Multilayer Control of the EMT Master Regulators

REVIEW Multilayer control of the EMT master regulators

H Zheng and Y Kang

Metastasis is the leading cause of cancer-associated death in most tumor types. Metastatic dissemination of cancer cells from the primary tumor is believed to be initiated by the reactivation of an embryonic development program referred to as epithelial– mesenchymal transition (EMT), whereby epithelial cells lose apicobasal polarity and cell–cell contacts, and gain mesenchymal phenotypes with increased migratory and invasive capabilities. EMT has also been implicated in the regulation of cancer stem cell property, immune suppression and cancer regression. Several transcription factors have been identiﬁed as master regulators of EMT, including the Snail, Zeb and Twist families, and their expression is tightly regulated at different steps of transcription, translation and protein stability control by a variety of cell-intrinsic pathways as well as extracellular cues. Here, we review the recent literature on the signaling pathways and mechanisms that control the expression of these master transcription factors during EMT and cancer progression.

Oncogene (2014) 33, 1755–1763; doi:10.1038/onc.2013.128; published online 22 April 2013 Keywords: cancer metastasis; EMT; Snail; Twist; Zeb; E3-ubiquitin ligase

EMT AND CANCER METASTASIS embryogenesis, EMT has a key role in the formation of various Ninety percent of cancer-related deaths are not caused by tissues and organs such as the neural crest, heart, musculoskeletal 7–9 the growth of primary tumors but rather by the spread of and peripheral nervous systems. In adult organisms, only a cancer cells to distant secondary organs.1–3 Despite its devastating certain subset of cells retains the ability to undergo EMT in specific consequence, cancer metastasis is in fact a very inefficient process physiological or pathological conditions, such as wound healing. that requires tumor cells to overcome a series of challenging However, during cancer progression, tumor cells often gain the hurdles as they spread from the primary tumor to secondary ability to reactivate the EMT program and use it for their own 12 organs.4,5 During the expansion of the primary tumor, benefits. Several in vitro cell culture models and mouse models accumulating genetic and epigenetic changes, as well as tumor of cancer progression have shown that reactivation of EMT by microenvironment cues, cause a small portion of tumor cells to tumor cells confers selective advantages, including enhanced disseminate from the tumor mass, invade through the basement motility and invasiveness during tumor progression that is membrane, escape the host immune surveillance, enter systemic believed to be pre-requisite for tumor cell dissemination and 7–9,13 circulation, and spread to secondary organs. Disseminated tumor metastasis. cells then extravasate through the capillary endothelium into surrounding tissues, find a way to survive in the new environment and eventually grow to become clinically detectable DYNAMICS OF EMT AND MESENCHYMAL-TO-EPITHELIAL macrometastases. TRANSITION IN METASTASIS As an early step in cancer metastasis, tumor cells need to gain In tumor biology, EMT has been extensively studied using in vitro their ability to disseminate from solid tumor mass and invade into cell line models, but direct observation of EMT in vivo, particularly the surrounding stromal tissues either as a group of cells (cohesive in clinical specimens, has been difficult and there has been migration) or as single cells (mesenchymal invasion or amoeboid skepticism regarding its pathological relevance.14,15 Analysis of invasion).6 As epithelial tumor cells are usually tightly associated clinical patient samples have demonstrated features of EMT with their neighboring cells via E-cadherin-containing adherens among cancer cells in the invasive front of a number of different junctions, tumor cells must break these intercellular junctions tumor types,16–19 but success in identifying EMT-transitioning cells before they can move out as single cells and invade stromal in vivo has been hampered by the spatial and temporal tissues. Epithelial tumor cells may undergo a process termed as heterogeneity of EMT in human cancer and lack of reliable EMT epithelial-to-mesenchymal transition (EMT) to facilitate this type of markers that can distinguish tumor cells having undergone EMT cellular invasion (Figure 1), which can be stimulated either by from surrounding stromal cells.20 One explanation for this is that extracellular cytokines, such as TGF-b, EGF, FGF, or by intracellular cancer cells only undergo a transient EMT, reverting back to the cues, such as oncogenic Ras or NFkB signaling.7–11 During EMT, epithelial state by mesenchymal-to-epithelial transition (MET) epithelial cells lose their epithelial polarity and cell–cell contacts once they have penetrated stromal tissue, which facilitates and gain mesenchymal phenotypes with increased migratory and colonization in the target organs.21 Indeed, recent experimental invasive capabilities. A hallmark of EMT is the functional loss of evidence showed that although EMT can increases tumor invasion E-cadherin, while additional cellular changes, such as reduced and cancer stem cell properties,22 tumor cells undergoing EMT are expression of epithelial markers cytokeratins and ZO-1, and the often growth arrested since many EMT-inducing transcriptional overexpression of mesenchymal markers N-cadherin, Vimentin factors can directly inhibit proliferation.16,23–27 Thus, tumor cells and Fibronectin, are also often observed. During metazoan must undergo MET to revert to the epithelial state to allow

Department of Molecular Biology, Princeton University, Princeton, NJ, USA. Correspondence: Professor Y Kang, Department of Molecular Biology, Princeton University, LTL255, Washington Road, Princeton, NJ 08544, USA. E-mail: [email protected] Received 9 February 2013; revised 27 February 2013; accepted 27 February 2013; published online 22 April 2013 Multilayer control of the EMT master regulators H Zheng and Y Kang 1756 metastatic growth in distant organs. The mechanisms underlying cancer—perhaps the seeds of metastasis have disseminated years the induction of MET are largely unknown but it is suspected that before the clinical detection of primary pancreatic cancer. withdrawal of EMT inducers, resulting from changes in the stromal In aggregate, recent advances in clinical and experimental studies cues provided by the surrounding environment, is a potential have firmly established the pathological relevance of EMTs in cause of MET. In support of the sequential EMT/MET model, cancer metastasis and underscore the importance of elucidating dynamic expression of E-cadherin has been documented in cancer molecular mechanism that controls such transitions. progression.28 Whereas those cells that undergo hypermethylation and repression of CDH1 in primary breast cancers are more invasive and metastatic, subsequent TRANSCRIPTION FACTORS IN EMT demethylation and reactivation of CDH1 during the process of As a crucial ‘gate-keeper’ of epithelial characteristics, E-cadherin is metastasis is observed. Consistent with the dynamic expression of mutated or downregulated in many cancer types and is associated CDH1, it has been reported that E-cadherin levels are elevated in with increased invasive behavior of tumor cells.8 For example, in lymph-node metastases relative to matched primary tumor breast cancer, at least 6% of breast cancer patients’ genomes samples, suggesting once again that EMT in primary tumors is 37 29–31 harbor E-cadherin null mutations, which permanently inactivate followed by MET at distant sites. Recently, multiple studies in the function of E-cadherin. Reversible suppression of E-cadherin is experimental mouse models demonstrated that tumor cells need mediated by epigenetic silencing, as well as a network of to downregulate EMT inducers and regain epithelial phenotype transcriptional repressors that target the CDH1 gene promoter.38 before they can form metastatic colonies in distant organs, These EMT transcriptional repressors include Snail (Snail1),39,40 indicating that a complete EMT–MET cascade is important for 41 13 42 43,44 45 46 23,32–35 Snail2 (Slug), Twist1, Zeb1, Zeb2, Goosecoid, FOXC2, tumor metastasis. FoxQ1,47 KLF8,48 Prrx132 and many others. In this review, we will EMT and tumor dissemination have long been believed to be focus on the Snail, Twist and Zeb families of transcription factors initiated at the advanced stage of tumor progression when and provide a focused discussion of the molecular network that changes of tumor microenvironment such as hypoxia and regulate their activities in controlling EMT and cancer metastasis. inflammation induce the invasive behavior of malignant cells. However, this concept has recently been challenged by an elegant animal model in which lineage tracking was used to identify The Snail zinc finger transcription repressor family epithelial cells that have undergone EMT and disseminate to Snail family members are the most intensively studied EMT distant organs.36 In this study, EMT and cell dissemination can transcriptional repressors. A wide ranges of signaling pathways be found at pre-neoplastic stage of pancreatic tissues, and have been found to induce Snail1 and Snail2 expression including pathological conditions such as pancreatitis dramatically TGF-b,49–51 Notch52,53 and Wnt pathways,54–56 reactive oxygen increase EMT and dissemination of otherwise non-malignant species,57 and hypoxia stress.58–60 Among them, TGF-b signaling pancreatic epithelial cells. The surprisingly early occurrence of EMT is one of the most prominent EMT inducing cytokines that in pre-neoplastic pancreas may offer a plausible explanation for activates an array of EMT transcriptional factors besides the rapid development of metastatic diseases in pancreatic Snail1/2.43,51 Hypoxia is believed to be another major Snail and EMT inducing signaling pathway.58,59 Under hypoxia condition, stabilized HIF1-a transcription factor directly upregulates Snail1 gene transcription.58 Notch signaling also has a central role RTK/Ras N in hypoxia-induced Snail1 protein upregulation. When Notch Wnt FB F N signaling is activated under hypoxia condition, Notch intracellular G ot T ch ia domain binds to the Snail1 promoter to induce Snail1 gene x O o th transcription. Notch intracellular domain also facilitates HIF1-a’s p e y r ZEB s binding to LOX gene promoter to upregulate lysyl oxidase (LOX), H Snail 53 Others which in turn stabilizes Snail1 protein. Interestingly, in a liver fibrosis model, stromal cells secrete large quantity of TGF-b under Twist hypoxia, which subsequently feeds back to tumor cells triggering Snail1 overexpression and initiate EMT,61 once again suggesting inter-connection of various EMT inducing signaling pathways. Several transcriptional factors have been reported to directly EMT bind to Snail2 promoter and repress its expression (Figure 2a). In chicken embryo, SOX3 and Snail2 have been identified Pathological impacts to reciprocally bind to each other’s promoter and inhibit the gene expression.11 In a TGFb-induced EMT prostate cancer Migration Immune model, FoxA1 and KLF4 can directly repress Snail2 gene & invasion suppression transcription.11,62,63 Two additional transcriptional repressors have also been identified recently as inhibitors of Snail2 in the Blocked Cancer context of mammary gland development, EMT and breast cancer 64,65 senescence stemness metastasis. In these reports, the short splice variant of the basic helix-loop-helix Per-Arnt-Sim transcription factor Singleminded-2 (SIM2s) and ELF5, an ETS (E twenty-six)-domain transcription factor family member, was shown to directly bind to Figure 1. Master transcriptional factors regulate EMT and cancer the Snail2 promoter to inhibit Snail2 transcription.64,65 metastasis. EMT is thought to be the initial and the most important Interestingly, both SIM2s and ELF5 are essential for mammary step in cancer metastasis cascade. Multiple signaling pathways can gland ductal development or alveologenesis during pregnancy, regulate EMT through modulating Snail, Twist and Zeb family and both are frequently lost during breast cancer transcription factors. Besides the roles in promoting migration and 64,65 invasion, these EMT master regulators also facilitate tumor development. Mammary gland-specific knockout of either metastasis through inducing cancer stem-cell phenotype, blocking SIM2s or ELF5 not only lead to defects in mammary gland oncogene-induced senescence and suppressing the host immune development, but also induces EMT-like changes in mammary surveillance system. epithelial cells. SIM2s and ELF5 knockdown in breast cancer cells

