Oncogene (2008) 27, 6920–6929 & 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $32.00 www.nature.com/onc REVIEW and cancer: how does dysfunction promote tumor progression?

A Jeanes1, CJ Gottardi2 and AS Yap1

1Division of Molecular Biology, Institute for Molecular Bioscience, The University of Queensland, Brisbane, Queensland, Australia and 2Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, USA

It has long been recognized that the cell– presented on neighboring cells (Leckband and Prakasam, receptor, E-cadherin, is an important determinant of 2006). On the other side of the plasma membrane, the tumor progression, serving as a suppressor of invasion cadherin cytoplasmic tails interact with a range of and metastasis in many contexts. Yet how the loss of that link the cadherin receptor to fundamental E-cadherin function promotes tumor progression is poorly intracellular processes, including the , understood. In this review, we focus on three potential cell signaling and trafficking (Yap and Kovacs, 2003; underlying mechanisms: the capacity of E-cadherin to Bryant and Stow, 2004; Mege et al., 2006). Indeed, regulate b- signaling in the canonical Wnt path- because classical cadherins lack intrinsic enzymatic way; its potential to inhibit mitogenic signaling through activity, their interactions with intracellular events occur growth factor receptors and the possible links between through these intermediary proteins. The best-under- cadherins and the molecular determinants of epithelial stood of these associated proteins are b-catenin, which polarity. Each of these potential mechanisms provides binds directly to the distal region of the cytoplasmic tail; insights into the complexity that is likely responsible for a-catenin, which is indirectly coupled to the cadherin the tumor-suppressive action of E-cadherin. complex by association with b-catenin; and p120-catenin Oncogene (2008) 27, 6920–6929; doi:10.1038/onc.2008.343 which binds to the membrane-proximal region of the cadherin cytoplasmic tail. The biochemistry and biology Keywords: E-cadherin; b-catenin; Wnt; growth factor of the have been thoroughly encompassed in a receptor; ; metastasis number of recent review articles (Anastasiadis and Reynolds, 2000; Scott and Yap, 2006; Daugherty and Gottardi, 2007). However, it is important to emphasize that these core catenins are not the only proteins that interact, Introduction physically or functionally, with classical cadherins. Indeed, a host of other cytoplasmic effectors can bind In this review, we aim to explore the relationship the cadherin directly or indirectly. These include cell between cadherin adhesion molecules and tumor pro- signaling molecules, among them tyrosine gression. Ever since the classic demonstration that kinases and protein tyrosine phosphatases (Perez- tumor cells adhere to one another less avidly than do Moreno et al., 2003; McLachlan and Yap, 2007); non-tumor cells (Coman, 1944; Brugmans et al., 1978), cytoskeletal regulators that affect nucleation, filament the question of how adhesive dysfunction contributes to dynamics, crosslinking and motors (Mege et al., cancer biology has been an active issue in cancer 2006) and determinants of (Laprise research. The discovery that cadherin cell–cell adhesion et al., 2004; Iden et al., 2006). Accordingly, it is perhaps receptors are key regulators of tissue architecture during not surprising that classical cadherins affect many development and in tissue homeostasis (Takeichi, 1991) aspects of tissue architecture. These include establishing provided molecular candidates to link cell–cell adhesion, and maintaining cell-to-cell cohesion, thereby linking and cancer. isolated cells into cohesive populations; cell–cell recog- Classical cadherins are versatile cell–cell adhesion nition during sorting and tissue reorganization; cell- receptors. As type 1 membrane , they upon-cell locomotion and epithelial polarity. function as dynamic membrane-spanning macro- It is unlikely, though, that cadherins interact with all molecular complexes (Goodwin and Yap, 2004). The these potential partners simultaneously. Instead, there is extracellular regions are responsible for adhesive recog- emerging evidence that associations may be transient nition, binding to the ectodomains of other cadherins and regulated. For example, the actin-based motor, Myosin VI, is recruited only to E-cadherin complexes relatively late in the biogenesis of epithelial monolayers, Correspondence: Dr AS Yap, Division of Molecular Cell Biology, Institute for Molecular Bioscience, The University of Queensland, as cell–cell contacts mature (Maddugoda et al., 2007). Brisbane, Queensland, Australia. The precise biological impact of cadherin receptors is E-mail: [email protected] therefore likely to be determined by the exact set of Cadherins and cancer A Jeanes et al 6921 proteins that they associate with, something that is, in 1995, 1996). Characteristically, these mutations are turn, likely to vary depending on cellular context and accompanied by loss of heterozygosity of the remaining cell signaling. Additionally, it is important to emphasize E-cadherin allele, and correlate with tumor progression that some proteins interact with the cadherin regardless toward an invasive and metastaticphenotype. Germline of whether the receptor is bound to other cadherins on mutations have also been reported in a number of neighboring cells. This is the case for b-catenin, which families with diffuse gastric carcinoma (Guilford binds the unoccupied receptor on the cell surface and et al., 1998), suggesting that for some cancers during intracellular membrane transport (Chen et al., cadherin mutations may act early in the natural history 1999). Conversely, other protein interactions appear to of the disease, indeed behaving as a classic tumor occur when the cadherin is engaged in a productive suppressor . adhesive interaction. This is consistent with the ob- Mutation of the coding sequence, however, probably servation that classical cadherins, such as E-cadherin, accounts for only a minority of cases of cadherin can behave as adhesion-activated signaling receptors dysfunction in cancer. More commonly, cadherin (Yap and Kovacs, 2003), stimulating intracellular expression is reduced by transcriptional silencing, signal-transduction pathways that ultimately affect notably through a variety of transcription factors that processes as diverse as the cytoskeleton (Mege et al., target the E-cadherin promoter (Thiery, 2002). These 2006) and cell proliferation (Perrais et al., 2007). include transcriptional repressors of the Snail/slug Their broad-ranging effects on physiological tissue family that are overexpressed in advanced carcinomas organization make cadherins attractive targets during (reviewed in Moreno-Bueno et al., 2008); and the tumorigenesis, the disruption of which might contribute transcriptional repressors SIP1 and ZEB1, the aberrant to the aberrant morphogenetic effects in cancer. Indeed, expression of which may, in turn, reflect loss of it is now clear that classical cadherin dysfunction is a inhibitory microRNAs (Gregory et al., 2008). Interest- major contributor to cancer progression (Birchmeier ingly, many of these transcriptional events also occur in and Behrens, 1994; Thiery, 2002). Because the majority development during epithelial-to-mesenchymal transi- of solid tumors are carcinomas that arise from epithelial tions, where E-cadherin expression is characteristically tissues, the key target is the prototypical epithelial lost (Yang and Weinberg, 2008). Given increasing cadherin, E-cadherin. This large literature has been evidence that aberrant epithelial-to-mesenchymal transi- thoroughly surveyed in several earlier reviews, to which tion participates in tumor progression, this lends we refer readers for more comprehensive discussion additional weight to the general concept that physiolo- (Birchmeier and Behrens, 1994; Berx et al., 1998; Thiery, gical processes regulating cadherin expression during 2002). development may be pathogenetically active to perturb Briefly, the major lines of evidence for this conclusion E-cadherin expression in carcinomas. include: (1) the pathologic demonstration that advanced Overall, then, these several lines of evidence demon- cancers often have abnormal E-cadherin expression. strate that E-cadherin function is perturbed in many Tumor progression correlates with loss of overall epithelial carcinomas, and that this cadherin dysfunc- E-cadherin expression or loss of its normal localization tion promotes cancer progression to invasion and at cell–cell contacts (Birchmeier and Behrens, 1994; metastasis. What is less clear, though, is what mechan- Yap, 1998). (2) Evidence from both cellular and animal isms allow E-cadherin dysfunction—especially loss of models that cadherin dysfunction promotes tumor cadherin expression—to promote tumor progression. progression to invasion and metastasis (Vleminckx This question is the subject of our review. It remains an et al., 1991). One notable example was the demonstra- open issue and we focus on three major areas: the role tion that expression of a dominant-negative cadherin for b-catenin signaling; the impact of cadherins on mutant in a mouse pancreatic b-cell tumor model mitogenic signaling and the capacity for cadherins to accelerated the rate at which adenomas converted regulate epithelial cell polarity. into carcinomas (Perl et al., 1998). Conversely, over- expression of E-cadherin in those cells retarded tumor progression. More recently, Derksen et al. (2006) demonstrated that conditional deletion of E-cadherin Cadherin loss and b-catenin signaling in p53-deficient mouse mammary promoted tumor initiation and progression to invasion and The elephant in the room is b-catenin. In vertebrate metastasis. Thus, experimental manipulation established systems, b-catenin was first identified as a binding an inverse relationship between E-cadherin function and partner for classical cadherins that interacted directly tumor progression. It is important to emphasize, with the distal portion of the cadherin cytoplasmic tail however, that in mouse models loss of E-cadherin alone (Nagafuchi and Takeichi, 1989; Ozawa et al., 1989). is not sufficient to induce tumor formation (Tinkle That it played a vital role in cadherin function was et al., 2004; Tunggal et al., 2005; Derksen et al., 2006). suggested by the demonstration that cadherin mutants (3) The identification of both somatic and germline lacking the b-catenin-binding domain were often poorly E-cadherin mutations in a number of cancers (Berx adhesive (Nagafuchi and Takeichi, 1989). In contrast, et al., 1998). Somaticmutations have been most the Drosophila homolog of b-catenin, armadillo, was thoroughly characterized in lobular breast cancer, but first identified as a component of the Wnt signaling have also been identified in other tumors (Berx et al., pathway (Peifer and Wieschaus, 1990; Peifer et al.,

