1997; the function of specific transcription factors that regulate Perrimon, 1994). Wnts are expressed in a tissue-specific target genes responsible for the ultimate biological response. manner, and mice that have mutations in genes whose products Although the individual pathways can be studied are involved in Wnt signalling show defects in the development independently of each other in separate experimental of a variety of organs (for a review see Cadigan and Nusse, systems, it is becoming increasingly clear that they are 1997). In Xenopus embryos, ectopic expression of activating frequently interconnected in the cell. This creates a highly components of the pathway induces dorsalisation of the complex network of signalling pathways. Although an embryo, which results in the formation of a secondary body increasing number of points at which cross-talk occurs are axis. Conversely, inhibitory components of the pathway being identified, their biological significance is only promote ventralisation and prevent axis formation. In beginning to be understood. Drosophila, the homologous Wingless pathway is involved in Here I concentrate on the Wnt signalling pathway and the establishment of segment polarity, wing formation and discuss cross-regulatory effectors that have been identified differentiation of the endoderm (Cadigan and Nusse, 1997). recently. A major focus is on the connection between Wnt The hierarchy of the Wnt pathway has been mainly established signalling and a MAP kinase pathway that involves the TAK1 through genetic epistasis experiments in Drosophila as well as and NLK serine/threonine kinases. I first introduce the two through overexpression studies in Xenopus. The Wnt system in pathways separately and then go on to describe the recent Caenorhabditis elegans is discussed in more detail below. evidence that indicates a connection between the two systems. In the overview given here, I use the mammalian names of Finally, I discuss the implications of the results as well as the Wnt pathway components (Fig. 1). In brief, Wnts act in a future research directions. paracrine fashion by binding to frizzled receptors, which are a 912 J. Behrens

ubiquitination by phosphorylation of specific serine and AB threonine residues in its N-terminal domain. Phosphorylated β- Porcupine MOM-1 catenin is recognized by the F-box protein slimb/βTrCP, which attracts ubiquitinating that transfer polyubiquitin Wnt MOM-2 chains to β-catenin (Hart et al., 1999; Jiang and Struhl, 1998;

Frizzled MOM-5 Kitagawa et al., 1999). Mutations of the phosphorylation sites result in stabilisation of β-catenin (Morin et al., 1997; Yost et al., 1996). Dishevelled The phosphorylation of β-catenin appears to take place in a multi-protein complex assembled by the scaffolding CK I KIN-19 component axin or the related protein conductin (Behrens et ?? GSK3β SGG-1 al., 1998; Ikeda et al., 1998; Kishida et al., 1998; Zeng et al., APCAxin/Conductin APR-1 1997). These proteins contain separate binding sites for the TAB1 TAP-1 tumor-suppressor protein APC (adenomatous polyposis coli), β-Catenin TAK1 WRM-1 MOM-4 the serine/threonine kinase GSK3β (glycogen synthase kinase 3β) and β-catenin (Fig. 1). Several lines of evidence point to a TCF NLK POP-1 LIT-1 role for these factors in the degradation of β-catenin. Mutations of APC, which occur in about 80% of sporadic colon Vertebrates/Drosophila C. elegans carcinomas and in the inherited disease familial adenomatosis β Fig. 1. Overview of the Wnt pathway and cross-talk to the polyposis, are correlated with increased levels of -catenin TAK1/NLK pathway in vertebrates and Drosophila (A) and C. (Munemitsu et al., 1995; Polakis, 1997). The mutations lead to elegans (B). Functional relationships between the components, rather deletions of the axin/conductin-binding SAMP repeats and of than biochemical interactions, are indicated. Homologous proteins 20 amino acid repeats that bind to β-catenin. Introduction of are shown in the same colour. Note that GSK3β and APC are wild-type APC into colon carcinoma cells induces degradation negative regulatory components in vertebrates and Drosophila and of β-catenin (Munemitsu et al., 1995). positive regulators in C. elegans. TCFs are activated by β-catenin in GSK3β can phosphorylate β-catenin, and phosphorylation is vertebrates and Drosophila but inactivated in C. elegans. CKI (casein stimulated by axin. This indicates that axin/conductin act by kinase I)/KIN-19 has been positioned between dishevelled and β β β bringing GSK3 into close proximity to -catenin (Ikeda et al., GSK3 in both vertebrates and C. elegans. Axin and conductin are 1998). GSK3β can also phosphorylate APC as well as related proteins showing similar domain structure and biochemical interactions. Both proteins bind to β-catenin, APC and GSK3β by conductin/axin (Rubinfeld et al., 1996; Willert et al., 1999), which increases the affinity of these components for β-catenin. using separate domains and are probably functionally redundant. An β β axin/conductin homologue has not been found in the genome A dominant-negative mutant of GSK3 stabilises -catenin sequence of C. elegans (Ruvkun and Hobert, 1998). In the C. elegans and induces double-axis formation in Xenopus (Yost et al., pathway, LIT-1 is activated by WRM-1, which suggests that both 1996). Both axin and conductin can induce β-catenin proteins cooperate, possibly in a complex with POP-1, to degradation in colon cancer cell lines and block Wnt-induced downregulate repressive function of POP-1. NLK/LIT-1 is activated stabilisation of β-catenin and axis formation in Xenopus by TAK1/MOM-4 and might thus be regulated by upstream stimuli embryos (Behrens et al., 1998; Hart et al., 1998; Ikeda et al., that act on these components (indicated by ‘?’). In Drosophila, it is 1998; Zeng et al., 1997). Mutation of the axin gene in mice unknown whether the TAK1-NLK system plays a role in Wnt leads to duplication of the body axis (Zeng et al., 1997). signalling. Drosophila homologues of axin and APC that have roles in Wnt signalling have also been described (Hamada et al., 1999; family of seven-transmembrane-span proteins (Bhanot et al., Yu et al., 1999). In summary, conductin/axin, APC and GSK3β 1996). Activation of these receptors leads to the stabilisation are negative regulators of the Wnt pathway, which act by of the cytosolic component β-catenin, which then enters the inducing the degradation of β-catenin. nucleus and interacts with HMG-box transcription factors of Little is known about signal transmission from frizzled the LEF-1/TCF family (below collectively referred to as TCF). receptors to the axin/conductin complexes (Fig. 1). It is clear The TCF-β-catenin complexes transmit the Wnt signal into the that the cytoplasmic phosphoprotein dishevelled becomes nucleus and act as transcriptional regulators of Wnt-target activated by frizzled receptors; this might involve genes (Behrens et al., 1996; Huber et al., 1996; Molenaar et phosphorylation of dishevelled, but the mechanism of al., 1996). β-Catenin not only is involved in Wnt signalling but activation is unknown. Activated dishevelled inhibits GSK3β, was originally identified in mammalian cells as a component possibly by interacting with the axin/conductin complex (Li et associated with cadherin cell adhesion molecules. It serves as al., 1999a). β-Catenin becomes hypophosphorylated and is no a link between the cytoplasmic domain of cadherins and the longer recognized by slimb/βTrCP and the ubiquitination cytoskeleton-associated protein α-catenin (Behrens, 1999; machinery, and consequently accumulates in the cells. Specific Hülsken et al., 1994). isoforms of have recently been implicated in Much progress has been made over the past years in our the stabilisation of β-catenin and appear to act between understanding of the two central regulatory events in the Wnt dishevelled and GSK3β (Peters et al., 1999; Sakanaka et al., pathway: the control of β-catenin stability and its interplay 1999). with TCF factors. In the absence of Wnts, cytosolic β-catenin After its stabilisation β-catenin is able to enter the nucleus is quite unstable, being rapidly ubiquitinated and cleared by either alone (Fagotto et al., 1998; Prieve and Waterman, 1999) proteasomes (Aberle et al., 1997). β-Catenin is marked for or complexed with TCF transcription factors (Behrens et al., MAP kinases in Wnt signalling 913

1996; Huber et al., 1996). The TCF proteins were originally The β-catenin-degradation machinery can be regulated by identified in the immune system, but clearly play important different means. The GSK3β-binding protein, GBP, from roles in embryonal development (Clevers and van-de-Wetering, Xenopus, which is related to the mouse proto-oncogene Frat1 1997). TCFs bind to specific sites in promoters of target genes (frequently rearranged in T cell lymphomas), blocks activity of via an HMG box and interact with β-catenin through their N- GSK3β, increases β-catenin levels and induces double-axis terminal domains. Several target genes that are activated by the formation in Xenopus embryos (Yost et al., 1998). The activity TCF-β-catenin complexes have been identified. These include of the β-catenin-destruction complex might also be regulated genes important in embryonal development, such as those that by dephosphorylation, given that subunits of protein encode ultrabithorax in Drosophila, and nodal-related 3, phosphatase 2A bind to both axin and APC (Hsu et al., 1999; siamois and twin in Xenopus, as well as others whose products Seeling et al., 1999). might be involved in the oncogenic effects of Wnt signalling Wnt signalling might also be influenced by cadherins (for a review see Nusse, 1999, and see below). (Behrens, 1999). β-Catenin has a core domain of twelve so- Apparently, two distinct mechanisms are involved in the called arm-repeats, and binding of this region to cadherins and activation of Wnt-target genes by TCF-β-catenin complexes TCF factors is mutually exclusive (Peifer et al., 1994; Sadot et (Bienz, 1998). First, TCFs can behave as ‘silent’ transcription al., 1998). Depletion of the cytosolic pool of β-catenin by factors that become converted to transcriptional activators upon overexpression of cadherins interferes with Wnt signalling binding to β-catenin. β-Catenin contains transcriptional in several experimental systems (Fagotto et al., 1996). activation domains in its N- and C-terminal parts (Hsu et al., Conversely, Wnt might also regulate cadherin-based cell 1998; van-de-Wetering et al., 1997), and genetic analysis adhesion by altering the levels of β-catenin (Hinck et al., 1994). indicates that additional transactivation domains might reside Whether the cross-talk between the cadherin and Wnt systems elsewhere in the molecule (Cox et al., 1999). Thus, a bipartite is physiological is unclear. Genetic evidence in Drosophila transcription factor is generated in which the DNA-binding and indicates that the cell adhesion and signalling roles of β-catenin transactivation functions are contributed by TCFs and β- can be functionally separated (Sanson et al., 1996). catenin, respectively. The integrin-linked kinase, ILK, might link integrin Second, TCFs can behave as active repressors of gene signalling and the Wnt pathway. When overexpressed in expression, and interaction with β-catenin relieves this epithelial cells, ILK leads to the downregulation of the repression. TCFs can repress gene expression by interacting epithelial E-cadherin and to nuclear translocation of β-catenin. with transcriptional co-repressors, such as TLE/groucho family Concomitantly, LEF-1 levels increase, which results in members and CtBP (Brannon et al., 1999; Cavallo et al., 1998; activation of TCF-β-catenin-dependent transcription. Roose et al., 1998). Genetic analysis in Drosophila shows that Interestingly, ILK can also block the activity of GSK3β, but a reduction in gene dosage of TCF suppresses the phenotype of this does not appear to alter β-catenin levels (Dedhar et al., wingless or armadillo mutations, indicating a negative role of 1999). In a similar manner to ILK, oncogenic Ras can promote TCFs (Cavallo et al., 1998). In Xenopus, TCF-β-catenin the loss of E-cadherin and induce translocation of β-catenin to complexes are active only in the dorsal part of the embryo and the nucleus. This probably involves interaction between β- locally activate siamois expression. Elimination of the TCF- catenin and phosphoinositide 3-kinase (PI3K; Espada et al., binding sites in the siamois promoter results in reduced dorsal 1999). but increased ventral expression of the gene (Brannon et al., 1997). Obviously, TCFs repress Wnt-target genes in the ventral part of the embryo. In C. elegans the TCF homologue POP-1 Wnt SIGNALLING IN CANCER also plays a negative role in Wnt signalling and becomes inhibited by the β-catenin homologue, WRM-1. Since WRM-1 The Wnt signalling pathway and the TCF-β-catenin interaction binds to POP-1 only inefficiently, it probably does not function appear to play an important role in the development of cancer. as a transcriptional activator in this process (see below). TCF-β-catenin complexes have transforming activity in vitro The transcriptional activity of TCF-β-catenin complexes is (Aoki et al., 1999; Orford et al., 1999), and transgenic regulated by different means. In Drosophila, the histone expression of stabilised versions of β-catenin lacking the N- acetyltransferase CBP/p300 acetylates TCF, which reduces terminal serine and threonine phosphorylation sites leads to interaction with β-catenin (Waltzer and Bienz, 1998). β- tumor formation in mice (Gat et al., 1998; Harada et al., 1999). Catenin also interacts with certain Sox proteins which are In human tumor samples and cell lines, β-catenin is frequently HMG-box factors related to TCFs. These interactions block the found in the nucleus and forms constitutive complexes with transcriptional activation by TCF/β-catenin complexes and TCFs (Herter et al., 1999; Morin et al., 1997; Porfiri et al., result in ventralisation of Xenopus embryos (Zorn et al., 1999). 