Oncogene (2014) 1755 – 1763 & 2014 Macmillan Publishers Limited Multilayer control of the EMT master regulators H Zheng and Y Kang 1757 induce EMT, and promote tumor invasion and metastasis with Similarly, another stem cell-related transcription factor Oct4 has increased expression of Snail2.64,65 Interestingly, in addition to also been reported to activate Snail1 transcription.67 Together, its well-established role in inducing EMT,8 Snail2 has recently these results suggest that genes important for stem cell function been shown to be a master regulator of mammary gland stem cell may also contribute to EMT and breast cancer progression, in line and breast cancer stem cell activities.66 Therefore, increased Snail2 with the recently discovered link between EMT and stemness.22 level in SIM2s and ELF5-low breast cancer may promote both Research in this field will lead to the discovery of additional genes EMT-dependent invasion and cancer stem cell activities, features and pathways that share critical functions in regulating stem cell that are essential for the metastatic spread of cancer. Consistent activity and cancer metastasis. with these functional and experimental studies, ELF5 and Snail2 At a post-transcriptional level, the expression and activity of expressions are positively and reversely associated with improved Snail family members are tightly regulated by several mechanisms. metastasis-free survival in breast cancer patients, respectively.65 A p53/miR-34 axis is reported to be important for Snail1 gene regulation.68–70 In colon, breast and lung carcinoma cell lines, loss of p53 function, either by mutation or by deletion, induces ELF5 Snail1-mediated EMT in these cells. Since p53 transcriptionally FOXA1 Promoter KLF4 activate miR-34 expression and miR-34 negatively regulates Snail1 SOX3 expression, loss of p53 is coupled to Snail1-driven EMT through Snail2 SIM2s miR-34. Interestingly, Snail1 protein can also bind to the miR-34 promoter region and repress miR-34 transcription, thus forming a double-negative feedback loop (Figure 2b; Table 1).70 Similarly, in non-small cell lung carcinoma, miR-30a was found to directly Snail1 mRNA 3UTR target Snail1 post-transcriptionally and the expression level of miR-34 miR-30a is negatively correlated with invasive phenotype.71 Additionally, let-7d, miR-1, miR-9, miR-200, miR-204 and miR-211 have all been shown to directly inhibit Snail2 and promote an p53 Snail1 epithelial cell type (Table 1).63,72–77 Like other transcription factors, Snail proteins have a very PKD1GSK3 p70S6K LATS2 PAK1 short half-life. Its protein turnover is tightly controlled by the E3 -Trcp ? ligase-mediated proteosomal degradation process. Depending on the model system used, at least two E3 ligases have been 11 97 101 203 246 S S S T S discovered that specifically ubiquitinate and degrade Snail1 78 55 C2H2 zinc finger region protein—FBXL14 (also known as Ppa) and b-Trcp (Figure 2c). Both Snail1 and Snail2 are known to be functionally important for SNAG domain Snail1 protein EMT during neural crest development and migration of neural ? crest cells.7 In the Xenopus neural crest model, FBXL14 was FBXL14 identified to have an important role in controlling Snail2 protein level by targeting it for ubiquitination and degradation.79 It is not Figure 2. Expression of Snail genes is controlled at several different uncommon that one E3 ligase can target multiple important layers. (a) Snail2 gene transcription is repressed by several transcription factors in cancer cells. For example, a well-known transcriptional factors. (b) miR-34 and Snail1 forms a double- F-box protein FBW780 can target multiple oncoproteins, including negative feedback loop, which is under the control of p53 tumor c-myc,81–83 Cyclin E,84 c-jun,85 KLF586,87 and Notch.80,88 It is suppressor. (c) Mechanisms regulate Snail1 protein phosphorylation, activation and degradation. There are at least five kinases can therefore not surprising that later findings suggested that FBXL14 can also target other major EMT master transcription factors for phosphorylate Snail protein on five distinct regions. These 78 modifications can activate Snail transcriptional activity or promote degradation, including Snail1, Twist1 and Zeb2 (SIP1). An Snail protein degradation. Two E3 ubiquitin ligases, b-Trcp and independent study demonstrated that FBXL14 is decreased FBXL14, can ubiquitinate and degrade SNAIL protein degradation. upon hypoxia, and stabilizes Snail1 protein in cancer cells.89

Table 1. MicroRNAs targeting Snail, Twist and Zeb factors

Gene MicroRNA Reference

Snail1 miR-9 Liu et al.73 miR-34 Kim et al.68, Nalls et al.69 and Siemens et al.70 Let-7d Chang et al.74 miR-30a Kumaraswamy et al.71 Snail2 miR-1/200 Liu et al.63 miR-124 Lee et al.75 and Xia et al.76 miR-203 Zhang et al.77 miR-204/211 Wang et al.72 Twist1 miR-214 Li et al.119 miR-580 Nairismagi et al.120 Let-7d Chang et al.74 Zeb1/2 miR-192 Krupa et al.151 and Jenkins et al.152 miR-200 Family Burk et al.140, Gregory et al.141, Korpal et al.142, miR-205 Park et al.143, Brackenet al.144 Zeb2 miR-138 Liu et al.145 miR-203 Saini et al.146 miR-215 White et al.147 miR-708 Saini et al.148

