Transcriptional regulation of cell polarity in EMT and cancer - Nature

Oncogene (2008) 27, 6958–6969

& 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $32.00 www.nature.com/onc

REVIEW

Transcriptional regulation of cell polarity in EMT and cancer G Moreno-Bueno, F Portillo and A Cano Departamento de Bioquı´mica, Universidad Autońoma de Madrid, Instituto de Investigaciones Biome´dicas ‘Alberto Sols’, Consejo Superior de Investigaciones Cientı´ficas-Universidad Autońoma de Madrid, Madrid, Spain

The epithelial-to-mesenchymal transition (EMT) is a crucial process in tumour progression providing tumour cells with the ability to escape from the primary tumour, to migrate to distant regions and to invade tissues. EMT requires a loss of cell–cell adhesion and apical–basal polarity, as well as the acquisition of a fibroblastoid motile phenotype. Several transcription factors have emerged in recent years that induce EMT, with important implications for tumour progression. However, their effects on cell polarity remain unclear. Here, we have re-examined the data available related to the effect of EMT related transcription factors on epithelial cell plasticity, focusing on their impact on cell polarity. Transcriptional and post-transcriptional regulatory mechanisms mediated by several inducers of EMT, in particular the ZEB and Snail factors, downregulate the expression and/or functional organization of core polarity complexes. We also summarize data on the expression of cell polarity genes in human tumours and analyse genetic interactions that highlight the existence of complex regulatory networks converging on the regulation of cell polarity by EMT inducers in human breast carcinomas. These recent observations provide new insights into the relationship between alterations in cell polarity components and EMT in cancer, opening new avenues for their potential use as therapeutic targets to prevent tumour progression. Oncogene (2008) 27, 6958–6969; doi:10.1038/onc.2008.346 Keywords: cell polarity; EMT; Snail; ZEB; tumour markers; interacting networks

Introduction Metastasis is the most important cause of morbidity and mortality in human cancers. The epithelial-to-mesenchymal transition (EMT) is an essential process during embryogenesis (Barrallo-Gimeno and Nieto, 2005) and its pathological activation during tumour development Correspondence: Professor A Cano, Instituto de Investigaciones Biome´dicas ‘Alberto Sols’ CSIC-UAM, Arturo Duperier 4, Madrid 28029, Spain. E-mail: [email protected]

can lead primary tumours to metastasize (Thiery, 2003; Gupta and Massague´, 2006). EMT is typically characterized by the loss of cell–cell adhesion and apical– basal cell polarity, as well as the increased motility of cells (Thiery, 2003). Cell adhesion and polarity in epithelia depends on the formation of adherens junctions in which E-cadherin is a key determinant, providing the physical structure for both cell–cell attachment and the recruitment of signalling complexes (reviewed by Knust and Bossinger, 2002; Perez-Moreno et al., 2003). One of the earliest steps in EMT is the loss of E-cadherin function, and in fact it is generally accepted that EMT-inducing factors initiate epithelial reorganization by impairing the expression or function of E-cadherin (Peinado et al., 2004; Jeanes et al., 2008). The characterization of E-cadherin regulation during EMT has provided important insights into the molecular mechanisms involved in the loss of cell–cell adhesion and in the acquisition of migratory properties during carcinoma progression (Peinado et al., 2007). To date, many different extracellular cues have been shown to trigger epithelial dedifferentiation and EMT, such as those involving transforming growth factor-b (TGF-b), Notch, fibroblast growth factor and Wnt signalling pathways (Barrallo-Gimeno and Nieto, 2005; De Craene et al., 2005a; Huber et al., 2005; Thiery and Sleeman, 2006). Most of the signals inducing EMT exert their action through the modulation of transcription factors that repress epithelial genes, such as those encoding E-cadherin and cytokeratins, and that activate transcription programmes that specify fibroblast-like motility and an invasive phenotype (Thiery and Sleeman, 2006; Peinado et al., 2007). Several transcription factors have been said to drive EMT, including members of the Snail and basic Helix Loop Helix (bHLH) families, and two double zinc finger and homeodomain (ZEB) factors (reviewed by Peinado et al., 2007). Extensive analysis of such transcription factors has clearly demonstrated their implication in the loss of cell–cell adhesion and in gaining motility, although an analysis of their impact on cell polarity is still pending. Here, we have re-examined the existing data on the effect and relationship of the EMT transcription factors on epithelial cell plasticity, focusing our attention on their impact on cell polarity mechanisms. We also summarize the data available on the expression of cell polarity genes and their transcriptional regulators in human tumours, which

Polarity regulation in EMT G Moreno-Bueno et al

6959

highlight the existence of complex regulatory networks that converge on the regulation of cell polarity in tumours. Apart from their potential as new tumour markers, the existence of complex networks involving EMT regulators that control cell polarity is becoming evident. Polarity-generating protein complexes at a glance Although our present understanding of the protein complexes involved in the generation of polarity has been dealt with in more specialized reviews (Dow and Humbert, 2007; Lee and Vasioukhin, 2008), to understand the significance of the alteration in the expression of cell polarity genes during tumour development, we will briefly summarize the epithelial cell polarity systems. Epithelial cells maintain two types of cell polarity, planar and apical–basal polarity (Dow and Humbert, 2007). Planar polarity is the polarization of cells across the two dimensions of the epithelial sheet and it will not be considered here (reviewed by Zallen, 2007). The apical–basal polarity of epithelial cells in an epithelium is evident from the presence of two specialized plasma membrane domains, the apical surface facing the lumen and a basolateral surface that contacts the adjacent cells and the underlying connective tissue. Apical–basal polarity may also be present in multilayer epithelial tissues, but in this case, the apical cell surface is in contact with the upper epithelial layer. The asymmetrical distribution of lipids and proteins between both apical and basal domains reflects their different functions, and it is the result of both polarized trafficking and the presence of a physical frontier established by the apical junctional complex comprised of tight junctions and adherens junctions. Tight junctions provide a tight seal between neighbouring cells, which is essential for the epithelium to function as a barrier, whereas adherens junctions maintain the adhesion between neighbouring cells. The protein composition and barrier properties of this apical junctional complex have recently been reviewed (Perez-Moreno et al., 2003; Perez-Moreno and Fuchs, 2006; Niessen, 2007; Hartsock and Nelson, 2008). As such, tight and adherens junctions are composed of transmembrane proteins that adhere to similar proteins in the adjacent cell. The transmembrane region of the tight junctions is composed mainly of claudins, tetraspan proteins with two extracellular loops. In mammals, the claudin gene family contains at least 24 members (Anderson et al., 2004) and the cytoplasmic domain of these proteins interacts with occludin and several ZO-proteins (ZO1-3) to form the plaque that associates with the cytoskeleton (Tsukita et al., 1997). Adherens junctions are mediated by Ca2 þ -dependent homophilic interactions of cadherins, single-span membrane proteins with more than 100 members in the vertebrate gene family grouped into six subfamilies (Nollet et al., 2000). Cadherins interact with cytoplasmic catenins that link the cadherin/catenin complex to the actin cytoskeleton (Perez-Moreno and Fuchs, 2006).