Oncogene Cadherins and cancer A Jeanes et al 6922 1991). Only subsequently was it discovered that might also contribute to tumor progression. And, by armadillo makes a genetically distinct contribution to implication, that components of the canonical Wnt cadherin function in flies (Sanson et al., 1996). pathway might be implicated in tumor biology. Indeed, b-Catenin/armadillo supports cadherin-adhesive func- activation of canonical Wnt signaling is a key early tion by acting as an adaptor for a range of cytoplasmic event in the vast majority of sporadicand familial proteins, many of which potentially interact, functionally colorectal cancers (Fodde and Brabletz, 2007). In the and/or physically, with the actin cytoskeleton. The best majority of cases, this hyperactive signaling occurs known is a-catenin, which incorporates into the cadherin either through inactivating mutations in APC or molecular complex by binding directly to b-catenin. mutations in b-catenin itself that render it resistant to a-Catenin has a clear and significant contribution to phosphorylation by GSK3b. Similar roles for b-catenin cadherin-based cell–cell interactions (Vasioukhin et al., signaling have been identified in skin, breast and 2001), but its mechanism of action is poorly understood hemopoietictumors (Fodde and Brabletz, 2007). Aber- (Scott and Yap, 2006). Once commonly thought to rant b-catenin signaling in the canonical Wnt pathway is anchor cadherin complexes directly to cortical actin thus a well-established contributor to tumor biology. filaments, this simple model has not been experimentally The fact that cells use this common component, verified. Alternatively, a-catenin may indirectly couple b-catenin, for the transactivation of TCF-dependent cadherins to actin filaments through proteins such as target and to support cadherin function, has long EPLIN (Abe and Takeichi, 2008), though a-catenin suggested a relationship between cadherins and Wnt can also modulate actin filament nucleation in a signaling. Intuitively, it has been attractive to postulate cadherin-independent manner (Drees et al., 2005). that cadherin loss promotes tumorigenesis by effectively Functional interaction with the actin cytoskeleton is releasing membrane-bound b-catenin into the cytosol, not the only way that b-catenin can contribute to hence stimulating canonical Wnt signaling. The key cadherin function, however, as it can bind many foundation for this idea is the demonstration that other proteins and is implicated in other aspects of binding to cadherins can antagonize Wnt signaling by cadherin biology, including post-Golgi transport of sequestering b-catenin at the membrane, thereby pre- newly synthesized cadherin to the cell surface (Chen venting it from entering the nucleus to transmit Wnt et al., 1999). signals. Thus, forced overexpression of cadherins In Wnt signaling, b-catenin acts independently of the antagonizes Wnt/b-catenin signaling in Xenopus, Dro- cadherin (Daugherty and Gottardi, 2007). Briefly, sophila and cell culture models (Heasman et al., 1994; b-catenin can function as a transcriptional co-regulator, Fagotto et al., 1996; Sanson et al., 1996; Orsulic et al., cooperating with transcription factors of the TCF 1999). Conversely, reductions in cadherin levels have (T-cell factor) family to determine gene expression. been shown to enhance b-catenin signaling events One key to the role of b-catenin in Wnt signaling is associated with fly and mouse development (Cox et al., regulation of the cytosolic pool of b-catenin that is 1999; Ciruna and Rossant, 2001). These phenotypicdata available to enter the nucleus and thereby modulate are supported by biochemical evidence that b-catenin transcription. In the absence of Wnt signaling, this uses the same binding interface to engage TCF and cytoplasmic pool is tightly controlled by a destruction cadherin ligands (reviewed in Gottardi and Gumbiner, complex, the key components of which include axin and 2001), and that the cadherin can preferentially compete the adenomatosis polyposis coli (APC) protein. Axin with TCFs for binding to b-catenin through superior binds to b-catenin directly, and through co-recruitment binding affinity of the cadherin–catenin interaction of CK1 and GSK3b promotes the phosphorylation of (Choi et al., 2006). Taken together, these data suggest serine residues in the N terminus of cytosolic b-catenin that cadherins effectively serve as a ‘sink’ for Wnt/b- required for polyubiquitylation and subsequent proteo- catenin signals, such that the level of cadherin expres- somal destruction (reviewed in Luo and Lin, 2004) sion in a cell sets a threshold over which b-catenin (Figure 1). APC is thought to out-compete Axin for protein levels must rise to gain access to the nucleus, b-catenin, promoting b-catenin flux through the bind TCFs and drive transcription. phosphorylation complex (Ha et al., 2004), and thus However, although these gain- and loss-of-function facilitating the degradation process. Thus, under basal experiments clearly demonstrate that cadherin levels can conditions free cytosolic b-catenin levels are kept low by affect b-catenin signaling, the notion that cadherin loss rapid degradation. Wnt signaling inhibits this degrada- is sufficient to promote b-catenin signaling by releasing tive process by phosphorylating and inhibiting GSK3b, it from the cadherin-bound pool is largely incorrect. For thereby allowing b-catenin to accumulate in the cytosol example, cancer cell lines that lack E-cadherin do not and enter the nucleus. manifest a corresponding upregulation of b-catenin b-Catenin signaling is an essential element in this signaling (Caca et al., 1999; van de Wetering et al., canonical Wnt pathway both during development and 2001). Moreover, depletion of E-cadherin in a mouse physiological tissue turnover in post-embryonic life model for pancreatic cancer (using the Rip-Tag system) (Cadigan and Nusse, 1997). Thus, the fact that in showed no role for b-catenin signaling in the progres- mammalian systems Wnts were first discovered in sion of these tumors (Herzig et al., 2007). This is because screens for oncogenes (Fung et al., 1985; Brown et al., although a cadherin can bind a newly synthesized 1986) suggested the attractive notion that the pathologic b-catenin, compete for its association with the Axin/ expression of these important developmental regulators APC degradation complex, and stabilize it (Sadot et al.,