& 2014 Macmillan Publishers Limited Oncogene (2014) 1755 – 1763 Multilayer control of the EMT master regulators H Zheng and Y Kang 1758 E3 ligase b-Trcp has also been demonstrated to ubiquitinate Hypoxia and degrade Snail1 protein in breast cancer cell lines, causing the inhibition of Snail1-induced EMT.55 The Snail1 protein contains at HIF1 least two GSK3b-mediated phosphorylation consensus sequences CD44 MAPK between amino acids 90–110 (Figure 2c). Phosphorylated Serines in this motif facilitate the binding of Snail1 by b-Trcp, which then adds a poly-ubiquitin chain to the Snail1 protein and target it for FBXL14 Twist p12 26S proteosomal degradation. As a downstream effector protein for Wnt signaling, GSK3b activity is tightly regulated. Upon canonical Wnt activation, Axin2 is upregulated at the transcriptional level and acts as the chaperone protein to promote GSK3b translocation from nucleus to cytoplasm, thus blocking GSK3b- CDH1 Bmi1 p16 mediated Snail1 protein degradation.90 Nuclear accumulated Let-7i miR-10b Snail1 protein along with other Wnt signaling downstream targets inhibits E-cadherin expression and induces an invasive breast cancer phenotype. At least four other kinases (Figure 2c) PDGF HOXD1 RAC1

have also been shown to phosphorylate Snail1 at distinct sites for Epithelial Oncogene induced Migration & invasion regulating Snail1 protein transcriptional activity, nuclear localiza- features senescence 91 tion and protein stability. Lats2 phosphorylates Snail1 protein at Figure 3. Twist’s function and regulation in cancer progression. Threonine-203, retains Snail1 in the nucleus and stabilizes the Besides its role in inducing EMT, Twist has also been reported to Snail1 protein. PAK1 phosphorylates Snail1 at Serine-246 and have important roles in inhibiting oncogene-induced premature promotes its transcriptional repressor function as well as its’ senescence and promoting tumor cell migration/invasion through retention in the nucleus.92 p70 S6 kinase induces EMT by both inhibiting p16, and microRNAs let-7i and miR-10b. An array of activating Snail1 mRNA transcription and Snail1 protein turnover; transcriptional factors have been identified to regulate Twist gene however, the exact mechanism for this process is yet to transcription, but only one E3 ligase, FBXL14 has been suggested to be understood.93 Furthermore, PKD1 kinase was found to ubiquitinate and degrade Twist protein. phosphorylate Serine-11 on the SNAG transcription repressor þ À domain on the N-terminus of the Snail1 protein.94–96 Upon immune surveillance by increasing CD4 CD25 Treg immune phosphorylation of this site, Snail1 protein is translocated from suppressing cells and impaired dendritic cell activity. It is believed nucleus to cytoplasm and blocks EMT progression. Since most that this immuno-editing function is achieved through the effects of Snail1 on Treg-inducible cytokines such as TGF-b and transcription factors couple their transcriptional activities with 99 their ubiquitination/degradation processes, their transcriptional thrombospondin-1. In Her2/Neu-induced mammary tumors, Snail1 has been demonstrated to be important for breast tumor activation and repressor domains are often also the interacting 100 domains with E3 uiquitin ligases. It is likely there is another E3 relapse as most of the recurrent tumors have upregulated ligase recognizes the phosphorylated SNAG transcription Snail1 gene expression. Interestingly, these recurrent tumors also repressor domain and target Snail1 protein for degradation. presented a strong EMT phenotype. A further analysis of human Less is known about how Snail2 protein is regulated at the post- breast cancer patients revealed that high levels of Snail1 expression are significantly correlated with poorer relapse-free transcriptional level. One report suggests that Snail2 protein 100 stability is regulated by the Wnt/GSK3b/b-Trcp signaling pathway. survival. Without Wnt signaling, Snail2 protein undergoes GSK3-dependent phosphorylation and subsequent b-Trcp-dependent ubiquitination.97 The E3 ligase FBXL14 can also ubiquitinate and degrade The Twist family of bHLH transcription factors Snail2 protein,78,79 although to what extent this is relevant in Twist1 and Twist2 genes belong to a basic helix-loop-helix cancer progression is not clear. Since F-box E3 ligases usually transcriptional factor family. Both Twist1 and Twist2 are initially require a substrate phosphorylation step, it would be interesting characterized in embryonic morphogenesis.101–103 Twist1 was to know whether GSK3b-mediated phosphorylation is also later found to be highly expressed in metastatic mammary required for interaction between FBXL14 and Snail2. tumor cell lines compared with its lowly metastatic isogenic It is worth noting that besides their roles in EMT, Snail1 and counterparts.13 Further functional analysis revealed that it is Snail2 also function through regulating other aspects of tumor a master regulator for cancer metastasis by inducing EMT progression. In Neu- or Ras-oncogene transformed HMLE cells, and increasing tumor cell migration and invasion as well as TGF-b induction, E-cadherin knock-down as well as Snail1 or Snail2 promoting invasion and inhibiting oncogene-induced senescence overexpression can induce typical EMT.22 Cells that have (Figure 3).13 Mechanistically, Twist1 can bind to the Snail2 undergone EMT acquire breast cancer stem cell characteristics promoter region and induce Snail2 gene transcription to (CD44high/CD24low) with significantly higher in vitro colony promote EMT in mammary epithelial tumor cell line.104 Also, formation ability and in vivo tumor initiation ability.22 These cells through directly binding to PDGFa promoter region and promotes also acquire drug resistance to traditional chemotherapy agents PDGFa and its downstream gene expression, Twist1 promotes the including paclitaxel,98 a common chemotherapeutic drug for formation of invadopodia, which are specialized membrane breast cancer treatment. Interestingly, Snail2, together with protrusions important for extracellular matrix degradation, tumor another transcription factor SOX9, has also been shown to be cell invasion and metastasis.105 Interestingly, another Twist family important for normal mammary stem cell renewal and cancer member, Twist2 is found to be crucial for overcoming oncogene- stem cell activities.66 These discoveries suggest that Snail family induced premature senescence.106 In the MMTV-Neu transgenic transcription factors are functionally important for normal and mouse model, mice develop spontaneous mammary gland tumors cancer stem cells. Besides its link to cancer stem cell, the Snail1 and eventually form lung metastasis. Twist2 was found to be gene has also been shown to have a role in immune evasion of significantly upregulated in lung metastasis samples. When Twist2 tumors. In B16-F10 melanoma cells, Snail1 overexpression induces and/or Twist1 were depleted from primary tumor cell lines, these EMT and promotes tumor metastasis in vivo. This effect is not tumor cells underwent cellular senescence, possibly due to the simply due to the migration and invasion effects conferred by activation of p53 and Rb signaling pathways and ultimately EMT, as tumor cells that had undergone EMT also suppressed host p16INK4a activation.106 This observation was then extended to