1997). As mentioned above, stabilisation of β-catenin in most colorectal tumors results from mutations in the APC tumor suppressor. In colorectal tumors that lack mutations in APC, MODULATION OF THE Wnt PATHWAY activating mutations of β-catenin that alter or delete the regulatory phosphorylation sites have been identified (Morin et Several components modulate Wnt signalling at various levels al., 1997). These mutations of β-catenin have also been found of the pathway. A number of factors, such as WIF-1, cerberus in a wide variety of other tumor types, which indicates that β- and frzB physically interact with Wnts and prevent binding to catenin can have widespread action as an oncoprotein in vivo frizzled proteins. In Drosophila, the porcupine gene product (reviewed by Behrens, 1999). Several target genes of TCF-β- regulates secretion of the Wnt homologue wingless (reviewed catenin complexes that might function in tumorigenesis have by Cadigan and Nusse, 1997). been identified. These include c-MYC and the gene encoding 914 J. Behrens cyclin D1 (He et al., 1998; Tetsu and McCormick, 1999), GF TGF-β BMP IL-1 TNFα which might drive cell proliferation, and genes involved in other aspects of tumor development, such as those encoding GFR TGF-βR BMPR IL-1R TNFαR the matrix metalloproteinase matrilysin (Crawford et al., 1999) and PPARδ (peroxisome proliferator-activated receptor δ), RAS XIAP TRAF6 which is a target of anti-cancer drugs (He et al., 1999). TAB1 TAB1 TAB1 β RAF Interestingly, TCF-1 is also a target of TCF-4- -catenin TAK1 TAK1 TAK1 complexes. Ablation of the gene that encodes TCF-1 in mice results in tumor formation. TCF-1 may act as a feedback repressor of TCF-4-activated genes and cooperate with APC to MEK1 MKK4/MKK7/SEK NIK suppress malignant transformation (Roose et al., 1999). IKK ERK SAPK/JNK axis formation IκB THE TGF-β-ACTIVATED KINASE TAK1 AND ITS ACTIVATOR TAB1 NFκB

Mitogen-activated protein kinases (MAPKs) play fundamental (a) (b) (c) (d) roles in various receptor-mediated signalling pathways. In the Fig. 2. MAP kinase signal transduction pathways involving TAK1. classical Ras/Raf pathway, which transmits signals from (a) The Ras/Raf pathway is shown for comparison. Ras activates a growth factor receptors, Ras activates the MAP kinase kinase sequence of kinases that includes the MAPKKK RAF, the MAPKK kinase (MAPKKK) Raf, which through phosphorylation MEK1 and the MAPK ERK. (b) Activation of TGF-β receptors stimulates the MAP kinase kinase MEK1. Finally, the MAP (TGF-βR) by TGF-β stimulates the kinase activity of TAK1 probably kinase ERK is activated by MEK1 and modifies the function via TAB1, which can lead to activation of various MAP kinase of various substrates by phosphorylation (Fig. 2a). Mammalian kinases, such as MKK4, MKK7 or SEK. These kinases can TAK1 is related to MAPKKKs both in function and in phosphorylate and activate SAPK/JNK, which will phosphorylate further substrates (Shibuya et al., 1996; Wang et al., 1997; sequence. TAK1 was initially identified by complementation Yamaguchi et al., 1995; Yao et al., 1999). (c) In Xenopus, TAK1 is assays in yeast, where it can substitute for the MAPKKK activated after binding of agonists to BMP receptors (BMPR). XIAP Ste11p in the yeast pheromone-induced MAPK pathway is an intermediate protein that links TAB1 to the BMP receptor. (Yamaguchi et al., 1995). TAK1 was subsequently implicated Activation of TAK1 leads to ventralisation of the embryo and blocks in TGF-β signalling: TAK1 can activate the TGF-β-dependent axis formation (Shibuya et al., 1998; Yamaguchi et al., 1999). plasminogen activator inhibitor (PAI) gene promoter, and a (d) The role of TAK1 in the stabilisation of NFkB by IL-1 and dominant negative (kinase-dead) mutant of TAK1 blocks TNFα. TAK1 can be activated by IL-1 or TNFα, which leads to the promoter activation by TGF-β (Yamaguchi et al., 1995). The stimulation of NIK and IKK. These kinases phosphorylate IKKα κ kinase activity of cellular TAK1 is stimulated by treatment with causing its degradation and the subsequent release of NF B TGF-β or with BMP (bone morphogenetic protein), a member (Ninomiya-Tsuji et al., 1999; Sakurai et al., 1999). GF, growth β factor; GFR, growth factor receptor; ERK, extracellular signal- of the TGF- superfamily (Fig. 2b,c). Intriguingly, TAK1 has regulated kinase; MKK, MAP kinase kinase; SEK, stress activated a role in the control of dorsal-ventral axis specification by protein kinase/ERK kinase; SAPK/JNK, stress-activated protein BMPs in Xenopus (Fig. 2c). Microinjection of BMPs leads to kinase/c-Jun amino-terminal kinase. TRAF, TNF receptor-associated ventralisation of Xenopus embryos. This effect is mediated by factor; NIK, NFκB-inducing kinase; IKK, IκB kinase, specific BMP receptors and appears to require functional TAK1: Xenopus TAK1 was shown to induce ventralisation of embryos, whereas a dominant negative mutant of TAK1 blocks BMP-induced ventralisation (Shibuya et al., 1998). versions of XIAP partially prevent ventralisation induced by Shibuya et al. (1996) identified the TAB1 protein as a BMP (Yamaguchi et al., 1999). These results indicate that specific interaction partner of TAK1 in a yeast two-hybrid XIAP is a component of the BMP signalling sytem that acts screen. TAB1 activates TAK1 by directly binding to its by recruiting the TAB1-TAK1 complex to the BMP receptor catalytic domain. In mammalian cells TAB1 stimulated the (Fig. 2c). In the Xenopus assays XIAP also prevents activation of the PAI reporter by TAK1. A deletion mutant of triggered by overexpression of TAB1 or TAK1. TAB1 lacking the TAK1-interaction domain repressed the Besides TGF-β signalling, TAK1 appears to be involved in TGF-β-induced activation of the reporter, which suggests that various other signal transduction pathways (Fig. 2d). It TAB1 participates in TGF-β signalling. It is not clear by which activates the NFκB transcription factor by stimulating the mechanism TAB1 activates TAK1, but it might block the NFκB-inducing kinase (NIK) or the IκB kinase (IKK; inhibitory activity that has been located to a short piece of the Ninomiya-Tsuji et al., 1999; Sakurai et al., 1998, 1999). This N-terminus of TAK1. could be of importance for interleukin 1 (IL-1) signalling, XIAP (for X-chromosome-linked inhibitor of apoptosis because a dominant negative TAK1 mutant prevents IL-1- protein) is another TAB1-binding protein (Yamaguchi et al., induced activation of NFκB (Ninomiya-Tsuji et al., 1999). 