Oncogene Cadherins and cancer A Jeanes et al 6923

Wnt LRP

Cadherin Frizzled

p120 p120

Catenin - β α α-Catenin -Catenin

GSK3β

APC Axin CK1

in aten β-C Brg1

nin Pontin ate -C Pygopus Hyrax β p300/CB BCL9

LEF/TC

nin ate β-C

Figure 1 Overview of b-catenin signaling and its regulation. Free cytosolic b-catenin is regulated by the combination of a destruction complex (adenomatosis polyposis coli (APC), Axin, GSK3b and CK1) and binding to the cytoplasmic domain of classical cadherin adhesion receptors. Both mechanisms limit the pool of free cytosolic b-catenin and hence tend to inhibit signaling through the canonical Wnt pathway. For further details, see text and http://stanford.edu/Brnusse?wntwindow.html.

1998; Herzig et al., 2007), cadherin loss appears to let exhibit b-catenin/TCF-dependent proliferative activity this process work in reverse, allowing the once cadherin- that can be potently attenuated by the cadherin (Eger bound b-catenin to be consumed by the degradation et al., 2000; Stockinger et al., 2001). Thus, it appears machinery. Simply, loss of the cadherin alone does not that the cadherin’s capacity to buffer b-catenin is most increase b-catenin stabilization and signaling when the critical during times when the cell is actively engaged in degradation pathway is intact. signaling. Furthermore, canonical Wnt signaling is often However, if cadherin loss alone is not sufficient to upregulated in tumors (Fodde and Brabletz, 2007). Wnt activate b-catenin signaling, loss of the cadherin ‘sink’ expression may be increased either from tumor cells can certainly amplify the cellular response to Wnt or themselves or secreted from stromal cells (notably Wnt-like signals (Cox et al., 1999; Ciruna and Rossant, macrophages). Alternatively, paracrine factors in the 2001). For example, colon cancer cells that manifest tumor environment, especially cytokines secreted by b-catenin signaling activation through mutations in the stromal cells, can amplify the intracellular Wnt signal- APC tumor suppressor gene show enhanced transcrip- transduction pathway. For example, tumor necrosis tional activation in the absence of E-cadherin (Gottardi factora secreted from macrophages stimulates Akt et al., 2001; Kuphal and Behrens, 2006). Similarly, (protein kinase B) that in turn phosphorylates and mammary epithelial cells expressing an estrogen-induci- inhibits GSK3b activity, which ultimately promotes the ble cFos estrogen receptor fusion protein (FosER) stabilization of b-catenin (Oguma et al., 2008).

Oncogene Cadherins and cancer A Jeanes et al 6924 Thus, when we consider E-cadherin’s role as a tumor metastasis (Bublil and Yarden, 2007). Indeed, a number suppressor and its relationship to b-catenin, it is perhaps of recent studies point to the potential for E-cadherin most accurate to consider that E-cadherin sets a to inhibit or modulate signaling through the EGF threshold for Wnt/b-catenin signaling: high cadherin receptor. expression—as is typically found in epithelial tissues–– Support for this idea comes from the following lines serves to keep Wnt signals off, whereas cadherin loss of evidence. First, E-cadherin co-accumulates with can potentiate Wnt signaling driven at many points in EGF-R at cell–cell contacts (Fedor-Chaiken et al., the pathway. There are, though, additional levels 2003) and can physically interact with the EGF receptor of complexity. Some intriguing studies revealed that (Hoschuetzky et al., 1994; Fedor-Chaiken et al., 2003; cadherin-mediated cell–cell adhesion and polarity Qian et al., 2004) and also with other members of the (rather than changes in cadherin expression) may affect ErbB receptor tyrosine kinase family (Ochiai et al., the strength or duration of Wnt/b-catenin signals 1994). EGF-R did not, however, interact with through a poorly defined mechanism (Greaves et al., N-cadherin, suggesting a degree of selectivity among 1999; Hamada and Bienz, 2002; McCartney et al., 2006). the classical cadherins (Fedor-Chaiken et al., 2003; Qian Further, in this discussion we have focused on the et al., 2004). The molecular basis for this interaction impact that overt loss of E-cadherin may have on remains controversial. Recombinant EGF-R can bind b-catenin signaling. A more subtle alternative is that the b-catenin in vitro (Hoschuetzky et al., 1994); however, in high-affinity cadherin–b-catenin interaction may be cells neither the b-catenin- or p120-ctn-binding regions disrupted during oncogenesis, thereby perturbing appeared to be necessary for E-cadherin and EGF-R to E-cadherin adhesion as well as releasing b-catenin to co-immunoprecipitate (Fedor-Chaiken et al., 2003; Qian signal (Nelson and Nusse, 2004). It is particularly et al., 2004). Instead, the interaction appeared to map to attractive to postulate that tyrosine phosphorylation of the extracellular domain of E-cadherin. Regardless, b-catenin and/or E-cadherin by oncogenic kinases, such E-cadherin and EGF-R have the capacity to interact as Src, may serve just this dual purpose. Both these functionally with one another. proteins, and other cadherin-binding proteins, are Second, E-cadherin can inhibit cell responsiveness to targets for tyrosine phosphorylation in cells (McLachlan EGF stimulation (Qian et al., 2004; Perrais et al., 2007). and Yap, 2007); tyrosine phosphorylation of recombinant This was first suggested by Qian et al. (2004) who b-catenin disrupts its binding to cadherin in vitro observed that mitogenicresponsiveness to EGF (mea- (Roura et al., 1999); and reduced binding of b-catenin sured by cell proliferation and activation of Ras to cadherin has been reported in growth factor- or signaling) decreased as cells grew to confluence. oncogene-transformed cells (Piedra et al., 2003; Bamji However, this desensitization was overcome by adding et al., 2006). How extensive a mechanism this may be antibodies that block E-cadherin function. Interdicting remains an open question: the cadherin–b-catenin the cadherin increased both EGF receptor autopho- interaction is not always perturbed despite phospho- sphorylation and EGF-induced DNA synthesis. This rylation of b-catenin (Piedra et al., 2003) and cadherin implied that E-cadherin exerted an effect at a receptor- adhesion is often regulated without detectable changes proximal point in the signaling pathway with ramifica- in cadherin–b-catenin association (Gumbiner, 2005). tions to downstream targets. Although these experi- Any increase in free b-catenin may also be transient ments identified a role for cadherin adhesion, they did unless accompanied by changes in b-catenin degradation not distinguish whether inhibition of EGF-R signaling through the destruction complex. Clearly, much remains occurred as a direct response to E-cadherin homophilic to be learnt about the relationship between E-cadherin ligation or more indirectly through the many other and b-catenin signaling in cancer. juxtacrine events that occur when cadherin adhesion brings cell surfaces into contact with one another. Perrais et al. (2007) addressed this issue directly, Cadherin regulation of growth factor signaling developing assays using recombinant cadherin ligands to test whether homophilicligation of E-cadherin was A second avenue for E-cadherin to influence tumorigen- sufficient to affect cell proliferation. They found that esis is by modulation of mitogenicsignaling. This notion E-cadherin ligation inhibited serum-stimulated cell was first suggested by observations that cadherin proliferation through a mechanism that required the adhesion could influence cell proliferation in the PC9 ability of cadherin to bind b-catenin, but did not appear lung carcinoma cell line (Watabe et al., 1994). Although to involve the canonical Wnt/b-catenin signaling path- these cells possess E-cadherin and b-catenin, they adhere way. Further, they demonstrated that E-cadherin poorly to one another due to lack of a-catenin. ligation inhibited EGF-induced cell proliferation. Inter- Restoration of a-catenin induced cell–cell cohesion estingly, in their experiments, E-cadherin binding did and also retarded cell proliferation, suggesting that not appear to affect either the ability of EGF-R to functional E-cadherin adhesion might participate in undergo autophosphorylation or to stimulate ERK contact inhibition of growth. Growth factor stimulation signaling. Instead, homophilicligation disrupted the is a major drive for cell proliferation and often ability of EGF-R to activate the Stat 5b signaling upregulated in tumors. Notably, many epithelial cancers pathway that leads to ERK-independent DNA synthesis. display high levels of EGF receptor (EGF-R, ErbB1), Taken together, these recent studies suggest the broad which is implicated in cell proliferation, invasion and capacity for E-cadherin to negatively regulate mitogenic