Oncogene (2014) 1755 – 1763 & 2014 Macmillan Publishers Limited Multilayer control of the EMT master regulators H Zheng and Y Kang 1759 human cancer cell lines, including melanoma and breast cancer progression have gained increasingly recognitions in recent years. cell lines, suggesting a general role of the Twist family in inhibiting Expression of Zeb proteins in epithelial cells strongly inhibits CDH1 oncogene-induced senescence. Twist2 is also responsible for gene expression and this mechanism has been implicated in EMT, TGF-b and p12 induced EMT through downregulating E-cadherin tumorigenesis and metastasis.42,126–131 Zeb1 and Zeb2 can expression in colon cancer cells.107 interact with DNA binding sites composed of bipartite E-box Bmi1, a polycomb-group repressor complex protein, which is regions (CACCT and CACCTG).132 The CDH1 gene promoter usually upregulated in stem cells and in cancer, is found to be a contains such E-box sequences and it has been shown that direct downstream gene of Twist1.108 Overexpression of Twist1 ectopic expression of Zeb factors in mammary gland epithelial induces Bmi1 gene transcription. Bmi1 in turn works in concert cells42 or MDCK cells43 is sufficient to induce dissociation of with Twist1 as a complex to repress E-cadherin and p16INK4a, adherens junctions42,43 and enhance cancer cell migration and providing another layer of mechanism for Twist1’s function in invasion.43 The Zeb factors can also regulate the expression of EMT, tumor initiation and senescence (Figure 3).108 Recent studies various EMT- and tumor-related genes. For example, they have also revealed an miRNA-dependent mechanism for the metastasis- been shown to repress the expression of genes encoding proteins promoting function of Twist1 (Figure 3).109,110 Twist1 induces the critical to maintain the epithelial phenotype such as E-cadherin, expression of mir-10b, which in turn inhibits HOXD1 expression plakophilin 2 and ZO3.44 Conversely, the Zeb factors can also and controls a set of HOXD1 downstream genes that are involved activate the expression of genes promoting tumor cell invasion in cell migration, invasion and metastasis.109 One of the prominent and metastasis, such as the pro-invasion MMP family members,133 genes is RHOC, a well-known gene involved in cell cytoskeleton especially MMP2.38,134 Surprisingly, Zeb2 mRNA has tumor remodeling and migration, as well as cancer metastasis.111 A later suppressive functions in melanoma by acting as a competitive study revealed yet another miRNA-mediated mechanism for Twist- endogenous RNA, a decoy mRNA for multiple miRNAs targeting induced migration-invasion cascade. In cooperation with Bmi1, tumor suppressor PTEN.135 Twist1 suppresses the expression of let-7i miRNA and ultimately The best-known signaling pathways that regulate Zeb factors at but indirectly leads to RAC1 activation, which promotes cell the transcriptional level are the TGF-b,43 HIF-1a,136 and the Rb migration and invasion.110 tumor suppressor signaling pathways.137 More recently, GRHL2 At the transcriptional level, Twist is regulated by hypoxia- was found to inhibit EMT partially through direct repression of signaling pathway. It is well known that the intratumoral region is Zeb1 gene promoter activity and through upregulating the hypoxic, and the transcriptional factor HIF1-a protein is stabilized miR-200b/c expression.138 Interestingly, GRHL2 expression was at the hypoxia condition.112 Stabilized HIF1-a protein induces EMT often found to be downregulated in claudin-low subclass of breast and promotes lung metastasis in MCF7 and FaDu human cancer cancer, which is known to display EMT and cancer stem cell cell lines through Twist1 upregulation.113 In contrast to Snail1 features. However, in most other cases, the exact transcriptional protein stabilization in hypoxia,53,89 HIF-1 activates Twist1 gene regulators for Zeb1 and Zeb2 are largely unknown. A different expression by directly binding to the Twist1 proximal promoter mechanism of regulating Zeb factor function is through region through a hypoxia-response element.113 Supporting these modulation of their interaction with transcriptional co-repressors. in vitro findings of HIF-1a/Twist1 signaling cascades, Twist In order for Zeb2 to efficiently repress downstream gene knockout and HIF-1a mutation in a mouse models displayed transcription, Zeb2 needs to bind to its co-repressor, C-terminal very similar phenotype.102,114,115 Interestingly, a cell surface binding protein (CtBP). Interestingly, the binding motif responsible marker CD44, which is linked to breast cancer stem cells,116,117 for CtBP interaction is often sumoylated, which blocks the is able to translocate from membrane to nucleus upon ligand- interaction between the two partners, causing Zeb2 to become induced activation and bind directly to the Twist1 gene promoter a very inefficient transcription repressor.139 to initiate transcription.118 Perhaps the most well-understood regulatory mechanism for Less is known about Twist gene post-transcriptional regulation. Zeb1 is the reciprocal negative feedback loop between Zeb Several microRNAs have been reported to target Twist1 mRNA, factors and miR-200 family members (Table 1).140–143 The miR-200 including miR-214, miR-580 and let-7d (Table 1). Downregulation family miRNAs consist of five members, which are located at two of these microRNAs is linked to invasive cancer phenotype, different clusters in two different chromosomes. Cluster 1 contains increased metastasis as well as chemoresistance.74,119,120 At post- three miRNAs, miR-200a, miR-200b and miR-429, while cluster 2 translational level, in ovarian epithelial cancer stem cells, Twist1 is contains miR-200c and miR-141. They all share either identical or constantly degraded but will be stabilized after differentiation.121 similar miRNA seeding sequences. As reported by several However, the exact signaling pathways controlling this turnover independent groups using a variety of cell models, miR-200 process are not clear. One of the E3 ligases mentioned earlier, family was found to be able to repress Zeb1 and Zeb2 expression FBXL14 can target a set of transcriptional factors involved in EMT, by targeting the miR-200 target sequences on the Zeb1 and Zeb2 including Twist1. FBXL14 specifically interacts with and mRNAs.140–143 Interesting, Zeb1 and Zeb2 can directly suppress ubiquitinates Twist1 protein at the C-terminal WR domain.78 miR-200 transcription, thus forming a mir-200-Zeb reciprocal Another interesting study suggested that the Twist protein negative feedback loop that controls cancer cell epithelial or is phosphorylated at Serine 68 by MAP kinase, which can mesenchymal phenotypes.140,144 As expected, overexpression of stabilize Twist1 and promotes breast cancer cell invasion.122 miR-200 enforces epithelial traits, inhibits EMT, and reduces tumor Since Ser-68 is located outside of the WR domain, it is unlikely this invasion. Surprisingly, elevated miR-200 can also promote phosphorylation affects FBXL14-mediated Twist ubiquitination metastatic colonization, in part through inhibiting Sec23a- and degradation. dependent secretion of metastasis suppressive proteins IGFBP4 and TINAGL1.34 Besides the miR-200 family, miR-138, miR-203, miR-215 and miR-708 can target Zeb2 gene and suppresses The Zeb family zinc finger transcription factors tumorigenesis, EMT and metastasis (Table 1).145–148 Zeb1 (Zfhx1a/Tcf8/Ef1/Nil-2) and Zeb2 (Zfhx1b/Sip1), members of the Zeb family, were initially found to be mutated in developmental diseases including Hirschsprung’s disease123 and CONCLUDING REMARKS posterior polymorphous corneal dystrophy.124 Pathological As the crucial initial step in cancer metastasis, the functional and examination and a mechanism study revealed that Zeb2 is clinical importance of EMT has been extended from tumor critical for neural crest epithelial sheet formation, a developmental invasion and migration to other aspects of tumor malignancy, process involving EMT.125 The functions of Zeb factors in cancer including tumor stemness, immune evasion, resistance to