1999) and belongs to a family of proteins that antagonize Moreover, TAK1 interacts with TRAF6, which is a direct apoptosis in insect cells. XIAP also interacts with the downstream effector of the IL-1 receptor. TAK1 might also cytoplasmic domains of BMP receptors and might link TAB1 play a role in the activation of the c-Jun N-terminal kinase to the receptor. In Xenopus embryos, XIAP enhances the (JNK) by TGF-β, TNFα or ceramides (Fig. 2; Shirakabe et al., ventralisation induced by TAK1 and TAB1, and truncated 1997; Wang et al., 1997; Yao et al., 1999). MAP kinases in Wnt signalling 915

AB P1 AB P1 AB P1

further cell divisions Fig. 3. Early development of the C. elegans embryo and phenotype after mutation or EMS P2 EMS P2 EMS P2 disturbance of Wnt-pathway further cell components. In wild-type C. divisions elegans the daughters of the EMS cell become commited to the mesodermal or endodermal MS E MS MS E E cell fate by a signal from P2 to EMS (left). Mutation or mesoderm endoderm mesoderm endoderm disturbance of mom, apr-1, kin- 19, sgg-1, wrm-1 or lit-1 genes results in formation of wild-typemom pop-1 mesoderm by both daughter apr-1 cells of the EMS cell (middle), kin-19 whereas mutation of pop-1 sgg-1 results in endoderm formation wrm-1 (right). lit-1

CROSS-TALK BETWEEN THE Wnt SIGNALLING β-catenin homologues in C. elegans, WRM-1, leads to AND THE MAP KINASE SYSTEMS mesodermal cell fates, which indicates that wrm-1 acts oppositely to pop-1 (Rocheleau et al., 1997). These findings The initial observations pointing to a role of TAK1 in Wnt indicate that the components of the Wnt pathway in C. elegans signalling came from genetic analyses in C. elegans. The Wnt and vertebrates are similar but have different activities: pathway controls anterior-posterior polarization and sister cell whereas APC and GSK3 are inhibitors in vertebrates, they fates in the early C. elegans embryo (Fig. 3). The fates of the behave as activators in C. elegans. Conversely, TCFs become two daughter cells produced after the first cell division are activated by β-catenin in vertebrates but are inactivated in C. already determined: one follows the anterior fate (AB cell), the elegans (Figs 1, 3). other the posterior fate (P1 cell). The P1 cell generates the It should be noted that the sequence similarity shared by the EMS and P2 cells, which are again programmed to become vertebrate and the C. elegans homologues of APC and β- anterior and posterior cells, respectively. A signal originating catenin is very low (~20%). There are also differences in the from P2 polarizes the EMS cell so that the anterior daughter molecular interactions of the proteins. WRM-1 does not cell becomes mesoderm (MS) and the posterior daughter cell interact efficiently with POP-1, and APR-1 appear to lack the becomes endoderm (E; Fig. 3; reviewed by Bowerman and 15- and 20-residue repeats present in APC that mediate binding Shelton, 1999). to β-catenin (Rocheleau et al., 1997). Interestingly, APR-1 It is at this stage that the Wnt signal appears to come into contains sequences similar to the SAMP repeats of APC, which play (see Fig. 1). Mutation of the mom (more mesoderm) genes bind to axin/conductin (Behrens et al., 1998; Rocheleau et al., blocks differentiation into the endoderm lineage, which leads 1997). This indicates that APR-1 interacts with an axin/ to the development of mesoderm from both daughters of the conductin-like protein in C. elegans, although such a EMS cell (Rocheleau et al., 1997; Thorpe et al., 1997). Some homologue has not been detected in the C. elegans genome moms encode homologues of Wnt-pathway components: mom- sequence (Ruvkun and Hobert, 1998). Apparently, the 1, mom-2 and mom-5 correspond to porcupine, wingless and biochemical and functional characteristics of these Wnt frizzled, respectively. MOM-4 is the homologue of TAK1 signalling components have diverged considerably during (Meneghini et al., 1999). Interestingly, a gene distantly related evolution. to that encoding mammalian APC, called apr-1, is also The results of protein expression studies are in line with involved in endoderm specification (Rocheleau et al., 1997). involvement of pop-1 in endodermal cell fate. POP-1 protein APR-1 behaves as an activator in this pathway, in a similar can be detected in large amounts in the nuclei of the MS cells manner to a GSK3β homologue, SGG-1 (Schlesinger et al., but only in small amounts in the E cells. Mutation of the mom- 1999). Importantly, POP-1, which is a TCF-like protein has the 2 gene or RNA interference with WRM-1 eliminates this opposite function Ð i.e. mutation of pop-1 leads to the difference and results in similarly high levels of POP-1 in both generation of endodermal cells from both EMS progenitors cell types (Rocheleau et al., 1997; Thorpe et al., 1997). This (Lin et al., 1998). Moreover, RNA interference with one of the suggests that Wnt signalling establishes the assymetry of POP- 916 J. Behrens