Oncogene Cadherins and cancer A Jeanes et al 6925 signaling through EGF-R. It remains to be seen whether Drosophila neuroepithelium in DE-cadherin (Shotgun) the differences between these two studies reflect the mutant embryos (Tepass et al., 1996). These findings are different experimental systems used or whether cadherin complemented by the observation that E-cadherin adhesion may affect EGF-R signaling by multiple supported assembly of tight junctions (Gumbiner pathways involving both direct pathways from homo- et al., 1988), thought to assist in segregating membrane philic ligation and possible juxtacrine effects. We do not domains, and by the recent demonstration that have a clear picture of how cadherin adhesion inhibits E-cadherin adhesions may contribute to the polarized EGF-R signaling. It is likely to be complex. Blocking targeting of basolateral membrane components (Nejsum cadherin function decreased receptor affinity for EGF and Nelson, 2007). This emphasizes that a major impact and also increased the apparent surface mobility of of cadherin may be on biogenesis of the basolateral EGF-R, as measured by FRAP (Qian et al., 2004). This membrane domain; indeed, epithelial cells begin to would be consistent with the observation that cadherin segregate apical membrane markers in response to modulated EGF-R autophosphorylation (Qian et al., substrate, even in the absence of cell–cell adhesion 2004), though not with a more secondary effect on Stat (Vega-Salas et al., 1987). However, the precise mechan- 5b signaling (Perrais et al., 2007). An interesting isms that allow E-cadherin adhesion to influence alternative model is presented in endothelial cells, where epithelial polarity remain incompletely understood. VE-cadherin interacts with the VEGF-receptor 2 The exciting advance has been the identification of (VEGF-R2). Here, VEGF stimulation induces the highly conserved groups of polarity-determining factors clathrin-dependent internalization of VEGF-R2, which that act as membrane proteins, scaffolds and signaling leads to sustained signaling from the internalized molecules to support polarized cellular asymmetry in receptors (Lampugnani et al., 2006). It is noted that many different contexts (Kemphues et al., 1988; Tepass VE-cadherin can associate with VEGF-R2 and inhibit et al., 1990; Bilder et al., 2000). There are three major its internalization. This suggests that VE-cadherin may groups of polarity determinants that function in inhibit mitogenicsignaling by VEGF-R2 by preventing epithelia; these are discussed in greater depth in other its internalization and entry into an endomembrane reviews in this issue. But briefly, the Crumbs/Stardust/ compartment. Whether something similar occurs for Discs lost complex (Crb/Sdt/Dlt) localizes to the apical E-cadherin and EGF-R is an interesting issue for future membrane and has a key function in apical membrane research. biogenesis (Tepass et al., 1990; Tepass and Knust, 1993; Bhat et al., 1999); the Par3/Par6/aPKC complex is found in the region of tight junctions at the apico-lateral E-cadherin, polarity and tumorigenesis interface (Kemphues et al., 1988; Tabuse et al., 1998; Petronczki and Knoblich, 2001) and the Scribble/Discs Finally, let us discuss the potential relationship between large/Lethal giant larvae complex (Scrib/Dlg/Lgl) is E-cadherin and polarity. Apico-basal polarization is a found at the lateral membrane (Bilder et al., 2000). defining feature of epithelial cells—most apparent in Some of these were first identified as tumor suppressor simple transporting epithelia. Ultimately, the functional genes in Drosophila (Bilder, 2004). Geneticand cellular asymmetry of membrane composition and cytoskeletal studies have implicated all these molecules in the control and organellar organization supports many aspects of of apico-basal epithelial polarity. They have also epithelial function, among them regulated transport demonstrated functional interplay, by positive and across tissue barriers in the body (Rodriguez-Boulan negative feedback, between the different complexes to and Nelson, 1989). Loss of normal plasma membrane maintain their asymmetric localization within cells polarization in epithelial cancers was an early observa- (Bilder and Perrimon, 2000; Benton and St Johnston, tion when antibodies to polarized membrane markers 2003; Tanentzapf and Tepass, 2003; Yamanaka et al., became readily available (Molitoris and Nelson, 1990). 2003). Further, loss of one protein in a complex often However, it was the subsequent discovery in Drosophila leads to mislocalization and disruption of the other of tumor suppressor genes that are physiological proteins in the complex, indicating interdependence determinants of epithelial polarity (Bilder, 2004; Lee within complexes (Tabuse et al., 1998; Bilder et al., and Vasioukhin, 2008) that has catalysed recent interest 2000). in the regulation of polarity as a determinant of The tantalizing question, then, is whether E-cadherin tumorigenesis. may influence epithelial polarity through one or several It has long been recognized that E-cadherin is of these polarity factors. If this were the case, it would functionally linked to the generation of a polarized suggest the possibility that cadherin dysfunction might epithelial phenotype. Early studies that used inhibitory promote tumorigenesis by disrupting the action of a antibodies to acutely block E-cadherin function demon- polarity determinant. There are, indeed, early hints of strated that this adhesion receptor was necessary for physical and functional links between E-cadherin and isolated cells to make productive contacts with one certain polarity factors. Consistent with their roles as another and for the subsequent processes of epithelial basolateral determinants, members of the Scrib/Dlg biogenesis, notably epithelial polarization (Gumbiner complex are found at cadherin-based cell–cell contacts. et al., 1988; Wheelock and Jensen, 1992). There is Dlg colocalizes with E-cadherin both during cellulariza- also genetic evidence to link E-cadherin and epithelial tion of the Drosophila ectoderm (Harris and Peifer, polarity: apical markers are mislocalized in the 2004) and in cultured mammalian epithelial cells, where