& 2014 Macmillan Publishers Limited Oncogene (2014) 1755 – 1763 Multilayer control of the EMT master regulators H Zheng and Y Kang 1760 senescence, chemoresistance and recurrence. It is therefore not 13 Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C et al. Twist, a surprising that the study of EMT has become a focal point in master regulator of morphogenesis, plays an essential role in tumor metastasis. cancer research in recent years. The Snail, Twist and Zeb families Cell 2004; 117: 927–939. of transcription factors have all been identified as master 14 Tarin D, Thompson EW, Newgreen DF. The fallacy of epithelial mesenchymal regulators promoting EMT and their pathological and clinical transition in neoplasia. Cancer Res 2005; 65: 5996–6000, discussion-1. relevance have been established by extensive animal model and 15 Thompson EW, Newgreen DF, Tarin D. Carcinoma invasion and metastasis: a role clinical association studies. Despite these advances, the inherent for epithelial-mesenchymal transition? Cancer Res 2005; 65: 5991–5995 difficulties in therapeutically targeting transcription factors pre- discussion 5. 16 Brabletz T, Jung A, Reu S, Porzner M, Hlubek F, Kunz-Schughart LA et al. Variable vent the development of pharmaceutical agents that can beta-catenin expression in colorectal cancers indicates tumor progression driven effectively block EMT by inhibiting the function of EMT transcrip- by the tumor environment. Proc Natl Acad Sci USA 2001; 98: 10356–10361. tion factors. Therefore, a thorough mapping of the regulatory 17 Franci C, Takkunen M, Dave N, Alameda F, Gomez S, Rodriguez R et al. Expression mechanisms controlling the expression and activity of these of Snail protein in tumor-stroma interface. Oncogene 2006; 25: 5134–5144. transcription factors is crucial for identifying ‘druggable’ ther- 18 Huang D, Du X. Crosstalk between tumor cells and microenvironment via Wnt apeutic targets in the EMT control network. For example, kinases, pathway in colorectal cancer dissemination. World J Gastroenterol 2008; 14: phosphatases, E3 ubiquitin ligases and deubiquitinases that 1823–1827. control the protein stability and turnover of EMT transcription 19 Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nat Rev factor might become appealing targets to inhibit EMT and the 2002; 2: 442–454. 20 Voulgari A, Pintzas A. Epithelial-mesenchymal transition in cancer metastasis: associated malignant features of invasive cancer. mechanisms, markers and strategies to overcome drug resistance in the clinic. Recent studies also highlighted the importance of the reverse Biochim Biophys Acta 2009; 1796: 75–90. process of EMT, MET, for tumor cells to colonize the distant 21 Brabletz T. To differentiate or not--routes towards metastasis. Nat Rev 2012; 12: 149,150 organs. The fact that most EMT-inducing changes are 425–436. transient in nature and do not involve permanent genetic 22 Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY et al. The epithelial- alterations in EMT regulators also support the notion that it is mesenchymal transition generates cells with properties of stem cells. Cell 2008; the overall epithelial–mesenchymal plasticity, rather than EMT or 133: 704–715. MET alone, that is important for tumor cells to overcome different 23 Tsai JH, Donaher JL, Murphy DA, Chau S, Yang J. Spatiotemporal regulation of hurdles in the metastatic cascade. In this regard, inhibiting either epithelial-mesenchymal transition is essential for squamous cell carcinoma metastasis. Cancer Cell 2012; 22: 725–736. EMT or MET alone may have unwanted adverse side effects, 24 Barr S, Thomson S, Buck E, Russo S, Petti F, Sujka-Kwok I et al. Bypassing cellular and the therapeutic window for applying EMT/MET targeting EGF receptor dependence through epithelial-to-mesenchymal-like transitions. treatments needs to be carefully evaluated with advanced animal Clin Exp Metastasis 2008; 25: 685–693. models of metastasis and careful clinical studies. 25 Vega S, Morales AV, Ocana OH, Valdes F, Fabregat I, Nieto MA. Snail blocks the cell cycle and confers resistance to cell death. Genes Dev 2004; 18: 1131–1143. 26 Mejlvang J, Kriajevska M, Vandewalle C, Chernova T, Sayan AE, Berx G et al. Direct CONFLICT OF INTEREST repression of cyclin D1 by SIP1 attenuates cell cycle progression in cells undergoing an epithelial mesenchymal transition. Mol Biol Cell 2007; 18: The authors declare no conflict of interest. 4615–4624. 27 Jung A, Schrauder M, Oswald U, Knoll C, Sellberg P, Palmqvist R et al. The invasion front of human colorectal adenocarcinomas shows co-localization of ACKNOWLEDGEMENTS nuclear beta-catenin, cyclin D1, and p16INK4A and is a region of low prolifera- We apologize to researchers whose studies we were unable to cite due to the space tion. Am J Pathol 2001; 159: 1613–1617. limitation of this review. We thank W Li and HA Smith for reading the manuscript and 28 Graff JR, Gabrielson E, Fujii H, Baylin SB, Herman JG. Methylation patterns of the providing insightful advises. Research in our laboratory is supported by grants from E-cadherin 5’ CpG island are unstable and reflect the dynamic, heterogeneous the National Institutes of Health R01CA134519 and R01CA141062, Komen for the loss of E-cadherin expression during metastatic progression. J Biol Chem 2000; Cure, the Brewster Foundation and the Champalimaud Foundation. H Zheng is a 275: 2727–2732. recipient of a Komen for the Cure Postdoctoral Fellowship (KG111164). 29 Jeschke U, Mylonas I, Kuhn C, Shabani N, Kunert-Keil C, Schindlbeck C et al. Expression of E-cadherin in human ductal breast cancer carcinoma in situ, invasive carcinomas, their lymph node metastases, their distant metastases, REFERENCES carcinomas with recurrence and in recurrence. Anticancer Res 2007; 27: 1 Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of cancer 1969–1974. cells in metastatic sites. Nat Rev 2002; 2: 563–572. 30 Park D, Karesen R, Axcrona U, Noren T, Sauer T. Expression pattern of adhesion 2 Nguyen DX, Bos PD, Massague J. Metastasis: from dissemination to organ- molecules (E-cadherin, alpha-, beta-, gamma-catenin and claudin-7), their specific colonization. Nat Rev 2009; 9: 274–284. influence on survival in primary breast carcinoma, and their corresponding 3 Valastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving axillary lymph node metastasis. APMIS 2007; 115: 52–65. paradigms. Cell 2011; 147: 275–292. 31 Chao YL, Shepard CR, Wells A. Breast carcinoma cells re-express E-cadherin 4 Fidler IJ. The pathogenesis of cancer metastasis: the ’seed and soil’ hypothesis during mesenchymal to epithelial reverting transition. Mol Cancer 2010; 9: 179. revisited. Nat Rev 2003; 3: 453–458. 32 Ocana OH, Corcoles R, Fabra A, Moreno-Bueno G, Acloque H, Vega S et al. 5 Sethi N, Kang Y. Unravelling the complexity of metastasis—molecular under- Metastatic colonization requires the repression of the epithelial-mesenchymal standing and targeted therapies. Nat Rev 2011; 11: 735–748. transition inducer Prrx1. Cancer Cell 2012; 22: 709–724. 6 Friedl P, Alexander S. Cancer invasion and the microenvironment: plasticity and 33 Celia-Terrassa T, Meca-Cortes O, Mateo F, de Paz AM, Rubio N, Arnal-Estape A reciprocity. Cell 2011; 147: 992–1009. et al. Epithelial-mesenchymal transition can suppress major attributes of human 7 Yang J, Weinberg RA. Epithelial-mesenchymal transition: at the crossroads of epithelial tumor-initiating cells. J Clin Investig 2012; 122: 1849–1868. development and tumor metastasis. Dev Cell 2008; 14: 818–829. 34 Korpal M, Ell BJ, Buffa FM, Ibrahim T, Blanco MA, Celia-Terrassa T et al. Direct 8 Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchymal transitions in targeting of Sec23a by miR-200s influences cancer cell secretome and promotes development and disease. Cell 2009; 139: 871–890. metastatic colonization. Nat Med 2011; 17: 1101–1108. 9 Nieto MA. The ins and outs of the epithelial to mesenchymal transition in health 35 Chaffer CL, Brennan JP, Slavin JL, Blick T, Thompson EW, Williams ED. and disease. Annu Rev Cell Dev Biol 2011; 27: 347–376. Mesenchymal-to-epithelial transition facilitates bladder cancer metastasis: role 10 Kang Y, Massague J. Epithelial-mesenchymal transitions: twist in development of fibroblast growth factor receptor-2. Cancer Res 2006; 66: 11271–11278. and metastasis. Cell 2004; 118: 277–279. 36 Rhim AD, Mirek ET, Aiello NM, Maitra A, Bailey JM, McAllister F et al. EMT and 11 Acloque H, Ocana OH, Matheu A, Rizzoti K, Wise C, Lovell-Badge R et al. dissemination precede pancreatic tumor formation. Cell 2012; 148: 349–361. Reciprocal repression between Sox3 and snail transcription factors defines 37 Polyak K, Metzger Filho O. SnapShot: breast cancer. Cancer Cell 2012; 22: embryonic territories at gastrulation. Dev Cell 2011; 21: 546–558. 562–562.e1. 12 Thompson EW, Williams ED. EMT and MET in carcinoma--clinical observations, 38 Peinado H, Olmeda D, Cano A Snail. Zeb and bHLH factors in tumour progression: regulatory pathways and new models. Clin Exp Metastasis 2008; 25: 591–592. an alliance against the epithelial phenotype? Nat Rev Cancer 2007; 7: 415–428.