The assembly and localization of the apical junctional complex require a set of conserved polarity-generating protein complexes. To date, three core interacting protein complexes have been identified in mammals that participate in apical–basal cell polarity and that influence the assembly and localization of the junctional complexes: (1) the PAR (Par6/Par3/atypical protein kinase C (aPKC)) complex; (2) the CRB (Crb/Pals/Patj) complex; and (3) the SCRIB (Scrib/Dlg/Lgl) complex (reviewed by Dow and Humbert, 2007; Assemat et al., 2008; Aranda et al., 2008; Humbert et al., 2008). The PAR complex was initially identified in Caenorabditis elegans mutants (for partitioned defective; Kemphues et al., 1988), and in mammals, it is composed of two scaffold proteins (PAR6 and PAR3) and an aPKC. To date, two PAR3 (PARD3, PARD3B), three PAR6 (PARD6A, PARD6B, PARD6G) and two aPKC (PRKCI, PRKCZ) genes have been identified in mammalian genomes. The PAR complex is localized to the apical junction domain, and significant evidence indicates that it has an important function in the assembly of tight junctions (reviewed by Goldstein and Macara, 2007; Ebnet, 2008). The CRB and SCRIB complexes were initially identified in Drosophila melanogaster when screening mutations responsible for epithelial defects (reviewed by Assemat et al., 2008). Mammalian CRB complex localizes to the apical membrane and are made up of the transmembrane protein Crumbs (Crb) and the cytoplasmic scaffolding proteins PALS1 (the homologue of Drosophila Stardust, Sdt, a MAGUK protein) and PATJ (Pals-associated tight junction protein, the homologue of Dpatj). At present, there are three CRB, two Pals and one PATJ gene known to exist in humans (Assemat et al., 2008). The composition and function of the CRB complex in the mammalian retina have been recently reviewed (Gosens et al., 2008), where mutations in the human CRB1 gene cause autosomal recessive retinitis pigmentosa and autosomal Leber congenital amaurosis (Richard et al., 2006). The mammalian SCRIB complex is localized in the basolateral domain of epithelial cells and it is comprised of three proteins, Scribble (Scrib), Disc large (Dlg) and Lethal giant larvae (Lgl) (reviewed by Vasioukhin, 2006). Four DLG (DLG1-4) and two LGL (LGL1, LGL2) have now been identified in the mammalian genomes. In addition to these three ‘core’ cell polarity complexes, further components and protein kinases are increasingly being implicated in the organization of cell polarity (Assemat et al., 2008). There is now convincing evidence that the establishment of cell polarity is the result of mutually antagonistic interactions between the PAR, CRB and SCRIB complexes, where the Par6/Par3/aPKC complex has a pivotal function and Rac1/Cdc42 GTPases activation triggers the process (reviewed by Suzuki and Ohno, 2006; Etienne-Manneville, 2008; Figure 1). Essentially, Cdc42 that is activated is recruited by Par6 to the Par complex where it causes the activation of aPKC. This kinase, in turn, phosphorylates Par3, thereby allowing the formation of an active Par complex at the apical domain and promoting the assembly of the junctional Oncogene


6960 Apical CRB PALS PATJ CDC42

TJ/AJ

Cell-cell interaction

TJ/AJ

PAR6 SCRIBB DLG

LGL

PAR3 aPKC

Basal Figure 1 Schematic view of polarity-generating mechanism. Cell polarity complexes PAR (Par3-Par6-aPKC) and CRB (CrumbsPals1-Patj) localize to the apical membrane and promote apical membrane identity. This process is triggered by Cdc42 activation, which is mediated by cell-cell interaction. SCRIB complex is localized at the basolateral domain counteracting the function of the PAR complex and promotes basal membrane identity. SCRIB complex is reciprocally inactivated by aPKC-dependent phosphorylation of Lgl protein.

structure. Localization of active Par complexes to the apical domain is stabilised by the CRB complex whose distribution is reciprocally dependent on the PAR complex. In this scenario, the basolaterally located Scrib complex functions as an antagonist of the apical localization of the active PAR complex. Lgl proteins of the Scrib complex compete with Par3 for binding to the Par complex, thereby sequestering the active Par complex away from the apical junction domain (Figure 1). Conversely, Lgl phosphorylation by aPKC inactivates the SCRIB complex (reviewed by Assemat et al., 2008; Lee and Vasioukhin, 2008).

Regulation of cell polarity genes during EMT EMT: a dynamic process underlying tumour progression Metastasis of epithelial tumours (carcinomas) is a complex process in which the tumour cells undergo a sequential series of events that initiate their exit from the primary tissue and lead to the formation of a secondary tumour focus in a distant tissue (Gupta and Massague´, 2006). EMT is presently recognized as a key event in tumour progression, as EMT can be exploited by tumour epithelial cells to acquire the ability to dissociate from their neighbours and migrate (Thiery, 2003). Thus, EMT is the first step of the metastatic cascade followed by carcinomas, although EMT can also occur at other stages of the metastatic process, such as at intra- or extravasation (Gupta and Massague´, 2006). As indicated above, the full accomplishment of EMT requires Oncogene

activation of a complex genetic programme that together with the loss of the epithelial character and epithelial polarity implies the acquisition of mesenchymal properties and motility. However, there are several important features of EMT and tumour progression that must be considered. The first aspect to bear in mind is the fact that, as in embryonic development, EMT is probably a transient event during tumour progression and the converse mesenchymal-to-epithelial transition can also occur during the course of metastasis (Thiery, 2003; Peinado et al., 2004; Peinado et al., 2007). Second, as EMT can be driven by many different signalling factors, most probably in a paracrine fashion (Huber et al., 2005; Thiery and Sleeman, 2006), only some tumour cells might be responsive to EMT-inducing cues acting over short distances in the tumour microenvironment (Scheel et al., 2007). Moreover, EMT can probably be considered as the most extreme manifestation of epithelial cell plasticity (Gru¨nert et al., 2003; Huber et al., 2005; Scheel et al., 2007), and the appearance of the full EMT phenotype might be difficult to observe within tumours. Indeed, the relevance of EMT in human tumours is still a matter of considerable debate (Tarin et al., 2005; Thompson et al., 2005; Christiansen and Rajasekaran, 2006; Scheel et al., 2007; Talmadge, 2007), and some of these considerations await clarification and the development of appropriate in vivo model systems, as well as extensive studies on human tumours. As most of the knowledge on EMT has so far been obtained from cell culture studies, we will first review the existing data on the transcriptional regulation of polarity genes in cell culture systems, before examining the data available from human tumours. EMT and the loss of cell polarity: strictly coordinated events Concomitant to the EMT process is the disappearance of the apical–basal polarity of epithelial cells. However, the mechanisms underlying the downregulation of cell polarity determinants and the potential coordination of this with the loss of E-cadherin and other epithelial markers is still poorly understood. Most of the present knowledge regarding the regulation of EMT has come from studies of the mechanisms underlying E-cadherin loss, which has led to the identification of several transcriptional repressors including members of the Snail superfamily, SNAI1 (previously known as Snail) and SNAI2 (previously known as Slug), of the bHLH family (E47/TCF3 and Twist), two ZEB factors, ZEB1 (also known as dEF1/TCF8) and ZEB2 (also known as SIP1/ZFHX1B; reviewed by Peinado et al., 2007) and FOXC2 (Mani et al., 2007). At present, all these E-cadherin repressors are recognized as key inducers of EMT. They not only induce EMT when overexpressed in epithelial cells, but indeed they drive genetic EMT programmes that regulate epithelial and mesenchymal genes, as well as genes involved in extracellular matrix remodelling, cytoskeletal reorganization, cell movements or even cell survival (BarralloGimeno and Nieto, 2005; De Craene et al., 2005b;