1 expression or activity in the EMS cell daughters and that this β-catenin. This blocks translocation of TCFs to the nucleus difference is responsible for the E-versus-MS cell-fate and/or prevents binding to the target gene promoters (Fig. 1). decision. Activation of LIT-1 by WRM-1 might be the means by which RNA interference studies implicate the C. elegans Wnt signalling in C. elegans blocks the negative activity of homologue of the Drosophila nemo gene, lit-1, in Wnt POP-1 Ð i.e. both proteins might actually cooperate in a signalling (Meneghini et al., 1999; Rocheleau et al., 1999). complex with POP-1 to downregulate its repressive function. LIT-1 is highly homologous to nemo and to the mammalian nemo-like kinase (NLK) in its kinase and C-terminal domains. Injection of dsRNA for lit-1 resulted in a mom phenotype Ð i.e. CONCLUSIONS AND PERSPECTIVES produced embryos lacking endoderm. Similar defects in endoderm development were also observed in lit-1 mutants These studies have a number of implications and raise several (Rocheleau et al., 1999). Importantly, mutations in pop-1 are questions. It will be interesting to determine whether TAK1 epistatic to mutations in lit-1 Ð i.e. lit-1 pop-1 double mutants and LIT-1 are directly regulated by Wnt signalling and/or have the pop-1 phenotype (Kaletta et al., 1997). This indicates whether they represent entry points by which other signal that LIT-1 is positioned upstream of POP-1 and negatively transduction pathways become connected to the Wnt pathway. regulates POP-1 activity. Moreover, the assymetric distribution The direct activation of the LIT-1 kinase by β-catenin indicates of POP-1 in wild-type worms is abolished in lit-1 mutants. that LIT-1 can be part of the Wnt pathway (Fig. 1). However, Biochemical analysis revealed a possible mechanism by which genetic evidence suggests that the MAP kinase and Wnt LIT-1 regulates POP-1: POP-1 is a substrate for the LIT-1 pathways act in parallel and converge at the level of WRM- kinase and phosphorylation appears to relocalize POP-1 from 1/β-catenin. For instance, Wnt-pathway mutants upstream of the nucleus to the cytoplasm (Rocheleau et al., 1999). wrm-1 exhibit only partially penetrant losses of endoderm and Importantly, LIT-1 binds to and becomes activated by the β- POP-1 assymetries in comparison with mom-4, lit-1 and wrm- catenin homologue WRM-1 (Rocheleau et al., 1999), which 1 mutants (Meneghini et al., 1999). Furthermore, mom-4 and explains how WRM-1 activated by the Wnt pathway blocks lit-1 mutants show defects in body wall muscle not observed POP-1 function. In this scenario, LIT-1 is actually a component with mom-1 and mom-2 mutants. It is therefore possible that of the Wnt pathway and is positioned between β-catenin and MOM-4/TAK-1 and LIT-1/NLK function in a pathway that is TCFs (Fig. 1). distinct from Wnt signalling and upstream of WRM-1/β- That LIT-1 might also be regulated by cross-talk from the catenin. A major goal is to identify the upstream inputs that MAP kinase pathway became apparent from analysis of the activate MOM-4 and LIT-1. functions of the TAK1 and TAB1 homologues MOM-4 and The LIT-1 kinase might be activated by the TGF-β/BMP, TAP-1. Lit-1 and mom-4 were shown genetically to be negative TNFα and IL-1 pathways, which all regulate TAK1 (Figs 1, 2). upstream regulators of pop-1 (Kaletta et al., 1997; Thorpe et It remains to be determined whether any of these potential al., 1997). Moreover, mutations in mom-4 lead to the loss of cross-links controls the Wnt signal and are functional in C. assymetry of POP-1 in MS-E sister pairs (Meneghini et al., elegans. The fact that TAK1 mediates the ventralising effects 1999; Shin et al., 1999). RNA-interference analysis also of BMPs in Xenopus embryos is therefore interesting in this showed that TAP1 is needed for MOM-4 activity in endoderm context. It is tempting to speculate that ventralisation is due to specification, which indicates that TAP1 positively regulates inhibition of TCF-β-catenin complexes by activated TAK1. MOM-4 (Meneghini et al., 1999). Indeed, TAP-1 binds to However, recent studies show that BMPs do not affect the MOM-4 and promotes its kinase activity. expression of known downstream target genes of Wnts, such How is the MOM-4/TAK1 signal relayed to the TCF-β- as siamois, Twin and Xnr3. Furthermore, blocking of the BMP catenin complexes? Experiments in mammalian cells indicate pathway in ventral blastomeres does not lead to the induction that, in the presence of TAB1, TAK1 can activate NLK, the of siamois, which makes a direct Wnt-BMP connection via LIT-1 homologue (Ishitani et al., 1999). Similarly, MOM-4 TAK1 questionable (for a review see Sokol, 1999). activates LIT-1 via an activation-loop-like motif that is also There are further links between Wnt pathway components found in other MAPKs; activation depends on the presence of and MAP kinases. Dishevelled activates JNK both in WRM-1 (Shin et al., 1999). Furthermore, Ishitani et al. (1999) Drosophila and in mammalian cells (Boutros et al., 1998; Li have shown that NLK interacts with and phosphorylates TCFs et al., 1999b). This activation requires the so-called DEP and that this is stimulated by the co-expression of TAK1 and domain of dishevelled, which is not needed for Wnt signalling. TAB1. When the DNA-binding activity of TCFs was analysed, The functions of dishevelled in both pathways are therefore it became apparent that NLK selectively blocks the DNA separate. Indeed, the activation of JNK by dishevelled appears binding of the TCF-β-catenin complexes but not that of TCF to be important for the establishment of planar cell polarity in alone. Moreover, β-catenin might be needed for formation of Drosophila, but not for stabilisation of β-catenin (Boutros et TCF-NLK complexes (Ishitani et al., 1999). NLK blocks al., 1998). Nevertheless, dishevelled might still modulate JNK double-axis formation in Xenopus induced by β-catenin but not activity by activating TAK1. Zhang et al. (1999) recently by the downstream components siamois and twin (Ishitani et showed that axin binds to the MAP kinase kinase kinase al., 1999). MEKK1 and activates JNK. The activation of JNK did not The picture that emerges from these studies can be require the APC-, GSK3β- and β-catenin-binding domains of summarized as follows. Genetic studies show that both MOM- axin, which indicates that the Wnt and MAP kinase pathways 4 (TAK1) and LIT-1 (NLK) are negative regulators of TCF have different axin-domain requirements. activity. Biochemical analysis indicates that MOM-4 (TAK1) As described above, constitutive Wnt signalling has been activates LIT-1 (NLK), which phosphorylates TCFs bound to strongly linked to the development of cancer. The negative MAP kinases in Wnt signalling 917 regulation of TCF-β-catenin by the TAK1/NLK system raises Clevers, H. and van-de-Wetering, M. (1997). TCF/LEF factor earn their the possibility that these kinases control tumorigenesis in wings. Trends Genet. 13, 485-489. humans Ð i.e. behave as tumor suppressors. To date only one Cox, R. T. and Peifer, M. (1998). Wingless signaling: the inconvenient complexities of life. Curr. Biol. 8, R140-144. study has analysed the expression and DNA sequence of TAK1 Cox, R. T., Pai, L.