Oncogene (2014) 1755 – 1763 & 2014 Macmillan Publishers Limited Multilayer control of the EMT master regulators H Zheng and Y Kang 1761 39 Batlle E, Sancho E, Franci C, Dominguez D, Monfar M, Baulida J et al. The epithelial-mesenchymal transition associated with up-regulation of slug and transcription factor snail is a repressor of E-cadherin gene expression in matrix metalloprotease 2. Mol Cell Biol 2008; 28: 1936–1946. epithelial tumour cells. Nat Cell Biol 2000; 2: 84–89. 65 Chakrabarti R, Hwang J, Andres Blanco M, Wei Y, Lukacisin M, Romano RA et al. 40 Cano A, Perez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ, del Barrio MG et al. Elf5 inhibits the epithelial-mesenchymal transition in mammary gland devel- The transcription factor snail controls epithelial-mesenchymal transitions by opment and breast cancer metastasis by transcriptionally repressing Snail2 repressing E-cadherin expression. Nat Cell Biol 2000; 2: 76–83. Nat Cell Biol 2012; 14: 1212–1222. 41 Hajra KM, Chen DY, Fearon ER. The SLUG zinc-finger protein represses E-cadherin 66 Guo W, Keckesova Z, Donaher JL, Shibue T, Tischler V, Reinhardt F et al. Slug and in breast cancer. Cancer Res 2002; 62: 1613–1618. Sox9 cooperatively determine the mammary stem cell state. Cell 2012; 148: 42 Eger A, Aigner K, Sonderegger S, Dampier B, Oehler S, Schreiber M et al. DeltaEF1 1015–1028. is a transcriptional repressor of E-cadherin and regulates epithelial plasticity in 67 Hu J, Guo H, Li H, Liu Y, Liu J, Chen L et al. MiR-145 regulates epithelial to breast cancer cells. Oncogene 2005; 24: 2375–2385. mesenchymal transition of breast cancer cells by targeting Oct4. PLoS One 2012; 43 Comijn J, Berx G, Vermassen P, Verschueren K, van Grunsven L, Bruyneel E et al. 7: e45965. The two-handed E box binding zinc finger protein SIP1 downregulates 68 Kim NH, Kim HS, Li XY, Lee I, Choi HS, Kang SE et al. A p53/miRNA-34 axis E-cadherin and induces invasion. Mol Cell 2001; 7: 1267–1278. regulates Snail1-dependent cancer cell epithelial-mesenchymal transition. JCell 44 Vandewalle C, Comijn J, De Craene B, Vermassen P, Bruyneel E, Andersen H et al. Biol 2011; 195: 417–433. SIP1/ZEB2 induces EMT by repressing genes of different epithelial cell-cell 69 Nalls D, Tang SN, Rodova M, Srivastava RK, Shankar S. Targeting epigenetic junctions. Nucleic Acids Res 2005; 33: 6566–6578. regulation of miR-34a for treatment of pancreatic cancer by inhibition of 45 Hartwell KA, Muir B, Reinhardt F, Carpenter AE, Sgroi DC, Weinberg RA. The pancreatic cancer stem cells. PLoS One 2011; 6: e24099. Spemann organizer gene, Goosecoid, promotes tumor metastasis. Proc Natl Acad 70 Siemens H, Jackstadt R, Hunten S, Kaller M, Menssen A, Gotz U et al. miR-34 and Sci USA 2006; 103: 18969–18974. SNAIL form a double-negative feedback loop to regulate epithelial-mesenchy- 46 Mani SA, Yang J, Brooks M, Schwaninger G, Zhou A, Miura N et al. Mesenchyme mal transitions. Cell Cycle 2011; 10: 4256–4271. Forkhead 1 (FOXC2) plays a key role in metastasis and is associated with 71 Kumarswamy R, Mudduluru G, Ceppi P, Muppala S, Kozlowski M, Niklinski J et al. aggressive basal-like breast cancers. Proc Natl Acad Sci USA. 2007; 104: MicroRNA-30a inhibits epithelial-to-mesenchymal transition by targeting Snai1 10069–10074. and is downregulated in non-small cell lung cancer. Int J Cancer 2012; 130: 47 Zhang H, Meng F, Liu G, Zhang B, Zhu J, Wu F et al. Forkhead transcription factor 2044–2053. foxq1 promotes epithelial-mesenchymal transition and breast cancer metastasis. 72 Wang FE, Zhang C, Maminishkis A, Dong L, Zhi C, Li R et al. MicroRNA-204/211 Cancer Res 2011; 71: 1292–1301. alters epithelial physiology. FASEB J 2010; 24: 1552–1571. 48 Wang X, Zheng M, Liu G, Xia W, McKeown-Longo PJ, Hung MC et al. Kruppel-like 73 Liu S, Kumar SM, Lu H, Liu A, Yang R, Pushparajan A et al. MicroRNA-9 factor 8 induces epithelial to mesenchymal transition and epithelial cell invasion. up-regulates E-cadherin through inhibition of NF-kappaB1-Snail1 pathway in Cancer Res 2007; 67: 7184–7193. melanoma. J Pathol 2012; 226: 61–72. 49 Peinado H, Quintanilla M, Cano A. Transforming growth factor beta-1 induces 74 Chang CJ, Hsu CC, Chang CH, Tsai LL, Chang YC, Lu SW et al. Let-7d functions as snail transcription factor in epithelial cell lines: mechanisms for epithelial novel regulator of epithelial-mesenchymal transition and chemoresistant prop- mesenchymal transitions. J Biol Chem 2003; 278: 21113–21123. erty in oral cancer. Oncol Rep 2011; 26: 1003–1010. 50 Romano LA, Runyan RB. Slug is an essential target of TGFbeta2 signaling in the 75 Lee MR, Kim JS, Kim KS. miR-124a is important for migratory cell fate transition developing chicken heart. Dev Biol 2000; 223: 91–102. during gastrulation of human embryonic stem cells. Stem Cells 2010; 28: 51 Thuault S, Tan EJ, Peinado H, Cano A, Heldin CH, Moustakas A. HMGA2 and 1550–1559. Smads co-regulate SNAIL1 expression during induction of epithelial-to- 76 Xia H, Cheung WK, Ng SS, Jiang X, Jiang S, Sze J et al. Loss of brain-enriched miR- mesenchymal transition. J Biol Chem 2008; 283: 33437–33446. 124 microRNA enhances stem-like traits and invasiveness of glioma cells. J Biol 52 Timmerman LA, Grego-Bessa J, Raya A, Bertran E, Perez-Pomares JM, Diez J et al. Chem 2012; 287: 9962–9971. Notch promotes epithelial-mesenchymal transition during cardiac development 77 Zhang Z, Zhang B, Li W, Fu L, Zhu Z, Dong JT. Epigenetic Silencing of miR-203 and oncogenic transformation. Genes Dev 2004; 18: 99–115. upregulates SNAI2 and contributes to the invasiveness of malignant breast 53 Sahlgren C, Gustafsson MV, Jin S, Poellinger L, Lendahl U. Notch signaling cancer cells. Genes Cancer 2011; 2: 782–791. mediates hypoxia-induced tumor cell migration and invasion. Proc Natl Acad Sci 78 Lander R, Nordin K, LaBonne C. The F-box protein Ppa is a common regulator of USA 2008; 105: 6392–6397. core EMT factors Twist, Snail, Slug, and Sip1. J Cell Biol 2011; 194: 17–25. 54 Bachelder RE, Yoon SO, Franci C, de Herreros AG, Mercurio AM. Glycogen syn- 79 Vernon AE, LaBonne C. Slug stability is dynamically regulated during neural crest thase kinase-3 is an endogenous inhibitor of Snail transcription: implications for development by the F-box protein Ppa. Development 2006; 133: 3359–3370. the epithelial-mesenchymal transition. J Cell Biol 2005; 168: 29–33. 80 Hubbard EJ, Wu G, Kitajewski J, Greenwald I. sel-10, a negative regulator of lin-12 55 Zhou BP, Deng J, Xia W, Xu J, Li YM, Gunduz M et al. Dual regulation of Snail by activity in Caenorhabditis elegans, encodes a member of the CDC4 family of GSK-3beta-mediated phosphorylation in control of epithelial-mesenchymal proteins. Genes Dev 1997; 11: 3182–3193. transition. Nat Cell Biol 2004; 6: 931–940. 81 Yada M, Hatakeyama S, Kamura T, Nishiyama M, Tsunematsu R, Imaki H et al. 56 Zhou BP, Hung MC. Wnt, hedgehog and snail: sister pathways that control by Phosphorylation-dependent degradation of c-Myc is mediated by the F-box GSK-3beta and beta-Trcp in the regulation of metastasis. Cell Cycle 2005; 4: protein Fbw7. EMBO J 2004; 23: 2116–2125. 772–776. 82 Welcker M, Orian A, Jin J, Grim JE, Harper JW, Eisenman RN et al. The Fbw7 tumor 57 Radisky DC, Levy DD, Littlepage LE, Liu H, Nelson CM, Fata JE et al. Rac1b and suppressor regulates glycogen synthase kinase 3 phosphorylation-dependent reactive oxygen species mediate MMP-3-induced EMT and genomic instability. c-Myc protein degradation. Proc Natl Acad Sci USA 2004; 101: 9085–9090. Nature 2005; 436: 123–127. 83 Welcker M, Orian A, Grim JE, Eisenman RN, Clurman BE. A nucleolar isoform of the 58 Imai T, Horiuchi A, Wang C, Oka K, Ohira S, Nikaido T et al. Hypoxia attenuates Fbw7 ubiquitin ligase regulates c-Myc and cell size. Curr Biol 2004; 14: 1852–1857. the expression of E-cadherin via up-regulation of SNAIL in ovarian carcinoma 84 Koepp DM, Schaefer LK, Ye X, Keyomarsi K, Chu C, Harper JW et al. Phosphor- cells. Am J Pathol 2003; 163: 1437–1447. ylation-dependent ubiquitination of cyclin E by the SCFFbw7 ubiquitin ligase. 59 Lester RD, Jo M, Montel V, Takimoto S, Gonias SL. uPAR induces epithelial- Science 2001; 294: 173–177. mesenchymal transition in hypoxic breast cancer cells. J Cell Biol 2007; 178: 85 Wei W, Jin J, Schlisio S, Harper JW, Kaelin Jr. WG. The v-Jun point mutation allows 425–436. c-Jun to escape GSK3-dependent recognition and destruction by the Fbw7 60 Lopez-Novoa JM, Nieto MA. Inflammation and EMT: an alliance towards organ ubiquitin ligase. Cancer Cell 2005; 8:25–33. fibrosis and cancer progression. EMBO Mol Med 2009; 1: 303–314. 86 Zhao D, Zheng HQ, Zhou Z, Chen C. The Fbw7 tumor suppressor targets KLF5 for 61 Copple BL. Hypoxia stimulates hepatocyte epithelial to mesenchymal transition ubiquitin-mediated degradation and suppresses breast cell proliferation. Cancer by hypoxia-inducible factor and transforming growth factor-beta-dependent Res 2010; 70: 4728–4738. mechanisms. Liver Int 2010; 30: 669–682. 87LiuN,LiH,LiS,ShenM,XiaoN,ChenYet al. The Fbw7/human CDC4 tumor 62 Liu YN, Abou-Kheir W, Yin JJ, Fang L, Hynes P, Casey O et al. Critical and reciprocal suppressor targets proproliferative factor KLF5 for ubiquitination and degradation regulation of KLF4 and SLUG in transforming growth factor beta-initiated pros- through multiple phosphodegron motifs. JBiolChem2010; 285: 18858–18867. tate cancer epithelial-mesenchymal transition. Mol Cell Biol 2012; 32: 941–953. 88 Gupta-Rossi N, Le Bail O, Gonen H, Brou C, Logeat F, Six E et al. Functional 63 Liu YN, Yin JJ, Abou-Kheir W, Hynes PG, Casey OM, Fang L et al. MiR-1 and interaction between SEL-10, an F-box protein, and the nuclear form of activated miR-200 inhibit EMT via Slug-dependent and tumorigenesis via Slug-indepen- Notch1 receptor. J Biol Chem 2001; 276: 34371–34378. dent mechanisms. Oncogene 2012; 32: 296–306. 89 Vinas-Castells R, Beltran M, Valls G, Gomez I, Garcia JM, Montserrat-Sentis B et al. 64 Laffin B, Wellberg E, Kwak HI, Burghardt RC, Metz RP, Gustafson T et al. The hypoxia-controlled FBXL14 ubiquitin ligase targets SNAIL1 for proteasome Loss of singleminded-2s in the mouse mammary gland induces an degradation. J Biol Chem 2010; 285: 3794–3805.