6961

Vandewalle et al., 2005; Bermejo-Rodrı´ guez et al., 2006; Moreno-Bueno et al., 2006; Peinado et al., 2007; Escriva` et al., 2008). These different EMT regulators share a similar basic molecular mechanism of repression, binding to conserved E-box sequences (mainly of the CAGGTG type) in the proximal promoter of E-cadherin and other epithelial genes (Peinado et al., 2004, 2007). The mechanisms underlying the upregulation of mesenchymal motility or survival genes are still poorly understood, but they probably involve indirect activation pathways (Vega et al., 2004; Jorda` et al., 2005; Grotegut et al., 2006; Beltran et al., 2008). Apart from E-cadherin, other epithelial genes initially downregulated by Snail and ZEB factors are components of the tight junctions, including occludin and several members of the claudin family (Ikenouchi et al., 2003; Ohkubo and Ozawa, 2004; De Craene et al., 2005b; Vandewalle et al., 2005; Martı´ nez-Estrada et al., 2006). In fact, direct binding of SNAI1/SNAI2 to conserved E-box elements in the corresponding promoters has been reported for occludin, claudin-1 and claudin-7 (Ikenouchi et al., 2003; Martı´ nez-Estrada et al., 2006). Interestingly, downregulation of claudin-4, the junctional adhesion molecule-1 (JAM-1/JAM-A) and Dlg3 has been detected in gene profiling studies of carcinoma cells, as well as in Madin-Darby canine kidney (MDCK) cells undergoing EMT after expression of SNAI1, SNAI2, E47 or other EMT regulators (De Craene et al., 2005b; Moreno-Bueno et al., 2006; Peinado et al., 2008). Because of the close relationship between tight junctions and the organization of epithelial cell polarity, these observations suggest that components of the core cell polarity complexes could also be direct targets of EMT transcriptional regulators. Cell polarity regulators as targets of EMT transcriptional inducers Until recently, very little information existed about the regulation of core polarity components by factors that induce EMT. Several recent reports have now provided evidence that members of the CRB and the SCRIB complexes are direct targets of EMT inducers in different cell systems. In a search for ZEB1 targets, CRB3, PATJ and HUGl2/LGL2 (the human homolog of lethal giant larvae 2), as well as several tight junctions components (JAM-1; occludin, claudin-7), were seen to be upregulated in undifferentiated breast carcinoma cells after ZEB1 silencing (Aigner et al., 2007). Promoter analysis and chromatin immunoprecipitation assays indicated that CRB3 and PATJ genes are direct targets of ZEB1, which bind at specific proximal E-box sequences leading to their repression (Aigner et al., 2007). Similarly, ZEB1 silencing upregulates the expression of several polarity genes in colorectal carcinoma cells, including CRB3 and LGL2, the LGL2 promoter being a direct target of ZEB1 (Aigner et al., 2007). Interestingly, silencing of ZEB1 in breast and colorectal carcinoma cells leads to partial reversion of the epithelial phenotype and, significantly, to the relocalization of CRB3 and/or LGL2 to the membrane, which is associated with a reorganization of apical–basal polarity

(Aigner et al., 2007; Spaderna et al., 2008). A causal relationship between ZEB1-induced EMT and the loss of cell polarity is also supported from these recent studies, as maintenance of cell polarity depends on LGL2 in ZEB1-silenced colorectal carcinoma cells (Spaderna et al., 2008). The biological significance of the downregulation of LGL2 by ZEB1 is further emphasized by the observation that ZEB1 silencing suppresses the metastatic ability of colorectal carcinoma cells (Spaderna et al., 2008). Notably, ZEB1 was detected at invasive regions in breast and colorectal tumours and at the tumour–stroma interface, regions that display a strong reduction of LGL2 protein and no signs of polarity (Aigner et al., 2007; Spaderna et al., 2008). These recent observations are in line with previous studies indicating that EMT occurs at invasive regions of colorectal tumours (Brabletz et al., 2005) where ZEB1 may have an important function (Spaderna et al., 2006). Snail factors have also recently been proposed to downregulate polarity genes. Indeed, downregulation of CRB3 and LGL2 expression by SNAI1/SNAI2 was observed in breast and colorectal carcinoma cells, although less potently than that of ZEB1 (Aigner et al., 2007; Spaderna et al., 2008). Interestingly, SNAI1 seems to preferentially act through distal E-box sequences of the CRB3 promoter in breast carcinoma cells, in contrast to the action of ZEB1 on proximal E-box elements of the CRB3 promoter in this cell system (Aigner et al., 2007), suggesting differential binding specificities for both repressors. However, direct binding of SNAI1 to the endogenous CRB3 promoter was not analysed in that study. SNAI1 also downregulated the CRB complex but not the PAR complex in MDCK cells (Whiteman et al., 2008), in which SNAI1 strongly represses the CRB3 promoter but less so the PALS and PATJ promoters. Indeed, direct binding of SNAI1 to proximal E-boxes of the CRB3 promoter was demonstrated in MDCK-SNAI1 cells (Whiteman et al., 2008), in contrast to previous suggestions from breast carcinomas (Aigner et al., 2007). As mentioned above, other core polarity genes like DLG3 are thought to be downregulated by several EMT inducers, including SNAI1/SNAI2 and bHLH factors (E47/TCF3, E2-2/ TCF4), both in MDCK and mouse carcinoma cells (Moreno-Bueno et al., 2006; Peinado et al., 2008; Sobrado V, Moreno-Bueno G et al., unpublished), although no functional promoter studies are yet available. Taken together, these observations suggest that different EMT transcriptional inducers act cooperatively to repress factors involved in determining polarity, which could reinforce the suppression of apical–basal polarity that is required for cells to undergo EMT. Alternatively, it is likely that the different EMT inducers could operate by silencing cell polarity genes specifically in different tumour types. Studies on the relationship between the expression of EMT transcriptional repressors and polarity genes in large series of tumours (see below) are clearly required to clarify these issues. Regardless of the relationship between the different transcriptional inducers of EMT, these recent reports Oncogene


6962

highlight that important polarity genes of the CRB and SCRIB complexes are transcriptional targets of different EMT inducers. Thus, it appears that there is a coordinated programme of gene regulation underlying EMT. Apart from the direct action of EMT factors on core polarity genes, there are additional possibilities for the transcriptional regulation of cell polarity in association with EMT. The distribution of phosphatidylinositol (4,5) bisphosphate at the apical surface is one of the requirements for the correct apical–basal polarity of epithelial cells (Martin-Belmonte et al., 2007). Remarkably, the PTEN lipid phosphatase that converts phosphatidylinositol (3,4,5) trisphosphate into phosphatidylinositol (4,5) bisphosphate was recently proposed to have an important function in the apical localization of phosphatidylinositol (4,5) bisphosphate lipids (MartinBelmonte et al., 2007), in addition to its well-known action counteracting the prosurvival effect of phosphatidyl inositol-3OH kinase. Therefore, alterations in the expression or localization of PTEN could also affect apical membrane organization. It was recently shown that PTEN is an additional target of SNAI1 and that it is strongly repressed when SNAI1 binds directly to the proximal E-boxes of the PTEN promoter (Escriva` et al., 2008). Thus, SNAI1-mediated repression of PTEN provides an indirect mechanism to affect both the loss of cell polarity and the survival properties of SNAI1expressing cells (Vega et al., 2004). This example of indirect regulation of epithelial cell polarity by SNAI1 makes it plausible that future studies will identify further indirect mechanisms of transcriptional regulation by SNAI1 and other EMT inducers that might affect cell polarity. Regulating the EMT regulators: transcriptional and post-transcriptional mechanisms Perhaps one of the crucial questions in the EMT field relates to the regulation of the different EMT inducers. Mechanisms underlying the modulation of EMT inducers are just beginning to be revealed. A plethora of signalling pathways are known to control Snai1/Snai2 expression both during development and in tumour cells (Thiery, 2003; Barrallo-Gimeno and Nieto, 2005). Importantly, although many of the signalling pathways (TGF-b, fibroblast growth factor, Wnt, Notch) impinge on the transcriptional regulation of the Snai1/Snai2 genes (De Craene et al., 2005a; Huber et al., 2005; Thiery and Sleeman, 2006), post-translational mechanisms affecting protein stability and/or nuclear transport of Snail factors are also emerging as crucial regulators of their activity (reviewed by Peinado et al., 2007). In contrast to the Snail factors, much less is known about the regulation of bHLH and ZEB factors, apart from the involvement of some of the aforementioned signalling pathways (Peinado et al., 2007). Very recently, significant insights have been obtained into the modulation of ZEB factors expression, which implicated noncoding RNAs in new post-transcriptional mechanisms. ZEB1/ZEB2 expression is downregulated in epithelial cells through selective targeting of their mRNAs by Oncogene