-M., Kirkpatrick, C., Stein, J. and Peifer, M. (1999). in human tumors (Kondo et al., 1998), but no mutations of Roles of the C terminus of armadillo in wingless signaling in Drosophila. TAK1 were detected in 49 samples of human lung cancer. Genetics 153, 319-332. Similar analysis of TAB1 and NLK in tumor types for which Crawford, H. C., Fingleton, B. M., Rudolph-Owen, L. A., Goss, K. J., β Rubinfeld, B., Polakis, P. and Matrisian, L. M. (1999). The a role for TCF- -catenin has been established might be metalloproteinase matrilysin is a target of β-catenin transactivation in performed. Note that mutations in the type II TGF-β receptor intestinal tumors. Oncogene 18, 2883-2891. have been identified in hereditary nonpolyposis colorectal Dedhar, S., Williams, B. and Hannigan, G. (1999). Integrin-linked kinase carcinomas (HNPCC), in which DNA-repair genes are (ILK): a regulator of integrin and growth-factor signalling. Trends Cell Biol. defective (Markowitz et al., 1995). The mutations block 9, 319-323. Espada, J., Perez-Moreno, M., Braga, V. M. M., Rodriguez-Viciana, P. and receptor activity and thereby might induce the escape of cells Cano, A. (1999). H-Ras activation promotes cytoplasmic accumulation and from TGF-β-mediated growth control. It will be of interest to phosphoinositide 3-OH kinase association of β-catenin in epidermal determine whether the mutations in the TGF-β receptor are keratinocytes. J. Cell Biol. 146, 967-980. correlated with activated TCF-β-catenin signalling in this Fagotto, F., Funayama, N., Glück, U. and Gumbiner, B. M. (1996). Binding to cadherins antagonizes the signaling activity of β-catenin during axis specific subset of colorectal carcinomas. Interestingly, a high formation in Xenopus. J. Cell Biol. 132, 1105-1114. frequency of mutations in β-catenin has been detected in Fagotto, F., Glück, U. and Gumbiner, B. M. (1998). Nuclear localization HNPCC tumors (Miyaki et al., 1999). A functional analysis of signal-independent and importin/karyopherin-independent nuclear import of the TAK1/NLK pathway in mammalian systems (e.g. through β-catenin. Curr. Biol. 8, 181-190. gene ablation in mice) will help to clarify the respective roles Gat, U., DasGupta, R., Degenstein, L. and Fuchs, E. (1998). De Novo hair follicle morphogenesis and hair tumors in mice expressing a truncated β- of these components in normal and tumour development. catenin in skin. Cell 95, 605-614. Hamada, F., Tomoyasu, Y., Takatsu, Y., Nakamura, M., Nagai, S., Suzuki, I thank the referees and editor for their most helpful comments, A., Fujita, F., Shibuya, H., Toyoshima, K., Ueno, N. et al. (1999). which improved the manuscript considerably. Negative regulation of Wingless signaling by D-axin, a Drosophila homolog of axin. Science 283, 1739-1742. Han, M. (1997). Gut reaction to Wnt signaling in worms. Cell 90, 581-584. Harada, N., Tamai, Y., Ishikawa, T., Sauer, B., Takaku, K., Oshima, M. REFERENCES and Taketo, M. M. (1999). Intestinal polyposis in mice with a dominant stable mutation of the β-catenin gene. EMBO J. 18, 5931-5942. Aberle, H., Bauer, A., Stappert, J., Kispert, A. and Kemler, R. (1997). β- Hart, M. J., de-los-Santos, R., Albert, I. N., Rubinfeld, B. and Polakis, P. catenin is a target for the ubiquitin-proteasome pathway. EMBO J. 16, 3797- (1998). Downregulation of β-catenin by human Axin and its association 3804. with the APC tumor suppressor, β-catenin and GSK3β. Curr. Biol. 8, 573- Aoki, M., Hecht, A., Kruse, U., Kemler, R. and Vogt, P. K. (1999). Nuclear 581. endpoint of Wnt signaling: neoplastic transformation induced by Hart, M., Concordet, J. P., Lassot, I., Albert, I., del-los-Santos, R., Durand, transactivating lymphoid-enhancing factor 1. Proc. Nat. Acad. Sci. USA 96, H., Perret, C., Rubinfeld, B., Margottin, F., Benarous, R. et al. (1999). 139-144. The F-box protein β-TrCP associates with phosphorylated β-catenin and Behrens, J., von-Kries, J. P., Kühl, M., Bruhn, L., Wedlich, D., Grosschedl, regulates its activity in the cell. Curr. Biol. 9, 207-210. R. and Birchmeier, W. (1996). Functional interaction of β-catenin with the He, T. C., Sparks, A. B., Rago, C., Hermeking, H., Zawel, L., da-Costa, L. transcription factor LEF-1. Nature 382, 638-642. T., Morin, P. J., Vogelstein, B. and Kinzler, K. W. (1998). Identification Behrens, J., Jerchow, B. A., Würtele, M., Grimm, J., Asbrand, C., Wirtz, of c-MYC as a target of the APC pathway. Science 281, 1509-1512. R., Kühl, M., Wedlich, D. and Birchmeier, W. (1998). Functional He, T.-C., Chan, T. A., Vogelstein, B. and Kinzler, K. W. (1999). PPARδ is interaction of an axin homolog, conductin, with β-catenin, APC, and an APC-regulated target of nonsteroidal anti-inflammatory drugs. Cell 99, GSK3β. Science 280, 596-599. 335-345. Behrens, J. (1999). Cadherins and catenins: role in signal transduction and Herter, P., Kuhnen, C., Müller, K. M., Wittinghofer, A. and Müller, O. tumor progression. Cancer Metast. Rev. 18, 15-30. (1999). Intracellular distribution of β-catenin in colorectal adenomas, Bhanot, P., Brink, M., Samos, C. H., Hsieh, J. C., Wang, Y., Macke, J. P., carcinomas and Peutz-Jeghers polyps. J. Cancer Res. Clin. Oncol. 125, 297- Andrew, D., Nathans, J. and Nusse, R. (1996). A new member of the 304. frizzled family from Drosophila functions as a Wingless receptor. Nature Hinck, L., Nelson, W. J. and Papkoff, J. (1994). Wnt-1 modulates cell-cell 382, 225-230. adhesion in mammalian cells by stabilizing β-catenin binding to the cell Bienz, M. (1998). TCF: transcriptional activator or repressor? Curr. Opin. Cell adhesion protein cadherin. J. Cell Biol. 124, 729-741. Biol. 10, 366-372. Hsu, S. C., Galceran, J. and Grosschedl, R. (1998). Modulation of Boutros, M., Paricio, N., Strutt, D. I. and Mlodzik, M. (1998). Dishevelled transcriptional regulation by LEF-1 in response to Wnt-1 signaling and activates JNK and discriminates between JNK pathways in planar polarity association with β-catenin. Mol. Cell Biol. 18, 4807-4818. and wingless signaling. Cell 94, 109-118. Hsu, W., Zeng, L. and Costantini, F. (1999). Identification of a domain of Bowerman, B. and Shelton, C. A. (1999). Cell polarity in the early Axin that binds to the serine/threonine protein phosphatase 2A and a self- Caenorhabditis elegans embryo. Curr. Opin. Genet. Dev. 9, 390-395. binding domain. J. Biol. Chem. 274, 3439-3445. Brannon, M., Gomperts, M., Sumoy, L., Moon, R. T. and Kimelman, D. Huber, O., Korn, R., McLaughlin, J., Ohsugi, M., Herrmann, B. G. and (1997). A β-catenin/XTcf-3 complex binds to the siamois promoter to Kemler, R. (1996). Nuclear localization of β-catenin by interaction with regulate dorsal axis specification in Xenopus. Genes Dev. 11, 2359-2370. transcription factor LEF-1. Mech. Dev. 59, 3-10. Brannon, M., Brown, J. D., Bates, R., Kimelman, D. and Moon, R. T. Hülsken, J., Birchmeier, W. and Behrens, J. (1994). E-cadherin and APC (1999). XCtBP is a XTcf-3 co-repressor with roles throughout Xenopus compete for the interaction with β-catenin and the cytoskeleton. J. Cell Biol. development. Development 126, 3159-3170. 