& 2014 Macmillan Publishers Limited Oncogene (2014) 1755 – 1763 Multilayer control of the EMT master regulators H Zheng and Y Kang 1762 90 Yook JI, Li XY, Ota I, Hu C, Kim HS, Kim NH et al. A Wnt-Axin2-GSK3beta cascade 117 Dontu G, Al-Hajj M, Abdallah WM, Clarke MF, Wicha MS. Stem cells in normal regulates Snail1 activity in breast cancer cells. Nat Cell Biol 2006; 8: 1398–1406. breast development and breast cancer. Cell Prolif 2003; 36(Suppl 1): 59–72. 91 Zhang K, Rodriguez-Aznar E, Yabuta N, Owen RJ, Mingot JM, Nojima H et al. Lats2 118 El-Haibi CP, Bell GW, Zhang J, Collmann AY, Wood D, Scherber CM et al. Critical kinase potentiates Snail1 activity by promoting nuclear retention upon phos- role for lysyl oxidase in mesenchymal stem cell-driven breast cancer malignancy. phorylation. EMBO J 2012; 31: 29–43. Proc Natl Acad Sci USA 2012; 109: 17460–17465. 92 Yang Z, Rayala S, Nguyen D, Vadlamudi RK, Chen S, Kumar R. Pak1 phosphor- 119 Li B, Han Q, Zhu Y, Yu Y, Wang J, Jiang X. Down-regulation of miR-214 ylation of snail, a master regulator of epithelial-to-mesenchyme transition, contributes to intrahepatic cholangiocarcinoma metastasis by targeting Twist. modulates snail’s subcellular localization and functions. Cancer Res 2005; 65: FEBS J 2012; 279: 2393–2398. 3179–3184. 120 Nairismagi ML, Vislovukh A, Meng Q, Kratassiouk G, Beldiman C, Petretich M et 93 Pon YL, Zhou HY, Cheung AN, Ngan HY, Wong AS. p70 S6 kinase promotes al. Translational control of TWIST1 expression in MCF-10A cell lines recapitu- epithelial to mesenchymal transition through snail induction in ovarian cancer lating breast cancer progression. Oncogene 2012; 31: 4960–4966. cells. Cancer Res 2008; 68: 6524–6532. 121 Yin G, Alvero AB, Craveiro V, Holmberg JC, Fu HH, Montagna MK et al. Con- 94 Du C, Zhang C, Hassan S, Biswas MH, Balaji KC. Protein kinase D1 suppresses stitutive proteasomal degradation of TWIST-1 in epithelial-ovarian cancer stem epithelial-to-mesenchymal transition through phosphorylation of snail. Cancer cells impacts differentiation and metastatic potential. Oncogene 2013; 32: 39–49. Res 2010; 70: 7810–7819. 122 Hong J, Zhou J, Fu J, He T, Qin J, Wang L et al. Phosphorylation of serine 68 of 95 Bastea LI, Doppler H, Balogun B, Storz P. Protein kinase D1 maintains the epi- Twist1 by MAPKs stabilizes Twist1 protein and promotes breast cancer cell thelial phenotype by inducing a DNA-bound, inactive SNAI1 transcriptional invasiveness. Cancer Res 2011; 71: 3980–3990. repressor complex. PLoS One 2012; 7:e30459. 123 Wakamatsu N, Yamada Y, Yamada K, Ono T, Nomura N, Taniguchi H et al. 96 Eiseler T, Kohler C, Nimmagadda SC, Jamali A, Funk N, Joodi G et al. Protein Mutations in SIP1, encoding Smad interacting protein-1, cause a form of kinase D1 mediates anchorage-dependent and -independent growth of tumor Hirschsprung disease. Nat Genet 2001; 27: 369–370. cells via the zinc finger transcription factor Snail1. J Biol Chem 2012; 287: 32367– 124 Liskova P, Tuft SJ, Gwilliam R, Ebenezer ND, Jirsova K, Prescott Q et al. Novel 32380. mutations in the ZEB1 gene identified in Czech and British patients with 97 Wu ZQ, Li XY, Hu CY, Ford M, Kleer CG, Weiss SJ. Canonical Wnt signaling posterior polymorphous corneal dystrophy. Hum Mutat 2007; 28: 638. regulates Slug activity and links epithelial-mesenchymal transition with epige- 125 Van de Putte T, Maruhashi M, Francis A, Nelles L, Kondoh H, Huylebroeck D et al. netic Breast Cancer 1, Early Onset (BRCA1) repression. Proc Natl Acad Sci USA Mice lacking ZFHX1B, the gene that codes for Smad-interacting protein-1, reveal 2012; 109: 16654–16659. a role for multiple neural crest cell defects in the etiology of Hirschsprung 98 Gupta PB, Onder TT, Jiang G, Tao K, Kuperwasser C, Weinberg RA et al. Identi- disease-mental retardation syndrome. Am J Hum Genet 2003; 72: 465–470. fication of selective inhibitors of cancer stem cells by high-throughput screen- 126 Imamichi Y, Konig A, Gress T, Menke A. Collagen type I-induced Smad-inter- ing. Cell 2009; 138: 645–659. acting protein 1 expression downregulates E-cadherin in pancreatic cancer. 99 Kudo-Saito C, Shirako H, Takeuchi T, Kawakami Y. Cancer metastasis is acceler- Oncogene 2007; 26: 2381–2385. ated through immunosuppression during Snail-induced EMT of cancer cells. 127 Maeda G, Chiba T, Okazaki M, Satoh T, Taya Y, Aoba T et al. Expression of SIP1 in Cancer Cell 2009; 15: 195–206. oral squamous cell carcinomas: implications for E-cadherin expression and 100 Moody SE, Perez D, Pan TC, Sarkisian CJ, Portocarrero CP, Sterner CJ et al. The tumor progression. Int J Oncol 2005; 27: 1535–1541. transcriptional repressor Snail promotes mammary tumor recurrence. Cancer Cell 128 Spaderna S, Schmalhofer O, Hlubek F, Berx G, Eger A, Merkel S et al. A transient, 2005; 8: 197–209. EMT-linked loss of basement membranes indicates metastasis and poor survival 101 Hebrok M, Wertz K, Fuchtbauer EM. M-twist is an inhibitor of muscle differ- in colorectal cancer. Gastroenterology 2006; 131: 830–840. entiation. Dev Biol 1994; 165: 537–544. 129 Spaderna S, Schmalhofer O, Wahlbuhl M, Dimmler A, Bauer K, Sultan A et al. The 102 Chen ZF, Behringer RR. Twist is required in head mesenchyme for cranial neural transcriptional repressor ZEB1 promotes metastasis and loss of cell polarity in tube morphogenesis. Genes Dev 1995; 9: 686–699. cancer. Cancer Res 2008; 68: 537–544. 103 Bialek P, Kern B, Yang X, Schrock M, Sosic D, Hong N et al. A twist code deter- 130 Spoelstra NS, Manning NG, Higashi Y, Darling D, Singh M, Shroyer KR et al. The mines the onset of osteoblast differentiation. Dev Cell 2004; 6: 423–435. transcription factor ZEB1 is aberrantly expressed in aggressive uterine cancers. 104 Casas E, Kim J, Bendesky A, Ohno-Machado L, Wolfe CJ, Yang J. Snail2 is an Cancer Res 2006; 66: 3893–3902. essential mediator of Twist1-induced epithelial mesenchymal transition and 131 Wellner U, Schubert J, Burk UC, Schmalhofer O, Zhu F, Sonntag A et al. The metastasis. Cancer Res 2011; 71: 245–254. EMT-activator ZEB1 promotes tumorigenicity by repressing stemness-inhibiting 105 Eckert MA, Lwin TM, Chang AT, Kim J, Danis E, Ohno-Machado L et al. Twist1- microRNAs. Nat Cell Biol 2009; 11: 1487–1495. induced invadopodia formation promotes tumor metastasis. Cancer Cell 2011; 132 Remacle JE, Kraft H, Lerchner W, Wuytens G, Collart C, Verschueren K et al. New 19: 372–386. mode of DNA binding of multi-zinc finger transcription factors: deltaEF1 family 106 Ansieau S, Bastid J, Doreau A, Morel AP, Bouchet BP, Thomas C et al. Induction of members bind with two hands to two target sites. EMBO J 1999; 18: EMT by twist proteins as a collateral effect of tumor-promoting inactivation of 5073–5084. premature senescence. Cancer Cell 2008; 14: 79–89. 133 Miyoshi A, Kitajima Y, Sumi K, Sato K, Hagiwara A, Koga Y et al. Snail and SIP1 107 Tsuji T, Ibaragi S, Shima K, Hu MG, Katsurano M, Sasaki A et al. Epithelial- increase cancer invasion by upregulating MMP family in hepatocellular carci- mesenchymal transition induced by growth suppressor p12CDK2-AP1 promotes noma cells. Br J Cancer 2004; 90: 1265–1273. tumor cell local invasion but suppresses distant colony growth. Cancer Res 2008; 134 Taki M, Verschueren K, Yokoyama K, Nagayama M, Kamata N. Involvement of 68: 10377–10386. Ets-1 transcription factor in inducing matrix metalloproteinase-2 expression by 108 Yang MH, Hsu DS, Wang HW, Wang HJ, Lan HY, Yang WH et al. Bmi1 is essential in epithelial-mesenchymal transition in human squamous carcinoma cells. Int J Twist1-induced epithelial-mesenchymal transition. Nat Cell Biol 2010; 12: 982–992. Oncol 2006; 28: 487–496. 109 Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated 135 Karreth FA, Tay Y, Perna D, Ala U, Tan SM, Rust AG et al. In vivo identification of by microRNA-10b in breast cancer. Nature 2007; 449: 682–688. tumor- suppressive PTEN ceRNAs in an oncogenic BRAF-induced mouse model 110 Yang WH, Lan HY, Huang CH, Tai SK, Tzeng CH, Kao SY et al. RAC1 activation of melanoma. Cell 2011; 147: 382–395. mediates Twist1-induced cancer cell migration. Nat Cell Biol 2012; 14: 366–374. 136 Krishnamachary B, Zagzag D, Nagasawa H, Rainey K, Okuyama H, Baek JH et al. 111 Clark EA, Golub TR, Lander ES, Hynes RO. Genomic analysis of metastasis reveals Hypoxia-inducible factor-1-dependent repression of E-cadherin in von Hippel- an essential role for RhoC. Nature 2000; 406: 532–535. Lindau tumor suppressor-null renal cell carcinoma mediated by TCF3, ZFHX1A, 112 Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, Cockman ME et al. and ZFHX1B. Cancer Res 2006; 66: 2725–2731. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxy- 137 Arima Y, Hayashi H, Sasaki M, Hosonaga M, Goto TM, Chiyoda T et al. Induction of gen-dependent proteolysis. Nature 1999; 399: 271–275. ZEB proteins by inactivation of RB protein is key determinant of mesenchymal 113 Yang MH, Wu MZ, Chiou SH, Chen PM, Chang SY, Liu CJ et al. Direct regulation of phenotype of breast cancer. J Biol Chem 2012; 287: 7896–7906. TWIST by HIF-1alpha promotes metastasis. Nat Cell Biol 2008; 10: 295–305. 138 Cieply B, Pt Riley, Pifer PM, Widmeyer J, Addison JB, Ivanov AV et al. Suppression 114 Ryan HE, Lo J, Johnson RS. HIF-1 alpha is required for solid tumor formation and of the epithelial-mesenchymal transition by Grainyhead-like-2. Cancer Res 2012; embryonic vascularization. EMBO J 1998; 17: 3005–3156. 72: 2440–2453. 115 Iyer NV, Kotch LE, Agani F, Leung SW, Laughner E, Wenger RH et al. Cellular and 139 Long J, Zuo D, Park M. Pc2-mediated sumoylation of Smad-interacting protein developmental control of O2 homeostasis by hypoxia-inducible factor 1 alpha. 1 attenuates transcriptional repression of E-cadherin. J Biol Chem 2005; 280: Genes Dev 1998; 12: 149–162. 35477–35489. 116 Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective 140 Burk U, Schubert J, Wellner U, Schmalhofer O, Vincan E, Spaderna S et al. identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA 2003; A reciprocal repression between ZEB1 and members of the miR-200 family 100: 3983–3988. promotes EMT and invasion in cancer cells. EMBO Rep 2008; 9: 582–589.