Oncogene Cadherins and cancer A Jeanes et al 6926 it can form a complex with the cadherin (Reuver and maintenance of polarity (Tunggal et al., 2005; Capaldo Garner, 1998; Laprise et al., 2004). Scribble, too, is and Macara, 2007). Further, polarity factors may found with E-cadherin at lateral membranes, though feedback on cadherin function (Qin et al., 2005), there is no evidence for a physical interaction to date adding a level of complexity to their functional inter- (Navarro et al., 2005; Qin et al., 2005). Moreover, relationship. expression of E-cadherin in cadherin-null cells induced the relocalization of both Dlg and Scribble into those contacts (Reuver and Garner, 1998; Navarro et al., Conclusions and future directions 2005), suggesting that cadherin adhesion can influence the subcellular localization of these polarity deter- Clearly, then, we cannot yet provide any simple answer minants. to the question of how E-cadherin dysfunction promotes E-cadherin also has links with elements of the Par tumor progression. Indeed, the many recent advances in complex. Although in mature mammalian epithelia, our knowledge would strongly suggest that there is no these proteins often show a restricted localization to the simple answer; the rich complexity of cadherin biology region of tight junctions, aPKC and Par3 can be found would perhaps make this an unreasonable ambition. in MTD1-A cells after punctate adherens junctions had Moreover, there are many other phenomena that we formed (Suzuki et al., 2002). In Drosophila, Bazooka have not discussed in this short overview, among them (Par3) also localizes with DE-cadherin in apical the potential role of other catenins (Yanagisawa et al., adherens junctions (Harris and Peifer, 2004). Moreover, 2008) and pathogeneticexpression of other cadherins cadherins can interact physically with elements of the (notably N-cadherin) when E-cadherin is lost (Wheelock Par complex. VE-cadherin was reported to bind Par3 et al., 2008). E-cadherin dysfunction is likely to affect and Par6 (Iden et al., 2006), whereas chick N-cadherin tumor cell biology in many ways. In this regard, how we formed a complex with Par3 (Afonso and Henrique, frame the relevant problems in tumor cell biology will 2006). Whether similar interactions occur with critically influence the analysis of cadherin dysfunction. E-cadherin remains to be determined. Much of the current literature, and our discussion, Finally, geneticstudies have provided perhaps the focuses on how E-cadherin may regulate tumor cell most striking evidence that places cadherin adhesion proliferation. But a major physiological impact of potentially upstream of polarity factors. In the early E-cadherin and other cadherins occurs by regulating Drosophila embryo, disruption of adherens junctions in morphogenesis (Gumbiner, 2005). This morphogenetic Armadillo mutant animals prevented Dlg from segregat- impact involves cellular processes such as cell–cell ing from the apical to the basolateral domain (Harris recognition, regulation of the cytoskeleton, dynamic and Peifer, 2004), and also perturbed the apical control of surface adhesion, many of these integrated localization of Bazooka in some studies (Bilder et al., through cell signaling. To what extent might dysregula- 2003), though not in others (Harris and Peifer, 2004). tion of mechanisms for morphogenesis contribute to Indeed, in the latter case, Bazooka was necessary for the tumor progression? The emerging notion that aberrant integrity of adherens junctions. This emphasizes the expression of epithelial-to-mesenchymal transitions con- likely complexity of genetic inter-relationships between tributes to tumor invasion and metastasis provides an cadherins and polarity determinants. A requirement for example of how a morphogeneticprocessactivein a functioning cadherin–catenin complex is also sug- development may be pathogeneticwhen aberrantly gested by studies in mice, where conditional knockout of expressed in tumors (Yang and Weinberg, 2008). a-catenin caused mislocalization of LGN, Par3 and Interestingly, the transcription factor Twist was recently aPKC (Lechler and Fuchs, 2005). observed to be upregulated when E-cadherin was Overall, then, there are encouraging signs that depleted in cultured epithelial cells (Onder et al., E-cadherin may contribute to epithelial polarization by 2008). As Twist can drive epithelial-to-mesenchymal regulating the subcellular localization of some of the transition (Yang et al., 2004), this provides a novel canonical polarity factors. The extent of such impact potential pathway for cadherin loss to drive morpho- and the mechanisms responsible are important issues geneticchanges in tumors. It is unlikely that this will be that need to be determined. Whether any such mechan- the last. ism contributes to the impact of cadherin dysfunction on tumorigenesis is also an open question. It is Acknowledgements important to emphasize that the relationship between E-cadherin and polarity is complex. Whereas acutely We thank our colleagues in our labs for their thoughtful perturbing E-cadherin can prevent cells from effectively comments and support. AJ and ASY were funded by the establishing a polarized phenotype, steady-state deple- National Health and Medical Research Council of Australia, tion of E-cadherin often does not overtly affect whereas CJG is funded by the NIH (GM076561).

References

Abe K, Takeichi M. (2008). EPLIN mediates linkage of the cadherin Afonso C, Henrique D. (2006). PAR3 acts as a molecular organizer to catenin complex to F-actin and stabilizes the circumferential actin define the apical domain of chick neuroepithelial cells. J Cell Sci belt. Proc Natl Acad Sci USA 105: 13–19. 119: 4293–4304.