several microRNAs (Gregory et al., 2008; Korpal et al., 2008; Park et al., 2008) or other non-coding RNAs (Beltran et al., 2008). MicroRNAs are short non-coding RNAs (20–22 nt long) that control gene expression at the post-transcriptional level by pairing their seed sequences (2–8 nt at the 50 -end) to complementary sequences located in the 30 -UTR region of target mRNAs (He and Hannon, 2004). This coupling leads to the degradation of the target mRNA or inhibition of its translation (Filipowicz et al., 2008). Significantly, evidence for the active participation of specific microRNAs in EMT regulation has now been provided and, in particular, the modulation of ZEB1/ZEB2 mRNA levels by five members of the miR-200 family and by the miR-205 RNA has been reported recently. ZEB1/ZEB2 mRNAs, initially described as targets of two members of the miR-200 family (Christoffersen et al., 2007; Hurteau et al., 2007), contain several binding sites for five miR200 members and miR-2005 at their 30 -UTR regions, which drive mRNA inactivation and prevent the induction of EMT (Gregory et al., 2008; Korpal et al., 2008; Park et al., 2008). In accordance with their proposed function as negative regulators of EMT, the expression of miR-200 and miR-205 RNAs is downregulated in cells that have undergone a full EMT under different stimuli (Gregory et al., 2008; Korpal et al., 2008; Park et al., 2008). Similarly, re-expression of individual miR-200 RNAs in mesenchymal cells induces partial reversion to an epithelial phenotype and reduces the migratory capacity of invasive mouse breast carcinoma cells (Gregory et al., 2008; Korpal et al., 2008). Significantly, expression of the miR-200 family is lost in invasive breast cancer cells and in metaplastic breast carcinomas, in conjunction with E-cadherin downregulation (Gregory et al., 2008). Moreover, there is a tight correlation between the expression of the miR200 RNAs and E-cadherin in a panel of 60 human carcinoma cells of the National Cancer Institute (NCI60), indicating that the miR-200 family may be a strong marker and determinant of the epithelial phenotype in cancer cells (Park et al., 2008). Collectively, these results emphasize the participation of the miR-200 family in the regulation of EMT though the tight control of ZEB1/ZEB2 factors level. Interestingly, regulatory feedback loops between miR-RNAs and EMT factors are also starting to emerge. The expression of two members of the miR200 family is downregulated at the transcriptional level by ZEB1 and Snai1 factors in carcinoma cells (Burk et al., 2008), providing a positive regulatory loop to maintain the expression of ZEB1. On the other hand, other miR-RNAs can act as pro-invasive and prometastatic agents, as exemplified by the miR-10b in breast cancer (Ma et al., 2007). Indeed, miR-10b transcription is upregulated by Twist, another key inducer of EMT, and its expression inhibits the translation of homeobox factor (HOXD10) mRNA, resulting in increased expression of the pro-metastatic gene RHOC (Ma et al., 2007). Although no studies have yet examined the effect of Twist on cell polarity genes, it is likely that they may also be downregulated during the


6963

EMT processes mediated by Twist. There is also evidence that Snai1 and ZEB2 can be upregulated by a distinct post-transcriptional mechanism (Beltran et al., 2008). In this case, SNAI1 mediates the induction of a natural antisense transcript of ZEB2 that prevents splicing of a 50 -UTR intron containing an internal ribosomal entry site, thereby favouring ZEB2 mRNA translation (Beltran et al., 2008). Intriguingly, ZEB1 and/or ZEB2 expression can be frequently found downstream of Snail factors in different model systems of EMT (Guaita et al., 2002; Moreno-Bueno et al., 2006; Sobrado V, Moreno-Bueno G et al., unpublished), supporting the existence of a close relationship between Snail and ZEB factors in the regulation of EMT. These observations, together with the potent repression of cell polarity genes by Snail and ZEB outlined above, support an active function of microRNAs or other non-coding RNAs in the regulation of apical– basal polarity. It will therefore be important to analyse the relationship between the expression of non-coding RNAs, EMT inducers and core polarity genes in carcinoma cell lines and tumours. Post-transcriptional regulation of the PAR complex during EMT The studies described above indicate that members of the CRB and SCRIB complexes can be transcriptionally regulated by known EMT inducers, such as the ZEB and Snail factors. Although still preliminary, the evidence available suggests that components of the PAR polarity complex are not direct targets of EMT transcriptional regulators, but rather they can be modulated in response to different oncogenes, tumour suppressors or EMT regulatory signals (reviewed by Lee and Vasioukhin, 2008). TGF-b is one of the more robust cues inducing EMT in different model systems (reviewed by Zavadil and Bottinger, 2005), leading to the upregulation of several EMT inducers, including SNAI1/SNAI2, ZEB1, Twist and the Id factors through Smad-dependent and Smad-independent pathways (Peinado et al., 2003; Kondo et al., 2004; Thuault et al., 2006; Moustakas and Heldin, 2007). Recent insights into the regulation of cell polarity by TGF-b during EMT have converged on the PAR complex. Par3 protein levels are diminished upon exposure to TGF-b in rat intestinal epithelial cells, concomitant to E-cadherin suppression and the induction of mesenchymal markers, such as a-smooth muscle actin (Wang et al., 2008). The decrease in Par3 mediated by TGF-b treatment results in a redistribution of the Par6–aPKC complex from the membrane to the cytoplasm and the subsequent disorganization of the Par complex, leading to a loss of polarity. By contrast, overexpression of Par3 strongly blocked the EMTmediated effects of TGF-b treatment, including the downregulation of E-cadherin and a-smooth muscle actin expression (Wang et al., 2008). Par6 is also affected by TGF-b, although through a distinct mechanism. In normal mammary cells, TGF-bRI interacts with Par6 in a complex located at the tight junctions. However, after

exposure to TGF-b, TGF-bRII is recruited to the apical junction complex where it phosphorylates Par6. Phosphorylated Par6 recruits Smurf1 (an E3 ubiquitin ligase) to the junctional complex that in turn targets the small GTPase RhoA for degradation, provoking EMT (Ozdamar et al., 2005). It is notable that the mechanisms behind Par6 phosphorylation and Smurf1 recruitment, and that are required for TGF-b induced EMT, are independent of the Smad pathway (Ozdamar et al., 2005). This further supports the participation of different pathways downstream of TGF-b in the regulation of cell polarity and EMT. Besides TGF-b, activation of receptor tyrosine kinases also modulates the epithelial phenotype in development and tumour cells (Huber et al., 2005; Thiery and Sleeman, 2006), suggesting that they may also participate in the regulation of cell polarity complexes. Indeed, activation of ErbB2 was shown to disrupt apical–basal polarity of mammary epithelial cells and of epithelial acini in three-dimensional cultures (Muthuswamy et al., 2001). ErbB2 negatively regulates cell polarity in epithelial cells by associating with Par6aPKC components, thereby dissociating Par3 and generating an inactive PAR complex (Aranda et al., 2006). Interestingly, the action of ErbB2 on cell polarity disruption seems to be independent of cell proliferation control (Aranda et al., 2006). These studies indicate that post-transcriptional regulatory mechanisms of the PAR complex are indeed important modulators of EMT and cell polarity. Collectively, the transcriptional and post-transcriptional regulatory mechanisms outlined above provide further evidence of the strict control of polarity components during EMT. Importantly, some of the regulatory pathways described so far are activated and implicated in human tumours. For example, ErbB2 has been associated with gynaecological tumours as well as other human cancer types (Hirohashi and Kanai, 2003), highlighting the biological relevance of the regulation of cell polarity in cancer. EMT, cell polarity and cancer stem cells The proteins that regulate cell polarity also have an important function in asymmetric cell division (Wodarz and Nathke, 2007), a critical event that guarantees the two essential properties of stem cells: self-renewal and differentiation. Studies in Drosophila neuroblasts (stem cell-like progenitors of the nervous system) have shown that the asymmetric localization of cell-fate determinants is controlled by the Par and Scrib complexes, mutations in their components provoking unrestricted tumour growth (reviewed by Lee and Vasioukhin, 2008). Recently, the existence of cancer stem cells has taken centre stage in cancer biology, as they are increasingly considered as the origin of cancer formation, acquiring the alternative denomination of cancer initiating cells. At present, the origin of cancer stem cells remains unclear, and the debate revolves around whether they derive from normal stem cells or if they arise from progenitor or more differentiated cells that acquire Oncogene