Oncogene (2014) 1755 – 1763 & 2014 Macmillan Publishers Limited Multilayer control of the EMT master regulators H Zheng and Y Kang 1763 141 Gregory PA, Bert AG, Paterson EL, Barry SC, Tsykin A, Farshid G et al. The miR-200 146 Saini S, Majid S, Yamamura S, Tabatabai L, Suh SO, Shahryari V et al. Regulatory family and miR-205 regulate epithelial to mesenchymal transition by targeting role of mir-203 in prostate cancer progression and metastasis. Clin Cancer Res ZEB1 and SIP1. Nat Cell Biol 2008; 10: 593–601. 2011; 17: 5287–5298. 142 Korpal M, Lee ES, Hu G, Kang Y. The miR-200 family inhibits epithelial- 147 White NM, Khella HW, Grigull J, Adzovic S, Youssef YM, Honey RJ et al. miRNA mesenchymal transition and cancer cell migration by direct targeting of proﬁling in metastatic renal cell carcinoma reveals a tumour-suppressor effect E-cadherin transcriptional repressors ZEB1 and ZEB2. J Biol Chem 2008; 283: for miR-215. Br J Cancer 2011; 105: 1741–1749. 14910–14914. 148 Saini S, Yamamura S, Majid S, Shahryari V, Hirata H, Tanaka Y et al. MicroRNA-708 143 Park SM, Gaur AB, Lengyel E, Peter ME. The miR-200 family determines the induces apoptosis and suppresses tumorigenicity in renal cancer cells. Cancer epithelial phenotype of cancer cells by targeting the E-cadherin repressors ZEB1 Res 2011; 71: 6208–6219. and ZEB2. Genes Dev 2008; 22: 894–907. 149 Thompson EW, Haviv I. The social aspects of EMT-MET plasticity. Nat Med 2011; 144 Bracken CP, Gregory PA, Kolesnikoff N, Bert AG, Wang J, Shannon MF et al. 17: 1048–1049. A double-negative feedback loop between ZEB1-SIP1 and the microRNA-200 150 Brabletz T. EMT and MET in metastasis: where are the cancer stem cells? Cancer family regulates epithelial-mesenchymal transition. Cancer Res 2008; 68: Cell 2012; 22: 699–701. 7846–7854. 151 Krupa A, Jenkins R, Luo DD, Lewis A, Phillips A, Fraser D. Loss of MicroRNA-192 145 Liu X, Wang C, Chen Z, Jin Y, Wang Y, Kolokythas A et al. MicroRNA-138 sup- promotes ﬁbrogenesis in diabetic nephropathy. J Am Soc Nephrol 2010; 21: 438–447. presses epithelial-mesenchymal transition in squamous cell carcinoma cell lines. 152 Jenkins RH, Martin J, Phillips AO, Bowen T, Fraser DJ. Pleiotropy of microRNA-192 Biochem J 2011; 440: 23–31. in the kidney. Biochem Soc Trans 2012; 40: 762–767.

& 2014 Macmillan Publishers Limited Oncogene (2014) 1755 – 1763