Oncogene Cadherins and cancer A Jeanes et al 6927 Anastasiadis PZ, Reynolds AB. (2000). The p120 catenin family: Cox RT, Pai LM, Kirkpatrick C, Stein J, Peifer M. (1999). Roles of the complex roles in adhesion, signalling and cancer. J Cell Sci 113: C terminus of Armadillo in Wingless signaling in Drosophila. 1319–1334. Genetics 153: 319–332. Bamji SX, Rico B, Kimes N, Reichardt LF. (2006). BDNF mobilizes Daugherty RL, Gottardi CJ. (2007). Phospho-regulation of beta- synaptic vesicles and enhances synapse formation by disrupting catenin adhesion and signaling functions. Physiology (Bethesda) 22: cadherin–beta-catenin interactions. J Cell Biol 174: 289–299. 303–309. Benton R, St Johnston D. (2003). Drosophila PAR-1 and 14-3-3 inhibit Derksen PW, Liu X, Saridin F, van der Gulden H, Zevenhoven J, Bazooka/PAR-3 to establish complementary cortical domains in Evers B et al. (2006). Somatic inactivation of E-cadherin and p53 in polarized cells. Cell 115: 691–704. mice leads to metastatic lobular mammary carcinoma through Berx G, Becker K-F, Hofler H, van Roy F. (1998). Mutations of the induction of anoikis resistance and angiogenesis. Cancer Cell 10: human E-cadherin (CDH1) gene. Hum Mutat 12: 226–237. 437–449. Berx G, Cleton-Jansen A-M, Nollet F, de Leeuw WJF, van de Vijver Drees F, Pokutta S, Yamada S, Nelson WJ, Weis WI. (2005). Alpha- MJ, Cornelisse C et al. (1995). E-cadherin is a tumour/invasion catenin is a molecular switch that binds E-cadherin–beta-catenin suppressor gene mutated in human lobular breast cancers. EMBO J and regulates actin-filament assembly. Cell 123: 903–915. 14: 6107–6115. Eger A, Stockinger A, Schaffhauser B, Beug H, Foisner R. (2000). Berx G, Cleton-Jansen A-M, Strumane K, de Leeuw WJF, Nollet F, Epithelial mesenchymal transition by c-Fos estrogen receptor van Roy F et al., (1996). E-cadherin is inactivated in a majority of activation involves nuclear translocation of beta-catenin and invasive human lobular breast cancers by truncation mutations upregulation of beta-catenin/lymphoid enhancer binding factor-1 throughout its extracellular domain. Oncogene 13: 1919–1925. transcriptional activity. J Cell Biol 148: 173–188. Bhat MA, Izaddoost S, Lu Y, Cho KO, Choi KW, Bellen HJ. (1999). Fagotto F, Funayama N, Gluck U, Gumbiner B. (1996). Binding to Discs Lost, a novel multi-PDZ domain protein, establishes and cadherins antagonizes the signaling activity of beta-catenin during maintains epithelial polarity. Cell 96: 833–845. axis formation in Xenopus. J Cell Biol 132: 1105–1114. Bilder D. (2004). Epithelial polarity and differentiation control: links Fedor-Chaiken M, Hein PW, Stewart JC, Brackenbury R, Kinch MS. from the Drosophila neoplastictumor suppressors. Genes Dev 18: (2003). E-cadherin binding modulates EGF receptor activation. Cell 1909–1925. Commun Adhes 10: 105–118. Bilder D, Li M, Perrimon N. (2000). Cooperative regulation of cell Fodde R, Brabletz T. (2007). Wnt/beta-catenin signaling in polarity and growth by Drosophila tumor suppressors. Science 289: cancer stemness and malignant behavior. Curr Opin Cell Biol 19: 113–116. 150–158. Bilder D, Perrimon N. (2000). Localization of apical epithelial Fung YK, Shackleford GM, Brown AM, Sanders GS, Varmus HE. determinants by the basolateral PDZ protein Scribble. Nature 403: (1985). Nucleotide sequence and expression in vitro of cDNA 676–680. derived from mRNA of int-1, a provirally activated mouse Bilder D, Schober M, Perrimon N. (2003). Integrated activity of mammary oncogene. Mol Cell Biol 5: 3337–3344. PDZ protein complexes regulates epithelial polarity. Nat Cell Biol 5: Goodwin M, Yap AS. (2004). Classical cadherin adhesion molecules: 53–58. coordinating cell adhesion, signaling and the cytoskeleton. J Mol Birchmeier W, Behrens J. (1994). Cadherin expression in carcinomas: Histol 35: 839–844. role in the formation of cell junctions and the prevention of Gottardi CJ, Gumbiner BM. (2001). Adhesion signaling: how b- invasiveness. Biochim Biophys Acta 1198: 11–26. catenin interacts with its partners. Curr Biol 11: R792–R794. Brown AM, Wildin RS, Prendergast TJ, Varmus HE. (1986). A Gottardi CJ, Wong E, Gumbiner BM. (2001). E-cadherin suppresses retrovirus vector expressing the putative mammary oncogene int-1 cellular transformation by inhibiting beta-catenin signaling in an causes partial transformation of a mammary epithelial cell line. Cell adhesion-independent manner. J Cell Biol 153: 1049–1060. 46: 1001–1009. Greaves S, Sanson B, White P, Vincent JP. (1999). A screen for Brugmans M, Cassiman JJ, van den Berghe H. (1978). Selective identifying genes interacting with armadillo, the Drosophila homo- adhesion and impaired adhesive properties of transformed cells. log of beta-catenin. Genetics 153: 1753–1766. J Cell Sci 33: 121–132. Gregory PA, Bert AG, Paterson EL, Barry SC, Tsykin A, Farshid G Bryant DM, Stow JL. (2004). The ins and outs of E-cadherin et al. (2008). The miR-200 family and miR-205 regulate epithelial to trafficking. Trends Cell Biol 14: 427–434. mesenchymal transition by targeting ZEB1 and SIP1. Nat Cell Biol Bublil EM, Yarden Y. (2007). The EGF receptor family: spearheading a 10: 593–601. merger of signaling and therapeutics. Curr Opin Cell Biol 19: 124–134. Guilford P, Hopkins J, Harraway J, McLeod M, McLeod N, Harawira Caca K, Kolligs FT, Ji X, Hayes M, Qian J, Yahanda A et al. (1999). P et al. (1998). E-cadherin germline mutations in familial gastric Beta- and gamma-catenin mutations, but not E-cadherin inactiva- cancer. Nature 392: 402–405. tion, underlie T-cell factor/lymphoid enhancer factor transcriptional Gumbiner B, Stevenson B, Grimaldi A. (1988). The role of the cell deregulation in gastric and pancreatic cancer. Cell Growth Differ 10: adhesion molecule uvomorulin in the formation and maintenance of 369–376. the epithelial junctional complex. J Cell Biol 197: 1575–1587. Cadigan KM, Nusse R. (1997). Wnt signaling: a common theme in Gumbiner BM. (2005). Regulation of cadherin-mediated adhesion in animal development. Genes Dev 11: 3286–3305. morphogenesis. Nat Rev Mol Cell Biol 6: 622–634. Capaldo CT, Macara IG. (2007). Depletion of E-cadherin disrupts Ha NC, Tonozuka T, Stamos JL, Choi HJ, Weis WI. (2004). establishment but not maintenance of cell junctions in Madin-– Mechanism of phosphorylation-dependent binding of APC to Darby canine kidney epithelial cells. Mol Biol Cell 18: 189–200. beta-catenin and its role in beta-catenin degradation. Mol Cell 15: Chen Y-H, Stewart DB, Nelson WJ. (1999). Coupling assembly of the 511–521. E-cadherin/b-catenin complex to efficient endoplasmic reticulum Hamada F, Bienz M. (2002). A Drosophila APC tumour suppressor exit and basal–lateral membrane targeting of E-cadherin in homologue functions in cellular adhesion. Nat Cell Biol 4: 208–213. polarized MDCK cells. J Cell Biol 144: 687–699. Harris TJ, Peifer M. (2004). -dependent and Choi HJ, Huber AH, Weis WI. (2006). Thermodynamics of beta- -independent steps in the establishment of epithelial cell polarity catenin–ligand interactions: the roles of the N- and C-terminal tails in Drosophila. J Cell Biol 167: 135–147. in modulating binding affinity. J Biol Chem 281: 1027–1038. Heasman J, Crawford A, Goldstone K, Garner-Hamrick P, Gumbiner Ciruna B, Rossant J. (2001). FGF signaling regulates mesoderm cell B, McCrea P et al. (1994). Overexpression of cadherins and fate specification and morphogenetic movement at the primitive underexpression of beta-catenin inhibit dorsal mesoderm induction streak. Dev Cell 1: 37–49. in early Xenopus embryos. Cell 79: 791–803. Coman DR. (1944). Decreased mutual adhesiveness, a property of cells Herzig M, Savarese F, Novatchkova M, Semb H, Christofori G. from squamous cell carcinomas. Cancer Res 4: 625–629. (2007). Tumor progression induced by the loss of E-cadherin