6964

specific mutations that endow them with stem cell properties, such as self-renewal (Reya et al., 2001; Bjerkvig et al., 2005). Recent data have shed new light on the relationship between EMT and stem cells, showing that EMT generates cells with stem cell properties (Mani et al., 2008). It is noteworthy that both non-transformed and transformed mammary cells that undergo EMT following SNAI1 or Twist expression acquire stem cell markers and properties, including self-renewal and the ability to form mammospheres in suspension cultures. Moreover, cells similar to stem cells isolated from mammary glands or mammary carcinomas express EMT markers that are associated with the increased expression of several key inducers of EMT (SNAI1/SNAI2, Twist and/or ZEB2) (Mani et al., 2008). For the first time, these data provide a direct link between EMT and the acquisition of stem cell properties. Although not formally proven, these data suggest a link between EMT and the generation of cancer stem cells that undergo self-renewal, contributing to the generation of secondary tumours at distant sites (Mani et al., 2008). This proposal is in accordance with the existence of cancer stem cells with migratory properties generated by EMT at invasive regions suggested previously (Brabletz et al., 2005) and with the induction by Snail and Twist of cell survival factors and local tumour recurrence (reviewed by Peinado et al., 2007; Cobaleda et al., 2007). Therefore, it is tempting to speculate that the alterations to cell polarity factors induced in tumour cells due to EMT may underlie the acquisition of a stem-like phenotype. Further studies into these phenomena in the near future will certainly clarify these important issues.

Cell polarity genes as tumour markers Early indications that cell polarity genes are involved in tumorigenesis came from observations in D. melanogaster. Mutations in three neoplastic genes, disc large (dlg), scribble (scrib) and lethal giant larvae (lgl), revealed a link between the regulation of cell polarity and cell proliferation. Indeed, because inactivating mutations in these three genes led to neoplastic overgrowth in imaginal discs, they were initially characterized as tumour suppressors in Drosophila (reviewed by Dow and Humbert, 2007; Wodarz and Nathke, 2007; Lee and Vasioukhin, 2008). In mammals, the Dlg, Scrib and Lgl proteins are highly conserved, although different isoforms of these proteins have been found. At least four DLG (DLG1-4) and two LGL (LGL1-2) genes have been described, whereas only one Scrib homologue was identified in humans (Assemat et al., 2008). These genes have been considered as tumour suppressors in humans, as they are downregulated in a variety of tumour types. DLG3 expression is virtually abolished in some oesophagus tumours (Hanada et al., 2000) and it is strongly repressed in gastric carcinoma (Liu et al., 2002). A decrease in DLG1 is a hallmark of gastric carcinomas and the additional decrease of DLG4 expression is Oncogene

correlated with more invasive types of diffuse gastric carcinomas (Boussioutas et al., 2003). Other recent data have shown decreased expression of DLG and/or SCRIB in colorectal tumours, associated with the lack of epithelial cell polarity and a disorganized tissue architecture (Gardiol et al., 2006), as well as in other tumour types (reviewed by Dow and Humbert, 2007). Decreased LGL1 expression has also been reported in several tumours, including colorectal carcinomas (Schimanski et al., 2005), endometrial carcinomas (Tsuruga et al., 2007) or melanomas (Kuphal et al., 2006). In several cases, the downregulation of LGL1 is associated with more advanced stages of the tumour, lymph node metastasis and/or poor survival, supporting its use as a marker of poor prognosis. In addition to expression data on core polarity genes, recent studies on LKB1, a Ser/Thr kinase involved in the regulation of AMPK and cell polarity (Baas et al., 2004), support its activity as a tumour suppressor and its involvement in the control of cell polarity in non-small cell lung cancer (Zhang et al., 2008) and endometrial carcinomas (Contreras et al., 2008). All these observations are evidence that cell polarity components participate in human tumours. However, there is little information regarding the loss of cell polarity during tumour progression and in relation to EMT regulators. Indeed, it is presently unclear whether alterations in the elements involved in maintaining cell polarity in human tumours are secondary consequences or if they are causally linked to tumour progression. An inverse correlation has been established between ZEB1, an inducer of EMT, and LGL2 expression in invasive regions of colorectal and breast carcinomas (Aigner et al., 2007; Spaderna et al., 2008), supporting a causal function for the loss of polarity in tumour progression. However, neither the degree of temporal regulation nor the extent of transcriptional programming of polarity components in human tumours has been studied previously. On the other hand, genetic links between human cancer and DLG, SCRIB and LGL homologues have not yet been reported. In the absence of studies to address these fundamental issues, we have reviewed the analyses of existing gene data sets for human breast carcinomas to define the correlation between the expression of core polarity genes and EMT inducers with the clinical pathological data where available. Networking of EMT inducers and cell polarity determinants in human tumours To understand complex biological processes such as the loss of cell polarity and cancer progression, it is important to consider differential gene expression in the context of complex molecular networks. To analyse the relationship between EMT inducers and the hypothetical function of polarity genes in human tumours, we have analysed gene expression profiles in two different data sets of breast adenocarcinomas (van’t Veer et al., 2002; Ma et al., 2004). We have used these models because they permit the expression of individual


6965 Table 1 Statistical signification between EMT inducer expression and polarity genes in two data sets from breast carcinomas (Ma et al., 2004, up; van’t Veer et al., 2002: bottom) SNAI1 expression (n %) DLG3 expression Negative (n ¼ 32) Positive (n ¼ 33)

15 (60.0) 10 (40.0) P ¼ 0.043 19 (67.9) 9 (32.1) P ¼ 0.019

DLG4 expression Negative (n ¼ 31) Positive (n ¼ 29) LLGL2 expression Negative (n ¼ 32) Positive (n ¼ 28) PARD6 Negative (n ¼ 32) Positive (n ¼ 28)

TCF8 expression (n %)

17 (68.0) 8 (32.0) P ¼ 0.048 18 (72.0) 7 (28.0) P ¼ 0.014 SNAI2 expression (n %)

DLG3 expression Negative (n ¼ 45) Positive (n ¼ 34) PARD6 Negative (n ¼ 43) Positive (n ¼ 36)

30 (75.0) 10 (25.0) P ¼ 0.001 29 (72.5) 11 (27.5) P ¼ 0.001

The microarray data from breast tumour samples in both series were available as background-corrected and normalized log-10 ratios. The expression values of genes related to cell polarity, EMT inducers and some cell adhesion molecules were extracted. We categorized the expression of the different genes in terms of each variable, considering a tumour as positive or negative using the 75th percentile value as the cutoff. The significance of the associations between categorical variables was tested in contingency tables with the Yates correction or Fisher’s exact test.

genes to be analysed in both non-metastatic and metastatic primary tumours in situ. The correlation between EMT repressors and cell polarity genes in both series was analysed using the categorized expression data test. Accordingly, although SNAI1 expression was associated with repression of DLG3, LLGL2 and PARD6 (Ma et al., 2004), the upregulation of TCF8/ZEB1 and DLG4 downregulation was observed (Table 1, up). However, no significant correlation was found between either the SNAI2 or the TCF3 transcription factors and any of the selected polarity genes. By contrast, in another series (van’t Veer et al., 2002), the only significant correlation detected was between SNAI2 expression and the downregulation of DLG3 and PARD6 genes (Table 1, bottom). These results suggest that the regulation of the selected polarity genes by EMT inducers depends on the tumour context. The correlation between the transcription factors and polarity genes indicated was reinforced by the identification of 9, 11 and 8 canonical E-boxes in the promoter