127, 2061-2069. Cadigan, K. M. and Nusse, R. (1997). Wnt signaling: a common theme in Ikeda, S., Kishida, S., Yamamoto, H., Murai, H., Koyama, S. and Kikuchi, animal development. Genes Dev. 11, 3286-3305. A. (1998). Axin, a negative regulator of the Wnt signaling pathway, forms Cavallo, R. A., Cox, R. T., Moline, M. M., Roose, J., Polevoy, G. A., a complex with GSK-3β and β-catenin and promotes GSK-3β-dependent Clevers, H., Peifer, M. and Bejsovec, A. (1998). Drosophila Tcf and phosphorylation of β-catenin. EMBO J. 17, 1371-1384. Groucho interact to repress Wingless signalling activity. Nature 395, 604- Ishitani, T., Ninomiya-Tsuji, J., Nagai, S., Nishita, M., Meneghini, M., 608. Barker, N., Waterman, M., Bowerman, B., Clevers, H., Shibuya, H. et 918 J. Behrens

al. (1999). The TAK1-NLK-MAPK-related pathway antagonizes signalling Porfiri, E., Rubinfeld, B., Albert, I., Hovanes, K., Waterman, M. and between β-catenin and transcription factor TCF. Nature 399, 798-802. Polakis, P. (1997). Induction of a β-catenin-LEF-1 complex by wnt-1 and Jiang, J. and Struhl, G. (1998). Regulation of the Hedgehog and Wingless transforming mutants of β-catenin. Oncogene 15, 2833-2839. signalling pathways by the F-box/WD40-repeat protein Slimb. Nature 391, Prieve, M. G. and Waterman, M. L. (1999). Nuclear localization and 493-496. formation of β-catenin-lymphoid enhancer factor 1 complexes are not Kaletta, T., Schnabel, H. and Schnabel, R. (1997). Binary specification of sufficient for activation of gene expression. Mol. Cell Biol. 19, 4503-4515. the embryonic lineage in Caenorhabditis elegans. Nature 390, 294-298. Rocheleau, C. E., Downs, W. D., Lin, R., Wittmann, C., Bei, Y., Cha, Y. Kishida, S., Yamamoto, H., Ikeda, S., Kishida, M., Sakamoto, I., Koyama, H., Ali, M., Priess, J. R. and Mello, C. C. (1997). Wnt signaling and an S. and Kikuchi, A. (1998). Axin, a negative regulator of the wnt signaling APC-related gene specify endoderm in early C. elegans embryos. Cell 90, pathway, directly interacts with adenomatous polyposis coli and regulates 707-716. the stabilization of β-catenin. J. Biol. Chem. 273, 10823-10826. Rocheleau, C. E., Yasuda, J., Shin, T. H., Lin, R., Sawa, H., Okano, H., Kitagawa, M., Hatakeyama, S., Shirane, M., Matsumoto, M., Ishida, N., Priess, J. R., Davis, R. J. and Mello, C. C. (1999). WRM-1 activates the Hattori, K., Nakamichi, I., Kikuchi, A., Nakayama, K. and Nakayama, LIT-1 protein kinase to transduce anterior/posterior polarity signals in C. K. (1999). An F-box protein, FWD1, mediates ubiquitin-dependent elegans. Cell 97, 717-726. proteolysis of β-catenin. EMBO J.18, 2401-2410. Roose, J., Molenaar, M., Peterson, J., Hurenkamp, J., Brantjes, H., Kondo, M., Osada, H., Uchida, K., Yanagisawa, K., Masuda, A., Takagi, Moerer, P., van-de-Wetering, M., Destree, O. and Clevers, H. (1998). The K., Takahashi, T. and Takahashi, T. (1998). Molecular cloning of human Xenopus Wnt effector XTcf-3 interacts with Groucho-related transcriptional TAK1 and its mutational analysis in human lung cancer. Int. J. Cancer 75, repressors. Nature 395, 608-612. 559-563. Roose, J., Huls, G., van Beest, M., Moerer, P., van der Horn, K., Li, L., Yuan, H., Weaver, C. d., Mao, J., FarIII, G. H., Sussman, D., Goldschmeding, R., Logtenberg, T. and Clevers, H. (1999). Synergy Jonkers, J., Kimelman, D. and Wu, D. (1999a). Axin and Frat1 interact between the tumor suppressor APC and the β-catenin-Tcf4 target Tcf1. with Dvl and GSK, bridging Dvl to GSK in Wnt-mediated regulation of Science 285, 1923-1926. LEF-1. EMBO J. 18, 4233-4240. Rubinfeld, B., Albert, I., Porfiri, E., Fiol, C., Munemitsu, S. and Polakis, Li, L., Yuan, H., Xie, W., Mao, J., Caruso, A. M., McMahon, A., Sussman, P. (1996). Binding of GSK3β to the APC-β-catenin complex and regulation D. and Wu, D. (1999b). Dishevelled proteins lead to two signaling of complex assembly. Science 272, 1023-1026. pathways. J. Biol. Chem. 274, 129-134. Ruvkun, G. and Hobert, O. (1998). The taxonomy of developmental control Lin, R., Hill, R. J. and Priess, J. R. (1998). POP-1 and anterior-posterior fate in Caenorhabditis elegans. Science 282, 2033-2041. decisions in C. elegans embryos. Cell 92, 229-239. Sadot, E., Simcha, I., Shtutman, M., Ben-Ze’ev, A. and Geiger, B. (1998). Markowitz, S., Wang, J., Myeroff, L., Parsons, R., Sun, L., Lutterbaugh, Inhibition of β-catenin-mediated transactivation by cadherin derivatives. J., Fan, R. S., Zborowska, E., Kinzler, K. W., Vogelstein, B. et al. (1995). Proc. Nat. Acad. Sci. USA 95, 15339-15344. Inactivation of the type II TGF-β receptor in colon cancer cells with Sakanaka, C., Leong, P., Harrison, S. D. and Williams, L. T. (1999). Casein microsatellite instability. Science 268, 1336-1338. kinase Iε in the wnt pathway: regulation of β-catenin function. Proc. Nat. Meneghini, M. D., Ishitani, T., Carter, J. C., Hisamoto, N., Ninomiya- Acad. Sci. USA 96, 12548-12552. Tsuji, J., Thorpe, C. J., Hamill, D. R., Matsumoto, K. and Bowerman, Sakurai, H., Shigemori, N., Hasegawa, K. and Sugita, T. (1998). TGF-β- B. (1999). MAP kinase and Wnt pathways converge to downregulate an activated kinase 1 stimulates NF-κB activation by an NF-κB-inducing HMG-domain repressor in Caenorhabditis elegans. Nature 399, 793-797. kinase-independent mechanism. Biochem. Biophys. Res. Commun. 243, Miller, J. R. and Moon, R. T. (1996). Signal transduction through β-catenin 545-549. and specification of cell fate during embryogenesis. Genes Dev. 10, 2527- Sakurai, H., Miyoshi, H., Toriumi, W. and Sugita, T. (1999). Functional 2539. interactions of transforming growth factor β-activated kinase 1 with IκB Miyaki, M., Iijima, T., Kimura, J., Yasuno, M., Mori, T., Hayashi, Y., kinases to stimulate NF-κB activation. J. Biol. Chem. 274, 10641-10648. Koike, M., Shitara, N., Iwama, T. and Kuroki, T. (1999). Frequent Sanson, B., White, P. and Vincent, J. P. (1996). Uncoupling cadherin-based mutation of β-catenin and APC genes in primary colorectal tumors from adhesion from wingless signalling in Drosophila. Nature 383, 627-630. patients with hereditary nonpolyposis colorectal cancer. Cancer Res. 59, Schlesinger, A., Shelton, C. A., Maloof, J. N., Meneghini, M. and 4506-4509. Bowerman, B. (1999). Wnt pathway components orient a mitotic spindle Molenaar, M., van-de-Wetering, M., Oosterwegel, M., Peterson-Maduro, in the early Caenorhabditis elegans embryo without requiring gene J., Godsave, S., Korinek, V., Roose, J., Destree, O. and Clevers, H. transcription in the responding cell. Genes Dev. 13, 2028-2038. (1996). XTcf-3 transcription factor mediates β-catenin-induced axis Seeling, J. M., Miller, J. R., Gil, R., Moon, R. T., White, R. and Virshup, formation in Xenopus embryos. Cell 86, 391-399. D. M. (1999). Regulation of β-catenin signaling by the B56 subunit of Moon, R. T. and Miller, J. R. (1997). The APC tumor suppressor protein in protein phosphatase 2A. Science 283, 2089-2091. development and cancer. Trends Genet. 