Oncogene Cadherins and cancer A Jeanes et al 6928 independent of beta-catenin/Tcf-mediated Wnt signaling. Oncogene Onder TT, Gupta PB, Mani SA, Yang J, Lander ES, Weinberg RA. 26: 2290–2298. (2008). Loss of E-cadherin promotes metastasis via multiple Hoschuetzky H, Aberle H, Kemler R. (1994). b-Catenin mediates the downstream transcriptional pathways. Cancer Res 68: 3645–3654. interaction of the cadherin–catenin complex with epidermal growth OrsulicS, Huber O, Aberle H, Arnold S, Kemler R. (1999). E-cadherin factor receptor. J Cell Biol 127: 1375–1380. binding prevents beta-catenin nuclear localization and beta-catenin/ Iden S, Rehder D, August B, Suzuki A, Wolburg-Buchholz K, LEF-1-mediated transactivation. J Cell Sci 112: 1237–1245. Wolburg H et al. (2006). A distinct PAR complex associates Ozawa M, Baribault H, Kemler R. (1989). The cytoplasmic domain of physically with VE-cadherin in vertebrate endothelial cells. EMBO the uvomorulin associates with three Rep 7: 1239–1246. independent proteins structurally related in different species. EMBO J Kemphues KJ, Priess JR, Morton DG, Cheng NS. (1988). Identifica- 8: 1711–1717. tion of genes required for cytoplasmic localization in early Peifer M, Rauskolb C, Williams M, Riggleman B, Wieschaus E. C. elegans embryos. Cell 52: 311–320. (1991). The segment polarity gene armadillo interacts with the Kuphal F, Behrens J. (2006). E-cadherin modulates Wnt-dependent wingless signaling pathway in both embryonicand adult pattern transcription in colorectal cancer cells but does not alter formation. Development 111: 1029–1043. Wnt-independent gene expression in fibroblasts. Exp Cell Res 312: Peifer M, Wieschaus E. (1990). The segment polarity gene armadillo 457–467. encodes a functionally modular protein that is the Drosophila Lampugnani MG, Orsenigo F, Gagliani MC, Tacchetti C, Dejana E. homolog of human . Cell 63: 1167–1176. (2006). Vascular endothelial cadherin controls VEGFR-2 inter- Perez-Moreno M, Jamora C, Fuchs E. (2003). Sticky business: nalization and signaling from intracellular compartments. J Cell orchestrating cellular signals at adherens junctions. Cell 112: 535–548. Biol 174: 593–604. Perl A-K, Wilgenbus P, Dahl U, Semb H, Christofori G. (1998). A Laprise P, Viel A, Rivard N. (2004). Human homolog of disc-large is causal role for E-cadherin in the transition from adenoma to required for adherens junction assembly and differentiation of carcinoma. Nature 392: 190–193. human intestinal epithelial cells. J Biol Chem 279: 10157–10166. Perrais M, Chen X, Perez-Moreno M, Gumbiner BM. (2007). Lechler T, Fuchs E. (2005). Asymmetric cell divisions promote E-cadherin homophilic ligation inhibits cell growth and epidermal stratification and differentiation of mammalian skin. Nature 437: growth factor receptor signaling independently of other cell 275–280. interactions. Mol Biol Cell 18: 2013–2025. Leckband D, Prakasam A. (2006). Mechanism and dynamics of Petronczki M, Knoblich JA. (2001). DmPAR-6 directs epithelial cadherin adhesion. Annu Rev Biomed Eng 8: 259–287. polarity and asymmetriccelldivision of neuroblasts in Drosophila. Lee M, Vasioukhin V. (2008). Cell polarity and cancer-cell and tissue Nat Cell Biol 3: 43–49. polarity as a non-canonical tumor suppressor. J Cell Sci 121: 1141–1150. Piedra J, Miravet S, Castano J, Palmer HG, Heisterkamp N, Garcia de Luo W, Lin SC. (2004). Axin: a master scaffold for multiple signaling Herreros A et al. (2003). p120 catenin-associated Fer and Fyn pathways. Neurosignals 13: 99–113. tyrosine kinases regulate beta-catenin Tyr-142 phosphorylation Maddugoda MP, Crampton MS, Shewan AM, Yap AS. (2007). and beta-catenin–alpha-catenin interaction. Mol Cell Biol 23: Myosin VI and cooperate during the morphogenesis of 2287–2297. cadherin cell–cell contacts in mammalian epithelial cells. J Cell Biol Qian X, Karpova T, Sheppard AM, McNally J, Lowy DR. (2004). 178: 529–540. E-cadherin-mediated adhesion inhibits ligand-dependent activation McCartney BM, Price MH, Webb RL, Hayden MA, Holot LM, Zhou of diverse receptor tyrosine kinases. EMBO J 23: 1739–1784. M et al. (2006). Testing hypotheses for the functions of APC family Qin Y, Capaldo C, Gumbiner BM, Macara IG. (2005). The proteins using null and truncation alleles in Drosophila. Develop- mammalian Scribble polarity protein regulates epithelial cell ment 133: 2407–2418. adhesion and migration through E-cadherin. J Cell Biol 171: McLachlan RW, Yap AS. (2007). Not so simple: the complexity of 1061–1071. phosphotyrosine signaling at cadherin adhesive contacts. J Mol Med Reuver SM, Garner CC. (1998). E-cadherin mediated cell adhesion 85: 545–554. recruits SAP97 into the cortical cytoskeleton. J Cell Sci 111(Part 8): Mege RM, Gavard J, Lambert M. (2006). Regulation of cell–cell 1071–1080. junctions by the cytoskeleton. Curr Opin Cell Biol 18: 541–548. Rodriguez-Boulan E, Nelson WJ. (1989). Morphogenesis of the Molitoris BA, Nelson WJ. (1990). Alterations in the establishment and polarized epithelial cell phenotype. Science 245: 718–725. maintenance of epithelial cell polarity as a basis for disease Roura S, Miravet S, Piedra J, Garcia de Herreros A, Dunach M. processes. J Clin Invest 85: 3–9. (1999). Regulation of E-cadherin/catenin association by tyrosine Moreno-Bueno G, Portillo F, Cano A. (2008). Transcriptional phosphorylation. J Biol Chem 274: 36734–36740. regulation of cell polarity in EMT and cancer. Oncogene 27: Sadot E, Simcha I, Shtutman M, Ben-Ze’ev A, Geiger B. (1998). 6958–6969. Inhibition of beta-catenin-mediated transactivation by cadherin Nagafuchi A, Takeichi M. (1989). Transmembrane control of derivatives. Proc Natl Acad Sci USA 95: 15339–15344. cadherin-mediated cell adhesion: a 94 kDa protein functionally Sanson B, White P, Vincent JP. (1996). Uncoupling cadherin-based associated with a specific region of the cytoplasmic domain of adhesion from wingless signalling in Drosophila. Nature 383: E-cadherin. Cell Regul 1: 37–44. 627–630. Navarro C, Nola S, Audebert S, Santoni MJ, Arsanto JP, Ginestier C Scott JA, Yap AS. (2006). Cinderella no longer: alpha-catenin steps et al. (2005). Junctional recruitment of mammalian Scribble relies out of cadherin’s shadow. J Cell Sci 119: 4599–4605. on E-cadherin engagement. Oncogene 24: 4330–4339. Stockinger A, Eger A, Wolf J, Beug H, Foisner R. (2001). E-cadherin Nejsum LN, Nelson WJ. (2007). A molecular mechanism directly regulates cell growth by modulating proliferation-dependent beta- linking E-cadherin adhesion to initiation of epithelial cell surface catenin transcriptional activity. J Cell Biol 154: 1185–1196. polarity. J Cell Biol 178: 323–335. Suzuki A, Ishiyama C, Hashiba K, Shimizu M, Ebnet K, Ohno S. Nelson WJ, Nusse R. (2004). Convergence of Wnt, beta-catenin, and (2002). aPKC kinase activity is required for the asymmetric cadherin pathways. Science 303: 1483–1487. differentiation of the premature junctional complex during epithelial Ochiai A, Akimoto S, Kanai Y, Shibata T, Oyama T, Hirohashi S. cell polarization. J Cell Sci 115: 3565–3573. (1994). c-erbB-2 gene product associates with catenins in human Tabuse Y, Izumi Y, Piano F, Kemphues KJ, Miwa J, Ohno S. (1998). cancer cells. Biochem Biophys Res Comm 205: 73–78. Atypical protein kinase C cooperates with PAR-3 to establish Oguma K, Oshima H, Aoki M, Uchio R, Naka K, Nakamura S et al. embryonicpolarity in Caenorhabditis elegans. Development 125: (2008). Activated macrophages promote Wnt signalling through 3607–3614. tumour necrosis factor-alpha in gastric tumour cells. EMBO J 27: Takeichi M. (1991). Cadherin cell adhesion receptors as a morpho- 1671–1681. geneticregulator. Science 251: 1451–1455.