region of the DLG3, LLGL2 and PARD6A genes, respectively. These data suggest that the expression of these genes could be modulated by the EMT-related transcription factors, although further studies are required to clarify this assumption. We also investigated the potential network between EMT inducers, cell adhesion and polarity genes at the level of direct interactions (Figure 2). We included 15 genes involved in these processes: DLG, LGL and PARD family members and SNAI1, SNAI2, TCF3, ZEB1 and ZEB2/TCF8, as well as five genes related to cell adhesion, such as CDH1, CLND1, CLDN7, OCLN and JAM1. These genes were used to produce a protein– protein interaction network for the direct interactions between their products (Figure 2). This analysis was carried out using PathArchitect software incorporating information about these genes from the literature or from the experimental data included in different databases. The protein–protein interaction network obtained is the first such network available to analyse the function of EMT modulators in the context of cell polarity. We found that the network is typical of the complex cell systems, and that whereas the majority of nodes have few links, a few nodes (CDH1, TCF8 or DLG4) have many links, ensuring that the elements are fully connected. Two different regions were identified in this network: the ‘cell polarity’ region and the ‘adhesion-EMT’ region. The former included interactions at the protein binding level between cell polarity and cell adhesion genes, highlighting the important function of the CLND1 and OCLN proteins as they are regulated by the LLGL2 or DLG2 proteins. In addition, this region marks a zone where the apical junction was regulated and other interactions between DGL1, 2 and 4, LLGL2 and PARD6A proteins were identified at the same level. In the ‘adhesion-EMT’ region, TCF8/ZEB1 appears to modulate the adhesion junctions by interacting with CDH1; interestingly, this interaction appears to be either directly or indirectly regulated by the DLG1 gene. Importantly, DLG1 and DLG4 seem to link both regions, and they could be considered as important members of this network that connect cell polarity and the EMT. More interesting results were obtained when the expression profile of the different transcription factors were compared with the clinical-pathological parameters of the data set. In the analysis of lymph node negative (N0) breast adenocarcinoma series (van’t Veer et al., 2002), TCF8 emerges as the most important factor influencing cell polarity. Indeed, the level of TCF8 expression was correlated with metastasis (Table 2) and almost all tumours studied that developed metastasis expressed TCF8 (30/30, Pp0.001). In addition, 26 out of 30 breast tumours that expressed TCF8 displayed a poor tumour differentiation grade (P ¼ 0.004). Moreover, a highly significant correlation was found between TCF8 expression and clinical survival (w2: 6.928, 1 df, P ¼ 0.014, Figure 3). These results indicate that TCF8 expression can be considered as a new marker of poor prognosis in N0 breast adenocarcinoma. Oncogene


6966

Figure 2 Epithelial-to-mesenchymal transition (EMT) and cell polarity protein networks at the direct interaction level. To obtain the protein–protein interaction network, we selected cell polarity genes (DLG, LGL and PARD family members), EMT inducers (SNAI1, SNAI2, TCF3, ZEB1/TCF8, ZEB2/SIP1) and cell adhesion-related (CDH1, CLND1, OCLN, JAM1) genes, and we used the PathArchitect software (Stratagene). (a) ‘Cell polarity region’, (b) ‘EMT-region’ (ellipses with a blue outline indicate the proteins included in the analysis, those without an outline represent proteins closely related according to the databases used). Symbol code indicates different interaction levels. Table 2 Statistical correlation between TCF8 expression and clinicopathological features

1,0

TCF8 expression (n %)

Grade WD (n ¼ 8) MD (n ¼ 20) PD (n ¼ 51)

0.8 TCF8-negative

0 30 (85.7) Pp0.001 0 4 (20) 26 (51.0) P ¼ 0.004

Probability

Metastasis Negative (n ¼ 44) Positive (n ¼ 35)

0.6

0,4 TCF8-positive 0,2

Well (WD), moderately (MD), poorly (PD) differentiated tumours. p=0.014

0,0

Concluding remarks and future directions There has been a significant advance in recent years regarding our understanding of the mechanisms involved in EMT and cell polarity regulation. There appears to be a direct link between EMT inducers and the transcriptional downregulation of several cell polarity genes. Indeed, direct and indirect mechanisms that regulate cell polarity components at the transcriptional and post-transcriptional level occur after induction of EMT. Moreover, functional interacting networks linking inducers of EMT to cell polarity and cell junction molecules have identified potential regulators in human breast carcinomas, proposing a pivotal function of ZEB1 factor. Finally, a link between EMT and stem cell properties has also emerged recently, which will certainly pave the way for future Oncogene

0.00

2,50

5,00

7,50

10,00

12,50

Follow up (years)

Figure 3 Kaplan–Meier analysis of TCF8/ZEB1 expression in lymph node-negative (N0) breast adenocarcinomas (van’t Veer et al., 2002). Positive (white) and negative (black) expression values of TCF8 mRNA were obtained from the average expression ratio. P-values were derived from log-rank tests.

studies to address the many questions that remain unresolved. In particular, we must define the relationship between EMT and cell polarity, and the cancer stem cell phenotype, as well as the establishment of metastasis and tumour recurrence. Nevertheless, these advances indicate that EMT inducers may be suitable targets for new therapeutic interventions to impair tumour progression.


6967 Acknowledgements We thank members of A Cano’s laboratory for their excellent work and encouraging discussions, and Gonzalo Gomez (CNIO, Spain) for helping with the protein network analysis. Our work is supported by grants from the Spanish Ministry of

Education and Science (SAF2007-63051, NAN2004-09230C04 and CONSOLIDER-INGENIO 2010 CSD2007-00017 to AC, and SAF2007-63075 to GMB), the EU (MRTN-CT-2004005428) to AC and from the Fundacioń Mutua Madrilenã (2006) to GMB. GMB is a junior researcher contracted on the Ramon y Cajal program, 2004.

References Aigner K, Dampier B, Descovich LM, Mikula M, Sultan A, Schreiber M et al. (2007). The transcription factor ZEB1 (deltaEF1) promotes tumour cell dedifferentiation by repressing master regulators of epithelial polarity. Oncogene 26: 6979–6988. Anderson JM, Van Itallie CM, Fanning AS. (2004). Setting up a selective barrier at the apical junction complex. Curr Opin Cell Biol 16: 140–145. Aranda V, Haire T, Nolan ME, Calarco JP, Rosenberg AZ, Fawcett JP et al. (2006). Par6-aPKC uncouples ErbB2 induced disruption of polarized epithelial organization from proliferation control. Nat Cell Biol 8: 1235–1245. Aranda V, Nolan ME, Muthuswamy SK. (2008). Par complex in cancer: a regulator of normal cell polarity joins the dark side. Oncogene 27: 6878–6887. Assemat E, Bazellieres E, Pallesi-Pocachard E, Le Bivic A, MasseyHarroche D. (2008). Polarity complex proteins. Biochim Biophys Acta 1778: 614–630. Baas AF, Smit L, Clevers H. (2004). LKB1 tumor suppressor protein: PARtaker in cell polarity. Trends Cell Biol 14: 312–319. Barrallo-Gimeno A, Nieto M. (2005). The Snail genes as inducers of cell movement and survival: implications in development and cancer. Development 132: 3151–3161. Beltran M, Puig I, Penã C, Garcı´ a J, Alvarez A, Penã R et al. (2008). A natural antisense transcript regulates Zeb2/Sip1 gene expression during Snail1-induced epithelial–mesenchymal transition. Genes Dev 22: 756–769. Bermejo-Rodrı´ guez C, Pe´rez-Caro M, Pe´rez-Mancera P, SańchezBeato M, Piris M, Sańchez-Garcı´ a I. (2006). Mouse cDNA microarray analysis uncovers Slug targets in mouse embryonic fibroblasts. Genomics 87: 113–118. Bjerkvig R, Tysnes BB, Aboody KS, Najbauer J, Terzis AJ. (2005). Opinion: the origin of the cancer stem cell: current controversies and new insights. Nat Rev Cancer 5: 899–904. Boussioutas A, Li H, Liu J, Waring P, Lade S, Holloway AJ et al. (2003). Distinctive patterns of gene expression in premalignant gastric mucosa and gastric cancer. Cancer Res 63: 2569–2577. Brabletz T, Jung A, Spaderna S, Hlubek F, Kirchner T. (2005). Opinion: migrating cancer stem cells—an integrated concept of malignant tumour progression. Nat Rev Cancer 5: 744–749. Burk U, Schubert J, Wellner U, Schmalhofer O, Vincan E, Spaderna S et al. (2008). A reciprocal repression between ZEB1 and members of the miR-200 family promotes EMT and invasion in cancer cells. EMBO Rep 9: 582–589. Cobaleda C, Perez-Caro M, Vicente-Duenas C, Sanchez-Garcia I. (2007). Function of the zinc-finger transcription factor SNAI2 in cancer and development. Annu Rev Genet 41: 41–61. Contreras CM, Gurumurthy S, Haynie JM, Shirley LJ, Akbay EA, Wingo SN et al. (2008). Loss of Lkb1 provokes highly invasive endometrial adenocarcinomas. Cancer Res 68: 759–766. Christiansen J, Rajasekaran A. (2006). Reassessing epithelial to mesenchymal transition as a prerequisite for carcinoma invasion and metastasis. Cancer Res 66: 8319–8326. Christoffersen NR, Silahtaroglu A, Orom UA, Kauppinen S, Lund AH. (2007). miR-200b mediates post-transcriptional repression of ZFHX1B. Rna 13: 1172–1178. De Craene B, Gilbert B, Stove C, Bruyneel E, van Roy F, Berx G. (2005b). The transcription factor snail induces tumor cell invasion through modulation of the epithelial cell differentiation program. Cancer Res 65: 6237–6244.