13, 256-258. Shibuya, H., Yamaguchi, K., Shirakabe, K., Tonegawa, A., Gotoh, Y., Morin, P. J., Sparks, A. B., Korinek, V., Barker, N., Clevers, H., Vogelstein, Ueno, N., Irie, K., Nishida, E. and Matsumoto, K. (1996). TAB1: an B. and Kinzler, K. W. (1997). Activation of β-catenin-Tcf signaling in activator of the TAK1 MAPKKK in TGF-β signal transduction. Science 272, colon cancer by mutations in β-catenin or APC. Science 275, 1787-1790. 1179-1182. Munemitsu, S., Albert, I., Souza, B., Rubinfeld, B. and Polakis, P. (1995). Shibuya, H., Iwata, H., Masuyama, N., Gotoh, Y., Yamaguchi, K., Irie, Regulation of intracellular β-catenin levels by the adenomatous polyposis K., Matsumoto, K., Nishida, E. and Ueno, N. (1998). Role of TAK1 and coli (APC) tumor-suppressor protein. Proc. Nat. Acad. Sci. USA 92, 3046- TAB1 in BMP signaling in early Xenopus development. EMBO J. 17, 1019- 3050. 1028. Ninomiya-Tsuji, J., Kishimoto, K., Hiyama, A., Inoue, J., Cao, Z. and Shin, T. H., Yasuda, J., Rocheleau, C. E., Lin, R., Soto, M., Bei, Y., Davis, Matsumoto, K. (1999). The kinase TAK1 can activate the NIK-IκB as well R. J. and Mello, C. C. (1999). MOM-4, a MAP kinase kinase kinase-related as the MAP kinase cascade in the IL-1 signalling pathway. Nature 398, 252- protein, activates WRM-1/LIT-1 kinase to transduce anterior/posterior 256. polarity signals in C. elegans. Mol. Cell 4, 275-280. Nusse, R. (1999). WNT targets. Repression and activation. Trends Genet. 15, Shirakabe, K., Yamaguchi, K., Shibuya, H., Irie, K., Matsuda, S., 1-3. Moriguchi, T., Gotoh, Y., Matsumoto, K. and Nishida, E. (1997). TAK1 Orford, K., Orford, C. C. and Byers, S. W. (1999). Exogenous expression mediates the ceramide signaling to stress-activated protein kinase/c-Jun N- of β-catenin regulates contact inhibition, anchorage-independent growth, terminal kinase. J. Biol. Chem. 272, 8141-8144. anoikis, and radiation-induced cell cycle arrest. J. Cell Biol. 146, 855-867. Sokol, S. Y. (1999). Wnt signaling and dorso-ventral axis specification in Peifer, M., Berg, S. and Reynolds, A. B. (1994). A repeating amino acid motif vertebrates. Curr. Opin. Genet. Dev. 9, 405-410. shared by proteins with diverse cellular roles. Cell 76, 789-791. Tetsu, O. and McCormick, F. (1999). β-catenin regulates expression of cyclin Perrimon, N. (1994). The genetic basis of patterned baldness in Drosophila. D1 in colon carcinoma cells. Nature 398, 422-426. Cell 76, 781-784. Thorpe, C. J., Schlesinger, A., Carter, J. C. and Bowerman, B. (1997). Wnt Peters, J. M., McKay, R. M., McKay, J. P. and Graff, J. M. (1999). Casein signaling polarizes an early C. elegans blastomere to distinguish endoderm kinase I transduces Wnt signals. Nature 401, 345-350. from mesoderm. Cell 90, 695-705. Polakis, P. (1997). The adenomatous polyposis coli (APC) tumor suppressor. van-de-Wetering, M., Cavallo, R., Dooijes, D., van-Beest, M., van-Es, J., Biochim. Biophys. Acta 1332, F127-147. Loureiro, J., Ypma, A., Hursh, D., Jones, T., Bejsovec, A. et al. (1997). MAP kinases in Wnt signalling 919

Armadillo coactivates transcription driven by the product of the Drosophila specifically activates the c-Jun N-terminal kinase signaling pathway. J. Biol. segment polarity gene dTCF. Cell 88, 789-799. Chem. 274, 2118-2125. Waltzer, L. and Bienz, M. (1998). Drosophila CBP represses the transcription Yost, C., Torres, M., Miller, J. R., Huang, E., Kimelman, D. and Moon, R. factor TCF to antagonize Wingless signalling. Nature 395, 521-525. T. (1996). The axis-inducing activity, stability, and subcellular distribution Wang, W., Zhou, G., Hu-MCT, Yao, Z. and Tan, T. H. (1997). Activation of β-catenin is regulated in Xenopus embryos by glycogen synthase kinase of the hematopoietic progenitor kinase-1 (HPK1)-dependent, stress- 3. Genes Dev. 10, 1443-1454. activated c-Jun N-terminal kinase (JNK) pathway by transforming growth Yost, C., Farr, G. R., Pierce, S. B., Ferkey, D. M., Chen, M. M. and factor β (TGF-β)-activated kinase (TAK1), a kinase mediator of TGF-β Kimelman, D. (1998). GBP, an inhibitor of GSK-3, is implicated in signal transduction. J. Biol. Chem. 272, 22771-22775. Xenopus development and oncogenesis. Cell 93, 1031-1041. Willert, K., Shibamoto, S. and Nusse, R. (1999). Wnt-induced Yu, X., Waltzer, L. and Bienz, M. (1999). A new Drosophila APC homologue dephosphorylation of axin releases β-catenin from the axin complex. Genes associated with adhesive zones of epithelial cells. Nature Cell Biol. 1, 144-151. Dev. 13, 1768-1773. Zeng, L., Fagotto, F., Zhang, T., Hsu, W., Vasicek, T. J., Perry, W. r., Lee, Yamaguchi, K., Shirakabe, K., Shibuya, H., Irie, K., Oishi, I., Ueno, N., J. J., Tilghman, S. M., Gumbiner, B. M. and Costantini, F. (1997). The Taniguchi, T., Nishida, E. and Matsumoto, K. (1995). Identification of a mouse Fused locus encodes Axin, an inhibitor of the Wnt signaling pathway member of the MAPKKK family as a potential mediator of TGF-β signal that regulates embryonic axis formation. Cell 90, 181-192. transduction. Science 270, 2008-2011. Zhang, Y., Neo, S. Y., Wang, X., Han, J. and Lin, S.-C. (1999). Axin forms Yamaguchi, K., Nagai, S., Ninomiya-Tsuji, J., Nishita, M., Tamai, K., Irie, a complex with MEKK1 and activates c-Jun NH2-terminal kinase/stress K., Ueno, N., Nishida, E., Shibuya, H. and Matsumoto, K. (1999). XIAP, activated protein kinase through domains distinct from wnt signaling. J. a cellular member of the inhibitor of apoptosis , links the Biol. Chem. 274, 35247-35524. receptors to TAB1-TAK1 in the BMP signaling pathway. EMBO J.18, 179- Zorn, A. M., Barish, G. D., Williams, B. O., Lavender, P., Klymkowsky, 187. M. W. and Varmus, H. E. (1999). Regulation of wnt signaling by sox Yao, Z., Zhou, G., Wang, X. S., Brown, A., Diener, K., Gan, H. and Tan, proteins: Xsox17α/β and Xsox3 physically interact with β-catenin. Mol. T. H. (1999). A novel human STE20-related protein kinase, HGK, that Cell 4, 487-498.