Oncogene Cadherins and cancer A Jeanes et al 6929 Tanentzapf G, Tepass U. (2003). Interactions between the crumbs, junctions while segregation of a basolateral marker requires cell–cell lethal giant larvae and bazooka pathways in epithelial polarization. interactions. J Cell Biol 104: 905–916. Nat Cell Biol 5: 46–52. Vleminckx K, Vakaet L, Mareel M, Fiers W, Van Roy F. (1991). Tepass U, Gruszynski-DeFeo E, Haag TA, Omatyar L, Torok T, Geneticmanipulation of E-cadherinexpression by epithelial tumor Hartenstein V. (1996). shotgun encodes Drosophila E-cadherin and cells reveals an invasion suppressor role. Cell 66: 107–119. is preferentially required during cell rearrangement in the neur- Watabe M, Nagafuchi A, Tsukita S, Takeichi M. (1994). Induction of ectoderm and other morphogenetically active epithelia. Genes Dev polarized cell–cell association and retardation of growth by 10: 672–685. activation of the E-cadherin–catenin adhesion system in a dispersed Tepass U, Knust E. (1993). Crumbs and stardust act in a genetic carcinoma line. J Cell Biol 127: 247–256. pathway that controls the organization of epithelia in Drosophila Wheelock MJ, Jensen PJ. (1992). Regulation of keratinocyte inter- melanogaster. Dev Biol 159: 311–326. cellular junction organization and epidermal morphogenesis by Tepass U, Theres C, Knust E. (1990). crumbs encodes an EGF-like E-cadherin. J Cell Biol 117: 415–425. protein expressed on apical membranes of Drosophila epithelial cells Wheelock MJ, Shintani Y, Maeda M, Fukumoto Y, Johnson KR. and required for organization of epithelia. Cell 61: 787–799. (2008). Cadherin switching. J Cell Sci 121: 727–735. Yamanaka T, Horikoshi Y, Sugiyama Y, Ishiyama C, Suzuki A, Thiery J-P. (2002). Epithelial–mesenchymal transitions in tumour Hirose T et al. (2003). Mammalian Lgl forms a protein complex progression. Nat Rev Cancer 2: 442–454. with PAR-6 and aPKC independently of PAR-3 to regulate Tinkle CL, Lechler T, Pasolli HA, Fuchs E. (2004). Conditional epithelial cell polarity. Curr Biol 13: 734–743. targeting of E-cadherin in skin: insights into hyperproliferative and Yanagisawa M, Huveldt D, Kreinest P, Lohse CM, Cheville JC, degenerative responses. Proc Natl Acad Sci USA 101: 552–557. Parker AS et al. (2008). A p120 catenin isoform switch affects Rho Tunggal JA, Helfrich I, Schmitz A, Schwarz H, Gunzel D, Fromm M activity, induces tumor cell invasion, and predicts metastatic disease. et al. (2005). E-cadherin is essential for in vivo epidermal barrier J Biol Chem 283: 18344–18354. function by regulating tight junctions. EMBO J 24: 1146–1156. Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C van de Wetering M, Barker N, Harkes IC, van der Heyden M, et al. (2004). Twist, a master regulator of morphogenesis, plays an Dijk NJ, Hollestelle A et al. (2001). Mutant E-cadherin breast essential role in tumor metastasis. Cell 117: 927–939. cancer cells do not display constitutive Wnt signaling. Cancer Res Yang J, Weinberg RA. (2008). Epithelial–mesenchymal transition: at 61: 278–284. the crossroads of development and tumor metastasis. Dev Cell 14: Vasioukhin V, Bauer C, Degenstein L, Wise B, Fuchs E. (2001). 818–829. Hyperproliferation and defects in epithelial polarity upon condi- Yap AS. (1998). The morphogenetic role of cadherin cell adhesion tional ablation of alpha-catenin in skin. Cell 104: 605–617. molecules in human cancer: a thematic review. Cancer Invest 16: Vega-Salas DE, Salas PJ, Gundersen D, Rodriguez-Boulan E. (1987). 252–261. Formation of the apical pole of epithelial (Madin–Darby canine Yap AS, Kovacs EM. (2003). Direct cadherin-activated cell kidney) cells: polarity of an apical protein is independent of tight signaling: a view from the plasma membrane. J Cell Biol 160: 11–16.

Oncogene