De Craene B, van Roy F, Berx G. (2005a). Unraveling signalling cascades for the Snail family of transcription factors. Cell Signal 17: 535–547. Dow L, Humbert P. (2007). Polarity regulators and the control of epithelial architecture, cell migration, and tumorigenesis. Int Rev Cytol 262: 253–302. Ebnet K. (2008). Organization of multiprotein complexes at cell–cell junctions. Histochem Cell Biol 130: 1–20. Escriva` M, Peiro´ S, Herranz N, Villagrasa P, Dave N, MontserratSentı´ s B et al. (2008). Repression of PTEN phosphatase by Snail1 transcriptional factor during gamma radiation-induced apoptosis. Moll Cell Biol 28: 1528–1540. Etienne-Manneville S. (2008). Polarity proteins in migration and invasion. Oncogene 27: 6970–6980. Filipowicz W, Bhattacharyya SN, Sonenberg N. (2008). Mechanisms of post-transcriptional regulation by microRNAs: are the answers in sight? Nat Rev Genet 9: 102–114. Gardiol D, Zacchi A, Petrera F, Stanta G, Banks L. (2006). Human discs large and scrib are localized at the same regions in colon mucosa and changes in their expression patterns are correlated with loss of tissue architecture during malignant progression. Int J Cancer 119: 1285–1290. Goldstein B, Macara IG. (2007). The PAR proteins: fundamental players in animal cell polarization. Dev Cell 13: 609–622. Gosens I, den Hollander AI, Cremers FP, Roepman R. (2008). Composition and function of the Crumbs protein complex in the mammalian retina. Exp Eye Res 86: 713–726. Gregory PA, Bert AG, Paterson EL, Barry SC, Tsykin A, Farshid G et al. (2008). The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targetin ZEB1 and SIP1. Nat Cell Biol 10: 593–601. Grotegut S, von Schweinitz D, Christofori G, Lehembre F. (2006). Hepatocyte growth factor induces cell scattering through MAPK/Egr-1-mediated upregulation of Snail. EMBO J 25: 3534–3545. Gru¨nert S, Jechlinger M, Beug H. (2003). Diverse cellular and molecular mechanisms contribute to epithelial plasticity and metastasis. Nat Rev Mol Cell Biol 4: 657–665. Guaita S, Puig I, Franci C, Garrido M, Dominguez D, Batlle E et al. (2002). Snail induction of epithelial to mesenchymal transition in tumor cells is accompanied by MUC1 repression and ZEB1 expression. J Biol Chem 277: 39209–39216. Gupta A, Massague´ J. (2006). Cancer metastasis: building a framework. Cell 127: 679–695. Hanada N, Makino K, Koga H, Morisaki T, Kuwahara H, Masuko N et al. (2000). NE-dlg, a mammalian homolog of Drosophila dlg tumor suppressor, induces growth suppression and impairment of cell adhesion: possible involvement of downregulation of beta-catenin by NE-dlg expression. Int J Cancer 86: 480–488. Hartsock A, Nelson WJ. (2008). Adherens and tight junctions: structure, function and connections to the actin cytoskeleton. Biochim Biophys Acta 1778: 660–669. He L, Hannon GJ. (2004). MicroRNA: small RNAs with a big role in gene regulation. Nat Rev Genet 5: 522–532. Hirohashi S, Kanai Y. (2003). Cell adhesion system and human cancer morphogenesis. Cancer Sci 94: 575–581. Huber M, Kraut N, Beug H. (2005). Molecular requirements for epithelial–mesenchymal transition during tumor progression. Curr Opin Cell Biol 17: 548–558. Oncogene


6968 Humbert PO, Grzeschik NA, Brumby AM, Galea R, Elsum I, Richardson HE. (2008). Control of tumourigenesis by the Scribble/ Dlg/Lgl polarity module. Oncogene 27: 6888–6907. Hurteau GJ, Carlson JA, Spivack SD, Brock GJ. (2007). Overexpression of the microRNA hsa-miR-200c leads to reduced expression of transcription factor 8 and increased expression of E-cadherin. Cancer Res 67: 7972–7976. Ikenouchi J, Matsuda M, Furuse M, Tsukita S. (2003). Regulation of tight junctions during the epithelium–mesenchyme transition: direct repression of the gene expression of claudins/occludin by Snail. J Cell Sci 116: 1959–1967. Jeanes A, Gottardi CJ, Yap AS. (2008). Cadherins and cancer: how does cadherin dysfunction promote tumor progression? Oncogene 27: 6920–6929. Jorda` M, Olmeda D, Vinyals A, Valero E, Cubillo E, Llorens A et al. (2005). Upregulation of MMP-9 in MDCK epithelial cell line in response to expression of the Snail transcription factor. J Cell Sci 118: 3371–3385. Kemphues KJ, Priess JR, Morton DG, Cheng NS. (1988). Identification of genes required for cytoplasmic localization in early C. elegans embryos. Cell 52: 311–320. Knust E, Bossinger O. (2002). Composition and formation of intercellular junctions in epithelial cells. Science 298: 1955–1959. Kondo M, Cubillo E, Tobiume K, Shirakihara T, Fukuda N, Suzuki H et al. (2004). A role for Id in the regulation of TGF-beta-induced epithelial–mesenchymal transdifferentiation. Cell Death Differ 11: 1092–1101. Korpal M, Lee E, Hu G, Kang Y. (2008). The miR-200 family inhibits epithelial–mesenchymal transition and cancer cell migration by direct targeting of E-cadherin transcriptional repressors ZEB1 and ZEB2. J Biol Chem 283: 14910–14914. Kuphal S, Wallner S, Schimanski CC, Bataille F, Hofer P, Strand S et al. (2006). Expression of Hugl-1 is strongly reduced in malignant melanoma. Oncogene 25: 103–110. Lee M, Vasioukhin V. (2008). Cell polarity and cancer—cell and tissue polarity as a non-canonical tumor suppressor. J Cell Sci 121: 1141–1150. Liu LX, Liu ZH, Jiang HC, Qu X, Zhang WH, Wu LF et al. (2002). Profiling of differentially expressed genes in human Gastric carcinoma by cDNA expression array. World J Gastroenterol 8: 580–585. Ma L, Teruya-Feldstein J, Weinberg RA. (2007). Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature 449: 682–688. Ma X, Wang Z, Ryan P, Isakoff S. (2004). A two-gene expression ratio predicts clinical outcome in breast cancer patients treated with tamoxifen. Cancer Cell 5: 607–616. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY et al. (2008). The epithelial–mesenchymal transition generates cells with properties of stem cells. Cell 133: 704–715. Mani SA, Yang J, Brooks M, Schwaninger G, Zhou A, Miura N et al. (2007). Mesenchyme forkhead 1 (FOXC2) plays a key role in metastasis and is associated with aggressive basal-like breast cancers. Proc Natl Acad Sci USA 104: 10069–10074. Martin-Belmonte F, Gassama A, Datta A, Yu W, Rescher U, Gerke V et al. (2007). PTEN-mediated apical segregation of phosphoinositides controls epithelial morphogenesis through Cdc42. Cell 128: 383–397. Martı´ nez-Estrada O, Cullere´s A, Soriano F, Peinado H, Bolo´s V, Martı´ nez F et al. (2006). The transcription factors Slug and Snail act as repressors of Claudin-1 expression in epithelial cells. Biochem J 394: 449–457. Moreno-Bueno G, Cubillo E, Sarrio´ D, Peinado H, Rodrı´ guez-Pinilla S, Villa S et al. (2006). Genetic profiling of epithelial cells expressing E-cadherin repressors reveals a distinct role for Snail, Slug, and E47 factors in epithelial-mesenchymal transition. Cancer Res 66: 9543–9556. Moustakas A, Heldin CH. (2007). Signaling networks guiding epithelial–mesenchymal transitions during embryogenesis and cancer progression. Cancer Sci 98: 1512–1520. Oncogene

Muthuswamy SK, Li D, Lelievre S, Bissell MJ, Brugge JS. (2001). ErbB2, but not ErbB1, reinitiates proliferation and induces luminal repopulation in epithelial acini. Nat Cell Biol 3: 785–792. Niessen CM.. (2007). Tight junctions/adherens junctions: basic structure and function. J Invest Dermatol 127: 2525–2532. Nollet F, Kools P, van Roy F. (2000). Phylogenetic analysis of the cadherin superfamily allows identification of six major subfamilies besides several solitary members. J Mol Biol 299: 551–572. Ohkubo T, Ozawa M. (2004). The transcription factor Snail downregulates the tight junction components independently of E-cadherin downregulation. J Cell Sci 117: 1675–1685. Ozdamar B, Bose R, Barrios-Rodiles M, Wang HR, Zhang Y, Wrana JL. (2005). Regulation of the polarity protein Par6 by TGFbeta receptors controls epithelial cell plasticity. Science 307: 1603–1609. Park S, Gaur A, Lengyel E, Peter M. (2008). The miR-200 family determines the epithelial phenotype of cancer cells by targeting the E-cadherin repressors ZEB1 and ZEB2. Genes Dev 22: 894–907. Peinado H, Moreno-Bueno G, Hardisson D, Pe´rez-Go´mez E, Santos V, Mendiola M et al. (2008). Lysyl oxidase-like2 as a new poor prognosis marker of squamous cell carcinomas. Cancer Res 68: 4541–4550. Peinado H, Olmeda D, Cano A. (2007). Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat Rev Cancer 7: 415–428. Peinado H, Portillo F, Cano A. (2004). Transcriptional regulation of cadherins during development and carcinogenesis. Int J Dev Biol 48: 365–375. Peinado H, Quintanilla M, Cano A. (2003). Transforming growth factor beta-1 induces snail transcription factor in epithelial cell lines: mechanisms for epithelial mesenchymal transitions. J Biol Chem 278: 21113–21123. Perez-Moreno M, Fuchs E. (2006). Catenins: keeping cells from getting their signals crossed. Dev Cell 11: 601–612. Perez-Moreno M, Jamora C, Fuchs E. (2003). Sticky business: orchestrating cellular signals at adherens junctions. Cell 112: 535–548. Reya T, Morrison SJ, Clarke MF, Weissman IL. (2001). Stem cells, cancer, and cancer stem cells. Nature 414: 105–111. Richard M, Roepman R, Aartsen WM, van Rossum AG, den Hollander AI, Knust E et al. (2006). Towards understanding CRUMBS function in retinal dystrophies. Hum Mol Genet 15: R235–R243. Scheel C, Onder T, Karnoub A, Weinberg R. (2007). Adaptation versus selection: the origins of metastatic behavior. Cancer Res 67: 11476–11479. Schimanski CC, Schmitz G, Kashyap A, Bosserhoff AK, Bataille F, Schafer SC et al. (2005). Reduced expression of Hugl-1, the human homologue of Drosophila tumour suppressor gene lgl, contributes to progression of colorectal cancer. Oncogene 24: 3100–3109. Spaderna S, Schmalhofer O, Hlubek F, Berx G, Eger A, Merkel S et al. (2006). A transient, EMT-linked loss of basement membranes indicates metastasis and poor survival in colorectal cancer. Gastroenterology 131: 830–840. Spaderna S, Schmalhofer O, Wahlbuhl M, Dimmler A, Bauer K, Sultan A et al. (2008). The transcriptional repressor ZEB1 promotes metastasis and loss of cell polarity in cancer. Cancer Res 68: 537–544. Suzuki A, Ohno S. (2006). The PAR-aPKC system: lessons in polarity. J Cell Sci 119: 979–987. Talmadge J. (2007). Clonal selection of metastasis within the life history of a tumor. Cancer Res 67: 11471–11475. Tarin D, Thompson E, Newgreen D. (2005). The fallacy of epithelial mesenchymal transition in neoplasia. Cancer Res 65: 5996–6000. Thiery J. (2003). Epithelial-mesenchymal transitions in development and pathologies. Curr Opin Cell Biol 15: 740–746. Thiery J, Sleeman J. (2006). Complex networks orchestrate epithelial– mesenchymal transitions. Nat Rev Mol Cell Biol 7: 131–142.


6969 Thompson E, Newgreen D, Tarin D. (2005). Carcinoma invasion and metastasis: a role for epithelial–mesenchymal transition? Cancer Res 65: 5991–5995. Thuault S, Valcourt U, Petersen M, Manfioletti G, Heldin CH, Moustakas A. (2006). Transforming growth factor-beta employs HMGA2 to elicit epithelial–mesenchymal transition. J Cell Biol 174: 175–183. Tsukita S, Furuse M, Itoh M. (1997). Molecular architecture of tight junctions: occludin and ZO-1. Soc Gen Physiol Ser 52: 69–76. Tsuruga T, Nakagawa S, Watanabe M, Takizawa S, Matsumoto Y, Nagasaka K et al. (2007). Loss of Hugl-1 expression associates with lymph node metastasis in endometrial cancer. Oncol Res 16: 431–435. Vandewalle C, Comijn J, De Craene B, Vermassen P, Bruyneel E, Andersen H et al. (2005). SIP1/ZEB2 induces EMT by repressing genes of different epithelial cell–cell junctions. Nucleic Acids Res 33: 6566–6578. van’t Veer LJ, Dai H, van de Vijver M, He Y, Hart A, Mao M et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer. Nature 415: 530–536. Vasioukhin V. (2006). Lethal giant puzzle of Lgl. Dev Neurosci 28: 13–24.

Vega S, Morales A, Ocanã O, Valde´s F, Fabregat I, Nieto M. (2004). Snail blocks the cell cycle and confers resistance to cell death. Genes Dev 18: 1131–1143. Wang X, Nie J, Zhou Q, Liu W, Zhu F, Chen W et al. (2008). Downregulation of Par-3 expression and disruption of Par complex integrity by TGF-beta during the process of epithelial to mesenchymal transition in rat proximal epithelial cells. Biochim Biophys Acta 1782: 51–59. Whiteman E, Liu C, Fearon E, Margolis B. (2008). The transcription factor snail represses Crumbs3 expression and disrupts apico-basal polarity complexes. Oncogene 27: 3875–3879. Wodarz A, Nathke I. (2007). Cell polarity in development and cancer. Nat Cell Biol 9: 1016–1024. Zallen JA. (2007). Planar polarity and tissue morphogenesis. Cell 129: 1051–1063. Zavadil J, Bottinger EP. (2005). TGF-beta and epithelial-to-mesenchymal transitions. Oncogene 24: 5764–5774. Zhang S, Schafer-Hales K, Khuri FR, Zhou W, Vertino PM, Marcus AI. (2008). The tumor suppressor LKB1 regulates lung cancer cell polarity by mediating cdc42 recruitment and activity. Cancer Res 68: 740–748.

Oncogene